Controlled Potential Enzymology of Methyl Transfer Reactions Involved in Acetyl-CoA Synthesis by CO Dehydrogenase and the Corrinoid/Iron-Sulfur Protein from CZostridium thermoaceticum*

fix COz via the Wood path- way of acetyl-CoA synthesis. dehydrogenase the corrinoid/iron-sulfur the synthesis of enzyme-bound have and methylation of CODH at controlled potentials. The rates of all these reactions except the exchange between methylated C/Fe-SP and methylated CODH are accelerated (from 1 to 2 orders of magnitude) when run at low potentials. Our results provide strong evidence for a nucleophilic redox-active metal center on CODH as the initial ac- ceptor of the methyl group from the methylated C/Fe- SP. This metal center also is proposed to be involved in the cleavage of acetyl-CoA in the reverse reaction.

We have performed these two exchange reactions, methylation of the C/Fe-SP, and methylation of CODH at controlled potentials.
The rates of all these reactions except the exchange between methylated C/Fe-SP and methylated CODH are accelerated (from 1 to 2 orders of magnitude) when run at low potentials. Our results provide strong evidence for a nucleophilic redox-active metal center on CODH as the initial acceptor of the methyl group from the methylated C/Fe-SP. This metal center also is proposed to be involved in the cleavage of acetyl-CoA in the reverse reaction.
Clostridium thermoaceticum and many other anaerobic bacteria perform CO, fixation via the acetyl-CoA pathway (Wood pathway) which involves a corrinoid/iron-sulfur protein (C/Fe-SP)' and carbon monoxide dehydrogenase (CODH) as central enzymes (see Refs. l-4 for recent reviews). This autotrophic pathway can be summarized by Equation  In this pathway, CO, is reduced to methyltetrahydrofolate (CH3-H4folate) via formate dehydrogenase and several H,folate enzymes. The individual steps which are involved in the synthesis of acetyl-CoA from CHa-Hdfolate involve the participation of enzyme-bound intermediates and are catalyzed by three enzymes: a methyltransferase, the C/Fe-SP, and CODH. The five genes encoding these three enzymes from C. thermoaceticum have been cloned and found to be part of an -11-kilobase gene cluster (5).
The methyl group of CHB-Hlfolate is transferred to the C/Fe-SP by methyltransferase forming enzyme-bound CH,cob(III)amide. Methylation of the C/Fe-SP involves reduction of both the cobalt and [4Fe-4S] centers from the 2+ to the l+ state (6), the 2+/l+ couples having midpoint reduction potentials at -504 and -523 mV, respectively (7). The exact mechanism of the cleavage of the CH,-N bond of CHs-Hlfolate has not been established; however, it appears that the reaction occurs by a nucleophilic attack of Co' on the methyl group (6,7). This reaction is analogous to the reaction mechanism of methionine synthase (8) which transfers the methyl group of CH3-H,,folate to homocysteine, forming methionine, via an enzyme-bound methyl-cob(III)alamin intermediate (8)(9)(10).
The final steps in the synthesis of acetyl-CoA occur on CODH. Based on studies of the exchange reaction between CO and the carbonyl of acetyl-CoA, it was proposed that CODH should be called acetyl-CoA synthase since it has the ability to bind the methyl group, the carbonyl group, and CoA at separate sites and then to condense these groups to form acetyl-CoA (11). When CODH binds CO or COZ (after reduction), an organometallic intermediate consisting of nickel, iron, and CO (4,12,13) is formed. This Ni-Fe-C intermediate has been studied by a number of spectroscopic methods, including EPR (12,13),2 M6ssbauer,3 electron nuclear double resonance,4 and x-ray absorption spectroscopies (17,18). Based on Mossbauer spectra, the magnetic properties of the iron associated with the Ni-Fe-C center resemble those of [4Fe-4S] clusters in the 2+ core oxidation state? The nickel site consists partly (17) or primarily (18) of Ni-S bonds and evidence for a Ni-Fe distance consistent with a Ni-X-Fe bridge has been presented (18). An analogous Ni-Fe-C intermediate apparently is involved in cleavage of acetyl-CoA by the CODH from methanogenic bacteria (19) and by the acetogenic CODH.5 CODH binds CoA at a site involving arginine and tryptophan residues (20,21) near the Ni-Fe-C site (12). Wood and coworkers6 have recently sequenced the residues surrounding the CoA binding site and found some homology to ATPbinding enzymes. Recently, Ramer et al. (22) characterized an exchange reaction between [3'-3ZP]coenzyme A and the CoA moiety of acetyl-CoA, earlier studied by Pezacka and Wood (23). This reaction occurs 6.3-fold faster than the exchange between CO and the carbonyl of acetyl-CoA (22).
CODH accepts the methyl group of the methylated C/Fe-SP, forming a methylated CODH intermediate. We favor a mechanism of methyl transfer to CODH involving a heterolytic cleavage via nucleophilic attack by a group on CODH on the methyl group of the methylcobalt (7). The chemistry involved in the methylation of CODH is the major focus of the work presented in this paper. Pezacka and Wood (24) were successful in methylating CODH with methyl iodide, methylated C/Fe-SP, and methyl-Blz as methyl donors. An enzyme-bound CH,-CODH intermediate was proposed to be S-methylcysteine based on amino acid analysis of 14CH3labeled CODH, was located on the /3 (71 kDa) subunit, and was converted to acetyl-CoA in the presence of CO and CoA (24). In the work presented in this paper, we have further characterized the reactions leading to methylation of CODH and also the subsequent demethylation reaction to form an acetyl-CODH intermediate, followed by the synthesis of acetyl-CoA. We propose that a nucleophilic metal center on CODH is the active site which accepts the methyl group from the methylated C/Fe-SP.
The synthesis of acetyl-CoA using methyl iodide as methyl donor is performed at 22 "C due to the volatility of methyl iodide. Even at this temperature, we required a correction for the concentration of methyl iodide in solution. Acetyl-CoA was synthesized by CODH and the C/Fe-SP in the presence of methyl iodide, CO, and CoA, as was reported earlier (20,24). The synthesis was linear with time for 30 min to 1 h. Quantitation of the amount of acetyl-CoA formed in the reaction mixture by the coupled enzyme assay and by the HPLC method gave approximately equivalent results. There was no detectable acetyl-CoA formed in the absence of CODH or the C/Fe-SP and methyl iodide, CO, and CoA were all essential. DTT or other reducing agents were not required in the synthesis. Acetate was formed as an additional product accounting for -20% of the total products and approximately the same amount of acetate was formed both in the absence or presence of CoA. Acetyl-CoA synthesis from methyl iodide was stimulated by ferredoxin with a maximum of -40% stimulation occurring with 0.8 pg of ferredoxin (1 nmol of ferredoxin/l5 nmol of CODH). Ferredoxin also stimulated the synthesis from CO, CoA, and CHs-Hdfolate (11). The K,,, value for CoA was determined from a Lineweaver-Burk plot to be 4.3 mM, which is consistent with the value for CoA (4.7 mM) reported before (34) with CH,-H,folate as the methyl donor, suggesting that the nonphysiological substrate, methyl iodide, does not alter the kinetic properties of the enzyme system. Inhibition of acetyl-CoA synthesis was ob-served when concentrations of CoA higher than 10 mM were used; 50% inhibition occurred at -15 mM. A K, value of 14.7 mM was determined for methyl iodide. Concentrations of methyl iodide higher than 10 mM were inhibitory to acetyl-CoA synthesis with 50% inhibition occurring at 17.5 mM; therefore, 5-7.5 mM methyl iodide was used in the acetyl-CoA synthesis reactions. The pH dependence of the synthesis followed a bell-shaped curve with an optimum pH between pH 7.2 and 7.5 and 50% of the activity seen at pH values of 5.2 and 9.3. At higher pH values, Pezacka and Wood (24) found that nonspecific methylation of CODH occurred and less than stoichiometric amounts of the bound methyl groups could be converted to acetyl-CoA in the presence of CO and CoA.
A specific activity of 15 nmol of acetyl-CoA formed min-' mg-' CODH was obtained. By varying the amount of the C/ Fe-SP from 20 to 100 pg under the same experimental conditions, we obtained a maximal velocity of 44 nmol of acetyl-CoA formed min-' (mg CODH)-' from a double reciprocal plot, which would be equivalent to -350 nmol min-' (mg CODH)-1 at 55 "C assuming a rate doubling for each 10 "C increase in temperature.

Determination
of Activity of the C/Fe-SP and the Essential Role of the C/Fe-SP in Acetyl-CoA Synthesis from Methyl Iodide, CO, and CoA-The assay of acetyl-CoA synthesis described above could be used to quantitatively measure the activity of the C/Fe-SP by adding enough CODH that the C/ Fe-SP was rate limiting.
When the amount of C/Fe-SP was in the range of IO-40 Kg the rate was proportional to the concentration of C/Fe-SP and the same specific activity (in terms of C/Fe-SP) was obtained ( Fig. 1). At higher levels of C/Fe-SP, the reaction was limiting in CODH and the specific activity decreased.
Although no detectable acetyl-CoA was formed without addition of either the C/Fe-SP or CODH under the experimental conditions described above, synthesis of acetyl-CoA was observed with the purified fraction of CODH alone at concentrations higher than 0.35 mg/lOO ~1 of reaction mixture. We investigated whether this was due to remaining C/ Fe-SP present in the purified CODH sample or if the methyl binding site of CODH actually was being directly methylated FIG Methyl Transfers in Acetyl-CoA Synthesis 3127 by methyl iodide as had been suggested earlier by Pezacka and Wood (24). First, we varied the amount of the C/Fe-SP and determined the amount of acetyl-CoA formed (Fig. 2). Extrapolation of the plots of the rates of acetyl-CoA synthesis against the amount of C/Fe-SP added gave intercepts of 12.5 and 8.2 pg on the abscissa for the reaction containing 0.36 and 0.216 mg of CODH, respectively. From these values the C/Fe-SP present in the CODH sample were calculated to be -3.5% (mg/mg). Second, by SDS-PAGE followed by staining with Coomassie Blue or by immunoblotting with purified antibody against the C/Fe-SP, similar amounts of the C/Fe-SP were found in several batches of the purified CODH. Third, from the acetyl-CoA synthesis activity (37 nmol/h) of 0.36 mg of CODH alone and its C/Fe-SP content (12 pg), a specific activity of 49 nmol of acetyl-CoA formed min-' (mg C/Fe-SP)-' was calculated, which is very close to the values shown in Fig. 1. Fourth, we obtained a highly purified CODH sample (but still containing 0.8% of the C/Fe-SP) after exhaustive washing of CODH bound to phenyl-Sepharose to attempt the total separation of C/Fe-SP from CODH. This sample showed no "C/Fe-SP-independent" activity except at very high levels of CODH (1 mg/lOO ~1).
3. Methylation of the C/Fe-SP with Methyl Iodide-This reaction is summarized in Equation 3. When poised electrochemically at --500 mV as described under "Experimental Procedures," the C/Fe-SP was rapidly methylated with methyl iodide as methyl donor. Usually a ratio of 0.9 to 1.0 methyl bound to C/Fe-SP was obtained and the UV-visible spectrum of the methylated C/Fe-SP was identical to that of the protein methylated by CHB-H4folate (6). The methylated C/Fe-SP was fully active in acetyl-CoA synthesis when assayed by the method described above. When the C/Fe-SP was methylated using higher concentration of methyl iodide and/ or longer (over 30 min) incubation times, significant inactivation was observed.
Although the midpoint potential for Co*+/Co+ couple of the C/Fe-SP is -504 mV (7), substantial methylation occurred at --350 mV (Fig. 3). We studied a series of methylation reaction at various potentials and at short reaction times (6 min) and found that the extent of methylation was redox potentialdependent (Fig. 3). From the results of Fig. 3 ylation of C/Fe-SP was obtained when the reaction time was extended from 6 to 12 min for the reaction at -400 mV. No significant methylation of the protein was found at redox potentials higher than -300 mV during this time period.
4. Methyl&ion of CODH with Methyl Iodide at Various Redon Potentials and Stimulation of the Methylation by the C/Fe-SP at Low Potentials-The measured potential of a solution of CODH dissolved in anaerobic Tris maleate buffer, pH 7.3, free of reducing agents was -0 mV. In a CODH sample containing a 3.5% contaminant of C/Fe-SP (-17 (nmol of CODH)/(nmol C/Fe-SP)), methylation by methyl iodide was seemingly independent of C/Fe-SP when the reaction was incubated at ambient potential, the standard conditions used by Pezacka and Wood (24) (Table I). Even after a l-h incubation, the methylation was still incomplete. The methylation at low potentials was complete within 10 min (Table I), yielding a ratio of -1 (0.9-1.2) methyl group bound per mol CODH, but still appeared to be independent of C/Fe-SP. With a highly purified CODH sample containing -0.8% C/Fe-SP ((76 nmol of CODH)/(nmol C/Fe-SP)) (Table II), at all potentials studied, even as low as -550 mV, the rate of methylation of CODH was stimulated by addition of the C/ Fe-SP and quantitative yields of methylated CODH were obtained. Thus, the methylation of CODH, as was shown for acetyl-CoA synthesis (above), is strictly dependent on the C/ Fe-SP to act as a methyl carrier. Methylated CODH generated by the reactions at low potentials was fully active in acetyl-CoA synthesis.
Since a cysteinyl thiol had been proposed to be the methyl acceptor site on CODH, we determined the effect of p-hydroxymercuribenzoate (pHMB) on methylation of CODH (Table I). When this reaction was studied earlier (24), 100% inhibition of methylation of CODH was observed at 0.05 mM concentration of pHMB. With our low potential system, we observed little inhibition even at final concentrations of 0.66 mM (molar ratio of pHMB/CODH of 50/l). We determined that the pHMB inhibition was not reversed during the experiment since 18-20 nmol of He-S/nmol CODH were bound to CODH both before and after poising the potential and performing the methylation. DTT was found to slowly demethylate CODH with 50 and 85% of the bound methyl groups lost on incubation with 0.5 mM DTT for 6 and 15 h, respectively.
When the methylated CODH which had been treated previously with pHMB was incubated with CO and CoA, the methyl group was converted to acetyl-CoA.
These results of the pHMB-treated enzyme indicate that thiol groups which are accessible to pHMB are not on either the The methylation was performed as described above under "Experimental Procedures" with 0.35 pmol of CHJ (1700 dpm/nmol). The redox potential was -550 mV (+20) and incubation time was 8 min. Highly purified CODH (containing 0.8% (mg/mg) C/Fe-SP) were used in Exots. 1 and 2. a The value was corrected for methylated C/Fe-SP (nmol) by assuming that C/Fe-SP added was 100% methylated. pathway of methylation of CODH or of acetyl-CoA formation from methylated CODH.
The dependence of the rate of methylation of CODH on redox potential was studied. Within 6 min, -50% of CODH was methylated when the poised potential was --390 mV (Fig. 3). This methylation reaction, like the methylation of the C/Fe-SP, was irreversible since complete methylation of CODH occurred at potentials lower than -300 mV after prolonged incubation.
These combined results indicate that the methyl acceptor site(s) on CODH has to be reductively activated before the methylation can occur. We also studied the effect of methylation of CODH on the activity of the exchange reaction between CO and acetyl-CoA and found that CODH which has been methylated with methyl iodide at -500 mV for 10 min (with 100% methylation) showed the same exchange activity (150 nmol of CO exchanged min-' mg-') as that of the non-methylated CODH. However, methylation of CODH with methyl iodide at ambient potential for 1 h as in Ref. 24 (with -85% methylation) resulted in -40% decrease in the exchange activity. Pezacka and Wood (24)

CH,CODH + [Co+]-C/Fe-SP
No substantial (less than 5%) transfer of the methyl group from ['*C]CH&/Fe-SP to CODH was observed when the two proteins were incubated at ambient potential in the absence or presence of ferredoxin and/or DTT (Table III), even after overnight incubation.
However, significant methyl transfer did occur within 15 min at potentials lower than --300 mV (Table III), clearly indicating that this reaction involves redox chemistry. There was more methyl transfer at --300 mV than at even lower potentials. The methyl transfer from methylated C/Fe-SP to CODH never exceeded 50%, even after longer incubation times (Table III). That a reduction is required to methylate CODH by the methylated C/Fe-SP is not inconsistent with the results of others (24), where a larger amount of CODH was methylated by CH&/Fe-SP in the presence of reduced ferredoxin than with native ferredoxin at ambient potential. There is a minor  discrepancy in that we were unable to methylate CODH with native ferredoxin at a measured potential of -0 mV. We do not know the reason for the discrepancy, but there were two major experimental differences between the two studies: (a) the ferredoxin we used was entirely in the oxidized form at -0 mV, while it is likely that the ferredoxin described in Ref. 24 was at least partially reduced (the rate was stimulated by both the "native" and reduced form of ferredoxin); and (5) the methylated C/Fe-SP (24) was not pure since the reaction mixture contained CODH, CO, ferredoxin, and methyltransferase and was subsequently chromatographed on Sephadex G-25 after the methylation.
Thus, other proteins were present during the methylation of CODH.  (Table IV). The calculated rate of this interprotein methyl exchange reaction was 0.08 nmol min-' (mg CODH)-'.
This was most likely not the maximal rate; however, our present methodology requires separation of the two proteins after the reaction, and 5 min is as fast as we were able to perform the reaction and the separation. After 5 min the reaction was at equilibrium (Table IV). Interestingly, the rate of methyl exchange reaction between the two methylated proteins is independent of the redox potential of the solution. i4CH,-CODH + CO + CoA + "CH3-CO-SCoA + CODH (6) and  (Table V). During the reaction, -50% of the methyl groups were converted to acetyl-CoA, -20% of the methyl groups remained bound to CODH, and another 20% was converted to acetate. The latter was detected as a radiolabeled peak by chromatography on HPLC immediately before the protein with a retention time of 2.5-3 min and coeluted with standard acetate. Thus the conversion of CODH-bound methyl groups to acetyl-CoA and acetate is higher than that (53%) reported earlier (24). In addition, -20% of the methyl group remained bound to CODH. In a separate experiment, after incubation under the same conditions as in Table V  . center(s) on CODH in the scheme below. Several metalloenzymes have been shown to be activated by reduction of their metal sites. The reduced Fe*+ state of globins is required for oxygen binding, and reduction of the heme sites of sulfite reductase (35), cytochrome P-450 (36), and nitrate reductase (37) increases the affinity of these enzymes for their substrate. Methyl-CoM reductase undergoes reductive activation, apparently to reduce the Ni*+ center to the active l+ state that can bind the methyl group of methyl-CoM (38). In addition, when purified aerobically, the Ni/Fe-S hydrogenase requires reductive activation (39) by a one-electron reduction of a redox center with an E, of -310 mV (40).
*CH,-CODH + CHZ-CO-SCoA d (7) CH,-CODH + *CHa-CO-SCoA [2-Y!]Acetyl-CoA was formed after incubation of CODH with acetyl-CoA, clearly indicating that the methyl group bound to CODH is exchangeable with the methyl group on acetyl-CoA (Table VI). This exchange reaction occurred to a significant extent only at low redox potentials.
However, as was found for the methylation of CODH (above), substantial exchange was observed at --400 mV within a few minutes incubation. Over 90% of the [14C]methyl group originally bound to CODH was transferred to acetyl-CoA after incubation at low potentials in the presence of much greater than a molar excess of acetyl-CoA (Table VI). An exchange rate of 3.7 nmol of methyl exchanged min-' (mg CODH)-' could be calculated from the data at -500 mV. This value probably does not reflect a maximum rate since it may not be an initial velocity measurement and increasing the amount of CH,-CODH (a substrate of the reaction) in the reaction mixture may enhance this rate. The maximum rate observed at -0 mV was 130-fold lower. We also determined that addition of nonmethylated CODH or nonmethylated C/Fe-SP had no detectable effect on the rate (1.5-min reaction time) of this exchange reaction. Furthermore, addition of the nonmethylated proteins to the reaction did not effect the final amount of exchange; i.e. 100% of the label from the methyl group of CH&ODH still exchanged with the methyl of acetyl-CoA. DISCUSSION Several reactions studied here require reduction of a site on CODH or the C/Fe-SP to observe either maximal activity or, in some cases, to observe even detectable activity. We do not yet know the actual oxidation states of the metal centers on CODH involved in catalysis or substrate binding, therefore it is premature to attempt to assign oxidation states to the Based on our results, Fig. 4 is a postulated scheme for the pathway of acetyl-CoA synthesis. The methyl transfer reactions involved in acetyl-CoA synthesis can be broken down into three steps: methylation of the C/Fe-SP, methyl transfer to CODH, and methyl migration to form an acetyl-enzyme intermediate.
Here we have studied the sequence of methyl transfer reactions in the pathway with methyl iodide as the methyl donor since this simpler system does not require methyltransferase, an enzyme purified earlier (41). All the reactions performed here were with an enzyme that lacks any protein with a subunit molecular weight of the CODH disulfide reductase. This is important since it was recently suggested that the active form of CODH contains three subunits (22). Thus, our results indicate that the active form of CODH is a two-subunit enzyme characterized earlier (42).

Acetyl-CoA Synthesis from Methyl
Iodide, CO, and CoA by CODH and the C/Fe-SP-We found, as before (23), that acetate and acetyl-CoA are both formed during the synthesis from a methyl donor, CO, and CoA in reactions that require both the C/Fe-SP and CODH. Our reaction conditions are highly simplified in that methyl iodide, CO, CoA, C/Fe-SP, and CODH are the only reactants required. Disulfide reduc- tase and/or reducing agents are not required. CoA has no effect on the rate or the total amount of acetate formed when the methyl and carbonyl donors are in excess. Formation of acetate is strong evidence for an enzyme-bound acetyl intermediate which we propose is hydrolyzed to acetate.
Formation of methylated C/Fe-SP appears to occur slower than transfer of the methyl from methylated C/Fe-SP to CODH (see Fig. 3). Formation of methyl-cob(III)amide from methyl iodide and Co+ apparently is irreversible since at longer times at these positive potentials the C/Fe-SP becomes completely methylated. In addition, the Co center is methylated at a potential -200 mV more positive than its midpoint potential. These results are highly significant with respect to methylation of the C/Fe-SP under physiological conditions. One can calculate, based on the midpoint potential (-504 mV; Ref. 6) of the Co*+/+ couple of the C/Fe-SP, that in the absence of methyl iodide 0.27% of the C/Fe-SP would be expected to be reduced from Co'+ to the active methyl-accepting Co' form at -350 mV; however, 30% of the C/Fe-SP is methylated within 6 min. Thus, under physiological conditions in which the reduced Co' species is immediately trapped by the methylation, the cell does not need to deliver electrons at very low potentials to generate methyl-cob(III)amide on the C/Fe-SP. Recently, Banerjee et al. (9) have studied the methylation of methionine synthase in which the Co*+'+ redox system is coupled to the irreversible methylation by S-adenosyl-L-methionine and found that the apparent midpoint potential of the redox couple was increased by several hundred mV. It will be interesting to determine the dependence on redox potential of methylation of the C/Fe-SP with CHs-H4folate as methyl donor. Methylation of CODH at Controlled Redon Potentiak-These results are interpreted according to Steps 2 and 3 of Fig. 4. The C/Fe-SP is required for acetyl-CoA synthesis from CHs-Hdfolate (6,43) and, thus, for methylation of CODH by CH,-H,folate.
Pezacka and Wood (24) proposed that the active site of CODH is directly methylated by methyl iodide without the mediation of the C/Fe-SP. Here, we have shown that the C/Fe-SP is required for both methylation of CODH and acetyl-CoA synthesis with methyl iodide as methyl donor. Very little C/Fe-SP is required for this methylation; e.g. less than a 1% contaminant of C/Fe-SP in CODH preparations is enough to catalyze significant methylation of CODH (Table  II).
There is a greater than lo-fold acceleration of the rate of methylation of CODH by methyl iodide at low redox potentials. That the rates of methylation of CODH and of the C/ Fe-SP are nearly identical (Fig. 3) suggests that the ratelimiting step in the methylation of CODH is methylation of the C/Fe-SP. Then, after formation of the methylated C/Fe-SP, this group is transferred to CODH. Based on this experiment alone, it is not clear whether CODH or the C/Fe-SP or both proteins require reduction since we cannot distinguish two separate reductions. Clearly the C/Fe-SP must be reduced from the Co*+ to Co' state to activate it for methylation (6,7). That the methylation of CODH from the methylated C/ Fe-SP is redox potential-sensitive is strong evidence that redox chemistry is not only required in the methylation of the Co center of the C/Fe-SP, but also in reductive activation of a nucleophilic site on CODH. Methylation of this site on CODH from the methylated C/Fe-SP never reaches a full equivalent/m01 of CODH; however, with methyl iodide as methyl donor, CODH is fully methylated.
In addition, methylation via methylated C/Fe-SP exhibits a peculiar dependence on redox potential. It requires a potential of --350 mV, however at even lower potentials, the reaction appears to be less complete. Our explanation for these findings involves reactions la, lb, 2, and 3 (Fig. 4). When excess methyl iodide is the methyl donor, both the C/Fe-SP and CODH can be stoichiometrically methylated, however, with approximately equimolar amounts of methyl-C/Fe-SP and CODH, the extent of methylation of CODH is determined by the equilibrium ratio of [methyl-C/Fe-SP] [CODH]/[C/Fe-SP] [methyl-CODH] as well as the redox equilibria between the Co*+'+ and CODH,,/CODH,d couples. At lower potentials, the C/Fe-SP is more extensively reduced and once the methyl group is transferred to CODH, at low potentials, the ratio of Co+/Co*+ becomes higher, thus the equilibrium could favor methylation of the C/Fe-SP. However, at higher potentials (in the -350 mV range), the equilibrium ratio of Co+/Co'+ is lower, thus, once the methyl transfer to CODH occurs, the initially formed Co' product would equilibrate with Co'+ which cannot accept a methyl group from CODH. This analysis implies that the methyl acceptor site on CODH is less nucleophilic than the Co' C/Fe-SP. It may also indicate that the redox couple of the methyl acceptor site on CODH has a more positive potential than the Co'+/+ couple (as shown in Fig. 3); however, there is not a strict relationship between nucleophilicity and reduction potential.
Pezacka and Wood (24) proposed that the methyl binding site on CODH is a cysteine. They incubated CODH with radiolabeled methyl iodide or the methylated corrinoid protein, acid-hydrolyzed the peptide bonds, chromatographed the amino acid residues, and identified a radiolabeled methylcysteine residue. When methylated CODH was incubated with CO and CoA and the same analysis repeated, this methylcysteine peak was absent. We now consider that the cysteinyl thiol may not be the essential methyl acceptor group on CODH for acetyl-CoA synthesis.
Based on our results, there are compelling reasons to propose a metal center rather than the cysteinyl thiol as the initial nucleophile.
First, there appears to be more than one methyl binding site on CODH based on three exchange reactions conducted with methylated CODH which was prepared by our standard conditions (at low potentials for short reaction times). The occurrence of the exchange between the methyl of the methylated CODH with the methyl group of acetyl-CoA (discussed below) and with the methyl of the methylated C/Fe-SP requires two methyl acceptor sites. In addition, methylated CODH is fully active in catalysis of the exchange reaction between acetyl-CoA and CO. Second, methylation of CODH occurs over a wide pH range from 5.5 to 7.5 (24). This pH profile is not consistent with the normal profile for cysteine ionization and could reflect changes in coordination state around a metal center although there are examples in which cysteine residues in proteins have low pK,, values. Third, the dependence on redox potential of the methylation of CODH is unlike that which would be expected for reduction of a dithiol to an active thiolate. Most disulfide/dithiol redox reactions occur in the range of -200 to -300 mV; yet, as discussed above the redox center which is methylated has a redox potential of <-350 mV. Fourth, methylation of CODH at low potentials is not inhibited by pHMB which is a potent thiol reagent. In addition, acetyl-CoA is synthesized from the pHMB-treated and methylated CODH in the presence of CO and CoA. It, however, is possible that the active thiol may not be accessible to pHMB.
Fifth, based on model chemistry, formation of an acetyl intermediate would be expected to involve a methyl migration (CO insertion) reaction. There is ample precedence for carbonyl insertions occurring on metal centers (see Refs. 14 and 44, for example), but we are not aware of such reactions occurring at a thioether. Raybuck et al. (15) also proposed a methylmetal intermediate based on the stereochemical retention of configuration during the acetyl-CoA/CO exchange. How, then, can one explain the findings which suggested that a cysteinyl residue is the methyl acceptor (24) in the light of our findings that suggest that a metal center is the active methyl acceptor? We suggest that the methyl group may migrate from a metal center to a cysteinyl residue. If formation of methylcysteine is due to this migration, it is likely that this migration would occur even in the active enzyme because we would not expect the acidic condition of the hydrolysis (for the amino acid analysis) of methylated CODH to promote this migration. Support for this methyl migration postulate is the observation that DTT can slowly demethylate CODH (above). The migration may be reversible and help explain our results that indicate there are two methyl binding sites on CODH.
Which metal center on CODH would be most likely to act as a methyl acceptor? One such candidate is the Ni-Fe-C center since it appears to be the binding site for the CO which will become the carbonyl of acetyl-CoA. The midpoint reduction potential for this center is --560 mV7 which is lower than the Co'+'+ couple. There are other metal centers in CODH,'," however, and further studies will be required to determine which one of these could be the methyl binding site.
Pezacka and Wood (24) considered both a reductive cleavage and a nucleophilic mechanism to explain the methyl transfer from the methyl-cob(III)amide to CODH. The reductive cleavage would generate a radical methyl intermediate. That the rate of the methyl-CODH/methyl-C/Fe-SP exchange reaction is independent of redox potential and the inability to reduce methyl-cob(III)amide to methylcob(II)a at redox potentials of -600 mV or greater (7) are inconsistent with the reductive cleavage mechanism and strongly supports a heterolytic cleavage of the methyl-cobalt bond. Methylation of CODH is proposed (reaction 3, Fig. 4) to occur by a nucleophilic methyl displacement reaction similar to the reaction of the methyl of CHs-Hlfolate to Co+ to form methyl-Co3+.
Reactions Performed by Methylated CODH-By performing an exchange reaction between methylated CODH and the methyl group of acetyl-CoA (Equation 6, above), a reaction that had never previously been proposed, we found that there is more than one methyl binding site on CODH. If there were only a single methyl binding site on CODH, then blockage of this site by methylation would preclude the ability of CODH to accept the methyl group from acetyl-CoA.
If, however, there is a second methyl acceptor site on CODH, then this reaction would be possible. Interestingly, this reaction requires reductive activation as do the methylation reactions described above. We are assured that small amounts of nonmethylated C/Fe-SP or CODH which could remain as contaminants in our reaction mixtures were not involved in catalysis of the exchange by serving as the proposed second methyl acceptor since addition of the nonmethylated proteins had no effect on the rate of the reaction.
Further studies will be required to propose a possible mechanism for this exchange. Either binding or cleavage of acetyl-CoA requires reductive activation of CODH. Recent experiments" indicate that the Ni-Fe-C EPR signal is generated upon reaction of CODH with acetyl-CoA at low potentials, indicating that reduced CODH can cleave acetyl-CoA into its component groups and the Ni-Fe-C signal is elicited by the ' C. M. Gorst, and S. W. Ragsdale, manuscript in preparation. " C. M. Gorst, and S. W. Ragsdale, manuscript in preparation. carbonylated form of the enzyme. Methylated CODH apparently can also perform the cleavage. For the methyl exchange to occur a methyl transfer is likely to occur. This transfer could be to another nearby metal center or possibly the cysteinyl residue proposed earlier (24). In the next step, the acetylmetal intermediate could occur with either the labeled or unlabeled methyl group and the addition of CoA would then generate labeled acetyl-CoA.
A second exchange reaction, between CH&ODH and methylated C/Fe-SP, another reaction that had never previously been proposed, also demonstrates that there probably are at least two methyl binding sites on CODH. Methylated CODH could only perform this exchange reaction if it can accept a second methyl group from the methylated C/Fe-SP and interconvert the C3H3 and the 14CH3 in 14CH3-CODH and C3H3-C/Fe-SP (Equation 5, above). In contrast to the methyl transfer reactions and also the exchange between methylated CODH and acetyl-CoA, this methyl exchange reaction is not dependent on the applied redox potential.
A possible explanation for this is that the [3H]methyl of methyl-cob(III)amide is transferred to the metal site on the ['4C]methyl-CODH through a process in which the methyl-cob(II1) amide and the methyl metal on CODH are extremely close, and bond breaking and bond making could be concerted.
It had been earlier shown that acetate was formed from CO and methylated C/Fe-SP in the presence of CODH and in the absence of CoA, indicating the existence of an acetyl-CODH intermediate (23,43). Ramer et al. (22) suggest that formation of an acetyl-CODH intermediate may be the rate-limiting step in synthesis of acetyl-CoA from the methylated C/Fe-SP, CoA, and CO. Here we have shown that acetate is formed during the synthesis of acetyl-CoA from methyl iodide and CO both in the presence and absence of CoA. We also have demonstrated that methylated CODH can be converted into acetate in the absence of CoA presumably via this acetyl-CODH intermediate.
Reaction 5 (Fig. 4) followed by hydrolysis (shown in dotted lines in Fig. 4) describes the formation and hydrolysis of this acetyl intermediate.

CONCLUSIONS
Step 1 involves the formation of the methylated C/Fe-SP from COZ and new results regarding these reactions are fully discussed above.
Step 2 is the reductive activation of a nucleophilic metal center which is the methyl acceptor site (M,,) on CODH. We do not know the midpoint potential of this center since all the reactions studied here couple a redox reaction to a chemical step which perturbs the redox equilibrium and, thus, is not amenable to a strict interpretation by the Nernst equation. Since several of the methylation and methyl exchange reactions studied here are significantly stimulated by poising the potential below --350 mV, it is clear that the midpoint potential of this nucleophilic metal center is fairly negative and could be isopotential to the Co2+'+ couple of the C/Fe-SP (-504 mV). As discussed above, however, it is likely that such low potentials would not be required in the cell, since the redox reaction is coupled to the methylation and then to the final steps of the synthesis.
Step 3 is the methylation of CODH by the methylated C/ Fe-SP which we propose involves the nucleophilic attack of a reductively activated metal center on CODH on the methyl group of methyl-cob(III)amide.
Step 4 involves the addition of CO to the methylated metal center. Earlier results indicate that the site for CO binding is a mixed metal center consisting of nickel and iron (12). This Ni-Fe center also may be the site represented by Mred in the