All buffers, media ingredients, and other reagents were acquired from Sigma-Aldrich (St. Louis, MO, USA) and, unless otherwise stated, were of the highest purity available. Solutions were prepared using Nanopure deionized water. N₂ (99.98%), CO (99.99%), argon (99.8%), H₂/CO₂ (80%/20%), and ultrahigh purity H₂ (99.999%) were obtained from cryogenic gases (Grand Rapids, MI, USA). 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 wt% solution in 2 N hydrochloric acid; Acros Organics, Morris Plains, NJ, USA) under anaerobic conditions and adjusting the pH to 7.0 with sodium bicarbonate (Zehnder and Wuhrmann, 1976). The concentration of Ti(III) citrate was determined routinely by titrating a methyl viologen solution. M. marburgensis (formally M. thermoautotrophicum strain Marburg) was obtained from the Oregon Collection of Methanogens catalog as OCM82. M. marburgensis was cultured on H₂/CO₂ (80%/20%) at 65°C in a 14-L fermentor (New Brunswick Scientific Co. Inc., New Brunswick, NJ, USA). Culture media were prepared as previously described (Kunz et al., 2006) with a slight modification of the sulfur and reducing source, by adding 50 mM sodium sulfide (instead of H₂S gas) at a flow rate of 1 ml min^-1 during the entire growth period.

Solutions were prepared and all manipulations were performed under strictly anaerobic conditions in a Vacuum Atmospheres (Hawthorne, CA, USA) anaerobic chamber maintained under nitrogen gas at <1 ppm of oxygen (monitored continually by a Teledyne detector). Cells were grown under H₂/CO₂ as described (Dey et al., 2010a) and, before harvest, the cells were treated with H₂ for 30 min in the fermenter (different gases were used in the experiments shown in Figure 1D). Then, the cells were harvested, transferred to the anaerobic chamber, and resuspended in 50 mM Tris, pH 7.6, 10 mM HSCoM, 0.1 mM Ti(III) citrate, and aliquoted into 150 ml serum-stoppered anaerobic bottles, as described (Dey et al., 2010a; Li et al., 2010). The headspaces of the vials containing the resuspended cells were then purged with CO or H₂ for 10 min, or at timed intervals right before samples were taken for analysis. This second (and all subsequent) purge was also omitted for the formate-activation experiment as described in the next paragraph to ensure that the gas treatment did not interfere with the formate incubation. To facilitate equilibration of the gas with the solution, the vials were shaken every 2 min. In most experiments, the cell suspension was transferred into smaller vials, where the headspace was purged again for 10 min with the test gas. Purges with H₂ were done inside of the chamber, while, for safety reasons, purges with CO were done in the fume hood outside of the chamber. To remove trace amounts of oxygen in the H₂ or CO line for aerating the serum vials, the line was fitted with a high pressure Oxy-Trap oxygen scrub (Alltech, Deerfield, IL, USA) followed by an indicating Oxy-Trap (Alltech) to monitor the efficiency/remaining capacity of the oxygen scrub. Between the purges, the serum vials were inverted in a water bath in the anaerobic chamber to minimize any gas escape. For monitoring the conversion of MCR to the MCR_red1 state, 200 μl samples were removed at different time points to be analyzed by EPR spectroscopy.

To determine whether formate activates MCR, the same protocol outlined above was followed (including the first 30 min H₂ purge in the fermenter), except that sodium formate was added directly to the cell suspension to a final concentration of 100 or 500 mM (pH did not change in the Tris buffer that is described above for resuspending cells). Thus, there was only one gas purge in this experiment. To inhibit carbon monoxide dehydrogenase (CODH) activity, NaCN was added to the cell suspension to a final concentration of 0.2 or 0.4 mM before aerating with CO. Then, the cell suspension was incubated as described above.

UV-visible spectra of MCR were recorded in the anaerobic chamber using a diode array spectrophotometer (model DT 1000A, Analytical Instrument Systems, Inc., Flemington, NJ, USA). All samples for EPR spectroscopy were prepared in 50 mM Tris–HCl buffer, pH 7.6, in a Vacuum Atmospheres anaerobic chamber maintained under nitrogen gas at <1 ppm of oxygen. EPR spectra were recorded on a Bruker EMX spectrometer (Bruker Biospin Corp., Billerica, MA, USA), 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 were: temperature, 70 K; microwave power, 10 mW; microwave frequency, 9.43 GHz; receiver gain, 2 × 10⁴; modulation amplitude, 10 G; and modulation frequency, 100 kHz. Double integrations of the EPR spectra were performed and referenced to a 1-mM copper perchlorate standard.

The purification of MCR_red1 from M. marburgensis was performed under strictly anaerobic conditions, as described previously (Kunz et al., 2006). The amount of MCR_red1 was calculated according to the UV-visible and EPR spectra (Kunz et al., 2006; Li et al., 2010). Protein concentration was determined by Rose Bengal assay (Elliott and Brewer, 1978) with lysozyme as the standard.

Methyl-coenzyme M reductase assays for methane formation from methyl-CoM and CoBSH were performed at 60°C, basically as described by Kunz et al. (2006). The standard assay mixture contained 10 mM ¹⁴CH₃-SCoM, 0.1 mM CoBSH, 1.8 mM aquacobalamin, 20 mM Ti(III) citrate, and 0.5 M Tris (pH 7.2) in a final volume of 0.2 ml. The reaction was started by adding MCR. MCR activity was measured by following the time-dependent loss of radioactivity from the methyl group of ¹⁴CH₃-SCoM, which is converted to highly volatile and insoluble [¹⁴C] methane (Kunz et al., 2006). Rates of methane formation were calculated from the linear portion of the time course. One unit of MCR activity is equal to 1 μmol of methane produced per minute.

To determine formate dehydrogenase (FDH), hydrogenase, and CODH activities, M. marburgensis cell extracts were prepared under anaerobic conditions as described (Kunz et al., 2008). All assays were performed in a 1-ml sealed solution at 60°C. The anaerobic assay mixture contained 20 mM methyl viologen, 50 mM Tris, pH 7.5, and cell extract. For the hydrogenase and CODH assays, oxygen-free hydrogen gas or carbon monoxide (99.99%), respectively, was bubbled into the assay mix in a serum-stopped cuvette for 5 min (~1 mM of each gas in the solution) before enzyme was added to start the reaction. In the FDH assay, 20 mM formate was used and either 20 mM methyl viologen or 10 mM NADP⁺ was included as an electron acceptor. The activities were determined by following the reduction of methyl viologen at 578 nm with an extinction coefficient of 9.78 mM^-1 cm^-1 or the reduction of NADP⁺ at 340 nm with an extinction coefficient of 6.27 mM^-1 cm^-1.

To date, the most effective way to activate MCR has been to incubate the cell suspension under a H₂ atmosphere (Rospert et al., 1991a; Kunz et al., 2006). While this is effective for generating MCR_red1 from organisms like M. marburgensis, for some methanogens, like Methanosarcina acetivorans, H₂ activation is inefficient because the hydrogenases are weakly expressed (Guss et al., 2009; Wang et al., 2011). Therefore, there must be other pathways for in vivo activation of MCR. We hypothesized that the low-potential reductant, CO, with a standard reduction potential (vs. NHE for the CO₂/CO couple) of -520 mV (in comparison to -420 mV for that of 2 H⁺/H₂; Lehn and Ziessel, 1982) should be able to activate MCR.

To measure the effectiveness of CO in activating MCR, we purged the M. marburgensis culture for 30 min with H₂ in the fermentor, harvested the cells, resuspended the cells in buffer and incubated the cell suspensions at 32°C in a water bath under a CO or H₂ atmosphere. The whole-cell EPR spectra of the H₂- and CO-treated cells at different time points after initiating H₂ or CO gas purge are compared in Figures 1A,B, respectively. The activity of MCR is linearly related to the MCR_red1 EPR signal, with g values at 2.24 and 2.065; thus EPR spectroscopy is a convenient method to assess the level of MCR activation achieved. It also reveals various other EPR-active forms of MCR that are present during the treatment – for example, MCR_ox1, which is an oxidized form of MCR that has a distinct EPR spectrum (g values at 2.23 and 2.16), can be converted into the MCR_red1 state by the addition of low-potential reductants like Ti(III) citrate, and MCR_red2, a Ni(I) state that is formed when the enzyme in the MCR_red1 form is treated with HSCoM and CoBSH (Mahlert et al., 2002). Therefore, we feel that assessing the level of MCR_red1 is the most accurate indicator of the level of activation achieved. The spectra of the cell suspension before the CO/H₂ purge reveal about two to three times more MCR_ox1 than MCR_red1, very little MCR_red2, and the remainder in the EPR-silent Ni(II) form. For the H₂-purged cells, after ~1 h of incubation, MCR_ox1 disappeared as MCR_red2 increased; then, over the next ~20 h, MCR_red1 appeared. Following a ~10 h lag time, the rate constant for H₂-dependent activation was 8 min^-1. The amount of MCR_red1 formed at the end of activation, when compared with the amount of MCR_ox1 in the beginning, varied slightly among enzyme preparations.

In contrast, when cells were incubated with CO during the timed purge, within 1 h, MCR underwent rapid activation to the MCR_red1 state without a noticeable lag time or intermediacy of MCR_red2. The enhanced activation by CO was noticeable by eye – the cell suspension began turning green immediately after aerating with CO. However, for H₂-dependent activation, this color development required overnight incubation. At the end of activation (when the whole-cell EPR signal reached maximum intensity), the CO-activated cell suspension was also greener than that of the H₂-activated cells (data not shown). Quantitative analysis of the activation confirmed the visual observations. The rate constant for CO-dependent activation was 130 min^-1, which is over 15-fold faster than that observed for activation by H₂; furthermore, the amount of accumulated MCR_red1 was ~twofold higher than with H₂ (Figure 1C). Furthermore, with CO activation, the amount of MCR_red1 formed at the end of activation was approximately two times more than the initial amount of MCR_ox1, indicating that CO activation converts the MCR_silent form to MCR_red1.

As described above, the CO-dependent activation experiments shown in Figures 1B,C involved an initial 30 min H₂ purge in the fermentor. To test whether the initial H₂ treatment is required, we purged the cell suspension with pure H₂, CO, or CO₂ in the fermentor, or omitted the initial 30-min purge before harvesting, and then bubbled the cell suspension with CO after resuspending the cells in Tris–HCl buffer as above. The initial EPR spectra significantly differed with different purge conditions. MCR_ox1 was only observed when the cells were initially purged with H₂ in the fermentor, while MCR_red1 was only noticeable with the CO treatment or when the initial H₂ purge was omitted. On the other hand, MCR appears to be completely in a Ni(II)-silent form when cells were initially purged with CO₂, because no detectable MCR signal was present. However, MCR underwent activation in all these conditions (i.e., H₂, CO, CO₂, no initial purge) after treating the cell suspensions with CO, and the activation rates and the final levels of activation were similar (Figure 1D). Thus, an initial H₂ purge is not required for CO-dependent activation of MCR. This experiment also clearly shows that the EPR-silent Ni(II)-MCR forms undergo conversion to MCR_red1 by CO in vivo.

We also tested the effect of addition of CO to the H₂-purged cells. Figure 1E shows that treatment with CO can further activate MCR in H₂-purged cells to nearly reach the level of MCR_red1 observed in CO-activated cells. On the other hand, treatment with H₂ does not increase the MCR_red1 level in CO-activated cells. These experiments indicate that H₂- and CO-dependent activations pathways converge after the initial transfer of electrons to the acceptor of reducing equivalents from hydrogenase or CODH.

H₂-dependent activation of MCR is very sensitive to the incubation temperature; it is most effective at 30–33°C (Kunz et al., 2006). To test whether CO-dependent MCR activation is also temperature-dependent, we incubated the cell suspension with CO at different temperatures and followed the whole-cell EPR spectrum over 12 h. As shown in Figure 2, CO-dependent activation, like the H₂-dependent process, is most effective at ~30°C. The complicating factor in these experiments is the inactivation of MCR, which occurs markedly faster as the temperature increases (Horng et al., 2001). At high temperatures (40–60°C), MCR_red1 decays after 2 h of CO-dependent activation, while, at lower temperatures (10–20°C), activation by CO occurs more slowly, but, because MCR_red1 decays more slowly, this form of the enzyme is more stable and accumulates to higher levels.

To test whether CODH activity is involved in CO-dependent activation of MCR, CN^-, which is a competitive slow-binding inhibitor of CODH with a K_i for CODH inhibition in the low micromolar range (Ragsdale et al., 1983; Ha et al., 2007), was added to the cell suspension. As shown in Figure 3, 0.2 mM CN^- blocked the CO-dependent activation of MCR, indicating that CODH activity is required for CO-dependent activation of MCR. In contrast, 0.2–0.4 mM CN^- does not block H₂-dependent activation of MCR (data not shown). We measured a specific activity of 3–5 U mg^-1 of CODH in cell extracts, which is significantly lower than that of hydrogenase (~80 U mg^-1), but is sufficient to drive the activation of MCR.

Methyl-coenzyme M reductase was purified from cells that had been undergone a 5-h incubation with CO. Based on the UV-visible (Figure 4A) and EPR spectra (Figure 4B) of the purified enzyme, MCR was about ~80% active. Based on the enzymatic assay for the conversion of methyl-SCoM to methane, the specific activity of the CO-activated enzyme was calculated to be 89 ±10 U mg^-1, indicating 89% conversion to the active MCR_red1 state. Thus, the CO-dependent activation is highly efficient in producing a stable Ni(I) form of the enzyme.

The standard redox potential of the CO₂/HCOOH half-cell (-430 mV, Eq. 2) is equivalent to that of the 2H⁺/H₂ half-cell (Eq. 3; Scherer and Thauer, 1978). Furthermore, the FDH gene (fdh) is present in the M. marburgensis genome and, based on sequence identity, is thought to use F₄₂₀ as an electron donor (Liesegang et al., 2010; Kaster et al., 2011a). Thus, we expected that formate could also activate the M. marburgensis MCR. However, as shown in Figure 5, we did not observe any increase in the level of MCR_red1 upon incubation of the cell suspension for over 20 h with formate. Correspondingly, we measured only a very low specific activity of FDH in cell extracts of M. marburgensis (3–4 mU mg^-1, using NADP⁺ as electron acceptor and 2–3 mU mg^-1, using MV), which is consistent with earlier reported values of 3–8 mU mg^-1, depending on the electron acceptor and the pH of the assays (Tanner et al., 1989). We assayed using both NADP⁺ and MV, which is an electron acceptor for the F₄₂₀-dependent enzyme, which the M. marburgensis enzyme is predicted to be (Kaster et al., 2011a). There are two forms of FDH: a tungsten- and a molybdenum-containing enzyme. The growth medium that we used has 1 μM molybdenum, but no tungsten (Kunz et al., 2006); however, addition of 1 μM tungsten and 1.1 μM selenium to the growth medium for M. marburgensis, also did not lead to activation of MCR in the presence of formate (data not shown). Apparently, the low level of FDH activity present in M. marburgensis is not sufficient to drive activation of MCR.