Dual Mechanismsof Tricarboxylate Transport and Catabolism

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, JUlY 1994, p. 2538-2544 Vol. 60, No. 70099-2240/94/$04.00+0Copyright © 1994, American Society for Microbiology

Dual Mechanisms of Tricarboxylate Transport and Catabolismby Acidaminococcus fermentansGREGORY M. COOK' AND JAMES B. RUSSELL' 2*

Section of Microbiology, Cornell University, 1 and Agricultural Research Service,U.S. Department ofAgrculture,2 Ithaca, New York 14853

Received 11 March 1994/Accepted 6 May 1994

Acidaminococcus fermentans utilized citrate or the citrate analog aconitate as an energy source for growth,and these tricarboxylates were used simultaneously. Citrate utilization and uptake showed biphasic kinetics.High-affinity citrate uptake had a K, of 40 ,uM, but the Vmax was only 25 nmol/mg of protein per min.Low-affinity citrate utilization had a 10-fold higher V.ax, but the Ks was greater than 1.0 mM. Aconitate wasa competitive inhibitor (K, = 34 ,uM) of high-affinity citrate uptake, but low-affinity aconitate utilization hada 10-fold-lower requirement for sodium than did low-affinity citrate utilization. On the basis of this largedifference in sodium requirements, it appeared that A. fermentans probably has two systems of tricarboxylateuptake: (i) a citrate/aconitate carrier with a low affinity for sodium and (ii) an aconitate carrier with a highaffinity for sodium. Citrate was catabolized by a pathway involving a biotin-requiring, avidin-sensitive,sodium-dependent, membrane-bound oxaloacetate decarboxylase. The cells also had aconitase, but this enzymewas unable to convert citrate to isocitrate. Since cell-free extracts converted either aconitate or glutamate to2-oxoglutarate, it appeared that aconitate was being catabolized by the glutaconyl-CoA decarboxylase pathway.Exponentially growing cultures on citrate or citrate plus aconitate were inhibited by the sodium/protonantiporter, monensin. Because monensin had no effect on cultures growing with aconitate alone, it appearedthat citrate metabolism was acting as an inducer of monensin sensitivity. A. fermentans cells always had a lowproton motive force (<50 mV), and cells treated with the protonophore TCS (3,3',4',5-tetrachlorosalicylanide)grew even though the proton motive force was less than 20 mV. On the basis of these results, it appeared thatA. fermentans was depending almost exclusively on a sodium motive force for its membrane energetics.

Grass tetany, a potentially fatal magnesium deficiency dis-ease of ruminants, has been correlated with the suddenaccumulation of aconitate in plants and its subsequent reduc-tion to tricarballylate by ruminal bacteria (1, 18, 21, 26-28, 31,32). Recent work showed that Acidaminococcus fermentanswas able to oxidize either aconitate or citrate (6). When A.fermentans was inoculated into a cow, a greater fraction of theaconitate was oxidized, and tricarballylate accumulation wasdecreased by 50% (7).The rumen has a high concentration of sodium (approxi-

mately 100 mM), and many ruminal bacteria are dependent ona sodium circulation for transport and energy derivation (30).Ruminants are often fed monensin, an antiporter which facil-itates an exchange of protons for sodium or other monovalentcations (20). Monensin increases the feed efficiency of feedlotcattle and the gain of grazing ruminants (29).

Gram-negative bacteria, which have an outer membrane toprotect the cell membrane, are generally more resistant tomonensin than are gram-positive species (29). The benefit ofmonensin as a feed additive has generally been ascribed to theshift of ruminal ecology to a predominantly gram-negativefermentation (29). A. fermentans is a gram-negative bacterium(24), but Buckel and coworkers reported that A. fermentanscould be inhibited by high concentrations of monensin (4, 36).

Preliminary experiments indicated that A. fermentans re-quired large amounts of sodium for citrate fermentation andthat citrate utilization was inhibited by monensin. Becausesodium and monensin had a much lesser effect when aconitatewas the energy source, we decided to define the mechanisms of

* Corresponding author. Mailing address: Wing Hall, Cornell Uni-versity, Ithaca, NY 14853. Phone: (607) 255-4508. Fax: (607) 255-3904.

tricarboxylate catabolism, the nature of the sodium require-ment, and the action of monensin in A. fermnentans.

MATERIALS AND METHODS

Cell growth. Acidaminococcus fermentans AO was grown at39°C in anaerobic basal medium containing (per liter) 292 mgof K2HPO4, 292 mg of KH2PO4, 480 mg of Na2SO4, 480 mg ofNaCl, 100 mg of MgSO4 * 7H20, 64 mg of CaCl2 - 2H20, 600mg of cysteine hydrochloride, vitamins and macrominerals (8),1 g of Trypticase (BBL Microbiology Systems, Cockeysville,Md.), and 20 mM trans-aconitate or citrate. The medium wasadjusted to pH 6.7, and the final pH was never less than 6.5.Growth was monitored spectrophotometrically at 600 nm.

Citrate and aconitate fermentation by washed cells. Expo-nential cultures were harvested by centrifugation (12,000 x g,5 min, 4°C) and were washed twice under anaerobic conditionsin 100 mM sodium-phosphate buffer (pH 6.5) containing 10mM potassium chloride and 2 mM dithiothreitol. The washedcell suspensions (2.2 mg of protein per 10 ml) were incubatedat 39°C, and substrate utilization was initiated by the additionof citrate or trans-aconitate (20 mM final concentration).Samples (1.5 ml) were taken at regular time intervals and werecentrifuged (1.5 ml, 13,000 x g, 2 min, 23°C). Fermentationacids in the cell-free supernatants were analyzed by high-performance liquid chromatography (HPLC) (see analysesbelow for column conditions and flow rate).

Transport assays. Cells were harvested by centrifugation(12,000 x g, 5 min, 4°C) during exponential growth (0.22 mg ofprotein per ml) and were washed twice in potassium-phosphatebuffer (50 mM, pH 6.5). The cell pellets were resuspended inthe same buffer to achieve a concentration of 14 to 20 mg ofprotein per ml. The concentrated cells (4 ,ul) were diluted

2538

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 11

Feb

ruar

y 20

22 b

y 14

8.71

.171

.138

.

TRICARBOXYLASE UTILIZATION BY A. FERMENTANS 2539

50-fold into 200,u of buffer (39°C), and transport was initiatedby the addition of 100 nCi of [1,5-14C]citrate (110 mCi/mmol)or [U-14C]glutamate (270 mCi/mmol). After 0 to 60 s, trans-port was terminated by the addition of ice-cold LiCl (2 ml, 100mM) and by rapid filtration (0.45-,um-pore-size cellulose-nitrate filter). Anaerobically prepared cells were harvestedunder N2, washed in potassium buffers (2 mM dithiothreitol),and assayed under N2.

Preliminary experiments indicated that the transport ratewas first order with respect to protein from 0 to 5 s, and initialrates were always measured over this time interval. The filterwas washed once with 2.0 ml of LiCl, dried for 20 min at 105°C,and counted by liquid scintillation. Preliminary experimentsindicated that virtually all of the ['4C]citrate either was trappedin the cells or remained in the extracellular supernatant. Theseresults indicated that little of the [14C]citrate was being lost asCO2 during the transport assay. Cells which were treated withmonensin (10,uM) or incubated in the absence of sodiumshowed essentially no ['4C]citrate or ['4C]glutamate uptake,and this result indicated that there was little nonspecificbinding to cells.

Artificial membrane potentials. An artificial membranepotential (At) was generated by loading valinomycin-treated(10,uM, 0°C, 30 min) cells with potassium (100 mM potassiumphosphate buffer containing 10mM MgSO4* 7H20, pH 6.5)and diluting them into sodium phosphate buffer (100 mMsodium phosphate buffer containing 10 mM MgSO4* 7H2O,pH 6.5). Acetate-loaded cells (80 mM potassium acetate, 20mM potassium phosphate, 10mM MgSO4* 7H20, pH 6.5, 0°C,30 min) were diluted into buffer lacking acetate (100 mMpotassium phosphate buffer, 10mM MgSO4 - 7H20, pH 6.5),to create an artificial pH gradient (ZApH). Transport wasinitiated by a 50-fold dilution of concentrated cells into buffercontaining the radioactive citrate (see above).Membrane vesicles. Right-side-out membrane vesicles were

prepared by mutanolysin-lysozyme digestion and osmotic lysisas previously described (16).Measurement of the electrochemical proton gradient (Ap).

Exponentially growing cultures were anaerobically transferredwith a hypodermic syringe (2.0 ml) to a tube (13 by 100 mm)which contained [3H]tetraphenylphosphonium ion (TPP+, 0.5,uCi, 30 ,uCi/,umol), [7-'4C]benzoate (1.0 ,uCi, 21.8 jiCi/jxmol),[U-14C]sucrose (1.0 ,uCi, 540 ,uCi/,umol), or 3H20 (4.00 ,uCi,3.6,uCi/,umol). After incubation for 5 min at 39°C, the cultureswere centrifuged through silicon oil (equal-part mixture ofDexter Hysol 550 and 560; Hysol Co., Olean, N.Y.) in micro-centrifuge tubes (13,000 x g, 5 min, 22°C), and 20-pul samplesof supernatant were removed. The tubes and contents werefrozen (-15°C), and the bottoms (cell pellets) were removedwith dog nail clippers. Supernatant and cell pellets wereresuspended in scintillation fluid and counted.The intracellular volume (2.35 pA/mg of protein) was esti-

mated from the difference between the partitioning of 3H20and [14C]sucrose. The electrical potential across the cell mem-brane (At) was calculated from the uptake of 3H-TPP+according to the Nernst relationship. Nonspecific TPP+ bind-ing was estimated from cells which had been treated withvalinomycin and nigericin (10 puM each). The ApH was deter-mined from the distribution of [14C]benzoate by using theHenderson-Hasselbalch equation (22), and ZApH was calcu-lated as 68 mV x ApH. AT and ZApH were corrected forextracellular contamination.Enzyme assays. One-liter cultures of A. fermentans grown

anaerobically on 20 mM citrate or aconitate were harvestedduring exponential growth (approximately 0.8 optical density,1-cm-diameter cuvette, 600 nm, 10,000 x g, 15 min, 5°C) and

were washed three times with potassium phosphate buffer (50mM, pH 6.5). The cell pellet (176 mg of protein per5 ml) wassonicated for 20 min at 0°C (Branson model 200 sonifier,maximum output, microtip, 50% duty cycle). Approximately90% of the cells were broken. Unbroken cells and cell debriswere removed by centrifugation (15,000 x g, 20 min, 5C) andthe cell extract was placed on ice. Specific activities weredetermined in a range in which linearity of activity with proteinconcentration was shown and were expressed as the amount ofenzyme catalyzing the conversion of 1 nmol of substrate permin per mg of protein. All assays were carried out in triplicatewith appropriate controls, and differences in replicates wereless than 5%. When available, commercial enzymes were usedas a positive control.

Citrate lyase was assayed by the method of Daron andGunsalus (9). The assay volume (1 ml) contained 100 mMKH2PO4 (pH 7.4), 0.1 mM MnCl2, 10 mM potassium citrate,and 50 pA cell-free extract (approximately 1 mg of protein).Activity was monitored continuously by the increase in A260.Acetate kinase was assayed by the method of Rose (25). Theassay volume (1 ml) contained 0.77 M potassium acetate, 48mM Tris-HCl, 9.7 mM MgCl2, 0.7 M KOH, 10 mM ATP, 4.9%(wt/vol) hydroxylamine hydrochloride, and 20 pA of cell extract(approximately 0.4 mg of protein). After 2 min (25°C), theconversion of acetate (via acetyl phosphate) to acetohydrox-amic acid was terminated by the addition of trichloroaceticacid (5% final concentration). The reaction mixture was cen-trifuged (13,000 x g, 5 min, 25°C), the supernatant was mixedwith 4 ml of ferric chloride solution (1.25% FeCl3 in 1.0 MHCl), and the color was measured at 540 nm. Commercialacetyl phosphate (Sigma, St. Louis, Mo.) was used as astandard.

Oxaloacetate decarboxylase activity. In order to determinethe localization of oxaloacetate decarboxylase, the cell crudeextract (1 ml) was fractionated by centrifugation (180,000 x g,60 min, 5C). The membrane pellet was washed once inpotassium phosphate buffer (50 mM [pH 7.0], containing 5 mMMgCl2) and was resuspended in a final volume of 0.5 ml. Theactivity of the washed-membrane and cytoplasmic fractionoxaloacetate decarboxylase was measured by monitoring thedecrease of enolic oxaloacetate at 265 nm as described byDimroth (11). The assay (1 ml at 25°C) contained 100 mMTris-HCl (pH 7.5), 20 mM NaCl, 1 mM oxaloacetate, and 0.05mg of biotin. A commercial preparation of oxaloacetate decar-boxylase (Sigma) was used as a control.

Other analyses. Protein from NaOH-hydrolyzed cells (0.2 MNaOH, 100°C, 15 min) and cell crude extracts were assayed bythe method of Lowry et al. (17). Citrate, aconitate, andfermentation acids in cell-free supernatant samples were ana-lyzed by HPLC with a Beckman 334 liquid chromatographwhich was equipped with a model 156 refractive index detectorand a Bio-Rad HPX-87H organic acid column. The samplesize was 20 pA, the eluant was 0.0065 M H2SO4, the flow ratewas 0.5 ml/min, and the column temperature was 50°C.

Statistics. All measurements were performed in triplicate,and the data were analyzed by Student's t test.

RESULTS

Tricarboxylate utilization. A. fermentans was able to utilizeeither aconitate or citrate as an energy source, but the maxi-mum specific growth rate and specific rate of tricarboxylateutilization were always greater for aconitate than for citrate(Table 1). Acetate was the predominant end product of citratefermentation, and the ratio of acetate to citrate was always 2 to1: C6H807 + H20--> 2C2H402 + 2CO2 + H2. Aconitate was

VOL. 60, 1994

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 11

Feb

ruar

y 20

22 b

y 14

8.71

.171

.138

.

2540 COOK AND RUSSELL

TABLE 1. Growth of A. fermentans on citrate and/or aconitate andits specific rates of tricarboxylate utilization

Specific rate of:

Tricarboxylate utilization (nmol/mgGrowth substrate Growth of protein per min)a(h-1)

Citrate Aconitate Total

Citrate 0.3 367 a 367Aconitate 0.5 600 b 600Aconitate + citrate 0.5 333 c 300 c 633

a Means followed by different letters differ significantly (P < 0.05).

converted to acetate, but small amounts of butyrate could alsobe detected: C6H606 + 1.6 H20-- 1.6 C2H402 + 2CO2 + 0.6H2 + 0.2 C4H802. When A. fermentans was provided with acombination of aconitate and citrate, the tricarboxylates wereutilized simultaneously at approximately the same rate (Table1).

Cells washed and incubated in sodium phosphate buffer didnot grow, but the rate of tricarboxylate fermentation was linearfor more than 60 min. On the basis of the Eadie Hofstee plots,the maximum velocity of aconitate fermentation (Vmax) was3.3-fold greater than that of citrate (Fig. la). The affinityconstant (Ks) was 18-fold lower for citrate than for aconitate.When the sodium content of the buffer was decreased, A.fermentans lost its ability to ferment both aconitate and citrate,but 10-fold-more sodium was needed for citrate fermentation(Fig. lb). Washed cells could ferment citrate under aerobic oranaerobic conditions, but aconitate could be fermented onlyanaerobically.Monensin and TCS resistance. The growth ofA. fernentans

was completely inhibited by monensin (10 ,uM) when citratewas the energy source (Fig. 2a). However, if aconitate was theenergy source, monensin had no effect (Fig. 2b). Cells whichwere provided with a combination of citrate and aconitate didnot grow in the presence of monensin (Fig. 2c). A. fermentanshad a proton motive force (Ap) of less than 60 mV when eithertricarboxylate was the growth substrate. The protonophoreTCS (3,3',4',5-tetrachlorosalicylanide) (10 ,uM) decreased theAp to less than 20 mV, but it had no effect on the growth ofA.fermentans on either citrate or aconitate (data not shown).

['4C]citrate transport. Cells which were washed twice inanaerobic sodium-potassium phosphate buffer (100 mM, pH6.5) transported [I C]citrate; however, little if any uptake wasobserved when the sodium was deleted (Fig. 3a). The kineticsof [14C]citrate transport did not follow typical Michaelis-Menten saturation kinetics, and the rate was abnormally highat high citrate concentrations. On the basis of the Eadie-Hofstee plot of both sodium-dependent [14C]citrate transportand unlabelled citrate utilization, it appeared that A. fermen-tans has both high and low affinity systems of citrate uptake(Fig. 3b). Similar rates of [14C]citrate transport were observedwhen the cells were grown on citrate and on aconitate (datanot shown).

Cells which were washed aerobically in potassium phosphatebuffer (50 mM, pH 6.5), resuspended in potassium phosphatebuffer with valinomycin (5 puM), and diluted 50-fold intosodium-potassium phosphate to create an artificial membranepotential (AT) and a chemical gradient of sodium (ApNa),transported [1 C]citrate (2 F.M) at an initial rate of 3 nmol/mgof protein per min (Fig. 4). An artificial ApNa was also able todrive [14C]glutamate transport, and in this case the initial ratewas 23 nmol/mg of protein per min (data not shown). No

500Z.r-0 E 400C

N *__ o 300

L-

L E 200

c E 100CZ c

0

400z Co -

- _ 300. c

" .__..-L 200

o E=

L- 100c-wi

_E

0 0.04 0.08 0.12v/S (nmol/mg protein/min/jaM)

.1 1 10SODIUM (mM)

100

FIG. 1. (a) Eadie-Hofstee plot of citrate and aconitate utilizationsby washed cells of A. fermentans which had 100 mM sodium in thebuffer. The K, and Vmax for aconitate utilization were 21.68 mM and777 nmol/mg of protein per min, respectively. The K. and Vm, forcitrate utilization were 1.2 mM and 230 nmol/mg of protein per min,respectively. (b) The effect of sodium on tricarboxylate utilization bywashed cells of A. fermentans which were provided with an initialconcentration of 20 mM citrate or aconitate.

['4C]glutamate or ['4C]citrate transport was observed whensodium was deleted from the dilution buffer (no ApNa). Therewas a 50% reduction in ['4C]citrate transport when potassiumwas deleted from the dilution buffer, but ['4C]glutamatetransport did not show a requirement for external potassium.Membrane vesicles which were loaded in a similar fashioncould not transport ["4C]citrate or ['4C]glutamate even whensodium and potassium were present in the dilution buffer (datanot shown).Enzyme activities. Cells grown on citrate or aconitate had

citrate lyase, oxaloacetate decarboxylase, and acetate kinaseactivities (Table 2). Most of the oxaloacetate decarboxylaseactivity of citrate-grown cells was membrane associated, andmembrane-associated activity was increased 10-fold by theaddition of sodium. The supernatant oxaloacetate decarboxyl-ase was not stimulated by sodium. Avidin (2 mg/ml) completelyinhibited both oxaloacetate decarboxylase fractions, but thisinhibition could be counteracted by the addition of biotin (50,ug/ml). Cells grown with citrate or aconitate had similaroxaloacetate decarboxylase activities. Cells grown on aconitatehad 12-fold-more aconitase activity than did cells grown oncitrate. The aconitase converted aconitate to isocitrate, but the

APPL. ENvIRON. MICROBIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 11

Feb

ruar

y 20

22 b

y 14

8.71

.171

.138

.

TRICARBOXYLASE UTILIZATION BY A. FERMENTANS 2541

1.5

1.0

0.5

CO0

%D%N00

I-cnz

-J

-.

u03L0D

0.01.5

1.0

0.5

0.02.5

2.0

1.5

1.0

0.5

0.00 2 4 6 8

TIME (h)FIG. 2. The effect of monensin addition (10 ,uM at the arrow) on

the growth of A. fermentans when either citrate (a), aconitate (b), orcitrate and aconitate (c) were the energy sources.

conversion of citrate to isocitrate was very low. Isocitratedehydrogenase activity was twice as high for aconitate-growncells as for citrate-grown cells.

DISCUSSION

A. fennentans was originally described as a chemoorganotro-phic bacterium which utilized "amino acids, especially gluta-mate, as a main energy source" (24). Buckel and Semmler (4)showed that A. fermentans required more than 0.5 mM sodiumto utilize glutamate as an energy source and that more than 20mM sodium was required for optimal growth. Rogosa (23)initially indicated that A. fernentans could not utilize citrateeven when 20 mM sodium was available, and our work

30 -

XL 20

z

i-EIJ O

F 240'-.200-i

*160-

o L 120-

E 80

40

-c

0-

0.0

50 100 150 200

Citrate (MM)

0.2 0.4 0.6 0.8

v/S (nmol/mg protein/min/MM)FIG. 3. (a) [14C]citrate transport by washed cells of A. fermentans

with or without 100 mM sodium. (b) Eadie Hofstee plot of sodium-dependent citrate transport and utilization. The high-affinity systemhad Vmaxof 25 nmol/mg of protein per min and a Kt of 40 ,uM, and thelow-affinity system had a Vmax of 230 nmol/mg of protein per min andaKtof 1.2mM.

indicated that more than 10 mM sodium was needed for citrateutilization (Fig. lb).

In Kiebsiella spp., Salmonella typhimurium, and Lactococcuslactis, citrate catabolism involves a sodium-dependent oxalo-acetate decarboxylase (10, 15), and it appeared that A. fermen-tans was catabolizing citrate by a similar pathway (Fig. 5).Acetate was the end product of citrate fermentation, and theratio of acetate to citrate was 2 to 1. Citrate lyase, oxaloacetatedecarboxylase, and acetate kinase activities were high (Table2). Because the oxaloacetate decarboxylase was largely mem-brane associated and sodium dependent, it appeared that theoxaloacetate decarboxylase was acting as a sodium pump.Because even late-stationary cultures (24 h) retained theirability to take up citrate (data not shown), it appeared that A.fermentans was able to maintain a sodium gradient for a longperiod.

Recent work by Van Der Rest et al. (35) indicated that thesodium-dependent citrate transport system of Kiebsiella pneu-moniae was a sodium/proton symport mechanism. Citrateutilization and transport by A. fermentans could be demon-strated only when sodium was present, but there was littleevidence for a proton/citrate symport. A. fermentans had a low

VOL. 60, 1994

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 11

Feb

ruar

y 20

22 b

y 14

8.71

.171

.138

.

2542 COOK AND RUSSELL

0.8

-

°@ 0 0.6Z L-a CLc 0.41- FE 04-4

LaJ11%.~I-

E£ 0.2_ Aconitateor

Ci trate

iGlutamate

0. 0'1'I0 4 8 12 16 20

TIME (s)FIG. 4. Transport of [14C]citrate by valinomycin-treated and potas-

sium-loaded A. fermentans cells. The cells were diluted into 100 mMTris buffer, 100 mM sodium phosphate, 100 mM potassium phosphateplus 100 mM sodium phosphate, or 100 mM potassium phosphate.

Ap, and TCS, a protonophore which dissipated Ap, had littleeffect on citrate fermentation or transport. Because the citratetransport of A. fermentans could be driven by an artificialsodium gradient under aerobic conditions, it did not appearthat the carrier per se was oxygen sensitive. Valinomycin-treated cells had higher rates of citrate transport when potas-sium was added to the dilution buffer containing sodium (Fig.4). Previous work indicated that the citrate transport system ofKlebsiella (Aerobacter) aerogenes was also stimulated by potas-sium, but the mechanism of this stimulation is as yet not clear(13).The sodium-dependent citrate uptake system of A. fermen-

tans showed biphasic kinetics, and only the low-affinity systemhad a velocity which could explain the citrate fermentation rateof growing cells. Because v/S was relatively constant, it ap-peared that the low-affinity system might be a facilitated

TABLE 2. Enzyme activities involved in the metabolism of citrateand aconitate by A. fermentansa

Activity" with growth substrateEnzyme of:

Citrate Aconitate

Citrate lyase 75.3 a 38.2 b

Oxaloacetate decarboxylaseCell crude extract 64.3 a 74.3 aSupernatant 37.6 NACMembrane fraction 85.7 NA

AconitaseAconitate hydratase 6.8 a 80.4 bCitrate dehydratase 1.7 a 1.2 a

Isocitrate dehydrogenase 10.6 a 20.0 b

Acetate kinase 7,143 a 3,366 b

a All enzyme activities are expressed as nanomoles of substrate converted perminute per milligram of protein at 25°C. Aconitase activity was determined withaconitate as a substrate (aconitate hydratase) or with citrate as a substrate(citrate dehydratase) for the enzyme.bMeans followed by different letters differ significantly (P < 0.05).c NA, enzyme activity not assayed.

FIG. 5. A hypothetical scheme of citrate, aconitate, and glutamatecatabolism in A. fermentans. CoA, coenzyme A.

diffusion mechanism. In S. typhimurium, tricarboxylate trans-port involves a periplasmic binding protein (34), and thisobservation may explain why we were unable to detect [I4C]ci-trate transport in membrane vesicles. Periplasmic bindingproteins would be lost in standard methods of vesicle prepa-ration.

Because mammalian aconitases catalyze the conversion ofaconitate to either citrate or isocitrate, we originally thoughtthat citrate and aconitate might be catabolized by the samepathway (Fig. 5). This hypothesis however, was not consistentwith the following observations: (i) citrate catabolism yieldedonly acetate, but aconitate catabolism yielded butyrate andacetate; (ii) citrate utilization occurred under either anaerobicor aerobic conditions, but aconitate utilization could be de-tected only anaerobically; and (iii) the aconitase convertedaconitate to isocitrate, but the extracts could not convertcitrate to isocitrate (citrate dehydratase activity was very low).The isocitrate dehydrogenase activity of A. fermentans was

twofold higher when the cells were grown on aconitate ratherthan on citrate (Table 2). Buckel and Semmler (3, 4) showedthat A. fennentans catabolized glutamate by a sodium-depen-dent glutaconyl-coenzyme A decarboxylase that was extremelysensitive to oxygen (2), and this previous result is consistentwith the observation that A. fermentans could ferment aconi-tate only under anaerobic conditions. The glutaconyl pathwayof glutamate catabolism normally produces acetate and bu-tyrate in a stoichiometry of 2 to 1 (14). Aconitate catabolismnever yielded this much butyrate, but hydrogen production wasfivefold greater with aconitate than with glutamate (data notshown). Hydrogen production allows bacteria to produce moreoxidized end products (e.g., acetate) at the expense of morereduced end products (e.g., butyrate; see reference 14).

High-affinity ['4C]citrate transport was competitively inhib-ited by unlabelled aconitate, but it was not possible to measureaconitate transport per se. [14C]aconitate is not commerciallyavailable, and standard methods of tritiation would destroy thedouble bond of trans-aconitate. Nonetheless, aconitate utiliza-tion required at least 10-fold-less sodium than did citrate

APPL. ENVIRON. MICROBIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 11

Feb

ruar

y 20

22 b

y 14

8.71

.171

.138

.

TRICARBOXYLASE UTILIZATION BY A. FERMENTANS 2543

utilization. On the basis of the difference in sodium require-ments, it appeared that A. fermentans probably has two systemsof aconitate uptake: (i) a tricarboxylate (citrate/aconitate)carrier with a low affinity for sodium and (ii) an aconitatecarrier with a high affinity for sodium.

Because monensin caused an almost immediate inhibition ofgrowth whenever citrate was present but had essentially noeffect when only aconitate was provided, it appeared thatcitrate utilization is an inducer of monensin sensitivity. Theouter membrane of gram-negative bacteria can protect the cellmembrane from ionophores like monensin, but not all gram-negative bacteria are ionophore resistant (29). Further work isneeded to ascertain if citrate catabolism invokes a change inthe outer membrane of A. fermentans.

Previous work indicated that A. fermentans is not normally apredominant ruminal bacterium, but ruminal inoculation withA. fermentans caused a marked decrease in aconitate-depen-dent tricarballylate accumulation in vitro (7). On the basis ofthese results, it appeared that the inoculation or the enrich-ment of A. fermentans by citrate might provide a method ofpreventing tricarballylate accumulation and grass tetany. Be-cause there is often an increase in the citrate concentration ofplants when aconitate accumulates (21), citrate-induced mon-ensin sensitivity may limit the use of A. fernentans in monen-sin-fed ruminants.A variety of bacteria are able to utilize sodium motive force

as a method of energy transduction (12), but primary sodiumpumps usually create a membrane potential (AT) as well as achemical gradient of sodium (ApNa). Energy transduction inClostridium fervidus, a thermophilic eubacterium, was reportedto be "exclusively coupled to sodium ions," but even thisbacterium was "partially inhibited" by protonophores whichdissipate At (33). A. fermentans was unaffected by the pro-tonophore TCS, even though the Ap was almost completelyabolished. To our knowledge only Clostridium aminophilum (5,19) shows such a complete dependency on a sodium gradient.

ACKNOWLEDGMENT

This research was supported by the U.S. Dairy Forage ResearchCenter, Madison, Wis.

REFERENCES1. Bohman, V. R., F. P. Horn, B. A. Stewart, A. C. Mathers, and D. L.

Grunes. 1983. Wheat pasture poisoning. I. An evaluation of cerealpastures as related to tetany in beef cows. J. Anim. Sci. 57:1352-1363.

2. Buckel, W. 1992. Unusual dehydrations in anaerobic bacteria.FEMS Microbiol. Rev. 88:211-232.

3. Buckel, W., and R. Semmler. 1982. A biotin-dependent sodiumpump: glutaconyl-CoA decarboxylase from Acidaminococcus fer-mentans. FEBS Lett. 148:35-38.

4. Buckel, W., and R. Semmler. 1983. Purification, characterisationand reconstitution of glutaconyl-CoA decarboxylase, a biotin-dependent sodium pump from anaerobic bacteria. Eur. J. Bio-chem. 136:427-434.

5. Chen, G., and J. B. Russell. 1990. Transport and deamination ofamino acids by a gram-positive, monensin-sensitive ruminal bac-terium. Appl. Environ. Microbiol. 56:2186-2192.

6. Cook, G. M., F. A. Rainey, G. Chen, E. Stackebrandt, and J. B.Russell. 1994. Emendation of the description of Acidominococcusfermentans, a trans-aconitate- and citrate-oxidizing bacterium. Int.J. Syst. Bacteriol. 44:576-578.

7. Cook, G. M., J. E. Wells, and J. B. Russell. 1994. Ability ofAcidaminococcus fermentans to oxidize trans-aconitate and pre-vent the accumulation of tricarballylate, a toxic end product ofruminal fermentation. Appl. Environ. Microbiol. 60:2533-2537.

8. Cotta, M. A., and J. B. Russell. 1982. Effect of peptides and aminoacids on the efficiency of rumen bacterial protein synthesis in

continuous culture. J. Dairy Sci. 65:226-234.9. Daron, H. H., and I. C. Gunsalus. 1962. Citratase and isocitratase.

Methods Enzymol. 5:622-627.10. Dimroth, P. 1981. Characterization of a membrane-bound biotin-

containing enzyme: oxaloacetate decarboxylase from Klebsiellaaerogenes. Eur. J. Biochem. 115:353-358.

11. Dimroth, P. 1986. Preparation, characterization and reconstitutionof oxaloacetate decarboxylase from Klebsiella aerogenes, a sodiumpump. Methods Enzymol. 125:530-540.

12. Dimroth, P. 1987. Sodium ion transport decarboxylases and otheraspects of sodium cycling in bacteria. Microbiol. Rev. 51:320-340.

13. Eagon, R. G., and L. S. Wilkerson. 1972. A potassium-dependentcitric acid transport system in Aerobacter aerogenes. Biochem.Biophys. Res. Commun. 46:1944-1950.

14. Gottschalk, G. 1986. Bacterial metabolism, p. 233. Springer Ver-lag, New York.

15. Hugenholtz, J., L. Perdon, and T. Abee. 1993. Growth and energygeneration by Lactococcus lactis subsp. lactis biovar diacetylactisduring citrate metabolism. Appl. Environ. Microbiol. 59:4216-4222.

16. Kaback, R. H. 1990. Active transport: membrane vesicles, bioen-ergetics, molecules, and mechanisms. In T. A. Krulwich (ed.),Bacterial energetics. Academic Press, Inc., New York.

17. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall.1951. Protein measurement with the Folin phenol reagent. J. Biol.Chem. 193:265-275.

18. Mayland, H. F., D. L. Grunes, and V. A. Lazar. 1976. Grass tetanyhazard of cereal forages based upon chemical composition. Agron.J. 68:665-667.

19. Paster, B., J. B. Russell, C. M. J. Yang, J. M. Chow, C. R. Woese,and R. Tanner. 1993. Phylogeny of ammonia-producing ruminalbacteria, Peptostreptococcus anaerobius, Clostridium sticklandii andClostridium aminophilum sp. nov. Int. J. Syst. Bacteriol. 43:107-110.

20. Pressman, B. C. 1976. Biological applications of ionophores.Annu. Rev. Biochem. 45:501-530.

21. Prior, R. L., D. L. Grunes, R. P. Patterson, F. W. Smith, H. F.Mayland, and W. Viselk 1973. Partition column chromatographyfor quantitating effects of fertilization on plant acids. J. Agric.Food Chem. 21:73-77.

22. Reibeling, V., R. K. Thauer, and K. Jungermann. 1975. Theinternal-alkaline pH gradient, sensitive to uncoupler and ATPaseinhibitor, in growing Clostridium pasteurianum. Eur. J. Biochem.55:445-453.

23. Rogosa, M. 1969. Acidaminococcus gen. n., Acidaminococcusfermentans sp. n., anaerobic gram-negative diplococci using aminoacids as the sole energy source for growth. J. Bacteriol. 98:756-766.

24. Rogosa, M. 1984. Genus II. Acidaminococcus, p. 684. In N. R.Kreig and J. G. Holt (ed.), Bergey's manual of systematic bacte-riology, vol. 1. The Williams and Wilkins Co., Baltimore.

25. Rose, I. A. 1955. Acetate kinase of bacteria (acetokinase). Meth-ods Enzymol. 1:591-595.

26. Russell, J. B. 1985. Enrichment and isolation of rumen bacteriathat reduce trans-aconitic acid to tricarballylic acid. Appl. Environ.Microbiol. 49:120-126.

27. Russell, J. B., and N. Forsberg. 1986. Production of tricarballylicacid by rumen microorganisms and its potential toxicity in rumi-nant tissue metabolism. Br. J. Nutr. 56:153-162.

28. Russell, J. B., and H. F. Mayland. 1986. Absorption of tricarbal-lylic acid from the rumen of sheep and cattle fed forages contain-ing trans-aconitic acid. J. Sci. Food Agr. 40:205-212.

29. Russell, J. B., and H. J. Strobel. 1989. Effect of ionophores onruminal fermentation. Appl. Environ. Microbiol. 55:1-6.

30. Russell, J. B., H. J. Strobel, and S. A. Martin. 1990. Strategies ofnutrient transport by ruminal bacteria. J. Dairy Sci. 73:2996-3012.

31. Russell, J. B., and P. J. Van Soest. 1984. In vitro ruminalfermentation of organic acids common in forage. Appl. Environ.Microbiol. 47:155-159.

32. Schwartz, R., M. Topley, and J. B. Russell. 1988. Effect oftricarballylic acid, a non-metabolizable rumen fermentation prod-uct, on Mg, Ca, and Zn utilization. J. Nutr. 118:183-188.

33. Speelmans, G., B. Poolman, T. Abee, and W. N. Konings. 1993.

VOL. 60, 1994

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 11

Feb

ruar

y 20

22 b

y 14

8.71

.171

.138

.

2544 COOK AND RUSSELL

Energy transduction in the thermophilic anaerobic bacteriumClostridium fervidus is coupled to sodium ions. Proc. Natl. Acad.Sci. USA 90:7975-7979.

34. Sweet, G. D., C. M. Kay, and W. W. Kay. 1984. Tricarboxylatebinding proteins of Salmonella typhimurium. J. Biol. Chem. 259:1586-1592.

APPL. ENVIRON. MICROBIOL.

35. Van Der Rest, M. E., D. Molenaar, and W. N. Konings. 1992.Mechanism of Na+-dependent citrate transport in Klebsiella pneu-moniae. J. Bacteriol. 174:4893-4898.

36. Wolfarth, G., and W. Buckel. 1985. A sodium ion gradient asenergy source for Peptostreptococcus. asaccharolyticus. Arch. Mi-crobiol. 142:128-135.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 11

Feb

ruar

y 20

22 b

y 14

8.71

.171

.138

.