[Manuscript received August 12, 1953]

THE PATHWAY FOR GLUCOSE OXIDATION BY ASTEROCOCCUS MYCOIDES, THE ORGANISM OF BOVINE PLEUROPNEUMONIA By A. W. RODWELL 1* and E. SHffiLEY RODWELL 1* [Manuscript received August 12, 1953] Summary Cell suspensions of A. mycoides oxidize glucose but do not attack it anaerobically. The pathway for glucose oxidation was investigated. Evidence was obtained for the presence of hexokinase and aldolase, and suspensions treated by freezing and thawing oxidized glucose-6-phosphate, frucose-6-phosphate, and fructose-1,6-diphosphate, but not gluconate or 6- phosphogluconate. In contrast to intact cell suspensions, cell suspensions treated by freezing and thawing attacked glucose anaerobically at a slow rate. The rate was considerably increased when the reaction was linked with the reduction of acetaldehyde by yeast alcohol dehydrogenase. In the presence of arsenite, an inhibitor of the pyruvate dismutation reaction, intact cell suspensions catalysed a homolactic glucose fermentation anaerobically at a rate comparable with the rate of glucose oxidation. It was concluded that glucose is broken down by A. mycoides by the Embden-Meyerhof pathway. Oxygen is required for the re-oxidation of reduced diphosphopyridine nucleotide. J. INTRODUCTION In the bacteria examined which oxidize glucose but are unable to attack it anaerobically, it has been found that glucose oxidation did not involve the Embden-Meyerhof scheme, but occurred either without phosphorylation (Barron and Friedemann 1941; Koepsall 1950; Stokes and Campbell 1951), or by a modification of the Warburg-Dickens hexose monophosphate pathway (Entner and Doudoroff 1952). It was shown previously (Rodwell and Rodwell 1954a) that cell suspensions of a strain of A mycoides oxidized glucose to acetate and carbon dioxide, but were incapable of attacking it anaerobically. Fructose and mannose were also oxidized, suggesting the participation of hexokinase in glucose degradation. Pyruvate was oxidatively decarboxylated, and was attacked anaerobically l:y the dismutation reaction. Evidence was presented (Rodwell and Rodwell 1954b) that the di:,;mutation system in A mycoides is similar to that described in Streptococcus faecalis, i.e. a diphosphopyridine-linked, pyruvic-lactic dehydrogenase system requiring cocarboxylase, Mg+ +, coenzyme A, and phosphate. Both the oxidation and the dis mutation systems were strongly inhibited by arsenite. (> Division of Animal Health and Production, C.S.I.R.O., Animal Health Research Laboratory, Melbourne.

38 A. W. RODWELL AND E. SHIRLEY RODWELL Glycerol, in the presence of added catalase, was oxidized to acetate and carbon dioxide, and was not attacked anaerobically. Methylene blue could not replace oxygen for glycerol oxidation. It was assumed that the pathway for glycerol oxidation resembled that described by Gunsalus and Umbreit (1945) for Strep. faecalis, i.e. glycerol is first phosphorylated to a-glycerophosphate which is oxidized by a flavoprotein enzyme to D-glyceraldehyde phosphate. Glyceraldehyde phosphate is then oxidized to pyruvate. In the organism studied by Gunsalus and Umbreit, pyruvate was reduced to lactate in order to regenerate DPN, whereas in A. mycoides it was assumed that pyruvate was further oxidized to acetate and carbon dioxide. II. METHODS The organism, growth conditions, and methods of preparation and standardization of cell suspensions, the preparation of frozen and thawed suspensions, and the manometric and chemical methods used have all been described previously (Rodwell and Rodwell 1954a). Cell-free extract was prepared by grinding lyophilized cells with "Alloxite" as described (Rodwell and Rodwell 1954b ). Hexokinase activity.-hexokinase activity was tested manometrically, by carbon dioxide evolution from bicarbonate buffer with carbon dioxide in nitrogen as the gas phase, in the presence of glucose, ATP, and fluoride (Colowick and Kalcker 1943). Alcohol dehydrogenase.-crystalline alcohol dehydrogenase was prepared from yeast by the method of Racker (1950). Glucose-6-phosphate.-Glucose-1-phosphate was prepared from starch by the action of potato phosphorylase (Hanes 1940) and then converted to glucose- 6-phosphate with a crystalline preparation of phosphoglucomutase (Najjar 1948), following the procedure of Neifakl and Grechishkina (1951). 6-Phosphogluconate.-Glucose-6-phosphate was oxidized with bromine as described by Robison and King (1931). Frttctose-1,6-diphosphate.-A commercial sample of the tetrabasic barium salt was purified as the acid barium salt by the method of Neuberg, Lustig, and Rothenberg (1943). Adenosine triphosphate (ATP).-Commercial A TP was purified by mercury precipitation. All barium salts were converted to the sodium salts immediately before use in the experiments. III. EXPERIMENTAL ( a) Action of Inhibitors on Glucose Oxidation The effect of iodoacetic acid and sodium fluoride on the rate of glucose oxidation by cell suspensions in phosphate buffer is shown in Table 1. Glucose oxidation was sensitive to inhibition by both substances.

GLUCOSE OXIDATION BY ASTEROCOCCUS MYCOIDES 39 (b) Hexokinase Activity Hexokinase activity was tested manometrically with suspension treated by freezing and thawing. The complete system contained fluoride (O.05M), Mg+ + (O.OIM ), frozen and thawed suspension, glucose, ATP, bicarbonate, and 5 per TABLE I EFFECT OF INHIBITORS ON RATE OF GLCCOSE OXIDATION BY SUSPENSIONS OF A. MrCOIDES Warburg vessels contained: M/5 phosphate buffer (ph 7 4) 1 25 ml; bacterial suspension 0 5 ml; inhibitor solution (or water) 0 25 ml (main compartment) ; M/20 glucose 0 25 ml (side bulb); 20% KOH (centre well) Results expressed as percentage inhibition of rate of oxygen uptake Inbibitor (M concn.) Iodoacetic acid 0 0001 Iodoacetic acid 0 00 I Iodoacetic acid 0 01 Sodium fluoride 0 005 Sodium fluoride 0 0075 Sodium fluoride 0 01 Inhibition (%) 29 82 100 77 92 100 cent. carbon dioxide in nitrogen (V Iv) in the gas phase (ph 7.5). The rate of carbon dioxide evolution was constant only during the first 15 min after adding glucose and ATP. The rates, calculated from the first I5-min interval, are set out in Table 2. TABLE 2 HEXOKINASE ACTIVITY OF A. MrCOIDES Complete system: NaHCOa (0 078M) 1 0 ml; NaF (0 5M) 0 2 ml; MgCI. (O'IM) 0 2 ml; frozen and thawed bacterial suspension 0 6 ml; glucose (O IM, side bulb) 0 2 ml; ATP (2 mg 7 min P/ml, side bulb) 0 4 ml; vol. 3 0 ml. Gas phase 7% CO 2 in N. (v/v). Rates expressed as fll CO./hr Components Complete system Complete system without ATP Complete system without glucose Complete system without MgCI. Complete system with cysteine (M/300) Rate 396 144 168 166 420 The results were consistent with the action of hexokinase. Carbon dioxide was evolved most rapidly in the complete system; Mg++ and cysteine both had a stimulatory effect. The results, however, were not conclusive because carbon

40 A. W. RODWELL AND E. SHIRLEY RODWELL dioxide was also evolved at a relatively rapid rate in the absence of either glucose or of ATP. The results were probably complicated by the presence of a fluoride-resistant ATP'ase, since acid-hydrolysable phosphorus (7 min P) had almost disappeared from the contents of all manometer vessels at the end of the experiment. ( c) Aldolase Activity A suitable dilution of a cell-free extract was incubated in the test system of Sibley and Lehninger (1949), as set out in Table 3. The approximate activity determined by the colorimetric method, by the use of the standard curve of Sibley and Lehninger, agreed with the activity determined by alkali-labile phosphate determinations (QHDP = 248). The extract possessed relatively high aldolase activity compared with the figures published for various tissues by Sibley and Lehninger. TABLE 3 ALDOLASE ACTIVITY OF A. MrCOIDES Test system: Cell-free extract 1 0 mi (=0 225mg protein); 2-amino-2-methyl I,3-propane dioi HCI buffer (O IM, ph 8 6) 1 0 ml; hydrazine solution (0'56M, ph 8 6) 0 25 ml; fructose-i, 6-diphosphate(HDP) (0 45M, ph 8 6) 0 25 m!. Incubation period: 30 min at 37 C. Reaction stopped and mixtures deproteinized by addition of2 mllo% (w/v) trichloracetic acid solution (TCA). Samples of TCA supernatants taken (a) for colorimetric method, 0'1, 0'2, 0 3 ml; (b) for alkali-labile P determinations, O 5 ml Colorimetric Method Alkali-labile P HDP split/mg protein/hr (,.1) 258 248 ( d) Oxidation of Phosphorylated Hexose Intermediates The oxidation of phosphorylated hexose intermediates by intact cell suspensions was tested in phosphate buffer. Oxygen uptake was rapid with glucose, but with glucose-6-phosphate, fructose-6-phosphate, hexosediphosphate, gluconate, or 6-phosphogluconate it was no higher than the endogenous rate. The oxidation of the phosphorylated hexoses by suspensions treated by freezing and thawing was tested in bicarbonate buffer with 5 per cent. (v Iv) carbon dioxide in nitrogen as the gas phase, and excess methylene blue in place of oxygen as electron acceptor. This was done because suspensions treated by freezing and thawing often had impaired ability to take up oxygen, but retained their ability to oxidize glucose with methylene blue as electron acceptor. With methylene blue as electron acceptor, the oxidation of 1 mole of glucose results in the evolution of 8 moles of carbon dioxide from bicarbonate buffer (Rodwell and Rodwell 1954a). Carbon dioxide evolution from glucose, glucose-6- phosphate, fructose-1,6-diphosphate, and 6-phosphogluconate is shown graphically in Figure 1.

GLUCOSE OXIDATION BY ASTEROCOCCUS MYCOIDES 41 Glucose, hexosediphosphate, and glucose-6-phosphate were attacked at comparable rates and the amount of carbon dioxide evolved approached the theoretical value, whereas the rate of carbon dioxide evolution from 6-phosphogluconate was no faster than in the absence of substrate. In other experiments fructose-6-phosphate was also shown to be attacked by frozen and thawed suspensions. 224 200 150 3-.. 8 100 50 o TIME (MIN) Fig. I.-Oxidation of phosphorylated hexose intermediates by frozen and thawed suspension of A. mycoides. Manometer vessels contained: frozen and thawed bacterial suspension 1.5 ml; NaHCOs 50 /tm; substrate 1.25 /tm; coenzyme concentrate (fraction A) 1 mg; methylene blue 5 flm. Volume of fluid 2.5 mi; gas phase 7% CO 2 in N2 (v/v). (e) Anaerobic Glucose Breakdown by Frozen and Thawed Suspensions In contrast to intact cell suspensions, suspensions treated by freezing and thawing were capable of glucose breakdown anaerobically. The rate was slow in comparison with the rate of.glucose oxidation. The rate of anaerobic hexose breakdown by suspensions treated by freezing and thawing may be compared with the rate for the oxidation of the substrates with methylene blue as electron acceptor, in Table 4 (for hexosediphosphate breakdown) and in Table 5 (for glucose breakdown). (f) Effect of the Addition of Pyruvate on the Rate of Anaerobic Hexose Breakdown The effect of the addition of pyruvate on the rate of anaerobic hexose degradation by suspension treated by freezing and thawing was tested by the rate of

42 A. W. RODWELL AND E. SHIRLEY RODWELL carbon dioxide evolution from bicarbonate buffer. Hexosediphosphate was used as hexose substrate in this experiment. The results are shown in Table 4. It may be seen that the rate of carbon dioxide evolution when both substrates were present (156 01!hr) was slightly greater than the sum of the rates for the pyruvate dis mutation reaction and for hexosediphosphate breakdown in tlie absence of added pyruvate (120 01!hr). TABLE 4 EFFECT OF THE ADDITION OF PYRUVATE 0]'; THE RATE OF ANAEROBIC BREAKDOWN OF HEXOSEDIPHOSPHATE BY FROZEN AND THAWED SUSPENSIONS OF A. MYCOIDES ''''arburg vessels contained: Main compartment: MilO NaHCO. 0 4 ml; frozen and thawed suspension 0 5 ml; M/30 iodoacetic acid (IAA) 0 25 ml; side bulb: M/20 hexosediphosphate (HDP) 0 2 ml; M/50 methylene blue (MB) 0 3 ml; M/50 pyruvate 0 2 ml; MilO NaHCI. O I m!. Volume fluid 2 5 m!. 5% CO2 in N2 in gas phase Components Rate (/-,1 CO 2/hr) No substrate Pyruvate HDP Pyruvate+HDP Pyruvate+HDP+IAA HDP+MB 20 68 52 156 20 286 ( g) E flect of Linking the Reduction of Acetaldehyde by Yeast Alcohol Dehydrogenase with Anaerobic Glucose Breakdown The complete system contained yeast alcohol dehydrogenase, frozen and thawed suspension, acetaldehyde, and bicarbonate buffer with 5 per cent. (v Iv) carbon dioxide in nitrogen as the gas phase. The rates of carbon dioxide evolution obtained in two experiments are set out in Table 5. It may be seen that, in the presence of both alcohol dehydrogenase and acetaldehyde the rate of glucose breakdown was increased considerably in both experiments. The addition of alcohol dehydrogenase alone increased the rate of glucose breakdown, slightly in one experiment, and considerably in the other experiment. The addition of acetaldehyde alone depressed the rate slightly. (h) Effect of Arsenite on Anaerobic Glucose Breakdown Both the oxidation and the anaerobic dismutation of pyruvate by intact cell suspensions were strongly inhibited by arsenite (Rodwell and Rodwell 1954b). In the presence of arsenite it was found that intact cell suspensions catalysed a homolactic glucose fermentation anaerobically. The effect of the concentration of arsenite on the rate of the reaction in bicarbonate buffer is shown in Figure 2. The rate of oxygen uptake in phosphate buffer is included in Figure 2 for comparison. It may be seen that in the presence of a concentration of

GLUCOSE OXIDATION BY ASTEROCOCCUS MYCOIDES 43 O.03M arsenite the rate of carbon dioxide formation approached the rate of oxygen uptake. That the fermentation in the presence of arsenite was homolactic was conrrmed in another experiment in which lactic acid was estimated colorimetrically in the contents of the manometer vessels. Of the carbon dioxide evolved from bicarbonate buffer in the presence of O.03M arsenite, 96 per cent. was accounted for as lactic acid. 200 150 3 6100 0: o (j u 50 o 10 20 30 TIME (MIN) Fig. 2.-Effect of arsenite on rate of anaerobic glucose breakdown by intact cell suspensions of A. mycoides. Anaerobic breakdown: 0.07.3M NaHC0 3 0.5 ml; bacterial suspension 0.5 ml; arsenite solution 0.25 ml; M/20 glucose 0.1 ml; water to 2.5 m!. 5% CO 2 in N2 in gas phase. Oxygen uptake: M/5 phosphate buffer (ph 7.4) 1.5 ml; bacterial suspension 0.5 ml; M/20 glucose 0.1 ml; 20% KOH 0.2 ml (centre well). IV. DISCUSSION The sensitivity of glucose oxidation to inhibition by iodoacetic acid implies the participation of a DPN or TPN dehydrogenase in glucose breakdown. Fluoride may not have exerted its effect by inhibiting enolase activity, but may have inhibited the biosynthesis of a factor required by the terminal respiratory system as was suggested for the inhibition of pyruvate oxidation (Rodwell and Rodwell 1954b).

44 A. W. RODWELL AND E. SHIRLEY RODWELL Although the presence of hexokinase was nqt demqnstrated cqnclusively, the range Qf hexqses attacked, the QxidatiQn Qf the phqsphqrylated hexqse intermediates Qf the Embden-MeyerhQf-Parnas scheme, and the presence Qf aldqlase provide further evidence. It was concluded that the initial step in the breakdqwn Qf glucqse, fructqse, and mannqse was a phqsphqrylatiqn to. glucqse-6- phqsphate, fructqse-6-phqsphate, and mannqse-6-phqsphate respectively, and that mannqse-6-phosphate WQuld be cqnverted to. fructqse-6-phqsphate by the actiqn Qf phqsphqhexqisqmerase. Since fructqse-l,6-diphqsphate was attacked, and the Qrganism was shqwn to. PQssess aldqlase activity, and since glucqnate Qr 6-phQsphoglucQnate were nqt attacked, it was cqncluded that the breakdqwn Qf hexqses prqceeds through fructqse-l,6-diphqsphate. TABLE 5 EFFECT OF LINKING ANAEROBIC GLUCOSE BREAKDOWX BY FROZEN AND THAWED SUSPENSIOK OF A. MYCOIDES WITH REDUCTION OF ACETALDEHYDE BY YEAST ALCOHOL DEHYDROGENASE Complete system: Frozen and thawed suspension == approx. 1 5 mg N; MgCI 2, 30fLM; alcohol dehydrogenase (AI D) solution in M/20 phosphate buffer containing 0 1% bovine albumin 0 4 ml; glucme 5JLM; acetaldehyde 10JLM; NaHCOa 60fLM; fluid vol. 3 0 ml; 7% CO. in N. (v/v) as gas phase Components Rate ( 1'1 CO./hr) Complete system Complete system without Al D* Complete system without CH 3CHO Complete system without Al D, CH 3CHO Complete system without glucose Complete system without glucose, Al D*, CH 3CHO Complete system without AI D, CHaCHO; with methylene blue (6 I'M) Expt. I 135 59 80 70 13 8 280 Expt.2 233 61 120 63 28 26 * Alcohol dehydrogenase solution replaced by M/20 phosphate buffer containing O I % bovine albumin. The pathway fqr the QxidatiQn Qf triqse phqsphate must nqw be cqnsidered. If glyceraldehyde phqsphate is Qxidized thrqugh the Embden-MeyerhQf scheme, i.e. by a triqse phqsphate dehydrogenase specific fqr DPN, an explanatiqn must be fqund fqr the absence Qf anaerqbic glucqse breakdqwn by intact cell suspensiqns. It was cqncluded previqusly that the pyruvate dismutatiqn system Qf A. mycoides resembled the system described in Qther Qrganisms. In the dismutatiqn system, DPNH 2 fqrmed by the QxidatiQn Qf pyruvate is re-qxidized by the reductiqn Qf pyruvate by lactic dehydrqgenase. It was difficult to. see why the reductiqn Qf pyruvate by lactic dehydrqgenase CQuid nqt serve to. re-qxidize DPNH 2 formed by the QxidatiQn Qf glyceraldehyde phqsphate in the glycqlysis system also.. AnQther possible explanatiqn fqr the absence Qf anaerqbic glucqse

GLUCOSE OXIDATION BY ASTEROCOCCUS MYCOIDES 45 breakdown might be that triose phosphate is oxidized by an enzyme requiring an electron carrier other than DPN, e.g. TPN (Tewfik and Stumpf 1951; Gibbs 1952; Arnon 1952), and that oxygen or other electron acceptor was required for the re-oxidation of TPNH2 This would not explain why suspensions treated by freezing and thawing were capable of a slow anaerobic glucose breakdown. No attempt was made to isolate a triose phosphate dehydrogenase from A. mycoides, and to determine its coenzyme specificity, but indirect evidence was obtained that triose phosphate is oxidized by a DPN-specific enzyme. When anaerobic glucose breakdown by suspensions treated by freezing and thawing was linked with an external system for the oxidation of DPNHz, viz. the reduction of acetaldehyde by purified yeast alcohol dehydrogenase, the rate of anaerobic glucose breakdown was very much increased. This experiment suggests that the rate of anaerobic glucose breakdown by frozen and thawed suspension is limited by the rate at which DPNH2 is re-oxidized. It might be expected that in this system, pyruvate would undergo dismutation to lactate, acetate, and carbon dioxide. The overall equation for the reaction would then be: C6H 120 6 + 2CH3CHO + H 20 = CHsCOOH + CH3CHOH.COOH + CO2 +2CH3CH20H. It would therefore be expected that for each mole of glucose broken down, 3 moles of carbon dioxide would be evolved from bicarbonate buffer, and 1 mole of carbon dioxide would be formed in phosphate buffer. The amounts of carbon dioxide found varied in different experiments and did not reach the theoretical value demanded by the equation. However, more than 2 moles in bicarbonate buffer and approximately 0.6 moles in phosphate buffer were recorded. The incomplete yield of carbon dioxide may have been due to aldehyde toxicity. Acetaldehyde, in the absen~e of alcohol dehydrogenase, slightly depressed the rate of the reaction. An analogy may be drawn between glucose fermentation in A. mycoides and Neuberg's forms of glycerol fermentation in yeast. In the third form of glycerol fermentation, in which the fermentation is carried out at an alkaline ph, acetaldehyde is removed by a dismutation reaction, DPNH 2 is then reoxidized by the reduction of dihydroxyacetone phosphate by a-glycerophosphate dehydrogenase. It is postulated that in A. mycoides, pyruvate is removed by the pyruvate dismutation reaction, but the organism possesses no other anaerobic mechanism for the reoxidation of DPNH 2, and fermentation cannot occur. The glycolytic and pyruvate dismutation systems compete for pyruvate, lactic dehydrogenase, and DPN. In the presence of added pyruvate, or of another system for the oxidation of DPNH2, the rate of anaerobic hexose breakdown by frozen and thawed suspensions was increased. In the presence of arsenite, a potent inhibitor of the pyruvate dis mutation reaction, intact cell suspensions catalysed a homolactic glucose fermentation at a rate comparable with the rate of glucose oxidation. A number of explanations are possible for the difference between organisms such as Strep. faecalis, which possess the pyruvate dismutation system and yet

46 A. W. RODWELL AND E. SHIRLEY RODWELL catalyse a homolactic glucose fermentation, and A. mycoides. There might be differences between the relative concentrations of pyruvic oxidase and lactic dehydrogenase, differences in the relative affinities of the enzymes for pyruvate, or differences in the organization within the cells of the two species. The latter possibility appears plausible. Treatment of suspensions by freezing and thawing resulted in a lowering of the rate of the pyruvate dismutation reaction, and suspensions so treated were capable of a slow anaerobic glucose breakdown. The freezing and thawing treatment may have resulted in a disturbance of the intracellular organization, which allowed pyruvate to combine with lactic dehydrogenase rather than pyruvic oxidase. V. ACKNOWLEDGMENT We wish to thank Dr. A. W. Turner, Assistant Chief of Division, for his interest and encouragement. VI. REFERENCES ARNON, D. I. (1952).-Science 116: 635. BARRON, E. S. G., and FRlEDEMANN, T. E. (1941).-J. Biol. Chem. 137: 593. COLOWICK, S. P., and KALCKAR, H. M. (1943).-J. Biol. Chem. 148: 117. ENTNER, N., and DOUDOROFF, M. (1952).-J. Biol. Chern. 196: 853. GIBBS, M. (1952).-Nature 170: 164. GUNSALUS, I. C., and UMBREIT, W. W. (1945).-J. Bact. 49: 347. HANES, C. S. (1940).-Proc. Roy. Soc. B 129: 174. "KOEPSELL, H. J. (1950).-1. Biol. Chem. 186: 743. NAJJAR, V. A. (1948).-J. Biol. Chem. 175: 281. NEIFAKL, S. A., and GRECHISHKINA, V. I. (1951 ).-Biochemistry, Leningr. 116: 444. NEUBERG, C., LUSTIG, H., and ROTHENBERG, M. A. (1943).-Arch. Biochem. 3: 33. RACKER, E. (1950).-J. Biol. Chem. 184: 313. ROBISON, R., and KING, E. J. (1931).-Biochem. J. 25: 323. RODWELL, A. W., and RODWELL, E. SHIRLEY (l954a).-.aust. 1. Biol. Sci. 7: 18. RODWELL, A. W., and RODWELL, E. SHIRLEY (1954b ).-Allst. T. Bioi. Sci. 7: 31. SIBLEY, J. A., and LEHNINGER, A. L. (1949).-J. Biol. Chern. 177: 859. STOKES, FLORA N., and CAMPBELL, J. J. R. (1951).-Arch. Biochem. 30: 121. TEWFIK, S., and STUMPF, P. K. (1951);-J. Bioi. Chern. 192: 527..