reductive activation of the coenzyme a/acetyl-coa isotopic

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1991 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 266, No. 6, Issue of February 25, pp. 3554-3564,1991 Printed in U. S A

Reductive Activation of the Coenzyme A/Acetyl-CoA Isotopic Exchange Reaction Catalyzed by Carbon Monoxide Dehydrogenase from Clostridium thermoaceticum and Its Inhibition by Nitrous Oxide and Carbon Monoxide*

(Received for publication, June 25, 1990)

Wei-Ping Lu and Stephen W. Ragsdale From the Department of Chemistry, University of Wisconsin, Milwaukee, Wisconsin 53201

The final steps in the synthesis of acetyl-coA by CO dehydrogenase (CODH) have been studied by following the exchange reaction between CoA and the CoA moiety of acetyl-coA. This reaction had been studied earlier (Pezacka, E., and Wood, H. G . (1986) J. Biol. Chem. 261, 1609-1615 and Ramer, W. E., Raybuck, S. A., Orme-Johnson, W. H., and Walsh, C. T. (1989) Biochemistry 28, 4675-4680). The CoA/acetyl-CoA exchange activity was determined at various controlled redox potentials and was found to be activated by a one-electron reduction with half-maximum activity occurring at -486 mV. There is -2000-fold stimulation of the exchange by performing the reaction at -575 mV relative to the rate at -80 mV. Binding of CoA to CODH is not sensitive to the redox potential; therefore, the reductive activation affects some step other than association/dissociation of CoA. We propose that a metal center on CODH with a midpoint reduction potential of 5-486 mV is activated by a one-electron reduction to cleave the carbonyl-sulfur bond and/or bind the acetyl group of acetyl-coA. Based on a comparison of the redox dependence of this reaction with that for methylation of CODH (Lu, W-P., Harder, S. R., and Ragsdale, S. W. (1990) J. Biol. Chem. 265, 3124-3133) and COz reduction and formation of the Ni-Fe-C EPR signal (Lindahl, P. A., Munck, E., and Ragsdale, S. W. (1990) J. Biol. Chem. 265, 3873- 3879), we propose that the assembly of the acetyl group of acetyl-coA, i.e. binding the methyl group of the methylated corrinoid/iron-sulfur protein, binding CO, and methyl migration to form the acetyl-CODH intermediate, occur at the novel Ni-Fe3-4-containing site in CODH. CO has two effects on the CoA/acetyl-CoA exchange: it activates the reaction due to its reductive capacity and it acts as a noncompetitive inhibitor. We also discovered that the CoA/acetyl-CoA exchange was inhibited by nitrous oxide via an oxidative mechanism. In the presence of a low-potential electron donor, CODH becomes a nitrous oxide reductase which catalytically converts NzO to Nz. This study combined with earlier results (Lu, W-P., Harder, S. R., and Ragsdale, S. W. (1990) J. Biol. Chem. 265, 3124-3133) estab- lishes that the two-subunit form of CODH is completely active in all reactions known to be catalyzed by CODH.

* This work was supported by Department of Energy Grant DE- FG02-88ER13875 (to S. W. R.) and a Shaw Scholars Award (to S. W. R.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Carbon monoxide dehydrogenase (CODH)’ or acetyl-coA synthase from Clostridium thermoaceticum catalyzes the synthesis of acetyl-coA from the methylated corrinoid/iron-sulfur protein (C/Fe-SP), CO, and CoA (Equation 1) and thus is the key enzyme involved in the autotrophic acetyl-coA pathway of COz or CO fixation, also called the Wood pathway (Fig. 1).

CHS-[C/Fe-SP] + CO + CoASH (1)

+ CH3-CO-SCoA + [C/Fe-SP]

This pathway, which occurs under anaerobic conditions, plays a major role in the global carbon cycle and has recently been reviewed (1-4). The Wood pathway involves the reduction of CO, to formate by formate dehydrogenase followed by a series of reactions catalyzed by tetrahydrofolate (H4folate)-dependent enzymes which reduce formate to methyl-H4folate. A methyltransferase then catalyzes the transfer of the methyl group of methyl-H4folate to the C/Fe-SP forming a methyl- cobamide intermediate. The final steps in the synthesis of acetyl-coA are catalyzed by CODH. For simplicity, Fig. 1 depicts CODH binding the methyl group and then CO in an ordered mechanism. In fact, it appears that CODH can bind the methyl and CO groups randomly. This is based on the observations that 1) methylated CODH can react with CO ( 5 ) and carbonylated CODH can react with the methylated C/ Fe-SP2 to form acetate, 2) methylation of CODH occurs in the absence of CO ( 5 ) , and 3) carbonylation of CODH occurs in the absence of the methylated C/Fe-SP (6, 7). Following the scheme of Fig. 1, CODH accepts the methyl group from the methylated C/Fe-SP, forming a methyl-CODH intermediate in a reaction which is accelerated at low redox potentials (5). Next, methylated CODH binds CO apparently at a metal center which contains nickel and three to four irons. In this novel metal center, the iron sites have magnetic properties resembling those of a [4Fe-4S] center, the nickel site is bridged to the iron components by a ligand, and the carbon of CO is bound to either the nickel or iron sites (6, 8-10). Methyl migration to form acetyl-CODH is proposed to be the next step in the synthesis ( 5 , 11, 12). Then acetyl-CODH binds CoA at a site which involves tryptophan and arginine residues (13-15) and apparently is near the Ni-Fe-C center (6, 15). Thus, binding the methyl, the carbonyl, and CoA groups results in assembly of the three moieties involved in the

The abbreviations used are: CODH, CO dehydrogenase; C/Fe- SP, the corrinoid/iron-sulfur protein; MV, methylviologen; DTT, dithiothreitol; TRIQUAT, N,N‘-trimethylene-2,2‘-dipyridinium bro- mide; H,folate, tetrahydrofolate; HPLC, high performance liquid chromatography; SDS, sodium dodecyl sulfate.

S. W. Ragsdale, P. A. Lindahl, and E. Munck, unpublished results.

3554

3556 CoAIAcetyl-CoA Exchange of CODH Ag wires 77 conneeled Io a gas line

,rubber 0 ring

AgIAgCI reference and counter electrodes

(C)

FIG. 2. Electrochemical cell used for controlled potential enzymology. See text for details. Side (a ) and top ( b ) view of the bottom half. Side (c) and top ( d ) view of the top half.

[3'-"P]CoA to start the reaction. For reactions performed at controlled potentials, the assay mixture (excluding CoA) was incubated in the electrochemical cell as described above and, in addition to MV, two more low potential redox mediators, 4,4'-dimethyl-TRIQUAT (30 pg) and TRIQUAT (30 pg), were included. That the redox dyes had no effect on the CoA/acetyl-CoA exchange was shown by comparing the rates of reactions performed under CO in the presence and absence of the three dyes. The temperature was maintained at 40 "C by wrapping Tygon tubing which was connected to a temperature- controlled water circulator around the bottom of the lower half of the cell. The analytical methods for application of a redox potential have been described in detail (21) and are outlined here. A potentiostat (model CV-lA, Bioanalytical Systems Inc.) was used to apply the potential, which was monitored with a voltmeter (Beckman, model 4410) across the working and reference electrodes. After the redox potential of the reaction mixture reached equilibrium, which required -10 min at potentials of around -520 mV, [3'-32P]C~A (in 5 pl) was added to initiate the reaction. Aliquots (50 pl) of the reaction mixture were removed either from the electrochemical cell under a high positive nitrogen gas flow or from the anaerobic vial a t desired time intervals and frozen immediately in liquid nitrogen. Less than a 10 mV drift in the redox potential of the reaction mixture was observed during removal of aliquots from the cell. The frozen samples were quickly melted and chromatographed on HPLC to isolate CoA and acetyl-coA on the same day that the reaction was performed as described before (5). Thawing and addition of the sample to the column required -1.5 min. Based on our observation of the rate of reaction at 20 "C under oxygen, less than 1% additional exchange could have occurred, and the values obtained are uncorrected for this. Fractions containing CoA or acetyl-coA were collected and analyzed for radioactivity, and the percentage of exchange was calculated by dividing the radioactivity in the acetyl-coA fraction by the total radioactivity found in the CoA and acetyl-coA fractions. The velocity ( u ) of the exchange reaction was calculated from Equation 5,

v = - [CoA][acetyl-CoA] 2.3 [CoA] + [acetyl-CoA] t

- log(1 - F ) (5)

where F is the fraction of isotopic equilibrium attained at time, t (23). At the concentrations of acetyl-coA and [3'-3z9P]CoA employed (2 and 0.2 mM, respectively), -95% of the radioactivity was recovered in acetyl-coA a t isotopic equilibrium which was reached after 2-8 min. The specific exchange activity was calculated from the initial velocity portion of the reaction curve. The exchange rate was linearly dependent on the CODH concentration. Turnover numbers are calculated based on the a/3 form of CODH (M, = 148,000).

Effect of CO on the CoAIAcetyl-CoA Exchange-The CoA/acetyl- CoA exchange reactions were performed a t various concentrations of CO in the electrochemical cell a t redox potentials of -520 (?lo) mV. The concentration of CO was varied by modulating the flow rate of

CO in cylinders containing 100 or 1.0% CO (Union Carbide Co., Linde Division, Danbury, CT) relative to that of N2 in a separate cylinder. The actual concentration of CO in solution was measured by a modification of the hemoglobin method (24). Buffers saturated with a gas phase containing 100,15, and 1% CO in N, gas were found to contain 850, 75, and 8 p M CO, respectively. At 850 p ~ , CO alone was sufficient to drive the potential of the reaction mixture to -520 mV within 6 min. At lower CO concentrations, the desired potentials were reached and maintained by flushing with the CO and nitrogen gas mixture while simultaneously poising the redox potential electrochemically, since it required from 20 min to 1 h to drive the potential of the reaction mixture below -500 mV with the gas mixture alone. The potential of the reaction mixture was monitored throughout the experiment.

CoA Binding to CODH-The binding of CoA to CODH was studied in the anaerobic chamber a t -25 "C. CODH (3-8 nmol) was incubated with [3'-32P]CoA (10 to 50 nmol, -1000 cpm/nmol) in 50 mM Tris- HCI, pH 7.0, in a final volume of 50-100 pl for about 15 min, and the reaction mixture was applied to a 1-ml Sephadex G-50 Penefsky column (25). The protein was rapidly separated from free [3'-32P] CoA by centrifugation, and the radioactivity and protein concentration of the eluate were measured to yield the ratio of bound CoA to CODH. Typically, a ratio of 0.54 was obtained when 10 nmol of CODH and 200 nmol of [3'-32P]CoA were incubated in a final volume of 100 p1 for 15 min. The ratio decreased to 0.4 after the second column treatment and to 0.28 after the third. Controls were performed in which the C/Fe-SP or lysozyme were incubated with CoA to rule out passage of free CoA through the spun column and nonspecific binding of CoA by proteins. The values of [CoA]bound, [CoA],,., and [CODH],, were calculated from the radioactivity measurements after the first Penefsky column and the concentrations of CODH and CoA added, and the data were plotted according to the Scatchard equation

Reactions Involving Nitrous Oxide and CODH-Oxidation of reduced CODH by N20 was performed a t room temperature (25 "C) in an anaerobic cuvette containing CODH (0.1 mg), MV (0.25 pmol), DTT (1 pmol), KPi (50 p ~ ) , pH 6.9, and KC1 (200 pmol) in a total volume of 1 ml. After bubbling with CO for 1-2 min to reduce the MV, CO was removed from the reaction mixture with five vacuum/ nitrogen cycles followed by bubbling with N2 for 1-2 min. The reaction mixture in the cuvette remained anaerobic during the CO removal as no oxidation of reduced MV was observed, i.e. the absorbance a t 604 nm remained stable. Alternatively, sodium dithionite (200 nmol) was used as reductant, instead of CO, to reduce MV. An aliquot (50 pl) of NzO-saturated 50 mM KPi buffer, pH 6.9, was added into the cuvette, and the initial decrease in absorbance a t 604 nm was recorded. For determination of the stoichiometry of the oxidation of reduced MV by N20, 1-5 p1 of an N20-saturated buffer were added and the absorbance at 604 nm was monitored. The mole ratio of MV oxidized/ N20 added was calculated from the decrease in absorbance at 604 nm due to oxidation of MV (Ac = 13.9 mM" cm") and the solubility of NzO (19.7 mM) a t 25 "C (26).

Identification of dinitrogen as the product of the reaction of NzO with CODH was accomplished by mass spectroscopy. The reaction was performed at 45 "C in the electrochemical cell containing CODH (2 mg), MV (50 nmol), DTT (0.3 pmol), KC1 (60 pmol), and Tris- HC1 (12 pmol), pH 7.6, in a total volume of 300 pl. The gas phase was 100% N 2 0 in a 5.5-ml volume. A redox potential of -500 mV was applied to the reaction mixture for -1 h however the measured potential only reached "320 mV. The reaction cell was connected to the mass spectrometer via a U-shaped glass tube which could be inserted into liquid nitrogen (and serve as a trap) and connected through a two-way valve to to a vacuum line. Initially the electrochemical cell was isolated from the trap while the connecter and spectrophotometer were evacuated for a few minutes. The trap was then isolated from the mass spectrometer. Sampling of the reaction in the electrochemical cell was performed by quickly opening and closing the outlet valve of the cell. This allowed the gas phase in the cell to enter the U tube trap. Since the boiling points for NZ0 and NP are -88 and -186 "C (26), respectively, nearly all of the NzO was trapped in the liquid form in the trap, whereas N2 remained in the gas phase and was introduced into the mass spectrometer (RMU-6E spectrometer, Hitachi/Perkin-Elmer). The ionization potential was set at 70 eV, and the gas inlet system was maintained at 200 "C. The mass range was scanned from 10 to 100 atomic mass units. To obtain the relative conversion of N 2 0 to Nz, the gas mixture left in the cell was introduced a second time into the trap after the electrochemical cell and the trap had been evacuated at room temperature. Then, the

(23).

CoAIAcetyl-CoA Exchange of CODH 3557

line from the electrochemical cell to the spectrometer was again opened and the gas mixture was analyzed by mass spectrometry. The second sampling represented the amount of the present in the liquid phase of the reaction mixture, since the gas phase had been removed from the cell during the first sampling and evacuation. The percentage of N,O converted to N, was calculated based on the solubilities of N,O (19.7 mM) and N2 (0.5 mM) in water at 30 "C (26) and the relative peak heights of N,O and NZ after correcting for the Nz formed as a result of N,O fragmentation during ionization in the spectrometer.

Reactions Involving the C/Fe-SP and Nitrous Oxide-The reaction between reduced MV and the C/Fe-SP was performed as described above except that 0.25 mg of the C/Fe-SP, instead of CODH, was used, and dithionite was used as reductant. Reactions were also performed at low redox potentials in the electrochemical cell containing C/Fe-SP (0.5 mg), MV (0.15 pmol), DTT (0.3 pmol), KC1 (60 pmol) and Tris-HC1 (15 pmol), pH 7.6, in a total volume of 0.3 ml equilibrated with 100% NzO. The potentiostat was set at -510 mV. UV-visible absorption spectroscopy of the reaction of the C/Fe-SP with NzO was performed in a UV-visible spectroelectrochemical cell. The spectroelectrochemical cell is of the same design as the electrochemical cell described above except that the lower part contains a fused quartz cuvette with a strip of gold foil as the working electrode. The reaction mixture contained C/Fe-SP (3 mg), DTT (1 pmol), KC1 (200 pmol), TRIQUAT (5 pg), and Tris-HC1 (50 pmol), pH 7.6, in a 1-ml volume. The solution was poised at -550 mV for -20 min to fully reduce the C/Fe-SP to the Co'+ form, and the spectrum was recorded. Then, 20 pl of NzO-saturated buffer (50 mM Tris-HCI, pH 7.6) was added, and the absorbance at 390 nm was monitored. The experiments were performed a t room temperature.

The effect of N,O on the activity of the C/Fe-SP in the total synthesis of acetyl-coA from CH3-Hafolate was examined. The reaction mixture (0.4 ml) contained C/Fe-SP (1 mg), MV (4 nmol), TRIQUAT (50 pg), KC1 (20 pmol), and Tris maleate, pH 7.3 (8 pmol). After poising the redox potential of the mixture at -550 mV at room temperature for 15 min under argon, the gas phase was changed to 100% N,O, and the potential of the reaction mixture was poised at -520 mV. At various times, aliquots (50 pl) were removed anaerobi- cally and transferred into the anaerobic chamber to concentrate and replace the buffer with 10 mM Tris maleate, pH 6.8, with an Amicon Minicon concentrator. This solution of N20-reacted C/Fe-SP was assayed for its activity in acetyl-coA synthesis. An aliquot (15-20 pg) of C/Fe-SP was added to a reaction mixture (24 pl) in a semm- stoppered micro-V vial (Pierce Chemical Co.) containing CODH (5 pg), methyltransferase (1.5 pg), CoA (13 nmol), and Tris maleate (0.24 wmol), pH 6.8. After incubation for 5 min under a CO gas phase at 55 "C, 1 p1 of 14CH3-H,folate (8.5 nmol, 17,000 dpm/nmol) was added to start the reaction. At various times, 5 pl of the mixture was removed, frozen immediately in liquid N,, and analyzed by HPLC as described before (5) except that 11% methanol was used in the eluting solvent and the flow rate was 1.8 ml/min. Under these conditions, acetyl-coA eluted a t 12 min, CH3-H,folate at 8 min, and CoA and Hlfolate a t 4.5 min. Fractions were collected and analyzed by liquid scintillation counting. The activity of the C/Fe-SP in the synthesis was 45 nmol min" mg" C/Fe-SP (0.15 s-').

Analytical and UV-visible Spectroscopic Techniques-Protein concentration was determined by the Rose-Bengal dye-binding assay (27) with lysozyme as standard. SDS-polyacrylamide gel electrophoresis was performed as described (28). UV-visible absorption spectra were recorded, stored, and manipulated using the PECUV software with a Perkin-Elmer Lambda 4C spectrophotometer. After preliminary es- timates by Lineweaver-Burk analyses, kinetic parameters with their standard deviations were obtained by an objective analysis of the data with a computer program based on the "Curfit" program of Bevington (56). This program is a nonlinear least squares fit to the appropriate kinetic equation. A modification of this program also was used to measure the binding constants of CODH for CoA.

RESULTS

Effect of CO and Redox Potential on the CoAIAcetyl-CoA Exchange-Under experimental conditions that are essen- tially the same as those described earlier (18), i e . pH 6.0 and 45 "C with 100% CO in the gas phase, exchange of the CoA moiety of acetyl-coA with free CoA was observed at a rate of 0.8 pmol min-' mg" CODH (2.0 s-'), which is comparable with the specific activity reported by Ramer et al., 2.5 pmol

min" mg" CODH (6.2 s-') (18). The redox potential of the reaction mixture in the presence of CO was "500 mV at pH 6. When CO was omitted and the gas phase was replaced with Nz, the measured redox potential in the absence of an applied potential was "80 mV. Under these conditions, -3% exchange was detected after 10 min incubation with 70 pg CODH, giving an activity of 8 nmol min" mg" (0.02 s-'). Thus, CO appeared to markedly stimulate the CoA/acetyl- CoA exchange. Since the requirement for CO in this exchange reaction was not obvious, we considered that it could function to activate CODH by reduction.

In order to determine the effect of redox potential on the rate of the CoA/acetyl-CoA exchange, reactions were performed at pH 7 and 40 "C at electrochemically poised potentials from -370 to -575 mV. The rate of the exchange reaction was extremely slow at potentials higher than "400 mV and markedly increased as the redox potential decreased (Fig. 3). At -400 mV, the specific activity was determined to be 0.08 pmol CoA exchanged min" mg" (0.2 s-') in comparison to 15.8 pmol min" mg" (39.5 s-') a t -575 mV. Thus, there is -2000-fold stimulation of the CoA/acetyl-CoA exchange reaction when the redox potential is lowered from -80 to -575 mV. We determined that the rate of acetyl-coA hydrolysis under these conditions is over 500-fold slower than the CoA/ acetyl-coA exchange and, thus, does not significantly affect our results.

The increased activity at lower potentials is due to a reversible process since the activity decreased in a predictable manner when the potential was raised. In addition, no matter what potential a reaction mixture had been previously poised, the exchange activity obtained was always dependent on the poised potential during the current reaction. The rate of exchange as a function of the poised potential exhibited Nernstian behavior, and the data were analyzed according to the Nernst equation (Fig. 3, inset). In this treatment, we assume that the highest activity is due to fully reduced and fully activated CODH, and the lower activities at higher redox potentials are due to fractionally active protein. Thus, the specific activity measured at -575 mV is termed the fully (100%) active form of the enzyme in analogy to the fully reduced form of a redox center in a redox titration. A value

TF 1 .c 10 E

5 - 1

0 ; m Y

-600 -0.70 -0.20 0.30 0.80 1.30

Log [active froction/(l-active froction)]

0

0

0

22 0 -600 -500

Redox Potential (rnV) -400

FIG. 3. Effect of redox potential on the CoA/acetyl-CoA exchange catalyzed by CODH. The redox potential of the reaction was controlled electrochemically under nitrogen gas in an electrochemical cell as described under "Experimental Procedures." The reaction was performed a t pH 7.0 and 40 'C, and 17.5 or 35 pg CODH was used. Inset, Nernst plot of the data. The specific exchange activity measured a t -575 mV is defined as the fully (100%) active form of the enzyme. The slope is 65 _t 3 mV and the y intercept is -486 f 6 mV.

3558 CoAIAcetyl-CoA Exchange of CODH

of the active fraction at a particular redox potential was obtained by dividing the value of the specific activity at that potential by the value of the specific activity measured at -575 mV. When the log of the ratios of activelinactive protein; i.e. (active fraction)/(l - active fraction), is plotted against the poised potential (Fig. 3, inset), the data nicely fit the Nernst equation. The slope of the plot, 65 f 3 mV, indicates that an one-electron reduction has occurred (a slope of 62 mV corresponds to a one-electron transfer at 40 "C) and the y intercept reveals that the potential at which half- maximal activity is observed is -486 & 6 mV.

Effect of p H or DTT on the CoAIAcetyl-CoA Exchunge- The optimum pH for the exchange reaction in the presence of CO was between 6.7 and 7 (Fig. 4). Tentative pK values for the functional groups ionized during the exchange reaction were calculated to be 6.7 and 7.3 assuming a rapid equilibrium diprotic system according to the method outlined by Dixon (59). Most of the experiments were performed at pH 7, since, due to the hydrogen overpotential, low redox potentials are easier to maintain with high uersus low pH values. Assays run at pH 7.0 in either KPi or Tris-HC1 showed the same exchange activity. Most of our reactions were performed in the presence of DTT, since CODH requires strictly anaerobic conditions for stability. Omission of DTT had no effect on the exchange velocity.

Inhibition of the CoAIAcetyl-CoA Exchange by CO-The rate of the exchange reaction performed in the presence of CO was lower than those of identical reactions at the same potential in which the CO had been replaced with N2 or Ar. Exchange reactions were carried out at -520 (f10) mV at concentrations of acetyl-coA from 0.5 to 3.5 mM at various fixed concentrations of CO (from 0 to 850 p M in solution). A double-reciprocal plot of the results (Fig. 5a) showed that the maximum velocities decreased as the CO concentration increased and the lines converged at the x axis. Objective analysis of the data yielded a K, value for acetyl-coA of 1.80

0.08 mM which was not affected by the concentration of CO. These results indicate that CO is a noncompetitive inhibitor of the CoA/acetyl-CoA exchange with respect to acetyl- CoA. Comparing the specific activity of 7 f 1 pmol min" mg" (17.5 s-') measured in the presence of 100% CO in the gas phase with the value of 28 f 3 pmol min" mg" (70 s-') obtained in the absence of CO, CO caused -75% inhibition of this exchange reaction. The Ki for CO was calculated to be 400 f 30 PM (Fig. 5b).

Kinetic Parameters for the CoAIAcetyl-CoA Exchange Re- action Measured at Low Redox Potential-In order to compare

c

h

.- E

' 1 Y O

0 0

0

0

0

0 1 I I

5.5 6.5 7.5 8.5 PH

FIG. 4. Effect of pH on the CoA/acetyl-CoA exchange. The assay was done as described under "Experimental Procedures" in the presence of 100% CO in the gas phase. KP, buffer (50 mM) was used at pH 7.0 and below and Tris-HC1 buffer (50 mM) at alkaline pH values.

' .OO 3" 1 /

400 -200 0 400 600 100

FIG. 5. Effect of CO on the CoA/acetyl-CoA exchange by CODH. a, the reaction was performed at pH 7.0, 40 "C, a controlled potential of -520 mV, and CO concentration of 0 (O), 8 (A), 75 (O), 400 (A), and 830 JIM (B). b, Dixon plot of the data in a with concentrations of acetyl-coA at 0.5 (O), 1 (01, and 2 (A) mM. For further details, see "Experimental Procedures."

kinetic parameters for the CoA/acetyl-CoA exchange with those obtained earlier for reactions performed in the presence of CO (18), CoA, and acetyl-coA were varied at a constant controlled redox potential of -520 (f10) mV at pH 7.0. Objective analysis of the data yielded K, values of 65 f 10 pM (data not shown) for CoA and 1.89 & 0.1 mM (Fig. 5a) for acetyl-coA which are comparable to those reported before, 50 p M and 1.5 mM for CoA and acetyl-coA, respectively (18). The specific activity of 28 pmol min" mg" (70 s-') (Fig. 5a) is over 10-fold higher than reported before (18). The activity of the enzyme is even higher at potentials lower than -520 mV. For example, based on Fig. 3, the enzyme is 26% more active at -575 than at -520 mV; therefore, at -575 mV, where the enzyme is 97% in the active (reduced) form based on the Nernst equation, the specific activity would be 35 pmol min" mg" (87.5 s-').

Binding of CoA by CODH-CoA binding to CODH was studied by a relatively rapid incubation of CODH with CoA followed by separation of free CoA from bound CoA with a Penefsky column. The observation that CoA coeluted with CODH from the Penefsky column and that 15-25% of the CoA was lost with each subsequent passage through the column indicates that CoA binds quite strongly to the protein but that covalent assocation is probably not involved. By varying the concentration of CoA in the incubation mixtures and plotting the data according to the Scatchard equation (23), the dissociation constants, Kd, and the ratio of bound

CoAIAcetyl-CoA Exchange of CODH 3559 0

1.0 - al E 0.9 - L - 10.8 - v) 5 0.7 - s 2 0 . 6 - c 0.5 - 3 2 0.4 - 10.3 - v, - 0 5 0.2 - Y

0.1 -

0 20 40 60 80 100 [CoASH] bound (4M)

FIG. 6. Scatchard plot of CoA binding to CODH. The binding experiment was performed as described under "Experimental Proce- dure." In this experiment 140 p~ CODH was used. A simulation ("-), based on the equation for a one ligand, two binding site model (57), consistent with the experimental points (O), was obtained using the following values: &(l) = 52 + 3 p M , n(1) = 0.47; &(2) = 2600 + 500 pM, n(2) = 0.40.

CoA to CODH was derived. The data were best fit using a two site model with approximately equal occupancy at each site (Fig. 6). The two Kd values for CoA was determined to be 52 and 2600 p~ with 47 and 40% occupancy in the high and low affinity sites, respectively. The value for the low affinity site is close to the K , values obtained from the enzymatic method (50 & 25 p ~ , Ref. 18 and 65 p ~ , above). The effect of the low redox potential on the binding of CoA to CODH also was tested. In this case, CoA and CODH were incubated a t a stable potential of -540 mV for -10 min and then separated by centrifugation through a Penefsky column. No significant difference in the ratio of CoA bound to CODH a t ambient potentials versus poised potentials was found, suggesting that the binding of CoA by CODH is not reductively a ~ t i v a t e d . ~

Effect of Various Inhibitors on the CoAIAcetyl-CoA Ex- change-N20 (10% in the gas phase) inhibited the CoA/ acetyl-coA exchange by -30% relative to the reaction performed at the same redox potential (-460 mV, Table I). However, compared to reactions performed a t potentials below -520 mV, the inhibition by N 2 0 was greater than 70%. CO, inhibited the CoA/acetyl-CoA exchange by -50% relative to a control reaction performed at the same potential (Table I). With either N20 (10% NzO, 90% N z ) or CO, (100%) in the gas phase, we could not obtain a stable (or equilibrated) redox potential lower than -460 mV. When a constant voltage of -610 mV was applied for over 30 min, the lowest stable potential we could poise was at -460 mV. This behavior with CO, had been observed earlier and was ascribed to the ability of CO, to serve as an electron sink (8). When CODH was omitted from the reaction mixtures containing either N 2 0 or COP, low stable potentials were obtained within 5 min.

We also investigated the effect of dephospho-CoA and desulfo-CoA on the rate of the CoA/acetyl-CoA exchange at low redox potentials (Table I). Desulfo-CoA is -10-fold less effective than dephospho-CoA, indicating that the sulfur group is an important determinant in the binding of CoA to CODH. These results are consistent with the observation by Raybuck et al. (17) that the inhibition constants for dephospho-CoA and desulfo-CoA for the CO/acetyl-CoA exchange

That the reaction mixture is not oxidized during the isolation on the Penefsky column is indicated by the observation that the blue color of the reaction mixture, due to reduced MV, remains on the column after completion of the separation of CoA from CODH.

TABLE I Effect of various inhibitors on the CoAlacetyl-CoA

exchange by CODH Reactions were performed at a controlled redox potential of -520

(+lo) mV as described under "Experimental Procedure" except for the N,O and CO, reactions (*) in which the lowest potential obtain- able was -460 (+20) mV. The control values used for the comparison were 12 and 6 pmol min" (mg CODH)" for reactions at -520 and -460 mV, respectively.

Inhibitors Concentration Specific activity Inhibition pmol min" mg" %

100% gas phase 3 + 1 50* N,O* 10% gas phase 4 f l 30* co,* CN" CN" 0.25 mM 8 + 1 CN"

30

CN" 2 + 1 80

1.2 mM 0.2 f 0.1 98 Dephospho-CoA 0.22 mM 6 + 1 50 Dephospho-CoA 0.44 mM 2.9 ? 0.5 75 Desulfo-CoA 1.25 mM 11 * 1 5 Desulfo-CoA 2.1 mM 8 f 2 30

0.05 mM 12 f 2 0

0.6 mM

were 35 and 6000 pM, respectively. The CoA/acetyl-CoA exchange reaction was much less sen-

sitive to inhibition by cyanide than the CO oxidation reaction, the CoA exchange reaction requiring 0.2 mM cyanide for 23% inhibition (Table I), whereas the formation of COz was completely inhibited at this concentration (29). CO oxidation and CO, reduction were found earlier to be inhibited competitively by cyanide with respect to CO (29, 30). Inhibition of the CO/ acetyl-coA exchange reaction by cyanide is similar to that of the CoA/acetyl-CoA exchange, requiring 0.25 mM cyanide for 26% inhibition (data not shown). Significant variation in the inhibition of CoA/acetyl-CoA exchange occurred at cyanide concentrations above 0.5 mM. In the presence of 1.2 mM cyanide, the total amount of CoA and acetyl-coA decreased to less than one-third of that seen in the absence of cyanide as judged by the radioactivity collected in the CoA and acetyl- CoA peaks, although the retention time from the HPLC remained the same. Due to this problem, we were unable to further study the mechanism of cyanide inhibition on the CoA/acetyl-CoA exchange but postulate that a side reaction(s) occurred a t high concentrations of cyanide. Based on the different sensitivies of cyanide inhibition for the CoA/ acetyl-coA and CO/acetyl-CoA exchange reactions versus the CO oxidation reaction and the competitive nature of the inhibition in the CO oxidation reaction, there could be two sites on CODH for CO binding, one involved in CO oxidation and another in acetyl-coA synthesis. However, since CO is not only a ligand but a reductant as well, another possibility is that the highest affinity cyanide binding site is not the CO binding site, but a site involved in electron transfer reactions.

Oxidation of Reduced CODH by Nitrous Oxide-Based on our inability to maintain a low redox potential in the presence of CODH and either CO, or N20, we postulated that N20 could act as an oxidant of CODH in a manner similar to CO,. When MV, reduced by either CO or dithionite, was incubated with CODH in the presence of N20 (0.62 mM), it was rapidly oxidized a t a rate of 0.4 pmol MV oxidized min" mg" (1.0 s-'). This value is -6-fold lower than that of CO oxidation measured under the same conditions. We should note that the CO oxidation and probably also the NzO reductase reaction are measured at suboptimal concentrations of MV and pH. The K , value for MV is 3.0 mM, and the optimal pH is -8.4 for the CO oxidation reaction (29). The reaction mixture could be repeatedly reduced and oxidized by addition of CO and NzO, respectively, without significant loss of either NzO reduction or CO oxidation activities, indicating that CODH


is not irreversibly inactivated by NzO. The stoichiometry of the oxidation was found to be 1.8 f 0.1 mol of MV oxidized per mol of N20 added, demonstrating that a two-electron transfer is involved in the reaction since MV is a one-electron carrier. This suggested that Nz was the product of the NzO reduction. Mass spectroscopic analysis of the gas mixture at the end of the reaction showed that a significant amount of NP was formed. After a 1-h incubation of CODH with NzO and reduced MV at an applied potential of -500 mV, approximately half of the NzO was found to be converted to dinitrogen, from which a specific activity of -1 pmol min" mg" protein (2.5 s-') was obtained.

Reaction of the C/Fe-SP with N20-Since cob(1)amide is known to react with N 2 0 to generate NZ (31,32), we examined the reactivity of the C/Fe-SP with N20. The C/Fe-SP was incubated with NzO and reduced MV under conditions identical to those described for CODH. Under these conditions, no NzO-dependent oxidation of reduced MV was detected. In addition, a stable and well equlibrated low potential (-480 mV) was established in the reaction system after being poised at -510 mV under 100% N 2 0 for just a few minutes, showing that the C/Fe-SP, in contrast to CODH, does not catalytically reduce NzO. Thus, NzO does not act as an oxidant in this system. It remained possible that the CO'+ form of the C/Fe- SP could react with N 2 0 and generate 1 mol of N2/mol of Co'+ and then become irreversibly inactivated. This possibility is ruled out by the results of another experiment. After generation in a spectroelectrochemical cuvette, the Co'+ form of the C/Fe-SP remained after the addition of N20 (0.25 mM) as seen by the stability of the 390-nm band, which is charac- teristic of the Co'+ form of cobamides (6). Furthermore, when we incubated the C/Fe-SP under an atmosphere of 100% NzO at -520 mV for 1 h, there was no significant decrease in activity in the synthesis of acetyl-coA from CH3-H4folate, CO, and CoA. Thus, the C/Fe-SP apparently does not react with N 2 0 .

DISCUSSION

Dependence of the Exchange Between CoA and Acetyl-coA on the Redox Potential-Considering steps 6 and 7 (Fig. 1) as reversible reactions, it becomes apparent that one can study the synthesis of acetyl-coA from the bound acetyl-CODH intermediate and free CoA by following the exchange reaction between CoA and the CoA moiety of acetyl-coA. We found that this exchange reaction is reductively activated which indicates that reduction of a site on CODH is required before it can catalyze the the CoA/acetyl-CoA exchange.' Since

One could consider that this one-electron reduction would be required either for each turnover of substrate to product (i.e. a redox reaction) or for reductive activation of the enzyme which has a certain activity determined by the redox potential which then catalyzes a nonredox reaction. As one reduces the oxidized system by poising the

disequilibrium until all the oxidized and reduced species equilibrate redox potential of the reaction mixture, the system is in a state of

at a value specified by that poised potential. Thus, when the potentiostat is turned off, the monitored potential drifts from the value poised at the electrode to the value actually in the solution. If the turnover event itself is a redox reaction requiring reduction (such as catalysis of CO, (8) or NzO (this work) reduction), then the redox potential will not stabilize until the substrates and products of the reaction reach equilibrium. When CODH is incubated with N20, for example, a reaction which occurs a t a rate 50-fold lower than the COA/acetyl-CoA exchange reaction, it is impossible to poise the reaction mixture below -450 mV as there are dramatic drops in the monitored potential. In the CoA/acetyl-CoA reactions, the protein reaches redox equilibrium with the dyes in solution before substrate is added. Then with no additional input of electrons ( i e . the potentiostat is switched off and we monitor the reaction with the voltmeter)

activation occurs only a t very low potentials, we assume that it involves a metal center.

Controlled potential enzymology has been defined as the study of the dependence of the rate of an enzymatic reaction on the redox potential at which the reaction is poised (5). By relating the redox potential of the different metal centers in CODH to the redox dependence of reactions involved in acetyl-coA synthesis, we can gain information about the active site metal center(s) involved in a particular reaction. CODH (per ap subunit) contains two nickel and -12 iron atoms (29) which are associated in redox active metal clusters (7-9). Three metal complexes in CODH have midpoint potentials which are in the same range as the redox transformation required for the CoA/acetyl-CoA exchange: the Ni-Fe-C species, a [4Fe-4S] center and a metal center of unknown com- position and EPR parameters similar to those of hemerythrin (8, 9). The midpoint potential of the Ni-Fe-C center was found to be 5-520 mV.5

Study of reactions in which an electron transfer step is followed by a chemical step(s) or a binding step is not straight- forward unless the thermodynamics and kinetics of both the redox and the subsequent steps are well characterized (53). If one considers only a binding event coupled to the redox reaction, the Nernst equation becomes more complex since it includes dissociation constants of the enzyme for the ligand. If the reduced form of a redox center has higher affinity for a ligand relative to the oxidized species, the apparent midpoint potential of the redox reaction is shifted toward a positive value since the effect of the coupled reaction is to lower the concentration of the reduced species relative to the inactive oxidized species. The opposite effect will be observed if the oxidized form of the enzyme binds the ligand. An example of substrate-induced alteration of the reduction potential is with the general acyl-CoA dehydrogenase where binding of acyl- CoA substrates increases the midpoint potential of the flavin by over 100 mV (54). The oxidized form of the iron protein of nitrogenase has a two order of magnitude greater affinity for Mg-ATP than the reduced enzyme leading to a shift in the apparent reduction potential by 130 mV in the negative direction upon binding Mg-ATP (55). If additional chemistry after substrate binding is involved, then the situation becomes even more complicated. However, qualitatively, the same prin- ciple holds. If the reduced species is involved in the coupled chemical reaction, then the apparent midpoint potential will be shifted in the positive direction. The magnitude of the shift depends on the how exergonic the coupled reaction is and how rapidly the coupled reaction occurs relative to the electron transfer reaction. Correspondingly, if one measures the dependence of the rate of a reductively activated reaction on the redox potential, the potential at which half-maximal activity occurs (the midpoint turnover potential) will be more positive than the actual midpoint potential for that active site redox center in the absence of the coupled reaction. Even if one cannot strictly determine the redox potential for the reductively activated center which is involved in catalysis and binding, it can be stated that the midpoint reduction potential for the reductively activated center must be 5 the midpoint

the redox potential remains constant during the reaction even though turnover is occuring a t rates of -60 s-*. Theoretically, the CoA/ acetyl-coA exchange should not require net oxidation-reduction chemistry, since it is an isotope exchange reaction and involves the freely reversible exchange of isotopically labeled substrates for unlabeled substrates. Thus, reductive activation of a site on CODH is the most likely explanation for our requirement for reductive activation and the requirement for reducing equivalents is to equilibrate CODH a t a low potential.

C. M. Gorst and S. W. Ragsdale, unpublished results.


turnover potential. In addition, as long as the coupled reaction does not also involve redox chemistry, the slope of the Nernst plot is not altered by the coupled reaction and indicates the number of electrons transferred during the oxidation-reduction process. When the CoA/acetyl-CoA exchange activity of CODH is related to the poised potential, the data fit the Nernst equation for a one-electron reduction of the system with midpoint turnover potential at -486 f 6 mV, indicating that the redox-active metal center has a midpoint potential of 5-486 mV.

One could argue that the rate enhancement observed as a result of degree of reduction could be due either to activation of a redox-active site which directly participates in the reaction or it could be due to an indirect effect such as reduction of a center on the enzyme which would then induce a confor- mational change in the protein to a state that is more active. At this point, we cannot rule out the possibility of an indirect effect; however, we favor the explanation that the reduction occurs at the actual site of catalysis or substrate binding. Our prejudice is based not only on the use of Occam's razor, but primarily on the properties of metal centers in enzymes which bind ligands and which are involved in formation of organo- metallic bonds. In several cases, redox changes in metal centers of enzymes are associated with enhanced catalysis or ligand binding. In other nickel-containing systems, the Ni2+ center of methyl-CoM reductase apparently requires reduction to the 1+ state to bind the methyl group of methyl-CoM (34) and a Ni/Fe-S hydrogenase requires reductive activation (35). Reductive activation of the cobalt centers of the C/Fe- SP (5, 21) and methionine synthase (see Ref. 36 for review) are required before these enzymes can be methylated. Reduc- tion of Fe3+ to the Fez+ state in the heme sites of several enzymes increases the affinity of these enzymes for their substrates (37-39). Both the copper (40) and the non-heme iron (41,42) containing phenylalanine hydroxylases are activated by one-electron reductions. In nonmetal redox systems, interconversion between the reduced dithiol and oxidized disulfide at specific sites in some enzymes has been shown to be important in their activity. Some examples include fruc- tose-1,6-bisphosphatase and NADP-malate dehydrogenase (43) and a valyl-tRNA synthetase (44). In addition, CODH from C. thermoaceticum also has been reported to be activated by reduction of disulfides by a disulfide reductase (11).

The CoA/acetyl-CoA exchange (Scheme 1) involves (i) redox chemistry, (ii) interaction of acetyl-coA, (iii) cleavage of the C-S bond forming an acetyl-CODH-CoA intermediate, (iv) release and rebinding of CoA, and (v) resynthesis of acetyl-coA. (i) The redox step apparently involves activation of a metal center on CODH. (ii) Interaction of acetyl-coA with CODH occurs through arginine and tryptophan residues

M

CODH e-

y/co(-486rrv %-C, ? FI Mred M

C H ~ C ~ A + ~ O D H + C ~ D H

Co iSH ,y

CH3-C

M

B CODH

e- FI y/E.C"6m" CH3-C CODH I

0 ye4 '?I /A c@SCoA'+ CODH === C e P H COA'SH

CoA'SH

SCHEME I

(13, 14), and since binding of CoA to CODH is unaffected by the redox potential, presumably acetyl-coA binding also is insensitive to the redox potential. (iii) Hence, either cleavage of the C-S bond of acetyl-coA or formation of the acetyl- CODH intermediate is assumed to be the redox sensitive step. Reaction of the reduced metal center with acetyl-coA is proposed to cleave the C-S bond and form an acetyl-metal intermediate. An acetyl-CODH intermediate has been proposed before based on the recovery of acetate as a minor product from reactions between methyl-H4folate, CO, and CoA (12) and between CH3-CODH, CO, and CoA (5). In addition, acetate is stoichiometrically formed in a single turnover reaction between CH3-CODH and CO in the absence of CoA (5). However, the acetyl-CODH intermediate is fairly stable since incubation of CODH with acetyl-coA during the exchange reaction at all redox potentials studied leads to only very low rates of acetate formation as had been reported before (18). In addition, the amount of acetyl-coA was not detectibly decreased during the CO/acetyl-CoA exchange reaction (12, 15, 17). (iv) In the CoA/acetyl-CoA exchange, unlabeled CoA is replaced by labeled CoA. Values which describe the strength of interaction of CoA with CODH are quite variable. Our data are best fit by a model which includes two CoA binding sites, with Kd values of 52 p M and 2.6 mM. The K d value of the low affinity site is close to the K, values (50-65 p ~ ) reported earlier (18) and here. Based on studies of tryptophan fluorescence quenching of CODH on binding CoA, a Kd of 0.1 p~ was calculated (13). CoA competitively inhibits the CO/acetyl-CoA exchange (15) with a Ki value of 7 p~ (17). K, values for CoA of 4.7 mM for the synthesis of acetyl-coA from CoA, CO, and either methyl iodide (5) or CH3-H,folate (58) have been measured. I t will require further experiments to explain the apparent existence of two CoA binding siteslap form of CODH. One possibility is that there are two forms of CODH, a high activity form which binds CoA tightly and a form which only weakly binds CoA. Alter- nately, a single form of CODH could have a high affinity catalytic CoA binding site and an additional site which could play a structural or regulatory role. Both explanations imply that under the conditions of our binding experiment, only approximately half of the CODH molecules in the a@ form bind CoA with high affinity. We have similar evidence for alternative forms of CODH by spectroscopic studies (8). The fractional occupancy could imply that the dimeric ~$3 form of CODH is not adequate for CoA binding. (v) The resynthesis of acetyl-coA then occurs which should regenerate reduced active CODH and labeled acetyl-coA.

Evaluation of the CoAIAcetyl-CoA Exchange Rates-The specific activity of the exchange between CoA and the CoA moiety of acetyl-coA a t 40 "C and at -520 mV and pH 7 is 28 pmol min" mg", yielding a turnover number of -70 s", based on an &3 subunit structure and a dimeric molecular weight of 150,000. Based on an activation energy of -26.8 kcal mol" (18), the rate expected at 55 "C is 7.2-fold higher than at 40 "C, yielding a value of 200 pmol min" mg" (500 s-'). This rate is -200-fold faster than the value (1.2 pmol min" mg", 3 s-') for the carbonyl exchange measured at 55 "C (17) and probably at similar redox potentials (since the reaction is performed under CO where the equilibrium potential is "520 mV). The rate of the CoAfacetyl-CoA exchange measured here is -10-fold higher than those reported recently by Ramer et al. (18), who measured their kinetic constants at suboptimal pH in the presence of CO, which we find is an inhibitor of the reaction. The rates reported here also are -14,000-fold higher than the values measured earlier by Pe- zacka and Wood (ll), presumbly due to the stimulatory effect


of the low redox potential.‘j The reducing systems used in their work were NADPH, reduced ferredoxin, or dithiothreitol, none of which would be capable of poising the potential of the solution below -400 mV, where a very low rate of exchange is observed.

Discussion of the Active Form of CODH-In the work reported here, the disulfide reductase is absent. This protein was postulated to stimulate the exchange between CoA and acetyl-coA and the synthesis of acetyl-coA by reducing key disulfide bonds in CODH (11). The disulfide reductase was proposed to be a subunit of CODH which dissociates during the purification of CODH and that the low activities reported by Pezacka and Wood were a result of partial reconstitution of the “three-subunit” form of CODH by the disulfide reductase (18). Ramer et al. (18) claimed that this dissociation was prevented in their purification procedure and thus their “three-subunit’’ CODH was more active. We considered the claims of a three subunit enzyme earlier (5) in studies of a number of reactions catalyzed by CODH. Earlier we had not studied the CoA/acetyl-CoA exchange which is important since a study of this reaction led to the “three-subunit” hypothesis. Here we have shown that, using conditions similar to those of Ramer et al. (18), we obtain similar rates of CoA/ acetyl-coA exchange with the two-subunit form of CODH. Other reactions which are catalyzed at high rates by the two- subunit enzyme include acetyl-coA synthesis from CO and CoA with methyl-CODH or methylated C/Fe-SP as methyl donor, methylation of CODH, and exchange reactions between methylated CODH and either acetyl-coA or the methylated C/Fe-SP (5). In addition, the COlacetyl-CoA exchange (15) and CO oxidation (29) are catalyzed at high rates by the purified CODH. Hence, all reactions known to be catalyzed by CODH have been performed with a form of the enzyme lacking the disulfide reductase and the two-subunit form of CODH which has been studied in various laboratories since 1983 (29) is fully active. Although we do not know the role, if any, of the disulfide reductase, it seems clear that claims that CODH is a “three-subunit” enzyme are unjustified.

Effect of CO on the CoAIAcetyl-CoA Exchange-Pezacka and Wood (11) observed 80% inhibition of the CoA/acetyl- CoA exchange by CO when reduced ferredoxin or NADPH was used as reductant. Ramer et al. (18) claimed that CO was not required for the CoA/acetyl-CoA exchange, but increased the reproducibility of the assay by scavenging trace amounts of oxygen. We have found that CO has two effects on CODH. First, CO is a powerful reductant which serves to reduce CODH. The midpoint potential for the reduction of COZ to CO is at -564 mV at pH 7.0, calculated using data from Ref. 26. Reduction of CODH by CO has been studied in some detail (see Ref. 4 for review). In the presence of redox mediators, CO will poise the redox potential of the reaction mixture at ”520 mV. Since low potentials stimulate the CoA/acetyl- CoA exchange, there is a strong stimulation of this reaction relative to the reaction performed in the absence of a reducing

Pezacka and Wood (11) measured -3.1 nmol of CoA exchanged into acetyl-coA in 30 min by 20 units of CODH (0.05 mg, -0.4 nmol of dimeric enzyme), which would give a minimum exchange activity of -2 nmol min” mg” (-14,000-fold lower than the rates measured here). Our estimate of this rate based on Pezacka and Wood’s data is only approximate since kinetic data for the exchange was not given. We feel that these values can probably be approximated as initial velocity values, since within 30 min only 8% of the CoA had exchanged and isotopic equilibrium would have been reached a t 50% exchange. We would expect that initial velocity would have continued until -25% exchange since we (also shown in Ref. 18 by Ramer et al.) have seen that the reaction does proceed fully to the calculated value of isotopic equilibrium and the initial velocity period is seen until -50% isotopic equilibrium is reached.

system. Second , CO acts as a noncompetitive inhibitor with respect to acetyl-coA in the CoA/acetyl-CoA exchange with a Ki value of -400 p~ (Fig. 5b) . This Ki value is very high in comparison with other measurements of the affinity of CO to CODH with K , values of 10 p~ in the CO/acetyl-CoA exchange and 25 p~ in the CO oxidation reaction (17). Based on the noncompetitive nature of the inhibition, it appears that CO can interact with similar affinity to either free or acetyl-CoA-bound CODH. The inhibitor site for CO could be the Ni-Fe-C center which appears to be the active site of CO binding.

Effect of Nitrous Oxide on CODH-N20 inhibits the CoA/ acetyl-coA exchange reaction by a mechanism which appears to involve oxidation of CODH according to Equation 6.

CODH,d + NZO + N, + CODH,, (6)

This was shown by coupling the oxidation of reduced MV to reduction of NzO and identification of the product of the reduction of NZO as Nz by mass spectroscopic methods.

NzO has been shown to inhibit methionine synthase leading to profound physiological disturbances (see Ref. 36 and references therein). The mechanism of inactivation is unclear, but is postulated to involve reaction of Co‘+ with NzO. NzO also has been shown to react with several transition metal complexes including the Co’+ form of vitamin B12, other Co’+ complexes, rhodium, and chromium (Refs. 31 and 32 and references therein). For the reaction with Co1+-BI2 the product was Coz+ and Nz. In contrast, the Co’+ state of the C/Fe-SP from C. thermoaceticum does not appear to react with NzO.

Our studies of the reactivity of CODH and inertness of the C/Fe-SP to NzO are important in light of a recent study of the inhibition by NzO of the exchange reaction between CO, and the carbonyl of acetyl-coA which is catalyzed by cell extracts of Methanosarcina barkeri with Hz as reductant (46). Although the system was not inhibited by propyl iodide, this inhibition by N20 was proposed to implicate a corrinoid protein in the exchange reaction. We offer an additional possibility that the inhibition in the methanogenic system could be via oxidation of an active site metal center on CODH as we have shown here with the C. thermoaceticum CODH rather than inhibition of a corrinoid protein. In acetogenic bacteria, reduction of COz to CO and the COjacetyl-CoA exchange clearly is catalyzed by CODH and addition of the C/Fe-SP has no effect on this exchange reaction. I t will be interesting to determine if the NzO inhibition of the C02/ acetyl-coA exchange activity in the methanogenic system could be related to oxidation of CODH.

A surprising result is that CODH can act as a NzO reductase at rates comparable with the rates of purified N2O reductases from some denitrifying bacteria, such as, Pseudomonas stutzeri (45). These enzymes have turnover numbers in the range of 50 to 280 s-’. Optimal conditions for determination of the MV/N20 reductase assay still have to be determined. Since CODH can act as a N 2 0 reductase, it would be interesting to determine how active C. thermoaceticum is in denitrification in uiuo. Denitrification is an eight-electron process in which nitrate is reduced to N2 via NOz, NO, and N20 as intermedi- ates. Copper enzymes from P. stutzeri (45) and Paracoccus denitrificans (47) have been purified and studied by enzymology and the copper sites analyzed by spectroscopic methods. Interestingly, as for CODH, N20 reductases are redox sensitive, with enzymes isolated under anaerobic conditions having the highest activity (45, 48). The copper enzyme from Rho- dopseudomonas sphaeroides fsp. denitrificans requires reductive activation for maximal activity (49). In addition, the NzO reductase from P. stutzeri reacts with CO which was proposed


to bind to the active site metal center and produce COZ (45). Assembly of Acetyl-coA at a Single Metal Center on CODH-

Based on the results presented here, it appears that a metal center on CODH with a redox potential 5-486 mV is reduced by one electron to become catalytically active in formation of the acetyl-CODH intermediate in the Wood pathway. Reduc- tive activation a t similar potentials is also required for methylation of CODH and for catalysis of the exchange reaction between methyl-CODH and the methyl group of acetyl-coA ( 5 ) . A study of these methyl transfer reactions indicate that the methyl acceptor is a reductively activated metal center with a midpoint reduction potential 5-400 mV (from a study of the former reaction) or 5-450 mV (from a study of the latter reaction). The pH profiles of the CoA/acetyl-CoA exchange and the synthesis of methyl-CODH from the methylated C/Fe-SP7 also are similar. In addition, the rate of reduction of CO, to CO by CODH exhibits a redox dependence which is similar to the methylation and CoA/acetyl-CoA exchange reactions. When the reduction of CO, and the formation of the Ni-Fe-C EPR signal of CODH were determined as a function of redox potential, both processes were found to exhibit a reduction potential for half-activity of "430 mV (8). The Ni-Fe-C EPR signal arises from a novel mixed metal complex which is formed upon reaction of CODH with CO. Thus, the CoA/acetyl-CoA exchange, methylation of CODH, CO, reduction, and the formation of the Ni-Fe-C EPR signal all occur at similar redox potentials. Based on the similar natures of the reductive activation, we argue that the same low-potential metal site on CODH binds the methyl group of the methylated corrinoid/iron-sulfur protein, CO, and the acetyl group of acetyl-coA. In adddition, we propose that the reductively activated metal center involved in these reactions is the metal center which reacts with CO to form the Ni-Fe-C center.

The Ni-Fe-C center of CODH has been studied by spectroscopic and electrochemical methods. This complex which is formed rapidly upon reaction of CODH with CO is paramag- netic with g-values a t 2.074 and 2.028 (6, 50) and has a reduction potential of 5-430 mV8 (8). Recent spectroscopic results (9, 10) suggest two possible types of models for the structures of the Ni-Fe-C center: a Ni-X-[4Fe-4S] complex, where the nickel is bridged to a [4Fe-4S] cluster, or a [Ni- 3Fe-4SI center, where nickel is incorporated directly into a cubane center. Recently, nickel has been incorporated as one of the components of a cubane structure (51). That the methyl, CO, and acetyl groups could bind at the same site is supported by model chemistry. Recently, nickel complexes in a sulfur-rich environment were shown to form methylnickel, nickel-carbonyl, and acetylnickel and, in the presence of thiols, thioesters could be formed (52).

' The pH profile referred to in Ref. 5 is for the synthesis of acetyl- CoA with methyl iodide, CO, and CoA in the presence of saturating amounts of C/Fe-SP and catalytic amounts of CODH. We showed that CODH is methylated by methyl iodide with the methylated C/ Fe-SP as an obligate intermediate. Therefore, the pH profile under these reaction conditions, reflects the rate-limiting methylation of CODH by the methylated C/Fe-SP. The only relevant pH dependent ionization of methyl iodide would be the formation of HI, which would occur a t much lower pH values (pK, of HI = 0.77, Ref. 26).

REarlier, we proposed that the redox potential for the Ni-Fe-C species responsible for formation of this EPR signal must have a redox potential within 70 mV of this half-activation potential, i.e. between -350 and -520 mV (8). However, since formation of the Ni- Fe-C EPR signal from CO, requires the reduced form of the Ni-Fe- containing species, the midpoint reduction potential of the Ni-Fe-C species is 5-430 mV, and the ranges cannot be determined from this kinetic experiment. This conclusion is based on the considerations discussed above.

Acknowledgments-We thank Dr. Frank Laib for performing the mass spectroscopy, Dr. Scott Harder and Robert Ponton for design and construction of the electrochemical cell, David Roberts for construction of the UV-visible spectroelectrochemical cell, and Carol Gorst for measuring the concentrations of CO in solution and her assistance in performing the curve-fitting analysis of CoA binding to CODH.

1.

2.

3. 4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25. 26.

27.

28.

29.

30.

31.

32.

33.

REFERENCES

Wood, H. G., Ragsdale, S. W., and Pezacka, E. (1986) Trends

Wood, H. G., Ragsdale, S. W., and Pezacka, E. (1986) Biochem.

Ljungdahl, L. G. (1986) Annu. Reu. Microbiol. 40,415-450 Ragsdale, W. S., Wood, H. G., Morton, T. A. Ljungdahl, L. G.,

and DerVartanian, D. V. (1988) in The Bioinorganic Chemistry of Nickel (Lancaster, J . R., ed) pp. 311-333, VCH, New York

Lu, W-P., Harder, S. R., and Ragsdale, S. W. (1990) J. Biol. Chem. 265,3124-3133

Ragsdale, S. W., Wood, H. G., and Antholine, W. E. (1985) Proc. Natl. Acad. Sci. U. S. A. 82, 6811-6814

Ragsdale, S. W., Ljungdahl, L. G., and DerVartanian, D. V. (1983) Biochem. Biophys. Res. Commun. 115,658-665

Lindahl, P. A., Munck, E., and Ragsdale, S. W. (1990) J. Biol. Chem. 265,3873-3879

Lindahl, P. A., Ragsdale, S. W., and Munck, E. (1990) J. Biol. Chem. 265,3880-3888

Fan, C.-L., Gorst, C. M., Ragsdale, S. W., and Hoffman, B. M. (1990) Biochemistry, in press

Pezacka, E., and Wood, H. G. (1986) J. Biol. Chem. 261, 1609- 1615

Hu, S.-I., Drake, H. L., and Wood, H. G. (1982) J. Bacteriol. 149,

Shanmugasundaram, T., Ragsdale, S. W., and Wood, H. G. (1988) BioFactors 1, 147-152

Shanmugasundaram, T., Kumar, G. K., and Wood, H. G. (1988) Biochemistry 27,6499-6503

Ragsdale, S. W., and Wood, H. G. (1985) J. Biol. Chem. 260, 3970-3977

Lebertz, H., Simon, H., Courtney, L. F., Benkovic, S., Zydowsky, L. D., Lee, K., and Floss, H. G. (1987) J. Am. Chem. SOC. 109,

Raybuck, S. A,, Bastian, N. R., Orme-Johnson, W. H., and Walsh,

Ramer, S. E., Raybuck, S. A., Orme-Johnson, W. H., and Walsh,

Ljungdahl, L. G., and Andreesen, J. R. (1978) Methods Enzymol.

Ragsdale, W. S., Lindahl, P. A,, and Munck, E. (1987) J. Biol.

Harder, S. R., Lu, W.-P., Feinberg, B. A., and Ragsdale, S. W.

Harder, S. R., Feinberg, B. A,, and Ragsdale, S. W. (1989) Anal.

Segel, I. H. (1975) Enzyme Kinetics, John Wiley & Sons, New

Bonam, D., Murrell, S. A., and Ludden, P. W. (1984) J. Bacteriol.

Penefsky, H. S. (1977) J. Biol. Chem. 252, 2891-2899 Weast, R. C . (ed) (1985) Handbook of Chemistry and Physics,

67th Ed., D-58, CRC Press, Cleveland Elliott, J. I., and Brewer, J. M. (1978) Arch. Biochem. Biophys.

97.37-40 Weber, K., and Osborn, M. (1975) in The Proteins (Neurath, H.,

and Hill, R. L., eds) 3rd Ed., Vol. 1, pp. 179-223, Academic Press, New York

Ragsdale, S. W., Clark, J. E., Ljungdahl, L. G., Lundie, L. L., and Drake, H. L. (1983) J. Biol. Chem. 258, 2364-2369

Diekert, G., Hansch, M., and Conrad, R. (1984) Arch. Microbiol.

Banks, R. G. S., Henderson, R. J., and Pratt, F. M. (1968) J .

Blackburn, R., Kyaw, M., and Swallow, A. J . (1976) J. Chem.

Bell, J. E., and Bell, E. T. (1988) Proteins and Enzymes, Prentice-

Biochem. Sci. 11, 14-18

Znt. 12, 421-440

440-448

3173-3174

C. T. (1988) Biochemistry 27, 7698-7702

C. T. (1989) Biochemistry 28,4675-4680

53,360-372

Chem. 262, 14289-14297

(1989) Biochemistry 28,9080-9087

Biochem. 181, 283-287

York

159,693-699

138,224-228

Chem. Soc. (Lond.) (A) 2886-2889

SOC. Faraday Trans. 73, 250-255

Hall, Inc., Enlewood Cliffs, NJ


34. Rouvicre, P. E., Bobik, T. A., and Wolfe, R. S. (1988) J. Bacteriol. 47. Snyder, S. W., and Hollocher, T. C. (1987) J. Biol. Chem. 262,

35. Fernandez, V. M., Aguirre, R., and Hatchikian, E. C. (1984) 48. Coyle, C. L., Zumft, W. G., Kroneck, P. M. J., Korner, H., and

36. Banerjee, R. V., and Matthews, R. G. (1990) FASEB J. 4, 1450- 49. Michalski, W. P., Hein, D. H., and Nicholas, D. J. D. (1986)

37. Janick, P. A,, Rueger, D. C., Krueger, R. J., Barber, M. J., and 50. Ragsdale, S. W., Ljungdahl, L. G., andDerVartanian, D. V. (1982)

38. Sligar, S. (1976) Biochemistry 15,6399-6406 51. Conover, R. C., Park, J.-B., Adams, M. W. W., and Johnson, M. 39. Walsh, T. A,, Johnson, M. K., Greenwood, C., Barber, D., Sprin- K. (1990) J. Am. Chem. SOC. 112,4562-4564

gall, J. P., and Thompson, A. J. (1979) Biochem. J . 177, 29- 52. Stavropoulos, P., Carrie, M., Muepterties, M. C., and Holm, R. 39 H. (1990) J. Am. Chem. SOC. 112, 5385-5387

40. Pember, S. O., Villafranca, J. J., and Benkovic, S. J. (1986) 53. Bard, A. J., and Faulkner, L. R. (1980) Electrochemical Methods, Biochemistry 25,6611-6619 John Wiley & Sons, New York

41. Marota, J. J. A,, and Schiman, R. (1984) Biochemistry 23, 1303- 54. Lenn, N. D., Stankovich, M. T., and Liu, H.-W. (1990) Biochem- 1311 istry 29,3709-3715

42. Wallick, D. E., Bloom, L. M., Gaffney, B. J., and Benkovic, S. J. 55. Watt, G . D., Wang, Z.-C., and Knotts, R. R. (1986) Biochemistry (1984) Biochemistry 23,1295-1302 25,8156-8162

43. Crawford, N. A., Droux, M., Kosower, N. S., and Buchanan, B. 56. Bevington, P. R. (1969) Data Reduction and Error Analysis for B. (1989) Arch. Biochem. Biophys. 271,223-239 the Physical Sciences, p. 237, McGraw-Hill Book Co., New

44. Black, S. (1983) J. Biol. Chem. 258, 2112-2114 York 45. Riester, J., Zumft, W. G., and Kroneck, P. M. H. (1989) Eur. J . 57. Feldman, H. A. (1972) Anal. Biochem. 48, 317-338

46. Fischer, R., and Thauer, R. K. (1990) Arch. Microbiol. 153, 156- 69

170,3946-3952 6515-6525

Biochim. Biophys. Acta 780, 1-7 Jakob, W. (1985) Eur. J . Biochem. 153,459-467

1459 Biochim. Biophys. Acta 872, 50-60

Siegel, L. M. (1983) Biochemistry 22, 396-408 Biochem. Biophys. Res. Commun. 108, 658-663

Biochem. 178, 751-762 58. Pezacka, E., and Wood, H. G. (1984) Arch. Microbiol. 137, 63-

162 59. Dixon, H. B. F. (1979) Biochem. J . 177, 249-250

reductive activation of the coenzyme a/acetyl-coa isotopic

Documents