JOURNAL OF BACTERIOLOGYVol. 87, No. 4, p. 844-851 April, 1964Copyright 0 1964 American Society for Microbiology

Printed in U.S.A.

2-KETOGLUCONATE FERMENTATION BY STREPTOCOCCUSFAECALIS

J. L. GODDARD AND J. R. SOKATCHDepartment of Microbiology, University of Oklahoma School of Medicine, Oklahoma City, Oklahoma

Received for publication 23 November 1963

ABSTRACT

GODDARD, J. L. (University of Oklahoma Schoolof Medicine, Oklahoma City), AND J. R. SOKATCH.2-Ketogluconate fermentation by Streptococcusfaecalis. J. Bacteriol. 87:844-851. 1964.-Streptococ-cus faecalis 1OCI did not grow with 2-ketogluco-nate alone as an energy source, but did grow whengluconate was added. More growth was obtainedthan could be accounted for by the gluconatealone. The requirement for gluconate in the stimu-lation of growth on 2-ketogluconate was found tobe stoichiometric, not catalytic. Glucose did notreplace gluconate in this phenomenon, apparentlyowing to the repression of the 2-ketogluconatepathway by glucose. Resting cells grown on a com-bination of gluconate and 2-ketogluconate didferment 2-ketogluconate without added gluconate.Fermentation balance studies with resting cellsdetected the following products in moles (per moleof 2-ketogluconate): carbon dioxide, 0.98; lacticacid, 0.19; formic acid, 1.42; acetic acid, 0.70; andethanol, 0.42. 2-Ketogluconate-1 -C14 and -2-C"4 wereprepared and fermented. The data were inter-preted to show that 90% of the substrate was de-carboxylated to carbon dioxide and pentose phos-phate. Pentose phosphate was then fermented topyruvate through the sedoheptulose diphosphatevariation of the pentose phosphate pathway foundin this organism. The other 10% of the substratewas converted to pyruvate by way of the Entner-Doudoroff pathway. Calculations of the energyavailable by the above combination of pathwaysindicated that about 2.3 moles of adenosine tri-phosphate per mole of 2-ketogluconate could beobtained if the energy available in acetate forma-tion is conserved through the acetokinase reaction.

It was previously reported that Streptococcusfaecalis 1OCl grown on gluconate was able to fer-ment 2-ketogluconate (Sokatch and Gunsalus,1957). 2-Ketogluconate was subsequently ruledout as an intermediate in gluconate fermentationbecause aged or dried cells lost the ability to fer-ment 2-ketogluconate, but retained the ability toferment gluconate. It was established that 1 mole

of carbon dioxide was produced per mole of2-ketogluconate fermented, although the otherproducts were not identified.

This problem was further investigated to es-tablish the relationship of the 2-ketogluconatefermentation to gluconate fermentation by thisorganism and to the 2-ketogluconate fermenta-tion by Leuconostoc mesenteroides (Blakley andBlackwood, 1957). Gluconate fermentation by S.faecalis occurs by a mixed pathway, where half ofthe substrate is oxidized to carbon dioxide andpentose phosphate and the other half is convertedto pyruvate and triose phosphate by way of 2-keto-3-deoxy-6-phosphogluconate. 2-Ketoglucon-ate fermentation by L. mesenteroides probablyoccurs by decarboxylation to pentose phosphateand carbon dioxide, followed by phosphoketolasecleavage of pentose phosphate to acetyl phos-phate and triose phosphate.

This report describes studies of growth yields,fermentation balances, and product labeling ob-tained during growth on, or resting cell fermenta-tion of, 2-ketogluconate by S. faecalis.

MATERIALS AND METHODS

Microbiological methods. The organism used inthese studies was S. faecalis 1OCl, which wasmaintained and grown in mass culture as previ-ously described (Sokatch and Gunsalus, 1957).Because no growth was obtained when 2-keto-gluconate alone was used, a combination of 30 gof sodium gluconate and 10 g of potassium2-ketogluconate was used per 15 liters of ACmedium (Wood and Gunsalus, 1942). Yield ofcell dry weight per Amole of 2-ketogluconate wasalso done according to published methods (So-katch and Gunsalus, 1957), with the use of thebasal medium of O'Kane and Gunsalus (1948).When carbon dioxide was trapped during growth(Table 1), the growth tube was fitted with agassing device which allowed the carbon dioxide-free nitrogen to be bubbled through the mediumduring growth. The gas leaving the growth tube

844

on May 24, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

VOL. 87, 1964 2-KETOGLUCONATE FERMENTATION BY S. FAECALIS

was bubbled through standard 0.25 N bariumhydroxide which was about 0.085 M in bariumchloride (Steele and Sfortunato, 1949). Ordinarily,three such traps were used in series, each con-

taining 3 ml of the barium hydroxide-bariumchloride solution. Residual barium hydroxide was

titrated with standard 0.1 N hydrochloric acid.Barium carbonate was collected on coarse sin-tered-glass frits with a Millipore XX10 025 00filtering assembly. All samples were corrected toinfinite thickness, and counts per minute were

converted to specific activity and expressed muc/mmole of carbon.

Fermentation balance studies. Resting-cell fer-mentations were carried out in a Warburg appara-

tus under nitrogen. A typical experimental vesselcontained 50 Mmoles of potassium 2-keto-gluconate, 300 Mmoles of potassium phosphatebuffer (pH 6.0), and cells, equivalent to about 80mg (dry weight) in a final volume of 3.0 ml. Thereaction was followed by measuring carbon diox-ide evolution. Large-scale experiments withlabeled substrates were done in glass tubes (25 by200 mm) fitted with a gassing attachment. Nitro-gen gas was bubbled through the reaction mix-ture, and the evolved carbon dioxide was trappedin barium hydroxide-barium chloride, as previ-ously described. A typical reaction mixture in thiscase would contain 500 ,moles of potassium2-ketogluconate, 4.0 mmoles of potassium phos-phate buffer (pH 6.0), and about 400 mg (dryweight) of cells in a final volume of 24 ml. Thefermentation was allowed to proceed for 4.5 hr;the cells were removed by centrifugation in thecold after the addition of 2 ml of 10 N sulfuricacid.

Chemical methods. Glucose-I- and -2-C'4 were

purchased from Nuclear-Chicago Corp. (DesPlaines, Ill.). Sodium gluconate was purchasedfrom Pfanstiehl Chemical Co., Waukegan, Ill.

Lactic acid was determined by the method ofBarker and Summerson (1941). Alcohol was sep-arated from the fermentation products by steamdistillation after adjusting the pH to 7.0. Alcoholwas then determined by oxidation to acetic acid.Samples of 50 ml of the steam distillate were col-lected in 250-ml flasks with ground-glass joints;34 ml of the oxidizing agent were added (34 g ofNa2Cr207 .2H20, 169 ml of 10 N H2SO4, and waterto make 250 ml). A ground-glass stopper was thenwired in place, and the flask was immersed in a

boiling-water bath for 30 min. The resulting acetic

TABLE 1. C1402 production from 2-ketogluconate byStreptococcus faecalis in the presence of other

energy sources*

CO2 C14 Per centEnergy source evolved content of C14

input

jsmoles ,uc

None .................... 59 -

Glucose ................... 74 -

Gluconate ................. 1102-Ketogluconate-1 -C14 ..... 53 0Glucose + 2-ketoglucon-

ate-1-C'4................ 58 0.09 0.72Gluconate + 2-ketoglu-

conate-1-C'4............. 181 7.7 62

* To 10 ml of the basal medium were added 100Asmoles of each substrate. Input of C14 was 12.4m/Ac of 2-ketogluconate-l-C'4.acid was then steam-distilled and titrated withstandard base.

Degradation of labeled compounds. Total oxida-tions were done with the persulfate method aspreviously described (Sokatch and Gunsalus,1957). Alcohol was separated from the fermenta-tion mixture, as described above, and oxidized toacetic acid for degradation. Acetic acid was de-graded according to the method of Phares (1951),except that carbon dioxide was absorbed directlyin the barium hydroxide-barium chloride mixture.After the removal of alcohol, the volatile acids,acetic and formic, were removed by steam dis-tillation and titrated with standard base. Thesteam distillate containing the acid salts was con-centrated in an oven at 100 C. Formic acid wasoxidized selectively according to the mercuricoxide procedure of Osburn, Wood, and Werkman(1933). This method does not produce oxidationof either acetic or lactic acids. Carbon dioxidefrom formic acid oxidation was absorbed directlyin the barium hydroxide mixture, as alreadydescribed. Acetic acid was recovered from themercuric oxide oxidation procedure by steam dis-tillation. The steam-distillate fractions weretitrated with standard sodium hydroxide, com-bined, and taken to dryness in an oven at 100 C.The dried sodium salt was used in the Phares(1951) modification of the Schmidt reaction asdescribed for alcohol. Lactic acid was degradedin the experiments in which arsenite was used.Lactic acid was extracted from the fermentationmixture by continuous ether extraction for 48

845

on May 24, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

GODDARD AND SOKATCH

hr. To separate lactic acid, the extract was chro-matographed on silicic acid (Steele, White, andPierce, 1954), and lactic acid was degraded aslpreviously described (Sokatch, 1960).

Preparation of potassiuml 2-ketogluconate. Be-cause rather large amounts of 2-ketogluconatewere required to grow the organism in mass cul-ture, the method of Stubbs et al. (1940) wasmodified to produce laboratory-scale amounts of2-ketogluconate. The organism used was a non-pigmented strain of Pseudowonas aeruginosa usedin other studies in this laboratory (Norton, Bul-mer, and Sokatch, 1963). The inoculum wasprepared by growing the organiosm in a mediumwhich contained 12 g of glucose and 0.5 g of -eastextract in 100 ml of water. After this medium hadbeen autoclaved, 2.7 g of calcium carbonate,wlhich had been sterilized sep)arately, were added.The medium was inoculated and incubated over-night at 37 C. The entire contents of the inoculumflask were added to the production flask. Themedium for the prodluction of 2-ketogluconatecontained 59 g of glucose, 2.5 g of yeast extract,125 mg of MgSO4-7H20, 300 mg of KH2PO4,and 450 ml of water.A solution of 1 g of urea in 50 ml of water and

13.5 g of calcium carbonate, both of which hadbeen autoclaved separately, was added to theproduction flask at the time of inoculation. Incu-bation was at 37 C, with aeration and stirring.When the labeled substrates were prepared, 50Pc of glucose-C14 were added the day after theproduction flask was inoculated. The medium wastested daily for the disappearance of glucose andthe appearance of 2-ketogluconate by chromato-graphing a sample of the flask contents on shortstrips of Whatman no. 1 paper in tubes (25 by200 mm) with methyl alcohol, ethyl alcohol, andwater, 9:9:2 (Stokes and Campbell, 1951). Thepapers were sprayed with o-phenylenediamine(De Ley, 1954). WVhen the glucose had disap-peared completely, generally after about 7 to 8days, the cells and calcium carbonate were re-moved by centrifugation at 12,100 X g in aServall RC-2 centrifuge for 10 min.The clarified medium was concentrated with

the aid of a rotarv evaporator to about 100 mland was then treated with Dowex 50 (acid form)to remove cations. The treatment with Dowex 50was repeated until there was no further decreasein pH. The solution was then brought to pH6.5 with solid potassium hydroxide, although the

final adjustment was made with 1 N potassiumhydroxide to prevent over-alkalinization. Thesolution was again reduced in -olume, this timeto about 30 to 40 ml. Crystalline potassium 2-ketogluconate appeared when the volume wasreduced to about 50 ml. This solution w-as allowedto cool, and the crystalline l)otassium 2-keto-gluconate was harvested by filtration. The filtercake was wsashed several times wNith small por-tions of cold 70% alcohol until the washings werecolorless. A second crop of cry-stals was., obtainedfrom the filtrate and washings bv again reducingthe volume. The yield was 30 to 33 g of l)otassium2-ketogluconate.

Elemental analysis of a typical samnple of po-tassium 2-ketogluconate for K, C, andl H yieldedthe following results: KC6H907(2.32.168); calcu-lated: K, 16.80; C, 31.00; H, 3.89. Obtained:K, 16.70; C, 31.36; H, 4.04. These results agreeclosely with the theoretical percentages for K,C, and H. The phenylhbdrazide hy(drazone waspreparedl according to the methodl of Ohle andBerend (1927) and melted at 109 to 110 C, thereported melting point being 108 C. The ana-lytical data fit either potassium 2-ketogluconateor 5-ketogluconate. The latter possibility wasruled out when the periodic acid oxiidation tech-nique of Juni and Heym (1962) was used. Inthis procedure, compounds such as gluconateand 5-ketogluconate, w-hich contain a hydroxylgroup at carbon 2, yield glyoxylic acid from car-bons 1 and 2. On the other hand, 2-ketogluconateyields glyoxylic acid only after reduction togluconate Nith borohvdride. This latter resultwas obtained, thus establishing the identity ofour material as )otassiun-l 2-ketogluiconate. Addi-tionally, it was lpossible to establish the limit ofcontamination of 2-ketogluiconate with gluconateas about 317.

RESULTS

Growth of S. faccalis on, glucosm, gluconate, and2-ketogluconate. Preliminary grow-tlh experimentswith S. faecalis in basal mediuim with 2-keto-gluconate as the energy source revealed that nogrowth occurred at the expense of this substrate.Consiidering the known pathway of 2-ketoglu-conate metabolism in Aerobacter cloacae (De Leyand Verhofstede, 1955) and P. fluorescens (Framp-ton and Wood, 1961), which involves reductionof 2-ketogluconate-6-p)hosphate to 6-phospho-gluconate, the possibility was considered that a

846 J . BACTERI OL.

on May 24, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

VOL. 87, 1964 2-KETOGLUCONATE FERMENTATION BY S. FAECALIS

catalytic amount of a reducing agent was re-

quired to initiate the fermentation. Growth ofS. faecalis in the presence of gluconate and 2-ketogluconate resulted in an increased yield ofcells, as compared with that obtained withgluconate alone. However, when glucose and2-ketogluconate were used, no greater cell yieldwas obtained than when glucose alone was used.The lack of stimulation of growth on 2-keto-gluconate by glucose was further investigated.Because it was known that resting-cell fermenta-tion of 2-ketogluconate by S. faecalis producescarbon dioxide, growth of the organism in thepresence of 2-ketogluconate-1-C"4 and eitherglucose or gluconate and measurement of theC14 content of the evolved carbon dioxide was

used as an index of 2-ketogluconate fermentation(Table 1). In the case of 2-ketogluconate-1-C'4alone, or in the presence of glucose, no C'4 was

recovered in the evolved carbon dioxide. In thecase of 2-ketogluconate-1-C'4 plus gluconate, 62%of the C'4 was recovered in the evolved carbondioxide, and the yield of carbon dioxide was alsoincreased. It thus appears probable that glucoserepresses the utilization of 2-ketogluconate.Growth of S. faecalis in graded amounts of

gluconate with 5 ,umoles/ml of 2-ketogluconateresulted in more growth than would be expectedfrom gluconate alone and with a slope abouttwice that obtained with gluconate alone (Fig. 1).Similarly when the gluconate concentration was

kept constant at 5 ,moles per ml and 2-keto-gluconate was varied, the increased amount ofgrowth was found to be proportional to the con-

centration of 2-ketogluconate. The slope of thisline was found to be 20 Aig (dry weight) per ,umole

oMOLES GLUCONATE OR 2-KETOGLUCONATE/ML

FIG. 1. Growth yields of Streptococcus faecalis on

gluconate and 2-ketogluconate.

TABLE 2. End products of 2-ketogluconatefermentation*

O/R value of/Amoles moles per productst

Products Amoles of 2ketoglu Oarot-onate -Oxid- Re-

ized duced

Carbon di-oxide . ....49 49 0.98 +1.96

Formic acid 71 71 1.42 +1.42Ethanol. 21 42 0.42 - -0.82Acetic acid 35 70 0.70 -

Lactic acid. 9. 27 0.19 -

* Fermentation balance studies were performedwith 50 ismoles of 2-ketogluconate, except for lac-tic acid determination which was done on a sepa-rate 10-,umole fermentation.

t Carbon recovery: 259/300 = 86%O.$ Total: oxidized, +3.38; reduced, -0.82. Sum

of oxidized and reduced products = +2.56; the-oretical for 2-ketogluconate, +2.00.

of 2-ketogluconate. These and the precedingdata suggest a relationship between the twosugar acids such that 1 mole of 2-ketogluconateis fermented to produce energy for each mole ofgluconate present. Hence, the role of gluconatein the fermentation of 2-ketogluconate is notcatalytic but stoichiometric. It should be pointedout that these results concern energy productionas measured by growth yields, and that restingcells ferment 2-ketogluconate without addedgluconate.End products of 2-ketogluconate fermentation.

The results of the studies with resting cells onthe fermentation of 2-ketogluconate are givenin Table 2. Inspection of these data reveals that1 mole of carbon dioxide was produced per moleof 2-ketogluconate fermented. The remainder ofthe carbon was found principally as the prod-ucts of the elastic reaction, formate, ethanol, andacetate, with surprisingly little lactic acid.The carbon recovery is somewhat low, approx-

imately 86%70, and summation of the oxidizedand reduced products reveals that the balance isshort in reduced products. This could be explainedby a failure to recover all the ethanol produced inthe fermentation. The balance is short in C2products because ethanol plus acetic acid totals1.1 moles per mole of 2-ketogluconate, whereas1.42 moles of formate (C1) per mole of 2-keto-gluconate were obtained. If the value for ethanol

GROWTH YIELDS

(2)Gluconate Voried- 300

5 j,moles/ml 2-Keto-

gIucooote

(l)looote 20 (3)2-Kelo ogluonate Vaioed -

Alone. 5 joooles/ml Gluconote./ 1 ~~~~~~~~~Slope * 20 pg/pjmolo

I0055polesml 2- Ketogluconate

Slope (2)-(l)* I9pg/pymole No Substretete

2 3 4 5 2 3 4 5

847

-3

wIt

on May 24, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

GODDARD AND SOKATCH

is adjusted accordingly, the fermentation balancefor 2-ketogluconate may be represented ideallyas: (1) 2-ketogluconate = 1.0 carbon dioxide +0.2 lactic acid + 1.47 formic acid + 0.73 ethanol+ 0.73 acetic acid.In the study of the gluconate fermentation, it

was observed that the addition of 10-3 M arsenitealtered the fermentation, so that the proportionof clastic products was very small (Sokatch andGunsalus, 1957). The mechanism by which thisoccurs is not clear because lipoic acid was notimplicated in this reaction. The same effect was

observed in the case of the 2-ketogluconate fer-mentation (Table 3). Carbon dioxide productionwas unaffected, but the yield of lactic acid was

raised from 0.19 moles per mole of 2-ketoglu-conate to 1.33 moles per mole of 2-ketogluconate.

SCCOOH 3 CO2

+ C*C-=O C

C C

3C 3C

CC C

CC C

C

*CC C

1

C C

C C

C

*C

IC+ Pentose C

I +C

Phosphate c

I C

C

C

TABLE 3. Effect of 10-3 M arsenite on gluconateand 2-ketogluconate fermentation

Carbon dioxide Lactic acid(,umoles) (,moles)

Additions- Ar- +Ar- -Ar- + Ar-senite senite senite senite

None................. 0.89 1.2 0.8 0.6Gluconate, 10 umoles.. 6.1 5.9 8.0 16.02-Ketogluconate, 10

,imoles... 10.4 10.4 1.9 13.3

TABLE 4. Product labeling in 2-ketogluconate-1-and -2-C14 fermentation

2-Ketoglu- 2-Ketoglu-conate-1-C04 conate-2-C14

ProductSpecific Isotope Sp ecific Isotopeactivity5 enrlichi- activity enrich-mentt ment

Carbon dioxide ... 120 5.8 1.9 0.09Formic acid. 15 0.7 23 1.09Ethanol-CH20H 0........ - 8.8 0.41-CH3.. 0 _ 34.3 1.62

Acetic acid-COOH ......... 0.9 0.04 15.3 0.72-CH3 ........... 0.2 0.01 34.1 1.61

* Specific activity expressed as m,uc/mmole ofcarbon. Specific activity of 2-ketogluconate-1-C14 = 20.6 m,uc/mmole of carbon. Specific activityof 2-ketogluconate-2-CI4 = 21.2 m,uc/mmole ofcarbon.

t Expressed as specific activity of product/specific activity of 2-ketogluconate.

FIG. 2. Formiiation of labeled hexo.,e and triosephosphates from 2-ketoglzuconate-2-C;('4; 3 moles of2-ketogluconate-2-C'4 are converted to 3 moles each ofcarbon dioxide and pentose phosphate-1-C14. Thelatter are converted to hexose and triose phosphate,with the indicated labeling pattern, by means of thesedoheptutlose diphosphate pathway found in thisorganism (Sokatch, 1960, 1962). The usual sequenceof transketolase and transaldolase would also providethe same labeling pattern. Hexose and triose phos-phates are then converted to pyruvate-1 ,3-C14 by theusiual Embden-Mleyerhof pathway enzymes.

Product labeling in 2-ketoglucon.ate-1 -C01 fer-mentation. Resting-cell fermentation of 2-keto-gluconate-1-C'4 resulted in production of carbondioxide with a C'4 concentration nearly six timesthat obtained in the total oxidation of 2-keto-gluconate (Table 4). About 10%7f of the totalisotope was found in formate, but no other com-

pound was significantlyI radioactive. Fermenta-tion of 2-ketogluconate-2-C'4 resulted in radio-active formate and methyl carbons of aceticacid and ethanol. Some of the ra(lio isotope was

also found in the carboxyl carbon of acetic acidand in the carbinol carbon of ethanol. Theseresults are best interpreted as (lecarboxylationof 2-ketogluconate to pentose and conversion ofthe pentose to hexose p)hosphate. Starting with2-ketogluconate-2-C'4, this pathway would resultin pyruvic acid with radioactivity in the methyland carboxyl carbons in the ratio of 2:1. Figure2 demonstrates the formation of labeled hexoseand triose phosphates from 2-ketogluconate-2-C'4.Pyruvate, with the indicated labeling pattern,

848 J. BACTERIOL.

on May 24, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

2-KETOGLUCONATE FERMENTATION BY S. FAECALIS

TABLE 5. Product labeling in 2-ketogluconate-1 - and2-C14 fermentation in the presence of

1O3 M arsenite*

2-Ketoglu- 2-Ketoglu-conate-1-C14 conate-2-C04

ProductSpecific Isotope Specifi sotopeaciiy*enrich- Spcfc enrich-atvt*ment activity ment

Carbon dioxide ... 107 5.2 1.6 0.07Lactic acid-COOH.12 0.58 22.3 1.05-CHOH... ot 4.4t 0.20-CH3 ........... Ot 39.2t 1.83

* All conditions identical to those in Table 4with the exception that fermentations were donein the presence of 10-3 M arsenite.

t Determined by persulfate oxidation of aceticacid from the permanganate oxidation of lacticacid.

t Determined by Schmidt degradation of aceticacid from the permanganate oxidation of lacticacid.

would then arise from these compounds by meansof the usual Embden-Meyerhof reactions.The same pathway with 2-ketogluconate-l-C'4

would result in the bulk of the isotope beingfound in carbon dioxide. Explanation of the sig-nificant amount of label found in formate in thiscase is most readily accomplished by assumingthat a portion of the 2-ketogluconate is routedthrough the Entner-Doudoroff pathway (Entnerand Doudoroff, 1952) which would producepyruvate labeled in the carboxyl group.

These same fermentation studies were repeatedin the presence of 10-3 M arsenite; in this case,the isotope distribution of lactic acid was studied(Table 5). In the case of 2-ketogluconate-1-C'4,again the isotope was found largely in carbondioxide. About 10%7, was recovered in the car-boxyl group of lactic acid corresponding to thatamount recovered in formate when arsenite wasnot used. In the case of 2-ketogluconate-2-C14,the methyl and carboxyl groups of lactic acidwere labeled in the proportion of 2:1 as expected.It is clear that arsenite did not change the majorcourse of the fermentation, but rather changedthe distribution of end products formed frompyruvate.

DISCUSSION

The data from the experiments with 2-keto-gluconate-1- and -2-C04 are best interpreted as a

decarboxylation of 2-ketogluconate to carbondioxide and pentose phosphate. This could occurby phosphorylation of 2-ketogluconate, reductionto 6-phosphogluconate, and oxidation to ribulose5-phosphate and carbon dioxide. As was foundin the case of the gluconate fermentation in thissame organism, pentose phosphate would thenbe converted to hexose phosphate and fermentedby way of the Embden-Meyerhof pathway. InS. faecalis, transaldolase appears to be lacking(Sokatch, 1960); however, its function may beserved by a combination of phosphofructokinaseand aldolase. Sedoheptulose 1,7-diphosphate isan intermediate and is converted to fructose1, 6-diphosphate by aldolase. The labeling patternand energy yield are identical with that obtainedwith the use of transaldolase (Sokatch. 1962).

(a) 3 2-ketogluconate + 3 adenosine triphos-phate (ATP) -- 3 carbon dioxide + 3 pen-tose phosphate

(b) 2 pentose phosphate - sedoheptulose7-phosphate + triose phosphate

(c) sedoheptulose 7-phosphate + ATP -+ sedo-heptulose 1,7-diphosphate + adenosinediphosphate

(d) sedoheptulose 1 ,7-diphosphate + triosephosphate -* fructose 1,6-diphosphate +erythrose 4-phosphate

(e) erythrose 4-phosphate + pentose phosphate-+ fructose 6-phosphate + triose phos-phate

Sum: 3 2-ketogluconate + 3 ATP = fructose1,6-diphosphate + fructose 6-phosphate+ triose phosphate

By this scheme, the isotope from 2-ketoglu-conate-1-C"4 would be recovered in the carbondioxide at a specific activity six times that ob-tained in the total combustion of 2-ketogluconate-1-C14. With similar reasoning, fermentation of2-ketogluconate-2-C14 should result in pyruvatelabeled in the methyl and carboxyl carbons at aspecific activity 2.4 and 1.2 times, respectively,that obtained in the total oxidation of 2-keto-gluconate (see also Fig. 2). The data presentedin Tables 4 and 5 are generally in agreement withthese expectations, assuming the usual reactionsleading from pyruvate to the obtained endproducts.

It is possible to calculate an energy yield permole of 2-ketogluconate fermented on the basisof the above interpretation of the labeling dataand the balance presented in equation 1. Thepathway from pentose to hexose phosphate

VOL. 87, 1964 849

on May 24, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

GODDARD AND SOKATCH

through sedoheptulose diphosphate yields 1.67moles of ATP per mole of pentose, as does thetransaldolase pathway (Elsden and Peel, 1955).Assuming that 90%0 of the 2-ketogluconate isfermented by this route, the energy yield wouldbe 0.9 X 1.67 = 1.51 moles of ATl' per mole of2-ketogluconate. By a similar process, the ATPcontribution from the Entner-Doudoroff path-way, which yields a net of 1 mole of ATP permole of hexose, would in this case be 0.1 moleof ATP per mole of 2-ketogluconate. Finally, 1mole of ATP is available from each mole of ace-tate through the action of the enzyme aceto-kinase. Based on the theoretical balance presentedin equation 1, 0.73 moles of ATP per mole of2-ketogluconate wN-ould be produced. The totalcalculated ATP yield would then be 2.34 molesper mole of 2-ketogluconate. The slope of thecurve in Fig. 1, obtainied by subtracting thegroN-th due to gluconate from that obtained inthe presence of both sugar acids is 19 ,g (dryweight) per,mole of 2-ketogluconate. Assuminga dry-weight yield of about 10 ,g/,umole of ATP(Bauchop and Elsden, 1960), then it may beestimated from this method that 1.9 ,moles netof ATP are produced per,moles of 2-ketoglu-conate, which is within reasonable range of thecalculations. The ATP yield in the gluconatefermentation may also be corrected for the ace-tate produced, in this case about 0.5 ,mole ofacetate per mole of gluconate. The final ATPyield from gluconate then w-ould be about 1.8,unoles per ,mole of gluconate fermented. Theslope of the curve for growth on gluconate alone(Fig. 1) is 20 Mg (dry- W(eight) per ,umole ofgluconate.The failure of S. faecalis to grow on 2-keto-

gluconate alone is not understood. The abilityof the organism to ferment 2-ketogluconate inthe resting-cell experiments wTithout added glu-conate indicates that a complete pathway existsfor the metabolism of 2-ketogluconate in thecells used. On the other lhand, the apparentstoichiometric requirement for gluconate duringgrowth suggests that gluconate is somehow in-volved in energy metabolism of 2-ketogluconate,possibly in the early stages of metabolism. Thismay cause some revision in the preceding cal-culations of energy yield.The conversion of a rather large proportion of

2-ketogluconate to ethanol, acetic acid, andformic acid recalls the fermentation of galactose

by S. pyogenes (Steele et al., 1954) and S. faecalis(Fukuyama and O'Klane, 1962). T'his phenom-enon was explained by Eisenberg and O'Kane(1963) as a result of a coIm1petitioIi for nico-tinamide adenine dlinucleotide (NAI)) by lacticdehydrogenase and uridine diphosphate galactose4-epimerase, with the resuilt that the NAD)-nonrequiring elastic system is favored in thecompetition for pyruvate. Ak similar situationcan be visualized in the case of the 2-ketoglu-conate fermentation, because there is reason tobelieve that the initial steps in the fermentationinvolve reduction of the substrate, probably witha pyridine nucleotide cofactor.

ACKNOWNLEDGMENTS

This research w-as supp)ortecl by V.S. PublicHealth Service research grant RG 6070 andResearch Career 1'rogram awarcl no. GAI K313343 from the Division of General AMedicalSciences.

LITERATURE CITED

BARKER, S. B., AND W. H. SUIMMERSON. 1941. Thecolorimetric determination of lactic acid inbiological material. J. Biol. Chem. 138:535-554.

BAUCHOP, T., AND S. R. ELSDEN. 1960. The growthof micro-organisms in relation to their energysupply. J. Gen. MIicrobiol. 23:457-469.

BLAKLEY, E'. R., .AND A. C. BLACKWOOD. 1957.The degradation of 2-keto-D-gluconate-C'4,D-gluconate-C'4 and D-fructose-C'4 by Leuco-nostoc miiesenteroides. Can. J. Microbiol. 3:741-744.

DE LEY, J., 1954. 2-Keto-D-gluconate-6-phosphate,a new intermediate in the carbohydrate me-tabolism of Aerobacter cloacae. Enzymologia17:55-68.

DE LEY, J., AND S. VERHOFSTEDE. 1955. The me-tabolism of 2-keto-D-gluconate-6-phosphate.Naturwissenschaften 42:584.

EISENBERG, R., AND D. J. O'KANE. 1963. Theeffect of DPN on the products formed fromthe galactose fermentation. Bacteriol. Proc.,p. 116.

ELSDEN, S. R., AND J. L. PEEL. 1958. MIetabolismof carbohydrates and related compounds.Ann. Rev. Microbiol. 12:145-202.

ENTNER, N., AND I. DOUDOROFF. 1952. Glucoseand gluconic acid oxidation of Pseudomnonassaccharophila. J. Biol. Chem. 196:853-862.

FRAMPTON, E. W., AND W. A. WOOD. 1961. Carbo-hydrate oxidation by Pseiudomnonasfluorescens.VI. Conversion of 2-keto-6-phosphogluconateto pyruvate. J. Biol. Chem. 236:2571-2577.

8501 J. BACTERIOL.

on May 24, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

2-KETOGLUCONATE FERMENTATION BY S. FAECALIS

FUKUYAMA, T. T., AND D. J. O'KANE. 1962. Galac-tose metabolism. I. Pathway of carbon in fer-mentation by Streptococcus faecalis. J. Bac-teriol. 84:793-796.

JUNI, E., AND G. A. HEYM. 1962. Determination ofcarbonyl acids formed upon periodate oxida-tion. I. Assay procedure. Anal. Biochem.4:143-158.

NORTON, J. E., G. S. BULMER, AND J. R. SOKATCH.1963. Oxidation of D-alanine by cell mem-branes of Pseudomnonas aeruginosa. Biochim.Biophys. Acta 78:136-147.

OHLE, H., AND G. BEREND. 1927. 2-Keto-glucon-saure. Ber. Deut. Chem. Ges. 60:1159-1167.

O'KANE, D. J., AND I. C. GUNSALUS. 1948. Pyruvicacid metabolism. A factor required for oxida-tion by Streptococcus faecalis. J. Bacteriol.56:499-506.

OSBURN, D. L., H. G. WOOD, AND C. H. WERKMAN.1933. Determination of formic, acetic andpropionic acids in a mixture. Ind. Eng. Chem.Anal. Ed. 5:247-250.

PHARES, E. F. 1951. Degradation of labeled propi-onic and acetic acids. Arch. Biochem. Bio-phys. 33:173-178.

SOKATCH, J. R. 1960. Ribose biosynthesis by

Streptococcus faecalis. Arch. Biochem. 91:240-246.

SOKATCH, J. R. 1962. Phosphorylation of sedohep-tulose 7-phosphate by Streptococcus faecalis.Arch. Biochem. 99:401-408.

SOKATCH, J. R., AND I. C. GUNSALUS. 1957. Aldonicacid metabolism. I. Pathway of carbon in aninducible gluconate fermentation by Strepto-coccus faecalis. J. Bacteriol. 73:452-460.

STEELE, R., AND T. SFORTUNATO. 1949. Techniquesin the use of C14. Brookhaven Natl. Lab. Bull.BNL-T-6.

STEELE, R. H., A. G. C. WHITE, AND W. A. PIERCE,JR. 1954. The fermentation of galactose byStreptococcus pyogenes. J. Bacteriol. 67:86-89.

STOKES, F. N., AND J. J. R. CAMPBELL. 1951. Theoxidation of glucose and gluconic acid bydried cells of Pseudomonas aeruginosa. Arch.Biochem. Biophys. 30:121-125.

STUBBS, J. J., L. B. LOCKWOOD, E. T. ROE, B.TABENKIN, AND G. E. WARD. 1940. Ketoglu-conic acids from glucose. Ind. Eng. Chem.32:1626-1631.

WOOD, A. J., AN) I. C. GUTNSALUS. 1942. The pro-duction of active resting cells of streptococci.J. Bacteriol. 4-4:333-341.

851VOL. 87, 1964

on May 24, 2018 by guest

http://jb.asm.org/

Dow

nloaded from