Isoxazole carboxylic acids as protein tyrosine phosphatase 1B (PTP1B) inhibitors

Article abstract of DOI:10.1016/j.bmcl.2004.08.063

Guided by X-ray crystallography, we have extended the structure-activity relationship (SAR) study on an isoxazole carboxylic acid-based PTP1B inhibitor (1) and more potent and equally selective (>20-fold selectivity over the highly homologous T-cell PTPase, TCPTP) PTP1B inhibitors were identified. Inhibitor 7 demonstrated good cellular activity against PTP1B in COS 7 cells. Guided by X-ray crystallography, we have extended the structure-activity relationship (SAR) study on an isoxazole carboxylic acid-based PTP1B inhibitor (1) and more potent and equally selective (>20-fold selectivity over the highly homologous T-cell PTPase, TCPTP) PTP1B inhibitors were identified. Inhibitor 7 demonstrated good cellular activity against PTP1B in COS 7 cells.

Full text of DOI:10.1016/j.bmcl.2004.08.063

Bioorganic & Medicinal Chemistry Letters 14 (2004) 5543–5546
Isoxazole carboxylic acids as protein tyrosine phosphatase
1B (PTP1B) inhibitors
Hongyu Zhao,a,* Gang Liu,a Zhili Xin,a Michael D. Serby,a Zhonghua Pei,a
Bruce G. Szczepankiewicz,a Philip J. Hajduk,b Cele Abad-Zapatero,b
Charles W. Hutchins,b Thomas H. Lubben,a Stephen J. Ballaron,a Deanna L. Haasch,a
Wiweka Kaszubska,a Cristina M. Rondinone,a James M. Trevillyana and
Michael R. Jirouseka,
aMetabolic Disease Research, Global Pharmaceutical Research and Development, Abbott Laboratories,
Abbott Park, IL 60064-6098, USA
bAdvanced Technology, Global Pharmaceutical Research and Development, Abbott Laboratories,
Abbott Park, IL 60064-6098, USA
Received 9 August 2004; accepted 31 August 2004
Available online 18 September 2004
Abstract—Guided by X-ray crystallography, we have extended the structure–activity relationship (SAR) study on an isoxazole carb-
oxylic acid-based PTP1B inhibitor (1) and more potent and equally selective (>20-fold selectivity over the highly homologous T-cell
PTPase, TCPTP) PTP1B inhibitors were identified. Inhibitor 7 demonstrated good cellular activity against PTP1B in COS 7 cells.
Ó 2004 Elsevier Ltd. All rights reserved.
Protein tyrosine phosphatase 1B (PTP1B), an intracellu-
lar PTPase, has been implicated in negative regulation of
the insulin and leptin signal transduction pathways
among other biological functions.1 Studies from two
laboratories have shown that PTP1B knockout mice ex-
hibit the phenotypes of increased insulin sensitivity, im-
proved glucose tolerance, and resistance to diet-induced
obesity.2,3 In a more therapeutically relevant study, Zin-
ker et al. have demonstrated that a PTP1B antisense
oligonucleotide normalizes blood glucose and improves
insulin sensitivity in diabetic mice via a mechanism of
lowering PTP1B protein expression.4 A synthetic small
molecule that selectively inhibits PTP1B action is ex-
pected to have similar beneficial effects in type 2 diabetic
and obesity patients.
hydrophilic and homologous to other PTPases.5,6 These
structural features were further substantiated by that
fact that most known PTP1B inhibitors are large
(MW > 500) and hydrophilic molecules containing mul-
tiple acid and peptide groups.7 Nonetheless, cell perme-
able PTP1B inhibitors have been reported in several
laboratories.8 High selectivity over PTPases other than
TCPTP9 has been routinely achieved, but only a few
compounds showed moderate selectivity over TCPTP
(ꢀ10-fold).10
O
HO
N
O
binds to the
catalytic site
binds to
site 2
F
MeO
O
O
linker
Two major challenges in developing potential therapeu-
tics targeting PTP1B are selectivity and cell permeabil-
ity. The X-ray crystal structure revealed that the
PTP1B catalytic site and vicinal binding sites are highly
OH
1
K
6.9 µM (PTP1B), K 164 µM (TCPTP)
i
i
Keywords: PTP1B; Selective; Cell permeable.
*
Corresponding author. Tel.: +1 847 935 4566; fax: +1 847 938
1674; e-mail: hongyu.zhao@abbott.com
We have previously reported the discovery of isoxazole
carboxylic acid-based PTP1B inhibitors such as 1.11
Present address: Pfizer Global R&D, La Jolla Laboratories, 10770
Science Center Drive, San Diego, CA 92121-1187, USA.
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.08.063

5544
H. Zhao et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5543–5546
Here we describe our efforts to extend the SAR studies
leading to more potent PTP1B inhibitors with similarly
high selectivity and cell permeability.
introduction of an amino group to the only chemically
modifiable position 4 of the isoxazole ring provided a
more active inhibitor 7 with good selectivity over
TCPTP (Ki 2.1lM, PTP1B; Ki > 30lM, TCPTP). The
calculated pKa values using ACD/pKa DB Version 7.0
for the carboxylic acid in 7 was 4.05 in comparison to
3.31 if the amino group was replaced by a hydrogen.
We think the increased pKa value of the acid in 7 was
at least partially responsible for its cell permeability
(vide infra). Alkylation or acylation of the amino group
in 7 provided only weaker inhibitors due to the limited
size of the catalytic site. Replacing the amino group with
a hydroxy methyl group led to a submicromolar inhibi-
tor 8 (Ki 0.92lM, PTP1B; Ki 19.2lM, TCPTP). How-
ever, the calculated pKa for the carboxylic acid in 8
was 2.95 and inhibitor 8 was inactive in COS7 cellular
assay.
Our SAR plan was directed by the X-ray crystal struc-
ture of PTP1B in complex with 1.11 It appeared a hydro-
gen bond donor on or attached to any of the three-
carbon atom in the linker would have a chance to inter-
act with the carboxyl group of Asp48.12 Unfortunately,
all efforts to introduce hydrogen bond donors on the
carbon linker to interact with Asp 48 resulted in less ac-
tive compounds. For example, replacing the CH@CH
group of the linker in 1 with an NHCO group only pro-
duced a weaker inhibitor (2, Ki 216lM).13 A hydroxy
group on the linker (3, Ki 122lM) did not seem to inter-
act with Asp48 favorably. Changing the nature of H-
bond donor by removal of the carbonyl oxygen in 2
led to a more potent inhibitor 4 (Ki 60.3lM), but weaker
than 1. Evidently, these hydrogen bond donors were
either not appropriately positioned or their introduction
altered the linker conformation too much. Cyclopropa-
nation of the linker double bond to fine tune the linker
conformation led to a less potent inhibitor (5, Ki 23lM).
The syntheses of 7 and 8 are described in Scheme 1.
Coupling of b-keto nitro compound (9a) or b-keto ester
(9b) with ethyl chlorooximido acetate followed by an
intramolecular dehydration process gave iodo com-
pounds 10a and 10b.14 Starting materials 9a15 and 9b16
were synthesized using known protocols. Reduction of
the nitro group in 10a gave amine 11a. Removal of
the tert-butyl group in 10b followed by mixed anhydride
formation and subsequent NaBH4 reduction, then pro-
tection of the resulting alcohol with an acetyl group pro-
duced 11b. Alcohol 12a and 12b could then be obtained
via a Stille coupling reaction between iodides 11a and
11b and 3-tributylstannyl-1-propenol.17 Finally, intro-
duction of the salicylate via a Mitsunobu reaction with
methyl 2,6-dihydroxylbenzoate produced 13a and 13b.
Selective hydrolysis of 13a and 13b led to inhibitors 7
and 8, respectively.
O
O
HO
N
HO
N
O
O
R1
R2
MeO
O
O
MeO
O
O
R1 R2
X
OH
OH
2 X=NH, R1, R2=O Ki 216 µM
6 R1, R2=CH2CH2 Ki 11.5 µM
7 R1=NH2, R2=H Ki 2.1 µM
8 R1=CH2OH, R2=H Ki 0.92 µM
3 X=CH2, R1=H, R2=OH Ki 122 µM
4 X=NH, R1=H, R2=H Ki 60.3 µM
5 X,R1=CHCH2, R2=H Ki 23 µM
O
O
EtO
EtO
Another observation from the crystal structure of the
PTP1B-inhibitor 1 complex was that the isoxazole ring
and the benzene ring were nearly co-planar. The unfav-
orable interactions between C–HÕs (or C–F) of the ben-
zene and isoxazole rings would be eliminated by
chemically forming a ring system between them. How-
ever, one molecule of this kind, tricyclic inhibitor 6,
did not show any improvement (Ki 11.5lM), even
though the crystal structure revealed no apparent inter-
ference between the added two methylene groups and
the protein surface.
N
O
N
O
R
O
R
R
i
ii
iii
I
I
I
9a R=NO2
10a R=NO2
11a R=NH2
9b R=CO2tBu
10b R=CO2tBu 11b R=CH2OAc
O
O
EtO
EtO
7 (from 13a)
8 (from 13b)
N
N
iv
v
O
O
R
R
The crystal structure of the PTP1B-inhibitor 1 complex
suggested that only small additional groups could be
accommodated in the catalytic site. Due to the hydro-
philic nature of PTP1B catalytic site, polar groups are
likely to provide more favorable interactions, but they
are also likely to have negative impact on cell permeabil-
ity. One way to possibly solve this dilemma is to intro-
duce polar groups in such a way that they could
increase the pKa value of the isoxazole carboxylic acid
and therefore increase the concentration of the neutral
form of the acid in physiological environment. Thus,
MeO
O
O
OH
OH
12a R=NH2
12b R=CH2OAc
13a R=NH2
13b R=CH2OAc
Scheme 1. Reagents and conditions: (i) ethyl chlorooximido acetate,
Et3N; (ii) 10a–11a: Fe, NH4Cl; 10b–11b: (a) TFA; (b) methylchloro-
formate, Et3N; (c) NaBH4; (d) AcCl, pyridine; (iii) 3-tributylstannyl-1-
propenol, Pd2(dba)3, (2-furyl)3P, CuI, DMF; (iv) Ph3P, DEAD, methyl
2,6-dihydroxylbenzoate; (v) NaOH.

H. Zhao et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5543–5546
5545
The PTP1B inhibitors disclosed here demonstrated
excellent selectivity over other PTPases. For example,
inhibitor 8 showed greater than 20-fold selectivity and
inhibitor 7 was inactive at 30lM (highest concentration
tested) toward TCPTP, the most homologous PTPase to
PTP1B. The earlier SAR results revealed that the second
site binder (methyl salicylate) contributed to 5–10-fold12
and the catalytic site binding moiety (isoxazole carboxy-
lic acid) provided another 2–4-fold11 to the selectivity
over TCPTP. These compounds showed no inhibitory
activity against LAR, CD45, cdc25, and SHP-2 at high-
est concentration tested (300lM). To the best of our
knowledge, this series of inhibitors represent the most
selective PTP1B inhibitors over TCPTP reported to
date.10
Figure 2. X-ray crystal structure of PTP1B in complex with inhibitor 8
(carbon is in orange and green for inhibitor 8, oxygen is in red,
nitrogen is in blue, and sulfur is in yellow).
Although these inhibitors possess a reasonable struc-
tural profile (MW < 450, contain only one carboxylic
acid, and few rotatable bonds), only inhibitors 7 showed
PTP1B inhibitory activity in COS7 cells as described be-
low. The beneficial effect of the amino group in 7 most
likely derived from its electron donating and hydrogen
bonding ability that reduce the acidity of the adjacent
carboxylic acid as discussed before.
site 2 of PTP1B.22 The isoxazole carboxylic acid binds
to the active site of PTP1B with WPD loop in the closed
conformation. A salt bridge between the carboxylate
and Arg221 provides a critical interaction. The methyl-
enehydroxyl group on the isoxazole ring forms a hydro-
gen bond with Asp181 carboxyl group. The isoxazole
and phenyl rings interact with a hydrophobic pocket
normally occupied by the phenyl ring of pTyr. The ether
oxygen of the salicylate forms a hydrogen bond with
Gln262. The salicylate carboxylate group adapts an
out of plane conformation, and is within hydrogen
bonding distance to Arg254, Tyr20, and Gln262. The
hydroxyl group of the salicylate hydrogen bonds with
Arg24 and Arg254. The aromatic portion of the salicy-
late lies on top of the hydrophobic side chain of
Met258, providing a complimentary van der Waals
interaction.
PTP1B has been shown to negatively regulate leptin
pathways by dephosphorylating (inactivating) Janus
kinase (JAK)2 both in vitro18 and in vivo.19 Inhibitor
7 was analyzed in COS7 cells that were transiently trans-
fected with exogenous PTP1B for its ability to reverse
PTP1B-induced signal transducer and activator of trans-
cription (STAT)3 dephosphorylation. The decreased
STAT3 phosphorylation caused by PTP1B transfection
(25ng) was attenuated by inhibitor 7 in a dose-depend-
ent manner (Fig. 1), similar to the effect exerted by a
nonspecific phosphatase inhibitor BpV(pic).20,11 The
more potent inhibitor 8 showed no activity in the
COS7 cellular assays, presumably due to the increased
polarity and decreased pKa value of the acid group com-
pared to 7.
In conclusion, guided by X-ray crystallography, we have
designed and synthesized several potent and highly
selective PTP1B inhibitors. These inhibitors contain
only one carboxylic acid and have relatively low mole-
cular weight and polarity. Inhibitor 7 showed good cell
permeability in COS7 cells.
The X-ray crystal structures of PTP1B (residues 1–322)
in complex with inhibitors 4–8 were determined and they
shared an almost identical binding mode. A representa-
tive structure (PTP1B/8) is shown in Figure 2.21 As we
expected, these compounds bind both catalytic site and
Acknowledgments
We thank Dr. Yvonne C. Martin for pKa value
calculations.
120
100
80
References and notes
´
1. (a) Cheng, A.; Dube, N.; Gu, F.; Tremblay, M. L. Eur. J.
Biochem. 2002, 269, 1050–1059; (b) Zhang, Z.; Lee, S. Y.
Expert Opin. Investig. Drug 2003, 12, 223–233.
2. Elchebly, M.; Payette, P.; Michaliszyn, E.; Cromlish, W.;
Collins, S.; Loy, A. L.; Normandin, D.; Cheng, A.;
Himms-Hagen, J.; Chan, C. C.; Ramachandran, C.;
Gresser, M. J.; Tremblay, M. L.; Kennedy, B. P. Science
1999, 283, 1544–1548.
60
40
20
0
P=0.001
P=0.022
100 uM
P=0.042
30 uM
0 uM
0 uM
10 uM
10uM
BpV(Pic)
3. Klaman, L. D.; Boss, O.; Peroni, O. D.; Kim, J. K.;
Martino, J. L.; Zabolotny, J. M.; Moghal, N.; Lubkin, M.;
Kim, Y. B.; Sharpe, A. H.; Stricker-Krongrad, A.;
Shulman, G. I.; Neel, B. G.; Kahn, B. B. Mol. Cell. Biol.
2000, 20, 5479–5489.
0 ng PTP1B
25 ng PTP1B
Figure 1. Reversal of PTP1B-dependent dephosphorylation of STAT3
in COS 7 cells by PTP1B inhibitor 7.

5546
H. Zhao et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5543–5546
4. Zinker, B. A.; Rondinone, C. M.; Trevillyan, J. M.; Gum,
S. J.; Haasch, D. L.; Kaszubska, W.; Rondinone, C. M.;
Trevillyan, J. M.; Jirousek, M. R. J. Med. Chem. 2003, 46,
4232–4235.
R. J.; Clampit, J. E.; Waring, J. F.; Xie, N.; Jacobson, P.;
Frost, L.; Kroeger, P. E.; Reilly, R. M.; Koterski, A.;
Opgenorth, T. J.; Ulrich, R. G.; Crosby, S.; Butler, M.;
Murray, S. F.; McKay, R. A.; Bhanot, S.; Monia, B. P.;
Jirousek, M. R. Proc. Natl. Acad. Sci. U.S.A. 2002, 99,
11357–11362.
12. Szczepankiewicz, B. G.; Liu, G.; Pei, Z.; Xin, Z.; Lubben,
T. H.; Trevillyan, J. M.; Stashko, M. A.; Ballaron, S. J.;
Hajduk, P. J.; Liang, H.; Huang, F.; Hutchins, C. W.;
Abad-Zapatero, C.; Jirousek, M. R.; Fesik, S. W. J. Am.
Chem. Soc. 2003, 125, 4087–4096.
13. The fluoro group in 1 has no effect on potency, see Ref. 11.
All inhibitors give satisfactory 1H NMR and high reso-
lution mass spectral data.
14. Piaz, V. D.; Pinzauti, S.; Lacrimini, P. Synthesis 1975,
664–665.
15. Lehr, F.; Gonnermann, J.; Seebach, D. Helv. Chim. Acta
1979, 62, 2258–2275.
16. Wierenga, W.; Skulnick, H. I. J. Org. Chem. 1979, 44,
310–311.
17. Jung, M. E.; Light, L. A. Tetrahedron Lett. 1982, 23,
3851–3854.
18. Kaszubska, W.; Falls, H. D.; Schaefer, V. G.; Haasch, D.;
Frost, L.; Hessler, P.; Kroeger, P. E.; White, D. W.;
Jirousek, M. R.; Trevillyan, J. M. J. Mol. Cell Endocrinol.
2002, 195, 109–118.
19. Zabolotny, J. M.; Bence-Hanulec, K. K.; Stricker-Kron-
grad, A.; Haj, F.; Wang, Y.; Minokoshi, Y.; Kim, Y. B.;
Elmquist, J. K.; Tartaglia, L. A.; Kahn, B. B.; Neel, B. G.
Dev. Cell 2002, 2, 489–495.
5. Barford, D.; Flint, A. J.; Tonks, N. K. Science 1994, 263,
1397–1404.
6. Sarmiento, M.; Puius, Y. A.; Vetter, S. W.; Keng, Y. F.;
Wu, L.; Zhao, Y.; Lawrence, D. S.; Almo, S. C.; Zhang, Z.
Biochemistry 2000, 39, 8171–8179.
7. (a) Liu, G.; Trevillyan, J. M. Curr. Opin. Investig. Drugs
2002, 3, 1608–1616; (b) Johnson, T. O.; Ermolieff, J.;
Jirousek, M. R. Nature Rev. Drug Discov. 2002, 1, 696–709.
8. (a) Liljebris, C.; Larsen, S. D.; Ogg, D.; Palazuk, B. J.;
Bleasdale, J. E. J. Med. Chem. 2002, 45, 1785–1798; (b)
Larsen, S. D.; Stevens, F. C.; Lindberg, T. J.; Bodnar, P.
M.; OÕSullivan, T. J.; Schostarez, H. J.; Palazuk, B. J.;
Bleasdale, J. E. Bioorg. Med. Chem. Lett. 2003, 13, 971–
975; (c) Xie, L.; Lee, S. Y.; Andersen, J. N.; Waters, S.;
Shen, K.; Guo, X.; Moller, N. P. H.; Olefsky, J. M.;
Lawrence, D. S.; Zhang, Z. Biochemistry 2003, 42, 12792–
12804.
9. Ibarra-Sanchez, M. D. J.; Simoncic, P. D.; Nestel, F. R.;
Duplay, P.; Lapp, W. S.; Trembley, M. L. Seminars
Immunol. 2000, 12, 379–386.
10. For references on the discovery of a difluorophosphonic
acid-based, highly potent PTP1B inhibitor (Ki = 2.4nM)
with 10-fold TCPTP selectivity, see: (a) Shen, K.; Keng,
Y.; Wu, L.; Guo, X.; Lawrence, D. S.; Zhang, Z.-Y.
J. Biol. Chem. 2001, 276, 47311–47319; (b) Sun, J.-P.;
Fedorov, A. A.; Lee, S.-Y.; Guo, X.-L.; Shen, K.;
Lawrence, D. S.; Almo, S. C.; Zhang, Z.-Y. J. Biol.
Chem. 2003, 278, 12406–12414.
20. Posner, B. I.; Faure, R.; Burgess, J. W.; Bevan, A. P.;
Lachance, D.; Zhang-Sun, G.; Fantus, I. G.; Ng, J. B.;
Hall, D. A.; Lum, B. S. J. Biol. Chem. 1994, 269, 4596–
4604.
21. The refined crystallographic coordinates have been depos-
ited in the Protein Data Bank (www.rcsb.org) with entry
code 1XBO.
22. Puius, Y. A.; Zhao, Y.; Sullivan, M.; Lawrence, D. S.;
Almo, S. C.; Zhang, Z.-Y. Proc. Natl. Acad. Sci. U.S.A.
1997, 94, 13420–13425.
11. Liu, G.; Xin, Z.; Pei, Z.; Hajduk, P. J.; Abad-Zapatero,
C.; Hutchins, C. W.; Zhao, H.; Lubben, T. H.; Ballaron,