Spectroscopic and Kinetic Studies of the Reaction of Bromopropanesulfonate with Methyl-coenzyme M Reductase*

Methyl-coenzyme M reductase (MCR) catalyzes the final step of methanogenesis in which coenzyme B and methyl-coenzyme M are converted to methane and the heterodisulfide, CoMS-SCoB. MCR also appears to initiate anaerobic methane oxidation (reverse methanogenesis). At the active site of MCR is coenzyme F430, a nickel tetrapyrrole. This paper describes the reaction of the active MCRred1 state with the potent inhibitor, 3-bromopropanesulfonate (BPS; I50 = 50 nm) by UV-visible and EPR spectroscopy and by steady-state and rapid kinetics. BPS was shown to be an alternative substrate of MCR in an ionic reaction that is coenzyme B-independent and leads to debromination of BPS and formation of a distinct state (“MCRPS”) with an EPR signal that was assigned to a Ni(III)-propylsulfonate species (Hinderberger, D., Piskorski, R. P., Goenrich, M., Thauer, R. K., Schweiger, A., Harmer, J., and Jaun, B. (2006) Angew. Chem. Int. Ed. Engl. 45, 3602-3607). A similar EPR signal was generated by reacting MCRred1 with several halogenated sulfonate and carboxylate substrates. In rapid chemical quench experiments, the propylsulfonate ligand was identified by NMR spectroscopy and high performance liquid chromatography as propanesulfonic acid after protonolysis of the MCRPS complex. Propanesulfonate formation was also observed in steady-state reactions in the presence of Ti(III) citrate. Reaction of the alkylnickel intermediate with thiols regenerates the active MCRred1 state and eliminates the propylsulfonate group, presumably as the thioether. MCRPS is catalytically competent in both the generation of propanesulfonate and reformation of MCRred1. These results provide evidence for the intermediacy of an alkylnickel species in the final step in anaerobic methane oxidation and in the initial step of methanogenesis.

Methanogens are strictly anaerobic microbes in the kingdom Archaea that form methane to generate cellular energy. Approximately 1 billion tons of methane are produced by these anaerobic microorganisms per year (1), which generates signif-icant amounts of a valuable fuel but poses a potential environmental problem because methane is a potent greenhouse gas. Methyl-coenzyme M reductase (MCR) 3 catalyzes the final step in biological methane formation: the reaction of methyl-coenzyme M (methyl-SCoM) with N-7-mercaptoheptanoyl-threonine phosphate (HSCoB) to generate methane and the mixed disulfide CoBS-SCoM (Reaction 1) (1,2). HSCoB serves as the two-electron donor (3). At the active site of MCR is a nickel tetrapyrrolic cofactor called coenzyme F 430 that is central to catalysis (4 -6). X-ray crystallographic studies reveal that coenzyme F 430 binds noncovalently to MCR at the bottom of a 30-Å hydrophobic channel (7). The phosphate group of HSCoB binds at the upper lip of this channel with its thiol group located 8.7 Å from the central nickel atom of F 430 .
Methyl-SCoM ϩ HSCoB 3 CH 4 ϩ CoBS-SCoM REACTION 1 ⌬G ϭ Ϫ30 kJ/mol (Eq. 1) MCR can exist in several states that differ in their nickel oxidation and/or coordination states (Fig. 1). The MCR red1 state is catalytically active (8 -10) based on the correlation between the amount of this enzyme form and MCR activity in Methanothermobacter marburgensis (10) and Methanosarcina thermophila (11). MCR ox1 appears to be the only state that can be converted in vitro to active MCR red1 , which is accomplished by incubating the MCR ox1 state with a strong reductant, titanium(III) citrate (Ti(III) citrate) (10). The MCR red1 state also can be generated in vivo when cells are bubbled with 100% H 2 before harvesting (8). MCR ox1 can be formed when the gas medium of growing cells is switched before harvesting from 80% H 2 , 20% CO 2 to 80% N 2 , 20% CO 2 (8) or by treating the growing cells with sodium sulfide (11).
High resolution crystal structures for three EPR-silent and inactive Ni(II) states of this enzyme have been determined: MCR silent , MCR ox1-silent , and MCR red1-silent (7,12,13). These states have several common features: the central nickel atom of F 430 is coordinated by four planar tetrapyrrole nitrogen atoms and a lower axial oxygen ligand contributed by the carbonyl oxygen of the side chain of Gln-␣Ј147. In the Ni(II) silent form of MCR, the upper axial nickel ligand is the sulfonate oxygen of CoBS-SCoM, whereas, in the Ni(II) ox1-silent form, this site is occupied by the thiol(ate) group of CoM-S(H) (Fig. 1) (7,12,13). A five-coordinate form of Ni(II)-MCR red1-silent , lacking an upper axial ligand, has also been observed in the crystal structure (13). The structures of "ready" MCR ox1 and "active" MCR red1 forms have not yet been determined, since they are quite labile. Based on an array of spectroscopic (x-ray absorption (14), UV-visible (15,16), EPR (14,17), pulsed EPR (17), and magnetic circular dichroism (15,16)) and computational methods (15), MCR ox1 is best described as a high spin Ni(II) coupled to a thiyl-radical (15,17) (Fig. 1). MCR red1 is known to be in the Ni(I) state (9 -11). Recent x-ray absorption results (14,18) indicate that the nickel is six-coordinate in the MCR ox1 state, yet there is still some controversy about the MCR red1 state, which is either five-coordinate with an open upper axial site (14) or sixcoordinate with two axial ligands coordinated through an oxygen atom (18) (Fig. 1). The coordination state of MCR red1 needs to be resolved, because having an open upper axial site would poise the Ni(I) for reaction with substrate, whereas an axial ligand would need to be replaced/removed before reaction could occur. Kinetic studies using several small molecules (CHCl 3 , CH 2 Cl 2 , and NO) and substrate analogs (including 3-bromopropanesulfonate (BPS)) support the requirement for ligand exchange (19); however, further studies are important in resolving this mechanistic question.
Two radically different mechanisms have been proposaled for the catalytic mechanism of MCR. In Mechanism I, Ni(I)-MCR red1 reacts with the methyl group of methyl-SCoM, forming a methyl-Ni(III) or methyl-Ni(II) intermediate (20 -23). Although the corresponding methyl-nickel species has not been observed on the enzyme, methane formation from the reaction of Ni(I)-derivatives of F 430 and activated methyl donors like methylsulfonium ions has been observed (24), and a methyl-Ni(II) form of the pentamethyl ester of F 430 has been characterized by NMR methods (25). Furthermore, reacting Ni(I)-octaethylisobacteriochlorin (a structural cousin of F 430 ) with various alkyl halides generates alkyl-Ni(III) species that undergo reduction to the alkyl-Ni(II) followed by protonolysis to yield the corresponding alkane (26). 4 On the basis of a computational study using density function theory (DFT), the proposed methyl-nickel species was considered to not be a feasible intermediate in methane synthesis, and Mechanism II was proposed as an alternative to Mechanism I (27). The key steps of Mechanism II include nucleophilic attack by Ni(I)-MCR red1 on methyl-SCoM to form a nickel-thiolate complex and a methyl radical, which abstracts a hydrogen atom from HSCoB to generate methane (27).
Unfortunately, it has been difficult to distinguish between these mechanisms, mainly because we have so far been unable to observe any spectrally distinct intermediates by rapid kinetics. 5 Another complication is that MCR red1 -catalyzed cleavage of the C-S bond of methyl-SCoM requires the other substrate, HSCoB, even under single turnover conditions (23). In the studies described here, we used BPS as a substrate analog, because, unlike methyl-SCoM, reduction of BPS in a single turnover reaction does not require the other substrate HSCoB; furthermore, EPR and UV-visible changes can be observed. When MCR red1 and MCR red2 are incubated with BPS, their EPR signals convert to the MCR PS signal, which does not exhibit measurable halogen-Ni(I) interaction (28) and has been assigned to a high spin Ni(II)/alkyl radical species (29). Here we show that this EPR spectroscopic change is accompanied by formation of an "MCR ox1 -like" UV-visible spectrum and the reduction of BPS to propanesulfonate.

EXPERIMENTAL PROCEDURES
Materials and Organisms-M. marburgensis (f. M. thermoauotrophicum strain Marburg) was obtained from the Oregon S-CH 2 -CH 2 -SO 3 is coenzyme M, -SO 3 -CH 2 -CH 2 -S-S-CoB is the heterodisulfide, and H 2 N-CO-is glutamine residue Gln-␣147. The crystal structures of the MCR ox1 , MCR PS , and MCR red1 states have not been solved. The oxidation state and coordination environment of nickel in these two states are described in the Introduction.
Methyl-SCoM was prepared from coenzyme M and methyl iodide (30). After methyl-SCoM was crystallized twice in acetone, its purity was checked by 1 H NMR spectroscopy. CoBS-SCoB was prepared as described from 7-bromoheptanoic acid (Karl Industries, Aurora, OH), N-hydroxysuccinamide, and o-phospho-L-threonine (31,32). HSCoB was generated from CoBS-SCoB by anaerobic reduction with NaBH 4 (33), and its purity was ascertained by HPLC. The concentration of HSCoB was checked routinely with Ellman's reagent (34) before using it. Ti(III) citrate solutions were prepared from a stock solution of 200 mM Ti(III) citrate, which was synthesized by adding sodium citrate to Ti(III) trichloride (30 weight % solution in 2 N hydrochloric acid) (Acros Organics, Morris Plains, NJ) under anaerobic conditions and adjusting the pH to 7.0 with sodium bicarbonate (35). The concentration of Ti(III) citrate was determined routinely by titrating a methyl viologen solution.
M. marburgensis Growth, Harvest, and Purification Conditions-M. marburgensis was cultured on H 2 /CO 2 /H 2 S (80%/ 20%/0.1%) at 65°C in a 14-liter fermentor (New Brunswick Scientific Co., Inc., New Brunswick, NJ) (10,36). Culture media were prepared as previously described (36). Once the culture had reached late log phase (A 578 ϳ 3.0), sodium sulfide was added directly to the growing cells (final sulfide concentration, 20 mM) prior to harvesting as described (11). After 15 min at 65°C, the culture was cooled to 25°C within 30 min and harvested anaerobically. MCR red1 used in Figs. 2,5,8,9, S1, and S2 was isolated (details below) from M. marburgensis cultured on H 2 /CO 2 /H 2 S (80%/20%/0.1%) at 65°C in a 14-liter fermentor (New Brunswick Scientific) (10,36). Culture media were prepared as previously described (36). MCR red1 was generated in vivo as described with the following modifications (8,37). Once the culture had reached A 578 ϳ 4.0, the H 2 /CO 2 /H 2 S (80%/20%/ 0.1%) was switched to H 2 /CO 2 (80%/20%). 30 min after the H 2 S gas flow was turned off, the CO 2 /H 2 (20%/80%) was switched to 100% H 2 . After purging for 20 min with 100% H 2 at 65°C, the cells were cooled to ϳ20°C within 10 min. After purging the cells with 100% H 2 for a total of 30 min, the cells were harvested anaerobically, and gas pressure was used to transfer the cells from the fermentor into a Cepa LE cell harvester (New Brunswick Scientific). After harvesting the cells, the rotor was immediately moved into the anaerobic chamber. The cells were removed from the rotor and resuspended in 200 ml (total volume) of 50 mM Tris-HCl, pH 7.6, 1 mM Ti(III) citrate, and 10 mM HSCoM (lysis buffer). The resuspended cells were aliquoted (50-ml aliquots) into four 100-ml serum vials (Alltech, Deerfield, IL) fitted with butyl rubber stoppers (Bellco Glass Inc., Vineland, NJ) and crimp-sealed with aluminum caps (Bellco Glass Inc.). The resuspended cells were placed under a 100% H 2 atmosphere by purging the headspace of the seal serum vials with 100% H 2 for 5 min, followed by vigorously shaking the resuspended cells by hand for 30 s and then purging with 100% H 2 for an additional 5 min. The serum vials were inverted to minimize H 2 escape, and the cells were kept under 10 p.s.i. of 100% H 2 at room temperature with replacement of the H 2 atmosphere every 6 -8 h. EPR spectroscopy was used to monitor in vivo generation of MCR red1 by drawing a 200-l subsample of resuspended whole cells every 6 -8 h. To remove trace amounts of oxygen in the H 2 line entering the anaerobic chamber, the line was fitted with a high pressure Oxy-Trap oxygen scrub (Alltech) followed by an indicating Oxy-Trap (Alltech) to monitor the efficiency/remaining capacity of the oxygen scrub. The cells were processed when the EPR intensity of MCR red1 stopped increasing, typically 24 h after cell harvest. MCR red1 and MCR ox1 were purified under strictly anaerobic conditions at 17°C in a Vacuum Atmospheres chamber maintained below 1 ppm oxygen, monitored continually with a Teledyne oxygen analyzer (model 317; Teledyne Analytical Instruments, City of Industry, CA).
MCR ox1 was purified as described, except that methyl-SCoM was omitted from the lysis buffer (10). For experiments requiring high concentrations of MCR, MCR ox1 was concentrated in a 50-ml ultrafiltration stirred cell (Amicon; Millipore Corp., Bedford, MA) with a 30 kDa molecular mass cut-off filter using high pressure argon that had been passed through an oxisorb column (Oxyclear; Supelco) to remove oxygen. MCR red1 used for the experiments presented in Figs. 2, 5, 8, 9, S1, and S2 was purified as described with the following modifications (10). The pellet from the 100% ammonium sulfate fractionation was resuspended in 60 ml of 50 mM Tris, pH 7.6, 10 mM HSCoM, 0.1 mM Ti(III) citrate (buffer A). The resuspended 100% ammonium sulfate pellet was applied at a flow rate of 5 ml/min to a 1.6 ϫ 25-cm high performance Q-Sepharose (Sigma) column (C 16/40 column fitted with an AC 16 flow adapter (GE Health)) equilibrated in buffer A. After loading the resuspended pellet, the column was washed with 60 ml of buffer A, and protein was eluted with a step gradient of 0.44 M NaCl in buffer A (60 ml) followed by 0.55 M NaCl in buffer A (90 ml). Five-ml fractions were collected and analyzed for MCR red1 by UV-visible spectroscopy. MCR isoenzyme I typically eluted at the beginning of the 0.55 M NaCl step. The most concentrated fractions were transferred to separate 9-ml serum vials (Alltech) and stoppered with butyl rubber stoppers (Bellco). The serum vials were stored in the anaerobic chamber on ice packs in Styrofoam containers. In this way, MCR red1 can be stored in the anaerobic chamber for 3-4 months before losing all activity. This purification method routinely generates 50 -80% MCR red1 as determined by UV-visible and EPR spectroscopy.
Spectroscopy of MCR-UV-visible spectra of MCR were recorded in the anaerobic chamber using a diode array spectrophotometer (model DT 1000A; Analytical Instrument Systems, Inc., Flemingron, NJ). EPR spectra were recorded on a Bruker ESP 300E spectrometer recently upgraded to an EMX, equipped with an Oxford ITC4 temperature controller, a Hewlett-Packard model 5340 automatic frequency counter, and Bruker gaussmeter. Unless otherwise noted, the EPR spectroscopic parameters included the following: temperature, 100 K; microwave power, 10 milliwatts; microwave frequency, 9.43 GHz; receiver gain, 2 ϫ 10 4 ; modulation amplitude, 12.8 G; modulation frequency, 100 kHz. Double integrations of the EPR spectra were performed and referenced to a 1 mM copper perchlorate standard. All NMR data were acquired at 298 K on a Bruker Avance DRX 500-MHz NMR instrument (Bruker Biospin Corp., Billerica, MA) equipped with a TXI cryoprobe in the UNL Chemistry Department.
Conversion of MCR ox1 to MCR red1 -For the experiments presented in Figs. 3, 4, 6, and 7, MCR ox1 was activated to the MCR red1 state in the absence of methyl-SCoM or HSCoM in a reaction mixture containing 20 mM Ti(III) citrate, 0.67 M TAPS (pH 10), 16.7 mM Tris-Cl (pH 7.6), and MCR (30 -200 M). The mixture was heated at 60°C for 40 min, cooled on ice, and then neutralized to pH 7.0 -7.1 by adding an equal volume of 2 M Tris-HCl buffer (pH 7.0) (11). Protein concentrations were determined by the Bradford method using the Bio-Rad reagent and bovine serum albumin as a standard (38). F 430 content was estimated using an extinction coefficient of 22,000 cm Ϫ1 M Ϫ1 at 420 nm (39).
Conversion of MCR red1 to MCR PS -MCR red1 used in Figs. 8, 9, and S1 was incubated for 5 min with a 10-fold excess of BPS in 50 mM Tris, pH 7.6. Unreacted BPS was removed from MCR PS by buffer exchange using Amicon Ultra-15 centrifuge filter units with a 50 kDa cut-off (Millipore). Typically, 300 -600 l of MCR PS /BPS reaction mixture was exchanged into 3-6 ml of 50 mM Tris, pH 7.6. This mixture was concentrated to 200 -300 l, and this process was repeated three times. The EPR spectroscopic parameters were as noted above except the following: temperature, 70 K; microwave power, 10.2 milliwatts; modulation amplitude, 5.0 G; scans, 4.
Measurement of MCR Activity-MCR assays were performed at 65°C in rubber-sealed 8-ml serum vials. The standard assay mixture contained 20 mM methyl-SCoM, 1.2 mM HSCoB, 0.6 mM aquocobalamin, 25 mM Ti(III) citrate, and 0.5 M MOPS (pH 7.0) in a final volume of 0.4 ml. The reaction was started by increasing the temperature from 4 to 65°C. The methane generated was determined by withdrawing gas samples at specific time points for analysis by gas chromatography (Varian model 3700 equipped with a flame ionization detector). Alternatively, MCR activity was measured by following the time-dependent loss of radioactivity as 14 CH 3 -SCoM was converted to [ 14 C]methane (23). Rates of methane formation were calculated from the linear portion of the time course. One unit of MCR activity is equal to 1 mol of methane min Ϫ1 .
Reactions between MCR red1 and Sulfonates-MCR red1 (typically 0.1-0. Stopped-flow Studies-Stopped-flow experiments were performed on an Applied Photophysics spectrophotometer (SX.MV18; Leatherhead, UK) equipped with a photodiode array detector. Constant temperature was maintained with a bath of nitrogen-bubbled water from a circulating pump to maintain anaerobicity. Rigorous measures were taken to purge oxygen from the stopped-flow instrument. The solutions of enzymes and inhibitors were made in the anaerobic chamber in 0.5 M TAPS, pH 10.0. The solutions were then loaded into tonometers, which had been incubated in the anaerobic chamber for 4 days and served as reservoirs for the drive syringes of the stopped-flow instrument. The drive syringes were maintained anaerobically at 25°C in a temperature-controlled bath of anaerobic water. MCR red1 and varied concentrations of BPS were rapidly mixed at 25°C in a 1:1 ratio. The reaction was monitored in the single wavelength mode by following the decay of MCR red1 at 388 and 715 nm, and MCR PS formation was followed at 422 nm. Data were fit to single exponential decay functions with software provided by Applied Photophysics (version SX MV.18). Reported rate constants are the average of at least five different rapid mixing experiments.
HPLC Connected to Conductivity Detector-The reaction between MCR red1 and BPS was initially followed by HPLC using a 3.9 ϫ 150-mm Bondapak C 18 analytical column (Waters, Milford, MA), which was developed with a 0 -80% methanol gradient in 10 mM ammonium formate (pH 3.3) (30 min at a flow rate of 1 ml/min). BPS decay was monitored using an absorbance detector set at 207 nm, whereas product formation was followed using a conductivity detector (Wescan Instruments Inc., Deerfield, IL). In this experiment, after activation in 0.5 M TAPS (pH 10.0), the MCR solution was exchanged with a solution containing 50 mM Tris-HCl (pH 7.6), 2 mM Ti(III) citrate using a 50-ml ultrafiltration stirred cell (Amicon) with a 30 kDa molecular mass cut-off. This step also removed any free F 430 that could have been released from the enzyme during the activation step. The reaction was started by injecting 20 l of 0.5 M BPS solution (50 mM Tris-HCl, pH 7.6) into the 2-ml enzyme mixture. At each time point, 200 l of reaction mixture was removed, frozen in EPR tubes, and observed by EPR spectroscopy. Next, the enzyme solution was transferred to a 1.5-ml microcentrifuge tube and exposed to O 2 for 30 min. After diluting the sample with the same amount of Tris-HCl buffer, enzyme and ligands were separated by using microconcentrators (Amicon) with 30 kDa cut-off. The clear filtrate, which lacked MCR, was injected directly into the HPLC with the conductivity detector. The C 18 column was pre-equilibrated with 50 mM formic acid (pH 3.0) and eluted with the same buffer at a flow rate of 0.5 ml/min. The retention times for BPS and PS were 6 and 5 min, respectively.
Chemical Quench Studies-Chemical quench experiments presented in Fig. 6 were performed at 20°C using an Update Instruments (Madison, WI) chemical/freeze-quench apparatus with a model 745 controller. Rapid reaction kinetic studies were performed with MCR red1 and BPS in separate 2-ml stoppedflow syringes. Solutions of MCR red1 and BPS (both in buffer containing 50 mM Tris-HCl, pH 7.6, and 5 mM Ti(III) citrate) were rapidly mixed by activating the ram to displace each syringe by 1.3 mm, generating a total reaction volume of 82 l (41 l from each syringe) per shot. The ram speed was varied from 0.8 to 8.0 cm/s and was shown not to affect the observed reaction rate. Typically, a ram speed of 4.0 cm/s was used with a 100-ms aging hose. For each data point, six shots were collected (492 l) in an 18-ml scintillation vial containing 0.2 ml of 0.5 N formic acid. After the chemical quench experiment, each sample was lyophilized and dissolved in 100 l of deionized water. The time course of each reaction was followed by monitoring the decrease in concentration of BPS and the formation of PS. The conditions and profile for eluting PS and BPS are the same as described above except that 80% methanol was used to wash out F 430 from the column after each run. Data were fit to single exponential equations using Sigma Plot (Point Richmond, CA).
MCR red1 activity was measured prior to the rapid reaction kinetic experiments. The typical specific activity of the MCR used for the stopped-flow or chemical quench studies was 20 -25 units/mg at 65°C, which would be equivalent to 60 -100 units/mg when corrected to 100% active MCR red1 . The amount of MCR red1 in each rapid reaction kinetic experiment was measured by double integration of the MCR red1 EPR signal before and after the stopped-flow or chemical quench experiments. Typical spin concentrations of the MCR red1 in the MCR samples for stopped-flow or chemical quench experiment ranged from 0.25 to 0.35 spin/mol of MCR, and less than 10% loss of this state occurred during the experiment.
MCR red1 used in the chemical quench studies presented in Fig. 5 was prepared by removing HSCoM by buffer exchange exchanged into 50 mM Tris, pH 7.6, 0.1 mM Ti(III) citrate with Amicon Ultra-15 centrifuge filter units with a 50 kDa cut-off (Millipore). The concentration of MCR red1 was determined by UV-visible spectroscopy using extinction coefficients of 27.0 mM Ϫ1 cm Ϫ1 and 9.15 mM Ϫ1 cm Ϫ1 at 385 and 420 nm, respectively, using a multiple wavelength calculation (40). These values were established from a 100% MCR red1 sample as determined by EPR using a 1 mM copper perchloric acid standard. The concentration of MCR silent was calculated using extinction coefficients of 22.0 and 12.7 mM Ϫ1 cm Ϫ1 at 420 and 385 nm, respectively (39). To ensure the accuracy of this method for determining the amount of MCR red1 in a solution containing a mixture of MCR red1 and MCR silent forms of the enzyme, the total protein concentration calculated with this method was compared with values determined using other methods. Protein concentrations calculated by this method were in good agreement (10 -15%) with the Bio-Rad reagent (Bio-Rad) using bovine serum albumin as a standard. Total protein concentrations were also measured by converting a heterogeneous mixture containing MCR red1/silent to a homogeneous mixture containing 100% MCR silent by exposing it to air and using an extinction coefficient of 22.0 mM Ϫ1 cm Ϫ1 to calculate the total MCR concentration (39).
Chemical quench experiments for Fig. 5 were performed with MCR red1 and BPS in separate 2-ml stopped-flow syringes. Solutions of MCR red1 and BPS were rapidly mixed by activating the ram to displace each syringe by 6.5 mm, generating a total reaction volume of 413 l (206.5 l from each syringe) per shot. A ram speed of 1.25 cm s Ϫ1 was used with a 500-ms aging hose. Each spectrum represents a total of 10 shots collected in a 30-ml serum vial (Alltech) containing 3 ml of 0.5 N formic acid to quench the reaction. After the reaction was quenched, enzyme and ligands were separated using a 50-ml ultrafiltration stirred cell concentrator fitted with a 50 kDa cut-off filter (Millipore). To separate BPS/PS from other ligands, mainly F 430 , the flowthrough collected from the stirred cell was applied to a 2-ml C 18 column equilibrated with water. The flow-through from the C 18 column was collected and lyophilized to dryness and dissolved in 550 l of 100% D 2 O/0.75% 3-(trimethylsilyl)-propionic acid-D4 by weight for NMR analysis.

RESULTS
Reaction of MCR red1 with BPS and Related Analogs, Studied by EPR-BPS is the most potent known inhibitor of MCR (41). When MCR red1 is reacted with BPS, a unique EPR signal called "MCR BPS " is observed, which, because of its air sensitivity and its similarity to the MCR red1 spectrum, was assigned as an Ni(I) state (42). A more recent study based on EPR data suggests that when MCR red1 reacts with BPS, bromide (or HBr) is presumably released, and a species is formed that can be described as an Ni(III)-propylsulfonate or a high spin Ni(II) with an alkylsulfonate radical (29). We confirmed this result (Table 1 and Fig. 2, inset) and have studied this reaction by various kinetic and spectroscopic methods (Reaction 2). Because the bromide has presumably undergone elimination, the designation "MCR BPS " is misleading, and we use the designation "MCR PS " instead. When MCR red1 is incubated with other structurally related sulfonates, an EPR signal nearly identical to MCR PS is observed (Table 1) (29). Compounds that elicit this spectral change include 3-chloropropanesulfonyl chloride (1 mM) and BBS (7 mM) ( Table 1). Unlike MCR ox1 and MCR red1 , these "MCR PSlike" EPR signals show no resolved hyperfine splitting, even when the modulation amplitude is decreased to 5.0 G. The MCR PS signal is relatively stable in the absence of oxygen or one of the reagents that react with it (below), decaying with a halflife of ϳ8 h at pH 7.6 in the anaerobic chamber. However, in the presence of oxygen, the signal is rapidly quenched within 5 min).
The requirement for the sulfonate group of BPS to elicit the MCR PS EPR signal was tested using 4-chlorobutyrate, in which a carboxylate group replaces the sulfonate. When reacted with MCR red1 , this BPS-like analog gave rise to an EPR signal nearly identical to MCR PS . A similar result was reported earlier using 4-bromobutyric acid (29). However, reaction of MCR red1 with propanesulfonate (70 mM), butanesulfonate (70 mM), and 3-mercapto-1-propanesulfonate (90 mM) neither decreased the amount of the MCR red1 signal nor elicited any EPR spectral change. The reaction of MCR red1 with BES quenched the MCR red1 EPR signal, as reported previously (29) ( Table 1).
We also tested the sulfonate-containing buffers MOPS, MES, CAPS, and CHES. Our initial data revealed that reacting MCR red1 with MOPS or CAPS gave the MCR PS signal (data not shown). The buffer MES and CHES quenched the MCR red1 signal, like BES (data not shown). A caveat to these experiments is the relatively high concentrations of MOPS and CAPS to elicit the MCR PS , 1.0 and 0.5 M, respectively. Further investigation led us to the description of the synthesis of these com-pounds, BPS as a starting material for MOPS and CAPS and BES for MES and CHES (43). This would explain the high concentrations needed to elicit the MCR PS signal, and the discrepancy in results from different batches of the same compound. For example, incubation of MCR red1 with 0.5 M of 98% pure CAPS at pH 7.0 elicited the MCR PS EPR signal, whereas incubating with 0.5 M 99% pure CAPS at pH 7.0 did not. Therefore, we suggest that the EPR signals observed upon incubating MCR red1 with MOPS or CAPS and quenching of the EPR signal upon incubating with MES or CHES is due to contamination from the BPS or BES remaining in the starting material, respectively.
Reaction of MCR red1 and MCR PS Formation, Followed by Rapid Kinetics-The UV-visible absorption spectra of EPR-active MCR PS (Fig. 2) resemble those of EPR-active MCR ox1 and EPR-silent Ni(II) MCR silent with absorbance maxima at 423 nm. In contrast, MCR red1 and Ni(I)F 430 exhibit a 40-nm blue shift relative to the MCR PS , MCR ox1 , and Ni(II) forms, which suggest that BPS causes a redox change in MCR. We followed the reaction of MCR red1 with BPS by absorption spectroscopy using a stopped-flow instrument (Fig. 3). Single exponential kinetics are observed although the reaction is not run under pseudofirst order conditions. Decay of the MCR red1 absorbance peaks at 388 and 715 nm matches the rate of formation of the MCR PS absorbance peak at 422 nm. This indicates that there is no intermediate between the MCR red1 and MCR PS states; alternatively, any intermediate that is formed is too transient to be observed. On the basis of the BPS concentration dependence of the UVvisible spectral changes (Fig. 4), the second order rate constant for MCR PS formation (or MCR red1 decay) is 1.6 ϫ 10 5 M Ϫ1 s Ϫ1 (at 20°C). This value is about 10-fold higher than the k cat /K m for methyl-SCoM (1.9 ϫ 10 4 M Ϫ1 s Ϫ1 at 65°C) and 100-fold larger than the k cat /K m for HSCoB (2.2 ϫ 10 3 M Ϫ1 s Ϫ1 ; 20°C) in methane formation (23). The rate constants at pH 7.6 (in 50 mM Tris-HCl) buffer are identical to those at pH 10.0 (in 0.5 M TAPS) (data not shown).
MCR red1 can also react with BBS but with significantly lowered efficiency. As shown in Fig. S2 (see supplemental material), the kinetic parameters for formation of MCR BS (in analogy with MCR PS ) and MCR red1 decay are as follows: k max ϭ 21.4 Ϯ 0.3 s Ϫ1 , K m ϭ 9.4 Ϯ 0.3 mM, k max /K m ϭ 2270 Ϯ 70 M Ϫ1 s Ϫ1 . Thus, the second order rate constant for reaction between MCR red1 and BBS is about 70-fold lower than that for reaction with BPS. In the case of BBS, the electrophilic carbon adjacent to the bromide leaving group is five bonds (ϳ6 Å) from the sulfonate group, which corresponds to the methyl group of methyl-SCoM.
Reaction of MCR ox1 with BPS-When MCR ox1 was incubated with BPS in the absence of Ti(III) citrate for several hours at room temperature, the MCR PS EPR signal appeared. The spectrum of MCR PS formed from MCR ox1 is indistinguishable from that formed from MCR red1 , yet the second order rate constant (2.4 M Ϫ1 s Ϫ1 ) (19) is 10 5 -fold slower than that for formation of MCR PS from MCR red1 (see above). This suggests, as proposed earlier (44), that the reactivity of MCR ox1 may be due to a very low amount (as low as 0.001%; i.e. 1/10 5 ) of MCR red1 present in the MCR ox1 solution. Such low amounts would be undetectable by any spectroscopic method. The rate of MCR ox1 decay matches the rate of MCR PS formation, and these rates are faster at pH 10.0 than at pH 7.6 (data not shown) (40), which also would be consistent with MCR red1 intermediacy in this process, since pH 10.0 is optimal for activation of MCR ox1 to MCR red1 with Ti(III) citrate (45). Furthermore, as described above, formation of MCR PS involves oxidation of MCR red1 to an "MCR ox1 -like" state and reduction of BPS (see above). Thus, the data indicate that MCR red1 , but not MCR ox1 , can generate the MCR PS state.
MCR red1 -catalyzed Propanesulfonate Formation from BPS-As noted above, the similarity between the UV-visible and EPR spectra of MCR PS and MCR ox1 indicates that MCR red1 is two electrons more reduced than MCR PS . The EPR spectra generated when MCR red1 is reacted with 3-iodo-or 3-chloropropanesulfonate is identical to MCR PS with no detectable hyperfine splitting from bromide or iodine (28), indicating that the halogen has been eliminated during formation of the MCR PS species. Rapid chemical quench methods were used to identify the alkylsulfonate group that is proposed to ligate to nickel (see Scheme 1). It was expected that quenching the reaction in acid would protonate the propylsulfonate ligand (forming HPS) and inactivate MCR. We removed HSCoM from the solution containing MCR red1 (MCR red1 is usually stored in the presence of HSCoM, because it stabilizes this form of the enzyme), because thiolates react directly with BPS in the absence of MCR. MCR red1 was rapidly mixed with BPS (in the absence of HSCoB and HSCoM) under single turnover conditions in the chemical quench apparatus, and the reaction was quenched into a solution containing 0.5 M formic acid. Then the small molecule products were separated from the protein by ultrafiltration and analyzed by NMR spectroscopy to measure the appearance of the product of this reaction. When MCR red1 and Ti(III) citrate (in a 2.4-and 1.4fold molar excess of the BPS, respectively) were reacted with BPS, HPS was detected as the product of the acid quench (Reaction 3; Fig. 5A). The reaction has a strict requirement for MCR red1 , since, when MCR red1 was omitted and Ti(III) citrate was present in a 2.9-fold molar excess (relative to BPS), the HPS peaks were absent, whereas peaks characteristic of BPS remained (Fig. 5B). Furthermore, when the MCR silent form of . MCR PS formed was exposed to air for 10 min until the EPR signal (see inset) of MCR PS was totally quenched (dashed line). Inset, representative EPR spectra of MCR red1 (top), MCR PS (middle), and MCR PS exposed to air (bottom). Note that the EPR signal in the bottom spectrum is not MCR PS but contaminating MCR ox1 , which is also seen in the spectra of MCR red1 and MCR PS . the enzyme was rapidly reacted with BPS under similar conditions, PS was not detected as a product (data not shown). These results demonstrate that the reaction of MCR red1 with BPS leads to debromination and formation of a propylsulfonate adduct (MCR PS ; Reaction 3) that can undergo protonolysis to form propanesulfonate (HPS; Reaction 4).
To determine the rate of propylsulfonate adduct formation relative to the rate of MCR PS generation (as measured by UVvisible and EPR spectroscopy), the MCR red1 -catalyzed conversion of BPS to PS was monitored under single turnover conditions by rapid chemical quench methods (quenched with formic acid). HPS formation and BPS depletion were measured by HPLC connected to a conductivity detector (Fig. 6). The product was identified by its elution at the same position as an authentic HPS standard. The rates and amplitudes for BPS decay and HPS formation were identical and equaled the rates of MCR PS formation and MCR red1 decay (above in Fig. 3).
These combined results indicate that the reaction of MCR red1 with BPS occurs according to Reaction 3 (i.e. the oxidation of MCR red1 is coupled to reduction of BPS to form MCR PS , which undergoes protonolysis to form propanesulfonate). Propanesulfonate appears to be the only product formed from BPS, as summarized in Scheme 1.
Evidence that MCR PS Can Be Reactivated to MCR red1 -The results described above demonstrate that Ni(I) of MCR red1 performs a nucleophilic attack on BPS to form MCR PS and bromide. This is an oxidative addition, much like the methyltransferase-catalyzed reaction of cob(I)alamin with CH 3 -H 4 -folate to form methyl-cob(III)alamin and H 4 -folate (46). The reaction of MCR with BPS has been thought to lead to irreversible inactivation of MCR. However, NMR results indicate that multiple turnovers of BPS conversion to propanesulfonate can occur.
PS formation was followed by HPLC under multiple turnover conditions in which MCR red1 was incubated with a 100-fold excess of BPS (Fig. 7). MCR red1 acts as a catalyst in this reaction, since the amount of propanesulfonate formed is 11-fold greater than the amount of enzyme. There are two major differences between the results of the steady-state reaction (Fig. 7) and the single turnover reaction (Fig. 6). The k obs for the single turnover reaction (2.6 s Ϫ1 ) (Fig. 6), which equals the rate constant for formation of MCR PS (Figs. 4 and 5), is much faster than the steady-state rate (2.2 h Ϫ1 , calculated based on either the exponential (k obs ) or the initial velocity (linear portion) in Fig. 7) of propanesulfonate formation, indicating that it is not chemistry but rather regeneration of MCR red1 or propanesulfonate release that is rate-limiting in the steady-state reaction. In addition, in the single turnover reaction, BPS is fully converted to propanesulfonic acid, whereas in the steady-state reaction, only 10% conversion occurs. This indicates that MCR red1 or MCR PS undergoes slow inactivation by conversion to a Ni(II) MCRsilent state during the steady state reaction. Thus, as shown in Scheme 1, MCR PS can suffer two alternative fates: reactivation to form MCR red1 or inactivation to form a Ni(II) silent state (or potentially a Ni(III) state   the reaction of MCR PS with HSCoM, we monitored the reaction by EPR (Fig. S1) and UV-visible spectroscopy (Fig. 8). Conversion of MCR PS to MCR red1 in the presence of HSCoM was determined by following the decrease of MCR PS at 420 nm and increase of MCR red1 at 385 nm. At HSCoM concentrations between 0.2 and 4 mM, ϳ80% conversion of MCR PS to MCR red1 was achieved (data not shown). The remaining 20% of MCR PS probably decays into MCR silent , which is indistinguishable from MCR PS by absorption spectroscopy but is EPR-silent. The rate    6). BPS and propanesulfonate were detected by reverse phase HPLC connected to a conductivity detector. The rate constants and the amplitudes were determined by fitting the data to single-exponential equations. There is no decay of BPS when MCR is excluded from the reaction mixture (triangle).
of MCR PS conversion to MCR red1 was dependent on the concentration of HSCoM with a second order rate constant of 3.99 M Ϫ1 s Ϫ1 (Fig. 8, inset), which is about 40,000-fold slower than the second order rate constant for MCR PS formation (1.6 ϫ 10 5 M Ϫ1 s Ϫ1 ). This large difference between the second order rate constants for the formation of MCR PS versus "reactivation" by forming MCR red1 was crucial in our ability to characterize MCR PS as an intermediate in PS formation. This rate mismatch also is what led to the earlier assignment of BPS as an irreversible inhibitor. The slow rate of reactivation of MCR PS to MCR red1 also indicates that the steady-state rate of PS formation will be very slow (as is shown below), because the multiple turnover reaction will be limited by the reactivation rate constant.
Ti(III) citrate is less efficient than CoMSH in converting MCR PS to MCR red1 . Within 10 min at pH 10, Ti(III) citrate converted only ϳ5% MCR PS to MCR red1 , whereas ϳ83% MCR PS converted to an EPR-silent form of the enzyme, and ϳ12% remained as MCR PS (Fig. 9). At pH 7.6, MCR PS was unaffected by Ti(III) citrate over the same period of time (data not shown).
The HSCoM-dependent conversion of MCR PS to MCR red1 is pHdependent, showing no conversion at pH 7.2 (data not shown). In an experiment otherwise identical to the one described in Fig. 8, 4 mM MeSCoM was used in place of HSCoM; under these conditions, MCR PS was not reactivated to MCR red1 (data not shown), demonstrating the importance of the thiol group of HSCoM for reactivation. A survey of thiols has not been performed; however, dithiothreitol and mercaptoethanol were found to reactivate MCR PS .
As summarized in Schemes 1 and 2, these results indicate that BPS is not a competitive inhibitor or irreversible inactivator but rather a reversible redox inactivator of MCR (reversible by HSCoM) that can serve as an alternative substrate and that MCR PS is a more oxidized state than MCR red1 , which can be converted back to the MCR red1 state by thiolate nucleophiles or, less efficiently, by Ti(III) citrate.

DISCUSSION
BPS has been described as a highly potent competitive inhibitor (apparent K i ϭ 50 nM) (41) and as an irreversible inhibitor (44) of MCR, and its high affinity has been exploited to titrate the active sites of this enzyme (11,47). Here we clarify the reaction of the active MCR red1 state with BPS by a combination of spectroscopic and kinetic studies. As shown in Schemes 1 and 2, this reaction occurs by the nucleophilic attack of Ni(I) on BPS to displace bromide and generate the EPR-active MCR PS species. MCR PS (formerly called MCR BPS ) was described by Rospert et al. (28) and assigned as a high spin Ni(II)-alkylsulfonate radical species (29,48). The lack of detectable hyperfine broadening from the halide of BPS (or any of the related compounds shown in Table 1 that give the MCR PS signal) provided evidence that the bromide undergoes elimination before or as the MCR PS state is formed. In fact, identical EPR spectra (MCR PS ) are formed by incubating MCR red1 with 3-bromo-(I ϭ 3/2), and 3-iodo-(I ϭ 5/2) propanesulfonate with no  detectable hyperfine splitting from these halide atoms, indicating that the nickel is not positioned close to the halogen groups of these ligands (28). Advanced EPR studies have defined the hyperfine coupling constants between the Ni(III) center and the alkyl ligand; however, in this complex, 75% of the spin is in the nickel d x 2 Ϫ y 2 orbital and ϳ7% on the attached methylene carbon atom (48). These values are in reasonable agreement with a DFT calculation on a methyl-Ni(III)-F 430 model (49). 6 In the studies described here, MCR red1 was reacted with BPS, and the propylsulfonate ligand was unambiguously identified (after acid quenching) as propanesulfonate by NMR spectroscopy and by HPLC analysis in both single and multiple turnover reactions. This experiment would not per se rule out the possibility that the various halopropane sulfonates ligate to nickel via the sulfonate oxygen(s), especially since similar MCR PS -like EPR spectra are generated when MCR red1 is incubated with bromobutanesulfonate and chloropropane sulfonate (Table 1). However, other sulfonates, including 3-mercapto-1-propanesulfonic acid, butanesulfonate, and propanesulfonate, do not induce MCR PS -like EPR signals. Furthermore, the reaction of MCR red1 with 4-bromobutyric acid (29) or 4-chlorobutyric acid generates an EPR signal nearly identical to MCR PS . Therefore, the sulfonate group is not a strict requirement for generation of the MCR PS signal, only its anionic character. In addition, the advanced EPR studies described above strongly indicate the formation of an organometallic alkyl-nickel species (49).
Besides its intermediacy in PS formation from BPS, MCR PS is interesting in sharing some striking similarities to the MCR ox1 state. These two states are functionally similar in that they can be activated to the MCR red1 state 7 ; however, it has so far proven impossible to activate other Ni(II) states of MCR. The "ready" nature of the MCR PS and MCR ox1 states must be related to their electronic structures, which have both been assigned as Ni(II) associated with a radical, a thiyl radical in the case of MCR ox1 (15,17) and an alkylsulfonate radical for MCR PS (29). The UVvisible spectra of these two species are indistinguishable, indicating that MCR ox1 and MCR PS share the same nickel oxidation state. The EPR spectra are similar (for MCR ox1 , the g values are 2.231, 2.153, and for MCR PS , they are 2.223, 2.115), which would be consistent with the two species sharing the same oxi-dation state but different coordination states. Since the four planar nitrogens of the macrocycle are unlikely to vary, the spectral differences must result from changes in the upper axial ligands: a thiyl radical for MCR ox1 versus an alkyl radical in the case of MCR PS .
In the presence of the low potential reductant Ti(III) citrate, the MCR PS state undergoes reductive activation and protonolysis to form MCR red1 and propanesulfonic acid. Some thiolates like HSCoM, dithiothreitol, and mercaptoethanol also react with MCR PS to regenerate MCR red1 and presumably a thioether product, whose characterization is under way. Thus, BPS can serve as an alternative substrate for MCR. In these reactions, the rate of MCR PS formation (k max /K m ϭ 190 mM Ϫ1 s Ϫ1 , k max ϳ17 s Ϫ1 at 20°C) is ϳ60-fold faster than the rate of methane formation from the natural substrates, methyl-SCoM and HSCoB (k cat /K m ϭ 3 mM Ϫ1 s Ϫ1 at 20°C, 50 mM s Ϫ1 at 60°C (23)). However, the elimination of PS and regeneration of MCR red1 occurs 1000-fold more slowly than methane formation from the natural substrates. The large difference between the rates of MCR PS formation and regeneration of MCR red1 explains the nearly stoichiometric accumulation of MCR PS and rationalizes why BPS acts as such a strong inhibitor of MCR.
Schemes 1 and 2 rationalize the seemingly disparate properties of BPS, which have led to its classification as a competitive inhibitor (41), as an irreversible inhibitor (44), and, as described here, as an alternative substrate and a redox inactivator. Based on steady-state kinetics, BPS would appear to be a competitive inhibitor, because the active enzyme (MCR red1 ) can react either with BPS or methyl-SCoM and because there is a path from MCR PS back to active MCR red1 . Therefore, in comparing the competitive inhibition Mechanism A with the redox inactivation/reactivation Mechanism B (Scheme 2), as long as the reactivation of MCR red1 is rapid and complete (i.e. the enzyme rapidly and quantitatively returns to its active state), one would not suspect the occurrence of a series of complex underlying reactions. However, by following enzyme-monitored turnover reactions, it is clear that BPS is not a traditional competitive inhibitor in a simple rapid equilibrium with free enzyme nor an irreversible inactivator, because there is a redox pathway back to the active enzyme form.
These results demonstrate conclusively that the MCR PS state is a catalytically competent intermediate in propanesulfonate formation and constitute the first spectroscopic observation of any intermediates in the MCR reaction that can be related to catalysis. The results could thus be considered to support the intermediacy of an alkylnickel intermediate in MCR-catalyzed methane formation from methyl-SCoM. In the rest of this discussion, the results are discussed in relation to Mechanisms I and II.
With respect to Mechanism I, the formation of the MCR PS intermediate at rates exceeding the rate of methane formation from methyl-SCoM indicates the possibility that an alkylnickel intermediate is involved in methanogenesis. Furthermore, any reaction mechanism must exhibit the property of microreversibility; therefore, as shown in Fig. 11, our results provide evidence for an alkylnickel intermediate in the final step of anaerobic methane oxidation. According to Mechanism I (Fig. 11A), this step involves the thiolytic cleavage of an alkylnickel bond to form the thioether, methyl-SCoM, and regenerate MCR red1 . Then proton transfer to the methyl group to form methane, followed by electron transfer to Ni(III) to generate Ni(II), would generate a disulfide anion radical, which is considered to be sufficiently reducing to convert Ni(II) back to the active MCR red1 state.
On the other hand, the relationship of the alkylnickel intermediate to Mechanism I deserves a few words of caution, since the properties of methyl-SCoM and BPS are quite different. Although the strength of the CH 3 -SCoM bond (293 kJ/mol (27)) is similar to that of the C-Br bond in BPS (ϳ270 kJ/mol), neither the methyl group (if Ni(I) attacked the thioether sulfur) nor the mercaptoalkylsulfonate group (CoMS Ϫ ) (if nickel attacked the methyl group) of methyl-SCoM is a good leaving group. However, bromide is an excellent leaving group. The standard measure of the goodness of a leaving group is the pK a of its conjugate acid (HBr versus CH 4 ); the lower the pK a , the better the leaving group ability. Methane is one of the weakest acids known, with a pK a of 50, whereas HBr is a very strong acid, with a pK a of Ϫ9. HSCoM would have an intermediate pK a of around 8. Therefore, an ionic reaction of MCR red1 with BPS to eliminate bromide and form Ni(III)alkylsulfonate is quite reasonable; however, it is unlikely that Ni(I) reacts with methyl-SCoM to eliminate a methyl anion and form Ni(III)-SCoM or to eliminate Ϫ SCoM and form methyl-Ni(III). Furthermore, computational results suggest that cleavage of a high energy methyl-SCoM (ϳ70 kcal/ mol) bond to form a weak methylnickel (less than 25 kcal/mol) bond is thermodynamically unreasonable, and a radical mechanism (Fig.  11C) was proposed that involves reaction of Ni(I) with the sulfur of methyl-SCoM to generate a Ni(II)thiolate and a methyl radical, which abstracts a hydrogen atom from HSCoB (27,50).
The role of HSCoB in the MCR reaction was assessed by single turnover kinetic studies using the chemical quench technique with radioactive [ 14 C-methyl]SCoM as the substrate. These transient kinetic studies demonstrated that formation of even a single equivalent of methane requires HSCoB (23). This reaction was repeated with an analog of HSCoB that is one carbon shorter in length (mercaptoHEXanoyl-threonine phosphate). A single exponential decay was observed in both reactions that depended on the concentration on HSCoB or the analog, which was shown to react 1000-fold slower than HSCoB. These results indicate that HSCoB is not required for proton donation, since the acidic conditions of the quench would have supplanted that role by protonolysis of the methylnickel intermediate to form methane and indicate that HSCoB has an integral role in a step(s) that promotes C-S bond cleavage. Conversely, with respect to the methyl radical mechanism, the results do not rule out a requirement for HSCoB in hydrogen atom abstraction or in promoting a conformational change that might alter the reaction pathway and/or the regiospecificity of the attack of Ni(I) on methyl-SCoM.  HSCoM is firmly held in the active site of MCR by interactions of the thiol group with nickel, one of the sulfonate oxygens with a backbone nitrogen in the ␣ subunit, and another sulfonate oxygen with Arg-120 of the ␥ subunit (Fig. 10). The interactions of the sulfonate oxygens greatly influence the type of substrate that can react with MCR. When the structures of the various sulfonate and carboxylate analogs shown in Table 1 are compared, it appears that generation of the MCR PS signal requires a good leaving group adjacent to an electrophilic carbon atom that is four bonds (ϳ4.8 Å) from the negatively charged oxygen of a carboxylate or sulfonate. When this distance is increased to ϳ6 Å with BBS, the second order rate constant for reaction of MCR red1 with the bromoalkanesulfonate decreases by 70-fold. Interestingly, the electrophilic carbon in BPS is in the same position as the thioether sulfur of methyl-SCoM and the bromine group of BPS is located at approximately the same position as the methyl group of methyl-SCoM. These results indicate that, for the natural substrate, steric considerations would favor attack of Ni(I) at the sulfur of methyl-SCoM. However, as mentioned above, it is possible that binding of HSCoB could alter the stereoselectivity of the active site.
By strict analogy with the BPS reaction and as shown in Fig.  11B, in which attack of Ni(I) on C-3 of BPS followed by elimination of the bromide would generate the MCR PS state, one might expect Ni(I) to react with the thioether sulfur of methyl-SCoM to generate a Ni(III)-thiolate and eliminate a methyl anion. However, the poor leaving group properties of a methyl group as discussed above would prevent such a reaction. On the other hand, the attack of Ni(I) on the thioether sulfur of methyl-SCoM to form a Ni(III)-thiolate and release a methyl radical, as described in Fig. 11C, would retain the steric constraints of the reaction, based on the substrate profile described in the legend to Fig. 1. This step, leading to the formation of a relatively strong Ni(III)-thiolate bond and release of a methyl radical, was proposed to be feasible, with a barrier of ϳ17-20 kcal/mol 8 (instead of ϳ45 kcal/mol for generation of the methyl-Ni) (27). Hydrogen atom abstraction from HSCoB by the methyl radical to form methane was predicted to have a very small barrier of ϳ1 kcal/mol (27).
As shown in Fig. 11C, the reaction of MCR PS with a thiolate to form the thioether and regenerate MCR red1 is similar to the proposed reaction of the CoBS radical with the Ni(III)-thiolate to generate the disulfide anion radical and a Ni(II) form of MCR (Fig. 11C). Electron transfer from the disulfide anion radical to Ni(II) would form the disulfide and Ni(I)-MCR red1 . Interestingly, in methane formation, cleavage of the C-S bond of methyl-SCoM appears to be rate-limiting, whereas in the BPS reaction, the first step (elimination of bromide) is rapid, and regeneration of MCR red1 is rate-limiting.
In conclusion, the results described here are relevant to the initial step in methane formation and to the final step in anaerobic methane oxidation. Further studies involving analogs of methyl-SCoM and HSCoB are under way to uncover details of the MCR catalytic mechanism and to discriminate between the two proposed mechanisms for this unique nickel enzyme.