INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Biodegradation of aromatic compounds is an important component of the carbon cycle in various anoxic environments. Despite the large number of natural and synthetic homocyclic aromatic compounds, anaerobic microorganisms initially channel all aromatic substrates into a few central intermediates prior to ring cleavage (20). Benzoyl coenzyme A (CoA) is the most important of these intermediates since a large number of compounds, such as chloro-, nitro-, and aminobenzoates, aromatic hydrocarbons, and phenolic compounds, are initially converted to benzoyl-CoA prior to ring reduction and cleavage (18). The central pathways for benzoate and benzoyl-CoA metabolism under anaerobic conditions have been studied primarily in two microorganisms, the phototrophic, purple, nonsulfur bacterium Rhodopseudomonas palustris and the nitrate-reducing bacterium Thauera aromatica (18, 19). After activation of benzoate to benzoyl-CoA (1, 17), benzoyl-CoA is reduced to cyclohex-1,5-diene carboxyl-CoA by a benzoyl-CoA reductase, which has been purified from T. aromatica (8, 9, 26). Based on DNA sequence homology, it is believed that a similar reductive reaction occurs in R. palustris (14). After ring reduction, the pathways diverge in the two organisms. In T. aromatica, cyclohex-1,5-diene carboxyl-CoA is hydrated to 6-hydroxycyclohex-1-ene carboxyl-CoA (28). The latter compound is oxidized to 6-ketocyclohex-1-ene carboxyl-CoA, which is then hydrolytically cleaved to 3-hydroxypimelyl-CoA (29). The pathway in R. palustris is similar except that cyclohex-1,5-diene carboxyl-CoA is most probably reduced to cyclohex-1-ene carboxyl-CoA. The latter compound is metabolized to 2-ketocyclohexane carboxyl carboxyl-CoA, which is hydrolytically cleaved to pimelyl-CoA. The C₇ ring cleavage products then undergo -oxidation, which yields three molecules of acetate and one molecule of CO₂.

Benzoate degradation also occurs under methanogenic conditions (37, 49). Tarvin and Buswell (49) observed degradation of benzoate in anoxic sediments with production of carbon dioxide and methane as the final end products. The discovery that methanogenic benzoate degradation to carbon dioxide and methane is mediated by a consortium of a fermentative (syntrophic) microorganism and hydrogen- and acetate-utilizing methanogens (15) and the subsequent isolation of the syntrophic partners (36) provided the opportunity to study the pathway for benzoate degradation under methanogenic conditions. So far, three species that syntrophically metabolize benzoate have been isolated (22, 36, 51), and all of these species belong to the genus Syntrophus. Benzoate degradation under syntrophic conditions has not been investigated as thoroughly as benzoate degradation under nitrate-reducing and phototrophic conditions due to the relatively slow growth rates and low cell yields of these organisms (3). However, recent studies have shown that benzoate is activated to benzoyl-CoA by an ATP-dependent ligase as the first step in benzoate metabolism (4, 46). Also, enzyme activities for glutaryl-CoA metabolism to acetate and CO₂ have been detected in cell extracts of Syntrophus gentianae (46), and glutaryl-CoA dehydrogenase and the enzyme activities responsible for crotonyl-CoA metabolism to acetate have been detected in Syntrophus buswellii GA (2). The ring reduction and cleavage steps required for syntrophic benzoyl-CoA metabolism have not been investigated yet.

In this study, we investigated the pathway for syntrophic benzoate metabolism in "Syntrophus aciditrophicus" strain SB by identifying and quantifying metabolites produced during growth of this organism with benzoate, cyclohexane carboxylate, and cyclohex-1-ene carboxylate in cocultures with hydrogen-utilizing partners and by measuring the key enzyme activities postulated to be involved in benzoate metabolism. Below, we describe transient production of cyclohex-1-ene carboxylate and relatively larger amounts of cyclohexane carboxylate during benzoate degradation. This is consistent with the hypothesis that benzoyl-CoA reduction during syntrophic benzoate metabolism may involve a four- or six-electron reduction (46, 47). We hypothesize that the difference in benzoyl-CoA metabolism from that observed in R. palustris and T. aromatica may be due to the energetic constraints imposed by syntrophic metabolism of aromatic substrates.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Microorganisms and media. "S. aciditrophicus" SB^T (= ATCC 700169^T) was isolated from a sewage treatment plant in Norman, Okla. (22). Methanospirillum hungatei JF1 and Desulfovibrio sp. strain G11 were obtained from the culture collection of M. P. Bryant (Urbana, Ill.). All media and stock solutions were prepared anaerobically by the techniques described by Balch and Wolfe (6). The organisms were grown in a basal medium (33) lacking rumen fluid. To grow "S. aciditrophicus" in pure culture, crotonate (40 mM) was added to the basal medium (7) and the headspace was pressurized to 172 kPa with an 80% N₂-20% CO₂ gas mixture. M. hungatei and Desulfovibrio sp. strain G11 were grown in the basal medium containing 2 mM sodium acetate in the presence of 243 kPa of 80% H₂-20% CO₂. Sodium sulfate (15 mM) was included in the medium when Desulfovibrio sp. strain G11 was present. M. hungatei and Desulfovibrio sp. strain G11 cultures were incubated in a shaking incubator (100 rpm). Cocultures of "S. aciditrophicus" and M. hungatei or "S. aciditrophicus" and Desulfovibrio sp. strain G11 were established by adding a 15 to 20% (vol/vol) inoculum of each microorganism to the basal medium containing 1.2 to 1.5 mM sodium benzoate, sodium cyclohexane carboxylate, or sodium cyclohex-1-ene carboxylate as the substrate in the presence of a headspace containing 80% N₂ and 20% CO₂ (172 kPa). All inoculations were performed by using sterile disposable plastic syringes and needles that were degassed with oxygen-free nitrogen gas. All cultures were incubated at 37°C. T. aromatica DSM 6984 was obtained from the Deutsche Sammlung von Mikroorganismen (Braunschweig, Germany) and was cultured anaerobically at 28°C in a benzoate-nitrate medium (50). R. palustris CGA009 was kindly provided by Caroline S. Harwood and cultured as previously described (17).

Detection and quantification of metabolites by GC-MS. Cocultures of "S. aciditrophicus" and either M. hungatei or Desulfovibrio sp. strain G11 were grown in 600 ml of basal medium with 1.4 mM sodium benzoate to detect metabolites of benzoate metabolism. Samples (60 ml) were withdrawn from the cultures at various times. The pH of each sample was brought to more than 12 for 30 min by stepwise addition of 1 N NaOH to hydrolyze putative thioester bonds. Each sample was then acidified to a pH of less than 2 with 12 N HCl. The samples were then extracted three times with 25-ml aliquots of ethyl acetate. The ethyl acetate extracts were filtered through anhydrous sodium sulfate to remove water, combined, and then concentrated to volumes of 2 to 3 ml under a vacuum. The concentrated ethyl acetate extract was then quantitatively transferred to 6-ml vials and evaporated to dryness under a stream of nitrogen gas. The dried ethyl acetate extract was then redissolved in 0.3 ml of ethyl acetate and derivatized with N,O-bis-(trimethylsilyl)triflouroacetamide (Pierce Chemicals, Rockford, Ill.). Each concentrated and derivatized extract was analyzed with a Hewlett-Packard 5890 series II gas chromatograph (GC) equipped with a Hewlett-Packard 5970 series mass spectrometer (MS) and a 30-m DB-5 fused silica capillary column (J & W Scientific, Folson, Calif.). Helium was used as the carrier gas at a flow rate of 0.8 ml/min. The oven temperature was held at 70°C for 5 min, increased at a rate of 10°C/min to 220°C, and then held at 220°C for 5 min. The controls for this experiment included heat-killed cocultures containing "S. aciditrophicus" and non-substrate-amended cocultures to differentiate between metabolites formed due to benzoate metabolism and metabolites present in the inoculum of crotonate-grown "S. aciditrophicus" cultures or H₂-grown M. hungatei and Desulfovibrio sp. strain G11 cultures. All treatments were performed in triplicate. The metabolites were identified by comparing their retention times and mass spectral profiles with the retention times and mass spectral profiles of trimethylsilyl (TMS)-derivatized chemical standards and were quantified by comparison to standard curves constructed with the TMS derivatives of the compounds of interest. The detection limits ranged from 0.07 µM (pimelic acid) to 0.14 µM (cyclohexane carboxylic acid) under the experimental conditions used. The benzoate concentrations calculated by GC-MS analysis at different times were within ±10% of the benzoate concentrations determined by high-performance liquid chromatography analysis. R. palustris and T. aromatica cultures were grown with similar benzoate concentrations, and the samples were processed as described above. A similar protocol was used for detection of metabolites in cyclohexane carboxylate- or cyclohex-1-ene carboxylate-grown "S. aciditrophicus"-M. hungatei cocultures.

NMR spectroscopy. Cocultures of "S. aciditrophicus" and M. hungatei were grown with 1.5 mM [ring-¹³C₆] benzoate. Samples (100 ml) were withdrawn from the cultures at various times and acidified to a pH of less than 2 by dropwise addition of 12 N HCl. Each sample was extracted three times with ethyl acetate, concentrated under a vacuum, and dried under an N₂ atmosphere as described above. The dried samples were then dissolved in deuterated chloroform (CDCl₃). Non-substrate-amended and heat-killed controls were included as described above. A coculture of "S. aciditrophicus" and M. hungatei with unlabeled benzoate was used to ensure that none of the peaks observed in the nuclear magnetic resonance (NMR) spectra were due to ¹³C impurities in the solvents used. The NMR spectra of the organic solvent-extracted samples were obtained with a Unity INOVA 400-MHz NMR spectrometer (Varian) with a ¹³C resonance frequency of 100.573 MHz. The ¹³C spectra were obtained at 30°C by using a standard inverse-gated pulse sequence. The following experimental parameters were used: sweep width, 24,140 Hz; acquisition time, 1.00 s; and recycle delay, 8 s. The number of scans ranged from 500 to 36,000 depending on the sample concentration. The data were processed with 1-Hz line broadening.

Other analytical procedures. Protein was determined as described previously (10). Growth was monitored by measuring absorbance at 600 nm. Benzoate was analyzed by high-performance liquid chromatography as described previously (21). Methane was analyzed by GC (23), and sulfate was analyzed by ion chromatography (32). Enzyme activities were determined with either a Beckman DU-64 spectrophotometer or a Shimadzu 2101-PC dual-beam spectrophotometer.

Preparation of cell extracts. Cells were harvested by centrifugation (12,000 × g, 20 min, 4°C) and were washed by resuspending and recentrifuging the cell pellets three times in anoxic 100 mM Tris-HCl buffer (pH 7.8). The final cell pellet was suspended in the same buffer (0.2 g of cells/ml) containing 1 mM MgCl₂, 2 mM dithiothreitol (DTT), and 0.2 mg of DNase per ml. Cells were broken under anaerobic conditions by two passages through a chilled French pressure cell at 110,400 kPa. Unbroken cells and cell debris were removed by centrifugation at 27,200 × g for 20 min at 4°C. The resulting supernatant, termed the cell extract, was used immediately in enzyme assays or was stored anaerobically in liquid nitrogen until it was used.

Enzyme assays. Acyl-CoA ligase activity was assayed by measuring the amount of AMP formed in the CoA ligase reaction by a coupled enzyme assay (4). The reaction was initiated by adding 10 to 50 µl of cell extract, and then oxidation of NADH was monitored at 340 nm. Formation of 1 mol of AMP corresponded to oxidation of 2 mol of NADH. The substrates tested included benzoate, 2-, 3-, and 4- chlorobenzoates, 2-, 3-, and 4-fluorobenzoates, 4-hydroxybenzoate, picolinic acid, phenyl acetate, crotonate, n-butyrate, isobutyrate, heptanoate, and hexanoate. Formation of benzoyl-CoA from benzoate and CoA in cell extracts was confirmed by using [phenyl-¹⁴C] benzoate (56.9 mCi/mmol) as the substrate and a previously described procedure (17). The reaction was stopped after 2 min, and the assay mixture was extracted twice with ethyl acetate.

Chemical synthesis. 2-Hydroxycyclohexane carboxylic acid was synthesized by reducing ethyl 2-cyclohexanone carboxylate with sodium borohydride in 95% ethanol as previously described (40). 2-Ketocyclohexane carboxylic acid was also synthesized from ethyl-2-cyclohexanone (12, 40). 2-Ketocyclohexane carboxyl-CoA, cyclohex-1-ene carboxyl-CoA, pimelyl-CoA, and glutaconyl-CoA were synthesized by reacting 2-ketocyclohexane carboxylic acid, cyclohex-1-ene carboxylic acid, pimelic acid, and glutaconic acid, respectively, with free CoA by using a procedure modified from the procedures of Merckel et al. (35) and Gallus and Schink (16). Crude preparations of the CoA thioesters were purified by using C₁₈ reversed-phase cartridges (Sep-Pak Plus; Millipore Corp., Midford, Mass.) as previously described (40). Ferricenium hexafluorophosphate was synthesized from ferrocine and sodium hexafluorophosphate (30).

Chemicals. Sodium benzoate was purchased from Sigma Chemical Co. (St. Louis, Mo.), cyclohexane carboxylic acid was purchased from Acros Organics (Fair Lawn, N.J.), and cyclohex-1-ene carboxylic acid was purchased from Aldrich Chemical Co. (Milwaukee, Wis.). [ring-¹³C₆]benzoate and CDCl₃ were purchased from Cambridge Isotope Laboratories (Andover, Mass.). S-Acetoacetyl-CoA, benzoyl-CoA, glutaryl-CoA, butyryl-CoA, octanoyl-CoA, CoA (sodium salt), ferrocine, sodium hexafluorophosphate, phosphoenolpyruvate, pyruvate kinase, myokinase, lactate dehydrogenase, citrate synthase, NADH, NAD⁺, crotonase, L-(+)-3-hydroxyacyl-CoA dehydrogenase, acetyl phosphate, and ATP were purchased from Sigma Chemical Co. All other chemicals used in this study were obtained from Sigma, Aldrich, or Fluka (Milwaukee, Wis.).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Metabolites produced during benzoate degradation. During growth of the "S. aciditrophicus"-M. hungatei cocultures with benzoate, cyclohexane carboxylate, cyclohex-1-ene carboxylate, pimelate, and glutarate were detected as their TMS derivatives. The TMS derivative of each compound had the same retention time and mass spectral profile as the the TMS derivative of the authentic chemical standard (Fig. 1). None of these compounds were detected in non-substrate-amended or heat-killed controls. Cyclohexane carboxylate (Fig. 1A) was transiently produced and accumulated to a maximum concentration of 260 µM in culture fluids (Fig. 2). The maximum concentration of cyclohexane carboxylate occurred when about 97.5% of the benzoate was consumed (Fig. 2). Cyclohex-1-ene carboxylate, pimelate, and glutarate (Fig. 1B to D) were also transiently produced and consumed, but these compounds were detected at much lower concentrations (Fig. 2). The maximum concentrations observed were 12.5, 6.4, and 11.3 µM for cyclohex-1-ene-carboxylate, pimelate, and glutarate, respectively, which represented 0.45 to 0.9% of the initial amount of benzoate. These four compounds were detected at similar concentrations regardless of whether the alkaline hydrolysis step was included and whether the entire culture or the cell-free culture fluid was analyzed.

View larger version (34K): [in this window] [in a new window]   FIG. 1.   Mass spectra of TMS derivatives of metabolites detected in "S. aciditrophicus"-M. hungatei benzoate-grown cocultures (upper spectra) and TMS derivatives of authentic chemical standards (lower spectra). (A) cyclohexane carboxylate-TMS; (B) cyclohex-1-ene carboxylate-TMS; (C) pimelate-TMS; (D) glutarate-TMS.

View larger version (23K): [in this window] [in a new window]   FIG. 2.   Transient detection of cyclohexane carboxylate (), cyclohex-1-ene carboxylate (), pimelate (), and glutarate () in "S. aciditrophicus"-M. hungatei cocultures grown with benzoate (). Each compound was detected and quantified as its TMS derivative.

Metabolism of cyclohexane carboxylate and cyclohex-1-ene carboxylate by "S. aciditrophicus"-M. hungatei cocultures. Cyclohex-1-ene carboxylate and cyclohexane carboxylate were metabolized without a lag by "S. aciditrophicus"-M. hungatei cocultures at rates slightly faster than the rate observed for benzoate (Table 2). Cocultures of "S. aciditrophicus" and M. hungatei grown with cyclohex-1-ene carboxylate and cyclohexane carboxylate were analyzed by GC-MS as described above. In cyclohex-1-ene carboxylate-grown cocultures, pimelate and glutarate were detected as their TMS derivatives at levels comparable to those found in benzoate-grown cultures; the maximum concentrations measured were 6.6 and 20.7 µM, respectively. Although cyclohexane carboxylate was detected (as its TMS derivative) in cyclohex-1-ene carboxylate-grown cultures, it was not produced at as high a concentration as observed in benzoate-grown cocultures. Instead, it was detected at levels comparable to the levels of pimelate and glutarate (e.g., at a maximum concentration of 4.2 µM). In addition, 2-hydroxycyclohexane carboxylate (Fig. 3) was detected as its TMS derivative in trace amounts (<0.5 µM). In cyclohexane carboxylate-grown cocultures, cyclohex-1-ene carboxylate, pimelate, and glutarate were detected. The maximum concentrations of these compounds in cyclohexane carboxylate-grown cultures were also comparable to those measured in benzoate-grown cocultures, (8.9, 6.1, and 15.7 µM for cyclohex-1-ene carboxylate, pimelate, and glutarate, respectively). None of the compounds mentioned above were detected in heat-killed or non-substrate-amended controls.

View larger version (19K): [in this window] [in a new window]   FIG. 3.   Mass specta of a TMS derivative of a metabolite detected in "S. aciditrophicus"-M. hungatei cyclohex-1-ene carboxylate-grown cocultures (A) and a synthesized TMS-derivatized 2-hydroxycyclohexane carboxylic acid standard (B).

Enzyme activities detected in cell extracts of "S. aciditrophicus." Cell extracts of benzoate-grown cocultures of "S. aciditrophicus" and Desulfovibrio sp. strain G11 contained an ATP-dependent benzoyl-CoA ligase activity (Table 3). Formation of benzoyl-CoA from benzoate was confirmed by the isotopic assay by using [phenyl-¹⁴C]benzoate (17). The results showed that 84% of the [¹⁴C]benzoate added was converted to benzoyl-CoA after 2 min. In addition to benzoate, 2-, 3-, and 4-fluorobenzoates, as well as 3-hydroxybenzoate, also served as substrates. No activity was detected with 4-hydroxybenzoate, 2-, 3-, or 4-chlorobenzoate, picolinic acid, phenylacetate, 4-hydroxyphenylacetate, crotonate, n-butyrate, isobutyrate, heptanoate, or hexanoate. No acyl-CoA ligase activity was detected in cell extracts of pure cultures of Desulfovibrio sp. strain G11 (data not shown). Similar levels of acyl-CoA activities were detected in crotonate-grown pure cultures of "S. aciditrophicus," suggesting that this activity was constitutively present in "S. aciditrophicus." Cell extracts of benzoate-grown cocultures contained very low levels of a benzoyl-CoA transferase activity. This activity was not detected in cell extracts of Desulfovibrio sp. strain G11 or pure cultures of "S. aciditrophicus."

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Using GC-MS analysis, we identified cyclohexane carboxylate, cyclohex-1-ene carboxylate, pimelate, and glutarate in benzoate-grown cocultures of "S. aciditrophicus" and M. hungatei based on a comparison of the retention times and mass spectra of their TMS derivatives to the retention times and mass spectra of authentic chemical standards (Fig. 1 and 2). In addition to pimelate and glutarate, we also detected cyclohexane carboxylate and 2-hydroxycyclohexane carboxylate (Fig. 3) in cyclohex-1-ene carboxylate-grown cocultures and cyclohex-1-ene carboxylate in cyclohexane carboxylate-grown cocultures. These compounds were transiently produced and were not detected in heat-killed or non-substrate-amended controls, providing strong evidence that these compounds, probably as their CoA derivatives, are intermediates in the metabolism of benzoate, cyclohexane carboxylate, and cyclohex-1-ene carboxylate. ¹³C NMR spectroscopy confirmed that cyclohexane carboxylate and cyclohex-1-ene carboxylate were produced from benzoate. The fact that cyclohexane carboxylate, cyclohex-1-ene carboxylate, pimelate, and glutarate were detected when the alkaline hydrolysis step was omitted and were found in whole-culture broth (cells plus medium) and in cell-free culture broth suggests that "S. aciditrophicus" may excrete potential intermediates of benzoate and alicyclic acid metabolism. Several studies have detected free acids corresponding to potential intermediates in phenol (5) and benzoate (13) metabolism. Since "S. aciditrophicus" cocultures and pure cultures contain enzymatic activities that metabolize the various alicyclic and aliphatic CoA intermediates of benzoate metabolism (Table 3), this finding suggests that the CoA derivatives of the compounds mentioned above and not the free acids are the functional intermediates involved in benzoate and alicyclic acid metabolism.

Cyclohexane carboxylate was originally thought to be an intermediate in benzoate metabolism by R. palustris (13). However, subsequent investigations of cyclohexane carboxylate metabolism in this microorganism (27), as well as a better understanding of the biochemistry and genetics of benzoyl-CoA metabolism in R. palustris (38-40) and T. aromatica (28, 29), led to exclusion of cyclohexane carboxylate from the benzoyl-CoA pathway in both microorganisms. It was suggested that the transient production of cyclohexane carboxylate in R. palustris is a physiological response to excess reducing equivalents or is a mechanism to maintain appropriate intracellular levels of free CoA (27). Some of the earlier studies on benzoate degradation by methanogenic consortia from sewage sludge, rumen fluid, or anaerobic mud (15, 24, 37) also suggested that cyclohexane carboxylate is a benzoate degradation intermediate based on cochromatography data and the ability of the consortia to degrade cyclohexane carboxylate without a lag period. Given the number of different organisms that are present in such consortia, it was difficult to reach definitive conclusions regarding potential intermediates involved in the degradation of benzoate in these studies. However, we clearly show that relatively large amounts of cyclohexane carboxylate, accounting for about 18% of the benzoate carbon, are formed and consumed by "S. aciditrophicus" during growth with benzoate. These data suggest that cyclohexane carboxylate may serve as a repository for reducing equivalents generated during benzoate metabolism.

Detection of cyclohexane carboxylate, as well as cyclohex-1-ene carboxylate, pimelate, and glutarate, at levels not encountered in R. palustris or T. aromatica suggests that the pathway for benzoate metabolism in syntrophic bacteria is different from that in phototrophs and nitrate reducers. This difference is not surprising for two reasons. First, members of the genus Syntrophus, which belong to the  subgroup of the class Proteobacteria, T. aromatica, which belongs to the  subgroup of the Proteobacteria, and R. palustris, which belongs to the  subgroup of the Proteobacteria, are phylogenetically distinct. Second, the energy constraints encountered in syntrophic aromatic metabolism (45-47) are not encountered by organisms that can obtain energy from respiration and photosynthesis. It is obvious that syntrophic microorganisms can obtain energy for growth on benzoate, yet it is not clear how net ATP production occurs if benzoate activation and ring reduction during syntrophic benzoate metabolism occur, as found in benzoate-degrading phototrophs and denitrifiers (18, 46, 47). This study and other studies (4, 46) demonstrated that benzoate activation proceeds by a benzoyl-CoA ligase reaction which uses two energy-rich bonds. We also detected a benzoate:acetyl-CoA transferase reaction that is similar to the succinyl-CoA:benzylsuccinate transferase reaction that has been suggested for benzylsuccinate activation in T. aromatica (31), which used only the equivalent of one energy-rich bond for benzoate activation. However, the level of activity of this enzyme in benzoate-grown cocultures is very low and is probably not sufficient to account for benzoate metabolism. While our evidence to date cannot exclude the possibility that there is a two-electron reduction of benzoyl-CoA to cyclohex-1,5-diene carboxylate, the transient production of large amounts of cyclohexane carboxylate suggests that ring reduction may differ in syntrophic metabolism. Thermodynamic calculations indicate that if the product of benzoyl-CoA reduction is cyclohex-1-ene carboxyl-CoA or cyclohexane carboxyl-CoA instead of cyclohex-1,5-diene carboxyl-CoA, the product observed in T. aromatica (9), then the reaction could proceed without ATP investment (44, 46). The detection of cyclohexane carboxylate and cyclohex-1-ene carboxylate in cell extracts is consistent with a four- or six-electron reduction step rather than a two-electron reduction step.

Although our work demonstrates that cyclohexane carboxylate is produced and consumed by benzoate-grown cultures of "S. aciditrophicus," the role of this compound (or its CoA derivative) in syntrophic benzoate degradation is unclear. Cyclohexane carboxyl-CoA might be the product formed by a six-electron reduction of benzoyl-CoA. Cyclohexane carboxyl-CoA could then be oxidized to cyclohex-1-ene carboxyl-CoA by cyclohexane carboxyl-CoA dehydrogenase, and the latter compound could be metabolized in a manner similar to that observed in R. palustris (Fig. 4). However, this does not explain why cyclohexane carboxylate is produced at a level far higher than the levels of other metabolites. Another possible explanation is that cyclohexane carboxylate is produced in a dismutation reaction in which the reducing equivalents produced during oxidation of one benzoyl-CoA molecule are used to reduce another benzoyl-CoA molecule to cyclohexane carboxyl-CoA (Fig. 4). This scheme would require two distinct enzymes to reduce benzoyl-CoA, one enzyme to reduce benzoyl-CoA to cyclohex-1-ene carboxyl-CoA and another enzyme to reduce benzoyl-CoA to cyclohexane carboxyl-CoA. It has recently been suggested that benzoate dismutation (simultaneous oxidation and reduction of benzoate) is an alternative mechanism for hydrogen removal in methanogenic, benzoate-degrading enrichments in which the electron flow to methanogenesis is inhibited by bromoethanesulfonic acid (25). However, if such a reaction occurs in "S. aciditrophicus," it is not clear what factors determine whether the reducing equivalents produced during benzoate metabolism form hydrogen or are used to reduce benzoyl-CoA to cyclohexane carboxyl-CoA. A third possibility is that benzoyl-CoA is reduced to cyclohex-1-ene carboxyl-CoA and then a dismutation reaction occurs in which oxidation of one molecule of cyclohex-1-ene carboxyl-CoA is coupled to reduction of three molecules of cyclohex-1-ene carboxyl-CoA (Fig. 4). However, the fact that cyclohex-1-ene carboxylate-grown cocultures do not accumulate cyclohexane carboxylate at levels comparable to those in benzoate-grown cultures argues against this possibility.

View larger version (24K): [in this window] [in a new window]   FIG. 4.   Proposed pathway for benzoate metabolism in "S. aciditrophicus." Enzyme activities detected in "S. aciditrophicus" cell extracts are indicated by boldface type. The asterisks indicate compounds detected as their TMS derivatives in "S. aciditrophicus"-M. hungatei coculture extracts. 1, Benzoate; 2, benzoyl-CoA; 3, cyclohexane carboxyl-CoA; 4, cyclohex-1-ene carboxyl-CoA; 5, 2-hydroxycyclohexane carboxyl-CoA; 6, 2-ketocyclohexane carboxyl-CoA; 7, pimelyl-CoA; 8, 2-heptenedioc acid-1-CoA; 9, 3-hydroxypimelyl-CoA; 10, 3-ketopimelyl-CoA; 11, glutaryl-CoA; 12, glutaconyl-CoA; 13, crotonyl-CoA; 14, 3-hydroxybutyryl-CoA; 15, acetoacetyl-CoA; 16, acetyl-CoA; 17, acetate. The question mark indicates that the exact benzene ring reduction mechanism in "S. aciditrophicus" is not completely understood yet.

Enzyme studies in which "S. aciditrophicus" cell extracts were used indicated that the organism metabolizes cyclohex-1-ene carboxyl-CoA by a pathway similar to that observed in R. palustris. It could be argued that the same enzymes could also catalyze cyclohex-1,5-diene transformation to 6-hydroxycyclohex-1-ene carboxylate, 2-keto-6-hydroxycyclohex-1-ene carboxyl-CoA, and 3-hydroxypimelyl-CoA (19) since the latter substrates were not tested in our enzyme assays because they were not available commercially. However, detection of cyclohex-1-ene carboxylic acid, 2-hydroxycyclohexane carboxylic acid, and pimelic acid in culture extracts is consistent with a pathway similar to the R. palustris pathway. Also, the twofold-higher level of pimelyl-CoA dehydrogenase in benzoate-grown cells than in crotonate-grown cells is consistent with production of pimelate as the ring cleavage product (40). This activity was specific for pimelyl-CoA since benzoate-grown cells of "S. aciditrophicus" lacked detectable activity when butyryl-CoA, octanoyl-CoA, and palmitoyl-CoA were used as substrates. Thus, it is unlikely that the enzyme activity detected in benzoate-grown cells was due to the presence of a nonspecific acyl-CoA dehydrogenase. The pimelyl-CoA dehydrogenase activity, which was demonstrated for the first time in a benzoate-grown anaerobic microorganism, was not detected in a nitrate-reducing isolate when it was grown with benzoate (16), probably because benzoate degradation in the isolate proceeds via transformation of cyclohex-1,5-diene carboxyl-CoA to 3-hydroxypimelate without the formation of pimelyl-CoA. The dehydrogenase activity was detected, however, when the same isolate was grown on pimelate (16). It could also be argued that the observed cyclohex-1-ene carboxyl-CoA hydratase activity is due to the action of an enoyl-CoA hydratase, which acts primarily on short-chain unsaturated fatty acids (28). However, detection of cyclohex-1-ene carboxyl-CoA hydratase at 28-fold-higher levels when the organism was grown with benzoate than when it was grown with crotonate and detection of enoyl-CoA hydratase at similar levels in benzoate- and crotonate-grown cells indicate that a specific cyclohex-1-ene carboxyl-CoA hydratase activity is present in "S. aciditrophicus."

An intriguing observation is the presence of both ¹²C and ¹³C carbons in the carboxyl and methyl moieties of acetate. Acetate molecules with ¹³C only in the methyl group most likely arose from pimelate in which one of the carboxyl groups was the original unlabeled carboxyl group of benzoate. The presence of acetate molecules that have ¹³C only in the carboxyl group is less easily understood since the use of a ring-¹³C₆-labeled compound should result in acetate molecules with both carbons ¹³C labeled, except for the situation discussed above. It is possible that the glutaconyl-CoA decarboxylation reaction is in equilibrium with the bicarbonate pool and that this reaction allows an exchange of ¹²C and ¹³C atoms to occur.

In conclusion, although syntrophic benzoate metabolism follows the main themes observed in nitrate reducers and phototrophs, it appears that a third variant of the benzoyl-CoA pathway in which cyclohexane carboxylic acid is produced from benzoate may operate in microorganisms that syntrophically degrade benzoate (Fig. 4). The variation is probably imposed by the strict energy constraints of syntrophic metabolism. Detailed investigations are still required to determine the exact function of cyclohexane carboxylic acid in syntrophic benzoate metabolism and to determine how syntrophic microorganisms are able to obtain net ATP production to support growth on benzoate. The fact that three variants of the benzoyl-CoA pathway have been detected in the three groups of microorganisms that have been studied so far raises the question of how similar or different benzoate metabolism is in other physiological groups of microorganisms, such as sulfate-reducing and iron-reducing bacteria that have yet to be studied.