Acetate biosynthesis by acetogenic bacteria. Evidence that carbon monoxide dehydrogenase is the condensing enzyme that catalyzes the final steps of the synthesis.

The purified carbon monoxide dehydrogenase from Clostridium thermoaceticum is the only protein required to catalyze an exchange reaction between carbon monoxide and the carbonyl group of acetyl-CoA. This exchange requires that the CO dehydrogenase bind the methyl, the carbonyl, and the CoA groups of acetyl-CoA, then equilibrate the carbonyl with CO in the solution and re-form acetyl-CoA. CoA is not necessary for the exchange and, in fact, inhibits the reaction. These studies support the view that CO dehydrogenase is the condensing enzyme that forms acetyl-CoA from its component parts. Carbon dioxide also exchanges with the C-1 of acetyl-CoA, but at a much lower rate than does CO. At 50 degrees C and pH 5.3, the optimal pH, the turnover number is 70 mol of CO exchanged per min/mol of enzyme. Low potential electron carriers are stimulatory. The Km app for stimulation by ferredoxin is 50-fold less than the value for flavodoxin. Neither ATP or Pi stimulate the exchange. The EPR spectrum of the CO-reacted enzyme is markedly changed by binding of CoA or acetyl-CoA. Arginine residues of the CO dehydrogenase appear to be involved in the active site, possibly by binding acetyl-CoA. Mersalyl acid, methyl iodide, 5,5-dithiobis-(2-nitrobenzoate), and sodium dithionite inhibit the exchange reaction. A scheme is presented to account for the role of CO dehydrogenase in the exchange reaction and in the synthesis of acetate.

exchanges with the C-1 of acetyl-coA, but at a much lower rate than does CO. At 50 "C and pH 5.3, the optimal pH, the turnover number is 70 mol of CO exchanged per min/mol of enzyme. Low potential electron carriers are stimulatory. The KmaPp for stimulation by ferredoxin is 50-fold less than the value for flavodoxin. Neither ATP or Pi stimulate the exchange. The EPR spectrum of the CO-reacted enzyme is markedly changed by binding of CoA or acetyl-coA. Arginine residues of the CO dehydrogenase appear to be involved in the active site, possibly by binding acetyl-CoA. Mersalyl acid, methyl iodide, S,S-dithiobis- (2nitrobenmate), and sodium dithionite inhibit the exchange reaction. A scheme is presented to account for the role of CO dehydrogenase in the exchange reaction and in the synthesis of acetate.
Acetogenic bacteria synthesize acetate from a number of carbon and energy sources, including hexoses and inorganic substrates such as CO or C02/H2 (1). The pathway of acetate biosynthesis in these bacteria involves a reduction of CO, to methyltetrahydrofolate, transmethylation to a corrinoid enzyme, and a condensation of the methyl group of the methylated corrinoid with coenzyme A and a C1 unit derived from CO,, CO, or the carboxyl of pyruvate (see Refs. 2,3,and 4 for review).
Carbon monoxide dehydrogenase was first implied to be a key intermediate in this sequence of reactions in investigations which showed that a protein fraction, F3, which contained a high level of CO dehydrogenase, was essential in the synthesis of acetyl-coA from pyruvate and methyltetrahydro-*This research was supported by Grant GM 24913 from the National Institutes of Health. The Varian E112 EPR spectrometer (Case Western Reserve University) used in these studies was provided by the National Institute of General Medical Science Grant 27519. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. folate (5) and from CO and methyltetrahydrofolate (6). Pezacka and Wood (7) demonstrated that CO dehydrogenase was required for the synthesis of acetyl-coA from CO, and HZ. This same F3 fraction was shown to be the only component which was required to carry out an exchange reaction between [l-"Clacetyl-CoA and carbon monoxide (Equation 1) and the authors (6) showed that during this [l-"Clacetyl-CoA + "CO e "CO + [l-'2C]acetyl-CoA (1) conversion, there was no net decrease in acetyl-coA. They postulated that a nickel-bound Cl unit is intermediate in the exchange reaction and that the corrinoid enzyme is the methyl acceptor. In later studies, the F B fraction was reported to contain a number of enzymes, including CO dehydrogenase, hydrogenase, a corrinoid enzyme, and formate dehydrogenase and the corrinoid enzyme was purified (3).
We have investigated the exchange reaction (Equation 1) in detail, first, since it required the least number of components (ie. methyl transferase, pyruvate-ferredoxin oxidoreductase, and phosphotransacetylase were not required), second, because it should give insight into the mechanism of the transformation of the methyl, carbonyl, and CoA moieties into acetyl-coA, and, third, because CO dehydrogenase had been shown to be a necessary component for the exchange reaction (6). CO dehydrogenase was recently purified to homogeneity and was shown to be a hexameric enzyme containing 6 nickel, 1-3 zinc, and approximately 33 iron and inorganic sulfur in the form of iron sulfur clusters (8). Our studies demonstrate that the purified CO dehydrogenase catalyzes the exchange reaction and based on these results, we present a scheme which explains the key role of CO dehydrogenase in the synthesis of acetyl-coA.

EXPERIMENTAL PROCEDURES' AND RESULTS
Comparison of CO Dehydrogenase and Exchange Reaction during Purification The bacteria were grown in the presence of =Ni and the CO dehydrogenase was purified as outlined in Table I  ND, we were unable to determine the exchange activity in the crude extract because the enzyme is dilute in pH 7.6 Tris-HC1 buffer and it is necessary to add a large amount of extract to the assay mixture, which caused an increase in pH to greater than 7.0 (see later for pH optimum). The concentration of protein in the phenyl-Sepharose fractions was too dilute to determine the exchange activity. e Only 5% of the HTP protein was applied to this column. Only 14% of the HTP protein was applied to this column.
purification. The most important value is the ratio of CO dehydrogenase activity to exchange activity, based on units and milliunits/mg of protein, which has an average value of 2.3, with very little variation during the purification procedure. Almost identical results are obtained based on the ratios of counts/min of 63Ni/~nit of CO dehydrogenase and per milliunit of exchange. These results indicate that the nickelcontaining CO dehydrogenase is responsible for both activities. Both activities are linear with concentration of CO dehydrogenase. We have found that some samples of enzyme lose exchange activity during storage faster than the CO dehydrogenase activity. In fact, one sample of CO dehydrogenase which was stored for a year without reducing agent maintained its CO dehydrogenase activity (150 units/mg, approximately 66% loss of activity); however, the exchange activity was totally lost. Therefore one can affect the exchange activity without necessarily affecting the CO dehydrogenase activity (see below for modification experiments).

Proof That Minor Contaminants Are Not Required as Catalysts for the Exchange Reaction
After hydroxylapatite chromatography, the minor contaminant (band 4) was approximately 33% of the protein (Lane I of Fig. 1). If the protein of band 4 was indeed stimulatory to the exchange reaction, then removal of this protein should cause a decrease in specific activity and addition of the protein back to the exchange reaction assay mixture should cause an increase in exchange activity. Gel filtration (Step 5) removed most of the protein giving rise to band 4 (Lane 3 of Fig. 1) and also caused an increase in specific activity of both the CO dehydrogenase and the exchange activities (Table I). Based on densitometry scanning, the protein, after Step 5, was 99% pure. Addition of the purified protein of band 4 (Lane 2 of Fig. 1) in fairly high amounts (33 and 66 pg) did not increase the activity. It is unlikely that the protein of band 4 was inactivated by the time of the assay, since the exchange activity was determined within 1 h after elution from the column. Therefore, the small amount of contaminant due to the protein of band 4 is not responsible for catalysis of the exchange reaction.
The other minor contaminant, giving rise to band 3, could be removed by agarose-CoA chromatography (Step 6), and, in fact, it binds very tightly to agarose-CoA. It is not removed with 2 M urea; however, 6 M urea elutes this protein. After elution, this protein also did not stimulate the exchange reaction and, as in the case of the removal of the protein of band 4 by gel filtration, the specific activity increased after this affinity chromatography step (Table I).
That CO dehydrogenase is the only catalyst necessary for the exchange reaction is in contrast to results found previously (8,9). The major source of the discrepancy is that previously 2 mM sodium dithionite was added to all enzyme solutions during the purification of the CO dehydrogenase, and dithionite is a strong inhibitor of the exchange reaction. In addition, since electron carriers, which stimulate the exchange reaction were not added, an even lower percentage of the maximum amount of exchange activity/mg of added protein could be detected. These results show conclusively that CO dehydrogenase is the only protein that is required for catalysis of the exchange reaction between CO and the C-1 of acetyl-coA.

Stimulation of the Exchange Reaction by Low Molecular Weight Electron Carriers
The exchange activity was approximately 5000-fold lower than the CO to COz reaction, so a large excess of CO dehydrogenase was necessary to show that the exchange reaction occurred. Therefore, we attempted to purify a component that would stimulate the exchange activity.
Isolation of Ferredoxin and Evidence ThabZt Stimulates the Exchange-An oxygen-sensitive factor which stimulated the exchange reaction was purified. The purified factor migrated on G-50 superfine gel with an elution volume similar to that of ferredoxin (Mr = 6000). The exchange reaction occurs without addition of electron carrier, but a large increase in activity is obtained with addition of ferredoxin (Fig. 2). The amino acid composition of the factor corresponded to ferredoxin I1 (10) and was clearly different from ferredoxin I (11) (see "Experimental Procedures"). However, we have recently isolated two peaks of ferredoxin from DEAE-Sephacel and found that both proteins stimulated the exchange activity. Two ferredoxins, I and 11, have previously been purified from C. thermoaceticum and extensively characterized (10, 11).
Test of Other Electron Carrier Proteins-Several different electron carriers, isolated from Clostridium formicoaceticum (12), were tested for stimulatory activity (Table 11). Only the low potential carriers stimulated the exchange reaction. Nonlinear Lineweaver-Burk plots of l/concentration uersus 1/ activity were obtained, and the K , values, determined from the linear portion of the graph (at high concentrations of electron carrier) were 22.2 p~ for flavodoxin and 0.43 p~ for ferredoxin ( Fig. 2B). Thus the K , for ferredoxin is 50-fold lower than the K,,, for flavodoxin. The carriers were reduced upon addition of CO dehydrogenase to the assay mixture. This was most obvious in the flavodoxin reaction when the bright yellow color of flavodoxin was bleached, corresponding to the reduction of FMN to FMNH,. No semiquinone intermediate was seen, indicating the low potential of the CO/CO, couple and the ability of CO dehydrogenase to use flavodoxin as a two electron acceptor (8,9). Methyl viologen also stimulated the exchange reaction between [1-"Clacetyl-CoA and CO; however, at high concentrations (greater than 0.5 mM), it was found to inhibit.

Role of Electron Carriers in the Exchange Reaction-When
CO wa8 replaced with CO, as the gas phase in the exchange reaction, 16% of the "C from the C-1 of acetyl-coA was exchanged with CO, as compared to 44% with CO as the gas phase. When Nz was the gas phase, 3% of the counts were lost from the l-14C during the reaction. These results indicate that CO, and not CO,, is the actual form of carbon undergoing exchange with the C-1 of acetyl-coA and that the role of the low potential electron carriers is to stimulate an internal electron transfer that occurs during the cleavage of the methyl and CoA bonds to the carbonyl and the resynthesis of acetyl-CoA (see below for a possible mechanism).

Effect of Modifying Reagents on the Exchange Reaction and the CO Dehydrogenase Reaction
Phenylglyoxal, methylglyoxal, and butanedione, which are potent arginine modifiers (13,14), all caused inhibition of the CO dehydrogenase exchange reaction (Table 111). When arginine reagents were added directly to the assay mixture, strong inhibition occurred within the time course of the assay. When the enzyme was preincubated with 24 mM butanedione and then added to a complete assay mixture at specified times, approximately 50% inhibition occurred within 5 min, and no further inactivation occurred up to 1 h. Under these conditions, the presence of CO in the enzyme solution had no effect on the inhibition. The CO to COZ activity of the CO dehydrogenase was not affected by the arginine reagents, even in high concentrations under conditions that caused 90% inhibition of exchange activity. When 63Ni-enzyme was reacted with 20 mM methylglyoxal and gel filtrated on the Bio-Gel P-6DG  The nanomole of acetyl-coA exchanged was calculated by subtracting the counts/min of acetyl-coA at the end of the reaction from the counts/min at time 0. This difference is divided by the initial specific activity of the acetyl-coA to obtain the nanomole exchanged. This value is therefore a miniumum value for the exchange activity since some of the CO that adds back to the enzyme is labeled.  column, 96% of the nickel was retained in the protein fraction.
Amino acid analysis of the phenylglyoxal-reacted enzyme demonstrated that approximately 6 arginines were modified. The actual number of arginines modified will be determined using radioactive phenylglyoxal.
There are 117 Arg in the complete protein and a small change from this large number cannot be determined accurately by amino acid analysis.
Mersalyl acid, which disrupts iron-sulfur clusters (15), causes complete inhibition of both the exchange and the CO dehydrogenase activities a t 10 mM concentration (Table 111). It seemed likely that this inhibition was due to destruction of the enzyme's iron-sulfur clusters; however, when the 63Nicontaining enzyme was incubated with 5 mM mersalyl acid for 50 min in 100 mM KPi buffer, 70% of the nickel was removed from the enzyme. In addition, a t 10 mM concentration, the sulfhydryl reagent, 5,5'-dithiobis-(2-nitrobenzoate), strongly inhibits both the exchange and the CO to COz  (-) and presence (---) of 12 mM CoA. Bottom spectra: 0.05 mM CO dehydrogenase in KPi/DTT (50 mM/5 mM), pH 6.8, in the absence (-) and presence ( reactions. At 40 mM concentration, methyl iodide inhibits both reactions as well. Interestingly, the nickel chelator, dimethylglyoxime, had no effect on either the exchange or the CO dehydrogenase activity after a 30-min incubation with 2 mM glyoxime. These results with the glyoxime reagent indicate that the nickel is chelated quite tightly in the enzyme. ATP was found to have no effect on the exchange reaction either in the presence or absence of ferrous ammonium sulfate, which is in contrast to the results obtained by Hu et al. (6) with a n impure preparation. It seemed likely that a phosphate source might be necessary so MES' (Na salt) buffer was substituted for phosphate buffer in the exchange reaction.
The replacement of phosphate by MES had absolutely no inhibitory effect on the exchange reaction.
Interaction of CO Dehydrogenase with CoA EPR Spectra-When the CO dehydrogenase is treated with CO, a paramagnetic nickel-carbon species is formed which has g values at 2.08 and 2.02 (16). We have studied a number of preparations of CO dehydrogenase and found, in many cases, a thirdg value at g = 2.05 (Fig. 3). The relative intensity of the 3 components varies significantly among different preparations of the enzyme. In the spectra shown at the top * The abbreviations used are: MES, 2-(A"morpholino)ethane SUIfonate; DTT, dithiothreitol. of Fig. 3, the g = 2.05 signal is the major component and in the bottom spectra, it is the minor component. A remarkable change occurs in the EPR spectrum when the enzyme is treated with CoA, and the symmetry becomes more axial (Fig.  3). The same change is seen upon treatment of CO dehgdrogenase with acetyl-coA (not shown). Enzyme incubated with CoA or acetyl-coA in the absence of CO is EPR silent. When the enzyme is treated with CoA, gel filtrated to remove all the nonbound CoA, and then treated with CO, the resulting spectrum has three g values and is identical to the untreated enzyme. Thus the change in EPR properties is reversible. Theory predicts that each line produced by paramagnets associated with IzC (I = 0) will be broadened or split into a doublet when the carbon is replaced by 13C (I = 1/2) (17). Substitution of 13C0 for "CO causes a substantial broadening of the line widths of all three components of the nickel-carbon signal, as was found earlier for the g = 2.08 and 2.02 components (16). The g = 2.08, 2.05, and 2.03 line widths are broadened by 2,5, and 13 G, respectively. The g, component (g = 2.03) clearly showed a splitting into a doublet as was found earlier (16). These results indicate that CoA and acetyl-CoA bind to the CO dehydrogenase near the paramagnetic nickel-carbon center.
Other Evidence of Acetyl-coA Binding to CO Dehydrogenase-We have found that CO dehydrogenase binds to hexane-CoA-Sepharose 4B. The enzyme is eluted with approximately 0.5 M NaC1; whereas, most other proteins elute in the wash (50 mM KPi or Tris-HC1 buffer).
The 'H of [acetyZ-3H]acetyl-CoA, after incubation with CO dehydrogenase is clearly separated from the enzyme peak by gel filtration (Fig. 4). When [1-"Clacetyl-CoA is reacted with CO dehydrogenase in the presence of CO, a significant amount of radioactivity is retained in the enzyme peak (Fig. 4). This complex between "C of C-1 acetyl-coA and CO dehydrogenase is, therefore, not due to binding of acetyl-coA to CO dehydrogenase, but is apparently due to binding of "CO which is formed from [l-'4C]acetyl-CoA by the exchange reaction. The lack of complete coincidence of the radioactivity and the CO dehydrogenase may be due to partial dissociation which occurs during the chromatography. Previously, Pezacka and Wood (7) showed that 14C02 plus Hz and the C-1 of pyruvate formed a C-1 complex with CO dehydrogenase that could be isolated by gel filtration. We have been unable to detect, by gel filtration techniques, an analogous complex of "CO with CO dehydrogenase by directly reacting CO dehydrogenase with 14C0.

DISCUSSION
Based on the results presented here, a new scheme for the biosynthesis of acetate is proposed (Fig. 5) which, in contrast to previous schemes (2)(3)(4)6), places CO dehydrogenase as the central enzyme in the pathway. This proposal is based on the observation that CO dehydrogenase is the only enzyme that is required for catalysis of the exchange between [ 1-'*C]acetyl-CoA and CO. The reactions involved in the exchange are enclosed by the dotted lines of Fig. 5. For the C-1 to be converted to CO, the bonds between C-2 and C-1 and between C-1 and CoA must be cleaved forming separate methyl, carbonyl (C1), and CoA groups which are bound to the enzyme as shown schematically in Equation 2, where X, Y, and 2 are sites on the CO dehydrogenase. In CHs-CO-SCOA X-CHI + Y-CO + Z-COA (2) addition, the Y-CO must equilibrate with CO in the solvent. CO has been shown to bind to CO dehydrogenase and form a paramagnetic nickel-carbon species (10, 16; Fig. 3). Although After 15 min reaction at 50 "C under a CO gas phase, the enzyme solution was applied to a 24-ml(l.O X 30 cm) Bio-Gel P-6DG desalting column equilibrated with Tris-HCl/KCl/DTT (50 mM/0.2 M/5 mM), pH 7.0, and 0.5-ml fractions were collected. Aliquots were assayed for CO dehydrogenase activity, radioactivity, and concentration of Pi. In a parallel experiment, [acetyl-'H]acetyl-CoA was run on the column in the absence of CO dehydrogenase. The elution pattern was identical to the migration of the radioactivity in A. The second peak of radioactivity is apparently a contaminant of the acetyl-coA or is due to an interaction of acetyl-coA with the gel. not shown in Fig. 5, the "Y" in Equation 2 apparently is the nickel site of the CO dehydrogenase and the paramagnetic complex is speculated to be the C, intermediate referred to as Y-CO. Since ATP and Pi are not needed for the exchange reaction, this Ni-CO species is probably not phosphorylated, as had been suggested by Hu et al. (6). Ferredoxin is postulated to be involved in this step of Fig. 5 since electron carriers stimulate the exchange and since this step involves a redox change in the nickel. We were able to trap this C1 intermediate by gel filtration by using the C-1 of acetyl-coA as the source of CO. However, we have been unable to detect an intermediate by gel filtration when CO dehydrogenase is reacted directly with 'CO. This is apparently due to the low solubility of CO and to exchange between solvent CO and enzymebound CO.
Our results demonstrate that the Cl, but not the methyl or CoA group, form a tightly bound complex with the CO dehydrogenase that can be isolated by gel filtration. Nevertheless, the methyl and CoA groups must bind to the CO dehydrogenase since there are no other acceptors for these groups during the exchange. That the exchange is inhibited by CoA is ferredoxin, respectively. F, is a protein which is necessary for acetate synthesis but whose function is unknown. CHaTr represents methyl transferase; F-THFS, 10-formyltetrahydrofolate synthetase; FD, formate dehydrogenase; and CODH, CO dehydrogenase. THF is tetrahydrofolatq and [Co]E is the corrinoid enzyme.
x CHXTr evidence that CoA is bound to the CO dehydrogenase. In addition, EPR results demonstrate binding of CoA and acetyl-CoA to the enzyme. Acetyl-coA and CoA modify the environment around the nickel-CO paramagnetic center from a rhombic to a more axially symmetric species (Fig. 3). Since the nickel in the enzyme is labile to mersalyl, sulfhydryl ligands to the nickel in CO dehydrogenase seem likely. Nickel (111) peptide complexes with all nitrogen or all sulfur ligands have EPR spectra with axial symmetry; whereas, the mixed N,S donor sets have rhombic spectra (18). It is possible, therefore, that the CoA sulfhydryl (formed from free CoA or acetyl-coA via the exchange) binds directly to the nickel creating a more symmetric environment around the nickel and the observed changes in the EPR spectrum. CoA analogues will be tested for similar effects on the EPR spectrum in order to determine which moieties of the coenzyme are primarily responsible for the change in the EPR spectra. Interestingly, acetyl-coA has no effect on either the K, or V,, of the CO oxidation reaction (8). We propose that the CoA may be bound to arginine groups of CO dehydrogenase. This is based on the inhibition of the exchange, but not the CO dehydrogenase reaction, by butanedione, methylglyoxal, and phenylglyoxal. Detailed studies by other workers (13, 14) have shown that the pyrophosphate bridge of coenzymes can be bound via the guani-din0 linkage of arginine residues and that butanedione, methylglyoxal, and phenylglyoxal prevent this binding.
Therefore, we propose that, in the exchabsge reaction, [l-Wlacetyl-CoA binds to the CO dehydrogenase and is separated into the methyl, 14Cl, and CoA moieties. In the next step, the 14C1 (Y-CO) exchanges with CO in the solution, and, in the final step, the groups recombine to form [l-lzC]acetyl-CoA.
In the synthesis of acetyl-coA from CO, and Hz, the required reductive capacity is supplied by the Hz, and, with CO, is supplied in the formation of CO, by the CO dehydrogenase. The CO, is converted via formate to formyltetrahydrofolate which is reduced to methyltetrahydrofolate. The reactions have been completely characterized by Ljungdahl and coworkers (2,19). A methyl transferase catalyzes the synthesis of a methylated corrinoid protein from methyltetrahydrofol-Pi + &P I ATP COe ate (3). This methyl group is then transferred to CO dehydrogenase. Then CoA and CO bind to the CO dehydrogenase, as shown in Fig. 5, and are condensed with the methyl group to form acetyl-coA, which is utilized in anabolic pathways. We do not know if free CoA or the methyl group of the methylated corrinoid protein are transferred directly to the CO dehydrogenase, or if the protein, F,, may function in one of these roles (20). This protein is required in acetate synthesis, but is not required or even stimulatory for the exchange reaction. Thus CO dehydrogenase plays four roles in acetate biosynthesis: (i) the oxidation of CO to COz with generation of required reducing capacity, (ii) the reduction of COz to CO, (iii) the formation of a C1 intermediate from CO, and (iv) the condensation of this C1 intermediate with the methyl group and the CoA group to form acetyl-coA. These roles for CO dehydrogenase are much greater than had been previously imagined. Our results directly contradict the recent hypothesis of Diekert et al. (21), based on whole cell studies that the CO dehydrogenase plays only the first and second but not the third and fourth roles.
Surprisingly, low potential electron carriers, like ferredoxin and flavodoxin, are stimulatory to the exchange reaction. It seemed likely that the electron carrier could play one of two roles: (i) if CO, was the actual form of C-1 that exchanged with the C-1 of acetyl-coA, then the electron carriers could stimulate the exchange by oxidizing the CO to the more active form of C-1, COz, or (ii) that the electron carrier could act by stimulating an internal electron transfer that occurred in the exchange reaction sequence. Since CO was more active in the exchange than COz, the stimulation by low potential electron carriers may occur by aiding in an internal electron transfer that occurs during the exchange reaction. As shown in Table  I, the ratio of the rates of the CO dehydrogenase (at pH 7.6) and exchange reactions (at pH 5.31, is approximately 2200. However, when the activity of CO oxidation to C02 is measured at 50 "C at pH 5.3 with l PM ferredoxin I1 as electron carrier, the turnover number is approximately 780 mol/mol of CO dehydrogenase, which is 1/200 the rate with 100 mM methyl viologen at pH 7.6 (see Refs. 9 and 22). The exchange reaction, under these conditions, has a turnover number of 70 mol of CO exchanged per min/mol of enzyme. Thus under the physiological conditions of growth in fairly acidic media, the two rates are more similar. Since the binding and the two electron oxidation of CO and release of COz is rapid in comparison to the exchange reaction, the rate-limiting step of the exchange is not the Y-CO e CO + Y reaction of Equation 2. We have been unable to detect other factors in the ammonium sulfate 0-40%, 40-60% saturation fractions, or the 60% supernatant which stimulate the CO dehydrogenase exchange reaction in an assay mixture containing ferredoxin. It should be noted that rubredoxin is an excellent carrier for the low potential CO/CO, reaction but does not stimulate the exchange reaction.
It is interesting to compare this model (Fig. 5 ) with the proposed mechanism for industrial synthesis of acetate from methanol and CO (23, 33). Methyl iodide is an intermediate in this conversion. In this process, a rhodium catalyst is proposed to bind the methyl and iodide groups of methyl iodide and carbon monoxide. Then, by an insertion reaction, acetylrhodium is formed as an intermediate that is cleaved to acetyl iodide and then hydrolyzed to acetate. The rhodium undergoes a +1/+3 redox interconversion during the catalysis. Thus the Monsanto process for acetate formation may have an analogue in nature and it is tempting to propose an acetylnickel intermediate in the CO dehydrogenase reaction. The model in Fig. 5 makes several readily testable predictions: that labeled CoA should exchange with unlabeled acetyl-coA and that the methyl group of acetyl-coA should exchange with the methylated corrinoid protein. These possibilities are under current investigation.