Two Pathways of Glutamate Fermentation Bacteria · pathway appears to be used only by species of Clostridium, whereas the hydroxyglutarate pathway is used byrepresentatives ofseveral

JOURNAL OF BACrERIOLOGY, Mar. 1974, p. 1248-1260Copyright O 1974 American Society for Microbiology

Vol. 117, No. 3Printed in U.S.A.

Two Pathways of Glutamate Fermentation by AnaerobicBacteria

WOLFGANG BUCKEL' AND H. A. BARKER

Department of Biochemistry, University of California, Berkeley, California 94720

Received for publication 22 October 1973

Two pathways are involved in the fermentation of glutamate to acetate,butyrate, carbon dioxide, and ammonia-the methylaspartate and the hydroxy-glutarate pathways which are used by Clostridium tetanomorphum and Pep-tococcus aerogenes, respectively. Although these pathways give rise to the sameproducts, they are easily distinguished by different labeling patterns of thebutyrate when [4- "4C Iglutamate is used as substrate. Schmidt degradation of theradioactive butyrate from C. tetanomorphum yielded equally labeled propionateand carbon dioxide, whereas nearly all the radioactivity of the butyrate from P.aerogenes was recovered in the corresponding propionate. This procedure wasused as a test for the pathway of glutamate fermentation by 15 strains (9 species)of anaerobic bacteria. The labeling patterns of the butyrate indicate thatglutamate is fermented via the methylaspartate pathway by C. tetani, C.cochlearium, and C. saccarobutyricum, and via the hydroxyglutarate pathwayby Acidaminococcus fermentans, C. microsporum, Fusobacterium nucleatum,and F. fusiformis. Enzymes specific for each pathway were assayed in crudeextracts of the above organisms. 3-Methylaspartase was found only in clostridiawhich use the methylaspartate pathway, including Clostridium SB4 and C.sticklandii, which probably degrade glutamate to acetate and carbon dioxide byusing a second amino acid as hydrogen acceptor. High levels of 2-hydroxygluta-rate dehydrogenase were found exclusively in organisms that use the hydroxy-glutarate pathway. The data indicate that only two pathways are involved in thefermentation of glutamate by the bacteria analyzed. The methylaspartatepathway appears to be used only by species of Clostridium, whereas thehydroxyglutarate pathway is used by representatives of several genera.

Several strictly anaerobic bacteria fermentglutamate to ammonia, carbon dioxide, acetate,butyrate, and small amounts of hydrogen (7).The pathway of this fermentation (Fig. 1) wasfirst elucidated in Clostridium tetanomorphum(7).

In the initial coenzyme B,2-dependent reac-tion, the linear carbon chain of glutamate isrearranged to the branched chain of 3-methylas-partate. Elimination of ammonia plus additionof water yields citramalate, which subsequentlyis cleaved to acetate and pyruvate. The reduc-ing equivalents produced by the subsequentoxidative decarboxylation of pyruvate to CO2and acetyl coenzyme A (CoA) are used either tosynthesize butyryl CoA from 2 mol of acetylCoA or to form molecular hydrogen. Finally,adenosine triphosphate (ATP) is generatedfrom the thio esters, probably via the acylphos-phates.

'Present address: Fachbereich Biologie, Universitiit Re-gensburg, 84 Regensburg, Germany.

Peptococcus aerogenes, which ferments gluta-mate to the same products as C. tetanomor-phum, uses another pathway. This was con-cluded by Whiteley (56) from the inability ofcrude extracts of P. aerogenes to metabolizemesaconate or citramalate. On the fermenta-tion of "C-labeled glutamates with cell suspen-sions of P. aerogenes, Horler et al. (26) foundthat the linear carbon chain of glutamate re-mains unchanged in butyrate. In addition,these authors (25) were able to convert gluta-mate to glutaconate by using cells of the sameorganism. In crude extracts, however, 2-hydroxy-glutarate was formed from glutamate (28).Recently, the enzymes catalyzing these reac-tions, glutamate dehydrogenase (EC 1.4.1.2)and 2-hydroxyglutarate dehydrogenase (EC1.1..x), were shown to be present at high levelsin crude extracts of P. aerogenes and Acidami-nococcus fermentans (29, 30, 33). The fate of2-hydroxyglutarate remains unclear. Probablyit is converted to crotonyl CoA via glutaconyl

1248

on January 16, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

GLUTAMATE FERMENTATION BY ANAEROBES

HOO1C 3CH2 5COOH CH4cH2-3CH2-4COOH\/ \4/2C 4CH2 butyrate

H NH2S-glutamate \

3CH-1COSCoA 34---(H) I~~~CH3~~~HH2 acetyl- CoA Ii

H. 3CH3 r i \CCC

NS33eH3lsott

Hc c -OH H CO C

H/

COOH CH2 \COOHmesaconate S- citramolate

FIG. 1. Pathway of glutamate fermentation by C.tetanomorphum. Compounds in boxes are final prod-ucts.

CoA. The mechanism of butyrate and acetateformation from crotonyl CoA may be similar tothat found in Clostridium kluyveri (6). Thesesuggestions and facts about this pathway aresummarized in Fig. 2, which in outline wasproposed by Barker in 1937 (4).This paper is concerned with the occurrence

of these two pathways in other glutamate-fer-menting bacteria. The results have a bearing onthe taxonomic and evolutionary relations ofthese organisms.

MATERIALS AND METHODSMedia. Three different media were used. Medium

A is the yeast extract-glutamate used for cultivatingC. tetanomorphum (13). Medium B is thiolbroth(Difco) supplemented with 1% sodium s-glutamate.(We are using the R/s system of Cahn-Ingold-Prelog[see reference 2] for designating the configuration ofoptically active compounds because some of thecompounds under consideration, such as (3S)- [4-'4C]citrate, cannot be identified adequately in anyother way. For many organic compounds with a singleasymmetric center, the designations R and s areequivalent to the more commonly used D and L,respectively.) It is rich enough in peptone to supportgrowth of non-glutamate-fermenting organisms aswell. Later it was replaced by medium C, which issimilar to that described by Lerud and Whiteley (33)for P. aerogenes. It consists of 0.5% sodium s-gluta-mate, 0.5%; yeast extract (Difco), 1%; peptone(Difco), 1%; 1.0 M potassium phosphate (pH 7.4), 2%(vol/vol); and sodium thioglycolate, 0.1%.

Culture techniques. Usually, bacteria were grownat 37 C in semisolid media (0.2% agar, 10 to 15 ml) inscrew-cap tubes (160 by 16 mm), or in liquid media(25 to 2,000 ml) in volumetric flasks filled up to theneck. For isolation or purity controls, petri dishes

containing 50 ml of solid media (2% agar) and coveredwith clay dishes were used. They were incubated in aBrewer Anaerobic Jar (BBL) under an atmosphere of84% argon, 11% hydrogen, and 5% carbon dioxide(vol/vol/vol). For the same purpose, shake cultures ina series of dilutions were sometimes used; oxygen wasremoved by means of alkaline pyrogallol. Tests foraerobic growth were performed on petri dishes con-taining a medium of 1% peptone, 1% yeast extract, 1%glucose, and 2% agar.

Bacteria. In Table 1, all organisms which wereused in this work are listed together with their sourceand the medium that supported good growth. Whenlittle or no growth on medium A was observed,medium B or C was used. Table 1 also gives theliterature reference where the strain was describedand/or fermentation of glutamate was indicated. Thelast column indicates ability to ferment glutamateunder the conditions described below.

Test of glutamate fermentation. Most of thestrains were tested manometrically with washed-cellsuspensions as described for the fermentation oflabeled glutamates. Gas formation after the additionof glutamate was a good indication of glutamateutilization. When this test failed, the growth mediumwas checked for high levels of butyrate by gas-liquidchromatography.The physiological characteristics of the bacteria

were tested by use of the media described by Spray (4;Table 2).

Special remarks on some strains. Strains Bi and

HOO'C 3CH2 5COOH\ / \ /2c/ 4CH2

H NH2S -glutomate

V NAD + H20

lNADH+ NH3

HOOC CH2 COOH\ /\ /

C CH2II0

2- ketoglutorate

g NADH

'NAD

HOOC CH2 COOH

ICHj3 CH2 ZCH2 COO

butyrote

2 - - (H)

4CH3-3COOH2CH 3-COOHacetate

C CH2 C 4CH33CH= 2CH- COSCoA

H OH crotonyl- CoA

R, 2- hydroxyglutorate

H

H005C 3C ICOSCoA

4CH2 2c

"Ox H

Iglutaconate g lutoconyl - CoAL

FIG. 2. Postulated pathway of glutamate fermen-tation by P. aerogenes. Compounds in boxes are finalproducts. Neither glutaconyl CoA nor crotonyl CoAhas been identified as an intermediate in the fermen-tation.

1249VOL. 117, 1974

on January 16, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

BUCKEL AND BARKER

C141 were isolated from garden soil with the aid ofenrichment cultures containing glutamate as themain source of carbon and energy (5). Both strainsproduce terminal spherical spores that cause conspic-uous swelling of the cell. Vegetative cells of strainC141 are noticeably thinner than those of strain Bi.The colonies of the two strains also differ signifi-cantly. The main physiological difference betweenthese strains is the ability of strain Bi to fermentglucose. According to the Spray tests, strains Bi andC141 resemble C. tetanomorphum and C. cochleari-um, respectively.

P. aerogenes had to be transferred at least monthly,otherwise it lost its ability to ferment glutamate. Themorphology of this altered or contaminant strain wassimilar to P. aerogenes, but the Spray tests and theenzyme content turned out to be very different(Tables 2 and 7).

A. fermentans VR4 ATCC 25085 fermented lactose,sucrose, and glucose, in contradiction to the report ofRogosa (39). The possibility that this was caused by acontaminant was not excluded.Enzymes and chemicals. Acetate kinase (EC

2.7.2.1), phosphotransacetylase (EC 2.3.1.8), malatedehydrogenase (EC 1.1.1.37), si-citrate synthase (EC4.1.3.7) from pig heart, isocitrate dehydrogenase (EC

1.1.1.42), and glutamate dehydrogenase (EC 1.4.1.3)were purchased from Boehringer Mannheim, Ger-many. D- and L-2-hydroxyglutarates (i.e., R- ands-enantiomers, respectively [121) were obtained fromSigma Chemical Co., and [2- "IC Jacetate, R, S- [1- 4C ]-and R, s- [5-4C ]glutamates were from New EnglandNuclear Corp., Boston, Mass. 2s, 3s-3-methylaspar-tate was prepared enzymatically by the method ofBarker et al. (9). Acetyl CoA and butyryl CoA weresynthesized by the method of Simon and Shemin (43).

Aconitase (EC 4.2.1.3) was prepared as describedby Siebert (42) and assayed by measuring the increaseof absorbance at 240 nm (e = 4.88 mM-' cm-')resulting from the conversion of citrate to aconitate.The test assay solution (22) contained 100 mMtris(hydroxymethyl)aminomethane (Tris)-hydrochlo-ride (pH 7.6), 100 mM sodium citrate, and enzyme.Under these conditions, the aconitase preparation hadan activity of 8 U/ml.

General analytical methods. Paper chromatogra-phy was done with two systems. Solvent I (ascending)consisted of 1-butanol-acetic acid-water (60:15:25,vol/vol/vol). Solvent II (descending) consisted of phe-nol-water (4: 1, wt/wt).

Paper ionophoresis was performed on the apparatusof Crestfield and Allen (20) with 50 mM sodium

TABLE 1. Bacteria investigated

Good GlutamateBacteria Source References growth on fermen-

mediuma tation

Clostridium tetanomorphum Hl This laboratory 4, 7, 57 A +Clostridium SB4 This laboratory 19 C (±)C. sticklandii This laboratory 48, 49 B (+)C. tetanomorphum NCTC 2909 G. C. Mead, Food Research Inst., 4, 7, 57 A +

Norwich, EnglandC. tetani NCTC 5404 G. C. Mead 17 A +C. cochlearium ATCC 17787 G. C. Mead 5 A +Clostridium GL 12A (resembling G. C. Mead _b A +

C. cochlearium)C. microsporum HB25, formerly G. C. Mead A +named Fusiformis biacutus

C. saccharobutyricum Inst. Pasteur, Paris 18 A +C. carnofoetidum Inst. Pasteur, Lille 11 CC. perfringens ATCC 10543 ATCC 58 CAcidaminococcus fermentans ATCC 33, 39 C +VR4 ATCC 25085

Peptococcus aerogenes ATCC 14963 ATCC and G. W. S. Westlake 26, 56 C +Fusobacterium fusiforme 2388 W. E. Moore, Virginia Polytech- - C +

nic Inst., and State Univ.,Blacksburg, Va.

F. fusiforme 4351 W. E. Moore - C +F. nucleatum 4355 W. E. Moore 27 C +F. nucleatum 4357 W. E. Moore 27 C +Clostridium Bi (resembling C. Isolated from soil - A +tetanomorphum)

Clostridium C141 (resembling C. Isolated from soil _ A +cochlearium)

a See Materials and Methods.bGlutamate fermentation was reported by G. C. Mead (personal communication).

1250 J. BACTERIOL.

on January 16, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

GLUTAMATE FERMENTATION BY ANAEROBES

formate (pH 3.8) as buffer. Usually, a potentialgradient of 35 V/cm was applied for 90 min. Whatmanno. 1 paper was used. Radioautographs were madewith Kodak X-ray film by exposure for 1 to 3 days,depending on the level of radioactivity. Citrate was

located on chromatograms by spraying the paper with0.1% KMnO4 in water. Amino acids were located byreaction with ninhydrin.

For quantitative determinations of a single aminoacid, the method of Moore and Stein was used (36).Ammonia was first removed by adding a slight excess

of 2 M K,COS and evaporating the solution to drynessunder reduced pressure in a desiccator containingconcentrated sulfuric acid. Finally, the sample was

dissolved in an amount of 5 N HCl equivalent to theadded alkali. In some cases, determinations were

performed with a Beckman/Spinco amino acid ana-lyzer (model 120; column, 0.9 by 55 cm; citrate buffer,pH 3.22). When this apparatus was used on a prepara-

tive scale, fractions of 2.3 ml (2 min) were collected.The citrate buffer was removed by the desalting

procedure of Dreze et al. (21) on Dowex 2. Thismethod also was used for purification of labeledglutamate.

Radioactive carbon was measured with a PackardTri-Carb scintillation spectrophotometer (model3214). The scintillation fluids of Bray (15) and Gold-mark and Linn (23) were used. Both systems (10 ml offluid plus 1 ml of aqueous sample) had the same

efficiency. Corrections were made for the lower effi-ciency of samples containing 14CO, in ethanolamine.When indicated, mixtures of fatty acids were ana-

lyzed by gas-liquid chromatography. The dried so-

dium salts were dissolved by a threefold excess of 18M phosphoric acid, followed by peroxide-free ether tomake the solutions 5 to 40 mM in acetic and propionicacids, and 2 to 30 mM in isobutyric and isovalericacids. Samples of about 2 uliters were injected at130 C into a stainless-steel column (1/8 inch by 7 feet[about 3.17 mm by 2.13 ml) containing 6% FFAP on

Chrom W 100/120 at 90 C; N2 flow was 24 ml/min.The separated acids were detected by flame ioniza-

TABLE 2. Physiological characteristics of the bacteria

Ml-rnGelatine Fermentation of NO2- IdBacterium Milk-iron Gflique- H2S G ruta-nfromBacterium ~test' faction produced Lactose Sucrose Glucose Glt- NO3 produced

mate

Clostridium tetanomor- V - + + + +phum HI

C. tetanomorphum 2909 V - + _ _ + + - +Clostridium Bl V - _ _ + + - +

C. cochlearium 17787 V - + - + - +Clostridium GL 12A V + - - _ - + -

Clostridium C141 V + _ - + -

C. saccharobutyricum V _ _ _ _ + + - +

C. sticklandii V _ + _ - + (±) _

Clostridium SB4 IV + + _ _ - (+) _

C. perfringens I + + + + + - +

C. carnofoetidum IV + + _ _ + _ _

C. microsporum V - _ + _ + + _

Peptococcus aerogenes II - + _ _ - + + +Unidentified coccus II - _ + + + - +

Acidaminococcus fer- II - + + + + + _mentans

Fusobacterium fusiforme V + + _ _ + + +2388

F. nucleatum 4357 V - +_ + + +

a I, Active gaseous fermentation with early coagulation; II, inactive gaseous fermentation with coagulationdelayed; IV, inactive gaseous fermentation with digestion and blackening; V, unchanged. Medium for groupreactions of R. S. Spray (44).

1251VOL. 117, 1974

on January 16, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

BUCKEL AND BARKER

tion; H, flow was 20 ml/min.Ammonia was determined by steam distillation

and titration with hydrochloric acid (3).Synthesis of s- [4-"1C]glutamate from [2-

"CC]acetate. The synthesis of s- [4- "Ciglutamatewas performed enzymatically in two steps via s-[4"C]citrate. For the first step, the method of Berg-meyer and Moellering (13), modified by Lenz et al.(32), was used according to equation 1.

[2-"Cjacetate + s-malate + ATP + NAD

s-[4-"C]citrate + NADH + ADP + P, (1)

A reaction solution (1.5 ml) containing 1.5 mM[2-"C]acetate (1.3 x 108 counts/min), 10 mMmalate, 100 mM Tris-chloride (pH 8.5), 4 mM nico-tinamide adenine dinucleotide (NAD), 4 mM ATP, 6mM MgCl,, 2 mM ethylenediaminetetraacetic acid(EDTA), 3 mg of CoA, 20 Mg of si-citrate synthase(prefix si indicates that the enzyme catalyzes addi-tion of an acetate group to the si-side of oxaloacetate[2] and therefore synthesizes (3s)-[4- "C]citrate from[2-"4C]acetyl CoA and oxaloacetate), 20 jug of malatedehydrogenase, 10 ug of phosphotransacetylase, and30 Mg of acetate kinase formed 24 Mmol of s- [4- 4C ]ci-trate in an 8-h incubation period. The product was

separated from residual acetate and malate by means

of a Dowex 1-X2 formate column; citrate was elutedwith 4 N formic acid after removal of other compo-

nents with 1 N formic acid. The product contained a

small amount of adenosine diphosphate (ADP) or

ATP (about 0.4 Mmol/2.4 gmol of citrate).In the second step, citrate was converted to gluta-

mate according to equation 2.

s-[4-"4C]citrate + NH4+

s-[4-"4C]glutamate + CO, (2)

The reaction solution (1.0 ml) contained 100 mMTris-chloride (pH 7.6), 1.0 mM MgCl,, 0.5 mM NADphosphate (NADP), 4 mM NH4Cl, 200 Mg of isocitricdehydrogenase, 200 kig of glutamic dehydrogenase, 0.4U of aconitase, and 2.4 mM s-[4- "C]citrate (1.28 x

108 counts/min). A 20-min incubation period gave an

overall 79% yield of s-[4-"4C]glutamate (1.04 x 10"counts/min) based on the starting acetate. The prod-uct was isolated by adsorption and elution fromDowex-2 (21). Paper chromatograph (two solvents)and paper ionophoresis showed that more than 98% ofthe "4C was present in glutamate and less than 2% was

present in pyrrolidone carboxylate.Purification of commercial R, s- [I-_4C- and R, S-

[5-"4C]glutamates. Paper chromatography (solvent I)indicated that both samples were contaminated withseveral minor radioactive compounds. [1-_4C]gluta-mate (10 Amol, 120 MCi) was purified by using an

amino acid analyzer. The glutamate obtained afterdesalting was pure as analyzed by paper chromatog-raphy (solvent I). [5-14C ]glutamate (50 MCi, 18.7Mmol) was purified by the Dowex 2 procedure (21) andnot analyzed further.

Test of the pathway of glutamate fermentation

with "4C-labeled glutamates. Fermentations wereperformed either with growing cultures or washed cellsuspensions. The cell suspensions were prepared bysuspending washed, stationary-phase cells, at 50 to100 times the concentration in the growth medium, in0,-free 50 mM potassium phosphate (pH 7.0) con-taining 0.1% sodium thioglycolate. From 0.1 to 2.0 mlof cell suspension was used to ferment 40 to 100 Mmolof sodium s- [14C Iglutamate (about 10" counts/min) ina total volume of 2.2 ml under 0,-free argon; gasevolution was followed until it ceased (30 min to 3 h)in a Warburg apparatus at 37 C.

Total volatile fatty acids obtained with bothmethods were estimated by titration (phenol red)after steam distillation. From a sample of the distil-late, the total amount of radioactivity converted tofatty acids was determined. The neutralized solutionwas evaporated in vacuo, and the fatty acids wereseparated by column chromatography (41) as follows.The red residue was dissolved in 0.5 ml of acetone-2-butanone-water (2:1:9, vol/vol/vol) and transferredto a column (75 by 0.9 cm) of Amberlite IRC 50. Onelution by the same solvent with a pressure differenceof about 100 cm of water, a yellow band of phenol redmoved through the white column. When this bandapproached the end of the column (4 to 6 h), fractionsof about 3 ml (40 drops) were collected; acetic andbutyric acids eluted in fractions 9 to 11 and 18 to 23,respectively. The specific radioactivities of the sepa-rated acetic and butyric acids were determined bysteam distillation followed by counting and titrationof samples.The Schmidt degradation of highly labeled butyric

acid was done by the method of Mosbach et al. (37).Carbon dioxide was trapped in 2.5 ml of ethanola-mine-methyl cellosolve (4:3, vol/vol) instead ofNaOH. To precipitate the CO2 as BaCO,, samples of1.5 ml were mixed with 1.0 M BaCl, and 10 ml of 0.1N NaOH and heated for 10 min in a boiling-waterbath. Propylamine was isolated and oxidized to pro-pionic acid (37), which was isolated by steam distilla-tion and separated from its by-product acetic acid(about 10%) by chromaography on Amberlite IRC 50(41). The relative amounts of "C in both acids arelisted in Tables 3 and 4. One sample of propionic acidwas degraded further to carbon dioxide and aceticacid (38).

Preparation of cell-free extracts. Stationary-phase cells were washed twice with standard bufferand frozen at -20 C. They were then thawed with 2ml of standard buffer per g of packed cells andsonically treated (four times for 30 s) with a smallsonic converter probe (Branson Sonifier) at maximalsound. The resulting homogenate was centrifuged for30 min at 50,000 x g. The clear supernatant waseither stored at -20 C or used immediately. Proteinwas determined by the method of Lowry et al. (34),with crystalline bovine serum albumin as standard.

Determination of enzymatic activities. 3-Methylaspartase (EC 4.3.1.2) was measured by themethod of Barker et al. (9), with 2s,3s-3-methylas-partate as substrate. 2-Hydroxyglutarate dehydrogen-ase and glutamate dehydrogenase (EC 1.4.1.2) were

1252 J. BACTERIOL.

on January 16, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

GLUTAMATE FERMENTATION BY ANAEROBES

measured as described by Lerud and Whiteley (33). Inthe 2-hydroxyglutarate dehydrogenase assay, R- ors-2-hydroxyglutarate was used as substrate.

For determination of the stereochemical purity of2-hydroxyglutarate, a 1.0-ml incubation solution wasused, containing: 10 mM potassium phosphate (pH7.4), 100 mM ammonium sulfate, 2 mM NAD, 20 mMR- or s-2-hydroxyglutarate, and crude extract of A.fermentans (300 yg of protein). After 75 min at 25 C,0.1 ml of 15% trichloroacetic acid was added. Afraction of the supernatant was used to determineglutamate with an amino acid analyzer. R-2-Hydroxy-glutarate gave 10.5 gmol (53%) of glutamate, whereasthe -s-enantiomer gave 0.23 umol (1%) of glutamate.Similar results were obtained in the fermentation ofthe two isomers of 2-hydroxyglutarate with washedcells of A. fermentans (see Table 5).

Glutamate mutase (EC 5.4.99.1) activity was mea-sured by both the spectrophotometric and anaerobicassays (8). In the latter assay, the ratio of glutamateto methylaspartate was used as a measure of mutaseactivity. Acetate and butyrate kinases (EC 2.7.2.1and EC 2.7.2.7) (52) were assayed by the method ofRose (40), with either acetate or butyrate as substrateat a concentration of 700 mM. The samples wereincubated at 25 C for 10 min. Phosphotransacetylase(EC 2.3.1.8) and phosphotransbutyrylase (EC2.3.1.19) (53) were measured by arsenolysis of thecorresponding acyl-CoA derivative at 232 nrn (46). Aquartz cuvette contained 50 mM Tris-hydrochloride(pH 7.4), 10 mM (NH4)2S04, 1 mM dithioerythritol,0.2 mM acetyl or butyryl CoA, and crude extract in atotal volume of 0.9 ml. The reaction was initiated bythe addition of 0.1 ml of 0.5 M sodium arsenate. Theassay of the CoA-transferase (EC 2.8.3.1) (45) is basedon the determination of acetyl CoA, which is pro-duced from butyryl CoA and acetate (10). The reac-tion mixture contained 100 mM Tris-hydrochloride(pH 8.0), 0.2 mM butyryl CoA, 100 mM sodiumacetate, and crude extract in a volume of 0.5 ml. Afterincubation for 20 min at 25 C, 50 uliters of 5 N HCIwas added. The supernatant was passed through acolumn (0.5 by 4 cm) with Dowex 50 H+ (8x, 200 to400 mesh). The eluate was lyophilized and dissolvedin 1.0 ml of water. A fraction (0.5 ml) was used for thedetermination of acetyl CoA (16). Controls were runwithout acetate or with acetyl CoA instead of butyrylCoA. The values obtained are corrected for losses ofacetyl CoA in this procedure (Table 4).

RESULTS AND DISCUSSION

Experimental approach. A good method todifferentiate between the two known pathwaysshould be the fermentation of "C-labeled gluta-mates. According to Fig. 1 and 2, carbon 1 ofglutamate should be incorporated into bothacetate and butyrate obtained with P.aerogenes, but only into acetate on fermenta-tion by C. tetanomorphum. With [4-14C]gluta-mate as substrate, C. tetanomorphum shouldproduce butyrate with a specific activity twice

as high as that formed by P. aerogenes. How-ever, the specific activities may be altered bya variety of secondary factors, such as dilutionof the labeled glutamate by glutamate fromyeast extract or peptone, dilution of the labeledfatty acids by endogenous fatty acids or byfermentation of amino acids other than gluta-mate present in complex media, and conversionof acetate to butyrate. Therefore, another morereliable test was used which was independentof the specific activities of the products. Thistest is based on the different labeling pattemsof the butyrate obtained with [4- "C] glutamate.The butyrate formed by C. tetanomorphum willbe labeled in carbons 1 and 3 since it is derivedfrom two identical acetyl units mainly derivedfrom glutamate carbons 3 and 4 (Fig. 1). Thebutyrate formed by P. aerogenes, on the otherhand, will be labeled only in carbon 4 since it isderived from glutamate carbons 1 to 4 withoutrearrangement. Schmidt degradation of thesebutyrates will give a clear indication of the path-way involved since propionate (carbons 2 to 4)and CO2 (carbon 1) obtained from the butyrateformed by C. tetanomorphum should be labeledequally, whereas the total radioactivity of thebutyrate from P. aerogenes should be present inpropionate after degradation and the CO2should be unlabeled.Test of the pathway of glutamate

fermentation. The reliability of the test waschecked first with the two organisms for whichthe labeling patterns had been described-C.tetanomorphum (54) and P. aerogenes (26, 35).The results agree well with those expected(Table 3). Fermentation of [4- 4C ]glutamate byC. tetanomorphum yielded butyrate which wasequally labeled at carbon 1 and in carbons 2 to4. On the other hand, when the same reactionwas performed with P. aerogenes, butyrate wasobtained which was almost exclusively labeledin carbons 2 to 4. These data also prove that theenzymatic synthesis of s- [14- "C ]glutamatefrom [2-1C ]acetate gave the expected product.When this method was applied to butyrate

formed by the other glutamate-fermenting bac-teria, only two labeling patterns were found,identical to those of C. tetanomorphum and P.aerogenes, respectively (Tables 3 and 4). Theorganisms can therefore be divided into twogroups, the C. tetanomorphum group and the P.aerogenes group. To get a more complete pic-ture, [1- 14C ]- and [5- 14C ]glutamates were usedas additional substrates. In agreement with Fig.2, over 90% of the radioactivity of the butyratewas found in the carboxyl group when [1-14C ]glutamate was fermented by organisms

1253VOL. 117,1974

on January 16, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

of the P. aerogenes group. With [5-1C Jgluta- (Table 3). This can be explained by activationmate, almost no radioactivity was incorpor- of the free acetate which is generated in theated into the fatty acids, as required by Fig. 1 citramalate pyruvate-lyase reaction (ECand 2. 4.1.3.22) (Fig. 1). Enzymes which catalyze thisButyrate synthesis from acetate. When [1- activation are either acetate kinase and phos-

"IC glutamate was fermented by organisms of photransacetylase, or CoA-transferase.the C. tetanomorphum group, a large participa- Therefore, Clostridium Bi and C.tion of carbon 1 of glutamate in butyrate tetanomorphum Hi, which incorporated differ-synthesis was observed with some species ent amounts of carbon 1 of glutamate into

TABLE 3. Fermentation of ['4Clglutamate by cell suspensionsa

Distribution of "C inGlutamate added Products formed butyrateb (%)

Organism Posi- Counts Acetate Butyrate '4 inOrganism < P LFC C~~~~~~~voatle| C2, -3, C3

tion pmol per min Relative Relative vacidsd c C 2an -4, an3-of "C per nmol mol sp actc Amol sp act" a (%) l

C. tetanomorphum Hi 1 40 21.3 51 67 18 14 98C. tetanomorphum Hi 4 40 23.3 54 21 16 138 94 50 49 44C. tetanomorphum Hl 5 40 50.6 58e 0.2C. tetanomorphum 2909 1 40 21.3 54 57 15 75 90C. tetanomorphum 2909 4 40 22.2 53 34 14 122 89 52 55 51Clostridium B1 1 40 21.4 54 51 18 46 92Clostridium B1 4 40 23.6 54 32 17 89 89 48 50 48C. cochlearium 1 40 21.4 45 80 >7 12 92C. cochlearium 4 40 23.6 44 17 14 159 74 48 50 44Clostridium GL 12A 1 40 21.4 49 69 12 25 92Clostridium GL 12A 4 40 23.6 48 25 11 150 83 47 50 46Clostridium 141 1 40 21.4 43 58 6 24 91Clostridium 141 4 40 22.0 53 29 7 137 80 37 49C. tetani 1 40 21.3 48 69 14 12 83C. tetani 4 40 22.2 52 20 14 131 78 49 50 47C. saccharobutyricum 1 40 19.9 49 49 12 36 95 55 50C. saccharobutyricum 4 40 19.3 52 28 15 92 92 51 51C. saccharobutyricum 5 100 18.5 153e 0.8P. aerogenes' 1 50 17.8 48 25 14 92 62 92 0.6P. aerogenes' 4 50 17.5 54 36 16 99 89 0.2 85A. fermentansg 1 40 21.3 40 38 16 117 93 70 1.4 1.3A. fermentansg 4 40 23.4 40 46 15 113 98 0.8 93 97A. fermentansg 5 40 50.6 58e 0.2F. nucleatum 4355' 1 40 19.9 41 18 26 68 69 97 3F. nucleatum 4355' 4 40 19.3 42 33 24 86 102 0.1 97F. nucleatum 4357' 1 40 19.9 33 22 11 108 76 94 2F. nucleatum 4357f 4 40 19.3 32 43 13 102 88 0.2 100F. nucleatum 4357' 5 100 18.5 72" 0.3C. microsporum 1 40 21.3 46 31 17 110 87 92 2C. microsporum 4 40 22.0 47 41 17 100 93 0.4 100 90

a The bacteria were grown on medium A unless otherwise indicated. Washed cells from 25 to 100 ml of growthmedium were allowed to ferment the indicated quantity of ["C]glutamate in 2.2 ml of 50 mM potassiumphosphate buffer containing 0.1% sodium thioglycolate under an argon atmosphere. Gas evolution was followedin a Warburg respirometer at 37 C until it ceased (30 min to several hours). The volatile fatty acids were thenseparated and examined by methods described in the text.

'4C levels in Cl, C2, -3, and -4, and C3 and -4 were obtained from CO2, propionate, and acetate formed bychemical degradation of butyrate (see Materials and Methods).

c Specific radioactivity (counts per minute per nanomole) of glutamate = 100%.d Percentage of ithe "C in the added glutamate.eTotal volatile acids.' Grown on medium B.9 Grown on medium C.

1254 BUCKEL AND BARKER J. BACTERIOL.

on January 16, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

GLUTAMATE FERMENTATION BY ANAEROBES

butyrate (46 and 14%, respectively; Table 3),also should contain different levels of at leastone of these enzymes. This was found foracetate kinase (0.1 U/mg versus <0.01 U/mg,respectively; Table 5), whereas the activities ofthe CoA-transferases were too low to be impor-tant. The levels of butyrate-activating enzymes,however, are high in both organisms.

If butyrate is synthesized from [1- 4C]gluta-mate via [1-14C]acetate, it should be labeledequally at carbon 1 and carbons 2 to 4. This wasconfirmed by degradation of the butyrate fromC. saccharobutyricum (Table 3). By a closerexamination of the labeling patterns of thebutyrates from the P. aerogenes group, up to 6%

of carbon 1 of glutamate was found at carbons 2to 4 of butyrate (Tables 3 and 4). A two-stepdegradation of the butyrate obtained from thefermentation of [4- '4C ]glutamate by C.microsporum indicated that this label is causedby a small synthesis of butyrate from C2 units.Although the main label (90%) was presented inthe acetate (carbons 3 and 4 of butyrate),formed by degradation of the propionate, 8%was found in the CO2 (carbon 2 of butyrate).Presumably a similar small synthesis of butyr-ate from C2 units also occurs in other membersof the P. aerogenes group.Enzymes specific for each pathway. The

tracer experiments indicate that only the two

TABLE 4. Fermentation of ['4Clglutamates by growing cultures"

Distribution of "4CGlutamate added Products formed

Organism Acetate Butyrate "4C inPosition Counts vltl 2 3Pofition mol per mm Relative RelativeX acidse C1 and -

per umol umol sp actc Mmol sp actc (%)

F. fusiforme 2388 1 200 4.00 419 10 204 33 62 97 4F. fusiforme 2388 4 200 3.70 422 21 208 49 106 0.2 91F. fusiforme 4351 1 400 0.85 390 10 261 29 36 95 6F. fusiforme 4351 4 400 0.86 360 22 326 56 89 0.2 98Clostridium SB4 1 150 2.90 307 14 35 3 38Clostridium SB4 4 150 3.20 277 19 22 13 56C. sticklandii 1 150 2.90 11C. carnofoetidum 1 490 1.60 310e 1C. carnofoetidum 4 490 1.50 320' 1

a F. fusiforme and C. carnofoetidum were grown in 10 ml of modified medium B; Clostridium SB4 and C.sticklandii were grown in 5 ml of modified medium C. Both media were modified to contain the indicatedquantity of "C-labeled glutamate. Incubation time varied from 2 to 10 days at 37 C. The volatile fatty acidswere then separated and examined by methods described in the text.

b-d See respective footnotes in Table 3.e Total volatile acids.

TABLE 5. Transferase activities of two strains of the C. tetanomorphum groupa

Activities (umol per min per mg of protein)

Organism Fatty acid kinases' Phosphotransacylasesc CoA-transferase'Butrae Acetyl CoA Butyryl CoA Butyryl CoAAcetate Butyrate arsenatee arsenatee acetatee

C. tetanomorphum Hi (d) <0.01 0.25 2.9 5.5 3 x 10-4C. tetanomorphum Hi (f) <0.01' 0.58Clostridium Bi (d) 0.10 0.38 3.5 1.8 5 x 10-3

a Activities are given in lsmoles per minute and mg protein at 25 C. They were measured in crude extractswhich were either dialyzed (d) over night against 0.01 M Tris-hydrochloride (pH 8.1) or used immediately aftercentrifugation (f, fresh). For test conditions see Materials and Methods.

" EC 2.7.2.1/7; donor was ATP.C EC 2.3.1.8/19.dEC 2.8.3.1.eAcceptor.

VOL. 117, 1974 1255

on January 16, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

BUCKEL AND BARKER

known pathways are involved in the fermenta-tion of glutamate by the bacteria studied.However, the existence of a third pathway is notcompletely excluded because identical labelingpatterns may arise from different pathways.Therefore, the occurrence of enzymes specificfor each pathway was determined in extracts.

For the pathway used by the C. tetanomor-phum group, the enzyme 3-methylaspartasewas selected. It is measured easily by a sensi-tive spectrophotometric assay, and its sub-strate, 2s,3s-3-methylaspartate, is not knownto participate in other metabolic pathways (9).Of enzymes specific for the pathway used by P.aerogenes group, there was not much choice,since only two are known, glutamate dehydro-genase and 2-hydroxyglutarate dehydrogenase,also called 2-ketoglutarate reductase (29, 33).The presence of the latter enzyme seems to be agood indication for this pathway because 2-hydroxyglutarate is not a very common sub-strate. Furthermore, evidence has been pro-vided for the first time that R-2-hydroxygluta-rate is an intermediate in the pathway used byA. fermentans. This acid was fermented toacetate and butyrate twice as fast as s-gluta-mate at the same concentration (Table 6),whereas s-2-hydroxyglutarate was used veryslowly. Both dehydrogenases were tested in one

assay with 2-ketoglutarate and reduced NADPwith and without ammonia. To exclude a simu-lated 2-hydroxyglutarate dehydrogenase activ-ity caused by a glutamate dehydrogenase andimpurities of ammonia, the back reactions weremeasured as well, with NAD and R- or s-2-hydroxyglutarate as substrates. The two assaysgave concordant results.The specific activities of these enzymes in

crude extracts of most of the organisms arelisted in Table 7. These results are completelyconsistent with the labeling patterns. Bacteria

from the C. tetanomorphum group contain highlevels of 3-methylaspartase and varying levels ofthe common enzyme glutamate dehydrogenase,but no 2-hydroxyglutarate dehydrogenase. Fur-thermore, all these organisms contain signifi-cant levels of the first enzyme of their pathway,the coenzyme B,2-dependent glutamate mu-tase. Although the activity of this enzyme isdifficult to measure precisely, its presenceserves as an excellent indicator of the fermenta-tion mechanism (Table 8).The organisms of the P. aerogenes group show

an entirely different enzyme pattern. 3-Methylaspartase is completely absent, whereasboth dehydrogenases are present at high levelsnot commonly found with other bacterial en-

zymes (Table 7). All the 2-hydroxyglutaratedehydrogenases are stereospecific for R-2-hydroxyglutarate. The low activities with thes-enantiomer are probably caused by contami-nation of the commercial product with smallamounts of racemic material (see Table 6 andMaterials and Methods). The inability of C.perfringens (not shown in Table 3) and C.carnofoetidum to ferment glutamate correlateswith the absence of 3-methylaspartase and2-hydroxyglutarate dehydrogenase in both spe-

cies. The high level of glutamate dehydrogenasein crude extracts of C. carnofoetidum is consist-ent with the findings of Beerens et al. (11).Surprisingly, the two clostridia, ClostridiumSB4 and C. sticklandii, which ferment gluta-mate very slowly and only in a complex medium(Table 4) contain 3-methylaspartase (Table 7)and glutamate mutase (C. SB4, Table 8). It isknown that C. sticklandii is capable of ferment-ing pairs of amino acids, i.e., the Sticklandreaction (48). Therefore, it is probable that inboth organisms pyruvate, derived from gluta-mate via 3-methylaspartate, serves as a reduc-tant, whereas a different amino acid, e.g.,

TABLE 6. Fermentations of 2-hydroxyglutarate and s-glutamate with whole cells of A. fermentansa

Products

Amount GasSubstrate (Amol) Maximal Total |NH Acetate Butyrate Acetate/

rate amount (Amol) (Mmol) (Mimol) butyrate(mm/min) (mm)

R-2-Hydroxyglutarate 40 30 280 0 23.4 8.0 2.9R-2-Hydroxyglutarate 20 14 150s-2-Hydroxyglutarate 40 10 0 0 0s-Glutamate 40 16 340 34.6 34.7 12.4 2.8

a Each fermentation was performed with 2 ml of a cell suspension obtained from 400 ml of a resting culture.The values are corrected with a blank determination (20 mm of gas, 18.7 ,umol of NH3, 6.5 Mmol of acetate, nobutyrate). For further conditions see Materials and Methods.

1256 J. BACTRIOL.

on January 16, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

GLUTAMATE FERMENTATION BY ANAEROBES

glycine, may act as an oxidant. This wouldexplain the higher acetate-butyrate ratio ob-tained with Clostridium SB4 than with most ofthe other organisms.The determination of enzymes in crude ex-

tracts of all these organisms substantiates theconclusion that only two pathways are involvedin the fermentation of glutamate. Althoughslight modifications of Fig. 1 are found withspecies which need, in addition to glutamate,another amino acid as source for carbon andenergy, a third pathway can be excluded. Fur-thermore, 2-hydroxyglutarate dehydrogenaseand 3-methylaspartase can be regarded as spe-cific for each pathway, which is important forthe analysis of other glutamate-fermenting or-

ganisms. The measurement of these enzymesrequires much less time than the application ofthe 14C method. It is therefore possible to namethe pathway by the specific intermediate in-volved. Thus, in the following discussion, thepathways used by the C. tetanomorphum andP. aerogenes groups are called the "methylas-

partate" and the "hydroxyglutarate" pathways,respectively.Relationships among glutamate-fermen-

ting bacteria. In the classification of bacteria,many physiological properties are used in addi-tion to morphological characteristics (44).These are tested most commonly by the disap-pearance of a substrate or the formation ofproducts, e.g., decrease in the pH of the me-

dium caused by the decomposition of a certainsugar. But the results presented in this paper

indicate clearly that such observations alonemay be inadequate for taxonomic purposes. Themetabolism of a given substrate by variousorganisms to identical products does not neces-

sarily involve identical enzymatic reactions. Inthe case of glutamate fermentation, two very

different pathways are used, each of whichprobably involves at least four to five specificenzymes which are not found in the other. Thenature of the pathway and the nature of par-ticipating enzymes are clearly more fundamen-tal taxonomic characteristics than the identities

TABLE 7. Activities of 2-hydroxyglutarate dehydrogenase (HGDH), 3-methylaspartase (MA), and glutamatedehydrogenase (GlDH) in crude extracts of most of the organisms

HGDH activitya activitya G1DH activitya

Organism 2-Ketoglutarate | NADI 2s, 3s-3- 2-Ketoglutarate s-Gluta-~~Methyl- + NH." mate'

NADH NADPH R-HGd s-HGd aspartate NADH NADPH (NAD-)

C. tetanomorphum Hl 0 0 9.4 0.17 0.17C. tetanomorphum 2909 0.21 0 7.1 0.43 0.09Clostridium B1 0 0 7.2 0.79 0.38C. cochlearium 1.31 0 0 0 9.0 6.3 0.03Clostridium GL 12A 0 0 0 0 9.0 0.08 0.06Clostridium C 141 0 0 2.6 0 0.13C. saccharobutyricum 0 0 5.9 0.10 0.03P. aerogenes 15 2.4 0.2 0 28 4.3fUnidentified coccus 0.015 0.002 0 0 0 0.016 0.020A. fermentans 44 1.6 6.1 0.7 0 49 1.5 7.4F. fusiforme 2388 26 0.8 0.25 0 29 3.0'F. nucleatum 4357 25 4.6 2.2 0.5 0 36 e 3.5tC. microsporum 61 0 3.4 0.2 0 27 0.06 1.5Clostridium SB4 (me- 0 0 2.5 30 0dium C)

Clostridium SB4 (lysine 0 10medium) (18)

C. sticklandii 0.56 0 2.3 6.1 0.29C. carnofoetidum 0 0 0 17 0.02C. perfringens 0 0 0 0 0.01

a Activity is measured in units per milligram of protein. Zero activity means <0.01 U/mg.Substrate has pH of 6.5.

c Substrate has pH of 8.5.dHG, 2-Hydroxyglutarate.'60% inhibition of HGDH activity after addition of NH4+.' No reaction with R-glutamate.

VOL. 117, 1974- 1257

on January 16, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

BUCKEL AND BARKER

TABLE 8. Activities of glutamate mutase in crudeextracts of organisms using the methylaspartate

pathway

Activity of mutase

Spectro-photo- Anaerobicmetric assay'

Organism assay'

(gmol per Medium Mediummin per c Amg)

C. tetanomorphum Hl 0.0085 OCC. tetanomorphum 2909 0.005 1 0.03Clostridium Bi 0.02 0.3 0.1C. cochlearium 10Clostridium Gl 12A 0.0087 10Clostridium C141 10C. saccharobutyricum 0.017 Od 0.1Clostridium SB4 0 Od 4

a Substrate was s-glutamate. Cells were grown onmedium A.

'Substrate was mesaconate plus NH4.. Measuredas glutamate/3-methylaspartate (mole/mole).

^ 3-Methylaspartate plus a trace of glutamate.d Only 3-methylaspartate.

of the substrate and products.The utilization of different enzymes and

pathways in the degradation of glutamate bytwo organisms must indicate a relatively distanttaxonomic relationship between them, and pre-sumably reflects a long evolutionary separation.However, the converse situation, the utilizationof identical pathways by different organisms,does not necessarily indicate a close taxonomicrelationship, because of the possibility of con-vergent evolution. Evidence for or against theorigin of identical pathways by the latter proc-ess could presumably be obtained by determin-ing the amino acid composition and sequence ofrelevant enzymes.The correlation between the usual taxonomic

criteria and the pathway of glutamate fermen-tation is good for organisms utilizing the me-thylaspartate pathway. All the organisms usingthis pathway are clostridia, and all clostridiatested, with the possible exception of C. micros-porum, use this pathway. Actually C.microsporum HB25 may not be a real exceptionto this generalization, since its true taxonomicposition is uncertain. This strain was originallyidentified as Fusiformis biacutus (G. C. Mead,private communication), and its "spores" werefound not to be viable after exposure to 80 C for10 min. Consequently it may not be a clostrid-ium.

In contrast, the bacteria using the hydroxy-

glutarate pathway are very heterogeneous. Theydiffer in many properties, including morphol-ogy, gram-stain reaction, apparent ability toform spores, and the guanine plus cytosinecontent of their deoxyribonucleic acids (24, 39,51). They belong to four genera in four differentfamilies, namely, Bacillaceae (C. micros-porum), Micrococcaceae (Peptococcus), Neis-sericaceae (Acidaminococcus), and Bacteri-odaceae (Fusobacterium). Since these orga-nisms cannot be closely related, it is possiblethat the hydroxyglutarate pathway developedindependently in one or more of these families.

In bacterial metabolism, one finds other ex-amples of the conversion of a certain substrateby two different pathways to identical products.In connection with this work, the fermentationof lactate and related three-carbon compoundsto propionate, acetate, and CO2 is remarkablebecause of the striking analogies to the fermen-tation of glutamate. As in the methylaspartatepathway, the degradation of lactate by Propio-nibacterium shermanii involves a coenzymeB ,2-dependent carbon-carbon rearrangement,i.e., the conversion of- succinyl CoA to methyl-malonyl CoA (8, 50). The analogy to the hydroxy-glutarate pathway is found in C. propionicum,where the linear carbon skeleton of lactate alsoremains unchanged in the product propionate(31). The crucial step of both latter pathwaysseems to be the dehydration of the correspond-ing 2-hydroxy acid to gluconate or acrylate,respectively (2, 47).

ACKNOWLEDGMENTSWe are indebted for the gifts of bacteria listed in Table 1.

We thank Barbara Baltimore for skillful technical assistance,Michael Herbst for determinations on the amino acidanalyser, and H. Eggerer and P. Willadsen (University ofRegensburg, Germany) for reading the manuscript.

This work was supported by Public Health Service re-search grant AI-00563 from the National Institute of Allergyand Infectious Diseases, and by funds from the CalifomiaAgricultural Experiment Station. One of us (W.B.) wassupported by a generous grant from the Deutsche For-schungsgemeinschaft.

LITERATURE CITED1. Alworth, W. L. 1972. Stereochemistry and its applica-

tions in biochemistry, p. 69-81. Wiley-lnterscience.New York.

2. Anderson, R. L., and W. A. Wood. 1969. Carbohydratemetabolism in microorganisms. Annu. Rev. Microbiol.23:539-578.

3. Ballentine, R. 1957. Determination of total nitrogen andammonia, p. 984-995. In S. P. Colowick and N. 0.Kaplan (ed.), Methods in enzymology, vol. 3. Aca-demic Press Inc., New York.

4. Barker, H. A. 1937. On the fermentation of glutamic acid.Enzymologia 2:175-182.

5. Barker, H. A. 1939. The use of glutamic acid for theisolation and identification of Clostridium cochleariumand Cl. tetanomorphum. Arch. Mikrobiol. 10:376-384.

1258 J. BACTERIOL.

on January 16, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

GLUTAMATE FERMENTATION BY ANAEROBES

6. Barker, H. A. 1956. Bacterial fermentations. John Wileyand Sons, Inc., New York.

7. Barker, H. A. 1961. Fermentations of nitrogenous organiccompounds, p. 151-207. In I. C. Gunsalus and R. Y.Stanier (ed.), The bacteria, vol. 2. Academic Press,Inc., New York.

8. Barker, H. A. 1972. Coenzyme B,2-dependent mutasescausing carbon chain rearrangements, p. 509-537. In P.0. Boyer et al. (ed.), The enzymes, vol. 6, 3rd ed.Academic Press Inc., New York.

9. Barker, H. A., R. D. Smyth, E. J. Wawszkiewicz, M. N.Lee, and R. M. Wilson. 1958. Enzymic preparatiori andcharacterization of an a-L-ft-methyl-aspartic acid.Arch. Biochem. Biophys. 78:468-476.

10. Barker, H. A., E. R. Stadtman, and A. Kornberg. 1955.Coenzyme A transphorase from Clostridium kluyveri,p. 599-602. In S. P. Colowick and N. 0. Kaplan (ed.),Methods in enzymology, vol. 1. Academic Press Inc.,New York.

11. Beerens, H., S. Sugama, and M. Tahon-Castel. 1965.Psychrotrophic clostridia. J. Appl. Bacteriol. 28:36-48.

12. Beilstein. 1962. Handbuch der Organischen Chemie, vol.3, 3rd supplement, p. 929. Springer-Verlag, Heidel-berg.

13. Bergmeyer, H. U., and H. Moellering. 1966. En-zymatische Bestimmung von Acetat. Biochem. Z.344:167-189.

14. Blair, A. H., and H. A. Barker. 1966. Assay and purifica-tion of (+)-citramalate hydro-lyase components fromClostridium tetanomorphum. J. Biol. Chem.241:400-408.

15. Bray, G. A. 1960. A simple efficient liquid scintillator forcounting aqueous solutions in a liquid scintillationcounter. Anal. Biochem. 1:279-285.

16. Buckel, W., and H. Eggerer. 1965. Zur optischen Bestim-mung von Citrat-synthase und Acetyl-Coenzym A.Biochem. Z. 343:29-43.

17. Clifton, C. E. 1942. The utilization of amino acids andrelated compounds by Clostridium tetani. J. Bacteriol.44:179-183.

18. Cohen, G. N., B. Nisman, and G. Cohen-Bazire. 1948.Sort des amino-acides d6grades par Cl.saccharobutyricum et Cl. sporogenes. Bull. Soc. Chim.Biol. 30:109-111.

19. Costilow, R. N., 0. M. Rochovansky, and H. A. Barker.1966. Isolation and identification of ,-lysine as anintermediate in lysine fermentation. J. Biol. Chem.241:1573-1580.

20. Crestfield. A. M., and F. W. Allen. 1955. Improvedapparatus for zone electrophoresis. Anal. Chem.27:422-425.

21. Dreze, A., S. Moore, and E. J. Bigwood. 1954. On thedesalting of solutions of amino acids by ion exchange.Anal. Chim. Acta 11:554-558.

22. Fansler, B., and J. M. Lowenstein. 1969. Aconitase frompig heart, p. 26-30. In J. M. Lowenstein (ed.), Meth-ods in enzymology, vol. 13. Academic Press Inc., NewYork.

23. Goldmark, P. J., and S. Linn. 1970. An endonucleaseactivity from Escherichia coli absent from certain rec-strains. Proc. Nat. Acad. Sci. U.S.A. 67:434-441.

24. Hill, R. L. 1966. An index to desoxyribonucleic acid basecompositions of bacterial species. J. Gen. Microbiol.44:419-437.

25. Horler, D. F., W. B. McConnell, and D. W. S. Westlake.1966. Glutaconic acid, a product of the fermentation ofglutamic acid by Peptococcus aerogenes. Can. J. Mi-crobiol. 12:1247-1252.

26. Horler, D. F., D. W. S. Westlake, and W. B. McConnell.1966. Conversion of glutamic acid to volatile acids byMicrococcus aerogenes. Can. J. Microbiol. 12:47-53.

27. Jackins, H. C., and H. A. Barker. 1951. Fermentative

processes of the fusiform bacteria. J. Bacteriol.61:101-114.

28. Johnson, W. M., and D. W. S. Westlake. 1969. Theproduction of a-hydroxyglutaric acid from glutamicacid by cell-free preparations of Peptococcus aerogenes.Can. J. Biochem. 47:1103-1107.

29. Johnson, W. M., and D. W. S. Westlake. 1972. Purifica-tion and characterization of glutamic acid dehydrogen-ase and a-ketoglutaric acid reductase from Peptococ-cus aerogenes. Can. J. Microbiol. 18:881-892.

30. Kew, 0. M., and C. A. Woolfolk. 1970. Preparation ofglutamate dehydrogenase from Micrococcus aerogenes.Biochem. Biophys. Res. Commun. 39:1126-1133.

31. Leaver, F. W., H. G. Wood, and R. Stjernholm. 1955. Thefermentation of three carbon substrates by Clostridiumpropionicum and Propionibacterium. J. Bacteriol.70:521-530.

32. Lenz, H., W. Buckel, P. Wunderwald, G. Biedermann, V.Buschmeier, H. Eggerer, J. W. Cornforth, J. W. Red-mond, and R. Mallaby. 1971. Stereochemistry of si-citrate synthase and ATP-citrate-lyase reactions. Eur.J. Biochem. 24:207-215.

33. Lerud, R. F., and H. R. Whiteley. 1971. Purification andproperties of a-ketoglutarate and reductase from Mi-crococcus aerogenes. J. Bacteriol. 106:571-577.

34. Lowry, 0. H., N. J. Rosenbusch, A. L. Farr, and R. J.Randall. 1951. Protein measurement with the Folinphenol reagent. J. Biol. Chem. 193:265-275.

35. McConnell, W. B., D. F. Horler, and D. W. S. Westlake.1967. Metabolism of histidine by Peptococcusaerogenes. Can. J. Microbiol. 13:143-146.

36. Moore, S., and W. H. Stein. 1954. A modified ninhydrinreagent for the photometric determination of aminoacids and related compounds. J. Biol. Chem.211:907-913.

37. Mosbach, E. H., E. F. Phares, and S. F. Carson. 1951.Degradation of isotopically labelled citric, a-ketoglu-taric and glutamic acids. Arch. Biochem. Biophys.33:179-185.

38. Phares, E. F. 1951. Degradation of labelled propionic andacetic acids. Arch. Biochem. Biophys. 33:173-178.

39. Rogosa, M. 1969. Acidaminococcus gen. n., Acidamino-coccus fermentans sp. n., anaerobic gram-negativediplococci using amino acids as the sole energy sourcefor growth. J. Bacteriol. 98:756-166.

40: Rose, I. A. 1955. Acetate kinase of bacteria (acetokinase),p. 591-595. In S. P. Colowick and N. 0. Kaplan (ed.),Methods in enzymology, vol. 1. Academic Press Inc.,New York.

41. Seki, T. 1958. Chromatographic separation of lower fattyacids. J. Biochem. (Tokyo) 45:855-860.

42. Siebert, G. 1962. Citrat und Isocitrat, p. 318-328. In H. U.Bergmeyer (ed.), Methoden der enzymatischenAnalyse. Verlag Chemie, Weinheim.

43. Simon, E. J., and D. Shemin. 1953. The preparation ofS-succinyl coenzyme A. J. Amer. Chem. Soc. 75:2520.

44. Spray, R. S. 1936. Semisolid media for cultivation andidentification of the sporulating anaerobes. J. Bacte-riol. 32:135-155.

45. Stadtman, E. R. 1953. The coenzyme A transphorasesystem in Clostridium kluyveri. J. Biol. Chem.203:501-512.

46. Stadtman, E. R. 1955. Phosphotransacetylase from Clos-tridium kluyveri, p. 596-599. In S. P. Colowick and N.0. Kaplan (ed.), Methods in enzymology, vol. 1.Academic Press Inc., New York.

47. Stadtman, E. R., and P. R. Vagelos, 1958. Propionic acidmetabolism, p. 86-92. In K. Ichiara (ed.), Proc. Int.Symp. Enzyme Chem., Tokyo-Kyoto. Academic PressInc., New York.

48. Stadtman, T. C. 1954. On the metabolism of an aminoacid-fermenting clostridium. J. Bacteriol. 67:314-320.

1259VOL. 117, 1974

on January 16, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

BUCKEL AND BARKER

49. Stadtman, T. C., and L. S. McClung. 1957. Clostridiumsticklandii nov. spec. J. Bacteriol. 73:218-219.

50. Stjernholm, R., and H. G. Wood. 1961. Methylmalonylisomerase, II. Purif'ication and properties of the enzymefrom propionibacteria. Proc. Nat. Acad. Sci. U.S.A.47:303-313.

51. Tonomura, B., R. Malkin. and J. C. Rabinowitz. 1965.Desoxyribonucleic acid base composition of clostridialspecies. J. Bacteriol. 89:1438-1439.

52. Twarog, R., and R. S. Wolf'e. 1962. Enzymatic phospho-rylation of butyrate. J. Biol. Chem. 237:2474-2477.

53. Valentine, R. C., and R. S. Wolf'e. 1960. Purif'ication androle of phosphotransbutvrvlase. J. Biol. Chem.235:1948-1952.

54. Wachsman, J. T., and H. A. Barker. 1955. Tracerexperiments on glutamate fermentation by Clos-

tridium tetanomorphum. J. Biol. Chem. 217:695-702.55. Waelsch, H. 1957. Chemical determination of glutamic

acid, p. 570-575. In S. P. Colowick and N. 0. Kaplan(ed.), Methods in enzvmology, vol. 3. Academic PressInc., New York.

56. Whitelev, H. R. 1957. Fermentation of' amino acids byMicrococcus aerogenes. J. Bacteriol. 74:324-330.

57. Woods, D. D., and C. E. Clifton, 1937. Studies in themetabolism of the strict anaerobes (genusClostridium). VI. Hydrogen production and amino-acid utilization by Clostridium tetanomorphum. Bio-chem. J. 31:1774-1788.

58. Woods, D. D., and A. R. Trim. 1942. Studies in themetabolism ot' amino-acids by Cl. welchii. Biochem. J.36:501-512.

1260 J. BACTERIOL.

on January 16, 2020 by guesthttp://jb.asm

.org/D

ownloaded from