Advances in Applied Microbiology, Volume 10

ADVANCES IN Applied Microbiology VOLUME 10 CONTRIBUTORS TO THIS VOLUME Paul M. Borick S. G. Bradley Allan H. Brown...

Author: Wayne W . Umbreit

12 downloads 1313 Views 18MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

ADVANCES IN

Applied Microbiology VOLUME 10

CONTRIBUTORS TO THIS VOLUME

Paul M. Borick S. G. Bradley

Allan H. Brown Henry R. Bungay, 111 Mary Lou Bungay Alex Ciegler Edward D. Garber

D. J . Kushner Gilbert V. Levin Eivind B. Lillehoj John W. Rippon

S. N . Sehgal Kartar Singh Robert L. Starkey

G. Stotzky

M. J . Thirumalachar Claude Vezina

ADVANCES IN

Applied Microbiology Edited by W A Y N E W. UMBREIT Department of Bacteriology Rutgers, The State University New Brunswick, New Jersey

D. PERLMAN School of Pharmacy The University of Wisconsin Madison, Wisconsin

V O L U M E 10

@

1968

ACADEMIC PRESS, New York and London

COPYRIGHT(8

1968,

BY

ACADEMICPRESS,

INC.

ALL RIGHTS RESERVED. NO PART OF THIS BOOK MAY BE REPRODUCED IN ANY FORM, BY PHOTOSTAT, MICROFILM, O R ANY OTHER MEANS, WITHOUT WRITTEN PERMISSION FROM THE PUBLISHERS.

ACADEMIC PRESS, INC. 1 I 1 Fifth Avenue, New York, New York 10003

United Kingdom Edition published by ACADEMIC PRESS, INC. (LONDON) LTD. Berkeley Square House, London W.l

LIBRARY O F CONGRESS CATALOG CARD NUMBER59-13823

PRINTED IN THE UNITED STATES OF AMERICA

LIST OF CONTRIBUTORS Numbers in parentheses indicate the pages on which the authors’ contributions begin.

PAUL M. BORICK,Ethicon, Incorporated, Somerville, New Jersey (291)

S. G. BRADLEY,Department of Microbiology, University of Minnesota, Minneapolis. Minnesota (101) ALLANH. BROWN,Depariment of Biology, University of Pennsylvania, Philadelphia, Pennsylvania ( 5 ) HENRYR. BUNGAY,111, Department of Environmental Systems Engineering, Clemson University, Clemson. South Carolina (269) MARY LOU BUNGAY,Clemson, South Carolina (269) ALEX CIEGLER,Northern Regional Research Laboratory, Peoria, Illinois ( 155) EDWARDD. GARBER,Department of Biology, The University of Chicago, Chicago, Illinois ( I 37)

D. J . KUSHNER,Department of Biology, University of Ottawa, Ottawa, Ontario, Canada (73) GILBERTV. LEVIN,Biopherics Research, Incorporated, Washington, D. C. (55) EIVINDB. LILLEHOJ, Northern Regional Research Laboratory, Peoria, Illinois ( 1 55)

JOHN W. RIPPON,Depariment of Medicine (Dermatology), The University of Chicago, Chicago, Illinois ( I 37) S . N . SEHGAL,Department of Microbiology, Ayerst Laboratories, Montreal, Quebec, Canada (221) KARTARSINGH,Department of Microbiology, Ayerst Laboratories, Monireal, Quebec, Canada (22 1 ) ROBERTL. STARKEY,Departmeni of Biochemistry and Microbiology, Rutgers. The State Universiiy, New Brunswick, New Jersey ( I )

G. STOTZKY, Kitchawan Research Laboratory of the Brooklyn Botanic Garden, Ossining, New York (17)* M. J. THIRUMALACHAR, Antibioiics Research Centre, Hindustan Antibiotics Limited, Plmpri, Poona, India (3 13) CLAUDEVBZINA,Department of Microbiology, Ayrest Laboratories, Montreal, Quebec, Canada (22 I )

*Present address: Department of Biology, New York University, New York, New York. V

This Page Intentionally Left Blank

PREFACE

This volume represents an expansion into areas of applied microbiology other than those solely related to the laboratory. The reviews on the problems involved in determining the occurrence and ecological roles of microorganisms on other planets forecast some of the new activities for microbiologists in the twentieth century. The more practical aspects of harnessing microorganisms to help feed our expanding population are developed by Thirumalachar in the evaluation of the use of antibiotics for the control of food-destroying plant pathogens and by the Bungay’s who are concerned with the application of engineering to mixed culture systems. The paper by Vezina on new techniques in microbial transformations of organic compounds and the physiology of spores and the one by Garber and Rippon on the characterization of organisms by their enzyme content conveniently collect diverse information not easily assembled. The problems involved in the preservation of pharmaceuticals using nontoxic chemicals are a continuing concern, and Borick’s evaluation of some new compounds promotes the hypothesis that “something new is always worthy of study.” Ceigler’s summary on mycotoxins certainly highlights the growing importance of microorganisms in food problems. It is interesting to contemplate that 100 years ago Pasteur was studying some of the same problems covered in these reviews. Has microbiology advanced in the century or have the problems been too much for us? No matter how you look at it, there is still plenty of work ahead.

W. W. UMBREIT D. PERLMAN August, 1968

vii


CONTENTS LIST OF CONTRIBUTORS ............................................................................... PREFACE....................................................................................................... CONTENTS OF PREVIOUS VOLUMES ...............................................................

v vii xiii

Detection of l i f e in Soil on Earth and Other Planets. Introductory Remarks ROBERTL. STARKEY Text ................................................................................................. Reference.. ........................................................................................

1 3

For What Shall W e Search? ALLANH. BROWN Text ................................................................................................. References ........................................................................................

5 1s

Relevance of Soil Microbiology to Search for l i f e on Other Planets G. STOTZKY I. Introduction ...................................................................................... 11. Factors Affecting Microorganisms in Soil ............................................... 111. Methods for Detecting Life in Terrestrial Soils ........................................ IV. Speculations on Numbers and Types of Soil Microorganisms..................... V. Conclusions. ...................................................................................... References ........................................................................................

17 19 36 41 44 45

Experiments and Instrumentation tor Extraterrestrial Life Detection GILBERTV. LEVIN I. Introduction

......................................................................................

11. Sampling ........................................................................................... 111. Life Detection Methods .......................................................................

IV. Automated Microbial Metabolism Laboratory ......................................... References ........................................................................................

ix

55 56 56 67 71

X

CONTENTS

Ha lophilic Bacteria D . J . KUSHNER

.

...................................................................................... I1 . Ecology and Classification of Halophiles ................................................ 111. The Nature of Halophilic Life .............................................................. 1 Introduction

. . .

1V Taxonomic Importance of Halophilic Lipids ........................................... V Growth and Survival of Extreme Halophiles ........................................... V1 Speculations on Moderate Halophiles .................................................... VII . On the Origin of Extreme Halophiles ..................................................... References ........................................................................................

73 75 78 91 92 94 96 97

Applied Significance of Polyvalent Bacteriophages S. G . BRADLEY I . Introduction ...................................................................................... I1 . Fundamental Aspects ..........................................................................

.

111 Infection of Fermentations

...................................................................

IV . Phage Typing ..................................................................................... V . Conversion ........................................................................................ V1; Transduction ..................................................................................... VII . Bacteriocins....................................................................................... VIII Tools for Applied Research .................................................................. IX . Conclusions ....................................................................................... References ........................................................................................

.

101

102 113 116 118

120 123 124 132 133

Proteins and Enzymes as Taxonomic Tools EDWARDD . GARBER A N D JOHN w . RIPPON . ...................................................................................... I1 . Zone Electrophoresis .......................................................................... I l l . Protein Profiles and Zymograms in Microbial Taxonomy ....................... IV . General Comments ............................................................................ 1 Introduction

References ........................................................................................

137 140 144 151

153

Mycotoxins A N D ElVIND B . LILLEHOJ ...................................................................................... I1 . Aflatoxin ........................................................................................... 111. Alimentary Toxic Aleukia (ATA).......................................................... IV . Ochratoxin ........................................................................................ V . Sporidesmin....................................................................................... VI . F-2 Estrogenic Factor (Zearalenone) ..................................................... V11 . Pink Rot Dermatitis ............................................................................ VIII . Slaframine (Slobber Factor) .................................................................. IX . Yellow Rice Toxins ............................................................................ X . Stachybotryotoxicosis ......................................................................... XI. Rubratoxins ....................................................................................... XI1 . Other Mycotoxins .............................................................................. XI11. Summary .......................................................................................... References ........................................................................................

ALEX CIEGLER

1. Introduction

155

156 195 197 199 200 201 202 203 205 206 207 209 2 10

xi

CONTENTS

Transformation of Organic Compounds by Fungal Spores CLAUDEVEZINA,S. N. SEHGAL,A N D KARTARSINGH I. Introduction ......................................................................................

.

V. Transformation of Other Organic Compounds ........................ ................ V1. Discussion and Conclusion .................................................................. References ................. .................

22 I 222 221 24 I 260 264 265

Microbial Interactions in Continuous Culture HENRYR. BUNGAY,I l l A N D MARYLou BUNGAY I. Introduction ........ ...... 11. Nomenclature ................. . ...... ................,......................,.................... I l l . Systems of Defined Microbial Composition ............................................. IV. Multiculture Systems .......................................................................... V. Computer Models ............................................................................... V1. Conclusion ............ .................................................. References .................................. .......

269 215 211 282 281 288 288

Chemical Sterilizers (Chemosterilizers) PAULM. BORICK I. Introduction ................................................................................ ...... 11. Methods for Evaluation of Chemosterilizers ............................................ 111. Gaseous Chemosterilizers ............................. .................. ................... IV. Liquid Chemosterilizers ....................................................................... V. Conditions for the Use of Chemosterilizers ........ ......... VI. Summary .......................................................................................... References ............................................................ ...........................

..

.

29 1 293 296 300 301 3 10 3 10

Antibiotics in the Control of Plant Pathogens M.. J. THIRUMALACHAR I. Introduction

....................................

Antibiotics against Plant Viruses ........................................... ............... Use of Antibiotics for Control of Fungal Disease .................................... Antibiotics for Preventing Postharvest Decay .... Advantages and Disadvantages of Antibiotics over Chemical Fungicides ..... References ........................................................................................

313 315 319 323 3 24 33 I 331 333

AUTHORINDEX................................................................................. SUBJECTINDEX .............. ................. CUMULATIVE A .............. ................. CUMULATIVE TITLEINDEX...........................................................................

339 359 362 365

11. General Information on Use of Antibio 111. Antibiotics against Bacterial Diseases of Plants ....................... ...............

1V. V. V1. VII.


CONTENTS OF PREVIOUS VOLUMES A Commentary on Microbiological Assaying F. Kavanagh

Volume 1

Protected Fermentation Milo's Herold and Jan Nekbsek

Application of Membrane Filters Richard Ehrlich

The Mechanism of Penicillin Biosynthesis Arnold L. Demain Preservation of Foods and Drugs by Ionizing Radiations W . Dexter Bellamy

Methods

in

the

Newer Development in Vinegar Manufactures Rudolph J. Allgeier and Frank M . Hildebrandt

The State of Antibiotics in Plant Disease Control David Pramer Microbial Synthesis of Cobamides D. Perlman Factors Affecting the Antimicrobial Activity of Phenols E . 0.Bennett Germfree Animal Techniques and Their Applications Arthur W. Phillips andJames E . Smith Insect Microbiology S.R. Dutky The Production of Amino Acids by Fermentation Processes Shukuo Kinoshita Continuous Industrial Fermentations Philip Gerhardt and M . C. Bartlett The Large-Scale Growth of Higher Fungi Radcllffe F. Robinson and R. S. Davidson AUTHOR INDEX-SUBJECT

Microbiol Control Brewery Gerhard J . Hass

INDEX

The Microbiological Transformation of Steroids T. H. Stoudt Biological Transformation of Solar Energy William J. Oswald and Clarence G. Golueke SYMPOSIUM O N ENGINEERING ADVANCES IN

FERMENTATION PRACTICE

Rheological Properties of Fermentation Broths Fred H. Deindoerfer and John M . West Fluid Mixing in Fermentation Processes J . Y.Oldshue Scale-up of Submerged Fermentations W .H. Bartholemew Air Sterilization Arthur E. Humphrey Sterilization of Media for Biochemical Processes Lloyd L. Kempe

Volume 2

Newer Aspects of Waste Treatment Nandor Porges

Fermentation Kinetics and Model Processes Fred H. Deindoerjer

Aerosol Samplers Harold W . Batchelor

xiii

xiv

CONTENTS OF PREVIOUS VOLUMES

Continuous Fermentation

AUTHOR INDEX-SUBJECT INDEX

W. D. Maron Volume 4

Control Applications in Fermentation

George]. Fuld AUTHOR INDEX- SUBJECT INDEX

Volume 3

Preservation of Bacteria b y Lyophilization

Robert J. Heckly Sphaerotilus, Its Nature and Economic Significance

Norman 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 . Chattgny Oxidation of Aromatic Compounds by Bacteria

Induced Mutagenesis in the Selection of Microorganisms S . 1. Alikhanian The Importance of Bacterial Viruses in Industrial Processes, Especially in the Dairy Industry

F. J. 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. Nickel1 AUTHOR INDEX- SUBJECT INDEX

Martin H. Rogoff Screening for and Biological Characterizations of Antitumor Agents Using Microorganisms

Frank M. Schabel, Jr., and Robert F. Pittillo 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 . Hulme

Volume 5

Correlations between Microbiological Morphology and the Chemistry of Biocides

Adrien Albert Generation of Electricity by Microbial Action

J. B. Davis Microorganisms and Biology of Cancer

the

Molecular

G . F . Cause Rapid Microbiological with Radioisotopes

Determinations

Gilbert V . Leuin The Present Status of the 2,3-Butylene Glycol Fermentation

Sterling K. Long ond Roger Patrick


Aeration in the Laboratory W . R. Lockhart and R. W . Squires

xv

AUTHOR INDEX-SUBJECT INDEX Volume 7

Stability and Degeneration of Microbial Cultures on Repeated Transfer Fritz Reusser Microbiology of Paint Films Richard T. Ross The Actinomycetes and Their Antibiotics Selman A. Waksman Fuse1 Oil A. Dinsmoor Webb and John L. Ingraham AUTHOR INDEX-SUBJECT INDEX Volume 6

Global Impacts of Applied Microbiology: An Appraisal Carl-Coran Hedbn and Mortimer P. Starr Microbial Processes for Preparation of Radioactive Compounds D. Perlman, Aris P. Bayan, and Nancy A. Ciuflre 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

Microbial Carotenogenesis Alex Ciegler Biodegradation: Problems of Molecular Recalcitrance and Microbial Fallibility M . Alexander Cold Sterlization Techniques John B. Opfell and Curtis E. Miller Microbial Production of Metal-Organic Compounds and Complexes D. Perlman 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. Mhatre The Microbiology of Freeze-Dried Foods Gerald J . Silverman and Samuel A. Goldblith Low-Temperature Microbiology Judith Fanell and A. H . Rose AUTHOR INDEX-SUBJECT INDEX Volume 8

Industrial Fermentations and Their Relations to Regulatory Mechanisms Arnold L. Demain Genetics in Applied Microbiology S. G . Bradley

xvi


Microbial Ecology and Applied Microbiology Thomas D. Brock The Ecological Approach to the Study of Activated Sludge Wesley 0.Pipes Control of Bacteria in Nondomestic Water Supplies Cecil W. Chambers and Normon A. Clarke The Presence of Human Enteric Viruses in Sewage and Their Removal by Conventional Sewage Treatment Methods Stephen Alan Kollins Oral Microbiology Heiner Hoffman Media and Methods for Isolation and Enumeration of the Enterococci Paul A. Hartman, George W.Reinbold, and Devi S . Saraswat Crystal-Forming Bacteria Pathogens Martin H. RogofF

as

Insect

Mycotoxins in Feeds and Foods Emanuel Borker, Nino F. Insalata, Colette P. Levi, and ]ohn S . Witzeman AUTHOR INDEX

- SUBJECT INDEX

Volume 9

The Inclusion of Antimicrobial Agents in Pharmaceutical Products A. D. Russell, ]une Jenkins, and I. H. Hawison Antiserum Production in Experimental Animals Richard M.Hyde Microbial Models of Tumor Metabolism C.F. Cause Cellulose and Cellulolysis Brigitta Norkrans Microbiological Aspects of the Formation and Degradation of Cellulosic Fibers L. Iuraiek, 1. Ross Colvin, and D. R. Whitaker The Biotransformation of Lignin to Humus -Facts and Postulates R. T. Oglesby, R. F. Christman, and C. H. Driver Bulking of Activated Sludge Wesley 0. Pipes Malo-lactic Fermentation Ralph E. Kunkee AUTHOR INDEX - SUBJECT INDEX

Detection of life in Soil on Earth and on Other Planets' Introductory Remarks2

ROBERT L. STARKEY Department of Biochemistsy and Microbiology, Rutgers, The State University, New Bnrnswick, New Jersey

The first four chapters are concerned principally with the microbiological aspects of the ventures to the moon and planets for the purpose of determining whether or not there is extraterrestrial life 01 substances that might be precursors of living things. They deal with a subject that would have had little relevance to reality until recent years. Now it is indeed realistic in that devices have been transported to the moon to sample and test its surface substance, and others are in advanced stages of development. Opportunity to examine the moon's soil material is important to microbiologists because in it one may find living things like or unlike those in the soil on Earth. Even if no life is detected, information will be obtained that is needed for investigations of Mars and other planets. There is some knowledge of the elements contained in the moon and the planets, but, since no material from them has been brought to Earth yet, physical and chemical properties of the substances and the presence of living things or primordial substances that might be precursors of living things are subjects for speculation. Certainly, the symposium from which these chapters were taken was unusual in that it was concerned little with factual information and the material under examination is far beyond our present reach. There is dependence on intelligent conjecture about what should be sought and on exceptional ingenuity to devise and construct highly sophisticated devices to obtain the desired information. Lack of facts promotes questioning and the questions are indicative of the information that will be sought. Many of the important questions are ans'wered or discussed in the following discourses. Some of the questions that come to mind are the following: 'This and the next three chapters are taken from a symposium of the Annual Meeting of the American Society for Microbiology, New York City, May 2, 1967. R. L. Starkey convenor. 'Paper of the Journal Series, College of Agriculture and Environmental Science Rutgers, The State University, New Brunswick, New Jersey.

1

2

ROBERT L. STARKEY

Why search for living things on extraterrestrial bodies? Based on what is known of the environments of the moon and the planets, what is the probability that these environments are suitable for living things? Is it possible that there are living things unlike those we know on Earth and, if so, can they be detected and recognized? Should it be assumed that bacteria will be present if there are any living things? What information has been obtained or can be obtained in laboratory research to provide a substantial basis for the investigations in space? What kinds of devices are needed to examine, sample, and test planetary substance and to transmit the information to Earth? Are there problems concerned with sterility of the equipment to be landed on the celestial bodies? Should the tests be made b y manned or unmanned vehicles? Are there problems inherent in bringing to Earth materials obtained from the moon and planets? Can presence of living things be detected by determination of metabolic activity and of products typical of terrestrial life? Assuming that the moon is the first area to be investigated, what follows? The scope of the subject is indicated in the chapter by Brown and the basic ideas and principles are discussed. These include the importance of the search, what to search for, how, when, and where to search, and the role of the scientist and the engineer. Much attention is devoted to soil in the thinking about life on the planets because soil on Earth harbors a great diversity of living things. It is a reasonable assumption that there will be living things in planetary soil if there is life in any form on the planets. The chapter by Stotzky deals with the knowledge of soil microbiology that may be applicable to studies of life on the planets. The concluding chapter by Levin deals with the kinds of devices that are being constructed for missions to obtain evidence of the existence of extraterrestrial life and its characteristics. Remarkable miniaturized equipment has been produced that effectively makes various determinations and transmits the results. Much thought and effort have been devoted already to exobiology and the program is evolving rapidly. It is anticipated that important definitive information will be obtained during the next few years.

DETECTION OF LIFE IN SOIL ON EARTH AND OTHER PLANETS

3

Hopefully, there will be evidence of life on the moon or substance from which living things originate. Even if there is no life there we can be confident that the search will continue on Mars and elsewhere because the probability is great that life exists many places in the universe. The following words of Shapley (1958)provide reassurance for those who would seek for life on celestial bodies and ascertain its characteristics:

.

. , whenever the physics, chemistry, and climates are right on a planet’s surface, life will emerge, persist, and evolve. . . . Life is a cosmos-wide occurrence. . . We should contemplate at least 10l4planetary situations for life. . . Many but not all of these planets probably have the plant-animal interdependence in which we ourselves participate. . . . Exactly where these life bearing planets are we cannot now say; perhaps we never can. . . Nor can we say what kind of organisms inhabit these other worlds.

.

.

.

REFERENCE Shapley, H. (1958). “Of Stars and Men,” 157 pp. Beacon Press, Boston, Massachusetts.


For What Shall We Search?

ALLAN H. BROWN Department of Biology University of Pennsylvania, Philadelphia, Pennsylvania

The central theme of our discussion in this and in the next two papers is the detection of exotic life, a topic which has received attention repeatedly in both the popular and scientific literature, especially during the past half dozen years (Firsoff, 1963; Keosian, 1964; Mamikunian and Briggs, 1965; Pittendrigh et al., 1966; Shklovski and Sagan, 1966; Shneour and Ottesen, 1966; Sullivan, 1965). While the present contribution must be in some sense introductory to the general topic, it will be appropriate to deal mostly with what we may term the design phase of our program to extend biological research to other worlds. I believe that the most critical part of almost any major research program is the design of the experiments and still more fundamentally the scientific justification for doing the experiments at all. Sometimes these things may be taken very much for granted and then the experiments are performed chiefly because available techniques make them possible or the research is carried out in response to the demands of funding patterns and political considerations. There is, of course, nothing immoral about accepting the realities of public support of science as long as we scientists keep the scientific ends in view even though the day-to-day efforts deal largely with means, not ends. How shall we conduct our search? How shall we design our experiments? What questions shall we ask? What approaches should we try? What sort of grand strategy ought we develop? These are questions which arise wherever research is planned and they are especially critical in what we have come to call exobiology chiefly because the scientific challenge, the difficulties and costs, and the potential rewards for a better understanding of life as we know it on the Earth and of man’s philosophical position in the universe are all much greater than for most other research efforts with which we may make comparisons. If we are to be successful in this grand venture (and it is not yet clear how much optimism is warranted) we shall have to overcome a series of great difficulties - scientific, technological, and socioeconomic. We shall have some hard choices to make following those to which we already are committed, decisions which often cannot be 5

6

ALLAN

n.

BROWN

made easily, for some will involve reaching an imaginative compromise between daring and feasibility. First, let us be clear about why we want to conduct the search for life in extraterrestrial environments. To say that we just want to see what’s there is almost trivial. Man’s strongest motivation here is his intellectual curiosity about himself and about the living world in which he dwells. As far as we know for certain, there is no life anywhere in the universe except on this planet we call Earth. Proof of the uniqueness of terrestrial life seems impossible but disproof that our life as we know it is unique may come within a half dozen years. That “disproof’ as I have termed it is most likely to come from biological and biochemical observations on the surface of Mars. I know of no single scientific event which could be more widely appreciated by the educated and less fortunate alike than would be this new insight on man’s identity. Thinking now as a biologist I must take a different tack. Of course, we shall be delighted to learn-perhaps in the early ’70’s-that there are life forms on Mars. We all would want to learn about such fascinating biological entities. But more basic than asking “Is life there?” would be the question: “What stage of organic evolution has been attained on the Martian surface?” If there is no life on Mars the chemical evidence of organic evolution from the raw constituents of the “primitive reducing atmosphere” through the abiotic synthesis of more and more complex organic compounds to the abundance and diversity presumably required for the appearance of life-all this evidence which on our planet was long ago destroyed by terrestrial life processes -would still be retained on a highly evolved but sterile Mars. Not only the details but even the broad-brush neo-Oparin treatment of how the organically rich environment evolved prebiotically on this planet, which another generation now reads in standard high school biology textbooks, is, after all, only a set of reasonable hypotheses. I should like to see nature’s control experiment-a planet in which chemical evolution is well advanced but where biopoesis has not yet occurred. To put it another way, if I were permitted to specify what kind of planet would yield the maximum in biologically interesting new information, I would unhesitatingly say two planets, one inhabited by endogenous life forms and the other one on which life has not yet gotten started. While it is hard to be quite objective about it, I cannot truthfully say which of these two hypothetical planetary research objectives would yield the more scientifically useful data relevant to

FOR WHAT SHALL WE SEARCH?

7

terrestrial biology. One thing is clear; the search for extraterrestrial life will not end in failure if it can be shown that Mars, Venus, the moon, etc., are certainly lifeless. Our search could be highly rewarding (if we ask the right questions) on a sterile planetary surface. Where would it be profitable to search? Our present knowledge of the solar system strongly suggests that Mars, the Earth’s moon, and Venus are the only likely abodes of life beyond the Earth and in that order. Mars, I feel very sure, ought to be the prime target for an intensive investigation. As ground-based telescopy and observations from near space, fly-by, and orbiter operations round out our store of remotely obtained information on atmospheric and surface conditions, we cannot expect to learn much about Martian life (if there is any) nor even to detect its presence. But we need the environmental data, including identification of the seasonal changes, to assess factors of ecological significance for our subsequent and more advantageous surface investigations. We also can expect that better data on gross surface characteristics and especially on the atmosphere will improve markedly our success with the first landing we attempt. The first steps then to an intensive survey of, say, Martian soil microbiology, will be through remote sensors rather foreign to the experience of most bacteriologists. Who among us has looked for a bacterium with an astronomical telescope? Six years and some hundreds of million dollars later we can begin to land a series of unmanned experimental capsules on the Martian surface. The program which will do this has been named Voyager and when Voyager becomes operational-as we now expect in the early ’70’s-we shall be in a position to ask all sorts of fascinating questions about what life forms may be there and about the micro and macro environments that harbor them. What questions? Just what shall we look for? If the crux of the whole enterprise rests on the design of our experiments, on deciding what questions to ask, we cannot shrug off this responsibility by thinking “When we are ready to land lifedetecting experiments on Mars biologists and especially microbiologists will come up with all kinds of good ideas.” Lead time for experiment development is long and the time for decisions is now! Discussions of Voyager usually focus on Mars but the program is broadly one of planetary, not just Martian, exploration. Doubtless the initial emphasis should be on Mars and should be principally biological. That is because Mars now seems the most likely of the nonterrestrial planets to harbor life and because biological studies must precede other kinds of investigations more likely to result in contami-

8

ALLAN H. BROWN

nation of the planet by terrestrial organisms. Such contamination, if achieved, would jeopardize both the scientific objectives of our biological search as well as the Martian environment and the resources it may contain. Only after we have more detailed knowledge of that environment can we be in a position to dismiss the latter hazard as inconsequential. How shall we search for life on Mars? From what we know and can deduce about the Martian environment, it has all the requisites for life as we know it. Although relatively inhospitable by our own standards, even some terrestrial organisms would surely survive and probably multiply on Mars. There seems little question that whatever kind of life may have started up the Martian evolutionary ladder, it should be suitably adapted to its environment and would not be leading a “cliff hanging” existence in unusual niches of greater than average environmental hospitality. I think it is a highly likely assumption that, if there is any life on Mars, it is widespread, diverse, and well adapted ecologically. Moreover, as has often been said, we can envision a Martian biota consisting only of microbes, but a biota without any microorganisms is inconceivable. Can there be any doubt that so-called “life detectors” whose function will be to identify correlative properties of microbial growth and metabolism, will be prominent among the experiments aboard the first Martian landing capsule? Nowadays, design studies and mission planning documents have a great deal to say about inferences to be drawn from detected changes in pH, specific light absorption, COZ release, oxygen metabolism, catalyzed isotope exchange, and numerous other quantities well suited to automated measurement. How definitive would such measurements be? When we define life (as we must for our students even if we ourselves are not designing experiments to be sent to Mars) our definition must be so general that no single operational criterion appears unique. If we want a quick yes-or-no answer to the question, “Does this sample contain anything alive”?, then we cannot expect our most definitive criteria to be necessarily applicable to the sample in question, while the broadest and most fundamental criteria are far less apt to be operationally useful (Quimby, 1964). It may be discouraging, but it is surely realistic to admit that no single test for life will suffice for our purpose but, if we can bring to focus enough different kinds of information about one or a group of exotic life forms, we can increase enormously our confidence in the mutually supporting results of our tests. Not only may we expect to


9

confidently detect life, if it is present, but also to assemble a large amount of descriptive information about it. One life detector cannot do all this; ten measurements different in kind can provide far more than 10 times as much information; 50 different tests with suitable qualitative diversity could yield so much more information at so much higher confidence levels that such results would be worth millions of times more than the data obtained from only ten kinds of measurements. To paraphrase a familiar aphorism, the Martian biota may mislead some of our life detectors all of the time and all life detectors some of the time, but not all life detectors all of the time. The moral is that we need a diversity of experimental approaches and we must bring different techniques to bear on the problems of identifying and describing extraterrestrial life and its ecosystem. A few experiments alone will not suffice except as publicity stunts; our science requires a multiplicity of measurements and accordingly the Voyager landing capsule must be larger than a bread box. We should be thinking in the size range of at least a Volkswagen. Hampered as we are by our familiarity with terrestrial examples when we try to predict what kind of macrobiota might accompany the Martian microbiota, we tend to think first of such things as lichens, mosses, or perhaps colonies of blue-green algae. Nevertheless, I cannot escape the belief that Mother Nature (or whatever may be the Martian counterpart of this pantheistic maternal image) has been much more imaginative. I strongly suspect that, if Mars is inhabited, it will even have a luxuriant macrobiota and so we may be in for some surprises. Putting aside for the moment the problem of how to investigate unknown, exotic microbes, when it comes to higher organisms I am very confident that I know how best to search for them effectively. The most fruitful method by far will be some brand of photography or, more generally, optical imagery. Excellent examples of how this can be done remotely are to be seen in Surveyor spacecraft photographs of the lunar surface features. Impressive as are these demonstrations, it is within even the present state of the art to obtain pictures at still higher effective magnification were that required. I believe that Martian higher organisms, if revealed in pictures returned to Earth, would be identified unmistakably on morphological criteria alone. I am not suggesting that we may be confronted with a “little green man” sticking his tongue out at us, but I feel certain that the exotic morphology of even the most bizarre plant-animal creature will be immediately recognizable as Martian life on candid camera. The lo5 or so bits of information needed for such a picture would, I believe,

10

ALLAN H. BROWN

convey more scientific knowledge than the results of all other kinds of tests which can be used on the macroscopic component of the (still hypothetical) Martian biota within the same bit limitation. One picture would indeed be worth a thousand words. If I suggest that more highly evolved life, intelligent life, may exist on Mars, the reader would be justified in rejecting that speculation as not worth serious consideration. Of course I share this conservative judgment. I would be willing to wager at very long odds that there are no intelligent beings anywhere in the solar system except on Earth and, if this sounds chauvinistic, be assured that I have no exaggerated respect for mankind's collective intelligence; I merely suggest that w e probably have no competition here in the solar system. However, one can drive the argument further, and I must truthfully admit even intelligent life on Mars is not impossible although extraordinarily improbable. In any case, if finding evidence of Martian intelligence is to be one of the grand surprises in store for us, we need not make any special effort to identify the Martians. We may be confident that their intelligence will manifest itself and, if it should be of a high order comparable, say, to our own with well-developed techniques of communication, we may reverse the aphorism and suggest that one word may be worth a thousand pictures-even if we cannot read the language of the message there contained. How, when, and where shall we research? Earlier I stated emphatically that Mars is our most promising objective. The near absence of a lunar atmosphere and what probably are prohibitively high surface temperatures on Venus make these two nearest neighbors less interesting biologically. When we can reach Mars economically depends on the relative positions of the sun, the Earth, and Mars, and on the motions of the two planets. As it turns out, about every two years a Martian mission becomes energetically feasible, although there are additional opportunities for combined Venus-Mars missions which may be considered also. Not all so-called minimum energy opportunities are equally advantageous. The best will occur in 1969 and 1971. Then missions will become more costly (in terms of energy needed) through the '70's and b y 1984 we shall have a launch opportunity as favorable as the '69 and '71 opportunities. The 15-17 year cycle will then repeat itself. With the long lead times necessary to launch a Voyager spacecraft toward Mars-lead times of the order of 4 to 5 years-it is obvious that w e have missed our best chances for Martian rendezvous for nearly a generation. But, discouraging as this may seem, we are


11

fortunate in having available a booster rocket capable of transferring a still significant payload to Mars even during some of the less favorable opportunities. The Saturn V vehicle developed for Apollo, a largely nonscientific manned conquest of the moon, will be used for Voyager, a completely scientific unmanned exploration of the other planets. What then are the pacing items in this venture? They are twofold: funding and lead times. The cost of Voyager may be about 1/10 the cost of Apollo, and Congress has not really made up its mind whether planetary exploration with initial emphasis on the search for and study of Martian life would be worth that cost. Only preliminary work and planning has been funded and is under way. But even if an unlimited fraction of the nation’s resources were to be devoted immediately to Voyager-and of course that will not happen-probably it would even so be impossible to launch an optimal scientific payload by the 1971 opportunity. Such long lead times are discouraging and frustrating, but they are part of the real world and w e must plan accordingly. Meanwhile, what must be done in the next half dozen years which particularly concern us as scientists? Several quite new things must be undertaken successfully. First, the complexity of the Voyager scientific payload is a giant step ahead of anything NASA has attempted before and it is significant that the capsule design requirements will be more exactingly determined by the scientific payload than has been the case for other major space science missions. A continuing intimate dialogue between scientists and engineers will be essential if the design is to optimize rather than to compromise the scientific observations. Wolfgang Panofsky (1966) has put this very aptly, although he was thinking of his experience with large accelerator development rather than with space research: Some [engineers] prefer to have the experimental physicist write down a series of requirements and then go away and not talk to the engineer until the product is delivered. The problem is, of course, that any advanced technological thing simply doesn’t get built that way. Usually the scientist doesn’t quite know what he wants. H e may in fact specify things that can’t be built. So he needs to work very closely with the engineer and find out that this particular solution won’t work or maybe another one will; then the scientist can change his mind as to how really rigid certain needs are. This constant back-and-forth communication is really the key to creating a new scientific technological advance. As soon as you put a barrier between the two people in which the scientist becomes the fellow who writes down what is needed and the engineer the one who solves the problem, then you automatically at best get something which is designed in an unimaginative way and which costs too much or at worst won’t work at all.

12

ALLAN H. BROWN

The NASA office of Space Science and Applications is accustomed to planning a space mission by first deciding on the vehicle, general spacecraft configuration, and numerous other mission constraints. Then it issues to the scientific community so called “Announcement of Flight Opportunities” and “Requests for Proposals.” These are screened by various criteria, including the ability of each experiment to be accommodated within the numerous engineering constraints. A selection is made and the experiments essentially fly piggy-back on a piece of well-engineered hardware. That procedure cannot be applied successfully to the development of the Voyager payload. NASA must generate a truly new mechanism to ensure a collaboration between engineers and biological scientists and to implement effective collaboration among the investigators. A new management concept is required and it is not clear whether that has yet been fully appreciated. It remains, even at this late date, as one of the things “for which we must search” -in the sense of the title of this paper - in order that we may effectively get on with the job of Voyager capsule design. The reason a well-integrated team effort is needed is partly because the various devices for tests, observations, and measurements which will make up the Voyager scientific payload will contain large amounts of redundant hardware, and economy of payload capacity demands that power supplies, telemetry channels, sample handling apparatus, reagent reservoirs, various sensors, much of the electronic data processing, and many other facilities must be shared in common. Even more significant is the manner in which the results of one set of observations would condition a choice among a large number of other possible measurements. Thus some test results will set priorities for other tests to follow on the same sample. As advantage is taken of this, an enormous economy will result in the total number of tests which would be performed. The concept requires, of course, either onboard preprogramming of many alternate test sequences or, more logically, reprogramming by experimenters on the earth. Conceptually, this payload of well-integrated and interrelated experiments has been referred to as an Automated Biological Laboratory, and it is this ABL on which Voyager science chiefly depends. (I understand that NASA official terminology has recently changed ABL to VBL, Voyager Biological Laboratory, so it is by that acronym that it will be referred to-at least until the next reorganization of the NASA glossary.) How to develop a VBL? By what mechanisms will NASA initiate, nurture, and coordinate this science engineering development. NASA


13

management must be innovative and imaginative if the VBL is to achieve its scientific potential. Still another new dimension in Voyager is the strict quuruntine which the world biological community looks to the U.S. and to the U.S.S.R. to exercise with respect to our Martian missions (Cospar Information Bulletin, 1966). Among the several very good reasons why Mars must be protected from contamination by terrestrial organisms at least during the earliest phase of exploration the most cogent are 1. Earth organisms, if they become established and propagate on Mars, would surely distort and effectively destroy the chemical evidence of prebiotic evolution, just as happened on the Earth. 2. Earth organisms, if successfully introduced into an exotic flora, would by definition have an influence on it. The effects could include the ultimate destruction of the Martian biota and, in view of our present ignorance about it, to risk such wanton destruction would be scientifically irresponsible. 3. The detection of Martian life forms by sensors transported on Voyager landers would become quite uncertain if the detector systems could become contaminated with microbial hitchhikers from the earth. We should realize that every kind of test we can perform with a VBL for Martian life will also be a test for terrestrial life. Ambiguity of results would be intolerable. We must have a high level of confidence that our life detectors operate on Mars with an acceptably low contamination risk. For these and other reasons the “acceptable risk” has been set at one chance per mil that contamination will result during the course of unmanned Martian exploration. This seems to be a conservative figure agreed to by various advisory scientific groups both in the U.S. and abroad. On any given mission a correspondingly smaller risk of contamination must be certified, and since we are dealing with large landing capsules which cannot be autoclaved and still be expected to function, it has been necessary to develop almost from scratch a whole new technology to ensure the sterility of these capsules. This is not the occasion for a thorough description of the very interesting history of this development. But I want to emphasize that in this area we have an example par excellence of how necessary it is for scientists and engineers to work together in an atmosphere of mutual respect. When spacecraft engineers were first confronted with the major problems involved in building and launching a sterile payload into space the usual reaction was that this would be (a) cause

14

ALLAN H. BROWN

for delay, (b) prohibitively expensive, (c) unnecessary, and (d) impossible. Even some basic scientists were at first unconvinced that sterilization would be required and they were willing to compromise in various ways to get on with some kind of exploration of Mars. Now, I think for an engineer to deny that space probe sterilization is essential was at worst merely grossly philistine, but for a scientistespecially for a biologist-to take this position is, I submit, a heinous act of scientific irresponsibility. Fortunately, more sober analysis and better communication between biologists and engineers has paid off. What some thought was impossible became a difficult challenge and today, after much intensive work to develop a sterilization technology adequate for the purpose, NASA is well along toward the needed technical solutions. Very briefly, the method is based on dry heating of the entire capsule within its protective shroud at relatively low temperatures but for long times. Various heating regimens from 105°C for 7 days to 135°C for 1 day have been examined. Whatever the final choice, it is clear that a careful selection of heat-tolerant components will be required. It is the selection, testing, monitoring, and design restrictions which will increase the cost of a sterilizable payload over that of an unsterilized capsule-an increase which may amount to from 10 to 30%. However, more than sterility is to be achieved by this method of building sterilizable space hardware. The selection and screening of heat-resistant components will also ensure a higher overall reliability for the many critical subsystems, and our chances for success for Voyager missions will be enhanced by these new procedures. Scientists who have been involved as experimenters on flight projects are well aware of the heavy price which must be paid in man years of effort as well as in dollars for each spacecraft failure and any source of additional reliability is sure to be most welcome. Contrary to what some of us thought earlier, the most difficult problem is not likely to be finding acceptable components which will suitably tolerate the rigors of dry heat sterilization. In fact, that set of problems no longer is deemed insurmountable. The principal difficulty which remains is how to design a poststerilization checkout and repair procedure to correct those operating defects which will be detected in the flight hardware after sterilization but prior to launch. Repairs must be made under rigorously aseptic conditions or else reheating will be required. Successive reheating of the hardware surely is contraindicated for this would lead to predictable degradation of performance. Sterile procedures for making repairs have yet to be


15

devised and tested although NASA has made a good beginning; the developmental effort required must be greatly accelerated. Periodically one hears the suggestion that a manned mission to Mars could be the best way to be certain whether or not there is life there. However, a manned landing on Mars probably will not be feasible for technological as well as other reasons before the mid19kO’s and surely such an expedition would be preceded by unmanned missions. Moreover, a manned landing seems nearly certain to contaminate Mars with Earth organisms. Therefore, the initial geochemical and biological explorations of Mars must be carried out by remote sensors. Man will make the observations, but he will be located here on Earth while he controls his automated laboratory across as much as several hundred million miles of interplanetary space. It is for this great scientific adventure that we must now plan in earnest. I have stressed that the earliest stages of experimental design are likely to be the most critical. May we be equal to the challenge! REFERENCES Cospar Information Bulletin (1966). No. 33, pp. 54-56. Committee on Space Res. Intern. Council of Scientific Unions, Paris. Firsoff, V. A. (1963).“Life Beyond the Earth.” Basic Books, New York. Keosian, J. (1964).“The Origin of Life.” Reinhold, New York. Mamikunian, G., and Briggs, M. H. eds. (1965). “Current Aspects of Exobiology.” Macmillan (Pergamon),New York. Panofsky, W. (1966). Sci. Res. I, no. 11, 36-37. Pittendrigh, C., Vishniac, W., and Pearman, P. eds. (1966).Natl.Acad. Sci.-Natl. Res. Council. Publ. 1296. Quimby, F. H., ed. (1964).NASA (Natl.Aeron. Space Admin.) SP-56. Shklovski, I. S.,and Sagan, C. (1966). “Intelligent Life in the Universe.” Holden-Day, San Francisco, California. Shneour, E. A., and Ottesen, E. A. (1966). Natl. Acad. Sci.-Natl. Res. Council Publ. 1296A. Sullivan, W. (1965). “We Are Not Alone.” McGraw-Hill, New York.


Relevance of Soil Microbiology to Search for l i f e on Other Planets'

G. STOTZKY~ Kitchawan Research Laboratory of the Brooklyn Botanic Garden Ossining, New York I. Introduction ................................................................ 11. Factors Affecting Microorganisms in Soil ........................ A. Substrates ............................................................ B. Mineral Nutrients ................................................. C. Growth Factors ..................................................... D. Antagonism .......................................................... E. Moisture .............................................................. F. Space .................................................................. G. Atmospheric Composition ...................................... H. pH ...................................................................... I. Eh ...................................................................... J. Ionic Strength ...................................................... K. Pressure ............................................................... L. Temperature ........................................................ M. Particulates .......................................................... 111. Methods for Detecting Life in Terrestrial Soils ............... IV. Speculations on Numbers and Types of Soil Microorganisms.. .......................................................... V. Conclusions ................................................................ References ..................................................................

17 19 19 22 23 23 25 26 27 29 30 31 31 32 33 36 41 44 45

I. Introduction The initial phases of the search for extraterrestrial life will apparently concentrate on detecting some form of microscopic life in the surface layers of other planets (147). The reasoning for this, syllogistically, is: the particulate surface layer of Earth -the soil - is the repository for and source of the widest diversity and greatest density of microbial life on our planet; the surfaces of other planets probably consist of some sort of particulate layers; therefore, we should search 'This discussion and some of our studies reported herein were supported in part by Public Health Service Research Grants AI-05810, from the National Institute of Allergy and Infectious Diseases, and AP-00440, from the National Center for Air Pollution Control and by Forest Service, U.S. DepartmentofAgriculture, Grant No. 1. 'Present address: Department of Biology, New York University, Washington Square, New York, New York.

17

18

G. STOTZKY

for microbial life on other planets in their surface layers. There is also a simple pragmatic reason for initially searching in the surface crust of other planets: this is where a life detection device will probably impact. Consequently, regardless of the validity and geocentricity of our logic, our concepts of terrestrial soil microbiology -at this point in time and space-have potential relevance to the search for life on other planets. We should, however, be prepared to have our theories of relevance exploded when the first data are received from the first successfully impacted life detection device. But, until then, let us briefly, and blissfully, examine some of the potentially relevant areas of soil microbiology. Reference is made to recent reviews of these and other aspects of soil microbiology (5, 6,

44,46,99,139,188,219,243). Soil, as an environment for microorganisms, possesses certain unique features, which, although probably not chahging the genotypic biochemical potentialities of microorganisms, probably influence their phenotypic expression. Many of the environmental parameters that influence microorganisms in soil also influence microorganisms in other habitats. Although attempts are usually made to examine each parameter individually, for reasons of simplicity in experimentation and presentation, the essentially unlimited permutations probably exert a greater effect than does any individual factor. These interactions assume major importance when evaluating the tolerances to environmental stresses of terrestrial microorganisms in relation to potential extraterrestrial life. For example, responses to temperature extremes are significantly modified by the ambient pressure and by the type and concentration of dissolved solutes

(145,316). Microbial life in soil exists in discrete microhabitats, the chemical, physical, mineralogical, and microbiological compositions of which undoubtedly differ. Consequently, the influence of various environmental factors on microorganisms is perhaps more difficult to evaluate in soil, as they vary less uniformly than in more homogeneous and less compartmentalized habitats. For example, the pH of soil, although easily determined, is only an average value and provides no indication of the variations in pH between and within the innumerable microenvironments comprising the sample. As food supply, moisture content, aeration status, composition of the microbiota, etc., influence the production of intermediary metabolites within the microhabitats, the pH and, as a result, the microbial com-

RELEVANCE OF SOIL MICROBIOLOGY TO SEARCH FOR LIFE ON OTHER PLANETS

19

position of each will differ. Unfortunately, critical investigation of individual microhabitats is presently not possible, and the soil microbiologist tacitly extrapolates to his particular concept of the microenvironment either from gross macroscopic observations of intact soil or from simple model systems. II. Factors Affecting Microorganisms in Soil

A. SUBSTRATES

The soil probably provides an array of substrates as vast as the microbial types contained therein. The complex residues of plants and animals, the products of these residues altered by microbes, by man, or abiologically, and those materials synthesized by man from primary elements all serve as substrates for one or another group of soil microorganisms, which in turn serve, either alive or dead, as food for others. Because of this apparent omnivorousness of the soil microbiota, there is the temptation to compare the soil to a digestive system composed of cells differing only in their degree of specificity. This analogy, however, is restrictive, even if the ability of the soil to incorporate or dispose of substrates is the sole consideration. T h e soil microbiota is composed of individual species, each with unique biochemical potentialities and ecological niches, which operate sometimes in concert but, probably more often, in antagonistic competition. Furthermore, the type of substrate, the spatial relationships between the substrate and the microbial cell, and the conditions prevailing in the microenvironment dictate which substrates are attacked by what organisms and at what rate (149,249,297, see 6). Nor is the soil as omnivorous as once assumed. Not only do some natural products accumulate under certain conditions [e.g., porphorins (191, 315), “humus” (228, 251)], but the persistence of some man-made molecules (e.g., certain detergents and pesticides) has cast doubt on the classical concept that, to paraphrase Alexander (8), the soil is the all-powerful and final incinerator which eventually converts all organic substances to their theoretical end products by virtue of its “microbial infallibility.” As recently reviewed by Alexander (7), “recalcitrant molecules” exist, which even the most adaptable of terrestrial microbial communities -the soil -does not decompose. In addition, terrestrial life has evolved using and producing substrates that usually contain only one optical antipode of a racemic mixture.

20

G. STOTZKY

The utilization of materials consisting of only the other enantiomorph would, therefore, be extremely slow or, possibly, not occur (see 305, 326). Even the heterogeneous soil microbiota exhibits stereospecificity, in that the L-antipodes in a racemic mixture of amino acids are preferentially decomposed (108). Because such recalcitrant compounds exist on Earth, and microbes, which have evolved on similar compounds composed of the same elements, are incapable of utilizing them, how successful will be the presentation of terrestrial substrates to extraterrestrial organisms with the aim of using their utilization as an index of the presence of life? Terrestrial microorganisms have presumably evolved in response to changing terrestrial conditions, which apparently differ greatly from current conditions on other planets (252,253). If primordial conditions on other planets also differed from those that existed on primeval Earth, then the evolutionary paths and the present genotypic and phenotypic expressions of extraterrestrial organisms might be different from those on Earth. Although intellectually attractive inferences can result from this type of reasoning (e.g., that extraterrestrial life may be based on substrates and energetics different from ours), conclusions derived from geocentric concepts of energetics and genetics indicate that if life exists elsewhere in the universe, it is based on elements and biochemical pathways similar to, if not the same as, those of terrestrial life (see 107,125,230,321). Furthermore, if primordial or present atmospheres on other planets bear a similarity to our primordial atmosphere (125),then the probability is high that the life forms that developed de nouo from these gases under the impact of universal polymerizing energies (e.g., high energy radiations) would be similar. Even if abiological synthesis occurred universally along the lines proposed by Miller and Urey (195),Fox and Yuyama (86),and others (see 211, 217), the subsequent polymerization of these building blocks into living organisms may have differed, resulting in different morphological and biochemical forms. For example, it has been suggested (26,211, 268) that terrestrial soil particles, especially clay minerals, served as templates during the sequential polymerization of abiologically synthesized organic moieties into replicating living forms. On Earth, clay minerals presently have a net negative charge (105)and most natural organic molecules have either no charge or also a net negative charge, depending upon the pH of the environment (145).As the pH of present terrestrial microbial environments


21

is usually above the isoelectric point of most ionizable organic molecules, it is difficult to envision how negatively charged clays served as effective templates for negatively charged molecules. The electrokinetic potential of clays, however, is modified by the ions present in the ambient environment (256,319),and, if the primordial terrestrial milieu contained an excess of ions, especially polyvalent cations, then negatively charged molecules could have been adsorbed by clays. If the primordial environment was not highly acidic or did not contain excess ions, sorption of nonpolar molecules may have been favored over that of polar molecules (an attractive possibility if early organic compounds were aliphatic), and the resultant life forms may have derived energies from rearrangement of intramolecular bonds rather than from reactions between terminal atomic groups. Alternatively, clays may not have functioned as templates until molecules of sufficient size were synthesized, which, by virtue of long carbon backbones, could be sorbed, despite their net negative charge. The possible role of clays and other minerals in the origin of optically active substances must also be considered (see 305). Similar reasoning must be applied if terrestrial life did not develop de nouo, but evolved from a universally distributed propagule, regardless whether its arrival on Earth was according to classical concepts of panspermia (12) or to the half tongue-in-cheek “garbage theory” (107, 254) (i.e., terrestrial life arose from the litter deposited on Earth by picnicking visitors from another world). Evolution of such a propagule into present life forms must have also been conditioned by the environmental pressures unique to Earth. Must it not then be considered that possible life on other planets evolved differently, as their environments are currently vastly different from ours? Terrestrial evolutionary pressures, however, may have had their major impact during primordial times, when conditions on Earth were different from what they are now. The increasing reports of fossil microorganisms (e.g., 18, 19, 20, 55, 128, 263, 264, 315), some presumably as old as 3 billion years (19) but with morphologies not too different from contemporary microbes, as well as reported isolations of living microbial “fossils” (73, 271), support this possibility. If the pristine environments of Earth and other planets and, hence, the primary evolutionary pressures on universally distributed propagules were similar, perhaps the life forms are also similar, even though current environments differ drastically.

22

C . STOTZKY

Regardless of the mechanisms and pathways involved, these and other possibilities must b e considered when expecting extraterrestrial life to metabolize terrestrial substrates.

B. MINERAL NUTRIENTS Despite the predominantly inorganic nature of soil, mineral nutrition can limit the development of microorganisms in soil. This limitation is usually caused not by a paucity of a specific nutrient but by its unavailability. The best-known example is nitrogen. A normal soil usually contains adequate total nitrogen to support maximum microbial development, but seldom enough readily available nitrogen. Most soil nitrogen is present in organic form, either in primary residues or in microbial products and protoplasm (33),and some may be fixed between lattices of expanding clay minerals (209, 246). Phosphorus (60) and sulfur (87), even though required in smaller quantities than nitrogen, may also be in forms not readily assimilable by an actively developing microbial population (297, 298). The availability of other essential nutrients is usually adequate to sustain high levels of microbial activity, although some soils will respond microbiologically to additions of potassium (291). The limiting effects of inorganic nutrients in soil are usually apparent only when substantial amounts of readily oxidizable carbonaceous materials are introduced (291, 296, 297, 298, 299). The level of limiting nutrients then determines the rate of substrate oxidation but, given adequate time, not the amount of oxidation. This turnover rate is determined by the chemical structure of the nutrient-containing compound. For example, the sulfur in thiamine and thioglycollate is more slowly available to a population rapidly developing on glucose than is that contained in cysteine, methionine, glutathione, or MgSO, (298). Mineral nutrients are also present as elements, in inorganic compounds, and on the exchange complex. The initial utilization of elemental forms is usually restricted to specific microorganisms (e.g., nitrogen fixers, sulfur oxidizers). Nutrients contained in water-soluble compounds and on the exchange complex are usually readily available, whereas those present in insoluble forms are less so. Even the transformations of relatively small molecular weight compounds, such as MnOz, may require specific microbes, either alone or in association (41, 199). The solubilization of more complex compounds, such as apatite and other minerals, is relatively slow and may be re-

RELEVANCE OF

sou MICROBIOLOGY TO SEARCH FOR LIFE ON OTHER PLANETS

23

stricted to certain microbial groups (30, 41, 72, 260, 282, 311, 335, 336). Nevertheless, as the bulk of mineral nutrients are present in these insoluble forms, they constitute important nutrient [and energy (321)] sources on Earth and, undoubtedly, extraterrestrially.

C. GROWTHFACTORS It is not unexpected that components of the complex soil microbiota require a diversity of growth factors. The extensive studies by Lochhead and co-workers have demonstrated not only the stimulatory effects of various vitamins and amino acids, but also the absolute requirement of certain soil bacteria for specific amino acids [e.g., the sulfur-containing group (328)], for specific vitamins [ e.g., BIZ (161, 162)], for a water-soluble, heat-stable soil substance dubbed the “terregens factor” (160). The need for agarized soil or for “fulvic acid” extracts for the isolation of unusual microbial forms (206, 208) emphasizes the importance of growth factors in evaluating the composition of the soil microbiota. In terrestrial soils, growth factors are provided by various microorganisms (159), by root excretions (248), and probably by organic residues, but their possible absolute exogenous requirement must be considered when concocting media with which to search for extraterrestrial life. Although it is doubtful that reproducing extraterrestrial organisms are incapable of synthesizing such factors or of deriving them from their neighbors, extraterrestrial environments may serve only as repositories for dormant propagules which may require growth factors for initiation of growth and, hence, detection. The rate of synthesis of growth factors mediates the sequential development of microbial species in soil, as forms not requiring such factors develop rapidly after the introduction of oxidizable substrates, followed by species dependent upon growth factors synthesized by the primary populations. The concepts of zymogenic and autochthonous microbiotas, related here to growth factors, should be considered when evaluating the density and metabolic activity of potential extraterrestrial organisms.

D. ANTAGONISM Antagonism undoubtedly defines the normal mode of microbial life in soil, and, as in other habitats, has many forms: e.g., competition for nutrients and space; predation; antibiosis, in the broadest sense; alteration of the environment. These factors must obviously be con-

24

C . STOTZKY

sidered in the search for extraterrestrial life, as any addition to or removal from extraterrestrial soil may enhance local antagonisms sufficiently to inhibit growth. For example, the addition of a terrestrial growth medium may alter the pH, Eh,or osmotic pressure of a highly poised extraterrestrial system to the extent that growth is inhibited, even if the medium is not directly toxic. Conversely, the medium may overcome local inhibition, perhaps similar to the reversal of soil mycostasis by readily available carbon (65, 157, 163) or the release of hatching factors for nematode larvae from roots or decaying residues (247, 318), and organisms normally not active may initiate growth. Antagonism assumes major importance in potential interplanetary contamination by improperly sterilized spacecrafts or their inhabitants. The contamination of a barren planet, or of one containing life forms incapable of withstanding the onslaught of terrestrial forms, could result in a blurring of the “biogeochemical” record, which would mean not only an irreplaceable scientific loss, but such intrusion would constitute the ultimate in pancolonialism (see 126, 147). Conversely, the potential contamination of the terrestrial environment with extraterrestrial life introduced by returning space voyagers constitutes a frightening prospect of pandemic disease on Earth. Regardless of the direction of the contamination, the success of the contaminant will depend, in part, on its ability to withstand the antagonistic reactions of the indigenous microbiota. Terrestrial soils appear to restrict the establishment and subsequent development of many organisms introduced from other and usually more specialized environments. The rapid decrease in numbers of enteric and pathogenic bacteria in soil (e.g., 89, 90,136, 337, 338) is but one example. Some soils, however, permit the proliferation of exogenous organisms [e.g., agents of respiratory mycotic disease (2, 301)]. The establishment and survival of plant pathogens in soils illustrate the influence of antagonistic and other forces in great detail (see 17,88,192). The numerous examples of synergystic (e.g., 41, 159,340) and symbiotic (e.g., 1, 213) relationships between microbes in terrestrial soil make it necessary to consider such cooperative relationships elsewhere. It is difficult to assume, however, that such a confluence of enzyme systems would be necessary for extraterrestrial life, as direct cooperation is the exception rather than the norm in terrestrial soils.

RELEVANCE OF

sou MICROBIOLOGY

TO SEARCH FOR LIFE ON OTHER PLANETS

25

E. MOISTURE Terrestrial microbes, regardless of their habitat, are aquatic creatures. Not only is water necessary for intracellular metabolism, but an adequate layer of external water is necessary for diffusion of substrates and toxic products, for maintenance of turgor, etc. Microbes may persist for long periods of time in the absence of water (145), usually by virtue of spores, cysts, or other dormant structures, but free water is necessary for reproduction and metabolism. Terrestrial soils are seldom devoid of moisture, and even in desert soils the particles retain some water against the extreme dryness of the atmosphere above the soil (25). Cameron (47) reported that the relative humidity in desert soils seldom ranged below 85%,even at the surface when the relative humidity of the air was less than 15%.This water, however, is not available to plants nor probably to microorganisms. The retention by soil of water against adverse concentration gradients and its unavailability at low levels to biological systems result from the high activity with which water is held by some soil particles. The charged surfaces of clay minerals induce a rearrangement of adjacent water molecules, resulting in a quasicrystalline structure within the surrounding water films (167, 168). This water exhibits greater viscosity than normal water, and temperatures exceeding 500°C. are necessary to remove it (105).Terrestrial microorganisms do not produce the energy to utilize this water, and, should their metabolic pathways be similar, neither would extraterrestrial forms. The crystalline nature of water appears to influence greatly a variety of metabolic processes (e.g., 168, 224). Metabolic anomalies, including those of microorganisms, associated with changes in temperature presumably result from thermal alterations in water structure (62, 63, 70, 71, 194, 212). The significance of these discontinuities in the evolution and activity of terrestrial life has not been determined. In fact, the occurrence of discrete thermal discontinuities in water has been questioned (83). A paucity of water influences the microbial ecology of some soil environments. In desert soils, for example, specific microbial groups may develop in water pockets formed from moisture retained by soil particles and deliquescent salts, and from condensation resulting from shadows, diurnal temperature cycles, and differences in opacities of surface crusts (47). Limitations in available water may also select for genetic types capable of persisting under drought conditions and

26

C:.

STOTZKY

of rapidly exploiting sporadic additions of water (47), similar to the apparent selection of some soil microbes capable of rapidly exploiting carbon substrates by virtue of having a higher metabolic rate than others (135, 244, 339). Cycles of wetting and drying, as well as freezing and thawing, markedly stimulate metabolic activities in soil (28,

93,146,197,223,280).

F. SPACE Moisture also influences microbial development in soil b y its effects on O2 content; on diffusion rates of nutrients, waste products, and gases; and on the space available for proliferation. Microbial activity is probably restricted to water films, contact rings, and capillaries between soil particles (21, 27, 75, 244), and, although filamentous forms can grow through the voids between water films (the filaments appear to be coated with water), feeding undoubtedly occurs only in sites of accumulated water. Under conditions of ample food supply, and if other environmental factors do not become restrictive first, the water films may become saturated with both living and dead cells, with further proliferation dependent upon the rate of decomposition of the dead cells. Such limitations in “active” space probably develop primarily when nutrient sources are in excess (297, 299), a situation that usually occurs only in localized zones, such as around plant roots and pieces of decaying organic matter. Although it is difficult to demonstrate unequivocally limitations of space, either in soil (297, 299) or in pure culture systems (15, 58), they are probably important in the sequential development of microorganisms in soil. Shortly after the introduction of substrates, the available space in the microhabitat may become virtually filled with cells of the primary populations, thereby restricting the development of slower growing or more fastidious populations. Similar limitations should be considered in relation to extraterrestrial habitats, especially if the primary question (i.e., does life exist?) is answered in the affirmative. Limitations of space probably do not occur when the soil is water saturated, as 0 2 then becomes limiting and cell proliferation does not exceed cell lysis. At low moisture contents, microbial development may be restricted by available water and space; at high moisture levels [usually above 65% of the water-holding capacity (251)], development may be restricted by O2 deficits and by toxic factors associated with anaerobic conditions (e.g., 40, 103). Griffin (104), in a recent review, concluded that moisture affects the growth of fungi


27

in soil primarily by influencing the aeration status, and that both high concentrations of COP and deficits of Or are involved. Diffusion rates of solutes and gases are also affected by moisture content and b y saturation of “active” space by microbial cells. Diffusion becomes increasingly more important as sites of microbial growth are approached. The diffusion rates of nutrients and Or towards, and of excreted metabolites and COz away from the cell are influenced not only by concentration gradients across the cell wall, but also by the degree of orderliness of the surrounding water (167, 168).

G. ATMOSPHERIC COMPOSITION As a result of microbial metabolism and impairment of gas exchange, the COZ content of the soil atmosphere is usually higher, and the On content lower, than of the atmosphere above the soil (251). This imbalance is accentuated when oxidizable substrates are introduced and is not corrected even by rapid aeration. For example, when glucose is added and soils are aerated with COn-free air at flow rates exceeding 15 liters/hour, the respiratory quotients during the period of active metabolism exceed the theoretical quotient of 1 (286). With large additions of glucose (e.g., 4% w./w.), quotients higher than 10 have been observed (291), even in soil containing 96% sand. Once the peak in respiratory activity has passed and the glucose has been oxidized to CO, and intermediary metabolites, the quotients decrease to 1 and below (286). Metabolism in soil apparently switches from predominantly aerobic to anerobic pathways when the 0 2 concentration drops below approximately 3pM (100, 101). The widespread distribution of facultative and obligate anaerobes in soil indicates that loci of anaerobiosis occur commonly and that the composition of the soil atmosphere, especially at sites of microbial activity, fluctuates between extremes of high and low tensions of both Or and COP. The tolerance of soil microorganisms to these fluctuations appears to be greater than that of less complex and more specialized microbial populations of other habitats (293, 294). Soil fungi are more tolerant, in general, of high COr and low 0 2 tensions than are bacteria and actinomycetes. The inhibition by high COz concentrations seems to be a function of the pCOr and not of the absence of 02: more bacteria, fungi, and actinomycetes are capable of developing under nitrogen than under COP tensions approaching loo%, and the presence of even a small amount of O2mitigates the in-

28

G . STOTZKY

hibitory effects of C o n (293, 294). Soil microbes also appear to be capable of “adapting” to high COz and low 0 2 tensions (294). It has not been established whether this adaptation is physiological or represents only an enrichment of segments within the heterogeneous population already tolerant to these conditions, but it is also caused by elevated COz concentrations rather than by reductions in 02. High levels of COZ also induce morphogenetic responses in fungi (22, 293), and some organisms, other than conventional anaerobes, appear to be isolated from soil or, at least, show a characteristic morphology, only under altered atmospheres (e.g., NH3-rich atmospheres, 270, 271, 275). Because the atmospheres of other planets are apparently devoid of 0 2 and, in the case of Mars and Venus, high in COS, or of the Jovian planets, high in NH3 (252), caution should be exercised in interpreting the morphologies of potential extraterrestrial forms in comparison to those of terrestrial microbes. Some organisms are tolerant of and even capable of metabolizing under atmospheric conditions even more stringent and divergent from normal terrestrial atmospheres than the high COZ and low 0 2 tensions mentioned above (252,274).Survival and growth of microbes (106, 218, 270), seed germination and early seedling growth (272, 273), and invertebrate activity (269) occur in atmospheres devoid of 0 2 and containing appreciable amounts of nitrogen oxides, hydrocarbons, NH3, CO, CSz, and/or A, as well as under stress conditions such as desiccation, temperature extremes, high salinity, and low atmospheric pressures. The results of such studies, and the tolerance and adaptability of soil microorganisms to SO2 (>13,000 p.p.m.), HnS (>10,000 p.p.m.), and O3 (>200 p.p.h.m.) (292), demonstrate the high “plasticity” of terrestrial microorganisms to adverse atmospheric conditions, and suggest that the absence of O2 and the prevalance of other gases may not be important limiting factors for life elsewhere. Consequently, if photosynthesis is the primary source of extraterrestrial energy, then microorganisms, perhaps similar to terrestrial photosynthetic bacteria, may be the energy-fixing agents on planets devoid of 0 2 . Possible metabolic pathways occurring in such environments have been discussed by Vishniac (320, 321). Although the absence of O2 in the ambient atmospheres of other planets should not be a decisive factor for life, the possibility of 0 2 at depths below the surface should not be excluded. On Earth, aerobic bacteria have been found at depths as great as 2000 m., where atmospheric O2 is presumably absent but where 0 2 may be produced from

RELEVANCE OF SOL MICROBIOLOGY TO SEARCH FOR LIFE ON OTHER PLANETS

29

water at localized points b y radiations from radium and mesothorium (145).The importance of this phenomenon in terrestrial, much less in extraterrestrial, soils is not known. Should it occur, however, then an interesting inverse situation in the distribution of microbes in terrestrial and extraterrestrial soils would exist, as the known surface soil microbiota is predominantly aerobic and the percentage of anaerobes increases with depth (5, 44, 139, 251), presumably in response to decreases in the Oz/COz ratio. H. PH Microorganisms presumably metabolize in soil at pH values either higher or lower than those necessary for the same processes in vitro: e.g., nitrification can occur in soils having an overall acidity closer to pH 4 than to the minimum of approximately pH 6 required in liquid media containing glass beads (334), although it has been suggested that the optimum pH for oxidation of NOz- to NO3- is approximately 0.5 pH units higher in soil (190); maximal oxidation of various substrates by bacteria presumably adsorbed on anionic resins (used as model soil systems) occurs at p H values that are at least one unit higher than those required by nonsorbed cells (113, 114); enzymes adsorbed on clay minerals have a higher p H optimum than enzymes in solution (183, 187); bacteria grow at lower pH values in the presence than in the absence of certain clay minerals (289). The requirement for higher p H values for microbial activity in soil than in liquid culture has been interpreted as a reflection of the lower pH at the surface of negatively charged particles, due to the concentration of positively charged ions within the double layer (112, 166). The measured pH, however, reflects predominantly the ambient soil solution and not the pH near the particle surface, which may be 100 times more acid than that of the solution (110,187).In fact, it has been deduced that the pH of a hydrogen or an aluminum-saturated montmorillonite surface is lower than 0.8 (309). An increase in the pH of the entire system, therefore, only appears necessary, and the pH optima may not differ between sorbed and nonsorbed cells or enzymes. This interpretation, however, does not explain the apparent ability of microbes to metabolize at lower pH values in the presence of particles than in their absence (289,334),especially as mechanisms of sorption between microbes and inanimate particles have not been adequately elucidated (see 257). Clarification is needed of the apparent discrepancies in results and interpretations of pH optima for microbes in structured environments such as soil.

30

G . STOTZKY

Even on a gross scale, however, pH influences the predominance of one group of organisms over another: in acidic soils (below approximately pH 5.5),fungi will predominate; in near-neutral or moderately alkaline soils (between approximately pH 6.0 and 8.0), bacteria and actinomycetes will predominate, not because fungi are intolerant of these pH values, but because bacteria and actinomycetes compete more effectively. Exceptions to this general scheme, of course, exist: e.g., iron- and sulfur-oxidizing autotrophs are tolerant of extremely low pH values, and lactic acid bacteria are tolerant of less extreme, but still low, pH values (145); some soil actinomycetes require pH values as high as 8.5 for optimum growth and are more prevalent in alkaline than in acid soils (307, 308). The pH also influences microbial growth by altering the availability of carbonaceous substrates and mineral nutrients. At pH values below their isoelectric point, amphoteric substrates will be sorbed by negatively charged clay minerals, but whether sorption decreases substrate availability is still a matter of controversy, as is the relative effectiveness of sorbed antibiotics (see Section M). At low pH values, hydrogen ions predominate on the exchange complex and replaced cations may b e leached from the soil. The solubilities of Fe, Al, and Mn are also increased and may inhibit microbial development by direct toxicity or by removing phosphorus from solution. At high pH values, minor elements and phosphorus may become limiting, as a result of decreased solubility (251).

I.

Eh

The oxidation-reduction potential in soil is influenced by the solubility and oxidation state of some minerals (e.g., Fe, Mn) and by the O2 content. There is a sequential reduction of various oxidationreduction systems in soil; e.g., the E, (corrected to pH 7) at which reduction begins is +224 mV. for nitrate (220), +122 mV. for ferric iron (221),and -150 mV. for sulfate (57);manganic manganese is reduced at potentials slightly lower than those necessary for nitrate (14, 222). Limited research has been conducted on the influence of Eh on soil microorganisms, but there is little doubt that it affects their ecology, both directly [e.g., sulfate reducers require an Eh of at least -200 mV. for initiation of growth (239), and sulfide producers function best at potentials below -75 mV. ( l l l ) ] and by its effect on production of toxic materials (e.g., reduced ions, organic acids). Under conditions of reduced Eh, anaerobes and organisms capable of using oxidized forms of certain elements a s terminal electron acceptors predominate,

RELEVANCE OF SOIL MICROBIOLOGY TO SEARCH FOR LIFE O N OTHER PLANETS

31

whereas under conditions of high E,, more energy storage would occur in species capable of exploiting this environment. In terms of microbial tolerances, Vallentyne (316) has listed -450 mV. at pH 9.5 for sulfate-reducing bacteria as the lower limit and +850 mV. at pH 3 for iron bacteria as the upper limit. The oxidationreduction potential of extraterrestrial environments must obviously be considered when speculating about the energetics of life based on elements and metabolic schemes different from those on Earth (see 320, 321).

J. IONIC STRENGTH

The ionic strength of the soil solution is influenced, in part, by pH through its effect on mineral solubility. The primary effect of ionic strength on the soil microbiota is probably through its mediation of the electrokinetic potentials of microbial cells and inanimate soil particulates. This aspect is discussed in Section M . K. PRESSURE Although the solute concentration of normal soil solutions is usually not high enough [approximately 0.05%,with osmotic pressures ranging from 0.2 to 1 atm. (251)l to suggest osmotic inhibition of the soil microbiota, the concentration probably increases to inhibitory levels during periods of reduced moisture, especially in saline and alkaline soils. Furthermore, the concentration in the microhabitats is undoubtedly higher than that measured in the soil solution, as solutes concentrate at particle-water interfaces, especially at charged surfaces. Inhibition of microbial activity in soil by hypertonic osmotic pressures has been demonstrated (297, 298, 299), and organisms isolated from soil are sensitive to elevated solute concentrations (289, 291, 302). Neither the limits of this sensitivity nor the adaptability of the soil microbiota to high salt concentrations have been adequately defined. For example, the possible evolution of marine microorganisms from soil forms (118, 133, 172), and the relative salt-dependence and salt-resistance of microorganisms from saline and nonsaline soils (48, 84, 118, 137, 308) have not been resolved. The role of the soil environment must also b e defined, as certain clay minerals appear to protect microorganisms from the detrimental effects of hypertonic osmotic pressures (291, 302). Some speculations have recently been presented (126) on the salt content of possible liquid water on Mars. Even though extremes in pressure are not normally encountered in soils, the survival of bacterial and fungal spores under vacuums

32

G . STOTZKY

to 2 X ranging from 1 X mm Hg (238, see 126,274,306) and under hydrostatic pressures as high as 1500 atm. (140,172,342) indicates that terrestrial microorganisms are remarkably tolerant of extremes in pressure.

L. TEMPERATURE Changes in soil temperature influence both the species composition and the metabolic activity of the microbiota, but there are apparently no known natural areas on the Earth's surface, with the possible exception of the interiors of volcanoes, where microbial life is absent as a result of temperature extremes (e.g., 25, 314). Microorganisms are present in the Gobi Desert, where diurnal temperature fluctuations can attain 50°C. (253), and at the polar caps, where temperatures may remain as low as -23" to -40°C. for extended periods (201). The adaptation of microorganisms to extremes in temperature is further demonstrated by their presence in composts with temperatures as high as 60°C. (59), in hot springs where temperatures may reach 90°C. (38, 266), in deserts (47, 253), in frozen soils, and in mountain snowbanks (201,253). Populations of microorganisms survive extremely low temperatures, although their numbers are reduced: e.g., cells and spores of bacteria survived diurnal variations from -60" to 30°C. (116); numerous fungi grew in soil exposed to diurnal extremes of -94" to 23°C. (61); spores of bacteria and fungi survived a 2-hour exposure to temperatures as low as 0.0047" K. (24); fungal spores have been successfully germinated after storage at -196°C. (64,94);Escherichia coli bacteriophage T1, poliovirus type 111, and Penicillium roqueforti survived temperatures ranging between -75" and -45°C. in rocket and balloon flights (127). Freezing of microbes in soil also does not appear to be particularly detrimental, even though temperatures of -20" to -30°C. may persist for several months. In fact, Krasil'nikov (139) reported that Azotobacter and root-nodule bacteria (which also survived 1 month at -180°C.) and yeasts exhibited more vigorous metabolism and reproduction after a 3-week exposure to -15" to -20°C. Aerobes have been reported to be more tolerant to diurnal freeze-thaw cycles than organisms capable of growing in a COP atmosphere without O2 (see 126). The critical factors determining survival of microbial cells in low temperatures are apparently the rates at which cells are frozen and subsequently thawed (e.g., 64, 94). Most vegetative cells are killed by temperatures approaching 85°C. (although the thermal death point of some thermophilic bacteria may

RELEVANCE OF S O U MICROBIOLOGY TO SEARCH FOR LIFE ON OTHER PLANETS

33

be as high as 110" to l2O0C.),but spores can resist temperatures exceeding 100°C. for extended periods (see 145). Spores of Bacillus anthracis have been recorded to survive a 30-second exposure to 400°C. (210). In general, microorganisms are more resistant to elevated temperatures when dried, and to dry than to moist heat (see 145). Because of their apparent tolerances, temperature extremes alone should not restrict the persistence, nor perhaps even the development, of terrestrial microbes in many extraterrestrial habitats.

M. PARTICULATES The environmental factor that primarily distinguishes soil from other microbial habitats is the predominance of a solid phase. Even though microbes are restricted physiologically to the liquid phase, they are influenced by the physicochemical properties of solids, especially by clay minerals and particulate organic matter. Because of their large particle size and relatively inert surfaces, sand- and siltsize particles do not long retain water films, and thereby do not maintain a permanent microbial population, nor are these particles efficient concentrating surfaces for nutrients. The permanence of organic particles is limited by their decomposition rate. Consequently, most microbial development in soil is probably associated with the clay mineral fraction. This fraction, which occurs primarily in aggregates or as coatings on larger particles (34), is characterized by high surface activity resulting from its large surface area and its chemical and mineralogical composition. Clay minerals vary greatly in their physicochemical characteristics (e.g., structure, specific surface, cation exchange capacity, surface charge density, water retention, swelling capacity) (see 105), and these differences must be considered when evaluating the many contradictory reports on the effect of clays on microbial ecology and activity. Clay minerals have been reported to influence microbial and enzymic transformation of a variety of substances, ranging from nitrogen and sulfur to carbohydrates, proteinaceous materials, organic phosphorus compounds, phenolics, vitamins, antibiotics, and complex plant residues. They have also been variously claimed to increase or decrease multiplication of cells; increase, decrease, or not affect microbial growth and activity; increase fixation of atmospheric nitrogen; and differentially sorb and inactivate antibiotics, enzymes, viruses, and intact cells (see below for details). The type, concentration, particle size, and cation saturation of

34

G. STOTZKY

clays, as well as the ambient environment, will influence their effect on microorganisms (255, 290). Microbial cells and clay minerals have a net negative charge at the pH of normal soils (256, 257), and sorption between these populations occurs in nitro only at p H values below the isoelectric point of the cells, or in the presence of polyvalent cations which either decrease the electrokinetic potentials or result in a differential charge reversal of the particles (255, 256, 257). Similar relationships appear to be involved in the adsorption of microbial metabolites b y clays (258). The contradiction between the apparent sorption of microbial cells in uiuo (e.g., removal of microbes from waters in percolation beds; failure to wash substantial numbers of microorganisms from soil in perfusion experiments) and the absence of sorption in d t r o , except under the conditions mentioned, can be reconciled by at least two possibilities: one is that microbes are retained by entrapment in narrow channels between particles and/or by surface tension, rather than by direct electrical attraction; the other is based on recent observations that the isoelectric point of some bacteria is dependent on the type of cations present in the ambient environment (256). For example, the isoelectric point of several bacteria in the presence of low concentrations (ionic strength = 3 x low4)of mono- or divalent chloride salts is pH 2.5 to 3.5, whereas, in the presence of La+++or Cr+++,it is approximately p H 5.0, and in the presence of Fe+++or Al+++,pH 7.0. The isoelectric points are also shifted to higher pH values as the cation concentrations increase. As polyvalent cations are present in the soil solution, microbes may be positively charged or, at least, have a lower net negative charge at normal soil pH values than at comparable p H values in uitro, which would facilitate sorption with negatively charged clays (256, 257). Clay minerals also appear to exert an influence on the ecology of microbes in soil. The rate of spread of Fusarium wilt of banana (295) and the geographic distribution of Histoplasma capsulatum in soils (301) has been correlated with the presence or absence of a specific clay mineral. This clay, an expanding three-layer silicate which tends toward montmorillonite in the montmorillonite-vermiculite sequence, also influences rates of growth of microbes through soil (288). In general, fungi grow faster through soils not containing this clay mineral, whereas bacteria grow faster and antagonistic effects between bacteria and fungi appear to be greater in soils containing this clay (291, 300). Autotrophic (nitrification) and heterotrophic (substrate decomposition) activities are also altered by adding various types of clay minerals to soil (176, 291).

RELEVANCE OF S O L MICROBIOLOGY TO SEARCH FOR LIFE ON OTHER PLANETS

35

In pure culture, montmorillonite markedly stimulates the respiration of bacteria, whereas other clay minerals have relatively little influence (289, 302). This stimulation is partly due to the ability of some clays to maintain the pH of the environment suitable for sustained growth and is related to their cation exchange capacity, possibly specific surface, but not particle size distribution (290). Montmorillonite also appears to protect bacteria against hypertonic osmotic pressures (289, 291, 302). Fungal respiration and spore germination are not significantly influenced by clays at low concentrations, but are markedly reduced at high concentrations, especially by montmorillonite, apparently as a result of impaired gas exchange caused by increased viscosities (259, 291, 303). Clays also influence the transformations of nitrogen (3, 92, 129, 150, 176, 193, 244, 291) and sulfur

(91, 283, 322). These pronounced differential effects of clay minerals on microorganisms indicate that they are important determinants of the activity, ecology, and population dynamics of microbes in soil. As most microbial activity probably occurs in water films adjacent to clay surfaces, for reasons already discussed, the types of clays present would be expected to influence associated microorganisms. One prerequisite for growth and development of an organism in a mixed population is its establishment as a member of the population. This requires either adaptability of the organism or an alteration in the indigenous population which enables successful competition by the organism. Most organisms are capable of some degree of adaptation, but this process is relatively slow and haphazard in contrast to population changes resulting from environmental stresses. If the environment does not mediate these stresses, portions of the indigenous population may be suppressed and an organism not a member of this population, but tolerant to the prevailing conditions, may become established. For example, if the intruding organism is a fungus, and the environment does not mediate reductions in pH, inhibition of populations intolerant to low p H may enable the fungus to become established before the pH returns to a range favorable for resumption of growth by the inhibited populations. Such a relationship may, in part, be responsible for the greater rate of spread of Fusarium wilt of banana in, and the apparent association of H.capsulatum with, soils not containing clay minerals of the montmorillonite type (295, 301). Although the implications of clay minerals in the development of microbes in soil, in potential interplanetary contamination, and as possible templates for abiological synthesis are obvious from such

36

G . STOTZKY

reasoning, the validity of the reasoning and the extent of the implications are dependent upon reconciliation of numerous apparent contradictions. Do clay minerals in soil (a) decrease or enhance the decomposition of proteins (9, 76, 77, 81, 169, 181, 185, 186, 231, 233, 310; see 80, 184), of carbohydrates (169, 170, 171; see 80), of nucleic acids and other organic phosphorus compounds (95,198),of phenolics (79), of vitamins (262),and of other organic materials (9, 82, 169; see 16, 31, 32)? (b) inhibit the activity of enzymes (79, 81, 181, 182, 186, 198, 232, 279; see 37, 74, 80, 122, 278, 337)? (c) sorb with microbial cells (225,243,256,257,279)?(d) increase, decrease, or not affect the multiplication and metabolic activity of cells (56, 75, 113, 114, 134, 144, 175, 177, 180, 225, 259, 289, 290, 302, 303, 341, 343)? (e)adsorb and inactivate antibiotics (35, 96, 120, 130, 141, 142, 178, 234, 235, 236,241,242,276,277,281,325;see 139,240)? (f) serve as reservoirs for viruses, which enables them to persist in the absence of a host (109, 117, 196)? Clarification of these and other questions require critical experiments and detailed reporting of the variables involved: e.g., the types of clay, organic compounds, or microbial cells; the pH and ionic status; the mechanisms and sites of adsorption (external or intermicellar); if sorbed intermicellarly, the orientation of the organic (vertical or horizontal to its major axis); the methods for measuring the extent of sorption; the treatment of the clay-organic complex prior to its exposure to microbial or enzymic degradation. Ill. Methods for Detecting Life in Terrestrial Soils There are basically two approaches for detecting and studying microorganisms in soil: the indirect and the more indirect approach. Unfortunately, the development of direct techniques, whereby microorganisms can be observed for extended periods in soil, has been slow. Direct microscopic observations are severely restricted by the opacity of soil; by the similarity in size, color, and shape of microorganisms and many soil particulates; by limitations in the resolving power, depth of focus, and working distance at the magnifications required; by common retention of many nonspecific dyes. Attempts have been made to overcome these limitations, but no technique has yet resolved them all. Soil sectioning methods (e.g., 4, 45, 119, 143) demonstrate the spatial distribution of microbes in natural soil, but, because the soil is removed and usually hardened with an impregnating resin, it is neither possible to observe the growth and distribution of organisms over a period of time nor to study the structure of cells, soil particles,


37

and their possible complexes. Fluorescent antibody techniques (78, 121,261) provide some specificity to staining methods, but the spatial integrity of the soil is disturbed and continuous observations are not possible. Furthermore, only one, or few, marker species can be used concomitantly, and the specificity of the staining reaction is usually not high. Although the visualization of bacteria in the presence of soil particles by infrared photography (49a) should facilitate certain types of ecological studies of bacteria in soil, the soil must be removed for observation, thereby also precluding continuous observations. The numerous modifications of immersion methods (52, 53, 200, 267, 312, 313, 323; see 66, 333), while providing some information about the type of organisms present, their relative location, and their rate and nature of colonization of the immersed object, are selective and have many limitations in common with those of other methods (see 288). Dissection of soil clods (66,98), while permitting examination of organisms in situ,is tedious and also not adapted to continual observations. Electron microscopy (131, 204, 208), although yielding much information on the structure of soil microbes, is limited by the need to remove the soil sample for examination and by the artifacts produced during preparation and examination. Even the scanning electron microscope (97), which has provided exciting pictures, with a three-dimensional effect, of microorganisms on sand and humus particles, is plagued by these problems. Incident light microscopy, coupled with “dipping cones,” overcomes some of the difficulties in opacity and in the working distance between the soil surface and the objective. However, the limited depth of focus, the necessity for immersion oil or similar fluid at adequate magnifications, and the difficulty in distinguishing between inert particles and microbial cells [which can apparently be reduced by immunofluorescent techniques (98)], still limit the adaptability of this technique to continuous observations. The application of aligned fiber optics is restricted by similar problems. Although slender probes (e.g., 19 gauge) have been successfully used to study circulation in muscle, joints, and brain (158, 164, 165a, 192a), their need to be fluid-coupled, their relatively low resolution (5-10 p ) , field of view, and depth of focus, and the high contrast required between particles and background, indicate that the art of fiber optics is not yet sufficiently advanced for use in soils. Nevertheless, because the insertion of slender probes would minimize soil disturbance and might permit continuous observations on specific

38

G . STOTZKY

soil loci, the applicability of this potentially invaluable tool for studying microbes in situ should be further investigated (291). Because the microbiologist cannot meander, like Alice in Wonderland, through soil to observe the workings of the microorganisms, indirect approaches will probably continue to be used in answering questions such as: Where are microbes located in soil? What do microbes look like in soil (are they protoplasts)? Do they form colonies? How do they move through soil? Similarly, indirect approaches should be initially more rewarding and easier than microscopic methods in the exploration for extraterrestrial life (147, 153,

154). The various dilution techniques (see 54) provide some information on types and numbers of microorganisms in soil, but not on their location and distribution. In the case of fungi, they are limited essentially to those species producing abundant spores (331, 332, 333). Similar limitations are inherent in methods based on incubating small quantities of soil directly with a nutrient medium (324, 329), although these methods, as do those which laboriously isolate hyphae directly from soil (330),detect nonsporulating or sparsely sporulating fungi. The main limitation in dilution plating methods is that they permit the development of only certain microorganisms, presumably because only these can develop on media in the laboratory environment or they prevent the development of slower growing and more fastidious forms. This limitation has resulted in the widely held view that organisms isolated by these methods constitute the major microbiota of soils. Recent descriptions of unusual microbial forms in soil (49, 50, 202, 203, 204, 208, 215, 216) are aiding in reversing this narrow view and may provide explanations for the long-observed discrepancies between the activities of microbes in soil and the numbers isolated (115, 132, 297, 317). The most widely used methods for studying microbes in soil are based on their metabolic activities. Of these methods, respiratory measurements appear to be favored for the same reasons that they are being considered for extraterrestrial life detection (153):namely, they are indicative of gross metabolic activity and not of specific microorganisms or biochemical pathways. Respiratory measurements are well correlated with other indices of microbial activity, such as nitrogen or phosphorus transformations, metabolic intermediates, pH, organic matter content, changes in soil weight (see 287). Maximum rates of respiration, however, usually precede by several days to


39

weeks the maximum numbers of microorganisms isolated (115, 132, 297, 317), suggesting that respiratory rates reflect the metabolic activity rather than numbers or growth of at least that portion of the soil microbiota enumerated by conventional dilution methods. Either C o n evolution or O2 uptake can be measured. The former, however, has wider application, as C o n is evolved from anaerobic as well as aerobic environments, and use of I4C-labeled substrates greatly enhances sensitivity, an obvious advantage for extraterrestrial application (153). Recent reviews (67, 287) have described the many techniques, with their advantages and disadvantages, for measuring soil respiration. Although respiration is usually indicative of a metabolizing microbiota, some of the COO evolved from soil may have nonbiological origins. Carbon dioxide may be produced by chemical decarboxylation (43), by cell-free, heat-stable enzymes (43), by the action on carbonates of added chemicals or organic acids produced during metabolism (51). Oxygen may be taken up during chemical oxidation (43), and both COZ and 0 2 may be adsorbed on soil and water interfaces (250). Even soil “sterilized” by ionizing radiation still “respires” (229), although the absence of growth on dilution plates does not constitute unequivocal proof that the soil is devoid of living organisms. The nonbiological production of C 0 2must, of course be considered when using this technique for life detection. The observation of a respiratory peak or a growth curve (see 287), especially following the addition of substrate, must be a requisite for presuming the presence of life. The primary limitation of respiratory methods for detection of extraterrestrial life is the presumption that terrestrial substrates will be metabolized. Microbial activity can also be evaluated by following the disappearance of substrates; appearance of intermediary metabolites; changes in pH; mineralization, immobilization, or volatilization of mineral nutrients; changes in numbers of specific microbial groups; heat production; etc. Methods for measuring many of these changes in soil have recently been compiled (29). Perfusing soils with specific substrates has provided much information about microbes which may be present in relatively low numbers or whose metabolic activities are unique (13, 102, 148, 177, 244). Adaptation of this technique to extraterrestrial environments may be indicated, should the density of potential life forms be low or their metabolism different from that of terrestrial forms, but not to the ex-

40

G . STOTZKY

tent that enzymes capable of using terrestrial substrates cannot be induced. The density of potential life forms and their rate of growth in extraterrestrial soils could be evaluated by adaptation of the replica plating technique (288). Considerable information on the influence of various chemical, physical, and mineralogical characteristics on the growth and interactions of microbes in soils has been obtained with this method (291, 300), which can be coupled with fluorescent staining (78, 121, 261) and autoradiological (39) techniques. Particle size analysis may also have application in the detection of extraterrestrial life. For example, the swelling of fungal spores can be critically and easily measured with an electrical sensing zone particle analyzer, even in the presence of large quantities of other particulates (255, 259). Active uptake of water by viable propagules could be distinguished from the swelling of hydrophilic inanimates by incorporating appropriate controls, a potential problem also in using turbidimetric measurements as an index of extraterrestrial microbial growth (42). Cell-free enzymes have been used as indicators of soil microorganisms (see 37, 74, 124, 237, 278), as enzymic activity in soil appears to be correlated with the concurrent or immediately past activity of intact cells. The presence of cell-free enzymes is usually deduced from the conversion of simple substrates in soil, after organisms capable of developing on various media have been killed by chemicals (23, 69, 123; see 278) or ionizing radiation (189; see 278). Few enzymes, however, have actually been isolated from soil and purified (36, 179). A wide spectrum of enzymes has been detected (e.g., 37,278), but, with the exception of some transaminases, each is restricted to a single-step catalysis involving only a single substrate. This apparent restriction is logical, as sequential enzymic activity requires a complex of enzymes, with accessories, in a spatial arrangement to enable the transfer of metabolic intermediates and energy from one enzyme system to another. Consequently, the activity of cell-free transaminases in soil is surprising, as these require concomitant spatial accessibility to two substrates, and, even more so, the suggestion that initial oxidation of glucose to C o t in soil is accomplished by cell-free enzymes (68).Although cell-free enzymes appear to be associated with, and perhaps adsorbed on, clays and organic matter (see 37, 74, 122, 237, 278), the enzyme activity in (presumably) sterilized soils may occur within killed cells (229),


41

possibly explaining the above activity of transaminases and the oxidation of glucose. In view of the many contradictions that exist in the literature (see 278), this and other questions relating to the mechanisms of sorption of enzymes, the activity of sorbed enzymes (e.g., are functional groups involved in enzymic activity also involved in sorption?), the persistence of enzymes in soil (e.g., whether and how they are protected from microbial degradation and yet still function), etc., require clarification. Nevertheless, enzymic assays for detection of extraterrestrial life have at least two aspects to recommend them: assays for specific enzymes are highly sensitive [McLaren and Peterson (188) quote a test that is sensitive at the one-enzyme-molecule level], and most are easy to conduct [e.g., fluorescent assay for phosphatase in the Multivator (156)l. A positive response by itself, however, is only presumptive evidence of life, as enzymes may persist in the absence of any current organisms. Nonetheless, such a response from extraterrestrial environments would be exciting, as it would stimulate further investigations and speculations about panspermia and the historicity of life on other planets. A negative response would be less exciting, as it would not be possible to distinguish between the absence of life and the absence of life with biochemical similarity to terrestrial forms. Similar conclusions would be made from positive or negative responses to the luciferin-luciferase assay for ATP (152) and to tests based on uptake or exchange of isotopes (138, 151, 155,

173). Clearly, devices for detection of extraterrestrial life must be multipronged and utilize tests at all levels of molecular complexity. Such thinking forms the background for various current life-detection packages (see 42,245).

IV. Speculations on Numbers a n d Types of Soil Microorganisms Although soil microbiology has been an active discipline for approximately 100 years, it still shares one major problem with the fledgling discipline of extraterrestrial biology: namely, the numbers and types of microorganisms in soils have not yet been definitively assessed. The literature is replete with numbers of bacteria, fungi, and other members of the microbiota in various soils, at various times of the year, under various plant covers, etc. These numbers have been

42

G. STOTZKY

obtained from both dilution techniques and from microscopic counts of soil smears, contact slides, thin sections, etc. The former methods normally yield about 1 to 10% ( lo6to lo8microbes/gm. oven-dry soil) of the numbers determined by microscopy ( lo8 to 10’O microbes/gm. oven-dry soil). These differences are usually ascribed to inability to distinguish accurately between living and dead cells [although this can be rectified somewhat by judicious interpretation of acridine orange-ultraviolet fluorescence microscopy (265, 304)] and to the failure of many soil microbes to grow on laboratory media, either because of nutritional fastidiousness, dormancy, inhibition by neighboring cells, or differences in physicochemical properties between laboratory and soil environments. Regardless of the reasons, there was little doubt, until recently, that either method, or both together, provided a reasonable picture of the composition of the soil microbiota. These assumptions are being shattered by demonstrations that the soil harbors not only a vaster microbiota, but also one that contains morphological (and probably physiological) forms that differ greatly from those usually observed on dilution plates or directly with light microscopy. The discovery of Bdellovibrio bacteriovorus (284, 285) was only the beginning in this revolution. Casida (49) isolated a catalase-negative, microaerophilic, coccoid microorganism from soil at dilutions of los and suggested that this organism, which may have some taxonomic relationships to the families Actinomycetaceae and Mycobacteriaceae, comprises the majority of the coccoid forms observed in soil. This organism, which is present in greater numbers than the entire soil microbiota normally isolated, is presumably excluded from soil dilution plates by less fastidious and faster growing bacteria. Casida and Wood (50) have also isolated a gram-negative Bacillus sp. from soil irradiated with 5 megarads from %o, which appears as a long, slender, tightly coiled cell bearing no morphological relationship to known bacteria on some synthetic media, but which, on an agarized-soil medium exhibits a palisade arrangement of cells with well-developed endospores similar to

Bacillus circulans. Bystricky and co-workers (202,203,215,216)have found an unusual form in soils incubated with a carbon-free medium. This organism, which is of the general size of bacteria, but strongly contoured with longitudinal rows of spherical subunits, has been observed in soils from various parts of the world (215, 216). Tentatively named “helicoidal polyspheroid,” this organism does not appear to be an artifact of the electron microscopy procedures used nor of the medium, as

RELEVANCE OF SOIL MICROBlOLOGY TO SEARCH FOR LIFE ON OTHER PLANETS

43

identical handling of “normal” soil bacteria has not yielded similar shapes (214). “Unusual” microbes in soil have also been observed in the laboratories of Nikitin, Aristovskaya, and Perfil’ev. The latter has developed a technique of preparing rectangular capillaries with small (3-5 p ) and variously shaped cross sections, which, after being inserted for various periods of time in soils, muds, or waters, are examined microscopically (226, 227; see 85). Because the capillaries, which become filled with the ambient solution, presumably simulate the surrounding nutritional and physicochemical environment (assuming similar restrictions in gas exchange), the organisms which develop presumably reflect the dominant forms present in that environment. With these techniques, Perfil’ev has discovered numerous new and bizarre-looking microorganisms, some of which, especially in aquatic sediments, appear to be primary agents in the precipitation and accumulation of iron and manganese deposits. Others, some of which form special appendages, are predators on other microbes. Aristovskaya (10, 11) has also applied these techniques to soil and has demonstrated the occurrence of hitherto unknown organisms. Nikitin (204,208)has uncovered a wide spectrum of new organisms by examining soil suspensions with the electron microscope. These organisms, some of which are similar in size to common bacteria and some which approach 0.1 p, have morphologies unlike those of microbes usually isolated from soil. These forms, which do not grow on most laboratory media, have been classified as bacterialike (forms with unusual appendages; threadlike with stalks), protozoalike (starlike; organisms with pores; rod-shaped with spherical convex protrusions), algaelike (small diatoms and algaelike), and viruslike (208). Many have long appendages, some the size of pili but some considerably thicker, which suggests that they may act as holdfasts to secure these organisms to soil particles. These forms presumably comprise 80 to 85% of the microbiota of podzols, which Nikitin (205) has calculated exceeds 2 X lo9 viable microbes/gm. soil. The number of these unusual forms is positively correlated with the “fulvic acid” fraction, as extracted by Nikitin et al. (206), and is inversely related to the number of usual forms. Fulvic acid, as well as agarized soil, serves as a laboratory medium for growth of some of these unusual forms. One of the organisms observed, and recently isolated in pure culture (207), appears identical in morphology to the “helicoidal polyspheroid” observed by Bystricky et al. (202, 203). Another organism

44

G . STOTZKY

is reminiscent of the Kakabekia-like forms isolated from soil under ammonia-rich atmospheres by Siege1 et al. (271,275), and which they suggest are relicts of Kakabekia umbellata Barghoorn, the 1 9 0 W O O X 10s-year-old Precambrian microfossil from the Gunflint formation (20). Other forms also bear a similarity to described fossil microorganisms. In view of the apparent ubiquity and density of the many novel and pleomorphic microbial forms now being discovered in soil, critical evaluation is necessary of the assumption that, because the morphology of some microorganisms bears no relationship to hitherto known microbial species, these are prehistoric living organisms. This precaution will become even more relevant if and when extraterrestrial forms are observed, especially in relation to examining returning space voyagers for potential contaminants. The numbers and diversity of terrestrial microbes are only now beginning to be plumbed, but this inventory should be as complete as possible before we can be sincerely surprised by possible strange morphologies and physiologies of potential extraterrestrial inhabitants. V. Conclusions

Soil microbiology shares several points of relevance with the search for life on other planets. The primary point is that, although a considerable amount is known about terrestrial soil microbiology, recent investigations are illuminating how much is not known; e.g., the limits of the soil’s capability to decompose organic molecules to their theoretical end products; the multitude and diversity, both in morphology and function, of the soil microbiota; the shapes of microbes in soil; where or how microbes metabolize in soil; how the physicochemical factors in soil interact and how they influence microbes in the microenvironment. On the positive side, perhaps enough is known to state what is not known and to formulate intelligent questions, if not always intelligent experiments. It is also becoming increasingly apparent that soil microbiological questions must be framed in the context of the soil. To paraphrase Wald’s statement that “a physicist is the atom’s way of knowing about atoms” (327), the microbiologist is the microbe’s way of knowing about microbes -but, not necessarily about soil microbes. Attainment of this knowledge requires a soil microbiologist, with equal emphasis on both parts of the binomial. What polynomial epithets will someday be necessary to describe the interlocutors of extraterrestrial organisms?

RELEVANCE OF SOIL MICROBIOLOGY T O SEARCH FOR LIFE ON OTHER PLANETS

45

REFERENCES 1. Ahmadjian, V. (1966). In “Symbiosis” (S. M. Henry, ed.), pp. 36-98. Academic Press, New York. 2. Ajello, L. (1967).Bacteriol. Reu.31,6-24. 3. Albrecht, W. A., and McCalla, T. M. (1937). Soil Sci. Soc. Am. Proc. 2, 263-267. 4. Alexander, F. E. S., and Jackson, R.M. (1954).Nature 174,750. 5. Alexander, M. (1961).“Introduction to Soil Microbiology,” 472 pp. Wiley, New York. 6. Alexander, M. (1964).Ann. Reu. Microbiol. 18,219-252. 7. Alexander, M. (1965).Aduan. Appl. Mtcrobiol. 7,35-80. 8. Alexander, M. (1965).Soil Sct. Soc. Am. Proc. 29,l-7. 9. Allison, F. E., Sherman, M. S., and Pinck, L. A. (1949).Soil Sct. 68,463-478. 10. Aristovskaya, T. V. (1963). In “Microorganisms in Agriculture” pp. 395-405. Publishing House of Moscow Univ., Moscow. (In Russ.) 13. Audus, L. H. (1951).Plant Soil 3,170-191. 14. Baas-Becking, L. G. M., Kaplan, I. R., and Moore, I. (1960).J . Geol. 68,243-284. 15. Bail, 0. (1929).Z. Immunitaetsforsch. 60,l-14. 16. Bailey, G. W., and White, J. L. (1964)J. Agr. Food Chem. 12,324-332. 17. Baker, K. F.,and Snyder, W. C., eds. (1965).“Ecology of Soil-Borne Plant Pathogens,” 571 pp. Univ. of California Press, Berkeley, California. 18. Barghoorn, E. S., and Schopf, J. W. (1965).Science 150,337-339. 19. Barghoorn, E. S., and Schopf, J. W. (1966).Science 152,758-763. 20. Barghoorn, E. S., and Tyler, S.A. (1965).Science 147,563-577. 21. Bartholomew, W. V., and Norman, A. G. (1946). Soil Sci. Soc. Am. Proc. 11,270279. 22. Bartnicki-Garcia, S.(1963).Bactertol. Reu. 27,293-304. 23. Beck, T., and Poschenrieder, H. (1963).Plant Soil 18,346-357. 24. Becquerel, P. (1950).Compt. Rend. 231,1392-1394. 25. Benoit, R. E., and Cameron, R. E. (1967).Bacteriol. Proc. A13. 26. Bernal, J. D. (1951). “The Physical Basis of Life,” 80 pp. Routledge & Kegan Paul, London. 27. Bhaumik, H. D., and Clark, F. E. (1947).Soil Sct. Soc. Am. Proc. 12,234-238. 28. Birch, H. F. (1958).Plant Soil 10,9-31. 29. Black, C. A., Evans, D. D., White, J. L., Ensminger, L. E., and Clark, F. E., eds. (1965). “Methods of Soil Analyses,” 1634 pp. Am. SOC. Agron., Madison, Wisconsin. 30. Boyle, J. R., and Voigt, G. K. (1967).Science 155,193-195. 31. Bremner, J. M. (1951).J. Soil Sci. 2,67-82. 32. Bremner, J. M. (1954).J. Soil Sci. 5,214-232. 33. Bremner, J. M. (1965).In “Soil Nitrogen” (W. V. Bartholomew and F. E. Clark, eds.), pp. 93-149. Am. SOC.Agron., Madison, Wisconsin. 34. Brewer, R. (1964). “Fabric and Mineral Analysis of Soils,” 470 pp. Wiley, New York. 35. Brian, P. W. (1957). In “Microbial Ecology” (R. E. 0. Williams and C. C. Spicer, eds.), pp. 168-188. Cambridge Univ. Press, London and New York. 36. Briggs, M. H., and Segal, L. (1963).Life Sct. 2,69-72. 37. Briggs, M. H., and Spedding, D. J. (1963).Sci. Progr. (London) 51,217-225. 38. Brock, T. D. (1967).Nature 214,882-885. 39. Brock, T. D., and Brock, M. L. (1966).Nature 209,734-736.

46

G . STOTZKY

40. Bromfield, S . M. (1960).AustralianJ.Agr. Res. 11,304-316. 41. Bromfield, S. M., and Skerman, V. B. D. (1950).Soil Sci. 69,337-348. 42. Bruch, C. W. (1966). I n “Biology and the Exploration of Mars” (C. S. Pittendrigh, W. Vishniac, and J. P. T. Pearman, eds.), pp. 487-502. Natl. Acad. Sci., Natl. Res. Council, Washington, D. C. 43. Bunt, J. S., and Rovira, A. D. (1955).J. Soil Sci. 6,129-136. 44. Burges, A. (1958).“Micro-organisms in the Soil,” 188 pp. Hutchinson, London. 45. Burges, A., and Nicholas, D. P. (1961).Soil Sci. 92,25-29. 46. Burges, A., and Raw, F., eds. (1967).“Soil Biology,” 532 pp. Academic Press, New York. 47. Cameron, R. E. (1966). In “Biology and the Exploration of Mars” (C. S. Pittendrigh, W. Vishniac, and J. P. T. Pearman, eds.), pp. 164-186. Natl. Acad. Sci., Natl. Res. Council, Washington, D. C. 48. Carlucci, A. F., and Pramer, D. (1959).Appl. Microbiol. 7,388-392. 49. Casida, L. E., Jr. (1965).Appl.Microbiol. 13.327-334. 49a. Casida, L. E., Jr. (1968).Science 159,199-200. 50. Casida, L. E., Jr., and Wood, R.T. (1967).Bacteriol. Proc.Al2. 51. Chase, F. E., and Gray, P. H. H. (1957).Can.J. Microbiol. 3,335-349. 52. Chesters, C. G. A. (1941).Brit. Mycol. Soc. Trans. 24,352-355. 53. Cholodny, N. (1930).Arch. Mikrobiol. 1,620-652. 54. Clark, F. E. (1965). In “Methods of Soil Analysis” (C. A. Black, D. D. Evans, J. L. White, L. E. Ensminger, and F. E. Clark, eds.), pp. 1460-1466. Am. SOC.Agron., Madison, Wisconsin. 55. Cloud, P. E., Jr. (1965).Science 148,27-35. 56. Conn, H. J., andConn, J. E. (1940)../.Bacteriol. 39,99-100. 57. Connell, W. E., and Patrick, W. H., Jr. (1966).Agron.Abstr.p. 68. 58. Contois, D. E. (1959).]. Gen. Microbiol. 21,40-50. 59. Cooney, D. G., and Emerson, R. (1964). “Thermophilic Fungi,” 188 pp. Freeman, San Francisco, California. 60. Cosgrove, D. J. (1967).In “Soil Biochemistry” (A. D. McLaren and G. H. Peterson, eds.), pp. 216-228. Dekker, New York. 61. Curtis, C. R. (1967).Nuture213,738-739. 62. Davey, C. B., and Miller, R.J. (1966).Nature209,638. 63. Davey, C. B., Miller, R. J., and Nelson, L. A. (1966).J. Bacteriol. 91,1827-1830. 64. Davis, E. E., Hodges, F. A., and Goos. R. D. (1966). Phytopathology 56, 14321433. 65. Dobbs, C. G., and Hinson, W. H. (1953).Nature 172,197-199. 66. Dobbs, C. G . , and Hinson, W. H. (1960).In “The Ecology of Soil Fungi” (D. Parkinson and J . s. Waid, eds.), pp. 33-43. Liverpool Univ. Press, Liverpool. 67. Domsch, K. H. (1962).Zentr. Bukteriol. Parusitenk.,Abt. IZ 116,33-78. 68. Drobnik, J. (1960a).Plant Soil 12,199-211. 69. Drobnik, J. (1960b).PZantSoil 14,94-95. 70. Drost-Hansen, W. (1956).Naturwissenschaften 43,512. 71. Drost-Hansen, W. (1965).Ann. N. Y. Acad. Sci. 125,471-501. 72. Duff, R.B., Webley, D. M., and Scott, R. D. (1963). Soil Sci. 95,105-114. 73. Dumbrowski, H. (1963).Ann. N . Y. Acad. Sci. 108,453-460. 74. Durand, G. (1965).Reu. Ecol. Biol. Sol. 2,141-205. 75. Ellinger, G.. and Quastel, J. H. (1948).Blochern.J. 42,214-218. 76. Ensminger, L. E., and Gieseking, J. E. (1939).Soil Sci. 48,467-471.

RELEVANCE OF S O U MICROBIOLOGY TO SEARCH FOR LIFE ON OTHER PLANETS

47

Ensminger, L. E., and Gieseking, J. E. (1942).Soil Sci. 53,205-209. Eren, J., and Pramer, D. (1966). Soil Sci. 101,39-45. Erickson, A. E. (1948). Ph.D. Thesis, Univ. of Illinois, Urbana, Illinois. Estermann, E. F., and McLaren, A. D. (1959).]. Soil Sci. 10,64-78. Estermann, E. F., Peterson, G. H., and McLaren, A. D. (1959). Soil Sci. SOC. Am. Proc. 23,31-36. 82. Evans, L. T., and Russell, E. W. (1959).]. Soil Sci. 10,119-132. 83. Falk, M., and Kell, G. S. (1966).Science 154,1013-1014. 84. Flannery, W. L. (1956). Bacteriol. Reo. 20,49-66. 85. Foster, J . W. (1964).BioScience 14,15-20. 86. Fox, S. W., and Yuyama, S. (1963).Ann. N . Y. Acad. Sci. 108,487-494. 87. Freney, J. R. (1967). In “Soil Biochemistry” (A. D. McLaren and G. H. Peterson, eds.), pp. 229-259. Dekker, New York. 88. Garrett, S. D. (1956). “Biology of Root-Infecting Fungi,” 293 pp. Cambridge Univ. Press, London and New York. 89. Glathe, H., Knoll, K. H., and Makawi, A. A. M. (1963a). Z. Pflanzenernaehr. Dueng. Bodenk. 100,142-150. 90.Glathe, H., Knoll, K. H., and Makawi, A. A. M. (196313). Z. Pflanzenernaehr. Dueng. Bodenk. 100,224-233. 91. Gleen, H., and Quastel, J . H. (1953).Appl. Microbiol. 1,70-77. 92. Goldberg, S. S., and Gainey, P. L. (1955).Soil Sci. 80,43-53. 93. Gooding, T. H., and McCalla, T. M. (1946).Soil Sci. Soc. Am. Proc. 10, 185-190. 94. Goos, R. D., Davis, E. E., and Butterfield, W. (1967).Mycologia 59,58-66. 95. Goring, C. A. I., and Bartholomew, W. V. (1952). Soil Sci. 74,149-164. 96. Gottlieb, D., and Siminoff, P. (1952).Phytopathology 42,91-97. 97. Gray, T. R. G. (1967). Science 155,1668-1670. 98. Gray, T. R. G., and Parkinson, D., eds. (1968). “The Ecology of Soil Bacteria,” Liverpool Univ. Press, Liverpool (in press). 99. Gray, T. R. G., Baxby, P., Hill, I. R., and Goodfellow, M. (1968). In “The Ecology of Soil Bacteria” (T. R. G. Gray and D. Parkinson, eds.), Liverpool Univ. Press, Liverpool (in press). 100. Greenwood, D. J . (1961).Plant Soil 14,360-376. 101. Greenwood, D. J. (1962).Plant Soil 17,365-391. 102. Greenwood, D. J., and Lees, H. (1959).Plant Soil 11,87-92. 103. Greenwood, D. J., and Lees, H. (1960).Plant Soil 12,69-80. 104. Griffin, D. M. (1963).Biol. Reo. Cambridge Phil. Soc. 38,141-166. 105. Grim, R. E. (1953). “Clay Mineralogy,” 384 pp. McCraw-Hill, New York. 106. Hagen, C. A., Hawrylewicz, E. J., and Ehrlich, R. (1964). Appl. Microbiol. 12, 2 15-218. 107. Haldane, J. B. S. (1954). In “New Biology” (M. L. Johnson, M. Abercrombie, and G. E. Foggs, eds.), Vol. 16, pp. 12-27. Penguin, London. 108. Halpern, B., Westley, J . W., Anderson, P. J., and Lederberg, J. (1966). Anal. Biochem. 17,179-182. 109. Harrison, B. D. (1965).Nature 178,553. 110. Harter, R. D., and Ahlrichs, J. L. (1967). Soil Sci. Soc. Am. Proc. 31,30-33. 111. Harter, R. D., and McLean, E. 0. (1965).Agron.]. 57,583-588. 112. Hartley, G. S., and Roe, J. W. (1940).Trans. Faraday Soc.36,101-109. 113. Hattori, T., and Furusaka, C. (1960).J.Biochem. 48,831-837. 114. Hattori, T., and Furusaka, C. (196l).J.Biochem. 50,311-315.

77. 78. 79. 80. 81.

48

6. STOTZKY

115. Hausenbuiller, R. L. (1950). Trans. 4th Intern. Congr. Soil Sct., Amsterdam 3, 89-93. 116. Hawrylewicz, E., Gowdy, B., and Ehrlich, R.(1962).Nature 193,497. 117. Hendrix, J. W. (1961).Phytopathology51,194. 118. Henis, Y., and Eren, J. (1963).Can.J.Microbiol. 9,902-904. 119. Hepple, S., and Burges, A. (1956).Nature 177,1186. 120. Hessayon, D. G. (1953).Soil Sci. 75,295-312. 121. Hill, I. R.,and Gray, T. R. G. (1967).J. Bactertol.93,1888-1896. 122. Hoffmann, G. (1959).2.Pflantenernaehr. Dueng. Bodenk. 85,97-104. 123. Hofmann, E., and Hoffmann, G. (1961).Plant Soil 14,96-99. 124. Hofmann, E., and Hoffmann, G. (1966).Adoan.Enzymol. 28,365-390. 125. Horowitz, N. H., and Miller, S. L. (1952).In“Fortschritte der Chemie organischer Naturstoffe” (L. Zechmeister, ed.), Vol. 20, pp. 423-459. Springer, Vienna. 126. Horowitz, N. H., Sharp, R. P., and Davies, R. W. (1967).SctencelS5,1501-1505. 127. Hotchin, J., Lorenz, P., and Hemenway, C. (1965).Nature 206,442-445. 128. Jackson, T. A. (1967).Science 155,1003-1005. 129. Jannasch, H. W. (1960).Arch. HydrobioZ.56,355-369. 130. Jefferys,E. G. (1952).J. Cen. Microbiol. 7,295-312. 131. Jenny, H., and Grossenbacher, K. (1963).Soil Sci. Soc. Am. Proc. 27,273-277. 132. Jensen, H. L. (1936).Proc. Ltnnean Soc. N.S.Wales 61,27-55. 133. Johnson, T. W., and Sparrow, F. K. (1961). “Fungi in Oceans and Estuaries,” 668 pp. Carmer, Weinheim. 134. Kalyushnyi, M. E. (1957).Mikrobtologiya 26,346-352. 135. Katznelson, H., and Bose, B. (1959).C Q C J .Microbiol. 5,79-85. 136. Klein, D. A., andcasida, L. E., Jr. (1967).Bacteriol. Proc.Al4. 137. Klevenskaya, I. L. (1960).Mtkrobiologtya 29,161-164. 138. Kok, B., and Varner, J. E. (1967).Sciencel55,1110-1112. 139. Krasil’nikov, N. A. (1958). “Soil Microorganisms and Higher Plants,” 474 pp. Akad. Nauk SSSR., Moscow. (Transl. from Russian by Israel Program Sci. Transl., Washington, D. C., for Natl. Sci. Found. and U.S.Dept. Agr., 1961.) 140. Kriss, A. E. (1963). “Marine Microbiology, Deep Sea” (transl. from Russian by J. M. Shewan and Z. Kabata), 536 pp. Wiley (Interscience), New York. 141. Kriiger, W. (1961).S. AfricanJ. Agr. Sci. 4,171-183. 142. Kriiger, W. (1961).S . AfricanJ.Agr. Sct. 4,301-315. 143. Kubiena, W. L. (1938). “Micropedology,” 243 pp. Collegiate Press, Ames, Iowa. 144. Lahav, N., and Keynan, A. (1962).Can.J. Microbiol. 8,565-572. 145. Lamanna, C., and Mallette, M. F. (1965). “Basic Bacteriology,” 1001 pp. Williams & Wilkins, Baltimore, Maryland. 146. Lebedjantzev, A. N. (1924).Soil Sci. 18,419-447. 147. Lederberg, J. (1960).Science 132,393-400. 148. Lees, H. (1949).Plant Soil 1,221-239. 149. Lees, H., and Porteous, J. W. (1950).Plant Sotl2,231-241. 150. Lees, H., and Quastel, J. H. (1946).Biochem.J.40,815-823. 151. Levin, G. V.(1963).Adoan.Appl. Microbiol. 5,116-133. 152. Levin, G. V., Clendenning, J. R.,Chappelle, E. W., Heim, A. H.. and Rocek, E. (1964).BioScience 14,3748. 153. Levin, G. V., Heim, A. H., Clendenning, J. R., and Thompson, M. F. (1962). Science 138,114-121.


49

154. Levin, G . V., Heim, A. H., Thompson, M. F., Horowitz, N. H., and Beem, D. R. (1964). In “Life Sciences and Space Research 11” (M. Florkin and A. Dollfus, eds.), pp. 124-132. North-Holland Publ., Amsterdam. 155. Levin, G . V., and Shapiro, J . (1965).J . Water Pollution Control Federution 37, 800-806. 156. Levinthal, E., Hundley, L., and Lederberg, J. (1964). I n “Life Sciences and Space Research 11” (M. Florkin and A. Dollfus, eds.), pp. 112-113. North-Holland Publ., Amsterdam. 157. Lingappa, B. T., and Lockwood, J. L. (1961).]. Gen. Microbiol. 26,473-485. 158. Lipson, R. (1967).Personal communication. 159. Lochhead, A. G . (1957).Soil Sci.84,395-403. 160. Lochhead, A. G., and Burton, M. 0.(1953).Can.]. Botany 31,7-22. 161. Lochhead, A. G., and Burton, M. 0. (1957).Can.]. Microbiol. 3,35-42. 162. Lochhead, A. G., and Thexton, R. H. (1952).]. Bacteriol. 63,219-226. 163. Lockwood, J. L. (1964).Ann. Reu. Phytopathol. 2,341-362. 164. Long, C. (1966). In “Reports, Highland View Hospital,” 11. 5. Ampersand, Cleveland, Ohio. 165. Long, C. (1967).Personal communication. 165a. Long, C., Brushenko, A., and Pontarelli, D. A. (1964).Appl. Optics 3,1031-1032. 166. LOW,P. F. (1954).Soil Sci. 77,29-41. 167. Low, P. F. (1961).Adoan. Agron. 13,269-327. 168. Low, P. F. (1962).Soil Sci. 93,6-15. 169. Lynch, D. L., and Cotnoir, L.J., Jr. (1956).Soil Sci. Soc. Am. Prvc. 20,367-370. 170. Lynch, D. L., Wright, L. M., and Cotnoir, L. J., Jr. (1956).Soil Sci. Soc. Am. Proc. 20,6-9. 171. Lynch, D. L., Wright, L. M., Hearns, E. E., and Cotnoir, L. J., Jr. (1957). Soil Sci. 84,113-126. 172. MacLeod, R. A. (1965).Bacteriol. Reu. 29,9-24. 173. MacLeod, R. A., Light, M., White, L. A., and Currie, J. F. (1966).Appl. Microbiol. 14,979-984. 174. Macura, J., and Malek, I. (1958).Nature182,1796-1797. 175. Macura, J., and Pavel, L. (1959).Folia Microbiol. (Prague)4, 82-90. 176. Macura, J., and Stotzky,G. (1967).Unpublished data. 177. Marshall, K. C., and Roberts, F. J. (1963).Nature 198,410-411. 178. Martin, N., and Gottlieb, D. (1952).Phytopathology 42,294-296. 179. Martin-Smith, M. (1963).Nature 197,361-362. 180. McCalla, T. M. (1940).]. Bacteriol. 40,33-43. 181. McLaren, A. D. (1954).]. Phys. Chem. 58,129-137. 182. McLaren, A. D. (1954). Soil Sci. Soc. Am. Proc. 18,170-174. 183. McLaren, A. D. (1957).Science 125,697. 184. McLaren, A. D., and Peterson, G . H. (1965). In “Soil Nitrogen” (W. V. Bartholomew and F. E. Clark, eds.), pp. 259-284. Am. Soc. Agron., Madison, Wisconsin. 185. McLaren, A. D., Peterson, G. H., and Barshad, I. (1958).Soil Sci. Soc. Am. PTOC. 22,239-244. 186. McLaren, A. D., and Estermann, E. F. (1956).Arch. Biochem. Biophys. 61,158173. 187. McLaren, A. D., and Estermann, E. F. (1957). Arch. Biochem. Biophys.68, 157160.

50

G . STOTZKY

188. McLaren, A. D., and Peterson, G. H., eds. (1967). “Soil Biochemistry,” 509 pp. Dekker, New York. 189. McLaren, A. D., Reshetko, L., and Huber, W. (1957).Soil Sci. 83,497-502. 190. McLaren, A. D., and Skujins, J. J. (1963).Can.J.Microbiol.9,729-731. 191. Meinschein, W. G., Barghoorn, E. S., and Schopf, J. W. (1964). Science 145,262264. 192. Menzies, J. D. (1963).Botan. Reo. 29,79-122. 192a. Mer, S. B., Derbyshire, A. J., Brushenko, A., and Pontarelli, D. A. (1967).Arch. Otoloryngol. 85,400-406. 193. Millbank, J. W. (1959).Plant Soil 11,293-311. 194. Miller, R. J., and Davey, C. B. (1967).Soil Sci. Soc. Am. Proc. 31,286-287. 195. Miller, S. L., and Urey, H. C. (1959).Science 130,245-251. 196. Miyamoto, Y. (1959).Virology 7,250-251. 197. Mortensen, J. L., and Martin, W. P. (1954). Soil Sci. Soc. Am. Proc. 18,395-398. 198. Mortland, M . M., and Gieseking, J. E. (1952).Soil Sci. Soc. Am. Proc. 16, 10-13. 199. Mose, J. R.,and Brantner, H. (1966). Zentr. Bakteriol. Parasitenk., Abt. I I 120, 480-495. 200. Mueller, K. E., and Durrell, L. W. (1957).Phytopathology 47,243. 201. Myers, G. H., Morrow, M. B., Wyss, O., and Littlepage, J. L. (1962). Science 138, 1103- 1104. 202. Nemec, P., and Bystricky, V. (1962).J. Gen. Appl. Microbiol. 8,121-129. 203. Nemec, P., and Bystricky, V. (1963).Naturwissenschaften 49,335-336. 204. Nikitin, D. I. (1964).Pochoouedenie 6,86-91. 205. Nikitin, D. I. (1966).Personal communication. 206. Nikitin, D. I., Lokhmacheva, R. A., and Vasil’eva, L. V. (1966). Intern. Congr. Microbiol., 9th Moscow, USSR. Abstr. C2ll5, p. 269. 207. Nikitin, D. I., and Vasil’eva, L. V. (1967). Rept. USSR Acad. Sci., Biol. Ser. 2, 296-301. (In Russ.) 208. Nikitin, D. I., Vasil’eva, L. V., and Lokhmacheva, R. A. (1966). “New and Rare Forms of Soil Microorganisms,” 117 pp. Nauka Publ., Moscow. (In Russ.) 209. Nommik, H. (1965).In “Soil Nitrogen” (W. V. Bartholomew and F. E. Clark, eds.), pp. 200-260. Am. SOC.Agron., Madison, Wisconsin. 210. Oag, K. R. (1940).J.Pathol. Bacteriol. 51,137-141. 211. Oparin, A. I. (1964). “Life: Its Nature, Origin, and Development” (transl. from Russ. by A. Synge), 207 pp. Academic Press, New York. 212. Oppenheimer, C. H., and Drost-Hansen, W. (1960).J. Bacteriol. 80,21-24. 213. Orenski, S. W. (1966). In “Symbiosis” (S. M. Henry, ed.), pp. 1-35. Academic Press, New York. 214. Orenski, S. W., and Bystricky, V. (1965).Personal communication. 215. Orenski, S. W., Bystricky, V., and Maramorosch, K. (1966).Can. J . Microbiol. 12, 1291-1292. 216. Orenski, S. W., Bystricky, V., and Maramorosch, K. (1966).Nature 210,221. 217. Orb, J. J. (1965). In “Current Aspects of Exobiology” ( G . Mamikunian and M. H. Briggs, eds.), pp. 13-76. Macmillan (Pergamon), New York. 218. Packer, E., Scher, S., and Sagan, C. (1963).Icarus 2,293-316. 219. Parkinson, D., and Waid, J. S., eds. (1960). “The Ecology of Soil Fungi,” 324 pp. Liverpool Univ. Press, Liverpool. 220. Patrick, W. H., Jr. (1960). Trans. 7th Intern. Congr. Soil Sci., Madison, Wisc. 2, 494-500.


51

221. Patrick, W. H., Jr. (1964). Trans. 8th Intern. Congr. Soil Sci., Bucharest, Rumania 4,605-609. 222. Patrick, W. H., Jr. (1966).Nature 212,1278-1279. 223. Paul, E. A., and Tu, C. M. (1965).Plant Soil 22,207-219. 224. Pauling, L. (1961).Science 134,15-21. 225. Peele, T. C. (1936).Cornell Unio.Agr. E x p t . Sta. Mem. 197,l-18. 226. Perfil’ev, B. V., and Gabe, D. R. (1961). “Capillary Methods of Studying Microorganisms,”534 pp. Akad. Nauk SSSR, Moscow. (In Russ.) 227. Perfil’ev, B. V., and Gabe, D. R. (1965). In “Applied Capillary Microscopy: Role of Microorganisms in the Formation of Iron-Manganese Deposits” (M. S. Gurevich, ed., trans]. from Russ. by F. L. Sinclair), pp. 1-55. Consultants Bureau, New York. 228. Perrin, R. M. S., Willis, E. H., and Hodge, C. A. H. (1964).Nature 202,165. 229. Peterson, G. H. (1962). Soil Sci. 94,71-74. 230. Pimentel, G. C., Atwood, K. C., Gaffron, H., Hartline, H. K., Jukes, T. H., Pollard, E. C., and Sagan. C. (1966). In “Biology and the Exploration of Mars” (C. S. Pittendrigh, W. Vishniac, and J. P. T. Pearman, eds.), pp. 243-251. Natl. Acad. Sci., Natl. Res. Council,Washington, D.C. 231. Pinck, L. A., and Allison, F. E. (1951).Science 114,130-131. 232. Pinck, L. A., and Allison, F. E. (1961).Soil Sci. 91,183-188. 233. Pinck, L. A., Dyal, R. S., and Allison, F. E. (1954).Soil Sci.78,109-118. 234. Pinck, L. A., Holton, W. F., and Allison, F. E. (1961).Soil Sci. 91,22-28. 235. Pinck, L. A., Soulides, D. A., and Allison, F. E. (1961). Soil Sci. 91,94-99. 236. Pinck, L. A., Soulides, D. A., and Allison, F. E. (1962).Soil Sci. 94,129-137. 237. Porter, L. K. (1965). In “Methods of Soil Analysis” (C. A. Black, D. D. Evans, J. L. White, L. E. Ensminger, and F. E. Clark, eds.), pp. 1536-1549. Am. SOC. Agron., Madison, Wisconsin. 238. Portner, D. M., Spiner, D. R., Hoffman, R. K., and Phillips, C. R. (1961).Science 134,2047. 239. Postgate, J. R. (1959).Ann.Reu. Microbiol. 13,505-520. 240. Pramer, D. (1958).Appl. Microbiol. 6,221-224. 241. Pramer, D., and Starkey, R. L. (1951).Science113,127. 242. Pramer, D., and Starkey, R. L. (1962).Soil Sci. 94,48-54. 243. Quastel, J. H. (1965).Ann. Reo. Plant Physiol. 16,217-240. 244. Quastel, J. H., and Scholefield, P. G. (1951).Bacteriol. Reo. 15,l-53. 245. Quimby, F. H., ed. (1964). “Concepts for Detection of Extraterrestrial Life,” 53 pp. Natl. Aeron. Space Admin., Washington, D. C. 246. Rodrigues, G. ( 1 9 5 4 ) ~Soil . Sci. 5,264-274. 247. Rohde, R. A. (1965).Phytopathology 55,1159-1162. 248. Rovira, A. D. (1965).Ann. Reo. Microbiol. 19,241-266. 249. Rovira, A. D., and Greacen, E. L. (1957).AustralianJ. Agr. Res. 8,659-673. 250. Runkles, J. R., Scott, A. D., and Nakayama, F. S. (1958).Soil Sci. SOC.Am. Proc. 22,15-18. 251. Russell, E. J. (1952). “Soil Conditions and Plant Growth,” 635 pp. Longmans, Green, New York. 252. Sagan, C. (1966). I n “Biology and the Exploration of Mars” (C. S. Pittendrigh, W. Vishniac, and J. P. T. Pearman, eds.), pp. 73-113. Natl. Acad. Sci., Natl. Res. Council, Washington, D. C. 253. Salisbury, F. B. (1962).Science 136,17-26.

52 254. 255. 256. 257. 258. 259. 260. 261. 262. 263. 264. 265. 266. 267. 268. 269. 270. 271. 272. 273. 274.

275. 276. 277. 278. 279. 280. 281. 282. 283. 284. 285. 286. 287.

288. 289. 290.

G. STOTZKY

Salisbury, F. B. (1967).BioSclence 17,15-24. Santoro, T., and Stotzky, G. (1967).Arch. Btochem. Biophys. 122,664-669. Santoro, T., and Stotzky,G. (1967).Bacteriol. Proc. A15. Santoro, T., and Stotzky, G. (1968).Can.J. Microbiol. 14,299-307. Santoro, T., and Stotzky,G. (1967).Soil Sci. Soc. Am. Proc. 31,761-765. Santoro, T., Stotzky, G., and Rem, L. T. (1967).Appl. Microbiol. 15,935-939. Schatz, A. (1963).J.Agr. Food Chem. 11,112-118. Schmidt, E. L., and Bankole, R. 0.(1962).Science 136,776-777. Schmidt, E. L., and Starkey, R. L. (1951).Soil Sci. 71,221-231. Schopf, J. W., Barghoorn, E. S., Maser, M. D., and Gordon, R. D. (1965). Science 149,1365-1367. Schopf, J. W., Ehlers, E. G., Stiles, D. V., and Birle, J. D. 1965).Proc. Am. Phil. SOC. 109,288-308. Seifert, Y. (1958).Pochoooedenie 2,50-54. Setchell, W. A. (1903). Science 17,934. Sewell, G. W. F. (1956).Nature 177,708. Siegel, S. M. (1957).Proc. Natl. Acad. Sci. U . S . 43,811-816 Siegel, S. M., Daly, O., Halpern, L., Giumarro, C., and Davis, G. (1963).Aerospace Med. 34,817-820. Siegel, S. M., and Giumarro, G. (1965).Icarus 4,37-40. Siegel, S. M., and Giumarro, G. (1966). Proc. Natl. Acad. Sci. U . S . 55,349-353. Siegel, S . M., Giumarro, C., and Halpern, L. (1964).Boton. Gaz. 125,241-245. Siegel, S. M., Giumarro, C., and Latterell, R. (1964). Proc. Natl. Acad. Sci. U . S . 52,ll-13. Siegel, S. M., Renwick, G., Daly, O., Giumarro, C., Davis, G., and Halpern, L. (1965). In “Current Aspects of Exobiology” (G. Mamikunian and M. H. Briggs, eds.), pp. 119-178. Macmillan (Pergamon),New York. Siegel, S. M., Roberts, K., Nathan, H., and Daly, 0. (1967). Science 156, 12311234. Siminoff, P., and Gottlieb, D. (1951).Phytopathology 41,420-430. Skinner, F. A. (1965).J. Gen. Microbiol. 14,393-405. Skujins, J. J. (1967).I n “Soil Biochemistry” (A. D. McLaren and G. H. Peterson, eds.), pp. 371-414. Dekker, New York. Skujins, J. J., Estermann, E. F., and McLaren, A. D. (1959). Can. J . Microbiol. 5, 631-639. Soulides, D. A., and Allison, F. E. (1961).Soil Sci. 91,291-298. Soulides, D. A., Pinck, L. A., and Allison, F. E. (1961).Soil Sci. 92,90-93. Sperber, J. (1957).Nature 180,994-995. Starkey, R. L., Jones, G. E., and Frederick, R. L. (1956).J . Gen. Microbiol. 15, 329-334. Stolp, H., and Petzold, H. (1962).Phytopathol. 2.45,364-390. Stolp, H., and Starr, M. P. (1965).Ann. Reo. Mtcrobiol. 19,79-104. Stotzky, G. (1960).Can.J. Mtcrobiol. 6,439-452. Stotzky, G . (1965). In “Methods of Soil Analysis” (C. A. Black, D. D. Evans, J. L. White, L. E. Ensminger, and F. E. Clark, eds.), pp. 1550-1570. Am. SOC.Agron., Madison, Wisconsin. Stotzky, G . (1965).Can.J.Microbiol. 11,629-636. Stotzky, G . (1966).Can.]. Microbiol. 12,831-848. Stotzky, G. (1966).Can.J. Microbiol. 12,1235-1246.

RELEVANCE OF SOIL MICROBIOLOGY T O SEARCH FOR LIFE ON OTHER PLANETS

291. 292. 293. 294. 295. 296. 297. 298. 299. 300. 301. 302. 303. 304. 305.

306. 307. 308. 309. 310. 311. 312. 313. 314. 315. 316. 317. 318. 319. 320. 321.

322. 323. 324. 325. 326. 327.

53

Stotzky, G. (1967).Unpublished data. Stotzky, G., and Dobson, A. (1966).Unpublished Data. Stotzky, G., and Coos, R. D. (1965).Can.J. Microbiol. 11,853-868. Stotzky, G.,and Coos, R. D. (1966).Can.J. Microbiol. 12,849-861. Stotzky, G., and Martin, R. T. (1963).Plant Soil 18,317-338. Stotzky, G., and Mortensen, J. L. (1957).Soil Sci. 83,165-174. Stotzky, G., and Norman, A. G . (1961).Arch.Mikrobiol.40,341-369. Stotzky, G., and Norman, A. G. (1961).Arch. Mikrobiol. 40,370-382. Stotzky, G., and Norman, A. G. (1964).Can.J. Microbiol. 10,143-149. Stotzky, G., and Post, A. H. (1965).Agron. Abstr. p. 88. Stotzky, G., and Post, A. H. (1967).Can.J. Microbiol. 13,l-7. Stotzky, G., and Rem, L. T. (1066).Can.J.Microbiol. 12,547-563. Stotzky, G., and Rem, L. T. (1967).Con.J.IMicrobiol. 13, 1535-1550. Strugger, S. (1948).Can../. Res. B26,188-193. Stryer, L. (1966). In “Biology and the Exploration of Mars” (C. S. Pittendrigh, W. Vishniac, and J. P. T. Pearman, eds.), pp. 141-146. Natl. Acad. Sci., Natl. Res. Council, Washington, D. C. Sussman, A. S., and Halvorson, H. A. (1966).“Spores, Their Dormancy and Germination,” 354 pp. Harper & Row, New York. Taber, W. A. (1959).Can.J. Microbiol. 5,335-344. Taber, W. A. (1960).Can.J. Microbiol. 6,503-514. Tahoun, S. A,, and Mortland, M. M. (1966).Soil Sci. 102,248-254. Talibudeen, 0.(1955).Trans. Faraday Soc. 51,582-590. Tardieux-Roche, A., Muller-Felter, S., and Ebel, J. P. (1964). Bull. Soc. Chim. B i d . 46,917-926. Thornton, R. H. (1952).Research(London) 5,190-191. Tribe, H. T. (1957). In “Microbial Ecology” (R. E. 0. Williams and C. C. Spicer, eds.), pp. 287-298. Cambridge Univ. Press, London and New York. Ugolini, F. C., and Starkey, R. L. (1966).Nature211,440-441. Urey, H. C. (1966).Science 151,157-166. Vallentyne, J. R. (1963).Ann. N. Y. Acad. Sci. 108,342-352. Vandecaveye, S. C. (1939).Trans. Third Comm. Intern. SOC. Soil Sci., New Bmnswick, N. J . A, 120-130. Van Gundy, S. D. (1965).Ann.Reo. Phytopathol. 3,43-68. van Olphen, H. (1963). “An Introduction to Clay Colloid Chemistry,” 301 pp. Wiley (Interscience),New York. Vishniac, W. (1960).Aerospace Med. 31,678-680. Vishniac, W., Atwood, K. C., Bock, R. M., GafFron, H., Jukes, T. H., McLaren, A. D., Sagan, C., and Spinrad, H. (1966). In “Biology and the Exploration of Mars” (C. S. Pittendrigh, W. Vishniac, and J. P. T. Pearman, eds.), pp. 229-242. Natl. Acad. Sci., Natl. Res. Council, Washington, D.C. Vogler, K. G., and Umbreit, W. W. (1941).Soil Sci. 51,331-337. Waid, J. S., and Woodman, M. J. (1957).Pedologie7, 155-158. Waksman, S. A. (1916).Science 44,320-322. Waksman, S. A., and Woodruff, H. B. (1942).Soil Sci. 53,233-239. Wald, G. (1957).Ann. N. Y. Acad. Sci. 69,352-368. Wald, G. (1958). Introduction in Henderson, L. J. “The Fitness of the Environment: An Inquiry into the Biological Significance of the Properties of Matter,” pp. 17-24. Beacon Press, New York.

54 328. 329. 330. 331. 332. 333. 334. 335. 336. 337. 338. 339. 340. 341. 342. 343.

G . STOTZKY

Wallace, R. H., and Lochhead, A. G. (1950).Can.]. Res. C28,l-6. Warcup, J. H. (1950).Nature 166,117. Warcup, J. H. (1955).Nature 175,953. Warcup, J. H. (1955).Brit. Mycol. SOC.Trans. 38,298-301. Warcup, J. H. (1957).Brit. Mycol. SOC.Trans. 40,237-262. Warcup, J. H. (1960).In “The Ecology of Soil Fungi” (D. Parkinson and J. S. Waid, eds.), pp. 3-22. Liverpool Univ. Press, Liverpool. Weber, D. F., and Gainey, P. L. (1962).Soil Sci. 94,138-145. Webley, D. M., Duff, R. B., and Mitchell, W. A. (1960).Nature 188,766-767. Webley, D. M., Henderson, M. E. K., and Taylor, I. F. (1963).J . Soil Sci. 14, 102-1 12. Welshimer, H. J. (ISSO).]. Bacteriol. 80,316-320. Wentz, M. W., Scott, R. A., and Vennes, J. W. (1967).Science 155,89-90. Zagallo, A. C., and Katznelson, H. (1957).]. Ba~teriol.73~760-764. Zavarzin, G. A. (1962).Mikrobiologiya 31,586-588. Zobell, C. E. (1943).]. Bacteriol. 46,39-56. Zobell, C. E., and Morita, R.Y. ( 1 9 5 7 )Bacteriol. ~ 73,563-568. Zvyagintsev, D. G . (1959).Mikrobiologiya 28,496-501.

Experiments and Instrumentation for Extraterrestrial Life Detection

GILBERT V. LEVIN Biospherics Research, Incorporated Washington, D.C . I. Introduction .................................................................

................................................................ n Methods ................................................ Gulliver ................................................................

A. B. Heterotrophic Photosynthesis .................................. C. Autotrophic Photosynthesis ..................................... D. Diogenes ............................................................... E. Phosphate Uptake ....... .................................... F. Sulfur Uptake ........................................................ IV. Automated Microbial Metabolism Laboratory .................. References ...................................................................

55 56 56 57 59 61 62 65 67 67 71

I. Introduction Any treasure hunt should begin with a description of the object sought. Yet, in the search for extraterrestrial life, the greatesr treasure hunt in history, this is not possible. Fortunately, a few considerations [largely discussed by Drs. Brown and Vishniac at this symposium (see footnote 1, p. I)] somewhat reduce this handicap in the impending quest: 1. It seems logical to concentrate the search on relatively primitive types of microbial life. It is possible to imagine, especially from what we already know about environmental conditions on the planets in our solar system, that life on Mars, for example, might be limited to such relatively primitive forms. On the other hand, it is very difficult to imagine an ecology in which only highly organized macroorganisms exist. Such a situation could not provide for the necessary degradation of complex biochemicals and recycling of the components. Regardless of the level of biological development in any ecosphere, it would seem that microorganisms would be essential if the life process were to be sustained. 2. Chemical considerations make carbon the most likely candidate element on which extraterrestrial life would be based. 3. Chemical considerations imply that all biochemical reactions probably take place in aqueous solution. 55

56

GILBERT V. LEVIN

4. Because the prime source of the energy required to sustain life for evolutionary periods is the sun, photosynthetic forms seem essential in any extraterrestrial population. They may or may not be accompanied by other forms of autotrophs or by heterotrophs. 5. Finally, it is likely that the microorganisms would be widespread over the planet. II. Sampling It is fortuitous that microorganisms are the principal life form of interest. This considerably reduces the complexity of obtaining a sample for testing in an automated instrument landed on the planet. However, this is not to imply that obtaining a sample of microorganisms will be simple. The samples should be taken from an area unaffected by the landing of the. spacecraft. If retrorockets are used during landing, contamination may be widespread. Samples should include vertical profiles from the surface to considerable depths. The mere taking of a sample of material, such as loosely conglomerated soil, may produce profound changes in the material. Equilibria and physical arrangements important to biological activity could be disrupted. Above all, great care must be taken against the possibility of contaminating the sample with terrestrial microorganisms. The stringent precautions necessary to prevent planetary contamination are discussed elsewhere by Mr. Hall.

111. life Detection Methods Once an extraterrestrial sample is obtained, how is it to be examined for evidence of life? Obviously, this is the most difficult part of the problem. An unknown life form, operating on unknown biological processes in an unknown environment presents an unprecedented challenge to detection. Six general categories of experiments are being developed for the National Aeronautics and Space Administration: (a) physical and chemical assays for simple organic compounds of biological interest, (b) morphological evidence of living forms, (c) assays for intermediate or complex biochemicals, (d) evidence of metabolism, (e) evidence of growth, and (f) evidence of reproduction. These attributes of life are cited in the increasing order of significance I would ascribe to them. Experiments seeking physical and chemical determinations of relatively simple compounds would be those most likely to yield positive results. However, the evidence thus obtained could not establish the presence of life. Morphological distinctions

EXPERIMENTS AND INSTRUMENTATION FOR EXTRATERRESTRIAL LIFE DETECTION

57

between living and nonliving forms at the microscopic level are frequently difficult on Earth. Interpretation of micrographs may be impossible when unfamiliar forms of life are encountered in a strange inorganic matrix. Organic compounds, including moderately complex ones, are known to be generated abiogenically. On the other hand, while those experiments seeking evidence of reproduction would yield the most conclusive results regarding the existence of life, they are the least likely to be successful. This is because they must be based on assumptions, such as those discussed earlier, concerning the general nature and function of the living systems sought. However, if successful, not only could these experiments establish the presence of life, they might determine metabolic rates, metabolic pathways, the mode of growth, and the generation period. Through such experiments, ultimately, the nature of the life encountered could be compared with terrestrial life for the paramount determination of whether the two forms of life are similar. It is in the categories of metabolism growth, and reproduction that I would like to describe some experiments that my co-workers and I have been developing over the past several years.

A. GULLIVER Gulliver (1, 2, 3) offers radioactive substrates containing 14C and 35S in aqueous solution to the sample. If organisms are present, and they can metabolize one or more of the labeled substrates, the production of radioactive gas is likely. A “getter” collects the gas so produced by chemically precipitating it on a surface monitored by a radiation counter. An exponential increase in the output from the counter is indicative of growth or reproduction. In the event metabolism occurs in the absence of growth or reproduction, this will be indicated by any significantly positive slope to the curve generated. The curve produced by the test unit is compared to that obtained from an identical, but inhibited, control unit. Inhibition is induced by the application of heat or a chemical antimetabolite. The object of the control is to differentiate between a metabolic response and an inorganic reaction with the extraterrestrial sample. Extensive laboratory and field tests (3) have supported the general applicability to terrestrial microorganisms of the media developed and the antimetabolite selected. The Mark I11 version of Gulliver, shown in Fig. 1, has been widely tested in the laboratory and field. Placed in a simulated capsule, as shown in Fig. 2, the instrument has performed in extreme terrestrial environments. One test was made at the 12,000-foot elevation of White

58

GILBERT V. LEVIN

FIG. 1. Mark 111 model of Gulliver. The projectiles deploy the string over the planetary surface. The string carries particulates back into the incubation chamber where the radioactive broth is applied. The geiger counter and electronics are housed in the cylinder resting on the display base.

Mountain, California. The temperature was below freezing at this barren location (Fig. 3) above the timber line. Positive results were obtained within 1 hour. Tests on a sand dune in Death Valley, California, (Fig. 4) produced counts significantly above those of the inhibited control within 2-3 hours. Similar results were obtained on the desert salt flats near the Salton Sea, California. During the desert tests, an experiment was performed in which the radioactive substrate was applied directly to the soil. An almost immediate, high level response resulted. After only several minutes of gas collection, the activity exceeded that of the Gulliver I11 instrument by approximately an order of magnitude. The advantage seems to accrue from the minimal disturbance imposed on the microenvironment by this in situ method. A new model of Gulliver, Mark IV (Fig. 5), was designed and fabricated to take advantage of this finding. As currently visualized, a number of these miniaturized instruments would be ejected from the landing capsule. The units might all be replicates or groups of replicates containing different media or antimetabolites. Each unit is self-righting and contains all necessary


59

FIG. 2. Mark 111 Gulliver in simulated planetary lander capsule. Projectiles fire through the assembled housing. Note that opposite end of unit is up compared to Fig. 1. Instrument operates in any position.

components, including a Geiger counter and associated electronics, to conduct the life detection test directly on the “soil.” Unlike, Mark 111, which samples only the surface material, Mark IV examines the site to the depth penetrated by the medium released. Power is supplied from the spacecraft through an umbilical cord which also serves to relay the data to the central capsule for processing and transmission to Earth.

B. HETEROTROPHIC PHOTOSYNTHESIS The probability that any life on Mars must obtain its ultimate energy from the sun is high. Tests for photosynthesis thus assume considerable importance. An experiment has been developed (3) in which the photosynthesis of algae is detected through the dark evolution of radioactive carbon dioxide derived from labeled glucose. The organisms assimilate the glucose heterotrophically. When exposed to light, they retain and fix the carbon dioxide produced. In the dark, however, endogenous respiration releases the recently fixed 14C02which is then detected in the Gulliver fashion. Data obtained from such an experiment are presented in Fig. 6. Evidence for photo-

60

GILBERT V. LEVIN

synthetic activity is revealed by the correlation of the evolution of I4CO2with the light and dark periods.

FIG. 3. Mark I11 Gulliver in test position on bleak, 12,000-foot elevation of White Mountain, California. Reproduced from Levin, G . V., and A. H . Heim (5).

FIG.4. Mark 111 Gulliver in test position on sand dune in Death Valley, California.


61

FIG.5. Mark IV Gulliver, in sttu model. Fired from the landed capsule, this unit is self-righting and contains the radioactive broth, the Geiger counter and associated electronics, and all the electromechanical components needed for automatic operation. Low voltage power is supplied through the umbilical cord, right, through which the data are also returned to the spacecraft.

C. AUTOTROPHICPHOTOSYNTHESIS This experiment (3) seeks evidence for strict phototrophs. It is probably the least geocentric of any of the experiments described in this report. Its only assumption is that carbon dioxide will participate in a gas exchange step of the photosynthetic process. The detection of large quantities of carbon dioxide in the Martian athosphere supports this possibility. In essence, a small portion of the planetary surface material and overlying atmosphere are enclosed. A trace quantity of I4COr is then introduced into the trapped atmosphere in the presence of light. Time is allowed for any photosynthetic organisms present to fix carbon dioxide, including some of that labeled. The atmosphere is then replaced with Martian atmosphere and the light excluded. The evolution of l4 C 0 2 through endogenous respiration by the photosynthetic organisms present is then monitored in the Gulliver fashion. Data obtained from a laboratory experiment with algae are given in Table I.

62

GILBERT V . LEVIN

X

loooE

FIG. 6. Heterotrophic photosynthesis experiment, laboratory test data. Medium, M, 10 pcurieslml.) Activity includes urea agar with sodium lactate-lJT (2 x sterile control and background. Activity is 14C02evolved by Chlorella pyrenoidow in Dark control. response to light and dark growth cycles. - . - - . - Light control. V-V-V Tests, -0-n-Tests: - - - I n lighti n dark. Reproduced from Levin et al. (5) p. 130.

*

D. DIOGENES The intermediate compound adenosinetriphosphate (ATP) is present in all living terrestrial cells. A very sensitive assay for ATP can be performed using the luciferase enzyme system present in the lantern of the firefly. These two established facts have been combined to produce a highly sensitive, general life detection test (4, 5, 6). A sample of the material to be tested for microorganisms is treated in a manner to release microbial ATP, for example, by extraction with dimethylsulfoxide. An aliquot of this extract is then injected into a solution containing the luciferase system extracted from the firefly


63

TABLE I LABORATORY DETERMINATIONS OF AUTOTROPHIC PHOTOSYNTHESIS DETECTIONEXPERIMENT. TEST ORGANISM: C . pyrenoidosa Net Radioactivity (C.P.M.) ~

Treatments (30 minutes each)

I4CO2Evolution Replicate

Mean

1

2

2398

2514

Live cells, continuous darkness

53

Killed cells, preilluminated

Live cells, preilluminated

Killed cells,

~ ~ ~ _ _ _ _ _ _ _ _ _ _

Net 14C02Fixation Replicate

Mean

1

2

2456

50,733

56,251

53,492

53

53

658

644

65 1

10

0

5

152

-

152

7

3

5

4

3

4

continuous darkness "Reproduced courtesy of: Proc. 12th Ann. Am. Astron. SOC., Meeting, Anaheim, California, May, 1966.

lantern (luciferase, luciferin, and magnesium ion in the presence of dissolved oxygen). Any ATP present will result in the production of light. The peak intensity of the light produced is directly proportional to the amount of ATP present. The reaction is monitored in an instrument containing a photomultiplier tube and the result may be displayed on an oscilloscope or recorded on a strip chart. The laboratory instrument built for this purpose is shown in Fig. 7. A typical response recorded by a Polaroid photograph of an oscilloscope is seen in Fig. 8. Developments in the biochemistry and instrumentation of this experiment make it possible to detect approximately 200 Escherischia coli or one yeast cell in a total elapsed time of less than 2 minutes. Figure 9 shows a feasibility model of an instrument developed for the Goddard Space Flight Center of NASA. Flight models of this instrument would be carried aboard rockets to make real-time determinations of microbial ATP collected in the upper atmosphere. The instrument can make four assays in a 2-minute period. The results would be transmitted to a ground station by radio. ATP has been abiogenically synthesized under supposed primitive

64

GILBEH’I’ V. LEVIN

Earth conditions (7). Thus, while its presence on another planet would be of great biological interest, it would not establish the exist-

FIG. 7. Laboratory model of instrument developed for assay of microbial ATP by firefly bioluminescent method. Left to right: oscilloscope for readout; instrument proper, housing reaction chamber and photomultiplier hibe; nulling circuit (foreground); power snpply. Reproduced from Levin, G. V. et u1. (&).

FIG. 8. Typical response of ATP assay, Polaroid photograph of oscilloscope tube. Off-scale response produced by ATP from approximately 1000 yeast cells. Full scale deflection = 10 crn.


65

ence of life. However, the incorporation of the “delta time” concept into the experiment does convert it into a life detection test. Thus, the determination of an increase in ATP content with time in a culture of the sample material would constitute almost unimpeachable evidence for life.

FIG. 9. Feasibility model of rocket-borne, ATP assay instrument. Instrument can extract ATP from particulates and conduct quadruplicate assays within total elapsed time of 2 minutes.

E. PHOSPHATE UPTAKE It is believed that all terrestrial organisms require inorganic orthophosphate for the production of ATP and nucleic acids. Accordingly, the uptake of orthophosphate from solution can be indicative of metabolism. Such uptake can take place even in the absence of growth or reproduction (8). A life detection test has been developed to the point where approximately 200 E . coli per milliliter of medium can be detected within 3-5 hours by following the disappearance of dissolved phosphate from the culture medium. As in the case with Gulliver, an inhibited control is used in the experiment. Experimental data obtained with this technique are presented in Fig. 10. The antimetabolite used in the control was 2,4-dinitrophenol which is known to uncouple oxidative phosphorylation. At the in-

66

GILBERT V. LEVIN

dicated intervals, aliquots of the cultures were removed and filtered. They were then assayed for orthophosphate by the stannous chlorideammonium molybdate method. Radioactive phosphorus was not used because its short half-life precludes its use in a Mars probe. It is

FIG. 10. Phosphate uptake experiment with wild sewage microorganisms. Note effect of 2,4-dinitrophenol (DNP) in uncoupling oxidative phosphorylation. Reproduced from Levin, G . V. and J. Shapiro (8a).

interesting to note that the control culture took up some orthophosphate, as would be expected in that 2,Cdinitrophenol uncouples oxidative phosphorylation only, permitting substrate phosphorylation to continue. When a general poison is used, no uptake is observed. The use of phosphate uptake as a technique for seeking to detect exterrestrial life brings into play another element essential for terrestrial life. Thus, the phosphate uptake test provides an opportunity to detect noncarbon-based life as well as carbon-based life. The possibility of the evolutionary incorporation of phosphate into any living system seems strongly directed by the high-energy capacity associated with the phosphate trimer.

EXPERIMENTS

AND INSTRUMENTATION FOR EXTRATERRESTRIALLIFE DETECTION

67

F. SULFUR UPTAKE This technique (9) seeks to detect the metabolic uptake of inorganic sulfur as an index for life. Of particular interest is the uptake of the sulfate ion. On the basis of chemical considerations, high-energy bonding associated with sulfate polymers is a good candidate to substitute for the role of phosphate polymers in biological energy transfer. This consideration, together with the fact that sulfur is an essential element for all forms of terrestrial life, provides another independent means for seeking extraterrestrial life. The half-life of sulfur is sufficiently long to permit radioisotopic techniques to be used in a Mars probe. Inorganic forms of sulfur other than sulfate may also be included in the test medium. The suspected organisms can be filtered and examined directly for incorporation of the isotope, or the medium can be dried and assayed for radioactivity as evidence of uptake by the microorganisms. A control is also incorporated as part of this experiment.

IV. Automated Microbial Metabolism laboratory

A major development in the preparations to search for extraterrestrial life has occurred over the past 2 or 3 years. This has been the realization by NASA and the biological community that the expense and importance of the planetary exploration program requires the integration of a number of individual experiments into a single instrument package (10, 11). This package would constitute an automated laboratory to be landed on the surface of the planet. In agreement with this philosophy, my co-workers and I have been developing a relatively simple version of such an automated biological laboratory which we have designated as the Automated Microbial Metabolism Laboratory. Development of the biological experiments and conceptual engineering of the AMML is underway (9, 12). The AMML is anticipated to weigh less than 25 pounds and could serve as the biological laboratory on a relatively small planetary lander. If, on the other hand, the first planetary lander will have a very large payload capacity, the AMML could serve as a subsystem of the total laboratory. In essence, the six metabolic experiments just described and six associated physical determinations of biological interest are incorporated into the AMML in a manner to make common use of various subsystems and to use standardized modules for others. The physical measurements serve two purposes. They are required for interpretation of the biological results and, with relatively minor

68

GILBERT V. LEVIN

modifications, they can serve to make measurements of the environment. Specifically, the parameters to be determined are: (a) temperature, (b) atmospheric oxygen, (c) pH of the surface material, (d) ambient light intensity, (e)background radiation, and (f) soluble phosphate content of the surface material. Temperature measurements would be made by means of a thermister and are required in the metabolic experiments for an assessment of the influence of temperature on the metabolic rates monitored. Oxygen will be determined by an oxygen electrode. This measurement is very important to the photosynthetic experiments to determine whether any photosynthesis detected is of the plant or bacterial type, i.e., whether oxygen is produced or not. A pH electrode would be used in culture experiments to help interpret the data obtained. The ambient light intensity incident to the planetary surface would be measured by superimposing neutral Sample In

I

*Control unit- -_ __

Either chamber may be selected for control

Coarse filters

31 0-0 preparation

control*

preparation

L-

pmparation control

___

__

Readout

"C 0,photosynthesis eaperiment

FIG. 11. Automated microbial metabolism laboratory, schematic. Reproduced from: Levin, G . V. and G. R. Perez (12).

density filters over the photomultiplier tube which serves as a central sensor for the metabolic experiments. Background radiation can be determined in conjunction with the radioisotope experiments. Solu-


69

ble phosphate content of the surface material can be obtained by a zero-time” measurement of the culture in the phosphate uptake experiment. An attempt is being made to convert all of the metabolic readouts to light pulses and thereby utilize a common sensor system, a photomultiplier tube circuit, for the six metabolic experiments. The isotopic experiments would use scintillators to transduce the beta particles into photons. The output of the ATP experiment is already in the form of light. Attempts are being made to convert the phosphate assay output into light. One possibility is to complex the phosphate with triethylamine containing labeled carbon. Triethylamine quantitatively precipitates orthophosphate (13).Through the use of carbon-labeled triethylamine, the orthophosphate can be determined by measurement of the radioactivity of the precipitated complex. As in the case of the other radioactive tests, the beta emissions would be converted to light pulses. A schematic of the proposed AMML is shown in Fig. 11. Figure 12 shows a conceptual layout of the instrument. In summary, this instrument would examine an extraterrestrial sample for metabolism, growth, or reproduction through monitoring the biological interface with the environment for the involvement of carbon, sulfur, oxygen, phosphate, and light. It would seek these interactions in both heterotrophic and autotrophic systems. Further, it would look for the production of the intermediate compound ATP. Each of these “windows” into the living process could possibly answer the question of whether life exists. However, incorporated in this fashion, the experiments create a sum greater than its parts. This is because, although separate and diverse, the experiments reinforce and extend each other. The results of one may permit an otherwise impossible interpretation of another which, by itself, might yield doubtful results. For example, the phosphate and sulfur tests might indicate the presence of life which, yielding negative results in the ATP test, would thereby be shown to possess an intermediary metabolism considerably different from terrestrial forms. The next step in the development of the AMML is the detailed design and construction of an operable breadboard. Then it will be possible to examine a number of microbial cultures and soil samples to test the integrated experiment concept. The data obtained will be used to refine the experiment further in preparation for the exciting biological opportunity opening to us. I‘

70 ~

Chemical getter and scintillator Antimetabolite

/ Pliable bellows

Dual in situ chamber

Cross section

Sample in ,-Culture

chamber

ATP preparotion system

- h sifu

experiments

multiplier tubes

:tion c hombers

I ..a"""",

FIG. 12. Automated microbial metabolism laboratory, conceptual design. Reproduced from Levin, G . V. and G . R. Perez (12).


71

ACKNOWLEDGMENTS

The Gulliver and AMML programs have been supported by the Bioscience Programs, Office of Space Science and Applications, National Aeronautics and Space Administration. Initial support for the ATP life detection method was given by the Bureau of Naval Weapons, Naval Testing Laboratory, Dahlgren, Virginia. The Goddard Space Flight Center, NASA, has supported and worked along with the Diogenes program. In particular, Dr. Norman H. MacLeod and Mr. Emmett w. Chappelle, Space Biology Branch, GSFC, have made scientific contributions to this effort. Dr. Norman H. Horowitz, Division of Biology, California Institute of Technology, is co-experimenter on the Gulliver program and, as such, devised the autotrophic photosynthesis experiment. In addition, the author wishes to express thanks for the scientific and technical assistance of his co-authors named on the various papers cited herein.

REFERENCES 1. Levin, G . V., Heim, A. H., Clendenning, J. R.,and Thompson, M.-F. (1962).Science 138,114. 2. Levin, G . V., Heim, A. H., Thompson, M.-F., Horowitz, N. H., and Beem, D. R. (1964). In “Life Sciences and Space Research 11” (M. Florkin and A. Dollfus, eds.). North-Holland Publ., Amsterdam. 3. Radioisotope Biochemical Probe for Extraterrestrial Life, Ann. Repts (1962-1965). NASA Contract No. NASr-10, Hazleton Lab., Falls Church, Virginia (Resources Res., Washington, D.C.). 4. Levin, G . V., Clendenning, J. R.,Chappelle, E. W., Heim, A. H., and Rocek, E. (1964).BioScience 14, No. 4. 5. Levin, G . V., and Heim, A. H. (1965). 1n“Life Sciences and Space Research 111” (M. Florkin, ed.). North-Holland Publ., Amsterdam. 6. The Design and Fabrication of an Instrument for the Detection of Adenosinetriphosphate (ATP), Final Rept (1965). Goddard Space Flight Center Contract No. NAS5-3799, Hazleton Lab., Falls Church, Virginia. 6a. Levin. G . V., Usdin, E., and Slonim, A. R.(1968).AerospaceMed. 3 8 , l . 7. Ponnamperuma, C., and Mack, R.(1965). Science 148.1221. 8. Levin, G . V. (1963). Ph.D. Thesis, Johns Hopkins Univ., Baltimore, Maryland. 8a. Levin, G . V., and Shapiro, J. (1965).J.Water Poll. Control Fed. 37,6. 9. A Study Toward Development of an Automated Microbial Metabolism Laboratory, Ann. Rept. (1967). NASA Contract No. NASW-1507, Hazleton Lab., Falls Church, Virginia. 10. Biology and Exploration of Mars, 1964 Summer Study, Space Science Board, U. S. Natl. Acad. Sci.-Natl. Res. Council, sponsored by NASA, as reported in Proposed Biological Exploration of Mars Between 1969 and 1973. (1965). Nature 206, No. 4988,974. 11. Young, R. S., Painter, R. B., and Johnson, R. D. (1965). “An Analysis of the Extraterrestrial Life Detection Problem.” Ames Res. Center, Natl. Aeron. Space Admin., Washington, D. C. 12. Levin, G . V., and Perez, G. R. (1966).Proc. 12th Ann. Meeting, Am. Astron. SOC., Anaheim, California, May. 13. Sugino, Y., and Miyoshi, Y. (1964)J. Biol. Chem. 239,2360.


Halophilic Bacteria

D. J. KUSHNER Department of Biology, Unioersity of Ottawa Ottawa, Ontario, Canada

Introduction ................................................................. 73 Previous Reviews ................. ...... 74 11. Ecology and Classificat ...... 75 75 A. Occurrence and Economic Importance ..................... 76 B. Classification of Halophilic Bacteria .......................... 78 111. The Nahire of Halophilic Life ....................................... 78 A. Resistance and 79 B. Enzyme Studies 81 C. Ribosomes and Protein Synthesis .............................. 83 D. External Layers of Extreme Halophiles ..................... E. Differential Effects of Na+ and K on Cells and 87 I.

+

IV. V. VI. VII.

of Halobacterium Envelopes .................................... Taxonomic Importance of Halophilic Lipids.. ................... Growth and Survival of Extreme Halophiles .................... Speculations on Moderate Halophiles .............................. On the Origin of Extreme Halophiles ............. References ...................................................................

88 91 92 94 96 97

I. Introduction Organisms able to grow in, or requiring, high or saturated salt concentrations excite the biologist’s imagination because of their inherent unlikeliness. The recent increase of interest in the extremely halophilic bacteria has been gratifying to this reviewer and, presumably, to the handful of others studying these unusual microorganisms. My present task is simplified by the excellent reviews that have already appeared on the biological and biochemical aspects of halophilic life. Instead of repeating material already covered in detail, I intend in this chapter to sift and summarize some of the main aspects of this life and to speculate on the evolutionary, physiological, and biochemical problems that halophiles are subject to or that they raise. In addition, this chapter will deal with material that is not covered by the most recent review (Larsen, 1967),or that is still unpublished. Indeed, unpublished (and even unpublishable) work has relatively great interest to amateurs of halophilic bacteria. We are few; our bacteria grow more slowly than those of other workers, and 73

74

D. J. KUSHNER

the salt they grow in presents us with certain very real practical problems; their nutrition is complex and probably ill-defined, this poor definition being helped by the apparent ability of some extreme halophiles to store large quantities of nutrient material; a good deal of what might be called the “basic bacteriology” of extreme halophiles is simply missing and must be filled in by the researcher interested in higher things. It is sometimes as useful to know what others have tried unsuccessfully as to know where they have succeeded. The word “halophilic” that is, salt-loving, is used in different ways. Marine bacteria have been described as halophilic -which they are, in that many if not most of them require up to 0.5 M NaCl for growth and survival (MacLeod, 1965). Salt-tolerant bacteria are also described as halophilic, and most bacteria that are able to grow in high salt concentrations, say 3.0 M , require at least some salt for growth. In this review, the nomenclature of Baxter and Gibbons (1956) will be followed, whereby the term “extreme” halophile will be reserved to those few (possibly only two) bacterial species that grow if the salt concentration is 3.0 M or higher. The term “moderate” halophile refers to those that can grow in from about 0.5-3.5 M NaCl (about 3-25% NaC1). Several distinct species of such bacteria have already been described, and it is easy to find moderate halophiles that have probably not been previously classified in seawater and in more salty environments. Finally, halotolerant bacteria are those that apparently require no salt for growth but may grow in NaCl concentrations of 10% or more. Since most complex media already contain some NaC1, the reported lack of a requirement for this salt is not always accurate. A synthetic medium may be needed to demonstrate the requirement for a small amount of NaCl (MacLeod, 1965). PREVIOUSREVIEWS

For the earlier literature on halophilic bacteria, the reviews of Flannery (1956), Ingram (1957), and Larsen (1962) should be consulted. Brown’s (1964) and MacLeod’s (1965) reviews are especially useful in bringing in the correlation between extremely halophilic bacteria and marine bacteria, and the former also contains very interesting speculations on moderately halophilic bacteria. Biochemical aspects of halophilic life are considered in these, and in more detail in Larsen’s (1962, 1967) two reviews. The second deals almost entirely with extreme halophiles. I discussed halophilic

HALOPHILIC BACTERIA

75

bacteria briefly in a longer review on resistance to various harsh environmental conditions (Kushner, 1964, 1966b). A current review by Gibbons (1968) discusses the practical details of growing halophilic bacteria and mentions several interesting halophilic species. This chapter is concerned with bacteria, but many other microorganisms also live in strong salt solutions (see Ingram, 1957; Larsen, 1962) and will be mentioned when they illustrate contrasting mechanisms of life.

I I . Ecology and Classification of Halophiles A. OCCURRENCE AND ECONOMICIMPORTANCE Extremely halophilic bacteria are found in salt lakes, the Great Salt Lake and the Dead Sea among others, and in pools and flats along the seashore where seawater has been concentrated by drying in the sun. Early research on these bacteria was stimulated by the fact that they spoil fish, bacon, and hides preserved in (presumably) solar salt prepared from seawater (Flannery, 1956; Ingram, 1957). The spoiled products turned red, just as in the open the presence of extremely halophilic bacteria is manifested by these bacteria’s bright red or pink pigments. A fine collection of extreme halophiles may be seen from the air in the salt flats at the southern end of San Francisco Bay; as these flats dry, their color progresses from pink to brick-red. Such color changes were used very long ago by the Chinese to indicate saturation of evaporating seawater, and other ancient writers commented on the red or purple color of salts from different sources, colors presumably attributable to halophiles (Baas-Becking, 1931). The extremely halophilic algae, Dunnaliella salina and D. viridis can also color salt seas red or green, respectively (Larsen, 1962; Zahl, 1967). The colors of halophilic bacteria are due to carotenoid pigments, and there is evidence that this pigmentation protects the bacteria from the bright sunlight to which they are naturally exposed (Larsen, 1962; Nandy and Sen, 1967). Many different moderately halophilic bacteria have been isolated, some from spoiled foods but most from curing brines and salted food products and some from saline soils (for several references see Flannery, 1956; Ingram, 1957; Keller and Henis, 1967; Larsen, 1962; Scott, 1957). Salt-tolerant yeasts, many of which are found in soy sauce whose NaCl content is 18%, are reviewed by Onishi (1963), and bacteria from soy sauce are reviewed by Ueno (1964) and Yoshii (1967).

76

D. J. KUSHNER

Some recently-studied halophiles of special interest include a bacterium that can grow in saturated LiCl (Siege1 and Roberts, 1966), a moderate halophile that is photosynthetic and a strict anaerobe (Raymond and Sistrom, 1967) and a strain of Chlamydomonas that grows best in the presence of 10% NaCl (Yamada and Okamoto, 1961; Yamamoto, 1967).

B. CLASSIFICATION OF HALOPHILICBACTERIA

The moderate halophiles include many different bacterial species, both Gram-positive and Gram-negative and of several different forms. halodenitri$cans and Vibrio costicolus The species M~CTOCOCCUS have been most studied. Few moderate halophiles have pigments. Their classification seems to have caused less controversy than that of the extreme halophiles. These have been given a bewildering variety of names, as have indeed many other bacterial species; but the taxonomy of the extreme halophiles has been especially confused by the ease with which some of them change shape on handling and on different culture conditions, and by the difficulties of staining them (see below). Historical reviews of the taxonomy of halophilic, especially extremely halophilic, bacteria are given by Flannery (1956), Ingram (1957), and Larsen (1962). The Seventh Edition of Bergey’s Manual (Breed et d., 1957) places the rod-shaped halophiles in the genus Halobacterium, family Pseudomonadaceae. This genus consists of rod-shaped, highly pleomorphic bacteria which are obligate halophiles, requiring at least 12% salt for growth and which live even in saturated brine. If motile, they have polar flagella. They are Gram-negative and usually contain carotenoid pigments, ranging from orange to brilliant red. They may or may not attack carbohydrates, but do not produce gas. The five species in this genus, Halobacterium salinarium, H . cutirubrum, H . halobium, H . marismortui, and H . trapanicum, are distinguished from each other by their ability to produce nitrites and gas from nitrate and to produce acid from glucose as well as by the intensity of their pigmentation. In addition, among the first three species (which have been most studied), H . cutirubrum is more proteolytically active than H . salinarium and does not rupture as rapidly when the salt concentration is reduced. H . hulobium is distinctive in containing intracellular structures that resemble gas vacuoles. These can be made to appear or

HALOPHILIC BACTERIA

77

disappear by changing the hydrostatic pressure of a bacterial SUSpension (Larsen et al., 1967). Colonies of H . halobium containing gas vacuoles are opaque, whereas those without vacuoles are translucent, as are colonies of other halobacteria. The halophilic cocci are placed in Bergey's Manual (Breed et al., 1957) in the family Micrococceaceae. The following species are described: Micrococcus morrhuae, Sarcina littoralis (with the note that these species may be identical) and S. morrhuae, also closely related. All are spheres, 1-1.5 p in diameter, occurring singly, in pairs, short chains, or packets. They are described as Gram-negative or Gram-variable obligate halophiles, colored various shades of red. M . morrhuae is described as having the ability to grow in from 9 to 30% NaCl -a wider range than that of other extreme halophiles -and S. littoralis as growing in the range 16-32% NaC1, that of the halobacteria. Colwell and Gibbons (1968) have recently reassessed the taxonomy of extremely and moderately halophilic bacteria. They subjected 63 strains of the former and 6 of the latter to taxonomic analysis by the computer method (Colwell and Liston, 1961; Colwell, 1963; Quadling and Colwell, 1964). The strains examined had been isolated in different parts of the world from solar salt, saline lakes, salted cow and buffalo hides, salt fish and beans, sausage casing, and ham-curing and bacon-curing brines. Some 185 different features were compared, including motility, staining, growth on various media, including those at different pH values and different salt concentrations; sensitivity to antibacterial substances; possession of specific enzymes, fermentative ability, and other biochemical reactions. The results suggested that there was much species synonymy within the halophiles. Only two genera were thought to be substantiated, Halobacterium, the rod-like Gram-negative halophilic bacteria, and Micrococcus, the aerobic halophilic cocci (for which Halococcus has been suggested as an appropriate generic epithet'). It was suggested that within the genus Halobacterium only the single species Halobacterium salinarium should be recognized and that H . cutirubmm and H . halobium should be considered as synonyms for the species. The results suggested tentatively that H . marismortui, H . trapanicum, and H . gibbonsii were also synonymous with H . salinarium, but more work on these was thought necessary. 'This term was also suggested by Larsen (1967).

78

D. 1. KUSHNER

Colwell and Gibbons also found that the moderate halophiles called

M . denitrijicans and M . halodenitrijicans were probably a single species, as were Vibrio costicolus and Pseudomonas beijerincki. A comparison was also made with 48 cultures of staphylococci, microcci, and streptococci. Most of these were halotolerant, being able to grow in from 0 to 10% NaCl. No significant interrelationships were found between these and the halophiles studied. This work brought out several physiological features of extreme halophiles: The rods had a wider pH growth range than the cocci; as a group, they liquefied gelatin slightly better than the cocci and hydrolyzed starch much less well. Sixty-seven percent of the halophilic rods, but none of the cocci, hydrolyzed casein. None of the extremely halophilic bacteria were sensitive to penicillin, chloramphenicol, terramycin, aueromycin, or erythromycin, though some of the moderate halophiles were sensitive to polymyxin B, tetracycline, and dihydrostreptomycin, sensitivity to these antibiotics occurring more frequently in the rodsa2 Colwell and Gibbons suggest that further work on the interrelations between halophiles should include studies of nucleic acid composition. It may be significant that H. salinarium contains about 20%of its DNA as a satellite band of lower density than the main component, and that H. cutimbmm contains about 10%of such DNA (Joshi et al., 1963). Larsen (1967) made the interesting suggestion that DNA with a high guanine cytosine content might be characteristic of extreme halophiles. This has been established thus far only for the halobacteria.

+

Ill. The Nature of Halophilic l i f e

A. RESISTANCE AND ADAPTATIONTO SALT One of the first questions still asked about halophiles is: Do these organisms have a very active pumping mechanism for excluding salts, so that their interior environment is of low ionic strength? This question has been studied in some detail (see reviews of Ingram, 1957; Larsen, 1962, 1967; Brown, 1964; Kushner, 1964a), resulting in the answer that in general these organisms do not exclude salt at all and that their internal environment has as high an ionic strength as their 2There is not a great deal of work available on growth-inhibitoryor lethal agents for extreme halophiles. Kushner et al. (1965) suggested that their great sensitivity to low pH values could be made the basis of a simple method of sterilizing solar salt.

79

HALOPHILIC BACTERIA

external environment, though they may actively select the ions found inside. This conclusion follows from measurements of the freezing points of cells grown in different salt concentrations, from lytic experiments on cells grown in different salt concentrations, and from chemical determinations of the internal ionic contents. One striking, if indirect piece of evidence is the simple fact that extreme halophiles can be centrifuged down. Ingram (1957) pointed out that a high internal salt concentration would be necessary to make such cells denser than the high salt medium in which they were suspended. An interesting example of another mechanism of adaptation was found in the ciliate, Tetrahymena pyriformis which, when adapted to 0.2 M NaCl could maintain a lower inner than outer salt concentration (Dunham, 1962). Although the overall internal salt concentration in halophilic bacteria is equal to that of the external medium, the ionic composition may be quite different. The small amount of data available on this very important point (summarized in Table I, from the paper of

INTRACELLULAR

Organism:

NaCl in medium ( M ) KCI in medium ( M ) Na+ in cells K + in cells C1- in cells

TABLE I IONICCONCENTRATIONS IN NONHALOPHILIC AND HALOPHILIC BACTERIA^

Staphylococcus auTeuS

Vibrio costicolus

Halobacterium salinarium

Sarcina morrhuae

0.15 0.025 0.098 0.680 0.008

1.0 0.0004 0.684 0.221 0.139

4.0 0.032 1.37 4.57 3.61

4.0 0.032 3.17 2.03 3.66

From Christian and Waltho (1962). Intmcellular concentrations are expressed as moles per milligram cell water.

Christian and Waltho, 1962) suggests that extremely halophilic bacteria have, in a very high degree, the ability displayed by other cells to concentrate K + and to exclude Na+. Christian and Waltho (1961) also found that in bacteria that tolerated but did not require salt, there was a correlation between internal K+ content and tolerance to NaC1.

B. ENZYMESTUDIES What we know so far of the physiology of extreme halophiles confirms that these organisms are not resistant to salt but rather are fully

80

D. J. KUSHNER

adapted to life in a very salty internal and external environment. This adaptation may be seen at the molecular level and at the level of subcellular particles. About twenty different enzymes of extreme halophiles have been studied. These include cytochrome oxidase, cysteine desulfhydrase, and several dehydrogenases of H. salinarium studied by Baxter and Gibbons (1954, 1956, 1957), who pioneered work in this field; the catalase and glutamic-aspartic transaminase of H. salinarium; these enzymes and dehydrogenases of S. morrhuae, studied by Aasmundrud and Larsen (Larsen, 1962);arginine desimidase, condensing enzyme, alkaline phosphomonoesterase and other enzymes of H. salinarium studied by Holmes et al. (1965);and nicotinamide adenine dinucleotide oxidase ofH. salinarium (Hochstein and Dalton, 1968). In general, all these enzymes can function in high salt concentrations, and most of them are inactive in the absence of salts. Potassium ions are usually more effective activators than sodium ions. Other cations, and other anions than chloride may also be effective in maintaining enzyme activity, though most of the work has been carried out with chlorides. Baxter (1959)proposed that salts decreased electrostatic repulsion between ionized groups in the molecule and thus affected protein conformation. Katchalsky (1954) pointed out that in polyelectrolyte gels neutral salts can form Debye atmospheres around the charged groups which can diminish both attractive and repulsive electrostatic forces. Changing such forces can make the gels perform mechanical work. The idea that salts act by shielding mutually repulsive groups on proteins is basic to studies of extreme halophiles; but, though it seems eminently reasonable, direct evidence on this point is still difficult to obtain because the proteins themselves have not yet been purified. To my knowledge, purification of any halophilic enzyme to the point that it can be used for sequence analysis has not yet been reported, though such work is being carried out in at least one laboratory (R. R. Colwell, personal communication). The inactivation of enzymes of extreme halophiles when salt is removed was first thought irreversible. However, Holmes and Halvorson (1965) showed that this inactivation can be reversed if the salt is added back slowly, as by dialysis against a salt solution, but not if it is added back quickly. Studies by this method of the rate of enzyme reactivation showed that it is a complex process, possibly involving conformational changes in the enzyme protein. The presence of high salt concentrations makes some of the standard techniques of protein separation, such as ammonium sulfate fractiona-

HALOPHILIC BACTERIA

81

tion and column chromatography, difficult or impossible, and this has delayed progress in studying enzymes of halophiles. Thus, the discovery of Holmes and Halvorson (1965) is of particular importance; by establishing a technique that made it practical to restore most of the activity of salt-free enzymes, these workers also made it possible to attempt a separation of the salt-free protein. They have purified the malic dehydrogenase of H . salinarium about 1000-fold by this method. Enzymes of moderate halophiles are, as might be expected, less salt-tolerant and less salt-dependent than those of extreme halophiles. Some are partly inhibited by the salt concentrations in which the cells grow best. However, many or most of them require some salt for activity (Larsen, 1962; Kushner, 1964a).

c. RIBOSOMES AND PROTEIN SYNTHESIS The fact that the extreme halophiles have a very high internal salt concentration prompted an investigation of the ribosomes of one of these organisms, H . cutirubrum (Bayley and Kushner, 1964). Their nature was of special interest, since it was reasoned that the type of ribosome found in other plants, animals, and microorganisms should be dissociated into protein and nucleic acid by such high salt concentrations. However, not only were the ribosomes of H . cutirubrum not dissociated by high salt concentrations, but they required them for stability. In view of the high K+ content inside their cells (see below), it seems significant that ribosomes specifically required a high KC1 concentration to remain in what is probably their physiologically active form. In 3-4 M KCl plus 0.1 M MgClz ribosomes existed in the 70 S form. If either the KCl or the MgClz concentration was lowered to 2.0 M or 0.01 M, respectively, the ribosomes dissociated into the 31 S + 52 S forms. It is well known that other ribosomes need Mg++ for stability, but in much lower concentrations. Those of Escherichia coli maintain the 70 S form in presence of only 0.001M Mg++,and need no other metallic cation. On greater dilution of the salts, the ribosomal subunits dissociated still further into soluble protein plus particles richer in RNA than the original ribosome (RNA: protein 80:20, as compared with 60:40). Electrophoresis of these showed that there were several protein components which, in contrast to proteins of other ribosomes, were acidic instead of basic. Bayley and Kushner (1964) proposed that the unusual behavior of these ribosomes, and their requirement for high salt concentrations for stability was due to their high content of acidic proteins and that

82

D. J. KUSHNER

K + and Mg++were necessary to shield mutually repulsive negative charges or to provide binding between negative groups. The requirement for K+ was quite specific. Substitution of Na+ for K+ led both to disaggregation and to a rather random assortment of particles heavier than the 70 S particles. Other monovalent cations could not substitute for K+. It was suggested that the size of the monovalent cation was very important to assure a proper fit together of the component parts of the ribosomes. Although most of the protein left the ribosomes on extensive dilution, and could be shown to be acidic, some remained. Bayley (1966a) found by amino acid analysis that the protein remaining bound to RNA was basic. He also found that K + was weakly bound, and Mg++ more strongly bound, to these particles, and suggested that Mg++ acted on the RNA rather than on the protein or on a combination of the two. On raising both the K+ and the Mg++concentrations, by dialysis against these salts, it was possible to reconstitute ribosomal particles from very dilute solutions into a mixture of 70 S, 52 S, and 31 S subunits (Bayley, 1966b). Though these had the same sedimentation values as the original ribosomes and subunits, the reconstituted particles still lacked some of the protein of the untreated ribosomes. Bayley and Griffiths (1968a) pointed out that cations are already known to be involved in protein synthesis. The structural integrity of ribosomes, the binding of messenger RNA to ribosomes, and the activity of transfer enzymes are all affected by Mg++.The binding of transfer RNA to ribosomes depends on NHs+ and K+. They suggested that the high ionic environment inside extremely halophilic bacteria might be expected to affect these structures (as was already known for ribosomes) and their interactions. They prepared a cell-free system from H. cutlmbrum that incorporated each of 20 amino acids into hot trichloroacetic acid-insoluble material. This incorporation apparently represents true polypeptide synthesis. It required ribosomes and other components generally needed in protein-synthesizing systems; it was sensitive to puromycin and ribonuclease, but not to deoxyribonuclease. Phenylalanine incorporation was stimulated by polyuridylic acid. The optimum Mg concentration for incorporation was 0.02-0.04M , which is not greatly different from that in other incorporating systems, but the system was halophilic in requiring 3.8 M KCl, 1.0 M NaCl and 0.4M NH&l for full activity. In further work on this system Bayley’s group (papers in preparation by Griffiths and Bayley; Bayley and Griffiths; Rauser and Bayley) ++

HALOPHILIC BACTERIA

83

has shown that distinct synthetase and transferase enzymes are involved in amino acid incorporation; that incorporation is associated with polyribosome-like material; that amino acyl synthetases, though halophilic, are less specific in their ionic requirements than the whole incorporating system. Bayley (1966a) suggested that the acidity of H. cutirubrum proteins might be caused by a salt-induced misreading of codons for basic amino acids. Though this has not been tested directly, Bayley and Griffiths (196813)found recently that the codon assignment for phenylalanine, proline and other amino acids was the same in H. cutirubrum as in nonhalophilic organisms. However, Na+ slightly affected the fidelity of translation.

D. EXTERNALLAYERSOF EXTREMEHALOPHILES Recently there has been much interest in the external layers of halobacteria, partly because of their relation to those of marine bacteria, but mainly because of the possibility, seriously and perhaps too hopefully considered by students of membrane physiology, that these layers represent pure cytoplasmic membranes of relatively simple structure. Practically all the work on external layers of extreme halophiles has been done on the halobacteria rather than on the halophilic cocci. The latter organisms do not lyse as dramatically as the halobacteria on dilution; they are much tougher and very difficult to break mechanically (R. R. Colwell, personal communication). In contrast, the halobacteria are so easy to break mechanically or sonically (I have found a 5-second burst of ultrasound to cause extensive cell disintegration of H. cutirubrum, for example) that the problem is not to break them, but rather to keep them from breaking into too small fragments. A proper general survey of outer layers of extreme halophiles should certainly include more work on the rather unobliging cocci.

1 . Active Transport The great concentration difference between internal and external

K + in extreme halophiles suggests that a very active transport system is in operation, but so far this has hardly been studied. In an earlier review (Kushner, 1964a) I pointed out that in Christian and Waltho’s (1962) work K + content had been determined only late in the growth phase and that it was known from other bacteria that the K + content could vary greatly during the growth cycle. We have since found, however (Gochnauer and Kushner, 1968) that the K + content of

84

D. J. KUSHNER

several halobacteria varies relatively little from the early logarithmic to the late stationary part of the growth cycle. When cells were washed and suspended in 25% NaCl, their K + content remained constant for several days and the contents of a sedimentable suspension only fell in proportion to the number of cells that had died. This could indicate that a high K + content is necessary for cells to live or, more simply, that death, as in many other cells, is accompanied by leakage. Adenosine triphosphatase (ATPase) is known to be closely associated with active transport in other cells (Kleinzeller and Kotyk, 1961). Preliminary work has been done on the function and properties of this enzyme in extreme halophiles. We found (Korngold and Kushner, unpublished) very low ATPase activity in H. cutirubrum envelopes broken mechanically in high salt solutions. However, De et al. (1966) found a trinucleotidase in the membranes of H. halobium that hydrolysed ATP, inosine triphosphate (ITP), and uridine triphosphate (UTP), at nearly equal rates and ADP at half this rate. This enzyme is activated by Mg++,Na+, or K + and from the brief -account given, it does not appear to be a so-called “Mg++dependent, Na+ K+-activated ATPase,” the enzyme complex which can be demonstrated under special conditions of preparation, and which is thought to be closely tied in with active transport (Kleinzeller and Kotyk, 1961). Drapeau and MacLeod (1963) found an ATPase in a marine bacterium, which was also activated either by Na+ or K+, but the effects of these two ions were not additive. De et al. (1966) made the important observation that exposing membranes to sufficiently low salt concentrations could completely and irreversibly inactivate their ATPase without causing membrane dissociation. Studies of the uptake of carbon compounds by extremely halophilic and marine bacteria have helped explain the role that Na+ may play in these bacteria. Both extremely halophilic and marine bacteria have a specific and usually high Na+ requirement for growth (Larsen, 1962, 1967; MacLeod, 1965), even though other ions and solutes may serve to protect the cells from lysis. MacLeod showed that in a marine pseudomonad Na+ was needed for the active transport of a-aminoisobutryric acid, a nonmetabolizable amino acid. Stevenson (1966) showed that glutamate uptake in H. salinarium, a process dependent on active transport, required a high external Na+ concentration. It is attractive to think that the special requirement for Na+ is due to its role in active transport of other substances, but more work is required

+

85

HALOPHILIC BACTERIA

with several other microorganisms before such a generalization can be made.

2. Cell Lysis and Envelope Disintegration Some of the more interesting properties of the external layers of the halobacteria are revealed when the salt concentration in which these organisms are suspended is lowered. At 3.0-4.0 M NaCl the organisms are rod-shaped. As the salt concentration is lowered, they become first swollen and irregular, then spherical. Below between 1 and 2 M NaC1, depending on the strain used, envelopes disintegrate and intracellular substances are released. If envelopes prepared by breaking cells mechanically in high salt concentrations are placed in water, the envelopes disintegrate partially or completely into particles too small to be seen in the light microscope or to sediment after about 1 hour at 15,000-20,000 g. Different workers (Brown, 1963; McClare, 1967; Stoeckenius and Rowen, 1967) found somewhat different amounts of H . halobium envelopes still sedimentable at these forces after the envelopes had broken up in water. McClare (1967) pointed out that the sedimentable fraction appeared mainly in envelopes from cells harvested late in the growth phase. The process of cell lysis and envelope disintegration has been studied in some detail (reviewed in Larsen, 1962, 1967; Brown, 1964; MacLeod, 1965) and the following generalizations can be made. 1. Osmotic forces play a very important part when extremely halophilic bacteria, which have a concentrated internal environment, are placed in a much more dilute solution. This may also be true of marine bacteria (MacLeod, 1965), which have a smaller concentration difference possible between the inside and the outside. However, the effect of different solutes in preventing lysis of intact cells cannot be explained entirely on the basis of the osmotic pressure they exert. Thus, high concentrations of sugar cannot preserve the integrity of extremely halophilic bacteria and may or may not be as effective as equiosmolar salts in protecting marine bacteria from lysis (Brown, 1964; Larsen, 1962, 1967; MacLeod, 1965; see also Korngold and Kushner, 1968, for a description of a marine psychrophile with very specific ionic requirements for integrity). Brown (1964) pointed out that it is difficult to separate the specific effects of ions from their water activity but that “where a relatively high concentration of a monovalent ion can be replaced by a relatively low concentration of a bi- or multivalent ion . . it is safe to conclude

.

86

D. J. KUSHNER

that the direct ionic effect is major and that water activity exerts, at most, a minor influence.” In extremely halophilic, as well as in marine, bacteria Mg++ salts are more effective than Na+ salts in preventing lysis of intact cells. To take one example, 0.5 M MgCl2 is more effective in preventing lysis of H . cutirubrum than is 1.0 M NaC1; furthermore, 0.05 M MgCl2 is more effective than 1.0 M NaCl in preventing dissolution of envelopes (Kushner, 1964b). Such experiments indicate a dual role of salts in halophilic bacteria: to maintain osmotic pressure and to maintain envelope structure. Apparently, the lower Mg++ concentrations are sufficient to prevent envelope dissolution (the manner in which they do this will be considered below) but not to overcome the physical forces of osmotic pressure. This phenomenon has been used by Brown et al. (1965) as the basis of a very simple method to prepare large quantities of H . halobium envelopes. Lysis of cells in 0.02 MgCl2 opens up the cells but leaves intact envelopes which resemble those prepared mechanically. This review deals only with bacteria, and some of the generalizations made may not apply to other microorganisms. A number of so called “osmophilic” yeasts require a high osmotic pressure, which can be supplied by sugars (Ingram, 1957; Onishi, 1963). Vaisey (1954) and Ormerod (1967) reported that the halophilic mold, Sporendonema epizoum could grow in between 0.8 M and 4.5 M NaCl, but that glucose could replace NaC1. 2. In water, the envelopes of halobacteria disintegrate very rapidly (though not necessarily completely-see above), even in the cold, into particles with 4 S-5 S sedimentation values. Raising the salt concentration again may cause a rather random agglutination of these particles, but certainly not a re-formation of envelopes (Onishi and Kushner, 1966). Because of the speed of dissolution, the fact that it does not involve peptide bond splitting, release of low molecularweight substances or breaking of protein-lipid bonds; and the fact that it is not prevented by several metabolic inhibitors, the process of dissolution is considered to be nonenzymic. 3. The proteins of the envelopes of all halobacteria studied ( H . cutirubrum, H . salinarium, H . halobium) are acidic, or at least have more acidic than basic amino acids. This has been shown by direct analysis (Brown, 1963; Kushner and Onishi, 1966; Larsen, 1967); Brown (1963) also showed by electrophoresis that the envelopes of H . halobium were negatively charged. Workers on halophilic bacteria generally agree that salts are needed to shield such negative charges,

HALOPHILIC BACTERIA

87

and that in the absence of salts mutual repulsion between these charges causes envelopes to break up. Kushner and Onishi (1966) pointed out that lipids could also contribute to the negative charge of envelopes, though in H. cutirubmm the lipids contained only 1/7 as many negatively charged groups as the proteins. They showed, however, that removing the lipid did not decrease the salt requirement for stability, but that removing the protein left residues, containing lipid and carbohydrate, which were quite stable in distilled water. These results suggested that mutual repulsion was due to groups on the proteins alone. 4. Halobacteria are unusual in that their envelopes lack muramic acid and hence the mucopeptide found in all other bacteria exc:ept the very plastic pleuropneumonia-like organisms (Salton, 1964). They contain some carbohydrate, but are predominately lipoprotein. These observations supported the first electron microscope studies that indicated that H. halobium and sometimes H . salinarium were surrounded only with a “unit” cytoplasmic membrane (Brown and Shorey, 1963; further work on this point is discussed below). Since in other bacteria mucopeptides provide mechanical rigidity and shape, these observations brought up the question, still unresolved, of how the halobacteria maintain their rod shape. Brown (1964) suggested that the high concentration of sodium ions alone could stiffen the lipoprotein envelope. Kushner and Onishi (1966) pointed out that on removal of protein and lipid from H . cutirubrum envelopes, fractions rich in carbohydrate remained, roughly the same shape as untreated envelopes. The function, if any, of carbohydrates in preserving cell shape is still not understood, but should be further investigated. Data on cell walls of the halophilic cocci are still lacking. M. Masui and E. Ohtani found that envelopes of a moderately halophilic coccus and of a pseudomonad contained both muramic acid and diaminopimelic acid (personal communication). E. DIFFERENTIAL EFFECTSOF NA+ AND K + ON CELLS AND ENVELOPES

Although high concentrations of monovalent salts are needed to prevent lysis of extreme halophiles, all salts do not act equally well. Abram and Gibbons (1961) found that NaCl and LiCl were much more effective in maintaining intact cells of H. cutimbrum than were KC1 or NH4C1. Kushner (1964b) found that NaC1, KC1 and NH4C1 were about equally effective in maintaining envelopes of H. cutirubrum.

88

D. J. KUSHNER

Soo-Hoo and Brown (1967) found a similar phenomenon in H . halobium, and Buckmire and MacLeod (1965) also observed this in a marine pseudomonad that is of special interest in that its envelopes also undergo a partial dissolution in the absence of salts, although they require a lower salt concentration for stability than do halophile envelopes. Different explanations have been advanced for this phenomenon. Onishi and Kushner (1966) suggested that Na+ and K+ acted on different sites on the external cell surface so that Na+ prevented leakage better than K+, whereas both preserved the overall structure of envelopes. In fact, it was found that after lysis in 2.0 M KCL intact envelopes of H . cutirubrum remained. Soo-Hoo and Brown (1967) criticized this interpretation and suggested that the difference was merely due to the difference in the ability of these ions to penetrate the cell. Several facts argue against their interpretation. One is that, in H. cutirubrum at least, a sufficiently high KC1 concentration (3-4 M ) can prevent lysis (Abram and Gibbons, 1961; Kushner, 1964b). If K+ were acting as a penetrating solute, one would not expect it to give osmotic protection however high the concentration used (see Davson, 1964). Second, extreme halophiles maintain very high internal K + concentration even when suspended in 25% NaCl. It is very difficult to envisage external K+ penetrating against a saturated inner K+ concentration. In trying to explain the different effects of Na+ and K + on intact cells, Kushner (1964b) did consider the possibility that an active K + uptake caused lysis but ruled this out because lysis in 2 M KCl took place as rapidly (that is, in less than a second) at 0°C at 37°C.

F. CELL FINE STRUCTURE AND MOLECULARARCHITECTURE OF HALOBACTERIUM ENVELOPES Extremely halophilic bacteria have always presented a challenge to the morphologist because of the difficulty of fixing them for microscopic examination. Simple heat fixation and staining is apt to result in nothing but a collection of stained salt crystals. Dussault (1955) proposed drying salt suspensions of bacteria on a glass slide and fixing the cells by immersing the slide in dilute acetic acid. Such cells can be stained but are still distorted. Because of the difficulties of fixation, all results should be checked b y comparing fixed cells with living ones seen under phase contrast. The first electron microscopic demonstration of surface structure of halobacteria was that of Houwinck (1956), who took advantage of the fact that H. halobiurn contains vacuoles and is therefore sufficiently

HALOPHILIC BACTERIA

89

thin in places for the electron beam to pass through. Later work (Kushner and Bayley, 1963; Kushner et al., 1964; Mohr and Larsen, 1963) used isolated envelopes or replica techniques on whole cells to show the surface structure. A regular arrangement of spherical particles was found on the surfaces of H. cutirubrum, H. salinarium, and H. halobium. All fixation methods used so far, to my knowledge, have caused some distortion of this regular arrangement, so that, despite the beautiful detail shown in the most recent and detailed work on thin sections ofH. halobium (Cho et al., 1967; Stoeckenius and Rowen, 1967) and the demonstration of structures found in more orthodox bacteria, the available photomicrographs should still be interpreted with caution. Brown and Shorey (1963)studied the fine structure of thin sections of H . halobium and H . salinarium and concluded that the former was surrounded only b y a unit membrane, though the latter might be surrounded by a more complex membrane. The suggestion that H. halobium could provide a source of unit membranes stimulated further work on this organism but, perhaps unfortunately, more recent fine structure studies, using other fixing techniques, indicate that H. halobium probably has a cell wall after all. Stoeckenius and Rowen (1967) found that their strain of H . halobium was surrounded by two layers, which look like those structures designated as cytoplasmic membrane and wall in other bacteria. They also found that stepwise reduction of the salt concentration in which cells were suspended caused a release of cell wall material before the cytoplasmic membrane began to disintegrate. They suggested that the 4 S component, which is known to appear when envelopes of H. halobium disintegrate in water (see above), comes from the wall rather than the cytoplasmic membrane. They also found evidence for curious intracytoplasmic membranes that seemed, from their distribution in different strains, to be related to gas vacuole formation. More recently, Larsen et al. (1967) reported the isolation of gas vacuoles from H. halobium. These appear to be bounded by a single layer. Cho et al. (1967)found that the envelope of H . halobium was more complex than a single unit membrane; they also suggested that the cell was surrounded by a cytoplasmic membrane and a wall. They found an unusual organelle inside the cytoplasm in the form of parallel striated strands. These may or may not be similar to the intracytoplasmic membranes described by Stoeckenius and Rowen (1967). Recently, attempts have been made to correlate chemical studies with electron microscope data and to understand the molecular

90

D. J. KUSHNER

architecture of halophile membranes. The latter will be discussed only briefly, since a detailed consideration of present theories of membrane structure, to which these studies contribute, is beyond the scope of this review. Larsen (1967) described the isolation and partial characterization, by workers in his group, of a homogenous lipoprotein particle fraction from H. salinarium envelopes dissolved in water. This fraction contained most of the protein in the envelope, though it had lost the glucosamine present in the whole envelope. The protein had a predominance of acid groups. In the electron microscope these lipoprotein particles appeared to be between 100-200 A in diameter, a size consistent with their being the spherical particles regularly arranged on the surface of H.salinarium. Further work seems necessary to clarify the relation between this lipoprotein, the 4-5 S particle released when H. halobium and H . cutirubrum envelopes disintegrate in water, and the spheres on cell surfaces. Brown (1965) examined the molecular distribution of acidic and basic groups in H . halobium envelopes by titration. He found that dissolution of envelopes in water led to the appearance of titratable basic groups, whereas dissolution did not change the titratability of the carboxyl groups. These studies suggested that the basic groups were buried below the membrane surface, possibly bound to phospholipids, whereas the acidic groups were at or near the surface. More recently, Brown and Netschey (1967) studied the proteins of H. halobium envelopes after removal of lipids and treatment with 8 M urea and found two types of protein, of molecular weights approximately 340,000 and 97,000. Studies of the interaction of Mg++with lipids and proteins of halophile envelopes have indicated the important structural role that this ion, required in relatively high concentration for growth, can play. Kushner and Onishi (1966) found that removal of lipids from H. cutirubrum envelopes greatly increased the Mg++ concentration required for stability. They took this to mean that lipids were involved as Mg++binding sites. Rayman et aZ. (1967) found that 1 mole of Mg++ was bound per mole of phosphatidyl glycerophosphate (diether analog) in cells of H. cutirubrum. McClare (1967) studied binding between lipid and protein, and the role of Mg++ in this binding, in H. halobium envelopes by examining the partition of these components on dialysis and on extraction with lipid solvents. His results suggested that lipids are bound hydrophobically to one group of proteins and are bound to another group of proteins by polar bonds. It seemed possible that Mg++formed a chelate between lysine and

HALOPHILIC BACTERTA

91

glutamic acid groups on the protein and lipid diphosphate head groups and that an ionic link also existed between the terminal lipid phosphate group and an arginyl residue on the protein. IV. Taxonomic Importance of Halophilic Lipids

Studies of the lipids of extremely and moderately halophilic bacteria have revealed what may be a characteristic of great taxonomic and evolutionary importance and have shown that some of the metabolic pathways in the extreme halophiles may be quite different from those in other bacteria. The lipids of H . cutirubrum consist almost entirely of derivatives of a glycerol diether (Kates et al., 1965a,b; Joo and Kates, 1968). The major component is a diether analog of phosphatidyl glycerophosphate and a minor component is a diether analog of phosphatidyl glycerol. In H . cutirubrum these compounds are located almost entirely in the cell envelope (Kushner et al., 1964). In addition, a diether-containing glycolipid and sulfolipid were found (Kates et al., 1967). Besides having the two ether groups attached to glycerol, these lipids are unusual in containing dihydrophytol groups rather than fatty acids. These two unusual characteristics appear limited in nature to lipids of extremely halophilic bacteria. Kates et al. (1966) carried out a survey of the lipid composition of several extremely halophilic and moderately halophilic bacteria. The results (Table 11) showed two major differences between extreme halophiles and all other bacteria: the extreme halophiles contained much higher amounts of unsaponifiable material (representing mainly the diether) and much lower, indeed only trace, amounts of fatty acids (see also Cho and Salton, 1966). Chromatographic and infrared spectroscopic studies showed that only extreme halophiles contained diethers. The colorless rod, A-31C, which had been maintained in 25% NaC1, was first taken to be an extreme halophile, but after it was shown to lack the diether, a reexamination of its salt requirements showed that it grew best in 7.5-2076 NaCl and hence could be considered a moderate halophile. The colorless rod, A-212, however, grew best in 25% NaCl and not at all below 15% NaCl, and hence could be considered an extreme halophile. The very low level of fatty acids in extreme halophiles is another striking characteristic of these organisms. Incorporation studies using growing cultures of H. cutirubrum have shown that acetate-2-14Cis strongly incorporated into lipids, most of the activity appearing in the dihydrophytyl chain but not in the fatty acids. Very little incorpora-

92

D. J. WSHNER

tion of malonate-2-14C,a straight chain fatty acid precursor, was observed (Kates et al., 1968). These observations are consistent with the unusual lipid composition of H. cutirubrum. Further work is being carried out on the pathways leading to the formation of the major diethers. TABLE I1 LIPIDSOF HALOPHILIC AND NONHALOPHILIC BACTERIA^ Lipid analyses (% of total lipids)

P Organism

Extreme halophiles H . cutirubrum H . halobtum M H . halobtum P H . salinarium S. littoralis Sarcina sp. Halophile A-2c Moderate halophiles Halophile A-31c M . halodenttrijicans (48-hour culture) M . halodenttrfjicans (20-hour culture) v. costicolus Nonhalophiles S . lutea

s.flava

Unsaponifiable material

Fatty acids

4.25 4.32 3.83 3.37 3.70 4.23 4.41

70.4 73.6 68.7 66.5 64.8 67.8 70.7

0.6 0.4 0.3 0.7 2.3 0.3 0.3

3.80 3.33 3.74 3.76

13.8 0.5 9.1 5.2

59.8 62.6 68.3 60.5

2.89 3.28

13.7 13.5

68.3 59.6

aFrom Kates et al. (1966).

One among other questions still to be answered about halophilic lipids is: How are these broken down in nature? So far, we have been able to find no evidence that H. cutirubrum breaks down its own phospholipids, nor that several species of marine bacteria can break down these phospholipids. The fate of these compounds is still unknown. V. Growth and Survival of Extreme Halophiles

Extreme halophiles grow slowly, with mean generation times of 7-15 hours, and they grow best at temperatures of 37°C. or higher. Defined growth media containing several amino acids have been described for H. salinarium by Dundas et al. (1963) and for H. cuti-

93

HALOPHILIC BACTERIA

rubrum by Onishi et al. (1965). Other modifications in this medium are described by Kushner (1966a); see also Larsen (1967). Onishi et al. (1965)found that vitamins and carbohydrates did not stimulate growth in their medium. The K+ content of this medium is relatively low, and we have found (M. B. Gochnauer and D. J. Kushner, unpublished) that if the K+ content is increased, both mixtures of vitamins and certain sugars stimulate growth. Growth has previously been measured only as turbidity. Measurements of viable counts showed that in several halobacteria there is no loss of viability for several days after the beginning of the stationary growth phase. This stability is also displayed in salt solutions without added nutrients. Viable counts of cell suspensions shaken at 37°C. in a solution containing NaCl, MgS04, KC1, and other salts did not decrease for over a week. Figure 1 shows the survival of H. halo-

I

\

Complete so115

T-K

- K Mg

b-Mg

Co, Fe

51

2

6

4

8

Days

FIG. 1. Survival of H . hnlobium suspended in different salt solutions. Cells were washed in 25% NaCl and suspended in solutions containing all salts (NaCI, 22.5%; KCI, 0.2%; MgS04.7Hz0,2.0%; FeSO4.7HzO,0.0049%; CaCl2.7H2O,40 p.p.m.) or with omissions as shown.

bium; similar results were found with H. salinarium and H . cutirubrum. Removing KCl did not cause any loss of viability, though removing Mg++ did (M. B. Gochnauer and D. J. Kushner, unpub1ished).

94

D. J. KUSHNER

VI. Speculations on Moderate Halophiles Brown’s (1964) review considers some of the properties associated with growth at different salt concentrations. His conclusions were limited by the scarcity of data, but he noted that a few examples showed a rough negative correlation between salt tolerance and amino sugar content. There was more convincing evidence for increase in the acidity of membrane proteins in organisms associated with higher salt concentrations. As the previous discussion shows, the increase in acidity of proteins is a very general characteristic of extreme halophiles. In comparing different species grown under different conditions it is not always possible to separate genetic from environmental effects. The moderate halophiles, which can grow over an &fold range of salt concentration, may help provide unequivocal answers to the question: What are the physiological responses to salt? In comparing two cultures of the same organism grown at different salt concentrations, one of the first questions to be answered is: Are such cultures really homogeneous in their salt tolerance or salt requirement, or does growth of an inoculum at different salt concentrations rather reflect a selection of cells best able to grow at each concentration? If there is such a selection, then we might expect inocula from cultures grown at higher salt concentrations to contain more cells able to grow at these concentrations (and perhaps fewer able to grow at lower concentrations) than inocula from cultures grown at low concentrations. We tested this possibility (M. Forsyth and D. J. Kushner, unpublished) by studying the growth pattern of Vibrio costicolus and Micrococcus halodenitr$cans in complex media (proteose peptone-tryptone) at different salt concentrations. The growth pattern of both was the same, whether the inoculum consisted of cells grown in 3.0 M or 0.5 M NaCl (Fig. 2 shows growth of M. halodenitrijicans; very similar results were found for V. costicolus). The same results were obtained when the inocula were transferred several times on high or low NaCl concentrations before the experiment. These experiments argue against a heterogeneity in salt response of these cultures. These two organisms were found previously to lack the characteristic lipids of extreme halophiles when grown in 1.0 M NaCl (Kates et al., 1967; see Table 11). We have found since (M. Forsyth, M. K. Wassef, and D. J. Kushner, unpublished) that even if these bacteria

95

HALOPHILIC BACTERIA

were grown in 3.0 M NaCl, a concentration in which extreme halophiles can grow, they did not form the diether-containing phospholipid.

Hours

Hours

FIG. 2. Growth of M. halodenitrijicans in different salt concentrations. Ten-milliliter lots of growth medium (proteose peptone-tryptone, containing the molar NaCl concentrations shown by each curve) were inoculated with 0.02 ml. of an overnight culture in 0.5 M or 3.0 M NaC1-proteose peptone-tryptone broth.

To my knowledge, no experiments exist on the effect of salts on the composition of proteins of moderate halophiles. L. Hochstein (personal communication) found that the salt response of NADHz oxidase of M. hulodenitrijicuns was the same for cells grown in several different salt concentrations. This may indicate that the acidity of the proteins involved is the same at each concentration. We have recently examined the effects of salts on the nutrition of V. costicolus (M. Forsyth and D. J. Kushner, unpublished). This organism was chosen partly because a simple amino acid medium exists for it (Flannery and Kennedy, 1962).We found that it can also grow in a glucose-ammonia-salts medium, in a very narrow range of salt concentration (near 1.0 M), and that, in general, the richer the medium, the wider the salt range permitting growth. The widest

96

D. J. KUSHNER

range of growth was possible only in a complex medium. These results suggested that some synthetic enzymes are inactivated by too high or too low salt concentrations, so that outside a narrow optimal range the end products of enzyme action must be supplied. VII. On the Origin of Extreme Halophiles This subject is discussed in some detail in recent reviews. Larsen (1967) describes an attempt by K. Middleton in his laboratory to transform a number of different bacteria into extreme halophiles. This exeriment, which seems more sophisticated than many of this type, in that mutagenic agents were used, gave negative results. Other negative experiments are reviewed by Larsen (1962) and some equivocal ones by Ingram (1957)and Kushner (1964a). A certain amount of adaptation to higher salt concentrations is possible (see reviews cited, and Limsong and Frazier, 1966) but so far no one has made an extreme halophile in the laboratory. Larsen (1967) points out that in order to give a cell an extremely halophilic character a large number of proteins must become acid, that is, a large number of mutations are required, which probably “have taken place in nature over a very long period of time.” Production and selection of the appropriate mutants in the laboratory is understandably difficult. Few attempts have been made to “train down” extreme halophiles to grow in a lower salt concentration, though this procedure has the theoretical advantage that the worker is starting with an organism he is interested in -an especially important point since there probably are only two species of extreme halophiles. I have made such an attempt with several halobacteria, so far with negative results. A few cells of H.cutirubrurn could survive on plates made up in 10% NaCl long enough to grow when the salt concentration was raised by the plates drying, but no bacteria could be adapted to grow in lower than 15% NaC1. Mutagenic agents were not used in these experiments. An unsuccessful attempt was also made by H. Brown in Gibbons’ laboratory several years ago (unpublished communication from Dr. N. E. Gibbons). In future studies of the origin of extreme halophiles, their lipid composition should certainly be kept in mind. The presence of a diether-containing phospholipid and the absence of fatty acids seem to be signatures of extremely halophilic character in bacteria. They should be used as a clue in any ecological survey.

HALOPHILIC BACTERIA

97

REFERENCES Abram, D., and Gibbons, N. E. (1961). Can. J . Microbiol. 7 , 741-750. Baas-Becking, L. G. M. (1931). Sci. Monthly 32,434-446. Baxter. R. M. (1959). Con.J . Microbiol. 5, 47-57. Baxter, R. M., and Gibbons, N. E. (1954). Can. J . Biochem. Physiol. 32, 206-217. Baxter, R. M., and Gibbons, N. E. (1956). Can.J . Microbiol. 2,599-606. Baxter, R. M., and Gibbons, N. E. (1957). Can.J . Microbiol. 3,461-465. Bayley, S . T. (1966a).J. Mol. Biol. 15, 420-427. Bayley, S. T. (1966b).J. Mol. Biol. 18, 330-338. Bayley, S. T., and Griffiths, E. (1968a).Biochemist% 7,2249-2256. Bayley, S. T., and Griffiths, E. (1968b). Can.]. Biochem. 46,937-944. Bayley, S . T., and Kushner, D. J. (1964).J . Mol. Biol. 9,654-669. Breed, R. S., Murray, E. G. D., and Smith, N. R. eds. (1957). “Bergy’s Manual of Determinative Bacteriology,” 7th Ed. Williams & Wilkins, Baltimore, Maryland. Brown, A. D. (1963). Biochim. Biophys. Acta 75,425-435. Brown, A. D. (1964). Bacteriol. Reo. 28,296-329. Brown, A. D. (1965). J . Mol. Biol. 12, 491-508. Brown, A. D., and Netschey, A. (1967). Biochem. J . 103, 24-28. Brown, A. D., and Shorey, C. D. (1963).]. Celi. Biol. 18,681-689. Brown, A. D., Shorey, C. D., and Turner, H. P. (1965).J. Gen. Microbiol. 41,225-231. Buckmire, F. L. A,, and MacLeod, R. A. (1965).Can.J . Microbiol. 11,677-691. Cho, K. Y., Doy, C. H., and Mercer, E. H. (1967).J . Bacteriol. 94,196-201. Cho, K. Y., and Salton, M. R. J. (1966).Biochim. Biophys. Acta 116,73-79. Christian, J. H. B., and Waltho, J. (1961).J. Gen. Microbiol.25, 97-102. Christian, J. H. B., and Waltho, J. (1962). Biochim. Biophys. Acta 65, 506-508. Colwell, R. R. (1963). Tarometrics 3,l-3. Colwell, R. R., and Gibbons, N. E. (1968). In preparation. Colwell, R. R., and Liston, J. (1961).J. Bacteriol.82, 1-14. Davson, H. (1964). “A Textbook of General Physiology,” 3rd Ed. Little, Brown, Boston, Massachusetts. De, K., Passow, H., Stoeckenius, W., and White, M. (1966).Arch. Ges. Physiol. 289, R15. Drapeau, G. R., and MacLeod, R. A. (1963).J . Bacteriol. 85,1413-1419. Dundas, I. D., Srinivasan, V. R., and Halvorson, H. 0. (1963). Can. J . Microbiol. 9, 6 19-624. Dunham, P. B. (1962). B i d . Bull. 123,462-463. Dussault, H. P. (1955).J.Bacteriol. 70, 484-485. Flannery, W. L. (1956). Bacteriol. Rev. 20, 49-66. Flannery, W. L., and Kennedy, D. M. (1962).Can.J . Microbiol. 8,923-928. Gibbons, N. E. (1968). In “Cultivation of Microorganisms” (J. R. Norris and D. W. Ribbons, eds.) Academic Press, New York Gochnauer, M. B., and Kushner, D. J. (1968). Bacteriol. Proc. 28-29. Hochstein, L. I., and Dalton, B. P. (1968).J . Bacteriol. 95, 37-42. Holmes, P. K., and Halvorson, H. 0. (1965).J . Bacteriol. 90, 316-326. Holmes, P. K., Dundas, I. D., and Halvorson, H. 0. (1965).J.Bacteriol. 90, 1159-1160. Houwinck, A. L. (1956).J . Gen. Microbiol. 15, 146-150. Ingram, M. (1957). Symp. SOC. Gen. Microbiol. 7 , 90-133. Joo, C. N., and Kates, M. (1968). Biochim. Biophys. Acta (in press).

98

D. J. KUSHNER

Joshi, J. C., Guild, W. R., and Handler, P. (1963).J . Mol. B i d . 6, 34-38. Katchalsky, A. (1954). Progr. Biophysics Btophys. Chem. 4, 1-59. Kates, M., Palameta, B., and Yengoyan, L. S. (1965a). Biochemistry 4, 1595-1599. Kates, M., Yengoyan, L., and Sastry, P. S. (1965b). Biochim. Btophys. Acta 98,252-268. khtes, M., Palameta, B., Joo, C. N., Kushner, D. J., and Gibbons, N. E. (1966). Biochemtsty 3, 4092-4099. Kates, M., Palameta, B., Perry, M. B., and Adams, J. (1967). Biochtm. Biophys. Acta 137, 213-216. Kates, M., Wassef. M. K., and Kushner, D. J. (1968). Can. J . Btochern. 46, 971-977. Keller, P., and Henis, Y. (1967). Can.J . Mtcrobtol. 13, 1427-1432. Kates, M., Wassef, M. K., and Kushner, D. J. (1968). Can. J . Biochem. 46, 971-977. Kleinzeller, A,, and Kotyk, A., eds. (1961). “Membrane Transport and Metabolism.” Academic Press, New York. Korngold, R. R.,and Kushner, D. J. (1968).Can./.Mtcrobtol. 14,253-263. Kushner, D. J. (1964a). In “Experimental Chemotherapy” (R. J. Schnitzer and F. Hawking, eds.), Vol. 2, pp. 114-168. Academic Press, New York. Kushner, D. J. (1964b).J . Bacteriol. 87, 1147-1156. Kushner, D. J. (1966a). Btotechnol. Btoeng. 8,237-245. Kushner, D. J. (196613). In “Experimental Chemotherapy” (R. J. Schnitzer and F. Hawking, eds.), Vol. 4, pp. 512-514. Academic Press, New York. Kushner, D. J., and Bayley, S . T. (1963). Can.J . Microbtol. 9,53-63. Kushner, D. J., and Onishi, H. (1966).J . Bacteriol. 91,653-660. Kushner, D. J., Bayley, S. T., Boring, J., Kates, M., and Gibbons, N. E. (1964). Can. J . Microbiol. 10,483-497. Kushner, D. J., Masson, G., and Gibbons, N. E. (1965). Appl. Microbtol. 13, 288. Larsen, H. (1962). In “The Bacteria: a Treatise on Structure and Function” (I. C. Gunsalus and R. Y. Stanier, eds.), Vol. IV, pp. 297-342. Academic Press, New York. Larsen, H. (1967). In “Advances in Microbial Physiology” (A. H. Rose and J. F. Wilkinson, eds.), Vol. I, pp. 97-132. Academic Press, New York. Larsen, H., Omang, S., and Steensland, H. (1967). Arch. Mikrobtol. 59, 197-203. Limsong, S., and Frazier, W. C. (1966). Appl. Mtcrobtol. 14,899-901. McClare, C. W. F. (1967). Nature 216,766-771. MacLeod, R. A. (1965). Bacteriol. Reu. 29,9-23. Mohr, V . ,and Larsen, H. (1963).J . Gen. Mtcrobtol. 31,267-280. Nandy, S. C., and Sen, S. N. (1967). IndianJ. Exptl. Btol. 3, 146-148. Onishi, H. (1963). Adoan. Food Res. 12,53-92. Onishi, H., and Kushner, D. J. (1966).J . Bacteriol. 91,646-652. Onishi, H., McCance, M. E., and Gibbons, N. E. (1965). Can.J.Microbiol. 11,365-373. Ormerod, J. G. (1967).Arch.Mikrobtol. 5 6 , 3 1 4 9 . Quadling, C., and Colwell, R. R. (1964).Deoelop. Ind. Mtcrobfol. 3, 151-161. Rayman, M . K., Cordon, R. C., and MacLeod, R. A. (1967).J . Bacteriol. 93, 1465-1466. Raymond, J. C., and Sistrom, W. R. (1967). Arch. Mtkrobiol. 59,255-268. Salton, M. R. J. (1964). “The Bacterial Cell Wall.” Elsevier, Amsterdam. Scott, W. J. (1957). Aduan. Food Res. 7,83-127. Siegel, S . M., and Roberts, K. (1966). Proc. Natl. Acad. Sci. U.S.36,1505-1508. Soo-Hoo, T. S . , and Brown, A. D. (1967). Biochtm. Biophys. Acta 139,164-166. Stevenson, J. (1966).Biochem.]. 99,257-260.

HALOPHILIC BACTERIA

99

Stoeckenius, W., and Rowen, R. (1967).J.Cell B i d . 34,365-393. Ueno, T. (1964). Bull. Unio. Osaka Prefect. Ser. B . 15,67-113. Vaisey, E. B. (1954).J.Fish. Res. Bd. Canada 11,901-903. Yamada, H., and Okamoto, H. (1961). Zeit. Allg. Mikrobiol. 1,245-250. Yamamoto, M. (1967).Zeit. Allg. Mikrobiol. 7,267-277. Yoshii, H. (1967). Nippon Jozo Kyokai Zasshi 62,43-49; 147-152. (In Japanese; cited in Chem. Abstr. 68, 1006~). Zahl, P. A. (1967). Natl. Geograph. 132,252-263.

This page intentionally left blank This Page Intentionally Left Blank

Applied Significance of Polyvalent Bacteriophages

S . G. BRADLEY Department of Microbiology University of Minnesota. Minneapolis. Minnesota

I . Introduction ............................................................... I1. Fundamental Aspects ...................................................

A . Structure and Composition .................................... B. Life Cycle ............................................................ C . Transfection ......................................................... D . Host-Controlled Modification ................................. E . Variation ............................................................... 111. Infection of Fermentations .......................................... A . Source and Manifestations of Phages for Lactic Acid Bacteria ......................................... B. Control of Bacteriophages for Lactic Acid Bacteria ........................................................ C . Actinophages in Fermentations .............................. IV. Phage Typing ............................................................ A . “Specific” Phages .................................................. B. Polyvalent Phages ................................................ V Conversion .................................................................. VI . Transduction ............................................................... A . Transduction in Staphylococcus .............................. B. Transduction in Bacillus ....................................... VII . Bacteriocins ............................................................... VIII Tools for Applied Research .......................................... A . Mechanism of Drug Action .................................... B. Immunologic Studies ............................................. IX . Conclusions ............................................................... References ..................................................................

.

.

101 102 102 106 108 109 111 113 113 114 115 116 116 117 118 120 120 122 123 124 124 125 132 133

.

1 Introduction

A bacterial virus. or bacteriophage. is an exogenous submicroscopic particle capable of multiplication only inside of specific bacterial cells Viral reproduction. however. is unique in that typical viral particles cannot be found inside of infected host cells for some finite period immediately following infection This eclipse period perhaps constitutes the only useful criterion for differentiating between large viruses and strict intracellular parasites such as rickettsia Viruses were recognized as transmissible. filterable pathogens by Beijerinck about 70 years ago; about 15 years later Twort and d’Herelle described viral diseases of bacteria . Since then. bacterial viruses have

.

.

.

101

102

S. G. BRADLEY

been described for nearly all groups of prokaryotic microbes (Stent, 1963), including blue-green algae (Safferman and Morris, 1963). The best-known bacterial viruses are those for Escherichia coli, and until recently, the T-even coliphages held the attention of the vast majority of the microbial virologists. As a result, the biological and chemical characteristics of the T-even coliphages have been adequately collated (Stent, 1963). Until recently, there has been a tendency to generalize that which is known about the T-even coliphages to all other bacterial viruses. This is clearly not justified in several respects; for example, T-even phages alter the normal metabolism of the host much more dramatically than most bacteriophages. In this essay, the diversity of bacteriophage behavior has been emphasized; accordingly, the examples have been chosen from a variety of microbial genera. This review is not exhaustive; rather, it reflects my prejudices as to what aspects of microbial virology are relevant to the practicing industrial microbiologist today. II. Fundamental Aspects

A. STRUCTUREAND COMPOSITION

Although the release of bacteriophage particleq from bursting bacteria was observed in the dark field microscope by d'Herelle in 1926, phage particles per se cannot be resolved by light microscopy. Study of phage morphology therefore had to await development of the electron microscope. All phages examined have had characteristic shapes; most phages resemble tadpoles, with tails attached to spherical, cylindrical, or polyhedral heads. A tail appendage, though varying in length from phage to phage, is present in nearly all bacteriophages and is the specialized organelle for attachment to the bacterial cell (Bradley and Ritzi, 1967). A limited number of bacterial viruses are spherical (Sinsheimer, 1959) and still fewer are rod-shaped (Kottman, 1942). Bacteriophages, like other viruses, are separable into at least two components: nucleic acid and protein. In some viruses, such as tobacco mosaic virus, the protein component contains only one molecular species of protein. Alternatively, coliphages T2, T4, and T6 contain many molecular species of protein. To a large extent, specific proteins have been related to specific structures, for example, tail fiber protein, tail sheath protein, tail core protein, and head membrane protein (Van Vunakis et al., 1958).In order to study the chemical composition of bacteriophages, the phage particles have to be sepa-

APPLIED SIGNIFICANCE O F POLYVALENT BACTERIOPHAGES

103

rated from the medium components and host debris. Classically, Schlesinger (1933) used filtration through membranes having graded pore sizes and high speed centrifugation. Later, Peterson and Sober (1956) applied cellulose chromatography to the same problem and Albertsson (1960) used two-phase aqueous polymer systems. Bradley and co-workers have used combinations of treatments including dialysis against polyethylene glycol, ion-exchange chromatography, partitioning between the two phases of aqueous polymer systems, gel filtration, precipitating agents, differential centrifugation and sedimentation through sucrose gradients (Bradley et al., 1963; Kolstad and Bradley, 1964; Bradley and Ritzi, 1967). Sequential application of dialysis against polyethylene glycol, centrifugation through a 60% sucrose solution and elution chromatography has routinely yielded milligram quantities of highly purified bacterial virus. Actinophage MSP2 purified in this manner contains, per 1OI2 plaque-forming units, 11.3 pg phosphorus, 35.3 pg nitrogen, and 142 pg deoxyribonucleic acid (DNA); has a dry weight of 246 pg; and gives an absorbancy at 260 m p of 3.1. Similar values have been found with purified actinophage MSPS (Kolstad and Bradley, 1966). The sequence of the four nucleotides in the DNA ultimately determines the sequence of amino acids in the cellular proteins, which in turn, confers upon an organism its specific attributes. A measure of the overall similarity of nucleotide sequences within DNA, therefore, provided evidence for genetic relatedness. Since relative similarity in nucleotide sequence was difficult to determine until recently, earlier work relied upon overall nucleotide composition of the DNA as an indication of relatedness. The DNA of two viruses, however, might possess the same nucleotide composition but have markedly different sequences; therefore, the simpler determination is less reliable. Nevertheless, gross DNA analyses have provided useful information. The serologically related T-even coliphages, for example, all have DNA containing about 37 mole percent guanine plus hydroxy methyl cytosine whereas the DNA from the unrelated coliphage T1 contains 49% mole percent guanine plus cytosine (%GC).The DNA of their host, Escherichia coli, contains 50% GC. Obviously, phage virulence does not require that the phage DNA have a nucleotide composition similar to that of the bacterial host. The temperate coliphage lambda, however, does have essentially the same DNA base composition as its host. Until recently, it was accepted that the DNA composition of temperate phages and transducing phages must be identical to that of the bacterial recipient. Cordes and Epstein (1961) have found, however, that the DNA of the temperate alpha

104

S. G. BRADLEY

phage of Bacillus megaterium contains 42% GC whereas the host DNA contains 37% GC. In addition, the DNA of the transducing phage PBS2 contains 28% GC whereas the DNA of its bacterial recipient, B . subtilis has 43% GC (Takahashi and Marmur, 1963).Nevertheless, present data indicate that the DNA of a temperate phage, in the process of lysogenization, is inserted into the bacterial genophore in a region where the phage and host DNA possess the same nucleotide sequence (Green, 1963). Accordingly, it should be possible to demonstrate at least a small amount of homology between a temperate phage genome and the host genome regardless of the DNA nucleotide composition. Actual nucleotide sequence analysis can be done only with great difficulty; fortunately though, there are several methods available to test for common nucleotide sequences in two DNA preparations of diverse origin. In the agar gel technique (Tewfik and Bradley, 1967), heat-denatured DNA is immobilized in an agar gel. Particles of the DNA-agar are then incubated with radioactive, denatured DNA fragments. Next, the labeled, denatured DNA fragments which have not hybridized to the immobilized DNA in the agar gel are removed. Finally, the labeled DNA fragments which have hybridized are eluted from the DNA agar, collected, and their radioactivity determined. In this procedure, self-renaturation of the DNA does not appreciably complicate the assay; accordingly, relatively small amounts of homology can be detected (Cowie and McCarthy, 1963). Applying these techniques to coliphage DNA, it has been established that T2, T4, and T6 phages possess extensive genetic homology with each other but they show essentially no homology with lambda phage. Lambda phage DNA shows about 50% homology with DNA from phage 434 and 20% homology with $80 DNA (Brenner et al., 1967). Moreover, about 8%of the lambda phage genome is complementary to that of its host, E. coli. In the reciprocal assay, it was found that 0.8% of the E. coli genome is capable of forming hybrids with denatured lambda phage DNA (Cowie, 1967). On the other hand, Tocchini-Valentini et al. (1963) reported that alpha phage DNA does not hybridize with DNA from its lysogenic host B . megaterium. Their data, although not definitive, may indicate that lysogeny may be established by alternative methods, and that the well-knownE. colblambda results may not apply universally. The composition and function of phage structural proteins have been analyzed biochemically and genetically. Fraser (1957) purified and analyzed the amino acid content of coliphages T2 and T3. Subsequently Van Vunakis and co-workers (1958)dissociated the protein


105

subunits of ghosts of bacteriophage T2. They were able to isolate tail fiber protein, tail sheath protein, tail core protein, and head membrane protein, Similarly, Jones (1964) prepared ghosts of an actinophage by treating purified phage particles with alkaline pyrophosphate at 65°C. The whole ghosts, consisting of head and tail pieces joined, were subsequently exposed to mercaptoethanol. The mercaptoethanoltreated material contained intact head membranes but the tail pieces were destroyed. These head membranes were isolated and their amino acid composition determined. The amino acid content of actinophage MSPB is distinctly different from that of its host Streptomyces uenezuelae and that of coliphage T2 (Table I). TABLE I AMINO ACID COMPOSITIONS OF ACTINOPHACE MSP2, ITS COMPONENTS, HOST si3, AND COLIPHAGE Tz‘ Mole % of total amino acids recovered Amino acid

MSPB Intact

MSP2 Ghost

MSP2 Head

Lysine Histidine Arginine Aspartate Threonine Serine Glutamate Proline Glycine Alanine Cysteine Wine Methionine Isoleucine Leucine Tyrosine Phenylalanine

1.1 0.5 5.4 9.4 6.0 4.2 9.2 7.6 12.5 11.8 1.9 8.8 1.9 5.7 7.6 2.5 3.9

1.2 0.7 5.0 9.7 7.7 4.8 9.6 7.6 9.0 12.8 2.1 8.5 1.2 5.3 7.5 2.4 4.0

1.1 0.7 5.6 9.7 8.2 4.8 10.1 8.0 8.4 13.1 0.9 8.5 1.4 5.3 7.6 1.7 4.1

S 13

T2

3.4 1.5 5.3 7.9 5.0 3.8 12.6 3.8 11.1 15.1

5.4 0.7 3.2 10.7 6.1 6.1 10.0 4.3 15.2 10.5

7.2 4.0 3.9 8.5 3.5 3.0

6.1 2.0 6.1 5.4 4.3 4.0

-

-

“Adapted from Jones (1964).

Phage mutants have proved useful for the analysis of structure and function, especially those called conditional lethal mutants. Two types of conditional lethal mutants of coliphage T4 have been studied extensively; temperature-sensitive mutants which form plaques at

106

S . G . BRADLEY

25°C. but not 42°C. and amber mutants which form plaques on E. coli C but not on E. coli B (Epstein et al., 1963). These mutants are especially useful in the study of phage biosynthesis because, under restrictive conditions, mutant infections are abortive. By comparing a given feature of abortive mutant growth with that of the wild type phage under comparable conditions, it is possible to determine the step in development at which a mutation exerts its effect. The mutations appear to fall into two major classes; one class exhibits defects in DNA synthesis whereas the second class exhibits defects in maturation. These two groups of mutations occupy two nonoverlapping regions of the phage genome. One mutation belonging to the first class is called DO; with this mutant, no detectable DNA synthesis occurs. Moreover, neither tail fiber antigen, morphological subunits, nor lysis of the infected cells has been detected. The infected cells lose the ability to form colonies. One series of mutations belonging to the second class affect the phages’ tail fibers. These mutants, under restrictive conditions, produce essentially normal-appearing viral particles that lack tail fibers. In some instances the tail fiber protein is not made; in other instances the tail fiber protein is made but not attached to the tail assemblage. In another series of maturation mutants, normal phage head membranes are not made. In the restrictive host, one of these mutant phages produces long protein cylinders which are composed of head protein subunits. Recently Edgar and Wood (1966) have mixed extracts of bacteria infected with mutant phages under restrictive conditions and have generated intact infective phage particles.

B. LIFECYCLE The typical lytic cycle of a bacterial virus may be divided into a number of discrete stages. First, the phage attaches to specific receptors on the bacterial cell wall. In general, two types of attachment can be observed: reversible and irreversible. Reversible attachment occurs instantaneously, and is relatively independent of temperature and phage concentration. Irreversible attachment is time dependent and temperature dependent. The controlling variables in extent of adsorption are the amount of host mycelium and the ionic content of the medium. Reversible attachment is probably an ion-exchange reaction whereas irreversible attachment may involve an enzymic process. Irreversible adsorption is followed by penetration of the phage genome into the bacterial host cell. New enzyme activities, which are not found in the uninfected bacterium, appear after in-

APPLIED SIGNIFICANCE OF POLYVALENT BACTERIOPHAGES

107

fection. These new activities are due to synthesis of new enzymes under the control of structural genes in the phage itself. In addition, synthesis of many bacterial enzymes and of ribosomal ribonucleic acid is arrested by some, but not all, invading viral genomes. Phageinduced enzyme synthesis proceeds in a characteristic sequence; phage lysozyme, for example, is made only during the latter part of the phage life cycle. Synthesis of new phage-induced enzymes stops, however, at about the time that viral structural proteins first appear. If the infecting phage is treated with ultraviolet light so that no DNA synthesis occurs, the production of phage-initiated enzymes does not stop. Apparently some of the phage's late functions repress genes that control the production of phage-induced enzymes. The composition of the phage structural proteins are determined by the viral genome. The genetic loci controlling several of these structural proteins have been mapped by phage crosses; these include phage head proteins, tail fiber protein and phage lysozyme. In addition, mutations that alter the morphology of the phage tail have been recognized. In summary, the sequence of events during the period of vegetative replication are (a) synthesis of messenger ribonucleic acid (m-RNA) from the phage DNA template; (b) synthesis of new enzymic activities, many of which are concerned mainly with DNA synthesis; and finally (c) synthesis ofthe phage structural components. Following vegetative replication, there is the maturation period during which new proteins must be made which initiate the folding of the DNA into a compact or condensed structure. The condensed DNA is then enclosed within a head membrane. Progressively, the tail assemblage, base plate, and tail fibers are added (Edgar and Wood, 1966; Weigle, 1966). The final step in the lytic life cycle is lysis of the bacteria cell and release of the mature phage particles. With temperate bacteriophages that can establish lysogeny, one of two alternative sequences of events follows penetration of the viral genome into the recipient bacterial cell: either there occurs an irreversible reaction that leads to the synthesis of new viral particles and lysis of the bacterium, or vegetative phage multiplication is arrested and the phage genome is reduced to a normal cell component. The latter process is known as lysogenization. The critical step in determining whether a temperate phage will lysogenize the host or initiate the lytic life cycle is production of a repressor which presumably blocks a function essential for vegetative replication (Luria, 1962). Later on, there will be a breakdown in immunity of the lysogenic bacterium, and the temperate prophage will initiate a lytic life

108

S. C. BRADLEY

cycle. Not only do temperate phages not destroy the genome of their host bacteria, but they actually share some degree of genetic homology with it. The lysogenic bacteria are immune to superinfection with homologous phages in the sense that the superinfecting phage, as well as the prophage, is prevented from initiating vegetative replication and maturation. As a rule, the superinfecting phage genome does not multiply and is randomly passed on to one daughter cell at cell division as an extraneous plasmid. Some phages initiate chronic viral infections which cannot be clearly defined as lytic or lysogenizing. These phages attach to the phage-receptor sites and inject their nucleic acid into the host cell in the same manner as virulent phages. The essential metabolic processes for bacterial growth, however, are not markedly inhibited. Consequently the host and vegetative phage multiply simultaneously but not necessarily at the same rate. Accordingly, an occasional bacterial progeny cell will not contain a vegetative phage particle; in other bacterial cells, virus replication may exceed the host’s capability to maintain itself, thereby leading to death of the host and liberation of the accumulated mature virus particles (Hohann-Berling and Maze, 1964). The cells which maintain balanced growth of the phage genome and its own essential constituents have been referred to as unstably lysogenic by some workers. Such phages can serve as vectors in genetic transduction (Takahashi, 1963).

C . TRANSFECTION Transfection is the infection of bacterial cells by isolated nucleic acid from a bacteriophage, resulting in the production of a complete virus. Transfection can be considered as analogous to bacterial transformation. In both genetic transformation and transfection, the biologically active DNA is destroyed by pancreatic deoxyribonyclease but not by phosphodiesterase from Escherichia cob, and DNA inactivated by ultraviolet irradiation can be repaired by yeast. photoreactivating enzyme. Moreover, susceptibility to infection and competence for bacterial transformation decrease concomitantly. Also, transfection and transformation show an initial linear dependence on concentration of added DNA, and saturation at high DNA concentrations. Finally, the addition of bacterial DNA reduces the incidence of transfection, indicating that the two types of DNA are competing with each other (Harm and Rupert, 1963). The most complete studies on transfection have been carried out with the bacteriophages of Bacillus subtilis (Foldes and Trautner,


109

1964; Reilly and Spizizen, 1965). In general, the viral DNA is extracted with phenol; infective DNA has a molecular weight of about lo8 daltons. Transfection is not inhibited by ribonuclease, trypsin, or antiserum directed toward the intact phage particle. There is no apparent lag in DNA adsorption by competent €3. subtilis cells. Only about 1% of the viable cells are ever in the competent state that will support transfection. About 15 minutes after DNA uptake are required before the first intact phage particles are formed. There is a higher frequency of recombinant phages produced by cells infected with DNA than by cells infected with intact virus. This observation indicates that infected cells are formed as a result of an interaction among fragmented viral DNA molecules. Green (1964) has calculated from the concentration dependence of infectivity that four viral DNA molecules must interact in order to form an infectious center. D. HOST-CONTROLLED MODIFICATION Host-controlled modification was originally defined as the process by which the phenotype of a bacteriophage is altered during growth in a particular host (Bertani and Weigle, 1953). The resulting change has usually been an apparent increased specificity for that propagating bacterial strain (Table 11). The host-controlled specificity described TABLE I1 HOST-CONTROLLED MODIFICATION OF ACTINOPHAGEWSP-8' Relative plaque-forming ability on alternative hosts Primary host

104

11

86

S. griseus 104

1 10-2 10-2

10-2 1 10-4

10-4 10-3 1

S. olioaceus 11 S. venezuelae 86

"Adapted from Bradley (1959).

by Luria and Human (1952)was readily reversible by propagating the modified phage on an alternative host; therefore, they referred to this process as nonhereditary host-induced modification of T-even coliphages. This phenomenon is clearly different from classical host-range mutations, with superimposed selection wherein new, stable geno'types arise and become the predominant type in a phage population. Host-controlled modification is also clearly differentiated from pheno-

110

S . G. BRADLEY

typic mixing, which gives rise to a small proportion of phage particles having the genotype of one phage strain and the attachment properties of a second phage strain. These phage mosaics, formed in bacterial cells infected simultaneously with two different phage strains, are presumably formed by packaging the DNA of one strain into the protein coat of the other. Such a phage particle has a transitory host specificity different from that prescribed by its genetic material. In phenotypic mixing, the attachment process is altered; in host-controlled modification, the rate and extent of phage attachment is not drastically reduced. Accordingly, the genetic material of the phage itself is the probable site affected. It has been repeatedly demonstrated that in host-controlled modification, the phage DNA is so altered that the bacterial cells are no longer able to recognize the viral DNA as foreign. The host that rarely allows for phage replication, called a restrictive host, degrades the phage DNA into small genetically inactive pieces whereas the permissive host supports the normal viral life cycle (Wood, 1966). In the lambda phage system, the modified phage particles contain DNA newly synthesized in the modifying host (Arber and Dussoix, 1962). The small number of lambda phage particles that are able to grow on the restricting host possess at least one DNA strand from the original phage particle infecting the modifying host (Arber, 1965a). This definitively establishes that the phage DNA is modified during the process of DNA replication in the modifying host. Clearly, host modification is not nonhereditary. The modifying factor may be a structural gene of the host itself (and accordingly, itself subject to mutation) or may be a prophage. In the case of phage P1, a strain of E. coli K12 infected with a high multiplicity of lambda phage, then superinfected with P1, all the progeny (even those containing DNA from the original unmodified phage particles) are modified with respect to host specificity (Lederberg, 1957). It has been proposed that the host specificity of lambda phage is due to methylation of the DNA bases (Arber, 196513).Host-controlled modification produces essentially the same effects on lambda-dg transducing phage as it does on normal lambda phage. Transduction to a restricting host yields a low proportion of gal+ transformed cells which are mostly heterogenotes, that is, the input transduced genetic fragment persists but is unable to integrate into the genophore of the bacterial recipient. Superinfection with nonrestricted lambda phage greatly increases the transduction frequency when lambda-dg is used with a restricting recipient cell type. Superinfection with restricted lambda phage does not appreciably increase the rate of transduction by


111

restricted lambda-dg unless the multiplicity of superinfection is high. The T-even coliphages are also subject to host-controlled modifications. T-even phages efficiently infect and replicate in wild type Escherichia coli and Shigella. The DNA of these T-even coliphages differs from that of the bacterial host DNA in that 5-hydroxymethyl cytosine replaces cytosine and the phage DNA is glucosylated. The glucosylation reaction requires uridine diphosphoglucose (UDPG), which is supplied by the host. In fact, the glucosylation of hydroxymethylcytosine occurs after DNA synthesis is complete. Mutants of E. coli that lack the ability to synthesize UDPG can still produce Teven phages but their DNA is incompletely glucosylated. Such modified phages do not generally propagate in E. coli B or its mutants but can grow in Shigella strains (Hattman and Fukasawa, 1963).The phage produced in Shigella are normal in that their DNA is glucosylated and the phage can replicate equally well in E. coli and Shigella strains. The rare instances in which modified T-even phage do grow in E. coli strains is the result of variation in the bacterial population rather than variation in the phage population. The exceptional bacteria may arise through physiologic variation in ability to restrict the growth of the modified phage or by mutation. It is significant that several episomes, including F-lac, F-gal, and probably resistance-transfer factor, are subject to host modification. In fact, the species and strain specificity of bacterial DNA may be partially attributable to glucosylation or methylation of preformed DNA. Restriction, as seen in host-controlled modification, may constitute another barrier to genetic exchange between diverse strains and species which possess substantial genetic homology.

E. VARIATION A given strain of a bacteriophage retains indefinitely its essential properties whether stored or maintained by serial propagation. Unexpected changes in characteristics or behavior of a phage preparation, however, are frequently encountered. Failure of the phage to form plaques on a sensitive indicator strain may be the result of any of a large number of rather trivial factors; these include a change of a few degrees in the incubation temperature, a change in medium composition, or a change in the amount of host added. Once these physiologic variables have been considered, variability in the phage preparation itself must be considered. In many families of bacteria, over half of the laboratory cultures are lysogenic. If the phage strain is propagated blindly for a number of subcultures, a virulent mutant

112

S. G. BRADLEY

of a temperate phage may become established as the principal member of the mixed viral population. The resulting lysate will give aberrant results in host range determinations; this problem plagued the early workers trying to phage-type staphylococci. A major source of variation of a given phage per se is host-controlled modification. Host-controlled modification reflects an alteration of the viral genetic material by host enzymes. These alterations usually involve addition of glucosyl or methyl groups to the phage DNA. The chemically altered DNA is inherited semiclonally thereafter if the phage is subsequently propagated on a host that does not modify the viral DNA, that is, the parental DNA remains altered but the newly synthesized DNA does not carry the characteristic chemical substitution. Host-controlled modification is distinctly different from phenotypic mixing which may occur when a bacterial cell is simultaneously infected with two different but related phages. In phenotypic mixing, some of the phage particles possess the attachment properties of one phage type and the genetic determinants of another phage type. As an example, an E. coli cell infected simultaneously with T2 and T4 will liberate normal coliphage T2, which can propagate in both E. coli B and E . coli B/4. Moreover, normal coliphage T4 will be liberated; this phage can propagate in E. coli B but not E. coli B/4. The predominant chimera phage type from a mixed infection with coliphages T2 and T4 can infect E. coli B but not E. coli B/4. The progeny of the chimera phage grown on E. coli B can infect both E. coli B and E . coli Bl4. From the recent work on phage maturation (Edgar and Wood, 1966), it seems probable that the phage chimera results from the attachment of a T4 tail assemblage to a T2 head. Bacterial viruses are subject to both spontaneous and induced mutation. In fact, bacteriophages have proven to be convenient biological material for the study of chemical mutagenesis. Mutations can be induced in intact, resting phage particles by alkylating agents, radiation, and nitrous acid. In order to induce mutations in bacteriophages with analogs of nucleic acid precursors, the chemical must be presented to the phage-infected cell. A unique chemical mutagen is acriflavine which, in the dark, is mutagenic for T-even phages but not mutagenic for most other bacterial viruses. Acriflavine, in the dark, induces aberration in the exchange of DNA segments during genetic recombination. Recombination occurs extensively among T-even bacteriophages. In order to detect recombination, an E. coli cell must be simultaneously infected with two recognizably different virulent


113

phage strains. Recombination may also occur between a temperate phage and its virulent mutant. Many bacteria carry unrecognized or defective temperate phages; therefore, it is not always easy to determine whether a new viral variant arose by mutation or recombination. Ill. Infection of Fermentations

A. SOURCE AND MANIFESTATIONSO F PHAGES FOR LACTICACID BACTERIA Lactic acid bacteria are employed extensively in the dairy industry, primarily for the manufacture of cheese and fermented milk products. Moreover, they have a determinative role in developing the flavor and texture of a large number of food products, including cucumber pickles, rye bread, and sauerkraut. In addition, the conversion of chopped plants, for example, corn stalks and alfalfa, to ensilage is primarily a lactic acid fermentation. Whitehead and Cox (1935) isolated phage from a culture of Streptococcus cremoris and were able to establish that it was a cause of slow acid production in milk inoculated with this started culture. Subsequently, there have been numerous reports of bacteriophages as the cause of unsatisfactory performance by starter cultures. Indeed, bacteriophages are probably the main cause of slow acid production in milk inoculated with lactic acid bacteria. Moseley and Winslow (1959) detected bacteriophages in more than 90% of samples from cheese factories that were having problems; surprisingly, they also found phages in about 75% of the samples from factories not experiencing difficulties. Lysis of the lactic acid bacteria during cheese manufacture was not marked when raw milk was used, probably because untreated milk contained many different bacterial strains of varied susceptibility to bacteriophages. As the use of pasteurized milk for cheese making has increased, so has the phage problem (Babel, 1962). Most starter cultures, however, contain more than one strain, and the individual components usually possess different phage sensitivities. Lysis of one component of a mixed starter culture does not grossly affect the coagulation of the milk. The bacteriophages that infect lactic acid fermentations seem to come from diverse sources. The air in cheese factories often contains phage. Another source of phage is the starter culture. Some starter cultures are carrier systems in which sensitive bacteria, resistant bacteria, and the viral population exist in an

114

S.

G. BRADLEY

equilibrium. In addition, a number of lactic acid bacteria are lysogenic. In fact, the incidence of lysogeny among some groups of streptococci is as high as 85%. The phage population may rise explosively when a lysogenic culture is mixed with a sensitive culture in preparing a starter culture composed of many strains or when a virulent mutant of a temperate phage develops. B. CONTROL OF BACTERIOPHAGES FOR LACTICACID BACTERIA Bacteriophages for lactic acid bacteria are generally inactivated at

75°C.within 15 minutes. Even though they are more resistant to heat than their hosts, heat is a satisfactory method for freeing milk of bacteriophages. Hypochlorite compounds are frequently used to cleanse and disinfect milk-processing equipment. Moreover, hypochlorites are potent viricides; 50 pg. sodium hypochlorite/ml. will inactivate bacteriophage in a matter of seconds. Hypochlorite compounds, however, are quite corrosive; therefore, their use is limited. Quaternary ammonium compounds, frequently used to cleanse and disinfect milk-processing equipment, are less corrosive than the hypochlorite compounds, but are also less effective in inactivating bacteriophages. Nevertheless, 100 p g detergent/ml. destroys phage in a matter of minutes. The viricidal action of quaternary ammonium compounds for S . cremoris phage is increased if ethylenediaminetetraacetate is added to the sanitizing solution. Bacteriophages for streptococci are inactivated by p-hydroxymercuribenzoate and mercuric ion; reduced glutathione restores plaque-forming capability. As noted previously, the starter cultures themselves often contain phage. However, in order to demonstrate phage in the inoculum, a sensitive indicator strain is required. Frequently, several hundred isolates must be tested before a suitable indicator strain is discovered. If the starter culture is a mixture of several bacterial strains of varied phage sensitivities and lytically propagating phage, the inoculum can be freed of phage. Each significant member of a starter culture should be purified by serial subculture from single colonies. Antiserum directed against the phage may be prepared and added to the culture medium to neutralize the phage. Once the inoculum is freed of phage, the cultures must be maintained in an environment free of airborne phage. This means that the culture room and equipment should be kept meticulously clean and separate from the processing plant. Air filters and regulation of movement of personnel from the factory to the culture room will reduce phage contamination. If


115

the starter cultures contain a lysogenic strain, elimination of that organism from the mixed inoculum is desirable. Alternatively, mutants resistant to the temperate phage must be selected from all strains that may come in contact with the lysogenic organism. That resistance to a given phage simultaneously confers resistance to closely related, but not all, phages merits emphasis. Rarely, a lysogenic culture can be freed of its temperate phage, but currently this process is tedious and only irregularly successful. A management practice that oftentimes successfully decreases the incidence of phage outbreaks is rotation of starter cultures. If the different inocula have unique susceptibilities to phage, a population of phage will not build up in the air and on the equipment. It must be emphasized that a novel phage which is virulent for one or more of the rotated starter cultures may enter the plant. Periodic checking of the individual members of the inocula for sensitivity to phages in the production vats is a wise precaution. Because bacteriophages for lactic acid bacteria require more calcium ion for growth than their hosts, starter cultures free of contaminating phage can be prepared, even though the air is heavily laden with phage, in media low in calcium content. Alternatively, it is possible to suppress phage growth by chelating the divalent cations with pyrophosphate or oxalate, but cost and standards for food additives limit this approach.

C. ACTINOPHACES IN FERMENTATIONS Actinophages have significance for the industrial microbiologist because they can destroy an antibiotic-producing culture during the fermentation process, causing a substantial loss in yield. The destructive action of the actinophages is not confined to the sporadic, dramatic “clearing” or “thinning” of a fermentation culture, but is also manifested as a chronic infection in which a partially phage-resistant population supports a substantial viral population in both laboratory and production cultures. Experience has established that diverse phage can attack a given strain; accordingly, selection of genetically resistant variants constitutes only one approach to the development of standard operating procedures that prevent phage outbreaks. Adequate medium, equipment, and air sterilization, coupled with general plant hygiene, are the first line of defense against both bacterial and viral contamination. Moreover, media have been devised which support host growth but not phage replication or are toxic for the phage but not the host (Bradley and Ritzi, 1967).

116

S. G . BRADLEY

IV. Phage Typing

A. “SPECIFIC”PHAGES Attempts to identify staphylococci by means of serologic techniques have not provided the definitive classification necessary for epidemiologic studies. Bacteriophage typing, however, has proven to be most useful when reliable phage-typing preparations are employed. The Central Public Health Laboratory at Colindale, London, is historically the source of reference phage-typing preparations. These phages may be propagated in the laboratory by a number of methods. Although propagation in broth is simple, this method does not give as high titered lysates as soft-agar methods (Blair and Williams, 1961). An efficient means for phage propagation by virtue of simplicity and quality of the resulting lysate employs a cellophane disk rather than an agar overlay (Liu, 1958).The titer of a particular phage lysate is a function of the phage being propagated and the technique used. Many difficulties beset the use of undiluted phage lysates, therefore, the preparation is diluted to give a “routine test dilution.” This step is necessary because many coagulase-positive cocci adsorb all of the typing phages irrespective of their lytic activity. This adsorption may be lethal to the bacterium even though phage replication does not occur. Moreover, the majority of the propagating strains are lysogenic and high-titered phage lysates prepared with such hosts may contain one or more temperate phages in addition to the intended phage. In addition, high-titered lysates may contain host-range mutants which produce plaques on a strain which the parental phage does not infect. Also, concentrated phage lysates may contain substances which induce lysogenic bacteria to liberate phage, resulting in plaque formation. Accordingly, phage lysates are diluted so that they contain from lo3to lo5 plaque-forming units. The real problem associated with phage propagation is not the preparation of high-titered lysates, but of maintaining phage with the proper host range specificities. The lytic pattern of each typing phage preparation should be tested against a set of reference strains. No typing-phage preparation should be used unless it conforms to the accepted pattern. Finally, new phage stocks should be obtained periodically from one of the International Reference Laboratories. Phage lysates can be freed of viable bacteria by filtration or by addition of bactericidal chemicals. Thymol has been used extensively, but residual bactericidal effects may result in false “positive” tests. Fortunately staphylococcal phages are relatively stable and undiluted


117

stock phage preparations may be stored in the refrigerator for a year without significant loss in titer. For long-term storage, the staphylococcal phage preparations may be lyophilized. Phage typing is a relatively simple procedure in which an agar plate is inoculated with sufficient bacterial culture to give a confluent growth; next, a drop of a “routine test dilution” of each phage is placed at predetermined positions on the agar surface. After incubation, the degree of lysis produced by each phage is recorded and the phage type determined from published patterns. It should be obvious that cultures from clinical specimens may contain mixtures of staphylococcal strains; unless these strains are separated and tested individually, a nontypable reaction may be observed. The practice of transferring confluent growth from what appears to be a pure primary or secondary culture of staphylococci on an agar plate, rather than from an isolated colony, is imprudent. The indicator lawn of bacteria may be obtained by flooding the plate with 1 ml. of a broth culture or a smaller inoculum may be spread with a sterile bent glass rod. The excess fluid in the inoculum should be removed and the agar surface should appear dry within an hour. The typing phage may be applied with a loop, a pipet, or a syringe and needle. Multiple inoculating devices that use loops, capillary tubes, or metal cylinders have been made to facilitate this step (Zierdt et al., 1960). About 20% of staphylococcal cultures examined do not show any lysis with the “routine test dilutions” (Wentworth, 1963). A number of workers retest such cultures using typing phage 1000-fold more concentrated than the “routine test dilution.” Moreover, many untypable cultures can be typed if grown at 45°C. rather than 37°C. Apparently a certain proportion of the nontypable cultures are in some sort of “refractory” state rather than inherently resistant to the typing phage.

B. POLYVALENT PHAGES The relationships of bacterial strains to one another can be deduced from host range patterns using polyvalent phages. In one method of analysis, results of phage tests may be considered as all-or-none phenomena and the arrangements based upon overlapping viral susceptibilities. Additionally, quantative as well as qualitative aspects of phage susceptibility may be used in developing a phylogenetic sequence. For example, if a particular phage lysate assayed on bacterial strains A, B, and C gave titers of lo@,lo7,and lo5 plaqueforming units/ml., respectively, then A is considered more closely allied to B than to C. Because viral susceptibility can be easily de-

118

S. G . BRADLEY

termined both qualitatively and quantitatively on a large scale, the data can be best analyzed with the aid of a computer. Affinities discovered by this method reflect phylogeny and evolution as accurately as any other system in use today (Jones and Bradley, 1964). Phage typing, especially with polyvalent phages, must be done with care and understanding. The relative power of a bacteriophage to produce plaques on a particular putative host is dependent upon the propagating strain used in preparing the stock lysate. Such host modification of virulence may represent nonhereditary alterations of the virus or selection for phage mutants. If a given phage is propagated on different host strains, the resulting lysates must be considered as unique typing reagents until adequate study establishes their identity. Numerous technical faults may make scoring of phagetyping assays difficult. Occasionally air bubbles leave clear, discrete spots that resemble plaques in the confluent host growth. Some bacterial contaminants produce thin, transparent colonies, which may or may not be surrounded by a zone of inhibition, and which simulate viral plaques. Excessive moisture on an assay plate can create false plaques by water-spotting the host lawn or may cause smearing and streaking of valid plaques. A dense inoculative suspension of the test bacterium may obscure plaque formation in the confluent bacterial growth. Bacteriophage susceptibility has not been maximally utilized as a taxonomic tool. Additional techniques could be used for detecting interactions among phage and bacteria. Phage reproduction is dependent on a number of sequential events: attachment, infection, replication, maturation, and lysis. Interference with any one of these steps will prevent plaque formation. Attachment alone can be used to determine whether specific receptor sites are present. In addition, normal attachment and infection reactions can be circumvented by infecting cells directly with phage nucleic acid and assaying for production of mature phage particles (Jones and Bradley, 1962).

V. Conversion The genus Salmonella consists of a large group of gram-negative, pathogenic organisms related to Escherichia coli and other enteric bacteria. The salmonellae are divided into species or serotypes on the basis of differences in the antigenic structure of the flagellar (H) antigens and the cell wall polysaccharides (somatic or 0 antigens). Antigenic variations in salmonellae are being recognized so fast that


119

over 1200 serotypes have been described, which is considerably higher than the 700 serotypes reported a few years ago (Robbins and Uchida, 1962). It has been shown that bacterial viruses play a prominent role in these changes. The following discussion will deal with those genetic changes related to bacteriophages that influence the distribution and composition of somatic or 0 antigens in salmonellae. There are two general properties of lysogenic bacteria which set them apart from their nonlysogenic counterparts. Every lysogenic cell carries phage genes as a normal cellular component; the presence of the integrated temperate phage genome renders the cell immune to attack by exogenous homologous phage. The immunity is of an intracellular type because penetration and infection are not affected. The second characteristic of a lysogenic bacterium is the ability to release mature phage particles sporadically. When a culture is “induced” with peroxides, ultraviolet radiation, or mitomycin C, many cells are lysed and mature phage particles are released. Lysogenic bacteria may acquire additional characteristics as a result of the introduction of the phage genome. This process is known as lysogenic conversion. For example, Iseki and Sakai (1953) observed that culture autolysates of Salmonella having group EZ somatic antigens were able to change the serotype of organisms having E (3,lO) somatic antigens to group E2(3,15).Subsequently, Uetake et al. (1958) demonstrated that Salmonella anatum having antigens 3,lO developed antigens 3,15 after lysogenization with the bacteriophage epsilon-15. The new antigen appears on the surface of S. anatum within 15 minutes after infection of the phage-sensitive cells with epsilon-15 phage whether the infecting virus enters the lytic cycle or the lysogenic cycle. S. anatum lysogenized with epsilon-15 phage sometimes segregates to give nonlysogenic progeny. The lysogenic progeny retains the new antigenic structure (3,15) whereas the nonlysogenic derivatives, after temporarily carrying the 3,15 antigens, return to the 3,lO serotype. The new serotype (3,15) also becomes sensitive to another bacteriophage, epsilon-34, whereas the original S . anatum (3,lO) is resistant to this phage. When the converted S.anatum cells (3,15) are lysogenized with epsilon-15 phage, the antigenic pattern (3,15,34) is developed. Uetake and Hagiwara (1961) isolated a culture, lysogenic for epsilon-34 phage and having the serotype E1(3,10).These cells did not produce antigen 34. When this strain was lysogenized with epsilon-15 phage, the cellular antigenic structure converted from 3,lO to 3,15,34. These results indicate that antigen 15 is a precursor of antigen 34.

120

S. G. BRADLEY

Numerous other examples of phage conversion have been described. Nonlysogenic strains of Escherichia coli K12, for example, support the growth of the rII mutants of the T-even coliphages, but K12 cultures 1ysogenized with lambda-phage do not. This alteration in host range is not directly due to the specific lysogenic immunity directed toward phages related to lambda. Lambda phage and T-even coliphages are unrelated antigenically and also differ markedly in morphology and chemical composition. Probably the most dramatic consequence of lysogenic conversion is the production of diphtheria toxin by Corynebacterium diphtheriae (Barksdale, 1959). Cultures of virulent C . diphtheriae invariably contain free phage, indicating that some fraction of the bacterial population is lysogenic. Lysogenization of a nontoxigenic strain with a phage from a toxigenic culture may confer toxigenicity upon the recipient bacterium. Toxin production occurs as a result of induction of the proper temperate phage, or as a result of infection with an appropriate lytic phage (Bradley, 1966).

VI. Transduction

A. TRANSDUCTION IN Staphylococcus The transfer of a limited amount of genetic information from a disrupted cell to an intact cell by a bacteriophage vector was first demonstrated by Zinder and Lederberg (1952) with mutant strains of Salmonella typhimurium. The majority of the subsequent studies on transduction have been done with Salmonella and Escherichia coli; however, considerable significant work has been carried out with

Staphvlococcus aureus. Morse (1959) used typing phage 53 to transduce the genes for resistance to streptomycin and novobiocin. He found that there was a delay in the phenotypic expression of streptomycin resistance, which took 4 to 6 hours to become expressed. Approximately half of the transduced genes for novobiocin resistance were expressed. immediately after the transduced genes had entered the recipient bacterium. Only one phage particle in l o 7 was able to bring about transduction. Almost all of the transformed cells were lysogenic, but most cells were infected by many different phage particles so it was not clear whether the transducing phage and the lysogenizing phage were one and the same. Korman and Berman (1962) used phage 53 to transduce genes controlling antibiotic resistance and sugar fermentation. The transduction frequency was enhanced by treating the phage


121

lysates with ultraviolet radiation. Korman (1966) has isolated and described a remarkable mutant of S. aureus that fails to elaborate coagulase, is resistant to a number of bacteriophages, is resistant to cycloserine and has altered ability to metabolize a number of carbohydrates, including mannitol. In transduction, the genetic lesion responsible for these manifold alterations is inherited as a mutation affecting a single locus. Biochemically, such mutants have increased amounts of serine and reduced amounts of glycine in their cell walls. Ritz and Baldwin (1961) were able to transduce ability to produce penicillinase using staphylococcus phage 80. Using phage 80, the transformed cells were not lysogenized, indicating that the transducing particle and the lysogenizing particle were different. Either sensitive or lysogenized staphylococcal strains could be used as recipients for transduction. None of the genes controlling penicillinase production, chlorotetracycline resistance, novobiocin resistance, streptomycin resistance, lactose fermentation, or mannose fermentation were linked in transduction analyses. Resistance to the macrolide antibiotics oleandomycin, erythromycin, magnamycin, or spiromycin was transferred as a block, indicating that resistance to these antimicrobial agents was controlled by a single locus (Patee and Baldwin, 1962). Mitsuhashi and co-workers (1965) not only confirmed that penicillinase production and resistance to macrolide antibiotics is transduced but showed that they could be jointly transduced. Not only is this pair of characters rarely separated in transduction experiments but they can be jointly eliminated by treatment with acriflavine. Mitsuhashi et al. (1965)proposed that this pair of genetic determinants are located on a single genetic element that is not integrated into the normal bacterial genophore. The observations of Barber (1949) are consistent with this view; he noted that naturally occurring penicillinase negative staphylococci have never been observed to mutate to penicillinase production but that penicillinase-producing staphylococci frequently give rise to penicillinase-negative variants which are stable. Novick and Morse (1967) have provided additional data suggesting that penicillinase production is controlled by an autonomous plasmid. This plasmid is not transferred upon cell contact as is the resistance-transfer factor in enterobacters but can be transduced. In fact, plasmids seem to be transduced more readily than genes in the normal staphylococcal genome. Many staphylococci contain temperate phage capable of mediating generalized transduction; it is not surprising, therefore, that spontaneous transduction

122

S. G. BRADLEY

resulting in plasmid transfer can occur in mixed cultures. Moreover, Jarolmen et al. (1965) have shown that the penicillinase plasmid can be transduced in uiuo.

B. TRANSDUCTION IN Bacillus When a culture of a potentially lysogenic bacterium is infected with an appropriate temperate phage, a certain fraction of the bacterial population lyses and a certain fraction of the population becomes lysogenic. The relative numbers of bacteria following each course are determined by genetic compositions of the phage and of the host, and the experimental conditions. Temperate phages of Bacillus follow this general pattern, but have an additional relationship with their host which can be designated as unstable lysogeny. As an example of this latter host-parasite interaction, some observations made by Thorne (1962) can be summarized. Bacillus subtilis W23 was infected with phage SP10. Colonies were picked from the center of turbid plaques and were subcultured in broth. The resulting population was transferred to potato extract medium to induce sporulation. The resulting spores were heated to kill free phage and vegetative bacteria. These spores were inoculated into nutrient broth and incubated until the stationary phase of growth was achieved. This culture filtrate contained lo8 phage SPlOlml. This titer is unusually high for a lysogenic relationship though both the original and final cultures displayed two phenotypes characteristic of lysogenic cultures; they were resistance to phage SPlO and continuously produced phage SP10. Takahashi (1963) described a similar situation with phage PBSl growing on B . subtilis. He observed that even though cultures were stably lysogenic, individual bacteria were not. Clones from lysogenic spores grown in the presence of antiphage serum yielded 70% sensitive cells and 30% that were lysogenic for phage PBS1. If antiserum was not present, only lysogenic and resistant bacteria were found; the phagesensitive organisms became infected by the free phage and were either lysed or lysogenized. These observations indicate that the phage-infected bacterium continues to grow and divide; concurrently the intracellular viral genome which is not integrated with the bacterial genophore replicates independently and is variably distributed to the daughter cells. Unless reinfection is prevented, the sensitive cells are systematically eliminated. Okubo et al. (1963) have extracted DNA from lysogenic B . subtilis cells and obtained DNA’s having two distinctly different buoyant densities in CsCl equilibrium gradient density centrifugation. This confirms that the phage DNA is


123

not covalently linked to the bacterial genophore. In some temperate Bacillus phages, the prophage may be associated with the bacterial DNA. Phage S1, for example, transduces some loci with a significantly greater frequency than other loci whereas the recombination frequency of the loci via transformation is the same (Takagi and Ikeda, 1962). Moreover, several Bacillus strains possess bacteriocinlike defective phages which replicate in synchrony with the bacterial genome (Seaman et al., 1964). These defective phages can be induced by treating the lysogenic bacillus strain with ultraviolet light or mitomycin C. Because lysogeny in Bacillus is generally an unstable state, transducing phage may be harvested from the supernatant fluid of a Bacillus culture grown from a heated spore inoculum. Alternatively, a donor culture may be lytically infected and the transducing phages harvested after the culture lyses. As is generally true, transducing ability is more stable to ultraviolet irradiation than is plaque-forming ability. In several instances transducing efficiency is actually enhanced by low doses of ultraviolet light. Takahashi (1965) has observed that certain asporogenous mutants which cannot be transformed by free DNA can undergo transduction. Transduction seems to be the only available method for genetic studies on the control of the sporulation process in Bacillus.

VI I. Bacteriocins Bacteriocins are a class of antibiotics which act only on strains of the same or closely related species. The bacteriocins are composed of protein; some of the colicins are components of the “0”somatic antigen of the producing strain. Bacteriocins kill susceptible bacteria after first adsorbing onto specific receptors on the bacterial surface. Bacteriocins and bacteriophages both appear to adsofb onto similar receptors, and there are a few known examples of cross-resistance between colicins and bacteriophages. Bacteriocins are produced by most groups of bacteria: the enteric bacteria, pseudomonads, bacilli, lactic acid bacteria, and the staphylococci. Bacteriocinogenic strains, although they possess the stable genetic potentiality to produce a bacteriocin, do not do so all of the time. Several bacteriocinogenic strains can be induced to produce bacteriocin by treatment with ultraviolet light or peroxides (Fredericq, 1963). The cells which actually produce colicin die in the process. Accordingly, there is a striking parallel between the behavior of bacteriocinogenic bacteria and lysogenic bacteria. For example, some bacteriocinogenic strains,

124

s. G .

BRADLEY

when grown together with appropriate nonbacteriocinogenic strains, can transfer to them the ability to produce bacteriocin. The transfer requires cell contact; the bacteriocin made by the recipient bacterium is of the same type as that of the donor strain. The acquired ability to produce a bacteriocin is regularly transmitted to the progeny. The new bacteriocinogenic strain resembles the nonbacteriocinogenic parent in all other respects (Reeves, 1965). The ability to produce bacteriocin, however, can be lost. These observations are consistent with the proposal that bacteriocinogenicity is controlled by a plasmid. These genetic determinants which are not associated with the bacterial genophore have been called C-factors. The “C” is taken from colicin, the best-known family of bacteriocins. C-factors, in addition to determining the ability to produce bacteriocins, also confer immunity to that bacteriocin and to bacteriocins of the same type, but the immunity may be incomplete. C-factors have been estimated to contain about lo5 nucleotide pairs. Certain C-factors can serve as a fertility factor for syncytic recombination. Moreover, acridine orange eliminates certain C-factors just as it eliminates F-factors (Nomura, 1967). Cfactors, therefore, resemble F-factor in a number of features (Watanabe, 1963); conceivably, C-factors have evolved from F-factor. Consistent with this hypothesis is the observation of Clowes (1963) that some Hfr strains can kill F-strains during syncytic recombination. VIII. Tools for Applied Research

A. MECHANISM OF DRUGACTION Most bacterial viruses contain DNA and structural protein but a number of bacteriophages contain RNA instead of DNA. The RNA viruses can replicate without the intervention of DNA polymerase, thymidine synthetase, or DNA-dependent RNA polymerase. Accordingly, antibiotics which inhibit DNA-containing viruses but not RNAcontaining viruses may affect one of these enzymes (Zinder, 1965). The mechanism of action of antimetabolites that prevent viral replication can frequently be partially elucidated by reversal experiments in which normal metabolites are tested for the capability to restore phage production. Most mechanism of action studies cannot be done so simply. Biochemical methods, especially cellular fractionation of isotopically labeled materials, electron microscopy and immunoassays may be needed to identify the vulnerable physiological locus.


125

Fortunately though, the biosynthetic sequences are well known for many bacteriophages and these events can be initiated synchronously at will in essentially 100 per cent of the population. Moreover, the entire viral life cycle is short, so that crucial experiments can frequently be initiated and completed during a single working day.

B. IMMUNOLOGIC STUDIES 1 . Bacteriophages as Antigens The serum of animals that have recovered from a viral infection contains specific antibodies that neutralize the virus. This can be demonstrated by mixing the serum of a convalescing animal with a preparation of the free virus and then assaying for infectious capability. The sample containing neutralized virus has decreased power to produce an infection. The exact nature of the virus-neutralization reaction is not completely understood. Because most animal viruses cannot readily be obtained pure and in large amounts, and because animal virus assays are relatively tedious, bacteriophages constitute a useful model system for studying the antigenic properties of virus (Overby, 1967). The normal infectious process of a bacteriophage is interrupted if it is exposed to specific antibody. The antibody reacts with the viral coat proteins and thereby prevents either adsorption or penetration. For a number of years, the strength of a phage-neutralizing antiserum has been expressed as the rate constant (k) of a first-order reaction

Po k=- 2.3 D log T

Pt

where D is the dilution factor, T is time in minutes, Po is the initial number of phage and Pt is the number of plaque-forming units at time T. However, it has been recognized for an equally long time that a significant fraction of a viral population is refractory to neutralization by antiserum (Burnet et al., 1937).There is as yet no satisfactory explanation for this puzzling observation, that is, viral neutralization generally proceeds rapidly at first but later on the rate decreases. Various explanations have been proposed to account for this deviation from first-order kinetics; for example, the reaction of a phage with an antiserum may be reversible. Accordingly, with increasing time, previously inactivated particles may become infective again through

126

S. G. BRADLEY

dissociation of the antigen-antibody complex. It has been shown that plaque-forming capability can be restored by removing the antibody from the viral particle by enzymic digestion or by urea. However, neutralized phage particles diluted in buffer do not regain plaqueforming ability, even on prolonged storage. Another explanation for the decreasing rate of phage neutralization is that the antiserum is heterogeneous or that it becomes limiting during the course of the reaction. Conversely, the phage population might be heterogeneous or heterogeneity might de.velop during the course of the reaction. In general, antibody activity does not become limiting during the course of the neutralization assay. This has been demonstrated by adding additional antiserum when the rate of phage neutralization has markedly decreased; the rate of viral inactivation after the addition of fresh serum follows the characteristic curve of the late phase rather than the rapid initial phase. In addition, preexistent genetic heterogeneity is not usually responsible for the deviation from first-order inactivation kinetics. The progeny of surviving phage from the late phase of the neutralization assay are neutralized at the same rate as the original population. The “tailing effect” discussed here is commonly observed. The change in slope may be abrupt or gradual. Many “tailing effects” can be attributed to physical factors such as aggregation of viral particles or adsorption to walls of the reaction vessels (Hiatt, 1964). These physical interactions can occur in uiuo, thereby allowing virus to escape complete annihilation even in the presence of ample antibody in the serum. In fact, low concentrations of bacterial viruses do persist within animals for several months after injection (Bradley et al., 1961).

2. Zmmunosuppression There are a variety of potential clinical applications for immunosuppressive drugs. They are needed to prevent immune rejection of a transplanted organ; they have utility in the treatment of autoimmune diseases. A number of other medical problems involve deleterious consequences of the immune response; these include allergy, delayed hypersensitivity, arthritis, and some of the symptoms of aging. Conversely, immunostimulants have merit for the prophylactic and therapeutic control of infectious diseases (Santos, 1967). The major objective of immunosuppressive therapy is selective inhibition. A number of observations have accumulated, indicating that it is possible to suppress differentially circulating antibody responses from delayed


127

hypersensitivity (Hitchings, 1967). Impairment of the immune response may occur naturally, as in hypogamma-globulin anemia, or may be induced by X-radiation, cortisone treatment, dietary deficiency, immune tolerance, or paralysis, and by a variety of metabolic inhibitors. A useful assay for detection of immunosuppression or immunostimulation employs bacteriophage as the antigen. Phages have the desirable feature of superb antigenicity; moreover, the phage neutralizing assay for antibody activity is exquisitely sensitive and precise. For extensive surveys of potentially useful drugs, mice are the animal of choice because tests for immunosuppression of humoral antibody and of the graft-versus-host reaction can be readily carried out. Graft-versus-host reactions can be invoked by injecting intraperitoneally suspensions of spleen cells from another mouse strain into recipient animals; after 1 week, spleen enlargement is measured and used as the index of the graft-versus-host reaction (Papermaster et al., 1962). To elicit humoral antibody production, each mouse is injected intraperitoneally with 1O'O plaque-forming units of purified phage. In studying the early immune response, the animals are injected once; for hyperimmunization, five weekly injections are given. Five days after antigenic stimulation, blood samples are drawn and phage neutralizing activity determined. Neutralizing activity is recorded as the inactivation rate constant (k) described previously. The test compounds are usually given at 0.2 LDso daily for 4 days. Variations of dosage and regimen may be dictated by the nature of the test substance. Groups of animals injected with phage alone and drug alone are included in each experimental series to ensure that any alteration in neutralizing activity is an effect on the animals and not on the phage directly. From such surveys, a number of immunosuppressants have been discovered. Unfortunately, most of these inhibitors are also broad-spectrum cytotoxic agents and are not specific suppressors of antibody formation. Moreover, the known immunosuppressants equally affect the early and the secondary immune responses (Cooney and Bradley, 1963).

3. Molecular Genetics of the Onset of Antibody Formation The antibody formed during the first several days after injection of a bacteriophage into a mouse is a globulin having a molecular weight of about 950,000 daltons. After a week, the animal produces the classical gamma globulin antibody which has a molecular weight of 180,000 daltons. These observations, however, reveal little about

128

S . 6. BRADLEY

the genetic control of antibody formation. Watson, Kim, and I have used phage as a necessary tool to attack the question of what controls the onset of antibody formation (Kim et al., 1966a,b, 1967a,b). The first question that we asked was: What is the relationship between the antigenic dose and the amount of antibody produced? At the same time, we asked: What is the smallest amount of antigen that will give a measurable immune response? These studies were carried out with the mouse. We found that as few as 100,000phage particles induced a measurable increase in antibody concentration. This amount of phage represents lo-" gm. of protein, or 5 x lo7molecules (assuming that the phage protein is entirely broken into its polypeptide subunits). Moreover, we found that the amount of antibody produced increased with antigenic dose, up to 2 mg. phage protein/ animal, which was the largest antigenic dose tested. In general, the amount of antibody produced increased 2- to 3-fold for each 10-fold increase in antigenic dose. These results told us that minute amounts of antigen would give a measurable immune response and that the amount of antibody produced is a function of the amount of antigen injected. The second question that we asked was: How long does it take after antigenic stimulation to detect an immune response? Concurrently, we wanted to know the rate at which antibody increased in the serum. Again we used the mouse. We found increased amounts of antibody in the serum within 6 hours after antigenic stimulation. Moreover, the antibody content of the serum doubled every 6 to 7 hours during the first 3 to 4 days after antigenic stimulation. Obviously the immune response to phage in the mouse occurs very rapidly. At this point in the investigation, Watson and Kim turned their attention to the chemical nature of the early and late antibodies whereas my collaborators and I began to examine various aspects of normal antibody. Normal antibody is the antibody found in normal animals who have not knowingly been exposed to antigen. We found that normal antibody is primarily of the macroglobulin class, that is, has a molecular weight of 950,000 daltons, as does early induced antibody. We also found that mice can be induced to make macroglobulin antibody by oral administration of the antigen. Finally, we found that chemical immunosuppressants that effectively inhibit the induced, early immune response also effectively reduce normal antibody levels. Immunosuppressants that are less effective in inhibiting the induced early response are less effective in inhibiting normal antibody levels. We were led to suspect that normal antibody is, in reality, antibody induced by exposure of the animal to the same and cross-reacting


129

antigens, especially in the food. By this time, Watson and Kim had isolated and purified the phage-induced early and late antibodies. They found, contrary to other workers, that the early and late antibodies were made of the same subunits, but differently arranged. Early and late antibodies, however, do differ in molecular weight, in susceptibility to mercaptoethanol, and in their behavior on DEAEcellulose chromatographic columns (Kim et al., 1964). It was not long before Watson and Kim discovered the basis of the discrepancy between their work and other investigators’ results. Most immunochemists have isolated and characterized a serum fraction composed of a mixture of /32M (also called IgM) and 19 S IgG proteins. Their chemical analyses reflect the composition of the 19 S IgM protein more than that of the 19 S IgG (the latter is the early antibody). This error is widespread in the literature, and most papers equate 19 S IgM and 19 S IgG. By this time, we were convinced that the adult mouse had been exposed to, and had responded to, practically every known antigen. If we were to study the onset of antibody formation, we needed an animal system that was “immunologically virgin.” We examined many sera for their normal antibody content. We found no measurable gamma-globulin in fetal swine serum. All of our subsequent research has been carried out with the Minnesota miniature swine (Kim et al., 1965). We found immediately that piglets born naturally possessed measurable amounts of antibody, obtained by ingestion of maternal blood liberated during the birth process. Next we discovered that piglets obtained by hysterectomy developed antibodies within 1week in the laboratory, even though not intentionally exposed to antigens. Indeed, we have found that piglets kept germfree and fed a nonallergenic diet, develop detectable amounts of antibodies to most antigens within 1 month. In order to study the onset of antibody formation, we keep the animals germ-free and antigen-free for 3 to 4 days; after that the germ-free animals are fed a nonallergenic diet for the duration of the experiment. The serum of piglets kept antigen-free and germ-free for 4 days contains no antibody or gamma-globulin, whereas piglets injected with antigen produce measurable antibody within 36 hours. If the sow is injected with bacterial endotoxin, or develops an infection during pregnancy, the serum of the piglets usually contains antibody. However, the piglets from healthy sows, hyperimmunized with phage, contained no phageneutralizing antibodies. Apparently, the selective permeability of the placenta is altered by stress and disease. On the basis of these experiments, we are now convinced that all antibody is induced by antigen.

130

S. G .

BRADLEY

Watson and Kim have shown that early, induced antibody from piglet serum is composed of the same protein subunits as late antibody. Fortunately, the piglet does not produce 19 S /32M (19 S IgM) until the animal is a week or two old. Accordingly, it is easy to demonstrate in swine that the early antibody is 19 S IgG distinct from 19 S IgM protein. We have immunized germ-free piglets with antigenic doses that do not induce a maximal antibody response. When a piglet is simultaneously stimulated submaximally with two unrelated antigens, the amount of each antibody produced is less than that produced by piglets immunized with a single antigen. Because the animals were not maximally stimulated antigenically, the observed competition is not for antibody precursors. What have these experiments told us about the genetic control of antibody formation? Presently, all theories concerning the genetic control of antibody formation can be assigned to either of two concepts: In the clonal selection theory, it is assumed that the immunocytes are genetically heterogeneous and that the role of antigen is to select preexistent clones of antibody-producing cells; the instructional theory holds that antigen induces antibody formation in genetically homogeneous immunocytes. The clonal selection theory supposes that each immunocyte produces a limited amount of antibody even in the absence of antigen. If each immunocyte in the piglet produced and released only one molecule of antibody, the gamma-globulin cancentration in the serum should be 0.01 pg./ml. We have shown that there is less than 0.1 pg. gamma-globulin/ml. in piglet serum. Although our present data do not definitively rule out the clonal selection theory, they raise serious doubt about the validity of this popular idea. The clonal selection theory supposes that there is a genetically heterogeneous population of immunocytes. If the cells able to respond to two unrelated antigens represent different cell lines, then submaximal doses of two unrelated antigens should each induce the same amount of antibody, whether presented separately or simultaneously. This is not the case. We are, therefore, supporting the instructional idea. How do we fit this hypothesis into our present concepts of molecular genetics? We propose that an antigen is taken up by a phagocyte (macrophage). In the phagocyte, the antigen is degraded into several determinant groups. Each determinant is joined to a messenger RNA, which contains the information for that part of gamma-globulin which is common to all gamma-globulin molecules. The determinant m-RNA is then transferred to an immature plasma cell. We suggest that undedicated immature plasma cells have pro-


131

duced 10 to 20 m-RNA’s which code for the polypeptide subunits of the combining groups of all antibody molecules. By analogy to the isozymes of lactic dehydrogenase, we propose that specificity of the antibody for the antigen is conferred b y a specific arrangement of two polypeptides. Note that pairwise combinations of 20 subunits yield 1,000,000 different kinds of antibody molecules. The m-RNA’s made by structural genes of the immature plasma cell would bind to a ribosome or ribosomal precursor. Each m-RNA ribosome would then direct the synthesis of one molecule of polypeptide which we propose would remain attached to the m-RNA ribosome. The role of an antigenic determinant, according to our theory, is to organize polysomes which will direct the synthesis of specific antibody. The antigenic determinant serves as a focus around which appropriate m-RNA ribosomes will be gathered by virtue of an interaction between the hapten moiety of the determinant m-RNA and the polypeptidem-RNA-ribosome complex. The polysome so formed will now produce specific antibody. The early antibody has a molecular weight of 950,000 daltons and a valency of about 5. The 19 S IgG is probably released only upon death of the antibody-producing cell. As the immature plasma cell replicates and differentiates, it probably develops a system for secreting antibody. The secretory mechanism may convert 19 S IgG to the 7 S IgG molecular configuration. This latter bivalent antibody is the classical gamma-globulin. In the absence of additional antigen, the polysomes would be inherited semiclonally. Upon secondary presentation of antigen, new induction would occur; in addition, dedicated, dormant plasma cells would start to grow again. We call this hypothesis a messenger-selection theory (Bradley, 1966). We consider the problem of self-recognition as another aspect of immunologic tolerance and immunologic paralysis. In the latter two instances, heterologous living cells or massive amounts of foreign antigen render the animal unable to respond upon subsequent presentation of the antigen. Immunologically tolerant or paralyzed animals do respond to unrelated antigens. A recent, significant finding is that relatively small amounts of extensively degraded antigen can induce immunologic tolerance. We believe that small determinantbearing fragments are taken up by the plasma cell. They bind with the polypeptide of the m-RNA-ribosome-polypeptide complex and prevent formation of polysomes. The situation would be something like antibody excess in the precipitin reaction, in which each antigen is saturated with antibody so that cross-linking between antigenantibody complexes cannot occur.

132

S. G. BRADLEY

IX. Conclusions

Our present views about the biochemical and biophysical organization of genetic information are extensively based upon researches employing bacterial viruses. Similarly, some of the elegant and critical experiments on the control of cellular metabolism have utilized phage-infected bacteria. Recently, insight into the mechanisms by which macromolecules are put together to form structural and functional organelles has been provided by analyses of the interactions between conditional lethal mutants of bacteriophages. These discoveries have great relevance for present and future researches in industrial laboratories. The project leader of an antimetabolite synthetic program, for example, must be familiar with the metabolic consequences that a given analog might reasonably be expected to have. Desirable antimetabolites include not only antimicrobial, antitumor, and antiviral agents, but also include immunosuppressants, antiobesity agents, and serum cholesterol lowering compounds. Bacterial viruses also constitute models for the treatment of viral infections, both chemotherapeutically and immunologically. Moreover, studies on genetic transduction may provide the fundamental knowledge and the impetus for comparable research with viruses of man. Clearly, the transfer of genetic information is not restricted to temperate viruses having substantial genetic homology with the host DNA. Repair of genetic lesions that result in metabolic diseases may some day be achieved by viral-mediated transfer of genetic information from a normal donor to a defective recipient. Because viruses possess tissue specificity, this mechanism may be much more efficient than genetic transformation, which uses free DNA and is therefore subject to extensive dilution in the tissue fluids and is limited by the ubiquitous nucl ease s. Bacterial viruses may have utility in controlling bacterial populations in nature. Blue-green algae, for example, are frequently responsible for undesirable odor and tastes in water, and may produce large amounts of slime. Phages for these organisms might be added to aquarium water, irrigation sluices, and paper-pulping water reservoirs. Conversely, phages in nature may present more of a problem than is generally recognized. Sewage disposal, particularly where the effluent contains a large amount of soluble organic material, constitutes a serious problem in today’s highly industrialized society. In order to prevent water pollution, the food processing and fermentation industries must cooperate in a concerted study of the ecological factors affecting waste disposal.


133

The role of bacterial viruses in the industrial research laboratory is not restricted to the long-range, and perhaps far-fetched areas discussed above. Bacterial viruses have established utility for culture identification. Bacteriophages do present a significant problem in a number of food, antibiotic, and amino acid fermentations. Methods for monitoring and controlling phage contamination should be an accepted part of the standard operating procedure of any bacterial fermentation. Bacteriophages constitute useful tools for biochemical and immunologic studies. Although bacterial viruses are work horses in the academic laboratory, their beneficial and detrimental qualities have not yet been fully appreciated or exploited by the industrial microbiologist.

REFERENcEs Albertsson, P. A. (1960). “Partition of Cell Particles and Macromolecules.” Wiley, New York. Arber, W. (1965a).Ann. Reo. Microbiol. 19,365-378. Arber, W. (1965b).Pathol. Microbiol. 28,71-72. Arber, W., and Dussoix, D. (1962).J. Mol. Btol. 5, 18-36. Babel, F. J. (1962).Adoan. Appl. Microbiol. 4, 51-75. Barber, M. (1949).J. Gen. Microbiol. 3,274-281. Barksdale, L. W. (1959). Bacteriol. Reo. 23,202-212. Bertani, G., and Weigle, J. J. (1953).J . Bacteriol. 65, 113-121. Blair, J. E., and Williams, R. E. 0. (1961). Bull. World Health Organ. 24, 771-784. Bradley, S. G. (1959).Mycologta 51,9-17. Bradley, S . G. (1966).Aduan. Appl. Microbiol. 8,29-59. Bradley, S. G., and Ritzi, D. (1967). Deoelop. Ind. Microbiol. 8,206-213. Bradley, S . G., Papermaster, B. W., Watson, D. W., and Good, R. A. (1961).Proc. SOC. Exptl. Biol. Med. 108, 79-82. Bradley, S. G., Lee, B. E., and Jones, L. A. (1963).Develop. Ind. Microbiol. 4,288-294. Brenner, D. J., Falkow, S., and Cowie, D. B. (1967).Science 156,535-536. Bumet, F. M., Keogh, E. V., and Lush, D. (1937).AustraZianJ.Exptl. Biol. Med. Sci. 15,227-247. Clowes, R. C. (1963). Genet. Res. 4, 162-165. Cooney, W. J., and Bradley, S. G. (1963).Antimicrobial Agents and Chemotherapy 1962,l-9. Cordes, S., and Epstein, H. T. (1961).Nature 191,1097-1098. Cowie, D. B. (1967). Science 156, 537. Cowie, D. B., and McCarthy, B. J. (1963). Proc. Natl. Acad. Sci. U.S. 50, 537-543. Edgar, R. S., and Wood, W. B. (1966).Proc. Natl. Acad. Sci. US.55,498-505. Epstein, R. H., Bolle, A., Steinberg, C. M., Kellenberger, E., Boy d e la Tour, E., Chevalley, R., Edgar, R. S., Susman, M., Denhardt, G. H., and Lielausis, A. (1963).Cold Spring Harbor Symp. Quant. Btol. 28,375-392. Foldes, J,, and Traufner, T. A. (1964). Z. Vererbungslehre 95, 57-65. Fraser, D. (1957).J. Biol. Chem. 227,711-715.

134

S . G. BRADLEY

Fredericq, P. (1963). Ergeb. Mikrobiol. Zmmunitaetsforsch. Exptl. Therap. 37, 114161. Green, D. M. (1964).]. Mol. Biol. 10,438-451. Green, M. H. (1963). Proc. Natl. Acad. Sci. U.S. 50, 1177-1184. Harm, W., and Rupert, C. S. (1963).2. Vererbungslehre 94, 336-348. Hattman, S., and Fukasawa, T. (1963). Proc. Natl. Acad. Sci. U.S. 50,297-300. Hiatt, C. W. (1964). Bacteriol. Reo. 28, 150-163. Hitchings, G. H. (1967). Federation Proc. 26,958-960. Hohann-Berling. H., and Maze, R. (1964).Virology 22, 305-313. Iseki, S., and Sakai, T. (1953). Proc. Japan. Acad. 29, 121-126. Jarolmen, H., Bondi, A., and Crowell, R. L. (1965).J. Bacteriol. 89,1286-1290. Jones, L. A. (1964). Ph.D. Thesis, Univ. of Minnesota, Minneapolis, Minnesota. Jones, L. A., and Bradley, S. G. (1962). Deuelop. Znd. Microbiol. 3,257-264. Jones, L. A,, and Bradley, S. G. (1964).Mycologia 56,505-513. Kim, Y. B., Bradley, S. G., and Watson, D. W. (1964).J. Zmmunol. 93,798-805. Kim, Y. B., Bradley, S. G., and Watson, D. W. (1965).“Swine in Biomedical Research,” pp. 273-284. Frayn Printing, Seattle, Washington. Kim, Y. B., Bradley, S. G., and Watson, D. W. (1966a).J . Zmmunol. 97, 52-63. Kim, Y. B., Bradley, S. G., and Watson, D. W. (1966b).J . Zmmunol. 97, 189-196. Kim, Y. B., Bradley, S. G., and Watson, D. W. (1967a).J. Zmrnunol. 98, 868-873. Kim, Y. B., Bradley, S. G., and Watson, D. W. (1967b).J . Zmmunol. 99, 320-326. Kolstad, R. A., and Bradley, S. G. (1964).J. Bacteriol. 87,1157-1161. Kolstad, R. A., and Bradley, S. G. (1966).J . Bacteriol. 91, 1372-1373. Korman, R. 2.(1966).J. Bacteriol. 92,762-768. Korman, R. Z., and Berman, D. T. (1962).J . Bacteriol. 84,228-236. Kottman, U. (1942).Arch. Ges. Virusforsch. 2,388-396. Lederberg, S. (1957). Virology 3, 496-513. Liu, P. V. (1958).Am. J . Clin. Pathol. 29, 176-180. Ldria, S. E. (1962). Science 136,685-692. Luria, S. E., and Human, M. L. (1952).J . Bacteriol. 64,557-569. Mitsuhashi, S., Hashimoto, H., Kono, M., and Morimura, M. (1965).J. Bacteriol. 89, 988-992. Morse, M. L. (1959). Proc. Natl. Acad. Sci. U.S. 45, 722-727. Moseley, W. K., and Winslow, R. L. (1959).J . Daity Sci. 42,906. Nomura, M. (1967). Ann. Reo. Microbiol. 21, 257-284. Novick, R. P., and Morse, S. I. (1967).J. Enptl. Med. 125, 45-58. Okubo, S., Stodolsky, M., Bott, K.,and Straws, B. (1963). Proc. Natl. Acad. Sci. U.S. 50, 679-686. Overby, L. (1967). BioScience 17,807-812. Papermaster, B. W., Bradley, S. G., Watson, D. W., and Good, R. A. (1962).J . Erptl. Med. 115, 1191-1210. Pattee, P. A., and Baldwin, J. N. (1962).J. Bacteriol. 84,1049-1055. Peterson, E. A., and Sober, H. A. (1956).J . Am. Chern. Soc. 78,751-755. Reeves, P. (1965).Bacteriol. Reo. 29,25-45. Reilly, B. E., and Spizizen, J. (1965).J . Bacteriol. 89,782-790. Ritz, H. L., and Baldwin, J. N. (1961).Proc. Soc. Erptl. Biol. Med. 107,678-680. Robbins, P. W., and Uchida, T. (1962). Federation Proc. 21, 702-710. Safferman, R. S., and Morris, M. E. (1963). Science 140,679-680. Santos, G . W. (1967). Federation Proc. 26,907-913.


135

Seaman, E., Tarmy, E., and Mannur, J. (1964).Biochemistry 3,607-612. Schlesinger, M. (1933).Biochem. 2. 264,6-12. Sinsheimer, R. L. (1959).Brookhawen Symp. Biol. 12,27-34. Stent, G . S. (1963).“Molecular Biology of Bacterial Viruses.” Freeman, San Francisco, California. Takagi, J., and Ikeda Y. (1962).Biochem. Biophys. Res. Commun. 7,482-485. Gen. Microbiol. 31, 211-217. Takahashi, I. (1963).J. Takahashi, I. (1965).J. Bacteriol. 89, 1065-1067. Takahashi, I., and Marrnur, J. (1963).Biochem. Biophys. Res. Commun. 10,289-292. Tewfik, E.,and Bradley, S. G . (1967).J. Bacterial. 94, 1994-2000. Thorne, C. B. (1962).J.Bacteriol. 83,106-111. Tocchini-Valentini, G. P., Stodolsky, M., Aurisicchio, A., Sarnat, M., Graziosi, G., Weiss, S. G., and Geiduschek, E. P. (1963).Proc. Natl. Acad. Sci. US.50,935-946. Uetake, H.,and Hagiwara, S. (1961).Virology 13,500-506. Uetake, H.,Luria, S., and Burrows, J. W. (1958).Virology 5, 68-91. Baker, W. H., and Brown, R. K. (1958).Virology 5, 327-336. Van Vunakis, H., Watanabe, T. (1963).Bacteriol. Reo. 27, 87-115. Weigle, J. (1966).Proc. Natl. Acad. Sci. U.S. 55, 1462-1466. Wentworth, B. B. (1963).Bacteriol. Rev. 27,253-272. Whitehead, H. R.,and Cox, G. A. (1935).New Zealand J . Sci. Technol. 16, 319-320. Wood, W. B. (1966).J. Mol. B i d . 16, 118-133. Zierdt, C. H., Fox, F. A., and Norris, G. F. (1960).Am. J . Clin. Pathol. 33,233-237. Zinder, N. D.(1965).Ann. Reo. Microbiol. 19,455-472. Zinder, N. D.,and Lederberg, J. (1952).J. Bacteriol. 64,679-699.

This page intentionally left blank This Page Intentionally Left Blank

Proteins and Enzymes as Taxonomic Tools

EDWARDD. GARBERAND

JOHN

w.

RIPPON

Departments of Biology and Medicine (Dermatology), Division of Biological Sciences, The University of Chicago, Chicago, lllinois

I. Introduction .......................................... A. The Species Concept in Microbiology

.....................

B. The Role of Biochemistry in Microbial Taxonomy ...... 11. Zone Electrophoresis ................................................... A. Purpose ............................................................... B. Variation in Electrophoretic Patterns ....................... C. Methodology ........................................................ 111. Protein Profiles and Zymograms in Microbial Taxonomy ... A. Fungi .................................................................. B. Bacteria ............................................................... C. Protozoa ............................................................... D. Algae ................................................................... IV. General Comments ...................................................... References ..................................................................

137 137 138 140 140 140 141 144 144 147 151 151 151 153

I. Introduction A. THE SPECIESCONCEPTIN MICROBIOLOGY According to the International Code of Nomenclature, organisms must be assigned to a species. When an organism is to be named, a dossier of its characteristics is prepared to determine whether the organism has already been described. In the event the organism does not fit an already described species, it must be described to receive recognition as a “new” species according to the appropriate Code. This ritual constitutes a memorial to Linnaeus who created order in the previously chaotic classification of plant and animal macroorganisms. In the beginning, microbial classification followed Linnaean protocol. Although considerable progress has been made in the preparation of increasingly more sophisticated dossiers, the microbial taxonomist seems to have produced increasingly greater Linnaean chaos in certain groups. Perhaps the situation is best appreciated by considering an opinion offered by a macrobial taxonomist (Heslop-Harrison, 1962): “Whatever other function might be attributed to taxonomy - such, for example, as investigating phylogeny -this one, the creation of a data storage and retrieval agency, is surely the inescapable one.” 137

138

EDWARD D . GARBER AND JOHN W . RIPPON

The Twelfth Symposium of the Society for General Microbiology was devoted to “Microbial Classification” (Ainsworth and Sneath, 1962). Although the protozoologists, phycologists, and a number of mycologists appeared to be comfortable with the “species problem,” the bacteriologists ranged from optimistic to anarchistic in their attitudes. One bacterial taxonomist (Cowan, 1962) used a provocative title to present his views: “The Microbial Species- A Macromyth?“ An interesting attitude concerning microbial taxonomy was revealed in Appendix I of the Symposium: “As there is no objective definition of species for most microorganisms and the tradition of species varies from one group to another the best advice to a microbiologist with little taxonomic experience who believes he is working with an undescribed organism which he wishes to propose as a new species is for him to consult a specialist on the taxonomy of the particular microorganism in question.” In other words, the proprietor of the latest and most extensive data storage and retrival agency may serve as a consultant for microbiologists with taxonomic problems. The paucity of morphological characteristics in bacteria, for example, is presumably responsible for the usual taxonomic skirmishes involving the “splitters” and the “lumpers,” the gross, fine, and ultrafine morphologists and the biochemists comparing small, large, and extralarge molecules, 01 the purists and the diagnosticians. The culturing of microorganisms, particularly fungi and bacteria, isolated from their native habitats had long sensitized microbiologists to the diverse nutritional requirements and to the responses of microbial populations growing in media with different components and supplements under different physical regimes. Finally, the animal and plant pathogens acquired unique status in that the means to control infectious disease demanded reasonably practical methods for classifying the pathogens with similar or different hosts which would exhibit different syndromes. In some cases, specialized and vested interests required a practical taxonomy which would serve as a useful data storage and retrieval agency. Unfortunately, the microbial geneticist interfered with this purpose by demonstrating a trivial basis (single gene-mutation) for a number of apparently meritorious taxonomic characters in a number of microbial groups. B. THE ROLE OF BIOCHEMISTRY IN MICROBIAL TAXONOMY

The biochemical and physiological characteristics of microbial populations have furnished many criteria for systematic purposes mostly at the level of the species and to some extent, the genus:

PROTEINS AND ENZYMES AS TAXONOMIC TOOLS

139

fermentation of sugars and the production of acid, the assimilation of organic carbon or nitrogen compounds, biosynthetic pathways and their associated enzymes, to name the obvious ones. Groups of illdefined taxonomic status were created in bacterial groups on the basis of reactions (lysis) to bacterial viruses and in fungal phytopathogens on the basis of host range. Serological methods such as immunodiffusion and immunoelectrophoresis, to mention the refined techniques, provided additional means to detect groups with close or distant affinity. It would seem that the cup of the microbial taxonomist is brimming with methods and techniques to characterize populations and therefore to construct reasonable taxonomic treatments. According to Cowan (1962), “The microbial species does not exist; it is impossible to define except in terms of a nomenclatural type, and it is one of the greatest myths of microbiology.” To counter this stark appraisal of the situation, he suggests that “On the other hand, better characterization and separation of microbial units will depend on the progress made by biochemists in analyzing the different parts of microbes and substituting a biochemical characterization for the old biological ones based on simple morphology, staining and sugar reactions, serology, and so on.” It might be reasonable to extend the nomenclatural Code for microbial populations to acceptable infraspecific categories rather than to apply phylogenetic significance to such categories and to seek additional biochemical criteria to characterize microbial populations. Biochemical reactions involving specific substrates, the assimilation of organic compounds as sources of energy, bacteriophage specificity, and serological relations were used as taxonomic criteria long before a relation between these criteria and the genic determinants was established. With the formulation of the one gene-one enzyme (one cistron-one polypeptide) hypothesis, characters with taxonomic value could be viewed as expressions of fundamental differences or similarities in the organization of the genetic information in taxa and related to differences or similarities for particular proteins. The taxonomic treatments, however, had presumed all-ornone expressions of characteristics so that differences for the allexpressed characters could not be detected. Practical techniques were necessary to compare proteins or enzymes which would yield meaningful observations with respect to differences or similarities. Zone electrophoresis seems to have provided a useful tool for comparing proteins and enzymes to determine whether differences or similarities might have taxonomic value.

140

EDWARD D. GARBER AND JOHN W . RIPPON

II. Zone Electrophoresis

A. PURPOSE Zone electrophoresis provides a means to characterize the proteins or enzymes from different sources such as cellular extracts or culture filtrates by comparing their number, location (anodic or cathodic), and relative mobility in terms of travel from the origin. It should be clear that identical electrophoretic mobility for proteins or enzymes from different sources does not necessarily imply an identity of these proteins or enzymes. Biochemical characters including electrophoretic patterns of proteins or enzymes are subject to the same limitations as the conventional characters used in systematic studies. Furthermore, certain biochemical characters have been useful in clarifying taxonomic problems in some groups but not in other groups. Electrophoretic patterns may be expected to have value for some groups and not others. “The more difficult a systematic problem is, the more likely that an incomplete solution will be obtained, regardless of whether the method is morphological or molecular. The advantage of using protein characteristics is that any given protein molecule is the ultimate product of only one or two cistrons; accordingly the protein systematist is close to the genotype and yet still concerned with the fundamental basis of phenotypes” (Manwell and Kerst, 1965).

B.

VARIATION IN

ELECTROPHORETIC PATTERNS

Related species may include populations with different electrophoretic patterns and yet share common sites of protein or enzyme activity. At least four explanations may be offered to account for such observations. First, the species may be valid but occasional interspecific hybridization (conjugation, transduction, transformation) may allow gene flow in either or both directions, that is, introgressive hybridization. This explanation is reasonable when barriers to hybridization are incomplete and the hybrids yield progeny. Second, the species is polymorphic and includes intraspecific categories which merit recognition. Third, polymorphism may have been present in the common ancestor and was maintained during the evolution of related species from this ancestor. Fourth, convergent evolution might have occurred by mutations in related species so that proteins or enzymes with identical electrophoretic mobilities arose. Obviously a reasonable survey of strains of a species as well as species of a genus should be made to determine the value of electrophoretic patterns as adjuncts in a systematic survey of a group of organisms.


141

C. METHODOLOGY A mixture of charged compounds may be separated into reasonably discrete regions when a direct current is applied to the system for a period of time. In the case of proteins and enzymes, the net charge depends on the amino acids of the primary structure of the polypeptides and of associated moieties as well as the pH of the buffer in the system. A protein or enzyme is neutral at its isoelectric point and has a net negative charge as the buffer pH approaches alkaline values or a net positive charge as the buffer pH approaches acidic values. For example, when two proteins differing in their isoelectric points (pH 4.7 and 7.2) are dissolved in a buffer of pH 8.6, the protein with the acidic isoelectric point will migrate farther than the protein with the basic isoelectric point. The shape and dimensions of the proteins and enzymes also play a role in their electrophoretic mobility so that the pore size of the solid supporting medium will have a sieving effect which must be considered. Various supporting media have been used for zone electrophoresis: paper, cellulose acetate membrane, and gels prepared from agar, starch, or acrylamide to mention the most common types. A symposium (Whipple, 1964) devoted to the theory and techniques of zone electrophoresis and to its application to clinical, biological, and enzymological problems and a laboratory manual (Nerenberg, 1966)have recently appeared. Although the cellulose acetate membrane provides a convenient supporting medium for the relatively rapid separation of proteins or enzymes, only relatively small volumes may be applied to the membrane. While this limitation may not be significant when solutions contain relatively few proteins or enzymes in relatively high concentration, membranes do not provide sufficient material for further study after elution. Consequently, gels are generally employed to handle relatively large volumes of material. Furthermore, gels may be cut along their long axis to yield a number of slices and each may be used to detect proteins as well as a number of different enzyme systems. The apparatus for zone electrophoresis may be fabricated or purchased. The relative merits of the different types of apparatus are generally determined by the supporting media which will be used, the need to detect proteins or enzymes or both, the number of samples to be examined and, not least, the versatility and cost. In general, larger volumes of material may be used for starch and acrylamide gel than for agar gel or cellulose acetate membrane. When small differ-

142

EDWARD D. CARBER AND JOHN W. RIPPON

ences in electrophoretic mobility must be established, the apparatus should allow for the addition of a number of samples to the same gel. Finally, a source of stable direct current with either constant voltage or amperage is necessary. The choice of buffers for preparing gels is usually determined rather empirically, as far as the components, pH, and ionic strength are concerned. Buffers of low ionic strength are associated with relatively rapid migration of proteins and enzymes, fuzzy zones, and little heating of the gel; those of high ionic strength with relatively slow migration, sharp zones, and some heating of the gel. The buffer used to prepare the gel may or may not differ in composition fiom that in the electrode compartments. In a continuous buffer system, the same buffer is used for both gel and compartments but with a lower molarity for the gel; in a discontinous buffer system, different buffers are used for the gel and for the compartments. An extensive literature is now available for seeking appropriate recipes for both types of buffer systems. Procedures for obtaining cellular extracts are usually determined by the purpose of the experiments and b y the enzyme systems to be studied. When extracts from a large number of cultures are to be surveyed, crude extracts may suffice and the procedures should be reasonably rapid and relatively efficient. When proteins are to be separated and the extracts are known to contain numerous proteins in concentrations sufficiently great so that the gels will contain many sites, it may be difficult to detect meaningful differences among the extracts. In such cases, the extracts may have to be treated and then compared. For example, fractions obtained by means of ammonium sulfate precipitation, column chromatography, or heating at a constant temperature for different periods of time could be used. The sites of the separated proteins may be visualized by immersing the gel or membrane in a solution containing an appropriate stain such as amido black. The relative concentrations of proteins or enzymes in the extracts may be so high that two proteins or enzymes differing only slightly in their electrophoretic mobility may yield only one detectable site. Broad zones of proteins or enzymes may be resolved b y diluting the extract, adding different inhibitors to the gel or reaction mixture in the case of enzymes, or heating the extract prior to electrophoresis. By varying the pH of the buffer used to prepare the gel, the voltage or amperage of the current, the duration of the time for separation, or the concentration of the starch or acrylamide in the gel, it may be possible to improve the resolution of the sites of protein or enzyme.


143

Hunter and Markert (1957) used a histochemical technique to visualize the sites of enzyme activity in the developed gels and the pattern of such sites has been termed a zymogram. Enzymes with similar or identical activity for a specific substrate but differing in their electrophoretic mobility have been termed isozymes or isoenzymes (Markert and Mhller, 1959). By coupling the enzymic reaction to another reaction system to obtain a colored compound, the sites of a number of enzymes may be visualized. Because the product rather than the protein is detected in the zymogram, relatively little protein is needed to obtain suitable zymograms. In fact, it is possible to detect sites of activity for an extract which would not yield detectable sites of protein after electrophoresis. Extracts which have been concentrated may yield more sites of activity than crude extracts. In screening many extracts, it may be necessary to use the crude extracts rather than attempt the effort needed to concentrate the extracts. The principle involved in preparing zymograms will be illustrated for lactic acid dehydrogenase (LDH). When the developed gel is immersed in the appropriate reaction mixture containing NAD as coenzyme, the sites of LDH activity catalyze the dehydrogenation of lactate to produce pyruvate. The liberated H+ reduces NAD to NADH which in turn is oxidized to NAD by the phenazine methosulfate in the reaction mixture. The H+ liberated from NADH then reduces the soluble, colorless tetrazolium salt to an insoluble, colored formazan. In practice, the gel is immersed in the reaction mixture and then incubated either at room or elevated temperature until sufficient color is generated at the sites of the enzymes. Once the gel exhibits sufficiently colored sites of enzyme activity, it is rinsed and photographed or translated into a diagram expressing relative electrophoretic mobility and intensity of color at each site. By storage under water, treatment with solutions, or wrapping in a thin plastic sheet, the stained gel may be saved for future reference. Sites of enzyme activity in developed gels may be detected without immersing the gels in a reaction mixture. Strips of filter paper saturated with the appropriate reaction mixture may be applied to the cut surface of the gel, incubated in situ,removed, and then processed to visualize the site of enzyme activity. The reaction mixture may be dissolved in clear, molten agar which is poured to give a thin layer over the surface of a sliced gel and then incubated until sites of enzyme activity can be detected. A zymogram may be obtained for certain enzymes when the reaction yields a compound detectable by ultraviolet light. In such cases, the sites of enzyme activity may be detected without resorting

144

EDWARD D. CARBER AND JOHN W . RIPPON

to certain expensive tetrazolium salts. It is reasonable to assume that the means to visualize sites of enzyme activity may be rather simple when the reaction is straightforward. For example, sites of amylase activity in starch gels developed in the cold room may be visualized by adding an iodine solution to gels which have been incubated so that the sites of enzyme activity will not yield the blue-black color of stained starch. Preparative acrylamide electrophoresis in conjunction with a fraction collector provides a means to detect isoenzymes when histochemical techniques are not available for zymograms. Racusen and Foote (1966)developed a relatively simple apparatus which was used to demonstrate isoenzymes of peroxidase and malate DH in extracts from bean leaves. Each fraction collected in test tubes provided sufficient material to assay for these enzymes and a plot of the activity in each tube gave a pattern equivalent to that observed in zymograms. Reaction mixtures and procedures for obtaining zymograms for a number of enzyme systems have been published by Hubby and Lewontin (1966) and a recent review by Shaw (1966) lists numerous publications which present additional reaction mixtures. Zymograms may be prepared for such enzyme systems as esterase, acid or alkaline phosphatase, leucine aminopeptidase. lactate DH, malate DH, glycerophosphate DH, xanthine DH, glucose 6-phosphate DH, 6phosphogluconate DH, alcohol DH, phosphoglucomutase, carbonic anhydrase, amylase, and catalase, and the list is being rapidly extended. 111. Protein Profiles a n d Zymograms in Microbial Taxonomy

We have surveyed the available literature to assess the claims made for the value of zone electrophoresis to separate proteins and enzymes. Cellular extracts or culture filtrates have been used for electrophoresis and the comparisons here involved either patterns of sites of stained protein (protein profiles) or patterns of sites of enzyme activity (zymograms). A. FUNGI 1 . Protein Profiles

Chang, Srb, and Steward (1962) distinguished four species of Neurosporu by their profiles, using mycelial extracts from one strain of each species. Different profiles, however, were obtained for a


145

wild-type and a mutant strain of N. crassa. These observations indicated that differences might be expected for protein profiles but a survey of strains would be necessary to determine the extent of intraspecific variation. Mycelial extracts from a number of strains of six species of Pythium gave protein profiles which exhibited little intraspecific variation and could be used to characterize each species (Clair, 1963).A comparison of these profiles with profiles from mycelial extracts of Pullaria pullulans, Fusarium oxysporum, and Saccharomyces cerevisiae indicated that each species had a characteristic profile. Similar observations have been reported for strains of three species of Phytophthora (Clare and Zentmyer, 1966; Gill and Powell, 1967). Profiles for mycelial extracts from 12 isolants of each of three species of Septoria displayed intraspecific qualitative variation but no evidence was offered in the preliminary report that the profiles might be used to characterize each species (Durbin, 1966). The profile for an isolant, however, remained constant when extracts were prepared from cultures grown for different periods of time in each of three defined media. Although significant intraspecific variation was detected for profiles of strains in each of four species of Ceratocystis, the interspecific differences were more pronounced than the intraspecific differences (Stipes, 1967). Shechter et al. (1966) compared profiles from extracts of one strain from two species of Microsporum, four of Trichophyton, and one of Epidermophyton which were grown in different complex media. The profiles for a strain grown in the different media had different patterns. Although profiles from different species grown in the same medium were compared in terms of the number of sites with the same electrophoretic mobility, the profiles were shown to characterize each species. Novacky and Macko (1966) found at least 23 common sites in profiles for extracts from two physiological races of Puccinia graminis var. tritici and claimed to be able to distinguish these races by a single difference. It is difficult to assess this claim because no mention was made of the number of replicates which were used for each race. Protein profiles were obtained for culture filtrates from cultures of two species of Microsporulh, four of Trichophyton, and one of E p i dermophyton grown in two different complex media (Shechter et al., 1966). Although the medium was shown to influence the pattern of sites of protein, each species could be characterized b y its profile when the cultures were grown in the same medium. Similar results

146

EDWARD D. GARBER AND JOHN W. RIPPON

were obtained for a more extensive survey involving seven species of Microsporum, 12 of Trichophyton, one of Keratinomyces, and one of Epidermophyton grown in the same complex medium (Schechter, 1967a). The range of intraspecific variation was determined for four to six isolants of M . gypseum, M . canis, T . mentagrophytes, T. tonsurans, T . rubrum, and E . floccosum (Schechter, 1967b). No significant intraspecific variation was detected and each species could be characterized by its protein profile.

2. Zymograms Tsao (1962) used mycelial extracts from seven strains of N. crassa grown in a defined medium or in two different complex media to obtain zymograms for malate DH, isocitrate DH, glucose 6-phosphate DH, and 6-phosphogluconate DH. No significant differences were detected for comparable zymograms from these strains nor for strains grown in the different media. Invertase zymograms for several strains of this species were usually found to have two sites of activity but one strain gave a single site (Eilers et al., 1964). Mycelial extracts of three strains of Phytophthora cinnamomi, two of P. palmiuora and one of P . citrophthora were used to obtain zymograms for glucose 6-phosphate DH, 6-phosphogluconate DH or peroxidase activity (Clare and Zentmyer, 1966). Although identical zymograms were observed for the strains of P . cinnamomi, differences were found for those of P . palmiuora. When cultures of P . palmivora and P. citrophthora were grown for different periods of time and mycelial extracts were used for glucose 6-phosphate D H zymograms, no differences in the zymograms for P . palmiuora were observed in contrast with the zymograms for P. citrophthora. Nealson and Garber (1967) used mycelial extracts from 32 strains representing 15 species of Aspergillus to prepare esterase, phosphatase, and leucine aminopeptidase zymograms. A sufficient number of strains was available for four species to determine the range af intraspecific variability. Although the strains gave different patterns of esterase and phosphatase activity, it was possible to detect invariant sites for strains assigned to the same species. A comparison of esterase and phosphatase but not leucine aminopeptidase zymograms revealed that each species could be associated with a characteristic pattern for both esterases and phosphatases. One strain of A. rugulosus differed in both its esterase and phosphatase zymograms from those for the other seven strains. Attempts to cross this strain with other strains were not successful. McClements and Garber (1967) surveyed 79


147

strains of A. nidulans using esterase and phosphatase zymograms to determine the extent of intraspecific variation. Only two strains differed from each other and from the remaining 77 strains for both esterase and phosphatase zymograms. A similar survey of 11 strains of A. heterothallicus indicated no intraspecific variation (Hayhome and Garber, 1967). Esterase and phosphatase zymograms were obtained for concentrated culture filtrates from eight formae speciales of Fusarium oxysporum, one strain of F . xylarioides, and one strain of Fusarium sp. (Meyer et al., 1964). Each strain had a characteristic array of sites with respect to their number and location (anodic or cathodic). Strains belonging to the same forma did not share more sites with similar electrophoretic mobility than strains belonging to different formae. Spontaneous morphological mutants gave zymograms which significantly differed from those of the wild-type strains. Furthermore, different esterase zymograms were obtained from the filtrates when strains were grown in a defined medium containing either sucrose or pectin as the sole source of organic carbon. The esterase rather than the phosphatase zymograms appeared to characterize each strain when cultures were grown in the same defined medium. Henry and Garber (1967) used concentrated culture filtrates from 23 strains representing nine physiological races of Colletotrichum lagenarium to obtain esterase zymograms. Although 20 strains yielded 12 sites which could be accommodated by eight anodic and seven cathodic sites, three strains gave unique patterns which did not conform with each other nor with any of the other strains. The absence of invariant sites precluded characterizing this species b y means of esterase zymograms.

B. BACTERIA 1. Protein Profiles The first report on the use of zone electrophoresis to characterize different microbial species appears to be that of van Riel et al. (1960) who found 16 detectable protein sites for extracts from one strain of Leptospira ictzrohaemorrhagica and seven sites from one strain of L. bijlexa. Huisingh and Durbin (1967) used the supernatant fluid from centrifuged cellular extracts to obtain profiles for the soluble proteins in determining whether differences could be detected for three species of Agrobacterium. No significant intraspecific differences were observed for four to six strains of A . tumefaciens, A. rhizogenes, or

148


A. radiobacter and the strains from each species could be characterized by their profiles. The soluble fraction of extracts of strains of Mycoplasma subjected to sonication was used to prepare protein profiles and each strain could be distinguished by comparing the patterns of sites (Fowler et al., 1963).Rottem and Razin (1967) isolated the cellular membrane from strains of five species of Mycoplasma and from four isolants which had not been assigned and treated the membranes with a solution of phenol and acetic acid to obtain soluble protein for electrophoresis. They reasoned that the membranes provided homologous structures so that the protein profiles would be more meaningful than the profiles from the heterogeneous cellular extracts. Each species was characterized by its profile and the groupings of strains agreed with those previously established by serological methods. For example, three strains assigned to the avian species M . gallisepticum formed one group on the basis of their profiles and serological affinity but two strains assigned to this species gave different profiles and serological reactions. The profiles and serological affinities for the strains were sufficiently related to warrant the conclusion that the profiles could be used for taxonomic purposes and could replace the expensive antisera usually required to distinguish strains and species of this genus. Gottlieb and Hepdon (1966) detected 7-12 major sites of soluble protein from cellular extracts of five strains of Streptomyces griseus, four of S . venezuelae, and one of S. orientalis, S . ramulosus, S . antibioticus, and Streptoverticillatum subsp. azocolatum by means of disk electrophoresis. The reasonably good similarity of protein profiles for the strains of s. griseus and of s. venezuelae indicated that these species might be characterized by their profiles. The authors appraised the value of protein profiles in the taxonomy of Streptomyces: “It is obvious that the polyacrylamide gel technique could not be used at present as a general procedure in which all protein bands from one extract were compared with all those from extracts of many other species. It can serve, however, to help separate or identify one streptomycete from another or a few other closely related streptomycetes.” Lund (1965)compared protein profiles for cellular extracts of strains of Streptococcus faecalis vars. zymogenes andliquifaciens, S.faecium, and S . durans. The strains of S. faecalis generally gave similar profiles for the major protein sites regardless of variety or serotype and could be distinguished from those for the strains of both S. faecium


149

and S. duruns. Furthermore, the profiles for the strains of the latter species were similar. These observations supported earlier evidence obtained from biochemical reactions and serology.

2. Zymograms Norris and Burges (1963) reported an interesting application of zymograms obtained from cellular extracts of the crystalliferous bacteria pathogenic for insects. The need to screen a large number of isolants and strains which were used in this investigation on the outbreaks of disease in insect populations was solved by preparing and comparing esterase zymograms using cellular extracts. Four esterase patterns accommodated the zymograms from the approximately 70 strains and isolants and the stability and reproducibility of the patterns were established by preparing zymograms for the strains or isolants on more than 20 separate occasions over a period of 1 year. According to the authors, “Esterase analysis was instrumental in unravelling the network of disease outbreaks described here.” In this investigation, the identification of strains isolated from diseased insects was determined solely by esterase zymograms. Cann and Wilcox (1965) used extracts to obtain catalase and esterase zymograms for seven strains of Mycobacterium rhodochrons, eight of M . phlei, nine of M . fortuitum, three of M . smegmatis, three of M . butyricum, and three which were not assigned to determine the extent of intraspecific and interspecific variation. Although the catalase zymograms could not be used to distinguish strains or species, the esterase zymograms were responsible for several significant observations. Two strains of M . butyricum had esterase zymograms which could not be distinguished from those for M . smegmutis and were found to have been misclassified. The third strain of M . butyricum gave a pattern which did not conform with that for any of the tested strains in the survey. One strain assigned to M.fortuitum had a pattern which was different from those of the other strains assigned to this species and was then shown to exhibit physiological characteristics which differed from those of the other strains. Although intraspecific variation was detected for the strains of M . rhodochrons, the esterase zymograms were more similar to each other than to those of other species, indicating that this species might include “bio-chemical races.” According to a personal communication from Norris and Laidlow to the authors, esterase zymograms for M . tuberculosis could be used to distinguish this species from others in the genus. Lund (1965) used cellular extracts from strains of two varieties of

150


Streptococcus faecalis, S. faecium, and S . durans to prepare esterase zymograms. The patterns of esterase sites detected for S . faecalis exhibited some variation which could not be related to the variety of serotype of the strains. The zymograms for this species, however, could be distinguished from those for both S . faecium and S . durans. One strain of S . faecium gave an esterase zymogram (and protein profile) which was significantly different from that for another strain. An investigation of the history of the “aberrant” strain revealed an error in the correlation of typing systems. This strain was then shown experimentally to belong to a different serotype. Different patterns of esterase activity were found in zymograms prepared from cellular extracts of two strains of Leptospira bijlexa and one strain of L. icterohaemorrhagica, L. canicola, and L. Pomona (Green and Goldberg, 1967). Green, Goldberg, and Blenden (1967) used cellular extracts or culture filtrates of 37 serotypes of Leptospira to obtain zymograms for a relatively large number of enzyme systems. The zymograms for extracellular esterases and for the naphthylamidases provided an acceptable basis to distinguish the pathogenic and saprophytic serotypes. Except for L. canicola and L. autumnalis, the isolants belonging to the same serogroup generally gave consistent zymograms. Finally, the zymograms for extracellular esterases exhibited a close correlation with the serological classification of the tested leptospiras. Although esterase, peroxidase, malate DH, glutamate DH, lactate DH, succinate DH, and alcohol DH zymograms were obtained from extracts of one to five strains of six species of Agrobacterium, Hsu and Chen (1967) did not comment in their preliminary report on the extent of intraspecific or interspecific variation. An extensive survey of strains and species in Serratia, Erwinia, Enterobacter, Klebsiella, Citrobacter, Arizona, Escherichia, Proteus, Pseudomonas, Alcaligenes, Shigella, Salmonella, Vibrio and Pasteurella using cellular extracts to prepare glucose 6-phosphate DH and 6-phosphogluconate D H zymograms was reported by Bowman et al. (1967). Considerable intraspecific variation was observed for both types of zymograms. For example, eight electrophoretic sites of glucose 6-phosphate DH and 11 sites of 6-phosphogluconate DH activity were found for the 14 strains of E . coli used in the survey and 12 of the 14 strains could be characterized by their pattern of sites of activity. Strains of P . pseudotuberculosis were also characterized by their zymograms. In the Salmonellae and Shigellae, identical zymograms were found for S . paratyphi and S . cholerosuis and for S . flexneri and S. sonnei, respectively.


151

C. PROTOZOA Allen (1960, 1961) reported different esterase zymograms for extracts from strains of Tetrahymena pyriformis and used these differences as material for genetic rather than taxonomic purposes. Kates and Goldstein (1964) were not able to distinguish among two strains of Amoeba proteus and one strain of A. discoides by their esterase or phosphatase zymograms but could readily identify Chaos chaos by such zymograms. Sherman (1966) used malate DH zymograms to determine the number of isoenzymes in erythrocytes of the host species before and after infection with the appropriate species of Plasmodium and in extracts from each pathogen. Zymograms for P . lophurae had one site of malate DH activity and those for P . berghei four sites. The erythrocytes from each host were characterized by sites of activity which differed from those of the parasite and the infected erythrocytes yielded sites which matched those for uninfected cells and for the parasite. Njogu and Humpryes (1967)were able to characterize two members of the Brucei subgroup of trypanosomes by their different profiles for the extracted soluble proteins. D. ALGAE

Keck (1961) used acid phosphatase zymograms to investigate the nuclear and cytoplasmic factors determining the species-specificity of protein in Acetabularia mediterranea and Acicularia schenkii by using grafts from one species to another so that the cap of one species was associated with the nucleus of the other species. This investigation could be accomplished because each species gave a single site of acid phosphatase activity in zymograms but the enzymes differed in their electrophoretic mobility. Two reports on the use of zymograms in algae involved a genus of blue-green algae which would have been presented in the section on bacteria. Frederick (1962)detected different patterns of phosphorylase activity in zymograms from extracts of two isolants of the blue-green alga Oscillatoria. A comparison of such zymograms for 0. princeps, Rhodymenia pertusu, and Spirogyra sp. revealed quantitative rather than qualitative differences (Frederick, 1964). IV. G e n e r a l Comments

The growing interest in using proteins and enzymes as they yield profiles or zymograms by zone electrophoresis for taxonomic purposes is evident from the number of reports which have appeared in the

152

EDWARD D. GARBER AND JOHN W. RlPPON

last 2-3 years. The profiles or the zymograms, however, should not be viewed as a means to evade the usual taxonomic criteria in identifying strains of species. Intraspecific variation for protein profiles and zymograms must be established by surveying a reasonable number of strains and species. When invariant sites are detected for all the tested strains of a species, such sites may be considered to be species-specific. When each species of the genus yields a characteristic array of invariant sites, it may be possible to construct an atlas of invariant sites for the genus. Conflicts will undoubtedly arise for the microbial taxonomist. For example, a preliminary survey of strains may indicate a number of invariant sites but a more comprehensive study yields strains lacking these sites. When the usual taxonomic criteria demand that all the strains belong to the same species, it may be necessary to recognize “biochemical races.” For the microbiologist interested in correct specific epithets for his experimental material, such deviant strains should be examined more closely to ensure against errors or misclassifications. When strains of a species indeed yield characteristic profiles or zymograms, the pattern of sites may serve as a useful fingerprint to identify the strains without resorting to a battery of tests. The paucity of data precludes an evaluation of the relative worth of protein profiles and zymograms as taxonomic tools at this time. Extracts usually yield many more sites of stained protein than of enzyme activity. Although apparatus is available to distinguish each site of stained protein, it is more difficult to make meaningful comparisons of protein profiles than of zymograms. Furthermore, comparisons should be made for samples applied to the same gel so that slight differences in electrophoretic mobility can be detected. It may be possible, by disk electrophoresis, to compare extracts or culture filtrates by adding a common protein to each extract or filtrate or by mixing two extracts. The choice of enzyme systems for zymograms is an important factor in detecting enzymes with different electrophoretic mobility. Although esterases and phosphatases appear to be used more frequently than other enzymes, the cost of the components in the reaction mixtures, particularly nitro blue tetrazolium salt, may have been a significant factor in choosing esterases and phosphatases for extensive surveys. The value of proteins and enzymes as taxonomic tools has been established for very few microbial genera. The fungi and bacteria


153

have received most of the attention and the protozoa and algae almost none. The relative simplicity of the techniques required to prepare protein profiles and zymograms should encourage more extensive investigations to determine their worth as adjuncts to the conventional criteria used to detect and to define strains and species. ACKNOWLEDGMENTS

Support for studies to evaluate the taxonomic value of protein profiles and zymograms has been provided by grants from the National Science Foundation (GB-24927, GB6665) and the Dr. Wallace C. and Clara A. Abbott Memorial Fund, University of Chicago, and by Contract Nonr-2121(14) from the Office of Naval Research to the first author.

REFERENCES Ainsworth, G. C., and Sneath, P. H. A., eds. (1963). “Microbial Classification,” 483 pp. Cambridge Univ. Press, London and New York. Allen. S. L. (1960).Genetics45,1051-1070. Allen, S. L. (1961).Ann. N. Y. Acad. Sci.94,753-773. Bowman, J. E., Brubaker, R. R., Frischer, H., and Carson, P. E. (1967).J . Bacteriol. 94, 544-551. Cann, D. C., and Wilcox, M. E. (1965)J.Appl. Bacteriol. 29,165-1 74. Chang, L. O., Srb, A. M., and Steward, F. C. (1962).Nature 193,756-759. Clare, B. G. (1963).Nature 200,803-804. Clare, B. G., and Zentmyer, G. A. (1966).Phytopathology 56,1334-1335. Cowan, S. T. (1962). In “Microbial Classification” (G. C. Ainsworth and P. H. A. Sneath, eds.), pp. 433-455. Cambridge Univ. Press, London and New York. Durbin, R. D. (1966).Nature210,1186-1187. Eilers, F. I., Allen, J., Hill, E. P., and Sussman, A. S. (1964).J . Histochem. Cytochem. 12,448-450. Fowler, R. C., Cable, D. W., Kramer, N. C., and Brown, T. McP. (1963).J.Bacteriol. 86, 1145-1 151. Frederick, J. F. (1962). Phytochem. 1,153-157. Frederick, J. F. (1964). Phyton 21,85-89. Gill, H. S., and Powell, D. (1967).Phytopathology57,812. Gottlieb, D., and Hepdon, P. M. (1966).J.Gen. Microbiol. 44,95-104. Green, S. S., and Goldberg, H. S. (1967).J.Bacteriol. 93,1739-1740. Green, S. S., Goldberg, H. S., and Blenden, D. C. (1967). Appl. Microbiol. 15, 11041115. Hayhome, B., and Garber, E. D. (1967). Unpublished data. Henry. C. E., and Garber, E. D. (1967).Acta Phytopath. 2,89-94. Heslop-Harrison, J. (1962). In “Microbial Classification” (G. C. Ainsworth and P. H. A. Sneath, eds.), pp. 14-36. Cambridge Univ. Press, London and New York. Hsu, J. C., and Chen, P. K. (1967).Bacteriol. Proc. 11. Hubby, J. L., and Lewontin, R. C. (1966).Genetics 54,577-594. Huisingh, D., and Durbin, R. D. (1967).Phytopathology 57,922-923.

154


Hunter, R. L., and Markert, C. L. (1957).Science 125,1294-1295. Kates, J. R.,and Goldstein, L. (1964)J. Protozool. 11,30-35. Keck, R. (1961).Ann. N. Y.Acad. Sci.94,741-752. Lund, B. M. (1965).J. Gen. Microbiol. 40,413-419. McClements, J. K., and Garber, E. D. (1967). Unpublished data. Manwell, C., and Kerst, K. V. (1966). Comp. Biochem. Physiol. 17,741-754. Markert, C. L., and Mdller, F. (1959). Proc. Natl. Acad. Sci. U.S. 45,753-763. Meyer, J. A,, Garber, E. D., and Schaeffer, S. G. (1964). Botan. Gaz. 125,298-300. Nealson, K. H., and Garber, E. D. (1967). Mycologia 59,330-336. Nerenberg, S. T. (1966). “Electrophoresis,” 272 pp. Davis, Philadelphia, Pennsylvania. Njogu, A. R., and Humphryes, K. C. (1967).Nature216,280-282. Nods, J. R.,and Burges, H. D. (1963)J. Znsect Pathol. 5,460-472. Novacky, A., and Macko, V. (1966).Naturwissenschaften 53,281. Racusen, D., and Foote, M. (1966).Can.J. Botany 44,1633-1638. Rottem, S., and Razin, S. (1967).J. Bucteriol. 94,359-364. Shaw, C. R.(1966). Science 149,936-943. Shechter, Y.(1967a). Personal communication. Shechter, Y.(1967b). Personal communication. Shechter, Y., Landau, J. W., Dabrowa, N., and Newcomer, V. D. (1966). Sabouruudia 5,144-149. Sherman, I. W. (1966).J. Protozool. 13,344-349. Stipes, R. J. (1967).Phytopathology57,833. Tsao, M. U. (1962).Science 136,42-43. van Riel, J., van Sande, M., and van Riel, M. (1960).Compt. Rend. 250,3235-3237. Whipple, H. E., ed. (1964). Gel Electrophoresis.Ann. N. Y. Acad. Sci. 121,305-650.

Mycotoxins

ALEX CIEGLER

AND

EIVINDB. LILLEHOJ

Northern Regional Research Laboratory Peoria, lllinois

............................................................... 155 ..................................................................... 156 156 A. Introduction ......................................................... B. Toxin-Producing Fungi .......................................... 158 C. Toxin-Affected Commodities ................................... 158 D. Production ........................................................... 159 E. Biological Activity ....... ................................... 160 ................................... 173 F. Biochemistry ................ G . Analyses ............................................................... 185 H. Safety Procedures .................................................. 192 I. Control and Detoxification ...................................... 192 Alimentary Toxic Aleukia (ATA).................................... 195 Ochratoxin ...................,.............................................. 197 Sporidesmin ............................................................... 199 F-2 Estrogenic Factor (Zearalenone) .............................. 200 201 Pink Rot Dermatitis ..................................................... Slaframine (Slobber Factor) .......................................... 202 203 Yellow Rice Toxins ...................................................... ....................... 205 Stachybotryotoxicosis ..................... ....................... 206 Rubratoxins.................................,. Other Mycotoxins ........................................................ 207 Summary .................................................................... 209 References ..............................................................._..210

I. Introduction 11. Aflatoxin

111.

IV. V. VI. VII. VIII. IX. X. XI. XII. XIII.

1. Introduction The stimulus given to mycotoxin research since the discovery of aflatoxicosis in England in 1960 is difficult to exaggerate. Forgacs in 1962 called mycotoxicoses the neglected diseases. Before 1960, however, intensive research had been carried out on mycotoxins in Russia and in Japan; language barriers probably kept the results from being given the proper attention they deserved in the West. Forgacs and his colleagues were responsible for much of the mycotoxin research in the United States before the aflatoxin outbreak in England. It is still difficult to accurately assess quantitatively the importance of mycotoxins in causing disease; conditions conducive to the production of mycotoxins in the laboratory may not prevail in nature. Mycotoxicoses have been well documented for animals, but their occurrence in humans has been definitely established for only a 155

156

ALEX CIEGLER AND ElVIND B. LILLEHOJ

limited number of toxins: alimentary toxic aleukia (ATA), the psoralens from diseased celery, yellow rice toxins, and stachyobotrytis toxin. The difficulty of research in this area is obvious. However, Forgacs and Carll (1962)and Townsend (1965) have noted certain useful diagnostic features that characterize outbreaks of mycotoxicoses; these were well summarized by Feuell (1966a): (1)The diseases are not transmissible; (2) drug and antibiotic treatment have little or no effect on the disease; (3) in field outbreaks, the trouble is often seasonal; (4) the outbreak is usually associated with a specific foodstuff; and (5)examination of the suspected food or feed reveals signs of fungal activity. The following excellent reviews and proceedings of symposia on various aspects of mycotoxins and mycotoxicoses have been pyblished: Forgacs (1962), Forgacs and Carll (1962), Kraybill and Shimkin (1964), McCalla and Haskins (1964), Wogan (1965, 1966a,b), Symp. Mycotoxins Foodstufs (1965), Borker et al. (1966), Feuell (1966b), Hesseltine et al. (1966), Legator (1966), Brook (1966), Wilson (1966), Hesseltine (1967), Mateles and Wogan (1967a,b), Uritani (1967), and Schoental(l967). II. Aflatoxin

A. INTRODUCTION

The history leading up to the discovery of the aflatoxins, a group of highly toxic secondary metabolites produced by various fungi, has been well documented by Borker et al. (1966). Briefly, aflatoxins were shown to be the causative agents for the high incidence of hepatomas throughout the world in hatchery-raised trout (Halver, 1962, 1965; Halver et al., 1962; Engebrecht et al., 1965) and to be responsible for many deaths among turkey poults, ducklings, chicks, and young pheasants (Sargeant et al., 1961a,b; Austwick and Ayerst, 1963; Spensley, 1963; Blount, 1961; Asplin and Carnaghan, 1961; Carnaghan and Allcroft, 1962; Kohler and Swaboda, 1962; Derzsy et al., 1961). The chemical characteristics of the aflatoxins were soon determined (Asao et al., 1963, 1965; van der Menve et al., 1963; Van Dorp et al., 1963; Cheung and Sim, 1964). Subsequently, Buchi et al. (1966) synthesized aflatoxin B1. Hydroxylated derivatives of aflatoxin were reported by Allcroft and Carnaghan (1963) and by De Iongh et al. (1964) to be in the milk of cows fed toxic peanut meal; these compounds, called aflatoxins M 1 and Mz, were as toxic as the aflatoxins from which they were derived

157

MYCOTOXINS

(Purchase, 1967). Their chemical structures were subsequently determined by Holzapfel et al. (1966) who showed them to be 4-OH-aflatoxin B, and 4-OH-aflatoxin BP.These hydroxylated derivatives were also isolated from the lactating rat (De Iongh et al., 1964), from rat liver (Butler and Clifford, 1965), and from sheep urine (Allcroft et al., 1966; Maxi et al., 1967); all these animals had been fed aflatoxin B1-containingdiets. Two new hydroxy-aflatoxins were described by Dutton and Heathcote (1966).These corresponded to the 2-OH derivatives of aflatoxin Bz and G z and were designated aflatoxins B% and G e , respectively. Aflatoxin Bk appears to be identical to the aflatoxin B1 hemiacetal of Buchi et al. (1966),to the aflatoxin-W of Andrellos (Pohland, Cushmac, and Andrellos, private communication), and to the low R, factor of Ciegler and Peterson (1967). Structural formulas for the various aflatoxins are shown in Fig. 1.

Mi

M2

B2a

G20

H

FIG. 1. Structural formulas of the various aflatoxins.

158

ALEX CIEGLER AND EIVIND B. LILLEHOJ

That other aflatoxin derivatives are readily formed is indicated by the work of Andrellos et al. (1967) who irradiated aflatoxins B 1and G1 with ultraviolet light and secured a variety of new fluorescent products.

B. TOXIN-PRODUCING FUNGI Numerous investigators have shown that Aspergillus flauus and A. parasiticus produce aflatoxin. However, there is some question as to whether or not aflatoxin-producing cultures of A. flauus should be classified as A. parasiticus; this point remains to be resolved. Hodges et al. (1964)reported that Penicillium puberulum, isolated from moldy peanuts, also produced aflatoxin. More recently, Scott et al. (1967) found that A. ostianus, a member of the A. ochraceus group, produced aflatoxin B1 and G 1 . The identity of these toxins was proved by cochromatography with authentic aflatoxins, derivative formation, and ultraviolet and infrared spectra. Kulik and Holaday (1966) isolated a wide variety of species of Aspergilli and Penicillia from corn grains and reported that they produced aflatoxin. These isolates included a number of strains of A. jlavus, A. niger, A. parasiticus, A. Tuber, A. wentii, P. citrinum, P . frequentans, P . puberulum, and P. uariabile. This broad spectrum of toxin producers has not yet been confirmed by other investigators. However, Wilson and his colleagues at Vanderbilt University (personal communication) reexamined many of the cultures other than A . flauus and A. parasiticus that had been reported to be aflatoxin producers in the literature and were unable to detect aflatoxin production by any of them. They concluded that only A. flauus and A. parasiticus were capable of producing aflatoxins. C. TOXIN-AFFECTED COMMODITIES It has been demonstrated that aflatoxin can occur in some moldinfested peanuts and cottonseed. In addition, this toxin has been found in other foods and feeds, including lower grades of corn not used for food. Loosmore et al. (1964)reported it in coconut. It should be stressed, however, that moldiness per se is not always accompanied by the production of mycotoxins. Aflatoxin M has been found in the milk of animals eating aflatoxincontaminated feed (Allcroft and Carnaghan, 1962; De Iongh et al., 1964). Japanese investigators detected aflatoxin in the dried fermented fish called katsubushi. Actual natural occurrence of aflatoxin in other foods is rare. Nevertheless, the ease with which A. flauus can be experimentally grown on a variety of agricultural products (cassava, Nartey, 1966; rice, wheat, corn, oats, Hesseltine et al., 1966; timothy, sweet clover, and

MYCOTOXINS

159

oat straw, Hesseltine et al., 1967) with good production of toxin suggests that, under the proper conditions, aflatoxin contamination could be widespread. However, such conditions are not likely to occur in nature or during normal handling of a harvested crop. Soybeans, red clover, and alfalfa support growth of A. flauus, but toxin production is poor or absent (Hesseltine et al., 1967). Lie and Marth (1967) found that 3-month-old Cheddar cheese would support growth and aflatoxin production by A. flauus and A. pamsiticus. However, there are no known reports of natural occurrence of aflatoxin in cheese.

D. PRODUCTION The production of aflatoxin for experimental studies has been investigated on a variety of agricultural substrates and on conventional microbiological media. The choice of substrate has usually been governed by the use to which the final product is to be put. Thus, for toxicological studies in which the toxins do not have to be extracted and purified, agricultural substrates have proved advantageous. Aflatoxin has been produced in good yield on peanuts (Codner et al., 1963; Ashworth et al., 1965); crushed wheat (Chang et al., 1963; Kraybill and Shimkin, 1964; Hesseltine et al., 1966; Stubblefield et al., 1967; Robertson et al., 1967); corn, rice, sorghum (Hesseltine et al., 1966; Shotwell et al., 1966); oats, millet, egg solids and skim milk powder (Kraybill and Shimkin, 1964). Curiously, soybeans and ather legumes have supported toxin production poorly even though A. flauus grows well on these substrates; various laboratories are currently investigating this aspect. Wildman et al. (1967) reported good toxin production on a variety of fruit juices, fruits, cheese, bread, cocoa beans, sterilized beef, and beef infusion; aflatoxin production did not occur on nonsterile beef. Rice appears to be the best substrate of the various commodities investigated and the production method, developed by Hesseltine and his co-investigators at the Northern Regional Research Laboratory, may have application in other fermentations as well. Slightly moistened sterile rice, after inoculation with a spore suspension of A. flauus, was incubated in Erlenmeyer or Fernbach flasks on a rotary shaker at 28C at 200 rpm. Yields of up to 1.5 mg. toxin/gm. substrate resulted after 5-6 days incubation. This heavily contaminated rice has proved useful in toxicology studies, the rice being blended with feedstuffs to give appropriate toxin concentrations. Various factors affecting growth and aflatoxin production by A. flauus have been investigated in submerged culture. Mateles and

160


Adye (1965) readily obtained aflatoxin yields of 60-90 mg./liter in shaken-flask culture using a medium containing sucrose, casamino acids, zinc sulfate, and various salts; only 20-50 mg./liter were produced on scale-up to 5-liter fermentors. Codner et al. (1963) using a modified Czapek-Dox medium to which corn steep liquor was added obtained 100-200 mg. toxin/liter in 100 ml. of medium, but no toxin resulted on scale-up to 3- and 20-liter fermentors. Aflatoxin production by A. parasiticus in Czapek's broth fortified with corn steep liquor increased proportionately as the concentration of corn steep liquor was increased from 0.5 to 8.0% (v./v.) until maximum growth was reached (Schroeder, 1966). Ciegler et al. (196613) made 200-300 mg./liter of aflatoxin in 20-liter fermentors under proper conditions of inoculum (well-dispersed growth) and aeration (0.5 v./v./min. of air, 300 r.p.m., 30 p.s.i. back pressure, baffles). Peak yields usually resulted in 72 hours, after which the aflatoxin concentration declined rapidly. The need of high aeration rates in fermentors to obtain good toxin yields was confirmed by Hayes et al. (1966). Stationary-flask studies in a semisynthetic medium (yield, 2-63 mg. toxin/100 ml.) were reported by Davis et al. (1966a). Several investigators have shown that the ratio to one another of the four major aflatoxins produced is temperature dependent; aflatoxin B tends to predominate at higher temperatures, but the proportion of G increases as incubation temperature is lowered (Ciegler et al., 1966b; Schindler et al., 1967; Schroeder and Hein, 1967; Sorenson et al., 1968). Optimum temperature for aflatoxin production varies considerably with the strain of mold used but is generally between 25"-30°C; at higher temperatures there is a rapid decline in toxin production. Since toxin production occurs down to at least lO"C, cold storage would not prevent toxin formation on agricultural commodities. Numerous investigators have shown that there is considerable variation in the capability of toxin production by various strains of A. flauus and A. parasiticus, some strains producing no detectable toxin. Genetic studies on aflatoxin synthesis have not been published and would be highly desirable.

E. BIOLOGICAL ACTIVITY

1. Efects on Animals Toxicity studies demonstrated that many animal species are susceptible to aflatoxins with L D ~ values o for aflatoxin B I ranging from 0.5

MYCOTOXINS

161

to 10.0 mg./kg. of body weight (Wogan, 1966a). Older animals resist aflatoxin more than young animals (Allcroft, 1965). Carnaghan et al. (1963) have investigated the relative toxicties of the four aflatoxins and found that the oral 7-day LDso in the day-old duckling, based on 50-gm. body weight, was: B1, 18.2 pg.; Bz, 84.8 pg.; G I , 39.2 pg.; and Gz, 172.5 pg. Several reports since then have verified the fact that aflatoxin B1is the most toxic of the four toxins (Wogan, 1966a). Examinations of animals following aflatoxin treatment have shown a general pattern of liver damage; histopathological studies demonstrate that the most consistent liver lesion was proliferation of the bile duct cells. Other pathological symptoms associated with aflatoxin poisoning have been summarized by Wogan (1966a). Among the animals most sensitive to aflatoxin B1 is the day-old duckling with an LDso of 0.5 mg./kg. Histopathological examination of the livers from ducklings fed B l for 5 days, followed by sacrifice on the seventh day, has been used as a biological assay for the presence of the toxin since a total dose of 2.5 pg. can be reproducibly determined. In addition to acute toxicity, chronic exposure for extended periods of time to aflatoxin has resulted in the development of hepatomas. Early studies with contaminated peanut meals fed to rats at sublethal levels demonstrated that malignant liver tumors developed within a 6-month period (Lancaster et al., 1961). Since that time, the carcinogenic action of aflatoxin has been verified by several research groups (Barnes and Butler, 1964; Butler, 1965; Wogan, 1966a; Newberne, 1965), and it has been established that continued feeding is not required for initiation of the liver tumors. In feeding trials with rats, 1.8 p.p.m. B1 produced 90% tumor incidence following 370 days of ingestion (Newberne, 1965). Furthermore, Barnes and Butler (1964) found that feeding 2.5 mg. of aflatoxin to rats over an 89-day period resulted in hepatomas up to 1 year later. The consensus of opinion is that aflatoxin is the most active hepatocarcingen known (Butler, 1965; Kraybill and Shimkin, 1964). In addition to the long-term effects of aflatoxin on the liver, Butler and Barnes (1966) reported that rats fed diets containing 3-4 p.p.m. aflatoxin developed carcinoma of the glandular stomach. R. C. Shank and G. N. Wogan (1966) examined the effects of feeding sublethal doses of aflatoxin B1 on the liver of rats and ducklings. Feeding the duckling 60 pg./kg./day for 5 successive days markedly reduced liver glycogen, enhanced lipids, had no effect on total protein content, and did not significantly alter the capacity of the liver to utilize leucine. Feeding the rat 600 pg./kg./day for 5 successive

162


days did not affect glycogen, lipid, or protein content; however, protein synthesis capacity measured by leucine uptake was enhanced. Leucine incorporated into the rat liver, following toxin treatment (LDzo),was biphasic and consisted of an initial inhibition in uptake followed by an enhanced rate of protein synthesis. They postulated that the inhibition might be due to an initial suppression of nuclear RNA synthesis. A stimulatory activity of protein synthesis by liver cells has also been observed b y workers investigating hepatocarcinogens other than aflatoxin (Hawtrey and Schirren, 1962; Arrhenius and Hultin, 1962). R. H. Smith (1963) reported that aflatoxin B l inhibited the in vitro incorporation of leucine into protein by rat liver slices. At levels of 0.5 pmole/ml. of B1,protein synthesis was entirely blocked. Further work b y Smith (1965) demonstrated that in vitro addition of B1 inhibited the amino acid-activating enzymes from rat liver slices and Escherichia coli. Various sulfhydryl compounds reversed the aflatoxin inhibition. Bassir and Osiyemi (1967) determined the bilary excretion pattern of labeled metabolites following intraperitoneal injection of 14Clabeled aflatoxin. Following an initial lag period of approximately 60 minutes, 30% of the labeled material was excreted in the bile 6 hours after injection. Ten percent of the bile-excreted label was chloroformsoluble and was a mixture of aflatoxin B1and M ,whereas 20% of the label was a conjugate of taurocholate and a material similar in nature to aflatoxin B l , which might have been degraded B1. Urine analysis showed that 26% of the labeled aflatoxin was excreted in the urine as a glucuronide conjugate. In earlier work, Falk et al. (1965) reported that aflatoxin and a series of fluorescent compounds were excreted in the bile following toxin treatment. G . N. Wogan’s group has examined the metabolic fate of labeled aflatoxin B 1 in the rat (Wogan, 1966a). They also observed about 25% excretion of the label in urine. The only site of major retention of the label was the liver, which contained 6-9% of the total counts 24 hours after dosage. In addition, they made the interesting observation that methoxy-labeled toxin was demethylated in the rat, yielding COZ whereas ring-labeled toxin produced no 14C C o t (Shank and Wogan, 1965).Their work suggested that the ring structure of aflatoxin was metabolically stable. An intriguing study has been carried out by Wogan’s group on the effects of aflatoxin B1 on the induction of rat liver tryptophan pyrrolase by hydrocortisone (Wogan, 1966a). At levels of hydrocortisone initiating a 4-fold increase in the level of tryptophan pyrrolase, a 1

MYCOTOXINS

163

mg./kg. dose of aflatoxin B1 blocked the enhanced enzyme activity. Induction of the catalyst by tryptophan was not modified by B 1. Since it is felt that hydrocortisone stimulates enzyme synthesis and tryptophan functions as an enzyme stabilizer, apparently hormonal stimulation of protein synthesis is particularly sensitive to the toxin. Butler (1964) has investigated cellular changes that occur in the liver after poisoning by aflatoxin. Administration of 15 pg. of toxin to ducklings induced extensive bilary proliferation and fatty degeneration of the peripheral parenchymal cells. Three days after dosage the parenchymal cells appeared to undergo a repair process. The toxic action suggested to him that the aflatoxin might be functioning as an alkylating agent. When rats were fed 0.1 mg. of B1, l-hour examination of the periportal cells showed that the rough endoplasmic reticulum was dilated and that some of the ribosomes appeared to be dislocated (Butler, 1966). In addition, formation of nucleolar caps was observed in aflatoxin-treated cells. The formation of nucleolar caps has also been observed by Svoboda et al. (1966, 1967) in liver cells of rats and monkeys dosed with aflatoxin B 1 , Twenty-four to 72 hours after an oral dose of 0.45 mg./kg. of aflatoxin B1 to rats, liver cells decreased in RNA and protein in both the nucleus and cytoplasm. The RNA-DNA ratio decreased 40% in the liver homogenate while the protein-DNA ratio decreased 8% after administration of the toxin. Sporn and Dingman (1966) also observed a decrease in nuclear RNA of rat liver cells treated with aflatoxin B Aflatoxin affects mitochondrial activity of rat liver cells as demonstrated by a decrease in the P:O ratio following treatment (Svoboda et aZ., 1966, 1967). Both phosphorylation and oxygen consumption were inhibited by the toxin. These observations are quite similar to those reported by Brown (1965) and Brown and Abrams (1965) on the toxic effects of aflatoxin B1 on mitochondrial process in liver cells of ducklings and chickens. Clifford and Rees (1967a) stated that mitochondria from rats dosed with 7 mg. Bl/kg. body weight up to 24 hours after treatment changed neither their respiratory capacity nor P:O ratios relative to controls. There was a 26-35% reduction in the ATP concentration of the liver but no change in ATPase activity following poisoning. Further examination of the effect of B1 on the liver cell showed that the toxin did not affect the ion transport mechanism. There is a difference in the statements made by Clifford and Rees (1967a) and those made by others (Svoboda et al., 1966,1967; Brown, 1965; Brown and Abrams, 1965) on the effects of aflatoxin on mito-

164

ALEX CIECLER AND EIVIND 8. LILLEHOJ

chondrial processes. It is hoped that future work will resolve this apparent discrepancy. Svoboda et al. (1966)reported that the hepatic lesions from monkeys treated with aflatoxin resembled human liver during acute hepatitis. A similar observation was made by Newberne et aZ. (1966b) on the liver of dogs fed subacute levels of aflatoxin B1 for extended periods of time. In this study, the experimentally induced disease resembled hepatitis X. Allcroft and Camaghan (1962) reported that extracts of milk from cows and lactating rats fed aflatoxin in the diet initiate liver lesions in ducklings similar to those attributed to aflatoxin. A later study (De Iongh et al., 1964) disclosed that the toxic factor in milk was a blue fluorescent material with an R f lower than that of B,. Since sheep appear remarkably insensitive to the action of toxin, which has been related to rapid excretion of aflatoxin (Allcroft, 1965), Allcroft et al. (1966) investigated and found the milk toxin in sheep urine dosed with aflatoxin. They named the milk toxin aflatoxin M. Characterization of the compound was carried out, and it was concluded that aflatoxin M is hydroxy aflatoxin B (Holzapfel et al., 1966; Masri et al., 1967) (Fig. 1). The question has been raised whether aflatoxin M is actually present in the toxic feeds administered to the animals with preferential excretion or if there is a metabolic conversion of aflatoxin B to M. (De Iongh et al., 1964; Masri et aZ., 1967). .Purchase (1967) has determined the acute toxicity of aflatoxin M 1 in ducklings and found that it is approximately as toxic as aflatoxin B1 with no observed effect on the kidneys. Determination of the levels of aflatoxin M 1 in milk (Masri et al., 1967) demonstrated that 2-3% of the ingested dose of B1 could be recovered as M 1, and only 0.3% of the original B1 was found in the milk. Measurements made of the excretion of aflatoxin from a sheep dosed with aflatoxin B1 indicate that aflatoxin M I is excreted in relatively large quantities in the milk, urine, and feces (Nabney et al., 1967). Feeding chickens a diet containing aflatoxin at 1.5 p.p.m. induced a rise in the fat content of the liver and a marked reduction in liver vitamin A (Allcroft, 1965). Further work by Brown (1965) and Brown and Abrams (1965) on ducklings and chickens fed aflatoxin B, investigated the toxic effect on some liver enzymes. These studies show that the activity of certain mitochondria1 dehydrogenases and of electron transfer catalysts is decreased during aflatoxin feeding. They attribute the reduced protein synthesis in treated bird livers to a toxic

MYCOTOXINS

165

event in the mitochondria which decreases the rate of ATP synthesis. Brown (1965)and Brown and Abrams (1965)also observed that liver damage produces an increase in the plasma lactic dehydrogenase, aldolase, glutamic-oxalacetic, or glutamic-pyruvic transaminase activity. The increased action of these enzymes in the serum was correlated with liver lesions; it was suggested that this enhancement be used as measure of liver damage. A similar increase in serum alkaline phosphatase and glutamic oxalacetic transaminase has been related to liver damage in calves and swine (Allcroft, 1965; Harding et al., 1963). Determination of enzyme activity in the serum of rats following ingestion of feed contaminated with A. jZauus showed an increase in lactic dehydrogenase, leucine-aminopeptidase activity and a decrease in cholinesterase (Bassir, 1964). Following administration of 7 mg. BJkg. to rats, Clifford and Rees (1967a) found a marked decrease in hepatic enzymes and a concomitant increase in the activities of the catalysts in the serum. Isocitrate dehydrogenase, malate dehydrogenase, and glutamate hydrogenase activities decreased markedly in the liver during the second 24-hour period after B t treatment while increasing in the serum. Furthermore, 96 hours after poisoning there was a significant increase in serum alkaline phosphatase and serum bilirubin. Theron (1965) has correlated the aflatoxin B1 damage to parenchymal liver cells of ducklings with histochemical studies of various enzymes. He found a progressive decrease following administration of toxin in activity of succinic dehydrogenase, alkaline phosphatase, adenosine triphosphatase, inosine diphosphatase, and thiamine pyrophosphatase, but an increase in activity of acid phosphatase. He suggested that the red blood cells are responsible for the transport of the toxic material and that the cytotoxic effect might be due to direct action of the toxin on liver cell membrane. In a study carried out by Newberne et al. (1966a), the effect of dietary modifications on aflatoxin response in ducklings was determined. Basal diets, with or without aflatoxin, were supplemented with 1.0% arginine, 0.8% lysine, and 4.0% methionine added separately and a combination of 1.0% arginine and 0.8% lysine. Results of this experiment are presented in Table I. Their work established that addition of lysine and arginine sensitized the ducklings to the effects of the toxin. Addition of the amino acids alone or in the absence of aflatoxin had no appreciable effect on liver morphology. The investigators postulated that the increased transamination and deamination activity required by the liver when lysine and arginine were

166


added could be responsible for the added sensitivity to the toxin. However, addition of 4% methionine, which is about the same amino nitrogen as the combination of 0.8% lysine and 1.0% arginine, did not facilitate the added sensitivity to toxin produced by the basic amino TABLE I EFFECTSOF SUPPLEMENTAL AMINO ACIDS O N DUCKLINGS FED AFLATOXIN-CONTAMINATED PEANUTMEAL'

Dietary treatment Basal diet D Basal diet B Basal diet B 4% methionine Basal diet B 1.0% arginine Basal diet B + 0.8% lysine Basal diet B 1.0% arginine 0.8% lysine

+

+ + +

Avg. Dietary 2-week Bile duct aflatoxin wt 2-Week No. of ducklingsb (p.p.m.) (gm.) mortality proliferationc 15 10 15 15 15

0.0 1.5 1.5 1.5 1.5

263 143 112 118 121

20

1.5

-d

O/l5 6/10 3/15 1/15 3/15

20/20

0.0 34.0 24.0 32.0 34.0 20.0

"Table from Newberne et al. (1966a). *Numbers represent 2 or 3 trials in each dietary treatment, results of which are combined; the total is presented in the table. "Graded 0 to 4 according to severity ranging from none (0) to severe (4), multiplied by 10 and averaged. dLosses began at 5 days with 15/20 dead by day 7 and 20/20 dead by dav 12.

acids. Their work seems to have broad implications, particularly as it relates to enhanced sensitivity of liver to aflatoxin in pyridoxinestarved animals (Foy et al., 1966) since pyridoxine is a potential coenzyme in transaminase enzymes. Furthermore, the increase in serum transaminases following aflatoxin poisoning may relate to release of the catalyst from the liver cells which interferes with liver amino acid metabolism and accelerates the toxic action. Liver lesions similar in nature to those produced by aflatoxin in the duckling have also been observed in rhesus monkeys (Tulpule et al., 1964) dosed with pure aflatoxin (60% B1, 40% G I ) . Feeding 0.5-1.0 mg./kg. body weight/day produced typical histopathological changes in liver generally associated with aflatoxin damage. When a total dose of 10-15 mg./kg. body weight was administered, all the test animals died within a 28-day period. In further studies on the aflatoxin effects on rhesus monkeys (Madhaven et al., 1965b), some particularly striking observations were made on animals treated with

MYCOTOXINS

167

1.0 mg./kg. body weightlday. The fat content of the liver was 34% by wet weight, and the organ was greatly enlarged. The kidneys were also enlarged, yellow, and showed an accumulation of fat. There was a marked elevation of the serum glutamic oxalacetic transaminase and of serum bilirubin and a fall in serum albumin. Madhaven et al. (1965a) have extended their studies on the effects of aflatoxin in monkeys by feeding the animals a diet fortified with 4% and 20% protein. All the test animals given 100 pg. aflatoxidday on the low protein diet died within 30 days with characteristic liver lesions, whereas the animals fed the upper level of protein plus aflatoxin did not develop lesions of the liver after 35 days of feeding. An intriguing corollary has been suggested between the hepatic injury observed in pyridoxine-deficient baboons and aflatoxin effects (Foy et al., 1966). Pyridoxine-deficient animals demonstrated severe disturbances in content of lipid, glycogen, RNA, and DNA of liver parenchymal cells accompanied by bile duct hyperplasia which were quite similar to the effects produced by aflatoxin. The investigators speculate on possible metabolic links between pyridoxine-starved or aflatoxin-treated livers to liver carcinoma in Africans. They suggest that the high incidence of liver cirrhosis and carcinoma in Africans may be due to: (1)diets either deficient in pyridoxine or containing pyridoxine antagonist and (2) diets sporadically may contain aflatoxin from mold-contaminated porridges or brews. The action of aflatoxin and pyridoxine starvation may function synergistically in hepatic injury. Since reduction in pyridoxine probably impairs the capacity of the cells to carry out transaminase reactions, loading transaminase requirements in the liver cells may produce a subsequent sensitization to the toxic effects of aflatoxin (Newberne et al., 1966a). In addition to the standard cellular changes that occur in the liver as a result of aflatoxin poisoning, Madhaven and Rao (1966) have observed that day-old ducklings fed 10 to 40 pg./day for 2-5 days developed hepatic infarction. An intense congestion of the vessels of the liver routinely observed following toxic action most likely is responsible for the disturbance. However, liver infarctions were not seen in monkeys, guinea pigs, or rats dosed with aflatoxin. Although reports of kidney damage resulting from aflatoxin are not extensive, Blount (1961) saw kidney lesions in turkeys fed aflatoxin while Newberne et al. (1964) observed damage to the kidney of ducklings fed aflatoxin. Madhaven and Rao (1967) reported that young guinea pigs treated with 250 pg. and 500 pg./day of aflatoxin died within 7 days, demonstrating typical liver damage; 7 of the 12 animals

168

ALEC CIEGLER AND EIVIND B . . LILLEHOJ

had severe renal lesions characterized by a reflux of the proximal tubular epithelium into Bowman’s spaces.

2. Effects on Tissue Culture Early work on the effects of aflatoxin on cells in tissue culture carried out by Juhasz and Greczi (1964) demonstrated that low concentrations of aflatoxin extracts from moldy peanut meal are toxic to calf kidney cells. Utilizing Chang liver cells and HeLa cells, Gabliks et al. (1965) found that 1-5 pg./ml. of aflatoxin B1 caused significant cellular destruction. The toxic effect was characterized by a decrease in cell division and an increase in protein, RNA, and DNA per cell; the increase suggested cellular enlargement. However, the total protein, RNA, and DNA per culture decreased with increasing levels of B1.Legator and Withrow (1964) have investigated the effects of aflatoxin on the cells from human lung grown in tissue culture. Five parts per million of aflatoxin completely blocked growth of the cells and exposure to 0.1 p.p.m. reduced mitosis 50% in generating giant cells. A concentration of 0.1 p.p.m. B1inhibited DNA synthesis 45%, whereas 1.0 p.p.m. reduced DNA production 80%.Aflatoxin was also tested as an inducing agent in development of bacteriophage in lysogenic bacteria (Legator, 1966). At 0.06 pg./ml. aflatoxin, an increase in lytic action was observed, and higher toxin concentrations produced increased level of induction. These observations suggested to the investigators that the earliest effect of aflatoxin is inhibition of DNA synthesis through the action of aflatoxin as an alkylating agent. Studying the effects of aflatoxin on thymidine kinase in tissue culture of the lung cells, Childs and Legator (1966) found that the activity of the enzyme increased two to three times in extracts of cells grown in 1 p.p.m. aflatoxin. Studies on the effects of aflatoxin on tissue cultures of human liver by Zuckerman et al. (1967a,b) reported that following incubation of cells in 10 p.p.m. aflatoxin B I , there was a complete loss of RNA from the cytoplasm and an apparent loss of chromatin from the nucleus. Addition of low concentrations of aflatoxin B1 to cell cultures resulted in an inhibition of both nuclear DNA and RNA synthesis. Acute toxicity studies of the human liver cells in culture demonstrated that the LDso for aflatoxin B I was 1.0 p.p.m., 5 p.p.m. for G1,and 16 p.p.m. for Gz. 3. Teratogenic Effects Aflatoxin B1 has been shown to be a strong teratogenic substance in hamsters (DiPaolo et al., 1967).A 4 mg./kg. dose administered on

MYCOTOXINS

169

the eighth day of pregnancy was the most effective level. The teratogenic influence was diminished by mixing the B1 with DNA before administration (Elis and DiPaolo, 1967).The livers of both the mother and offspring demonstrated typical damage associated with aflatoxin poisoning. When hamsters were sacrificed on day nine of pregnancy, following administration of toxin on day eight, 29.4%of the fetuses were malformed, and 17.6%were dead (DiPaolo et al., 1967). If the teratogenic dose was given subsequent to organogenesis, no malformed fetuses occurred. Preliminary experiments with mice showed that administering 8-12 mg./kg. of B1 on day eight of pregnancy resulted in death or resorption of 90% of the fetuses, and there was no observable difference in the livers of the offspring. Dosing rats with 25% of the LDsolevel of aflatoxin B1 early in pregnancy had no effect on the fetus or placenta (Butler and Wigglesworth, 1966).These tests suggest that aflatoxin is species-specific in its action as a teratogenic agent. 4. Effects on Plants Some interesting observations have been made on the effect of aflatoxin in higher plants. Schoental and White (1965)found that 100 p.p.m. of aflatoxin completely inhibited the germination of cress seed and 10 p.p.m. of the toxin induced a chlorophyll deficiency in the seedlings. Several other fungal metabolites were also tested for their effectiveness as inhibitors of germination in cress seeds and in inducing albinism. Coumarin inhibited the germination of the seeds at 25 p.p.m. but did not demonstrate so striking an effect on chlorophyll inhibition as the aflatoxins. This comparison seems particularly pertinent since aflatoxin contains a coumarin nucleus and coumarin has been implicated in other plant growth regulatory processes (Mayer and Poljakoff-Mayber, 1961). It has been shown that the inhibitory effect of coumarin on the germination of kale seeds can be reversed by kinetin, whereas the gibberellic acid system seems to be only indirectly involved (Knypl, 1967). Furthermore, coumarin inhibition of root development has been related to a striking pertubation in the nucleic acid content of the affected cells (De Greef, 1964). Since the current trend in plant auxin studies postulates that the major site of auxin action involves nucleic acid metabolism, particularly control of RNA synthesis (Van Overbeek, 1966; Armstrong, 1966), the inhibitory effect of aflatoxin and coumarin on germination suggests that they are functioning as antiauxins by inhibiting RNA synthesis. Since aflatoxin is an inhibitor of chlorophyll synthesis, the mechanism of action of bleaching compounds might be instructive in

170

ALEX CIEGLER AND EIVIND B. LILLEHOl

determining the toxic mode of action. One of the most widely studied bleaching compounds is streptomycin. Germination of barley seed in the presence of this antibiotic produces seedlings that are either partially or completely lacking in chlorophyll (De Deken-Grenson, 1955). Streptomycin appears to interfere with the messenger RNA by binding to the phosphate groups of ribosomal RNA (Cox e t al., 1964; Zimmer et al., 1967).The bleaching effect of streptomycin on Euglena gracilis has been extensively studied (Provasoli et d.,1948). The induction of permanent albinism in Euglena by streptomycin has been attributed to the deletion from the cells of chloroplast DNA (Edelman et al., 1965). Kirk (1964) has presented evidence that the chloroplast contains a unique DNA-RNA and protein-synthesizing system which is more sensitive to actinomycin than is the nuclear mechanism. The divergence in sensitivity between the chloroplast and other cellular components of Euglena has also been observed in the action of degreening by various antibiotics blocking protein synthesis and the DNA inhibitor hydroxyurea (Linnane and Stewart, 1967; Buetow and Mego, 1967). Another group of compounds, the nitrofurans, are of particular interest as bleaching agents because of their structural similarity to aflatoxin, which also contains furan groups. Twelve derivatives of 5-nitrofuran have demonstrated a bleaching activity against E . gracilis (McCalla, 1965), while studies with other furans demonstrated that the nitro group in the 5-position was required for degreening activity (Ebringer et al., 1967). McCalla (1965) postulates that the action of nitrofurans in bleaching of Euglena is caused by damage to the chloroplast DNA or to the DNA-synthesizing system. The degreening studies suggest that aflatoxin is acting as a bleaching agent by preferentially interfering with metabolic processes of the chloroplast. The inhibition of chlorophyll synthesis is probably due to an initial binding of the toxin to a nucleic acid constituent that causes a malfunctioning of the transcription-translation process. Black and Altschul (1965) reported that gibberellic acid-induced synthesis of lipase and a-amylase in germinating cottonseed and barley was inhibited by a mixture of aflatoxin B and G (45 pg./ml.). The hormonal effect of gibberellic acid system in inducing enzyme synthesis was also inhibited by actinomycin D, which suggested that the aflatoxin was inhibiting the DNA-dependent RNA synthesis. However, further work on the effects of aflatoxin on the gibberellic acid induction of lipase in cottonseed (Jones et al., 1967) demonstrated that low concentrations of aflatoxin actually stimulated lipase synthesis, whereas higher concentrations inhibited the gibberellic

MYCOTOXINS

171

acid-induced production. These effects suggested that higher concentrations of aflatoxin were interfering with the messenger. A similar type of stimulatory-inhibitory response has been reported for the action of coumarin (Verbeek and Dumitru, 1964). The study demonstrated that 0.68 M coumarin inhibited the formation of a-amylase in germinating barley, whereas at lower concentrations the enzyme production was stimulated. The coumarin inhibition could be reversed by gibberellic acid. Thus, it appears again that aflatoxin and coumarin are functioning biologically in an analogous manner. An interesting parallel exists between the inhibition by aflatoxin of the hormonal stimulation of enzyme production in cottonseed and the induction of tryptophan pyrrolase by hydrocortisone in rat liver (Wogan, 1966a). The similarities in the inhibitory responses suggest that aflatoxin is operating through a common mechanism of blocking messenger RNA synthesis in plant and animal systems. However, more evidence is required before this supposition can be verified.

5. Efects on Znsects Examination of the effects of aflatoxin on the yellow fever mosquito (Aedes aegypti), house fly (Musca domestica), and fruit fly (Drosophila melanogaster ) demonstrated that the mycotoxin produced a response in the insects (Matsumura and Knight, 1967). Feeding 0.005-0.03% aflatoxins in the diet of adult insects reduced both the number of eggs and the percentage of eggs that hatched. In another study, larvae of Heliothis virescens exhibited a significant sensitivity to the toxic action of aflatoxin (Gudauskas et al., 1967).

6. Efects on Microorganisms The effect of aflatoxin on microorganisms has been studied because it offers the research worker an opportunity to examine the toxic mode of action at the cellular level without interference from the complexities of multicellular control mechanisms. Furthermore, the possibility exists that aflatoxin can be microbially degraded, which might provide a practical tool for decontamination of mold-infected foods and feedstuffs. Burmeister and Hesseltine (1966) reported that out of 329 microorganisms tested, 12 species of Bacillus, a clostridium, and a streptomycete were inhibited by aflatoxin. Another study showed that aflatoxin effectively inhibited growth of several species of Streptomyces and Nocardia (Arai et al., 1967). In addition, Nocardia asteroides converted a portion of the aflatoxin B1 in liquid culture to undefined products. Growth of several species of Aspergillus and Penicillium was in-

172

ALEX CIECLER AND EIVIND B. LILLEHOJ

hibited by 20 pg./ml. aflatoxin B1 (Lillehoj et al., 1967~).It seemed particularly noteworthy that growth of A flauus was partially inhibited by the toxin since this organism produces the aflatoxin. Knypl(l963) reported that coumarin inhibited the growth of A. niger and P. glaucum, which he attributed to coumarin interference with the energetic pathways of the organisms. Ciegler et al. (1966a) reported that cells of Flauobacterium aurantiacum effectively removed aflatoxin from a liquid medium. Since, l O I 3 cells were required for removal of 600-700 pg. of toxin, the result suggested that the cells were not degrading the toxin but simply taking it up. Cells of F. aurantiacum that had been incubated with aflatoxin B1 and had removed the toxin from solution were fed to ducklings with no harmful effect. Further studies on the effect of aflatoxin B1 on F. aurantiacum demonstrated that growth of the cells was inhibited by 10 pg./ml. of B1 (Lillehoj et al., 1967b). Incubation of 1.0 X 10" resting cellslml. with 7.0pg.lml. B1 completely removed the toxin. Considering the amount of toxin removed per cell suggested that aflatoxin Bl was being taken up by the cells in antibiotic quantities. Autoclaved cells and cell wall preparations of F. aurantiacum removed toxin from a liquid medium, but the B1could be recovered by washing with water while the toxin taken up by intact cells could not be extracted into the aqueous phase. These observations suggested that aflatoxin was initially loosely bound to the cell wall surface followed by a tighter binding of the toxin which required the architecture of the intact cell. Since ruptured cell preparations did not modify the toxin, it was concluded that no enzymic systems were present in the cells that were capable of attacking the toxin directly. Cells grown in the presence of subinhibitory levels of aflatoxin produced aberrant forms. Since the aberrant forms were similar to those observed by other workers as a result of cell wall inhibition, the tentative postulate was offered that aflatoxin interfered with cell wall synthesis. This question was later resolved when it was found that growth of an L-form of an aflatoxin-sensitive strain of Bacillus subtillis was inhibited by the toxin (Burmeister and Hesseltine, 1968). Further studies on the effect of aflatoxin on F . aurantiacum demonstrated that aflatoxin G1 was also taken up by the cells (Lillehoj et al., 1967a). Growth of the organism was inhibited 90% by 37.5 pg./ml. G I after 48 hours, whereas 15 pg./ml. of aflatoxin B1 was required for the same degree of growth inhibition. A similar diminished toxicity of G I relative to B1 occurred in the induction of aberrant morphological forms of F. aurantiacum by aflatoxin G Teunisson and Robert-

MYCOTOXINS

173

son (1967)have recorded an abnormal form of Tetruhymenu pyriformis when the cells were grown in the presence of G 1 . Wragg et al. (1967)reported that 1-5 pg./ml. of aflatoxin B1 inhibited growth of Escherichia coli. The ratio of DNA to protein was reduced by 5 pg./ml. B1 about the same extent as when the cells were grown in 1 pg./ml. mitomycin C. The incorporating capacity of DNA polymerase from E . coli cells grown in aflatoxin B1 or mitomycin C was significantly reduced. They suggested that the inhibition of DNA synthesis in aflatoxin- and mitomycin C-treated cells was related to reduction in DNA polymerase activity. Cells of E. coli grown in the presence of aflatoxin B1 produced elongated filaments similar to those reported for alkylating agents, including mitomycin C (Curry and Greenberg, 1962; Kilgore and Greenberg, 1961).However, azaserine, nitrofurazone, nitrosoguanidine, and proflavine have been classified as radiomimetic compounds since they share with bifunctional alkylating agents and ultraviolet light the capacity of mutagenicity and filament induction, and since they are not inhibitory toward radiation-resistant mutants of E . coZi (Terawaski and Greenberg, 1965). Azaserine and nitrosoguanidine are monofunctional alkylating agents, whereas proflavine and nitrofurazone are not alkylating agents. Therefore, probably neither alkylation nor cross-linking of the DNA is required for the induction of filamentous forms, although the evidence suggests that the radiomimetic agents share a similar mechanism of action probably associated with the DNA. Although all radiomimetic agents inactivate the transforming systems of €3. subtilis when added in viuo, they do not selectively block the synthesis of DNA which has been reported for the bifunctional alkylating agents (Harold and Ziporin, 1958). Terawaki and Greenberg (1965) examined the inhibitory effects of the radiomimetic agents that are not bifunctional alkylating agents on protein synthesis (5070%), RNA production (50-65%), and DNA synthesis (6046%). Their experiences are quite similar to those reported on the effects of aflatoxin B1 in the macromolecular synthesis of F. uuruntiacum (Lillehoj and Ciegler, 1967). Fifty micrograms per milliliter B1 reduced DNA synthesis 80%, RNA production 48%, and protein synthesis 32%.

F. BIOCHEMISTRY Postulates pertaining to the role of aflatoxin as a cytotoxin have largely been generated from experiments in biological systems. These seem to fall into the broad category of an involvement by the toxin in

174

ALEX CIEGLER AND EIVIND 8. LILLEHOJ

processes related to DNA replication, transcription of messenger RNA, or translation of the messenger RNA into protein. The area of research attempting to elucidate the mode of action of aflatoxin offers a stimulating opportunity to investigate regulatory mechanisms at the molecular level. Another exciting area of investigation is the biosynthesis of aflatoxin. The literature relating to biosynthesis has been comprehensively reviewed by Mateles and Wogan (196713).

1 . E f e c t on Messenger RNA Synthesis Clifford and Rees (1966) observed that orotic acid incorporation into RNA and protein synthesis was inhibited in an in vitro system containing 0.032 mmoles aflatoxin B1 and rat liver slices. One of the striking aspects of their study was the immediate blockage of orotic acid incorporation while protein inhibition occurred 15 minutes later (Figs. 2,3). Since these results were identical to the action of 2500

1

-Control \

c 3 u 0

I

1

I

rat

-

2000-

c

-

.t 1500n 0)

-

0

.-c

n P)

-

-

0

15

30 45 Time [min.]

60

75

FIG. 2. Time curve of the incorporation of leucine-"C into the proteins of liver slices. 0 ,Control. 0 ,Aflatoxin B,-treated rat. (From Clifford and Rees, 1967a.)

actinomycin D (Trakatellis et al., 1964), it was felt that the inhibitory facet of aflatoxin B1 might be similar to that of actinomycin D. Further evidence supporting this concept was offered by tests of rat liver cell

175

MYCOTOXINS

nuclei which demonstrated that orotic acid was incorporated in the nucleotide pool of the nuclei but not into RNA. These observations suggested that aflatoxin B was interfering with RNA polymerase activity. Tests on the binding of aflatoxin B1 to DNA demonstrated

e u

z

-

.U aa

-

.-c -

FIG.3. Time curve of the incorporation of orotic acid-6°C into the RNA of liver slices. 0 , Control. 0 , Aflatoxin B,-treated rat. (From Clifford and Rees, 1967a.)

that the nucleic acid reduced the absorbancy of aflatoxin at 355 mp while an increase was observed at 385 mp (Fig. 4). The net conclusion of this work was that B1 inhibited messenger RNA synthesis by attachment to DNA and that the resulting decreased level of protein synthesis might be responsible for hepatic necrosis. At approximately the same time as the work of Clifford and Rees was published, Sporn et al. (1966) reported a verification of the finding that aflatoxin B 1 binds with DNA. The DNA caused a shift in the absorption maximum of B1 from 362-364 mp to 366-368 mp with a distinct hypochromism at 362 mp. RNA from E . coli also produced a shift in the B1 spectrum, but no change was seen with native or heattreated bovine serum albumin or calf thymus histone. Equilibrium dialysis experiments utilizing DNA treated with aflatoxin B1 demonstrated that 600 moles of native DNA-phosphorus bound 1 mole of afla-

176


toxin BI. Extending the earlier finding that rat liver tryptophan pyrrolase induction by hydrocortisone was inhibited by B1, Sporn et al. (1966) found that tryosine transaminase induction by hydrocortisone was also inhibited by B1. Since aflatoxin B1 lowered the nuclear ratio of RNAlDNA in rat liver cells, it appeared that the toxin exerted a greater inhibition on RNA synthesis than on DNA production. It was concluded that aflatoxin B1 was inhibiting the transcription process of rat liver nuclei in the same manner as actinomycin D (Dingman and Sporn, 1965).

340

360 380 400 Wavelength Imp]

FIG. 4. Difference spectra of aflatoxin BI with single (a) and double ( 0 )stranded DNA. (From Clifford and Rees, 1967b.)

In an extensive study on the effects of aflatoxin B1 in rats, Clifford and Rees (1967a) demonstrated that liver slices from B1-poisoned rats (7 mg./kg.) inhibited protein synthesis in vitro by 42% 1 hour after poisoning, but no effect was observed on in viuo incorporation of amino acid following aflatoxin B1treatment. They postulated that the difference in protein synthesis levels of intact livers and liver slices from B1-dosed animals was due to a greater requirement for newly formed messenger RNA in liver slices. Verification of earlier work demonstrated that hydrocortisone induction of tryptophan pyrrolase was inhibited by B1. Considering that the hormone-induced enzyme

MYCOTOXINS

177

was blocked by the toxin while in uiuo protein synthesis was quite insensitive to aflatoxin, apparently the two protein synthesizing mechanisms were operating independently of each other. A provocative explanation of the difference in protein synthesis of inducible and constitutive enzymes has been proposed by Moses and Calvin (1965). Their work suggests that the messenger RNA produced for normally inducible catalysts is more labile than that for Constitutive enzymes. Thus, the inhibition of messenger RNA would manifest itself initially in the blockage of induced enzymes while the constitutive protein synthesis would continue for some time. Such an explanation could rationalize the observed discrepancies between aflatoxin inhibition of induced enzyme and the absence of the effect on “total” protein synthesis. Liver slices from rats poisoned with hepatoxin, carbon tetrachloride (1.25 ml./kg.), demonstrated an inhibition of in uiuo amino acid incorporation, and in uitro the microsomal preparations from the livers of the treated animals also showed a decrease of protein synthesis (Clifford and Rees, 1967a). However, dosing the animals with B1 had no effect on the protein production of microsomal preparations derived from liver. It was suggested that the difference in inhibitory facility of carbon tetrachloride and aflatoxin B1 was a manifestation of the toxic action of carbon tetrachloride at the ribosomal level (translation) and the B1 at the transcription site. Clifford and Rees (1967a) also determined the distribution of aflatoxin B in various fractions of liver 30 minutes after treatment with the toxin. The results showed that the nuclear fraction of the liver cells contained the second largest concentration of B1 (29%) while the largest amount was found in the cell sap (42%). Another report on the effects of aflatoxin on rat liver RNA polymerase demonstrated that intraperitoneal injection of.l.0 mg. Bl/kg. followed by sacrifice of the animals at various times produced a decrease in polymerase activity (Gelboin et al., 1966). Maximum inhibition of 70% was observed at 30 minutes and persisted for 2 hours, followed by a partial recovery that caused a 10% decrease in enzyme activity at 12-24 hours. This study proved that aflatoxin rapidly interferes with RNA synthesis in the rat liver. Although there is a growing body of information supporting the view that aflatoxin inhibits RNA polymerase action, King and Nicholson (1967) reported that the addition of 25-42 pmoles of aflatoxin B1 to a 0.25 ml. incubation mixture for assaying E. coli RNA polymerase activity did not interfere with the RNA synthesis, whereas 10 pg./ml.

178


actinomycin D inhibited the polymerase action 90%. This study also showed that aflatoxin interacted with DNA but that the binding was relatively weak since it could be significantly decreased by adding 0.15 M sodium chloride to the test solution.

2. Efect on DNA Synthesis Although there seems to be mounting evidence supporting the concept that aflatoxin acts through suppression of the messenger RNA synthesis, several studies indicate that DNA metabolism is also affected. Gabliks et al. (1965) and Legator (1966) observed a decrease in DNA synthesis and mitosis in tissue culture cells propagated in the presence of aflatoxin, and Zuckerman et al. (1967a,b) reported a decrease in both nuclear DNA and RNA in tissue culture cells grown in the presence of the toxin. A significant decrease in mitosis and an increase in abnormal anaphase chromosomes have been observed in seedling roots of Viciafuba incubated with aflatoxin (Lilly, 1965). Rogers and Newberne (1967) found that following a 3 mg. Bl/kg. body weight dosage to rats, a marked depression occurred in mitosis and also in DNA synthesis in hepatic parenchymal cells. Aflatoxin B1 inhibits both DNA and RNA synthesis of rat liver cells following partial hepatectomy (De Recondo et al., 1965) even though the enzymes of DNA synthesis do not appear to be affected. Further work (De Recondo et al., 1966) suggested that B1was interfering with the nucleic acid synthesis by binding to the DNA and inhibiting its primer activity in the polymerase reaction. However, Wragg et al. (1967) recorded a 60-63% decrease in DNA of E . coli grown in the presence of B1 (5 pg./ml.) which they have related to a decrease in DNA polymerase activity independent of primer. The differences in conclusions on aflatoxin inhibition of DNA polymerase could reflect a fundamental difference between the toxic action in mammalian and microbial systems; this question deserves further consideration.

3. Binding to Histones Although Sporn et al. (1966) did not note a shift in the aflatoxin B1 spectrum when reacted with calf thymus histone, further work by Black and Jirgensons (1967) detected that aflatoxin B1 increased the viscosity of one of the two main fractions of calf thymus histone and of DNA. Equilibrium dialysis studies of the histone and of aflatoxin B 1 revealed that the two protein fractions bound 33 and 65 molecules of toxin per molecule of histone, respectively. Studies were also carried out utilizing equilibrium dialysis to determine the binding

179

MYCOTOXINS

of B1 to DNA. These tests demonstrated that one molecule of the toxin was bound for each five nucleotides. This level of binding corresponded quite closely to that reported b y Sporn et al. (1966). It was concluded that the binding forces of aflatoxin to histone are equal to or greater than for DNA. Since histones have been implicated in the replication-transcription processes (Hnilica and Billen, 1964; Liau et al., 1965), the authors postulate that the conformational changes brought about in the DNA and histones by aflatoxin could produce a change in nucleic acid coding. Black and Jirgensons (1967) also examined the melting profiles of DNA in the presence of aflatoxin B1 and found no differences due to the toxin which indicated that there was no change in the secondary structure of DNA due to BI (Fig. 5). I

I *Presence

I

I

of Aflatoxin

B1

..

1.11..1.1.1.1.1.1.

20"

40"

60" Temperature

80"

FIG.5. Thermal hyperchromicity of DNA in the presence (0)and absence ( 0 )of aflatoxin BI. (From Clifford and Rees, 1967b.)

4. Binding to DNA

Clifford et al. (1967) compared difference spectra from aflatoxins B1, G 1 , and G z , after they had been incubated with DNA to the

180


spectra of the toxins with no DNA. The change in spectra of G1 and G2 is significantly smaller than that observed with B1. The inhibition of orotic acid incorporation into RNA by rat liver slices by 50 pmoles of B1, GI, and Gz, in 3.0 ml. of reaction mixture was SO%, 29%, and 15%, respectively. Leucine incorporation into protein of rat liver slices was determined to measure the comparative aspects of the three aflatoxins on this synthetic mechanism. After 1 hour of incubation in 3.0 ml. volume, 20 pmoles of B1, 150 pmoles of GI, and 230 pmoles of Gz inhibited protein production 32%, 35%, and 38%, respectively. Clifford and Rees (1967b) determined the interaction of various nucleosides with aflatoxins. The difference spectra from aflatoxins B1, G I , and G2 showed the largest spectral changes with deoxyguanosine, deoxyadenosine, and adenine. Therefore, it was concluded that the greatest interaction occurred between the toxins and purine bases. Tests with various amino-substituted purines demonstrated that the amino groups aided the interaction. Although aflatoxin G 1 and G2 had a smaller difference spectra than B1 in the presence of DNA, this was not true with the nucleosides. In some instances GI gave the same magnitude of spectral change with the nucleosides as did B1. Analysis of melting profiles verified Black and Jirgensons’ observation that there is no interaction between DNA strands mediated by aflatoxin B1 (Fig. 5). Difference spectra from B1 incubated with double- or single-stranded DNA were essentially identical (Fig. 4). Examination of the stability of the BI-DNA complex showed that the two components separated without signs of the complex on a Sephadex G-50 column.

5. Comparison of Effects with Actinomycin D Examination of Table I1 comparing the action of aflatoxin B1 with actinomycin D demonstrates that most of the differences between the two compounds can be explained by the weaker binding of B1to DNA. Since several investigators observed a marked DNA inhibition in aflatoxin-treated systems, this phenomenon also deserves attention. Although it is generally accepted that actinomycin D selectively inhibits RNA polymerase action, DNA synthesis is also affected (Goldberg and Reich, 1964). Evidently DNA polymerase is less sensitive to actinomydin D than is RNA polymerase. It has been proposed that initially actinomycin D binds to DNA and blocks the surface of the template necessary for RNA polymerase activity while higher levels of the antibiotic affect the DNA polymerase through another mechanism (Goldberg and Reich, 1964). Actinomycin appears

TABLE I1 COMPARISON OF AFLATOXIN BI EFFECTSWITH THOSEOF ACTINOMYCIND

Condition

Aflatoxin B1 Reaction

Block messenger RNA synthesis Binds to DNA Inhibit enzyme induction by hydrocortisone Inhibit in uiuo protein synthesis Inhibit DNA synthesis following hepatectomy Inhibit protein synthesis in microsomes LDM dose produces liver necrosis Complex with DNA changes T, Interact with single stranded DNA Binding to DNA dissociated by salts DNA complexes separated by Sephadex G-50 Requires helical form DNA for binding

+

+

+

-

Reference

Reference

Clifford and Rees (1966) Clifford and Rees (1966)

+

Trakatellis et al. (1964) Goldberg and Reich (1964)

Wogan (1966a) Clifford and Rees (1966)

+

Greengardet al. (1963) Greengard et al. (1963)

De Recondo et al. (1965,1966) Clifford and Rees (1967a) Butler (1965) Clifford and Rees (1967b) Clifford and Rees (196%) King and Nicholson (1967)

+

Actinomycin D Reaction

+

-

+ + -

+

-

Hartman et al. (1963)

Clifford and Rees (1967b) Black and Jirgensons (1967)

Philips et al. (1960) Staehelin et al. (1963) Schwartz et al. (1965) Goldberg and Reich (1964) Goldberg and Reich (1964) Goldberg and Reich (1964)

+

Goldberg and Reich (1964)

35

:: 3

182


to bind in a groove of a DNA helix containing guanine. Phlemycin, another antibiotic, specifically inhibits DNA polymerase by binding to DNA containing adenine-thymine (Falaschi and Kornberg, 1964). The type of binding expressed by these two antibiotics seems highly specific, whereas the binding of proflavine to DNA inhibits both RNA and DNA synthesis (Terawaki and Greenberg, 1965). Evidence has been presented that proflavine binds to DNA by intercalating between adjacent base pairs (Lerman, 1963). These observations reveal that various types of binding to DNA affect the replication-transcription process in different ways. It is evident that more information is required on the nature of aflatoxin B 1 inhibition before the site of action can be categorically stated.

6. Lactones Since aflatoxin contains a lactone in the coumarin nucleus and since many lactones are biologically active, the carcinogenicity of the toxin has been considered from these structural characteristics. Dickens (1964) has observed that several lactones are carcinogenic. This activity has been related to the alkylating capacity of some lactones fundamentally associated with their electrophilic properties (Goldschmidt, 1965). Dickens and Jones (1963) found that injection subcutaneously of 50 pg. of aflatoxin twice weekly in rats produced rapidly growing tumors after 21 weeks, some of which were successfully transplanted. Later studies demonstrated that the same test system produced tumors at levels as low as 2 pg./dose (Dickens and Jones, 1965). These studies disclosed that aflatoxin was as active a carcinogen as P-propiolactone (Dickens, 1964). Although these studies suggested that aflatoxin was acting as a toxin b y alkylation, Dickens and Cooke (1965) related carcinogenicity of lactones to their chemical interaction with the sulfhydryl group of cysteine. Since aflatoxin did not react, it appeared that the toxin was not functioning as an alkylating agent. Several other workers have thought that aflatoxin functioned as an alkylating agent (Legator et ul., 1965; Butler, 1964; Wragg et ul., 1967). However, recent work on the instability of the DNA-aflatoxin complex shows that toxin bound to the DNA can be removed on a Sephadex-50 column (Clifford and Rees, 1967b) and other reports also suggest that the binding of the toxin to DNA is weak (Black and Jirgensons, 1967; King and Nicholoson, 1967). It would appear that, in uitro, the aflatoxins are not effective alkylating agents. However, the possibility exists that they are activated in uiuo,

MYCOTOXINS

183

producing toxic products. The in vivo activation of mitomycin C yields a bifunctional alkylating compound (Szybalski and Iyer, 1964). Another group of lactones has recently been implicated as carcinogenic agents. These are compounds containing the lactone of isocoumarin. Buu-Hoi et al. (1966) have synthesized 5-oxo-5H-benzo(e)isochromeno (4,3-b) indole which contains an isocoumarin system. A 0.6 mg. dose of this substance was injected subcutaneously into the flanks of mice three times with a month between injections (Lacassagne et al., 1967).Of the 54 mice tested, 52 developed sarcomas, and tumors were detected as early as 78 days following the first injection. These workers believe that the isocoumarins found in natural products such as ellagic acid and the fungal metabolite alternariol should be considered as possible sources of carcinogenicity. Dickens et al. (1966) have investigated the carcinogenic properties of sterigmatocystin, a metabolite of Aspergillus versicolor. The compound consists of a xanthone nucleus attached to a bifuran structure somewhat similar to the structure of aflatoxin in which the bifuran is attached to the coumarin nucleus. Dickens et al. (1966) found that in rats sterigmatocystin had l/250 of the tumorigenic activity of aflatoxin. Lillehoj and Ciegler (1968)have determined the biological activity of sterigmatocystin utilizing the duckling test as a comparison. They found that aflatoxin B1 is 125 times more effective than sterigmatocystin in initiating bile duct hyperplasia.

7. Nitrofurans It seems particularly interesting that some of the nitrofurans appear to function biologically in a manner quite similar to aflatoxin. Two of these compounds studied extensively are 3-amino-6-(2-5-nitrofuryl)vinyl)-1,2,4 triazine (NFT) and 5-nitro-2-furaldehyde semicarbazone (nitrofurazone).Endo et al. (1963)have shown that NFT is an inhibitor of DNA synthesis at levels that have little effect on RNA and protein production. Treatment of a lysogenic strain ofE. coli with NFT (0.2 pg./ml.) induced phage production and cellular lysis (Kato et al., 1966). A comparison of the NFT action with the activity of nitrofurazone appears instructive particularly as related to aflatoxin effects. Nitrofurazone is a mutagen (Zampieri and Greenberg, 1964) and has other radiomimetic properties (Terawaki and Greenberg, 1965), but it does not induce phage formation in a lysogenic strain of E. coli (McCalla, 1964) at 10 pg./ml. nor does it selectively inhibit D N A synthesis (Terawaki and Greenberg, 1965). The two nonalkylating

184


nitrofuran compounds are radiomimetic agents but differ in selectivity of DNA inhibition and phage induction. McCalla (1964) has offered the provocative idea that both compounds may inhibit DNA production similarly, but that nitrofurazone concomitantly inhibits respiratory processes which in turn block RNA and protein synthesis. This concept would also explain the absence of phage lysis by nitrofurazone since it can be assumed that virus multiplication depends on host metabolic energy. A dual inhibition of this sort, which is related to concentration and possibly environmental conditions, might explain the divergence of results obtained from aflatoxin-inhibited systems. 8. Furocoumarins

An interesting group of structural analogs of aflatoxins are the furocoumarins. Structures of naturally occurring coumarins and their related physiological activities have been extensively reviewed by Soine (1964) and compared with the aflatoxin effects (Schoental, 1967). Some furocoumarins are highly active photosensitizing agents capable of inducing sunburn and augmenting skin pigmentation (Musajo and Rodighiero, 1962).The tests are performed by both painting the compounds on the skin or injecting them intradermically followed by irradiation. Psoralen, 8-methoxy-psoralen (xanthotoxin) and 5-methoxy-psoralen (bergapten) (Fig. 10) appear to be the most photodynamic compounds of the furocoumarins tested (Musajo and Rodighiero, 1962). The simultaneous exposure of various species of bacteria to furocoumarins and visible light killed the microogranisms (Fowlks et al., 1958). Reportedly, a mixture of furocoumarins inhibited the growth of several fungal strains at 200 p.p.m. or less (Martin et al., 1966). Examination of the photosensitization process conferred by 8-methoxy-psoralen or Sarcina h t e a established that the photo-induced furocoumarin was functioning as a mutagen (Mathews, 1963). This finding suggested that the compound was interfering with DNA metabolism. Colombo et al. (1965) found that kidney cells irradiated (3650 A in the presence of psoralen were inhibited in RNA, DNA, and protein synthesis. Furthermore, irradiated psoralen solutions inhibited the propagation of DNA viruses but not RNA viruses. Investigation of the binding of bergapten to DNA showed that there was a weak binding in the dark, but following irradiation (3650 A) a stable chemical linkage was formed (Musajo et al., 1966). Thus, it could be that the biological action of furocoumarin may be due to a binding with DNA.

MYCOTOXINS

185

G. ANALYSES Specific chemical analyses for aflatoxin have not yet been developed. Qualitative and quantitative assays for the various aflatoxins depend on thin-layer chromatography with authentic samples of known materials for reference. The presence of toxin is usually verified, when necessary, by a bioassay.

1 . Chemical Analysis Chemical assays are complicated by several factors: (1)Aflatoxins are usually found in agricultural commodities, necessitating a preliminary extraction procedure; (2) extraction procedures adequate for one commodity are inadequate for another; (3)the amount of toxin is small, usually in the microgram per kilogram range; (4) distribution in natural products is usually extremely uneven, requiring large samples for toxin detection; and (5) the assay is nonspecific and requires further confirmatory tests. The most obvious difficulty in detecting aflatoxin in contaminated materials is adequate sampling. The heterogeneous distribution of toxin makes it desirable to test the entire batch of a suspected commodity, but obviously, practical considerations limit the size of sample to be tested. Usually 5-kg. samples are taken of peanuts and l-kg. samples of other substances. These samples are ground to give homogeneous distribution of toxin, then 50-gm. portions are taken for toxin analysis. a. Peanuts and Peanut Products

The early methods of assay developed in the United States and in Europe have been summarized well by Borkeret al. (1966). In general, methods involve defatting, extraction of toxin, and thin-layer chromatography. Pertinent references to this literature are: Coomes and Sanders (1963), Coomes et al. (1964), Campbell et al. (1964), Trager et al. (1964), Nesheim (1964), Nesheim et al. (1964), Nabney and Nesbitt (1965), and Coomes and Feuell (1965). The method currently used in the United States is as follows (Anonymous, 1966): Peanuts or the peanut product is defatted in a blender with hexane-methanol-water; the slurry is centrifuged and the aqueous methanol layer added to acid-washed Celite 545; the Celite slurry is transferred to a chromatographic column, washed with hexane, then with hexane-chloroform (1: 1); the eluate from the chloroform-hexane wash contains the aflatoxin; the solvent is evap-

186


orated off to near dryness and fresh chloroform added to a given volume; this solution is spotted on Silica Gel G-HR thin-layer chromatographic plates along with a series of internal standards and developed in chloroform-methanol (93:7); fluorescence of the unknown is matched to the standard spot having the same degree of fluorescence. The amount of toxin present in the original peanut sample is determined by the following formula: pg./kg.

=

S x Y x V/X x W

where S = microliters of aflatoxin B1 standard equal to unknown; Y = concentration of aflatoxin B 1 standard, pg./ml.; V = volume in microliters of final dilution of sample extracted; X = p1 of extracted sample spotted giving fluorescent intensity equal to S , the B1 standard; and W = grams of sample applied to Celite column. A new solvent system for thin-layer chromatography (TLC) using acetone-chloroform (1:9) in place of methanol-chloroform (7:93) gives better resolution of the four principal aflatoxins (Eppley, 1966a). In addition, a newly developed assay procedure for aflatoxin in peanuts and peanut products seems to have advantages over the method outlined (Eppley, 1966b). In the improved procedure direct extraction of a water-wetted sample by chloroform is combined with the use of a silica gel column for defatting and cleanup. This method can be applied to samples ranging from 50 gm. to 1kg. It permits the detection and estimation of as little as 1pg. aflatoxinslkg. of a peanut product. Andrellos and Reid (1964) devised a confirmatory test for aflatoxin B1 by derivative formation in anhydrous solutions with trifluoracetic acid, formic acid-thionyl chloride, and glacial acetic acid-thionyl chloride; the derivatives have altered chromatographic behavior. This procedure was studied collaboratively in 19 laboratories and no false identifications were made (Stoloff, 1967).

b. Corn and Corn Products Various methods developed for other commodities also appear to be applicable to corn and corn products (Pons et al., 1966; Wiley, 1966). However, according to Smith (Campbell, 1967), methods satisfactory for corn are not satisfactory for the analysis of the primary wet-milling products of corn. Interfering substances that were solvent extracted are not removed during cleanup procedures and actually prohibit detection of aflatoxin at levels as high as 200 pg./kg. Smith modified current procedures by extracting the primary product with chloro-

MYCOTOXINS

187

form in the presence of water. The chloroform extract is cleaned up by hexane extraction, lead acetate precipitation, or chromatography on a Celite or silica gel column with diethyl ether. c . Cottonseed and Cottonseed Meal

Methods for the analysis of these products have been developed by Engebrecht et al. (1965), Chen and Friedman (1966), Pons et al. (1966), and Stoloff et al. (1966). The last three involve essentially an extraction with acetone or methanol followed by removal of interfering substances with lead acetate; a further cleanup by column chromatography follows in which the toxin is eluted and its concentration determined by thin-layer chromatography. d . Other Products

Shotwell et al. (1968) found that oats extracted by the procedure of Campbell and Funkhouser (1966) contained fluorescing substances in the hulls which behaved like aflatoxins B I and G I on thin-layer chromatographic plates. Several methods of purification, including solvent distribution, chromatography on silica gel, and lead acetate precipitation failed to remove these materials. The interfering substances did not exhibit aflatoxinlike activity in ducklings and failed to form derivatives by the procedure of Andrellos and Reid (1964). Development of Silica-Gel G-HR coated thin-layer plates in 5 or 7% methanol in chloroform adequately separated these “oat factors” from aflatoxins B 1 and G 1. Yokotsuka et al. (1966) reported that a number of molds used in food fermentations produce nontoxic substances resembling the aflatoxins in their fluorescent and chromatographic behavior. Interfering substances have also been reported for peanuts (Andrellos and Reid, 1964). Little progress has been reported in the analysis of cocoa, tea, or coffee for aflatoxin (Campbell, 1967). Procedures for the analysis of aflatoxin produced by fermentation of various grains in a semidry state have been described (Shotwell et al., 1966; Stubblefield et al., 1967). The yellow pigment that is usually produced in these fermentations and is extracted along with aflatoxin can be removed by absorption onto insoluble basic green copper carbonate (Wiseman et al., 1967). Recovery and analysis of aflatoxin from deep culture fermentations have been described by Ciegler et al. (1966b). The whole beer is sonified after addition of methanol-chloroform (30:70 v./v.) and the toxins are recovered from the chloroform layer. Davis et al. (196613)

188


described the separation of aflatoxin from small-scale shaken-flask fermentations by paper chromatography. The method of Lee (1965) for the quantitative determination of aflatoxin in peanut products was applied by Denault and Underkofler (1967) for detecting toxin in enzyme preparations produced by fungal fermentation. Although no aflatoxin was detected, all products tested showed the presence of a fluorescing spot having an R, higher than aflatoxin B1 on thin-layer chromatography. Estimation of aflatoxin M in milk has been reported by Purchase and Steyn (1967). They found extraction of toxin to be most efficient when an azeotropic mixture of acetone-chloroform-water was used in a Soxhlet apparatus.

2. Znstrumental Analysis The visual estimation of aflatoxin on thin-layer chromatographic plates involves a deviation of f 17% with an accuracy within 20%. Densitometric methods using commercially available fluorodensitometers have proved more accurate in determinations than the visual method with average deviations of about k 2% (Ayres and Sinnhuber, 1966; Stubblefield et al., 1967). Because these instruments are comparatively expensive, Peterson et al. (1967) developed an economical combination of a simple darkroom densitometer connected to an available recorder which gave more reliable data for routine aflatoxin analyses than visual observation. A spectrophotometric method for determining the aflatoxins, especially Bl, based on intensity of ultraviolet absorption at 363 mp. after purification by thin-layer chromatography, was developed by Nabney and Nesbitt (1965). It has not been adopted for general usage because of inherent limitations. Gajan et a1. (1964) found that aflatoxins B and G 1 gave characteristic oscillopolarographic traces having peak potentials at - 1.33 and -1.25 k 0.02 V. versus a silver wire electrode for B1 and G1, respectively, in an electrolyte containing (CH&N Br and LiCl in aqueous methanol. The diffusion current is proportional to concentration. The method requires some purification of the toxins but not to the same extent as that required for infrared analysis.

3. Biological Analyses a. Duckling The lack of a definitive chemical test has necessitated confirmation of suspected aflatoxin contamination by biological tests. Of the

MYCOTOXINS

189

various laboratory animals tested, the duckling represents the most susceptible species toward acute lethality. The duckling test is based on the work of various laboratories (Sargeant et al., 1961a,b; Carnaghan et al., 1963; Armbrecht and Fitzhugh, 1964; Wogan, 1965). The test is based, not on lethality, but on the degree of hyperplasia of the bile duct epithelium. The degree of hyperplasia is related roughly to the amount of toxin fed to the bird up to a point. Generally, l-dayold ducklings are used. Samples are given via a capsule or by the use of stomach tubes over a 4- to 5-day period. Samples are usually dissolved in water or propylene glycol. Control ducklings are fed a total of from 0 to 16 pg. aflatoxin B1; aflatoxin G1 requires about a 10-fold higher dose to give comparable results. After the last dose of toxin is given, the birds are held 2 additional days, sacrificed, a section of liver tissue collected and fixed for histological examination. The degree of hyperplasia is usually scored as 0 to 4 or 5-t. Typical results from feeding aflatoxins B1, B2, and GI are given by Wogan (1965). Carnaghan et al. (1963)compared toxicity of the four main aflatoxins and reported the oral 7-day LDso for each compound to be: Aflatoxin B1, 18.2 pg.; B2, 84.8 pg.; G1, 39.2 pg.; and G2, 172.5 pg. (all data on a 50-gm. body weight basis). Although an LD50 was not obtained for aflatoxin B1 hemiacetal (aflatoxin Bza, aflatoxin B-W), Ciegler and Peterson (1967) found it to be considerably less toxic than aflatoxin B1 to ducklings. Aflatox.in M 1reportedly has an LD50 of 16 pg.; M2, 61.4 pg. (Purchase, 1967). Sterigmatocystin, a compound closely related to aflatoxin and believed to be in its metabolic pathway of biosynthesis, was nontoxic to ducklings, evoking no bile duct hyperplasia (Lillehoj and Ciegler, 1968). It should be noted that the hyperplasia invoked by aflatoxin is not specific since other toxic agents cause a similar response. However, in conjunction with the thin-layer chromatographic examination, hyperplasia gives a reasonable confirmation. Although the duckling assay is the generally accepted biological test used, it is comparatively expensive, time consuming, and requires highly trained technicians. Other biological systems have been examined for aflatoxin bioassay.

b. Mollusk Eggs Townsley and Lee (1967) report that aflatoxin B1 inhibits cell cleavage in fertilized mollusk (Bankia setacea) eggs without preventing fertilization or nuclear division. In the presence of aflatoxin the fertilized eggs remain unicellular and become multinuclear, whereas

190


the controls are multicellular. The bioassay requires a minimum of technique and training and is sensitive to concentrations of 0.05 pg. aflatoxin per milliliter. When the reaction is observed with the aid of a low-powered microscope (70X),the concentration of aflatoxin required can be reduced considerably below 0.05 pg.; e.g., 0.0025 pg./50 pl. final volume of solution when observing egg division or 0.0025 pg./50 p1. final volume of solution when observing the presence or absence of the larval swimming state. The test requires 2 to 4 hours for completion when observing egg division and 18 hours for swimming larvae. The limiting features of this test are the requirement for seawater, availability of the particular mollusk, and the need to work at low temperatures (10"-20°C) for normal larval development. However, Townsley and Lee suggest that other mollusks might also work although they did not investigate this aspect.

c. Embryonated Eggs Injection of aflatoxin into the yolk of 5-day-old chicken embryos by Platt et ul. (1962) caused deaths; the quantity of toxin required was only 1/200 of that needed for a positive result in the day-old duckling. Verrett et u1. (1964) found that chicken embryos were feasible as a test organism for the assay of aflatoxin toxicity. Best results came from toxin injection into the yolk or air cell before incubation; toxicity was greater on injection via the air cell route than the yolk. A dose response was exhibited in that the toxicity of the samples was related to the mortality at the time of hatching. The LDsoat 21 days' incubation was 0.048 pg. for the yolk and 0.025 pg. for the air cell route. Aflatoxin G I (1.0 pg.) gave 60% mortality at 21 days. Nonsurviving embryos injected with B1 showed a severe growth retardation-in most cases edema, hemorrhage, underdevelopment of the mesencephalon in embryos that died before the seventh day, mottled and granular liver surface, short legs, and slight clubbing of the down. Gabliks et al. (1965) determined the LDsOfor 15-day-old duck and 10-day-old chick embryos that were injected into the chorioallantoic sac and incubated for 48 hours: 0.5-1.0 pg./ml. for duck embryos; 2.0-5.0 pg./ml. for chick embryos. The greater susceptibility of duck embryonated eggs is in agreement with the relative susceptibility of young ducklings and chickens tested in vivo (Asplin and Carnaghan, 1961).

d . Tissue Culture Addition of aflatoxin to calf kidney monolayer cultures destroyed

MYCOTOXINS

191

the cytoplasm and nucleus in affected cultures (Juhasz and Greczi, 1964). Toxin concentrations of 0.1-0.5 p.p.m. caused cell destruction after 48 hours’ incubation. Extractions of up to a dilution of aflatoxin-contaminated peanuts gave similar results that could only

be attributed to the presence of toxin. Legator and Withrow (1964) noted that crude aflatoxin mixtures, as well as crystallized aflatoxin B1 , suppressed mitotic division in heteroploid and diploid human embryonic lung cells. Aflatoxin inhibited mitosis between the fourth and sixth hours after exposure with maximum effect between 8 and 12 hours. These investigators then determined the feasibility of applying the measurement of mitotic suppression as a practical bioassay for aflatoxin. A concentration of 0.01 pg. could be detected and a level of 0.03 pg. produced a 51% reduction in mitosis. After 48 hours’ incubation, aflatoxin from 0.05 to 1.0 p.p.m. reduced the growth of cultured heteroploid human embryonic lung cells (Legator et al., 1965); cells did not grow at 5.0 p.p.m. After exposure of 8-12 hours to 1 p.p.m. BI, there was a 92% increase in giant cells over the control. In studies on Chang liver cultures, there were also decreases in cell number, protein, RNA, and DNA with increasing aflatoxin B1 concentrations (Gabliks et al., 1965).

e. Plant Albinism Schoental and White (1965) found increasing degrees of albinism in the leaves of cress (Lepidium sativum) subjected to from 1 to 10 pg. aflatoxin per milliliter; seed germination was prevented a t 25 pg./ml. These workers suggested that these effects could be elaborated into a simple test for aflatoxin detection in suspect materials.

f. Microorganisms Burmeister and Hesseltine (1966) surveyed 329 microorganisms (including bacteria, fungi, algae, and one protozoan) for their sensitivity to aflatoxin. They reasoned that a sensitive microorganism might supplement the duckling assay in establishing toxicity levels of aflatoxin in contaminated feeds. Sensitive cells could also be used in an assay when toxin levels were in excess of those required to inhibit the test organism. A strain of Bacillus brevis and two of B . megaterium were most sensitive to aflatoxin, being inhibited at 10 and 15 pg./ml., respectively. This level appears to be too high to be useful for a quantitative assay. Arai et al. (1967), reported a minimal inhibitory concentration of 7.5 pg. aflatoxin per milliliter against Streptomyces viginiae and S . netropsis but did not attempt to develop an assay.

192


g. Fish

Trout are extremely sensitive to aflatoxin (Halver, 1965). This response suggests that smaller fish, such as guppies, might also prove to be sensitive and could serve as a low-cost test animal. We are not aware of any studies in this area.

4 . Standards The standards used routinely in most laboratories are chloroform solutions of the four aflatoxins (B,, Bz,G I , and G z ) .Unless elaborate precautions are taken, however, chloroform evaporates rapidly, leading to decreasing values given to aflatoxin in unknown solutions. Peterson and Ciegler (1967) circumvented these difficulties by substituting water for chloroform. H. SAFETYPROCEDURES The extreme toxicity and demonstrated carcinogenicity of aflatoxin for various domestic and laboratory animals, including primates, have dictated stringent safety procedures in laboratory experiments with this family of toxins. Dry toxin is electrostatic and is best transferred, when necessary, in laboratory hoods. All equipment coming in contact with the toxin should be decontaminated by washing with 5% NaOCl (Fishback and Campbell, 1965). Work surfaces should also be decontaminated periodically with this bleach. Procedures for operating and decontaminating fermentors in which toxin has been produced are described by Ciegler et al. (1966b). A t the Northern Laboratory, disposable paper laboratory coats are worn and can be discarded should they become contaminated. Face masks and rubber gloves are worn when dry toxin is handled. In the case of accidental spillage or breakage, the contaminated area can be detoxified by the liberal application of hypochlorite. Where hypochlorite is unavailable, strong alkali, e.g., NH40H, can be substituted. When pipetting solutions of toxin, it is best to use a rubber bulb pipetting device or some other vacuum system. A high intensity portable ultraviolet light should be used to monitor work surfaces occasionally.

I. CONTROL AND DETOXIFICATION Obviously, the best method to control aflatoxin is prevention. The knowledge and technology on how to sharply curtail mold contamination is available and should be applied at all stages of growth, harvest, transportation, and processing of crops. Detailed discussion on these aspects has been published by Goldblatt (1966). Prevention of

MYCOTOXINS

193

mechanical damage during harvesting and reduction of moisture content of susceptible crops below the level necessary for spore germination are probably the most important factors for control of mold growth and aflatoxin production. It should be noted that all parts of the stored commodity must be at the correct moisture level; an average lowmoisture level is not sufficient protection against mold infestation. Several approaches have been investigated for detoxification of aflatoxin-contaminated foods and feeds: (1)physical or mechanical, (2) chemical, and (3)biological. Where applicable, good progress has been made in the use of culling devices, such as photoelectric sorters, to remove contaminated cottonseed and peanuts (Dollear and Gardner, 1966). Since mold-contaminated peanuts are usually darker in color than clean nuts, suspect nuts could be removed from the mass by a combination of electronic and manual picking procedures before and after roasting. Density differences between viable and nonviable cottonseed have also been used as a basis of separating contaminated material (Dollear and Gardner, 1966). Extensive investigation on detoxification by heat and various chemical agents, including acids, bases, salts, oxidizing and reducing agents, and various gases, has been carried out at the Southern Regional Research Laboratory of the US. Department of Agriculture. Mann et al. (1968) obtained an 80% reduction of the 214 p.p.b. aflatoxin present in cottonseed meal (with a moisture content of 20%)by heating 120 minutes at 100°C; only 34% reduction in aflatoxins resulted from a similar treatment of a contaminated peanut meal. However, Coomes et al. (1966) succeeded in reducing the aflatoxin content of 10%moisture peanut meal from 7,000 p.p.b. to 370 p.p.b. by heating at 120°C for 4 hours. Previously, attempts to eliminate toxin from peanuts and peanut meal by heat were unsuccessful (Carnaghan, 1964; Pomeranz, 1964). Trager and Stoloff (1967) investigated reactions of the four main aflatoxins with a series of selected reagents to determine their possible usefulness in detoxification procedures; detoxification was determined by bioassay. Reactions appear to be primarily addition and oxidation involving the olefinic double bond of the terminal furan ring and oxidation involving the phenol formed on opening of the lactone ring. Lack of unsaturation of the terminal furan ring of aflatoxins B2 and G2 makes them less susceptible than B1 and G I to electrophilic attack. Gaseous chlorine, chlorine dioxide, nitrogen dioxide, and 5% NaOCl detoxified the four toxins.

194


Removal of aflatoxin from contaminated seeds by polar solvents has been effective. A solvent system of acetone-hexane-water removed aflatoxin and oil readily and quantitatively from peanuts and peanut meal (Robertson et al., 1965); toxin in the oil is destroyed on processing (Parker and Melnick, 1966). Dollear and Gardner (1966) investigated removal of aflatoxin by various solvents from prepressed peanut cake on a pilot-plant scale; 90-99% toxin was removed, depending on the solvent and process used. Pons and Eaves (1967) reported that aqueous acetone (acetone-water, 70:30) removed 96-98% of the aflatoxin from cottonseed flakes. Inactivation of aflatoxin in peanut meal and cottonseed flakes by chemical means has been reported by Dollear and Gardner (1966) and on a pilot-plant scale by Dollear (personal communication); sodium hydroxide, methylamine, ozone, and ammonia in combination with heat and moisture were particularly effective. The effectiveness of ammonia was confirmed by Masri (1965). However, there was some loss of nutritive value of the detoxified substances. Destruction of toxin in contaminated material by irradiation appears to be ineffective (Feuell, 1966b),although instability of aflatoxin on exposure to ultraviolet light has been noted by many workers. Sreenivasamurthy e t al. (1965) used calcium chloride solutions to extract aflatoxin from peanuts; they suggested that precipitation of the protein with calcium chloride at neutral pH instead of acid precipitation at isoelectric pH eliminated 80% of the toxin. Use of gaseous chlorine destroyed 90% of the toxin in peanut meal (Fishbach and Campbell, 1965) but left a residual odor of chlorine. Sreenivasamurthy et al. (1967) adjusted the pH of contaminated meal to 9.5 to open the lactone ring of the toxin and then treated the meal at 80°C with hydrogen peroxide to effect detoxification. The destruction of aflatoxin was confirmed by bioassay with ducklings and duck embryos. Biological detoxification of contaminated agricultural commodities has been reported. Ashworth et al. (1965) found that addition of various molds to naturally contaminated peanuts resulted in partial diminution of the aflatoxin present. Several investigators have observed that toxin concentration in a production medium often declines after peak yields are attained (Ciegler et al., 1966b; Davis et al., 1966a; Schroeder, 1966). This degradation was studied in detail by Ciegler et al. (1966b) in 20-liter fermentors. They noted that toxin loss was correlated with mycelial lysis. The percentage and rate of toxin degradation were independent of the toxin concentration and appeared to be nonenzymic and nonspecific. Degradation simulating

MYCOTOXINS

195

that occurring in the fermentor was achieved by reacting aflatoxin with peroxidized methyl esters of vegetable oil; initial degradation was rapid and appeared to involve a complex series of reactions. Ciegler et al. (1966a) screened approximately 1000 microorganisms for their ability to either destroy or transform aflatoxin B1 and G1. Some molds and mold spores partially transformed aflatoxin B1 to new fluorescing compounds. Only one of the bacteria assayed, Flauobacterium sp. NRRL B-184, removed aflatoxin from solution. Both growing and resting cells of this culture took up toxin irreversibly. Toxin-contaminated milk, oil, peanut butter, peanuts, and corn were rapidly and completely detoxified at 28°C; contaminated soybean was 86% detoxified after 12 hours’ incubation. Duckling assays showed that detoxification of aflatoxin solutions by B-184 was complete, with no new toxic products being formed. Conversion of aflatoxin B1 to a new fluorescing compound in the presence of various acid-producing molds proved nonspecific (Ciegler and Peterson, 1967) and represented an acid-catalyzed conversion of B to hydroxy-dihydro-aflatoxin B1 (Biichi et al., 1966). This compound is relatively nontoxic. Teunisson and Robertson (1967) reported that Tetrahymena p y r i formis W decreased the concentration of 2 pg. aflatoxin Bl/ml. 67% in 48 hours with production of a bright blue fluorescent substance. The toxicity of this new compound was not determined. Other microorganisms have also been reported to remove aflatoxin from solution, but detoxification studies were not carried out (Burmeister and Hesseltine, 1966; Arai et al., 1967). None of the processes described for detoxification of contaminated foods or feeds appears to be completely satisfactory for industrial use. Although further research is needed, it should be realized that any process adopted will require approval by the Food and Drug Administration. To gain this approval will necessitate proof that the “nutritive value of the feed is unimpaired and all toxins are destroyed” (Barnes, 1966). 111. Alimentary Toxic Aleukia (ATA)

The intensive investigation on mycotoxins that followed the outbreak of aflatoxicosis in England in 1960 has tended to obscure the research that was already underway on mycotoxins before then. Much of this work was carried out in Russian laboratories. A comprehensive review of the Soviet work between 1932 and 1953 on the clinical aspects of alimentary toxic aleukia (ATA) and on the botany, phytopathology, and toxicology of Russian cereal food was published by

196


Mayer (1953a,b). Mayer termed the disease endemic panmyelotoxicosis, but the official Soviet designation remains ATA. The disease produces a wide range of manifestations in humans but most include the clinical symptoms of leukemia, agranulocytosis, necrotic angina, hemorrhagic diathesis, sepsis, and aregenerative exhaustion of the bone marrow. The disease has not been observed outside of Russia and is always associated with grain overwintered in the field. These grains include wheat, rye, oat, buckwheat, and millet; this last grain is particularly suspect since it is widely grown and late ripening causes large amounts of it to be left to overwinter. The disease was particularly rampant during World War I1 when manpower shortages caused unusually large amounts of grain to be left to overwinter in the fields. Studies in depth on the etiology of the disease were carried out by Joffe from 1943 to 1950; this work was later published in a series of papers (Joffe, 1960a,b, 1962, 1963, 1965). In addition, these papers contain an extensive bibliography of his publications on the same subject previously published in Russia. The causes of alimentary toxic aleukia were identified as fungi (Joffe, 1960a). Large numbers of toxic fungi were isolated from cereals that had been overwintered in the Orenburg district, U.S.S.R., where the disease had been particularly prevalent (more than 10% of the population was affected, with many deaths recorded). Isolates from summer-harvested cereals were not toxic. Symptoms of ATA were best reproduced in cats, but mice and skin tests on rabbits, calves, and other animals were also used. The fungal species, of which toxic and highly toxic isolates were most common, were Fusarium poae, F. sporo trichioides, and Cladosporium epiphyllum. Most of the toxic fungi isolated from overwintered cereals were characterized by cryophilic properties, growing satisfactorily at -2" to -10°C. Sharp temperature fluctuations with freezing and thawing gave intense production of toxin in both pure and mixed cultures. Cultures of Fusarium and Cludosporium were most toxic at the time of abundant spore formation when grown at -2" to - 10°C; no toxin was produced at 23"-25"C. Mixtures of toxic cultures grown together appeared to be more toxic than cultures grown singly; this condition indicates synergistic action. The toxins produced by F . sporotrichioides, F. poae, C . epiphyllum, and C .fugi were identified by Bekker (1963; cited by Joffe, 1965) and named sporofusariogenin, poaefusariogenin, epicladosporic acid, and fagicladosporic acid, respectively (Fig. 6).

197

MYCOTOXINS

%ix CH-CHc.CH-!O

“’i i i“‘

H3C

OH (c~HPo~)(c~H~oos)z-O-H

Furariogenin CH,(CH~)PCH=CH(CH~)~-C’

Epicladosporic a c i d

0

‘”

C H J ( C H Z ) ~ C H - C H ( C H ~ )‘SH ~-C/~

Fagicladosporic acid

FIG. 6. Structural formulas of toxins produced by species of Fusosium and Cladosporium associated with alimentary toxic aleukia (ATA).

IV. Ochratoxin

Aspergillus ochruceous is a ubiquitous mold found in soil, moist cereals, legume crops, and spice; it constitutes part of the flora of katsuobushi, a Japanese fermented fish preparation. van der Menve and his co-workers (1965) inoculated a variety of natural substrates (corn, wheat, sorghum, rye, rice, buckwheat, soybean, and peanuts) which were subsequently fed, with lethal results, to day-old Peking ducklings, weaned white mice, and rats. The toxic factor was extracted from inoculated corn with methanolchloroform. The extract was fractionated with aqueous bicarbonate; the isolated acids were chromatographed on neutral silica, followed by separation on Dowex 1. The separated compounds were named ochratoxin A, B, and C. Ochratoxin A has the same order of toxicity as aflatoxin B1;the LD50 in 50-gm. ducklings was &25 pg. However, Nel (1967) reported that the LDsoof ochratoxin A in ducklings was 250 pg./duckling, a 10-fold difference from the South African results. Livers from toxin-fed ducklings showed acute fatty infiltration; there was no evidence of necrosis. Ochratoxins B and C were nontoxic at a thousandfold higher dosage. It is not yet known whether or not ochratoxin A is carcinogenic. Toxicology of ochratoxin has been reviewed by Purchase and Nel (in Mateles and Wogan, 1967a). van der Menve et al. (1965) determined the structures of ochra-

198


toxins A, B, and C. Ochratoxin A was shown to be 7-carboxy-5chloro-8-hydroxy-3,4-dihydro-3R-methylisocoumarin.Ochratoxins B and C were characterized as the dechloro and ethyl ester derivatives of ochratoxin A (Fig. 7).

CI

FIG.7.

Structural formulas of ochratoxin A, B, C.

A rapid method for the detection and estimation of ochratoxin was developed by Steyn and van der Menve (1966). It involved extraction of. the moldy material with chloroform-methanol (1:1) in a Soxhlet, extraction of the extract with aqueous sodium bicarbonate and then acidification of the aqueous layer followed by chloroform extraction. Portions of the latter extract are spotted on TLC (silica gel) along with a standard and developed in 3:l benzene-acetic acid. Under ultraviolet light, ochratoxin A fluoresces green at R ~ 0 . 5ochratoxins ; B and C fluoresce blue-green and pale-green at RJ 0.35 and 0.65, respectively. A method for the detection and semiquantitative estimation of ochratoxin A in flour and other cereal products, which could be used in conjunction with the analysis for aflatoxin, was described by Scott and Hand (1967). It involved extraction of the sample with aqueous methanol and n-hexane, partition of the toxin on a Celite column, separation on TLC, and estimation of the intensity of fluorescence compared with that of a reference standard. By this procedure, 25 pg. in 1 kg. of cereal foodstuff could be detected. The method is applicable to whole wheat flour, corn meal, barley cereal, and rice cereal.

199

MYCOTOXINS

Nesheim (1967) found that 4 out of 14 cultures of A. ochraceous produced aflatoxin A. Ochratoxin A was readily produced in shaken-flask culture, 10-liter and 100-liter fermentors (Ferreira, 1966). Yields of 50-100 mg. toxin/ liter resulted on a medium containing sucrose and glutamate. V. Sporidesmin Sporidesmin is the toxin produced by the saprophytic mold, Pithomyces chartarum (formerly Sporidesmium bakeri), growing on pasture grasses. It is responsible for the hepatotoxic disease, facial eczema, of sheep and cattle, particularly in Australia and New Zealand (Thornton and Percival, 1958, 1959; Percival and Thornton, 1958). Eczema and edema of areas of the skin exposed to sunlight are typical signs of the disease (Cunningham et al., 1942; Clare, 1944; Done et al., 1960; Hore, 1960). Published evidence indicates that photosensitization secondary to obstructive damage to bile ducts is the major cause of the skin lesions (Mortimer and Taylor, 1962). However, repeated skin contact of sheep to sporidesmin also caused skin lesions without exposure to sunlight (Gallagher, 1964). Synge and White (1959, 1960) isolated, purified, and characterized the toxin from ether extracts of the mold and from its spores by methanol extraction. They found sporidesmin to be resinous; it would, however, form a crystalline adduct with carbon tetrachloride and other solvents. The molecular structure of sporidesmin and sporidesmin B, also produced by P. chartarum, was determined after considerable effort by several laboratories (Fridrichsons and Mathieson, 1962; Hodges et al., 1963; Ronaldson et al., 1963). Sporidesmin was found to be Cl8H20C1N30& and sporidesmin B, C1BH20C1N305S2 (Fig. 8). These compounds resemble, structurally, the antibiotic,

Gliotoxin H3

h,OH

FIG.8. Structural formulas of sporidesmin and gliotoxin.

200


gliotoxin. The chemistry and biochemistry of sporidesmins and other 2,5-epidithia-3,6-dioxopiperazines have been exhaustively reviewed by Taylor (in Mateles and Wogan, 1967a). Bioassays for the presence of toxin are tedious and time consuming. Sporidesmin causes mild to moderate liver damage to guinea pigs at the 50-100 pg. level and severe damage at the 100-200 pg. level. The animals are dosed for 7 days and examined 21 days from the initiation of dosing (Te Punga and MacKinnon, 1961).MacKinnon and Te Punga (1961)developed a grading system for bilary epithelial hyperplasia, damage to the interlobular bile ducts, vascular lesions, and focal coagulative parenchymal necrosis caused by feeding P. chartarum to guinea pigs. Fastier (1961)unsuccessfully attempted to demonstrate in vitro toxin neutralizing antibodies in the sera of guinea pigs dosed with P. chartarum spores, in rabbits hyperimmunized with fungal material, or in sheep suffering from facial eczema. Fastier did show that sporidesmin was highly toxic to tissue cultures and an assay procedure was developed on this basis. A prominent biochemical characteristic of the enterophepatic disturbance preceding bilary obstruction in toxicated animals is an accumulation of liver triglycerides; this phenomenon was interpreted as a malfunction of the hepatic triglyceride-secreting mechanism (Peters, 1963;Peters and Smith, 1964).Peters (1966)later found that sporidesmin had a comparable effect in the rabbit (a nonruminant) as in sheep. His studies led him to the following conclusions: (1) the early action of sporidesmin is similar in its effects to those of many other hepatotoxins. (2)The early action of sporidesmin results in changes in the triglyceride equilibrium of serum and liver, which are probably initiated by transitory disturbances of hepatic mitochondrial protein synthesis. (3)The pathological and biochemical changes that follow recovery from the cytotoxic effect of sporidesmin may result from accumulation of bile acids in the liver and in the circulatory system. VI. F-2 Estrogenic Factor (Zearalenone)

Several reports have appeared in the literature of genital involvement in animals fed moldy grain (McNutt et aZ., 1928; Pullar and Lerew, 1937; Koen and Smith, 1945; McErlean, 1952; Stob et aZ., 1962; Christensen et al., 1965). The syndrome is characterized by vulvar hypertrophy, occasional vaginal eversion in females, preputial enlargement in castrated males, and prominent mammary glands in

201

MYCOTOXINS

both sexes. Stob et al. (1962) isolated Gibberella zea (Fusarium graminearurn) from toxic moldy corn and showed this microorganism to be responsible for the estrogenic syndrome in pigs. Similar effects were produced in mice fed corn inoculated with G. zea, thus confirming the toxicity of this microorganism previously reported by Christensen and Kernkamp (1936) and by Mitchell and Beadles ( 1940). Stob and his colleagues produced toxin on a variety of substrates, including corn, wheat, barley, and a synthetic medium composed of salts and carbohydrate. In addition to the estrogenic factor, a second water-soluble metabolite that was toxic to mice was isolated. Toxin was recovered by alcohol extraction followed by phase partition and chromatography on magnesium silicate. Purified toxin fluoresces blue-green under ultraviolet light. Andrews and Stob (1965) reported on the chemical isolation and structure of this toxin (Fig. 9) and found that partially purified concentrates of the active factor were effective in improving growth rate and feed efficiency in sheep.

c-0 HO

FIG.9. Structural formula of the estrogenic factor, F-2.

Christensen et al. (1965) also isolated the toxin fromG. zea-contaminated corn and named it F-2; it was partially characterized. In an extensive study Mirocha et al. (1967) reported on the production, isolation, chromatography, crystallization, and physical and chemical properties of F-2. They found F-2 to be identical with the compound reported by Andrews and Stob (1965). Urry et al. (1966) recently gave this substance the name zearalenone. Zearalenone has been chemically synthesized by Taub et al. (1967) and by Girotra and Wendler (1967). VII. Pink Rot Dermatitis

Pink rot is a disease of celery caused by the fungus Sclerotinia sclerotiorum. In the harvest of celery, a blistering cutaneous disorder often affects the exposed skin of workers coming in contact with diseased plants (Birmingham et al., 1961). Perone et al. (1963) showed that the skin on coming in contact with diseased portions of the plants and with subsequent exposure to sunlight (or ultraviolet radiation of

202


A

3200-3700 with a peak at 3650 A) produced the blistering lesion characteristically seen among celery harvesters. Scheel et al. (1963) subsequently isolated from celery infected with S . sclerotiorum two crystalline compounds identified as 4,5'8trimethylpsoralen and 8-methoxypsoralen (Fig. 10). Both of these cn 4,5:8 -trimethylpsoralen

8 -rnethoxypsorolen

FIG. 10. Structural formulas of psoralens from diseased celery.

compounds were bioassayed by the rabbit skin test, which on exposure to light gave a positive reaction. The compounds were not detected in nondiseased celery. Scheel and his colleagues, however, astutely noted that the phytotoxic psoralens were not necessarily produced by the fungus but may have been produced by the plant in response to the fungus. This type of response has been observed when sweet potatoes are invaded by a variety of fungi (Uritani, 1967).

VIII. Slaframine (Slobber Factor) Cattle, dairy cattle, and sheep eating second cutting red clover hay reportedly salivate excessively (O'Dell et al., 1959; Byers and Broquist, 1960, 1961). In addition to excess slobbering, animals went off feed, developed diarrhea, bloat, stiff joints, and sometimes died. Rainey et al. (1965) found that swine, chickens, guinea pigs, rats, and mice were also sensitive. The syndrome was found to be caused by forage infested with Rhizoctonia leguminicola, the fungus responsible for black-patch of red clover and other legumes (Elliott, 1952; Gough and Elliott, 1956; Smalley et al., 1962; Crump et al., 1963). Crump et al. (1963) concluded that the fungus and its products, not the red clover hay per se, caused the mycotoxicosis; the mycotoxin was termed slobber factor. Research on the isolation, purification, and identification of the slobber factor is centered at the Universities of Illinois and Wisconsin. Investigators at both schools published simultaneously on

MYCOTOXINS

203

the crystallization and tentative identification of the slobber factor as an alkaloid (Aust and Broquist, 1965; Rainey et al., 1965).The methods used by the two teams for the production, extraction, bioassay, chromatography, and crystallization of the toxic factor were different. Subsequently, Aust et al. (1966) proposed the following structure for the slobber factor which they named slaframine:

H,N

0 II 0-C-CH,

In uitro experiments showed that slaframine is neither a cholinesterase inhibitor nor a direct stimulator of cholinergic fibers; rather, it appears to hypersensitize smooth muscle preparations to acetylcholine. Atropine reverses its action in uivo and in uitro. IX. Yellow Rice Toxins In 1891, the possible toxicity of molds growing on rice was suspected by Sakaki; he attributed the cause of beriberi to toxic substances produced by these molds. Since then, the Japanese have intensively investigated toxicoses associated with moldy rice. Unfortunately, much of this work has been published in Japanese, making it unavailable to most Western scientists. A brief review on toxic moldy rice has been published in English by Kinosita and Shikata (1965) and Miyake and Saito (1965). Yellow rice results from the presence of a number of toxin-producing penicillia, including P . islandicum, P. citrinum, P . toxicarum, P . citreouiride, and P . rugulosum. Formulas of toxins produced by these fungi are shown in Fig. 11. P . islandicum produces two toxins, islanditoxin and luteoskyrin. Islanditoxin is a colorless, hydrophilic, powerful hepatotoxin causing rapid death with severe liver damage and hemorrhage. It is an unusual chlorine-containing cyclopeptide with a molecular formula, C25H3sOeN5C12. Dechlorination removes toxicity. The LDsofor 10-gm. mice ranges from 3.4 to 65.5 pg., depending on route of administration. It is toxic to Chang’s liver cell tissue culture. Luteoskyrin (C30H22012)is a yellow crystalline, lipophilic, relatively slow-acting anthraquinone derivative, 1,l-bis(2,4,5,8-tetrahydroxy-7-methyl-2,3-dihydroanthraquinone. In mice at high doses

204

Luteorkyrin

Citrinin HOOc&Ch

n

n Skyrin

Hw

Rugulrin

H3

0)\ T)$

d c \ t / < C it re ou v ir id i n : ,

C itreom y c e tin

FIG. 11. Structural formulas of mycotoxins found in toxic moldy rice.

MYCOTOXINS

205

it causes centrolobular necrosis and fatty metamorphosis of the remaining liver cells; it may cause hepatomas at lower dosage levels. The LD50 for 10-gm. mice ranges from 66.5 to 2210 pg., depending on the administration route. Extraction of the toxin from the mycelium also removes the pigments responsible for yellowing the rice: erythroskyrin, rubroskyrin, islandicin, catenarin, irridoskyrin, and skyrin. Penicillium toxicarium, P. ochrosalmoneum, and P . citreoviride produce a yellow fluorescent polyene, citreoviridin (Sakabe et al., 1964). This compound localizes in the central nervous system, adrenal cortex, liver, and kidneys where it causes paralysis and respiratory failure. Depending on route of administration, the minimum lethal dose for rats is 8-30 mg./kg. Citreoviridin is inactivated by ultraviolet light. Moldy rice by this fungus becomes nontoxic when exposed to the sun for 2 days. Penicillium citrinum has been found on rice in all parts of the world and occasionally on barley and dried fish. This microorganism produces citrinin, a nephrotoxin that is also elaborated by other penicillia and by Aspergillus terreus and A. candidus. Its LD50 in mice in 35 mg./kg. subcutaneously. Citrinin is alcohol soluble with a molecular formula of C13H1405. Citreomycetin, a related compound produced by several penicillia and Citromyces, causes chronic kidney damage on prolonged dosing. X. Stachybotryotoxicosis Stachybotryotoxicosis is a mycotoxicosis of horses resulting from ingestion of fodder contaminated with the fungus, Stachybotrys alternans (atra).Other animals affected by the mycotoxin are sheep, calves, swine, mice, guinea pigs, rabbits, chicks, and dogs. In horses, the disease occurs in two forms, typical (resulting from ingestion of sublethal amounts of toxin over a prolonged period of time) and atypical (caused by ingestion of large quantities of toxin and characterized by neurological disorders). The pathological picture of equine stachybotryotoxicosis is characterized by severe hemorrhage and necrosis in many tissues. Death often occurs in both forms of the disease. Man is sensitive to the mold or the purified toxin from the mold; a severe dermatitis occurs on contact. Naturally occurring cases of dermatitis have been reported among persons coming in contact with infected hay in regions where the equine disease occurs. Cases have also been reported where contaminated hay has been burned. The disease is characterized in man by a rash, particularly on body parts

206


that prespire the most. The rash frequently progresses to a moist dermatitis which later crusts over with dried serous exudate. There may be a severe pharyngitis, burning sensations in the nose, a bloody nasal exudate, and moderate to severe cough. The toxin can be extracted from contaminated hay by diethyl ether. It is soluble in many organic solvents and is stable to heat (120°C) and acids. It has been reported to have been crystallized and its structure determined by Russian investigators. Stachybotryotoxicosis is one of the earliest and most intensively investigated animal mycotoxicosis.The work was pioneered in Europe by the Russians and in this country by Forgacs and his colleagues. Pertinent references upon which this very brief review is based are: Drobotko (1945), Drobotko et al. (1946), Gajdusek (1953), Forgacs (1962,1965),Forjacs and Carl1 (1962),and Fojacs et al. (1958a). These references also furnish a good introduction to the Russian literature on stachybotryotoxicosis. XI. Rubratoxins

Burnside et al. (1957) reported that a disease of cattle and swine was caused by ingestion of corn molded with Penicillium rubrum. Pathological findings included congestion and hemorrhage in various organs and toxic cellular changes in the liver and kidneys. Forgacs et al. (1958b) fed chicks grain infected with P. rubrum and P. purpurogenum and obtained a hemorrhagic syndrome. The toxin could be extracted with water and was heat stable. Wilson and Wilson (1962a,b) studied the production, extraction, partial purification, and some chemical properties of the toxins isolated from P. rubrum. They obtained good yields of toxin (1gm./165 gm. substrate) after 20 days’ growth of P. rubrum on a corn-sucrose medium. Toxin could also be produced in Sabouraud and Czapek liquid medium and on moistened rice, oats, rye seed, and timothy hay. The toxins were extracted with methanol after a preliminary extraction with petroleum ether to remove lipoidal material. Toxin could be partially precipitated out of water by adjusting the pH to 1.8; two toxic fractions were recovered. Initial chemical and physical tests, although giving only limited information, established the presence of a carboxylic acid, as well as a hydroxyl and other groups. Administration of the toxin to mice, guinea pigs, rabbits, and dogs produced hepatotoxic and hemorrhagic syndromes. Townsend et al. (1966) obtained two toxins from P. rubrum by growing the mold in stationary culture on a Raulin-Thom medium

MYCOTOXINS

207

supplemented with 2.5%malt extract. The more easily obtained toxin, called rubratoxin A, had the following characteristics: m.p. 214°C (decomp.); [ a ] P 86.6; C, 59.6; H, 6.1; 0, 34.6%. The ultraviolet absorption spectrum in ethanol had a broad peak at 206-209 m p . and in alkali a peak at 262 mp., a shoulder at 300 mp., both of which disappeared on acidification. It has an LDS0of 3.5 mg./kg. given intraperitoneally in mice. Rubratoxin B was not so well characterized. Moss and co-workers (1967) later found rubratoxin B was the major constituent of the crude toxin. It recrystallized from ethyl acetatebenzene mixtures as fine needlelike laths with an m.p. 168"-170°C (decomp.); [a]do 67; molecular weight, 518; a probable formula of

+

+

C d L O i1. Structural studies are currently in progress at the Tropical Products Institute in England and at the Massachusetts Institute of Technology. XII. Other Mycotoxins Other toxin-producing molds have been found on a variety of foods and feeds, but investigations of them have been limited. Fusarium roseum infects barley, wheat, and rice and reportedly causes poisoning in animals and man; the syndrome includes nausea, vomiting, and even death. Fusarium niuale-infected cereals cause nausea and vomiting in man (Tsunoda, 1963). Kinosita and Shikata (1965) reported that a Rhizopus sp. (probably R. oryzae) produced toxins in soybean tempeh cakes that were lethal to mice. Repeated injections of patulin (produced by Aspergillus clavatus and A. giganteus) into rats caused sarcomas at the site of injection (Dickens and Jones, 1961). A. fumigatus produces the toxic metabolites, fumagillin, gliotoxin, and helvolic acid (Spector, 1957). A. oyzae, grown on media containing malt extract, produces the mycotoxin, maltoryzine; this compound causes muscular paralysis (Iizuka and Iida, 1962). A. candidus, which is sometimes used b y the Japanese for baking, produces the toxins, citrinin, kojic acid, and candidulin (Spector,

1957). In Japan, a culture of A. sydowi was found in contaminated glucoseRinger's solution that had been given patients who became feverish, with some deaths resulting (Wooley et al., 1938; Kobayashi et al.,

1951).

208


Scott (1965) in South Africa reported that 46 strains of 22 species of molds were toxigenic based on the duckling test; these strains included cultures of Aspergillus, Fusarium, Paecilomyces, Penicillium, and Trichotheceum. Diplodia zea on corn has been recorded as being toxic to cattle and sheep in Africa, causing salivation, incoordination, and occasionally death (Watt and Breyer-Brandwijk, 1962). Penicillium decumbens produces a dihydroxy-a,p-unsaturated lactone, decumbin, which laboratory tests have shown to be toxic in rats. (Singleton et al., 1958). Wilson et al. (1967) failed to find aflatoxin in a culture of P . puberulum but isolated a new mycotoxin when the culture was grown on corn, wheat, peanut, millet, or oats. The substance was extracted with methanol or acetone and fluoresced bright green under ultraviolet light. Administration of nonlethal quantities of crude extract to mice and ducklings caused incoordinated locomotion with apparent stiffness and exaggerated movements of the limbs; albino mice exhibited darkening of eye color with cyanotic coloring of the mouth, feet, and tail. 6-Methoxy-melleinYoriginally thought to be produced by carrots in response to invasion by various fungi, has been produced by submerged cultures of Sporomia bipartis (Mauli and Sigg, 1966). This isocoumarin, although not known to be a mycotoxin, is related to ramulosin (Benjamin and Stodola, 1960) and possesses the lactone structure common to many toxic compounds. P-Nitropropionic acid proved to be the toxic principle of the legume Zndigofera endecaphylla, which when introduced into Hawaii and Latin America as a forage and cover crop produced severe toxic symptoms in dairy cattle (Morris et al., 1954). Since then it has been found to be produced by the ubiquitous mold, Aspergillus oryzae (Nakamura and Shimoda, 1954), Penicillium atrovenetum (Raistrick and Stossl, 1958),and A. avenaceus (Brookes et al., 1963). Tall fescue is a valuable forage grass in the southern and western areas of the United States. Cattle grazing on fescue pasture are sometimes affected with “fescue foot,” a noninfectious disease characterized by lameness of the hind quarters and a predisposition to dry gangrene of the extremities (Yates, 1962, 1963; Jacobson et al., 1963; Yates and Tookey, 1965). The disease has the characteristics of a mycotoxicoses, but the causative microorganism has not yet been isolated with certainty. A culture of Fusarium nivale found on tall fescue hay from a pasture where fescue toxicity had been reported

MYCOTOXINS

209

produced a toxic butenolide characterized as 4-actamido-4-hydroxy2-butenoic acid-y-lactone (Yates et al., 1967):

R

HC=CH / \ O=C CH-NH-C-CH, ‘O/

The pharmacological properties of this compound are under investigation to determine its role, if any, in fescue toxicity. Extracts of the fungus, Fusarium tricinctum, isolated from moldy corn toxic to cattle, produced severe edema and intradermal hemorrhage when applied to rat skin. A crystalline material, CI7Hz4O6, was isolated that appeared to be a hydroxy ester or lactone (Gilgan et al., 1966). Subsequent investigation indicated that the compound was 4-15-diacetoxy-scirpen-3-01, C 1YH2607, which is also produced by F. equiseti (Brian et al., 1961; Dawkins, 1966). A second toxin was isolated from another strain of F. tricinctum that was similar to but not identical with, diacetoxyscirpenol (Marasas et al., 1967). This second toxin was lethal to trout with an approximate L D ~ of o 6.1 mg./kg. Pathological symptoms included a shedding of the intestinal mucosa and severe edema with fluid in the body cavities and behind the eyes. A strain of F. culmorum was isolated from moldy corn that had caused loss of appetite, marked decrease in milk production, and symptoms of staggering in dairy cows (Fisher et al., 1967). Extracts of corn on which F. culmorum were grown caused a severe reaction in the rabbit skin test with death ensuing between the fifth and seventh day of the test. Toxic antimetabolites are occasionally produced by plants and raw foodstuffs in response to invasion by various fungi. The current literature in this area of toxicoses has been reviewed by Uritani (1967).

XIII. Summary Obviously, it is impossible to condense the salient features of all mycotoxins discussed in a review. However, several points deserve special comment. In general, it appears that many mycotoxicoses arise from improper harvesting, handling, and storage of various foods and feeds. More research on preventive measures is required. Studies in depth have been carried out on only a limited number of mycotoxins. Undoubtedly, many diseases of unknown etiology in both man and animal have been caused by mycotoxins, pointing up the need

210


for better methods of detection and analysis. Since mycotoxins mainly are produced in minute quantities and may be unevenly distributed, adequate sampling of suspect commodities presents a real problem. A rapidly increasing world population demands that methods, other than condemnation and destruction, be found for practical and lowcost detoxification of contaminated foods and feeds. The multiplicity and complexity of the problems associated with mycotoxins call for a multidisciplinary approach. REFERENCES Allcroft, R. (1965). In “Mycotoxins in Foodstuffs” (G. N. Wogan, ed.), pp. 153-162. M.I.T. Press, Cambridge, Massachusetts. Allcroft, R.,and Carnaghan, R. B. A. (1962). Vet. Record 74,863-864. Allcroft, R., and Carnaghan, R. B. A. (1963). Vet. Record 75,259-263. Allcroft, R., Rogers, H., Lewis, G., Nabney, J., and Best, P. E. (1966). Nature209, 154- 155. Andrellos, P. J., and Reid, G. R. (1964).J . Assoc. Ofic. Agr. Chemists 47,801-803. Andrellos, P. J., Beckwith, A. C., and Eppley, R. M. (1967). J . Assoc. Ofic. Anal. Chemists 50, 346-350. Andrews, F. N., and Stob, M. (1965).U.S. Patent 3,196,019. Anonymous (1966).J . Assoc. Ofic. Anal Chem. 49,229-231. Arai, T., Tatsuya, I., and Koyama, Y. (1967).J . Bacteriol. 93, 59-64. Armbrecht, R. H., and Fitzhugh, 0. G. (1964). Toricol. Appl. Pharmacol. 6,421-426. Armsbong, D. J. (1966).Proc. Natl. Acad. Sci. US.56,64-66. Arrhenius, E., and Hultin, T. (1962).CancerRes. 22,823-834. Asao, T., Biichi, G., Abdel Kader, M. M., Chang, S. B., Wick, E. L., and Wogan, G. N. (1963).J.Am. Chem. S O C . 85,1706-1707. Asao, T., Biichi, G., Abdel Kader, M. M., Chang, S. B., Wick, E. L., and Wogan, G. N. (1965)J.Am. Chem. Soc. 87,882-886. Ashworth, L. J., Schroeder, H. W., and Langley, B. C. (1965). Science 148,1228-1229. Asplin, F. D., and Carnaghan, R. B. A. (1961). Vet. Record73, 1215-1219. Aust, S. D., and Broquist, H. P. (1965).Nature 205,204. Aust, S. D., Broquist, H. P., and Rinehart, K. L. (1966). J . Am. Chem. SOC. 88, 28792880. Austwick, P. K., and Ayerst, G. (1963).Chem. Znd. (London) pp. 55-61. Ayres, J. L., and Sinnhuber, R. 0.(1966).J.Am. Oil Chemists SOC. 43,423-424. Barnes, D. (1966).Food Technol. 20, 755-756. Barnes, J. M., and Butler, W. H. (1964). Nature 202, 1016. Bassir, 0. (1964). W. African]. B i d . Appl. Chem. 8, 5-15. Bassir, O., and Osiyemi, F. (1967).Nature 215,882. Bekker, Z. E. (1963). “Physiology of Fungi and Its Application in Practice,” p. 268. Moscow Univ. (In Russ.) Benjamin, C. R., and Stodola, F. H. (1960). Nature 188,662-663. Birmingham, D. J., Key, M., Tubich, G. E., and Perone, V. B. (1961). Arch. Dermatol. 83, 73. Black, H. S., and Altschul, A. M. (1965).Biochem. Biophys. Res. Commun. 19,661-664.

MYCOTOXINS

211

Black, H. S., and Jirgensons, B. (1967). Plant Physiol. 42, 731-735. Blount, W. P. (1961). Turkeys 9,52, 55-58,61-71. Borker, E., Insalata, N. F., Levi, C. P., and Witzeman, J. S. (1966).Aduan. Appl. Microbiol. 8, 315-351. Brian, P. W., Dawkins, A. W., Grove, J. F., Hemming, H. G., Lowe, P., and Norris, G. L. F. (1961).J . Exptl. Botany 12, 1-12. Brook, P. J. (1966). Ann. Reo. Phytopathol. 4, 171-194. Brookes, D., Tidd, B. K., and Turner, W. B. (1963).J . Chem. SOC. pp. 5385-5391. Brown, J. M. M. (1965). Symp. Mycotoxins Foodstuffs, Agr. Aspects, Unio. Pretoria, February pp. 33-46. Ind. Res. Develop. Div., Council Sci. Ind. Res., Dept. Agr. Tech. Serv., Pretoria. Brown, J. M. M.. and Abrams, L. (1965).OnderstepoortJ. Vet. Res. 32, 119-145. Biichi, G., Foulkes, D. M., Kurono, M., and Mitchell, G. F. (1966).J . Am. Chem. SOC. 88,4534-4536. Buetow, D. E., and Mego, J. L. (1967). Biochim. Biophys. Acta 134,395-401. Burmeister, H. R., and Hesseltine, C. W. (1966). Appl. Microbiol. 14, 403-404. Burmeister, H. R., and Hesseltine, C. W. (1968).J.Bacteriol. (in press). Burnside, J. E., Sippel, W. L., Forgacs, J., Carll, W. T., Atwood, M. B., and Doll, E. R. (1957). Am. J . Vet. Res. 18, 817-824. Butler, W. H. (1964).J . Pathol. Bacteriol. 88, 189-196. Butler, W. H. (1965). In “Mycotoxins in Foodstuffs” (G. N. Wogan, ed.), p. 175. M.I.T. Press, Cambridge, Massachusetts. Butler, W. H. (1966).Am. 1.Pathol. 49, 113-128. Butler, W. H., and Barnes, J. M. (1966). Nature 209,90. Butler, W. H., and Clifford, J. I. (1965).Nature 206, 1045. Butler, W. H., and Wigglesworth, J. S. (1966). Brit. J . Erptl. Pathol. 47, 242-247. Buu-Hoi, N. P., Mangane, M., and Jacquignon, P. (1966)J. Chem. SOC. pp. 50-52. Byers, J. H., and Broquist, H:P. (1960).J . Dairy Sci. 43,873-874. Byers, J. H., and Broquist, H. P. (1961).J. Dairy Sci. 44,1179. Campbell, A. D. (1967).J . Assoc. Ofic. Anal. Chemists 50,343-346. Campbell, A. D., and Funkhouser, J. T. (1966). J . Assoc. Ofic. Anal. Chemists 49, 730-739. Campbell, A. D., Dorsey, E., and Eppley, R. M. (1964).J . Assoc. Ofic. Agr. Chemists 47,1002-1003. Carnaghan, R. B. A. (1964). Proc. Roy. SOC. Med. 57,414. Camaghan, R. B. A., and Allcroft, R. (1962).Vet. Record 74,925-926. Camaghan, R. B. A., Hartley, R. D., and O’Kelly, J. (1963). Nature 200, 1101. Chang, S. B., Abdel Kader, M. M., Wick, E. L., and Wogan, G. N. (1963).Science 142, 1191-1 192. Chen, S.-C., and Friedman, L. (1966).J.Assoc. Ofic. Anal. Chemists49, 28-33. Cheung, K. K., and Sim, G. A. (1964). Nature 201, 1185-1188. Childs, V. A., and Legator, M. S. (1966).Life Sci. 5, 1053-1056. Christensen, C. M., Nelson, G. H., and Mirocha, C. J. (1965).Appl. Microbiol. 13, 653-659. Christensen, J. J., and Kernkamp, H. C. H. (1936). Minn. Unio. Agr. E r p t . Sta., Tech. Bull. 113. Ciegler, A., and Peterson, R. E. (1967). Bacteriol. Proc. p. 17. Ciegler, A., Lillehoj, E. B., Peterson, R. E., and Hall, H. H. (1966a).Appl. Microbiol. 14,934-939.

212


Ciegler, A., Peterson, R. E., Lagoda, A. A., and Hall, H. H. (1966b).Appl. Microbiol. 14,826-833. Clare, N. T. (1944). New Zealand]. Sci. Technol. A25,202-220. Clifford, J. I., and Rees, K. R. (1966). Nature209, 312-313. Clifford, J. I., and Rees, K. R. (1967a). Biochem. J. 102,65-75. Clifford, J. I., and Rees, K. R. (1967b). Biochem. J. 103, 467-471. Clifford, J. I., Rees, K. R.,and Stevens, E. M. (1967). Biochem. J . 103, 258-261. Codner, R. C., Sargeant, K., and Yeo, R. (1963). Biotechnol. Bioeng. 5, 185-192. Colombo, G., Levis, A. G., and Torlone, V. (1965). Intern. Symp. Radiosensitizers Radioprotect. Drugs, l s t , Milan, 1964 pp. 392-399. Coomes, T. J., and Feuell, A. J. (1965).Trop. Prod. Inst. Rept. G 13. Trop. Prod. Inst., London. Coomes, T. J., and Sanders, J. C. (1963). Analyst 88,209-213. Coomes, T. J., Crowther, P. C., Francis, B. J., and Shone, G. (1964).Analyst89,436437. Coomes, T. J., Crowther, P. C., Feuell, A. J., and Francis, B. J. (1966). Nature 209, 406-407. Cox, E. C., White, J. R.,and Flaks, J. G. (1964). Proc. Natl. Acad. Sci. U.S.51,703-709. Crump, M. H., Smalley, E. B., Henning, J. N., and Nichols, R. E. (1963).J . Am. Vet. Med. Assoc. 143,996-997. Cunningham, I. J., Hopkirk, C. S. M., and Filmer, J. F. (1942). New Zealand J. Sci. Technol. A24, 185. Curry, J., and Greenberg, J. (1962).J. Bacteriol. 83, 38-42. Davis, N. D., Diener, U. L., and Eldridge, D. W. (1966a).Appl. Microbiol. 14,378-380. Davis, N. D., Hayes, A. W., Eldridge, D. W., and Diener, U. L. (1966b).J.Assoc. Ofic. Anal. Chemists 49, 1224-1255. Dawkins, A. W. (1966).J.Chem. SOC.(C)pp. 116-123. De Deken-Grenson, M . (1955). Biochim. Biophys. Acta 17,35-47. De Greef, J. A. (1964). Enzymologia 27,311-325. De Iongh, H., Vles, R. O., and van Pelt, J. C. (1964). Nature 202,466-467. Denault, L. J., and Underkofler, L. A. (1967).Cereal Chem. 44, 1-7. De Recondo, A. M., Frayssinet, C., Lafarge, C., and Le Breton, E. (1965). Compt. Rend. 261,1409-1412. De Recondo, A. M., Frayssinet, C., Lafarge, C., and Le Breton, E. (1966). Biochim. Biophys. Acta 119,322-330. Derzsy, D., Meszaros, J., Prokopovitsch, L., and Toth-Baranyi, I. (1961).Magy. Allutoru. Lapja 17,49. Dickens, F. (1964). Brit. Med. Bull. 20, 96-101. Dickens, F., and Cooke, J. (1965). Brtt. J. Cancer 19,404-410. Dickens, F., and Jones, H. E. H. (1961). Brit. J . Cancer 15, 85-100. Dickens, F., and Jones, H. E. H. (1963). Brit. J. Cancer 17,691-698. Dickens, F., and Jones, H. E. H. (1965). Brit. J. Cancer 19,392-403. Dickens, F., Jones, H. E. H., and Waynforth, H. B. (1966).Brit. J . Cancer 20,134-144. Dingman, C. W., and Sporn, M. B. (1965).Science 149,1251-1254. DiPaolo, J. A., Elis, J., and Erwin, H. (1967). Nature 215,638-639. Dollear, F. G., and Gardner, H. K., Jr. (1966). Proc. 4th Natl. Peanut Res. Con$. Tifton, Georgia pp. 72-78. Done, J., Mortinier, P. H., and Taylor, A. (1960).Res. Vet. Sci. 1, 76.

MYCOTOXINS

2 13

Drobotko, V. G. (1945). Ann. Reo. Sooiet Med. 2,238-242. Drobotko, V . G., Marushenko, P. E., Aizeman, B. E., Kolesnik, N. G., Kudlai, D. B., Iatel, P. D., and Melnichenko, V. D. (1946). Vrachebnoe Delo 26, 125-128. Dutton, M. F., and Heathcote, J. G . (1966). Biochem. J . 101, 21-22. Ebringer, L., Krkoska, R., Macor, M., Jurasek, A., and Kada, R. (1967). Arch. Mikrobiol. 57,61-67. Edelman, M., Schiff, J. A., and Epstein, H. T. (1965). J . Mol. Biol. 11, 769-774. Elis, J., and DiPaolo, J. A. (1967). Arch. Pathol. 83,53-57. Elliott, E . S. (1952). West Va. Uniu. Agr. Erpt. Stu. Bull. 351T, 47-60. Endo, H., Ishizawa, M., Kamiya, T., and Kuwano, M. (1963). Biochim. Biophys. Acta 68,502-505. Engebrecht, R. H., Ayres, J . L., and Sinnhuber, R. 0. (1965). J . Assoc. Ofic. Agr. Chemists 48, 815-818. Eppley, R. M. (1966a).J. Assoc. Ofic. Anal. Chemists 49,473-474. Eppley, R. M. (1966b). J . Assoc. Ofic. Anal. Chemists 49,1218-1223. Falaschi, A., and Kornberg, A. (1964). Federution Proc. 23,940-945. Falk, H. L., Thompson, S . J., and Kotin, P. (1965). Am. Assoc. Cancer Res. Proc. 6, 18. Fastier, L. B. (1961). New Zenland J . Agr. Res. 4, 72-82. Ferreira, N. P. (1966). A m . Chem. Soc., Program Abstr., 152nd Meeting, New York p. A-17. Feuell, A. J. (1966a). Trop. Sci. 8, 61-70. Feuell, A. J. (196613).Can. Med. Assoc. J . 94,574-581. Fishbach, H., and Campbell, A. D. (1965). J . Assoc. Ofic. Anal. Chemists 48,128. Fisher, E. E., Kellock, A. W., and Wellington, N. A. M. (1967).Nature 215,322. Forgacs, J. (1962). Feedstufs 34, 124-134. Forgacs, J. (1965). in “Mycotoxins in Foodstuffs” ( G . N. Wogan, ed.), pp. 87-103. M.I.T. Press, Cambridge, Massachusetts. Forgacs, J., and Carll, W. T. (1962). Adoan. Vet. Sci. 7, 273-382. Forgacs, J., Carll, W. T., Herring, A. S., and Hinshaw, W. R. (1958a).Truns. N.Y. Acad. Sci. 20, 787-808. Forgacs, J., Koch, H., Carll, W. T., and White-Stevens, R. H. (1958b).Am. J . Vet. Res. 19,744-753. Fowlks, W. L., Griffith, D. G., and Oginsky, E. L. (1958).Nature 181,571-572. Fay, H., Gillman, T., Kondi, A., and Preston, J. K. (1966). Nature 212,150-153. Fridrichsons, J., and Mathieson, A. McL. (1962). Tetruhedron Letters 26,1265-1268. Gabliks, J., Schaeffer, W., Friedman, L., and Wogan, G . (1965).J. Bacteriol. 90,720723. Gajan, R. J., Nesheim, S., and Campbell, A. D. (1964). J . Assoc. Ofic. Agr. Chemists 47,27-28. Gajdusek, D. C. (1953). Med. Sci. Puhl., Army Med. Seru. Grud. School, Walter Reed Army Med. Center 2, 107-111. Gallagher, C. H . (1964). Nature 201, 1293-1294. Gelboin, H. V., Wortham, J. S., Wilson, R. G., Friedman, M., and Wogan, G . N. (1966). Science 154, 1205-1206. Gilgan, M. W., Snialley, E. B., and Strong, F. M. (1966).Arch. Biochem. Biophys. 114, 1-3. Girotra, N . N., and Wendler, N. L. (1967). Chem. Ind. (London) p. 1493. Goldberg, I. H., and Reich, E. (1964). Federation Proc. 23,958-964.

2 14


Goldblatt, L. A. (1966). Aduan. Chem. Ser.57,216-227. Goldschmidt, B. M. (1965). In “Mycotoxins in Foodstuffs” (G. N. Wogan, ed.), p. 287. M.I.T. Press, Cambridge, Massachusetts. Gough, F. J., and Elliott, E. S. (1956). West Va. Uniu. Agr. E x p t . Sta. Bull. 387T. Greengard, O., Smith, M. A,, and Acs, G. (1963).J . Biol. Chem. 238, 1548-1551. Gudauskas, R. T., Davis, N. D., and Diener, U. L. (1967).J . Inuertebrate Pathol. 9, 132- 133. Halver, J. E. (1962). Proc. Working Con5 Trout Hepatoma, 1961. Dept. Health Educ. Welfare Rept., Natl. Inst. Health, Bethesda, Maryland. Halver, J. E. (1965). In “Mycotoxins in Foodstuffs” (G. N. Wogan, ed.), pp. 209-234. M.I.T. Press, Cambridge, Massachusetts. Halver, J. E., Johnson, C. L., and Ashley, L. M. (1962).Federation Proc. 21,390. Harding, J. D. J., Done, J. T., Lewis, G., and Allcroft, R. (1963). Res. Vet. Sci. 4, 217229. Harold, F. M., and Ziporin, Z. Z. (1958). Biochim Biophys. Acta 28, 482-503. Hartman, G., Coy, U., and Kniese, G. (1963). 2. Physiol. Chem. 330,227-233. Hawtrey, A. O., and Schiiren, V. (1962). Biochim Biophys. Acta 55, 785-786. Hayes, A. W., Davis, N. D.. and Diener, U. L. (1966).Appl. Microbiol. 14,1019-1021. Hesseltine, C. W. (1967). fi;.alth Lab. Sci. 4,222-228. Hesseltine, C. W., Shotwell, 0. L., Ellis, J. J,, and Stubblefield, R. D. (1966).Bacteriol. Rev. 30,795-805. Hesseltine, C. W., Shotwell, 0. L., Smith, M. L., Shannon, G. M., Vandegraft, E. E., and Goulden, M. L. (1967). Bacteriol. Proc. p. 17. Hnilica, L. S., and Billen, D. (1964). Biochim. Biophys. Acta 91,271-280. Hodges, F. A., Zust, J. R., Smith, H. R., Nelson, A. A., Armbrecht, B. H.,andCanipbell, A. D. (1964).Science 145, 1439. Hodges, R., Ronaldson, J. W., Taylor, A., and White, E. P. (1963).Chem. Ind. (London) PI].42-43. Hohapfel, C. W., Steyn, P. S., and Purchase, I. F. H. (1966). Tetrahedron Letters25, 2799-2803. Hore, D. E. (1960). Australian Vet.J . 36,172. Iizuka, H., and Iida, M. (1962).Nature 196, 681-682. Jacobson, D. R., Miller, W. M., Seath, D. M., Yates, S. G . , Tookey, H. L., and Wolff, I. A. (1963).J . Dairy Sci. 46,416-422. Jcffe, A. Z. (1960a). Bull. Res. Council Israel D8, 81-95. Joffe, A. Z. (1960b). Bull. Res. Council Israel D9, 101-126. Joffe, A. Z. (1962).Mycopathol. Mycol. Appl. 16,201-221. Joffe, A. Z. (1963). Plant Soil 18, 31-44. Joffe, A. Z. (1965).In “Mycotoxin in Foodstuffs” (G. N. Wogan, ed.), pp. 77-85. M.I.T. Press, cambridge, Massachusetts. Jones, H. C., Black, H. S., and Altschul, A. M. (1967). Nature214, 171-172. Juhasz, S., and Greczi, E. (1964). Nature203,861-862. Kato, K., Sugino, Y., and Endo, H. (1966). Biochim. Biophys. Acta 119,309-321. Kilgore, W. W., and Greenberg, J. (1961).J . Bacteriol. 81, 258-265. King, A. M. Q., and Nicholson, B. H. (1967). Biochem. J . 104,69~-70p. Kinosita, R., and Shikata, T. (1965).In “Mycotoxins in Foodstuffs” (G. N. Wogan, ed.), pp. 111-131. M.I.T. Press, Cambridge, Massachusetts. Kirk, J. T. 0. (1964). Biochem. Biophys. Res. Commun. 14,373-377. Knypl, J . S. (1963).Nature 200,800-802.

MYCOTOXINS

215

Knypl, J. S. (1967). Planta 72, 292-296. Kobayashi, Y.,Uraguchi, K., Sakai, F., and Miyaki, I. (1951).Japan J. Pharmacol. 1, 1. Koen, J. S., and Smith, H. C. (1945). Vet. Med. 40,131. Kohler, H., and Swaboda, R. (1962).Wien. Tieraerztl. Monatsschr. 49,205. Kraybill, H. F., and Shimkin, M. B. (1964). Aduan. Cancer Res. 8, 191-248. Kulik, M. M., and Holaday, C. E. (1966). Mycopathol. Mycol. Appl. 30(2), 137-140. Lacassagne, A., Buu-Hoi, N. P., Zajdela, F., Jacquigmon, P., and Mangane, M. (1967). Science 158,387-388. Lancaster, M. C., Jenkins, F. P., and Philp, J. M. (1961).Nature 192, 1095-1096. Lee, W. V. (1965).Analyst 90,305-307. Legator, M. S. (1966). Bacteriol. Rev. 30,471-477. Legator, M. S., and Withrow, A. (1964).J.Assoc. O ~ CAgr. . Chemists 47,1007-1009. Legator, M. S., Zuffante, S. M., and Harp, A. R. (1965).Nature 208,345-347. Lerman, L. S. (1963). Proc. Natl. Acad. Sci. U S . 49, 94-102. Liau, M. C., Hnilica, L. S., and Hurlbert, R. B. (1965). Proc. Natl. Acad. Sci. U.S. 53, 626-633. Lie, L. L., and Marth, E. H. (1967).J.Dairy Sci. 50,1708-1710. Lillehoj, E. B., and Ciegler, A. (1967).J. Bacteriol. 94,787-788. Lillehoj, E. B., and Ciegler, A. (1968), Mycopathol. Mycol. Appl. (in press). Lillehoj, E. B., Ciegler, A., and Hall, H. H. (1967a).Can. J. Microbiol. 13, 629-633. Lillehoj, E. B., Ciegler, A., and Hall, H. H. (1967b).J. Bacteriol. 93, 464-471. Lillehoj, E. B., Ciegler, A., and Hall, H. H. (1967~). Experientia 23,187-188. Lilly, L. J. (1965).Nature 207,433-434. Linnane, A. W., and Stewart, P. R. (1967).Biochem. Biophys. Res. Commun. 27,l-6. Loosmore, R. M., Allcroft, R., Tuttan, E. A,, and Carnaghan, R. B. A. (1964).Vet.Record 76,64-65. McCalla, D. R. (1964). Can.J. Microbiol. 42, 1245-1247. McCalla, D. R. (1965).J. Protozool. 12, 34-41. McCalla, T. M., and Haskins, F. A. (1964).Bacteriol. Reu. 28,181-207. McErlean, B. A. (1952).Vet Record 64,539-540. MacKirinon, M. M., and T e Punga, W. A. (1961). New Zealand]. Agr. Res. 4, 141-146. McNutt, S. H., Punvin, P., and Murray, C. (1928).J. Am. Vet. Med. Assoc. 73,484. Madhaven, T. V., and Rae, K. S. (1966). Arch. Pathol. 81, 520-524. Madhaven, T. V., and Rao, K. S. (1967).J.Pathol. Bacteriol. 93, 329-331. Madhaven, T. V., Rao, K. S., and Tulpule, P. G. (1965a). ZndianJ. Med. Res. 53, 985990. Madhaven, T. V., Tulpule, P. G., and Gopalan, C. (1965b). Arch. Pathol. 5, 466-469. Mann, G. E., Godifer, L. P., Jr., and Dollear, F. G . (1968).J.Agr. Food Chem. (in press). Marasas, W. F. O., Smalley, E. B., Degurse, P. E., Bamburg, J. R., and Nichols, R. E. (1967).Nature 214,817-818. Martin, J. T., Baker, E. A., and Byrde, R. J. W. (1966).Ann. Appl. Biol. 57,501-508. Masri, M. S. (1965). Western Expt. Sta. Collaborators’ Con5 Importance Mold Metabolites Agr. Prod., Albany, California Abstr. pp. 9-10. U.S. Dept. Agr., ARS, Western Util. Res. Develop. Div. Masri, M. S., Lundin, R. E., Page, J. R., and Garcia, V. C. (1967). Nature 215,753-755. Mateles, R. I., and Adye, J. C. (1965).Appl.Microbiol. 13,208-211. Mateles, R. I., and Wogan, G. N. (1967a). In “Biochemistry of Some Foodborne Microbial Toxins” (R. I. Mateles and G. N. Wogan, eds.). M.I.T. Press, Cambridge, Massachusetts .

216


Mateles, R. I., and Woyan, G. N. (1967b).Aduan. Microbiol. Physiol. 1,l-31. Mathews, M. M. (1963).J. Bacteriol. 85, 322-328. Matsumura, F., and Knight, S. G. (1967).J. Econ. Entomol. 60, 871-872. Mauli, R. A. R., and Sigg, H. (1966). Experientia 15, 575. Mayer, A. M., and Poljakoff-Mayber, A. (1961). In “Plant Growth Regulation” (R. M. Klein, ed.), p. 735. Iowa State Univ. Press, Ames, Iowa. Mayer, C. F. (1953a). Military Surgeon 113, 173-189. Mayer, C. F. (1953b). Military Surgeon 113, 295-315. Mirocha, C. J., Christensen, C. M., and Nelson, G. H. (1967). Appl. Microbiol. 15, 497-503. Mitchell, H. H., and Beadles, J. R. (1940).J.Agr. Res. 61, 135. Miyake, M., and Saito, M. (1965). In “Mycotoxins in Foodstuffs” (G. N. Wogan, ed.), pp. 133-146. M.I.T. Press, Cambridge, Massachusetts. Morris, M. P., Pagan, C., and Wannke, H. E. (1954). Science 119, 322-323. Mortimer, P. H., and Taylor, A. (1962). Res. Vet. Sci. 3, 147-160. Moses, V., and Calvin, M. (1965).J . Bacteriol. 90, 1205-1217. Moss, M. O., Robinson, F. V., Wood, A. B., and Morrison, A. (1967).Chem. Ind. (London) pp. 755-757. Musajo, L., and Rodighiero, G. (1962).Experientia 18, 153-161. Musajo, L., Rodighiero, G., Breccia, A,, Dall’Acqua, F., and Malesani, G. (1966). Experientia 22, 75. Nabney, J., and Nesbitt, B. F. (1965).Analyst 90, 155-160. Nabney, J., Burhage, M. B., Allcroft, R., and Lewis, G . (1967).Food Cosmet. Toxicol. 5,ll-17. Nakamiira, S., and Shimoda, C. (1954). Nippon Nogei Kayaku Kaishi 28,909. Nartey, F. (1966).Physiol. Plantarum 19,818-822. Nel, W. (1967). Cited in Scott and Hand (1967), 11. 367. Nesheim, S. (1964).J . Assoc. Ofic. Agr. Chemists 47,1010-1017. Nesheim, S. (1967).J . Assoc. Ofic. Anal. Chemists 50,370-371. Nesheim, S., Banes, D., Stoloff, L., and Campbell, A. D. (1964).J.Assoc. Ofic. Agr. Chemists 47,586. Newberne, P. M. (1965). In “Mycotoxins in Foodstuffs” (G. N. Wogan, ed.), pp. 187208. M.I.T. Press, Cambridge, Massachusetts. Newberne, P. M., Carlton, W. W., and Wogan, G. N. (1964). Pathol. Vet. (Basel) 1, 105-132. Newberne, P. M., Wogan, G. N., and Hall, A. (1966a).J . Nutr. 90,123-130. Newberne, P. M., Russo, R., and Wogan, G. N. (196613).Pathol. Vet. (Basel) 3,331-340. O’Dell. B. L., Regan, W. O., and Beach, T. J . (1959).Missouri Uniu. Agr. Expt. Sta., Res. Bull. 702, 1. Parker, W. A., and Melnick, D. (1966).J . Am. Oil Chemists’ Soc. 43,635-638. Percival, J . C., and Thornton, R. H. (1958). Nature 182, 1095-1096. Perone, V. B., Scheel, L. D., and Meitus, R. J. (1963).J.Inuest. Dermatol. 42,267-272. Peters, J. A. (1963). Nature 200, 286. Peters, J. A. (1966). Nature 210, 601-603. Peters, J. A., and Smith, L. M. (1964). Biochem. J.92,379-385. Peterson, R. E., and Ciegler, A. (1967).J . Assoc. Ofic. Agr. Chemists 50, 1201-1202. Peterson, R. E., Ciegler, A., and Hall, H. H. (1967).J . Chromatog. 27, 304-307. Philips, F. S., Schwartz, H. S., Sternberg, S. S., and Tan, C. T. C. (1960).Ann. N.Y.Acad. Sci. 89, 348-360. Platt, B. S., Stewart, R. J. C., and Gupta, R. (1962). Proc. Nutr. SOC. (Engl. Scot.)30,21.

MYCOTOXINS

217

Pomeranz, Y. (1964).Cereal Sci. Today 9,93. Pons, W. A., Jr., and Eaves, P. H. (1967).J . Am. Oil Chemists’ SOC. 44, 139A. (Abstr. 115.) Pons, W. A., Jr., Cucullu, A. F., Lee, L. S., Robertson, J. A., Franz, A. O., and Goldblatt, L. A. (1966).]. Assoc. Ofic. Anal. Chemists 49,554-562. Provasoli, L., Hunter, S. H., and Schatz, A. (1948). Proc. SOC. Exptl. Biol. Med. 69, 279-282. Pullar, E. M., and Lerew, W. M. (1937).Australian Vet. J . 13,28. Purchase, I. F. H. (1967). Food Cosmet. Toxicol. 5, 339-342. Purchase, I. F. H., and Steyn, M. (1967).J . Assoc. Ofic. Anal. Chemists 50,363-366. Rainey, D. P., Smalley, E. B., Crump, M. H., and Strong, F. M. (1965). Nature205, 203-204. Raistrick, H., and Stossl, A. (1958). Biochem. J . 68, 647-653. Robertson, J. A., Jr., Lee, L. S., Cuculla, A. F., and Goldblatt, L. A. (1965).J . Am. Oil Chemists’ SOC. 42, 467-471. Robertson, J. A., Jr., Pons, W. A., and Goldblatt, L. A. (1967).J . Agr. Food Chem. 15, 798-801. Rogers, A. E., and Newberne, P. M. (1967).Cancer Res. 27,855-864. Ronaldson, J. W., Taylor, A., White, E. P., and Abraham, R. J. (1963).J. Chem. Soc. pp. 3172-3180. Sakabe, N., Goto, T., and Hirata, Y.(1964).Tetrahedron Letters 27,1825-1830. Sakaki, J. (1891). 2. Tokio Ges. 5, 1097. Sargeant, K., O’Kelly, J., Carnaghan, R. B. A., and Allcroft, R. (1961a). Vet. Record73, 1219-1223. Sargeant, K., Sheridan, A,, O’Kelly, J., and Carnaghan, R. B. A. (1961b). Nature192, 1096-1097. Scheel, L. D., Perone, V. B., Larkin, R. L., and Kupel, R. E. (1963).Biochemistry 2, 1127- 1131. Schindler, A. F., Palmer, J. G., and Eisenberg, W. V. (1967).Appl. Microbiol. 15, 10061009. Schoental, R. (1967). Ann. Reu. Pharmacol. 7,345-356. Schoental, R., and White, A. F. (1965).Nature 205, 57-58. Schroeder, H. W. (1966). Appl. Microbiol. 14, 381-385. Schroeder, H. W., and Hein, H., Jr. (1967). Appl. Microbiol. 15,441-445. Schwartz, H. S., Sodergren, J. E., Garofalo, D., and Strenberg, S. S. (1965).Cancer Res. 25,307-317. Scott, d e B. (1965). Mycopathol. Mycol. Appl. 25,213-222. Scott, P. M., and Hand, T. B. (1967)J. Assoc. Ofic. Anal Chemists 50,366-370. Scott, P. M., Van Walbeek, W., and Forjacs, J. (1967).Appl. Microbiol. 15,945. Shank, R. C., and Wogan, G. N. (1965). Federation Proc. 24,627. (Abstr. 2752.) Shank, R. C., and Wogan, G. N. (1966). Toxicol. Appl. Pharmacol. 9,468-476. Shimkin, M. B., and Kraybill, H. F. (1964). Aduan. Cancer Res. 8, 1991. Shotwell, 0. L., Hesseltine, C. W., Stubblefield, R. D., and Sorenson, W. G. (1966). Appl. Microbiol. 14,425-428. Shotwell, 0. L., Shannon, G. M., Goulden, M. L., Milburn, M. S., and Hall, H. H. (1968).Cereal Chem.(In press.) Singleton, V. L., Bohonos, N., and Ullstrup, A. J. (1958). Nature 181, 1072-1073. Smalley, E. B., Nichols, R. E., Crump, M. H., and Henning, J. N. (1962).Phytopathology 52,753. Smith, R. H. (1963).Biochem. J . 88,50P-51P.

218


Smith, R. H. (1965).Biochem. J . 95,43-44. Soine, T. 0. (1964).J . Pharm. Sci. 53,231-264. Sorenson, W. G., Hesseltine, C. W., and Shotwell, 0. L. (1968). Mycopathol. Mycol. Appl. 13,49-55. Spector, W. S. (1957). “Handbook of Toxicology, 11. Antibiotics.” Saunders, Philadelphia, Pennsylvania. Spensley, P. C. (1963). Endeavour 22,75-79. Sporn, M. B., and Dingman, C. W. (1966).Cancer Res. 26,2488-2495. Sporn, M. B., Dingman, C. W., Phelps, H. L., and Wogan, G. N. (1966). Science 151, 1539-1541. Sreenivasamurthy, V., Jayaraman, A., and Parpia, H. A. B. (1965).In “Mycotoxins and Feedstuffs” (G. N. Wogan, ed.), pp. 251-260. M.I.T. Press, Cambridge, Massachusetts. Sreenivasamurthy, V., Parpia, H. A. B., Srikanta, S., and Shankar, A. (1967).J. Assoc. Ofic. Anal. Chemists 50, 350-354. Staehelin, T., Wettstein, F. O., and NOH, H. (1963). Science 140, 180-183. Steyn, P. S., and van der Menve, K. J. (1966). Nature 211,418. Stob, M., Baldwin, R. S., Twite, J., Andrews, F. N., and Gillette, K. G. (1962).Nature 196, 1318. Stoloff, L. (1967).J . Assoc. Ofic. Anal. Chemists 50, 354-360. Stoloff, L., Graff. A,, and Rich, H. (1966).J . Assoc. Ofic. Anal Chemists 49, 740-743. Stubblefield, R. D., Shotwell, 0. L., Hesseltine, C. W., Smith, K. L., and Hall, H. H. (1967).Appl. Microbiol. 15,186-190. Svoboda, D. J., Grady, H., and Higginson, J. (1966). Am. J . Pathol. 49, 1023-1051. Svoboda, D., Racela, A., and Higginson, J. (1967). Biochem. Pharmacol. 16,651-657. Symp. Mycotoxins Foodstuffs (1965). Agr. Aspects, Univ. Pretoria, February. Ind. Res. Develop. Div., Council Sci. Ind. Res., Dept. Agr. Tech. Serv., Pretoria. Synge, R. L. M., and White, E. P. (1959). Chem. Ind. (London) p. 1546. Synge, R. L. M., and White, E. P. (1960).New Zealand]. Agr. Res. 3,907-921. Szybalski, W., and Iyer, V. N. (1964). Federation Proc. 23,946-957. Taub, D., Cirotra, N. N., Hoffsommer, R. D., Kuo, C. H., Slates, H. L., Weber, S., and Wendler, N. L. (1967).Chem. Commum. p. 225. Te Punga, W. A., and MacKinnon, M. M. (1961).New Zealand]. Agr. Res. 4,147-150. Terawaki, A., and Greenberg, J. (1965). Biochim. Biophys. Acta 95,170-173. Teunisson, D. J., and Robertson, J. A. (1967).Appl. Microbiol. 15,1099-1103. Theron, J. J. (1965).Lab Znuest. 14, 1586-1603. Thornton, J. C., and Percival, R. H. (1958).Nature 182,1095. Thornton, J. C., and Percival, R. H. (1959).Nature 183, 63. Townsend, R. J. (1965). Symp. Microbiol. Deterioration Tropics, London, April 8-9. Townsend, R. J., Moss, M. O., and Peck, H. M. (1966).J.Pharm. Pharmacol. 18,471473. Townsley, P. M., and Lee, E. G. H. (1967).]. Assoc. Ofic. Anal. Chemists50,361-363. Trager, W. T., and Stoloff, L. (1967).J . Agr. Food Chem. 15,679-681. Trager, W. T., Stoloff, L., and Campbell, A. D. (1964).J . Assoc. Ofice. Agr. Chemists 47,993-1001. Trakatellis, A. C., Axelrod, A. E., and Montjar, M. (1964). Nature 203, 1134-1136. Tsunoda, H. (1963). Food Scl. Technol. May, Suppl., p. 161. Tulpule, P. G., Madhaven, T. V., and Gopalan, C. (1964).Lancet 1,962-963. Uritani, 1. (1967).J. Assoc. Ofic. Anal. Chemists SO, 105-114.

MYCOTOXINS

219

Uny, W. H., Wehrmeister, H. L., Hodge, E. B., and Hidy, P. H. (1966). Tetrahedron Letters 27, 3109-3114. van der Merwe, K. J., Fourie, L., and Scott, de B. (1963). Chem. Ind. (London)pp. 1660- 1661. van der Merwe, K. J., Steyn, P. S., and Fourie, L. (1965).J . Chem. Soc. pp. 7083-7088. Van Dorp, D. A., van der Zijden, A. S. M., Berrthuis, R. K., Sparreboom, S., Ord, W. O., de Jong, K., and Keuning, R. (1963).Rec. Tray. Chim. 82,587-592. Van Overbeek, S. (1966). Science 152, 721-731. Verbeek, R., and Dumitru, I. F. (1964). Arch. Intern. PhysioZ. Biochim. 72, 799-818. Verrett, M. J., Marliac, J. P., and McLaughlin, J., Jr. (1964).J.Assoc. Ofic. Agr. Chemists 47, 1003-1006. Watt, J. M., and Breyer-Brandwijk, M. G . (1962). “The Medicinal and Poisonous Plants of Southern and Eastern Africa,” pp. 1127-11.28. Livingstone, Edinburgh and London. Wildman, J. D., Stoloff, L., and Jacobs, R. (1967). Biotechnol. Bioeng. 9, 429-437. Wiley, M. (1966).J . Assoc. Ofic. Anal. Chemists49, 1223-1224. Wilson, B. J. (1966).Bacteriol. Reu. 30,478-484. Wilson, B. J., and Wilson, C. H. (1962a).J. Bacteriol. 83,693. Wilson, B. J., and Wilson, C. H. (1962b).J.Bacteriol. 84,283-290. Wilson, B. J., Harris, T. M., and Hayes, A. W. (1967). J . Bacteriol. 93, 1737-1738. Wiseman, H. G., Jacobson, W. C., and Harmeyer, W. C. (1967). J . Assoc. Ofic. Anal. Chemists 50, 982-983. Wogan, G. N., ed. (1965). “Mycotoxins in Foodstuffs.” M.I.T. Press, Cambridge, Massachusetts. Wogan, G. N. (1966a). Bacteriol. Reu. 30,460-470. Wogan, G . N. (1966b).Federation Proc. 25, 124-129. Wooley, D. W., Berger, J,, Peterson, W. H., and Steenbock, H. (1938).J . Nutr. 16, 465-476. Wragg, J. B., Ross, V. C., and Legator, M. S. (1967). Proc. Soc. Exptl. B i d . Med. 125, 1052-1055. Yates, S. G. (1962). Econ. Botany 16,295-303. Yates, S . G. (1963).J . Chromatog. 12,423-426. Yates, S . G., Tookey, H. L. (1965).AustraZianJ.Chem. 18, 53-60. Yates, S. G., Tookey, H. L., Ellis, J. J., and Burkhardt, H. J. (1967). Tetrahedron Letters 7,621-625. Yokotsuka, T., Sasaki, M., Kikuchi, T., Assao, Y., and Nobuhara, A. (1966). Am. Chem. SOC., Program Abstr., 152nd Meeting, New York p. A-016. Zampieri, A., and Greenberg, J. (1964). Biochem. Biophys. Res. Commun. 14, 172-176. Zimmer, C., Triebel, H., and Thrum, H. (1967). Biochim. Biophys. Acta 145,742-751. Zuckerman, A. J., Rees, K. R., Inman, D., and Petts, V. (1967a). Nature214,814-815. Zuckerman, A. J., Tsiquaye, K. N., and Fulton, F. (196713). Brit. J . Exptl. PathoZ.46, 20-27.


Transformation of Organic Compounds

by Fungal Spores

CLAUDE V ~ Z I N AS. , N. SEHGAL,AND URTAR SINGH Department of Microbiology, Ayerst Laboratories, Montreal, Quebec, Canada

I. Introduction ......................

..........................................

221 222 222 224 B. Spomlation ........................................................... C. Chemical Composition ................ ........................_.. 225 226 D. Dormancy ......... 226 E. Germination ...... 227 F. Chemical Activity. ........................ 227 Characteristics of Transformation with Spores.. A. Similarities a n d Differences between the Us 227 and the Use of Gro 229 B. General Technique ............................................... Transformation of Steroids ............................................ 24 1 A. Types of Transformations ...................................... 24 1 B. Transformation of Steroids on Large Scale ............... 250 C. Carbohydrate Metabolism and Steroid Transfor25 1 mations .......................................... 260 Transformation of Other Organic Compou 260 A. Formation of Ketones from Fatty Acids ............... B. Transformation of Triglycerides .............................. 263 C. Transformation of Penicillin into 6-Aniinopenicillanic Acid ............................,....................................... 263 D. Transformation of Carbohydrates ............................ 264 264 Discussion and Conclusion ........ ........................... References .................................................................. 265

11. The Fungus Spore ....................................................... A. Mycelium and Spores ...................

111.

IV.

V.

VI.

1. introduction

The capacity of microorganisms to synthesize rapidly large quantities of protoplasm and reserve materials, and to secrete a large variety of useful metabolites has led to industrial processes which can be classified arbitrarily into three categories. Microorganisms are grown on suitable media and harvested for their cells or for their reserve materials: the production of yeasts destined for bakery, the preparation of microbial fats, carbohydrates, proteins for human and animal consumption, etc. Microorganisms are also cultivated for the valuable excretory metabolites they sometimes accumulate, such as alcohols, organic acids, vitamins, antibiotics, and enzymes. These 22 1

222

CLAUDE V ~ Z I N A .s. N. SEHGAL, AND KARTAR SINGH

products, after appropriate purification, are used as solvents in industry, supplements in human food or animal feed, therapeutic agents in medicine, or as reagents (enzymes) to modify organic molecules (hydrolysis of starch by amylases), etc. Finally, actively growing cells in a nutrient medium, or cells washed free of nutrients, are used to effect specific transformations of chemical compounds. Transformation of steroids, alkaloids, fatty acids, polyols, hydrocarbons, and hydrolysis of penicillins into 6-aminopenicillanic acid belong to this group of microbial processes. Molds, yeasts, and bacteria (including streptomycetes) are commonly used in these transformations. Vegetative cells or spores serve to inoculate a suitable medium containing, besides nutrients, the chemical compound to be converted; the compound is transformed by the growing cells and the product recovered. It is sometimes advantageous to cultivate the microorganisms to a certain density, when its enzymic activity is optimum, before adding the transformable substrate. In general, transformation is complete before the organism sporulates to any extent. In transformations catalyzed by spores (“spore process”), to which this review refers, spores are first produced according to suitable methods and separated from the mycelium; spore suspensions are then brought in contact with the substrate until maximum yield of the product is obtained. Therefore, transformation with spores is essentially a two-step process. Needless to say, it is limited to sporulating microorganisms. The scope of this review is to describe the “spore process” as applied to the transformation of steroids and some other organic compounds, and to point out factors involved in the production of sufficient quantities of spores. Although this process has been studied mainly for fungal spores, it can also be applied to the spores of certain streptomycetes.

II. The Fungus Spore The fungus spore, in contrast to the vegetative mycelium, f s a nucleate portion delimited from the thallus, characterized by cessation of cytoplasmic mouement, small water content and slow metabolism, lack of vacuoles, and specialized for dispersal, reproduction or survival (Gregory, 1966).

A. MYCELIUMAND SPORES Spores represent a turning point in the history of a fungus; they are both the initial and the final stages of its life cycle, except in Mycelia Sterila. Under suitable conditions, spores germinate, i.e., grow a

TRANSFORMATION OF ORGANIC COMPOUNDS BY FUNGAL SPORES

223

germ tube, which becomes a hypha; hyphae elongate by deposition of new protoplasm at hyphal tips, and branch below growing tips. A mycelium of hyphae is eventually formed which constitutes the thallus in the mycelial fungi. A few families in lower phycomycetes do not produce a mycelium and their thallus is a single cell, i.e., holocarpic. In the mycelial fungi, hyphae are either aseptate or septate; phycomycetes have characteristically aseptate hyphae in their growing stage; septa may later form in older hyphae. Ascomycetes, basidiomycetes and fungi imperfecti have regularly septate hyphae; cross-walls, i.e., septa, have a small central pore which allows protoplasmic movement between cells. In both aseptate and septate hyphae a coenocytic condition prevails and cytoplasmic circulation is mainly unidirectional, i.e., toward the growing tips; in the meantime, cytoplasm retracts from the old part of the hypha, which becomes skeletal. The cytoplasmic mass continues to move toward the growing region, and, when hyphal elongation ceases, it incorporates mainly in the reproductive structures, the spores. Hickman (1965), Mandels (1965), and Raper (1966) have recently reviewed structure, growth, and life cycles in fungi. Reproduction in fungi is characterized by the formation of a large variety and a great number of spores. According to their origin, they can be classified into two groups: sexual spores and asexual spores. The former contain nuclei which are the product of meiosis (Olive, 1965) and serve as a basis of classification of the perfect classes oE fungi: resting spores (such as oospores and zygospores) in phycomycetes, ascospores in ascomycetes, basidiospores in basidiomycetes. For a detailed account of the origin of sexual spores, the reader is referred to general textbooks of mycology and to the recent reviews of Ainsworth and Sussman (1965-1966) and Madelin (1966b). Asexual spores are produced by phycomycetes, ascomycetes, basidiomycetes, and fungi imperfecti (except Mycelia Sterila). They can be uni- or multinucleate, and their nuclei are never the product of meiosis. However, multinucleate spores may be heterokaryotic, i.e., they may contain genetically different nuclei from two or more parents which have exchanged nuclei through hyphal anastomosis. Mutation of a nucleus in a multinucleate spore leads to the same condition; for reviews on heterokaryosis, see Vezina (1956) and Davis (1966). In heterokaryons, nuclei divide independently, can complement nutritionally, but never fuse. Exceptionally, nuclei in a heterokaryotic cell can fuse to yield heterozygous diploids by a process known as “parasexuality” (Roper, 1966).

224

CLAUDE VkZINA, S. N. SEHGAL, AND KARTAR SINGH

Asexual spores are extremely varied both in origin and shape and have served very early as a taxonomic criterion (Saccardo’s Sylloge Fungarurn, cited by Tubaki, 1966; Vuillemin, 1910, 1912). Many systems of classification have emerged (Tubaki, 1966; Madelin, 1966a); for the present review the following distinctions are sufficient. Sporangiospores are asexual spores formed within a sporangium and are usually present in phycomycetes. Thallospores, which include arthrospores and blastospores, are asexual spores which originate by septation and breaking u p of hyphae or by enlargement of hyphal tips. Finally, conidiospores or conidia (conidium verum of Vuillemin, 1912) are asexual spores borne externally on specialized hyphae, the conidiophores; these are borne singly from the mycelium or grouped in an acervulus or enclosed in a pycnidium.

B. SPORULATION The capacity of fungi to produce spores (sometimes many types are present in the same fungus) is a hereditary attribute which does not necessarily express itself. Phenotypic expression, qualitative and quantitative, greatly depends on the environment. Hawker ( 1966a) has recently reviewed the influence of external factors on sporulation. Factors most often recorded to play an important role in the production of various spores are: 1. Nature and concentration of nutrients: carbon and nitrogen sources, N/C ratio, growth factors, mineral elements (calcium and zinc especially), 2. Water content of the medium and relative humidity of the atmosphere surrounding the substratum, 3. pH and temperature, 4. Aeration and accumulation of volatile substances in the atmosphere above the substratum, 5. Illumination, inhibitors, presence of other organisms. All fungi do not respond identically to these factors, and great discrepancy is usually observed between species of the same genus and even between strains of the same species. This is reflected by the contradictory results often reported in the literature for the influence of given factors. The biochemical mechanism of sporulation is not known, but, in genetically competent strains, induction, repression, or feedback inhibition may well be responsible for these variations. At sporulation, considerable reorganization of nitrogenous components (probably including enzymes) has been observed (Turian, 1966). Therefore, when metabolites accumulate, certain inducible


225

enzymes probably appear, others are repressed or derepressed in their synthesis, still others are inhibited in their activity. A delicate balance must establish between various metabolic pathways. Nutrients in the environment which can penetrate into the protoplasm may also have a great influence on these metabolic pathways, i.e., they may contribute to maintain or disrupt the equilibrium. Two strains of the same species may be apparently identical, but fine analysis could reveal nonidentical biochemical blocks or nonidentical enzymes (isoenzymes) which have the same activity but different susceptibility to metabolites, nutrients, or physical-chemical conditions (pH, temperature, etc.). Therefore, the response to environmental conditions would vary considerably in these apparently identical strains. Despite these discrepancies, enough consistent data are available to make the following generalizations (Hawker, 1966a): 1. “Conditions favoring maximum mycelial growth may or may not also favor asexual reproduction but are usually unfavorable to sexual reproduction,” and to asexual sporulation in submerged culture (see Section 111, B,l,b); 2. “The range of any particular external condition which will allow sporulation is usually narrower than that permitting mycelial growth and is sometimes narrower for sexual than for asexual reproduction.” C. CHEMICALCOMPOSITION Lilly (1965) and Birkinshaw (1965) have recently reviewed the chemical composition of fungi, which contain and excrete a rich variety of chemical compounds. Qualitatively, spores do not differ greatly from mycelium. Quantitatively, notable differences are observed which may have an important influence on their respective behavior. The usually lower water content of spores may be partly responsible for their lower permeability and higher resistance to various external factors. In Neurosporu crussu, Bianchi and Turian (1967) reported a higher content of phospholipids in conidia than in mycelium; although phospholipids were slowly consumed as an energy source in ungerminated conidia, they were not actively metabolized during germination, and no specific role could be attributed to these compounds. In the same organism, Hill and Sussman (1965) and Sussman (1966b) reported significant differences in the lipid content of conidia and ascospores; phospholipids predominate in the former, esters in the latter. Both conidia and ascospores contain more trehalose than the mycelium; on germination, conidia uti-

226


lize trehalose rapidly, whereas ascospores first utilize lipids, then, after they are activated, trehalose and lipids. Trehalose synthesis is by the mycelium, which accumulates it until conidiation begins; it is possible that this sugar is also necessary for conidiospore formation. Trehalose is the main endogenous substrate of dormant ascospores.

D. DORMANCY Sussman (1966a) has defined dormancy as “any rest period or reversible interruption of the phenotypic development of an organism.” He distinguishes between constitutional and exogenous dormancy; the former appears as an inherent property of the dormant stage which depends on impermeability, or presence of inhibitors (such as trimethylethylene); exogenous dormancy results from unfavorable chemical or physical conditions of the environment and comes to an end as soon as these conditions are removed. Constitutively dormant spores will germinate only after an activation treatment is applied: heat-shock, light, sodium glycocholate, sodium taurocholate, wetting agents such as alcohols and acetone, or transfer to distilled water or nutrient medium. Combination of two or more treatments may be necessary. E. GERMINATION The germination process, according to Manners (1966), can be described as including three phases: 1. Pregermination phase in which morphological and physiological changes occur within the confines of the spore wall, 2. Germination proper, which is the act of protrusion of the germ tube from the spore wall, 3. Elongation of the germ tube. In the pregermination stage, the first detectable change is an increase in oxygen consumption, followed by the swelling of the spore. Increase in volume may be as high as 20 times (Hawker, 1966b) and is always greater for spores unable to germinate without addition of glucose (“heterophagous” spores, Cochrane, 1966). It appears that constitutively dormant spores have lower respiratory rates than exogenously dormant spores. During the activation period, however, increase in respiration is much greater in the former than in the later (Sussman, 1966b). Most enzymes responsible for respiratory pathways have been detected in spores (Cochrane, 1966). During the activation period, fermentative characteristics are observed, which are soon replaced by


227

a strong oxidative type of metabolism (Cochrane, 1966; Sussman, 196613). Synthesis of macromolecules is also observed in the later stage of activation (Gottlieb, 1966). Conidia of N . crassa, during activation, show de novo synthesis of nicotinamide-adenine dinucleotide phosphate (NADP) specific glutamic acid dehydrogenase, although no net protein synthesis is observed during the same period (Tuveson et al., 1967). Triglycerides are rapidly utilized during germination of conidia of Fusarium culmorum (Marchant and White, 1967); atmospheric COZ is fixed; glucosamine was proposed as the initiator of swelling.

F. CHEMICALACTIVITY It is long-experienced observation that spores, constitutively or exogenously dormant (Sussman, 1966b), are endowed with low, but steady, metabolism: respiratory reactions (Cochrane, 1966), biosynthetic pathways (Gottlieb, 1966), exchange of gases, and excretion of metabolites are operative. The first observation that spores can transform compounds added to the medium and apparently unrelated to assimilation or dissimilation processes dates back to the report of Gehrig and Knight (1958) that octanoic acid is converted to 2-heptanone by conidia of Penicillium roqueforti. In submerged culture, ketone formation was directly proportional to the degree of sporulation; in spore suspensions, carefully freed of mycelial cells, oxygen uptake increased directly with ketone formation, and spores did not germinate. Germinated conidia and mycelium were inactive. Gehrig and Knight (1961) extended their observation to other molds and recognized possible applications; soon after, Schleg and Knight (1962) reported the lla-hydroxylation of progesterone by conidia of Aspergillus ochraceus. The pioneer work of Knight and his collaborators laid the foundations of the “spore process” which is described in the next section. 111. Characteristics of Transformation with Spores

A. SIMILARITIES AND DIFFERENCES BETWEEN THE USE OF SPORES AND THE USE OF GROWING OR RESTING MYCELIUM Transformation of organic compounds by microorganisms differs to a great extent from the so-called classical fermentations. In the latter, the products (antibiotics, organic acids, solvents) are the result of complex biosynthetic processes, whereas in “microbiological transformations,” a chemical compound added to the culture usually

228

CLAUDE VkZINA, S. N. S E I G A L , AND KARTAR SlNGH

undergoes a simple modification. These transformations, however, are similar to classical fermentations as far as the microbiological technique is concerned. The microbe of choice is maintained according to standard methods; an inoculum is built up which is used to inoculate a production medium in a shake flask or a fermenter; the organism is grown until a desired concentration of active cells is reached; the substrate (steroid, alkaloid, etc.) is added and fermentation continued until maximum yield of the product(s) is obtained, after which extraction of the product is completed. Aseptic conditions are maintained throughout the entire process of transformation. As a modification, the organism is grown to the desired concentration, cells are centrifuged or filtered off and resuspended in buffer containing certain nutrients (if needed) before the substrate is added. Sometimes, antibiotics are used as a precaution against contamination. Another modification consists in diluting the microbial culture and adjusting the p H before adding the substrate. The optimum procedure is usually arrived at from previous experience and by empirical studies. For a more detailed description of the microbiological technique involved in steroid transformation, the interested reader is referred to the reviews by Prescott and Dunn (1959), Stoudt (1960), Kogan (1962), Peterson (1963), capek et al. (1966), and Charney and Herzog (1967), and to the voluminous patent literature since the patent of Murray and Peterson (1952). Transformation with spores has a dual nature. In the first step, a spore inoculum is prepared aseptically from a slant or a Roux bottle seeded with a chosen (sporulating) organism, and used to inoculate an artificial or a natural medium in surface or submerged culture. Conditions are such that the inoculum germinates rapidly and a “dominant” (not necessarily pure) mycelial culture establishes which yields an optimum number of spores in a few days. Spores are harvested, separated from the mycelium, and collected as a paste which can be used immediately or stored in the freezer until needed. In the second step, spore pastes are suspended in a suitable buffer, and the suspensions adjusted to the desired concentration of spores. Glucose is added, depending on the transformation reaction; to check contamination, antibiotics may also be used, if the transformation is expected to last more than 48 hours. The chemical compound to be converted is then charged into the mixture, and incubation continued until maximum yields are obtained. Since transformation is aerobic,


229

the reaction is carried out in shake flasks or open tanks equipped with agitation-aeration facilities. During the course of transformation, samples are withdrawn periodically and analyzed for substrate and product, and sometimes for glucose; dissolved oxygen is also determined in tanks. If necessary, the course of transformation can be corrected at any time by adding nonaseptically glucose, spores, or by increasing aeration-agitation. During the entire process, spores remain ungerminated, and no effort is made to work under aseptic conditions.

B. GENERAL TECHNIQUE 1 . Production of Spores Sporulation of fungi has been studied extensively, and the influence of several factors is relatively well known (see Section II,B), hence standard procedures are available for preparation of spore suspensions for inoculum purposes. However, quantitative results are seldom reported in the literature, and it is not possible to predict for a particular organism the number of spores which can be obtained under certain conditions. This information becomes necessary when the spore process is contemplated, since for transformation a specific spore concentration is required (Vezina et al., 1963). The authors studied quantitatively the relative value of various techniques in mass production of spores. Surface and submerged culture techniques will be considered.

a. Surface Sporulation. i. Artificial media. Artificial media dispensed in Roux bottles are usually suitable and sufficient on a laboratory scale for screening work and for studying factors involved in transformation. At this preliminary stage, aseptic conditions can be observed throughout the entire process. The technique is as follows (V6zina et al., 1963; Singh et al., 1965):various media in Roux bottles are inoculated from a soil stock (Greene and Fred, 1934)or a slant, and individual bottles incubated at various temperatures in a well-aerated incubation room maintained at 55-60% R.H. After suitable periods of incubation, usually between 5 and 15 days, spores are harvested by brushing gently the surface of agar media with a sterile brush in presence of a 1:10,000 solution of Tween 80 or sodium monolauryl sulfate. When liquid media are used, the sporulated mycelial mat is removed carefully, suspended in a wetting solution, and macerated in a Waring

230


blendor. Suspensions are then filtered through sterile glass wool to remove mycelial fragments, to break chains and clumps of spores, and to obtain homogenous suspensions. Conditions for optimum spore yields vary considerably from one strain to another; the most important factor for a given strain is the medium. The authors compared a great many media reported in the literature and modifications therefrom. For A. ochraceus, yields vary from a few spores on several media to 1.2 x 10” spores per Roux bottle when W.L. Nutrient Medium (Difco Manual, 1953) is used; Sabouraud Dextrose Agar yields 8 X lo9 spores per bottle. Optimal media for Septomyxa afinis and Mucor griseo-cyanus are reported by Singh et al. (1965, 1967a). The disadvantage of artificial media is the difficulty in scaling u p the process, i.e., producing conveniently and at reasonable cost enough spores for large-scale transformation.

ii. Natural media. Cereal grains are cheap media and a great number of fungi sporulate well on these. Hard wheat bran, “pot” barley (Ogilvie Flour Mills, Montreal) and cracked corn are satisfactory. Singh et al. (1968)have described the procedure for A. ochraceus and S . afinis. Two hundred grams of “pot” barley or 100 gm. wheat bran are dispensed in a 2.8 liter Fernbach flask, and moistened with tap water (80 ml. and 120 ml. for A. ochraceus and S . afinis on barley; 50 ml. on wheat bran); flasks are sterilized at 121°C. for 1 hour and cooled to suitable temperature while shaking to break clumps; each flask is inoculated with 2-5 ml. of a spore suspension containing about 4 X lo8 conidia per milliter, and incubated at 28”C., 98% R.H. for 5-7 days for A. ochraceus, or at 25”C., 98% R.H. for 10-12 days for S . afinis. Conidia are harvested by adding 500 ml. of water containing 0.01% Tween 80 to each flask; the cotton plug is replaced by a rubber stopper and the flask shaken to dislodge spores from grains. The spore suspension is decanted through a copper screen which fits the flask neck and keeps the grains in the flask; a second harvest is effected using 300 ml. of water. Combined spore suspensions are filtered through glass wool, spores collected in a Sharples centrifuge, and washed by resuspending the paste in water and centrifuging. Each flask yields 5-6 gm. of spore paste of A. ochraceus or S . afinis: dry weight is about 30%, total yield is 2.5-5.0 X 10” conidia. The water content of the medium and the relative humidity maintained in the incubator have a great influence on sporulation and optimum condi-


231

tions must be determined for each organism. Influence of varying amounts of water added to barley on spore production is given in Table I for six fungi. The addition of various supplements, such as TABLE I ADDITION OF WATER TO “POT” BARLEY ON SPORULATION VARIOUSMOLDS (CONIDIA x 101l/F~~~~)U

INFLUENCE OF

OF

Water Aspergillus Aspergillus Mucor Penicillium Septomyra Stachylidium addedb ochraceus niger griseocyanusc chrysogenum afinis theobromae (ml./flask) NRRL 405 ATCC 9142 ATCC 1207-A WIS 53-414 ATCC 6737 ATCC 12,474

40 60 80 100 120 140 160

r

TABLE X METABOLISM OF GLUCOSEBY SPORES

Substrate

Gluco~e-l-~~C Glucose-&'% Glucose-l-'% Gluco~e-l-~~C

Incubation time (min.) 120 120 230 230

COZ evolved

OF

LU

A . ochrceus"

Glucose utilized

COab Ratio: - glucose % ~

pmoles

coundmin.

10.7 10.7 21.0 21.0

3,020

640 7,350 2,320

pmoles 3.7 3.7 11.5 11.5

coundmin. 6,790 6,410 21,100 19,920

E CJC6

44 10 35 12

"Spores, 5 x lo8and glucose, 54.4 pmoles per Warburg flask; phosphate buffer,pH 5.0; temperature, 28°C.; volume, 2.5 ml. * "COZ as percent of glucose-"C utilized.

2

z

4.4 2.9

TABLE XI EFFECT OF ANTIBIOTICS ON PROGESTERONE CONVERSION AND GLUCOSE UTILIZATION BY CONIDM OF A.

ochraceus

r!

8-

Antibiotic

Control Polymyxin B Polymyxin B Tyrothricin Valinomycin Oligomycin Oligomycin Antimycin A "Conidia, 5 x

Concentration Gg.lml.)

20 200 20 20 2.7 13.5 20

Progesterone conversion" (% of control)

Glucose utilized" (% of control)

100

100 86 28 7% 75 41 36 78

62 18 24 97 24 0 83

lo8;progesterone, 2.5 mg.; incubation, 16 hours; temperature,28°C. (Singh, unpublished).

O2 Uptake in 5 hours

s

(pliters)

2

625 517 137 583 720 749 606 615

0

i

8 n 0

8 9

2 m .e

260

CLAUDE VkZIh'A, S. N. SEHGAL, AND KARTAR SINGH

18% inhibition of glucose utilization. Studies with inhibitors suggest the requirement for SH groups and the involvement of the cytochrome system in spores. However, antimycin A (20 pg./ml.) had a very limited effect on steroid conversion or oxygen uptake with glucose (0.1 pg./ml. inhibits oxygen uptake in yeast: Ahmad et al., 1950; Ramachandran and Gottlieb, 1961). One might conjecture that the organism (spores), while having a typical cytochrome-linked electron transport system, does not possess an antimycin-sensitive (AS) site in its electron transport chain (Rieske, 1967),as far as glucose utilization and steroid conversion are concerned. Oxidation of acetate is, however, strongly inhibited by antimycin. With spores of S. afinis, on the other hand, C-l-dehydrogenation, glucose and acetate oxidation are strongly inhibited (Singh et al., 1965). 2. Cell-Free Preparations Cell-free preparations of washed and starved A. ochraceus and S . ufinis spores were prepared by three different procedures: (1)grinding with sand or glass powder in pestle and mortar; (2) treatment in a 10-kc ultrasonic oscillator; and (3)releasing spore suspensions from a pressure of 20 to 40,000 p.s.i. through a narrow orifice in a Ribi cell disintegrator. Cell-free extracts of A. ochraceus spores exhibited good phosphoglucose isomerase, glucose 6-phosphate dehydrogenase, and phosphoribose isomerase activities. The extracts also exhibited hexokinase, 6-phosphogluconic dehydrogenase, phosphoglucomutase, and transketolase-transaldolase. Aldolase, glyceraldehyde 3-phosphate dehydrogenase and enolase, some of the key enzymes of the E M pathway, could not be detected (Singh et al., 1966). An active electron transport system in spores was indicated by the presence of NADH-oxidase and NADH-cytochrome-c reductase in cell-free extracts of A. ochraceus. Cochrane and Cochrane (1966) found every known enzyme of the EM pathway in the ungerminated conidia of Fusarium solani.

V. Transformation of Other Organic Compounds

A. FORMATION OF KETONES FROM FATTY -ACIDS Gehrig and Knight (1958) demonstrated the ability of Penicillium roqueforti to transform fatty acids to 2-ketones. Their observation differs from the transformation of steroids with spores of fungi in that spores were reported to be solely responsible for this reaction and transformation could not be brought about by the growing spore-free


261

mycelium. Gehrig and Knight (1961)reported that spores of 9 out of 11 aspergilli, 9 out of 12 penicillia and related fungi, including Paecilomyces uarioti and Scopulariopsis breuicaulis, rapidly converted caprylate to 2-heptanone. However, spores of mucorales (Mucor hiemalis, M . mucedo, Phycomyces blakesleanus, Rhizopus stolon$er, Thammidium elegans) were not capable of catalyzing this reaction. Knight (1966) postulated the pathway of ketone formation as follows. (1)

R-CHz-CHz-COOH-

-2H HrO

R-CH=CH-COOH

(2)

R-CH=CH-COOH-

(3)

R - CHOH- CHz- COOH __* R-CO- CHz- COOH R - CO- CHz - COOH R -CO - CH, COz

(4)

R-CHOH-CHz-COOH

-

-2H

SUM: R - C H ~ - C H ~ COOH 3R +HI0

+

co - CH$+ coZ

The general procedure for transformation of fatty acids with spores of P. roqueforti is very similar to that used for steroid transformation. A suspension of spores containing 1.5 X loDspores/ml. in 0.1 M sodium phosphate buffer is made. Fatty acids are added as their sodium salts. Lawrence (1966) added 0.5 mg./ml. streptomycin to prevent any bacterial contamination. The methyl ketones are estimated as their 2,4-DNP-hydrazones and identified by paper chromatography (Lawrence, 1966). The optimum pH for transformation of fatty acids by spores is between 5.5 and 7.0 and optimum temperature is 27°C. (Fig. 8). Aeration of spore-fatty acid suspensions is essential for transformation. The ability of washed spores of P. roqueforti to oxidize octanoic acid decreases markedly with age (Fig. 9), presumably because of inactivation of oxidative enzymes. However, young spores (3-4 days old) can be stored for over 36 months at 4°C. without any loss in their ability to convert octanoic acid to 2-heptanone (Gehrig and Knight, 1963). A lag period exists in oxidation of fatty acids by spores of P. roqueforti. By increasing the concentration of the fatty acid, or by decreasing the number of spores, the lag period becomes longer. The ratio of spore number to initial concentration of acid determines the rate of oxidation as shown in Fig. 8. The lag phase can be completely eliminated by preincubating the spores with low concentration of octanoic acid for 2-3 hours with subsequent additions of higher concentrations of octanoic acid (Lawrence, 1966). Lawrence suggested that the oxidative enzymes are activated during the lag phase and not induced,

262

CLAUDE VBZINA, S. N. SEHGAL, AND KARTAR SINGH

FIG.8. Penfcillfumroquefortf. The effect of spore concentration on the rate of formation of heptan-2-one from octanoic acid (3 pmoles) at pH 5.5 (100 pmoles phosphate buffer). Total volume 3 ml. Spore suspension A contained 1.5 X lo8 spores. B, C, D and E contained one half, quarter, eighth and sixteenth, respectively, of this spore concentration. From Lawrence, 1966; reproduced from The Journal of General Microbiology 44, p. 398, with permission of Cambridge University Press. 4 days

1.5-

-ia.

5 days

1.0-

g N I

C

f

0.5-

I

0

2

4 Time (hours)

6

FIG. 9. Effect of age of washed spores of Penicfllfumroqueforti (1.5 x loB)on ability to form heptan-2-one from octanoic acid (3 pmoles) at pH 5.5 (100 pmoles phosphate buffer). Total volume 3 ml. From Lawrence, 1966; reproduced from The Journal of General Microbfology 44, p. 396, with permission of Cambridge University Press.


263

since chloramphenicol, a specific inhibitor of protein synthesis, did not prevent formation of methyl ketones. He also showed that small amount of sugars, such as glucose, galactose, xylose, and sucrose, and amino acids, such as L-alanine, L-serine, and L-proline, stimulated this activation and shortened the lag phase. Lawrence (1966) and Knight (1966) reported the inhibition of fatty acid oxidation by cyanide, azide, 2,3-dimercapto-l-propanol, 2,4-dinitrophenol, mercuric chloride, etc. Similar observations were made in steroid transformations by fungal spores (Vezina et al., 1963; Singh et al., 1965).

B. TRANSFORMATION OF TRIGLYCERIDES Oxidation of triglycerides to methyl ketones by spores of Penicillium roqueforti has recently been demonstrated by Lawrence (1967). oxidases

esterases

Triglyceride-

fatty acid-

methyl ketones

Transformation with washed spores proceeds very slowly, but addition of casamino acids, L-proline or Lalanine, and certain sugars, such as D-galaCtOSe, D-xylose, or D-glUCOSe, stimulates the reaction. Since nitrogen-free spore systems can oxidize fatty acids to methyl ketones (Knight, 1966; Lawrence, 1966), it is likely that a nitrogen source is necessary to transform triglycerides into fatty acids. The factors which stimulate the formation of methyl ketones from triglycerides also stimulate the germination of spores (Lawrence, 1967); Lawrence has proposed a role of fungal esterases in spore germination. C. TRANSFORMATION OF PENICILLININTO 6-AMINOPENICILLANIC ACID 6-Aminopenicillanic acid (6APA) has become an important starting material for the synthesis of new penicillins. It has been obtained by the action of microbial amidases on penicillin G and penicillin V (Batchelor et al., 1961). Singh et al. (196%)have reported the tranformation of penicillin V into GAPA with spores of Fusarium conglomerans and F. monilifomne. However, penicillin G is not transformed. To a suspension containing 2 X lo8to 1X logspores per milliliter of F. moniliforme in phosphate buffer (1%)at pH 7.8, 2 mg./ml. penicillin V are added and the mixture is incubated at 28°C. on a rotary shaker. In 48 hours, about 80% of penicillin V is converted into 6APA. Only a trace amount of penicillin V is transformed into GAPA in stationary culture. Spores can be reused 2 or 3 times without significant loss in activity

264


to transform penicillin V. The addition of glucose or another source of energy is not necessary. This transformation can be carried out on a large scale. D. TRANSFORMATION OF CARBOHYDRATES Johnson et al. (1967) have reported that the amylase of conidia of AspergiZZus wentii NRRL 2001 converts soluble starch to glucose at an initial rate faster than the rate at which glucose is oxidized. Amylase activity reaches a peak in 3 days and then rapidly declines. Optimal pH for amylase activity is from 4 to 5, with only slight activity at pH 7. Production of glucose from starch is directly proportional to the concentration of conidia. Spore age, temperature of transformation, and substrate concentration influence the rate of transformation.

VI. Discussion and Conclusion In recent years spores, generally considered as a dormant stage in

the life cycle of fungi, have been shown to contain enzymes necessary for a variety of metabolic activities. Before the observation of Gehrig and Knight (1958), spores had been primarily used for maintenance and inoculation purposes by the industrial microbiologist, and the possibility of using them for transformation of organic compounds had not been visualized. Further studies by Lawrence (1966)on transformation of fatty acids and work in our laboratory on transformation of steroids have demonstrated the widespread and high activity of fungal spores in transforming compounds apparently unrelated to assimilation and catabolic processes. During transformation, even though they are maintained in the early pregermination stage (before swelling takes place), spores are 3 to 10 times more active than mycelium on a dry weight basis. This is not too surprising, if one considers spores as finely divided cytoplasm acquired from the mycelium. Although spores have only been used to transform steroids, fatty acids, triglycerides, penicillin V, and carbohydrates, there is no reason to believe that they will not transform other substrates, such as alkaloids, hydrocarbons, etc. The “spore process,” on the other hand, is limited to sporulating fungi and streptomycetes. Feasibility of producing spores and transforming on a large scale has already been established (Singh et al., 1968). The use of spores offers certain advantages: spores of various organisms can be stored frozen for a long period of time without significant loss in activity and are readily available for transformation of compounds. Homogeneity of spore suspensions and simplicity of transformation media


265

prevent excessive foaming and make extractions easier; homogenous spore suspensions are also easy to standardize, and activity can be expressed in terms of a unit number of spores. Fungal spores offer interesting possibilities as biochemical catalysts and reagents in a variety of biochemical studies; this field has not been fully explored and deserves further attention. ACKNOWLEDGMENTS

We are deeply indebted to the late Dr. Stanley G. Knight who made the original observation which led to the development of the “spore process” for the transformation of steroids. Until his premature death, Dr. Knight constantly advised us not only on the use of spores, but also on experiments to widen horizons on the nature of the fungal spore. We wish to acknowledge Dr. Roger Gaudry, Dr. Donald A. Buyske, and Dr. W. G. Hendrickson for their sustained encouragement in the course of these studies. We also thank all our colleagues and collaborators at Ayerst Laboratories for their valuable help in all phases of this work.

REFERENCES Ahmad, K., Schneider, H. G., and Strong, F. M. (1950).Arch. Biochem. 28,281-294. Ainsworth, G. C., and Sussman, A. S., eds. (1965-1966). “The Fungi-An Advanced Treatise,” Vols. I & 11. Academic Press, New York. Batchelor, F. R., Chain, E. B., and Rolinson, G. N. (1961). Proc. Roy. Soc. (London) BlM, 478-531. Bianchi, D. E., and Turian, G . (1967). Nature 214,1344-1345. Birkinshaw, J. H. (1965). In “The Fungi-An Advanced Treatise” (G. C. Ainsworth and A. S.Sussman, eds.), Vol. I, p. 179. Academic Press, New York. Blumenthal, H. J. (1965). In “The Fungi- An Advanced Treatise” (G. C. Ainsworth and A. S. Sussman, eds.), Vol. I, p. 229. Academic Press, New York. Bobbitt, J. M. (1963).“Thin Layer Chromatography,” p. 169. Reinhold, New York. Bush, I. E. (1961). “The Chromatography of Steroids.” Macmillan (Pergamon), New York. Caltrider, P. G., Ramachandran, S., and Gottlieb, D. (1963). Phytopathology 5 3 , 8 6 9 2 . Capek, A., Hanf, O., and Tadra, M. (1966). “Microbial Transformations of Steroids.” Academia, Prague, and Dr. W. Junk, Publ., The Hague. Carvajal, F. (1947). Mycologia 34,426440. Casas-Campillo, C., and Bautista, M. (1965).Appl. Mtcrobiol. 13,977-984. Charney, W., and Herzog, H. L. (1967). “Microbial Transformations of Steroids.” Academic Press, New York. Chiang, C., and Knight, S. G. (1960). Biochem. Biophys. Res. Commun. 3,554-559. Cochrane, V. W. (1966). In “The Fungus Spore” (M. F. Madelin, ed.), p. 201. Butterworths, London and Washington, D. C. Cochrane, V. W., and Cochrane, J. C. (1966).PlantPhysiol. 41,810-814. Cochrane, V. W., Cochrane, J. C., Collins, C. B., and Serafin, F. G. (1963a). Am. J . Botany 50,806-814.

266

CLAUDE VbZINA, S. N. SEHGAL, AND KARTAR SINGH

Cochrane, V. W., Cochrane, J. C., Vogel, J. M., and Coles, R. S., Jr. (1963b)J Bacteriol. 86,312-319. Davis, R. H. (1966). In “The Fungi-An Advanced Treatise” (G. C. Ainsworth and A. S. Sussman, eds.), Vol. 11, p. 567. Academic Press, New York. Deghenghi, R., Boulerice, M., Rochefort, J. G., Sehgal, S. N., and Marshall, D. J. (1966). J . Med. Chem. 9,513-516. Dulaney, E. L., Stapley, E. O., and Hlavac, C. (1955).Mycologta 47,464-474. Eppstein, S. H., Meister, P. D., Peterson, D. H., Murray, H. C., Leigh Osborn, H. M., Weintraub, A., Reineke, L. M., and Meeks, R.C. (1958).]. Am. Chern. Soc. 80,33823389. Farkas, G. L.,and Ledingham, G. A. (1959). Can.]. Microbiol.5,141-151. Foster, J. W., McDaniel, L. E., Woodruff, H. B., and Stokes, J. L. (1945).J . Bacteriol. 50,365-368. Gehrig, R.F., and Knight, S. G. (1958). Nature 182,1237. Gehrig, R. F., and Knight, S. G. (1961). Nature 192,1185. Gehrig, R.F., and Knight, S. G. (1963).Appl. Mtcrobtol. 11,166-170. Gottlieb, D. (1966). In “The Fungus Spore” (M. F. Madelin, ed.), p. 217. Buttenvorths, London and Washington, D. C. Greene, H. C., and Fred, E.B. (1934). Ind. Eng. Chem. 26,1297-1299. Gregory, P. H. (1966).In“The Fungus Spore” (M. F. Madelin, ed.), p. 1. Buttenvorths, London and Washington, D. C. Hawker, L. E. (1966a). In “The Fungi-An Advanced Treatise” (G. C. Ainsworth and A. S. Sussman, eds.), Vol. 11, p. 435. Academic Press, New York. Hawker, L. E. (1966b). In “The Fungus Spore” (M. F. Madelin, ed.), p. 151. Butterworths, London and Washington, D. C. Hickman, C. J. (1965). In “The Fungi-An Advanced Treatise” (G. C. Ainsworth and A. S. Sussman, eds.), Vol. I, p. 21. Academic Press, New York. Hill, E. P., and Sussman, A. S. (1965)J. Bacteriol.88,1556-1566. Johnson, D. E., Nelson, G. E. W., and Ciegler, A. (1967).Bacteriol. Proc. p. 34. Kakac, B., and Vejdelek, Z. (1963).“Handbuch der Kolorimetrie,” Bd. 11. Fischer, Jena. Knight, S. G. (1966).Ann. N . Y. Acad. Sci. 139,8-15. Kogan, L. M. (1962). Russ. Chem. Rev. (English Transl.)31,294-308. Kosmol, H., Kieslich, K., Vossing, R., Koch, H. J., Petzoldt, K., and Gibian, H. (1967). Ann. Chem. 701,198-205. Lardy, H. A., Johnson, D., and McMurray, W. C. (1958).Arch. Biochem. Btophys. 78, 587-597. Lawrence, R.C. (1966).J. Gen. Microbiol. 44,393-405. Lawrence, R.C. (1967).J. Gen. Mtcrobtol. 46,65-76. Lilly, V. G. (1965).In “The Fungi -An Advanced Treatise” (G. C. Ainsworth and A. S. Sussman, eds.), Vol. I, p. 163. Academic Press, New York. Madelin, M. F. (1966a). In “The Fungus Spore” (M. F. Madelin, ed.), p. 15. Butterworths, London and Washington, D. C. Madelin, M. F., ed. (196613). “The Fungus Spore,” Buttenvorths, London and Washington, D. C. Mandels, G. R. (1965). In “The Fungi-An Advanced Treatise” (G. C. Ainsworth and A. S. Sussman, eds.), Vol. I, p. 599. Academic Press, New York. Manners, J. G. (1966). In “The Fungus Spore” (M. F. Madelin, ed.), p. 165. Butterworths, London and Washington, D. C. Marchant, R., and White, M. F. (1967)J. Gen. Mtcrobiol. 48,65-77. Meyers, E., and Knight, S. G. (1958).Appl. Microbiol.6,174-178.

TRANSFORMATIONOF ORGANIC COMPOUNDS BY FUNGAL SPORES

267

Meyers, E., and Knight, S. G. (1961).Mycologia 53,115-122. Morton, A. G. (1961).Proc. Roy. Soc. (London) B153,548-569. Murray, H. C., and Peterson, D. H. (1952).US.Patent No. 2,602,769. Olive, L. S. (1965).In “The Fungi-An Advanced Treatise” (G. C. Ainsworth and A. S. Sussman, eds.), Vol. I, p. 143. Academic Press, New York. Owens, R. G. (1955).Contrib. Boyce Thompson Inst. 18,125-152. Perlman, D., Weinstein, M. J., and Peterson, G. E. (1957).Can.]. Microbiol. 3,841-846. Peterson, D. H. (1963). I n “Biochemistry of Industrial Microorganisms” (C. Rainbow and A. H. Rose, eds.), p. 538. Academic Press, New York. Prescott, S. C., and Dunn, C. G. (1959). “Industrial Microbiology,” pp. 723, 844. McCraw-Hill, New York. Ramachandran, S., and Gottlieb, D. (1961).Biochim. Biophys. Acta 53,391-396. Randerath, K. (1963). “Thin Layer Chromatography,” p. 111. Academic Press, New York. Raper, J. R. (1966). In “The Fungi-An Advanced Treatise” (G. C. Ainsworth and A. S. Sussman, eds.), Vol. 11, p. 473. Academic Press, New York. Raper, K. B., and Alexander, D. F. (1945).Mycologia 37,499-525. Rieske, J. S. (1967). In “Antibiotics-Mechanism of Action” (D. Gottlieb and P. D. Shaw, eds.), Vol. I, p. 542. Springer, New York. Roper, J. A. (1966). In “The Fungi-An Advanced Treatise” (G. C. Ainsworth and A. S. Sussman, eds.), Vol. 11, p. 589. Academic Press, New York. Samiinakova, A. (1966).]. Inuert. Pathol. 8,395-400. Schleg, M. C., and Knight, S. G. (1962).Mycologia54,317-319. Sehgal, S. N., Singh, K., and Vezina, C. (1963).Steroids 2,93-97. Sehgal, S. N., Schilling, G., Singh, K., and Vkzina, C. (1966a). Intern. Congr. Microbiol., 9th, Moscow p. 242. (Abstr.) Sehgal, S. N., Singh, K., and Vezina, C. (1966b).U.S.Patent No. 3,294,647. Sehgal, S. N., Singh, K., andVezina, C. (1968).Can.]. Microbiol. 14,529-532. Singh, K., and Rakhit, S. (1967).Biochim. Biophys. Acta 144,139-144. Singh, K., Sehgal, S. N., and VBzina, C. (1962). Intern. Congr. Microbiol., 8th, Montr e d p. 21. (Abstr.) Singh, K., Sehgal, S. N., and VBzina, C. (1963).Steroids 2,513-520. Singh, K., Sehgal, S. N., and VBzina, C. (1965).Can.]. Microbiol. 11,351-364. Singh, K., Sehgal, S. N., and VBzina, C. (1966). Intern. Congr. Microbiol., 9th, Moscow p. 140. (Abstr.) Singh, K., Sehgal, S. N., and Vezina, C. (1967a).Can.]. Microbiol. 13,1271-1281. Singh, K., Sehgal, S. N., and VBzina, C. (1967b).U.S. Patent No. 3,305,453. Singh, K., Sehgal, S. N., and Vezina, C. (1968).Appl. Microbiol. 16, 393-400. Staples, R. C. (1957).Contnb. Boyce Thompson Inst. 19,19-31. Stoudt, T. H. (1960).Aduan. Appl. Microbiol.2,183. Sussman, A. S. (1966a).In “The Fungi-An Advanced Treatise” (G. C. Ainsworth and A. S. Sussman, eds.), Vol. 11, p. 733. Academic Press, New York. Sussman, A. S. (1966b). I n “The Fungus Spore” (M. F. Madelin, ed.), p. 235. Butterworths, London and Washington, D. C. Tokoro, M., and Yanagita, T. (1966).]. Gen. Appl. Microbiol. 12,127-145. Tubaki, K. (1966). In “The Fungi-An Advanced Treatise” (G. C. Ainsworth and A. S. Sussman, eds.), Vol. 11, p. 113.Academic Press, New York. Turian, G . (1966). In “The Fungi-An Advanced Treatise” (G. C. Ainsworth and A. S. Sussman, eds.), Vol. 11, p. 339. Academic Press, New York. Tuveson, R. W., West, D. J., and Barratt, R. W. (1967).J. Gen. Microbiol. 48,235-248.

268

CLAUDE VkZINA, S. N. SEHCAL, AND KARTAR SINGH

Underkofler, L. A. (1954).In “Industrial Fermentations” (L. A. Underkofler and R. J. Hickey, eds.), Vol. 11, p. 97.Chem. Publ. Co.,New York. Van Etten, J. L., Molitoris. H.P., and Gottlieb, D. (lssS).J. Bacteziol. 91,169-175. Vbzina. C. (1956).Ph.D. Thesis, Univ. of Wisconsin, Madison, Wisconsin. Vbzina, C., Sehgal, S. N., and Singh, K. (1963).Appl. Mfcrobfol. 11,50-57. Vbzina, C., Singh, K.. and Sehgal, S.N. (1965).Mycologia 57,722-736. Vuillemin, P. (1910).Bull. SOC. Sci. Nancy 11,129-172. Vuillemin, P. (1912). “Les Champignons, Essai de Classification.” Doin, Paris. Waksman, S. A. (1959).“The Actinomycetes,” Vol. 1, p. 17.Williams & Wilkins, Baltimore, Maryland.

Microbial Interactions in Continuous Culture

HENRYR . BUNGAY.I11 Department of Environmental Systems Engineering Clemson Universitig. Clemson. S C

. .

and

MARY Lou BUNGAY I . Introduction................................................................

. Historical ............................................................. B. Research Approaches ............................................. C. Theory ................................................................. I1 . Nomenclature.............................................................. A . Definitions ........................................................... B. The Symbiotic Index ............................................. A

C. Class Designations ................................................ 111. Systems of Defined Microbial Composition .................... A. Pure Cultures ....................................................... B Competition ......................................................... C Commensalism ..................................................... D Inhibition ............................................................. E Predation ............................................................. F. Mutualism ............................................................ G . Neutralism ........................................................... H Parasitism............................................................. I Synergism ............................................................ J Complex Interactions............................................. IV. Multiculture Systems ................................................... A Mixture of Defined Cultures ................................... B Undefined Mixed Cultures ..................................... V Computer Models ........................................................ VI Conclusion .................................................................. References ..................................................................

. . . .

. .

. . . . .

269 271 272 272 275 275 276 276 277 277 278 279 279 280 281 281 281 282 282 282 282 284 287 288 288

.

1 Introduction

Mixed culture phenomena are not merely composites of the pure culture behavior of the organisms present . The performance of a complex microbial process depends upon interactions between its species and strains. Winogradsky and Beijerinck. key figures in mixed culture research. began about 70 years ago. but until now technical difficulties prevented a concerted attack on the problems of mixed culture environments . Major barriers to such studies have been the tedious counting procedures available to the bacteriologist and the complex mathematical analysis needed to uncover significant relationships . 269

270

HENRY R. BUNGAY, 111 AND MARY LOU BUNGAY

However, new techniques for enumeration, modern instrumentation, and computers for simulation and data treatment permit more sophisticated approaches. A recent volume of this series of Advances had three separate commentaries on the relevance of mixed culture phenomena to various areas of microbiology (Brock, 1966a; Pipes, 1966; Hoffman, 1966). Nevertheless, very little is known about fundamental characteristics of mixtures of microorganisms; basic research is needed on growth rates, survival, population dynamics, physiology, and ecology. A proper emphasis is required for mixed culture research. It would be wrong to follow the early ecologists in simply isolating and characterizing the constituents of a mixture. While identification is essential, pure culture studies are of limited relevance to the mixed population. The aim is environmental control of the activity of populations of the desired organisms (Hawkes, 1963). To achieve this, we need to know (a) what the desired organisms are, (b) what events are rate or yield limiting, (c) how energy is transferred. The concept of energy flow and energy balances in microbial communities is discussed by Brock (1966b). A one-sided approach cannot shed much light on basic phenomena. For example, research on biological waste treatment processes usually concentrates on chemical details without identifying or describing the microbiological flora and fauna. The volume or weight of biological solids may be recorded, but there is seldom any measure of what is living or dead, much less any attempt at classification. Microorganisms have vastly difkrent nutritional requirements. They exhibit a wide range of growth rates and thrive at various conditions of pH, temperature, ionic strength, and the like. In spite of this diversity, species persist in nature throughout times of hardship and can flourish when favorable conditions occur. Thus the occurrence of unusual physical or chemical conditions or of rare nutritional circumstances leads to population changes in which different species achieve predominance so that relatively efficient microbial processes can continue. This adaptability of mixed populations is a key to the study of the microbiology of natural environments. Understanding of the mechanisms underlying this adaptation must come from investigations of population dynamics and microbial interactions. Complex mixed culture systems present a formidable problem for research because of the multitude of parameters. Some simplifications which can be made are (a) fixing or controlling such factors as pH, temperature, dissolved oxygen; (b) feeding defined media either con-

MICROBIOL INTERACTIONS IN CONTINUOUS CULTURE

27 1

tinuously or on a schedule; (c) restricting the number of species either by natural selection or by inoculating known organisms into an aseptic system. Although both batch and continuous flow conditions are important in nature, continuous flow is often better for drawing valid conclusions in the laboratory because either a steady state is obtained or else the variables stay within relatively narrow ranges. In contrast, batch cultures undergo drastic and simultaneous changes in substrate concentration, product accumulation, and possibly in pH, dissolved oxygen, and other key parameters. A major complication is the extreme range of physiological ages from new cells to old or dead cells in a batch system, while continuous cultures at reasonable dilution rates have most cells washed out before they become very old. Because of the difficulties in interpreting batch culture experiments, this article will concentrate on continuous culture techniques.

A. HISTORICAL Over the years there have been sporadic reports of batch cultures inoculated with two types of organisms. Some food and beverage fermentation use mixed inocula intentionally to improve flavor, lower pH to keep out undesired species, or to achieve some other result (Bungay and Krieg, 1966). Very few observations of batch cultures will be mentioned here because only quite gross conclusions can be drawn in such changing environments. Nurmikko (1954) showed that an association exists between different strains of lactic acid bacteria in a medium of known chemical composition. When certain vitamins and amino acids essential for the growth of two lactic acid bacteria were omitted from the synthetic medium, each strain produced the chemical factors needed by the other. Nurmikko (1955) demonstrated the same relationship of these strains when separated by a dialysis membrane, and populations and biochemicals could be determined for each compartment. Schaumburg and Kirsch (1966) have devised a technique for recirculation of media to a membrane-segmented vessel and have demonstrated the action of Escherichia coli in maintaining anaerobiasis for the culture of Methanobacillus omelianski. Azuma (1961) studied symbiotic relationships between Rhodopseudomonas capsulatus and Axotobacter vinelandii. The combination fixed three times more molecular nitrogen than did either pure culture. Mixed cultures are included in a series of reviews on continuous culture (Malek and Fencl, 1961; Malek and Beran, 1962; Malek and

272

HENRY R. BUNGAY, I11 AND MARY LOU BUNGAY

Ricica, 1964, 1965, 1966). There are reviews on microbial ecology by Hungate (1960,1962)and books by Wood (1965)and Brock (1966b).

B. RESEARCH APPROACHES Some of the methods for studying mixed cultures in continous flow are 1. Observing a natural environment such as a stream, a ruminant animal, etc. 2. Modeling a mixed culture environment, i.e., small-scale activated sludge units, rumen models, etc. 3. Operating an open vessel with ambient reinoculion 4. Operating an aseptic system with complex mixed inoculum 5. Operating an aseptic system with inoculation of a few known cultures. Various apparatus configurations could be multistage units, recycle of part of the organisms from the overflow, and various schemes for the flow regime or the aeration. The theory for some of these arrangements has been considered by Herbert (1961).Mathematically, the simplest research device for continuous culture is a small, wellmixed unit in which the overflow may be assumed to have the same concentration as the homogeneous contents of the vessel. It is not practical to attempt perfect mixing in a large vessel because the violent agitation required would damage the organisms. All units that are not perfectly mixed violate a key assumption in the customary mathematical formulations, and this may be important or not, depending upon the particular circumstances and interpretations of the data.

C. THEORY Pure culture phenomena must be considered in order to provide a foundation for understanding behavior in mixed culture. An excellent introduction has been given by Powell (1965).The best-known treatment is that of Herbert, Ellsworth, and Telling (1956)based upon work of Monod (1942). In a perfectly mixed vessel with a sterile feed solution, mass balance considerations give the equations:

dx -= dt

px - DX


273

where x = mass concentration of organisms; D = dilution rate; p = specific growth rate; S = substrate concentration (So = concentration in feed); Y = yield constant. The specific growth rate is a function of the limiting substrate concentration as approximated by the equation:

where hx is the maximum growth rate and K is a constant. A more detailed treatment would consider such factors as cell death, maintenance energy, and variations in the yield constant. The relation of p to S must be obtained experimentally. Equation (3) fits the experimental data for specific growth rate versus concentration fairly well, but it is sometimes more accurate and convenient in computer work to use the data themselves in a function generator expressing p =f(S). The theory for competition between a contaminant or a mutant strain and the initial strain has been developed (Powell, 1958; Renneboog-Squilbin, 1967). There are mass balances such as Eq. (1) for each constituent and additional utilization terms in Eq. (2). The relationship of its specific growth rate to substrate concentration is different for each strain. Competition in continuous flow apparatus should lead to establishment of the faster growing strain and complete washout of the other (Powell, 1958). Unfortunately, some of the theory for mixed cultures has been more concerned with mathematics than with reality. There has been considerable development of the theory of prey-predator systems based on the theory of Volterra (1931). In continuous culture, Volterra's approach gives the equations: -= dH

dt

pH - DH - klHP

-= ddtp

k2HP - DP

(4)

(5)

where H = concentration of prey (hosts); P = concentration of predators (parasites); p = specific growth rate of prey; D = dilution rate; kl = killing efficiency based upon encounters; kz = predator growth constant; t = time.

274


The assumptions are that predator growth and death of prey both depend upon encounters between prey and predators as indicated by the terms containing H P . The specific growth rate is assumed constant because predation can keep the prey concentration low enough that critical nutrients are little depleted and p is near its maximum. The terms DH and DP represent overflow of organisms from the vessel. Except at very low dilution rates, organisms probably leave before becoming old, and death rate terms can be neglected. Oscillatory population levels are predicted, and this agrees with observations of real systems (see later sections). It has been pointed out that this model is equivalent to a predator growth rate linearly related to its food concentration (Gause, 1934), and a better approach is that there is a maximum growth rate for predators as found by Proper and Gamer (1966) for the ciliate Colpoda stenii feeding on E . coli. The equations based on a Monod type of relationship of predator growth rate to substrate (prey)concentration are:

dH dt

-=

PPP ~ H -HDH -Y

where p H= specific growth rate of prey; p p = specific growth rate of predators; Y = yield constant, prey consumed per predator division. The first two equations stem from mass balances, and Eq. (8) would be similar to Eq. (3). These equations also permit oscillatory solutions with shapes similar to those from the Volterra model (Bungay and Cline, 1967). Effects of limiting substrate concentration on the growth rate of prey can be built into the model. Models for other types of interactions can also be developed from mass balance considerations, but at some point a relationship such as Eq. (8)is needed for growth rate versus vitamin or inhibitor concentration or temperature or some other factor. The model is a quick and powerful tool for testing ideas about these relationships. A particular shape can be assumed or values can be guessed at, and the computer solution can be compared to real data. A fit must not be construed as


275

proof of the validity of the model. The Volterra model with its false assumption about predator growth rate fits data about as well as the model with the Monod relationship. II. Nomenclature

The terminology for interactions between organisms is especially prolific and confusing, and the type of interaction may change as the population levels change. Brock (1966b) does not feel that rigid criteria are useful or even possible, but that the important thing is to understand what is happening to the population. This view seems questionable because anything must be defined before it can be quantitated. A. DEFINITIONS Many terms for microbial interactions have been proposed but most of those shown in Table I had early origin (de Bary, 1879) and are well known. TABLE I COMMON TERMSFOR MICROBIALINTERACTIONS Interaction

Paraphrased definitions

Neutralism Commensalism Mutualism Competition Amensalism Parasitism Predation

Lack of interaction One member benefits while the other is unaffected Each member benefits from the other A race for nutrients and space One adversely changes the environment for the other One organism steals from another One organism ingests the other

Note that the term symbiosis is not in the list. A broad definition of symbiosis (Greek roots, “living together”) to designate any type of coexistence is preferable to using it for beneficial associations [Webster’s International Dictionary, 19661, A serious limitation of this terminology is the difficulty of stating exactly what is happening in complex multiculture systems where several interactions can occur simultaneously. Interactions can change, as in the case of one organism stimulating another to the extent that it becomes a competitor. Should this be called competitive commensalism, competitive stimu-

276

HENRY R. BUNCAY, III A N D MARY LOU BUNGAY

lation, or some other term? Such terminology is qualitative, imprecise, and unsuitable for detailed study of important systems.

B. THE SYMBIOTIC INDEX We wish to propose a method for reporting population data from mixed culture experiments; it provides a quantitative measure of how much a species is helped or hindered by the other symbionts. Consider a symbiotic index defined as: S.Z. =

growth of a species in mixed cultures growth of that species in pure culture (same initial medium and conditions)

In a batch culture, the S.Z. should change with time. As growth begins, no interactions would be evident because of low populations and thus S.Z. would equal 1. Later on, various biochemicals would be elaborated, and inhibitions or stimulations could be observed. Subscripts could denote the times, and at extended times when growth is complete, the S.Z. should become constant. If there is no detectable growth in pure culture, the population should be assumed equal to 1.0 so that S.Z. is not infinity. The S.Z. would be a quantitative term describing growth, and any anomalies could be studied in a meaningful way. There would be a symbiotic index for each member of a mixed culture. The symbiotic index would apply very nicely to continuous cultures that achieve steady state, but there would be difficulties whenever the population levels oscillate. We suggest using a weighted average value between the population extremes. Of course, when population dynamics of mixed cultures is the whole point of the research, graphical presentation of the time history of the populations would be required. C. CLASSDESIGNATIONS For reporting research on multiculture systems in either batch or continuous culture environments, there is a very real problem in writing an accurate title that does not use several lines of words. A system is needed for succinctly stating what type of mixed culture was studied. We propose the class designations shown in Table 11. The use of such symbols can be illustrated by an example from our laboratories. We have a persisting mixed culture in which a yeast elaborates vitamin factors for two different bacteria and all three


277

are ingested by a predatory myxameba. A possible title might be “Oscillatory Populations in a Four-Member Continuous Culture, Class 1S2, 1P3.” This would be deciphered as one species stimulated two others and one species preyed upon three others. TABLE I1 PROPOSEDCLASS DESIGNATIONS“ Symbol

Meaning

S

Stimulated Inhibited Predatory Parasitic Synergistic effects

I P W E

” Numbers would be used to indicate degrees of involvement 111. Systems of Defined Microbial Composition

Except for the dynamic aspects of adjustment to stresses, the behavior of a pure culture in a chemostat is reasonably well understood. A brief discussion of pure cultures will serve as an introduction to mixed cultures.

A. PURE CULTURES The classic treatment of Herbert et al. (1956) has been tested against laboratory data for several bacterial species in continuous culture, and the results usually confirm the theory quite well. However, cases have been found in which the variation in steady-state population with dilution rate differs markedly from their predictions. While some of these deviations result from poor mixing or from growth attached to surfaces, others arise from the effects of endogenous metabolism and from diversion of substrate into pathways for the synthesis of reserve materials. Diauxie or catabolite repression (Magasanik, 1961) seems to occur in continuous culture as well as in batch culture. When a normal substrate such as glucose is present, a less common substrate such as galactose is used very slowly or not at all until the utilization of glucose reduces its concentration to a low level. Cell size and composition change with dilution rate and with temperature (Dicks and Tempest, 1966). The variations with dilution

278


rate reflect the differences in young cells and old cells because of the various average physiological ages established. The effects of different limiting substrates were studied by Dean and Rogers (1967). Dynamic analysis of pure cultures has particular importance to mixed culture systems. Although pure cultures at fixed dilution rates tend to steady states, most mixed culture systems appear to fluctuate in populations at fixed dilution rates. Thus the interpretation of rate phenomena in a mixture will depend upon some knowledge of how a pure culture responds to upsets. Herbert (1964) made the important observation that specific growth rate adjusts almost instantaneously to a change in dilution rate. Looking into this more deeply, Mateles et aZ. (1965) found for E . coli an almost instantaneous change in growth rate but the final adjustment took several hours and included a period of overshoot. The bacteria increased their rate of protein synthesis immediately after a shift up in dilution rate. Gilley and Bungay (1967) found a decaying oscillation in numbers of the yeast Saccharomyces cerevisiae in establishing a new population level following an abrupt change in dilution rate. Frequency response analysis of pure cultures has been attempted with some success (Fuld et aZ., 1961; Gilley and Bungay, 1968). Such studies should provide the coefficients and differential equations for microbial growth responses to stresses.

B. COMPETITION Competition between bacterial strains has been studied fairly extensively because of its importance to the isolation of mutants. Here the purpose is to enrich the population in numbers of mutants so that they may more easily be isolated. The topic of enrichment has been reviewed for both batch and continuous culture by Schlegel and Jannasch (1967). Analyses by Powell (1958) and Moser (1958) of the fate of contaminants or of mutants in a continuous culture were concerned with competition for a single limiting substrate. Provided the contaminant or mutant was not washed out before its first few divisions, the more rapidly growing organism should take over, and the slower growing one should be washed out. In our laboratories, some studies of competition have showed that adherence to the walls of the vessel can be a key factor in competition (unpublished). A slower growing species can persist in appreciable numbers in competition with rapidly growing organisms if the slow grower is continually reinoculated from the wall growth into the main liquid bulk.


279

Jannasch (1967) has studied competition leading to enrichment for cases in which the curves for specific growth rate versus limiting nutrient concentrations cross. The organism having the higher growth rate at the selected conditions will predominate. In a chemostat at steady state, dilution rate and specific growth rate are equal while substrate concentration must be adjusted. Therefore the organism to be favored depends upon the dilution rate range. C. COMMENSALISM Elsworth (Powell, 1958) observed that Pseudomonas pyocyanae grew very poorly alone in the medium but was a tenacious contaminant requiring products of the metabolism of Bacterium cloacae. Contois and Yango (1964) found a yeast-bacterium commensalism in which an unidentified yeast required factors produced by Aerobacter aerogenes. At dilution rates greater than about .03 hours-l, the yeast was washed out of the system. It would appear that this was not competition for a single substrate, and at low dilution rates various nutrients could accumulate for the yeast. At higher dilution rates, a different nutritional pattern could render the slower growing yeast unable to compete successfully with the bacteria. Shindala et al. (1965) found that S. cerevisiae was not overgrown by Proteus vulgaris on a medium deficient in niacin, and the yeast population was the same in pure as in mixed culture. Adding niacin or related biochemicals allowed the bacteria to take over. Another situation has been reported by Frederickson and Tsuchiya (1967)with Lactobacillus caseii and S . cereuisiae on a medium deficient in riboflavin; however, this was not a strict commensalism because the yeast population was lower in mixed culture than in pure culture. It would seem that a commensalism in continuous culture tends to a steady state even though there is competition for a single substrate.

D. INHIBITION Inhibition of one microbial species by another is basic to the study of antibiotics. An antibiotic can inhibit at trace concentrations, but many other types of toxic factors could accumulate during metabolic activities. A common application of antibiosis in mixed culture is not intentional but has great economic importance. Fermentation processes for antibiotics are often challenged by contaminants but are self-protecting to a considerable extent. The early hours before much antibiotic accumulates are most critical to contamination, and narrowspectrum antibiotics are much less protective than those with a broad

280

HENRY R. BUNCAY, III AND MARY LOU BUNCAY

spectrum. Thus certain industrial fermentations are much more susceptible to contamination than are others. Shindala (1964)and Gilley (1965)studied competition between E. coli and S. cerevisiae. When inoculated into a thriving yeast culture, the bacteria took over rapidly, but the decline in yeast numbers was more precipitous than expected. In a two-stage system with a pure culture of yeast being fed to a mixed culture of yeast and E. coli, the exit stream had the same numbers of yeast as in the feed stream of the second stage. This suggests that E. coli inhibits the growth of S . cereuisiae. Furthermore, a different yeast, Torula sphaerica, persists with Proteus vulgaris in mixed culture because the slower growing yeast is reinoculated from films on the vessel surfaces. In contrast, E. coli added to a thriving T . sphaerica continuous culture took over rapidly and caused the wall growth of yeast to be sloughed off (unpublished). Ransom et al. (1961)have found Vibrio cholerae to be suppressed when grown in continuous culture with enterococci; this inhibition was not demonstrable by classic plating techniques. The pure culture and mixed culture behavior of Serrutia marcescens and an unnamed yellow organism were studied by Leal (1964). In the mixture, S. marcescens predominated and was at the same concentration observed in pure culture. The concentrations of the yellow organisms were about one-tenth that obtained in pure culture. Self-inhibition caused an interesting interaction in a mixed culture of lactobacillus strains (Contois and Yango, 1964).The faster growing strain caused a lowering of the pH to a range which was inhibitory to itself but not to the other strain. This second strain continued to grow but did not produce organic acids. With the pH inhibition removed by dilution, the first strain again became predominant to initiate a new cycle of changes in populations and in pH. A mutual inhibition of Streptococcus salivarius and Veillonella alcalescens was reported by Parker and Snyder (1961).Pure cultures grown in separate chemostats were fed to a mixed culture vessel along with a feed of rich, fresh medium. The dilution rate in this vessel was greater than the maximum specific growth rate of either species. There was clear evidence of interaction because each grew considerably more slowly in the mixture than in pure culture controls.

E. PREDATION Contois and Yango (1964)reported a persisting mixed continuous culture of Aerobacter aerogenes and the myxameba Dictutosteliurn discoideum but there were insufficient enumerations to detect any


28 1

oscillations in populations. Drake et al. (1966), Chen (1967), and Koelling (1966) studied these myxameba with E. coli and other bacteria and found cyclic variations in population such as Gause (1934) had found in batch predation systems. The predators consume so many bacteria that they cut off their own food supply and decrease in numbers. Reduced predation then allows the bacteria to achieve high numbers which stimulate myxameba growth and another cycle starts. The oscillations in predator growth lag behind those of the prey.

F. MUTUALISM Yeoh (1967), using a medium which would not support Proteus uulgaris or Bacillus polymyxa singly but in which a combination grew well, found that P. vulgaris would grow alone if niacin were added to the basal medium and B. polymyxa grew alone if biotin were added. At several dilution rates continuous symbiotic populations oscillated in numbers and in relative predominance; no steady states were achieved. P. vulgaris produces a substance which accumulates to inhibit growth of B . polymyxa. Cessation of growth of the latter halts growth of P. vulgaris as well because of the vitamin interdependency. Continued feeding of the continuous culture diluted the inhibitory substance so that rapid growth of both organisms started a new cycle of the oscillation. The inhibitor was quite labile and was also destroyed by proteolytic enzymes. Addition of chymostrypsin to the continuous culture stopped the oscillations, the populations of each member rose and then achieved a steady state. G. NEUTRALISM It is not inconceivable that two or more species might thrive wholly independently in mixed culture, though such a relationship would be very difficult to confirm. Yogurt starter strains of Lactobacillus and of Streptococcus grown together in a chemostat have been shown to achieve similar population values as the same organisms grown separately (Lewis, 1967).

H. PARASITISM Contois and Yango (1964) describe the introduction of a coliphage into a continuous culture of E. coli. The phage titer rose rapidly to a maximum, then declined rapidly to a level that was fairly constant with time. The E. coli showed a marked drop to a very low population level at which they persisted indefinitely. It would be interesting to examine this steady state to see whether all the E. coli are infected or if

282

HENRY R . BUNGAY, 111 AND MARY LOU BUNGAY

a certain percentage is resistant but yields some progeny which maintain the coliphage. I. SYNERGISM

No reports of synergism in continuous culture have come to our attention. This class of interaction differs from all the others in that there is not necessarily any population response, and a biochemical output serves for the definition. There is great economic potential for microbial synergisms to produce substances which cannot be synthesized by a single type of organism. The synthesis of carotenes in batch culture by the cooperative metabolism of plus and minus strains of Blakeslea trispora illustrates a type of synergism.

J. COMPLEXINTERACTIONS It should be noted that several of the interactions which have been discussed do not fit any of the simple definitions. Mutualism complicated by inhibition, and commensalism leading to competition are only two of the many possible interactions. Even neutralism, the absence of interaction, could not occur except at low populations because there would be competition for living space as one species began to crowd the other. A dependence of self-inhibition on population level was observed in hydra by Davis (1966).When high populations cause crowding, hydra release a factor which inhibits growth and regeneration.

IV. Multiculture Systems

A. MIXTURE OF DEFINEDCULTURES

1. Single Vessel Many natural environments are probably dominated in population by relatively few species. Half a dozen or so species or strains might constitute 95% of the population; this would hold true especially where selective enrichment factors are operating, such as high temperature, extreme pH, or a toxic factor. Inoculating several known cultures into a continuous culture vessel would provide a model system whose behavior might be meaningful in relation to complex natural environments. Unless there is something other than competition for a single limiting substrate, the expected result would be that one species would win out and the others would wash out. Some preliminary studies by our group used various interrelationships to es-


283

tablish persisting mixed cultures (Olsen, 1966). The results are summarized in Table 111. TABLE 111

PERSISTINGCONTINUOUS MULTIPLE CULTURES Multiculture System a. Saccharomyces cereuisiae

Class

Comments

1S1, 1P2

The bacteria cannot displace the yeast because the yeasts provide an essential vitamin. The myxameba ingest both. All populations are oscillatory.

1s2

Each bacterial species requires a different vitamin elaborated by the yeast. All populations are oscillatory.

1S2, 1P3

Same vitamin dependencies as above for bacteria. Myxameba feed on all three. Populations are oscillatory.

Proteus vulgaris Dictytostelium discoideum

b. S. cereuisiae P. vulgaris B . polymixa c. S. cereoisiae

P . vulgaris B . polymxa D. discoideum

Since the few defined multiculture systems that have been studied are oscillatory in populations and some of the binary continuous cultures are also oscillatory, it is not surprising that natural environments display such wide fluctuations in relative abundance of species.

2. Multivessel Apparatus has been described for feeding six pure cultures, some of which can be cultivated anaerobically, to the mixed culture vessel (Parker, 1966). These pure cultures were not grown in strict chemostasis (i.e., all nutrients in excess except the limiting nutrient) but rather the nutrients were all nearly exhausted so that the mixed cultures would not receive leftovers from the various pure cultures and thus have an unknown nutritional pattern. Fresh nutrients fed to the mixed cultures permitted some defined program of nutrition, and populations were adjustable by varied pumping rates of the pure cultures. The dilution rate in the mixture culture vessel was greater than the maximum specific growth rate of any of the constituent species. In a culture of Staphylococcus aureus, Streptococcus salivarius, and Veillonella alcalescens, the generation time for S. aureus was little different than in pure culture while the growth rate for V. alcalescens increased by a factor of two, and that of S . salivarius decreased. A

284

HENRY R. BUNCAY, I11 AND MARY LOU BUNGAY

remarkably short generation time of 8.1 minutes was observed for S . saliuarius in the mixture; 30 runs confirmed this rapid growth rate.

B. UNDEFINED MIXED CULTURES Almost any research on biological waste treatment units or the rumen has relevance to microbial interactions in continuous culture. We have chosen to be highly selective in reviewing such work because it is covered elsewhere and was usually not performed in a manner which gave information about population dynamics. 1 . Population Fluctuations

Cassell et al. (1966) fed powdered milk suspension continuously to well-mixed, open culture vessels. The walls were scrubbed periodically to remove attached oganisms. Low dilution rates typical of waste treatment units were selected at 0.22, 0.16, 0.085, 0.028,and 0.013 hour-’ and held for periods of several weeks. Various pigments could be extracted from the biological growth, and these were used as indices of types of bacteria. Each pigment concentration fluctuated over a wide range of values. Typically, a certain pigment might be pronounced for several days then become almost undetectable as another became predominant. Photomicrographs confirmed that the types of bacteria in the units were shifting in relative predominance. Conventional performance indices such as suspended solids and soluble effluent COD also fluctuated. In large-scale mixed culture processes, flow changes, loading variations, and diurnal changes would accentuate the oscillations observed in the laboratory systems. Gaudy et al. (1967)fed a model activated sludge unit continuously with a glucose-ammonium sulfate-salts solution at considerably higher dilution rates than in the experiments mentioned above. Unfortunately the feed concentrations varied somewhat, but measurements of biological solids confirmed that mixed populations have gross, random changes in numbers. Chian and Mateles (1966) observed oscillations of cell concentration and chemical oxygen demand in a heterogeneous population in single-stage continuous culture. They pointed out that the design of an activated sludge uqit should be based on averages taken over extended periods of time rather than values obtained from a steady-state assumption. These population fluctuations are different from “ecological succession” (Hawkes, 1963)observed in batch cultures in which new species gain predominance as others produce nutrients or die and release cell substances.


285

2. Sequential Use of Substrate Mateles (1967) provided an undefined mixture of organisms in which Pseudomonas seemed predominant with a continuous flow of medium containing both glucose and lactose. At low dilution rates, both sugars were consumed. At higher dilution rates, however, glucose was used preferentially so that lactose was little depleted. This is another example of “catabolite repression” (Magasanik, 1961); batch mixed cultures have also demonstrated this effect (Gaudy, 1962; Bhatla and Gaudy (1965);Stumm-Zollinger, 1966). 3. Population Enrichment This aspect of mixed cultures has been reviewed by Schlegel and Jannasch (1967).High dilution rates in an ambiently inoculated mixed culture tend to favor attached organisms because free-swimming organisms are washed out before dividing (Edwards, 1966). Prakasam and Dondero (1967) have used certain differential reactions as indications of population changes in a model activated sludge unit adapting to sorbitol. Enrichment would tend to reduce heterogeniety of a natural population. However, other factors must be combatting enrichment since there is a wide diversity of species in nature. Fluctuating population behavior may well have implications for the patterns of microbial evolution. Larger metabolic pools and faster control mechanisms would all tend to reduce fluctuations and would seem quite possible types of mutations. Since these oscillations have persisted to this point in the history of life on our planet, it would appear that there is some reason that individual species have not achieved greater efficiency. Systems analysts speak of suboptimization of system components not always giving an optimum total system; this may be a microbiological example of this phenomenon. Eliminating fluctuations would probably give optimum behavior of a species but not of a complex multiculture system. Without fluctuations and mutual inhibitions, the faster growing and least fastidious species would crowd out the others to the point where so little adaptability would remain that a change in substrate or conditions could cause the entire culture to be lost. Other factors which aid in maintaining heterogeneity are layered growth on surfaces; clumping of organisms to provide zones of varied oxidation-reduction potential; varied respiratory properties at different physiological ages; and cysts, spores, or other resting forms to retain dormant species in the culture.

286


The normal fluctuations of mixed cultures may help to explain malfunctions of waste treatment units such as “bulking” of activated sludge and “sour” anaerobic sludge digestors. A bad pickle fermentation or other industrial mixed-culture process may not result from an error by the operating personnel but rather from the peaks of some interactions reinforcing each other to get an undesired species off to a rapid start.

4 . Growth on Surfaces in Continuous Culture Growth on surfaces is studied because of its importance in trickling filters for treatment of wastes and in the benthos of lakes and streams. In situations where the respiration in a slime on a surface is limited by the diffusion of oxygen, biochemical indices may be little affected by population dynamics in the slime (Saunders, 1966; Maier et al., 1967). However, there are extreme fluctuations in population both in the slime and in medium in contact with it (Saunders, 1967). While population dynamics of organisms adhering to surfaces may be a minor factor for some dihsion-limited situations, it may be of great importance to events in the bulk of the liquid medium (see Section 111,B). After adding glass wool to increase the surface in continuous cultures of Serratiu marcesens, Larsen and Dimmick (1964)estimated that u p to 90% of the cells in the liquid may be progeny of adhering organisms. Munson and Bridges (1964) report “take over” of continuous cultures of tryptophan-requiring mutant E. coli by revertant prototrophs which become attached to the vessel surfaces. Growth of attached organisms is of great importance to soil microbiology (Alexander, 1964). Since a cell at a surface cannot exchange equally on all sides as it can when suspended in the medium, its respiration rate will usually be reduced. On the other hand, the microenvironment in the slime or film may be different from conditions in the bulk of the liquid. The different pH or oxygen tension may provide a refuge for certain organisms and thus be a major factor in their ecology. Continuous flow perfusion of soil is discussed by Macura (1964). This interesting approach is quite different from the other systems covered here, and space does not permit its inclusion.

5. Enuironmental Efects A laboratory model of the rumen often is not run as a typical chemostat, but rather as a continuous flow of a salivalike solution with intermittent scheduled addition of solid nutrients. Changes in the p H of the artificial saliva are reflected by pH changes in the model rumen;


287

low pH gave less fermentation as measured by volatile acids and lower populations using DNA concentration as an indicator (Slyter et al., 1966). While populations at pH 6.0 to 6.7 had many of the organisms commonly seen in a normal rumen, the populations at pH 5.0 showed a high proportion of organisms different from the major species in the rumen. A protozoa-free rumen model was produced by adding copper sulfate (Slyter and Wolin, 1967). Methane production decreased and propionate production was slightly increased. Since there were relatively few protozoa before treatment, the copper sulfate probably affected the bacteria directly rather than disturbing the population dynamics. Other aspects of microbial ecology of the rumen are reviewed by Hungate (1960). A secondary effect of predation in laboratory activated sludge units was reported by Cooke and Ludzak (1958). In the absence of fungi which had been preying upon rotifers, there was decreased efficiency of removal of nitriles. It was postulated that a shift in the bhcterial population took place as the level of predation by rotifers changed. Continuous culture of fecal organisms to detect interactions was conducted by Zubrzycki and Spaulding (1958).

V. Computer Models Computer models provide methods for testing new ideas far more rapidly than by experiments with living cultures. Constructing the models forces the investigator to develp an understanding of his systems, and discrepancies between the model and actual data point out weaknesses in the theory. Conflicting theories may be tested in such a way that microbiological experiments can be designed based on the computer results. Analog and digital simulation have been used. Analog techniques appear better suited when a broad range of parameters are to be tested, but digital methods have the great advantage of not requiring scaling to fit a voltage span. Simulation of some microbial interactions by a small analog computer has been shown to give population curves which resemble data from experiments with living systems (Bungay, 1968). Such simulations permit rapid testing of parameters and can pilot research into fruitful channels by predicting which variables are most significant and by indicating how much of a shift in a variable will produce a measurable response. The repetitive operation feature of analog computers allows an oscilloscope display of the results

288

HENRY R.

BUNGAY,

111 AND MARY LOU BUNGAY

so that the curves appear to flex and move as the coefficient knobs are turned. Points from actual data can be marked on the oscilloscope screen and fitted rapidly by “tuning” the coefficients. These approaches provide considerable insight, and construction and improvements of the mathematical models force the investigator to think deeply about his systems. VI. Conclusion Research on mixed cultures is in its infancy. However, the importance of the rumen, biological waste treatment, the oral cavity, natural waters, soil microbiology, some industrial mixed culture processes, and other mixed culture environments ensures that this research will expand. This will be sophisticated work requiring many types of talent and interdisciplinary cooperation. ACKNOWLEDGMENT

Grant Number ES-00025from the U.S. Public Health Service has enabled the senior author to devote time to mixed culture studies for several years. Mr. E. G. Willard at Virginia Polytechnic Institute provided technical assistance that was invaluable. REFERENCES Alexander, M. (1964).Ann Reo. Microbiol. 18,217-252. Azuma, 0.(1961).Nature 192, 1207-1208. Bhatla, M. N., and Gaudy, A. F., Jr. (1965).Appl. Microbiol. 13,345-347. Brock, T. D. (1966a).Adoan. Appl. Microbiol. 8,61-75. Brock, T. D. (1966b).“Principles of Microbiol Ecology.” Prentice-Hall, Englewood Cliffs, New Jersey. Bungay, H. R., I11 (1968).Chem. Eng. Prog. Symp. Ser. in press. Bungay, H. R.,111, and Cline, D. M. (1967).Abstr. Papers, Am. Chem. Soc., 154th Meeting. Bungay, H. R., 111, and Krieg, N. R. (1966).Chem.E n g . Progr. Symp. Ser. 62,68-72. Cassell, E. A., Sulzer, F. T., and Lamb, J. C., 111 (1966).J . Water Pollution Control Federatton 38,1398-1409. Chen, S.J. (1967).M. S.Thesis, Virginia Polytechnic Inst., Blacksburg, Virginia. Chian, S. K., and Mateles, R. I. (1966).Abstr. Papers, Am. Chem. Soc., 152nd Meeting. Contois, D. E., and Yango, L. D. (1964).Abstr. Papers, Am. Chem. Soc., 148thMeeting. Cooke, W. B., and Ludzak, F. J. (1958).Sewage Ind. Wastes 30,1490-1495. Davis, L. V. (1966).Nature 212,1215-1217. Dean, A. C. R., and Rogers, P. L. (1967).Biochim. Biophys. Acta 148,267-279. de Bary, A. (1879).Reo. Intern. Sci. 3,301-309. Dicks, J. W.,and Tempest, D. W.(1966)J.Cen. Mfcrobtol.45,547-556. Drake, J. F., Jost, J. L., Tsuchiya, H. M., and Frederickson, A. G. (1966).Abstr. Papers, Am. Chem. Soc., 152nd Meeting.


289

Edwards, H. R. (1966).M. S. Thesis, Virginia Polytechnic Inst., Blacksburg, Virginia. Frederickson, A. G., and Tsuchiya, H. M. (1967).Abstr. Papers, Am. Chem. Soc., 154th Meeting. Fuld, G. J., Mateles, R. I., and Kusmierck, B. W. (1961). SOC.Chem. Ind. (London) Monograph 12,54-67. Gaudy, A. F., Jr. (1962).Appl. Microbiol. 10,264-271. Gaudy, A. F., Jr., Ramanathan, M., and Rao, B.S. (1967).Biotechnol. Bioeng. 9,387-411. Cause, G. F. (1934). “The Struggle for Existence.” Williams & Wilkins, Baltimore, Maryland. (Reprinted: Hafner, New York, 1964.) Gilley, J. W. (1965).Ph.D. Thesis, Virginia Polytechnic Inst., Blacksburg, Virginia. Gilley, J. W., and Bungay, H. R., 111 (1967).Biotechnol. Bioeng. 9,617-622. Gilley, J. W., and Bungay, H. R., 111(1968).Biotechnol. Bioeng. 10,99-101. Hawkes, H. A. (1963). “The Ecology of Waste Water Treatment.” Macmillan, New York. Herbert, D. (1961).SOC.Chem. Znd. (London) Monograph 12,21-53. Herbert, D. (1964). Continuous Cultioation Microorganisms, Proc. 2nd Symp., Prague, 1962 pp. 121-131. Czech. Acad. Sci., Prague. Herbert, D., Elsworth, R., and Telling, R. C. (1956).J.Gen. Microbiol. 14,601-622. Hoffman, H. (1966).Adoan. Appl. Microbiol. 8,195-251. Hungate, R. E. (1960).Bacteriol. Rev. 24,353-364. Hungate, R. E. (1962). In “The Bacteria” (I. C. Gunsalus and R. Y. Stanier, eds.), Vol. 4, pp. 95-119. Academic Press, New York. Jannasch, H. W. (1967).Arch. Mtkr0bfol.59~165-173. Koelling, H. R. (1966).M. S. Thesis, Virginia Polytechnic Inst., Blacksburg, Virginia. Larsen, D. H., and Dimmick, R. L. (1964).J. Bacteriol. 88,1380-1387. Leal, L. L. (1964).Thesis, Oklahoma State Univ., Norman, Oklahoma. Lewis, P. M. (1967).J. Appl. Bacteriol. 30,406-409. Macura, J. (1964). Continuous Cultioation Microorganisms, Proc. 2nd Symp., Prague, 1962 pp. 121-131. Czech. Acad. Sci., Prague. Magasanik, B. (1961).Cold Spring Harbor Symp. Quant. Btol. 26,249-256. Maier, W. J., Behn, V. C., and Gates, C. D. (1967).J . Sanit. Eng. Dio., Am. SOC.Cioil Engrs. 93,91-112. Malek, I., and Beran, K. (1962).Folia Microbiol. (Prague) 7,388-411. Malek, I.. and Fencl, Z. (1961).Folia Microbiol. (Prague) 6,192-209. Malek,I., and Ricica, J. (1964).Foltu Microbiol. (Prague) 9,321-344. Malek, I., and Ricica, J. (1965).Folia Microbiol. (Prague) 10,302-323. Malek, I., and Ricica, J. (1966).Folia Microbiol. (Prague) 11,479-535. Mateles, R. I. (1967).Abstr. Papers,Am. Chem. Soc., 154th Meeting. Mateles, R. I., Ryu, D. Y., and Yasuda, T. (1965).Nature 308,263-265. Monod, J. (1942).“La Croissance des Cultures Bacteriennes.” Hermann, Paris. Moser, H. (1958).Carnegie Inst. Wash. Pub1.614. Munson, R.J., and Bridges, B. A. (1964).J. Gen. Microbiol.37,411-418. Nurmikko, V. (1954).Ann.Acad. Sci. Fennicae Ser. A 1 Z 54,7-58. Nurmikko, V. (1955).Acta Chem. Scand. 9,1317-1322. Olson, R.C. (1966).M. S . Thesis, Virginia Polytechnic Inst., Blacksburg, Virginia. Parker, R. B. (1966).Biotechnol. Btoeng. 8,473-488. Parker, R. B., and Snyder, M. L. (1961).Proc. SOC.Erptl. Bid. Med. 108,749-752. Pipes, W. 0.(1966).Adoan. Appl. Microbiol. 8,77-103. Powell, E. 0. (1958).J. Gen. Microbiol. 18,259-268.

290


Powell, E. 0. (1965).Lab. Pract. 14,1145-1149. Prakasam, T. B. S., and Dondero, N. C. (1967).Appl.Microbiol. 15,1128-1137. Proper, G., and Carver, J. C. (1966). Biotechnol. Bioeng. 8,287-296. Ransom, J. P., Finkelstein, R. A., Cedar, R. E., and Formal, S. B. (1961). Proc. Soc. Exptl. Biol. Med. 107,332-336. Renneboog-Squilbin, C. (1967).J.Theoret. Biol. 14,74-101. Saunders, W. M., 111 (1966).Air Water Pollution 10,253-272. Saunders, W. M., 111 (1967).Personal communication. Schaumburg, F. D., and Kirsch, E. J. (1966).Appl.Microbiol. 14,761-766. Schlegel, H. G., and Jannasch, H. W. (1967).Ann. Reu. Microbiol. 21,49-70. Shindala, A. (1964).Ph.D. Thesis, Virginia Polytechnic Inst., Blacksburg, Virginia. Shindala, A,, Bungay, H. R., 111, Krieg, N. R.,and Culbert, K. (1965).]. Bacteriol. 89, 693-696. Slyter, L. L., and Wolin, M. J. (1967).Appl.Microbiol. 15,1160-1164. Slyter, L. L., Bryant, M. P., and Wolin, M. J. (1966).Appl. Microbiol. 14,573-578. Stumm-Zollinger, E. (1966).Appl.Microbiol. 14,654-664. Volterra, V. (1931). “Lecons sur la Theorie Mathematique d e la Lutte pour la Vie.” Gauthier-Villars, Paris. Wood, E. J. F. (1965).“Marine Microbial Ecology.” Reinhold, New York. Yeoh, H. T. (1967). Ph.D. Thesis, Virginia Polytechnic Inst., Blacksburg, Virginia. Zubrzycki, L., and Spaulding, E. H. (1958).j.Bacteriol. 75,278-282.

Chemical Sterilizers (Chemosterilizers)

PAUL M. BORICK Ethicon, lnc., Somerville, New Jersey

I. Introduction .............. .......................................... 291 11. Methods for the Evalu of Chemosterilizers .............. 293 293 A. Sporicidal Tests .................................................... B. Additional Test Methods ........................................ 295 111. Gaseous Chemosterilizers ............................................. 296 A. Ethylene Oxide ....................................... B. Gases and Vapo C. Consideration of Other Antimicrobials as Chemosterilizers ........ IV. Liquid Chemosterilizers ........ A. Strong Acids and Alkalies B. Phenolics ............................................................. 300 C. Halogens .............................................................. 301 D. Formaldehyde ..... E. Glutaraldehyde ............................ ................. 302 F. Other Chemosterilizers .......................................... 306 V. Conditions for the Use of Chemosterilizers ..................... 307 VI. Summary .................................. References

I. Introduction The continuous introduction into world markets of new materials that cannot be heat sterilized necessitates the use or development of other means of sterilization. A major modern method for this purpose is based on the use of chemical agents. Selectivity must be exercised in the employment of chemical compounds, however, as only those which show sporicidal activity can be classified as chemical sterilizers. A wide variety of antimicrobial agents is available, but in most instances these do not kill resistant bacterial spores. Microbicides are specifically limited to the destruction of the type of organism prefixed to cide, e.g., bactericide refers to killing of bacteria, fungicide to fungi, viricide to viruses and sporicide to spores, both bacterial and fungal. Since bacterial spores are the most difficult to destroy, sporicides may be considered synonymous with chemosterilizers. These may be defined as chemical agents which, when utilized properly, can destroy all forms of microbiological life, including bacterial and fungal spores, tubercle bacilli, and viruses. The term chemosterilizer should be distinguished from the term 291

292

PAUL M. BORICK

chemosterilant as reported in “Chemical Processing,” (Anonymous, 1964). Chemosterilizers are chemical compounds which are used to destroy all forms of microbiological life whereas chemosterilants are chemical substances used to sterilize insects and render them incapable of reproduction on mating with nonsterile partners. The term chemosterilant is unfortunately sometimes employed in chemical sterilization literature; it is suggested that such misuse of the term be avoided, and the term chemosterilizer be substituted to prevent further confusion. Although, by definition, disinfectants and antiseptics should be capable of destruction of pathogenic microorganisms, they do not destroy the spores of pathogenic clostridial or aerobic bacilli. Chemicals intended for this use then cannot be classified as chemosterilizers, and it should be realized that they find only limited application. The chemical sterilizer, in addition to destroying fungi and bacterial spores, must be effective in destroying all types of microorganisms, including Mycobacterium tuberculosis within the recommended time. It should also be effective in destroying viruses, thus preventing the occurrence of viral diseases such as hepatitis when used to sterilize contaminated instruments, needles, or other medical equipment. Since only man is known, at present, to be susceptible to hepatitis viruses A and B and no laboratory methods are available for the evaluation of chemical sterilizers, it must be assumed that if other resistant viruses are destroyed, the hepatitis virus will also be killed. In this country, commercial use of chemical sterilizers is governed by the Federal Insecticide and Rodenticide Act and its control falls within the jurisdiction of the U.S. Department ofAgriculture (USDA). Since test methods are not available for hepatitis virus, the USDA is most interested in this problem to ensure that chemical sterilizer claims are justified. In the evaluation of a chemosterilizer, one must take into account the intended use and the conditions associated with its use. These applications may include wide areas, e.g., medical, hospital, and pharmaceutical use, food technology or sterilization of interplanetary space objects. If proven processes such as heat sterilization or irradiation cannot be employed, the proper chemical sterilizer should be selected. One must consider, then, the circumstances under which the chemosterilizer is to be utilized. Time, concentration, numbers and types of microorganisms as well as the type of material to be sterilized are important factors. If ethylene oxide gas were to be used, the

CHEMICAL STERILIZERS (CHEMOSTERILIZERS)

293

limitations associated with this sporicide must be borne in mind. For example, many hours may be required with proper conditions of humidity, temperature, pressure, and gas concentration at the site of the microorganism. If the material to be sterilized were wrapped in a gas-impermeable membrane, the sporicidal agent, no matter how effective, could not reach the site to be sterilized. Although chlorine is recognized by microbiologists as being a highly active antimicrobial agent with sporicidal properties, caution must be exercised in the presence of organic matter since the chlorine concentration, with a concomitant loss in activity, rapidly diminishes upon interaction with any form of organic material. Hence, sufficient active chlorine may not actually be available for adequate sterilization. The manufacturer most probably has had the greatest experience with his product; it is wise then to follow his recommendations and accept the limitations of that particular agent. The practical applications of the use of chemical sterilizers should also be considered. Harsh or difficult conditions may make it impossible to achieve chemical sterilization under any circumstances. It is necessary then to consider chemical, physical, and biological factors in effecting sterilization and achieving sterility. It is the intent of this paper to discuss some of these objectives and present available information on the use of chemicals which can be regarded as effective chemosterilizers. It is also the purpose of this publication to clarify those issues which are borderline and tend to be misleading. The popular use of alcohols, for example, as antimicrobial agents gives one the impression that they can be used effectively as chemosterilizers and are sporicidal in nature. This is not so; alcoholic suspensions of viable bacterial spores have been retained for long periods of time and grown readily after transfer to the appropriate growth media. II. Methods for Evaluation of Chemosterilizers

A. SPORICIDAL TESTS Although the broth dilution or serial dilution method and the agarcup plate technique are widely employed for the rapid screening and evaluation of antimicrobial agents, their use in the appraisal of sporicidal agents is not recommended. Time is not a factor in these techniques and bacterial or fungal spores can revert to the vegetative state. Thus, it would be doubtful whether the chemical compounds tested were effective against the spore or the vegetative cell.

294

PAUL M. BORICK

Tests for determining the presence or absence of sporicidal activity and potential effectiveness against various spore-forming bacteria under different conditions are recommended by the AOAC (Official Methods of Analysis, 1960). Either suture loops or porcelain pennycylinders are infected with spores of bacilli or clostridia under the recommended conditions and exposed to the test agent for prescribed periods of time. The spore inoculum is removed from the test reagent, added to tubes of growth culture media, and incubated. The lack of positive cultures is considered as satisfactory evidence of the desired response, and it is expected that the use of these materials may be extrapolated to products of a similar nature under practical “in use” situations. Resistance of bacterial spores with this method is determined by exposure to 2.5 N HC1. The test spores are expected to resist the HCl for at least 2 minutes, although many will resist the HCl for 30 minutes and longer. Vegetative cells should not show any significant resistance against this concentration of strong acid. Ortenzio et al. (1953) used constant boiling hydrochloric acid for short periods to destroy bacterial spores. The procedure was developed to minimize the hazard in handling dried spores of BaciZZus anthracis or Clostridium tetani and to provide the bacteriologist with a constant source of dried spores for sporicidal evaluations. These workers showed the high degree of resistance of bacterial spores, and confirmed the work of Curran (1952) which indicated that the capacity of bacterial spores to withstand destructive agents is not equaled by any other living thing. The test method for determination of sporicidal activity with suture loops is difficult to evaluate since Spaulding (1964) has pointed out that the suture loop is extremely resistant to sterilization by germicides. When the severity of the test condition was increased by using a heavy inoculum, some bacteria still showed long survival times. Why contaminated silk sutures should be so resistant to sterilization is not known, but it should be borne in mind that this type of situation, perhaps, cannot be extrapolated in practical terms to an actual in-use condition. Stuart (1966) has reported on the current status of AOAC sporicidal test and has recommehded changes in the procedure which would make it more precise. Tests with spores showed a high degree of resistance to chemical sterilizers. Rehydration treatments of spores prior to chemosterilizer exposure showed less resistance. Studies in


295

this area were also conducted by Gilbert et al. (1964). It was pointed out that these procedures were applicable in evaluating the sporicidal activity of both liquids and gases, and that there can be no double standard for liquids or gases as sterilizing agents. This test is considered as satisfactory evidence of the desired response when applied to the carrier giving the lowest result; however, for a confidence level of 95%, killing in 59 out of 60 replicates is expected.

B. ADDITIONAL TESTMETHODS Although the AOAC sporicidal method can be used readily for laboratory evaluations of chemosterilizers, Opfel and Miller (1965) have stated that a valid demonstration must be based on results and experiments duplicating the anticipated environment for each cell in the contaminating population and using microorganisms whose resistance to the sterilizing agent resembles that of the species in the contaminating population. Hence, a chemical sterilizer employed for medical instruments must be capable of destroying those microorganisms considered part of the normal flora of a hospital environment and practical “in-use” test evaluations must be performed subsequent to any initial laboratory evaluations. A chemosterilizer, in addition to destroying bacterial spores and vegetative cells, must also be effective in (a) destroying Mycobacterium tuberculosis, which is known to be more resistant than other commonly encountered vegetative pathogenic bacteria, and (b) in destroying viruses, including proper sterilization of instruments and needles, for the prevention of human hepatitis. Moessel (1963) developed a method for determining the overall rapid evaluation of a microbicide by providing a mixed culture of bacteria, spores, and yeasts. The suspension is treated with the antimicrobial; thus, if the manufacturer claims his agent is sporicidal, a viable spore count is required to determine this activity. Portner et al. (1954) reported on the efficacy of membrane filters for testing the microbiological activity of formaldehyde solutions. This method offers the advantage that bacteria entrapped on membrane filters can be washed free of bacteriostatic material prior to subculture. In the adaptation of this test to sporicidal evaluations, clostridia or other bacterial spores may be used with sporistasis determined by inoculating the medium with spores of the test strain. Growth of the bacterium indicated that the test solution was not sporicidal.

296

PAUL M. BORICK

111. Gaseous Chemosterilizers

A. ETHYLENEOXIDE

The literature on sterilization by ethylene oxide is voluminous and a review of the subject would include hundreds of papers. One of the earliest extensive reviews regarding the bactericidal effectiveness and physical properties of ethylene oxide gas was published by Phillips and Kaye (1949). Phillips (1949) and Kaye and Phillips (1949) presented data which showed the effects of temperature, concentration, exposure time, and moisture upon sterilization with ethylene oxide. Methodology and procedures used in handling ethylene oxide for sterilization purposes in the pharmaceutical industry, surgical dressings and appliances, research institutions and hospitals were explored by Perkins (1956). This work and the properties of gaseous ethylene oxide mixtures were further reviewed by Lloyd and Thompson (1958). One of the first detailed reports regarding the inactivation of viruses by ethylene oxide was published by Ginsberg and Wilson (1950). Various investigators have reported on the use of ethylene oxide for Sterilization of a wide variety of objects, e.g., the sterilization of bacteriological media and other fluids by Wilson and Bruno (1950) and viruses suspended in protein solutions by Auerswald and Doleschel (1962). Polley (1952) used ethylene oxide in the preparation of stable noninfective influenza and mumps antigen and Perkins (1960) employed ethylene oxide on a broad spectrum of pharmaceutical products. Ethylene oxide has also been used extensively in the food industry, e.g., for spices; however, there are limitations to its use in this industry since it can react with vitamins and other amino acids. For more recent reviews on the subject, the reader is referred to the work of Bruch (1961), Mayr (1961), Phillips (1961), Stierli et aZ. (1962), and Lloyd (1963). The above literature demonstrates that ethylene oxide is widely used commercially for sterilization. It is not the intent of this article to pursue this matter, but rather to point out some of the restrictions in its applications. Serious limitations to its widespread use in hospitals were reported by Spaulding et aZ. (1958). It was also shown by Spaulding (1964) that the portable type of ethylene oxide sterilizer was less rapidly cidal than the autoclave type. Ernst and Shull(l962) showed that the relationships of reaction temperatures and concentration of gaseous ethylene oxide to the time required for inactivation of air-dried spores were highly complex and that use conditions of


297

ethylene oxide should be selected on the basis of experimental support. It was further reported by Opfell et al. (1959) that the absence of hygroscopic substances appeared to increase the resistance of B. globigii spores to gaseous ethylene oxide sterilization. B. globigii spores were also shown to be more resistant than other bacterial spores by Lloyd et al. (1967). Although Walter and Kundsin (1959) reported on the faulty functioning of a table model ethylene oxide sterilizer, sterilization problems are also frequently encountered by users of large autoclaves. During these test periods, it may be necessary to explore the many parameters involved in ethylene oxide sterilization. These include temperature, pressure, relative humidity, gas concentration, numbers and kinds of microorganisms to be destroyed, nature and permeability of the material to be sterilized, and, finally, the process and equipment used. If the material to be sterilized is a solid, and the gaseous sterilizer cannot penetrate it, some other means must be employed to achieve sterility. Doyle and Ernst (1967) showed, for example, that occluded spores could not be inactivated with ethylene oxide when a crystalline matrix of calcium carbonate acted as a barrier. It was pointed out by these investigators that spores entrapped in insoluble materials are extremely resistant to ethylene oxide, as well as to moist or dry heat. The package used for sterilization with ethylene oxide is also of major importance. Many materials are impermeable to ethylene oxide, and where this is a problem, a vent may be included in the package and aseptic sealing applied after sterilization. If this is not practical, some other means for achieving sterility must be employed. The use of ethylene oxide for the preparation of sterile products was recommended by Guthrey (1967), using a polyester bag as a portable sterile hood. B. GASESAND VAPORS 1 . Proplylene Oxide (C3HeO) According to Bruch (1961), propylene oxide has weak penetrating ability but strong microbicidal properties, and is frequently used as a substitute for ethylene oxide in sterilization. Russell (1965) has shown that spores of B. subtilis and B . stearothermophilus were destroyed after exposure to gaseous propylene oxide, and that B. subtilis was the more resistant of the two. Although Bruch and

298

PAUL M. BORICK

Koesterer (1961) reported that the activity of propylene oxide decreased with increasing relative humidity, Himmelfarb et al. (1962) found that the death rate of Bacillus subtilis spores substantially increased with this increase in relative humidity. Propylene oxide can be used in conjunction with ethylene oxide for chemical sterilization and, like ethylene oxide, requires dilution with inert gases for safety. Carbon dioxide was previously combined with these chemosterilizers; more recently, mixtures with Freon gas have come into popular usage. Flammability limits in air were reported by Bruch (1961) to be 2.1% to 21.5%by volume. Use of the above-mentioned inert gaseous mixtures can help to eliminate the flammability hazard. The acceptable concentration recommended for sterilization is in excess of 800 mg. per liter.

2. P-Propiolactone P-Propiolactone, a heterocyclic ring compound, is a colorless, pungent liquid, vapors of which have been used for sporicidal activity. Hoffman and Warshowsky (1958) showed that P-propiolactone was lethal for cells of B . subtilis var. niger. Sporicidal activity was dependent on the concentration of the vapor and relative humidity above 70%. Bruch (1961) recommended the use of this chemical for sterilization in various areas and reported that tests performed with spore samples on paper strips showed that this method was used as an indicator for sterilization. Allen and Murphy (1960) showed that P-propiolactone has a broad field of application provided that its limitations are recognized and that it is handled like other bactericidal agents with due regard for its toxic properties. Lo Grippo (1961),however, was able to sterilize biological material without toxic or allergic manifestations, employing P-propiolactone in the liquid state. The use of a specially purified P-propiolactone without the usual impurities of acrylic acid, acetic anhydride, and polymers was stressed. Vischer et al. (1963),Fellowes (1965), Barbieto (1966), and Pritchard et al. (1966), among others, have reported on sterilization by P-propiolactone and the advantages and limitations in its use. One of the major factors in limiting its development as a sterilizing agent is attributed to its carcinogenic and other physiologically hndesirable properties as reported by Wisely and Falk (1960). However, the hydrolysis product- P-hydroxy propionic acid-does not appear to have these disadvantages as it is lower in toxicity and is noncarcinogenic.


299

3. Formaldehyde Vapor Formaldehyde, in the vapor phase, has long been recognized for its chemical sterilizing ability. Formaldehyde vapor cabinets have been used to sterilize lensed instruments which cannot be subjected to heat. Although they have been employed effectively, a high degree of humidity and a prolonged exposure time is required if the formaldehyde is to exert its effect. Nordgren (1939) has summarized much of the earlier work together with his own experimental data on the action of formaldehyde vapor. Although formaldehyde vapors have a high degree of microbiological activity, their ability to penetrate is weak, with the result that their application should be limited to surface sterilization. This method of sterilization appears to be losing its popularity and is being rapidly replaced by ethylene oxide or other gaseous sterilizers. The use of formaldehyde solutions as chemosterilizers will be dealt with separately in this article and the use of formaldehyde vapor in combination with other sporicides is discussed in a later section.

C. CONSIDERATION OF OTHERANTIMICROBIALS AS CHEMOSTERILIZERS A number of other chemical compounds have been reported as chemosterilizers, but the reader is cautioned in their use. As was pointed out earlier, in order to be considered a chemosterilizer, the anti-microbial agent must be sporicidal. Unfortunately, in most instances the microbiological population is unknown. Since spores may be present, it becomes necessary to destroy them as well. Thus, an agent may very well induce sterile conditions where vegetative bacterial cells are present, but not so in the presence of spores. Methyl bromide and ozone, for example, have weak microbicidal properties and, therefore, find limited application. In comparing methyl bromide vapors with ethylene oxide, Opfel et al. (1967) showed significant differences in the susceptibility of microbial populations to these agents, but the mixed vapors sterilized both spores and desiccated cells of S . epidermidis. Ethylene imine was reported by Phillips (1949) to have a higher degree of microbiological activity than ethylene oxide or its methyl, chloromethyl or bromomethyl derivatives. This compound has not gained the wide acceptance of ethylene oxide nor has it been completely explored despite the fact that the work was reported almost

300

PAUL M. BORICK

two decades ago. A possible reason for the lack of popularity may be the occurrence of corrosion with the use of this chemosterilizer. IV. liquid C hemosterilizers

A. STRONG ACIDS AND ALKALIES Because of their sporicidal nature, strong acids or alkalies may be used as chemosterilizers. However, the acidic or caustic nature of these solutions places a strong limitation on their use. Solutions of 2.5 N hydrochloric acid are employed in AOAC sporicidal tests to determine the resistance of spores to this solution. Bacterial !pores are rapidly destroyed within minutes and do not normally withstand an exposure limit beyond 30 minutes, although our own experiments with pure cultures of spores at concentrations of 1 X lo5to 1 X lo6per milliliter showed survival times in excess of 1 hour. This high degree of resistance is indicative of the complex nature of this biological material and its ability to survive under extremely difficult circumstances. Sodium hydroxide also may require long exposure times to destroy bacterial spores. Although spores of Clostridium sporogenes were destroyed within 3 hours when exposed to 2.5 N NaOH, spores of BacilZus megaterium survived a 3-hour exposure to this concentration of alkali. It would appear that while strong caustic solutions may be used for the destruction of spores, long exposure times may be necessary to achieve sterility.

B. PHENOLICS Phenolics have been widely used for disinfection by microbiologists; if they are to be effective as sporicidal agents, however, they must be employed in conjunction with heat. Klarmann (1956)reported. that destruction of resistant bacterial spores could be effected by boiling with dilute aqueous solutions of synthetic phenolics of low volatility. This procedure was recommended for the sterilization of surgical instruments and it was shown that spores of Cl. sporogenes were destroyed by this method without corroding the instruments. While this procedure may be applicable in certain areas, it obviously cannot be utilized for sterilization of delicate-lensed instruments, e.g., bronchoscopes or cystoscopes. The use of 2%phenol and cresol by this method was contraindicated because a pungent odor intensified by heating was found to be objectionable. In accordance with the above, it appears that heating of cultures of sporulating organisms


30 1

to kill vegetative cells will weaken the spores so that they will be more easily killed. For a review of the microbiological activity of the phenolics, the reader is referred to the works of Klannann (1957) and Sykes (1965). C. HALOGENS Chlorine has long been employed as a chemosterilizer. Weber and Levine (1944)were able to show that hypochlorous acid was sporicidal over a pH range of 6 to 9. In a study of the effect ofN-chloro compounds on Bacillus metiens spores, Marks et al. (1945)also found that, at each pH value, the log of time required for kill plotted against the log concentration gave a straight line, and the rate of sterilization decreased rapidly as the pH increased. This author’s experience with dichloro-S-triazinetione, an N-chloro compound, has also shown it to possess sporicidal activity. Where a concentration in excess of 400-500 p.p.m. available chlorine was used at neutral pH, bacterial spore populations were destroyed in a few hours. Engelhard et al. (1961) studied the effects of an organic hypochlorous acid in a phosphate buffer and found that it was able to destroy various microorganisms including Bacillus stearothermophilus and Clostridium sporogenes. Chandler et al. (1957) had reported earlier, however, that hypochlorous acid compounds were not sporicidal when tested by two different methods. The activity of chlorine was markedly reduced in the presence of protein matter and storage temperatures in excess of 70°F. Discrepancies regarding the sporicidal action of iodine, as with chlorine, are reported. The use of iodine as a chemical sterilizing agent was recommended by Gershenfeld and Witlin (1952). These authors found that 2% iodine solutions destroyed spores of Bacillus subtilis, Bacillus anthracis, Bacillus mesentericus and Clostridium tetani when tested on both absorbent and nonabsorbent surfaces. The relatively greater resistance of B . subtilis spores to an iodophor as compared with NaOCl was unexpected by Cousins and Allen (1967) because of the generally accepted view that iodine has greater activity against vegetative bacterial cells. Iodine is frequently used for the destruction of microorganisms on the skin. Although it is effective in destroying a major percentage of microbes, the nature of skin is such that complete destruction of microbes cannot be achieved, hence, these solutions cannot be considered as chemosterilizers. The use of chemical sterilization, for the present at least, is limited to inanimate objects.

302

PAUL M. BORICK

D. FORMALDEHYDE Since alcohols are not sporicidal, they cannot be considered as chemosterilizers. However, they are frequently used as carriers for other chemicals such as formaldehyde. Spaulding (1963) recommended the use of 8% formaldehyde or 20% formalin for the destruction of bacterial spores. His findings showed that formaldehyde was superior to iodophors insofar as sporicidal activity was concerned. The higher concentration of formaldehyde was essential since 1%or 2% formaldehyde solutions are not good sporicides. A solution of 8% formaldehyde in 70% isopropyl alcohol was sporicidal, and sterility was achieved in 3 hours according to Spaulding (1963). Willard and Alexander (1964) concluded from their studies of the sterilizing properties of formaldehyde in methanol and water solutions that formaldehyde in methanol effected sterilization in 24 hours; formaldehyde in water sterilized in shorter periods of time and has a longer shelf life than formaldehyde in methanol. The test inoculum used in these studies was a water suspension of spores of BaciZZus subtilis var. niger. According to these authors, the most recent suggested application of this chemical is for the sterilization of space vehicles. Sykes (1965) reported that a formaldehyde-alcohol mixture is a more effective sterilizing agent than an aqueous solution of formaldehyde and that a 5 % solution of formalin in alcohol will effectively sterilize all spores in 24 hours at 25°C. A 1% solution was also reported to kill in 24 hours at the higher temperature of 37°C. The kind of alcohol used as solvent was not specified. Although formaldehyde has been widely employed for many years for the destruction of microorganisms, and my own experience has shown it to be a good chemosterilizer, one of the chief objections to its use is the pungent odor of the vapors given off by the sterilizing solutions. E. GLUTARALDEHYDE Glutaraldehyde, a saturated dialdehyde having the formula CHO-CH2CH2CH2-CH0, has been known for many years, but it is only recently that it has been developed commercially as a chemosterilizer. Borick (1965) reported that acid solutions stored at room temperature in a closed system remain stable whereas alkalinized glutaraldehyde solutions show a drop in pH and a decrease in glutaraldehyde concentration after 2 weeks (see Fig. 1). The alkaline solutions show a higher degree of sporicidal activity, however, and

303


for this reason are preferred for chemical sterilization (see Table I). A comparison of 1% and 2% glutaraldehyde solutions with 4% formaldehyde solutions by Rubbo and Gardner (1965) showed that

1001

7

21

147

28 0

Time in days

FIG. 1. Glutaraldehyde concentration and pH of alkalinized glutaraldehyde after standing 4 weeks. Reprinted from Borick (1965) courtesy John Wiley & Sons (Interscience).

TABLE I ABILITY OF Bacillus subtilis SPORES TO SURVIVE 2% AQUEOUS GLUTARALDEHYDE SOLUTIONS AT DIFFERENT PH

Exposure time (minutes)

0 30 60 90 180

PH 5

PH 7

PH 9

Microbial Percent count survivors



1.1 x 106 4.0 X lo5 7.3 x 103 0

1.7 X lo8 3.7 x loJ 7.2 x 104

2.1 x 1.7 X 1.0 x 5.3 x 0

106

lo6 lo6 104

100 83 48 3 0

0

100

36 1 0 0

0 0

100 60 4 0 0

glutaraldehyde was able to destroy spores of Bacillus anthmcis more rapidly than was formaldehyde. A comparison of the sporicidal activities of alcoholic glutaraldehyde and formaldehyde solutions

304

PAUL M. BORlCK

(as seen in Table 11) showed that a lower concentration of glutaraldehyde killed aerobic and anaerobic spores in a shorter period of time than a higher concentration of formaldehyde. The microbiological activity of saturated alcoholic dialdehyde solutions was reported by Pepper and Chandler (1963). TABLE I1 SWIUCIDAL ACTIVITIES OF 1% ALKALINE ALCOHOLIC GLUTARALDEHYDE AND 8% FORMALDEHYDE SOLUTIONS USING PENICYLlNDERS CONTAMINATED WITH A MIXED SUSPENSION OF VARYING NUMBERSOF SPORES' Exposure timeb Bacterial counP Test solutions

1 hour

0 hour

3 hours

6 hours

Aerobes Anaerobes Eug.d Thio.' EugadThio.e Eug.d Thio.' Eug." ThioSe

1XGlutaraldehyde l%Glutaraldehyde 1% Glutaraldehyde

380,000 210,000 2,100 3,000 lo 100

8WFormaldehyde 8%Formaldehyde 8%Formaldehyde

380,000 2,100 100

210,000 3,000

l

+ +

+

+

+

o+

+

-

+ +

+ +

+ +

o

o

+

'Reprinted from Pepper and Chandler (1963)courtesy bKey:+ indicates growth and - indicates no growth.

+

+ + + +

+

-

-

+ -

+ +

+

+

+

+

-

-

+

+

+

-

+

+

-

-

+

+

-

of American Society of Microbiology.

CSporecounts were estimated from heat-shocked samples and shown as numbers of spores for cylinder. dEugon broth. CThioglycollatebroth. dKey: indicates growth and - indicates no growth.

+

A wide spectrum of microorganisms was destroyed by aqueous glutaraldehyde solutions, according to various investigators: Borick (1964a,b, 1965), Klein and DeForest (1963), Spaulding (1963), Snyder and Cheatle (1965), and Stonehill et a2. (1963).Since various test methods were employed by these investigators, details will not be reiterated here. The types of microorganisms destroyed can be seen from a perusal of Table 111. A wide range of microorganisms, including vegetative bacterial and fungal cells, viruses, tubercle bacilli, and bacterial and fungal spores, were destroyed by 2% aqueous glutaraldehyde. The ability of glutaraldehyde solutions to maintain their activity in the presence of organic matter was further reported by Borick(1964a,b).

305


Serum was neither precipitated nor coagulated in the presence of glutaraldehyde. When,2% glutaraldehyde was tested in the presence of 20% blood serum, its activity remained the same as that in the absence of organic matter. TABLE I11 EFFECTS OF AQUEOUS ALKALINE 2% GLUTARALDEHYDE SOLUTIONS ON VARIOUSMICROORGANISMS"

Microorganisms

Source

Staphylococcus aureus Streptococcus pyogenes Diplococcus pneumoniae Escherichia coli Pseudomonas aeruginosa Serratia marcescens Proteus vulgaris Klebsiella pneumoniae Trichophyton interdigitale Micrococcus lysodeikticus Mycobacterium tuherr:itlusfs H37Rv Puliurnyrlttts t y p c g I & 11 Echo type 6 Coxsackie B-1 Herpes simplex Vaccinia Influenza A-2, Asian Adeno type 2 Mouse hepatitis (MHV3) Bacillus subtilis spores Bacillus megaterium spores B . globigii spores C1. tetani spores CZ. pe$ringens spores

ATCC 6538 ATCC 12384 U. of Michigan ATCC 6880 ATCC 10145 Rutgers U. ATCC 6380 ATCC 132 ATCC 640 Ohio State U. ATCC 76W Templc: U. Temple U. Temple U. Temple U. Temple U. Temple U. Temple U. Temple U. USDA U. of Texas USDA USDA USDA

Killing time

< < < < < < < < <

Advances in Applied Microbiology, Volume 10

Advances in Applied Microbiology, Volume 10

Advances in applied microbiology

Advances in Applied Microbiology, Volume 26

Advances in Applied Microbiology, Volume 42

Advances in Applied Microbiology, Volume 37

Advances in Applied Microbiology, Volume 49

Advances in Applied Microbiology, Volume 73

Advances in Applied Microbiology, Volume 14

Advances in Applied Microbiology, Volume 60

Advances in Applied Microbiology, Volume 45

Advances in Applied Microbiology, Volume 17

Advances in Applied Microbiology, Volume 9

Advances in Applied Microbiology, Volume 18

Advances in Applied Microbiology, Volume 60

Advances in Applied Microbiology, Volume 39

Advances in Applied Microbiology, Volume 11

Advances in Applied Microbiology, Volume 49

Advances in Applied Microbiology, Volume 68

Advances in Applied Microbiology, Volume 56

Advances in Applied Microbiology, Volume 6

Advances in Applied Microbiology, Volume 36

Advances in Applied Microbiology, Volume 72

Advances in Applied Microbiology, Volume 12

Advances in Applied Microbiology, Volume 63

Advances in Applied Microbiology, Volume 47

Advances in Applied Microbiology, Volume 75

Advances in Applied Microbiology, Volume 8

Advances in Applied Microbiology, Volume 65

Advances in Applied Microbiology, Volume 48

Advances in Applied Microbiology, Volume 43

Advances in Applied Microbiology, Volume 10