MICROBIAL ANTAGONISMS AND ANTIBIOTIC SUBSTANCES LONDON GEOFFREY CUMBERLEGE OXFORD UNIVERSITY PRESS Penicillin crystals Microbial Antagonisms AND Antibiotic Substances SELMAN A. WAKSMAN PROFESSOR OP MICROBIOLOGY, RUTGERS UNIVERSITY; MICROBIOLOGIST, NEW JERSEY AGRICULTURAL EXPERIMENT STATION ^La vie emfeche la vie" — Pasteur NEW YORK THE COMMONWEALTH FUND 1947 COPYRIGHT, 1945, BY THE COMMONWEALTH FUND FIRST PRINTING MARCH 1 945 SECOND PRINTING DECEMBER 1 945 SECOND EDITION, REVISED AND ENLARGED COPYRIGHT, 1947, BY THE COMMONWEALTH FUND PUBLISHED BY THE COMMONWEALTH FUND 41 EAST 57TH STREET, NEW YORK 22, N.Y. PRINTED IN THE UNITED STATES OF AMERICA BY E. L, HILDRETH & COMPANY, INC. This hook is ajfectionately dedicated to BOBILI who has stimulated me in moments of defressiony who has been at all times an inspiration in the search for the unknown, my constant associate and antagonist PREFACE TO THE FIRST EDITION On the basis of their relation to man, the microscopic forms of life may be classified in two major groups: pathogenic forms that attack living systems, especially those useful to man and to his domesticated plants and animals } and saprophytic forms that attack inanimate matter, in- cluding the universal scavengers and the organisms utilized in industry and in the preparation of foodstuffs. Between true parasitism — one or- ganism living in or upon the body of another — and true saprophytism — one organism merely destroying the waste products and the dead cells of another — are groups of relationships that may be designated as antagonistic and associative. In the first of these, one organism is in- jured or even destroyed by the other, whereas in the second, one or- ganism assists the other and may in turn be benefited by it. The antagonistic interrelationships among microorganisms have at- tracted attention since the early days of bacteriology. Following the discovery by Pasteur that microbes are responsible for certain human, animal, and plant diseases, it was established that other organisms, later designated as antagonists, are able to combat and even destroy the dis- ease-producing agents. At first the soil was believed to be the natural habitat of the bacteria that cause epidemics and disease as a whole, but after careful study the fact was definitely established that very few of these bacteria survive for long in the soil. On the contrary, the soil was found to be the natural medium for the development of antagonists chiefly responsible for the destruction of pathogens. The saprophytic organisms that influence in various ways the disease-producing bacteria and fungi were found to inhabit, in addition to the soil, various other natural substrates, such as manure heaps and water basins. The activities and potentialities of these antagonistic microbes still present many problems. Little is known about the nature and mode of formation of the antibiotic substances they produce, and even less about the mode of their action. The substances vary greatly in their physical and chemical properties. Some are soluble in water, others in ether, alco- hol, or other solvents. Some are thermolabile, others are thermostable. Some are sensitive to alkalies or to acids, others are not. Some are viii PREFACE TO THE FIRST EDITION readily oxidized and destroyed, others are not. Some are subject to de- struction by specific enzymes. The substances are largely bacteriostatic in action, to a lesser extent bactericidal. They are selective in their ac- tion upon bacteria, some affecting largely gram-positive organisms and others acting alike upon certain gram-positive and certain gram-nega- tive forms. Some are also fungistatic and fungicidal. Differences are largely quantitative rather than qualitative. Some of the substances are highly toxic to animals. Others are either nontoxic or of limited toxicity and are active in vivo. Some hemolyze red blood cells, others do not. Those that are hemolytic and moderately toxic may be useful for application to local infections. Those that are neither hemolytic nor toxic and are active in vivo may have great im- portance in combating certain diseases in animals and man. Some substances are formed by only a few specific organisms, others may be formed under proper conditions of nutrition by many different organisms. Some antagonists produce only one type of antibiotic sub- stance, others form two or even more chemically and biologically dif- ferent substances. The ability of an antagonist or its products — antibiotic substances — to destroy a parasitic microorganism in vivo is influenced by the nature of the host as well as by the type and degree of the infection. The manner in which antagonists destroy or modify parasites varies greatly, depend- ing frequently upon the nature of the antibiotic substances produced. It is thus clear that the subject is extremely complicated, involving numerous interrelationships among different biological systems of both higher and lower forms of life. In the following pages an attempt is made to present the broad inter- relationships among microorganisms living in association, either in sim- ple mixed cultures or in complex natural populations, with special at- tention to the antagonistic effects. Emphasis is laid upon the significance of these associations in natural processes and upon their relation to dis- ease production in man and in his domesticated plants and animals. The chemical nature of the active — antibiotic — substances produced by vari- ous antagonists is described and the nature of the antagonistic action as well as its utilization for practical purposes of disease control is dis- cussed. However, because concepts of the significance of these phenom- PREFACE TO THE FIRST EDITION ix ena are changing so rapidly, no pretense has been made of examining completely the practical applications of this important subject. Due to the fact that more detailed studies have been made on the production, nature, and utilization of penicillin, more information is presented about this than about any of the other substances. However, this should not be construed as desire on the author's part to emphasize this substance. The subject of antagonistic effects of microorganisms has been re- viewed in both general treatises (706, 944) and special papers (268, 440, 443, 449, 580, 621, 730, 836, 867, 986) J special attention has been paid to the occurrence of such organisms in the soil (316, 670, 941). Advantage was taken of these reviews in the preparation of the com- prehensive bibliography presented at the end of this monograph. At- tention is directed also to a recent complete review of the literature on the nature and formation of penicillin, the historical development of our knowledge of this agent, method of assaying, and clinical applica- tion (410). The author expresses his sincere appreciation to the members of the staff of the Microbiology Department, New Jersey Agricultural Ex- periment Station ; to members of the Department of Research and De- velopment of Merck & Co. and of the Merck Institute for permission to use reproductions of their work, especially the photograph of strep- tomycin crystals J to members of the staff of E. R. Squibb & Sons for supplying the photograph of the penicillin-sodium crystals used as the frontispiece to this volume j to Mrs. Herminie B. Kitchen for her care- ful editing of the manuscript j and to the many investigators in the field whose work has been freely cited both in the form of text or tabular matter and as illustrative material. S. A. W. November i^, ig44 PREFACE TO THE SECOND EDITION The manuscript of the first edition of this book was completed less than three years ago. Since then the subject of antibiotics has made phe- nomenal progress. A number of new substances have been isolated. Several of those known previously as crude preparations have been purified, and some have been crystallized. Penicillin has risen from a metabolic product of certain fungi, promising but difficult to produce, to one of the most important chemotherapeutic agents now available to the medical world, and its yield has been increased a hundredfold by the selection of new strains and by the development of more suitable media and better conditions of growth. Its chemistry has been com- pletely elucidated, and the existence of a number of different forms varying in chemical nature and biological activity has been established. Streptomycin was a laboratory curiosity late in 19435 now it occupies an important place as a promising chemotherapeutic agent for the treatment of certain diseases resistant to penicillin and the sulfa drugs. This rapid progress of our knowledge of the formation, isolation, and utilization of antibiotics makes it advisable to bring out a revised edition of this book. A great deal of new material has been added, but in order to avoid enlarging the book excessively, it was decided to omit a number of references, mostly earlier articles of purely historical in- terest for which the reader is referred to the first edition, and those dealing with the clinical application of penicillin. Several excellent volumes on penicillin dealing with its use for disease control have re- cently been published. S. A. W. February 75, 1^4^ CONTENTS 1. Soils and Water Basins as Habitats of Microorganisms i 2. Human and Animal Wastes 19 3. Interrelationships among Microorganisms in Mixed Popula- tions 36 4. Isolation and Cultivation of Antagonistic Microorganisms j Methods of Measuring Antibiotic Action 53 5. Bacteria as Antagonists 85 6. Actinomycetes as Antagonists 108 7. Fungi as Antagonists 1 30 8. Microscopic Animal Forms as Antagonists 154 9. Antagonistic Relationships between Microorganisms, Viruses, and Other Nonspecific Pathogenic Forms 1 63 10. Chemical Nature of Antibiotic Substances 170 1 1 . The Nature of Antibiotic Action 2 1 8 12. Utilization of Antibiotic Substances for Disease Control 261 13. Microbiological Control of Soil-borne Plant Diseases 300 14. The Outlook for the Future 314 Classification of Antibiotic Substances 329 Glossary 331 Bibliography 233 Index of Microorganisms 395 General Index ' 403 62i9i CHAPTER I SOILS AND WATER BASINS AS HABITATS \. tt OF MICROORGANISMS Although microorganisms inhabit a variety of substrates, from the dust in the atmosphere, the surface of living plants and plant residues, and numerous foodstuffs to the living systems of plants and animals, their natural habitations are soils and water basins. The soil is by no means an inert mass of organic and inorganic de- bris. On the contrary, it fairly teems with life. The organisms inhabit- ing the soil range from those of ultramicroscopic size to those readily recognizable with the naked eye. Many thousands of species, capable of a great variety of activities, are represented in the soil. The physical nature and chemical composition of the soil, the climate, the plant vege- tation, and the topography influence greatly both the composition of the microbiological population of the soil and its relative abundance. One gram of soil contains hundreds, even thousands, of millions of bac- teria, fungi, actinomycetes, protozoa, and other groups of microorgan- isms. Under certain conditions, especially when the supply of fresh or- ganic matter in the form of plant and animal residues is increased, the number may be much greater. This varied microbiological population renders the soil capable of bringing about a great variety of chemical and biological reactions. Through its diverse activities, the microscopic population inhabiting soils and water basins forms one of the most important links in the chain of life on earth. However, its great influence upon numerous phases of human endeavor has been recognized only within recent years. All plants and all animals, including man himself, are dependent upon these organisms to bring about some of the processes essential to the continuation of life. The growth of annual and perennial plants, the supply of food for man and animals, and the provision of clothing and shelter depend largely upon the activities of these microorganisms, especially the transformations brought about in the state of such ele- ments as carbon, nitrogen, sulfur, and phosphorus. 2 MICROORGANISMS IN SOILS AND WATER BASINS Soils and water basins may be regarded as the primary reservoirs for all living systems inhabiting this planet. Whereas the great majority of microorganisms are saprophytic in nature, living upon inorganic ele- ments and compounds and upon the dead residues of plant and animal life, others have become adapted to a parasitic form of existence and have learned to thrive upon the living tissues of plants and animals. Many of these parasites find their way into the soil and into water basins and may be able to survive there for long periods of time or even in- definitely. Although the following discussion is limited primarily to the micro- biological population of the soil, it also applies, to a greater or lesser extent, to the microorganisms that inhabit manures made up of animal excreta, household wastes, and artificially prepared composts and to those that inhabit water basins, including rivers, lakes, and seas. There are, however, marked differences in the nature of the microbial popu- lation of waters and of soils because of the physical and chemical differ- ences in the composition of these two substrates. Nevertheless, some of the underlying principles apply to all substrates. There are, for exam- ple, marked differences in the nature and abundance of the populations of soil and water and those of milk, sewage, and foodstuffs. Whereas microorganisms multiply in the latter substrates at a very rapid rate, those in the soil and in water basins are more nearly static, since the rate of their multiplication is much slower except under very special condi- tions, such as the addition of fresh, undecomposed plant and animal residues or a change in the environment or in the chemical nature of the substrate. PHYSICAL PROPERTIES OF THE SOIL The soil — the surface layer of the earth's crust — comprises three dis- tinct phases, the gaseous, the liquid, and the solid. The last is largely inorganic in nature, with varying concentrations of organic constituents originating from plant and animal residues and found in the soil in dif- ferent stages of decomposition. The organic substances together with the living and dead cells of microorganisms that inhabit the soil make up what is known as soil organic matter or, more often, soil humus. The soil as a medium for the development of microorganisms is thus mark- PHYSICAL PROPERTIES OF THE SOIL 3 edly different from the common artificial laboratory media, whether these be synthetic or consist of products of animal or plant life, upon which these organisms are grown. The inorganic soil particles are surrounded by films of colloidal ma- terials, which are both inorganic and organic in nature. As a rule, the microorganisms inhabiting the soil adhere to these films, although some move freely in the water surrounding the particles. Water and air play essential roles in the soil system and control the nature and extent of the soil population. The nature and size of the mineral and organic soil fractions, as well as the phenomena of adsorption, also influence the abundance, nature, and distribution of microorganisms in the soil. Sandy soils are better aerated than heavy clay soils j they are, therefore, more favorable for the growth of aerobic bacteria and fungi. However, since such soils lack the high water-holding capacity of the heavier soils, they are more readily subject to the process of drying out, which may result in a reduction in microbial activities. Oxygen, another important factor in microbial development in the soil, becomes available to microorganisms by gaseous diffusion. The oxygen supply diminishes with increase in depth of the soil. When an excess of free water is present in the soil, gaseous oxygen cannot pene- trate very deeply and soil organisms then become dependent upon the dissolved oxygen which diffuses into the soil solution. Since the rate of oxygen diffusion is extremely slow, waterlogged soils tend to become depleted of oxygen. Under these conditions, there are marked changes in the microbiological population of the soil: the fungi and actinomy- cetes tend to decrease, and the bacteria, especially anaerobic types, pre- dominate. Peat bogs are examples of soils in a perpetual anaerobic state j the microbial population is quite distinct from that of mineral soils. Semiarid soils, with a much greater diffusion of oxygen into the deeper soil layers, possess a population which is largely aerobic j in these and other mineral soils the abundance and nature of the organic matter exert a decided influence upon the abundance and nature of the microorgan- isms present. Th^ microbiological populations of soils, composts, and water basins are also influenced markedly by seasonal and temperature changes. Certain microorganisms are capable of active life at temperatures ap- 4 MICROORGANISMS IN SOILS AND WATER BASINS preaching the freezing point of water j others, known as thermophilic forms, can withstand very high temperatures, some being active even at 60° to 70° C. The reaction of the soil is also a factor influencing the nature of the population. Many microorganisms are active within a very limited range of /)H values j others, notably many of the fungi, are adapted to much wider ranges of reaction. In acid soils, larger numbers of fungi are present, because of the fact that they tolerate more readily the more acid reactions, which limit bacterial competition. On the other hand, actinomycetes comprise a large percentage of the microbial popu- lation of dry and alkaline soils. CHEMICAL COMPOSITION OF THE SOIL The solid part of the upper or surface layer (20 to 30 cm.) of the soil commonly is made up of i to 10 per cent organic matter and 90 to 99 per cent inorganic or mineral matter. The concentration of organic mat- ter may be even less than i per cent, as in desert and poor sandy soils, or more than 10 per cent, as in certain virgin prairie soils and, especially, peat lands which consist of 50 to 99 per cent organic matter, on a dry basis. The organic matter of the soil is markedly different in chemical na- ture from that of plant and animal materials. It contains much less cellu- lose and hemicelluloses than the majority of plants and is higher in lignins and proteins. It is characterized by a narrow ratio of the two important elements carbon and nitrogen, usually about 10: i j it is much more resistant to microbial decomposition than are plant and animal residues. It is black, is soluble to a considerable extent in alkalies, and is partly reprecipitated by acids. These alkali-soluble constituents have often been designated as "humic acids" or "humic bodies," thus impart- ing the idea that soil organic matter is made up largely of these "acids" (94^). The inorganic constituents of f he soil comprise largely sand, silt, clay, and, to a more limited extent, a number of soluble and insoluble salts, notably phosphates, sulfates, and silicates of calcium, magnesium, potas- sium, iron, aluminum, manganese, zinc, copper, and others. Some of BIOLOGICAL STATE OF THE SOIL 5 the chemical elements comprise the framework of the soil and are used to only a limited extent by plant and microbial life. Others form im- portant nutrients (for example, C, N, S, P, H, and O) or serve as cata- lysts for the continuation of life (Zn, Fe, Mn, Cu, Mo, B, and even K are often considered as belonging in this category). The function of most of these elements in the life of microorganisms is not fully under- stood. In view of the fact that some of the elements in the latter group have been found to form important constituents of certain enzyme sys- tems, the difference between the two functions is not significant. BIOLOGICAL STATE OF THE SOIL The abundance of higher plant and animal life in and upon the sur- face of the soil influences considerably the nature and extent of the microbiological population. Certain plants harbor in their roots specific microorganisms that act as symbiontsj this is true of the root nodule bacteria of leguminous plants and the mycorrhiza-forming fungi found in orchids, evergreens, and many other plants. Higher plants also offer a favorable environment for the growth of certain other types of bacteria and fungi, this specific environment being designated as the rhizosphere. The bacterial population of the rhizosphere is not very different qualitatively from that found some distance away from the plants, except that certain types of bacteria are more prominently repre- sented. The growth of plants results in the production of waste materials and residues left in and upon the soil in the form of roots, leaves, needles, and other products, all of which offer favorable nutrients for microbial development. The root systems of plants also bring about bet- ter aeration of the soil, thus making conditions more favorable for the development of aerobic organisms. The presence of higher plants often leads to the development of certain types of bacteria, fungi, and nema- todes that are pathogenic to the plants, such as the causative agents of root rots, damping-off diseases, root-galls, and various others. Some of the pathogens may become well established in the soil and may persist there long after the specific host plants have been removed. They may even be able to attack other hosts. Plant life thus exerts a variety of in- 6 MICROORGANISMS IN SOILS AND WATER BASINS fluences upon the nature and abundance of the soil-inhabiting micro- organisms. Higher animals also influence the soil microbiological population. Cattle and horses on pastures contribute, through their droppings, energy sources and various other essential nutrients for the develop- ment of microorganisms. After death, the bodies of animals, from the smallest insects to man, the lord of creation, also offer available nutri- ents for the growth of numerous microorganisms. Many animals living in the soil, such as insects and rodents, become carriers of certain bac- teria and fungi that are destructive to their hosts j this phenomenon is often utilized for combating injurious animals. Finally, the numerous animals living on the surface of the soil leave waste products rich in bac- teria, fungi, and invertebrate animals, some of which are capable of causing serious animal diseases (945). NATURE AND COMPOSITION OF THE SOIL MICROBIOLOGICAL POPULATION The microorganisms inhabiting the soil can be divided, on the basis of their systematic position in the biological kingdom, into the following eight groups: bacteria, actinomycetes, fungi, algae, protozoa, worms, insects and other near-microscopic animals, and ultramicroscopic forms. The last group comprises bodies that range from living systems to products of living organisms j they possess the property of activating similar substances and imparting to them their specific activities, as in the case of phages and viruses. Five methods are commonly employed for determining the abun- dance of the various groups of microorganisms inhabiting the soil 3 namely, plate culture, selective culture, direct microscopic methods, contact slide, and mechanical separation. Each of these has certain ad- vantages and certain limitations. In many cases, special methods have been devised to supplement the more common methods. The plate method is based upon principles similar to those employed in other branches of bacteriology. Various media are used, both organic and synthetic. The soil microbiologist has attempted to produce media that either allow the development of the greatest number and the great- SOIL MICROBIAL POPULATION 7 est variety of organisms or are particularly favorable for the growth of certain special types of organisms. None of the media so far employed allows the growth of the total soil population. The plate method is often supplemented by the selective culture method, in which a great variety of media are used in order to obtain a representative picture of the soil population. Since the number of media required to enable all soil micro- organisms to develop is virtually limitless, the enrichment methods can only give a proximate idea of the nature and abundance of the micro- biological population. Because of the development on the plate of cer- tain organisms that exert a toxic effect upon others, the plate method often shows excessive variation in the numbers of bacteria and fungi (256). The microscopic methods have been introduced to fill this gap, since by them the relative abundance of the various groups of organisms found in soils, composts, or other natural substrates can be established. Unfortunately, these methods do not allow any differentiation between living and dead cells, nor do they permit a differentiation between the various physiological types of microorganisms such as pathogens and nonpathogens. A further limitation, especially of the contact slide, is that the fast-growing forms cannot be prevented from overgrowing the slide and repressing the slow-growing types. The mechanical separation methods are based upon the use of special sieves or water emulsions and are utilized for the study of the larger forms such as insect larvae and nematodes. The relative abundance of the different groups of microorganisms in a given soil, as determined by any one of the foregoing methods, varies with the nature of the soil, amount of organic matter, oxygen sup- ply, moisture content, temperature, acidity, and buffering capacity (Table i), as well as with the nature of the higher plants growing in the given soil (Table 2). Despite all these factors, the microbiological population of the soil throughout the world has certain definite and common characteristics and comprises certain well-defined, specific types. The bacteria usually range in number from a few hundred thou- sand to several hundred million per gram of soil, though many species do not develop on the common plate. Fungi are found in the form of mycelial filaments and as spores and may therefore constitute as large 8 MICROORGANISMS IN SOILS AND WATER BASINS a mass of living matter as do the bacteria ; their actual number, as deter- mined by the plate method, may vary from a few thousand to several hundred thousand per gram of soil. The significance of these results is TABLE I. INFLUENCE OF SOIL TREATMENT ON NUMBER OF MICROORGANISMS REACTION TREATMENT OF SOIL OF SOIL MICROORGANISMS FOUND* fH Bacteria Act :inomycetes Fungi Unfertilized and unlimed 4.6 3,000 1,150 60 Lime only added 6.4 5,410 2,410 23 Potassium salts and phosphates added 5-5 5^360 1,520 38 Salts and ammonium sulfate added 4.1 2,690 370 1X2 Salts, ammonium sulfate, and lime added 5.8 6,990 2,520 39 Salts and sodium nitrate added 5-5 7,600 2,530 47 Stable manure and salts added 5-4 8,800 2,920 73 From Waksman (945). * In thousands per gram of soil as determined by plate method. TABLE 2. INFLUENCE OF GROWING PLANTS ON NUMBER OF MICROORGANISMS IN THE SOIL SAMPLE OF PLANT SOIL TAKEN MICROORGANISMS FOUND* Bacteria Actinomycetes Fungi Rye Near roots 28,600 4,400 216 Away from roots 13,200 3,200 162 Corn Near roots 41,000 13,400 178 Away from roots 24,300 8,800 134 Sugar beet Near roots 57,800 15,000 222 Away from roots 32,100 12,200 176 Alfalfa Near roots 93,800 9,000 268 Away from roots 17,800 3,300 254 From Starkey (877). • In thousands per gram of soil. SOIL MICROBIOLOGICAL POPULATION 9 not always clear, since a given colony may have originated from a hyphal filament, a mass of mycelium, or a single spore. Determination by the plate method of the number of actinomycetes is subject to the same limitations j these organisms usually constitute from lo to 50 per cent of the colonies appearing on common bacterial agar plates. Algae are numerous in the surface layers of soil only. Protozoa are present in the soil in an active vegetative or trophic state and in the form of cysts. The active cells appear when excessive water is present, even for a few hours j in dry soil, the cysts predominate. Flagellates are represented by the largest numbers, sometimes approaching a mil- lion individuals per gram of soil j amebae are next in abundance j cili- ates are usually found to the extent of a few hundred to several thou- sand per gram of soil. Nematodes, rotifers, earthworms, and larvae of numerous insects are also abundant, often forming a large part of the bulk of the living mass of cell substance. By means of the selective and enrichment culture methods, several physiological classifications of bacteria have been recognized. The fol- lowing descriptive terms are commonly used to designate these groups : autotrophic vs. heterotrophic, aerobic vs. anaerobic, motile vs. non- motile, pathogenic vs. saprophytic, psychrophilic and mesophylic vs. thermophilic, symbiotic vs. nonsymbiotic, and antagonistic vs. non- antagonistic. The fungi may be classified into three types: saprophytic and free- living, mycorrhiza-producing, and plant pathogenic. The most com- mon groups of soil fungi are found in the genera Rhizo-pus, Mucor, Penicillium, Aspergillus y Trichoderma, Fusarmm, Cladosforium, and Cefhalosforium. The soil often harbors an abundant population of yeasts and fleshy or mushroom fungi. The latter may produce an ex- tensive mycelium in the soil, binding the particles together and pre- venting their falling apart. Various bacteriolytic agents, including specific phages, have also been demonstrated in the soil. The phage of root-nodule bacteria is of par- ticular interest. It is readily adsorbed by the soil, but its presence can easily be established. The repression of spore-forming bacteria and the abundance of Pseudomonas jiuorescens may be due to the antagonistic action of the latter. 10 MICROORGANISMS IN SOILS AND WATER BASINS SOILS AND WATER BASINS AS CULTURE MEDIA Microorganisms require for their growth and respiration certain energy sources and certain nutrients, as well as certain conditions favor- able for their development. Different organisms show considerable variation in this respect. The mineral elements required for growth and multiplication are almost invariably present in the soil and to a large extent also in many water basins. The available energy supply may be limited, however, and thus usually becomes the most important factor regulating the abundance and activities of microorganisms in natural substrates. The autotrophic bacteria depend on the supply of oxidizable minerals such as ammonium salts, nitrite, sulfur, iron, and manganese, the oxidation of which makes energy available for their growth. The heterotrophic organisms are dependent on the carbon compounds brought into the soil in the form of plant and animal residues as well as the bodies of many insects, earthworms, and other small animals. The roots of plants also supply an abundance of easily available sub- stances for microbial nutrition. Every organic compound produced in nature finds its way, sooner or later, into the soil or into lakes and rivers, where it serves as a source of energy for microorganisms. This energy becomes available to some of the organisms through anaerobic or fermentative transformation and to others through aerobic or oxidative processes. The net change in the energy produced by any one organism or group of organisms is accom- panied by a loss of free energy by the system to which the culture is TABLE 3. MULTIPLICATION OF COLIFORM BACTERIA IN STERILE SOIL BACTERIA ORGANISM INOCULATED* BACTERIA RECOVERED* After 10 days After 26 days Escherichia coli in soil alone 2,600 149,000,000 138,000,000 Aerobacter aero genes in soil alone 109,000 48,000,000 42,600,000 in soil and glucose 109,000 1,660,000 240,000,000 From Waksman and Woodruff (978). * Per gram of soil. NUTRITION OF MICROORGANISMS 11 confined. The synthesis of new cell material by microorganisms is ac- companied by a gain of free energy, which must be supplied by other chemical transformations. Ordinary soils, however, contain microbial nutrients in concentrations sufficient to support a large number of living cells. This can be illustrated by the fact that when a soil is sterilized and then inoculated with a pure culture of bacteria rapid multiplication takes place (Table 3). When fresh water taken from a lake or the sea is kept in the laboratory for one or two days, a great Increase in its bac- terial population occurs. There is considerable variation in the ease with which a specific or- ganism can be isolated from a natural substrate and consequently in the techniques employed. Some microorganisms may be present in abundance and can be readily isolated. Others are found only in limited numbers and can be obtained only with considerable difficulty and by the use of special procedures. Still others can be isolated only after the natural substrate is treated in such a manner as to favor the multiplica- tion of the specific organism ; this can be done by enriching the soil with a nutrient or substance which the particular organism is able to utilize, or by changing conditions of reaction, by aeration, or by other treat- ment. Such treatment sometimes results in the development of special strains or races adapted to the special conditions. NUTRITION OF MICROORGANISMS IN NATURAL SUBSTRATES It was at first assumed that bacteria and other microorganisms possess a simpler type of metabolism than do higher plants and animals j al- though some can obtain all the nutrients required for cell synthesis and energy from simple elements and compounds, others need for their nu- trition certain highly complicated organic substances. Recently it has been recognized that various "growth-promoting" substances or vita- mins play an important role in the nutrition of many microorganisms. It has also been established that highly complicated enzyme systems are produced by these lower forms of life, and that many interrelationships exist among their metabolic processes, the composition of the medium, and the environmental conditions. One thus begins to realize that the L 5 3 3 A 12 MICROORGANISMS IN SOILS AND WATER BASINS metabolism of these microbes is also highly complicated. Most of the in- formation on their nutrition is based upon their growth on artificial cul- ture media. In nature, however, these organisms live in associations and vary considerably in the degree of their interdependence. As yet no laboratory method has been developed that duplicates these conditions. Microorganisms vary considerably in their nutrition and energy utilization, as well as in the breakdown and transformation of the avail- able nutrients. Certain elements or compounds are required for cell synthesis. In some cases, certain trace elements as well as varying con- centrations of growth-promoting substances are also essential. Among the nutrient elements, nitrogen occupies a prominent place. Consider- able variation exists in the ability of microorganisms to utilize different types of nitrogen compounds: some can obtain their nitrogen from a wide variety of substances j others are restricted to the use of a single group of compounds such as proteins, amino acids, urea, ammonia, or nitrate J a few are able to use atmospheric nitrogen. The variety of or- ganic nitrogenous bodies supplied to microorganisms in soils and in water basins is limited only by the number of such compounds synthe- sized by plants and animals. The complex forms of nitrogen are broken down to simpler compounds j these may be assimilated by organisms and again built up into complex forms, or they may be utilized only by other organisms. Microbial activity thus regulates the state of the nitro- gen in natural substrates and is responsible for the continuous stream of ammonia and nitrate forming the available sources of nitrogen that make possible the growth of higher plants. THE GROWTH OF THE MICROBIAL CELL IN PURE CULTURE AND IN MIXED POPULATIONS When nutrients are available in sufficient concentration and when the environmental conditions are favorable for the development of the microbial cell in pure culture, growth follows a definite sigmoid-shaped curve. Slow multiplication is followed by rapid development, until a certain maximum number of cells within a given volume of medium is reached} the rate of growth then diminishes. The maximum population of Aerohacter aero genes grown in a medium containing lactose and DISEASE-PRODUCING ORGANISMS IN THE SOIL 13 ammonium tartrate increases at first in proportion to the concentrations of these nutrients but later becomes independent of them. The onset of the stationary phase may be due to several factors: exhaustion of sub- stances necessary for growth, change in the reaction of the medium to one unfavorable for further development, accumulation of toxic prod- ucts. When the nutrients in the medium are exhausted, addition will restore growth. When an unfavorable change in reaction has taken place, the addition of acid or alkali will render the medium again favor- able. The production of toxic substances in the medium can be counter- acted usually by the use of heat or by treatment with charcoal, though some of the injurious bodies may be heat-resistant. In the presence of other microorganisms, a certain organism may show reactions markedly different from those obtained in pure culture: it may produce substances that are either favorable or injurious to the other cells, it may compete with the other organisms for the available nutrients or it may render the medium more favorable for their de- velopment. Some bacteria like Bacillus cereus can attack native proteins but not amino acids, whereas others like Ps. jiuorescens can attack amino acids but not proteins j when these two organisms are placed together in the same medium, their activities supplement one another. Numerous other instances are found in soil and water of an organism preparing the substrate for another, ranging from distinct symbioticism, where one organism depends absolutely for its living processes upon the ac- tivities of another (symbiosis), to association, where one organism merely is favored by the growth of another (metabiosis), to the injury of one organism by another (antagonism), and finally, to the actual destruction of one by another (parasitism). INTRODUCTION OF DISEASE- PRODUCI NG ORGANISMS INTO THE SOIL Ever since higher forms of life first made their appearance on this planet they have been subject to attack by microbes. These microscopic organisms must have gained, at an early stage in the development of the higher forms, the capacity of attacking them in one manner or an- other. There is no plant or animal now living that is not subject to in- 14 MICROORGANISMS IN SOILS AND WATER BASINS fection by different bacteria, fungi, and protozoa. The more advanced the animal body is in the stage of evolution, the more numerous are its ills, most of which are caused directly or indirectly by microorganisms. The microbial agents causing thousands of diseases of plant and ani- mal life have now been recognized and even isolated and described. In many cases these disease-producing agents are closely related morpho- logically to others that lead a harmless existence in soils or water basins i many of the saprophytes, for instance, are found to be of great benefit to man and to his domesticated plants and animals. This sug- gests the probability that pathogenic microorganisms represent certain strains of soil and water-inhabiting types that have become adjusted to a parasitic existence. During their life in the host, they multiply at a rapid rate and produce substances toxic to the body of the host. The re- sult is that the host is incapacitated for a certain period of time, until it succeeds in building up resistance against the invading organisms. It may thus overcome the injurious effect of the pathogen or it may be killed if such resistance cannot be effected. In the first instance, a tem- porary or permanent immunity against the specific disease-producing microbe or its close relatives may result. The host is often able to sur- vive the attack without being able to destroy the invading microbes j if it again attains a normal form of life, it is designated as a carrier of the disease-producing agent. Pathogenic organisms pass their existence in the living body of the plant or animal. They spread from one host to another by contact or through a neutral medium, such as water, milk, or dust where they may remain alive and active for varying lengths of time, or they reach the soil or water basins in the excreta of the host. If the host is killed by the infecting microbes, they may survive for some time upon the rem- nants of what was once a living animal or plant and thus find their way into the soil and water basins. Considering the millions of years that animals and plants have ex- isted on this planet, one can only surmise the great numbers of microbes causing the numerous diseases of all forms of life that must have found their way into the soil or into streams and rivers. What has become of all these pathogenic bacteria? This question was first raised by medical bacteriologists in the eighties of the last century. The soil was searched SAPROPHYTIC ORGANISMS IN THE SOIL 15 for bacterial agents of infectious diseases. It was soon found that, with very few exceptions, organisms pathogenic to man and animals do not survive very long. This was at first believed to be due to the filtration effect of the soil upon the bacteria. It came to be recognized, how- ever, that certain biological agents are responsible for the destruction of the pathogenic organisms. These investigations led to the conclusion that the soil can hardly be considered as a carrier of most of the infec- tious diseases of man and animals. The fact that many pathogens can grow readily in sterilized soil but do not survive long in normal fresh soil tends to add weight to the theory of the destructive effect upon pathogens of the microbiological population in normal soil. INTRODUCTION OF SAPROPHYTIC ORGANISMS INTO THE SOIL It often becomes necessary to inoculate the soil with organisms not usually found there. The common practice of inoculating soil with bac- teria capable of forming root nodules on leguminous plants is a case in point. It is essential, therefore, to know how long these organisms will survive. The survival period is influenced greatly by the presence of a host plant that protects the specific bacteria from attack by antagonistic organisms. In the absence of the host plant, the bacteria seem to disap- pear gradually, and reinoculation becomes advisable when the host is again planted in the given soil. It has been observed also that specific strains of bacteria tend to deteriorate in the soil, and that it is necessary to reinoculate the soil with more vigorous strains of the organisms in question. Some bacteria, notably members of the Azotobacter group, are able to fix nitrogen independently of host plants but these organisms are absent from many soils. The suggestion was made that such soils might benefit from inoculation. However, it has been found that when soils and peats are inoculated with A. chroococcum large-scale destruction of the latter often occurs (814), due, it is believed, to the presence in th9 soil of antagonistic organisms as well as toxic substances (492, 687, 980). Certain fungi are unable to grow in fresh nonsterilized soil but are 16 MICROORGANISMS IN SOILS AND WATER BASINS capable of growing in heated soil. This was found to be due to the fact that normal soils contain certain substances that render the growth of the fungus impossible j these substances are destroyed by heating. An extract of fresh soil acts injuriously upon the growth of the fungus Pyronema; the injurious effect is partly removed on boiling. Certain forest soils contain not only antifungal but also antibacterial factors (630a) which are dialyzable and thermostable j other thermolabile and nonfilterable substances may be present which neutralize the effect of the antibiotics. The survival of microorganisms added to soil or water is thus influ- enced by the nature of the native soil or water population, the organ- isms added, the composition of the substrate, and various environ- mental conditions. SAPROPHYTIC AND PATHOGENIC NATURE OF CERTAIN SOIL MICROORGANISMS Various fungi and actinomycetes causing animal diseases, notably skin infections, appear to resemble very closely the corresponding soil saprophytes. It was therefore suggested that many of the dermato- phytic fungi normally lead a saprophytic existence in the soil but are also capable of developing on epidermal tissue and bringing about in- fection of the tissues. This was found to be true especially of species of S-porotrkhum, various actinomycetes such as those causing lumpy jaw of cattle, and certain other organisms. Henrici (406) divided fungus in- fections of animals into two groups : first, superficial mycoses, compris- ing moniliases and dermatomycoses, that are caused by a variety of fungi widely distributed in nature 5 and, second, deep-seated infections, namely, aspergillosis, sporotrichosis, and blastomycosis, with a marked tendency to restricted distribution. The latter were said to be caused primarily by saprophytic forms, including varieties capable of chance survival and of multiplication when accidentally introduced into ani- mal tissues. Walker (981) suggested that the partly acid-fast coccoid, diph- theroid, and actinomycoid organisms that have been cultivated repeat- edly from leprosy are merely different stages in the life cycle of the SAPROPHYTIC AND PATHOGENIC MICROORGANISMS 17 same form. The causative agent of leprosy, like certain pathogenic actinomycetes, is believed to be a facultatively parasitic soil organism, probably of wide but irregular distribution. Leprosy was thus looked upon primarily as a soil infection, brought about presumably through wounds i a secondary means of infection by contagion was not excluded. A comparison of cultures obtained from rat leprosy, human leprosy, and bacteria of soil origin led to the conclusion that the strains from all three sources were identical j human and rat leprosy were said to have the same etiology and endemiology, finding a normal habitat in the soil. An interesting relationship has been shown to exist between Texas fever and the capacity of cattle tick {Boofhilus bovis)y the parasite car- rier, to persist in the soil (865). The causative agent is an organism with protozoan characteristics. It persists in southern pastures where the carriers survive from one season to the next and keep the cattle con- tinuously infected. The disease is of little importance in northern re- gions, the ticks being destroyed during the winter. When northern cat- tle are moved to southern pastures, they become subject to the disease. Pathogenic microorganisms capable of surviving in the soil have pre- sented important economic problems to farmers raising hogs, cattle, poultry, and other domestic animals, but disease incidence through this source has been greatly diminished by the proper practice of sanitation. The rotation of crops has been utilized for the purpose of overcoming these conditions, several years usually being required to render infected pastures safe for use. The fact that most pathogenic organisms rapidly disappear when added to the soil makes this problem rather simple j the prevention of infectious diseases would have presented far more diffi- cult problems were the infecting agents to remain indefinitely virulent in the soil. The few disease-producing agents that are capable of per- sisting, such as anthrax, blackleg, and coccidiosis, have been the cause, however, of considerable damage to animals. Of greater economic importance than the survival in the soil of hu- man and animal pathogenic agents is the fact that the soil harbors a number of plant pathogens, including not only fungi, bacteria, and actinomycetes, but also nematodes and insects. Fortunately, the con- tinued development of these organisms in the soil also leads to the ac- cumulation of saprophytic organisms destructive to them. 18 MICROORGANISMS IN SOILS AND WATER BASINS The extent to which virus diseases persist in the soil is still a matter for speculation. It has been demonstrated that the phage of legume bacteria may persist and become responsible for a condition designated as "alfalfa-sick soils" and "clover-sick soils" (178, 49o). In order to overcome this condition, the breeding of resistant varieties of plants has been recommended. CHAPTER 2 HUMAN AND ANIMAL WASTES And a flace shall thou have without the campy ivhither thou shalt go forth abroad: and a sfade shalt thou have with thy weapons; and it shall be, when thou sittest abroad, thou shalt dig therewith, and shalt afterward cover that which cometh from thee. — Deuteronomy 2^:1^ and 14. Human and animal excreta and other waste products, which are or fre- quently become both offensive and dangerous to public health, sooner or later find their way into the soil and water basins. The soil also re- ceives the many residues of growing crops that are annually left on the land, together with the waste materials of the farm and the home (942), These wastes contain substances partly digested by man and ani- mals, and their metabolic waste products, as well as freshly synthesized material in the form of microbial cells. The microbial population of such waste materials comprises agents of digestion, some microbes that are present accidentally, and some that possess the capacity of causing human, animal, and plant diseases. These waste materials do not remain long in an unaltered form and do not accumulate in or on the surface of the soil or in water basins j otherwise both soil and water long ago would have been rendered un- sightly, disagreeable bodies, which man would not dare to tread upon or enter. On the contrary, the soil and the water are capable of di- gesting all these cast-off materials and of completely destroying their undesirable characteristics. Through all past ages, the waste products of plant and animal life have disappeared, whereas the soil and the water in the rivers, lakes, and seas have remained essentially the same, except under very special conditions such as those that brought about the pro- duction of peat in water-saturated basins and, in past geological ages, the formation of coal. The capacity of soil and water to destroy these of- fensive wastes is due entirely to the microorganisms that inhabit the substrates. The important ultimate products of destruction are am- monia, carbon dioxide, and water j often hydrogen and methane are 20 HUMAN AND ANIMAL WASTES produced j various mineral compounds, such as phosphates, sulfates, and potassium salts are also liberated. These mineralized substances are es- sential for the continuation of plant and animal life on this earth. Largely because of the activities of the microorganisms inhabiting soils and water systems, man does not need to worry about the disposal of plant and animal wastes. These activities need only be regulated, in order to accomplish the breakdown of complex substances with the greatest efficiency and the least loss of valuable nutrient elements. The following principal objectives are usually to be attained: first, the de- struction of plant and animal pathogens, including pathogenic bacteria and fungi and disease-producing protozoa, worms, and insects j second, the liberation of the essential elements required for plant nutrition in available forms, especially carbon, nitrogen, and phosphorus j and, third, the formation of certain resistant organic substances, known col- lectively as humus, which are essential for the improvement of the physical, chemical, and biological condition of the soil. STABLE MANURES AND FECAL RESIDUES IVLicrobial Pofulation Fresh excreta of animals and man are rich in fecal bacteria, consisting, on the average, of 5 to 20 per cent bacterial cells. Lissauer (575) calculated that the bacterial substance of feces ranges from 2,5 to 15.7 per cent of the dry weight, with an average of 9 per cent. Bacteria were reported to make up 9 to 42 per cent of the bulk of animal stools, the percentage depending on the composition of the foodstuffs, the nature of the animal and its condition of health, and other factors {^66). Since i mg. of dry bacterial substance contains about 4 billion bacterial cells, the number of these organisms in fecal excreta can be seen to be very large, although many, if not most, of the cells are no longer in a living state. By suitable methods of cultivation, human feces were found (626) to contain 18 billion bacteria per gram. About 100 billion bacteria may be produced daily in the human intestine. Human feces are made up, on an average, of 32.4 per cent bacterial cells amounting to 2,410 million bacteria per milligram of moist material. Feces of healthy STABLE MANURES AND FECAL RESIDUES 21 persons were shown (301 ) to contain 8.2 to 24.2 per cent bacterial cells j in those of persons suffering from intestinal disturbances the percent- age was 20.1 to 40.2. With the development of the microscopic tech- nique for counting bacteria, much larger numbers of cells were shown to be present than could be determined by the plate method. The urine of healthy persons is sterile or very low in bacteria. Be- cause of the ability of many bacteria to utilize the chemical constituents of urine, rapid bacterial multiplication takes place in fresh urine, espe- cially when mixed with animal feces and bedding (811). The microbiological population of animal excreta is characteristic. In addition to the common fecal bacteria, it contains fungi, thermophilic bacteria, and, in herbivorous animals, anaerobic cellulose-decomposing bacteria (581). The bacterial population of fresh cow manure was found (833) to consist of 47.5 per cent cocci {Streptococcus fyogenes, Sarcina sp., and Micrococcus candicans) ,21.2 per cent coli-like colonies {Escherichia coli, A. aero genes y and S. sefticemiae), and many dark colony-forming types. Other groups represented were Bacteroides, Flavobacteriumy Pseudomonas, Bacillus^ various anaerobic bacteria, Oidiuniy and many others. When the manure was allowed to decompose, yellow rods, fluorescent bacteria, and mesentericus types took the place of the strep- tococci. The following heterotrophic bacteria have been demonstrated (811) in manure : Bacillus subtilis, Bacillus mesentericus. Bacillus cereus, Ba- cillus tumescenSy Bacillus fetasitesy Pseudomonas fluorescenSy Pseudo- monas futiduy Salmonella enteritidisy Escherichia coliy Proteus vul- garis y Micrococcus luteusy Micrococcus candicans y Staphylococcus albusy Sarcina -flavay Streptococcus pyo genes y and others. Anaerobic bacteria are also abundant (337). Pathogenic bacteria may also occur frequently in human feces and in stable manure 5 Mycobacterium tuberculosis and various hemolytic streptococci (860), as well as pathogenic anaerobes including Clos- tridium welchiiy CI. septicumy CI. oedematisy and CI. fallax have been fpund (484). The protozoa capable of developing in manure and in urine include not only saprophytic forms but also certain parasites, such as Tricho- 22 HUMAN AND ANIMAL WASTES mastric and Trichomonas^ capable of living and even of multiplying in excreta. The coprophilic protozoa comprise various flagellates, cer- tain amebae, and ciliates. The liquid part of the manure is considerably richer than the solid in total number of protozoa as well as in species, including Polytofna uvellay Cryftochilum nigricans^ and T etramUus rostratus. These protozoa nearly all feed upon bacteria. The infusoria may feed upon smaller protozoa, so that forms like Colpdium may not destroy bacteria at all. Human and animal excreta also contain a large population of fungi, chiefly in a spore state. Schmidt (837) divided the manure-inhabiting fungi into three groups : Those found only in manure ; their spores are swallowed with the feed, and they pass unchanged through the digestive tract, though they are favorably influenced toward germination by the body heat and digestive fluids of the animal. Their natural multiplication by spores is impossible without the physiological action of the digestive proc- esses. Those that do not have to pass through the digestive tract of an animal in order to germinate and develop. The representatives of this group occur in nature only in manure, although some are able to grow also on other substrates. They can be cultivated both on manure and on other media, mostly at ordinary temperatures. Organisms found both in manure and on other substrates. They grow readily at room temperature on a number of media. Comfosition and Decomposition The chemical composition of human and animal excreta, and of stable manures in general, varies considerably, depending on the nature of the animal, its age, mode of nutrition, and composition of food- stuffs (463). As soon as voided, manure begins to undergo rapid de- composition. This results in the formation of ammonia and various other nitrogenous degradation products. These give rise to offensive sm.ells, which are controlled by the conditions of decomposition. From a sanitary point of view, it is essential that decomposition should be accompanied by the destruction of the injurious organisms present in the manure. The fecal organisms gradually disappear and their place STABLE MANURES AND FECAL RESIDUES 23 is taken by a population concerned in the decomposition of cellulose, hemicelluloses, and proteins. The decomposition of complex plant and animal residues leads to a rapid reduction in carbohydrates and is accompanied by the evolution of considerable heat, the temperature of the compost reaching as high as 75° C, as shown in Figure i. In order to hasten the decomposition of manure, conditions must be favorable to the activities of microorganisms. It must be properly 2z 1- LU «/5q 70 _ t^^' §0 60 1 / o^- •'*• J ' / . z fej^so - \\ A : ^.-^ NO BEDDING .WITH BEDDING 15 20 25 30 35 40 45 COMPOSTING PER.10D IN DAYS Figure i. Influence of straw bedding upon temperature changes in the composting of manure. Circles indicate times of turning composts. From Waksman and Nissen (961). aerated and well moistened but not saturated with water. By placing the manure, together with the waste materials of the farm and the home, in heaps, designated as composts, the decomposition processes can be controlled so as to lead to heat liberation j this results in the destruction of the injurious organisms and the conservation of the plant nutrient elements. When not properly regulated, the decomposition processes may be wasteful, unsanitary, and unsightly, and may even become a source of infection to man and his domesticated animals. 24 HUMAN AND ANIMAL WASTES SEWAGE Disposal of sewage and other home wastes is one of the important sanitary problems of men living in industrial and residential centers. Haphazard methods of disposing of sewage not only lead to conditions most unpleasant to human habitation but they are dangerous from the standpoint of infectious diseases. Sewage abounds in microorganisms that originate not only from hu- man excreta but also from other household and industrial wastes. The various saprophytic bacteria present in sewage rapidly attack the or- ganic constituents and bring about their gradual mineralization. The destructive action of saprophytic organisms greatly reduces the number of pathogens (342, 343). Activated sludge, for example, has been shown (882) to possess a definite and consistent bactericidal action against the colon bacteria. In addition to antagonistic organisms, active bacteriophages against nearly all types of intestinal bacteria are present in sewage. The destruction of pathogens by bacteriolysis thus readily finds a place in the activated-sludge method of sewage purification. Dissolved oxygen is generally present when sewage is diluted with water. As the destruction of the organic matter proceeds rapidly, the oxygen becomes depleted, so that none is left after a few hours. The predominant bacterial flora of the water may then become anaerobic, with the result that the chemical processes of decomposition are com- pletely changed J hydrogen sulfide, mercaptans, and other foul -smell- ing substances are then formed. This is accompanied by a typical anaerobic breakdown of carbohydrates, leading to the formation of vari- ous organic acids, carbon dioxide, hydrogen, and methane. The nitro- gen in the protein and urea is transformed to ammonia and various amines. When sewage is aerated, the anaerobic processes gradually give way to aerobic processes, as the oxygen diffuses into the liquids or as the sewage is diluted with water containing dissolved oxygen. When sewage is freed from solids by sedimentation before discharge, or when it is aerated sufficiently to maintain the concentration of dis- solved oxygen, decomposition proceeds rapidly without the production of the bad odors usually associated with the anaerobic breakdown. The destruction of the pathogenic bacteria results largely through the ac- GARBAGE 25 tivities of the saprophytes (809, 1008). For the purpose of promoting the development of aerobic bacteria, processes employing the use of intermittent sand filters, broad irrigation, contact beds, trickling filters, and activated sludge are applied. The modern methods of sewage purification are based on the long- known fact that the soil is a destroyer of offensive wastes. In early days, in fact, the soil handled all sewage problems. Sewage disposal plants in modern cities are so operated that microorganisms found to be so effi- cient in the soil are able to act under optimum conditions, resulting in rapid purification. Sewage freed from most of its organic constituents can be discharged into a stream and will not deplete the water of its dis- solved oxygen. Chlorine is frequently employed in the final treatment to assure the complete destruction of the pathogens. GARBAGE The processes involved in the disposal of garbage from the home are similar to those utilized in the disposal of stable manure rather than of sewage. At present, garbage usually is destroyed by burning, which results in great economic waste, or is dumped outside cities, thus creat- ing centers of infection and unpleasant appearance. More logical and less wasteful processes are based upon the principle of composting. Sev- eral of these processes are now utilized in India and China, where eco- nomic pressure is greatest. By proper handling, a product is formed that is free from injurious insects, parasitic worms, and bacteria, and that conserves all the valuable elements essential for plant growth. DESTRUCTION OF INJURIOUS MICROORGANISMS Improper methods of disposal of human and animal wastes were responsible, in the early history of mankind, for many epidemics of cholera, typhoid, plague, and other diseases. Only in recent years, after man learned the nature of the spread of these diseases, were proper methods developed for disposing of human wastes. Fecal-borne diseases rank with venereal disease and tuberculosis as the most important infectious diseases of China, because the people 26 HUMAN AND ANIMAL WASTES do not maintain proper sanitation and because human excreta are used as fertilizers. Any successful system for the control of these diseases must be sanitary and at the same time profitable. Of 1,190 persons examined, 81 per cent were positive for ascaris, with an average egg count of 14,000 per cubic centimeter. Children had a higher count than adults, and females a higher count than males. The life habits of the Chinese people are highly favorable for the spread of ascaris. By a special process of composting of feces, sufficient heat was produced to destroy disease-producing organisms and their reproductive bodies. The compost thus produced is highly effective as a fertilizer (1023). SURVIVAL OF HUMAN AND ANIMAL PATHOGENS IN SOIL AND WATER During the period 1878 to 1890 following the brilliant work of Pasteur, when bacteriology was still in its infancy, medical bacteriolo- gists took much interest in soil microbes. This was due largely to the belief that causative agents of disease that find their way into the soil may survive there and thus become a constant and important source of infection. The introduction by Koch, in i88i-, of the gelatin plate method placed in the hands of the investigator a convenient procedure for measuring the abundance of the soil population and determining the survival in the soil of agents causing serious human diseases. In spite of the fact that this method revealed only a very small part of the soil population, it enabled the medical bacteriologist to establish beyond doubt that such organisms tend to disappear in the soil. This resulted in definite conviction on the part of the public health and medical world that the soil is seldom a source of infection. It was soon demonstrated that disease-producing agents die out in the soil at a rather rapid rate, depending on the nature of the organisms, the soil, climate, and other conditions. Organisms that Survive for Long Periods Only a few disease-producing microorganisms are able to survive in the soil for any considerable periods of time. These few include the or- ganisms causing tetanus, gas gangrene, anthrax, certain skin infections, SURVIVAL OF PATHOGENS IN SOIL AND WATER 27 actinomycosis in cattle, coccidiosis in poultry, hookworm infections, trichinosis, enteric disorders in man, blackleg in cattle, and Texas fever. To these may be added the botulinus organism and others producing toxic substances, as well as bacteria, actinomycetes, and fungi that cause plant diseases such as potato scab, root rots, take-all disease of cereals, and damping-off diseases. Anthrax, a scourge of cattle and sheep, is a persistent survivor in soil } spores of this organism are known to retain their vitality and viru- lence for fifteen years. Anthrax survives particularly well in damp re- gions, especially in soils rich in decomposing organic matter j the hay and feed from these lands may transmit the disease to animals. The fact that certain fields carry anthrax infection ("anthrax pastures") was recognized in Europe long before the nature of the disease was known. Human infection results from contact with diseased animals or animal products. The anaerobic, spore-forming bacteria that cause gas gangrene are widely distributed in nature. They are found extensively in soils and in decomposing plant and animal residues. The causation of disease by these organisms received particular attention during the first world war, which was fought chiefly in trenches. Another important pathogenic anaerobe able to survive in soil for long periods of time is CI. chauvoei, the causative agent of blackleg in cattle J southern pastures are said to be better carriers of blackleg than northern pastures. CI. tetani is also widely distributed in the soil and appears to be associated with the use of stable manures. Wounds in- fected with soil may lead, therefore, to the development of tetanus or gas gangrene and must be treated accordingly. The botulinus organism not only may remain alive in the soil for a long time (642), but it may also produce there a potent toxin that causes much loss of water fowl and other wild life. Aeration of the soil results in the destruction of this toxin by aerobic bacteria (742). Orgamsms that Survive for Brief Periods . Other pathogenic bacteria, however, are able to survive in the soil only for limited periods of time. They are eliminated sooner or later from the soil, either because of their inability to compete with the soil 28 HUMAN AND ANIMAL WASTES population or because of their actual destruction by the latter. Although the pathogens seem to possess considerable resistance toward unfavor- able soil conditions, they are unable to multiply at rates permitting their indefinite survival in the soil. The anthrax bacillus and certain other parasites infesting domesticated and wild animals belong to this group. Certain insect and animal carriers make possible the survival and spread of many pathogens in the soil. The great majority of disease-producing bacteria, however, are able to survive only for very brief periods outside their respective hosts, especially in soil and water. It is sufficient to cite the fact that typhoid and dysentery bacteria, which are known to contaminate watersheds and water supplies, disappear sooner or later. It has been estimated, for example, that in sewage sludge free to undergo normal digestion, typhoid bacteria probably survive for less than 7 days. It was sug- gested, therefore, that sludge held in a digestion tank for about 10 days might be applied to the soil for fertilizer purposes without detriment to public health. The gram-negative bacteria of the typhoid-dysentery group die out rapidly in septic material; the typhoid bacteria survive for about 5 days, the Flexner type of dysentery for about 3 days, and the Shiga bacillus dies out even in a shorter period. If decomposition in the tank has not advanced far enough, as shown by low alkalinity, the organisms may survive for a much longer period. The efficiency of ripe tank ef- fluent to destroy bacteria is believed to be due to both the alkaline re- action and the presence of antagonistic metabolic products. The destruc- tion of typhoid and dysentery bacteria in the soil depends on a number of factors, chief among which are the moisture content and reaction, and the nature and abundance of the microbiological population. In moist or dry soils, most of the pathogenic bacteria were found to die within 10 days (510). Numerous other pathogenic agents, including those causing some of the most deadly human and animal scourges — tuberculosis, leprosy, diphtheria, pneumonia, bubonic plague, cholera, influenza, mastitis and abortion in cattle, the many poxes — constantly find their way into the soil in large numbers. They disappear sooner or later, and no one now SURVIVAL OF PATHOGENS IN SOIL AND WATER 29 ever raises the question concerning the role of the soil as the carrier of these disease-producing agents or as the cause of epidemics. This rapid disappearance of disease-producing bacteria in the soil may be due to a number of factors: (a) unfavorable environment j (b) lack of sufficient or proper food supply j (c) destruction by predacious agents such as protozoa and other animals j (d) destruction by various sapro- phytic bacteria and fungi considered as antagonists j (e) formation by these antagonists of specific toxic or antibiotic substances destructive to the pathogens J (f ) in the case of some organisms at least, increase of the bacteriophage content of the soil resulting in the lysis of some bacteria, especially certain spore-formers (50). The course of survival of only a few disease-producing organisms outside the host has been studied in detail. Sufficient information has been accumulated, however, to justify certain general conclusions. When E. colt is added to sterile soil, it multiplies at a rapid rate (Table 3, p. 10), but when added to fresh, nonsterile soil it tends to die out quickly (Table 4). The rate of its disappearance is independent of re- action of the soil and of incubation temperature. In order to illustrate the fate of certain important disease-producing TABLE 4. SURVIVAL OF BACTERIA ADDED TO SOIL AND THEIR EFFECT UPON THE SOIL MICROBIOLOGICAL POPULATION INOCULUM INCUBATION ORGANISMS RECOVERED* Number Tem- Coliform of days perature Total bacteria Control soil 5 28° C. 21,400 <200 E. colt addedf 5 28° c. 25,600 6,800 E. colt added$ 5 28° c. 39>700 3,500 E. coli added 5 37° C. 22,800 4,700 Control soil 33 28° c. 5,900 'i days, but not for 100 days j on ice, virulence was still evident after 102 days but not after 153 days (657). Pure cultures of the bovine organism mixed with cow manure and ex- posed in a 2-inch layer in a pasture remained virulent for 2 months in sunlight and longer in the shade. Tubercle bacteria were still alive in a garden soil on the 213th day and dead on the 230th day. They were alive in buried tuberculous guinea pigs on the 71st day, and dead on the 99th day. In running water, they survived for more than a year (86). Mycobacterium tuberculosis survived for 309 days in sputum kept in darkness, even when completely desiccated j in decomposing sputum, living organisms could be isolated after 20 days but not after 25 days (871a). Under conditions prevailing in southern England, it was found ( 1020) that the tubercle organism may remain alive and viru- lent in cow's feces exposed on pasture land for at least 5 months dur- ing winter, 2 months during spring, and 4 months during autumn j in summer, no living organisms were demonstrated even after 2 months. Under protection from direct sunlight, the survival period was longer. Feces protected from earthworms yielded viable cells even after 5 months. Virulent bacteria were still present in stored liquid manure at least 4 months after infection, though during this time a gradual reduc- tion in virulence of the organism was observed. The addition of manure to soil was found to favor the survival of the tubercle bacteria, as indicated by a higher proportion of test animals becoming tuberculous when the amount of manure added to the soil was increased (613). Positive tests were obtained for soil and manure after 178 days but not later. The organism survived on grass for at least 49 days. Rhines (780) found that M. tuberculosis multiplied in sterile soil 34 HUMAN AND ANIMAL WASTES as well as in the presence of certain pure cultures of bacteria j however, a fungus was found to check the development of the pathogen, especially in manured soil. In nonsterile soil, the pathogen was slowly destroyed, the plate count being reduced to about one sixth of the original in I month. In a study of the survival of avian tubercle bacteria in sewage and in stream water, there was a reduction, in 73 days, from 48,000 to 1,400 per milliliter in sewage and to 4,200 in water (779). Other Disease-f reducing Microorganisms A study of the viability of Brucella rjtelitensis in soil and in water in Malta brought out the fact that the organism survived in sterile tap water 42 days and in unsterile tap water only 7 days. It survived 25 days in soil and 69 days in dry sterile soil, but only 20 days in unsterile manured soil, 28 days in dry natural road dust, 20 days in dry sterile sand, and 80 days on dry cloth (334, 446). The rapid destruction of cholera bacteria added to soil was first pointed out by Houston (451). Similar rapid destruction of the diph- theria organism was also noted. Serraiia, however, retained its vitality for 158 days. Vibrio comma also survived for a short time only in feces (362), different strains showing considerable variability; temperature was an important factor. During the hot season in Calcutta, the viable period was somewhat longer than a day, as compared to 7 or 8 days during the cold season ; the critical cholera months were found to fol- low directly the cool months. The organism did not survive very long in fresh water, although the time appeared to be long enough to cause occasional serious epidemics. It remained alive for 47 days in sea water (459). The conclusion was reached that although the organism is ordi- narily destroyed rapidly in water as a result of competition with other microbes, it may survive in certain instances for some time. As a result of the evidence presented above and of other information not reported here, it has gradually become established that the soil has an enormous purification or sanitation effect upon the pathogenic bacteria brought into it either by direct excreta, in sewage, or in other- wise contaminated waters. This effect is of a double kind: (a) physical adsorption of the bacteria upon the soil, light, porous, sandy soils being SURVIVAL OF PATHOGENS IN SOIL AND WATER 35 far less efficient in removing the bacteria than heavy loam or clay soils j (b) biological effect, resulting in the destruction of the bacteria in the soil. As a result of early studies on the survival of the cholera organism in the soil, certain soils became recognized as "cholera immune" or as "cholera destroying" (736). CHAPTER 3 INTERRELATIONSHIPS AMONG MICROORGAN- ISMS IN MIXED POPULATIONS It must not be jor gotten that there are extremes in another di- rectton, where one of the two associated organisms is injuring the other, as exemflified by many farasites, but these cases I leave out of account here. This state of affairs has been termed antibiosis. — H. M. Ward. The antagonistic effects of one organism upon another were observed by many of the early microbiologists. It is sufficient to cite here three striking examples based upon totally different approaches to the sub- ject. In 1876, Tyndall (919), on the basis of the growth of wild cultures of bacteria and fungi in organic media, spoke of "the struggle for existence between the Bacteria and the PenkilUum. In some tubes the former were triumphant j in other tubes of the same infusion the latter was triumphant. The Bacteria which manufacture a green pigment appear to be uniformly victorious in their fight with the Penicillium." In 1877, Pasteur (710) noted that the production of anthrax in sus- ceptible animals can be repressed by the simultaneous inoculation with B. anthracis and various other bacteria. This led him to make the foUov/ing significant suggestion: ". . . on peut introduire a profusion dans un animal la bacteridie charbonneuse sans que celui-ci contracte le charbon: il suffit qu'au liquide qui tient en suspension la bacteridie on ait associe en meme temps des bacteries communes." In 1879, DeBary (172) emphasized the significance of the antag- onistic interrelations among microorganisms j when two organisms are grown on the same substrate, sooner or later one overcomes the other and even kills it. These and other observations thus laid the basis for a study of mutu- alistic effects of microorganisms in natural and in artificial environ- ments. SYMBIOSIS AND ANTIBIOSIS 37 SYMBIOSIS AND ANTIBIOSIS Microbes grow and bring about many metabolic reactions in natural substrates, such as soils and water basins, in a manner quite different from those in pure cultures where they are not influenced by the growth of other organisms. In artificial and natural media, whether these be synthetic materials, complex organic mashes and infusions used for the preparation of industrially essential products, or the bodies of plants and animals, pure cultures of microbes are free from the asso- ciative and competitive effects of other microbes found in natural sub- strates. In mixed populations, a number of reactions that do not com- monly take place in pure cultures are involved. Even in the case of mixed infections, a pathogen may be preceded or followed by one or more saprophytes, whereby the processes of destruction brought about in the living animal or plant body are alleviated or hastened. In the mixed populations found in natural substrates, the ecological relation- ships are largely responsible for many of the essential differences in the behavior and metabolism of the microbes, as compared with the same organisms growing in pure culture. Almost all microorganisms inhabiting a natural milieu, such as soil or water, are subject to numerous antagonistic as well as associative, or even symbiotic, interrelations. Every organism is influenced, directly or indirectly, by one or more of the other constituent members of the complex population. These influences were at first visualized as due primarily to competition for nutrients. This was well expressed by Pfeffer, who said that "the entire world and all the friendly and an- tagonistic relationships of different organisms are primarily regulated by the necessity of obtaining food." It was soon recognized, however, that this explanation does not account fully for all the complex inter- relations among microorganisms in nature. Symbiotic, or mutualistic, and antagonistic relationships among mi- croorganisms indicate whether advantages or disadvantages will result to the organisms from the particular association j the first are beneficial and the second are injurious and may even be parasitic (41, 982). When two organisms are capable of utilizing the same nutrients but are differently affected by environmental conditions such as reaction, air 38 INTERRELATIONSHIPS AMONG MICROORGANISMS supply, and temperature, the one that finds conditions more suitable for its development will grow more rapidly and in time be able to suppress the other. According to Porter (729), the effects produced by fungi in mixed culture are due either to exhaustion of nutrients or to the formation of detrimental or beneficial products. When two or more organisms live in close proximity they may exert antagonistic, indifferent, or favorable effects upon one another. These potentialities were later enlarged (1046) to include stimulating, inhibiting, over- growing, and noninfluencing effects. After considerable experimenta- tion and speculation, Lasseur (548) came to the conclusion that antago- nism is a very complex phenomenon and is a result of numerous and often little-known activities. Antagonism influences the morphology of the organisms, their capacity for pigment production, and other physiological processes. No sharp lines of demarcation can be drawn between associative and antagonistic effects. Well-defined effects of two symbionts may change during the various stages of their life cycles or as a result of changes in the environment. It is often difficult to separate strictly symbiotic phe- nomena from associations of less intimate nature, frequently desig- nated as commensalisms. The various stages of transition from obligate parasitism to true saprophytism can be represented as follows: Obligate parasitism (cer- Facultative parasitism (spe- Modified parasitism; tain bacteria, smut fungi) — > cies of Fusarium, Rhizoc- —^ hosts may derive some — > tonia, and Actinomyces') benefit (certain mycor- rhiza) Balanced parasitism (vari- True symbiosis (root- True saprophytism (auto- ous mycorrhiza) —^ nodule bacteria, lichen — > trophic and heterotrophic formations) bacteria and fungi). The phenomena of antagonism do not fit exactly into the above scheme but are parallel with it: the injurious effects of one organism upon another range from antagonism of varying degrees of intensity to the actual living or preying of one organism upon another. The lat- ter may be classified with the phenomena of parasitism and disease pro- duction. Microorganisms inhabiting the soil live in a state of equilibrium SYMBIOSIS AND ANTIBIOSIS 39 and any disturbance of this equilibrium results in a number of changes in the microbial population, both qualitative and quantitative. The ecological nature of this population found under certain specific con- ditions, as well as the resulting activities, can be understood only when the particular interrelationships among the microorganisms are recognized. Because of its complexity, the soil population cannot be treated as a whole, but some of the processes as well as some of the interrelations of specific groups of organisms can be examined as sepa- rate entities. Some have received particular attention, as the relations between the nonspore-forming bacteria and the spore-formers, the ac- tinomycetes and the bacteria, the bacteria and the fungi, the protozoa and the bacteria, and the relations of the bacteria and the fungi to the insects. The term "synergism" has been used to designate the living together of two organisms, resulting in a change that could not be brought about by either organism alone (440). Microbes living in association fre- quently develop characteristics which they do not possess when living in pure culture. For example, Schiller (835) found that when beer yeasts are placed together with tubercle bacteria in a sugar-containing but nitrogen-free medium, the yeasts develop antagonistic properties toward the bacteria and use the latter as a source of nitrogen by secret- ing a bacteriolytic subtance that is also active outside their cells. Various bacteria are able to kill yeasts when they are inoculated into suspensions of the latter in distilled water. The destruction of the fungus Ofhiobolus, the causative agent of the take-all disease of cereals, by soil organisms was believed to be a result of the need of a source of nitrogen by the latter. Papacostas and Gate (706) suggested applying the term "antibiosis" to interactions in mixed cultures in vitro and "antagonism" to mixed infections in vivo. In order to obviate a possible concept that the two types of interaction, namely in the test tube and in the living body, are different, it is more appropriate to apply the term "antagonism" to the unfavorable effects of one living system upon another and "antibiosis" to the production by one organism of specific chemical substances which have an injurious effect upon another organism. 40 INTERRELATIONSHIPS AMONG MICROORGANISMS THE NATURE OF A MIXED MICROBIAL POPULATION A mixed microbial population is made up of a great variety of bac- teria, and often also of fungi, actinomycetes, and protozoa j to these are added, under certain conditions, various algae, diatoms, nematodes and other worms, and insects. The specific nature and relative abundance of the various microorganisms making up a complex population in either a natural or an artificial environment depend upon a number of factors, which can be briefly summarized as follows : The physical nature of the medium in which the population lives: soil, compost, or manure pile; river, lake, or ocean; sewage system; or peat bog. The nature, concentration, and availability of the chemical constituents of the medium used by the microbes as nutrients, especially the ma- terials used as sources of energy and for the building of cell sub- stance. Various organic and inorganic substances, whether complex or simple in chemical composition, favor the development of specific groups of microorganisms capable of utiHzing them. For example, sulfur favors the development of specific sulfur bacteria, and cellu- lose favors such organisms as are capable of attacking this complex carbohydrate as a source of energy. In many instances there is con- siderable competition for the available food material. Organisms that possess a greater capacity for attacking the particular compound, or are capable of preventing the development of other organisms by the formation of substances injurious to the latter, usually become pre- dominant. Proteins, starches, and sugars can be acted upon by a great variety of microorganisms. The predominance of one group may depend not only upon the chance presence of the particular or- ganism or its capacity for more rapid growth, but also upon its ability to form alcohols, acids, and other products that influence the growth of other organisms. Environmental conditions favorable or unfavorable to the development of specific organisms. Of particular importance in this connection are temperature (thermophilic vs. mesophih'c organisms), oxygen supply (aerobic vs. anaerobic organisms), moisture content (bac- teria and fungi vs. actinomycetes), reaction (acid-sensitive vs. acid- ASSOCIATIVE INTERRELATIONSHIPS 41 tolerant forms), as well as the physical conditions of the substrate as a whole. The presence and abundance of organisms that produce substances having a favorable and stimulating or an injurious and toxic effect upon other organisms, or that may compete for the available nutrients. The equilibrium in the microbiological population in a natural me- dium such as soil or water may be upset by the introduction of spe- cific nutrients, as well as by treatment with chemical and physical agents whereby certain organisms are destroyed and others stimu- lated. The presence of specific microorganisms in a natural medium may be con- siderably influenced by the presence of certain parasitic or phagocytic agents. The role of protozoa in controlling bacterial activities by consuming the cells of the bacteria has been a subject of much specu- lation. The presence of bacteria, fungi, and nematodes capable of destroying insects is of great importance in human economy. Many other relationships, such as the presence of phages against specific organisms, are often found greatly to influence the nature and com- position of a specific population. ASSOCIATIVE INTERRELATIONSHIPS Numerous instances of associative interrelationships among micro- organisms are found in nature. These may be grouped as follows: Preparation or modification of the substrate by one organism whereby it is rendered more favorable or more readily available for the growth of another organism. As an illustration one may cite the breakdown of cellulose by specific bacteria, thereby making the particular en- ergy source available to noncellulose-decomposing organisms, in- cluding not only certain bacteria and fungi but also higher forms of life such as ruminant animals (herbivores) and insects (termites, cockroaches), which carry an extensive cellulose-decomposing micro- biological population in their digestive systems. Another illustration is the breakdown of complex proteins by proteolytic bacteria, result- ing in the formation of amino acids and polypeptides, which form f favorable substrates for peptolytic bacteria. The ammonia liberated from proteins and amino acids supplies a source of energy for nitrify- 42 INTERRELATIONSHIPS AMONG MICROORGANISMS ing bacteria and a source of nitrogen for many fungi. The ability of bacteria to concentrate in solution those nutrients that are present only in mere traces enables animal forms (protozoa) to exist at the expense of the bacteria ( io6). Influence upon the oxygen concentration available for respiration. This involves the phenomenon first observed by Pasteur (709) of con- sumption of oxygen by aerobic bacteria, thus making conditions fa- vorable for the development of anaerobes. Symbiotic interrelationships, where both organisms benefit from the asso- ciation. The three most important examples found in nature are: (a) the phenomenon of symbiosis between root-nodule bacteria and leguminous plants; (b) mycorrhiza formations between certain fungi and higher plants; (c) lichen formation between algae and fungi. Certain other interrelations are not strictly symbiotic, but are found to fall between groups a and c; here belong nitrate reduction accompanied by cellulose decomposition and nitrogen-fixation with cellulose decomposition, carried out in each case by two specific groups of organisms. Production by one organism of growth-promoting substances that favor the development of other organisms. The formation of riboflavin by anaerobic bacteria in the digestive tract of herbivorous animals is an interesting and highly important phenomenon in the nutrition of such animals. The production of bacterial growth stimulants by yeasts and the beneficial action of mixed populations upon nitrogen-fixation by Avcotohacter are other illustrations of this general phenomenon. The presence of specific bacteria has been found necessary for the sporulation of certain yeasts and for the formation of perithecia by various Aspergilli (825). Various other processes of association have also been recognized (940). Destruction by one microorganism of toxic substances produced by an- other, thereby enabling the continued development of various mem- bers of the microbiological population. Modification of the physiology of one organism by another. In the presence of certain bacteria, Clostridium granulobacter-fectinovorum forms lactic acid instead of butyl alcohol (873). The presence of Clos- tridium acetobutylicum in cultures of bacteria producing dextro-lactic acid and laevo-lactic acid causes such bacteria to form the inactive lac- tic acid (897) ; intimate contact of the bacteria is essential, the use of membranes preventing this effect. Pigment formation by Ps. ASSOCIATIVE INTERRELATIONSHIPS 43 aeruginosa may be weakened when the latter is grown together with other organisms. E. colt may lose the property of fermenting sugars when grown in the presence of paratyphoid organisms (462). Some associations of microorganisms are not so simple. The complex system of animal infection by more than one organism, with the result- ing complex reactions in the animal body, is a case in point. The effect of one organism upon the activities of another can be illus- trated by the results of the decomposition of complex plant material by pure and mixed cultures of microbes (Table 5). Trichoderma, a fungus TABLE 5. DECOMPOSITION OF ALFALFA BY PURE AND MIXED CULTURES OF MICROORGANISMS TOTAL HEMICELLU- CELLU- ALFALFA DE- LOSES DE- LOSE DE- NH3-N ORGANISM COMPOSED COMPOSED COMPOSED PRODUCED Per cent Per cent Per cent mgm. Trichoderma 9-3 4-7 61 Rhizopis 6.6 12.8 2.9 53 Trichoderma + Rhizofus 13-7 22.6 10.6 63 Trichoderma + Cunninghamella 15.0 15.4 5-7 47 Trichoderma + ?s. fuorescens 10.5 14.5 6.4 32 Streftomyces 3065 16.6 43-0 23.2 52 Trichoderma + Streftomyces 3065 12.5 14.6 4.8 56 Soil infusion 28.4 40.9 50.8 21 From Waksman and Hutchings (960). known to be an active cellulose-decomposing organism, did not attack at all the cellulose of alfalfa and decomposed the hemicelluloses only to a limited extent} however, the organism utilized the proteins rapidly, as illustrated by the amount of ammonia liberated. RhizofuSy a non- cellulose-decomposing fungus, attacked largely the hemicelluloses in the alfalfa and some of the protein } a small reduction in cellulose was recorded, probably because of an analytical error. When Trichoderma was combined with Rhizofus y the former attacked readily both the cel- lulose and the hemicelluloses. The same effect upon the activity of Trichoderma was exerted by other noncellulose-decomposing organ- 44 INTERRELATIONSHIPS AMONG MICROORGANISMS isms, such as the fungus Cunninghamella and the bacterium Ps. fluores- cens. On the other hand, when Trichoderma was combined with a cellu- lose-decomposing Stre'ptomyces, there was considerable reduction in the decomposition of the total plant material as well as of the cellulose and hemicelluloses. These results further emphasize the fact that two or- ganisms may either supplement and stimulate each other or exert an- tagonistic effects. The total soil population is far more active than any of the simple combinations of microorganisms. COMPETITIVE INTERRELATIONSHIPS The following competitive relations among the microscopic forms of life inhabiting the sea have been recognized ( 15) : Competition among chlorophyll-bearing diatoms for the available nutri- ent elements in the water Competition among the copepods for the available particulate food mate- rials, notably the diatoms Competition between individuals belonging to one species and individuals belonging to another Competition between young, growing, and reproducing cells and older, respiring cells Food competition and space competition Competition between transitory and permanent populations Competition between sedentary or sessile organisms and free-moving forms This list has been enlarged (943) to include other factors that are par- ticularly prominent in nonaquatic environments: Degree of tolerance of the immune or resistant varieties and of the less re- sistant or more sensitive forms to attack by disease-producing or- ganisms Fitness for survival of microbes that are able to adapt to a symbiotic form of life, such as leguminous plants or mycorrhiza-producing plants, and those that are not so adapted Survival of parasitic forms that require living hosts for their development, as contrasted with saprophytes that obtain their nutrients from min- eral elements or from dead plant, animal, and microbial residues ANTAGONISTIC INTERRELATIONSHIPS +5 Various special types of competition, for example, competition between Strains of root-nodule bacteria (Rhizobium), whereby one strain checks completely the multiplication of other strains, even outside the plant, the dominant strain then becoming responsible for all the nodules produced (679). These phenomena of competition are found not only in natural sub- strates, such as soil and water, but also in artificial media. When several microbes are growing in the same culture medium, some will be re- pressed in course of time whereas others will survive and take their place. This is due to the fact that these microbes compete for the use of the same nutrients or that conditions, such as reaction, oxygen supply, and temperature, are more favorable to some organisms than to others. Another phenomenon may also be involved, that some organisms may produce toxic substances that repress the growth of others. In artificial media, slowly growing tubercle bacteria, diphtheria organisms, and others will be repressed by the rapidly growing saprophytes. Under aerobic conditions, aerobic bacteria and fungi will repress yeasts and anaerobic bacteria, whereas under anaerobic conditions the reverse will take place. An alkaline reaction will favor the development of bacteria, an acid reaction will favor the growth of fungi. ANTAGONISTIC INTERRELATIONSHIPS When two or more organisms live together, one may become antag- onistic to the others. The composition of the medium and the conditions of growth influence the nature and the action of the antagonist j the metabolism and cell structure of the antagonized organism may be modified or the cell itself may be destroyed (184). In urine, for example, staphylococci may become antagonistic to E. coli or vice versa, depending on the initial numbers of the two groups, on the formation of metabolic products, or on the exhaustion of nutrients (246). The toxic substances produced by the antagonists comprise a variety of com- pounds, ranging from simple organic acids and alcohols to highly com- plex bodies of protein or polypeptide nature. Various types of antagonism are recognized. Nakhimovskaia (670) L 5 B R A 46 INTERRELATIONSHIPS AMONG MICROORGANISMS concluded that all phenomena of antagonism among microorganisms can be conveniently classified into four groups : 1. Antagonism in vivo vs. antagonism in vitro. According to some inves- tigators, only the inhibitive forms of antagonism (in vitro) may be designated as true antagonisms; the in vivo forms w^ere designated as phenomena of antibiosis. As pointed out above (p. 38), this differentiation is no longer recognized. 2. Repressive, bactericidal, and lytic forms of antagonism. One may fur- ther distinguish between bacteriostatic and bactericidal, fungistatic and fungicidal forms of antagonism, as well as between antagonism of function and antagonism of growth. 3. Direct, indirect, and true antagonism. 4. One-sided and two-sided antagonism; antagonism between strains of the same species and antagonism among strains of different species. Duclaux (212) was the first to demonstrate that the growth of a fungus upon a certain medium renders the medium unfavorable for the further growth of the same organism. Kiister (541) has shown that culture solutions in which fungi have grown are not suitable for the germination of freshly inoculated spores but are improved by boiling. This effect was observed as a result of the growth not only of the same organism but also of other species. Similar observations were made for bacteria: Marmorek (620) reported, in 1902, that the growth of hemolytic streptococci in broth rendered the medium unsuitable for subsequent growth of the same organism. The production of spores by bacteria was believed to be caused by the formation of toxic, thermola- bile organic substances j upon the destruction of these by boiling, the medium was again made favorable for the growth of bacteria and bac- terial spores were once more able to germinate. Some of the toxic sub- stances appeared to be thermostable (668). Fungi are capable of producing not only growth-inhibiting but also growth-promoting substances. By means of certain procedures, it was found possible to separate the two (690). The tendency of fungus hyphae to turn away from the region in which other hyphae of the same fungus were growing was explained as a negative reaction to chemical substances produced by the growing fungus (306). This nega- ANTAGONISTIC INTERRELATIONSHIPS 47 tive chemotropism was shown to be due to thermolabile staling sub- stances (358). The phenomenon of staling was often spoken of as vacci- nation of medium (45), and was ascribed to the action of protein degra- dation products. These and other experiments led to the conclusion that many micro- organisms are capable of producing substances that are injurious to their own development (iso-antagonistic) or, and sometimes much more so, to other organisms growing close to them (hetero-antagonistic). The growth of certain fungi and bacteria in practically pure culture, even in a nonsterile environment, was believed to be due to this phenomenon. It is sufficient to mention the lactic and butyric acid bacteria, the citric- acid-producing species of As-pergillus, the lactic and fumaric-acid- producing species of RMzofus, and the alcohol-producing yeasts. The chemical substances produced by these organisms in natural substrates may be looked upon as protective metabolic products of microorgan- isms in their struggle for existence. Such products play a highly sig- nificant part in the metabolism of various organisms, especially those that grow parasitically upon living plant and animal bodies. Among the various types of antagonism, the one resulting in the pro- duction of active substances that can be isolated and purified has re- ceived the greatest consideration recently. These substances have been designated as toxins, poisons, antagonistic agents, bacteriostatics, and antibiotics. The chemical nature of some has been elucidated, but that of many others is still unknown. Some of these substances are destroyed by boiling, by exposure to light, or by filtration, whereas others are re- sistant to heat and to ultraviolet raysj some are readily adsorbed by certain filters, from which they can be removed by the use of special solvents such as ether, alcohol, chloroform, and acetone. The concen- tration of the antagonistic substances produced by many fungi and bac- teria is greatly influenced by the energy and nitrogen sources in the medium and by environmental conditions, such as temperature and aeration. The three important types of antagonism are (a) the repressive, in- hibitive, or bacteriostatic, (b) the bactericidal, and (c) the bacteriolytic. When one bacterium is inoculated into the filtrate of another, the growth of the first is slower than that of the control. Certain types of 48 INTERRELATIONSHIPS AMONG MICROORGANISMS antagonism express themselves in the destruction by the antagonist of the other organisms present in the mixed culture, with or without lysis. B. mesentericusj for example, is capable not only of depressing but also of killing the cells of diphtheria and pseudodiphtheria (1052). The lytic form of antagonism is illustrated by the action of Ps. aeruginosa, B. brev'ts, and certain other antagonists upon micrococci and various spore-forming bacteria. In differentiating between "direct antagonism" and "passive antag- onism," attention was directed (670) to the fact that the latter depends not upon the direct action of the antagonist but upon the changed con- ditions of culture under the influence of the antagonist's growth. This may comprise a change in ^H and r¥L of medium or an impoverish- ment of some of the nutrient constituents. "Direct antagonism" was often distinguished (677) from "indirect antagonism," the first being limited to those phenomena in which the antagonistic action is con- nected with the direct action of the living cell, whereas in the second the metabolic products produced by one organism are injurious to others. Intestinal bacteria were found (367, 369) to repress the anthrax organ- ism only when the former were in an active living state. Other investi- gators (418) designated the action of the living cell itself as "true antagonism." Bail (32) suggested that for every bacterium there is a typical constant number of cells capable of living in a given space. When this concentration (M) is reached, multiplication comes to a standstill, in- dependently of exhaustion of the nutrients or formation of toxic sub- stances. The same phenomenon was believed to hold true when two bacteria live together (1013): if the limiting cell-in-space concentra- tions are different for the two organisms, the one with a higher M value represses the other; however, the weaker species may check the stronger when planted in sufficient excess (243). It has been suggested (370) that certain physiological properties of the individual organisms, desig- nated as "biological activity" and "competitive capacity," must also be taken into consideration in evaluating this relationship. The fact that the number of yeast cells reaches a maximum independently of the ini- tial number of cells added or the concentration of nutrients in a given ANTAGONISTIC INTERRELATIONSHIPS 4-9 volume of medium has been explained (91) by the amount of oxygen originally present. Garre (315) deserves the credit for having first noted that antago- nism may be either one-sided or two-sided. In the first case, one organ- ism represses another that is not antagonistic to itj in the second case, both organisms repress each other. A one-sided antagonism may become two-sided under certain conditions of culture. E. coli is antagonistic to E. typhosa; however, if the latter is inoculated into a medium some- what earlier than the former, E. tyfhosa becomes antagonistic to E. coli (936). Although the most common antagonisms are between organisms of different species, there are numerous instances where one strain of a species may be antagonistic toward another strain of the same species (53j 370? 651)- Certain strains may develop antagonistic properties in the presence of other strains (74). Nonflagellated variants of typhoid bacteria were repressed by a flagellated form, smooth variants of para- typhoid bacteria by rough forms, and so on. The fact that all bacterial cultures stop growing after a certain period of time has been interpreted to be a result of the antagonistic action of some cells upon others. When the filtrates of such cultures are added to fresh nutrient media they may stop the growth of the same species as well as that of other species. Certain organisms produce pigments in the presence of others j these pigments are believed to be in some way associated with the phenome- non of antagonism. In the presence of S. lutea^ V. comma forms a dark violet pigment that is accompanied by an increase in agglutination and in virulence (670). The destruction of Dictyostelimn muco- roides by a red-pigment-forming bacterium was accompanied by an in- crease in intensity of the pigment (723)5 the blue pigment of Bac- terium violaceumy however, only delayed the growth of the fungus. Penicillium ajricanum produces a more intense pigment in contact with other fungi such as Asfergillus niger; this pigment accumulates in the mycelium of the latter, which may thereby be killed (186). P. luteum and Sficaria furfurogenes produce a pigment that is used not only for purposes of protection, but also for attack upon other organisms, whereby the latter are killed and stained (669). 50 INTERRELATIONSHIPS AMONG MICROORGANISMS THEORIES OF THE NATURE OF ANTAGONISTIC ACTION The various theories proposed to explain the mechanism of antago- nistic effects of microorganisms may be summarized under the follow- ing processes: Exhaustion of nutrients Physicochemical changes in medium Pigment action Action at a distance Space antagonism Enzyme action, either directly by the antagonist or as a result of cell autolysis, under the influence of the antagonist Production and liberation of antibiotic substances Pasteur (710) ascribed the antagonistic effect that aerobic bacteria have upon the anthrax organism to the consumption of the oxygen by the former} the unfavorable influence of normal blood upon the growth of anthrax was believed to be due to competition for the oxygen by the red blood corpuscles. Freudenreich (298) considered the antagonism between Ps. aeruginosa and Bacillus anthracis as due to exhaustion of nu- trients by the former. These studies were soon followed by numerous other investigations in which the exhaustion of nutrients in the media was believed to be responsible for the phenomenon of antagonism j the onset of the stationary phase in bacterial growth was believed (579) to belong here. Change in ^H of medium, exhaustion of nutrients, and accumulation of toxic products were also found to be limiting factors. It thus became apparent, even in the early days of bacteriology, that certain changes are produced by microbes in the medium in which they grow which render it unfit for the growth of other organisms. It also was soon recognized that the problem is more complicated than the mere exhaustion of nutrients. The changes in relationship produced by changes in surface tension, in oxidation-reduction potential, in reaction, and in osmotic pressure were suggested as explanations. Among the classical examples of the effect of reaction upon the growth of other organisms is the acidification of milk by lactic acid bacteria. Metchnikov emphasized the fact that Lactobacillus bulgaricus acts antagonistically NATURE OF ANTAGONISTIC ACTION 51 not only by means of the lactic acid that it produces but also by the formation of special substances. The production by bacteria of alkali- reaction products that have an injurious effect upon the further growth of the organisms has also been demonstrated (342). These substances were found to correspond to amino compounds, liberated in the process of cellular disintegration. Numerous other physical and physicochemi- cal factors influence the growth of an organism in an artificial medium. It is to be recalled that the rate of survival of bacterial cells in water or in salt solution is markedly influenced by the colloids present, the con- centration of electrolytes, the reaction, and the temperature. Microbial antagonism was thus looked upon largely as a result of a series of physical factors, including various radiations such as mytoge- netic rays, -pH changes, conductivity, electric charge, and surface ten- sion (525). Most antagonisms, however, can be explained by the production of antibiotic substances by the antagonists. Because of the thermolability of some, sensitivity to chemical reagents, or adsorption on bacterial filters, considerable difficulty has been experienced in isolating the active substances. Many of these substances are iso-antagonistic, where- as others are able to act upon different bacteria. Most of them have been found to be thermostable. The first antibiotic recognized as such was pyocyanase, produced by Ps. aeruginosa (235). Other organisms that produce such substances are Serratia marcescens (229), Ps. jluorescens {S^^)-, B. mesentericus (1052), B. mycoides, B. subtilis, and other spore-forming bacteria. Since the early work at the turn of the century and especially during the last five or six years, many new antibiotics have been isolated or demonstrated. These will be discussed in detail later. The production of these antibiotics by microorganisms is greatly influenced by reaction, temperature, and aeration of substrate, as well as by the presence of other organisms. Evidence is still lacking as to whether these substances may accumulate in the soil and in water, whether the antagonized organisms are able to overcome their effect, and whether they are destroyed by other members of the soil or water microbiological population (365, 976). Different organisms possess different degrees as well as different 52 INTERRELATIONSHIPS AMONG MICROORGANISMS mechanisms of antagonism. Often one organism may completely check the growth of another j later, growth may be resumed, although it will not be quite normal. Antagonism stimulates spore-production and brings about deformed growth of the mycelium in fungi or the forma- tion of gigantic cells in bacteria. The morphological effects produced by the antagonists comprise changes in form, size, and structure of hyphae, direction of growth, complete cessation of growth, and ab- breviation of hyphal segments. CHAPTER 4 ISOLATION AND CULTIVATION OF ANTAGONISTIC MICROORGANISMSj METHODS OF MEASURING ANTIBIOTIC ACTION In nearly all the earlier work and even in many recent investigations on the antagonistic properties of microorganisms and the production of antibiotic substances, two procedures were employed : indiscriminate testing of pure cultures of bacteria and fungi, commonly taken from culture collections, for antagonistic effects against one another or against certain specific or test organisms} and isolation of occasional antagonistic organisms from old plate cultures, as air contaminants, or from mixed infections. These studies were carried out either by medical bacteri- ologists interested in agents capable of suppressing bacterial pathogens or by plant pathologists interested in organisms capable of inhibiting the growth of fungi, principally those concerned in the causation of plant disease. They resulted in the accumulation of considerable infor- mation concerning antagonistic organisms, the nature of the phenome- non of antagonism, and, to a more limited extent, the mechanisms in- volved. Neither of these methods, however, is suitable for a systematic study of antagonism as a natural process. The last decade has witnessed a number of systematic attempts to de- termine the distribution of antagonists in nature, to isolate specific or- ganisms capable of bringing about the desired reactions, and to estab- lish the mechanism involved in these reactions. These studies were undertaken by a group of Russian investigators interested largely in fungi and actinomycetes as agents antagonistic to other microorganisms chiefly causing plant diseases, and by American and British investigators interested in agents active against bacterial pathogens of man. The early significant, but unrecognized, investigations of Schiller (835) on forced antagonisms and the studies of Gratia and his asso- ciates (356, 357) on mycolysates were in direct line of the more re- cent studies of Dubos (201), who made a systematic attempt to isolate from specially enriched soils bacteria capable of destroying specific 54 ISOLATION AND CULTIVATION OF ANTAGONISTS pathogenic organisms. Although it had been previously established that many spore-forming bacteria are capable of producing substances that have antibacterial properties, as shown by the work of Pringsheim (738), Much (664), and others, Dubos was the first to succeed in iso- lating in crystalline form the active substances involved and in demon- strating their chemical nature. He utilized for the isolation of the or- ganisms the soil enrichment culture method. This consisted in adding repeatedly various pathogenic bacteria to a soil in which, as a result, antagonistic organisms developed that were capable of destroying the bacteria i these organisms were then isolated by appropriate procedures. These investigations, as well as the work of Fleming (261 ) later fol- lowed by other British investigators (5) on the antibacterial properties of molds belonging to the Penkillium notatum group, served as the di- rect stimulus to numerous studies. The entire series of studies led to the development of simple methods for the systematic isolation of microorganisms capable of inhibiting the growth of fungi and bacteria, both pathogenic and saprophytic, and for separating many of the anti- biotic substances produced by these organisms. METHODS OF ISOLATING ANTAGONISTIC MICROORGANISMS Several methods are now available for the isolation of antagonistic microorganisms from natural substrates such as soil, stable manure, composts, sewage, water, and food products. These methods are dif- ferent in nature, but they are all based on the same principle, that of bringing a living culture of a bacterium or fungus into close contact with a mixed natural population, thereby allowing certain members of this population to develop at the expense of the added culture. Soil Enrichment Method By this method a soil is enriched with known living pathogenic bac- teria. Fresh garden or field soil is placed in glass beakers or pots, and the moisture of the soil is adjusted to optimum for the growth of aerobic bacteria, which is about 6s per cent of the water-holding capacity of the soil (20 to 50 per cent of the moist soil) j the containers are covered METHODS OF ISOLATION 5 5 with glass plates and placed in an incubator at 28° or 37° C. Washed suspensions of living bacteria are added to the soil at frequent intervals, care being taken to avoid puddling it with an excess of the fluid, so con- ditions will not be made anaerobic. Samples of the enriched soil are removed at intervals and tested for the presence of organisms antag- onistic to the bacteria added. Fresh washed suspensions of the living bacteria are inoculated with the enriched soil as soon as the presence of antagonistic organisms is demonstrated j this results in the development of the antagonistic organisms and the destruction of the bacteria in sus- pension. Transfers are then made to fresh suspensions of the bacteria, resulting in an enrichment of the antagonist, which can finally be iso- lated in pure culture (201, 207, 442). The significance of the soil enrichment method and its application to the isolation of specific antagonistic organisms has been questioned (969). It was suggested that whereas there is no question concerning the multiplication of microorganisms capable of decomposing a given substance or of secreting enzymes active upon such a substance in re- sponse to its introduction into the soil, there is still doubt whether specific antagonistic organisms develop as a result of the introduction of living cells into the soil. The reason for this was based upon the fact that antibiotic reactions produced by antagonistic organisms do not affect bacteria by simple digestive or oxidative mechanisms. Bacterial A gar Plate Method This method was first used by Gratia and Dath (357) for the isola- tion of antagonistic agents, actinomycetes having been found readily by it. To isolate antagonistic bacteria, agar (1.5 per cent) is washed in dis- tilled water, then dissolved in water supplemented by i per cent glucose and 0.2 per cent K0HPO4. Ten-milliliter portions of the sugar- phosphate agar are placed in glass tubes and sterilized. The sterile agar is melted, and the tubes are placed in a water bath kept at 42° C. A washed, centrifuged suspension of living bacteria, grown on solid or in liquid media, is then added and thoroughly mixed with the agar. This "bacterial agar" is poured into a series of Petri plates containing one- milliliter portions of fresh or enriched soil, diluted i : lOO to i : 10,000 56 ISOLATION AND CULTIVATION OF ANTAGONISTS times with sterile water. The contents of the plates are thoroughly mixed in order to distribute the diluted soil suspension in the bacterial agar. The plates are inverted and incubated at 28° or 37° C. After I to 10 days' incubation, depending on the nature of the or- ganism used for the preparation of the plates, the presence of antago- nists is manifested by the formation of clear zones surrounding their colonies (Figure 2). The organisms are isolated from these colonies and are retested for antagonistic properties, either by transfer to fresh bacterial agar plates or by inoculating solidified agar plates and cross- streaking with test organisms (956, 978). In the isolation of antagonistic fungi the same method is followed, except that it is preferable to make the bacterial agar acid by using KH2PO4 in place of K2HPO4. The resulting acidity (pH 4.5) inhibits the growth of bacteria and actinomycetes. Since the soil contains fewer fungi than bacteria, lower dilutions of soil are employed for this pur- pose (1:10 to 1:1,000). This method, like the soil enrichment method, does not always yield desirable results. As shown in Table 6, some of the most important antagonists, such as Ps. aeruginosa, S. antibioticus, A. jiavus, and P. notatum-y do not develop on such a plate since they cause only limited lysis of bacteria. On the other hand, B. brevis, S. griseus, A. fumigatus, and A. clavatus cause extensive lysis of gram-positive bacteria and so can readily be isolated. Crowded Plate Method Ordinary field or garden soil is plated out on common nutrient (beef- peptone) agar, very low dilutions (1:10 to 1:1,000) being used to enable a large number of bacterial colonies to grow on the plate. The resultant crowding of these colonies allows the development on the plate of potential antagonists that are normally present in the soil. The production of antibacterial substances by these antagonists inhibits the growth of bacteria in close proximity to them and, in consequence, clear zones are formed around the colonies (Figure 3). It is possible, by means of this method, to demonstrate that many strains of spore-form- ing bacteria possessing antagonistic properties are present in the soil and can readily be isolated from it. METHODS OF ISOLATION 57 TABLE 6. GROWTH OF ANTAGONISTIC ORGANISMS ON BACTERIAL WASHED AGAR MEDIA AND LYSIS OF BACTERIA ANTAGONISTIC ORGANISM Bacteria: B. brevis B. simflex Ps. aeruginosa Actinomycetes: 5. antibioticus S. griseus S, lavendulae Micromonosfora sp. N. gardneri Fungi: A. clavatus A. flavus A. fumigatus Glioclaiium sp. P. notatum MEDIUM CONTAINING WASHED CELLS OF E. coli S. lutea B. subtilis Growth Lysis Growth Lysis Growth Lysis O O o o O O O O O o O O O O O O h+ o H- O h+ O h+ o H- O From Waksman and Schatz (969). Note, o indicates no growtii of antagonist or lysis of test bacterium as shown by formation of clear zone on plate; ± indicates trace; + to I I I I indicates increasing amounts of growth or lysis. Direct Soil Inoculation Method Nutrient agar plates are inoculated with the bacteria or fungi for which antagonists are to be found, and the plates are incubated for 24 to 48 hours at 28° or 37° C. Particles of fresh or enriched soil placed on the surface of the bacterial or fungus growth on the plate will give rise to antagonistic organisms that will bring about the killing or even the lysis of the original culture. By this method, organisms antagonistic to many bacteria and fungi causing plant and animal diseases have been isolated (683,685). For the isolation of bacteria antagonistic to fungi, the latter are grown on potato agars until they have spread over the plate j particles of moist soil are then placed on the surface of the mycelium, and the plates are incubated in a moist chamber. Bacteria lysogenic to the fungi 58 ISOLATION AND CULTIVATION OF ANTAGONISTS grow out of the soil and gradually dissolve the mycelium until the en- tire surface of the plate becomes free of the hyphae of the fungus. By transferring some of the material from the lysed spots, pure cultures of bacteria have been obtained that are capable of producing destructive effects upon the fungi, similar to the action of the particles of soil. To these four methods may be added the "forced antagonism" method of Schiller (835), previously referred to, which consists in feed- ing a culture of an organism with another one, thereby forcing the sec- ond to develop the capacity of destroying the first. By means of the foregoing methods, as well as various modifications of them, it was possible to demonstrate that soils, composts, and water basins contain an extensive population of microorganisms possessing antibacterial and antifungal properties. When E. coli was used as the test organism, it was found that although this organism was capable not only of surviving but actually of multiplying in sterile soil, it died off very rapidly when added to fresh soil. The rate of its destruction was greatly increased with every subsequent addition of washed bacterial cells to the soil. This was accompanied by the development of certain antagonistic microbes capable of destroying E. coli in pure culture. A large number of fungi, actinomycetes, and bacteria possessing an- tagonistic properties have thus been isolated. The nature of the test or- ganism was found to be of great importance in this connection. When Stafhylococcus aureus , S. luiea, and B. subtilis were used, a large num- ber of antagonists could readily be isolated. With E. coUy however, a much smaller number of microbes thus isolated possessed antagonistic properties. Certain other gram-negative bacteria, like Brucella abortuSy were more sensitive than E. coli, whereas certain gram-positive bac- teria, like B. mycoides and B. cereus, were less sensitive than B. subtilis (956,958). Bacteria destructive to fungi, or possessing fungistatic and fungicidal properties, have also been isolated from soils as well as from the surface of plants, such as flax, by transferring small sections of soil or plant stem to plates of fungi growing on potato agarj transfers made from the lytic spots yielded antagonistic bacteria (686). By the use of this Figure 2. Development of antagonistic fungi on bacterial-agar plate. From Waksman and Horning (956). Figure 3. Bacterial plates made from soil, showing clear zones surround- ing colonies of antagonistic organisms. From Stokes and Woodward (885). Antagonistic action of .S. ant'i- b'toticus upon S. lutea Antagonistic action of S. ant'i- btoticus upon B. myco'ides Bacteriostatic action of actino- mycin upon 5. lutea Bacteriostatic action of actino- mycin upon B. myco'tdes Figure 4. Antagonistic effects of living organisms and their products. From Waksman and Woodruff (974). METHODS OF TESTING ANTAGONISTIC ACTION 59 method, Chudiakov (143) isolated various bacteria antagonistic to Fusarium. The antagonists were found abundantly in cultivated soils, but not in flax-sick soils rich in Fusarium. Bamberg (35) demonstrated, in the soil, bacteria capable of bringing about in 10 days complete de- struction of Ustiliago zeae and other fungi. Myxobacterium was also found (473) capable of bringing about the disintegration of fungus mycelium. Nonspore-forming bacteria, similar to the cultures of Chudiakov, were isolated and shown to be able to attack a number of fungi, including species of Fusarium, Sclerotinia, Gleosforium, Acro- stalagmus, Alternaria, and Zygorhynchus (729). METHODS OF TESTING THE ANTAGONISTIC ACTION OF MICROORGANISMS Once antagonistic organisms have been isolated, it is essential to es- tablish their bacteriostatic spectrum — that is, their ability to inhibit the growth of various specific microorganisms. Usually these antagonists do not affect alike all bacteria and fungi, some acting primarily against gram-positive bacteria and against only a few gram-negative forms (mostly cocci), others acting upon certain bacteria within each of these two groups. A considerable number of methods have been developed for meas- uring these antagonistic effects. They measure the selective nature of the antagonistic action and they can also give quantitative information concerning the intensity of this activity. Because of the great differences in the degree of sensitivity of bacteria to the action of the antagonists, the proper selection of one or more test organisms is highly essential. S. aureus has been employed most commonly, different strains of this organism having been found to vary greatly in their sensitivity even to the same substance. Stre-ptococcus viridans, B. subtilis, Micrococcus ly- sodeikticus, S. luiea, E. coli, and E. tyfhosa are other organisms that are frequently employed for testing the activity of antagonists. Al- though for purposes of concentration and purification of a known sub- stance a single test organism is sufficient, it has been found advisable during the isolation of antagonistic organisms and the study of the na- 60 ISOLATION AND CULTIVATION OF ANTAGONISTS ture of the antibiotic substance or substances that they produce to use more than one test bacterium, including one or more gram-positive and one or more gram-negative bacteria. Most of the methods for testing antagonistic action are based upon the growth of the test organisms in the presence of the living antago- nists or of the antibiotic substances produced by them in liquid and on solid nutrient media. Only a few of these methods are now utilized, most of them being chiefly of historical interest. Liquid Media Several methods using liquid media have been proposed for testing the antagonistic activities of microorganisms: Simultaneous inoculation of the medium with the antagonist and the test organism. Inoculation of the medium with the antagonist first, followed after 6 to 48 hours by inoculation with the test organism. Inoculation of the medium with the test organism first, followed, after a certain interval, by the antagonist. Effect of the metabolic products of the antagonist upon various micro- organisms. In 1888, Freudenreich (298) first filtered the culture through a Chamberland candle and inoculated the filtrate with the test organisms. The culture filtrate is usually added to the fresh me- dium, either previously inoculated with the test organism for the purpose of establishing the lytic effect of the filtrate, or followed by the test organism, whereby the bacteriostatic action is measured. Placing a porcelain filter or cellophane membrane between the cultures of the antagonist and of the test organism. Frankland and Ward (295) used a filter of the Pasteur-Chamberland type partly filled with broth and placed in a beaker containing the same kind of broth; the antagonist and test organism were inoculated into the two lots of broth, and the effect of each upon the growth of the other was de- termined. Frost (303) emphasized, however, that, although theo- retically this is an ideal method, it is open to criticism since motile bacteria are usually able to grow through the filter after a certain lapse of time. Collodion sac method. Collodion sacs, prepared by means of test tubes from which the bottoms have been cut out, are partly filled with METHODS OF TESTING ANTAGONISTIC ACTION 61 broth and placed in a flask containing the same kind of broth. The test organism is inoculated into the medium inside the sac, and the antagonist into the flask (303). Solid Media Solid media have also been used extensively for testing the action of antagonists. These media offer certain advantages over liquid media. The following methods are most commonly used: Simultaneous inoculation of antagonist and test organism. This method, introduced by Garre (315) in 1887, consists in streaking the an- tagonist and the test organism on the surface of a solidified agar or gelatin medium. The streaks are alternate and may be parallel, radi- ating from a common center, or intersecting at right angles (Fig- ure 4). If the active substance produced by the antagonist does not diffuse for any considerable distance into the medium, the method is not satisfactory. Frost (303) modified this method by inoculating the whole medium with the test organism and, when the medium had hardened, streaking the antagonist across the surface. The first of these came to be known as the anaxogramic method ; the second is often spoken of as the implantation method. The spotting of the two organisms on the plate is illustrated in Figure 5. Successive inoculation of the test organism, after the antagonist has al- ready made some growth, so as to enable the active substance to dif- fuse. Double plate methods (303). A Petri dish is divided into two parts by means of a small glass tube or rod. After sterilization, one tube of molten agar is heavily inoculated with the antagonist and poured into one half of the plate. When the agar has hardened, another tube of sterile agar is poured into the other half of the plate. Both sides are then streaked with the test organism, each side being equally inocu- lated by separate streaking. This can be done by using a loop bent at nearly right angles; the charged loop is moved from the circumfer- ence toward the glass rod. The loop is then sterilized, recharged with the test culture, and the streak continued on the other side of the plate. The inoculation with the test organism may be made soon ( after the plate is poured, or the antagonist may be given an opportu- nity to develop before the test organism is streaked thus making the 62 ISOLATION AND CULTIVATION OF ANTAGONISTS Uelminthosforium (A and B) in- Pestalozzia (A) inhibited by one hibited by Fusarium (C) species of Penic'tlUum (C) but not by another (B) Helminthos for turn (A) inhibited Helminthosfortum (A and B) in- by a bacterium (C) hibited by a white yeast (C) Figure 5. Inhibition of fungus development by antagonists. From Porter (729)- METHODS OF TESTING ANTAGONISTIC ACTION 63 antagonistic effect more striking. This method has also been used (258) for testing the antibiotic properties of fungus cultures. Mixed culture inoculation. The cultures of the antagonist and the test or- ganism are mixed and inoculated upon the surface of the solidified agar or before the molten agar has been added to the plate. The colo- nies of the antagonist will be surrounded by clear sterile zones, free from any growth of the test organism. Spot inoculation of the antagonist upon an actively growing culture of a bacterium or fungus on an agar plate. This method is particularly convenient for detecting antagonists that possess lytic properties. A layer of molten sterile agar is used to cover the surface of an antagonist that has made some growth in a plate, and the surface of the agar layer is then inoculated with the test organism. The active substance produced by the antagonist will diffuse through the agar and reduce the growth of the test bacterium (609). Semisolid media are used for testing the action of antagonists upon the motility of bacteria (182). A number of other methods, usually modifications of those outlined above, have been used for testing the ability of fungi to produce anti- biotic substances (724, 1016). Some of these methods, notably the agar diffusion (cup, paper disc, cylinder) test, are used for the quantitative estimation of the concentration of the antibiotic in the medium and for isolation purposes. These methods can indicate the formation not only of growth-inhibiting but also of growth-promoting substances (99). Raper et al. (765) removed plugs of agar of constant dimensions from the fungus cultures being tested and placed them on the surface of plates seeded with S. aureus. The plates were incubated at 37° C, and the amount of penicillin present was estimated by the size of the zones of inhibition. For the purpose of screening many cultures, a modified Czapek's solution agar, i per cent by volume of corn steep liquor {^$^ per cent solids) was used} the solution was adjusted to /jH 7.0, and 2 per cent agar was added. Twenty-milliliter portions were placed in tubes, sterilized, and poured into sterile Petri dishes. The plates were selected to insure that the agar layers were of uniform depth. Single colonies were established by suspending spores of the cul- ture to be tested in melted agar at 45 ° C. The agar was allowed to so- 64 ISOLATION AND CULTIVATION OF ANTAGONISTS lidify and small amounts were placed with an inoculating needle in the centers of the agar plates. The plates were incubated at 24° C. for 6 days J then 4 or 5 plugs were removed radially from the agar, the first being adjacent to the colony margin, and tested as described above (838). Various other methods have been proposed for measuring the rate of production or secretion of antibiotic substances by fungi (726, 963). METHODS OF GROWING ANTAGONISTIC ORGANISMS FOR THE PRODUCTION OF ANTIBIOTIC SUBSTANCES Once the antagonistic action of any organism has been established, the next step is to determine the nature of the substance produced and to measure quantitatively its antibiotic action. Before this can be done, however, the organism must be grown upon suitable media under conditions favorable for the maximum production of the antibiotic substance. The media used for the production of antibiotics can be classified into two groups: synthetic and complex organic media. The first contain a source of carbon, usually glucose, sucrose, or starch (2 to 6 per cent) j a source of nitrogen, usually nitrate or ammonia sulfate (0.2 to 0.6 per cent) ; several salts, namely, K0HPO4 orKHoP04 (o.i to 0.2 percent), MgS04.7H20 (0.05 per cent), KCl (0.05 per cent), and FeS04.7H20 (0.00 1 per cent). Certain supplementary materials such as yeast ex- tract, meat extract, or corn steep, or trace elements such as ZnS04, MnS04, or CUSO4 (i to 2 ppm.) may also be added. The organic media contain a complex form of nitrogen, such as tryptone, peptone, casein digest j either no other source of carbon is used or a carbohydrate is added in the form of lactose, glucose, dextrin, starch, brown sugar, molasses, or similar products as well as several salts similar to those listed above. Some media are supplemented with CaCOo, others are not, depending upon the extent of acidity produced by the organism. The medium may be solid (agar or bran) or liquid, the latter being the more common. Several types of culture vessels are used, depending on the condition of aeration. Since so far as is known all the micro- METHODS OF GROWING ANTAGONISTIC ORGANISMS 65 organisms capable of producing antibiotic substances are aerobic, either shallow layers of medium ( 1.5 to 2 cm. in depth) are placed in station- ary vessels (flasks or trays), or shaken cultures are used. In the case of deep vessels or tanks, the medium is properly stirred and aerated by forced draft with sterilized and filtered air. The optimum temperature required for the growth of the antagonis- tic organisms and the production of antibiotic substances ranges be- tween 20° and 30° C. The length of incubation varies from 2 to 6 days for submerged cultures and from 3 to 20 days for stationary cultures. A knowledge of the preliminary treatment of the inoculum or spore material is essential. For the growth of spore-forming bacteria, the use of a pasteurized spore suspension is advisable in order to avoid the vari- able factor due to vegetative cells. Actinomycetes and fungi are grown on agar slants in order to obtain abundant spore material for the inocu- lation of stationary cultures. For submerged cultures, special spore sus- pensions are produced by growing the organisms in shaken cultures. The cultures must be tested carefully in order to establish the opti- mum activity when the culture filtrate is cooled and extraction of active substance is started. Tyrothrkin For the production of tyrothricin, shallow layers of medium are used most frequently. The media contain complex sources of nitrogen, such as tryptone, casein hydrolysate, soybean meal digest, and pressed juice of waste asparagus. Simple substances, such as glutamic acid, aspara- gine, ammonium salt, plucid citric or malic acid, are also effective in presence of 0.2 per cent tryptone. Glucose, mannitol, or glycerol (3 to 5 per cent) can be used as the source of carbon, and calcium, magne- sium, and manganese as required mineral. Maximum yields of more than 2 gm. per liter are obtained in 10 to 16 days' incubation at 35° C. (564). Penicillin ^ For the production of penicillin, the composition of the medium is highly important. At first a simple glucose-nitrate solution known as Czapek-Dox medium was used. It was later found that when yeast ex- 66 ISOLATION AND CULTIVATION OF ANTAGONISTS tract or corn steep liquor was added and brown sugar was used in place of glucose, the growth of the organism and the production of penicillin were greatly facilitated (5, 281). The ratio of C and N sources is sig- nificant. It has been shown, for example, that penicillin is produced in organic media when the ratio sucrose-peptone is less than i .0, and in inorganic media when sugar-NaNOg ratio is i.o or lo.o (591). The following was found (838) to be a suitable medium for maxi- mum production of penicillin : Lactose 40.00 gm. NaNOg 3.00 gm. MgS04.7H20 0.25 gm. KH0PO4 0.50 gm. ZnSOg 0.0 1 gm. Corn steep liquor 90 ml. Distilled water to make 1000 ml. This medium has been variously modified, as by reducing the lactose to 20 mg. per liter and the corn steep to 40 ml. or by using in its place 25 mg. of dried steep liquor solids (764). The need for a specific penicillin-promoting substance, such as might be found in corn steep or in other plant extracts (555)j in order to in- crease appreciably the yield of the antibiotic agent is of particular in- terest. Certain amino acids, namely, arginine, histidine, and glutamic acid, in concentrations of 0.3, 0.3, and 0.4 gm. per liter, respectively, appear to provide a large part of this stimulating effect (1004). A proper balance of the concentration of the ions POf , SOf , NO -3 , and Mg+-^ is also essential. The proportions of the essential three salts in optimum solution were found (733) to be KH2PO4 — O.475, MgS04.7HoO— 0.05, and NaN03— 0.475 j different strains show marked differences in their response to a change in balance of these three salts. This led to the development of different synthetic media, such as the following (735) J the amounts are given on a liter basis: Starch 5.0 gm. Lactose 25.0 gm. Glucose, crude 5.0 gm. METHODS OF GROWING ANTAGONISTIC ORGANISMS 67 Glacial acetic acid 6.0 gm. NaoHPO^ 1.6 gm. K,PO, 2.0 gm. NH4NO3 4.0 gm. (NH,),S03 i.O gm. KNO3 i.O gm. MgSO^.yHoO 0.25 gm. FeS04.7H20 0.2 gm. MnS04.7HoO 0.04 gm. CUSO4.5H..O 0.005 gin- Cr (as KoCroO,) 3 Mg Phenylacetic acid and its derivatives have a marked effect upon peni- cillin yields 5 frequently amide derivatives are just as effective as the corresponding acids (887). With the introduction of the submerged process for the production of penicillin, it became necessary to find a simple means of obtaining large numbers of spores. For this purpose, a medium high in calcium salt appears to be essential (282). Such a medium is as follows: Sucrose or brown sugar 20.0 gm. NaNOo 6.0 gm. KH2PO4 1.6 gm. MgS04.7HoO 0.5 gm. CaClo 25.0 gm. Tap or distilled water to make lOOO ml. The culture is grown for 4 to 6 days with continuous aeration and agi- tation. Various other methods are used for spore production, for the purpose of inoculating large batches of medium. For surface growth, dry spores are mixed with a floating and spreading agent, such as whole wheat flour. Other media, such as bran (762), have been utilized to a limited extent for the production of penicillin. However, the submerged proc- ess, accompanied by agitation and aeration, using one of the above liquid media has now come into general use for large-scale production of penicillin. 68 ISOLATION AND CULTIVATION OF ANTAGONISTS Streftothrkin and Streftomycin For the production of streptothricin, a tryptone medium with starch or glucose is used. A typical medium is given here: Glucose or starch 10.00 gm. Tryptone 5.00 gm. K2HPO4 2.00 gm. NaCl 2.00 gm. FeSO^ 0.0 1 gm. Tap water to make 1000 ml. For stationary cultures, 0.25 per cent agar may be added. For streptomycin, certain specific organic precursors are required. The precursors are present in meat extract, in corn steep, and in the cells of certain microorganisms such as yeasts and actinomycetes. A typi- cal medium, on a liter basis, consists of: Glucose 1 0.0 gm. Peptone 5.0 gm. Meat extract 5.0 gm. NaCl 5.0 gm. Final fn 6.S to 7.0 Tap water to make 1000 ml. For spore production, a simple synthetic medium may be used, such as glucose-asparagine agar, consisting of: Glucose lO.O gm. Asparagine 0.5 gm. K2HPO4 0.5 gm. Agar 15.0 gm. Distilled water to make 1000 ml. A synthetic medium has also been suggested (905a) for streptomy- cin production, consisting of: Glucose 7.4 gm. Ammonium lactate 5.4 gm. KH,P04 2.38 gm. METHODS OF MEASURING ANTIBIOTIC ACTIVITY 69 K0HPO4 S-65 gm. MgS04.7HoO 0.98 gm. ZnS04.7H20 o.oi 1 5 gm. FeS04.7HoO o.oiii gm. CUSO4.5H.O 0.0064 gm. MnClo^HoO 0.0079 gm. Distilled water to make lOOO ml. fU 6.95 METHODS OF MEASURING THE ACTIVITY OF ANTIBIOTIC SUBSTANCES It has long been recognized that the evaluation of bacteriostatic and bactericidal substances is controlled to a considerable extent by the methods employed. These methods are based upon the following fac- tors: (a) proper selection of the test organism, (b) composition of the medium used for testing activity, (c) time of action, (d) conditions of carrying out the test, and (e) nature of the active substance. The results obtained in a comparison of substances containing the same active prin- ciple may not be very reliable when different agents are compared, since these vary greatly in their specific action upon different bacteria. This is especially true of antibiotics. In most of the work on chemical disinfectants, which are primarily bactericidal agents, the death rate of the viable cells has been used as a basis for evaluation. Different substances have been compared with a standard, ordinarily phenol. Since antibiotic and chemotherapeutic substances are primarily bacteriostatic in action, the inhibition of the growth and multiplication of the test organism is commonly used as a basis for their evaluation. In any attempt to select a single standard method for measuring quantitatively the activity or potency of an antibiotic substance, it is es- sential to recognize several pertinent facts, which may be briefly sum- marized as follows: (Antibiotic (antibacterial, antimicrobial) substances are primarily bac- teriostatic (or fungistatic) in their action; some substances are also markedly bactericidal (or fungicidal). 70 ANTIBIOTIC ACTION OF ANTAGONISTS Antibiotic substances are selective in their action ; they are able to inhibit the growth of some bacteria in very low concentrations, whereas much larger amounts are required to affect other bacteria and some organisms may not be inhibited at all by the particular substance even in very high concentrations. Conditions for the bacteriostatic activity of different antibiotic substances vary greatly. Some substances are not active at all, or their activity is greatly reduced in some media because of the neutralizing effect of certain constituents of the media, such as peptone, />-amino-benzoic acid, or glucose. Other agents require the presence in the medium of specific constituents for their activity to become effective. The activ- ity of some is reduced at an acid reaction, whereas that of others is not affected. The mechanism of the action of different antibiotic agents is different. Some agents interfere with bacterial cell division, others with bac- terial respiration, and still others with utilization by the bacteria of essential metabolites. Many antagonistic organisms produce more than one antibiotic substance. Ps. aeruginosa produces pyocynnase and pyocyanin; B. brevisy grami- cidin and tyrocidine; P. notatuni, penicillin and notatin; A. fumiga- tuSy spinulosin, fumigatin, fumigacin, and gliotoxin; A. flavus, asper- gillic acid and penicillin. The culture filtrate of an antagonistic or- ganism often differs, therefore, in its activity from that of the isolated active substance. The course of production of antibiotic substances by two typical antago- nistic organisms is illustrated in Figures 6 and 7. In view of the bacteriostatic nature of antibiotic substances, few of the methods commonly used for testing the efficiency of antiseptics and germicides can be employed. This is particularly true of the "phenol coefficient test," which measures the germicidal action of phenol upon E. tyfhosa. The limitations of this method, based on the bactericidal ac- tion of a single substance on a single organism, even as applied to chemi- cal antiseptics have long been recognized (810). A number of methods have been developed for determining the ac- tivity of antibiotic substances. They vary greatly, each having its limita- tions and advantages. Because of lack of uniformity in the methods, the results obtained by one are not always comparable with those obtained sy3xnn-im ooi aad •s■l'Nva^Di-nI^N ni NIDiaHiO^AX aNV NBOOailN -)Viai31DV8 U1 O lO o CM (M - - ^ Oq 1 1 1 1 1 r~ ^ /° X \ / / \/ / s ^ V / A / /\\ / / \ Y O m / v^ (0 8 O °\ ^ ''>;- \ ^ ^^ ^^"^ '^^vX \ ^v^ \\ z o 2 (0 1- < D O z ^>\>.^ - o fVJ ^^\ \ *N \\N 1 1 I 1 I ^ o o o o o o o o o o o o «o ^ n c\j — syainniitN ooi a3d swvyoiTiiKN ni aakNnsNOD qdv oiiNvxmo 1 1 1 1 1 o o o o o c 3 o o o o o O > 5 75 \ O 5 50 2 < - / / \ 25 / 1 1 1 1 1 30 36 42 48 54 60 66 72 INCUBATION PERIOD IN HOURS 84 96 Figure 7. Production of gliotoxin by Trlchoderma. From Weindling (989). by another. The most important methods at present in use are briefly summarized in the following pages (583). A gar Streak-Dilution Method If an unknown antibiotic substance is tested, it is essential to employ several test organisms in order to throw light upon the selective activity of the substance on different bacteria, and thus to determine the anti- biotic spectrum characteristic of each substance. Nutrient agar media have usually been employed. Sterility is not absolutely essential for this method, although it is desirable. The unknown substance is diluted to various concentrations (i, 0.3, o.i, etc.; or i, 0.5, 0.25, etc.) ; these dilutions are added and thoroughly mixed with definite volumes (10 ml.) of sterile agar medium, melted and cooled to 42° to 45° C. The agar is allowed to solidify, and is streaked with three or four test bacteria, among the most common of which are E. coli, B. mycoides, B. suhtilisy S. aureus, M. lysodeikticus, S. lutea, M. fhlei, as well as various other bacteria and fungi. The age of the cultures (16 to METHODS OF MEASURING ANTIBIOTIC ACTIVITY 73 24 hours) is important. The plates are incubated at 28° or 37° C. for 16 to 24 hours, and readings are made. The highest dilution at which the test organism fails to grow is taken as the end point. Activity is ex- pressed in units, using the ratio between the volume of the medium and the end point of growth or the dilution at which growth is inhibited (964). The bacteriostatic and fungistatic activity of several antibiotic sub- stances is shown in Table 7. Serial Dilution Method Once a substance is characterized as regards its selective action upon specific bacteria, its activity or concentration can be measured more ac- curately by the liquid dilution or titration method. One test organism is selected, usually a strain of S. aureus. Different strains may vary in their action. Definite volumes of the test medium are placed in test tubes and sterilized (sterility is essential in this method), and various dilutions of the active substance are added. The dilutions can range in order of 3 :i, 2:1, or even narrower, namely in series of 1.2:1, 1.5:1, etc. The tubes are inoculated with the test organism and incubated for 16 to 24 hours. In some cases the medium is inoculated before it is distributed into the tubes. The highest dilution of the antibiotic giving complete inhibi- tion of growth, as expressed by a lack of turbidity of medium, is taken as the end point. Activity is expressed in units as above. The dilution method has several disadvantages: every assay takes much time 5 during chemical fractionation, the substance may become contaminated with bacteria not sensitive to the active substances 5 only one organism can be used in a single series of tests. One modification of the method has been adapted for measuring the activity of penicillin. Several dilutions of the active agent are prepared and 0.5 ml. portions added to 4.5 cm. quantities of liquid medium in test tubes. These are inoculated with a standard drop (0.04 ml.) of a 24-hour culture of the test organisms. Complete or partial inhibition is shown by the absence of turbidity after 24 hours of incubation at 37° C. Dilutions higher than those required for complete or partial inhibition gave, after 24 hours of incubation, only a retarding effect (1,5)5 ^ "^i" croscopic examination (311) indicated defective fission of the bacteria, t ^ o o o o o o O 'ih O O i^ o o ^ o ° V °- A o Q o o o LO o o o o o o ^ O o o q oo q V ^ o" ""l- r<^ '^ o " Ti- 8 V t<^ ^ 8 ir> vO O o O o o i^ o o o O o o Tj- so o q q q o 6 V V o V o V o o Q o o ir> o o o o o o 'd- O o o o o o V °- o ^ U-1 N o ^ o o V N N o V o o o o q q o <-<^ O r4 O vO o o .s .s •G .^ ^ In B -5 5 i o 2 'G o .2 o o Oh O, dJ (U < u ^ 3 1-. )-l O o g < O U fe O uo in u:;^ METHODS OF MEASURING ANTIBIOTIC ACTIVITY 75 even though the macroscopic appearance of the culture did not show any inhibition. Pneumococci and S. viridans show marked strain differences by this method. In one experiment with Salmonella tyfhiy partial in- hibition was obtained in a dilution of i : 1 0,000 j however, elongation of the cells was detected in a dilution of i : 60,000, a concentration which was considered as a therapeutic possibility (Table 8). The other modifications of this method either use different test cul- tures, such as B. sub tilts (285), or have been developed to meet the re- quirements of the clinician when only small amounts of blood or other body fluids are available, in which case a very sensitive strain of S. hemolyticus is used (751). The use of Klebsiella fnemnoniae for as- saying streptothrycin and streptomycin permits the determination of as little as 0.05 Mg/ml., giving somewhat more rapid and more accurate results (191). Agar Diffusion {Cuf, Pafer Disc, Cylinder) Method (5, 173, 283, 285, 390) This method, first employed for measuring antiseptics qualitatively (810), was later developed for quantitative use. A suitable agar me- dium is inoculated with a test organism {S. aureus or B. subtilis), the active agent being placed upon the agar within a groove or in a special small glass cup with an open bottom from which the substance diffuses into the medium. The rate of diffusion of the antibiotic is parallel to its concentration. Potency can be calculated by measuring the zone of in- hibition and comparing it with that of a known standard preparation. Various modifications of this method have recently been introduced (286, 937). This method has the advantage of simplicity and con- venience, since it does not require sterile material and several prepara- tions or duplicates can be tested on the same plate. The method also possesses certain disadvantages, however, since it cannot be used for comparing different substances but is limited to the measurement of activity of only one type of substance; it cannot be used for the study of unknowns until a standard has been established for each j it cannot be used for substances that are not water soluble. Nutrient agar containing 5 gm. NaCl, 3 gm. meat extract, 5 gm. 76 ANTIBIOTIC ACTION OF ANTAGONISTS TABLE 8. BACTERIOSTATIC SPECTRUM OF PENICILLIN DILUTIONS AT WHICH INHIBITORY ORGANISM AFFECTED EFFECTS WERE OBSERVED Complete Partial None N. gofiorrhoeae^ 2,000,ooo > 2,000,000 > 2,000,000 "N . meningitidis 1,000,000 2,000,000 4,000,000 S. aureus 1,000,000 2,000,000 4,000,000 S.fyogenes 1,000,000 2,000,000 4,000,000 B.anthracis 1,000,000 2,000,000 4,000,000 A.bovis 1,000,000 2,000,000 4,000,000 CI. tetani\ . i ,000,000 CI. welchii 1,500,000 CI. sefticum 300,000 1,500,000 7,500,000 CI. oedematiens 300,000 1,500,000 S.viridansX 625,000 3,125,000 Pnetimococcus\ 250,000 500,000 1,000,000 C. difhtheriae (miiis) 125,000 625,000 C. difhtheriae {gravis) 32,000 64,000 128,000 S. gartneri 20,000 40,000 8 0,000 S.tyfhi 10,000 30,000 90,000 Pneu7nococcus\ 9,000 27,000 Anaerobic Streptococcus^ 4,000 8,000 16,000 P. vulgaris 4,000 32,000 6o,000 S. viridans% 4,000 8,000 1 6,000 P.festis 1,000 100,000 500,000 S. iyfhimurium <^ 1,000 8,000 1 6,000 S.faratyfhiB • - iij 1- u 8 ^^^"'^^ ^^^^-""^ 1 2^ 5 O ■^.''^^^'^-^^ s Q z ..---^ ^^ "--^-tjfo. - 0.8 < o ^--'**''^ "'""'--- 1 -« 1 1 1 . 1 1 1 0.4 .05 .10 .15 .20 .25 .30 PENICILLIN IN B PER. MILLILITER .35 .40 Figure 9. Relation between penicillin concentration and inhibition of Stafhylococcus aureus. The penicillin preparation contained 42 Oxford units per milligram, and the incubation period was sixteen hours at 37° C. From Foster (280). pared with the control, or it can be measured by a convenient turbi- dimeter. The results obtained by this method are more nearly quantita- tive than those obtained by other methods, as shown in Figure 9. By proper modifications, the length of time required to obtain a satisfac- tory reading can be reduced to four hours (482, 610), or even to 90 minutes (280, 554). The turbidimetric method has also found application in the stand- ardization of streptomycin. For this purpose, certain noncapsulated cultures of K. fneumoniae are used. Nutrient broth is inoculated from a slant and incubated at 37° C. for 22 to 24 hours. A series of dilutions of streptomycin in nutrient broth is prepared and one-milliliter portions are added in duplicate to tubes containing 9 cc. of broth inoculated with 2 per cent of the culture. The tubes are incubated at 37° C. for 3 hours, 3 drops of formalin are added to stop growth, and turbidity is deter- mined in a photoelectric colorimeter. The turbidity measurements are METHODS OF MEASURING ANTIBIOTIC ACTIVITY 81 plotted against the concentrations of streptomycin and compared with the standard. S fecial Methods Special methods were found to be specific for measuring the action of certain substances. The ability of tyrothricin to hemolyze red blood cells served as the basis for measuring the potency of this substance ( 1 8 1 ) : the tyrothricin content is calculated from the amount of hemoly- sis by the unknown and is read from a standard curve. The inhibition of growth of a (3-hemolytic streptococcus, group A, as measured by hemolysin production has been used for assaying the potency of peni- cillin (747, 1022). Penicillin can also be estimated by its inhibition of nitrite production by 6". aureus cultures (350). The antiluminescent test has been utilized not only for measuring the activity of certain sub- stances but also for determining their possible usefulness. The results of a comparative study of a number of antibiotic substances by this and the dilution method are brought out in Table 9. Other methods suggested for measuring the activity of antibiotic substances are based upon interference with a given physiological func- tion of the test organism such as dehydrogenase activity and respiration (562) upon the prevention of growth of the test organism (pneumo- coccus) in semisolid tissue culture medium (392), or upon the diffu- sion of the antibiotic placed upon the surface of inoculated semisolid in tubes and measurement of the depth of clear zone (27). Although only a single method is usually employed in the concentration and standard- ization of a given antibiotic such as penicillin or streptomycin, it is often advisable to check the results by another method, especially where sev- eral test organisms are used, in order to obtain an antibacterial spectrum that will demonstrate that one is still dealing with the same type of chemical compound. The results obtained by the various methods for determining the relative potency of different antibiotics lend themselves readily to statistical analysis (68, 513, 514). ' Methods have also been developed for detection of chemotherapeu- 82 ANTIBIOTIC ACTION OF ANTAGONISTS TABLE 9. ANTILUMINESCENT AND ANTIBACTERIAL ACTIVITIES OF VARIOUS ANTIBIOTIC SUBSTANCES SMALLEST AMOUNT SHOWING ACTIVITY, IN MICROGRAMS al/ab ratio Antiluminescent test Antibacterial test* Tolu-p-quinone O.I I Gramicidin .002 Tolu-p-quinone .002 Pyocyanase 3 Tyrothricin .008 Pyocyanase .07 Clavacin I II Penicillin II Penicillin I Flavatin .0156 .06 .256 Clavacin I .18 Aspergillic acid 15 Gramidinic acid •23 Sodium clavacinate .18 Gliotoxin 17 AP2it •31 Clavacin II .19 Clavacin II 22 Actinomycin •54 Sulfanilamide Phenol <.56 •5 Pyocyanin 47 Aspergillic acid 2.0 Pyocyanin 1-7 Actinomycin 54 Gliotoxin 2.1 Lauryl sulfate 4.6 Streptothricin 56 Streptothricin 2.8 Aspergillic acid 7-5 Sodium clavacinate 94 Fumigacin 13-0 Gliotoxin 8.0 Flavatin 256 Fumigacin 273 Pyocyanin 27.0 Streptothricin 20.0 Lauryl sulfate 273 Pyocyanase 42.0 Fumigacin 21.0 Phenol 1 1 70 Tolu-p-quinone 55.0 Actinomycin Flavatin 1 00.0 1 000.0 Penicillin I 1650 Lauryl sulfate 59.0 AP2it >i630 Sulfanilamide 3940 Clavacin I 63.0 Gramidinic acid >2i75 Gramicidin >500 Clavacin II 1 13.0 Penicillin I 27,500 Gramidinic acid >5oo Sodium clavacinate 500.0 Tyrothricin >62,5oo Tyrothricin >500 Phenol 2300.0 Gramicidin > 2 50,000 AP2it >5oo Sulfanilamide >7000.0 Penicillin II >325,ooo Penicillin II >50oo From Rake, Jones, and McKee (748). * Streptococcus pyogenes used as test organism. t A tyrothricin-llke preparation. tic substances in tissues and their secretions, as by the use of jfluorescent microscopy, penicillin giving a green fluorescence (403). By utilizing the inactivating effect of penicillinase upon penicillin, it was possible to work out a method for the evaluation of different forms of penicillin in blood serum (130). Several methods are commonly employed for measuring bactericidal METHODS OF TESTING IN VIVO ACTIVITY 83 action of antibiotic substances. A suspension of washed bacterial cells in saline or other suitable solution, or a 5-to-i2-hour-old broth culture of the test organism, is treated with various dilutions or concentrations of the active substance. After incubation at 37° C. for i to 24 hours, the number of living cells is determined. If the active substance has lytic properties or if the test organism undergoes lysis readily, the readings are simplified. If no lysis occurs, the treated bacterial suspension or cul- ture is streaked or plated out. The streaking procedure gives only a relative idea of the extent of bactericidal action. If 50 to 90 per cent killing of cells Is to be taken as a unit of measurement, the culture Is plated on a suitable medium and the actual number of surviving cells are determined. Some of the foregoing methods can also be utilized for measuring the fungistatic and fungicidal properties of antibiotic substances. Pro- tective fungicides may first function as fungistatic agents, others func- tion better either as fungicidal or as fungistatic agents, and still others show either a high or a low for both. The growth of Ceratostomella ulmi was inhibited by actinomycin, clavacin, and hemipyocyanin in con- centration of 1:100,000 (771, 949). METHODS OF TESTING THE IN VIVO ACTIVITIES OF ANTIBIOTIC SUBSTANCES Ordinary pharmacological, bacteriological, and pathological proce- dures are used for testing the toxicity and activity of antibiotic sub- stances In the animal body. In order to determine the amount of an antibiotic required for the treatment of a certain infection, It is essential to know not only the re- sponse of the organism causing the Infection but also the sensitivity of the particular strain Involved. It is also essential to determine the concentration of the antibiotic In the body fluids. A number of methods have been proposed for this purpose, especially for penicillin and strep- tomycin. Most of these represent various modifications of the agar diffusion and serial dilution methods, using a hemolytic streptococcus or some other suitable test organism, such as S. aureus or B. subtilis. In some 84 ANTIBIOTIC ACTION OF ANTAGONISTS cases the serial dilution has been combined with the turbidimetric method. The ability of penicillin to inhibit hemolysin production by streptococci has also been utilized (8ooa). A comparison of the agar diffusion, turbidimetric, and serial dilution methods led to the conclu- sion that the last is the most suitable for routine clinical work (508). Penicillin-containing material may also be spread over a given area of a nutrient agar surface, allowing time for absorption of the liquid, and streaking the surface with standard strains of S. aureus of known sensi- tivity j on comparing with standard penicillin preparation, it is possible to determine both the concentration of penicillin and the degree of sen- sitivity to penicillin of the infecting agent (152). Fleming (266) proposed a method using a hemolytic streptococcus as test organism and blood (preferably group O) from which leucocytes have been removed or inactivated and treated with a clot-inhibiting substance as a medium. Hemolysis of blood is measured either in a slide cell or in a capillary tube. The penicillin concentration in the blood is estimated by the serial dilution method. This method has been vari- ously modified for clinical assays of penicillin. A convenient method for measuring the concentration of streptomy- cin in body fluids is to use the agar diffusion method with an alkaline medium, low in salt, and a carefully selected test organism (582, 879). CHAPTER 5 BACTERIA AS ANTAGONISTS Following the work of Pasteur in 1877 on the antagonistic ef- fects of bacteria against the anthrax organism, considerable attention has been centered upon bacteria as agents possessing antibacterial prop- erties. A systematic study of this phenomenon was first made by Babes in 1885 (155), who demonstrated that this antibacterial action is due to the formation of definite chemical substances. Garre (315) first in- troduced, in 1887, suitable methods, such as the streak test, for demon- strating the antagonistic effect of one organism upon another. The first antibiotic substance, pyocyanase, was isolated in 1 899 by Emmerich and Low (235). Freudenreich (298) found in 1888 that when certain bacteria were grown in a liquid medium, the filtrate obtained by passing the culture through a porcelain candle supported the growth of the typhoid or- ganism not at all or only very feebly. Garre (315) observed that Ps. futida inhibited the growth of S. aureus, E. tyfhosa, and Bacillus muco- sus-cafsulatus but not of B. ant hr acts and other bacteria. It was soon reported {S^'i)j however, that B. anthrach was also killed by the Pseudomonas antagonist, whereas the growth of S. aureus and V . comma was only retarded j no effect at all was exerted upon E, typhosa or E. coli. In consequence, the antagonist was claimed to be active against B. anthrach but not against other bacteria. Olitsky (691) con- cluded that Ps. -fluorescens inhibited the growth not only of E. typhosa but also of B. anthracis, V. comma, S. marcescens, and S. aureus. These and other apparently contradictory results were undoubtedly due to differences in the specific nature of the strains of the organisms used by the various investigators and to different methods of cultivation. The presence of Ps. fluorescens in sewage was found (551 ) to reduce greatly the period of survival of the typhoid organism. The latter did not develop even in gelatin upon which Ps. fluorescens had previously grown, and it could not be detected in sterile sewage in which the an- tagonist was present for seven days. According to Frost (303), E. ty- 86 BACTERIA AS ANTAGONISTS fhosa can be antagonized by a number of different soil bacteria, of which Ps. fluorescens exhibits the strongest effect. He observed that al- though P. vulgaris acted more rapidly, the active substance did not dif- fuse to so great a distance in the medium, thus pointing to a different in- hibition mechanism. Mixed cultures showed greater activity than pure cultures, either because the latter lost their antibiotic property when grown for a long time on artificial media or because mixed cultures com- prise two or more species with a greater combined action. The antago- nistic substances produced by these bacteria were active at 37° C, where- as at ice-chest temperature the action was delayed so that the pathogen had an opportunity to develop. This was believed to offer a possible ex- planation for the fact that when water supplies become contaminated in cold weather, their power of producing infection is retained for a longer time than when the contamination takes place in warm weather. Frost concluded that the phenomenon of antagonism results in checking the growth of E. ty fhosa as well as in killing the pathogen. Evidence that antagonistic substances exist in an active state in the soil or in water appeared to be lacking j rather, the results suggested that formation of such substances depends on the actual development of specific antagonistic organisms. Changes in environment, such as tem- perature, oxygen supply and reaction of the medium, and nature and concentration of nutrients, were believed to have little or no influence on the production of the antibiotic substances j these were produced under conditions favoring growth of the antagonists. The activity of the influenza organism was found ( 1 02 5 ) to be largely dependent on the presence of accompanying bacteria. Some of these, especially micrococci, are favorable to the growth of this organism whereas others, such as Ps. aeruginosa and B. subtilis, are injurious. According to Lewis {S^^)-, luxuriant growth of Ps. fluorescens in manured soil and in protein solution containing B. cereus is due to an- tagonistic action of the former organism against the latter. Ps. fuo- rescens also inhibits the growth of B. anthracis, B. megatherium^ V. comma, Chrom^obacterium, violaceum^, and Rhodococcus. Other species of the genera Bacillus, Eberthella, Sarcina, Neisseria, and Phytomonas are somewhat more resistant to the action of Ps. fluorescens. Salmonella species are less sensitive, whereas E. coli, A. aerogenes, and S. marces- SPORE-FORMING BACTERIA 87 cens are highly resistant. Ps. fluorescens produces a thermostable sub- stance which is toxic to all bacteria except the green fluorescent forms and which is active against actinomycetes but not against fungi. This substance is water-soluble and dialyzable through collodion and other membranes. In addition to the aforementioned bacteria, numerous other groups were found to contain strains which had strong antagonistic properties toward bacteria as well as fungi. Some of the antagonists were highly specific, such as those acting upon the various types of pneumococcij others were less selective, such as certain soil bacteria that can bring about the lysis of living staphylococci and inhibit the growth of various gram-positive and gram-negative bacteria. S. m,arcescens was antagonis- tic to various spore-forming bacteria. These, in turn, were antagonistic to sarcinae, bringing about their lysis, to V. comma, and to various other bacteria. It was further found that the antagonists modified the physiology of the antagonized organism. When two bacteria were planted, for example, in the same medium, metabolic products were formed that were not produced in the culture of either organism alone, whereas certain decomposition processes were either hastened or re- tarded (674). The various antagonistic bacteria can be divided into several groups, on the basis of their morphological and physiological properties. SPORE-FORMING BACTERIA Many aerobic spore-forming bacteria possessing antagonistic proper- ties have been isolated from a great variety of sources, such as soil, sew- age, manure, and cheese. Among these, B. subiilis, B. mycoides, B. mesenterkus, and B. brevis occupy a prominent place, as shown in Table 10. Duclaux (212) isolated antagonistic spore- forming bacteria from cantal cheese, the organisms having been designated as Tyrothrix. Nicolle (680) obtained from the dust in Constantinople a strain of B. subtilis that had decided bacteriolytic properties against members of the pneumococcus group and various other bacteria such as the typhoid, an- thrax, and Shiga organisms. E. coU and V. comma were most readily 88 BACTERIA AS ANTAGONISTS TABLE 10. SPORE-FORMING BACTERIA ANTAGONISTIC TO OTHER BACTERIA ANTAGONIST ORGANISM AFFECTED KNOWN PROPERTY REFERENCES B. ant kr acts Anthrax, typhoid, and lactic acid bacteria 298, 819 B. brevis Gram-positive bacteria Produces tyrothricin 201, 202, 208 B. mcsentericus Many bacteria Bacteriolytic 419 B. mesentericus Diphtheria bacteria Bactericidal 31.984 B. mesentericus C. difhtheriae Substance thermola- 738 vulgatus bile, nonfilterable B. my c aides 7 to 20 species of bacteria Lytic 664 B. mycoides, Most pathogens and 292 var. cytolyticus many nonpathogens B. subtilis Various bacteria Bacteriolytic 680 B. subtilis Various bacteria, espe- cially certain plant pathogens Produces subtilin 453,460 B. subtilis M. tuberculosis, E. tyfhosa, etc. 927 B. subtilis M. tuberculosis Thermostable sub- 693, 816 and other bacteria stance produced B. subtilis- Mostly living gram- Lytic 806, 808 mescntericus positive bacteria and dead gram-negative bacteria B. thermofhilus 5. lutea Suppresses growth 864 acted upon, staphylococci were less affected, and B. suifestijer least. The filtrate of the organism grown in peptone broth had strong anti- biotic properties J it liquefied gelatin and hemolyzed red blood cor- puscles. When various bacteria cultivated on a solid medium were sus- pended in physiological salt solution and seeded with the antagonist, the latter developed abundantly and the bacterial suspensions became SPORE-FORMING BACTERIA 89 clarified. The lysed solutions of pneumococcus prepared by the use of the filtrate of B. subtilis could be used for purposes of vaccination. In this connection, Nicolle spoke of the work of Metchnikoff who had proved, in 1897, that organisms belonging to the B. subtilis group are capable of destroying various bacterial toxins. Rosenthal (806) isolated, from soil and from fecal matter, facultative thermophilic antagonistic bacteria belonging to the B. mesentericus group capable of dissolving both living and dead bacteria. The simul- taneous growth of the antagonist with V. comma and other bacteria brought about the clarification of the culture of the latter in about 5 or 6 days. These bacteriolytic organisms were designated as "lysobacteria." It was recognized that the action of antagonists is different from that of phage in several respects: (a) the filtrate of the antagonist is active against other bacteria j (b) both living and dead cultures of bacteria are dissolved J (c) antagonistic action is not so specific as that of phage j (d) races of E. coli resistant to phage are dissolved by the filtrate of the an- tagonist. The active substance was believed to be of the nature of an enzyme. Friedlander's bacillus was not acted upon, possibly because of the formation of a pellicle by the bacillus. The active substance was formed in 4 to 5 days but increased in activity after 2 to 3 weeks. It was essential that a surface pellicle of the organism be maintained. Sub- merged growth was less favorable. Fresh filtrates had the greatest ac- tivity, the property being lost after storage for 3 months. The substance was thermolabile, activity being destroyed at 70° C. The filtrate of an organism dissolved by the action of the antagonist proved to be as ac- tive as the filtrate of the culture of the antagonist. It acted injuriously upon intestinal bacteria not only in vitro but also in vivo. Much and associates (664) isolated several strains of B. •mycoides that possessed strong antagonistic properties. The active strains were said to be found only rarely in nature. They gave a mesentericus-like growth, producing a pellicle and no turbidity in bouillon. One strain was able to lyse 20 species of bacteria, another acted upon 18, a third on 12, and a fourth on only 7. Marked differences were shown to exist in the degree of antagonistic activity of the different strains. P. vulgaris, E. tyfhosa, and V. comma were lysed in 24-hour bouillon cultures as a result of adding pieces of agar containing colonies of the 90 BACTERIA AS ANTAGONISTS antagonist. A lytic effect was also exerted upon staphylococci (824) and gram-negative bacteria (504, 505). The substance was precipitated by 10 per cent tungstic acid and lead acetate and was thermostable. Much and Sartorius (664) came to the conclusion that B. mycoides Flugge comprises two groups of organisms. One produces branching colonies on agar and forms no pellicle in meat broth, the flaky growth dropping to the bottom and the medium remaining more or less clear. The second group forms flat surface growth similar to that of B. mes- enterkus on agar and a pellicle on the surface of liquid media. Many of the pellicle-forming strains have the capacity, in varying degrees, of dissolving various cultures of bacteria. This is not due to their proteo- lytic activity, since members of the first group may be more actively proteolytic. The culture filtrate of the antagonist dissolves the bacteria but does not destroy their antigenic properties. The lytic substance, designated as Much-lysin, was said to have a double effect: one, bound to the living cells of the organism, had nothing to do with phage, and the other, found in the bacteria-free filtrate, had an apparent similarity to phage but was distinct from it. The idea that in the case of bacterial antagonists one is dealing with specific strains rather than with distinct species was further emphasized by Franke and Ismet (292). Various strains of B. mycoides, desig- nated as cytolitkusy were shown to be able to lyse many pathogenic and nonpathogenic bacteria but not their own cells j the same action was exerted by the culture filtrate (Table 11). The lytic action of strains of B. subtilis upon different bacteria, including M. tuberculosis (927), pneumococci, typhoid, diphtheria, and other organisms, has also been definitely established. Pringsheim (738) isolated a strain of B. mesentericus-vulgatus that had a decided inhibiting effect upon a variety of bacteria, particularly Corynebacterium difhtheriae. On agar plates the antagonist produced a circular zone of inhibition, just beyond which was a ring of larger colonies, indicating a stimulating effect. It was suggested that the an- tagonist produced a toxin that was stimulating in small doses and in- jurious in larger concentrations. The active substance was thermolabile and nonfilterable. The antagonistic properties appeared to be inherent in the particular strain of an organism and were not increased by serial SPORE-FORMING BACTERIA 91 passage. The action of filtrates of B. mesenterkus against diphtheria organisms was considered (984) as highly specific. Other strains of this organism were reported to be active against Pasteurella festis (244). Living gram-positive bacteria were found (806) to be more susceptible than gram-negative organisms to the antagonistic action of spore-form- ing aerobes J in the case of dead organisms, the reverse was true. Plates were heavily seeded with the test bacteria and the centers of the plates TABLE I. LYSIS OF PATHOGENIC BACTERIA BY VARIOUS STRAINS OF A SPORE-FORMING ANTAGONIST (cYTOLYTICUs) ORGANISM LYSED E. tyfhosa Paratyphoid A Paratyphoid B Shigella Y bacillus E. coli C. diphtherias Ps. pyocyaneus S. aureus S. alius S. citreus S. viridis S. /laemolyticus S. mucosus P. vulgaris (Weil-Felix) Pneumococcus STRAIN NUMBER OF CYTOLYTICUS II III VI VII VIII IV o +-H- O o + o + o From Franke and Ismet (292). O no clearing. + trace but no true clearing. ++ clearing, slight sediment. +++ clearing without sediment. inoculated with the antagonist. Inhibition of growth and lysis were used as measures of antagonistic action. Hettche and Weber (419) isolated 41 strains of B. mesenterkus from 25 samples of soil. These were streaked on blood agar, and the diphtheria organism was used for testing their effect. In 18 strains the antagonistic action was detected in 24 hours j there was no parallelism 92 BACTERIA AS ANTAGONISTS between inhibition and hemolysis. Of the 1 8 active strains, 1 1 lost the property after two transfers and 2 were exceedingly active. More recently, beginning with the work of Dubos, considerable at- tention has been devoted to spore-forming bacteria, resulting in the iso- lation of a number of substances or preparations that have been desig- nated as tyrothricin, gramicidin S, subtilin, bacitracin, bacillin, sim- plexin, subtilysin, and endo-subtilysin (815a). These substances are largely active against gram-positive bacteria j however, some also aifect gram-negative bacteria and fungi. Dubos (201) obtained from a soil enriched with various living bac- teria a gram-negative, spore-bearing bacillus (B. brevis) that had a marked lytic effect against gram-positive bacteria, including staphy- lococci and pneumococci. The antagonist was grown for 3 to 4 days in shallow layers of peptone media at 37° C. The bacterial cells were re- moved by centrifuging, and the filtrate was acidified, giving a precipi- tate from which a highly active substance (tyrothricin) was isolated. On crystallization, two preparations were obtained, namely gramicidin and tyrocidine, these making up only a fraction of the tyrothricin com- plex. Natural substrates, such as soil, sewage, manure, and cheese, were found (209) to contain various spore-forming bacteria that have marked antagonistic properties against various gram-positive and gram- negative bacteria. Hoogerheide (442) obtained from the soil an aero- bic, spore-forming bacterium that produced a highly active bactericidal substance J it also prevented the formation of capsules by Friedlander's bacterium. This substance appeared to be similar to gramicidin. Grami- cidin S is, however, more like tyrocidine. Further studies definitely established that strains of spore-forming bacteria possessing antagonistic properties are widely distributed in the soil and possess certain physiological characteristics that differentiate them from the inactive strains. This is brought out in Table 12. The production of the antibiotic is a function of the growth of the bacterial cell. The yield of the antibiotic is influenced by the composition of the medium j the substance is bound to a protein, the bond between the two being destroyed by trypsin (523). B. sub tills has been reported by many investigators to exert an an- SPORE-FORMING BACTERIA 93 tagonistic effect upon many pathogenic bacteria, bringing about their complete lysis. The time required for such lysis was 48 hours for gram- positive cocci, 5 days for the typhoid and paratyphoid organisms, and 8 to 12 days for E. coli and M. tuberculosis (693). The action of B. TABLE 12. BIOCHEMICAL CHARACTERISTICS OF ACTIVE AND INACTIVE STRAINS OF SPORE-FORMING SOIL BACTERIA LIQUE- HY- pro- FAC- DROLY- STRAIN ACID PRODUCTION FROM duction TION OF SIS OF GRAM Dextrose Lactose Sucrose OF HoS GELATIN STARCH STAIN Active Strains A-2 - + + - - A-5 - + + - - A-io - + + - - A-21 - + + - - A-23 - + + - - A.27 + + - - - A-34 - + + - - Inactive Strains A-15 + - - + + A-31 - + + - + + + A-32 + + + - + + + From Stokes and Woodward (885). — reaction becoming alkaline. + acid produced. subtilis upon various bacteria is also growth-inhibiting. This property is due to the formation of one or more antibiotics which have been de- scribed in the literature under several different names. Cultures of B. subtilis found (453) to have a high activity against plant pathogenic bacteria yield an antibiotic that was designated (460) subtilin. Other preparations designated by the same name (759) have a strong bacteriostatic, bactericidal, and lytic effect upon a variety of bac- teria, including B. anthracis, C. di-phtheriae, and Sh. dysenteriae; the activity of the culture filtrate was about 4 to 1 6 units and there was a marked parallelism between the antibacterial properties of the filtrate and its proteolytic action. A water-soluble, nontoxic, relatively heat-stable compound was iso- 94 BACTERIA AS ANTAGONISTS lated from other strains of B. subtiUs and named bacitracin (469). Still another strain of B. sub tills isolated from soil enriched with M. tuber- culosis yielded an antibiotic designated bacillin (284). This substance is produced in manganese-containing media and is mostly found in the cell-free filtrate of the culture. It is adsorbed on norite and eluted with 90 per cent ethyl alcohol, concentrated in vacuo and taken up in water. It is active against both gram-positive and gram-negative bacteria. Blood and certain other complex organic materials reduce or destroy its activity. This is due to the presence of a substance which was desig- nated antibacillin and which was found to be a peptide (1030). B. Ucheniformisy related to B. subtilis, was found (107) to produce an effect against M. tuberculosis. The active substance was present in the cells of the organism when grown on a synthetic medium. The cul- ture was acidified to f¥L 2.5 and treated with 3 volumes of 95 per cent ethanol. The coagulum was autoclaved and extracted on boiling with 0.5 volume of 0.4 per cent acetic acid for 45 minutes. The substance had an activity against M. fhlei and S. aureus in i : 80,000 dilution, but not against E. coU. M. tuberculosis hominis was inhibited in i : 20,000 dilution. Preparations of greater purity had an activity of i : 80,000/ gm. The preparation, which was not very toxic to mice, was considered as a mixture of several substances. Various other antibiotics have been reported for aerobic spore-form- ing bacteria. Some of these substances are active against both gram- positive and gram-negative bacteria. This is true, for example, of colis- tatin (323a). Spore-forming bacteria are also able to produce antibiotics antago- nistic to fungi. B. simplex was found (154) to be antagonistic to Rhizoctonia solaniy an important plant pathogen. It produced a thermostable agent that inhibited the growth and even caused the death of the fungus. When the active substance was added to the soil it controlled to some extent seed decay and damping-off disease of cucumbers and peas. It was also active against bacteria (491). It is ad- sorbed on norite and eluted with methyl alcohol 5 the latter is evapo- rated in vacuo and the residue is taken up in water. This preparation was designated as simplexin (287). B. mesentericus produced on artificial media an antibiotic that sup- NONSPORE-FORMING BACTERIA 95 pressed the growth of H elminthosforium sativum. It increased sporu- lation of the fungus, inhibited or retarded spore germination, caused abnormal hyphal development, and induced mutations in certain strains of the fungus. The substance was thermostable and diffusible. It passed through a Berkfeld filter, was absorbed by infusorial earth, withstood freezing and desiccation, and did not deteriorate readily. It was destroyed by alkalies but not by acids. It was inactivated or de- stroyed, however, by certain fungi and bacteria (142). Various other spore-forming bacteria were found capable of inhibit- ing the growth of bacteria, fungi, and other lower forms of life. In many instances, only little is known of the nature of the active agent in- volved. It is sufficient to illustrate this by an observation that B. h'lru- denses, growing abundantly in the digestive fluids of leeches and con- sidered as a symbiont of these animals, exerted a marked inhibitory effect upon the growth of various bacteria and fungi (845). Antagonistic relations among entomogenous bacteria have been demonstrated for the foul brood of the honeybee (441). This inter- action between B. fopilliae and B. lentimorbus was believed to explain the mutually exclusive development of the two types of milky disease in Japanese beetle groups. NONSPORE-FORMING BACTERIA: PS. AERUGINOSA, Among the nonspore-forming bacteria, those belonging to the fluo- rescent, green-pigment and red-pigment producing groups have prob- ably received the greatest attention as antagonists. Bouchard (78) was the first to report, in 1888, that the pyocyaneus organism {Ps. aerugi- nosa) was antagonistic to the anthrax bacillus. It was soon found (131, 298) that when grown on artificial media, this organism affected bac- teria, including E. tyfhosa, Pfeiferella mallei, V. comma, and Bac- terium tyrogenes. The growth of staphylococci, micrococci, diplococci, and spore-forming rods was also reduced. The antagonist inhibited its own growth as well. These early observations were amply substantiated (Table 13). Ps. aeruginosa was shown to be active against E. coli, M. tuberculosis, and 96 BACTERIA AS ANTAGONISTS a variety of other bacteria. The addition of top minnows (Gambusia) to water polluted with E. colt caused the disappearance of the bacteria j this was shown to be due to the inhibiting effect of the pyocyaneus or- ganism present in the intestinal flora of Gambusia. The presence of this antagonist in water renders the colon index of the water an unreliable guide to pollution (388). When a mixture of the antagonist and the colon organism was incubated, the former tended to outgrow the latter after 24 hours. Even after sterilization, media in which Ps. aeruginosa had grown depressed the growth of other microorganisms including TABLE 13. NONSPORE-FORMING BACTERIA AS ANTAGONISTS TO BACTERIA ANTAGONIST ORGANISMS AFFECTED Ps. aeruginosa B. anthracis, E. typhosa, V. comma, etc. Ps. aeruginosa Gram-negative bacteria, M. tuberculosis, and yeasts Ps. fiuorescens E. coli, S. marcescens, C. difhtheriae, B. anthracis, etc. Ps. fiuorescens Actlnomycetes S. marc esc ens CI. chauvoei, B. anthracis. staphylococci, micrococci S. marcescens Gram-positive but not gram- negative bacteria E. colt Typhoid, paratyphoid, diph- theria, staphylococci, and proteolytic bacteria E. colt Other E, coli strains E. coli B. anthracis and other spore- forming bacteria A . aerogenes B. anthracis, P. festis E. tyfhosa E. tyfhosa, Ps. fiuorescens, E. coli, B. anthracis S. far at yf hi E. coli, B. anthracis, P. festis KNOWN PROPERTY Thermostable, filter- able substance Depresses growth Thermostable, filter- able substance Lytic action Colorless, thermo- stable, lytic sub- stance Alcohol-soluble pigment Growth-inhibiting REFERENCES 64, 78, 131,235, 236, 298, 557 451a, 798, 800, 801 244, 303, 315, 334, 417, 418, 420, 421, 446, 563, 566,692 593 51, 229,777 420 53,55, 132,368, 515,685,769,912, 983 3573,681 no, 344,367,369, 457,485, 819,923 244, 367, 369 315,354,923, 936 244, 462, 810, 923 NONSPORE-FORMING BACTERIA TABLE 13 {continued) 97 ANTAGONIST Streptococci Streptococci Staphylococci Micrococci Diplococci and pneumococci K. -pneumoniae P. vulgaris P. avicida Myxobacteria Anaerobic bac- teria ORGANISMS AFFECTED B. anthracis, C. difhtheriae B. anthracis, Ph. tumefaciens, S. lactis, P. festis, L. bul- garicus Gram-positive bacteria, C. difhtheriae, P. festis V. comma, M. tuberculosis, E. tyfhosa, Br. melitensis Various bacteria B. anthracis, C. difhtheriae, P. festis B. anthracis, P. festis, CI. sforogenes B. anthracis, E. tyfhosa Plant-disease-producing bacteria M. tuberculosis, B. anthracis KNOWN PROPERTY Activity not associ- ated with hemoly- sis or virulence Thermostable, non- filterable substance Thermolabile sub- stance Active filtrate Thermostable lytic substance REFERENCES 53, 110, 187, 233, 303,670,711, 836 70, 244, 802, 1007 53, 155,215,244, 247 213, 214, 580, 625, 670 213, 214, 243, 244, 370, 580,677, 766 677>7"5 853 36, 244, 923, 985 440, 708 S. marcescensy Ps. fluorescenSj and Sacckaromyces cereviseae; spore formation by the last was favored (800). The specific antagonistic action of Ps. aeruginosa upon various bac- teria was found by early investigators to be due to the production of an active heat-resistant substance. By filtering the culture through a Berk- feld, evaporating to a small volume, dialyzing through a parchment membrane, precipitating with alcohol, and drying over sulfuric acid, a preparation was obtained which was designated as pyocyanase (see p. 51). It had, even in very low concentrations, a marked destructive effect upon diphtheria, cholera, typhus, and plague organisms, as well as on pyogenic streptococci and staphylococci. It rapidly dissolved V. 98 BACTERIA AS ANTAGONISTS comma cells and in a few seconds rendered inactive such bacterial toxins as that of diphtheria. Since the bacteriolytic action of pyocyanase was in direct proportion to the time of its action and concentration, and in in- verse proportion to the numbers of bacteria acted upon, its enzymatic nature was believed to be substantiated. The preparation withstood heating in flowing steam for 2 hours. It has been established that pyocyanase has a lytic effect against the diphtheria organism, streptococci, meningococci, the typhoid organism, pneumococci, P. festis. Vibrio metchnikovi, V. commas and many other bacteria. There has been considerable disagreement, however, concern- ing the chemical nature and therapeutic action of pyocyanase, due largely to the variation in the nature of the preparations obtained. Kramer, for example, has shown (529) that the activity of the sub- stance depends on three factors: nature of strain, not all strains being equally effective} composition of medium, glycerol-containing media being most favorable} and method of extraction of active substance from culture media. The enzymatic nature of pyocyanase was not uni- versally accepted, largely because of the thermostability of the sub- stance, its solubility in organic solvents, and the fact that temperatures of o to 37° C. fail to influence its activity (59, 420, 737). Vs. aeruginosa produces, in addition to pyocyanase, a blue pigment, pyocyanin. Both substances possess lytic properties, i : 1,000 dilution of the pigment being able to lyse E. coli in 6 hours. Pyocyanin was said to be more effective in younger cultures, and pyocyanase in older. Pyo- cyanin had a bactericidal action also upon S. hemolyticusy S. albus, S. aureus, C. dl-phtheriae , M. tuberculosis, V. metchnikovi, and the Y-Ruhr bacillus, but not upon P. vulgaris, E. coli, or the typhoid organ- ism. In general, gram-positive bacteria were largely affected. Numer- ous other substances have been isolated from the cells of the organism or from the culture medium of Ps. aeruginosa. It is sufficient to men- tion the pyo-compounds and pyolipic acid. In order to test the action of Ps. aeruginosa upon other bacteria, Kramer (529) placed a drop of a suspension of this organism upon a plate inoculated with M. tuberculosis or with V. metchnikovi. In 24 hours, a sterile zone surrounded the colony of the antagonist, the width COLON-TYPHOID BACTERIA 99 of the zone depending upon the moisture content of the medium, the degree of diffusion of the active substance, its concentration, and the resistance of the test bacteria. When either of the two pathogens was inoculated into liquid media and the antagonist was introduced simul- taneously or within 24 hours, the latter had a decided bactericidal effect. No less extensive is the literature on the antagonistic action of the fluorescent group of bacteria, first established by Garre (315) in 1887 and later by others. Its bacteriostatic spectrum is illustrated in Table 14. The active substance is thermostable, dialyzes through a membrane, passes through Seitz and Berkfeld filters and is said to be soluble in chloroform (418, 566). Aerobic culture conditions are favorable to its accumulation. Members of this chromogenic group of bacteria were also found to be able to bring about the lysis of infusoria (134). S. marcescens exerts antagonistic effects against a number of bacteria, including diphtheria, gonococci, anthrax, and CI. chauvoeiy as well as fungi causing insect diseases (624). The formation of antibiotic sub- stances by this organism has been demonstrated by various investi- gators. These substances are active not only in vitro but also in vivo. Their formation was believed not to be associated with the production of the pigment by the organism. Hettche (420), however, asserted that the bactericidal action of Serratia is closely related to pigment produc- tion. The pigment was extracted with alcohol and was found capable of dissolving dead gram-positive bacteria but not gram-negative organ- isms. Eisler and Jacobsohn (229) ascribed the antagonistic action of Serratia not to the pigment but to certain water-soluble, thermostable (70° C. for 30 minutes) lytic substances. THE COLON-TYPHOID BACTERIA Members of the colon-typhoid group are not typical soil inhabitants, although they find their way continuously into the soil and into water basins. Various organisms belonging to this group have been said to possess antagonistic properties (440). Bienstock {ss) reported, in 1899, that proteolytic bacteria are repressed by the presence of E. coli and A. aero genes. Tissier and Martelly (912) emphasized that this phenomenon occurs only in the presence of sugar, the effect being due 100 BACTERIA AS ANTAGONISTS TABLE 14. ANTAGONISTIC ACTION OF PS. FLUORESCENS UPON VARIOUS MICROORGANISMS ORGANISM PERCENTAGE OF AGED MEDIUM IN THE AGAR 0.5 I.O 2.5 5.0 10 15 20 30 40 50 B. cereus — — + B. mycoides — — + B. ant hr ads — + B. vulgatus — - + B.subtilis - - + B, megatherium — + R. cinnebareus — + R. roseus - - + M. fiavus — - — + N. catarrhalis — — — + Ps. aeruginosa __________ Ps. fiuorescens __________ S.lutea - - - + S. marcescens _____ + S.albus - - + S. aureus — - — + S. citreus — — + K. fneumoniae - - - + V. comma — + Ch. violaceum — + £'. tyfhi - - + 5A. faradysenteriae — — + 5. enteritidis — — — + 5. suisfestifer — — — + 5. pillorum — - — + £. ^o/i ______ + ^ . aero genes ______ 4. ?>^. bowlesii - — + 5^^:. marianus __________ 5H value of the medium or in the oxidation-reduction potential or by a direct enzy- matic effect. In some cases thermolabile, filterable substances were dem- COCCI 103 onstrated (369, 618). These substances have been considered to be either autotoxins (148) or proteolytic enzymes (719). The filtrate of E. coli was reported (836) to be highly selective in its action, depress- ing only the dysentery organism of Shiga. Gundel (372) isolated from a bouillon culture of E. coli thermostable lipoids capable of bringing about the lysis of the colon organism and other bacteria. The antago- nistic relations between E. coli and V. comma are well established. The cholera organism also possesses antagonistic properties (308, 499). The typhoid organism is also capable of exerting an antagonistic ac- tion against itself as well as against Ps. jiuorescenSy E. coli, and various other bacteria, including B. anthracis. The nature of the action is not clearly understood. Salmonella faratyfhi possesses antagonistic proper- ties against E. coli, B. anthracis, P. -pestis, and various other bacteria. COCCI Numerous cocci have been found to possess antagonistic properties against other bacteria. Doehle (187) first demonstrated in 1889 that streptococci are able to antagonize B. anthracis, especially on solid media. Similar action was exerted against diphtheria bacteria j this ac- tion was not correlated with the hemolytic properties or the virulence of the antagonist. Further studies established the effect of various streptococci against anthrax. This effect was found (no) to be more pronounced in liquid than in solid media, and to be highly specific as regards the strain. S. pyogenes was shown to be antagonistic, in vivo, to B. anthracis and to Phytomonas tumejaciens, even to the extent of sup- pressing vegetative malformations brought about by the last named (70). 5. cremoris was active against S. lactis ( 1007), 5. mastidis against 5. lactis and L. acidophilus, and Streptococcus mucosus against P. pestis. Rogers (802) reported an antagonistic effect of S. lactis against L. bul- garicus; the active substance was thermostable and would not pass through a bacterial filter. More recently, certain streptococci were found (625) to produce a very potent antibiotic which was thermo- stable and dialyzablej it was active against various gram-positive but not gram-negative bacteria j it was well tolerated on subcutaneous and 104 BACTERIA AS ANTAGONISTS intravenous injection, and was believed to offer promise as a chemo- therapeutic agent, Freudenreich (298) first emphasized the antagonistic action o£ staphylococci against various bacteria. The list was later enlarged to in- clude gram-positive acid-resisting forms, corynebacteria, and the plague organism. Some of these antagonists were found to be able to lyse the dead cells of their own kind as well as those of various other organisms. Gundel (372) isolated from staphylococci an active lipoid which had bactericidal properties. A water-soluble, alcohol-insoluble substance, said to be an enzyme capable of bringing about the lysis of corynebac- teria, was also isolated from a strain of staphylococcus (215). Various micrococci possess strong antagonistic properties. Lode (580) isolated a micrococcus which affected a variety of microorganisms three or more centimeters away, the active substances being dialyzable. An organism related to Micrococcus tetragenus and described as M. an- tibioticus was found to possess a strong antagonistic action against V. comma y M. tuberculosis y E. tyfhosaj Ph. tumejaciensy Br. melitensisy various spore-forming bacteria, numerous cocci, and others. Diplococci exerted an antagonistic action against various bacteria, in- cluding pyogenic staphylococci and streptococci in the sputum, spore- formers, and gram-negative bacteria. They produced, under aerobic conditions only, a filterable substance that was heat resistant. The antagonistic action of pneumococci has definitely been estab- lished. The active substance of these organisms was said to be thermo- labile, since it was destroyed at 80° to 85° C.j it was produced only under aerobic conditions. In reviewing the literature on the longevity of streptococci in symbiosis, Holman (440) observed that many chances of error are inherent in mixed cultures, particularly with closely similar organisms j pneumococci, for example, were found to be able to live for long periods in association with nonhemolytic streptococci. Peculiar antagonistic relations between pneumococci and staphylococci were also reported (13). Adaptive alterations could be expected in the growth of bacteria in mixed cultures (32). Which of the two organisms antagonizes the other was believed to depend frequently upon the nu- merical abundance of one or the other (243). OTHER BACTERIA 105 OTHER AEROBIC AND ANAEROBIC BACTERIA The antagonistic action o£ K. pneumoniae against B. anthracis has been reported. Freudenreich (298) found that the filtrate of this an- tagonist repressed the growth of a number of bacteria, including the diphtheria and plague organisms. Other aerobic bacteria were found capable of exerting antagonistic effects against one or more organisms, these effects varying considerably in nature and intensity. It is sufficient to mention the action of P. vul- garis against B. anthracis and P. festis; of Ps. aviseftica against B. an- thracis and E. tyfhosa; of Bacterium lactis aerogenes against B. an- thracis and P. festis. B. anthracis is capable of iso-antagonism and of antagonizing certain other organisms, including E. tyfhosa and Bac- terium acidi lactici (786). Certain Myxobacteriales have been shown to be capable of bringing about the lysis of various plant-disease-producing bacteria 3 a thermostable lytic substance, passing through cellophane but not through a Seitz filter, was obtained. Although certain bacteria like Achromobacter lifolyticum were found capable of reducing the patho- genicity of M. tuberculosis, no active cell-free extract could be ob- tained (79). M. tuberculosis produces a water-soluble substance, designated phthiocol, which in concentrations of 0.05 to o.i per cent inhibited the growth of various gram-positive and gram-negative bacteria, but not Ps. aeruginosa (568a). Bacillus larvae, a gram-negative rod, was found (441) capable of inhibiting the growth of various gram-positive and gram-negative bac- teria. The human and bovine strains of M. tuberculosis were also in- hibited but not the avian strain. The antibiotic was soluble in water but not in organic solvents. It was adsorbed on activated charcoal but no eluent could be found. It was moderately heat stable. Its antibiotic ac- tivity was inhibited by glucose but not by cysteine or sucrose. The morphology of one bacterium may be considerably modified by the presence of another. Living cultures of L. bulgaricus influenced the variation of E. coli from the "S" to the "R" phase, inhibited develop- ment of the organism, and even brought about its lysis. No active sub- 106 BACTERIA AS ANTAGONISTS stance could be demonstrated} the lactic acid itself had only a limited effect (9). Korolev (528) has shown that when a yellow sarcina was added to solid media a stimulating effect was exerted on the growth of species of Brucella {Br. melitensisj Br. abortus, Br. suis)-, in liquid media, however, the activities of these species were repressed, the sar- cina thus acting as an antagonist. A white staphylococcus exerted an an- tagonistic action on Brucella species both in liquid and on solid media. Certain acid-producing aerobes were found capable of inhibiting toxin production by Clostridium hotulinum in glucose but not in noncarbo- hydrate media (373). Since acid itself cannot bring about this effect, Holman (440) suggested that the acid must be active in a nascent state. A mixture of a Clostridium sforogenes and CI. Botulinum also inter- fered with the development of the toxin j it was even thought possible that the first anaerobe might cause the disappearance of toxin already produced (164, 165). S. aureus, E. coU, P. vulgaris, and other bac- teria permitted the growth of CI. hotulinum in aerobic cultures, accom- panied by toxin production (290). However, Streptococcus thermofhi- lus inhibited the growth of CI. hotulinum, the toxin of the latter being gradually destroyed (493). Passini (708) claimed that Bacillus futrificus verrucosus destroyed M. tuberculosis in nine days. The effect of other anaerobes on the sur- vival of anthrax spores in dead animals has been extensively studied (440). Novy (688) reported that the injection into guinea pigs of P. vulgaris and Clostridium oedem^atiens resulted in rapid death of the animals and extensive growth of the anaerobe in the animal bodies; however, the simultaneous inoculation of CI. sforogenes and P. vul- garis did not result in putrid lesions. According to Barrieu {'^6), the presence of P. vulgaris and certain nonpathogenic spore-bearing aerobes in wounds favors, through their proteolytic activity, the virulence of pathogenic bacteria. Pringsheim (738) grew CI. welchii with Alka- ligenes fecalis for ten generations on agar slants and could easily detect in the growth of the latter the opaque colonies of the anaerobe. A lique- fying sarcina allowed CI. welchii and CI. butyricum to grow in open tubes. Many war-wound infections were believed (985) to be due to an association of P. vulgaris with anaerobes, since the former increased the virulence of CI. ferjringens and others. OTHER BACTERIA 107 The antagonistic effects of lactic acid bacteria of the L. bulgarkus and L. acidophilus groups have received considerable attention, especially in regard to their action against intestinal bacteria. This was believed to be due to the production of acid by the bacteria rather than to the formation of specific antagonistic substances. This phenomenon aroused particular interest because of the function of some of these organisms in replacing bacterial inhabitants of the human digestive system (526). Various bacteria also have a marked destructive effect upon plant pathogenic fungi, as will be shown later. Some produce stable, heat- resistant, antifungal substances (731 ). CHAPTER 6 ACTINOMYCETES AS ANTAGONISTS AcTiNOMYCETES are found in large numbers in many natural sub- strates. They occur abundantly in soils, composts, river and lake bot- toms, in dust particles, and upon plant surfaces. Certain species are capable of causing serious animal and plant diseases. Actinomycetes, like fungi, produce a mycelium, but they are largely unicellular organisms of dimensions similar to those of bacteria. Some of the constituent groups are closely related to the bacteria, others to the fungi. On the basis of their morphology, the order A ctinomycetales has been divided into three families, Mycohactenaceae^ Actinomy- cetaceae^ and Streftomycetaceaey comprising the genera Mycobacte- riumy Actinomyces y Nocardiay StreftomyceSy and Micromonosfora. These genera are represented in nature by many thousands of species, of which several hundreds have been described. A few are shown in Figure lo. Comparatively little i^ known of the physiology of actinomycetes. Some produce certain organic acids from carbohydrates j others prefer proteins and amino acids as sources of energy, many species being strongly proteolytic. Some are able to attack starch, with the production of dextrins and sugar, accompanied by the formation of diastatic en- zymes. Many reduce nitrates to nitrites. Some attack sucrose and form the enzyme invertasej others, however, do not. Certain species are able to utilize such resistant compounds as rubber and lignin. Synthetic media are favorable for the production of a characteristic growth and pigmentation. Among the pigments, the melanins have received par- ticular attention. They range from the characteristic brown to various shades of black and deep green and are formed in protein-containing and in some cases also in protein-free media. The other pigments range from blue, yellow, and orange to various shades of grey. According to Beijerinck (41), the process of pigment production by actinomycetes in gelatin media is associated with the formation of a S. antibiotic lis y important antagonist. From Waksman and Woodruff (974) S. hivcjidulary important antagonist ' 4^ -^li Submerged growth of S. lavendulae. From Woodruff" and Foster (1031) S. grlseus, streptomycin-producing str; Prepared by Waksman and Schat2 *p V 4 r > 4 X- { ^ M. vulgaris. From Waksman. Cordon, ar id Hulpoi (953) "X^r Streftomyces 3042, showing close spi type of branching. Prepared by Starl Figure 10. Types of actinomycetes. ACTINOMYCETES AS ANTAGONISTS 109 quinone, which turns brown at an alkaline reaction and in the presence of oxygen. The action of quinone in the presence of iron was found to be similar to that of the enzyme tyrosinase. Since an excess of oxygen is required for the formation of quinone, only limited amounts are found in deep cultures. The quinone is believed to be formed from the pep- tone in the medium j although good growth was produced on media containing asparagine, KNO3, and ammonium sulfate as sources of ni- trogen, only traces of quinone, if any, were found. The tyrosinase reac- tion is not involved in the production of all black pigments by actinomy- cetesj some species produce such pigments in purely synthetic media, in the complete absence of peptone. Actinomycetes grow in liquid media in the form of flakes or small colonies, usually distributed either on the bottom and walls of the con- tainer or throughout the liquid j often a ring is formed on the surface of the medium around the wall of the vessel. In many cases, a full sur- face pellicle is produced, which may be covered with aerial mycelium. As a rule, the liquid medium does not become turbid, even in the pres- ence of abundant growth. When grown on solid media, actinomycetes form small, compact, soft to leathery colonies j a heavy lichen-shaped mat is produced that may become covered by an aerial mycelium. The addition of a small amount of agar (0.25 per cent) to a liquid medium is highly favorable to growth, especially in large stationary containers. Actinomycetes can also be grown in liquid media in a submerged con- dition, with suitable agitation and aeration in order to supply oxygen j the medium may also be kept in shaken state. Growth occurs in the form of a homogeneous suspension of discrete colonies and mycelial fragments throughout the liquid. Responses in growth and biochemical activities as a result of treatments may thus be obtained under more homogeneous physiological conditions. Although most actinomycetes are aerobic, some are anaerobic, and many can grow : t a reduced oxygen tension. The aerobic actinomycetes commonly found on grasses and in soil are said (511) never to have been isolated from animal infections. Mixed infections consisting of anaerobes growing at body temperature together with aerobes have often been demonstrated. Certain aerobic species also are capable of no ACTINOMYCETES AS ANTAGONISTS causing infections in man and other animals, and certain plant diseases (potato scab, sweet potato pox) are caused by aerobic species of actino- mycetes. ANTAGONISTIC PROPERTIES Many actinomycetes have the ability to antagonize the growth of other microorganisms, notably bacteria, fungi, and other actinomycetes; this is brought out in Tables 1 5 and 1 6. The antagonistic species are not limited to any one genus but are found among three genera, Nocardiay StreftomyceSy and Mkromonosfora, Gasperini (322) first demonstrated, in 1890, that certain species of the genus Strepomyces had a marked lytic effect upon other micro- organisms. He emphasized that "Streftothrix develops habitually in a spontaneous manner upon the surface of bacteria and fungi, upon which it lives to a limited extent in the form of a parasite, due to the faculty that its mycelium possesses to digest the membrane from these lower fungi." Greig-Smith (364, 365) found that soil actinomycetes are an- tagonistic to not only bacteria but also certain fungi j since actinomy- cetes grow abundantly in normal soils, it was suggested that they may become an important factor in limiting bacterial development. Lieske established (571) that specific actinomycetes are able to bring about the lysis of many dead and living bacterial cells; they are selec- tive in their action, affecting only certain bacteria such as S. aureus and S. fyogenesy but not S. lutea, S. marcescens, or Ps. aeruginosa. Rosenthal (805) isolated from the air an actinomyces species which he designated as the true biological antagonist of the diphtheria or- ganism. He inoculated the surface of an agar plate with an emulsion of the bacteria and inoculated the actinomyces into several spots. At the end of two days, the plate was covered with the diphtheria organisms, but the colonies of the actinomyces were surrounded by large trans- parent zones. In another method utilized, agar was mixed with an emul- sion of the diphtheria bacteria killed by heat, and the mixture was poured into plates. After solidification of the medium, the antagonist was inoculated in several spots on the plates. Its colonies gradually be- came surrounded by clear zones, thus proving that it produced a lytic ANTAGONISTIC PROPERTIES 111 TABLE 15. ANTAGONISTIC PROPERTIES OF VARIOUS ACTINOMYCETES ANTAGONIST ORGANISMS AFFECTED KNOWN PROPERTY REFERENCES S. alius Pneumococci, strepto- Thermolabile sub- 354, 357> cocci, staphylococci, stance, causes lysis looi, 1002 Ps. aeruginosa, etc. of dead cells S. alius Various fungi Protein, enzyme, causes lysis of dead and certain living bacteria 10-12 S. antibioticus All bacteria and fungi, Thermostable sub- 976 especially gram-posi- tive types stance, bacterio- static S. griseus Gram-positive and gram- negative bacteria, not fungi or anaerobic bacteria Produces streptomy- cin 830 S. lavendulae Various gram-positive Produces streptothri- 979 and gram-negative bacteria cin S. fraecox S. scabies 644 Streftomyces sp. Bacteria and fungi Lytic action 322 Streftomyces sp. Diphtheria Growth inhibition 805 Streftomyces sp. B. mycoides, proactino- mycetes, mycobacteria Bactericidal action, with or without lysis 76,534 Streftomyces sp. Fusarium Lytic action 633 A'', gardneri Gram-positive bacteria Bacteriostatic action 313,958 Micromonosfora Gram-positive bacteria Thermostable active substance produced 958 Actinomycetes Dead and living bacteria Lysis 571 Actinomycetes Spore-forming bacteria Repression of growth 364, 1000 Actinomycetes Gram-positive bacteria Thermostable sub- stance, produced on synthetic media, resembles lysozyme 536,671 Actinomycetes Pythium Thermostable sub- stance 908 2 ACTINOMYCETES AS ANTAGONISTS TABLE 1 6. ANTIBACTERIAL SPECTRUM OF CERTAIN ANTAGONISTIC ACTINOMYCETES TEST ORGANISM ZONE OF INHIBITION. , IN MILLIMETERS S. violaceus 5. aurantiacus S. griseus S. gl obis for us N. rubra 35 32 N. corallina 40 45 22 10 N. alba 40 25 M. rubrum 40 33 10 M. citreum 38 37 M. tuberculosis 8 10 M. smegmae 10 8 O M. fhlei 20 25 Corynebacterium sp. 12 10 E. coli S. aureus 25 19 M. ruber 35 28 B. mycoides 30 10 B. megatherium 33 5 B. mesentericus 30 2 B. subtilis 23 I B. tumescens 22 Ps. fluorescens Ps. aeruginosa O P. vulgaris o S. marcescens o M. luteus 30 25 M. candicans 37 22 M. roseus 42 27 M. lysodeikticus 38 33 o S. lutea 30 27 A%, vinelandii 3 o Az. chroococcum 5 Rh. leguminosarum Radiobacter From Krassilnikov and Korenlako (534)- substance that diffused through the agar and dissolved the diphtheria cells. Gratia and Dath (357) suspended dead cells of staphylococci and other bacteria in 2 per cent agar and exposed the plates to the air. A cul- ANTAGONISTIC PROPERTIES 113 ture of a white actinomyces developed on the plates, each colony being surrounded by a clear zone of dissolved bacterial cells. By transferring this culture to a suspension of dead staphylococci in sterile saline, a characteristic flaky growth was produced, the bacterial suspension be- coming clarified in 36 hours. When the lysed emulsion was filtered, the filtrate could again dissolve a fresh suspension of dead staphylococci. This culture was found able to attack all staphylococci tested as well as certain gram-negative bacteria, such as Ps. aeruginosa; however, it was inactive against M. tuberculosis and E. coli. Some antagonistic strains could also attack E. coli, though this property was readily lost. This type of antagonism was believed to be widely distributed in na- ture and to be directed against many bacteria, pathogenic and sapro- phytic. The culture of the antagonist in bouillon gave a very active agent, whereas the lysed bacterial suspension was weaker in its action. The active substance was present extensively in old cultures and was fairly stable. The material obtained by lysing the suspension of bacteria by means of an antagonist was designated as "mycolysate." It did not possess the toxicity of the nonlysed suspension but it preserved its anti- genic properties {2>S^)' Gratia (354) also reported that actinomycetes were able to attack living cells of bacteria, except E. coli and E. tyfhosa which had to be first killed by heat before they could be dissolved. Welsch (100 1, 1002) made a detailed study of the lytic activity of an actinomyces culture, presumably identical with that employed by Gratia and later described as Actinomyces alhus. The culture was grown in different media, the best results being obtained in very shallow layers of ordinary bouillon. The active substance present in the filtrate was designated as "actinomycetin." It was able to dissolve, at least partly, all dead bacteria, whether killed by heat or by chemicals, gram-positive or gram-negative, though gram-negative bacteria were, as a rule, more susceptible. The growing culture of the antagonist brought about better clarification (lysis) of the bacterial suspension than the filtrate. The solubilizing properties of the active agent, its susceptibility to heat and to ultraviolet rays, its size as measured by ultrafiltration, suggested its protein nature. The kinetics of its action pointed to its being an enzyme. It was precipitated by acetone, alcohol, and ammonium sulfate. Most of the gram-negative bacteria were not attacked either by actinomycetin 114 ACTINOMYCETES AS ANTAGONISTS or by the living culture of the antagonist. Only a few of the gram- positive bacteria, including certain pneumococci and streptococci, could be dissolved by sterile actinomycetin. A definite parallelism in the ac- tivity of the preparation against dead bacteria and of the living culture against living bacteria suggested that the same substance is concerned in both cases. The bacteria were therefore divided ( looo), on the basis of their relation to actinomycetin, into three groups : Bacteria that were lysed by the culture filtrate; these included pneumo- cocci and hemolytic streptococci Bacteria that were not dissolved even by the most active soluble sub- stance, but which were depressed by the mycelium of the living ac- tinomyces; these comprised various sarcinae and fluorescens types Bacteria that were not acted upon by either the living culture or the actinomycetin preparation ; these included the colon-typhoid and the pyocyaneus groups, though when the latter were killed by heat or inactivated by radium emanations, as in the case of E. colt, or were placed under conditions unfavorable to multiplication, they were dis- solved by the lytic substance. The first detailed survey of the distribution of antagonistic organisms among actinomycetes was made by a group of Russian investigators. According to Borodulina (76), actinomycetes are able to antagonize various spore-forming bacteria and to bring about the lysis of their liv- ing cells. A thermostable substance was produced on agar media. The activity of this substance was greatly reduced at an alkaline reaction, whereas an acid reaction favored it. When B. mycoides and an antago- nist were inoculated simultaneously into peptone media, no inhibitive effect was obtained, because the bacterium changed the reaction of the medium to alkaline, thereby making conditions unfavorable for the production of the antibiotic substance by the antagonist. When the an- tagonist was first allowed to develop in the medium, before the bac- terium was inoculated, a strong antagonistic effect resulted, which led to the elongation of the vegetative cells of B. mycoides; this was due to a delay in fission and was accompanied by the suppression of spore formation. Krassilnikov and Koreniako (534) found that many species of actino- mycetes belonging to the genus Streftomyces but not Nocardia pro- ANTAGONISTIC PROPERTIES 115 duced a substance that possessed a strong bactericidal action against a large number of microorganisms. This substance was particularly active against nocardias, mycobacteria, and micrococci ; it was less active upon spore-bearing bacteria and had no action at all on nonspore-forming bacteria, as illustrated in Table i6. Under the influence of the anti- biotic factor, the microbial cells were either entirely lysed or killed with- out subsequent lysis. The action upon spore-bearing bacteria was bac- teriostatic but not bactericidal. The nonspore-forming bacteria, includ- ing species of Rhizobiutn and Azotobacter, not only were not inhibited but were actually able to develop in filtrates of the antagonists. Of 80 cultures of actinomycetes isolated from different soils, 47 pos- sessed antagonistic properties, but only 27 of them secreted antibiotic substances into the medium (Table 17). These agents were capable of inhibiting the growth of gram-positive but not of gram-negative bac- teria or fungi. The nature of the action of the various antagonists was TABLE 17. OCCURRENCE OF ANTAGONISTIC ACTINOMYCETES IN DIFFERENT SOILS TOTAL STRAINS NUMBER OF STRAINS W^HICH OF ACTINOMY- ANTAGONISTIC LIBERATED TOXIC NATURE OF SOIL CETES TESTED STRAINS SUBSTANCES Chernozem 24 10 9 Podzol 11 Solonets 4 High altitude soil 9 Sandy soil 6 Dry desert soil 5 River bank meadow 14 Cultivated soil 7 — — — Total 80 47 27 From Nakhimovskaia (671). found not to be identical. Some excreted water-soluble substances into the medium, others did not. All the antibiotic agents were thermo- stable, since heating for 30 minutes at 1.5 atmospheres only reduced somewhat their activity. For those antagonists which did not excrete 116 ACTINOMYCETES AS ANTAGONISTS any substance into the medium, the presence of the growing antagonist was essential in order to bring about an inhibition of bacterial develop- ment. On the basis of their sensitivity to the antibiotic substance of actinomycetes, mycobacteria could be differentiated from nonspore- forming, especially nodule-forming, bacteria. The production of the antibiotic substance was highest in synthetic media and was rather weak or even totally absent in media that contained proteins. The substance was filterable and was able to resist the effect of radiation. It was further reported (671) that the antagonistic effects of actino- mycetes were manifested not only in artificial media but also in soil, the interrelations here being much more complex. Some of the strains that produced antagonistic effects in artificial nutrient media were ineffec- tive under soil conditions. The antagonistic action was more intense in light podzol soils and was greatly reduced in heavy or chernozem soils. One of the factors that resulted in a decrease in the antagonistic properties of actinomycetes in the heavy soils was apparently the high content of organic matter. By adding peptone to a light soil, the antago- nistic action of the actinomycetes was greatly weakened. When actino- mycetes were allowed to multiply in the soil before inoculation with B. mycoidesy the antagonistic effect was highly pronounced even in the presence of high concentrations of peptone. An attempt to isolate an antibiotic substance from some of the soil actinomycetes was made by Kriss {S'i^)- On the basis of its properties he was led to conclude that this substance could be classified definitely with lysozyme. It was insoluble in ether, petroleum ether, benzol, and chloroform, and was resistant to the effects of light, air, and high tem- peratures. The optimum reaction for the production of this substance by Streftomyces violaceus was found to be ^H 7.1 to 7.8, the activity not being increased by selective cultivation. On the basis of its properties, this substance could hardly be classified with egg-white lysozyme. It must be concluded also that the differences in the antibiotic properties of the various antagonistic actinomycetes isolated by the Russian investi- gators definitely point to the fact that more than one antibiotic substance was involved. In a more recent survey (958) of the distribution of antagonistic ac- tinomycetes in soils and in composts, it was found that of 244 cultures ANTAGONISTIC PROPERTIES 117 isolated at random from different soils, 49, or 20 per cent, of the cultures were actively antagonistic} 57, or 23 per cent, showed some antagonistic properties; and 138, or 57 per cent, possessed no antagonistic action at all (Table 18). A somewhat similar distribution of antagonistic prop- erties was observed among a group of well-identified species taken from a type culture collection, embracing 161 pure strains. Only one of the members of the genus Nocardia proved to be antagonistic ; only one of the Micromonospora forms was active. Most of the antagonists were found among the members of the genus Stre-ptomyces. These cultures were also examined for bacteriolytic properties, living S. aureus being TABLE 15. ISOLATION OF ANTAGONISTIC ACTINOMYCETES FROM VARIOUS SUBSTRATES GROUP I GROUP n GROUP III GROUP IV TTITAT. Percent- Percent- Percent- Percent- SOURCE OF CULTURES Cul- age of Cul- age of Cul- age of Cul- age of ORGANISMS ISOLATED tures total tures total tures total tures total Fertile, ma- nured, and limed soil 74 20 27.0 5 6.8 I 1-3 48 64.9 Infertile, un- manured. limed soil 75 11 14.7 8 10.7 4 5.2 52 69-3 Potted soil 13 I 7-7 I 7-7 II 84.6 Potted soil, en- riched with E. coli 21 I 4.8 4 19.0 4 19.0 12 57.2 Potted soil, en- riched with mixtures of bacteria 15 12 80.0 2 13-3 I 6.7 Lake mud 9 3 33.3 4 44.4 2 22.2 Stable-manure compost 37 ' 2.7 20 54.0 4 10.8 12 324 Total 244 49 20.1 44 18.0 13 5-3 138 56.6 From Waksman, Horning, Welsch, and Woodruff (958). Note. The organisms in group I were the most active antagonists, those in groups II and III had more limited antigonistic properties, and those in group IV showed no antibacterial effects with the methodg used. 118 ACTINOMYCETES AS ANTAGONISTS used as the test organism. On this basis, 87 cultures (53.1 per cent) were found to be inactive, 53 cultures (32.3 per cent) were moderately active, and 24 cultures (14.6 per cent) were highly active. The conclu- sion was reached (1000) that bacteriolytic activities against killed bac- teria and living gram-positive bacteria are widely distributed among the actinomycetes. Growth-inhibiting properties of actinomycetes were found to be significantly associated with bacteriolytic action upon living gram-positive bacteria. Certain actinomycetes also show antagonistic activities against fungi (10-12, 908). S. albus was capable of inhibiting the growth of all the species of fungi tested, an effect shown to be due to the production of an active substance. By the use of a culture of Colletotrkhum gloe- osforioideSj the antagonistic activities of 80 type cultures of actino- mycetes were measured. The antagonist was allowed to grow for 5 days on maltose agar, at fH 7.4, and the fungus was then inoculated. The cultures of actinomycetes were divided, on this basis, into three groups: strong, weak, and noninhibitors. The first group comprised 17.5 per cent of the cultures j the second, 38.8 per centj and the third, 43.7 per cent. These results are surprisingly similar to those reported for the distribution of actinomycetes possessing antibacterial properties, includ- ing those that were isolated at random from the soil and those taken from a culture collection. Meredith (633) made a survey of the distribution of organisms an- tagonistic to Fusarium oxys forum cubense in Jamaica soils j most of these antagonists belong to the actinomycetes. The antagonists were not evenly distributed in the various soil samples, 10 of the d^i samples giv- ing 44 per cent of the antagonistic organisms. Those actinomycetes that were antagonistic to Fusarium when grown in their own soil-solution agar were not always antagonistic when tested in soil-solution agar pre- pared from other soil. A culture of actinomyces isolated from a compost produced lysis of the Fusarium. When spores of both organisms were mixed in an agar medium, the fungus developed normally for two days but began to undergo lysis on the fifth day, large sections of the my- celium disappearing. On the seventh day only chlamydospores were ob- served. In 9 days the fungus completely disappeared, the actinomyces making a normal growth. NATURE OF ANTIBIOTIC SUBSTANCES 119 NATURE OF ANTIBIOTIC SUBSTANCES It has already been established that antagonistic actinomycetes read- ily produce a variety of different types of antibiotic substances. Some of these have been isolated and even crystallized and information has been gained concerning their chemical nature. Others have been obtained in the form of crude but highly active preparations. Still others are known but they have not been isolated as yet and have, therefore, been rather insufficiently studied. So far, eight substances have been definitely recognized: actinomycetin, actinomycin, streptothricin, streptomycin, proactinomycin, micromonosporin, litmocidin, and mycetin. Among the various antagonistic actinomycetes, five species have been studied in detail and, therefore, deserve particular attention, namely, S. antibioticus (974), S. lavendulae (979), 5. griseus (830), A'', gardneri (313), and S. albus (1000). S. antibioticus produces a highly active antibiotic substance that has been isolated and described as actinomycin. It was shown to be antago- nistic to all species of bacteria tested as well as to many fungi (Table 19). Actinomycin is not affected by heat. It is soluble in ether and in alcohol as well as in other solvents, but in water only in very high dilu- tions. It is highly toxic to animals. Several species of actinomycetes are capable of producing actinomy- cin in both organic and synthetic media, the yield varying with the or- ganism (955a, 1002a). The addition of 0.25 per cent agar to stationary cultures increases considerably the growth of the organism and the pro- duction of actinomycin. The presence of a small amount of starch, phos- phate, and sodium chloride was also found to be favorable. Actinomy- cin-producing forms are strictly aerobic, and are able to produce actino- mycin only when grown either in very shallow layers or under aerated or agitated submerged conditions. S. lavendulae is capable of inhibiting the growth of many gram- negative and gram-positive bacteria. It produces an antibiotic substance designated as streptothricin. For the production of streptothricin, the tryptone can be replaced by a variety of simple nitrogenous compounds, such as glycine (Table 20), alanine, aspartic acid, asparagine, and glutamic acidj the carbo- 120 ACTINOMYCETES AS ANTAGONISTS TABLE 19. BACTERIOSTATIC SPECTRUM OF ACTINOMYCIN ACTINOMYCIN ADDED, MILLIGRAMS ORGANISM GRAM STAIN PER LITER OF MEDIUM O.I I.O 10 100 S. marcescens - 3 3 3 3 A . aero genes - 3 3 3 3* E. coli (intermediate) - 3 3 3 3* E. coli - 3 3 3 I* Ps. aeruginosa - 3 3 3 Ps. fluorescens - 3 3 3 Br. abortus - 3 3 3 N. catarrhalis - 3 3 2 E. carotovora - 3 3 2 SA. ga/linarum - 3 2 2 A . stutzeri - 3 2 I H. fertussis - 3 3 Az. vinelandii - 3 S. cellulosae + 3 2 I S. calif ornicus + 3 3 2 M. tuberculosis + 3 3 CI. welchii + 3 B. macerans .+ 3 3 B. megatherium + 3 B. folymyxa + 3 B. mycoides + I B. mesentericus + I B. cereus + I B. subtilis I + B. subtilis II + • G. tetragena + S. lutea + Streptococci and staphylococci + From Waksman and Woodruff (975). Note, o indicates no growth; i, trace of growth; 2, fair growth; 3, good growth. * If 200 mg. per liter were added the results were usually as follows: for A. aero genes, fair; for E. coli (intermediate), trace; for E. coli, no growth. hydrate may be left out completely, with only limited reduction in ac- tivity. No growth of the organism is obtained on tryptophane, phenyl alanine, and certain other forms of nitrogen. Good growth may be ob- tained with ammonium sulfate or sodium nitrate, but the production of NATURE OF ANTIBIOTIC SUBSTANCES 121 TABLE 20. GROWTH AND PRODUCTION OF STREPTOTHRICIN BY S. LAVENDULAE TREAT- DAYS GROWTH IN MG, ACTIVITY IN UNITS SOURCE OF MENT OF OF INCU- PER 100 ML. E. ~ ' U. sub- NITROGEN CULTURE BATION OF MEDIUM colt tilis Tryptone Shaken 2 346 150 1,000 Tryptone Shaken 5 253 ICO 1,000 Glycine Shaken 2 162 30 30 Glycine Shaken 5 266 100 500 Tryptone Stationary 8 245 20 200 Glycine Stationary 8 239 25 150 From Waksman (946). Note. The organism was grown in i per cent starch medium. the active substance is limited unless the organism is grown under sub- merged conditions. Iron appears to play an essential role in the produc- tion of the active substance. An increase in growth as a result of an in- crease in the amino-acid concentration, with the same amount of carbo- hydrate, causes an increase in the production of streptothricin. An in- crease in growth as a result of an increase in carbohydrate concentration does not. When the medium contains one amino acid as the only source of car- bon and nitrogen, there is gradual increase in the alkalinity of the medium, resulting in the destruction of the streptothricin. Neither the growth of the organism nor the production of the streptothricin, how- ever, is influenced by the reaction of the medium, within certain limits, even between ^H 4.4 and 8.0 ( 1028). The metabolism of S. lavendulae and the course of production of streptothricin under stationary and sub- merged conditions are illustrated in Figure 1 1 . The bacteriostatic spec- trum of streptothricin is shown in Table 2 1 . It has a certain delayed, even if limited, toxicity to animals and is active in vivo against both gram-positive and gram-negative bacteria (792). Different strains of S. lavendulae differ greatly in their ability to pro- duce streptothricin. The possibility that other species of Streftomyces zre also capable of producing streptothricin or closely related com- pounds, as indicated by somewhat different antibiotic spectra, has also aamnniiAi ai3d simn 1 j / 1 ( > Q^ < z o 1- d 1- X \ < o \ 1- u < w 5 \ i > \ \ Q^ \ K - Q X \ ^ \ \ ^4. \ li^-A - 1 1 1 \ aaniino jo s?i3±nn"iii^ 99 2i3d sNvajDiiii^M TABLE 21. INHIBITORY EFFECT OF STREPTOTHRICIN UPON GROWTH OF VARIOUS BACTERIA CRUDE STREPTOTHRICIN ADDED, ORGANISM B. subtilis B. mycoides B. macerans B. megatherium B. folymyxa B. cereus M. lysodeikticus S. muscae S. lutea A . aerogenes* A . aero genes E. coU\ E. colt (4348) S. marcescens S. m^cescens Ps. fluorescensX Sh. gallinarum P. fseudotuberculosis Br. abortus S. cholerasuis S. schottmulleri S. abortivoequtTia S. tyfhimurium H. suis H. influenzae Br. abortus Az. agile Az. vinelandii Az. chroococcum Az. indicu?n M. fhlei CI. butyricum^ L. casei^ S. a/bus S. violaceus-ruber S. lavendulae ;ram :S per 10 CUBIC CENTIMETERS AGAR I 0.3 0.1 0.03 O.OI I 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 I 2 I 2 2 I 2 2 Tr 2 2 Tr I 2 I 2 2 2 I 2 2 2 2 2 2 2 2 Tr 2 2 Tr 2 I 2 2 Tr 2 2 2 2 2 2 2 2 I 2 2 2 2 Tr 2 2 2 2 2 I 2 2 2 2 2 2 2 2 2 2 I 2 2 I 2 2 2 2 From Waksman and Woodruff (979). Note, o indicates no growth; I, limited growth; 2, good growth; Tr, trace of growth. • Representing 3 distinct strains. t Representing 5 strains of E. coll obtained from different sources. t Representing 4 strains. § Cultured anaerobically. 124- ACTINOMYCETES AS ANTAGONISTS been established (969). Other members of the genus are capable of forming different antibiotics (498). S. grlseus produces an antibiotic substance, designated as streptomy- cin, that is also active against both gram-positive and gram-negative bacteria. It is similar in its solubility and certain chemical properties to streptothricinj however, it acts readily against B. mycoides and is more active than streptothricin against certain gram-negative bacteria, such as M. tuberculosis, S. marcescens, and Ps. aeruginosa. The bacteriostatic spectrum of streptomycin is given in Table 22. Streptomycin is also active in vivo against a variety of other bacteria, as shown later. It is also active against spirochetes, but it is not active against fungi, ana- erobic bacteria or viruses (791). TABLE 22. BACTERIOSTATIC SPECTRUM OF STREPTOMYCIN ORGANISM n ORGANISM f^S A. aero genes 0.5-2.5 M. tuberculosis, hominis 0.15 B. anthracis 0.375 N. gonorrhoeae 5.0 B. megatherium 0.25-3.0 P. pestis 0.75-1.5 B. mycoides 0.1-3.8 P. tularensis 0.15-0.3 B. subtilis 0. 1 2-1.0 Ph. fruni 0.25 Br. abortus 0.5-3-75 Pr. vulgaris 0.4-3-0 Br. suis 0.5 Ps. aeruginosa 2.5-25.0 CI. butyricum 8.0 Ps. fiuorescens 12.5 CI. tetani >i04 S. lutea 0.25 C. difhtheriae 0.375-3.75 5. marcescens I.O D. pneumoniae 8.0 5. enteritidis 0.5 E. tyfhi I -0-37.5 S. schottmillleri 2.0 Er. muriseftica 2-5 Sh. paradysenteriae 0.25-3.75 E. coli 0.3-3-75 S. aureus o.5->i6.o H. influenzae 1.56-5.0 S. hemolyticus 2.0->l6.0 H. -pertussis 1.25-3-0 S. viridans >i6 L. monocytogenes 2.5 V. comma 6.0-37.5 K. -pneumoniae 0.625-8.0 A. bovis 3-75 M. mallei 10->10.0 N. asteroides 12.5 M. avium lO.O S. antibioticus h6, P. chrysogenum also certain gram-nega- in vivo, low tox- 262, 437, 770, tive {Neisseria, Gone- icity 956 coccus') bacteria P. notatum All bacteria tested, in Notatin, penatin. 157,517,521. presence of glucose penicillin B, E. coli factor 786, 956 P. piberulum Various bacteria Penicillic acid 61,698,703 and P. cyclofium P. resdculosum Various bacteria Crude metabolic product 62 Penicillium sp. Gram-negative as well as gram-positive bacteria Penicidin 29 134 FUNGI AS ANTAGONISTS giving fresh lots of penicillin in about half the time required for the initial growth. Crude penicillin cultures are capable of inhibiting the growth of staphylococci in dilutions of i : 8oo. Recently, much more active preparations have been obtained ( i : 1 0,000 ). Chain et al. (123) were the first to succeed in isolating from the cul- ture medium of P. notatum a water-soluble, stable, brown powder which had marked antibacterial activity. This preparation inhibited, in dilutions of i to several hundred thousand, the growth of many aerobic and anaerobic bacteria. The active material was relatively nontoxic to laboratory animals. Intravenous and subcutaneous injections of 10 mg. or more to mice had little or no effect. The material was active m vivo, subcutaneous injections saving the lives of mice injected intraperitone- ally with S. pyogenes or S. aureus. Intramuscular infections of mice with CI. seftkum were also successfully treated by repeated subcutane- ous injections of penicillin. An extensive literature soon began to accumulate on the production (127, 128, 171, 281), isolation, and identification of penicillin. The course of its formation in the culture of the organism is illustrated in Figure 13. Conditions of nutrition were found to be particularly im- portant. Preparations having an activity of 2,000 Oxford units or 100,- 000,000 dilution units have been obtained. The importance of the dual nature of P. notatum (the culture being composed of two distinct cell constituents) must be recognized for maximum penicillin production (34, 377). The low toxicity of penicillin, its solubility in water, and its in vivo activity make it an ideal agent for combating disease caused by gram-positive bacteria. P. notatum represents an extremely variable group of organisms, some strains producing considerable penicillin, others producing little penicillin but large amounts of a second factor, designated as penatin or notatin. Some strains of a closely related fungus, P. chrysogenumy are also capable of producing penicillin that is apparently the same as the penicillin of P. notatum. The P. notatum-chrysogenum group of fungi is widely distributed in nature, having been isolated from different soils and from various moldy food products j however, only a few strains produce enough penicillin to justify their use for the commercial pro- duction of this substance. ■b3inn"iitN 2)3d siNvaomiiAj ni avons ivnaisaa C\J O H Penicillin /.H Penicillin fU Penicillin />H O.U./ml. O.U./ml. O.U./ml. O.U./ml. 832 Stock 35 7-9 65 8.1 61 8.3 40 8.5 832.A2 47 7-9 92 8.0 83 8.3 50 8.4 832.A,(6) 37 8.1 62 8.0 75 8.2 52 8.4 832.B3 29 8.2 37 7-7 42 8.1 35 8.4 From Raper and Alexander (764). TABLE 26. PRODUCTION OF PENICILLIN IN SURFACE CULTURE BY NRRL I249.B2I, 1950, 1978, AND TWO SUBSTRAINS OF 1 978, A AND B FOURTH DAY FIFTH DAY SIXTH : DAY SEVENTH DAY CULTURE Penicillin fn Penicillin fn Penicillin fn Penicillin fH O.U./ml. O.U./ml. O.U./ml. O.U./ml. I249.B2I 76 6.4 185 7.2 177 7-9 135 8.2 1950 Stock 98 7-3 139 7.7 103 8.0 81 8.3 1978 stock 120 6.9 233 7-4 114 7-9 162 8.3 1978A 109 7-3 154 7.6 131 8.0 85 8.3 1978B 124 6.9 262 7-3 246 7.8 190 8.2 From Raper and Alexander (764). FUNGI AS ANTAGONISTS □ su eFACE CULTURE SUBMERGED CULTURE D. 20 PENICILLIUM SUBSTRAIN Figure 15. Comparative production of penicillin by substrains of P. chry- sogenum 1951.B25 in surface culture and submerged culture. From Raper and Alexander (764). As a result of these studies the following conclusions were reached (763,764): 1. The capacity to produce penicillin as a metabolic product is a group- specific rather than a strain-specific character. 2. Different members of the P. notatum-chry so genum group vary greatly in their capacity to produce penicillin. 3. Special strains are particularly suited for certain types of penicillin production. For surface production of penicillin, no strain was found to be better than the original Fleming culture that has been freed from degenerate, ANTIBACTERIAL EFFECTS 139 mutant strains. This freeing of mutants must be carried out continu- ously, in order to avoid the degeneration of the culture. By strain se- lection and improvement of medium (addition of corn steep liquor, use of lactose), the penicillin yield of such cultures has been increased from ■2 to 6 to more than 200 O.U./ml. (662). For submerged cultures, how- ever, strains of P. notatum and P. chrysogenum are used which are not related to the Fleming strain. The best medium for tank production is about half the concentration of the nutrients used for surface culture. Pregerminated inocula are used {66'^). In a study of the metabolism of the penicillin-producing fungi (308a) it was established that the most important factors for high yields of penicillin are the nature of the culture, aeration, temperature, and proper balance among the nutrients of the medium, especially the relation between the carbon and nitrogen sources (591 )• Specific amino acids have an influence on the yield of penicillin, especially in sub- merged culture (375). Yields ranging from 90 to 900 units of peni- cillin were obtained under submerged conditions of growth. The rate of utilization of different sugars and oxygen consumption by penicillin-producing strains of a submerged culture is brought out in the following summary (519) : RATE OF UTILIZATION O2 UPTAKE SUGAR gm./l/hour ml./l/hour Lactose 0.32 109 Sucrose 0.46 150 Glucose 0.71 300 The lactose is more slowly utilized than the glucose, and less oxygen is required for penicillin production. The addition of boron to the me- dium favors lactose utilization and results in a higher rate of respiration and nitrogen utilization, with a less abundant mycelium, lower am- monia levels, and higher penicillin yields (520). At least four different penicillins have been isolated, namely, F, G, X, and K, two or more being found in the same culture broth. These penicillins differ in their chemical characteristics, in their antibacterial spectra, and also in their chemotherapeutic utilization. For example, penicillin X is more effective in the treatment of gonorrhea than G. 140 FUNGI AS ANTAGONISTS Various cocci are 6 to 8 times more sensitive to X than to commercial penicillin, which is largely G (695). Penicillin or penicillin-like substances are also produced by A. flavus, A. farasiticus (149), A. giganteus (722), and a variety of other fungi, largely species of Aspergillus and Penicillium, including A. niger, A. nidulans, A. oryzae, P. citreo-roseum (281), P. cms to sum (1039), and others, such as A. jlavifes ( 1005). Certain species of PenicilUum are also capable of producing other antibacterial substances, namely, citrinin, penicillic acid, and clavacin, the first of which is also produced by species of Aspergillus belonging to the candidus group (906). Atkinson (29) tested 68 cultures of PenicilUum and found that 18 possessed antibacterial properties. These cultures were divided into two groups : first, those largely active against gram-positive bacteria and producing substances like penicillin and citrinin 5 second, those active also against gram-negative bacteria and producing substances of the penicillic acid and penicidin types. Aspergillus -fiavus-oryzae Group The. A. oryzae members of this group possess only limited antagonis- tic properties. Many of the A. flavus strains, however, apparently have the property of producing at least two antibacterial substances when grown on suitable media and under suitable conditions. White and Hill (1006) isolated from cultures of a strain oi A. flavus grown on tryptone media a crystalline substance, aspergillic acid, that showed antibacterial activity against certain gram-negative as well as gram-positive bacteria. The substance was produced when the organism was grown on organic media, but not on synthetic. It was soluble in ether, alcohol, acetone, or acetic acid, but not in petroleum ether j it was soluble in dilute acid or alkaline aqueous solutions, and was precipitated by phosphotungstic acid. Aspergillic acid proved to have relatively high toxicity, and showed no protective action against hemolytic streptococci or pneumococci infections in mice. Glister isolated a culture (338) that also produced an antibacterial agent with a wide range of activity, both gram-positive and gram-nega- tive bacteria being inhibited by the culture filtrate. An extract was ob- ANTIBACTERIAL EFFECTS 141 tained that inhibited the growth of these bacteria in a dilution of ap- proximately 1 : 200,000. A. flavus was found (46) to produce frequent variants j two of these consistently gave far higher yields of aspergillic acid than those re- ported by White. The substance was found to have wide activity, especially against gram-positive cocci, but was less active against the anaerobes of gas gangrene and the gram-negative bacteria. Bush and Goth (103) isolated from A. flavus a second substance designated as flavicin. They grew the organism for 6 to 8 days on a nitrate-glucose medium containing 2 per cent corn steep. The filtrate was acidified to -pH 2.5 to 3.0 with phosphoric acid and extracted with purified isopropyl ether. The ether was treated with a slight ex- cess of o.2A^ NaHCO.j (5 to 10 cc. per liter of culture), giving a yield of 75 to 100 per cent of active material obtained. Purification was ob- tained by acidification of the NaHCOg extract with H3PO4 to ^H 2 to 3 and removal of the precipitate, the latter containing most of the toxicity (due no doubt to aspergillic acid) and the filtrate most of the activity. The filtrate was treated with ice-cold isopropyl ether, satu- rated with COo, washed with cold distilled water, and reextracted. The combined extracts were distilled at 0° C. to dryness under COo. A yel- low-orange glassy residue was obtained. It had a low toxicity and was active in vivo. The similarity to penicillin of the second antibiotic substance pro- duced by A. flavus has been definitely established (605, 606) by chemi- cal isolation and composition, solubility and stability, biological be- havior, low toxicity to animals, and therapeutic activity. A sodium salt assaying 240 O.U./mg. was obtained chromatographically and gave the following composition: 45.36 per cent C, 4.16 per cent H, 3.02 per cent N, and 13.36 per cent Na, [aj^ = + 108° (in water). Under submerged conditions, A . flavus thus produces two substances, one of the aspergillic acid type and the other of the penicillin type. Some strains produce little or no activity in submerged cultures, and most strains produce very little activity in stationary cultures. No ac- tivity is produced in synthetic media (950). The culture filtrate of A. flavus grown on lactose-peptone media was active against Af . tubercu- losis and other acid-fast bacteria in vitro. 142 FUNGI AS ANTAGONISTS Aspergillus jufnigatus Group Four antibacterial substances were isolated from strains of A . fumi- gatus: the two pigments, spinulosin and fumigatin (702), which are not selective in their action against bacteria, the colorless fumigacin that is active largely against gram-positive organisms (957), and glio- toxin (339, 631). Helvolic acid, isolated from a strain of A. jumigatus (126, 161), was found (631, 955) to be identical with purified fumi- gacin. Fumigacin is active against S. aureus in dilutions of i : 200,000 to 1 : 750,000 and is very stable. The pigment fumigatin, however, was said to deteriorate on standing, inhibition of S. aureus being reduced from 1 150,000 to 1 : 25,000 in 7 days. Fumigacin has a certain degree of resistance to high temperatures. Boiling in aqueous solution for 5 to 10 minutes reduced but did not destroy completely its activity. Heat- ing at 80° C. for 15 minutes reduced the activity only slightly. When fumigacin was dissolved in alcohol and precipitated by addition of nine volumes of water, the alcohol-water solution was found to contain 0.25 mg. per ml. A comparison of the antibacterial activity of fumigacin with that of the other substances produced by A. jumigatus is given in Table 27. A number of fungi, largely Aspergilli and usually members of the A . jumigatus group, have been found to be able to inhibit the growth TABLE 27. CHEMICAL PROPERTIES AND BACTERIOSTATIC ACTIVITY OF FOUR ANTIBIOTIC SUBSTANCES PRODUCED BY ASPERGILLUS FUMIGATUS MELTING CRYSTALLI- POINT SUBSTANCE ZATION ° C. FORMULA BACTERIOSTATIC ACTIVITY IN DILUTION UNITS B. sub- E. coli S. aureus tills Spinulosin Purplish-bronze plates 201 CgHgOg Fumigatin Maroon-colored needles 116 CgHgOg Fumigacin Very fine white needles 215-220 C32H44O8 Gliotoxin Elongated plates 195 C13H14O4N2S2 6,000 1,500,000 750,000 1,200 200,000 40,000 1,200 2,000,000 100,000 ANTIBACTERIAL EFFECTS 143 of M. tuberculosis. As pointed out previously, Vaudremer recorded in 1 913 (934) that the fungus produces a thermostable substance which is responsible for the antituberculosis effect. Zorzoli (1051) reported in 1940 that A. fumigatus produces a thermostable substance (100° C. for I hour) which interferes with the growth of M. tuberculosis. Ashes- hov and Strelitz (27) observed a marked action oi A. fumigatus prepa- rations upon the B.C.G. but not upon the avian strain of M. tuberculo- sis; the bacteriostatic activity was greater against M. tuberculosis B.C.G. than against staphylococci, although the bactericidal activity was lower. Culture filtrates and extracts of various unidentified fungi were found capable of inhibiting the growth of the organism (647). One such extract was designated as mycocidinj its effect upon the human tubercle bacillus was both bacteriostatic and bactericidal (328). Jen- nings (464) reported that helvolic acid (fumigacin), one of the anti- biotics produced by A. fumigatus y in concentrations of 1:10,000 in- hibited completely and in i : 100,000 only partly, the growth of the tuberculosis organism isolated from sputum. A . ustus produces in ordinary Czapek-Dox medium with 4 per cent glucose and o.i per cent yeast extract, after 14 to 19 days' incubation, a substance that inhibits the growth of M. tuberculosis and M. ranae (539). This antibiotic can be extracted from the medium with ether and other organic solvents. The ether residue is dissolved in phosphate buffer of fH. 1 1 .0. On acidification of the alkaline solvent, a yellow flocculent precipitate is obtained. This substance inhibited the growth of M. ranae in a dilution of 1:150,000, and the acid precipitate in 1 : 300,000 dilution. By means of a "countercurrent distribution" the active agent was separated into two crystalline and one partially crys- talline preparations (438). The mycelium of A. ustus was found (188) to contain a group of antibiotics, one of which was designated as ustin. This substance was active against gram-positive, including acid-fast, bacteria (1:500,000). It is inhibited by serum albumins and by lipids. Aspergillus clavatus Group This comprises a number of strains that produce highly active anti- biotic substances. By treating the culture filtrate with charcoal and 144 FUNGI AS ANTAGONISTS eluting the active substance with ether, Wiesner (1012) obtained a preparation having a bactericidal potency in dilutions of i : 100,000. This activity was not inhibited by serum, pus, or urine j strains of bac- teria that proved to be resistant to sulfonamides or mandelic acid were inhibited by this material. The active substance was designated (957) clavacin. It is active against E. colt and other gram-negative bacteria, as well as against gram-positive bacteria. It is different in this respect from fumigacin. Whereas the latter acts much more readily upon B. mycoides than B. subiilis, clavacin shows the opposite effect — greater activity against B. subtilis than against B. mycoides. Clavacin possesses a high bactericidal action, as compared with other antibiotic substances. A detailed study of its production by a variety of strains of A. clava- tus was made (968). The marked differences in the physiology of the different strains of A . clavatus were said to explain the differences in the production of clavacin by different strains. Those that change the re- action of the medium to alkaline, for instance, tend to inactivate the clavacin. Since clavacin is produced by a number of different fungi, it has re- ceived a number of designations, including patulin formed by P. fatu- lum (744), claviformin by P. claviforme (124, 125), and clavatin (47). It is also produced by strains of P. exfansum-y P. urticaey A. ter- reuSyA. giganteusy GymnoascuSy and others (24, 486, 501). For species of Penlcilliufn it was found (578) that glucose as a source of carbon, an incubation temperature of 20° C, stationary culture, and a source of iron offer optimum conditions. Trichoderma and Gliocladmm Grouf Certain strains of fungi of the genera Trichoderma and Gliocladium were found to exert a marked antagonistic action against various fungi and bacteria. An antibiotic substance designated as gliotoxin was iso- lated and found (82, 989) to be highly bactericidal. In order to produce this substance, the fungus is grown in a submerged condition in shake- cultures. An abundant supply of oxygen and a high acidity (/)H 5.0 or lower) are essential. Ammonium salts as nitrogen sources give better results than peptone or nitrates. Glucose and sucrose were found to be ANTIBACTERIAL EFFECTS 145 good carbon sources. It is of particular interest to note that whereas penicillin and flavicin are produced in media containing complex or- ganic materials as sources of nitrogen, fumigacin, clavacin, and glio- toxin are produced in synthetic media, the presence of complex nitrogen sources often being deleterious. Gliotoxin was isolated from the culture filtrate by the use of lipoid solvents, chloroform being most effective. Nonsterilized media ad- justed to fH. 2.5 to 3.0 could be used for large-scale production, the high acidity reducing the effect of contaminants (992). Gliotoxin is stable in neutral and acid solutions at room temperature j at alkaline reactions, it is very unstable, the rate of decomposition increasing with increasing alkalinity and temperature. At ^H 2.4, heating to 122° C. for 30 min- utes did not affect the active substance. With decreasing acidity, espe- cially at ^H 5.0, it became less thermostable. Gliotoxin is also produced by a number of other fungi, including P. obscurum {66$) a.nd A. fumigaius (631). Certain species of Trichoderma, including T. viridis, produce another antibiotic substance that is particularly active against fungi, designated as viridin (84). It is produced when the organism is grown in shallow layers of nitrate-containing media for 4 to 6 days at 25° C. j the cultures are characterized by a bright yellow color. It is isolated from the cul- ture filtrate by extraction with chloroform, evaporation, and recrystalli- zation from alcohol or benzene. It is stable only in acid solution. Fusarium Grouf The ability of species of Fusarium to produce antibiotic substances was first observed in a survey of the antibacterial properties of fungi, as pointed out above (p. 131). F. oxys forum was found (112) to pos- sess antibacterial properties. One of the organisms, namely F. javanl- cum', was studied in detail. A substance, designated as javanicin, was isolated (26) from the medium by the use of ether or benzene. It was removed from the solvent by extraction with aqueous NasCOs. It con- tained a quinone group but no carboxyl. It was active against gram- positive, including acid-fast, bacteria in concentrations of i : 50,000 to 1 : 400,000 but had little activity against gram-negative bacteria. It was relatively nontoxic. 146 FUNGI AS ANTAGONISTS Basidtomycetes The larger Basidiomycetes produce bacteriostatic substances that compare favorably with those formed by Aspergilli and Penicillia. The testing of the sporophore extract alone may be indicative, but it is not a fully reliable test for a positive result j the fungus must be cultured and a strip test made (1014). Of 700 species tested, about 70 gave a strong positive reaction and lOO a weak reaction against S. aureus and/or E. coU (1019). In a comparison of 72 genera, one or more species of 43 genera produced some antibiotic activity j none, however, was more active than P. notatum and none affected gram-negative forms (785, 785a). Polyporin, produced in the culture filtrate and in the sporophores of Polystktus sanguineus y is a thermostable substance not affected by ^H changes between 2.0 and 8.0. It passes through a Seitz filter, is not af- fected by body fluids, is nontoxic, and is active in vitro and in vivo against various gram-positive {S. aureus j S. viridans) and gram-nega- tive bacteria {E. ty^hosa, V. comma, etc.). Clitocybe gigantea var. Candida, a member of the Agaricus group, contains in its cell material a substance, designated as clitocybin, which is soluble in water, chloro- form, acetone, and ether. It is destroyed on heating at 70° to 80° C. It inhibits the growth of various gram-negative bacteria, such as E. coli, Ps. aeruginosa, E. tyfhosa, and Br. abortus, various gram-positive bac- teria, and M. tuberculosis. It is fairly toxic to animals: i gm. of the dry fungus substance is treated for 24 hours with 10 ml. water j i ml. of this extract will kill a 300 gm. guinea pig in 48 hours. It is effective in ar- resting the development of tuberculosis in guinea pigs (439). Several species of Cortinarius and one of Psalliota inhibited various gram-posi- tive and gram-negative bacteria (30). Other Groufs Various other fungi, including A . albus, A . niger, and Monilia albi- cans, were found (1051) to exert a marked antibacterial action against human and bovine tubercle bacteria 5 active filtrates were obtained, but the specific agents were not isolated. Certain dermatophytes, especially strains of Trichophyton mentagrofhytes, also produce an antibiotic ACTION AGAINST FUNGI 147 substance when grown in glucose-peptone media. This substance is simi- lar to penicillin in that it is favored by the addition of corn steep, and in its antibiotic spectrum, its sensitivity to reaction and temperature, and its destruction by penicillinase preparations (714). A number of unidentified molds have been reported to produce pig- ments which have antibiotic activity against various bacteria (807). This is true, for example, of P. c'mnah annus. The red pigment ex- tracted from the mycelium of this fungus inhibited S. aureus and S. pyogenes in a dilution of i : 5,000. The extract was slightly hemolytic, although not very toxic (637). A study of the distribution of antibiotic properties among the fungi revealed the fact that the Aspergilli and Penicillia are most active and the Phycomycetes least (Table 28). TABLE 28. DISTRIBUTION OF ANTAGONISTIC PROPERTIES AMONG THE FUNGI TOTAL PERCENTAGE ORDER OR NUMBER PERCENTAGE WEAKLY PERCENTAGE GENUS EXAMINED ACTIVE ACTIVE INACTIVE Phycomycetes 30 - - 100 Ascomycetes 20 - - 100 Aspergillus 150 30 20 50 Penicillium 200 20 30 50 Basidiomycetes 730 10 20 70 From Wilkins and Harris (1017, 1018, 1019). ANTAGONISTIC ACTION OF FUNGI AGAINST FUNGI Numerous fungi were found to exert antagonistic effects either against fungi belonging to the same species or against other fungi (Table 29). This phenomenon is particularly important in connection with the study of plant diseases. The effects are selective. The hyphae of Peziza will kill various Mucorales, whereas different species of As- fergillus and Penicillium are able to kill Peziza. A single spore of P. luteum was found capable of germinating in cultures of Citromyces [48 FUNGI AS ANTAGONISTS TABLE 29. ANTAGONISTIC INTERRELATIONSHIPS AMONG DIFFERENT FUNGI ANTAGONIST ORGANISMS AFFECTED REFERENCES Acrostalagmus s^. Rhizoctonia 990 Alternaria tenuis Ofhiobolus 89 A . clavatus Various fungi 949 A . fiavus Peziza 773 A . niger Peziza, Rhizoctonia 773. 933^990 Botrytis allii Monilia, Botrytis, etc. 933 Botrytis cinerea Rhizoctonia 990 Cefhalothecium roseum H elviinthosforium 359 Cunninghamdla elegans Monilia 933 Fusarium laieritium Rhizoctonia 990 Fusarium sp. Deuterofhofna 827 Gliocladium sp. H elminthosforium, Mucor, etc. 729 H elminthosforium sp. Colletotrichum, Fusarium, Botrytis, etc. 729 H. teres Fusarium, Ustilago, Helmintho- sforium, etc. 729 H. sativum Ofhiobolus 89 Mucor sp. Ofhiobolus, Mucor 89> 837 Penicillium sp. Peziza, Rhizoctonia, etc. 773 Penicillium sp. Ofhiobolus, Fusarium, etc. 89 Peziza sclerotiorum Mucor, Trichothecium, Dematium, 773 Peziza trifoliorum etc. Peziza 773 Sclerotium rolfsii H elminthosforium 729 SterigTnatocystis sp. A Iternaria 729 Thamnidium elegans Mucor 837 Torula suganii Asfergillus, Monascus, etc. 690 Torulosis sp. Blue-staining fungi 630 Trichoderma lignorum Rhizoctonia, Armillaria, Phy- tofhthora, etc. 989, 990 T. lignorum Rhizoctonia, Pythium, etc. 14^63,933 Verticil Hum sp. Rhizoctonia 990 From Novogrudsky (683). and of bringing about their destruction. P. luteum-furfurogenum pro- duces a thermostable substance, soluble in ether and in chloroform, that is antagonistic to the growth and acid production of A. niger (705). ACTION AGAINST FUNGI 149 Coniofhora cerehella was inhibited by a species of PenkilUumy its my- celium being considerably modified j however, in time the former or- ganism adapted itself to the latter and overgrew it, its rate of growth being eventually more rapid than that of a pure culture (380). Certain fungi are able to parasitize other fungi. The germination of the spores of one fungus may be reduced by the presence of spores of another {SS3)- Different fungi produce different types of fungistatic and fungicidal substances, some of which are stable, others unstable. These are formed particularly by the lower fungi or the molds, with the exception of the Phycomycetes that have so far not been found to produce any antibiotic substances. Their action consists in modifying or killing the mycelium of the other fungus, or merely in preventing spore germination. Brom- melhues (89), studying the effects of H. sativum and Penicillium sp. against Ofhiobolus graminis, emphasized that the inhibitory action was due to a toxic substance that was thermostable and diffusible in agar. In some cases, no relation could be observed between the acidity pro- duced by one organism and its ability to influence the growth of another ( 1046) i in other cases, as in the mutualistic effects of Sderotium rolfsii and Fusarium vasinfeaum, the first overgrew completely the second at f¥L 6.9, whereas in alkaline ranges the reverse took place (804). Random isolations oi Penicillium cultures and of other soil-inhabiting fungi were tested for their effects on the virulence of H. sativum on wheat seedlings grown in steam-sterilized soil (823). Some forms ex- erted a marked degree of suppression, some had no effect, and others increased the virulence of the pathogeny marked variations in activity were observed among the different species of Penicillium. Because Hyphomycetes were found to be capable of parasitizing the oospores of Pythium (196), Hyphomycetes were believed to serve as effective agents in promoting soil sanitation. Various species of Torulosis, in addition to certain bacteria, are capable of inhibiting the growth of Dematiaceae, fungi that cause the blue staining of wood pulp (630). A species of Penicillium (P. gladioli) was found (8ia) to produce an antibiotic (gladiolic acid) which is actively fungistatic but only weakly bacteriostatic. Certain fungi may affect the reproduction of others. Melanosfora 150 FUNGI AS ANTAGONISTS -pamfeana, for example, normally does not form any perithecia in cul- ture but is able to do so in the presence of Basis-porium gallarum or Fusarium monilijorme. This effect was ascribed to a special substance that resists heating at i io° C. Different fungi have a special influence on the germination of spores of various ascomycetes and of other fungi (28, 776), these effects being characteristic of the antagonists. The edible mushroom Psalliota camfestris exerts a definite antago- nism against the parasitic fungus Mycogone (135). This phenomenon has been looked upon as a case of antibody formation. Species of Fusch rium are able to antagonize the mushroom fungus 5 however, an actively growing culture of the latter may become antagonistic to the former (1026). In the destruction of paper pulp by fungi, a marked antago- nism was shown (341) to take place between different organisms, especially by Tr'ichoderma lignorum against various species of Fusa- rium and other fungi, as illustrated in Figures 16 and 17. Certain species of Trkhoderma and Gliocladium are able to inhibit the growth of various plant pathogenic fungi, especially R. solani, as well as of Blastomycoides dermatitidis , a causative agent of human skin diseases. The active substance, gliotoxin, is liberated during the early stages of growth. The mycelium of older cultures contains another sub- stance that is soluble in acetone j this has only an inhibiting effect and is not fungicidal as is gliotoxin. The fungicidal effect of gliotoxin upon the germinating spores of Sclera tinia americana and hyphae of R. solani was found to be greater than that of CUSO4 and less than that of HgCU. Various other fungi are able to exert antagonistic effects against plant pathogens. T. lignorum and A . niger restricted the growth of the fungi Macrofhomina fhaseoli and R. solaniy which produce cotton root rot, and reduced the activity of the filtrates of the pathogens causing wilting of the plants. Satoh (826) has shown that Ofhiobolus m^iyabeanus produces both growth-promoting and growth-retarding substances, the first of which is heat stable and passes through a Chamberland filter j the second is inactivated at 100° C. and does not pass through a filter. The formation of two substances by Torula suganii, both of which were thermostable, however, was also demonstrated (690). Figure i6. Antagonistic effect of one fungus, Ps. "zonatum (in center), upon another, T. I'lgnorum. From Goidanich et al. (341). Figure 17. Attack of an antagonistic fungus. T. llgnorum, upon another fungus, F. sambiicinum (in center). From Goidanich et al. (341). ACTION OF BACTERIA AGAINST FUNGI 151 ANTAGONISTIC EFFECTS OF BACTERIA AND ACTINOMYCETES AGAINST FUNGI Various bacteria and actinomycetes have marked selective fungistatic and fungicidal effects (Table 30). Bacteria active against U. zeae were isolated from corn, these bacteria being capable of destroying the colo- nies of the smut fungi. The widespread distribution of such bacteria in the soil was believed to check the multiplication of the pathogenic fungi. Four types of bacteria antagonistic to smuts and to certain other fungi have been described (470). Some of these bacteria produce en- zymes that are able to dissolve the chemical constituents of the cell walls of the fungus sporidia; they were also found to be active in the soil against the specific fungi. Brown (93) observed that H. sativum and a certain bacterium produced thermostable, mutually mhibitmg substances. The bacterium as well as its metabolic products inhibited the TABLE 30. ANTAGONISTIC EFFECTS OF BACTERIA AGAINST FUNGI ANTAGONIST Achromobacter %^. Al. faecalis Bacillus "/)" B. anthracis B. mesentericus B. mycoides B. simflex B. subtilis Bacterium sp. Bacterium sp. Myxobacterium P. vulgaris Ps. aeruginosa Ps. juglandis Ps. fhaseoli Ps. translucens Ps. vulgaris S. marcescens Spore-forming bacteri From Novogrudsky (683). ORGANISMS AFFECTED REFERENCES Fusariumy Sclerotinia 143 Helminthosforium 729 Ustilago, Penicillium 35 S. cerevisiae 525 Helminthosforium 142.729 Helminthosforium 729 Rhizoctonia 154 Cefhalothecium roseum II Fusarium, Sclerotinia, etc. 729 Ustilago 470 Ustilago 247, 470 Basisforum, Phytofhthora, etc. 506,729 Saccharomyces 525 Dothiorella 247 Fusarium 63,247 Ofhiobolus 87 Ofhiobolus Beauveria, etc. 87 10, 11, 12,624 a Fungi 35>729 152 FUNGI AS ANTAGONISTS growth not only of the particular fungus but also of other members of the same genus, but not of Fusarium- conglutinans . These bacteria pro- duced a diffusible agent that inhibited the growth of H. sativum (i 15a). The active substance was not destroyed by autoclaving^ it dif- fused into fresh agar and water, producing "stale water" that was in- hibitory to the fungus. Chudiakov (143) isolated from the soil two bacteria that were capable of bringing about the lysis of different species of Fusarium as well as other fungi. These bacteria were found to be widely distributed in most soils J they were absent, however, in flax-sick soils, in spite of the abun- dance of Fusarium. When this fungus was added to the soil containing antagonistic bacteria, it did not develop, and the plants did not become diseased. The antagonistic action of a variety of other bacteria against plant pathogenic fungi has been definitely established, as in the case of B. simflex against Rhizoctoniay P. vulgaris against Phyio-phthora (488), and B. mesenlericus against H elminthosforiuin (142). B. sim- plex was grown (491) for 7 days at 28° C. in potato-dextrose medium containing i per cent peptone, and the active substance was removed by charcoal and dissolved in alcohol. Different fungi differed in the de- gree of tolerance to this substance. The majority were repressed when 10 per cent concentration of the stale medium was added to fresh medium. The ability to produce a thermostable substance toxic to the plant- disease-producing fungus Rhizoctonia is widespread among spore-form- ing bacteria. The toxic substance is insoluble in ether, chloroform, and benzol, but is soluble in ethyl alcohol. It passes through collodion, cellophane, and parchment membranes. It is readily destroyed on boil- ing in alkaline media but is more resistant in acid media. Nakhimovskaia (672) found that various bacteria are able to inhibit the germination of rust spores. Nonspore-forming bacteria, such as Ps. ■fluorescens and S. marcescenSj prevented the germination of the spores of Ustilaga avenaey Ustilaga hordeiy Ustilaga nuda^ and Ustilaga reae. Spore-forming bacteria, including B. mycoides and B. 'mesenlericus ^ as well as sarcinae (5. ureae, S. lutea), exerted no antagonistic action on the rust spores. The presence of these bacteria, however, influenced the nature of the germination of the spores, which gave rise to mycelium- ACTION OF BACTERIA AGAINST FUNGI 153 like forms with great numbers of copulating filaments, whereas in the control cultures yeast-like forms prevailed and copulating cells were rarely encountered. The presence of a certain concentration of bacterial cell substance was essential to this antagonistic effect. With a more lim- ited amount of cell material, the bacteria ceased to inhibit the germina- tion of the spores but influenced the germination process in the same manner as do nonantagonistic bacteria, that is, they stimulated the sex- ual process. An increase in concentration of cell substance, even of non- antagonistic organisms, would inhibit spore germination. The common occurrence of the fungus Pyronema conjiuens in freshly burned-over soils, but not in natural soils, was explained (684) as due to the destruction of the bacterial antagonists by heating of the soil. Ps. fuorescens was particularly effective as an antagonizing agent. A com- parative study of the fungistatic action of substances of bacterial origin (883) has shown these to be more active than common disinfectants. Tyrothricin inhibited the growth of animal pathogens in dilutions of 1:5,000 to 1:20,000, pyocyanin in 1:2,000 to 1:5,000, and hemi- pyocyanin in i : 20,000 to i : 60,000. Actinomycetes may also exert a marked depressive effect upon the growth of fungi. The active substances produced by these organisms show considerable selective action just as in the case of the bacteria. Actinomycin was found (974) to inhibit the growth of Penicilliuniy Aspergillus J Ceratosiomella, and yeasts in concentrations of i :50,000 j larger amounts (1:10,000) were required to inhibit other fungi, in- cluding Rhizofus and Trichoderma. Streptothricin is less effective against fungi, although it inhibits the growth of certain yeasts (1031). In general, antibiotics vary as much in their antifungal as in their antibacterial effects. Some, like gliotoxin and actinomycin, were found to be highly active against both parasitic and saprophytic fungi, whereas others, like chaetomin and streptomycin, had little if any activity. Since some of the substances, like actinomycin, have a highly toxic effect upon animal tissues, the selection of a suitable antifungal agent for chemotherapeutic purposes is limited to a very few promising mate- rials j among these gliotoxin and streptothricin were mentioned (771). CHAPTER 8 MICROSCOPIC ANIMAL FORMS AS ANTAGONISTS The microscopic animal world inhabiting the soil and water basins com- prises protozoa, insects and insect larvae, nematodes and other worms. Their relationships to the microbiological flora of soils and waters are varied. Many, if not most, of these animals feed upon the bacteria and fungi, as well as upon the smaller animal forms. Some carry a bacterial population in their digestive tract and appear to depend upon these bacteria for some of the digestion processes. Many of the animal forms are parasitized by bacteria and fungi. Some of these forms are subject to the action of specific substances produced by microbial antagonists. No detailed discussion will be presented of these varied relationships, but attention will be directed to a few specific phenomena which have a bearing on the subject under consideration. The ability of higher ani- mals to produce antibacterial substances has been amply established. Some of these substances are well characterized, as in the case of ly- sozyme found in mammalian tissues and secretions and inhibins found in fresh human urine (189). RELATIONS OF PROTOZOA TO BACTERIA The lower animal forms inhabiting the soil, manure piles, and water basins often utilize bacteria in the synthesis of their foodstuffs. Al- though many of the smallest organisms, namely the protozoa, are able to obtain their nutrients from simple organic compounds and mineral salts, they frequently depend upon the bacteria to concentrate the nu- trients present in dilute forms in the natural substrate. It has been shown (106), for example, that when carbohydrates are present in water in very low concentration, the protozoa may not be able to use them in that dilute formj however, the bacteria can assimilate these carbohydrates and can build up extensive cell substance, and the pro- tozoa are then able to multiply by consuming the bacteria. Protozoa are apparently also able to destroy pathogenic bacteria (781). RELATIONS OF PROTOZOA TO BACTERIA 155 The fact that some of the protozoa feed upon bacteria served as the basis for a theory designated as the "protozoan theory of soil fertility" (812). According to this theory, the capacity of protozoa to consume bacteria is responsible for the limited fertility of certain soils. The bac- teria were viewed as the sole agents responsible for the liberation of nutrients in the decomposition of soil organic matter and for the trans- formation of these nutrients into forms available to higher plants. The protozoa, because of their capacity to digest bacteria, were looked upon, therefore, as the agents injurious to soil fertility. The increased fer- tility which results from the treatment of soil with heat and with cer- tain chemicals was believed to be due to the destruction of the protozoa, considered as the "natural enemies of the bacteria." Subsequent investigations did not support this theory. When proto- zoa were added to cultures of bacteria responsible for certain specific processes they did not exert any detrimental effect upon the particular reactions brought about by the bacteria, despite the fact that they fed upon and thereby considerably reduced the numbers of these bacteria. In many cases, the effect of protozoa upon bacterial activities may actu- ally be considered beneficial (163). This was found to be true for such processes as the fixation of atmospheric nitrogen, the liberation of ammonia from proteins, and the formation of carbon dioxide from car- bohydrates. Failure to confirm the protozoan theory of soil fertility was due pri- marily to the fact that several assumptions were made that were not fully justified, namely, (a) that bacteria are the only important soil or- ganisms responsible for the decomposition of the soil organic matter j (b) that protozoa, by consuming some of these bacteria, are capable of restricting bacterial development and, if so facto, organic matter de- composition. The fact was overlooked that the soil harbors, in addition to the bacteria, many fungi and actinomycetes capable of bringing about the decomposition of plant and animal residues, resulting in the liberation of ammonia, and that this could take place even if all the bac- teria were completely eliminated from the soil. The favorable effect of partial sterilization of soil upon fertility still remains to be explained. Various other theories have been proposed, the most logical of which is one based upon a soil condition designated as 156 MICROSCOPIC ANIMAL FORMS AS ANTAGONISTS "microbiological equilibrium" (972). It has also been suggested (527) that the phenomenon is due to the disappearance of the bacterial antago- nists in the soil as a result of partial sterilization. In many cases, however, protozoa are responsible for bringing about extensive destruction of bacteria. This may find a practical application in the purification of water and sewage. The action of the protozoa is due in this case to the actual ingestion of the bacteria (452). The theory that protozoa may favor soil processes because of the stimulation of bacterial development and hence the accelerated trans- formation of soil materials is not always justified. The assumption is usually made that these processes take place in the soil in a manner simi- lar to those brought about in artificial culture media, a generalization that may be justified only in very special cases. No consideration is given to the fact that the presence of numerous other organisms in the soil may modify considerably the activities of the protozoa. The use of arti- ficial media gives only a one-sided conception of the significance of pro- tozoa in soil processes. Although the more recent claim concerning the function of protozoa in the soil is based upon more direct experimental evidence, it is still inadequate, because it gives insufficient consideration to the numerous elements involved in the complex soil population. The protozoa make up only a small portion of the soil population, both in numbers and in the actual amount of cell substance synthesized. Their ability to reduce bacterial numbers in normal soil is not very sig- nificant. The indirect method of studying protozoa in solution media, whereby the types observed and the activities obtained are quite differ- ent from those occurring in the natural soil, has been largely responsible for the exaggerated importance attached to these organisms. One may conclude that the protozoa, by consuming some of the bac- teria, keep these organisms at a high state of efficiency, thus assisting in the breakdown of the plant and animal residues in the soil. In other words, the rate of energy transformation brought about by bacteria and even the total amount of change produced in the substrate are increased by the presence of protozoa. Thus, an interrelationship among micro- organisms which was at first thought to be antagonistic actually has proved to be associative. The protozoan Oikomonas termo was found to be capable of living at the expense of a large number of bacteria. RELATIONS OF PROTOZOA TO BACTERIA 157 namely 83 per cent of those tested. The fact that Oikomonas causes many species of bacteria to flocculate was suggested as explanation for the ability of the protozoa to digest these bacteria (381). The ability of protozoa to destroy bacteria was said (426) to be re- sponsible for the protection of certain plants against attack by plant pathogenic bacteria and fungi. This was said to hold true of attack of potatoes by Bacterium aroideae and of other plants by Pseudomonas hyacinthi and Pseudomonas ckri, as well as by species of Fusarium and Penicill'ium,. Various bacteria may exert a toxic action upon protozoa, thus limiting the development or bringing about the destruction of the latter (133, 584). Certain plant pathogenic bacteria inedible by amebae were found to produce a toxin that was harmful to these amebae. In some cases, the protozoa were capable of developing a certain resistance to specific bacterial products (721). The toxic action of some bacteria against Paramecium could be overcome by the presence of a flagellated proto- zoan Oikomonas (382). On the basis of the ability of protozoa to utilize bacteria as food, Singh classified (855) the latter into 3 groups: (a) edible forms, (b) inedible but harmless to protozoa, (c) forms toxic to protozoa. Pig- ment-producing bacteria are inedible and some are toxic j these comprise the Ps. aeruginosa and the S. m^arcescens groups. Since some amebae, like Hartmanella castellanii, function as phago- cytes, they are believed (545) to offer excellent material for the study of the effect of antibiotic substances upon pathogenic bacteria in the presence of these amebae, the latter not being affected, as by penicillin, for example. Certain factors in the medium seem to affect the encystment of pro- tozoa (9C0) J it remains to be determined to what extent these factors can be classified with antibiotic substances. Myxamoebae of the slime mold Dictyostileum discoideum also live upon bacteria. They are able to utilize the gram-negative somewhat better than the gram-positive types, with certain few exceptions. Bac- terial spores are also ingested by these organisms, but they are not di- gested. 158 MICROSCOPIC ANIMAL FORMS AS ANTAGONISTS RELATIONS OF PROTOZOA TO FUNGI The presence of Colfoda and other infusoria in an active form was found to repress the growth of VerticilUum dahUae in culture media and to prevent infection of tomato plants by this pathogen j Colfoda was also active in soils and reduced the incidence of wilting (88). The ability of various fungi to destroy protozoa and nematodes has been studied in detail by Drechsler (194, 195). MALARIAL AND TRYPANOSOME PARASITES In connection with the recent interest in antibiotic substances, con- siderable work has also been done on the effect of these substances upon different strains of Plasmodium causing malaria and upon different trypanosomes causing various tropical diseases. Weinman found (993) that the general correlation between the gram-stain of bacteria and their sensitivity to gramicidin also extends to protozoa (Leishmania, Trypanosoma) and to the Leftosfira tested. Tyrocidine had a marked effect, in concentration of 5 pg/ml., upon the flagellates j they remained active for many hours, gradually losing their motility j a few escaped, giving rise to delayed growth. Levaditi and Twort (561) demonstrated that trypanosomes are de- stroyed by B. sub tills and are also partly destroyed by E. coli, but not by B. frodigiosus, B. m^esentericus, B. fyocyaneus. The active substance, designated as trypanotoxin, was found to be produced by B. subtilis in the culture filtrates and in centrifugates. The washed cells of the or- ganism were inactive. The substance is thermolabile and is destroyed at 70° C. in 20 minutes. It does not pass collodion dialysis membranes. It is also active in high concentrations against the tic-fever Sfirillum and Leishmania J but not against Borrelia gallinarum. It is apparently not very active in vivo, since it did not protect mice against trypanosomes. Contact between trypanotoxin and trypanosomes in vitro led to the de- velopment of toxo-resistant strains of the latter. This resistance was maintained for many generations j however, the new strains do not be- come more resistant to pyocyanase and other anti-trypanosome re- agents. Further studies (560) brought out the following facts: resistant MICROBIAL CONTROL OF INSECT DISEASES 159 strains did not adsorb the toxin, as did the susceptible strains} the susceptible trypanosomes were destroyed completely by antiserum, whereas the resistant forms were also resistant to this antiserum. A lipid-like substance produced by species of Phycomyces was ac- tive against Tryfanosoma equiferdum in vitro but not in vivo (830a). MICROBIAL CONTROL OF INSECT DISEASES Insects are subject to attack by various groups of microorganisms, including bacteria, fungi, protozoa, nematodes, and other insects. Many attempts have been made to control insect pests by the use of pure or mixed cultures of microorganisms. In this connection the following re- lationships must be considered: the receptivity of the insect to microbial attack during its various stages of development} the environmental conditions favoring the attack on the insect by the disease-producing organism; the influence of environment upon the virulence of the at- tacking microbe; the manner in which the parasite attacks the host; the coordination of the optimum activity of the disease-producing agent with the abundance of the host and the proper stage of its develop- ment. The microbial agents that keep in check the spread of insects, some of which are highly injurious to plants and animals, are far more impor- tant than any other methods of control. These microbial agents can be classified into three groups, depending upon the nature of the host: (a) microbes that attack economically useful insects and that must be con- trolled in order to avoid important losses from disease; (b) microbes that attack injurious insects and that must therefore be favored and en- couraged; (c) microbial agents infectious to plants, animals, and man that are spread by insects. Various bacterial diseases that formerly caused considerable destruc- tion of silkworms and bees have been controlled, once the nature of the organisms concerned was established. One of Pasteur's important con- tributions to microbiology was the control of Flacheria among silk- worms. However, most of the problems of control of injurious insects have been difficult to solve. A great number of bacterial, fungus, and virus diseases of insects are now known, but the many attempts to em- 160 MICROSCOPIC ANIMAL FORMS AS ANTAGONISTS ploy these pathogens in combating the insect hosts have not alwaj's been successful. The investigations so far carried out in this important field may be considered as at a very primitive stage. Metalnikoff (634) compared the bacterial treatment of caterpillars of Pect'mophora gossypiella with the action of arsenical poisoning. The dry spores of Bacterhrrn efhest'iae, Bacterium gelechiaey Bacterium 5, and Bacterium cazaubon, in powder form, were mixed with water at the rate of i to 4 ounces to 2^-2 gallons of water, with the addition of 4 per cent of molasses J this preparation was sprayed on the plants t^^'o to four times, at regular intervals, at the rate of 196 gallons or less per acre. The best results were obtained for plants treated with B. ephestiaej the infestation being reduced by about 50 per cent as compared with the controls. A slightly smaller reduction occurred on plots sprayed with B. cazauborij while B. gelechiae reduced the infestation by less than 40 per cent. Those plants that were treated with the arsenical spray showed a reduction of only 18 per cent. Recently microorganisms have been used for the control of the larvae of Japanese and other beetles in the soil. A variety of bacteria, fungi, and nematodes were found capable of destro\-ing these larvae. Once the attacking microorganisms have become established in the soil, the larvae and the beetles themselves tend to disappear. Glaser {'^2)5) utilized for this purpose Neoaflectana glaseri. This parasite possesses great repro- ductive capacity and is capable of destro}-ing large numbers of grubs. Glaser demonstrated the presence of this nematode also in localities where the grub was not present. Dutk\' (222) described two spore-forming bacteria {Bacillus fo- filUae and Bacillus lentimorbus) which cause the milky disease of the larvae of the Japanese beetle. These bacteria are grown in the larvae and then inoculated into soil. They are capable of infecting the grub, and are said to be responsible for the reduction in the beetle population. Bacteria pathogenic to the citrus red scale have also been isolated from the soil (868). Fungi have also been utilized for the control of insects. Sweetman (891 ) emphasized the importance of entomogenous fungi as destructive enemies of insects. A limitation to their practical importance in the fight against insects is that the fungi require special conditions for develop- MICROBIAL CONTROL OF INSECT DISEASES 161 ment, especially high humidity and favorable temperature, which are not always found under natural conditions. Glasgow (22^ established that some of the caecal bacteria of Het- eroptera show a marked antagonism toward other bacteria and proto- zoan parasites that occur in the intestines of these insects. The caecal system of the insects was removed and dropped into nutrient bouillon, where it remained for a month or more without showing any bacterial growth. This was believed to be proof of the fact that the caecal bac- teria are antagonistic to ordinary saprophytic and parasitic bacteria and prevent their development} also they apparently kill these bacteria when they invade the alimentary canal of the insect. According to Duncan (216), the bactericidal principle found in dif- ferent insects and ticks shows differences in regard to the types of bac- teria affected and the degree of their susceptibility. The gut-contents of Argas and Stomoxys show the widest range of action j that of bugs, the least. Spore-forming bacteria are especially affected by material from Stomoxys, whereas staphylococci appear to be more susceptible to the action of Argas material. The gut-contents of ticks was found to have a weak activity upon P. festis, whereas the contents of certain in- sects favored the growth of the latter. This phenomenon may have a bearing upon the function of the plague flea. The action of the lethal principle is greater and more rapid at 37° C. than at room temperature. The lethal principle has been found to be active for at least six months when kept in a dry state. It is thermostable, resisting temperatures as high as 120° C, and is not destroyed by proteolytic enzymes. It appears to be bound to proteins, since it is precipitated from solution by alcohol and acetone, but it is not affected by these reagents. It is insoluble in the common fat solvents. It becomes inactivated when allowed to act upon bacteria and appears to be adsorbed by killed bacteria, even by species that are not destroyed by it. This substance does not have the properties of either bacteriophage or lysozyme. The presence in certain insects of a variety of other substances, such as allantoin, which affect bacterial activities has also been established. These observations give rise to the hope that man may in time succeed in developing and utilizing microorganisms for the biological control of injurious insects (881). 162 MICROSCOPIC ANIMAL FORMS AS ANTAGONISTS RELATION OF NEMATODES TO SOIL MICROORGANISMS Nematode worms are represented in the soil by a number of sapro- phytes as well as by many plant and animal parasites. The latter vary greatly in their relation to the host. The larvae of the cereal parasite Tylenchus tritici penetrate the wheat seedlings between the leaf sheaths, near the growing or apical points. When the head is formed, the larvae enter the flowering parts and form galls. They become sexu- ally mature, mate, and lay eggs which hatch in the galls, and then be- come dormant. When the galls fall to the ground and decompose, the larvae are liberated and proceed to find and attack new plants. Other nematodes attack plants by feeding upon the roots. The methods of control require, therefore, a knowledge of their life history. Some species produce resistant forms or cysts that may survive in the soil for many years, even in the absence of the host plant. Soil sterilization by steam or by chemicals is frequently employed as a measure of nematode extermination. Antagonistic relationships may be utilized for the control of nema- todes. Linford et al. (572) found that the root-knot nematode of pine- apple {Heterodera marioni) may be controlled by heavy applications of organic material. The decomposition of this material results in a greatly increased population of saprophytic nematodes in the soil. The decom- posed organic residues also support large numbers of such other soil microorganisms destructive to the parasitic nematodes as the nema- capturing fungi (180, 196), the nontrapping fungal parasites, the predaceous nematodes, the predaceous mites, and different bacteria ca- pable of destroying nematodes. BACTERICIDAL ACTION OF MAGGOTS Surgical maggots are said to have a bactericidal effect in wounds, in addition to removing necrotic debris. The presence of an active bacteri- cidal substance which is thermostable and active against 5. aureus^ hemolytic streptococci, and CI. welchii has been demonstrated (854) in the maggot LuciUa serkata. CHAPTER 9 ANTAGONISTIC RELATIONSHIPS BETWEEN MICROORGANISMS, VIRUSES, AND OTHER NONSPECIFIC PATHOGENIC FORMS Antagonistic phenomena in relation to viruses have been but little in- vestigated. It has been established, however, that certain microorgan- isms are capable of destroying viruses, and particularly that some vi- ruses possess the capacity of antagonizing other viruses. The rapid in- activation of poliomyelitis virus in the process of aeration of sewage sludge has also been indicated ( 1 1 1 ) . BACTERIA AND VIRUSES The ability of certain strains of B. subtilis to inactivate the virus of rabies has long been recognized. The activity was found to be due to a substance produced in the culture filtrate j limited experimental evi- dence pointed to the effectiveness of this substance not only in vitro but also in vivo (619). When a mixture of the culture filtrate of B. subtilis and the virus was injected into rabbits, the activity of the virus was sup- pressed (619). It has been suggested (774a) that this action upon the virus of rabies and of equine encephalitis is due not to a true antibiotic but to a proteinase similar to the one which destroys bacterial toxins. An inactivating effect of B. subtilis upon the virus of vesicular stoma- titis as well as staphylococcus phage, when in contact with them for 1 5 to 18 hours at 35° C, was also reported (750). This phenomenon has been explained as due to the process of adsorption. The facts that it is selective, that the phage cannot be reactivated, and that the virus is rendered impotent by the action of the bacterium also point to a pos- sible antagonistic effect. However, different specific antibiotics, includ- ing penicillin, tyrothricin, and subtilin, when used either alone or in combination with sulfonamides or acridine, have failed to prevent in- fection of mice with influenza virus (537). A "nontoxic" inactivator has been defined (307) as a substance that 164 ANTAGONISMS BETWEEN NONSPECIFIC PATHOGENS inactivates plant viruses and is not detrimental to most forms of life. Various microorganisms are capable of producing such inactivators. Plant viruses differ in their sensitivity to "nontoxic" inactivators. Ac- cording to Johnson (473) various microorganisms are capable of form- ing such inactivators against tobacco-mosaic virus. The inactivators pro- duced by A. aero genes and A. niger are particularly effective against a variety of plant viruses, but not against all of themj the inactivators produced by the two organisms appear to be similar. They are com- paratively heat stable but are slowly destroyed by certain organisms. They can be concentrated by evaporation of medium. A substance which was capable of rapidly inactivating the tobacco-mosaic virus was isolated (895) from yeast. A chemical reaction between the inactivating principle and the virus was therefore suggested. The inactivator in this instance was destroyed by heating with i TV NaOH solution, but not by 2 N HCl. It was not a protein and gave on analysis 39.7 per cent C and 5.85 per cent H. The substance was said to be a polysaccharide. A. niger was also found (307) to form in the medium a substance capable of in- activating a number of different plant viruses; the effect of the inactiva- tor was found to be exerted upon the virus itself and not upon the plant. Of 150 organisms, comprising bacteria, fungi, and actinomycetes, isolated from different natural substrates as well as from soil enriched with virus concentrates, only three showed some inactivation of the fowl pox virus, and, in one case, of the laryngotracheitis virus. The active principle of one of these organisms was actinomycin, an anti- bacterial substance known to be highly toxic to animals (477). ANTIBIOTIC SUBSTANCES, VIRUSES, AND PHAGES The first recorded observation on the effect of antibiotics upon vi- ruses is that of Fukuhara (304) who demonstrated that pyocyanase, after having been in contact with the viruses of vaccinia, rabies, and chicken pest, brought about their inactivation, as shown by the fact that when viruses so treated were inoculated into experimental animals the respective diseases did not develop. Most of the viruses, however, appear to be resistant to the action of ANTIBIOTICS, VIRUSES, AND PHAGES 165 antibiotics J this was found to be true of penicillin and clavacin against fowl pox inoculated into the chorioallantoic membrane of the chick embryo (784). Penicillin was also found (707) to be without effect on the virus of vaccinia, encephalitis, and equine encephalonigelitisj how- ever, it had an effect, when used in large doses, on the course of infec- tion of chick embryos with psittacosis and meningopneumonitis. The possible effect of other antibiotics, such as aspergillin, upon certain viruses has also been indicated (375). In a study of phage inactivation, it was found that streptothricin, streptomycin, and clavacin exerted an effect, whereas penicillin and actinomycin did not. There was no correlation between the suscepti- bility of the host cells and that of the phage to an antibiotic agent. In the case of E. colt host and phage, a concentration of the antibiotic great enough to inactivate all the viable cells showed progressive decrease in 24 hours of phage added to such mixtures. With lower concentra- tions of the antibiotic, the phage multiplied only when the cells were increasing. Phage in suspensions of streptomycin-treated cells was not TABLE 31. EFFECT OF PENICILLIN AND STREPTOMYCIN ON S. AUREUS PHAGE AND ITS HOST. RESULTS X 10° BACTERIAL CELLs/mL. plaques/ml. AFTER TIME SPECIFIED AFTER TIME SPECI- TREATMENT UNI Ts/m L. AT 37° c.^ FIED AT 37° c. 3 24 48 3 24 48 hours hours hours hours hours hours Culture control 320 3300 570 Cells -\- streptomycin 2 .01 4.25 4200 Cells + penicillin 10 .15 .275 55 Cells + phage o .01 4700 Phage + broth 7.5 3-7 .001 Phage -\- streptomycin 2 120 .04 Phage -\- penicillin lO 100 .98 Cells + phage -f- streptomycin 2 .01 .001 ,294 .41 .2 Cells + phage + penicillin 10 .09 .001 ,025 From Jones (476). * Number of cells at start, 166 ANTAGONISMS BETWEEN NONSPECIFIC PATHOGENS reactivated by dilution after prolonged incubation (Table 31). Peni- cillin and streptomycin acting on S. aureus phage and its host, at concen- trations of the substances which had no destructive effect on the phage alone, showed that no reduction of the phage occurred when placed in the presence of penicillin-treated cells, whereas a definite decrease took place in the case of streptomycin-treated cells (476). A mixture of phage and penicillin caused more rapid killing and lysis of staphylococci than either alone, thus indicating that the penicillin- resistant organisms were killed by the phage and vice versa. Penicillin itself did not affect phage multiplication and did not interfere with its lytic action (425). The formation of antiphage agents can be studied by a group of methods, making use of the phage agar plate, phage streak, and agar- diffusion or cup tests. Growth of the antagonist upon the phage-seeded agar, or the diffusion of the antiphage agent into the agar, is followed by flooding the surface with host-seeded agar. Antiphage action is in- dicated by a reduced number of plaques or by a zone of bacterial growth surrounding either the antagonist or the cup containing the antiphage substance (466). The use of antibiotics in combating true viruses has so far given only little encouragement. However, the inhibition of growth of typhus rickettsiae by penicillin has been established (361). RELATIONSHIPS AMONG VIRUSES The cultivation of influenza virus in a simple tissue-culture was found (20) to render the culture unable to support the growth of a biologically distinct strain of the virus added 24 hours later. The tissue- culture, however, was still capable of supporting multiplication of a related virus such as that of lymphogranuloma venereum. When two strains of the influenza virus were added to the tissue-culture simul- taneously, the one added in larger concentration suppressed the growth of the other. Numerous reports have been made concerning the interference of one virus by another, and even of inactivated bacteriophage with the active agent of the same strain (1047, 1048). Henle and Henle (404) have RELATIONSHIPS AMONG VIRUSES 167 shown that even an inactivated virus, whether a homologous or a heterologous strain, is capable of suppressing the development of the influenza virus. Jungeblut and Sanders (483) suggested that poliomyelitis in ani- mals may be aborted by the injection of another virus. A strong antago- nism was observed between a murine virus mutant (virus passed through mice for many generations) and the parent strain of the virus. The murine virus was capable of counteracting large paralytic doses of poliomyelitis j the two viruses virtually counterbalanced each other. Other types of antagonism between viruses include that of canine dis- temper or lymphocytic chorio-meningitis virus against experimental poliomyelitis (169). An intramuscular injection of a neurotropic strain of yellow fever virus was found to protect animals against simultaneous infection with a highly pathogenic viscerotropic strain (447). The an- tagonistic agent was believed to be a chemical substance produced by the murine virus, for which the term "poliomyelitis inhibition" was pro- posed by Jungeblut. The "interference phenomenon" of two viruses can be used to advantage in bringing about immunity reactions. The suppression of one strain of yellow fever virus by another, as well as of equine encephalomyelitis virus and of influenza A by yellow fever virus, belongs to the same group of phenomena. No neutralizing antibodies or nonspecific antiviral substances were found in the yellow fever virus (558). A similar type of antagonism is frequently observed also among plant viruses. Yellow mosaic virus will not grow in the tobacco tissue cells al- ready infected with the agent causing common mosaic disease (608). Other antagonistic phenomena between plant viruses have been re- ported (612). The virus of peach-yellow prevented invasion by the virus of little-peach and the latter prevented invasion by the former (540). The conclusion was reached (608), therefore, that virus domi- nation in a plant may be looked upon as a type of antagonism, quantita- tive in nature, the degree of domination by a given virus being influ- enced by the host. Many other instances of virus antagonism have been reported, as when one virus prevents the multiplication of another and actually re- places it in plants in which it is established (37). Certain vitamins, such 168 ANTAGONISMS BETWEEN NONSPECIFIC PATHOGENS as ascorbic acid and thiamin, and certain other organic compounds, such as cysteine, inhibit the formation of necrosis produced by tobacco-mo- saic virus. This reaction is reversible, since necroses begin to develop when the tobacco leaves thus treated are placed in pure water (815). The ability of bacterial phages to interfere with the development of other phages has been studied in detail by Delbriick and Luria (175, 586). They have shown that a certain phage, after inactivation by ultra- violet radiation, retained its ability to interfere with the growth of an- other phage acting upon the same host. The partly inactivated first phage is adsorbed by the sensitive bacteria and inhibits their growth without producing lysis. The partly inactivated phage interferes also with the growth of the active phage. This interference between bac- terial phages was explained as due to competition for a "key-enzyme" present in limited amount in each bacterial cell. This enzyme was also believed to be essential for bacterial growth. In order to explain the "mutual exclusion effect" of one virus by another, a "penetration hypothesis" was proposed ( 1 74) . According to this hypothesis, the penetration of one virus into the cell renders the cell membrane impermeable to any other virus j each virus has a char- acteristic penetration time, and a change of permeability occurs at the end of this time. The depressor effect consists in competition between the two viruses for the same substrate. The function of a co-factor, like tryptophane, was considered to be either that of a cement substance acting in a specific combination be- tween virus and host receptive spots or as a coenzyme which enables the virus particles, during their encounters with the host cells, to be- come attached to them and attack them (21). MICROBES AND TUMORS The ability of certain microbes to bring about hemorrhage in tumors (455a, 1044) may also be classed among the antagonistic phenomena. The hemorrhage-producing agent is a polysaccharide and is isolated only from gram-negative bacteria. Laszlo and Leuchtenberger (549) described a rapid test for the de- tection of tumor-growth inhibitors. Inhibition was judged by comparing ANTITOXIC PROPERTIES OF ANTIBIOTICS 169 tumor sizes and weights in treated and untreated groups of mice bear- ing sarcoma, after a period of 48 hours of growth. The groups were matched as to initial size of the tumors. The selective damage said to be caused by penicillin to sarcoma cells as compared with normal cells (156) was later shown (567) to be due not to the pure penicillin itself but to some impurity present in crude penicillin preparations. The hemorrhagic effect upon the tumors is highly selective, being characteristic of the sarcoma cells only and does not occur in normal tissues, with a few minor and slight exceptions. The phenomena of hemorrhage and necrosis are followed in some cases by a complete and permanent regression of the tumor. The curative effects of such treat- ments are still open to question, however (94). The same may be said of the effect upon tumors of trypanosomes or of the "factors" produced by them. The effect of penatin upon sarcoma has been tested and found to be negative (113). ANTITOXIC PROPERTIES OF ANTIBIOTICS The ability of various microorganisms to destroy or neutralize bac- terial toxins has been definitely established. The substance involved was designated as an antidotic (759). It is produced by B. subtilis and P. notatum; however, isolated penicillin had no such effect, although large doses of this antibiotic protected mice against the action of gono- coccal endotoxin (685). Clavacin was also found (675) capable of neu- tralizing tetanus toxin. CHAPTER 10 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES Antimicrobial agents are of either chemical or biological origin. The first comprise inorganic (heavy metals, halogens) and organic (phenols, arsenicals, dyes, aromatic oils) compounds. The second include a variety of products of higher plants (quinine, chaulmoogra oil, wheat flour pro- tein, allicin), higher animals (lactenin, lysozyfne), and microorgan- isms, to which the term "antibiotic" Is specifically applied. The property possessed by culture filtrates of many bacteria of inhib- iting the growth of bacterial cells has long been recognized. The sug- gestion has even been made that all bacteria, when tested at the right age and under proper conditions of culture, are able to produce anti- bacterial substances (71). It is now definitely established, however, that this property Is characteristic of only certain strains of specific bac- teria, fungi, and actinomycetes. The production of antibiotic substances by microorganisms is influ- enced by the strain of the organism, composition of the medium, incu- bation temperature, age of the culture, aeration, and certain other factors. The more Important antibiotic substances are described briefly in Table 32. They may be classified on the basis of their origin from spe- cific microorganisms, their chemical properties, or their biological ac- tion. Differences between various compounds may often be in degree rather than In kind. Different organisms may produce the same anti- biotic j frequently the substance may show minor variations from the general type, these variations being both chemical and biological. Some organisms are able to produce more than one antibiotic: B. brevis pro- duces tyrocldlne and gramicidin j P. notatum, penicillin and penatin; A. fumigatusy fumlgatin, fumlgacin, splnulosin, and gllotoxinj A. fiavusj aspergilllc acid and penicillin. Since the name of an antibiotic often designates only a crude prepara- tion, considerable confusion has arisen because different names have been given to the same preparation, or the same name has been applied n3 .2 2 n c o a 1 J .2 .2 o C c !2 •^ u a E o ^ « c o .5? *-5 i £>0 'T3 C 1 'i .2^ <45 a O 1 13" > O < < 1 > 15 tj i t3 J3 60 •S s G .2 B r> c -C o c 1 Sd u G i So c i 1 (J o 6 > tj 60 G i S u _> i o a, o O G O 3 .2 > 03 J 13 u 1 OS 1 E c u J G OJ K 1 g. 1) l-H too 1 c 1 3 t;5 c ^ .> •l > 4J t4 u *5b TJ i o .2 -i '1 a. 1 ■§ 1 .« o ^ w O O 5 Oh O i 'c o pa "1 O G 1 < 1 c 6 E CI o .s a, i 6 13 J i O S3 1 1 u .. ^ s V c o 1 1 1 J o ^ Oh ^^ c ^ B:' 'Z G S 1 & T < 4-. U S "u G ^ -z iXI Oh 3 O ^ O O vO U o .^ 6 -a '■^ a, o sjo 'o .S Ui O M ^ iJ ° 2 *^ G O u . N O N CO -^ 8 2 ^ o a S 2 o > c^ U 4 1^1 8 -s •§ I s I 6 Q .s g } y o .5 , U e '3 5 w) O E O o W) ^ c o 1 8 1» c & _> 1) 4) ■^i "cS !5 s "1 B ■_Q a n o ^ 6 C 60 6 a X B "S ^ *< 4J o rt Oh 1> .^ .r. 3 rt o o to J5 r£i o o u JJ (U u '« Vh 'T3 4) 1 «o 3 3 n "o o ^ ^ -S ■-tH "TS O °_ "J O rt N C « o Ox 3 o 6 ^ 'H *S P I I u ■7? -^ -^3 J2 « S •K o U I-i Ou, (U '-5 B 60 O S a <«- a 3 8 c I .2 (73 -Q o ^ • ?s a a '8 f| :S 1^ .1 .1 1 60 ^ "i 1- 1 1 "2 1^- 03 ^- 1 a-* t o a. B -^ >i fc O c 60 'T5 > c 3 ^ 's .2 6 o o 1 Oh C o _re u c o o 6 > IS .> 'G o C1-, & o y ■§ c re •g o -g ° !& o Q-, -s a ?" CL, o u 'o D-i 6 .2 6 ^ "So 51 u uo 6 1 2 O Oh c a h^ ^3 . 3 S .9 c 3 ^ C o Jj -50 -2 " .t; -^ ^ -^ E -^ St jTn ^ G O .S O 8 ;^' :^ a, as a. O O ^ n c . • -S o c ^ ^ ^'^ b (U c C O ^ g •^ o Dh X o 're :5 s >^ o C C o ^ z f^:. B o 1; rt O o U •;: ^ n .o ^ E OJ o -Q > W & r2 > ^ '§ C '35 s. g. 1 6 1 1 > o 60 1) « > Es t5 -5 o o •r; ^ .- O ^ e •r; rt « 1 o ^ vO o to ? di ^ B ,S? K « £! r 6 y ^ U o .s M u if. 1 o ■g. a Oh C -6 B ;i 3 3 J tuO 3 — 5 r^ < a .S S 3 -^ ^ J^ ^ (>o "I -s g bo V3 .^ i- I !> .- a: ^ *-; JD .ti •'^ rt « 1 .> B^-g- ^ C o > a 11 1 1 a, 2 i J ^ o 6 nzl — ;s o -C != '^ I^' u ^ r^ c >^ S cq > ^ -S O ■*n '■S c (J 3 o <+-! ?^ ^ ^ i too 6 ^ K CM S t) g 13 o C -^ r^ ■" c u O ^^ o 'a. 6 o T3 i3 3 & 4_, ^ Ph O c i:. S e rs .-. 4) -1 > ^ Oh o ^ J3 3 CL, J" 3 o .s '^ 's W) « o _N .^ j_, _o Ml "13 13 ns "13 g ^ S CO ■§ JJ T) ^ ^ "& o ^ 3 >;- V >s c 'W) ^ 2 J3 o U «J ts t; W) o -13 *i o _D c s ■&, ta > o O re 'c i^ (5 e vO X 're ^ g re 3 1 .2 .S ? u C C o: (55 Og laq 3 3 re tj CX, O 2 re 13 3 'u > > 176 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES by different workers to different preparations even when these have been obtained from the same organism. Witness, for example, the designation "subtilin" that has been applied in different countries to preparations obtained from different strains of members of the B. suh- t'tlis group grown on media of different composition. The fact that other names, like "bacitracin," "subtilysin," "endo-subtilysin," and "bacillin," are given to certain preparations of B. sub tilts does not necessarily indicate that the substances are different. We must await further in- formation concerning the chemical and biological properties of these preparations before their identity can be definitely established. The name "aspergillin" has been applied to at least four preparations, in spite of the fact that it was first used to designate the black pigment of A . niger. Various names have been used to designate indefinite preparations produced by unknown organisms. This is true, for example, of "my- cocidin" produced by certain fungi and active against M. tuberculosis (328), and of "fungin" and "my coin," terms used to designate anti- biotics of fungi and actinomycetes, as well as of the term "inhibin" to designate antibacterial substances present in honey. On the basis of their solubility, antibiotics may be divided into four groups : Group A. Soluble in water at different reactions, and insolutle in ether. These substances usually represent proteins, organic bases, or adsorp- tion compounds on protein molecules. Some have been isolated in a pure state. They comprise the bacterial enzymes acting upon micro- bial polysaccharides, actinomycetin, microbial lysozyme, streptothri- cin, streptomycin, penatin, and pyocyanin. Group B. Soluble in ether and in water at proper reactions. Here belong many of the important antibiotic substances so far isolated and de- scribed, namely, penicillin, flavicin, citrinin, clavacin, proactinomy- cin, penicillic acid, and aspergillic acid. Group C. Insoluble in ether and in water. These include gramicidin, ty- rocidine, subtilin, and simplexin. Group D. Soluble in ether and insoluble in water. These include fumi- gacin, fumigatin, gliotoxin, actinomycin, pyocyanase, and others. Some of the antibiotic substances have been crystallized, and infor- CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES 177 mation has been gained concerning the approximate chemical nature of others j many others are still imperfectly known. On the basis of their chemical nature, the antibiotic substances may be divided as follows : Lipoids and various microbial extracts removed by organic solvents, such as pyocyanase, pyolipic acid, and others Pigments, namely pyocyanin, hemipyocyanin, prodigiosin, fumigatin, chlororaphin, toxoflavin, actinomycin, litmocidin, and others Polypeptides, comprising tyrothricin, gramicidin, tyrocidine, colicines, subtilin, bacillin, and actinomycetin Sulfur-bearing compounds, namely the different penicillins, gliotoxin, and chaetomin Quinones and ketones, namely, citrinin, spinulosin, clavacin, and peni- ciHic acid Organic bases, including streptothricin, streptomycin, and proactinomycin Oxford (701) classified the known antibiotic substances on the basis of their chemical structure. Most of the antibiotic substances can thus be grouped as follows : I. Compounds containing C, H, and O only 1. Ce group: C6H6O4 — kojic acid 2. C7 group: C7H6O4 — clavacin 3. Cg group: CgHgOe — puberulic acid C8H8O4 — fumigatin C8H10O4 — penicillic acid 4. Cio group: C10H00O3 — pyolipic acid 5. Ci3 group: C13H14O5 — citrinin 6. Ci5 group: C15H14O6 — javanicin 7- Ci7 group: C17H20O6 — mycophenolic acid 8. C20 group: CgoHieOe — viridin 9. C32 group: C32H44O8 — fumigacin, helvolic acid Various other compounds belonging to this group have been isolated, such as gladiolic acid, CnHjoOg. II. Compounds containing C, H, O, and N * I. C12 group: C10H8ON2 — hemipyocyanin C12H8O4N2 — iodinin C10H20O2N2 — aspergillic acid 178 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES 2. Ci3 group: CjaHioONg — pyocyanin 3. C21 group: C2iH37_390i2N7 — streptomycin 4. C34 group: C34H4(;04N2 — pyo II 5. C41 group: C4iH5gOiiN8 — actinomycin 6. Ci4e group: High molecular weight compounds, such as grami- cidin and tyrocidine; diplococcin may also be in- cluded in this group III. Compounds containing C, H, O, N, and S 1. C9+, namely the penicillin group of compounds which is desig- nated by the formula C9H11O4SN2.R 2. Ci3 group: C13H14O4N0S0 — gliotoxin IV. Other compounds, many of which have as yet not been fully identi- fied. Here belongs ustin, C19H15O5CI3. On the basis of their toxicity to animals, antibiotic substances may also be divided into three groups: Compounds that are nontoxic or but slightly toxic ; here belong penicillin, streptomycin, flavicin, polyporin, and actinomycetin Compounds of limited toxicity, including gramicidin, tyrocidine, citrinin, streptothricin, and fumigacin Highly toxic compounds, such as actinomycin, gliotoxin, aspergillic acid, and clavacin Many of the antibiotic substances are thermostable, others are ther- molablle ; some pass readily through Seitz and other filters, others are adsorbed. The various methods of isolation of these substances are based upon their chemical nature, solubility, and properties of adsorption. SUBSTANCES PRODUCED BY BACTERIA Lifoids and Pigments Ps. aeruginosa, discovered by Gessard in 1882 (329), and formerly known under the names of Bacterium fyocyaneum and Bacillus fyo- cyaneusy produces several antibiotic agents, the colorless lipid pyocya- nase, the pigment pyocyanin, and an alcoholic extract of the bacterial cells. Pyocyanase, the first antibiotic substance to be isolated, has had a SUBSTANCES PRODUCED BY BACTERIA 179 rather interesting history. Emmerich believed that it is an enzyme (233, 236). Later it was found (766) that all the active substance could be extracted with lipid solvents j the extraction of the cells of Ps. aeruginosa with alcohol also gave active antibacterial preparations. A crystalline product was finally obtained (448) 5 it was soluble in organic solvents and had a bactericidal effect upon B. anthracisy S. albus, C. difhtheriae, and a number of other organisms. In the course of time it was recognized that all the antibacterial ac- tivity of the lipoid extracted from the medium was due to the presence of fatty acids, so that the term pyocyanase is now used to designate the antibiotic lipid, found in the medium and containing unsaturated fatty acids. Certain well-defined compounds have recently been isolated, such as pyolipic acid (50a). Schoenthal (843) obtained three compounds that possessed antibac- terial properties, namely, pyocyanin, oxyphenazine, and an active oil that formed insoluble salts with calcium, barium, and heavy metals. The last appeared to be similar to what had previously been described as pyocyanic acid, a substance highly active against V. comma. All three compounds were isolated by extraction with chloroform. Different strains of Ps. aeruginosa may produce either pyocyanase or pyocyanin or both, the production of the two not proceeding in a paral- lel manner. Among the amino acids, alanine and tyrosine were found to be favorable to pyocyanin production (346), although the effect of tyrosine is not very significant (346, 461, 573). The determination of the nature of the antibacterial substances of Ps. aeruginosa can be carried out in the following manner (418) : the organism is grown in bouillon for 14 daysj the cultures are heated for a half hour at 75° C. to kill the living cells j they are then centrifuged, the liquid is treated with chloroform which extracts the pigment, and the chloroform solution is concentrated in vacuo at 50° C. j the aqueous solution remaining after chloroform extraction is acidified with hydro- chloric acid and again shaken five times with chloroform, thus extract- ing the fatty acids. It was found that the antibacterial properties are yery little diminished by removal of the pigment j however, when both the pigment and the fatty acids are removed, no antibacterial action is left in the culture. S. aureus is commonly used as the test bacterium. 180 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES The broth culture of the organism may also be first extracted with ether, giving pyocyanase, and the residue treated with chloroform, yielding pyocyanin. The solution left after the removal of the blue chloroform extract may be again treated with ether, giving a yellow pigment, which also has some activity (529). This pigment is a deriva- tive of pyocyanin and is often designated (1036) as hemipyocyanin. It may also be obtained by acidifying pyocyanin with acetic acid and heat- ing. The fluorescin remaining in the culture after the ether and chloro- form extraction was found to be inactive. In old cultures, pyocyanin is changed into a brown pigment, pyoxanthose. A fourth pigment, which is yellow in transmissible light and fluorescent-green in reflected light, is produced under certain conditions. It was excreted into the medium as a leuco base. Pyocyanase is soluble in ether, benzol, benzene, and petrol ether (766). It can be separated into several lipoids, the action of which shows slight variation. This preparation consists of a phosphatide, a neutral fat, and a free fatty acid. The antibacterial properties have been attributed to the last constituent (421 ). A definite relation has been ob- served between the number of double bonds and the activity of the sub- stance (59, 420). According to Dressel (197), most fatty acids exert bactericidal and bacteriolytic effects upon gram-positive bacteria, whereas gram-negative organisms are not lysed. Pyocyanase acts upon various bacteria, including the colon-typhoid group, though the ability of the substance to inhibit the growth of this group of bacteria has been denied by some workers (372). B. mesenter'icus and other spore-forming bacteria also produce anti- biotic agents of a lipoid nature. The substance is not affected by heating for 30 seconds at 100° C. but is weakened at 1 15° C. for 10 minutes. It is considered similar in its bactericidal properties to pyocyanase. Alcohol and acetone extracted from B. mesentericus a weakly active substance (419) that diffused through a cellophane membrane and could be partly absorbed on a Berkfeld filter. When shaken directly with ether, the culture lost its antibacterial properties. The ether extract was concentrated and ammonia added, and the solution was treated with 50 per cent alcohol. The alcohol was then removed, and the residue was acidified and treated with petrol ether, which brought the active sub- SUBSTANCES PRODUCED BY BACTERIA 181 stance into solution. The active substance was again dissolved in alcohol and taken up in ether. The ether solution was washed with water, evapo- rated, and dried. One liter of a 30-day-old culture of B. mesentericus gave 1 62 mg. of petrol-ether-soluble fatty acids and an oily substance of a brownish color. It was neutralized with NaOH solution and tested. The extract diluted to 1:7,500 killed diphtheria} a 1:1,000 dilution was required to kill staphylococci. Iso-valerianic acid and oleic acid, isolated from this material, had a similar bactericidal action. Weaken- ing of the substance by heating was demonstrated and was believed to be due to a break in the double bond of the oleic acid. E. coli exerts an antagonistic effect in vivo when injected subcutane- ously or when used for feeding. It produces (367, 369) a thermolabile substance that was considered to be a lipoid in character. However, some of the antibiotics of E. coli, namely the colicines, appear to be definitely proteins or polypeptides. Pyocyanin is a dark blue pigment, red in acid solution, m.p. 133° C, water soluble and amphotheric. It is extracted with chloroform, then reextracted by acidulated water. It is characterized by a wide antibiotic spectrum and high toxicity to animals. This pigment was first studied by Fordos in 1 860 (277). Since then many contributions have appeared dealing with formation and nature of this pigment. Several formulae have been suggested for pyocyanin (461, 935, 1036), one of which is shown in Figure 18. The structure of pyocyanin has considerable simi- larity to chlororaphin and iodinin, obtained from Chromobacterium (596), and two synthetic compounds, phenazine and acridine (939). Since Ps. aeruginosa is an extremely variable organism, the nature and abundance of the pigment are variable. Keeping the organism for 5 minutes at 57° C. or cultivating it in liquid egg-albumin has been found to result in destruction of some of its pigment-producing properties (330,557). Hemipyocyanin is found in old cultures of Ps. aeruginosa (843) and is synthesized (1036) from pyocyanin. It is a yellow pigment, m.p. 158° C, with basic and phenolic properties. It is moderately bacterio- static and strongly fungistatic (883). Prodigiosin is produced by S. marcescens. It is insoluble in water and is active against B. anthracis (1035). 182 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES CH3O.C C=CHp H HC CH3 I I I c ZC C=CH\ HOC CH ri2C CH.CO O / COOH PENICILLIC ACID CLAVACIN HC C.CH2OH KOJIC ACID PHENAZINE FUMIGATIN CONH2 CHLORORAPHIN Figure 18. Structural formulae of some antibiotic substances. lodinin is a deep purple-bronze pigment, m.p. 236° C. (Figure 18). It is produced by Ch. iodinum {S9l) and is excreted into the medium. It is insoluble in water and in acids but soluble in alkali, and is phenolic in character. It is dissolved in chloroform. It is active against S. hemo- lyticus, less so against S. aureus and other bacteria. Violacein, a purple pigment, is produced by Ch. violaceum. This pigment is active against gram-positive bacteria, except CI. welchii; it SUBSTANCES PRODUCED BY BACTERIA 183 has little effect upon the gram-negative bacteria, except the meningo- cocci. Among the fungi, only Blastomyces dermatiditis is susceptible. The action of the pigment is greatly affected by serum (569). It may be added here that certain aromatic oils and various fatty acids possess marked bactericidal properties (518). Unsaturated acids are more active than saturated acids (1040). Ordinary peptones have also been found to contain a substance that is active against various bacteria, especially when small amounts of inoculum are used (202). The active substance is thermostable and is associated with an acid-precipitated fraction that is pigmented and changes color upon oxidation and reduc- tion. The bacteriostatic effect of this material can be corrected by the addition of reducing agents, such as thioglycollic acid. The bacterio- static action of dyes is well known and need hardly be discussed here. It is sufficient to mention, for example, methylene blue and indophenols in oxidized forms. Pyo-compounds. Doisy and his collaborators (389) centered their attention upon the antibiotics present in the Ps. aeruginosa cells. This group of compounds was designated as Pyo I, Pyo II, Pyo III, and Pyo IV. The culture of the organism was incubated for 5 weeks, cooled, and acidified with HCl to ^H 't^.S'-, it was centrifuged, and the precipitate was extracted with hot 95 per cent ethyl alcohol. The alcohol extract was diluted with water to 80 per cent alcohol and treated with petro- leum ether, to remove the fats and fatty acids. The alcoholic solution was evaporated and the aqueous residue extracted with ether. The ex- tract was separated into the four fractions listed above, which repre- sented pure, crystalline, active substances. These fractions were struc- turally related and were more active against the gram-positive than the gram-negative bacteria. They were nontoxic to animals. Polysaccharidases Among the antibiotic substances of microbial origin may also be in- cluded the enzyme systems that have the capacity of decomposing the capsular substance of certain bacteria, thereby rendering them more readily subject to destruction in the blood stream or in other substrates. The first enzyme of this type was isolated by Dubos and Avery (204, 184 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES 205, 207) from certain soil bacteria. These enzymes are highly specific, some being able to act only upon one type of pneumococci. As a result of their action, the pneumococcus cell is rendered susceptible to destruc- tion by phagocytosis. This enzyme was produced by the soil bacteria under selective conditions of culture, that is, when the capsular polysac- charide of the pneumococcus was present in the medium j the only other substance that could be used for its production was aldobionic acid, which was derived from the above polysaccharide. Yields of the en- zyme were increased by increasing the concentrations of the specific substrate in the medium from 0.0 1 to o.i per cent. Above o.i per cent, the yields decreased, 0.3 to 0.4 per cent inhibiting the growth of the bac- terium. The addition of 0.1 per cent yeast extract favored the produc- tion of the enzyme j proper aeration was essential, the bacterium mak- ing the best growth in shallow layers of medium. The enzyme was concentrated by distillation in vacuo and by ultrafiltration. Toxic sub- stances accompanying the active preparation could be largely removed by the use of an aluminum gel. The enzyme is associated with a protein which passes through a collodion membrane with an average pore size of 10.6 (J, but is held back by pores having a diameter of 8.2 |j. After filtration, the enzyme can be recovered in solution by immersing the membrane in distilled water or in physiological salt solution. Dubos (199) believed that it is possible to develop "adaptive" bac- terial enzymes against many organic substances. These enzymes exhibit a great degree of specificity, as in the case of the enzyme that hydrolyzes the capsular polysaccharide of the pneumococcus. The cell of this or- ganism contains an enzyme that changes the cell from the gram-positive to the gram-negative state, but is ineflrective against streptococci or staphylococci. Active preparations of the enzyme protected mice against infection with as many as i ,000,000 lethal doses of the specific pneumococcus. The enzyme retained its activity for 24 to 48 hours after its injection into normal micej it also exerted a favorable influence on the outcome of an infection already established at the time of treatment. A definite rela- tionship was found to exist between the activity of the enzyme in vitro and its protective power in the animal body. SUBSTANCES PRODUCED BY BACTERIA 185 Polyfepides and Proteins From the tyrothricin complex group of antibiotics produced by B. brevisy two crystalline compounds have been isolated. They are poly- peptides resistant to the action of proteolytic enzymes (201, 208, 450, 909). The organism is grown in shallow layers of a suitable medium, such as one containing i per cent casein digest or tryptone and 0.5 per cent NaCl in tap water, adjusted to /)H 7.0. After inoculation, the medium is heated for 20 minutes at 70° C, in order to kill the vegeta- tive cells of the bacteria, leaving only the spores to develop. The cul- ture is allowed to grow for 72 hours. The reaction of the culture is then adjusted to /)H 4.5 by the use of about 3 or 4 cc. concentrated HCl per liter of culture. A precipitate is formed which is removed by filtration through paper J it is then suspended in 95 per cent alcohol (20 cc. of alcohol per liter of culture) and allowed to stand 24 hours. The active substance is dissolved and is separated from the residue by filtration j when the alcoholic solution is diluted with 10 volumes of i per cent NaCl, the substance is precipitated out. It carries all the activity and can be desiccated in vacuo, over P2O5, giving a yield of about 100 mg. of final dry substance per liter of culture medium. The protein-free, alco- hol-soluble active material is tyrothricin. When an attempt was made to produce tyrothricin in aerated submerged cultures, none was obtained in complex nitrogenous media j however, simple amino compounds, like asparagine, gave good growth and yielded the antibiotic substance. The presence of cystine in the mixture of amino acids appeared to in- hibit growth (884). Gramicidin is obtained by treating tyrothricin with a mixture of equal volumes of acetone and ether, evaporating, and dissolving in boiling acetone. On cooling, it crystallizes out as spear-shaped colorless platelets, melting at 228° to 230° C, with a yield of about 10 to 15 grams from 100 grams of the crude material. Gramicidin is soluble in lower alcohols, acetic acid, and pyridine, and moderately soluble in dry acetone and dioxanej it is almost insoluble in water, ether, and hydro- carbons. When a solution containing 20 to 50 mg. per milliliter alcohol is diluted to i mg. per milliliter, with distilled water or with glucose solution, an opalescent solution is produced without flocculation. On dilution with electrolyte solutions, an immediate flocculation occurs. 186 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES The specific rotation of gramicidin in 95 per cent alcohol solution is approximately [a]^5 _ _|_ ^o^ Qj^ analysis, it gives 62.7 per cent C, 7.5 per cent H, and 13.9 per cent N. The molecular weight, as determined in camphor, is about 1,400. The empirical formula of C74H106O14N14 has been suggested. On further study, the molecular weight of gramicidin was found (911) to present an anomaly in that it appeared to depend on the nature of the solvent and on the concentra- tion of the solute, giving values from 600 to 1,200 j isothermal distil- lation in methanol, however, indicated a molecular weight of 2,700 to 3,100, with an approximate formula of C146H000O2N30. It gave neither free amino nor carboxyl groups j it contained 10 molecules of a-amino acids, of which two or three were tryptophane residues. These and a saturated aliphatic acid, with 14 to 16 carbons, account for about 85 to 90 per cent of the weight of the substance. Amino acids that have definitely been identified are /-tryptophane, /^-leucine, /-alanine, dl- valine, and glycine (345, 892). A study of the configuration of the di- peptide valyvaline separated from gramicidin brought out the fact that only valines of like configuration have been joined together by the bac- terium (137). About 45 per cent of the a-amino acids gave the d con- figuration (449, 450). An unknown hydroxyamino compound has also been indicated. The presence of ethanolamine (2-aminoethanol-l) as a component of gramicidin hydrolysates, which reacts with periodate to yield formal- dehyde and NHo, has been definitely indicated (893). Actually two ethanolamine residues may occur in gramicidin, since their destruction during acid hydrolysis of gramicidin is considerable. The liberation of some of these amino acids during hydrolysis, such as valine and trypto- phane, can be measured by their availability to L. arah'mosus (139). Tyrocidine hydrochloride is moderately soluble in alcohol, acetic acid, and pyridine ; it is sparingly soluble in water, acetone, and dioxane, and is insoluble in ether and hydrocarbon solvents. An alcohol solution can be diluted with water to give a clear solution containing 5 to 10 mg. per milliliter; electrolytes produce an immediate precipitate. A solu- tion in distilled water containing i mg. or even less per milliliter has a low surface tension and behaves like a soap or detergent solution. Un- SUBSTANCES PRODUCED BY BACTERIA 187 like gramicidin, it precipitates a number of soluble proteins in a manner similar to some of the cationic detergents. Tyrocidine is dissolved in four times its weight of boiling absolute alcohol, to which is added alcoholic HCl (o.i mol. per liter). On cool- ing, a precipitate is formed. This is recrystallized from absolute metha- nol plus small amounts of HCl ; clusters of microscopic needles are ob- tained, melting at 237-239° C, with decomposition; the specific rota- tion is [a]^5 = — 102° ( I per cent in 95 per cent alcohol). Tyrocidine analyzes: 59.4 per cent C, 6.8 per cent H, 13.5 per cent N, 2.7 per cent CI. The molecular weight is about 1,260 or a multiple of this number. Tyrocidine is a salt of a polypeptide having free basic amino groups. The ^-amino acids make up 20 per cent of its a-amino groups. The most probable molecule was shown to contain two amino groups, three amide groups, and one weakly acidic carboxyl or phenolic group, with a molec- ular weight of 2,534. Among the amino acids, tryptophane, tyrosine, and dicarboxylic-amino acids have been detected j concentration of some of these acids has been established: aspartic acid, 5.1 per centj valine, 7.6 per cent J and leucine, 8.2 per cent (138, 140). Summaries of the chemical and biological properties of gramicidin and tyrocidine were made by Hotchkiss (449) and Hoogerheide (443). The tyrothricin-type of antibiotic substance appears to be widely dis- tributed among spore-forming aerobic soil bacteria (442, 444, 885). Preparations obtained from different bacteria appear to be markedly different in chemical nature and biological activity. This is true, for example, of the preparation obtained by the following method: A seven-day-old bacterial culture was treated with 2 to 5 per cent of an electrolyte and HCl added to give a fH of 4.0. A precipitate was formed which was centrifuged and extracted with 95 per cent alcohol, until no more turbidity could be observed after dilution with an equal volume of water. The alcoholic extracts were evaporated to dryness and extracted with ether, petroleum ether, and benzol, in which the active substances are insoluble. The residue was then dissolved in absolute alcohol, and the concentrated solution dialyzed for 24 hours against running tap-water and for 24 hours against distilled water. The active substance was obtained partly in a precipitated form and partly in a 188 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES colloidal solution in the dialysis bag. Upon evaporation of the water, a highly active, grayish-white powder was obtained. One hundred liters of medium gave 15 grams of purified active substance. The activity could be tested by inhibition of encapsulation of Friedlander's bac- terium j this was brought about by the addition of 4 mg. to i ml. of cul- ture medium. This preparation was later found to be identical with gramicidin (443,911). Gramicidin S (43, 324) is related to the tyrothricin complex, being tyrocidine in nature. It was found (894) to be a cyclopeptide, with a stoiochiometric minimum unit formed from one residue each of /-orni- thine, /-proline, /-valine, /-leucine, and ^-phenylalanine. The unit pos- sesses one free amino group, no free carboxyl groups, and one residue of chloride. Bacitracin is formed by certain strains of B. subtilis grown in shallow layers of media. A heavy surface pellicle is produced after 3 to 5 days' incubation at 37° C. The medium is extracted with normal butanol and concentrated by steam distillation in vacuo, giving a grayish-white powder. The substance is neutral and water soluble and withstands heat- ing for 15 minutes at 100° C. without significant loss of activity. It is stable in acid solution but unstable in alkaline solution above fH 9, and is not digested by proteolytic enzymes. It is active chiefly against gram- positive organisms, but the gonococcus and meningococcus are also susceptible to it. It is active in vivo against experimentally produced hemolytic streptococcus and gas gangrene infections (469). Subtilin is produced by certain strains of B. subtilis. It is a polypep- tide and is readily digested by proteolytic enzymes. It is most active at /»H 2.2 and gradually becomes inactivated with decreasing acidity. It is active against various gram-positive bacteria, acid-fast bacteria, and certain pathogenic fungi (816). Eumycin, produced by certain strains of B. subtilis and active largely against fungi, actinomycetes, and myco- bacteria, although showing little effect against staphylococci may also belong to this group (471). Subtilysin was reported to have a lytic ac- tion against gram-negative bacteria, none against cocci (925). Some of the subtilin preparations also have the capacity of inactivating bacterial toxins, such as diphtheria, tetanus, and others. This property was as- SUBSTANCES PRODUCED BY ACTINOMYCETES 189 cribed to the presence of a heat-stable substance designated as antidotic (759). Diplococcin Is produced by certain lactic acid streptococci. It is a pro- tein synthesized in the bacterial cells from the amino acids in the me- dium, and is extracted with cold dilute acetic acid. The active protein is precipitated by 60 per cent saturation with ammonium sulphate (701 ). It is active against gram-positive cocci and Lactobacillus species, but not against gram-negative bacteria. A thermostable substance was obtained (154) from B. simplex, an organism capable of bringing about the destruction of various patho- genic fungi. This antibiotic was later designated as simplexin. It was produced by the bacterium grown both on synthetic and on organic media. It can be adsorbed on activated charcoal and recovered from the latter by the use of hot alcohol. To what extent substances of bacterial origin that are toxic to brain tissues, like toxoflavin (C6H6N4O2), are also effective against bacteria and other microorganisms still remains to be determined. Toxoflavin, formed by Bacterium cocovenenans , is extracted from the culture satu- rated with salt by means of chloroform j from this it is recovered by an aqueous solution and purified (931, 932). Other bacterial toxins, like botulinus toxin, various amines and purine bases, and numerous toxins produced by bacteria in living plant and animal systems, are beyond the scope of this treatise. SUBSTANCES PRODUCED BY ACTINOMYCETES Actinom^ycin Actinomycin is an ether-soluble and alcohol-soluble pigmented sub- stance produced by certain actinomycetes, notably S. antibioticus. The culture medium is treated with ether, giving an orange-colored extract. The residue is evaporated and treated with petrol ether (975). The purification of actinomycin was effected by chromatographic adsorption, followed by fractionation of eluate. The orange-brown resi- due left after treatment with petroleum ether was dissolved in benzene, filtered, and allowed to pass through a tower packed with aluminum oxide. On washing the tower with large amounts of benzene, a chro- 190 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES matogram slowly developed. The column was then washed with a solu- tion of 1 5 parts acetone to 85 parts benzene until the yellow-orange band approached the bottom of the column. The elution of the pigment from the column was accomplished finally by further washing with 30 per cent acetone in benzene until the eluate was faintly yellow in color. The later eluates were found by assays to contain all the active pigment, whereas all previous eluates, as well as the fractions remaining on the adsorbent, showed no bacteriostatic or bactericidal activity. Pure actinomycin was obtained by concentrating the 30 per cent acetone-benzene eluates to dryness, then recrystallizing the red solid residue from acetone-ether mixtures or from ethyl acetate. From these solvents, the pigment separated as vermilion-red platelets which melted at 250° C, with slow decomposition. The pigment is readily soluble in chloroform, benzene, and ethanolj moderately in acetone and hot ethyl acetate j and slightly in water and ether. The color of the solid pigment depends on its state of subdivision j when ground very fine, its color is orange-red (910). Actinomycin is optically active, a solution of 5 mg. in 2 cc. ethanol in a I dm. tube having a rotation — i.6o°j [a]'^ = —320° ± 5. Its molecular weight was found to be around 1,000. Cryoscopic measure- ments in cyclohexanol and in phenol gave molecular weights of 768 to 780 and 813, respectively. The approximate molecular formula was found to be C41H56O11N8. Actinomycin exhibits characteristic ab- sorption in the visible and ultraviolet regions. In ethyl alcohol, it shows strong absorption at 450 (E| ^ = 200) and between 230 and 250. Actinomycin is not soluble in dilute aqueous alkali or in dilute min- eral acids. It is soluble in 10 per cent hydrochloric acid and appears to be regenerated by diluting such solutions with water. With strong alco- holic alkali a purple color is formed, which rapidly disappears. Actino- mycin is readily reduced by sodium hydrosulfite and by stannous chloride, but is unaffected by sodium bisulfite. With sodium hydro- sulfite the reduction is characterized by a change in color from red to pale yellow. The color change is reversed by exposing the reduced pig- ment to air. The same reversibility of color occurs when the pigment is subjected to catalytic hydrogenation in the presence of platinum oxide. The pigment has one or more functional groups capable of re- SUBSTANCES PRODUCED BY ACTINOMYCETES 191 versible reduction-oxidation (probably quinone in nature) and several others capable of acetylation (probably hydroxyls). The quinone-like structure of the pigment is borne out by its sensitivity to alcoholic alkali, and to hydrogen peroxide in the presence of sodium carbonate. In the latter instance, the color rapidly disappears and a cleavage seems to occur. Actinomycin in alcohol-water solutions is resistant to the action of heat, being able to withstand boiling for 30 minutes. When such solu- tions are made acid, however, boiling has a destructive effect upon the activity of the substance, the extent of destruction being directly pro- portional to the concentration of acid. The effect of alkali, however, is much greater. Dilute alkali changes the color of the substance to light brown, accompanied by a reduction in activity, which can be largely re- stored when the solution is made neutral again. At a higher alkalinity (0.25 N), especially at boiling temperature, the activity and reversibil- ity are destroyed. Exposure of solutions to light for i to 3 months re- duces the activity of the pigment very little. Streftothrkin Streptothricin is produced by Streftomyces lavendulae grown in a medium containing glucose or starch ( i per cent) as a source of energy, and tryptone, glycocoU, glutamic acid, or other organic nitrogenous compound (0.3 to 0.5 per cent) as a source of nitrogen. Sodium nitrate is a somewhat less favorable source of nitrogen. The organism is grown in stationary, shallow cultures containing starch as a source of carbon or glucose and a small amount of agar, or in submerged cultures. The optimum temperature for the production of streptothricin is 23° to 25° C. (946). The relation between growth of the organism and pro- duction of the antibiotic substance is brought out in Table 't^'t,. Streptothricin is soluble in water and in dilute mineral acids, but is destroyed by concentrated acids. It is insoluble in ether, petrol ether, and chloroform. In the crude culture-filtrate and in the alcohol-precipi- tated form, streptothricin is thermolabile, whereas in the purified state it is thermostable, withstanding 100° C. for 15 minutes. Treatment with proteolytic enzymes does not reduce its activity. On adjusting the reaction of the medium, when growth is completed, to ^H 3.5 with 192 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES TABLE 23- GROWTH OF STREPTOMYCES LAVENDULAE AND PRODUCTION OF STREPTOTHRICIN ON TRYPTONE-STARCH MEDIUM DRY WEIGHT NITROGEN ACTIVITY INCU- OF MYCE- IN MYCE- IN UNITS BATION STARCH LIUM IN LIUM IN E. B.sub- AERATION IN DAYS LEFT MILLIGRAMS MILLIGRAMS coli tilis Shaken 2 +++ 10 5 Shaken 3 + 225 18.2 10 50 Shaken 4 O 293 26.2 75 250 Shaken 6 o 231 17.3 100 300 Shaken 8 75 200 Shaken 12 o 142 9.6 30 50 Stationary 7 +++ 50 200 Stationary 10 Tr 235 18.8 50 300 Stationary H Tr 60 250 From Waksman (946). acid, a precipitate is produced, the filtrate containing virtually all the activity. Streptothricin is completely adsorbed, at neutrality, on charcoal, from which it can be removed by treatment for 8 to 1 2 hours with dilute mineral acid or acid alcohol. The acid extract is neutralized and con- centrated in vacuo y at 50° C, just to dryness j the residue is extracted with absolute alcohol, filtered, evaporated, and taken up in water. It can also be precipitated from the neutralized solution with ether or acetone. Further concentration and reduction in ash content can be ob- tained by subsequent treatments. On electrodialysis, the active sub- stance moves to the cathode at fH. 7.0. Streptothricin has been crystallized as the Reinecke salt (300). The crystals consist of a cluster of fine platelets which decompose at 192° to 194° C. after sintering at 184° C. The molecule was found to corre- spond to the di-reineckate of a base C13H25O7N5 j the a-amino nitrogen was 20 to 22 per cent of the total nitrogen. The molecule of streptothri- cin is thus believed to contain at least five nitrogen atoms, two of which are present as salt-forming basic groups j it is free of O-methyl, N- methyl, and hydrolyzable acetyl groups. Streptothricin is stable be- tween fH I and 8.5, but is destroyed by high alkalinity. The activity of SUBSTANCES PRODUCED BY ACTINOMYCETES 193 the sulfate is 500-530 [ig/mg. One of the more recent modifications (713a) of the method of isolation of streptothricin comprises the fol- lowing steps: charcoal adsorption, elution with formic acid in methyl alcohol-water, partial concentration in vacuo, precipitation with picric acid, conversion to hydrochloride, chromatography over aluminum oxide, and precipitation with methyl orange as helianthate. This preparation had an activity of 830 ng/mgl., with a specific rotation Streftomycin Streptomycin is produced, in stationary and shaken cultures, in a medium containing meat extract, corn steep, soy bean meal, or some other suitable material. Its maximum production occurs in shaken cul- tures in 2 to 3 days, and in stationary cultures in 7 to 10 days (830, 971), as shown in Table 34. TABLE 34. GROWTH OF S. GRISEUS AND PRODUCTION OF STREPTOMYCIN Incubation /^g />Hof Growth medium in mg.* Shaken cultures 2 days 10 . 7.8 270 3 days 70 7-7 185 4 days 60 7.8 - 7 days 70 8.2 - Stationary cultures' 3 days 6 7-7 73 5 days 12 7.8 171 7 days 53 7.9 163 9 days - 8.3 264 1 2 days 55 - - From Schatz, Bugie, and Waksman (830). * Weight of dry m>'celium produced by S. griseus. Streptomycin is also a base, like streptothricin, but differs from it in cl^emical composition, antibacterial spectrum, and lower toxicity for animals (830, 952). It is highly active against the gram-negative en- teric group of bacteria and related organisms. A detailed discussion of 194 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES the nature of the antibiotic action of streptomycin and its utilization for chemotherapeutic purposes is presented elsewhere (pp. 287-296). It was at first suggested (947) that 3 units of activity be recognized for measuring streptomycin: an S unit, or the amount of material that will inhibit the growth of a standard strain of E. coU in i ml. of glu- cose-free nutrient agar or broth j an L unit, inhibition on a liter basis j a G unit, inhibition on a dry weight basis of crystalline material. Since streptomycin base was found to be 1,000 S units per i mg., it was de- cided to accept the weight of streptomycin as a basis of standardization: I S unit is thus equivalent to i microgram of the pure base. Streptomycin can be isolated from the medium by several proce- dures. In one method (115), culture filtrates of S. griseus assaying 100 to 180 units of streptomycin per ml. served as the starting material. Several common adsorption agents, such as charcoal, can be used to re- move the active material from the culture. The substance is then eluted with hydrochloric acid in 95 per cent ethanol. Anhydrous hydrogen chloride in methanol is a more convenient reagent, since the crude streptomycin can be precipitated directly from the methanol solution with ether. The filtrate is clarified at pH 2 with 0.5 per cent carbon j this is followed by removal of the streptomycin, at ^H 7, with i per cent carbon, which is washed successively with water, neutral ethanol, and neutral methanol, and the streptomycin is eluted by two or three extractions with o.i N methanolic hydrogen chloride. The alcoholic extracts are combined and 2 to 3 volumes of ether added, precipitating the crude streptomycin chloride as a light-brown amorphous powder. When the methanol solution contains much water, a sticky gum results. The recovery of the streptomycin by this method varies from 30 to 50 per cent, the product assaying from 150 to 300 micrograms. For further purification, a faintly acid solution of crude streptomycin chloride in 70 to 80 per cent methanol is percolated over a sulfuric acid-washed alumina column (^H 5 to 6) j an inactive fraction giving a positive Sakaguchi test first appears, followed by a Sakaguchi-negative fraction. This test parallels the antibiotic action of the fractions. A small amount of active material remains on the column and can be washed through by lowering the methanol content of the solvent. This ma- terial contains sulfate ion but no chloride. The streptomycin sulfate SUBSTANCES PRODUCED BY ACTINOMYCETES 195 passes through the column less rapidly, since it is less soluble than the chloride in methanol. The various streptomycin fractions obtained from the column are concentrated and lyophilized, giving white amorphous powders. The most active fractions range from 600 to 900 jjg/mg., and amount to approximately 80 per cent of the total. Satisfactory results are obtained only if the crude streptomycin has an activity of about 200 Mg/mg. or higher. Preparations of lesser purity contain substances which interfere with the development of the chromatogram. The chloride is soluble in methanol, less soluble in ethanol, practically insoluble in butyl alcohol, acetic acid, and pyridine. The sulfate is only slightly soluble in metha- nol and practically insoluble in the other solvents. Streptomycin gives a positive Sakaguchi test, the presence of a guani- dine group being indicated by the fact that alkaline hydrolysis results in the formation of ammonia and the disappearance of the Sakaguchi test. Streptomycin also gives a positive test for an hydroxyl group. Negative tests are obtained in the amino nitrogen, Hopkins-Cole, Mil- Ion, xanthoproteic, biuret, and Pauly diazo tests. The presence of a carboxyl group is considered as questionable, since the streptomycin chloride, obtained by precipitation from methanolic hydrogen chloride with ether, gives approximately neutral solution. The ultraviolet spec- trum of streptomycin showed only end-absorption below 230 my, which makes improbable the presence of an aromatic ring or conjugated double bonds. Streptomycin is inactivated rapidly by o.i N sodium hydroxide at room temperature. It is relatively stable over a -pH range of i to 10 but is inactivated by i N hydrochloric acid. Streptomycin was first crystallized as the reineckate salt from water, in the form of thin plates which decomposed at 162°-! 64° C, the basic component being (CioHi907_8N3)^. The antibiotic potency of pure streptomycin lies between 800 and 910 Mg per mg. (299). Streptomycin can also be isolated (537a) by the method described above for strep- tothricin. This includes charcoal adsorption, elution with methanolic formic acid, precipitation with picric acid, conversion to the hydro- chloride, chromatography with aluminum oxide, and final conversion to the crystalline helianthate. 196 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES In another procedure (538) the crystalline salt of streptomycin and /)-(2-hydroxy-l-naphthylazo)-benzenesulfonic acid is prepared from streptomycin HCl, and orange II. The salt has an activity of 300 Mg/mg. Crystalline streptomycin sulfate was found to have an activity of 520 ng/mg. On chromatographic purifications, streptomycin concentrates yielded a crystalline double salt of streptomycin trihydrochloride and calcium chloride (Figure 19). This preparation showed that streptomycin has the composition CsiHg^.ggN^Ois- The double salt is characterized by constant biological, chemical, and physical properties. It is more satis- factory than the hydrochloride which is obtained by precipitation. The double salt can be prepared from streptomycin hydrochloride or from the crystalline streptomycin helianthate. A cryoscopic molecular weight determination on streptomycin trihydrochloride in water gave about 800 for the free base, necessary corrections having been made for the chloride ion and the non-ideal cryoscopic behavior of the trivalent streptomycin ion (712). Further studies on the chemistry of streptomycin revealed the fact that it has the general constitution of a hydroxylated base (streptidine) attached through a glycosidic linkage to a nitrogen-containing disac- charide-like molecule. The latter group of the streptomycin molecule contains a free or potential carbonyl group and a methyl-amino group (85). The reaction of the streptomycin with one molecule of water can be presented as follows: C.,H3,_3oN,0,, + H,0 ^ CsH.sNeO, + Ci3H,,_,3NO, Streptomycin Streptidine Streptobiosamlne Ci3H,3NO, -f H3O -> CeHioO, + C,Hi,N05 Streptobiosamine Streptose N-methyl-a- /-glucosamine The basic nitrogen atom in the streptobiosamine is not present as a primary amino group. The streptomycin molecule was presented graphically as follows : Tyrocidine hydrochloride. From Hotchkiss (449) Gramicidin, From Hotchkiss (449) e--^:^ A^'U !^^^!^^^^*^^^ Fumigacin. From Waksman and Geiger (955) Gliotoxin. From Waksman and Geiger (955) i/MI^^Hi T^^v^^^^^^l r^^'^^^^H SI Pi Citrinin. Prepared hy Tii Actinomycin. Prepared by Tischler Figure 19. Crystalline preparations of antibiotic substances. FiciURE 20. StiLptoiiiycin crystals. SUBSTANCES PRODUCED BY ACTINOMYCETES 197 CnHasNAs + H.O- Streptomycin - NH + C.jH.jNO, CHNHC - NH2 CHOH -> Streptidine Streptobiosamine C,3H,3NO, + H,0 H HOC I HOCH I HOC - CHO I HC H,C Streptobiosamine >■ Streptose CHOH I CH3NHCH O HCOH I HOCH I CH I CH,OH N-methyl/glucosamine Streptidine was characterized by the following crystalline salts: dipicrate, sulfate, carbonate, dihydrochloride, dihydroiodide, dihelian- thate, di-<:^-camphorsulfonate, and chloroplatinate. Streptidine appears to contain one or more hydroxyl groups, but no primary amino, car- boxy, methoxy, or carbonyl groups. It formed an octaacetyl derivative (713). Streptidine has the molecular formula C8H18N6O4. A further study of this compound has been made by Carter et al. (114). Streptomycin hydrochloride was completely inactivated on standing 24 hours in an- hydrous i.O N methanolic hydrogen chloride without forming a new basic group. The addition of two volumes of ether completely precipi- tated the guanidine, which was previously reported by Carter et al. (115) as one of the functional groups of streptomycin. From the supernatant solution there is readily obtained an amorphous, optically active hydrochloride of a nonguanidine base whose properties agree 198 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES with those of the methyl streptobiosaminide dimethyl acetal hydro- chloride reported by Brink et al. (85). The addition of picric or sulfuric acid to an aqueous solution of the guanidine hydrochloride gives in- soluble crystalline salts that are readily recrystallized from hot water. The analytical data for the salts agree with those of a diguanidine base of the composition C8H18N6O4. This compound has the same empirical formula as that suggested by Brink et al. for streptidine, and is pre- sumed to be identical with it. Streptidine sulfate was also obtained by allowing a solution of strep- tomycin chloride in i N sulfuric acid to stand at 37° C. for 45 hours. The sulfate was precipitated in crystalline form by adding 3 to 5 vol- umes of acetone to the reaction mixture. Streptidine was hydrolyzed by refluxing for 48 hours with 6 N alkali yielding four moles of ammonia and a new base, for which the name streptamine was proposed. This base was isolated as the slightly soluble sulfate by neutralizing the hydrolysis mixture with sulfuric acid and adding an equal volume of methanol. The sulfate was purified by re- crystallization from aqueous methanol. The hydrolysis of streptidine proceeded as follows: CsHjsNeO^ -f 4H2O -> CoHi.NoO^ + 4NH3 -f 2CO2 Streptidine Streptamine These results appeared to establish the fact that the six nitrogen atoms of streptidine are present as two monosubstituted guanidine groups which are replaced by two primary amino groups in strepta- mine. Further treatment with benzoyl chloride in pyridine yielded a product melting at 350° to 351 ° C, the analyses of which agreed fairly well for hexabenzoylstreptamine. Streptidine reduced two moles of periodatej streptamine, sixj diben- zoylstreptamine, twoj no formaldehyde was formed from any of these compounds. The fact that streptamine required six moles of periodate suggested to Carter et al. (114) that the four hydroxyl and two amino groups are located on adjacent carbon atoms, pointing to a cyclic struc- ture, since an open chain molecule should have yielded at least two moles of formaldehyde and required only five moles of periodate. Streptidine and streptamine were assigned the following formula : SUBSTANCES PRODUCED BY ACTINOMYCETES 199 NH-X A OH HO— *v i-NH-X OH Streptamine X = H Streptidine X= — C^ ^tt When streptomycin chloride is hydrolyzed with i.o N sodium hy- droxide, for three minutes at ioo° C. or for eighteen hours at 40° C, a weakly acidic substance, m.p. 161° -162° C. is obtained. It has been characterized as maltol, namely, The maltol gives a brilliant violet color with ferric chloride and a positive iodoform testj it reacts rapidly with nitric acid, and sublimes readily, even at 100° C. The benzoate melts at 114°-! 15° C. It has been isolated from hydrolyzates of streptomycin salts ranging in purity from 280 to 800 Mg/mg. The yields of maltol were about 30 per cent if one mole was derived from one mole of streptomycin. It was suggested that the formation of maltol by alkaline hydrolysis of streptomycin, measuring the ultraviolet absorption in acid solution, be used as an assay procedure, for the absorption produced is propor- tional to the initial antibiotic activity in preparations having a potency of 50 to 800 Mg/mg. The ferric chloride color reaction also appeared to be useful for this purpose. Streptomycin can be distinguished from streptothricin by inactiva- tion with cysteine. This property is not due to the sulfhydryl group alone. On oxidation of the cysteine, the substance is reactivated (179). In view of the specific sensitivity of different bacteria to streptothricin 200 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES and streptomycin, not only can they be readily differentiated from one another by their bacteriostatic spectra, but the admixture of one with the other can actually be measured quantitatively. B. mycoides can be used to measure the presence of a small amount of streptomycin with streptothricin, whereas bacteria resistant to streptomycin can be utilized for detecting the presence of a small amount of streptothricin or other antibiotic (858). Proactinomycin is produced by A^, gardneri grown in soft agar media, from which it is extracted by organic solvents, such as ether, amyl acetate, benzene, and carbon tetrachloride. It can be re-extracted in water by adjusting the fH to 4.0 with HCl or H2SO4. The aqueous extract is concentrated in vacuo and evaporated to dryness from the frozen state. A white powder, very easily soluble in water, is obtained. The yield of the material is 60 mg. from i liter of culture. The sub- stance is fairly stable, though boiling for 10 minutes at /)H 2.0 or -pYL 7.0 results in a small loss of activity. Boiling at fH. lO.O destroys the greater part of the antibacterial activity. Proactinomycin has basic prop- erties and is precipitated from aqueous solution by such base precipitants as picric acid, picrolonic acid, and flavianic acid. Proactinomycin is active in a dilution of i : 500,000 or more against gram-positive cocci, B. ant hr acts y and N. meningitidis ; it is much less active against gram-negative bacteria and is not very toxic to animal tis- sues, but definitely more so than penicillin or streptomycin. When given by mouth it can confer a considerable degree of protection against intraperitoneal infection with hemolytic streptococci. It is excreted in the urine and bile, and is absorbed from the alimentary canal. Repeated injections cause fatty changes in the livers of mice (273). SUBSTANCES PRODUCED BY FUNGI Penicillin Penicillin is produced by various strains of P. notatum and P. chry- sogenunty as well as by a variety of other fungi. The penicillin-like nature of an antibiotic substance is usually established by its chemical and biological properties: extraction in organic solvents at ^H 2 and re-extraction in water at ^H 7 ; inactivation by acid and alkali j partial inactivation by heating at 100° C. and ^H 7 for 15 minutes j complete SUBSTANCES PRODUCED BY FUNGI 201 inactivation by penicillinase and by copper ionsj inactivation by methyl alcohol J characteristic antibiotic spectrum, such as activity against S. aureus and not against E. coU (270). The strain of the organism used, the composition of the medium, and the conditions of growth greatly influence not only the yield of penicil- lin but also its chemical nature. Complex organic media containing glu- cose or brown sugar as a source of carbon are essential. Nitrate is used as a source of nitrogen j the medium also must contain a phosphate and certain other minerals. The supplementary addition of a stimulating substance in the form of yeast extract, corn steep, or certain vegetable juices is essential for the maximum production of penicillin. Since the organism produces an acid, probably gluconic, in the medium, some CaCOg must also be added. The metabolism of P. notatum in relation to penicillin production is illustrated in Figure 13 (p. 135). Four methods have been proposed for the growth of the fungus and the production of penicillin. These are: Surface growth in shallow liquid media; usually flasks, bottles, and other containers are employed, the depth of the medium being 1.5 to 2.0 cm. Submerged growth in liquid media; the vessels must be provided with proper stirrers and aeration Surface growth upon semi-solid media, including grain and bran (762) Circulation of medium through a column, the supporting material being made up of wood shavings or pebbles; the rate of flow of the me- dium is very important Since the various strains of penicillin-producing organisms vary greatly in their optimum conditions for the production of this antibiotic substance, different strains must be used for different conditions of cul- tivation. Penicillin is formed in the medium when active growth begins and reaches a maximum soon after the growth maximum, which occurs in 7 to 14 days in stationary cultures and in 3 to 7 days in submerged cul- tures, at 20° to 25° C. Penicillin is soluble in ether, acetone, esters, and dioxanej it is mod- 202 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES erately soluble in chloroform, slightly soluble in benzene and in carbon tetrachloride. It is soluble in water to the extent of 5 mg./ml. It is inactivated by oxidation and by evaporation at 40° to 45° C. in acid and in alkaline solutions, although it is fairly stable at fH 5 to 6. If the solutions are adjusted to ^H 6.8, it retains its potency for 3 months. The crude penicillin does not dialyze through a collodion membrane and resists heating at 60° to 90° C. for short periods; it remains active when heated at 100° C. for 5 minutes but not for 10 min- utes. The methods of isolation of penicillin from the culture media can be classified under the extraction and adsorption procedures. Fleming first reported that penicillin is insoluble in ether. This was found (146) to be due to the alkaline reaction of the filtrate; for at ^H 2.0 ether removes completely the antibacterial substance. The ether extract is evaporated with some water in vacuo at 40° to 45° C, the residual water containing the active substance, which is extremely labile. For practical purposes, penicillin is extracted from the acidified cul- ture by means of different organic solvents, such as ether or amyl ace- tate (6). It is then removed from the solvent by shaking with phos- phate buffer or with water at ^H 6.7. Since penicillin is rapidly de- stroyed at a high acidity, the first extraction must be carried out very quickly and at a low temperature. In the presence of the solvents, peni- cillin is stable for several days. The aqueous extract may be partly de- colorized by shaking with charcoal and filtering. The solution is cooled, acidified, and extracted several times with ether or amyl acetate; the extracts are passed through an adsorption alumina column, or through a 2.5 per cent precipitate of an alkaline earth carbonate on silica gel. Water may often contain a pyrogenic or heat-producing substance that must be removed from the penicillin. The following four main zones were recognized in the chromato- grams, beginning from the top : 1. A dark brownish-orange layer, the depth of which is inversely propor- tional to the amount of charcoal used for the decolorization ; this zone contains some penicillin 2. A light yellow layer containing most of the penicillin but none of the pyrogen SUBSTANCES PRODUCED BY FUNGI 203 3. An orange layer which contains some penicillin and some or all of the pyrogen 4. A brownish or reddish-violet layer which contains almost no penicillin; the pigment disappears on exposure to light The fourth fraction is discarded, and the others are eluted with M/15 phosphate buffer (-pH 7.2). The penicillin is again extracted with ether, then with water, sodium hydroxide being used to adjust the fH. Since penicillin is destroyed readily in alkaline solution, care must be taken in adding the alkali. The "nonpyrogenic" or "therapeutic" fraction, which contains about 80 per cent of the penicillin, is extracted with pyrogen-free water. It is a deep reddish-orange liquid, yellow in dilute solution, with a characteristic smell and bitter taste. Another method for obtaining penicillin has been suggested (638). In this method, the culture medium was adjusted to fH 3 to 4, satu- rated with ammonium sulfate and extracted with chloroform. The con- centrated chloroform extract was treated with phosphate buffer at fH 7.2 to remove the active substance. This process was repeated, the less active substance being separated from the active fraction by extraction with chloroform at different ranges. By precipitating the concentrated extracts from petroleum ether, the free acid form of penicillin was ob- tained. By saturating the chloroform-benzol solution with dry am- monia gas, an ammonium salt was obtained which gave a dark yellow microcrystalline powder. The salt was more stable than the acid form. By acetylating or benzoylating the ammonium salt a further increase in stability was obtained. This penicillin was strongly dextrorotatory and had an adsorption maximum of 2,750 A°. The preparation had an ac- tivity of 32,000,000 dilution units against hemolytic streptococci, which corresponds to about 240 Oxford units per milligram. When ether is used, the medium is adjusted to fH 3, extracted sev- eral times, the ether extract treated with dilute NaHCOa, the aqueous solution acidified and again extracted with ether j this is followed by shaking with excess of BaCOo, separating aqueous phase, filtering, and evaporating in frozen state (145). By the adsorption method, activated charcoal or fuller's earth is used (20 gm./L). The solution is first acidified to /)H 3.6, filtered, neutral- ized, treated with charcoal, and filtered. Ethanol is used to remove the 204 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES penicillin from the charcoal j the extract is evaporated, acidified, and treated with ether. Various procedures for large-scale production and recovery of penicillin have been described (503, 607, 768). Various modifications of these methods may be employed. In some cases, «-butyl alcohol is used for extraction. The culture filtrate is ad- justed to fH 6.4 and ammonium sulfate added, and the penicillin is extracted. When light petroleum ether and dilute sodium bicarbonate solution are added to the butyl alcohol extract, the penicillin is brought back into aqueous solution (48). The problem of drying is very im- portant (276). The barium salt was at first considered as the most suitable form for general use. In this form, penicillin retains its antibacterial activity for an indefinite period. It is soluble in absolute methyl alcohol but in- soluble in absolute ethyl alcohol. However, the Na and Ca salts are the common forms now used. Penicillin forms water-soluble salts with most heavy metals, except Fe"^+. Penicillin is unstable and readily inactivated by a number of reagents, including heavy metal ions, especially Cu, Pb, Zn, and Cd. Penicillin is stable toward light and atmospheric oxygen, but is oxidized by H^Oo and KMn04, the antibacterial activity being lost. In assaying penicillin, both biological and chemical tests are used. Of the former, the cup assay method is most commonly employed, al- though the turbidimetric and other tests are also frequently used. The different forms of penicillin are recognized by the differences in their effect upon various bacteria, notably S. aureus and B. subtilis. The world standard for penicillin has been defined as that activity which is present in 0.6 micrograms of the international penicillin stand- ard (384). One mg. of crystalline penicillin will thus contain 1,667 Oxford units (O.U.) and will be comparable to 84 million dilution units against S. aureus (168, 454). The chemical method for assaying penicillin is based upon the acidity produced by the action of a standard penicillinase solution upon the penicillin preparation and titrated to ^H 8.0 {666). The colorimetric method is based upon its interaction with an intensely colored primary amine, N-(l-naphthyl-4-azobenzene)-ethylenediamine to give amidic products containing acidic groups (845a). SUBSTANCES PRODUCED BY FUNGI 205 Tests are also made for sterility, moisture content, presence of pyro- genic substances, and toxicity (248). By means of adsorption, distribution between solvents, and reduc- tion, a barium salt or penicillin was at first obtained (5, 6) which was homogeneous by chromatographic analysis and gave 450 to 500 Oxford units per milligram of dry material. The active substance was found to be a salt of a strong dibasic acid with fH values approximately 2.3 and 2.5. The molecule contained one carboxyl, one latent carboxylic, two acetylatable, at least five C-Me groups, and no easily reducible double bond. The penicillin thus prepared was more sensitive to oxidizing agents than to reducing agents j it was unstable toward dilute acids and alkalies, and to heat (loss of COo), primary alcohols, and various heavy metal ions. Tentative suggestions were made concerning its chemical nature as follows: (a) a polysubstituted hydroaromatic ring structure j (b) the acidic groups (carboxyl) not conjugated with the chromophore responsible for the absorption j (c) the possible presence of a trisubsti- tuted a-unsaturated ketone grouping. With the introduction of new cultures for the production of penicil- lin, with the development of new methods for the growth of the or- ganism, as submerged vs. stationary, and especially with the employ- ment of synthetic media, it was found that several forms of penicillin are produced (735). P. chrysogenum x 1,612 was found to yield about 100 O.U./ml. The penicillin molecule is readily synthesized, especially when a phenyl linkage has been supplied. The addition of 3.3 gm./L of phenylacetic acid to the medium gave a maximum yield of 244 O.U./ml. P. notatum 1,984-A yields 40 to 50 O.U./ml. of penicillin on a purely synthetic medium, in presence of such factors as indole acetic acid or naphthalene acetic acid. The production of penicillin takes place in the presence of the following groups: I. Cysteine (or cystine in presence of a suitable reducing agent such as sulfite waste liquor) . 2. The — C — C — N — chain with the proper linkage at each end II I O H 206 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES 3. The phenyl ring, or preferably 2 and 3 combined as phenylacetu- rates, a-phenylacetamide or 3-phenylethylamine. Several forms of penicillin or "natural penicillins" have been iso- lated (715). They were all found to have the empirical formula C9H11O4SN2.R. These forms may be presented as follows: O H CH3 I S C CH3 P = CH3 CH2 CH--CH CH2 C — N— CH — CH * I \ 0=C N CHCO2H CH3 I OH S CCH3 Q = ^_^CH2C-N-CH-CH 0=C N CHCO2H X = HO^ -/ CH3 I OH S C CH3 ^ / / CH2C-N-CH-CH I \ 0=^C N CHCO2H CH3 I O H S CCH3 K^ / / = CH3 (CH2)6C-N-CH-CH I \ 0=C N CHCO2H In accordance with the nomenclature employed in the forthcoming monograph on the chemistry of penicillin, which is being prepared under the auspices of the National Academy of Sciences, penicillin G is designated benzylpenicillin j penicillin K, F, and X are designated, re- spectively, ;?-heptylpenicillin, A"-pentenylpenicillin, and ^-hydroxy- benzylpenicillin. SUBSTANCES PRODUCED BY FUNGI 207 The penicillins are strong monobasic acids of fK about 2.8. On treat- ment with hot dilute mineral acids, penicillins give i molecule of COo, an amino acid (penicillamine) and other products. The penicillamine (^-(3,(3-dimethylcysteine) belongs to the d or "unnatural" series of a-amino acids. Penicillin G yields on hydrolysis phenaceturic acid, phenylacetamide, and an aldehyde, CiqHuOoN. Phenylacetic acid was also identified as a hydrolytic product. The acidic group of penicillin was found to be identical with the carboxyl group in penicillamine j the addition of water to penicillin gives a second carboxyl j the new carboxyl breaks down to COo by the action of hot dilute mineral acid. The dicarboxylic acid thus produced, designated as penicilloic acid, is presumably the product of the action of penicillinase on penicillin. The molecular weight of penicillin ranges from 490 to 510. Penicillin is very unstable, especially in acid solutions and in an iso- lated form. Penatin P. notatum produces, in addition to penicillin, a second antibacterial substance designated as E. coli factor, penatin, notatin, and penicillin B. This substance is a flavo-protein and acts as a glucose-oxidase, oxygen being required. It is characterized by its action not only upon gram- positive but also upon many gram-negative bacteria, and by the fact that the presence of glucose is required for its activity. Its action is inhibited by the presence of catalase (157, 517, 786). Flavian Flavicin, a substance similar in every respect to penicillin, is pro- duced by A. flavus (950) j another substance, gigantic acid, is produced by A . giganteus {"]ii). Preparations of flavicin have also been designated as flavatin (748),aspergillin (103, 876), and flavacidin (606). This sub- stance was found to be largely penicillin F, with a potency of 1,400 O.U./mg. against S. aureus and a B. subtilis-S. aureus ratio of 0.72. The variable R group in the general penicillin formula is represented by — CHo.CHo.CH^CH.CHg. A small amount of penicillin G (R = — CHo.CgHg) was also detected (299). 208 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES As fer gillie Acid Aspergillic acid is produced by A. flavus. It is extracted from the tryptone medium by adsorption on norite and elution with ether. The pure acid has an m.p. of 93° C. (84° to 96° C.) and an optical activity of [a]D = +i4°. The formula C12H20O2N2 has been proposed for this substance. It possesses a hydroxyl group which gives it its acid na- ture (/)K 5.5). It is stable under acid and alkaline conditions and can be distilled with steam or in vacuo without loss of activity. When grown in brown-sugar-containing media, a closely related substance is formed, having the formula C12H20O3N2 and an m.p. of 149° C, with lower biological activity. It can also be extracted from the medium, at low ^H, by organic solvents (benzene or heptane). The pure aspergillic acid (m.p. 96.5° to 97.5° C.) was found to account for only 5 per cent of the yield, whereas another fraction (m.p. 118° to 121° C.) had about 75 per cent of the antibiotic activity of the medium (102, 220). Aspergillic acid is moderately active against various gram-positive and gram-negative bacteria, including S. aureus y E. coli, and M. tuber- culosis, as well as fungi. The addition of blood to the medium greatly reduces this activity. However, the antibiotic potency and the toxic effect rapidly disappear after systemic or intrathecal administration, the material being excreted in the urine in very small amounts (102). The addition of cobaltous ions (i:ioo,ooo) greatly enhances the tubercu- lostatic properties of aspergillic acid, especially in media low in iron (349). Bromo-aspergillic acid is 8 to 10 times as active against S. hemolyticus as aspergillic acid. Further studies of the chemical nature of the mate- rial showed it to be a pyrazine derivative (220). Citrinin Citrinin is produced by P. citrinum and other fungi ( 745 ) on a syn- thetic medium, with inorganic salts of nitrogen and with glucose as a source of carbon. The culture filtrate is acidified with HCl, and the sub- stance crystallized from boiling alcohol. Citrinin forms a monosodium salt which, at ^H 7.0 to 7.2, gives a virtually colorless solution in water. It is a yellow crystalline solid, m.p. 170° to 171° C. (decomp.). It is SUBSTANCES PRODUCED BY FUNGI 209 /-rotatory (in alcohol) and nearly insoluble in water. It changes in color from lemon-yellow at /)H 4.6 to orange-pink at ^H 5.6 to 5.8 and to cherry-red at fH 9.9. The addition of FeCls to the culture solution gives a heavy buff-colored precipitate, which dissolves in an excess of reagent to give an intense iodine-brown solution. Citrinin has little if any activity against gram-negative bacteria and about 50,000 dilution units against B. sub tills and S. aureus (33, 907). Penicillic Acid Penicillic acid was first isolated in 19 13 by Alsberg and Black (16) as a metabolic product of P. fuberulum. It is also produced by P. cyclofium. A limited air supply and an acid reaction of the medium favor the production of this acid. It is isolated (698, 703) by evapora- tion of the culture solution, the crude acid crystallizing on cooling. It is purified by recrystallization from hot water. Yields greater than 2 gm. per liter of culture were obtained. It is a monobasic acid, stable, colorless, appreciably soluble in cold water, giving a series of colorless and readily soluble salts (61, 487). It is optically inactive, and its m.p. is 87° C. (anhydrons), 64° to 65° C. (+ HoO). It has limited activity against gram-positive and gram-negative bacteria. It is also active against yeasts, and is toxic to animals when injected subcutaneously in concentrations of 0.2 to 0.3 gm. per kilogram weight. Fumigatin Fumigatin is a 3-hydroxy-4-methoxy-2,5-toluquinone or C8H8O4 (Figure 18). It forms maroon-colored crystals, has an m.p. of 116° C, and is water-soluble (23, 743). It is extracted with chloroform from acidified medium. It has limited activity against gram-positive bacteria. All quinones have been divided into three groups on the basis of their action on Staphylococcus: (a) those that have a markedly weaker anti- bacterial action than fumigatin, including toluquinone and some of its derivatives; (b) those that are somewhat more effective than fumi- gatin, including 3:4 dimethoxytoluquinone ; (c) those with activity perceptibly greater than that of fumigatin (methoxytoluquinones). The introduction of -OCH3 into the quinone nucleus results in an in- 210 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES crease in antibacterial activity. The introduction of an OH or the re- placement of -OCH3 by OH results in a decrease in activity. None of these quinones, however, has any very striking action on gram-negative bacteria, such as E. coli (325, 697). Clavacin Clavacin is anhydro-3-hydroxy-methylene-tetrahydro-Y-pyrone-2- carboxylic acid (Figure 18). It is produced by a number of fungi, and has also been designated claviformin, patulin, clavatin, and expan- sin. It is colorless, optically inactive, neutral, and readily soluble in water and most common organic solvents (445, 489) j it has an m.p. of 111° C. It is isolated either by preliminary adsorption on norite fol- lowed by removal with ether or chloroform, or by the direct treatment of the culture with ether. The extract is evaporated, leaving a brown substance J this is treated with a small amount of water, and the aqueous solution again extracted with ether. Clavacin crystallizes when the ether solution is concentrated, or after preliminary purification over a silica gel column. Clavacin is about equally active against gram-positive and gram- negative bacteria, its growth inhibition being about 200,000 dilution units. It is also strongly fungistatic. It is toxic to animal tissues, its lethal action upon mice being about 25 mg. per kilogram body weight when given intravenously or subcutaneously. Clavacin neutralizes the action of tetanus toxin and can thus be dis- tinguished from isoclavacin and its derivatives. This specific action was ascribed (739a) to the position of one double bond in clavacin. Fumigacin Fumigacin is produced by different strains of A. jumigatus. It is a colorless, monobasic acid, m.p. 212° C, /-rotatory in chloroform. It is insoluble in water except as sodium salt, sparingly soluble in methyl and ethyl alcohols, and readily soluble in acetone, ether, chloroform, and other organic solvents. It is extracted from the medium either by preliminary adsorption on charcoal followed by treatment with ether and alcohol, or by direct extraction of culture in accordance with the following method (631) : The culture filtrate is acidified to ^H 2 with SUBSTANCES PRODUCED BY FUNGI 211 phosphoric acid and extracted three times with ether, the combined ex- tracts equalling the volume of the filtrate. The ether is evaporated to one-tenth of its volume and the concentrate is shaken repeatedly with saturated sodium bicarbonate solution, which removes a dark-red pig- ment. The solution is then exhaustively extracted with 6 per cent so- dium carbonate solution. The ether phase, on evaporation, yields glio- toxin. The sodium carbonate solution is acidified and distributed several times with benzene 5 the partly crystalline residue from the benzene (7-12 mg. per i L of culture filtrate), on repeated recrystallization from methanol, yields pure fumigacin in the form of filamentous needles. Fumigacin melts with some decomposition at 2i5°-220° C, depending on the rate of heating. [oi]'J — — 132 ±: 2° (0.41 per cent in chloroform). The ultraviolet absorption curve shows only strong end absorption below 260 mp with E 'J^^^ = 298 at 234 mp. Fumigacin is markedly bacteriostatic against gram-positive bacteria, but not against the gram-negative forms. It is also active against tuber- cle bacilli, giving complete inhibition in a dilution of i : 10,000, and par- tial inhibition in i :ioo,000 dilutions (464). It is not very toxic to ani- mals (126). Gliotoxin Gliotoxin is produced by various species of Trichoderma, Glio- cladium, Asfergillus {A.fumigatus) and PenicilUum {P. obscurum) j2iS well as various other fungi. It has been analyzed as C13H14O4N2S2 (474). It is rapidly produced in an acidified (/)H 3.0 to 3.5) synthetic medium when grown in a submerged or shaken condition for 2 to 4 days. It is extracted from the culture medium by the use of chloroform. The latter is distilled off, and the residue is taken up in a small amount of hot benzene or 95 per cent alcohol, from which, on cooling, silky white needles crystallize. It is recrystallized from benzene or alcohol. It has an optical rotation of WY^ — — 2,39°, and an m.p. of 121° to 122° C. (991). Gliotoxin is frequently accompanied by one or more other antibiotic substances, A. jumigatus producing as many as three others. P. ob- scurum also produces one other. The removal of both from the medium is brought about by extraction, at ^H 2, with benzene. The addition of 212 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES equal parts of petroleum ether to the concentrated extract results in the separation of a crude preparation of gliotoxin. Purification is ac- complished by repeated crystallizations (66$). Gliotoxin is sparingly soluble in water and readily soluble in alcohol. It is unstable, particularly in alkaline solutions, and is sensitive to oxidation and to heating (988) j it is inactivated by heating for 10 min- utes at 100° C. ( 14). Its potency was found to be destroyed by bubbling oxygen for 5 minutes. Gliotoxin is active against various bacteria and fungi. It is toxic to Rh'fzoctoma hyphae in a dilution of i : 300,000, which is about two- thirds of the toxicity of HgCls- The crystals, as well as the crude ma- terial, were found to be toxic also to Trkhoderma, but the minimum lethal dose was about 40 times greater than that required for Rhizoc- tonia. Viridin Viridin is produced by Trkhoderma viride. It crystallizes in the form of colorless rod-like prisms, which decompose without melting at 217° to 223° C. It is extracted from the medium with chloroform, evaporated under reduced pressures, and crystallized from ethyl alco- hol. It is optically active 5 a i per cent solution in chloroform gives [a] ^^ = — 222°. The addition of phloroglucinol and HCl to a dilute alcoholic solution gives a deep reddish-violet color. It is unstable in aqueous solutions, but stable at ^H 3.5. It is highly fungistatic but not very bacteriostatic (84). Other Substances A number of other antibacterial substances have been isolated from fungi, but have not been adequately studied either chemically or bio- logically. It is sufficient to mention the following: Puberulic acid, a colorless, optically inactive, water-soluble dibasic acid, with an m.p. of 316° to 318° C, and puberulonic acid, a bright- yellow acid with an m.p. of 298° C, are produced (62, 704) by various species of PenkilUum (P. fuberulum) . The first is a quinol and the second is quinonoid. They have moderate activity against gram-positive bacteria. P. fuberulum also produces a photosensitive compound. SUBSTANCES PRODUCED BY FUNGI 213 C17H10N2O2, with an m.p. of 220° C. j it appears in the mycelium after 5 weeks' incubation and has certain antibiotic properties (108). Penicidin was isolated (29) from a species of PenkilUum. It is soluble in ether, alcohol, chloroform, and dilute acids, but not in petrol ether. It is destroyed by bases, and is adsorbed on active charcoal. It is similar to aspergillic acid in its antiluminescent properties. It is active against E. tyfhosa. Chaetomin is produced by a species of Chaetomium {Ch. cochliodes) grown in complex organic media. It is active largely against gram-posi- tive bacteria (948). Much larger concentrations of the material are found in the mycelium of the organism than in the culture filtrate j it is extracted from the former with acetone and from the latter with ethyl acetate. It is purified by washing with sodium carbonate, treated with petroleum ether, followed by chromatographic absorption. Chae- tomin contains nitrogen and sulfur, but it differs in biological activity from penicillin and from gliotoxin (326). Kojic acid (Figure 18) is produced by various species {A. oryzae and A. efusus). It possesses definite, even if limited, antibacterial proper- ties and is more active against gram-negative than gram-positive bac- teria j its antibiotic activity is not inhibited by serum (150, 465). It is particularly active against species of Leftos-pira (660). Polyporin is produced by Polystktus sanguineus grown for two to three weeks in various synthetic media. It is present in both the culture filtrate and the sporophores of the fungus. It is water soluble, and is ac- tive (bacteriolytic) against various gram-negative {E. tyfhosa^ V. comma) and gram-positive (6". aureus^ bacteria. It is nontoxic and nonhemolytic (77). Its activity is not affected by oral administration, by passage through a Seitz filter, or by pus and other body fluids and tissues. It protected animals against V . comma and E. tyfhosa infec- tions, and neutralized typhoid vaccine (77a). Mycophenolic acid is produced by P. brevi-com factum. It was so named by Alsberg and Black (16) in 1913, although it was first iso- lated by Gosio in 1 896 and is said (268) to be the first antibiotic to have 'been crystallized. It has only limited activity upon certain gram-posi- tive bacteria, but it has a considerable effect in inhibiting the growth of 214 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES fungi, especially those pathogenic to man and to plants. This effect is largely fungistatic, giving titers up to i :8o,000 (271). Glutinosin is produced by Metarrhizimn glutinosum grown on syn- thetic media (83a). The substance is extracted with ether, n-h\i.ty\ alco- hol, or petroleum ether. The solvent is evaporated and the material crystallized from ethyl alcohol in the form of thin, colorless plates, free from S and N. It does not melt at 300° C. It has specific antifungal ac- tivity, inhibiting spore germination, but does not possess any antibac- terial properties. SUBSTANCES PRODUCED BY YEASTS According to Fernbach (252), certain yeasts produce volatile sub- stances which are toxic not only to other yeasts but also to bacteria. Rose yeasts {Torula suganii), either fresh or heated to 120 to 130° C, were found (690) to contain a substance which has an antagonistic action against fungi, especially in the young mycelial stage, but not against yeasts ; the growth of A . niger was reduced by 60 to 70 per cent and that of A. oryzae by 25 to 30 per cent. The substance was not found in the ash of the organism and was not secreted in the filtrate, but re- mained in the yeast cells. An alkaline reaction was unfavorable to its formation and action. The active substance was soluble in acetone, alco- hol, ether, and chloroform, and was adsorbed by kaolin, Seitz filter, paper, and by the fungus mycelium. It could be removed from the kaolin by treatment with ether or acetone. Acetone-treated yeast no longer had an antagonistic effect, but only a stimulating one. According to Schiller (835), yeasts produce a bacteriolytic substance only in a state of "forced antagonism," that is, in the presence of staphy- lococci and certain other bacteria. The substance is thermolabile, since it is destroyed at 60° C. It is active also outside the cell. More recently ( 151 ), the active substance of yeast was concentrated. In a crude state, the active material was found to be nonvolatile and readily soluble in water, in 95 per cent alcohol, and in acetone containing a trace of water. It was stable at 100° C. at fH 7.3. It contained nitrogen but no sulfur. Although a positive biuret reaction was obtained, it appeared that the protein was present as an impurity. SUBSTANCES PRODUCED BY ANIMALS 215 The ability of Torulosfora utilis var. major to inhibit the growth of various gram-negative and other bacteria has also been demonstrated (112). SUBSTANCES PRODUCED BY ANIMALS To what extent antibacterial substances produced by animals and plants should be classified with the true antibiotics is open to question. In view of the fact, however, that these substances behave in a manner similar to antibiotics, they can be mentioned here. Lysozyme Fleming (260) found that egg white contains an enzyme, designated as lysozyme, that is active against certain bacteria, notably micrococci, bringing about their lysis. It is soluble in water and in dilute NaCl solu- tion. It is precipitated by chloroform, acetone, ether, alcohol, and tol- uene. It is not acted upon by pepsin or trypsin. Lysozyme has been demonstrated in most mammalian tissues and secretions, in certain vegetables, and in bacteria (905). It was found to be a polypeptide containing 16 per cent nitrogen and 2 to 3 per cent sulfur and having a molecular weight of 18,000 to 25,000. It is soluble and stable in acid solution, insoluble and inactivated in alkaline solu- tions, and inactivated by oxidizing agents (641 ). It diffuses in agar and through cellophane, and thus is markedly different from bacteriophage (332). It is fixed on the bacterial cells. It acts primarily upon the cell membrane of bacteria, the highly viscous component of the bac- terial cell (the mucoids), especially the sugar linkages of the complex amino-carbohydrates, being disintegrated by the enzyme. The degrada- tion of the bacterial polysaccharide to water-soluble products (N-acety- lated amino-hexose and a keto-hexose) by lysozyme is accompanied by complete lysis of some of the bacteria. In the case of other lysozyme- sensitive bacteria, such as B. suhtillsy no lysis occurs j apparently the morphological structure of these bacteria does not depend exclusively on the unaltered state of the substrate for lysozyme (242). The formation of a lysozyme-like material was also demonstrated (259) for a coccus isolated from dust. A sarcina susceptible to egg-white 216 CHEMICAL NATURE OF ANTIBIOTIC SUBSTANCES lysozyme also was found (641 ) to produce an autolytic enzyme similar to it. It has been suggested that the antibacterial action of saliva may be due to the presence in it of antagonistic bacteria. The lysozyme of saliva is known to act primarily upon gram-positive bacteria. A large number of bacteria were tested (31) for their ability to antagonize diphtheria and pseudo-diphtheria organisms. Only the spore-forming B. mesentertcus and B. suhtills groups produced antagonistic substances, but these bacteria were not found in the saliva. Cultures of bacteria isolated from the saliva had no antagonistic effect, thus proving that the action of saliva need not be due to its bacterial content. Milk Milk was found (694) to contain several thermolabile bactericidal substances and two thermostable compounds which acted injuriously upon lactic acid bacteria. Orla-Jensen emphasized that the growth of bacteria in milk is influenced by a combination of activators or growth- promoting substances and of inhibitors, the predominance of one or the other being determined by various conditions. These substances influ- ence the development of specific lactic acid bacteria during the spon- taneous souring of milk. Other Substances Various other animal tissues and fluids contain substances which pro- duce a bacteriostatic effect upon gram-positive bacteria. The method of isolation of these substances and their selective action are similar to those of tyrothricin (523a). SUBSTANCES PRODUCED BY HIGHER PLANTS It has been known for a long time that certain alkaloids and other plant extracts possess bactericidal properties. Sherman and Hodge (851) demonstrated in 1936 that the raw juices of cabbage, turnips, and horseradish possess antibacterial properties. The active substance in the juice could be adsorbed on activated carbon and by passage through fine Berkfeld filters. The substance was thermolabile, being destroyed at 60° C. in 10 minutes. SUBSTANCES PRODUCED BY HIGHER PLANTS 217 Osborn (696) examined as many as 2,300 different flowering plants, of which 134 species gave positive results. The activity against gram- positive bacteria was far greater than against gram-negative forms 5 cabbage, cauliflower, broccoli, and kohlrabi gave the greatest effect upon both groups of bacteria. The Ranunculaceae were most active of all the plants. The stability of the substances as regards wilting and dry- ing varied greatly. Extracts of honeysuckles {Lonicera tartar jia), espe- cially of the roots, are very active (583). Antibacterial substances are thus found to be widely distributed among higher plants. Some of these substances have been isolated and even crystallized. Allicin, a diallyl sulfoxide, was isolated from Allium sativum (iij). Crepin, an a3-unsaturated lactone, has been isolated from Crefis taraxacifoUa; the empirical formula C14H16O4 has been suggested for it. It is active against both gram-positive and gram- negative bacteria (391). Protoanemonin, obtained from buttercups, Ranunculus J is active against various bacteria and fungi (847). Toma- tin, an antibiotic occurring in the tomato plant, was found (456a) to be active against a variety of gram-positive bacteria and fungi, including both plant and animal pathogens. Lichens were also found capable of producing antibacterial activity. Of about 100 forms tested, 52 could inhibit either B. subtilis or S. au- reus or both ; gram-negative bacteria are generally not susceptible. Al- though certain lichen constituents were found to possess antibiotic prop- erties, no specific agent has been isolated (100). Spanish moss also pro- duces an antibiotic effect (999). Unbleached wheat flour was shown (889) to contain a protein which had bacteriostatic and bactericidal activity in vitro; although its activ- ity was greater against gram-positive organisms, it also had some ac- tivity against gram-negative types. The antimicrobial action of this protein can be neutralized by means of a phosphatide (1033), a reac- tion which may be due to the formation of a lipoprotein that has no longer any antibiotic activity. CHAPTER I I THE NATURE OF ANTIBIOTIC ACTION Sulfanilamide, -penicillin, and gramicidin can be clearly set afart from, the classical antiseptics which are general protoplasmic poisons. All three substances are primarily bacteriostatic rather than bactericidal in their action. Since they do not destroy the res- piration of bacteria, one may assume that the inhibition of growth which they cause depends not upon interruption of the celltdar metabolism as a whole, but rather upon some subtle interference with certain individual reactions. To interrupt the pathogenic career of an infectious agent, therefore, it is not necessary to kill the invading cell, but only to block one step in its metabolic path by some specific inhibitor. — Dubos. ANTIBIOTIC SUBSTANCES AND CHEMICAL DISINFECTANTS Since antibiotic substances vary greatly in their origin and in their chemical nature, they may be expected to vary also in their mode of action upon the cells of bacteria and other microorganisms, and in the effect upon the animal tissues when these agents are used for chemo- therapeutic purposes. Comparatively little is known concerning these mechanisms. It is known, however, that antibiotic substances act chiefly by interfering with the growth of the bacterial cell, although in many cases they are able to bring about the lysis of the cell as well. Because of the first effect, it has been assumed that antibacterial agents are struc- turally related to bacterial metabolites that usually function as co- enzymes (600). In this connection, the following properties of anti- biotic agents are of particular significance: Most antibiotics are strongly bacteriostatic and only weakly bactericidal, though a few are also strongly bactericidal and some are even bac- teriolytic. Some substances act primarily in vitro and only to a limited extent in vivo because of interference of the body tissues with their action; others, however, act readily upon bacteria in vivo. ANTIBIOTICS AND CHEMICAL DISINFECTANTS 219 A few antibiotic agents are fairly nontoxic to the animal body ; others are somewhat more toxic but can still be utilized; and some are so highly toxic that they offer little promise as chemotherapeutic agents. Antibiotic agents differ greatly in their solubility : some are water soluble ; others are alcohol soluble and only slightly soluble in water; and some are acids and react with alkali solution to form soluble salts. Some antibiotic agents are stable under a variety of conditions, whereas others are unstable. Some antibiotic substances are hemolytic, others have apparently no in- jurious effect upon blood cells. The latter can be used for general body treatment, whereas the former are suitable only for local ap- plications. Since antibiotic substances are selective in their action upon microorgan- isms, none can be expected to be utilized as general agents against all bacteria. This also points to the remarkable physiological differences in the morphology and physiology of bacterial cells, and to the dif- ferences in mode of action of the different antibiotics upon various bacteria. A comparison of the antibacterial action of the antibiotic substances produced by two bacteria will serve to illustrate some of the foregoing points. Pyocyanin, produced by Ps. aeruginosa, inhibits the growth of many gram-positive and gram-negative bacteria in dilutions as high as I : lOOjOOOj pyocyanase and hemipyocyanin have less activity upon the bacteria, but yeasts are more sensitive to them than to pyocyanin. Ty- rothricin, produced by B. brevis, is far more specific in its action, which is limited largely to gram-positive bacteria. The sensitivity of patho- genic fungi to these compounds also differs markedly. Some other striking differences are found on comparing two types of antibiotic sub- stances produced by fungi, namely, penicillin and clavacln, and two substances produced by actlnomycetes, namely, streptothricin and ac- tlnomycln. The bacteriostatic spectra of these four substances are re- corded In Table 't^S- The first of each pair has limited toxicity to ani- mals, and the second is highly toxic. Whereas penicillin acts largely upon gram-positive bacteria and only upon a few gram-negative organ- isms, streptothricin acts alike upon certain bacteria within each group. Clavacln and actlnomycin, both of which are highly toxic, differ simi- larly In their action upon bacteria, the first being largely active against 220 NATURE OF ANTIBIOTIC ACTION gram-positive and the second active against members of both groups. These four compounds show various other differences in the nature of their antibacterial action. Differences in the bactericidal properties of other antibiotic substances are brought out in Table 36. Various attempts have been made to compare the antibacterial action of antibiotic substances with that of organic antiseptics. According to Suter (890), the bactericidal action of a compound depends upon cer- tain physical and chemical characters} a property that determines the bactericidal action of the compound upon E. ty-phosa may be relatively unimportant in the case of another organism such as S. aureus. A sub- stance may have the same activity, as expressed by the phenol coeffi- cient, against two organisms and still differ markedly in its relative TABLE 35. BACTERIOSTATIC SPECTRA OF FOUR ANTIBIOTIC SUBSTANCES GRAM TEST ORGANISM STAIN S. aureus + S. aureus + S. lutea + B. subtilis 4- B. megatherium + B. mycoides + CI. welchii + Actinomyces sp. + Neisseria sp. - Br. abortus - Sh. gallinarum — Pasteurella sp. — Hemofhilus sp. - S. schottmiilleri — S. aertrycke - Ps. -fiuorescens — 5. marcescens — A . aero genes — E. coli E. coli - PENI- CILLIN 9,500* I jOOof 38,000* 19,000* 1,900* 5* i,50ot i,ooot 2,000t it Gluconic acid + HoOo. The antibiotic substance favors certain lytic mechanisms in the cell, whereby the latter is destroyed; this mechanism may be either sec- ondary or primary in nature. The antibiotic substance affects the surface tension of the bacteria, acting as a detergent; tyrocidine lowers the surface tension of the bacterial cell, thereby causing its death, possibly by forming a stable complex with it (200). The antibiotic substance may interfere with the sulfhydryl group which is essential for cell multiplication. This was shown by Fildes (254) to hold true for mercurials and other chemical antiseptics as well as for true antibiotics such as clavacin and penicillic acid (325 ). The interaction of sulfhydryl-containing compounds with antibiotics depends on the nature and concentration of the latter. It has been sug- gested (118) that the activity and specificity of an antibiotic are func- tions of several factors, such as its diflusibility into the microbial cell, its adsorption by various enzyme systems, its reaction with sulfhydryl groups of the enzymes or with other sulfhydryl-containing substances adsorbed by the enzyme. Gliotoxin and the active principles of Allium sativum and Arctium minus showed little specificity in reactivity toward the thiols, whereas penicillin, streptornycin, and the Asarum canadense antibiotic reacted more readily with those sulfhydryl compounds which contained basic amino groups in the vicinity of the — SH. Pyocyanin had intermediate properties (115b). 226 NATURE OF ANTIBIOTIC ACTION The theory of inner antagonism has been suggested (509). The bac- terial cell is said to contain two antagonistic groups, namely coagulants and lysinsj when the correlation between these two groups is disturbed the result is either agglutination and precipitation by the first or lysis by the second. The phage is given as an example of a free inner antago- nist, the lysinj reproduction of the phage is thus explained by the lik^ formation of lysins in multiplying cells. The action of antibiotic sub- stances and resulting cell lysis were also explained by the inner antago- nism. It has been postulated (170) that the action of growth-inhibiting substances may consist in prolongation of the lag phase, reduction of the growth rate, lowering of stationary population, or hastening the death of the bacteria. A bactericide has all these effects, whereas a bacterio- static agent may affect one stage selectively. When organisms are al- lowed to grow in the presence of an antibacterial agent, a greater con- centration of the latter is required to bring about a given effect upon the bacterial culture. On the other hand, bacteria subjected to the action of an antibiotic substance may develop mechanisms that render them resistant to the action of the substance, and some bacteria and fungi even may produce an enzyme, such as penicillinase, that brings about the destruction of the antibiotic substance. The antibacterial action of gramicidin was found (412) to be in- hibited by a cationic detergent, phemerol, whereas penicillin was not affected by either gramicidin or two cationic detergents, phemerol and zephiran. When gramicidin and penicillin were used together, their effect was only slightly additive (394) j however, penicillin and strepto- thricin exerted a marked additive effect upon bacteria sensitive to both of these substances (287). The inhibition of the antibacterial action of sulfanilamide by /)-amino-benzoic acid has been explained by the fact that the latter is a growth factor in bacterial nutrition. Competition for this growth factor between the bacterial cell and the bacteriostatic agent is responsible for the inhibition of the agent. In a similar manner pantoyltaurine, which is related to pantothenic acid as sulfanilamide is to /)-amino-benzoic acid, will inhibit the growth of hemolytic streptococci, pneumococci, MECHANISM OF ANTIBIOTIC ACTION 227 and C. difhtheriae, by preventing the utilization of pantothenic acid by these bacteria, for which it is an essential metabolite. Fildes (253) em- phasized that "chemotherapeutic research might reasonably be directed to modification of the structure of known essential metabolites to form products which can block the enzyme without exhibiting the specific action of the metabolite." Since /)-amino-benzoic acid has no such action on penicillin, it is assumed that its mode of action upon bacteria is dif- ferent from that of sulfanilamide. However, it was suggested that penicillin as well acts by inhibiting directly one or more enzymes, the difference being merely one of degree (648). The antibacterial activity of iodinin is neutralized by quinonesj this is probably due to the destruction of the iodinin, since the N-oxide is reduced by the organism (596, 597, 1009). Different anti-inhibitors are known for other antibiotic substances, as shown later. Numerous other examples of metabolite-antagonism can be cited. Since the nature and function of the various metabolites are so diverse, and there are so many ways of modifying their structure, the principle of interference with biological processes through the use of analogs of essential metabolites is considered as established (994). The interfer- ence is sometimes explained as a direct competition between the metabo- lite and its analog for some cellular component for which they both have great affinity. However, in addition to competition, other factors also operate. The majority of the interferences involve organisms that are unable to synthesize the essential metabolite the function of which is disturbed. Mcllwain recommended the use of an antibacterial index to repre- sent the minimal value of Ci/Cm, or the ratio of concentration of in- hibitor (Ci) just sufficient to prevent the growth of the organism, to the concentration of metabolite (Cm) present. The smaller the antibacterial index the more effective is the compound, therefore, as an inhibitor. With S. hemolytkus, the homopantoyltaurine was found to have an index of 20,000, the pantoyltauramide 2,000, and the pantoyltaurine 500. The indices vary for different organisms. E. coU and P. vulgaris synthesize their own pantothenate and are not inhibited by these ana- logs of pantothenic acid. The mechanism of the resistance is at present unknown. 228 NATURE OF ANTIBIOTIC ACTION The concentration of the active substance and the composition of the medium are highly important in modifying the activity of the sub- stance. Some antibiotic substances, like penicillic acid, lose considerable bacteriostatic activity when incubated with sterile broth or with sterile peptone water at fH 7 and 37° C. for i to 3 days (700) j a similar ef- fect was observed with certain simple amines and amino acids. The con- centration of the substances reacting with penicillic acid is diminished on autoclaving the peptone broth in the presence of 2 per cent glucose. The neutralizing or anti-inhibiting agent interacts with the antibiotic substance and neutralizes its antibacterial effect either in the absence or in the presence of the organism. Since not all antibiotics of microbial origin have been isolated in a crystalline state, confusion often resulted from the use of crude prepa- rations. Concentrated and partly purified actinomycetin had no appre- ciable lytic action upon living cells j however, the presence of a small amount of a highly bactericidal substance, which was especially active against gram-positive bacteria, resulted in the lysis of living bacteria by actinomycetin. This action was thus a result of the activity of at least two different agents present in one preparation (1002). ANTIBACTERIAL ACTION Chain and Florey (122) divided all antibiotic substances into two groups: 1. Antibiotics which react with protoplasmic constituents and kill both bacterial and animal cells, comparable to the action of "antisep- tics." These antibiotics can be further subdivided into (a) those that are active against both gram-positive and gram-negative bacteria, and (b) those that exert a selective antibiotic action, usually against gram-posi- tive organisms, such as gramicidin and actinomycin. The selectivity is not absolute, since gramicidin acquires strong bactericidal activity against gram-negative bacteria in the presence of protamines, due to the fact that protamines remove phospholipids, which inhibit the antibac- terial action of gramicidin. 2. Antibiotics which react with substances having a specific signifi- cance in the bacterial cell only. Some of these substances are largely ANTIBACTERIAL ACTION 229 growth inhibiting and can, therefore, be designated as "bacteriostatics." The bacteriostatics may be expected to be relatively nontoxic to animal cells. Antibiotics of this class have possibilities as chemotherapeutic agents for general administration and for the treatment of systemic in- fections. In order to determine whether an antibacterial substance has chemo- therapeutic potentialities, the effect of the antibiotic on bacterial respira- tion can be determined by using the Barcro ft- Warburg apparatus. If respiration is stopped by addition of the antibiotic in dilution of 1 : 1,000, the organisms may be said to have been killed, the substance being an antiseptic which will be toxic to animal tissues. If, however, the antibiotic produces little or no effect on respiration of the bacteria, there is a probability that the substance has chemotherapeutic possi- bilities. Chain and Florey further suggested that observations be made on: (a) the toxicity of the antibacterial substance to leucocytes, a wide gap between a toxic concentration and a bacteriostatic effect suggesting that the substance may be useful, at least for local application j (b) the effect of blood, pus, and tissue extracts on the bacteriostatic activity, inhibition of activity being due to chemical combination between the active sub- stance and a tissue constituent or to an inhibitory mechanism similar to that of /)-amino-benzoic acid for the sulfonamides; (c) the toxicity of the substance to mice when injected intravenously. Any therapeutically active substance will be excreted unchanged or little changed in the urine, since it does not combine with the tissue cells. Although Dubos (206) believed that none of the in vitro metabolic screening methods at present available is satisfactory in a search for new chemotherapeutic agents, Chain and Florey emphasized that those antibiotics which pass the above biological tests can be expected to be effective as general chemotherapeutic agents and to be worth further investigation with mouse protection tests. Tyrothricin The phenomenon of antibiotic action by a specific substance can best be illustrated by the action of tyrothricin upon bacterial cells. Five dis- tinct stages have been described (201 ) : 230 NATURE OF ANTIBIOTIC ACTION 1. Inhibition of growth. Certain gram-positive bacteria are inhibited by as little as i microgram or less of the substance per lo milliliters of nutrient broth or agar, thus giving an activity of i : 10,000,000 or more. 2. Bactericidal action consists in the killing of the bacterial cells, either in a washed state and suspended in saline, or in a growing state in broth culture. 3. Lytic activity comprises the rate of lysis of a suspension of bacterial cells. Streptococci, for example, are readily lysed by gramicidin, whereas staphylococci are acted upon more slowly and less com- pletely. 4. Inhibition of enzyme activity includes dehydrogenases or enzymes of respiration. Gram-positive cocci, incubated at 37° C, lose their abil- ity to reduce methylene blue in the presence of glucose, upon addi- tion of gramicidin. Since inactivation of the dehydrogenase takes place before any morphological changes are observed in the cells, lysis was believed to be a secondary process, following cell injury; hydrolytic enzymes, however, remained unaffected. 5. Protection of animals by the antibiotic substance against infection. Gramicidin and tyrocidine differ in chemical properties and in bio- logical activity. Gramicidin acts only against gram-positive bacteria, in- cluding pneumococci, streptococci, staphylococci, diphtheria bacteria, and aerobic spore-forming bacilli j meningococci and gonococci are not readily acted upon. Tyrocidine affects both gram-positive and gram- negative organisms. Gramicidin causes hemolysis of washed red cells, this hemolytic action being destroyed on heating. Tyrocidine causes lysis of many bacterial species. This action, however, is secondary, autolysis following the death of the cells. Peptones and serum inhibit the action of tyrocidine, but gramicidin is affected only to a limited ex- tent by these agents (617). Tyrocidine behaves as a general protoplasmic poison. The effect of gramicidin, on the other hand, is reversible. Staphylococci "killed" with gramicidin and no longer able to grow on organic media can be made to grow in the presence of certain tissue components. Gramicidin Is, therefore, not considered as a gross protoplasmic poison, but retains a good deal of its activity in animal tissues. When applied locally at the site of infection, gramicidin exhibits definite action against pneumococci ANTIBACTERIAL ACTION 231 and streptococci. When injected intravenously, however, it is almost completely inactive against systemic infection. . It was demonstrated by tissue culture technique (412) that the he- molytic effect of tyrothricin was due to the presence of gramicidin. When tyrothricin or gramicidin was heated in an aqueous suspension there was a loss of hemolytic and bactericidal activity. Tyrocidine, which is not very hemolytic, showed no marked toxic effect upon the leucocytic elements of the human blood in amounts up to 100 mg. per milliliter for 8 hours. Other investigators (757) have reported that the hemolytic activity of tyrothricin is inherent rather in the tyrocidine fraction, although gramicidin also exhibits a definite hemolytic action. The addition of glucose causes only slight inhibition of the hemolytic effect. Treatment with formaldehyde results in the lowering of the hemo- lytic and toxic activity of gramicidin, without reduction of antibacterial properties J this was interpreted as signifying that these properties do not necessarily depend upon the same molecular configuration (S^S)- Gramicidin was found to be effective, in amounts as low as i mg., upon a billion gram-positive organisms, whereas tyrocidine acted in 25 to 50 times that concentration in the absence of inhibitors (449, 450). Tyrocidine appeared to block all the oxidative systems of the bacteria studied, whereas gramicidin seemed to affect only certain individual reactions. Tyrothricin was reported (67) to inhibit enzymatic dehydrogenation not only of glucose but also of a number of other compounds, such as lactic acid, fumaric acid, and glutamic acid. Inhibition of dehydrogenase was parallel to inhibition of growth. Both substances were found to exert a protective antibacterial action in mice infected intraperitoneally with susceptible bacteria j gramicidin protected the animals at a level one-fiftieth as high as that required for tyrocidine. Both substances are toxic to animals when injected into the blood stream j they show little toxicity when applied locally by the sub- cutaneous, the intramuscular, or the intrapleural route j oral adminis- tr^ition is not accompanied by toxic effects, but such treatment is in- effective (758). Gramicidin remains active in the blood stream, but it has only weak 232 NATURE OF ANTIBIOTIC ACTION bacteriostatic properties and no bactericidal action. Tyrocidine is strongly bactericidal but it is inactivated by blood serum, hence it is limited to local applications. No specific effect was exerted by these sub- stances on respiratory or circulatory systems (793). According to Dubos (200), the retention of the stain by gram-posi- tive bacteria indicates a peculiar property of the cell wall of these or- ganisms. The addition of one microgram of gramicidin to a billion pneumococci, streptococci, and staphylococci is considered sujEficient to inhibit the growth of these organisms on subsequent transfers. This effect v/as said to be due not to an alteration of the protoplasm but to some specific interference with an essential metabolic function. Bacterial cells which have been inhibited by the action of gramicidin become viable again when cephalin is added to the medium. It was suggested that the ineffectiveness of gramicidin against gram-negative bacteria may be due to the presence of a phospholipid in these organisms. Different strains of S. aureus differ in their susceptibility to the ac- tion of tyrothricin. There is apparent adaptation of the organism to in- creasing concentrations of the substance. A marked increase in resist- ance of the infecting organism, after several weeks of therapy, was ob- served (752). Staphylococci grown in the presence of increasing con- centrations of gramicidin become resistant to inhibition by this sub- stance (81, 720). Both gramicidin and tyrocidine are said (206) to be surface-active compounds, their antibacterial action being inhibited by phospholipids. Tyrocidine behaves like a cationic detergent j it is bactericidal in buffer solutions for all bacterial species so far tested, with the exception of the tubercle bacillus. Gramicidin influences some energy-using process which would normally allow carbohydrate and phosphate storage. This effect is specific, since penicillin and sulfanilamide do not have the same effect upon the phosphate metabolism of staphylococci. On the other hand, like many surface detergents, tyrocidine modifies the surface of the bacterial cell in such a manner that vital soluble metabolites, such as nitrogen compounds, inorganic phosphate, and phosphate esters are washed out of the cell. Hotchkiss (449) concluded that although ty- rothricin and its constituents are more active against gram-positive than gram-negative organisms, Neisseriae respond more like gram-positive ANTIBACTERIAL ACTION 233 cocci, and gram-positive, spore-forming bacteria are insensitive to gramicidin J tyrocidine has more activity against gram-negative organ- isms and is more bactericidal, whereas gramicidin is primarily bacterio- static. Tyrocidine destroys immediately and irreversibly the metabolic ac- tivity of the bacteria, such as oxygen uptake and acid production. For most tissue cells, with the exception of spermatozoa, gramicidin is much less toxic than tyrocidine. It behaves like a specific inhibitor of certain metabolic reactions. It retains much of its activity in vivo. Tyrocidine brings about rapid cytolysis of the cells. There is a quan- titative relation between the concentration of the antibiotic and the number of cells lysed, namely i mg. for lo^ and o.i mg. for lo^ cells. The amino acid decarboxylases are not inhibited even by concentrations of tyrocidine of 0.3 mg./ml. (309). Other Antibiotics from S-p ore-forming Bacteria The other antibiotic substances isolated from spore-forming bacteria are characterized by bacteriostatic spectra quite difFerent from that of tyrothricin. This is brought out in Table 37. Some of these substances, TABLE 37. COMPARATIVE ANTIBIOTIC SPECTRA OF SUBSTANCES PRODUCED BY AEROBIC SPORE-FORMING BACTERIA TEST ORGANISM BACILLIN SUBTILIN SIMPLEXIN 5". aureus I.O 1.0 96.0 M. conglomeratus 2.0 1.0 - D. pteumoniae III 3-0 - 0.4 S. faraiyfhi 0.25 1 0.0 96.0 Pasteur ell a sp. 1.0 - 0.4 E. coli 2.7 1 0.0 2.7 E. tyfhosa 1.4 lO.O 2.7 From Foster and Woodruif (284). Note. Unit of activity is the amount of antibiotic required to inhibit S. aureus as test bacterium. like subtilin, are capable of destroying various bacterial toxins, such as diphtheria, tetanus, and others, as well as hemolysin (759). Subtilin was found to be similar to gramicidin in its effect upon sur- face tension, in producing hemolysis, even if more delayed, in killing 234 NATURE OF ANTIBIOTIC ACTION Entamoeba histolytica in 1 1400,000 dilution, and in cytolyzing T . equi- ferdum in i :2,000 dilution (20). Extracts of cells of B. subtilis with ether or chloroform in an acid medium (pH 2.5), redissolved in an aqueous alkaline solution (^H 8.5), were found to be active against staphylococci, E. coli, and M. tuberculosis (693). Penicillin In his first description of penicillin, Fleming recorded (261) that "it was noticed that around a large colony of a contaminating mould the staphylococcus colonies became transparent and were obviously under- going lysis." Penicillin was referred to by Fleming as a bactericidal agent and the conclusion was reached that it belonged to the group of slow-acting antiseptics, since staphylococci were completely killed only after an interval of 4^ hours, even in a concentration 30 to 40 times that required for complete inhibition of the culture in broth. Florey and Florey (275), however, concluded, as a result of in vitro experi- ments, that penicillin is bacteriostatic and not bactericidal, at least in concentrations suitable for chemotherapeutic purposes. This led Garrod to state that "penicillin is in a true sense an antiseptic rather than a germicide: it does not kill bacteria quickly." The action of penicillin was found to be affected by changes in temperature, reaction of the sub- strate, and age of the bacterial culture (320, 321 ). In addition to its marked bacteriostatic effect, penicillin has also been found to be decidedly bactericidal j this is accelerated by an increase in temperature from 4° to 42° C. but is impaired by an increase in acidity of medium between f¥l 7.0 and 5.0. The rapid drop in the number of bacteria within the first 15 minutes after application of the penicillin was interpreted (594) as indicative of its bactericidal action in vivo. Young cells are particularly susceptible, whereas mature cells are neither lysed nor readily killed. The bacteriolytic action of penicillin upon sensitive organisms is greatest at the maximum rate of multiplica- tion (507, 512). The lysis of bacteria by penicillin depends upon their ability to produce autolysin. Bacteria are resistant to the lysin when liv- ing and become sensitive to it after the cells have been killed by penicil- lin or by other agents. The rate of bacteriolysis is thus controlled by bacterial multiplication and production of autolysin (913). ANTIBACTERIAL ACTION 23 5 Penicillin is markedly sporicidal against sensitive organisms j this action is greater in milk than in water, especially if preceded by sub- lethal heating of the spores (162). Penicillin is active against spiro- chetes (399), including Treponema fallidum (225, 296). Penicillin is thus found to be actively bactericidal in a medium and an environment in which active multiplication of the bacteria occurs, since it acts best in good culture media such as broth or serum and poorly in water or saline solutions. Although penicillin kills large numbers of sensitive bacteria, it does not always kill all the bacteria present, but leaves a few cells that are resistant to its action. These soon begin to multiply, giving rise to a resistant culture. Penicillin affects a metabolic function of the bacteria during the early stages of their development. Certain antibacterial substances, like hel- volic acid, neutralize the effect of penicillin on the bacteria, whereas others, like sulfanilamide, have a synergistic effect. The latter is espe- cially well marked with strains of staphylococci that are naturally re- sistant to penicillin (121). Although penicillin is active primarily on gram-positive bacteria, it also has an effect on certain gram-negative bacteria, but not on the colon organism. Hemophilus, or Brucella. The gram-negative cocci can be divided into two groups, on the basis of their sensitivity to penicillin : N. gonorrhoeae, N. intracellular, and A^. catarrhalis, which are sensi- tive j and A^. -flava and other nonpathogenic Neisseriae, which are not sensitive. Some species of Hemophilus, such as H. ducreyi, are as sensitive to penicillin as is S. aureus, although less so than S. hemolyticus (659). High potency preparations of penicillin were found (431) to have an inhibitive effect even on E. coli. The susceptibility of gram-negative bacteria to penicillin is much greater in synthetic than in complex or- ganic media; in the case of the latter, various polypeptides and pos- sibly some amino acids appear to neutralize the effect of penicillin upon E. coli, the antagonism being partly removed by methionine (852). In studies on the effect of penicillin on bacteria in urine, it was shown (402) that 90 times the dose required to eliminate S. aureus will affect S. faecalis, 240-fold increase will act on P. vulgaris, and 900-fold will act on £. coli (880). Although Salmonella strains were inhibited by 236 NATURE OF ANTIBIOTIC ACTION only 2 units of penicillin per i ml. and P. vulgaris, E. ty-phosa, Shigella, Escherichia, and Aerobacter showed even greater resistance, it was still believed that concentrations of penicillin in the urine can be attained to inhibit the growth of these organisms (904). Penicillin is not active against pathogenic fungi, the growth of which may actually be stimulated by this antibiotic. However, it has some activity against A. bovis, the growth of which was inhibited by 0.0 1 O.U./ml. (496), and against certain other actinomycetes (193). Various forms of penicillin differ in their action upon specific bacteria. Welch et al. (998) have shown that penicillin X is more effective than commercial penicillin against certain bacteria but not against others, not only in the test tube but also in the animal body. These results have been confirmed, as shown in Table 38. S. aureus and B. subtilis are more sensitive to penicillin G than to X, on a weight basis j however, peni- cillin X is more effective than G on certain other bacteria. TABLE 38. INHIBITION OF GROWTH OF DIFFERENT BACTERIA BY TWO FORMS OF PENICILLIN (MICROGRAMS OF PENICILLIN PER Ml) RATIO G ORGANISM PENICILLIN G PENICILLIN X X S. aureus .040 .060 0.7 B. subtilis .059 .098 0.6 Pneumococcus Type I .019 .016 1.2 Pneumococcus Type II .007 .005 1-4 Streptococcus Group D 2.400 1.700 1.4 Streptococcus Group B .120 .066 1.8 Streptococcus Group A .010 .006 1.7 Er. rhusifathiae .097 .049 2.0 E. coli 81.000 46.900 1-7 From Libby and Holmberg (568). Note. Unit of activity calculated on the basis of 1,650 units/mg. for pure penicillin G, and 1,000 units/mg. for pure penicillin X. Different strains of the same organism show marked variations in their sensitivity to penicillin. For example, a study of 40 strains of hemolytic S. aureus isolated from patients in an Army hospital (725) gave a range of sensitivity from complete tolerance of 4 O.U./ml. to ANTIBACTERIAL ACTION 237 Inhibition by 0.002 O.U./ml. More than 40 per cent of the strains thus Isolated could be called resistant j this was especially true of the strains isolated from patients who received penicillin. The oxygen uptake of suspensions of staphylococci was not inhibited to any extent by the action of penicillin for 3 hours. In a concentration of I : ijOOOj after incubation for 24 hours at 37° C, the bacteria gave larger numbers of colonies on plating (5). Although 0.0 1 to o.i mg. of penicillin per milliliter was found (432-437) to be sufficient to inhibit the growth of 2,500,000 hemolytic streptococci (Group A), no con- clusion could be reached as to whether its action is truly bactericidal or merely bacteriostatic. Penicillin inhibits fibrinolysis by sensitive strains of S. -pyogenes; this phenomenon is believed to be connected with growth inhibition (183). A comparison was made of the amounts of crude penicillin and gramicidin required to bring about total inhibition of growth of bac- teria, on the basis of micrograms per milliliter of culture medium (rabbit's plasma and a serum extract of chick embryo) . The results were as follows: Penicillin Gramicidin D. fneumoniae 2.5-5.0 0.5-1.0 S. fyogenes 2.5 5.0 S. sal'tvarius 20-40 2.5-60 S. jae calls 200* 40-60 S. aureus 2.5-10 300* * Inhibition not complete at these figures. The two substances appeared to be as effective against bacteria in cul- tures containing growth tissue as in cultures in which no tissue was present (394,395). Inhibition of growth of 2 to 4 million hemolytic streptococci was ob- tained by the use of 0.03 \\g penicillin with an activity of 240 to 250 O.U./mg. (432, 437). No inhibition was obtained with peptone, /)-amino-benzoic acid, blood, or serum. The fact that both penicillin and sulfonamides act upon some bacteria and are ineffective upon others suggests a similarity In their mode of action (648). A marked differ- ence was found, however, in the action of penicillin and sulfonamides, / p t>- 0/ /! ryi ^ V) 1 /."*- Iq Q? Vr 1 - So S • J ''■ . . n u *-» k • - 2 z Ni.. \ _ 2 nTNv. *x <. ^ CM H \«?^^^'v ^'^ < >V ^N ^"^^^^ '■» -^^ N^^^-;^;^"-!!^^ •^' "^ 1 — '■ P o o>ooi^^ <») — H-l 2 - SP _ z 9 K y^ y'' ^-- (J < l_ 1 i,->^-r* .'' 1 1 1 1 1 1 o ^^x^^lll^ -^20^ s-aoAiA-^ns jo nHiiavoon 0^ ;*