Hakko 927 E S D Manual Dexterity

Bachelder, J. M. Great expectations 1946 torrent kickass. , Harkins, D. (1995) Do occupational therapists have a primary role in low vision rehabilitation? American Journal of Occupational Therapy, 49, 927 – 930. Google Scholar, Crossref, Medline: Bouchard, D., Tetreault, S. (2000) The motor development of sighted children and children with moderate low vision aged 8–13.

Windows Server 2008 R2 Overview • 1. Windows server 2008 r2 x86 portugues isosorbide 64.

• Fara reclame ascunse, usor de downloadat. De ce sa ne alegi pe noi? Ai nimerit bine, aici gasesti cele mai noi melodii mp3, iti punem la dispozitie cea mai noua muzica din toate genurile muzicale, categoriile de baza fiind: manele, muzica romaneasca, muzica straina, muzica house dar si albume mp3 cat si videoclipuri in format HD. Download bella si edward stay with me zippy. • Site-ul numarul #1 cand vine vorba de Muzica Noua • Simplu de folosit, acum cu un nou design usor de navigat si in varianta mobila.

Gldirect 5.0 2 fully executed contract. Just paste the urls you'll find below and we'll download file for you! If file you want to download is multipart you can use our to check whether multiple download links are still active before you start download.

• Hakko 927 E S D Manual Dexterity Activities
• Hakko 927 E S D Manual Dexterity Pdf
• E&s Academy

Hakko 927 E S D Manual Dexterity Activities

Hakko 927 E S D Manual Dexterity Pdf

E&s Academy

ADVANCES IN
Applied Micro biology VOLUME 77
CONTRIBUTORS TO THIS VOLUME
I. Campbell William A. Clark
P. S. S. Dawson Dorothy H. Geary William E. Gledhill J. M. T. Hamilton-Miller
K. L. Phillips David L. Pruess James P. Scannell Norman Shaw G. G. Stewart Wayne W. Umbreit
E. J. Vandamme J. P. Voets B. J. B. Wood
F. M. Yong
ADVANCES IN
Applied Microbiology Edited by D. PERLMAN School of Pharmacy The University of Wisconsin Madison. Wisconsin
VOLUME 17
@
1974
ACADEMIC PRESS, New York and London A Subsidiary of Harcourt Brace Jovanovich, Publishers
COPYRIGHT 0 1974, BY ACADEMIC PRESS,INC. ALL RIGHTS RESERVED. NO PART OF THIS PUBLICATION MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM OR BY ANY MEANS, ELECTRONIC OR MECHANICAL, INCLUDING PHOTOCOPY, RECORDING, OR ANY INFORMATION STORAGE AND RETRIEVAL SYSTEM, WITHOUT PERMISSION IN WRITING FROM THE PUBLISHER.
ACADEMIC PRESS, INC.
111 Fifth Avenue, New York. New York 10003
United Kingdom Edition published by ACADEMIC PRESS, INC. (LONDON) LTD. 24/28 Oval Road. London NWl
LIBRARY OF
CONGRESS CATALOG CARD NUMBER:
ISBN 0-12-002617-
1
PRINTED IN THE UNlTED STATES OF AMERICA
59-13823
CONTENTS LIST
OF
CONTRIBUTORS .......................
ix
Education and Training in Applied Microbiology
WAYNEW . UMBREIT I. I1. I11. IV.
Introduction ......................... Education and Training in the United States ........... Training in Applied Microbiology in the Developing Countries . . . . . . Summary ..........................
1 1 11 17
Antimetabolites from Microorganisms
DAVIDL. PRUESSAND I. I1. I11. IV . V. VI . VII.
JAMES
P. SCANNELL
Introduction ......................... Definition of an Antimetabolite ................. Detection Methods ...................... Mechanisms of Action and Reversal ............... Antimetabolites Isolated from Microorganisms ........... Incidence of Antimetabolite Production .............. Chemotherapeutic Applications ................. Note Added in Proof ..................... References ..........................
19 19 21 22 24 50 51 53 54
lipid Composition as a Guide to the Classification of Bacteria
NORMANSHAW I. I1. I11. IV . V. VI.
Introduction ......................... 63 The Major Types of Lipids Found in Bacteria ........... 64 Analysis of Bacterial Lipids .................. 71 Distribution of Lipids in Bacteria ................ 74 Correlation of Lipid Composition with Taxonomic Classification ... 98 Conclusions ......................... 103 References .......................... 104 V
vi
CONTENTS
Fungal Sterols and rhe Mode of Action of the Polyene Antibiotics
J . M . T . HAMILTON-MILLER I. I1. I11. IV. V. VI . VII . VIII .
Introduction ......................... Sterols in the Microbial Kingdom ................ Factors Affecting the Sterol Content of Fungal Cells . . . . . . . . . . . . . . . Physiological Role of Sterols .................. Sterols and the Mode of Action of Polyenes ............ Resistance to the Polyene Antibiotics .............. The Role of Sterols in Resistance to the Polyenes . . . . . . . . . . . . . . . . . . Conclusions ......................... References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
109 110 112 114 115 126 128 130 131
Methods of Numerical Taxonomy for Various Genera of Yeasts
I. CAMPBELL I. I1. I11. IV .
Theory of Classification of Yeasts ................ Numerical Taxonomy Applied to Yeasts .............. Comparison of Numerical and Classical Taxonomy . . . . . . . . . . . . . . . . . Application of Computer Techniques to Identification . . . . . . . . . . . . . . . References ..........................
135 137 146 150 154
Microbiology and Biochemistry of Soy S.auce Fermentation
.
F. M . YONG AND B . J . B WOOD I. Introduction ......................... I1. Fermented Soy Products .................... 111. History of Soy Sauce Production ................ IV . Chemical Composition of Soy Sauce ............... V . Raw Materials ........................ VI . Treatment of Raw Materials .................. VII . Koji ............................. VIII. Culturing the Koji ....................... IX . Mash (Moromi) ....................... X . Pressing ........................... XI. Pasteurization ......................... XI1. “Chemical” Soy Sauce .................... XI11. Semichemical Soy Sauce, or Shinshiki Shoyu ........... XIV . Future Developments in the Soy Sauce Industry .......... xv. Conclusions .......................... References ..........................
157 159 161 163 165 169 171 173 175 182 182 183 184 185
188 188
CONTENTS
vii
Contemporary Thoughts on Aspects of Applied Microbiology
P. S . S . DAWSONAND K . L . PHILLIPS I. I1. I11. IV . V.
Introduction ......................... Biological Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nonbiological Considerations .................. Mathematical Considerations .................. General Considerations .................... References ..........................
195 196 213 224 227 230
Some Thoughts on the Microbiological Aspects of Brewing and Other Industries Utilizing Yeast
.
G G . STEWART
. .
I I1 I11. IV . V. VI . VII . VIII IX. X. XI . XI1. XI11. XIV.
.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Historical Survey of the Industrial Uses of Yeast . . . . . . . . . . . . . . . . . . . Present-Day Industrial Uses of Yeasts . Outline of Traditional Brewing Processes ............. New Developments in Batch Processes for Brewing . . . . . . . . . . . . . . . . . Continuous Fermentation in Brewing .............. Comparison of the Available Modern Fermenting Systems in Brewing . . Selection. Behavior. and Efficiency of Yeasts . . . . . . Baker’s Yeast ................. The Fed-Batch Process ..................... Continuous Processes .............. Distiller’s Yeast . . . . . .......... ....... Food and Fodder Yeas .................. Biochemicals froin Yeast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References ............ .............
233 235 238 239 241 244 247 248 250 252 253 254 256 260 262
linear Alkylbenzene Sulfonate: Biodegradation and Aquatic Interactions
WILLIAME . GLEDHILL I. I1. I11. IV .
Introduction ......................... LAS Biodegradation Studies .................. LAS and the Aquatic Environment ............... Summary and Conclusions ................... References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
265 266 281 286 289
The Story of the American Type Culture CollectionIts History and Development ( 1 899-1973)
WILLIAMA . CLARKAND DOROTHY H . GEARY I. Founding and Early Years ................... I1. Expansion ..........................
295 299
viii
CONTENTS
I11. Permanent Facilities ...................... References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
303 309
Microbial Penicillin Acylases
E . J . VANDAMME AND J . P . VOETS I. I1. I11. IV . V. VI . VII . VIII . IX. X. XI . XI1.
Introduction ......................... History ........................... Biotransformation of Penicillins into 6-APA ............ Screening Procedures ..................... Penicillin Acylases ....................... Physiological Role of the Penicillin Acylases ............ Coexistence of Acylase and 8-Lactamase ............. Penicillin Acylase and Cephalosporins .............. Nonmicrobial Penicillin Acylases ................. Aspecific Amidases ....................... Chemical Transformation of Penicillins into 6-APA ......... Concluding Remarks ...................... References ..........................
SUBJECTINDEX........................... CONTENTSOF PREVIOUS VOLUMES...................
311 312 316 323 325 347 352 354 356 356 358 359 361
371 377
LIST
OF CONTRIBUTORS
Numbers in parentheses indicate the pages on which the authors' contributions begin.
I. CAMPBELL,Department of Brewing and Biological Sciences, Heriot-Watt University, Edinburgh, Scotland ( 135) WILLIAMA. CLARK,American Type Culture Collection, Rockville, Maryland (295) P. S. S. DAWSON, National Research Council of Canada, Prairie Regional Laboratory, Saskatoon, Saskatchewan, Canada ( 195)
DOROTHY H . ~GEARY, American Type Culture Collection, Rockville, Maryland (295) WILLIAME. GLEDHILL, Monsanto Company, S t . Louis, Missouri (265) J . M . T. HAMILTON-MILLER, Department of Medical Microbiology, University of London, Royal Free Hospital, London, Great Britain (109)
K . L. PHILLIPS,National Research Council of Canada, Prairie Regional Laboratory, Saskatoon, Saskatchewan, Canada ( 195) DAVIDL. PRUESS,Chemical Research Department, Hoffmann-La Roche Inc., NutZey, New Jersey (19) P. SCANNELL, Chemical Research Department, Hoffmann-La Roche Inc., Nutley, New Jersey (19)
JAMES
NORMANSHAW, Microbiological Chemistry Research Laboratory, School of Chemistry, University of Newcastle upon Tyne, Newcastle upon Tyne, England ( 63 ) G. G. STEWART,Beverage Science Department, Labatt Breweries of Canada Ltd., London, Ontario, Canada (233) WAYNEW. UMBREIT,Department of Microbiology, Rutgers University, New Brunswick, New Jersey ( 1 ) E. J . VANDAMME,Laboratory of General and Industrial Microbiology, University of Gent, Gent, Belgium (311) J. P. VOETS, Laboratory of General and Industrial Microbiology, University of Gent, Gent, BeEgium (311)
B. J . B. WOOD,Department of Applied Microbiology, University of Strathclyde, Glasgow, Scotland (157)
F. M . YONG,*Department
of Applied Microbiology, University of Strathclyde, Glasgow, Scotland ( 157)
* Present address: Singapore Institute of Standards and Industrial Research, Republic of Singapore. ix
This Page Intentionally Left Blank
ADVANCES IN
Applied Microbiology VOLUME 77
This Page Intentionally Left Blank
Education and Training in Applied Microbiology
WAYNEW. UMBREIT Department of Microbiology, Rutgers University, New B r u m i c k , New Jersey I. Introduction ................... 11. Education and Training in the United States ..... 111. Training in Applied Microbiology in the Developing Countries .................... IV. Summary ....................
1 1 11 17
I. Introduction The education and training of microbiologists is a complex problem with many facets. This informal essay is a personal opinion, not documented with surveys and statistics, but based upon experience with applied microbiology on both sides of the fence. It is divided into two sections. One is concerned with the needs and problems in the United States. These considerations might apply to other developed countries, but I do not myself have the necessary experience to handle this subject. The other is the problem of training in the developing countries. It is clearly unlikely that people training or employing applied microbiologists in developing countries will ever see this essay, so that it is primarily designed to explain to microbiologists in developed countries some of the problems faced by their colleagues elsewhere. 11.
Education and Training in the United States
A discussion of the training could well begin with a discussion of the question: “What do we expect an applied microbiologist to do?” Clearly, it will be rather difficult to design and conduct a training program without knowing its ultimate objective. However, microbiology is applied in many fields: in public health, in water purification, in soils, in foods, in industrial fermentations, in evaluation of clinically useful drugs, in the “fine” chemical industry, in the pharmaceutical industry, in the cosmetic industry, in waste disposal mechanisms, in eutrophication studies, in the oil refinery, and even in mining and metallurgy. These are clearly not all alike and results in different kinds of work that require markedly different skills. A training program for applied microbiologists ought to leave room for a variety of talents since its practitioners undertake a wide variety of tasks. That is, there are different kinds of applied microbiologists and these may well require different types of training. 1
2
WAYNE W. UMBREIT
There is a rarely appreciated aspect to the training of applied microbiologists, namely, that it is relatively rare for the beginning college student to decide upon a career in applied microbiology at the start of the program. Rather, as knowledge develops, an interest in microbiology may develop, and finally, conditioned by ability, interest, and opportunity, a career in applied microbiology develops. While there has been much discussion about “choosing a career,” it is my own opinion that most people, myself included, “fall into” a career rather than deliberately “choose” it by making use of the opportunities that arise in the course of growing older and more experienced. But this means, in fact, that a career in applied microbiology, as a professional microbiologist, is rarely “chosen” even in college, and that, therefore, a program labeled “applied microbiology,” even at the undergraduate college level, will probably not attract many students. The time available to the undergraduate is usually somewhat limited and, depending upon the students’ objectives, may be very limited indeed. Because so many different kinds of skills are required in the many fields of applied microbiology and because young people do not, as a rule, deliberately choose applied microbiology as a career, one cannot set up (in a practical sense) a training program for applied microbiologists as such. One may design programs with more limited objectives, and one can devise intensive programs for food microbiologists, or water microbiologists, or others. But it is clearly difficult to provide a training program at the undergraduate level, and even at the graduate level, in the general field. Indeed, who is to tell in which subject a position may open when the student has completed the work and needs such a position? Two other aspects of applied microbiology should be emphasized. First, an applied microbiologist always works under, and is normally acutely aware of, economic restraints. No one expects a molecular biologist to pay his way-i.e., that the product of his work should be economically more valuable than its cost. Yet this is continually required of the applied microbiologist. Even in such subsidized situations as water purification or sewage disposal, his work is expected to be of direct application and direct economic benefit to the overall operation. Clearly, of course, there are sometimes indirect applications, such as meeting the requirements of regulatory agencies, but these too have an economic vector. It is this constant awareness of economic constraints that seems to be most difficult for the present generation of students to assimilate. Possibly, having been brought up in an affluent society, in which essentially all the activities, especially in the schools, have been heavily subsidized, they have had minimal contact with the economic facts of life. Possibly an era of scarcity will change this viewpoint, but as it presently
EDUCATION AND TRAINING IN APPLIED MICROBIOLOGY
3
exists, a student expects to be supported and subsidized in work in the future, as in the past, and to “get a grant” to do the research. It rarely, if ever, occurs to him that he might develop a problem to the stage of economic value, and from this provide support. A second constraint of the applied microbiologist is that one must take the situation as one finds it, that one cannot choose the ideal circumstances. One cannot, for example, make citric acid from pure cane sugar (it is not economic, one does not have it, the apparatus is missing, etc.), but rather one must make it from molasses or even corn stalks, because that constitutes the raw material that is available and economically feasible. Furthermore, one must make it with as little modification of available equipment as possible. Microbiologists are trained at three levels. First, there are technicians who do valuable, but probably mostly routine, work. They should, understand what they are doing, but one ought not to expect very much beyond the “cook book approach. There is a tendency to downgrade the importance of the technician in the sciences, even to the extent that no training at all is acceptable or, rather, that “on the job training” is all that is required. For many technician jobs this might be true, provided that the organization has other people who can interpret the results, who can carry out the innovative functions, and who know how to move into a new routine and how to train individuals for the new operation when it is required. Our educational system does not provide adequate numbers of beginning level microbiological technicians. What seems to happen, in the majority of cases, is that the positions are filled by persons with college degrees in fields other than microbiology or by people without any extensive college training. They are trained on the job. Most of the results are not very satisfactory. True, a few technicians become quite good, but the majority stay mediocre, protected usually by their union, and show little interest in further accomplishment, or indeed, further training. Because usually their training has been limited to the job at hand, they are quite unaware of other applications of microbiology and tend to stick closely to a routine. With advancement via seniority, with the accumulation of dependents and mortgages, it is difficult for them to move to any higher intellectual level, and few make the attempt. This is not to say that they may not do valuable work, but normally they have had no training in microbiology per se except what they were given in a busy laboratory or plant. But if they wanted further training in microbiology, where would they get it? Unfortunately, there are few places for such training. In the high schools there is none. In the normal course of events, the first place they can acquire microbiology is the junior year in college and they must have chosen a science background to do that; i.e., they must
4
WAYNE
W. UMBREm
have had general chemistry, organic chemistry, calculus, and general biology. Many applied oriented individuals never go to college or go to majors other than science. Now, I do not think that training for a microbiology technician need be all that difficult, and I am sure that it can be done without years of chemistry or advanced mathematics, but there is no place known to me where such training is, in fact, available. The requirements of the community colleges and of technical or vocational institutes are every bit as complex as those of the colleges. Furthermore, the total number of technicians in microbiology (aside from the hospital-related or diagnostic laboratories) is not great enough so that special schools can be set up for them. But what, in fact, do these people need? I think there are two needs. One is a regular course in general microbiology, possibly without detailed intermediary metabolism, with laboratory training so that they will know some microbiology in addition to that which they see in their own laboratories. I would suggest that the industries of a given area cooperate with a local college or university, to provide such a course when necessary, preferably in a laboratory away from the place of employment, and- that they hire a suitable member of the teaching staff to provide this “special” course, which is not to be restricted by university entrance requirements, etc. The other need is some counseling or contact with the teaching staff of the local institutions so that opportunities available for their further training can be brought to their attention. I would suppose that their union might have an interest in such opportunities for its members, although, possibly I have been away from such contacts too long. The second level of training is the B.A., B.Sc. (or M.A., M.Sc.) level. In theory, at least, it is the function of individuals at the higher level to see that things are properly done and to adjust, modify, and improve operations as required. Here at least, the training is in microbiology and, in my opinion, is generally pretty good. A typical program, to the bachelor’s level, involves two years of chemistry (through organic chemistry), mathematics through calculus, about two years of general biology and genetics, with additional microbiology represented by a course in general and pathogenic microbiology plus one or more advanced courses. These are frequently listed as food microbiology, dairy microbiology, soil microbiology, water microbiology, etc., or they may be combined into one course, called “applied microbiology.” All these courses are generally accompanied by laboratory sessions which rarely have more than 20-25 students per laboratory instructor ( a graduate student). Furthermore, the standards are actually more than adequate, and a bachelor’s degree from most schools in the United States represents a reasonably standardized training of rather good quality. One hears two complaints. One is that individuals do not have ade-
EDUCATION AND TRAINING IN APPLIED MICROBIOLOGY
5
quate laboratory experience, i.e., that they “can’t do a plate count,” “can’t do a chemical determination,” “must be trained in our laboratory,” etc. There is some substance to this type of complaint. It is not that an individual does not know how to do a plate count, but rather that he or she has not had much experience with such counts, maybe has done no more than two or three. Further, these counts were done in an undergraduate teaching laboratory, where facilities and expertise were quite different from those in the industrial laboratory, so he or she does not know how to do it as the employer wants it done. This the employer will have to teach and to allow time to acquire the experience in the hands which the more experienced technician has already acquired. After all, how did the student know who was to be his employer? And how was he supposed to know that the employer wanted plate counts done in a particular way with particular equipment? Most laboratory or microbiological or chemical determinations involve operations which, while essentially the same in principle, differ in the details of the operations; thus one must always be prepared for a “break in” period for each new individual added to a group. The second kind of complaint is that the college product is inadequately trained in the communication skills: “he can’t write a report,” “he can’t give a clear discussion,” “he can’t approach the problem realistically,” etc. It is, of course, true that the present generation seems to communicate mostly by way of monotones at a high decible and it is a mystery to me why we need such high fidelity stereo, 8-channel reproduction of this nonsense. And it is quite true that verbal communication and written work are poorly handled, although folksongs seem to substitute. I am afraid that there is little that microbiologists can do about it, and the protests should be directed to our colleagues in the liberal arts. While I think that the product at the bachelor’s level in the developed countries of the world is quite adequate (given a lack of experience), I do see a serious difficulty in the training of applied microbiologists in our colleges and universities. These so regularly advocate “fundamental” work, research, and “pure” science that the notion that application of science might also be a reasonable alternative to “men good and true,” seems to be mentioned but rarely, if at all. Application is almost submerged in the exciting “new” discoveries of fundamental research. If fundamental science is “pure,” then there must be an “impure” science, presumably applied. Thus it is rather difficult to attract students to “applied microbiology” when they can take “molecular biology” or “fundamentals of the gene.” Further, in most universities the majority of the faculty and a vast majority of the graduate students regard as secondary any application of their work. “If it can be applied, let some-
6
WAYNE W. UMBREIT
one else apply it. After all, it’s not in my research grant and besides I’m not interested.” There seems to be little one can do about this attitude. It is clearly predominant in the graduate schools, and since these by and large provide the personnel for the undergraduate schools, the same attitude prevails. I believe that this attitude is the principal difficulty encountered in the training of microbiologists for the applied fields. It is somehow regarded as beneath the “important” work of fundamental science. As a secondary issue, I believe that we are teaching too much biochemistry in our beginning microbiology, and most popular texts reflect this error. I think that intermediary metabolism should be relegated to advanced courses, and only an absolute minimum be placed in the general beginning microbiology. The time thus released should be devoted to more discussion of applied microbiology, and if there are any modifications we need to make, over and above attempting to change the attitude toward application, it is to increase the applied areas in microbiology at the expense of metabolic studies. I can hear the cry-that one cannot understand how a cell works unless one understands its intermediary metabolism-but I could submit that many students (and instructors) have a fairly good notion of how a cell grows and how it controls the enzymes within it, without having a very good understanding of intermediary metabolism. I would ( and do ) place intermediary metabolism much later in the sequence of subjects. But such sequences do not really matter too much. I think that we might improve the product a little, but I doubt that we shall make any great change. With respect to persons with the bachelor’s degree, there still remain two problems. The first is that the individual who takes a job in the applied fields at the bachelor’s degree level normally does so because he is not doing too well intellectually and he has given up the idea of medical school or graduate school because his record is not adequate. We tell such individuals that there is indeed a place for them in microbiology, but that if they wish to go farther they must take either of two routes. One is to obtain more education and training to the Ph.D. level if at all possible. Since this is not, in fact, what they do well, we try to point out that they should consider adding to their background other types of activity-sales, or business management, or whatever they might be good in, so that they can develop a career in which their science is a valuable background but not their sole asset and in perhaps a decade they may leave the bench entirely. Still we do have the “late bloomer,” the person who went into application directly for financial or other reasons, and at the bachelor’s level we certainly do have individuals who may have ceased their formal training without reaching their intellectual maximum. Many of these do indeed become bored with
EDUCATION AND TRAINING
IN APPJJED
MICROBIOLDGY
7
their activities and yearn to go to graduate school and obtain advanced degrees. This route is, in fact, a long hard road, but for some few it is desirable. The second problem has to do with the training of persons to the bachelor’s degree. It is our opinion that work at the bench is the “bread and butter” of applied microbiology. We therefore think that such courses should have laboratory instruction and that a considerable effort should be made to make the laboratory the prime vehicle of the learning experience. In this opinion we are opposed by both administration and students. This is an expensive way to teach, not that it requires more materials, but it does require more personnel and more back up facilities, and the number of students taught per dollar expended is considerably lower. We have been forced to cut laboratory instruction partly because the extra personnel required was simply eliminated on the grounds that other departments, not doing at least as much laboratory teaching, were doing a good job without these individuals and therefore we could do it too, “if we only put our mind to it.” It turns out that this is true-if we only lecture (“put our mind to it”) we certainly can teach more cheaply. It is putting our “hands” to it that costs money. Next the students do not like the idea; it takes too long, it’s too confining, indeed it is work. It is only near the end of their training period, when many realize that some credentials must be presented that will lead to a means of employment, that they begin to regard the laboratories in a more favorable light. Frankly, I think most university departments need insistence and assistance (and assistants) on the laboratory type of teaching and that, if more personnel is required, it should somehow be provided. Such training is not exactly the same as producing a product (although we are doing that); one should not try to produce an acceptable product as cheaply as one can, but one should try to produce the best product one can. By the time graduate work is undertaken, the student is committed to a career in microbiology. In certain individuals, this career is directed toward applied microbiology. More frequently, however, the entering graduate student does not know enough about the subject nor about opportunities, to make a reasonable judgment, so that he tends to work in the areas either where there is support or where there is activity that appeals to him, either because of the subject, the personality of the professor, or the ambiance of the laboratory. For the past two decades the support picture has been somewhat unfavorable for applied microbiology in that much of the federal money going into departments of microbiology has been channeled into fundamental rather than applied training. I think it proper to point out that such funding was not due to the mere “glamour” of the rising science of molecular biology, which
8
WAYNE W. UMBREIT
drew so heavily upon microbiology for its materials. In one sense, one could say that the past quarter century was dominated by Escherichia coli just as it is likely that the next decade ( a quarter century is too long to predict) will doubtless be dominated by HeLa. In both cases, I think the reason is the development of simple methods for obtaining quantities of cells. The situation fifteen years ago was that the federal government, through the National Institutes of Health, set about to build a research establishment second to none for a variety of reasons, most of which were quite valid, The NIH used both an internal research laboratory development and a system of external research grants and by the early 1960s had built a research organization in this country that was without question the very best in the world; and they did it economically, with an enormous amount of freedom for the individual investigator, and with due and appropriate concern for the young men and women just beginning their careers. Of the National Institutes of Health, some provided training grants for graduate education which offered, in contrast to the research grants, the freedom of choice for the individual to study what he liked. From about 1965 on, however, the research establishment in microbiology possibly reached the point of diminishing returns; too many people were involved, too large a bureaucracy was required to “run” it, too many restrictions on the freedom of the investigator, too many projects of marginal value, too many reports, too much documentation, etc. There was a breakdown in communication and, in addition to increased cost and a small, well publicized wastage, there was the accumulation within the system of people whose objective seemed to be to destroy it. The system did not, in fact, shrink, but it did not continue to expand. The impact of all of this development upon applied microbiology is that, until about 1965, money was easily available for the support of graduate students and for research, but “fundamental” research was the main beneficiary. Applied areas remained at a low level, and industrial money, which might have been a source of support, was not easily forthcoming, partly on the ground that there was SO much government money. At least such support failed to make any major impact. As of today, the educational facilities are staffed mostly by people brought up during the easy money period (it actually was called “soft” money and encouraged the expansion of many departments far beyond the resources of their university to maintain them), and these people are not, from their training or experience, oriented to application. During this same period, however, the opportunities in applied microbiology outside of the training institution did not markedly expand. The fermentation industry faced serious problems from petrochemicals.
EDUCATION AND TRAINING IN APPLIED MICROBIOLOGY
9
Further, many functions traditionally those of microbiologists were taken over by engineers, biochemists, etc. whose training in microbiology, if any, was minimal. I think that we have now reached the end of this period of doldrums in the applied field and that we will see a marked expansion of applied microbiology. In this essay I have earlier argued that one does not “choose” a career and that applied microbiology is not normally “chosen” in the early stages of work. The same applies to graduate students. The period of graduate work develops research ability, develops a knowledge in depth of one very narrow area of the field, and provides an overall view of the state of microbiological science as contemporary as possible. But an individual develops his (the “or her” is understood) talents and becomes aware of his strengths, and inclinations, in addition to acquiring a vast repository of knowledge. He is a very different individual after the four or more years of graduate work then he was at the start. It is rare that we, his instructors, can tell at the very start of the program whether or not his talents lie in applied microbiology; before long, however, the courses he selects, the kinds of things that interest him, what he studies on his own without being required to do so, begin to shape his acquired knowledge, and these may well orient him toward application in spite of the pressures of graduate schools to other directions. It is at this point, somewhere well along into the graduate program, that an option for applied work should be available. But this should not be a rigid training program. One must seriously consider the nature of the product desired at the end of the training period. In a program designed to train applied microbiologists, one may end up with an individual well aware of today’s technology and problems, but what one really wants is a person capable of developing new approaches and able to face future problems. It happens too frequently to be sheer chance; a man trained in one field makes really significant contributions to another. In a sense he does not know that it can’t be done, so he does it. of course the gap between the two fields must be small; it is rare for a major in English literature to contribute anything to microbiology. But perhaps we are dealing with what Pasteur pointed out as “chance favors the prepared mind.” Pasteur was a good example of this transfer of knowledge. The knowledge of the “diseases” of silk worm, or wine, resulted from the application of principles learned in highly theoretical chemistry. Therefore, a good graduate program should not concentrate on too narrow a range of scientific awareness. Indeed, it is a period of exploration, of filling the store of knowledge that an individual carries, of developing the habit of life-long learning, and of developing that unpleasant but necessary critical facility that questions all statements of fact or all generally ac-
10
WAYNE W. UMBRElT
cepted concepts until experience proves them valid. A good graduate program should have a minimum of time, perhaps one thorough thoughtprovoking course, devoted to all aspects of applied microbiology (from industrial production of antibiotics to lignin decomposition in soils, for example) and a specialty course as, for example, food microbiology, for the training in specific specialties in which the department is engaged, but the rest of the training shouId not be so oriented but should provide more than a smattering of the fundamentals so that the student develops sufficient knowledge to be flexible in his future work. A second factor, in addition to inclination and ability, which determines entry into the applied area is opportunity. A young man who has invested at least four years of his life, who is financially not very well off, will, in a market where jobs are scarce, take whatever offers present themselves, and this is conditioned by the market place. It is certainly unreasonable to train an individual to run a butyl fermentation plant when there are no opportunities to run such a plant. Certainly this was the case four to five years ago when the present crop of Ph.D.’s began their training. The current shortage of petroleum makes butyl fermentation a more economically attractive possibility, and, while there may be few men specifically trained in this area, there are enough good men who know enough of related matters to learn what they need to know rapidly when and if opportunity opens and there is a demand for their services. Two years ago one would not have thought of these possibilities. As such, in the United States, it is my contention that our graduate programs are for the most part doing the right thing by concentrating on fundamentals, but as with the undergraduate programs, the attitude that only fundamentals are important should be changed. And there should be at least one course on applied microbiology taught by a man with experience in one or more of the applied fields. The program should emphasize fundamentals, while developing the relevant aspects of the applied areas. But for the rest I would not tinker with the system; it has produced good men. There is no serious demand for change. We have gone through a period of “innovative” teaching promoted largely by people who were bored with what they were doing. To the best of our knowledge it has not improved the product appreciably, and in some ways it has lowered its quality. Nor is it reasonable to set up specific training programs, partly because an individual’s talents are not clearly evident at the start, and he thus chooses a field when he lacks the very knowledge necessary to make such a choice, and partly because the demand of the market place is continually changing and with a four-year lag in the training program it is impossible to know where the opportunities will be when an individual completes his train-
EDUCATION AND TRAINING I N APPLIED MICROBIOLOGY
11
ing. Since in many cases the inclinations to applied microbiology are not evident until late in the graduate program and since a broad applied training is normally not possible, one solution may be postdoctoral training for a specific applied field. That is, once a career in applied microbiology is in the cards, one knows which part of applied microbiology one is then engaged in, and one could obtain specific “mission-oriented training or indeed experience in the specific applied field (large-scale fermentation, for example, or waste water treatment). A further possibility is to provide to the B.Sc. cheinical engineer, who is almost certainly oriented to the applied field, microbiological training useful in his profession. Surely this training would not need to be nearly as broad based because the outlines of the career and the orientation of the individual are already somewhat delineated. It is my feeling that the training institutions need to be much more flexible in their approach to inultidiscipline or cross-discipline offerings, and rules and regulations of admission to colleges, graduate schools, and specific programs need to be more imaginative and accommodating. Ill.
Training in Applied Microbiology in t h e Developing Countries
The problems of training in applied microbiology in the developing countries are at least different, if not more complex, than those of the developed countries. Further, things are not always what they seem, and the Western eye may read into situations and progranis something quite different from reality. Nevertheless, in devcloping countries, “ppIied microbiology is an iniportniit field and their rather limited resources can frequently be improved by microbial methods. The following observations are based on some practical experience i n teaching in some of the developing countries, of rather close observation of their educational systems and of the environment in which the individual must work. It is aIso derived from the open and very frank discussions of education in niicrobiology in the developing countries at the Congresses of the Global Impacts of Applied Microbiology held at Addis Abnba (1967), Bombay (1969), Sao Paulo, Brazil (1973) as well as the similar discussion held at the Tenth International Congress of Microbiology, Mexico City (1970). I am further indebted to a wide variety of microbiologists who took time from their busy schedules to write to me about their views and who further spent much time and patience with me explaining the situation in their own countries. The distillate of the problems and the solutions I have prepared represents an average opinion, not a consensus--colored, of course, by my own background and experience. Clearly, not all developing countries are alike, and for a Westerner
12
WAYNE W. WMBREIT
to understand any one requires a reasonably long period of living in the country and attempting to work within the system. But there are some problems, which seem to be common to most if not all developing countries, that should be pointed out. The workers in the countries are quite aware of them; indeed, they must live with them. But the rest of the world is not so aware of them; so that in calling attention to the problems we mean no criticism of the developing world, but would hope to assist them. The first problem evident in most countries is lack of maintenance. Excellent buildings of which one can be really proud are, after a year or two, almost nonfunctional. The window is never washed, the faucet leak is never repaired, there is no gas in the supply line, the centrifuge has rusted, and the electricity is unreliable. For the research laboratories this is not insurmountable; the inhabitants of the laboratory can do it themselves or improvise. Whether they do or not depends on the leader and upon the customs of the country. But teaching laboratories, which are frequently shared by several departments, have no such regular inhabitants, and usually no one takes the slightest responsibility. Very similar things happen in the laboratories of the developed countries as well. It is very difficult to teach skills and precision under these circumstances. One solution is to channel more money into maintenance, and in most developing countries maintenance personnel are available; indeed there is usually an excess of untrained personnel-“bearers,’’ “porters,” “assistants,’’ etc. Unfortunately, these individuals often do little more than make tea or coffee (not always of drinkable quality) and, indeed, frequently hamper the work of the department, but they cannot be replaced for a variety of understandable reasons. This means that the department head must set the tone, and I have seen places run under these very conditions that are models of upkeep. The department head must see to it that the laboratory is maintained, that a budget is provided for paint and plumbing, for soap and scrub brushes, and that the men available to do the work, do it right; and at times, the head must actually do the maintenance himself. The credentials of the department head in microbiology may be excellent, but unless he is also a good plumber and spends a great deal of his time seeing that the housekeeping details are adequately taken care of, chaos soon results. A second serious difficulty is the lack of technical infrastructure. This means not only the lack of technicians but of the whole concept of a technical competence, of skill, persistence, adaptability, and responsibility-indeed, of pride in work of scientific and technical nature. People with these skills are poorly paid, are not treated with the respect they deserve, are regarded as merely a pair of hands, replaceable, without
EDUCATION AND TRAINING IN APPLIED MICROBIOLOGY
13
a status above, that of the “bearer.” Yet they are considerably above the “bearer” status in terms of education, skills, and background. There are, it is true, all too few of them, and they ought, therefore, to be regarded with more respect and attention. It is futile to set up schools for training technicians when technicians are so poorly treated after such training. There are both surface and subsurface reasons for this situation. Sometimes the cause arises from deep-seated attitudes of the culture, but clearly it constitutes a difficulty. A third difficulty in most developing countries is the actual size of the undergraduate classes. A general microbiology course, for example, may be required for all majors in premedicine, agriculture, biological sciences, pharmacy, and a variety of others. A class may consist of 1000 students and may be arbitrarily increased to 2000 due to some regulation in the “ministry” of education without any change in personnel. The course must have a laboratory (usually specified in the regulations) and must occupy X number of hours. Compared to Western standards this is not desirable. It is clearly impossible to teach laboratory skills under these circumstances. Hence, much is done by demonstration. It is remarkable indeed that the students learn anything at all, and transfer to a graduate school in a developed country must be a terrifying experience. The more advanced courses are usually also overcrowded or, at times, completely neglected. In many countries the course content is also controlled by a syllabus-prepared in many cases by the ministry of education. Unfortunately, there seems to be Iittle input by microbiologists (and especially little application). It has been suggested frequently that a worldwide textbook on microbiology backed by some international organization might be useful in changing the course content. While the results of such efforts may be dubious, the difficulties in writing such a text are real. A fourth difficulty is that opportunities for effective positions are limited, and we see countries struggle to train hundreds of people even though they have an excess of people already trained but unemployed or working at alternative positions. The microbiologist working as a cab driver is not at all uncommon, and salaries of even well established individuals are barely enough to live on, so moonlighting is also common. These four difficulties-lack of maintenance, lack of technical infrastructure, excessive size of classes, and lack of opportunity once training is complete-seem to be characteristic of the majority of the developing world although clearly there are exceptions. Let us, however, make it clear that it is not the fault of the microbiologists involved, and they are quite aware of these difficulties. In most cases to overcome one of the difficulties is only to exaggerate another, and usually the cause of the problem lies in some unapproachable ministry or administrative
14
WAYNE W. UMBREIT
unit. At times it is sheer inertia and the custom of centuries that one must reform, and frequently this is not only difficult but also dangerous. Since it is unlikely that many in the developing world will read this essay (one limitation being that most do not have funds to buy publications of this sort), I am clearly not speaking to them, but I am pointing out that these are the conditions under which they must work and solutions or programs of the developed world are not directly transferable unless these factors are taken into account. Applied microbiology in the developing countries has three primary functions. One is the improvement of existing applications of microbiological science-legume inoculation or composting, for example. A second is to develop new resources: antibiotics from the sisal plant, for example, or citric acid from the local sage brush. And the third is the transfer of such portions of world technology as are applicable to their situation-for example, the production of butyl from local molasses or the control of powdery mildew. Within the limits of the system and such reforms as are locally possible, there seems to be a reasonable approach to applied microbiological training based upon two basic principles. The first is the “need to know”; i.e., it is not necessary for every microbiologist to know everything. Especially at the technician level, one may train individuals for certain tasks (public health laboratories, for example) without a complete background in mathematics or organic chemistry. Where there are specific objectives in mind, the training program can be concentrated upon the requirements of those objectives rather than broadly scattered. The second principle is that while the introductory material may be widely disseminated, only a small percentage with talent and imagination need to be trained further. At the college level, a class of perhaps 1000 students in the beginning course can be taught by way of lecture or lecture-demonstration only and those who wish to go on into microbiology obtain their first laboratory in the second course. These two principles, together with some sincere emphasis on maintenance, would go a long way to improve the technical and undergraduate programs of most developing countries. It seems unreasonable to expect that undergraduate training can be accomplished outside of the country for any but a small number. With respect to graduate education, there is a sharp division of opinion as to whether it should be accomplished within the country or outside it. Many of those who have obtained their Ph.D.’s externally (and some of those with internal Ph.D.’s) are strongly of the opinion that a local Ph.D. is of low merit. But there are numerous difficulties found by the student in obtaining his Ph.D. externally in addition to the sometimes insurmountable one of financing its cost. First, his laboratory training is generally most inadequate in preparing him to compete with the
EDUCATION AND TRAINING IN APPLIED MICROBIOLOGY
15
average graduate student of the Western world. Next, he must do his work in a language not his own, and this is normally given very little consideration. He also becomes involved in a research project (since a Ph.D. is a research degree) usually chosen by his professor; he does well at it, it becomes his absorbing interest; he returns to find that he really cannot work on this project or anything comparable and that there are other limitations he had forgotten while abroad. Thus, a significant percentage of the returned Ph.D.’s, who certainly represent an important potential for the country, are bitterly unhappy and frustrated. Frequently, also, upon return, they are arbitrarily assigned to positions that in many cases bear little relation to their training. In the external country itself, the chance of a student getting to the right place, i.e., where he can get the training useful to his country, is relatively small and depends to a large extent upon those his undergraduate professors happened to know when they were out of the country twenty years ago. Further, on moving to the external country such students undergo “cultural shock and upon return have become alienated from their own country. To avoid this difficulty it has been proposed that institutes for the training of applied microbiologists should be established in locations not too far removed from the countries requiring them, to minimize the culture shock and the cost while yet retaining the advantages of foreign experience. For example, an institute in Beirut (where it could make use of university facilities) teaching in both English and French might be suitable for the countries of Africa. A similar applied institute in Hong Kong or Singapore would be suitable for the far east, and one in Puerto Rico or Mexico for the Central and South American countries. Many countries in these geographical areas have adequate facilities, but for those who do not, and for those who hold that experience in the “outer world is important, such institutes could provide both the basic training necessary and a transition phase before further study elsewhere. Such proposals are favorites of the international congresses, but success in organizing and financing them seems rather remote. A student from the developing countries entering the graduate schools of microbiology in the United States will be given extensive training in biochemistry, physiology, molecular biology, and genetics but rather less, if any at all, in applied microbiology. Yet in many developing countries the feeling is that these subjects are expensive and useless toys and that the need for such knowledge has not yet reached the point of relevancy. Many countries select and send the very best of their young people (although the selection i s rarely for dedication to applied aspects), and then the students given such training are “useless” (or worse) upon their return. “All of this wonderful training is wasted
16
WAYNE W. UMJ3ElElT
is the complaint from the individuals themselves. There is something to be said for this position. However, the programs of any country are designed for its own students, and they are not specifically oriented toward a country a thousand miles away with a different culture and requirements, Such training rarely pretends to provide any solution to the problems of the developing nations. The bright young men sent to the prestige universities become fascinated by the subjects offered, and these are what they study, Some, it is true, on return cannot readapt to a clearly different situation in which they are now placed. And while it is senseless to suggest that developing countries ought to concentrate only on applied matters, it is probably equally senseless to suggest that the training obtained should be essentially applied. Applied training is only rarely available, is likely to concentrate on advanced technology perhaps not feasible for the country of origin, and is likely not to produce men of vision, who are the ones most needed. It is thought by many that foreign travel and training is a necessary part of the Ph.D. training. An internal degree program with perhaps a six months sojourn in a developed country, or perhaps a mission-oriented program, if such were developed, might be a solution to the problem. But surely some of the best men should be Westernized, should “think Western” to provide the necessary push to modify the system and gradually overcome its present difficulties. Some of the negative aspects of training in the developed countries could be overcome if the student were assisted in getting to the right place, where the useful training for his country and his talents was available. On the other hand, there are those who feel that an internal degree is more useful since it is more attuned to facilities and needs of the country. Certainly a great deal of money (both internal and external) has been put into the higher education and facilities of many countries, and some of them are excellent. Further, excellent men have been (and thus can be) produced by these internal facilities when circumstances are right, granting that a certain percentage are not very satisfactory. A reasonable solution to the dilemma of culture shock for those countries having the facilities would again be to concentrate upon the internal degree, and then, a year or two later, when the man has settled into a position, send him abroad on a mission-oriented visit to one or more institutions for a relatively short period, perhaps 3 months at a given location. As a general compromise one could suggest the following: 1. A B.Sc. at home 2. An MSc. with emphasis on the applied aspects at an institute for applied microbiology ( Beirut, Puerto Rico, or Singapore) 3. Two years of work in applied microbiology in the home country
EDUCATION AND TRAINING IN APPLIED MICROBIOLOGY
17
4. Either: ( a ) a Ph.D. abroad at an institution providing the applied training ( b ) a local Ph.D. with a postdoctoral (of perhaps 6 months) in a developed country in a field related to the work he is expected to do. This kind of “mission-oriented work could be most useful. The advantages of foreign training could be provided by step 2, and the transfer of developed nation technology could be assisted by step 4, especially step 4b. IV.
Summary
For the United States, there is a lack of facilities and opportunities for the training of technicians. There is at the college level neglect of training in applications; at the graduate level, there is the general lack of broad training in applied microbiology although specialty training is adequate, Each training level has certain problems, but the general assessment is that there is reasonably adequate response to the demand. Training in the developing nations is a somewhat different problem and may require somewhat different solutions. Some of the problems have been outlined and some solutions have been proposed. ACKNOWLEDGMENTS
I wish to especially thank Dr. B. W. Koft and Dr. R. W. Thoma for very pertinent suggestions. While I am indebted to many foreign colleagues, I should like to mention especially the following: (Argentina) M. D.’Aquino, G. W. Marengo, T. S. Molina, S. Soriano; (Brazil) W. Borzani, L. P. De C Lima, J. S. Furtado, A. Neghme, W. Schmidell Netto; (Costa Rica) F. Montero-Gei; (Egypt) Badr El-Din, A. S. El Nawawy, M. A. Foda, M. A. Samie, S. Taha; (Ethiopia) R. M. Baxter; (Ghana) G. Tewari; (India) J. V. Bhat, F. Femandes, K. G. Gopal Krishanan, S. Gorhe, C . Rangaswami, M. S. S. Rao, T. N. R. Rao, D. V. Rege, S. R. Saksina, G. D’Souza, S. R. Vyas; (Kenya) M. G. Rogoff; (Mexico) C. CasasCampillo, M. Servin-Massieu; (Pakistan) A. Anwar; (Thailand) S. Sirisinha; (Uganda) J. Mukiibi; (Venezuela) A. Divo, J. Gomez-Ruiz, J. J. Gutierrez-Alfaro.
This Page Intentionally Left Blank
Antimetabolites from Microorganisms
DAVDL. PRUESSAND
JAMES
P. SCANNELL
Chemical Research Department, Hofmann-La Roche Znc., Nutley, New Jersey
I. Introduction
...................
11. Definition of an Antimetabolite ........... 111. Detection Methods ................
IV. Mechanisms of Action and Reversal . . . . . . . . . . . . . . . . . V. Antimetabolites Isolated from Microorganisms ..... A. Amino Acids and Related Compounds ...... B. Nucleosides and Related Compounds ....... C. Vitamin Antimetabolites ............ D. Miscellaneous Antimetabolites . . . . . . . . . . . . . . . . . . VI. Incidence of Antimetabolite Production ....... VII. Chemotherapeutic Applications ........... Note Added in Proof ............... References ...................
I.
19 19 21 22 24 25 38
44 48 50 51 53 54
Introduction
Although almost 2000 antibiotics had been described by the rnid-l960’s, only about 10 of these were commonly regarded as antimetabolites. Since then, several microbiological screens have been developed specifically designed to detect antimetabolites, and the number of antimetabolite antibiotics has increased to about 70 compounds. The purpose of this review is to describe the screening methods, to catalog the antimetabolites, and to summarize what is known about the biochemistry of these compounds. We hope, thereby, to stimulate an interest in the biosynthesis, mechanism of action, and basis of selective activity of this group of compounds. II.
Definition of an Antimetabolite
Woolley (1963) defined an antimetabolite as ‘a structural analog of an essential metabolite, vitamin, hormone, or amino acid, etc., which is able to cause signs of deficiency of the essential metabolite in some living thing or in some biological reaction.” This definition was based on the theory that antimetabolites act as isosteric enzyme inhibitors ( Woolley, 1952). Accordingly, reversible inhibition would be relieved by the addition to the test system of either the product or an excess of substrate. In the case of irreversible inhibition, only addition of product would be effective. Thus, experimental reversibility has become commonly accepted as evidence for antimetabolite activity, although this is not explicitly required by the original definition. The association of 19
20
DAVID L. PRUESS AND JAMES P. SCANNELL
experimental reversibility with the original definition of an antimetabolite has led to two major difficulties: (1) Antibiotics, such as penicillin and puromycin, can be considered as structural analogs of essential metabolites which cause signs of deficiency of the metabolites. However, since the metabolites are often complex macromolecules ( cell wall components, proteins, nucleic acids), they cannot be effectively supplied to an intact cell, hence reversibility cannot be demonstrated. (2) Often there is no apparent structural analogy between reversant and antimetabolite. This can sometimes be explained by the complexity of metabolic pathways and control mechanisms. It is possible to devise screens to detect reversible activities but structural analogy can be made only after isolation and structure determination. Thus, a definition based on reversal of activity rather than structural analogy is more useful in the search for naturally occurring antimetabolites. In this review an antimetabolite is defined operationally as an organic compound which at a low concentration has a deleterious effect on the growth or viability of an organism, the effect of which can be prevented by concurrent administration of one or more common biochemicals. It should be noted that an in vivo effect is required by this definition; thus, compounds that have only been shown to be in vitro enzyme inhibitors are not included. However, a successful program has been developed ( Umezawa, 1973 ) for screening fermentation broths for inhibitors of specific enzymes. This approach is more narrowly focused than the antimetabolite screen since feedback inhibitors and repressors cannot be detected. However, it has the advantage that inhibitors of nonmicrobial enzymes can be directly discovered. The screening of fermentation samples for antibiotic activity is usually based on observing growth inhibition of one or more test organisms. The reason antimetabolites are seldom discovered by this method is that the media used for the growth of the test organisms usually contain substances such as yeast extract and meat peptones. These media provide all the common biochemicals and therefore would be expected to reverse the activities of any antimetabolites present in the tested samples. Many microorganisms possess all the enzymes necessary to synthesize the common amino acids and vitamins from simple nutrients and therefore can grow well on a glucose-salts medium with ammonium ion as the sole source of nitrogen. Growth of a microorganism in such a medium can be prevented by the presence of an antimetabolite since there are no external sources of metabolites which can reverse the activity of the antimetabolite. The technique of growing test organisms on a defined minimal medium has been extensively applied to investigations of the properties of synthetic antimetabolites ( Richmond, 1962; Kaiser, 1960), but until five years ago the method was infrequently applied to the
ANTIMETABOLlTES FROM MICROORGANISMS
21
detection of antimetabolites in fermentation broths. Not only are antimetabolites detected by this method, but they are also differentiated from antibiotics since the activity of the latter is not reversed by the addition of nutrients to the assay medium. Moreover, the metabolic pathway affected can be determined by the selective addition of individual nutrients to the assay medium. For example, the antimetabolite activity of L-3(2,5-dihydrophenyl) alanine (XX ) is reversed by addition of L-phenylalanine or L-tyrosine to the assay medium, but not by the addition of other amino acids ( Fickenscher and Zahner, 1971). Ill.
Detection Methods
Techniques to demonstrate antimetabolite reversibility have been known for some time and have been reviewed by Woolley (1952) and Kaiser ( 1960). The techniques stem primarily from studies on synthetic antimetabolites. Of special interest is the first description of the counterdiffusion agar-well test developed by Cuthbertson et al. (1956) in their studies of vitamin B, antimetabolites, and which was later modified by Smith (1965) and Simon (1969) to measure indexes of inhibition. Outgrowth of Neurospma CTUSSU from an agar well into minimal medium has been used to measure the effect of many synthetic antimetabolites ( Fuerst and Skellenger, 1958; Fuerst, 1964). Techniques of screening microbial products for antimetabolite activity have been developed independently in several laboratories. A very sensitive spread-plate method developed by Foster and Pittillo (1953a,b) to detect reversal of antibiotics by metabolites has been used to select synergistic pairs of antimetabolites ( Pittillo and Foster, 1954). However, this technique does not appear to have been used widely by others. Schabel and Pittillo (1961) described a method to detect antimetabolite activity in those antitumor compounds which inhibit the growth of one or more microorganisms in minimal agar medium. The antitumor compound is incorporated into seeded agar at a concentration sufficient to inhibit completely the growth of the test microorganism. Filter paper disks impregnated with various metabolites are placed on the agar surface. After suitable incubation, growth around one or more of the metabolite-containing disks denotes antimetabolite activity. A counterdiffusion disk assay has been used by Stapley et al. (1968) to screen for antimetabolites in fermentation broths. In their method, paper disks containing metabolites are first applied to seeded agar and then disks containing fermentation broth are placed near the metabolite disks. If a zone of inhibition caused by the fermentation broth is distorted or eliminated by one or more of the metabolites, antimetabolite activity is indicated.
22
DAVW
L. PRUESS AND JAMES P. SCANNELL
The “Kreutztest” used in the laboratory of Professor Zahner (Zahner et al., 1960; Keller-Schierlein et al., 1969) also employs the counterdiffusion principle. Two paper strips, one containing sample and the other containing metabolite, are placed crosswise on seeded minimal agar. From the shape of the zone of inhibition along the sample strip, a reversal can be demonstrated to be competitive or noncompetitive in nature. A method which probably evolved from work on the microbiological assay ( Hanka and Burch, 1960) and mechanism of action (Hanka, 1960) of psicofuranine has been described by Hanka (1968). Broth samples are placed separately on nutrient-agar and minimal-agar plates which have been inoculated with the same test microorganism. Antimetabolite activity is recognized as a zone of inhibition occurring on only the minimal-agar plate. Leads are then evaluated in a second stage where broth samples are tested on four inoculated minimal-agar plates, containing low concentrations of one of the following: common purines and pyrimidines, amino acids, vitamins, or Krebs cycle intermediates. Samples which give no inhibition on one or more of these plates are chosen for the third and final stage of testing where individual members of the metabolite groups are used to supplement the agar. By means of this three-stage procedure, it is possible to detect antimetabolites and to differentiate them on the basis of the metabolites which cause reversal. IV.
Mechanisms of Action and Reversal
The counterdiffusion method may be used to differentiate two types of reversal mechanisms. A competitive reversal is shown in Fig. 1A.
A
B
FIG. 1. Counterdiffusion antimetabolite assay. ( A ) Competitive reversal. ( B ) Noncompetitive reversal.
A'I'lMETAB0LITES FROM MICROORGANISMS
23
The boundary between the area of inhibition and the area of reversal is a straight line. At all points on this line the ratio of inhibitor to reversant is constant. In contrast, a noncompetitive reversal is shown in Fig. 1B. In this case, all points on the boundary are equidistant from the center of the disk from which the reversant diffused. At that distance the concentration of reversant is sufficient to supply the minimum requirement for growth, and the concentration of antimetabolite which varies along the boundary is not a factor. A noncompetitive reversal does not imply that an enzyme is inhibited in a non- or uncompetitive manner. In a whole-cell system, an exogenous supply of the product of an inhibited enzyme or metabolic pathway is often the most efficient reversant. Such product reversals are noncompetitive. The type of reversal may be related to the mechanism of action of the antimetabolite. An isosteric enzyme inhibitor is reversed noncompetitively in vivo by the product of the inhibited reaction and often by subsequent compounds in the metabolic pathway. Sometimes these compounds are not structurally related to the enzyme inhibitor, as, for example, certain glutamine analogs which are reversed by purines. The substrate and possibly its metabolic precursors, which also may be structurally dissimilar, will reverse competitively. However, in cases where the inhibitor is very tightly bound, substrate reversal may not be demonstrable. Many metabolic pathways are regulated by the end product of the pathway by two mechanisms. The first enzyme in the metabolic sequence may be inhibited allosterically; i.e., the end product of the pathway inactivates the first enzyme in the pathway by binding at a site other than the active site of the enzyme. This is commonly referred to as feedback inhibition. When an end product analog causes inhibition, synthesis of the end product and its precursors is stopped, and an antimetabolite effect is observed which can be reversed noncompetitively by the end product and many of its precursors. Alternatively, an end product or end product analog may repress the synthesis of many or all of the enzymes in a metabolic sequence. If such is the case, only the end product of the metabolic sequence will reverse, and the reversal will be noncompetitive. When intersecting metabolic pathways are controlled by such mechanisms, unusual reversal effects can be observed. For example, uracil inhibits the growth of an Escherichia coli mutant, and this growth inhibition is relieved by addition of L-arginine or L-citrulline to the medium (Novick and Maas, 1961). The mechanism underlying this effect is well understood. Uracil is converted to uridylic acid, which represses the biosynthesis (Anderson and Meister, 1966; Pierard et al., 1965) and inhibits the activity (Pierard et al., 1965) of carbamyl phosphate synthetase. These feedback mechanisms not only prevent the un-
24
DAVID L. PRUESS AND JAMES P. SCANNELL
necessary biosynthesis of pyrimidine nucleotides by shutting off the sup-
ply of carbamyl phosphate, but also prevent the carbamylation of ornithine to citrulline, the immediate precursor of the essential amino acid arginine. This example of the complexity of metabolic controls provides a partial explanation for two apparent anomalies sometimes observed with antimetabolites: (1) lack of structural analogy to reversant; ( 2 ) antagonism between common essential metabolites. In addition, this example illustrates lethal synthesis whereby the inhibited organism biosynthesizes the active antimetabolite ( uridylic acid) from the administered precursor (uracil). Another mechanism of antimetabolite action which has been postulated (Mathews, 1958) is the incorporation of the antimetabolite into macromolecules which then have altered activities. One would expect a competitive reversal for this effect. Transport may be involved in antimetabolite reversals. Thus, normal metabolites may interfere with the uptake of a toxic substance thereby preventing growth inhibition. Such reversals can generally be recognized when two structurally related but metabolically unrelated compounds ( e.g., arginine and lysine ) relieve inhibition competitively. It should be noted that even though reversal occurs by a transport mechanism this does not necessarily mean that the mechanism of action of the toxic substance is at the transport level. Reversal patterns are not always clear cut and may vary markedly with different test organisms. At best, they can be taken only as a guide to the enzyme studies needed to define a mechanism. Detailed reviews have appeared on the mechanisms of action of amino acid (Fowden et al., 1967; Nass et al., 1971; Woods, 1964), water-soluble vitamin ( Woolley, 1963), and nucleoside ( Suhadolnik, 1970; Wagner, 1971) antimetabolites.
V.
Antimetabolites Isolated from Microorganisms
Seventy antimetabolites are described in detail in this section, Of these, 8 are vitamin structural' analogs, 14 are nucleosides or related compounds, 25 are amino acids, 8 are di- or tripeptides and 15 are of undetermined or miscellaneous structure. Excluded are antibiotics which react chemically with broth constituents to give apparent reversal of activity. Among these are naphthomycin, which is inactivated by thiols (Balerna et al., 1969); aranciamycin, by amino acids (Keller-Schierlein et al., 1970); and penicillic acid, by amines ( Oxford, 1942). Cell membrane active antibiotics, which are reversed by a variety of sterols, lipids, and phospholipids (Halder and Bose, 1971), are also excluded. It should be noted that a somewhat
ANTIMETABOLITES FROM MICROORGANISMS
25
bewildering array of reversants, including certain amino acids, have been reported for the actinomycins (Modest et al., 1963). Several other antibiotics are reported to be reversed by various substances. Gottlieb and Shaw (1967) have discussed such reversals for sarkomycin, mitomycin, puromycin, cycloheximide, tetracycline, and patulin. In addition a reversal of iodinin by naphthoquinones and anthraquinones ( McIlwain, 1943) and a reversal of daunomycin by xanthine (Pittillo and Hunt, 1968) have been reported. Other reversals in the older literature could well have been overlooked. The sideromycins, iron-binding peptide hydroxamate antibiotics, are reversed by the structurally related sideramines. These have been reviewed elsewhere (Emery, 1971; Maehr, 1971). Also excluded are the many antagonisms (Meister, 1965) between the common naturally occurring amino acids such as that reported b e tween glycine and alanine (Snell and Guirard, 1943). The following antimetabolites are arranged somewhat arbitrarily by chemical structure. Any structural uncertainties, synthetic procedures or resolutions have been noted. If not otherwise indicated, proof of structure is contained in the publication listed under “Source.” The comments under “Mechanism” usually refer to in vitro observations, but in only a few cases has a definitive in vivo mechanism been established. In general, the biosynthesis and mode of action of many of these compounds remain to be investigated. Only the nucleoside antibiotics have been intensively studied in this respect, and these compounds have been thoroughly reviewed elsewhere ( Suhadolnik, 1970 ) . Four compounds (XXII, XXVII, XXVIII, and duazomycin C ) closely related to known antimetabolites have been included although reversal studies have not been reported. It should be noted that about half of the compounds in this list, particularly the nucleoside antimetabolites, were first detected in antibiotic or antitumor screens and only later shown to be antimetabolites. A. AMINO ACIDSAND RELATEDCOMPOUNDS
1. Aliphatic Amino Acids
~-2-Amino-4,4,-dichlorobutanoic acid c‘
NH2 0 I II CH-C HZ - C H - C - 0 H
/
CI’
Reversant: Source:
(I)
L-Leucine in Proteus mirabilis, Eschenichia coli, P. vulgaris, P. rettgeri ( Argoudelis et al., 1967) Streptomyces armentosus ( Argoudelis et al., 1967)
26
DAVID L. PRUESS AND JAMES P. SCANNELL
L- ( threo ) -2-Amino-3,4-dihydroxybutanoic acid NHZ 0 I II HO- C H z - C H - C H - C - O H I OH (III
Reversant: Source: Synthesis:
L-Serine, L-threonine in Escherichia coli ( Westley et al., 1971) Streptomyces species ( Westley et al., 1971) ( Hamel and Painter, 1953)
L-2-Amino-3-dimethylaminopropionic acid ( L-4-azaleucine) CH
NH2 0 'N-CH2-CH-C-OH I II
/
CH3
(rn)
L-Leucine in Proteus mildenbergii, P. rettgeri, and Sarcina lutea (Argoudelis et al., 1967), L-leucine ( competitive) in Escherichia coli and Leuconostoc dextranicum (Smith et al., 1963), L-leucine in Salmonelkz typhimurium (Stieglitz and Calvo, 1971) Source: Streptomyces neocaliberis ( Argoudelis et al., 1967) Mechanism: Inhibition of cY-isopropylmalate synthetase, the first enzyme unique to leucine biosynthesis; and incorporation into protein (Stieglitz and Calvo, 1971) Synthesis: Racemate (Smith et at., 1963) Reversant:
~-2-Amino-4-methoxy-trans-3-butenoic acid NH2 0 I II C-OH
H I
CH3 - 0 - C = C - C H I
H
cm, Reversant:
Source:
D-Alanine, L-alanine, D-ghtamic acid, D-aspartic acid, D-2-aminobutanoic acid, L-2-aminobutanoic acid in Bacillus sp. 1283B (Scannell et al., 1972a). Methionine, homocysteine, cystathionine in Escherichia coli (Sahm et al., 1973). Pseudomonas aeruginosa (Scannell et al., 1972a; Sahm et al., 1973).
ANTIMETABOLITES FROM MICROORGANISMS
27
Mechanism: The unusual reversal pattern (Scannell et al., 1972a) may indicate the compound affects enzymes involved in cell wall biosynthesis. In E . coli pseudo feedback effects on homoserine O-succinyltransferase are probably involved ( Sahm et al., 1973).
~-2-Amino-4-methyl-5-hexenoicacid Y 3 NH2 I O II HzC=CH- CH-CHZ-CH-C-OH
IP I
Leucine in Staphylococcus aureus, Bacillus subtilis, Sarcinu lutea (to a lesser extent by valine and isoleucine) (Kelly et al., 1969) Streptomyces species ( Kelly et al., 1969), Boletus Source: ixocomus nudi (New Guinea fungus) (Rudzats et al., 1972) Stereochemistry: Unknown at C-4 (Kelly et al., 1969; Rudzats et al., 1972) Reversant:
L-2-Amino-4-pentynoicacid ( propargylglycine ) NH2 0 I I1 H-CEC-CH2-CH-C-OH lpIi
L-Methionine, L-leucine in Bacillus subtilis ( Scannell et al., 1970a); L-valine in Saccharomqces cerevisiae (Shapira and Dittmer, 1961) Streptomyces species (Scannell et al., 1971a) Source: Synthesis: Racemate (Gershon et al., 1949, 1954; Schlijgl, 1958; Shapira and Dittmer, 1961) Synthesis and resolution: (Scannell et al., 1971a) Reversant:
L-S-Ethyl homocysteine ( L-ethionine )
Reversant:
L-Methionine in Escherichia coli (Harris and Kohn, 1941) and the rat (Dyer, 1938); adenine and ATP in the rat ( Villa-Trevino et al., 1963)
28
DAVID L. PRUESS AND JAMES P. SCANNELL
Source:
Synthesis: Mechanism:
Escherichia coli, Bacillus megaterium, Pseudomonas aeruginosa and Aerobacter aerogenes ( Fisher and Mallette, 1961) (Dyer, 1938) Detailed review (Fowden et al., 1967)
L-Selenomethionine CH3
- Se -CH2-CH2
Y 2:
-CH-C-OH
[rn) Reversant: Source: Synthesis: Review:
L-Methionine in Chlorellu vulgaris ( Schrift, 1954, 1961) Escherichia coli (Tuve and Williams, 1961) (Painter, 1947) (Fowden et al., 1967)
2. Basic Amino Acids L-2-Amino-3-( 2-aminoethoxy ) propionic acid ( L-4-oxaly sine ) NH2 0 I II H2N-CH2-CH2 - 0 - C H 2 - C H - C - 0 H
(Ix)
Reversant:
L-Lysine, diaminopimelic acid and L-lysylglycine in Escherichia coli (Stapley et al., 1968); L-lysine in several species of Lactobacillus and Streptococcus (McCord et al., 1957); L-lysine (competitively) in Leuconostoc destranicum ( McCord et al., 1957) Source: Streptomyces chartreusis and Streptomyces erythrochromogenes (Stapley et al., 1968) Synthesis : Racemate (McCord et al., 1957; Tesser and Nefkens, 1959) Resolution : (Tesser et al., 1962) L-2-Amino-4-( 2-aminoethoxy ) -trans-Sbutenoic acid NH2
H
I
I
H2N-CH2-CH2 - 0 - C = C -
0 II
CH-C-OH I
H
(XI
Reversant:
Source:
L-Cysteine ( noncompetitive), L-arginine, and L-lysine (competitive) in Streptomyces cellulosae (Pruess et al., 1974) Streptomyces species (Pruess et al., 1974)
ANTIMETABOLITES FROM MICROORGANISMS
29
2-Amino-4-( 2-amino-3-hy droxypropoxy ) -tram-3-butenoic acid ( rhizobitoxine) H NH2 0 NH2 I I I 11 HC-CH2-O-C=C-~~-C-O~ I I CH2-OH n
(XI)
Homocysteine, L-methionine in Salmonella typhimurium (Owens et al., 1968); L-methionine partially reverses inhibition of ethylene biosynthesis in apple and sorghum (Owens et al., 1971). Rhizobium japonicum (Owens and Wright, 1965) Source: (Owens et al., 1972) Structure: Mechanism: Inhibition of p-cystathionase ( Giovanelli et al., 1971) Reversant:
L-N~( l-iminoethyl ) ornithine NH2 0 y 3 I II HN=C - NH - C H ~C H ~ C - H -~C H - C - OH
-
Reversant: Source: Synthesis:
L-Arginine, L-citrulline, L-ornithine, and L-NZ-acetylornithine in Escherichia coli (Scannell et al., 1972b) Streptomyces species (Scannell et al., 1972b) (Scannell et al., 1972b)
3. Acidic Amino Acids ~-2-Amino-3( N-nitroso) hydroxaminopropionic acid ( alanosine ) N=O NH2 0 I I I1 HO-N-CH2-CHC-OH
(Xm)
Reversant: Source: Mechanism:
L-Aspartic acid and adenine in Candida albicans (Gale et al., 1968) S,treptomyces alanosiruicus ATCC 15710 (Murthy et al., 1966; Coronelli et al., 1966) Postulated inhibition of aspartate transcarbamylase and adenylosuccinate synthetase (Gale et al., 1968). Inhibition of glutamic acid carboxy-lyase has been demonstrated (Gale and Atkins, 1968)
30 (
DAVID L. PRUESS AND JAMES P. SCANNELL
+) -2,SDiaminosuccinic acid 0 NH2 NH2 0
I l l I II HO-C-CH-CH-C-OH
mr
Reversant: Source: Synthesis:
L-Aspartic acid in Escherichia coli (meso form). The optically active form has apparently not been tested (Shive and Macow, 1946) Streptomyces rimosw ( Hochstein, 1959) Review of dl and meso forms (McKennis and Yard, 1958)
N-Formyl-N-hydroxyglycine ( hadacidin ) 0 II
0
CH
II HO-N-CH2-C-OH I
(XPI
L-Aspartic acid and adenine in KB cells (Neuman and Tytell, 1963), Escherichia coli (Shigeura, 1963), and chloroplasts from Euglena (Mego, 1964) Source: A wide variety of Penicillia including the following species: P. frequentans (Gittennan et al., 1962), P. purpurescens (Gray et al., 1964), P . aurantio-uiolaceum, P . caseicolum, P . cr&osum, P. implicatum, P. janthinellum, P . liuidum, P . spinulowm, P . trzebinskii, and P . turbatum (Dulaney and Gray, 1962) Synthesis: (Kaczka et al., 1962) Mechanism: Inhibition of adenylosuccinate synthetase ( Shigeura and Gordon, 1962) Review: (Shigeura, 1967) Reversant:
4. Cyclic Amino Acids L-%Amino-( cis-2,5-dihydro-5-methyl)furan-2-acetic acid ( furanomycin; threomycin ) 0
II H2N - C H - C - O H
I
^. ,.H HC’IILo HCiC’
/ .. CH3
ANTlMETABOLITES FROM MICROORGANISMS
31
L-Isoleucine ( competitive), L-valine (noncompetitive) in Escherichia coli (Katagiri et al., 1967) Source: Streptomyces (Katagiri et al., 1967) Mechanism: Inhibition of isoleucyl tRNA synthetase at high antimetabolite concentration (Tanaka et al., 1969) Reversant:
l-Amino-2-.nitrocyclopentanecarboxylic acid
(Xm)
Reversant: Source: Synthesis: Stereochemistry:
Leucine in pea root (Brian et a,?.,1965) Aspergillus wentii (Burrows and Turner, 1966) (Burrows and Turner, 1966) Unknown
2- ( l-Cyclohexen-3( R ) -yl ) -S-glycine and l-cyclohexen-3( R ) -yl glyoxylic acid NH2 0
CH-cHz
//

I
l
l
CH -CH-C-OH HC / ‘cH~-cH~
(Xm)
Competitively by L-homoserine, L-threonine, a-ketobutyrate, a-keto-p-methylvalerate, a-ketoisovalerate, L-valine, a-ketoisocapronate, and L-leucine, and noncompetitively by L-isoleucine in Bacillus subtilis ( Keller-Schierlein et al., 1969) Source: Streptomyces antibioticus ( Celmer, 1964) Synthesis: (Edelson et al., 1958) Mechanism: Competitive inhibition of threonine deaminase after lethal synthesis of the amino acid from the keto acid ( Keller-Schierlein et al., 1969; Zahner and Poralla, 1970; Kruepe and Poralla, 1972) Review: (Nass et al., 1971)
Reversant:
32
DAVID L. PRUESS AND J A M E S P. SCANNELL
~ - 32,s-Dihydrophenyl ( ) alanine
(xx) Reversant:
L-Phenylalanine, L-tyrosine in Eschericlaia coli ( Shoulders et al., 1968; Scannell et al., 1970; Fickenscher et al., 1971),Leuconostoc dextranicum (Snow et al., 1968;Yamashita et al., 1970),Bacillus subtilis ( Scannell et al., 1970),Pseudomonas aeruginosa (Yamashita et al., 1970), Lactobacillus plantarum ( Shoulders et al., 1968) and Sacchuromyces cerevisiae ( Genghof,
1970) Streptomyces, several strains ( Yamashita et al., 1970; Scannell et al., 1970;Fickenscher et al., 1971) Synthesis: ( Snow et al., 1968;Shoulders et al., 1968;Ressler, 1972) Mechanism: Feedback inhibition of 3-deoxy-~-arabinoheptulosonSource:
Review:
ate-7-phosphate synthetase and prephenate dehydratase (Fickenscher et al., 1971; Fickenscher and Zahner, 1971) (Nass et al., 1971)
~-2,3-Epoxy-4-oxohexahydrophenylalanine ( anticapsin ) 0 I1 H2N- C H - C - O H I
1XXI1
Reversant:
Glutamine, glucosamine, and N-acetylglucosamine in hyaluronic acid capsule of Streptococcus p yogenes ( Whitney and Funderburk, 1970) Source: Streptomyces griseoplanus (Shah et al., 1970) Structure: (Neuss et al., 1970) Mechanism: Inhibition of ~-glutamine-~-fructose-6-phosphate amidotransferase ( Whitney and Funderburk, 1970)
ANTIMETABOLITES FROM MICROORGANISMS
33
~-Alanyl-~-2,3-epoxy-4-oxohexahydrophenylalanine ( bacilysin, tetanine) 0 0 I1 I1 H2N-CH-C-NH-CH-C-OH I I CH3 y 2
Reversant: Source: Structure: Mechanism: Bacillin Reversant: Source: Structure:
Not demonstrated; presumed to be the same as for anticapsin; active toward Staphylococcus aureus Bacillus subtilis (Rogers et al., 1965; Florey et al., 1949); B . pumilus (Kamiliski and Sokolowska, 1973) (Walker and Abrahams, 1970) Inhibition of bacterial cell wall biosynthesis ( Chmara and Borowski, 1973) N-Acetylglucosamine in gram-positive organisms (Walton and Rickes, 1962) Bacillus szibtilis ( Foster and Woodruff, 1946) Unknown; may be related to or identical with bacilysin
5. D-Amino Acids ~-4-Amino-3-isoxazolidinone( cycloserine ) H2N
CH-C
//”
c m 1
Reversant:
Source:
Synthesis: Review:
and L-Alanine in Bacillus species (Harned et al., 1955; Harris et al., 1955; Shull and Sardinas, 1955; Kurosawa, 1952) Several specics of Streptomyces ( Harned et al., 1955; Harris et al., 1955;Shull and Sardinas, 1955;Kurosawa, 1952) (Stammer et al., 1955) (Neuhaus, 1967;Stammer, 1971) D-
DAVID L. PRUESS AND JAMES P. SCANNELL
34
0-Carbamyl-D-serine 0
NH2 0
II I I1 H2N-C-0-CH2-CH-C-OH
(xxm) Reversant: Source: Synthesis: Review:
D-Alanine and L-alanine in Bacillus species and Streptococcus faecalis (Hagemann et al., 1955) Streptomyces polychromogenes ( Hagemann et al., 1955) (Skinner et al., 1955) (Neuhaus, 1967)
6. Diazo Amino Acids
0-Diazoacetyl-L-serine ( azaserine )
(xxp)
Reversant:.
L-Glutamine, adenine, aromatic and basic amino acids (Pittillo and Hunt, 1967) Streptomyces fragilis (Stock et al., 1954; Anderson Source: et al., 1956) Synthesis: (Moore et al., 1954) Review: (Pittillo and Hunt, 1967) Mechanism: Inhibition of de nooo purine biosynthesis (Pittillo and Hunt, 1967)
6-Diazo-5-oxo-~-norleucine ( DON ) 0 NH2 0 @ I1 I II N = N =C H - C - C H 2 -.C H 2 - C H - C -OH
Q
L-Glutamine, purine bases and nucleosides, various amino acids in several organisms (reviewed by Pittillo and Hunt, 1967) Streptomyces species (Ehrlich et al., 1956) Source: ( DeWald and Moore, 1958) Synthesis: Mechanism : Inhibition of de nooo purine biosynthesis ( reviewed by Pittillo and Hunt, 1967) Review: (Pittillo and Hunt, 1967)
Reversant:
L-Alanyl- ( 6-diazo-5-0x0) -L-norleucyl-( 6-diazo-5-0x0) +norleucine
( alazopeptin )
ANTIMETABOLITES FROM MICROORGANISMS
0
0 I1 H2 N -CH -C-NH
I
0 II
II
- CH - C - N H
-CH- C-OH I
I
CH2 I CH2 I
CH3
35
CH2 I CH2
I
c=o
c=o
I CH II
I CH II @N II 8 N
BN II
BN
(rn) Reversant: Source: Structure:
Presumed to be same as for DON Streptomyces griseoplanus ( DeVoe et al., 1957) (Patterson et al., 1966)
N *-Acetyl-( 6-diazo-5-0x0) +norleucine ( N-acetyl-DON, duazomycin A ) 0 I1
H N - C -CH3
1 :
8
@ 0 II N=N=CH-C-CH2-CH2-CH-C-OH
(m) Reversant: Source: Structure:
Presumed to be same as for DON Streptomyces ambofaciens (Rao et al., 1960) (Rao, 1962)
5-~-Glutamyl( 6-diazo-5-0x0) -L-norleucyl-( 6-diazo-5-0x0) -L-norleucine ( duazomycin B) 0
0 0 II II II H2N-CH-CH~-CH2-C-NH-CH-C-NH-CH-C-OH I I I C-OH CH2 CH2 I1 I I 0
CH2
1
c=o I
CH I1
@N II ON
CH2 I
c=o
I CH II [email protected] II
NQ
36
DAVID L. PRUESS AND JAMES P. SCANNELL
Reversant: Source: Structure:
L-Glutamine in lymphocytes (Hersh and Brown, 1971); presumed to be same as for DON Streptomyces ambofaciens (Rao et al., 1960) (Rao, 1963)
Duazomycin C Reversant: Source: Structure:
Presumed to be same as for DON Streptomyces ambofaciens (Rao et al., 1960) Unknown
7. Di- and Tripeptides 2-Amino-4-( 3-hydroxy-2-0~0-3-azetidinyl) butanoyl-L-threonine (wild fire toxin, tabtoxin) 0 II
0 II
HZN-CH-C-NH-CH -C-OH I I CH-OH CHZ I I CHZ CH3 I +o HO-C-C I I HzC-NH
(xxx) Reversant:
L-Glutamine in tobacco leaves (Sinden and Durbin, 1968),L-methionine in Chlorella vulgaris (Braun, 1950) Source: Pseudomonus tabaci (Johnson and Murwin, 1925; Woolley et al., 1965);Pseudomom sp. (Taylor et al., 1972) Structure: Revised (Stewart, 1971) Mechanism: Glutamine synthetase inhibition ( Sinden and Durbin, 1968)
2-Amino-4-methylphosphinobutanoyl-~-alanyl-~-alanine ( phosphinothricyl-alanyl-alanine )
CH2 I $.HZ
no-
P=O I CH3
(XXXI)
Reversant:
L-Glutamine in Escherichia coli, Bacillus subtilis, Botrytis cinerea, Pseudornonas saccharophilia ( Bayer et al., 1972)
ANTIMETABOLITES FROM MICROORGANISMS
37
Streptomyces viridochromogenes ( Bayer et al., 1972) Source: Mechanism: Inhibition of glutamine synthetase after enzymatic degradation to phosphinothricin (Bayer et aZ., 1972) Fumarylcarboxamido-~-2,3diaminopropionyl-~-alanine O
O H I l l
H2N - C
- C=C-
I H
I1
C -NH-CHz
NH2 0 I I1
CH3 0 I II
- C H - C - N H -CH-C-OH
(XXXLII
Reversant: Source:
D-Ghcosamine and N-acetylglucosamine in gram-negative microorganisms (Molloy et aZ., 1972) Streptomyces collinus, Lindenbein ( MoIloy et al., 1972)
( S,S ) -N-Methylisoleucyl-~-2-amino-4( 4-amino-2,5-cyclohexadien1-yl ) butanoic acid (stravidin, MSD 235 S,)
Reversant: Source: Structure:
Biotin in Escherichia coli and other gram-negative microorganisms (Stapley et al., 1964) Streptomyces avidinii, S. Lavendulae (Stapley et al., 1964) (Baggaley et al., 1969)
L-( N5-Phosphono) methionine-( S ) -sulfoximinyl-L-alanyl-L-alanine ) 0
0
0
I1 I1 11 H2N-CH-C-NH-CH-C-NH-CH-C-OH I I I CH3 CH3
CH3
OH
DAVID L. PRUESS AND JAMES P. SCANNELL
38 Reversant: Source: Structure:
L-Glutamine in Serratia sp. and Bacillus subtilis (Pruess et al., 1973) Streptomyces sp. (Pruess et al., 1973) (Scannell et al., 1972c)
B. NUCLEOSIDES AND RELATED COMPOUNDS 1. Pyrimidine Type SAzacytidine
d
HOCH2
on
OH
Cytidine, uridine in Escherichia colt (Cihik and Sorm, 1965; Hanka et al., 1967) Streptooerticillium kzahkanus ( Bergy and Herr, 1967) Source: Synthesis: (Piskala and Sorm, 1964) Mechanism: ( Reviewed by Suhadolnik, 1970) (Suhadolnik, 1970) Review :
Reversant:
Emimycin ( 1,2-dihydro-2-oxopyrazine-4-oxide)
Reversant: Source: Structure: Synthesis:
Uracil in Escherichia coli ( DeZeeuw and Tynan, 1969) Streptomyces sp. (Terao et at., 1960) (Terao, 1963); confirmation by X-ray analysis of 2-bromo derivative (Tamura, 1963) Improved method (Bobek and Block, 1972)
ANTIMETABOLITES FROM MICROORGANISMS
39
Oxazinomycin ( minimycin )
HoH2w OH
Reversant: Source: Structure:
OH
Deoxycytidine in Staphylococcus aureus ( Haneishi et al., 1970) Streptomyces tunesashinensis ( Haneishi et al., 1971); Streptomyces hygroscopicus ( Kusakabe et al., 1972) (Haneishi et al., 1971; Sasaki et al., 1972)
Polyoxins
II C-HN-CH I H 2 N-CH I HC- R
I OH OH 0 I II CHz-O-C-NHz
H0 - C H
Peptides in Pellicularia sasakii ( Mitani and Inoue, 1968) Source: Streptomyces cacaoi var. asoenis (Suzuki et al., 1965) Structure: (Isono et al., 1969) Mechanism: Inhibitor of chitin synthetase; it is competitive with UDP-N-acetylglucosamine ( Endo et al., 1970) Review: ( Suhadolnik, 1970) Reversant:
40
DAVID L. PRUESS AND J A M E S P. SCANNELL
Pyrazomycin 0
HoH2H OH OH
Uridine and uridylic acid in vaccinia virus (Streightoff et al., 1969) Streptomyces candidus ( Streightoff et at., 1969) Source: Structure: (Gerzon et al., 1969) Mechanism: 5’-Phosphate ester is inhibitor of orotidylic decarboxylase (Gerzon et al., 1971) Review : (Gerzon et al., 1971) Reversant:
Showdomycin
OH OH
Ribo- and deoxyribonucleosides and sulfhydryl compounds (Nishimura and Komatsu, 1968) Streptomyces showdoensis ( Nishimura et al., 1964) Source: (Nakagawa et al., 1967; Darnall et al., 1967) Structure: Mechanism: Reversal by nucleosides shown to be due to use of same transport system ( Roy-Burman and Visser, 1972); inhibits formation of deoxyribonucleotides from
Reversant:
ANTIMETABOLITES FROM MICROORGANISMS
Review:
41
corresponding ribonucleotides ( Komatsu and Tanaka, 1971) (Suhadolnik, 1970; Gerzon et al., 1971)
2. Purine Type Aristeromycin
Reversant:
Source: Structure: Synthesis: Review:
Adenosine, and to a lesser extent by adenine, deoxyadenosine and inosine in Xanthomonus oryxae (Kusaka, 1971) Streptomyces citrocolor (Kusaka et al., 1968) Absolute configuration determined by X-ray analysis (Kishi et al., 1972) (Shealy and Clayton, 1966,1969) ( Suhadolnik, 1970)
Decoyinine ( angustmycin A )
OH OH
[mu) Reversant:
Guanosine, guanine in Bacillus subtilis and Mycobacterium sp. (Tanaka et al., 1960) and Streptococcus faecalis (Bloch and Nichol, 1964)
42
DAVID L. PRUESS A N D JAMES P. SCANNELL
Source: Structure: Review:
Streptomyces hygroscopicus var. angustmyceticus (Yuntsen et al., 1956) and Streptomyces hygroscopicus var. decoyicus (Hoeksema et al., 1964) (Hoeksema et al., 1964) (Suhadolnik, 1970)
Cordycepin ( 3'-deoxyadenosine)
Reversant: Source: Structure: Review:
Guanosine, adenosine in Bacillus subtilis ( Rottman and Guarino, 1964) Cordyceps militans (Cunningham et al., 1951); Aspergillus nidulans (Kaczka et al., 1964) (Kaczka et al., 1964) ( Suhadolnik, 1970)
Fonnycin B OH
ANTIMETABOLITES FROM MICROORGANISMS
Reversant: Source: Structure:
P3 Adenosine, inosine, thymidine, and uridine in Xanthomonm oryzae (Hori et al., 1968) Nocardia interforma ( Koyama and Umezawa, 1965) (Koyama et al., 1966)
Psicofuranine ( angustmycin C)
OH OH
(XLTI
Reversant:
Source:
Structure: Review:
Guanine and guanosine in Staphylococcus aureus (Hanka, 1960); Escherichia coli (Slechta, 1960); Bacillus subtilis (Tanaka et al., 1960) Streptomyces hygroscopicus var. angustmyceticus (Yuntsen et al., 1956); Streptomyces hygroscopicus var. decoyicus (Eble et al., 1959) ( Yuntsen, 1958; Schroeder and Hocksema, 1959) ( Suhadolnik, 1970)
Thioguanine SH
Reversant: Source: Synthesis : Review:
Adenine or guanine in Escherichia coli (Scannell et al., 1971b) Pseudomoms sp. GH (Scannell et al., 1971b) ( Elion and Hitchings, 1955) (Hitchings and Elion, 1963)
44
DAVID L. PRUESS AND JAMES P. SCANNELL
Toyocamycin
OH OH
Reversant: Adenosine and inosine ( D. L. Pruess, unpublished observation ) Streptomyces toyocaensis ( Nishimura et al., 1956) Source: Structure: (Ohkuma, 1961) Review: (Suhadolnik, 1970) Tubercidin
Reversant: Source: Structure: Review:
Adenosine, 2’-deoxyadenosine, uridine, cysteine, pyruvate, ribose-5’-phosphate in Streptococcus faecalis (Bloch et al., 1967) Streptomyces tubercidicus ( Anzai et al., 1957) (Suzuki and Marumo, 1961) ( Suhadolnik, 1970) C. VITAMINANTIMETABOLITES
1. Biotin Reversed Actithiazic acid
ANTIMETABOLITES FROM MICROORGANISMS
45
0
LNH (XLIX)
Reversant:
Source:
Structure : Synthesis: Review:
Biotin in Mycobacteria (Grundy et al., 1952; Umezawa et al., 1953; Kawashima et al., 1953; Pittillo and Foster, 1954); Aerobacter aerogenes (PittilIo and Foster, 1954); Bacillus subtilis (Stim et al., 1959) Streptomyces virginiae (Grundy et al., 1952); S. acidomyceticus ( Miyake et al., 1953); S. lavendulae (Tejera et al., 1952); S . cinnamonensis (Maeda et al., 1952) (McLamore et al., 1953; Schenck and DeRose, 1952; Miyake et al., 1953) (Clark and Schenck, 1952; McLamore et al., 1953) (Caltrider, 1967)
a-Dehydrobiotin
Reversant: Source: Synthesis:
Biotin in Escherichia coli (Hanka et al., 1966) Streptomyces lydicus (Hanka et al., 1966) (Field et al., 1970)
,-Methylbiotin U
(LI)
Reversant: Source: Synthesis:
Biotin in Mycobacteria and presumably Escherichia coli and Bacillus subtilis (Hanka et al., 1972) Streptomyces lydicus (Martin et al., 1971) (Martin et al., 1971)
46
DAVID L. FRUESS AND JAMES P. GCANNELL
a-Methyldethiobiotin
Reversant: Source:
Biotin in Mycobacteria and presumably Escherichia coli and Bacillus subtilis (Hanka et al., 1972) Streptomyces lydicus (Martin et al., 1971)
2. Ergocalciferol Reversed Cerulenin H
H
Ergocalciferol in Candida stellaloidea and Saccharomyces cerevisiae (Nomura et al., 1972b) Source: Cephalosporium caerulens ( Sano et al., 1967) Structure: (Omura et al., 1967) Mechanism: Inhibits p-ketoacyl-acyl carrier protein synthetase thus blocking fatty acid synthesis (Nomura et al., 1972a; DAgnolo et al., 1973)
Reversant:
3. Nicotinic Acid or Nicotinamide Reversed Albocycline
ANTIMETABOLITES FROM MICROORGANISMS
Reversant: Source:
Structure:
47
Nicotinic acid, nicotinamide, and quinolinic acid in Bacillus subtilis (Reusser, 1969) Streptornyces bruneogriseus ( Nagahama et al., 1967); S . roseocinereus and S . roseochromogenes (Furumai et al., 1968); S . maizeus (Reusser, 1969) (Nagahama et al., 1971)
Melinacidins Reversant: Source: Structure:
Nicotinic acid and nicotinamide in Bacillus subtilis ( Reusser, 1968) Acrostalagmus cinnabarinus var. melinucidinus ( Argoudelis and Reusser, 1971) Family of 3,6-epidithiadiketopiperazines, exact structure unknown ( Argoudelis, 1972)
Streptozotocin HOH2C
H O & - - -H O$ NH
I c=o I
N- NO
I CH3 (LP)
Nicotinamide reverses diabetogenic action in mouse and rat (Schein et al., 1967; Schein and Loftus, 1968) Source: Streptomyces achromogenes (Vavra et al., 1960; Herr et al., 1960 Structure: (Herr et al., 1967) Synthesis: (Herr et al., 1967; Hessler and Jahnke, 1970) Mechanism: Presumed to block NAD biosynthesis in pancreas (Schein and Loftus, 1968) Review: (Rudas, 1972) Reversant:
4. Thiamine Reversed Bacimethrin
48
DAVID L. PRUESS AND JAMES P. SCANNELL
Reversant: Source: Structure:
Thiamine and pyridoxine in Bacillus subtilis, Escherichia coli, and several yeasts ( Tanaka et al., 1961, 1962a) Bacillus mgatherium ( Tanaka et al., 1961,1962a) ( Tanaka et al., 1962b)
5. Vitamin B I 2 Reversed Descoboltocorrinoids Reversant:
Source:
Structure:
Vitamin B, in vitamin B1,-requiring strains of Escherichia coli, Euglena gracilis, and Arthrobacter duodecadis (Perlman and Toohey, 1968) Chromatiurn, Rhodospirillum rubrum, and R. palustris (Toohey, 1965, 1966); Streptomyces olivaceus ( Sato et al., 1970) Exact structures not known. However, from Chromatiurn phenylhydrogenobamide and hydrogenobyric acid were isolated (Koppenhagen and Pfiffner, 1970), and from Chromatiurn grown in medium containing 5,6-dimethylbenzimidazole a-( 5,6-dimethylbenzimidazolyl) hydrogenobamide was isolated ( Koppenhagen and Pfiffner, 1971)
D. MISCELLANEOUS ANTIMETABOLITES Borrelidin CH3 I /CH-CH2
CH3 I -CH-CH2-
C H j OH I I
- -
CH CH CH CH2 C ‘ H ‘CH CH3 I OH
/
-
0 II
- c
P
0
/CH-CH-CH-C-OH I I C = CH - CH=CH-CH2 CH2 CH2 I C I N CH2
II
I
(LVlIl
L-Threonine, L-homoserine in Bacillus subtilis ( Hutter et al., 1966) Streptomyces rochei (Berger et al., 1949) Source: Structure: ( Keller-Schierlein, 1967) Mechanism: Inhibition of threonyl-tRNA synthetase ( Nass et al., 1969) Review: (Nass et al., 1971)
Reversant:
ANTIMETABOLITES FROM MlCROORGANISMS
49
Griseofulvin
CH30 CH3
(LYD)
Reversant:
Adenylic acid, guanylic acid in Microsporum canis ( McNall, 1960) Penicillium griseofuluum and other species ( Oxford et al., 1939) Structure and synthesis (Grove, 1963); mechanism of action (Huber, 1967)
Source: Review: Lycomarasmin
0
0 II HO-C-CHZ -CH-NH-CH~-CH-NH-CHZ-C-NH~ I I II
c =o
c=o I OH
I
OH
(LEI
Reversant:
Source: Structure:
Serylglycylglutamic acid and other peptide growth factors in tomato plant and Lactobacillus casei ( Woolley, 1946) Fusarium oxysporum, F. lycopusici ( Plattner and Clauson-Kaas, 1945) Revised (Hardegger et al., 1963)
Mycophenolic acid
Reversant: Source: Structure:
Guanosine, guanylic acid, deoxyguanylic acid in viral tissue culture test (Cline et al., 1969) Penicu’llium brevi-compactum ( Clutterbuck et al., 1932); P. stoloniferum (Williams et al., 1968) (Logan and Newbold, 1957)
DAVID L. PRUESS AND JAMES P. SCANNELL
50
Mechanism: Inhibition of guanylic synthetase ( Estermann and Sweeney, 1970) and inosine dehydrogenase ( Franklin and Cook, 1969) Pentalenolactone ( PA-132) y 3
(LXI)
Reversant: Adenine (Takeshima et al., 1969) Source: Streptomyces species (Koe et al., 1957; English et al., 1957) Structure: (Martin et al., 1970; Takeuchi et al., 1969) Tunicamy cin Reversant: Source: Structure:
N-Acetylglucosamine and related compounds ( Takatsuki and Tamura, 1971) Streptomyces lysosuperificus (Takatsuki et al., 1971) Unknown
Histidomycin Reversant: Source: Structure:
VI.
Mixture of amino acids in Escherichia coli (Stapley et al., 1967) Nocardia histidam (Stapley et al., 1967) Unknown, a complex of antimetabolites; one component, C25H3BNeOloC1, contains histidine ( Kaczka et al., 1967) Incidence of Antimetabolite Production
Hanka (1938) observed that 23% of the fermentation broths tested showed antimetabolite activity against E. coli and B . subtilis. Korobkova et al. (1970) reported that of 756 actinomycetes which showed no zones of inhibition on complex media, 25 produced antimetabolite activity in a minimal medium which was reversed by leucine. Gause et al. (1972) tested 2160 actinomycete cultures by an agar block method against two yeasts and a mutant derived from each yeast. The mutants, which had impaired respiration and enhanced glycolysis were found to be much more sensitive to antimetabolites than their parents; detection of antimetabolites was increased from 0.04%to 1.64: for Candida utilis and 0.0% to 5.00%
ANTIMETABOLITES FROM MICROORGANISMS
51
for Torulopsis globosa. In our experience, 5% of the actinomycetes screened produced detectable antimetabolite activity against at least one of five test organisms. As is the case for antibiotics, actinomycetes give a higher incidence of antimetabolite production compared to other orders of microorganisms. Antimetabolites which are reversed by aliphatic amino acids, diamino acids, or glutamine appear to be much more prevalent than those reversed by heterocyclic amino acids. Antimetabolites for proline and tryptophan have not as yet been reported as microbial products. A wide variety of nucleoside antimetabolites is now known with structural variations in either the sugar or the base moieties. With the exception of the biotin analogs, few antimetabolites structurally related to vitamins have been discovered. It is clear that the search for antimetabolites from natural sources has yielded compounds which had not been previously synthesized in spite of the intensive efforts of organic chemists in this field. Of the compounds listed in Section V, only 13 had been previously synthesized. Some of the novel compounds could well have been synthesized, but in other cases problems of stability or the multiplicity of optical or geometrical isomers would have presented rather difficult synthetic problems. The most interesting cases, however, are those in which the structural analogy between antimetabolite and reversant is so remote that the compounds would never have been the prior object of a deliberate synthetic effort, as for example, borrelidin (LVII) and mycophenolic acid ( LX) which are reversed by threonine and guanosine, respectively. Of particular interest to peptide chemists is the finding that the two phosphorus-containing tripeptide antimetabolites ( XXXI and XXXIV) are much more potent microbial growth inhibitors than are the constituent unusual amino acids (Bayer et al., 1972; Pruess et al., 1973), even though the amino acids are better enzyme inhibitors (Bayer et al., 1972; A. Meister, personal communication), A similar effect has been demonstrated ( Smith and Dunn, 1970) with the synthetic phenylalanine analog, P-2-thienylalanine. This probably indicates that the peptide is more readily taken into the cell and there enzymatically hydrolyzed to the active enzyme inhibitor. VII.
Chemotherapeutic Applications
The chemotherapeutic uses of antimetabolites have been the subject of a vast review literature. Advocates of a “rational approach to chemotherapy” are particularly interested in antimetabolites, since the comparative biochemistry of metabolic pathways is, in general, well understood and an elaborate theory of enzyme kinetics and inhibition has been
52
DAVID L. PRUESS AND JAMES P . SCANNELL
developed. Since essential amino acids and vitamins are produced by microorganisms, not by mammals, a favorable selective activity would be expected if biosynthesis were blocked by an antimetabolite. Accordingly, one might expect that antimetabolites for amino acids and vitamins would be an especially promising group of compounds for chemotherapeutic application. Nevertheless, only a few vitamin or amino acid antimetabolites are useful drugs, perhaps the most notable examples being D-cycloserine and the sulfonamides. An analysis of the consequences of antimetabolite action in terms of the mechanisms discussed in Section IV partially explains this lack of success. If the mechanism of action of an antimetabolite is simple isosteric competitive inhibition of an enzyme reaction, the substrate concentration might be expected to increase until it relieved the block. It has been suggested ( Baker, 1967) that noncompetitive irreversible inhibitors would not be subject to this drawback. A more serious problem, however, is reversal by the product of the inhibited reaction or metabolic pathway. Except for certain topical infections, one would expect the product to be exogenously supplied to the infecting organism; consequently, regardless of how effectively biosynthesis was shut off, the viability of the organism would be unaffected. If the mechanism of antimetabolite inhibition is at the level of control, either by repression of enzyme synthesis or by allosteric pseudo-feedback inhibition of enzyme action, the outlook is even less promising. Not only will an exogenous supply of the end product of the blocked metabolic sequence relieve growth inhibition, but the end product itself will also shut off its own biosynthesis, thus making the action of the antimetabolite redundant. Transport inhibition is only likely to be effective against organisms lacking the relevant biosynthetic pathway and thus dependent on an exogenous supply of an essential metabolite. In organisms where exogenous and endogenous supplies of an essential metabolite are available, the antimetabolite would either have to serve double duty in inhibiting both uptake and biosynthesis, or two antimetabolites would have to be administered simultaneously in order to cut off both sources of supply. This would present an additional problem since the uptake inhibitor would very likely also inhibit the transport of the biosynthesis inhibitor. Perhaps this difficulty could be obviated in the case of amino acid antimetabolites by incorporation of the biosynthesis inhibitor into a di- or tripeptide. There are cases in which inhibition of biosynthesis is effective by itself since an exogenous supply of the essential metabolite is either not available to the parasite because it is not produced by the host organism, or not taken up by the parasite because a transport capability
ANTIMETABOLlTES FROM MICROORGANISMS
53
is lacking. In many of the cases where antimetabolites have proved to be therapeutically useful, lack of access to an exogenous supply of product reversant appears to be important. For instance, the effectiveness of D-cycloserine and the sulfonamides may be ascribed to the unavailability of an exogenous supply of the products of the inhibited reactions, D-alanine and D-alanyl-walanine in the case of D-cycloserine ( Neuhaus, 1967) and folic acid, which apparently does not enter the bacterial cells effectively (Woods et aE., 196l), in the case of the sulfonamides. Thus, in the search for new therapeutically useful antimetabolites, special attention should be directed to agents that block the biosynthesis of metabolites which are not exogenously available to the parasite.
NOTEADDEDIN
PROOF
The following antimetabolites have appeared in the literature during
1973: ~-N~-Hydroxyarginine NH OH 11 I HzN-C-N-CH2-CHz
Reversant:
Source:
NH2 0 I II -CHz-CH-C-OH
L-Arginine (noncompetitive), L-citrulline ( competitive) in Escherichia coli (Fischer et al., 1973; Maehr et al., 1973) Bacillus species XB-13248 (Maehr et al., 1973); Mannizzia gypsea (Fischer et al., 1973)
0-[ 1-Norvalyld]-’isourea NH2 0 I II -CH2 -CH2 -CH-C-OH
NH II HzN-C-O-CHz
Reversant: L-Arginine, L-ornithine, and L-citrulline in Escherichia coli and Bacillus subtilis (Konig et al., 1973a) Unidentified gram-negative bacterium, TU 222 ( Konig Source: et al., 1973a)
~S,5S-~-Amino-3-chloro-2-isoxazoline-5-acetic acid ’N II
CI -C-CH2
Reversant: Source: Structure:
NH2 0 I I1 ‘CH-CH-C-OH
0
I
Histidine in Bacillus subtilis and Escherichia coli (Hanka and Dietz, 1973; Hanka et al., 1973) Streptomyces wiceus (Hanka and Dietz, 1973) (Martin et al., 1973)
54
DAVID L. PRUESS A N D J A M E S
P. SCANNELL
L-Arginyl-D-do-threonyl-L-phenylalanine y 3
0 HO-CH
II H2N-CH-C -NH I y 2
CH2 I CH2 1 NH I C=NH
I
0
0 I1
II
- CH-C-NH - CH - C-OH I
FH2
ACH
CH
I
CH
II
I
Reversant: L-Histidine, L-threonine in Paecilomyces varioti ( Konig et al., 1973b) Source: Keratinophyton terrum (Konig et al., 1973b) Synthesis: L-Arginyl-D,L-albthreonyl-L-phenylalanine ( Konig et al., 1973b)
A9145 Reversant: Source: Structure:
Undefined components of Sabouraud’s medium in Candida albicans (Gordee and Butler, 1973) Streptomyces griseo2us (Boeck et al., 1973) Unknown; the compound is weakly basic, has a molecular weight of 510, and contains adenine (Hamill and Hoehn, 1973) ACKNOWLEDGMENTS
We wish to acknowledge the helpful discussion and critical review provided by Drs. T. C. Demny and P. A. Miller. REFERENCES Anderson, L. E., Ehrlich, J., Sun, S. H., and Burkholder, P. R. (1956). Antibiot. Chemother. (Washington, D.C.)6, 100. Anderson, P. M., and Meister, A. (1966). Biochemistry 5, 3164. Anzai, K., Nakamura, G., and Suzuki, S. (1957). J. Antibiot. 10, 201. Argoudelis, A. D. (1972). J. Antibiot. 25, 171. ArgoudeIis, A. D., and Reusser, F. (1971). J. Antibiot. 24, 383. Argoudelis, A. D., Herr, R. R., Mason, D. J., Pyke, T. R., and Zierserl, F. J. ( 1967). Biochemistry 6, 165. Baggaley, K. H., Blessington, B., Falshaw, C . P., OW, W. D., Chaiet, L., and Wolf,F. J. (1969). Chem. Cmmun. p. 101. Baker, B. R. ( 1967). “Design of Active-Site-Directed Irreversible Enzyme Inhibitors.” Wiley, New York. Balema, M., KelIer-Schierlein, W., Martius, c., Wolf, H., and Zahner, H. (1969). Arch. Mikrobiol. 65, 303.
ANTIMETABOLITES FROM MICROORGANISMS
55
Bayer, E., Gugel, K. H., Hagele, K., Hagenmaier, H., Jessipow, S., Konig, W. A., and Ziihner, H. ( 1972). Helv. Chim. Acta 55, 224. Berger, J., Jampolsky, L. M., and Goldberg, M. W. (1949). Arch. Biochem. 22, 476. Berg, M. E., and Herr, R. R. (,1967).Antimicrob. Ag. Chemother. p. 625. Bloch, A., and Nichol, C. A. (1964). Fed. Proc., Fed. Amer. SOC. Exp. B i d . 23, 324. Bloch, A., Leonard, R. J., and Nichol, C. A. (1967). Biochim. Biophys. Acta 138, 10. Bobek, M., and Bloch, A. (1972). J . Med. Chem. 15, 164. Boeck, L. D., Clem, G. M., Wilson, M. M., and Westhead, J. E. (1973). Antimicrob. Ag. & Chemother. 3, 49. Braun, A. C. (1950). Proc. Nat. Acud. Sci. U.S. 36, 423. Brian, P. W., Elson, G. W., Hemming, H. G., and Radley, M. E. (1965). Nature (London) 207, 998. Burrows, B. F., and Turner, W. B. (1966). J. Chem. Soc., C . p. 255. Caltrider, P. G. (1967). In “Antibiotics” (D. Gottlieb and P. D. Shaw, eds.), Vol. 1, p. 666. Springer-Verlag, Berlin and New York. Celmer, W. D. ( 1964). Belgian Patent 644,682; Chern. Abstr. 63,9836a ( 1965). Chmara, H., a?d Borowski, E. (1973). Biochem. Biophys. Res. Commun. 52, 1381. CihBk, A., and Sorm, F. (1965). Collect. Czech. Chem. Commun. 30, 2091. Clark, R. K., Jr., and Schenck, J. R. ( 1952). Arch. Biochem. Biophys. 40, 270. Cline, J. C., Nelson, J. D., Gerzon, K., William, R. H., and DeLong, D. C. (1969). Appl. Microbiol. 18, 14. Clutterbuck, P. W., Oxford, A. E., Raistrick, H., and Smith, G. (1932). Biochem. J. 26, 1441. Coronelli, C., Pasqualucci, G. R., Tamoni, G., and Gallo, G. G. ( 1966). Famaco, Ed. Sci. 21, 269. Cunningham, K. G., Hutchinson, S. A., Manson, W., and Spring, F. S. (1951). J . Chem. Soc., London p. 2299. Cuthbertson, W. F. S., Gregory, J., OSullivan, P., and Pegler, H. F. (1956). Biochem. 1. 62, 15p. D’Agnolo, G., Rosenfeld, I. S., Awaya, J., Omura, S., and Vagelos, P. R. (1973). Biochim. Biophys. A d a 326, 155. Darnall, K., Townsend, L., and Robins, R. (1967). Proc. Nut. Acad. Sci. US. 57, 548. DeVoe, S. E., Rigler, W. E., Shay, A. J., Martin, J. H., Boyd, T. C., Backus, E. J., Mowat, J. H., and Bohonos, N. (1957). Antibiot. Annu. p. 730. DeWald, H. A, and Moore, A. M. (1958). J. Amer. Chem. SOC. 80, 3941. DeZeeuw, J. R., and Tynan, E. (1969). J . Antibiot. 22, 386. Dulaney, E. L., and Gray, R. A. (1962). Mycologia 54, 476. Dyer, H. M. (1938). J. Biol. Chem. 124, 519. Eble, T. E., Hoeksema, H., Boyack, G., and Savage, G. M. ( 1959). Antibiot. Chemother. (Washington, D.C.) 9, 419. Edelson, J., Fissekis, J. D., Skinner, C. G., and Shive, W. (1958). 1. Amer. Chem. SOC. 80, 2698. Elulich, J., Coffey, G. L., Fisher, M. W., Hillegass, A. B., Kohberger, D. L., Machamer, H. E., Rightsel, W. A., and Roegner, T. R. (1956). Antibiot. Chemother. (Washington, D.C.) 8, 487. Elion, G. B., and Hitchings, G. H. (1955). J . Amer. Chem. SOC. 77, 1676. Emery, T. (1971). Aduan. Enzymol. 35, 135. Endo, A., Kakiki, K., and Misato, T. (1970). J . BacterioZ. 104, 189. English, A. R., McBride, T. J., and Lynch, J. E. (1957). Antibiot. Annu. p. 676. Esterman, M. A., and Sweeney, M. J. (1970). Proc. Amer, Ass. Cancer Res. 11, 91.
56
DAVID L. PRUESS AND JAMES P. SCANNELL
Fickenscher, U., and Zahner, H. (1971).Arch. Mikrobiol. 76, 28. Fickenscher, U., Keller-Schierlein, W., and Zahner, H. ( 1971). Arch. Mikrobiol. 75, 346. Field, G . F., Zally, W. J., and Sternbach, L. H. (1970).J. Amer. Chem. SOC. 92, 3520. Fischer, B., Keller-Schierlein, W., Kneifel, H., Konig, W. A, Loeffler, W., Miiller, A., Muntwyler, R.,and Zahner, H. (1973).Arch. Mikrobiol. 91, 203. Fisher, J. F., and Mallette, M. F. (1961).J. Gen. Physiol. 45, 1. Florey, H. W., Chain, E., Heatly, N. G., Jennings, M. A., Sanders, A. G., Abraham, E. P., and Florey, M. E. (1949).“Antibiotics,” Vol. I, p. 457. Oxford Univ. Press, London and New York. Foster, J. W., and Pittillo, R. F. (1953a).J. Bacteriol. 65, 361. Foster, J. W., and Pittillo, R. F. (195313).J. Bacteriot. 66, 478. Foster, J. W., and Woodruff, H. B. (1946).J. Bacteriol. 51, 363. Fowden, L., Lewis, D., and Tristram, H. (1967).Aduan. Enzymol. 29, 89. Franklin, T.J., and Cook, J. M. (1969).Biochem. J. 113, 515. Fuerst, R. (1964). Proc. Int. Congr. Chemother., 3rd, 1963 Vol. I, p. 983. Fuerst, R., and Skellenger, W. M. (1958).Antibiot. Chemother. (Washington, D.C.) 8, 76. Furumai, T., Nagahama, N., and Okuda, T. (1968).J. Antibiot. 21, 85. Gale, G. R., and Atkins, L. M. ( 1968). Proc. SOC. Exp. Biol. Med. 129, 842. Gale, G. R., Ostrander, W. E., and Atkin, L. M. (1968).Biochem. Pharmacol. 17, 1823. Gause, G. F., Laiko, A. V., Kusovkova, L. I., and Selesneva, T.I. ( 1972).Experientia 28, 608. Genghof, D. S. (1970).Can. J . Microbiol. 16,545. Gershon, H., Meek, J. S., and Dittmer, K. (1949).1. Amer. Chem. SOC. 71, 3573. Gershon, H., Shapira, J., Meek, J. S., and Dittmer, K. (1954). J. Amer. Chem. SOC. 76,3484. Gerzon, K., Williams, R. H., Hoehn, M., Gorman, M., and DeLong, D. C. (1969). Proc. Int. Congr. HeterocycE. Chern., 2 4 , 1969 Abstract 30-C. Gerzon, K.,DeLong, D. C., and Cline, J. C. (1971).Pure Appl. Chem. 28, 489. Giovanelli, J., Owens, L. D., and Mudd, S. H. (1971). Biochim. Biophys. Acta 227, 671. Gitterman, C. O., Dulaney, E. L., Kaczka, E. A., Hendlin, D., and Woodruff, H. B. (1962).Proc. SOC. E x p . Biol. Med. 109, 852. Gordee, R. S., and Butler, T.F. (1973).J. Antibiot. 26, 466. Gottlieb, D., and Shaw, P. D., eds. ( 1967). “Antibiotics-Mechanism of Action,” Vol. 1, Springer-Verlag, Berlin and New York. Gray, R. A., Gauger, G. W., Dulaney, E. L., Kaczka, E. A., and Woodruff, H. B. (1964).Plant Physiol. 39, 204. Grove, J. (1963).Quart. Rev., Chem. SOC. 17, 1. Grundy, W. E., Whitman, A. L., Rdzok, E. G., Rdzok, E. J., Hanes, M. E., and Sylvester, J. C. ( 1962). Antibwt. Chemother. (Washington, D.C.)2, 399. Hagemann, G., Penasse, L., and Teillon, J. (1955). Biochim. Biophys. Acta 17, 240. Halder, A., and Bose, S. K. ( 1971).J . Antibiot. 24, 779. Hamel, E. E., and Painter, E. P. (1953). J . Amer. Chem. SOC. 75, 1362. Hamill, R. L., and Hoehn, M. M. (1973).J. Antibiot. 26, 463. Haneishi, T.,Nomura, M., Okazaki, T., Naito, A., Seki, D., Arai, M., Hata, R., and Tamura, C. (1970).174th Meet. Jap. Antibiot. Res. Ass.
ANTIMETABOLITES FROM MICROORGANISMS
57
Haneishi, T., Okazaki, T., Hata, T., Tamura, C., Nomura, M., Naito, A., Seki, I., and Arai, M. ( 1971). J. Antibiot. 24, 797. Hanka, L. J. (1960). 3. Bacteriol. 80, 30. Hanka, L. J. (1968). PTOC.Int. Congr. Chemother., Sth, 1967 Vol. HI, p. 351. Hanka, L. J., and Burch, M. B. (1960). Antibiot. Chemother. (Washington, D.C.) 10, 484. Hanka, L. J., and Dietz, A. (1973). Antimicrob. Ag. 6.Chemother. 3, 425. Hanka, L. J., Bergy, M. E., and Kelly, R. B. (1966). Science 154, 1667. Hanka, L. J., Evans, J. S., Mason, D. J., and Dietz, A. (1967). Antimicrob. Ag. Chemother. p. 619. Hanka, L. J., Martin, D. G., and Reinke, L. M. (1972). Antimicrob. Ag. (?1 Chemother. 1, 135. Hanka, L. J., Martin, D. G., and Neil, G. L. (1973). Cancer Chemother. Rep. 57, 141.
Hardegger, C., Liechti, P., Jackman, L. M., Butler, A., and Plattner, P. A. (1963). Helu. Chim. Acta 46, 60. Harned, R. L., Hidy, P. H., and LaBow, E. K. (1955). Antibiot. Chemother. (Washington, D.C.) 5, 204. Harris, D. A., Ruger, M., Reagon, M. A., Wolf, T. J., Peck, R. L., Wallick, H., and Woodruff, H. B. (1955). Antibiot. Chemother. (Washington, D.C.) 5, 183. Harris, J. S., and Kohn, H. I. ( 1941). J . P ~ U T ~ Q CExp. O ~ . Ther. 73, 383. Herr, R. R., Eble, T. E., Bergy, M. E., and Jahnke, H. K. (1960). Antibiot. Annu. p. 236. Herr, R. R., Jahnke, H. K., and Argoudelis (1967). J . Amer. Chem. SOC. 89, 4808. Hersh, E. M., and Brown, B. W. (1971). Cancer Res. 31, 834. Hoeksema, H., Slomp, G., and van Tamelen, E. E. (1964). Tetrahedron Lett. p. 1787. Hessler, E. J., and Jahnke, H. K. (1970). J. Org. Chem. 35,245. Hitchings, G. H., and Elion, G. M. (1963). In “Metabolic Inhibitors” (R. M. Hochster and J. H. Quastel, eds.) Vol. 1, p. 215. Academic Press, New York. Hochstein, F. A. (1959). I. Org. Chem. 24, 679. Hori, M., Wakashiro, T., Ito, E., Sawa, T., Takeuchi, T., and Umezawa, H. (1968). 1. Antibiot. 21, 264. Huber, F. M. (1967). In “Antibiotics” (D. Gottlieb and P. Shaw, eds.) Vol. 1, p. 181. Springer-Verlag, Berlin and New York. Hiitter, R., Poralla, K., Zachau, H. G., and Zahner, H. (1966). Biochm. Z. 344, 190. Isono, K., Asahi, K., and Suzuki, S. (1969). J . Amer. Chern. SOC. 91, 7490. Johnson, J., and Munvin, H. F. (1925). Wis., Agr. Exp. Sta., Res. Bull. 62, 1. Kaczka, E. A, Gitterman, C. O., Dulaney, E. L., and Folkers, K. (1962). Biochemistry 1, 340. Kamka, E. A., Dulaney, E. L., Gitterman, C. O., Woodruff, H. B., and Folkers, K. (1964). Biochem. Biophys. Res. Commun. 14, 452. Kaczka, E., Miller, T., Tausig, F., and Wolf, F. (1967). Antimicrob. Ag. Chemother. p. 603. Kaiser, C. ( 1960). In “Medicinal Chemistry” (A. Burger, ed.), 2nd ed., pp. 89-132. Wiley (Interscience), New York. Kamihski, K., and Sokolowska, T. (1973). J . Antibiot. 26, 186. Katagiri, K., Ton, K., Kimura, Y., Yoshida, T., Nagasaki, T., and Minato, H. (1967). J . Med. Chem. 10, 1149.
58
DAVID L. PRUESS AND J A M E S P. BCANNELL
Kawashima, M., Hamada, Y., and Fujii, S. (1953). Phamt. BUZZ. I, 94. Keller-Schierlein, W. (1967). Helo. Chim. Acta 50, 731. Keller-Schierlein, W., Poralla, K., and Ziihner, H. (1969). Arch. Mikrobiol. 87, 339. Keller-Schierlein, W., Sauerbier, J., Vogler, U., and Zahner, H. (1970). Helo. Chim. Acta 53, 779. Kelly, R. B., Martin, D. G., and Hanka, L. J. (1969). Can. J . Chem. 47,2504. Kishi, T., Muroi, M., Kusaka, T., Nishikawa, M., Kamiya, K., and Mizuno, K. ( 1972). Chem. Phamz. Bull. 20, 940. Koe, B. K., Sobin, B. A., and Celmer, N. D. (1957). Antibiot. Annu. p. 672. Komatsu, Y., and Tanaka, K. ( 1971). Agr. Biol. Chem. 35, 526. Konig, W. A., Kneifel, H., Bayer, E., Miiller, G., and Zahner, H. (1973a). J. Antibiot. 20, 44. Konig, W. A, Loeffler, W., Meyer, W. H., and Uhmann, R. (1973b). Chem. Ber. 108, 816. Koppenhagen, V. B., and Pfiffner, J. J. (1970). J. Bid. Chem. 245, 5865. Koppenhagen, V. B., and Pfiffner, J. J. (1971). J. Biol. Chem. 246, 3075. Korobkova, T. P., Terekhova, L. P., and Maksimova, T. S. (1970). Antibiotiki (Moscow) 15, 936. Koyama, G., and Umezawa, H. (1965). 1. Antibiot. 18, 175. Koyama, G., Maeda, K., Umezawa, H., and Sitaka, Y. (1966). Tetrahedron Lett. No. 6, p. 597. Kruepe, H., and Poralla, K. (1972). Arch. Mimobiol. 85, 253. Kurosawa, H. (1952). J. Antibiot. 5, 682. Kusaka, T. (1971). J . Antibiot. 24, 756. Kusaka, T., Yamamoto, H., Shibata, M., Muroi, M., Kishi, T., and Mizuno, K. (1968). J . Antibiot. 21, 255. Kusakabe, Y., Nagatsu, J., Shibuya, M., Kawaguchi, O., Hirose, C., and Shirato, S. (1972). 1. Antibiot. 25, 44. Logan, W. R., and Newbold, G. T. (1957). J. Chem. SOC., London p. 1946. McCord, T. J., Ravel, J. M., Skinner, C. G., and Shive, W. (1957). J. Amer. Chem. SOC. 79, 5693. McIlwain, H. ( 1943). Biochem. J . 37, 265. McKennis, H., and Yard, A. S. ( 1958). J . Org. Chem. 23, 980. McLamore, W. M., Celmer, W. D., Bogert, V. V., Pennington, F. C., Sobin, B. A, and Solomons, I. A. (1953). J. Amer. Chem. SOC. 75, 105. McNall, E. G. (1960). AMA Arch. Dermutol. 81, 657. Maeda, K., Okami, Y., Kosaka, H., Taya, O., and Umezawa, H. (1952). J . Antibiot. 5. 572. Maehr, H. (1971). Pure Appl. Chem. %8, 603. Maehr, H., Blount, J. F., Pruess, D. L., Yarmchuk, L., Kellett, M. (1973). J. Antibiot. 28, 284. Martin, D. G., Slomp, G., Miysak, S., Duchamp, D. J., and Chidester, C. G. (1970). Tetrahedron Lett.No. 56, p. 4901. Martin, D. G., Hanka, L. J, and Reineke, L. M. (1971). Tetrahedron Lett. No. 41, p. 3791. Martin, D. G., Duchamp, D. J., and Chidester, C. G. (1973). Tetrahedron Lett. No. 27, p. 2549. Mathews, R. E. F. (1958). PharniacoZ. Rev. 10, 359. Mego, J. L., (1964). Biochim. Biophys. Acta 79, 221. Meister, A. (1965). “Biochemistry of the Amino Acids,” 2nd ed., Vol. 1, p. 239. Academic Press, New York.
ANTIMETABOLITES FROM MICROORGANISMS
59
Mitani, M., and Inoue, Y. (1968). J. Antibiot. 21, 492. Miyake, A., Mormato, A., and Kinoshita, T. (1953). Phurm. Bull. 1, 84. Modest, E. J., Foley, G. E., and Farber, S. (1963). I n “Metabolic Inhibitors” (R. M. Hochster and J. H. Quastel, eds.), Vol. 1, pp. 75-129. Academic Press, New York. MolIoy, B. B., Lively, D. H., Gale, R. M., Gorman, M., Boeck, L. D.,Higgens, C. E., Kastner, R. E., Huchstep, L. L., and Neuss, N. (1972). J. Antibiot. 25, 137. Moore, J. A., Dice, J. R., Nicolaides, E. D., Westland, R. D., and Wittle, C. L. (1954). J. Amer. Chem. SOC.76, 2884. Murthy, Y. K. S., Thiemann, J. E., Coronelli, C., and Sensi, P. (1966). Nature (London) 211, 1198. Nagahama, N., Suzuki, M., Awataguchi, S., and Okuda, T. (1967). J. Antibiot. 20, 261. Nagahama, N., Takamori, I., and Suzuki, M. (1971). Chem. Phurm. BuU. 19, 660. Nakagawa, Y., Kano, H., Tusukuda, Y., and Koyama, H. (1967). Tetrahedron Lett. No. 42, p. 4105. Nass, G., Poralla, K., and Zahner, H. ( 1969). Bioohem. Biophys. Res. Commun. 34, 89. Nass, G., Poralla, K., and Zahner, H. (1971). Naturwissenschaften 58, 603. Neuhaus, F. C. (1967). In “Antibiotics” ( D . Gottlieb and P. D. Shaw, eds.), Vol. 1, p. 40. Springer-Verlag, Berlin and New York. Neuman, R. E., and Tytell, A. A. (1983). Proc. Soc. Exp. BWZ. Med. 112, 57. Neuss, N., Molloy, B. B., Shak, R., and DeLaHiguera, N. (1970). Biochem. J. 118, 571. Nishimura, H., and Komatsu, Y. (1968). J. Antibiot. 21, 250. Nishimura, H., Kahgiri, K., Sato, K., Mayama, M., and Shimaoka, N. (1956). J . Antibiot. 9, 60. Nishimura, H., Mayama, M., Komatsu, Y., Kato, H., Shimaoka, N., and Tanaka, Y. 3 (1964). J . Antibiot. 17, 148. Nomura, S.,Horiuchi, T., Hata, T., and Omura, S. (1972a). I. Antibiot. 25, 365. Nomura, S., Horiuchi, T., Omura, S., and Hata, T. ( 1972b). J. Biodhem. (Tokyo) 71, 783. Novick, R. P., and Maas, W. K. (1961). J. BacterwZ. 81, 236. Ohkuma, K. ( 1961). J . Antibiot. 14, 343. Omura, S., Katagiri, M., Nakagawa, A, Sano, Y., Nomura, S., and Hata, T. (1967). 1. Antibiot. 20, 349. Owens, L. D., and Wright, D. A. (1965). Plant Physiol. 40, 927. Owens, L. D., Guggenheim, S., and Hilton, J. (1968). Biochim. Biophys. Acta 158, 219. Owens, L. D., Lieberman, M., and Kunishi, A. (1971). Plant Physiol. 48, 1. Owens, L. D., Thompson, J. F., Pitcher, R. G., and Williams, T. (1972). Chem. Commun. p. 714. Oxford, A. E. (1942). Biochem. J. 36, 438. Oxford, A. E., Raistrick, H., and Simonart, P. (1939). Biochem. J . 33, 240. Painter, E. P. (1947). J . Amer. Chem. SOC.69, 232. Patterson, E. L., Johnson, B. L., DeVoe, S. E., and Bohonos, N. (1966). Antimicrob. Ag. Chemother. p. 115. Perlman, D., and Toohey, J. I. (1968). Arch. Biochem. Biophys. 124, 462. Pierard, A., Glansdorff, N., Mergeay, M., and Wiame, J. M. (1965). J. Mol. Biol. 14, 23. Ciskala, A., and Sorm, F. (1964). Collect. Czech. Chem. Commun. 29,2060.
60
DAVID L. PRUESS AND JAMES P. SCANNELL
Pittillo, R. F., and Foster, J. W. ( 1964). J . Baderiol. 67, 53. Pittillo, R. F., and Hunt, D. E. (1967). In “Antibiotics” (D. Gottlieb and P. D. Shaw, eds.), Vol. 1, p. 481. Springer-Verlag, Berlin and New York. Pittillo, R. F., and Hunt, D. E. (1968). Cancer Chemother. Rep. 52, 707. Plattner, P. A., and Clauson-Kaas, M. (1945). Helu. Chim. Acta 28, 188. Pruess, D. L., Scannell, J. P., Ax, H. A., Kellett, M., Weiss, F., Demny, T. C., and Stempel, A. (1973). 1. Antibiot. 26, 261. Pruess, D. L., Scannell, J. P., Kellett, M., Ax, H. A., Janecek, J., Williams, T. H., Stempel, A., and Berger, J. (1974). J. Antibiot. (in press). Rao, K. V. (1962). Antimicrob. Ag. Chemothu. p. 178. Rao, K. V. (1963). Antimicrob. Ag. Chemother p. 179. Rao, K. V., Brooks, S. C., Kugelman, M., and Romano, A. A. (1960). Antibiot. Annu. p. 943. Ressler, C. ( 1972). J. Org. Chem. 37, 2933. Reusser, F. (1968). J . Bacteriol. 96, 1285. Reusser, F. (1969). J. Bacteriol. 100, 11. Richmond, M. H. (1962). Bacteriol. Reo. 26, 398. Rogers, H. J., Newton, G. G. F., and Abraham, E. P. (1965). Biochem. 1. 97, 573. Rottman, F., and Guarino, A. J. (1964). Biochim. Biophys. Acta 80, 632. Roy-Burman, S., and Visser, D. W. (1972). Biochim. Biophys. Acta 282, 383. Rudas, B. ( 1972). Arzneim.-Forsah, 22, 830. Rudzats, R., Gellert, E., and Halpern, B. (1972). Biochem. Siophys. Res. Commun. 47, 290. Sahm, U., Knobloch, G., and Wagner, F. (1973) J . Antibiot. 26, 389. Sano, Y., Nomura, S., Kamio, Y., Omura, S., and Hata, T. (1967). J. Antibiot. 20, 344. Sasaki, K., Kusakabe, Y.,and Esumi, S. (1972). J. Antibiot. 25, 151. Sato, K., Shimizu, S., and Fukui, S. (1970). Biochem. Biophys. Res. Commun. 39, 170. Scannell, J. P., Pruess, D. L., Demny, T. C., Williams, T., and Stempel, A. (1970). J. Antibiot. 23, 619. Scannell, J. P., Pruess, D. L., Demny, T. C., Weiss, F., Williams, T., and Stempel, A. (1971a). J . Antibiot. 24, 239. Scannell, J. P., Pruess, D. L., Kellett, M., Demny, T. C. and Stempel, A. (1971b). 1. Antibiot. 24, 328. Scannell, J. P., Pruess, D. L., Demny, T. C., Sello, L. H., Williams, T., and Stempel, A. (1972a). J. Antibiot. 25, 122. Scannell, J. P., Ax, H. A, Pruess, D. L., Williams, T., Demny, T. C., and Stempel, A. (1972b). J. Antibiot. 25, 179. Scannell, J. P., Pruess, D. L., Demny, T. C., Ax, H. A, Weiss, F., Williams, T., and Stempel, A. ( 1 9 7 2 ~ )In . “Chemistry and Biology of Peptides” (J. Meinhofer, ed.), p. 415. Ann Arbor Sci. Publ., Ann Arbor, Michigan. Schabel, F. M., Jr., and Pittillo, R, F. (1961). Adoan. Appl. Microbial. 3, 223. Schein, P. S., and Loftus, S. (1968). Cancer Res. 28, 1501. Schein, P. S., Cooney, D. A., and Vernon, M. L. (1967). Cancer Res. 27, 2324. Schenck, J. R., and DeRose, A. F. (1952). Arch. Biochem. Biophys. 40, 263. Schlogl, K. (1958). Monutsh. Chem. 89, 377. Schrift, A. (1954). Amer. J. Bot. 41, 345. Schrift, A. (1961). Fed. Proc., Fed. Amer. SOC. Exp. Biol. 20, 695. Schroeder, W., and Hoeksema, H. (1959). J. A m r . Chem. SOC. 81, 1767. Shah, R., Neuss, N., Gorman, M., and Boeck, L. D. (1970). J . Antibiot. 23, 613.
ANTIMETABOLITES FROM MICROORGANISMS
61
Shapira, J., and Dittmer, K. (1961). J. Bacteriol. 82, 640. Shealy, Y. F., and Clayton, J. D. (1966). J. A m . Chem. SOC.88,3885. Shealy, Y. F., and Clayton, J. D. (1969). J. Amer. Chem. SOC. 91, 3075. Shigeura, H. T. (1963). Arch. Biochem. Biophys. 100, 472. Shigeura, H. T. (1967). In “Antibiotics” (D. Gottlieb and P. D. Shaw, eds.), Vol. 1, p. 451. Springer-Verlag, Berlin and New York. Shigeura, H. T., and Gordon, C. N. (1962). J . Biol. Chem. 237, 1937. Shive, W., and Macow, J. (1946). J. Biol. Chem. 162, 451. Shoulders, B. A., Gipson, R. M., Jandacek, R. J., Simonsen, S. H., and Shive, W. ( 1968). J . Amer. Chem. SOC. 90, 2992. Shull, G. M., and Sardinas, J. L. (1955). Antibiot. Chemother. (Washington, D.C.) 5, 398. Simon, A., (1969). Anal. Biochem. 31, 532. Sinden, S. L., and Durbin, R. D. (1968). Nature (London) 219,379. Skinner, C. G., McCord, T. J., Ravel, J. M., and Shive, W. (1955). J. Amer. Chem. SOC.78, 2412. Slechta, L. (1960). Biochem. Phurmuco2. 5, 98. Smith, E. L. ( 1965). “Vitamin Ba,” p. 94. Wiley, New York. Smith, R. L., and Dunn, F. W. (1970). J . Biol. Chem. 245, 2962. Smith, S. S., Bayliss, N. L., and McCord, T. J. (1963). Arch. Biochem. Biophys. 102, 313. Snell, E. E., and Guirard, B. M. ( 1943). Proc. Nut. Acad. Sci. US. 29, 66. Snow, M, L., Lauinger, C., and Ressler, C. (1968). J. Org. Chem. 33, 1774. Stammer, C. H. ( 1971). In “Chemistry and Biochemistry of Amino Acids, Peptides and Proteins” (B. Weinstein, ed.), Vol. I, p. 23. Dekker, New York. Stammer, C. H., Wilson, A. N., Holly, F. W., and Folkers, K. (1955). J. Amer. Chem. SOC. 77, 2346. Stapley, E. O., Mata, J. M., Miller, I. M., Demny, T. C., and Woodruff, H. B. (1964): Antimicrob. Ag. Chemother. p. 20. Stapley, E. O., Demny, T., Miller, A. K., and Woodruff, H. B. (1967). Antimicrob. Ag. Chemother. p. 595. Stapley, E. O., Miller, T. W., Mata, J. M., and Hendlin, D. (1968). Antimicrob. Ag. Chemother. p. 401. Stewart, W. W. ( 1971). Nature (London) 229, 174. Stieglitz, B., and Calvo, J. M. (1971). J . Bacteriol. 108, 95. Stim, T. B., Arnwine, B. C., and Foster, J. W. (1959). J . Bacteriol. 77, 566. Stock, C. C., Reilly, H. C., Buckley, S. M., Clarke, D. A., and Rhoads, C. P. (1954). Nature (London) 173, 71. Streightoff, F. J., Nelson, J. A., Cline, J. C., Gerzon, K., Hoehn, M., Williams, R. H., Gorman, M., and DeLong, D. C. (1969). Conf. Antimicrob. Ag. Chemother., gth, 1969 p. 8. Suhadolnik, R. J. ( 1970). “Nucleoside Antibiotics.” Wiley, New York. Suzuki, S., and Marumo, S. ( 1961). J . Antibiot. 14, 34. Suzuki, S., Isono, K., Nagatsu, J., Mizutani, T., Kawashima, Y., and Mizuno, T. (1965). 1. Antibiot. 18, 131. Takatsuki, A., and Tamura, G. (1971). J . Antibiot. 24, 232. Takatsuki, A., Arima, K., and Tamura, G. (1971). J . Antibiot. 24, 215. Takeshima, H., Kotani, M., Yagisawa, T., Yamagishi, M., Nomura, S., Omura, S., and Hata, T. ( 1969). 42nd Ann. Meet. Jap. Bacteriol. SOC. Takeuchi, S., Ogawa, Y., and Yonehara, H. (1969). Tetrahedron Lett. No. 32, p. 2737. Tamura, C. (1963). BUZZ. Chem. SOC. lap. 36, 1187.
62
DAVID L. PRUESS AND JAMES P. SCANNELL
Tanaka, F., Takeuchi, S., and Tanaka, N. (1961). J. Antibiot. 14, 161. Tanaka, F., Tanaka, N., Yonehara, H., and Umezawa, H. (1962a). I. Antibiot. 15, 191. Tanaka, F., Takeuchi, S., and Yonehara, H. (1962b). J. Antibiot. 15, 197. Tanaka, K., Tamaki, M., and Watanabe, S. (1969). Biochim. Biophys. Act4 195, 244. Tanaka, N., Miyairi, N., and Umezawa, H. (1960). J . Antibiot. 13, 265. Taylor, P. A., ‘Schnoes, H. K., and Durbin, R. D. (1972). Biochim. Biophys. Acta, 286, 107. Tejera, E., Backus, E. J., Dann, M., Ervin, C. D., Shakofski, A. J., Thomas, S . O., Bohonos, N., and Williams, J. H. ( 1952). Antibiot. Chemother. (Washington, D.C.) 2, 333. Terao, M. (1963). J. Antibiot. 14, 182. Terao, M., Karasawa, K., Tanaka, N., Yonehara, H., and Umezawa, H. (1960). 1. Antibiot. 13, 401. Tesser, G . I., and Nefkens, G. H. L. (1959). Rec. Trau. Chim. Pays-Bas. 78, 404. Tesser, G. I., Nivard, R. J. F., and Gruber, M. (1962). Rec. Truu. Chim. Pays-Bas. 81, 713. Toohey, J. I. ( 1965). Proc. Nat. Acad. Sci. U.S. 54, 934. Toohey, J. I. (1966). Fed. Proc., Fed. Amer. SOC. Exp. Biol. 25, 1628. Tuve, T., and Williams, H. H. ( 1961). J. Biol. Chem. 236, 597. Umezawa, H. (1973). Pure AppZ. Chem. 33, 129. Umezawa, H., Oikawa, K., Okami, Y., and Maeda, K. (1953). J. BacterioZ. 68, 118. Vavra, J. J., Deboer, C., Dietz, A., Hanka, L. J., and Sokolski, W. T. (1960). Antibiot. Annu. p. 230. Villa-Trevino, S., Schull, K. H., and Farber, E. (1963). J. Bid. Chem. 238, 1757. Wagner, G. ( 1971). P b a m z i e 26,377. Walker, J. E., and Abraham, E. P. (1970). Biochem. I . 118, 557. Wakon, R. B., and Rickes, E. L. (1962). 1. Bacterial. 84, 1148. Westley, J. W., Pruess, D. L., Volpe, L. A., Demny, T. C., and Stempel, A. (1971). 1. Antibiot. 24, 330. Whitney, J. G., and Funderburk, S. S. (1970). Abstr. Int. Congr. Microbiol. loth, 1970, p. 101. Williams, R. H., Lively, D. H., DeLong, D. C., Cline, J. C., Sweeney, M. J., Poore, G. A., and Larsen, S. H. ( 1968). 1. Antibiot. 21, 463. Woods, D. D. (1964). Proc. Int. Congr. Chemother., 3rd, 1963 Vol. I, p. 227. Woods, R. C., Ferone, R., and Hitchings, G. H. ( 1961). Biochem. Pbarmucol. 6, 113. Woolley, D. W. (1946). J. Biol. Chem. 166, 783. Woolley, D. W. (1952). “A Study of Antimetabolites.” Wiley, New York. Woolley, D. W. (1963). In “Metabolic Inhibitors” (R. M. Hochster and J. H. Quastel, eds.), Vol. 1, p. 445. Academic Press, New York. Woolley, D. W., Schaffner, G., and Braun, A. C. (1955). J. BioZ. Chem. 215, 485. Yamashita, T., Miyairi, N., Kunugita, K., Shimizu, K., and Sakai, H. (1970). J. Antibiot. 23, 537. Yiintsen, H. (1958). J. Antibiot. 11, 244. Yiintsen, H., Ohkuma, K., Ishii, Y., and Yonehara, H. (1956). J. Antibiot. 9, 195. Zahner, H., and Poralla, K. (1970). Proc. Znt. Congr. Chemother., 6th, 1969 Vol. 11, p. 1111. Ziihner, H., Hiitter, R., and Bachmann, T. ( 1960). Arch. MicrobioZ. 38,325.
lipid Composition as a Guide to the Classification of Bacteria
NORMAN SHAW Mimoblological Chemisty Research Laboratory, School of Chemistty, University of Newcastle upon Tyne, Newcastle upon Tyne, England
I. Introduction ................... 11. The Major Types of Lipids Found in Bacteria ..... A. Apolar Components .............. B. Phospholipids ................ C. Glycolipids .................. D. Neutral Lipids ................ E. Nonextractable Lipids ............. 111. Analysis of Bacterial Lipids ............ IV. Distribution of Lipids in Bacteria .......... A. Fatty Acids .................. B. Phospholipids ................. C. Glycolipids .................. V. Correlation of Lipid Composition with Taxonomic Classification ................... A. Pseudomonadales ............... B. Eubacteriales ................. C. Actinomycetales ............... D. Mycoplasmatales ............... VI. Conclusions ................... References ....................
I.
63 64 64 67 69 69 70 71 74
75 81 81 98 98
99 102 103 103 104
Introduction
The first attempt to correlate lipid composition with taxonomic classification was made by Abel et al. (1963), who showed that qualitative fatty acid analyses could be used to differentiate between various organisms. In the following year Kates (1964) reviewed the'information then available on all types of bacterial lipids, and he was able to extend these correlations to many more bacteria. In the ensuing years, owing to an increasing awareness of the importance of lipids in membrane processes and a tremendous improvement in lipid methodology, information on bacterial lipids has increased logarithmically and shows no sign of diminution. This information has been discussed extensively in two monographs ( Asselineau, 1966; O'Leary, 1967) and several reviews ( Op den Kamp et al., 1969b; Lennarz, 1970; Shaw, 1970). The purpose of this review is to reassess in the light of this intensive work the value of lipid composition as a guide to the classification of bacteria. The ideal chemotaxonomic method has three basic criteria. First, it should be applicable to as large a number of organisms as possible; second, the required information should be readily obtained; and third, 63
64
NORMAN SHAW
the parameters utilized should differ as widely as possible from one genus or family to the next. After reading this review, I hope you will share my view that the use of lipid composition as a chemotaxonomic method goes further toward satisfying these criteria than the other methods presently in use. Lipids are universally present in bacteria, they are easily and specifically extracted, and they are readily identified. Moreover, the great variety of lipid species present in bacteria satisfies in large measure the third criterion and is probably as near to an ideal solution as we are likely to reach. There are three parameters which nay be used in a taxonomic study of lipid composition-namely, phospholipids, glycolipids, and their respective apolar constituents. Each of these three classes has many different members and the presence or absence of any individual type may be significant. After a brief discussion of the types of lipids present in bacteria and their analysis, the distinctive taxonomic patterns which have clearly been established will be described. II.
The Major Types of Lipids Found i n Bacteria
Lipids are most simply defined as those natural products which may be isolated from biological materials by extraction with organic solvents and which are usually insoluble in water. This broad definition would include sterols, terpenoids, isoprenoids, waxes, and pigments, many of which are not present in bacteria. The lipids commonly found in bacteria are amphiphatic molecules, i.e., they consist of distinct polar and apolar regions. The apolar part is usually a diacylglycerol derivative-diglyceride-in which long-chain fatty acids are esterified to carbon atoms 1 and 2 of glycerol. The polar group linked to carbon atom 3 of glycerol may be a phosphate ester-phospholipid-or a carbohydrate derivative-glycolipid.
A. THE APOLARCOMPONENTS
1. Fatty Acids The fatty acids most frequently found in bacteria are between C, and C, in length although those between C , , and C, are usually the most prevalent. They may be divided into four major groups: straightchain, branched-chain, unsaturated, and cyclopropane ( Fig. 1). Their common names and shorthand notation are shown in Table I. The straight-chain fatty acids are those found elsewhere in nature, C12, Cli, Cle, and CIS. However, the corresponding p-hydroxy derivatives, in particular p-hydroxymyristic acid (3-OH C , , ) , are not normally
CHEMOTAXONOMY OF BACTERIAL LIPIDS
CH,(CH,)n COOH
65
CH,(CH& CHCH,COOH
I
OH
p - Hy droxy
Saturated
CH,?H(CH,),
COOH
CHS is0
CH, ( CH,) CH= CH ( CH,) COOH
CH,CH,CH(CH,), COOH I CHS
anteiso
CH,( CH,) I CH‘C’
CH(CH,), COOH
H,
Unsaturated
Cyclopropane
FIG.1. Formulas of the major types of fatty acids.
found in higher organisms, but are common components of the bound lipids present in the walls of gram-negative bacteria ( Section 11,E). The branched-chain fatty acids are of two types: the is0 form in which the methyl group is located on the penultimate carbon atom and the anteiso in which the methyl group is located on the antepenultimate TABLE I THESYSTEMATIC NAMES,COMMON NAMES, AND SHORTHAND FORMULAS OF THE FATTYACIDSFOUNDI N BACTERIA Shorthand formula. Saturated:
Dodecanoic acid Tetradecanoic acid CIS Hexadecanoic acid C18 Octadecanoic acid C*O Eicosanoic acid cl G: I*’ cis-9-Hexadecenoic acid ClS:l*~ cis-9-Octadecenoic acid clS:l*ll cis-1 1-Octadecenoic acid is0 Cl€.br 13-Methyl tetradecanoic acid anteiso Clbbr 12-Methyl tetradecanoic acid anteiso C 17br 14Methylhexadecanoic acid C1, (or CIgv) cis-11,12-Methylene octadecanoic acid c1*
Cl4
Unsaturated
Branched chain
Cyclopropane
Systematic name
Common name Lauric acid Myristic acid Palmitic acid Stearic acid Arachidic acid Palmitoleic acid Oleic acid Vaccenic acid
Lactobacillic acid
I n unsaturated acids, the number preceding the colon indicates the number of carbon atoms, and that following the number of double bonds. The second number indicates the position of the double bond, counting from the carboxyl end of the chain.
68
NORMAN SHAW
H,COCH=CH(CH,), CH, CH,(CH2),coo& I H2COP0,C~CH,NH,
FIG.2. Plasmalogen analog of phosphatidylethanolamine.
carbon atom. The unsaturated acids found in bacteria are monounsaturated; the double bond in these is most frequently found between carbon atoms 11 and 12. The polyunsaturated fatty acids present in higher organisms are completely absent in bacteria. Conversely, the cyclopropane fatty acids, rarely found in higher organisms, are frequently encountered in bacteria. The biosynthetic precursors of these fatty acids are the corresponding monounsaturated fatty acids; as a consequence the major example is cis-ll,12-methylenehexadecanoicacid ( lactobacillic acid).
2. Alk-1-enyl Ethers An increasing number of anaerobic bacteria, in particular Clostridium (Goldfine, 1964), have been shown to contain “plasmalogens,” which are lipids containing an alk-1-enyl ( a,p-unsaturated ) ether substituent usually on C-l of glycerol, together with a normal fatty acid in ester linkage at C-2 (Fig. 2 ) . The alk-I-enyl ether substituent is released as the corresponding aldehyde upon acid hydrolysis, and aldehydes corresponding to the major types of fatty acids have been identified.
3. Alkyl Ethers The lipids of extremely halophilic bacteria do not contain normal fatty acid components but have long-chain alcohols bound by an ether linkage (Kates et al., 1966). This type of linkage is extremely resistant to chemical hydrolysis and presents considerable difficulty for structural analysis. These di5culties may be responsible in part for the dearth of information on the distribution of this type of lipid. The principal alcohol so far described is dihydrophytyl alcohol (Fig. 3 ) .
4. More Complex Fatty Acids Mycobacteria and related organisms possess considerably more complex fatty acids containing up to 90 carbon atoms. It is interesting to note that these most complex lipids, whose structures are only now being established by the use of modern techniques, particularly nuclear magnetic resonance spectroscopy and mass spectrometry, were among the CH,CHCH,CH,(CH,CHCH,CH,),CH,CHCH,CH,OH I I I CH, CH, CH,
FIG.3. Dihydrophytyl alcohol.
CHEMOTAXONOMY OF BACTERIAL LIPIDS
67
first bacterial lipids to be isolated and studied in the pioneering work of Anderson and his colleagues. Mycolic acids are derivatives of longchain p-hydroxy fatty acids, which may be unsaturated or contain cyclopropane rings, containing an additional long-chain alkyl substituent on the carbon atom carrying the carboxylic acid (Fig. 4 ) . Related compounds have been isolated from some Corynebacterium spp. (GO-GO) and Nocardia (around C,) and are of great taxonomic value (see Section V,C ) . OH I CH,(CHJn,CH= CH(CHJ,~CH=CHCH(CHJn~CHCHCOOH I
CH,
I
CZZ%,
FIG.4. A mycolic acid.
B. PHOSPHOLIPIDS The simplest phosphohpid is 1,2-di-O-acyl-sn-glycerol 3-phosphate, “phosphatidic acid,” which although an important intermediate in biosynthesis, rarely occurs itself in bacteria. The common phospholipids are based on phosphatidic acid containing an additional group, X, forming a phosphodiester bond, and are usually named “phosphatidyl-X (Fig. 5). Phosphutidylglycerol is probably the most widely occurring bacterial phospholipid although it was first isolated from plants. Bisphosphutidylglycerol usually co-occurs with phosphatidylglycerol, and a biosynthetic relationship between these two phospholipids has been established (Short and White, 1972). This lipid was originally isolated from beef heart, hence the trivial name cardiolipin. It is the most apolar of the phospholipids, a property resulting from the presence of four acyl residues in the molecule instead of the more normal two. Aminoacyl phosphatidylglycerol This recently discovered class of phospholipids ( Macfarlane, 1962b) is apparently unique to bacteria. An amino acid is esterified to a hydroxyl group of the glycerol moiety in phosphatidylglycerol. The range of amino acids found so far is limited, alanine and lysine being predominant. Phosphutidylethunolamine ( cephalin ) is a very common phospholipid and its N-methyl and N-dimethyl derivatives also occur. Phosphatidylcholine (lecithin), although widely distributed in higher organisms, is comparatively rare in bacteria. Phosphutidylinositol. This lipid is not a common constituent of bacteria, but derivatives containing mannose, phosphatidylinositol mannosides, are prevalent in mycobacteria and related organisms.
68
NORMAN S M W
c:
H, OCOfCHJnCH, CH,(CH,), COOCH I H,COPO,-X
Phosphatidyl- X
X =
Name
H
Phosphatidic acid
CH,CH,NH,
Phosphatidy lethanolamine
CH,CH&CHJ~
Phosphatidylcholine
CH,CH(OH)CH,OH
Phosphatidy lgly cerol
F
CH,CH(OH) H, 0 I RCH-C=O
Aminoacyl phosphatidylglycerol
I
NH, CH,CH(OH)CH20POsCH,CHCHzOCOR Bisphosphatidylglycerol I
(-J
OCOR
HoQ-Mmose
HO HO
0-
Phosphatidylinositol
Mannoee 0 ~
Phosphatidylinositol dimannoside
FIG.5. Structures of the major phospholipids.
Minor phospholipids. Phosphatidylserine is an intermediate in the synthesis of phosphatidylethanolamine but is rarely present itself. Another intermediate which rarely accumulates is phosphatidylglycerophosphate. Trace amounts of Zysophospholipids are also occasionally found. These are monoacyl derivatives of the major phospholipids; i.e., one of the acyl substituents has been lost. The sphingolipids, commonly found in animal tissues, were until recently thought to be absent from bacteria but they have now been found in Bacteroides melaninogenicus (Rizza et al., 1970). It is interesting to note that the polar groups of these lipids resemble those of normal bacterial phospholipids as distinct from the animal sphingolipids which are usually based upon phosphorylcholine. Another recent development has been the isolation and characterization of a new class of glycophospholipids which are glycerophosphoryl or phosphatidyl derivatives of diglycosyl diglycerides, the latter being common glycolipid components of gram-positive bacteria ( Section II,C ) .
CHEMOTAXONOMY OF BACTERIAL LIPIDS
69
( The term “glycophospholipid is used for those phospholipids which also contain carbohydrate constituents, thus the phosphatidylinositol mannosides mentioned above are also glycophospholipids.) Although they have been found in several bacteria (Shaw and Stead, 1972), it is doubtful that they will prove to be of taxonomic importance as they are always structurally related to the diglycosyl diglyceride which is present in the same organism, usually in larger concentrations.
C. GLYCOLIPJDS Glycolipids are derivatives of fatty acids and carbohydrates and do not contain phosphorus. Two major types occur in bacteria: diglycosyl diglycerides and acylated sugars. A comprehensive review on all aspects of bacterial glycolipids has recently been published ( Shaw, 1970). Diglycosyl diglycerides consist of a disaccharide moiety glycosidically bound to the hydroxyl group of a diglyceride. Five major types have so far been characterized: a-diglucosyl, p-diglucosyl, dimannosyl-, digalactosyl-, and galactosylglucosyl diglyceride ( Fig. 6). The related mono-, tri-, and tetraglycosyldiglycerides have also been found, but usually in much smaller quantities. Acylated sugars. These glycolipids do not contain glycerol, and the fatty acids are esterfied directly to the sugar. They vary in complexity from the simple acylated glucoses to the diacyl trehalose and diacyl inositol mannoside (Fig. 7).
D. NEUTRAL LIPIDS
Glycerides. Unlike higher organisms, glycerides do not occur in large amounts in bacteria. Those present are mono- and diglycerides; no conclusive evidence has been presented for the occurrence of triglycerides. Poly-P-hydroxybutyric acid. This polymeric lipid can constitute a large proportion of the dry cell weight of some bacteria and presumably functions as an energy reserve. It is of no taxonomic importance, but its presence in large amounts can complicate the purification and identification of the other lipids. Sterols and carotenoids. Bacteria either do not contain sterols or they are present in such small amounts that it is impossible to exclude contamination as the probable source. Although bacteria are capable of synthesizing isoprenoids and utilize such derivatives as intermediates in cell wall synthesis (Rothfield and Romeo, 1971), they apparently cannot perform the cyclization steps to form sterols. Carotenoids occur universally in photosynthetic bacteria and in some aerobic nonpigmented bacteria.
70
&-
NORMAN SHAW
YH,OH
oy,
HO
OH
CHOCOR I CH,OCOR
HO OH
CH,OH YH,OH
CHOCOR I CH,OCOR
PCH, CHOCOR I CH,OCOR
OH OH
(d)
CH,OH
CHOCOR I
HOG H & O /
CH,OCOR (e)
FIG.6. Structures of the five major types of diglycosyl diglycerides: ( a ) a-diglucosy1 diglyceride; ( b ) p-diglucosyl diglyceride; ( c ) galactosylglucosyl diglyceride; ( d ) digalactosyl diglyceride; ( e ) dimannosyl diglyceride.
E. NONEXTRACTABLE LIPIDS In many bacteria, acid or alkaline hydrolysis of the cells after exhaustive extraction with lipid solvents releases further quantities of fatty acids. The nature of the material from which these fatty acids emanate is in many instances unknown. In gram-negative bacteria, a lipid moiety called “lipid A is an important constituent of the lipopolysaccharide present in the cell walls. Lipid A is a complex molecule composed of glucosamine, phosphate and fatty acids of which there is a large proportion of p-hydroxymyristic acid, The detailed structure of this lipid is
71
CHEMOTAXONOMY OF BACTERIAL LIPIDS
$!H,OCOR
(-J
RCOO
CH,OCOR
0 I OH
I
I
RCOO
I
RCOOFH,
OH
CH,OCOR
(b)
FIG. 7. Some glycolipids of the acylated sugar type: ( a ) an acylated glucose; ( b ) diacyltrehalose ( “cord-factor” ) ; ( c ) diacylinositolmannoside.
under active investigation and preliminary results would suggest a striking similarity in the basic structure of lipid A from Salmonella, Escherichia, Pseudomonm, and Shigella ( Luderitz, 1970). Another lipid-polysaccharide polymer has recently been isolated from several gram-positive bacteria, This “lipoteichoic acid” consists of a teichoic acid and covalently bound lipid which may be a phosphatidyl diglycosyl diglyceride or a related compound (Toon et al., 1972). However, in both gram-positive and gram-negative organisms these lipid containing polymers are extremely difficult to isolate and purify and therefore from both a chemical and a practical standpoint do not appear to have taxonomic potential. Ill.
Analysis of Bacterial Lipids
The expansion of our knowledge of lipids has been effectively controlled by the availability of suitable practical techniques for their investigation. The present exponential growth was initiated by the advent of gas-liquid chromatography and thin-layer chromatography, and by a realization by biochemists and microbiologists that lipids, as major components of cell membranes, must play a fundamental part in cell
72
NORMAN S H A W
processes. A study of lipid composition using these chromatographic techniques which are now (or should be) routinely available in aI1 laboratories, is a comparatively undemanding exercise which compares favorably with other chemotaxonomic methods. Numerous books and review articles have been devoted exclusively to lipid technology (e.g., Marinetti, 1967); indeed one is forced to the conclusion that much more effort has been devoted to devising lipid techniques than actually using them to obtain results. The premise that the analysis of lipids is a “black art” best left to practising mystics is certainly a myth that should be immediately dispelled. All chromatographic techniques rely on the availability of authentic materials to enable adequate comparisons to be made and many pure lipids are now available from commercial suppliers although often at inflated prices. To overcome the latter, or when a particular lipid is not commercially available, two alternatives present themselves; either to request a sample from a colleague known to possess some, or to isolate the lipid from a known source. The former alternative is easier and usually successful, the latter can provide valuable practical experience. A complete analysis (Fig. 8 ) requires ( i ) isolation of the lipids, (ii) determination of fatty acid composition, and ( iii ) identification of individual lipid components. The total lipid present in bacteria usually represents at least 3-5% of the dry cell weight, although it can be much higher, so a gram or so of dry cells is sufficient. The lipids are extracted by stirring at room temperature with organic solvents, usually chloroform-methanol mixtures, and evaporation of the solvents leaves the crude lipid mixture. This can be analyzed directly, but it is preferable to remove any impurities present, the procedure using Sephadex (Wells and Dittmer, 1963) being extremely efficient and convenient. This step effectively removes nonlipid contaminants, e.g., amino acids, sugars which are usually present albeit in small quantities, but which can lead to erroneous identifications. Treatment of a sample of the total lipid with sodium methoxide liberates the fatty acids as their methyl esters which may then be analyzed by gas-liquid chromatography. If no further lipid analyses are to be carried out the fatty acid composition can be examined without prior removal of the lipids. Acids or alkaline hydrolysis of the cells followed by extraction with organic solvents, usually ether, effectively removes the liberated fatty acids which following esterification can be analyzed by gas-liquid chromatography, However, it is important to note that this latter procedure also removes the fatty acids present in the nonextractable lipids (see Section 11,E).The analysis of the individual lipid components is conveniently carried out by thin-layer chromatography coupled with the use of specific spray reagents. Phospholipids can be detected with the Zinzadze reagent (Dittmer and Lester, 1964),
73
CHEMOTAXONOMY OF BACTEXIAL LIPIDS
I Dry rells extraction with chloroform/ methanol
acid or alkaline hvdrolvsis followed by extraction with ether
Fatty acids identified by GLC
I
Crude lipid
i
removal of contaminants with Sephadex
alkaline 4
Pure lipid silicic acid chromatography
hydrolysis
Neutral ripids
-1
Glycolipids
Water-soluble gly cosides, etc., identified by paper chromatography
I-
Preliminary identification of major components by TLC
Phospholipids
alkaline hydrolysis
Fatty acids
5
Fatty acids
Water-soluble phosphate esters identified by paper chromatography
FIG.8. An analytical sequence for the separation a n d identification of bacterial lipids. GLC, gas-liquid chromatography; TLC, thin-layer chromatography.
and free amino groups with the ninhydrin reagent (Marinetti, 1962). Thus phosphatidylethanolamine would react with both these reagents. Glycolipids and other lipids containing vicinal hydroxyl groups (-CH*OH-CH.OH-) can be detected with the periodate-Schiff reagents ( Shaw, 1968b), which also give characteristic colors with some phospholipids. Many solvent systems have been described, but the system chloroform-methanol-water, 65 :25 :4, is particularly useful for bacterial lipids. Two-dimensional analysis gives a much better resolution and Fig. 9 illustrates a typical separation of a hypothetical mixture (Minnikin and Abdolrahimzadeh, 1971). The lipids may be separated into neutral, glycolipid, and phospholipid components by chromatography on a column of silicic acid (Vorbeck and Marinetti, 1965), and the components are further characterized, The water-soluble phosphate esters produced by deacylation of phospholipids may be identified by paper chromatography ( Dawson, 1960), as may the diglycosylglycerols similarly produced from diglycosyl diglycerides ( Brundish and Baddiley, 1968).Their sugar
74
NORMAN SHAW
FIG. 9. A separation of the major types of lipids found in bacteria by twodimensional thin-layer chromatography. BPG, bisphosphatidylglycerol, PG, phosphatidylglycerol; APG, aminoacyl phosphatidylglycerol; PE, phosphatidylethanolamine; PC, phosphatidylcholine; PI, phosphatidylinositol; DGDG, diglycosyl diglyceride. The arrows indicate the direction of development. Q , [email protected] , Phospholipid; , phospholipid containing NH, group; teristic reaction with the periodate-Schiff reagents.
composition is easily established by acid hydrolysis and the anomeric configuration by hydrolysis with specific glycosidases. Separation into glycolipid and phospholipid components enables individual fatty acid analyses to be carried out, but the procedure is unnecessary in the present context. The identification of new lipids would certainly require a much more rigorous approach, but even the observation of an unknown lipid, irrespective of its identity, could be of great taxonomic significance. IV.
Distribution of lipids in Bacteria
Before considering the distribution of the various types of lipids a cautionary note is advisable concerning the comparison of data obtained from bacteria grown under unspecified conditions. The chemical composition of many bacterial components, including lipids, can be affected by a variety of external factors, e.g., temperature of growth, substrate
CHEMOTAXONOMY OF BACTERIAL LIPIDS
75
composition, pH of environment, time of harvesting, etc. Fortunately, from a taxonomic standpoint this is not usually a particularly serious situation with respect to lipid composition as taxonomic conclusions are derived mainly from qualitative rather than quantitative differences and it is usually the latter that are most affected by external factors. Many bacteria will incorporate into their membranes exogenous lipid, but fortunately the media routinely used for cultivation are free from the types of lipids usually found in bacteria. The other external factors usually affect quantitative composition, and it is extremely rare, except under extreme conditions, for one particular lipid species to disappear and to be replaced by another different lipid. However, it is obviously preferable when comparative studies are to be made to grow the bacteria under as near identical conditions as possible. The lipid composition of bacteria is summarized in Tables II-VI. The classification scheme adopted throughout is that used in the seventh edition of Bergey’s Manual (Breed et al., 1957). I have not attempted to cite every reference to bacterial lipids which has appeared in the literature, but I have chosen those which are either the most recent, comprehensive, and in the author’s opinion, most reliable. The tables have been divided into sections for phospholipids, glycolipids, and fatty acids. In order to keep them as clear as possible, I have not attempted to include quantitative determinations; only the positive identification of a particular lipid is recorded. Of the class Schizomycetes, information is available on the genera of the Orders Pseudomonadales, Hyphomicrobiales (Table 11), Eubacteriales (Tables I11 and IV), Actinomycetales (Table V), Spirochaetales, and Mycoplasmatales (Table VI). The most significant feature to emerge from the data presented is the clear distinction between the lipid composition of gram-negative and gram-positive bacteria. The majority of gram-negative organisms have a simpler lipid composition although they have lipid in their cell walls as well as their cytoplasmic membrane. The lipids of gram-positive bacteria are generally more complex and therefore offer more scope for taxonomic correlations.
A. FAIT ACIDS A consideration of the number of different fatty acids theoretically possible in any particular lipid assuming a chain length of between ten and twenty carbon atoms for all of the four major types, straightchain, branched-chain, unsaturated, and cyclopropane, will reveal that it is extremely high. White and Frerman (1967) were able to positively identify no less than 64 different fatty acids in the lipids of Staphylococcus aureus. Fortunately, two or three fatty acids usually represent
Tables 11-VI: The following abbreviations are used in these tables. PG, phosphatidylglycerol; BPG, bisphosphatidylglycerol; APG, aminoacylphosphatidylglycerol;PE, phosphatidylethanolamine; PC, phosphatidylcholine; PI, phosphatidylinositol; PIM, phosphatidylinositol mannosides; S, straight-chain fatty acids; B, branched-chain fatty acids; U, unsaturated fatty acids; C, cyclopropane fatty acids. TABLE I1 LIPIDSOF PSEUDOMONADALES AND HYPOMICROBIALES Phospholipids PG PSEUDOMONADALES Thiorhodaceae Chromatium strain D Athiorhodaceae Rhodopseudomonas spheroides R . capsulata R . gelatinosa R. palustrw Rhodospirillum rubrum R . capsulatus Chlorobacteriaceae Chlorobium sp. Chloropseudomonaa ethylieum Nitrobacteraceae Nitrosocystis oceanus Nitrosomonaa europaea
+ +
BPG APG
+
+
+ + + + + + + +
PE
+ +
PC
Fatty acids
PI
GlymOther lipid
S
B
U
+' +
+
+ + + + + +
+ +
+ + +
+ + +
1-
+' + +
+ +
+ +
Tr
References
Steiner et al., 1970a
fC
+
C Other
Miscellaneous
+
+*
fb
Woodet al., 1965 Gorchein, 1968 Steiner et al., 1970b Wood et al., 1965 Wood et al., 1965 Wood et al., 1965 Wood et al., 1965 Hirayama, 1968 Hirayama, 1968 Cruden and Stanier, 1970 Constantopoulis and Bloch, 1967 Hagen et al., 1966 Hagen et al., 1966
z
$j
E
P
Methanomonadaceae Hydrogenomonas sp. Thioacteriaceae Thiobacillus neapolitanus T . thioparus T . intermedius T . novellus
T . thiooxidans Pseudomonadaceae Pseudomonas acidovorans P. aureofaciens P . aeruginosa
+ + + + + + + + + + + + + + +
+ + + + +
+d
+
+ + + + + + + +
+ +
P . alcaligenes P. chloroaphis P . denitri$cans P . diminzlta
+ + + + +
+ +
P . fluorescens
+ +
+
Thiele, 1971
+'
+
+g
+h
+ + + +
+ + + +
+
+
+b
+' +'
+'
+b
Barridge and Shively, 1968 Barridge and Shively, 1968 Barridge and Shively, 1968 Barridge and Shively, 1968 Shively and Benson, 1967 Levin, 1971 Wayne Moss et al., 1972 Wilkinson, 1970 Hancock and Meadow, 1969 Wayne Moss et al., 1972 Wayne Moss et al., 1972 Wilkinson, 1970 Wilkinson, 1970 Wilkinson 1968a, 1969 Wilkinson and Bell, 1971 Cullen et al., 1971 Welsh et al., 1968 (continued)
2
TABLE I1 (Continued) Phospholipids PG
BPG APG
PSEUDOMONADALES Pseudomonadaceae (Continued) P . frugi P . kingii P . maltophilia
+ +
P . mucidolas P . multworans P . ovalis P . pavonacea P . putida P . rubescens P . stutzeri
P . synoynea P . synxantha
+ + + +
PE
PC
PI
Other
Glycolipid
S
+a
+ + +
+ + + +' + +'
+ + +
+ + + +' +
+ + +
+ + +
+ +
+ + +
.+ +
+ + + +
+ +
MiScellaC Other neous
Fatty acids
Tr
+'
+ +a
B
U
+
+ +
+ +
+ +
+ +
+ +
Wilkinson, 1970 Ssmuels et al., 1973 Wilkinson, 196th Wayne Moss et al., 1972 Wilkinson, 1970 Samuels et al., 1973 Wilkinson, 1970
+' +b
+'
References
+b
+*
Phizackerley et al., 1966 Wilkinson et al., 1973 Wilkinson, 1970 Wayne Moss et al., 1972 Wilkinson, 1968a,b, 1972 Wilkinson et al., 1973 Wikinson, 1972 Wayne Moss et al., 1972 Wilkinson et d.,1973 Wilkinson, 1972 Wilkinson, 1970
t
P 5 z
E
8
P . taetrolens P . testosteroni Halobacterium cutirubrum H . hulobium
H . salinarium
Halobacterium sp (moderately halotolerant) Siderocapsaceae Ferrobacillus ferroozidans Spirillaceae Vibrio costicolus V .fetus
+ +
+
I-
+
+'
+
+'
+'
+'
+
+'
+&
+'
+ f
+
+' +
+
+
+
+'
+'
+&
+'
Wilkinson, 1970 Wayne Moss et al., 1972 Katea et al., 1966, 1967 Hancock and Kates, 1972 Kates et al., 1966, 1967 Hancock and Kates, 1972 Marshall and Brown, 1968 Kates et al., 1966, 1967 Hancock and Kates, 1972 Peleg and Tietz, 1971
BK 2 0
3
5
8 m b
3L! Tr
Short et al., 1969
+d
P !i!
+
+
+
+ +
+
+ +' + f
V . cholerae But yrwibrio jibrisolvens
+ +
+
+
?
+
+ + + +
Kates et d.,1966 Tornabene and Ogg, 1971 Brian and Gardner, 1968 Kunsman, 1970 (continued)
8
3
00
0
TABLE I1 (Continued) Phospholipids PG HYPOMICROBIALES Hypomicrobiaceae Hypomicrobium vulgare Rhoabmicrobium vanniellii
BPG APG
PE
Fatty acids
PI
PC
GlycoOther lipid
S
U
C
Other
References
z
+
+
+
+
+ +
n
8
+' +
+
Glucosyl diglyceride. Ornithine lipid. c Galactosyl diglyceride unknown containing galactose and rhamnose. d N-Methylphosphatidylethanolamine. Hydroxy fatty acids. J Glycophospholipid. 0 Glycolipid containing uronic acid.
Ir
b
i
+
+
Goldfine and Hagen, 1968 Park and Berger, 1967
Acylated sugar. Phospholipid containing glucosamine. j Dialkyl ether analog of phosphatidylglycerophosphate. Glycolipid sulfate. 2 Dialkyl ethers. Digalactosyl diglyceride. Dimethylphosphatidylethanolamine.
a
6
B
MiScellaneous
E
CHEMOTAXONOMY OF BACTERIAL LIPIDS
81
a major proportion of the total. A summary of the distribution of the major types of fatty acids is shown in Table VII. While there are major differences at the level of Order, significant differences can also often be observed within a particular genus. Here a comparison between the trace from a gas-liquid chromatographic separation of fatty acids and an infrared spectrum is not inappropiate. With the latter it is usual to identify only the major absorptions, but the “fingerprint” region of minor but unidentified absorptions is also important when one spectrum is compared to another. Such a situation is applicable to gas-liquid chromatographic traces of fatty acids from related species. The relative amounts of minor, unidentified, fatty acids can enable organisms to be subdivided within a genus. At this level, however, it is obviously important that growth conditions be as uniform as possible. The isolation of enzymes which specifically hydrolyze the fatty acids esterified at either carbon atoms 1 or 2 of a diglyceride has enabled significant results to be obtained concerning the distribution of specific fatty acids at these two sites. However, a discussion of these results is outside the scope of this review.
B. PHOSPHOLIPIDS The distribution of the various phospholipids is summarized in Table VIII. There is little qualitative difference between many gram-negative organisms of the Orders Pseudomonadales and Eubacteriales, the major component being phosphatidylethanolamine, and this is certainly the least satisfactory area from a taxonomic standpoint. Phosphatidylethanolamine does not usually occur in gram-positive bacteria with the notable exception of the Bacillaceae. Phosphatidylinositol does not appear until the Corynebacteriaceae except for a few reports of its occurrence in two micrococci. In the Actinomycetales phosphatidylinositol mannosides are the major phospholipids. Ikawa ( 1967) has discussed possible phylogenetic relationships which emerge from a study of phospholipid composition. C. GLYCOLIPIDS The presence of glycolipids, particularly in gram-positive bacteria, has been a comparatively recent discovery, but one of great taxonomic significance. .Of the two types of glycolipids, insufficient information is available on the structure and distribution of acylated sugars to enable general taxonomic conclusions to be drawn with certain exceptions, e.g., the diacyl inositol mannoside in propionic acid bacteria. The distribution of diglycosyl diglycerides is summarized in Table IX. This type of gly-
TABLE I11 LIPIDS OF EWACTERIALES (GRAMNEQATIVE) Phospholipids PG Azotobacteraceae Awtobacter agilis A . vinelandii Rhizobiaceae Agrobacterium tumefaciens Chromobacterium violaceurn Rhiwbium japonicum Achromobacteraceae Ackomobacter sp. Flwobacterium thermophilus Enterobacteriaceae Escherichia coli B
E. coli K12 Serratia marcescens Proteus vulgaris
BPG APG
PE
PC
+ +
+ T r
+
+
+
+ + + + +
+ + + + +
Fatty acids
PI
GlycoOther lipid
+'
+
+ + + + + +
+ + +
B
U
C Other
+ + +
+'
cellaneous
References
Randle et al., 1969 Kaneshiro and Man, 1962 Jurtshuk and Schlech, 1969 Randle et al., 1969 Randle et al., 1969 Bunn and Elkam, 1971
+*
+ +
S
Mis-
+' +'
+ + +
+d
+
+ + +
Jauch et aZ., 1970 Oshima and Yamakawa, 1972 Kanfer and Kennedy, 1963 Welsh et al., 1968 Brennan et al., 1970a Ames, 1968 Kates et al., 1964 Randle et al., 1969
*z
! i 8
Proteus mirablis Proteus P18 Salmonella typ himurium
S. paratyphi Brucellaceae Brucella abortus 3. melitensis Hemop hilus parainjluenzae Bacteroidaceae Bacteroides melaninogenicus Neisseriaceae Neisseria gonorrheae
+ + + + + +e
+ +
+ + +
Trc
+ +
+'
+ + +
+ + + + + +
+
+ + + + +'
Tr
Sud and Feingold, 1970 Nesbitt and Lennare, 1965 Ames, 1968 Macfarlane, 1962a Olsen and Ballou, 1971 Modak et al., 1970
+g
+
+
+
+ +
+ +
+ +
+
Tr
+
N . meningitidis Brevibacteriaceae Brevibacterium jlavum B . thiogenitalis N-Methyl- and N-dimethylphosphatidylethanolamine. Tetraglycosyl diglyceride. Phosphatidylserine. Acylated glucose.
+
+
d
Thiele et al., 1968 Prome et al., 1969 White, 1968
I Risea et al., 1970
+
+
+
+
+
+
8
*
U
Wayne Moss et al., 1970 N . Shaw, unpublished Wayne Moss et al., 1970 Otsuka and Shiio, 1968 Okaeaki et al., 1969
A triacyl derivative of phosphatidylglycerol also present. Hydroxy fatty acids. g Ornithine lipid. 50-70 % phosphosphingolipids. e
3
I r
f
R
TABLE I V LIPIDS OF EUBACTERIALES (GRAMPOSITIVE) Phospholipids PG
BPG APG
PE
PC
Fatty acids
PI
GlycoOther lipid
S
B
U
C Other
Miscellaneous
References ~
Micrococcaceae Micp.ocoecus caseol yticus M . cerijicans M . conglomeritus M . denitrificans M . 1ysodeikticus
M . roseus M . tetragenus M . varians Staphylococcus aureus
+ + +
+ + + + + + +
+
+ + + +
+ + + +
m
+ + + + +
+ +
+' +
+b
+
+' +' +'
+d
+ + + + + + + +
~~
Whiteside et al., 1971 Makula and Finnerty, 1970 Whiteside et al., 1971 Wilkinson el al., 1972 Lennarz and Talemo, 1966 Whiteside et al., 1971 Macf arlane, 1961,a,b Whiteside et aZ.,1971 Whiteside et al., 1971 Whiteside et al., 1971 Gould and Lennarz, 1970 Brundish and Baddiley, 1968 Short and White, 1970 Macfarlane, 1962s. Joyce et d.,1970 Houtsmuller and van Deenen, 1965 White and Frerman. 1967
*
P8
Sarcina Java
+
+
+
+ + + + +' + +
S. lutea
+ + +
+
+' + +
S. aureus ( L form)
+
+
+
+d
S. lactis
S . maxima
s. V e T t T k U l i Sporosareina ureae Lactobacillaceae Pneumoeoccus type1 Pneumococcus type XIV Streptococcus faecalis
S. lactis
S . hemolyticus
+d
+ + + +
+ +
+
+ +
+' +'
+ + + +
+f
+f
f
+a
+
+
+a
+
+
+
+c
+'
+h,i
+A
+h
+
+ +
Ward and Perkins, 1968 Brundish et al., 1967b Huston et al., 1965 Kates et al., 1966 Whiteside et al., 1971 Hunter and Thirkell, 1971 Kates et al., 1966 Whiteside et al., 1971 Whiteside et al., 1971 Whiteside et al., 1971 Whiteside et al., 1971 Brundish et al., 1965, 1967a Kaufmann el al., 1965 Fischer and Seyferth, 1968 Dos Santos Mota et al., 1970 Ambron and Pieringer, 1971 Pieringer, 1968 Welsh et al., 1968 Fischer and Seyferth, 1968 Fischer, 1970 Ishizuka and Yamakawrt, 1968 (continued)
a
5
2o
3
I 8 m c
3
2
6
P
+4
03
(x,
TABLE I V (Continued)
m
Phospholipids PG
S . pyogenes
+
BPG APG
PE
PC
Fatty acids
PI
GlycoOther lipid
+
+h
Pediococcus acidilacti P . cerevisiae P . halophilus
P . homari P . urinae-equi
P . pentosaceus Leuconostoc mesenteroides Lactobacillus acidophilus
L . arabinosum L . buchneri L. bulgaris L . casei
S
B
U
C Other
+ + + + + + +
+ + + + + + + + + Tr
+
+ +
+' +
+
+
+ + + + + + + + - I -
f'
+' +' +
+-I-
Miscellaneous
References Cohen and Panos 1966 Uchida and Mogi, 1972 Uchida and Mogi, 1972 Uchida and Mogi, 1972 Uchida and Mogi, 1972 Uchida and Mogi, 1972 Uchida and Mogi, 1972 Shaw and Stead, 1972 Exterkate et al., 1971 Shaw, 1070 Uchida and Mogi, 1972 Exterkate et al., 1971 Shaw, 1970 Exterkate et al., 1971 Exterkate et al., 1971 Shaw et al., 1968b Uchida and Mogi, 1972
z
0 in
E
' ! 8
3
+ + +
+ +
3
3 3 3
+ + +
+ ++ + + + + + + + +
3
+ +
+
.' + +
+
CHEMOTAXONOMY OF BACTERIAL LIPIDS
+ &
+ + + +
+ + +
2
b W
2 'zl eid
+ + +
.
+ + +
87
TABLE IV (Continued) Phospholipids PG
BPG APG
PE
PC
PI
B
+' + +
C . coelicolor C . diphtheriae
+ + +
++
+l
C . equi
+
i-
+'
+i.n
+'
+ +'
c. ovis
+
f
C . palvum C . xerosis
+
+ +'
Listeria
+ +
Microbacterium lacticum M . thermosphaetum
U
+ + +
C . hofmanii
monocytogenes
MiScellaC Other neous
Fatty acids GlycoOther lipid
+ + + +
+'
+'
+
+' + [email protected]
+' +hi
+ +
+ +
+ +
References Whiteside et aZ., 1971 Senn et al., 1967 Brennan et al., 1970b Gomes et al., 1966 Brennan and Lehane, 1971 Brennan and Lehane, 1971 We1by-Gieusse et aZ., 1970 Brennan and Lehane, 1971 Etemadi, 1963 Brennan, 1968 Brennan and Lehane, 1969,1971 Carrroll et al., 1968 Raines et aZ., 1968 Shaw and Stead, 1972, and unpublished Shaw, 1968a Shaw and Stead, 1970
+ +
+
+'
+d
+ +
C . jimi
+
+
+
+'
+d
+ +
A rthrobacter crystallopoites A . globiformis
+ +
+ +
+ +
+' +'
+ +
+ +
+ +
+ +
+'
+
Cellulomonas biazotea
-4. pascens
A . simplex Bacillaceae Bacillus alvei B . anthracis B . cereus
B . circulans B . licheniformia B . macerans B . megdherium
+
+
+
+
+ + + +
+d
+c
+r
+ + + + + + + + +
+f
Shaw and Stead, 1972, and unpublished Shaw and Stead, 1972, and unpublished Shaw and Stead, 1971 Shaw and Stead, 1971 Walker and Bastl, 1967 Walker and Fagerson, 1965 Shaw and Stead, 1971 Yano et al., 1970b
' A
*1
m
c
Kaneda, 1967 Kaneda, 1968 Lang and Lundgren, 1970 Saito and Mukoyama, 1970 Kaneda, 1968 Kaneda, 1967 Kaneda, 1967 Kaneda, 1967 Op den Kamp et al., 1965 (continued)
3it! r
3
$g
TABLE IV (Continued Phospholipids PG
B. natto B. polymyza B. pumilis B. stearothermophilus B . subtilis
B. thurangiensis Clostridium botulinum C . butyricum
BPG APG
+ +
+
+
+
+
+
+
+
PE
PC
Fatty acids
PI
GlywOther lipid
+ + +
+
+
+
+d
+g
S
B
U
C Other
+ + + +
+ + + + + +' + + +
Miscellaneous
References MacDougall and Phizackerley, 1969 Phizackerley et al., 1972 Kaneda, 1967 Utakami and Umetanik, 1968 Matches et al., 1964 Kaneda, 1967 Kaneda, 1967 00 and Lee, 1972 Op den Kamp et al., 1969a Bishop et al., 1967 Brundish and Baddiley, 1968 Kaneda, 1967 Kaneda, 1968 Kimble et al., 1969 Baumann et al., 1965 Goldfine and Panos, 1971 Matsumoto et al., 1971
3
P
Z
C . bifermentans
C. histolyticum
C . perfringens C . sporogenes C . welchii
+
+
+
Glycolipid of unknown structure. b Dimannosyl diglyceride. 8 Glycophospholipid. d pDiglucosy1 diglyceride. 0 Unidentified phospholipid as major component. f Hydroxy fatty acids. 0 Galactosylglucosyl diglyceride. a
1
a-Diglucosyl diglyceride. Acylated glucose.
+”
+ + + + +
+ +
Wayne Moss and Lewis, 1967 Wayne Moss and Lewis, 1967 Wayne Moss and Lewis, 1967 Wayne Moss and Lewis, 1967 Macfarlane, 1962b
Glycolipid containing galactofuranose. k Diacylinositol mannoside. 1 Phosphatidylinositol mannosides. * Diacyltrehalose. * Corynemycolic acids. Mixture of digalactosyl and dimannosyl diglycerides. P Glucosaminyl diglyceride. c “Plasmalogen” phospholipids. i
0
Bg
i 0
3
5
$ W
ii
F r
TABLE V LIPIDSOF ACTINOMYCETALES Phospholipids
Gly-
Fatty acids
co-
PG
BPG APG
PE
PC
PI
PIM Other lipid
My cobacteriaceae Mycobacterium aurum M . aviurn M . brevicale
S
B
U
+b
M . bovis M . butyricum M . chitae
+
+
+
+ + +
M. diernhofera M . flavescens M . fortuitum
+ +
+b
MiscellaC Other neous
References
+'
Lechevalier et al.,
+'
Miquel et al., 1963 Lechevalier et al.,
+'
Lechevalier et al.,
1971
1971 1971
+' +'
Okuyama et al., 1967 Lechevalier et al.,
+'
Lechevalier et al.,
+' +'
197 1 1971
Lechevalier et al., 1971
Vilkas and Rojas, 1964
Lechevalier et a1 ., M . gallinarum
M . gastri
M . intracellulare
+' +' +'
1971
Lechevalier et al., 1971
Lechevalier et al., 1971
Lechevalier et al., 1971
3
E 8
M . kansasii M . marinum M . nmum M . parafinicum M . pellegrino M . phlei
M . rhodochrous
+ + +
+ + +
+ +
+
b
+
+
+
+'
+
+'
M . smegmatis
M . terrae M . thermoresistible
M . tuberculosis
M . vaccae
+' +' +' +' +' +'
+
+
+
+'+++
+' +' +' t'
Lechevalier et al., 1971 Lechevalier et al., 1971 Lechevalier et al., 1971 Lechevalier ei al., 1971 Laneelle et al., 1965 Lee and Ballou, 1965 Brennan and Ballou, 1967 Azuma et al., 1962 Lechevalier et al., 1971 Laneelle et al., 1965 Lechevalier et al., 1971 Brennan et al., 1970a Azuma et al., 1962 Lechevalier et al., 1971 Lechevalier et al., 1971 Lechevalier et al., 1971 Lee and Ballou, 1965 Brennan et al., 1970a Lechevalier et al., 1971 Lechevalier et al., 1971 (continued)
[ g
2 X
0
3
5
$
ti
1 r
!I P
co
w
W
rp
TABLE V (Continued) Phospholipids PG
BPG APG
PE
PC
PI
GlycoPIM Other lipid
Actinomycetaceae Nocardia aateriodes N . braailiensis
Fatty acids
S €3
+'
+b
+
+
+
U
MiscellaC Other neous
+O
N . carneus
+O
N . cwiae
+O
N . coeliaca
+
+ + +
+ +' + ' + +
N . corallina
+b
N . erythropolis
+b
N . eppingerii N . farcinica
+b
N . kirovani N . leishmanii
+ +
+
+ +
+O
+' +'
+ +
+c
+a
References
Ioneda et al., 1970 Lechevalier et al., 1971 Laneelle et al., 1965 Lechevalier et al., 1971 Lechevalier et al., 1971 Lechevalier et al., 1971 Maurice el al., 1971 Yano et al., 1969 Khuller and Brennan, 1972b Yano et al., 1971 Batt et al., 1971 Yano et al., 1971 Maurice et al., 1971 Yano et al., 1971 Lechevdier et al., 1971 Vacheron et al., 1972 Yano et al., 1970a
3
E
5
N . pol ychromogenes
+
+
+
+
+h
+*
N. rhodochrozis N. rubra N. Sumatra
S . coelicolor S. jlavovirens S. gelaticus
S. sioyaensis S. viridochromogenes Mycolic acids. Diacyltrehalose. Nocardic acids. Acylated glucose. 6 Phosphatidylserine.
a
+e
+'
+b
Streptomycet aceae Streptomyces LA7017 S . aureofaciens
S . griseus
+
+
+
+
+
+ +
+ +
+
+'
+ + + + +-I-
+
+g
+#
$0
+ Glycolipids containing glucose. Hydroxy fatty acids. * Diglucosyl diglyceride. Glycosyl diglyceride containing uronic acid.
f
Yano et al., 1968 Khuller and Brennan, 1972b Ioneda et al., 1970 Yano et al., 1971 Maurice et al., 1971 Bergelson et al., 1970 Laneelle et al., 1968 Ballio et al., 1965 Ballio et al., 1965 Ballio et al., 196.5 Laneelle et al., 1968 Ballio et al., 1963 Kutaoka and Nojima, 1967 Laneelle et al., 1968 Ballio et al., 1965 Kimura et al., 1967 Ballio et al., 1966
g g
g
3
I 5
[ 5
3 4
VI
TABLE VI LIPIDSOF SPIROCHAETALES AND MYCOPLASMATALES Phospholipids PG SPIROCHAETALES Treponemataceae Treponema pallidum
T . zuelzerae MYCOPLASMATALES A choleplasma laidlawii
Mycoplasma m ycoides M . hominis
M. avian sp. strain J
M. T strain 0
+
BPG APG
PE
+
+
PC
Fatty acids
PI
GlycoOther lipid
+'
+ + +
+ + + + + +
Choline plasmalogen.
* Galactosyl diglyceride. Glucosyl diglyceride. Glycophospholipid. ~~-Diglucosyl diglyceride.
+
fd
S
B
U
+ +< +
+ +
+' +
+
+b
C Other
+' +o
+
+'
+
+
+
+
+h
Galactofuranosyl diglyceride. Phosphatidylglycerophosphate. h Acyl glucose. i Phosphatidic acid. p
Miscellaneous
References
Livermore and Johnson, 1970 Livermore and Johnson, 1970 Shaw et al., 1968a, 1972 Smith and Henrikson, 1965 Plackett, 1967 Smith and Mayberry, 1968 Smith and Mayberry, 1968 Romano et al., 1972
2
8$
i! 8
CHEMOTAXONOMY OF BACTERIAL LIPIDS
DISTRIBUTION OF
97
TABLE YII MAJOR TYPJ~S OF FATTY ACIDS
THE
Universal, but minor components in many species Widespread, particularly in gram-negative bacteria; absent or very minor components in most gram-positive bacteria except Lactobacillaceae Usually cooccur with unsaturated f a t t y acids in gram-negative Cyclopropane bacteria; also found in some members of the Lactobacillaceae Branched chain Absent or very minor components in gram-negative bacteria; widely found in gram-positive bacteria except Lactobacillaceae Alk-1-enyl ethers Clostridium and some other anaerobes Alkyl ethers Extreme halophiles Higher fatty acids Corynebacterium and Actinomycetales Straight chain Unsaturated
TABLE VIII DISTRIBUTION OF THE MAJORPHOSPHOLIPIDS PhosphatidylglycerQl Bisphosphatidylgl ycerol
Phosphatidylethanolamine Phosphatidylinositol Phosphatidylinositol mannosides
Present in all bacteria except Actinomycetales Present in most bacteria together with phosphatidylgly cerol Major component of all gram-negative bacteria; also present in Bacillaceae Propionibacterium, Corynebacterium, Arthrobacter, Nocardia Corynebacterium and Actinomycetales; may be present in Propionibacterium
TABLE IX DISTRIBUTION OF DIGLYCOSYL DIGLYCERIDES 8-Diglucosyl diglyceride a-Diglucosyl diglyceride Galactosylglucosyl diglyceride Dimannosyl diglyceride Digalactosyl diglyceride
Staphylococci, bacilli Streptococci, Acholeplasma laidlawii Lactobacilli, pneumococci, Listeria monocytogenes Micrococcus lysodeikticus, Corynebacteraum aquaticum, Microbacterium sp., Arthrobacter A rthrobacter
colipid is found in most gram-positive bacteria of the Order Eubacteriales and is genus specific. Thus all lactobacilli investigated contain the galactosylglucosyl diglyceride, and all streptococci contain the a-diglucosyl diglyceride. The same glycolipid may occur in another genus, e.g., the p-diglucosyl diglyceride is found in both staphylococci and bacilli, but species of the same genus have the same glycolipid.
98 V.
NORMAN SHAW
Correlation of Lipid Composition with Taxonomic Classification
Sufficient information is available on genera of the Orders Pseudomonadales, Eubacteriales, Actinomycetales, and Mycoplasmatales to enable meaningful comparisons to be made. The taxonomic correlations which emerge from a consideration of the lipid composition of the genera and families of these Orders will now be assessed. The relevant information is presented in the tables together with the references so the latter will not be cited again unnecessarily in the following discussion. A. PSEUDOMONADALES The genera of this Order which have received most attention are the photosynthetic bacteria, the sulfur bacteria, the pseudomonads, and the obligate halophilic bacteria. The lipids of Halobacterium are extremely characteristic and although organisms of this genus are gram-negative, they do not possess a typical cell wall. There are two significant features of the lipids of halophilic bacteria. First, the majority are negatively charged either by the presence of an additional phosphate group or a sulfate group; second, their lipids do not possess ester-linked fatty acids but have dihydrophytyl alcohol linked by an ether linkage to the glycerol. Both these deviations from the norm are undoubtedly modifications developed to contend with their unusual environment. A dialkyl phosphatidylglycerophosphate is the major phospholipid, with smaller amounts of dialkyl phosphatidylglycerol and a sulfated derivative thereof. A glycosyl diglyceride is also present as a sulfated derivative. Such a composition, ignoring the two unique modifications, is more typical of a gram-positive organism; no dialkyl derivative of phosphatidylethanolamine has been observed. The phospholipid composition of the remaining members of the Pseudomonadales is fairly typical of gram-negative bacteria; phosphatidylethanolamine ( or its N-methylated derivative) is the major lipid. Phosphatidylcholine has been found in all Rhodopseudomonas species but not Rhodospirillum. The photosynthetic bacteria also contain glycosy1 diglycerides, the most prevalent being the galactosyl diglyceride normally found in the chloroplasts of eukaryotic cells. This has a structure different from that of the galactosyl diglycerides found in grampositive bacteria. Thiobacilli have appreciable quantities of the N-methylated phosphatidylethanolamine, and Thiobacillus thiooxidans have an unusually high proportion of a CIS cyclopropane fatty acid. Considerable controversy has raged over the nature of the extracellular “wetting agent” which enables these organisms to utilize the sulfur. This agent was originally
CHEMOTAXONOMY OF BACTERIAL LIPIDS
99
thought to be phosphatidylinositol, but this lipid has not been identified as a cellular lipid component. At this juncture it is pertinent to note that the investigation of extracellular lipids produced by prokaryotic organisms is virtually an unexplored area. The Pseudomonas genus is an extremely large and heterogeneous group of organisms, and this is reflected in their lipid composition. Much of the information in this area has been obtained by Wilkinson and his colleagues during a systematic investigation of the cell walls of those Pseudononus species sensitive to EDTA. Unfortunately, the lipid compositions reported are also those of the cell walls and therefore may not be representative of the total lipids and also makes comparisons with other reports difficult. P. diminuta, P. maltophilia, and P . rubescens all have unusual lipid compositions, and it is therefore interesting to note that their taxonomic position has already been questioned ( Wilkinson, 1967). All three organisms contain unusual glycosyl diglycerides with glucose and glucuronic acid as the constituent sugars. P. diminuta is notable for the absence of phosphatidylethanoldmine although it must be stressed that this is from the cell wall, not necessarily the cytoplasmic membrane. Wayne Moss et al. (1972) have carried out a survey of the fatty acids, including those of the Iipopolysaccharides, of ten representative species of Pseudomonas and have noted significant differences particularly in the type of hydroxy fatty acid present. This is a good example of the value of fatty acid “spectra” in grouping members of a genus. Although much more work is necessary, it is becoming apparent that lipid analysis will be able to contribute significantly to the classification of this important group.
B. EUBACTERIALES The Order Eubacteriales is another very large and diverse group of bacteria which may be conveniently divided into two on the basis of the gram stain. Where information is available, bacteria belonging to gram-negative families, e.g., Azotobacteraceae, Rhizobiaceae, Enterobacteriaceae and Neisseriaceae all have qualitatively very similar lipid compositions and are not readily distinguishable one from another. Most of these organisms have a relatively simple lipid composition the major component being phosphatidylethanolamine which can account for up to 90% of the total lipid. Phosphatidylcholine has been detected in Azotobacter agilis, Agrobacterium tumefaciens, Rhizobium japonicum and is the major lipid in BrmceEla abortus. Glycolipids are usually absent in this group of gram-negative organisms. Bacteroides meluninogenim has a very unusual lipid pattern composed mainly of phosphosphingolipids. Whether this is typical of this genus remains to be established.
100
NORMAN SHAW
The lipids of many genera of the gram-positive families have been investigated, and several genus-specific features have emerged. Undoubtedly the most significant of these is the presence of diglycosyl diglycerides. Although glycolipids are known to be present in many micrococci, only the dimannosyl diglyceride in M . lysodeikticus has been characterized, and information on other members of this genus would be of interest, particularly because staphylococci contain a diglucosyl diglyceride. Which glycolipid would those organisms which fall midway between these two overlapping genera possess? Perhaps both? The phospholipid composition of some micrococci show minor variations, and two species have been shown to contain phosphatidylinositol, which is common to higher bacteria. Here is an area obviously in need of further investigation. The lipid composition of Micrococcus dentrificans, e.g., the presence of phosphatidylethanolamine and phosphatidylcholine each possessing unsaturated fatty acids is quite atypical and supports the suggestion that this organism is the same as Thiobacillus novellus (in Breed et al., 1957, p. 463). The lipid composition of Spmosarcina urae also supports its reclassification in the genus Bacillus (Whiteside et al., 1971). The Lactobacillaceae are differentiated from the other families of gram-positive bacteria by their fatty acid composition, which is similar to that of many gram-negative organisms in having large quantities of unsaturated and cyclopropane fatty acids. The two most common genera, Streptococcus and Lactobacillus, have different glycolipids; all streptococci have the a-diglucosyl diglyceride, whereas the lactobacilli have the galactosylglucosyl diglyceride. It has been suggested that Lactobacillus bifidus should constitute a separate genus Bifidobacterium, and the lipids of the latter are certainly very different from those of lactobacilli. Bifidobacterium bifidum has a glycolipid containing galactofuranose and an unusual polyglycerophosphatide of ill-defined structure. The genus Propionibacterium possesses a unique glycolipid, diacylinositol mannoside, which has not yet been found in any other organism and is the only glycolipid known which contains inositol. It is structurally related to the phosphatidylinositol mannoside present in many corynebacteria and mycobacteria, but the presence of these phospholipids in Propionibacterium has yet to be conclusively demonstrated. With the exception of Erysipelothrix, considerable information is available on genera of the Cornynebacteriaceae, and they may be readily distinguished. The lipids of Listeria monocytogenes have attracted considerable attention, as it was originally thought that the “monocyte-producing agent” was a lipid, a view no longer held. This organism contains a galactosylglucosyl diglyceride identical to that found in Lactobacillus and the transfer of this genus into the Lactobacillaceae has been sug-
CHEMOTAXONOMY OF BACTERIAL LIPIDS
101
gested. Microbacterium lacticum contains a dimannosyl diglyceride which also probably occurs in small amounts in M . thermosphactum, the classification of which has been questioned by some authors who consider it more related to the Lactobacillaceae. Despite the presence of phosphatidylethanolamine, its lipid composition is more akin to the Corynebacteriaceae; branched-chain fatty acids predominate and unsaturated acids are absent. Dimannosyl diglycerides have not been found in any Lactobacillaceae. Two strains of the gram variable CeZZu2omona.s species C . jimi and C. biazotea have been examined and have similar compositions reminiscent of the bacilli; phosphatidylethanolamine and p-diglucosyl diglyceride are both present. Arthrobacter is characterized by the presence of phosphatidylinositol and two diglycosyl diglycerides, digalactosyl and dimannosyl diglyceride. The co-occurrence of two unrelated diglycosyl diglycerides is a phenomenon as yet unique to Arthrobacter. The corynebacteria have received considerable attention probably because of the presence of the “diphtheria bacillus” in this genus. C. diphtheriae i s closely related to the mycobacteria in that both have very complex cell walls of similar constitution containing free fatty acids of considerable complexity. Little information is available concerning the distribution of these “corynemycolic acids” in other corynebacteria. They have been detected in C. hofmanii but are absent from many other corynebacteria. C. diphtheriae also contains diacyltrehalose, “cord-factor,” commonly found in mycobacteria but so far not reported in any other member of this genus. Another distinguishing feature of C. diphtheriw is the presence of appreciable quantities of unsaturated fatty acids esterified to the phospholipids whereas other corynebacteria have predominantly branched-chain fatty acids. Although acylated glucoses are prevalent, glycolipids of the diglycosyl diglyceride type have been found only in C. aquaticum. This organism is also unusual in containing phosphatidylethanolamine. Phosphatidylinositol mannosides have been found in many corynebacteria. The anaerobic C . acnes has also been classified as Propionibacterium acnes. Recent opinion, however, favors the former classification and the lipid composition is in agreement with this view. c. acnes does not possess the diacylinositol mannoside found in Propionibacterium, and moreover the ratio of is0 C, to anteiso C, fatty acids is very much higher in C. acnes than in Propionibacterium species. However no evidence could be found for the presence of phosphatidylinositol mannosides, which are prevalent in most corynebacteria but are not found in any other genus of the ,Corynebacteriaceae. Much of the above confirms the already acknowledged unsatisfactory state of the taxonomy of the corynebacteria, and as yet no clear picture is emerging except
102
NORMAN SHAW
to note that C. diphtheriae is reasonably distinct from other members of the genus. The lipids of the spore-forming Bacillaceae are extremely characteristic. Members of the genus Bacillus are characterized by the presence of a p-diglucosyl diglyceride and phosphatidylethanolamine. The amount of the latter present is usually less than that found in gram-negative bacteria. A unique glucosaminyl diglyceride has been found in one strain of B. megatherium together with a glucosamine-containing phospholipid. The fatty acid “spectra” of many representative bacilli show minor differences. Less information is available on Clostridium species, but this genus also contains phosphatidylethanolamine. Glycolipids have been observed, but are not well characterized. High concentrations of plasmalogens are a characteristic feature of Clostridium.
C . ACI~NOMYCETALES The lipids of the Order Actinomycetales are the most complex found in bacteria and have been the subject of intensive studies. Phospholipids, glycolipids, complex fatty acids, sulfolipids, lipopolysaccharides, mycosides, and waxes make up the lipid content of these organisms (Lederer, 1967; Goren, 1972). The most important class of lipids, however, from a taxonomic point of view are the high molecular weight fatty acids known collectively as “mycolic acids.” They are found as the free acids in the cell wall or esterified to wall polymers and as the apolar constituents of glycolipids. Those present in mycobacteria, mycolic acids, are the most complex, with carbon numbers around 80; those present in nocardia, nocardic acids, are smaller, with carbon numbers around 50, and are thus intermediate between the my colic acids and the corynemycolic acids found in C. diphtheriae. Streptomyces do not contain these large fatty acids, but apparently they have a-hydroxy acids of normal length. Nocardia and rapidly growing mycobacteria are extremely difficult to differentiate by the usual morphological and physiological tests. Moreover their GC ratio and wall composition are very similar. The identification of mycolic or nocardic acids is therefore of great importance, and Lechevalier et al. (1971) have used this method to classify many mycobacteria and nocardia. Thus they were able to show that M . rhodochrous contained nocardic acids and therefore should be reclassified as N . rhodochrous. A rapid method based upon solubilities for differentiating mycolic and nocardic acids has been described (Kanetsuda and Bartoli, 1972), and it is not necessary to carry out a full chemical structural elucidation. The latter is an extremely complex process which requires all the modern techniques available to the organic chemist. However such studies have shown that, within the genus Myco-
CHEMOTAXONOMY O F BACTERIAL LIPIDS
103
bacterium, species specificity is shown in mycolic acid structure. Thus M. tuberculosis, M . avium, M . smegmutis and M . phlei all have different mycolic acids (Lederer, 1971). However, these differences cannot be routinely used for classification until simpler methods are available for their characterization. The Mycobacterium, Nocardia, and Streptomyces all contain the complex phosphatidylinositol mannosides and phosphatidylethanolamine. These phospholipids, however, do not contain mycolic acids as their apolar constituents but have the normal straight-chain and unsaturated fatty acids. One branched-chain fatty acid, 10-methyl octadecanoic acid ( tuberculostearic acid) is a characteristic component of these lipids. Diacyl trehalose, cord-factor, has been found in Mycobacterium and Nocardia, but not yet in Streptomyces. D. MYCOPLASMATALES The mycoplasma are gram-negative owing to the absence of a normal cell wall but, like the halobacteria, their lipid composition is more akin to that of gram-positive bacteria. These organisms have been very popular with microbial biochemists interested in membrane phenomena, as they show a marked ability to incorporate exogenous lipid into their membranes thereby enabling studies of the influence of fatty acid composition on membrane permeability to be made. Many mycoplasma will not grow on a completely synthetic, lipid free, medium so it is impossible to determine their true lipid composition. The most detailed studies have been carried out on Acholephma laidlawii whose lipids closely resemble those of streptococci. Appreciable quantities of the a-diglucosyl diglyceride are present and the structurally related glycophospholipid both of which are also found in streptococici. Phosphatidylglycerol and its amino acid ester derivatives have also been isolated. The major glycolipid of an avian strain of Mycoplasma is a triacyl glucose. VI.
Conclusions
There are clear, well defined patterns emerging from studies on lipid composition as a guide to the classification of bacteria. Although many anomalies have already been observed, it is significant that in nearly all cases the taxonomy of the organism in question has previously been disputed using other criteria. Assuming no prevous knowledge of bacterial classification, except perhaps the gram stain, the classification of bacteria solely on lipid composition would in effect produce a scheme which is basically very similar to the one presently accepted. There are obviously very large groups of organisms for which information
104
NORMAN SHAW
about their lipid composition is virtually nonexistent. It is the hope of the author that this review will stimulate those microbiologists presently studying these organisms to undertake such studies, and that this method will become accepted as a valuable guide to bacterial taxonomy.
REFERENCES Abel, K., deschmertzing, H., and Peterson, J. L. (1963). 1. Bacteriol. 85, 1039. Ambron, R. T., and Pieringer, R. A. (1971). J.-Bid.Chem. 246,4216. Ames, G. F. ( 1968). J . Bacteriol. 95, 833. Asselineau, J. ( 1966 ) . “The Bacterial Lipids.” Holden-Day, San Francisco, California. Azuma, I., Nayasuga, T., and Yamamura, Y. (1962). J. Biochem. (Tokyo) 52, 92. Ballio, A., Barcellona, S., and Boniforth, L. (1965). Biochem. J. 94, 11C. Barridge, J. K., and Shively, J. M. (1968). J . Bacteriol. 95, 2182. Batt, R. V., Hodges, R., and Robertson, J. G. (1971). Biochim. Biophys. Acta 239, 368. Baumann, N. A., Hagen, P.-O., and Golfine, H. (1965). J . B i d . Chem. 240, 1559. Bergelson, L. D., Batrakov, S. G., and Pilipenko, T. V. (1970). Chem. Phys. Lipids 4, 181. Bishop, D. G., Rutberg, L., and Samuelsson, B. ( 1967). EUT.J. Biochem. 2, 448. Breed, R. S., Murray, E. G. D., and Smith, N. R. (1957). “Bergey’s Manual of Determinative Bacteriology,” 7th ed. Williams & Wilkins, Baltimore, Maryland. Brennan, P. J. (1968). Biochem. J . 109, 158. Brennan, P. J., and Ballou, C. E. ( 1967). J. Biol. Chem. 242, 3046. Brennan, P. J., and Ballou, C. E. (1968). Biochem. Biophys. Res. Commun. 30. 69. Brennan, P. J., and Lehane, D. P. (1969). Biochim. Biophys. A d a 176,675. Brennan, P. J., and Lehane, D. P. (1971). Lipids 6, 401. Brennan, P. J., Flynn, M. P., and Griffin, P, F. S. ( 1970a). FEBS Lett. 8,322 Brennan, P. J., Lehane, D. P., and Thomas, D. W. (1970b). Eur. J. Biochem 13, 117. Brian, B. L., and Gardner, E. W. (1968). J. Bacterial. 96, 2181. Brundish, D. E., and Baddiley, J. ( 1968). Carbwhyd. Res. 8,308. Brundish, D. E., Shaw, N., and Baddiley, J. (1965). Biochem. J. 97, 158. Brundish, D. E., Shaw, N., and Baddiley, J. (1967a). Biochem. J. 104,205. Brundish, D. E., Shaw, N., and Baddiley, J. (196710). Biochem. J . 105,885. Bunn, C. R., and Elkam, G. H. ( 1971). Can. J. Microbiol. 17,291. Carroll, K. K., Cutts, J. H., and Murray, E. G. D. (1968). Can. J. Biochem. 46, 899. Cohen, M. and Panos, C. (1966). Biochemistry 5,2385. Constantopoulos, G., and Bloch, K. (1967). J. Bacteriol. 93, 1788. Cruden, D. L. and Stanier, R. Y. (1970). Arch. MikrobioZ. 72, 115. Cullen, J., Phillips, M. L., and Shipley, G. G. (1971). Biochem. J . 125, 733. Dawson, R. M. C. (1960). Biochem. J. 75, 45. Dittmer, J. C. and Lester, R. L. (1964). J. Lipid Res. 5, 126. Dos Santos Mota, J. M., Op den Kamp, J. A. F., Verheij, H. M., and van Deenen, L. L. M. (1970). J . Bacterial. 104, 611. Etemadi, A. H. (1963). Bull. Soc. Chim. Biol. 45, 1423.
CHEMOTAXONOMY OF BACTERIAL LIPIDS
105
Exterkate, F. A., Otten, B. J., Wassenberg, H. W., and Veerkamp, J. H. (1971). J. Bacteriol. 106, 824. Fischer, W. (1970). Biochem. Biophys. Res. Commun. 41, 731. Fischer, W., and Seyferth, W. (1968). Hopper-Seyler's Z. Physiol. Chem. 349, 1662. Goldfine, H. (1964). J. Biol. Chem. 239,2130. Goldfine, H., and Hagen, P-0. (1968). J. Baderiol. 95, 367. Goldfine, H., and Panos, C. (1971). J. Lipid Res. 12, 214. Gomes, N. F., Ioneda, T., and Pudles, J. ( 1966). Nature (London) 211, 81. Gorchein, A. (1968). Biochim. Biophys. Acta 152, 358. Goren, M. B. ( 1972). Bacteriol. Rev. 36, 33. Gould, R. M., and Lennarz, W. J. (1970). J. Bacteriol. 104, 1135. Hagen, P-O., Goldfine, H., and Williams, P. J. le B. (1966). Science 151, 1543. Hancock, A. J., and Kates, M. (1972). Chem. Phys. Lipids 8, 87. Hancock, I. C., and Meadows, P. M. (1969). Biochim. Biophys. Acta 187, 366. Hirayama, 0. ( 1968). Agr. Biol. Chem. 32, 34. Houtsmuller, U. M. T., and van Deenen, L. L. M. (1985). Biochirn. Biophys. Acta 106, 564. Hunter, M. I. S., and Thirkell, D. ( 1971). J. Gen. Microbiol. 65, 115. Huston, C. K., Albro, P. w., and Grindley, G. B. ( 1965). J. Bacteriol. 89,768. Ikawa, M. (1967). Bactem'ol. Rm. 31,54. Ioneda, I., Lederer, E., and Rozanis, J. ( 1970). Chem. Phys. Lipids 4, 375. Ishizuka, I. and Yamakawa, T. (1968). J. Biochem. (Tokyo) 64,13. Jauch, R., Schiele, U., and Lynen, F. (1960). Arch. Mikrobiol. 72, 186. Joyce, G. H., Hammond, R. K., and White, D. C. (1970). J. Bacteriol. 104, 323. Jurtshuk, P. and Schlech, B. A. (1969). J. Bacteriol. 97, 1507. Kaneda, T. ( 1967). J. Bacteriol. 93, 894. Kaneda, T. ( 1968). J. Bucteriol. 95, 2210. Kaneshoiro, T., and Marr, A. G. (1962). J. Lipid Res. 3, 184. Kanetsuna, F., and Bartoli, A. (1972). J. Gen. Microbiol. 70, 209. Kanfer, J., and Kennedy, E. P. (1963).3. Biol. Chem. 238,2919. Kates, M. (1964). Aduan. Lipid Res. 2, 17. Kates, M., Adams, G. A., and Martin, S. M. (1964). Can. J . Biochem. 42, 461. Kates, M., Wassef, M. K., Kushner, D. J, and Gibbons, N. E. (1966). Biochemistry 5, 4092. Kates, M., Palameta, B., Perry, M. P., and Adams, G. A. (1967). Biochim. Biophys, Acta 137, 213. Kaufmann, B., Kundig, F. D., Distler, J., and Roseman, S. (1965). Biochem. Biophys. Res. Commun. 18, 312. Khuller, G. K., and Brennan, P. J. (1972a). Biochem. J. 127,369. Khuller, G. K., and Brennan, P. J. (1972b). J. Gen. Microbiol. 73, 409. Kimble, C. E., McCollough, M. L., Paterno, V. A., and Anderson, A. W. ( 1969). Appl. Microbiol. 18, 883. Kimura, A., Kawanami, J., and Otsuka, H. ( 1967). J. Biochem. (Tokyo) 62,384. Kunsman, J. E. (1970). J. Bacteriol. 103, 104. Kutaoka, T., and Nojima, S. (1967). Biochim. Biophys. Acta 144, 681. Laneelle, M. A., and Asselineau, J. (1968). C. R. Acud. Sci., Ser. D 266, 1901. Laneelle, M. A., Asselineau, J., and Castelnuovo, G. (1965). Ann. Inst. Pasteur, Paris 108, 69. Laneelle, M. A., Asselineau, J., and Castelnuovo, G. (1968). Ann. Inst. Pusteur, Paris 114, 305. Lang, D. R., and Lundgren, D. G. (1970). J. Bucteriol. 101,483. ~
106
NORMAN SHAW
Lechevalier, M. P., Horan, A. C., and Lechevalier, H. (1971). J . Bacteriol. 105, 313. Lederer, E. (1967). Chem. Phys. Lipids 1, 294. Lederer, E. (1971). Pure Appl. Chem. 25, 135. Lee, Y. C., and Ballou, C. E. (1965). Biochemistry 4, 1395. Lennarz, W. J. (1970). In “Lipid Metabolism” (S. J. Wakil, ed.), pp. 154-184. Academic Press, New York. Lennarz, W. J., and Talemo, B. (1966). J. Biol. Chem. 241,2707. Levin, R. A. ( 1971). J . Bacteriol. 108, 992. Livermore, B. P., and Johnson, R. C. (1970). Biochim. Biophys. Acta 210, 315. Luderitz, 0. (1970). Angew. Chem., Int. Ed. Engl. 9,649. MacDougall, J. C., and Phizackerley, P. J. R. (1969). Biochem. J. 114, 381. Macfarlane, M. G. (1961a). Biochem. J. 79, 4P. Macfarlane, M. G. (1961b). Biochem. J. 80, 45P. Macfarlane, M. G. (1962a). Biochem. 1. 82, 40P. Macfarlane, M. G. ( 196213). Nature (London) 196, 136. Makula, R. A, and Finnerty, W. R. (1970). J . Bucteriol. 103, 348. Marinetti, G. V. (1962). J. Lipid Res. 3, 1. Marinetti, G. V. ( 1967). “Lipid Chromatographic Analysis,” Dekker, New York. Marshall, C. L., and Brown, A. D. (1968). Biochem. J . 110,441. Matches, J. R., Walker, H. W., and Ayres, J. C., (1964). 1. Bacteriol. 87, 16. Matsumoto, M., Tamiya, K., and Koizumi, K. (1971). J. Biochem. (Tokyo) 69, 617. Maurice, M. T., Vacheron, M. J., and Michel, G. ( 1971). Chem. Phys. Lipids 7, 9. Minnikin, D. E., and Abdolrahimzadeh, H. (1971). J . Chromutogr. 63, 452. Miquel, A.-M., Ginsburg, H., and Asselineau, J. (1963). Bull. Soc. Ghim. B i d . 45, 715. Modak, M. J., Nair, S., and Venkataraman, A. (1970). J . Gen. Microbiol. 60, 151. Moss, C. W., Dowell, V. R. Lewis, V. J., and Schekter, M. A. (1967). J. Bacteriol. 94, 1300. Nesbitt, J. A, and Lennarz, W. J. (1965). J . Bacteriol. 89, 1020. Okazaki, H., Sugino, H., Kanzaki, T., and Fukuda, H. (1969). Agr. Biol. Chem. 33, 764. Okuyama, H., Kankura, T., and Nojima, S. (1967). J . Biochem. (Tokyo) 81, 732. O’Leary, W. M. (1967). “The Chemistry and Metabolism of Microbial Lipids,” World Publ. Co., Cleveland, Ohio. Olsen, R. W., and Ballou, C. E. ( 1971). J. Biol. Chem. 246, 3305. 00,K. C., and Lee, Y. C. ( 1972). J. Biochem. (Tokyo) 71, 1081. Op den Kamp, J. A. F., Houtsmuller, U. M. T., and Van Deenen, L. L. M. ( 1965). Biochim. Biophys. Actu 106, 438. Op den Kamp, J. A. F., Redai, I., and van Deenen, L. L. M. (1969a). J . Bacteriol. 99, 298. Op den Kamp, J. A. F., van Deenen, L. L. M., and Tomasi, V. ( 1969b). In “Structural and Functional Aspects of Lipoproteins in Living Systems” (E. Tria and A. M. Scanu, eds.), p. 227. Academic Press, New York. Oshima, M., and Yamakawa, T. (1972). Biochem. Biophys. Res. Commun. 49, 185. Otsuka, S. I., and Shiio, I. (1968). J . Gen. Appl. Microbiol. 14, 135. Park, C. E., and Berger, L. R. (1967). J . Bacteziol. 93, 221. Peleg, E., and Tietz, A. ( 1971). FEBS Lett. 15, 309. Phizackerley, P. J. R., MacDougall, J. C., and Francis, M. J. 0. (1966). Biochem. J., 99, 21c.
CHEMOTAXONOMY OF BACTERIAL LIPIDS
107
Phizackerley, P. J. R., MacDougall, J. C., and Moore, R. A. (1972). Biochem. J. 126, 499. Pieringer, R. A. (1968). J. Biol. Chem. 243, 4894. Plackett, P. ( 1967). Biochemistry 6, 2746. Prome, C., Lacave, C., and Laneelle, M. A. (1969). C . R. A c d Sci., Ser. c 269, 1664. Raines, L. J., Wayne Moss, C., Farshtchi, D., and Pittman, B. (1968). J. Bacteriol. 96, 2175. Randle, C . L., Albro, P. W., and Dittmer, J. C. (1969). Biochim. Biophys. Acts 187, 214. Rizza, V., Tucker, A. N., and White, D. C. (1970). J. Bacterial. 101, 84. Romano, N., Smith, P. F., and Mayberry, W. R. ( 1972). J. Bacteriol. 109, 565. Rothfield, L., and Romeo, D. (1971). Bacteriol. Reu. 35, 14. Saito, K., and Mukoyama, K. ( 1970). J. Biochem. (Tokyo) 69,83. Samuels, S. B., Moss, C. W., and Weaver, R. E. (1973). J. Gen. Microbiol. 74, 275. Senn, M., Iondea, T., Pudles, J., and Lederer, E. (1967). Eur. J. Biochem. 1, 33. Shaw, N. (1968a). Biochim. Biophys. Acta 152,427. Q 435. Shaw, N., (196813). Biochim. Biophys. A C ~164, Shaw, N., (1970). Bacterial. Reu. 34, 365. Shaw, N., and Dinglinger, F. (1968). Biochem. J. 112, '769. Shaw, N, and Stead, D. (1970). J . Appl. Bacterial. 33, 470. Shaw, N., and Stead, D. (1971). J. BQcteTiOl.107, 130. Shaw, N., and Stead, A. (1972). F E B S Lett. 21, 249. Shaw, N., Smith, P. F., and Koostra, W. L. (1968a). Biochem. J. 107,329. Shaw, N., Heatherington, K., and Baddiley, J. (1968b). Biochem. J. 107, 491. Shaw, N., Smith, P. F., and Verheij, H. M. (1972). Biochem. J. 129, 167. Shively, J. M., and Benson, A. A. ( 1967). J. Bncteriol. 94, 1679. Short, S. A., and White, D. C. (1970). J. Bacteriol. 104, 126. Short, S. A., and White, D. C. (1972). J. Bacterial. 109, 820. Short, S. A., White, D. C., and Aleem, M. I. H. ( 1969). J. Bacteriol. 99, 142. Smith, P. F., and Henrikson, C. V. (1965). J. Lipid Res. 6, 106. Smith, P. F., and Mayberry, W. R. (1968). Biochemistry 7, 2706. Steiner, S., Burnham, J. C., Conti, S. F. and Lester, R. L. (1970a). J. Bacteriol. 103, 500. Steiner, S., sojka, G. A., Conti, S. F., Gest, H., and Lester, R. L. (1970b). Biochim. Biophys. A C ~203, Q 571. Sud, I. J., and Feingold, D. (1970). 3. Bacteriol. 104, 289. Thiele, 0. W., (1971). Experientia 27, 1268. Thiele, 0. W., Busse, D., and Hofmann, K. ( 1968). EUT.J. Biochem. 5,513. Toon, P., Brown, P. E., and Baddiley, J. (1972). Biochem. J. 127, 399. Tornabene, T. G., and Ogg, J. E. (1971). Biochim. Biophys. Acts 239, 133. Uchida, K., and Mogi, K. ( 1972). J. Gen. Appl. Microbiol. 18, 109. Utakami, C., and Umetani, K. (1988). Biochim. Biophys. Act0 164, 64. Vacheron, M. J., Guinand, M., Michel, G., and Ghuysen, J.-M. (1972). EUf. J. Biochem. 29, 156. [email protected] J. H. (1970). Biochim. Biophys. Acta 210, 267. Vilkas, E., and Rojas, A. (1964). Bull. Soc. Chim. B i d . 46, 689. Vorbeck, M. L., and Marinetti, G. V. ( 1965). J. Lipid Res. 6, 3. Walker, R. W., and Bastl, C. P. (1967). Carbohyd. Res. 4, 49. Walker, R. W., and Fagerson, P. (1965). Can. J. Microbiol. 11, 229. Ward, J. B., and Perkins, H. R. ( 1968). Biochem. J. 106, 391. Wayne Moss, C., and Lewis, V. J. (1967). Appl. Microbiol. 15, 390.
108
NORMAN SHAW
Wayne Moss, C., Kellog, D. S., Farshy, D. C., Lambert, M. A., and Thayer, J. D. ( 1970). J. Bacteriol. 104, 63. Wayne Moss, C., Samuels, S. B., and Weaver, J R . E. (1972). Appl. Microbiol. 24, 596. Welby-Gieusse, M., Laneelle, M. A., and Asselineau, J. (1970). EUT. J. Biochem. 13, 164. Wells, M. A., and Dittmer, J. C. (1963). Biochemistry 2, 1259. Welsh, K., Shaw, N., and Baddiley, J. (1968). Biochem. J. 107, 313. White, D. C. (1968).J. Bucteriol. 96, 1159. White, D. C., and Frerman, F. E. ( 1967). J. Bacteriol. 94, 1854. Whiteside, T. L., de Siervo, A., and Salton, M. J. R. (1971). J. Bacteriol. 105, 957. Wilkinson, B. J., Morman, M. R., and White, D. C. (1972). J. Bacteriol. 112, 1288. Wilkinson, S. G. (1967).J. Gen. Microbiol. 47, 67. Wilkinson, S . G. (1968a). Biochim. Biophys. Acta 152, 227. Wilkinson, S. G. (196813). Biochim. Biophys. Actu 164, 148. Wilkinson, S. G. (1969). Biochim. Biophys. Acta 187, 492. Wilkinson, S. G. (1970). J. Bacteriol. 104, 1035. Wilkinson, S. G. (1972). Biochim. Biophys. Acta 230, 1. Wilkinson, S. G., and Bell, M. E. ( 1971). Biochim. Biophys. Acta 248, 293. Wilkinson, S. G., Galbraith, L., and Lightfoot, G. A. (1973). Eur. J. Biochem. 33, 158. Wood, B. J, B., Nichols, B. W., and James, A. T. (1965). Biochim. Biophys. Acta 106, 261. Yano, I., Furakawa, Y., and Kusunose, M. ( 1968). J. Biochem. (Tokyo) 63, 133. Yano, I., Furakawa, Y., and Kusunose, M. (1969). J. Bucteriol. 98, 124. Yano, I., Furakawa, Y., and Kusunose, M. (1970a). Biochim. Biophys. Acta 202, 189. Yano, I., Furakawa, Y., and Kusunose, M. (1970b). Biochim. Biophys. Acta 210, 105. Yano, I., Furakawa, Y., and Kusunose, M. (1971). J. Gen. Appl. Microbiol. 17, 329.
Fungal Sterols and the Mode of Action of the Polyene Antibiotics
J. M. T. HAMILTON-MILLER Department of Medical Microbiology, Uniuersity of London, Royal Free Hospital, London, Great Britain
.
I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Sterols in the Microbial Kingdom . . . . . . . . . . . . . . . . . . A. Bacteria .................. B. Yeasts and Fungi . .. . . . . . . . . . . . . . 111. Factors Affecting the Sterol Content of Fungal Cells . . A. Aerobiosis . . . .. .... . . . . . . . . . . . . . B. Growth and Morphological Phase . . . . . . . . . . . . . . C. Medium Composition . . . . . . . . . . . . . . . . . . . . . . . . D. Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Physiological Role of Sterols . . . . . . . . . . . . . . . . . . . , . . V. Sterols and the Mode of Action of Polyenes . . . . . . . . . . A. Studies in Solution . . . . . . , . . . . . . . . . . . . . . . . . . . B. Differential Calorimetry . . . . . . . . . . . . . . . . . . . . . . C. Experiments with Monolayers . . . . . . . . . . . . . . . , . D. Experiments with Bilayers . . . . . . . . . , . . . . . . . . . . E. Liposomes ................. F. Studies with Disrupted. Natural Membranes . . . . . . G. Experiments with Whole Cells and Functional Subcellular Fractions . . . . . . . . . . . . . . . . . . . . . . . . . . . . H. Discussion ................. VI. Resistance to the Polyene Antibiotics . . . . . . . . . . . . . . . . VII. The Role of Sterols in Resistance to the Polyenes . . . . . VIII. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . . . .
1.
109 110 110 110 112 112 112 113 113 114 115 116 117 118 119 120 122 123 125 126 128 130 131
Introduction
When I started work on mycological topics in 1970, I was given the brief to investigate the mode of action of polyene antibiotics and resistance to them among pathogenic yeasts. After a few weeks in the library, I came to two conclusions: first, that the mode of action was already known; and second, that resistance did not occur (at least in nature). These same conclusions could very well be reached by anyone else who studied the question for a short time only, and they could be excused for turning to some apparently more fruitful field. However, on more mature reflection, it became apparent that my first conclusion was untrue, and my second (apparently true) of potentially staggering import. Further reading and subsequent and concurrent experimental work showed me that a full study of the mode of action of the polyene antibiotics leads to a biophysical world entirely removed from conventional medical mycology. Experiments designed to probe the mode of action 109
110
J. M. T. HAMILTON-MILLER
of polyenes have revealed, often for the first time, vital information on basic membrane function and structure, and have thus paved the way for fundamental discoveries in this most crucial field. In much the same way, studies on antibiotics that inhibit protein synthesis have vastly increased knowledge on protein synthesis in general. It is gratifying for workers in mycological disciplines, who are often regarded as the Cinderellas of Microbiology, that such experiments are the result of their early and continuing interest in these antibiotics. Problems of drug resistance have been a continual worry to the medical bacteriologist since antimicrobial therapy started in earnest in the 1930's with the introduction of the sulfonamides. Mycologists have been spared this problem, and an explanation of why this might be so is of the greatest interest and importance. A short analysis of this problem is presented toward the end of the present review. Various aspects of sterols in relation to polyene activity are considered here; first, the distribution, nature, and function of sterols is discussed; then their role in the mode of action of the polyenes; and finally the part they have to play in resistance mechanisms in yeasts.
II. Sterols in the Microbial Kingdom
A. BA-IA Sterols are notably absent from bacteria ( Asselineau and Lederer, 1960) with the exception of the nitrogen-fixing Azotobacter chroococcum which has a sterol content of 10 mg/100 gm. Schubert et al. (1964, 1968) were able to isolate minute amounts of cholesterol, campesterol ( a homolog of cholesterol), p-sitosterol, and stigmasterol from Escherichia coli, but the total sterol yield was only 400 pg/lOO gm. Sterols are important in the classification of Mycoplasmataceae, the two genera Mycoplasma and Acholeplasma being distinguished by the ability of the latter to grow in the absence of added cholesterol. However, A. laiduwii is capable of incorporating exogenous sterol, a phenomenon which has helped to adduce important evidence as to the mode of action of the polyene antibiotics (see Section V ) (Feingold, 1965; Weber and Kinsky, 1965).
B. YEASTSAND FUNGI The existence of sterols in the fungal kingdom has been known for many years. The most widely distributed sterol, ergosterol, was so named
FUNGAL STEROLS AND POLYENE ANTIBIOTICS
111
after the mold ergot, from which it was first isolated more than 90 years ago (Tanret, 1879); before the end of the 19th century, ergosterol had also been found in Penicillium and Mucor. As ergosterol is a commercially important natural product, being a precursor of vitamin D, much attention has been paid to finding the best source of this compound. Dulaney et al. (1954) investigated 558 yeast cultures, and found that Sacchuromyces carlsbergensis, S. cerevisiae, and S . ovif ormis contained at least 2% by weight of sterols; brewer’s yeast and ale yeasts were the most consistent in this respect. Appleton e,t al. (1955) screened more than 50 strains from a total of more than 30 genera of molds in all except Neuroand yeasts and found a low sterol content ( <=' 5=' %=' total=' sterol),=' fecosterol,=' an=' unidentified=' c?,=' diunsaturated=' components.=' authors=' that=' fraction=' 8=' 0=' 4=' composed=' esterified=' material.=' adams=' parks=' (1967)have=' s=' .=' ceremae=' resists=' extraction=' conventional=' means=' alkaline=' hydrolysis=' followed=' into=' organic=' solvent)=' ;=' this=' requires=' acid=' treatment=' before=' being=' extractable.=' actual=' used=' for=' commercial=' (e.g.,=' manufacture=' baker’s=' yeast)=' accounts=' 80%of=' sterol.=' laboratory=' conditions,=' however,=' 24(28)-dehydroergosterol=' may=' come=' 74%=' breivik=' owades,='>
112
J . M.
Ill.
T. WMILTON-MILLER
Factors Affecting the Sterol Content of Fungal Cells
A. AEROBIOSIS It is now clearly established that sterol metabolism is dependent upon oxygen availability. The reason for this seems to be that the conversion of squalene to sterols requires oxygen; both the oxidative cyclization of squalene to lanosterol and the conversion of the latter to cholesterol require molecular oxygen. Thus, squalene can be shown to accumulate in yeast cells grown anaerobically, and is converted to sterol when aerobiosis is resumed. Sacchuromyces cerevisiae requires a suitable sterol in order to grow anaerobically; a very wide range of sterols can substitute for the naturally occurring ergosterol. Requirements for such sterols are that the molecule must be planar, have a long C-17 alkyl side chain and a hydroxyl group (Proudlock et al., 19SS). By no means all yeasts, however, are incapable of growing anaerobically; Schixosaccharomyces jnponicus grows very well anaerobically unsupplemented, but anaerobic cells are virtually devoid of ergosterol ( Bulder, 1971). A similar situation obtains for Mucor genevensis, in which squalene accumulates anaerobically (Gordon et al., 1971).
B. GROWTH AND MORPHOLOGICAL PHASE There are several instances of reports of alterations in sterol content, both qualitative and quantitative, occurring during the growth cycle of various yeasts and fungi. In M . genevensis, Gordon et al. (1971) found that ergosterol comprised about 80%of the total sterol content during the exponential phase of growth, and that the stigmasterol content rose to 40% in the stationary phase. Capek and Simek (1972) found that the sterol content (mainly ergosterol) of the dermatophytes Microsporum gypseum and Tdchophyton mentagrophytes increased as the cultures aged. Bianchi (1967) reported a higher proportion of the lipid fraction from juvenile ( logarithmic phase) than from mature yeast cells of Candida albicans was composed of sterol and sterol esters; four sterols were separated by thin-layer chromatography, but only cholesterol was tentatively identified. Between them, these two fractions always outweighed the other lipid components of the cell wall (triglycerides, free fatty acids, and phospholipids). The same author has also compared the sterol contents of C. albicans cells in the yeast form with those of the filamentous phase; although the total sterol content was greater when the organism was in the mycelial form, the proportions of the various lipid fractions were markedly different. Anaerobiosis, which markedly decreases the ergosterol content of M. geneuensis cells (see
FUNGAL STEROLS AND POLYENE ANTIBIOTICS
113
previous section), is one of the factors which predisposes toward the yeastlike development of this species (Gordon eat al., 1971). However, it is the opinion of these workers that this decrease in sterol, and the concomitant alteration in fatty acid content, is an effect rather than the cause of the morphological change. C. MEDIUM COMPOSITION Inasmuch as a certain amount of dimorphism may be induced by alteration of medium [e.g., use of a liquid medium to obtain the filamentous form of C. albicans: (Bianchi, 1967; Hamilton-Miller, 1972a)], the physical composition of medium may be said to affect the sterol content of yeast cells, but more attention in this respect has been paid to the affects of different sources of carbon and nitrogen. Thus, various authors have claimed the following carbon sources to be the most favorable for the production of sterols by yeast: maltose, glucose, sucrose, and pyruvate; similarly, many nitrogen sources have also been investigated, NH,’ being favored by some, peptone by others (see El-Refai and ElKady, 1968a,b). Concentrations of nutriments may also be crucial. The situation is complicated not only by the multiplicity of sources of nitrogen and carbon available-giving the possibility of virtually limitless permutations and combinations to be tested-and the differing strains and cultural conditions, but also by the fact that ergosteroI is a valuable commercial product and hence what has often been sought, understandably, is maximum yields of sterol per volume of culture rather than maximum yield per gram dry weight of cells. It is thus clear that the true importance of the nature and concentration of nutriments and the production of sterols by yeast and fungi is very difficult to assess. Dulaney (1957) has shown how the ergosterol content of S. cerevisiae may be raised to as much as lo%, w/w; this was done by growing the organisms to maximum cell yield, harvesting and resuspending in fresh medium for a “refermentation” phase during which ergosterol enrichment occurs.
D. TEMPERATURE The true effect of alterations in temperature on the cellular processes in microorganisms is very difficult to assess except by the use of continuous culture. In batch culture, changing the incubation temperature is accompanied by an alteration in growth rate, so that any overall effects may be the result of either the altered growth rate, the altered temperature, or a combination of the two. It should be noted that the above
114
J . M . T. HAMILTON-MILLER
argument also applies, although perhaps not to such an extent, to alterations in nutriments. Hence, from the strictly scientific point of view, one should reject virtually all the studies made on the effects of altering the incubation temperature on sterol production, while bearing in mind that, from an applied point of view (obtaining the maximum overall yields of sterol), such studies are of obvious importance. The problem of dissociating changes in growth rate from alterations in temperature is solved by the use of continuous culture, using a fermenter-a piece of apparatus regrettably much more familiar to scientists in industry than to those in universities. By this technique, the growth rate (dilution rate) of the culture can be adjusted independently of the temperature, so that a constant growth rate can be achieved over a range of temperatures and, conversely, growth rates can be varied at any one desired temperature. An excellent illustration of the application of this technique to the current problem is the work of Hunter and Rose (1972). They found that, although in batch culture lowering the growth temperature had little or no effect on the content of sterols and sterol esters, lowering the growth rate at fixed temperature caused an increase in the amount of sterol esters, whereas decreasing the incubation temperature while maintaining the same growth rate brought about a decrease in total and esterified sterols. Parks and Starr (1963) found that growing S. cerevisiae at 42°C caused a sharp reduction in sterol content. IV.
Physiological Role of Sterols
Sterols seem to have only one major function in the majority of fungi, a structural role in maintaining membrane integrity and permeability. Studies at the molecular level have shown that sterols have a marked condensing effect upon experimental membranes composed of phospholipid, reducing their permeability to polyo1 compounds, such as glucose, glycerol, and erythritol, and to water (see Demel et al., 1972b). Incorporation of cholesterol into the membrane of cells which usually lack sterols was found to decrease the leakiness of the cells ( Child et al., 1969). In one group of fungi, however, the oomycetes, sterols do have an additional role. Members of this group, typified by the genera Pythium and Phytophthora, do not contain sterols under normal cultural conditions (Schlosser et al., 1969); vegetative growth can occur in the absence of sterols, but they are required for sexual reproduction and the formation of normal sporangia and zoospores. Hendrix (1964) found that ergosterol, stigmasterol, phytosterol, and cholesterol allowed the formation of oospores, and that the latter two compounds also increased growth rates. Divalent cations (Lenney and Klemmer, 1966) are additionally
FUNGAL STEROLS AND POLYENE ANTIBIOTICS
115
required for sexual maturation. As they contain no sterols, oomycetes are resistant to the antifungal action of polyenes, but Defago et al. (1969) reported that polyenes in high concentrations ( 0 . 5 1 mg/ml) induced oogenesis in Pythium, and that this effect was reversed by cholesterol. Sterols can be incorporated into the cell membranes of Pythium, and the resulting cells are sensitive to polyenes (Schlosser and Gottlieb, 1966). The general necessity for sterol biosynthesis to fungal growth is shown by the fact that, when S. cerevisiae is grown anaerobically to a state of total ergosterol depletion, no intracellular membranous organelles can be seen (Morpurgo et al., 1964). It is of interest, too, that certain drugs known to inhibit sterol biosynthesis in animals also possess some antifungal activity (Elliott, 1969; Hamilton-Miller, 1973b). It is interesting to speculate whether the fact that oomycetes lack sterol confers any ecological advantage on them over other fungal types. It could be argued that the oomycetes will be at a disadvantage due to a slower growth rate and their inability to reproduce sexually in the absence of sterols, but that they will enjoy a telling advantage in the presence of Streptomyces spp., which produce polyene antibiotics. While there is no doubt that organisms capable of producing polyenes are widespread in the soil (see Hamilton-Miller, 1973a), the question of whether antibiotics are actually produced under natural conditions is still far from resolved (see Hill, 1972).
V.
Sterols and the Mode of Action of Polyenes
It is now firmly established that the presence of sterols is a prerequisite for sensitivity to polyene antibiotics. The evidence for this has been reviewed frequently, most recently by the present author (HamiltonMiller, 1973a), and it would not serve any useful purpose in the present context to repeat it. At the multicellular level, the action of polyenes on a wide variety of organisms has been studied: mycoplasmas, yeasts and filamentous fungi, chloroplasts, protozoa, flatworms, flies and moths, amphibian tissues (muscle fibers, skin, and bladder), chick intestine, and various mammalian tissues. It is at the fractionated-cell and molecular levels, however, that the greatest strides have been made recently toward an understanding not only of the intimate details of the nature of the interaction of polyene and sterol, but also of membrane function as a whole. It is the results of this type of experiment that will be reviewed here. These studies, most of which emanate from the Utrecht Laboratory of L. L M. van Deenen, are chiefly biophysical in nature, and interpretation of the findings made may offer some difficulty to the reader who
116
J . M. T. HAMILTON-MILLER
is more usually occupied with medical or microbiological matters. It will therefore be appreciated that the following summary is much condensed and simplified. Various systems, at different levels of organization, have been used in the studies, and the order of review (in ascending order of complexity), although not chronologically correct, appears apt. It is noticeable that the majority of these studies involve filipin, which has been conclusively shown to be much the most active of the common polyene antibiotics as far as membrane activity is concerned in in vitro experiments. Much of the work was done before it was reported (Bergy and Eble, 1968) that filipin was a highly heterogeneous mixture composed of several compounds with similar molecular structures but distinct properties. The main component of the complex, filipin 111, comprises about 53%of the total. Some of the latter work was done using filipin 111, but some workers have continued to use the complex. Thus, all results with “filipin” may hot be strictly comparable. Some pertinent differences between the biological properties of the various filipin fractions are discussed by Sessa and Weissmann ( 1968b).
A. STUDIESIN SOLUTION The earliest work carried out on the problem of sterol-polyene interaction was based on the observation of changes in absorption spectra that occurred when sterols were added to aqueous preparations of polyenes. Thus, Lampen et al. (1960) and Gottlieb et al. (1961) deduced that cholesterol and ergosterol combined with nystatin, antimycoin, and filipin, greatest spectral changes occurring with filipin. Norman et al. ( 1972a,b) have recently repeated these experiments, and found that filipin, lucensomycin ( Etruscomycin ) , amphotericin B, nystatin, and pimaricin interact with cholesterol, the magnitude of the interaction being in the order given. The precise chemical nature of the sterol greatly affected the degree of action, an optimum effect being obtained with compounds containing a cholestane ring and a A-22 double bond. Thus, it appears that the interaction is hydrophobic. Sterol esters were found to be of low activity in this respect. Polyenes are known to form suspensions of micelles rather than true solutions in aqueous solvents; e.g., Lampen et al. (1959) and Norman et al. (1972b) found that Em for filipin was consequently lower in water than in methanol, and the Beer-Lambert law is not obeyed. Thus, experimental results based on spectrophotometric measurements are open to the criticism that changes in spectra may be due merely to the fact that polyenes are less soluble in dilute aqueous solutions of sterols than in pure water, However, change in micellar size may be argued to be
FUNGAL STEROLS AND POLYENE ANTIBIOTICS
117
a direct effect, and, if specific for a certain sterol, is clearly of potential physiological significance; similarly, where qualitative changes occur the only explanation is interaction. These hypothetical difficulties may be overcome, however, by measuring a parameter which is (unlike extinction coefficient) not concentration dependent. Such a parameter is partial quantum efficiency, measured spectrofluorimetrically. As a criterion for fluorescence appears to be an uncharged molecule, original experiments were carried out using the nonionized pentaenes filipin and Iagosin (Norman et aE., 1972a; Schroeder et al., 1971; Bittman and Fischkoff, 1972). The combination of sterols with these compounds was found in general to decrease fluorescence and partial quantum efficiency, thereby offering direct evidence for the interaction of such compounds as cholesterol, p-stigmasterol, ergosterol, and P-cholestenol with the antibiotics. The intrinsic fluorescence of amphoteric polyenes, such as nystatin, amphotericin B, pimaricin and its homolog lucensomycin, is extremely low. However, it was found that fluorescence increased ( in some cases substantially) when the antibiotics combine with cholesterol and other P-OH sterols (Crifo et al., 1971; Strom et al., 1973a,b; Schroeder et al., 1972; Norman et al., 1972a), thus again providing evidence in favor of the sterol-polyene interaction hypothesis. By means of another fluorimetric parameter, fluorescence polarization, Bittman and Fischkoff (1972) and Strom et al. (1973a), were able to show that when polyene antibiotics combine with the sterols the rigidity of the polyene molecule increases. Schroeder et al. (1972) reported that the stoichiometry of the filipin/cholesterol interaction was 1:1; they also found that filipin may, after recrystallization from organic solvents, exist in forms that react with sterols only after some considerable period of time in aqueous solution.
B. DIFFERENTIAL CALORIMETRY Norman et al. (1973a,b) have investigated the effects of the incorporation of cholesterol into, and the subsequent addition of polyene antibiotics to, dispersions of lecithin. The lecithins used in these cases, l-oleoyI-2-stearoyI-sn-glycero-3-phosphorylcholine and 1,2-dielaidoyl-snglycero-3-phosphorylcholine, each display a phase change at about 13OC which is characterized by a sharp absorption of heat as the lecithin changes from a crystalline to a liquid crystalline structure. The heat is required to melt the hydrocarbon chains. When sterols are incorporated into the lecithin dispersion, the phase change becomes less endothermic (de Kruyff et al., 1972); it is interesting that cholesteroI had a much greater effect on the energy content of this phase transition than did epicholesterol or sterols lacking a 3-OH group or a 17 side chain.
118
J . M.
T. HAMILTON-MILLER
Demel et al. (1972b) further showed that such sterols-which have also been shown to be those that interact the best with the polyenes-also have the greatest effect on the permeability of liposomes (see next paragraph). This system clearly offers scope for the investigation of the interaction of polyenes with sterols, as the combination of the two will result in a greater or lesser negation of the effect of the sterol on the heat content of the phase transition of lecithin. By differential scanning calorimetry, Norman et al. (1972a,b) have shown that filipin, lucensomycin, pimaricin, nystatin, and amphotericin B all interact with cholesterol, and that the antibiotics by themselves had no effect on the phase transition of lecithin. Under the conditions used (viz. highly concentrated solutions ), the stoichiometry of the interaction varied from 1.2 to 7.9 molecules of cholesterol complexed per polyene molecule (Norman et al., 1972a).
C. EXPERIMENTS WITH MONOLAYERS The interaction of polyenes and sterols can also be quantitated by measuring changes in surface pressure occurring when polyenes ( which behave as soluble compounds under the experimental conditions obtaining) are injected beneath monolayers of lipids or lipid mixtures at varying initial surface pressures. These changes are brought about because interactions alter the spatial orientation of the lipid molecules with respect to the air-water interface; Demel et al. (1967) discovered that cholesterol had a condensing effect on monolayers of certain lecithins. Demel et al. (1965), using this technique, found that nystatin was less active than filipin, which reacted more strongly with cholesterol than with ergosterol, and virtually not at all with phospholipids. Monolayers prepared from material extracted from bacterial protoplast membranes were not affected by filipin, in contrast to similar monolayers made from erythrocyte membranes. In an extension of this study, Demel et al. (1968a) found that lucensomycin, amphotericin B, and pimaricin behaved like filipin and nystatin; they also showed that the activity of filipin was decreased if the phospholipid: cholesterol ratio of the monolayer was increased. Filipin at high concentrations interfered with monolayers of lecithin alone, but this interaction, also reported by Sessa and Weissman (1967) using a different system (see below), was not considered to be of physiological significance. The facts that ( i ) monolayers of cholesteryl acetate were virtually unaffected by filipin, (ii) filipin at low concentrations reacted with monoand (iii) 5 M urea, layers of cetyl alcohol (CH,.(CH2),.CH,0H), which is known to disrupt hydrogen bonds, diminishes the effect of filipin on cholesterol monolayers, suggested that, in this heterogeneous system as
FUNGAL STEROLS AND POLYENE ANTIBIOTICS
119
opposed to in solution (see above), hydrophilic bonding may be important in the interaction of sterols with polyenes (see Demel et al., 1972a).
D. EXPERIMENTS WITH BILAYERS Van Zutphen et al. (1966) were the first to exploit the fact that stable bilayer lipid films possess certain dimensional, mechanical, electrical, and permeability properties, which are characteristic of the membranes of living cells, in order to investigate the sterol-polyene interaction. Such bilayer films had been developed a few years previously; they are formed from solutions of lipids across a small hole in a membrane, contained in a small cell. The cell is then filled with an appropriate solution, such as 0.1 M NaCl, and the passage of ions across the film can be studied by introducing the marker on one side of the membrane, and measuring its concentration subsequently on the other side. In their early work, van Zutphen et al. (1966) in fact merely studied the macroscopic stability of bilayer films in the presence of filipin; they found that lecithin films were stable for more than 1 hour in the presence of c. 4 x M filipin and nystatin, while bilayers comprised of equal parts of cholesterol and lecithin disrupted after 5 and 25 minutes, respectively, when these polyenes were added. Films made of 10 parts of lecithin to 1 of cholesterol were more stable. The concentration of polyene used in these experiments was about 10 times that required to prevent the growth of Candida albicans (see Hamilton-Miller, 1973a), and some 5000 times greater than that which had an effect on lipid monolayers (see preceding section) ; this discrepancy is doubtless due to the rather crude criterion of activity. In a later, more extended study, van Zutphen et al. (1971) investigated the effects of filipin, nystatin, lucensomycin, pimaricin, and amphotericin B on bilayers of various compositions. In addition to showing the disrupting effects of the antibiotics, which roughly paralleled their physiological activities in terms of hemolytic abilities (Kinsky, 1963), it was found that the electrical dc resistance of lecithinlcholesterol black films could also be markedly decreased ( u p to 1000-fold) by polyene treatment. Bilayers exposed to filipin (components 11, 111, or the complex) became selectively permeable to Ca'+ ions, while those treated with nystatin or amphotericin B became, under appropriate conditions, anion selective, although being of increased permeability to cations as well. High concentrations of filipin (0.1 mM) or nystatin ( 10 mM) also slightly lowered the resistance of, and disrupted, bilayers composed of lecithin alone. Lippe (1968) found that the permeability of mixed bilayers to an uncharged molecule, thiourea, was increased, while the film remained completely stable, 10fold and 4-fold by 1 p M and 0.2 p M amphotericin B, respectively [cf.
120
J . M. T. HAMILTON-MILLER
antiyeast activity of amphotericin B c. 0.5 pM ( Hamilton-Miller, 1973a)1. Andreoli and Monahan (1968) and Andreoli et al. (1969) used bilayers made from sheep erythrocyte lipids, and their observations on the action of nystatin and amphotericin B on such films are in general agreement with those made by van Zutphen et al. (1966, 1971)-thus, electrical resistance was decreased one millionfold, permeability to C1- and CH,.COO- was increased greatly and to Na+, K+, and Li+ to a lesser extent. On the basis of their findings, they calculated that amphotericin B causes the formation in this type of membrane of pores of equivalent radii 0.7 to 1 nm. The results of Finkelstein and Cass (1968) are also generally in line with those listed above; they made the further observations that, in nystatin-treated bilayers, membrane conductance decreased extraordinarily with increasing temperature, the process having a Ql0 of about 1000, and also that the conductance was proportional t o the tenth power of the nystatin concentration. Both the latter authors and Andreoli et al. (1968, 1969), however, were unable to detect changes in ion permeability or electrical properties after filipin treatment, although bilayers were readily lysed by this compound. Cass et al. (1970) and Holz and Finkelstein (1970) investigated the changes in permeability to ions, water, and nonelectrolytes in lipid membranes induced by amphotericin B and nystatin, and concluded that the behavior of such systems was consistent with the formation of nonstatic pores, of diameter about 0.8 nm, which coalesce and reform continuously. The permeability to water and small hydrophilic solutes of polyene-treated bilayers approximated closely to that of normal erythrocytes.
E. LIPOSOMES
If 'lecithin is allowed to swell in water, or if solutions (in solvents such as decane or chloroform) of lecithin, dicetylphosphate, and, if desired, sterols are mixed, allowed to evaporate and then suspended in an aqueous menstruum, the result is the formation of spherical structures consisting of several bimolecular layers of mixed lipids separated by aqueous compartments, varying in diameter (according to the method of homogenization) from 0.5 to 50 pm (Sessa and Weissmann, 1968a). Such bodies, now generally called liposomes, are very useful as model membranes, and have been widely utilized for the study of the interaction of polyenes and sterols. Liposomes are, by nature, much less permeable to cations than to anions, and are liable to osmotic swelling, as water is freely diffusible through the membranes. The permeability of lecithin liposomes is reduced by the incorporation of cholesterol, but this sterol effect is markedly dependent upon the chemical nature of
FUNGAL STEROLS AND POLYENE ANTIBIOTICS
121
the lecithin (Demel et al., 1968b; de Gier et al., 1968). Cholesterol seems to make the phospholipid molecules pack more tightly, thereby increasing the stability of the membrane. It is, presumably, this packing effect with which polyenes interfere in order to have their effect on permeability. One of the most convenient methods of applying liposomes involves the trapping of a marker such as glucose, Cr042-,or H,PO,-, removal of external marker by dialysis or gel filtration, and measurement of its transfer from the disperse to the continuous phase after treatment of the suspension with a polyene. A disadvantage inherent in the use of liposomes is that high concentrations of antibiotics ( > 10 p M ) are necessary, a fact which gave rise to some early confusion when Weissmann and Sessa (1967) reported that filipin was active against liposomes composed of lecithin only-a cognate phenomenon to that found with bilayers. Reaction also occurred with liposomes made from sphingomyelin, cardiolipin, and phosphatidylcholine, and was not due to the nature of the charge on the liposome (Sessa and Weissmann, 1967). This paradox was resolved when it was found that filipin 11, which comprises 252 of the filipin complex, was responsible for this phenomenon (Sessa and Weissman, 1968b). The results obtained with this techinque using liposomes of lecithin and lecithin /cholesterol in general provide evidence which supports the sterol-polyene interaction theory; thus, using a sensitive assay for glucose, Kinsky et al. (1968) showed that the filipinmediated release of this substance was much more rapid from cholesterol/lecithin liposomes than from lecithin liposomes. Neither they nor Weissmann and Sessa (1967), however, were able to show interaction between either pimaricin or lucensomycin and cholesterol-containing liposomes. Recent experiments by HsuChen and Feingold (1973) indicate the importance of the nature of phospholipid; they found that cholesterol incorporation into egg lecithin liposomes enhanced the sensitivity of the liposome to amphotericin B (as found above), but that the sensitivity of dipalmitoyl lecithin liposomes was in fact decreased by the incorporation of cholesterol. This phenomenon is explained by postulating that the membrane must be in an ordered state for a polyene to cause permeability changes therein, and that cholesterol, while making liposomes of egg lecithin more ordered, decreases the order of dipalmitoyl lecithin liposomes. Liposomes containing sterols have been shown both to bind and to alter the ultraviolet absorption spectra of polyene antibiotics. Norman et al. (1972a) demonstrated this for filipin, lucensomycin, and amphotericin B, but the phenomenon was not apparent using nystatin or pimaricin. The same workers showed that the presence in the sterol of a 3P-OH group is an absolute requirement for interaction with polyenes, in conand 3 keto trast to the situation in solution (see above), where 3 ~ 0 H
122
J . M. T. HAMILTON-MIL,LEXI
sterols, as well as sterol esters, do react, although less well than 3 p O H compounds. Similarly, liposomes of appropriate composition alter the fluorescence properties of polyenes: Drabikowski et al. (1973) and Strom et al. (1973a) reported that the fluorescence intensities of solutions of filipin and lucensomycin, respectively, were enhanced in the presence of liposomes made from cholesterol and lecithin. The finding in respect to filipin is of interest in view of the quenching of fluorescence by “soluble” cholesterol ( see above). Liposomes can also be made from material extracted from naturally occurring membranes. De Kruyff et al. (1972) and McElhaney et al. ( 1973) have indicated that such liposomes have permeability properties very similar to those of the cells from which they were derived-in these cases, A. laidlawii. This model, in the hands of Demel et al. (19721,) and de Kruyff et aE. (1972, 1973), shows quite clearly that the structural requirements for sterols to give optimal reductions in permeability to noneIectrolytes ( e.g., erythritol and glycerol) as well as electrolytes (e.g., Rb’), in both whole cells of A. laidlawii and derived liposomes, are identical with those for maximum interaction with the polyenes. This is indeed an intriguing finding, which indicates why polyenes can have such a devastating effect on cellular permeability, combining most avidly with the very steroIs whose presence most affect permeability.
F. STUDIESWITH DISRUPTED NATURAL MEMBRANES Most of these studies have been carried out using erythrocyte ghosts, which are very easy to prepare in bulk, and are known to contain about 50 mole I% of their total lipid as cholesterol and are thus equivalent to liposomes composed of equal parts of cholesterol and lecithin, or membrane preparations from A. laidlawii. The convenience of the latter is that it may be preconditioned to contain a particular sterol, or none. Crifo et al. (1971) and Strom et ul. (1973b) found that the fluorescence of lucensomycin was enhanced by erythrocyte ghosts; treatment of the ghosts with formaldehyde or extraction with methanoI/chloroform did not remove this effect. The enhancement occurred much more rapidly than when free cholesterol was used. Norman et al. (1972a,b) reported that erythrocyte ghosts alter the absorption spectra of filipin, lucensomycin, and amphotericin B, as did membranes from A. laidluwii enriched with cholesterol; membranes lacking cholesterol, or enriched with the 3a isomer epicholesterol, did not bring about spectral changes. Electron microscope examination of red cells lysed with filipin and negatively stained revealed the presence of pits about 12 nm in diameter (Kinsky et al., 1966). This sort of evidence is fully consistent with the
FUNGAL STEROLS AND POLYENE ANTIBIOTICS
123
theory (Kinsky et al., 1967b) that filipin acts by inducing a lamellar to micellar phase change in membranes. These electron micrographic studies have also been able to show conclusively that the action of saponins, which are hemolytic agents whose activity is blocked by sterols, is entirely different to that of polyenes. A later study (Verkleij et al., 1973), using a freeze-etch technique as opposed to negative staining, showed that the pits were not in fact pores which penetrated the membrane, but were rather areas of aggregated material, presumably composed of filipin-cholesterol complexes. Lesions of an identical type were seen in A. laidlawii membranes and liposomes damaged by filipin. No parallel phenomenon was observed in membranes which had been in contact with amphotericin B, but the postulated size of “pores” produced under these conditions-0.5-1 nm in diameter (Andreoli et al., 1969)-would not be visible after freeze-etching. Membranes from A. Zaidawii, if they contained cholesterol, caused changes in the absorption spectrum of filipin, lucensomycin, and amphotericin B and also bound the antibiotics (Norman et al., 1972a,b). Similar findings were made for filipin interacting with membranes obtained from disrupted rabbit sarcoplasmic reticulum ( Drabikowski et al., 1973); such binding was associated with an increase in the .fluorescence of filipin.
G. EXPERIMENTS WITH WHOLE CELLSAND FUNCTIONAL SUBCELLULAR FRACTIONS 1 . Erythrocytes
Kinsky et al. (1962) were the first to show that polyenes caused hemolysis; filipin was the most active compound, followed by amphotericin B and nystatin. In further experiments, Kinsky (1962, 1963) showed that several polyenes lysed erythrocytes and caused permeability changes to N . crassa protoplasts, but did not affect bacterial protoplasts; the polyenic compound vitamin A (retino1)-which, it should be noted, is not a macrolide-lysed red cells and bacterial protoplasts, but not fungal protoplasts. Kinsky et al. (1967a) correlated the hemolytic and chemical properties of various filipin derivatives, the perhydro compound (obtained by catalytic hydrogenation), the irradiated derivative and the saponified adduct (the lactone ring having been broken by aqueous NaOH ) . All these derivatives had drastically decreased hemolytic and antifungal activities, and were later shown (e.g., Demel et al., 1968a) also to have diminished power to combine with sterols. A further degradation product of filipin (structurally much less altered than the compounds mentioned above), the tetraenic epoxide ( Rickards et al., 1970), is also devoid of sterol-combining power (Schroeder et al., 1972). Sessa
124
J.
M. T. HAMILTON-MILLER
and Weissman (1968b) found that the hemolytic and sterol-combining activities of the four components of filipin ran in parallel. An important point, stressed by Kinsky et al. (1967a), is that it is the ratio of antibiotic to sterol, rather than the absolute concentration of antibiotic, which determines the rate and extent of hemolysis. 2. Mitochondria The function of mitochondria, isolated from a sensitive organism, N . crassa, was not affected by polyenes, although the antibiotics were bound (Kinsky et al., 1965). This lack of sensitivity was attributed to the high phospholipid to sterol ratio in the mitochondria1 membrane (c. 40: 1)) compared to microsomes ( 7 : 1 ) , the latter fraction accounting for the bulk of antibiotics binding in this species. These phospholipid: sterol ratios are virtually identical to those reported in mammalian tissues (Parsons and Yano, 1967), and should be compared with the figures for parenchymatous cell membranes (c. 1.3:1 Dod and Gray, 1968), and erythrocyte membranes ( 1:1) . Weissman et al. (1966) reported that polyenes disrupted lysosomes, but did not affect mitochondria, from rabbit cells. 3. Tissues from Cold-Blooded Animals
Lippe and Giordana (1967) reported a differential effect on amphotericin B on the large and small intestine of Testudo; the mucosal bur not the serosal side of the large intestine became more permeable to thiourea after exposure to the antibiotic. A similar type of effect was noted using toad bladder; Lichtenstein and Leaf (1965) observed that amphotericin B acted on the mucosal side of the bladder only, mimicking to some extent the action of antidiuretic hormone. Leung and Eisenberg (1973) have reported an effect of nystatin on frog skeletal muscle, the result of which is an increase in membrane conductance.
4. Insect Larvae The experiments of Sweeley et al. (1970) and Schroeder and Bieber (1971/1972) have shown that filipin kills the larvae of the house fly and interferes with cholesterol metabolism in the Iarvae of the wax moth. The administration of cholesterol can reverse some of these deleterious effects. The disturbance of phospholipid metabolism reported may not be a direct effect of polyene, but a consequence of the altered cholesterol metabolism.
5 . Microorganisms Experiments elucidating the role of sterols in the mode of action of the polyenes will be described very briefly, as they have been covered
FUNGAL STEROLS AND POLYENE ANTIBIOTICS
125
in detail elsewhere. That exogenously added sterols can protect sensitive organisms from the action of polyenes has been demonstrated by Lampen et al. (1960) and Gottlieb et al. (1960) for S. cerevisiw, and by Zygmunt and Tavormina (1966) for C . albicans. Resistant organisms that do not contain sterols become sensitive when they have incorporated sterols into their membranes: this was shown for A. Zaidlawii by Feingold (1965) and Weber and Kinsky (1965), for Schizosaccharomyces iaponiCW by Bulder (1971), and for Pythium spp. by Schlosser and Gottlieb (1966) and by Child et al. ( 1969).
H. DISCUSSION All the diverse and ingenious systems which have been used to probe the mode of action of the polyene antibiotics have given consistent results and have satisfied their users that their behavior resembles that of living cells. Filipin, the antibiotic which shows the greatest membrane-disrupting activity, in terms of lysis of erythrocytes, protoplasts, and black films, and alteration in pressure relationships in monolayers and permeability changes, is also the compound which reacts most strongly with sterols, and whose antiyeast activity is reversed most easily by sterols (Zygmunt and Tavormina, 1966). The general order agreed upon by the various groups using the different techniques is roughly filipin >> amphotericin B ( >lucensomycin, pimaricin ) >nystatin; lucensomycin and pimaricin are in parentheses because by no means all the studies included them. While this order correlates to some extent with the relative toxicities of the antibiotics (thus, filipin is too toxic for human use; nystatin is used topically whereas amphotericin B can be used parenterally), it does not tally at all with the order of antiyeast activity, which is amphotericin B >> nystatin > filipin = pimaricin (HamiltonMiller, 1973a). One is inevitably led to the conclusion that the results of the experiments described above throw more light quantitatively upon the toxicity of the polyenes than upon their precise mode of antimicrobial activity. It is still not clear which of the many sequelae of membrane damage is ultimately responsible for yeast cell death (Lampen, 19f39). The atomic architecture of the sterols with which polyenes preferentially complex has been worked out very clearly. The differences between requirements in solution (where the interaction was primarily by hydrophobic forces) and on the surface of membranous structures (where it is hydrophilic forces that seem to predominate) are of particular interest, While something is known of the stoichiometry of the interaction, there is Iittle information about the precise chemical and physical nature of the sterol-polyene complex. Gale (1973) has recently put forward
126
J. M. T. HAMILTON-MILLER
a tentative scheme for the reaction of polyenes and membranes in V ~ V O . From a physiological point of view, and with the medical microbiologist in mind, perhaps the most significant findings in this respect are that cholesterol and ergosterol are good “receptors” for polyenes, and that cholesteryl esters are much less effective. Again, there is not total agreement when the “polyene avidity” league table for sterols is compared with the ability of sterols to inhibit the antiyeast activity of polyenes. Thus, Gottlieb et al. (1960) and Zygmunt and Tavormina (1966) found that cholesterol acetate effectively reversed the action of filipin on C . albicans, and that stigmasterol, the most active compound in in vitro tests in models, was not as good an antipolyene agent as was ergosterol. It is of passing interest to note that the actions of two other antibiotics, mycobacillin ( a polypeptide) and pyrrolnitrin ( a halogenated pyrrole), are to some extent nullified by the presence of certain lipids, including phospholipids and sterols ( Haldar and Bose, 1973). Another point which bears strongly upon the physiological activity of the polyenes and which is brought out by the experiments described above is that it is not only the qualitative and quantitative nature of the sterol in a membrane that is important, but also the stero1:phospholipid ratio, and the chemical nature of that phospholipid. Further, the multiplicity of polyene to sterol molecules also affects the final issue. With all these variables in mind, it can be seen that mechanisms of selective toxicity will be at best difficult and at worst impossible to work out precisely. Overall, however, the quality and the quantity of the research done on the mode of the action of the polyenes is most impressive, if only as an exercise in applied biophysics, and has given rise to two, at least, potentially valuable spin-offs, namely, the nystatin-modified erythrocyte (Cass and Dalmark, 1973) and the muscle fiber with high conductance (Leung and Eisenberg, 1973), both of which may prove to be of extreme value to the physiologists. VI.
Resistance to the Polyene Antibiotics
A search of the literature has revealed only two examples of the occurrence of apparent polyene resistance: Bodenhoff (1968) reported on two cases in which drug-resistant C. albicans appeared in vivo after the prolonged use of polyenes, and Hejzlar and Vymola (1970) describe an apparent increase in the incidence of low-level nystatin resistance in clinically isolated yeasts. Thus, from a clinical point of view, the problem of polyene resistance is nonexistent. There are no other antibiotics in common therapeutic use about which a similar statement could be made. Before experimental work is reviewed in this section, it seems worth while to speculate briefly on this remarkable fact.
FUNGAL STEROLS AND POLYENE ANTIBIOTICS
127
First, the extent of its veracity must be probed. It must be accepted that, at least in the United Kingdom, most medical mycology is carried out in laboratories geared to microbiological investigations in general, the bulk of which involves bacteria and viruses. The presence in such routine laboratories of a trained mycologist is not regarded as necessary. Bearing in mind that the speciation of yeasts is certainly not as easily carried out as is bacterial identification, there is a distinct chance that strains of potentially pathogenic yeasts may be misidentified (or, most commonly, not identified at all), and therefore not reported upon. Further, sensitivity testing to the polyenes is not always routinely carried out, as it is generally assumed that sensitivity is universal. Perhaps the advent of new antimycotics, such as clotrimazole and 5-fluorocytosine, to which there is a range of sensitivities ( Hamilton-Miller, 1972c), may lead to a reappraisal of the whole question of the sensitivity testing of clinically isolated yeasts. The problem of misidentification, however, is by no means so easy to rectify. As pointed out previously by the reviewer ( Hamilton-Miller, 1972a), if resistant variants have even slightly unusual properties ( and the limited evidence available suggests that this might indeed be so), they may very well be misidentified even by experienced mycologists. However, whatever doubts one may cast on the lack of resistance to polyenes, for practical purposes it must be assumed that the happy situation exists whereby the vast majority, at least, of pathogenic yeasts and fungi have retained their sensitivity throughout 15 years or more of polyene usage. Why, then, has resistance not arisen so far? It is known that many (but probably not all) Candida strains are capable, genetically, of giving rise to mutants with markedly lower sensitivity to the polyenes, especially to amphotericin B. Multistep resistant mutants, made in the usual laboratory fashion of serial subculture in the presence of increasing concentrations of antibiotics, have been characterized from several Candida species. Also, single-step mutants of C. albicans have been isolated either with or without the necessity of mutagenesis (see Hamilton-Miller 1973a for reference to the above), but the multistep pathway seems much more common. The only reports of polyene-resistant dermatophytes have so far come from one group (Capek and Simek, 1972), and under in vitro conditions only. Perhaps dermatophyte infections are more often treated with Whitfields ointment or with griseofulvin than with nystatin, but nevertheless there must have been considerable opportunities for the emergence of resistant mutants, which the pathogens do not seem to have grasped. The lack of resistance is even more surprising when one considers the enormous topical use of antifungal antibiotics, topical use of an antibacterial compound often sounding its death knell. Official figures (Annual Report for 1970, 1971) show that 2.9
128
J . M. T. HAMILTON-MILLER
million prescriptions, worth ;E900,000, were written in 1970 in the United Kingdom for “antibacterial and fungicidal agents” prepared for use on the skin and mucocutaneous junctions: a substantial proportion of those prescriptions must have been for preparations containing nystatin. Any attempted explanation for the lack of acquisition of resistance to the polyenes must take into account two predominant factors-first, the opportunistic nature of the pathogen, and second, the physical properties of the antibiotics. While a true study of the nature of polyene resistance is, by its very nature, impossible, such observations as have been made on strains made resistant in vitro tend to support the view that resistance ( a t least the multistep variety) is associated with diminished virulence ( see Athar and Winner, 1971, and references therein). A further lowering of an already low virulence may not allow the organism to continue its parasitic existence, even in the compromised host, so that any resistant celIs that appear under selection during antibiotic treatment may succumb to lost bodily defense mechanisms. As to the second point: the only clear qualitative difference between the polyenes, to which resistance does not occur, and all other antibiotics, to which resistance does occur, is that the polyenes are virtually insoluble in water. This fact may have very important repercussions; as is well known, multistep resistance is encouraged by the use of too small doses of antibiotics, and prevented by the use of massive dosages. Even in very dilute aqueous solution, it seems probable that polyene antibiotics exist in micelles, i.e., packets or quanta of molecules. Thus, a subinhibitory concentration of a polyene antibiotic may well be one in which there are fewer quanta than there are organisms, and one can visualize the situation that any cell that comes in contact with a quantum will perish. In this way, an all-or-none phenomenon will exist, despite superficial appearances to the contrary, as cells will either die or not be exposed at all to the antibiotic, thus effectively preventing the selection of mutants. It is interesting in this respect that Lampen et al. (1959) found that yeast cells absorb far more nystatin than is necessary to kill them. If this hypothesis is correct, namely, that the lack of resistance to polyenes is a consequence of their lack of solubility in water, then the introduction of soluble derivatives of amphotericin B (see Keim et al., 1973) is to be deplored. Time alone will tell.
VII.
The Role of Sterols in Resistance to the Polyenes
Polyene resistance has been demonstrated not only in Candida spp. and dermatophyte strains, as discussed in the preceding section, but has also been found in other organisms of mycological interest, namely Sacchromyces cerevisiae and Neurospora mmsa. In the former species,
FUNGAL STEROLS AND POLYENE .ANTIBIOTICS
129
nystatin-resistant mutants were selected, without prior mutagenesis, by Ahmed and Woods (1967) and Patel and Johnston (1968), while Molzahn and Woods (1973) used ethyl methanesulfonate to obtain their mutants. Grindle ( 1973) isolated polyene-resistant N . crmsu variants after treatment with nitrous acid or ultraviolet light. Patel and Johnston (1971) suggested that the low-level (3.5-fold) resistance occurring in tetraploid S. cerevisiae was due to the difference in surface area to volume ( A I V ) ratio in these cells as compared to haploids, rather than to a biochemical mechanism. Unfortunately, this interesting suggestion was not supported by any hard evidence. In an attempt to resolve this question, the present reviewer ( Hamilton-Miller, 1973c) calculated that AIV ratios for tetraploid yeast cells were 2.7 times larger than for haploid cells, so this “physiological” theory of nystatinresistance is indeed a possibility for polyploid strains of S . cere’euisiae, but cannot be proved in the absence of sterol analyses. In all other cases of reported polyene resistance, where sterol assays have been carried out, changes, qualitative, quantitative, or both, in sterol patterns have been observed. Thus, Woods (1971) and Molzahn and Woods (1972) found a new sterol in nys-1 and pol-1 mutants, and reduction in the content ol ergosterol and its dehydro analog, with replacement by a second new sterol in nys-3 mutants. The pol-2 mutants (Molzahn and Woods, 1972) appeared to lack sterol altogether. Mutants studied by Bard (1972) lacked ergosterol and his nyr-15 strain resembled the nys-3 mutant mentioned above. Most of the N . crussu mutants examined by Grindle (1973) lacked ergosterol: types I and I1 were devoid of sterols; type I11 had greatly reduced levels of ergosterol; and types IV, V, and VI lacked ergosterol and contained other sterols instead. In C . albicuns mutants, Hamilton-Miller ( M 2 b ) found that sterol patterns in strains YL and YS resembled those in the nys-1 mutants of Woods (1971), but that all 4 mutant strains studied contained more ergosterol than the wild-type. The multistep mutants of various Cundidu species studied by Athar and Winner (1971) seemed to be ergosterol-deficient,but lack new sterols (Athar, 1969). It is apparent from the above that alterations in sterol patterns play a crucial role in resistance to the polyenes. Lampen et ul. (1959) showed that the binding of polyene was a prerequisite for the killing of cells, and that a low-level resistant strain of C. albicans absorbed less antibiotic than did a sensitive strain, It is logical, therefore, to consider the ways in which alterations in sterol content, both qualitative and quantitative, could cause changes in the amount of polyene absorbed by yeast cells. The simplest example is given by the nyr-1 and -3, pol-2, -4, and -5 mutants of S. cerevisiue, as well as petites (Woods, 1971), the types I, 11, and I11 N . crassu strains, and the “trained variants of C~ndida
130
J . M. T. HAMILTON-MILLER
studied by Athar and Winner (1971). All these strains either contain decreased amounts of ergosterol compared with the wild type, or none at all, and there are no other changes in sterol composition. Clearly, if it is to ergosterol that polyenes absorb, such organisms will have increased resistance by virtue of having fewer binding sites. Second, there are the strains in which some or all of the ergosterol has been replaced by another sterol, one which binds polyenes less well. Examples in this category are the nys-1 and -3, pol-1 and -3, nyr-15, ole 2-1, ole 2-2 and ole301e4 S . cerevisiae mutants, and the N . cra~sa IV, V, and VI mutants. It has been suggested (Thompson et al., 1971; Bard, 1972) that the nys-3 and the ole series contain zymosterol and lanosterol, respectively. Both these sterols differ sufficiently from ergosterol to mean that the binding of polyenes will be substantially decreased. The remaining strains-the mutagen-induced C. albicans variantswere found to contain more ergosterol than the wild type, whether the sterol content was expressed as gravimetrically, as molecules per cell, or as molecules/pm2.0f cell surface ( Hamilton-Miller, 1973c), and strains Y L and YS contained a new sterol as well (Hamilton-Miller, 1972b). In order to explain resistance in these cases on a biochemical basis-the physiological hypothesis of Pate1 and Johnston ( 1971) being unlikely to apply in this instance ( Hamilton-Miller, 1973~)-it must be postulated that the ergosterol is reoriented (or, in the case of strains YS and YL, masked by the new sterol) so that polyene binding is made more difficult on steric and/ or thermodynamic grounds. Alternatively, one can postulate a change in the ratio of phospholipid to sterol in the membrane, as such changes have been seen to alter the lability of membranes to polyenes. Again, the ergosterol may be esterified in such strains. It must be stressed that the above is only a tentative effort to delineate possible ways in which changes in sterol patterns may affect resistance; it is unlikely that any single explanation applies to any one particular case. Further work needs to be done, especially on uptake of antibiotic by resistant organisms, before any hard and fast statements can be made. In this respect, much more information can be obtained by working with laboratory-induced resistant strains rather than naturally occurring variants, as one has the wild type as a control throughout.
VI I I.
Conclusions
Sterols in yeast have taken on a new importance in the past few years. Previously, as indicated in Section 111, interest in this topic was almost completely a commercial one, as S . cereuisiae was a valuable and easily accessible source of ergosterol, a precursor of many important products in the pharmaceutical industry. Now that it has been established
FUNGAL STEROLS AND POLYENE ANTEIIOTICS
131
that polyene-resistant mutants of yeasts may have altered sterol patterns, the pharmaceutical industry has new access to a potentially rich source of rare chemicals and important precursors. Furthermore, the presence of sterol-deficient mutants is inevitably a bonus to workers in the field of sterol and steroid biosynthesis, whose understanding of the relevant metabolic pathways will be furthered by a study of various polyeneresistant and sterol-dependent ( Karst and Lacroute, 1973) mutants. In the long run, however, it may transpire that the greatest contribution made to the fund of human knowledge from a study of the mode of action of the polyene antibiotics and the involvement therein of sterols will lie in the field of membrane structure and function, which is perhaps the most vital and challenging problem in biology today. REFERENCES Adams, B. G., and Parks, L. W. (1967). J. Cell. Physiol. 70, 161-168. Ahmed, K. A., and Woods, R. A. (1967). Genet. Res. 9, 179-193. Andreoli, T. E., and Monahan, M. ( 1968). I . Gen. Physiol. 52, 300-325. Andreoli, T. E., Dennis, V. W., and Weigh, A. M. (1969). 1. Gen. Physiol. 53, 133-156. Annual Report for 1970. (1971 ). Department of Health and SociaI Security. HM Stationery Office, London. Appleton, G. S., Kieber, R. J., and Payne, W. J. (1955). A w l . Microbiol. 3, 249-251. Asselineau, J., and Lederer, E. ( 1960). In “Lipide Metabolism” (K. Bloch, ed.), pp. 337-406. Wiley, New York. Athar, M. A. (1969). Ph.D. Thesis, University of London. Athar, M. A., and Winner, H. I. (1971). J. Med. Microbiol. 4, 505-517. Bard, M. (1972). J. Bacteriol. 111, 649-657. Barton, D. H. R., Kempe, U. M., and Widdowson, D. A. (1972). J. Chem. SOC., Perkin Trans. I, 513522. Bergy, M. E., and Eble, T. E. (1968). Biochemistry 7,653-658. Bianchi, D. T. (1967). Antonie van Leeuwenhoek; 1. Microbiol. Serol. 33, 324-332. Bittman, R., and Fischkoff, S. A. (1972). Proc. Nat. Acad. Sci. US.69, 37954799. Bodenhoff, J. (1968). Odontol. Tidskr. 76, 279-294. Breivik, 0.N., and Owades, J. L. (1957). Agr. Food Chem. 5,360363. Bulder, C. J. E. A. (1971). Antonie uan Leeuwenhoek; J . Microbiol. Serol. 37, 353-358. Capek, A, and Siniek, A. ( 1972). Folia Microbiol. (Prague) 17, 239-240. Cass, A., and Dalmark, M. (1973). Nature (London),New Biol. 244, 4 7 4 9 . Cass, A, Finkelstein, A., and Krespi, V. (1970). 1.Gen. Physiol. 56, 100-124. Child, J. J., Defago, G., and Haskins, R. H. (1969). Can. J. Microbiol. 15, 599-603. Crifo, C.,Strom, R., Santoro, A, and Mondovi, B. (1971). FEBS Lett. 17, 121-126. Defago, G., Child, J. J., and Haskins, R. H. (1969). Can. J. Microbiol. 15, 509-514. de Gier, J., Mandersloot, J. G., and van Deenen, L. L. M. ( 1968). Biochim. Biophys. Acta 150, 666-675. de Kruyff, B., Demel, R. A., and van Deenen, L. L. M. (1972). Biochim. B i o p h p A d a 255, 331-347. de Kruyff, B., De Greef, W. J., van Eyk, R. V. W., Demel, R. A., and van Deenen, L. L. M. (1973). Biochim Biophys. Acta 298, 479-499.
132
J. M. T. HAMILTON-MILLER
Demel, R. A, van Deenen, L. L. M., and Kinsky, S. C. ( 1965). J. BioZ. Chem. 240, 2749-2753. Demel, R. A., van Deenen, L. L. M., and Pethica, B. A. (1967). Biochim. Biophys. Actu 135, 11-19. Demel, R. A., Crombag, F. J. L., van Deenen, L. L. M., and Kinsky, S. C. (1968a). Biochim. Biophys. Actu 150, 1-14. Demel, R. A., Kinsky, S. C., Kinsky, C. B., and van Deenen, L. L. M. (1968b). Biochim. Biophys. Actu 150, 655-665. Demel, R. A., Bruckdorfer, K. R., and van Deenen, L. L. M. (1972a). Biochim. Biophys. Actu 255, 311-320. Demel, R. A., Bruckdorfer, K. R., and van Deenen, L. L. M. (1972b). Biochim. Biophys. Acta 255, 321-330. Dod, B. J., and Gray, G. M. (1968). Biochim. Biophys. Actu 150, 397-404. Drabikowski, W., Lagwinska, E., and Sarzala, M. G. (1973). Biochim. Biophys. Acta 291, 61-70. Dulaney, E. L. (1957). U.S. Patent 2,817,624. Dulaney, E. L., Stapley, E. O., and Simpf, K. (1954). AppZ. Microbiol. 2, 371-379. Elliott, C. G. (1969). J. Gen. Microbiol. 56, 331-343. El-Refai, A. H., and El-Kady. I. A. (1968a). 2. AZZg. Mikrobiol. 8, 55-60. El-Refai, A. H., and El-Kady, I. A. ( 1968b). Z. AZZg. MikrobioZ. 8, 361-366. Feingold, D. S. (1965). Biochem. Biophys. Res. Commun. 19, 261-267. Finkelstein, A., and Cass, A. (1968). J. Gen. Physiol. 52, 145s-172s. Gale, E. F. (1973). Brit. Med. J. 3, 33-38. Gordon, P. A., Stewart, P. R., and Clarke-Walker, G. D. (1971). J. Bucteriol. 107, 114-120. Gottlieb, D., Carter, H. E., Wu, L. C., and Sloneker, J. H. (1960). Phytopathology 50, 594-603. Gottlieb, D., Carter, H. E., Sloneker, J. H., Wu, L. C., and Gaudy, E. (1961). Phytopathology 51, 321-330. Grindle, M. ( 1973). Mol. Gen. Genet. 120, 283-290. Hddar, A., and Bose, S. K. (1973). 1. Antibiot. 26,358-361. Hamilton-Miller, J. M. T. (1972a). J. Med. Microbid. 5, 425-440. Hamilton-Miller, J. M. T. ( 1972b).J. Gen. Microbiol. 73, 201-203. Hamilton-Miller, J. M. T. ( 1 9 7 2 ~ )Subouruudiu . 10, 276-283. Hamilton-Miller, J. M. T. (1973a). Bacteriol. Rm. 37, 16CL196. Hamilton-Miller, J. M. T. (1973b). Chemotherapy (Basel) 18, 154-161. Hamilton-Miller, J. M. T. ( 1 9 7 3 ~ )Microbios . 8, 209-213. Hejzlar, M., and Vymola, F. (1970). J. Hyg., EpidemioZ., Microbiol., Immunol. 14, 211-213. Hendrix, J. W. ( 1964). Science 144, 1028-1029. Hill, P. (1972). J. Gen. Microbiol. 70, 243-252. Hob, R., and Finkelstein, A. (1970). J. Gen. Physiol. 56, 125-145. HsuChen, C-C., and Feingold, D. S. (1973). Biochem. Biophys. Res. Commun. 51, 972-978. Hunter, K., and Rose, A. H. (1972). Biochim. Biophys. Acta 280, 639-653. Karst, F., and Lacroute, F. (1973). Bwchem. Biophys. Res. Commun. 52, 741-747. Keim, G. R., Poutsiaka, J. W., Kirpan, J., and Keysser, C. H. (1973). S C ~ ~ ~ C S 179, 584-585. Kinsky, S. C . (1982). J. Bacteriol. 83, 351458. Kinsky, S. C. (1963). Arch. Biochem. Biophys. 102, 180-188. Kinsky, S. C., Avruch, J, Permutt, M., Rogers, H. B., and Schonder, A. A. (1962). Biochem. Biophys. Res. Commun. 9, 503-507.
FUNGAL SrEROLS AND POLYENE ANTIBIOTICS
133
Kinsky, S. C., Gronau, G. R., and Weber, M. M. (1965). Mol. Pharmucol. 1, 190-201. Kinsky, S. C., Luse, S. A., and van Deenen, L. L. M. (1966). Fed. Proc., Fed. Amer. SOC. Exp. Biol. 25, 1503-1510. Kinsky, S . C., Demel, R. A., and van Deenen, L. L. M. (1967a). Biochim. Biophys. Acta 135, 835-843. Kinsky, S . C., Luse, S. A., Zopf, D., van Deenen, L. L. M., and Haxby, J. (1967b). Bioohim. Biophys. Acta 135, 844-861. Kinsky, S. C., Haxby, J., Kinsky, C. B., Demel, R. A., and van Deenen, L. L. M. (1968). Bwchim. Biophys. A d a 152, 174-185. Lampen, J. 0. (1969). Amer. 3. Clin. Pathol. 52, 138-145. Lampen, J. O., Morgan, E. R., Slocum, A., and Amow, P. (1959). 3. Bacteriol. 78, 282-289. Lampen, J. O., Arnow, P. M., and Safferman, R. S. ( 1960). 1. Bacteriol. 80, 200-206. Lenney, J. F., and Klemmer, H. W. ( 1966). Nature (London) 209, 1265-1366. Leung, J., and Eisenberg, R. S. (1973). Biochim. Biophys. Acta 298, 718-723. Lichtenstein, N. S., and Leaf, A. ( 1965). 3. Clin. Invest. 44, 1328-1342. Lippe, C. (1968). J . Mol. Biol. 35, 635-637. Lippe, C., and Giordana, B. (1967). Biochim. Biophys. Acta 135, 966-972. McElhaney, R. N., de Gier, J., and Ven der Neut-Kok, E. C. M. (1973). Biochim. Biophys. Acta 298, 500-512. Molzahn, S., and Woods, R. A. (1972). 3. Gen. Microbiol. 72, 339-348. Morpurgo, G., Serlupi-Crescenzi, G., Tecce, G., Valente, F., and Venettacci, D. (1964). Nature (London) 201, 597-899. Norman, A. W., Demel, R. A., de Kruyff, B., Geurts van Kessel, W. S. M., and van Deenen, L. L. M. (1972a). Biochim. Biophys. Acta 290, 1-14. Norman, A. W., Demel, R. A., de Kruyff, B., and van Deenen, L. L. M. (1972b). 3. Biol. Chem. 217, 1918-1929. Parks, L. W., and Starr, P. R. (1963). 3. Cell. Comp. Physiol. 611, 61-65. Parsons, D. F., and Yano, Y. (1967). Biochim. Biophys. Acta 135,362-364. Patel, P. V., and Johnston, J. R. (1968). Appl. Microbial. 16, 164-165. Patel, P. V., and Johnston, J. R. (1971). 3. Appl. Bacteriol. 34, 449458. Proudlock, J. W., Wheeldon, L. W., Jollow, D. J., and Linnane, A. W. (1968). Biochim. Biophys. Acta 152, 434-437. Rickards, R. W., Smith, R. M., and Golding, B. T. (1970). J . Antibiot. 23, 603-4312. Schlosser, E., and Gottlieb, D. ( 1966). J . Bacteriol, 91, 1080-1084. Schlosser, E., Shaw, P. D., and Gottlieb, D. (1969). Arch. Microbiol. 66, 147-153. Schroeder, F., and Bieber, L. L. (1971/1972). Chem. Biol. Interact. 4, 239-249. Schroeder, F., Holland, J. F., and Bieber, L. L. (1971). 3. Antibiot. 24, 84t?-849. Schroeder, F., Holland, J. F., and Bieber, L. L. (1972). Biochemistry, 11, 3105-3111. Schubert, K., Rose, G., Tiimmler, R., and Ikekawa, N. (1984). Hoppe-Seyler's Z . Physiol. C h m . 339,293-296. Schubert, K., Rose, G., Wachtel, H., Horhold, C., and Ikekawa, N. (1968). E m . 1. Biochem. 5, 245251. Sessa, G., and Weissmann, G. (1967). Biochim. Biophys. Acta 135, 416426. Sessa, G., and Weissmann, G. (1968a). 1. Lipid &s. 9, 310-318. Sessa, G., and Weissmann, G. (19681,). 3. Biol. C h m . 243, 4364-4371. Strom, R., Crifo, C., and Bozzi, A. (1973a). Biophys. 3. 13,568-580. Strom, R., Crifo, C., and Santoro, A. (1973b). Biophys. 1. 13, 581593. Sweeley, C. C., O'Connor, J. D., and Bieber, L. L. (1970). Chem. Biol. Interact. 2, 247-253. Tanret, C. (1879). Ann. Chim. Phys. [S] 17, 493-512.
134
J . M. T. HAMILTON-MILLER
Thompson, E. D., Starr, P. R., and Parks, L. W. ( 1971). Biochem. Biophys. Res. Commun. 43, 1304-1309. van Zutphen, H., van Deenen, L. L. M., and Kinsky, S. C. (1966). Biohem. Biophys. Res. Commun. 22, 393-398. van Zutphen, H., Demel, R. A., Norman, A. W., and van Deenen, L. L. M. (1971). Biochim. Biophys. Acta 241, 310-330. Verkleij, A. J., de KruyE, B., Gerritsen, W. F., Demel, R. A., van Deenen, L. L. M., and Ververgaert, P. H. J. (1973). Biochim. Biophys. Acta 291, 577581. Weber, M. M., and Kinsky, S. C. (1965). J . Bacteriol. 89,306-312. Weissmann, G., and Sessa, G . (1967). J . Biol. Chern. 242, 616-625. Weissmann, G., Hirschhorn, R., Pras, M., and Bevans, V. ( 1966). Fed, Proc., Fed. AWT. Soc. Exp. Biol. 25, 358. Woods, R. A. (1971). J . Bacteriol. 108,69-73. Zygmunt, W. A., and Tavormina, P. A. (1966). Appl. Microbiol. 14, 865-869.
Methods of Numerical Taxonomy for Various Genera of Yeasts
I. CAMPBELL Depattment of Brewing and Biological Sciences, Heriot-Watt University, Edinburgh, Scotland
I. II. 111. IV.
Theory of Classification of Yeasts .......... Numerical Taxonomy Applied to Yeasts ....... Comparison of Numerical and Classical Taxonomy ... Application of Computer Techniques to Identification . . References ....................
1.
135 137 148 150 154
Theory of Classification of Yeasts
The methods of classification of yeasts are fully described elsewhere (Kreger-van Rij, 1969; van der Walt, 1970a), but some introductory comment is required. The current system of classification depends on the following characters (see also Table I ) . 1. Form of asexual reproduction: by multilateral and narrow-based, or bipolar and broad-based budding, or binary fission. Usually no difficulties arise in this test, but Bamett (1960) has noted the occasional difficulties in determining the difference between multilateral and bipolar budding. 2. Formation of spores. Classification of yeasts has evolved from the methods of mycology rather than of bacteriology. Therefore the mycologists' obsession with the type of sexual spores, if formed, is of fundamental importance in yeast taxonomy. This is unfortunate for several TABLE I TESTSFOR IDENTIFICATION OF YEASTS' Morphology
Characteristics of vegetative reproduction Size and shape of cells Size, shape, and number of spores Formation of pseudomycelium Appearance of colonies
Physiology
Pellicle formation on liquid media Fermentation of sugars Assimilation of carbon compounds Assimilation of nitrate (and possibly nitrite, ethylamine, or lysine) Tests of minor importance Growth in absence of vitamins Maximum temperature of growth Growth in high osmotic pressure Resistance to cycloheximide (Actidione)
0
After van der Walt (1970a).
135
136
I. CAMPBELL
reasons: up to 3 weeks’ incubation may be required to demonstrate spores; sporulation is greatly dependent on cultural conditions, particularly the type of medium used; and poor sporulation may be missed. Most sporing yeasts have an exact nonsporing equivalent, and in their simplified identification scheme, Beech et al. (1968) suggested ignoring sporulation and offering two identifications of the culture: the sporing and the nonsporing species which best fit the results of the identification tests. One therefore avoids the uncertainty and delay of waiting for sporulation before identifying the genus according to Lodder’s (1970) key. 3. Assimilation of nitrate. Species of the genera Citeromyces, Hansenula, and Pachysolen grow on defined medium in which nitrate is sole nitrogen source: two species of Endomycopsis also have this property, and, in the two species of the genus Dekkera, assimilation of nitrate is variable. We observe, therefore, that nitrate assimiliation may define genera, or define species within a genus, or be completely irrelevant to classification. The consequences of such inconsistency are discussed later. 4. Fermentation of sugars. Genera such as Kluyveromyces and Saccharomyces are characterized by active fermentation of glucose and, often, other sugars. Species of the genera Debaryomyces and Pichia may ferment sugars, but if so, only weakly. Recent reclassification, e.g., transfer of D . globosus to the genus Sacchuromyces as S. kloeckerianus ( Kreger-van Rij, 1970a; van der Walt, 1970c) has helped classification, but nevertheless some experience is required to judge whether fermentation is “active” or “weak or “slow.” Taxonomic criteria for the nonsporing yeasts are completely different ( Lodder, 1970). The absence of spores having been assumed and the isolate therefore allocated to the group of imperfect yeasts, the genus is identified by production of mycelium ( Leucosporidium, Rhodosporidium ) , pseudomycelium ( Candida ) , or no mycelium ( Cryptococcus, Rhodotorula, or Torulopsis ) . Production of pseudomycelium is variable according to cultural, and possibly other, conditions, and the distinction between the genera Candida and Tomlopsis, both of which include fermentative, nonfermentative, nitrate-assimilating, and nonnitrate-assimilating species is therefore not always clear (van Uden and Buckley, 1970; van Uden and Vidal-Leiria, 1970). Although yeast genera are difficult to identify reliably, identification of species is usually more consistent. Each species is defined by physiological properties and by the morphology of cells, including pseudomycelium and spores, if formed. Although all these properties are variable in most species, the variation is normally within narrow limits. Physiological tests include 5-13 sugar fermentation tests (there is some lack of standardization between authorities on this point ) , assimilation
NUMERICAL TAXONOMY OF YEASTS
137
of 31 carbon compounds as sole source of carbon in synthetic medium, assimilation of 1-4 nitrogen compounds, and growth in high sugar or salt concentrations, at 37OC, and in the absence of vitamins. It is rare indeed for all known isolates allocated to a single species to be identical in all these properties: standard descriptions of species include an allowance for a variation in properties which appears to be in proportion to the number of strains available for study. The description of S. cerevisiae includes 23 variable characters in the above physiological tests; the morphological description of the species also includes generous allowance for variation (van der Walt, 1 9 7 0 ~ ) . Similar variation is also noted in, for example, S. bayanus, S. rouxii, S. uvarum (van der Walt, 1970c), Cryptococcus (Phaff and Fell, 1970), and Rhodotorula (Phaff and Ahearn, 1970). Such variation is partly due to allowance for latent assimilation or fermentation reactions; slow or acquired reactions previously caused the transfer of cultures from one species to another (Scheda and Yarrow, 1966, 1968; Gilliland, 1969). The range of carbon compounds for assimilation tests includes pairs or groups of substances metabolized by a common pathway (Barnett, 1960, 1968). Thus, pentoses and their corresponding alcohols do not justify a test for each compound. It is very much easier to make the foregoing criticisms of the present system than to suggest a practicable alternative; even the authors of Lodder’s ( 1970) widely accepted classification admit imperfection. Alternative tests to those generaIly used have been attempted, but not yet on a sufficiently wide range of yeasts to be seriously considered for classification. I refer especially to such methods as proton magnetic resonance spectra of cell wall polysaccharides (see, e.g., Gorin and Spencer, 1970), GC ratio of DNA bases (Meyer and Phaff, 1969; Stenderup and Leth Bak, 1968; Nakase and Komagata, 1970, 1971a-d) and serological properties of whole cells (Tsuchiya et al., 1965a,b; Campbell, 1970) or of cell wall extracts (Ballou, 1970; Sandula et al., 1964, 1972; Summers et al., 1964; Suzuki et al., 1968). The available results are in moderately good agreement with currently accepted classification, but numerous examples exist, some mentioned in Section 111, of identical reactions between supposedly different species, or of widely different reactions between related species. Numerical analysis may help to solve some of these problems of yeast taxonomy. II.
Numerical Taxonomy Applied to Yeasts
Classification by a large number of properties of equal importance was first proposed by Adanson in 1763 (Sneath, 1957), but the use of such a system has become practicable only with computer analysis
138
I. CAMPBELL
of the copious data necessarily involved. Sneath’s reviews (1962, 1972) are excellent introductions to numerical taxonomy; a more detailed treatment is provided by Sokal and Sneath‘s (1963) book. Similarity judged by a large number of unrelated properties must give a better indication of relationship between cultures than a few “important” properties, especially when, as in yeast taxonomy, the importance of a character varies with the genus to which it is applied. Following these ideas, numerous authors have published classification schemes of groups of microorganisms, although few of these papers have been concerned with yeasts. Two possible quantities for evaluation of relationships between species are the similarity coefficient and matching coefficient. Most extensive examination of matching Coefficients between yeasts has been conducted by Kockova-Kratochvilova and her co-workers ( 1966, 1968, 1969a, 1970). The analyses reported by Kockova-Kratochvilova et al. up to 1970 have been based on matching coefficients between strains, defined by Sneath ( 1962) as S,
=
(a
+ d ) / ( a + b + c + d ) X 100
Where a = number of characters positive in both strains; d = number of characters negative in both strains; b = number of characters positive in the first strain and negative in the second, and c = number of characters negative in the first strain and positive in the second. Previously only the similarity coe5cient S , was considered acceptable (see Sneath, 1957,1962); 8, = a / ( a
+ b + c)
X 100
where a, b, and c are defined above. Negative matches were considered inappropriate, since two strains could be negative for different reasons and therefore not equivalent in these characters. However, subsequent experience indicated no practical objections to the reasonable use of negative matches (see Sneath, 1962; Sokal and Sneath, 1963, for discussion). It is normal practice (Sneath, 1957, 1962) to ignore characters that are identical in all strains under test. Therefore in analysis of the genera Saccharomyces and Khyveromyces (Campbell, 1972a), we omitted assimilation of nitrate and production of hat-shaped spores, always negative, and fermentation and assimilation of glucose, always positive. In more recent comparisons between Debaromyces, Hamenula, and Pichia (Campbell, 1973a) these properties were introduced and were considered in all comparisons, whether applicable or not; for example, in a set of comparisons of Debaryomyces and Pichia only, results of nitrate tests were included, although always negative. In practice if only a few
139
NUMERICAL TAXONOMY OF YEASTS
characters are consistently positive, or negative, in a set of tests the effect on matching coefficient is negligible, e.g., 50 similarities
58 characters
= 86.2%
By addition of 4 consistently positive characters we have 54 similarities = 87.1% 62 characters
Reduced to its simplest terms, numerical analysis consists of tabulating the descriptions of each strain, scoring (or 1) for all characters which the strain possesses, and-(or 0, or no score) for characters which are absent, e.g., no fermentation of maltose. At least 50, and preferably over 100, characters are required for valid determination of similarity or matching coefficient ( Sneath, 1962; Sokal and Sneath, 1963). Not all characters are conveniently expressed in a simple two-state (+/- or 1/0) systems. Properties may be graded, as multistate characters. In analysis of Saccharomyces species, the tolerance of high concentrations of alcohol may be expressed as in Table I1 (Campbell, 1970); the difference between two strains could be up to 3 units. Allocation of 3 characters to alcohol tolerance was justified in a specialized comparison of fermentative yeasts, but was reduced to 2 units in subsequent comparisons of other genera ( Campbell, 1971,1972a). On the other hand, Kockova-Kratochvilova et al. (1966, 1967, 1968, 1969a,b, 1970) used a nonadditive code as shown in Table 111, using the symbol “no comparison,” but at the expense of a more complex program. By their system, the difference between two strains with respect to alcohol tolerance did not exceed one character. There are no obvious rules for deciding which or -; e.g., assimilation of characters should be expressed as simply trehalose was scored as +/- by Kockova-Kratochvilova et al. (1966, 1967, 1968, 1969a,b, 1970) and Campbell (1972a), but on a 0-5 scale by Poncet ( 1967, 1970).
+
+
TABLE I1 SCORING OF GROWTH I N VARIOUS CONCENTRATIONS OF E T H A N O L ~ Property No growth in 8 % ethanol Tolerance of up to 8% ethanol Tolerance of up to 12% ethanol Tolerance of up to 16% ethanol ~
a
After CampbeH (1970).
Code 0 1 1 1
0 0 1 1
0 0 0 1
140
I. CAMPBELL
TABLE I11 SCORING OF GROWTH I N VARIOUS CONCENTRATIONS OF ETHANOLQ Property Tolerance of up Tolerance of up Tolerance of up Tolerance of up
to to to to
1% ethanol 4% ethanol 8% ethanol 12% ethanol
Code 0 Nb N 1 0 0 1 1 0 1 N 1
0. After Kockova-Kratochvilova et al. (1966, 1969a). * N = no comparison.
Having provided the computer with information on each of the strains the matching between the first and second, first and third, etc., is determined. When all other strains have been matched against the first, the procedure is repeated by comparing the second strain with the third, fourth, etc. Thus the matching coefficient is determined for all pairs of strains. The analysis is illustrated by Tables IV-VII, from data used by Campbell ( 1972a). The morphological and physiological properties of a strain are punched on a card, one to each strain (Table IV); 1 usually represents a positive reaction. Comparison of a series of cards provides data, a sample of which is shown as Table V. When large numbers of strains are compared, this form is difficult to interpret and TABLE I V PROPERTIES OF 10 STRAINS OF Saccharomyces
no. Strain N543 N616 379 3081 4575 5155 5378 N161 1108 1116
c
I
Morphology 00101110000 00101010000 00101111000 00100111000 00101010000 00100110000 00100110000 00101110000 00100111000 00100101000
I
FOR
Fermentation and assimilation 11111100011010101101100111010001 11111100011010111101100111010000 11111000000010011101110101010001 11010110000000111101110101010001 11011110011010111001000001000001 11110110000000111101100101010000 10000000000000001101100101000001 10011110000010001011011111011000 11111110000110001001100101010000 11111110000100001001000101010000
ANALYSIP Miscellaneous tests 00001000010111111 00001000011011111 10001000010110111 01100000000110111 01100000000011111 00010000000100111 01100000010110111 00001000011111111 01100000011110011 01000000011111111
Each line of information represents a separate card. The 10 strains are selected from the 72 strains described by Campbell (1972a), Fig. 4. The first four characters identify the strain; the remainder represent its morphological and physiological properties.
141
NUMERICAL TAXONOMY OF YEASTS
TABLE V UNSORTED RESULTSOF ANALYSIS OF STRAINS LISTEDIN TABLEI V PERCENTAGE PERCENTAGE PERCENTAGE PERCENTAGE PERCENTAGE PERCENTAGE PERCENTAGE PERCENTAGE PERCENTAGE PERCENTAGE PERCENTAGE
MATCHING OF STRAIN MATCHING OF STRAIN MATCHING OF STRAIN MATCHING OF STRAIN MATCHING OF STRAIN MATCHING O F STRAIN MATCHING OF STRAIN MATCHING OF STRAIN MATCHING OF STRAIN MATCHING OF STRAIN MATCHING OF STRAIN
N543 WITH STRAIN N543 WITH STRAIN N543 WITH STRAIN N543 WITH STRAIN N543 WITH STRAIN N543 WITH STRAIN N543 WITH STRAIN N543 WITH STRAIN N543 WITH STRAIN N616 WITH STRAIN N616 WITH STRAIN
N616 = 92.19 C379 = 84.37 3081 = 75.00 4575 = 78.12 5155 = 78.12 5387 = 76.56 N161 = 78.12 1108 = 75.00 1116 = 75.00 C379 = 79.69 3081 = 73.44
and so on, to PERCENTAGE MATCHING OF STRAIN N161 WITH STRAIN 1108 = 75.00 PERCENTAGE MATCHING OF STRAIN N161 WITH STRAIN 1116 = 78.12 PERCENTAGE MATCHING OF STRAIN 1108 WITH STRAIN 1116 = 90.62
TABLE VI RESULTSOF ANALYSISOF STRAINS LISTEDIN TABLE IV, ARRANGEDAT 5% INTERVALS IN ORDEROF MATCHINQ COEFFICIENTS 90-95% 90-95% 85-90% 85-90% 8045% 80-85% 80-85% 75-80 %
MATCHING OF STRAIN MATCHING OF STRAIN MATCHING OF STRAIN MATCHING OF STRAIN MATCHING OF STRAIN MATCHING OF STRAIN MATCHING OF STRAIN MATCHING, etc.
N543 1108 3081 3081 N543 C379 3081
WITH WITH WITH WITH WITH WITH WITH
STRAIN STRAIN STRAIN STRAIN STRAIN STRAIN STRAIN
N616 1116 5155 5387 C379 3081 1108
92.19 90.62 = 87.50 = 85.94 = 84.37 = 81.25 = 81.25 = =
TABLE VII TRIANGULAR MATRIXOF RESULTSOF ANALYSISOF STRAINS N543 N616 c379 3081 4575 5155 5378 N161 1108 1116
0 0 0 0 0 0 0 0 0 0 N543
8 0
6 5 0 0 0
0 0 0 0 0 0 0 0
0 0 0 0 0
N616
C379
4 4 6 0 0 0 0 0 0 0
3081
5 Ti
3 5 0 0 0 0 0 0 4575
5 5 5 7 4 0 0 0 0 0 5155
5 3 5 7 4 5 0 0
5 5 4 3 3 3 3 0
4 4 5 6 4 5 5 4
0
0 0
0 0
0
5378
N161
1108
4
4 4 5 4 5 4
5 8 0 1116
142
I. CAMPBELL
is therefore rearranged by the computer in order of highest matching coefficient, in Table VI at 5%intervals, and also in the form of a triangular matrix (Table VII) (Sneath, 1962; Sokal and Sneath, 1963) in which 0 represents S, loo%, or no test, 9 = S, 95-99.994&, etc., as far as is necessary. From these data a dendrogram (Fig. 1) is prepared; computer assistance is possible (Sneath, 1957), but we have preferred to perform this stage manually. Thus from the punched-card data the relationships between strains may be expressed as a dendrogram (“taxonomic tree”) which indicates the relationships between species. Figure 1 is a twodimensional representation of a multidimensional relationship ( see Sokal and Sneath, 1963, for discussion) and therefore introduces some distortion of the cluster, but this is tolerated for the sake of simplicity. Although the system illustrated by Table IV-VII was used to measure relationships between yeast strains (Campbell, 1971) or species (Campbell, 1972a), a more sophisticated treatment is usual. Analyses by Kockova-Kratochvilova et al. (1969a, 1970) made extensive use of nonadditive multistate characters; e.g., cell morphology was expressed in terms of 24 characters, whereas in our analyses 7 (additive) characters sufficed (Campbell, 1971). The choice of characters should be based on the widest possible range of properties of the organism: Kockova-Kratochvilova et al. (1966, 1967, 1968, 1969a,b, 1970) considered morphology, fermentation, assimilation of carbon compounds, tolerance of alcohol, osmotic effects, resistance to autolysis, resistance to Actidione and hop oils, and serological reactions. Our analyses ( Campbell, 1971, 1972a) considered the utilization of a wider range of carbon sources, and also the ability to use various 543 616
319 161
noa 1116
3081 5155 5378 4575 70
a0
90
100%
Matching coefficient
FIG.1. Dendrogram of strains analyzed in Tables IV-VII.
NUMERICAL TAXONOMY OF YEASTS
143
amino acids as nitrogen source, and requirement for individual growth factors, in addition to an otherwise shortened version of Kockova-Kratochvilova’s list. It was important to use only independent characters; e.g., lack of fermentation of glucose implied lack of fermentation of any other carbohydrate, and such comparisons were therefore not valid. The genera Kluyueromyces and Sacchuromyces were not affected by that restriction, but for analysis of fermentative and nonfermentative species of Hansenula and Pichia ( Poncet, 1967, 1970; Campbell, 1973a), a “no comparison” entry should have been used (but was not used, introducing the possibility of up to 5 erroneous negative results). Such scoring was justified (Campbell, 1973a) as additional weighting for the property of fermentation. Barnett’s (1968) criticism of related tests in identification is equally applicable to numerical taxonomy; metabolism of the three p-glucosides arbutin, cellobiose, and salicin by the same enzyme system suggests that only one of these compounds should be compared. Attempts to broaden the range of assimilation tests to hydrocarbons (Komagata et aZ., 1964; Scheda and Bos, 1966) and aromatic compounds (Mills et al., 1971) have been received with little enthusiasm. Various nitrogen sources have been investigated for possible use in taxonomy: amines (Brady, 1965), amino acids (Brady, 1965; Campbell, 1970; Morris and Eddy, 1957; Wickerham, 1946), and purines and pyrimidines (Larue and Spencer, 1968). These investigations have been limited to small groups of yeasts. Variable results between strains of the same species, or, worse, between tests with the same strain, have discouraged extensive investigation. Sneath and Johnson ( 1972) demonstrated the serious effects on a numerical taxonomy routine of variable results in identification tests. The most extensive program of numerical taxonomy of yeasts has been performed by Kockova-Kratochvilova and her colleagues ( 1966, 1967, 1968, 1969a,b, 1970). They analyzed strains of Saccharomyces which fermented glucose, sucrose, maltose, and raffinose and strains which fermented galactose and melibiose in addition to these sugars. Their strains represented the range of industrially important yeasts. Analysis of their selection of strains provided results expressed as a dendrogram in Fig. 2 ( Kockova-Kratochvilova et al., 1966, 1968, 1969a, 1970); the results differ from our analyses only in the similarity levels between species, due mainly to the different methods of computation. Kockova-Kratochvilova et al. determined the mean matching coefficient, while the results in Fig. 3 (Campbell, 1973b) were calculated at the highest matching coefficient between strains. It is interesting to note the number of distinct groups which, in our opinion, may be counted as separate species (Figs. 2 and 3). In van der Walt’s ( 1 9 7 0 ~ )classifica-
144
I. CAMPBELL
I 1 I
-I -i
S . uvarum S . car1sberRensi.s S. carlsbergensis
S . ellipsoideus S . ellipsoideus
S . ellipsoideus S. cerevisiae
S . cerevisiae S. oviformis ’
I I I 70
S. heterogenicus
S . willianus
80
90
100%
Matching coefficient
FIG. 2. Taxonomic relationships of common brewery culture yeasts (Succhuroet ul. (1966, 1968, 1969a, 1970).
myces) and contaminants, deduced from results of Kockova-Kratochvilova
tion of the genus Saccharomyces, the formerly separate species S. cerevisiae and S . willianus were combined as a single species 3, cerevisiae, in spite of morphological and physiological differences. S. carlsbergensis, S . logos, and S. uvarum, separate species according to Lodder and Kreger-van Rij (1952) were combined as a single species by van der Walt ( 1 9 7 0 ~ )as S. uvarurn, to the consternation of brewery microbiologists (e.g., Gilliland, 1971). Van der Walt’s insistence on allocating isolates of the same sugar fermentation reactions to the same species, irrespective of other properties, required the descriptions of s. bayanus, S. cerevisiae, and s. uuarum to include many variable properties. Analysis of the descriptions by van der Walt ( 1 9 7 0 ~ )indicated such wide variation and such extensive overlap of the species that logical classification as 3 species seemed impossible (Campbell, 1972a). Taxonomic studies on yeasts have also been performed by factor analysis. The factors are theoretical concepts deduced from the relationships, or correlations, among the strains, which are therefore allocated to their position in multidimensional space by computer analysis of these factors. Gyllenberg ( 1965) suggested identification ( of bacteria) by examining whether an unknown culture lay in the region of a recognized species; if not, the species to which it showed greatest affinity could be recognized. Gyllenberg‘s concept of position in three-dimensional space is easier
NUMERICAL TAXONOMY OF YEASTS
145
to visualize than the more accurate situation in six-dimensional space (Poncet, 1967, 1970) or indeed a space of one dimension for every * character, which is strictly correct but not in practice required (Poncet, 1967). Poncet (1967) analyzed a strain of each of 30 species of the genus Pichia by factor analysis using 36 characters scored on a 05 scale. All but 7 of the tests were assimilation and fermentation of carbohydrates and organic acids. A two-dimensional diagram of the distance between species indicated three main groups of Pichia. Boidin et aZ. (1965) had previously classified the genus Pichia as three groups on the basis of physiological properties, assessed without computer aid, but the groups of Poncet’s analysis did not match exactly with Boidin’s groups. A similar analysis of the genus Hamenula (Poncet, 1970) again produced some discrepancies from previously accepted groupings, and from the proposed phylogenetic relationships by Wickerham ( 1970). Poncet’s system was criticized by Kreger-van Rij (1969) for relying on a single strain to represent species which included a wide extent of variation. However, repetition of Poncet’s analyses ( 1967, 1970) , using the standard descriptions provided by Kreger-van Rij ( 1 9 7 0 ~ )and Wickerham (1970), S . bayaws S. carlsberEensis
S . carlsbergensis S. heterogenicus
S . inusitatus
S . italicus S . ouiformis S . steineri S . uvarum S . willianus
S . ellipsoideus S . ellipsoideus K . veronae S . cereuisiae
Miscellaneous 70
80
90
100%
Matching coefficient
FIG.3. Taxonomic relationships of common brewery culture yeasts (Sacchuromyces and Kluyoeromyces) and contaminants, summarized from (mainly unpublished) results used by CampbelI ( 1972a).
146
I. CAMPBELL
and therefore allowing for all known variations, gave results that were essentially the same ( Campbell, 1973a). Such differences as occurred between Poncet’s and our assessments of the genera Pichia and Hansenula were probably a result of the difficulty, discussed elsewhere ( Campbell, 1973a), of drawing dividing lines between groups. The separation of the genera Pichia and Hansenula by ability to assimilate nitrate has been criticized by various taxonomists. Indeed the definitions of the genera given by Lodder and Kreger-van Rij (1952) or by Kreger-van Rij (197Oc) and Wickerham (1970) suggest that assimilation of nitrate is the only difference. Attempts to classify by other properties produced groups essentially the same as before (Mrak et al., 1942; Kudriavzev, 1960). This view was confirmed by numerical analysis of the two genera (Campbell, 1973a), although there was certainly no clear distinction. Numerical analyses of the 25 species of Hansenula defined by Wickerham (1970) suggested that only 15 were sufficiently different for classification as separate species; it was also suggested that Kreger-van Rij’s ( 1 9 7 0 ~ )35 species of Pichia be reduced to 23. The analyses also included new species described since the preparation of Lodder’s ( 1970) book. Sixty species of Debaryomyces, Endomycopsis, Pichia, and the new genus Ambrosioxyma (van der Walt, 1972) were reduced to 49 species, although the genus to which they could be allocated was not obvious. In the reclassification of yeastlike fungi by von Arx ( 1972), nitrate-assimilating species of Endomycopsis ( E . bispora, E. muscicola, and E. platypodis) were left in the genus Hansenula, as Wickerham ( 1970) had suggested. Nitrate-negative species of Endomycopsis in Kreger-van Rij’s (1970b) list were dispersed to the genera Ambrosioxyma, Endomyces, Guilliermondella, and Saccharomycopsis. Unfortunately, numerical analysis of Pichia and related filamentous genera (campbell, 1973a) indicated that the distinction between genera was not as clear as von Arx (1972) had suggested. Analyses of the genera Hansenula and Pichia (Campbell, 1973a) showed that although Endomycopsis and Pichia species were sufficiently close to be synonymous, there were no sufficiently close relationships to suggest synonymy of species at present allocated to Hansenula and Pichia. The genera were closely related, as shown in Fig. 4. Species of Debaryomyces included in that analysis (Campbell, 1973a) were synonymous with Pichia species, and a merger of these genera was therefore proposed (Fig. 5). Ill.
Comparison of Numerical and Classical Taxonomy
Taxonomy of yeasts was originally based mainly on the morphological properties of the vegetative cells and spores, with only a restricted range
NUMERICAL TAXONOMY OF YEASTS
147
H . wingei H . bimundalis H . cnli f o m i c n
P. wickerhamii P . rhoilanensis
P . ioletnna
1- '
P. hovis
0H. plntypodis A . philen toma
E . monospm-a
P. acacine E . .fibidigel-n I
I
I
70
80
90
I
100%
Matching coefficient
FIG.4. Relationships between species of Ambrosiozyma, Endomycopsis, Hansenuk, and Pichia (Campbell, 1973a).
of biochemical tests to distinguish fermentative from nonfermentative or poorly fermentative genera. The range of assimilation tests proposed by Wickerham (1946; Wickerham and Burton, 1948), especially of up to 40 carbon compounds as sole source of carbon, furnished a more D. cantarellii
D . phaffii
I
1
70
1
80
I 90
1 P. pseudopolymorpha 1
D . uanrijii
I
P. polymorpha
]
D . marama
I
D. hansenii
I 100%
Matching coefficient
FIG. 5. Relationships between species of Debaryomyces, Pichia, and Wingea (Campbell, 1973a).
148
I. CAMPBELL
extensive series of identification properties, but even so Lodder and Kreger-van Rij (1952) used only assimilation and fermentation of seven sugars, assimilation of nitrate and ethanol, and splitting of arbutin as biochemical tests. The 1952 system avoided the criticism (Barnett, 1965) that groups of carbon compounds of Wickerham and Burton’s (1948) list are linked through related or identical metabolic pathways, a fault which has persisted to Lodder’s (1970) current classification scheme. It is easy to imagine, from differences between strains in one, two, three, or four reactions in a large number of tests that one has different species. In reality, four differences represent only a 6% difference if a total of 66 characters is considered. Yet numerical analysis of the genus Saccharomyces ( Campbell, 1972a) demonstrated several examples of authors defining separate species from the same environment when the differences on subsequent analysis were too slight to justify more than one species. Thus, a group of four melibiose-fermenting species, S. oleaceus, S. oleaginosus, S. hienipiensis, and S . norbenis (van der Walt, 1970c), all described as contaminants in the manufacture of olive oil, were related at over 90%matching coefficient and therefore were classified as one species, S. oleaginosus (Campbell, 1972a). Kluyveromyces africanus, K . phafii, S . dairensis, S. saitoanus, S. transvaalensis, and S. unisporus were insufficiently different to be classified as separate species (Campbell, 1972a), despite their classification (van der Walt, 1970b,c) in separate genera, and were therefore classified as S. dairensis, the description of which had to be modified in only four properties to include the five formerly separate species. K . veronae of van der Walt’s (1970b) classification was shown to be closer to the S . cerevisiae group than to other Kluyveromyces species. ( Campbell, 1972a) . Indeed, the genus Kluyveromyces, separated from Saccharomyces mainly by properties of spores (van der Walt, 1970b,c) was indistinguishable from the genus Saccharomyces. Although there are other examples of apparently separate genera which overlap in numerical analysis, e.g., Pichia and Hansenulu ( Campbell, 1973a),in the case of Kluyveromyces and Saccharomyces there were no properties, equivalent to assimilation of nitrate by Hansenula, which could be associated with either genus. Van der Walt (1970a,b) listed reniform spores or early liberation of spores from the asci as typical of Kluyveromyces, but since these properties were associated with species typical of Saccharomyces in all other respects, such a distinction was no longer valid. The remaining 12 of the 18 former species of Kluyveromyces were distinct from Saccharomyces species, at S, approximately 70%,but this was regarded as insufficient to justify a separate genus (Campbell, 1972a). Sneath (1962) suggested subdivision of taxa at 85%and 65%similarity coefficient. Silvestri et al. (1962) similarly suggested a division at approx-
NUMERICAL TAXONOMY OF YEASTS
149
imately 85%.Neither Sneath nor Silvestri stated whether they considered species or varieties to be distinguished at these levels. Such caution is wise in bacteriology, as it is difficult to decide what constitutes a bacterial species ( Cowan, 1962, 1970). Even in yeasts, in which a test of interbreeding is possible, the definition of species is difficult. Successful hybridization ( Mortimer and Hawthorne, 1969) between S . bayanus, S . carlsbergensis, S. chevalieri, S. chodati ( = S. italicus), S . diastaticus, S. italicus, and S. oviformis ( = S. bayanus) may indicate that they constitute a single species. A separate mating group K . lactis, K . fragilis, K . dobzhanskii and Zygosaccharomyces ashbyii ( = S . marxianus) (Wickerham and Burton, 1956; Mortimer and Hawthorne, 1969) may also be a single valid species. The close interrelationships within each group (Campbell, 1972a) is in accord with this supposition. Groups within the genus Saccharomyces at 85-90% matching (Campbell, 1970, 1972a) may be varieties rather than distinct species. It may eventually be possible to define species by numerical taxonomy, but at present such a system is incompatible with the current (Lodder, 1970) form of standard description of species ( Campbell, 1972a). All species except S. bayanus, S. cerevisiae, S. uvarum, S. rouxii, and K . lactis were defined as clusters within 85-90% matching coefficient ( Campbell, 1972a). K . Zactis, K . marxianus ( including K . fragilis), and K . dobzhanskii ( including K . drosophilarum) were transferred to the genus Saccharomyces from van der Walt’s (1970b) genus Kluyveromyces, but as three independent species. There is a strong possibility that in separation of species as clusters at 90% matching, related at approximately 85% matching coefficient, the division was placed too high. The combined genus Saccharomycesl Kluyveromyces could equally well have been described as 15 species or clusters at 85% matching, separated at S, 80%(Campbell, 1972a), but with such a classification, 15 satisfactory descriptions of species could not be provided. An alternative form of description of species, by the average properties around which limited variation is tolerated (Campbell, 1973b), may allow a revision of the genus Saccharomyces more in accord with genetic evidence ( Mortimer and Hawthorne, 1969). If 8590% matching separates species, then the 65% limit suggested by Sneath (1962) should separate genera. This is obviously not true in yeasts. Although experience limited to the genus Saccharomyces suggested delimitation of genera at around 6570% (Campbell, 1972a), studies on nonfermentative genera (Campbell, 1973a) showed such an extensive overlap of genera that no satisfactory division could be suggested. Hamenula and Pichia overlapped so extensively that certain species of the two genera were related at S, 80% (e,g., H . bimundalis, H . californica, P . bovis, P . rhodunensis, and P. toletanu) (Campbell, 1973a,
>
150
I. CAMPBELL
and Fig. 4). Since the two genera were distinguishable by the nitrate test, the species remained separate, but such separation was an obvious contradiction of the Adansonian rule of equal weight to all characters. The genera Debaryomyces and Pichia showed an even closer relationship (Fig. 5), e.g., D. uanrijii and P . polymorpha were synonymous. These findings, and the difficulty of defining the genus Debaryomyces (Kreger-van Rij, 1970a,c), prompted the suggestion of synonymy of the genera Debaryomyces and Pichia (Campbell, 1973a), supported by the similarity of paramagnetic resonance (PMR) spectra (Gorin and Spencer, 1970) and DNA base ratios (Nakase and Komagata, 1970, 1971a-d) of Debaryornyces and closely related Pichia species. The species of HansenulQ which are closely related to Pichia species have similar PMR spectra and GC ratios, but it was not convenient to merge the genera. The group of poorly fermentative yeasts could not be divided clearly; there was a zone in the spectrum of species where species had equal right, in terms of numerical analysis, to a place in Pichia or Hansenula, and were allocated to one genus solely by assimilation of nitrate. There is no satisfactory definition, to the computer, of a genus, and the separate identity of Hansenula and abolition of Debaryomyces were proposed for the nonmathematical reason of convenience. The difference by assimilation of nitrate is easily determined in the laboratory. We may also comment here on the status of spores in classification of yeasts, since the number of characters contributed by the spores is small: none (Poncet, 1967) to five (Campbell, 1973a, in classification tests only). Kockova-Kratochvilova et al. ( 1966, 1968, 1969a, 1970) scored the number of spores per ascus and the percentage of cells which sporulated. Analysis of strains of corresponding sporing and nonsporing species (Campbell, 1971); e.g., P . membranaefaciens and Candida mycoderma (now C. ualida) indicated S , > 95%, i.e., identity. Dekkera bruxellensis and Brettanomyces bruxellensis were also synonymous ( Campbell, 1973a). IV.
Application of Computer Techniques to Identification
Various systems of computerized identification have been described in the last decade. The simplest is a punched-card system to match unknown strains with known species (Steel, 1962). The concept of each species at some point in multidimensional space was developed by Gyllenberg (1965) and Rypka et al. ( 1967) : an unknown culture was identified as the species which occupied the space to which the unknown was allocated by its observed properties; if the unknown was not within the range of a species it was classified as the closest species. Lapage et al. (1970) described a system for identification of gram-negative bacte-
NUMERICAL TAXONOMY OF YEASTS
151
ria by matching with known properties of each species. The analysis was based on those routine diagnostic tests which were easy to read and gave consistent results. Each unknown was identified as the species with the greatest likelihood of fitting the observed properties of the unknown. Therefore limited variation in the standard properties of each species was tolerated, which was an advantage over a simple punchedcard system. Where two or more species fitted the description of the unknown, the computer described a low likelihood of the strain belonging to any of these species, and further tests had to be performed. While this may be advantageous on occasion, the punched-card system allows the operator to allocate the unknown to the most closely-matched species. In a punched-card system for the genus Saccharomyces, equivocal results were resolved by allocating the unknown strains to the species which it matched in serological properties and fermentation tests ( Campbell, 1973b). Computer identification of gram-negative bacteria is simplified by the availability of diagnostic tables in which all properties are of equal importance (e.g., Edwards and Ewing, 1971). Identification of yeasts is confused by the lack of such information and by the inconsistent importance ascribed to such tests as assimilation of nitrate, and formation of pseudomycelium and spores (see Section I ) . At present there are difficulties in identification by what may seem to be the most logical system: identification of the genus at a first stage, and then identification of the species to which the isolate belongs. The close relationship between Hansenula and Pichia discussed above cannot be resolved without suitable weighting of nitrate assimilation as a test of fundamental importance. Unfortunately, having increased the contribution of nitrate assimilation to the identification routine, the computer is less able to deal with genera in which nitrate separates species ( Candida, Torulopsis) or is unimportant ( Brettanomyces) . Spores account for few characters in the program; this too is inconsistent with the importance of sporulation in currently accepted classification ( Lodder, 1970), but is in agreement with the criticism of importance of spores by Barnett (1960, 1971), Beech et al. (1968), and Campbell ( 1971). It appears, therefore, that computerized identification of yeasts must operate by recognition of species in a single operation ( Campbell, 1973a,b) . Barnett ( 1971), using a computer technique, analyzed the importance of the standard tests of Lodder (1970) in identification. The number of species differentiated by each of Lodder's tests showed fermentation of glucose to be a valuable test, but fermentation of galactose, sucrose and other commonly tested sugars to be of little diagnostic value. Assimilation tests provided more useful information. The best 13 tests were, in order of usefulness, fermentation of glucose and assimilation of nitrate
152
I. CAMPBELL
and maltose > assimilation of cellibiose, galactose, and raffinose > assimilation of xylose > assimilation of mannitol, erythritol, succinate, lactate, and sucrose and requirement for vitamins. Barnett’s analysis, which considered the whole range of yeasts, may not be applicable to a specialized group such as the genus Saccharomyces, in which species are differentiated by fermentative properties which are not necessarily reflected in the assimilation results (van der Walt, 1 9 7 0 ~ ) On . the other hand, it has been suggested as a result of numerical analysis of the genus Saccharomyces ( Campbell, 1972a) that these minor differences are insufficient justification for separate species, thereby supporting Barnett’s hierarchy of useful tests. Buhagiar and Barnett (1971) described identification of the yeast species of strawberries by such computer-selected tests; we may regard this as a first step to complete identification of yeast isolates by computer. Numerical analysis of yeast species should lead to a more logical taxonomy, from the computer’s point of view. Examination of Saccharomyces, Hansenula, and Pichia ( Campbell, 1973a,b) has yielded promising results. Having defined species by an alternative form of standard or -, and never description in which all properties are regarded as variable, slow, latent, or acquired, new isolates were easily allocated to the most closely matched species. This system has been shown to work by comparison between punched cards of known species and unknown isolates (Campbell, 1973b); but we have not reexamined the strains by the more sophisticated comparison of Lapage et al. (1970). In our present system of computer identification (see Campbell, 1973a,b) two decks of cards are used, one for species which ferment glucose (and possibly also other sugars) and one for nonfermentative species. It is usually obvious from the results of identification tests which set of cards should be used. Batches of up to 40 unknown yeasts may be processed simultaneously, but in normal practice the number is very much less. The printed identification is represented in Fig. 6. If no suitable identification is obtained at S, > 85%, a culture is retested with the alternative set of cards. Failure to identify on this second test means either that the results of the identification tests on the unknown are faulty in some respect, or that the species is outside the range for which cards have been prepared. Since our collection of cards at present (Campbell, 1973a,b) does not include the genera Candidu, Cryptococcus, Rhodotorula, or Torulopsis, or any of the bipolar or fission yeasts, failure to match all unknowns with species is not surprising. The distinction between classification and identification is obscure at times. In recent years, techniques of numerical taxonomy have been applied to ecological work, thus reducing the temptation to define new species on the basis of a few differences from existing species. Dorfwirth
+
NUMERICAL TAXONOMY OF YEASTS
153
95-1008 Matching of Strain N566 with Strain N585 = 95.24 95-100% Matching of Strain N585 with Strain SROS = 95.24 N566 with Strain N566 with Strain N696 with Strain N700 with Strain
90-95% 90-95% 90-958 90-958
Matching of Strain Matching of Strain Matching of Strain Matching of Strain
85-90% 85-90% 85-908 85-90% 85-90%
Matching of Strain N415 with Strain Matching of Strain N415 with Strain Matching of Strain N415 with Strain Matching of Strain N543 with Strain Matching of Strain N566 with Strain
N696 = 90.48
SROS = 93.65 SROS = 90.48 SITA = 92.06 SBAY = 85.71 SHET = 88.89
SWIL = 87.30 SKLY = 87.30 SDLB = 85.71
FIG.6. Specimen printout of identification of Saccharomyces species ( Campbell, 197313). Identification of strains: N415 = S. heterogenicus, N543 = S . kluyueri, N566, N585 and N696 = S . msei, N700 = S . ita2icus. Note ( a ) that only the first 5 lines of the 8 5 4 0 % matches are printed, but the computer continued to print all matches to 75%; ( b ) that all matches are printed for any one strain, and the highest is accepted as identification. Therefore N415 is S . heterogenicus, not S . bayanus or S . willianus.
(1971) described the spread of properties of the various strains isolated from mead. A similar range of properties of single species was shown by numerical analysis of yeasts from sherry flor (Campbell, 1972b). Pignal, defining P. heimii (1970), justified its status as a new species by its taxonomic distance from existing species. Other taxonomists could profitably follow her example. Much work remains to be done in numerical taxonomy of yeasts and on the use of this information in computerized identification. The culture yeasts and common contaminants of the fermentation industries have obtained adequate attention ( Kockova-Kratochvilova et al., 1966, 1968, 1969a, 1970; Campbell, 1970, 1972a). Other genera have been virtually neglected. The limited attention to the genus CandicEa has shown that species have been defined unnecessarily, e.g., C. lusitaniae and C . obtusa were indistinguishable from C. parupsilosis ( Kockova-Kratochvilova et al., 1969b), and C . albicans and C . stellatoidea were indistinguishable (Schechter et al., 1972). In an earlier examination of the C. tropicalis group by Kockova-Kratochvilova et al. (1967)) C . intermedia was shown not to be a true species, but a collection of borderline forms of various related species of the group. Also, C . tropicalis var. larnbica did not belong to the parent species; it is interesting that van Uden and Buckley (1970) allocated the variety lambica to C. sake. This encouraging start with the genus Candida suggests that in the near future a classification
154
I. CAMPBELL
and identification scheme based on numerical analysis will be available for all yeast genera, and routine identification of yeasts should be simplified. Since the industrial microbiologist should benefit from such an improvement, you now know why a paper, apparently on theoretical aspects of yeast taxonomy, has appeared in Advances in Applied Microbiology. REFERENCES Ballou, C. E. (1970). J . Biol. Chem. 245, 1197-1203. Barnett, J. A. (1960). Nature (London) 186, 449451. Barnett, J. A. (1968). In “The Fungi” (G. C. Ainsworth and A. S. Sussman, eds. ) Vol. 3, pp. 557-595. Academic Press, New York. Bamett, J. A. ( 1971). Nuutre (London) 232, 221-223. Beech, F. W., Davenport, R. R., Goswell, R. W., and Burnett, J. K. (1968). In ‘‘Identification Methods for Microbiologists: Part B” (B. M . Gibbs and D. A. Shapton, eds.), pp. 151-175. Academic Press, New York. Boidin, J., Pignal, M.-C., and Besson, M. (1965). Bull. SOC. Mycol. Fr. 81, 566-606. Brady, B. L. (1965).Antonie van Leeuwenhoek; 1. Microbiol. Serol. 31, 95-102. Buhagiar, R. W. M., and Bamett, J. A. (1971). 1. Appl. Bucteriol. 34, 727-739. Campbell, I. (1970). 1. Gen. Microbiol. 63, 189-198. Campbell, 1. (1971). I. Gen. Microbiol. 67, 223-231, Campbell, I. (1972.a). 1. Gen. Microbiol. 73, 279-301. Campbell, 1. (1972b). J. Inst. Brew. 78, 491-496. Campbell, I. (1973a). J. Gen. Microbiol. 77, 427-441. Campbell, I. (197313). 1. Gen. Microbiol. 77, 127-135. Cowan, S . T. (1962). Synap. SOC.Gen. Microbiol. 12, 433-455. Cowan, S. T. (1970). I . Gen. MicrobioZ. 61, 145-154. Dorfwirth, K. ( 1971). Mitteilungen Versuchssta. Gaerungsgewerbe Wien 25, 208-211. Edwards, P. R., and Ewing, W. H. (1971). “Identification of Enterobacteriaceae,” 3rd ed. Burgess, Minneapolis, Minnesota. Gilliland, R. B. (1969). Antonie uan Leeuwenhoek; J . Microbiol. Serol. 35, 13-23. Gilliland, R. B. (1971). I. Inst. Brew. 77, 276-284. Gorin, P. A. J., and Spencer, J. F. T. (1970). Aduan. Appl. Microbiol. 13, 25-89. Gyllenberg, H. G . (1965). J. Gen. Microbiol. 39 401-405. Kockova-Kratochvilova, A., Pokorna, M., and kandula, J. ( 1966). Folia Microbiol. (Prague) 11, 188-199. Kockova-Kratochvilova, A, iandula, J., Vojtkova, A., Pokorna, M., and Stuchlik, V. ( 1967). Folia Microbiol. (Prague) 12, 327-344. Kockova-Kratochvilova, A., Vojtkova-LepSikova, A, Sandula, J., and Pokorna, M. ( 1968). Folia Microbiol. (rrague) 13, 300-309. Kockova-Kratochvilova, A., Sandula, J., Sedlarova, L., Vojtkova-LepBikova, A., and Kasmanova, M. ( 1969a). “Taxometric Study of the Genus Sacchuromyces ( Meyen) Reess,” Part I. Slovak Aca$ Sci., Bratislava. Kockova-Kratochvilova, A., Sandula, J., and Vojtkova-LepBikova, A. ( 196913). FoZilr Microbiol. (Prague) 14, 239-250. Kockova-Kratochvilova, A., Sedlarova, L., Vojtkova-Lepgikova, A., and iandula, 3. ( 1970). “Taxometric Study of the Genus Saccharomyces ( Meyen) Reess. Sacchuromyces cereuisiue and Related Species,” Part 2. Slovak Acad. Sci., Bratislava.
NUMERICAL TAXONOMY OF YEASTS
155
Komagata, K., Nakase, T., and Katsuya, N. (1964). 1. Gen. Appl. Microbiol. 10, 313-321. Kreger-van Rij, N. J. W. (1969). In “The Yeasts” (A. H. Rose and J. S. Harrison, eds. ), Vol. 1, pp. 5-78. Academic Press, New York. Kreger-van Rij, N. J. W. (1970a). In “The Yeasts, a Taxonomic Study” ( J . Lodder, eds. 1, 2nd ed., pp. 129-165. North-Holland Publ., Amsterdam. Kreger van Rij, N. J. W. (1970b). In “The Yeasts, a Taxonomic Study” (J. Lodder, ed.), 2nd ed., pp. 166-208. North-Holland Publ., Amsterdam. Kreger-van Rij, N. J. W. ( 1 9 7 0 ~ )In . “The Yeasts, ‘a Taxonomic Study” (J. Lodder, ed. ), 2nd ed., pp. 455-554. North-Holland Publ., Amsterdam. Kudriavzev, V. I. (1960). “Die Systematik der Hefen.” Berlin [cited by Kreger-van Rij (1970a) and Wickerham (1970)l. Lapage, S. P., Bascomb, S., Willcox, W. R., and Curtis, M. A. (1970). In “Automation, Mechanization and Data Handling in Microbiology” (A. Baillie and R. J. Gilbert, eds.), pp. 1-22. Academic Press, New York. Lame, T. A., and Spencer, J. F. T. (1968). Can. 1.Microbiol. 2nd ed. 14, 79-86. Lodder, J., ed. (1970). “The Yeasts, a Taxonomic Study.” North-Holland Publ., Amsterdam. Lodder, J., and Kreger-van Rij, N. J. W. (1952). ‘The Yeasts, a Taxonomic Study,” 1st ed. North-Holland Publ., Amsterdam. Meyer, S. A., and Phaff, H. J. (1969). I. Bacteriol. 97, 52-56. Mills, C., Chiid, J. J., and Spencer, J. F. T., (1971). Antonie van Leeuwenhoek; I. Microbiol. Serol. 37, 281-287. Morris, E. O., and Eddy, A. A. (1957). 1. Inst. Brew. 63, 34-35. Mortimer, R. K., and Hawthorne, D. C. (1M9). In “The Yeasts” ( A . H. Rose and J. S. Harrison, eds.), Vol. 1, pp. 385460. Academic Press, New York. Mrak, E. M., Phaff, H. J., Vaughn, R. H., and Hansen, H. N. ( 19‘42). J. Bucteriol. 44, 441450. Nakase, T., and Komagata, K. (1970). J. Gen. Appl. Microbiol. 16, 511-521. Nakase, T., and Komagata, K. (1971a). J. Gen. Appl. Microbiol. 17, 43-50. Nakase, T., and Komagata, K. (1971b). J. Gen. Appl. Microbiol. 17, 77-84. . Gen. Appl. Microbiol.17,227-238. Nakase, T., and Komagata, K. ( 1 9 7 1 ~ )1. Nakase, T., and Komagata, K. (1971d). I. Gen. Appl. Microbiol. 17, 259-279. Phaff, H. J., and Ahearn, D. G. (1970). In “The Yeasts, a Taxonomic Study” (J. Lodder, ed.), 2nd ed. pp. 1187-1223. North-Holland Publ., Amsterdam. Phaff, H. J., and Fell, J. W., (1970). In “The Yeasts, a Taxonomic Study” (J. Lodder, ed. ), 2nd ed. pp. 1088-1145. North-Holland Publ., Amsterdam. Pignal, M.-C. (1970). Antonie uan Leeuwenhoek; 1. Microbiol. Serol. 36, 525-529. Poncet, S. (1967). Antonie van Leeuwenhoek; J . Microbiol. Serol. 33, 345358. Poncet, S. (1970). Ann. Inst. Pasteur, Paris 119, 232-248. Rypka, E. W., Clapper, W. E., Bowen, I. G., and Babb, R. (1967). 1. Cen. Microbiol. 46, 407-424. $and&, J., Kockova-Kratochvilova, A., and Zameknikova, M. ( 1964). Brauwissenschclft 17, 130-137. hndula, J., Masler, L., Sikl, D., and Vojtkova-LepBikova, A. (1972). In “Proceedings of the First Specialized Symposium on Yeast, Bratislava” ( A. Kockova-Kratochvilova and E. Minarik, eds.), pp. 195-206. Slovak Acad. Sci., Bratislava. Schechter, Y., Landau, J. W., and Dabrowa, N. (1972). Mycologia 64, 841-851. Scheda, R., and Bos, P. (1966). Nature (London) 211,660. Scheda, R., and Yarrow, D. (1966). Arch. Mikrobiol. 55, 209-225. Scheda, R., and Yarrow, D. (1968). Arch. Mikrobiol. 61, 310416.
158
I. CAMPBELL
Silvestri, L., Turri, M., Hill, L. R., and Gilardi, E. (1962). Symp. SOC. Gen. Microbiol. 12, 333-360. Sneath, P. H. A. (1957). J. Gen. Microbwl. 17, 201-226. Sneath, P. H. A. (1962). Symp. SOC.Gen. Microbiol. 12, 289-332. Sneath, P. H. A. (1972). In “Methods in Microbiology” (J. R. Norris and D. W. Ribbons, eds.), Vol. 7A, pp. 29-98. Academic Press, New York. Sneath, P. H. A., and Johnson, R. (1972). J. Gen. Microbiol. 72, 377-392. Sokal, R. R., and Sneath, P. H. A. (1963). “Principles of Numerical Taxonomy.” Freeman, San Francisco, California. Steel, K. J. (1962). Symp. SOC. Gen. Microbiol. 12, 405432. Stenderup, A., and Leth Bak, A. (1968). J. Gen. Microbiol. 52, 231-236. Summers, D. F., Grollman, A. P., and Hasenclever, H. F. (1984). J. Immunol. 92, 491499. Suzuki, S., Sunayama, H., and Saito, T. (1988).Jup. J. Microbiol. 12, 19-24. Tsuchiya, T., Fukazawa, Y., and Kawakita, S. ( 1985a). Mycopathol. Mycol. Appl. 26, 1-15. Tsuchiya, T., Fukazawa, Y., and Kawawakita, S., Imai, M., and Shinoda, T. (196513). lap. 1. Microbiol. 9, 149-159. van der Walt, J. P. (1970a). In “The Yeasts, a Taxonomic Study” (J. h d d e r , ed. ), 2nd ed., pp. 34-113. North-Holland Publ., Amsterdam. van der Walt, J. P. (1970b). In “The Yeasts, a Taxonomic Study,” ( J . Lodder, ed.), 2nd ed., pp. 316-378. North-Holland Publ., Amsterdam. van der Walt, J. P. ( 1 9 7 0 ~ ) .In “The Yeasts, a Taxonomic Study” (J. Lodder, ed. ), 2nd ed., pp. 556718, North-Holland Publ., Amsterdam. van der Walt, J. P. (1972). Mycopathol. Mycol. Appl. 46, 305-316. van Uden, N., and Buckley, H. (1970). In “The Yeasts, a Taxonomic Study“ (J. Lodder, ed.), 2nd ed., pp. 893-1087. North-Holland Publ., Amsterdam. van Uden, N., and Vidal-Leiria, M. (1970). In “The Yeasts, a Taxonomic Study” (J. Lodder, ed.), 2nd ed., pp. 1235-1308. North-Holland Publ., Amsterdam. von Arx, J. A. ( 1972). Antonie uan Leeuwenhoek; J . Microbiol. Serol. 38, 289-309. Wickerham, L. J. (1946). J . Bacteriol. 52,293-301. Wickerham, L. J. (1970). In “The Yeasts, a Taxonomic Study,” ( J . Lodder, ed.), 2nd ed. pp. 226415. North-Holland Publ., Amsterdam. Wickerham, L. J., and Burton, K. A. (1948). J . Bacterwl. 56, 363471. Wickerham, L. J., and Burton, K. A. (1956). J . Bacteriol. 71, 290-295.
Microbiology and Biochemistry of Soy Sauce Fermentation
F. M . YONG’ AND B. J . B. WOOD Department of Applied Microbiology. Uniuersity of Strathclyde. Glasgow. Scotland Introduction ................... Fermented Soy Products ............. History of Soy Sauce Production ......... Chemical Composition of Soy Sauce ........ Raw Materials ................. A. Soybeans .................. B . Wheat ................... C. Ratio of Soybeans to Wheat ......... D . Salt ..................... E . Substitute Raw Materials ........... VI . Treatment of Raw Materials ............ VII . Koji ...................... VIII . Culturing the Koji ................ IX . Mash (Moromi) ................. A . Preparation (Mashing) ............ B . Control of the Mash ............. C. Aging ................... D . Microbiology of Mash ............ X. Pressing .................... XI . Pasteurization .................. XI1. “Chemical” Soy Sauce .............. XI11. Semichemical Soy Sauce. or Shinshiki S h o p ..... XIV . Future Development in the Soy Sauce Industry ... xv. Conclusions ................... References ...................
I. I1. I11. IV. V.
157 159 161 163 165 165 167 167 168 168 169 171 173 175 175 176 177 177 182 182 183 184 185 188 188
.
I Introduction Soy sauce has occupied a place of honor as a condiment in orientaI cuisine since time immemorial and is now finding its way into the occidental kitchen In the course of our studies on the biochemistry and microbiology of this fascinating fermentation. we became aware that there is no complete modern review of the subject readily available to the Western worker. The sequel is an attempt to correct this lack. True soy sauce is the product of a complex fermentation in which a mixture of soybeans and wheat flour are inoculated with a mold; the mixture is incubated for about 3 days. and. when a good growth of mold mycelium has taken place. the resulting mass is placed in a brine An anaerobic fermentation. in which yeasts and bacteria participate. then occurs. The liquid obtained from this stage is the soy sauce of commerce
.
.
.
Present address: Singapore Institute of Standards and Industrial Research. Republic of Singapore .
158
F. M. YONG AND B. J. B . WOOD
Inferior products which lack the rich, satisfying flavor of true soy sauce are prepared by acid hydrolysis of bean and wheat mixtures; unfortunately a substantial proportion of the “soy sauce” sold in the United Kingdom is of this variety. The fermentation may perhaps owe its origin to the need, in time of plenty, for preserving excess foodstuffs against periods of scarcity. Drying and smoking of meat were man’s first methods of conserving his food supplies, discovered purely by chance. Later, it is likely he discovered the use of salt (sodium chloride) in combination with the drying and smoking methods. Not knowing what actually did take place one may speculate that it is also by accident that mankind discovered that meat and plant materials could be preserved for long periods by the addition of salt. Thus, probably the first fermentations were discovered accidentally when salt was incorporated with the food material and the salt selected certain harmless microorganisms which fermented the raw material to give a nutritious and acceptable food. Speculating along these lines, we might expect that the first fermented product in the Orient may have been fish sauce. With the advent of religions in which meat was excluded from the diet, the use of salt and fermentation techniques was adapted to plant materials. The discovery of soy sauce fermentation may have been due to the introduction of Buddhism to China, Korea, and Japan. Soy sauce is the term applied to a salty condiment made by a two-stage fermentation, in which the first stage is aerobic and the second stage is anaerobic. In the first stage, a mixture of soybean and wheat is inoculated with an Aspergillus culture. The mold grows and releases enzymes that break down the proteins and oligosaccharides present in the substrate, The second stage is begun when the mold growth has reached a desired level. The material is then added to a salt solution and an anaerobic fermentation is initiated in which lactobacilli and yeasts predominate. At the end of this stage, the liquid which is pressed off and clarified is soy sauce. Soy sauce fermentation was at one time a household art and secret. The people at that time had no idea what happened except that soybeans could be fermented into a product with a completely different odor, appearance, and taste, but the product remained wholesome. Although soy sauce fermentation is still to some extent carried out as small operations in the traditional way in the home, the art being passed on from father to son, it is one of the most modern and sophisticated fermentation industries in the Orient, especially in Japan. Most of the research on soy sauce has been and is being conducted in Japan. As early as the thirteenth century, especially during the Muromachi period ( 1328-1573), many investigations were carried out on the kinds
SOY SAUCE FERMENTATION
159
and relative quantities of raw materials and on the fermentative process, with the result that the product was improved considerably (Yokotsuka, 1960). There is currently an extensive amount of published work on different aspects of soy sauce fermentation. Judging from recent reviews by Umeda (1957)) Mogi and Iguchi (1958, 1965, 1966), Mogi (1959), Yokotsuka (1960), Akira (1967), Onishi ( 1967, 1968), and personal communications with colleagues from Korea, Japan, and Hong Kong, the modern methods now in use there are only ad hoc modifications of the traditional process. The present method is still time-consuming, low in efficiency, and without adequate microbiological control. The authors are of the opinion that the next few years will see rapid progress in the development of improved technology for the production of soy sauce, leading to a process which is much more economical of time, space, and labor, but without any sacrifice of the flavor and quality of the product. Soy sauce fermentation is at present a two-stage batch process involving the biochemical activities of three types of microorganisms: mold, bacteria, and yeast. The first stage is done by growing Aspergillus oryzae or Aspergillus soyae on a mixture of soybeans and wheat. The extracellular amylases and more importantly the proteases hydrolyze the carbohydrates and proteins in the raw materials, respectively. When mold growth has reached the desired level, the mixture of soybean and wheat covered by mold mycelium is placed into 18%w/v sodium chloride solution. The mixture undergoes lactic acid and yeast fermentations for at least one year at ambient temperatures to give a product of good quality. ti.
Fermented Soy Products
The use of various bacteria, yeasts, and fungi to make excellent fermented foods from soybeans or a mixture of soybeans and cereals, such as rice or wheat, has been known in the Orient for centuries. Hesseltine ( 1965, 1966), Hesseltine and Wang ( 1967), and Gray ( 1970) have given comprehensive accounts of the different kinds of fermented soybean, soybean-and-wheat, and soybean-and-rice products; the types of micrcqorganisms; and the processes of making these foods. Table I, adapted from Hesseltine (1985)) lists the name( s ) of the fermentation, the chief organism( s ) used, the fermentation substrate, the nature of the product, and the country or countries where the fermentation is now carried on. From Table I, two very evident conclusions can be drawn: ( a ) fungi are the organisms most frequently used to process soybeans; and ( b ) the areas where the articles mentioned are of commercial interest lie in the Orient.
160
F. M. YONG A N D B. J. B. WOOD
TABLE I FOODFERMENTATIONS' Name of food Organism(s) used
Substrate
Nature of product
Soybeans
Solid
Soybeans
Solid
Area of commercial use
Of fungal origin
Tempeh Sufu (Soybean cheese)
Rhizqpus ( R . oligosporus) Principally Actinomucor elegans, Mucor species Aspergillus oryzae, Saccharomyces rouzii
Miso (breakfast food and soup base) Shoyu (Soy Aspergillus oryzae, Lactobacillus, sauce) Hansenula, and Saccharomyces
Rice and other Paste cereals plus soybeans
Indonesia and vicinity China, Formoea
Soybeans, wheat
Liquid
Japan, China, and some other parts of the Orient China, Japan, Philippines, and some other parts of the Orient
Soybeans
Solid
Japan
Of bacterial origin Natto (condiment) a
Bacillus subtilis
Adapted from Hesseltine (1965).
The Orient is distinguished from the Occident not only on geographical and ethnological grounds, but also on the basis of the types of food that are processed by fungi. In the Occident, fungi are used in the processing of milk protein (i.e., certain cheeses) whereas in the Orient fungi are used mainly to process soybeans, although a variety of other materials, including rice, wheat, peanuts, copra, and fish, are also so processed. Moreover, the major past and present contributions of fungi to man's food supply must be sought in the Orient (Gray, 1970). Of the many different fermented foods from soybeans, the soy sauce fermentation is both very typical and probably the most widespread. It is an article of immense commercial importance in the Orient and is gradually but certainly being accepted by the peoples of the Occident. The total annual production of soy sauce in Japan was estimated by Yokotsuka (1964) to be about one million kiloliters. This would mean that the yearly consumption per capita is about 10 liters in Japan. Unfortunately no details are available with regard to the total annual production in the People's Republic of China, Taiwan, and the nations of South-
SOY SAUCE FERMENTATION
161
east Asia. It has, however, been reported that soy sauce represents one of China’s largest uses of soybeans (Smith, 1949). Soy sauce is used daily in the kitchen in the preparation of food and also as a table condiment. At most meals a dish of soy sauce is placed on the table, and certain foods are dipped in the soy sauce for seasoning. The sauce contains about 18%w/v of salt, hence serving as a salting agent as well as being a flavor accentuator. Although soy sauce is the product of the Oriental practice of modifying soybeans with fungi, lactic acid bacteria, and yeasts, which is best known in America and Europe, very few scientific reports on soy-sauce fermentation are available in Western journals. What actually is soy sauce? Hesseltine (1965), Hesseltine and Wang (1967), and Gray ( 1970) have described soy sauce under traditional fermented foods. However, in the strict sense of the word, soy sauce is not a food. It is not a solid nourishment for human consumption. It is a liquid food condiment which is used to add flavor and color to the bland, basic Oriental diet including rice, raw fish, beancwd, fermented beans, and boiled vegetables. To be more specific, it is a combination of hydrolytic products of soybeans alone or soybeans and wheat. Among the Chinese soy sauce is also known as Ch‘au yau “drawing oil,” or Pak yau “white oil” (Groff, 1919). The term s h o p is preferred in Japan and it appears for the first time in Japan in 1596 and in China in 1618 (Yokosuka, 1964). In the territories formerly known as the Dutch East Indies and Malaysia, soy sauce is known as ketjap (Stahel, 1946). 111.
History of Soy Sauce Production
The written records of the Chinese show that they have been using soy sauce for over three thousand years. In the book of Chau Lai (one of the Thirteen Classics of Confucius) written before 10o0 BC, we read that the king’s cook used twenty jars of soy sauce for the ceremonial rites of the Chau Dynasty (Groff, 1919). Hesseltine (1965) noted that, according to information supplied by a major shoyu manufacturer, shoyu has been produced in Japan for over lo00 years. Soy sauce fermentation probably started in Japan as a result of the introduction of Buddhism from China and the consequent change to a vegetable diet in 552 AD (Hesseltine, 1985). Buddhism was already well established in China and Korea by the fourth century before it was introduced into Japan between 500 and 600 AD (Bush, 1959). Although the making of soy sauce originated in China, Japan has probably the largest soy sauce plant in the world, the Noda Mati plant of the Kikkoman Shoyu Company Limited which was founded in 1764.
162
F. M. YONG AND B. J. B. WOOD
It has been estimated that the annual production of shoyu at Noda in 1948 was as much as 23 million gallons, consuming 30,000 metric tons of soybeans, 27,000 metric tons of wheat, and 29,000 metric tons of salt (Smith, 1949). He also notes that it had been claimed that the general occurrence, throughout the small city of Noda, of the mold used in fermenting shoyu suppresses other microbial life and maintains a high health standard for the city. Stahel (1946) made reference to the soy sauce made in the former Dutch East Indies but gave no indication as to when the fermentation of soy sauce was first started there. Descriptions of the process of producing soy sauce by the traditional, or orthodox, fermentative procedure are to be found in books written in Chinese more than 1500 years ago (Lockwood and Smith, 1950-1951). However, Hoffmann (1874) was the first to make available to the Occidental nations, directions, though very brief, for the manufacture of soy sauce by the fermentation process. Thereafter short essays appeared in Western journals on soy sauce and its manufacture (Kellner, 1888; Kita, 1913; Prinsen-Geerlings, 1917a,b; Waksman and Davidson, 1926; AIlen, 1926; Morikawa, 1926; Dyson, 1928; Ramsbottom, 1936; Minor, 1945; Stahel, 1946; Blaisten, 1947; Lockwood, 1947; Lockwood and Smith, 1950-1951; Hoogerheide, 1954; Hesseltine, 1965; Sakurai, 1965; Hesseltine and Wang, 1967,1968; Gray, 1970). In 1919 Groff gave a very comprehensive description of the then current procedures of soy sauce manufacture in Kwangtung, China. Thirty years later Smith (1949) published a report on Oriental methods of using soybeans as food. In his report he gave figures which illustrated the importance of soy sauce in the diet of people in China, Japan, and Korea, and also compared the state of technology in soy sauce making in these countries. Church (1923) published perhaps the most detailed piece of work on soy sauce found in Western literature. Yokotsuka (1960) has given the best review article on soy sauce to date in English. In this review he covered not only the various ingredients in soy sauce that are responsible for giving the sauce its particular meaty flavor and aroma, but also the technology and microbiology of soy sauce production in Japan. Formerly, the technology of soy sauce fermentation was a closely guarded family art passed on from father to son. Even now some manufacturers point with pride to the fact that their factories have been operated as family enterprises for 5 centuries. The technology was developed before the biochemical processes involved in the degradation of the raw materials by microorganisms to give soy sauce were known. Even now there is still much to be learned about the microbiology of soy sauce fermentation. The major steps involved in the manufacture
163
SOY SAUCE FERMENTATION
Soybeans (50 parts)
Mold (Aspergillus oryzae) grown on Whole wheat grains polished rice; 1-2 parts of approx. S d a y (50 parts) 1 Soaked in wat.er for culture used 12 hours Roasted
1
1 Boiled until soft
-------
1
1
Mixed
1
Spread on trays to a depth of 1-2 inches
.1 Incubated for approximately 72 hours in 30' room
1 Transferred to 50-gallon capacity earthen vessel
1
Approximately 20 % salt brine added (200 parts)
1
Lactic acid and yeast fermentation lasting for 1-3 years
1 Filtered -+ Residue -+ Used in animal feed
1 Filtrate .1
Pasteurized
2 Soy sauce
FIG.1. Flow sheet for the fermentation of soy sauce by the conventional method.
of soy sauce are no longer a secret, but the finer and important points are. Soy sauce is manufactured by two basic processes. The traditional method involves a fermentation technique; the quality of the product will in this case depend greatly on the personal experience of the human involved. The other, chemical method, came into existence at a much later date. Prior to 1912 soy sauce took one to three years to mature. It is therefore a time-consuming and expensive process (Yokotsuka, 1960). One can understand why it took such a long time by looking at Fig. 1, which gives all the essentials of the fermentation process. IV.
Chemical Composition of Soy Sauce
There are two major kinds of shoyu in the Orient, the Chinese type and the Japanese type. In China most kinds of shoyu are made from
164
F. M. YONC AND B. J . B. WOOD
soybeans alone or from a mixture of soybeans and wheat with a higher percentage of the soybeans; Japanese shoyu is made from equal amounts of soybeans and wheat. Table I1 (from Smith, 1949) shows the chemical composition of Chinese and Japanese soy sauce. Japanese law has regulated the quality of the shoyu exported from Japan since 1950 (Yokotsuka, 1964). The inspection of quality consists of both organoleptic and analytical tests involving the determination of specific gravity (degrees Baumk ), sodium chloride, extract, and total nitrogen. Good quality Japanese shoyu is said to contain about 1.5% w/v total nitrogen; of this about 45%is lower peptides, 45%is amino acids (including every essential one), and the remaining 10% is ammonia nitrogen. Yokotsuka (1964) is of the opinion that although shoyu is so rich in amino acids, its absorption by the gastrointestinal tract is limited by its high salt content, but this opinion has never been corroborated by experiments, and it would be interesting to know more of the extent of its nutritional contribution when it is a regular component of the diet. Good quality Chinese-type soy sauce is of high specific gravity and viscosity and high nitrogen conTABLE I1 CHEMICAL COMPOSITION OF CHINESE AND JAPANESE SOY
Item
Chinese Japanese soy sauce soy sauce from from Nanking Noda (gm/100 mi) (gm/100 ml)
Total solids Mineral matter Sodium chloride Phosphoric acids (as PIOa) Total nitrogen Protein nitrogen Nonprotein nitrogen Amino nitrogen Volatile acids (m acetic) Nonvolatile acids (as lactic) Total acidity Sugar (as glucose) Dextrins Vkcosityb Hydrogen ion concentration (pH) Specific gravity (at 15°C) a
b
Modified, after Smith (1949).
By Ostwald's U-tube viscometer at 25°C. Not determined.
SAUCE'
32 NDc 16 ND 1 ND ND ND 0.5 ND
38 20 18 0.50 1.5 0.1 1.4 0.7 0.1 0.6
0.7
ND
2 ND ND ND 1
6
1 5 4.5 1
165
SOY SAUCE FERMENTATION
tent; it is dark in color and is sometimes sweetened with cane sugar. This is different from Japanese-type shoyu, which is lower in viscosity and in total nitrogen content and is a lighter (but beautiful) red. Yokotsuka (1964)claims that in spite of the lower nitrogen content of Japanese shoyu, amino acids (especially glutamic acid) content is higher than in the Chinese type. Most characteristic of Japanese shoyu is its aroma and flavor. The Japanese have attributed this to the use of much wheat as raw material and to the strong yeast fermentation.
V.
Raw Materials
The raw materials used-soybeans, cheap and easily available.
wheat, and salt-are
relatively
A. SOYBEANS The soybean known botanically as Glycine m x ( L ) Merr. is also called soya bean, soja bean, Chinese pea, and Manchurian bean. It is native to Eastern Asia. The name soya came from the Chinese through the Japanese. It is taken from Chinese “Chiang-yiu,” which means soy sauce and is pronounced “show-yu” in Japanese. The Japanese contracted it into “so-ya” but with the fundamental characters for “chiang-yiu.” “So-ya” was further corrupted into “soy-a” or “soya” and in English into soybean ( Markley, 1950). There are slight differences in the chemical composition of the beans from different countries, as shown in Table III (adapted from Yokotsuka,
1980). Either whole beans or defatted soybean meal can be used in sauce TABLE I11 COMPOSITION OF SOYBEANS FROM VARIOUS SOURCES“
Country Japan China U.S.A.
Water content (%) 13 11 10
Total Protein nitrogen nitrogen6 (%) (%) 6 6 6
38 38 38
Crude fat (%)
Invert sugars
16 19 20
15 19 17
( %)
Adapted from Yokotsuka (1960). The carbohydrate content waa not determined by Yokotsuka (1960) but Kawamura (1967) has found it to be about 34% based on dry basis. Obtained by multiplying total nitrogen with the factor 6.25 (Lillevik, 1970). (I
166
F. M. YONG AND B. J. B. WOOD
manufacture. When whole beans are used an oily fraction can be separated from the fermented soy mash. This fraction is known as “soy-sauce oil,” which is used for making lower grade soap and as a source of linoleic acid. It was found by Kubo (1947) that soy sauce oil contains esters, chiefly ethyl esters of higher fatty acids produced during fermentation by the exchange of glycerol with ethyl alcohol. Soy sauce oil also contains 3040% free fatty acids (Yokotsuka, 1964). These esters and free fatty acids are inhibitory to yeasts. The soy sauce oil is separated from soy sauce by decantation before pasteurization. Little, if any, work seems to have been done on the changes in soybean oil occurring during the fermentation. For exampIe, it would be interesting to know if any change in degree of unsaturation of the fatty acids occurs. For the past 60 years defatted soybeans have been used to replace whole beans in the production of soy sauce. At first, pressed soybean meal was used to replace whole beans; now, however, solvent-extracted soybean meal is widely used. In recent years defatted soybean meal has replaced whole beans in about 75%of the shoyu production in Japan (Hesseltine and Wang, 1968). The main reason for using defatted soybean is that the cost is very much lower. Furthermore the utilization of nitrogen in defatted soybeans is higher. It has been found by practical experience that the fermentation of defatted soybean meal requires a shorter period; it takes about 15 months at ambient temperatures for whole beans, but only about 10 months for defatted soybean meal (Yokotsuka, 1960). The difference in this rate of fermentation has never been explained; it has only been observed. It may be due to the surface layer of oil affecting the rate of transfer of oxygen through the fermented soy mash, thus slowing down yeast growth, or to the yeast-static compounds that result when whole soybeans are used (Fukai, 1928), or to whole beans being less rapidly digested than the more finely divided meal. It is important to note that soy sauce made from whole beans is said to be more stable than that produced from defatted soybean meal. Yokotsuka (1960) did not explain exactly what he meant by stability. Presumably stability here would mean resistance to microbiological spoilage rather than chemical deterioration of the soy sauce on storage. Good-quality fermented soy sauce does not have a white (yeast) film on the surface when exposed to the atmosphere because it contains “yeast-static” or “yeast-cidal” compounds. Yokotsuka et al. ( 1958) recognized that soy sauce produced from whole beans has a greater resistance to yeast invasion than that produced from defatted bean meal. They found that raw soy sauce made from whole bean (meaning soy sauce that had not undergone any treatment either by heating or addition of preservatives) has the same storage stability as a pasteurized product.
SOY SAUCE FERMENTATION
167
Fukai (1928) and Fukai and Komatsu (1934a,b) studied the ethyl esters of fatty acids and free fatty acids, such as caproic acid, in soy sauce, and the toxicity of these compounds toward undesirable yeasts which may infect soy sauce. B. WHEAT
Either whole wheat or wheat flour may be used. Usually only soft wheat flour, i.e., wheat flour with about 8-9%w / w protein, is used. When wheat bran is used as a substitute for wheat flour or whole wheat, the resulting soy sauce is not satisfactory (C. T. Yeow, personal communication). Smith (1949) has reported that when whole wheat or wheat flour was unobtainable or too expensive in China, the Chinese substituted wheat bran, oats, kaoliang ( a root vegetable), and rye. Barley was not favored. Rice as a substitute was tried by Yenko and Baens-Arcega (1940) in the Philippines. The different starting materials modify to some extent the flavor of the resulting soy sauce. It is preferable to have the grain rich in starch and low in fiber. Yokotsuka (1964) listed the roles of wheat in soy sauce manufacture as follows: (1) To make the moisture content of the material to be cultured with mold just adequate for mold growth; it must be about 45% in order to minimize the damage due to the growth of undesirable bacteria. Cooked soybeans have about 60% moisture, so roasted and crushed wheat serves to decrease the moisture of the material. ( 2 ) To assist in obtaining the highest proteolytic activity from the koji; the activity is highest when the starting material is an equal mixture of soybeans and wheat, along with greater growth of the mold. ( 3 ) TO serve as the major source of carbohydrates as the precursor of sugars, alcohol, and organic acids. ( 4 ) To serve as the source of lignin and glycosides, the precursors of vanillic flavor of shoyu. (5) To serve as a rich source of glutamic acid.
C. RATIO OF SOYBEANS TO WHEAT The use of wheat decreases the total nitrogen content of soy sauce but it contributes aroma, flavor, and glutamic acid. The best soy sauce is generally believed to be made from a soybean-to-wheat ratio of 50:50 by weight based on the weight of the raw materials as received. The present method of adjusting the ratio of soybean-to-wheat has certain disadvantages. These arise because the moisture content and, more importantly, nitrogen and carbohydrate contents of each consignment may differ; perhaps it would be better to have a ratio based on protein-tocarbohydrate.
168
F. M. YONG AND B. J. B. WOOD
Considerable work has been done to find out the working ratio of soybeans-to-wheat that would provide conditions favorable for maximum mold growth but unfavorable for bacterial growth, so giving a good quality soy sauce ( Church, 1923; Ohara and Moriguchi, 1955a,b).
D. SALT Sodium chloride acts as a preservative and exerts a selective action on the microorganisms which grow in the fermentation. If salt were not present, a dangerous anaerobic bacterial fermentation would occur in the traditional process, and one might expect such organisms as Botulinus to grow on such neutral or slightly alkaline mixtures, just as they can on imperfectly processed cans of green beans, for example. But the role of salt is not confined to the purely negative one of preventing the growth of dangerous organisms. It is clear from many studies, including our own (Yong, 1971), that the salt is necessary to permit the exclusive development of the flavor and aroma-forming yeasts and lactic acid bacteria. Usually commercial, not chemically pure, sodium chloride is used for making soy sauce. Church (1923) noted that according to Japanese authority, experimental work had been successful with purified sodium chloride only occasionally, and commercial practice, never. It may well be that raw salt will carry a larger inoculum of halophilic and halotolerant bacteria than pure salt. In commercial salt a larger quantity of essential “impurities” necessary for yeast and bacterial growth may be present. Foreign substances other than basic salts of calcium and magnesium, often found in even a fair grade of bulk commercial salt, do not seem to interfere with shoyu making. Sea salt is used as a rule in Japan. In our own studies (Yong, 1971)) we employed British Drug Houses Limited’s “Laboratory Reagent Grade” sodium chloride, and found it to be entirely successful on a laboratory scale. The fact that we employed fermentations with added pure cultures of yeasts and Lactobacilli may, however, have influenced our results.
E. SUBSTITUTERAW MATERIALS There have been several attempts to make “soy sauce” using raw materials other than soybeans and wheat, but they were unsuccessful commercially. Church (1923) experimented with peanut press cake instead of the soybean and wheat mixture. The sauce obtained was called peanut sauce, not soy sauce, since “the taste of peanut was retained to such an extent that those accustomed to judging peanut products by tasting were not deceived, even when uninformed as to the ingredients
SOY SAUCE FERMENTATION
169
of the sauce.” The fermentation of peanut press cake on an industrial scale would not be easy. The control of bacterial growth and activity of bacteria harmful to the process is more difficult in a koji of peanut press cake than in a soybeans and wheat mixture. Soybeans are produced in the Philippines only in limited quantities; to meet the requirements of the manufacturers of soy sauce, soybean milk, and soybean oil, considerable quantities of beans are annually imported ( Baens-Arcega, 1969). As a cheap source of proteinaceous material Baens-Arcega (1966) claimed successful use of a 50-50 mixture of copra meal and soybeans in the mold process of manufacturing soy sauce. Kato and Matsumoto (1964) used soybean hull to make soy sauce. The Japanese were so desperate for raw materials during the Second World War that they even tried to produce soy sauce from garbage (Tsukahara, 1948). Oda et al. (1949) prepared a “substitute soy sauce” using acorns and wheat with an Aspergillus culture which produces tannase. VI.
Treatment of Raw Materials
Whole soybeans or defatted soybean meal can be used. Soybean flour is seldom used and would involve the preparation from the flour of tiny cubes before it could be used for fermentation (Yamaguchi, 1958). The whole beans or defatted soybean meal are best prepared by soaking in water for 10-12 hours at 29°C. The soaking can be done either by using running water or with still water which is changed every 2 or 3 hours. Unless the water is changed during soaking a rapid, undesirable fermentation, due to spore-forming rods (Church, 1923), occurs. These bacilli as spores are found on the beans as they come from the field. Beans soaked in unchanged water become warm, even hot, and sour at the bottom of a mass 5-6 inches deep in 2 or 3 hours at an initial temperature of 22°C. Such beans, even after autoclaving, are sour to the taste. It is the customary factory practice in Japan to soak the beans with changes of water at intervals of several hours. Nakaya (1934) found that beans should preferably be soaked until there is a 2.10 to 2.15 times increase in weight. Before the introduction of the autoclave the soaked beans were cooked in a large open iron pan until they were soft enough to be easily pressed flat between the thumb and finger. This desired softness can be obtained by autoclaving at 10 psi for 1 hour as against the much longer cooking in an open pan. Autoclaving under pressure has the additional advantage of sterilizing the material. A slight excess of water, just more than enough to cover them, is added to the beans before autoclaving. The beans
170
F. M. YONG
AND B. J. B. WOOD
are drained as soon as the autoclave pressure runs down. The pressure and time used for autoclaving the beans are very important. Kawano (1938) found that the digestibility of soybeans decreased as the pressure in the autoclave was increased. Evans et al. (1961) have reported that autoclaving soybean protein formed lysine-aspartic acid and lysine-glutamic acid linkages that are resistant to mild hydrolysis. Yokotsuka (1957) studied the influence of cooking soybeans on the quality of soy sauce and found that too long a cooking time causes a decrease in total, amino, ammonia, and tannin-precipitable nitrogen, acidity, volatile acids, and glycerol. Recently Tateno and Umeda (1955) and the Noda Soy Sauce Co., Ltd. (1955), claimed that total nitrogen in the soybean is best utilized when the beans are soaked in water for 10-12 hours at room temperature and then autoclaved at 10-13 psi for about 1 hour. Immediately after cooking the material was cooled to less than 4OOC. Rapid cooling of the beans on an industrial scale is done by spreading out the cooked beans to cool in about a 1-foot layer on a large traylike platform, the beans being turned over from time to time to hasten the cooling. Or they may be spread in wire trays and cooled with the draft of air from an electric fan. The rapid cooling of the hot sterile beans prevents the growth of microorganisms collected from utensils and handling after the beans are taken out of the autoclave. Details of the autoclaving process differ slightly depending on whether roasted cracked wheat or steamed wheat flour are to be employed. In the former case, the soaked beans are autoclaved immersed in water; in the latter case the soaked beans are drained before autoclaving, so giving a dryer product. Either whole wheat, wheat flour, or wheat bran, may be used; the choice will depend on their availability and price. When whole wheat is used it is first roasted and then coarsely crushed. The roasting should be continued until the wheat is crisp but not tough, and is browned to give a slight charred flavor. It is said that some manufacturers of soy sauce roast the wheat only slightly whereas others char the cereal. The browned wheat is believed by the Japanese to add flavor to the finished product through the formation of maltol due to the activity of the yeasts during the moromi stage of the fermentation (Church, 1923). It also adds a desired brown color. In the roasting of the wheat, practically all microorganisms present on it are killed. After roasting, the wheat is crushed, the crushing being carried to the extent of breaking the grains into large pieces. Furthermore, the crushing should be of such a character as to reduce some portions of the kernel to a fine powder or wheat dust. A supply of roasted wheat may be kept on hand and crushed as needed. Wheat bran and wheat flour on the other hand are generally steamed instead of being roasted. The properties of the
SOY SAUCE FERMENTATION
171
finished sauce will of course differ, depending on whether roasted grain or steamed flour has been employed, the latter giving a paler sauce of milder flavor. Asao and Yokotsuka (1957) reported on the changes produced in phenolic compounds in wheat and wheat bran by steaming. The characteristic products produced by cooking wheat were the guaiacyl series compounds, such as vanillin, vanillic acid, and ferulic acid. These phenolic compounds were claimed to be associated with the aroma and flavor of soy sauce. Wheat bran which had been subjected to ether extraction was cooked under 13 psi for 10, 20, 30, and 60 minutes, and the free and conjugated phenolic compounds in the cooked bran were determined colorimetrically. The free phenols increased with heating, but the conjugated phenols were at a minimum when cooked for 20 minutes. Paper chromatography of the ether extract of cooked wheat showed the presence of vanillin, ferulic acid, vanillic acid, and 4-ethylguaiacol, Among these, vanillin was the most abundant. These workers concluded that the phenolic compounds came from the degradation of lignin and glycosides by heating. The sterile softened beans and the crushed wheat are mixed in large trays or on mixing tables under nonsterile conditions. The beans are cooled below 28°C before being mixed with the wheat. The beans and the wheat need to be thoroughly mixed in such a way that the beans are held apart. The angular pieces of wheat when evenly and thoroughly distributed among the beans serve as a mechanical means of separating the wet smooth beans which would naturally pack much closer. The interstices are filled with the finer wheat particles to a certain extent, but not enough to check aeration. It is well to have the wheat, rather than the beans, on the surface. Furthermore, these two ingredients need to be thoroughly mixed so that the wheat dust may form a coat over each bean. The surfaces of the beans treated in this manner have a lower water content than when the precaution of thorough mixing is not taken. The dry wheat dust takes up the moisture readily. The lower water content thus induced on the exterior of the beans is better adapted to mold growth than to bacterial growth. VII.
Koji
Etymologically the word “koji” is an abbreviation of “kabi-tachi,” meaning something like “bloom of m o l d (Tamiya, 1958). Koji is just an enzyme preparation produced by growing a mold, such as AspergiZlus orynae, on steamed rice or other cereals or sometimes on steamed pulses. Historical documents show that Japan learned the use of koji from China more than 1700 years ago (Ono, 1941). The use of koji is analogous
172
F. M. YONG AND B.
J . B. WOOD
to the use of malt in the Western nations. Its use as a starter is not exclusive to the shoyu industry, it is used in the brewing of the Japanese rice-wine (sake) and fermentation of “miso” ( a sort of porridge or mush made by mixing mashed steamed soybeans, salt, and koji, and by allowing the mixture to ferment). Seed koji is produced by culturing A. soyae or A. oryzae on either steamed, polished rice (usual practice in Japan) or a mixture of wheat bran and soybean flour (China). It is usually incubated for 72 hours in small boxes or trays in a warm room. To avoid confusion the authors would like to use the term seed koji for the starter used to inoculate the very much larger quantity of soybean-wheat mixture to be fermented and koji for the large quantities of inoculated soybean-wheat mixture. The term koji is currently used to designate the starter as well as the whole mixture of soybean-wheat after being inoculated without any distinction and will cause confusion to someone learning about soy sauce fermentation for the first time. The seed koji after being incubated for about 72 hours is added to the soybean-wheat already prepared and thoroughly mixed. Usually 1-2% w/ w of the seed koji is used as the starter. Some morphological comparison of the Aspergillus strains used by the soy sauce industry has been reported by Kibi (1926), Sakaguchi and Yamada (1945a,b), and others. Besides A. oryzae and A. soyae, A. ochraceus, A. mellius, and A. niger have been tested in laboratories but have shown no practical value (Yokotsuka, 1960). A good culture mold must have high proteolytic activity and be easy to culture. Yokotsuka (1960) has also stated that a good culture mold must give the characteristic aroma and flavor to the soy sauce. This requirement would therefore imply that the mold is responsible for imparting the flavor and aroma to soy sauce. The authors have not been able to find any conclusive experiments to support this contention. There can, however, be no doubt as to the practical importance of the extracellular hydrolases produced by the mold. The mold Aspergillus oryzae was first isolated from koji in 1878 by a teacher of natural history from Germany, Dr. Ahlburg (Ahlburg and Matsubara, 1878). A strong diastatic activity was first demonstrated in 1881 by a teacher of applied chemistry from England, Dr. Atkinson (Atkinson, 1881). Takamine succeeded in improving the diastatic activity of the mold and prepared from it a drug having a strong digestive activity-Takadiastase ( Takamine, 1894). One of the major components of koji mold is a starch-liquefying amylase, Taka-amylase A. It was crystallized by Akabori (Akabori et al., 1951, 1953a,b, 1954) and was also purified by Sawasaki ( 19eOb). Okazaki ( 1954, 1955) and Sawasaki (1960a) claimed to have purified a saccharogenic amylase, Taka-amylase B, although its crystallization was not achieved.
SOY SAUCE FERMENTATION
173
Many investigators ( Kibi, 1951; Matsushima, 1954; Kageyama and Sugita, 1955; Sakamoto and Shuzui, 1957; Miura, 1958; Matsushima, 1959; Suzuki et al., 1959; Morimoto and Terui, 1960; Matsushima and Shimada, 1962; Yasui, 1964) have studied the proteases produced by Aspergillus oryzae and tried to identify the components of the mixture. Nowadays, it seems to be clear that three kinds of proteases-acid, neutral, and alkaline-exist in crude enzyme preparations of A. oryzae. Some workers (Morimoto and Terui, 1960; Matsushima and Shimada, 1962; Nunokawa, 1963; Morihara et al., 1!368) have reported on the fact that there are two kinds each of acid and neutral proteases. Matsushima (1959) indicated the presence of a “semi-acid” protease, but it has not been isolated in any form as yet, so its existence appears to be still in doubt. Recently, Misaki et at. (1970) reported the isolation of a “semi-alkaline” protease. Thus the constitution of A. oryxae proteases is so complicated and diverse that the separation and purification of the proteases have not been sufficiently performed except in the case of alkaline protease (Crewther and Lennox, 1950; Akabori et al., 1953a,b; Miura, 1955; Hayashi et al., 1967b). Tsujita (1967) has succeeded in crystallizing acid protease, but in the case of neutral protease there is‘little information in respect to its purification. The mixture of proteases from Aspergillus sojae have also been extensively studied (Sakaguchi and Yamada, 1944; Yamamoto, 1957b,c; Hayashi et al., 1967a,c; Sekine et al., 1970). The constitution of A. sojae proteases is very similar to that of A. oryxae. Iguchi (1949, 1950a,b, 1951, 1952a,b,c, 1953, 1955a,b, 1956) and Iguchi and Yamamoto (1955a,b) obtained by X-ray irradiation a mutant strain of A . sojae K.S. which had two or three times the proteolytic activity of the other strains tested. More recent workers had tried to increase even further the protease productivity of this mutant (Sekine et al., 1969; Nasuno et al., 1971). Although a lot of work had been done on the isolation and purification of the proteases and amylases of A. oryzae and A. sojae, and the mutation of parent strains to produce higher levels of proteases, there is very little work on the levels of proteases and amylases in the koji and SOY mash throughout its whole period of fermentation ( Sakaguchi, 1958a).
VIII.
Culturing the Koji
The effects of physical environmental factors such as temperature and moisture on the culturing of mold has been investigated by Kinoshita and Matsumoto (1935), Murakami (1951a,b), Harada (1951), and Yamamoto ( 1957a,b,c). In industrial practice the whole mass of koji is distributed in small flat trays, made from bamboo strips, stacked above each other but sep-
174
F. M. YONG AND B. J. B. WOOD
arated by a gap of about 4 inches. Each tray is filled up to a depth of about 2-3 inches. The trays are placed in tiers along the walls of the room. In the incubation of koji, Church (1923) has stressed that “favorable aeration . . . is extremely important in shoyu-koji fermentation because moisture and the lack of oxygen induce the development of Mucors and bacteria, and are said to cause the diastatic enzyme to develop at the expense of the proteolytic enzyme.” The control of temperature and moisture is equally important ( Yokotsuka, 1960; Kibi, 1938).The temperature of the culturing room is usually kept at 2535°C. When the mold grows, the temperature of the material rises. When the temperature becomes too high, that is, above 40”C, the culture must be cooled. Cooling is usually done twice, at about 20 and 40 hours after inoculation. A rather high moisture content is required at the beginning, when the mold is growing rapidly; lower moisture content is desirable at the later stages, when spores are being formed. Kinoshita and Matsumoto (1935) reported that the cooling must be done before the temperature rises above 40°C since mold growth is inhibited above this temperature. Harada (1951) studied koji cultured under various temperature conditions and found that the highest proteolytic activity was obtained when the culture was cooled rapidly during the second cooling operation. Cooling is done by stirring the koji, the bottom being brought to the top and vice versa. Thorough stirring is also necessary, especially on the first day of culture when the beans are bound together into a mass by the mycelia or white threads of the shoyu mold. Usually 3 days of mold growth will be sufficient; with a shorter incubation period the enzymes produced would be inadequate. On the other hand, if the growth period is extended, excessive sporulation occurs and undesirable flavor may be imparted to the sauce (Lockwood and Smith, 1950-1951; Baens-Arcega, 1970). Mature koji has a clear yellow to yellowish-green color on the surface and throughout the whole mass. The koji may at times become infected with Rhizopus nigricans if the atmosphere of the koji chamber is moist to the point of condensation as drops. A little Mucor or Rhizopus is disregarded in the material, unless a bad flavor or odor is also present. It is poor practice, however, to allow the Rhizopus to enter. If allowed to gain a foothold, its fruiting to any extent may be prevented by breaking up the koji into chunks and turning these chunks bottomside up. Instead of exposing a large surface area as in the case of bacterial infection, care should be taken to have only surfaces where Rhizopus has secured no firm footing exposed to the air. “Trays of koji infected with Rhizopus should be stacked in a cool, dry place until the material is mature or needed for the shoyumoromi” ( Church, 1923).
SOY SAUCE FERMENTATION
175
The koji stage of mold fermentation brings about the enzymatic breakdown of proteins to peptides and amino acids by proteolytic enzymes, especially by neutral and alkaline proteases. Starches are hydrolyzed to dissaccharides and monosaccharides by a-amylase secreted by the mold. There have been no reports on the breakdown of soybean oil at this stage of fermentation. Maximal proteolytic activity and hence maximum degradation of insoluble protein into soluble protein, polypeptides, peptides, and amino acids, is obtained from a koji of equal parts of wheat and soybeans (Yamamoto, 1957b). We have found that there is a rapid increase in reducing sugar during the very early stages of mold growth. This is due to the activity of extracellular invertase hydrolyzing the considerable quantities of sucrose present in the beans (Yong, 1971). There is evidence for the production of @-amylaseand cellulase by the mold (S. K. Goel and B. J. B. Wood, unpublished data). Our studies have shown that Aspergillus isolated from soy koji also produces variable but appreciable amounts of lipase ( Yong, 1971). Apparently it is not yet known whether the deamination of amino acids to yield the free ammonia, which is always present in small amounts in mature koji and may attain considerable proportions in overripe koji, is the result of intra- or extracellular enzyme activity, and this is among our current subjects of study. Continuous or mechanically controlled mold culturing is greatly desired but as yet has not been successful (Yokotsuka, 19f30). IX.
Mash (Moromi)
The mature shoyu-koji is mixed with about an equal amount of brine to form the mash, or “moromi.” The mash is kept in a large container, which can be either a wooden tub or a concrete tank such as is used in modern Japanese factories (Sakurai, 1965), for about one year at ambient temperatures, or for 3-4 months if warmed. “Ordinarily the mash is stirred by compressed air” (Yokotsuka, 1960). Major chemical changes in this process are degradations of protein and carbohydrate caused by enzymes derived from koji. According to Yokotsuka, a lactic acid fermentation occurs in the first stage, an alcoholic fermentation by yeasts in the second stage, and an aging or completion of fermentation in the last stage. The color of the mash gradually becomes more intense.
A. PREPARATION (MASHING) Formerly, equal volumes of salt water (17-19%w/v of sodium chloride) and koji were used, but recently the volume of salt water has been increased to 110-120% of the raw material. Mixing koji with more water
176
F. M. YONG AND B. J. B. WOOD
causes a better utilization of the total nitrogen of the raw material; however, this may have some undesirable effect on the composition of the soy sauce. It is regarded as dangerous to use a solution of less than 16% because of the danger of putrefaction. Usually the solution of salt added to the koji is controlled by the specific gravity of the saline solution, the range being between 20° and 22OBaumB (Be). Ohara and Moriguchi (1955a,b) mixed salt solutions of 19O, 20°, and 23OBe with koji, and found that the lower the specific gravity the better the utilization of total nitrogen and amino nitrogen, also the better the fermentation; in addition, less residual sugar and higher acidity were obtained in the soy sauce.
B. CONTROL OF
THE
MASH
In natural fermentation, where the temperature of the mash is not controlled, a fermentation period of over 12 months is necessary, since the mash must be fermented by yeasts for one summer to give good aroma and flavor. To shorten the fermentation period, the temperature must be controlled artificially; if warmed to about 3540°C it will take only 3-4 months. Recently, S. I. Tan (personal communication, 1970) has successfully made soy sauce of good quality at the Singapore Institute of Standards and Industrial Research by incubating the mash at 4OoC for 1 month. Some lactobacilli will grow at temperatures up to 47°C in more norma1 media, but it seems surprising that they will tolerate the combination of high temperature and high salinity. It is even more remarkable that yeasts can survive such conditions, especially when one considers 'the low pH (4.5-5) which they require in order to grow at all in these saline conditions. In our own studies, we routinely employed a temperature of 4OoC, and once the mash pH had dropped from its initial level of 6.5 to around 4.8 the yeast grew rapidly, increasing in number of viable cells 30-60 times in the space of 3-4 days. It would be most interesting to examine the effects of yet higher temperatures. In the traditional method or in some cottage industries, the mash is stirred by wooden paddles once each day in the initial stage of the fermentation, then once a week, and finally less frequently toward the end of the fermentation. In modern factories in Japan, the mash is aerated and agitated by compressed air at 6-10 psi for 5-10 minutes. Yokotsuka (1960) did not specify the frequency of the stirring and aeration by compressed air nor any other related data. Yamada and Furusaka (1954) have reported that too much stirring, especially in summer, hinders the fermentation of mash. Stirring the mash is very important and a difficult procedure, and no final conclusion has been reached regarding the frequency of stirring.
SOY SAUCE FERMENTATION
177
C. ACING The initial pH of the new mash is usually 6.0-7.0. It decreases rapidly, especially in summer, to about 4.5, the point at which alcoholic fermentation begins if the temperature is suitable. Usually, in 3 or 4 months at room temperatures, the total nitrogen dissolved in the mash increases to a constant value. Amino acid content increases at almost the same rate as the total nitrogen. Udo (1931b) analyzed the chemical composition of mashes aged for varying periods at ambient temperature. He found under the conditions used by him that glutamic acid was at a maximum after 15 months, and then it started to decrease. Udo concluded that the optimum aging period could be determined by analyzing for free glutamic acid content. He was using whole beans and, more recently, Umeda et aE. (1953), working with defatted soybeans, reported the same finding. The maximum glutamic nitrogen content was obtained in 10 or 11 months. In another experiment, they found that the glutamic nitrogen content of a mash aged 58 months was equal to that of a mash aged for only 1 month. Yokotsuka (1960) has noted that higher temperatures cause a rapid decrease of mash acidity and rapid inactivation of enzymes.
D. MICROBIOLOGY OF MASH Yokotsuka (1960) stated that, besides Aspergillus oryzae or A. soyae, Monilia, Penicillium, and Rhizopus are sometimes found in the mash, but that these molds are believed to have no relation to proper aging. It is unlikely that these molds would be growing in such a high salt concentration, and it has been found that sodium chloride inhibits the growth of Aspergillus oryzae (Ichikawa, 1954a). It is likely that the molds exist as spores in the mash, not in the vegetative form, which would most likely have autolyzed in 18%wlv salt solution. There are two conflicting reports on the effect of koji on the flavor of soy sauce. Sugita ( 1956) studied the relationship between the organoleptic evaluation of cultured koji and the quality of soy sauce made from it, and recognized a fairly good relation between them. However, Sakaguchi (1959a) fermented soy sauce in a microbiologically closed system in which koji mold (that is, either A. oryzae or A. sojae) was the only microorganism participating and obtained a product which has no characteristic flavor and taste. It might be possible that the cultured koji obtained by Sugita was contaminated with 'flavor-producing' microorganisms. The amino acids, especially glutamic produced by the mold fermentation will, however, contribute to the taste, and hence quality, of
178
F. M. YONG AND B. J. B. WOOD
the product obtained. Our own studies support the views advanced by Sakaguchi. Soy sauce made with mold alone, or mold plus Lactobacillus, lacked the full, rich aroma of a good sauce. The participation of the yeast was essential, and the best flavor was invariably obtained from fermentations where the moromi had been either soured with lactic acid, then fermented with yeast, or fermented with a mixture of yeast and lactobacilli. A mold-koji was, of course, employed in all cases. It was realized as early as 1915 that “complicated chemical changes are going on during the ripening process of shoyu-moromi” and that “. . . these chemical changes are entirely due not only to the various enzymic actions of Aspergillus oryzae in koji, but also to the vegetative and autolytic actions of many kinds and numbers of microbes in moromi” (Yukawa, 1915). The “many kinds and numbers of microbes” refer to the bacteria and yeasts present in the soy mash. Much of the published work on the microbiology of soy mash has been concerned with the isolation and identification of bacteria and yeasts from soy mashes obtained from the factories ( Mogi, 1949).
1. Yeasts The first report on soy yeasts appeared in 1906, when Saito isolated five strains of salt-tolerant yeasts from soy mashes in the Choshi district in Japan and classified them as Saccharomyces soja, Zygosaccharomyces japonicus, Pichia farinosa, Mycoderma sp. and Torula sp. Subsequently, many taxonomic studies were conducted on soy yeasts (Mitsuda, 1910; Nishimura, 1910a-g, 1911, 1912a,b; Kita, 1911; Ishimaru, 1935; Sakasai and Yoshida, 1966; Onishi and Suzuki, 1970). These workers also isolated Z. major, 2. sulsus, Torulopsis, and Monilia species from soy mash. However, none of these observations explained which yeast( s ) played an important role in the ripening of soy mashes and how the yeast( s ) improves the quality of the product. Takahashi and Yukawa (1911-1915) showed that 2. maior and Z. soyae were useful in ripening soy mashes, giving the characteristic taste and flavor through their fermentation. Z. soyae and Z. major are believed to be the most indispensable yeasts in normal fermentation (Asao et al., 1969). Takahashi and Yukawa (1911, 1914), Ishimaru (1935), and Mogi et al. (1951, 1952) showed that three kinds of yeasts are harmful to soy sauce on keeping: film-forming yeasts, such as Zygosaccharomyces s u h s , 2. japonicus, and Pichia; ring-forming Torulopsis which grows on the surface of the soy sauce around the edge of the container; and bottom yeasts belonging to Zygosaccharomyces. 2. mafor, 2. sofa, 2. s u h s , and 2. japonicus were included in one species, Saccharomyces rouxii, by the Lodder and Kreger-van Rij’s system (1952). These yeasts are also known as soy yeasts. They are found
SOY SAUCE FERMENTATION
179
in soy mashes because of their tolerance to high concentrations of sodium chloride and osmotic pressure. Reliable information on the biochemical activity of the yeasts in the soy mash has not yet been obtained. In recent years, extensive studies has been carried out by Onishi (1954a,b, 1957a,b,c, 1958a,b,c, 1960a-d, 196l), Onishi, et al. ( 1961), and Onishi and Saito ( 1961, 1962). The growth of soy yeasts was found to involve a process of physiological adaptation. The nutritional requirements for nitrogen compounds and vitamins in two contrasting media (sodium chloride-free and a high concentration of sodium chloride) were also studied. Probably the most striking finding on the applied aspects of soy yeasts was that the pH range for the growth of soy yeasts in a sodium chloride-free medium is very wide (pH 3.0-7.0), while that in an 18%w/v sodium chloride medium is very narrow (pH 4.05.0). Saccharomyces rouxii was found to ferment glucose and maltose but not galactose, saccharose, and lactose. In a high sodium chloride medium (under aerobic conditions) S . rmxii produced high yields of glycerol whereas it produced only small amounts of glycerol in an ordinary medium. As much as 4030% of the glucose fermented was converted into glycerol in the saline medium under aerobic conditions. Yamada and Furusaka (1954) claimed that more than 10%w/v of sodium chloride is necessary for the growth of 2. major in soy mash. Obata (1955), however, found that soy sauce aroma is produced when the sodium chloride content of the medium is higher than 5%w/v. Takeda (1954) has reported that the addition to soy mash of a yeast inoculum grown up in shake-flask culture, would produce a relatively good quality soy sauce. However, Yauchi et al. (1955) observed that excessive fermentation induced by cultured yeasts could result in inferior soy sauce. Recently, Yokotsuka et al. (1967a) isolated 300 strains of yeasts from several kinds of shoyu mash which were considered to have good flavor. Thirteen strains of non-film-forming yeasts were divided into three groups according to the speed of alcohol fermentation in soy-koji extract supplemented with 5% wlv glucose and 18%wlv sodium chloride. The three groups are: ( I ) good and rapid fermentative; (11) bad fermentative; and (111) slow but good fermentative. The strains of groups I and I11 were evaluated as superior strains with respect to good flavor production in actual fermentation of soy mash. The strains of group I11 were observed to produce a characteristic flavor like that of aged soy mash. The production of this characteristic flavor was showed by Yokotsuka et al. (1967a) to be due to the formation of 4-ethylguaiacol, 4-ethylphenol, and 2-phenylethanol. Torulopsis species were the only type of yeast that produced these alkyl phenols during the fermentation
180
F. M. YONG AND B. J. B. WOOD
of soy sauce (Asao et al., 1967; Yokotsuka et al., 1967b). The alkyl phenols are not produced in soy mash fermented by S. rouxii alone ( Asao et al., 1969). Furfuryl alcohol, one of the components producing flavor in soy sauce, was found to be produced by yeasts, for example, S. rouxii and Pichia furinosa, by reduction of furfural present in koji extract (Morimoto and Matutani, 1969). Mori and Onishi ( 1967) have reported successful diploid breeding of salt-tolerant S. rouxii, a heterothallic haploid, for soy sauce and miso fermentation. The factors affecting the stability of the diploids obtained were also investigated ( Mori and Onishi, 1969). It is clear that, as with so many other aspects of the product of soy sauce, the reports in the literature on the functions of yeasts are confusing and sometimes completely contradictory. This may reflect the influence of local or regional differences in the practice of soy sauce manufacture, but there clearly exists a great need for detailed studies to resolve these problems. There are also conflicting reports on the ecology of mold, yeasts, and bacteria in the fermentation of soy sauce. Lockwood and Smith (1950-1951) are of the opinion that a good yeast growth in the koji before the mold growth becomes apparent will result in a product of superior quality. In this case, yeast would be added to the steamed soybeans about a day before mixing them with parched wheat, and the yeast would start to grow before the mold gets under way. It has been the normal accepted procedure to let the mold grow and degrade the raw materials before mixing the koji with brine and inoculating with yeasts. Since the mold fermentation is not under aseptic conditions in practice, one would expect yeasts and bacteria to be present in the koji. The majority of workers consider that these microorganisms have not been shown to be of importance in the making of a good quality koji, thus disagreeing with the contentions of Lockwood and Smith. Yokotsuka (1960) and Onishi (1963) among many other Japanese workers are of the opinion that a lactic acid fermentation is followed by a yeast fermentation. This hypothesis is supported by the fact that after lactic acid bacteria have propagated in the young moromi of pH 6.0 to 7.0, lowering the pH to below 5.0, the yeasts will grow vigorously. It has been found by Onishi (1957b, 195810) that the yeast isolated from soy sauce mash would not grow at a pH above 4.5 in artificial media containing a high concentration of salt. It is possible that the growth of yeast in soy mash which has an approximate salt concentration of 18%is controlled by the pH value of the mash. Hesseltine and Wang (1968) noted that it was not known for sure where the lactic acid bacteria and the yeast enter the fermentation.
SOY SAUCE FERMENTATION
181
On balance, therefore, the most likely sequence of dominant microorganisms would seem to be fungus, then lactic acid bacteria, and finally yeasts. Further work is needed ( a ) to establish this sequence, ( b ) to investigate whether it is better to have mixtures rather than using pure strains of each of the major classes of microorganisms, and ( c ) to determine whether the numerically less important organisms have a part to play, or are merely contaminants exerting no significant effect on the fermentation. In our own work, convenience constrained us to use pure cultures of a few selected species. Even so, employment of A. oryzae under aseptic conditions for the koji, and of one strain each of S . rouxii and of Lactobacillus delbrueckii under aseptic conditions for the moromi stage, gave a product judged entirely acceptable by ourselves and by colleagues from Hong Kong, Malaysia, Korea, and Thailand.
2. Bacteria Saito (1907) first isolated homofermentative (that is, giving lactic acid as sole product of the fermentation of glucose), salt-tolerant, tetradforming cocci from soy mash. He named the organism Sarcinu hamuguchiae. A similar coccus has also been isolated by Sugimori and Joze (1970) and they called it Tetracoccus soyae. Ishimaru (1930) reported the presence of Pediococcus acidihctici var. soya in soy mash and described its effect on soy-sauce brewing. Sakaguchi (1954, 1958b) studied the same kind of bacterium isolated from soy mash and proposed the name P. soyae for it. Independently, and at almost the same time, Iizuka and Yamazato (1959a,b) proposed the same species name for an identical organism which they isolated from a so)- mash. The growth factor requirements of P . soyae have been extensively examined by Sakaguchi. He has isolated from soy-koji extract, a peptide which increased the organism’s rate of cell division and the cell yield (Sakaguchi, 1959c,d). Addition of glycylbetaine or of carnitine to defined media also increased the growth rate and cell yield (Sakaguchi, 1960a). He has also examined the requirements for amino acids and vitamins exhibited by the organism ( Sakaguchi, 1960b); the effect on its growth of highly concentrated solution of inorganic salts; and the organism’s role in soy sauce fermentation (Sakaguchi, 1959a,b), which he found to be entirely concerned with the production of acid. Sakaguchi and Mori (1969) have suggested that P. soyae, P. halophilus [which was isolated from salted anchovy pickles by Mees (1934)], and P. homari [which was isolated from meat curing brines by Deibel and Niven ( 1960)] are sufficiently similar in their morphological, physiological, and nutritional characteristics to place them into a single species. Matsumoto (1925) isolated various Bacillus strains from soy mash,
182
F. M. YONG AND B. J . B. WOOD
and Ishimaru (1933a,b) also reported on many species of salt-tolerant bacilli and a few strains of Micrococcus in soy mash. Sakaguchi (1958a) found very few vegetative cells of Bacillus in soy mash and he therefore assumed that bacilli mainly grow in the koji and survive in the soy mash as spores. Lockwood (1947) and Lockwood and Smith (1950-1951) reported on the use in soy sauce fermentation of Lactobacillus delbrueckii which was isolated from soy mash. The Japanese workers report on the presence of Lactobacillus in the moromi, but do not seem to attach much significance to it, whereas Lockwood (1947) claims to have produced good soy sauce using L. delbrueckii as the only bacterium in the fermentation. The growth of lactic acid bacteria in soy mash would result in the formation of organic acids and make the mash acidic, a condition which is necessary for “sound fermentation, to remove undesirable flavors, and add indispensible good flavors to the soy-sauce” ( Sakaguchi, 195913). However, the lowering of the pH will most probably also result in the depression of neutral and alkaline proteinase activities, followed by their inactivation; thus the percentage of protein solubilization would decrease. X.
Pressing
In ancient China and in more primitive factories, the liquid is removed by “drawing” (Groff, 1919) or siphoning off the liquid on top. If the method of drawing is used, fresh salt solution is added to the “teng shi” or the beans remaining in the jar from the first drawing, and lactic acid and yeast fermentations are allowed to occur for another 1-2 months before the second drawing. This method of drawing may be done 4 times, and each “drawing” represents a different grade; the first drawing being the best and the last drawing the cheapest and poorest grade. In modern factories the liquid part of the mash is separated from the residue with a hydraulic press. The residue, which has a moisture content of about 40%,is sometimes used as animal feed. The oily layer of the filterate from soy mash is separated from the aqueous layer by decantation.
XI.
Pasteurization
Raw soy sauce is pasteurized, and this process not only kills the vegetative form of microorganisms but also denatures enzymes and brings about the sedimentation of incompletely degraded protein compounds by coagulation. Matsumoto (1923)) Matsumoto and Murakami (1941),
SOY SAUCE FERMENTATION
183
Ohara and Orishi ( 1951), Ueno and Omori ( 1951), and Kosaka (1956) have reported on the compounds coagulated by pasteurization. Usually the raw soy sauce is pasteurized at 65OC. Alum (1 ounce to every 80 gallons) or kaolin (1 ounce to 1 gallon) is then added for clarification. The precipitate is allowed to settle overnight, and the sauce is finally filtered ( Baens-Arcega, 1970). It is not clear what steps, if any, are taken to prevent the kaolin from introducing an inoculum of undesirable microorganisms. Chemical preservatives are very often added to pasteurized soy sauce to prevent growth of microorganisms, especially film-forming yeasts, such as Zygosaccharornyces sulsus, 2. japonicus, and Pichia; ring-forming Torulopsis which grows on the surface of soy sauce around the edge of the container; and bottom yeasts belonging to the Zygosaccharornyces. The latter two species may be present in dilute soy sauces. Reports on these organisms had been made by Takahashi and Yukawa (1911, 1914), Ishimaru (1935), and Mogi et al. (1951, 1952). The chemical preservative most widely used in Japan is butyl-p-hydroxybenzoate at a concentration of 0.005% An alternative choice is usually sodium benzoate which is used in the region of 0.02%concentration. XII.
“Chemical” Soy Sauce
In order to lower production costs, several attempts were made to shorten the long fermentation period required to produce a relatively good soy sauce and at the same time to obtain a product of more constant quality. At first various lots of sauces which had been fermented for different periods of time were blended and sold. It needs no one to point out that this was not an improvement in soy sauce technology. The Japanese, unlike their Chinese rivals in the same industry, did not cling tenaciously to their traditional technology, which was established when the role of the beneficial “contaminants” of mold, lactic acid bacteria, and yeast was still yet unknown to them, but were more courageous and aggressive in trying to modify the fermentation process then in existence. Whereas the Chinese process was carried out in 50-gallon earthen jars out in the open, the Japanese as early as 1948 were constructing special cement tanks housed in large buildings, so providing cover and a degree of control over the environmental conditions (Smith, 1949). Early in this century experiments began to be made on acid rather than enzymatic hydrolysis of the mixture of soybeans and wheat. Chow (1935) has given a review of the chemistry and manufacture of chemical soy sauce in mainland China. Then from 1950 onward (Yokotsuka, 1960), there was a growing tendency to make soy sauce by acid hydrolysis in order to increase the
F. M. YONC
184
AND B. J. B. WOOD
Soybeans
Wheat
Hydrolyzed with 20% hydrochloric acid
1 Filtered
1 Fil t,rate
1 Neutralized wit,h sodium hydroxide
1 Pasteurized
1 Soy sauce
FIG.2. Flow sheet for production of chemical soy sauce.
yield and reduce production costs. This chemical method is illustrated in Fig. 2. According to Minor (1945), a mixture of soybean (meal) and wheat is hydrolyzed by refluxing with constant boiling hydrochloric acid (20% solution) until a maximum concentration of amino acid nitrogen has been obtained. Sufficient hydrolysis would have taken place after 12-16 hours. After hydrolysis, the preparation is neutralized gradually with a 50%sodium hydroxide to pH 4-5. The sauce is then ready to be placed in hardwood storage tanks for aging prior to bottling. When properly manufactured, the sauces will have a final salt concentration of 18% w/v sodium chloride. The greatest problem here is that the carbohydrate in the raw materials is not oiily hydrolyzed faster than the protein, but also a small, but important, percentage of it is converted into undesirable compounds such as “dark color of humus,” levulinic acid, and formic acid. Moreover, the complete decomposition of tryptophan, the formation of an excess of lower boiling point sulfur compounds, giving bad odors, and the lack of fermented flavor, are further major disadvantages of this method. There is no market for a purely acid-hydrolyzed product in Asia, although it finds a market in Europe.
XIII.
Semichemical Soy Sauce, or Shinshiki Shoyu
The next move made by the Japanese was to take the advantage of both methods-fermentative and chemical. The idea of hydrolyzing the raw materials first and then subjecting
SOY SAUCE FERMENTATION
185
the whole to fermentation was known in the 1930’s; however, at that time the microorganism added to ferment the hydrolyzed substrate was a mold of the Aspergillus flauus-oryxae group (Kodama and Kimura, 1931; Snelling, 1937; Itikawa and Huzi, 1939). This method was not very successful in producing a good quality soy sauce, and therefore was not seriously taken up. Noda Shoyu Company Limited managed to develop a process involving chemical hydrolysis followed by fermentation with lactic acid bacteria and yeast just after World War I1 (Yokotsuka, 1964). In this method, defatted soybeans are first partially degraded by using dilute hydrochloric acid, (about 7-8$), then neutralized with sodium hydroxide. Next rather large amounts of wheat bran or copra-meal koji are added to the partially hydrolyzed defatted soybeans. Only yeast is used in the fermentation. Using this “intermediate” method the fermentation period was shortened to 2 or 3 months, but a small amount of the characteristic furfurol and sulfide-like odors were unavoidable in the final products. This method is widely applied in Japan, and the amount of shoyu made by this method is estimated to be 20%of the total production ( Yokotsuka, 1964). The soy sauce made by first hydrolyzing the raw materials and then fermenting them with bacteria and yeasts is called “shinshiki” or “semichemical” soy sauce, Investigations have been conducted into the best conditions for hydrolyzing the raw materials, either together or separately, and the components present in this type of soy sauce (Ueno and Kuramochi, 1960, 1961; Ueno and Nobuhara, 1960a-d; Ishigaki and Nagase, 1964a,b) . XIV.
Future Developments in the Soy S a u c e Industry
When it was realized that the fermentation process was definitely essential in producing a superior product with superior quality, efforts were made to shorten the traditional or conventional fermentation process. Soybeans which were once boiled in an open pan are now autoclaved, and a lot of work went into getting the best conditions for autoclaving to get the maximum digestibility of soybeans (Kawano, 1938; Tateno and Umeda, 1955; Noda Soy Sauce Co. Ltd., 1955). In recent years defatted soybeans have replaced whole soybeans in about 75% of the shoyu produced in Japan. This is because defatted soybeans are much cheaper, and, more importantly, the fermentation period is much shorter. The use of defatted soybeans reduced the fermentation period from about 15 months to approximately 10 months. A lot of work has been done by Japanese workers, such as Kinoshita and Matsumoto ( 1935), Harada ( 1951), and Murakami (1951a,b), on
186
F. M. YONG AND B. J. B. WOOD
elevated and controlled temperature conditions of incubation to reduce the fermentation period. In the early days of soy sauce fermentation, the soy mash was left out in the sun during the summer to mature and there was no device to maintain it around 37°C for fermentation to take place faster and uninterrupted by variations of temperature resulting from seasonal variations of climate. Work on the effect of the moisture content of koji throughout the 72 hours of incubation had been conducted by Yamamoto (1957b), and this factor has been found to be of some importance in producing the correct soy sauce flavor and of great importance in preventing excessive bacterial contamination. He found that a higher moisture content (50-75%) favored mycelial growth, and a lower moisture favored spore formation. At the very beginning of soy sauce fermentation, chance inoculation from the surroundings was relied upon to provide the necessary useful microflora. This would in a way account for the 3-year fermentation period required in the early days of fermentation. When it was realized and proved that soy sauce making is essentially a microbial process, attempts, which were successful, were made to isolate, identify and to use appropriate organisms as pure inocula (Lockwood, 1947; Lockwood and Smith, 1950-1951). The use of pure culture inocula was a really significant step in understanding the process and has the following two main advantages: ( a ) Contaminating microorganisms are not carried over from one fermentation to another. ( b ) Since most types of soy sauce require at least 6 months to mature, it is apparent that an old moromi would be deficient in the appropriate microorganisms, giving rise to problems if it were used to inoculate a new moromi. With pure cultures, young, active cells in the correct proportions can be used to inoculate each new batch of moromi. It had been found that sodium chloride at such high concentrations as 18%wlv, will not only decrease the rate of growth of lactic acid bacteria and yeasts, but also their physiological activities. The fermentation of soy sauce may actually have come into existence as a result of efforts to preserve cooked soybeans with a strong solution of sodium chloride. It has been found by Ichikawa (1954a,b) that fermentation of the koji in a medium without sodium chloride would result in a soy sauce with excessive amounts of ammonia due to uncontrolled growth of mold in the soy mash. If a process could be developed whereby one need not use Aspergillus orzyzae or A. soyae to break down the proteins and oligosaccharides in the raw materials, and also whereby the use of sodium chloride could be reduced or eliminated during moromi fermentation, the fermentation period might be drastically reduced. The use of acids to hydrolyze the raw materials is too drastic and results in the production of undesirable compounds such as levulinic
SOY SAUCE FERMENTATION
187
and formic acids which would give an inferior quality product (Yokotsuka, 1964). The use of commercially available enzymes to cause effective, selective, and rapid hydrolysis of the raw materials is certainly a very interesting method for replacing the use of molds in the koji stage. If a liquid multistage continuous soy sauce fermentation process can be developed, the production cost will be reduced by more than one-half because the mold fermentation stage takes up about 40%of the production cost and another 40%goes into the filtration and processing of the mash. The high cost of filtration results from the fact that it is necessary to press-filter the soy mash at very high pressure. Using a lower pressure means a longer filtration time for expression of the raw soy sauce and subjects the raw soy sauce to greater chances of oxidation of the oxidizable ingredients of the shoyu, which are the flavor-producing agents in shoyu. To develop a multistage continuous soy sauce fermentation process, one needs to have some basic and important data on soy sauce fermentation. There have been a lot of investigations into soy sauce fermentation, mainly by the Japanese, and therefore there is a voluminous amount of literature on soy sauce in Japanese journals. However, in spite of and because of this superabundance of literature, one gets confused when one first begins to read about soy sauce fermentation. The investigations were mainly carried out on koji or soy mashes obtained from the factories rather than those fermented under controlled conditions in the laboratories. Published research on soy sauce fermentations can be divided into approximately eight areas : 1. The isolation and identification of the beneficial microorganisms from koji and soy mash, and maintaining pure cultures of the isolates. 2. The mutation of molds to give highly proteolytic strains to degrade the proteins in the soybeans. 3. The use of raw materials other than wheat and soybeans. 4. Preservation. 5. The development of technology to improve yield and quality; the modifications have so far only resulted in alterations in the equipment already present in the factory, without any daring and far-sighted attempt to develop an entirely new process which will require radically new equipment and processes. 6. Possible production of aflatoxin by molds used in soy sauce manufacture. Hesseltine et al. (1966) and Hesseltine (1966) reported on tests carried out on extracts of soy sauce from Japan and Taiwan for aflatoxin at the levels of 3-5 parts per thousand million and have found it to be negative. They also found that none of 53 strains of A. oryzae
188
F. M. YONC A N D B. J. B. WOOD
used in food fermentations, such as miso and soy sauce, produced detectable quantities of aflatoxin. 7. The flavor of soy sauce. Work in this particular aspect of soy sauce fermentation has been carried out since 1887 by Tawara (1887), Yukawa ( 1916), Ikeda and Kawaguchi ( 1922), Ishida ( 1925a,b), Taira ( 1925, 1926), Shoji (1927), Udo (1931a,b), Akabori and Kaneko (1936), Kihara (1940), Nakajima et aZ. (1949), Yokotsuka (1949, 1951a-d, 1952a,b,c, 1953a-d, 1954, 1957, 1960), Obata and Yamanishi ( 1951a,b,c, 1953), Asao and Yokotsuka ( 1958, 1961a,b, 1963), Yokotsuka ( 1960), and others. 8. Color of soy sauce. The color of soy sauce is very important since it is said to be associated with flavor. It has been studied by Kurono and Katsume (1927) and Kurono et aZ. (1927). They found that the color of soy sauce is caused by the so-called “nonenzymatic browning action.” Very recently, Okuhara et aZ. (1969a,b, 1970), Okuhara and Saito ( 1970a,b), and others have carried out extensive investigations on the color of soy sauce. XV.
Conclusions
It is clear that soy sauce both in its production and its chemistry is a most complex material, offering much to fascinate the biologist, microbiologist, chemist, biochemist, and chemical engineer. Our own studies (Yong, 1971; Yong and Wood, 1973) are an attempt at a systematic attack on some of these problems. Yet soy sauce is only one of a range of similar products, mostly unfamiliar to the Occidental palate, but affording the investigator challenges and problems very similar to those described for soy sauce (Hesseltine, 1965; Wood and Yong, 1973). We are convinced that, so far as soy sauce is concerned, the application of modern technology will yield a process for the production of soy sauce by continuous fermentation techniques, employing enzyme extracts to replace the present koji stage, and continuous yeast fermentation under aseptic conditions, to give a low-cost product of very high quality indeed. REFERENCES Ahlburg, H., and Matsubara, S. (1878). Tokyo lji Shinshi 24, 12 (as cited in Tamiya, 1957 ) . Akabori, S., and Kaneko, T. (1936). 1. Agr. Chem. SOC. lap. 57, 832-836. Akabori, S., Hagihara, B., and Ikenaka, T. (1951). Proc. lap. A d . 27, 350-351. Akabori, S., Hagihara, B., and Ikenaka, T. (1953a). Proc. Jup. Acud. 8, 48 (as cited in Tamiya, 1957). Akabori, S., Hagihara, B., Ikenaka, T., and Sakoda, N. (1953b). Symp. Enzyme Chem. 8, 49-54.
SOY SAUCE FERMENTATION
189
Akabori, S., Ikenaka, T., and Hagihara, B. (1954). J. Biochem. Tokyo 41,577-582. Akira, K. ( 1967). Chomi Kagaku 14,42-53. Allen, P. W. (1926). In “Industrial Fermentations,” pp. 123-127. Chem. Catalog Co. (Tudor), New York. Asao, Y., and Yokosuka, T. ( 1957). Bull. Agr. Chem. SOC. Jap. 21, 622-628. Asao, Y., and Yokotsuka, T. (1958). J. Agr. Chem. SOC. Jap. 32, 622-628. Asao, Y., and Yokotsuka, T. (1961a). J. Agr. Chem. SOC. lap. 35, 1211-1218. Asao, Y., and Yokotsuka, T. (1961b). J . Agr. Chem. SOC. Jap. 35,831-837. Asao, Y.,and Yokotsuka, T. ( 1963). J. Agr. Chem. SOC.Jap. 37,563-568. Asao, Y., Sakasai, T., and Yokotsuka, T. (1967). J. Agr. Chem. SOC. lap. 41, 434-441. Asao, Y., Sakasai, T., and Yokotsuka, T. (1969). J. Ferment. Technol. 47, 318-326. Atkinson, R. W. (1881).Proc. Roy. SOC., London 32, 299-332. Baens-Arcega, L. ( 1966 ) . Philippines Patent 2553. Baens-Arcega, L. ( 1969). Biotechnol. Bioeng. Symp. 1, 53-62. Baens-Arcega, L. (1970). Process Biochem. 5, 50-51 and 56. Blaisten, R. J. (1947). Reu. SOC.Bioquim. Argent. 14, 191-192. Bush, L. ( 1959). “Land of the Dragonfly.” Hale, London. Chow, T. Y. (1935). Chemistry (China) 2,280-295. Church, M. B. ( 1923). U.S.,Dep. Agr., Bull. 1152, 1-26. Crewther, W. G., and Lennox, F. G. (1950). Nature (London) 165, 680. Deibel, R. H., and Niven, C. F., Jr. (1960). J . Bacterial. 79, 175-180. Dyson, G. M. (1928). Pharm. J. 121,375-377. Evans, R. J., Banderer, S. L., and Bauer, D. H. ( 1961). J . Food Sci. 26, 663-669. Fukai, T. (1928). J. Agr. Chem. SOC. Jap. 5, 458-470. Fukai, T., and Komatsu, S. (1934a). Rep. Gout. Inst. Brew. Jap. 119, 209-217. Fukai, T., and Komatsu, S. (1934b). Rep. Gout. Inst. Brew. Jap. 119,219-229. Gray, W. D. ( 1970). “The Use of Fungi as Food and in Food Processing,” pp. 225-329. Butterworth, London. Groff, E. H. (1919). Philipp. 1. Sci. 15, 307-316. Harada, Y. (1951). Rep. Tatsuno Inst. Soy Sauce 2, 51-55. Hayashi, K., Fukushima, D., and Mogi, K. (1967a). Agr. Biol. Chem. 31, 642-643. Hayashi, K., Fukushima, D., and Mogi, K. (1967b). Agr. Biol. Chem. 31, 1171-1178. . Biol. Chem. 31, 1237-1241. Hayashi, K., Fukushima, D., and Mogi, K. ( 1 9 6 7 ~ )Agr. Hesseltine, C. W. (1965). Mycobgia 57, 149-197. Hesseltine, C. W. (1966). PTOC.Int. Conf. Soybean Protein Foods, 1966 pp. 170-179. Hesseltine, C. W., and Wang, H. L. (1967). Biotechnol. Bioeng. 9, 275-288. Hesseltine, C. W., and Wang, H. L. (1968). Dry Bean Res. Conf., 9th, 1968 Paper, pp. 1-10. Hesseltine, C. W., Shotwell, 0. L., Ellis, J. J., and Stubblefield, R. D. (1966). Bacterial. Reu. 30, 795-805. Hoffmann, F. B. (1874). Mitt. Ges. Oesterr. 1, 8-11. Hoogerheide, J. C. ( 1954). In “Industrial Fermentations” (L. A. Underkofler and R. J. Hickey, eds. ), Chem. Publ. Co., New York. Ichikawa, K. (1954a). J . Ferment. TechnoZ. 32, 379-384. Ichikawa, K. (1954b). 1.Ferment. Technol. 32,501-506. Iguchi, N. ( 1949). J . Agr. Chem. SOC. Jap. 23, 16-18. Iguchi, N. (1950a). J. Agr. Chem. SOC.Jap. 24, 16-18. Iguchi, N. (1950b). J. Agr. Chem. SOC. lap. 24,283-286. Iguchi, N. ( 1951). J. Agr. Chem. SOC. Jap. 25, 81-84. Iguchi, N. (1952a). J. Agr. Chem. SOC. Jap. 25,459-465.
190
F. M. YONG A N D B. J . B. WOOD
Iguchi, N. (195213).J. Agr. Chem. Soc. Jop. 26, 146. Iguchi, N. ( 1 9 5 2 ~ )J. Agr. Chem. SOC. Jap. 27,229-233. Iguchi, N. (1953). J. Agr. Chem. Soc. Jap. 27, 541-544. Iguchi, N. (1955a). J. Agr. Chem. SOC. lap. 29, 67-73. Iguchi, N. (1955b). J. Agr. Chem.Soc. lap. 29, 73-78. Iguchi, N. (1956). J. Ferment. Technol. 34, 553-559. Iguchi, N., and Yamamoto, K. (1955a). J. Agr. Chem. Soc. Jap. 29, 178-179. Iguchi, N., and Yamamoto, K. (1955b). J. Agr. Chem. SOC. lap. 29, 394-399. Iizuka, H., and Yamazato, K. (1959a). Nippon Nogei Kagaku Kaishi 33, 379 and 383. Iizuka, H., and Yamazato, K. (195913). J. Gen. Appl. Microbiol. 5, 58 (as cited in Sakaguchi and Mori, 1969). Ikeda, K., and Kawaguchi, M. (1922). 1. Chem. Soc. Jap., Pure Chem. Sect. 43, 956-981. Ishida, A. (1925a). J. Soc. Brew., Tokyo 20, 33-35. Ishida, A. (1925b). Rep. Gout. Inst. Brew. lap. 99,342-345. Ishigaki, R., and Nagase, I. ( 1964a). Nippon Jozo Kyokai Zasshi 59(2), 161-162. Ishigaki, R., and Nagase, I. ( 1964b). Nippon Jozo Kyokai Zasshi 59( 6), 507-509. Ishimaru, Y. (1930). Jozogaku Kaishi 11, 439, 576, 689, 778, 882, 986, and 1076 (as cited in Sakaguchi and Mori, 1969). Ishimaru, Y. (1933a). J. Agr. Chem. SOC. Jap. 9, 859-904 and 953-1004. Ishimaru, Y. (1933b). 1. Agr. Chem. Soc. lap. 9, 1143-1183. Ishimaru, Y. (1935). J. Soc. Brew., Tokyo 13, 295-318. Itikawa, S., and Huzi, S. (1939). Japanese Patent 132, 906. Kageyama, K., and Sugita, 0. (1955). J. Ferment. Technol. 33, 100 (as cited in Misaki et aE., 1970). Kato, T., and Matsumoto, K. ( 1964). Nippon ] o m Kyokai Zasshi 59( 5), 431-434. Kawamura, S. (1967). U S . Dep. Agr., ARS ARS 71-35, 249-254. Kawano, Y. (1938). 1. Brew. 16, 755-767. Kellner, 0. (1888). CoZZ. Agr. Tokyo. Uniu. 5, 9 (as cited in Hoogerheide, 1954). Kibi, M. (1926). J. Agr. Chem. Soc. Jap. 12, 885. Kibi, M. (1938). J. Agr. Chem. Soc. Jap. 14, 1094-1102. Kibi, M. (1951). J. Agr. Chem. SOC. Jap. 24, 278-281. Kihara, K. (1940). J. Chem. SOC. Jap., lnd. Chem. Sect. 43, 876-883. Kinoshita, A., and Matsumoto, K. (1935). Rep. Gout. lnst. Brew. Japan 122, 205-211. Kita, G. (1911). J. Chem. Soc. Jap., lnd. Chem. Sect. 14, 109-128. Eta, G. (1913). Wochewchr. Brau. 30, 549, 552. Kodama, S., and Kimura, T. (1931). Japanese Patent 91, 407. Kosaka, Y. (1956). Res. Rep. Noda Soy Sauce Co., Ltd. 1, 81-89. Kubo, T. (1947). “The Composition of Soy-sauce Oil” (unpublished data). Noda Soy Sauce Co. Ltd. (as cited in Yokotsuka, 1960). Kurono, K., and Katsume, E. ( 1927). J. Agr. Chem. Soc. lap. 3,594-613. Kurono, K., Fukai, T., and Tateno, M. (1927). J. Agr. Chem. SOC. Jap. 3,1193-1205. Lillevik, H. A. (1970). In “Methods in Food Analysis” (M. A. Joslyn, ed.), 2nd rev. ed., Chapter XX, pp. 601-616. Academic Press, New York. Lockwood, L. B. (1947). Soybean Dig. 7, 10-11. Lockwood, L. B., and Smith, A. K. (195&1951). Yearb. Agr. ( U S Dep. Agr.) pp. 357361. Lodder, J., and Kreger-van Rij, N. J. W. (1952). “The Yeasts, A Taxonomic Study.” North-Holland Pub., Amsterdam. Markley, K. S. (1950). “Soybeans and Soybean Products,” Vol. I. Wiley (Interscience), New York.
SOY SAUCE FERMENTATION
191
Matsumoto, K. (1923). J. SOC. Brew., Tokyo 19, 14-18. Matsumoto, K. (1925). Rep. Gout. Inst. Brew. Japan 99, 1-187. Matsumoto, K., and Murakami, H. (1941). 1. Soc. Brew., Tokyo 36, 63-71. Matsushima, K. (1954). J. Ferment. Technol. 32, 14-19. Matsushima, K. (1959). Nippon Nogei Kagaku Kaishi 33, 116 and 120 (as cited in Misaki et al., 1970). Matsushima, K., and Shimada, K. ( 1962). Nippon Nogei Kagaku Kaishi 36, 193-197. Mees, R. H. (1934). Ph.D. Thesis, Delft University. Minor, L. J. (1945). Food Ind. 17, 758-760. Misaki, T., Yamada, M., Okazaki, T., and Sawada, J. (1970). Agr. Biol. Chem. 34, 1383-1392. Mitsuda, R. (1910). J. Agr. SOC.lap. 96,43454. Miura, Y. ( 1955). Ann. Sankyo Res. Lab. 6, 14 (as cited in Misaki et al., 1970). Miura, Y. ( 1958).Nippon Nogei Kagaku Kaishi 32, 486490. Mogi, M. (1949). Nogaku 3, 92-96. Mogi, M. ( 1959). Hakko Kyokaishi 17, 78-86. Mogi, M., and Iguchi, N. ( 1958). Nippon JOZO Kyokai Zasshi 53,282-293. Mogi, M., and Iguchi, N. ( 1965). Nippon Jmo Kyokai Zasshi 60( 3), 218-239. Mogi, M., and Iguchi, N. (1966). Nippon Jozo Kyokai Zasshi 61( 3 ) , 231-250. Mogi, M., Nakajima, H., and Ando, T. ( 1951). J. Ferment. Technol. 29,317-321. Mogi, M., Nakajima, H., and Ando, T. (1952). J. Ferment. Technol. 30, 49-59, 152-159, and 194-204. Mori, H., and Onishi, H. (1967). Appl. Microbiol. 15, 928-934. Mori, H., and Onishi, H. (1969). Rep. N o h Inst. Sci. Res. 13, 13-15. Morihara, K., Tsuzuki, H., and Oka, T. (1968). Arch. Biochem. Biophys. 123, 572-588. Morikawa, K. ( 1926). Unpublished Thesis, Massachusetts Institute of Technology, Cambridge, Massachusetts. Morimoto, S., and Matutani, N. (1969). J. Ferment. Technol. 47,518-528. Morimoto, T., and Terui, G. ( 1960). J. Ferment. Technol. 38, 307 (as cited in Misaki et ul., 1970). Murakami, H. (1951a). J. SOC. Brew., Tokyo 46,52-56. Murakami, H. (1951b). J. SOC. Brew., Tokyo 46,90-93. Nakajima, M., Imaki, T., and Takai, S. (1949). J. Chem. SOC. lap., Pure Chem. Sect. 70, 40-42. Nakaya, K. (1934). “Studies on the Treatment of Soybean for the Soy-sauce Fermentation,” Rep. of Noda Soy Sauce Co. Ltd., Vol. 4, pp. 83-108. Noda-shi, Chiba-ken, Japan. Nasuno, S., Ohara, T., and Iguchi, N. (1971). Agr. Biol. Chem. 35, 291-295. Nishimura, T. (1910a). J. Agr. SOC. Jap. 92, 19-37. Nishimura, T. (1910b). J. Agr. SOC.Jap. 93, 1-28. Nishimura, T. ( 1 9 1 0 ~ )J. Agr. SOC.lap. 94, 42-65. Nishimura, T. (1910d). J. Agr. SOC. Jap. 95, 1-17. Nishimura, T. (1910e). J. Agt. SOC.Jap. 98, 18-50. Nishimura, T. (191Of). J. Agr. Soc. Jap. 99, 1-60. Nishimura, T. (191Og). J. Agr. SOC. Jap. 100, 33-62. Nishimura, T. (1911). J. Agt. Soc. Jap. 104, 23-63. Nishimura, T. (1912a). J. Agr. SOC.Jap. 115, 1-10. Nishimura, T. (1912b). J. Agr. SOC. Jap. 116, 1-24. Noda Soy Sauce Co., Ltd. ( 1955). “The Treatment of Soybean in Soy-sauce Fermentation.” Noda-shi, Chiba-ken, Japan (as cited in Yokotsuka, 1960). Nunokawa, Y. (1963). Hdcko Kyokuishi 21(9), 379-385.
192
F. M. YONG AND B. J . B. WOOD
Obata, Y. (1955). Soy-Sauce Stud. 1, 29-55. Obata, Y., and Yamanishi, T. (1951a). J. Agr. Chem. SOC. Jap. 24, 226-229. Obata, Y., and Yamanishi, T. (1951b). J. Agr. Chem. SOC. Jap. 24, 334-338. Obata, Y., and Yamanishi, T. ( 1 9 5 1 ~ )J. Agr. Chem. SOC.Jap. 24, 479-481. Obata, Y., and Yamanishi, T. (1953). J. Agr. Chem. SOC. JQP. 27, 652-654. Oda, M., Ikeda, K., and Tanimoto, M. (1949). J. Ferment. Technol. 27, 16-23. Ohara, H., and Moriguchi, S. (1955a). Seasoning Sci. (Japan)3(3 ) , 12-17. Ohara, H., and Moriguchi, S. (1955b). Seasoning Sci. (Japan)3( 3 ) , 18-26. Ohara, H., and Orishi, R. ( 1951). Rep. Tatsuno Inst. Soy Sauce 2, 1-2. Okazaki, H. ( 1954). 1. Agr. Chem. SOC. Jap. 28, 48 and 54 (as cited in Tamiya, 1957) . Okazaki, H. (1955). J. Agr. Chem. SOC. Jap. 29, 181, 273, and 277 (as cited in Tamiya, 1957). Okuhara, A., and Saito, N. (1970a). J. Ferment. Technol. 48, 177-189. Okuhara, A., and Saito, N. (1970b). J. Ferment. Technol. 48, 190-201. Okuhara, A., Harada, M., Saito, N., and Yokotsuka, T. (1969a). J. Ferment. Technol. 47, 50-56. Okuhara, A., Nakajima, T., Tanaka, T., Saito, N., and Yokotsuka, T. (1969b). 1. Ferment. Technol. 47, 57-58. Okuhara, A., Tanaka, T., Saito, N., and Yokotsuka, T. (1970). J. Ferment. Technol. 48, 228-236. Onishi, H. (1954a). J. Agr. Chem. SOC. lap. 28, 134-140. Onishi, H. ( 195413). J. Agr. Chem. SOC. lap. 28,546-550. Onishi, H . (1957a). Bull. Agr. Chem. SOC.Jap. 21, 137-142. Onishi, H. (1957b). Bull. Agr. Chem. SOC. ]up. 21, 143-150. Agr. Chem. SOC. Jap. 21, 151-156. Onishi, H. ( 1 9 5 7 ~ )Bull. . Onishi, H. (1958a). Bull. Agr. Chem. SOC.Jap. 23, 332-339. Onishi, H. (195813). Bull. Agr. Chem. SOC. Jap. 23, 351-358. Agr. Chem. SOC. ]up. 23, 359-363. Onishi, H. ( 1 9 5 8 ~ )Bull. . Onishi, H. (1960a). Bull. Agr. Chem. SOC. Jap. 24, 126-130. Onishi, H. (1960b). Bull. Agr. Chem. SOC. ]up. 24, 131-140. Onishi, H. ( 1 9 6 0 ~ )Bull. . Agr. Chem. SOC. ]up. 24, 226-230. Onishi, H. ( 1960d). Bull. Agr. Chem. SOC. Jap. 24, 3,86-396. Onishi, H. (1961). Agr. Biol. Chem. 25, 341-349. Onishi, H. ( 1963). Aduan. Food Res. 12,53-94. Onishi, H. (1967). Nippon Jozo Kyokai Zasshi 62( 3 ) , 291-307. Onishi, H. ( 1968). Nippon Jozo Kyokai Zasshi 63(3 ) , 342-359. Onishi, H. and Saito, N. ( 1961). Agr. Biol. Chem. 25, 768-772. Onishi, H., and Saito, N. ( 1962). Agr. Biol. Chem. 26, 245-251. Onishi, H., and Suzuki, T. (1970). Rep. Noda. Inst. Sci. Res. 14, 71-72. Onishi, H., and Saito, N., and Koshiyama, I. (1961). Agr. Biol. Chem. 25, 124-130 (as cited in Onishi, 1963). Ono, A. (1941). “Nihon Sangyo Hattatsu-Shi no Kenkyu (Studies on the Development of Industries in Japan).” Shibun-do, Tokyo. Prinsen-Geerlings, H. C. (1917a). 2. Angew. Chem. 30,256-257. Prinsen-Geerlings, H . C. (1917b). J. SOC. Chem. Ind. London 36, 662-663, Ramsbottom, J. (1936). Brit. Ass. Aduan. Sci., Annu. Rep. pp. 189-218. Saito, K. (1906). “Mikrobiologische Studien uber die Soya” (as cited in Onishi, 1963). Saito, K. (1907). Zentrdbl. Bakteriol., Parasitenk., Infektimskr., Abt. Z, 18, 154 (as cited in Sakaguchi and Mori, 1969). Sakaguchi, K. (1954). J. Agr. Chem. SOC. Jap. 28, 758-764. Sakaguchi, K. (1958a). Bull. Agr. Chem. SOC. lap. 22,345452.
SOY SAUCE FERMENTATION
193
Sakaguchi, K. (195813). Bull. Agr. Chem. SOC. Jap. 22, 353-362. Sakaguchi, K. (1959a). Bull. Agr. Chem. SOC.Jap. 23,22-28. Sakaguchi, K. (195913). Bull. Agr. Chem. Soc. Jap. 23, 100-106. . Agr. Chem. SOC. Jap. 23, 438442. Sakaguchi, K. ( 1 9 5 9 ~ )Bull. Sakaguchi, K. ( 1959d). Bull. Agr. Chem. SOC. lap. 23,443-450. Sakaguchi, K. (1960a). Bull. Agr. Chem. SOC. Jap. 29, 489496. Sakaguchi, K. (1960b). Bull. Agr. Chem. SOC. Jap. 24, 638-643. Sakaguchi, K., and Mori, H. (1969). J. Gen. Appl. Microbial. 15, 159-167. Sakaguchi, K., and Yamada, K. (1944). J. Agr. Chem. SOC. Jap. 20, 64, 141 (as cited in Sakaguchi, 195913). Sakaguchi, K., and Yamada, K. (1945a). J . Agr. Chern. SOC. lap. 20, 65-73. Sakaguchi, K., and Yamada, K. (194513). J. Agr. Chem. SOC. Jap. 20, 141-154. Sakamoto, M., and Shuzui, K. (1957). J. Ferment. Technol. 35, 137-142. Sakasai, Y., and Yoshida, T. (1966). J. Ferment. Technol. 44, 61-72. Sakurai, Y. (1965). In “Food Technology the World Over” ( M . S. Peterson and D. K. Tressler, ed.), Vol. 11, pp. 381-383. Avi, Westport, Connecticut. Sawasaki, T. (1960a). Nippon Nogei Kagaku Kaishi 34,l-5. Sawasaki, T . (1960b). Nippon Nogei Kagaku Kaishi 34, 5-8. Sekine, H., Nasuno, S., and Iguchi, N. (1969). Agr. Biol. Chem. 33, 1477-1487. Sekine, H., Nasuno, S . , and Iguchi, N. (1970). Agr. B i d . Chem. 34, 1690-1696. Shoji, K. (1927). J. Agr. Chem. SOC. Jap. 3, 1126-1136. Smith, A. K. (1949). “Oriental Methods of Using Soybeans as Food, with Special Attention to Fermented Products,” pp. 1-40, U.S. Bur. Agr. Ind. Chem. Snelling, W. 0. (1937). U.S. Patent 2,107,133. Stahel, C. (1946). J . N. Y. Bot. Gard. 47, 261-267. Sugimori, T., and Joze, H. (1970). J. Ferment. Technol. 48,218-227. Sugita, N. (1956). 1. Ferment. Technol. 34, 607-612. Suzuki, M., Nunokawa, Y., and Ito, Y. (1959). Nippon Nogei Kagaku Kaishi 33, 857 (as cited in Misaki et al., 1970). Taira, T. ( 1925). J. Agr. Chem. SOC. Jap. 2, 734-743. Taira, T. (1926). J. SOC. Brew., Tokyo 21, 31. Takahashi, T., and Yukawa, M. (1911). 1.Agr. SOC. Jap. 112, 1-26. Takahashi, T., and Yukawa, M. (1914). J. Agr. SOC. Jap. 133, 1-25. Takahashi, T., and Yukawa, M. (1915). J. Imp. Uniu. Jap., Mem. Coll. Agr. 5, 227-26 1. Takamine, J. (1894). U.S. Patents 525,823 and 525,825. Takeda, R. ( 1954). Seasoning Sci. ( J a p a n )2, 23-28. Tamiya, H. (1958). Proc. Int. Symp. Enzyme Chem., 1957 pp. 21-24. Tateno, M., and Umeda, I. (1955). Japanese Patent 204,858. Tawara, Y. (1887). I. Pharm. SOC.Jap. 61, 80-89. Tsujita, Y. (1967). Ann. Sankyo Res. Lab. 19, 144-146. Tsukahara, T. ( 1948). Hakko Kyokaishi 6, 49-51. Udo, S. (1931a). J. Agr. Chem. SOC. Jap. 7, 322-327. Udo, S. (191313).J. Agr. Chem. SOC.Jap. 7, 852-858. Ueno, T., and Kuramochi, T. (1960). Nippon Nogei Kagaku Kaishi 34, 187-193. Ueno, T., and Kuramochi, T. (1961). J. Agr. Chem. SOC. Jap. 35, 454458. Ueno, T., and Nobuhara, A. (1960a). Nippon Nogei Kagaku Kaishi 34,223-228. Ueno, T., and Nobuhara, A. (1960b). Nippon Nogei Kugaku Kaishi 34,229-233. . Nogei Kagaku Kaishi 34,566-569. Ueno, T., and Nobuhara, A. ( 1 9 6 0 ~ )Nippon Ueno, T., and Nobuhara, A. (1960d). Nippon Nogei Kagaku Kaishi 39,569-572. Ueno, T., and Omori, S. ( 1951). Rep. Tatsuno Inst. Soy Sauce 2, 51-55. Umeda, 1. ( 1957). Nippon Nogei Kagaku Kaishi 31, A48-A52.
194
F. M. YONG AND B. J. B. W O O D
Umeda, I., Ishizuka, Z., and Saito, N. (1953). Seasoning Sci. (Japan) 1 ( l ) , 29-40. Waksman, S. A, and Davidson, W. C. ( 1926). “Enzymes, Properties, Distribution, Methods, and Applications,” pp. 266-267. Baillibre, London. Wood, B. J. B., and Yong, F. M. (1974). In “The Filamentous Fungi” ( J . D. Smith and D. R. Berry, eds.), Vol. I, Academic Press, New York. Yamada, K., and Furusaka, T. (1954). Seasoning Sci. (Japan) 1 ( 3 ) , 14-19. Yamaguchi, M. (1958). Japanese Patent 1,598. Yamamoto, K. (1957a). Bull. Agr. Chem. SOC. l a p . 21, 308-312. Yamamoto, K. (1957b). Bull. Agr. Chem. SOC. Jap. 21, 313-318. . Agr. Chem. SOC. Jap. 21, 318-324. Yamamoto, K. ( 1 9 5 7 ~ )Bull Yasui, T. (1964). Nippon Nogei Kagaku Kaishi 38,361-366. Yauchi, J., Takita, K., and Amemiya, K. (1955). Seasoning Sci. (Japan) 3 ( 4 ) , 18-22. Yenko, F. M., and Baens-Arcega, L. (1940). PhiZipp. J. Sci. 71, 1-3. Yokotsuka, T. (1949). J. Agr. Chem. SOC. Jap. 23, 200-205. Yokotsuka, T. (1951a). J. Agr. Chem. SOC. Jap. 24,355458. Yokotsuka, T. (1951b). J. Agr. C h m . SOC. Jap. 24, 358362. Yokotsuka, T. ( 1 9 5 1 ~ )J. Agr. Chem. SOC. Jap. 24,402-406. Yokotsuka, T . (1951d). J. Agr. Chem. SOC. Jap. 24, 407-409. Yokotsuka, T. (1952a). J. Agr. Chem. SOC. Jap. 25,206-210. Yokotsuka, T. (1952b). J. Agr. Chem. SOC. Jap. 25, 401-405. Yokotsuka, T. ( 1 9 5 2 ~ )J. Agr. Chem. SOC. Jap. 25,446450. Yokotsuka, T. (1953a). J. Agr. Chem. SOC. Jap. 27,276-281. Yokotsuka, T . (1953b). J. Agr. Chem. SOC. Jap. 27,334-339. Yokotsuka, T . ( 1 9 5 3 ~ )J. Agr. Chem. SOC. Jap. 27, 359-363. Yokotsuka, T. (1953d). J. Agr. Chem. SOC. Jap. 27, 549-553. Yokotsuka, T. (1954). J. Agr. Chem. SOC. Jup. 28, 114-118. Yokotsuka, T. (1957). J . Agr. Chem. SOC. Jap. 32, 23-26. Yokotsuka, T. (1960). Adoan. Food Res. 10, 75-134. Yokotsuka, T. (1964). Int. Symp. Oilseed Protein Foods, 1964 pp. 31-48. Yokotsuka, T., Okuhara, A., and Takimoto, K. (1958). Seasoning Sci. (Japan) 6, 18-29. Yokotsuka, T., Sakasai, T., and Asao, Y. (1967a). J. Agr. Chem. SOC. Jap. 41, 428-433. Yokotsuka, T., Asao, Y., and Sakasai, T. (1967b). J. Agr. Chern. SOC. Jup. 41, 442447. Yong, F. M. ( 1971). M.Sc. Thesis, University of Strathclyde, Glasgow. Yong, F. M., and Wood, B. J. B. (1974). Trans. Brit. Mycol. SOC. (in press). Yukawa, M. (1915). J. Imp. Uniu. Jap. Coll. Agr. 5,291-299. Yukawa, M. (1916). J. Chem. SOC. Jap., Pure Chem. Sect. 38, 358-374.
Contemporary Thoughts on Aspects of Applied Microbiology'
P. S. S. DAWSON AND K. L. PHILLIPS National Research Council of Canada, Prairie Regional Laboratory, Saskatoon, Saskatchewan, Canada I. Introduction ................... 11. Biological Considerations .............. A. Cultivation Methods for Microbes (and Cells) ... B. Influence of the Cultivation System on Microbial Growth ................... C. Physiological Aspects of Growth ......... D. Further Problems Relating to Cell Metabolism ... E. Rationale of Growth .............. 111. Nonbiological Considerations ............ A. Environmental Factors ............. B. Physical ................... C. Chemical .................. D. Interplay of Factors .............. E. Equipment-Design and Operation ....... IV. Mathematical Considerations ............ V. General Considerations .............. References ....................
I.
195 196 196 202 205 209 211 213 213 214 218 219 221 224 227 230
Introduction
Applied microbiology, like all biological disciplines, is not an exact science: it lacks the firm base which theorems and principles give to mathematics and the physical sciences. An empiricism, which permeates and predominates throughout, pervades theory, operations, and equipment, and prevents a rational development of the subject in practice. It would be advantageous, in many ways, to be able to make things quantitative, predictable, and absolute. Peripherally, applied microbiology embraces and encroaches upon many different disciplines and specialties; physiology, engineering, biochemistry, adsorption, tissue culture, extractions of various kinds, to mention just a few. All these activities, related in some way to the central core of cellular activity, have little or no obvious interconnections. Inevitably problems originating in these interdisciplinary areas, especially those of a specialist nature, become oriented along the axis of the specialist-the physicist sees the problems as those of physics, the engineer, biologist, or chemist likewise-so that problems, projects, processes, and procedures evolve, develop, are reported, and become accumulated as empirical data in a burgeoning literature. This vast store
' Issued as NRCC No.
13827. 195
196
P. S. S. DAWSON AND K. L. PHILLIPS
of information and knowledge lacks the cohesion, integration, and reproducibility necessary to build upon it in a positive manner. One has only to think of the inadequacy of batch data for the development of successful continuous culture processes, of the difficulties of scale up, or even of the sometimes erratic reproducibility in yield and performance of batch cultures, to see how frustrating the present situation can be for advancing applied microbiology. Contemplating the present extensive scope of applied microbiology, one realizes that although it is difficult to recognize and delineate its boundaries, at the center, and easily recognizable, are the microbes and their activities. Recognition of this microbial base does not solve, but rather pinpoints, the problems of applied microbiology, and locates them in the empiricism that largely arises from this biological center: to attempt any rationalization, one must inevitably start from there. The problem is not so simple as that, however, because applied microbiology, being essentially multidisciplinary in extent, involves other physical, chemical, and mathematical factors too. However, as organisms are the essential prerequirements to the subject, biological considerations are primary, and it is logical and convenient to consider them first. Microbiology, as the necessary preliminary to applied microbiology, requires some initial scrutiny or attention.
II.
Biological Considerations
A. CULTIVATION METHODSFOR MICROBES( A N D CELLS) In the past, as the study of the single cell or microbe was impracticable, pure cultures of organisms grown in batch culture were, of necessity, used instead. This procedure influenced the development of microbiology in two important ways: first, the batch culture was consolidated as a foundation upon which the discipline presently rests; and second, the expediency of substituting the cell population for the cell as the basis for study complicated and confused many problems associated with microbial growth. It is academic and irrelevant to question that pure cultures are, in the first instance, artifacts, for without them our knowledge of microbiology would not have progressed very far; but it is becoming increasingly relevant now, and realistic too, to question the suitability of the batch culture to continue serving in its long established role as the universal provider and arbiter of microbial knowledge.
ASPECTS OF APPLlED MICROBIOLOGY
197
It may appear to be unimportant that populations, instead of individuals, have been employed to acquire much of the knowledge of how cells perform, but studies of microbial growth made during the last two decades have revealed certain differences that conflict with some of the ideas derived or based up011 experiences with batch cultures. The dominance of the batch culture in microbiology is challenged now by new techniques which encourage novel, and somewhat heretical, developments for microbiological studies. The new techniques of continuous and synchronization methods for cultivating cells, developed during the past twenty years, permit the study of the cell alongside that of the population in the culture and thereby release the microbiologist from his previous total subservience to the cell population. Most significant in these developments is the emergence of new experimental parameters, previously unrecognized, which show a pronounced shift from the population to the cell as the basis for study and application. Thus, new dimensions, wider perspectives and promising developments exist, which have opportunities for reassessing and advancing some of the trends currently held in microbiology; these have implications for advancing applied microbiology too. It is necessary to consider these different developments a little more closely now. 1. Batch Culture There was little fundamental understanding of what happened in a culture until Monod (1942) published the results of his now classical researches on the growth of bacterial cultures and revealed the importance of the substrate in controlling the growth obtained in a batch culture. For several decades prior to this event, many workers had recognized the existence of a general pattern in the changes taking place during the development of batch cultures, but without avail. The basic pattern, usually outlined as a growth curve, was associated with the so-called “growth cycle”-a term that had come to be used to describe the progression in the batch culture of a newly inoculated seed of organisms which multiplied and developed into a population, that thrived, flourished, declined and then, if transplanted before dying, repeated the procedure. The growth curves of batch cultures were, in fact, simply empirical traces that recorded the transient histories of the cultures. In each culture, fresh medium was consumed and converted into products by the growing cell population, and entailed accompanying changes in morphology, mass and numbers, besides alterations in the physiology of the organisms. Two relatively distinct areas of different cellular activity were recognized : the early stage of multiplication by
198
P. S. S. DAWSON AND K. L. PHILLIPS
“proliferating cells” and the later period of quiescent change by so-called “resting cells.” Invariably, however, these serial performances were not exactly reproducible. Consequently, during this early period of studies in microbial growth, no tangible basis could be recognized to account for the kaleidoscopic variations in growth cycle patterns, and “a consensus of opinion considered (microbial) growth to be a complex phenomenon” (Van Niel, 1949). The growth curves became meaningful, and of use quantitatively as analytical tools, when Monod (1942) showed that the amount of growth and the rate of growth in a culture were related to the amount and concentration of the limiting substrate (i.e., the nutrient component in least supply) present in the medium. Monod described growth constants specific for an organism and the substrate, and showed also that one limiting substrate could follow the depletion of another. Thus, in different nutrient media, changes in the pattern of the growth curve would reflect the pattern of usage and exhaustion of the growth-limiting substrates: singly in simple chemically defined nutrient media or successively in complex media. When adaptation (induction) to an alternative substrate occurred in the culture, such that an interim lag period was involved, Monod recognized the condition of “diauxie.” The lag phase at the start of a batch culture was likewise seen as a period of initial adjustment which enabled the cells to approach their state of maximal growth rate. In a subsequent review on growth, Monod (1949) summarized the study of batch culture at the time when the technique was at its zenith. Thus, it became clear that a batch culture demonstrates the ability of a population of cells to attain and maintain a maximal rate of growth on a substrate until depletion and exhaustion of that substrate in the culture terminates growth. Variations in growth rate can occur that reflect other extraneous environmental influences too, as for example temperature effects, but at any moment, it is the concentration of the specific limiting substrate in the culture that is of primary importance for the growth rate of the cells. The significance of these developments was 2-fold: primarily, the real basis of the batch culture had been defined; and second, the development of the theory of continuous culture was originated. Empirical applications of batch cultures are still widespread in microbiology, but the method is now being used more circumspectly. For meaningful work it has two main uses: (1) for pursuing growth studies during the relatively uniform period of growth at constant (maximal) rate in the logarithmic or exponential phase, and ( 2 ) for indicating the overall range of an organism’s abilities; expediency is usually the sole justification for its use in most other applications.
ASPECTS OF APPLIED MICROBIOLOGY
199
2. Continuous Culture Having recognized the part played by the nutrient concentration in the control of growth, Monod realized that if the substrate concentration could be maintained indefinitely constant in a growing culture, then it should be possible to continue growing cells indefinitely, so that a continuous culture would exist; furthermore, if the substrate concentration could be maintained at any fixed value, then continued growth at a corresponding fixed rate should be obtained, so that continuous culture at any desired growth rate would be practicable. As an extension of his growth constant formulation, Monod (1950) was able to develop and propose his theory of continuous culture. For the practical realization of this idea a simple device, that of constant dilution at constant volume, was used to obtain a condition of equilibrium in the culture vessel. In this steady-state condition, a constant residual concentration of nutrient was established in the culture which controlled the growth rate of the cells, and concomitantly, by continuous removal of the growth increase with the flow of medium producing it maintained the cell population at a constant level. Independently, the same arrangement was evolved by Novick and Szilard (1950) for their chemostat, publication of which coincided with that of Monods bactogen. A variation of this arrangement, the turbidostat of Bryson (1952), was also contemporary with these developments. Other spasmodic and empirical attempts at continuous-flow culture had been made previously ( Whalley, 1955), but lacking the basic information necessary for successful operation, such endeavors had to await the advent of the Monod theory for their fruition. However, having produced his theory and shown practically that continuous cultivation of bacteria could be obtained, Monod had little further interest in his bactogen. It was fortunate, therefore, that at this point Herbert and the Porton School played a crucial part by substantiating, developing and expounding the Monod theory (Herbert et al., 1956)-so that it became more widely known and increasingly appreciated. With practical advantages to be obtained from continuous methods of operation, which had previously been demonstrated by chemical engineers in chemical industries, together with the novel opportunities for studying growth in a condition of equilibrium, many microbiologists and technologists soon realized that the technique had much potential, and the study of continuous cultures actively commenced. An early blossoming of continuous culture methods which followed soon produced many diverse experimental arrangements ( Ricica, 1958), sometimes with results quite different to those expected or obtained with batch techniques.
200
P. S. S. DAWSON AND K. L. PHILLIPS
For instance, it was largely anticipated initially, on the basis of Monod‘s theory, that by the systematic operation of a simple chemostat at different rates, the spectral change common to the sequence of a batch culture might be obtained as an array of steady states. In practice this was not wholly realized: reasonable agreement was obtained between batch and continuous cultures at faster growth rates, but wide discrepancies in yield and metabolic performance generally occurred at slower growth rates. Herbert (1959) showed that some of these differences could be accounted for in a number of ways, for example, by assimilation or endogenous maintenance occurring in the cells, or due to inadequate fermenter design and performance. But other differences, especially the formation of so-called secondary metabolites, like the production of many antibiotics, could not be explained. Thus, a dichotomy in growth studies became evident: at the faster rates of growth, a quantitative agreement between batch and continuous cultures which showed little divergence from the Monod theory, but at slower growth rates, an ever-widening qualitative disparity in the physiological properties of the cells grown by the two techniques. There appeared to be something lacking in Monod’s theory, and it seemed likely that an understanding of physiological aspects of growth could be important for the desired operation and use of continuous culture. This view was strongly championed by Malek (1958) and his co-workers in Prague. Malek considered the “physiological state”-the summation of changes taking place in a culture-to be basically important to the function and successful operation of continuous cultures. The important distinction between multiplication and development, well known to exist in batch culture studies, was not likely to be detected in simple ( single stage) cultivation systems where multiplication was an obligatory condition for their successful operation. Consequently, once the possible importance of a qualitative nature to growth was realized, multistage arrangements were introduced ( Powell and Lowe, 1962; Herbert, 1962; Hroncek, 1962; Kozesnik, 1962). These are now being widely employed in the search for an answer to what appears to be the outstanding problem-the resolution of Malek‘s physiological state and its connection with growth rate (see Section I1,C).
3. Synchronization Culture Following closely upon these revolutionary developments in methods for cultivating microbes was another; the introduction of synchronization followed that of continuous culture within a decade (Campbell, 1957; Maal@e,1962), In these synchronization methods, the cells comprising the population in a culture grow “in step” and proceed through their
ASPECTS OF APPLIED MICROBIOLOGY
201
cycles of replication simultaneously; so that all the cells are doing the same thing at the same time in the culture. The earliest methods devised, those of synchronized (Scherbaum and Zeuthen, 1954) and synchronous culture ( Maruyama and Yanagita, 1956), sometimes alternatively referred to as induced and selected synchrony, respectively (James, 1966), show developmental characteristics of batch culture. In both of these methods a population of cells growing “in synchrony” is used as an inoculum to initiate a batch culture, the first few generations of which retain the synchrony and thus permit examination of the cell cycles. The first methods to be developed for attaining synchrony, those of synchronized culture, relied upon the application of various constraints to align, and thus synchronize, the cells in a randomly dividing population, so that upon release from the constraint, the entrained cells grew and multiplied together: oscillations of temperature, additions of threshold concentrations of growth inhibitors, or withholding nutrients were some of the constraints used (James, 1966). Some objections to the use of constraints, of their nature or means by which they were used to align the cells, were made: and it was suggested that upon release from the constraint cells were “convalescent” and not functioning in a nomial manner ( Abbo and Pardee, 1960). Selection by inert physical methods, of filtration ( Maruyama and Yanagita, 1956) or centrifugation ( Mitchison and Vincent, 1965), or later by membrane filtration (Helmstetter and Cummings, 1963) upon an actively growing population was proposed to obtain a fraction of cells at the same stage of growth which could then be used as an inoculum for growth of a synchronous culture, and was developed as an alternative method, supposedly nondisturbing to the cells. It is doubtful if this advantage is always achieved in practice, however, because frequently the manipulations used are not inconsequential (Hattori et al., 1972). In either synchronous or synchronized types of culture, the inoculum grows batchwise giving several ( usually 2.3) generations of cells in synchrony, although rarely in an exactly reproducible manner. The technique can only be used to examine cells growing at one rate, that maximal for the medium being used. In the continuous methods of synchrony, evolved during the past ten years, cell performance in successive generations remains the same and can be examined at any desired growth rate. Two general types can be recognized: (1)pulsed methods of synchrony, in which a chemostat culture grows on a basal medium deficient in the limiting nutrient but receiving the limiting nutrient at intervals corresponding to the doubling time of the cells (Hansche, 1969; Goodwin, 1969), and ( 2 ) phased culture, where the complete medium is added at doubling time intervals in a continual manner (Dawson, 1965). In the pulsed method,
202
P.
S. S. DAWSON AND K. L. PHILLIPS
a continuous flow of medium dilutes away some of the culture synchronized by the nutrient doses; but in phased culture the whole population is retained during its multiplication in the cell cycle, which occurs under temporarily “closed conditions. The overall operation of both pulsed and phased methods is continuous, however, and the systems are “open” ones (see Section II,B,l). In most synchronization methods it is possible to examine the cells only as they are multiplying, so that only aspects of cultivation related to multiplication can be followed. The phased method, however, has the advantage of enabling the postcycle activity (i.e., the behavior of the cell population after completing the cell cycle) to be observed (Dawson, 1970). This application corresponds to the second and later stages of multistage chemostat systems, but employs a homogeneous and synchronized population instead of the heterogeneous populations of variable age structure found in the latter (see Section II,C,2). In phased cultures, the cell cycle can therefore be considered to reflect the cellular activities involved in the multiplication of the cell, and the post-cycle the later developmental activities of the cell (Dawson, 1972). OF THE CULTIVATION SYSTEM B. INFLUENCE ON MICROBIALGROWTH
Since the introduction of continuous culture methods, about 1950, and the more recent developments of synchronization techniques, it has become apparent that the system and method of cultivation used for growing microbes is largely responsible for the type of result to be obtained from a culture. If it is not yet absolutely clear how these results come about, it is becoming clearer, and for this clarification, we are largely indebted to the contributions of three investigators: to Monod, for discovering the quantitative basis of microbial growth; to Herbert, for discerning the coordination with environment; and to Malek for realizing the qualitative nature of these changes.
1. Closed and Open Systems Initially, the theoretical conception and practical realization of continuous culture demonstrated the general validity and application of the Monod growth constants, and established the common basis for wide and seemingly different manifestations of microbial growth. Herbert’s contributions were very important in this connection, and especially those that subsequently distinguished between “closed” and “open” systems, for they showed how these systems, by controlling the nutrient
ASPECTS OF APPLIED MICROBIOLOGY
203
environment in a culture, thereby regulated and controlled the growth of the cells (Herbert, 1961a,b), In “closed” systems microbial cells are retained within the system and changes take place with elapse of time, so that in batch culture-a closed system-the developing population continuously modifies its environment and, by depleting the nutrient substrate, itself becomes modified by the changes in growth rate and metabolism of the cells produced. In “open” systems, cells continually leave the system at the same rate at which the nutrient inflow occurs and, when new cells are generated within it at the same rate, steady-state operation is possible because there is no overall change within the system with passage of time. This situation is common to continuous-flow cultures, and in the chemostat the flow rate establishes the residual nutrient concentration in the vessel by which the growth rate of the cell population is maintained constant. In this way, the apparent paradox of growth being a constant, and yet changing phenomenon was explained by its relationship to the concentration of the nutrient available in the culture; and the so-called “growth cycle” of batch culture was revealed as an artifact of growth taking place in a closed system ( Herbert, 1961a). With this basic distinction between batch and continuous culture established, Herbert (1961b) proceeded to show that the many different methods available for the cultivation of microbes could be divided into the two groups of closed and open systems, both of which, in turn, could be subdivided into homogeneous and heterogeneous ( including mixed) types, comprised of single or multiple stages. Homogeneous systems have a uniform composition throughout ( e.g., stirred vessels ) , but heterogeneous systems exhibit concentration gradients of cells and of substrates within the system (e.g., pipe flow). In operation, the kinetics of homogeneous and heterogenous types of open systems differ markedly and are important. In homogeneous systems, uniform steady-state conditions exist in the culture; whereas in heterogeneous systems, a stabilized sequence of change akin to the so-called “growth cycle” of batch culture may be obtained and along which the organisms pass: the former permits steady-state operation in an unchanging environment with a fixed physiological state, but in the latter a stabilized change of environment gives a serial change of physiological state in the culture. Theoretically, these differences offer the possibility for comparing continuous and batch cultures directly, but unfortunately, as tube fermenters can be satisfactorily operated only under anaerobic conditions, the opportunity is very restricted and largely lost. Homogeneous types of open systems are necessarily single-phase systems, but they may be joined serially into multistage arrangements.
204
P.
S. S. DAWSON AND
K. L. PHILLIPS
Heterogeneous systems may exist in single phase (pipe flow) or multiphase (packed towers) types and may be compounded together, and with homogeneous types, into multistage assemblies. Some arrangements have more importance than others, but any of them may be elaborated further by partial recirculation of cells and additions of supplementary feeds at different points. Herbert (1961b) has discussed some of these arrangements in detail, and developed them in a formal manner as pro: jections of the Monod theory. With the introduction of synchronization, it is apparent that the general properties relating to “open” and “closed systems of asynchronous cultures apply to synchronous ones too. Recently, Dawson (1972) has indicated these relationships (Table I). TABLE I TECHNIQUES FOR THE CULTIVATION OF MICROBES Type
System
Technique
Asynchronous
Closed Open
Synchronous
Closed
Batch Chemostat Turbidostat Synchronous Synchronized Pulsed Phased
Open
2. Relationships of the Culture with Time Other developments, related to the operation of closed and open systems, appeared when the early results from work with continuous cultures began to accumulate. The fundamental difference that exists between batch and continuous cultures in relation to time was soon recognized. Batch cultures develop with time and have a history; each generation differs from that which precedes it, so that changes are related to the age of the culture. Continuous cultures, although developing in time, are independent of time and have no history; and in the absence of genetical change, each generation resembles that which precedes and follows it: the culture is without age. Nevertheless, common to both batch and continuous cultures is the division cycle (or cell cycle)-the reproduction period of the cell; a fundamental property of the organism, applicable in any culture, and itself a function of time. In batch culture, this function changes according to circumstance, and is difficult to measure, but in continuous and syn-
ASPECTS OF APPLIED MICROBIOLOGY
205
chronization cultures it is a critical experimental parameter-the operational time base of the method, constant and very easily measured. Hence, continuous culture automatically relates growth on an absolute basis-in terms of the doubling time-to the cell, while synchronization methods can extend the relationship further to that of the cell throughout the cell cycle. Thus, different cultivation methods confer different modes of growth on the cells, and also different dimensions.
3. Cell vs. Populations The newer techniques found reluctant acceptance generally in microbiology (Tempest, 1970), so that progress has been slower than it might have been: only now are continuous-flow cultures becoming widely used, but acceptance of synchronization as a general technique is still tardy despite its proved capacity for revealing many new perspectives for cell growth. However, the relatively small amount of work so far performed with the newer techniques makes it evident that cells function in a variable manner during the cell cycle and that such functions change and alter with circumstances. It would appear that results obtained in various cultivation systems originate from the many variable ways in which microbial units can perform and be compounded. Some of these developments were anticipated some time ago by James (1961). This brings into focus the differences that can arise from using the cell, as distinct from the cell population, as the basis for study: a precise resolution of the unit, attainable on the one hand, has to be distinguished from, and compared with, the mediocrity of the average or randomized mean obtained from the complexity of the other. It also follows that to sort out the difficulties associated with the cell growth and function, one should perhaps consider the unit and its integrations, rather than, as heretofore, the population with its complex factions and functions. Before considering these possibilities, it is necessary to look first at some of the discrepancies, largely physiological, presently associated with cell growth and function observed in different cultivation systems.
C. PHYSIOLOGICAL ASPECXS OF GROWTH
1. Recognition of Physiological Aspects of Growth The discrepancies observed in performance of cells grown at different rates in batch and continuous cultures, already mentioned (Section II,A,2), ultimately drew attention to the significance of physiological
206
P. S. S. DAWSON AND K. L. PHILLIPS
aspects of microbial growth-the importance of which had long been neglected. An interest in physiological aspects of cell development had been evident during the two decades preceding Monod‘s investigations, but its effects had been peripheral rather than central to the problem of growth. Thus, the sequential phases of the growth curve (Buchanan, 1918) had encouraged greater interest in the characteristics of the cells in the different stages and led to concepts of “physiological youth (Sherman and Albus, 1923) and “senility,” of morphological change (Henrici, 1928), studies with “proliferating and “resting” cells in the Cambridge School, and various other endeavors (Winslow and Walker, 1939). However, all these developments did not greatly influence the basic study of growth which was still seen to be related to the quantitative incremental changes taking place in the mass and number of cells in a batch culture. Monod’s analysis, which stands as the most distinguished example, reflected this general and major trend of studies into microbial growth performed at that time. Possibly these developments were related to the experimental techniques available to investigators at that time-the experimental parameters of mass and number were the most obvious and easiest to measure, while the other less tangible changes in metabolism were not so easily recognized and methods for evaluating them, being rudimentary or nonexistent, tended to diminish their significance. Thus, later, when Monod (1950) developed his theory of continuous culture, although aware of the existence of physiological change (Monod, 1942, 1949, 1950), he chose to assume that cells in a culture developed in a uniform manner and thereby limited any universal application of the theory in practice. For, by making this simple assumption, any implications arising from physiological change were avoided and the theory became primarily concerned with multiplication of the cells in a culture. Similar reservations apply to the work of Novick and Szilard (1950) and other contemporary workers. A deepening interest in physiological aspects of growth was nevertheless developing in the realm of batch cultures during this period. Several attempts at continuous cultivation of cells were being recorded (McClung, 1949). These early excursions into continuous culture were seen, by their originators, to be extensions of the batch method, although Utenkov, and later Malek, considered the procedures to be a new experimental method especially useful for investigating the physiology of cells (Malek, 1958). However, in the nature of pioneers, these early steps were very empirical and, lacking any rational basis on which to build, remained of little value until the advent of the Monod theory. With the recognition of continuous culture as a separate technique,
ASPECTS OF APPLIED MICROBIOLOGY
207
Malek was soon able to establish that physiological considerations were important in cell growth: he indicated that both qualitative and quantitative changes were involved, and introduced the concept of the physiological state (Malek, 1958). Thus it came to be seen that microbial growth could no longer be considered as a simple exercise in mass and numbers, but must involve problems of cell development coordinated with those of cell multiplication. This fact, which would appear to be self evident now, nevertheless often goes unheeded. Cells must enlarge before they can divide, so that although, a priori, growth entails problems of mass and number, cells growing at different rates produce mass increases that are different both quantitatively and qualitatively. It should be remembered too, that cells which simply increase in size are no longer vitally active but moribund. 2. Investigations of the Physiological State The important distinction between multiplication ( the quantitative change in numbers) and development (the qualitative changes in mass), long appreciated in batch culture studies, is partly obtained, by segregation, in multistage systems of continuous culture; where multiplication largely takes place in the first (propagator) stage and development in the subsequent stages. The achievement is not a precise and completely predictable one however. The introduction of multistage systems of continuous culture, as an attempt to resolve qualitative aspects of microbial growth-Malek‘s socalled “physiological state,” has met with only limited success, as yet. Some practical bridging of the gap between batch and continuous culture has been reached, but in an empirical manner rather than by precise experiment. As the proceedings of the last three International Symposia (Malek, 1967, 1969; Powell, 1969, 1972) on continuous culture concede, an adequate theoretical translocation is still unattained. There would appear to be a number of reasons contributing toward these difficulties. ( a ) At the outset, the useful treatment of multistage systems by Herbert (1961b), already referred to, suffers from the same simple assumption which restricts the application of Monods theory-no allowance is made for the effects of physiological change taking place in the cells growing in different stages of continuous culture systems. But there are other difficulties too. ( b ) The stepwise passage of cells through multistage systems will result in changes taking place in the cells transferred, similar to those reported by Kjeldgaard et al. ( 1958) : for instance, “stepdown” effects must be expected to occur in cells progressing along a chain of homolo-
208
P. S . S . DAWSON AND K . L. PHILLIPS
gous reactors, and “step up” changes in cells being recirculated to earlier vessels in a system. These effects will be more pronounced in the cells as the distance between the stages widens. ( c ) The age structure of the cell population, and of the accompanying cellular physiology, will vary enormously in different systems too. The simple randomization that exists in a homogeneous population growing at a fixed rate in the single-stage chemostat will become complicated by the stepwise changes and the effects of flow through succeeding stages in multistage systems. It is possible to calculate statistically the distribution and mean of cell residence times in such systems (Painter and Marr, 1968), but the corresponding physiological activities of the cell populations (Malek‘s physiological state) cannot be so easily divined. Variations in the proportions of different types of cells and of cellular activity present in cell populations, though yielding apparently similar mean growth rates, can produce widely different cultures: such changes probably account for some of the hysteresis type changes now being reported (Tempest, 1970). ( d ) Variations in the viability of cells constituting the cell population in these different systems, or stages of a system, will likewise complicate the net metabolic activity at any point in the system. The effects will differ too between the stepwise changes of homogeneous (cascade) and the graded changes in heterogeneous ( pipeflow) type systems. ( e ) Another fundamental difference, often overlooked, between batch and continuous culture is that posed by the nutrient supply. In the chemostat, fresh medium is always present in the culture, but in batch cultures many nutrient constituents are removed progressively with the passage of time. It follows therefore, that repression effects removed during batch culture are likely to remain in the chemostat by virtue of the incoming flow of fresh nutrient medium into the culture. ( f ) In addition to these complications, there remains that, so far neglected, of distinguishing between the behavior of the microbial unit, the microbe, of which the cell population is formed, and the average mean of that population. The effects of changes undergone by a cell during the cell cycle have still to receive adequate consideration in relation to the many anomalies that exist in performance between different types of culture. It is erroneous to assume that a statistical mean, or net average, of a cell population faithfully reflects cell function, and should be accepted without question as the basic arbiter of cell growth. The invariant performance of a chemostat steady state, or “balanced growth” in the exponential phase of batch culture, is not completely representative of the operational unit-the microbial cell. The general average level of the total individual operations, seen to be constant in the chemostat or changing during the batch sequence is important
ASPECTS OF APPLIED MICROBIOLOGY
209
in itself, and reflects the performance of the culture, but fails to indicate how the individual cells perform, which is the crucial factor that decides how the whole population ultimately behaves in toto in any culture. It would appear, therefore, that the physiological state, as the summation of the metabolic activities in a culture, is dependent upon the individual cells, growing in the culture under the influence of the cultivation system; and that resolution of its complexity will depend upon these factors. As growth is ultimately the expression of cell metabolism, it follows that the methods for studying growth may be used for examining cellular metabolism. Thus, the metabolism of the culture, and hence the physiological state, becomes approachable in terms of the cell-in its various ordered and disordered arrays-by employing the various appropriate cultivation methods to investigate it.
RELATINGTO CELL METABOLISM D. FURTHERPROBLEMS It is pertinent at this point to consider other aspects of growth and metabolism in relation to these new- cultivation techniques, for they have important bearing on several matters related to the cell. From what has been presented above in relation to growth, it will be apparent that the results obtained from using different methods of cultivation will reflect different trends in cell metabolism, and consequently differences in ideas or interpretation of cell metabolism, as of growth, could result. Thus, it must not be forgotten that the batch culture has held a predominant part in gathering together, and forming, many of our present ideas relating to microbial metabolism. It has been noted above that in recent years the study of growth has shown some of the early discovered concepts to be misleading or inadequate, so that it is likewise possible that certain aspects currently relating to cellular metabolism and physiology are not necessarily correct and require reconsideration or reassessment too. It is a fact that much of our knowledge of cellular metabolism has been acquired haphazardly and empirically from poorly defined material of doubtful growth and origin. This information has been gleaned by in vitro methods of examination of variously extracted cell populations or tissues and, unquestioned, often provides data for predictions of precise cell function in uiuo. It is a transposition that is very tenuous and may often be stretched too far. Currently there are certain concepts and tenets, previously accepted unquestioned, that have been, or are now seen to be, in need of modification. In enzymology, for instance, the purity of crystalline enzymes is prejudiced by the recognition of isoenzymes; the generality of the reversi-
210
P. S.
S. DAWSON AND K. L. PHILLIPS
bility of enzyme actions derived from in vitro studies is challenged by their often unidirectional performance in vivo and by the increasing appearance of hysteresis effects; there are also many examples where in uitro cell mechanisms do not exhibit the rates observed in viuo functions-protein synthesis, for instance, is several orders short. Anaplerotic pathways now augment the anabolic and catabolic mechanisms of earlier years. Concepts of primary and secondary metabolism exist, and maps of various metabolic pathways. It must be remembered, however, that maps are only as good as the surveying upon which they are based. New methods would seem likely and desirable for describing in vivo cell functions. Until we have prospected cell metabolism in cellular dimensions at the cellular level it is naive and dangerous to assume that the empirical measures we presently employ are necessarily adequate or correct. What, for example, constitutes primary and secondary metabolism in a cell? Is one vital, or the other essential? How do such components contribute toward cell function under different conditions of growth rate or nutrient limitation? It is known to the authors, for example (P. S. S. Dawson and D. W. S. Westlake, unpublished observations), by release of radioactive 14C02from I4C1, 14C,:1 and I4C6labeled glucose, that the relative contributions of the Embden-Meyerhof-Parnas and pentose pathways change during the cell cycle of Candidu utilis when grown under different conditions of growth rate and nutrient limitation. Other published results have indicated other changes: for instance, how the so-called secondary metabolite ( glycogen) may be variously produced in the cell cycle under different conditions for growth (Dawson, 1970), and of changes in phosphorus metabolism (Glattli and Dawson, 1971). Molecular biologists aspire to describe, in molecular terms, how the cell performs, an endeavor closely linked to that of exploring cellular metabolism and to resolving the nature of the physiological state. For these tasks, it is essential that suitable methods and materials be used. It is most important that the seemingly uniform condition of cells growing under variously so-called “balanced growths” be recognized as having the composite variations which their uniformity conceals: randomized cells still have variations within their cell cycles (James, 1961; Dawson, 1972). A fixed growth rate does not prescribe a common performance in cells growing under different nutrient limitations, nor an invariant one ( Dawson and Gliittli, 1972). The cell is extremely responsive to its environment, so that strict control of the conditions for cultivation and production of experimental materials is mandatory in all cell studies, whether performed by the microbiologist, biochemist, cell physiologist, or biologist. The flexible operations of a cell, which are the basis of the phenotypic expression
ASPECTS OF APPLIED MICROBIOLOGY
211
of the genotype, are always ready to intrude upon the cell performance, so that it cannot be dogmatically assumed that the cell behaves in a constant, stereotyped manner. The cell machinery may appear to function smoothly under many different conditions, but it is modulated in doing so, with the modulations, which cannot be assumed or ignored, varied and adjusted to the circumstances. The extent in range of the phenotypic expressions of the genotype is not fully appreciated in microbiology-mainly because it has never been properly evaluated. For similar reasons it is also doubtful whether different genotypes have been compared under identical conditions. Systematic studies of cells cuItivated CircumspectIy by appropriate techniques could permit some rational exploration of these changes upon which the phenotypic expression depends. It will be apparent from what has gone before that the range of expression within a genotype is considerable. The changes that might be effected by change of genotype are also considerable, but these are within the realms of genetics and will not be discussed here. Needless to say, the already wide range would be extended in ways parallel to those already discussed. Thus far, only relatively simple aspects of growth that apply to unicellular microorganisms have been considered; more complicated problems, such as those relating to mycelial organisms; higher plant or animal cells, and their associated problems of differentiation, introduce further complications. Before addressing such tasks, however, further clarification of the lower, simple levels of the hierarchical development of cellular growth would appear to be necessary. Complications soon appear, even at these lowest levels; for instance, One might consider initially the three general types of growth considered by Jerusalimsky (1966): ( a ) binary fission, ( b ) the terminal extension of hyphae, ( c ) basipetal formation of conidia. In these, different relationships exist between multiplication and development, so that different expressions of physiological state are to be expected. As the degree of organization becomes more complex, as in multicellular organisms, the physiological machinery becomes more complicated and hierarchical considerations exert their influences: these intricacies are relatively unknown. Regulation within the cell is extended, SO that regulation between cells has to be considered. Our present knowledge of these matters is rudimentary, and it is not proposed to develop them speculatively here.
E. RATIONALEOF GROWTH
It is evident that, uniquely, the cultivation systems influence and thereby determine the development of a cell population, largely by controlling
212
P. S. S. DAWSON AND K. L. PHILLIPS
the environment through the manner in which the nutrient is supplied to the cells. The combinations provided by three pairs of alternative possibilities define systems for growth that permit the wide range of microbial growth to be appreciated. By growing cells as asynchronous or synchronized populations, in closed or open systems, under homogeneous or heterogeneous conditions, it is possible to demonstrate, examine, and utilize the plasticity of cell growth and metabolism quite simply on a single, chemically defined medium, or subsequently, by suitably changing that medium, to expand the range considerably. It follows, therefore, that in this way a systematic approach to exploring the interrelationship between the cell and the cell population is attainable. Systematic studies become possible for exploring cellular metabolism, and for examining the machinery of phenotypic expression in a positive manner. Direct methods can replace indirect ones, empiricism can be reduced or perhaps eventually eliminated. Thus general adoption of open and synchronous methods would permit of reproducible, rational, and more systematic progress to be achieved, instead of persisting with futile attempts to divine the intricate machinery of cell operations and performance from randomized populations always undergoing temporal change in closed systems where the inherent confusion dictates that empiricism is inevitably the only exercise possible. As pointed out earlier ( see Section II,A), the study of microbial growth by studying the cell population rather than the cell has created rather than solved its problems. Cell populations have still to be used, but the rationalization of their composition and constitution, now recognizable and realizable, permits microbiology to reorient its base to that of the unit-the microbial cell. Working from this base line, which ultimately could be absolute, permits of positive integration and projection, so that the behavior of the population becomes predictable: for with the unit performance defined, the behavior of the whole can be more easily foreseen, and known. The reverse, as we know, is not true. It would appear that microbiology has a need to take a hard look at these heretical possibilities, which could bring it into close alignment with molecular biology, for a considerable reorientation is required in outlook and application. Applied microbiology, together with other sister disciplines, will obviously be affected too. The biochemical or microbial engineers will probably realize now that they have to recognize another, the most fundamental, addition to their list of unit processes-that of the operation of the microbiaI cell. This leads to other “nonbiological” aspects of cell cultivation-physical, chemical and mathematical considerations-which perhaps should be
ASPECTS OF APPLIED MICROBIOLOGY
213
less empirical and hence more easily defined. They constitute some of the important environmental parameters for the microbe.
Ill.
Nonbiological Considerations
A. ENVIRONMENTAL FACTORS In the culture, which underlies most objectives in applied microbiol-
ogy, there are two components-the cells and their environmental milieu. In this system, there are, on the one hand, the largely biological considerations of the biochemist and microbiologist with the cell, and on the other, the nonbiological factors of the environment which now have to be considered. These nonbiological factors are mainly chemical and physical in nature and are of primary importance to the engineer who, by designing and operating his equipment, arranges, provides for, and controls the environment for the cells. It might be thought that the nonbiological parameters, lying within the realm of the exact sciences, might be easily and precisely defined, but this is not always so because the problems are usually involved and of an interdisciplinary nature. Invariably the problems, especially in relation to applied microbiology, overlap and combine the interests and skills of both the biologist and the engineer. It is not always appreciated how mutually involved these are. The engineer usually considers his problems in relation to his unit type operations (Brown, 1950), in terms of physical and chemical parameters, and the formulations and stoichiometries these involve. But in biological systems, these are all inadequate if the biological dimensions are not considered too-because ultimately all depends on the well-being of the organisms. The dominating importance of the biological influence in these matters was recently acknowledged by the late Sir Harold Hartley, the father of biochemical engineering, who had become inclined to this view after having long held the opposite one (Hartley, 1967). The physical factors, such as temperature, pressure, agitation, light, viscosity-and chemical ones, of nutrient supply and availability-may be considered separately; but they have to be considered also in conjunction with the cells, because of the changes these can undergo within the experimental systems: conditions in a culture involve cells plus environment, either, or both of which may change. As already discussed in relation to growth, the population will develop according to the cultivation system in a homogeneous or heterogeneous manner, which may or may not change with elapse of time.
214
P. S. S. DAWSON AND K. L. PHILLIPS
Besides the biological influences, there are the conventional interdisciplinary problems with which the engineer is already acquainted-problems of scale-up, materials of construction, and equipment design. But all these may require further modification too, because of the biological implications. New methods and new dimensions in cultivation techniques give novel aspects to older problems; for example, the introduction of synchrony and cell cycle developments require new rapid methods for harvesting, immobilizing, analyzing, and processing the cells. The time required to transfer materials into and out of the fermentation system must be taken into consideration; if neglected, this factor may lead to the appearance of undesirable artifacts or to canceling any advantages to be gained from the new methods. It would be pertinent to consider some of these environmental problems more fully now-especially from the standpoint of the interrelationships with the cell.
B. PHYSICAL
1. Physical State The physical state of the environment is most important to the organism: usually this is liquid or solid, but as gases can serve as nutrients, these may be significant too. Combinations of these, as suspensions, emulsions, or foams, also exert special influences; the surface area of colloids may be important for cellular activities. Certain physical processes associated or related to these conditions, such as diffusion, solubility, adsorption, absorption, have considerable effects in the environment. The textbook principles of these different entities do not concern us here, only their interactions among themselves and the cells. a. Solids
In the environment, solids have considerable influence as physical supports, as agar gels in the laboratory, beech shavings in vinegar tanks, or clinkers in filter beds. Solids have two effects in the system: (1) their surface area for growth of the organisms is usually important, and ( 2 ) they tend to encourage heterogeneity rather than homogeneity in a system. The effects of surface area are largely self-evident; but the implications of heterogeneity in a system are usually more subtle and less obvious, like the spatial zones of development in a trickling filter bed; or the gradients of nutrient diffusion and of morphological and physiological development occurring in the diff erent-sized colonies developing on an agar plate (Trinci, 1971).
ASPECTS OF APPLIED MICROBIOLOGY
215
b. Liquids The liquid environment generally facilitates mobility and promotes homogeneous conditions. As the medium is usually water, most of the important properties relate to that ubiquitous solvent. The characteristic physical properties relevant to solubility-the nature and amount of solute, flocculation, precipitation, viscosity, surface tension, and osmotic effects, besides problems of insolubility, are most significant in this context. Universally recognized as being the ideal natural solvent, there are nevertheless considerable problems due to insolubility, of suspensions, emulsions and foams when solids, liquids, and gases do not readily dissolve in water. In many cases, the cultivation system employed affects these ( see Section III,E ). Surface area relationships, important in suspensions, emulsions, and foams, have enhanced significance when colonization of the interfaces by cells takes place: such effects have importance for the production of single-cell proteins from liquid or gaseous hydrocarbons (Humphrey, 1967)) the leaching of ores (Torma et al., 1970), the unbalanced removal of cells or constituents from a culture by foam flotation-especially in harvesting chemostats by overflow methods, or the blanketing of gaseous exchange in foam-filled vessels. Fermentation media may be free-flowing Newtonian type fluids, or viscous, non-Newtonian-type liquids. The latter will present additional problems in mixing, pumping, and mass transfer operations. Donovick (1960) has suggested that for viscous beers containing large amounts of intertwined filamentous mycelia, even though the bulk mixing intensity may appear sufficient to give complete homogeneity, a concentration gradient of nutrients and waste products in the vicinity of the cell surface may be a limiting condition for mass transfer due to insufficient mixing within the mycelial mass. c. Gases
The recognition of gases as substrates, realized for a long time with oxygen and carbon dioxide, is generally more acceptable now that the possibility of single cell protein production from natural gas has been demonstrated (Coty, 1969). Problems of solubility arise and are general to those discussed under aeration ( see Section II1,D). 2. Physical Processes
In these different environmental conditions, problems of solubility, diffusion, and mixing predominatc, and involve processes of absorption, adsorption, agitation, and heat and mass transfer. Such physical processes have to be considered by themselves, in relation to the environment, and also in conjunction with the cells. There are often many difficulties
216
P. S. S. DAWSON AND K. L. PHILLIPS
involved, and sometimes the initial problem( s), or even of recognizing the true nature of the problem( s ) , is in question. In many experimental situations it is not easy to tackle the problem; the cells developing in a population (as, inevitably, they always do in batch or closed systems) change and modify the environment and thereby alter the experimental system under investigation. It is not yet appreciated in these matters how useful and imperative the employment of steady state cultures could be. The engineer could benefit from studying the system in equilibrium by ensuring that the biological system itself was in equilibrium (see Section II1,D). 3. Physical Properties a. Temperature
Temperature has a profound effect on the growth of microbes and is characterized by a limiting range between minimal and maximal extremes, with an optimal value in between: these properties may change slightly with altered conditions of environment, so that while being characteristic for an organism, they are not inviolably constant. The extremes in temperature are exploited in sterilizing media and equipment and in immobilizing the cells at harvesting, but in the intervening range of active growth a number of important conditions arise which the new cultivation methods have been instrumental in revealing. Most studies of temperature have been performed with batch cultures and have examined the change of growth rate with temperature, so that these effects have necessarily been linked to each other; the change of cell behavior at a fixed growth rate but with change of temperature, and vice versa, were not possible. The advent of continuous-flow culture in the chemostat has changed this: studies with Candida utilis by Dawson and Craig (1966) and by Brown and Rose (1969) have shown how the proportion of saturated and unsaturated fatty acids varies with growth rate changes at a fixed temperature, and with temperature changes when the organism is grown at a fixed growth rate, respectively. Another temperature effect, exploited, but not generally explained, is that of differential effects of temperature on cell growth-several manifestations exist of this effect; Scherbaum and Zuethen’s (1954) synchronization of protozoan cells by alternating temperature changes; the sharpening of synchrony in phased cultures by s. J. Pirt and co-workers (private communication) using pulsed temperature changes; and the production of oscillating conditions in chemostat steady states by relaxed control of temperature by thermostat differentials (Pa S. S. Dawson, unpublished observations) . The effect of temperature on microbial metabolism is unexplained, but the flexible use of the chemostat should facilitate exploration in
ASPECTS OF APPLIED MICROBIOLOGY
217
this area. Characteristic properties relevant to temperature effects on cells might be expected from comparative investigations of mesophiles, thermophiles, and psychrophilic organisms. The empirical importance of temperature changes in certain industrial processes, as for example the different attemperation levels in brewing and fermentation practice, could be used more predictably.
b. Pressure Pressure cannot be considered as having a direct influence as a physical factor in the environment in general practice, but indirectly it is of significance for the cells. Most organisms are not generally affected by raising or lowering the pressure in a culture, but rapid changes of pressure as in cavitation zones or in disruptive cell presses can damage or kill the cells (Hughes, 1951,1961; Edebo, 1969). In most cultures the main effect of pressure, which is an indirect one, is that of affecting the solubility of gases, especially oxygen. Much of the work dealing with the growth of microorganisms under hyperbaric pressure is more concerned with the effect of increased oxygen tension on the growth rate rather than of higher pressure per se. Fermentations operated at increased air pressure can maintain the mass flow rate of air and at the same time decrease the volumetric flow through the fermentor so that tendency for foaming is decreased (Phillips, 1969). In addition to hydrostatic effects of pressure are those due to osmotic pressure. Most organisms can adjust to moderate changes in the concentration of solutes in the environment, but some extremes are notable; halophiles, for example, require high concentrations of salt. Plant and animal cells are more susceptible to osmotic change than are microbes, but the latter can be affected too. The system of cultivation, whether closed or open, can also exert some influence; the chemostat has the advantage of being able to operate under constant conditions at all times, although the conditions may be changed as desired; but closed systems may change considerably and often uncontrollably. c.
Light
Light is mainly of consideration for photosynthetic organisms, where intensity and wavelength are important. The periodicity of the illumination is sometimes critical and has been used for obtaining synchronization (Tamiya et al., 1953) and perturbation effects (Tamiya, 1964). 4. Physical-Chemical
There are a number of physical-chemical factors, such as pH and Eh, that are susceptible to different interpretations according to the manner and direction of their application. pH can be defined in a number
218
P. S. S. DAWSON AND K. L. PHILLIPS
of ways (Ogston, 1947), and its effects are usually considered directionally. In the environment, pH and Eh have to be considered as empirical parameters and used accordingly: the precise interpretation of their effects is often doubtful for pH, and usually meaningless for Eh. Technically the measurement can be difficult, but recent developments have produced much improved equipment ( Munro, 1970; Jacob, 1970)-this is useless if not properly employed. Homogeneity is essential, otherwise location of probes may invalidate or alter the results. The system of cultivation will require consideration-the chemostat again has advantage for control. Related to pH and electrochemical considerations in the environment are the amphoteric properties and electrostatic changes that affect flocculation and influence cell distributions; these may be affected by cell metabolism too. EIectricaI phenomena are becoming increasingly significant in the development of specific probes, in physiological studies (Allen, 1972), for determining various ions ( Hinke, 1969), dissolved oxygen ( Beechey and Ribbons, 1972), utilizing fixed enzymes in assay procedures (Guilbault, 1972) and other measurements.
C. CHEMICAL
As an active factor in the environment, a chemical may be nutritive, inhibitory, or toxic in its effect on the cells. This effect may alter depending upon the concentration of the chemical compound; a substance may be nutritive in low concentrations but inhibitory at higher values. For growth of the organisms a number of nutrients are required and, as discussed earlier (see Section II,A,l), this growth may be controlled and limited by the concentration of these constituents. Different nutrients have different effects, but it is outside the scope of this contribution to consider these at length; however, a number of points must be mentioned. A defined medium is mandatory for meaningful work and should be used whenever possible; no one has established this point better than Herbert (1961a). It is not always necessary to use defined media for all purposes: in applied work, for example, it is usual, economically expedient, and often justified, to use undefined substrates. The method by which the nutrients are supplied to the cells is crucial for directing their development; as outlined earlier, the addition may be initially complete, intermittent, or continuous, giving batch, synchronous or continuous cultures, respectively. Precursors may also be added in like fashion at any time, but it is best to ascertain how this may be done most advantageously. It is point-
ASPECTS OF APPLIED MICROBIOLOGY
219
less to add precursors in excess of the amount required or at an inappropriate time, since in all probability the precursor could be used as a superfluous energy source. Certain nutrients have specific effects, others have general influences: some serve primarily as energy sources, some as building blocks, and others for both functions in cellular metabolism. The ratio of different nutrient components supplied to the cells can be important too; for example, the ratio of carbon to nitrogen for influencing assimilating processes, or the supply of oxygen for aerobic or microaerophilic activities: critical concentrations might serve as switch points in cell metabolism ( Dawson, 1972). Most important, especially in continuous cultures, is the need for homogeneous conditions-adequate mixing is essential. In empirical practice, the errors of commission in this section are often profound yet unrecognized.
D. INTERPLAY OF FACTORS In the above sections a number of different environmental factors have been mentioned independently, but rarely is this experienced in practice; usually the factors are markedly influenced by each other pH greatly influences temperature effects, for example, and pH is profoundly affected by chemical environmental factors. It is therefore dficult to define optimal conditions unless rigid specification of the environmental circumstances is made. Nevertheless it is important to know how the factors behave, both individually, and in conjunction with each other. Aeration is perhaps one of the most important in this respect and for this reason may be considered a little more fully now.
Aeration The study of aeration is an excellent example of how complex the evaluation, control, and arrangement of some physical parameters may be. For many organisms oxygen is an essential requirement for growth, and in submerged culture the organisms have to obtain their needs from the dissolved oxygen in the nutrient medium. The solubility of oxygen from air in water is low, only around 7 ppm at 3OoC at atmospheric pressure; its solubility is adversely affected by other solutes. It is also affected by temperature and pressure. Technically, there is the problem of getting oxygen into solution. Growing organisms quickly use up this dissolved oxygen, so that a dissolved oxygen content has to be maintained to meet the organisms requirements. The input of oxygen must be at least equal to, and preferably greater than, the output.
220
P.
S. S. DAWSON AND K. L. PHILLIPS
To obtain solution of the oxygen, the medium must be mixed and agitated to provide intimate contact of the air or oxygen with the liquid; either by stirring and sparging, or by film formation and rapid renewal of surfaces. These operations need to be evaluated, methods for oxygen uptake must account for physical, chemical, and biological considerations. Methods used to evaluate and specify oxygen supply or demand often disagree. For example, sulfite oxidation results may or may not be directly applicable to establish the quantitative transfer of oxygen to be expected for biological growth ( Finn, 1954). Aeration may be defined in either of two ways. The first method requires a knowledge of the quantitative requirements of the cell population, and is defined as the amount of oxygen required by unit weight of cells per unit time, or the amount of oxygen required by unit volume of fermentation medium per unit time. The second method maintains a predetermined concentration of dissolved oxygen, usually above that considered critical for normal metabolism of the cells, throughout the complete fermentation cycle. The latter method is more applicable to continuous fermentation systems. Cell requirements vary during their growth and development; this is plainly evident in batch culture. But during such an investigation the cell population and the environment undergo change too, so that cause and effect are confused. The chemostat however, with its steady state, affords a stable test bed for observing oxygen utilization rates under equilibrium conditions that can be systematically examined ( Pirt, 1957; Harrison and Pirt, 1967; Herbert et al., 1965; Tempest et al., 1967). Dissolved oxygen content of a culture varies, but probes are now technically developed to a point where they can be considered reliable-however, changes in dissolved oxygen need to be linked to cell metabolism, because this changes too under different oxygen tensions (Harrison, 1972). It is evident that the main problem is to be able to stabilize and maintain all experimental parameters constant except for those directly under observation. In aeration studies the needs are (1) to estimate the oxygen parameters, ( 2 ) to stabilize experimental conditions, ( 3 ) to evaluate the specific oxygen requirements, and ( 4 ) to identify the mechanism and usage. Oxygen uptake rates can be evaluated either by calculating the material balance based on the composition of the known flow rate of gas entering and leaving the fermentation system, or by measuring the amount of dissoIved oxygen in the medium. The latter system will define whether an excess of oxygen is available or whether the fermentation is limited by oxygen supply, but unless the oxygen-limiting characteristics
ASPECTS OF APPLIED MICROBIOLOGY
221
of the fermentor are known, will give no quantitative measure of the amount of oxygen being used. It is possible to stabilize growth in continuous culture at different levels of dissolved oxygen by manipulating ( 1 ) nutrition, ( 2 ) growth rate, ( 3 ) agitation, ( 4 ) temperature, or ( 5 ) oxygen partial pressure. Batch cultures, as already mentioned, give an indication of change in the physiology and morphology of the cells with changing dissolved oxygen concentration. Use of the chemostat to obtain steady growth states reveals this effect more directly (Harrison and Pirt, 1967). When cells were grown using the phased culture technique, oxygen requirements were shown to vary during the cell cycle, and also were dependent on the type of growth obtained (Miiller and Dawson, 1968; Dawson, 1971). The design of aerobic fermentation systems, whether for batch, continuous, or phased operation, must ensure an adequate supply of oxygen at all times. Interruption of this supply, even for a short period of time, can greatly alter the physiology of the cells. The critical need for maintaining a continuing supply of oxygen to the microorganisms has been dramatically described by Finn (1967), who stated that during active respiration in an aerobic fermentation “even if the broth could be saturated with air, it would contain only a 15-second supply of oxygen for the culture; it is as if the crowded occupants of a well-ventilated room were at all times just four deep breaths away from suffocation.” In this respect, also, mixing of the culture medium should be intense enough that a homogeneous system is obtained with regard to dissolved oxygen; otherwise, gradients in oxygen tension may be present, even to the extent that in some parts of large fermentation vessels, the organisms may be subject to anaerobic conditions.
E. EQUIPMENT-DESIGN AND OPERATION For the engineer, the culture vessel must initially be the central focus of his concern with his tasks in microbiology. It is with this device that he provides and controls the environment for his biological agents. Second, he is concerned about the type of system in which the vessel is used, for this largely decides its purpose. The scope of his problems is wide and their number countless; details do not concern us here, but rather certain important basic approaches to problems of geometry, of system, their scope and size, and of matters associated with these.
1. Geometry In design, endless variations of the culture vessel exist ranging from the bulk tank, on the one hand, to the thin trickling film through a
222
P. S. S. DAWSON AND K. L. PHILLIPS
filter bed on the other. In principle, however, there are limits, and Herbert’s classification ( 1961b) conveniently summarizes the general possibilities. It is generally preferable, however, to have a versatile, adaptable equipment giving a homogeneous performance and universal application, and for some time the conventional stirred agitated tank has sufficed, rendering reasonable service. The model is widely used and serves many purposes, but no design is perfect, and it does have two deficiencies; it promotes (1) wall growth or surface accretions, and ( 2 ) foam production. If care is not shown in its design and operation, imperfect mixing and unrepresentative sampling also intrude. These imperfections are detrimental for continuous and synchronization techniques. Alternative designs like those of rotating drums (Phillips et aZ., 1961) and cyclone columns (Dawson, 1963; Dawson et al., 1971) have been suggested to circumvent these deficiencies. Ideally, a culture vessel should have the following characteristics: ( 1 ) be of simple design, inexpensive to build, economic to operate; ( 2) provide homogeneous conditions-temperature, aeration, mixing, light, etc.; ( 3 ) provide for batch, continuous, or synchronous growth; (4)be adaptable to unicellular or mycelial organisms; ( 5 ) operate under aerobic or microaerophilic, or strictly anaerobic conditions; ( 6 ) should be adaptable to multistage operations; ( 7 ) be reliable in use and capable of long-term operation; ( 8 ) permit control of experimental parameters, probes, etc.; ( 9 ) provide representative sampling/harvesting; ( 10) avoid foam production, accretions, and wall growth.
2. System As discussed earlier, whether the cultivation system is open or closed, single or multistage, homogeneous or heterogeneous, decides the mode of development and the nature of the results obtained from a system. The engineer, or any other worker, should be able to judge what system is being used or is required for a specific purpose. Too little attention is paid to these matters, perhaps because of equipment already installed and immediately available to the investigator. It might be better to take the time and trouble to design and use the most suitable system for the desired purpose.
3. Scope The magnitude or potential of a new project is often problematical, and when considered in new dimensions is not readily appreciated at first contact. Newcomers to continuous culture, for example, are often surprised by the medium-making capacity required to maintain even a small chemostat in operation.
ASPECTS OF APPLIED MICROBIOLOGY
223
It is important to recognize the size of the apparatus required, and in chemostat work for experimental purposes, this often needs only to be large enough to provide samples for analysis. However, for production purposes scale-up is necessary, and those problems well known to the engineer intrude. It is essential in such matters that the geometrical ratios be considered-that the balance of area to volume, for example, be maintained-remembering that increase in area is a function of the square, while the volume increases as the cube of the linear dimensions. Unfortunately such good intentions are not easily rewarded; maintaining geometrical ratios will not necessarily balance accompanying changes in the dimensionless variables. Allowing for the increase in radius of an impeller, to maintain the same tank to impeller diameter ratio at a fixed stirring speed will unbalance the system with respect to shear, since the tip speed is increased porportionally; shear effects which are harmless in a small fermenter become damaging in a large one. Under certain circumstances, cavitation phenomena may develop. There is a point where scale-up may be self-defeating: excessively large-batch fermentations may require too much time for charging the batch, sterilizing, excessive holding time at the conclusion of the fermentation before processing of the product can be completed, as well as the limitations imposed on the fermentation system which might preclude adequate mixing and aeration. There is much to be said for a small properly des’igned system balancing its productivity goal by dimensional means, as, for example, the substitution of a closed system by an open one. Instead of a massive batch culture, which is actively functional only during its peak period, a much smaller, efficiently designed continuous culture, operating continually on stream at an optimal production might advantageously be substituted. In these matters, the reassessment of certain operational parameters might be advisable-of dissolved oxygen concentration, for example, instead of power input per unit volume, as the arbiter in controlling the fermentation. Besides such matters it is also important to consider the experimental facilities of the batch, continuous and synchronous systems and of particular problems reIated to them. For example, wall growth and foam enrichment effects do not intrude on most batch-culture operations-but they can seriously affect the steady states of chemostats or the synchrony of phased cultures. It must not be forgotten that heterogeneity can arise within a homogeneous system (within the pellets growing in a stirred tank) or homogeneity within a heterogeneous system (zones in a pipe flow system).
2.24
P. S. S. DAWSON AND X. L. PHILLIPS
The impact of new methods and techniques shows a greater degree of control and sophistication in the equipment and methods as one progresses from batch to continuous to synchrony for example. In application, the same philosophy applies; the empiricism of the bulk solvent fermentations gives way to the production of specific metabolites in new applications. In the development of equipment and techniques an increasing interest in cellular dimensions is developing-probes for enzyme assay, or new processes for immobilizing, harvesting, fractionating, and processing of cells or cell components. In conclusion, it would appear that when investigating the nonbiological or environmental factors, the biological effects should be neutralized or eliminated from the system as much as possible. The open system of the chemostat would appear to be the simplest and most satisfactory way of stabilizing these interferences and observing the physical and environmental parameters, which may be examined singly, or in suitable combinations, possibly employing the EVOP method to do this most economically ( Barnett, 1960). IV.
Mathematical Considerations
It will be appreciated from the foregoing observations on biological and iionbiological aspects of microbial growth that a cursory approach to its problems, or ignorance of the fundamentals, although frequently proving successful, might equally be disappointing for attaining a desired result. The empirical use of microorganisms is unpredictable in practice: it sometimes leads to erratic performance of batch cultures; or in other circumstances, requires palliative statistical treatments to avoid or explain capricious variations and shortcomings often attributed to “biological error.” After considering the biological and nonbiological implications of growth, it might be pertinent now to consider the relationship of mathematics to microbial growth and its applications. The association between mathematics and studies of microbial growth is closely linked to the former inability of the microbiologist, without his microscope, to examine the microbe as an individual, and of his consequent expedient of having to use the population in its stead. This inconvenience produced complications, and mathematics has been used to help resolve, or correct, the anomalies produced. From the start of growth studies, the mathematical description of growth has been a continuing exercise in microbiology that has progressed from one formulation to another, as model has succeeded model. But mathematical models, as Tsuchiya (1970) has pointed out, are representations made up of equations; precise mathematical statements
ASPECTS OF APPLIED MICROBIOLOGY.
225
that define exactly. The preciseness, however, depends upon the model and the complexity or simplicity of the object decides how complicated this model, and its mathematics, might be. To the average microbiologist this is very disconcerting, but it is important nevertheless that a sense of the mathematical proportions and perspectives be appreciated. Thus, initially, we observe that the early workers, seeing growth in terms of a simple increment of mass or numbers in a population, attempted to seek a simple mathematical equation which might give a quantitative description of the growth curve. With the development of microbiology, however, it became evident that no unique curve existed: instead, variations accumulated and reflected the many new experimental advances occurring in growth studies. Thus, with progress, the simple models proceeded through modifications and changes until the reasonably successful general formulation of Monod ( 1942) appeared. Like all modeIs, however, this could only be as good as its limitations allowed; so that the formulation which described and predicted the mode of exponential growth in a batch culture did not distinguish between the different changing phases of growth. Hence, when the formulation was used subsequently as the basis for developing the theory of continuous culture (Monod, 1950), the new model was limited to the cells developing in a constant manner, so that aspects of physiological change in the cells were excluded, and the model therefore excluded physiological change. Work with continuous ( exponential) growth in chemostat cultures gave practical emphasis to the limitations of the Monod formulation-and it was evident that the simple mathematics of the Monod theory needed to acquire more ample dimensions. Thus Powell (1967) attempted to make adjustments for physiological change and give both quantitative and qualitative expression to the basic formulation. But such improvements can progress only as far as the physiological changes can be defined in mathematical terms. This involves or invokes a mathematical description of the physiological state: a problem still unresolved. Various models for growth can be obtained by recapitulating in a mathematical form many of the different variations possible in growth dimensions that have already been outlined earlier in relation to cell populations; of their homogeneity or heterogeneity, and of the randomness or synchrony of their cells. Indeed, a considerable array of such models is to be found in the literature: models, for example, relating to the distribution of cell size and cell age in a population, and of the changes that these undergo during the further development of the culture. However, these different models would appear to have a very limited usefulness-if only because simple models have extremely restricted applications and complex models are self-defeating because the
226
P. S. S. DAWSON AND K. L. PHILLIPS
necessary information to ensure their validity is lacking. Until the physiological state can be defined in realistic terms, it would seem that mathematical modeling for growth is essentially an exercise in mathematical manipulation. This is not to say that mathematical models and theories are useless, but rather that they have to be considered within their correct perspective. We have seen that growth entails the cell, its metabolism and physiology, and its development as a population; all of which are influenced by the environment and its changes. To be able to define such a system mathematically is not a task within the bounds of reality at present; but to approach the problems conceptually, using mathematical principles to do so, is reasonable and attainable. Consequently, it is not necessary at this point, to develop precise models, because our knowledge is inadequate for this, but rather to make use of mathematical concepts in attempts to delineate and outline the mechanism for growth. The thesis of Dean and Hinshelwood (1963, 1966) exemplifies this approach. In their view, the modulations of the cellular machinery are integrated and linked to the environment and thereby form the basis of growth in the cell-a cyclic network of integrated mechanisms continually responding and adjusting in a self-regulating development. In this conceptual model, the physiological state is coordinated with growth and, as conditions change, adjusts to accommodate the change; so that it reflects the operations of phenotypic change. In contrast to this approach, there is still, on the cellular base, the dogma of the molecular biologist and the more rigid mechanical expression of the genotype in its functions. Such theoretical considerations between mathematics and cell function lie beyond the present state of experimental endeavor, but interest in this direction is encouraged by the efforts of Hansche (1969), Goodwin (1963, 1969), and others. Of more practical importance, in the close association of mathematics with cell growth and function, is the growing interest in possibilities for the computer control of microbial processes (Nyiri, 1971; Humphrey, 1971). The scope of this subject is beyond the present paper, but in the context of the latter, one or two points may be mentioned here. Initially the computer can perform only within the limits imposed by its program-the mathematical model that controls its actions. The object of the model is therefore of supreme importance, and it is essential that it should be clearly identified and defined. In applied microbiology or fermentation practice certain possibilities and limitations occur: but initially programs are likely to be restricted to very compact areas for the obvious reason that our present lack of knowledge cannot avoid
ASPECTS OF APPLIED MICROBIOLOGY
227
it. Physical parameters and equipment performance must necessarily predominate at first, as these are much simpler than the biological areas. It will be apparent, for instance, that whereas the chemical engineer can program quite definitely on the stoichiometry of a chemical equation and physical parameters of the reactants, the biochemical engineer will have to rely on the mathematical concepts of microbial growth and the variability of the physiological state. Therefore, for processes involving biological parameters the controls will be largely empirical-because of our present inadequate knowledge-so that absolute control cannot be entertained for sometime yet. However, certain points relevant to our previous discussions begin to appear: the suitability of “open” systems rather than “closed systems for obtaining stabilized conditions and easy control; the need to distinguish between average mean conditions and individual cell characteristics for the finer aspects of cellular control; the complex interplay of variables in changing environments; the need to distinguish between “biological” and “nonbiological” factors-always bearing in mind the possible auxiliary presence of biological influences under nonbiological conditions.
V.
General Considerations
As the variety and range of microbial activities is virtually limitless, no brief selection can be representative, so that it is more practical to consider the various applications generally and in relation to their usage. Long before man was aware of the existence of microbes, they were being used to improve or destroy his amenities and livelihood; enriching the soil, destroying his crops, fermenting his drinks, spoiling his food, removing his waste, affecting his health and general well-being. Over the millenia such processes continued relatively undisturbed until about a century ago, when man began to be cognizant of these activities. Then, by empirical procedures and an increasing inquisitiveness, he began to accumulate some knowledge of these matters-so that today three broad areas of development can be recognized: empirical practices, basic knowledge, and the combination of the two-applied microbiology, with which we are concerned. The developments of the last hundred years are sufficient testimony to the impact of basic knowledge on empirical practice: of the importance of fundamental research for successful practical applications. The development of new cultivation techniques during the past two decades, for example, stems from a seemingly academic drudgery-a study of growth in the batch culture; or, to think of another, the discovery of
228
P. S.
S. DAWSON AND
K. L. PHILLIPS
penicillin, as an example that amply demonstrates the essential requirement of basic knowledge for successful application. There will always be the dissenting voice calling attention to the old, well-tried methods-but even beer frequently comes from continuous fermentations these days. It is therefore with the application and use of basic knowledge, and of continually improving upon it, that applied microbiology can develop itself quite rationally, rather than by resorting to empirical “try it and hope” expectations. The tactics, as we have seen, are reasonably straightforward and are available. However, some developments in future might require changes in strategy, not necessarily of a microbiological choosing. Applied microbiology has in the past, like microbiology, been compartmented in its development: medical, veterinary, industrial, agricultural-each largely segregated from the rest and usually characterized by a particular outlook or hierarchical control. Today, the position is changing; barriers are eroding and recognition of comparative similarities and differences tend to integrate approaches and problems. New orientations appear-economic, social, political, and natural; alien confines with which applied microbiology will have to be related in future, for it is likely that projects, besides producing a desired product, will need to fit into the other general schemes of things. Industrial processes, besides being sound technically and economically, will have to be accommodated socially and politically, so that ultimately a natural harmonious balance is attained. Often in the past, vested interests have allowed narrow ends to justify the means-attainment of product has been expediently, if not efficiently, obtained. Such policies have had adverse side effects of which the overexploitation of natural resources, whether by denudation or pollution, is one that confronts us daily. In former times economic considerations solely applied as external controls on a process, but now social and political factors are increasingly becoming involved. The production of single-cell protein from hydrocarbons, for example, would have been considered purely in terms of technical feasibility and economics until quite recently. Now it becomes of increasing social importance in providing food for the hungry of the world, and probably of political significance too, because of the concern about whether irreplaceable fossil fuel reserves should be squandered in irreversible procedures rather than be sustained by cyclic operations based upon the products of anaerobic digestion processes that can be maintained naturally. In another aspect of single-cell protein production, which can be engineered in various ways, questions arise as to whether large factories operating under the complex and highly sophisticated methods and dis-
ASPECTS OF APPLIED MICROBIOLOGY
229
tribution channels of highly developed countries should be used, instead of thousands of primitively run village kitchen operations at the source of the raw materials and the place required of product consumption in underdeveloped lands ( Heden, 1969; Langlykke, 1969). Such problems reflect the wide diversity of the scope and range of applied microbiology-but the critical factor in each problem is the same-what are the significant parameters involved? In the past, this answer had to be met in an empirical way, but today this is not always necessary, and tomorrow, perhaps not at all, if we are diligent and do our work properly. It is folly to expect that things will turn out all right in the end: as Herbert (1961a) has pointed out, meaningful work cannot be expected from “placing some medium in a glass container, inoculating it with microorganisms and allowing events to run their course,” even if this dodge is useful sometimes. In our deliberations so far, it will be apparent that we have considered the scope and problems of microbiology and applied microbiology in relation to the activities of pure cultures, and largely biased in the direction of what may loosely be described as that concerned with fermentation practice. There are other important areas too: medical, agricultural, veterinary-many of these begin to move into still more complicated dimensions; of interactions of different populations, and between organisms and host environments, areas which are still largely undeveloped and unexplored. We do not include these here because space does not allow them to be discussed at length-but they are of extreme importance for the applications of microbiology in natural environments; with problems of host-parasite, symbiotic, and ecological relationships. To appreciate some of the problems lying in these areas it will be necessary to understand the operations of cell mechanisms more closely, of how the cell behaves on its own and how it is affected by others. How, for example, does the hierarchical ordering of organisms in the rumen function, or what metabolic shift( s ) cause( s ) the cutoff in nitrification mechanisms? Answers to such questions might permit further dimensions of applied microbiology to be added at the subcellular level. Synchronization, for example, could permit the cell cycle mechanisms to be investigated in a manner analogous to that of the engineer as he examines his Carnot cycles on the test bench: by disassembling the cells, maybe at temporally significant points, into fractions, or possibly of organelles, that, having special activities, might enable them to be used in parts for reassembling into biochemical machines-perhaps as specifically ordered arrays of immobilized enzymes, for instance. These are some of the perspectives for applied microbiology, which has at its heart the cell and its activities. Such cells and their activities may be used empirically in bulk and in an unordered fashion to produce
230
P. S. 6. DAWSON AND K. L. PHILLIPS
useful materials quite simply; or alternatively perhaps-in ordered, and ingeniously sophisticated, regimentations to carefully provide us with valuable, precious compounds or some of the secrets of life processes. REFERENCES Abbo, F. E., and Pardee, A. B. ( 1960). Biochim. Biophys. Acta 39,478. Allen, M. J. (1972). In “Methods in Microbiology” ( J . R. Norris and D. W. Ribbons, eds.), Vol. 6B, pp. 247-283. Academic Press, New York. Barnett, E. H. (1960). Ind. Eng. Chem. 52, 500. Beechey, R. B., and Ribbons, D. W. (1972). In “Methods in Microbiology” ( J . R. Norris and D. W. Ribbons, eds.), Vol. 6B, pp. 25-53. Academic Press, New York. Brown, C. M., and Rose, A. H. (1969). J. Bacteriol. 99, 371. Brown, G. G. (1950). “Unit Operations.” Wiley, New York. Bryson, V. (1952). Science 116, 48. Buchanan, R. E. (1918). J. Infec. Dis. 23, 109. Campbell, A. (1957). Bacteriol. Reo. 21, 263. Coty, V. F. (1969). Biotechnol. Bioeng. Symp. 1, 105. Dawson, P. S. S. (1963). Can. I. Microbiol. 9,671. Dawson, P. S. S. ( 1965). Can J. Microbiol. 11, 893. Dawson, P. S. S. ( 1970). Can. J. Microbiol. 16, 783. Dawson, P. S. S. (1971). Biotechnol. Bioeng. 13, 877. Dawson, P. S. S. (1972). J. Appl. Chem. BiotechnoL 22, 79. Dawson, P. S. S., and Craig, B. M. (1966). Can. J . Microbiol. 12, 775. Dawson, P. S. S., Anderson, M., and York, A. E. (1971). Biotechnol. Bioeng. 13, 865. Dean, A. C. R., and Hinshelwood, Sir C. (1963). Nature (London) 199, 7. Dean, A. C. R., and Hinshelwood, Sir C. (1966). “Growth, Function, and Regulation in Bacterial Cells,” pp. xvi & 439. Oxford Univ. Press, (Clarendon), London and New York. Donovick, R. (1960). Appl. Microbiol. 8, 117. Edebo, L. ( 1969). In “Fermentation Advances” (D. Perlman, ed.), pp. 249-271. Academic Press, New York. Finn, R. K. (1954). Bacteriol. Rev. 18, 254. Finn, R. K. ( 1967). In “Biochemical and Biological Engineering Science” ( N . Blakebrough, ed.), Vol. 1, Chapter 4, p. 70. Academic Press, New York. Glattli, H., and Dawson, P. S. S. (1971). Can. J. Microbiol. 17, 339. Goodwin, B. C. ( 1963). “Temporal Organization in Cells.” Academic Press, New York. Goodwin, B. C. ( 1969). Eur. J. Biochem. 10,511. Guilbault, G. G. (1972). Biotechnol. Bioeng. Symp. 3, 361. Hansche, P. E. ( 1969). J. Theor. Biol. 24, 335. Harrison, D. E. F. (1972). 1. Appl. Chem. Biotechnol. 22, 417. Harrison, D. E. F., and Pirt, S. J. (1967). J. Gen. Microbiol. 46, 193. Hartley, Sir H. (1967). Process Biochem. 2, 3. Hattori, R., Hattori, T., and Furusaka, C. (1972). J. Gen. Appl. Microbiol. 18, 271. Heden, C. G. (1969). In “Fermentation Advances” ( D . Perlman, ed.), pp. 861-882. Academic Press, New York.
ASPECTS OF APPLIED MICROBIOLOGY
231
Helmstetter, C. E., and Cummings, D. J. (1963). Proc. Nut. Acad. Sci. U.S.50, 767. Henrici, A. T. (1928). “Morphologic Variation and Rate of Growth of Bacteria.” Thomas, Springfield, Illinois. Herbert, D. ( 1959). In “Recent Progress in Microbiology” (G. Tunevall, ed.), pp. 381-396. Almqvist I% Wiksell, Stockholm. Herbert, D. ( 1961a). Symp. Soc. Gen. Microbiol. 11, 393. Herbert, D. (1961b). SCI ( SOC. Chem. Ind., London), Monogr. 12,21. Herbert, D. ( 1962). In “Continuous Cultivation of Microorganisms” (I. Malek, K. Beran, and J. Hospodka, eds.), pp. 23-44. Publ. House Czech. Acad. Sci., Prague. Herbert, D., Elsworth, R., and Telling, R. C. (1956). J. Gen. Microbiol. 14, 601. Herbert, D., Phipps, P. J., and Tempest, D. W. (1965). Lab. Practice 14, 1150. Hinke, J. A. M. (1969). In “Glass Microelectrodes” ( M . Lavallee, ed.), pp. 349-375. Wiley, New York. Hroncek, J. ( 1962). In “Continuous Cultivation of Microorganisms” (I. Malek, K. Beran, and J. Hospodka, eds.), pp. 73-82. Publ. House Czech. Acad. Sci., Prague. Hughes, L. E. ( 1951). Brit. J. E x p . Pathol. 32, 97. Hughes, L. E. ( 1961). J. Biochem. Microbiol. Technol. Eng. 3, 405. Humphrey, A. E. (1967). Biotechnol. Bioeng. 9, 3. Humphrey, A. E. ( 1971). In “Computer Control of Fermentation Processes” (Proceedings, Labex International Symposium, Earls Court, London), pp. 1-15. Fermentation Design Inc., Bethlehem, Pennsylvania. Jacob, H. E. (1970). In “Methods in Microbiology” (J. R. Norris and D. W. Ribbons, eds.), Vol. 2, pp. 91-123. Academic Press, New York. James, T. W. (1961). Annu. Reu. Microbiol. 15, Z7. James, T. W. ( 1966). In “Cell Synchrony” ( I . L. Cameron and G. M. Padilla, eds.), pp. 1-13. Academic Press, New York. Jerusalimsky, N. D. ( 1966 ), “Fundamentals of the Physiology of Microorganisms,” pp. 135-136. Academy of Sciences, Institute of Microbiology, USSR ( Israel translation into English, Jerusalem, 1966). Kjeldgaard, N. O., Maalge, O., and Schaecter, M. (1958). J . Gen. Microbiol. 19, 607. Kozesnik, J. ( 1962). In “Continuous Cultivation of Microorganisms” (I. Malek, K. Beran, and J. Hospodka, eds.), pp. 59-68. Publ. House Czech. Acad. Sci., Prague. Langlykke, A. F. ( 1969). In “Fermentation Advances” (D. Perlman, ed.), pp. 883-893. Academic Press, New York. Maalge, 0. (1962). In “The Bacteria” (I. C. Gunsalus and R. Y. Stanier, eds.), Vol. 4, pp. 1-32. Academic Press, New York. McClung, L. S. (1949). Annu. Reo. Microbiol. 3, 395. Malek, I. ( 1958). In “Continuous cultivation of Microorganisms” ( Proceedings of a Symposium), pp. 11-28. Publ. House Czech. Acad. Sci., Prague. Malek, I. ( 1967). In “Microbial Physiology and Continuous Culture” (E. 0. Powell et al., eds.), pp. 1-10. H.M. Stationery Office, London. Malek, I. ( 1969). In “Continuous Cultivation of Microorganisms” ( I . Malek et d.,eds.). pp. 23-45. Academic Press, New York. Maruyama, Y., and Yanagita, T. (1956). J . Bacteriol. 71, 542. Mitchison, J. M., and Vincent, W. S. (1965). Nature (London) 205, 987. Monod, J. ( 1942). “Recherches sur la croissance des cultures bacte‘riennes.” Hermann, Paris.
232
P. S. S. DAWSON AND K. L. PHILLIPS
Monod, J. ( 1949). Annu. Reu. Microbiol. 3, 371. Monod, J. ( 1950). Ann. Inst. Pasteur, Paris 79,390. Miiller, J., and Dawson, P. S. S. (1968). Can. J. Microbiol. 14, 1127. Munro, A. L. S. (1970). In “Methods in Microbiology” (J. R. Norris and D. W. Ribbons, eds.), Vol. 2, pp. 38-89. Academic Press, New York. Novick, A., and Szilard, L. (1950). Science 112, 715. Nyiri, L. K. ( 1971 ). In “Computer Control of Fermentation Processes” ( Proceedings, Labex International Symposium, Earls Court, London), pp. 16-26. Fermentation Design Inc., Bethlehem, Pennsylvania. Ogston, A. G. ( 1947). Physiol. Rev. 27, 228. Painter, P. R., and Marr, A. G. ( 1968). Annu. Reu. Microbiol. 22, 519. Phillips, K. L. ( 1969). In “Fermentation Advances” (D. Perlman, ed.), pp. 465-490. Academic Press, New York. Phillips, K. L., Sallans, H. R., and Spencer, J. F. T. (1961). Ind. Eng. Chem. 53, 749. Pirt, S. J. (1957). J. Gen. MicrobioZ. 16, 59. Powell, E. 0. (1967). In “Microbial Physiology and Continuous Culture” (E. 0. Powell et al., eds.), pp. 34-56. H.M. Stationery Office, London. Powell, E. 0. ( 1969). In “Continuous Cultivation of Microorganisms” ( I. Malek et ul., eds.), pp. 275-282. Academic Press, New York. Powell, E. 0. (1972). 1. Appl. Chem. Biotechnol. 22, 71. Powell, E. O., and Lowe, J. R. (1962). In “Continuous Cultivation of Microorganisms” (I. Malek, K. Beran, and J. Hospodka, eds.), pp. 45-53. Publ. House Czech. Acad. Sci., Prague. Ricica, J. ( 1958). In “Continuous Cultivation of Microorganisms” (Proceedings of a Symposium), pp. 75-105. Publ. House Czech. Acad. Sci., Prague. Scherbaum, O., and Zeuthen, E. (1954). Exp. Cell Res. 6, 221. Sherman, J. M., and Albus, W. R. (1923). J . Bacteriol. 8, 127. Tamiya, H. (1964). In “Synchrony in Cell Division and Growth” (E. Zeuthen, ed. ), pp. 247-305. Wiley (Interscience), New York. Tamiya, H., Iwamura, T., Shibata, K., Hase, E., and Nihei, T. (1953). Biochim. Biophys. Acta 12, 23. Tempest, D. W. (1970). In “Advances in Microbial Physiology” (A. H. Rose and J. F. Wilkinson, eds.), Vol. 4, pp. 223-250. Academic Press, New York. Tempest, D. W., Herbert, D., and Phipps, P. J. (1967). In “Microbial Physiology and Continuous Culture” ( E . 0. Powell et al., eds.), pp. 240-253. H.M. Stationery Office, London. Torma, A. E., Walden, C. C., and Branion, R. M. R. (1970). Biotechnol. Bioeng. 12, 501. Trinci, A. P. J. ( 1971). 1. Gen. Microbiol. 67, 325. Tsuchiya, H. M. (1970). Biotechnol. Bioeng. 12, 645. Van Niel, C. B. (1949). In “The Chemistry and Physiology of Growth” (A. K. Parpart, ed. ), pp. 91-105. Princeton University Press, Princeton, New Jersey. Whalley, M. E. ( 1955). Nut. Res. Counc. Can., T.I.S. Rep. 45. Winslow, C. E. A., and Walker, H. H. (1939). Bacteriol. Reu. 3, 147.
Some Thoughts on the Microbiological Aspects of Brewing and Other Industries Utilizing Yeast
G. G . STEWART Beverage Science Department, Labatt Breweries of Canada Ltd., London, Ontario, Canada
I. 11. 111. IV. V. VI. VII. VIII. IX. X. XI. XII. XIII. XIV.
Introduction .................. Historical Survey of the Industrial Uses of Yeast Present-Day Industrial Uses of Yeast ........ Outline of Traditional Brewing Processes ...... New Developments in Batch Processes for Brewing .. Continuous Fermentation in Brewing ........ Comparison of the Available Modem Fermenting Systems in Brewing ................. Selection, Behavior, and Efficiency of Yeasts ..... Baker’s Yeast .................. The Fed-Batch Process .............. Continuous Processes ............... Distiller’s Yeast ................. Food and Fodder Yeasts ............. Biochemicals from Yeast ............. References ...................
...
233 235 238 239 241 244 247 248 250 252 253 254 256 260 262
Research and development in the Brewing Industry differs significantly in emphasis from R. and D. in most other industries. Normally the major objective of industrial R. and D. is the creation of new products from existing raw materials, and such new products tend t o have a short life cycle. On the other hand, beer is a product of which the life cycle is measured in decades or even centuries, and brewing R. and D. is therefore to a considerable extent aimed at producing the same product from varying raw materials. The R. end of the R. and D. spectrum therefore tends to be concentrated on fact-finding and the D. end tends to be concentrated largely on process development and investigation ( Dalgliesh, 1972 ).
I.
Introduction
There are radical developments occurring in the fermentation techniques employed in brewing and other yeast-utilizing industries, and as a consequence achieving brevity in the discussion of such developments has been something of a Herculean task. In recent years there has been an increase of interest in yeasts from the industrial side as well as from a basic biochemical and microbiological research approach. In biochemical research, yeasts are convenient to handle, and for genetical investigations they offer the possibility of quick results that may in some cases be of economic importance. Another reason for the increased research in this group of microorganisms is the development 233
234
G. G. STEWART
of scientific approaches by such yeast-using industries as brewing and baking; furthermore, the realization that yeast is a valuable food source, in an ever increasing world population, may result in its use as a valuable adjunct to the diet. The past few years have seen the appearance of a number of textbooks and monographs devoted to yeast; three such works of special merit are those edited by Rose and Harrison ( 1970), Reed and Peppler ( 1973), and Lodder (1970). The first is a three-volume treatise and attempts a broad survey of the whole subject of yeast; the second is a broad introduction to the technology of yeast in the food and beverage industries; and the third is devoted to yeast classification and is an updated revision and expansion of the first edition of this work originally published in 1952. There have recently been published two books dealing with the scientific aspects of brewing (Findlay, 1971; Hough et al., 1971) together with a comprehensive review of the subject (Kleyn and Hough, 1971) . It is the intention of this review to consider modern developments in brewing and similar yeast-utilizing industries; fermentation technology in both batch and continuous processes will be discussed after a brief description of older, more traditional and well-tried procedures. Modern developments have been contrasted with their ancestors, and by this means it is hoped that the reader will be given as complete a picture as possible of the “state of the art” in this area. The group of microorganisms known as the “yeasts” is by traditional agreement limited to fungi in which the unicellular form is predominant ( Lodder and Kreger-van Rij, 1952). Vegetative reproduction is usually, but not always, by budding. This group does not constitute a taxonomic unity although it comprises subdivisions of narrowly related species. The diversity of the yeasts is illustrated by the fact that 39 genera and some 350 species are recognized. The main characteristics used for the classification of yeasts have been morphological or physiological, and to a lesser extent genetical. The most important taxon, the species, is based on similarities between strains, though some variability is permitted. Hence the delineation of a species is rather subjective, largely depending upon the insight and ideas of the taxonomist studying the group. The Adansonian principles and numerical methods applied by Sneath (1957) to bacterial taxonomy have been applied to yeast taxonomy by Kockovd-Kratochvkova et aE. ( 1965, 1967). Initially these Czech workers studied strains of Sacchuromyces cerevisiae and S . carlsbergensis using characteristics that are of value in brewing in order to choose the most promising strains for that purpose. In Poncet (1967) numerical methods were used to classify those Candida species which can ferment sucrose
BREWING AND OTHER INDUSTRIAL USES OF YEASTS
235
and maltose, considering in particular the species C . tropicalis, C. pelliculosa, and C . Tobusta. To date, the results of applying numerical taxonomy to the yeasts have not been spectacular. One problem is that insufficient data are known about the interdependence of the characters used for classification, and attaching equivalent weight to them all may give misleading results ( Barnett, 1968). However, numerical classification might usefully be applied to the larger genera, especially in the future when more is known about the background of the characteristics used. 11.
Historical Survey of the Industrial Uses of Yeast
The application of yeasts for preparing alcoholic beverages goes back far beyond the dawn of recorded history. It is known that beer was produced by the Sumerians more than 7000 years B.C. and wine by the Assyrians 3500 years B.C. It is difficult to imagine how such products tasted, but it is a fact that the techniques of malting and brewing were already highly developed among the Babylonians without any knowledge of the underlying biochemistry, the existence of enzymes or of yeast, and the role these agents play in the malting or fermentation process. One of the early forms of beer was a sort of fermented wet bread. Perhaps the coincidence of bread, a rain shower, and some yeast cells produced the first beer. The “bread beer” played a great role in old Mesopotamia, where part of the employee’s salary was paid in the form of beer. The same kind of beer was also produced in the Egypt of the Pharaohs, where it was regarded as a holy gift from Osiris, the God of the Dead. It was considered to be an enviable sign of wealth if one could buy enough beer to become intoxicated. Even children in Egypt at the time were given beer to take with them to school; beer was considered to have a certain therapeutic value in preventing stomach and kidney diseases. In the Israel of the Old Testament and Caesarean Rome beer was also drunk, although in Rome, unlike wine, it was forbidden to women. The Celtic people living in Central Europe before the invasion of the Germanic tribes made a beer that probably tasted somewhat better than bread beer, and the Germanic people probably learned brewing from them; in turn, the Huns learned the process from the Germanic people. It is said that Attila, the famous leader of the Huns, was convinced that his enormous personal consumption of beer was largely responsible for his victories. Beer was an extremely popular beverage at the time of Attila, but its taste must have been very dull because it lacked what is now considered to be the essential flavor of beer, the hops. Hops had been known
236
G. G. STEWART
as a drug since before the 8th century A.D. and were believed to be a good remedy against a number of diseases, including angina pectoris, sour stomach, and kidney stones. It was during the Middle Ages that hops were used as an addition to beer, as were many other flavors and spices in different “medicinal beers.” These beers were sold in the apothecary shops, which were more or less converted into bars, or at least something resembling the American drugstore. Hops have now won such a complete victory over other spices that a drink without hops would not be called beer, although nowadays the enormous rise in the cost of hops together with a consumer preference for lighter products has resulted in a considerable reduction in utilization of hops in beer than was the case even a short while ago. The use of yeasts for the rising of dough for baking, or “leavening,” was a process already used thousands of years ago for preparing bread. Undoubtedly wild yeast strains were at least partly responsible for the gas development that occurrcd. By retaining part of the fermented dough for the next baking, the process could be perpetuated. Several hundred years ago it was discovered that dough rise could be accomplished by adding the creamy foam of a top beer fermentation derived from the breweries of those days. Beer yeast was used extensively during the Middle Ages for bread making. Unfortunately, the bitter taste due to the presence of hop resins in such yeast foam was quite a disadvantage. Around 1780 a number of Dutch distillers in the town of Schiedam, famous for its gin production, started to experiment with the waste yeast of their distilleries for possible use in bread baking. This became successful, and soon it was found that it was possible and economically feasible to combine the production of gin with the recovery of a yeast suitable for baking purposes. There was a limitation on the system: the amount of yeast available from this process was restricted and tied directly to the amount of gin that couId be sold. This so-called “Dutch process” of baker’s yeast manufacture was a rather complex one and was very similar to the brewery process of today. The main ingredients were cooked rye flour and barley malt, which were digested together. After settling, the supernatant was decanted, and this clean sugar-containing medium was used for yeast production. The yeast used as the original inoculum was frequently obtained from a neighboring brewery. When fermentation was completed, the yeast settled to the bottom, and also a large part accumulated in the foam lager on top. The layer between, with comparatively few cells, was removed and distilled for alcohol and gin. The yeast slurry thus obtained was then passed through a silk screen for the removal of grain residues, and after washing and filtration the yeast was recovered as a cake. A Viennese brewer, A. I. Mautner, later developed the so-called Vienna
BREWING AND OTHER 1NDUSTRIAL USES OF YEASTS
237
process, which was similar to the Dutch process except that the whole mash, without removal of the grain residues, was fermented in shallow containers. The yeast-rich foam, which formed during the fermentation process, was continuously removed, and the yeast was recovered by filtration after extensive washing with water. Whereas the Dutch process gave yields of around 7-8 kg of pressed baker’s yeast per 100 kg of grain, the Vienna process gave 9-10 kg, and later even 13-14 kg per 100 kg of grain (under present conditions a yield of 120 kg is possible). A Dutchman, Antonie van Leeuwenhoek, is credited with being the first man to have observed yeasts microscopically. In 1680, he sent descriptions and drawings of yeast cells to the Royal Society in London. Leeuwenhoek was, unfortunately, not a scientist-he was a draper by profession-and it was left to the botanists of the day to lay the firm foundations of yeast microbiology. However, it was not until the first half of the nineteenth century that significant progress was made toward understanding the biology of yeasts, and through this to an appreciation of yeast physiology and biochemistry. Cagniard-Latour in 1837 demonstrated that beer contains spherical bodies that are able to multiply and which belong to the vegetable kingdom. In this, he received support from Schwann, who termed yeast “Zuckerpily” or “sugar fungus” from which the name “Saccharomyces” originates. This cellular or vitalistic theory of fermentation was, however, vehemently, and occasionally mockingly, attacked by a trio of chemistsLiebig, Wohler, and Berzelius. The euphoria which stemmed from the success of the newly recognized branch of organic chemistry in explaining hitherto complex and mysterious organic processes convinced this trio that chemical reactions, rather than the activities of living cells, could perfectly well explain the alcoholic fermentation of sugars. It finally fell to Louis Pasteur to prove that fermentation is due to living cells-thereby winning the argument for a vitalistic theory of fermentation. Pasteur continued to make masterly contributions to yeast microbiology and to understanding fermentation in general, particularly in his “Etudes sur le Vin” (Pasteur, 1866) and “Etudes sur la Bikre” (Pasteur, 1876) in which he cIarified much concerning the effect of oxygen on alcoholic fermentation by yeast. It can be seen, therefore, that the association of yeast with the science of biochemistry is an old and well-established one and that in fact biochemistry was born out of yeast technology. The Buchner brothers, working in Germany in 1897, were interested in preparing extracts of yeasts for medicinal purposes. To do this, they ground brewer’s yeast with sand and then squeezed out the juice. In an effort to preserve their extract, they tried adding large amounts of sucrose; since it was known that solutions containing high concentrations of sugar are less prone
238
G. G. STEWART
to microbial infection. To their surprise, they discovered that the sugar was rapidly fermented by the yeast juice. This chance encounter set off a whole series of studies in an effort to determine the nature of the steps in the fermentation of glucose to ethanol and carbon dioxide. This work spanned a number of years and is associated with the names of many outstanding biochemists. Ill.
Present-Day Industrial Uses of Yeasts
The major industrial applications of yeast have changed very little over the centuries. Yeasts still play the major role (Table I ) in several industries ( Hoogaheide, 1969): ( 1) the production of beer, wine, cider and other fermented beverages, such as ginger beer, the pulque of Mexico, and the sake of Japan; ( 2 ) the production of distilled alcoholic beverages such as whiskey, gin, vodka, brandy, and liquors; ( 3 ) in the baking industry, for the production of bread, rolls, and other bakery products. In latter years other applications have been found to add to those mentioned above, namely: ( 4 ) the production of industrial alcohol, mainly from sulfite waste liquor, wood hydrolyzates, or waste molasses; ( 5 ) the applications of yeast as food or feed ingredient either as such or frequently after autolysis to yeast extracts; (6) as pharmaceutical products mainly for its high vitamin B content; ( 7 ) as ingredients for the production of biochemicals, such as ergosterol, ribonucleic acid, and enzymes, products which in turn find use as starting materials for the production of certain vitamins, such as vitamin D, or as flavor enhancers, such as AMP and IMP. It is not possible in this presentation to discuss in detail all the applications described above for two reasons: ( 1 ) too much space would be TABLE I UTILIZATION OF YEASTS INDUSTRIAL Cell constituents Dry Macrowhole molecular yeast constituent Fodder Food
Lipids Proteins
Extraction compounds
Breakdown products
Excretion products
Enzyme-substrate interaction
Coenzymes Vitamins
Amino acids Purines Pyrimidines
Beer Wine Cider Spirits Glycerol
Whey utilization (8.fragilis) Starch utilization (Symba process) Maltotriose prdn. (8. uvarum)
cot
BREWING AND OTHER INDUSTRIAL USES OF YEASTS
239
required; ( 2 ) the author considers himself to be unqualified to consider all aspects of yeast use. As a consequence, particularly in the light of the author’s background, the use of yeast in beer production will be considered in some depth and the other applications will be reviewed briefly. IV.
Outline of Traditional Brewing Processes
The manufacture of beer is a biological process whereby agricultural products, such as barley and hops, are converted into beer by control of the biochemical reactions in malting, mashing, and fermentation. It has been suggested that malting and mashing may not always be a part of the manufacture of beer (Lewis, 1968) ; however, fermentation has at the present time and for the foreseeable future no prospect of being replaced. There are four main ingredients required in brewing: water, barley, hops, and yeast. All of these are equally important, and unless the quality of all four is constantly monitored and controlled, a variable product will result. Beer production is divided into three quite clear-cut processes: malting, mashing, and fermentation. The first two processes together produce a medium known as wort; to this yeast is added, and fermentation is allowed to proceed. Wort is essentially an aqueous extract of malted barley, the primary raw material in the manufacture of beer. A typical approximate chemical analysis of malt, expressed as percentage dry matter, is as follows: insoluble carbohydrates, 15; protein ( N X 6.25)) 2.5; starch, 58; ash, 2.5; sugars, 10 (Thorne, 1957). Some 3040%of the malt nitrogen is soluble, and about 20-25’K; of the soluble nitrogen is amino nitrogen. Among .its enzymes, malt contains significant amounts of CY- and p-amylase and protease, which play a vital role during the mashing processes. Malt nowadays usually is supplemented with unmalted cereals (adjunct), such as corn, wheat starch, cane sugar, or rice, to increase the carbohydrate content of the wort. The process of mashing need not be considered in detail here; suffice to say that it is an enzymatic process whereby most of the nonfermentable carbohydrates or proteins of the malt are converted into fermentable materials. This statement represents a gross oversimplification of the process, and the reader is referred to the recent textbook of Hough et al. (1971)for detailed consideration of the mashing process. The two main beer types are lager and ale, which are fermented with strains of Succharomyces carlsbergensis or S . cerevisiae, respectively, Traditionally lager is produced by a bottom fermentation procedure whereas ale is produced by a top fermentation-this means that at the end of fermentation bottom yeasts flocculate to the bottom of
240
G. G. STEWART
the fermenter and top yeasts tend to be somewhat less flocculent and form a yeast head on the wort surface. As a consequence thereof, top yeasts are collected for reuse from the surface layer of the beer, whereas bottom yeasts are collected fronr the fermenter bottom. The secrets of using bottom-fermentation yeasts were held by Bavarian brewers, notably in Munich, and until the middle of the nineteenth century the rest of the world used top-fermentation yeasts. The yeast and fermentation techniques were smuggled to Czechoslovakia by a Bavarian monk in 1842 and so helped to establish Pilsen as a premier brewing center. Only three years later, a Danish brewer, Jacobsen took bottom-fermentation from Munich to Copenhagen and improved Danish beer so that Copenhagen, too, bccame a world-renowned center of brewing. About the same time, bottom-fermentation was introduced into Pennsylvania and spread through the country, largely because of the immigation of German brewmasters. After the spread of bottom-fermentation, the traditional top-fermentation techniques were largely discarded except in Britain. However, a proportion of the beer produced in Australia, Belgium, Canada, the United States, and West Germany is of this type; in the last few years, this proportion has increased. Some brewers in the United States producing both Iager and ale use only one yeast, a strain of S . carlsbergensis, for both beer types, thereby eliminating the problems in keeping two yeast cultures separate-ales produced in this manner are designated “bastard ales.” A lager yeast used to ferment ale is usually not reused because of enhanced antolysis due primarily to the higher fermentation temperature of ale (15-22°C) as compared to lager beers (6-15°C). The taxonomic characteristic used to distinguish these two species is that S . cerevisiae can ferment only one-third of the raffinose molecuIe whereas S . carlsbergensis can ferment raffinose completely. As has been previously stated, this distinction happens to coincide in many cases with the ability of S. cerevisiae to form a yeast head toward the end of fermentation, whereas strains of S. carlsbergensis fall to the bottom of the fermenter. This coincidence is by no means universal, and it is common to find strains of S . cerevisiae that behave as bottom yeasts, just as it is possible, in some rare circumstances, to find strains of S. carlsbergensis that form a head during fermentation. The formation of a yeast head depends on the fermentation temperature and on an adequate depth of liquid in the fermentation vessel, as well as on the yeast strain. The traditional method of ale fermentation over a long period of time has been a batch process, taking place in shallow rectangular vessels in which the bulk of the yeast rises to the top of the fermenting wort at the completion of the fermentation. The yeast is skimmed off and
BREWING AND OTHER INDUSTRIAL USES OF YEASTS
Ul
stored in the cold either as a slurry or in the pressed form until it can be reused. During the course of a fermentation it is usual to skim 2 or 3 times. The first (or dirty) head, contaminated with wort particles, is rejected; the middle head, thrown up toward the end of active fermentation, is collected for use in subsequent brews or the final head, which forms a protective cover for the beer against airborne infection, is also rejected. Once the skimming is complete, the remaining yeast can be encouraged to come out of suspension by cooling; it is common practice to reduce the temperature of the fermented wort to 2 4 ° C by the end of fermentation. The total fermentation time for an ale is usually 5-8 days. The feature which principally distinguishes traditional bottom fermentation (i.e., lager) from top, other than the yeast species and the direction of yeast flocculation at the end of the fermentation, is the lower temperature at which it is conducted. The initial temperature is within the range of 615°C (the lower temperatures being especially preferred in Germany), and the fermentation lasts for 7-14 days. A wide range of conditions are used in different countries. In Denmark it is usual to ferment for 8 days at 9-10°C; in Western Germany the time of fermentation may range from 5 to 14 days at a temperature between 5 and 10°C. American and Canadian practice is to ferment for 6-9 days at temperatures ranging between 10 and 15°C but being reduced to as low as 2°C before the completion of fermentation. Yeast recovered from bottom fermentation is often passed through an oscillating fine sieve and then washed with cold water to remove wort solids, bacteria, and dead yeast cells. However, with the advent of more aseptic methods, bacterial contamination in most large breweries is no longer a serious problem. The denser living cells are retained on the sieve and stored under water, beer, or wort at about 2°C. Yeast rooms are usually refrigerated, but water-jacketed storage vessels are also used. Only a proportion of the yeast is retained for reuse; the middle layer of yeast sediment in the fermentation vessel is the one that usually receives the above treatment. Before being inoculated into a fresh fermentation, the yeast is slurried in either wort or water and injected into the wort leading into the fermenters.
V.
N e w Developments in Batch Processes for Brewing
When developments on the fermentation side of the brewing industry are considered, continuous processes are at once brought to mind. It should be remembered, however, that a number of brewing companies have made the decision to remain with the batch process. Although developments with batch systems have not been so radical as with the
242
G.
G. STEWART
continuous processes, there are developments in this area that are worthy of note. It was in 1880 that Emil Christian Hansen devised methods for isolating single cells of brewery yeasts by repeated dilution of a yeast suspension. He was therefore in a position to separate the component strains of a yeast mixture and study them in isolation. The technique provided an opportunity to free the yeast from attendant bacteria and wild yeasts and also provide the brewer with a pure culture. As a result of this work, the practice of using a pure strain in lager production was soon adopted, but in ale-producing regions this radical innovation met severe opposition, the method being regarded merely as a means to reduce infection by wild yeast and bacteria. Over the years, however, the use of pure yeast strains has increased in ale-producing areas, and it is interesting to note that it was reported by Hough (1959a) that of 39 yeast cultures in use commercially in Britain, 12 contained only a single strain, 16 had two major strains, and the rest had three or more components. If a similar survey were undertaken today, it is probable that the percentage of pure strains being used for beer production would be considerably higher. The use of pure cultures in brewing naturally involves considerable attention to the management of the yeasts employed. Pure culture brewing, although it removes some of the anxieties about the changing characteristics of the yeast culture (i.e., changes in the proportion of the yeast strains one to another, thus leading to possible changes in fermentation characteristics) and the invasion of gross infection, does not eliminate all the brewer’s worries. Even with the yeast strain derived from an original single cell or from an ascus, mutation is possible. Recent studies on yeast mutation as it affects the brewing industry will be discussed in a later section of this review. There have been some modifications in fermenter design for batch culture, and allied with this the use of bottom fermenting strains of S. cerevisiae for ale production. A number of breweries in the United Kingdom have begun using Nathan (cylindrical, conical based) fermenters for ale production ( Wakerbauer, 1969; Thompson, 1970). A considerable amount of developmental work was done by one particular brewing company prior to the installation of Nathan fermenters in a new brewery, and much of their experience in this venture has been well documented (Shardlow, 1972). Their first problem right at the start of planning was whether to replace the conventional batch system with a continuous fermenter. Initially a continuous system was favored, and considerable progress was made toward developing a suitable process. It was eventually rejected, because it was not sufficiently flexible to allow for the fermentation of the different types of ale with annual
BREWING AND OTHER INDUSTRIAL USES OF YEASTS
24.3
production varying from 15,000 to 150,000 barrels to be produced in the new brewery. The problem then became one of deciding upon the most economical batch fermentation that could be devised to take full advantage of inplace cleaning of fermenters and the requirement to devise similar methods of yeast handling so as to reduce labor costs and beer losses to a minimum. Such considerations led to the conclusion that enclosed vertical cylindrical vessels provided with a conical base would be the most suitable for their purposes. It was found that for best results it was necessary to use a yeast which sedimented to the base of the vessel at the completion of fermentation, and for this purpose a suitable strain was isolated. During the next 2 years the potentialities of the process were fully evaluated, and at the completion of this developmental phase a number of conclusions had been reached: 1. The angle of the cone at the base of the vessel must be reasonably steep to assist in yeast removal and to give a clearly defined yeastlfermented wort interface, a 90' cone proved much more satisfactory in this respect than did the shallower types tried. 2. The centrifuging of green beer at the end of attenuation was found to produce unacceptable changes of palate. 3. The beer produced without the use of a centrifuge was completely acceptable with regard to flavor. This method of fermentation has been used for over 3 years, and all reports indicate that the operation is satisfactory and has led to substantial savings in labor costs, plant buildings, installation costs, and beer losses. The general trend is for faster fermentation. There are many traditional processes which are based on the idea of partially emptying a vessel, in which fermentation is active into a second vessel and replenish the first with fresh medium. In this way, the lag period of growth of the yeast is reduced or even eliminated in the replenished vessels ( a process known as 'drauflassen') . Further increases in fermentation rate can often be achieved by: (1) keeping the yeast free of infection, ( 2 ) stirring so that the yeast is kept in suspension, ( 3 ) raising the temperature to a maximum of about 28'C, ( 4 ) increasing the yeast concentration, ( 5 ) increasing the dissolved oxygen concentration of the wort, ( 6 ) optimizing the composition of the wort for maximum rate of fermentation. Speeding up fermentation rates is relatively simple for most breweries, but achieving it without change or even deterioration of the beer quality is far more difficult. The acceleration is quite likely to alter the levels of beer constituents which are important in beer flavor and aroma. This is particularly true of such substances as esters, diketones, fuse1 oils, and hydrogen sulfide (Hough, 1961; Watson and Hough, 1969). General
2/44
G . G . STEWART
experience seems to indicate that if the amount of yeast produced, per unit volume of beer, is changed significantly, important alterations in the character of the beer occur (Portno, 1969). This does not mean that acceptable beer cannot be produced if the yeast crop is smaller or greater than normal, but the beer will tend to differ appreciably from the normal beer in taste and aroma. The next logical progression from a batch system is a semicontinuous one; such a system may be fed with wort either continuously or batchwise. In either case the beer can be discharged in batches. There are logistic reasons for preferring such a system to the less flexible fully continuous methods requiring continuous feed and discharge. In one example of a semicontinuous system ( Haboucha et al., 1969), cylindrical-conical vessels are 5%filled with wort (12.5'P) at 10-12°C and seeded with yeast at a rate of 144 pounds/bushel. After 4.5 hours, the time needed for 75% attenuation, fresh wort is fed in at such a rate that the vessel is full in a further 4448 hours. Providing that the yeast is kept in suspension, the attenuation of the beer is satisfactory. Beers produced in this manner are similar to those derived from traditional fermentation processes. It is known from other studies (Harris and Watson, 1971) that the wort may be fed at a uniform rate into the partially filled vessel, or at a variable rate to maintain a given degree of attenuation. A flocculent strain may be selected so that the sedimentation of the yeast is rapid when the vessel is full, The beer can then be run off comparatively free of yeast, leaving a large quantity of yeast for the next delivery of wort into the vessel. One of the disadvantages of the technique is the relatively high percentage of dead cells when high concentrations of yeast are maintained, and this has caused some concern among brewing scientists. It is now known why viability is low under these conditions but it would appear likely that ( 1 ) yeast cells in high concentrations do not bud because the wort is not a suitable medium for such cell numbers and ( 2 ) yeast cells that fail to bud for a considerable period of time tend to have weak fermenting ability. VI.
Continuous Fermentation in Brewing
Patented methods of continuous fermentation have existed since the beginning of the century (Barbet, 1905; Van Rijn, 1906). In many respects, they resemble the more recent methods, but advances in vessel design and construction and in microbiological techniques have given these new methods better chances of survival. To the brewer, the most important advantages of continuous fermentation are that present production can be achieved with much smaller plant, uniform quality of product can be maintained with lower labor costs, and the early slower
BREWING AND OTHER INDUSTRIAL USES OF YEASTS
245
stages of batch fermentation are eliminated so that according to some investigations (Righelato and Elsworth, 1970), the most rapid rate of a batch system can be continually maintained or even exceeded. It should be emphasized, however, that the enthusiasm for continuous fermentation which existed some 10 years ago has somewhat abated. In the intervening decade it has been tried in one form or another in many breweries, and a good deal of sobering, and even painful, experience of the process has accumulated (Portno, 1973). Continuous fermentation on a full production scale was pioneered in New Zealand. Encouraged by this venture, and by the generally favorable climate of opinion at the time, several major brewing concerns have, during the past 15 years, undertaken extensive trials of the process. Although the available information as to their experience is scanty most breweries decided to remain with the batch process in one form or another. That is not to say that the continuous process has been completely rejected by the brewing industry; to the contrary, a large percentage of the total beer production of New Zealand and of the United Kingdom is now produced by continuous processes ( Coutts, 1957). There are two basic types of system of continuous culture of microorganisms (Emeis, 1965): (1) Turbidostatic systems in which the cell concentration is kept constant by varying the flow-rate, which is controlled by a photoelectric device. The cell population may be maintained in the exponential phase of growth. These systems are particularly suitable for metabolic studies and operate under physiological conditions quite different from those in chemostatic systems. ( 2 ) Chemostatic systems, in which the growth rate is regulated by the dilution rate and is kept well below the maximum value by maintaining concentrations of growth-limiting factors. The medium is introduced at a constant rate into the culture vessel and the cell population is maintained in the retardation phase of growth which precedes the stationary phase. Continuous fermentation systems of commercial interest to brewers are chemostatic systems. The efficiency of these systems is usually measured as the flow-rate of wort in unit volumes per day per unit volume capacity of the apparatus (Hough and Rudin, 1958); or by the number of times per day the vessel contents are completely changed (Hough, 1959b). A number of different types of continuous fermentation systerps are presently in use or are being considered in the brewing industry.
1. OverfEow System These are “open” continuous fermentations either in one stirred and temperature-controlled vessel, or in a series of such vessels, into which wort is pumped at one end while the beer overflows at the other. A
246
G. G . STEWART
two-vessel system has been found to be more efficient than a single vessel, but a three-vessel system is more advantageous (Hough and Rudin, 1959). In these systems, the yeast is maintained dispersed in wort by agitation, while fresh wort is introduced at the same rate as beer and suspended yeast are discharged, the concentration of yeast in the effluent beer being the same as that within the stirred vessel. In some systems, this yeast is returned to the fermenting vessel so as to increase the yeast concentration and thereby the rate of fermentation. A V-tube apparatus has been devised (Hough and Ricketts, 1960) which exploits the sedimenting properties of flocculent yeast to retain the yeast within the vessel, so that problems of yeast separation of recycling are minimized, saving high fermentation rates and low rates of yeast reproduction. The apparatus consists of a stirred vertical tube connected at its base to an unstirred tube set to an angle of 25 degrees of arc. The vertical tube is charged with pure flocculent yeast, and wort is pumped into the top. The yeast is kept in suspension by stirring, but it settles to the bottom of the unstirred inclined tube, while clean beer flows off from a point near the top.
2. The Tower System The tower system consists essentially of a vertical tube charged with a high concentration of yeast (Klopper et al., 1965). Wort enters at the bottom, passes through a plug of yeast, and beer overflows at the top, after passing through a section designed to disengage gas and separate yeast, the latter falling back into the tower. As a result of the high concentration of yeast maintained in the tower, fermentation rates are very high, with careful selection of the correct yeast strain, ale and lager worts have been fermented satisfactorily in 4 and 8 hours, respectively. The tower system is a heterogeneous, virtually “closed fermenter, in which little yeast growth occurs. With normal brewery wort, this leads to the production of beer with a higher nitrogen content than usual, however, this can be compensated for by using worts containing a low concentration of nitrogen-containing compounds. It has been recently reported ( Ault et al., 1969), however, that if the tower is operated with a degree of aeration, which will promote more yeast reproduction, beers with normal nitrogen contents can be obtained. 3. “Gradient Tube” Fermenter
This is designed to operate as a nearly perfect heterogeneous “open system,” incorporating a true fermentation sediment, which reproduces, under continuous conditions, the sequential changes of a batch fermentation. It is essentially a long narrow ( 0.25-inch bore) temperature-con-
BREWING AND OTHER INDUSTRIAL USES OF YEASTS
2,447
trolled tube, into which controlled proportions of pasteurized wort and yeast are fed, so that the fermentation proceeds as the mixture passes along the tube. A mixture of beer, gas, and yeast is ultimately discharged into a separating vessel, from which settled yeast is recirculated to enable efficiencies greater than those usual for open systems to be attained.
4. Centrifugal Continuous Fmmentation Systems In closed fermenters, such as the tower, control of yeast concentration depends on the opposing factors of rate of upward movement of liquid and the sedimentation characterizations of the yeast. In addition, there is a hazard that high concentrations of flocculent yeast may form cohesive plugs, which ferment inefficiently because o’f the limited surface of the yeast exposed to the wort, and which may be expelled from the fermenter on a cushion of fermentation gas. The centrifugal fermenter (Portno, 1967) attempts to overcome these difficulties and enables greater control to be exerted over the escape of yeast. It consists of a single stirred vessel from which beer escapes through passages radiating from a control spinning rotor. In escaping, yeast suspended in the beer is partially or entirely removed according to the speed of rotation of the rotor and its length of radius. The apparatus enables less flocculent yeast to be used than would be acceptable for tower operation. It operates as a homogeneous fermenter in which, under steady state conditions, the yeast remains in a constant metabolic state in a medium almost devoid of usable nutrients.
VII.
Comparison of the Available Modern Fermenting Systems in Brewing
Without in any way denying the place of traditional batch fermentation in modern brewing practice, employing modern technological advances, there is no question that there are considerable economic and quality advantages to be achieved in both vertical conical-bottom fermenters and continuous fermentation systems. Ricketts (1971) considers that the conical-bottom fermenter is the easiest departure from the traditional batch processes. Its advantages can be calculated in terms of ( a ) reduced building costs, ( b ) reduced beer losses by virtue of eliminating the flat-bottom vessel, ( c ) reduced labor costs by virtue of eliminating skimming, and ( d ) increased turnover of vessels by virtue of reduced fermentation time. The “open” continuous fermentation system offers in addition ( a ) approximately 258 lower hop rate, ( b ) further reduced labor costs, ( c ) collection of all the CO, in 100%pure condition, and ( d ) further reduced
2A8
G. G. STEWART
building and vessel costs by virtue of fermentation times being so much shorter. Additional costs of the “open” system, particularly of those for which operating costs are available, include electrical stirring and cooling costs when the wort is chilled prior to pasteurization. Also, comparatively large volumes of fermenter contents require substantid cooling during close-down and reduced production. The tower fermenter offers yet further advantages over the “open” system. These include: ( a ) reduced beer losses by virtue of less yeast growth (not so important, because of the excise system, in Canada as it would be in Britain), ( b ) cooling costs similar to those for batch fermentation, ( c ) probable increased alcohol yield, and ( d ) possible savings in maturation. VIII.
Selection, Behavior, and Efficiency of Yeasts
For all systems, whether batch or continuous, the primary considerations of yeast selection must be that it produce a palatable beer and that it ferment wort vigorously in the conditions of the chosen system. In view of the very wide variation in properties of different strains of brewing yeasts, it may appear an immense task to select a strain that is adequate in all characteristics. If, however, suitable starting material is used, and if a systematic elimination plan of undesirable strains is followed, it is not too difficult. The first requirement is a list of the properties of the ideal yeast; the second is a set of tests by which these properties can be detected and measured; and the third is a source of suitable yeast strains in which the ideal strain may be found. The list of properties should be divided into “essential requirements” and “desired characteristics.” For example, it might be considered essential for a top fermentation brewery that the yeast should give a satisfactory top crop, that the fermentation be completed within a given time, that the yeast have a normal attenuating limit, and that the beer produced should have the desired flavor characteristics. On the other hand, the “desired characters’’ are generally measurable quantities, which could be improved. For example, it may be decided that the rate of fermentation should be increased, or that the total yeast crop should be reduced, or that the amount of isohumulones removed during fermentation should be reduced, or that the resistance of the yeast to autolysis should be increased. It also helps in the choice of strain if a relative order of importance can be given to the various characters, of if one character can be selected which it is most desired to improve (Gilliland, 1971). A suitable series of tests must be devised to determine how any strain
BREWING AND OTHER INDUSTRIAL USES OF YEASTS
249
measures up to the requirements, and the most important of these tests is a fermentation of brewery wort. In a preliminary survey, anything from a few dozen strains to many hundreds of strains may be tested, depending on whether the object is simply to select a strain from one’s own yeast or to extend the range of the search to all available sources. First, the microscopical appearance, flocculation characteristics, and attenuation limit of each strain are recorded. Then the behavior of each strain during fermentations in wort in 2-liter tubes is studied with emphasis on the following parameters: fermentation rate, duration of fermentation, separation of yeast, final attenuation of the beer, total yeast crop, brightness of the beer, pH of the beer, loss of isohumulones during fermentation, and volatiles produced. These results allow the selection of a much smaller number of strains, which should then be subjected to a series of tests in which the appropriate yeast crop (skimmings or bottom yeast) is harvested and used to pitch successive laboratory fermentations. Finally, the strains which have proved to be most promising can be tested in the brewery. Several sources of yeast strains are available. The most obvious of these is the yeast already in use, particularly if it is an old established, or mixed, culture. A second source is yeast from other breweries, preferably from those that use a yeast cropping system similar to that for which the yeast is required. Culture collections can also be useful sources of yeast strains, where advice may be available, or which strains would best fit the requirements. Another source of yeast strains is production by hybridization. It is difficult, however, to hybridize brewing yeasts, and it has recently been suggested (Fowell, 1969) that a breeding program simply designed to produce a “better” brewing yeast was unlikely to succeed, but a program designed to improve one specific characteristic should have a good chance of success (Gilliland, 1962). Recently, however, a mass mating procedure has been used to simply try to improve the overall performance of brewing yeasts, and a fair degree of success has been achieved (Clayton et al., 1972). If a yeast is required for a new type of fermentation, then a breeding program should be worthwhile. For example, there has been no lengthy natural selection operating in continuous fermentation; however, by means of a breeding program, a hybrid with improved ability for continuous fermentation was produced (Johnston, 1963). The production of a hybrid between top and bottom brewing yeasts has been reported (Enebo et al., 1960), which was better than the parent yeasts for producing good aroma and flavor in the beer. A breeding program has been used by Windisch and Emeis (1969) to produce a yeast that has a very low attenuation limit. They first made
250
G. G . STEWART
a hybrid between a brewing yeast and a yeast with the superattenuating characteristics of S . diastaticus. This hybrid was heterozygous for dextrin fermentation. Single spore cultures of the hybrid tested for superattenuation and for suitability for brewing suitable strains were noted to give diploid cells homozygous for superattenuation. A further selection of these diploid was necessary to obtain the most suitable strain with respect to yeast handling and the products of a palatable beer.
IX.
Baker’s Yeast
The function of yeast in breadmaking (Burrows, 1970) is 4-fold: (1) to increase dough volume by the evolution of gas during the fermentation of the available carbohydrates in the flour; ( 2 ) to develop structure and texture in the dough by the stretching effect of the expansion due to gas production; ( 3 ) to impart a distinctive flavor; and ( 4 ) to enhance the nutritive value of the bread. The last is a very minor function. It is estimated that at present at least 700,000 tons of baker’s yeast is produced worldwide per annum ( Hoogaheide, 1969). Whereas in bygone years production occurred in thousands of small yeast plants, evenly distributed over the bread-consuming areas of the world, there is at present a tendency to consolidate production in large factories. This becomes possible because of the availability of better storage-resistant strains and more efficient and faster distribution systems. To illustrate this: Germany in 1891 had over 1100 yeast plants. At present, notwithstanding double the baker’s yeast production of 1891, this number has been reduced to 26, and of these, the three largest produce 50% of the total output. It is unlikely that baker‘s yeast production will expand very much in the future, especialIy in the developed countries. Statistics show that bread consumption per head of population first increases with improvements in living standards and then, with further improvements, decreases. A considerable and steadily increasing part of baker’s yeast at present is produced in the form of live dried yeast, with a moisture content of 5-102, When properly stored under conditions excluding air, live dried yeast retains its fermentative power for years instead of weeks and is thus eminently suitable for use in areas where it would not be economically feasible to produce fresh baker’s yeast. Special resistant yeast strains and a special culturation technique yielding a low protein product are required in order to obtain a yeast that can withstand the drying process without too much loss in fermentative power. Drying is usually done by pressing a paste of fresh yeast through small openings, yielding spaghetti-like strands, and drying such strands either in slowly rotating
BREWING AND OTHER INDUSTRIAL USES OF YEASTS
251
drums with an air current passing through the moving yeast particles or a moving belt drier. More recently a fluid-bed drying process with the particles suspended in an air current has come into use. Also careful spray-drying of a yeast is reported to give an acceptable product. A drawback of the use of live dried yeast in the bakery is the fact that rehydration before use has to be performed quite carefully, preferably by adding the dried yeast directly to a 5% sugar solution preheated to 35°C;otherwise part of the cell content leaks out of the cell before rehydration is completed, causing serious loss of activity. Baker’s yeast is almost always a strain of S . cereuisiae and is most easily grown on a medium comprising simple hexoses, including glucose and fructose and dissaccharide, usually sucrose or maltose. S. hctk, which, unlike S. cereuisiae, can ferment lactose, has been grown commercially on whey as a baker’s yeast (Pyke, 1958) but is not widely used. Some commercial baker’s yeasts are derived from naturally occurring strains selected by trial and error. In certain instances, strains of brewer’s yeast may be found, after appropriate treatment to free them from bittering substances and other contaminating materials, to be suitable for propagation as baker’s yeast. Alternatively, new strains of baker’s yeast may be produced by mating existing types (Fowell, 1966).It is interesting to note that breeding programs to produce baking strains of yeast have been far more successful than similar attempts to produce brewing strains. The reason for this must be that the parameters of a good baking strain are well defined whereas those of a good brewing strain are still not completely understood. Yeast production ;s achieved by an intensive vegetative multiplication of yeast cells resulting in an enormous synthesis of living matter from a few simple substances. The energy required for this is obtained by aerobic carbohydrate breakdown, at the same time that the carbohydrate supplies carbon, hydrogen, and oxygen atoms for synthesis. The most important source of carbohydrate for yeast production nowadays is still molasses. Beet molasses content is characterized by approximately 4540% sucrose, a small amount of hexose, and 1.5-2% nitrogen (of which only about one-third consists of assimilable compounds). Cane molasses has generally a somewhat higher sucrose content, 55-654 of which a large part may be hexose. The nitrogen content of cane sugar is generally lower than in beet molasses-less than 1%. The amount of assimilable nitrogen present in all forms of molasses is far from adequate, and consequently simple ammonium salts are usually added to the substrate. A mixture of ammonia and ammonium sulfate may be used, and, by varying the ratio between these two chemicals during cultivation, a ready means of regulating pH is available. Phosphate must
252
G.
G. STEWART
also be added, usually in the form of diammonium hydrogen phosphate. Previously the cheaper chemical, superphosphate, was often used, and it still is in some countries, but as superphosphate often contains considerable amounts of fluoride its use should be avoided. Large quantities of yeast are required as inocula for the final commercial propagation stage when batch processes are used. This “seed yeast” must be prepared as aseptically as possible, preferably commencing with a pure culture. Complete asepsis after the laboratory stage is not necessarily practiced, but precautions are taken to reduce contamination to a minimum. The vital seed stages are frequently conducted under conditions of limited aeration, and are true batch fermentations in the sense that all the nutrients are present in the vessel at the time of inoculation.
X.
The Fed-Batch Process
A suitable amount of clean water, various chemicals (phosphate, magnesium, and ammonium compounds), a small part of treated molasses, and a suitable amount of seed yeast are added to a steam-sterilized, cooled container for the final propagation phase ( Rosen, 1968). The seed yeast may have been stored as yeast cream under refrigeration (2-4°C) for 1-2 weeks, and the necessary amount of yeast has to be pumped into the fermenters when the fermentation is to commence, possibly after a brief treatment with acid at pH 7. The yeast is not damaged by this treatment, but most of the contaminating microorganisms are impeded-to such an extent that they cannot multiply during the subsequent fermentation process. The air inlet is opened and molasses and chemicals are admitted in quantities corresponding. to the immediate requirements for yeast growth. The flow of these may take place according to an exponential curve. The basis for this is the observation that all yeast cells divide at a constant rate ( r ) , when there is sufficient nutrient matter available. The amount of yeast produced ( d x ) in time ( d t ) may then be expressed by the equation dx = rxde; x is the amount of yeast in the vat at time t. By integration we have for a batch process. x = xo exp ( r t ), when xo is the amount of pitching yeast. During the fermentation, the temperature is maintained around or just below 30°C. The pH of the fermenting liquid is kept between 4.0 and 5.5. The lower pH values cause the yeast to become darker, but the growth of contaminating microorganisms is impeded when the addition of molasses and chemical feed are completed, aeration is continued for another 30-60 minutes, whereby the yeast is “ripened,” i.e., the yeast is brought into a more stable state with better keeping qualities.
BREWING AND OTHER INDUSTRIAL USES OF YEASTS
XI.
253
Continuous Processes
In the fed-batch process, such as has been described, the growing yeast cells can be maintained in virtually a steady-state environment. That is, the concentration of all nutrients and the near physical parameters, such as temperature, can be kept constant. The whole system, however, is not strictly in a steady state since either the yeast concentration or the volume of the medium is increasing. In the continuous process, as discussed when brewing processes were described, the whole system is in steady state with medium entering and yeast leaving the system at a constant rate, although there are abrupt changes of environment as the individual cells pass from one stage to the next. It is clear that the change from a fed-batch to a continuous process is not so great as from a cIassical batch to a continuous process, where a completely nonsteady state is replaced by a truly steady-state system. There has been a singular lack of interest, however, in continuous systems to produce baker’s yeast commercially, only one instance being quoted in the literature (Sher, 1960, 1961). There are probably several reasons for this lack of interest. Infection problems have frequently troubled yeast factories when using the batch process, but even greater dangers from infecting organisms are inherent in continuous systems. Incomplete understanding of the theory of steady-state continuous processes has also hindered their application. Until about 1950 the mathematical theory of the process had not been developed, and no published method contained any reference to the need for strict control of growth rate. The term growth rate was sometimes used erroneously, in the sense of the hourly increment of the fed-batch process. It was applied to the building up of the necessary yeast concentration in the system before the onset of steady-state conditions, but there was no clear description or understanding of how conditions in the steady-state phase were controlled. It is now well known that, in a steady-state process at equilibrium, considerations of material balance lead to the conclusion that the growth rate of the organism is the same, to a close approximation, as the dilution rate (Monod, 1950; Maxon, 1955). It follows, therefore, that the control of dilution rate is essential, and is equivalent to the correct preparing of the molasses feed in the batch process. Two methods of maintaining the system in a steady state are available. In the first, dilution water is fed at a predetermined rate, and yeasted medium is pumped out of the vessel at such a rate that the volume of liquid in the vessel remains constant. Transfer of yeast in this way is controlled by a level detector attached to the vessel. By the second method, yeast medium
254
G . G . STEWART
is removed from the vessel at a constant rate, and addition of dilution water is controlled by the level detector. Further processing of the yeast, whether produced by a batch or continuous process, includes: ( 1) centrifugation, whereby the yeast is separated from the major part of the liquid and becomes a yeast cream; ( 2 ) washing, whereby the yeast cream is diluted with abundant amounts of water; and ( 3 ) centrifugation for removal of the washing water. As a result of these procedures a yeast cream is obtained containing 650-800 gm of yeast per liter. The yeast cream is cooled by plate heat exchangers, in a countercurrent of cold saline or an alcohol-water mixture, and is then collected in closed isolated containers with slowly rotating stirrers and kept at a temperature of 2-4°C. The removal of surplus water was formerly done by filtering through filter presses, but these have now practically been replaced everywhere by vacuum filter dehydraters. The filtered yeast is formed, divided, and packed in amounts of 1 kg or 0.5 kg for bakers, and possibly in smaller units-50 gm-for domestic use. The packed yeast should be stored in refrigerators at 2-6°C. If it is distributed quickly and treated properly by the bakers, the yeast will normally keep for a few weeks or more. Improper treatment, however, will lead to autolysis, and the dough-raising properties will be gradually lost. XII.
Distiller's Yeast
Although methods for the concentration of spirit from weaker brews have been known since the time of the Egyptians, the history of the use of special yeasts for the preparation of spirit prior to distillation dates only from about the beginning of the present century. Previously brewer's and wine yeasts were used. When it became apparent, however, that benefits would be gained by manufacturing industrial spirit in the greatest possible yield at high concentration in the shortest time, for economic reasons, attempts were made to select suitable yeast strains from those available. It is now the common practice to select, adapt, or breed special strains, which are maintained as pure cultures, for most of the specialized processes by which distilled spirit is manufactured. The virtues of particular yeasts for the production of the different forms of potable spirit are difficult to assess because of the subjective nature of the required properties (Harrison and Graham, 1970). For the commercial production of industrial spirit, however, a quantitative estimation can be made on theoretical grounds of the total available carbon-containing compounds in the fermentation medium which can be converted to ethanol. As the manufacture of spirit for industrial purposes became
BREWING AND OTHER INDUSTRIAL USES OF YEASTS
255
more competitive, owing to the development of large-scale chemical synthesis, much attention was given to this subject, and yields of ethanol very close to the theoretical maximum can now be obtained. The economic importance of ethanol is considerable, the annual world output of potable spirit being equivalent to about 20 million hectoliters (1.5 million tons) of pure ethanol, and of industrial alcohol several million tons, of which 1.3 million tons were made by fermentation processes in 1963. Yeast therefore still plays a very significant part in producing ethanol for large-scale manufacturing purposes in spite of synthetic competition, and it contributes indirectly considerable sums of money, of the order of billions of dollars, to national exchequers in the form of excise duty. In many countries, for instance, France, Scotland, and the West Indies, spirituous liquors represent a high proportion of the export business. It has been mentioned previously that the ultimate criterion of the suitability of a yeast strain must be the result of its performance under working conditions. Because of the complexity of the raw materials, fermentation, distillation and maturation processes, and the judgment of the properties of the final product, the choice of cultures for the preparation of potable spirits is nearly as difficult as for brewing purposes. On the other hand, for industrial spirits, where only high ethanol content and possibly freedom from particular chemical contaminants is required, the problem can be approached in a straightforward scientific manner. In the latter case, simulated conditions of time, temperature, pH value, etc., can be applied in the laboratory, using the raw materials grown to be commercially available, and by careful analysis of the fermented liquor the yield and quality of the spirit can be found relatively quickly. Laboratory studies of alcohol and sugar tolerance and growth rate can be carried out. The alcohol tolerance for instance, can vary from 5-82 to 11.62 for different strains (Stark, 1954). If necessary, pilot plant tests can be used to confirm the results on closer strains. Yeasts can be screened and selected in this way, and, if required, new types can be produced by adaptation, breeding, or mutation. The problem with potable spirits is different. In the case of brandy, where many types of wine made by various fermentation processes can be available for distillation, it is possible to select the most suitable source. This has taken place over the years, and no doubt accounts for the original localization of brandy distilling in certain areas. Many factors contribute to different types of wine (grape variety, soil, rainfall, sunshine, etc.), but a major factor is the variation in yeast population in different districts. The special and unusual characteristics of the pot stills used for the distillation of many potable spirits make it impossible to judge the effects
256
G. G. STEWART
of any change in the manufacturing process, including the yeast strain, without repeated full-scale tests. In many cases, in modern practice, selected culture yeasts are specially grown up for use as the inoculum for fermentation processes in distilleries; it can therefore be inferred that changes that are observed under those circumstances cannot be due to instability in the genetic makeup of the yeast employed. For these reasons it is difficult to express quantitative views on the effect of yeasts or spirit quality. It is possible, however, that as a result of the great efforts being made to understand the biochemical, physical, and physiological mechanisms involved in the production and quality of distilled products, it will become possible to select yeast strains by sophisticated laboratory techniques for these purposes. The final development, to tailor-make strains with the desired organoleptic properties, remains for the future. XIII.
Food and Fodder Yeasts
“During the last 50 years people have become so used to the excellent results produced by the chemical industry that they consider everything possible and therefore look upon synthetic proteins as the cheap and common food of tomorrow. This hope was illustrated in a paper under the heading ‘Nutrition from Coal’ by a splendid picture showing the conversion of coal into various delicious dishes in a sophisticated restaurant connected to a coal mine. Unfortunately the soberly thinking chemist cannot tolerate such daring expectations” (Fisher, 1907). These were the words of the German chemist Emile Fisher, 65 years ago. Today, there are quite a number of realistic microbiologists who do not hesitate to draw the picture of a coal mine or oil well coupled to a restaurant by a fermentation and food technology unit. The engineering problems of this futuristic view have largely been solved, owing to the development of industrial processes for the large-scale cultivation of microorganisms on natural gas and petroleum. The microorganisms that are considered in this connection are microalgae, bacteria, yeast, and filamentous fungi. The remarks in this review will be confined to the yeasts; recently, however, a review dealing with all forms of singlecell protein has been published ( Kihlberg, 1972). Large quantities of yeast slurries, by-products of breweries and distilleries, are dried and sold as high-protein fodder. In many countries fermentation industries produce large quantities of other types of yeast, cultivated specifically to serve as protein sources in order to improve the quality of feed for farm animals. As long as sufficient quantities of legumes are produced by agriculture, which after oil extraction yield a meal that nutritionally is practically equal in quality to dried yeast,
BREWING AND OTHER INDUSTRIAL USES OF YEASTS
257
this product must compete in price with soya flour, cottonseed meal, and fish meal, which have a world market value of approximately 10-12 cents per pound of protein. For this reason, the production of fodder yeast from molasses is usually too expensive, except maybe in areas close to cane-sugar factories where backstrap molasses often is an unwanted by-product and where high-protein fodder would have to be imported. The development of fodder-yeast production has obtained its major stimulus from local availability of large amounts of so-called sulfite waste liquor of paper mills. In many countries, severe restrictions have been imposed on the paper mills in regard to dumping such waste liquors in rivers and streams, and several large paper mills went into feed-yeast production primarily to reduce the biochemical oxygen demand of their waste liquor. Fodder-yeast production at present predominates in the eastern European countries such as the USSR, East Germany, Poland, and Czechoslovakia, although West Germany, France, Italy, the United States, Cuba, Taiwan, and Japan are also fairly large producers. It is estimated that worldwide approximately u)O,OOO tons of dried feed yeast is produced, primarily from sulfite waste liquors or wood hydrolyzates and some from molasses or whey. Strains of Cundida utilis are usually employed since this species grows fast, is not fastidious as to growth conditions, and is able to utilize the pentose sugars present in large amounts in such waste liquors and wood hydrolyzates. In Japan Cundidu yeast is first extracted in order to remove its ribonucleic acid content, leaving practically all protein intact. Ribonucleic acid is used as a source of two important flavoring agents, the nucleotides guanylic monophosphate ( GMP) and inosinic monophosphate (IMP). The remaining protein-rich residue is then sold as high-protein feed. This makes a feed yeast production process far more attractive. The production of the protein gluten from wheat leaves starch as a by-product for which markets are not readily available. There is a possibility of fermenting the starch through a modification of the Swedish Symba Process (Tveit, 1967). This process involves two organisms, Endomycosis fibuliger and Candidu utilis; the former organism is high in a- and @-amylaseactivity and converts waste starch materials into fermentable sugars, which are then utilized by the Cundidu to give fodder yeast. There is a species of yeast-S. diastuticus-that could be used instead of the above two species. It produces an extracellular glucoamylase that would hydrolyze the starch to fermentable sugar which would in turn be used as a carbon source to produce yeast protein. Another waste material that could be used as a substrate for yeast is whey. Large amounts of this material are produced in cheesemaking which could be fermented with the lactose-fermenting yeast S . frugilis,
258
G . G . STEWART
and thereby yield a good supply of yeast protein. A number of detailed studies have been done on this process, but although the disposal of whey is a serious problem in many countries (the United States and Canada being countries with a serious problem), there are only a few systems in commercial operation. Although it has been known for many years that certain yeast species could metabolize hydrocarbons, it was not until the early 1950s that special attention was paid to the possibility of producing yeast protein from hydrocarbon fractions. British Petroleum (in Britain) and the Centre National de Recherche Scientifique (in France) then began a scientific investigation on the possibility of large-scale production of yeasts or petroleum fractions in the late 1950’s, and in the early 1960’s laboratories in the United States, Europe, and Japan began similar research, not only restricting their studies to yeasts, but also aiming at the production of bacterial protein from hydrocarbons, including methane, the major hydrocarbon present in natural gas. It is beyond the scope of this review to discuss in detail the facets of petroleum fermentation, a substrate practicalIy insoluble in water and with an energy potential far greater than that of the usual carbohydrate energy sources. It suffices to mention that it is possible to produce protein easily and in good yield using a fermentation process wherein hydrocarbons of various types are the source of carbon and of energy for the microorganism. Typical fermentation industries have in general stayed aloof of yeast production from hydrocarbons. However, several oil companies, such as Esso of New Jersey, Gulf Oil, Sun Oil, and Humble Oil in the United States, are actively engaged in such research. British Petroleum is probably the furthest advanced and has built two fermentation plants, one at its Gangemouth (Scotland) refinery, where n-paraffins ( Clo-Cls) are being used as the substrate, and the other at LavBra, France, where gas oil (piped from North Africa) is being used. The total production from the two plants will be of the order of 300,000 tons of protein per annum (Laine, 1971). To surmount the problem of harvesting the yeast, new machinery has had to be developed because nothing on the scale required was commercially available. For example, enlarged centrifuges were developed which resulted in food recovery of the yeast from the nonmetabolizable hydrocarbon. The yeast protein produced by British Petroleum has been subjected to acute toxicity tests with a variety of experimental animals. The results to date have all proved to be negative and are therefore encouraging. Permission has been obtained in the past two years for the use of this protein for animal food and some of it is now on the market for cattle and poultry feed.
BREWING AND OTHER INDUSTRIAL USES OF YEASTS
259
The use of yeast protein as a food for man is also receiving much attention at the present time. Our present world population is approximately 3.5 billion, 70% located in the developing countries (Johnston, 1971). It is estimated that in the next 30 years the world population will nearly double what it is today. In highly industrialized countries this increase will be restricted to approximately 10%per 10 years, but in the Asiatic, Central and South American countries, which already have grave difficulty in supporting their population, the increase over the same period will be of the order of 2030%. This all amounts to the fact that in 30 years food production must be at least doubled, and it is by no means certain that agriculture, together with breeding and fisheries will reach this goal. However, yeast protein could supply a large part of the approximately 50% of the daily intake of protein required to maintain good health. The use of yeast protein in any quantity in the human diet has one serious drawback. It has long been recognized (Funk et al., 1916; Wintz, 1916) that, owing to the high content of nucleic acid, a diet of yeast can increase urinary uric acid to serious levels. Man and higher apes lack the enzyme uricase, which catalyzes the oxidation of uric acid to the more soluble allantoin, so that in individuals with a genetic tendency to primary overproduction of uric acid, there may be precipitation of uric acid crystals in joints (gout) and soft tissues (trophi) or the formation of stones in the urinary tract. The increase in uric acid, secondary to a high intake of dietary purines and other erogenous factors, will probably have similar effects. It follows, therefore, that, if all protein from yeast is to be used as a primary protein source for human consumption, the nucleic acid content will have to be reduced to a safe level. In a recent dietary study (Edozian et at., 1970) involving feeding Candidu utilis to humans, it has been found that the maximum intake must be in the range of 2 gm of nucleic acid per day. Methods for reducing the nucleic acid content of yeast have recently been published (Mateles and Tannenbaum, 1968; Maul et al., 1970; Canepa et al., 1972). Most of the methods involve heat treatment, with an initial heat shock phase at 90-100°C followed by incubation at 5040°C for approximately 15 minutes. Some of the methods also use exogenous ribonuclease along with the heat treatment. The most serious problem, which has not been solved as yet, is the acceptability of yeast as food. Experience has shown that few people consume food because of its nutritional value. It has to fulfill certain specifications of taste, structure, and color, and even then it is extremely difficult to adjust food habits. Much progress has already been made in giving vegetable proteins a structure and taste which imitates the taste and structure of the expensive animal proteins. This problem is, however,
260
G. C. STEWART
far from being solved, and up to the present most attention has been focused on soya protein instead of yeast protein as starting material. A closer cooperation between the yeast manufacturer and the food technologist will be required to overcome this serious obstacle of poor acceptability. Properly prepared yeast autolyzates have already found an extensive use as an ingredient of dry soups and other precooked foods as a substitute for meat extract (Peppler, 1970). This is only an interesting side-track, however, and the main attention should be directed to transforming yeast into an acceptable food product of such a taste and structure that it can replace part of the animal proteins and could become an acceptable and inexpensive protein adjunct. XIV.
Biochemicals from Yeast
Substances that can be obtained from yeast can be divided into several groups : enzymes, coenzymes, cellular components, excretion products, and chemical substances produced by special reactions ( Harrison, 1968). A further important division of the various compounds formed by the action of microorganisms is into bulk (cheap) and low-yielding (expensive) products. The substances fall naturally into these two categories: examples of the former are ethanol and whole cells, and the latter includes purified enzymes and vitamins. Antibiotics also fall into this latter class, but no commercial use of yeast for this purpose has been demonstrated. A number or enzymes are obtained from yeast, invertase, amylase, lactase, alcohol dehydrogenase, glucose-6-phosphate dehydrogenase, glutathione reductase, and hexokinase, to mention only a few. One of the commonest reactions controlled by enzymes is hydrolysis. A practical example, of this is hydrolysis by invertase, or p-fructofuranosidase, of sucrose to its component hexose moieties, fructose and glucose, thus inverting the rotation of the solution from dextro to levo. Invertase occurs in many yeast strains, including S. cereuisiae or S . carlsbergensis, and is closely associated in the cell wall with the polysaccharide mannan. Commercial preparations may be mads by growing the yeast aerobically on a sucrose (molasses) medium to obtain an active material and separating the yeast, which is plasmolyzed by treatment with toluene and digested with papain, and the solids are removed. The chilled solution is adjusted to pH 4.5, the enzyme is precipitated by addition of ethanol, separated, and taken up again in water. The activity is protected by adding glycerol to a concentration of about 55%. Alternatively, a freezedried powdered preparation can be made. One of the main uses of invertase is in the manufacture of soft-centered candy (fondant); the centers are made in a cold state with a sucrose and invertase mixture, which is hard when first coated but later becomes soft when glucose
BREWING AND OTHER INDUSTRIAL USES OF YEASTS
261
and fructose, which are more soluble than sucrose, are formed by the action of the enzyme. Another enzyme made by yeast is lactase, or p-D-galactosidase, which is formed when S. fragilis is grown in whey with aeration and a limited supply of assimilable nitrogen (Young and Healey, 1957). The yeast is dried in such a way, e.g., by spray-drying, that the fermentative activity is destroyed while the lactase remains effective (Meyers and Stimpson, 1956)) and sucrose may be incorporated in the product to ensure high stability. Lactase preparations, which are usually only partially purified, are employed primarily to hydrolyze lactase in milk products for ice cream bases, frozen milk, and animal feeds (Pomeranz, 1964). Coenzymes comprise another important group of compounds that can be obtained from yeast. Although quite a number of these substances are grown and can be prepared in reasonable quantities by modern techniques, they have found little commercial use, as their function is specific to the intercellular reactions in which they take part. Furthermore, in spite of their comparatively simple chemical composition, they cannot normally be assimilated into living cells, and therefore have little application, whatever apparent advantages they may possess in specific metabolic disorders. The nucleic acids in baker’s and brewer’s yeasts amount to about 8% of the dry weight. They can be extracted by alkali, followed by immediate neutralization and filtration, and precipitated by acid from the cooled filtrate. After washing, the solids are again dissolved and reprecipitated, and may be left as the free acids or converted to the copper, iron, or magnesium salts (Pyke, 1958). Various clinical applications have been suggested, but none has found wide use whereas, by mild alkaline hydrolysis with 1 N sodium hydroxide at 37°C) individual nucleic acids can be obtained; more severe treatment results in the production of the simpler purines and pyrimidines. The derivatives can be separated as described above, but this puts them into a high cost market, and they are not used in great quantity. Disodium guanylate, disodium inosinate, and inosinic acid are also produced in some quantity in Japan from Candida yeast (Peppler, 1967). Lipids normally occupy only about 3% of the dry weight of yeast, but certain species, notably Rhodotorula gracilis and Torulopsis lipofere, when grown under special conditions of nitrogen and phosphate feed, may contain as much as 70% (Prescott and Dunn, 1940). This property has been used, particularly under war conditions when animal and vegetable raw materials were difficult to obtain as a commercial source of fats. A particularly useful lipid is ergosterol, which can be produced by dehydrating yeast and extracting with ethanol or a mixture of lower alcohols and saponifying the sterol esters in the extract ( Petzoldt, 1967).
262
G. G. STEWART
Yeasts contain a number of growth factors. These include thiamine, riboflavin, niacin, biotin, pantothenic acid, pyridoxine, inositol, folic acid, p-aminobenzoic acid, and choline. The first three are necessary for human nutrition, and the remainder are implicated in animal and microbial metabolism. Thiamine, or vitamin B1,the classical preventive of beriberi, consists of two moieties 2-methyl-4-amino-5-hydroxymethylpyrimidine and 4-methyl-5-( 2-hydroethyl )thiazole, which are normally produced with the yeast cell in the amounts required for metabolic purposes. If the two halves of the complete molecule are supplied to growing yeasts as solutions of the synthetic parts, the intercellular enzymes complete the synthesis of the vitamin; in this way much higher concentrations can be produced (as much as 1 mg of thiamine per gram of dry matter). The other growth factors occurring in yeast are not usually isolated, but are used in the crude form mixed with other cell components (Hansen, 1967). REFERENCES Auk, R. G., Hampton, A. N., Newton, R., and Roberts, R. H. (1969). 1. Inst. Brew., London 75, 260-277. Barbet, E. (1905). British Patent 16, 233. Barnett, J, A. (1968). In “The Fungi” ( G . C. Ainsworth and A. S. Sussman, eds.), Vol. 3, pp. 557-595. Academic Press, New York. Burrows, S. (1970). In “The Yeasts” (by A. H. Rose and J. S. Hamson, eds.), Vol. 3, pp. 349-420. Academic Press, New York. Canepa, A., Pieber, M., Romero, C., and Lohe, J. C. (1972). BiotechnoL Bioeng. 14, 173-199. Clayton, E., Howard, G. A., and Martin, P. A. (1972). Amer. SOC. Brew. Chem. Conu. pp. 78-81. Coutts, M. W. (1957). British Patents 872,391-879,400. Dalgliesh, C. E. (1972), Brew. Trade Reu. 87, 22. Edozian J. C., Udo, U. U., Young, V. R., and Schrimshaw, N. S. (1970). Nature (London) 228, 181. Emeis, C. C. (1965). Monutschr. Brau. 18, 224-228. Enebo, L., Johnsson, E., Nordstrom, K., and Muller, A. (1960). Su. Bryggeri tidskr. 12, 270-275. Findlay, W. P. K., ed. (1971). “Modem Brewing Technology.” Macmillan, New York. Fisher, E. ( 1907). Sitzungsber. Preuss. Akad. Wiss., Phys.-Math. K1. pp. 35-38. Fowell, R. R. (1966). Process Biochem. I, 25-28. Fowell, R. R. ( 1969). In “The Yeasts” (A. H. Rose and J. S. Harrison, eds.), Vol. 1, pp. 303-348. Academic Press, New York. Funk, C., Lyle, W. G., and McCaskey, D. ( 1916). J . Bwl. Chem.27,173-186. Gilliland, R. B. ( 1962). J. Inst. Brew. London 68,271-275. Gilliland, R. B. ( 1971). In “Modem Brewing Technology” (W. P. K. Findlay, ed.), pp. 108-162. Macmillan, New York. Haboucha, J., Jenard, H., Devreux, A., and Masschelein, C. A. (1969). Eur. Brew. Conu., Proc. Congr. 12, 241-251.
BREWING AND OTHER INDUSTRIAL USES OF YEASTS
263
Hansen, A. M. ( 1967). In “Microbiological Technology,” pp. 222-255. Van NostrandReinhold, Princeton, New Jersey. Harris, J. O., and Watson, W. (1971). Eur. Brew. Conu., Proc. Congr. 13, 273-286. Harrison, J. S. ( 1968). Process Biochem. 3,59-62. Harrison, J. S., and Graham, J. C. J. (1970). In “The Yeasts” (A. H. Rose and J. S. Hamson), Vol. 3, pp. 283-348. Academic Press, New York. Hoogaheide, J. C. (1969). Antonie uan Leeuwenhoek; I. Microbiol. Smol. 35, Suppl., 71-82. Hough, J. S. (1959a). J . Inst. Brew. London 65, 479-482. Hough, J. S. (1959b). Brew. Dig. 34, 39-44. Hough, J. S. (1961). Eur. Brew. Conu., Proc. Congr. 8,160-179. Hough, J. S., and Ricketts, R. W. (1960). J . Inst. Brew., London 66,301-304. Hough, J. S., and Rudin, D. (1958). J . Inst. Brew., London 64, 301-312. Hough, J. S., and Rudin, D. (1959). Eur. Brek. Conu., Proc. Congr. 7 , 208-216. Hough, J. S., Briggs, D. E., and Stevens, R. (1971). “Malting and Brewing Science.” Chapman, & Hall, London. Johnston, J. R. (1963). Eur. Brew. Conu., Proc. Congr. 9, 412-421. Johnston, M. J. (1971). Proc. Int. Congr. Microbiol. loth, 1971 p. 86. Kihlberg, R. (1972). Annu. Reu. Microbiol. 26, 427466. Kleyn, J., and Hough, J. S. (1971). Annu. Reu. Microbiol. 25,583-608. Klopper, W. J., Roberts, R. H., Royston, R. G., and Auk, R. G. (1965). Eur. Brew. Conv., Proc. Congr. 10, 238-259. Kochovi-Kratochvkova, A., Stuchlik, V., and Toniasek, K. ( 1965). Brauwiss 18, 328-336. Kockov6-Kratochvkova, Sandula, J., Vojtkovii, M., Pokorna, A., and Stuchlik, V. ( 1967). Folk Microbiol. (Prague) 12, 327-341. Laine, B. (1971). Proc. Int. Congr. Microbiol., loth, 1970 pp. 82-84. Lewis, M. J. (1968). Tech. Quart., Master Brew. Ass. Amer, 5, 103-113. Lodder, J. ( 1970). “The Yeasts,” 2nd ed. North-Holland Publ., Amsterdam. Lodder, J., and Kreger-van Rij, N. J. W. (1952). “The Yeasts,” 1st ed. North-Holland Publ., Amsterdam. Mateles, R. I., and Tannenbaum, S. R. (1968). “Single-Cell Protein.” MIT Press, Cambridge, Massachusetts. Maul, S. B., Simskey, A. J., and Tannenbaum, S. R. (1970). Nature (London) 228, 181. Maxon, W. D. (1955). Appl. Microbiol. 3, 110-117. Meyers, R. P., and Stimpson, E. G. (1956). U.S. Patent 2,762,749. Monod, J. (1950). Annl. Inst. Pasteur, Paris 79, 390-395. Pasteur, L. (1866). “Etudes sur le Vin,” pp. 264-266. Imprimeuss Impkrials, Paris. Pasteur, L. (1876). “Etudes sur la Bithe,” pp. 383-385. Gauthier-Villars, Pans. Peppler, A. H. ( 1967). In “Microbiological Technology,” pp. 161-189. Van NostrandReinhold, Princeton, New Jersey. Peppler, H. J. (1970). In “The Yeasts” (A. H. Rose and J. S. Harrison, eds.), Vol. 3, pp. 421-462. Academic Press, New York. Petzoldt, K. (1967). West German Patent 1,252,674. Pomeranz, Y. (1964). Food Technol. 18,682-686. Poncet, S . (1967). Antonie uan Leeuwenhoek; J . Microbiol. Serol. 33, 347-351. Portno, A. D. (1967). J . Inst. Brew., London 73, 4 3 5 0 . Portno, A. D. (1969). J . Inst. Brew., London 75,468-471. Portno, A. D. (1973). Brew. Dig. 102,33-35. Prescott, S. C., and Dunn, C. G. (1940). “Industrial Microbiology,” pp. 169-182. McGraw-Hill, New York.
264
G . G . STEWART
Pyke, M. (1958). In “The Chemistry and Biology of Yeasts” (A. H. Cook, ed.), pp. 535-586. Academic Press, New York. Reed, G., and Peppler, H. J. (1973). “Yeast Technology.” Avi Publ. Co., Westport, Connecticut. Ricketts, R. W. ( 1971). In “Modern Brewing Technology” (W. P. K. Findlay, ed.), pp. 83-112. Macmillan, New York. Righelato, R. C., and Elsworth, R. (1970). Aduan. Appl. Microbiol. 13, 399-465. Rose, A. H., and Harrison, J. S., eds. (1970). “The Yeasts,” Vol. 3. Academic Press, New York. Rosen, K. ( 1968). Process Biochem. 3, 45-47. Shardlow, P. J. (1972). Tech. Quart., Master Brew. Ass. Amer. 9, 1-5. Sher, H. N. ( 1960). British Patent 845,315. Sher, H. N. (1961). Sci. ( SOC. Chem. Ind., London) Monog. 12,94-108. Sneath, P. H. A. (1957). J . Gen. Microbiol. 17, 184-196. Stark, W. H. (1954). In “Industrial Fermentation” (L. A. Unkenkofler and R. J. Hickey, eds.), Vol. 1, pp. 235-246. Academic Press, New York. Thompson, C. C. (1970). J. I&. Brew., London 76,423. Thorne, R. S. W. ( 1957). “Yeasts,” pp. 79-126. Junk, The Hague. Tveit, M. (1967). In “Biology and the Manufacturing Industries” ( M . Brook, ed.), pp. 155171. Academic Press, New York. Van Rijn, L. A. ( 1906). British Patent 18,045. Wakerbauer, K. (1969). Eur. Brew. Conu., Proc. Congr. 12, 523-537. Watson, T. G . and Hough, J. S. (1969). J. Inst. Brew., London 75, 359-363. Windisch, S., and Emeis, C. C. (1969). West German Patent 1,442,311. Wintz, H. ( 1916). Muenchen. Med. Wochenschr. 63, 454-461. Young, H., and Healey, R. P. (1957). U.S. Patent 2,776,928.
linear Alkylbenzene Sulfonate: Biodegradation and Aquatic Interactions
WILLIAME. GLEDHILL Monsanto Company, S t . Louis, Missouri
I. Introduction ................... 11. LAS Biodegradation Studies ............ A. LAS Structure ................ B. Terminology ................. C. Analytical Methods .............. D. Biodegradation Test Procedures ......... E. Molecular Structure of LAS vs Biodegradability .. F. Preferential Isomer Degradation ......... G. Laboratory vs Field Studies .......... H. Salt Water Environments ........... I. Anaerobic Systems .............. J. Metabolism-Mixed Cultures .......... K. Metabolism-Pure Cultures ........... 111. LAS and the Aquatic Environment ......... A. Acute LAS Toxicity .............. B. LAS Toxicity vs Biodegradation . . . . . . . . . . . . . . . . C. Chronic LAS Toxicity ............. IV. Summary and Conclusions ............. References ....................
I.
265 266 266 266 267 269 270 272 273 275 276 276 277 281 281 282 284 286 289
introduction
During the 1950's alkylbenzene sulfonates ( ABS ) became the major surfactant material employed in detergent formulations. The branched nature of the ABS side chain, however, rendered the molecule fairly resistant to bacterial degradation under ordinary sewage treatment conditions. Consequently, the foaming property of ABS remained upon discharge from sewage plants into natural waters. This adverse environmental effect prompted development of more readily biodegradable compounds, such as ABS molecules with linear side chains. During the early 1960's linear alkylbenzene sulfonates ( LAS ) gradually replaced their branched-chain ABS predecessors. Since that time an immense volume of literature has accumulated supporting the environmental acceptability of LAS products. The following review article is presented in order to consolidate the current knowledge of LAS from the biodegradation and aquatic standpoints and to stimulate interest in specific areas for which additional research would provide a challenge to several disciplines. 265
WILLIAM E. GLEDHILL
266
II.
LAS Biodegradation Studies
A. LAS STRUCTURE Linear alkylbenzene sulfonates are prepared by sulfonation of linear alkylbenzenes. Commercial linear alkylbenzene products are currently manufactured via alkylation of benzene with either chlorinated linear paraffins or linear olefins. In the chlorination process, all possible monochloro isomers of the starting paraffin are found in the intermediate chloroparaffins. Alkylation of benzene with the chloroparaffin intermediate yields a mixture of primarily the 2-7 phenyl isomers plus relatively small quantities of the l-phenyl isomers. With mono olefins as the starting material, alkylation of benzene yields all possible secondary isomers regardIess of the double-bond position. l-Phenyl isomers are not formed in the olefin process. Since the initial paraffin or olefin material usually consists of a mixture of homologs, such as Clo-ls, Cll-,4,or C12-15,the resultant number of possible alkylbenzene isomers and homologs becomes large. With alkylbenzenes as the starting material, alkylbenzene sulfonates are prepared by reaction with oleum or sulfur trioxide. The sulfonic acid which results is generally neutralized to yield the sodium salt along with more or less Na,SO,. The sulfonation occurs primarily at the para position of the benzene ring; however, minor amounts of the ortho isomer may be formed depending on the configuration of the alkyl side chain. The final LAS product may also contain small amounts of various other minor components, such as dialkyl indane and dialkyl tetralin species which result from cyclization of the alkyl side chain during the alkylation step. Also branched alkylbenzenes from nonlinear impurities in the initial paraffin or olefin may be present. Thus, LAS prepared by different manufacturers will vary somewhat in composition depending upon the nature of the starting material and the chemical process involved. The complexity of this mixture must be kept in mind when environmental problems are being examined.
B. TERMINOLOGY The term biodegradation, the conversion of chemical compounds to simpler substances through the action of microorganisms, has been employed in several contexts ( Standard Method Committee-Subcommittee on Biodegradation, 1967). Primary biodegradation is the minimum extent necessary to change the identity of a compound. The loss of a positive
LAS: ENVIRONMENTAL BEHAVIOR
267
methylene blue test in LAS biodegradation exemplifies this use. Clearly, the molecule has been changed; however, biodegradation may have proceeded only to a limited extent. The loss of the ability to produce foam is another example of primary biodegradation. Here again foaming properties may be lost, but biodegradation may not be complete. Ultimate biodegradation, the complete conversion of a compound to C02, H,O, and other inorganic compounds (mineralization), is of importance today in surfactant technology, and in the future it will become increasingly more so. During sewage treatment complete mineralization of organic materials is not usually achieved. Rather, conversion of a compound to primarily COP, H 2 0 , and microbial cells should be expected so that an innocuous nontoxic, low biological oxygen demand effluent leaves the treatment plant.
C. ANALYTICAL METHODS The following is a brief summary of some of the more common analytical techniques employed for assessing the degree of biodegradation of LAS. A more complete list of methods is discussed in detail by Swisher ( 1970). 1. Methylene Blue Analysis ( Methylene Blue Active SubstuncesMBAS) Perhaps the most widely employed routine assay for assessing LAS biodegradation has been the methylene blue procedure (Longwell and Maniece, 1955). An anionic surfactant-methylene blue salt is readily extracted into organic solvents making colorimetric assay possible. The test is quite easy to run, but suffers from the drawback that it may only measure initial (or primary ) biodegradation, Oxidation of the terminal methyl group of the alkyl side chain to the carboxyl group or removal of the sulfonate group from the benzene ring (either of which have been reported to occur early in LAS biodegradation) may eliminate the methylene blue reaction. Consequently, using uniformly labeled 14Cdodecylbenzene sulfonate, it was found that when 90% MBAS was gone, 75%14Cwas still present (Allred et al., 1964).
2. Foam Generation The loss of the ability to produce foam has been employed as a qualitative measurement of biodegradation ( Bacon, 1966). Foaming potential may also be lost with oxidation and minor reduction in the alkyl chain length of the LAS molecules, so here again its loss does not necessarily indicate anything beyond primary biodegradation.
268
WILLIAM E. GLEDHILL
3. Biological Oxygen Demand ( B O D ) , Chemical Oxygen Demand ( C O D ) ,RespiTometTy, and Organic Carbon Measurement Both biological and chemical oxygen demand tests have been widely employed for estimating the extent of biodegradation. These, however, suffer from a number of drawbacks (Sheets and Malaney, 1956; Janicke, 1968; Swisher, 1970) including poor sensitivity, interferences, length of time required for testing, and various other problems. Measurements of oxygen uptake via Warburg respirometry (Hunter and Henkelekian, 1964)) and modifications, are more precise than the BOD method and, if properly adapted cultures are employed, can be quite informative about the extent of biodegradation and metabolic pathways employed for biodegradation. Measurements of total organic carbon ( Busch, 1966) by use of commercially available carbon analyzers has recently become more prevalent. The inorganic carbon values ( C0,-carbonates ) are generally subtracted from the total carbon content to give the organic carbon. Inorganic carbon may also be removed by acidification and blowing with C0,-free air or by barium precipitation before analysis. Biodegradation can, consequently, be followed by the disappearance of dissolved organic carbon.
4. CO, Evolution
CO, evolution from a substrate can give information similar to 0, uptake for assessment of biodegradation (Hunter and Henkelekian, 1964; Brink and Meyers, 1966). In addition, complete release of theoretical amounts of CO, from a substrate and from the bacterial cells that have developed on the substrate is an indication of complete or ultimate biodegradation (mineralization) ( Ludzack et al., 1959; Thompson and Duthie, 1968; Sturm, 1973). The procedure is quite adaptable for measuring biodegradation of 14C-labeled substrates in natural environments.
5. 9-Labeled LAS 35S-Labeled LAS has been employed in studies where conventional extraction methods are unsatisfactory for the recovery of U S . Thus, the recovery of 'SO, from natural soils, sewage, anaerobic digesters, septic tanks, etc., has proved a sensitive tool to measure the extent of surfactant degradation in complex systems. The uncertainty of measurements of the complete biodegradation is also inherent with this assay procedure since the sulfonate group has been reported to be removed from the benzene ring both prior to and at the same time as ring cleavage (Cain et al., 1971; Oba, 1971).
LAS: ENVIRONMENTAL BEHAVIOR
269
6. Ultraviolet ( U V ) and Infrared ( I R ) Analysis LAS molecules have both characteristic UV absorbance and infrared patterns. Since ring degradation occurs just prior to complete mineralization of LAS in aerobic systems, UV monitoring of LAS ring biodegradation has been successfully employed to measure secondary or terminal LAS degradation (Setzkorn et al., 1964; Swisher, 1967, 1972a). IR measurement has also been used to follow LAS disappearance; however, with undefined systems many interfering compounds are usually present.
7. Specific Measurement of Substrate Desulfonation gas chromatography allows individual isomers and homologs of LAS to be followed during biodegradation (Huddleston and Allred, 1963; Swisher, 1963a) and intermediate degradation products to be detected (Swisher, 196313). The method has also been vital in elucidation of the metabolic pathways of LAS biodegradation (Cain et al., 1971) . D. BIODEGRADATION TESTPROCEDURES
A variety of laboratory tests have been employed to measure the biodegradation of a test compound. These range from a dilute river or lake water die-away test to tests employing more concentrated numbers of microorganisms. The standard procedures are described in more detail by Swisher ( 1970) and by Huddleston and Allred ( 1970). 1. River or Lake Water Die-Away Test This procedure employs a natural water source (river, lake, ocean) to which low levels of LAS are added ( 1-10 ppm). The procedure may require an extended period of time because of the low numbers of organisms present and the time required for acclimatization. More rapid procedures, involving the supplementation of the natural water with washed, acclimatized cells, have been reported ( Hitzman, 1964).
2. Shake Flask Test The shake flask test employs a standard inorganic salts medium to which 50-300 pprn yeast extract are added (Soap and Detergent Association, 1965). Surfactant is generally added at 30 ppm and the units receive an acclimated inoculum from soil, sewage, etc. In a comparison of the river water die-away, shake flask, semicontinuous, and continuous activated sludge tests, Orgel and Rupp (1967) found the shake flask method to be more reproducible.
270
WILLIAM E. GLEDHILL
3. Actiuated Sludge Test Of the standard tests, the heaviest microbial population is present in the activated sludge test. Batch, semicontinuous (24 hr fill-draw) and continuous laboratory tests ( German Government, 1962; Swisher et al., 1964) are employed routinely in assessing LAS biodegradation. These tests most closely represent the biodegradation capabilities of the sewage treatment plants. 4. Miscellaneous
Various systems, septic tanks, cesspools, anaerobic digesters, are employed for studies of anaerobic biodegradation. Trickling filters ( sewage running over an aerated, porous rock bed) and oxidation ditches are also employed in aerobic biodegradation studies. Numerous variations in the test units, media, and inocula make comparison of data from laboratory to laboratory difficult. Substitution of what would appear to be a minor difference, such as dried meat extract instead of paste, can lead to great differences in the apparent degree of biodegradation of LAS compounds (Vaicum, 1969). In addition, the organisms in the systems are quite sensitive to temperature (Halvorson and Ishaque, 1969) and to substrate concentration ( Burnop and Bunker, 1960) I
E. MOLECULAR STRUCTWXE OF U S vs BIODEGRADABILITY Throughout the remainder of this report references will be made to specific examples of LAS biodegradation. The reader must keep in mind that these are not constant values, but are those found under a given set of conditions. The authors may not have employed fully acclimated systems. Also, the exact identity of the various LAS preparations and their purity is often uncertain. In addition, the average molecular weight (chain length) and percentage of various phenyl isomers varies from product to product. Such variations may explain the lower or higher degree of apparent biodegradation from laboratory to laboratory. Since their introduction, numerous studies have attested to the superiority in biodegradation of the linear alkylbenzene sulfonates compared to their branched alkyl chain counterparts. Extensive tables on the extent of biodegradation of individual LAS isomers and homologs and commercial LAS products have been presented by Swisher (1970). A variety of biological systems ranging from septic tanks to activated sludge units have demonstrated the greater percentage of degradation of LAS products than of branched-chain ABS products (Pitter, 1967). Activated sludge units were shown to remove 95%LAS (MBAS) and only 4560% branched chain ABS ( Pitter, 1967). Rotary trickling filters demonstrated
LAS : ENVIRONMENTAL BEHAVIOR
271
the ratio of surfactant (10 ppm) removal to BOD removal was 0.4-0.5 for branched-chain ABS and 1.0 for LAS (Renn, 1965). Branched-chain ABS ( 35S-labeled) undergoing 70% primary biodegradation ( MBAS) in activated sludge units were shown to release only 12%of the label as T O 4 ; whereas, the LAS product in this study underwent 98%primary biodegradation and released 3050% of its sulfur (Sweeney and Foote, 1964). Aerobic biodegradation of alkylbenzene sulfonates is dependent on the configuration of both the alkyl and benzene species of the molecule. Alkyl chain branching inhibits biodegradation ( Hammerton, 1955). Fujiwara et al. (1968), employing NMR and GC analysis of chain branching and its relationship to biodegradability, have shown that an inverse linear relationship is obtained when percentage of chain branching is plotted against percentage of biodegradation. Sawyer and Ryckman (1957) examining the effect of C, through C, carbon chains on primary ( linear ), secondary, and tertiary ABS molecules, reported both the primary and secondary alkyl linkages to be degraded completely with the release of SO,'- whereas, the tertiary alkyl linkages were resistant (incomplete oxidation, no SO,'- release). Substituents to the aromatic ring also alter the ease of biodegradation. Alexander and Lustigman ( 1966), examining simple benzene compounds, reported biodegradation to become progressively more difficult in going from benzoic acid to phenol to aniline to anisole to benzene-sulfonate to nitrobenzene. Biodegradation required 1, 1, 4, 8, 16, and more than 64 days, respectively. Substitution of a second sulfonate group at the ortho position of benzenesulfonate extended the time required for biodegradation from 16 to more than 64 days. Symons and DelValle-Rivera (1961) also reported sufonated benzene to be considerably more difficult to degrade than benzoic acid. Substitution of methyl, ethyl or hydroxy groups on the benzene ring of dodecylbenzene sulfonate resulted in the persistence of 844% MBAS at 21 days whereas, MBAS of unsubstituted dodecylbenzene sulfonate was completely gone at 8 days (Borstlap and Kortland, 1967). LAS biodegradation is dependent on alkyl chain length and position on the alkyl chain of the benzene sulfonate group. In general, biodegradation is facilitated by increasing the distance between the sulfonate group and the most remote methyl group of the alkyl chain. In regard to alkyl chain length, several studies employing pure LAS homologs have indicated the C, homolog to be most rapidly degraded, with alkyl chain lengths shorter or longer causing the molecule to be slightly more resistant to attack (Swisher, 1963a; Allred et al., 1964; Durand, 1965; Smith et al., 1966). Thus, studies employing the river water die-away test and pure LAS homologs have indicated the rate of biodegradation (loss of MBAS )
272
WILLIAM E. GLEDHILL
to increase as the alkyl chain length increases from LAS species with C, to C1, alkyl groups and to decrease with the C, to C, homologs, and to again increase, but not to the extent of CI2,between C,-C, (Swisher, 1963a; Durand, 1965). With mixtures of LAS homologs, such as C, through C, LAS, this result is not observed; rather the longer chain length homologs undergo more rapid biodegradation ( Swisher, 1963a). Tarring (1965) also reported the half-life of LAS in solution to decrease as alkyl chain length increased. With this LAS species, for alkyl chains of C&, the half-life was 3 days; CIo-13,2 days; C11-15,1.5 days; and C16-18, 1.25 days. Another study has indicated biodegradation of mixed isomers of pure C, to C, homologs of LAS to occur as long as levels for each homolog were below a certain limit concentration (Ciattoni and Scardigno, 1968). For example, in one river water the limit concentrations were determined as being 140, 55, 22, 8, 2.8, and 1.2 ppm for the Clo,C, C, C, C,*, and CIShomologs, respectively. With regard to the position of the sulfophenyl group, a longer acclimation period and slower rate of biodegradation are noticed as the sulfophenyl group is moved toward the center of the molecule (Huddleston and Allred, 1963; Swisher, 1963a, 1972; Allred et aZ., 1964; Setzkurn et al., 1964; Sweeney, 1964; Durand, 1965; Smith et al., 1966). These effects are generally observed only in dilute systems, where the internal, short chain-length isomers and homologs degrade more slowly ( Allred et al., 1964). However, in systems such as activated sludge, where heavier bacterial populations are present, the preferential rate of isomer and homolog biodegradation is not as evident ( Swisher, 1972a). Carboxylated 2- and 3-phenyl-substituted alkyl benzene sulfonates with side chains of P7 carbon atoms in length have been implicated as transient intermediates in river water systems (Swisher, 1963b), but no evidence of the accumulation of these intermediates in activated sludge systems was presented. C, LAS molecules sulfonated at the para-position of the benzene ring are degraded faster than those substituted at the ortho position (Swisher, 1963a). A single methyl side group at the near or far end of the alkyl chain was found to have little effect on the rate of biodegradation (Swisher, 1963a).
F. PREFERENTIAL ISOMER DEGRADATION In contrast to the work above (Swisher, 1972a) several studies with activated sludge units have indicated poor degradation of specific isomers and homologs of LAS. In one study, high molecular weight LAS products (MW 358) have been observed to be poorly degraded (Mann, 1969). Poor degradation of the higher molecular weight homologs may have been due to an inhibitory effect on the bacterial population (Swisher,
LAS : ENVIRONMENTAL BEHAVIOR
273
1963a) or to the use of poorly acclimated cells. Degradation of the various isomers and homologs of Marlon@A was investigated by gas chromatography in a sewage treatment plant consisting of both primary and secondary treatment facilities ( Kruger, 1966; Bock and Wickbold, 1967). The C, and C, isomers were degraded to the extent of 92-96% through the unit, with the exception of 6-sulfophenyl dodecane, which was only 81% degraded. Removal of the various C, and C, isomers ranged from 5249%. In this system, 5-sulfophenyl decane and 6-sulfophenyl undecane were the slower degrading isomers, being removed only to the extent of 52 and 58%. 5-Sulfophenyl decane was also the main residual component in the activated sludge biodegradation studies of Wickbold (1967). The low rate of degradation of these specific isomers, however, is usually not observed in acclimated systems (Swisher, 1968a).
G. LABORATORY vs FIELD STUDIES Laboratory experimentation is in agreement as to the improved ease of degradation of LAS over the branched-chain ABS products and the greater ease of the more terminally phenyl-substituted homologs vs the internal phenyl isomers. However, from the previous sections it can be noted that different degrees of biodegradation of various LAS isomers and homologs occur from one set of conditions to another. Commercial preparations consist of a random mixture of the various phenyl isomers of the C, to C , , homologs. Laboratory activated sludge units have yielded data on the biodegradation of commercial LAS products from the mid 80%to the upper 90%range. Ryckman and Sawyer (1957), by formation, and disappearance of measurement of oxygen uptake, benzene rings, have reported 2-sulfophenyl decane to be completely degraded in activated sludge. Eleven laboratories averaged 87%removal of Dobane JNX (MBAS) (Eden et al., 1968). Pitter (1963, 1964b)c) obtained 90-91% removal of MBAS for various synthetic products in activated sludge. The sulfonate groups were almost completely liberated as SO,2- in 20 days in BOD water. UV analysis, a sensitive indicator of benzene rings, indicated commercial C12 LAS and the 1-, 3-, and 6-phenyl dodecanesulfonates to undergo 85-92% ring degradation (Swisher, 1967). Although ring degradation can be seen to occur after alkyl side chain attack in dilute systems (Foster and Fields, 1964), the time differential between the two processes in activated sludge units is considerably shortened. Kruger ( 1964 ) reported 54% biodegradation of LAS benzene ring during a 3-hour retention in activated sludge. Complete disappearance of LAS has also been established by measurement of soluble organic carbon employing the carbon analyzer method devel-
274
WILLIAM E. GLEDHILL
*
oped by Janicke (1971). Swisher (1973) was able to obtain 95.7 1.7% carbon removal in 24 hour semicontinuous activated sludge units. In a standardized CO, evolution procedure, Sturm ( 1973) has reported LAS to release 65-70% of the theoretical carbon as CO,. Field studies, in general, support laboratory findings on the extent of LAS biodegradation. The study by Kriiger (1966) with Marlon@A was described above. Mann and Reid (1971) reported laboratory biodegradation values (MBAS) for 3 commercial LAS products of 71-91%, 88-98%, and 90-97% for shake flasks, semicontinuous, and continuous activated sludge tests, respectively. They found that these same three products underwent 8694%removal (MBAS) in a full-scale activated sludge treatment plant. With a final aeration stage after activated sludge treatment, the products were 94-97% degraded. Gas chromatographic and infrared analysis of the influent sewage, settled sewage, and sewage effluent from this plant revealed homologs of longer chain length, CIg to CI4, to decrease most rapidly (Mann and Reid, 1971). The C, to C , , homologs decreased at a slower rate. With regard to sulfophenyl position, the 2-sulfophenyl isomers of the C, to C , , homologs decreased most rapidly followed by the 3, and then the 4-7 phenyl isomers. The commercial products, consisting of a complex mixture of homologs and isomers, thus substantiated the laboratory data of Swisher ( 1963a). These commercial products also contained small amounts of branched chain material which was found to be degraded to the extent of 90%.In contrast, Simko et al. (1965) in field studies reported no preferential degradation of the various LAS isomers and homologs using mass spectroscopic analysis. His results were questioned (Swisher, 1970) since effluents from plants receiving only branched-chain ABS yielded mass spectroscopic data which appeared to indicate 3-phenyl LAS homologs to be present. Renn et al, (1964, 1965), examining LAS product biodegradation in an extended aeration activated sludge plant, reported LAS removal equal to or slightly better than the total BOD removal. In this plant study, an average 97.2%MBAS and 94.6%BOD removal was achieved. Detergent loads in sewage from the community had a range 2-4 times that which is normally encountered. Influents ranging from 10.5 to 37.5 ppm were degraded substantially, yielding effluent MBAS of 0.3 to 0.7 ppm. Spiking influent LAS streams to achieve 54 to 73 ppm MBAS resulted in the effluent concentration of MBAS being raised from 0.51 to only 1.36 ppm. Others have reported no significant change in percentage of LAS removal to occur with influent COD values being raised 195580 (McGauhey and Klein, 1967). Additional field studies indicated the change from branched-chain ABS to LAS products resulted in MBAS removal increasing from 3352%
U S : ENVIRONMENTAL BEHAVIOR
275
to 85-943 (Hanna et al., 1965). In this study wastewater temperature also appeared to affect the extent of biodegradation. Halvorson and Ishaque (1969) reported influent stream temperatures of 10°C to significantly inhibit biodegradation of C, LAS, whereas at 2OC no degradation occurred. These studies were based on unaerated sewage lagoons in Canada which were ice covered for approximately 5 months a year. Consequently, poor biodegradation was attributed to both low temperature and lack of oxygen. Also, other compounds, such as glucose, urea, acetate, palmitate, and casamino acids, which were readily degraded at 25OC showed poor or no biodegradation at 2OC. Such temperature problems are not encountered in U.S. activated-sludge plants. Significant biodegradation of LAS (21-34%) was also found to occur in sewage pipes before entrance into the sewage treatment plant (Spohn, 1964, 1967; Knapp and Morgan, 1965). Direct measurement with 35Slabeled LAS also confirmed 15%LAS biodegradation in a 4.17 mile sewer during a 170-minute retention time ( Standing Technical Committee on Synthetic Detergents, 1967). Spohn ( 1964, 1967) reported better LAS degradation in trickling filters than in activated sludge units (8593% vs 7040%). Poor degradation in these activated sludge units may have been due to poor acclimation or low residence time (high flow through). Low residence time in full-scale activated sludge plants has caused LAS degradation to decrease to as low as 48% (Kelly et aE., 1965). In contrast, Kumke and Renn (1966) reported only 801%MBAS removal in trickling filters. LAS biodegradation in oxidation ditches can be quite substantial. MBAS removal of 80-921% have been reported in these systems (Spohn, 1964, 1967; Huber, 1968). MBAS load and the length of contact period (residence time) are the most important factors in obtaining substantial degradation in oxidation ditches (Huber, 1968). In this system 2.1 gm of MBAS per kilogram dried solids per day was the maximum MBAS load which could be efficiently removed. Other field studies involving aeration type plants have reported 87-99% MBAS removal (Southgate and Eden, 1961; DeJong, 1964; Knapp and Morgan, 1965; Knopp et al., 1965a,b). H. SALT WATERENVIRONMENTS Biodegradation studies of LAS, as well as most industrial chemicals, in brackish or sea waters have been few. In the future, however, more emphasis will be placed on such studies. LAS has been found to undergo equivalent biodegradation in fresh, brackish, and sea waters (in excess of 90!%MBAS removal) (Mann, 1970; Bock and Mann, 1971). Examination at the mouths of several German rivers revealed highest LAS concentrations (0.3 mg/liter) only in the immediate vicinity of certain waste-
276
WILLIAM E. GLEDHILL
water outfalls (Mann, 1970; Bock and Mann, 1971). MBAS levels were below detection in other areas examined.
I. ANAEROBICSYSTEMS Thus far, data presented has been obtained from aerobic systems, activated sludge, trickling filters, shake flasks, oxidation ditches, etc., and indicates the relative ease of LAS biodegradation. However, a portion of the population is not served by aerated sewage systems, and must rely on systems, such as cesspools, septic tanks, where only low levels of dissolved oxygen, or none at all, may be present. In such systems accumulation of organic compounds may occur because of the absence of suitable electron acceptors for metabolic reactions ( 02,NO3-, SO,'), formation of microbial inhibitors (acid, aromatics), or the absence per se of oxygen for oxygenase-type reactions (Alexander, 1965). Initial oxidation of a hydrocarbon chain requires molecular oxygen. Consequently, under strict anaerobic systems LAS alkyl chain oxidation would not be expected. In addition, molecular oxygen is incorporated into the benzene ring in the primary mode of attack. Nonoxygen-requiring reactions, hydroxylations, have also been reported to occur with aromatic substrates (Allred et al., 1964; Taylor et al., 1970). These, however, have not been reported for aliphatic hydrocarbon chains. In anaerobic conditions we might, therefore, anticipate a minimum amount of LAS ring degradation, but little alkyl side-chain oxidation. This, however, has not been experimentally verified. Numerous experimental instances verify the low degree of LAS biodegradation in anaerobic systems (Straus, 1963; Wayman and Robertson, 1963; Pitter, 1964a,d, 1967; Klein and McGauhey, 1965; Maurer et al., 1965, 1971; Oba et al., 1967; Rismondo and Zilio-Grandi, 1968; Pitter et al., 1971). In those systems where a low degree of degradation did occur, a certain amount of dissolved oxygen may have been present or LAS adsorption onto solids was not accounted for. In any event, LAS has not been observed to inhibit digestion in anaerobic systems until abnormal levels are achieved (Pitter, 1964a,d, 1967; Maurer et al., 1965). However, septic tanks in combination with soil percolation fields are quite efficient in LAS biodegradation (90-97% MBAS removal) (Straus, 1963; McGauhey and Klein, 1964; Klein and McGauhey, 1965; Oba, 1965; Pitter, 1967). J.
METABOLISM-MIXEDCULTURES
Laboratory studies of mixed cultures can yield generalities about the metabolic pathways employed for biodegradation. However, uncontrol-
LAS: ENVIRONMENTAL BEHAVIOR
277
led mixed populations are usually unreliable for studying metabolic details (Prakasam and Dondero, 1970) due to the variety of organisms present and the array of metabolic pathways employed for substrate and intermediate product degradation. Oba (1971) isolated 7 different genera of microorganisms from a mixed culture (Micrococcus, Aerobacter, Fluuobacterium, Paracolobacterium, Alcaligenes, Pseudomonas, and Citrobacter ) capable of assimilating portions of, or all of, the LAS molecule. The Pseudomonas species was capable of carrying out 97%degradation (MBAS). A mixture of 3 cultures, Micrococcus, Aerobacter, and Pseudomonas, gave the maximum rate and extent of biodegradation. A mixture of 4 cultures was capable of removing 1831%MBAS under anaerobic conditions in 18 days. Huddleston and Allred (1963) examined the oxidation of 2-sulfophenyl dodecane by mixed cultures of bacteria and found, in this particular case, 13%of the original substrate to accumulate as sulfophenyl-substituted acids. Small amounts of sulfophenyl dodecanoic and sulfophenyl decanoic acids were present, but predominantly sulfophenyl octanoic and hexanoic acids were found. Since no odd chain length acids were found, the results substantiated the premise that the primary mode of alkyl chain degradation was via w-oxidation followed by p-oxidation. The findings of Swisher (1963b, 1968a,b, 1972a) supported these results, and in addition presented evidence for the occurrence of different pathways involved in ring degradation. Sulfate formation during LAS biodegradation was shown to parallel ring biodegradation in the latter study ( Swisher, 1972a).
K. METABOLISM-PURE CULTURES Simultaneous adaptation studies with Pseudomonas C,B provided insight into both the primary (alkyl) and secondary (ring) metabolism of even and odd alkyl chain length LAS molecules (Heyman and Molof, 1967) . Benzoic acid, p-toluene sulfonate, and p-pentylbenzene sulfonate adapted cells (and cell extracts) were found capable of immediate oxidation of benzoic acid, catechol, and p-ketoadipic acid, indicating metabolism of the side chain of odd carbon length chains via p-oxidation to benzoic acid, followed by either decarboxylation and ring oxidation to catechol, or to direct ring oxidation to protocatuate, and ortho cleavage with subsequent degradation to p-ketoadipid acid. Ethylbenzene sulfonate and p-dodecylbenzene sulfonate-adapted cells were capable of immediate oxidation of phenylacetic acid and homogentisic acid. C, LAS-adapted cells were found to be adapated to both the benzoic acid and phenylacetic pathways. An intermediate degradation product, p-sulfophenylmalonic acid, was postulated to be formed from the 2-,
278
WILLIAM E. GLEDHILL
4-, and 6-sulfophenyl isomers of C, LAS; whereas, 3-p-sulfophenylglutaric acid was the postulated intermediate from the 3- and 5-sulfophenyl isomers (Fig. 1).The p-sulfophenyl malonic acid was thought to degrade via the phenylacetic acid pathway to homogentisic acid. The 3-p-sulfophenylglutaric acid was postulated to undergo an a-oxidation and decarboxylation to yield p-sulfophenylpropionic acid, which in turn was metabolized through the benzoic acid pathway to catechol and p-ketoadipic acid. In contrast, C, LAS-adapted cells were found to be adapted only to the benzoic acid pathway. The intermediate 2-p-sulfophenyl succinic acid was postulated as being formed from all phenyl isomers of C, LAS. The 2-p-suIfophenyl succinic acid was metabolized via the benzoic acid pathway. Cain and Farr (1968) reported sulfonate was released as sulfite from the benzene ring of ,benzene sulfonate and p-toluene sulfonate. Sulfite was found to be converted to sulfate in the medium by nonenzymatic H HOOC-C-COOH C, LAS
2 - , 4 - , and 6Phenylisomers-
Q -co, Q CH,-
Phenylacetic acid pathway
so;
0
so,
HOOC- CH,-~H-CH,-COOH 3- and 5-Phenyl is0 me rs
a-oxidatio?
- co,
so,
Benzoic acid pathway
SO,
HOOC-CH--.CH,-COOH
- c)
CH,-COOH
I
so,
2-, 3 - , 4 - , 5 - , and 6-Phenyl isomers
,$ FH,-
I
C, LAS
COOH
-T
p - oxidation
-coz
-
Benzoic acid pathway
so,
FIG. 1. Metabolic pathways deduced from simultaneous adaptation studies of C, and CIz linear alkylbenzene sulfonate (LAS) niolecules ( Heyman and Molof, 1967).
LAS: ENVIRONMENTAL BEHAVIOR
279
oxidation. Benzene sulfonate was metabolized by a Pseudomonas species to catechol followed by either meta cleavage to 2-hydroxymuconic semialdehyde and subsequent degradation to formate, acetaldehyde, and pyruvate, or ortho cleavage to cis,cis-muconic acid which was further oxidized to p-ketoadipic acid. Paratoluene sulfonate was metabolized to 4-methylcatechol followed by meta cleavage to 2-hydroxy-5-methylmuconic semialdehyde which upon further degradation yielded propionate and pyruvate. Focht and Williams ( 1970), on the other hand, reported their Pseudomonas species to form 3-methylcatechol ( not 4-methylcatechol ) from p-toluene sulfonate, which subsequently degraded to acetate and pyruvate. 35S-Labeledp-toluene sulfonate indicated oxygen uptake and 35S0,2-release were constant during growth. No 35S aromatic intermediates were observed, indicating the sulfonate group to be removed prior to ring cleavage. Extensive studies of LAS metabolism have been presented by Cain et al. (1971) and Willetts and Cain (1972a,b). Four species ( 3 genera) of organisms were isolated and found to metabolize primary phenyl-substituted LAS molecules in different manners. A Vibrio species (Bird and Cain, 1972) was found t o metabolize 1-phenyldodecane to sulfophenylacetic and -butyric acids. 1-Sulfophenylundecane was oxidized to sulfophenyl propionate by this species. This isolate was incapable of carrying out sulfite release or ring cleavage. Bacillus sp. C, was capable of side-chain oxidation with immediate removal of sulfite from 1-phenylundecane and 1-phenyldodecane sulfonates. Ring attack was initiated after side-chain oxidation. Products from 1-phenylundecane sulfonate included p-hydroxyphenylvaleric, -propionic, and benzoic acids and protocatuate. The 1-phenyldodecane sulfonate yielded p-hydroxyphenylbutyric and -acetic acids. Further degradation is shown in Fig. 2. Bacillus sp. J 1 oxidized 1-sulfophenylundecane with sulfite release to p-hydroxyphenylhexanoic, -butyric, and -acetic acids and l-sulfophenyldodecane to p-hydroxyphenylvaleric, -propionic, and -benzoic acids. No ring oxidation was noted. It was concluded that an a-oxidative decarboxylation (1 carbon removal) must have occurred as the initial side-chain oxidation. Swisher (1963b, 1964) also presented evidence for the occurrence to a minor extent of such a process in LAS biodegradation in river water. A Cladosporium sp., C. resinae CM 188968, oxidized the substrates in the same manner as Bacillus C,. The initial products after sulfite removal, however, were not hydroxylated indicating sulfite removal to be reductive rather than hydrolytic. In another study, 2 Pseudomonas and 2 Nocardia were isolated (Treccani, 1971; Baggi et al., 1972). The 2 pseudomonads oxidized the ring
280
WILLIAM E. GLEDHILL
COOH
COOH
COOH
OH OH
COOH
OH
CO,
Succinate acetyl CoA
0 side
Phenyl LAS
?-=
CH,-
(CH,),-COOH
CH,COOH
CH,COOH
I
I
OH
OH
CH,COOH I
FIG. 2. Degradation of even and odd alkyl chain linear alkylbenzene sulfonate ( L A S ) molecules by Bacillus sp. C, (Cain et aZ., 1971; Willetts and Cain, 1972a).
portion of 2-phenylbutane, 3-phenylpentane, and 4-phenylheptane ( and the p-sulfo analogs) before the side chain. Meta cleavage yielding the hydroxymuconic semialdehydes was noted. The Nocardia oxidized 3-phenyldodecane to 2-phenylbutyric, -valeric, and -hexanoic acids. Ripen et al. (1971) also noted initial attack on the benzene ring with sulfate release ( not sulfite ) by Pseudomonas testosteroni growing on benzene sulfonate, p-toluene sulfonate, and p-ethylbenzene sulfonate. Pure cultures of bluegreen and green algae have been reported to oxidize the aliphatic chain and to release the sulfonate group of LAS molecules ( Davis and Gloyna, 1969). LAS intermediate degradation products with toxicity toward pure cultures of Pseudomonas have been postulated to be formed during degradation of LAS isomers under certain laboratory conditions (Payne, 1963; Payne and Feisal, 1963). Pseudomonas C,B was found to display good growth on the 2-, 3-, and 6-phenyl isomers of dodecylbenzene sulfonate. However, growth on the 4- and 5-phenyl isomers resulted in an inilia1 growth period followed by loss of cell viability. This phenomenon was not observed with the mixed isomer, dodecylbenzene sulfonate product; however, cultures displayed a long lag period before growth. The possibil-
LAS: ENVIRONMENTAL BEHAVIOR
281
ity of the formation of a toxic intermediate from the 4- and 5-phenyl isomers was presented; however, these experiments were conducted with extremely high LAS levels (7000 ppm). These toxic effects were not seen in mixed culture systems ( Setzkorn et al., 1964). Evidence for the anaerobic degradation of the benzene nucleus has also been presented (Taylor et al., 1970). Pseudomonas PN-1 was able to oxidize p-hydroxybenzoate with the release of 4 5 moles of CO, per mole of benzoate. Protocatuate was not an intermediate, but it was proposed that oxidation proceeded via hydroxylation at the 2 and 5 positions of p-hydroxybenzoate followed by cleavage to yield dihydroxypimelic acid. The existence of anaerobic mechanisms for LAS metabolism, however, have not been reported, but work of this nature does not preclude the possibility.
Ill.
LAS and the Aquatic Environment
A. ACUTE LAS TOXICITY Acute toxicity values indicate the potential of a chemical to affect an organism under a given set of laboratory conditions. With readily biodegradable compounds, such as LAS, these values are often irrelevant under natural environmental conditions where adsorptive and biodegradative processes are involved. Acute aquatic toxicity studies will be summarized briefly in order to develop a background for future studies concerned with LAS biodegradability and its relationships to aquatic toxicity. The most prominent manifestation of the acute toxicological effect of surfactants is on the gill tissue of fish. Destruction of the gill epithelium is regarded as a consequence of the reduction of surface tension by the presence of surfactants ( Bock, 1966). Consequently, most molecules with surfactant properties should elicit this effect. The toxic effects begin as the surface tension is reduced to 40-50 dyneslcm according to Bock (1967) and Kruger (1964). At surface tensions above this value gross toxicological effects are not apparent. Experimentally, it has been shown that exposure of fish to elevated levels of surfactants causes multiple hematomas to develop on the gill tissue resulting in diminished oxygen uptake and impairment of salt balance (Schmid and Mann, 1961, 1962). This, in turn, is followed by death. Acute aquatic toxicity studies have centered, primarily, on various fish species and on the establishment of that level of test compound which does not kill 50%of the population (median tolerance limit, TL,). Several acute studies have established 96-hr TL, values of various commercial LAS products to range from approximately 3-7 ppm for several adult fish, young fish, and eggs; somewhat more toxic than corresponding
282
WILLIAM E. GLEDHILL
values for branched-chain ABS material ( Hirsch, 1963a; Pickering, 1966; Thatcher and Santner, 1966; Wurtz-Arlet, 1967). Commercial LAS products of lower average molecular weight are slightly less toxic than products of higher molecular weight (Swisher et al., 1964; Thompson et al., 1971). Acute LAS toxicity is dependent primarily upon the homolog distribution, increasing with alkyl chain length (Hirsch, 1963a,b; Swisher et al., 1964; Borstlap, 1967; Marchetti, 1968; Thompson et d.,1971). To a lesser extent, phenyl sulfonate position on the side chain also affects acute toxicity ( Borstlap, 1967; Hirsch, 1963a,b). The 2-phenyl isomers of LAS molecules containing C,-C, side chains are slightly more toxic than the more internal phenyl isomers. The reverse appears true, however, for the 'Cl4 and C, LAS molecules. Isolated studies have also indicated both dissolved oxygen concentration and temperature to significantly alter the toxicity of LAS to fish. Marchetti (1968) noted a reduction in the 6-hour lethal concentration of C, LAS from 8.4 to 1.8 ppm for goldfish (Carassius) with a temperature rise from 15OC to 28°C. The lethal concentration for C, LAS went from 7.0 ppm to 0.1 ppm with the same temperature rise. The acute toxicity to bluegill fingerlings ( Lepornis) receiving LAS in Mississippi river water was found to increase 5-fold (48 hour TL, 2.2-0.4 ppm) as the dissolved oxygen was decreased (Hokanson and Smith, 1971). Both temperature and dissolved oxygen changes represent physiological stress to aquatic organisms. It appears that stress renders an organism more susceptible to the toxic effect of intact LAS, an effect which similarly occurs with other toxicants. Sublethal LAS exposures to the fathead minnow (Pirnephles) have also been reported to promote the toxic effects of other chemicals, or vice versa. Such reports are not surprising since toxicities of chemicals are often additive. Levels of 1 ppm LAS significantly increased the toxicity of various insecticides and pesticides ( Solon and Nair, 1970). The 48-hour mean percentage of survival of fish exposed to 0.8 ppm parathion was 5, 38.5, and 97.5% in the presence of 1, 0.5, and 0.25 ppm of LAS (Solon et aZ., 1969), respectively. Goldfish in tanks maintained at a 4-ppm LAS level for a 2-month period were more susceptible (more deaths per unit time) to DDT when compared to fish not living in the presence of LAS ( Dungan, 1967).
B. LAS TOXICITY vs BIODEGRADATION Although the toxicity of intact LAS toward fish is somewhat greater than that of the branched-chain ABS material, the effective hazard of LAS is much less, primarily because the LAS molecules are more rapidly
LAS : ENVIRONMENTAL BEHAVIOR
283
degraded and the partially degraded residues demonstrate progressively lower toxicity toward individual species than the original material (Niemitz and Pestlin, 1962; Swisher et al., 1964; Borstlap, 1967; Thompson et al., 1971). Thompson et al. (1971) noted LAS samples which had biodegraded to the extent of 90-95% had a 96 hour TL, value of greater than 10 pprn for guppies (Poeciliu) and harlequins (Rasbotu). Residues remaining after 90% degradation in activated sludge receiving 100 ppm of LAS were not toxic to fish when concentrated to 21-26 pprn (Niemitz and Pestlin, 1962). Borstlap (1967), in an attempt to generate the maximum amount of LAS intermediate degradation products, fed a static activated sludge culture 500 pprn dodecylbenzene sulfonate ( DOBS-JN Technical). After 4-6 weeks approximately 90-95% biodegradation had occurred ( MBAS ) . The biodegradation intermediates, poorly characterized, and representing 50-60 weight percent of the original LAS, were nonfoaming at concentrations of less than 60 ppm. The TL, value of these “products” for guppies (Lebistes) was in excess of 750 ppm. Swisher et al. (1964) fed continuous activated sludge units 100 ppm C, LAS and 165-200 ppm C , , LAS (in excess of 100 times the 96 hour TL, value for bluegill fingerlings, Lepomis). With LAS retention times of 3 and 7 hours, no toxic effects were observed for up to 9 days on fish living in 3-liter containers directly receiving the treated LAS effluent. Mixed isomers of sulfophenylundecanoic acid, representing an early stage high molecular weight intermediate biodegradation product, had a TL, value of 75 ppm for bluegills (Swisher et al., 1964). Also, in this study, fish living in “acclimated tanks,” in static effluents from the activated sludge process, were able to tolerate 1.5 x the TL, value without apparent toxic effects. In another study, acclimated aquaria receiving a single initial addition of LAS of 1 pprn were found to support growth and reproduction of guppies (Lebistes),white clouds (Cainchthys), and catfish (Cotydoros and Plecostomus) for up to 9 months without observable toxic effects ( Sharman, 1965). During the initial phase of the experiment, the LAS concentration decreased from 1 to less than 0.01 ppm. Studies of this nature indicate the effectiveness of an acclimated system for removing toxic effects of LAS molecules. Rapid biodegradation of LAS in acclimated systems coupled with the significantly lower toxicity of the slower degrading isomers and the much lower toxicity of the intermediate biodegradation products greatly diminishes the chance for upsets in the aquatic environment. The major source of concern would be not from use in household detergents, but from the isolated instances of large spillage at high concentration into unacclimated aquatic systems. In such cases, biodegradation may be delayed several days resulting in toxic effects being expressed.
284
WILLIAM E. GLEDHILL
C. CHRONIC LAS TOXICITY Studies on the temporary survival of aquatic organisms (acute TL, values) and gross pathological effects of elevated (unnatural) levels of LAS are of limited value. Chronic studies with sublethal concentrations of LAS for the purpose of detecting aberrant behavior, failure to reproduce, poor growth rates, histological abnormalities, population interactions, etc., are necessary. Unfortunately, a relatively limited number of such studies have been reported with LAS, none of which have incorporated the effects of biodegradation. Apart from the gross histological manifestations of toxic levels of LAS, more subtle effects are noted with sublethal concentrations. Respiratory folds of the gills of the trout are caused to stick together with loss of mucous cells upon exposure to 5 ppm LAS (Schmid and Mann, 1961, 1962). Sublethal concentrations of LAS have also been found to damage the chemoreceptors of lctalurus species ( Bardach et al., 1965 ). Damage, a function of exposure time, could not be detected histologically, but only via electrophysical measurements of olfactory function. Impairment of fish reproductive function as a function of detergent concentration was reported by Mann and Schmid (1961). Pickering and Thatcher ( 1970) examined four different sublethal levels of LAS (0.34, 0.63, 1.2, and 2.7 pprn). They reported the two higher levels not to affect hatchability of fathead minnow (Pimephales) eggs, but to inhibit 5-week weight gain and to lower survival. A 2.7 pprn LAS exposure (96 hour TL, value, 4.35 ppm) to the first generation fish resulted in no deaths and indicated LAS not to have a cumulative effect on mortality. Pickering and Thatcher concluded that LAS concentrations up to 0.63 pprn represented safe exposure levels for this species. Hokanson and Smith ( 1971) also reported the most sensitive developmental stage to LAS was the feeding sacfry (84 hour T L , 3.5 ppm). In this study with bluegills (Lepomis) 3 ppm of LAS was found to inhibit spermatozoa1 movement and, consequently, fertilization by 38%. Concentrations of 4.6-5.5 ppm, approaching the adult 48-hour TL, value (6.4 ppm) also inhibited egg hatching. Hokanson and Smith recommended the maximum safe level of LAS to fall between 0.63 and 1.2 ppm (14-18% of the 96 hour TL, value of 4.35 pprn). A 5 ppm concentration of LAS was also found to inhibit trout sperm movement, and prolonged exposure to 2-5 ppm inhibited egg development (Niemitz and Pestlin, 1962). Various isolated studies of the toxicological effects of LAS to other members of the aquatic ecosystem have appeared. Algae vary quite substantially in their response to LAS. Growth of Stichococcus, Nan-
LAS : ENVIRONMENTAL BEHAVIOR
285
nochloris, and Tetraselmis was inhibited by as little as 1 ppm of LAS (Ukeles, 1965; Renzoni, 1971), and Chlorella withstood 100-200 ppm without growth inhibition ( Matulova, 1964). Scenedesmus and Chlamydomonus were intermediate in their response, displaying growth inhibition between 2.5 and 10 ppm (Matulova, 1964). Growth of protozoans (Paramecium, Coccidium, and Trichomonus) was inhibited by C, LAS concentrations ranging from 3 to 100 ppm ( Duperray, 1969). Larvae of various marine invertebrates ( Arabacia, Asterias, Spisula, Chaetopterus, and Molgula) displayed gross developmental abnormalities and egg damage with LAS levels ranging from 0.7 to 3 pprn (Moffett and Grosch, 1967). Female brine shrimp, exposed to 5 pprn of LAS for 8 hours, produced fewer offspring, with fewer of these offspring surviving to maturity ( Moffett and Grosch, 1968). Fertilized egg development, larval survival, and growth of clams (Mercenaria) and oysters (Crassostrea) were found to be affected by levels of LAS ranging from 1 to 10 ppm (Hidu, 1965). The %-hour TL, value for 2 snails (Campeloma and Physa) and an Amphipod (Gammarus) were established as 27, 9, and 7 ppm, respectively (Arthur, 1970). A chronic 6-week study of these species with different LAS levels indicated the maximum concentration tolerated without apparent toxicological effects was 0.4 ppm for Gummarus, 1.0 ppm for Campeloma, and in excess of 4.4 pprn for Physa. The flow-through system employed was operated at a 0.3- to 0.8-hour retention time. Consequently, LAS degradation was only 18%(MRAS). Another extensive study of the acute and chronic effects of LAS was conducted by Swedmark et al. (1971). The studies were concerned with marine species, fish (cod, Gadus; flounder, Pleuronectes) crustaceans ( Balanus, Crangon, Leander, Eupagurus, Carcinus, and Hyas), and bivalves (mussel, Mytilus; oyster, Ostrea; and cockles, Astarte and Cardium ) and were conducted in a continuous flow-through apparatus with a series of surfactants including LAS. Specifically, 96-hour TL, values were recorded and behavior patterns (swimming, shell closure response, siphon retraction, breathing rate) were noted. In general, the more active species were most sensitive to LAS. The 96-hour TL, values for marine fish ranged from 1 to 5 ppm; for the crustacean larvae, 3 to 9 ppm; for the adult crustaceans, 50 to 100 ppm; and for the bivalves 5 to 100 ppm. Fish behavior was affected in the 0.1-1 ppm range, bivalves in the 100-ppm range, and crustacean larvae in the 1-10 ppm range. The animals first reacted with increased activity (avoidance) and at high concentration of surfactant with impaired function followed by death. The ability of an animal to recover normal behavior after exposure to LAS decreased with increasing surf actant concentration and increasing length of exposure.
286
WILLIAM E. GLEDHILL
IV.
Summary and Conclusions
Laboratory studies have documented the complete LAS biodegradation (alkyl chain and benzene sulfonate ring) in acclimated systems ( activated sludge, shake flasks). Biodegradation studies of LAS isomers and homologs from LAS species with C , to CI2 alkyl chain lengths in dilute, unacclimated aqueous systems have indicated the time required for biodegradation to decrease as the distance from the sulfonate group to the most remote methyl group on the alkyl chain is increased. Biodegradation times for longer chain homologs (C13 to C18) may be either somewhat less or the same as the C , , homolog depending on whether they are added separately or as mixtures to the system. In dilute river water systems, the rate of biodegradation decreases and the time required to gain acclimation increases with the C, to C, LAS homologs as the phenylsulfonate group is moved toward the center of the molecule. Transient intermediate biodegradation products have been observed in river water and activated sludge systems. Evidence for the accumulation of carboxylated 2- and 3-sulfophenyl isomers of butyric through heptanoic acid was presented. In activated sludge various internal sulfophenyl shorter chain length homologs have been observed to degrade more slowly in various studies. In the latter systems, however, degradation was somewhat slower than would be expected since none of the products have been detected in effluents from acclimated activated sludge units. The determination of the extent of LAS ring biodegradation and the establishment of the types of lower molecular weight biodegradation products present, if any, in natural systems from specific commercial LAS preparations have not been carried out. The verification of laboratory data in these areas, perhaps by use of 14C-ring-labeled LAS and GC, in relationship to acclimation would be important. With the nationwide sampling programs now underway, there have been and will be occasional “high MBAS values reported. No studies have been directed to establish the MBASILAS ratio in such isolated instances. Therefore, reexamination of high MBAS samples from natural waters or sewage effluents should be conducted in order to establish whether the MBAS value is due to LAS and, if so, which isomers are present. Also, since extensive LAS biodegradation may occur prior to its entry into sewage plants, knowledge as to the nature of those isomers and homologs which reach the plant and the effectiveness of the plant in removing both the LAS components and the other sewage components (BOD) would be of value. Field studies have generally supported laboratory data. In general, the extent of biodegradation reported for various LAS products ranges
LAS: ENVlRONMENTAL BEHAVIOR
287
from the mid 80% to upper 90% level (MBAS-GC, I R ) under typical plant operating conditions. Comparisons of data from different laboratories is often difficult because in many of the studies either the exact composition of the LAS products was unknown, products varied in isomer and homolog distribution, or the amount and nature of nonlinear byproducts in the LAS preparation was unknown. The extent of acclimation of sewage plants is uncertain in certain cases. Thus, reports of preferential isomer and homolog degradation, in perhaps poorly acclimated systems, have been presented. In contrast, other work indicates no preferential isomer or homolog biodegradation. LAS influents, in field studies, as high as 73 ppm (MBAS) have been reduced to 1.4 ppm effluents. Low temperatures and residence times have been seen to contribute to lower LAS biodegradation rates. Metabolic studies with pure and mixed cultures have indicated a variety of pathways for and organisms capable of degradation of part, or the total, LAS molecule. Several mechanisms of alkyl chain oxidation, sulfonate removal, and ring degradation have been documented. Additional work with postulated intermediates, such as the p-sulfophenylmalonic, -glutaric, and -succinic acids from odd- and even-chain LAS species, will further substantiate the mechanism of LAS degradation. Further research will, no doubt, continue to find other microorganisms involved in U S biodegradation and to define the enzymatic processes involved in LAS metabolism. Anaerobic studies, laboratory and field, indicate LAS to be poorly degraded. However, in anaerobic systems many organic molecules accumulate. Manmade anaerobic systems were designed primarily for solids removal, not soluble organic biodegradation. Anaerobic systems, such as subterranean soils, septic tanks, etc., present isolated environmental situations where accumulation of soluble organics, including LAS, does not seriously affect the outside world. Upon leaving such systems and entering aerobic environments, they would usually be substantially biodegraded. In isolated areas where surfactant accumulation ( LAS?), and probably other organic accumulation, has been questioned, such as in Long Island ground water, the need for adequate aerobic sewage treatment is indicated. Removal of synthetic surfactants from detergent formulations will have little effect upon the other chemical and biological pollutants entering the ground water with the sewage. At this time the consequences of and the solution to the anaerobic accumulation of organic compounds are unknown. However, in order to place LAS in its proper perspective, knowledge of its biodegradation under controlled microaerophilic conditions (various low oxygen concentrations ) in comparison to the degradation rate of other surfactants and organics would be beneficial.
288
WILLIAM E. GLEDHILL
Aquatic toxicity studies of LAS have centered primarily on acute experiments with fish. LAS toxicity is dependent, primarily, upon alkyl side chain length and to a lesser extent the phenyl sulfonate position. Acute toxicities (96 hour TL,) of commercial products toward various fish generally range from 3 to 7 ppm. The lowest reported lethal concentration for a mixed isomer LAS homolog was 0.1 pprn for C, LAS; however, this was observed at an elevated temperature (28OC) (Marchetti, 1968). Such data were obtained from systems in which LAS biodegradation was prevented, or greatly minimized; consequently, their applicability of validity under natural conditions is questionable. Assuming an average of ( a ) 10%C,., LAS in the LAS portion of detergents, ( b ) 10 ppm of LAS in sewage influent, ( c ) 90% biodegradation through the system, and ( d ) %o dilution of the effluent into a stream, it can be calculated the maximum anticipated C, LAS reaching fish in the stream would be 0.01 ppm. This calculation, indicating a 10-fold safety margin for the most toxic LAS homolog, is probably in error on the high side, since with this extent of biodegradation intact C, LAS would not exist, rather carboxylated, nontoxic, low molecular weight species would comprise the remaining material. Experimentally, the nontoxicity to fish of effluents from laboratory continuous activated sludge units receiving 200 ppm from C, LAS with retention times in the units of as little as 3 hours indicates the relative ease by which toxicity is lost. Field studies have tended to support the above calculations (Swisher, 1970, pp. 5 and 197). The reported MBAS levels of the Illinois River, an extreme example in the sense that it receives effluent from a very large metropolitan area (Chicago), before and after the branched-chain ABS to LAS switch over, are exemplary. Before the LAS switchover (1959-1965) the MBAS levels at the Pekin-Peoria reach, 160 miles downstream from Chicago, averaged 0.5 ppm. From 1965 to 1966 this level dropped to 0.2 pprn and by 1968 had further decreased to only 0.05 ppm. Less than 20% of the MBAS was attributed to the presence of LAS. Consequently, LAS levels were below 0.01 ppm, ten times less than the lowest reported lethal concentration for any LAS product or component. Summaries of the results of field studies in Germany, England, and the United States also attest to low MBAS levels in sewage effluents (Brenner, 1968; Heinz and Fischer, 1968; Husmann, 1968; Waldmeyer, 1968). Continued samplings are required in order that a correlation between the actual environmental LAS levels and aquatic toxicity can be made. Supplemental studies involving LAS as it exists in the aquatic environment (obtained by extraction from natural systems) should further strengthen the LAS toxicological picture. Young fish ( f r y ) and invertebrate larvae are the species most suscepti-
LAS: ENVIRONMENTAL BEHAVIOR
289
ble to sublethal concentrations of LAS. Maximum subacute toxicities (functional abnormalities) are generally reported in the 0.4-1 ppm range for various aquatic vertebrates and invertebrates; however, functional impairments have been noted with as little as 0.1 ppm. Nevertheless, adequate sewage treatment and subsequent dilution should reduce the surfactant concentration to below the level where sublethal effects are noted. Here again, the literature is lacking on suitable subacute toxicity studies employing “real-world conditions. More specifically, knowledge as to the rate of isomer and homolog biodegradation of specific current LAS products ( U V and GC analysis) in several natural systems (activated sludge, river water, lake water, etc.) under different stages of acclimation, and the direct relationship of this rate and state of acclimation to acute and chronic effects on suitable aquatic species would be desirable. There have been perhaps, only three significant papers relating to the sublethal effects of LAS toward aquatic species (Arthur, 1970; Pickering and Thatcher, 1970; Swedmark et al., 1971). In assessing the effect of LAS products on the aquatic ecosystem a complete food chain from algae and protozoans to fish should be examined under conditions where LAS is entering as a sewage component. Concentrations of surfactants which exert sublethal effects on individual laboratory species may elicit more or less pronounced toxic effects in a natural environmental situation where there may be ( a ) a delicate balance between individuals, ( b ) the presence of other “mildly toxic” materials, ( c ) a variation of individual response to introduced chemicals, and ( d ) bacteria capable of degrading the added compound. Also, the presence of soil, suspended soil particles, and sewage components in natural waters provides a surface for interaction with a surfactant, perhaps diminishing its toxic effects. Data on algae and protozoans are sketchy; species sensitivity to LAS varies greatly. The invertebrate population, especially the larvae, is susceptible to low levels of LAS. Sublethal effects of LAS products before and after sewage treatment on these species are lacking. Isolated literature reports also allude to significant effects of oxygen tension and temperature on the toxic effects of LAS and to the additive effects of other toxicants to LAS toxicity. Supplementary studies, of these effects in relationship to LAS biodegradation, would be beneficial, but highly complex without full knowledge of the experiments discussed previously. REFERENCES Alexander, M. (1965). Aduan. Appl. Microbial. 7, 38-80. Alexander, M., and Lustigman, B. K. (1966). 1. Agr. Food Chem. 14, 410-413. Allred, R. C., Setzkom, E. A., and Huddleston, R. L. (1964). J. Amer. Oil Chem. SOC.41, 13-17.
290
WILLIAM E. GLEDHILL
Arthur, J. W. (1970). Water Res. 4, 251-257. Bacon, R. L. (1966). I. Amer. Oil Chem. SOC. 43, 18-24. Baggi, G., Catelani, D., Galli, E., and Treccani, V. (1972). Biochem. 1. 126, 1091- 1097, Bardach, J. E., Fujiya, M., and Holl, A. (1965). Science 148, 1605-1607. Bird, J. A., and Cain, R. B. ( 1972). Biochem. J., Proc. 127, 46P. Bock, K. J. (1966). Arch. Fischereiwiss. 17, 68-77. Bock, K. J. (1967). Muenchner Beitr. Abwasser-, Fisch. FlussbioI. 9, 110-117. Bock, K. J., and Mann, H. ( 1971). Arch. Fischereiwiss. 22, 287-292. Bock, K. J., and Wickbold, R. (1967). Vom Wasser 33, 242-252. Borstlap,-C. (1967). Proc. Int. Congr. Surface Actiue Subst., 4th, 1964 Vol. 3, pp. 891-901. Borstlap, C., and KortIand, C. (1967). J. Amer. Oil Chem. SOC. 44, 295-297. Brenner, T. E. (1968). J. Amer. Oil Chem. SOC. 45, 433-436. Brink, R. H., and Meyers, J. A. (1966). J. Amer. Oil Chem. SOC. 43, 449-451. Burnop, V. C. E., and Bunker, H. J. (1960). Cent. Belge Etude Doc. Eaur, BuZl. Mens. 4, 262-268. Busch, A. W. ( 1966). Water Resour. Res. 2, 59-69. Cain, R. B., and Farr, D. R. ( 1968). Biochem. J. 106, 859-877. Cain, R. B., Willetts, A. J., and Bird, J. A. (1971). Proc. Int. Biodeterior. Symp., 2nd, 1971, p p . 136-144. Ciattoni, P., and Scardigno, S. (1968). Riu. Ital. Sostanze Grasse 45, 15-26. Davis, E. M., and Gloyna, E. F. ( 1969). J. Amer. Oil Chem. SOC. 46,604-608. DeJong, A. L. (1964). Int. J. Air Water Pollut. 8, 591-608. Dungan, P. R. (1967). J. Water Pollut. Contr. Fed. 39, 63-71. Duperray, J. N. (1969). Proc. Int. Congr. Surface Active Subst., 5th, 1968 Vol. 3, pp. 243-250. Durand, R. R. (1965). Chim. Ind. (Paris) 94, 345-365. Eden, G. E., Truesdale, G. A., and Stennett, G. V. (1968). Water Pollut. Contr. 67, 107-116. Focht, D. D., and Williams, F. D. (1970). Can. J. Microbiol. 16, 309316. Foster, D. J., and Fields, R. R. (1964). Soap Chem. Spec. 40, 49-52. Fujiwara, Y., Takezono, T., Kyona, S., Sakayanigi, S., Yamasato, K., and Iizuka, H. (1968). Yukagaku 17, 396-399. German Government ( 1962). Bundesgesetzblatt (Bonn) 49, Part I , 698-706. Halvorson, H., and Ishaque, M. (1969). Can. J. Microbiol. 15, 563-576. Hammerton, C. (1955). J. Appl. Chem. 5, 517. Hanna, G. P., Sheets, W. D., Weaver, P. J., and Gerhold, R. M. (1965). Proc. Ind. Waste Conf. 20, 725-734. Heinz, €I. J., and Fischer, W. K., (1968). Hydrocarbon Process. 47, 96-102. Heyman, J. J., and Molof, A. H. (1967). Environ. Sci. Technol. 2, 773-778. Hidu, H. (1965). J. Water Pollut. Contr. Fed. 37, 262-270. Hirsch, E. (1963a). Vom Wasser 30, 249-259. Hirsch, E. (196313). Fette, Seifen, Anstrichm. 65, 814-818. Hitzman, D. 0. ( 1964). J. Amer. Oil Chem. SOC. 41, 593-595. Hokanson, K. E . F., and Smith, L. L. (1971). Trans. Amer. Fish. SOC. 100, 1-12. Huber, L. (1968). Tenside 5, 65-76. Huddleston, R. L., and Allred, R. C. (1963). Develop. Ind. Microbiol. 4, 24-37. Huddleston, R. L., and Allred, R. C. (1970). In “Soil Biochemistry” ( A . D. McLaren and G. H. Peterson, eds.), Dekker, New York. pp. 343-370. Hunter, J. V., ‘and Henkelekian, H. ( 1964). Proc. Ind. Waste Conf. 19, 616-627.
U S : ENVIRONMENTAL BEHAVIOR
291
Husmann, K. (1968). Water Pollut. Contr. 67, 80-90. Janicke, W. (1988).Gus-Wusserfaoh 109, 246-249. Janicke, W. (1971). Water Res. 5, 917-931. Kelly, R. J., Konecky, M. S., Shewmaker, J. E., and Bernheimer, R. (1965). SCI. ( SOC. Chem. Ind., London) Monogr. 19, 286-304. Klein, S. A., and McGauhey, P. H. (1965). 1. W&r Pollut. Contr. Fed. 37, 857-866. Knapp, J. W., and Morgan, J. M. (1965). Proc. Ind. Waste Conf. 20,737-745. Knopp, P. V., Uhren, L. J., Rohlich, G. A., and Nichols, M. S. (1965a). Proc. Ind. Waste Conf. 20, 745-754. Knopp, P. V., Uhren, L. J., Rohlich, G. A., and Nichols, M. S. (1965b). J. Amer. Oil Chem. SOC. 42, 867-873. Kriiger, R. (1964). Fette, Seifen, Anstrichm. 66, 217-221. Kriiger, R. (1966). Lichtbogen 15, 24-27. Kumke, G. W., and Renn, C. E. (1966). J. Amer. Oil Chem. SOC. 43, 92-94. Longwell, J., and Maniece, W. D. (1955). Analyst 80,167-171. Ludzack, F. J., Schaffer, R. B., Bloomhuff, R. N., and Ettinger, M. B. (1959). Sewage Ind. Wastes 31, 33-44. McGauhey, P. H., and Klein, S. A. (1964). Proc. Ind. Waste Conf. 19, pp. 1-8. McGauhey, P. H., and Klein, S. A. (1967). Proc. Int. Conf. Water Pollut., 3rd, 1968 Vol. 1, pp. 353-374. Mann, A. H. (1969). Proc. Int. Congr. Surface Actioe Subst., 5th, 1968 Vol. 1, pp. 103-1 14. Mann, A. H., and Reid, V. W. (1971). J. Amer. Oil Chem. SOC. 48, 588-594. Mann, H. (1970). Fette, Seifen, Anstrichm. 72, 981-983. Mann, H., and Schmid, 0. J. (1961). lnt. Reo. Cesamten Hydrobiol. 46, 419-426. Marchetti, R. (1968). Riu. Ital. Sostanze Grasse 45, 2734. Matulova, D. (1964). Sb. Vys. Sk. Chem. Technol. Praze, Technol. Vody. 8, 251-301. Maurer, E. W., Cordon, T. C., Weil, J. K., Nunez-Ponzoa, M. U., Auk, W. C., and Stirton, A. J. (1965). J. Amer. Oil Chem. SOC.42, 189-192. Maurer, E, W., Cordon, T. C., and Stirton, A. J. (1971). J. Amer. Oil Chem. SOC. 48, 163-165. Moffett. D. S., and Grosch, D. S. (1967). Biol. Bull. 133, 476-477. Moffett, D. F., and Grosch, D. S. (1968). Biol. Bull. 135, 429. Niemitz, W., and Pestlin, W. (1962). Staedtehyg. 13, 231-232. Oba, K. (1965). Yukagaku 14, 565-571. Oba, K. ( 1971). Nippon Eiseigaku Zasshi 25, 494-511. Oba, K., Yoshida, Y., and Tomiyama, S. (1967). Yukagaku 16, 27-33. Orgel, G., and Rupp, F. A. (1967). Proc. Int. Congr. Surface Actioe Subst., 4th, 1964 V O ~3, . pp. 131-140. Payne, W. J. (1963). BiotechnoL Bioeng. 5, 355-365. Payne, W. J., and Feisal, V. E. (1963). Appl. Microbiol. 11, 239-244. Pickering, Q. H. (1966). Int. J. Air Water Pollut. 10, 385-391. Pickering, Q. H., and Thatcher, T. 0. (1970). 1. Water. Pollut. Contr. Fed. 42, 243-254. Pitter, P. (1963). Sb. Vys. Sk. Chem. Technol. Praze, Technol. Vody. 7, 19-32. Pitter, P. (1964a). Sb. Vys. Sk. Chem. Technol. Praze, Technol. Vody. 8, 1-7. Pitter, P. (196413). Sb. Vys. Sk. Chem. Technol. Praze, Technol. Vody. 8, 13-39. Pitter, P. ( 1 9 6 4 ~ )Sb. . Vys. Sk. Chem. Technol. Praze, Technol. Vody 8, 167-174. Pitter, P. ( 1964d). Sb. Vys. Sk. Chem. Technol. Praze, Technol. Vody 8, 175-193. Pitter, P. (1967). Proc. Int. Congr. Surface Actioe Subst., 4th, 1964 Vol. 3, pp. 861-869.
292
WILLIAM E. GLEDHILL
Pitter, P., Veselsby, J., and Ciperova, M. (1971). Sci. Pap. Inst. Chem. Technol., Prague 16, 3 7 5 1 . Prakasam, T. B. S., and Dondero, N. C. (1970). App2. Miorobiol. 19, 663-680. Renn, C. E. (1965). Pap. 38th S o a p Detergent Ass. Conv. Renn, C. E., Kline, W. A., and Orgel, G. (1964). J. Water Pollut. Contr. Fed. 36, 864-879. Renn, C. E., Kline, W. A., and Orgel, G. (1965). Proc. Ind. Waste Conf. 20, 734-737. Renzoni, A. (1971). Rev. Int. Oceanogr. Mediterr. 24, 50-52. Ripen, M. J., Noon, K. F., and Cook, T. M. (1971). Appl. Microbiol. 21, 495-499. Rismondo, R., and Zilio-Grandi, F. (1968). Riu. Ital. Sostanze Grasse 45, 116-121. Ryckman, D. W., and Sawyer, C. N. (1957). Proc. Ind. Waste Conf. 12,270-284. Sawyer, C. N., Ryckman, D. W. ( 1957). J. Amer. Water Works Ass. 49, 480-489. Schmid, 0. J., and Mann, H. ( 1961). Nature (London) 192, 675. Schmid. 0. J., and Mann, H. (1962). Arch. Fishereiwiss. 13, 41-51. Setzkorn, E. A., and Huddleston, R. L. (1965). J. Amer. Oil Chem. SOC. 42, 1081-1084. Setzkorn, E. A, Huddleston, R. L., and Allred, R. C. (1964). J. Amer. Oil Chem. SOC. 41, 826-830. Sharman, S. H. (1964). Nature (London) 201, 704-705. Sheets, W. D., and Malaney, G. W. (1956). Sewage Ind. Wastes 28, 10-16. Simko, J. P., Emery, E. M., and Blank, E. W. (1965). J. Amw. Oil Chem. SOC. 42, 627-6.29. Smith, F. D., Stirton, A. J., and Nunez-Ponzoa, M. V. (1966). J . Amer. Oi2 Chem. SOC.43,501-504. Soap and Detergent Association. ( 1965). J. Amer. Oil Chem. SOC.46, 432440. Solon, J. M., and Nair, J. H. (1970). Bull. Environ. Contam. Toxicol. 5, 408-413. Solon, J. M., Lincer, J. L., and Nair, J. H. (1969). Water Res. 3, 767-775. Southgate, B. A, and Eden, G. E. (1961). C h m . Process Eng. 42, 5-10. Spohn, H. (1964). Tenside 1, 1-8 and 18-26. Spohn, H. (1967). Tenside 4, 241-247. Standard Method Committee-Subcommittee on Biodegradation. ( 1967). J . Water Pollut. Contr. Fed. 39, 1232-1235. Standing Technical Committee on Synthetic Detergents. ( 1967). “Ninth Progress Report.” H.M. Stationery Office, London. Straus, A. E. (1963). Science 142, 244-245. Sturm, R. N. (1973). J. Amer. Oil Chem. Soc. 50, 159-167. Swedmark, M., Braaton, B., Emanuelson, E., and Granmo, A. ( 1971). Mar. Biol. 9, 183-201. Sweeney, W . A. (1964). Soap Chem. Spec. 30, 45-48. Sweeney, W. A, and Foote, J. K. ( 1964). J. Water Pollut. Contr. Fed. 36, 14-37. Swisher, R. D. (1963a). J . Water Pollut. Contr. Fed. 35, 877-892. Swisher, R. D. (1963b). J. Water Pollut. Contr. Fed. 25, 1557-1564. Swisher, R. D. (1964). Chem. Eng. Progr. 60, 41-45. Swisher, R. D. (1967). 3. Amer. Oil Chem. SOC.44, 717-724. Swisher, R. D. (1968a). Develop. Ind. Microbiol. 9, 270-279. Swisher, R. D. (196813). Id.Chim. Belge 32, Part 111, 719-722. Swisher, R. D. ( 1970). “Surfactant Biodegradation.” Dekker, New York. Swisher, R. D. (1972). Yukagaku 21, 130-142. Swisher, R. D. (1973). Progr. Water Tedhnol. 3, 303-306. Swisher, R. D., O’Rourke, J. T., and Tomlinson, H. D. (1964). J. Amer. Oil Chem. SOC.41,746-752.
U S : ENVIRONMENTAL BEHAWOR
293
Symons, J. M., and DelValle-Rivera, L. A. (1961). Proc. Ind. Waste Conf. 16, 555-571. Tarring, R. C. (1965). Int. J. Air Water Pollut. 9, 545-552. Taylor, B. F., Campbell, W. L., and Chinoy, I. (1970). J. Bacteriol. 102, 430-437. Thatcher, T. O., and Santner, J. F. (1966). Proc. Ind. Waste Conf. 21, 996-1003. Thompson, A., Mann, A. H., Shepherd, E. L., Brown, V. K. H., Box, V. L., and Hubble, D. (1971). Report M(T)-6-71, pp. 1-9. Shell Res. Ltd., London. Thompson, J. A., and Duthie, J. R., (1968). J. Water Pollut. Contr. Fed. 40, 306-3 19. Treccani, V. ( 1971). Riv. Ital. Sostanze Grasse 48, 233-235. Ukeles, R. (1965). 1. Phycol. 1, 102-110. Vaicum, L. (1969). Proc. Int. Congr. Surface Active Subst. Sth, 1968 Vol. 1, pp. 111-114. Waldmeyer, T. (1968). W a k r Pollut. Contr. 67, 66-79. Wayman, C. H., and Robertson, J. B. (1963). Biotechnol. Bioeng. 5, 367484. Wickbold, R. (1967). Proc. Int. Congr. Surface Active Subst., 4th, 1964 Vol. 3, pp. 903-912. Willetts, A. J., and Cain, R. B. (1972a). Biochem. J. 129, 389-402. Willetts, A. J., and Cain, R. B. (1972b). Antonie von Leeuwenhoek; J . Microbiol. Serol. 38, 543-555. Wurtz-Arlet, J. (1967). Proc. Int. Congr. Surface Active Subst., 4th, 1964 Vol. 3, pp. 937-943.
This Page Intentionally Left Blank
The Story of the American Type Culture CollectionIts History and Development (1 899-1 973)
WILLIAMA. CLARK AND DOROTHY H. GEARY American Type Culture Collection, Rockville, Maryland
I. Founding and Early Years
.............
11. Expansion .................... 111. Permanent Facilities ...............
....................
References
295 299 303 309
The role of the infinitely small in nature is infinitely great. -Louis Pasteur
I.
Founding and Early Years
The American Type Culture Collection was bred and nurtured by the scientific community and plays a versatile .role in that community today. While serving as a museum, a bureau of standards, and a national repository for microorganisms, it is dedicated to service, research, and education. The history of the American Type Culture Collection is the story of individuals working tirelessly, giving of their time, energy, and expertise because they believed in the basic role the Collection should play in advancing the science of microbiology. Some worked for little or no compensation, many without recognition; yet each person’s contribution was essential to the development of this national resource. This history cites but a few of these people; it is in no way intended as a definitive list of those who have helped the ATCC. Today this private nonprofit organization houses the largest collection of diverse microorganisms and related materials in the world. Behind this assertion is a story of struggle and determination that has covered over sixty years. But it was at the end of the last century that the story really began. On December 1899 bacteriologists from all over the United States convened in New Haven, Connecticut to found the Society of American Bacteriologists (now the American Society for Microbiology). The charter members had a lively interest in taxonomy and nomenclature and had long felt the need for a central distribution agency for bacterial cultures. One of these members was C.-E. A. Winslow of the Bacteriology Department at the Massachusetts Institute of Technology, who later became curator of the Museum of Natural History in New York. It was Winslow’s idea to found a “museum of living bacteria for the benefit 295
296
WILLIAM A. CLARK AND DOROTHY H. GEARY
C.-E. A. Winslow
L. A. Rogers
R. D. Coghill
AMERICAN TYPE CULTURE COLLECTION
297
of working laboratories all over the world.” In 1911 he established the Bacteriological Collection and Bureau for the Distribution of Bacterial Cultures at the Museum of Natural History and sent this announcement to the leading laboratories of the country: “The Department of Public Health at the Museum of Natural History has equipped a laboratory to serve as a central bureau for the preservation and distribution of bacterial cultures of both pathogenic and non-pathogenic organisms. . . . It is hoped that . . . those engaged in biochemical work of all sorts will furnish the museum with cultures at present in their possession. . . . The laboratory is ready to receive and care for such cultures.” The response was gratifying. Cultures began arriving from all over the United States and Canada. The famous KrAl Collection in Vienna made arrangements to exchange cultures with the new collection. By December 1, 1912 the collection included 578 strains, representing 374 different named species. During the first two years the collection distributed cultures to 122 different colleges and research laboratories, all without charge. In 1913 the embryonic collection published its first catalogue, listing about 350 strains. This was long before the first edition of Bergey’s Manual appeared to help standardize bacterial names, and species names were interspersed with common names. The catalogue listed “Bacillus of Canned Clams” and “Bacillus of Ropy Milk along with the very proper Bacillus subtilis. After Winslow left the Museum for a position at the Yale University Medical School, the collection deteriorated from lack of supervision. In 1922 the Society of American Bacteriologists (SAB) assumed responsibility for the collection, thereby beginning a formal involvement that has continued through the years. The Society decided to move the collection, which now contained only 175 strains, to the Army Medical Museum in Washington, D.C., Lore A. Rogers, then president of SAB, moved the meager collection to Washington in a suitcase. The Army Medical Museum, located at that time in an old brick building on the Mall, had in its courtyard a secondary building which was turned over to the collection. As landlord the Museum was generous, providing space and supplies free, and even allowing the collection the use of its franking privileges to mail cultures. There was still no charge for the cultures. W. R. Albus, a bacteriologist in Lore Rogers’ laboratory at the USDA’s Bureau of Dairy Industry, served as Acting Curator, overseeing the c01lection on his own time. Everyone realized that such an arrangement was merely a holding operation, and that a permanent organization must be established for the maintenance and distribution of cultures. Thus on October 23, 1924, a joint committee of the Divisions of Medical Sciences and of Biology
298
WILLIAM A. CLARK AND DOROTHY H. GEARY
and Agriculture of the National Academy of Sciences met to organize a committee to manage the collection. Frederick P. Gay, Chairman of the Division of Medical Sciences, helped obtain a $24,000 grant from the General Education Board of the Rockefeller Foundation to be used over a period of five years. The joint committee appointed a “Committeein-Charge” with the members representing various scientific societies. Society representatives thus began their roles as advisers, mentors, and members of the governing body of the ATCC. The list of the charter members of the Committee included some of the foremost scientists of the time: Lore Rogers (Chairman) and S. J. Nichols (on Nichols’ death, R. E. Buchanan), representing the Society of American Bacteriologists; C. A. Kofoid, American Zoological Society; C. L. Shear, American Phytopathological Society, and F. P. Gay, American Association of Pathologists and Bacteriologists. In 1925 the Committee-in-Charge proposed the establishment of an agency “to preserve cultures of microorganisms that have historic and scientific interest and to provide a center for obtaining cultures needed in education and research.” The collection was incorporated as a nonprofit scientific institution and called by the same name that it bears today: The American Type Culture Collection. The John McCormick Institute of Infectious Diseases offered to sponsor the fledgling Collection. So in January 1925 the remaining 175 cultures were moved to Chicago, under the general direction of Professor Ludwig Hektoen of the McCormick Institute. George Weaver, Curator, and his assistant Leila Jackson, comprised the entire staff. Later W. R. Albus and a clerk were added to the staff. In just one year the holdings increased to 722 strains. The first notable addition was made by E. 0. Jordan of the University of Chicago, who gave 139 subcultures from his private collection. R. S. Breed of the New York State Experiment Station, Geneva, New York, also contributed some strains. Several of the special collections were not housed in Chicago, but were made available by the cooperation of individuals who maintained private collections : yeasts, F. V. Tanner, University of Illinois; actinomycetes, S. A. Waksman, New Jersey Agricultural Experiment Station; anaerobic bacteria, I. C. Hall, University of Colorado Medical School; Fzlsarium, C. D. Sherbahoff, University of Tennessee, Agricultural Experiment Station; nodule bacteria of legumes, E. B. Fred, College of Agriculture, University of Wisconsin; molds and plant pathogens, Charles Thom and Margaret B. Church, Department of Agriculture, Washington, D.C. Other established collections, such as the National Collection of Type Cultures in London and the Centralbureau voor Schimmelcultures in Baarn, The Netherlands, contributed strains, gratis. (This practice of a free exchange of strains between the major culture collections of the world is still followed today.)
AMERICAN TYPE CULTURE COLLECTION
299
In 1930 the Rockefeller Foundation gave the Collection an additional grant of $10,000 with the understanding that when that amount was exhausted no more funds would be forthcoming. To meet expenses the ATCC began charging for cultures. “In only a few instances have there been any objections to the payment of the price of one dollar for cultures,” said the curator in his first annual report. Catalogues were issued in 1927, 1928, and 1934 and their appearance materially increased the sale of cultures. So, through income from cultures and various contributions, the budget was balanced until the depression years. During those crucial years the SAB faithfully continued its financial support, and the drug and allied industries also made contributions. In desperation culture fees were raised to an unheard of two dollars and the staffs salaries were adjusted month by month according to income. But the ATCC survived. It remained in Chicago 12 years and increased the collection to 2000 strains. During those years there was an increased demand for special cultures for commercial purposes. However, the depression had sorely affected the McCormick Institute, and it was no longer able to sponsor the Collection. So in the midst of its new growth the Collection was compelled to look for a new home.
II.
Expansion
The Committee-in-Charge decided that Washington, D.C. would be the best location, but no institution was anxious to sponsor the Collection and there seemed to be no suitable laboratory space available anywhere in the city. Finally the Georgetown University Medical School offered quarters and in October, 1937, the ATCC was moved to its new home at the University. Mario Mollari, Professor of Bacteriology and Preventive Medicine at Georgetown, became curator. The Collection shared laboratory space with the Department of Bacteriology. An additional small room housed the fungi collection. With this move the staff began to grow. Katherine Alvord was employed as a secretary that same October. She capably served the American Type Culture Collection as a key lay person long and loyally, retiring from her post as Secretary-Manager in 1960. During her early years with the ATCC, as part-time scientists came and left, Katherine Alvord, sometimes almost single-handedly, kept the Collection functioning on a day-to-day basis. During this era P. A. Hansen, Paul Hegarty, 0. A. Bushnell, and Isabelle Bouldin were part of the organization. A full time mycologist, Isabel Christison, also was added to the staff. In 1939 Lore Rogers and Paul Hegarty began experimenting with various techniques for freeze-drying cultures. Rogers ( 1914) had pio-
300
WILLIAM A. CLARK AND DOROTHY H. GEARY
neered in freeze-drying and was among the first to freeze-dry bacteria successfully. By 1940, 90%of the bacterial stocks in the ATCC were maintained in the freeze-dried state, but cultures were not freeze-dried for distribution. Still wary of the freeze-drying process, the Collection maintained duplicate cultures in test tubes. ( I t was not until the early 1960's that most of the cultures were distributed in a freeze-dried state.) During this period the Collection began expanding and its financial situation improved. Sales to commercial firms increased by 35%in 1942 as a result of the growing antibiotics industry. The Collection benefited from the war economy and, in turn, aided the war effort. At the request of the Surgeon Genera1 of the United States Public Health Service the ATCC became a central depository for cultures isolated during epidemics. By 1947 the burgeoning Collection was maintaining almost 3000 strains and outgrowing the Georgetown University facilities. The search began again for another home. The ATCC seemed doomed to a peripatetic existence. This time, however, the Collection decided it needed independent quarters. After considerable searching the Committee-in-Charge rented a small two-story apartment building at 2029 M Street in Washington. It was in poor condition and required extensive remodeling. Lore Rogers and the staff of nine did most of the work themselves, spending many weekends wielding paint brushes, hammers, and saws. Rogers, like Pasteur, even made the incubators. When the remodeling on the old apartment building was finally completed, it contained an office, a media room, and chemistry, bacteriology, and mycology laboratories. In September 1947 the Collection moved into its new quarters with Ruth Gordon as curator. The more fastidious cultures were stored at USDA's Plant Industry Station in nearby Beltsville, Maryland, and part of Gordon's work was done there in Nathan R. Smith's laboratory. Smith aided immeasurably in the work at the ATCC. Not only did he make his unique collection of strains of Bacillus species available to the ATCC, but prepared subcultures of the Azotobacter species for orders. He gave his time and counsel unstintingly to the incredibly diverse problems of the ATCC. It was at this time that Gordon collaborated with Smith and Francis Clark in monographing the genus Bacillus (Smith et al., 1952). Gordon began verification of the specific identity of every strain in the collection aided by scientists in many laboratories in the Washington area: in the Department of Agriculture A. B. Crawford (animal pathogens), L. A. Burkey, H. R. Curran and R. P. Tittsler (streptococci and lactobacilli), J. A. Stevenson, W. W. Diehl, R. W. Davidson, and C. Drechsler (fungi and actinomycetes), and K. B. Raper and L. J. Wickerham (filamentous fungi and yeasts). The laboratories of the National Institutes of Health were also a source
AMERICAN TYPE CULTURE COLLECTION
301
of assistance: C. W. Emmons (fungi of medical importance), Sara E. Branham (Ndsseriu), Bernice E. Eddy ( pneumococci), Margaret Pittman (Hemophilus),and Elizabeth Verder (streptococci). Microbiologists of the Army Medical School, particularly A. Abrams ( Salmonella and ShigeZZu), and Harriette D. Vera, Baltimore Biological Laboratories ( clostridia), also helped. R. E. Buchanan paid this tribute: “The Collection owes Ruth Gordon a debt of gratitude for making it an organization from which one might expect real contributions to knowledge of the taxonomy of the bacteria.” The growing organization demanded more staff members. One of these was Marion T. Alexander, age 18, just graduated from high school. Gordon encouraged him to get his college degree, which he did by attending classes at night and on a part time basis. Today he is ATCC‘s Chief of Facilities and Microbiological Services. In 1948 another financial crisis hit the Collection. Remodding expenses had exceeded expectations and the reserve funds were depleted. Costs of operation were going up. The sale of cultures was good, but not sufficient to meet total maintenance costs. It became clear that it was impossible for a culture collection to be self-supporting on a fee-for-culture basis, since the rate of acquisition of potentially important strains always exceeded the contemporary demand for cultures. There was a real danger of the collection going bankrupt. The staff was informed that funds were available for only two more months. The financial situation became so acute that the Committee-in-Charge approached the National Research Council for help. In January 1949, the NRC formed an ad hoc committee with Henry Welch, of the USDA’s Food and Drug Administration as chairman, and R. D. Coghill, Abbott Laboratories, as vice-chairman. As a result of the committee’s recommendations the constitution and by-laws were revised and the Commiteee-in-Charge became a Board of Trustees. Representatives from the National Academy of Sciences, the Mycological Society of America, and the American Institute of Biological Sciences were added to the Board, and culture fees were raised to $10 (with a 701%reduction for educational institutions). An appeal letter sent to friends in industry raised some $15,000 that served as a contingency fund. In 1949 Lore Rogers retired as Chairman of the Board after 24 years of association with the ATCC. R. E. Buchanan wrote in 1964: “Had it not been for the devotion, foresight, executive ability, ingenuity, and manual dexterity of Lore Rogers, I am sure that the collection would not have survived.” Carl Lamanna, of the Johns Hopkins University, followed Rogers as Chairman of the Board, and in 1950, when Ruth Gordon accepted a position at Rutgers University, Freeman A. Weiss became the new cura-
302
WILLIAM A. CLARK AND DOROTHY H. GEARY
tor. Under Weiss’ management, the ATCC entered a decade of growth. During these years the United Nations Educational Scientific and Cultural Organization ( UNESCO) contributed to the Collection’s support. The ATCC also initiated a “patent culture” deposit service which later was to provide a significant source of income. The Society of American Bacteriologists appointed a Technical Advisory Committee to advise ATCC on techniques of culture maintenance and laboratory practice. This advisory committee was the forerunner of today’s advisory committees composed of some 90 specialists from pertinent fields who give guidance in various aspects of service and research. These were times of a growing awareness that the old methods of test tube culture maintenance, though necessary in the early days because better techniques were not available, were becoming obsolete. Freezing and freeze-drying not only would stabilize ever-changing microbial forms but also would save time and money. Improved powdered media for propagating bacteria and fungi also began to make daily life in the cuIture collection easier. In 1950 the Viral and Rickettsia1 Registry was established at the ATCC by a group of scientists engaged in the study of viral and rickettsia1 diseases. Joseph E. Smadell of the Army Medical School was largely responsible for its formation, with financial help from NIH. The aim was to preserve important strains of these organisms and to provide a convenient means for their distribution. Members of the group each prepared a number of ampules of one or more agents and deposited them, along with documentation, in the ATCC. Randall Thompson, Morris Schaeffer, Francis Gordon, and Edwin Lennette helped carry on this effort. A start was made, in 1951, in forming a collection of plant viruses. The specimens were prepared, maintained, and shipped by individual collaborators. The ATCC, aided by a committee composed of concerned plant virologists led by H. H. McKinney, published catalogues in 1953 and in 1958, listing 103 strains. William Arthur Clark joined the staff as Assistant Curator under Weiss in 1954. A short time earlier the Viral and Rickettsia1 Hegistry Committee had recommended the establishment of a bacteriophage collection. The National Science Foundation awarded the ATCC a grant to support this collection, and Clark was put in charge. The seven T phages from S. E. Luria and the collection of I. N. Asheshov formed the nucleus of the collection. Space was scarce at the M Street quarters, but fortunately the National Canners’ Association provided laboratory space for Clark‘s work. R. E. Buchanan, long time Board member, Professor of Bacteriology
AMERICAN TYPE CULTURE COLLECTION
303
and Dean of the Graduate School at Iowa State College, was a familiar figure at the Collection in those days. He maintained his interest to the present time, and with his expertise in bacterial systematics had great influence on the Collection. John Stevenson of the USDA's National Fungus Collections in Beltsville, was another ATCC advocate. He maintained the ATCC's fungus collection in the early days when it was housed at the USDA. His recommended revision of the ATCC constitution, made in 1962 while he served on the Board of Trustees, still stands today. In 1954, when Kenneth Raper became Chairman of the Board, the ATCC maintained in its 2029 M Street quarters an impressive inventory of strains: 114 actinomycetes, 23 algae, 1889 bacteria, 21 bacteriophages, 1358 filamentous fungi, 2 myxomycetes, 6 protozoa, 68 animal viruses, and 511 yeasts, all from a meager nucleus of 175 cultures in 1925. The seams were indeed beginning to bulge and it became obvious that the ATCC was going to have to move once again. Fortunately the collection was able to buy and renovate a brownstone apartment building at 2112 M Street, N. W., one block from the old quarters. The new headquarters were located in a neighborhood which was to undergo renovation, and when this property was sold in 1964, the profit became the nucleus for a small Foundation Fund that was to play an important role in the Collection's financial future. Weiss retired as Curator in 1960. His quiet, sincere ways made many friends for the ATCC. The Board of Trustees retired Weiss with a stipend (the Collection had no retirement policy at that time). He promptly donated it to the ATCC's Foundation Fund. Ill.
Permanent Facilities
William Clark succeeded Weiss as Curator in 1960, and when the Board authorized a reorganization, he became the Director. The ATCC at this time became subdivided into individual collections, each responsible to the Director. No actual physical division was possible, however, because of lack of laboratory space. Office space was rented in nearby buildings, and an abandoned fire house next door was used for storage. In 1959 the data presented at three ad hoc conferences at the National Research Council, the National Cancer Institute, NIH, and the Rockefeller Institute revealed widespread contamination and admixture of cell cultures that cast serious doubts on the validity of much research conducted with cultured animal cells. Furthermore, it was evident that cultured cell strains were constantly subjected to selection processes and a variety of changes with continuous serial passage.
304
WILLIAM A. CLARK AND DOROTHY H. GEARY
The National Cancer Institute asked J. T. Syverton to organize a national collection of characterized animal cell strains to serve as reference standards. A grant funded a Cell Culture Collection Coordinating Committee composed of experts to advise and aid in the development of the collection; to formulate policies, procedures, and standards; to accept and certify cell lines which satisfied established standards; to facilitate cooperation among participating laboratories and investigators; and to encourage research in characterization and preservation of cell lines. The Child Research Center of Michigan, Detroit, the South Jersey Medical Research Foundation, Camden, the Naval Biological Laboratory, University of California, Berkeley and the ATCC were designated cooperating laboratories. ATCC also was designated as the center for storage and distribution of reference seed cultures. John E. Shannon, Jr. became the first curator. At the same time, plans also were set in motion to seek funds for permanent facilities for the ATCC to accommodate the Animal Cell Culture Collection and to house adequately the other flourishing collections. The ATCC requested funds from NIH and NSF to construct a permanent laboratory building. NIH awarded a $90,000 grant, with the understanding that it be matched by the ATCC. R. D. Coghill offered to lead a drive to obtain the matching funds. Through Coghill’s tireless efforts, $157,000 were raised from industry. The National Science Foundation awarded the ATCC a building grant of $865,000 along with a general logistics grant for support of the Collection’s program. In May 1961 the Board purchased a two-acre building site in Rockville, Maryland, a Washington suburb. The ground-breaking ceremonies were held on March 15, 1963. Francis L. Schmehl, Chief of NIH’s Health Research Facilities Branch, called the projected new facility “the first real home for one of our nation’s most precious resources.” A building specially designed for culture collections of diverse microorganisms and cell lines presented some unique challenges in design and construction. Architect Anthony Harrer, of Ronald Senseman Associates, in consultation with the director, the curators, and the collaborators of ATCC, designed a building which was to be virtually problem free. The million dollar building, comprising 35,000 square feet of space, was built in less than a year by Victor R. Beauchamp Associates for $29 per square foot, equipped. Well designed and well-built, it is admirably maintained now by the special expertise of building engineer George Mays. Moving day came February 23, 1964. The cultures, 8496 of them, went by van this time (instead of in a suitcase) to the permanent home in Rockville.
AMERICAN TYPE CULTURE COLLECTION
305
On May 1 and 2, 1964, to mark the dedication of the new facilities, the ATCC sponsored, with the American Society for Microbiology (as an adjunct to its 1964 annual meeting in Washington, D.C. ), a symposium on the topic “Stability in Dynamic Microbial Systems.” On May 2 the ATCC formally dedicated its facilities. More than 300 microbiologists and friends gathered at the new building for the ceremony. An address by Colin M. MacLeod, Office of Science and Technology, Executive Office of the President, titled “Some Thoughts on Microbial Taxonomy,” and one by Carl Lamanna, “Microbiology, Museums, and the American Type Culture Collection,” climaxed the event. So began a new era for the ATCC. The Collection was now deeply aware of its growing responsibilities in the advancement of the biological sciences and was determined to meet the challenge of the years ahead with imaginative and constructive programs. With permanent facilities some of its potentials could be realized. The possibility now existed for attracting new staff to improve the service potential and expand research in culture preservation and systematics. There too was the possibility of visiting scientists coming and studying the collections, bringing from time to time “new blood and new expertise to the organization. In the years immediately following, a number of scientists achieved milestones in ATCC’s history. In 1960 Shuh-wei Hwang began experimenting at the ATCC with the use of liquid nitrogen for the preservation of fungal cultures. She found that storage in liquid nitrogen provided a secure long-term preservation method for practically all the fungi tested (Hwang, 1966). Today all but a few of the 6000 strains in the Collection of Fungi have been successfully preserved in liquid nitrogen. In 1969 Hwang engaged in a cooperative study with J. P. San Antonio (San Antonio and Hwang, 1970) of the Mushroom Research Laboratory in Beltsville, Maryland, and several commercial spawn makers, which proved the value of liquid nitrogen in the long-term preservation of mushroom spawn. This was the first example of liquid nitrogen storage for the preservation of a commercial crop. H. H. McKinney ( McKinney et aZ. 1961) discovered that certain plant viruses in leaf tissue remained viable and unchanged in liquid nitrogen. This study and the experimental freeze-drying of plant viruses at the Plant Pathology Laboratory, Harpenden, England, laid the basis for a centralized Collection of Plant Viruses at the ATCC. W. Q. Loegering performed freezing and storage tests on Puccinia graminis var. tritici with great success (Loegering et al., 1961). He and his co-workers subsequently established a unique rust fungus collection at the ATCC. R. P. Hall, S. H. Hutner, and A. M. Elliott pointed to the need for a centralized protozoan collection in the early 60’s. L. S . Diamond (1964)
306
WILLIAM A. CLARK AND DOROTHY €I GEARY .
reviewed numerous published experiments describing successful freezepreservation of parasitic protozoa. Hwang, Davis, and Alexander ( 1964) of ATCC first succeeded in recovering the free-living protozoan, Tetrahymena pyriformis, from the frozen state. In 1967 limited funds became available to begin to develop the protozoan collection at the ATCC. In 1960 Rudolph Hugh, of the George Washington University, instituted a broadened program for morphological and biochemical testing and taxonomic evaluation of the bacteria at the ATCC. His system, expanded, is still in practice at ATCC today. P. A. Hansen, professor of bacteriology at the University of Maryland, was the ATCC’s first visiting investigator. In 1964, with a grant from the National Science Foundation, he organized an international taxonomic subcommittee on lactobacilli under the International Committee on Nomenclature of Bacteria. The ATCC served as a clearing house for receiving candidates for type and neotype strains of Lactobacillus species and for distributing subcultures to members of the subcommittee for study and evaluation. Out of this five-year effort Hansen (1968) produced a monograph titled “Type Strains of Lactobacillw Species.” Since Hansen’s time the ATCC has been host to a series of visiting scientists: Einar Leifson of Loyola University, Chicago to teach flagella staining; Colin Booth of the Commonwealth Mycological Institute in England to study Fusarium strains; Karen Schmidt, Gottingen, Germany to freeze photosynthetic bacteria; V. B. D. Skerman, Brisbane, Australia to examine gliding bacteria; John Stevenson, National Fungus Collections, Beltsville, Maryland to help update the nomenclature of the fungi for the ATCC catalogue; Norbert Weiss, Botanical Institute of the University of Munich to learn culture collection techniques; and Anand S. Saxena, University of Delhi, India to study taxonomy of thermophilic fungi. Today ATCC has a resident staff of 64. Half of these workers are trained biologists, the remainder constituting an experienced supportive unit. There are eight collections at the ATCC, each maintained in wellequipped, separate laboratories. Cultures are maintained by freezing and freeze-drying. Though primarily a service organization, the ATCC does research in culture preservation and systematics. The Collection of Animal Viruses, Rickettsiae and Chlamydiae is the largest single collection of such human and animal agents in the world. It is stocked with some 700 strains contributed by over 150 of the world’s most eminent virologists. T. 0. Berge became this collection’s first curator in 1964; on his retirement in 1971, D. A. Stevens became acting curator. The Collection of Bacteria, under the curatorship of Erwin F. Lessel, with R. L. Gherna as assistant curator, is, from a taxonomic standpoint, the worlds most comprehensive assemblage of bacteria. In the collection
AMERICAN TYPE CULTURE COLLECTION
307
are 9500 strains representing 1843 named species, as well as cultures with unusual characteristics useful in teaching, research, and commerce. Because it is not feasible for the Collection to employ all the many experts needed for its proper development, it has employed the concept of extramural curators. These experts in the field are invited, without compensation, to assist the ATCC in developing specific segments of the collections. As a current example, Ruth G. Wittler, of the Walter Reed Army Institute of Research, initiated in 1960 collections of Mycophmatales and L-phase variants. Other extramural curators are A. D. Alexander, M. P. Bryant, Elizabeth P. Cato, P. Hirsch, L. V. Holdeman, N. R. Krieg, Daisy A. Kuhn, W. E. C. Moore, M. Rogosa, R. M. Smibert, and S. W. Watson. The Collection. of Anamal Cell Lines, with J. E. Shannon as curator, is a repository for well-characterized, contaminant-free animal cell cultures for research in cancer, virology, biochemistry, and cytogenetics. This unique collection contains over 160 cell lines derived from over 30 different animal species. These include normal diploid cells, malignant cells, those grown in chemically defined media, cells with special chromosome configurations, biochemical markers, and virus susceptibilities, and those that produce specialized products, such as hormones, immunoglobulins, and pigments. In addition, this collection contains some 100 cell lines of human skin fibroblasts derived from individuals afflicted with various genetic disorders and other disease states, and the most comprehensive collection in the world of xeroderma pigmentosum cell lines. The Collection of Fungi with S. C. Jong as curator, includes 6000 cultures, which are vital tools in agriculture, industry, education, and medicine. S. W. Hwang and Roger Goos formerly were curators of this collection. Recently Sally Meyer joined the staff to develop the yeast collection and to do taxonomic research on DNA homologies in these organisms. The Collection of Bacteriophages consists of over 400 strains for many host genera. These bacteria-specific viruses are recognized as invaluable models for basic studies on the life cycle of viruses, as well as tools in epidemiology and genetics. The Collection of Plant Viruses and Plant Virus Antisera at the ATCC has grown to include 160 specimens of important plant disease agents and antisera for them, J. W. Blizzard is developing and maintaining this collection with help and advice from concerned plant pathologists. The Collection of Algae, with Shuh-wei Hwang in charge, is collaborating with the Indiana University Collection of Algae to study the preservation in liquid nitrogen of these plant forms, which are potential food sources and pollution indicators. The Collection of Protozoa is maintained by Roger Zieg with guidance
308
WILLIAM A. CLARK AND DOROTHY H. GEARY
from an active advisory group of parasitologists, protozoologists, and zoologists. Today 138 strains of free-living and parasitic protozoa are made available to educators and investigators. The collections at the ATCC still depend heavily upon scientists in outside institutions for their development. The compensation for such collaborators is their satisfaction in contributing to the improvement of this national culture collection. Overseeing, guiding, and advising the ATCC is the Board of Trustees, made up of representatives from 13 different scientific societies concerned with microbiology, and four trustees-at-large. Trustees-at-large are appointed by the Board, without reference to scientific society affiliation, to furnish special expertise. The Board is the policy-making body for the Collection, and it provides continuity. Board members serve without remuneration, and each gives six years of service to the ATCC, representing his society’s interests in the Collection. The ATCC now maintains 16,000 strains of microorganisms, viruses, and animal cell lines and stores around 500,000 ampules of culture material. In 1972, over 21,000 cultures were shipped. Clients are largely from universities, hospitals and industry, although workers in goverment and research institutions account for a significant portion of the requests. The ATCC supplies a wide range of cultures for industrial use in the production of antibiotics, food and beverages, flavoring agents, bacterial insecticides, drugs, and vitamins. The ATCC also serves as a distribution center for specialist collections, such as that of the Institute of Microbiology at Rutgers University and M. P. Starr’s International Collection of Phytopathogenic Bacteria at the University of California at Davis. The Food and Drug Administration and certain industrial companies also deposit strains in the ATCC for distribution. Three separate catalogues are published every two years: the Catalogue of Strains (bacteria, fungi, bacteriophages, plant viruses, algae, and protozoa); the Registry of Animal Cell Lines; and the Registry of Viruses, Rickettsiae and Chlamydiae. These serve not only as lists of extant living cultures but also as useful dictionaries of names and uses of microorganisms and as manuals of maintenance methods. The staff also publishes annual reports, scientific papers, and monographs on pertinent subjects. With the annual budget now at 1.1 million dollars, present financial support for the ATCC is roughly one-half from sale of reference cultures and fees from other services performed, and half from grants and contracts from the National Institutes of Health, the National Science Foundation and the U.S. Department of Agriculture. In 1971 a reduction and the projected phasing out of NIH annual
AMERICAN TYPE CULTURE COLLECTION
309
support precipitated still another financial crisis. Twelve staff positions were eliminated and stringent belt-tightening procedures were instituted. The Trustees have launched a Development Program to find a permanent and stable source of support for the Collection. The program is directed primarily toward broadening the base of government support, with industry, professional societies, and individuals being asked to provide interim maintenance funds, and philanthropic foundations being asked to support specific projects. The Board believes that the financial support of this national collection represents a proper function of government, because the strains assembled here are part of our country’s scientific wealth and the general public benefits because use of the cultures produces advances in research, education, agriculture, industry, medicine, and public health. The ATCC‘s obligation to the future is to preserve its growing wealth of cultures for posterity and to extend and expand the work of those individuals who have labored through the years to keep the Collection viable. After over half a century of struggle and service the ATCC is prepared to meet the needs of the future with courage, vision, and enthusiasm. REFERENCES
Diamond, L. S. (1964). Cryobiology 1, 95. Hansen, P. A. ( 1968). “Type Strains of Lactobacillus Species.” American Type Culture Collection, Rockville, Maryland Hwang, S.-w. (1966). Appl. Microbiol. 14, 784-788. Hwang, S.-w., Davis, E. E., and Alexander, M. T. (1964). Science 144, 64. Loegering, W. Q., McKinney, H. H., Harmon, D. L., and Clark, W. A. ( 1961). Plant Dis. Rep. 45, 384. McKinney, H. H., Greeley, L. W., and Clark, W. A. ( 1 9 e l ) . Plant Dis. Rep. 45, 755. Rogers, L. A. (1914). J. Infec. Dis. 1, 100. San Antonio, J. P.,’andHwang, S.-w. (1970). J . Amer. SOC. Hort. Sci. 95, 565. Smith, N. R., Gordon, R. E., and Clark, F. (1952). “Aerobic Mesophilic Sporeforming Bacteria,” USDA, Agr. Monogr. No. 17. U.S. Govt. Printing Office, Washington, D.C.
This Page Intentionally Left Blank
Microbia I Penicillin Acylases E. J. VANDAMMEAND J. P. VOETS Laboratoy of General and Indwtrial Microbiology, Uniuersity of Gent, Gent, Belgium I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. History ..................... 111. Biotransformation of Penicillins into 6-Aminopenicillanic Acid (6-APA) . . . . . . . . .. . . . . . . . . . . . .. . A. General Considerations . . . . . . . . . . . . . . . . . . . . . . . B. Penicillin Acylase-Producing Strains . . . . . . . . . . . C. The Substrate: Penicillins . . . . . . . . . . . . . . . . . . . . . D. The End Product: 6-Aminopenicillanic Acid (6APA) ................... E. Detection of 6-APA and Determination of Penicillin Acylase Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Screening Procedures . . . . . . . , . . . . . . . . . . . . . . . . . . . . V. Penicillin Acylases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Penicillin V Acylases . . . . . . . . . , . . . . . . . . . . . . . . . C. Penicillin C Acylases . . . . . . . . . . . . . . . . . . . . . . . . D. Ampicillin Acylases . . . . . . . . . . . . . . . . . . . . . . . . . . , VI. Physiological Role of Penicillin Acylases . . . . . . . . . . . . VII. Coexistence of Acylase and p-Lactamase . . . . . . . . . . . . VIII. Penicillin Acylase and the Cephalosporins . . . . . . . . . . . . IX. Nonmicrobial Penicillin Acylases . . . . . . , . . . . . . . . . . . . x. Aspecific Amidases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XI. Chemical Transformation of Penicillins into 6-APA . . . XII. Concluding Remarks . . . , . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
I.
311 312 316 316 316 317 320 322 323 325 325 326 336 346 347 352 354 356 356 358 359 361
Introduction
There has been marked success in utilizing microorganisms in the preparation and transformation of intermediates for the synthesis of steroids, but less interest has been directed to microbial transformations of known antibiotics needed to prepare new and therapeutically useful ones. Many antibiotics have been modified, and thereby inactivated or degraded, by microbial enzymes, but no practical application has been found for most of the products (Jarvis and Berridge, 1969; Sebek and Perlman, 1971) . Microbial transformation of antibiotics into useful compounds is mostly concentrated upon the production of 6-aminopenicillanic acid ( 6-APA ) by enzymatic hydrolysis of biosynthetic penicillins. 6-APA is the starting material for the production of the semisynthetic penicillins and is produced on an industrial scale by enzymatic removal 311
E. J . VANDAMME AND J. P. VOETS
312
of the side chain of penicillin G or V ( Carrington, 1971). The enzyme responsible is penicillin acylase, which is found in many bacteria and fungi (Cole, 1967). The availability of the penicillin nucleus has made it possible to prepare semisynthetic penicillins with superior clinical effectiveness (Price, 1969).
It. History The antibacterial substance, detected-by Fleming (1932) in the fermentation broth of Penicillium notatum, was found to display an inhibitory effect against a wide range of gram-positive bacteria. Fleming (1932) called this substance penicillin. Some years later, Chain et al. (1940) obtained a relatively pure penicillin preparation and demonstrated in animals its antibiotic activities in vivo. Soon after, Abraham et al. (1941) and Florey and Florey (1943) demonstrated that penicillin administered to man displayed a low toxicity and was therapeutically effective. These experiments marked the start for the development of high penicillin-yielding strains. By selection of appropriate Penicillium chrysogenum strains and by submerged fermentation in special growth media, high titers of penicillin were obtained. These antibiotic-rich growth media allowed easier purification of the product and at the same time revealed that always a mixture of chemically related penicillins was produced. These natural penicillins are summarized in Table I. It was now evident that natural penicillins all have a common nucleus TABLE I PENICILLINB FOUND IN Penicillium chrysogenum FERMENTATION BROTHS
Penicillin
R
p-Hydroxybenzyl (XI 2-Pentenyl (F)
CH3CH2CH=CHCH&0
n-Amy1 (dihydro-F)
CH,CH,CH,CH,CH&O
n-Heptyl (K)
CH,CH, CH,CH, CH,CH, CH, CO
L-
6- Aminoadipyl (iso-N)
6-Arninopenicillanic acid
6- HOOCCH(NH,) (CH,),CO
H
MICROBIAL PENICILLIN ACYLASES
313
and differ chemically only in their side-chain structure. Furthermore, exactly these variations exert a marked influence on their antibacterial activity. Moyer and Coghill (1946, 1947) demonstrated that the relative amounts of the naturally occurring penicillins G, X, F, dihydro-F, and K were dependent on the composition of the fermentation medium. Subsequently, they succeeded in the selective fermentation of penicillin G by adding to the fermentation broth phenylacetic acid as a precursor. Behrens et aZ. (1948) tested a wide range of products as possible side chain precursors, and in this way the biosynthetic penicillins-penicillin V, penicillin 0, and penicillin S-were obtained. Particularly penicillin V was widely applied, as it was found to be acid stable. However, soon it was realized that only monosubstituted acetic acid derivatives could act as effective precursors. Besides penicillin G and V, none of the penicillins obtained in this way were found to have higher qualifications. Attempts to change the biosynthetic penicillins chemically into more effectiveantibiotics failed also. Indeed, neither 20 years of fermentation research nor chemical modification processes resulted in the production of a penicillin displaying more attractive qualities than did “Fleming’s penicillin.” In addition, Abraham and Chain (1940) described the inactivation of the biosynthetic penicillins by a bacterial enzyme : penicillinase (p-lactamase) (Ambler and Meadway, 1969; Citri and Pollock, 1966; Pollock, 1971; Saz, 1970; Richmond and Sykes, 1973). An ever increasing number of gram-positive and gram-negative microbes synthesize this enzyme, which hydrolyzes the p-lactam structure of the penicillin molecule, resulting in the formation of the inactive penicilloic acid. Especially the wide application of the biosynthetic penicillins in therapy, combined with the transferable nature of antibiotic resistance (R.factors), resulted in the selection and spreading of p-lactamase producing-hence penicillin-resistant-bacteria (Anderson, 1968; Bach et al., 1966; Datta and Kontomichalou, 1965; Fullbrook et al., 1970; Garber and Friedman, 1970; Hamilton-Miller, 1968; Morrin and Malamy, 1970; Novick, 1969; Silver and Falkow, 1970; Sykes and Richmond, 1970; Richmond and Sykes, 1973). In view of this resistance, the need for new penicillins became highly urgent. The elucidation of the structure of penicillin N (cephalosporin N or synnematin B ) yields initial indications that variations in the side chain structure influenced not only the antibacterial activity, but also the antibacterial spectrum, of penicillins (Abraham et aZ., 1955). Stimulated by the fact that side chain modifications might influence in a positive way the fundamental properties, such as antibacterial spec-
314
E. J . VANDAMME AND J . P. VOETS
trum, acid stability, and intrinsic activity, and forced by the unfortunate spread of p-lactamase-producing bacteria, intensive research resulted in the late 1950s in a solution of “the penicillin problem.” Sheehan and Henery-Logan (1959) succeeded in the total chemical synthesis of penicillin V, but the complexity and the low yield of the process precluded industrial application. Batchelor et al. (1959) and Ballio et al. (1961) reported the isolation of the penicillin nucleus from a Penicillium chrysogenum fermentation medium to which no precursor had been added. In this way, they obtained by fermentation a penicillin without side-chain structure and named it 6-aminopenicillanic acid (6-APA) . Six years earlier, Kato (1953) had reported the same phenomenon. However, already in 1950 Sakaguchi and Murao had demonstrated that 6-APA could be produced more easily by hydrolyzing penicillin G with an enzyme (penicillin amidase) present in the mycelium of a sulfathiazole-resistant Penicillium chrysogenum and Aspergillus oryzae, but these findings did not elicit full attention (Murao, 1955; Sakaguchi and Murao (1950) and, Murao (1955). Batchelor et al. (1961b) were the first to obtain crystalline 6-APA by a fermentation process, and they proved the structure of the compound by converting it with phenylacetylchloride into penicillin G. Thus, by adding chemically the appropriate side chain to 6-APA, it would be possible to synthesize even new penicillins. The direct fermentation of 6-APA by Penicillium chrysogenum grown in a complex and costly medium was the original procedure adopted for the production of the first semisynthetic penicillin: phenethicillin. However, fermentation of 6-APA was a laborious process, yielding low broth potencies of 6-APA invariably accompanied by penicillin production. The isolation and purification of the 6-APA involved complex procedures, so the biosynthetic route of 6-APA production is now obsolete ( Carrington, 1971 ) . However, as a result of the isolation of 6-APA from fermentation broths, the enzymatic deacylation procedure of Sakaguchi and Murao (1950) revived interest. As a result of intensive screening programs, deacylation of penicillins by specific enzymes was reported to occur in bacteria, molds, yeasts, actinomycetes, and even in plant and animal tissues ( Hamilton-Miller, 1966). Soon it became evident that higher 6-APA levels, coupled with an easier isolation procedure, were obtainable by the simple transformation process described earlier by Sakaguchi and Murao ( 1950) and Murao (1955). Now high-productive penicillin acylase strains are used to produce &APA on an industrial scale, starting from the biosynthetic penicillins
MICROBIAL PENICILLIN ACYLASES
315
G and V. In this way, a wide range of semisynthetic penicillins were prepared (Doyle et al., 1961a,b), and at this moment ampicillin, methicillin, cloxacillin, oxacillin, and carbenicillin are widely applied in therapy (Acred et al., 1967; Bodey and Terrell, 1968; Nayler et al., 1962; Price, 1969). Semisynthetic penicillins are synthesized by direct addition to 6-APA of an appropriate side chain acid chloride or anhydride. Side chain groups are also coupled to 6-APA with a good yield in the presence of N,N-bicyclohexyl carbodiimide ( Sheehan and Henery-Logan, 1959) or N,N-dimethylchloroformiminium chloride ( Novak and Weichet, 1965). In addition, 6-APA is also acylated enzymatically on the assumption that penicillin acylase can catalyze the reverse reaction too (Bauer et al., 1960; Cole, 1969c,d; Nara et al., 1971a; Okachi et al., 197313; Rolinson et al., 1960). Attempts to prepare new semisynthetic penicillins are still proceeding, as is evidenced by the recent synthesis of carboxyl-3-thienylmethylpenicillin, substituted ureidopenicillins, and 6-p-amidinopenicillanic acids, which are now intensively being tested as to antibacterial properties (Bodey and Deerhake, 1971; Bodey and Stewart, 1969, 1971; Lund and Tybring, 1972; Neu and Winshell, 1971). In addition to the obsolete biosynthesis of 6-APA by precursor-free fermentation processes and the widely used enzymatic hydrolysis of penicillins into 6-APA, purely chemical procedures were recently applied to the preparation of 6-APA on an industrial scale (Carrington, 1971; Fosker et al., 1971; Weissenburger and Vanderhoeven, 1970). As the structure of cephalosporins is analogous to that of penicillins (Abraham and Newton, 1961), microbial acylases were applied to convert cephalosporin C into its nucleus 7-aminocephalosporanic acid (7-ACA). However, as yet no acylase has been found that can carry out in an economic way this important transformation (Demain et al., 1963; Hamilton-Miller et al., 1970), as 7-ACA is the starting material for the industrial synthesis of semisynthetic cephalosporins ( Axelrod et al., 1971; Chauvette et al., 1962; Griffith and Black, 1964; Loder et al., 1961; Morin et al., 1962; Sassiver and Lewis, 1970; Walton, 1964; Wick, 1967) . From these considerations, it appears that penicillins may still be widely used-although the longest known antibiotics and in spite of the discovery of many other antibiotics-largely because of the discovery of the enzyme penicillin acylase, which is capable of removing the side chain of penicillin V or G, thus providing an easy way to obtain the 6-APA structure for the industrial production of semisynthetic penicillins (Chain, 1971).
E. J . VANDAMME AND J . P. VOETS
316
111.
Biotransformation of Penicillins into 6-APA
A. GENERALCONSIDERATIONS Biotransformation processes of penicillins into 6-APA have been described that use, as a source of acylase, intact bacterial cells, fungal mycelium, fungal spores, crude cell extracts, purified enzyme preparations, stabilized enzymes, and supernatant liquid of extracellular acylaseproducing microorganisms. Although a large number of microorganisms have been found to display penicillin acylase activity, detailed studies of penicillin acylases have been focused on only a few species, and most of the research has been carried out with suspensions of intact microorganisms as the enzyme source. Only recently, attempts have been made to isolate and purify penicillin acylases and to stabilize these enzyme preparations (Dinelli, 1972; Marconi et al., 1973; Ryu et al., 1972a,b; Self et al., 1969; Warburton et al., 1972). In view of recently developed chemical procedures for the preparation of 6-APA (Fosker et al., 1971), the introduction of methods of insolubilization of enzymes, which give stability and prolonged high potency activity resulting in a continuous enzyme-reaction procedure, seems bound to improve the economics of enzymatic processes that may be on the borderline of success or be in competition with chemical procedures (Faith et al., 1971; Smiley and Straidberg, 1972; Wingard, 1972).
B. PENICILLIN ACYLASE-PRODUCING STRAINS Large-scale surveys have indicated the rather small incidence of penicillin acylase-producing microorganisms; nevertheless, this enzymatic activity is not restricted to a few genera, but appears to be distributed widely among the different genera of microorganisms. There are also indications that penicillin acylase is produced by individual strains and is not a property of the species. Starting from “natural” penicillin acylase-producing strains, the penicillin industries have obtained high-yielding constitutive mutants: once a suitable microorganism has been isolated, mutation and genetic techniques are employed to select organisms with improved abilities. Concerning the industrial strains and their applications, this paper is bound to present an incomplete view of the true situation. Competition within the fermentation industries is rather fierce, and the resultant
317
MICROBIAL PENICILLIN ACYLASES
secrecy is so great that nobody knows exactly what kind of investigations and processes are in progress ( Carrington, 1971). As a result of this lack of information and the inaccessibility of patent literature, much of the work cited here inevitably is based on published data.
C. THE SUBSTRATE: PENICILLINS The penicillin molecule consists of a nucleus linked with a side-chain group. The nucleus, known as 6-aminopenicillanic acid, is built up of two amino acids, L-cysteine and D-valine, twisted together biogenetically into a cyclic dipeptide. However, the biosynthesis process of the natural and biosynthetic penicillins is far from being elucidated (Abraham, 1971; Demain, 1966; Maier and Groger, 1972). Chemically, penicillins are a condensation product of a thiazolidine ring, a p-lactam ring, and a suitable side chain (Fig. 1). Penicillins in general are quite stable in the dry state, but once in solution, different degrading products can arise, depending on environmental conditions. Rapson and Bird (1963) and Hou and Poole (1971) reported on the ionization constants of penicillins (pK, I+ 2.73) and on their hydrolysis products. The effect of ionic strength on the stability of penicillin G was studied by Lindsay and Hem (1972). The influence of acids, bases, and alcohol (Dennen, 1967; Schwartz, 1965a,b; Schwartz and Buckwalter, 1962; Segelman and Farnsworth, 1970), and also of water, is important in regard to stability. Dennen and Davis (1961), Schwartz ( 1965a,b), and Kinget and Schwartz (1969) described the kinetics of formation of degradation products in acidic solutions, and Finholt et al. (1965) reported on the catalytic effect of buffers on the degradation of penicillin G in aqueous solution. Recently, Cole et al. (1973) investigated the fate of different penicillins in healthy humans as a preliminary to studying what happens in humans with various types
j
L-Qsteine
Phenoxyacetic acid
o-vdine
e 2 - C O O H ,/
/
co
O $-C -(&.
I
Side chain
0-b'ctam Thiaeolidine ring ring
FIG.1. Basic structure of penicillin (penicillin V).
318
E. J. VANDAMME AND J. P. VOETS
of bacterial infections. Hou and Poole (1969, 1971) recently reviewed the physicochemical properties and the degradation reactions in relation to the structure of penicillins. The reaction of cysteine and related compounds with penicillins, resulting in a loss of antibacterial activity, is described by Wagner and Gorman ( 1971). Polymerization of penicillins during storage as concentrated aqueous solutions was reported by Smith and Marshall ( 1971). The formation of antigenic polymers in aqueous solutions of the penicillins was recently studied by Dewdney et al. (1971) and Ottens et al. (1971). Although the stability of the penicillins in the dry state and in aqueous solution has been well established, less is known about their stability in microbiological agar media. According to Ryan et al. (1970), the stability of different antibiotics, including penicillins, in agar plates is very good and comparable with that of the dry product. Spontaneous degradation of the penicillins in acid solution consists mainly in a rearrangement into penillic acids. In aqueous and alkaline solution, degradation results in the formation of penicilloic acids, which can further decarboxylate into penilloic acids. p-Lactamase activity also ends in the formation of pcnicilloic acids. Both processes result in a loss of antibacterial activity of the antibiotics. Acylase activity results in the formation of 6-APA, although according to Batchelor and Cameron-Wood ( 1962), penicillins can spontaneously hydrolyze into 6-APA, however, only for a small percentage (0.2%). Apart from the action of p-lactamases and acylases, which can be produced by the same microbial strain, penicillins could be transformed by the action of some Enterobacteriaceae into penicillamine during the spheroplast-induction process ( Dulong de Rosnay et al., 1970), These different degrading patterns, involving spontaneous chemical and enzymatic degradation, are represented in Scheme 1. The estimation of penicillins and penicillin destruction has been reviewed by Hamilton-Miller et al. ( 1963). Paper chromatographic detection of penicillins, combined with bioautographic methods were widely used (Cole, 1967; Thomas, 1961). Thin-layer chromatography (TLC), whether or not combined with bioautographic methods, was intensively used to separate individual pcnicillins from a mixture of penicillins ( McGilveray and Strickland, 1967; Nussbaumer, 1962; Pan, 1973) . However, the separation and identification of the different degradation products of one kind of penicillin by TLC has only recently been carried out (Birner, 1970; Fooks and Mattok, 1969; Nara et al., 1971a; Vandamme and Voets, 1972a ) . Colorimetric methods were described by Ford ( 1947), Boxer and Ever-
319
MICROBIAL PENICILLIN ACYLASES
H,N-HC-C/ H S c/'~s H,C ,C-CH-COOH H,C I SH
I I OC-N-C,
z
I'CHs H COOH
6-APA
Penicillamine
0 -CH&O-NH-HC-C
H,S
I I OC-N-C
/ /CHS ?CHs
HCOOH
Penicillin V
Penillic acid
Penicilloic acid
Penilloic acid
SCHEME1. Degradation patterns of penicillin.
ett (1949), Pan (1954), and Holm (1972) for the determination of penicillins containing an intact p-lactam ring. Bundgaard and Ilver (1972) described for the determination of penicillins a new spectrophotometric method based on the measurement of penicillenic acid mercuric mercaptides of the penicillins, Chapman et al. (1964) identified penicillins and penicilloic acids by their infrared spectra. Pruess and Johnson ( 1965, 1967), Spencer ( 1968), and Gatenbeck and Brunsberg (1968) detected penicillins and degraded products by radiochromatographic procedures, using S35--labeledpenicillins as substrates. Gas chromatographic determination of penicillins was described by Evrard et al. (1964) and Hishta et al. (1971). Iodometric titration methods ( Alicino, 1961; Perrett, 1954 ), colorimetric methods ( Pan, 1954; Sargent, 1968), and spectrophotometric methods (Lepidi and Nuti, 1971) have been described for the quantitative determination of penicilloic acids. Pal (1969) described a colorimetric dosage of penicillamine, and Thomas and Broadbridge ( 1970) described the electrophoretic separation of penicillins and penicilloic acids. The properties of the end product of penicillin acylase activity, 6-APA, and methods for its detection will be reviewed in more detail below.
320
E. J . VANDAMME AND J . P. VOETS
D. THEENDPRODUCT:6-AMINOPENICILLANIC ACID (6-APA) By the hydrolytic action of penicillin acylases, penicillins are hydrolyzed into 6-APA. Batchelor et al. (1959) found that 6-APA also could be produced by fermentation in precursor-free penicillin-fermentation media, and Sheehan and Henery-Logan ( 1959) synthesized 6-APA chemically. Recently nonenzymatic conversions of penicillins into 6-APA are described by Fosker et al. (1971), Weissenburger and Vanderhoeven (1970), and Chauvette et al. (1972). 6-APA is in fact an amino acid. Rapson and Bird determined by electrometric titration the pK values 2.30 and 4.91 (1963). All penicillins could be considered to be N-acyl derivatives of the common nucleus 6-APA. Although antibacterial properties of penicillins are related to the structure of their side chain and 6-APA can be considered as a penicillin without side chain, it displays antibacterial activity (Rolinson and Stevens, 1961). The mechanism of action of 6-APA as an antibiotic is identical to that of penicillin G (Highton and Hobbs, 1971; Kats et al., 1968a,b; Mainer and Perkins, 1970; Lawrence and Strominger, 1970; Strominger et al., 1971; Tipper and Strominger, 1968). According to Batchelor et al. (1961c), 6-APA is an inducer of p-lactamase; Walton ( 1964) suggested that 6-APA inhibits p-lactamases, hydrolyzing cephalosporin C. Also 6-APA contains a p-lactam ring, and hence it is susceptible to p-lactamase activity and to spontaneous and alkaline hydrolysis. These processes result in the formation of penicic acid ( D-4carboxy-5,5-dimethyl-~-amino-2-thiazolidine acetic acid ) , the penicilloic acid of 6-APA, also named 6-aminopenicilloic acid ( Hamilton-Miller et al., 1963; Savitskaya et al., 1972). By decarboxylation, penicic acid is transformed into CDAT ( ~-4-carboxy-5,5-dimethyl-2-aminomethylthiazolidine) (Nys et at., 1972, 1973). Less is known about the stability of 6-APA in aqueous solution. Dennen (1967), Vandamme et al. (1971a), and Libinson (1971) reported on the degradation kinetics of 6-APA in buffer solutions in function of pH and temperature and found that 6-APA in solution displays maximal stability in the pH region of 8. Spontaneous degradation of 6-APA consists of p-lactam hydrolysis resulting in the formation of penicic acid, coupled with a dimerization process (Grant et al., 1962). According to Grant et al. (1962), spontaneous and enzymatic p-lactam hydrolysis of 6-APA is associated with a polymerization of the produced penicic acid. 6-APA is in some aspects more stable than penicillins as it is relatively stable toward acids. However, COP, which has no infiuence on penicillin stability, reacts readiIy
321
MICROBIAL PENICILLIN ACYLASES
with 6-APA, resulting in the formation of 8-hydroxypenillic acid (Batchelor et al., 1961e; Johnson and Hardcastle, 1961). Moss and Cole (1964) showed that 6-APA reacts readily with reducing sugars, carbohydrates and related substances, and frequentin. Godtfredsen et al. (1967) were able to demonstrate photochemical transformation of 6-APA into an unknown compound by ultraviolet treatment of aqueous 6-APA solutions. Electrophoretic separation of 6-APA revealed that crystalline 6-APA preparations contain penicillin-related substances ( named factors 1, 2, and 3 ) , which display each a different antibacterial activity (Batchelor et al., 1962). Recently, Shaltiel et al. (1970, 1971) and Ottens et al. (1971) found that crystalline 6-APA contains protein impurities responsible for allergic reactions, The antigenic impurity found in 6-APA appeared to be derived from the penicillin acylase preparation used for the removal of the side chain from penicillins. They were able to reduce the immunological manifestations of 6-APA by treatment with an insolubilized protease from Streptomyces griseus (Shaltiel et al., 1971). The degradation patterns of 6-APA are represented in Scheme 2.
I
6-APA
OH
';'.
HOOC N-C, H
H COOH
Penicic acid
H S HSC, 3 , ,c cH-c-~-NH--HC-CA I I1 I I YkH, H,C I I C-NH H 0 OC -N -C, H COOH
6-APA Dimer
HPN-HPc,H,SC/cH, C I
HN-C.
I.cH, H COOH
CDAT
I
Poly- 6- APA
SCHEME2. Degradation patterns of 6-aminopenicillanic acid (6-APA). (CDAT, ~-4-~arboxy-5,5-dimethyl-2-aminomethyfthiazolidine ).
322
E . J . VANDAMME AND J. P. VOETS
OF 6-APA E. DETECTION
AND DETERMINATION OF PENICILLIN ACYLASEACTIVITY
At present, penicillin acylase activity is demonstrated by the detection of its end product: 6-APA. After paper-chromatographic separation of penicillins and 6-APA, the 6-APA produced can be treated with phenylacetyl chloride and converted again into penicillin G, which is then demonstrated by bioautographic methods. Despite the unstable nature of the products involved, this time-consuming bioautographic phenylacetylation technique was until recently the only one widely applied that allows a quantitative 6-APA determination (Batchelor et al., 1961b; Lemke and Nash, 1972; Uri and Sztaricskai, 1961). The colorimetric hydroxylamine method of Ford (1947) and Boxer and Everett (1949) for penicillin determination is also applicable to 6-APA determination if penicillins are previously extracted from the mixture at pH 2 with n-butyl acetate. This quantitative, although timeconsuming, 6-APA determination method is still widely used (Nara et al., 1971a). Ivashkiv (1964) and Bomstein and. Evans (1965) described an automated colorimetric 6-APA dosage method for fermentation media. Niedermayer (1964) and Chiang and Bennett (1967) determined the liberated side chain by means of gas chromatography as a measure of the penicillin acylase activity. Alicino ( 1961) proposed an iodometric titration method; this method, however, was useful only for quantitative determination of pure 6-APA. Chapman et al. (1964) used infrared spectroscopy to determine 6-APA, after paperand Pruess and Johnson ( 1965), working with peni~illin-~~S, chromatographic separation could detect 35S-labeled6-APA by means of radioautography. Enzymatic preparation of 35S-labeled 6-APA was described by Nathorst-Westfelt et al. (1963) and Pruess and Johnson ( 1965,1967) . Huang et al. (1960) extracted the nonconverted penicillin from the reaction mixture and treated the 6-APA with a p-lactamase. Subsequently, the resulting penicic acid was colorimetrically determined by the ninhydrin reaction. Nys et al. (1973) described a direct spectrophotometric estimation of 6-APA in benzylpenicillin hydrolyzates; the estimation was based on the reaction of the 6-APA amino group with p-dimethylaminobenzaldehyde in acid medium. A spectrophotometric method for the determinaton of 8-hydroxypenillic acid, a degradation product of 6-APA, was described by Hull et al. (1964).
MICROBIAL PENICILLIN ACYLASES
323
Recently, a method was described whereby 6-APA can be determined spectrophotometrically, D-glucosamine being used as reagent, in the presence of other penicillin-derived compounds without prior separation ( Shaikh et al., 1973). Cole (1964, 1966, 1969a), Sjoberg et al. (1967), and Vanderhaeghe et al. ( 1968) estimated 6-APA quantitatively by titration of the liberated side-chain acid. Findlater and Orsi (1973) described an indicator method for penicillin acylase assay; in this procedure, proton release was measured during the hydrolysis process. Bondareva et al. (1969a,b), Vandamme et al. (1971a,b), Vandamme and Voets (1972a,b), Nara et al. (1971a), and Okachi et al. (1973b) detected 6-APA qualitatively in the presence of intact penicillins and other degrading products by means of thin-layer chromatography. Serova et al. (1973) used this technique for quality control of 6-APA samples. A semiquantitative method for the determination of 6-APA, based on thin-layer chromatography, was reported by Korchagin et at. ( 1971). Oostendorp (1972) described a microbiological plate assay for the quantitative determination of 6-APA, using a strain of Serratia marcescens ( ATCC 27117) as test organism. Intact biosynthetic penicillins do not interfere with the assay, so extraction procedures are no longer needed. Recently Bauer et a2. (1971) described a simplified determination of the enzyme penicillin acylase. I t is based on the ability of the acylase to hydrolyze phenylacetyl-L-asparagine to L-asparagine and phenylacetic acid. As quickly as these products are formed, the L-asparagine is hydrolyzed into aspartic acid and ammonia by an excess of added L-asparaginase. The released ammonia is determined with Nessler’s reagent. However, it is not clear whether this indirect penicillin acylase determination is applicable to all penicillin acylases. Probably, this procedure is useful for Escherichia c d i ATCC 9637 and ATCC 11105 acylases, and for acylases whose specificity is not closely connected with the nucleus structure of the penicillin molecule. However, in view of the unstable nature of the products involved and because a convenient and rapid quantitative 6-APA determination method is not yet available, indirect quantitative determination procedures deserve full attention.
IV.
Screening Procedures
Originally penicillin acylase was thought to occur preferentially in penicillin-producing fungi. But this enzymatic activity was soon reported also in fungi, yeasts, actinomycetes, and bacteria that produced no penicillin.
324
E. J . VANDAMME A N D J. P. VOETS
Several authors tried to select on well-defined principles microorganisms that produce penicillin acylase. Holt and Stewart (1964b) argued that penicillin acylase activity is correlated with penicillin resistance or degradation. However, penicillin acylase activity is not so widespread as p-lactamase activity, as is evidenced by the results of the screening programs elaborated by Ayliffe (1963, 1965), Demain et al. (1963), Cole and Sutherland ( 1966), Hamilton-Miller ( 1966), Rozansky et al. ( 1969), Vandamme ( 1972), and Nara et al. (1971a,b). Ayliffe (1965) was not able to detect acylase activity among 148 bacterial strains, although 55 of them displayed p-lactamase activity. Demain et al. (1963) also failed to detect penicillin acylase among 201 microorganisms, although 52 of them possessed a p-lactamase. Cole and Sutherland (1966) tested 148 bacterial strains, only 10 displayed penicillin acylase activity, although 70 were p-lactamase positive. Vandamme ( 1972) tested 184 microorganisms, including fungi, yeasts, actinomycetes, and bacteria; 9 molds and 1 bacterial strain were found to produce a penicillin acylase, and 38 microorganisms displayed p-lactamase activity. Rozansky et al. ( 1969) tested 125 p-lactamase-producing strains and found acylase activity in only 10 of them after subculturing with phenyl acetate as inducer. Nara et al. (1971a) screened 251 strains of bacteria, 229 of actinomycetes, 2 of yeasts, and 37 of basidiomycetes. They found that most of the organisms decomposed penicillin G or V to some extent or completely; the resulting product was mostly penicilloic acid. Among this large number of strains, only 9 bacteria and 5 actinomycetes were selected as penicillin acylase producers. Kameda et al. (1961) proposed a selective isolation procedure based on growth on a mineral medium containing benzylpenicillin or phenylacetic acid as sole source of carbon, and they correlated hydrolysis of acyl-m-amino acids with the presence of penicillin acylase. Batchelor et al. (1961d) isolated 215 species of fungi, yeasts, and actinomycetes and selected 38 penicillin acylase-producing strains after growth on corn steep liquor (CSL) media and related penicillin fermentation broths. Huang et al. (1963) detected 60 species of bacteria and actinomycetes displaying penicillin acylase activity, among 329 isolates grown on media containing CSL, molasses, or yeast extract. Walton (1964) used amide substrates in a soil-enrichment procedure for isolating microorganisms that would liberate 7-aminocephalosporanic acid (7-ACA) from cephalosporin C in order to detect cephalosporin C acylases. As a screening procedure and assay system for the detection of penicillin acylase-producing strains, an indirect method was applied, based on the liberation of yellow p-nitroaniline from the colorless N-phenylacetylnitroaniline. Haupt and Thrum (1967) isolated different Streptomyces species dis-
MICROBIAL PENICILLIN ACYLASES
325
playing amide synthetase activity as possible penicillin acylase-producing strains. On the assumption that compounds with a structural similarity or an identical stereo configuration to the penicillin molecule, or to its side chain p-lactam ring part, might behave as selective agents in the screening for penicillin acylase-producing microorganisms, Vandamme and Voets (1973) and Vandamme (1973) were able to isolate acylaseproducing microorganisms from soil. They used a mineral medium to which N-acetylglycine, N-glycylglycylglycine, or phenylacetamide was added as sole source of carbon and nitrogen. Recently, Golub et al. (1973) described mutants of Escherichia coli with damaged glucose transport as organisms that produce penicillin acylase. It stemmed clearly from the results of these test programs that penicillin acylase is a rather scarce enzymatic activity of microorganisms, and detection of a penicillin acylase-producing strain as yet is difficult to predict and is associated with a bit of luck. However, as none of the above-mentioned methods appear to be promising, the most productive screening programs are effected by the trial and error method. The most satisfactory procedures for selecting microorganisms that produce penicillin acylase are those involving the direct detection of 6-APA, by one of the methods described above, when the organism, after isolation and culturing, is incubated in the presence of penicillins. However, the indirect determination of E . coli penicillin acylase activity as recently described by Bauer et al. (1971) needs full attention and might be extended and applied to detect new acylaseproducing microorganisms.
V.
Penicillin Acylases
A. NOMENCLATURE Not only is the penicillin molecule susceptible to enzymatic hydrolysis of the p-lactam ring (by p-lactamase), but the peptide linkage, which joins the side chain to the nucleus of the molecule (6-APA) is also hydrolyzable by enzymatic action. The enzyme responsible for this enzymatic cleavage, which results in the formation of 6-APA and the side chain acid, has been given different names: penicillin amidase, penicillin acylase, penicillin amidohydrolase, penicillin splitting and synthesizing enzyme, penicillin acyltransferase, penicillin deacylase, and penamidase ( Hamilton-Miller, 1966). We refer to the enzyme throughout this paper as penicillin acylase (EC 3.5.1.11). Penicillin acyIases hydrolyze preferentially penicillin G or penicillin
326
E. J . VANDAMME AND J. P. VOETS
V. Penicillin G acylases mostly occur in bacteria, and penicillin V acylases are usually found in molds and actinomycetes, although the number of exceptions to this general rule is rising quickly (Cole, 1964; Nara et al., 1971a, 1972; Vandamme et al., 1971b; Voets and Vandamme, 1972; Okachi et al. (1973a) . Moreover, Nara et al. (1972) and Okachi et al. (1973a) recently described a novel type of penicillin acylase, namely, ampicillin acylase. The division of the penicillin acylases in two classes, based on the type of microorganism involved, as proposed by Claridge et al. (1963) seems no longer to be vaIid in view of this increasing number of exceptions. Here, the penicillin acylases are classified according to the type of penicillin which is preferentially hydrolyzed. So far, three types of penicillin acylase can be clearly recognized: penicillin V acylase, penicillin G acylase, and ampicillin acylase. However, other “unusual” acylases have been described, but as none of them has been purified to elucidate their substrate spectrum, it is not clear whether they interfere with the above-mentioned classification or whether they should be added as novel types (Cole, 1966, 1969a; Pruess and Johnson, 1965; Vandamme et al:, 1971b; Vanderhaeghe et al., 1968). In this way, a survey of strains that produce penicillin acylase is given in Tables 11-IV. As practically each strain that produces penicillin acylase displays its own properties with respect to growth pattern and enzyme production, a condensed view would omit important features. In this connection, detailed description of these fascinating microbes and their peculiar enzymes is fully justified.
B. PENICILLIN V ACYLASES 1. Hydrolytic Action a. Molds Soon after the discovery and the purification of 6-APA from fermentation broths, Rolinson et al. (1960) reported that 6-APA was also produced by enzymatic activity of molds on penicillin V. Cole and Rolinson ( 1961) demonstrated penicillin amidase activity in the mycelium of Cephalosporium sp. and Emericellopsis minima, after they had found 6-APA in the fermentation broth of these molds during penicillin N production. To the washed mycelium suspended in a 0.05 M phosphate buffer of pH 8, 1%penicillin V, 1%penicillin G, and 1% penicillin N was added. After 18 hours of incubation at 2OoC, the mix-
327
MICROBIAL PENICILLIN ACYLASES
TABLE I1 MICROORGANISMS THATPRODUCE PENICILLIN G ACYLASE Microorganisms Bacteria Rhodopseudomonas spheroides KY 4112, Pseudomonas aeruginosa KY 3591, KY 8501, P. cruciviae KY 3960 P . desmolytica KY 3981 Pseudomonas sp. Xanthomonas sp. Alcaligenes jaecalis A-9424 A. jaecalis BRL 1237, 1238 Bacterium jaecalis alcaligenes 415 Flavobaeterium sp. Azotobacter chroococcum Beij .C12Pr. Escherichia sp. E. coli ATCC 9637 (NCIB 8666) E . coli N, 1/3-67 E . coli NCIB 9465 E . coli BMN, KY 8219, KY 8268, KY 8275, KY 8289 E. coli I 187 E . coli N, I / 3 E. coli BRL 351, BRL 1360 E. coli NCIB 8743, 8743A E. coli ATCC 11105 (NCIB 8878) E. coli XG3A 9455 E . coli 0111 B, E. coli NCIB 8134, 8879, 8949 E . coli NCIB 8741, 8742, 8744 Aerobacter cloacae Erwinia sp. Serratia sp. Proteus morganii KY 4035, KY 4051 Proteus rettgeri FD 13424 P. rettgeri ATCC 9919, 9250 Bordetella sp. Micrococcus lysodeikticus Micrococcus sp. M . roseus ATCC 516 M . ureae KY 3967, M. luteus KY 3781 Sarcina sp., Corynebacterium sp., Cellulomonas sp. Kluyvera citrophila KY 3641, PL-10, PL-21, Kl. noncitrophila KY 3642, KY 8991 Bacillus subtilis var. niger B. megaferium ATCC 14945 Actinomycetales Mycobacterium phlei Nocardia F D 46973, ATCC 13635 Streptomyces ambojaciens SPSL-15 Fungi Neuraspora crassa FSD 987, DGC 757, R F 424, RWB 622, FGSC 262, 3a6A
References
Nara et al. (1971a) Okachi et al. (1973b) Huang el al. (1960) Huang el al. (1963) Claridge et al. (1963) Cole and Sutherland (1966) Gotovtseva et al. (1965) Huang el al. (1963) h p i d i et al. (1970) Rolinson et al. (1960) Kaufmann and Bauer (1960) Ssentirmai (1984) Holt and Stewart (1964a) Okachi et al. (1973b) Cole and Sutherland (1966) Nyiri (1967) Sjoberg el al. (1967) Cole (1969a) Bauer et al. (1971) Claridge et al. (1963) Dulong de Rosnay et al. (1970) Cole (1967) Cole and Sutherland (1966) Claridge et al. (1960) Huang el al. (1963) Okachi el al. (197313) Huang el al. (1963) Cole (1967) Huang el al. (1960) Claridge et al. (1960) Huang el al. (1960) Pruess and Johnson (1965) Nara el al. (1971a) Huang et al. (1963) Nara et al. (1971a); Okachi et al. (1973b) Claridge et al. (1960) Chiang and Bennett (1967) Claridge et aZ. (1960) Huang el al. (1960) Nara el al. (1971a) Rossi et al. (1973)
328
E. J. VANDAMME AND J. P. VOETS
TABLE I11 MICROORGANISMS THATPRODUCE PENICILLIN V ACYLASE Microorganisms Fungi a. Mycelium Penicillium chrysogenum Q176 P. chrysogenum A-9342 P . chrysogenum W5120 P. chrysogenum Wis. 49-408, P . chrysogenum P-5009 P . chrysogenum SC 3576 P . chrysogenum 51-20F3 Penicillium BRL 807 Emericellopsis m i n i m a (Stolk) I M I 69015, Cephalosporium salmosynnematum Cephalosporium CMI 49137 C. acremonium ATCC 11550 Aspergillus niger sp. A. ochraceus BRL 731 Epidermophyton interdigitale, E. Jloccosum, Trichophyton gypseum, 2'. mentagrophytes, T . interdigitale Alternaria, Epicoccum, Mucor, Phuma, Trichoderma Pleurotus ostrealus Botrytis cinerea Fusarium sp. 75-5 F . avenaceum F . semitectum
F . semitectum BC 805 Gibberella fujikuroi b. Conidia Fusariunt conglutinans AYF 254 F. moniliforme AYF 255, CBS 24064 CBS 44064, CBS 26654 Yeasts BRL 809, Torulopsis, Zygosaccharomyces, Debaryomyces, Torula Cryptococcus, Saccharomyces, Trichosporon Rhodotorula glutinis var. glutinis Bacteria Erwinia aroideae NRRL B-138 Achromobacter BRL 17.55 (NCIB 9424) Micrococcus ureae KY 3767 Actinomy cetales Nocardia globerula KY 3901 Streptomyces lavendulae BRL 198 S . netropsis 2814, S . erythreus JA 4143 S . ambofaciens SPSL-15 Actinoplanes utahensis
References
Sakaguchi and Murao (1950) Claridge et al. (1963) Batchelor et al. (1959) Erickson and Bennett (1965) Erickson and Dean (1966) Spencer and Maung (1970) Cole (1966) Cole and Rolinson (1961) Claridge et al. (1963) Dennen et al. (1971) Vandamme et al. (1971a) Cole (1966) Uri et al. (1963) Batchelor et al. (1961d) Brand1 (1965) Batchelor et al. (1961a) Thadhani et al. (1972) Vanderhaeghe et al. (1968) Waldschmidt-Leitz and Bretzel (1964) Baumann et al. (1971) Vasilescu et al. (1969) Singh et al. (1969) Vandamme et al. (1971a) Cole (1966, 1967) Batchelor et al. (1961d) Vandamme and Voets (1973) Voets and Vandamme (1972) Cole (1964) Nara et al. (1971a) Nara et al. (1971a) Batchelor el al. (1961d) Haupt and Thrum (1967) Nara et al. (1971a) Dennen et al. (1971)
MICROBIAL PENICILLIN ACYLASES
329
TABLE IV A. MICROORQANISYS THATPRODUCE AMPICILLIN ACYLASE Microorganisms
References
Pseudommas melanogenum KY 3987, K Y 4030, KY 4031 Pseudommas ovalis K Y 3962
Nara et al. (1972), Okachi et al. (1973b) Okachi et al. (1973b)
B. MICRVORQANISMS REPORTED TO PRODUCE UNUSUAL PENICILLIN ACYLASES Microorganisms Yeast BRL 809 Penicillium chrysogenum W49-408 Erwinia aroideae NRRL-B138 E . coli NCIB 8743A
Substrate
References
Methylpenicillin
Cole (1966)
n-Undecylpenicillin Cloxacillin, methicillin p-Hydrvxy benzylpenicillin
Vanderhaeghe et al. (1968) Vandamme et al. (1971b) Cole (1969a)
tures were paper chromatographically checked for the presence of 6-APA. Only penicillin V was hydrolyzed into 6-APA and phenoxyacetic acid. Batchelor et al. (1961a)d) showed that the addition of benzylpenicillin to a submerged Botrytis cinerea culture results in the formation of traces of 6-APA. More 6-APA was produced by adding a mixture of biosynthetic penicillins: heptyl, 2-heptenyl, pentyl, benzyl, p-hydroxybenzyl, and phenoxymethylpenicillin. Batchelor et al. used this mixture of penicillins as a model-substrate and concluded that the penicillin acylase present in molds preferentially hydrolyzes penicillin V. They screened 215 different molds and found 17%to be acylase producers, including Alternaria, Aspergillus, Epicoccum, Mucor, Penicillium, Phoma, Fusarium, and Trichoderma strains. Claridge et al. (1963) described the substrate specificity of the penicillin acylases present in Penicillium chrysogenum A-9342 and Cephalosporium CMI 49137. At pH 7 at 37OC, the Cephulosporium acylase hydrolyzes only penicillin V into 6-APA, whereas the Penicillium acylase preferentially hydrolyzes penicillin V, but also to a lesser degree carboxypenicillins, ampicillin, and penicillin G. In 1957 Uri et al. observed that Trichophyton mentagrophytes and Epidermophyton interdigitale strains were able to produce penicillin, and as a result they examined 70 dermatophytes for production of 6-APA. Endocellular, as well as exocellular, penicillin acylase was produced, penicillin V being the best substrate. However, the mycelium possessed a higher activity than the fermentation broth. The acylase activity could
330
E. J. VANDAMME AND J. P. VOETS
be stimulated by the addition of 0.2%phenoxyacetic acid to the growth medium ( Uri et al., 1963,1964). Waldschmidt-Leitz and Bretzel ( 1964) detected in Fusarium semitectum an intracellular penicillin amidase with a pH optimum 7.5 and penicillin V as substrate. They liberated the enzyme by extracting acetone-dried mycelium with 0.2 N sodium acetate solution at pH 6. After treatment with DEAE-cellulose at pH 8, separation on Sephadex G-25, and lyophilization, the extract was redissolved in 0.005 N sodium citrate. After precipitation with acetone (44%)and -chromatography on Amberlite IRC-50 and Sephadex G-25, a 300-fold purified enzyme extract was obtained. Ultracentrifugation experiments indicate a molecular weight of about 65,000. The nitrogen content was 16.7%, and two atoms of zinc per molecule were found, indicating the role of this metal as a cofactor. Brandl (1965, 1972) studied the penicillin amidase activity of Fusarium semitectum and Pleurotus ostreatus. As substrate, penicillins G and V and phenylmercaptotnethylpenicillin were tested. At pH 8 and an optimum temperature of 5OoC, penicillin V was transformed preferentially into 6-APA. As K , value for this Fusarium penicillin acylase, a value of 2.50-2.82 m M was found. Recently, Brandl tested 52 different biosynthetic and semisynthetic penicillins as possible substrates for the purified penicillin acylase of Fusarium semitectum ( Brandl, 1972). He concluded that bulky substitutions and the presence of double bonds in the side chain of the penicillin molecuIe cause steric hindrance that results in a lowered binding capacity to the active site of the enzyme. Erickson and Bennett (1961, 1965) examined the acylase activity of two Penicillium chrysogenum strains-a nonpenicillin-producing strain and a high-producing one: Wis. 45-408 and P-5009. As enzyme source, they used washed mycelium grown in penicillin fermentation broths. The acylase activity was located intracellularly and was optimal at pH 8.5 at 30°C. Penicillins V, K, F, and G were transformed into 6-APA, the relative rates being approximately 6, 3.5, 2.5, 1. Acetone-dried mycelium also displays activity. The acylase activity of the high penicillin-producing and the nonproducing strains were found to be very similar; this indicates that penicillin acylase production does not always result in penicillin production, although so far all the penicillin-producing fungi possess acylase activity. Shimi and Imam (1966) reported on the breakdown of penicillin G by preformed mycelium mats of Penicillium chrysogenum Q176 and Aspergillus fluuus. Degradation of the penicillin was initiated by cleavage of the side chain structure. The 6-APA produced was further degraded into cysteine, acetone, and glycine. Cole ( 1966) described penicillin acylase activity of different molds:
MICROBIAL PENICILLIN ACYLASES
331
Trichophyton mentagrophytes, Aspergillus ochraceus, Cephalosporium species, and three Pencillium species. Washed mycelium displayed high acylase activity with penicillin V as substrate. Aspergillus ochraceus acylase activity was considerably increased by culturing the mold in the presence of phenoxyacetic acid, confirming the stimulating effect earlier described by Uri et al. (1964). Cole concluded that at least penicillin V acylase activity is present in all penicillin-producing fungi. Vasilescu et al. ( 1969) used the fungus Gibberella fujikuroi to hydrolyze penicillin V into 6-APA and phenoxyacetic acid. Washed mycelium could be recovered from the reaction mixture and used for 7 or 8 further hy drolyses. Vanderhaeghe et al. (1968) described penicillin acylase in crude mycelial extracts of Penicillium chrysogenum Wis. 49-408, a strain that produces no detectable amounts of penicillin or 6-APA, and in extracts of Fusarium avenaceum and Fusarium orysporum. Acylase activity was optimal at pH 7.5 at 37°C with penicillin V as substrate. Aliphatic penicillins were hydrolyzed into 6-APA in relation to the number of carbon atoms in the side chain (the more C, the more 6-APA). Penicillin N and isopenicillin N were unaffected by the enzyme, but their esters were slightly hydrolyzed. Brunner and Rohr (1965) and Brunner et al. (1966a,b, 1968) reported on the enzymatic hydrolysis of penicillin V and a wide range of N-acy amino acids by FusaTium semitectum. They reported on amino acylases present in Penicillium chrysogenum 4176 mycelium which display no penicillin acylase activity at all. However, after growth of the mold in the presence of N-phenoxyacetyl amino acids as sole source of carbon, low-penicillin V acylase activity could be detected. Recently Baumann et al. (1971) reported on the substrate specificity of a penicillin amidase from Fusarium semitectum BC 805. When cultivated in media that contain phenoxyacetic acid as an inducer, the intact mycelium is able both to hydrolyze penicillin V into 6-APA and to deacylate a number of N-phenoxyacetyl amino acids. Baumann et al. showed that this broad substrate-specificity pattern is due to the action of two different enzymes. By extraction of the mycelium with distilled water and 0.2 M NaCI, and by precipitation with acetone, ultracentrifugation and chromatography on calcium phosphate gel, the two enzymes were separated. The penicillin V acylase was purified 20.5-fold with a recovery of 43%.It displayed most pronounced specificity toward the hydrolysis of penicillin V, but lacked the ability to split either N-phenoxyacetyl amino acids or benzylpenicillin. The K , for the hydrolysis of phenoxymethylpenicillin was 4.75 mM. The other enzyme, called phenoxyacetylase was purified 15-fold and preferentially deacylated N -
332
E. J. VANDAMME AND J. P. VOETS
phenoxyacetyl-L-alanine and other N-acetylamino acids at pH 7-8 at 40°C but displays no hydrolytic action on penicillins. Recently, Thadhani et al. (1972) reported also on the specificity of a purified fungal penicillin acylase. The enzyme specifically hydrolyzing phenoxymethylpenicillin to 6-aminopenicillanate and phenoxyacetate was purified 200-fold from the culture filtrate of Fusarium sp. and was homogeneous on polyacrylamide disc-gel electrophoresis. The enzyme was formed adaptively when the organism was grown in the presence of phenoxyacetate or its derivatives. The amount of enzyme induced depended on the nature of the substitution of phenoxyacetate and on the amino acid moiety of the phenoxyacetyl L-amino acid. The acylase induction was highest with phenoxyacetyl-L-alanine and decreased with higher homologs of aliphatic amino acid. The presence of phenyl, heterocylic or hydroxyl groups on the side chain or the absence of the a-carboxyl group in the amino acid moiety did not decrease the inducer effect of the compound. Phenoxyacetyl-D-ahnine failed to induce the enzyme formation by the organism, indicating that the enzyme formation exhibited marked stereospecificity toward the amino acid moiety of phenoxyacetyl amino acids. Although the phenoxyacetyl amino acids served as good inducers of the enzyme, none of them was acted upon as a substrate by the enzyme. The penicillin acylase preparation, purified from Penicillium chrysogenum 51-2OF3, also displayed a high specificity toward the penicillin molecule ( Spencer and Maung, 1970). These reports indicate at least that these penicillin V acylases are highly specific and are not aspecific deacylating enzymes, as is proposed for other penicillin acylases by many authors experimenting with intact cells or crude preparations as enzyme source.
b. Fungal Spores Hitherto most of the fungal acylase activity has been detected during fermentation processes or in experiments with mycelial preparations. In 1969, Singh et al. reported on the penicillin acylase activity of Fusarium moniliforme AYF 255 and Fusarium conglutinans AYF 254 conidia. Indeed, fungal spores, generally regarded as being metabolically inert, have now been shown to exhibit a wide range of enzymatic activities ( Caltrider and Gottlieb, 1963; Van Etten et al., 1969). Apparently, fungal spores are capable of a wide range of substrate conversions, some of which could prove to be valuable in the fermentation industry (CasasCampillo and Bautista, 1965; Gehrig and Knight, 1958; Hafez-Zedan and Plourde, 1971; Johnson et al., 1968; Lawrence, 1966, 1967; Singh
MICROBIAL PENICILLIN ACYLASES
333
et al., 1968; Vezina et al., 1965, 1968). In addition, transformation of organic compounds by spores offers certain advantages not provided by growing cell populations: large numbers of spores are obtained in an easy way by growth on “waste” materials, spores can be stored for a long period without loss of activity, and the simplicity of transformation by the “spore process” avoids contamination. The “spore process” and its applications were reviewed in detail by Vezina et al. (1965, 1968). Penicillin V acylase activity of fungal spores was studied by Vandamme et al. (1971a). The Fusurium moniliforme spore acylase was characterized by its specific localization in the conidia and was not detectable within the mycelium. Aspergillus, Penicillium, and Botrytis sp. with acylase-active mycelium produced spores without this activity. The spore acylase is highly specific toward penicillin V at pH 8, and its activity is stimulated by the addition of phenoxyacetic acid. Other penicillins or derivatives are not affected, but several aliphatic and aromatic amides are also-although rather slowly-hydrolyzed. Resynthesis into penicillin V was not detected. Spores obtained by surface sporulation methods and from submerged cultures were equal in their acylase activity, although the submerged culture technique allowed the stimulation of acylase biosynthesis by addition of phenoxyacetic acid. With intact spores as enzyme source, a K , value of 5.75 mM was found. c. Yeasts
Batchelor and associates (1961d) reported on the penicillin V acylase activity of Cryptococcus, Saccharomyces, and Trichosporon strains. Penicillin acylase activity was also detected in Torulopsis sp., Torula sp., Zygosaccharomyces, and Debaryomyces sp. (Cole, 1967). Cole also mentioned on a yeast BRL 809, displaying methylpenicillin acylase activity (Cole, 1966). Attempting selective screening procedures for microorganisms that produce penicillin acylase, Vandamme ( 1973) and Vandamme and Voets (1973) isolated a Rhodotorub glutinis var. glutinis from a soil sample, using a mineral medium containing N-acetylL-glycine as a sole source of carbon and nitrogen. The intracellular enzyme hydrolyzed specifically penicillin V at pH 6.3 at 28°C. A crude enzyme preparation was entrapped in polyacrylamide gel and tested as to its capacity to act as a continuous insoluble enzyme reactor. The transformation capacity was rather low, however.
d. Actinomycetales Already in 1961 Batchelor et al. ( 1961d), detected penicillin acylase activity in the culture filtrate of Streptomyces luvendulae BRL 198. Heptyl, 2-pentenyl, and phenoxymethylpenicillin were rapidly transformed into 6-APA. The optimum pH was found to be approximately 9.0 with maxi-
334
E. J. VANDAMME AND J. P. VOETS
mum enzyme stability at pH 10. They demonstrated that the acylase activity of Streptomyces lauendulae BRL 198 was highly influenced by the composition of the growth medium. The exocellular acylase was partially purified from the culture broth by precipitation with cold acetone and (NH,),SO,. Using this crude extract as enzyme source, a Z L of 10.3 mM was found. Addition of phenoxyacetic acid to the reaction mixture has no influence on the activity, while 6-APA totally inhibits the acylase. This phenomenon was also mentioned for the penicillin acylase of Actinoplanes utahnsis (Dennen et al., 1971). In 1967, Haupt and Thrum detect penicillin acylase in Streptomyces sp., possessing an aspecific amide synthetase activity. Acetone-dried mycelium, and also the concentrated culture fluid of Streptomyces erythreus JA 4143 and Streptomyces netropsis 2814, displayed acylase activity. Only penicillin V was hydrolyzed into 6-APA at pH 7.5 at 28OC. They concluded that this acylase is also able to hydrolyze other amide substrates and the enzyme is involved in the detoxification process of aromatic acids by Streptomyces species. Recently, Nara et al. (1971a) described a Streptomyces ambofaciem SPSL-15, which displayed both intra- and exocellular penicillin acylase activity. The exocellular enzyme hydrolyzed both penicillin V and G, and the intracellular one preferentially split penicillin G. Nocardia globerula KY 3901 and three unidentified actinomycetales were isolated after growth on n-paraffins and found to split preferentially penicillin V into 6-APA.
e. Bacteria Cole ( 1964) reported on an Achromobacter BRL 1755 (NCIB 9424) strain that preferentially hydrolyzes penicillin V and phenoxyacetylglycine into 6-APA. However, more information about this acylase is not available. Recently Nara et al. (1971a) described a Micrococcus u r e a KY 3767, specifically hydrolyzing penicillin V into 6-APA at pH 7.4 at 35OC and Vandamme et al. (1971b) detected an Erwinia aroideae penicillin V-acylase. This microorganism was previously extensively studied for its L-asparaginase activity (Beyaert and Voets, 1970; Peterson and Ciegler, 1969a,b; Liu and Zajic, 1972). Cell suspensions of the bacteria rapidly hydrolyze penicillin V and phenethicillin into 6-APA at pH 5.6, but penicillin G hardly at all. However, on further incubation, after a hydrolysis period, part of the produced 6-APA is again transformed into penicillin by the action of another, probably induced, enzyme, not by the reversed action of the penicillin V acylase. At pH 8, the cells hydrolyze different amides and also, although more slowly, the semisynthetic penicillins cloxacillin and methicillin. At this pH value, the activity on peni-
MICROBIAL PENICILLIN ACYLASES
335
cillin V and G is very low. The enZymatic activity is strictly associated with the cells. The Erwinia aroideae acylase seems to be a constitutive enzyme as it is not influenced in its biosynthesis or activity by the composition of the medium. The enzyme was extracted and purified 98-fold by gel filtration. The purified penicillin acylase is highly specific for penicillin V, although slow hydrolysis of tripeptides containing the glycylglycine moiety was also observed. Phenoxyacetic acid had an inhibitory effect on the acylase activity. A K , value of 35 mM was calculated. The molecular weight as estimated by thin-layer gel chromatography was 62,000. The purified acylase displays neither synthetic, nor amido hydrolase, nor transacylase activity ( Vandamme, 1972; Voets and Vandamme, 1972). 2. Enzymatic Synthesis of Penicillins by Penicillin V Acylases In 1966, Erickson and Dean demonstrated that mycelial preparations of Penicillium chrysogenum Wis. 49-408 and SC 3576 were able-in addition to hydrolysis into 6-APA-to synthesize penicillin by acylation of 6-APA. By adding 6-APA and phenoxy- or phenylacetic acid to the mycelium extract, after 48 hours of incubation at 25°C at pH 6.8, penicillin V or G was formed. They proposed this enzyme or enzyme system has a role in the mechanism of penicillin biosynthesis. However, whether there are two enzymes involved, one that deacylates penicillin V into 6-APA (penicillin acylase) and a second one that acylates 6-APA into penicillins, was not confirmed owing to a lack of highly purified extracts. Also Cole (1966) reported on the acylation of 6-APA into penicillins by Penicillium BRL 807 and a yeast strain BRL 809. Brunner et al. (1968) examined the reversibility of the acylase action on penicillin V. They used cell-free extracts of Penicillium chrysogenum and described the activation of the acyl group by phenoxyacetic acidcoenzyme A-ligase and the subsequent binding of this activated R group to 6-APA by a transacylase. According to these authors, it is evident that resynthesis into penicillin V is effected by an enzyme quite different from penicillin acylase. Also Spencer (1968), Spencer and Maung ( 1970), and Spencer and Burnham ( 1973) demonstrated in Penicillium chrysogenum 51-20F3 extracts enzymatic activity able to bind phenylacetyl coenzyme A to 6-APA, resulting in the formation of penicillin G. From these results, owing to a lack of experiments with highly purified penicillin V acylases, the reversible character is still to be proved and deserves further attention. However, Spencer and Maung (1970) recently claimed the identity of these enzymes in Penicillium chrysogenum 51-20F3, indicating that
336
E. J. VANDAMME AND J . P. VOETS
penicillin acylase is an enzyme with multiple, but related activities, each of them expressed in well defined conditions. A survey of the properties of different penicillin V acylases is presented in Table V. C. PENICILLIN G ACYLASES
1. Hydrolytic Action a. Bacteria In 1960 Rolinson and co-workers obtained 6-APA by hydrolysis of penicillin G by bacterial enzymes. Cell suspensions of Escherichia coli and Alcaligenes species transformed penicillin G into 6-APA at pH 8 and at 4OOC. The rate of penicillin V transformation was about five times slower. Claridge et al. (1960, 1963) described penicillin amidase activity in Aerobacter cloacae, Bacillus subtilis var. niger, Micrococcus lysodeikticus, and Alcaligenes faecalis. By ultrasonic destruction, they prepared a crude extract of Alcaligenes faecalis cells and demonstrated hydrolysis of penicillin G at pH 7.5 at 37OC. Penicillin K-although slower-was also hydrolyzed into 6-APA. They were able to reuse cell suspensions four times without appreciable loss in activity, indicating the stability of the enzyme in “resting cells.” They tested the substrate specificity of the penicillin acylases found in Alcaligenes faecalis A-9424 and in E. coli XG-3A-9455 cells. Both acylases hydrolyzed preferentially penicillin G into 6-APA, but also penicillin V, ampicillin, and phenethicillin. Huang et al. (1960) detected penicillin acylase activity among Bordetella, Alcaligenes, Micrococcus, and Pseudomonas species. Kaufmann and Bauer (1960) showed that a washed cell suspension of E . coli ATCC 9637, suspended in a phosphate buffer of pH 7.5, saturated with toluene was able to transform penicillin G into 6-APA. They demonstrated that the penicillin G acylase activity of this strain was stimulated by culturing the bacteria at 24°C in a medium, rich in yeast extract-, supplied with phenylacetic acid. Also other substrates containing acyl groups were hydrolyzed, and the activity was not dependent only on the presence of the 6-APA nucleus. They concluded that the enzyme involved is an aspecific enzyme localized on the cell membrane ( Kaufmann, 1964; Kaufmann and Bauer, 1964a,b). In 1964, Sikyta and Slezak described the continuous culture of E . coli ATCC 9637. They demonstrated that addition of 0.0154: ammonium phenylacetate stimulated the acylase activity. According to Lucente et al. (1965), the E . coli ATCC 9637 acylase is an aspecific amidase, hydro-
MICROBIAL PENICILLIN ACYLASES
337
lyzing phenylacetyl L-amino acids and their derivatives, in this way confirming the experiments of Cole (1964) and Kaufmann and Bauer ( 1964b). Nevertheless, all these experiments were carried out with intact cells or crude extracts, so eventually different enzymes might be found to be involved in the hydrolysis process. Pruess and Johnson (1965) prepared acetone-dried E. coli ATCC 9637 cells and demonstrated that this preparation has an optimal acylase activity at pH 9 at 35°C with penicillin G as substrate. Benzylpenicilloic acid and benzylpenilloic acid were slowly hydrolyzed. Levitov et al. (1967) examined the stimulating effect of phenylacetic acid and its derivatives on the penicillin acylase biosynthesis by E. coli ATCC 9637. A concentration of 1 mg/ml of phenylacetic acid results in an 8-fold stimulation of the acylase activity. Derivatives of phenylacetic acid and analogous compounds were inferior in this respect, Devcic and Divjak (1968) demonstrated that the concentration of lactic acid and of peptone in the growth medium have a markable influence tm the penicillin acylase biosynthesis of E. coli ATCC 9637. Kleiner and Lopatnev (1972) showed that the aeration was an important factor for penicillin acylase biosynthesis by this strain. Badr-Eldin and Attia (1973) described the deacylation of penicillin G by E. coli at pH 7.5. A K, value of 17.5 mM was reported, and, in addition to benzylpenicillin, methyl benzyl penicillinate was transformed into methyl-6-APA. Bondareva et al. (1969a,b) isolated the penicillin G acylase of E . coli ATCC 9637 cells by ultrasonic disruption and purified the extract about 100 times. The preparation was electrophoretically homogeneous and catalyzed not only the hydrolysis of penicillin G and other phenylacetyl substrates, but also their synthesis. Maximal hydrolytic activity was found in the pH range 7.8-8 and the optimal temperature was 50-52OC. It was found that, in reactions of hydrolysis and resynthesis, the purified enzyme is specific in regard to phenylacetic acid. The maximum hydrolytic reaction rate was found when phenylacetylated compounds were used as substrate. All alterations in the structure of the acyl part of the substrate molecule (benzylpenicillin) resulted in decreased acylase activity. Besides penicillin G and ampicillin, the purified enzyme also hydrolyzed a number of phenyl and phenoxyacetylated amino acids and therefore seems not to be an enzyme of narrow specificity. The rate of synthetic reactions-acylation of 6-APA with carboxylic acids-increased significantly when derivatives of these acids were used with amino acids and thioglycolic acid. Self et al. (1969) extracted the penicillin amidase from E. coli ATCC 9637 grown on phenylacetic acid and glutamate, and they purified the
TABLE V RELEVANT PROPERTIES OF PENICILLIN V ACYLASES Optimum p H
Microorganisms Molds Penicillium chrysogenum Cephalosporium sp. Emericellopsis minima (Stolk) I M I 69015 Trichophyton mentagrophytes, Epidermophyton interdigitale Fusarium semitectum Intact cells Enzyme Pleurotus ostreatus
Hydrol- Synysis thesis
Optimum temp. ('C)
Localization
K,
Molecular weight
References m
8 8
20 -
-
-
Intracellular (mycelium)
Cole and Rolinson (1961), Cole (1966)
P
-
8 8
Intra- and ex tracellular (mycelium)
7.5
-
8
50
8
50
Penicillium chrysogenum Wis. 49408, 8 . 5 P . chrysogenum P5009 Fusarium avenaceum 7.5
30
F . conglutinans AYF' 254
8
28
F . moniliforme AYF 255
8
28
37
In tracellular (mycellium)
-
4
3
4
Brandl (1965)
2.50-2.82 mM 65,000
Intracellular (mycellium) Intracellular (mycelium) Intracellular (mycelium) Intracellular (spores)
Uri et al. (1963)
-
Waldschmidt-Leitz and Bretzel (1964) Brandl (1965)
-
Erickson and Bennett (1965) Vanderhaeghe et al. (1968)
-
-
Singh et al. (1969)
5.75mM
-
Vandamme et al. (1971a)
2
Penicillium chrysogenum 51-20F3 (enzyme) Fusarium semitectum BC 805 (enzyme) Yeasts Rhodotorula glutinis
8
7.4
28
Intracellular (mycelium)
7.5
-
37
Intr acellular
6.5
-
28
Intracellular
x
-
Spencer and Maung (1970)
4.75mM
-
Baumann et al. (1971)
5 . 1 mM
-
Vandamme and Voets
1.67
1 0 - 4 ~
(1973)
Actinomycetes Streptomyces lavendulae BRL 198 S. erythreus JA 4143, S. netropsis
9 7.5
-
28
S. ambofaciens SPSL-15 Nocardia globerula KY 3901 Bacteria Achromobacter BRL 1755 Erwinia aroideae sp.
7.4 7.4
-
35 35
Ex tracellular Intra- and ex tracellular Ex tracellular Intracellular
5.5
-
28
Intracellular Intracellular
35mM
Micrococcus ureae KY 3767
7.4
-
35
Intracellular
-
2814
28
10.3 mM -
-
-
-
Batchelor et al. (1961d) Haupt and Thrum (1967) Nara et al. (1971a)
Cole (1964) 62,000 Vandamme (1972) ; Voets and Vandamme (1972) Nara et al. (1971a)
340
E. J . VANDAMME AND J. P. VOETS
enzyme by fractionation with streptomycin sulfate, ammonium sulfate, and polyethyleneglycol, followed by chromatography on DEAE-cellulose. The purification factor was about 100-200, and the overall yield was about 35%.Self et al. found a K , value of 7.7 mM. The enzyme was chemically attached to derivatives of cellulose and the kinetics of these insolubilized penicillin amidase preparations in a penicillin amidase reactor were investigated. An optimum pH of 7.5 was found at a temperature of 37OC. A DE52-cellulose reactor and a cellulose-sheet reactor were active during 11 weeks of operation and no loss of activity was observed. The free enzyme, under the same conditions, lost 4.2% of its activity in one day. The ability to carry out the reverse reaction (resynthesis) was checked by adjusting effluent from the preceding experiment to pH 5 and passing it through the reactor again. The presence of benzylpenicillin in the effluent confirmed that the reverse reaction was occurring. The possibility was reconsidered of producing 6-APA together with penicillin by fermentation (Fuska et al., 1972). In this process, the penicillin, used as the crude butylacetate extract from the culture filtrate, was mixed with a suspension of E . coli ATCC 9637 cells at pH 7.2-7.4. The purity and yields of the recovered 6-APA were better compared with enzymatic hydrolysis of a pure penicillin solution. Recently, a purified preparation of the penicillin acylase from E . coli ATCC 9637 cells was entrapped in cellulose triacetate fibers, which were packed in a column in order to obtain an insoluble enzyme reactor (Marconi et al., 1973). The acylase fibers were found to be quite stable. No loss of activity by casual contamination was found, because the entrapped enzyme was protected against microbial proteolytic attack. A remarkable reduction of the hydrolysis time was observed, and a conversion yield higher than 90% was reached. It was found that the acylase fibers were a better catalyst at low temperature than the free enzyme. Even whole microbial cells, trapped in wet spun synthetic fibers, displayed a high and lasting activity. This novel system was able to produce 6-APA continuously over a period of 4 months (Dinelli, 1972). The penicillin acylase activity of the strain E . coli ATCC 9637 is so far the most intensively studied. From numerous experiments it is clear that this bacterial strain contains an enzyme which is specific for phenylacetylated compounds and is not a specific penicillin G acylase. However, this statement is a p~iorinot valid for other penicillin acylases; further investigations, including purification procedures, are needed to elucidate the real nature of these enzymes. Kameda et al. (1961) isolated from soil, bacteria that could hydrolyze acyl derivatives of m-amino acids. Some strains were found to produce penicillin G acylase and could grow on a mineral medium, with penicillin
MICROBIAL PENICILLIN ACYLASES
341
G as a sole source of carbon. Kameda et al. postulated that this is a convenient way to isolate penicillin acylase positive bacteria. Huang et al. (1963) proposed the name penicillin acylase, as an N-acyl group was split off from an N-acyl 6-APA-derivative,a reaction analogous to those of the kidney acylases, which specifically hydrolyze N-acyl L-amino acids. Special attention was drawn to Proteus rettgeri FD 13424. As enzyme source they used lyophilized cells suspended in a toluenesaturated phosphate buffer solution. The intracellular enzyme hydrolyzed preferentially penicillin G at pH 8. Penicillin side chain variations strongly influenced the substrate specificity; indeed substitution in the phenyl ring of penicillin G (p-nitro, p-hydroxy) or on the methylene C atom reduced drastically the acylase activity. Insertion of 0 or S between the phenyl ring and the C atom also diminished the acylase activity. Penicillin V was partially, and penicillin N not at all, hydrolyzed. However, modifications of the penicillin nucleus, such as penicillinamide or methyl ester, scarcely influenced the activity. Szentirmai (1964) detected an E. coli NyI/3-67 with a low acylase activity. By the addition of aromatic carbonic acid and phenylacetic acid and its derivatives, and phenoxyacetic acid, which themselves strongly inhibit the function of this specific enzyme, the biosynthesis of the penicillin G acylase has been increased 10-fold. The production of this enzyme was effectively repressed with metabolic carbohydrates and polyalcohols (glucose, fructose, glycerol). Penicillin G was hydrolyzed 8 times faster than penicillin V. The specific activity of the bacterial cells depended upon temperature, low temperatures being favorable for acylase production. Holt and Stewart ( 1964a) detected in the concentrate of the supernatant liquid of an E. coli C15 (NCIB 9465) culture a penicillin G acylase with a pH optimum of 5.5, although normally optimum pH for hydrolysis is 8 and at pH 5.5 usually resynthesis of penicillins occurs. Cole (1964) described an E. coli BRL 1040 that possesses an acylase that is not specific for penicillin, but is for phenylacetyl groups; he demonstrated that a series of phenylacetyl-L-a-amino acids were hydrolyzed faster than penicillin G. He proposed the name phenylacetylase and compared it with other nonspecific acylases. Pruess and Johnson (1965) studied an intracellular penicillin G amidase found in acetone-dried cells of Micrococcus roseus ATCC 516. EDTA (1 mM) totally inhibited the activity of the cells, and calcium ions ( 2 mM) relieved the EDTA inhibition, suggesting that a metal is necessary for full enzymatic activity. Neither thioglycolate (10 mM) nor p-chloromercuribenzoate affected the rate of deacylation, indicating that no free thiol group is necessary for activity. Brand1 (1965) described an E. coli species extract displaying penicillin
342
E. J. VANDAMME AND J. P. VOETS
G acylase, with an optimum pH 7 at 3035°C and a K , value of 1.50 mM. Gotovtseva et al. (1965) detected penicillin acylase activity in Bacterium faecalis alcaligenes No. 415 at pH 7.6 and 42OC, cultured in a medium supplied with corn extract. They showed that the acylase activity is dependent on the composition of the growth medium and that abundant growth is accompanied by a low enzymatic activity of the culture. It was found that acetic and pyruvic acids supplied to the nutrient medium greatly reduced the production of penicillin acylase, while peptone and casein hydrolyzate have a stimulatory effect. More particularly, they found that tryptophan, other indole derivatives, and anthranilic acid stimulated acylase biosynthesis ( Gotovtseva and Levitov, 1965; Gotovtseva et al. 1968). Chiang and Bennett ( 1967) purified the extracellular penicillin amidase of Bacillus megaterium ATCC 14945 approximately 96-fold by means of two cycles of adsorption on, and elution from, Celite followed by further fractionation on carboxymethyl ceIlulose. From ultracentrifugation results, they concluded that they had obtained a homogeneous preparation with an apparent molecular weight of approximately 120,000. The enzyme is specific for benzylpenicillin and has a pH optimum between 8 and 9. Other penicillins, and nonpenicillin derivatives of phenylacetic acid, such as phenylacetamide and derivatives, were hydrolyzed at a much slower rate. The enzyme displayed no activity on other amides, dipeptides, polypeptides, and proteins. At higher substrate concentrations, hydrolysis of benzylpenicillin was inhibited by the reaction products: 6-APA acts as a noncompetitive inhibitor whereas phenylacetic acid is a competitive inhibitor. They found a K , value of 4.5 mM. Ho and Humphrey (1970) used this enzyme system to illustrate the optimization of an enzymatic process subject to enzyme deactivation. They selected temperature and pH control policies that would maximize the 6-APA yield and minimize the enzyme loss, Recently, Ryu et al. (1972a,b) compared the kinetic properties of the soluble and the bentonite-immobilized form of this extracellular enzyme and found that they exhibited significantly different inhibition constants. A continuous enzyme reactor was designed, consisting of a continuous-flow stirred tank and an ultrafiltration unit; it enables recirculation of the enzyme and continuous removal of the end products. A kinetic model was derived that was based on the inhibition effects of the end products on the enzymatic action. This model was then used, with the aid of a computer, to simulate the performance of a continuous enzyme reactor system and to optimize the productivity of the reactor in terms of process variations (Ryu et al., 1972b). Cooney and Acevedo (1972) used this enzyme as a model system
MICROBIAL PENICILLIN ACYLASES
343
to study the effect of various growth-limiting nutrients on enzyme production in continuous culture; they demonstrated a preferential excretion of penicillin amidase over other proteins during sulfur-limited growth and showed that the enzyme is produced as a growth-associated metabolite ( Acevedo and Cooney, 1973). Nyiri (1967) compared the deacylation of penicillin G and phenylacetylglycine by E. coli NyI/3. He found that acylase activity to be inherent to the exponential growth phase and is strictly endogenous. On the basis of the activity pattern, he suggested that only one enzyme is responsible for hydrolysis of both products. Sjoberg et al. (1967) examined the acylase activity of E. coli BRL 351 on penicillins and cephalosporins, stimulated by the results of Huang et ul. (1963), who demonstrated also that cephalosporins with phenoxyacetyl and phenylmercaptoacetyl as side chain structures were hydrolyzable. However, this acylase with an optimum p H 8 was rather specific for the phenylacetyl group, although penicillin G and 2-thienylacetylpenicillin were preferentially hydrolyzed. Benzylcephalosporin, cephalothin, and cephaloridine also were hydrolyzed; penicillin N and cephalosporin C were unaffected. Together with Cole (1964), they found that the rate of hydrolysis was greatly affected by structural and steric differences in the acylated part of the molecule, steric factors around the amide bond appearing to be of major importance. Kulhanek and Tadra (1963, 1968) demonstrated penicillin G acylase activity in Alcaligenes faeculis VUFB 572. The enzyme was able to hydrolyze also the semisynthetic ampicillin into 6-APA. Cole (1969a-d) reported in detail on the penicillin acylase activity of E . coli NCIB 8743A cells. This strain produced, after induction with phenylacetic acid, an intracellular acylase with a pH optimum of 8.2 at 50°C. As enzyme preparation, a suspension of E. coli cells treated with 1%n-butyl acetate was used. The specificity was directed to the acyl group of the penicillin molecule: p-hydroxybenzylpenicillin was the best substrate, followed by benzylpenicillin; also ampicillin was hydrolyzed into 6-APA, and some semisynthetic cephalosporins were transformed into 7-aminocephalosporanic acid ( 7-ACA ) . Next to penicillin hydrolysis, a wide range of amides, N-acylglycines, and acylated L-&-amino acids were hydrolyzed, indicating that the enzyme involved is rather an amidohydrolase. A K, value of 30 mM for penicillin G hydrolysis was found, A 100-fold purification process results in an enzyme preparation revealing that the penicillin acylase activity is still associated with amidohydrolase activity. In these aspects, the E. coZi NCIB 8743A enzyme closely resembles the E. coli ATCC 9637 enzyme. Penicillin acylase was extracted on a large scale from E. coli
344
E. J . VANDAMME AND J. P. VOETS
NCIB 8743A by mechanical disruption in a Manton-Gaulin homogenizer and purified according to Self et al. (1969). The enzyme was shown to be inhibited by benzylpenicillin and by both the products of hydrolysis. The kinetic and inhibition constants were measured at different pH values. At 37°C and at pH 8.0, a K , value for 6.7 X M was found (Balasinghan et al., 1972). Inhibition by phenylacetic acid was competitive, and 6-APA inhibits noncompetitively, an inhibition pattern which is similar to that found for the Bacillus megaterium enzyme, described by Chiang and Bennett (1967) and Ryu et al. (1972a,b). The E. coli NCIB 8743A enzyme has been immobilized by covalent binding to DEAE-cellulose, using 2-amino-4,6-dichloro-s-triazine. The preparation retained 45-818 of its activity observed before attachment. For the enzyme reaction kinetics, there was no evidence of diffusional limitation of the reaction rate. Using the insoluble enzyme, a K , value of 6.3 x lo-* M was calculated at 37°C at pH 8. The optimum pH for penicillin G hydrolysis was 7.65; while the free enzyme displayed maximal activity at pH 8.2 (Lilly et al., 1972; Warburton et al., 1972). The use and stability of this insoluble enzyme preparation in batch and continuous-flow stirred tank reactors for the conversion of benzylpenicillin into 6-APA was recently described, and a rate equation has been derived to describe this microbial transformation process (Warburton et al., 1973). Lepidi et al. (1970) studied the enzymatic degradation and the morphogenetic effects of penicillin G on Axotobacter chroococcum Beij. C12 Pr and found that this microorganism splits penicillin G into 6-APA and phenylacetic acid. The detection of penicillin acylase activity among nitrogen-fixing bacteria, however, deserves further attention. Recently Nara et al. (1971a,b) and Okachi et al. (1973b) described different bacteria producing an intracellular penicillin G acylase : KluyVera citrophila, K1. noncitrophila, Fseudomonas desmolytica, P. aeruginosa, F. cruciviae, Micrococcus luteus, Proteus morganii, E . coli sp., and Rhodopseudomonas spheroides. The penicillin acylase activity of K . citrophila KY 3641 was studied in detail (Okachi et al., 1972a,b, 1973a; Takasawa et al., 1972). Induction by phenylacetic acid and sodium glutamate and repression by glucose, fructose, maltose, and lactose were typical features, previously described in E . coli by Szentirmai (1964). Penicillin G was hydrolyzed preferentially at pH 7.5 at 35°C. The cells also hydrolyzed ampicillin and, to a much lesser extent, penicillin V and T. Some semisynthetic cephalosporins were hydrolyzed into 7-aminocephalosporanic acid ( 7-ACA) . However, under anaerobic growth conditions, the cells also displayed penicillinase activity. By isolating penicillinase-deficient mutants of this strain, better yields of 6-APA were achieved ( Okachi et at., 1973a).
MICROBIAL PENICILLIN ACYLASES
345
b. Actinomycetales Claridge et at. (1960) reported on a Mycobacterium phlei strain producing a penicillin G acylase. Huang et al. (1960) described a Nocardia FD 46973, preferentially splitting penicillin G into 6-APA at pH 8 at 28OC, although penicillin 0 and penicillin V were also hydrolyzed partially. Recently Nara et al. (1971a) detected an intracellular penicillin G acylase in Streptomyces ambofaciens SPSL-15; this microorganism produced in addition an extracellular penicillin V acylase.
c. Fungi Only recently, fungi were described displaying penicillin G acylase activity. Indeed, Rossi and associates (1973) reported on the enzymatic hydrolysis of penicillin G and other phenylacetylamino compounds by six Neurospora crassa strains. A purified cell extract transformed penicillin G preferentially at pH 7 at 37%. Penicillin V was hydrolyzed much slower, but hydrolysis of a range of N-phenylacetyl-L-a-amino acids was comparable to penicillin C hydrolysis. The substrate specificity of these acylases can be compared with that of E . coli ATCC 9637 (Lucente et al., 1965). 2. Enzymatic Synthesis of Penicillins by Penicillin G Acylases Many penicillin acylase producing bacterial strains not only catalyze the hydrolysis of penicillin G, but also the reverse reaction starting from 6-APA and phenylacetic acid or its derivatives. Synthetic reaction occurs at acid pH values (4.5-5.5), when the 6-APA concentration is raised. This was demonstrated with E . coli ATCC 9637 (Bauer et al., 1960; Bondareva et al., 1969a,b; Kaufmann and Bauer, 1960; Self et al., 1969), E. coli sp. (Rolinson et al., 1960), Alcaligenes faecalis (Claridge et al., 1960), E. coli NCIB 8743A (Cole, 1969c,d), and Kluyvera citrophila (Nara et al., 1971b) cell suspensions or crude extracts. The substrate specificity of the hydrolytic and synthetic reaction was identical, so many authors concluded that only one enzyme is involved in these processes. The synthetic activity of the acylase of E . coli NCIB 8743A was studied in more detail by Cole (1969c,d). The optimum pH for resynthesis was 5.0. The rate of penicillin G synthesis improved in the presence of energy-rich compounds when phenylacetylglycine was added to 6-APA instead of phenylacetic acid. According to Brunner et al. (1968), resynthesis can occur also in a nonenzymatic way, if at least the appropriate energy-rich side chain group is available. Penicillin G, ampicillin, and hydroxypenicillin were synthesized in this way. This acylase also cata-
346
E. J . VANDAMME AND J . P. VOETS
lyzed the acylation of hydroxylamine by acids or amides to yield hydroxamic acids. Whether this transacylase activity, and the resynthetic activity, are due to the action of different enzymes or to the multiple activities of an aspecific amidase will not be known until extensive enzyme purification has been carried out. However, Bondareva et a,?. (1969b) were able to demonstrate both hydrolytic and synthetic activity with a purified, electrophoretically homogeneous acylase preparation from E . coli ATCC 9637 cells. Recently, attempts were made to isolate ampicillin synthesizing enzymes. Nara et al. (1971a,b, 1972) reported that Kluyvera citrophila KY 3641 and its penicillinase-deficient mutants were able to produce D- ( - ) -a-aminobenzylpenicillin from 6-APA and phenylglycine derivatives in good yields. The optimum pH for synthesis was 6.5. Among various phenylglycine derivatives examined as substrates, D-phenylglycine methyl ester was the best compound, giving high ampicillin levels (Okachi et al., 1972a,b, 1973a; Takasawa et al., 1972). Profiting from this synthetic activity of penicillin G acylases, enzymatic synthesis of many semisynthetic penicillins has been achieved ( Bondareva et al., 1969b; Cole, 1969c,d; Kaufmann and Bauer, 1960; Nara et al., 1972; Okachi et al., 1972a,b). However, the rather low yield and the reversibility of the process have precluded so far that this enzymatic synthesis of penicillins might result in an industrially applicable process, although the use of insolubilized acylases for this synthesis looks promising (Self et al., 1969). Relevant properties of penicillin G acylases are represented in Table VI.
D. AMPICILLINACYLASES Recently, microorganisms that produce penicillin acylase have been selected which display a novel substrate spectrum. Indeed Nara et al. (1972) and Okachi et al. (197313) reported on an ampicillin acylase so far encountered especially among Pseudomonas strains. These enzymes form 6-APA only from ampicillin, but not from penicillin G, V, or N or other semisynthetic penicillins. Hence, this enzyme can be considered as a novel type of penicillin acylase. At pH 7.5, specific hydrolysis of ampicillin into 6-APA was observed with four strains: Pseudomonas melanogenum KY 3987, KY 4030, and KY 4031 and P. ovalis KY 3962 ( see Table VI ) , Especially Pseudomonas melanogenum KY 3987 received full attention (Nara et al., 1972; Okachi et al., 1973b). The enzyme was purified by the use of Sephadex chromatography and electrofocusing methods. In addition to a p-lactamase, this strain possessed an ampicillin acylase
MICROBIAL PENICILLIN ACYLASES
347
producing rather high levels of ampicillin at pH 5.5-6 at 34OC from 6-APA and DL-phenylglycine methyl ester. Penicillinase-deficient mutants were derived from the parent strain and were found to produce ampicillin in better yield ( Okachi et al., 1973a,b). These enzymes or their insolubilized enzyme reactors seem promising in view of the enzymatic synthesis of semisynthetic penicillins, especially ampicillin, on an industrial scale. VI.
Physiological Role of the Penicillin Acylases
The relatively widespread occurrence of the enzyme among different genera suggests that its acyl-hydrolyzing function cannot be restricted to the penicillin-type structures but might be of more general significance in microbial metabolism. However, lack of substrate spectra of a wide range of highly purified enzymes precluded arriving at a correct conclusion. About the physiological role of the enzyme found in microorganisms that do not produce penicillin, little is known, and it can be interpreted as a cometabolic function of an aspecific enzyme, probably an amidohydrolase. However, the narrow substrate specificity of the few penicillin V acylases which have been purified to homogeneity does not support this theory. The specific localization of penicillin V acylase in Fusarium moniliforme conidia suggests a role in the dormancy state of these spores ( Vandamme, 1972; Vandamme et al., 1971a), The Streptomyces acylases detected by Haupt and Thrum (1967) are thought to be in fact aspecific enzymes having a function in the detoxification process of aromatic acids, and they should be considered as aspecific amidases. This detoxification process of toxic aromatic acids might consist in a transformation into corresponding amides or in a binding to amino acids, among them 6-APA provided by penicillin acylase action. Holt and Stewart (196413) proposed that penicillin acylase has a role in the resistance of bacteria against penicillins, this hypothesis being soon rejected on solid grounds by Cole and Sutherland (1966) and Sutherland ( 1964) : 1. Penicillin acylases display a low affinity for their substrates. In uitro, hydrolysis may be a rather fast process, but with substrate concentrations obtained in vivo the activity is strongly reduced. 2. The pH optimum for an effective deacylation process of the penicillin lies in the alkaline range 8-9, so in uiuo maximal activity will never or rarely be obtained. 3. Enzyme biosynthesis is inhibited by growth of the microorganism
TABLE VI RELEVANT PROPERTIES OF PENICILLIN G ACYLASES A N D AMPICILLINACYLASES Optimum pH
Microorganisms
Hydrolysis
Synthesis
Optimum temp. ('C)
Localization
K,
Molecular weight
References
Penicillin G acylases Bacteria Alcaligenes sp. Alcaligenes faecalis Escherichia coli ATCC 9637 Intact cells Enzyme Insoluble enzyme E . coli NCIB 87438 Intact cells Enzyme .Insoluble enzyme
8 7.5
40 37
Intracellular Intracellular
Claridge et al. (1960) Claridge et al. (1963)
Intracellular
Kaufmann and Bauer (1960) Bondareva et al. (1969a,b) Self et al. (1969)
7.5
4.5-5.5
30
7.8-8 7.5
5
50-52 37
7.7mM
8.2 8.2 7.65
50 37 -
Intracellular
E . coli sp. E . coli sp.
7 7.5
30-3 5 -
In trace11ul ar Intracellular
E . coli C15 (NCIB 9465) Proteus rettgeri FD 13424 Micrococcus roseus ATCC 516 Bacterium faecalis alc. 415 B . megaterium ATCC 14945 Enzyme
5.5 8 9 7.8
35 42
Intracellular Intracellular Intracellular
8.5
-
Extracellular
6.7 6.3
30 mM x 10-4 M x 10-4 M
1.35-1.59mM 17.5 mM 4.00 mM
4.5mM
Cole (1969a-d) Balasingham et al. (1972) Warburton et al. (1972), Lilly et al. (1972) Brand1 (1965) Badr-Eldin and Attia (1973) Holt and Stewart (1964a) Huang et al. (1963) Pruess and Johnson (1965) Gotovtseva et al. (1965) 120,000 Chiang and Bennett (1967)
m ?
Insoluble enzyme Azotobacter chroococcum Beij. C12Pr. Kluyvera citrophila KY 3641 Actinomycetales Nocardia F D 46973 Streptomyces ambofaciens SPSL-15 Fungi Neurospora crassa
6 . 0 mM 7.2 7.5
6.5
25 35
Intracellular Intracellular
Ryu et al. (1972a,b) Lepidi el al. (1970) Nara et al. (1971a), Okachi et al. (1972b)
8 7.4
-
28 28
Intracellular Intracellular
Huang et al. (1960) Nara et al. (1971a)
7.0
-
37
Intracellular
Rossi et al. (1973)
7.5 7.5
5.5-6
z
Ampicillin acylases Pseudomonas melanogenum Pseudomonas ovalis
34
Intracellular
Okachi el al. (1973b)
P
F
sri W
350
E. J . VANDAMME A N D J. P. VOETS
at temperatures above 3OoC and by the presence of some carbohydrates and polyalcohols. 4. The end product of penicillin hydrolysis, 6-APA, is in many cases as active as penicillin. 5. Recently, Cole et al. (1973) followed the fate of penicillins in healthy humans and could detect only penicilloic acid formation. Only traces of 6-APA were found. It was shown by Kameda et al. (1961) and Szentirmai (1964) that many bacteria that produce penicillin acylase are able to use phenylacetic acid as a sole source of carbon. According to Cole (1964), Kaufmann and Bauer (1964b), and Lucente et al. (1965),the enzyme is a phenylacetylase or an amidase and should interfere in the metabolism of acylamides and derivatives. In the presence of benzylpenicillin, these bacteria should be nutritionally favored in comparison with strains not possessing an acylase. Gotovtseva et al. (1965, 1968) found indications that the penicillin acylase of Bacterium faecalis alcaligenes 415 is involved in the tryptophan metabolism of this strain. Szentirmai (1964) and Kaufmann (1964) observed that low concentrations of biologically active penicillins inhibit the activity of the Escherichia coli acylase. As a result, Kaufmann (1964) suggested that this acylase probably has a role in cell wall biosynthesis of gram-negative bacteria and is the target of the antibacterial action of the penicillins. According to Lepidi et al. (1970) the hydrolysis of penicillin G into 6-APA is coupled with the transfer of the phenylacetyl group on cellular acceptors, mainly on cell wall constituents. Such transfer appears to involve the intermediate formation of phenylacetyl coenzymes, according to the following reactions:
+
+
penicillin G coenzyme A -+ phenylacetyl-coenzyme A BAPA phenylacetyl-coenzyme A acceptor + phenylacetylacceptor coenzyme A
+
+
(1) (2)
Nyiri (1967) and Vandamme ( 1972) supposed penicillin acylase to be connected with the protein metabolism of the cell, and Vandamme (1972) and Vandamme and Voets (1973) found indications that the enzyme might be involved in transpeptidase reactions as described by Strominger et aZ. ( 1971). According to Hamilton-Miller ( 1966), penicillin acylase activity might be an incidental property of a structural protein. The detection of penicillin acylases able to synthesize penicillins from 6-APA and an appropriate side chain structure among p-lactam antibiotic producing molds has led to the suggestion that these enzymes might be involved in the biosynthesis process of these antibiotics (Cole, 1966; Erickson and Dean, 1966; Spencer and Maung, 1970).
MICROBIAL PENICILLIN ACYLASES
351
Enzymes with analogous action have been detected especially among penicillin-producing molds; they were named penicillin acyltransferase or transacylase and are capable of transferring the side chain from a penicillin to 6-APA. In this case, the penicillin can be seen as an activated side chain molecule, transferable to 6-APA by a modified acylase (Abraham, 1971; Brunner et al., 1968; Gatenbeck and Brunsberg, 1968; Lemke and Brannon, 1972; Pruess and Johnson, 1967; Spencer, 1968). It has been suggested by Lemke and Brannon (1972) that the transacylase is a membrane or particle-bound form of the acylase enzyme. However, these transacylases (and the acylases) cannot transfer or exchange or hydrolyze the side chain structure from isopenicillin N, the penicillin structure now believed to be the origin of the other penicillins ( Demain, 1966; Abraham, 1971). However, the complete pathway of penicillin biosynthesis is not yet known (Abraham, 1971; Arnstein and Morris, 1960a,b; Banks et al., 1969; Bauer, 1970; Bodansky and Perlman, 1969; Demain, 1966; Katz, 1971; Maier and Groger, 1972). Nevertheless, the real involvement of the acylase enzyme in penicillin biosynthesis recently received support from the experiments of Spencer and Maung ( 1970) and of Abraham (1971). Spencer and Maung (1970) observed four enzymatic activities in Penicillium chrysogenum 51-20F3: ( 1) penicillin acyltransferase, which catalyzes exchange of side chains from penicillins into 6-APA; ( 2 ) 6-APA transacylase, which synthesizes penicillin from acyl coenzyme A and 6-APA; ( 3) penicillin acylase, which hydrolyzes penicillin into 6-APA and the side-chain acid; (4) phenylacetyl-coenzyme A hydrolyzing enzyme. Disruption of the mycelium, precipitation with (NH,),SO, and fractionation on Sephadex G-100 and DEAE-cellulose yield a 130-fold purification of each of the activities. The ratios of the activities remained constant throughout the various purification stages and were detected in the same fraction. On polyacrylamide-gel electrophoresis, all four activities were found in the same band. Influence of pH and temperature on enzymatic activity have an identical pattern. Spencer and Maung concluded that these four activities are originated by one and the same enzyme, whatever its name. This consideration suggested that penicillin acylase has indeed a role in the biosynthetic mechanism of penicillins, however, and the transformation of penicillin into 6-APA is one of its multiple activities. The recent findings of Abraham (1971) and Abraham and Loder (1972) showing that both isopenicillin N and 6-APA stimulate penicillin G biosynthesis might be another positive indication of the biosynthetic potential of the acylases. According to Cooper ( 1972) , a derivative of L-cysteinybhydrovaline ( thiazoline-azetidinone ) might represent an alternative model as a possi-
352
E. J . VANDAMME AND J . P. VOETS
ble intermediate in the biosynthesis of the p-lactam antibiotics; the compound might be converted by enzymatic action (possibly acylase or transacylase activity) of Penicillium sp. into penicillins. In addition to fungi, a wide range of Streptomyces sp. have recently been described to be able to produce p-lactam antibiotics (Gauze et al., H72; Higgens and Kastner, 1971; Miller et al., 1962; Nagarajan et al., 1971; Stapley et al., 1972). As penicillin acylases are known to be present in all the fungal strains producing p-lactam antibiotics, the presence of this enzyme among the antibiotic-producing Streptomyces sp. is not yet established. On the other hand, Streptomyces sp. have been described that could produce penicillin acylase, but were not checked for p-lactam antibiotic production ( Batchelor et al., 1961d; Dennen et al., 1971; Haupt and Thrum, 1967; Nara et al., 1971a). Although some evidence is emerging about the role of penicillin acylases in the biosynthetic process of the p-lactam antibiotics, this is far from true for their function in the physiology of non-penicillin-producing molds, actinomycetales, yeasts, and bacteria. As suggested by the differences in the properties of the described penicillin acylases, this activity can hardly be considered to result from one particular enzyme widely distributed throughout the microbial biochemistry. Only a detailed study of purified penicillin acylases will show whether the penicillin acylases are functionally and structurally identical or whether they are different enzymes displaying a common cometabolic activity. Vll.
Coexistence of Acylase and @-Lactamase
p-Lactamase activity was first described by Abraham and Chain (1940). p-Lactamases hydrolyze the amide bond of the p-lactam ring of 6-APA and its derivatives. Penicillins ( and cephalosporins) are thereby transformed into penicilloic ( cephalosporoic) acids, compounds without any antibacterial activity. This enzyme has been detected as well in gram-positive as in gramnegative bacteria and is extensively described (Citri and Pollock, 1966; Jack and Richmond, 1970a,b; Jack et al., 1970; Kuwabara, 1970; Malik et al., 1970; Richmond and Sykes, 1973; Vandamme and Voets, 1971). According to several authors, p-lactamase activity was found to occur together with penicillin acylase activity in one and the same microbial strain. This coexistence was demonstrated by Batchelor et al. (1961d) and by Haupt and Thrum (1967) in Streptomyces lavendulae BRL 198 and Streptomyces netropsis 282%; by Cole and Sutherland (1966) in Escherichiu coli I 187; by Nyiri (1967) in Nocardiu sp., by Pruess and Johnson
MICROBIAL PENICILLIN ACYLASES
353
(1965) in E. coli ATCC 9637; by Kulhanek and Tadra (1968) in Alcaligenes faecalis; by Arcos et al. (1968) in E. coli 0127:K63 (B8):H; by Rozansky et al. (1969) in E . coli, Alcaligenes, and Proteus sp.; by Dennen et al. (1971) in Cephalosporium acremonium ATCC 11550; and by Nara et al. (1972) in Pseudomonas melanogenum KY 3987 and Kluyvera citrophila KY 3641. Recently, Okachi et al. (1973a,b) reported on the isolation of penicillinase-deficient mutants of Kluyvera citrophila and Pseudomom melanogenum, displaying higher acylase levels compared with the parent strains. The coexistence of the two enzymes in one strain has until now precluded accurate determination of both activities. The presence of penicilloic acid, 6-APA, and eventually penicic acid in the reaction mixture is a qualitative confirmation of this coexistence. On the other hand, spontaneous and chemical hydrolysis of the p-lactam bond can wrongly be considered as a result of p-lactamase activity (Scheme 3 ) . However, quantitative determination of one enzyme in the presence of the other would be possible if one activity could be blocked without interfering with the other. In these respects inhibiting p-lactamases by adding, besides the substrate, an acylase-resistant semisynthetic penicillin or cephalosporinwell known for their synergistic action-might allow an accurate acylase determination (Bach et al., 1966; Bobrowski and Borowski, 1971; Cole et al., 1972; Hamilton-Miller, 1971a,b; OCallaghan and Morris, 1972). The blocking of acylase activity can be performed by adding, as substrate for p-lactamases, acylase-resistant penicillins, 6-APA included, or by al-
Penicillins
Penicilloic acids
6- Aminopenicillanic acid
Penicic acid
SCHEME3. Combined action of acylase and p-lactamase on penicillins.
E . J . VANDAMME AND J . P. VOETS
354
lowing the enzymatic reaction to proceed at acid pH values, since most penicillin acylases display maximal activity at pH 8. VIII. Penicillin Acylase and Cephalosporins Although Huang et al. (1963), Sjoberg et al. (1967), and Cole (1969a) described E. coli, Proteus, and Nocardia penicillin acylases, and Nara et al. (1971b) described Kluyvera citrophila acylase capable of hydrolyzing synthetic N-acyl derivatives of 7-aminocephalosporanic acid ( 7-ACA ) in addition to penicillin G (see Table VII ), as yet no penicillin acylase has been reported which can remove the side chain of the natural cephalosporin C in order to obtain 7-ACA (Scheme 4). However, it is relevant that, although penicillin acylases surely are not cephalosporin C acylases, many p-lactamases ( penicillinases) display cephalosporinase activity ( Ayliffe, 1965; Demain et al., 1963; Garber and Friedman, 1970; Hamilton-Miller et al., 1970; Richmond and Sykes, 1973). TABLE VII ENZYMATIC PRODUCTION O F 7-AMINOCEPHALOSPORANIC ACID (7-ACA) FROM SEMISYNTHETIC CEPHALOSPORINS Microorganisms
Substrates
References
Escherichia coli BRL 351 Nocardia F D 4697, Proteus rettgeri F D 13424 E . coli NCIB 8743A, E. coli BRL 351 Kluyvera citrophila K Y 3641
Benaylcephalosporin N-Phenoxyacetyl-7-ACA, N-Phenylmercapto-7-AC A Cephalothin, cephaloridin
Sjoberg el al. (1967) Huang el al. (1963)
~
Cephalothin
THz HOOC -CH(CH,),CONH-HC
Sjoberg et al. (1967), Cole (1969a) Nara et al. (1971b)
H S -C’ ‘YH2 I I OC-N, ,C c’ ‘-CH,OCOCH, I COOH
Cephalosporin C
1 H , N - a C I- 3 S t y H , 0C-N‘- C +C CH,OCOCH, I
COOH
-
7 Aminocephalosporani c acid
SCHEME 4. Transformation of cephalosporin C into 7-aminocephalosporanic acid.
MICROBIAL PENICILLIN ACYLASES
355
The distribution of a cephalosporin C acylase hydrolyzing cephalosporin C into 7-ACA has been limited only to a few species of Brevibacterium, Achromobacter ATCC 14696, and Flavobacterium (Cole, 1967; Sebek and Perlman, 1971); this is in contrast to the well distributed and industrially applied penicillin acylase. In addition to this limited number of cephalosporin C acylases, during their hydrolytic action the acetoxy group is lost also, so that deacetyl-7-ACA (7-ADCA) is produced rather than 7-ACA. The 7-ACA required for the preparation of semisynthetic cephalosporins, because of a lack of specific cephalosporin C acylases, is instead obtained by an efficient chemical hydrolysis of cephalosporin C (Chauvette et al., 1962; Loder et al., 1961; Morin et al., 1962, Sassiver and Lewis, 1970). As a result of the successful enzymatic synthesis of semisynthetic penicillins, recently attempts were made to prepare semisynthetic cephalosporins enzymatically. Indeed, Japanese workers have detected bacterial enzymes able to synthesize cephalosporins, starting from 7-ACA and an appropriate side chain ( Takahashi et al., 1972). As side-chain structures, a-D-phenylglycine methyl ester and analogs (glycine ethyl ester, D-alanine ethyl ester, D-leucine ethyl ester, and the methyl esters of D-a-cyclohexenylglycine, D-a-hydroxyphenylglycine, and D-a-cyclohexylglycine) were coupled to 7-ACA or 7-ADCA at pH 6 at 37OC by enzymatic action of bacteria belonging to the Pseudomonadaceae, including Xanthomonas oryxae IF03995, X . citri IF03835, Acetobacter pasteurianus ATCC6033, A. turbiduns ATCC9325, Glucombacter suboxydans ATCC621, Pseudomonm mdanogenum IF012020, Mycoplana dimorpha IF013213, and Protaminobacter aluoflauus IF013221. This cephalosporin acylase activity was subsequently found among a wide range of bacteria: Aeromonas, Alcaligenes, Achromobacter, Beneckea, Eschem'chia, Staphylococcus, Arthrobacter, Proteus, Corynebacterium, Flavobacterium, Clostridium, Spirillum, and Bacillus. However, the enzymatic synthesis of cephalosporin C starting from 7-ACA and a-aminoadipic acid or its derivatives was not reported. The enzymatic activity was clearly different from that of penicillin acylase, and the strains involved displayed no penicillin acylase activity at all, indicating once more that penicillin acylases display no cephalosporin C acylase activity, and vice versa. On the other hand, successful attempts were made to prepare cephalosporins chemically starting from the penicillin molecule, this approach being evolved from the structural similarities of the penicillins and cephalosporins (Barton and Sammes, 1971; Heusler, 1972; Cooper et al., 1973) .
356
E. J . VAND-ME
IX.
AND J. P. VOETS
Nonmicro bia I Penicillin Acyla ses
Penicillin acylase activity was also found in higher organisms. Weitnauer detected in tissue homogenates, prepared from kidney, liver, spleen, or lung of cattle and pigs, penicillin V acylase activity ( Hamilton-Miller, 1966). Cole ( 1964) reported on penicillin V acylase activity in pig kidney extracts, which also can hydrolyze phenoxyacetylglycine. However, kidney preparations have long been known as a source of aspecific deacyIating enzymes, which are highly active against N-acylated amino compounds ( Birnbaum et al., 1952). Holt and Stewart (1964a) prepared from pig gastric mucosa a crude aminopeptidase extract capable of transforming penicillin G into 6-APA. Alburn, Grant, and Clark transformed penicillin V into 6-APA, using a proteolytic enzyme, ficin, isolated from trees of the genus Ficus (Hamilton-Miller, 1966) . According to Cole (1964), these enzymes are not specific for the penicillins mentioned and can be classified as broad-spectrum amidases.
X.
Aspecific Amidases
By several authors, penicillin acylases are named penicillin amidases. However, this name only leads to confusion. Real penicillin amidases, which hydrolyze penicillin amides into free penicillins and ammonia, were detected in Rhodotorula gracilis and Torulopsis sp. (Huang et al., 1963). Real penicillin acylases transform these penicillin amides into B-APAamide (Cole, 1969a). Penicillin amidases belong to this class of enzymes, specifically acting on penicillins, substituted in the 3-carboxyl function ( Hamilton-Miller, 1966). Next to this special penicillin amidase, many aspecific amidases or amidohydrolases are described (Arens et al., 1970; Behal and Carter, 1971; Brady, 1969; Brammar and Clarke, 1964; Clarke, 1970; Irnada et al., 1973; Tiller and Brunneman, 1970; Wade et al., 1971). The enzyme described by Kelly and Kornberg (1964) of Pseudomonas aeruginosa catalyzed the hydrolysis of aliphatic amides and the transfer of the liberated acyl group to hydroxylamine, resulting in the formation of hydroxamic acids. Experiments with Mycobacterium smegmatis (Draper, 1967) revealed that both activities of the amidase and acyltransferase are closely related. Clarke (1970) studied in detail the aliphatic amidases of Pseudomonas aeruginosa 8602 and detected hydroIytic as well as transferase activity.
MICROBIAL PENICILLIN ACYLASES
357
Tiller and Brunneman ( 1970) detected amidohydrolase activity among a large number of Enterobacteriaceae able to hydrolyze aliphatic and aromatic amides, and Wade et al. (1971) and Imada et al. (1973) reported the asparaginase and glutaminase activities of different bacteria, actinomycetes, fungi, and yeasts. Van Heijenoort and Van Heijenoort ( 1971) described amidase activity in Escherichia coli cells, hydrolyzing the N-acetylmuramyl-L-alanine amide bond in the bacterial cell wall peptidoglycan. This amidase, which has a role in the cell wall biosynthetic process, might possibly be related to the penicillin acylase of Escherichia coli ATCC 11105, which is located in the cell wall membrane complex. Fry and Lamborg (1967) described in Escherichia coli ATCC 9637 extracts an amidohydrolase specifically hydrolyzing N-formyl-L-methionine. This strain produces also a penicillin G acylase (Kaufmann and Bauer, 1960) and a L-asparaginase (Arens et al., 1970). As is evidenced by Arens et al. (1970), and Bauer et al. (1971), the L-asparaginase displays no penicillin acylase activity, but whether the amidohydrolase described by Fry and Lamborg (1967) is analogous to penicillin acylase or to the asparaginase is not known. Nevertheless, Cole ( 1969b,c) has described an Escherichia coli NCIB 8743 A penicillin G acylase, which also displays acyltransferase activity, and which binds phenylacetyl groups and hydroxylamine with formation of hydroxamic acids. However, these experiments were carried out with suspensions of intact cells as enzyme source, not with highly purified enzymes. A wide range of fungi also display amidase activity (Imada et al., 1973). Lanzilotti and Pramer (1970) detected an acylamidase in FUSUSium solani extracts, specifically hydrolyzing N-acetyl arylamines. Hynes and Pateman (1970) reported on the amidase activity of Aspergillus niduluns. Penicillium chrysogenum amidase was described by Imada et al. (1973). Brady ( 1969) purified the amidohydrolase of Candida uti2is ICG-3093 60-fold. The preparation hydrolyzed aliphatic as well as aromatic amides and in addition displayed transferase activity. Dennen et al. (1971) described arylamidase activity in Cephalosporium acremonium (which also produces penicillin N and cephalosporin C ) , hydrolyzing L-leucyl-p-naphthylamide. The 300-foId purified exocellular enzyme catalyzed also the hydrolysis of the p-Iactam ring of cephalosporin, but not that of penicillins. This amidase clearly displays p-lactamase activity and is evidently different from the penicillin acylase, already described within this species by Cole ( 1966). From these results, it can be stated that penicillin hydrolysis into 6-APA can hardly be considered to be a result of the activity of aspecific
358
E. J . VANDAMME AND J . P. VOETS
amidases or amidohydrolases. However, most studies are concerned with intact cells as enzyme source and do not warrant conclusions about specificity and substrate spectrum. Nevertheless, the few purified preparations of amidase tested as to this property revealed no relationship with the penicillin acylases. In addition to these microbial amidases, tissue amidases were isolated. Birnbaum et al. (1952) prepared an amino acid acylase from dog kidneys, splitting N-acylated amino acids. Ellis and Perry (1966) described arylamidases extracted from bovine pituitary glands, and Lugay and Aronson ( 1969) purified an aminoacylase from Palo Verde (Parkinsonia aculeata L. ) seeds, which hydrolyzed specifically N-formyl-Lmethionine and some related N-acylamino acids, an activity also found in the penicillin G acylase-producing Escherichia coli ATCC 9637 strain (Fry and Lamborg, 1967; Kaufmann and Bauer, 1964b). If these tissue amidases and transferases also display penicillin acylase activity, as is suggested by Cole (1969a,b) for Escherichia coli NCIB 8743 A amidase was not examined. Here again, one must wait for experiments with purified enzymes for proof of these statements.
XI.
Chemical Transformation of Penicillins into 6-APA
6-APA was originally produced by fermentation (Batchelor et al., 1961c), and Sheehan and Henery-Logan ( 1959) synthesized 6-APA in a purely chemical way. However, enzymatic transformation of biosynthetic penicillins into 6-APA is now an industrial process of economic importance as the main source of 6-APA used in the preparation of semisynthetic penicillins. Recently a chemical alternative for this enzymatic hydrolysis has been described which might be of practical importance, since it would compete with the enzymatic process. Weissenburger and Vanderhoeven ( 1970) reported on a successful application of the chemical transformation process of cephalosporin C into 7-ACA on the more labile benzylpenicillin. The selective cleavage of the amide bond has been accomplished by treating the silyl ester of benzylpenicillin with phosphorus pentachloride at -40°C. The iminochloride produced was converted into the iminoether with n-lbutanol, and the latter compound was hydrolyzed by pouring the reaction mixture into water. After the pH was adjusted to 4.1, 6-APA crystallized from the heterogeneous system as a pure white material in an excellent yield of 9%. These reactions are presented in Scheme 5. Fosker et al. (1971) described a nonenzymatic conversion of penicillin G into the semisynthetic penicillins cloxacillin, phenethicillin, and methi-
359
MICROBIAL PENICILLIN ACYLASES
H C6H,CH,CONp,S
0
CHs Me,SiCl, cHs PCl, COOH -40°C
*
COO
I
&Me,
COOH
SCHEME5. Chemical transformation of penicillin G into 6-aminopenicillanic acid according to Weissenburger and Vanderhoeven ( 1970).
cillin in a yield of about 60%, without the need for isolating the intermediate 6-APA. Very efficient and large-scale 6-APA production by both the enzymatic and chemical transformation process is now well established. However, the recent development of continuous penicillin acylase reactions and insolubilized penicillin acylase preparations certainly favor the enzymatic procedure (Carrington, 1971; Faith et al., 1971; Lilly et al., 1972; Ryu et al., 1972a,b; Self et al., 1969; Smiley and Strandberg, 1972; Warburton et al., 1973; Wingard, 1972).
XII.
Concluding Remarks
Industrial production of 6-APA by direct fermentation processes is now outdated in favor of the enzymatic hydrolysis of penicillins into 6-APA, which in turn might in the near future compete with chemical hydrolysis procedures. Both processes require penicillin G or V as substrate, and the industrial fermentation of these penicillins and their recovery are now a well established process, delivering a cheap and reliable source of raw material for transformation into 6-APA (Chain, 1971). Enzymatic transformation into 6-APA, catalyzed by penicillin acylases, is from the point of view of the fermentation industry nearly a completely solved procedure, although the introduction of insoluble enzyme reactors has given rise to novel problems. Since 6-APA became available, many semisynthetic penicillins have been prepared. Thereby penicillins with improved oral absorption, resistance to p-lactamase, and to a lesser degree increased activity toward gram-negative bacteria, have been made available. Many other N-substitutions are possible, however, but these have not so far resulted in useful compounds.
360
E. J . VANDAMME A N D J . P. VOETS
From the academic viewpoint, however, several striking problems still await clarification. In the search for new acylase-producing strains, one is limited by the fact that no specific selection procedures are available. Selective and specific screening procedures should be tested out, but this might be possible only if the role of this enzyme in microbial physiology is further elucidated. As is evidenced by recent experiments, both fungi and bacteria were found to produce penicillin V as well as penicillin G acylases, and ampicillin acylases were found among bacteria. The screening procedures until recently applied allowed many acylases to remain undetected. As a result the classification of acylases in type I and type I1 (Claridge et al., 1963; Hamilton-Miller, 1966) seems no longer to receive support. The specificity of the acylases is clearly correlated to the side-chain structure of the penicillin molecule, although the penicillin nucleus also has to be intact, as is evidenced by lack of activity or very low activity on penicilloic and penilloic acids. It seems, however, that penicillin acyIases are rather specific enzymes, not aspecific amidases, as is suggested by experimenting with intact cell suspensions as the enzyme source. The few purification processes on penicillin acylases all illustrate that the substrate specificity of these enzymes is quite narrow, especially in the case of the penicillin V acylases, irrespective of the type of microorganism producing it. Nevertheless, relevant data, such as molecular weight, amino acid composition, or even sequence, would allow comparison of the different acylases. Substrate pattern and localization studies would help to raise the curtain on the role of this enzyme in microbial biochemistry. Insolubilization procedures are now widely tested upon penicillin acylases, using intact cells as well as partially purified acylases; it seems likely that in the near future the use of penicillin acylase supported on cellulose by chemical coupling with triazines will be employed commercially in the production of 6-APA. Concerning the reverse action of the enzymes, recent experiments with purified acylases indicate that one enzyme with multiple activities is responsible, although this reversibility is not a common property (Bondareva et aZ., 1969b; Nara et al., 1972; Okachi et al., 1973a,b; Ryu et al., 1972b; Spencer and Maung, 1970). Quite another approach, which deserves full attention, is the use of the synthetic ability of strains that produce acylase. Indeed, chemical coupling of bulky side chains to 6-APA in order to prepare semisynthetic penicillins invariably is associated with the need for chemical protection of certain groups and bonds. This laborious process might become com-
MICROBIAL PENICILLIN ACYLASES
361
petitive with, or even be displaced by, an enzymatic coupling procedure using soluble or insolubilized acylases. In addition to the coexistence of penicillin acylase and p-lactamase, some authors have indicated that some microorganisms might possess more than one penicillin acylase (Cole, 1966; Nara et al., 1971a; Pruess and Johnson, 1965; Vandamme et al., 1971b). The role of the enzyme in the biosynthetic process of the penicillins is not yet fully understood, although recent experiments indicate that the enzyme might have multiple activities in the process. However, many fascinating problems await further resolution of the properties of microbial enzymes, which are responsible for restoring the health and saving of the life of so many people.
REFERENCES Abraham, E. P. (1971). Pure Appl. Chem.28,399. Abraham, E. P., and Chain, E. B. (1940). Nature (London) 146,837. Abraham, E. P., and Loder, P. B. (1972). Abstr., Int. Ferment. Symp., 4th, 1972, p. 82. Abraham, E. P., and Newton, G. G. F. ( 1961). Biochem. I . 79,377. Abraham, E. P., Chain, E. B., Fletcher, C. M., Florey, H. W., Gardner, A. D., Heatley, N. G., and Jennings, M. A. (1941). Lancet 2, 177. Abraham, E. P., Newton, G. G. F., Schenk, J. R., Hargie, M. P., Olson, B. H., Schuurmans, D. M., Fischer, M. W., and Fusari, S. A. (1955). Nature (London) 176, 551. Acevedo, F., and Cooney, C. L. ( 1973). Biotechnol. Bioeng. 15,493. Acred, P., Brown, P. M., Knudsen, E. T., Rolinson, G. N., and Sutherland, R. ( 1967). Nature ( London) 215,25. Alicino, J. F. (1961). Anal. Chem. 33, 648. Ambler, R. P., and Meadway, R. J. (1969). Nature (London) 222, 24. Anderson, E. S . (1968). Annu. Rev. Microbiol. 22, 131. Arcos, J. M., Ruiz, M. C., and Mugica, J. D. (1968). I. Bacteriol. 96, 1870. Arens, A., Rauenbusch, E., Irion, E., Wagner, O., Bauer, K., and Kaufmann, W. ( 1970). Hoppe Seykr’s 2. Physiol. Chem. 351, 197. Arnstein, H. R. V., and Morris, D. ( 1960a). Biochem. J. ‘76, 323. Arnstein, H. R.V., and Moms, D. (1960b). Biochem. J. 76, 357. Axelrod, J., Meyers, B. R., and Hirschman, S. Z. (1971). Appl. Microbiol. 22, 904. Ayliffe, G. A. J. ( 1963). f. Gen. Microbiol. 30, 339. Ayliffe, G. A. J. (1965). J. Gen. Microbiol. 40, 119. Bach, J. A., Buono, N., Chisholm, D., Price, K. E., Pursiano, T. E., and Gourevitch, A. (1966). Antimicrob. Ag. Chemother. p. 328. Badr-Eldin, S. M., and Attia, M. M. (1973). A d a Microbiol. Pol., Ser. B 5, 43. Balasingham, K., Warburton, D., Dunnill, P., and Lilly, M. D. (1972) Biochim. Biophys. Acta 276, 250. Ballio, A., Chain, E. B., Dentice di Accadia, F., Mauri, M., Raver, K., Schlesinger, K. J., and Schlesinger, S. ( 1961). Nature (London) 161, 909.
362
E. J . VANDAMME AND J. P. VOETS
Banks, G. T., Buck, K. W., Chain, E. B., Darbyshire, J. E., and Himmelweit, F. (1969). Nature (London) 222, 89. Barton, D. H. R., and Sammes, P. G. ( 1971). Proc. Roy. SOC., Ser. B 179,345. Batchelor, F. R., and Cameron-Wood, J. (1962). Nature (London) 195, 1000. Batchelor, F. R., Doyle, F. P., Nayler, J. H. C., and Rolinson, G. N. (1959). Nature (London) 183, 257. Batchelor, F. R., Chain, E. B., and Rolinson, G. N. (1961a). Proc. Roy. SOC., Ser. B 154, 487. Batchelor, F. R., Chain, E. B., Hardy, T. L., Mansford, K. R. L., and Rolinson, G. N. (1961b). Proc. Roy. Soc., Ser. B 154, 498. Batchelor, F. R., Cameron-Wood, J., Chain, E. B., and Rolinson, G. N. ( 1 9 6 1 ~ ) . Proc. Roy. SOC., Ser. B 159, 514. Batchelor, F. R., Chain, E. B., Richards, M., and Rolinson, G. N. (196ld). Proc. Roy. SOC., Ser. B 154, 522. Batchelor, F. R., Gazzard, D., and Nayler, J. H. C. (1961e). Nature (London) 191, 910. Batchelor, F. R., Cole, M., Gazzard, D., and Rolinson, G. N. (1962). Nature (London) 195, 954. Bauer, K. (1970). Z. Naturforsch. B 25, 1125. Bauer, K., Kaufmann, W., and Offe, H. A. (1960). Naturwissenschaften 47,469. Bauer, K., Kaufmann, W., and Ludwig, S. A. (1971). Hoppe Seyler’s 2. Physiol. Chem. 352, 1723. Baumann, F., Brunner, R., and Rohr, M. (1971). Hoppe Seyler’s 2. Physiol. Chem. 352, 853. Behal, J., and Carter, R. T. (1971). Can. J . Microbiol. 17, 39. Behrens, 0. K., Corse, J., Edwards, J. P., Garrison, L., Jones, R. G., Soper, Q. F., Vanabeele, F. R., and Whitehead, C. W. (1948). I. Biol. Chem. 175, 793. Beyaert, G., and Voets, J. P. (1970). Meded. Rijksfac. Landbouwwetensch., Gent 35, 1091. Birnbaum, S. M., Levintow, L., Kingsley, R. B., and Greenstein, J. (1952). J. Biol. Chem. 194, 455. Birner, J. (1970). J. Pharm. S c i . 59,757. Bobrowski, M., and Borowski, E. ( 1971). J. Gen. Microbiol. 68, 263. Bodansky, M., and Perlman, D. (1969). Science 163, 352. Bodey, G. P., and Deerhake, B. (1971). AppZ. MicrobioZ. 21, 61. Bodey, G. P., and Stewart, D. (1969). Appl. Microbiol. 18, 76. Bodey, G. P., and Stewart, D. (1971). Appl. Microbiol. 21, 710. Bodey, G. P., and Terrell, L. M. (1968). J. Bacteriol. 95, 1587. Bomstein, J., and Evans, W. G. (1965). Anal. Chem. 37, 578. Bondareva, N. S., Levitov, M. M., and Goryachenkova, E. V. (1969a). Biokhimiya 34, 96. Bondareva, N. S., Levitov, M. M., and Rabinovich, M. S. (1969b). Biokhimiya 34, 478. Boxer, G. E., and Everett, P. M. (1949). Anal. Chem. 21, 670. Brady, B. L. (1969). J. Gen. Microbiol. 59, 47. Brammar, W. J., and Clarke, P. H. (1964). J . Gen. Microbiol. 37, 307. Brandl, E. ( 1965). Hoppe Seyler’s Z. Physiol. Chem. 342, 86. Brandl, E. (1972). Sci. Pharm. 40, 1. Brunner, R., and Rohr, M. (1965). Abstr., 2nd Meet., Fed. Eur. Biochem. SOC. 99, 68. Brunner, R., Rohr, M., and Baumann, F. (1966a). Allg. Prukt. Chem. 17,66.
MICROBIAL PENICILLIN ACYLASES
363
Brunner, R., Rohr, M., and Zinner, M. (1966b). Monutsh. Chem. 97,952. Brunner, R., Rohr, M., and Zinner, M. (1968). Hoppe Seyler’s Z. Physiol. Chem. 349, 95. Bundgaard, H., and Ilver, K. ( 1972). J . Pharm. Phurmacol. 24,790. Caltrider, P., and Gottlieb, D. (1963). Phytopathology 53, 1021. Carrington, T. R. (1971). Proc. Roy. SOC., Ser. B 179, 321. Casas-Campillo, C., and Bautista, M. (1965). Appl. Microbiol. 13, 977. Chain, E. (1971). Proc. Roy. SOC., Ser. B 179,293. Chain, E., Florey, H. W., Gardner, A. D., Heatley, N. G., Jennings, M. A., Orr-Ewing, J., and Saunders, A. G. (1940). Lancet 2, 226. Chapman, M. J, Holt, R. J., Mattocks, A. R., and Stewart, G. T. (1964). J. Gen. Microbiol. 36, 215. Chauvette, R. R., Flynn, E. H., Jackson, B. G., Lavaglino, E. R., Morin, R. B., Mueller, B. A., Pioch, R. P., Roeske, R. W., Ryan, C. W., Spencer, J. L., and Van Heyningen, E. (1962). J. Amer. Chem. SOC. 84, 3401. Chauvette, R. R., Hayes, H. B., Huff, G. L., and Pennington, A. P. (1972). J. Antibiot. 25, 248. Chiang, D., and Bennett, R. E. (1967). J . Bacterial. 93,302. Citri, N., and Pollock, M. R. (1966). Advan. Enzymol. Relut. Subj. Biochem. 28, 237. Claridge, C. A., Gourevitch, A., and Lein, J. ( 1960). Nuture (London) 187, 337. Claridge, C. A., Luttinger, J. R., and Lein, J. (1963). Proc. SOC. Erp. Biol. Med. 113, 1008. Clarke, P. H. ( 1970). Aduan. Microbial: Physiol. 4, 179. Cole, M. ( 1964). Nature (London) 203, 519. Cole, M. (1966). Appl. Microbiol. 14, 98. Cole, M. ( 1967). Process Biochem. 2, 35. Cole, M. (1969a). Biochem. I. 115, 733. Cole, M. (1969b). Biochem. J . 115,741. Cole, M. ( 1 9 6 9 ~ )B. i o c h a . J. 115,747. Cole, M. (1969d). Biochem. J. 115, 757. Cole, M., and Rolinson, G. N. (1961). Proc. Roy. Soc., Ser. B 154, 490. Cole, M., and Sutherland, R. S. (1966). 1. Gen. Microbiol. 42, 354. Cole, M., Elson, S., and Fullbrook, P. D. (1972). Biochem. 1. 127, 295. Cole, M., Kenig, M. D., and Hewitt, V. A. (1973). Antimicrob. Ag. Chemother. 3, 463. Cooney, C. L., and Acevedo, J. (1972). Proc. Int. Ferment. Symp., 4th, 1972 p. 247. Cooper, R. D. G. (1972). J. Amer. Chem. SOC. 94, 1018. Cooper, R. D. G., Hatfield, L. D., and Spry, D. 0. (1973). Accounts Chem. Res. 6, 32. Datta, N., and Kontomichalou, N. (1965). Nuture (London) 208, 239. Demain, A. L. (1966). In “Biosynthesis of Antibiotics” (J . F. Snell, ed.), Vol. 1, p. 20. Academic Press, New York. Demain, A. L., Walton, B., Newkirk, J. F., and Miller, I. M. (1963). Nature (London) 199, 909. Dennen, D. W. (1967). J . Phurm. Sci. 56, 1273. Dennen, D. W., and Davis, W. W. (1961). Antimicrob. Ag. Chemother. p. 531. Dennen, D. W., Allen, C. C., and Carver, D. D. (1971). Appl. MicrobioZ. 21, 907. Devcic, B., and Divjak, S . (1968). Acta Biol. Jugoslav. Mikrobiol. 5, 189. Dewdney, J. M., Smith, H., and Wheeler, A. W. (1971). Zmmunology 21, 517.
364
E. J . VANDAMME AND J . P. VOETS
Dinelli, D. ( 1972). Process Biochem. 7, 9. Doyle, F. P., Nayler, J. H. C., Smith, H., and Stove, E. R. (1961a). Nature (London) 191, 1091. Doyle, F. P., Long, A. A. W., Nayler, J. H. C., and Stove, E. R. (1961b). Nature (London) 191, 1183. Draper, P. (1967). J. Gen. Microbiol. 46, 111. Dulong de Rosnay, C., Castagnou, F., and Latrille, J. (1970). Ann. Inst. Pasteur, Paris 118, 277. Ellis, S., and Perry, M. (1966). J. Biol. Chem. 241, 3679. Erickson, R. C., and Bennett, R. E. (1961). Bacteriol. Proc. p. 65. Erickson, R. C., and Bennett, R. E. (1965). Appl. Microbiol. 13, 738. Erickson, R. C., and Dean, L. D. (1966). Appl. Microbiol. 14, 1047. Evrard, E., Claesen, M., and Vanderhaeghe, H. (1964). Nature (London) 201, 1124. Faith, W. F., Neubeck, C. E., and Reese, E. T. (1971). Aduan. Biochem. Eng. 1, 78. Findlater, J. D., and Orsi, B. A. (1973). Fed. Eur. Biochem. Soc. Symp. Abstracts, p. 181. Finholt, P., Jurgensen, G., and Kristiansen, H. ( 1965). 1. Pharm. Sci. 54, 387. Fleming, A. (1932). J. Pathol. Bocteriol. 35, 831. Florey, M. E., and Florey, H. W. (1943). Lancet 1, 387. Fooks, H. R.,and Mattok, G:L. (1969). J. Pharm. Sci. 58, 1357. Ford, J. H. (1947). A d . Chem. 19, 1004. Fosker, G. R., Hardy, K. D., Nayler, J. H. C., Seggery, P., and Stove, E. R. (1971). 1. Chem. Soc., C 10,1917. Fry, K. T., and Lamborg, M. R. (1967). J. Mol. Biol. 28,423. Fullbrook, P. D., Elson, S. W., and Slocombe, B. (1970). Nuture (London) 226, 1054. Fuska, J, Jakubova, A., Halama, D., and Bucko, M. (1972). Antibiotiki (Moscow) 17, 775. Garber, N., and Friedman, J. (1970). J. Gen. Microbiol. 64, 343. Gatenbeck, S., and Brunsberg, U. (1968). Acta Chem.’Scand. 22, 1059. Gauze, G. F., Chugasova, V. A., Laiko, A. V., and Terekhova, L. P. (1972). Antibiotiki (Moscow) 17, 963. Gehrig, R. F., and Knight, S. G. ( 1958). Nature (London) 182,1237. Godtfredsen, W. O., von Daehne, W., and Vandegal, S. (1967). Experientia 23, 280. Golub, E. I., Garaev, M. M., and Romanova, N. B. (1973). Antibiotiki (Moscow) 18, 882. Gotovtseva, V. A, and Levitov, M. M. (1965). Mikrobiologiya 34, 857. Gotovtseva, V. A., Yudina, 0. D., and Levitov, M. M. (1965). Mikrobiologiya, 34, 180. Gotovtseva, V. A., Gorchenko, 6. L., and Suvorov, N. N. (1968). Mikrobiologiya 37, 683. Grant, N. H., Clark, D. E., and Alburn, H. E. (1962). J. Amer. Chem. Soc. 84, 867. Griffith, R. S., and Black, H. R. (1964). J. Amer. Med. Ass. 189, 823. Hafez-Zedan, H., and Plourde, R. ( 1971). Appl. Microbiol. 21, 815. Hamilton-Miller, J. M. T. (1966). Bacteriol. Reu. 30, 761. Hamilton-Miller, J. M. T. (1968). FEBS Lett. 1, 86. Hamilton-Miller, J. M. T. (1971a). J. Med. Microbiol. 4, 227.
MICROBIAL PENICILLIN ACYLASES
365
Hamilton-Miller, J. M. T. (1971b). J. Pheor. Biol. 31, 171. Hamilton-Miller, J. M. T., Smith, J. T., and Knox, R. (1963). J. Phurm. Pharmucol. 15, 81. Hamilton-Miller, J. M. T., Newton, G. G. F., and Abraham, E. P. ( 1970). Biochem. J. 116, 371. Haupt, I., and Thrum, H. ( 1967). 2. Allg. Mikrobiol. 7, 343. Heusler, K. ( 1972). Helu. Chim. Acta 85, 388. Higgens, C. E., avd Kastner, R. E. (1971). Int. 1. Syst. Bacteriol. 21, 32. Highton, P. J., and Hobbs, D. G. (1971). 1. Bacteriol. 106, 646. Hishta, C., Mays, D. L., and Garofalo, M. ( 1971). Anal. Chem. 43, 1530. Ho, L. Y., and Humphrey, A. E. (1970). Biotechnol. Bioeng. 12, 291. Holm, K. A. (1972). Anal. Chem. 44,795. Holt, R. J., and Stewart, G. T. (1964a). Nature (London) 201, 824. Holt, R. J., and Stewart, G. T. (1964b). 1. Gen. Microbiol. 36, 203. Hou, J. P., and Poole, J. W. (1969). J. Pharm. Sci. 4, 447. Hou, J. P., and Poole, J. W. (1971). J. Pharm. Sci. 60, 503. Huang, H. T., English, A. R., Seto, T. A, Shull, G. M., and Sobin, B. A. (1960). 1. Amer. Chem. SOC. 82, 3790. Huang, H. T., Seto, T. A., and Shull, G. M. (1963). Appl. Microbiol. 11, 1. Hull, D. A., Shepp, J. M., Weaver, W. S., Reschke, R. F., and Bomstein, J. ( 1964). Anal. Chem. 36, 599. Hynes, M. J., and Pateman, J. A. (1970). J. Gen. Microbiol. 63, 317. Imada, A, Igarasi, S., Nakahama, K., and Isono, M. (173). J. Gen. Microbiol. 76, 85. Ivashkiv, E. (1964). Anal. Chem. 36, 2506. Jack, G. W., and Richmond, M. H. (1970a). J. Gen. Microbiol. 61, 43. Jack, G. W., and Richmond, M. H. (1970b). FEBS Lett. 12, 30. Jack, G. W., Sykes, R. B., and Richmond, M. H. (1970). Postgrad. Med. J., Suppl. 46, 41. Jarvis, B., and Berridge, N. J. (1969). Chem. Ind. (London) p. 1721. Johnson, A. D., and Hardcastle, G. A. (1961). J. Amer. Chem. SOC. 83, 3534. Johnson, D. E., Nelson, G. E. N., and Ciegler, A. (1968). Appl. Microbiol. 16, 1678. Kameda, Y., Kimura, Y., Toyoura, E., and Omori, T. ( 1961). Nature (London) 191, 1122. Kato, K. (1953). J. Antibiot. 23, 212. Kats, L. N., Moroz, A. F., and Vaintraub, E. A. (1968a). Mikrobiologiya 37, 78. Kats, L. N., Vaintraub, E. A., and Moroz, A. F. (1968b). Mikrobiologiya 37, 263. Katz, E. ( 1971). Pure Appl. Chem. 28,551. Kaufmann, W. (1964). Biochem. Biophys. Res. Commun. 14, 458. Kaufmann, W., and Bauer, K. (1960). Naturwissenschaften 47, 474. Kaufmann, W., and Bauer, K. (1964a). J. Gen. Microbiol. 35, IV. Kaufmann, W., and Bauer, K. ( 1964b). Nature (London) 203,520. Kelly, M., and Kornberg, H. L. (1964). Biochem. J. 93, 557. Kinget, R. D., and Schwartz, M. A. (1969). J. Pharm. Sci. 58, 1102. Klainer, A. S., and Perkins, R. L. (1970). J. Infec. Dis. 122, 323. Kleiner, G. I., and Lopatnev, S. V. ( 1972). Prikl. Biokhim. Mikrobiol. 8, 554. Korchagin, V. B., Serova, L. I., and Kotova, N. I. (1971). Antibiotiki (Moscow) 16, 702. Kulhanek, M., and Tadra, M. (1963). Folia Microbiol. (Prague) 8, 301.
366
E. J . VANDAMME AND J . P. VOETS
Kulhanek, M., and Tadra, M. (1968). Folia Microbiol. (Prague) 4, 340. Kuwabara, S. ( 1970). Biochem. J. 118,457. Lanzilotti, R. L., and Pramer, D. (1970). Appl. Microbiol. 19, 307. Lawrence, R. C. (1966). J. Gen. Microbiol. 44, 393. Lawrence, R. C. (1967). J. Gen. Microbiol. 46, 65. Lawrence, P. J., and Strominger, J. L. ( 1970). J. Biol. Chem. 245, 3653. Lemke, P. A., and Brannon, D. R. (1972). In “Cephalosporin and Penicillin Compounds: Their Chemistry and Biology” (H. Flynn, Ed.), p. 370. Academic Press, New York. Lemke, P. A, and Nash, C. H. (1972). Can. J. Microbiol. 18, 155. Lepidi, A. A., and Nuti, M. P. (1971). Mycopdhol. Mycol. Appl. 43, 11. Lepidi, A. A., Nuti, M. P., and De Bertoldi, M. (1970). Ann. Microbiol. 21, 125. Levitov, M. M., Klapovskaya, K. I., and Kleiner, G. I. (1967). Mikrobiologiya 36, 768. Libinson, G. S . (1971). Khim. Farm. Zh. 5, 37. Lilly, M. D., Balasingham, K., Warburton, D., and Dunnill, P. (1972). Proc. Int. Ferment. Symp., 4th, 1972 p. 379. Lindsay, R. E., and Hem, S. L. (1972). J. Pharm. Sci. 61, 202. Liu, F. S., and Zajic, J. E. (1972). Can. J. Microbiol. 18, 1953. Loder, B., Newton, G. G . F., and Abraham, E. P. (1961). Biochem. J. 79,408. Lucente, G., Romeo, A., and Rossi, D. (1965). Experientia 21, 317. Lugay, J. C., and Aronson, J. N. (1969). Biochim. Biophys. Acta 191, 397. Lund, F., and Tybring, L. (1972). Nature (London), New Biol. 236, 135. McGilveray, I. J., and Strickland, R. D. (1967). J. Pharm. Sci. 56,77. Maier, W., and Groger, D. (1972). Pharmzie 27, 491. Malik, P. S., Ledford, R. A., and Parish, M. L. (1970). J. Dairy Sci. 53, 25. Marconi, W., Cecere, F., Morisi, F., Della Penna, G., and Rappuoli, B. (1973). J . Antibiot. 26, 228. Miller, I. M., Stapley, E. O., and Chaiet, L. ( 1962). Bacteriol. Proc. p. 32. Morin, R. B., Jackson, B. G., Flynn, E. H., and Roeske, R. W. (1962). J . Amer. Chem. Soc. 84, 3400. Morrin, T. G., and Malamy, M. H. (1970). J. Bacteriol. 103, 81. Moss, M. O., and Cole, M. (1964). Biochem. J . 92, 643. Moyer, A. J., and Coghill, R. D. (1946). J. Bacteriol. 51, 57. Moyer, A. J., and Coghill, R. D. (1947). J. Bacteriol. 53, 329. Murao, S. (1955). J. Agr. Chem. Soc. Jap. 29, 404. Nagarajan, R., Boeck, L. D., Gorman, M., Hamill, R. L., Higgens, C. E., Hoehn, M. N., Stark, W. M., and Whitney, V. G. (1971). J. Amer. Chem. Soc. 93, 2308. Nara, T., Misawa, M., Okachi, R., and Yamamoto, M. (1971a). Agr. Biol. Chem. 35, 1676. Nara, T., Okachi, R., and Misawa, H. (1971b). J. Antibiot. 24, 321. Nara, T., Okachi, R., and Kato, F. (1972). Abstr., Int. Ferment. Symp., 4th, 1972 p. 207. Nathorst-Westfelt, L., Schatz, B., and Thelin, H. (1963). Acta Chem. Scand. 17, 1164. Nayler, J. H. C., Long, A. P. W., Brown, D. M., Acred, P., Rolinson, G. N., Batchelor, F. R., Stevens, S., and Sutherland, R. (1962). Nature (London) 195, 1264. Neu, H. C., and Winshell, E. B. (1971). Appl. Microbiol. 21, 66. Niedermayer, A. 0. (1964). Anal. Chem. 36, 938.
MICROBIAL PENICILLIN ACYLASES
367
Novak, L., and Weichet, J. (1965). Erperientiu 21, 360. Novick, R. P. (1969). Bacteriol. Reu. 33, 210. Nussbaumer, P. A. (1962). Pharm. Actu Helv. 37, 65. Nyiri, L. (1967). Nature (London)214, 1347. Nys, P. S., Savitskaya, E. M., and Kolygina, T. S. (1972). Antibiotiki (Moscow) 17. 313. Nys,.P. S., Savitskaya, E. M., and Kolygina, T. S. (1973). Antibiotiki (Moscow) 18, 270. O’Callaghan, C. H., and Morris, A. (1972). Antimicrob. Ag. Chemother. 2, 442. Okachi, R., Misawa, M., and Nara, T. (1972a). Agr. Biol. Chem. 36, 925. Okachi, R., Misawa, M., Deguchi, T., and Nara, T. (1972b). Agr. Biol. Chem. 36, 1193. Okachi, R., Kawamoto, I., Yamamoto, M., Takasawa, S., and Nara, T. (1973a). Agr. Biol. Chem. 37, 335. Okaihi, R., Kato, F., Miyamura, Y., and Nara, T. (197313). Agr. Biol. Chem. 37, 1953. Oostendorp, J. G. (1972). Antonie uan Leeuwenhoek, J . Microbiol. Serol. 38, 201. Ottens, H., De Haan, E., and Sengers, C. H. J. ( 1971). Int. Arch. Allergy Appl. Immunol. 41, 575. Pal, P. R. (1959). J. Biol. Chem. 234, 618. Pan, S. C. (1954).Anal. Chem. 26, 1438. Pan, S. C. (1973). J. Chromatogr. 79, 251. Perrett, C.J. ( 1954). Nature (London) 174,1012. Peterson, R. E., and Ciegler, A. (1969a). Appl. Microbiol. 17, 929. Peterson, R. E., and Ciegler, A. (1969b). Appl. Miorobiol. 18, 64. Pollock, M. R. (1971). Proc. Roy. SOC. Ser. B 179, 385. Price, K. E. (1969).Aduan. Appl. Microbiol. 11, 17. Pruess, D. L., and Johnson, M. J. (1965). J . Bacteriol. 90, 380. Pruess, D. L., and Johnson, M. J. (1967). J. Bacteriol. 94, 1502. Rapson, H. D. C., and Bird, A. E. ( 1983). J. Pharm. Phurmucol. 25, 222. Richmond, M. H., and Sykes, R. B. (1973). Aduan. Microbial PhysioZ. 9, 31. Rolinson, G. N., and Stevens, S. (1961). Proc. Roy. Sac., Ser. B 145, 509. Rolinson, G. N., Batchelor, F. R., Butterworth, D., Cameron-Wood, J., Cole, M., Eustace, G. C., Hart, V., Richards, M., and Chain, E. B. (1960). Nature (London) 187, 236. Rossi, D., Romeo, A, Lucente, G., and Tinti, 0. (1973). Farmaco, Ed. Sci. 28, 262. Rozansky, R., Biano, S., Clejan, L., Frenkel, N., Bogokovsky, B., and Altmann, G. (1969). IN. J. Med. Sci. 5, 297. Ryan, K. J., Needham, G. M., Dunsmoor, C. L., and Sherris, J. C. (1970). Appl. Microbiol. 20, 447. Ryu, D. Y., Bruno, C. F., and Lee, B. K. (1972a). Abstr., Int. Ferment. Symp., 4th, 1972 p. 53. Ryu, D. Y., Bruno, C. F., Lee, B. K., and Venkatasubramanian, K. (1972b). Proc. Int. Ferment. Symp., 4th, 1972 p. 307. Sakaguchi, K., and Murao, S. (1950). J . Agr. Chem. SOC. Jap. 23, 311. Sargent, M. G. (1968). I. Baderiol. 95, 1493. Sassiver, M. L., and Lewis, A. (1970). Aduan. Appl. Microbiol. 13, 163. Savitskaya, E. M., Nys, P. S., Kolygina, T. S., Kleiner, G. I., and Elizarovskaya, L. M. ( 1972). Antibiotiki (Moscow) 4 , 308. Saz, A. K. (1970). J . Cell. Physiol. 76, 397.
368
E. J. VANDAMME AND J. P. VOETS
Schwartz, M. A. (1965a).1. Pharm. Sci. 54,472. Schwartz, M. A. (1965b). 1. Pham. Sci. 54, 1308. Schwartz, M. A, and Buckwalter, F. M. (1962). J. Phurm. Sci. 51, 119. Sebek, 0. K., and Perlman, D. ( 1971). Aduan. Appl. Microbiol. 14, 123. Segelman, A. B., and Farnsworth, N. R. ( 1970). J. Phurm. Sci. 59,726. Self, D. A., Kay, G., Lilly, M. D., and Dunnill, P. (1969). Biotechnol. Bioeng. 11, 337. Serova, L. I., Korchagin, V. B., and Kotova, N. I. (1973). Antibiotiki (Moscow) 18, 687. Shaikh, K., Talati, P. G., and Gang, D. M. (1973). Antimicrob. Ag. Chemother. 3, 194. Shaltiel, S., Mizrahi, R., Stupp, Y., and Sela, M. (1970). Eur. J . Biochem. 14, 509. Shaltiel, S., Mizrahi, R., and Sela, M. (1971). Proc. Roy. Soc., Ser. B 179, 411. Sheehan, J. C., and Henery-Logan, K. R. (1959). J . Amer. Chem. Soc. 81,5838. Shimi, I. R., and Imam, G. M. (1966). Biochem. J. 101,831. Sikyta, B., and Slezak, J. (1964). Biotechnol. Bioeng. 6, 309. Silver, R. P., and Falkow, S. (1970). J. Bacterial. 104, 331. Singh, K., Seghal, S. N., and Vezina, C. (1968). Appl. Microbiol. 16, 393. Singh, K., Seghal, S. N., and Vezina, D. (1969). AppZ. Mimobiol. 17, 643. Sjoberg, G., Nathorst-Westfelt, L., and Ortengren, B. (1967). Acta Chem. Scand. 21, 547. Smiley, K. L., and Strandberg, G. W. (1972): Aduan. Appl. Microbiol. 15, 13. Smith, H., and Marshall, A. C. (1971). Nature (London), New Biol. 232, 45. Spencer, B. (1968). Biochem. Biophys. Res. Commun. 31, 170. Spencer, B., and Burnham, M. (1973). Fed. Eur. Biochem. Soc. Symp. Abstracts, p. 186. Spencer, B., and Maung, G. (1970). Biochem. J . 118,29. Stapley, E. D., Jackson, M., Hernandez, S., Zimmerman, S. B., Currie, S. A, Mochales, S., Mata, J. M., Woodruff, H. B., and Hendlin, D. (1972). Antimicrob. Ag. Chemother. 2, 122. Strominger, J. L., Blumberg, P. L., Suginaka, H., Umbreit, J., and Wickus, G. G. ( 1971). Proc. Roy. Soc., Ser. B 179, 369. Sutherland, R. ( 1964). J . Gen. Microbiol. 34, 85. Sykes, R. B., and Richmond, M. H. (1970). Nature (London) 226, 952. Szentirmai, A. ( 1964). Appl. Microbiol. 12, 185. Takahashi, T., Yamazaki, Y., Kato, K., and Isono, M. (1972). J . Amer. Chem. Soc. 94, 4035. Takasawa, S., Okachi, R., Kawamoto, I., Yamamoto, M., and Nara, T. (1972). Agr. Biol. Chem. 36, 1701. Thadhani, S. B., Borkar, P. S., and Ramachandran, S. (1972). Biochem. J . 128, 49. Thomas, A. H., and Broadbridge, R. A. (1970). Anulyst 95, 459. Thomas, R. (1961). Nature (London) 191, 1161. Tiller, A., and Brunneman, H. (1970). Arch. H y g . Bakteriol. 154, 148. Tipper, D. J., and Strominger, J. L. (1968). J. Biol. Chem. 243, 3169. Uri, J., and Sztaricskai, F. ( 1961). Nature (London) 191, 1223. Uri, J., Szathmary, S., and Herpay, 2. (1957). Nature (London) 179, 1029. Uri, J., Valu, G., and Bekesi, 1. (1963). Nature (London) 200, 896. Uri, J., Valu, G., and Bekesi, I. (1984). Naturwissenschajten 51, 298. Vandamme, E. J. (1972). Doctoral Thesis, University of Ghent, Belgium.
MICROBIAL PENICILLIN ACYLASES
389
Vandamme, E. J. (1973). Fed. Eur. Bwchem. SOC. Symp. Abstracts, p. 53. Vandamme, E. J., and Voets, J. P. ( 1971). 2. Allg. Mikrobiol. 11, 155. Vandamme, E. J., and Voets, J. P. (1972a). 1. Chromatogr. 71, 141. Vandamme, E. J., and Voets, J. P. (1972b). Meded. Rijksfac. Landbouwwetensch., Gent 37, 1185. Vandamme, E. J., and Voets, J. P. (1973). 2. Allg. Mikrobiol. 13, 701. Vandamme, E. J., Voets, J. P., and Beyaert, G. (1971a). Meded. Rijksfac. Landbouwwetensch., Gent 36, 577. Vandamme, E. J., Voets, J. P., and Dhaese, A. (1971b). Ann. Inst. Pasteur, Paris 121, 435. Vanderhaeghe, H., Claesen, M., Vlietinck, A., and Parmentier, G. (1968). Appl. Microbiol. 16, 1557. Van Etten, J. L., Koski, R. K., and El-Olemy, M. M. (1969). J. Bacteriol. 100, 1182. Van Heijenoort, Y., and Van H,eijenoort,K. ( 1971). FEBS Lett. 15, 137. Vasilescu, I., Vociu, M., Voinescu, R., Birladeanu, R., Sasarman, E., and Rafirotu, I. ( 1969). I n “Antibiotics” ( M . Herold and Z. Gabriel, eds.), Butterworth, London. p. 518. Vezina, C., Singh, K., and Seghal, S . N. (1965).Mycologia 57, 722. Vezina, C., Seghal, S. N., and Singh, K. (1968). Advan. Appl. Microbiol. 10, 221. Voets, J. P., and Vandamme, E. J. (1972). Abstr., Int. Ferment. Symp., 4th, 1972 p. 246. Wade, H. E., Robinson, H. K., and Phillips, B. W. ( 1971). J . Gen. Microbiol. 69, 299. Wagner, E. S., and Gorman, M. ( 1971). J . Antibiot. 24, 647. Waldschmidt-Leitz, E., and Bretzel, G. (1964). Hoppe Seyler’s Z . Physiol. Chem. 337, 222. Walton, R. B. (1964). Science 143, 1138. Warburton, D.,Balasingham, K., Dunnill, P., and Lilly, M. D. (1972). Biochim. Biophys. Acta 284, 278. Warburton, D., Dunnill, P., and Lilly, M. D. (1973). Biotechnol. Bioeng. 25, 13. Weissenburger, H. W. O., and Vanderhoeven, M. G. (1970). Rec. Trav. Chim. Pays-Bas. 89, 1081. Wick, W. E. (1967). Appl. Microbiol. 15, 765. Wingard, L. E. (1972). Advan. Biochem. Eng. 2, 1.
This Page Intentionally Left Blank
SUBJECT INDEX L-2-amino-4-methoxv-trans-3butenoic acid, 26 N-Acetylglucosamine, reversal of inhibi6-Aminopenicillanic acid tion of analyses for, 322 anticapsin, 32 chemical and biological properties, 320 bacillin, 33 chemical process for production from bacilysin, 33 penicillin, 358-359 fumarylcarboxamide-~-2,3,-diaminopiproduction process, history, 3 1 4 4 1 5 melic acid-~-alanine,37 Ampicillin acylase, 329 tunicamycin, 50 Antimetabolite L-Nz-Acetylornithine, reversal of inhibidefinition, 19-20 tion of detection of, 21-22 L-N'-( 1-iminoethyl)omithine, 34 mechanisms of action and reversal, Actinomycins, 25 22-24 Adenine, reversal of inhibition of screening for, 20-21 alanosine, 29 Applied microbiology training aristeromycin, 41 biochemistry in, 6 azaserine, 34 in developing countries, 11-12 hadacidin, 30 in the United States, 1, 3-5 pentalenolactone, 50 job possibilities, 7 thioguanine, 43 Aranciamycin, 24 Adenosine, reversal of inhibition of L-Arginine, reversal of inhibition of aristeomycin, 41 L-2-amino-4- ( 2-aminoethoxy f-transcordycepin, 42 3-butenoic acid, 28 formycin, 43 ~-N'-hydroxyarginine, 53 toyocamycin, 44 L-N'-( 1-iminoethyl)ornithine, 29 tubercidin, 44 0-( 1-norvalyl-5 ) -isourea, 53 kdenylic acid, reversal of inhibition of D-Aspartic acid, reversal of inhibition of griseofulvin, 49 L-2-amino-4-methoxy-trans-3butenoic D-Alanine, reversal of inhibition of acid, 26 L-2-amino-4-methoxy-trans-3butenoic L-Aspartic acid, reversal of inhibition of acid, 26 alanosine, 29 O-carbamyl-~-serine,34 2,3-diaminosuccinic acid, 30 cycloserine, 33 hadacidin, 30 L-Alanine, reversal of inhibition of L-2-amino-4-methoxy-trans-3butenoic B acid, 26 0-carbamyh-serine, 34 Biotin, reversal of inhibition of cycloserine, 33 actithiazic acid, 44 American Type Culture Collection a-dehydrobiotin, 45 1931 expansion, 299-302 a-methylbiotin, 45 founding and early years, 295-299 a-methyldethiobiotin, 46 present organization, 303-309 stravidin, 37 D-2-Aminobutanoic acid, reversal of inBrewing hibition of comparison of systems, 247-248 A
371
372
SUBJECT INDEX
continuous processes, 244-247 culture selection for, 249-250 new developments, 241-244 traditional methods, 239-241 C
L-Citrulline, reversal of inhibition of ~-N~-hydroxyarginine, 53 L-N’-( I-iminoethyl )ornithine, 29 0-(1-norvalyl-5 )-isourea, 53 Cystathionine, reversal of inhibition of ~-2-amino-4-methoxy-trans-3-butenoic acid, 26 L-Cysteine, reversal of inhibition of L-2-amino-4-( 2-aminoethoxy )-trans3-butenoic acid, 28 tubercidin, 44 Cytidine, reversal of inhibition of 5-azacytidine, 38 D
Deoxyadenosine, reversal of inhibition of tubercidin, 44 Deoxycytidine, reversal of inhibition of minimycin ( oxazinomycin), 39 Deoxyguanylic acid, reversal of inhibition of
mycophenolic acid, 49 Diaminopimelic acid, reversal of inhibition of L-2-amino-3-( 2-aminoethoxy ) propionic acid, 28 E
Ergocalciferol, reversal of inhibition of cerulenin, 46 F
Fatty acids in bacteria, 64-66, 75 distribution in organisms, 97 G
D-GlUtamiC acid, reversal of inhibition of ~-2-amino-4-methoxy-trans-3-butenoic acid, 26 Glutamine, reversal of inhibition of alazopeptin, 34 anticapsin, 32 azaserine, 34 DON, 34 duazomycin A, 35 duazomycin B, 35 duazomycin C , 34 L-( N’-phosphono)methionine-( 5 ) sulfoximinyl-~-alanyl-~-alanine by, 37 Glycolipids in bacteria, 69, 81 distribution, 97 Guanine, reversal of inhibition of decoyinine (angustmycin A ) , 41 psicofuranine (angustmycin C ) , 43 thioguanine, 43 Guanosine, reversal of inhibition of cordycepin, 42 decoyinine (angustmycin A ) , 41 mycophenolic acid, 49 psicofuranine ( angustmycin C ), 43 Guanylic acid, reversal of inhibition of griseofulvin, 49 mycophenolic acid, 49 H
L-Histidine, reversal of inhibition of aS,5S-a-amino-3-chloro-2-isoxazoline5-acetic acid, 53 L-arginyl-D-abthreon yl-L-phenylalanine, 54 Homocysteine, reversal of inhibition of ~-2-amino-4-methoxy-trans-3-butenoic acid, 26 rhizobitoxine, 29 L-Homoserine, reversal of inhibition of borrelidin, 48 2-( l-cyclohexen-3( R)-y1)-S-glycine, 31
Glucosamine, reversal of inhibition of anticapsin, 32 1 bacilysin, 33 fumarylcarboxamide-~-2,3-diaminopro- Inosine, reversal of inhibition of pionyl-~-alanine,37 aristomycin, 41
373
SWJECT INDEX
formycin, 43 toyocamycin, 44 L-Isoleucine, reversal of inhibition of ~-2-amino-4-methyl-5-hexenoic acid, 27 2-( 1-cylohexen-3( R ) -yl)-S-glycine, 31
L-Lysine, reversal of inhibition of L-2-amino-3-( 2-aminomethoxy )propionic acid, 28 ~-2-amino-4-(2-aminomethoxy )-trans3-butenoic acid, 28 L-Lysylglycine, reversal of inhibition of L-2-amino-3- ( 2-methoxy )propionic acid, 28
K
a-Ketoglutaric acid, reversal of inhibition of 2( l-cyclohexen-3( R ) -yl)-S-glycine, 31 Koji, 171-175 L
L-Leucine, reversal of inhibition of ~-2-amino-4,4-dichlorobutanoic acid, 25 L-2-amino-3-methylaminopropionic acid, 26 ~-2-amino-4-methyl-5-hexenoic acid, 27 1-amino-2-nitrocyclopentanecarboxylic acid, 31 L-2-amino-4-pentynoic acid, 27 2 ( 1-cyclohexen-3( R ) -yl ) -S-glycine, 31 Linear alkylbenzene sulfonates, degradation analytical methods, 267-269 biodegradation tests, 269-270 effect of environment on degradation anaerobic conditions, 276-277 mixed cultures, 276 pure cultures, 277-281 salt water, 275-276 laboratory vs. field studies, 273-275 structures, 266 structures vs. degradation, 270-273 toxicity to aquatic environment, 281-285 Lipids as taxonomic aid, 63 for Actinomycetales, 92-95, 102-103 analytical procedures, 71-74 for Eubacteriales, 82-91, 99-102 for Pseudomonadales and Hypomicrobiales, 76-80, 98-99 for Spirochaetales and Mycoplasmatales, 96, 103
M
L-Methionine, reversal of inhibition of L-2-amino-4-methoxy-trans-3butenoic acid, 26
2-amino-4-methylphosphinobutano ylL-alanyl-L-alanine, 36 L-ethionine, 27 rhizobitoxine, 29 L-selenomethionine, 28 wildfire toxin, 36 Miso, 160 N
Naphthomycin, 24 Natto, 160 Neutral lipids in bacteria, 70 Nicotinamide, reversal of inhibition of albocycline, 47 melanicidin, 47 streptozotocin, 47 Nicotinic acid, reversal of inhibition of albocycline, 47 melanicidin, 47 0
L-Ornithine, reversal of inhibition of 0( 1-norvalyl-5 ) isourea, 53 P
Penicillin synthesis by acylases, 335, 345-347 types by fermentation, 312-313 Penicillin acylases analyses for, 3 2 2 4 2 3 nomenclature, 325-326 occurrence, 316, 3 2 3 4 3 2 substrate specificity, 3 2 2 3 2 3
374
SUBJECT INDEX
Penicillin G acylases in Actinomycetales, 345 in bacteria, 33-37, 339-344 in molds, 3 3 8 3 4 5 Penicillin V acylases in Actinomycetales, 333-334 in bacteria, 334-345 in fungi, 326,328-332 occurrence, 326-328 in spores, 332 in yeasts, 333 L-Phenylalanine, reversal of inhibition of L-3-( 2,5-dihydrophenyl) alanine, 21, 32 Phospholipids in bacteria, 67-68, 81 distribution, 97 Pilot plant technology biological consideration, 196-213 chemical factors, 218-221 environmental factors, 213 equipment, 2212224 mathematical models, 224-227 physical factors, 214-218 Polyenes, mode of action relationship of sterols, 115-125 Polyenes, resistance to, 126-128 role of sterols in resistance, 128-130 Pyridoxine, reversal of inhibition of bacimethrin, 48 Pyruvate, reversal of inhibition of tubericidin, 44 Q
Quinolinic acid, reversal of inhibition of albocycline, 47 S
L-Serine, reversal of inhibition of L- ( threo ) -2-amino-3,4-dihydroxybutanoic acid, 26 Serylglycylglutamic acid, reversal of inhibition of lycomarasmin, 49 Shoyu, 160 Soybeans, composition, 165-167 Soy sauce chemical composition, 163-164 chemical process for producing,
183-185 history, 161-163 mashing procedure, 175-178 microbiology, 177-182 raw materials for, 165-169 Sterols factors influencing biosynthesis, 112-114 occurrence in bacteria, 110 occurrence in yeasts and fungi, 110-111 physiological role in microorganisms, 114-1 15 Sufu, 160 T
Tempeh, 160 Thiamine, reversal of inhibition of bacimethrin, 47 L-Threonine, reversal of inhibition of L-argininybdo-threonyl-L-phenylalanine, 54 borrelidin, 48 2-( 1-cyclohexen-3( R )-y1)-S-glycine, 31 L- ( threo )-2-amino-3,4-dihydroxybutanoic acid, 26 Thymidine, reversal of inhibition of formycin B, 43 L-Tyrosine, reversal of inhibition of L-3( 2,5-dihydrophenyl)alanine, 32 U
Uracil, reversal of inhibition of emimycin, 38 Uridine, reversal of inhibition of 5-azacytidine, 38 formycin B, 43 pyrazomycin, 40 tubercidin, 44 Uridylic acid, reversal of inhibition of pyrazomycin, 40 V
L-Valine, reversal of inhibition of ~-2-amino-4-methyl-5-hexenoic acid, 27 L-2-amino-4-pentynoic acid, 27
SUBJECT INDEX
2- ( l-cyclohexen-3( R )-yl) -S-glycine, 31
threomycin ( furanomycin), 31 Vitamin B, reversal of inhibition of descobaltocorrins, 48 Y
Yeast baker’s, 250-252 batch process, 252 continuous process, 253-254
375
biochemicals from, 260-262 distiller’s, 254-256 food and fodder, 2 5 6 2 6 0 industrial uses historical uses, 235-238 present uses, 238-239 taxonomy application of computer techniques, 150-157 comparison of numerical and classical taxonomy, 1 4 6 1 5 0 numerical methods, 137-146 tests useful for, 135-137
This Page Intentionally Left Blank
CONTENTS OF PREVIOUS VOLUMES A Commentary on Microbiological As-
Volume 1
saying F . Kauanagh
Protected Fermentation Milo3 Herold and Jan NeEdsek
Of
The Mechanism of Penicillin Biosynthesis Arnold L. Demuin
Richard Ehrlich
Preservation of Foods and Drugs by Ionizing Radiations W . Dexter Bellumy
Microbial Control Brewery Gerhard 1. Hass
The State of Antibiotics in Plant Disease Control David Pramer
Newer Development in Vinegar Manufactures Rudolph J. Allgeier and Frank M . Hildebrandt
Microbial Synthesis of Cobamides D. Perlman
Methods
in
the
The Microbiological Transformation of Steroids T. H. Stoudt
Factors Affecting the Antimicrobial Activity of Phenols E . 0. Bennett Germfree Animal Techniques and Their Applications Arthur W . Phillips and James E. Smith
Biological Transformation of Solar Energy William J. Oswald and Clarence G. Colueke
SYMPOSIUM IN
Insect Microbiology S. R. Dutky The Production of Amino Acids by Fermentation Processes Shukuo Kinoshita Continuous Industrial Fermentations Philip Gerhurdt and M . C. Bartlett The Largescale Growth of Higher Fungi Radcliffe F. Robinson and R. S. Dauidson
ON ENGINEERING ADVANCES FERMENTATTON PRACTICE
Rheological Properties of Fermentation Broths Fred H . Deindoerfer and John M . West Fluid Mixing in Fermentation Processes 1. Y. Oldshue Scale-up of Submerged Fermentations W . H. Bartholemew Air Sterilization Arthur E. Humphrey
AUTHOR INDEX-SUB JECT INDEX
Volume 2
Newer Aspects of Waste Treatment Nandor Porges Aerosol Samplers Harold W . Batchelor '
Sterilization of Media for Biochemical Processes Lloyd L. Kempe Fermentation Kinetics and Model Processes Fred H. Deindoerfer 377
378
CONTENTS OF PREVIOUS VOLUMES
Continuous Fermentation W . D. Maxon Control Applications in Fermentation George J. Fuld AUTHOR INDEX-SUB J E C r INDEX
Volume 3
Preservation of Bacteria by Lyophilization Robert J. Heckly
Sphaerotilus, Its Nature and Economic Significance N o m n C. Dondero Large-Scale Use of Animal Cell Cultures Donald J. Merchant and C. Richard Eidam Protection Against Infection in the Microbiological Laboratory: Devices and Procedures Mark A. Chatigny Oxidation of Aromatic Compounds by Bacteria Martin H . Rogof Screening for and Biological Characterizations of Antitumor Agents Using Microorganisms Frank M. Schabel, Jr., and Robert F. Pittill o The Classification of Actinomycetes in Relation to Their Antibiotic Activity Eli0 Baldacci The Metabolism of Cardiac Lactones by Microorganisms Elwood Titus Intermediary Metabolism and Antibiotic Synthesis J. D. Bu’Lock Methods for the Determination of Organic Acids A. C . H d m e AUTHOR INDEX-SUB JECT INDEX
Volume 4
Induced Mutagenesis in the Selection of Microorganisms S . 1. Alikhunian The Importance of Bacterial Viruses in Industrial Processes, Especially in the Dairy Industry F . 1. Babel Applied Microbiology in Animal Nutrition Harlow H . Hall Biological Aspects of Continuous Cultivation of Microorganisms T . Holme Maintenance and Loss in Tissue Culture of Specific Cell Characteristics Charles C. Morris Submerged Growth of Plant Cells L. G. Nickell AUTHOR INDEX-SUB JECT INDEX
Volume 5
Correlations between Microbiological Morphology and the Chemistry of Biocides Adrien Albert Generation of Electricity by Microbial Action J . B. Davis Microorganisms and the Molecular Biology of Cancer G. F. Game Rapid Microbiological with Radioisotopes Gilbert V. Levin
Determinations
The Present Status of the 2,B-Butylene Glycol Fermentation Sterling K . Long and Roger Patrick Aeration in the Laboratory W . R . Lockhart and R. W . Squires
CONTENTS OF PREVIOUS VOLUMES
Stability and Degeneration of Microbial Cultures on Repeated Transfer Fritz Reusser Microbiology of Paint Films Richard T . Ross
379
Volume 7
Microbial Carotenogenesis Alex Ciegler
The Actinomycetes and Their Antibiotics Selman A. Waksmun
Biodegradation: Problems of Molecular Recalcitrance and Microbial Fallibility M . Alexander
Fuse1 Oil A. Dinsmoor Webb and John L. lngmham
Cold Sterilization Techniques John B. Opfell and Curtis E. Miller
AUTHOR INDEX-SUB J E C r INDEX
Microbial Production of Metal-Organic Compounds and Complexes D. Perhun
Volume 6
Global Impacts of Applied Microbiology: An Appraisal CarLCoran H e d h and Mortimer P. Starr Microbial Processes for Preparation of Radioactive Compounds D. Perlman, Aris P. Bayan, and Nancy A. Giuffre Secondary Factors in Fermentation Processes P. Margalith Nonmedical Uses of Antibiotics Herbert S. Goldberg Microbial Aspects of Water Pollution Control K. Wuhrmann Microbial Formation and Degradation of Minerals Melvin P. Silverman and Henry L. Ehrlich Enzymes and Their Applications Irwin W . Sizer
A Discussion of the Training of Applied Microbiologists B. W . Koft and Wayne W . Umbreit AUTHOR INDEX-SUBJECT
INDEX
Development of Coding Schemes for Microbial Taxonomy S. T . Cowan Effects of Microbes on Germfree Animals Thomas D. Luckey Uses and Products of Yeasts and Yeastlike Fungi Walter J . Nickerson and Robert G. Brown Microbial Amylases Walter W . Windish and Nagesh S . Mhutre The Microbiology of Freeze-Dried Foods Gerald J. Silverman and Samuel A. Goldblith Low-Temperature Microbiology Judith Farrell and A. H . Rose AUTHOR INDEX-SUB JECT INDEX
Volume 8
Industrial Fermentations and Their Relations to Regulatory Mechanisms Arnold L. Demain Genetics in Applied Microbiology S . G. Bradley
380
CONTENTS OF PREVIOUS VOLUMES
Microbial Ecology and Applied Microbiology Thomas D. Brock
Cellulose and Cellulolysis Brigitta Norkrans
The Ecological Approach to the Study of Activated Sludge Wesley 0. Pipes
Microbiological Aspects of the Formation and Degradation of Cellulosic Fibers L. JurSek, J. Ross Coluin, and D. R. Whitaker
Control of Bacteria in 'Nondomestic Water Supplies Cecil W. Chambers and Norman A. Clarke
The Biotransformation of Lignin to Humus-Facts and Postulates R. T. Oglesby, R . F . Christman, and C. H . Driver
The Presence of Human Enteric Viruses in Sewage and Their Removal by Conventional Sewage Treatment Methods Stephen Alan Kollins Oral Microbiology Heiner Hogman
Malo-lactic Fermentation Ralph E. Kunkee AUTHOR INDEX-SUB JECT INDEX
Media and Methods for Isolation and Enumeration of the Enterococci Paul A. Hartmun, George W . Reinbold, and Devi S . Saraswat Crystal-Forming Bacteria Pathogens Martin H . Rogof
Bulking of Activated Sludge Wesley 0. Pipes
as
Insect
Mycotoxins in Feeds and Foods Emnuel Borker, Nino F . Insahta, Colette P. Levi, and John S. Witzeman AUTHOR INDEX-SUB JECT INDEX
Volume 9
The Inclusion of Antimicrobial Agents in Pharmaceutical Products A. D. Russell, June Jenkins, and I. H . Harrison Antiserum Production in Experimental Animals Richard M . Hyde Microbial Models of Tumor Metabolism G. F. Guuse
Volume 10
Detection of Life in Soil on Earth and Other Planets. Introductory Remarks Robert L. Starkey For What Shall We Search? Allan H . Brown Relevance of Soil Microbiology to Search for Life on Other Planets G . Stotzky Experiments and Instrumentation Extraterrestrial Life Detection Gilbert V . Leuin
for
Halophilic Bacteria D. J . Kudhner Applied Significance of Polyvalent Bacteriophages S. G . Bradley Proteins and Enzymes as Taxonomic Tools Edward D.Garber and John W. Rippon
CONTENTS OF PREVIOUS VOLUMES
Mycotoxins Alex Ciegkr and Eivind B. Lillehoi
Ergot Alkaloid Fermentations William J. Kelleher
Transformation of Organic Compounds by Fungal Spores Claude VBzina, S . N. Sehgal, and Kartar Singh
The Microbiology of the Hen's Egg R. G. Board
Microbial Interactions in Continuous Culture Henry R. Bungay, H I and Mary Lou Bungay
381
Training for the Biochemical Industries 1. L. Hepner AUTHOR INDEX-SUB JECT INDEX
Volume 1 2
Chemical Sterilizers ( Chemosterilizers ) Paul M . Borick Antibiotics in the Control of Plant Pathogens M . 1. Thirumalachar
History of the Development of a School of Biochemistry in the Faculty of Technology, University of ManChester Thomas Kennedy Walker
Fermentation Processes Employed Vitamin C Synthesis Milo5 Kulhdnek CUMULATIVEAV~HOR INDEX-CUMULATIVE TITLEINDEX Flavor and Microorganisms P. Margalith and Y. Schwartz AUTHOR INDEX-SUB JECT INDEX
in
Volume 1 1
Successes and Failures in the Search for Antibiotics Selman A. Waksman Structure-Activity Relationships of 'Semisynthetic Penicillins K. E . Price Resistance to Antimicrobial Agents 1. S . Kiser, G. 0. Gale, and G. A. Kemp
Micromonospora Taxonomy George Luedemunn Dental Caries and Periodontal Disease Considered as Infectious Diseases William Gold The Recovery and Purification of Biochemicals Victor H. Edwards
Mechanisms of Thermal Injury in Nonspomlating Bacteria M . C . Allwood and A. D. Russell Collection of Microbial Cells Daniel 1. C. Wang and Anthony I . Sinskey Fermentor Design R. Steel and T. L. Miller The Occurrence, Chemistry, and Toxicology of the Microbial PeptideLactones A. Taylor Microbial Metabolites as Potentially Useful Pharmacologically Active Agents D. Perlman and G. P. Peruzzotti AUTHOR INDEX-SUB JECT INDEX
382
CONTENTS OF PREVIOUS VOLUMES
Volume 13
Chemotaxonomic Relationships Among the Basidiomycetes Robert G. Benedict Proton Magnetic Resonance Spectroscopy -An Aid in Identification and Chemotaxonomy of Yeasts P. A. I . Gorin and 1. F. T. Spencer Large-Scale Cultivation of Mammalian Cells R. C. Telling and P. J . Radlett Large-Scale Bacteriophage K. Sargeant
Production
Microorganisms as Potential Sources of Food Jnanendra K. Bhattachariee Structure Activity Relationships Among Semisynthetic Cephalosporins M . L. Sassioer and Arthur Lewis Structure-Activity Relationships in the Tetracycline Series Robert K. Blackwood and A&ur R. English Microbial Production of Phenazines J. M . Ingram and A. C. Blackwood The Gibberellin Fermentation E . G. Jefferys Metabolism of Acylanilide Herbicides Richard Bartha and David Pramer Therapeutic Dentrifices 1. K. Peterson Some Contributions of the U.S. Department of Agriculture to the Fermentation Industry George E. Ward Microbiological Patents in International Litigation John V . Whittenburg
Industrial Applications of Continuous Culture: Pharmaceutical Products and Other Products and Processes R. C. Righelato and R. Elsworth Mathematical Models for Fermentation Processes A. G. Fredrickson, R. D. Megee, I I l , and H. M . Tsuchiya AUTHOR INDEX-SUBJECT
INDEX
Volume 14
Development of the Fermentation Industries in Great Britain John I . H . Hastings Chemical Composition as a Criterion in the Classification of Actinomycetes H. A. Lecheoalier, Mary P. Lecheoalier, and Nancy N. Gerber Prevalence and Distribution of Antibiotic-Producing Actinomycetes John N. Porter Biochemical Activities of Nocardia R. L. Raymond and V . W . Jamison Microbial Transformations of Antibiotics Oldrich K. Sebek and D. Perlman In
Vivo Evaluation of Antibacterial Chemotherapeutic Substances A. Kathrine Miller
Modification of Lincomycin Barney J. Magerlein Fermentation Equipment G . L. solmnons The Extracellular Accumulation of Metabolic Products by Hydrocarbon-Degrading Microorganisms Bernard J . Abbott an$ William E. Gledhill AUTHOR INDEX-SUB JECT INDEX
CONTENTS OF PREVIOUS VOLUMES
Volume 15
Medical Applications Enzymes Irwin W . Sizer
383
Volume 16
of
Microbial
Immobilized Enzymes K. L. Smiley and G. W. Strandberg Microbial Rennets Joseph L. Sardinas Volatile Aroma Components of Wines and Other Fermented Beverages A. Dinsmoor Webb and Carlos J. Mulbr
Public Health Significance of Feeding Low Levels of Antibiotics to Animals Thomas H . Jukes Intestinal Microbial Flora of the Pig R. Kenworthy Antimycin A, a Piscicidal Antibiotic Robert E. Lennon and Claude Vkzina
Correlative Microbiological Assays Ladislav J. Hahka Insect Tissue Culture W . F . Hink Metabolites from Animal and Plant Cell Culture Irving S . Johnson and George B. Boder Structure-Activity Relationships in Coumerm ycins John C. Godfrey and Kenneth E . Price Chloramphenicol Vedpal S. Malik Microbial Utilization of Methanol Charles L. Cooney and David W . Levine Modeling of Growth Processes with Two Liquid Phases: A Review of Drop Phenomena, Mixing, and Growth P. S. Shah, L. T. Fan, I. C. Kao, and L. E . Erickson
A B C D E F G H
4 5 6 7 8 9 O 1
1 2 J 3
Ochratoxins Kenneth L. Applegate and John R. Chipley Cultivation of Animal Cells in Chemically Defined Media, A Review Kiyoshi Higuchi Genetic and Phenetic Classification of Bacteria R. R. Colwell Mutation and the Production of Secondary Metabolites Arnold L. Demuin Structure-Activity Relationships in the Actinomycins Johannes Meienhofer and Eric Atherton
Microbiology and Fermentations in the Prairie Regional Laboratory of the National Research Council of Canada 1946-1971 R. H . Haskins
Development of Applied Microbiology at the University of Wisconsin William B. Sarles
AUTHOR INDEX-SUBJECT INDEX
AUTHOR INDEX-SUB JECT INDEX
This Page Intentionally Left Blank