A pentacoordinated Di-N-carboxamidodithiolato- O-sulfinato-iron(III) complex related to the metal site of nitrile hydratase

Article abstract of DOI:10.1002/(SICI)1521-3773(19991203)38:23<3526::AID-ANIE3526>3.0.CO;2-Z

Postcoordination oxidation by dioxygen of one of the thiolate groups in a pentadentate N₂S₃ ligand results in an iron(III) complex with two N- carboxamido, two thiolato, and one O-sulfinato ligands (see the CAMERON representation). T

Full text of DOI:10.1002/(SICI)1521-3773(19991203)38:23<3526::AID-ANIE3526>3.0.CO;2-Z

COMMUNICATIONS
diethyl 2-methyl-2-benzylthiomethyl malonate (2)[3a] with
2-methyl-2-(benzylthio)propylamine (3)[3b] in toluene in the
presence of trimethylaluminum gave diamide 4. Upon sub-
sequent reductive deprotection of the thiol groups with
sodium in liquid ammonia, 1 was obtained in 72% yield
(Scheme 1).
A Pentacoordinated Di-N-carboxamido-
dithiolato-O-sulfinato-iron(iii) Complex
Related to the Metal Site of Nitrile Hydratase**
Laurent Heinrich, Yun Li, Jacqueline Vaissermann,
Á
Genevieve Chottard, and Jean-Claude Chottard*
Coordination of iron(iii) to the deprotonated ligand L5
gave (NEt4)2[FeIII(L-O2)] (5), whose X-ray structure shows
postcoordination oxidation of the L ligand (Figure 1).[4] The
intermediate complex formed before oxidation is presumably
the tris-thiolato complex; attempts to isolate and characterize
this species under argon have failed so far.
Nitrile hydratases (NHases) are enzymes that contain non-
heme iron or non-corrinoid cobalt and catalyze the hydration
of nitriles to amides.[1a] One NHase is used for industrial
production of acrylamide.[1b] A recent X-ray structure of the
active form of the NHase from Rhodococcus sp.R312
indicated that the iron center is bound to three cystein
thiolate groups and two nitrogen atoms from peptide bonds of
the protein main chain.[1c] A more precise X-ray structure of
the inactive form of the NHase from Rhodococcus sp.N-771
provided additional information: Two of the coordinated
cystein groups were posttranslationally modified to cystein
sulfinic and sulfenic moieties that are both bound through
sulfur, and one nitric oxide ligand occupies the sixth
position.[1a] Electron nuclear double resonance (ENDOR)
measurements suggested that the sixth ligand in the active
form is probably a hydroxide group that was not seen at the
resolution of the first structure.[1d]
Such an unusual FeIII coordination sphere invites inves-
tigation to better understand its electronic properties and its
catalytic activity in nitrile hydration. This, together with the
mechanism of posttranslational oxidation of bound cysteins,
can be achieved from the study of relevant mimetic com-
plexes. A (N2S)2FeIII complex with two aromatic thiolato, two
aromatic N-carboxamido, and two pyridine ligands has been
reported recently,[2a] and during the writing of the current
paper the same group described the bis-S-sulfinato derivative
of this FeIII complex.[2b]
Figure 1. CAMERON representation (ellipsoid at 30% probability level)
of the anion of complex 5. The H and S32[4] atoms are omitted for clarity.
Selected bond distances [Š] and angles [8]: Fe1 S1 2.223(3), Fe1 S2
2.213(3), Fe1 O3 2.005(6), Fe1 N1 1.946(7), Fe1 N2 1.957(8); S1-Fe1-S2
87.5(1), S1-Fe1-N1 86.7(2), N1-Fe1-N2 91.5(3), S2-Fe1-N2 86.2(3), S1-Fe1-
O3 102.2(2), S2-Fe1-O3 107.5(2), N1-Fe1-O3 96.6(3), N2-Fe1-O3 97.6(3).
To mimic the metallic site of NHase, we need a pentacoor-
dinated FeIII complex with a free sixth coordination site. Such
a complex was synthesized from 1 (H5L; see Scheme 1), a new
H5N2S3 ligand containing three aliphatic thiol groups and two
aliphatic amide moieties. The coupling reaction between
In complex 5, the iron center is above the plane containing
the two deprotonated N-carboxamido and two thiolato
ligands, whereas the O-sulfinato ligand occupies the apical
‡
position. Moreover, the presence of two NEt4 counterions
per iron atom in the crystal lattice indicates that the iron is still
in the ‡3 oxidation state. The Fe N and Fe S bond lengths
are comparable to those reported for complexes showing N-
carboxamido and thiolato coordination to FeIII.[2a] The
FeIII O bond from the O-sulfinato moiety, with a length of
2.005(6) Š, has no precedent to our knowledge.
Complex 5 is soluble in polar solvents (MeCN, DMF),
yielding red solutions. The EPR spectra of 5 at 10 K in MeCN
(Figure 2) or DMF glasses confirm the ‡3 oxidation state of
iron and reveal an intermediate spin state of S ˆ 3/2 with axial
symmetry, in agreement with the magnetic moment of 3.8 mB
(polycrystalline sample) between 2 and 300 K. Complex 5 is
stable under reductive conditions, whereas it undergoes an
irreversible oxidation at E1/2 ˆ 150 mV vs. SCE in DMF, which
might correspond to a ligand oxidation or to the formation of
an FeIV complex. The N-carboxamido ligands are known to
stabilize high oxidation states of transition metal com-
pounds,[5] but the cis position of the two thiolato ligands
might facilitate intramolecular disulfide formation, leading to
irreversible oxidation.
Scheme 1. Synthesis of ligand 1. Bn ˆ benzyl.
[*] Prof. Dr. J.-C. Chottard, Dr. L. Heinrich, Dr. Y. Li
Laboratoire de Chimie et Biochimie
Pharmacologiques et Toxicologiques
Â
Â
Universite Rene Descartes, UMR 8601 CNRS
Á
45, rue des Saints-Peres, F-75270 Paris Cedex 06 (France)
Fax: (‡33)1-42-86-83-87
E-mail: jean-claude.chottard@biomedicale.univ-paris5.fr
Dr. J. Vaissermann, Dr. G. Chottard
Â
Laboratoire de Chimie des Metaux de Transition
Â
Universite Pierre et Marie Curie
ESA 7071, F-75252 Paris Cedex 05 (France)
[**] This work was supported by the European Commission (TMR-
NOHEMIP research network no. ERB FMRX-CT98 ± 0174).
3526
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1433-7851/99/3823-3526 $ 17.50+.50/0
Angew. Chem. Int. Ed. 1999, 38, No. 23

COMMUNICATIONS
1
(EtOAc/cyclohexane 1/3) to afford 4 (410 mg, 79%). H NMR (250 MHz,
CDCl3): d ˆ 7.31 ± 7.19 (m, 17H), 3.72 (s, 2H), 3.69 (s, 4H), 3.25 (d, 4H, J ˆ
5.7 Hz), 2.97 (s, 2H), 1.48 (s, 3H), 1.26 (s, 12H); MS (CI, NH3): m/z: 609
‡
(MH ). 2) To a solution of 4 (400 mg, 0.66 mmol) in THF (4 mL) and liquid
ammonia (6 mL) cooled to 458C were added small portions of Na
(ꢀ91 mg, 4 mmol) until the blue color remained for 30 mn. After addition
of solid NH4Cl (400 mg) and evaporation of ammonia, HCl (2m) was added
(pH 1). Extraction with CH2Cl2 followed by chromatography over silica gel
(EtOAc/cyclohexane 1/4) gave ligand
1
(203 mg, 91%). 1H NMR
(250 MHz, DMSO): d ˆ 7.83 (t, 2H), 3.25 (d, 4H), 3.01 (d, 2H), 2.72 (s,
1
2H), 2.15 (t, 1H), 1.42 (s, 3H), 1.24 (s, 12H); IR (CHCl3): nÄ ˆ 1675 cm
‡
ˆ
(C O); high-resolution MS (CI, CH4): m/z calcd for C13H27O2N2S3 (MH ):
339.1235, found: 339.1234.
5: To a solution of 1 (622 mg, 1.84 mmol) and NaH (224 mg, 9.35 mmol) in
DMF (11 mL) cooled to 58C was added a solution of FeCl3 (296 mg,
1.83 mmol) in DMF (1.5 mL). The deep red solution was stirred at 58C
for 1 h, and then solid Et4NCl (610 mg, 3.68 mmol) was added. After
addition of EtOAc (145 mL), the solution was kept at 208C for one night.
The remaining operations were carried out in air. The red precipitate was
isolated, dried under vacuum, and dissolved in CH3CN. The solution was
then filtered over celite. The solvent was evaporated to afford 5 as a highly
hygroscopic, deep red solid (1.1 g, 87%).
Figure 2. EPR spectrum of 5 in MeCN at 10 K (g? ˆ 3.75, gk ˆ 2.01).
The IR spectrum (KBr pellet) of 5 shows a characteristic
nCO frequency at 1575 cm 1 for coordinated N-carboxamido[2a]
1
and a strong band at 1040 cm corresponding to the nSO
stretching of the O-coordinated sulfinate.[6] The UV/Vis
spectrum of 5 in MeCN exhibits an absorption band at
Received: May 10, 1999
Revised version: July 22, 1999 [Z13397IE]
German version: Angew. Chem. 1999, 111, 3736 ± 3738
1
475 nm (e ˆ 5000 Lmol 1 cm ), which is responsible for the
red color. The resonance Raman spectrum of 5 in MeCN with
an excitation at 476.5 nm shows three bands at 604 (Fe O
1
stretching[7]), 978, and 1160 cm (S O stretching[6]), which
Keywords: iron ´ N ligands ´ nitrile hydratases ´ O ligands ´
support O-sulfinato coordination to the iron as observed in
the solid state. Moreover, the excitation profiles of the three
bands in the region of 450 ± 530 nm enable us to assign the
band at 475 nm in the UV/Vis spectrum to a sulfinate-to-metal
charge-transfer transition.
S ligands
[1] a) S. Nagashima, M. Nakasato, N. Dohmae, M. Tsujimura, K. Takio,
M. Odaka, M. Yohda, N. Kamiya, I. Endo, Nature Struct. Biol. 1998, 5,
347 ± 351, and references therein; b) H. Yamada, M. Kobayashi,
Biosci. Biotechnol. Biochem. 1996, 60, 1391 ± 1400; c) W. Huang, J. Jia,
J. Cumming, M. Nelson, G. Schneider, Y. Lindqvist, Structure 1997, 5,
691 ± 699; d) P. E. Doan, M. J. Nelson, H. Jin, B. M. Hoffman, J. Am.
Chem. Soc. 1996, 118, 7014 ± 7015.
[2] a) J. C. Noveron, M. M. Olmstead, P. K. Mascharak, Inorg. Chem.
1998, 37, 1138 ± 1139; b) L. A. Tyler, J. C. Noveron, M. M. Olmstead,
P. K. Mascharak, Inorg. Chem. 1999, 38, 616 ± 617.
To our knowledge, complex 5 is the first example of an iron
complex with such a mixed coordination sphere and the first
example to result from air oxidation of an iron-coordinated
thiolato ligand to an O-bound sulfinato ligand. Oxidation of
thiolato ± iron complexes by dioxygen is known to give
noncoordinated disulfides or thiolato m-oxo complexes.[8]
A
few examples of sulfinato ± nickel complexes have been
reported that result from nucleophilic attack of bound
thiolate(s) on dioxygen.[9] We do not know yet whether the
oxidation to give 5 starts by activation of dioxygen by the
metal or by reaction of dioxygen with the bound thiolate. Very
recently, dioxygen oxidation of a thiolato ± CoIII complex to a
S-sulfenato and S-sulfinato complexes was reported,[10] and a
S-sulfinato ± FeIII complex has been obtained by H2O2 oxida-
tion of a thiolato ± FeIII complex.[2b] Moreover, O-sulfinato
complexes are known to rearrange into their more stable S-
sulfinato form.[11] In complex 5, the central side chain of the
polydentate ligand might be too short to allow stable
coordination through S. This question is under examination,
as is the possible oxidation of a second thiolato ligand.
[3] a) H. Böhme, H. G. Greve, Chem. Ber. 1952, 85, 409 ± 414; b) F. I.
Carroll, J. D. White, M. E. Wall, J. Org. Chem. 1963, 28, 1240 ± 1243.
[4] X-ray crystal structure analysis of 5: A red single crystal (0.5 Â 0.5 Â
0.5 mm3) of 5, (C13H21FeN2O4S3)(NEt4)2, grown from DMF/Et2O, was
glued in araldite; Nonius CAD4 diffractometer, Tˆ 253 K. Ortho-
rhombic space group P212121, a ˆ 10.912(4), b ˆ 17.723(6), c ˆ
18.699(7) Š, Vˆ 3616(2) Š3, Z ˆ 4, 1calcd ˆ 1.25 gcm 3. Anomalous
dispersion terms and correction of secondary extinction were applied.
The structure was solved by SHELXS-86 and refined by least-squares
analysis using anisotropic thermal parameters for the atoms of the
anionic complex and isotropic thermal parameters for the cation
‡
Et4N . The hydrogen atoms were introduced in calculated positions;
2335 reflections [Fo > 3s(Fo)], R1 ˆ 0.0727, wR2 ˆ 0.0877, 292 least-
squares parameters. The programs CRYSTALS and CAMERON
were used. Two equivalent positions S31 and S32 are found in the crystal
lattice due to positional disorder resulting from the coordination of
either O3 or O4. Crystallographic data (excluding structure factors)
for the structure reported in this paper has been deposited with the
Cambridge Crystallographic Data Centre as supplementary publica-
tion no. CCDC-120197. Copies of the data can be obtained free of
charge on application to CCDC, 12 Union Road, Cambridge CB2
1EZ, UK (fax: (‡44)1223-336-033; e-mail: deposit@ccdc.cam.ac.uk).
[5] T. J. Collins, K. L. Kostka, E. Münck, E. S. Uffelman, J. Am. Chem.
Soc. 1990, 112, 5637 ± 5639.
Experimental Section
All operations were carried under argon, unless mentioned otherwise, with
standard Schlenk techniques. Solvents were dried and distilled before use.
1: 1) A solution of 2[3a] (263 mg, 0.85 mmol) was added under argon at 08C
to a mixture of 3[3b] (500 mg, 2.56 mmol) and AlMe3 (2m in heptane,
1.7 mL) in toluene (8 mL). After the mixture was heated at reflux for one
night, the reaction was quenched by HCl (2m). Extraction with EtOAc
gave a crude product, which was purified by chromatography over silica gel
[6] G. Vitzthum, E. Lindner, Angew. Chem. 1971, 83, 315; Angew. Chem.
Int. Ed. Engl. 1971, 10, 315 ± 326.
[7] a) R. Y. N. Ho, G. Roelfes, B. L. Feringa, L. Que, J. Am. Chem. Soc.
1999, 121, 264 ± 265; b) T. C. Higgs, D. Ji, R. S. Czernuscewicz, C. J.
Carrano, Inorg. Chim. Acta 1999, 286, 80 ± 92.
Angew. Chem. Int. Ed. 1999, 38, No. 23
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1433-7851/99/3823-3527 $ 17.50+.50/0
3527

COMMUNICATIONS
[8] G. Musie, C. H. Lai, J. H. Reibenspies, L. W. Sumner, M. Y. Dare-
nsbourg, Inorg. Chem. 1998, 37, 4086 ± 4093.
[9] C. A. Grapperhaus, M. Y. Darensbourg, Acc. Chem. Res. 1998, 31,
451 ± 459, and references therein.
[10] K. Yamanari, T. Kawamoto, Y. Kushi, T. Komorita, A. Fuyuhiro, Bull.
Chem. Soc. Jpn. 1998, 71, 2635 ± 2643.
[11] F. M. D. Akhter, M. Hirotsu, I. Sugimoto, M. Kojima, S. Kashino, Y.
Yoshikawa, Bull. Chem. Soc. Jpn. 1996, 69, 643 ± 653.
The First Cyclodiasteromeric [3]Rotaxane**
Roland Schmieder, Gosia Hübner, Christian Seel, and
Fritz Vögtle*
Cycloenantiomerism in mechanically bonded molecules
was first predicted by Frisch and Wassermann in 1961,[1] and
the first catenanes and molecular knots displaying this
property were synthesized by Sauvage et al. almost three
decades later.[2, 3] As early as 1971, Schill proposed stereo-
isomeric [2]rotaxanes,[4] whose chiral information can be put
down to a directed segment sequence in the macrocycle as
well as in the axle. The axle and wheel are not chiral
themselves; the mechanical connection, however, would yield
a cycloenantiomeric rotaxane.
During the last decade several chiral [2]rotaxanes have
been synthesized whose stereoisomerism is based on the
central chirality.[5, 6] In 1997 we prepared the first cyclo-
enantiomeric [2]rotaxanes and [1]rotaxanes as well as topo-
logically chiral pretzelanes, which were separated into their
enantiomers by means of chiral HPLC and chiroptically
characterized.[6e] Recently several achiral [3]rotaxanes have
been reported.[7] To our knowledge stereoisomeric [3]rotax-
anes, however, have not been described yet.
We now succeeded in preparing a chiral [3]rotaxane
containing two achiral wheels, which are mechanically bonded
onto an achiral axle that is not directed. Analogous to the
covalently linked tartaric acid, we obtained a cyclodiastereo-
meric compound.[8] Applying our new efficent trapping
synthesis (chemical threading) for rotaxanes with diether
axles,[9] we allowed the dibromide 2 to react with the stopper 3
in presence of the wheel 4. Besides small amounts of the free
axle and the [2]rotaxane (10%), the [3]rotaxane 1 was
obtained in 29% yield (Scheme 1).[10]
Scheme 1. Schematic representation of the synthesis of the cylodiastereo-
meric [3]rotaxane as a pair of enantiomers (1b, 1c) and as the meso form
(1a).
meso form (1a) and a pair of enantiomers (1b, 1c). The
orientation of the macrocycles on the axle is caused by a
different sequence of the three amide groups and the
sulfonamide group. The two wheels 4 can have the same or
opposite direction. In the case of the same orientation the
meso compound 1a is obtained, and the opposite direction of
the wheels leads to 1b and 1c.
The high conformational flexibility of the moleculeÐin the
sense of the extended translational and rotational movement
of the wheels on the axleÐraises the questions of whether the
stereoisomers can be separated and whether chiroptical
properties show significant differences. The results obtained
hitherto on our separated chiral [2]- and [1]rotaxanes and
catenanes indeed fueled our hopes for success. The enan-
tiomers were successfully separated employing chiral HPLC
(Chiralpak AD; Figure 1).[11, 12] First of all the (‡) enantiomer
was eluted in base line separation, followed by the ( )
enantiomer. The retention time of the latter, however,
differed only slightly from that of the meso form 1a (Fig-
ure 1). Whereas the separation factor a(
of the enan-
),(‡)
If there is only one achiral wheel on a symmetrical axle, a
[2]rotaxane with no stereoisomerism results. Rotaxanes like 1
bearing two wheels, the atomic sequences of which can be
arranged clockwise or counterclockwise, should occur in a
[*] Prof. Dr. F. Vögtle, R. Schmieder, G. M. Hübner, Dr. C. Seel
Â
Kekule-Institut für Organische Chemie und
Biochemie der Universität
Gerhard-Domagk-Strasse 1, D-53121 Bonn (Germany)
Fax: (‡49)228-735-662
E-mail: voegtle@uni-bonn.de
[**] We thank Björna Windisch and Frank Schwanke as well as Dr.
Thomas Dünnwald, Kernforschungszentrum Jülich GmbH, for initial
work and help, and the Deutsche Forschungsgemeinschaft for
financial support (DFG,Vo145/47 ± 2).
Figure 1. Chromatogram for the enantiomeric separation. Retention
times: (‡) enantiomer: 10.5 min, ( ) enantiomer: 21.5 min, meso form:
23.5 min.[12]
3528
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1433-7851/99/3823-3528 $ 17.50+.50/0
Angew. Chem. Int. Ed. 1999, 38, No. 23