.. Brewing .. .. .. * . . . . . . . .. .. .. . . . . .. . . .. .. 137 FERMENTATION-THE LAST TEN YEARS AND THE NEXT TEN YEARS L. M. Miall, M.A., F.R.I.C. Fermentation Development Dept, Pfizer Ltd, Sandwich, Kent Yeast Hydrocarbon fermentations, 138 Acids Lactic acid, 140 Acetic acid, 141 Citric acid, 142 Amino acids, 144 135 .. 140 . . .. . . . , * . * . 147 148 150 153 158 159 . . . . .. . . . . . . . . . . .. .. . * . . . . Ribonucleotides . . . I ) Vitamins and provitamins Transformations . , Antibiotics Conclusion. . .. * . References . . . . * . .. . , .. . . . . . . . . .. . . . . .. .. . . .. . . . . . . Enzymes . . .. .. .. . . . . . . .. . . 157 . . . . . . . . .. . . In this review the word ‘fermentation’ is used in its modern sense; it cannot be strictly defined, but is roughly equivalent to the production of compounds by micro-organisms.Usually the word is extended to cover production of the micro-organisms themselves. In attempting to forecast likely developments in the next decade it is obviously wise to consider what has happened in the last 10 years or so and to try to make an intelligent extrapolation. In such an article it is possible to give only a very brief account of this vast and specialized technology; for those who want to go deeper, the references given are mostly either to more detailed reviews of specific aspects of the subject or to very recent papers not covered by reviews. BREWING The production of beer and other alcoholic beverages is by far the biggest branch of the fermentation industry; the subject is so vast that it would need a review to itself, and this discussion will be confined to one of the most likely future developments.Until recently, brewing has been a batch process, but much work on continuous brewing has been carried out in the last decade following the original work at the Brewing Industry Research Foundation.1,2 Only the fermentation step will be considered here. The conventional batch fermentation begins with the cultivation, on a laboratory scale, of a small quantity of the appropriate micro-organism. 135 Miall . . . * . . .. This is added to a larger vessel containing the appropriate medium in which the micro-organism reproduces itself and, after a suitable time, this is itself added to a still larger vessel, which may or may not be the final fermenter. At the end of the production run the process begins again.In continuous culture, having once established a process on the largest required scale, fresh medium is added continuously, the final brew is drawn off continuously, and the process is kept as near to equilibrium as possible-the micro- organisms reproduce at a rate determined by the addition of nutrients. Brewing seems an obvious process to be operated continuously, but it has required many years of development in many breweries before large-scale introduction. It was first operated on a production scale in New Zealand, and recently Watney Mann Ltd, one of the biggest of the British brewing combines, has announced a continuous fermentation capacity of 20 000 barrels a week.A barrel is 36 gallons or 163.3 litres, so that one way of bringing this figure home is to realize that roughly five and three-quarter million pints of continuously fermented beer are being consumed per week in this country. It is a fairly safe assumption that this figure will increase considerably in the next 10 years. When fermentation technologists meet at symposia and similar occasions, the talk is apt to turn to problems of continuous culture, with the academic microbiologists often accusing their industrial counterparts of dragging their feet in the introduction of such an obvious (from the engineering point of view) technique.The majority of fermentations follow the type of curve shown in Fig. I where production of product is plotted against time. The total fermenter residence time (t), plus the down-time (d) when the fermenter is being emptied, cleaned, filled and sterilized, should be compared with the time (x) during which production is at maximum rate. But continuous culture a high proportion of ( t + d ) so that there is a relatively small potential saving does not always give such an obvious saving. In a slow fermentation, x may be in continuous culture. With a rapid fermentation and a concentrated medium, the total occupation time of the fermenter may be small, and if the fixed costs of the process are small compared with material costs, the potential saving will again be small.Against the potential time-saving, other factors have to be taken into account. In many fermentations the cost of the substrate necessitates as complete a conversion to product as possible, so the fermentation is con- tinued to the end. This is also necessary when complete removal of the sub- strate facilitates subsequent recovery of the product. In continuous culture, this would be equivalent to operating at point B instead of point A (Fig. I ) and would immediately remove the advantage. The problem could be over- come by having a second vessel in line, with a longer retention time, but two vessels cost more than one! Similarly, the rate of reproduction of the micro- organisms may not be in phase with their desired activity, or a different medium may be required for growth.This would necessitate another fer- menter, this time before the main production one. The problems of contamination and of mutation in continuous culture are real, but have probably been exaggerated; the contaminant or mutant would have to have a rate of reproduction greater than the parent strain, or R.I.C. Reviews 136 Product YEAST Ti me Fig. I. Typical fermentation curve. it would be washed out. The problems with micro-organisms with mycelial growth are much greater than with bacteria and yeasts which reproduce by simple division or budding, though penicillin fermentations have been run continuously on a laboratory scale.3 Each fermentation has to be run in continuous culture long enough for the costs of continuous versus batch culture to be determined, and in most cases these favour batch culture.But beer, yeast from petroleum, baker’s yeast and vinegar have been made or are made on a large scale by continuous culture and undoubtedly the same will apply in the future to other substances. Some fermentations that might be particularly suitable to continuous culture, such as industrial alcohol production, have failed to compete with purely synthetic production. From brewing, it is natural to pass on to consider yeast production-not only the long established manufacture of baker’s yeast, but also the more recent and potentially more interesting production of food or fodder yeast from petroleum products, with all that this implies for helping to solve the problem of protein for the underfed millions.It is difficult to see any major Miall 137 developments in the production of baker’s yeast, other than the introduction, or more strictly re-introduction, of continuous culture. The Distillers Com- pany plant at Dovercourt used to operate a type of continuous culture procedure, using six fermenters in series,* but the plant only operated a 53-day week, reputedly because of weekend labour problems (a statement most firms in the fermentation industry find difficult to accept), and it was subsequently shut down and sold. It is easy for those outside the details of a process to be critical, but baker’s yeast production seems to be a process ideally suited for continuous culture and one day it should be so operated again.HjJdrocarbon fermentations The most exciting development in the last 10 years, the one with the greatest future potential, and the one that has revived interest in an industry that was to some extent becalmed after the great expansion in antibiotics production in the late 40s and ~ O S , has been the proof that many micro- organisms can grow and reproduce efficiently on hydrocarbons. Although papers on hydrocarbon fermentations had appeared spasmodically for many years,5 hydrocarbons were not seriously considered as substrates for micro- biological conversions until the work of Champagnat and colleagues of the Societk Franqaise des Pktroles, Bp.6~7 Essentially the BP workers developed two processes.One involves growth of specially selected strains of yeast on heavy gas oil. The yeast grows preferentially on normal unbranched straight- chain alkanes which have to be removed in refinery operations anyway, at the same time converting the carbon to its own biomass. This necessarily involves very large fermenters, since only about 7 per cent of the heavy alkanic crude oil substrate is unbranched and so metabolized by the yeast. But it has been stated that the upgrading of the gas oil alone makes the operation worth while. This process has been operated for over five years in a pilot plant making a dried product with 65-68 per cent protein (fish meal has 65 per cent) and rich in essential amino acids and vitamins.The other process involves the separation of the paraffins by a molecular- sieve technique and the growth of yeast on the Clo to C18 straight-chain fraction. Both processes are operated continuously in single fermenters and published information indicates that they can be run under non-aseptic conditions. That neither of these processes is economically much better than the other is shown by the fact that large plants are being erected to work both processes: at Lavera in France, a 16 700 tons a year plant to operate on gas oil, and at Grangemouth in Scotland, a 4000 tons a year plant to use normal alkanes. The former is expected to be in operation early in 1971, the latter, late in 1970. For once fermentation know-how is being passed to Japan.The Kyowa Hakko Co. has bought the rights of the process from BP and is to produce protein from normal alkanes at the rate of 1000-1500 tons a year. Since World War 11, the Japanese have put so much effort into fermentation technology that now one is almost surprised at new develop- ments coming from anywhere but Japan. Because of the difficulty of completely separating the yeast from petroleum residues and doubts about the long term toxicity of such residues, it has been 138 R.I. C. Reviews necessary to conduct very lengthy toxicological testing in rats and various other animals.8 The products of the plants being built are to be used as animal foodstuffs and only after considerably more experience with animals are they likely to be included in human foods.Yeasts will not grow on alkanes with fewer than six carbon atoms, but many bacteria will grow on lower alkanes, including methane. 9 The metabolic pathway by which methane is utilized by micro-organisms is an unusual one which has been worked out by Quayle and his colleagues.1° Briefly, methane is oxidized to formaldehyde which reacts with ribose-5-phosphate to form the ketohexose allulose-6-phosphate-this then isomerizes to fructose-6-phos- phate. Five out of every six molecules of fructose-6-phosphate rearrange through a rather complicated sequence to give six molecules of ribose-5- phosphate and the cycle is complete. The overall result is to get one molecule of fructose-6-phosphate from six molecules of methane. Shell Research Ltd is working actively in this field.11 They have done a lot of work on the conversion of gaseous hydrocarbons to bacterial cells, but little has been published.The Agricultural Division of Imperial Chemical Industries Ltd has been studying a process for making fodder from North Sea gas. It has been stated that, during 1970, the company will be in a position to decide on the manufacture of protein on a semi-technical scale either in the form of bacteria from natural gas or of yeast from higher molecular weight petroleum products. An expenditure of several million pounds is contemplated for a development effort running through the 1970s and maybe beyond. What is surprising, and potentially disturbing, is to find that while certain micro-organisms have been found to be well tolerated in animal foods, they cause digestive disturbances when fed at the same level to man.Only straight-chain aliphatic hydrocarbons can be used for making micro- bial protein. Aromatic hydrocarbons have more obvious specific uses as substrates for the preparation of compounds on their metabolic pathway. Thus cumk acid can be made microbiologically from p-cymenelz and salicylic acid from naphthalene. The latter process was developed some years ago and was said to be nearly competitive with production from sodium phenate. CH CH OH COOH CH3 cumic acid p-cymcne a - aCooH A recent paper13 describes how a three- to four-fold increase in salicylic acid yield was obtained using mutant strains of a Corynebacter.A strain of Pseudomonas aeruginosal4 has given a 94 per cent weight by weight conversion of naphthalene to salicyclic acid. This is obviously a process that might well 139 Miall be developed further if the economics are right. Many other transformations of aromatic hydrocarbons have been shown to occur.15 These conversions are on obvious pathways of the breakdown of the respec- tive hydrocarbons. What at one time would have been regarded as more surprising is the finding that a-ketoglutaric acid, a tricarboxylic cycle acid on the standard carbohydrate breakdown pathway, can be made in good yield from a mixture of normal alkanes by the yeast Candida lipolytica.16 Similarly, L-glutamic acid17 can be made from normal alkanes by strains of Corynebacter.In both these conversions the level of thiamine in the medium is critical. With the yeast it has been shown that there is an a-ketoglutarate decomposing system that is actuated by thiamine; presumably the same will apply to the Corynebacter. While complete proof of the pathway remains to be found, it is generally accepted that the alkane is broken down to acetic acid which enters the normal tricarboxylic acid cycle as acetyl coenzyme A. This effect of thiamine illustrates an important technique that the fermenta- tion technologist employs to get accumulation of an intermediate breakdown product-to inhibit the action of an enzyme by starving it of coenzyme. Biotin has a similar effect in the production of glutamic acid from carbohydrate sources, and in many other fermentations the levels of trace metals are highly important.Other methods of controlling enzymes are by direct poisoning and by the use of mutant strains of micro-organisms. Enzymes are often controlled by repressor mechanisms, and to influence such enzymes an alteration to the repressor mechanism may be necessary. The last decade has seen the gradual fading out of some of the traditional fermentation processes, at least in the more industrialized countries. Ethanol, except in relatively crude form for drinking purposes, is no longer made by fermentation in the UK or the US, though in countries like India, with no petroleum industry and with the virtual necessity to use indigenous raw materials, its microbiological manufacture survives.Much the same applies to the acetone-butanol fermentation, which from many aspects-political, technical, biological, chemical, engineering and economic-has the most interesting history of all.lB This still survives on a manufacturing scale in Japan, but in the UK and the US both acetone and butanol are now made only from petrochemical sources. ACIDS Lactic acid Lactic acid is now made both microbiologically and chemically and it remains to be seen whether the long established fermentation process can compete with a purely chemical process. The only lactic acid producer in this country, Messrs Bowman (now part of Croda International), uses a fermentation process involving the action of a LactobaciZlus on a starch hydrolysate.19 The process is run at about 50°C and does not involve aeration, so it has con- siderable advantages over many other microbiological processes, both in relative freedom from contamination and in simplicity of plant.Lactic acid, a liquid at room temperature, cannot be purified by conventional crystalliza- R. I.C. Reviews 140 tion processes. A number of purification procedures have been studied, but solvent extraction is favoured for large-scale operation. In the US, Monsanto Chemical Co. makes lactic acid by hydrolysis of lactonitrile, which is a by-product of the manufacture of acrylonitrile from acetylene, but not of the more favoured route from propylene. Lactonitrile can also be made from acetaldehyde and hydrogen cyanide and this process for making lactic acid is operated in Japan.At one time it looked as if the salvation of the fermentation process might lie in possible regulations requir- ing the use of the naturally occurring L(+)-lactic acid in foodstuffs. The microbiological process makes both isomers, but could be adapted without too much difficulty to make the L-isomer, whereas resolution of a racemate produced in the chemical process would be expensive. The introduction of such a regulation now seems unlikely. This anaerobic bacterial process which is relatively free from contamination might be thought to be particularly suitable for continuous operation and, indeed, the process has been run on a laboratory scale for as long as 64 days.If the process could be made to work satisfactorily on a large scale the mean fermenter residence time for converting a 9 per cent maize or barley meal hydrolysate would be reduced from the present five days of a batch process to two days. But no news has been divulged that this has been done. Acetic acid Vinegar is a fermentation product that is now made by a continuous process.20 Its production is a two-step process-first the production of a suitable alcohol solution and then its oxidation to acetic acid by strains of Acetobacter. In Britain, the vinegar made is malt vinegar, using a fairly conventional yeast fermentation of malted barley. Wine producing countries normally convert poor quality wines to vinegar; in general, any suitably available crude sugar solution can be used.In the past, the acetification step was carried out by pumping the alcoholic liquor through tubs packed with materials such as birch twigs or beechwood shavings, which acted as a support for the bacteria. It is now effected by a continuous submerged culture technique in which the feed is pumped into the vessel at a rate adjusted to balance the rate of aeration and to keep the Acetobacter population in a steady state. This step now goes at nearly theoretical efficiency, compared with about 65-70 per cent in the older process. Vinegar production in England exceeds 10 Mgallons a year of material containing about 5 per cent w/v of acetic acid. Acetic acid is manufactured commercially by the liquid phase oxidation of acetaldehyde or butane.In certain circumstances, and on a relatively small scale, a submerged fermentation process such as that described for vinegar can be operated on a dilute aqueous solution of ethanol with the required salts added. The acetic acid is recovered by extraction with ethyl acetate and separated by distillation. It is claimed that 905 kg of alcohol give 1000 kg of acetic acid, and that there are factories operating at twice this daily capacity in Turkey and in Spain. But there does not seem to be any future in the produc- tion of acetic acid by fermentation, except in these rather restricted circum- stances. MiaN 141 Citric acid The pure chemical made in the largest tonnage by a fermentation process is citric acid; total western European capacity is about 60000 tons a year.Much of it is still made by growing Aspergillus niger or related moulds on the surface of a molasses medium in shallow pans or trays. That this is still done by companies which make other products by more modern submerged culture processes shows that the surface culture process is not as obsolete as it sounds. Indeed, the Japanese use their ‘koji’ process, growing moulds on bran in surface culture, to make a number of products. Production of citric acid in submerged culture, first shown to be possible by Perquin21 in 1938, is gradually coming into greater use. John & E. Sturge Ltd is the latest company to announce the use of submerged culture, which it has been operating on a small scale for three years.It is probably also significant that the French company, Usines de Melle, and the Dutch com- pany, Noury van der Lande, have announced that they are integrating their activities and building a new plant. In the past, Usines de Melle has operated a submerged process, and Noury van der Lande, which is much the bigger producer, a surface culture process. Pfizer Inc., by far the largest manufac- turer of citric acid in the world, is building a new plant in Eire for its produc- tion. Most of the citric acid manufacturers keep their processes a close secret and relatively little has been published that throws light on the details of one of the most interesting and the most difficult of all the fermentation processes.Meyrath has published a short, but useful and critical review22 which in turn refers to other reviews and papers. He very rightly criticizes much of the earlier published work on the grounds that insufficient attention was paid to trace metal contamination. For example, what use is a comparison of sucrose and glucose as carbon sources, if their purities are not specified, when it was realized as far back as 1910 that the ‘addition of a trace of iron salt . . . resulted in re-utilization of the citric acid with accumulation of oxalic acid’ P 3 A more recent paper points out that the addition of only two parts per thousand million of manganese to beet molasses treated with ferrocyanide cuts the citric acid yield by 10 per ~ e n t .2 ~ The pathway from sugar to citric acid was largely worked out in the 1950s by Johnson and his co-workers at Wisconsin.25 The Embden-Meyerhof- l coo- CHg coo- J H z I HO-C-COO- co. j3-h I coo- CH3 l /to coo- \ (!OO\iH;3/ GO I S-COA H.I.C. Reviews 1 42 Parnas pathway is followed to give ~-glyceraldehyde-3-phosphate and dihydroxyacetone phosphate, which are interconvertible, and which, by a further series of reactions, give pyruvate. Pyruvate is either decarboxylated to acetate (in the form of acetyl coenzyme A) or adds on carbon dioxide to give oxalacetate; these two products react together to form citrate. The enzyme aconitase, responsible for the conversion of citrate to aconitate and isocitrate, is inhibited, largely by depriving it of the necessary metal coenzymes, so that the tricarboxylic cycle does not operate and citric acid accumulates.Most other acids on or related to the tricarboxylic cycle (Fig. 2) can be CoA I coo coo-. I HOC-COO-- I citrate coo- I co I CH2 I coo- oxalacetate malate co I -0oc epoxysuccinate -~ ~- Fig. 2. The tricarboxylic acid (Krebs') cycle. Miall coo- 5 coo - l 5H.3 itaconate /+-coo- CH. coo- CH? l I ' c-coo- cis-aconitate coo- H-C-COO- isocitrate coo- coo- l CH? coo- I I HC-COO- co I coo- oxalosuccinate glutamate 143 made by fermentation processes, but are mostly either not in demand or can be made more cheaply by chemical synthesis. Fumaric acid can be made in very high yield by strains of Rhizopus26 and, until quite recently, it was assumed that fumarate is made not by the full tricarboxylic cycle, but via the glyoxalate by-pass.The latter involves the splitting of isocitrate with isocitrate lyase to succinate and glyoxalate, and the condensation of gly- oxalate with acetyl coenzyme A to L-malate; fumarate could be formed from either succinate or malate. However, it has recently been shown that isocitrate lyase is strongly repressed under typical fermentation conditions, and alterna- tive evidence is presented for formation of fumarate from carbon dioxide and a three-carbon compound.27 On an industrial scale, fumaric acid is made from maleic anhydride by a purely chemical method.~-Malic acid is made by many micro-organisms, but the only demand is for the much cheaper racemic acid which can be made chemically. Itaconic acid is of interest because it is a compound of relatively simple structure, not optically active, which is made most efficiently by fermentation using Aspergillus terreus or Aspergillus itaconicus. Problems in its manufacture are very similar to those in the manufacture of citric acid.28 Itaconic acid and its derivatives have various special uses in the plastics and artificial fibres industries. Aspergilhs fumigatus and some other moulds make trans L- epoxysuccinic acid in good yield;29 and a-ketoglutarate can be made by many micro-organisms-its production from hydrocarbons has already been mentioned.Amino acids The route to glutamate involves the amination of a-ketoglutarate and therefore largely follows the tricarboxylic cycle pathway. Glutamic acid, used on a vast scale as a flavouring agent in the form of monosodium glutamate, is another of the major fermentation products. In Europe, there are now three plants in Italy with a total yearly capacity of about 11 000 tons and one in France, with a capacity of 5000 tons. Total world production and consumption of mono- sodium glutamate is believed to be over 100 000 tons a year-about two-thirds of this is produced and half consumed in Japan. It should be pointed out that figures quoted for total capacity are largely meaningless. The recovery process for monosodium glutamate can utilize fairly standard equipment and the fermentation process needs completely standard fermentation equip- ment, so many companies with fermentation plant could utilize as much or as little of their capacity as they like for the manufacture of a compound such as monosodium glutamate.Glutamic acid is made using Micrococcus glutamicus or other gram- positive, non-motile, non-spore forming bacteria. The processes using carbohydrate as the substrate fall into three types-those using glucose and a restricting level of biotin to control glutamic acid production, and those using molasses with the addition either of polyoxyethylene fatty acid or of penicillin to control glutamic acid production. These latter compounds are assumed to eliminate a permeability barrier in the cell membrane and so stop intracellular accumulation of glutamic acid ;30 thus they obviate the need for feedback control.Biotin is also assumed to act on the cell membrane. R.I.C. Reviews 144 One Japanese company, Sanraku-Ocean Co., has recently switched from molasses to acetic acid as its starting material. Acetic acid is an interesting starting material for microbiological processes. It is taken into the standard metabolic pathways via the glyoxalate and the tricarboxylic acid cycles as acetyl coenzyme A. In the IJK its price does not make it appear likely that it could compete with cheap carbohydrate sources such as molasses, but special circumstances such as isolation problems may affect a straight price compari- son, and relative costs may be very different in Japan.The manufacture of glutamic acid by fermentation of hydrocarbons has already been mentioned ; here the level of thiamine is one critical factor, and, again, the addition of penicillin enhances glutamate prod~ction.~~ 932 Monosodium glutamate can be made by a number of purely chemical processes. The most favoured is that used by the Ajinomoto Co. It starts with acrylonitrile which is subjected to an 0x0 reaction to give p-cyanopropion- aldehyde. This is treated with hydrocyanic acid and ammonia to give an aminonitrile, which is then hydrolysed to glutamic acid.33 1 I CH2:CH.CN -+ OHC.CH2.CH2.CN + NC.CH.CH2.CH2.CN -+ HOOC.CH.CHz.CH2.COOH N H2 N H2 Naturally, this gives the racemic acid, and patent~3~735 have been taken out for obtaining the required L-isomer by seeding with its crystals, a procedure that sounds most uncertain for large-scale operation to one who has not experienced it.An enzymic process for obtaining the required isomer sounds much more attractive. Enzymes from a strain of Pseudomonas can convert the L-glutamic acid present in a racemic mixture to L-Zpyrrolidone-5-carboxylic acid.36 CH2 CHz- CH2 CH2- I I CH .COOH CO CH.COOH HOOC NH:! / I ‘NH’ i At the same time, enzymes from certain Lactobacilli racemize the D-giutamic acid to the equilibrium mixture.37 Ultimately the DL-glUtamiC acid is all con- verted to L-pyrrolidone carboxylic acid, which can be converted readily back to L-glutamic acid. An alternative process involves the production of DL- pyrrolidone carboxylic acid chemically and its conversion, in 90 per cent yield, to L-glutamic acid by a strain of Pseudomonus alcaligenes.38 The synthetic process is apparently only competitive with the fermentation process when worked on a large scale and to full plant capacity. Recently there has been overcapacity and this has hit synthetic production more than production by fermentation.However, the pattern for the future production of many optically active compounds is likely to involve largely chemical syntheses with final microbiological involvement in obtaining the required isomer. Another example of such a microbiological step is the conversion of D-phenylalanine to the L-isomer with Pseudomonas Jluorescens, a conversion that is thought to go via phenylpyruvic a ~ i d .~ g Miall 145 The other amino acid that has been made on a large scale by fermentation processes is lysine-its biosynthetic pathway is shown below. COOH CHzNHz [CHZ]:~ CH. NHe I I I CH. NHz COOH CH.NH2 CHO COOPOzH COOH I I CHc CH2 - I CH. NHz CHI NHn CH. NH2 / I !OOH Aspartyl semialdehyde I --.) YHZ I COOH Aspartyl phosphate Diaminopimelic acid \ CHn I I COOH Aspartic acid - 1 [iHnl:, I COOH CH2OH - I CH. NH;, I I I COOH I I CH. NHs COOH Threonine Lysine CH:$ CHOH hionine Homoserine \Met Lysine was first produced by fermentation in a two-step process-the produc- tion of diaminopimelic acid by Escherichia coli, and its decarboxylation by Aerobacter aerogenes or another strain of Escherichia coli.Subsequently a homoserine auxotroph of Micrococcus glutamicus was developed which made lysine directly.*O (An auxotroph is a mutated strain of micro-organism which will not grow in the absence of a specific factor.) This is a typical example of feedback inhibition and has been explained by the theory that there are three distinct aspartyl semialdehyde dehydrogenases inhibited respectively by lysine, threonine and homoserine. This has been shown to be true for Escherichia coli, and if it also operates in Micrococcus glutamicus, if the production of homoserine and threonine is blocked, two out of three enzymes would operate without control and lysine made by these pathways would accumulate.A purely synthetic process to make lysine was developed by the Dutch State Mines Company, but their plant has recently been shut down. A com- bined chemical and microbiological process has been developed by workers at E. I. du Pont de Nemours & C O . ~ ~ This involves the manufacture of DL- diaminopimelic acid from 2-ethoxy-3,4-dihydropyran by a highly efficient four-step process which goes in 80-90 per cent yield. The racemic diamino- pimelic acid is selectively decarboxylated to L-lysine using an enzyme from Bacillus sphaericus (see next page). This looks like being the process for the future. If other amino acids are required on a large scale, microbiological or combined chemical and microbiological methods can no doubt be developed for their manufacture.Cereal proteins are generally deficient in lysine and, to a lesser extent, may also be deficient in threonine or tryptophan. Proteins R. I. C. Reviews 146 CHO CN I I --+ CHOH [CH2]:3 --.) [CHiJs CHO CHOH I 1 I CN I NH:! I I CH. COOH I + [CH& CH . COOH I N Hz 00. C2Hj NH-CO, NH I t CH-CO’ I [CHe]:i NH I CH-CO, I NH-CO’ made by microbiological processes tend to be deficient in sulphur-containing amino acids, but racemic methionine is as active, biologically, as the L- isomer and hence purely chemical methods of synthesis suffice. Processes for the preparation of both L-threonine and L-tryptophan by the action of various micro-organisms on the corresponding- bhydroxycarboxylic acid have been worked 0 ~ t 4 ~ and L-threonine can also be made by a direct fermen- tation using an Escherichia coli mutant.43 RIBONUCLEOTIDES From amino acids, we turn to nucleic acid derivatives.The chief reason for the current interest in producing ribonucleotides on a large scale is that guanosine-5-monophosphate (GMP), inosine-5-monophosphate (IMP) and xanthosine-5-monophosphate (XMP) have, in that order, a strong flavour- enhancing effect. This operates, particularly, as a monosodium glutamate sparing effect. Thus the addition of 5 per cent of a GMP/IMP mixture to monosodium glutamate reduces the total amount of glutamate to a fifth of that required to give the same flavour in the absence of nucleotide.Most of the work on purine nucleotide production has been done by Japanese workers but the subject has recently been reviewed by Demain.44 To date IMP has been made by direct extraction from fish and by the enzymic hydro- lysis of ribonucleic acid from yeast. Many micro-organisms break down ribonucleic acid, which may constitute up to a fifth of their dry weight, under a number of conditions of stress, such as heating, cooling and treating with detergents; but undoubtedly the processes for the future will involve direct synthesis of the nucleotides. The biosynthetic pathway to the nucleotides has been worked out and essentially involves step-by-step synthesis of the purine ring system with ribose present throughout.The first step is the synthesis of phosphoribosyl pyrophosphate from ribose-5-phosphate and adenosine triphosphate : the subsequent pathway is shown on the next page (several steps have been omitted). Miall 11 147 phosphoribosyi pyrophorphate H,N-CO H,N-(' ,N - CH + aminoimidazole ribotide \ t - N 11 + aminoirnidazole carboxamide ribotide (AICAR) From IMP paths lead to AMP and via XMP to GMP. Each nucleotide loses phosphate to give the corresponding nucleoside and loses ribose to give the purine. Production of nucleotides by micro-organisms involves the use of auxo- trophs with requirements for adenine, guanine or purines in general, so that feedback inhibition can be avoided. AICAR, which is said to have flavour en- hancing properties, can be made either by inhibiting the next step in the syn- thetic sequence, the transfer of a formyl group, by sulphonamides or by using an Escherichia coli mutant lacking the transformylase. More recently mutants have been obtained from Bacillus megaterium that accumulate AICAR in concentrations up to 11 gl-l.459*6 This can be converted to disodium 5- guanylate in five chemical steps.To date, the best process for nucleotide production uses an adenine requiring strain or auxotroph of Brevibacterium ammoniugenes which, in shaken flasks, gives up to 12.8 g 1-1 of lMP.47 / inorinic acid (IMP) 148 phorphoribosylamine OH OH glycinarnidc ribotide t\l -CH R.I.C. Reviews VITAMINS AND PROVITAMINS Nothing worthy of note has been published recently on the fairly long established procedures for makiw , riboflavine and vitamin Bl2 micro- biologically.The former was reviewed in 195948 and the latter in 1964.49 However, considerable work has been carried out on the preparation of /?-carotene using the mould Blakeslea t r i ~ p o r a . ~ ~ There are two particularly interesting aspects of this fermentation; one is the constitution of the medium; and the other the use of a mixture of two mating types of mould. Originally, it was found that addition of 0.1 per cent of /?-ionone was necessary for maximum yield. Later it was shown that this could be replaced by cheap citrus by-products (citrus oil, citrus pulp or citrus molasses) giving yields of carotene of the order of 1 g 1-l.Indications are that it is the presence of limonene that is responsible. Spent mycelium of Blakeslea trispora from a previous fermentation was found to be an even better source of carotenoid precursers. A number of other mould, yeast and bacterial cells have the same effect, as have solvent extracts of these. Other compounds, which have been shown to act as activators of /?-carotene production, include isonicotinyl hydrazine and various amides, imides and lactams. The process involving the use of citrus molasses was scaled up to a 10 1 operating volume, and in 1963 production costs for the crude dried solids made at a rate of 5000 tons per year were estimated to be $31.35 kg-1 of contained /?-carotene. It was soon realized that 10-15 times as much carotene was obtained by using mixed mating types of Blakeslea trispora, rather than unmated cultures.Mated cultures produce a number of compounds (the so-called ,&factor) which can stimulate carotene production in unmated cultures.51 The + strain is believed to produce the /?-factor and the - strain the extra /?-caro- tene. At one time it was believed that this hormonal action of soluble com- pounds would not give the same stimulation as mixed cultures, but this is not the case;52 stimulation with equivalent amounts of the hormones induces as much carotene production in the - strain as was obtained with mixed strains. The hormones have been shown to be trisporic acids.53 (-)Trisporic acid B, the most active component, has the formula: 0 CH3 H3C COOH CH, (-)Trisporic acid B What, if any, significance can be attached to the various terpenoid com- pounds that stimulate carotene production remains to be seen.The accepted biosynthetic pathway to the carotenoids involves the condensation of two molecules of geranylgeranyl pyrophosphate to give an open chain c40 compound, which then ring closes at each end, so none of these ring com- pounds are likely to be carotene precursors. The UK market for carotene is not large (in 1965 it was in the region of Miall 149 Limonene @Carotene E75 000, primarily for use in margarine), and much work would be needed to work out a process for vitamin A production that is competitive with synthetic production.Unfortunately, conversion of carotene to vitamin A is an inefficient process in many animal species;54 cows can suffer from vitamin A deficiency on a diet with plenty of carotene. Micro-organisms do not have any use for vitamin A either and hence they lack the enzyme required to make it from carotene. Enzymic conversion would have to be done with an enzyme from a mammalian source such as hog intestinal mucosa, an enzyme which has not as yet been extracted and shown to work in vitro. B-lonone TRANSFORMATIONS I CHOH + CO CH20H Transformations are usually understood to be relatively simple one or two step alterations in molecules, which are themselves fairly complicated, carried out by micro-organisms. Probably the simplest transformation that has been done on a commercial scale is the production of dihydroxyacetone from glycerol by Acetobacter suboxydans.CHzOH I I I CH~OH CHzOH The production of gluconic acid from glucose is another example. This can be effected by Aspergillus niger, Penicillium notatum, Acetobacter sub- oxydans, Pseudomonas strains and indeed any organism that makes glucose oxidase. Gluconic acid is made on a large scale for conversion to glucono-6- lactone, which is used as a raising agent and has various other uses in the food industry; and for making sodium gluconate, used as a sequesterant, and iron and calcium gluconates, used as sources of these elements in veteri- nary and human medicine. However, when thinking of microbiological transformations one naturally thinks of ster0ids.~5 Micro-organisms have been found that can hydroxylate either in the a- or p- configuration almost every position in the steroid molecule.The most important are hydroxylation in the lla position by R. I. C. Reviews 150 CH,OH CH,OH i Prednisolone Triamcinolone I co Rhizopus nigricans and in the 1 l p position by Curvularia Iunata, and intro- duction of a double bond in the 1 : 2 position by Corynebacterium simplex to give prednisolone and related compounds. Other steroid transformations of industrial importance involve hydroxyla- tion in the 16a position by Streptomyces roseochrornogenes (this is involved in the manufacture of triamcinolone), oxidation of a 3-hydroxyl group to a keto group, and splitting off the side chain from position 17. A fairly recent development is the demonstration that the spores of many moulds have the ability to effect such transformations.For example, the spores of Aspergillus ochraceus will effect 1 1 a-hydroxylations and those of Septomyxa afinis can be used for 1-dehydrogenations.56 The spores can be grown on bran or barley and stored at -20°C for a year or more without losing their activity. Reactions have been run on a 200 gallon scale under non-sterile conditions and in a simple medium, so that recovery of the highly expensive products is easier and more efficient. If spores could be physically fixed in, say, polyacrylic beads in the same way as enzymes are now being fixed, they could be packed in a column and so make the whole process very easy; but this remains to be demonstrated.Until now, microbiological transformations have been used in the manu- facture of the anti-inflammatory steroids, but not of those in much greater use as oral contraceptives. These have used oestrone as the starting product, and oestrone, made from diosgenin by a sequence of reactions ending with a pyrolysis step at over 5OO0C, is a highly expensive raw material. Oestr 3r.e 19-Hyd roxycholesterol-3-acetate Miall 151 /I-Sitosterol Sih and his co-workers have shown that 3-/I-acetoxy- 19-hydroxycholest-5- ene, which is conveniently prepared chemically from cholesterol acetate, can be converted to oestrone in good yield by a species of Nocardia.57 The same conversion can be done on 19-hydroxy-4-stigmasten-3-0ne,5~ which is easily obtained from p-sitosterol, a compound that occurs widely in plants. The Upjohn Co.has a stockpile of over 100 tons of P-sitosterol which has been separated as a waste product from soya bean sterols used in progesterone production. It is now appreciated that such microbiological transformations are not restricted to steroids. There are a number of simple conversions, particularly hydroxylations, that micro-organisms can do much more efficiently than chemists. Recent examples are the conversion of acetanilide to the 2-hydroxy derivative by the higher fungus Amanita muscaria and to the 4-hydroxy derivative by Streptomyces species,59 and the hydroxylations of N-(3-chloro- sclerotiorurn.The latter gives hycanthone, which is used as a schistosomacide. NH-CH2-CH2-N(C2H,J2 (C, H 4 2 N H-C H2-C H 2-N I CI I CI Q 4-methylpheny1)-N,N’-diethylenediamine6 O and of lucanthone61 by Aspergillus -9 CH,OH CH3 CH3 CH20H The conversion of phenylalanine to tyrosine comes under this heading,62 as does the conversion of tyrosine to ~-3,4-dihydroxyphenylalanine (L-dopa), which shows promise in the treatment of Parkinson’s disease.63 Both these conversions can be carried out by a number of micro-organisms. This last conversion is of interest in that the normal first step in the micro- biological decomposition of tyrosine is deamination, so the amino group must be chemically protected. This can be done by preparing an N-formyl or similar derivative.It is also necessary to add ascorbic acid to stop melanin 152 R.I.C. Reviews COOH COOH I I CHNH, CHNHl CHNH, I ~ I I OH Tyrosine 6APA I Phenylalanine ANTIBIOTICS H COOH HO 4 OH Dopa formation. This ability of micro-organisms to effect a specific step or steps in a sequence of reactions looks likely to be of very considerable use in the next decade, in the manufacture of pharmaceutically-active compounds. It was the discovery of penicillin and the proof of its remarkable antibacterial properties that triggered off the post-war expansion in the fermentation industry. After beer, it is possibly antibiotics that will immediately come to most chemists’ minds when thinking about fermentation products. But, in fact, compared to some of the fermentations already discussed, there is less intrinsic interest in the production of antibiotics-the interest lies more in their use and method of operation.Production is often improved more by hit and miss procedures such as alterations in medium and mutations without consideration of any specific auxotroph requirements. Metabolic pathways are seldom completely worked out. Antibiotics are a miscellaneous collection of molecules, mostly fairly complex, which are classed together only because they are made by micro-organisms and adversely effect other micro-organisms. The reason for their production is not even clear; it is unlikely that many of them are naturally produced at concentrations at which other micro-organisms are inhibited.Bu’Lock’s classification of secondary r n e t a b ~ l i t e s ~ ~ treats them essentially as storage products, made primarily to keep the micro-organisms’ enzyme systems in good working order. Recent work has been partly to find new antibiotics-there are still gaps in the armoury-and partly to improve existing ones by either chemical or biological modifications. The supreme examples are the semi-synthetic penicillins, all arising out of the discovery of the enzyme penicillin acylase, which is made by a variety of micro-organisms and which splits micro- biologically-produced penicillins to 6-aminopenicillanic acid (6APA). This is the essential step in the production of the newer penicillins, all of which are made chemically from 6APA.Of these, ampicillin, the first broad spec- trum penicillin, still has the greatest use. Benzyl penicillin R = C,H,CH,CONH- R = NHZ- R == D( - )C6HS. CH. CO. NH- Ampicillin I -COOH Miall 153 Closely related, structurally, to the penicillins are the cephalosporins. Again manufacture involves both microbiological and chemical steps. The mould Cephalosporium acremonium makes cephalosporin C. ( R = -NHCO(CH2)3CH 6-Methylpretetramid COOH NH; X = -CH20COCH3 )(=-cH,-N+ The metabolic pathway to the penicillins and cephalosporins has been elucidated by a number of workers.65 That to the tetracyclines has been partly worked out thanks to the very elegant work of McCormick and his col- leagues,66 and it is now known in detail how the substituted naphthacene, 6-methylpretetramid, is converted to tetracycline or to the 5-hydroxy or 7-chloro derivatives.Cephaloridine NH - x == - cH,oco--cH, 0 OH 0 and Cephalothin This is split chemically to 7-aminocephalosporanic acid (R = NH2, X = CH2QCOCH3) from which the cephalosporins used in medicine are made. These have the structures R=C)-CH,-c--NH-- OH OH OH OX 0 'coo- 0 R = Q- CH,-c-- *; C-NH, 6-Methylpretetramid is believed to be built up of acetate or malonate units, but evidence for the details of the synthetic route is lacking. As with the penicillins, new tetracyclines have been made by chemical modification of the molecule and not by new microbiological processes. Other than the penicillins, cephalosporins and tetracyclines there are about 154 H / ll II W:HNH2 OH ti 0 3 CH,, ,CH3 CH, OH OH N OH 0 Oxytetracycline R.1. C. Reviews 20 antibiotics in such regular use in Britain as to be included in the British Pharmacopeia or the British Pharmaceutical Codex. These fall into a number of groups : polypeptides or substituted polypeptides containing D-amino acids, mostly made by bacteria, such as bacitracin;67 the macrolides, com- pounds made by species of Streptomyces and characterized by having one or more unusual sugars such as desosamine in the molecule-an example is erythromycin ;6S the polyene antibiotics with a number of conjugated bonds in the molecule, such as n y ~ t a t i n ; ~ ~ and unusual glycosides, such as strepto- mycin.70 's' / NH, H,N-L-lieu-L-Cys-L-Leu-D-Glu-L-lleu-L-Lys-D-Orn-L-lleu-D-Phe-L-His-L-Asp-L-Asp HN I c=o ' 'cook4 155 OH OH H Streptomycin Useful antibiotics that do not fit happily under any of these headings are cycloserine,71 chlorarnphenic01,~2 novobiocin,73 griseofulvin74 and fusidic acid,75 which has a steroid structure.The majority of antibiotics are active against gram-positive bacteria. The Miall 11* Bacitracin A 0 CH, Erythromycin o+ CH, 1 N H 2 - C O O v CH’ OCH, Griseofulvin 61 tetracyclines, chloramphenicol and certain of the new penicillins, such as ampicillin, are broad spectrum antibiotics acting against both gram-positive and gram-negative bacteria. A few, mostly peptides, are used only against gram-negative bacteria.Certain antibiotics are in particular use in treating tuberculosis ; these include streptomycin, cycloserine and viomycin (a cyclic peptide whose formula has only recently been completely elucidated). Griseofulvin is specifically antifungal. One gap has been recently filled by carbenicillin, a new penicillin which is active against Pseudumunas infections, but there are still others. It can be predicted that the search for new and better antibiotics will continue, though the law of diminishing returns has already been in action for some years. The state has now been reached in which, once an antibiotic has been discovered, its point of attack can usually fairly soon be worked out and at least a partial explanation given for its action.This should lead ultimately to the position where a particular molecule can be designed to attack a particu- lar synthetic step in a particular micro-organism. It is probable that in the next few years the search may be concentrated on antiviral compounds, though it is very doubtful that compounds will be found which have the spectacular effects in the antibacterial field of penicillin or even the sul- phonamides. Antiviral activity has been claimed for a number of antibiotics, but this is usually because of their action on some step in protein synthesis, an action which normally makes them too toxic for medicinal use. Rifampicin, one of the rifam~cins,~~ has recently been shown to act against pox virus, vaccinia virus and trachoma virus, but it remains to be seen how effective it is clinically.Another potential use of antibiotics is as anti-tumour com- pounds; examples are actinomycin D,77 d a u n ~ m y c i n ~ ~ and streptonigrin.79 156 ?H “0 0 Novobiocin I H Cycloserine R. I . C. Reviews The recent ban on the use of antibiotics used in human medicine as animal growth promotants means that new growth promotants are urgently needed. All those used to date have antibacterial properties, so that once again it is natural to look for suitable compounds amongst those made by micro- organisms. Bacitracin, which is only used topically in human medicine, is already used as a growth promotant in a number of countries. Flavomycin, made by a species of Streptomyces and described as belonging to a new group of compounds, is the first antibiotic specifically prepared for this purpose.Yet another example of the miscellaneous uses of antibiotics is provided by nisin,@J which is used in cheese processing. This is a mixture of cyclic peptides with antibacterial activity produced by some strains of Strepto- coccus lactis, and contains the unusual amino acids lanthionine, /?-methyl- lanthionine and dehydroalanine. Thus there are many possibilities for the use of antibacterial substances made by micro-organisms other than in human medicine. ENZYMES It is widely predicted that the next 10 years or so will see a greatly increased industrial use of enzymes. A high-powered panel was recently considering the future of the British chemical industry by the ‘Delphi’ method and more than one of the panel believed that enzyme-based chemical processes would be among the major chemical innovations in the next 10-15 years.The Science Research Council is very actively supporting enzyme research through its Enzyme Chemistry and Technology Committee and has made a &200000 grant to University College London to support such work for five years, as well as grants to other institutions. It seems highly likely that all this interest will generate new uses for enzymes and that the enzymes concerned will largely be made by microbiological methods. Already fungal amylases, made by Aspergillus niger and species of Rhizopus, are widely used in breadmaking, in the production of soluble carbohydrate solutions from starch and in other applications.81 Various types of subtilisin, the alkaline protease of B. subtilis, are used in detergentsS82 Another proteo- lytic enzyme, keratinase, believed to be made by Streptomyces jradiae, is used in the leather industry and yet another type of protease has very similar properties to animal rennet and is made commercially by the moulds Endothia parasitica and Mucor pusillus. 83 Pectinase is made by various Aspergilli,g4 and glucose oxidase by many micro-organisms, in particular Penicillium notatum and Aspergillus niger. There are also many medicinal, analytical and other academic uses of microbial enzymes. Two of the more interesting recent developments are the use of L-asparaginase from Escherichia coIi,85 Erwinia carotovora86 and other micro-organisms in the treatment of certain forms of cancer; and the experiments involving the use of dextranase from Penicilliurn funiculosum for the prevention of dental caries.8 7 9 8 8 It is probable that a greatly increased production of enzymes will be required because of recent developments in the use of immobilized enzymes. Enzymes can be immobilized either chemically, by specifically binding them to insoluble supports such as various cellulose derivatives or silica, or physi- cally, by adsorption on to inert carriers or by entrapping them in Miail 157 polymers.89990 Such treatment means that enzymes can be packed in columns and used in the same way as one might use many inorganic catalysts, or that, if used in batch reactions, they can relatively easily be recovered and used again.Immobilization has the further advantage of greatly increasing the stability of the enzyme; it opens up many new possibilities for the use of enzymes on an industrial scale. As with antibiotics the development of processes for enzyme production is largely by hit and miss methods, such as screening for suitable micro- organisms, treatment of these to get better mutants, ad hoc variation of the medium. Some background knowledge is applied, but it does not ob- viously follow, for example, that to make high concentrations of protease it is necessary to have high concentrations of protein in the medium.CONCLUSION To review the whole subject of industrial fermentations would take a book. All that has been attempted here is to highlight some recent developments and to do a little crystal gazing. Many relevant topics have been omitted and some of these are now listed together with references. Amongst simple sub- stances the production of gluconic acid and kojic acid have not been men- tioned, nor has 2,3-butylene glycol.91 The microbiological step from sorbitol to sorbose is part of the standard process to make ascorbic acid: a possible alternative route to ascorbic acid incorporates two fermentation steps-the conversion of glucose to 5-ketogluconic acid with Acetobacter suboxydans, and of L-idonic acid to 2-keto-~-gulonic acid with Pseudomonas Jluorescens. Other steps are purely chemical.A process recently proposed for making xylitol (used as a sweetening agent) from glucose involves three fermentation steps-glucose to D-arabitol with Debaryomyces hansenii, D-arabitol to D-xylulose with Acetobacter suboxydans and D-xylulose to xylitol with Candida guillierrnondii.92 Other polyols, including glycerol, mannitol, arabitol and erythritol can all be made by fermentation processes;93 much work has been done on the production of microbial polysaccharide gums, as well as d e ~ t r a n . 9 ~ Production of the ergot alkaloids should be menti0ned.9~ So should that of the gibberellin plant hormones.96 Bacillus thuringiensis is cultured for its insecticidal use.97 A future possibility, though not a proba- bility, is the cultivation of nitrogen fixing micro-organisms for addition to the soil.About six years ago, J. J. H. Hastings said ‘. . . ever since 1 can remember the fermentation industry has been dying . . . it is the pharmacist who has halted the coffin of the fermentation industry’.98 But six years have altered the picture very considerably. The future of the fermentation industry now looks very healthy and it no longer relies on the pharmaceutical industry to keep it alive. There has recently been considerable expansion in the industry and one can predict with confidence a greater expansion in the next decade. The production of protein from hydrocarbons, wider uses of antibiotics, greatly increased uses of microbiologically-made enzymes and a far wider use of microbiological processes as steps in synthetic sequences, particularly for hydroxylating and for making optically active isomers; all these should keep fermentation technologists active and happy in the future. R.I.C.Reviews 158 REFERENCES 1 J. S. Hough and A. D. Rudin, J. Inst. Brew., 1958, 64,404, 2 J. S. Hough and R. W. Ricketts, J. Inst. Brew., 1960, 66, 302. 3 J. Pirt and D. S. Callow, J. appl. Bact., 1960, 23, 87. 4 A. J. C. Olsen, SCZ Monogr., 1961, 12,81. 5 A. C. Van der Linden and G. J. E. Thijsse, Adv. Enzymol., 1965, 27, 469. 6 A. Champagnat, Br. Pat. 914 567. 7 A. Champagnat, C. Vernet, B. Lain6 and J. FiIosa, Nature, Lond., 1963, 197, 13. 8 D. A. B. Llewelyn in Microbiology (ed.P. Hepple), 63. London: Institute of Petroleum, 1968, 9 R. Whittenbury, ProceJs Biochem., 1969, 4 (l), 51. 1968. 13 I. D. Hill and A. Gordon, Biotechnol. Bioengng, 1967, 9, 91. 10 M. B. Kemp and J. R. Quayle, Biochem. J., 1967,102,94. 11 D. W. Ribbons in Microbiology (ed. P. Hepple), 47. London: Institute of Petroleum, 12 K. Yamada, S. Horiguchi and J. Takahashi, Agric. biol. Chem., 1965, 29, 943. 14 T. Ishikura, H. Nishida, K. Tanno, N. Miyachi and A. Ozaki, Agric. biol. Chem., 1968, 32, 12. 15 R. L. Raymond, Process Biochem., 1969,4 (9), 71. 16 R. Tsugawa, T. Nakase, T. Kobayashi, K. Yamashita and S . Okumura, Agric. biol. Chem., 1969,33,929. 17 Y. Imada, J. Takahashi, K. Yamada, K. Uchida and K. Aida, Biotechnol. Bioengng, 1967, 9, 45.18 D. Ross, Prog. ind. Microbiol., 1961, 3, 71. 19 G. Machell, Ind. Chem., 1959, 35, 283. 20 J. White, Process Biochem., 1966, 1, 139. 21 L. H. C. Perquin, Bljdrage tot de kennis der oxydative dissimilafie van Aspergillus niger van Tieghem. Delft: Meinema, 1938. 22 J. Meyrath, Process Biochem., 1967, 2 (lo), 25. 23 C. Wehmer, Technical mycology (ed. F. Lafar), 350. London: Charles Griffin, 1910. 24 D. S. Clark, K. Ito and €4. Horitsu, Biotechnol. Bioengng, 1966, 8,465. 25 W. W. Cleland and M. J. Johnson, J. biol. Chem., 1954, 208,679. 26 R. A. Rhodes, A. A. Lagoda, T. J. Meisenheimer, M. L. Smith, R. F. Anderson and R. W. Jackson, Appl. Microbiol,, 1962, 10,9. 27 A. H. Romano, M. H. Bright and W. E. Scott, J. Bact., 1967, 93, 600.28 R. C. Nubel and E. J. Ratajak, Br. Pat. 950 570. 29 L. J. Wilkoff and W. R. Martin, J . biol. Chem., 1963, 238, 843. 30 M. Shibukawa, M. Kurima, S. Okabe and T. Ohsawa, Agric. biol. Chem., 1968,32,641. 31 J. Takahashi, K. Kobayashi, Y. Imada and K. Yamada, Appl. Microbiol., 1965, 13, 1. 32 I. Shiio and R. Uchio, J . gen. appl. Microbiol,, Tokyo, 1969, 15, 65. 33 C. W. Huffman and W. G. Skelly, Chem. Rev., 1963, 63,625. 34 Ajinomoto & Co. Inc., Br. Pat. 844 952. 35 Ajinomoto & Co. Inc., Br. Pat. 865 31 1. 36 Kyowa Hakko Co. Ltd, Br. Pat. 884 415. 37 Kyowa Hakko Co. Ltd, Br. Pat. 933 802. 38 Y. Kawai and T. Uemura, Agric. biol. Chem., 1966, 30,438. 39 I Chibata, T. Tosa and R. Sano, Appl. Microbiol., 1965, 13, 618. 41 B.S. Gorton, J. N. Coker, H. P. Browder and C. W. Defiebre, Znd. Engng Chem. Prod. 40 H. T. Huang, Prog. ind. Microbiol., 1964, 5, 70. 42 S. Takesue, T. Yokouchi and H. Wada, US Pat. 3 133 868. Res. Dev., 1963, 2, 308. 43 I. Shiio and S. Nakamori, Agric. biol. Chem., 1969, 33, 11 52. 44 A. L. Demain Prog. ind. Microbiol., 1968, 8, 35. 45 K. Kinoshita, T. Shiro, A. Yamazaki, 1. Kumashiro, T. Takenishi and T. Tsunoda, 46 I. Kumashiro, A. Yamazaki, T. Meguro, T. Takenishi and T. Tsunoda, Biotechnol. Biotechnol. Bioengng, 1967, 9, 329. Bioengng, 1968, 10, 303. 47 A. Furaya, S. Abe and S. Kinoshita, Appl. Microbiol., 1968, 16, 981. 48 T. W. Goodwin, Prog. ind. Microbiol., 1959, 1, 137. 49 L. Mervyn and E. L. Smith, Prog. ind. Microbiol., 1964, 5, 151.50 A. Ciegler, Adv. appl. Microbiol., 1967, 7 , 1. 53 L. Caglioti, G. CaineIli, B. Camerino, R. Mondelli, A. Prieto, A. Quilico. T. Salvatori 51 R. P. Sutter and M. E. Rafelson, J. Bact., 1968, 95, 426. 52 H. van den Ende, J. Bact., 1968,96, 1298. and A. Selva, Tetrahedron, 1966, Suppl. no. 7, 175. Mia11 I 59 54 W. M. Beeson, Fedn Proc. Fedn Am. SOCS exp. Biol., 1965, 24,924. 55 W, Charney and H. L. Herzog, A handbook of microbial transformations of steroids. New York, Academic, 1967. 56 C. VCzina, S. H. Sehgal and K. Singh, Adv. appl. Microbiol., 1968, 10, 221. 57 C. J, Sih, S. S. Lee, Y. Y . Tsong and F. N. Chang, J . Am. chem. SOC., 1965, 87, 2765. 58 C. J. Sih and K. C. Wang, J . Am. chem. SOC., 1965,87, 1387. 59 R. J. Theriault and T. H. Longfield, Appl. Microbiol., 1967, 15, 1431. 60 D. Rosi, T. R. Lewis, R. Lorenz, H. Freele, D. A. Berberian and S. Archer, J. med. Chem., 1967,10, 877. 61 D. Rosi, G. Peruzzotti, E. W. Dennis, D. A. Berberian, H. Freele, B. F. Tullar, and S. Archer, J. med. Chem., 1967, 10, 867. 62 P. Chandra and L. C . Vining, Can. J. Microbiol., 1968, 14, 573. 63 C. J. Sih, P. Foss, J. Rosazza and M. Lemberger, J. Am. chem. SOC., 1969, 91,6204. 64 J. D. Bu’Lock, Adv. Appl. Microbiol., 1961, 3,293. 65 E. P. Abraham, G. G. F. Newton and S. C . Warren in Biogenesis of antibiotic substances (ed. Z . Vanek and Z. Hostalek), 179. London: Academic, 1965. 66 J. R. D. McCormick in Biogenesis of antibiotic substances (ed. Z . Vanek and Z . Hosta- lek), 73. London: Academic, 1965. 67 R. J. Hickey, Prog. ind. Microbiol. 1964, 5, 93. 68 W. M. Stark and R. L. Smith, Prog. ind. Microbiol., 1961, 3, 211. 69 D. Perlman, Prog. ind. Microbiol., 1967, 6, 1. 70 D. J. D. Hockenhull, Prog. ind. Microbiol., 1960, 2, 131. 71 F. C. Neuhaus, in Antibiotics (ed. D. Gottlieb and P. D. Shawl vol. 1, 40. Berlin, Heidelberg and New York: Springer, 1967. 72 C. G. Smith and J. W. Hinman, Prog. ind. Microbiol., 1963, 4, 137. 73 H. Hoeksema and C. G. Smith, Prog. ind. Microbiol., 1961, 3, 91. 74 A. Rhodes, Prog. ind. Microbiol., 1963, 4, 165. 75 W. 0. Godtfredsen and S. Jahnsen, Process Biochem., 1969 4 C12), 11. 76 P. Sensi and J. E. Thiemann, Prog. ind. Microbiol., 1967, 6, 21. 77 E. Katz and H. Weissbach, Prog. ind. Microbiol., 1967, 6, 61. 78 A. Di Marco, Antibiotics, 1967, 1, 190. 79 K. V. Rao, K. Biemann and R. B. Woodward, J . Am. chem. SOC., 1963, 85, 2532. 80 N. J. Berridge, G. G. F. Newton and E. P. Abraham, Biochem. J., 1952, 52, 529. 81 G. T. Banks, F. Binns and R. L. Cutcliffe, Prog. ind. Microbiol., 1967, 6, 95. 82 M. H. J. Zuidweg, A. J. Vroemen and R. Beukers, Process. Biochem., 1969, 4 (8), 19. 83 J. L. Sardinas, Process. Biochem., 1969, 4 (7), 13. 84 L. Nyiri, Process Biochem., 1968, 3 (8), 27. 85 L. T. Mashburn and J. C . Wriston, Archs Biochem. Biophys., 1964, 105,450. 86 H. E. Wade, R. Elsworth, D. Herbert, J. Keppie and K. Sargeant Lancet, 1968, ii, 776. 87 H. M. Tsuchiya, A. Jeanes, H. M. Bricker and C. A. Wilham, J. Bact., 1952, 64, 513. 88 R. J. Fitzgerald, D. M. Spinell and T. H. Stoudt, Archs oral Biol., 1968, 13, 125. 89 1. H. Silman and E. Katchalski, A. Rev. Biochern., 1966, 35, 873. 90 G. Kay, Process. Biochem., 1968, 3 (8), 36. 91 S. K. Long and R. Patrick, Adv. appl. Microhiol.. 1963, 5, 135. 92 H. Onishi and T. Suzuki, Appl. Microbiol., 1969, 18, 1031. 93 J. F. T. Spencer, Prog. ind. Microhiol., 1968, 7, 1. 94 K. L. Smiley, Fd Technol., Lond., 1966, 20, 112. 95 M. Abe and S. Yamatodani, Prog. ind. Microbioi., 1964, 5, 203. 96 J. F. Grove, Q, Rev. chem. SOC., 1961, 15, 56. 97 J. R. Norris, J. appl. Bact., 1970, 33, 192. 98 J. J. H. Hastings, Chemy Znd., 1964, 147.5. R.I.C. Reviews 160