Redox regulation of protein tyrosine phosphatase 1B (PTP1B): A biomimetic study on the unexpected formation of a sulfenyl amide intermediate

Article abstract of DOI:10.1021/ja070410o

The effect of steric and electronic environments around the sulfur and nitrogen atoms and the role of nonbonded S...O/N interactions on the cyclization reactions of amide substituted benzene sulfenic acids are described. The reaction profiles and the role

Full text of DOI:10.1021/ja070410o

Published on Web 06/22/2007
Redox Regulation of Protein Tyrosine Phosphatase 1B
(PTP1B): A Biomimetic Study on the Unexpected Formation
of a Sulfenyl Amide Intermediate
Bani Kanta Sarma and Govindasamy Mugesh*
Contribution from the Department of Inorganic and Physical Chemistry, Indian Institute of
Science, Bangalore 560 012, India
Received January 19, 2007; E-mail: mugesh@ipc.iisc.ernet.in
Abstract: The effect of steric and electronic environments around the sulfur and nitrogen atoms and the
role of nonbonded S‚‚‚O/N interactions on the cyclization reactions of amide substituted benzene sulfenic
acids are described. The reaction profiles and the role of different substituents on the cyclization are
investigated in detail by theoretical calculations. It is shown that the synthetic thiols having ortho-amide
substituents may serve as good models for the enforced proximity of the amide and cysteine thiol groups
at the active site of protein tyrosine phosphatase 1B (PTP1B). However, some of the sulfenic acids derived
from such models do not effectively mimic the cyclization of protein sulfenic acids. This is mainly due to
the requirement of very high energy for breaking the S-O bond to form a planar five-membered ring of
isothiazolidinone. It is shown that the sulfenic acid having two substituentssan amide moiety and a
heterocyclic groupsin the ortho-positions undergoes a rapid cyclization reaction to produce the corre-
sponding sulfenyl amide species. These studies reveal that the introduction of a substituent at the 6-position
of the benzene ring enhances the cyclization process not only by facilitating a closer approach of the -OH
group and the backbone -NH moiety but also by increasing the electrophilicity of the sulfur atom in the
sulfenic acid.
Introduction
Protein tyrosine phosphatases (PTPs) regulate signal trans-
duction pathways involving tyrosine phosphorylation, and these
enzymes have been implicated in the development of cancer,
diabetes, rheumatoid arthritis, and hypertension.1 In particular,
the structure and function of the mammalian protein tyrosine
phosphatase 1B (PTP1B) have attracted significant interest in
recent years due to the importance of this enzyme in insulin
signaling.2 The reaction mechanism of PTP1B involves a
nucleophilic attack of a catalytically active cysteine residue
(Cys215) at a phosphotyrosine (pTyr) substrate, resulting in a
covalent phosphocysteine intermediate that is subsequently
hydrolyzed by a water molecule (Figure 1A).3
Figure 1. (A) Dephosphorylation cycle of PTP1B. (B) Redox regulation
cycle of PTP1B. (C) Participation of the unexpectedly formed sulfenyl amide
in the reversible redox control of PTP1B.
Recent studies suggest that the cellular redox state is involved
in regulating tyrosine phosphatase activity through the reversible
oxidation of the catalytic cysteine to the corresponding sulfenic
acid (Cys-SOH).4 The sulfenic acid reacts with thiols such as
glutathione (GSH) to regenerate the thiol, and this process
maintains the redox state of the protein (Figure 1B).5 The
importance of sulfenic acids in biology is well recognized in
recent years, and the diverse role of this species in various redox
reactions involving proteins and synthetic compounds has been
shown to be extraordinary.5-7 Although many enzymes protect
their active site cysteines from overoxidation and irreversible
inactivation by forming an internal disulfide bond with a second
(1) (a) Neel, B. G.; Tonks, N. K. Curr. Opin. Cell Biol. 1997, 9, 193-204. (b)
Hunter, T. Cell 2000, 100, 113-117. (c) Zhang, Z. Y. Annu. ReV.
Pharmacol. Toxicol. 2002, 42, 209-234. (d) Tonks, N. K. Cell 2005, 121,
667-670.
(2) (a) Johnson, T. O.; Ermolieff, J.; Jirousek, M. R. Nat. ReV. Drug DiscoVery
2002, 1, 696-709. (b) Hooft van Huijsduijnen, R.; Sauer, W. H. B.;
Bombrun, A.; Swinnen, D. J. Med. Chem. 2004, 47, 4142-4146.
(3) (a) Denu, J. M.; Tanner, K. G. Biochemistry 1998, 37, 5633-5642. (b)
Barrett, W. C.; DeGnore, J. P.; Keng, Y.-F.; Zhang, Z.-Y.; Yim, M. B.;
Chock, P. B. J. Biol. Chem. 1999, 274, 34543-34546. (c) Chiarugi, P.;
Fiaschi, T.; Taddei, M. L.; Talini, D.; Giannoni, E.; Raugei, G.; Ramponi,
G. J. Biol. Chem. 2001, 276, 33478-33487. (d) Lee, S.-R.; Yang, K.-S.;
Kwon, J.; Lee, C.; Jeong, W.; Rhee, S. G. J. Biol. Chem. 2002, 277, 20336-
20342. (e) Sohn, J.; Rudolph, J. Biochemistry 2003, 42, 10060-10070.
(4) Barford, D. Curr. Opin. Struct. Biol. 2004, 14, 679-686.
(5) Seth, D.; Rudolph, J. Biochemistry 2006, 45, 8476-8487, and references
therein.
9
8872
J. AM. CHEM. SOC. 2007, 129, 8872-8881
10.1021/ja070410o CCC: $37.00 © 2007 American Chemical Society

Redox Regulation of Protein Tyrosine Phosphatase 1B
A R T I C L E S
Figure 2. A part of the crystal structure of PTP1B showing the sulfenyl
amide (PDB Code: 1OEM).8a
cysteine residue within the active site or a mixed disulfide bond
with external thiols,5,6 the sulfenic acid intermediate produced
in response to PTP1B oxidation is rapidly converted into a
sulfenyl amide species (Figure 2), in which the sulfur atom of
the catalytic cysteine is covalently bonded to the main chain
nitrogen of an adjacent residue (Figure 1C).8 This novel and
unusual protein modification has been shown to protect the
active site cysteine from irreversible oxidation to sulfinic (E-
SO2H) or sulfonic (E-SO3H) acids.8 Similarly to the disulfide
bonds in proteins, the formation of sulfenyl amide is reversible
and the cleavage of the S-N bond by cellular thiols such as
glutathione (GSH) converts the inactivated protein back to its
catalytically active form. Furthermore, the inactivated PTP1B
can also be reactivated enzymatically by cysteine-containing
redox proteins such as thioredoxin or glutaredoxin.5
Recently, the conversion of a benzanilide-derived sulfenic
acid (3), generated by thermolysis of the corresponding â-sulfi-
nyl propionic acid ester, to the corresponding 3-isothiazolidinone
(5) has been shown to be a chemical model for the cyclization
reaction.9 Although the nitrogen atoms of amides are generally
considered poor nucleophiles due to the conjugation of the amide
nitrogen with the carbonyl group,9,10 this interesting model study
suggested that the sulfenic acid residue may have sufficient
electrophilicity to drive the cyclization with an amide backbone.
This model study also suggested that the further conversion of
the sulfenic acid to a sulfenyl peroxide or a mixed disulfide
may not be required for the oxidative transformation of PTP1B
Figure 3. Proposed mechanism showing the participation of His214 in
the cyclization of sulfenic acid state of PTP1B to the sulfenyl amide.
to its inactive 3-isothiazolidinone form. During our attempts to
mimic the action of PTP1B with synthetic models, we observed
that the cyclization of aromatic sulfenic acids bearing mono-
substituted ortho-amide substituents to the corresponding sulfe-
nyl amides requires a very large activation energy, which can
be brought down dramatically by introducing an additional
substituent having heteroatoms at the 6-position that can interact
with the sulfur atom. Therefore, it is important to consider the
steric and electronic effects around sulfur11 in PTP1B for
designing synthetic models for the active site of the enzyme.
In this paper, we describe an excellent model for the redox
regulation of PTP1B and show the first example of how the
S‚‚‚O/N nonbonded interactions involving sulfur and other
heteroatoms in the sulfenic acid intermediate modulate the
cyclization process. We also show that the formation of a
sulfenyl amide species occurs even in the presence of thiols
when the electrophilicity of the sulfur in the sulfenic acid
intermediate is enhanced by S‚‚‚N interactions and the -S-
OH group of the sulfenic acid adopts a strained conformation.
Results and Discussion
As mentioned in the introduction, the oxidation of cysteine
residue in PTP1B by H2O2 produces the corresponding sulfenic
acid intermediate, which undergoes an unusual modification to
produce a sulfenyl amide species.8 In these studies, Barford et
al. and Jhoti et al. have independently shown that the protein
modification occurs through oxidation of Cys215 to sulfenic
acid, followed by a nucleophilic attack of the backbone nitrogen
atom of Ser216 on the sulfur atom of Cys215. This cyclization
reaction has been shown to be facilitated by a hydrogen bonding
between the carbonyl oxygen atom of Cys215 and Nδ1 atom
of the His214, which increases the partial charge on the
backbone nitrogen atom of Ser216 (Figure 3).8
(6) (a) Yeh, J. I.; Claiborne, A.; Hol, W. G. J. Biochemistry 1996, 35, 9951-
9957. (b) Giles, G. I.; Jacob, C. Biol. Chem. 2002, 383, 375-388. (c) Jacob,
C.; Giles, G. I.; Giles, N. M.; Sies, H. Angew. Chem., Int. Ed. 2003, 42,
4742-4758. (d) Noguchi, T.; Nojiri, M.; Takei, K.; Odaka, M.; Kamiya,
N. Biochemistry 2003, 42, 11642-11650. (e) Jacob, C.; Holme, A. L.;
Fry, F. H. Org. Biomol. Chem. 2004, 2, 1953-1956.
(7) (a) Nakamura, N. J. Am. Chem. Soc. 1983, 105, 7172-7173. (b) Goto, K.;
Tokitoh, N.; Okazaki, R. Angew. Chem., Int. Ed. 1995, 34, 1124-1126.
(c) Saiki, T.; Goto, K.; Tokitoh, N.; Okazaki, R. J. Org. Chem. 1996, 61,
2924-2925. (d) Ishii, A.; Komiya, K.; Nakayama, J. J. Am. Chem. Soc.
1996, 118, 12836-12837. (e) Goto, K.; Holler, M.; Okazaki, R. J. Am.
Chem. Soc. 1997, 119, 1460-1461. (f) Okazaki, R.; Goto, K. Heteroat.
Chem. 2002, 13, 414-418. (g) Goto, K.; Shimada, K.; Nagahama, M.;
Okazaki, R.; Kawashima, T. Chem. Lett. 2003, 32, 1080-1081. (h) Goto,
K.; Shimada, K.; Furukawa, S.; Miyasaka, S.; Takahashi, Y.; Kawashima,
T. Chem. Lett. 2006, 35, 862-863.
In our study, we first evaluated the conversion of the
benzanilide-derived sulfenic acid (3) to the corresponding
3-isothiazolidinone (5) in acetonitrile by using an HPLC method.
(8) (a) Salmeen, A.; Anderson, J. N.; Myers, M. P.; Meng, T.-C.; Hinks, J.
A.; Tonks, N. K.; Barford, D. Nature 2003, 423, 769-773. (b) van Montfort,
R. L. M.; Congreeve, M.; Tisi, D.; Carr, R.; Jhoti, H. Nature 2003, 423,
773-777.
(11) (a) Steric and electronic factors have been shown to play important roles
in the stability of sulfenic acids. Davis, F. A.; Jenkins, L. A.; Billmers, R.
L. J. Org. Chem. 1986, 51, 1033-1040. (b) For an excellent article on
sulfenic acids and sulfur-nitrogen compounds, see: Davis, F. A. J. Org.
Chem. 2006, 71, 8993-9003 and references therein.
(9) Sivaramakrishnan, S.; Keerthi, K.; Gates, K. S. J. Am. Chem. Soc. 2005,
127, 10830-10831.
(10) Gajda, T.; Zwierzak, A. Synthesis 1981, 1005-1008.
9
J. AM. CHEM. SOC. VOL. 129, NO. 28, 2007 8873

A R T I C L E S
Sarma and Mugesh
Scheme 1. Conversion of Thiols to the Corresponding Sulfenyl
Amides in the Presence of H2O2
of Domagala et al. that 2,2′-dithiobisbenzamides exit in equi-
librium with the corresponding 1,2-benisothiazolin-3(2H)-ones
in solution.14 Therefore, the conversion of 1 to the sulfenyl
amide 5 by reaction with H2O2 was studied in acetonitrile to
avoid the possible conversion of the disulfide (7) to compound
5. In this case, the reaction sequence involving oxidation of
the thiol moiety followed by cyclization of the resulting sulfenic
acid was found to be extremely slow and the disulfide was
isolated from the reaction in nearly quantitative yield, supporting
the assumption that the disulfide 7 is the major precursor for
compound 5. The yield of the disulfide 7 was unaffected when
the reaction was performed in the presence of an excess amount
of methyl iodide in acetonitrile. Although methyl iodide blocks
the potential dimerization of the sulfenic acid 3 to the corre-
sponding thiosulfinate (RS-(O)-SR),9 this does not block the
reaction between the thiol 1 and the sulfenic acid 3 to produce
the disulfide 7.15 Interestingly, the sulfenic acid 3 does not
produce the disulfide 7 when it is generated by a route that
does not involve a thiol species, e.g., the thermolysis of the
corresponding â-sulfinyl propionic acid ester.9 It should be
mentioned that the oxidation of PTP1B-SH does not produce
any disulfide species and the formation of a protein-bound
sulfenyl amide from the thiol, therefore, should occur through
the generation of the corresponding sulfenic acid as shown in
Figure 1.
Table 1. Initial Rates (v0) for the Cyclization of 1 and 2 by H2O2
at 23 °Ca
initial rate, v0
1
entry
substrate
(µM
‚
min-
)
1
2
3
4
1
0.056
0.118
0.015
0.124
1 + H2O2
2
2 + H2O2
a The rates were calculated by following the initial formation of the
sulfenyl amides using a reversed-phase HPLC method. Conditions: [1 or
2] ) 0.17 mM, [H2O2] ) 0.45 mM. The reactions were carried out in MeCN.
To understand the effect various substituents on the cycliza-
tion, we have chosen to study compound 2 having an alcoholic
moiety attached to the amide side chain. Our initial assumption
was that the alcoholic group in 2 might mimic the structural
features of the serine residue (Ser216, Figure 3) attached to the
nitrogen atom of the amide backbone in PTP1B. Similarly to
the oxidation of 1, the addition of H2O2 to 2 produced the cyclic
compound 6, probably through the formation of the sulfenic
acid intermediate (4). As in the case of 1, the sulfenic acid
derived from 2 could not be isolated due to its instability in
solution and possible conversion to other oxidized species such
as sulfinic and sulfonic acids. However, the rate of cyclization
of 2 in the presence of H2O2 was found to be identical with
that of 1 (Table 1), indicating that the replacement of the phenyl
group in 1 by an alcoholic moiety does not have any significant
effect on the cyclization process. In agreement with our
observations on compound 1, the NMR and HPLC studies
indicated the formation of disulfide 8 as the major product,
which disproportionated in solution to afford the sulfenyl amide
6 as a stable product. The disulfide 8 was isolated and
characterized by NMR and mass spectral techniques. The single-
crystal X-ray structure of compound 8 reveals unusual S‚‚‚O
noncovalent interactions between the sulfur and amide backbone,
which probably facilitate the disproportionation reaction in
solution.
The addition of H2O2 to thiol 1 produced the cyclic compound
5, presumably Via the sulfenic acid 3 derivative (Scheme 1).9
In agreement with the previous report,9 the sulfenic acid 3 could
not be isolated due to the instability of this species in solution.
However, the reaction of 1 with H2O2 to produce the sulfenyl
amide 5 was found to be extremely slow (Table 1). On the other
hand, the reaction of 1 with H2O2 in acetonitrile afforded the
disulfide 7 as the major product. The identity of 7 was confirmed
by comparing the NMR data and HPLC profiles for the major
product of this reaction with that of the authentic sample of 7.
The disulfide was found to be unstable in solution, and this
compound underwent a disproportionation reaction12 to produce
the sulfenyl amide 5. A similar type of cyclization has been
previously observed for some cystinyl peptide disulfides, which
are oxidized readily by bromine to give the corresponding
dipeptide isothizolidinones.13 The disproportionation reaction
leading to the formation of 5 is found to be highly solvent
dependent. To understand the effect of various solvents on the
disproportionation reaction, we have determined the rate of
conversion of the disulfide 7 (0.05 mM) to the cyclic compound
5 in water, acetonitrile, and methanol and found that the initial
rate for the disproportionation of the disulfide 7 to the sulfenyl
amide 5 in acetonitrile (0.049 µM‚min-1) is much slower as
compared to that in methanol (0.244 µM‚min-1) and water
(0.094 µM‚min-1). This is in accordance with the observation
To further understand the reason for the poor reaction rates
observed for the cyclization of 3 and 4 to generate the
corresponding sulfenyl amides, we carried out density functional
theory (DFT) calculations using the B3LYP/6-31G(d) level of
theory (see Supporting Information for all optimized geom-
(12) The disproportionation of the disulfide 7 in phosphate buffer (0.25 M, pH
7.5 at 25 °C) affords a 1:1 mixture of the sulfenyl amide 5 and the thiol 1.
In the presence of air and H2O2, the thiol produced in this reaction undergoes
an auto-oxidation and H2O2-mediated oxidation, respectively, to reproduce
the disulfide, and this process continues until most of the thiol/disulfide
mixture is converted to the sulfenyl amide 5. It should be noted that the
formation of the disulfide 7 in the H2O2-mediated oxidation of 1 is mainly
due to the reaction between the thiol 1 and the sulfenic acid 3 and not due
to the dimerization of the sulfenic acid. The formation of compounds 1
and 5 in the disproportionation of 7 was followed by HPLC, and the
products were compared with authentic samples.
(14) Domagala, J. M.; Bader, J. P.; Gogliotti, R. D.; Sanchez, J. P.; Stier, M.
A.; Song, Y.; Prasad, J. V. N. V.; Tummino, P. J.; Scholten, J.; Harvey,
P.; Holler, T.; Gracheck, S.; Hupe, D.; Rice, W. G.; Schultz, R. Bioorg.
Med. Chem. 1997, 5, 569-579.
(15) For involvement of sulfenic acids in the oxidation of thiols to produce
disulfides, see: Davis, F. A.; Billmers, R. L. J. Am. Chem. Soc. 1981,
103, 7016-7018.
(13) Shiau, T. P.; Erlanson, D. A.; Gordon, E. M. Org. Lett. 2006, 8, 5697-
5699.
9
8874 J. AM. CHEM. SOC. VOL. 129, NO. 28, 2007

Redox Regulation of Protein Tyrosine Phosphatase 1B
A R T I C L E S
etries). These calculations indicated the presence of strong
intramolecular S‚‚‚O interactions in the sulfenic acids 3 and 4
with S‚‚‚O distances of 2.46 and 2.45 Å, respectively, which
are much shorter than the sum of the van der Waals radii of
sulfur and oxygen atoms (3.25 Å). These interactions force the
-OH group to adapt a position trans to the S‚‚‚O interaction,
leading to an almost linear arrangement of the O‚‚‚S-O moiety
(bond angle: 176.6° for 3 and 176.8° for 4). As a result of this
unusual arrangement, the cyclization partners, i.e., the -NH-
and -OH functionalities are positioned in opposite directions,
which may hamper a facile cyclization involving the -NH-
and -OH groups. Further insights into the relative rate of
cyclization of the sulfenic acids were obtained from the DFT
calculations by comparing the activation energy profiles for the
conversion of the sulfenic acids to the corresponding sulfenyl
amides. These calculations suggest that the cyclization of a
sulfenic acid having an amide moiety in the ortho position needs
to overcome several energy barriers before its possible conver-
sion to the sulfenyl amide. The major steps for the cyclization
of sulfenic acids include the following: (i) breaking of the
intramolecular S‚‚‚O nonbonding interaction and rotation of the
amide -NH- group toward the sulfur atom, (ii) rotation of the
-OH group toward the -NH- group of the amide moiety, and
(iii) elimination of a water molecule to form the expected cyclic
sulfenyl amide. The optimized geometries of various intermedi-
ates and transition states (TS) involved in the cyclization reaction
of 4 are summarized in Figure 4. The relative electronic energies
(including the ZPVE correction) in the gas phase calculated at
the B3LYP/6-31G(d) level of theory for the cyclization of 4
are summarized in Table 2, and the corresponding energy profile
is shown in Figure 5.
In the first step of conversion of 4 to 6, it is assumed that the
amide nitrogen atom moves toward the sulfur atom by breaking
the S‚‚‚O nonbonded interaction to form 4a. The conversion of
4 to 4a through the transition state 4TSI involves weakening
of the S‚‚‚O nonbonded interaction, which leads to an increase
in the energy by 4.6 kcal/mol. This conversion also brings down
the distance between sulfur and the amide nitrogen atom from
4.36 Å to 2.94 Å. The energy for the 4TSI was calculated to
be 7.76 kcal/mol higher than that of 4. In the second step, the
hydroxyl group moves toward the amide nitrogen via 4TSII,
leading to the formation of 4b, which is only 0.67 kcal/mol
more stable than the sulfenic acid 4a. The conversion of 4b to
4c is found to be the rate-determining step for the overall pro-
cess. The closer approach of the -OH group toward the amide
nitrogen leads to the elimination of a water molecule. At this
stage, the formation of a hydrogen bonding was observed be-
tween the sulfur atom and the -NH- group with a N-H‚‚‚S
distance of 2.42 Å. The conversion of 4b to 4c takes place
via 4TSIII, for which the energy was calculated to be 49.29
kcal/mol higher than that of 4b. The transition state 4TSIII is
raised very high in energy and the formation of a water molecule
is almost complete as evidenced by a very short O-H distance
of 1.03 Å, which is very close to that of a covalent O-H bond.
The most interesting feature in 4TSIII (see Figure 4) is that
the nitrogen atom keeps itself away from the sulfur atom
(N‚‚‚S ) 2.68 Å) with a 15.2° deviation from the aryl-C-S
plane. In the structure 4c, the water molecule is bound to the
product via two strong hydrogen bonds. Interestingly, the
hydrogen bonds appear to stabilize the sulfenyl amide. There-
Figure 4. B3LYP/6-31G(d) level optimized geometries of different
intermediates and transition states involved in the cyclization of the sulfenic
acid 4 to the cyclic sulfenyl amide 6.
Figure 5. Energy (∆Eelec + ZPVE) profiles for the cyclization of 4 to the
corresponding sulfenyl amide 6.
fore, the removal of the water molecule to form the final product
6 is found to be an endothermic process by 10.7 kcal/mol.
These observations reveal that some of the amino acid
residues in the active site of PTP1B may facilitate the cyclization
by providing steric and/or electronic effects. As shown in Figure
9
J. AM. CHEM. SOC. VOL. 129, NO. 28, 2007 8875

A R T I C L E S
Sarma and Mugesh
Table 2. Relative Electronic ∆(E + ZPE) and Relative Gibbs Free
Energy ∆(G + ZPE) for the Cyclization of Sulfenic Acid 4 to the
Corresponding Sulfenyl Amides 6 Calculated at the B3LYP/
6-31G(d) Level of Theory
4
f
6
+
H2O
4TSIII
4
4TSI
4a
4TSII
4b
4c
6 + H2O
∆(E + ZPE)
∆(G + ZPE)
0
0
7.76 4.60 7.51 3.92 53.21 -9.23
1.47
8.02 4.47 7.68 4.36 54.02 -8.57 -7.21
3, the hydrogen bond between the carbonyl oxygen atom of
Cys215 and the Nδ1 atom of His214 not only increases the
partial negative charge on the nitrogen but also may prevent
the possible S‚‚‚O interactions in the sulfenic acid intermediate.
A careful analysis of the reported crystal structure of PTP1B-
SOH reveals that the carbonyl oxygen atom lies away from the
sulfur atom (S‚‚‚O: 4.518 Å),8 confirming the absence of any
S‚‚‚O interactions. The theoretical calculations on sulfenic acids
3 and 4 also suggest that the absence of S‚‚‚O interactions alone
is not sufficient for an effective cyclization, as the sulfenic acid
intermediate such as 4b, in which the S‚‚‚O interactions are
Figure 6. B3LYP/6-31G(d) level optimized geometries of different
intermediates and transition states involved in the cyclization of the sulfenic
acid 10 to the cyclic sulfenyl amide 11.
absent, still requires a large activation energy (49.29 kcal‚mol-1
)
to produce the sulfenyl amide. Therefore, the formation of the
transition state 4TSIII from the intermediate 4b is the most
important step in the cyclization process. As a large amount of
energy is required to break the S-O bond and to form the planar
five-membered isothiazolidinone ring in the transition state
4TSIII, the introduction of any substituent that can assist the
S-O bond cleavage may lower the transition state and enhance
the cyclization process.
Scheme 2. Proposed Mechanism for the Formation of Sulfenyl
Amide 11 from Thiol 9
Our experimental and theoretical studies on the cyclization
of 3 and 4 and the analysis of PTP1B active site features suggest
that the oxidation of thiols to the corresponding disulfides and
the electronic environment around the sulfenic acid moiety
(-SOH) need to be taken into account while designing
biomimetic models for the unusual protein modification at the
active site of PTP1B. In this regard, we synthesized thiol 9, in
which the 6-H in compound 2 has been replaced by an oxazoline
moiety. The introduction of this five-membered heterocycle may
prevent the S‚‚‚O interactions between the sulfur atom and the
oxygen atom of the amide backbone in the sulfenic acid
intermediate. The introduction of steric hindrance in the thiol
is expected to prevent the oxidation of the thiol to the
corresponding disulfide. Crucially, the possible S‚‚‚N interac-
tions between the sulfur atom and the five-membered hetero-
cyclic nitrogen atom in the sulfenic acid may force the -OH
moiety to approach the -NH- group.16 Subsequently, this
arrangement may facilitate the cyclization reaction, leading to
the formation of the expected sulfenyl amide. With these aspects
in mind, we treated the thiol 9 with hydrogen peroxide in
acetonitrile and followed the reaction by HPLC. Interestingly,
this reaction resulted in a quantitative conversion of the thiol 9
to the sulfenyl amide 11 (Scheme 2), indicating that the
additional substituent at the 6-position plays an important role
in the cyclization reaction. This is in contrast to thiols 1 and 2,
which produce the corresponding sulfenyl amides in small
amounts. As expected, the formation of disulfide 12 was not
detected during the entire process, which is in agreement with
the PTP1B protein modification that does not produce any
disulfide species during the formation of the sulfenyl amide.
The DFT calculations show that the sulfur atom in sulfenic
acid 10 is not involved in an interaction with the carbonyl
oxygen. This is in sharp contrast to compounds 3 and 4, which
all exhibited strong S‚‚‚O interactions. The absence of such
interactions is, however, not very important in this case as the
energy required to break these interactions is expected to be
much lower than that of the overall reaction. The most important
feature in the sulfenic acid 10 is the interaction of the hetercyclic
nitrogen present in the five-membered oxazoline ring with the
sulfur atom, which participates in the cyclization process and
dramatically lowers the activation energy barrier (Figure 6). This
type of neighboring group participation, which is not possible
with monosubstituted sulfenic acids such 3 and 4, is crucial for
the formation of the sulfenyl amide and the facile elimination
of a water molecule. The plot of energy vs reaction coordinates
indicates that the conversion of 10 to 11 is energetically more
(16) In contrast to the S‚‚‚O interactions that are well described in the literature,
compounds having nonbonded S‚‚‚N interactions are extremely rare. For
some well-defined examples of nonbonded interactions between divalent
sulfur and nitrogen, see: (a) Chivers, T.; McGarvey, B.; Parvez, M.; Vargas-
Baca, I.; Ziegler, T. Inorg. Chem. 1996, 35, 3839-3847. (b) Chivers, T.;
Krouse, I.; Parvez, M.; Vargas-Baca, I.; Ziegler, T.; Zoricak, P. Inorg. Chem.
1996, 35, 5836-5842. (c) Mugesh, G.; Singh, H. B.; Butcher, R. J. Eur. J.
Inorg. Chem. 1999, 1229-1236. (d) Iwaoka, M.; Takemoto, S.; Okada,
M.; Tomoda, S. Bull. Chem. Soc. Jpn. 2002, 75, 1611-1625.
9
8876 J. AM. CHEM. SOC. VOL. 129, NO. 28, 2007

Redox Regulation of Protein Tyrosine Phosphatase 1B
A R T I C L E S
Table 3. Relative Electronic ∆(E + ZPE) and Relative Gibbs Free
Energy ∆(G + ZPE) for the Cyclization Reaction of 10 Calculated
at the B3LYP/6-31G(d) Level of Theory
10
f
11
+
H2O
10
TS10
10a
11 + H2O
∆(E + ZPE)
∆(G + ZPE)
0
0
28.63
30.12
-23.40
-23.02
-15.69
-24.20
occupy the position of the 6-substituent, a strain is imposed in
the molecule, which leads to the facile cyclization. (ii) The S‚
‚‚N interactions between the sulfur atom and the oxazoline
nitrogen atom in the sulfenic acid brings the -OH moiety close
to the -NH- group, which may facilitate the cyclization
process. In agreement with these assumptions, the DFT calcula-
tions on the sulfenic acid 10 clearly indicate that the sulfur atom
does not interact with the carbonyl oxygen, but it interacts
strongly with the nitrogen atoms present in the five-membered
oxazoline ring. The resulting strained molecular conformation
of 10 brings the cyclization to a single step process, which is
in contrast to the cyclization of 4 that needs to cross three
different energy barriers for the conversion. The relative
electronic energies (including the ZPVE correction) calculated
at the B3LYP/6-31G(d) level of theory for the cyclization of
10 to 11 are summarized in Table 3, and the corresponding plot
is shown in Figure 7. These data indicate that the total energy
required for the conversion of 10 to 11 is almost half when
compared with the energy required for the cyclization of 3 or 4.
Figure 7. Energy profiles (∆Eelec + ZPVE) for the cyclization of 10
to 11.
Scheme 3. Reduction of Sulfenyl Amide Bond in 11 by GSH.
favored than the conversion of 4 to the sulfenyl amide 6 (Figure
7). These calculations also indicate that the cyclic sulfenyl amide
11 is 15.69 kcal‚mol-1 more stable than the sulfenic acid 10.
The higher stability of the product may also arise from the
stabilizing interactions between the heterocyclic nitrogen and
the sulfur atoms in compound 11. The presence of such
interactions was unambiguously confirmed by single-crystal
X-ray studies on compound 11, which showed a strong S‚‚‚N
interaction (S‚‚‚N distance: 2.760 Å). The S-N covalent bond
length of 1.73 Å in 11 was found to be comparable with that of
the sulfenyl amide of PTP1B (1.70 Å).8b
In the sulfenic acid 10, the -OH group is very close to the
-NH- group and the hydrogen atom present on the -NH- is
hydrogen bonded to the -OH group (HO‚‚‚HN: 1.79 Å) (Figure
6). The complete elimination of a water molecule was observed
for compound 10, which leads to the generation of 10a via
transition state TS10. As a result, the water molecule is bound
to the sulfenyl amide Via two strong hydrogen bonds. The first
one is observed between the amide carbonyl oxygen atom and
the hydrogen atom that was previously in the -SOH moiety
(1.85 Å). The second hydrogen bond is observed between the
oxygen atom of the eliminated water molecule and the hydroxyl
hydrogen atom of the -CH2OH moiety in the side chain (1.83
Å). The strong S‚‚‚N interaction between the oxazoline nitrogen
and the sulfur atom is retained in this structure as evidenced by
a short S‚‚‚N distance (2.76 Å). The conversion of 10 to 10a
proceeds through the transition state TS10 for which the energy
was calculated to be 28.63 kcal/mol higher than that of 10. In
the transition state TS10, the water molecule is almost formed
as observed from the short distance (1.04 Å) between the NH
hydrogen atom and the oxygen atom of the OH group, which
is very close to the O-H covalent bond length. However, the
hydrogen atom remains hydrogen bonded to the nitrogen (1.54
Å). The oxygen atom is placed at a distance of 2.46 Å from the
sulfur atom, which is much longer than the S-O covalent bond
in 10 (1.72 Å). In the transition state TS10, the amide nitrogen
atom is positioned away from that of the sulfur atom (3.15 Å)
and the nitrogen atom is placed at about -19.8° out of the plane
of the phenyl ring attached to the sulfur atom. The most
interesting feature in TS10 is that the nitrogen atom of the
oxazoline ring is located almost within covalent bonding
distance with the sulfur atom (S‚‚‚N ) 1.79 Å).
Despite its higher stability, compound 11 undergoes facile
reaction with thiols such as benzene thiol (PhSH), glutathione
(GSH), or dithiothreitol (DTT), to produce the sterically hindered
thiol 9. The dithiol DTT was found to be a better reducing agent
than the monothiols. These reactions are expected to proceed
through the formation of a mixed disulfide (e.g., 13, Scheme
3) as previously described for 5.9 However, the sulfenyl amide
11 was found to exist in equilibrium with the corresponding
thiol (9). A large amount of reducing thiol was, therefore,
required for the quantitative conversion of 11 to 9. This is in
agreement with the report of Barford et al. that the sulfenic acid
derivative of PTP1B undergoes cyclization even in the presence
of thiols to produce the corresponding sulfenyl amide.8a Our
observations, therefore, support the assumption that the enzyme
PTP1B would produce sulfenyl amide species in the reducing
environment of the cell.
The detailed DFT investigations to understand the reason for
the rapid cyclization of sulfenic acid 10 to the sulfenyl amide
11 led to some interesting observations. We find that the model
system based on 9 offers two distinct advantages over the model
based on thiol 2 in the cyclization process. (i) The five-
membered oxazoline moiety prevents the S‚‚‚O interactions
between the sulfur atom and the oxygen atom of the amide
backbone in the sulfenic acid intermediate through steric
hindrance. As the -OH moiety would have to approximately
Based on the experimental observations and theoretical
calculations, we suggest an SN2 type elimination of the -OH
9
J. AM. CHEM. SOC. VOL. 129, NO. 28, 2007 8877

A R T I C L E S
Sarma and Mugesh
Figure 8. Participation of the oxazoline nitrogen atom in the cyclization
of the sulfenic acid 10 to the cyclic sulfenyl amide 11.
Scheme 4. Effect of Isosteric Replacement of O and N Atoms of
Oxazoline Ring in the Cyclization.
Figure 9. Energy profiles (∆Eelec + ZPVE) for the cyclization of 14
to 15.
group (Figure 8). According to this model, the nucleophilic
oxazoline nitrogen atom attacks at the electrophilic sulfur center
from the backside. As a result, the -OH group is eliminated as
a water molecule by abstracting the required proton from the
-NH- group. Consequently, the deprotonation leads to the
formation of a highly nucleophilic nitrogen on the amide moiety.
The simultaneous attack of this negatively charged nitrogen
moiety at the sulfur center and the cleavage of the S-N bond
between the sulfur and the oxazoline nitrogen lead to the
formation of sulfenyl amide 10a. The entire process can be
viewed as a neighboring group participation of the oxazoline
nitrogen followed by two successive elimination reactions
(Figure 8). The participation of the neighboring group brings
down the barrier for cyclization. The energy barrier for the
cyclization of 10 to 10a was calculated to be 28.63 kcal‚mol-1
,
Figure 10. HOMO and LUMO of TS10 calculated at the B3LYP/6-31G-
(d) level of theory.
which is much lower than that for the conversion of 4 to 6 (49.29
kcal‚mol-1). In the final step, the departure of the water molecule
from 10a to produce the final compound 11 is found to be an
that the introduction of sterically bulky groups that can enforce
the proximity of sNHs and sOHs groups is not sufficient
for a rapid cyclization of sulfenic acids to sulfenyl amides.
To gain better insight into the role of the oxazoline nitrogen
atom during the cyclization reaction, we have obtained the
HOMO and LUMO of TS10 by using the B3LYP/6-31G(d)
level of theory (Figure 10). The HOMO represents an orbital
delocalized over the whole amide moiety, whereas the LUMO
is generated by factors from both the phenyl and the oxazoline
rings. Furthermore, the LUMO shows contribution from a π*
(S-N) orbital. The electrons are expected to flow from the
HOMO to the π* (S-N) orbital, which would lead to the
cleavage of the S-N (oxazoline ring) bond and the formation
of a new S-N (amide) bond leading to the formation 11.
Therefore, the conversion of the sulfenic acid 10 to the sulfenyl
amide 11 is both kinetically as well as thermodynamically more
favored than the conversion of the sulfenic acid 4 to the
endothermic process by 7.71 kcal‚mol-1
.
The importance of the nitrogen in the five-memberd oxazoline
ring in 10 during an efficient cyclization process was further
verified by isosteric replacements of the nitrogen and oxygen
atoms in 10 with sCHs and sCH2s groups, respectively
(compound 14). These replacements are expected to retain the
steric hindrance at the 6-position but alter the electronic
properties of the sulfur. As expected, our theoretical calculations
show that the introduction of the five-membered ring having
no heteroatoms does maintain the steric bulkiness of the
6-substituent in the sulfenic acid and this steric hindrance is
sufficient to prevent any nonbonded S‚‚‚O interactions between
the sulfur and carbonyl oxygen in compound 14 (Scheme 4).
However, the activation energy barrier calculated for the
conversion of the sulfenic acid 14 to the sulfenyl amide 15
(45.77 kcal‚mol-1, Figure 9) was found to be much higher than
that of the conversion of 9 to 11 (28.63 kcal/mol) but
comparable with that of the conversion of 4 to 6 (49.29 kcal/
mol). These observations clearly suggest that the oxazoline
nitrogen participation, which assists the SsO bond cleavage to
lower the level of the TS, is more important than the S‚‚‚Od
C< nonbonded interactions.17 These observations also suggest
(17) Although simple sulfenic acids can act as either electrophiles or nucleo-
philes, the sulfenic acids 4, 10, and 14 in the present study act as
electrophiles towards amide functionality. The NBO charges on the sulfur
atoms of these compounds suggest that the sulfur atom in sulfenic acid 10
is more electrophilic than that of 4 and 14. This is due to the involvement
of oxazoline nitrogen in compound 10. This is also responsible for the
weakening of the S-O bond in the TS10.
9
8878 J. AM. CHEM. SOC. VOL. 129, NO. 28, 2007

Redox Regulation of Protein Tyrosine Phosphatase 1B
A R T I C L E S
MHz) NMR spectra were obtained on a Bruker Avance 400 NMR
corresponding sulfenyl amide (6). Owing to the presence of a
strong S‚‚‚N interaction in 11 and the steric crowding in 10,
compound 11 is 15.69 kcal/mol more stable than the sulfenic
acid 10. On the other hand, the presence of S‚‚‚O interaction in
4 and the absence of such interactions in 6 indicate the lower
thermodynamic stability of the product sulfenyl amide (6)
compared with the starting sulfenic acid 4. The HOMO-LUMO
calculations confirm our previous observations that the chelating
oxazoline ring leads to the stabilization of the transition state
(TS10) and, therefore, the barrier for cyclization is only half of
that calculated for the sulfenic acid 4.
The oxidation of thiol moieties in 1, 2, and 9 by hydrogen
peroxide and the facile reduction of sulfenyl amides 5, 6, and
11 by thiols prompted us to investigate the antioxidant activity
of the sulfenyl amides. The reduction of H2O2 was studied by
using PhSH as a thiol cosubstrate. The conversion of PhSH to
PhSSPh was followed by reversed-phase HPLC, using methanol-
water as the eluent. Although the thiols 1, 2, and 9 react readily
with H2O2 to produce the corresponding sulfenic acids, the
sulfenyl amides 5, 6, and 11 exhibited poor antioxidant activity
in the presence of PhSH. A detailed study on the antioxidant
behavior and catalytic chemistry of the sulfenyl amides is under
investigation.
1
spectrometer using the solvent as an internal standard for H and 13C.
Chemical shifts (1H, 13C) are cited with respect to tetramethylsilane
(TMS). Thin-layer chromatography analyses were carried out on
precoated silica gel plates (Merck), and spots were visualized by UV
irradiation. Column chromatography was performed on glass columns
loaded with silica gel or on an automated flash chromatography system
(Biotage) by using preloaded silica cartridges. High performance liquid
chromatography (HPLC) experiments were carried out on a Waters
Alliance System (Milford, MA) consisting of a 2690 separation module,
a 2690 photodiode-array detector, and a fraction collector. The assays
were performed in 1.8 mL sample vials, and a built-in autosampler
was used for sample injection. The Alliance HPLC System was
controlled with EMPOWER software (Waters Corporation, Milford,
MA). Anthranilic acid was first diazotized and was added to disodium
disulfide to produce dithiosalicylic acid, which was treated with freshly
distilled thionyl chloride to produce the 2-(chlorothio)benzoyl chloride
as a yellow solid. The cyclic sulfenyl amide 6 was prepared by treating
2-(chlorothio)benzoyl chloride with 2-amino-2-methylpropan-1-ol in
dry acetonitrile.
Synthesis of 7: To a stirred solution of benzanilide (1.0 g, 5.1 mmol)
in dry THF (35 mL) under nitrogen at 0 °C was added n-butyllithium
(6.4 mL, 10.2 mmol, 1.6 M solution in hexane). After 30 min elemental
sulfur (0.16 g, 5.1 mmol) was added to the orange-red solution while
a brisk stream of nitrogen was passed through the open system. The
solid material was almost consumed after 1 h, and the resulting
yellowish-green solution was added to a beaker containing K3Fe(CN)6,
(1.70 g, 5.2 mmol) in water (150 mL). The precipitated material was
filtered off to give, after washing with methylene chloride and drying,
Conclusions
In the present study, biomimetic models were developed for
the unusual formation of a sulfenyl amide species from the
sulfenic acid intermediate at the active site of PTP1B. This study
reveals that the aromatic sulfenic acids having a monosubstituted
amide backbone need to cross several energy barriers for the
possible conversion to the corresponding sulfenyl amides. The
introduction of an additional substituent at the 6-position in a
benzamide-derived sulfenic acid enhances the cyclization
process. The introduction of a sterically bulky substituent that
can force the -OH group to approach the backbone -NH-
moiety and a heteroatom that can interact with the sulfur in the
sulfenic acid intermediate to enhance the electrophilicity of
sulfur lead to the formation of a sulfenyl amide even in the
presence of thiols, supporting the assumption that the enzyme
PTP1B would produce the sulfenyl amide species in the reducing
environment of the cell. Although the thiol groups in the
synthetic models are readily oxidized by hydrogen peroxide and
reduced back by thiols, the sulfenyl amides possess a weak
antioxidant activity. This is in accordance with the fact that the
major function of phosphatases such as PTP1B is not associated
with the antioxidant activity.
1
1.40 g (60%) of the disulfide. H NMR (DMSO-d6) δ 10.56 (s, 1H),
7.73-7.78 (m, 4H), 7.52 (t, J ) 7.2 Hz, 1H), 7.36-7.41 (m, 3H), 7.13
(t, J ) 7.6 Hz, 1H); 13C NMR (DMSO-d6) δ 165.7, 138.9, 136.5, 134.7,
131.5, 128.8, 128.5, 126.4, 126.3, 124.0, 120.1. MS (TOF MS ES+)
m/z 479.0856 [M + Na]+.
Synthesis of 1: To a stirred solution of 7 (40.7 mg, 0.088 mmol) in
dry methanol (10 mL) was added sodium borohydride (33 mg, 0.88
mmol) in 5 mL of dry methanol. The reaction mixture was stirred for
24 h at 25 °C, and the solvent was evaporated to dryness and
dichloromethane was added to the reaction mixture. The dichlo-
romethane solution was then added to a dilute HCl solution, stirred for
30 min, and then extracted with dichloromethane. The combined
dichloromethane extracts were then evaporated under reduced pressure
to give colorless oil, which was then purified by flash chromatography
using petroleum ether/ethyl acetate as eluent to give the expected thiol
as a white solid in 90% yield. 1H NMR (CDCl3) δ 7.82 (s, 1H), 7.57-
7.62 (m, 3H), 7.35-7.39 (m, 3H), 7.31 (t, J ) 7.6 Hz, 1H), 7.15-7.22
(m, 2H), 4.55 (s, 1H); 13C NMR (CDCl3) δ 165.9, 136.8, 132.7, 132.3,
130.6, 130.3, 128.4, 127.3, 124.7, 124.1, 119.5. MS (TOF MS ES+)
m/z 252.0462 [M + Na]+. Anal. Calcd for C13H11NOS : C, 68.09; H,
4.84; N, 6.11. Found: C, 68.24; H, 4.90; N, 6.15.
Synthesis of 5: This compound was prepared by following the
literature method with minor modifications.18 To a stirred solution of
7 (90 mg, 0.2 mmol) in dichloromethane (5 mL) was added bromine
(25 µL, excess mmol), and the reaction mixture was stirred for 12 h at
25 °C. Activated basic alumina (670 mg, excess mmol) was then added,
and stirring was continued for an additional 3 h. The solvent was then
evaporated under reduced pressure, and the resulting material was
purified by column chromatography using neutral aluminum oxide to
Experimental Section
General Procedure: n-Butyllithium was purchased from Acros
Chemical Co. (Belgium). Sulfur powder, sodium borohydride (NaBH4),
and benzanilide were obtained from local suppliers. Methanol was
obtained from Merck and dried before use. All other chemicals were
of the highest purity available. All experiments were carried out under
dry and oxygen-free nitrogen using standard Schlenk techniques for
the synthesis. Due to unpleasant odors of several of the reaction mixtures
involved, most manipulations were carried out in a well-ventilated fume
hood. Mass spectral (MS) studies were carried out on a Q-TOF Micro
mass spectrometer with electrospray ionization MS mode analysis. In
the case of isotopic patterns, the value given is for the most intense
peak. Elemental analyses were performed on a ThermoFinigan FLASH
EA 1112 CHNS analyzer. Liquid-state NMR spectra were recorded in
CDCl3, d4-MeOH, or d6-DMSO as solvent. 1H (400 MHz) and 13C (100
1
give a white solid in 80% yield. The H and 13C NMR spectral data
are in accordance with the literature values.18 1H NMR (CDCl3) δ 8.11
(d, J ) 8.0 Hz, 1H), 7.58-7.72 (m, 4H), 7.43-7.50 (m, 3H), 7.33 (t,
J ) 7.6 Hz, 1H); 13C NMR (CDCl3) δ 164.1, 139.8, 137.2, 132.3, 129.3,
127.2, 127.0, 125.8, 124.8, 124.6, 120.1; MS (TOF MS ES+) m/z
250.0389 [M + Na]+.
(18) Kamigata, N.; Hashimoto, S.; Kobayashi, M. Org. Prep. Proc. Int. 1983,
15, 315-319.
9
J. AM. CHEM. SOC. VOL. 129, NO. 28, 2007 8879

A R T I C L E S
Sarma and Mugesh
1
Synthesis of 6: To a stirred solution of 2-amino-2-methylpropan-
1-ol (480 µL, 5 mmol) in dry acetonitrile (50 mL) was added dropwise
a solution of 2-(chlorothio)benzoyl chloride (1 g, 4.83 mmol) in
acetonitrile (15 mL). The reaction mixture was stirred for 5 h at 25
°C, and the resulting precipitate was filtered off. The filtrate was
evaporated under reduced pressure to give a colorless oil, which was
then purified by column chromatography using petroleum ether/ethyl
acetate (5:1) as eluent to give compound 6 as a white solid in 80%
product in 85% yield. H NMR (CDCl3) δ: 16.10 (b, 1H), 9.26 (s,
1H), 8.13 (d, J ) 9.6 Hz, 1H), 7.94 (d, J ) 9.6 Hz, 1H), 6.98 (t, J )
8 Hz, 1H), 5.44 (b, 1H), 4.52 (s, 2H), 3.74 (s, 2H), 1.67 (s, 6H), 1.44
(s, 6H); 13C NMR (CDCl3) δ 170.2, 167.5, 161.4, 137.9, 134.9, 132.2,
118.8, 117.8, 79.4, 69.5, 62.2, 59.4, 55.8, 26.6, 24.6, 23.7, 13.2.
HPLC Assay: In this assay, we employed a mixture containing a
2:1 molar ratio of PhSH and H2O2 in dichloromethane/methanol (95:
5) at room temperature as our model system. Runs with and without
10 mol % added test compounds were carried out under the same
conditions. Periodically, aliquots were removed and the concentrations
of the product diphenyl disulfide (PhSSPh) were determined from the
detector response, using pure PhSSPh as an external standard. The initial
rates (V0) for the conversion of thiols to the corresponding sulfenyl
amides were calculated from the first 5-10% of the reactions, and the
amounts of sulfenyl amides formed in these reactions were determined
from the calibration plots of authentic sulfenyl amides. The reduction
of H2O2 by sulfenyl amides in the presence of PhSH was studied by
following a similar method using the PhSSPh as an external standard.
The amount of disulfide formed during the course of the reaction was
calculated from the calibration plot for the standard.
1
yield. H NMR (CDCl3) δ 7.96 (d, J ) 8.0 Hz, 1H), 7.61 (t, J ) 7.6
Hz, 1H), 7.51 (d, J ) 8.0 Hz, 1H), 7.40 (t, J ) 7.6 Hz, 1H), 3.95 (s,
2H), 1.61 (s, 6H). 13C NMR (CDCl3) δ 166.4, 140.1, 131.8, 126.3,
126.0, 125.5, 119.7, 69.6, 63.8, 24.9; MS (TOF MS ES+) m/z 246.0529
[M + Na]+.
Synthesis of 8: To a stirred solution of 6 (100 mg, 0.45 mmol) in
CH2Cl2 (10 mL) was added excess benzenethiol (460 µL, 4.5 mmol).
The reaction mixture was stirred for 48 h at ambient temperature, and
the reaction mixture was kept for aerial oxidation for another 48 h.
The solvent was then removed in Vacuo, and the disulfide was then
purified by column chromatography by using ethyl acetate/pet ether
1
1:1 as eluent to give compound 8 as a white solid in 60% yield. H
X-ray Crystallography. X-ray crystallographic studies were carried
out on a Bruker CCD diffractometer with graphite-monochromatized
Mo KR radiation (λ ) 0.710 73 Å) controlled by a Pentium-based PC
running on the SMART software package.19 Single crystals were
mounted at room temperature on the ends of glass fibers, and data were
collected at room temperature. The structures were solved by direct
methods and refined using the SHELXTL software package.20 All non-
hydrogen atoms were refined anisotropically, and hydrogen atoms were
assigned idealized locations. Empirical absorption corrections were
applied to all structures using SADABS.21,22
NMR (MeOH-d4) δ 7.73 (d, J ) 8.0 Hz, 1H), 7.46 (d, J ) 7.6 Hz,
1H), 7.36 (t, J ) 7.2 Hz, 1H), 7.24 (t, J ) 7.6 Hz, 1H), 3.69 (s, 2H),
1.40 (s, 6H); 13C NMR (MeOH-d4) δ 169.2, 136.3, 135.6, 132.3, 131.1,
130.4, 129.7, 127.5, 127.2, 126.7, 126.3, 67.8, 55.8, 22.7; MS (TOF
MS ES+) m/z 471.1411 [M + Na]+.
Synthesis of 2: To a stirred solution of the disulfide 8 (20.5 mg,
0.045 mmol) in dry methanol (10 mL) was added sodium borohydride
(17 mg, 0.45 mmol) in 5 mL of dry methanol. The reaction mixture
was stirred for 24 h at 25 °C, and the solvent was evaporated to dryness
and dichloromethane was added to the reaction mixture. The dichlo-
romethane solution was then added to a dilute HCl solution, stirred for
30 min, and then extracted with dichloromethane. The combined
dichloromethane extracts were then evaporated under reduced pressure
to give colorless oil, which was then purified by flash chromatography
using petroleum ether/ethyl acetate as eluent to give compound 2 as a
Crystal data for 6: C11H13NO2S; Mr ) 223.28, monoclinic, space
group P2(1)/c, a ) 10.0765(56) Å, b ) 15.6555(69) Å, c ) 7.3410-
(33) Å, â ) 111.081(7)°; V ) 1080.56(33) Å3, Z ) 4, Fcalcd ) 1.37
g/cm, Mo KR radiation (λ ) 0.710 73 Å), T ) 273(2) K; R1 ) 0.059,
wR2 ) 0.181 (I > 2σ(I)); R1 ) 0.062, wR2 ) 0.187 (all data). Crystal
data for 11: C16H20N2O3S; Mr ) 320.4, triclinic, space group P1h, a )
9.4859(14) Å, b ) 9.7695(15) Å, c ) 10.9072(17) Å, R ) 90.233(2)°;
1
white solid in 90% yield. H NMR (CDCl3) δ 7.42 (d, J ) 6.8 Hz,
1H), 7.32 (d, J ) 7.6 Hz, 1H), 7.26 (t, J ) 6.8 Hz, 1H), 7.16 (t, J )
â ) 112.003(2)°; γ ) 117.918(2)°, V ) 807.99(16) Å3, Z ) 2, Fcalcd
)
6.8 Hz 1H), 6.07 (b, 1H), 4.40 (s, 1H), 3.72 (s, 2H), 1.41 (s, 6H); 13
C
1.32 g/cm, Mo KR radiation (λ ) 0.710 73 Å), T ) 293(2) K; R1 )
0.038, wR2 ) 0.113 (I > 2σ(I)); R1 ) 0.047, wR2 ) 0.120 (all data).
The structures were solved by a direct method (SIR-92)20 and refined
by full-matrix least-squares procedures on F2 for all reflections
(SHELXL-97).21-22
NMR (CDCl3) δ 168.4, 133.2, 130.8, 130.1, 129.7, 126.9, 126.6, 124.5,
69.2, 55.8, 23.6; MS (TOF MS ES+) m/z 248.0717 [M + Na]+.
Synthesis of 11: To a stirred solution of 2,6-bis(4,4-dimethyl-2-
oxazoline-2-yl)benzene (0.817 g, 3 mmol) in 30 mL of dry benzene
were added TMEDA (1.35 mL, 9 mmol) and LDA (4.5 mL, 9.0 mmol).
The reaction mixture was stirred for 4 h, and the resulting precipitate
was dissolved in dry THF (30 mL). After the solution cooled to -15
°C, elemental sulfur (0.100 g, 3 mmol) was added. The stirring was
continued for 12 h at ambient temperature, and then the mixture was
kept for 5 days for aerial oxidation by adding THF from time to time.
The solution was then extracted with CH2Cl2. The organic phase was
dried over anhydrous sodium sulfate, and the solvent was removed in
Vacuo to afford dark oil. The unusually cleaved product 11 was purified
by column chromatography as a stable compound on silica gel with
petroleum ether/ethyl acetate (1:1) as eluent in 50% yield. The product
was recrystallized from chloroform to give colorless crystals. 1H NMR
(CDCl3) δ 8.09 (d, J ) 7.6 Hz, 1H), 7.97 (d, J ) 6.8 Hz, 1H), 7.48 (t,
J ) 7.6 Hz, 1H), 5.58 (b, 1H), 4.22 (s, 2H), 3.94 (d, J ) 3.6 Hz, 2H),
1.68 (s, 6H), 1.44 (s, 6H); 13C NMR (CDCl3) δ 166.2, 160.4, 141.0,
130.1, 128.6, 127.5, 125.2, 119.7, 80.2, 70.5, 68.1, 63.3, 28.6, 25.0;
MS (TOF MS ES+) m/z 321.1295 [M + H]+. Anal. Calcd for
C16H20N2O3S: C, 59.98; H, 6.29; N, 8.74; S, 10.01. Found: C, 60.31;
H, 6.38; N, 8.87; S, 10.13.
Computational Methods. All calculations were performed using
the Gaussian98 suite of quantum chemical programs.23 The hybrid
Becke 3-Lee-Yang-Parr (B3LYP) exchange correlation functional
was applied for DFT calculations.24 Geometries were fully optimized
at the B3LYP level of theory using the 6-31G(d) basis sets. Transition
states were located using Schlegel’s synchronous transit-guided quasi-
Newton (STQN) method.25,26 Transition states were searched by using
the QST3 keyword, and the resultant conformation was optimized using
the TS keyword. Furthermore, the transition state and the stable
conformers were characterized by the presence or absence of a single
imaginary mode. The activation energies are the difference in the zero-
point vibrational energy corrected electronic energy between the
transition state and the stable conformations.
(19) SMART, version 5.05; Bruker AXS: Madison, WI, 1998.
(20) Altomare, A.; Cascarano, G.; Giacovazzo, C.; Gualardi, A. J. Appl.
Crystallogr. 1993, 26, 343-350.
(21) Sheldrick, G. M. SHELX-97. Acta Crystallogr., Sect. A 1990, 46, 467-
473.
(22) Sheldrick, G. M. SHELX-97, Program for the Refinement of Crystal
Structures; University of Go¨ttingen: Go¨ttingen, Germany, 1997.
(23) Gaussian98; Gaussian, Inc.: Pittsburgh, PA, 1998. The full reference is
given in the Supporting Information.
Synthesis of 9: To a stirred solution of 11 (20 mg, 0.062 mmol) in
dry THF (5 mL) was added dithiothreitol (DTT) (19 mg, 0.124 mmol),
and the reaction mixture was allowed to stir at room temperature under
nitrogen for 48 h. The solvent was removed under Vacuo, and the
expected thiol was purified by flash chromatography to give the final
(24) (a) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785-789. (b)
Becke, A. D. J. Chem. Phys. 1993, 98, 5648-5652.
(25) Gonzalez, C.; Schlegel, H. B. J. Chem. Phys. 1989, 90, 2154-2161.
(26) Gonzalez, C.; Schlegel, H. B. J. Phys. Chem. 1990, 94, 5523-5527.
9
8880 J. AM. CHEM. SOC. VOL. 129, NO. 28, 2007

Redox Regulation of Protein Tyrosine Phosphatase 1B
A R T I C L E S
Acknowledgment. This study was supported by the Depart-
ment of Science and Technology (DST), New Delhi, India. We
also thank the DST for the CCD single-crystal X-ray diffraction
facility and Dr. M. Nethaji for his help in solving the X-ray
structures. We are grateful to the Alexander von Humboldt
Foundation, Bonn, Germany for the donation of an automated
flash chromatography system. G.M. acknowledges the DST for
the award of Ramanna fellowship, and B.K.S. thanks the Indian
Institute of Science for a research fellowship.
Supporting Information Available: Archive entries for all
the optimized geometries reported in this paper and full reference
of ref 23. This material is available free of charge via the Internet
at http://pubs.acs.org.
JA070410O
9
J. AM. CHEM. SOC. VOL. 129, NO. 28, 2007 8881