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

-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

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: 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

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

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

; 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

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
˚

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

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

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
˚

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

(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

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

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

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-

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