FIG. 1 2 . Pentose fermentation—3-2 cleavage.
termediate, D-xylulose-5-phosphate, by means of a series of inducible isom-erases, kinases, and epimerases.3 3 6 · 3 3 7 D-XyluJose-5-phosphate is then cleaved by phosphoketolase to form acetyl phosphate and D-glyceraldehyde-3-phosphate.2 3 9 This process superficially resembles the phosphorolytic cleavage of p y r u v a t e in t h a t diphosphothiamine and inorganic phosphate are required a n d t h a t an acetyl compound is formed. Acetyl phosphate is then converted to acetate, presumably by transfer of the phosphoryl group t o adenosine diphosphate; glyceraldehyde-3-phosphate is metabolized to lactate via the lower portion of the Embden-Meyerhof p a t h w a y (Fig. 12).
W i t h the discovery of phosphoketolase in Lactobacillus pentosus by H e a t h et αί.,2 3 9 all of the steps in this p a t h w a y of pentose fermentation became known. F u r t h e r , t h e known reactions accounted for t h e labeling p a t t e r n s of the products obtained in t h e fermentation of 1-C1 4 pentoses.
I n the 3-2 cleavage route, one mole of adenosine triphosphate is required t o form the pentose phosphate and a t least two moles of adenosine triphos
p h a t e are formed in the conversion of glyceraldehyde-3-phosphate to pyru
v a t e . If a third A T P is formed in t h e conversion of acetyl phosphate t o ace
t a t e , t h e yield would be the same as obtained from fermentation of an equivalent amount of hexose, via Embden-Meyerhof p a t h w a y and one more t h a n fermentation of hexose b y H M P p a t h w a y (see Chapter 1). For pentose:
pentose + 2 iP + 2 ADP - * lactate + acetate + 2 ATP + 2 H20
114 W. A. WOOD
b. Pathway Involving Hexose Synthesis. I n contrast t o lactobacilli, pentose fermentation b y other organisms is more complex. T h e same products are formed, b u t the relative amounts are altered.3 3 8 ·3 3 9 I n some instances, how-ever, more t h a n one mole of lactate is produced per mole of pentose.3 4 0 - 3 4 1 T h u s it appears t h a t t h e hexose p a t h w a y is followed, a t least in t h e final stage. T h e postulate has been advanced t h a t pentose is first converted t o hexose, perhaps by addition of a single carbon atom, a n d t h a t t h e hexose is then fermented in t h e normal fashion. Particularly illustrative are t h e dis-cussions of Stanier a n d A d a m s1 9 0 concerning D-xylose fermentation b y Aeromonas hydrophila. Recent research has borne out t h e general idea of this hypothesis.
T h e first evidence for a pathway of pentose fermentation involving initial hexose synthesis was obtained by Horecker et aZ.,3 4 2 who showed t h a t D-ri-bose-5-phosphate utilization b y a liver preparation yielded heptulose a n d hexose phosphates. Further, when pentose phosphate-2,3-C1 4 was metabo-lized, 2,3,4-labeled hexose phosphate was formed. Direct evidence for a par-ticipation of a similar system in pentose fermentation was obtained by Neish and Simpson3 4 3 from t h e fermentation of D-arabinose-l-C1 4 and L-arabinose-l-C1 4 b y Aerohacter aerogenes. Later, similar experiments were conducted by Altermatt et aZ.3 4 4 with D-xylose 1-C1 4, 2-C1 4, 5-C1 4, and with ribose 1-C1 4. W i t h all of t h e pentoses t h e same products were obtained (Table X V I I ) . When different pentoses with label in t h e same position were compared, t h e labeling patterns in t h e products were identical. T h e methyl group of lactate, ethanol, acetate, a n d 2,3-butanediol contained 30 t o 4 0 % of t h e activity of carbon a t o m 1 of pentose-l-C1 4, whereas t h e carbon dioxide, formic acid, a n d carboxyl group of lactate contained 15 to 20 % of t h e original specific activity.
Similar labeling patterns for pentose-l-C1 4 fermentations were obtained by Gibbs and associates. Resting cells of aerobically grown E. coli K - 1 23 4 6 fermented D-xylose-l-C1 4 a n d L-arabinose-l-C1 4 t o lactate 1,3-C1 4, acetate 2-C1 4, and formate-C1 4. T h e methyl groups of lactate and acetate were simi-larly labeled and h a d 30 t o 50 % of t h e specific activity of t h e lactate car-boxyl or of formate. T h e fermentation of ribose-l-C1 4 b y yeast extracts yielded ethanol and carbon dioxide.3 4 6 T w e n t y t o 27 % of t h e specific activity of carbon 1 of ribose was present in t h e methyl groups a n d 15 % was present in carbon dioxide. Inhibition of product formation caused a n accumulation of hexose monophosphate. W h e n D-ribose-l-C1 4 was fermented b y Clos-tridium perfringens a n d D-xylose-l-C1 4 was fermented b y C. beijerincki a n d C. butylicum111 1.2 t o 1.3 moles of carbon dioxide were produced per mole of pentose fermented; t h e specific activity was about 2 0 % of t h a t of t h e first carbon of t h e pentose employed. I n addition, acetate a n d ethanol were similarly labeled in contrast t o earlier findings with glucose fermentations,1 1 6
2. FERMENTATION OF CARBOHYDRATES 115
and were methyl-labeled to t h e extent of 40 % of t h e initial specific activity of carbon 1 of pentose.
Pentose fermentation by propionibacteria also results in the formation of the same products as obtained with hexoses (Table X V I I ) , with the pro
portion of products varying in some cases whereas in others it remained the same as for hexoses.2 7 4 , 3 3 8 · 3 4 7 T h e fermentation of L-arabinose-l-C1 4 by Propionihacterium arabinosum studied b y Rappoport and B a r k e r8 4 8 did not yield clear-cut evidence for the existence of either p a t h w a y of pentose fer
mentation. T h e highest specific activity (37 % of C-l of pentose) was found in t h e acetate methyl group. However, the carboxyl group of acetate, all carbons of propionate, and t h e carbon dioxide were labeled as well. Al
though t h e high label in t h e methyl group was considered indicative of a 3-2 cleavage, as in the lactic acid bacteria, more recent evidence on t h e
116 W. A. WOOD
mechanism of pentose fermentation suggests t h a t the same result would also be obtained in the hexose synthesis p a t h w a y . T h e experiments of Leaver et al.m show t h a t carbon atoms 2 and 3 of lactate a n d p y r u v a t e enter carbon dioxide and both carbon atoms of acetate and propionate.
T h u s the labeling p a t t e r n of the pyruvate, indicating t h e p a t h w a y involved, cannot be ascertained by examining the fermentation products. I n addition, the labeling d a t a obtained from glucose-C1 4 fermentation2 8 7 (Section I I I , C, 1, a) shows t h a t if a hexose monophosphate cycle involving transaldolase and transketolase were involved, highly complicated labeling p a t t e r n s could arise which would not be interpretable in regard t o the p a t h w a y in
volved. I t appears likely as in the glucose fermentation by propionibacteria t h a t clear-cut answers cannot be obtained from fermentation of labeled substrates.
I n the hexose synthesis pathway, although the products formed vary greatly as a function of the organism involved, all of the products from pentose-l-C1 4 are labeled as if derived from pyruvate containing 40 % of the specific activity of carbon 1 of the pentose in the methyl group a n d 20 % of the specific activity of carbon 1 in the carboxyl group (Fig. 13). These specific activities are then carried into the products (Fig. 14). Neish and Simpson3 4 3 and Gibbs et aZ.3 4 5 independently postulated t h a t such a labeling pattern can arise, as shown in Fig. 5. Three moles of pentose phosphate-1-C1 4 are converted t o 5 moles of pyruvate. T h e three moles of unlabeled pyruvate arise from glyceraldehyde-3-phosphate and from the lower half of fructose-6-phosphate. The fourth mole of pyruvate, methyl-labeled, and t h e fifth mole, methyl- and carboxyl-labeled, arise from carbon atoms 1,2, and 3 of fructose-6-phosphate.
SPECIFIC ACTIVITY-100%
^12 r
TRANSKETOLASE9 — 9 9 I 1
I I I 9 I * ~*
R-5-P X u - 5 - P C
ERYTHR0SE-4-P FRUCT0SE-6-P
TRANSKETOLASE , • . TRANSALDOLASE
*CH3-C-^00H
ς SPECIFIC ACTIVITY
X G-3-P 4 0 % 2 0 %
9 ( 2 I n 5 ) (I In 5) SED0HEPTUL0SE-7-P
FIG. 13. Pyruvate labeling via hexose resynthesis (distribution of carbon).
2. FERMENTATION OF CARBOHYDRATES 117 I3-CHOH-CHOH-CH3