INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Of the chlorinated aliphatic hydrocarbons, only chloromethane and dichloromethane are known to support growth of strictly anaerobic bacteria (17). The homoacetogenic bacterium strain MC grows with chloromethane but not with dichloromethane as an energy source (27), and the strictly anaerobic gram-positive bacterium Dehalobacterium formicoaceticum utilizes only dichloromethane as a growth substrate (19). Dehalogenation of chloromethane by strain MC has been shown to be catalyzed by a chloromethane dehalogenase which transfers the methyl group of chloromethane onto tetrahydrofolate and thereby produces inorganic chloride and methyl tetrahydrofolate. In crude cell extracts, some properties of this enzyme have been determined (22), and the further metabolism of methyl tetrahydrofolate to acetate via the reactions of the CO dehydrogenase pathway has been established (20).

Much less is known about the metabolism of dichloromethane by D. formicoaceticum. The compound has been shown to be converted by growing cultures to formate and acetate in a molar ratio of 2:1 and to biomass. This observation and the fact that the key enzymes of the CO dehydrogenase pathway are present in cell extracts of D. formicoaceticum have led us to postulate the pathway for dichloromethane utilization shown in Fig. 1 (19). In this scheme, dichloromethane and tetrahydrofolate are converted by one or more unknown enzymatic reactions to methylene tetrahydrofolate and inorganic chloride. Two-thirds of the methylene tetrahydrofolate formed is then oxidized to formate by the enzymes of the acetyl coenzyme A (acetyl-CoA) pathway. The reducing equivalents generated by this oxidation are used by methylene tetrahydrofolate reductase and CO dehydrogenase in the formation of acetate from methylene tetrahydrofolate and CO₂. Here we report on experiments that support the hypothesis that this pathway is used for the anaerobic metabolism of dichloromethane. They show that the carboxyl group of acetate originates from carbonate present in the medium and not from dichloromethane, and they demonstrate an activity in cell extracts that converts dichloromethane to methylene tetrahydrofolate.

View larger version (16K): [in this window] [in a new window]   FIG. 1.   Proposed pathway for the metabolism of dichloromethane by D. formicoaceticum.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Culture conditions. D. formicoaceticum DSM 10151 was grown in a 16-liter stainless-steel fermentor (Bioengineering, Wald, Switzerland) with the MOPS (morpholinepropanesulfonic acid)-buffered medium described earlier (19). Experiments on the sodium dependence of the organism were performed with a carbonate-buffered medium (19) in which NaHCO₃ was replaced by KHCO₃. This medium was reduced with titanium(III) nitrilotriacetic acid solution (100 mM TiCl₃ [23]) prepared with KOH and K₂CO₃ instead of NaOH and Na₂CO₂, respectively. Sodium concentrations in the cultures were analyzed with an AA-646 atomic absorbance spectrometer (Shimadzu, Kyoto, Japan).

Transformation of [¹³C]dichloromethane by cell suspensions. Cell suspensions were prepared as described earlier (19). Experiments were performed anaerobically in 12.4-ml serum flasks containing 3.3 ml of cell suspension [390 µg of protein/ml in 50 mM Tris (pH 7.5), 5 mM NaHCO₃, and 1.4 mM titanium(III) citrate (29))]. The reaction was started by the addition of 1.2 mM [¹³C]dichloromethane, and the cell suspension was incubated at 30°C with shaking. After the complete disappearance of dichloromethane as monitored by gas chromatography (GC) analysis, 1.8-ml samples were transferred anaerobically to a 5-mm nuclear magnetic resonance (NMR) tube, supplemented with 0.2 ml of D₂O, and analyzed on an AMX-500 spectrometer (Bruker, Karlsruhe, Germany) (5) at 125 MHz with broad-band proton decoupling. Spectral parameters were as follows: a 45° pulse angle, a 1.3-s acquisition time, and a 1.0-s relaxation delay. A total of 25,000 scans were acquired with 80,000 data points and 1-Hz line broadening.

Dichloromethane dehalogenation by cell extracts. For the preparation of cell extracts, cells were suspended in 40 mM Tris-SO₄, pH 7.0, containing 2 mg of lysozyme/ml, 0.05 mg of DNase I/ml, 5 mM dithiothreitol (DTT), and 1 mM phenylmethylsulfonyl fluoride. After incubation for 1 h at room temperature, the suspension was centrifuged (100,000 × g, 30 min), yielding an extract containing approximately 15 mg of protein/ml. Cell extracts (1.5 ml) were desalted with a PD-10 column G-25M (Sephadex; Pharmacia, Uppsala, Sweden). The experiments were performed with the first 0.9 ml of the eluate. Assays were performed at 30°C in 12.4-ml serum flasks containing 4.1 mg of tetrahydrofolate, 2.5 ml of buffer (100 mM Tris-SO₄ [pH 7.0], 5 mM DTT, and 1 mM MgSO₄), 0.3 ml of desalted cell extract, 30 µl of ATP (100 mM), and 15 µl of methyl viologen (10 mM) under an atmosphere of N₂ plus H₂ (95:5 [vol/vol]; 170 kPa) unless indicated otherwise. The pH of the incubation mixture was adjusted with 1 M KOH. After 15 min of preincubation, the reaction was started by the addition of dichloromethane. All manipulations were performed under strictly anaerobic conditions. Figures and tables show values for single experiments which were repeated twice and yielded very similar results each time.

ATPase activity. For the determination of ATPase activity, cells from the 0.9-ml cell suspension were harvested by centrifugation (10,000 × g, 10 min), resuspended aerobically in 0.75 ml of Tris-SO₄ (100 mM, pH 7.0; 1 mM DTT), and broken by sonication (eight cycles of 15 s each at 40 W with 45 s of cooling between cycles). The crude extract was centrifuged (10,000 × g, 25 min). The formation of ADP from ATP was monitored with a coupled enzymatic assay (12) in 50 mM potassium phosphate buffer (pH 7.0) with 0.5 mM MgSO₄ and 10 mM ATP. Activities were corrected for background NADH consumption.

Analytical methods. Dichloromethane was measured by GC (18), with a relative error of 9%, and its concentration in the liquid phase was calculated with the Henry's law constant (10). Methanol was quantified with the same instrumental setup but with an oven temperature of 110°C and a 3-µl liquid sample. Tetrahydrofolate, methylene tetrahydrofolate, and other derivatives of tetrahydrofolate were measured by high-pressure liquid chromatography (HPLC), with a relative error of 1% (20). The compounds were detected with a UV detector at 320 nm. To avoid autoxidation of tetrahydrofolate and its derivatives, care was taken that no air bubbles were flushed through samples. ATP, ADP, and AMP in crude extracts were measured by HPLC with a C₁₈ reversed-phase Nucleosil 100-7 column (250 by 4.6 mm) according to instructions provided by Stagroma (Wallisellen, Switzerland). Protein was measured according to the method of Bradford (1) with a dye reagent from Bio-Rad (Munich, Germany) and bovine gamma globulin as a standard.

Chemicals. All chemicals used were of reagent grade or better and were purchased from Fluka (Buchs, Switzerland). [¹³C]dichloromethane (99% ¹³C) was obtained from Cambridge Isotope Laboratories (Andover, Mass.). Tetrahydrofolic acid trihydrochloride (FH₄) was purchased from Dr. Schircks Laboratorium (Jona, Switzerland), and ATP was from Sigma (St. Louis, Mo.).

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Products formed in cell suspensions from [¹³C]dichloromethane. The pathway for anaerobic dichloromethane metabolism shown in Fig. 1 implies that formate and the methyl group of acetate are derived entirely from dichloromethane, whereas the carboxyl group of acetate originates from CO₂. To test whether this was the case, 1.2 mmol of [¹³C]dichloromethane was subjected to transformation by a cell suspension of D. formicoaceticum. When the compound had disappeared from the incubation mixture, the products formed from the labeled substrate were analyzed by ¹³C NMR. As shown by the spectrum obtained (Fig. 2), formate, the methyl group of acetate, and carbonate were labeled, but there was no signal corresponding to the carboxyl group of acetate, which would have shown a chemical shift of 181.7 ppm. This result is in accordance with the pathway shown in Fig. 1. A further signal which corresponded to that of methanol was also detected (Fig. 2), and the formation of methanol from dichloromethane was subsequently confirmed by GC analysis. In cell suspensions, about 30% of the dichloromethane transformed was converted to methanol, whereas in growing cultures this fraction amounted to 3%. This observation suggests that methanol is produced by a side reaction when dichloromethane metabolism is uncoupled from growth. The labeling of carbonate in the ¹³C NMR experiment may have resulted from an exchange reaction between carbon dioxide and formate catalyzed by formate dehydrogenase.

View larger version (16K): [in this window] [in a new window]   FIG. 2.   NMR analysis of products formed from [13C]dichloromethane (1.2 mM) by a cell suspension of D. formicoaceticum. The identities of formate and the methyl group of acetate were confirmed by analyzing a sample supplemented with an excess of unlabeled acetate and formate (containing 1.1% 13C). The signal marked "buffer" was also detected in a control experiment performed with unlabeled dichloromethane.

Formation of methylene tetrahydrofolate from dichloromethane by cell extract of D. formicoaceticum. The hypothetical pathway for dichloromethane metabolism (Fig. 1) also stipulates that methylene tetrahydrofolate be formed as the first detectable intermediate. Figure 3 shows that this is the case. When dichloromethane and tetrahydrofolate were incubated in the presence of ATP and desalted cell extract, the substrates were consumed concomitantly with the formation of methylene tetrahydrofolate. Minor amounts of methenyl tetrahydrofolate were also formed. With nondesalted cell extract, up to 30% of the dichloromethane degraded yielded methenyl tetrahydrofolate (results not shown). Methyl tetrahydrofolate and other derivatives of tetrahydrofolate were not detectable. D. formicoaceticum thus possesses an enzyme or enzymes which transfer the methylene group of dicholoromethane onto tetrahydrofolate. Although we have not demonstrated the release of inorganic chloride from dichloromethane, the corresponding activity is termed dichloromethane dehalogenase. Dehalogenase activity in cell extracts was estimated from the linear initial rate of dichloromethane consumption. Depending on the extract, the specific activity ranged between 6 and 16 nmol/min · mg of protein, which is equal to the degradation activity observed in cell suspensions and corresponds to about 20% of the degradation activity measured in growing cells (19).

View larger version (19K): [in this window] [in a new window]   FIG. 3.   Formation of methylene tetrahydrofolate from dichloromethane plus tetrahydrofolate (FH4) by cell extract of D. formicoaceticum. The assay mixture (see Materials and Methods) contained 0.72 mg of protein/ml, 0.8 mM CH2Cl2, 2.4 mM FH4, and 1 mM ATP. CH2Cl2 was measured by GC, and FH4 and its derivatives were measured by HPLC. Symbols: , CH2Cl2; , FH4; , methylene FH4; , methenyl FH4.

Requirement of ATP for the in vitro dehalogenation of dichloromethane. The data in Table 1 show that dichloromethane dehalogenase activity was dependent on the presence of ATP and hydrogen in the incubation mixture. No reaction was observed in the absence of these components. Omission of methyl viologen from the incubation mixture reduced the dechlorination rate by 60%. Since the carbon atom of dichloromethane retains its oxidation state upon transformation to methylene tetrahydrofolate, the requirement of ATP, methyl viologen, and molecular hydrogen for dichloromethane dehalogenation is not obvious.

Light-reversible inhibition of dehalogenation by propyl iodide. We have previously reported that propyl iodide, an inhibitor reacting with corrinoids in the superreduced Co(I) state, reversibly inhibits the metabolism of dichloromethane by D. formicoaceticum cell suspensions (19). Although this observation did not allow conclusions regarding the dehalogenation mechanism in the whole-cell system, it is compatible with a reduced corrinoid acting as a nucleophile in dichloromethane dehalogenation. The requirement of molecular hydrogen and of the electron carrier methyl viologen for the in vitro system (Table 1) is also compatible with a role for a reduced corrinoid in dichloromethane dehalogenation, since electrons from hydrogen may be needed to keep the catalyst in the reduced state. To test the possible involvement of a corrinoid in the transformation of dichloromethane to methylene tetrahydrofolate in vitro, the effect of propyl iodide on this process was examined. Figure 4 shows that 20 µM propyl iodide indeed inhibited the conversion of dichloromethane to methylene tetrahydrofolate and that this inhibition was relieved by exposure of the incubation mixture to light. The dehalogenation rate in the reactivated mixture was about 50% of that in a noninhibited control mixture (data not shown). This experiment thus provided support for the participation of a Co(I) corrinoid in dichloromethane dehalogenation, as did the strong sensitivity of the system to oxygen (data not shown).

View larger version (19K): [in this window] [in a new window]   FIG. 4.   Light-reversible inhibition of dichloromethane degradation by 20 µM propyl iodide. The test was performed in the absence of DTT. After preincubation of the crude extract (0.62 mg of protein/ml of assay mixture) with inhibitor for 30 min, dichloromethane was added. At the time indicated by arrows, one vial (open symbols) was exposed to the light of a slide projector lamp (250 W) for 2 min (25), while a control vial (closed symbols) was kept in the dark. Symbols:  and , CH2Cl2;  and , methylene FH4.

Detection of ATP synthase activity in D. formicoaceticum. While the data presented so far provide strong evidence for dichloromethane metabolism according to the scheme shown in Fig. 1, they do not explain the previously observed inability of D. formicoaceticum to grow with typical acetogenic substrates such as CO₂ plus H₂ (20:80 [vol/vol]) or formate (19). The limitation of the organism in this respect might be due to its inability to gain ATP by a chemiosmotic mechanism. At the substrate level, there is no net formation of ATP during acetogenesis from CO₂ plus H₂ or from formate. Growth on these substrates thus depends on energy conservation via a chemiosmotic mechanism (4). In contrast, utilization of dichloromethane results in the net formation at the substrate level of 1 mol of ATP per mol of dichloromethane metabolized, by acetate kinase and by the reverse formyl tetrahydrofolate synthase reaction (Fig. 1). To test whether these observations might explain the restriction of the substrate range of D. formicoaceticum to dichloromethane, cell extract of the organism was examined for ATPase activity. A specific activity of 30 nmol/min · mg of protein was detected. N,N'-Dicyclohexylcarbodiimide (0.1 mM) inhibited this activity by 36%, and 1 mM sodium azide inhibited it by 32%. This suggests the presence of an F₀F₁-type ATP synthase in the organism. Since D. formicoaceticum exhibited normal growth at sodium concentrations as low as 0.8 mM, energy conservation appears to be sodium independent.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

This communication describes experiments with cell suspensions and cell extracts to elucidate the metabolism of dichloromethane in D. formicoaceticum. It provides experimental evidence for two major features of a degradation pathway for dichloromethane which was previously proposed on the grounds of carbon and electron balances (19). First, the pattern of ¹³C-labeled products formed by cell suspensions from [¹³C]dichloromethane confirmed that dichloromethane is metabolized by the enzymes of the acetyl-CoA pathway to formate and the methyl group of acetate (Fig. 1). Second, transfer of the methylene group of dichloromethane onto tetrahydrofolate, the key reaction of anoxic dichloromethane utilization, was demonstrated to occur in vitro and subjected to preliminary characterization.

Most of the enzyme reactions leading from dichloromethane to formate and acetyl-CoA have been demonstrated in this communication and in a previous report (19). There was, however, no evidence for the occurrence of methylene tetrahydrofolate reductase activity in crude extracts of D. formicoaceticum when this enzyme was measured with reduced methyl viologen or NADH as the electron donor. We now have observed that undesalted crude extracts, incubated under reducing conditions, converted methylene tetrahydrofolate to methyl tetrahydrofolate, which was identified by HPLC analysis (19a). It thus appears that methylene tetrahydrofolate reductase activity is present, but that the enzyme is dependent on an as-yet-unidentified endogenous electron donor. This view is supported by the observation that growth of D. formicoaceticum does not depend on sodium. The organism thus would appear to fall into the group of acetogenic bacteria generating a proton diffusion potential for ATP synthesis during acetogenesis (24). In these organisms, which include Clostridium thermoaceticum and Clostridium thermoautotrophicum (13), cytochromes are thought to be involved in the delivery of electrons to methylene tetrahydrofolate reductase. The group of sodium-dependent acetogens includes Acetobacterium woodii (11), Peptostreptococcus productus (9), and strain MC (21). They might use an electrochemical sodium gradient for energy conservation. In the latter two organisms, the methylene tetrahydrofolate reductase has been shown to be dependent on NADH (21, 28).

The ability of a cell-free activity to degrade dichloromethane at nearly physiological rates was dependent on the presence of tetrahydrofolate. Slight consumption of dichloromethane in the absence of tetrahydrofolate was also measured. Since corrinoids are known to catalyze the nonenzymatic dehalogenation of dichloromethane (6, 16), this might have been caused by corrinoids present in the cell extract of D. formicoaceticum. The conversion of dichloromethane to methylene tetrahydrofolate also depended on the presence of ATP, methyl viologen, and hydrogen in the incubation mixture. Since crude extract of D. formicoaceticum contains strong methyl viologen-dependent hydrogenase activity (19), the latter two components provided reducing conditions. A requirement of ATP has also been reported for a number of methyl transfer reactions of acetogenic bacteria. Transfer of the methyl group of chloromethane onto tetrahydrofolate by cell extract of the homoacetogen strain MC depended on ATP, as did the O-demethylase of this organism (22), and similar observations have been reported for tetrahydrofolate-dependent O-demethylating activities in cell extracts of acetogenic bacteria (7, 8, 15). As with dichloromethane dehalogenation (Table 2), some of these systems require ATP in catalytic amounts (22, 26). This suggests that ATP, together with hydrogen and methylviologen, might play a role in the activation of a corrinoid protein involved in methylene or methyl transfer. Activation of a corrinoid-dependent methyltransferase by hydrogen and ATP has recently been described (2).

There are major differences between the dichloromethane dehalogenation activity described in this communication and the methyl transfer activities discussed above. The enzyme system of D. formicoaceticum must cleave two carbon-chloride bonds to yield a C₁ tetrahydrofolate product, whereas chloromethane dehalogenase of strain MC, a representative of the methyltransferases, delivers a C₁ tetrahydrofolate intermediate by cleavage of a single carbon-halogen bond. Dichloromethane is much less reactive to nucleophilic displacement than chloromethane. This raises the question of whether an enzyme can sufficiently activate tetrahydrofolate and dichloromethane for a direct reaction. A corrinoid in its superreduced state is a strong nucleophile and thus could act in an intermediary fashion, leading to a metabolism analogous to the corrinoid-dependent metabolism of methanol proposed for Sporomusa ovata (26) and other corrinoid-dependent methyltransferases (14, 15). In such a scheme, dichloromethane would react with a reduced Co(I) corrinoid to form a chloromethyl-Co(III) corrinoid plus chloride. The latter would then act as a donor for a methyltransferase, generating 5-chloromethyl tetrahydrofolate. In a third step, 5-chloromethyl tetrahydrofolate would nonenzymatically rearrange to yield 5,10-methylene tetrahydrofolate and inorganic chloride. Other than an indication, by propyl iodide inhibition, that a reduced Co(I) corrinoid may be involved in dichloromethane dehalogenation, there is no experimental evidence for such a mechanism, and the system must await detailed analysis.