The role of zinc in the methylation of the coenzyme M thiol group in methanol:coenzyme M methyltransferase from Methanosarcina barkeri New insights from X-ray absorption spectroscopy Markus Kru¨er 1 , Michael Haumann 2 , Wolfram Meyer-Klaucke 3 , Rudolf K. Thauer 1 and Holger Dau 2 1 Max-Planck-Institut fu ¨ r terrestrische Mikrobiologie and Laboratorium fu ¨ r Mikrobiologie, Fachbereich Biologie der Philipps- Universita ¨ t, Marburg, Germany; 2 Freie Universita ¨ t Berlin, Fachbereich Physik, Berlin, Germany; 3 DESY, EMBL Outstation, Hamburg, Germany Methanol:coenzyme M methyltransferase from methano- genic archaea is a cobalamin-dependent enzyme composed of three different subunits: MtaA, MtaB and MtaC. MtaA is a zinc protein that catalyzes the methylation of coenzyme M (HS-CoM) with methylcob(III)alamin. We report zinc XAFS (X-ray absorption fine structure) results indicating that, in the absence of coenzyme M, zinc is probably coordinated by a single sulfur ligand and three oxygen or nitrogen ligands. In the presence of coenzyme M, one (N/O)-ligand was replaced by sulfur, most likely due to ligation of the thiol group of coenzyme M. Mutations in His237 or Cys239, which are proposed to be involved in ligating zinc, resulted in an over 90% loss in enzyme activity and in distinct changes in the zinc ligands. In the His237 fi Ala and Cys239 fi Ala mutants, coenzyme M also seemed to bind efficiently by ligation to zinc indicating that some aspects of the zinc ligand environment are sur- prisingly uncritical for coenzyme M binding. Keywords: 11 zinc enzymes; methanogenic archaea; methyl transferases; thiol group alkylation; EXAFS. Methanosarcina barkeri and other Methanosarcina species can grow on methanol as carbon source which is dispro- portionated to CH 4 and CO 2 [1]. The first step in this metabolic pathway is the formation of methyl-coenzyme M (CH 3 -S-CoM) from methanol and coenzyme M (HS-CoM) [2]. CH 3 OH þ HS-CoM ! MtaABC CH 3 -S-CoM þ H 2 O ð1Þ DG°¢ ¼ )27.5 kJÆmol )1 The reaction is catalyzed by methanol:coenzyme M methyltransferase which is composed of the three subunits MtaA (35.9 kDa), MtaB (50.7 kDa) and MtaC (27.9 kDa), of which MtaC is a corrinoid protein. They catalyze the following partial reactions [3–7]. CH 3 OH þ MtaC ÀÀ* )ÀÀ MtaB CH 3 -MtaC þ H 2 O ð1aÞ DG°¢ ¼ )7kJÆmol )1 CH 3 -MtaCþHS-CoM ÀÀ* )ÀÀ MtaA CH 3 -S-CoMþMtaC ð1bÞ DG°¢ ¼ )20.5 kJÆmol )1 MtaA is a zinc protein [3,7,8] that also catalyzes the methylation of coenzyme M with methylcob(III)alamin [9]. Several isoenzymes of MtaA, designated MtbA and MtsA have been found [9–11]. The methylation of coenzyme M to methyl-coenzyme M is a reaction in which a thiol group is alkylated. Enzymes catalyzing alkyl transfers to thiols have all been found to be zinc proteins [12]. They include the E. coli Ada protein [13], the cobalamin-dependent methionine synthase MetH [14], the cobalamin-independent methionine synthase MetE [15], betaine:homocysteine S-methyltransferase [16], S-methyl- methionine:homocysteine methyltransferase [17], epoxy- alkane:coenzyme M transferase [18] and protein farnesyl transferase [19]. The postulated role of zinc in these enzymes is that of a Lewis acid that activates the thiol group to be alkylated. Coordination of the thiol group to the active site zinc has been shown by extended X-ray absorption fine structure (EXAFS) spectroscopy [14,15], by UV spectros- copy [20] and in the case of protein farnesyl transferase by crystal structure analysis [19]. It results in a decrease in the pK value of the thiol group as shown by the release of a proton upon binding of the substrate to the zinc enzyme [21]. MtaA does not share sequence similarity to any of the other zinc enzymes catalyzing thiol group alkylation [8,22]. Correspondence to H. Dau, Freie Universita ¨ t Berlin, Fachbereich Physik, Arnimallee 14, D-14195 Berlin, Germany. Fax: + 49 30 838 56299, Tel.: + 49 30 83853581, E-mail: holger.dau@physik.fu-berlin.de or R. K. Thauer, Max-Planck-Institut fu ¨ r terrestrische Mikrobiologie, Karl-von-Frisch-Strasse, D-35043 Marburg, Germany. Fax: + 49 6421 178209, Tel.: + 49 6421 178200. Abbreviations: HS-CoM, coenzyme M; CH 3 -S-CoM, methyl-coen- zyme M; EXAFS, extended X-ray absorption fine structure; Mta, methanol:coenzyme M methyltransferase; MtaA, protein subunit of Mta; XANES, X-ray absorption near edge structure; XAS, X-ray absorption spectroscopy. (Received November 2001, revised 15 February 2002, accepted 28 February 2002) Eur. J. Biochem. 269, 2117–2123 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.02860.x In the sequence, however, the motif HXCX n C is found whichinMetEhasbeenshowntobethezincbindingmotif [15]. It has been suggested that the thiol group of coenzyme M is activated by MtaA via the same mechanism as proposed for the other zinc enzymes [7]. This suggestion is mostly based on the nding that MtaA contains per mol one mol of zinc and that upon binding of coenzyme M one mol of H + is released [7]. The zinc EXAFS results reported here indeed now show an increase of one in the number of sulfur ligands to zinc upon formation of the MtaA- coenzyme M complex. After completion of this manuscript a paper was published (30 October 2001) describing zinc EXAFS results for the MtaA isoenzyme MtbA from M. barkeri and coming to the same conclusion with respect to the sulfur ligation upon binding of coenzyme M [23]. The isoenzyme is involved in methyltransfer from methylamines to coen- zyme M. CH 3 ị 2 NH ỵ 2 ỵ HS-CoM ! MtbABC CH 3 -S-CoM ỵ CH 3 NH ỵ 3 2ị DGÂ ẳ )2kJặmol )1 MtbA and MtaA show only 40% sequence identity [22] and cannot substitute for each other in the catalysis of reactions 1 and 2 [6,8,24,25]. MATERIALS AND METHODS M. barkeri strain Fusaro (DSM 804) was obtained from the Deutsche Stammsammlung fu ă r Mikroorganismen und Zellkulturen (Braunschweig). Methylcob(III)alamin, coen- zyme M and methylmethanethiosulfonate were purchased form Sigma, 4-(2-pyridylazo)resorcinol from Fluka. The QuickChange site-directed mutagenesis kit, Escherichia coli XL1 Blue MFR and Pfu polymerase were from Strategene. T4 DNA Ligase was from Roche. Oligonucleotides were obtained from MWG Biotech. Heterologous overproduction and purication of the MtaA proteins Heterologous overproduction and purication of the MtaA proteins His 6 tagged at the N-terminus was performed as described by Sauer & Thauer [7]. The mtaA wild-type gene was expressed heterologously in E. coli M15 and the mutated mtaA genes were expressed heterologously in E. coli XL1 Blue MFR. From 500 mL cell culture, approximately 10 mg highly puried MtaA was obtained. For X-ray absorption spectroscopy (XAS) the enzyme was equilibrated with 50 m M Mops/KOH pH 7.0 by Centricon ultraltration. Site-directed mutagenesis Site-directed mutagenesis was performed using the Quick- Change site-directed mutagenesis kit following the protocol given by the manufacturer. Two different MtaA mutants were made and conrmed by DNA sequencing: in the rst mutant, His237 was exchanged for Ala using the mutagenic primers 5Â-CCGTGACTGTACTCgcCATCTGTGGTAA GG-3Â (sense) and 5Â-CCTTACCACAGATGgcGAGTAC AGTCACGG-3Â (antisense); in the second mutant, Cys239 was exchanged for Ala using the primers 5ÂCGTGACTGT ACTCCACATCgcTGGTAAGGTTAACGC (sense) and 5ÂGCGTTAACCTTACCAgcGATGTGGAGTACAGTC ACG (antisense). The mutated bases are given in lower-case letters. The mutated mtaA genes were obtained by ampli- fying the expression plasmid pUH28 [22] harbouring mtaA using the respective primers. The resulting linear strands were treated with DpnI, which digested the nonmutated template DNA. The linear strands were ligated using T4 DNA ligase following the manufacturers instructions and subsequently transformed into E. coli XL1 Blue MFR which had been grown and induced as described above. The resulting plasmids were designated as pMK2 for the His237 Ala exchange and pMK5 for the Cys239 Ala exchange. Determination of specic activity Methylcob(III)alamin:coenzyme M activity was determined at 37 C and pH 7.0 by following the demethylation of methylcob(III)alamin (50 l M ) photometrically at 520 nm (De ẳ 6.3 m M )1 ặcm )1 ) [7]. One unit ẳ 1 lmol methyl- cob(III)alamin demethylated per min under the assay conditions used in this paper. (Note that 1 unit ẳ 20 lmol methylcob(III)alamin demethylated per min under the assay conditions described by Gencic et al. [23].) Protein was determined using the Bradford method and bovine serum albumin as standard [26]. Determination of zinc content The zinc content was in principle determined as described by Zhou et al. [15]. The zinc concentration was determined from the absorption change at 500 nm associated with a zinc complex formed with 4-(2-pyridylazo)resorcinol. With ZnCl 2 as a standard an De ẳ 55 m M )1 ặcm )1 was deter- mined. To avoid zinc contamination, all plastic tubes and pipette tips were rinsed with 10 m M EDTA and distilled H 2 Obeforeuse. XAS measurements and data analysis Zinc K-edge X-ray spectra were collected at beamline D2 of the EMBL Hamburg outstation (HASYLAB, DESY, Hamburg, Germany). The liquid samples were injected into 1-mm cuvettes with capton foil windows. Fluorescence- detected X-ray absorption spectra were measured at 20 K as described previously [27] (monochromator detuning to less than 60% of maximum intensity; scan range: 9400 10 700 eV). An absolute energy calibration was performed by monitoring the Bragg reections of a crystal positioned at the end of the beamline [28]. For each element of the used 13-element solid-state detector the total count-rate was kept below 30 000; the output signal was corrected for detector saturation. The spot size of the X-ray beam on the sample was 4.75 ã 1.13 mm 2 ; not more than four scans of 1-h duration were taken on the same spot of the sample. Comparison of the rst and fourth scan revealed no evidence for radiation damage to the samples. Ten to 12 scans were averaged for each EXAFS spectrum. Spectra were normalized and EXAFS oscillations were extracted as described in Schiller et al. [27]. The energy scale of all collected EXAFS spectra was converted to a k-scale 2118 M. Kru ă er et al. (Eur. J. Biochem. 269) ể FEBS 2002 using an E 0 of 9660 eV; k 3 -weighted spectra were used for curve-fitting and calculation of Fourier transforms. For the shown Fourier transforms, for k values ranging from 1.8 to 15.3 A ˚ )1 , the data was multiplied by a fractional cosine window (10% cosine fraction at low and high R-side); for curve-fitting the energy range was 20–900 eV. For simula- tion of k 3 -weighted spectra, complex backscattering ampli- tudes were calculated using FEFF 7 [29]; the used value of S 2 0 , the amplitude reduction factor, was 0.9. For the least- squares curve-fitting of unfiltered k space the in-house software SIMX was used. By curve-fitting of various EXAFS spectra we found consistently that DE 0 refined to a value of % 9665 eV; this value has been fixed and used in all simulations discussed in this work. For further details see figure captions. The Ôedge energyÕ used to describe the position of the absorption edge is the energy value corresponding to 0.65 units of normalized absorption. (The value of 0.65 was chosen because it corresponds to roughly half of the magnitude of the maximum absorption on top of the edge.) Other approaches for determination of edge energies resulted in similar edge-shifts. RESULTS Typical ligands to zinc atoms in proteins are nitrogen and oxygen atoms provided by the imidazole side chains of histidines and carboxylic side chains, respectively, as well as coordinated H 2 Ospecies(H 2 O, OH – ). Furthermore, the sulfur atoms of the thiol groups of cysteine and methionine are potential ligands. Distances between zinc and (N/O) (meaning nitrogen or oxygen) of 2.03–2.12 A ˚ and between zinc and sulfur of 2.25–2.36 A ˚ have been observed in proteins [14,15,23,30,31] and in synthetic zinc compounds [32,33]. EXAFS of wild-type MtaA Wild-type MtaA from M. barkeri had a zinc content of 0.91 mol/mol and exhibited a specific activity of 0.3 U per mg under our assay conditions. The enzyme was studied in the absence and presence of coenzyme M (apparent K m ¼ 40 l M )byXASatthezincK-edge.EXAFSand XANES spectra are shown in Figs 1 and 2, respectively. The Fourier transforms of the EXAFS spectrum of wild- type MtaA (Fig. 1A-I) reveals two prominent and closely spaced peaks at reduced distances of about 1.6 A ˚ (peak 1) and 1.9 A ˚ (peak 2). Peaks at these reduced distances are indicative of zinc-ligand distances of % 2.0 and % 2.3 A ˚ , suggesting the presence of oxygen or nitrogen ligands around 2.0 A ˚ and sulfur ligands at about 2.3 A ˚ . Addition of coenzyme M resulted in a slight decrease of Peak 1 and strong increase of Peak 2 (Fig. 1A-II,III). Thus, already the visual inspection of the EXAFS data suggests that the number of sulfur atoms ligated to the zinc of MtaA increases upon formation of the MtaA–coenzyme M com- plex explainable by ligation of the thiol sulfur of coen- zyme M to the active-site zinc. We simulated the experimental spectra in Fig. 1A using two shells of backscattering ligands, namely one shell of N or O ligands and one shell of sulfur ligands to zinc. (In general a discrimination between N and O ligands by EXAFS analysis 22 is almost impossible because (a) the typical bond-lengths are similar and (b) the EXAFS oscillations of these two Ôlow-ZÕ atoms are of similar shape.) Without using any constraints, a fit of the spectrum of wild-type MtaA without coenzyme M resulted in coordination numbers of 3.4 and 0.7 for the (N/O)-ligand and the sulfur ligands, Fig. 1. Fourier transforms (FT) of k 3 -weighted EXAFS spectra of wild- type MtaA and of two MtaA mutants (Cys239 fi Ala, His237 fi Ala) in the absence and presence of coenzyme M. Experimental spectra (thin lines) and simulations (thick lines) are shown. (A) The wild-type en- zyme in the absence (trace labeled by I) and presence of 0.5 mol/mol HS-CoM (II) or 2 mol/mol HS-CoM (III). (B) In the absence of HS-CoM, comparison of wild-type (I) and mutants (IV – Cys239 fi Ala, V – His237 fi Ala). (C) In the presence of 2 mol/mol HS-CoM, comparison of wild-type (III) and mutants (VI – Cys239 fi Ala, VII – His237 fi Ala). The insets show the corresponding EXAFS oscilla- tions in the k-space (same top-to-bottom sequence as used for the Fourier-transformed spectra). Ó FEBS 2002 Zn XAFS of methanol:coenzyme M methyltransferase (Eur. J. Biochem. 269) 2119 respectively (R F ¼ 19.5%, Table 1), immediately suggest- ing the presence of three (N/O)-ligands and one sulfur ligand. Fixing the number of zinc ligands to the nearest integer values (i.e. 3 and 1 for (N/O) and sulfur ligands, respectively) yields a fit result of comparable quality (R F ¼ 20.9%; see Table 1, parameters in parenthesis), whereas a combination of 2 · (N/O) plus 2 · Sresultsin unsatisfactory fits (R F ¼ 29.3%) and ÔproblematicÕ Debye-Waller factors (2r 2 too high for (N/O)-shell, but too low for sulfur shell). Constraining the sum of the two coordination numbers to a value of 3 or 5 results in poorer fits and significantly higher R F values (26.0% and 27.8%) confirming that zinc is tetra-coordinate. We did not simulate the weak features seen in the Fourier transform at (reduced) distances greater than 3.2 A ˚ (Fig. 1A-I). Similar features have been ascribed to single and multiple scattering effects resulting from imidazole ligation to zinc [14,30]. In conclu- sion, the EXAFS data suggests that in the wild-type the zinc is tetra-coodinated with one sulfur and three (N/O)-ligands. The EXAFS spectra are significantly altered (Fig. 1A) by addition of HS-CoM to wild-type MtaA. When the ratio of MtaA to coenzyme M was 1 : 0.5, the spectrum is best simulated with 2.8 (N/O)-ligands and 1.3 sulfur ligands (Table 1, Fig. 1A-II, thick line). These numbers compare reasonably well with the results expected for high-affinity binding of HS-CoM resulting in substitution of one (N/O)- ligand by 1 sulfur ligand in 50% of all MtaA proteins [i.e. 2.5 (N/O) plus 1.5 sulfur at the MtaA–HS-CoM stoichi- ometry of 1 : 0.5]. At a MtaA to HS-CoM ratio of 1: 2, unconstrained fit results suggest 1.7 (N/O)-ligands and 2.3 sulfurs (Fig. 1A-III). These numbers are close to the ideal values for substitution of one (N/O)-ligand by a sulfur ligand in all zinc-containing MtaA proteins [i.e. 2.0 (N/O) plus 2 (S)]. We interpret these results as indicating the binding of the thiol sulfur of HS-CoM to the zinc atom of MtaA in a way that it replaces one of the nonsulfur ligands, possibly the oxygen of coordinated H 2 O or hydroxide. EXAFS of two MtaA mutants, Cys239 fi Ala and His237 fi Ala His237, Cys239 and possibly Cys316 have been proposed to be involved in zinc coordination in MtaA based on sequence comparisons of MtaA with MetE [15]. By site-directed mutagenesis, we therefore exchanged His237 and Cys239 to alanine. The His237 fi Ala mutated enzyme and the Cys239 fi Ala mutated enzyme had zinc contents of 0.25 mol/mol and 0.4 mol/mol, respectively, and exhibited specific activities of 0.02 UÆmg )1 and 0.01 UÆmg )1 . Figure 1B shows the Fourier transforms of the experi- mental EXAFS spectra (thin lines) for wild-type MtaA (trace I), for Cys239 fi Ala MtaA (trace IV), and for His237 fi Ala MtaA (trace V), in the absence of coen- zyme M. The two mutant proteins exhibit EXAFS spectra that are clearly different from those of the wild-type, strongly suggesting that His237 and Cys239 are involved in zinc coordination. In comparison to the wild-type, in the spectrum of Cys239 fi Ala (Fig. 1B-IV) Peak 2 is shifted to shorter distances and a new peak appears at a reduced distance of % 2.4 A ˚ (corresponding to a zinc-ligand distance of % 2.7 A ˚ ). The latter feature, which indicates the presence of relatively strong backscattering by atoms at % 2.7 A ˚ ,is not found in the wild-type. Only a poor simulation (R F ¼ 44%) of the EXAFS of the Cys239 fi Ala mutant is obtained with similar parameters as used for the wild- type. A specific alternative approach yields satisfactory results [(N/O) at % 2.0 A ˚ ,(N/O)at% 2.2 A ˚ ,Sat% 2.7 A ˚ , Table 1, Fig. 1B-IV)], whereas other simulation approaches using three shells of backscattering atoms did not result in high-quality fits. Interestingly, in Cys239 fi Ala the zinc seems to be penta-coordinated. These results are highly suggestive that, in the wild-type, Cys239 is a zinc ligand. In the Cys239 fi Ala mutant, the thiol group of Cys239 seems to be replaced by two (N/O)-ligands, presumably the oxygens of two H 2 O molecules. Furthermore, there may be a thiol group of an amino-acid residue which moves closer to the active-site zinc due the structural rearrange- ment resulting from the mutation. In the His237 fi Ala mutant, peak 2 is strongly increased (in comparison to the wild-type, see Fig. 1B-V) proving a significantly modified ligand environment of the active-site zinc. A simulation with two shells [(N/O), S] yielded coordination numbers of 2.4 and 1.8, respectively, Fig. 2. Normalized XANES spectra of wild-type MtaA and of two MtaA mutants (Cys239 fi Ala, His237 fi Ala)intheabsenceand presence of added coenzyme M. (A) Wild-type MtaA (thick line), wild- type MtaA + 0.5 mol/mol HS-CoM (line of medium thickness), wild- type MtaA + 2 mol/mol HS-CoM (thin line). (B) Wild-type MtaA (solid line), Cys239 fi Ala (dotted line), His237 fi Ala (dashed line). (C) His237 fi Ala (thick dashes), His237 fi Ala+2mol/mol HS-CoM (thin dashes). (D) Cys239 fi Ala (thick dots), Cys239 fi Ala + 2 mol/mol HS-CoM (thin dots). (E) ZnCl 2 in aqueous solution (dash-dotted line). 2120 M. Kru ¨ er et al. (Eur. J. Biochem. 269) Ó FEBS 2002 pointing towards the presence of two (N/O)-ligands and two sulfur ligands in the mutant (Table 1). Possibly, one of the (N/O)-ligands in the wild-type, probably His237, is replaced byasulfurligandintheHis237fi Ala mutant. It is tempting to speculate that this sulfur ligand is the same that is observed in the Cys239 fi Ala mutant at a zinc-sulfur distance of % 2.7 A ˚ . In Cys239 fi Ala and His237 fi Ala, the pronounced changes in the EXAFS spectra resulting from HS-CoM addition are indicative of an increase in the number of sulfur ligands (visual inspection of Fig. 1C). Thus, in both mutants the HS-CoM binding seems to involve ligation of the thiol group of HS-CoM to the active-site zinc. For the Cys239 fi Ala mutant plus HS-CoM the simulation results suggest the presence of tetra-coordinated zinc with two (N/O) and two sulfur ligands (Fig. 1C-VI, Table 1) showing that the HS-CoM addition may lead to an increase in the number of sulfur ligands of 2. We tentatively propose that either more than 1 HS-CoM molecule binds to a single zinc site or, alternatively, one HS-CoM sulfur as well as the sulfur detectable at % 2.7 A ˚ in the absence of HS-CoM become direct ligands in its presence. In His237 fi Ala, in the presence of HS-CoM the spectrum is well simulated using two shells of backscatterers [(N/O), S] with coordination numbers of 1.4 and 2.5 (Fig. 1C-VII, Table 1); fixing the coordination numbers to integer values results in a satisfactory simula- tion for 1 · (N/O) and 3 · S. The clear increase in the coordination number for sulfur upon HS-CoM addition indicates that also in His237 fi Ala the HS-CoM binding results in sulfur ligation to the active-site zinc. XANES spectra of wild-type and mutants As shown in the following, the interpretation of the EXAFS of the mutant enzymes is confirmed by a comparative analysis of the zinc K-edge spectra (XANES spectra, see Fig. 2) of the wild-type and the mutants (His237 fi Ala, Cys239 fi Ala) in the presence and absence of added HS-CoM. For the wild-type, the EXAFS analysis unambiguously indicates that HS-CoM addition results in an increase in the number of sulfur atoms in the first coordination sphere of zinc. This increase in the number of sulfur ligands is accompanied by the following changes in the XANES (Fig. 2A, Table 1): (a) the edge position shifts down from 9663.5 eV to 9663.1 eV; (b) the magnitude of an absorption peak at the top of the edge (% 9671 eV) becomes reduced (Fig. 2A). Seemingly, additional sulfur ligands result in down-shift of the edge-energy as well as in reduced absorption at the top of the edge. In inorganic models, peptide models and zinc proteins, Penner-Hahn and coworkers observed the same relations between XANES spectra and the number of sulfur ligands [14,23,30]. In His237 fi Ala (Fig. 2B, dashed line), in comparison to the wild-type enzyme (Fig. 2b, solid line) the K-edge is shifted to lower energies (Table 1) and the absorption on top of the edge is reduced. In the Cys239 fi Ala mutant (Fig. 2B, dotted line), the K-edge energy is increased (Table 1) and the absorption on top of the edge is significantly increased. We conclude that not only the EXAFS but also the XANES spectra are suggestive of a reduced number of sulfur ligands in the Cys239 fi Ala and an increased number of sulfur ligands in His237 fi Ala (in comparison to the wild-type enzyme). As a model for the zinc in the absence of any sulfur ligand we use ZnCl 2 dissolved in H 2 O (Fig. 2E, dash-dotted line). The Cys239 fi Ala (Fig. 2D, thick dots) and the ZnCl 2 spectra (Fig. 2E) are obviously very similar (edge energy of 9663.7 eV, single edge peak at % 9669 eV) confirming the absence of any sulfur ligands in the Cys239 fi Ala mutant. For both mutants (Fig. 2C,D), comparison of edge positions and edge-peak magnitudes strongly suggests that coenzyme M addition (Fig. 2C,D; thin lines) results in additional sulfur ligands in the first coordination sphere of zinc. In the Cys239 fi Ala mutant, coenzyme M addition Table 1. Zinc EXAFS simulation parameters and edge positions for wild-type (WT) MtaA and mutated MtaA (Cys239 fi Ala, His237 fi Ala) without and in the presence of 0.5 mol/mol coenzyme M (+ ½ HS-CoM) or 2 mol/mol coenzyme M (+ 2 HS-CoM). The values in parenthesis result from simulations using coordination numbers fixed to integer values. The edge position, E K-edge , was determined for the XANES spectra shown in Fig. 2; EXAFS data and simulations are shown in Fig. 1. Simulation parameters: N, coordination number (coord. no.); 2r 2 , Debye-Waller parameter; R F , the filtered R-factor [34] representing the deviation between data and simulation (in percentage) for reduced distances ranging from 1to2.8A ˚ . Sample E K-edge (eV) Shell Coord. no. Distance, R (A ˚ )2r 2 (A ˚ 2 )R F (%) WT MtaA 9663.5 (N/O) 3.4 (3.0) 2.02 (2.02) 0.009 (0.007) 19.5 (20.9) S 0.7 (1.0) 2.32 (2.31) 0.003 (0.006) WT MtaA + ½ HS-CoM 9663.3 (N/O) 2.8 (2.5) 2.02 (2.03) 0.010 (0.009) 24.6 (25.6) S 1.3 (1.5) 2.32 (2.32) 0.005 (0.006) WT MtaA + 2 HS-CoM 9663.1 (N/O) 1.7 (2.0) 2.06 (2.06) 0.005 (0.005) 11.4 (13.3) S 2.3 (2.0) 2.32 (2.32) 0.005 (0.004) Cys239 fi Ala 9663.6 (N/O) 2.9 (3.0) 2.04 (2.04) 0.004 (0.004) 21.4 (21.3) O 2.2 (2.0) 2.17 (2.17) 0.003 (0.003) S 0.9 (1.0) 2.72 (2.72) 0.006 (0.007) Cys239 fi Ala + 2 HS-CoM 9663.1 (N/O) 2.1 (2.0) 2.07 (2.07) 0.007 (0.007) 17.5 (18.0) S 1.9 (2.0) 2.32 (2.32) 0.005 (0.005) His237 fi Ala 9663.0 (N/O) 2.4 (2.0) 2.03 (2.03) 0.006 (0.005) 16.5 (17.1) S 1.8 (2.0) 2.29 (2.29) 0.010 (0.009) His237 fi Ala + 2 HS-CoM 9662.6 (N/O) 1.4 (1.0) 2.07 (2.06) 0.003 (0.003) 13.8 (14.1) S 2.5 (3.0) 2.32 (2.32) 0.005 (0.006) Ó FEBS 2002 Zn XAFS of methanol:coenzyme M methyltransferase (Eur. J. Biochem. 269) 2121 results in particularly pronounced changes. The absorption at 9671 eV and the edge position reach the values found for the wild-type in the presence of coenzyme M (Fig. 2C, thin dotted line). These findings point towards two sulfur ligands in the Cys239 fi Ala mutant supplemented with coen- zyme M. In summary, the XANES analysis fully confirms the EXAFS results on the number of sulfur ligands in the first coordination sphere of the active-site zinc. DISCUSSION Zinc EXAFS spectra of wild-type MtaA from M. barkeri and of two mutant proteins, Cys239 fi Ala and His237 fi Ala, in the absence and presence of coen- zyme M, indicate how zinc interacts with its substrate coenzyme M and how zinc is most probably coordinated in the active site of this methyltransferase. Upon binding of coenzyme M to MtaA the number of sulfur ligands to zinc increased by at least one whereas the total number of ligands remained four; one (N/O) ligand was replaced by a sulfur ligand. This, and the previous finding that upon coenzyme M binding to MtaA one mol of protons was released [7], is consistent with an activation of coenzyme M via coordination of its thiol group to the active site zinc. The EXAFS data reported for the isoenzyme MtbA show that the isoenzyme uses the same activation mechanism [23]. The EXAFS data do not prove this activation mecha- nism definitely as they only show that upon coenzyme M binding to MtaA the number of sulfur ligands increased from 1 to 2. However, this increase could also be due to the binding of a cysteine residue to the active site zinc upon enzyme–substrate complex formation. Final proof could come from EXAFS studies using seleno coenzyme M as substrate [31]. The simulation of the EXAFS spectrum of wild-type MtaA revealed that zinc is likely coordinated by 1 sulfur ligand and 3 (N/O) ligands, its total coordination number is 4. In both mutants, zinc is quite differently coordinated, namely by 2 (N/O) and 2 S ligands in His237 fi Ala, and by 5 (N/O) ligands in the Cys239 fi Ala mutant. These results, together with the features of the XANES spectra likely indicate that both residues, Cys239 and His237 provide direct ligands to zinc in wild-type MtaA. Cys239 was probably replaced by 2 H 2 O molecules and His237 by a thiol group from a cysteine residue of which the enzyme contains six. In the E. coli Ada protein, which is a zinc dependent methyltransferase, the active site zinc is coordinated to four cysteine sulfur ligands [13]. In the methyltransferase MetH zinc is coordinated to 3 S and 1 (N/O) ligands and in MetE to 2 S and 2 (N/O) ligands [14]. In protein farnesyl transferase the active site zinc is coordinated by 1 S and 3 (N/O) ligands, as revealed by the crystal structure [19]. Zinc thus can be quite differently coordinated and still activate thiol groups for alkylation. Consistent with the interpreta- tion is our finding that both MtaA mutants still bind coenzyme M and exhibit some activity although in both mutants the coordination of zinc differs from the situation in the wild-type enzyme. Apparently, other ligands to zinc (H 2 O thiol groups of cysteine residues) were able, at least in part, to functionally replace Cys239 and His237. The EXAFS and XANES spectra recently reported for the MtaA isoenzyme MtbA from M. barkeri revealed that in the active site of MtbA zinc is coordinated by 2 S and 2 (N/O) ligands [23]. These spectra are indeed clearly different to our spectra of MtaA showing that the active site zinc in MtaA is most probably coordinated by 1 S and 3 (N/O) ligands. MtaA and MtbA are only 40% sequence identical [22] but they share the putative HXCX n Czinc binding motif [15]. The latter finding indicates that zinc coordination in MtaA and MtbA should be the same. The finding that the number of sulfur ligands to zinc can vary in different S-alkyl transferases from 1 to 4 (see above) and that both mutated MtaA proteins still showed some activity indicates, however, that some aspects of the ligand envi- ronment of zinc are surprisingly uncritical for thiol group activation. It could therefore well be that in MtaA the zinc is ligated by 1 S and 3 (N/O) and in MtbA by 2 S and 2(N/O). In comparing the results for MtaA (this work) and MtbA [23] it has to be considered that the MtaA samples analysed by X-ray absorption spectroscopy were His tagged (whereas the MtbA samples were not). Adventitiously picked up zinc by the His tag might contribute to a higher number of (N/O) ligands and thus to a relatively lower number of S ligands. The finding, however, that upon addition of coenzyme M to MtaA the number of S ligands increases from 1 to 2 whereas in the case of MtbA the number increases from 2 to 3 suggests that it is unlikely that the His tagging is responsible for the differences between the MtaA- and MtbA-EXAFS results (although this possibility cannot be totally excluded). MtbA is inactivated when Cys241 or Cys316 of the putative zinc binding motif H239XC241X n C316 are mutated [23]. In case of MtaA the mutation of Cys239 in the H237XC239X n C316 motif leads to an enzyme with only a few percent activity. It is not known whether MtaA becomes inactive when Cys316 is mutated and how this mutation affects the Zn ligand environment. The apparent absence of any sulfur in the first coordination sphere of Zn in the Cys239 fi Ala mutant suggests that Cys316 is not a direct Zn ligand in MtaA. It would be interesting to investigate the XAFS of a Cys316 mutant to support (or disprove) the model of 1 S plus 3 (N/O) ligands in the resting state of MtaA. ACKNOWLEDGEMENTS We gratefully acknowledge financial support by the Max-Planck Society, by the Deutsche Forschungsgemeinschaft, and by the Fonds der Chemischen Industrie. REFERENCES 1. Thauer, R.K. (1998) Biochemistry of methanogenesis: a tribute to Marjory Stephenson. Microbiology 144, 2377–2406. 2. Keltjens, J.T. & Vogels, G.D. (1993) Conversion of methanol and methylamines to methane and carbon dioxide. Methanogenesis (Ferry, J.G., ed.), pp. 253–303. Chapman & Hall, New York London. 3. Sauer, K. & Thauer, R.K. (1997) Methanol:coenzyme M methyltransferase from Methanosarcina barkeri. Zinc dependence and thermodynamics of the methanol:cob(I)alamin methyl- transferase reaction. Eur. J. Biochem. 249, 280–285. 4. Sauer, K., Harms, U. & Thauer, R.K. (1997) Methanol:coenzyme M methyltransferase from Methanosarcina barkeri. Purification, 2122 M. Kru ¨ er et al. (Eur. J. Biochem. 269) Ó FEBS 2002 properties and encoding genes of the corrinoid protein MT1. Eur. J. Biochem. 243, 670–677. 5. Sauer, K. & Thauer, R.K. (1998) Methanol:coenzyme M methyltransferase from Methanosarcina barkeri – identification of the active-site histidine in the corrinoid-harboring subunit MtaC by site-directed mutagenesis. Eur. J. Biochem. 253, 698–705. 6. Sauer, K. & Thauer, R.K. (1999) Methanol:coenzyme M methyltransferase from Methanosarcina barkeri – substitution of the corrinoid harbouring subunit MtaC by free cob(I)alamin. Eur. J. Biochem. 261, 674–681. 7. Sauer, K. & Thauer, R.K. (2000) Methyl-coenzyme M formation in methanogenic archaea. Involvement of zinc in coenzyme M activation. Eur. J. Biochem. 267, 2498–2504. 8. LeClerc, G.M. & Grahame, D.A. (1996) Methylcobamide: coen- zyme M methyltransferase isozymes from Methanosarcina barkeri. Physicochemical characterization, cloning, sequence analysis, and heterologous gene expression. J. Biol. Chem. 271, 18725–18731. 9. Grahame, D.A. (1989) Different isozymes of methylcobalamin: 2-mercaptoethanesulfonate methyltransferase predominate in methanol- versus acetate-grown Methanosarcina barkeri. J. Biol. Chem. 264, 12890–12894. 10. Burke, S.A. & Krzycki, J.A. (1997) Reconstitution of mono- methylamine:coenzyme M methyl transfer with a corrinoid pro- tein and two methyltransferases purified from Methanosarcina barkeri. J. Biol. Chem. 272, 16570–16577. 11. Yeliseev, A., Gartner, P., Harms, U., Linder, D. & Thauer, R.K. (1993) Function of methylcobalamin:coenzyme M methyl- transferase isoenzyme II in Methanosarcina barkeri. Arch. Microbiol. 159, 530–536. 12. Matthews, R.G. & Goulding, C.W. (1997) Enzyme-catalyzed methyl transfers to thiols: the role of zinc. Curr. Opin. Chem Biol. 1, 332–339. 13. Wilker, J.J. & Lippard, S.J. (1997) Alkyl transfer to metal thiolates – kinetics, active species identification, and relevance to the DNA methyl phosphotriester repair center of Escherichia coli Ada. Inorg. Chem. 36, 969–978. 14. Peariso, K., Goulding, C.W., Huang, S., Matthews, R.G. & Penner-Hahn, J.E. (1998) Characterization of the zinc binding site in methionine synthase enzymes of Escherichia coli – the role of zinc in the methylation of homocysteine. J. Am. Chem Soc. 120, 8410–8416. 15. Zhou, Z.S., Peariso, K., Penner-Hahn, J.E. & Matthews, R.G. (1999) Identification of the zinc ligands in cobalamin-independent methionine synthase (MetE) from Escherichia coli. Biochemistry 38, 15915–15926. 16. Breksa, A.P. III & Garrow, T.A. (1999) Recombinant human liver betaine-homocysteine S-methyltransferase: identification of three cysteine residues critical for zinc binding. Biochemistry 38, 13991– 13998. 17. Thanbichler, M., Neuhierl, B. & Bock, A. (1999) S-methyl- methionine metabolism in Escherichia coli. J. Bacteriol. 181, 662– 665. 18. Allen, J.R., Clark, D.D., Krum, J.G. & Ensign, S.A. (1999) A role for coenzyme M (2-mercaptoethanesulfonic acid) in a bacterial pathway of aliphatic epoxide carboxylation. Proc. Natl Acad. Sci. USA 96, 8432–8437. 19. Strickland,C.L.,Windsor,W.T.,Syto,R.,Wang,L.,Bond,R., Wu, Z., Schwartz, J., Le, H.V., Beese, L.S. & Weber, P.C. (1998) Crystal structure of farnesyl protein transferase complexed with a CaaX peptide and farnesyl diphosphate analogue. Biochemistry 37, 16601–16611. 20. Rozema, D.B. & Poulter, C.D. (1999) Yeast protein farnesyl- transferase. pK a s of peptide substrates bound as zinc thiolates. Biochemistry 38, 13138–13146. 21. Goulding, C.W. & Matthews, R.G. (1997) Cobalamin-dependent methionine synthase from Escherichia coli: involvement of zinc in homocysteine activation. Biochemistry 36, 15749–15757. 22. Harms, U. & Thauer, R.K. (1996) Methylcobalamin:coenzyme M methyltransferase isoenzymes MtaA and MtbA from Methano- sarcina barkeri. Cloning, sequencing and differential transcription of the encoding genes, and functional overexpression of the mtaA gene in Escherichia coli. Eur. J. Biochem. 235, 653–659. 23. Gencic,S.,LeClerc,G.M.,Gorlatova,N.,Peariso,K.,Penner- Hahn, J.E. & Grahame, D.A. (2001) Zinc-thiolate intermediate in catalysis of methyl group transfer in Methanosarcina barkeri. Biochemistry 40, 13068–13078. 24. Ferguson, D.J., Krzycki, J.A. & Grahame, D.A. (1996) Specific roles of methylcobamide:coenzyme M methyltransferase isozymes in metabolism of methanol and methylamines in Methanosarcina barkeri. J. Biol. Chem. 271, 5189–5194. 25. Tallant, T.C., Paul, L. & Krzycki, J.A. (2001) The MtsA subunit of the methylthiol:coenzyme M methyltransferase of Methano- sarcina barkeri catalyzes both half-reactions of corrinoid-depen- dent dimethylsulfide:coenzyme M methyl transfer. J. Biol. Chem. 276, 4485–4493. 26. Bradford, M.M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254. 27. Schiller, H., Dittmer, J., Iuzzolino, L., Dorner, W., Meyer- Klaucke, W., Sole, V.A., Nolting, H. & Dau, H. (1998) Structure and orientation of the oxygen-evolving manganese complex of green algae and higher plants investigated by X-ray absorption linear dichroism spectroscopy on oriented photosystem II mem- brane particles. Biochemistry 37, 7340–7350. 28. Pettifer, R.F. & Hermes, C. (1985) Absolute energy calibration of X-ray radiation from synchrotron sources. J. Appl. Crystallogr. 18, 404–412. 29. Zhabinsky, S.I., Rehr, J.J., Ankudinov, A., Albers, R.C. & Eller, M.J. (1995) Multiple-scattering calculations of x-ray-absorption spectra. Phys Rev. B. 52, 2995–3009. 30. Clark-Baldwin, K., Tierney, D.L., Govindaswamy, N., Gruff, E.S., Kim, C., Berg, J., Koch, S.A. & Penner-Hahn, J.E. (1998) The limitations of X-ray spectroscopy for determining the sructure of zinc sites in proteins. When is a tetrathiolate not a tetrathiolate. J. Am. Chem Soc. 120, 8401–8409. 31. Peariso,K.,Zhou,Z.S.,Smith,A.E.,Matthews,R.G.&Penner- Hahn, J.E. (2001) Characterization of the zinc sites in cobalamin- independent and cobalamin-dependent methionine synthase using zinc and selenium X-ray absorption spectroscopy. Biochemistry 40, 987–993. 32. Corwin, D.T. & Koch, S.A. (1988) Crystal structures of Zn (SR) 2 complexes: Structural Models for the proposed [Zn(cys-S)2(his)2] center in transcription factor IIIA and related nucleic acid binding proteins. Inorg Chem. 27, 493–496. 33. Gruff, E.S. & Koch, S.A. (1989) A trigonal planar [Zn(SR)3]1- omplex. A possible new coordination mode for zinc-cysteine centers. J. Am. Chem Soc. 111, 8762–8763. 34. Meinke, C., Sole, V.A., Pospisil, P. & Dau, H. (2000) Does the structure of the water-oxidizing photosystem II-manganese com- plex at room temperature differ from its low-temperature struc- ture? A comparative X-ray absorption study. Biochemistry 39, 7033–7040. Ó FEBS 2002 Zn XAFS of methanol:coenzyme M methyltransferase (Eur. J. Biochem. 269) 2123 . The role of zinc in the methylation of the coenzyme M thiol group in methanol :coenzyme M methyltransferase from Methanosarcina barkeri New insights from. Characterization of the zinc binding site in methionine synthase enzymes of Escherichia coli – the role of zinc in the methylation of homocysteine. J. Am. Chem Soc.