Reconstitution of Monomethylamine:Coenzyme M Methyl Transfer with a Corrinoid Protein and Two Methyltransferases Purified fromMethanosarcina barkeri *

Methanogenesis from methylamines requires the intermediate methylation of 2-mercaptoethanesulfonate (CoM). In vitro reconstitution of CoM methylation with monomethylamine was achieved with three purified proteins: a monomethylamine corrinoid protein (MMCP), the “A” isozyme of methylcobamide:CoM methyltransferase (MT2-A), and a newly isolated protein termed monomethylamine methyltransferase (MMAMT).MMAMT is a 170-kDa protein with 52-kDa subunits. The MMAMT polypeptide was rate-limiting for methyl transfer until at a 2-fold molar excess over MMCP. MMAMT is a monomethylamine:MMCP methyltransferase, since methylation of MMCP required MMAMT but not MT2-A. MMCP and MMAMT formed a complex detectable by size exclusion high pressure liquid chromatography. Methyl group transfer from methyl-MMCP to CoM was mediated by MT2-A, since methyl iodide:CoM methyl transfer by MMCP and MT2-A did not require MMAMT. MT2-M, an isozyme of MT2-A, was inactive in MMCP-dependent methyl transfer. Immunodepletion of MMCP from the extract inhibited CoM methylation with monomethylamine but not dimethylamine. Purified MMCP reconstituted activity in immunodepleted extracts. These results show that MMCP is the major corrinoid protein for methanogenesis from monomethylamine detectable in extracts and that it interacts with two methyltransferases. MMAMT functions as a MMA:MMCP methyltransferase, while MT2-A functions as a methyl-MMCP:CoM methyltransferase.

and 7-mercaptoheptanoylthreonine phosphate (HTP) are then converted to methane and the heterodisulfide of CoM and HTP by methylreductase (4,5). The free thiols of CoM and HTP are regenerated with reducing equivalents gained from the concomitant oxidation of the methylotrophic substrate to CO 2 . The reduction of the heterodisulfide is a major site of energy conservation in methanogens (6).
Determination of the pathways of CoM methylation by methylotrophic substrates is therefore key to understanding methanogenesis from these substrates. The work of Vogels and co-workers (7,8) with Methanosarcina barkeri demonstrated that two enzymes were required to methylate CoM with methanol, designated methyltransferase I (MT1) and methyltransferase II (MT2). MT1 is a two-subunit enzyme that binds 5-hydroxybenzimidazolylcobamide, the major corrinoid found in this organism (9). The carbon of methanol undergoes nucleophilic attack by the central cobalt atom of the enzyme-bound corrinoid when in the Co(I) state. The methylated Co(III) corrinoid then serves as a substrate for MT2, which acts as a methyl-cobamide:CoM methyltransferase with the subsequent formation of methyl-CoM and Co(I) corrinoid (10). Three isozymes or homologues of MT2 have been found (11,12), but only the "M" isozyme, or MT2-M, of these interacts with MT1 in methanol metabolism (13). In addition to MT1 and MT2-M, several ancillary proteins can act in methanol:CoM methyl transfer. Oxidation of MT1 can leave the corrinoid in the inactive Co(II) form. A cellular activation system reactivates MT1 in a reductive process that also requires methanol and ATP (14 -16). Most studies have required relatively crude protein fractions containing hydrogenase, ferredoxin, and a methyltransferase activation protein for study of in vitro methanol: CoM methyl transfer, although recently the pathway was reconstituted with only purified MT1 and MT2-M (17).
Methylamines utilized by M. barkeri include trimethylamine (TMA), dimethylamine (DMA), and monomethylamine (MMA). Cells grown on TMA produce intermediate pools of DMA and MMA, which are subsequently consumed (18). Studies with cell suspensions and fractions indicate that the conversion of each methylamine to methyl-CoM involves distinct proteins (19 -21). However, a single MT2 isozyme, the "A" or amine isozyme (MT2-A), can be used for CoM methylation from all three methylamines (13). MT2-A was first identified by the general ability of MT2 isozymes to methylate CoM with free methylcobalamin (12). MT2-A shares 50% sequence similarity with MT2-M (10,22), and both are 37-kDa zinc-binding polypeptides (10). MT2-A is present in high amounts in cells grown on TMA (23). MT2-A was shown to stimulate MMA:CoM methyl transfer in extracts of M. barkeri MS (20). Separation of MT2-A from extracts with affinity-purified IgG specific for MT2-A inhibited MMA-, DMA-, and TMA-dependent CoM methylation (13). Activity was completely restored by the addition of highly purified MT2-A. MMA-or DMA-dependent methylation of CoM required MT2-A, while TMA:CoM methyl transfer utilized MT2-M or MT2-A.
The involvement of MT2-A indicated that corrinoid proteins exist in M. barkeri that could be methylated with TMA, DMA, or MMA and interact with MT2-A to effect the methylation of CoM. A corrinoid protein from M. barkeri Fusaro supported either TMA-or DMA-dependent CoM methylation in the presence of a 65% ammonium sulfate cut of cell extract containing proteins for reductive activation (21). In a more resolved system from M. barkeri MS, the ATP-dependent cellular reductive activation system was replaced with Ti(III)-citrate and methyl viologen (24). A protein termed TMA-MT, consisting of loosely associated 26-and 52-kDa polypeptides, was purified that mediated TMA:CoM methyl transfer when supplemented with homogeneous MT2-A or MT2-M. MT2-A was the preferred isozyme. The 26-kDa polypeptide bound corrinoid and was designated as the trimethylamine corrinoid protein (TCP).
In contrast, the pathways of DMA-and MMA-dependent CoM methylation have yet to be reconstituted with purified proteins. A corrinoid protein required for DMA:CoM methyl transfer has been partially purified (21). A monomeric 29-kDa corrinoid protein (designated the MMA corrinoid protein, or MMCP) is involved in MMA-dependent CoM methylation (20). This corrinoid protein was first identified as one of several proteins methylated following inhibition of methylreductase (25,26). Although MMCP and MT2-A are involved in MMA: CoM methyl transfer, mixtures of the two proteins will not catalyze the reaction. In this study we purified from cell extracts a protein designated MMA methyltransferase (MMAMT), which was essential for methyl transfer from MMA to CoM. MMAMT, MMCP, and MT2-A were the only proteins required for in vitro reconstitution of CoM methylation from MMA.

EXPERIMENTAL PROCEDURES
Cell Cultures and Extract Preparation-M. barkeri MS (DSM 800) was cultured under anaerobic conditions in a phosphate-buffered medium supplemented to 80 mM with TMA, MMA, or methanol as described previously (20,27). Cell extracts were prepared and stored at Ϫ70°C in hydrogen-filled serum vials until use (25).
Enzyme Assays-All assays and protein purifications were carried out under anaerobic conditions unless indicated otherwise. Gases were purchased from Linde Specialty Gases (Columbus, OH), and trace amounts of oxygen removed by passage through an R3-11 catalyst (Chemical Dynamics Corp., South Plainfield, NJ) before use. Routine assay of the MMA:CoM methyl transfer activity with purified components or during isolation of the MMAMT was carried out in N 2 -flushed 1.5-ml glass vials sealed with rubber stoppers. A typical reaction mixture contained the following components added to vials on ice in a total volume of 100 l: 50 mM MOPS-HCl buffer, pH 7.0, 1.5 mM Ti(III)citrate adjusted to pH 8.0 with a saturated Tris solution (28), 0.5 mM methyl viologen, 50 mM MMA, 2 mM CoM, 10 g of purified MMCP, 10 g of purified MT2-A, and a 50-l column fraction or specified amount of purified MMAMT. The reaction was started by placing the vials in a 37°C water bath. Samples (10 l) were removed and analyzed for the free thiol of CoM using 5,5Ј-dithiobis(2-nitrobenzoic acid) (16,29). Previously, this measure has been shown to reflect the CoM-dependent disappearance of methylamine in whole and fractionated extracts and is therefore an accurate measure of CoM methylation by MMA (20).
Methylation of MMCP with MMA was carried out using the same conditions described for the assay of MMAMT except that MT2-A and CoM were omitted from the reaction mixture, and 200 g of MMAMT and 120 g of MMCP were used. The reaction was incubated at 37°C for 30 min.
DMA:CoM or MMA:CoM methyl transfer activities of antibodytreated extracts were assayed by measurement of substrate-dependent disappearance of the free thiol of CoM as described previously (20).
Purification of Corrinoid Proteins and Methyltransferases-Purification of MMAMT was performed in an anaerobic chamber (Coy Laboratories) with buffers and column materials made anaerobic by repeated cycles of evacuation and flushing with N 2 . Protein purification was initiated by applying 25 ml (13.6 mg/ml) of an extract of cells grown on MMA to a 10 ϫ 2.5-cm chromatography column containing DE-52 (Whatman Inc., Fairfield, NJ) that had been equilibrated with 50 mM NaCl in 50 mM MOPS-HCl buffer, pH 7.8. Proteins bound to the matrix were eluted with a 250-ml linear gradient of 50 -600 mM NaCl in the same buffer at a flow rate of 2 ml/min. Fractions that supported MMA: CoM methyl transfer activity when supplemented with MMCP and MT2-A eluted between 180 and 294 mM NaCl. The pool of fractions was diluted 10-fold in 50 mM MES-NaOH, pH 6.0, and applied to a Mono-Q HR 10/10 (Pharmacia Biotech Inc.) column equilibrated with 50 mM NaCl in 50 mM MES-NaOH, pH 6.0. Protein was eluted with an 80-ml gradient of 50 -500 mM NaCl at a flow rate of 3 ml/min. Active fractions eluted between 180 and 200 mM NaCl. The fractions were pooled (4 ml) and then concentrated to 300 l using Centricon-10 microconcentrators (Amicon Inc., Beverly, MA) and applied to a Superdex-200 gel filtration column (Pharmacia) equilibrated with 50 mM MOPS-HCl, pH 6.5, 100 mM NaCl. The peak of activity eluted at 58 ml. The active fractions were combined and applied to a Mono-Q HR 10/10 column equilibrated with 50 mM MOPS-HCl, pH 6.5, 100 mM NaCl. Fractions eluting between 205 and 221 mM NaCl of an 80-ml linear gradient of 100 -500 mM NaCl were active in the MMA:CoM methyl transfer assay. MMCP in the methylated or unmethylated state was isolated from extracts of M. barkeri as described previously (20,26). MT2-A and MT2-M were aerobically purified from TMA-or methanol-grown cells, respectively, essentially as outlined by Yeliseev et al. (23). The two isozymes could be distinguished by their elution profiles on hydroxylapatite (20,23). The specific activities at pH 7.0 of MT2-A and MT2-M were 34 and 6 mol of methylcobalamin consumed/min/mg, respectively, as measured by the cyanide derivatization assay described by Grahame (12). A homogenous preparation of TCP was the kind gift of D. J. Ferguson and was isolated by a modification of his published procedure (24). MT2-M, MT2-A, and MMCP were homogenous preparations, and SDS-PAGE gels documenting their level of purity can be found in this paper and a previous paper (13).
Immunodepletion of MMCP from Extracts-Antiserum against MMCP was raised in rabbits inoculated with purified MMCP using standard techniques (30). IgG against MMCP was then further purified by affinity chromatography. The IgG fraction from 75 ml of antiserum was precipitated by the addition of ammonium sulfate to 50% of saturation. The precipitate was collected by centrifugation, and the pellet was resuspended in 150 mM NaCl in 10 mM potassium phosphate buffer, pH 7.0. The sample was dialyzed against phosphate-buffered saline overnight and passed through a 0.2-m sterile filter. Purified MMCP (1.7 mg) was coupled to a 1-ml column of N-hydroxysuccinamide-activated Sepharose High Performance matrix (Pharmacia) according to the procedure recommended by the manufacturer. The IgG fraction (500 mg) was then applied in five aliquots to the affinity column following the procedure outlined by the manufacturer. Adsorption and subsequent elution of IgG was monitored by the absorbance at 280 nm of the column effluent. Unbound IgG was reapplied to the column in three or four subsequent runs until the effluent showed no ␣-MMCP activity as judged by an enzyme-linked immunosorbent assay (30). Unbound IgG that eluted from the column was pooled and used as nonspecific IgG in control experiments as described below. Approximately 11 mg of ␣-MMCP IgG were recovered from 500 mg of the crude IgG fraction. MMCP was removed from extract using these antibodies in the following manner. An extract (3 mg of protein) of cells grown on TMA were added to 1.2 mg of affinity-purified ␣-MMCP IgG, and the mixture was incubated on ice for 3 h. A suspension (0.2 ml) of protein A-Sepharose (Sigma) containing approximately 1.2 mg of cross-linked protein A was added, and the mixture was incubated for 2 h at room temperature with gentle mixing. The protein A-Sepharose beads were removed by centrifugation at 10,000 ϫ g for 3 min, and the supernatant was concentrated by ultrafiltration using a Centricon 10 microconcentrator. Control samples were also prepared using 1.2 mg of nonspecific IgG or an equivalent volume of 50 mM MOPS-HCl buffer in place of ␣-MMCP IgG. Both control extracts were subjected to the same incubations, treatment with protein A-Sepharose, and concentration steps so that direct comparisons could be made.
Analytical Procedures-SDS-PAGE was carried out according to the method of Laemmli (31) using 12% acrylamide and a Mini-Slab electro- ␣-MMCP activity in antibody fractions was determined by enzymelinked immunosorbent assay as described by Harlow and Lane (30). Horseradish peroxidase-conjugated sheep ␣-rabbit IgG (Sigma) was used as secondary antibody and was detected by reaction with 2,2Јazino-bis-(3-ethylbenzthiazoline 6-sulfonic acid). Absorbance at 410 nm was measured using an MR700 enzyme-linked immunosorbent assay plate reader (Dynatech Laboratories, Inc.).
Size estimation of MMAMT and the complex formed with MMCP in solution was performed using size exclusion HPLC. G4000SW XL and G3000SW XL (Supelco, Bellefonte, PA) columns in series were equilibrated with anaerobic 50 mM potassium phosphate buffer, pH 7, 100 mM NaCl. The individual proteins to be tested were incubated at 22°C individually and in mixtures in 50 mM MOPS-HCl, pH 7.0, 100 mM NaCl under a nitrogen atmosphere for 1 h. Samples were injected onto the column being eluted at 1.0 ml/min with the equilibration buffer. Fractions (0.5 ml) were collected for further analysis by SDS-PAGE. The corrinoid cofactor bound to the MMCP used in these experiments was in a mixture of the Co(II) and hydroxylated forms. Molecular mass standards (Sigma) were thyroglobulin (669 kDa), apoferitin (450 kDa), ␤-amylase (200 kDa), alcohol dehydrogenase (150 kDa), bovine serum albumin (66 kDa), and carbonic anhydrase (29 kDa).
Corrinoid cofactor from MMCP was extracted with ethanol with 50 -60% yield (25). Reversed phase HPLC of corrinoid cofactor was performed essentially as described previously (25) with a Microsorb-MV C-18 column (4.6 mm ϫ by 25 cm) (Rainin, Woburn, MA) equilibrated with 20% methanol in 25 mM sodium acetate, pH 6.0. Samples were eluted with a linear gradient of 20 -100% methanol over 60 min. The retention time of methylcobalamin in this system was 22.5 min, while that of hydroxycobalamin was 12.6 min. As noted previously, methylated and hydroxylated corrinoid isolated from MMCP elute with very similar retention times to the analogous cobalamin derivatives (25).
UV-visible spectroscopy was performed using a Hewlett-Packard 8453 photo diode array spectrophotometer. Protein was assayed using bicinchoninic acid (33) with bovine serum albumin as standard.

Isolation of MMAMT, a Protein Essential for MMA:CoM
Methyl Transfer-Previous experiments indicated MMCP and MT2-A required an unidentified protein in cell extracts to effect CoM methylation with MMA (20). An assay for this putative protein was devised that consisted of purified MT2-A, MMCP, Ti(III)-citrate, and methyl viologen. Fractions from anaerobic column chromatographic separations of extract prepared with MMA-grown cells were added to this assay mixture, and MMAdependent CoM methylation was monitored.
A single peak of MMA:CoM activity was detected upon initial fractionation of an extract by anion exchange chromatography. This peak was designated as the MMA methyltransferase activity or MMAMT. Methylation of CoM with MMA was absolutely dependent on the pooled fraction, which was further purified ( Table I). The active peak from the final Mono-Q column was analyzed by SDS-PAGE and was comprised of a single 52-kDa polypeptide (Fig. 1, lane 4). This band corresponded to a major protein band that was observed in the extract of MMA-grown cells (Fig. 1, lane 3) but was much less intense in the extract of cells grown on methanol (Fig. 1, lane  2). A polypeptide that comigrated with MMCP in SDS gels was one of the last contaminants to be removed during the purification of the MMAMT polypeptide (not shown).
Purified MMAMT from the Mono-Q fraction was used in the remaining experiments described here. The isolated protein was colorless; its UV-visible spectrum taken under aerobic conditions possessed only a single peak at 280 nm, indicating that this protein bound no detectable corrinoid.
Purified MMAMT, MMCP, and MT2-A Are Sufficient for MMA:CoM Methyl Transfer-MMA-dependent methylation of CoM by MT2-A and MMCP (see Fig. 1, lanes 5 and 6 for the purity of these preparations) required the presence of the purified MMAMT. Increasing amounts of the MMAMT polypeptide were added to vials containing 0.27 nmol of MT2-A and 0.35 nmol of MMCP (Fig. 2). This resulted in a linear increase in the MMA:CoM methyl transfer reaction rate until the approximate molar ratio of the MMAMT polypeptide to MMCP was 2, after which saturation of the rate of methyl transfer was observed. The specific activity at this ratio of MMCP, MT2-A, and MMAMT was 0.6 mol/min/mg of total protein.
With this ratio of proteins the requirements for MMA:CoM methyl transfer were further investigated. MMCP and MT2-A were absolutely required for the MMA-dependent methylation of CoM, and no reaction was observed when these proteins were omitted. MMA was the only methylamine to function as a significant methyl donor in the reaction. DMA or TMA (50 mM) did not serve as substrates when substituted for MMA. No reaction was observed when Ti(III)-citrate was omitted. Reaction rates were less than 2% of the maximally observed rate when methyl viologen was not added. ATP (10 mM) and MgCl 2 (20 mM) were not required and did not stimulate the reaction rate  MMA was not methylated as judged by reverse phase HPLC. However, the addition of MMAMT resulted in the methylation of the corrinoid, since the cofactor now eluted with the retention time of the methylated form. No hydroxylated corrinoid was detected following incubation with MMAMT and MMA. Methylation of MMCP with MMA in the presence of the MMAMT polypeptide was confirmed by UV-visible spectroscopy (Fig. 3). The corrinoid from MMCP was extracted from samples of Ti(III)-citrate/methyl viologen-reduced MMCP incubated with MMAMT and MMA. The spectrum had the characteristic absorbance peaks of methylated corrinoid (34). The spectrum was not affected by the addition of cyanide (not shown), as expected for methylated corrinoid. In the absence of MMA, the spectrum of the extracted aerobic corrinoid was that of hydroxylated corrinoid. This form was sensitive to cyanolysis.  the approximate retention time of monomeric MMCP. Analysis of this peak by SDS-PAGE indicated that it was primarily composed of MMCP (Fig. 4B, lanes 5 and 6). However, the first peak eluted with a retention time of 16.5 min, faster than observed for MMAMT chromatographed by itself. This corresponded to an apparent molecular mass of 238 kDa. SDS-PAGE of fractions collected from this peak indicated the presence of MMAMT and MMCP, indicating that this peak was comprised of a complex of the two proteins (Fig. 4B, lanes 2-4). Densitometry of the Coomassie-stained gel indicated the ratio of the MMAMT polypeptide to MMCP in the peak was 3.5:1.
A stable association of MT2-A with either MMCP or MMAMT could not be demonstrated. MT2-A had no apparent effect on complex formation between MMAMT and MMCP or on their individual elution times. MT2-A continued to elute as a monomeric protein even in the presence of the other two proteins.
Methyl Transfer from MMCP to CoM Requires Only MT2-A-MT2-A can demethylate free methylcobalamin and methylate CoM. Presumably, the role of MT2-A in MMA:CoM methyl transfer is the demethylation of MMCP following MMCP methylation by MMAMT. This was directly tested by monitoring the suitability of methyl-MMCP as a substrate for CoM methylation by MT2-A in the absence of MMAMT. Methyl iodide serves as a nonphysiological methyl donor for many corrinoid proteins and will readily undergo nucleophilic attack by Co(I) corrinoid (35)(36)(37). Neither MT2-A nor MMCP could carry out methyl iodide:CoM methyl transfer when tested alone; however, in combination the two proteins mediated the reaction at a rate of 875 nmol/min/mg of total protein (Fig. 5). In order for MMCP to enter the Co(I)/methyl-cobalt catalytic cycle, it was added to the reaction mixture in the methylated form. MMCP in the Co(II) form was inactive in methyl transfer. This is consistent with the methylation of Co(I) corrinoid by methyl iodide, followed by catalysis of methyl transfer from methyl-MMCP to CoM by MT2-A.
MT2-M does not participate in MMA metabolism (20). Methyl iodide:CoM methyl transfer via MMCP was also specific for MT2-A; no detectable reaction occurred with MT2-M (Fig. 5) However, the final reconstituted system with these three purified polypeptides mediated the reaction at a rate of 0.6 mol/min/mg. It is possible that the MMCP and its associated methyltransferases were a minor pathway of CoM methylation from MMA. To determine if MMCP and its associated methyltransferases are the major route of MMA dependent CoM methylation, MMCP was specifically removed from extracts. Affinity-purified ␣-MMCP IgG was incubated with an extract of TMA-grown cells, then removed with Sepharose 6B coupled to protein A. Removal of MMCP protein was nearly complete, as indicated by Western analysis (Fig. 6B, lane 3). Extract treated with the nonimmune IgG fraction retained MMCP at a level similar to that of untreated extracts, as judged by densitometry of the corresponding signals on the Western blot (Fig. 6B, lanes 1 and  2). The MMCP-depleted extract had 9% of MMA:CoM methyl transfer activity remaining relative to extract treated with the nonimmune IgG fraction (Fig. 7). Activity could be completely restored to original levels by the addition of purified MMCP to the depleted extract.
The corrinoid-binding polypeptide involved in DMA:CoM methyl transfer has yet to be clearly identified. Since MT2-A was demonstrated by immunodepletion of extracts to be involved in both DMA-and MMA-dependent CoM methylation, it remained a possibility that the same corrinoid protein was also involved in both catabolic reactions. The extracts immunodepleted of MMCP above were also tested for DMA:CoM methyl transfer activity (Fig. 7). DMA-dependent CoM methylation was virtually unaffected by removal of MMCP, and the levels of activity were similar in extracts treated with either immune or non-immune IgG. This indicates that the role of MMCP in methylamine metabolism is specifically MMA utilization and that the corrinoid-binding polypeptide involved in DMA metabolism is distinct from MMCP. DISCUSSION This is the first reconstitution of MMA:CoM methyl transfer using only purified proteins. The involvement of MMCP and MT2-A had been previously demonstrated, but these two proteins were insufficient to reconstitute MMA:CoM methyl transfer in vitro (20). In this work, we were able to identify and purify to homogeneity from extracts of cells grown on MMA an abundant 52-kDa polypeptide, MMAMT, which In many of the studies of methyltransferases involved in methylotrophic methanogenesis, heterogeneous protein fractions have been used to supply a methyltransferase activation protein and reducing power to effect the ATP-dependent reductive activation of corrinoid proteins (14,16). We were able to reconstitute methyl transfer with only three proteins, and activity was not dependent on ATP. This was possible since Ti(III)-citrate and methyl viologen were used for reductive activation of MMCP. Ti(III)-citrate is capable of effecting the reduction of methyl viologen to the neutral species (24), which has a midpoint potential of Ϫ730 mV (38). This is sufficient to effect reduction to the Co(I) state in the corrinoid-dependent methyltransferases whose midpoint potentials have been measured, which range from Ϫ426 to Ϫ526 mV (39 -43). Methyl viologen greatly stimulates the MMA:CoM methyl transfer reaction and may circumvent a kinetic barrier in the direct reduction of MMCP to Co(I) with Ti(III)-citrate.
The specific activity that we have observed for the MMA: CoM methyl transfer reaction is lower than expected from the rate found with unfractionated extract. This accounts for the low apparent recoveries of MMAMT during purification. Similarly low specific activities have been observed with the purified proteins and protein fractions catalyzing CoM methylation with either TMA (21,24) or in most cases with methanol (7,15,21,24). Several explanations for the low specific activity with the MMA methyltransferases (and the resultant low apparent recovery of total activity from the extract) are possible. Reductive activation of MMCP to the Co(I) state with Ti(III)-citrate and methyl viologen could be incomplete, leaving only a fraction of MMCP in the active state. Alternately, a protein stimulating the reaction could be lost during purification. Our results indicate that such a protein would not be involved in direct methyl transfer from MMA to CoM but, for example, could promote the reductive activation of MMCP. In addition, the components of the in vitro reaction have not yet been optimized. However, despite the low specific activity of the reconstituted system, MMCP is the major route of methanogenesis from MMA detectable in cell extracts. Removal of MMCP from extracts with affinity-purified IgG significantly inhibited MMA-dependent methylation of CoM. Supplementation of the MMCP-depleted fraction with highly purified MMCP resulted in reaction rates even higher than those observed before removal of MMCP.
The current data allow the roles of MMAMT and MT2-A in the MMA:CoM methyl transfer reaction to be further defined. The role of MMAMT in MMA:CoM methyl transfer is the methylation with MMA of the corrinoid cofactor bound to MMCP, i.e. the protein acts as an MMA:MMCP methyltransferase. Methylation of MMCP did not require the presence of MT2-A. The MMAMT polypeptide could facilitate MMCP methylation with MMA in several ways. MMAMT binds MMCP, and it could further bind MMA and activate the methylamine so that nucleophilic attack by Co(I) of MMCP is facilitated. The interaction of MMAMT and MMCP could also result in kinetic and/or thermodynamic effects that favor the Co(I) state of the MMCPbound corrinoid. In any case, the involvement of MMAMT in methanogenesis from monomethylamine is clear. We recently found that MMAMT and MMCP lie adjacent on the chromosome and are co-transcribed. 2 The complex formed between MMCP and MMAMT is consistent with MMAMT function as an MMA:MMCP methyltransferase. However, it must be stressed that the stoichiometry observed in this complex (3-4 MMAMT monomers/MMCP monomer) may only represent the most stable complex detectable by the means employed. Complexes with higher amounts of MMCP may form that dissociate during size exclusion HPLC. Supporting this possibility is the observation that MMAMT-to-MMCP monomer ratios of 2 yielded maximal rates in the reconstituted MMA:CoM methyl transfer reaction.
Once methylated by MMA, MMCP serves as methyl donor to CoM. The role of MT2-A in the MMA:CoM methyl transfer reaction is that of a methyl-MMCP:CoM methyltransferase. This reaction can be carried out in the absence of MMAMT, as demonstrated by the methyl iodide:CoM methyl transfer reaction mediated by MMCP and MT2-A. That MT2-A serves to methylate CoM with methyl-MMCP was not unexpected, since MT2 isozymes such as MT2-A and MT2-M catalyze rapid methylation of CoM with non-protein-bound methylcobalamin (7,12), but direct methyl transfer from MMCP to CoM had not been previously demonstrated. The specificity of MMCP for MT2-A in either the methyl iodide-or MMA-dependent reactions is striking. Another MT2 isozyme, MT2-M, did not act as a methyl-MMCP:CoM methyltransferase, indicating that specific interaction of determinants on MMCP and MT2-A are required for the enzymatic transfer of the methyl group from the corrinoid to CoM. It was interesting in this regard to observe that MT2-M could inhibit methyl transfer from MMCP to CoM by MT2-A. The simplest explanation for inhibition is that MT2-M can also interact with MMCP but does not allow the proper orientation of CoM and methyl-corrinoid needed for methyl transfer. The inhibition by MT2-M of MMA:CoM methyl transfer would not present a problem in vivo, since MT2-A is by far the predominant MT2 isozyme found in cells grown on methylamines (23).
A direct comparison of the methyltransferases involved in CoM methylation from both MMA and TMA is now possible (Fig. 8). Like MMA:CoM methyl transfer, the TMA:CoM methyl transfer reaction requires three polypeptides (24). Aside from MT2-A, the TMA-dependent reaction requires a corrinoid-binding protein (TCP), and a 52-kDa polypeptide (TMA-52). However, both sets of methylamine methyltransferases are very specific for their substrates.  the same protein, MT2-A, the specificity for the methylamine substrate must lie in the polypeptides that are distinct between the two pathways. The corrinoid-binding polypeptides, TCP and MMCP, are distinct but possibly homologous proteins. The N-terminal sequences are 35% identical over 34 amino acids. 2 They are also functionally different. TCP cannot substitute for MMCP in the MMA:CoM methyltransferase reaction, indicating the inability of TCP to act as a methyl acceptor with MMAMT, the MMA:MMCP methyltransferase. The N termini of the largest polypeptides involved in the MMA and TMA reactions (MMAMT and TMA-52) have also been obtained; they share almost no homology over 20 amino acids. 3 It thus appears that both elucidated paths of CoM methylation from the methylamines can use the same MT2 isozyme but that the difference in substrates has dictated divergence in the corrinoid-binding polypeptides and the largest polypeptides of the TMA and MMA pathways.
The requirement for MT2-A in DMA:CoM methyl transfer (13) indicates that a separate corrinoid-binding polypeptide also exists in this pathway. This corrinoid polypeptide cannot be TCP, since TCP interacts with either MT2-A or MT2-M in this strain of M. barkeri (13,24). A corrinoid-containing protein fraction that supports only DMA:CoM, but not TMA:CoM, methyl transfer activity was recently reported (21). This protein was not characterized or further purified, but the current data show that MMCP cannot be a part of this reaction. Removal of MMCP from extracts did not affect DMA:CoM methyl transfer.
The TMA-52 polypeptide and TCP can be isolated in a loosely associated complex (24). MMAMT and MMCP are easily isolated as separate proteins; however, they also form a complex. This invites direct comparison of the TMA and MMA pathways of CoM methylation to MT1, which, with MT2-M, catalyzes methanol:CoM methyl transfer. MT1 is composed of 53-and 34-kDa subunits that do not readily dissociate. The genes encoding these subunits were recently sequenced by Sauer and Thauer and respectively designated mtaB and mtaC (17). The smaller subunit has the corrinoid binding motif found in cobalamin-dependent methionine synthase (44) and the corrinoid-binding subunit of a recently discovered CoM methylase (11,35). The sequences of MtaB and MtaC do not contain the N-terminal sequences we have obtained for TCP, TMA-52, MMAMT, and MMCP. However, homology of the N termini of TCP and MMCP to MtaC is detectable. 3 The similarities among methylotrophic paths to the methylation of CoM are striking and indicate that a general model of methylotrophic methanogenesis can be proposed. Central to each pathway is a distinct corrinoid protein. The corrinoid protein appears to determine specificity for the MT2 isozyme, which methylates CoM with the methylated corrinoid protein.
Aside from the corrinoid protein and the MT2 isozyme, a third polypeptide is found in each methyl transfer system. In the MMA system, this larger polypeptide is MMAMT, which acts to methylate the corrinoid protein with the methylotrophic substrate. This may also be the role of the largest polypeptides of the methanol:CoM and the TMA:CoM methyl transfer proteins.
FIG. 8. Comparison of known enzymes for the methylation of CoM with methylotrophic substrates. This scheme does not take into account higher ordered structures formed by the components of a methyl transfer pathway; rather, it is meant to indicate roles that may be played by analogous polypeptides. Each pathway possesses a distinct corrinoid-binding polypeptide that is methylated with a methylotrophic substrate by the largest polypeptide and demethylated by an MT2 isozyme. The implied involvement of the largest polypeptides of the methanol and TMA pathways in corrinoid methylation by the substrate is based upon our current results with the MMA pathway. The shaded polypeptides of the DMA:CoM methyl transfer pathway have not been isolated but are predicted from the composition of the other three pathways and the demonstrated involvement of MT2-A in DMA metabolism. Further details are found in the discussion.