Synthetic models for the zinc sites in the methionine synthases

Article abstract of DOI:10.1021/ic000505q

The syntheses and molecular structures of a series of tetrahedral zinc complexes designed to model the active sites in Escherichia coli methionine synthases are reported. [PhTt(tBu)]ZnBr (PhTt(tBu) = phenyltris((tert-butylthio)-methyl)borate) was prepared and characterized crystallographically to provide entry into [S₃]ZnX complexes. Metathesis with KSPh yielded the phenylthiolato complex, [PhTt(tBu)]Zn(SPh), which represents a structural mimic of the homocysteine ligated form of the enzyme. Alternatively, [S₂N]ZnX (X = Br, CH₃, SPh) species were prepared using the new mixed-donor ligands, [Ph(pz)Bt(tBu)] (phenyl(pyrazolyl)bis((tert-butylthio)methyl)borate) and [Ph(pz(tBu))Bt(tBu)] (phenyl(3-tert-butylpyrazolyl)bis((tert-butylthio)methyl)borate). Protonolysis of [Ph(pz(tBu))-Bt(tBu)]Zn(CH₃) by PhSH in toluene yielded [Ph(pz(tBu))Bt(tBu)]Zn(SPh), a synthetic analogue of the homocysteine ligated form of cobalamin-independent methionine synthase (Met E). The average Zn-S bond distance in [Ph-(pz(tBu))Bt(tBu)]Zn(SPh) of 2.37 A compares well with the EXAFS-derived distance of 2.31 A found in the homocysteine-bound form of Met E.

Full text of DOI:10.1021/ic000505q

Inorg. Chem. 2000, 39, 4347-4353
4347
Synthetic Models for the Zinc Sites in the Methionine Synthases
Show-Jen Chiou, Julie Innocent, Charles G. Riordan,* Kin-Chung Lam,
Louise Liable-Sands, and Arnold L. Rheingold
Department of Chemistry and Biochemistry, University of Delaware, Newark, Delaware 19716
ReceiVed May 11, 2000
The syntheses and molecular structures of a series of tetrahedral zinc complexes designed to model the active
sites in Escherichia coli methionine synthases are reported. [PhTttBu]ZnBr (PhTttBu ) phenyltris((tert-butylthio)-
methyl)borate) was prepared and characterized crystallographically to provide entry into [S3]ZnX complexes.
Metathesis with KSPh yielded the phenylthiolato complex, [PhTttBu]Zn(SPh), which represents a structural mimic
of the homocysteine ligated form of the enzyme. Alternatively, [S2N]ZnX (X ) Br, CH3, SPh) species were
prepared using the new mixed-donor ligands, [Ph(pz)BttBu] (phenyl(pyrazolyl)bis((tert-butylthio)methyl)borate)
and [Ph(pztBu)BttBu] (phenyl(3-tert-butylpyrazolyl)bis((tert-butylthio)methyl)borate). Protonolysis of [Ph(pztBu)-
BttBu]Zn(CH3) by PhSH in toluene yielded [Ph(pztBu)BttBu]Zn(SPh), a synthetic analogue of the homocysteine
ligated form of cobalamin-independent methionine synthase (Met E). The average Zn-S bond distance in [Ph-
(pztBu)BttBu]Zn(SPh) of 2.37 Å compares well with the EXAFS-derived distance of 2.31 Å found in the
homocysteine-bound form of Met E.
Introduction
nature of the protein ligands defines the diverse catalytic
function remains unclear. Small molecule analogues would
Zinc is an essential metal ion in biological processes, playing
a seminal role in the structure and function of a wide range of
metalloproteins. Structural zinc sites include, for example, zinc
finger proteins which regulate gene transcription and a structural
ion in the enzyme liver alcohol dehydrogenase.1 A common
feature of structural zinc sites is tetrahedral coordination in
which all four donors are protein residues. In this sense the metal
ion is coordinatively saturated. Alternatively, when zinc has been
sequestered for a catalytic function, an exogenous (and labile)
donor such as water occupies one of the four coordination sites.
Examples include mononuclear sites as in carbonic anhydrase
(CA), liver alcohol dehydrogenase (LADH), and the methionine
synthases.2 In these proteins the zinc is coordinated by three
side chain residues and an exogenous donor. Presumably,
catalysis requires a labile coordination site to allow for substrate
binding in a four-coordinate structure. However, this is far from
certain given the propensity with which zinc can access higher
coordination numbers. For example, the catalytic zinc in the
endopeptidase astacin is five coordinate as deduced crystallo-
graphically.3 While the common feature that three protein
residues ligate zinc in many catalytic sites has emerged, the
identity of these residues varies greatly. In general, the sites
may be characterized as Zn[NxOySz] (x + y + z ) 3 or 4). In
CA, the three donors are histidines while in LADH the ligands
are two cysteines and a histidine.4 In farnesyl transferase all
three donor atom types are present: histidine, cysteine, aspartate
(bidentate), and a water molecule.5 Despite the growing
structural database of zinc proteins, the extent to which the
clearly aid in developing a deeper understanding of relevant
zinc coordination chemistry with the goal of developing
functional (catalytic) models.6-11
The de novo biosynthesis of the amino acid methionine from
homocysteine (Hcys) requires methyl group transfer, a reaction
catalyzed by the metalloenzyme methionine synthase.2 The
reaction involves CH3-H4folate as the one-carbon donor and
Hcy as the acceptor, Scheme 1. Nature has evolved two
pathways for this process. One, termed the cobalamin-dependent
methionine synthase (Met H), involves the intermediacy of
methylcobalamin. In this two-step process the methyl group is
transferred first from CH3-H4folate to cob(I)alamin, Scheme
2. In a distinct step methylcobalamin alkylates Hcys yielding
Met and regenerating cob(I)alamin. The other pathway does not
employ the cobalamin cofactor. Instead the reaction proceeds
by direct transmethylation from CH3-H4folate to Hcys (cobal-
amin-independent methionine synthase, Met E). Matthews has
demonstrated recently that zinc is required for both enzymes.12,13
Several elegant biophysical studies have been reported, resulting
in proposals for the active site structures and function of the
zinc ion.14-16 In Met H, the zinc is ligated by three sulfur donors,
presumably cysteines and/or methionines, and a nitrogen or
(6) Looney, A.; Han, R. V.; McNeill, K.; Parkin, G. J. Am. Chem. Soc.
1993, 115, 4690-4697.
(7) Bergquist, C.; Parkin, G. J. Am. Chem. Soc. 1999, 121, 6322-6323.
(8) Brand, U.; Rombach, M.; Vahrenkamp, H. Chem. Commun. 1998,
2717-2718.
(9) Grapperhaus, C. A.; Tuntulani, T.; Reibenspies, J. H.; Darensbourg,
M. Y. Inorg. Chem. 1998, 37, 4052-4058.
(1) Lipscomb, W. N.; Stra¨ter, N. Chem. ReV. 1996, 96, 2375-2434.
(2) Matthews, R. G.; Goulding, C. W. Curr. Opin. Chem. Biol. 1997, 1,
332-339.
(10) Hammes, B. S.; Carrano, C. J. Inorg. Chem. 1999, 38, 4593-4600.
(11) Kimblin, C.; Hascall, T.; Parkin, G. Inorg. Chem. 1997, 36, 5680-
5681.
(3) Bode, W.; Gomis-Rueth, F. X.; Huber, R.; Zwilling, R.; Stoecker, W.
Nature 1992, 358, 164-167.
(12) Gonzalez, J. C.; Peariso, K.; Penner-Hahn, J. E.; Matthews, R. G.
Biochemistry 1996, 35, 12228-12234.
(13) Goulding, C. W.; Matthews, R. G. Biochemistry 1997, 36, 15749-
15757.
(14) Gonzalez, J. C.; Goulding, C. W.; Jarrett, J.; Peariso, K.; Penner-
Hahn, J. E.; Matthews, R. G. Protein Eng. 1997, 57.
(4) Benach, J.; Atrian, S.; Gonzalez-Duarte, R.; Ladenstein, R. J. Mol.
Biol. 1998, 282, 383-399.
(5) Park, H.-W.; Boduluri, S. R.; Moomaw, J. F.; Casey, P. J.; Beese, L.
S. Science 1997, 275, 1800-1804.
10.1021/ic000505q CCC: $19.00 © 2000 American Chemical Society
Published on Web 08/26/2000

4348 Inorganic Chemistry, Vol. 39, No. 19, 2000
Chiou et al.
Scheme 1
Scheme 2
investigation and comparison of the zinc sites in both Met H
and Met E. Contained herein is the preparation of the necessary
ligands and their zinc derivatives, including monomeric zinc
thiolates, that serve as models for the Hcy-bound forms of the
methionine synthases. Outcomes from these studies have broad
implications to a wide range of zinc protein catalyzed trans-
alkylations to thiols including methyl reductase,22 the Ada
protein,23 farnesyl transferase,5 betaine homocysteine methyl-
transferase,24 and epoxyalkane coenzyme M transferase.25
Consequently, we aim to bring our understanding of thiol
activating zinc proteins in line with the much more established
roles of zinc in maintaining structural integrity and catalyzing
hydrolysis reactions.
Results
Ligand Syntheses. The preparation of the mixed-donor [NS2]
ligands, [Ph(pz)BttBu]H and [Ph(pztBu)BttBu]H, proceeds smoothly
employing a one-pot, stepwise protocol.26 Using PhBCl2 as the
borane source, the tert-butyl thioether substituents were intro-
duced first by addition of 2 equiv of LiCH2StBu. Then the
appropriate pyrazolate was added as its Li+ salt. Aqueous
workup yielded gram quantities of the “free acid” ligand. The
compounds are white, crystalline solids stable in air and soluble
1
oxygen ligand, in a four-coordinate structure. Upon incubation
with Hcy the N/O donor is replaced by homocysteine yielding
an all-sulfur environment. In Met E, the zinc is initially in a
[S2(N/O)2] environment that changes to [S3(N/O)1] when the
enzyme is exposed to Hcy. The changes for each enzyme have
been interpreted to indicate that homocysteine binds to zinc
displacing the light atom (N or O) ligand.15 Metal binding
presumably activates homocysteine for transmethylation by
rendering it ionized (as the thiolate) at neutral pH; thiolate anions
are more nucleophilic than neutral thiols. However, it is not
known to what degree a thiolate’s nucleophilicity is modulated
by coordination to a metal ion. Recent work by our group17
and others9,18 has begun to quantify this effect.
in a range of organic solvents. Their H NMR spectra exhibit
clearly resolved resonances for the pyrazolyl and phenyl
hydrogens. Most characteristic are the downfield resonance for
the “acid” proton at δ 13.5 and the diastereotopic methylene
protons that appear as doublets at δ 2.2-2.0. The “free acid”
form of the ligands proved suitable for the direct addition to
organozinc precursors.
Preparation of Metal Complexes. The synthesis of the
tetrahedral zinc complexes is outlined in Scheme 3. The
[PhTttBu] derivatives were accessed using ZnBr2 as starting
material. Addition of [PhTttBu]Tl20 to anhydrous ZnBr2 in diethyl
ether produced [PhTttBu]ZnBr. The 1H NMR spectrum of
[PhTttBu]ZnBr contains single resonances for the methylene and
This paper details our initial efforts to develop a better
understanding of the reaction pathway of the methionine
synthases, specifically the ability of zinc to activate homocys-
teine for attack by electrophiles, through the study of synthetic
model complexes. As shown in Scheme 3, the borato ligands
developed in our laboratories are ideally suited for this
pursuit.19-21 They provide the necessary ancillary ligand
frameworksface-capping array of donor groupss to permit
(19) Ge, P.; Riordan, C. G.; Haggerty, B. S.; Rheingold, A. L. J. Am. Chem.
Soc. 1994, 116, 8406-8407.
(20) Schebler, P.; Riordan, C. G.; Guzei, I.; Rheingold, A. L. Inorg. Chem.
1998, 37, 4754-4755.
(21) Chiou, S.; Ge, P.; Liable-Sands, L. M.; Rheingold, A. L. Chem.
Commun. 1999, 159-160.
(22) Ermler, U.; Grabarse, W.; Shima, S.; Goubeaud, M.; Thauer, R. K.
Science 1997, 278, 1457-1462.
(23) Myers, L. C.; Terranova, M. P.; Ferentz, A. E.; Wagner, G.; Verdine,
G. L. Science 1993, 261, 1164-1167.
(15) Peariso, K.; Goulding, C. W.; Huang, S.; Matthews, R. G.; Penner-
Hahn, J. E. J. Am. Chem. Soc. 1998, 120, 8410-8416.
(16) Zhou, Z. S.; Peariso, K.; Penner-Hahn, J. E.; Matthews, R. G.
Biochemistry 1999, 38, 15915-15926.
(17) Ram, M. S.; Riordan, C. G.; Ostrander, R.; Rheingold, A. L. Inorg.
Chem. 1995, 34, 5884-5892.
(18) Wilker, J. J.; Lippard, S. J. Inorg. Chem. 1997, 36, 969-978.
(24) Breksa, A. P., III; Garrow, T. A. Biochemistry 1999, 38, 13991-
13998.
(25) Allen, J. R.; Clark, D. D.; Krum, J. G.; Ensign, S. A. Proc. Natl.
Acad. Sci. U.S.A. 1999, 96, 8432-8437.
(26) PhTttBu, phenyltris((tert-butylthio)methyl)borate; [Ph(pz)BttBu], phenyl-
(pyrazolyl)bis((tert-butylthio)methyl)borate; [Ph(pztBu)BttBu], phenyl-
(3-tert-butylpyrazolyl)bis((tert-butylthio)methyl)borate.

Zinc Sites in the Methionine Synthases
Inorganic Chemistry, Vol. 39, No. 19, 2000 4349
Scheme 3
tert-butyl protons consistent with C3-symmetric binding of the
borato ligand. The molecular structure of [PhTttBu]ZnBr, as
determined by X-ray diffraction, is described below. An
analogous reaction between [PhTttBu]Tl and CdCl2 yielded
dimeric {[PhTttBu]CdCl}2. Synthesis of a monomeric zinc
thiolate was achieved by metathetical replacement of bromide
in [PhTttBu]ZnBr with phenylthiolate. [PhTttBu]Zn(SPh) is highly
soluble in low-polarity solvents including pentanes and diethyl
ether.
Zinc complexes of the [NS2] ligand [Ph(pz)BttBu] were
prepared in a similar fashion. [Ph(pz)BttBu]Na was generated
in situ via the addition of [Ph(pz)BttBu]H to NaH in THF. Sub-
sequent addition of 1 equiv of ZnBr2 yielded [Ph(pz)BttBu]ZnBr.
The 1H NMR spectrum of [Ph(pz)BttBu]ZnBr exhibits resonances
within ca. 0.1 ppm of those for the “free acid” ligand, absent
the acid proton resonance. Again, the methylene hydrogens
appear as two second-order doublets, a consequence of their
diastereotopic relationship. Reaction of [Ph(pz)BttBu]ZnBr with
KSPh leads to production of [Ph(pz)BttBu]Zn(SPh). However,
this complex is unstable in solution. Over a period of days it
decomposed to a mixture of products that included [Zn(SPh)4]2-.27
The generation of the disproportionation product indicated that
this route would not be satisfactory for the preparation of [Ph-
(pz)BttBu]Zn(SPh). Consequently, all subsequent transformations
utilizing [NS2] borato ligands focused on the sterically more
demanding ligand, [Ph(pztBu)BttBu]. The additional tert-butyl
substituent on the pyrazolyl ring was designed to provide
sufficient steric bulk so as to preclude disproportionation
reactions leading to [Ph(pztBu)BttBu]2Zn and [Zn(SPh)4]2-. As
detailed below, this design works as intended.
(pztBu)BttBu]Zn(SPh) as a white crystalline solid. This compound
was characterized by 1H NMR spectroscopy and X-ray diffrac-
tion. Unlike the less hindered [Ph(pz)BttBu]Zn(SPh), this material
is quite robust, showing no evidence of disproportionation.
Molecular Structures. The solid-state structures of [PhTttBu]-
ZnBr, {[PhTttBu]CdCl}2, and [Ph(pz)BttBu]ZnBr have been
elucidated by X-ray diffraction. Molecular structures, presented
as thermal ellipsoid plots, are contained in Figure 1. Selected
metric parameters are contained in Table 2. [PhTttBu]ZnBr is
isostructural with the previously reported [PhTttBu]NiCl and
[PhTttBu]CoCl.20 This complex is a rare example of a four-
coordinate zinc thioether species that has been structurally
authenticated. The face-capping ligand provides three thioether
donors with an average Zn-S bond length of 2.360 Å. This
distance compares favorably to the EXAFS-derived distance of
2.31 Å for the three sulfur ligands in Met H. Each of the three
tert-butyl groups is canted in the same direction to minimize
van der Waals contacts. The resulting pocket at the metal ion
is rather shallow due to this canting. The Br completes the
approximate tetrahedral zinc geometry residing within the cavity
with Zn-Br of 2.310(1) Å.
{[PhTttBu]CdCl}2 crystallized as a dimer with bridging
chloride ligands, Figure 1B. The molecule resides on a crys-
tallographically imposed inversion center rendering the metal
ligation spheres metrically equivalent. The Cd coordination
geometry is trigonal bipyramidal. One sulfur (S3 or S3A) and
one chloride occupy the axial positions with the remaining
sulfurs and chloride equatorial. Expectedly, the S3-Cd-Cl1A
angle is nearly linear, 174.54(1)°. The equatorial Cd-S distances
of 2.580(1) and 2.592(1) Å are shorter than the axial Cd-S
distance, 2.730(1) Å. This molecule is only the second five-
coordinate Cd thioether complex to be characterized crystallo-
graphically.28
The molecular structure of [Ph(pz)BttBu]ZnBr is contained
in Figure 1C. The [NS2] borato ligand provides one nitrogen
and two sulfur donors in a facial arrangement. The Zn-S
distances, 2.374(1) and 2.402(1) Å, are longer slightly than those
found in [PhTttBu]ZnBr. The Zn-N distance, 2.002(1) Å, is
consistent with bond lengths reported for TpRZnX complexes.6,7
In comparison, the bond distances for the zinc site in Met E
Reaction of [Ph(pztBu)BttBu]H with (CH3)2Zn led to clean
formation of [Ph(pztBu)BttBu]Zn(CH3) and methane, Scheme 3.
In contrast to the spectra of the “free acid” ligands and their
metal complexes, the 1H NMR spectrum of [Ph(pztBu)BttBu]Zn-
(CH3) displayed a single, sharp resonance for the methylene
hydrogens. This observation is interpreted as coincident degen-
eracy of the diastereotopic proton resonances. The remaining
signals for [Ph(pztBu)BttBu]Zn(CH3) are unexceptional including
the zinc-bound methyl protons, δ -0.27. Low-temperature
protonolysis of [Ph(pztBu)BttBu]Zn(CH3) with PhSH yielded [Ph-
(27) Holah, D. G.; Coucouvanis, D. J. Am. Chem. Soc. 1975, 97, 6917-
(28) Griffith, E. A. H.; Charles, N. G.; Rodesiler, P. F.; Amma, E. L.
6919.
Polyhedron 1985, 4, 615-620.

4350 Inorganic Chemistry, Vol. 39, No. 19, 2000
Chiou et al.
these tert-butyl sulfur substituents. Presumably, the minimal
steric requirements of the planar pyrazolyl group favor this
conformation.
The structure of monomeric [Ph(pztBu)BttBu]Zn(SPh) is de-
picted in Figure 2 with selected metric parameters contained in
Table 2. The [NS3] coordination sphere includes two thioether
and one pyrazolyl donor provided by the borato ligand and the
phenylthiolato group in a distorted tetrahedral geometry. Given
the proclivity with which thiolates occupy bridging positions,
the monomeric structure is undoubtedly a consequence of the
steric protection provided by the three tert-butyl substituents
of [Ph(pztBu)BttBu]. The average thioether Zn-S bond length,
2.43 Å, is longer than the thiolato Zn-S, 2.265(1) Å. The
thioether distances are also slightly longer than those found in
[Ph(pz)BttBu]ZnBr. The average of the three Zn-S bond
distances, 2.37 Å, is strikingly close to that in the Hcy bound
form of Met E, 2.31 Å, as determined by EXAFS.9 However,
the Zn-N distance is essentially unchanged at 2.033(3) Å. Again
the two tert-butyl groups on the S donors are canted in opposite
directions, resulting in a small S1-Zn-S2 angle of 86.01(4)°.
The thiolato ligand is bent away from the tert-butyl pyrazolyl
group as evidenced by the large S3-Zn-N angle of 126.4(1)°.
The phenyl substituent of this ligand is on the opposite side of
the Zn-S3 vector from the pyrazole, placing it in a position
where it essentially bisects the S1-Zn-S2 angle; N1-Zn-
S3-C29 dihedral angle, 171.9°. This conformation minimizes
interaction between the 3-tert-butylpyrazolyl substituent and the
phenyl thiolate.
Discussion
To date most efforts in modeling catalytic zinc sites have
focused on CA and LADH. These endeavors have commonly
utilized tripodal ligands that provide nitrogen donors to mimic
the native protein residues. Given that many of the catalytic
sites contain a mixture of donor types, we have chosen to pursue
face-capping ligands that contain donors that more faithfully
model the protein residues. In the present study this requires
either a [S3] ligand for Met H or an [NS2] ligand for Met E.
Our tris((alkylthio)methyl)borato ligands have proven successful
for the former target.19,20 For the latter, recently we developed
a mixed-donor ligand containing one nitrogen and two sulfur
substituents.21 Related mixed-donor ligands for modeling zinc
enzymes have been reported by Parkin,11,30 Vahrenkamp,31 and
Carrano.10 As delineated below, our ligands are distinct and
provide certain advantages we seek to exploit.
The borato ligands provide multiple thioether donors to the
metal. The choice of thioether donors vs thiolate donors allows
for greater control or “tunability” of the ligand’s steric and
electronic properties. The thioether substituent may be system-
atically varied to impart the desired properties at the metal
center. In the present case, the tert-butyl groups prevent the
formation of octahedral [PhTttBu]2Zn, thereby allowing for
[PhTttBu]ZnX complex formation. In the TpR ligands substituent
changes are most commonly employed at the 3-pyrazole position
to direct steric control. As this position is on the carbon adjacent
to the donor nitrogen, the electronic changes provided are
somewhat attenuated. However, in the [PhTtR] ligands substitu-
ent changes have a greater impact on the electronic properties
of the ligand as these groups are bound directly to the donor
sulfur. Furthermore, the thioether donors have a greatly reduced
propensity to bridge metals compared to thiolate ligands.
Figure 1. Thermal ellipsoid plots of (A) [PhTttBu]ZnBr, (B) {[PhTttBu]-
CdCl}2, and (C) [Ph(pz)BttBu]ZnBr. Thermal ellipsoids are drawn at
the 30% probability level, and hydrogen atoms are omitted for clarity.
are Zn-S, 2.31 Å, and Zn-N, 2.03 Å.15 Similar distances are
found for the catalytic site in LADH in which zinc is coordinated
by two cysteines, one histidine, and an exogenous water ligand.29
The S-Zn-S bite angle of 86.91(2)° is significantly smaller
than the S-Zn-N bite angles of 102.85(6) and 100.63(6)°. The
smaller S-Zn-S angle may be a consequence of an unusual
feature of the structure. That is, the two tert-butyl groups are
canted in opposite directions. When the structure is viewed down
the Zn-Br vector, one tert-butylthio group is seen to be rotated
clockwise while the other is rotated counterclockwise. This
conformation greatly minimizes van der Waals contacts between
(29) Ramaswamy, S.; Eklund, H.; Plapp, B. V. Biochemistry 1994, 33,
5230-5237.
(30) Ghosh, P.; Parkin, G. Chem. Commun. 1998, 413-414.
(31) Brand, U.; Vahrenkamp, H. Inorg. Chem. 1995, 34, 3285-3293.

Zinc Sites in the Methionine Synthases
Inorganic Chemistry, Vol. 39, No. 19, 2000 4351
Table 1. Crystallographic Data for [PhTttBu]ZnBr, {[Ph(pz)BttBu]CdCl}2, [Ph(pz)BttBu]ZnBr and [Ph(pztBu)BttBu]Zn(SPh)
[PhTttBu]ZnBr
{[Ph(pz)BttBu]CdCl}2
[Ph(pz)BttBu]ZnBr
[Ph(pztBu)BttBu]Zn(SPh)
formula
fw
space group
a, Å
b, Å
c, Å
C21H38BBrS3Zn
542.78
P21/n
9.74863(2)
21.2707(3)
12.7166(2)
90.00
98.7525(7)
90.00
C44H80B2Cd2Cl6S6
1260.56
P1h
C19H30BBrN2S2Zn
506.66
P1h
C29H43BN2S3Zn
592.01
Pbca
15.6479(2)
18.0911(2)
22.4118(2)
90.00
90.00
90.00
6344.4(1)
8
9.6736(2)
12.3338(2)
13.7419(2)
65.2684(4)
77.4098(4)
77.2047(2)
1437.60(2)
1
9.8538(2)
9.9162(2)
12.0741(2)
95.4226(2)
93.2470(2)
100.1906(8)
1152.74(3)
2
R, deg
â, deg
γ, deg
V, Å3
2606.20(7)
4
Z
cryst color, habit
D(calc), g cm-3
µ(Mo KR), cm-1
temp, K
diffractometer
radiation
R(F), %a
R(wF2), %a
colorless rod
1.383
27.21
colorless plate
1.456
12.65
colorless block
1.460
29.86
colorless plate
1.240
9.91
173(2)
173(2)
173(2)
213(2)
Siemens P4/CCD
Mo KR (λ ) 0.71073 Å)
4.38
10.03
2.58
11.14
3.13
13.92
6.65
11.56
a Quantity minimized ) R(wF2) ) ∑[w(Fo - Fc2)2]/∑[(wFo )2]1/2; R ) ∑∆/∑(Fo), ∆ ) |(Fo - Fc)|, w ) 1/[σ2(Fo ) + (aP)2 + bP], P ) [2Fc2
2
2
2
+ Max(Fo,0)]/3.
Table 2. Selected Bond Lengths (Å) and Bond Angles (deg) for
[PhTttBu]ZnBr, {[Ph(pz)BttBu]CdCl}2, [Ph(pz)BttBu]ZnBr, and
[Ph(pztBu)BttBu]Zn(SPh)
Zn-Br
118.87(3)
97.65(3)
119.80(3)
96.90(3)
98.56(3)
Zn-S1
Zn-S2
Zn-S3
Br-Zn-S1
{[Ph(pz)BttBu]CdCl}2
Cd1-S1
2.5919(5)
2.5802(5)
2.7299(5)
2.5053(5)
2.6870(5)
Cl1-Cd1-Cl1
82.71(2)
94.77(2)
100.55(2)
91.92(1)
87.57(1)
84.12(1)
174.54(2)
Cd1-S2
S2-Cd1-Cl
S1-Cd1-Cl
Cl1-Cd1-S3
S2-Cd1-S3
S1-Cd1-S3
Cl1-Cd1-S3
Cd1-S3
Cd1-Cl1
Cd1-Cl1A
Cl1-Cd1-S2
Cl1-Cd1-S1
S2-Cd1-S1
126.38(2)
138.21(2)
95.08(2)
Zn-Br
102.85(6)
123.58(2)
100.63(6)
122.61(2)
86.91(2)
Zn-N1
Zn-S1
Figure 2. Thermal ellipsoid plot of [Ph(pztBu)BttBu]Zn(SPh). Thermal
ellipsoids are drawn at the 30% probability level, and hydrogen atoms
are omitted for clarity.
Zn-S2
N1-Zn-Br
[Ph(pztBu)BttBu]Zn(SPh)
Zn-S1
2.418(1)
2.441(1)
2.265(1)
2.033(3)
126.42(1)
N1-Zn-S1
101.7(1)
118.99(5)
98.4(1)
116.52(5)
86.01(4)
Zn-S2
S3-Zn-S1
N1-Zn-S2
S3-Zn-S2
S1-Zn-S2
Zn-S3
Zn-N1
N-Zn-S3
Consequently, the ability to control nuclearity outside of a
protein matrix is enhanced by using thioether ligands. Finally,
while cysteine thiolates are the most prevalent sulfur ligands
in proteins, they are often stabilized by hydrogen bonding,
Figure 3. This interaction serves to reduce the nucleophilicity
of the cysteine, to dissipate charge buildup and to modulate
redox properties in cases where cysteines are coordinated to
redox active metal ions. We contend that these properties are
accurately mimicked with thioether donors in aprotic solVents.
The discrete, monomeric thiolates, [PhTttBu]Zn(SPh) and [Ph-
(pztBu)BttBu]Zn(SPh), represent the most accurate small molecule
analogs to date of the zinc active sites in the methionine
synthases. This is because the borato ligands provide appropriate
coordination spheres, replicating the ligand atoms provided by
the protein. [PhTttBu], coordinating through three thioethers,
mimics the three sulfur donors in Met H. The ligand [Ph(pztBu)-
Figure 3. Comparison of coordinated thioether and coordinated thiolate
that also participates in hydrogen bonding.
BttBu] presents the [NS2] coordination environment found in the
cobalamin-independent Met E. In each of these complexes, the
fourth ligand is phenylthiolate. As such it serves to represent
homocysteine coordination to the zinc. Homocysteine coordina-
tion to zinc has been proposed to serve as an activating
mechanism rendering the thiol highly acidic at physiological
pH.2 That is, coordination results in deprotonation making the
homocysteine sufficiently nucleophilic to be alkylated. At
present it is not well understood to what extent metal coordina-
tion reduces the nucleophilicity. Lippard and Wilker demon-
strated that [Zn(SPh)4]2- is alkylated by phosphotriesters via a
dissociative pathway.18,32 Their kinetic analysis is consistent with

4352 Inorganic Chemistry, Vol. 39, No. 19, 2000
Chiou et al.
7.51 (d, 4 H), 7.40 (br, 1 H), 6.93 (t, 1 H), 6.85 (t, 3 H), 6.75 (t, 1 H),
1.83 (q, BCH2, 6 H), 1.20 (s, C(CH3)3, 27 H).
the active species being “free thiolate”. In other words,
dissociated SPh- is at least 2 orders of magnitude more reactive
than the zinc-bound thiolate in [Zn(SPh)4]2-. Alternatively, we
showed through kinetic studies that [Ni(tmc)SPh]+ reacts with
alkyl halides in a bimolecular mechanism in which the nickel-
bound thiolate is alkylated.17 In the latter case, “free thiolate”
reacts more rapidly than bound thiolate, but the dissociation
equilibrium lies in favor of coordinated SPh-.
[Ph(pz)BttBu]H. CH3StBu (12 mL, 90 mmol) and TMEDA (2 mL)
were placed in a 300 mL flask that was vented through an aqueous
solution of NaOCl. BuLi (36 mL, 2.5 M in hexanes) was added
dropwise via syringe over 5 min. After 4 h the solution was cooled to
-78 °C and 40 mL of THF added. The resulting solution was cooled
to -78 °C and added dropwise over 40 min to PhBCl2 (6.1 mL, 45
mmol) in 100 mL of THF. The mixture was allowed to warm to 25 °C
and stirred for an additional 30 min. Lithium pyrazolate, prepared in
situ from pyrazole (3.1 g, 45 mmol) and BuLi (18 mL, 2.5 M in
hexanes), was added to the borane solution at -78 °C. After stirring
for 1 h the reaction was quenched by addition of 100 mL of H2O.
Volatile organics were removed under reduced pressure. The resulting
yellow semisolid was dissolved in CH2Cl2 and dried over CaCl2. The
white product was precipitated from a concentrated solution of CH2-
Cl2 by addition of pentanes. Yield: 5.7 g (35%). [Ph(pz)BttBu]H is
soluble in THF, acetone, chlorinated hydrocarbons, methanol, and
pentane. 1H NMR (CDCl3): δ 13.75 (s, NH, 1 H), 7.94 (d, 5-pz, 1 H),
7.67 (d, 3-pz, 1 H), 7.14 ((o-C6H5)B and (p-C6H5)B, m, 3 H), 6.93 (t,
(m-C6H5)B, 2 H), 6.52 (t, 4-pz, 1 H), 2.23 (d, BCH2, 2 H), 2.04 (d,
BCH2, 2 H), 1.27 (s, C(CH3)3, 18 H). Anal. Calcd (found) for C19H31-
BN2S2: C, 63.0 (62.6); H, 8.62 (8.66); N, 7.73 (7.75).
Summary and Prospectus
This report presents synthetic strategies to access several zinc
complexes including two monomeric zinc thiolates, [PhTttBu]-
Zn(SPh) and [Ph(pztBu)BttBu]Zn(SPh). These complexes repre-
sent the most accurate structural models to date for the active
sites of Met E and Met H in the homocysteine-bound form of
the proteins. Additionally, we include preliminary synthetic
success in the preparation of a related cadmium halide that
serves as a necessary precursor for entry into monomeric
cadmium thiolates. Production of the chloride-bridged dimer,
{[PhTttBu]CdCl}2, highlights the differences between Zn and
Cd coordination preferences, most notably, the greater ease with
which Cd forms five-coordinate complexes. These differences
are clearly significant in the context of activity of cadmium-
substituted zinc proteins. For example, cadmium-substituted Met
E is completely inactive.2 In contrast, when the zinc in the DNA
repair Ada protein is replaced by cadmium, activity is main-
tained.23 Our current goals are directed toward the preparation
of monomeric Cd thiolates and the alkylation of [PhTttBu]Zn-
(SPh) and [Ph(pztBu)BttBu]Zn(SPh) with biologically relevant
carbon electrophiles.
[Ph(pztBu)BttBu]H. The free acid ligand was prepared in a manner
analogous to that used for H[Ph(pz)BttBu], with 3-tert-butylpyrazole
used in place of pyrazole. The solubility of [Ph(pztBu)BttBu]H in pentanes
is greater than that of H[Ph(pz)BttBu]. Additionally, the free acid is
soluble in THF, acetone, methanol, and chlorinated hydrocarbons.
Yield: 5.9 g (31%). 1H NMR (CDCl3): δ 13.58 (s, NH, 1 H), 7.73 (d,
5-pz, 1 H), 7.16 ((o-C6H5)B and (p-C6H5)B, m, 3 H), 6.92 (t, (m-C6H5)B,
2 H), 6.23 (t, 4-pz, 1 H), 2.24 (d, BCH2, 2 H), 2.05 (d, BCH2, 2 H),
1.37 (s, 3-C(CH3)3-pz, 9 H), 1.28 (s, C(CH3)3, 18 H). Anal. Calcd
(found) for C23H39BN2S2: C, 66.0 (66.3); H, 9.39 (9.40); N, 6.69 (6.86).
[Ph(pz)BttBu]ZnBr. [Ph(pz)BttBu]H (3.62 g, 10 mmol) was depro-
tonated using NaH (0.24 g, 10 mmol) suspended in THF to form [Ph-
(pz)BttBu]Na. [Ph(pz)BttBu]Na (115 mg, 0.3 mmol) in 5 mL of CH2Cl2
was added to ZnBr2 (67.5 mg, 0.3 mmol) in 15 mL of CH2Cl2/Et2O,
yielding a clear solution and white solid. After 1 h the solution was
filtered and the solvent removed under reduced pressure. The resulting
white solid was purified by recrystallization from CH2Cl2/Et2O. Yield:
Experimental Section
Materials and Methods. All reagents were purchased from com-
mercial sources and used as received, unless otherwise noted. Solvents
were distilled under N2 and dried as indicated. Toluene, hexanes, and
diethyl ether were freshly distilled over Na/benzophenone. TMEDA
was distilled over 3 Å molecular sieves under reduced pressure. tert-
Butyl methyl sulfide33 and 3-tert-butylpyrazole34 were prepared as
described previously. Manipulations involving organometallic reagents
were performed under a nitrogen atmosphere either on a Schlenk line
or in a Vacuum Atmospheres glovebox. Elemental analyses were
performed by Desert Analytics, Inc., Tuscon, AZ. NMR spectra were
recorded on a 250 MHz Bru¨ker spectrometer.
1
116 mg (77%). H NMR (CDCl3): δ 7.69 (d, 5-pz, 1 H), 7.42 ((o-
C6H5)B, s, 2 H), 7.32 ((p-C6H5)B and (m-C6H5)B, m, 3 H), 7.10 (d,
3-pz, 1 H), 6.17 (t, 4-pz, 1 H), 2.20 (d, BCH2, 2 H), 2.37 (d, BCH2, 2
H), 1.33 (s, C(CH3)3, 18 H). Anal. Calcd (found) for C19H30BBrN2S2-
Zn: C, 45.0 (44.9); H, 5.97 (6.01); N, 5.53 (5.42).
[Ph(pztBu)BttBu]Zn(CH3). Zn(CH3)2 (0.3 mL, 0.6 mmol) was added
to [Ph(pztBu)BttBu]H (250 mg, 0.6 mmol) in 20 mL of toluene. The
solution was stirred for 1 h, and then the solvent was removed under
reduced pressure. The resulting white solid was washed with 15 mL
of pentane and purified by recrystallization from CH2Cl2/Et2O. Yield:
CAUTION. Dimethyl sulfide and tert-butyl mercaptan are flam-
mable liquids that present a pungent odor. The deprotonation reaction
should be Vented through an aqueous solution of NaOCl and NaOH.
[PhTttBu]ZnBr. [PhTttBu]Tl20 (500 mg, 0.83 mmol) in 10 mL of CH2-
Cl2 was added to ZnBr2 (187 mg, 0.83 mmol) in 15 mL of Et2O,
yielding a clear solution and white solid. The solution was stirred for
1 h. After filtration the solvent was removed under reduced pressure,
yielding an off-white solid. The product was washed with Et2O (2 ×
10 mL). The resulting white solid was purified by recrystallization from
1
274 mg (92%). H NMR (CDCl3): δ 7.57 (d, (o-C6H5)B, 2 H), 7.37
(m, (m-C6H5)B and (p-C6H5)B, 3 H), 7.01 (s, 5-pz, 1 H), 6.00 (s, 4-pz,
1 H), 2.28 (s, BCH2, 4 H), 1.47 (s, 3-pz-C(CH3)3, 9 H), 1.37 (s, SC-
(CH3)3, 18 H), -0.27 (s, Zn-CH3, 3 H). Anal. Calcd (found) for C24H41-
BN2S2Zn: C, 57.9 (57.1); H, 8.30 (8.53); N, 5.63 (5.62).
[Ph(pztBu)BttBu]Zn(SPh). Thiophenol (1.72 mL, 0.87 M in toluene,
1.5 mmol) was added to [Ph(pztBu)BttBu]Zn(CH3) (500 mg, 1 mmol) in
20 mL of CH2Cl2 at 0 °C. The solution was stirred for 1 h, and then
the solvent was removed under reduced pressure. The resulting white
solid was washed with 20 mL of pentane and purified by recrystalli-
1
CH2Cl2/Et2O. Yield: 310 mg (68%). H NMR (CDCl3): δ 7.37 (d,
o-C6H5, 2 H), 7.20 (t, m-C6H5, 2 H), 7.10 (t, p-C6H5, 1 H), 2.23 (br,
BCH2, 6 H), 1.45 (s, C(CH3)3, 27 H). Anal. Calcd (found) for C21H38-
BBrS3Zn: C, 46.5 (46.1); H, 7.06 (6.98).
1
zation from CH2Cl2/Et2O. Yield: 520 mg (88%). H NMR (CD2Cl2):
[PhTttBu]Zn(SPh). [PhTttBu]ZnBr (120 mg, 0.22 mmol) in 5 mL of
THF was added to KSPh (33 mg, 0.22 mmol) in 5 mL of THF, yielding
a clear solution. After 12 h, the white precipitate that formed was
removed by filtration. The solvent was removed under reduced pressure,
δ 7.70 (d, (o-C6H5)S, 2 H), 7.55 (d, (o-C6H5)B, 2 H), 7.40 (m, (m-
C6H5)B + (p-C6H5)B, 3 H), 7.25 (t, (m-C6H5)S, 2 H), 7.09 (m, (p-
C6H5)S + 5-pz, 2 H), 6.08 (s, 4-pz, 1 H), 2.44 (d, BCH2, 2 H), 2.23 (d,
BCH2, 2 H), 1.53 (s, 3-C(CH3)3, 9 H), 1.28 (s, SC(CH3)3, 18 H). Anal.
Calcd (found) for C29H43BN2S3Zn: C, 58.8 (58.7); H, 7.32 (7.54); N,
4.73 (4.72).
{[PhTttBu]CdCl}2. [PhTttBu]Tl (550 mg, 0.91 mmol) in 5 mL of CH2-
Cl2 was added to CdCl2 (167 mg, 0.91 mmol) in 15 mL of Et2O,
yielding a clear solution and white solid. The solution was stirred for
2 h. After filtration the solvent was removed under reduced pressure,
1
yielding a white solid. Yield: 92 mg (73%). H NMR (CD3CN): δ
(32) Wilker, J. J.; Lippard, S. J. J. Am. Chem. Soc. 1995, 117, 8682-
8683.
(33) Vogel, A. T.; Cowan, D. M. J. Chem. Soc. 1943, 21, 16-24.
(34) Trofimenko, S.; Calabrese, J. C.; Thompson, J. S. Inorg. Chem. 1987,
26, 1507-1514.

Zinc Sites in the Methionine Synthases
Inorganic Chemistry, Vol. 39, No. 19, 2000 4353
yielding a light yellow solid. The solid was washed with (2 × 10 mL)
portions of Et2O to yield a white solid. Yield: 250 mg (51%). 1H NMR
(CDCl3): δ 7.38 (d, o-C6H5, 2 H), 7.30 (t, m-C6H5, 2 H), 7.15 (t, p-C6H5,
1 H), 2.42 (br, BCH2, 6 H), 1.50 (s, C(CH3)3, 27 H). Anal. Calcd (found)
for C42H76B2Cd2Cl2S6: C, 46.2 (46.0); H, 7.02 (7.15).
on an inversion center. All non-hydrogen atoms were refined with
anisotropic displacement coefficients, and hydrogen atoms were treated
as idealized contributions.
All software and sources of the scattering factors are contained in
the SHELXTL (5.10) program library (G. Sheldrick, Siemens XRD,
Madison, WI).
Crystallographic Structural Determinations. Crystal, data col-
lection, and refinement parameters are given in Table 1. For [PhTttBu]-
ZnBr and [Ph(pztBu)BttBu]Zn(SPh), the systematic absences in the
diffraction data are uniquely consistent for the reported space group,
and no evidence of symmetry higher than triclinic was observed in the
diffraction data of {[PhTttBu]CdCl}2 and [Ph(pz)BttBu]ZnBr. E-Statistics
and the value of Z suggested the centrosymmetric space group option,
P1h, which yielded chemically reasonable and computationally stable
results of refinement. The structures were solved using direct methods,
completed by subsequent difference Fourier syntheses, and refined by
full-matrix least-squares procedures. Empirical SADABS absorption
corrections were applied to all data sets. The asymmetric unit of
{[PhTttBu]CdCl}2 contains half of the cadmium complex, which lies
Acknowledgment. We gratefully acknowledge the financial
support provided by the National Science Foundation (NYI and
CHE-9974628 to C.G.R.) and the University of Delaware.
Supporting Information Available: Tables giving structure de-
termination summary, atomic coordinates, bond lengths and bond
angles, anisotropic thermal parameters, and hydrogen atom parameters
for [PhTttBu]ZnBr, {[PhTttBu]CdCl}2, [Ph(pz)BttBu]ZnBr, and [Ph(pztBu)-
BttBu]Zn(SPh). This material is available free of charge via the Internet
at http://pubs.acs.org.
IC000505Q