Design, synthesis and insulin-sensitising effects of novel PTP1B inhibitors

Article abstract of DOI:10.1016/j.bmcl.2013.02.073

Fifteen novel sulfathiazole-related compounds were designed as PTP1B inhibitors based on a previously reported allosteric inhibitor (1) of PTP1B. These compounds were synthesized and evaluated against human recombinant PTP1B. Six compounds (3, 4, 8 and 14

Full text of DOI:10.1016/j.bmcl.2013.02.073

Bioorganic & Medicinal Chemistry Letters 23 (2013) 2313–2318
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design, synthesis and insulin-sensitising effects of novel PTP1B
inhibitors
Yan-Bo Tang a, Dianyun Lu b, Zheng Chen a, Chun Hu b, Ying Yang a, Jin-Ying Tian c, Fei Ye c, Li Wu d,
Zhong-Yin Zhang d, Zhiyan Xiao a,
⇑
a Beijing Key Laboratory of Active Substance Discovery and Druggability Evaluation, Institute of Materia Medica, Chinese Academy of Medical Sciences and Peking Union
Medical College, Beijing 100050, PR China
b School of Pharmaceutical Engineering, Shenyang Pharmaceutical University, Shenyang 110016, PR China
c Beijing Key Laboratory of New Drug Mechanisms and Pharmacological Evaluation Study, Institute of Materia Medica, Chinese Academy of Medical Sciences & Peking Union
Medical College, Beijing 100050, PR China
d Department of Biochemistry and Molecular Biology and Chemical Genomics Core Facility, Indiana University, School of Medicine, 635 Barnhill Drive, Indianapolis, IN 46202, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Fifteen novel sulfathiazole-related compounds were designed as PTP1B inhibitors based on a previously
reported allosteric inhibitor (1) of PTP1B. These compounds were synthesized and evaluated against
human recombinant PTP1B. Six compounds (3, 4, 8 and 14–16) exhibited significant inhibitory activity
against PTP1B. The most active compound (16) showed IC50 value of 3.2 lM and kinetic analysis indicated
that it is a non-competitive inhibitor of PTP1B. Furthermore, compound 16 demonstrated excellent selec-
Received 27 December 2012
Revised 11 February 2013
Accepted 14 February 2013
Available online 26 February 2013
tivity to PTP1B over other PTPs. It also displayed in vivo insulin sensitizing effect in the insulin resistant
mice.
Keywords:
PTP1B inhibitors
Selectivity
Ó 2013 Elsevier Ltd. All rights reserved.
In vivo insulin-sensitising effects
Protein tyrosine phosphatases (PTPs) play an important regula-
tory role in the intracellular phosphorylation state of proteins.1
Among the PTPs, protein tyrosine phosphatase 1B (PTP1B) acts as
a key negative regulator in both insulin and leptin signaling path-
ways, thereby modulates both glucose and lipid metabolism. The
dual function of PTP1B makes it an attractive target for anti-dia-
betic drugs.2,3
catalytic loop of PTP1B, so as to stabilize the inactive conformation
and prevent the formation of active conformation. The allosteric
site is also characterized by its non-conservative and electronically
neutral constitution, which implies that targeting the allosteric site
might be a feasible strategy to achieve selectivity and in vivo
activity.
Benzbromarone derivative 1, which was identified as a selective
PTP1B inhibitor acting at the allosteric site, enhances insulin recep-
tor phosphorylation in a cell-based assay and displays good selec-
tivity over several other phosphatases.7 These observations
support the approach to overcome the selectivity and cellular per-
meability issues by designing and developing allosteric PTP1B
inhibitors. The discovery of novel PTP1B inhibitors via pharmaco-
phore-oriented scaffold hopping from the template structure 1 is
reported herein.
The crystal structure of 1-PTP1B complex reveals the binding
mode of 1 in the allosteric site.7 The benzofuran core is accom-
modated in the hydrophobic pocket formed by the side chains of
the residues Leu192, Phe196 and Phe280. Two hydrogen bonds
are observed between the ketone oxygen in 1 and the side chain
of Asn193 as well as the sulfonamide nitrogen and the carboxyl
of Glu276. An additional water-mediated hydrogen bond is de-
tected between the phenol group and the main chain carbonyl
of Phe196.
Extensive medicinal chemistry efforts, especially structure-
based drug design (SBDD) approaches have successfully identified
a variety of potent PTP1B inhibitors (Fig. 1). However, the endeavor
to develop PTP1B inhibitors into therapeutic drugs is largely
unsuccessful. The attrition is due to the highly conservative and
cationic nature of the PTP1B catalytic site. Most known inhibitors
incorporate pTyr mimetics interacting with the catalytic site of
PTP1B, thus are inherent with poor PTP1B selectivity and inade-
quate in vivo activity due to low cellular permeability.3–6 There-
fore, it is imperative to find novel PTP1B inhibitors acting at
alternative binding sites.
An allosteric site has been described for PTP1B in 2004.7 Small-
molecule inhibitors that occupy this site block the mobility of the
⇑
Corresponding author. Tel.: +86 10 63189228; fax: +86 10 63017757.
E-mail address: xiaoz@imm.ac.cn (Z. Xiao).
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.02.073

2314
Y.-B. Tang et al. / Bioorg. Med. Chem. Lett. 23 (2013) 2313–2318
HO
O
O
CF2PO3H2
OH
O
COOH
N
O
O
CF2PO3H2
O
O
H
N
HO
S
O
O
N
H
II
Br
III
I
Br
OH
Br
O
O
O
H
N
N
H
NH2
H
N
O
O
CH3
S
O
H
N
O
O
O
S
N
S
NH
O
O
S
IV
1
O
Figure 1. Structures of representative PTP1B inhibitors.
R1
Compound 1 also adopts a ‘U’ shape conformation to fit the allo-
steric site, which allows the thiazole and the benzofuran rings to
sandwich the residue Phe280 and presumably presents p–p stack-
ing effects. Accordingly, a hypothetic pharmacophore for PTP1B
A
R2
X
L
R3
N
B
O
H
N
O
allosteric inhibitors is proposed (Fig. 2).
C
N
S
Novel structures with the general formula illustrated in Figure 3
were designed by preserving the pharmacophric features while
altering the structural skeleton of the template molecule. Virtually,
the sulfathiazole part was kept intact, but the benzofuran core was
replaced by substituted phenyl to improve chemical accessibility
and ligand efficiency. The rest parts of the template molecule were
subjected to gradual structural alteration to investigate the impor-
tance of individual pharmacophoric features. First, an amide link-
age (L = –NH–) between rings A and B was maintained, and
structural elements affecting the ‘U’ shape conformation were
introduced at X, R2 and R3. The geometry of the sulfonamide moi-
ety (X = –SO2–, R2 = H, R3 = H) would allow an anti-parallel orienta-
tion of the thiazole and the phenyl (B) rings as observed in the
active conformation of 1. The amide linkage (X = –CO–, R2 = H,
R3 = H) may also achieve a ‘U’ shape conformation, though is prob-
ably more energy demanding. The inclusion of salicylanilide moi-
ety (X = –CO–, R2 = –OH, R3 = H) would adjust the overall
conformation of the molecules via an intramolecular hydrogen
bond between the amide oxygen and the para-hydroxyl on ring
O
S
X = -SO2-, -CO- or -CH2-
L = -NH-, -CH=CH-, or -CH2CH2
R1 : various substituents, R2 = H or OH, R3 = H or CH3
Figure 3. General formula of the target compounds.
B. While N-methylation of the amide bond (X = –CO–, R2 = H,
R3 = –CH3–) would also tune the general conformation of the mol-
ecules to adapt the steric hindrance posed by the methyl group.
The secondary amine (X = –CH2–, R2 = H, R3 = H) linkage with an
sp3 carbon, which would presumably be unfavorable for the ‘U’
shape conformation, was also introduced. After the optimal X, R2
and R3 were determined, the ‘L’ group was further elongated
(L = –CH@CH– or –CH2–CH2–) to explore the possibility of forming
a hydrogen bond directly between R1 and the enzyme instead of
the water-mediated interaction detected in the crystal structure
of 1-PTP1B complex. Various substituents were also introduced
at R1 to examine their effect on PTP1B inhibition.
Target molecules 2–12 were synthesized as shown in Schemes 1
and 2. Sulfonic acid 17 and carboxylic acids 19, 20 were prepared
from substituted benzoic acids, and were then successively reacted
with oxalyl dichloride in the presence of DMF and sulfathiazole or
4-(methylamino)-N-(thiazol-2-yl)benzene-sulfonamide to give the
target compounds 2–10 (Scheme 1). Reductive amination of 4-
nitrobenzaldehyde with sulfathiazole was followed by reduction
of the nitro group to yield 22, which reacted with 4-methoxybenz-
aldehyde or 4-methoxybenzoyl chloride to provide target com-
pounds 11 and 12, respectively (Scheme 2).
Aromatic group
"U" shape conformation
hydrogen
hydrogen
bond
bond
acceptor
or donor
acceptor
or donor
Compounds 13–16 were synthesized as shown in Scheme 3.
Compounds 23 were obtained via aldol condensation. Interest-
ingly, catalytic hydrogenation of 23a in the presence of 10% Pd/C
not only saturated the double bond as expected, but also reduced
the carbonyl group to hydroxyl. Such a byproduct was also ob-
Hydrophobic
hydrogen
bond
aromatic group
acceptor
Figure 2. Hypothetic pharmacophore for PTP1B inhibitors acting at the allosteric
site.

Y.-B. Tang et al. / Bioorg. Med. Chem. Lett. 23 (2013) 2313–2318
2315
R1
R1
H
N
H
N
H
e
N
O
O
S
H
N
O
O
N
HO3S
S
2 R1 = OCH3
17
O
O
S
R1
a
H
N
R1
R1
R3
N
H
N
R1
H
N
e
O
d
b
H
N
O
N
O
O
HO2C
S
C2H5O2C
HO2C
O
O
R1 = OCH3, R3 = H
3
S
18
19
4 R1 = CN, R3 = H
5 R1 = F, R3 = H
c
R1
6
7
H
N
R1 = OCH3, R3 = CH3
R1 = CN, R3 = CH3
R2
R1
H
e
H
N
R2
N
O
H
N
O
R1 = OCH3, R2 = OH
R1 = CO2CH3, R2 = H
N
O
8
9
10
S
HO2C
O
O
S
f
1
2
20
R = CO2H, R = H
Scheme 1. Synthesis of compounds 2–10. Reagents and conditions: (a) (i) (COCl)2, DMF (cat), CH2Cl2, rt, (ii) 4-aminobenzenesulfonic acid, 0.5 M NaOH, Na2CO3, NBu4Br, THF,
12 h, (iii) 1N HCl, 69%; (b) (i) (COCl)2, DMF (cat), CH2Cl2, rt, (ii) ethyl 4-aminobenzoate, NEt3, CH2Cl2, rt, 12 h, 84-87%; (c) (i) (COCl)2, DMF (cat), CH2Cl2, rt, (ii) 4-amino benzoic
acid, pyridine, THF, rt, 12 h, 44%; or 4-amino-2-hydroxybenzoic acid, acetone, rt, 12 h, 61%; (d) (i) 1 M NaOH, EtOH, 60 °C, (ii) 10% HCl, 71–88%; (e) (i) CH2Cl2, (COCl)2, DMF
(cat), rt, (ii) sulfathiazole, pyridine, acetone, rt, for 2 and 9, 68–76%; or sulfathiazole, NEt3, THF, rt, for 3–5, 52–70%; or 4-(methylamino)-N-(thiazol-2-yl)benzene-sulfonamide,
pyridine, acetone, rt, for 6 and 7, 50–62%; or sulfathiazole, NEt3, THF, rt, for 8, 16%; (f) (i) 0.3 M NaOH, EtOH, 60 °C, (ii) 10% HCl, 50%.
NO2
NH2
H
N
H
N
h
g
H
N
N
+
OHC
NO2
S
N
S
O
O
S
21
O
O
S
OCH3
NH2
H
N
H
N
L
H
N
i
H
N
N
H
N
S
N
O
O
11 L = -CO-
12
S
22
S
O
O
L = -CH2-
S
Scheme 2. Synthesis of compounds 11–12. Reagents and conditions: (g) EtOH, reflux, 3 h, then NaBH4, reflux, 30 min, 73%; (h) 10% Pd/C, NaBH4, 2 M NaOH, H2O, rt, 2 h, 91%;
(i) 4-methoxybenzoyl chloride, NEt3, acetone, for 11, 30%; 4-methoxybenzaldehyde, EtOH, reflux, 2 h, then NaBH4, reflux, 1 h, for 12, 35%.
O
O
O
H
CHO
l
N
j
R1
H3C
H
N
+
HO2C
R1
O
O
N
R1
HO2C
23a R1 = OCH3
23b R1 = CH3
S
13 R1 = OCH3
14 R1 = CH3
O
O
O
S
1 = CN
R
1 = CN
15
R
23c
O
k
H
N
O
R1
l
H
N
N
S
HO2C
R1
O
S
24 R1 = OCH3
R1 = OCH3
16
Scheme 3. Synthesis of compounds 13–16. Reagents and conditions: (j) (i) NaOH, H2O/EtOH, rt, 10 h, (ii) 20% HCl, 35–71%; (k) H2, Pd/C, diphenyl sulfide (cat), MeOH, rt, 6 h,
96%; (l) (i) SOCl2, DMF (cat), CH2Cl2, rt, (ii) sulfathiazole, NEt3, CH2Cl2, rt, 4 h, 58–81%.

2316
Y.-B. Tang et al. / Bioorg. Med. Chem. Lett. 23 (2013) 2313–2318
Table 2
served in significant amount even with shortened reaction time.
The catalyst poison diphenyl sulfide was then added to decrease
the catalytic reactivity of Pd/C, and selective reduction of the
double bond provided ketone 24 as the only product. The target
compounds 13–16 were then prepared from 23 or 24 and sulfat-
hiazole via acylation.
PTP1B inhibitory activity of compounds 14–1610
Compd
IC50 M)
14
15
16
1a
(
l
8.3 0.3
8.1 0.3
3.2 0.1
8
a
The IC50 value of compound 1 was taken from Ref. 7.
All target compounds were characterized by melting point, 1H
NMR, and mass spectra analyses.8
Inhibitory activity against human recombinant PTP1B: For the first
round, eleven compounds (2–12) were prepared and evaluated
against the human recombinant PTP1B. Most compounds showed
apparent inhibition against PTP1B, and compounds 3, 4 and 8
exhibited significant PTP1B inhibitory activity (Table 1). The results
suggested preliminary structure–activities relationship (SAR)
clues. Compound 2, which had minimal structural alteration from
the template structure 1, showed moderate inhibition against
PTP1B and supported the strategy of pharmacophore-oriented
scaffold hopping. An amide instead of sulfonamide linkage be-
tween rings B and C improved the PTP1B inhibitory activity (2 vs
3). However, introducing of a methyl group at the amide nitrogen
obviously decreased the activity against PTP1B (3 vs 6), which im-
plies that an indispensable hydrogen bond interaction might be
blocked by methylation, or alternatively, the steric hindrance at
this position could impair inhibitor-enzyme interactions. The for-
mation of intramolecular hydrogen bond exhibited no noticeable
effect on PTP1B inhibition (3 vs 8). Furthermore, the loss of activity
of compounds 11 and 12 indicated the essentiality of amide or sul-
fonamide linkage to ensure the active conformation. R1 might also
affect the PTP1B inhibitory activity (3, 4 vs 5).
Figure 4. Kinetic analysis of the inhibition against PTP1B-catalyzed pNPP hydro-
lysis by compound 16. The tested concentrations were 0 (d), 2.5 (s), 5 (N), and 7.5
(D) lM, respectively.
Table 3
Selective inhibitory activity against PTPs10
The optimal X, R1, R2 and R3 were selected as X = –CO–, R1 = –
OCH3– and R2 = R3 = H from the first round of synthesis. With these
molecular parts fixed, the effect of the ‘L’ group on PTP1B inhibi-
tory activity was further explored by replacing the amide linkage
(L = –NH–) between rings A and B with vinyl or ethylene group
(L = –CH@CH– or –CH2–CH2–), and compounds 13–16 were syn-
thesized. When tested against human recombinant PTP1B, com-
pounds 14–16 showed significant inhibition (Table 1), which is
probably because an elongated L group brings R1 to the vicinity
of Phe196 and favors the formation of hydrogen bond interaction.
Compounds 14–16 were further determined for their IC50 values
against PTP1B (Table 2). All of them showed IC50 values compara-
ble or even superior to that of 1.
IC50 (M)
14
15
16
PTP1B
SHP2
VHR
LYP
PTP
8.3 0.3
6.9 0.6
9.0 0.9
10.4 0.3
NDa
8.1 0.3
7.8 0.8
9.2 0.7
8.0 0.4
ND
3.2 0.1
8.0 1.0
>10
>10
>10
TC-PTP
CDC25
ND
ND
ND
ND
40 10
>100
a
ND: not determined.
LYP (lymphoid-specific protein tyrosine phosphatase), PTP
tein tyrosine phosphatase
a
(pro-
a
), TC-PTP (T-cell protein tyrosine phos-
To determine whether compound 16 inhibits PTP1B in a non-
competitive manner, its effect on PTP1B-catalyzed pNPP hydrolysis
was examined.10 As evidenced by the Lineweaver-Burk plot
(Fig. 4), compound 16 is indeed a non-competitive inhibitor with
phatase) and CDC25 (cell division cycle 25 phosphatases).10 The
IC50 values of compounds 14–16 against the seven PTPs were re-
ported in Table 3. Impressively, compound 16 showed significant
selectivity to PTP1B over TC-PTP, the most homologous enzyme
of PTP1B in the PTPs family.
a Ki value of 5.8 0.3 lM.
Selectivity of compounds 14–16 against a panel of PTPs: Com-
pounds 14–16 were evaluated against a panel of PTPs, including
PTP1B, SHP2 (Src homology2 domain containing phosphotyrosine
phosphatase 2), VHR (human vaccinia H1-related phosphatase),
A docking study11 of compound 16 indicated that it interacted
with the allosteric site of PTP1B in a way similar to that of the tem-
plate structure 1. The phenyl B ring situates in the hydrophobic
pocket formed by the side chains of Ala189, Leu192 and Phe280.
p–p stacking interaction was observed between the thiazole ring
and Phe280. Hydrogen bond interactions are also monitored be-
tween the ketone oxygen in 16 and Asn193 as well as the amide
nitrogen and Glu276 (Fig. 5).
Table 1
Percentage inhibition of the tested compounds against human recombinant PTP1B9
Compd Percentage
inhibitiona
(%)
Compd Percentage
inhibitiona
(%)
Compd Percentage
inhibitiona
(%)
In vivo insulin-sensitising effects: Compound 16 was further
examined for its in vivo insulin-sensitising effects with the hype-
rinsulinemic–euglycemic clamp test.9 Insulin resistant model
(IRM) obesity mice were induced by high-fat diet (HFD) as previ-
ously described.9 The insulin sensitizer Rosiglitazone was used as
a positive control. As illustrated in Fig. 6, the glucose infusion rate
(GIR) values in the IRM model mice were decreased significantly as
compared with those of mice in the blank control group (Con)
(12.8 2.8 vs 2.1 0.2 mg/min/kg), which indicated the validity
of the HFD-induced insulin resistance model. While oral adminis-
tration of Rosiglitazone (Rosi) and compound 16 to IR mice in-
2
3
4
5
6
45.3
81.9
87.1
53.1
24.6
7
8
9
23.8
78.1
NA
12
NA
13b
14b
15b
16b
20.7
79.8
68.9
75.7
10
11
NA
NA
NA: Not active, no noticeable inhibition was observed under the test concentration.
The percentage inhibition was measured under the concentration of 100 lM,
except other specified.
a
b
The percentage inhibition was measured under the concentration of 10 lM.

Y.-B. Tang et al. / Bioorg. Med. Chem. Lett. 23 (2013) 2313–2318
2317
Figure 5. Interaction mode of compound 16 with the allosteric site of PTP1B.
ences without modification) associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/
j.bmcl.2013.02.073.
16
14
12
10
8
References and notes
1. Barr, A. J. Future Med. Chem. 2010, 2, 1563.
2. Montalibet, J.; Kennedy, B. P. Drug Discovery Today Ther. Strateg. 2005, 2, 129.
3. Zhang, S.; Zhang, Z. Drug Discovery Today 2007, 12, 373.
4. Combs, A. P. J. Med. Chem. 2010, 53, 2333.
5. Lee, S.; Wang, Q. Med. Res. Rev. 2007, 27, 553.
6. Nichols, A. J.; Mashal, R. D.; Balkan, B. Drug Dev. Res. 2006, 67, 559.
7. Wiesmann, C.; Barr, K. J.; Kung, J.; Zhu, J.; Erlanson, D. A.; Shen, W.; Fahr, B. J.;
Zhong, M.; Taylor, L.; Randal, M.; McDowell, R. S.; Hansen, S. K. Nat. Struct. Mol.
Biol. 2004, 11, 730.
6
4
###
2
0
Con
IRM
Rosi
16
8. Compound 2: Yield 68%; whitish solid; mp 239–241 °C; 1H NMR (300 MHz,
DMSO-d6): d 3.84 (s, 3H), 6.80 (d, 1H, J = 4.2 Hz), 7.07 (d, 2H, J = 8.1 Hz), 7.18–
7.25 (m, 3H), 7.66 (d, 2H, J = 8.1 Hz), 7.79 (d, 2H, J = 8.7 Hz), 7.94 (m, 4H), 10.44
(s, 1H), 10.75 (s, 1H), 12.67 (s, 1H); FAB-MS: m/z 545 [M+H]+. Compound 3:
Yield 59%; whitish solid; mp >350 °C; 1H NMR (300 MHz, DMSO-d6): d 3.85(s,
3H), 6.82 (d, 1H, J = 3.9 Hz), 7.09 (t, 2H, J = 8.7 Hz), 7.26 (d, 1H, J = 3.9 Hz), 7.79
(d, 2H, J = 8.7 Hz), 7.93–8.01 (m, 8H), 10.38(s, 1H), 10.46 (s, 1H), 12.71 (s, 1H);
FAB-MS: m/z 509 [M+H]+. Compound 4: Yield 52%; light yellow solid; mp 301–
304 °C; 1H NMR (300 MHz, DMSO-d6): d 6.83 (d, 1H, J = 4.5 Hz), 7.26 (d, 1H,
J = 4.5 Hz), 7.79 (d, 2H, J = 8.7 Hz), 7.93–8.07 (m, 8H), 8.14 (d, 2H, J = 8.4 Hz),
10.48 (s, 1H), 10.76 (s, 1H), 12.71 (s, 1H); FAB-MS: m/z 504 [M+H]+. Compound
5: Yield 70%; whitish solid; mp 307–309 °C; 1H NMR (300 MHz, DMSO-d6): d
6.83 (d, 1H, J = 4.5 Hz), 7.26 (d, 1H, J = 4.5 Hz), 7.40 (t, 2H, J = 9.0 Hz), 7.79 (d,
2H, J = 9.0 Hz), 7.93–8.10 (m, 8H), 10.47 (s, 1H), 10.55 (s, 1H), 12.70 (s, 1H);
FAB-MS: m/z 497 [M+H]+. Compound 6: Yield 62%; whitish solid; mp 208–
212 °C; 1H NMR (300 MHz, DMSO-d6): d 3.39 (s, 3H), 3.83 (s, 3H), 6.83 (d, 1H,
J = 4.5 Hz), 7.05 (d, 2H, J = 8.4 Hz), 7.24–7.32 (m, 5H), 7.65 (d, 2H, J = 7.8 Hz),
7.68 (d, 2H, J = 8.7 Hz), 7.93 (d, 2H, J = 8.4 Hz), 10.17 (s, 1H), 12.76 (s, 1H); EI-
MS: m/z 522 [M]+. Compound 7: Yield 50%; whitish solid; mp 248–251 °C; 1H
NMR (300 MHz, DMSO-d6): d 3.39 (s, 3H), 6.83 (d, 1H, J = 4.5 Hz), 7.25 (d, 1H,
J = 4.5 Hz), 7.28 (d, 2H, J = 8.4 Hz), 7.31 (d, 2H, J = 8.4 Hz), 7.64 (d, 2H,
J = 8.4 Hz), 7.67 (d, 2H, J = 8.4 Hz), 8.02 (d, 2H, J = 8.1 Hz), 8.07 (d, 2H,
J = 8.1 Hz), 10.56 (s, 1H), 12.75 (s, 1H); FAB-MS: m/z 518 [M+H]+. Compound
8: Yield 16%; whitish solid; mp 309–311 °C; 1H NMR (300 MHz, DMSO-d6): d
3.84 (s, 3H), 6.82 (d, 1H, J = 4.5 Hz), 7.08 (d, 2H, J = 7.8 Hz), 7.24 (d, 1H,,
J = 4.5 Hz), 7.34 (d, 1H, J = 8.7 Hz), 7.65 (s, 1H), 7.78 (d, 2H, J = 7.8 Hz), 7.86 (d,
2H, J = 7.8 Hz), 7.90–8.05 (m, 3H), 10.28 (s, 1H), 10.51 (s, 1H), 11.88 (s, 1H),
12.70 (s, 1H); FAB-MS: m/z 525 [M+H]+. Compound 9: Yield 76%; whitish solid;
Figure 6. Glucose infusion rates (GIR) in hyperinsulinemic–euglycemic clamp test
(n = 8).9 ###P <0.001 versus Con; ⁄⁄⁄p <0.001 versus IRM.
creased the GIR values to 243% and 148%, respectively. These
observations implicated that similar to Rosiglitazone, compound
16 could ameliorated the impaired insulin sensitivity in IRM mice.
In conclusion, a series of novel PTP1B inhibitors were designed
based on the template structure 1. Biological evaluation identified
compound 16 as a new and non-competitive PTP1B inhibitor.
Excellent selectivity over a panel of PTPs and significant in vivo
insulin sensitizing effect were achieved. The current results sup-
port the strategy to target the allosteric binding site of PTP1B as
an approach to overcome the poor selectivity and inadequate
in vivo activity associated with most of the known potent PTP1B
inhibitors. Further hit evolution following this molecular design
are undergoing and will be reported in due course.
Acknowledgments
This research work was financially supported by National Nat-
ural Science Foundation of China (20972192) and Beijing Natural
Science Foundation (7102117) awarded to Z.X., National S & T Ma-
jor Project (2012ZX09103-101-063) to F.Y. and National Institutes
of Health Grant (CA152194) to Z.-Y.Z.
mp >350 °C; 1H NMR (300 MHz, DMSO-d6):
d 3.91 (s, 3H), 6.66 (d, 1H,
J = 4.5 Hz), 7.12 (d, 1H, J = 4.5 Hz), 7.76 (d, 2H, J = 8.7 Hz), 7.89 (d, 2H,
J = 8.7 Hz), 7.96 (d, 2H, J = 8.7 Hz), 8.01 (d, 2H, J = 8.7 Hz), 8.12 (s, 4H), 10.43
(s, 1H), 10.75 (s, 1H); ESI-MS: m/z 537 [M+H]+. Compound 10: Yield 50%;
whitish solid; mp >350 °C; 1H NMR (300 MHz, DMSO-d6): d 6.68 (d, 1H,
J = 4.2 Hz), 7.14 (d, 1H, J = 4.2 Hz), 7.75 (d, 2H, J = 8.7 Hz), 7.89 (d, 2H,
J = 8.7 Hz), 7.94–8.05 (m, 8H), 10.44 (s, 1H), 10.67 (s, 1H); ESI-MS: m/z 523
[M+H]+. Compound 11: Yield 50%; whitish solid; mp 233–236 °C; 1H NMR
(300 MHz, DMSO-d6): d 3.84 (s, 3H), 4.27 (d, 2H, J = 5.4 Hz), 6.61 (d, 2H,
J = 8.7 Hz), 6.74 (d, 1H, J = 4.5 Hz), 6.99 (t, 1H, J = 5.4 Hz), 7.05 (d, 2H, J = 8.7 Hz),
7.18 (d, 1H, J = 4.5 Hz), 7.29 (d, 2H, J = 8.7 Hz), 7.47 (d, 2H, J = 8.7 Hz), 7.70 (d,
Supplementary data
Supplementary data (supplementary information on general
preparation procedures is available online. Biological assays were
performed by following experimental protocols in the cited refer-

2318
Y.-B. Tang et al. / Bioorg. Med. Chem. Lett. 23 (2013) 2313–2318
2H, J = 8.7 Hz), 7.95 (d, 2H, J = 8.7 Hz), 10.06 (s, 1H), 12.42 (s, 1H); ESI-MS: m/z
Yield 81%; whitish solid; mp 245–247 °C; 1H NMR (300 MHz, DMSO-d6): d 3.00
(t, 2H, J = 7.2 Hz), 3.37 (t, 2H, J = 7.2 Hz), 3.82 (s, 3H), 6.81 (d, 1H, J = 4.5 Hz),
7.02 (d, 2H, J = 8.7 Hz), 7.24 (d, 1H, J = 4.5 Hz), 7.44 (d, 2H, J = 8.1 Hz), 7.76 (d,
2H, J = 9.0 Hz), 7.85–7.98 (m, 6H), 10.45 (s, 1H), 12.69 (s, 1H); HR-ESI-MS: m/z
522.1135 [M+H]+ (calcd for C26H24N3O5S2: 522.1157).
495 [M+H]+. Compound 12: Yield 35%; mp 111–113 °C; 1H NMR (300 MHz,
DMSO-d6): d 3.71 (s, 3H), 4.07 (d, 2H, J = 6.0 Hz), 4.15 (d, 2H, J = 6.0 Hz), 6.09 (t,
1H, J = 6.0 Hz), 6.52 (d, 2H, J = 8.7 Hz), 6.58 (d, 2H, J = 8.7 Hz), 6.73 (d, 1H,
J = 4.5 Hz), 6.76 (t, 1H, J = 6.0 Hz), 6.86 (d, 2H, J = 8.7 Hz), 7.00 (d, 2H, J = 8.7 Hz),
7.18 (d, 1H, J = 4.5 Hz), 7.25 (d, 2H, J = 8.7 Hz), 7.44 (d, 2H, J = 8.7 Hz), 12.42 (s,
1H); ESI-MS: m/z 481 [M+H]+. Compound 13: Yield 58%; light yellow solid; mp
9. The measurements were performed by following the experimental protocols
described in this cited reference Ma, Y.-M.; Tao, R.-Y.; Liu, Q.; Li, J.; Tian, J.-Y.;
Zhang, X.-L.; Xiao, Z.-Y.; Ye, F. Mol. Cell. Biochem. 2011, 357, 65.
235–237 °C; 1H NMR (300 MHz, DMSO-d6):
d 3.87 (s, 3H), 6.82 (d, 1H,
J = 4.5 Hz), 7.10 (d, 2H, J = 8.7 Hz), 7.25 (d, 1H, J = 4.5 Hz), 7.73–7.81 (m, 3H),
7.94 (d, 2H, J = 8.7 Hz), 7.99–8.10 (m, 5H), 8.19 (d, 2H, J = 8.7 Hz), 10.62 (s, 1H),
12.70 (s, 1H); HR-ESI-MS: m/z 520.0983 [M+H]+ (calcd for C26H22N3O5S2:
520.1001). Compound 14: Yield 74%; light yellow solid; mp 220–223 °C; 1H
NMR (300 MHz, DMSO-d6): d 2.41 (s, 3H), 6.82 (d, 1H, J = 4.5 Hz), 7.25 (d, 1H,
J = 4.5 Hz), 7.39 (d, 2H, J = 7.8 Hz), 7.75–7.81 (m, 3H), 7.94 (d, 2H, J = 8.7 Hz),
7.99–8.11 (m, 7H), 10.62 (s, 1H), 12.70 (s, 1H); HR-ESI-MS: m/z 504.1027
[M+H]+ (calcd for C26H22N3O4S2: 504.1052). Compound 15: Yield 78%; light
yellow solid; mp 260–262 °C; 1H NMR (300 MHz, DMSO-d6): d 6.81 (d, 1H,
J = 4.5 Hz), 7.24 (d, 1H, J = 4.5 Hz), 7.77–7.86 (m, 4H), 7.94 (d, 2H, J = 8.4 Hz),
8.01–8.09 (m, 6H), 8.31 (d, 2H, J = 8.7 Hz), 10.62 (s, 1H), 12.62 (br, 1H); HR-ESI-
MS: m/z 515.0836 [M+H]+ (calcd for C26H19N4O4S2: 515.0848). Compund 16:
10. The measurements were performed by following the experimental protocols
described in the cited references (a) Liu, S.; Zeng, L.-F.; Wu, L.; Yu, X.; Xue, T.;
Gunawan, A. M.; Long, Y.-Q.; Zhang, Z.-Y. J. Am. Chem. Soc. 2008, 130, 17075; (b)
He, R.; Zeng, L.-F.; He, Y.; Zhang, S.; Zhang, Z.-Y. FEBS J. 2013, 280, 731.
11. The structure of compound 16 was generated and molecular docking was
performed with the Discovery Studio 2.5 software package (Accelrys, San
Diego, USA).The complex structure of 1-PTP1B was obtained from the Protein
Data Bank (PDB code: 1t4j). The docking calculation was carried out with the
LibDock protocol. Default settings were used. All calculations were performed
on a DELL Precision T5500 workstation. The Ligand Interactions tool in MOE
(Chemical Computing Group, Inc., Montreal, Quebec, Canada) was used to plot
the 2D interaction modes.