Marine Biological Laboratory ^,pA July 26, 1958 Accession No. ~. D John Wiley and Sons, Inc. Clven By- NtfW York City - Place, f BACTERIAL FERMENTATIONS C I B A LECTURES IN MICROBIAL BIOCHEMISTRY 1956 H. A. Barker, Bacterial Fermentations 1957 E. P. Abraham, Biochemistry of Some Peptide and Steroid Antibiotics C I B A LECTURES IN MICROBIAL BIOCHEMISTRY BACTERIAL FERMENTATIONS By H. A. BARKER 1956 NEW YORK • JOHN WILEY & SONS, INC. London • Chapman & Hall, Ltd. CONTENTS CHAPTER 1 Biological Formation of Methane CHAPTER 2 The Chemistry of Butyric Acid-Butanol Fermentations 28 CHAPTER 3 Fermentations of Nitrogenous Compounds 57 7381 I vii CHAPTER 4° .._ V ■<& BIOLOGICAL FORM 1 OF METHANE Methane is the characteristic product of the microbial decomposition of organic compounds in the absence of oxy- gen. The escape of bubbles of marsh gas from decaying vegetation in swamps and lake beds is a conspicuous phe- nomenon which can hardly be overlooked by anyone who observes rivers or lakes, particularly in the late summer when the water is warm and microbial activity is high. The gas consists largely of methane with smaller amounts of carbon dioxide and sometimes a little hydrogen. Historical Background Although the escape of combustible gas from various places on the earth's surface was described in Roman times by Pliny and was undoubtedly known much earlier, credit for the discovery of the common formation of such gas in nature must be given to the Italian physicist Alessandro Volta.1 On November 14, 1776, Volta wrote a letter to a friend describing his unexpected discovery that "combusti- 1 BACTERIAL FERMENTATIONS ble air" was being formed continuously and in substantial quantities in all the lakes, ponds, and streams in the vicinity of Como in northern Italy. The initial observation was made in Lake Verbano. Volta found that to detect the gas it was necessary only to disturb the bottom sediments, whereupon a shower of bubbles would rise to the surface. He further recognized a close relation between the abun- dance of combustible gas and the amount of plant material in the sediment, and concluded that the former was derived from the latter. Volta collected samples of the gas from many different sites and determined the relative proportion of gas and air that would give the most vigorous explosion. Chemical knowledge of 1776 did not permit the char- acterization of Volta's inflammable gas. This was first accomplished in 1806 by William Henry, who showed that Volta's gas was apparently identical with the main constitu- ent of synthetic illuminating gas, which was later called methane. During the next sixty years several eminent scientists, including Bunsen and Boussingault, became interested in methane formation. Their work, which consisted mainly of collecting and analyzing gas samples from various sources, served to corroborate Volta's conclusion that methane is formed abundantly in nature and is generally associated with the decay of plant materials covered with water, but no new concepts were developed. The first definite indication that methane is formed by a microbiological process was obtained in 1868 by Bechamp,2 a student of Pasteur. Bechamp had previously studied the decomposition of sugar and starch that occurred when these substances were added to a simple inorganic medium con- taining chalk and incubated in the absence of oxygen. He ascribed the resulting fermentation to a living "ferment" which he thought was originally present in the chalk and BIOLOGICAL FORMATION OF METHANE therefore called Microzyma cretae. The description of this organism was remarkably vague; it was reported to be so small that even when examined under the highest magnifi- cation of the microscope it appeared only as a motile point. For some reason, now not altogether clear, Bechamp decided to find out whether the chalk and its microbe could also cause an anaerobic decomposition of ethyl alcohol. He was rewarded after some weeks by seeing the development of a vigorous fermentation which produced a large amount of methane, a little carbon dioxide, and a mixture of vari- ous fatty acids. Bechamp reported that when the gas evo- lution stopped the only "ferment" to be seen was M. cretae! Despite the naive interpretation of the microscopic obser- vations, Bechamp's experiment clearly showed that methane can be formed from ethyl alcohol and calcium carbonate by a process probably caused by microorganisms. This was the first demonstration of biological methane formation from simple carbon compounds. Unfortunately, Bechamp's work was overlooked by most subsequent investigators. More adequate proof of the microbiological origin of methane was provided by Tappeiner in 1882.3 He set up three identical anaerobic cultures provided with plant materials as substrate and with considerable amounts of the intestinal contents of ruminants as an inoculum and a possible source of soluble catalysts. One of the three cultures was treated with an antiseptic to inhibit bacteria without inactivating soluble "ferments," the second was boiled to destroy both bacteria and "ferments," and the third was left untreated. The microbial nature of the fermentation was deduced from the observation that only the untreated culture produced methane. During the last quarter of the 19th century, interest in the methane fermentation centered largely on the utiliza- tion of cellulose as a substrate. Since cellulose is the most BACTERIAL FERMENTATIONS abundant constituent of plants it was reasonable to sup- pose that it must be a major source of methane in decom- posing plant materials. Furthermore, methane had been shown to be formed in large amounts in the digestive tracts of herbivorous animals; therefore the possible role of methane-producing bacteria in cellulose digestion was of some concern to animal physiologists. The experiments of Popoff,4 Tappeiner, and Hoppe- Seyler5 on the utilization of cellulose by crude enrichment cultures of bacteria obtained from soil or the digestive tracts of herbivorous animals demonstrated that cellulose is in fact decomposed under anaerobic conditions, fre- quently with the formation of methane and other products including carbon dioxide, hydrogen, and acetic and butyric acids. At first this was taken to mean that the bacteria which attacked cellulose also form methane. However it was soon realized that another interpretation is possible, namely, that the methane is formed not by the cellulose- decomposing bacteria but by the action of other associated microorganisms on one or more of the products of the cellulose fermentation. This interpretation was supported by two lines of evidence. First, certain cellulose-fermenting cultures were found to produce carbon dioxide and hydro- gen but no methane. This proved that cellulose fermenta- tion is not necessarily associated with methane formation. Second, Hoppe-Seyler, Omelianski, and particularly Sohn- gen1 demonstrated that the products of cellulose fermenta- tion, such as formate, acetate, butyrate, ethanol, and even hydrogen and carbon dioxide, can be readily used as sub- strates by methane-producing bacteria. Although these results were consistent with the conversion of cellulose to methane by a two-stage process, they did not exclude the existence of methane-producing bacteria that attack cellulose. Omelianski (1895-1904) 6 tried to resolve BIOLOGICAL FORMATION OF METHANE this problem by isolating pure cultures of cellulose-fer- menting bacteria derived from soil. This approach proved to be unexpectedly difficult and was finally abandoned. Omelianski then tried more indirect methods. By analysis of the gas formed by various cellulose-fermenting enrich- ment cultures, he found that the ratio of hydrogen to methane varied greatly from culture to culture. Ultimately, by a combination of heating and transferring procedures, he was able to obtain cellulose-fermenting cultures pro- ducing either methane without hydrogen or hydrogen without methane. Both cultures appeared to contain only spore-forming bacteria which were similar in appearance. The organism presumed to be responsible for the formation of methane from cellulose was given the name of Bacillus methanigenes. Omelianski's conclusion that he had demonstrated the existence of a cellulose-fermenting, methane-forming bac- terium may be correct but is open to doubt. The descrip- tion of Omelianski's methods shows that he was not dealing with pure cultures. Consequently the possibility cannot be excluded that the observed methane formation was the result of the action of two types of bacteria, similar in morphology but differing in physiology. Subsequent research has provided no confirmation of Omelianski's con- clusion; none of the cellulose-fermenting bacteria since isolated in pure culture has been found to produce methane. The Methane Bacteria The formation of methane in decomposing organic mate- rials is the result of the action of a specialized physiological group of bacteria, often referred to as the "methane-pro- ducing bacteria" or, more simply, as the "methane bacteria." These bacteria are entirely different from the aerobic methane-oxidizing bacteria. The former produce methane BACTERIAL FERMENTATIONS from various organic and inorganic compounds; the latter oxidize methane to carbon dioxide and water. Enrichment and Isolation. The methane bacteria have not been studied as extensively as most other groups of bacteria of comparable scientific and practical importance. The reason for this is readily apparent. In order to study the biology and biochemistry of bacteria most effectively, it is necessary to use pure cultures, that is, to study one species at a time. Unfortunately, with the methane bac- teria, this elementary but basic requirement has been diffi- cult and in many instances impossible to achieve. Until 1936 all attempts to isolate pure cultures, or even to obtain growth of colonies in solid media, were unsuc- cessful.7 Consequently all the early studies and many of the more recent studies of methane bacteria were of neces- sity carried out with enrichment cultures, i.e., cultures in which the substrate and environmental conditions were chosen in such a way as to favor the development of certain species of methane bacteria, without, however, excluding a substantial and frequently significant number of other bacterial species, both methane-producing and non-produc- ing types. By the use of enrichment cultures it was possible to obtain considerable information about the existing mor- phological types of methane bacteria, the general environ- mental conditions that favor their development, the kinds of substrates attacked and the over-all chemical reactions involved, but many basic points of biology, nutrition, and biochemistry could not be studied effectively with these cultures. Since 1936 four species of methane bacteria (Metha- nobacillus omelianskii/ Methanobacterium formicicum* Methanosarcina barkerii,9 Methanococcus vannielii10) be- longing in three different genera have been isolated in pure culture. An additional four species (Mbact. suboxy- BIOLOGICAL FORMATION OF METHANE dans,11 Mbact. sohngenii,7 Ms. methanica, Mc. mazei) have been purified by isolation and reisolation of colonies from agar media until they were free of other species of methane bacteria, but they may still have been contaminated by other anaerobic bacteria. The complete or partial purifica- tion of these eight species indicates that the isolation of pure cultures of other species should also be possible. It must be emphasized, however, that the isolation of methane bacteria is generally a difficult and time-consuming task requiring great patience and considerable experience with anaerobic culture techniques. So far, pure cultures have been obtained in only three laboratories. Even after a pure culture has been isolated its preservation is no easy problem. Most of the pure cultures have been lost after a few years. Only one species, Mbac. omelianskii, has been kept in pure culture for a long period of time. Because of the difficulties encountered in the isolation and maintenance of pure cultures of methane bacteria only a rather limited amount of information has been obtained concerning the biological and physiological characteristics of the group as a whole. Nevertheless some tentative gen- eralizations can be made that are useful in defining the group and indicating the probable behavior of its members under a variety of circumstances. Relation to Oxygen. All of the species are strictly anaerobic; they develop only in the absence of oxygen and in the presence of a suitable reducing agent.7 The methane bacteria are much more sensitive to oxygen or certain other oxidizing agents, such as nitrate, than are most other an- aerobic bacteria. For this reason they can be grown more easily in liquid or semisolid media than on the surface of an agar medium. Even with liquid media not fully pro- tected from air, sufficient oxygen may leak in to inhibit cultures that have passed the peak of their activity. The BACTERIAL FERMENTATIONS beneficial effect on the growth of methane bacteria caused by the addition of solid sediments, such as diatomaceous earth and shredded asbestos, to a liquid medium may be attributed at least partially to a mechanical shielding of the bacteria from dissolved oxygen, although other explana- tions are possible. The most satisfactory reducing agents for pure cultures are generally sodium sulfide and hydrosulfite. Care must be used in adjusting the concentration of these substances since above the optimal level they become toxic to some species. Mylroie and Hungate12 have recently described the successful use of hydrogen and palladium chloride as a reducing system in media for Mbact. formicicum. The effectiveness of this reducing system supports the conclusion that a low oxidation-reduction potential in the medium is more important than the presence of a specific concentra- tion of a reduced sulfur compound per se. Energy Metabolism. A second characteristic of methane bacteria is their energy metabolism which is specialized for a process that produces methane as a major product. A little later we shall take a closer look at the nature of the energy-yielding reactions. Now it is sufficient to empha- size that the ability to form methane is not a common and widely distributed property of anaerobic bacteria but is restricted to a specialized group. A few possible excep- tions to this generalization are recorded in the literature, particularly in the work of Laigret.13 This investigator reported that Clostridium perfringens, which normally does not produce methane, can be induced to do so in a peptone- formate medium by the addition of a small amount of iodine. However, because of the absence of experimental details, the validity of Laigret's observations is difficult to assess. A thorough reinvestigation of this problem is nec- essary before a definite conclusion can be reached concern- BIOLOGICAL FORMATION OF METHANE ing the ability of CI. perfringens and other species to form methane. Compounds Fermented. Methane bacteria specialize not only in the chemical mechanisms of energy metabolism but also in respect to the types of substrates which they utilize. For some obscure reason, all of the methane bacteria whose substrate requirements have been studied effectively by the use of pure or nearly pure cultures are unable to decompose the more usual substrates for bacteria such as carbohydrates and amino acids. Early reports of cellulose- or glucose-fermenting methane bacteria have not been confirmed; probably they were based on observations with mixed cultures containing both carbohydrate-ferment- ing and methane-producing species. Of course the possi- bility that still unknown species of carbohydrate- or amino acid-fermenting, methane-producing bacteria do exist can- not be excluded, but convincing evidence for such organisms is still lacking. As a group the methane bacteria appear to be restricted to the utilization of relatively simple organic and inorganic compounds, many of which are products of the better- known types of bacterial fermentations. Table 1 gives a list of the oxidizable substrates that have been shown by studies with pure or nearly pure cultures to be decomposed by methane bacteria. These substrates fall into three groups, the lower normal fatty acids containing from one to six carbon atoms, the normal and iso alcohols contain- ing from one to five carbon atoms, and three inorganic gases, hydrogen, carbon monoxide,14 .and carbon dioxide. In Table 2 are listed some additional compounds that are probably decomposed by methane bacteria, judging from observations made with enrichment cultures. These com- pounds include several long-chain fatty acids, dicarboxylic acids, aromatic compounds, acetone, and butylene glycol. 10 BACTERIAL FERMENTATIONS They have all been shown, mainly by Buswell and his collaborators,15-16 to be converted more or less quantita- tively into methane and carbon dioxide by crude cultures TABLE 1 Compounds Known to Be Decomposed by Methane Bacteria (Based on experiments with pure or purified cultures) Fatty Acids Alcohols Gases Formic Methanol Hydrogen Acetic Ethanol Carbon monoxide Propionic w-Propanol Isopropanol Carbon dioxide n-Butyric n-Butanol Isobutanol n-Valeric n-Pentanol n-Caproic derived from sewage sludge. Since these compounds are not known to be attacked by other anaerobic bacteria under the test conditions used, they are probably decomposed by methane bacteria. TABLE 2 Some Compounds Probably Decomposed by Methane Bacteria (Based on experiments with enrichment cultures) n-Caprylic acid n-Capric acid Isobutyric acid Stearic acid Oleic acid Benzoic acid Phenylacetic acid Hydrocinnamic acid Cinnamic acid Oxalic acid Succinic acid Acetone 2,3-Butylene glycol In the older literature concerned with anaerobic decom- position processes in nature it is often stated that many other compounds such as sugars, cellulose, proteins, purines, glycerol, lactate, and citrate are also suitable substrates for BIOLOGICAL FORMATION OF METHANE n a methane fermentation. This is true in the sense that these substrates are decomposed by crude cultures with the formation of methane and carbon dioxide. However, since these compounds are rapidly decomposed by many common bacteria that do not produce methane, it is virtually certain that the conversion of these substrates to methane is a two- or multistage biological process in which various bacteria that cannot form methane convert the substrates to volatile fatty acids, alcohols, and other common fermentation products which are then transformed by methane bacteria to their characteristic products. Species Substrate Specificity. In addition to the apparently severe limitations of the methane bacteria as a group with respect to utilizable substrates, each species characteristically is restricted to the use of a few com- pounds. A few examples of the substrate specificity of different species are shown in Table 3. Methanobacterium TABLE 3 Substrate Specificity of Some Methane Bacteria Species Oxidizable Substrates Methanobacterium for micicum H2, CO, formate Methanobacillus omelianskii H2, ethanol, primary and secondary alcohols Methanobacterium suboxydans Butyrate, valerate, caproate Methanosarcina barkerii H2, CO, methanol, acetate formicicum9 oxidizes only hydrogen, carbon monoxide and formate. Methanobacillus omelianskii cannot oxidize car- bon monoxide or formate but specializes in the decom- position of primary and secondary short-chain aliphatic alcohols17 and hydrogen.18 Methanobacterium suboxydans cannot attack any of the previously mentioned compounds except possibly hydrogen which was not tested, but spe- 12 BACTERIAL FERMENTATIONS cializes in the oxidation of C4 to C6 fatty acids to acetate and propionate.11 Finally, Ms. barkerii is restricted to the decomposition of hydrogen, carbon monoxide, methanol, and acetate.9 This species does not attack longer chain alcohols or fatty acids. Since all the available evidence indicates that these species are representative of the group, it is apparent that methane bacteria show a rather extreme substrate specificity. This implies that several species of methane bacteria must be required for the complete methane fermentation of the variety of compounds that are present in sewage and decomposing plant materials. Even for the complete fer- mentation of so simple a compound as valeric acid, as many as three species of methane bacteria may be required. Valerate is oxidized by Mbact. suboxydans to acetate and propionate, which are not further attacked by this or- ganism.11 Methane is formed in a coupled reduction reaction. A second species, Mbact. propionicum, converts the propionate to acetate, carbon dioxide, and methane. Since this organism cannot attack acetate, a third species such as Mc. mazei is required to ferment the acetate. The need to establish a balanced population of bacteria to participate in this type of fermentation sequence undoubt- edly explains why considerable time is always required to develop a culture capable of causing a rapid and complete fermentation of complex mixtures of organic compounds. When such a culture is obtained, it is capable of main- taining itself more or less indefinitely when a fresh supply of organic materials is added continually, because the major products of the fermentation are gases which escape from the medium leaving behind very little in the way of poten- tially toxic byproducts. Nutrition. The general nutritional requirements of methane bacteria appear to be very simple, although this BIOLOGICAL FORMATION OF METHANE 13 can be said with certainty only for those species which have been studied in pure culture. These species all grow satisfactorily in media containing the usual nutritive salts, carbon dioxide, a reducing agent, a single oxidizable com- pound suitable for the organism, and an ammonium salt as a source of nitrogen.8 One species, Mbac. omelianskii, has been shown to utilize nitrogen gas.19 The general physiological properties of the methane bacteria suggest, by comparison with other nitrogen-fixing bacteria, that the ability to fix nitrogen may be a common characteristic of the group. The addition of extracts containing amino acids, growth factors, and other nutritional supplements to synthetic media does not have a beneficial effect on the rate or magnitude of growth of the few species that have been studied in this respect. Special attention should be drawn to the carbon dioxide requirement of methane bacteria. Qualitatively, there is nothing unusual in this since all microorganisms require small amounts of carbon dioxide to initiate and maintain growth. However, several species of methane bacteria need a large amount of carbon dioxide which they use as a major substrate.20 The specific role of carbon dioxide in the energy metabolism of methane bacteria is considered in a later section. pH Range. Another environmental factor that is im- portant for the methane bacteria is the hydrogen ion concentration. As a general rule these bacteria are most active in the pH range from 6.4 to 7.2. Below pH 6 and above pH 8 the growth rate of methane bacteria falls off rapidly. At the high pH this may be the result of the greatly diminished concentration of free carbon dioxide. There is one exception to the generalization that an alka^ line medium is unfavorable. The formate-fermenting species Mc. vannielii grows best between pH 8 and 9.10 14 BACTERIAL FERMENTATIONS So far no methane bacteria preferring acid media have been reported. Nevertheless the slow formation of methane in peat bogs having a pH as low as 4.0 suggests that acid tolerant species probably exist. Taxonomy. The system of classifying methane bacteria depends upon the relative importance given to morphology and physiology. The editors of Bergey's Manual of Deter- minative Bacteriology21 have placed primary emphasis upon TABLE 4 Classification of Methane Bacteria Family : M ethanobacteriaceae A. Rod-shaped cells I. Non-sporulating : Methanol acterium 1 . Mbact. formicicum : formate, carbon monoxide, hydrogen 2. Mbact. propionicum : propionate 3. Mbact. sohngenii: acetate, butyrate II. Sporulating: Methanobacillus 1. Mbac. omelianskii: primary and secondary alcohols, hy- drogen B. Spherical cells I. Cells not in sarcina arrangement: Methanococcus 1. Mc. mazei: acetate, butyrate 2. Mc. vannielii'. formate, hydrogen II. Cells in sarcina arrangement: Methano sarcina 1. Ms. barkerii: methanol, acetate, carbon monoxide, hydro- gen 2. Ms. methanica: acetate, butyrate (?) morphology, and consequently have dispersed the methane bacteria among the families and genera of better-known bacteria of similar forms. In contrast, those investigators who have had personal experience with the methane bac- teria have been impressed by the striking physiological characteristics of all members of the group, and have pre- ferred to think of them as belonging to a physiological BIOLOGICAL FORMATION OF METHANE 15 family, which might well be called Methanobacteriaceae. Within the physiological family, four genera can be recog- nized on the basis of morphology (Table 4) . The known methane bacteria are either rods or cocci. The rod-shaped organisms may be further subdivided into two groups on the basis of sporulation. The non-sporulating types are placed in the genus Methanobacterium, and the sporulating types in the genus Methanobacillus* The spherical organ- isms whose cells occur singly, in pairs, or in irregular masses are put in the genus Methanococcus, whereas the organisms with more or less spherical cells arranged in cubical sarcina packets are included in the genus Methano- sarcina. The species in these four genera differ with respect to size and arrangement of cells, presence or absence of flagella, or the ability to utilize various substrates. Sub- strate specificity has been of particular value in species identification; therefore the substrates known to be attacked by different species are included in Table 4 following the species name. So far, a total of eight species of methane bacteria have been described in these four genera. These species have been obtained entirely from sewage sludge and black mud. Undoubtedly many more species will be dis- covered as more adequate isolation methods are developed, and as the organisms associated with other habitats, such as the digestive tracts of ruminants, are systematically investigated. Chemistry of Methane Fermentations The early studies on the chemistry of methane fermenta- tions were done entirely with enrichment cultures contain- ing a single substrate and a mixture of organisms. The # The one species of this genus that has been studied extensively, Mbac. omelianskii, has previously been included in the genus Methano- bacterium. 16 BACTERIAL FERMENTATIONS main result of such studies was the demonstration that a variety of compounds can be more or less quantitatively converted to methane and carbon dioxide.22 Equations summarizing some of the observed reactions are given in Table 5. The equations show that fermentations of acetate, TABLE 5 Fermentation of Organic Compounds by Mixed Cultures CH3COOH — > CH4 + C02 4CH3CH2COOH + 2H20 — > 7CH4 + 5C02 2CH3GH2CH2COOH + 2H20 — > 5CH4 + 3C02 2CH3CH2OH — > 3CH4 + C02 CH3COCH3 + H20 — > 2CH4 + C02 propionate, butyrate, ethanol, or acetone give the same products, the ratio of methane to carbon dioxide depending on the oxidation state of the substrate. The remarkable aspect of this result is that the nature of the products is independent of the structure of the substrate. A plausible explanation for the carbon dioxide was easily found: carbon dioxide might be formed by the complete oxidation of part of the substrate. But it was difficult to understand why methane was always the other product. For one sub- strate, acetate, a simple and reasonable explanation was suggested, namely, that the acid is decarboxylated to form carbon dioxide from the carboxyl group and methane from the methyl group.22 Unfortunately this explanation is of no help with propionate which, according to the hypothesis, should yield ethane on decarboxylation. Actually no ethane could be detected by ordinary analytical methods. No generally applicable chemical explanation for the origin of methane was developed until 1934, when van Niel BIOLOGICAL FORMATION OF METHANE 17 suggested the so-called "carbon dioxide reduction theory."20 This theory postulates that the organic compounds fer- mented by methane bacteria are oxidized completely to carbon dioxide, and that this oxidation is coupled with a reduction of some or all of the carbon dioxide to methane. The application of this idea to the fermentation of acetate is shown in the following equations. Oxidation: CH3COOH + 2H20 — > 2C02 + 8H (1) Reduction : 8H + CO2 — ■> CK4 + 2H20 (2) Net: CH3COOH — > CH4 + C02 (3) The oxidation of acetate, according to the theory, gives 2 moles of carbon dioxide and 8 equivalents of hydrogen. One mole of carbon dioxide is reduced to methane and water. The net result, the conversion of 1 mole of acetate to 1 mole each of methane and carbon dioxide, is in accord- ance with the experimental facts. The main evidence for the general theory at the time it was developed was provided by the early observations of Sohngen22 on the fermentation of a mixture of hydrogen and carbon dioxide. Sohngen found that enrichment cul- tures could couple the oxidation of hydrogen with the reduction of carbon dioxide to methane according to equa- tion 4. Sohngen's experimental observations were later 4H2 + C02 — > CH4 + 2H20 (4) confirmed in several laboratories by the use of pure cultures.14,18 Much of the subsequent work on the chemistry of the fermentation has been concerned with determining whether the carbon dioxide reduction theory also applies to fer- 18 BACTERIAL FERMENTATIONS mentations of organic compounds. By now a considerable amount of information has accumulated which shows that the theory is correct for fermentations of many, but not all, substrates. Examples of the reduction of carbon dioxide to methane are provided by the fermentations of ethanol by Mbac. omelianskii, of butyrate by Mbact. suboxydans, and of pro- pionate by Mbact. propionicum. With all three of these species, the interpretation of the experimental data is greatly simplified by the fact that they oxidize their substrates to acetate which is not further decomposed. Methanobacillus omelianskii oxidizes ethyl alcohol almost quantitatively to acetate according to equation 5.17 The 2CH3CH2OH + C02 — ► 2CH3COOH + CH4 (5) oxidation of alcohol is completely dependent upon the supply of carbon dioxide; when the carbon dioxide is all consumed the oxidation of alcohol stops. Other oxidants, such as sulfate and nitrate, cannot replace carbon dioxide. Now the above equation strongly indicates that the acetate is derived from the alcohol, and that the methane is derived from carbon dioxide. To confirm this interpretation by an independent method, an experiment was done in which unlabeled ethyl alcohol was incubated with C14-labeled carbon dioxide.23 At the end of the fermentation, the C14 content per mole of the methane was found to be essen- tially equal to that of the carbon dioxide, thus proving that the methane produced in this fermentation is derived entirely from carbon dioxide. The fermentation of butyrate by Mbact. suboxydans is very similar in principle.11 The oxidation of 2 moles of butyrate to 4 moles of acetate is coupled with the reduc- tion of 1 mole of carbon dioxide to methane according to BIOLOGICAL FORMATION OF METHANE 19 equation 6. In this fermentation also tracer experiments have shown that at least 98% of the methane is derived from carbon dioxide. 2CH3CH2CH2COOH + 2H20 + C02 — > 4GH3COOH + CH4 (6) In the preceding examples of methane fermentations involving carbon dioxide reduction, essentially no carbon dioxide is formed by the oxidation of the organic sub- strate. The fermentation of propionate by Mbact. pro- pionicum, shown in the following equations, is somewhat more complicated because it involves both carbon dioxide formation and utilization.11 Oxidation : 4CH3CH2COOH + 8H2O — > 4GH3COOH + 4C02 + 24H (7) Reduction : 3C02 + 24H — > 3CH4 + 6H20 (8) Observed : 4CH3CH2COOH + 2H20 — > 4CH3COOH + C02 + 3CH4 (9) The observed reaction, shown in equation 9, results in the production of 1 mole of carbon dioxide from 4 moles of propionate. However, although 1 mole of carbon dioxide accumulates, it is reasonable to assume, on the basis of what is known about the oxidation of propionate in other bio- logical systems, that the oxidation of 4 moles of propionate initially gives 4 moles of acetate, 4 moles of carbon dioxide, and 24 "hydrogens" as is shown in equation 7. If these 24 "hydrogens" serve to reduce carbon dioxide, 3 moles would be converted to methane as is shown in equation 8. This accounts for the observed yields of methane and carbon dioxide (equation 9) . 20 BACTERIAL FERMENTATIONS The postulated intermediate role of carbon dioxide was tested by experiments with C14-carbon dioxide or C14- propionate. Without discussing the experimental data in detail, it may be stated that the results could be interpreted to indicate that approximately 1 mole of carbon dioxide is formed per mole of propionate, in accordance with equa- tion 7, and that carbon dioxide is a precursor of most, if not all of the methane, as indicated in equation 8. Buswell and his associates24 have also done tracer experi- ments on the fermentation of propionate using enrichment cultures, presumably containing a mixture of species, capa- ble of converting propionate completely to carbon dioxide and methane without the accumulation of acetate. These cultures were shown to use carbon dioxide and to convert all three carbon atoms of propionate to both methane and carbon dioxide in varying degrees. Part of these results can be interpreted in terms of the reactions already discussed in combination with a secondary decomposition of acetate. However, some additional reaction would be required to account for the preferential conversion of the a carbon of propionate to methane and a substantial conversion of the /? carbon to carbon dioxide. These results could be caused by the formation of a symmetrical intermediate like suc- cinate from propionate. This would allow a randomization of the a and ft carbons of propionate and therefore would account for the similarity in their behavior. Other expla- nations are also possible. This type of problem probably ( nnot be resolved until pure cultures of all of the organ- i ms involved are available. The results which I have mentioned, as well as others, show conclusively that, with many substrates and at least several species of methane bacteria, methane is formed mainly or wholly by reduction of carbon dioxide. How- ever, at least two compounds, methyl alcohol and acetate, BIOLOGICAL FORMATION OF METHANE 21[ are converted to methane by chemical pathways not involv- ing carbon dioxide. Schnellen,9 working in Kluyver's laboratory, showed that methyl alcohol is readily fermented by a species of Methano- sarcina according to equation 10. At first this reaction was 4CH3OH — > 3CH4 + CO2 + 2H20 (10) thought to be the result of a complete oxidation of the alco- hol coupled with the reduction of three-fourths of the resulting carbon dioxide. However, this possibility was excluded by tracer experiments which showed that less than 1% of the methane was derived from carbon dioxide.25 Therefore the methane must be formed more or less directly by reduction of methanol. With acetate, which is decomposed by several species of methane bacteria, the situation is similar. The fermenta- tion of acetate is equivalent to a decarboxylation. As we have seen, according to van Niel's theory, acetate should be completely oxidized to carbon dioxide, half of which is simultaneously reduced to methane. This mechanism was first critically tested by Buswell and Sollo26 in 1948 by the use of C14-labeled carbon dioxide. They demonstrated that essentially none of the methane is derived from carbon dioxide. Subsequently Thressa Stadtman and I23-25 showed by the tracer method that the methane is formed entirely from the methyl carbon, and the carbon dioxide exclusively from the carboxyl carbon, of acetate (equation 11) . These G*H3C°OOH — -> C*H4 + C°02 (1 1) results are inconsistent with the carbon dioxide reduction theory since the methyl carbon of acetate is not oxidized to carbon dioxide and carbon dioxide is not a precursor of methane. 22 BACTERIAL FERMENTATIONS After the fate of the carbon atoms of acetate was estab- lished, it became of interest to find out what happens to the hydrogen atoms attached to the methyl group. In par- ticular, it was important to determine whether some or all of these hydrogen atoms are removed by an oxidative reaction during the course of the fermentation of acetate or whether the methyl group is incorporated intact into methane. This question was investigated by Pine.27 In the experiment illustrated by equation 12 acetate labeled CD3COOH -5$. CD3H + co2 (12) in the methyl group with deuterium was fermented and the amount of deuterium per mole in the evolved methane was compared with that in the substrate. Mass analysis showed that the deuterium contents of the acetate and methane were essentially equal. In another experiment illustrated by equation 13 unlabeled acetate was fermented in the CH3COOH -2£ CH3D + co2 ( 1 3) presence of D20. The data indicate that approximately one atom of deuterium per molecule of methane was de- rived from the solvent. These results demonstrate that the methyl group is transferred from acetate to methane as a unit, without the loss of attached hydrogen or deu- terium. In other words, one or more transmethylation reactions occur during the fermentation. Pine28 later investigated the same problem with methanol by allowing unlabeled methanol to ferment in the presence of a mixture of D20 and H20 (equation 14) . The meth- ane formed in this experiment contained a large percentage of CH3D, although somewhat less than the percentage of deuterium in the water. This was interpreted to mean BIOLOGICAL FORMATION OF METHANE 23 that the methyl group of methanol, like that of acetate, is transferred intact into methane. The lower percentage of deuterium in the methane could be attributed to a prefer- ential utilization of the mass 1 hydrogen from the water. 4CH3OH -2$. 3CH3D + C02 (14) The specific reactions involved in methane formation from carbon dioxide, acetate, or methanol are not known. The simplest mechanism proposed for the conversion of carbon dioxide to methane is a stepwise reduction involving formate or carbon monoxide, formaldehyde, and methanol as successive intermediates. A faint suggestion of such a sequence was provided by Stephenson and Stickland,29 who claim to have isolated a bacterium capable of reducing carbon dioxide, carbon monoxide, formate, formaldehyde, and methyl alcohol to methane using hydrogen as a reduc- tant. However, their experiments do not provide any substantial evidence for a sequential reduction of these compounds. Carbon monoxide, formaldehyde, and meth- anol were reduced far too slowly to be intermediates. Formate was shown to be converted to hydrogen and carbon dioxide more rapidly than to methane; conse- quently carbon dioxide actually appeared to be an inter- mediate in formate reduction rather than the reverse. Kluyver and Schnellen14 later reinvestigated the fermen- tation and reduction of one-carbon compounds in the presence of hydrogen by pure cultures of three species of methane bacteria. Cell suspensions of Mbac. omelianskii were unable to reduce any one-carbon compound at a sig- nificant rate, except carbon dioxide, which reacted rapidly. Methanosarcina barkerii, which ferments methanol, ace- tate, and carbon monoxide, was unable to reduce formate or formaldehyde. A mixture of carbon monoxide and 24 BACTERIAL FERMENTATIONS hydrogen was converted to methane according to equa- tion 15. CO + 3H2 —> CH4 + H20 (15) However, convincing evidence was obtained that this re- action involves two steps with carbon dioxide being an intermediate (equations 16 and 17) . Methanobacterium CO + H20 — *- C02 + H2 (16) C02 + 4H2 — > CH4 + 2H20 (17) formicicum ferments both carbon monoxide and formate, and reduces carbon dioxide with hydrogen, but is unable to reduce formaldehyde or methanol. These observations of Kluyver and Schnellen clearly exclude formate, carbon monoxide, formaldehyde, and methanol as intermediates in carbon dioxide reduction despite the fact that all of these compounds, except possibly formaldehyde, can be metab- olized by some species. Therefore it is virtually certain that the intermediates in methane formation are not one- carbon compounds. The only available information which seems to provide evidence concerning the path of carbon dioxide reduc- tion to methane is derived from the previously mentioned studies on the acetate and methanol fermentations, which curiously enough do not involve carbon dioxide. The argu- ment might be made that these fermentations are entirely unrelated processes and have no bearing on the path of carbon dioxide reduction. However, there is a good reason for rejecting this view, namely, the fact that the Methanosarcina species which ferments acetate and meth- anol can also reduce carbon dioxide to methane with hydrogen;9 it seems highly improbable that one organism BIOLOGICAL FORMATION OF METHANE 25 would possess several types of chemical machinery for making methane from different substrates. A schematic representation of the possible pathways of carbon in methane formation, which combines the infor- mation obtained from studies of the acetate and methanol fermentations with the carbon dioxide reduction theory of van Niel, is presented in Fig. 1. Carbon dioxide is postu- C02 + XH — *• XCOOH CH3OH + XH CH3COOH +XH4- XCH3 XH + CH4 Fig. I. Possible Pathways of Methane Formation. lated to combine with an unidentified organic compound XH to form a carboxylated derivative of X which is then reduced by three successive steps to a methyl derivative of X which on further reduction yields methane and regener- ates the carbon dioxide acceptor XH. Methanol and acetate are postulated to react also with XH to give the intermediate XCH3 of the carbon dioxide reduction path- way. The direct formation of this intermediate from the organic methyl donors might well inhibit carbon dioxide utilization by lowering the concentration of XH or in other ways and thus account for the absence of carbon dioxide reduction with these substrates. 26 BACTERIAL FERMENTATIONS This scheme is to be regarded only as a rough working hypothesis. Undoubtedly the actual reaction mechanism is considerably more complicated. For example, a source of energy, such as adenosine triphosphate, would be required for the reduction of XCOOH to XCHO, and for the meth- ylation of XH by methanol. But these are only matters of biochemical detail that could be worked out once the main pathway is established. At present the most important point to determine is the nature of the hypothetical com- pound XH, which could be a metabolite, as in photo- synthesis, or a coenzyme as in fatty acid oxidation. The possibility must not be overlooked that the one-carbon unit is passed from one carrier to another as its oxidation level changes, so that XH may be not one but several compounds. The material presented in this chapter shows that knowl- edge of the biology, physiology, and biochemistry of meth- ane bacteria has developed slowly over a long period of time. Unfortunately this knowledge is still incomplete and superficial in many respects. Much time and effort will be required to bring our understanding of this group to a level comparable with that of many other groups of micro- organisms. REFERENCES 1. Sohngen, N. L., Het ontstaan en verdwijnen van waterstof en methaan onder den invloed van het organische leven, disserta- tion, Technical University, Delft, 1906 (J. Vis, Jr., Delft, pub- lisher) . 2. Bcchamp, E., Ann. chim. phys., 13, 103 (1868) . 3. Tappeiner, W., Ber. deut. chem. Ges., 15, 999 (1882) . 4. Popoff, Leo., Arch. ges. Physiol., 10, 142 (1875) . 5. Hoppe-Seyler, F., Z. physiol. Chem., 10, 201, 401 (1886) . 6. Omelianski, W., Centr. Bakteriol. Parasitenk., II, 15, 673 (1906) . 7. Barker, H. A., Arch. Mikrobiol., 7, 420 (1936) . BIOLOGICAL FORMATION OF METHANE 27 8. Barker, H. A., Leeuwenhoek, 6, 201 (1940) . 9. Schnellen, C. G. T. P., Onderzoekingen over de methaangisting, dissertation, Technical University, Delft, 1947 (De Maasstad, Rotterdam, publisher) . 10. Stadlman, T. C, and H. A. Barker, ]. BacterioL, 62, 269 (1951) . 11. Stadtman, T. C, and H. A. Barker, /. BacterioL, 61, 67 (1951) . 12. Mylroie, R. L., and R. E. Hungate, Can. J. Microbiol., 1, 55 (1954). 13. Laigret, J., Compt. rend. acad. sci., Paris, 221, 359 (1945). 14. Kluyver, A. J., and C. G. T. P. Schnellen, Arch. Biochem., 14, 57 (1947) . 15. Buswell, A. M., and W. D. Hatfield, Illinois State Water Survey Bull. 32, 1936. 16. Buswell, A. M., and H. F. Mueller, Ind. Eng. Chem., 44, 550 (1952) . 17. Barker. H. A., /. Biol. Chem., 137, 153 (1941) . 18. Barker, H. A., Proc. Nat. Acad. Sci. U.S., 29, 184 (1943) . 19. Pine, M. J., and H. A. Barker, /. BacterioL, 68, 589 (1954) . 20. Barker, H. A., Arch. MikrobioL, 7, 404 (1936) . 21. Breed, R. S., E. G. D. Murray, and A. P. Hitchens, Bergey's Manual of Determinative Bacteriology, sixth edition, Williams & Wilkins Co., Baltimore, 1948. 22. Sohngen, N. L., Rec. trav. chim., 29, 238 (1910) . 23. Stadtman, T. C., and H. A. Barker, Arch. Biochem., 21, 256 (1949) . 24. Buswell, A. M., L. Fina, H. F. Mueller, and A. Yahiro, /. Am. Chem. Soc, 73, 1809 (1951). 25. Stadtman, T. C., and H. A. Barker, /. BacterioL, 61, 81 (1951) . 26. Buswell, A. M., and F. W. Sollo, /. Am. Chem. Soc, 70, 1778 (1948) . 27. Pine, M. J., and H. A. Barker, /. BacterioL, 71, 644 (1956) . 28. Pine, M. J., and H. A. Barker, BacterioL Proc, 1954, 98. 29. Stephenson, M., and L. H. Stickland, Biochem. J., 27, 1517 (1933) . CHAPTER THE CHEMISTRY OF BUTYRIC ACID-BUTANOL FERMENTATIONS The butyric acid fermentation of sugar and lactate was discovered by Pasteur in 1861. During the next half century Fritz, Beijerinck, Duclaux, Winogradsky, Weizmann, and others studied various butyric acid fermentations and showed that they are caused mainly by several species of Clostridia that are widely distributed in nature and as a group are able to decompose a great variety of substrates, including sugars, polysaccharides, and hydroxy acids. The main products of these fermentations were identified as acetic and butyric acids, rz-butanol, ethanol, acetone, some- times isopropanol, carbon dioxide, and hydrogen. Also techniques were devised for the commercial exploitation of the fermentation of starch by Clostridium acetobutylicum for the production of acetone and butanol. Early studies in the chemistry of the butyric acid-butanol fermentations were hampered by the variety of the products and the difficulties associated with their quantitative estima- tion. Precise and extensive quantitative information con- 28 BUTYRIC ACID-BUTANOL FERMENTATIONS 29 cerning the fermentation products did not become available until the period between 1925 and 1932, when several fun- damental studies such as those of Donker,1 van der Lek,2 and Peterson and Fred3 were published. This information provided a basis for reasonable deductions and specula- tions concerning the chemical pathways involved in the conversion of glucose and other substrates to fermentation products. The state of knowledge of the chemistry of butyric acid- butanol fermentations that existed in 19314 is summarized schematically in Fig. 1. It was generally accepted without 1. C6H1206 —> 2C3 (?) — > 2C2 (?) + 2Ci (?) 2. 2CH3CHO (?) — > CH3CHOHCH2CHO (?) — > CH3CH2CH2COOH 3. 2CH3COOH (?) — > CH3COCH2COOH — ^ CH3COCH3 + C02 4. CH3CH2CH2COOH -^> CH3CH2CH2CH2OH 5. HCOOH (?) —> H2 + C02 Fig. I. Chemistry of Butyric Acid-Butanol Fermentation, 1931. proof that the early stages of hexose decomposition fol- lowed the general pattern of alcoholic fermentation. This involved the fission of the hexose into two C3 compounds each of which was converted to a C2 and a C± compound. Butyric acid and acetone were believed to be formed by a condensation of C2 units, followed in the case of acetone by the removal of one carbon as carbon dioxide. Reilly et al.5 and Speakman6 had provided evidence that acetate and butyrate are intermediates in the formation of acetone and butanol, respectively. The acids were shown to accu- 30 BACTERIAL FERMENTATIONS mulate in the early stages of the fermentation and then decrease simultaneously with the rapid formation of the solvents. Addition of acetate to a starch fermentation resulted in an almost theoretical increase in the yield of acetone, whereas addition of butyrate increased the yield of butanol. The suggested occurrence of acetoacetate as an intermediate between acetate and acetone was supported by the observation that added acetoacetate was converted to acetone and carbon dioxide. Several parts of the 1931 picture were largely speculative. For example, the identity of the postulated C3 compound was not known. Both methyl glyoxal and pyruvate were sug- gested as possibilities, but definitive evidence foi either was lacking. Also the C2 precursor of butyrate was not known. It was commonly assumed to be acetaldehyde because the aldol condensation in non-biological systems gave a product which might be converted to butyric acid. However, evi- dence for the occurrence of acetaldehyde as an intermediate in the fermentation was slight; neither added acetaldehyde nor acetaldol increased the yield of butyric acid. The identity of the primary Cx compound was also uncertain. It was thought to be either formate or carbon dioxide. Kluyver4 favored the view that formate was formed first and then decomposed to hydrogen and carbon dioxide, but again the evidence was not conclusive. This brief summary indicates some of the uncertainties concerning the chemistry of these fermentations that existed twenty-five years ago. Since that time considerable advances have been made in this field. In the following sections I shall try to review some of these developments and pre- sent the current picture of the chemistry of the butyric acid-butanol fermentations. The Role of Acetate. One of the most significant devel- opments in this field was the recognition of the key role of BUTYRIC ACID-BUTANOL FERMENTATIONS 31^ acetate or acetate derivatives in the formation of butyric acid. The first demonstration that acetate can be a pre- cursor of butyrate and butanol was made by Wood et al.7 They used CI. acetobutylicum to ferment starch in the presence of acetate -1-C13 and determined the distribution of the C13 in the products. /^ TT ^ n CH3C*OOH (C6Hn05)n > c*o2 CH3C*OCH3 CH3G*H2CH2C*OOH CH3C*H2CH2C*H2OH They found, as was to be expected on the basis of earlier balance studies, that the isotope was present in carbon dioxide and the carbonyl group of acetone. This labeling is consistent with a pathway involving acetoacetate, shown in equation 3 of Fig. 1. The additional finding of C13 equally distributed between the 1 and 3 positions of butyr- ate and butanol was quite unexpected. It necessitated the conclusion that the C4 compounds also are formed by a con- densation of two molecules of acetate or some compound that is easily formed from acetate. At about the same time a role of acetate in butyrate synthesis was established independently by nutritional and fermentation balance experiments carried out with two other butyric acid bacteria, namely CI. kluyveri and CI. tyrobutyricum. Clostridium kluyveri is a somewhat atypical butyric acid bacterium, which was isolated from enrichment cultures containing ethanol as the only organic compound.8 This organism was shown to have unusual substrate require- ments. It could not decompose carbohydrates, amino acids, or even pyruvate, but it grew well with ethanol and a com- 32 BACTERIAL FERMENTATIONS pound present in yeast autolyzate prepared in a special way.9 The concentration of yeast autolyzate necessary to support moderate growth was so high that it soon became evident that the essential compound probably was not an amino acid or a growth factor. Eventually, by fractionation of the yeast autolyzate the major active component was shown to be acetate. Fermentation balance experiments then showed that the energy-yielding process used by CI. kluyveri is a conversion of ethanol and acetate to butyrate and caproate.10 When acetate is present in excess, butyrate is the main product (equation 1) , whereas when ethanol is in excess the main product is caproate (equation 2) . It may be noted that CH3CH2OH + CH3COOH — > CH3CH2CH2COOH + H20 (1) 2CH3CH2OH + CH3COOH — > CH3CH2CH2CH2CH2COOH + 2H20 (2) no ethanol is utilized under the anaerobic conditions suit- able for growth of this organism unless acetate is also sup- plied. These balance experiments demonstrated by direct chemical analysis the conversion of acetate to the C4 and C6 fatty acids. CI. tyrobutyricum is a species originally isolated by van Beynum and Pette11 from low-grade silage. It was believed to be responsible for the conversion of lactate to the butyrate which is commonly present in silage that has not developed a sufficiently high acidity to inhibit spoilage organisms. When mixed cultures of CI. tyrobutyricum and other silage bacteria were allowed to grow in a complex medium containing lactate and yeast extract, the organism appeared to be responsible for a vigorous fermentation of lactate. However, pure cultures fermented lactate very BUTYRIC ACID-BUTANOL FERMENTATIONS 33 feebly and then only when the medium was supplied with a very high concentration of yeast extract. The high yeast extract requirement for the growth of CI. tyrobutyricum on lactate was reminiscent of the similar requirement found earlier with CI. kluyveri. Consequently Bhat and I12 reinvestigated the substrate requirements of CI. tyrobutyricum (CI. lactoacetophilum) particularly in relation to a possible role of acetate. We found that glucose or pyruvate is readily fermented when supplied alone in a medium containing only the usual nutrients required by butyric acid bacteria. In contrast lactate or glycerol cannot be fermented in such a medium unless a roughly stoichio- metric amount of acetate is also provided. When this is done, lactate or glycerol is fermented almost as readily as glucose. Balance experiments showed that acetate is actu- ally consumed during the fermentation of lactate and is converted to butyrate just as in the CI. kluyveri fermenta- tion of ethanol and acetate. A schematic explanation for the necessity of additional acetate in the lactate fermentation is given by the following equations. CH3CHOHCOOH + H20 -^£ CH3COOH + CO2 (3) 4-4H 2CH2COOH — -> CH3CH2CH2COOH + 2H20 (4) CH3CHOHCOOH + CH3COOH — > CH3CH2CH2COOH + H20 (5) The oxidation of lactate to acetate (equation 3) makes available four equivalents of hydrogen for the reduction of acetate or its derivations to butyrate. Two moles of acetate are required for reaction 4, whereas only 1 mole of acetate is formed. Therefore, a second mole of acetate must be supplied in the medium as a source of an oxidant so that 34 BACTERIAL FERMENTATIONS the fermentation can proceed according to reaction 5. Actu- ally the stoichiometric relations are not quite so simple because some hydrogen gas is evolved; consequently some- what less than 1 mole of acetate is consumed per mole of lactate. Acetate is usually required in butyric acid fermentations of lactate, glycerol, and other compounds more reduced than carbohydrate. Whenever the ability of an organism to ferment such compounds is to be tested acetate should be supplied in the medium. In the past, failure to do this has resulted in erroneous conclusions concerning substrate utili- zation and the taxonomic relations of certain butyric acid bacteria. Although acetate is usually required for the butyric acid fermentation of lactate, at least one exception to this gen- eralization is known. Butyribacterium rettgeri readily fer- ments lactate in the absence of added acetate; in fact both acetate and butyrate are produced.13 This is possible be- cause B. rettgeri possesses an additional oxidative mech- anism, by which part of the carbon dioxide it forms is reduced to acetate.14-15 This type of oxidative mechanism occurs in several anaerobic bacteria.16'17'18 Use of Enzyme Preparations. Most of the early work on the chemistry of butyric acid fermentations was done with growing cultures or washed cell suspensions. Experi- ments with such material provided information concerning the over-all chemical changes in these fermentations and permitted some rough conclusions about the course of the metabolic pathways. However, it eventually became appar- ent that a complete elucidation of the fermentation reac- tions could only be achieved by the use of cell-free extracts and purified enzyme systems. The first extensive use of cell-free extracts in the study of the butyric acid fermentation was made by Koepsell and BUTYRIC ACID-BUTANOL FERMENTATIONS 35 Johnson.19 They found that when a cell paste of CI. butylicum was kept frozen for a week or more and then allowed to thaw, a soluble enzyme preparation was obtained that catalyzed the conversion of pyruvate to acetate, carbon dioxide, and hydrogen. Phosphate was required for pyru- vate decomposition, and later20 evidence was obtained for the formation of acetyl phosphate (equation 6) . This pro- CH3COCOOH + HOPO3H2 — -> CH3COOPO3H2 + C02 + H2 (6) vided an indication that an acetyl compound might be a precursor of butyrate. Actually the extracts used by Koepsell and Johnson were unable to form butyrate from pyruvate under the conditions used. However, they did catalyze a reaction by which acetyl phosphate and butyrate were apparently converted to acetate and butyryl phosphate (reaction 7), thus opening the possibility that the latter Acetyl phosphate + Butyrate — > Acetate + Butyryl phosphate (7) compound was involved in butyrate or butanol synthesis. The origin of the hydrogen, which is a conspicuous product of most butyric fermentations, was somewhat clarified by this work. Extracts that produced hydrogen readily from pyruvate (equation 6) were found to be unable to decom- pose formate. This eliminated the theory that formate is an obligatory intermediate in the formation of hydrogen from pyruvate. The success of Koepsell and Johnson in obtaining enzy- matically active extracts from CI. butylicum suggested that similar methods might be used with other species. Conse- quently Stadtman and I21 tried preparing cell-free extracts of CI. kluyveri by drying the cells and then extracting with 36 BACTERIAL FERMENTATIONS dilute buffer. This simple procedure gave a soluble prepa- ration that contained an amazingly complete fermentation system. Clostridium kluyveri extracts were found to cata- lyze the oxidation of alcohol via acetaldehyde to acetyl phosphate (equation 8) , the oxidation of butyrate to acetyl phosphate and acetate (equation 9) , and the synthesis of butyrate from acetyl phosphate and acetate using hydrogen as the reducing agent (equation 10) . CH3CH2OH + H3PO4 + 02 — > CH3COOPO3H2 + 2H20 (8) CH3CH2CH2COOH + H3PO4 + 02 — > CH3COOPO3H2 + CH3COOH + H20 (9) CH3COOPO3H2 + CH3COOH + 2H2 — > CH3CH2CH2COOH + H3PO4 + H20 (10) The formation of acetyl phosphate by the oxidation of acetaldehyde, which was comparable to its formation from pyruvate by CI. butylicum extracts, indicated that acetyl phosphate might play a key role in butyric acid fermenta- tions. This idea was supported by the observation that under the conditions of these experiments acetyl phosphate was required for butyrate synthesis; it could not be replaced by acetate. A comparison of reactions 9 and 10 also indi- cated that the oxidation and synthesis of butyrate probably involve the same sequence of reversible reactions. Early Studies on the Path of Butyrate Synthesis in Clostridium kluyveri. After the discovery of the enzy- matic synthesis of butyrate from acetyl phosphate and ace- tate, the mechanism of this synthesis became the central problem of the butyric acid fermentation and indeed of fatty acid metabolism in general. The formation of butyr- ate from C2 precursors was obviously a complex reaction BUTYRIC ACID-BUTANOL FERMENTATIONS 37 since it must involve a minimum of three reactions: a con- densation and two reduction steps. The simplest possible mechanism of butyrate synthesis and oxidation, used as a working hypothesis in subsequent experimentation, is shown in the following reaction sequence: CH3COOPO3H2+CH3COOH ^=± CH3COCH2COOH 1 ±2H ±2H \Y CH3CH2CH2COOH ^=± CH3CHOHCH2COOH ±H20 In this sequence acetyl phosphate was postulated to con- dense with acetate to form acetoacetate which was reduced first to /?-hydroxybutyrate and then to butyrate. Both of the postulated intermediates were known to accumulate during the oxidation of fatty acids in animals under certain circumstances. The above sequence could readily be tested with CI. Kluyveri extracts by finding out whether acetoacetate and /?-hydroxybutyrate react as the scheme predicts. Stadtman22 found that they do not so react. /3-Hydroxybutyrate could not be reduced to butyrate nor oxidized to acetyl phosphate. Acetoacetate could undergo a phosphoroclastic conversion to acetyl phosphate and acetate, but this reaction was too slow and occurred under the wrong conditions to partici- pate in butyrate oxidation. Furthermore, although aceto- acetate could be reduced enzymatically with molecular hydrogen as the scheme predicts, the product was not butyr- ate but /3-hydroxybutyrate, which, could not be further reduced. These observations eliminated acetoacetate and /J-hydroxybutyrate as intermediates in butyrate synthesis. A variety of other four-carbon acids at the oxidation levels of acetoacetate and /J-hydroxybutyrate were then tested.22-23 The only one of these acids that proved to be 38 BACTERIAL FERMENTATIONS reactive in the CI. kluyveri system was vinylacetic acid (3-butenoic acid) . This compound at first showed the reactions expected of an intermediate. Like butyrate, it was oxidized to acetyl phosphate and acetate (reaction 11) or, in the absence of orthophosphate, to acetoacetate (reac- tion 12). CH2==CHCH2COOH + H3P04 + \02 — > CH3COOPO3H2 + CH3COOH (11) CH2=CHCH2COOH + J02 -^ CH3COCH2COOH (12) CH2=CHCH2COOH + H2 — > CH3CH2CH2COOH (13) 2CH2=CHCH2COOH + H3PO4 + H20 — > CH3CH2CH2COOH + CH3COOH + CH3COOP03H2 (14) In the presence of hydrogen, vinylacetate was reduced to butyrate (reaction 13) and in a nitrogen atmosphere it underwent a typical dismutation (reaction 14) . Further- more, all of these reactions occurred at rates equal to or greater than the over-all rates of butyrate synthesis and oxidation. However, all attempts to demonstrate the actual formation of vinylacetate during the synthesis or oxidation of butyrate were decisively negative. For example, when C14-butyrate was oxidized in the presence of a pool of unlabeled vinylacetate, essentially no C14 was incorporated into the vinylacetate. This and other experiments excluded vinylacetate as a normal intermediate in butyrate metab- olism. The negative results obtained with vinylacetate, aceto- acetate, and other acids at the same oxidation levels, ap- peared to exclude all C4 acids as intermediates in butyrate metabolism. This led inevitably to the hypothesis that the actual intermediates are compounds in which the BUTYRIC AC1D-BUTANOL FERMENTATIONS 39 C4 acids are combined with another substance, possibly coenzyme A.23-24 Role of Coenzyme A in Fatty Acid Activation, The suggestion that the intermediates in butyrate metabolism are coenzyme A derivatives was based on the work of Lipmann, Lynen, and others on the role of this cofactor in enzymatic acetylation reactions in animal systems, and on the work of Stadtman and his associates on several reactions catalyzed by CI. kluyveri extracts. Coenzyme A had been shown to be required for the acety- lation of sulfanilamide by pigeon liver extracts in a system containing acetate and ATP.25 Methods for the purifica- tion of coenzyme A had been developed, and the structure of the compound had been studied. Not all the structural details had been worked out at this time, but coenzyme A was known to contain adenosine-5'-phosphate, two addi- tional phosphate groups, pantothenic acid, and thioetha- nolamine in a terminal position.26 The mode of action of coenzyme A in acetate activation was discovered by Lynen et al.,27 who isolated acetyl coenzyme A from yeast, demon- strated that the acetyl group is attached to the sulfur atom, and showed that acetyl coenzyme A acts as an enzymatic acetylating agent. This new information concerning coenzyme A was soon applied to the CI. kluyveri system. Stadtman et al.28-29 showed that extracts of CI. kluyveri contain an enzyme called phosphotransacetylase that catalyzes a reversible transfer of the acetyl group from acetyl phosphate to coenzyme A. This enzyme was first detected as an activity that caused a rapid decomposition of acetyl phosphate in the presence of inorganic arsenate ("arsenolysis") and also catalyzed a rapid exchange between the phosphoryl group of acetyl phosphate and orthophosphate. When these reac- tions were found to be completely dependent upon the 40 BACTERIAL FERMENTATIONS presence of coenzyme A, the phosphotransacetylase reaction was formulated as follows: Acetyl phosphate + HSCoA ^=±: Acetyl-SCoA + Phosphate (15) This reaction accounts for the observed phosphate exchange. Since arsenate and phosphate have similar properties the reverse reaction in the presence of arsenate would be expected to give acetyl arsenate. This hypothetical com- pound, which has never been prepared or isolated, pre- sumably hydrolyzes rapidly to acetate and arsenate and thus accounts for the arsenolysis reaction. Further evidence for the correctness of reaction 15 was provided by the use of substrate amounts of coenzyme A and the identification of the product with Lynen's acetyl coenzyme A. This was done in part by showing that the product of reaction 15 could serve as an acetyl donor in the enzymatic acetylation of sulfanilamide and in the condensation with oxaloacetate to give citrate. The easy interconversion of acetyl phosphate and acetyl- SCoA by phosphotransacetylase opened the question as to which of these compounds is the immediate product of ethanol and acetaldehyde oxidations. Burton and Stadt- man30 investigated these reactions and found that ethanol oxidation involves two enzymes. Reaction 16 is catalyzed CH3CH2OH + DPN+ ^=± CH3CHO + DPNH + H+ (16) CH3CHO + HSCoA + DPN+ ^=±: CH3COSC0A + DPNH + H+ (17) by a DPN-specific alcohol dehydrogenase. The further oxidation of the resulting acetaldehyde by acetaldehyde dehydrogenase (reaction 17) is dependent upon DPN and BUTYRIC ACID-BUTANOL FERMENTATIONS 41^ coenzyme A but not upon orthophosphate. The product of this reaction is acetyl-SCoA which accumulates when substrate amounts of coenzyme A are used. With only catalytic amounts of coenzyme A acetaldehyde oxidation ceases as soon as the coenzyme is mostly converted to the acetylated form. The further addition of orthophosphate in the presence of phosphotransacetylase permits the trans- fer of the acetyl group from SCoA to phosphate, giving acetyl phosphate and regenerating free coenzyme A so that the oxidation of acetaldehyde can continue. Of course, any other reaction that will consume acetyl-SCoA and regenerate coenzyme A will serve the same function. As we shall see, the normal acetyl-SCoA-consuming reaction in the CI. kluyveri fermentation is that leading to the formation of acetoacetyl-SCoA. The formation of acetyl phosphate appears to be a side reaction which may be useful in the storage of acetyl groups and in the generation of ATP by the action of the enzyme acetokinase31 (reaction 18) . CH3COOPO3H2 + ADP =j=± CH3COOH + ATP (18) The formation of acetyl-SCoA by oxidation of acetalde- hyde is a peculiarity of CI. kluyveri. Other sugar-fermenting butyric acid bacteria probably form acetyl-SCoA during the phosphoroclastic decomposition of pyruvate (reaction 6) . This reaction in CI. butyricum is evidently quite complex since it requires at least three cofactors: cocarboxylase, coenzyme A, and ferrous ions.32 The requirement for coen- zyme A, the eventual formation of acetyl phosphate, and the presence of phosphotransacetylase in the enzyme ex- tracts indicate the intermediate formation of acetyl-SCoA. This compound is known to be a product of pyruvate oxida- tion by Escherichia coli.3i The role of coenzyme A in the activation of fatty acids other than acetate has been studied in some detail with 42 BACTERIAL FERMENTATIONS butyric acid bacteria. It has already been mentioned that Koepsell et al.20 obtained evidence that extracts of CI. butyli- cum catalyze the synthesis of butyryl phosphate from acetyl phosphate and butyrate (reaction 7) . Clostridium kluyveri extracts catalyze a similar reaction in which the phosphoryl group is transferred from acetyl phosphate to other fatty acids containing from three to eight carbon atoms.34 In a further study of this reaction using acetyl phosphate and propionate, Stadtman35 established that it is completely dependent upon the presence of coenzyme A and phos- photransacetylase. These requirements suggested that the apparently simple phosphoryl transfer (reaction 22) actu- ally involves three distinct enzymatic steps shown in the following equations: Acetyl phosphate + HSCoA ^^ Acetyl-SCoA + Phosphate (19) Acetyl-SCoA + Propionate ^=^ Acetate + Propionyl-SCoA (20) Propionyl-SCoA + Phosphate ^±= Propionyl phosphate + HSCoA (2 1 ) Acetyl phosphate + Propionate ^=^ Acetate + Propionyl phosphate (22) Reaction 19 is the usual phosphotransacetylase reaction, and reaction 21 is the analogous reaction involving pro- pionyl phosphate, catalyzed by the same enzyme. Reaction 20 represents a reversible transfer of the SCoA group from acetate to propionate. This transfer results in the acti- vation of propionate and the deactivation of acetate. The enzyme catalyzing this reaction has been partially purified and named coenzyme A-transphorase. In the presence of this enzyme the SCoA group may be transferred to several BUTYRIC ACID-BUTANOL FERMENTATIONS 43 carboxylic acids in addition to acetate and propionate, namely formate, n-butyrate, rc-valerate, rc-caproate, lactate, and vinylacetate. These include all the compounds that are formed or readily metabolized by CI. kluyveri extracts. In contrast, crotonate, /3-hydroxybutyrate, and acetoacetate, which are relatively unreactive in the CI. kluyveri system, do not serve as substrates for coenzyme A-transphorase. This indicates that only those acids which can be converted to their coenzyme A derivatives are highly reactive in this system. Further evidence for the participation of the coenzyme A derivatives of the four carbon acids in the metabolism of CI. kluyveri has been obtained by studying the oxidation of butyrate and the reduction of vinylacetate. Crude ex- tracts of CI. kluyveri readily oxidize butyrate. However, Stadtman21 found that after these extracts were dialyzed to remove soluble cofactors, butyrate was not oxidized unless catalytic amounts of acetyl phosphate, a source of coenzyme A, and other cofactors were added. This suggested that butyrate had to be activated by conversion to either butyryl- SCoA or butyryl phosphate before it was susceptible to oxidation. Butyryl phosphate was excluded as an inter- mediate by the observation that it was oxidized more slowly than butyrate. Finally Stadtman36 prepared butyryl-SCoA and showed that it was oxidized rapidly in the absence of free coenzyme A to acetoacetyl-SCoA, and in the presence of coenzyme A to acetyl-SCoA. The evidence regarding the role of coenzyme A in the reduction of vinylacetate to butyrate is similar though less extensive. Peel37 has found that a catalytic amount of either acetyl phosphate or acetyl-SCoA is essential for vinyl- acetate reduction. This points to the participation in this reaction of vinylacetyl-SCoA which is known to be formed from acetyl-SCoA and vinylacetate in the SCoA-transphorase 44 BACTERIAL FERMENTATIONS reaction. Later experiments by Bartsch38 have shown that CI. kluyveri extracts contain an enzyme called vinylacetyl isomerase which converts vinylacetyl-SCoA to crotonyl-SCoA. CH2=CHCH2COSCoA — > CH3CH=CHCOSCoA (23) Vinylacetyl-SCoA Crotonyl-SCoA The latter compound may be the substance actually re- duced, as it appears to be in animal systems.39-40 A key reaction in the synthesis of butyrate is the reversible formation of the C4 compound at the oxidation level of acetoacetate from its C2 precursor. The elucidation of this reaction required a series of investigations most of which were done with animal enzyme preparations. Studies carried out in Lipmann's laboratory with a soluble multi- enzyme preparation from pigeon liver showed that ATP, coenzyme A, and acetate are required for acetoacetate syn- thesis.41 This suggested that a coenzyme A derivative of acetate was a precursor of acetoacetate, but it was not known whether one molecule of "active" acetate condensed with a molecule of acetate to give acetoacetate directly or whether more complicated reactions were involved. This situation was clarified by Stadtman et al.,42 who studied acetoacetate synthesis in a system containing acetyl phos- phate, acetate, coenzyme A, phosphotransacetylase from CI. kluyveri, and a pigeon liver extract. In separate experi- ments, either the acetyl phosphate or the acetate was labeled with C14 and the incorporation of the isotope into aceto- acetate was studied. The data showed that free acetate was not used, but on the contrary both halves of the acetoacetate molecule were derived from acetyl phosphate presumably via acetyl-SCoA (reaction 24) . 2CH3COSCoA + H20 — > CH3COCH2COOH + 2HSCoA (24) BUTYRIC AC1D-BUTANOL FERMENTATIONS 45 Now the above reaction was catalyzed by a relatively crude pigeon liver extract, and there was no certainty that it rep- resented a single enzymatic step. Another possibility24,43 was a two-step reaction sequence in which one enzyme catalyzed the reversible transfer of an acetyl group from one molecule of acetyl-SCoA to another to give acetoacetyl- SCoA plus free coenzyme A (reaction 25) , and another enzyme hydrolyzed acetoacetyl-SCoA to acetoacetate (reac- tion 26) . 2CH3COSCoA ^^= CH3COCH2COSC0A + HSCoA (25) CH3COCH2COSC0A + H20 — > CH3COCH2COOH + HSCoA (26) The postulated formation of acetoacetyl-SCoA was consist- ent not only with all the available information concerning the pigeon liver acetoacetate-forming system but also with certain observations made on the CI. kluyveri system. For example, at that time the necessity of the activation of butyrate prior to its oxidation was known and the activated compound was suspected to be butyryl-SCoA. The oxida- tion of butyryl-SCoA would be expected to give acetoacetyl- SCoA, and the reverse reaction would also be anticipated. More direct evidence for the reversible formation of acetoacetyl-SCoA from acetyl-SCoA was first obtained by Lynen et al.,44-45 with soluble enzymes from sheep liver. This was made possible by the discovery that thioesters of acetoacetate such as S-acetoacetyl-N-acetyl-thioethanolamine and acetoacetyl-SCoA show a strong ultraviolet absorption peak at 303 m^ in alkaline solution. By means of this opti- cal property the formation and decomposition of acetoacetyl- SCoA according to reaction 25 were demonstrated. The enzyme catalyzing this reaction, called acetoacetyl thiolase 46 BACTERIAL FERMENTATIONS or just thiolase for short, was later purified from animal sources and studied extensively.46-47,48 Several other enzymes catalyzing specific reactions in the reversible conversion of acetoacetyl-SCoA to butyryl-SCoA have also been purified from animal tissues during the last few years.46'49-50,51,52 Current Knowledge of the Path of Butyrate Synthesis in Bacteria. The accumulated information obtained with animal and bacterial enzyme systems indicates that butyr- ate formation in CI. kluyveri involves the reactions shown in Fig. 2. Reactions I, II, and III have already been dis- ±HSCoA r>2CH3COSCoA ^ CH3COCH2COSC0A 2CH3CHO CH3CHOHCH2COSCOA il T2H Vl ±H20 2CH3CH2OH CH3CH=CHCOSCoA V1 1 ±2H vii Ik LM2H3COSC0A+CH3CH2CH2COOH z^±r CH3CH2CH2COSC0A +CH3COOH Fig. 2. Butyric Acid Synthesis in Clostridium kluyveri. cussed; reaction IV is the reduction of acetoacetyl-SCoA to /?-hydroxybutyryl-SCoA catalyzed by the enzyme /?-hydroxy- butyryl-SCoA dehydrogenase; reaction V is the dehydration of yg-hydroxybutyryl-SCoA to crotonyl-SCoA catalyzed by the enzyme crotonase; reaction VI is the reduction of cro- tonyl-SCoA to butyryl-SCoA by the enzyme butyryl-SCoA dehydrogenase; and reaction VII is the transfer of the SCoA group from butyrate to acetate catalyzed by the fatty acid SCoA-transphorase. Some evidence for the occurrence of these reactions in CI. kluyveri, in addition to that already discussed, has accumulated. The oxidation of butyryl-SCoA to aceto- BUTYRIC ACID-BUTANOL FERMENTATIONS 47 acetyl-SCoA in the absence of free coenzyme A (reactions VI, V, and IV) has been demonstrated by spectrophoto- metric methods.36-53 When free coenzyme A was added to such a reaction mixture, the accumulated acetoacetyl-SCoA was rapidly converted to acetyl-SCoA by reaction III. Unfractionated CI. kluyveri extracts were found by Bartsch38 to possess acetoacetyl thiolase activity corresponding to the decomposition of approximately 2 micromoles of acetoacetyl- SCoA per minute per milligram of protein. /?-hydroxy- butyryl-SCoA dehydrogenase activity (reaction IV) has also been demonstrated qualitatively by Bartsch by using a DPNH-generating system to reduce the N-acetyl thioeth- anolamine analogue of acetoacetyl-SCoA. The reaction was followed by the decrease in ultraviolet absorption of the analogue at 303 m^. The reduction product has not been characterized in this system. The corresponding enzyme from rat liver mitochondria has been shown to produce l ( + ) -/?-hydroxybutyryl-SCoA.54'55 Crotonase (reaction V) activity is very high in CI. kluy- veri extracts.38 Using crotonyl-SCoA as a substrate and following the hydration of the double bond by means of its ultraviolet absorption at 263 m^,, the crotonase activity was shown to correspond to the utilization of 570 micromoles of crotonyl-SCoA per minute per milligram of crude protein. This is roughly 20 times the specific activity of ox liver extracts, a rich animal source of the enzyme.40 Butyryl-SCoA dehydrogenase activity (reaction VI) has been observed by coupling the oxidation of the pantotheine analogue of butyryl-SCoA with the reduction of indo- phenol,38 and by coupling the reduction of crotonyl cysta- mine with the oxidation of leucosafranine.56 By means of the latter technique it was found that crude extracts of CI. kluyveri were twice as active with respect to butyryl- SCoA dehydrogenase as the corresponding animal enzyme 48 BACTERIAL FERMENTATIONS which had been purified 350-fold from pig liver extracts. The available evidence indicates that butyrate synthesis in CI. kluyveri involves reactions similar to or identical with those shown in Fig. 2. However, the precise identities of the postulated unsaturated and ^-hydroxy compounds, here referred to as crotonyl-SCoA and /?-hydroxybutyryl- SCoA, have not been established. The unsaturated com- pound might be crotonyl-, isocrotonyl-, or vinylacetyl-SCoA; and the hydroxy compound might be l(+)- or d( — )- /3-hydroxybutyryl-SCoA. Despite the considerable information that is available concerning the chemistry of the CI. kluyveri fermentation, at least one major problem remains unsolved, namely the nature of the mechanism by which the organism obtains useful energy from the conversion of ethanol and acetate to butyrate and caproate. The oxidation of ethanol to acetyl-SCoA provides an energy-rich compound which might be used to generate ATP by means of the phosphotrans- acetylase and acetokinase reactions. However, most of the acetyl-SCoA is converted to butyrate and caproate, and therefore is not directly available as a source of energy for ATP synthesis. Indeed, acetyl-SCoA must be used in this way in order to produce the electron acceptors, acetoacetyl- SCoA, crotonyl-SCoA, and their six carbon analogues, with- out which the oxidation of ethanol and acetaldehyde would be impossible. Since acetyl-SCoA seems to be excluded as a primary energy source for synthetic reactions, some other source of energy must be available to the organism. One possible method of providing ATP is by the coupling of an electron transport reaction with a phosphorylation as occurs in the so-called oxidative phosphorylation in aerobic organisms. In the CI. kluyveri system the most suitable electron trans- port system for this purpose is that between ethanol or BUTYRIC ACID-BUTANOL FERMENTATIONS 49 acetaldehyde and crotonyl-SCoA. The potential span between the ethanol-aldehyde couple [£'0 (pH 7) = — 0.204 volt] and the crotonyl-SCoA-butyryl-SCoA couple [E'0 (pH 7) = 0.19 volt], for example, is 0.39 volt which corresponds to a free-energy change of approximately —18 kcal. under physiological conditions. This is more than sufficient to provide the energy (8 to 10 kcal.) needed for the synthesis of 1 mole of ATP from ADP and orthophosphate. However, so far attempts to demonstrate a coupling of phos- phorylation with crotonyl-SCoA reduction by a DPNH- generating system have not been successful. The enzymatic reactions involved in butyrate formation from acetyl-SCoA in other butyric acid bacteria are prob- ably similar to those demonstrated in CI. kluyveri and animal tissues. Although relatively little specific informa- tion has been obtained with other species, Stern et al.49 have recently reported that extracts of CI. acetobutylicum and CI. sticklandii (strain HF) show enzymatic activities attrib- utable to crotonase, l(+) -/3-hydroxybutyryl-SCoA dehydro- genase, and thiolase. Formation of C5— C7 Fatty Acids. Clostridium kluyveri forms rz-caproate from butyrate and ethanol, and forms n-valerate and n-heptanoate from propionate and ethanol.57 The formation of caproate undoubtedly starts with a con- densation of butyryl-SCoA, derived from butyrate or cro- tonyl-SCoA, with acetyl-SCoA derived from ethanol to give /?-ketocaproyl-SCoA by a reaction analogous to the formation of acetoacetyl-SCoA from acetyl-SCoA. The further enzy- matic steps in the formation of caproate are probably the same as those involved in butyrate synthesis. Experimental evidence for the participation of a C6 /?-keto acid derivative was obtained by showing that the oxidation of caproate, under conditions which prevent the thiolase reaction, results in the accumulation of /?-ketocaproate.58 50 BACTERIAL FERMENTATIONS The formation of the C5 to C7 fatty acids involves a condensation between acetyl-SCoA and the thioester of a longer-chain fatty acid. In these condensations the C2 moiety always appears at the carboxyl end of the resulting fatty acid.57 This result is consistent with a synthetic path involving /?-keto acid derivatives. formation of Acetone and Butanol. Acetone and butanol are formed from carbohydrates by CI. acetobutyli- cum and other butyric acid bacteria mainly toward the end of the fermentation when the pH of the medium drops to about 4.0. As we have seen, the accumulation of solvents coincides with a decrease in the concentration of volatile acids. The conversion of acetic acid to acetone and butanol and the conversion of butyric acid to butanol are clearly established. Acetone is undoubtedly formed by the decarboxylation of acetoacetate (reaction 27) . Johnson et al.59 first demon- CH3COCH2COOH — > CH3COCH3 + C02 (27) strated this reaction with growing cultures and cell suspen- sions; later Davies60 extracted and partially purified the decarboxylase responsible for the reaction and showed that it is most active at pH 5. Acetoacetate is almost certainly formed from acetoacetyl- SCoA, although with bacteria only indirect evidence for this is available. In some CI. kluyveri extracts acetoacetate accumulates during the oxidation of butyrate when the normal conversion of acetoacetyl-SCoA to acetyl phosphate is prevented by a lack of either coenzyme A or orthophos- phate.23-36 This can be interpreted to mean that aceto- acetate is formed from its coenzyme A derivative. The nature of the reaction by which acetoacetyl-SCoA is con- verted to acetoacetate in Cl. kluyveri has not been investi- gated in detail but it appears to be an hydrolysis. This is BUTYRIC ACID-BUTANOL FERMENTATIONS 51^ very wasteful of energy since the hydrolytic cleavage of the thioester bond involves a free-energy change of about —8 kcal. In CI. kluyveri this is unimportant since little or no free acetoacetate is formed during its normal metabolism. A more efficient way of forming acetoacetate would be a transfer of SCoA from acetoacetyl-SCoA to acetate (reaction 28) to give acetoacetate and acetyl-SCoA. The latter CH3COCH2COSC0A + CH3COOH =^± CH3COCH2COOH -f CH3COSC0A (28) compound can be used to make more acetoacetyl-SCoA without a further expenditure of energy. The SCoA trans- ferase of animal tissues catalyzes a reversible reaction of this type between acetoacetyl-SCoA and succinate.61 How- ever, this enzyme probably does not occur in butyric acid bacteria, and the bacterial fatty acid SCoA transphorase, at least that which occurs in CI. kluyveri, does not react with acetoacetate.35 Nevertheless, it seems probable that bacteria like CI. acetobutylicum which form large amounts of acetone via acetoacetate will be found to conserve the thioester bond energy of acetoacetyl-SCoA by making use of reactions similar to reaction 28. The formation of n-butanol from butyrate undoubtedly involves butyryl-SCoA as an intermediate. Butyryl-SCoA can be formed from butyrate and acetyl-SCoA, and probably also by the reaction of butyrate with ATP and HSC0A.34 ATP + Butyrate + HSCoA ^±= AMP + PP 4- Butyryl-SCoA (29) In the first demonstration of the enzymatic formation of butanol with CI. kluyveri, butyryl phosphate was the sub- strate. The path of conversion of butyryl phosphate to butyryl-SCoA has not been determined, but a possible 52 BACTERIAL FERMENTATIONS reaction sequence would be a transfer of the phosphoryl group of butyryl phosphate to ADP (reaction 30) followed Butyryl phosphate + ADP z^±: Butyrate + ATP (30) by a reaction of the resulting ATP with butyrate and coenzyme A to give butyryl-SCoA as shown in reaction 29. It may be noted that phosphotransacetylase cannot catalyze a direct transfer of the butyryl group from butyryl phos- phate to coenzyme A. The reduction of butyryl-SCoA to butyraldehyde by DPNH (reaction 31) is catalyzed by the aldehyde dehydro- genase of CI. kluyveri30 and similar enzymes probably occur CH3CH2CH2COSC0A + DPNH + H+ z^z± CH3CH2CH2CHO + HSCoA + DPN+ (31) in other butyric acid bacteria. The CI. kluyveri enzyme also reacts with acetyl-SCoA and propionyl-SCoA to give acetaldehyde and propionaldehyde respectively. The alde- hydes are finally reduced to the corresponding alcohols by DPNH under the influence of alcohol dehydrogenase (reac- tion 32) . In carbohydrate fermentations the DPNH is probably supplied by the oxidation of triose phosphate. CH3CH2CH2CHO + DPNH + H+ ^=^= CH3CH2CH2CH2OH + DPN+ (32) Certain interrelations between acetone and butanol formation are of some interest. In the fermentation of carbohydrates by CI. acetobutylicum acetone, butanol, and ethanol are formed simultaneously.3 The reason for this is not hard to find. The conversion of acetoacetyl-SCoA to acetone results in the loss of two electron-accepting reactions, BUTYRIC ACID-BUTANOL FERMENTATIONS 53 namely, the reduction of acetoacetyl-SCoA to /?-hydroxy- butyryl-SCoA and the reduction of crotonyl-SCoA to butyryl-SCoA. If the oxidation of triose phosphate is to continue, these reactions must be replaced by other electron- accepting reactions. The only alternative reactions of this type available to these organisms appear to be the reduc- tions of acetyl- and butyryl-SCoA and the corresponding aldehydes to alcohols. Therefore, the formation of acetone is necessarily linked with the formation of alcohols. The trigger mechanism for solvent formation must be a step in the conversion of acetoacetyl-SCoA to acetone. If the conversion of acetoacetyl-SCoA to acetoacetate is in fact a reversible reaction, the irreversible decarboxylation of acetoacetate is probably the key reaction. The observation of Davies60 that solvent formation is associated with a rapid increase in the level of acetoacetate decarboxylase in CI. acetobutylicum supports this conclusion. The material presented in this chapter shows that sub- stantial progress has been made in understanding the complex chemical reactions involved in butyric acid- butanol fermentations. However, despite an early start and rapid initial progress, knowledge of the enzymatic reactions of fatty acid metabolism in bacteria is now less complete than that of similar processes in animal tissues. Only a few bacterial enzymes concerned with fatty acid metabolism have been purified fairly extensively and studied in some detail. More information is needed concerning the enzymes and intermediates functioning between acetoacetyl- SCoA and butyryl-SCoA in butyrate synthesis, and between acetoacetyl-SCoA and acetoacetate in acetone formation. We need to learn something about the mechanisms by which the energy derived from butyrate synthesis and associated reactions is made available to the cell. Also in the carbo- hydrate fermentations, the steps in the conversion of sugars 54 BACTERIAL FERMENTATIONS to pyruvate need to be carefully investigated in the light of the recent discovery of multiple pathways of carbo- hydrate breakdown. These are all problems for the future. REFERENCES 1. Donker, H. J. L., Bijdrage tot de kennis der boterzuur-, butylal- cohol- en acetongistingen, dissertation, Technical University, Delft, 1926 (W. D. Meinema, publisher) . 2. van der Lek, J. B., Onderzoekingen over de butylalkoholgisting, dis- sertation, Technical University, Delft, 1930 (W. D. Meinema, publisher) . 3. Peterson, W. H., and E. B. Fred, Ind. Eng. Chem., 24, 237 (1932) . 4. Kluyver, A. J., The Chemical Activities of Micro-Organisms, Uni- versity of London Press, London, 1931. 5. Reilly, W. J., W. J. Hickinbottom, F. B. Henley, and A. C. Thaysen, Biochem. ]., 14, 229 (1920) . 6. Speakman, H. B., /. Biol. Chem., 43, 401 (1920) . 7. Wood, H. G., R. W. Brown, and C. H. Werkman, Arch. Biochem., 6, 243 (1945) . 8. Barker, H. A., and S. M. Taha, /. Bacteriol, 43, 347 (1942) . 9. Bornstein, B. T., and H. A. Barker, /. Bacteriol, 55, 223 (1948) . 10. Bornstein, B. T., and H. A. Barker, /. Biol. Chem., 172, 659 (1948) . 11. van Beynum, J., and J. W. Pette, Zentr. Bakteriol. Parasitenk., II, 93, 198 (1935). 12. Bhat, J. V., and H. A. Barker, /. Bacteriol., 54, 381 (1947) ; 56, 777 (1948). 13. Barker, H. A., and V. Haas, J. Bacteriol., 47, 301 (1944) . 14. Barker, H. A., M. D. Karaen, and V. Haas, Proc. Natl. Acad. Sci. U. S., 31, 355 (1945) . 15. Pine, L., and H. A. Barker, /. Bacteriol., 68, 216 (1954) . 16. Barker, H. A., and M. D. Kamen, Proc. Natl. Acad. Sci. U. S., 31, 219 (1945). 17. Barker, H. A., S. Ruben, and J. V. Beck, Proc. Natl. Acad. Sci. U. S., 26, 477 (1940). 18. Barker, H. A., B. E. Volcani, and B. P. Cordon, /. Biol. Chem., 17, 149 (1948). 19. Koepsell, H. J., and Marvin J. Johnson, J. Biol. Chem., 145, 379-386 (1942) . BUTYRIC ACID-BUTANOL FERMENTATIONS 55 20. Koepsell, H. J., M. J. Johnson, and J. S. Meek, /. Biol. Chem., 154, 535 (1944). 21. Stadtman, E. R., and H. A. Barker, /. Biol. Chem., 180, 1085, 1095, 1117 (1949). 22. Stadtman, E. R., and H. A. Barker, J. Biol. Chem., 181, 221 (1949) . 23. Kennedy, E. R., and H. A. Barker, /. Biol. Chem., 191, 419 (1951) . 24. Barker, H. A., "Recent Investigations on the Formation and Utiliza- tion of Active Acetate," Phosphorus Metabolism, Vol. 1, edited by W. D. McElroy and B. Glass, The Johns Hopkins Press, Bal- timore, 1951, p. 204. 25. Kaplan, N. O., and F. Lipmann, /. Biol. Chem., 174, 37 (1948) . 26. Novelli, G. D., "The Structure of Coenzyme A," Phosphorus Metab- olism, Vol. 1, edited by W. D. McElroy and B. Glass, The Johns Hopkins Press, Baltimore, 1951, p. 414. 27. Lynen, F., E. Reichert, and L. Rueff, Ann. Chem., 574, 1 (1951) . 28. Stadtman, E. R., G. D. Novelli, and F. Lipmann, /. Biol. Chem., 191, 365 (1951) . 29. Stadtman, E. R., /. Biol. Chem., 196, 527, 535 (1952) . 30. Burton, R. M., and E. R. Stadtman, /. Biol. Chem., 202, 873 (1953) . 31. Rose, T. A., M. Greenberg-Manago, S. R. Korey, and S. Ochoa, /. Biol. Chem., 211, 737 (1954) . 32. Wolfe, R. S., and D. J. O'Kane, /. Biol. Chem., 205, 755 (1953) . 33. Korkes, S., A. del Campillo, I. C. Gunsalus, and S. Ochoa, J. Biol. Chem., 193, 721 (1951). 34. Stadtman, E. R., and H. A. Barker, /. Biol. Chem., 184, 769 (1950) . 35. Stadtman, E. R., J. Biol. Chem., 203, 501 (1953) . 36. Stadtman, E. R., Federation Proc., No. 3, 12, 692 (1956) . 37. Peel, J. L., and H. A. Barker, Biochem. ]., 62, 323 (1956) . 38. Bartsch, R. G., Reactions of Unsaturated Fatty Acids in Clostridium kluyveri Extracts, dissertation, University of California, 1956. 39. Crane, F. L., S. Mii, J. G. Hange, D. E. Green, and H. Beinert, J. Biol. Chem., 218, 701 ( 1956) . 40. Stern, J. R., A. del Campillo, and I. Rau, /. Biol. Chem., 218, 971 (1956) . 41. Soodak, M., and F. Lipmann, /. Biol. Chem., 175, 999 (1948) . 42. Stadtman, E. R., M. Douderoff, and F. Lipmann, /. Biol. Chem., 191, 377 (1951). 43. Lynen, F., E. Reichert, and L. Rueff, Ann. Chem., 574, 1 (1951) . 44. Lynen, F., L. Wessely, O. Wieland, and L. Rueff, Angew. Chem., 64, 686 (1952). 45. Lynen, F., Harvey Lectures, 48, 210 (1952-1953) . 56 BACTERIAL FERMENTATIONS 46. Lynen, F., and S. Ochoa, Biochem. et Biophys. Acta, 12, 290 (1953) . 47. Goldman, D. S., /. Biol. Chem., 208, 345 (1954) . 48. Stern, J. R., M. J. Coon, and A. del Campillo, /. Biol. Chem., 221, 1 (1956). 49. Stern, J. R., and A. del Campillo, /. Biol. Chem., 218, 985 (1956) . 50. Wakil, S. J., and H. R. Mahler, /. Biol. Chem., 207, 125 (1954) . 51. Wakil, S. J., D. E. Green, S. Mii, H. R. Mahler, /. Biol. Chem., 207, 631 (1954). 52. Green, D. E., S. Mii, and H. R. Mahler, /. Biol. Chem., 206, 1 (1954) . 53. Stadtman, E. R., Record Chem. Progr., 15, 1 (1954) . 54. Lehninger, A. L., and G. D. Greville; Biochem. et Biophys. Acta, 12, 188 (1953). 55. Stern, J. R., A. del Campillo, and A. L. Lehninger, J. Am. Chem. Soc, 77, 1073 (1955) . 56. Stadtman, E. R., unpublished observation. 57. Stadtman, E. R., T. C. Stadtman, and H. A. Barker, /. Biol. Chem., 178, 677 (1949) . 58. Lieberman, I., and H. A. Barker, /. Bacteriol., 68, 329 (1954) . 59. Johnson, M. J., W. H. Peterson, and E. B. Fred, /. Biol. Chem., 101, 145 (1933). 60. Davies, R., Biochem. J., 37, 230 (1943) . 61. Stern, J. R., M. J. Coon, A. del Campillo, and M. C. Schneider, J. Biol. Chem., 221, 15 (1956) . CHAPTER FERMENTATIONS OF NITROGENOUS COMPOUNDS In 1934 the biochemical transformations involved in the anaerobic breakdown of nitrogenous compounds had not been extensively studied. Various proteinaceous materials were known to undergo extensive decomposition in the absence of oxygen with the formation of ammonia, carbon dioxide, fatty acids, hydrogen sulfide, and an assortment of other volatile substances frequently having unpleasant or repulsive odors. Such "putrefactive" processes had been found to be caused mainly by Clostridia, several species of which had been shown to grow in media containing proteins and amino acids but no carbohydrates. However, very little information was available concerning the specific reactions catalyzed by individual species. The reasons for this are not hard to find. With some conspicuous exceptions, most of the studies in this area had been done either by chemists who disregarded one of the basic requirements for sound microbiological work, namely the use of pure cultures, or by bacteriologists who were unequipped to handle the fre- 57 58 BACTERIAL FERMENTATIONS quently difficult problems of chemical analysis. The chemists typically inoculated a nutrient solution containing a relatively pure amino acid with a piece of decaying meat, waited several months for the microbial activity to cease, and then identified the products. The bacteriologists typi- cally inoculated a pure culture of a Clostridium into a complex protein hydrolyzate and in the end were unable to identify either the specific amino acids decomposed or all of the products formed. One of the most substantial contributions in both a chemical and microbiological sense to knowledge of the anaerobic decomposition of amino acids was made by Stickland in 1934.1 He studied the action of washed sus- pensions of Clostridium sporogenes on amino acids. By the use of both manometric and dye-coupling techniques he demonstrated that, whereas most single amino acids are not readily attacked by CI. sporogenes, certain pairs of amino acids are decomposed rapidly by a coupled oxidation- reduction reaction; one member of the pair is oxidized while the other member is reduced. The simplest example of this type of reaction is that involving alanine and glycine. As the following equations show, the oxidation of alanine to acetic acid, carbon dioxide, and ammonia is coupled with the reduction of 2 moles of glycine to acetic acid and ammonia. CH3CHNH2COOH + 2H20 — > CH3COOH + C02 + NH3 + 4H (1) 2CH2NH2COOH + 4H — > 2CH3COOH + 2NH3 (2) CH3CHNH2COOH + 2CH2NH2COOH + 2H20 — >- 3CH3COOH + 3NH3 + C02 (3) Certain other amino acids react in similar ways. Other FERMENTATIONS OF NITROGENOUS COMPOUNDS 59 amino acids that can be oxidized include leucine and valine which are converted to the branched-chain fatty acids with one less carbon atom. Some of the amino acids that serve as oxidants, such as arginine, are reduced like glycine with the formation of ammonia. Others do not give ammonia but are reduced with the opening of a ring; proline for example is converted to S-aminovaleric acid (equation 4) . CH2— CH2 CH2— CH2 CR2 CH— COOH + 2H — > CH2 CH2— COOH (4) \/ \ NH NH2 Proline 5-Aminovaleric acid These reactions provide mechanisms for the decomposition of a variety of amino acids. Stickland in fact suggested that they could account for most if not all of the anaerobic decomposition of amino acids. In support of this idea, Clifton, Cohen, Nisman, and others2 later showed that some fifteen species of Clostridia, including CI. botulinum, CI. butyricum, and CI. histolyticum, make use of the Stickland reaction. However, the occurrence of this type of amino acid decomposition in many Clostridia in no way excluded the possibility that other mechanisms for amino acid break- down could also be operative and have quantitative significance. Considerably before the work of Stickland, the ferment- ability of single amino acids had already been observed by Naviasky, by Brasch, and by Liebert. Naviasky3 was per- haps the first investigator to observe the anaerobic decom- position of a purified nitrogenous compound by a single bacterium. In 1908 he showed that a pure culture of Bacil- lus proteus vulgaris could ferment asparagine with the for- 60 BACTERIAL FERMENTATIONS mation of several products among which succinate, acetate, carbon dioxide, and ammonia were positively identified. Two years later, Brasch4 isolated from a medium containing glutamate as the only carbon and nitrogen source an obligate anaerobic bacterium which he identified with Bienstock's Bacillus putrificus and reported that it could ferment glutamate with the formation of ammonia, formate, and butyrate. At about the same time Liebert,5 working in Beijerinck's laboratory, described a fermentation of uric acid by an organism he had isolated from a uric acid enrich- ment culture and named Bacillus acidi-urici. Other experi- ments on the anaerobic decomposition of single nitrogenous compounds made during the same period were mostly carried out with mixed cultures. They showed that several other amino acids including lysine, arginine, aspartate, serine, and alanine probably can be fermented. Following this flurry of research on amino acid fermenta- tions, interest in this subject subsided and was not reawak- ened until some twenty-five years later when van Niel6 began to investigate the possibility that amino acids and purines can serve as energy sources for some fermentative bacteria just as carbohydrates do for others. He also made use of the enrichment culture method in a search for organisms that can attack various nitrogenous compounds in the absence of oxygen, and he found, in confirmation of Brasch and Liebert, that glutamate and urate are particularly sus- ceptible to decomposition. With both substrates Clostridia appeared to be the predominant organisms in the enrich- ment cultures. My associates and I continued the study of the anaerobic bacteria that can be obtained from soil and decomposing organic materials by the use of enrichment cultures supplied with single nitrogenous compounds, and over a period of years we isolated a number of species capable of fermenting FERMENTATIONS OF NITROGENOUS COMPOUNDS M several such compounds. A list of these species and the substrates used for their isolation are given in Table 1. As has already been mentioned, glutamate is readily fermented by Clostridia which are abundant and widely dis- tributed in soils.7 All of the glutamate-fermenting strains we have isolated belong to one of two closely related species TABLE 1 Some Amino Acid-, Purine-, or Pyrimidine-Fermenting Bacteria Isolated by the Enrichment Culture Method Enrichment Substrate Species Glutamate CI. tetanomorphum CI. cochlearium Histidine CI. tetanomorphum Glycine Diplococcus glycinophilus Alanine CI. propionicum Uric acid CI. acidi-urici CI. cylindrosporum Allantoin Streptococcus allantoicus Orotic acid Zymobacterium oroticum which have been known for a long time, namely CI. coch- learium and CI. tetanomorphum. The latter species has been isolated also by enrichment with histidine.8 Besides decomposing glutamate and histidine, CI. tetanomorphum is also known to ferment cysteine, serine, aspartate, and tyrosine,9 but the efficacy of these compounds as enrichment substrates has not been tested. Enrichment with glycine10 has given cultures in which a non-motile coccus Diplococcus glycinophilus is the pre- dominant organism. This species is highly specialized with respect to its energy source since it appears to be limited to the use of glycine and some simple peptides containing glycine; it is unable to attack a variety of other amino acids 62 BACTERIAL FERMENTATIONS or carbohydrates. Such a restricted diet is unusual even among bacteria. Almost nothing is known about the dis- tribution of this organism. Alanine is a good substrate for the enrichment of an anaerobic spore former which we have called CI. propioni- cum.10 This organism, which also ferments serine, threo- nine, lactate, pyruvate, and acrylate but not glucose, appears to be rather widely distributed in soils since it has been isolated in England, New Zealand, and several parts of the United States. Uric acid enrichments from soil or mud invariably lead to the development of Clostridia similar to Liebert's Bacillus acidi-urici. Out of thirteen strains isolated from different soils,11 twelve proved to be indistinguishable from Liebert's organism and therefore were called Clostridium acidi-urici in accordance with modern nomenclature. The remaining strain was physiologically similar to the other twelve but was quite different morphologically, and there- fore was called CI. cylindrosporum. Both species decom- pose uric acid and certain other purines with great facility but do not attack carbohydrates or amino acids, except glycine and serine. In view of their highly developed ability to attach purines and their wide distribution in soils, these species may be largely responsible for the anaerobic decomposition of uric acid in nature. Several attempts to obtain anaerobic bacteria capable of growing on allantoin from several soil and mud samples were unsuccessful, indicating that such organisms are not common in nature; but eventually a vigorous fermentation of allantoin was obtained starting with an inoculum of black mud, and the causative organism was found to be a large streptococcus which we have called 5. allantoicus.12 This is one of the few species of the genus that is known to obtain energy for growth from a nitrogenous compound. FERMENTATIONS OF NITROGENOUS COMPOUNDS 63 Zymobacterium oroticum was isolated from mud by Romberg13 by enrichment with orotic acid. This organ- ism grows slowly at the expense of orotic acid in a relatively complex medium containing tryptone and yeast extract. It also ferments glucose and several other sugars, but not glutamate or lactate. Zymobacterium oroticum has been isolated only once, so nothing can be said about its natural occurrence. This organism is of special interest because it has been used by Lieberman and Romberg14-15 to eluci- date some of the enzymatic steps in pyrimidine biosynthesis. The application of the enrichment culture technique has clearly established the existence of a variety of bacteria capable of fermenting single nitrogenous compounds. Sev- eral of these organisms were not previously known. In general, these bacteria show a considerable specialization for the utilization of one or a few structurally related nitrogenous substrates. This specialization is most highly developed in D. glycinophilus and the purine fermenting Clostridia, and least developed in CI. tetanomorphum, which ferments a number of amino acids and other compounds. The substrates mentioned in Table 1 are not the only nitrogenous compounds suitable for the enrichment of anaerobic bacteria. Serine and threonine are readily fer- mented in enrichment cultures, although the species mainly responsible for their decomposition under such conditions have not as yet been identified. We have also observed that leucine, valine, cysteine, aspartate, uracil, and thymine can serve as anaerobic enrichment substrates. Probably still other compounds would also be suitable. Further explora- tion of amino acid, purine, and pyrimidine decomposition by the enrichment culture technique will undoubtedly lead to a better understanding of both the biochemistry and microbiology of fermentative processes. Of course, knowledge of amino acid fermentations has 64 BACTERIAL FERMENTATIONS not been obtained only by the study of organisms isolated from highly selective enrichment media. Several investi- gations have been carried out on anaerobic bacteria obtained by direct isolation from soil, spoiled foods, infected wounds, and similar sources. Species that apparently derive most of their energy from the fermentation of single amino acids include CI. tetania6'17 CI. saccharobutyricum,18 Micrococ- cus aerogenes,19 and Fusobacterium nucleatum,20 which use glutamate, histidine and serine, sometimes threonine and cysteine, and occasionally aspartate, lysine, and tyrosine. Other species in this group are M. anaerobius and M. vari- abilis,21 which like D. glycinophilus appear to be restricted to the utilization of glycine. In addition, a number of bacteria, both obligate and facultative anaerobes, are able to decompose some or all of the amino acids which readily undergo a non-oxidative deamination, namely serine, cys- teine, threonine, and arginine. These bacteria include CI. welchii,22 CI. sporogenes,23 CI. botulinum,24 E. coli25-26 Proteus vulgaris,27'28 P. morgani,29 and an anaerobic coccus isolated from the rumen of the sheep.30 A conversion of lysine to acetate, butyrate, and ammonia by CI. sticklandii has been reported by Stadtman,31 but this process alone does not support growth. Chemistry of the Fermentations of Nitrogen Compounds Fermentations of Histidine and Glutamate by Clos- tridium tetanomorphum. Clostridium tetanomorphum ferments histidine and glutamate with great facility. In view of the evidence that the decomposition of histidine goes by way of glutamate or glutamate derivatives in both aerobic bacteria32 and mammalian liver33 it seemed desir- able to find out whether the same or similar reactions are utilized by anaerobic bacteria. FERMENTATIONS OF NITROGENOUS COMPOUNDS 65 The pathway of L-histidine decomposition in Pseudo- monas fluoresceins32 involves the following sequence: tt. . f. -NH3tt +2H20^ . . -NH3 L-Histidme — >■ Urocanate — > Formimmoglutamate — > Formyl glutamate — > Formate + L-Glutamate Clostridium tetanomorphum carries out a similar sequence of reactions with certain modifications (Fig. 1) . Urocanate has been shown to be formed from histidine and further decomposed by washed cells.34-35 The evidence for formiminoglutamate (formamidinoglutarate) is less con- clusive, although possibly adequate. Both washed cells and an acetone powder have been observed to convert histi- dine to a compound which on alkaline hydrolysis yields 1 mole each of glutamate and formate and 2 moles of ammonia; these products are known to be formed under the same conditions from formiminoglutamate. When chromatographed on paper the compound has the same R; as the formiminoglutamate formed enzymatically by liver enzymes, and gives the same color with a nitroprusside- ferricyanide reagent that reacts with the formimino group. The enzymatic decomposition of formiminoglutamate has been observed but has not been extensively investigated in CI. tetanomorphum. Presumably this is a tetrahydrofolic acid-dependent reaction as are the analogous reactions in liver36 and in CI. cylindrosporum.37 The immediate prod- ucts of formiminoglutamate decomposition have not been determined. However, glutamate has been detected in small amounts during the fermentation of histidine, and formamide* accumulates as a major end product of the fermentations of both histidine and urocanate. Tracer experiments have demonstrated that formamide is formed * Formamide is also formed from histidine by Aerobacter aero- genes.z* 66 BACTERIAL FERMENTATIONS from the 8 nitrogen and carbon atom 2 of the imidazole ring of histidine.8 These results are consistent with the reaction sequence involving formiminoglutamate shown in Fig. 1. HC- -N HC- \ CH -NH •NH3 G- ■N V CH -NH +2H20 CH2 CH II CH 1 CHNH2 1 COOH COOH L-Histidine Urocanic acid COOH COOH CH— NH— CH=NH CH— NH2 H20 CH2 I CH2 CH2 I CH2 + HCO— NH2 COOH COOH Formiminoglutamic acid L-Glutamic acid Formamide Fig. I. Conversion of Histidine to Glutamate by Clostridium tetanomorphum. The available information indicates that the main dif- ference in the paths of histidine decomposition by CI. teta- nomorphum and Ps. fluorescens probably lies in the reac- tions of formiminoglutamate. Whereas CI. tetanomorphum removes the terminal nitrogen and carbon atoms of the formamidine side chain simultaneously, Ps. fluorescens removes them by successive hydrolytic reactions with formyl glutamate as an intermediate. FERMENTATIONS OF NITROGENOUS COMPOUNDS 67 The final products of the fermentation of glutamate by CI. tetanomorphum are butyrate, acetate, carbon dioxide, ammonia, and hydrogen.39 These products suggest that the terminal stages of the glutamate fermentation are sim- ilar to those of the butyric acid fermentation of carbohy- drates and other non-nitrogenous substrates. However, one unusual and significant feature appears when the quan- titative relations are considered, namely, only 1 mole of carbon dioxide is formed per mole of glutamate decom- posed. This is significant because in most organisms that have been studied glutamate is converted to acetate via a-ketoglutarate and the tricarboxylic acid cycle, and this reaction sequence produces not 1 but 5 molecules of carbon dioxide per molecule of glutamate decomposed; conse- quently the observed yield of carbon dioxide excludes the participation of the tricarboxylic acid cycle functioning in the forward or oxidative direction in the glutamate fermentation. In order to define the path of glutamate fermentation more clearly, Wachsman investigated the fate of the indi- vidual carbon atoms of glutamate by means of tracer experi- ments.40 Samples of glutamate labeled with C14 in specific positions were fermented by washed cells, and the products were isolated and degraded to locate the isotope. Figure 2 summarizes the results. Carbon atoms 1 and 2 of glutamate are converted mainly to acetate. Carbon atoms 3 and 4 are converted mainly to butyrate, carbon atom 3 appearing in the a and y positions, and carbon atom 4 in the carboxyl and ft positions; carbon atoms 3 and 4 also go to the methyl and carboxyl positions of acetate to a small extent. Carbon atom 5 of glutamate appears entirely in carbon dioxide. These results show that two types of C2 units are formed from glutamate, one of which, derived from carbon atoms 1 and 2, is converted directly to acetate, whereas the other, 68 BACTERIAL FERMENTATIONS derived from glutamate carbon atoms 3 and 4, gives rise to an active acetyl group which is preferentially converted to butyrate. The acetyl group and carbon dioxide might well be formed from pyruvate derived from carbon atoms 5, 4, and 3. Therefore, a fission of the glutamate carbon ^OOH 1 2CHNH2 f I J 3CH2 I 4CH2 5COOH ^OOH 2CH3 ■3CH3 ■ I 4CO L5coohJ IT 1 L4co— xj rCH3COOH CH3CH2CH2COOH 4CO + 5co2 Fig. 2. Fermentation of C14-Glutamate by Clostridium fetanomorphum. chain between carbon atoms 2 and 3 to give acetate and pyruvate is consistent with the tracer experiments. The results exclude the occurrence of a tricarboxylic acid cycle in a reverse as well as a forward direction. The operation of the cycle in the reverse direction, i.e., from a-ketogluta- rate to oxalacetate and acetate via citrate, would give an acetyl group in which the methyl carbon atom was derived from glutamate carbon atom 4, instead of carbon atom 3 as observed. The further exploration of the pathway of glutamate fermentation has been done with enzyme preparations. Wachsman41 found that cell-free extracts of CI. tetano- FERMENTATIONS OF NITROGENOUS COMPOUNDS 69 morphum can readily be obtained which convert glutamate in the absence of oxygen to ammonia, carbon dioxide, hydrogen, acetate, and several minor products which have been identified as pyruvate, «-ketoglutarate, mesaconate, and citramalate.42 Mesaconate, which is a branched-chain unsaturated dicarboxylic acid, was first detected by paper chromatography as an acidic spot which also appeared as a dark ("quenching") spot when viewed by ultraviolet light. The characteristic ultraviolet quenching of mesaconate is attributable to a conjugated system of double bonds in the molecule. Subsequently both mesaconate and citramalate were isolated in quantity and characterized more fully. A variety of enzymatic and tracer experiments have dem- onstrated that mesaconate and citramalate are intermediates in the fermentation of glutamate. As shown in Fig. 3, glutamate is converted to mesaconate and ammonia, the double bond of mesaconate is hydrated to give citramalate, and citramalate is converted to pyruvate and acetate. The formation of acetate, butyrate, carbon dioxide, and hydro- gen from pyruvate presumably involves reactions of the types discussed in Chapter 2. The formation of mesaconate from glutamate and its further decomposition have been shown to occur at rates as great as the over-all rate of glutamate fermentation by cell-free extracts. Under conditions which block its hydra- tion, mesaconate can be made to accumulate in almost theoretical yield, i.e., 1 mole per mole of glutamate. The formation of mesaconate from glutamate is of considerable interest, because it involves the interconversion of straight- and branched-chain compounds without loss of carbon atoms. This process undoubtedly is the result of a fairly complex reaction sequence, the details of which have not yet been worked out. However, some information concern- ing the origin of the methyl group of mesaconate has been 70 BACTERIAL FERMENTATIONS obtained. Munch-Petersen43 has studied the conversion of glutamate-4-C14 to mesaconate and has shown that the C14 is all present in the carbon atom adjacent to the methyl group (see Fig. 3) . This means that the methyl carbon must be derived from carbon 3 of glutamate and excludes any mechanism of methyl group formation involving a iCOOH iCOOH 2CHNH2 2CH =FNH3 s ±h2o 3CH2 ^=±r 4C— COOH ^^ 4CH2 3CH3 5COOH Glutamate Mesaconate iCOOH i !COOH 2CH2 2CH3 4| 5 HOG— COOH + 5 : 4COCOOH 3CH3 3CH3 (+)-Citramalate Acetate and pyruvate Fig. 3. Role of Mesaconate and Citramalate in Glutamate Decomposition. transfer of the carboxyl carbon 5 of glutamate from carbon 4 to carbon 3, which would be similar to the interconver- sion of methyl malonate and succinate reported by Flavin et al.44 The conversion of glutamate to mesaconate appar- ently involves a transfer of carbon atoms 1 and 2 of gluta- mate from carbon atom 3 to carbon atom 4. This would leave carbon atom 3 in the position occupied by the methyl group of mesaconate. The conversion of mesaconate to citramalate involves a FERMENTATIONS OF NITROGENOUS COMPOUNDS 71 hydration of the double bond. The product of the reaction has been identified by chemical and enzymatic methods as (+) -citramalate. Cysteine and Fe++ are cofactors for the reaction, and orthophenanthroline and other reagents that combine strongly with Fe++ are powerful inhibitors. As in the aconitase and fumarase reactions, the equilibrium favors the formation of the hydroxy acid. Citramalate is cleaved by a reversible aldolase-type reac- tion to acetate and pyruvate. The enzyme citramalase which catalyzes this reaction shows a high degree of speci- ficity since it attacks (+) -citramalate but neither the levo- rotatory isomer nor several structurally related compounds including citrate, isocitrate, and malate. The equilibrium of the reaction favors the cleavage of citramalate. The analysis of the CI. tetanomorphum system has estab- lished the existence of a new pathway of glutamate metab- olism which may be regarded as a substitute for a part of the tricarboxylic acid cycle in the sense that it permits a conversion of glutamate to pyruvate, acetate, and carbon dioxide. All of the reactions between glutamate and pyru- vate are readily reversible and involve only small free- energy changes. Therefore this sequence differs from the oxidative part of the tricarboxylic acid cycle in not provid- ing energy for the synthesis of adenosine triphosphate. The occurrence of the mesaconate pathway in other organisms has not yet been investigated. However, indirect evidence for this pathway in several anaerobic bacteria, in- cluding CI. botulinum, CI. tetani, CI. saccharobutyricum, and Fusobacterium nucleatum, is provided by the observa- tion that these species, like CI. tetanomorphum, form 1 mole of carbon dioxide per mole of glutamate fermented. The occurrence of mesaconate in sugar cane and cabbage, and of (— ) -citramalate in apples, indicates that these compounds also have a role in the metabolism of higher plants. 72 BACTERIAL FERMENTATIONS Glycine Fermentation by Diplococcus glycinophilus. The fermentation of glycine45 proceeds according to equa- tion 5. In addition, a small amount of hydrogen is pro- 4CH2NH2COOH + 2H20 — > 4NH3 + 2C02 + 3CH3COOH (5) duced, but since the evolution of hydrogen is readily re- versed by a partial pressure of the gas exceeding one-third of an atmosphere, the formation of hydrogen soon ceases in stationary cultures in which the medium quickly becomes saturated with hydrogen. The glycine fermentation presents some interesting prob- lems of intermediary metabolism which so far have been studied only by tracer methods. The results show plainly that the chemistry of glycine decomposition by D. glyci- nophilus is entirely different from that catalyzed by Clos- tridium sporogenes. Since the latter organism reduces glycine to acetate and ammonia, one might expect that the glycine fermentation by D. glycinophilus would consist of a complete oxidation of 1 mole of glycine to carbon dioxide coupled with reduction of 3 moles of glycine to acetate. This possibility was tested by fermenting glycine- 1-C14, glycine-2-C14, and unlabeled glycine plus C14-carbon diox- ide and determining the distribution of isotope in the products. The results show that the carbon dioxide is derived entirely from carboxyl carbon of glycine, whereas both carbons of acetate are derived partly from the meth- ylene carbon of glycine and partly from carbon dioxide. These results exclude both a direct reduction of glycine to acetate and a complete oxidation of glycine to carbon dioxide. They suggest an oxidative conversion of glycine to acetate and carbon dioxide, possibly via serine, coupled with a reduction of carbon dioxide to acetate. FERMENTATIONS OF NITROGENOUS COMPOUNDS 73 Alanine and Threonine Fermentations by Clostridium prop/on icu/n. Clostridium propionicum is able to ferment alanine, serine, threonine, lactate, pyruvate, and a:rylate.45 With all these substrates except threonine, the main prod- ucts are acetate, propionate, and carbon dioxide; ammonia is also formed from the amino acides. Equation 6 describes 3CH3CHNH2COOH + 2H20 — > 3NH3 + 2CH3CH2COOH + CH3COOH + CO2 (6) the fermentation of alanine. Except for ammonia, these products are the same as those formed by the propionic acid bacteria (genus Propionibacterium) . Despite the simi- larities in the products, a considerable amount of evidence indicates that the chemical reactions in the CI. propionicum and Propionibacterium fermentations are substantially different. Succinate is a characteristic product of the Pro- pionibacterium fermentation, and Whiteley46 has shown that succinate is converted to succinyl-CoA and then decar- boxylated to propionyl-CoA, the immediate precursor of propionate. Clostridium propionicum on the contrary neither forms nor decarboxylates succinate.47 Further evi- dence against the participation of succinate is the absence of randomization of the a and ft carbons of lactate during its conversion to propionate.48 Also Johns47 has shown that CI. propionicum, unlike the propionic acid bacteria, is unable to incorporate carbon dioxide into propionate. Perhaps the first indication of a distinctive pathway of propionate formation in CI. propionicum was provided by the discovery that acrylate is fermented at least as rapidly as alanine or lactate.45 This led to the suggestion that the latter substrates might be converted by loss of ammonia or water to acrylate, which might then be reduced to pro- pionate as follows: 74 BACTERIAL FERMENTATIONS CH3CHNH2COOH . NHj p CH2=CHCOOH -ini CH3CHOHCOOH ^H2° CH3CH2COOH (7) The occurrence of such a sequence in the reverse direction was investigated by Stadtman49 by studying the oxidation of propionate by cell-free extracts of CI. propionicum. He found that propionate was oxidized only after being con- verted to propionyl CoA. The supposed product of the oxidation, acrylyl CoA, could not be directly identified. However, indirect evidence for this compound was provided by the finding that acrylyl thioesters react enzymatically with ammonium ion to form /?-alanyl thioesters and ulti- mately /^-alanine itself in this system (equation 8) , and +nh3 H2O CH2=CHCOSR --> CH2NH2CH2COSR -4- CH2NH2CH2COOH + HSR (8) that /^-alanine is also formed in the oxidation of propionyl thioesters. Presumably the acrylyl thioester is so reactive that it cannot accumulate to an appreciable extent under the experimental conditions used. These experiments sug- gest that acrylyl CoA may be a precursor of propionate in CI. propionicum. However, as yet a conversion of alanine or lactate to acrylyl CoA has not been demonstrated nor is there any indication of the role of /^-alanine in these fermentations. Threonine is exceptional among the substrates fermented by CI. propionicum in that it gives rise to butyrate and propionate (equation 9) instead of propionate and ace- tate.50 This difference is obviously dependent on the fact that threonine is the only four-carbon compound attacked. Evidently the carbon chain of 1 mole of threonine is reduced FERMENTATIONS OF NITROGENOUS COMPOUNDS 75 to butyrate whereas 2 moles of threonine are oxidized to propionate and carbon dioxide. The specific reactions in- 3CH3CHOHCHNH2COOH + 2H20 — *- CH3CH2CH2COOH + 2CH3CH2COOH + 2C02 + 3NH3 (9) volved in tllese processes have not been investigated but it is probable that a-ketobutyrate is an intermediate in the fermentation. AHantoin Fermentation by Streptococcus allantoicus. The decomposition of allantoin by S. allantoicus cannot be represented by a simple equation. The major products are ammonia, urea, oxamic acid, and carbon dioxide, and the minor products are formic, acetic, lactic, and glycolic acids.12 Oxamic acid, the monoamide of oxalic acid, has not been found in any other fermentation. The chemistry of the allantoin fermentation has not been extensively studied. The first step is known to be the open- ing of the ring with the formation of allantoic acid. Beyond this little is known with certainty except that the nitrogen NH2 CO— NH / CO CO + h2o NH CH— NH Allantoin NH2 COOH NH2 CO CO (10) NH CH NH Allantoic acid of oxamic acid is not derived from ammonia. This has been shown by experiments with N15 ammonia. Since ammonia is formed in the fermentation and the organism 76 BACTERIAL FERMENTATIONS cannot convert urea to ammonia, it may also be concluded that at least one of the ureido groups of allantoic acid is decomposed before being separated from the two-carbon chain. The latter is ultimately oxidized in part to oxamic acid, carbon dioxide, and formate, and partly reduced to glycolate and acetate by paths that are yet to be determined. Orotic Acid Fermentation by Zymobacterium orofl- cuiti. Not all of the fermentation products of orotic acid have been identified, but ammonia, carbon dioxide, acetate, and one or more four-carbon dicarboxylic acids appear to be formed. The early reactions in the fermentation of orotic acid by this bacterium have been extensively studied by Lieberman and Romberg14-15 by the use of cell-free extracts. The first reaction is the DPN-linked reduction of orotic acid to L-dihydroorotic acid. This is followed by a hydrolytic cleavage of the pyrimidine ring to give L-ureido- succinic acid. HN— CO HN— CO II II OC CH +DPNH+H+ :^=± OC CH2 +DPN+ I II II HN— C— COOH HN— CH— COOH Orotic acid L-Dihydroorotic acid ±H20 COOH I I I CH2 +CO2+NH3 OC CH2 I I I H2N— CH— COOH ^- HN— CH— COOH L-Aspartic acid L-Ureidosuccinic acid The ureido group of the latter compound is degraded with the formation of ammonia, carbon dioxide, and L-aspartic acid. The further transformations of aspartic acid have not been analyzed in detail, but it is probable in view of the FERMENTATIONS OF NITROGENOUS COMPOUNDS 77 nature of the observed fermentation products that aspartic acid is converted either to fumarate or oxalacetate which is then partly oxidized to acetate and carbon dioxide, and partly reduced to succinate. The energy-yielding steps could be the decomposition of the ureido group and the oxidation of pyruvate; however, no phosphorylation reac- tions have yet been demonstrated in this organism. Both the reduction of orotic acid and the conversion of the latter to ureidosuccinic acid are readily reversible reac- tions and therefore permit the synthesis of the pyrimidine orotic acid from non-pyrimidine precursors. Considerable evidence now indicates that this reaction sequence repre- sents a major path of pyrimidine synthesis in living organisms.51 Purine Fermentations by Clostridia. Clostridium acidi-urici and CI. cylindrosporum readily ferment xan- thine, uric acid, guanine, and 6,8-dihydroxypurine, and decompose hypoxanthine, guanosine, and inosine more slowly or after a period of adaptation.52 Chemical studies of purine fermentations by growing cultures have estab- lished that the main products formed by CI. acidi-urici are ammonia, carbon dioxide, and acetate; in addition a small N=COH I I HOC C N \ CH+6H2O — ►■ / N— C— NH 4NH3+3CQ2+lCH3COOH (11) amount of formate accumulates. If, as a first approxima- tion, we disregard the formate, the decomposition of xan- thine is represented by equation 11. With uric acid, the yield of carbon dioxide is increased and the yield of acetate 78 BACTERIAL FERMENTATIONS is decreased; with hypoxanthine the reverse is true. With all three substrates, the purine nitrogen is converted essen- tially quantitatively to ammonia, without any detectable accumulation of urea. The same products are formed by CI. cylindrosporum, except that the yield of formate is higher, approximately 1 mole per mole of purine, and a considerable amount of glycine accumulates.53 An early attempt to get some evidence concerning the path of uric acid fermentation was made by investigating the utilization of various compounds previously implicated in the biological oxidation of uric acid.52 In particular, allantoin and urea, products of uric acid oxidation by both animals and aerobic bacteria, were found not to be attacked by cell suspensions of CI. acidi-urici which rapidly fer- mented uric acid. This appeared to exclude the only known pathway of uric acid decomposition via allantoin and indi- cated that the aerobic and anaerobic pathways must be fundamentally different. An indication of the actual mechanism of the purine fermentation was provided by the discovery that glycine is not only formed by CI. cylindrosporum but is also decom- posed by both species under appropriate conditions. The activation of glycine was first shown by the methylene blue reduction technique. Of a considerable number of purines and amino acids tested as reducing agents, using cell suspensions, glycine was by far the most active. When cell suspensions were allowed to act upon glycine alone under anaerobic conditions, essentially no decomposition of the amino acid could be observed. However, when glycine and uric acid were supplied simultaneously, both compounds were decomposed and the amount of glycine consumed was dependent upon the amount of uric acid fermented. This relation, combined with the demonstra- tion that glycine could act as a reducing agent for methylene FERMENTATIONS OF NITROGENOUS COMPOUNDS 79 blue, suggested that the oxidation of glycine was coupled with the reduction of urate. Evidence was also obtained that the decomposition of uric acid under some conditions was dependent on the presence of glycine. Certain thor- oughly washed and aged cell suspensions of CI. acidi-urici could ferment uric acid only after a long lag period. The lag could be abolished completely by the addition of glycine.54 Further investigation of the role of glycine in the purine fermentation was delayed for several years. During this interval, Sonne, Buchanan, and Delluva55 studied the syn- thesis of uric acid in the pigeon and established by means of tracer experiments with C14 that various carbon atoms in the purine are derived from carbon dioxide, formate, and glycine. Uric acid carbon atoms 2 and 8 are derived from formate, carbon atoms 4 and 5 from the carboxyl and methylene groups of glycine respectively,56 and carbon atom 6 from carbon dioxide (Fig. 4) . The nitrogen in position 'NH— 6C0^— C02 HCOOH — ^2C0 pC— -NH! 8 3 1 iJL /CO <— HCOOH 3NH-t4C^-9NH T CH2— NH2 I COOH Fig. 4. Uric Acid Synthesis in Pigeons. 7 was shown also to be derived from glycine.57 Therefore glycine appeared to be incorporated intact into the adjacent positions 4, 5, and 7 of the purine^ In view of the obvious similarities between the bacterial purine fermentations and purine synthesis in the pigeon, it seemed probable that the two processes would follow the same general pathway but in opposite directions, particu- 80 BACTERIAL FERMENTATIONS larly since fatty acid oxidation and synthesis in various organisms had been shown to be related in this way. Con- sequently we investigated the origin of the fermentation products by use of purines specifically labeled with C14. 1NH— CO— ^C02 C02-< — 2CO |5C— 7NH' I j || T" /CO— ^HCOOH 3NH-t4Cj-9NH CH2— NH2 I COOH Fig. 5. Uric Acid Breakdown by Clostridia. The results, summarized qualitatively in Fig. 5, show that the degradation of purines by Clostridia bears a close resem- blance to the general pattern of purine biosynthesis. Thus, carbon atoms 1 and 2 and the nitrogen atom of glycine are converted to purine carbon atoms 4, 5, and 7 respectively in the pigeon, and are formed from these same purine atoms in the fermentation. Carbon atoms 6 and 8 of purine arise from carbon dioxide and formate, respectively, in the biosynthesis and yield the same compounds in the fermen- tation. The only significant difference between the two processes is in the origin and fate of carbon atom 2 of purine; in biosynthesis this carbon atom is derived from formate, whereas in the fermentation it is converted to carbon dioxide. This difference is attributable to the fact that the primary synthetic product in the pigeon is hypo- xanthine ribotide, which contains a reduced carbon atom in the 2 position, whereas the purine that actually under- goes cleavage in the fermentation is probably xanthine, which contains an oxidized carbon atom in this position. Experiments on the fermentation of unlabeled purines in the presence of C14-labeled glycine, formate, or carbon FERMENTATIONS OF NITROGENOUS COMPOUNDS |H dioxide have confirmed the conclusion, already reached by other methods, that these compounds are not stable end products but are extensively further metabolized.53'58 The carboxyl carbon of glycine is converted mainly to carbon dioxide, whereas the methylene carbon is extensively incor- porated into both carbons of acetate. Formate is partly oxidized to carbon dioxide and partly incorporated into the methyl group of acetate. Finally, carbon dioxide is rapidly reduced to formate and is extensively incorporated into the carboxyl group of glycine and into both carbon atoms of acetate. Acetate itself is not metabolized to a sig- nificant extent. In addition to establishing the roles of carbon dioxide, formate, and glycine as intermediates, these observations also emphasize the complexity of the fermentation. I have already mentioned that xanthine is the purine which probably undergoes ring cleavage in the purine fer- mentation. The evidence for this, which is now fairly conclusive, was obtained in several ways. The first definite indication for the key role of xanthine was obtained by Beck59 in experiments on the influence of the growth sub- strate on the relative rates of decomposition of xanthine, guanine, uric acid, and hypoxanthine. He found that uric acid, hypoxanthine, and to a lesser extent guanine are de- composed by adaptive enzyme systems, whereas xanthine is rapidly utilized by cells grown on any of the four purines. This suggested that uric acid, guanine, and hypoxanthine are broken down via xanthine as follows: Guanine ., ■NH3 Hypoxanthine ~2> Xanthine <2H Uric acid i- i End products BACTERIAL FERMENTATIONS In support of this scheme Beck showed that cells grown on guanine contain an active guanase capable of converting guanine to xanthine.60 He also found that the decomposi- tion of uric acid by cells grown on this purine is inhibited by 2,4-dinitrophenol without affecting the fermentation of xanthine. Radin and Barker,61 using both dried-cell prep- arations and cell-free extracts of CI. acidi-urici, further showed that, whereas the decomposition of uric acid was increased by the presence of reducing agents and inhibited by oxidizing agents, the decomposition of xanthine was almost independent of such reagents. These and other types of evidence support the above scheme. The study of the enzymatic steps in the decomposition of xanthine was initiated by Radin and Barker61 and inde- pendently by Bradshaw and Beck.62 Both groups found that crude cell-free extracts readily convert xanthine to ammonia, carbon dioxide, formate, and glycine, or products which yield these substances in the analytical procedures used. Radin also made the important observation that one or more compounds which react in the Pauly and Bratton- Marshall tests were formed from xanthine in small amounts. The nature of the tests indicated that they were probably aminoimidazoles. These observations suggested that the pyrimidine ring of xanthine was first attacked in such a way as to form an imidazole derivative which was then further degraded. The individual enzymatic steps in the conversion of xan- thine to glycine have been extensively studied by Rabino- witz and his collaborators, who have developed methods for the isolation, analysis, and characterization of three imidazole derivatives and have clarified the reactions in- volved in the tetrahydrofolic acid dependent formation of glycine and formate. The decomposition of xanthine occurs by a hydrolytic FERMENTATIONS OF NITROGENOUS COMPOUNDS 83 cleavage of the pyrimidine ring between positions 1 and 6 and results in the formation of 4-ureido-5-imidazolecar- boxylic acid63 (see Fig. 6) . The accumulation of this com- NH— CO CO C— N NH2 COOH H20 \ CH NH— C— NH Xanthine COOH H20 -NH,, -C02 C— N \ CO C— N )CH NH— C— NH 4-Ureido-S -imidazolecarboxylic acid CH— N Sd H2Q -NH3 CH / H2N— C— NH 4-Amino-5 -imidazolecarboxylic acid [X] H20 H2N— C NH 4-Aminoimidazole CH2— NH V CH C HN /\ O OH Formiminoglycine Fig. 6. Conversion of Xanthine to Formiminoglycine. pound in crude cell-free extracts is made possible by the addition of sequestering agents such as ethylenediamine- tetraacetic acid which remove the Fe++ or Mn++ ions required for its further decomposition. The next enzymatic step is the cleavage of the ureido group to yield carbon dioxide, ammonia, and 4-amino-5-imidazolecarboxyl acid.64 The latter compound accumulates in an alkaline reaction mixture. The mechanism of this reaction has not been ex- amined in detail, but since orthophosphate is not required, the reaction appears to be hydrolytic. The aminocarboxy- 84 BACTERIAL FERMENTATIONS imidazole is enzymatically decarboxylated to 4-aminoimida- zole which accumulates in the presence of ethylenediamine- tetraacetic acid. The aminoimidazole is readily decomposed to ammonia and formiminoglycine by crude extracts of CI. cylindrosporum*5 Formiminoglycine (FIG) is also the product accumulating in the decomposition of xanthine by crude extracts. In earlier experiments on xanthine decom- position by extracts, FIG was mistakenly identified as a mixture of glycine, formate, and ammonia, because these compounds are readily formed non-enzymatically from FIG under conditions commonly used for their estimation. The conversion of 4-aminoimidazole to FIG is apparently a com- plex reaction involving more than one step. The enzyme responsible for 4-aminoimidazole decomposition has been partially purified and shown not to form formiminoglycine directly, but an unidentified compound (X in Fig. 6) which can be hydrolyzed non-enzymatically to glycine in the pres- ence of either acid or alkali. This compound is not an intermediate in the fermentation, since it is not converted to formiminoglycine by crude extracts. Probably it is formed from the true intermediate by a non-enzymatic side reaction. One of the most interesting parts of the purine fermenta- tion is the conversion of FIG to glycine, ammonia, and formate. This conversion, which evidently requires several enzymes, can provide useful energy to the organism in the form of ATP. Rabinowitz and Pricer66 have shown that FIG + ADP + P» — > GLYCINE + HCOOH + NH3 + ATP (12) under certain conditions the over-all reaction may be repre- sented by equation 12. This reaction sequence is depend- ent upon the presence of tetrahydrofolic acid (THFA) or FERMENTATIONS OF NITROGENOUS COMPOUNDS 85 other coenzyme form of folic acid. The initial step is a reversible transfer of the formimino group to THFA. The Formiminoglycine + THFA ^=^: Glycine + 5-Formimino-THFA (13) first evidence for this reaction consisted in the demonstra- tion by Sagers et al.67 that certain cell-free extracts catalyze an exchange between C14-glycine and formiminoglycine which is more rapid than the formation of ammonia and formate. Recently Rabinowitz and Pricer68 have purified the enzyme catalyzing this reaction and have partially char- acterized the product as 5-formimino-THFA. They have also shown that a second enzyme removes ammonia from 5-formimino-THFA to give 5,10-anhydroformyl-THFA (an- hydroleucovorin) which in turn is converted by a third enzyme to 10-formyl-THFA. The decomposition of 10-formyl-THFA can be coupled with the formation of ATP by the following reversible reaction: 10-Formyl-THFA + ADP + P, ^±: THFA + HCOOH + ATP (14) Perhaps this reaction also consists of more than one step since it involves an activation of orthophosphate (P^) and a transfer of the phosphate group to ADP. The above reactions account satisfactorily for the decom- position of xanthine to glycine, fprmate, carbon dioxide, and ammonia. However, in the fermentation of purine by living cells of CI. acidi-urici very little glycine and formate accumulate; the main organic product is acetate. There- fore it is necessary to account for the conversion of glycine and formate or their known precursors to acetate. 86 BACTERIAL FERMENTATIONS The most likely pathway for the formation of acetate is via serine and pyruvate as follows: CH2NH2COOH + "HCHO" — *- CH2OHCHNH2COOH j-NH3 CH3COOH + C02 ^ CH3COCOOH H2O This scheme postulates an hydroxymethylation of glycine by an active formaldehyde group which could be derived from the formimino group of formiminoglycine via formyl- THFA. The formation of serine has not yet been experi- mentally demonstrated in this system. The conversion of L-serine to pyruvate and the oxidation of pyruvate to acetate and carbon dioxide are catalyzed by both intact cells and extracts.69 Presumably the oxidation of pyruvate can be coupled with the formation of ATP although this has not yet been shown. The above scheme is further supported by a variety of tracer experiments.69,70 The carboxyl groups of both glycine and formiminoglycine are converted to carbon dioxide; glycine-2-C14 yields pyruvate-2-C14; both glycine-2-C14 and formiminoglycine-2-C14 yield acetate-1-C14; and formimino- glycine-C14 labeled in the formimino group is converted to acetate-2-C14. All of these results and others which will not be mentioned are consistent with the above scheme. We have seen that the formimino carbon of formimino- glycine can undergo two types of reactions. (1) It can be converted to formate via formyl-THFA with the simulta- neous formation of ATP, or (2) it can be reduced and transferred, presumably via the THFA derivatives, to glycine to give serine or an analogous compound. The relative importance of these two processes obviously de- FERMENTATIONS OF NITROGENOUS COMPOUNDS 87 pends upon the relative activities of the enzyme systems. With CI. cylindrosporum, glycine and formate accumulate in substantial amounts and therefore the enzyme system leading to ATP is probably more active. With CI. acidi- urici, on the contrary, nearly all of the glycine moiety of formiminoglycine is converted to acetate and carbon diox- ide, indicating a very active formyl transferring system. The studies summarized above have established the general pathway of the clostridial purine fermentation, al- though obviously several aspects of the process, particularly the incorporation of carbon dioxide into acetate and gly- cine71 and the assumed formation of serine, need to be investigated more fully. The ability to ferment purines is not restricted to Clos- tridia. Two anaerobic micrococci, M. aerogenes™ and M. lactilyticus,72 are able to grow in a complex medium at the expense of certain purines and carry out a modified propionic acid-type fermentation. These purine fermenta- tions have not been studied extensively but the experiments of Whiteley show that M. aerogenes forms considerable amounts of uracil and thymine during the decomposition of xanthine. This indicates that the initial enzymatic attack may be on the imidazole ring in contrast to what has been found with the purine fermenting Clostridia. These studies of the fermentations of nitrogen com- pounds have led to the discovery of several metabolic path- ways not previously encountered in other organisms. There is no reason to believe that these pathways are restricted to anaerobic bacteria. Indeed the whole history of biochem- istry indicates that any reaction discovered in one organism is ultimately found to occur in many others. Anaerobic bacteria obviously provide excellent and often neglected biological material for the study of certain types of meta- 88 BACTERIAL FERMENTATIONS bolic pathways. Consequently it can be anticipated that the further investigation of such bacteria will contribute significantly to a general understanding of the elaborate chemical machinery of living organisms. REFERENCES 1. Stickland, L. H., Biochem. J., 28, 1746 (1934) ; 29, 889 (1935) . 2. Nisman, B., Bacteriol. Revs., 18, 16 (1954). 3. Naviasky, P., Arch. Hyg., 66, 209 (1908) . 4. Brasch, W., Biochem. Z., 18, 380 (1909) . 5. Liebert, F., Proc. Koninkl. Akad. Wetenschap., Amsterdam, p. 990, April 23, 1909. 6. van Niel, C. B., unpublished observations, 1934. 7. Barker, H. A., Arch. MikrobioL, 10, 376 (1939) . 8. Wachsman, J. T., and H. A. Barker, /. Bacteriol, 69, 83 (1955) . 9. Woods, D. D., and C. E. Clifton, Biochem. ]., 31, 1774 (1937) . 10. Cardon, B. P., and H. A. Barker, /. Bacteriol, 52, 629 (1946) . 11. Barker, H. A., and J. V. Beck, /. Bacteriol, 43, 291 (1942) . 12. Barker, H. A., /. Bacteriol, 46, 251 (1943) . 13. Wachsman, J. T., and H. A. Barker, /. Bacteriol, 68, 400 (1954) . 14. Lieberman, I., and A. Kornberg, Biochem. et Biophys. Acta, 12, 223 (1953) . 15. Lieberman, I., and A. Romberg, /. Biol Chem., 207, 911 (1954) . 16. Clifton, C. E., /. Bacteriol, 44, 179 (1942) . 17. Pickett, M. J., /. Biol Chem., 151, 203 (1943) . 18. Cohen, G. N., B. Nisman, and G. Cohen-Bazire, Bull soc. chim. biol, 30, 109 (1948). 19. Whiteley, H. R., /. Bacteriol, 63, 163 (1952) . 20. Jackins, H. C, and H. A. Barker, /. Bacteriol, 61, 101 (1951) . 21. Douglas, H. C, /. Bacteriol, 62, 517 (1951). 22. Woods, D. D., and A. R. Trim, Biochem. J., 36, 501 (1942) . 23. Woods, D. D., Biochem. J., 30, 1934 (1936) . 24. Clifton, C. E., /. Bacteriol, 39, 485 (1940) . 25. Chargaff, E., and D. B. Sprinson, J. Biol. Chem., 151, 273 (1943) . 26. Wood, W. A., and I. C. Gunsalus, /. Biol Chem., 181, 171 (1949) . 27. Tarr, H. L. A., Biochem. J., 27, 759 (1933) . 28. Fromageot, C, "Desulfhydrases," The Enzymes, Chemistry and Mechanism of Action, Vol. 1, edited by J. B. Sumner and K. FERMENTATIONS OF NITROGENOUS COMPOUNDS 89 Myrback, Academic Press, New York, 1951, Part 2, Chapter 41, p. 1237. 29. Kallio, R. E., and J. R. Porter, /. BacterioL, 60, 607 (1950) . 30. Lewis, D., and S. R. Elsden, Biochem. J., 60, 683 (1955) . 31. Stadtman, T. C, "On the Anaerobic Degradation of Lysine," Amino Acid Metabolism, edited by W. D. McElroy and B. Glass, The Johns Hopkins Press, Baltimore, 1955, p. 493. 32. Tabor, H., "Degradation of Histidine," Amino Acid Metabolism, edited by W. D. McElroy and B. Glass, The Johns Hopkins Press, Baltimore, 1955, p. 373. 33. Waelsch, H., and A. Miller, "Certain Aspects of the Enzymatic Breakdown of Histidine," Amino Acid Metabolism, edited by W. D. McElroy and B. Glass, The Johns Hopkins Press, Balti- more, 1955, p. 407. 34. Wickremasinghe, R. L., and B. A. Fry, Biochem. J., 58, 268 (1954) . 35. Wachsman, J. T., Histidine and Glutamic Acid Metabolism of Clostridium tetanomorphum, doctoral dissertation, University of California, 1955. 36. Miller, A., Federation Proc, 15, 316 (1956) . 37. Rabinowitz, J. C, and W. E. Pricer, /. Am. Chem. Soc, 78, 4176 (1956) . 38. Magasanik, B., and H. R. Bowser, "Pathways of L-Histidine Degra- dation in Microorganisms," Amino Acid Metabolism, edited by W. D. McElroy and B. Glass, The Johns Hopkins Press, Balti- more, 1955, p. 398. 39. Barker, H. A., Enzymologia, 2, 175 (1937) . 40. Wachsman, J. T., and H. A. Barker, /. Biol. Chem., 217, 695 (1955) . 41. Wachsman, J. T., /. Biol. Chem., 223, 19, 1956. 42. Barker, H. A., and R. M. Wilson, unpublished observations. 43. Munch-Petersen, A., and H. A. Barker, unpublished observations. 44. Flavin, M., P. J. Ortiz, and S. Ochoa, Nature, 176, 823 (1955) . 45. Cardon, B. P., and H. A. Barker, Arch. Biochem., 12, 165 (1947) ; /. Biol. Chem., 173, 803 (1948) . 46. Whiteley, H. R., Proc. Nat. Acad. Sci. U. S., 39, 772, 779 (1953) . 47. Johns, A. T., /. Gen. Microbiol., 6, 123 (1952) . 48. Leaver, F. W., H. G. Wood, and R. Stjernholm, /. BacterioL, 70, 521 (1955). 49. Stadtman, E. R., Bull. soc. chim. biol., 37, 931 (1955) . 50. Barker, H. A., and T. Wiken, Arch. Biochem., 17, 149 (1948) . 51. Yates, R. A., and A. B. Pardee, /. Biol. Chem., 221, 743 (1956). 52. Barker, H. A., and J. V. Beck, /. Biol. Chem., 141, 3 (1941) . BACTERIAL FERMENTATIONS 53. Rabinowitz, J. C, and H. A. Barker, J. Biol. Chem., 218, 147 (1956) . 54. Barker, H. A., S. Ruben, and J. V. Beck, Proc. Nat. Acad. Sci. U. S., 26, 477 (1940) . 55. Sonne, J. C, J. M. Buchanan, and A. M. Delluva, /. Biol. Chem., 173, 69, 81 (1948) . 56. Karlsson, J. L., and H. A. Barker, /. Biol. Chem., 177, 597 (1949) . 57. Shemin, D., and D. Rittenberg, /. Biol. Chem., 167, 875 (1947) . 58. Karlsson, J. L., and H. A. Barker, /. Biol. Chem., 178, 891 (1949) . 59. Beck, J. V., Bacteriol. Proc, 1950, 136. 60. Rakosky, J., and J. V. Beck, /. Bacteriol, 69, 563, 566 (1955) . 61. Radin, N. S., and H. A. Barker, Proc. Nat. Acad. Sci. U. S., 39, 1196 (1953) . 62. Bradshaw, W. H., and J. V. Beck, Bacteriol. Proc, 1953, 86. 63. Rabinowitz, J. C, and W. E. Pricer, /. Biol. Chem., 218, 189 (1956) . 64. Rabinowitz, J. C, and W. E. Pricer, /. Biol. Chem., 218, 175 (1956) . 65. Rabinowitz, J. C, and W. E. Pricer, /. Biol. Chem., in press, 1957. 66. Rabinowitz, J. C, and W. E. Pricer, /. Am. Chem. Soc, 78, 1513 (1956) . 67. Sagers, R. D., J. V. Beck, W. Gruber, and I. C. Gunsalus, /. Am. Chem. Soc, 78, 694 (1956) . 68. Rabinowitz, J. C, and W. E. Pricer, /. Am. Chem. Soc, 78, 4176 (1956) . 69. Sagers, R. D., and J. V. Beck, /. Bacteriol., 72, 199 (1956) . 70. Rabinowitz, J. C, and W. E. Pricer, Federation Proc, 15, 332 (1956) . 71. Barker, H. A., and S. R. Elsden, /. Biol. Chem., 167, 619 (1947) . 72. Whiteley, H. R., and H. C. Douglas, /. Bacteriol., 61, 605 (1951) . INDEX Acetaldehyde dehydrogenase, 40, 46, 52 Acetaldehyde in butyrate forma- tion, 29, 30, 40, 46 Acetaldol, 29, 30 Acetate, formation from glycine and formate, 85-87 role in butyrate formation, 30- 33,37,46,51 role in lactate and glycerol fer- mentations, 32-34 Acetoacetate, conversion to ace- tone, 50, 51 phosphoroclastic decomposition of, 37, 38 role in butyrate formation, 29, 30, 37, 43 synthesis of, 44, 45, 50, 51, 53 Acetoacetate decarboxylase, 50, 53 Acetoacetyl coenzyme A, 41, 43, 45, 46, 47, 48, 50, 51, 53 Acetoacetyl thiolase, 45, 46, 47 Acetokinase, 41, 48 Acetone, formation of, 50, 51, 52, 53 Acetyl coenzyme A, 39, 40, 41, 42, 46,48 Acetyl phosphate, conversion to butyrate, 36, 37 formation from acetaldehyde, 36 formation from pyruvate, 35 reaction with ADP, 41 Acrylate, decomposition by CI. propionicum, 73, 74 Acrylyl coenzyme A, 74 Alanine, fermentation of, 60, 61, 62,73,74 oxidation of, 58 ^-Alanine, formation of, 74 Alcohol dehydrogenase, 40, 46, 52 Alcohols, decomposition by meth- ane bacteria, 9-11, 18 91 92 INDEX Allantoic acid, 75 Allantoin, fermentation of, 61, 62, 75,76 4-Aminoimidazole, 83-84 4-Amino-5-imidazolecarboxylic acid, 83 5-Arainovalerate, 59 5,10-Anhydroformyltetrahydro- folic acid, 85 Arginine, fermentation of, 59, 60 Arsenolysis of acetyl phosphate, 39, 40 Asparagine, fermentation of, 59 Aspartate, fermentation of, 60, 63 formation from orotic acid, 76 ATP synthesis, by CI. kluyveri, 48, 49 in formiminoglycine metabolism, 84-85, 86, 87 Bacillus acidi-urici, see Clostrid- ium acidi-urici Bacillus me thani genes, 5 Bacillus proteus vulgaris, 59 Bacillus putrificus, 60 Butyraldehyde, 52 Butyrate, oxidation of, 36, 43, 46- 49 synthesis of, 36, 46-49 Butyribacterium rettgeri, 34 Butyric acid-butanol fermenta- tions, chemistry of, in 1931, 29, 30 Butyryl coenzyme A, 43, 45, 46, 49, 50, 51 Butyryl coenzyme A dehydrogen- ase, 46, 47, 48 Butyryl phosphate, 35, 42, 43, 52 Caproate formation, 32, 49, 50 Carbon dioxide, conversion to ace- tate, 34, 81 reduction theory of methane formation, 16-21 utilization by methane bacteria, 13 Carbon monoxide decomposition, 9-11,23,24 Cellulose fermentation, 3-5, 9 Citramalate, role in glutamate fer- mentation, 69, 70, 71 Clostridium acetobutylicum, 28, 31, 49, 50, 51, 52, 53 Clostridium acidi-urici, 61, 62, 77- 87 Clostridium botulinum, 59, 64, 71 Clostridium butylicum, 35, 36, 42 Clostridium butyricum, 41, 59 Clostridium cochlearium, 61 Clostridium cylindrosporum, 61, 62, 77, 78, 84 Clostridium histolyticum, 59 Clostridium kluyveri, 31, 32, 33, 35, 36, 37, 38, 39, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52 Clostridium lactoacetophilum, 33 Clostridium perfringens, 8, 9 Clostridium propionicum, 61, 62, 73-75 Clostridium saccharobutyricum, 64, 71 Clostridium sporogenes, 58, 64, 72 Clostridium sticklandii, 49, 64 Clostridium tetani, 64, 71 Clostridium tetanomorphum, 61, 63, 64-71 Clostridium tyrobutyricum, 31, 32, 33 Clostridium welchii, 64 INDEX 93 Coenzyme A, 39, 40, 41, 42, 46, 47, 48,51 Coenzyme A transphorase, 42, 43, 46,51 Crotonase, 46, 47 Crotonate, 43 Crotonyl coenzyme A, 44, 46, 47, 48, 49 Cysteine, 63 Dihydroorotic acid, 76 6,8-Dihydroxypurine, 77 Diplococcus glycinophilus, 61, 63, 72 Energy metabolism of CI. kluyveri, 48, 49, 50, 51 Enrichment cultures, for bacteria fermenting nitrogenous com- pounds, 60, 61, 62, 63 for methane bacteria, 6 Escherichia coli, 41, 64 Ethanol, use by CI. kluyveri, 31- 32, 36, 40, 49 Fatty acids, accumulation in bu- tanol fermentation, 29-30, 50, 51 decomposition by methane bac- teria, 9-12 Formamide, 65, 66 Formamidinoglutarate (see for- miminoglutamate) Formate, formation of, 65, 78, 80, 81 role in H2 formation, 29, 30, 35 Formiminoglutamate, 65, 66 Formiminoglycine, role in purine fermentation, 83-87 5-Formiminotetrahydrofolic acid, 85 Formyl glutamate, 65, 66 10-Formyltetrahydrofolic acid, 85 Fusobacterium nucleatum, 64, 71 Glutamate, fermentation by cell- free extracts, 68-71 fermentation of, 60, 61, 64, 67- 71 formation of, 65, 66 Glycine, fermentation of, 61, 72 reduction of, 58 role in purine fermentation, 78, 79, 80, 81 Guanine, 77, 81 Heptanoate synthesis, 49 Histidine, decomposition of, 61, 64, 65, 66 £-Hydroxybutrate, 37, 43 0-Hydroxybutyryl coenzyme A, 46, 47, 48, 53 /3-Hydroxybutyryl coenzyme A de- hydrogenase, 46, 47 Hypoxanthine, oxidation to xan- thine, 81 Imidazoles, role in xanthine de- composition, 82-84 /3-Ketocaproate, 49 /3-Ketocaproyl coenzyme A, 49 a-Ketoglutarate, formation from glutamate, 69 Lactate fermentation by butyric acid bacteria, 32, 33, 34 Leucine, fermentation of, 63 Lysine, fermentation of, 60, 64 Mesaconate, role in glutamate fer- mentation, 69, 70, 71 94 INDEX Methane, microbial origin of, 2, 3 Methane bacteria, enrichment and isolation, 6, 7 influence of reducing agents on, 8 morphology of, 14-15 nutritional requirements of, 12, 13 pU range of, 13, 14 relation to oxygen, 7, 8 specialization in energy metabo- lism, 8, 9 substrate specificity, 9-12 taxonomy of, 14-15 Methane fermentation, of acetate, 16, 17, 21, 22, 23 of butyrate, 18-19 of ethanol, 3, 16, 18 of hydrogen, 9-11, 17 of methanol, 21, 22, 23 of propionate, 19, 20 Methane formation, by mixed cul- tures, 16 early history, 1-5 from carbon dioxide, facts and theories, 23-26 from cellulose, 3 from plant materials, 2 in ruminants, 4 Methanobacillus omelianskii, 6, 7, 11, 14, 18,23 Methanobacteriaceae, 14, 15 Methanobacterium formicicum, 6, 8,11, 14,24 Methanobacterium propionicum, 12,14, 18,19 Methanobacterium sdhngenii, 7, 14 Methanobacterium suboxydans, 6, 7, 11, 12, 14, 18 Methanococcus mazii, 7, 12, 14 Methanococcus vannielii, 6, 13, 14 Methanosarcina barkerii, 6, 11, 12, 14,23 Methanosarcina methanica, 7, 14 Methyl glyoxal, 30 Methyl group transfer, 22 Micrococcus aerogenes, 64, 87 Micrococcus anaerobius, 64 Micrococcus lactilyticus, 87 Micrococcus variabilis, 64 Microzyma cretae, 3 Nitrogen fixation, by methane bacteria, 13 Orotic acid, fermentation of, 61, 63, 76, 77 Oxamic acid from allantoin, 75, 76 Phosphotransacetylase, 39, 46, 48, 52 Proline, conversion to 5-aminoval- erate, 59 Propionate, use by CI. kluyveri, 49 Propionibacterium, 73 Propionyl coenzyme A, 42, 52, 74 Proteus morganii, 64 Proteus vulgaris, 64 Pseudomonas fluorescens, 65, 66 Purine fermentation, by Clostridia, 77-87 by micrococci, 87 Pyrimidine biosynthesis, 63, 77 Pyruvate, formation from gluta- mate, 68, 69 role in butyrate formation, 30, 35, 41 Serine, fermentation of, 60, 63 INDEX 95 Serine, role in purine fermenta- tion, 86-87 Stickland reaction, 58, 59 Streptococcus allantoicus, 61, 62, 75, 76 Tetrahydrofolic acid, role in for- miminoglycine decomposition, 84-85 Thiolase, see Acetoacetyl thiolase Threonine, fermentation of, 63, 73-75 Thymine, 63, 87 Tricarboxylic acid cycle, 67, 68 Uracil, 63, 87 4-Ureido-5-imidazolecarboxylic acid, 83 Ureidosuccinic acid, 76 Uric acid, fermentation of, 60, 61, 62 reduction to xanthine, 81 synthesis in pigeons, 79-80 Urocanate, 65 Valerate synthesis, 49 Valine, 63 Vinylacetic acid, role in butyrate synthesis, 38, 43 Vinylacetyl coenzyme A, 43, 44, 48 Vinylacetyl isomerase, 44 Xanthine, fermentation of, 77 role in purine fermentation, SO- BS Zymobacterium oroticum, 61, 63 I Li** &\ /«»