mMoles per 100 mmoles glucose fermented Products
Escherichia coli160* 164
Photobac- Pseudo-terium monas phospho-
formi-reum166 cans16*
2,3-Butanediol 0.3« 0.26* c 0.004
Acetoin 0.059 0.190 — Trace —
Glycerol 1.42 0.32 — — —
Ethanol 49.8 50.5 77.0 80.7 64.0
Formic 2.43 86.0 121.0 95.5 105.0
Acetic 36.5 38.7 78.0 61.8 62.0
Lactic acid 79.5 70.0 20.0 68.2 43.0
Succinic acid 10.7 14.8 39.0 8.85 22.0
Carbon dioxide 88.0 1.75 — 73.5 —
Hydrogen 75.0 0.26 — 54.8 —
Carbon recovered, % 91.2 94.7 108.0 115.6 96.0
O/R balance 1.06 0.91 1.04 1.16 1.02
• pH 6.2.
h pH 7.8.
e Resting cells, grown aerobically.
istence of an Embden-Meyerhof scheme was the precise study by U t t e r and W e r k m a n1 6 1 of the "zymohexase," aldolase, and triosephosphate isomerase equilibria. Currently, however, it cannot be stated with certainty t h a t other pathways do not also contribute to product formation.
As in the propionic and clostridial fermentations, the complexity of prod-ucts and the diversions produced b y alteration of conditions reflect vari-ations in pyruvate metabolism. For example, in E. coli, (a) phosphorolytic cleavage of pyruvate to acetyl coenzyme A and formate; (6) reduction of acetyl phosphate to ethanol; (c) conversion of formate to hydrogen and carbon dioxide; and (d) the inability to form acetoin from p y r u v a t e are characteristic. Conditions have a great influence, however, particularly upon the hydrogenlyase reaction (c above). I n Aerobacter aerogenes t h e same pathways exist, b u t with diminished importance because the pro-duction of acetoin and 2,3-butanediol from pyruvate are of major impor-tance and compete effectively for the p y r u v a t e (see Section I I I , A, 6 and Fig.
4).
T h e cleavage and oxidation of p y r u v a t e occurs in two similar systems:
(a) in Clostridia wherein acetyl coenzyme A (or acetyl phosphate), hydro-gen, and carbon dioxide are produced and formate is not an intermediate,
8 6 W. A. WOOD
and (b) in the Enterobacteriaceae which form acetyl coenzyme A, or acetyl phosphate, and formate.1 2 4 I n these systems the acetyl coenzyme A gener
ated is the source of acetate, ethanol, butyrate, and acetone. Each of these systems requires coenzyme A and diphosphothiamin.1 5 T h e reversible na
ture of the clastic reaction has been demonstrated by showing an incorpo
ration of labeled formate and carbon dioxide into t h e carboxyl group of pyruvate even when there was a net decrease in p y r u v a t e .1 6 2"1 6 6 Under the same conditions, however, acetyl phosphate does not exchange.
T h e energy available in the thiol ester bond of acetyl coenzyme A enters the high energy phosphate pool by means of phosphotransacetylase and acetokinase. Thus,' an additional mole of A T P (7.7 kcal.) is generated for each mole of acetate formed.
Pakes and Jollyman1 6 7 demonstrated t h a t formate is converted to carbon dioxide and hydrogen by m a n y microorganisms. Subsequent studies, in
cluding those b y Quastel and W h e t h a m ,1 6 8 and b y Stephenson and Strick
l a n d ,1 6 9 revealed the association of formic dehydrogenase and hydrogenase with hydrogenlyase activity. Therefore, the idea was advanced t h a t hydro
genlyase activity was due to the combined action of formic dehydrogenase and hydrogenase as f o l l o w s :1 7 0 , 1 7 1
HCOOH + A formic dehydrogenase > H2A + CO2 H2A ——: • H2 + A
hydrogenase
HCOOH hydrogenlyase ^2
Hydrogenlyase is i n d u c i b l e ;1 5 6 , 1 7 2 its formation is prevented b y aerobio-s i aerobio-s ,1 2 4 , 1 5 6·1 7 3 by the growth in a synthetic m e d i u m ,1 7 4-1 7 5 by a high ρ Η ,1 5 3·1 5 4 and b y the addition of n i t r a t e .1 7 3 , 1 7 6 T h e function of formic dehydrogenase and hydrogenase as components of hydrogenlyase is not clearly understood, and is not generally accepted. Certain organisms such as B. dispar,177 anaerogenic strains of E. coli,17* ·1 7 9 Salmonella,166 Eberthella162 and Shigella162 do not cleave formate to CO2 and H2. I n several instances the organisms which lack the lyase contain ample formic dehydrogenase and hydrogen
a s e .1 5 6 , 1 7 7 Thus, the hypothesis t h a t these enzymes, functioning together, make up the lyase system is either false or requires modification to include a role of additional enzymes or f a c t o r s .1 5 6 , 1 7 9 - 1 8 1 Since growth in iron-de
ficient media prevents lyase induction,1 8 2 and α-α-bipyridyl inhibits lyase action,1 8 3 it appears t h a t a component of hydrogenlyase requires iron.
6 . 2 , 3 - B U T A N E D I O L FERMENTATIONS
Several groups of organisms produce butanediol in fermentations which are otherwise of the mixed acid type. T h e process m a y be considered a
com-2. FERMENTATION OF CARBOHYDRATES 87 posite of a series of subfermentations of glucose, two of which yield ethanol and lactate, and the remainder composed of one or more of the following butanediol fermentation types.1 8 4
"Diol hydrogen"
glucose -> 2,3-butanediol + 2 C 02 + H2 (1)
"Diol-formic acid17
glucose —> 2,3-butanediol + formate CO2 (2)
"Diol-glycerol"
3 glucose 2,3-butanediol + 2 glycerol + C 02 (3) F u r t h e r distinctions are based on the configuration of butanediol (D,L, or
meso) produced.
a. Aerobacter aerogenes and Related Species. This fermentation was first recognized by H a r d e n and Walpole1 8 5 and some of the details have been re-corded b y Walpole,1 8 6 Scheffer,1 5 2 W e r k m a n and a s s o c i a t e s ,1 9 2 - 1 9 4 and H a r -den and N o r r i s .1 8 7'1 8 8 Aerobacillus polymyxa1*9 Aeromonas (Pseudornonas) hydrophila190 Erwinia carotovora197 and several species of Aerobacter,191"195 Serratia19* and Bacillus198'201 display variants of the same fermentation.
D u e to the conversion of appreciable carbon to the neutral 2,3-butanediol (accompanied by increased C 02 production), fewer acids are produced and small amounts of acetoin usually accumulate. These changes relative to the E. coli fermentation provide the basis for the methyl red test for acid pro-duction, the Voges-Proskauer test for acetoin, and the gas ratio test which are employed to differentiate the Escherichia and Aerobacter genera.
Altermatt et al.202 has found t h a t glucose and allose fermentations by Aerobacter aerogenes yielded the same products (Table I X ) . F r o m hexose-1-C1 4, t h e acetic acid, ethanol, lactic acid, and 2,3-butanediol were methyl-labeled, whereas with hexose-2-C1 4, hydroxymethyl groups of 2,3-butane-diol, lactate, and ethanol, and the carboxyl group of acetate were the only groups labeled. These results are in line with the p a t t e r n expected from utilization of methyl-labeled and carbonyl-labeled pyruvate (derived from hexose-l-C1 4 and hexose-2-C1 4, respectively) exclusively via the E m b d e n -Meyerhof pathway.
T h e interconversion of products during the fermentation and the effect of added hydrogen acceptors have been recorded by Werkman and associ-a t e s .1 9 2 , 2 0 5-2 0 7 Under aerobic conditions, as would be anticipated, a greater proportion of acetoin accumulates.1 9 4
p H has a marked effect upon the 2,3-butanediol fermentation. Above p H 6.3, acetic and formic acids accumulates and the production of hydrogen, carbon dioxide, acetoin, and 2,3-butanediol is prevented. Below p H 6.3,
88 W . A. WOOD
« Including 26.8 mmoles of carbon in cells.
f Hydrogen not determined.
acetic acid is converted to acetoin and 2,3-butanediol, and hydrogen pro
duction is suppressed.2 0 8 These d a t a indicate t h a t a gas ratio ( H2: C 02) of 0.5, once considered to be characteristic of Aerohacter, is fortuitous. M o r e recent studies by Neish and L e d i n g h a m ,1 8 4-2 0 9 utilizing automatic p H con
trol, revealed a broad maximum in rate of fermentation between p H 7.6 and p H 6.5 to 6.0 for both A. aerogenes and Aerobacillus polymyxa. Serratia
TABLE IX BUTANEDIOL FERMENTATIONS
2. FERMENTATION OF CARBOHYDRATES 89 marcescens and B. subtilis exhibit a much narrower range for maximum rate of 2,3-butanediol production centered at p H 6.2. At p H 6.2 to 6.3, both fermentation rate and butanediol production are a t a maximum.
b. Serratia. Pederson and B r e e d1 9 6 and later Sigurdsson and W o o d2 1 0 showed t h a t Serratia resembled A. aerogenes in t h a t a mixture of organic acids, acetoin, 2,3-butanediol, and ethanol are produced. Because of the possibility t h a t organisms which produce butanediol, b u t not hydrogen, might be high producers of glycerol, Neish et al.m investigated four strains of S. marcescens which produce little hydrogen. However, 60 % of the glucose utilized under anaerobic and aerobic conditions (Table I X ) followed the
"diol-formate" fermentation (Eq. 1) and t h e following oxidation:
glucose + y2Oi -> 2,3-butanediol + H20 + 2 C 02 (4) T h e remainder of t h e glucose yielded L-lactate, ethanol, and a small a m o u n t
of glycerol. T h u s the lack of hydrogen production is due to the absence of hydrogenlyase rather t h a n to glycerol formation. A formic dehydrogenase is present, however, as demonstrated b y its oxidation in the aerobic fer
mentation.
I n a further search for a butanediol-glycerol fermentation a survey of virtually all of the known Serratia species was made b y Neish et aZ.2 0 3 T h e fermentation balances (Table I X ) permitted a division into three groups based upon differences in fermentation under anaerobic conditions. I n t h e first group S. marcescens, S. anolium, and S. indica follow equations 1 and 4 for 2,3-butanediol formation. A second group containing S. plymuthicum and unnamed strains of A. aerogenes carries out the same process except t h a t hydrogenlyase is not present. S. kielensis, however, does not produce butanediol, b u t resembles E. coli in t h a t is produces acetate, CO2, and H2. T h e fermentation of the soft rot organism Erwinia carotovora191 (Table I X ) resembles t h a t of Serratia marcescens.
c. Bacillus. B. subtilis,1*4'198·199 Β. mesentericus,212 Β. anlhracis,201 and B.
cereus201 ferment glucose to 2,3-butanediol, acetate, ethanol, glycerol, and carbon dioxide; in addition, traces of formate and succinate are formed (Table I X ) . T h e fermentation differs from t h e types displayed b y Serratia and Aerobacter in t h a t formate and hydrogen are essentially absent. Instead, the available hydrogen is utilized to form glycerol (Eq. 3). Under aerobic conditions glycerol formation is suppressed.2 1 8
I n B. subtilis (both " F o r d " and " M a r b u r g " strains), thiamine promotes t h e formation of carbon dioxide and other products of p y r u v a t e metabolism, whereas in its absence lactate becomes t h e major product. B. subtilis (Mar
burg), grown anaerobically on a complex medium, displays a homolactic fermentation; aerobically, acetate, acetoin, and carbon dioxide are formed.2 1 4 Cells grown in synthetic medium also have lost fermentative capacity, b u t do oxidize glucose.2 1 6
90 W. A. WOOD
T h e behavior of carbon dioxide, formate, and acetate during glucose fer
mentation by B. subtilis (Ford) has been studied by Neish.2 0 4 As with A.
indologenes, added acetate was readily metabolized with an increased yield of butanediol and ethanol and a marked decrease in glycerol production.
Carboxyl-labeled acetate was readily converted to labeled ethanol, b u t not to 2,3-butanediol. Thus, the stimulatory effect of acetate upon butanediol production was an indirect result of its ability to act as a hydrogen ac
ceptor, as shown for Leuconostoc mesenteroides.216 Under the same conditions added formate was relatively inert, being recovered as such. I n addition, C1 4-formate did not yield labeled fermentation products. C1 402, on the other hand, was incorporated into succinic acid, the lactate carboxyl group, and formate at a slow rate (3 % ) ; much higher formate incorporation was ob
served with S. marcescens (38%) and A. aerogenes ( 5 4 % ) .2 0 4
T h e effect of p H upon the fermentation resembles S. marcescens as de
scribed above. As in the case of A. aerogenes, Neish found t h a t t h e fermenta
tion of glucose-l-C1 4 yielded methyl-labeled 2,3-butanediol and lactate, and glycerol labeled in the primary alcohol groups; the carbon dioxide was not labeled. I n line with the reasoning already presented for A. aerogenes, this labeling pattern constitutes evidence for the operation of the E m b d e n -Meyerhof glycolytic system in B. subtilis.204
T h e relative amounts of the different fermentation types has been p u b lished by Ledingham and Neish1 8 4 (Table X ) . For the purpose of compari
son, all of the organisms were grown in the same medium containing calcium carbonate. Under other conditions a nearly pure fermentation type m a y be
T A B L E X
RELATIVE IMPORTANCE OF THE VARIOUS FERMENTATION REACTIONS IN SPECIES OF