Discovery of a potent, selective protein tyrosine phosphatase 1B inhibitor using a linked-fragment strategy

Article abstract of DOI:10.1021/ja0296733

Protein tyrosine phosphatase 1B (PTP1B) is an enzyme that downregulates the insulin receptor. Inhibition of PTP1B is expected to improve insulin action, and the design of small molecule PTP1B inhibitors to treat type II diabetes has received considerable attention. In this work, NMR-based screening identified a nonselective competitive inhibitor of PTP1B. A second site ligand was also identified by NMR-based screening and then linked to the catalytic site ligand by rational design. X-ray data confirmed that the inhibitor bound with the catalytic site in the native, "open" conformation. The final compound displayed excellent potency and good selectivity over many other phosphatases. The modular approach to drug design described in this work should be applicable for the design of potent and selective inhibitors of other therapeutically relevant protein tyrosine phosphatases.

Full text of DOI:10.1021/ja0296733

Published on Web 03/12/2003
Discovery of a Potent, Selective Protein Tyrosine
Phosphatase 1B Inhibitor Using a Linked-Fragment Strategy
Bruce G. Szczepankiewicz,* Gang Liu,† Philip J. Hajduk,‡ Cele Abad-Zapatero,‡
Zhonghua Pei,† Zhili Xin,† Thomas H. Lubben,§ James M. Trevillyan,§
Michael A. Stashko,§ Stephen J. Ballaron,§ Heng Liang,‡ Flora Huang,‡
Charles W. Hutchins,‡ Stephen W. Fesik,| and Michael R. Jirousek
Contribution from the Global Pharmaceutical Research and DeVelopment Organization,
Abbott Laboratories, 100 Abbott Park Road, Abbott Park, Illinois 60064
Received December 10, 2002; E-mail: bruce.szczepankiewicz@abbott.com
Abstract: Protein tyrosine phosphatase 1B (PTP1B) is an enzyme that downregulates the insulin receptor.
Inhibition of PTP1B is expected to improve insulin action, and the design of small molecule PTP1B inhibitors
to treat type II diabetes has received considerable attention. In this work, NMR-based screening identified
a nonselective competitive inhibitor of PTP1B. A second site ligand was also identified by NMR-based
screening and then linked to the catalytic site ligand by rational design. X-ray data confirmed that the
inhibitor bound with the catalytic site in the native, “open” conformation. The final compound displayed
excellent potency and good selectivity over many other phosphatases. The modular approach to drug design
described in this work should be applicable for the design of potent and selective inhibitors of other
therapeutically relevant protein tyrosine phosphatases.
Introduction
oral hypoglycemic agents such as metformin, sulfonlyureas, the
thiazolidinediones, and R-glucosidase inhibitors.9 However, no
Reversible protein phosphorylation is the predominant strat-
egy used to control the activity of proteins in eucaryotic cells.
Approximately 30% of the 10 000 proteins in a typical mam-
malian cell are thought to be phosphorylated.1,2 Many cellular
functions could be artificially manipulated if one could exog-
enously control the activity of protein kinases and phosphatases.3
This has led to intense interest in identifying small molecules
capable of inhibiting the action of specific kinases4,5 or
phosphatases.6,7 Such agents offer a great deal of promise as
new therapies for a wide variety of human diseases including
cancer, inflammation, and diabetes.
combination of these therapies is completely successful in
ameliorating type II diabetes for many patients, and more
efficacious agents are needed. A major goal of new therapies
for type II diabetes is to potentiate the action of insulin. One of
the key proteins involved in insulin signaling is the insulin
receptor. When insulin binds to its receptor, changes in the
intracellular conformation of the receptor result in the O-
phosphorylation of three specific tyrosine residues.10 This serves
as the first step in insulin signaling, and it is followed by a
cascade of intracellular events that mediate the physiological
effects of insulin.11 There is compelling evidence that protein
tyrosine phosphatase 1B, or PTP1B, is primarily responsible
for the dephosphorylation of the insulin receptor and, therefore,
acts to downregulate insulin signaling.12-16 A PTP1B inhibitor
would be expected to increase the half-life of the phosphorylated
insulin receptor and enhance the effects of insulin. The most
intriguing data to support this hypothesis come from the
knockout mouse studies reported by two different laborato-
Type II diabetes afflicts over 130 million people worldwide,
and the incidence is expected to grow steadily over the next
several years.8 Current treatments include injectable insulin and
* Address correspondence to this author at R4MC AP10/LL12, Metabolic
Disease Research, Abbott Laboratories, 100 Abbott Park Road, Abbott Park,
IL 60064-6098. Phone: (847) 935-1559. Fax: (847) 938-1674.
† R4MC Metabolic Disease Research.
‡ R46Y Advanced Technologies.
§ R47R Metabolic Disease Research.
| R460 Cancer Research.
(9) Nuss, J. M.; Wagman, A. S. In Annual Reports in Medicinal Chemistry;
Doherty, A. M., Ed.; Academic Press: San Diego, CA, 2000; Vol. 35, pp
211-220.
Current address: Pfizer Global Research and Development, La Jolla
Laboratories, 10770 Science Center Drive, San Diego, CA 92121.
(1) Alberts, B.; Bray, D.; Lewis, J.; Raff, M.; Roberts, K.; Watson, J. Molecular
Biology of the Cell, 3rd ed.; Garland: New York, 1994; p 202.
(2) Cohen, P. Trends Biochem. Sci. 2000, 25, 596.
(10) Kahn, C. R. Diabetes 1994, 43, 1066.
(11) Czech, M. P.; Corvera, S. J. Biol. Chem. 1999, 274, 1865.
(12) Kenner, K. A.; Anyanwu, E.; Olefsky, J. M.; Kusari, J. J. Biol. Chem.
1996, 271, 19810.
(3) Ahn, N. Chem. ReV. 2001, 101, 2207.
(4) Toledo, L. M.; Lydon, N. B.; Elbaum, D. Curr. Med. Chem. 1999, 6, 775.
(5) Rewcastle, G. W.; Denny, W. A.; Showalter, H. D. H. Curr. Org. Chem.
2000, 4, 679.
(13) Chen, H.; Wertheimer, S. J.; Lin, C. H.; Katz, S. L.; Amrein, K. E.; Burn,
P.; Quon, M. J J. Biol. Chem. 1997, 272, 8026.
(14) Ahmad, F.; Li, P. M.; Meyerovitch, J.; Goldstein, B. J. J. Biol. Chem. 1995,
270, 20503.
(6) Serine threonine phosphatase inhibitors: McCluskey, A.; Sim, A. T. R.;
Sakoff, J. A. J. Med. Chem. 2002, 45, 1151.
(15) Chen, H.; Cong, L. N.; Li, Y.; Yao, Z. J.; Wu, L.; Zhang, Z. Y.; Burke, T.
R., Jr.; Quon, M. J. Biochemistry 1999, 38, 384.
(7) Protein tyrosine phosphatases: Zhang, Z.-Y. Annu. ReV. Pharmacol. Toxicol.
2002, 42, 209.
(16) Walchli, S.; Curchod, M. L.; Gobert, R. P.; Arkinstall, S.; Hooft van
Huijsduijnen, R. J. Biol. Chem. 2000, 275, 9792.
(8) Zimmet, P., Alberti, K. G. M. N.; Shaw, J. Nature 2001, 414, 782.
9
10.1021/ja0296733 CCC: $25.00 © 2003 American Chemical Society
J. AM. CHEM. SOC. 2003, 125, 4087-4096
4087

A R T I C L E S
Szczepankiewicz et al.
Figure 2. Structure of diaryloxamic acid 1.
Scheme 1 a
Figure 1. 13C/1H HSQC NMR spectra of δ-13CH3 Ile-labeled PTP1B in
the absence (black) and presence (red) of 0.1 mM diaryloxamic acid 1.
Significant chemical shift perturbations in the δ-methyl resonance of Ile219
were observed upon addition of ligand.
ries.17,18 PTP1B deficient mice were viable, healthy, and lean.
They displayed enhanced insulin sensitivity and resistance to
diet-induced obesity. This provides important evidence that
PTP1B inhibition would be an effective diabetes therapy. The
development of PTP1B inhibitors began in the early 1990s and
continues today.19-22
In an effort to develop a small, potent, and selective PTP1B
inhibitor, we used a combination of NMR-based screening and
iterative structure-based drug design to identify and optimize a
lead compound. This compound is a potent, competitive inhibitor
of PTP1B with a novel binding mode that selectively inhibits
PTP1B versus many other phosphatases.
a Reagents and conditions: (a) Cu(OAc)2, i-PrOH, reflux (43%); (b) (i)
EtOCOCOCl, Et3N, DMF, (ii) aqueous NaOH (16%).
PTP1B (Figure 2). This compound caused chemical shift
changes similar to those for phosphotyrosine (pTyr). Significant
perturbations in the 15N amide resonances were observed for
Val49, Gly220, and Gly 218 and also for the δ-13CH3 resonance
of Ile219. In addition, the Kd value of compound 1 as measured
by NMR (Kd ) 100 µM) compared favorably with the activity
of this compound in a PTP1B-mediated para-nitrophenyl
phosphate (pNPP) hydrolysis assay (Ki ) 293 ( 174 µM). These
data suggested that diaryloxamic acid 1 was binding to the active
site of PTP1B and could serve as a pTyr mimetic for lead
development. Two other groups have reported monoaryloxam-
ates as phosphotyrosine surrogates.25,26
Improving the Catalytic Site Ligand. Reasoning that
occupation of more space within the active site would lead to
a more potent inhibitor, we prepared naphthyloxamic acid 5
via a three-step sequence (Scheme 1). NMR titration data yielded
a Kd of 26 µM for naphthyloxamic acid 5, which correlated
well with the Ki value of 39 ( 14 µM (n ) 3) observed in the
pNPP hydrolysis assay. The inhibition kinetics were consistent
with competitive and reversible inhibition,27 indicating that
naphthyloxamic acid 5 did not form any covalent bonds to
PTP1B and that it did not permanently alter PTP1B in any other
way.
Results and Discussion
Discovery of a Catalytic Site Ligand. To identify novel
scaffolds that could be used in the design of potent PTP1B
inhibitors, NMR-based screening of a 10,000 compound library
was employed. Screening was accomplished using a truncated,
292 amino acid version of PTP1B either uniformly labeled with
15N or selectively labeled with δ-13CH3 Ile. Ligand binding was
1
monitored by observing chemical shift changes in H/15N or
1H/13C-HSQC in the presence of test compounds. 23,24 As shown
in Figure 1, high-resolution HSQC spectra could be obtained
on the protein, and ligand binding could be detected readily.
The NMR screen identified diaryloxamic acid 1 as a ligand for
(17) 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.
(18) Klaman, L. D.; Boss, O.; Peroni, O. D.; Kim, J. K.; Martino, J. L.; Zabotny,
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.
(19) Ripka, W. C. In Annual Reports in Medicinal Chemistry; Doherty, A. M.,
Ed.; Academic Press: San Diego, CA, 2000; Vol. 35, pp 231-250.
(20) Johnson, T. O.; Ermolieff, J.; Jirousek, M. Nat. ReV. Drug DiscoVery 2002,
1, 696.
An X-ray structure of compound 5 bound to PTP1B was
obtained which clearly showed how the inhibitor binds to the
(25) (a) Iversen, L. F.; Andersen, H. S.; Møller, K. B.; Olsen, O. H.; Peters, G.
H.; Branner, S.; Mortensen, S. B.; Hansen, T. K.; Lau, J.; Ge, Y.; Holsworth,
D. D.; Newman, M. J.; Møller, N. P. H. Biochemistry 2001, 40, 14812. (b)
Peters, G. H.; Iversen, L. F.; Branner, S.; Andersen, H. S.; Mortensen, S.
B.; Olsen, O. H.; Møller, K. B.; Møller, N. P. H. J. Biol. Chem. 2000, 275,
18201. (c) Andersen, H. S.; Iversen, L. F.; Jeppesen, C. B.; Branner, S.;
Norris, K.; Rasmussen, H. B.; Møller, K. B.; Møller, N. P. H. J. Biol.
Chem. 2000, 275, 7101.
(26) Beaulieu, P. L.; Cameron, D. R.; Ferland, J.-M.; Gauthier, J.; Ghiro, E.;
Gillard, J.; Gorys, V.; Poirier, M.; Rancourt, J.; Wernic, D.; Llinas-Brunet,
M. J. Med. Chem. 1999, 42, 1757.
(27) Cornish-Bowden, A. Fundamentals of Enzyme Kinetics, revised ed.; Portland
Press, Ltd.: London, 1995; Chapter 5.
(21) Blaskovich, M. A.; Kim, H.-O. Expert Opin. Ther. Pat. 2002, 12, 871.
(22) Shen, K.; Keng, Y.-F.; Wu, L.; Guo, X.-L.; Lawrence, D. S.; Zhang, Z.-Y.
J. Biol. Chem. 2001, 276, 47311. Compound 40 in this work gave a Ki of
2.4 ( 0.2 nM for PTP1B.
(23) Shuker, S. B.; Hajduk, P. J.; Meadows, R. P.; Fesik, S. W. Science 1996,
274, 1531.
(24) Hajduk, P. J.; Augeri, D. J.; Mack, J.; Mendoza, R.; Yang, J.; Betz, S. F.;
Fesik, S. W. J. Am. Chem. Soc. 2000, 122, 7898.
9
4088 J. AM. CHEM. SOC. VOL. 125, NO. 14, 2003

Diaryloxamic Acid-Based PTP1B Inhibitors
A R T I C L E S
Figure 3. X-ray structures of PTP1B complexed with inhibitor 5 (a), inhibitor 12 (b), and inhibitor 23 (c). Protein color scheme: orange ) carbon; red )
oxygen; blue ) nitrogen; yellow ) sulfur. Inhibitor carbon atoms are shown in green.
active site of PTP1B (Figure 3a). Naphthyloxamic acid 5 binds
to the “open” form of the enzyme, where the WPD loop
(Trp179-Ser187) retains a conformation similar to the unoc-
cupied enzyme.28 This is different from that observed for most
other inhibitors, where the WPD loop closes down toward the
active site cysteine (Cys215).29 As a result of binding to the
open conformation, the naphthyloxamates have access to a much
larger active site. The benzoic acid carboxylate group interacts
with Arg221, and the benzene ring makes hydrophobic contacts
with the side chain of Gln262. These critical interactions
stabilize the open form of the enzyme and preclude the WPD
loop from closing. The amide carbonyl oxygen and the oxanilic
acid anion are positioned to form hydrogen bonds to a ring of
backbone amides defined by the residues Ser216-Gly220.
Another hydrophobic interaction between the naphthalene
portion and Tyr46 also contributes to binding.
particular, the phosphate and oxamate groups of the two
compounds are similarly positioned, and the naphthalene system
occupies the same region of the binding site as the benzene
ring of phosphotyrosine30 (Figure 4). The overlaid X-ray
structures indicated that a diamide bearing chain might extend
out of the active site from the 5-naphthyl position. Initial
attempts to substitute the 5- or 6-naphthyl position resulted in
the loss of inhibitory activity (data not shown). However,
incorporation of a diamido chain at the 4-position gave
compound 12 (Scheme 2) and resulted in nearly a 40-fold boost
in potency, with a Ki value of 1.1 ( 0.5 µM. The active site
contacts observed with compound 5 were preserved with
inhibitor 12, but the X-ray structure showed that the naphthyl
ring system had flipped 180°. This ring flip occurs to accom-
modate two interactions between inhibitor 12 and PTP1B that
are not possible with compound 5 (Figure 3b). The diamide
chain extends out of the active site and reaches Asp48, where
it forms two hydrogen bonds to the side chain carboxylate. The
pentyl chain continues past Asp48 and is pointed directly toward
Installation of a Catalytic Site-Site 2 Linker. Despite its
unusual binding mode, naphthyloxamic acid 5 showed some
similarities to phosphotyrosine when bound to PTP1B. In
(28) Barford, D.; Flint, A. J.; Tonks, N. K. Science 1994, 263, 1397.
(29) Groves, M. R.; Yao, Z.-J.; Roller, P. P.; Burke, T. R., Jr.; Barford, D.
Biochemistry 1998, 37, 17773.
(30) Salmeen, A.; Andersen, J. N.; Myers, M. P.; Tonks, N. K.; Barford, D.
Mol. Cell 2000 6, 1401.
9
J. AM. CHEM. SOC. VOL. 125, NO. 14, 2003 4089

A R T I C L E S
Szczepankiewicz et al.
Figure 5. Selected PTP1B site 2 ligands identified using NMR screening.
represents a region that could impart enzyme selectivity.31 Thus,
the next stage in the design strategy was to utilize the diamide
bridge to position groups in the noncatalytic site that would
not only improve the potency but also impart greater PTP1B
specificity to the series.
NMR screening was again utilized to identify a ligand for
site 2. Binding to the noncatalytic site was specifically monitored
by labeling the protein with 13C-methionine and following the
resonance of Met258, an amino acid present in site 2 (see Figure
3b).24 Screening 10,000 compounds identified several capable
of binding weakly (Kd > 1 mM) to this site. Many of these
ligands were small fused-ring aromatic acids, such as 2-ben-
zofurancarboxylic acid (13), 2-benzothiophenecarboxylic acid
(14), or 2-quinolinecarboxylic acid (15) (Figure 5). Having two
ligands that bound to proximal sites on the protein along with
a suitable linker provided a straightforward application of the
linked-fragment SAR-by-NMR approach.23 The nature of the
site 2 ligands identified by screening implied that a naphthoic
acid might also be a suitable site 2 ligand. Since 3-hydroxy-2-
naphthoic acid methyl ester (18) could easily be linked to
compound 12, inhibitor 23 was prepared according to the route
depicted in Scheme 3. Whereas compound 12 was racemic,
inhibitor 23 was prepared with complete stereocontrol starting
from N-Boc-(S)-3-iodoalanine methyl ester.
Figure 4. Overlay of the PTP1B (yellow ribbon)-naphthyloxamic acid 5
(green) complex with catalytically inactive PTP1B (Cys215Ser, cyan ribbon)
bound to a phosphotyrosine residue (magenta). The WPD loop in the
catalytically inactive protein (“closed” position) is shown in magenta. Asp48
and Tyr46 are highlighted in orange.
Scheme 2 a
The Ki for inhibitor 23 was 22 ( 6 nM (n ) 3). The second
site ligand therefore gave a 25-fold improvement in inhibitory
activity, with a concurrent 2-fold improvement from the
asymmetric catalytic site ligand. The conformations of the
naphthyloxamate and diamide linker regions of inhibitor 23 were
essentially identical to those of compound 12, and the naphthoic
acid was positioned in site 2 (Figure 3c). The naphthoic acid
carboxylate group was positioned 2.6 Å from Arg254 and 2.8
Å from Arg24. Either or perhaps both Arg residues could
therefore be participating in a salt bridge or H-bonding interac-
tion with the naphthoic acid. The naphthalene system forms a
weak (4.1 Å) hydrophobic interaction with Met258.
Selectivity Against Other Phosphatases. To determine the
selectivity of the new compounds, we measured their inhibitory
activity against a panel of several phosphatases including
CD45,32 LAR,33 SHP-2,34 calcineurin,35 and TCPTP.36 On the
basis of amino acid sequence homology, TCPTP is the phos-
phatase most closely related to PTP1B. Selectivity for PTP1B
over TCPTP inhibition would therefore represent the highest
level of specificity for the desired phosphatase over another.
The Ki values of inhibitors 1, 5, 12, and 23 against the
phosphatases tested are shown in Table 1, and selectivity ratios
are given in Table 2. Inhibitor 5 showed similar activity against
PTP1B (Ki ) 39 ( 14 µM), TCPTP (Ki ) 44 ( 8 µM), and
LAR (Ki ) 34.8 ( 1.6 µM). This compound showed little or
a Reagents and conditions: (a) methyl 2-acetamidoacrylate, Pd(OAc)2,
P(o-tolyl)3, Et3N, DMF, 110 °C (76%); (b) 10% Pd/C, i-PrOH, 4 atm of
H2 (100%); (c) 3 N aqueous NaOH, MeOH, 25 °C (74%); (d) TBTU,
HOBT, amylamine, Et3N, DMF, (88%); (e) 2 (see Scheme 1), Cu(OAc)2,
i-PrOH, 80 °C (83%), (f) ClCOCO2Et, i-Pr2NEt, CH2Cl2, 0 °C f 25 °C;
(g) NaOH, EtOH, 25 °C (29%).
a second, noncatalytic phosphotyrosine binding site (site 2).31
Thus, the diamide linkage not only imparted additional binding
energy but also served as a bridge between the catalytic and
noncatalytic binding sites.
Discovery and Incorporation of a Site 2 Ligand. Accessing
the noncatalytic phosphotyrosine binding region was an impor-
tant goal. The catalytic domain is generally conserved among
many tyrosine phosphatases, so achieving selectivity for PTP1B
over other phosphatases was likely to be difficult without also
accessing a region outside the active site. As first reported by
Zhang and Lawrence, the second, noncatalytic site is much less
homologous among different tyrosine phosphatases and so
(32) Penninger, J. M.; Irie-Sasaki, J.; Sasaki, T.; Oliveira-dos-Santos, A. J. Nat.
Immunol. 2001, 2, 389.
(33) Cheng, A.; Dube, N.; Gu, F.; Trembley, M. L. Eur. J. Biochem. 2002,
269, 1050.
(34) Qu, C.-K. Biochim. Biophys. Acta 2002, 1592, 297.
(35) Rusnak, F.; Mertz, P. Physiol. ReV. 2000, 80, 1483.
(36) Ibarra-Sanchez, M.d. J.; Simoncic, P. D.; Nestel, F. R.; Duplay, P.; Lapp,
W. S.; Trembley, M. L. Semin. Immunol. 2000, 12, 379.
(31) 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.
9
4090 J. AM. CHEM. SOC. VOL. 125, NO. 14, 2003

Diaryloxamic Acid-Based PTP1B Inhibitors
A R T I C L E S
Scheme 3 a
a Reagents and conditions: (a) (i) N-Boc-(S)-3-iodoalanine, Zn, DMF, (ii) Pd(OAc)2, o-tol3P, 6 (55%); (b) NaOH, aqueous CH3OH (83%); (c) N-Boc-
5-amino-1-pentanol, PPh3, DEAD, THF (56%); (d) 4 M HCl, dioxane (95%); (e) EDCI, HOBT, NMM, CH2Cl2 (90%); (f) DPIC (2), Cu(OAc)2, i-PrOH
(100%); (g) (i) EtOCOCOCl, Et3N, DMF, (ii) aqueous NaOH; (h) (i) CF3COOH, (ii) aqueous NaOH, (iii) Ac2O (28%).
Table 1. Ki of Inhibitors 1, 5, 12, and 23 against Various Phosphatasesa
compd
phosphatase
1
5
12
23
PTP1B
TCPTP
LAR
SHP-2
CD45
293 ( 175 (3)
164 ( 36 (2)
>900 (2)
39 ( 14 (3)
44 ( 8 (2)
34.8 ( 1.6 (2)
>300 (2)
1.1 ( 0.5 (6)
1.1 ( 0.9 (3)
6.7 ( 4.7 (4)
29.6 ( 3.6 (2)
417 ( 271 (7)
>300 (2)
0.022 ( 0.006 (3)
0.049 ( 0.006 (3)
1.3 ( 0.5 (2)
2.49 ( 0.19 (2)
53.6 ( 5.7 (2)
>300 (2)
>900 (2)
>360 (2)
176 (1)
calcineurin
>300 (2)
a Ki values are in units of µM ( standard deviation, with the number of determinations in parentheses.
Table 2. Selectivity Ratios for Inhibitors 1, 5, 12, and 23 over
( 6 nM) over TCPTP (Ki ) 49 ( 6 nM). Inhibitory activity
against LAR (Ki ) 1.3 ( 0.5 µM), SHP-2 (Ki ) 2.49 ( 0.19
µM), CD45 (Ki ) 53.6 ( 5.7 µM), and calcineurin (Ki > 300
µM) demonstrated excellent selectivity for PTP1B.
Various Phosphatases
compd
12
phosphatase
1
5
23
The inhibitory data for compound 23 vindicated the hypoth-
esis that PTP1B selectivity against closely related phosphatases
was possible by taking advantage of the site 2 interactions. We
did not have access to the crystal structure of TCPTP,37 but a
homology model indicated that TCPTP also bore a second
phosphotyrosine binding site similar to the PTP1B second site.
In particular, Arg254 and Arg24 are conserved between PTP1B
and TCPTP (see Figure 3c). However, the exact shape of the
second site is expected to vary because of two point mutations,
Cys32 in PTP1B to His32 in TCPTP and Phe52 in PTP1B to
Tyr52 in TCPTP. We believe that the differences in the second
site alter the interaction of the naphthoic acid with the two
arginine residues, affording some selectivity for PTP1B over
PTP1B
TCPTP
LAR
SHP-2
CD45
1
1
>3
>3
>1
1
1
1
>7
>4
>7
1
1
6
27
380
1
2
36
104
2700
>13000
calcineurin
>270
no activity against SHP-2 (Ki > 300 µM), CD45 (Ki > 176
µM), and calcineurin (Ki > 300 µM). Inhibitor 12 showed
similar activity against PTP1B (Ki ) 1.1 ( 0.5 µM) and TCPTP
(Ki ) 1.1 ( 0.9 µM) but modest selectivity over LAR (Ki )
6.7 ( 4.7 µM). Selectivity for inhibitor 12 was good over SHP-2
(Ki ) 29.6 ( 3.6 µM), CD45 (Ki ) 417 ( 271 µM), and
calcineurin (Ki > 300 µM). As we had hoped, inhibitor 23
proved to be the most PTP1B selective inhibitor. This compound
showed modest but significant selectivity for PTP1B (Ki ) 22
(37) Iversen, L. F.; Møller, K. B.; Pedersen, A. K.; Peters, G. H.; Petersen, A.
S.; Andersen, H. S.; Branner, S.; Mortensen, S. B.; Møller, N. P. H. J.
Biol. Chem. 2002, 277, 19982.
9
J. AM. CHEM. SOC. VOL. 125, NO. 14, 2003 4091

A R T I C L E S
Szczepankiewicz et al.
25 mL) and brine (1 × 25 mL), dried over MgSO4, filtered, and
concentrated to a solid. This was recrystallized from 40 mL of ethyl
acetate to give 1.3 g (54%) of 2-((2,3-dimethylphenyl)amino)benzoic
acid39 as colorless crystals: 1H NMR (400 MHz, DMSO-d6) δ 12.95
(bs, 1H), 9.44 (bs, 1H), 7.89 (d, 1H, J ) 7.1 Hz), 7.30 (t, 1H, J ) 7.2
Hz), 7.11 (m, 2H), 7.03 (t, 1H, J ) 4.5 Hz), 6.69 (m, 2H), 2.29 (s,
3H), 2.10 (s, 3H); 13C NMR (100 MHz, DMSO-d6) δ 170.1 (C), 148.7
(C), 138.3 (C), 137.8 (C), 134.1 (CH), 131.7 (CH), 131.2 (C), 126.3
(CH), 125.9 (CH), 122.1 (CH), 116.2 (CH), 113.0 (CH), 111.2 (C),
20.1 (CH3), 13.6 (CH3); MS (ESI) m/z 242 ([M + H]+).
TCPTP. We also observed a modest improvement in inhibition
of LAR and CD45 upon installation of the site 2 ligand on an
asymmetric core, while we saw a more substantial improvement
in activity against SHP-2. These data indicate that the naphthoic
acid may make some binding contacts with each of these
phosphatases, albeit to widely varying degrees. Modification
of the second site ligand, or alternate second site ligands, could
result in still better selectivity for PTP1B.
Conclusion
To an ice cooled solution of 2-((2,3-dimethylphenyl)amino)benzoic
acid (343 mg, 1.55 mmol) and triethylamine (503 µL, 3.56 mmol) in
dichloromethane (5 mL) was added ethyl oxalyl chloride (401 µL, 3.56
mmol) over 30 min. The reaction was warmed to ambient temperature
and stirred for 16 h. The reaction mixture was then treated with 1 M
HCl (10 mL) and extracted with dichloromethane (2 × 15 mL). The
combined extracts were washed with brine, dried over Na2SO4, filtered,
and concentrated in vacuo. The residue was dissolved in 5 mL of
methanol, and 4.5 mL (4.5 mmol) of 1 M NaOH(aq) was added. After
being stirred for 2 h at ambient temperature, the basic reaction mixture
was diluted with 1 M HCl (10 mL). The product was isolated via reverse
phase HPLC to provide 235 mg (48%) of 2-[(2,3-dimethylphenyl)-
oxalylamino]benzoic acid (1) as a light brown solid: 1H NMR (300
MHz, DMSO-d6) mixture of rotamers δ 8.21 (dd, 1H [minor], J )
8.1, 1.0), 8.17 (d, 0.5H, J ) 6.6 Hz), 8.05 (dd, 0.5, J ) 7.7, 1.5 Hz),
7.94 (dd, 1H [minor], J ) 7.7, 1.8 Hz), 7.81-7.22 (m, 5.5H), 7.07 (d,
0.5H, J ) 7.3 Hz), 6.90 (d, 1H [minor], J ) 8 Hz), 2.54 (s, 1.5H),
2.51 (s, 3H [minor]), 2.49 (s, 1.5H), 2.45 (s, 1.5H), 2.42 (s, 1.5H),
2.33 (s, 3H [minor]), 2.29 (s, 3H [minor]); 13C NMR (100 MHz, DMSO-
d6) δ 167.1 (C), 166.5 (C), 163.3 (C), 162.0 (C), 140.5 (C), 140.2 (C),
139.2 (C), 138.6 (C), 138.4 (C), 137.7 (C), 135.2 (C), 134.0 (C), 133.2
(CH), 132.4 (C), 132.2 (CH), 131.3 (CH), 130.2 (CH), 130.1 (CH),
130.0 (CH), 129.1 (CH), 129.0 (C), 128.9 (CH), 128.1 (C), 127.7 (CH),
127.5 (CH), 126.5 (CH), 126.1 (CH), 125.5 (CH), 125.1 (CH), 20.1
(CH3), 20.0 (CH3), 14.5 (CH3); MS (ESI) m/z 314 ([M + H]+), 331
([M + NH4]+), 336 ([M + Na]+).
Starting with low affinity leads from NMR-based screening,
the inhibitor series was incrementally constructed by fully
utilizing the structural and NMR screening data to rationally
improve both potency and selectivity. Compound 23 is among
the most potent nonpeptidic PTP1B inhibitors reported to date,
and it displays excellent selectivity against LAR, SHP-2, CD-
45, and calcineurin. The modest selectivity against TCPTP is
noteworthy given the extremely high sequence identity between
these two phosphatases. It is also interesting that while the site
2 ligand endowed compound 23 with improved selectivity for
PTP1B, it also improved inhibitory activity to a lesser extent
against TCPTP and SHP-2. Since these phosphatases appear to
accommodate a second site ligand, it is possible that selectivity
for these and other tyrosine phosphatases can be achieved by
choosing an appropriate second site ligand for each. The NMR
screening method we employed to discover a PTP1B selective
site 2 ligand could serve to find selective, noncatalytic site
ligands for other tyrosine phosphatases as well. Such a strategy
could greatly reduce the discovery time for identifying specific
tyrosine phosphatase inhibitors. These inhibitors would be useful
tools to probe the role of phosphatases in different cellular
processes and may ultimately lead to new therapeutic agents
for many human diseases.
2-(Naphthalen-1-ylamino)benzoic Acid (4). To 3.42 g (10.0 mmol)
of diphenyliodonium-2-carboxylate monohydrate (2), 1.43 g (10.0
mmol) of 1-aminonaphthalene (3), and 73 mg (0.40 mmol) copper(II)
acetate was added 40 mL of 2-propanol. The mixture was stirred at
reflux under N2 for 2.5 h and then concentrated in vacuo. The solid
residue was taken up in 100 mL of ethyl acetate and extracted with
water (1 × 25 mL) and brine (1 × 25 mL), dried over MgSO4, filtered,
and concentrated to a solid. This was recrystallized from 20 mL of
ethyl acetate to give 1.12 g (43%) of a gray solid: 1H NMR (300 MHz,
DMSO-d6) δ 10.05 (bs, 1H), 7.98 (m, 3H), 7.78 (m, 1H), 7.55 (m,
4H), 7.33 (td, 1H, J ) 7.3, 1.0 Hz), 6.92 (d, 1H, J ) 8.5 Hz), 6.77 (t,
1H, J ) 7.5 Hz); 13C NMR (100 MHz, DMSO-d6) δ 148.7 (C), 136.2
(C), 134.4 (C), 134.2 (CH), 131.9 (CH), 128.6 (C), 128.4 (CH), 126.3
(CH), 126.1 (CH), 124.5 (CH), 121.8 (CH), 119.7 (CH), 117.1 (CH),
113.8 (CH); MS (ESI) m/z 264 ([M + H]+).
2-(Naphthalen-1-yloxalylamino)benzoic Acid (5). To an ice-cooled
solution of 100 mg (0.380 mmol) of 2-(naphthalen-1-ylamino)benzoic
acid (4) in 2 mL of DMF was added 400 µL (2.87 mmol) of
triethylamine and then 200 µL (1.79 mmol) of ethyl chlorooxoacetate.
The ice bath was removed, and the reaction was stirred at ambient
temperature. After 2 h, 3 mL of 2 M NaOH(aq) was added, the solution
was diluted with 10 mL water, and the mixture was allowed to stir for
2 h. Another 10 mL of water was added to clear the solution, and then
the reaction was stirred for an additional 2 h. The mixture was treated
with 3 mL of 6 M HCl and extracted with ethyl acetate (3 × 5 mL).
The combined ethyl acetate layers were back-extracted with water (2
× 5 mL) and brine (1 × 5 mL), dried over MgSO4, filtered, and
Experimental Section
Chemistry. General Methods. Reactions were dried over MgSO4
or Na2SO4, filtered through a fritted glass funnel or a plug of cotton,
and concentrated with a rotary evaporator at ca. 15 mmHg, warming
when necessary. Thin-layer chromatography systems were the same
as those used for column chromatography, with Rf approximately 0.3.
Thin-layer chromatography was performed using Merck F-254 silica
gel plates. Column chromatography was performed under 5 psi N2 using
230-400 mesh silica gel. Analytical HPLC traces and compounds
purified by HPLC were obtained by elution through a C18 reverse-
phase column with a gradient of 5 mM aqueous ammonium acetate/
acetonitrile or 0.1% aqueous TFA/acetonitrile as eluants. Concentration
gradients started at 0% acetonitrile and 100% aqueous medium. Starting
materials were obtained from commercial suppliers and used without
further purification. 1H NMR data are tabulated in the following
order: multiplicity (s, singlet; d, doublet, t, triplet; m, multiplet, b,
broad), number of protons, coupling constants in hertz. Where mixtures
1
of stereoisomers or rotamers were obtained, the H NMR resonances
of the major species are reported, unless otherwise indicated.
2-[(2,3-Dimethylphenyl)oxalylamino]benzoic acid (1). (An alter-
nate route to oxamic acid 2 has been described previously.38) To 1.21
g (10.0 mmol) of 2,3-dimethylaniline, 3.42 g (10.0 mmol) of diphe-
nyliodonium-2-carboxylate monohydrate (2), and 73 mg (0.40 mmol)
of copper(II) acetate was added 40 mL of 2-propanol. The mixture
was stirred at reflux under N2 for 23 h and then concentrated in vacuo.
The solid residue was taken up in 100 mL of ethyl acetate, with gentle
heating to aid dissolution. The solution was extracted with water (2 ×
(38) Parke, Davis and Co. U.S. Patent No. 3238201, 1966.
(39) Scherrer, R. A.; Beatty, H. R. J. Org. Chem. 1980, 45, 2127.
9
4092 J. AM. CHEM. SOC. VOL. 125, NO. 14, 2003

Diaryloxamic Acid-Based PTP1B Inhibitors
A R T I C L E S
concentrated to an oil. The product was purified via reverse phase
HPLC, eluting with a gradient of 0-70% acetonitrile in 0.1% (v/v)
aqueous trifluoroacetic acid. Product-containing fractions were con-
centrated in vacuo below 40 °C to give 54 mg (16%) of a white
powder: 1H NMR (3:2 mixture of rotamers) (500 MHz, DMSO-d6) δ
8.37 (d, 1H, J ) 8.7 Hz), 8.05-8.03 (m, 1H), 8.01-7.95 (m, 1H),
7.87 (dd, 1H, J ) 1.8, 7.7 Hz), 7.68-7.73 (m, 1H), 7.63-7.59 (m,
1H), 7.57-7.49 (m, 2H), 7.45-7.38 (m, 1H), 7.34 (dt, 1H, J ) 1.1,
7.5 Hz), 7.21 (d, 1H (rotamer), J ) 7.7 Hz), 6.89 (d, 1H, J ) 7.3 Hz);
13C NMR (100 MHz, DMSO-d6) δ 167.5 (C), 166.8 (C), 163.6 (C),
163.3 (C), 162.4 (C), 141.6 (C), 138.6 (C), 138.3 (C), 136.5 (C), 134.4
(C), 134.1 (C), 133.4 (CH), 132.3 (CH), 131.5 (CH), 130.3 (CH), 130.1
(C), 129.6 (CH), 129.5 (C), 129.0 (C), 128.7 (CH), 128.5 (CH), 128.4
(CH), 128.2 (CH), 128.2 (CH), 128.1 (CH), 127.3 (CH), 126.8 (CH),
126.8 (CH), 126.5 (CH), 126.0 (CH), 125.9 (CH), 125.6 (CH), 125.3
(CH), 123.9 (CH), 122.9 (CH); MS (ESI) m/z 336 ([M + H]+), 353
([M + NH4]+), 358 ([M + Na]+).
2-(Acetylamino)-3-(4-aminonaphthalen-1-yl)acrylic Acid Methyl
Ester (7). To a mixture of 4-bromo-1-naphthylamine (6) (2.5 g, 11.3
mmol), Pd(OAc)2 (140 mg, 0.63 mmol), and P(o-tolyl)3 (570 mg, 1.87
mmol) in anhydrous N,N-dimethylformamide (10 mL) in a pressure
tube was added methyl 2-acetamidoacrylate (2.1 g, 14.7 mmol) and
triethylamine (5.3 mL, 37.5 mmol).40 The mixture was flushed with
nitrogen for 3 min and then sealed and heated at 110 °C for 4 h. The
reaction mixture was cooled to ambient temperature and then partitioned
between ethyl acetate and water. The aqueous layer was extracted once
with ethyl acetate, and the combined organic layers were washed with
brine, dried (Na2SO4), filtered, and concentrated in vacuo. The crude
residue was purified via silica gel chromatography, eluting with ethyl
acetate to provide 2.5 g (76%) of compound 7 as a yellow solid: 1H
NMR (400 MHz, DMSO-d6) δ 9.34 (s, 1H), 8.15 (d, 1H, J ) 8.0 Hz),
7.90 (d, 1H, J ) 8.0 Hz), 7.78 (s, 1H), 7.63 (d, 1H, J ) 8.0 Hz), 7.51
(t, 1H, J ) 7.6 Hz), 7.42 (t, 1H, J ) 7.6 Hz), 6.72 (d, 1H, J ) 8.0 Hz),
6.29 (s, 2H), 3.73 (s, 3H), 1.94 (s, 3H); 13C NMR (100 MHz, DMSO-
d6) δ 169.4, 166.0, 147.1, 136.2, 129.4, 129.2, 126.7, 123.8, 123.4,
122.9, 122.1, 116.4, 106.9, 51.9, 22.4; MS (ESI) m/z 285 ([M + H]+).
2-(Acetylamino)-3-(4-aminonaphthalen-1-yl)propionic Acid (8).
To a solution of 2-acetylamino-3-(4-aminonaphthalen-1-yl)acrylic acid
methyl ester (7) (2.5 g, 8.8 mmol) in 1:1 (v/v) ethyl acetate/methanol
(50 mL) under N2 was added Pd/C (10%, 250 mg). The reaction flask
was capped with a hydrogen balloon and heated at 60 °C for 18 h. The
mixture was filtered through diatomaceous earth, and the filter bed was
washed with 1:1 (v/v) ethyl acetate/methanol (2 × 25 mL). The filtrate
was concentrated in vacuo to provide the propoinate methyl ester (2.5
g, 100%). The methyl ester was dissolved in methanol (50 mL), and
then 3 N NaOH (4.75 mL, 14.3 mmol) was added dropwise. After the
reaction was stirred for 3 h at ambient temperature, the methanol was
removed in vacuo, and then the remaining aqueous solution was
acidified to pH ∼ 4.5 with 3 N HCl. The mixture was concentrated to
dryness in vacuo, taken up in 10% (v/v) ethanol/dichloromethane (25
mL), and filtered through diatomaceous earth. The filter cake was
washed with additional 10% (v/v) methanol/dichloromethane (10%, 25
mL). The filtrate was concentrated under reduced pressure to provide
acid 8 as a brown solid (1.75 g, 74%): 1H NMR (400 MHz, DMSO-
d6) δ 8.17 (d, 1H, J ) 8.0 Hz), 8.08 (d, 1H, J ) 8.0 Hz), 7.99 (d, 1H,
J ) 8.0 Hz), 7.48 (t, 1H, J ) 7.8 Hz), 7.38 (t, 1H, J ) 7.8 Hz), 7.12
(d, 1H, J ) 7.8 Hz), 6.60 (d, 1H, J ) 7.8 Hz), 1.77 (s, 3H), 4.45 (td,
1H, J ) 4.8, 9.2 Hz), 3.43 (dd, 1H, J ) 4.8, 14.0 Hz), 3.05 (dd, 1H,
J ) 9.2, 14.0 Hz); 13C NMR (100 MHz, DMSO-d6) δ 173.8, 169. 3,
143.8, 132.3, 128.3, 125.7, 123.4, 123.3, 123.0, 120.7, 107.1, 53.6,
34.3, 22.4; MS (ESI) m/z 273 ([M + H]+).
propionic acid (8) (500 mg, 1.84 mmol), 1-[3-(dimethylamino)propyl]-
3-ethylcarbodiimide hydrochloride (EDCI) (493 mg, 2.57 mmol),
1-hydroxybenzotriazole hydrate (360 mg, 2.21 mmol), and n-penty-
lamine (320 µL, 2.75 mmol) in anhydrous N,N-dimethylformamide (10
mL) was adjusted to pH ∼6 by the addition of triethylamine, and then
the reaction was stirred for 5 h at ambient temperature. The reaction
was diluted with water and extracted with ethyl acetate (3 × 15 mL).
The combined ethyl acetate layers were washed with water and then
brine, dried over Na2SO4, filtered, and concentrated in vacuo. The
product was purified via silica gel chromatography, eluting with 5%
(v/v) MeOH/EtOAc to provide amide 9 as a yellow solid (552 mg,
1.62 mmol, 88%): 1H NMR (400 MHz, DMSO-d6) δ 8.15 (d, 1H, J )
8.0 Hz), 8.07 (s, 1H), 8.04 (s, 1H), 7.87 (t, 1H, J ) 5.9 Hz), 7.45 (t,
1H, J ) 8.0 Hz), 7.36 (t, 1H, J ) 8.0 Hz), 7.08 (d, 1H, J ) 8.0 Hz),
6.58 (d, 1H, J ) 8.0 Hz), 5.52 (s, 2H), 4.27 (q, 1H, J ) 5.9 Hz), 3.28
(dd, 1H, J ) 5.9, 14.0 Hz), 2.94-3.06 (m, 3H), 1.77 (s, 3H), 1.18-
1.39 (m, 4H), 1.14 (heptet, 2H, J ) 7.8 Hz), 0.84 (t, 3H, J ) 7.5 Hz);
13C NMR (100 MHz, DMSO-d6) δ 171.1, 169.0, 143.6, 132.4, 128.2,
125.4, 123.9, 123.1, 122.8, 120.7, 107.0, 54.0, 38.9, 38.4, 35.0, 28.5,
28.5, 22.5, 21.8, 13.9; MS (ESI) m/z 342 ([M + H]+).
2-[(4-(2-(Acetylamino)-2-(pentylcarbamoyl)ethyl)naphthalen-1-
yl)amino]benzoic Acid (10). To a stirred suspension of 2-(acety-
lamino)-3-(4-aminonaphthalen-1-yl)-N-pentylpropionamide (9) (552 mg,
1.62 mmol) and diphenyliodonium-2-carboxylate monohydrate (2) (580
mg, 1.78 mmol) in N,N-dimethylformamide (10 mL) was added
anhydrous Cu(OAc)2 (14.6 mg, 0.081 mmol). The mixture was heated
at 95 °C for 1.5 h. The reaction mixture was then concentrated in vacuo,
and the crude residue was purified by reverse phase preparative HPLC
to give benzoic acid 10 as a light brown solid (619 mg, 83%): 1H
NMR (400 MHz, DMSO-d6) δ 13.11 (s, 1H), 9.94 (s, 1H), 8.29 (d,
1H, J ) 8.4 Hz), 8.18 (d, 1H, J ) 8.4 Hz), 7.98 (d, 1H, J ) 8.4 Hz),
7.94 (dd, 1H, J ) 1.2, 7.6 Hz), 7.86 (t, 1H, J ) 5.9 Hz), 7.61 (t, 1H,
J ) 7.6 Hz), 7.54 (t, 1H, J ) 7.6 Hz), 7.41 (d, 1H, J ) 7.6 Hz), 7.37
(d, 1H, J ) 7.6 Hz), 7.26 (td, 1H, J ) 1.2, 7.6 Hz), 6.78 (d, 1H, J )
8.4 Hz), 6.73 (t, 1H, J ) 5.9 Hz), 4.59 (q, 1H, J ) 7.6 Hz), 3.42 (dd,
2H, J ) 7.2, 13.6 Hz), 3.23 (dd, 1H, J ) 7.6, 14.0 Hz), 3.03 (qd, 1H,
J ) 7.6, 14.0 Hz), 2.91 (heptet, 1H, J ) 6.4 Hz), 1.81 (s, 3H), 1.13-
1.37 (m, 4H), 1.07 (heptet, 2H, J ) 5.9 Hz), 0.79 (t, 3H, J ) 5.9 Hz);
13C NMR (100 MHz, DMSO-d6) δ 170.5, 170.3, 169.1, 149.0, 135.0,
134.1, 132.9, 131.7, 130.9, 129.2, 127.5, 126.3, 125.9, 124.7, 122.5,
120.0, 116.7, 113.6, 111.7, 53.8, 38.5, 35.3, 28.5, 28.4, 22.5, 21.7, 13.8;
MS (ESI) m/z 462 ([M + H]+), 484 ([M + Na]+).
N-Acetyl-4-[(carboxycarbonyl)(2-carboxyphenyl)amino]-N-pen-
tyl-1-naphthylalaninamide (12). To an ice-cooled, stirred solution of
2-[(4-(2-(acetylamino)-2-(pentylcarbamoyl)ethyl)naphthalen-1-yl)ami-
no]benzoic acid (10) (619 mg, 1.34 mmol) and triethylamine (680 µL,
5.13 mmol) in dichloromethane (10 mL) was added ethyl oxalyl chloride
(452 µL, 4.04 mmol) slowly, over 30 min. The reaction was warmed
to ambient temperature and stirred for 16 h. After this time, 1 N HCl
(10 mL) was added, and the mixture was extracted with dichlo-
romethane (2 × 20 mL). The combined dichloromethane layers were
washed with brine, dried over Na2SO4, filtered, and concentrated in
vacuo. Crude ester 11 was hydrolyzed without any further purification.
To a solution of crude 2-{[4-(2-(acetylamino)-2-(pentylcarbamoyl)-
ethyl)naphthalen-1-yl](ethoxyoxalyl)amino}benzoic acid (11) (950 mg)
in methanol (7 mL) at ambient temperature was added 1 N NaOH (3.9
mL, 3.9 mmol). The basic mixture was stirred at ambient temperature
for 2 h, and then 1 N HCl (10 mL) was added. The product was purified
by reverse phase preparative HPLC to provide 210 mg of compound 3
as a light brown solid (0.39 mmol, 29% yield over two steps):
1H NMR (400 MHz, DMSO-d6) (mixture of rotamers) δ 13.13 (s,
1H), 8.43 (t, 1H, J ) 7.8 Hz), 8.32 (d, 1H, J ) 8.4 Hz), 8.22 (dd, 1H,
J ) 2.8, 8.4 Hz), [8.18 (d, J ) 8.4 Hz), 8.08 (d, J ) 8.4 Hz), 1H in
total], 8.01-7.83 (m, 3H), 7.75-7.12 (m, 5H), 6.86-6.80 (m, 1H),
4.58 (q, 1H, J ) 7.4 Hz), 3.67-2.82 (m, 6H), [(1.81, s), (1.78, s) (1.75,
s), 3H in total], 1.39-1.00 (m, 4H), 0.92-0.69 (m, 3H); 13C NMR
2-(Acetylamino)-3-(4-aminonaphthalen-1-yl)-N-pentylpropiona-
mide (9). A solution of 2-acetylamino-3-(4-aminonaphthalen-1-yl)-
(40) Dygos, J. H.; Yonan, E. E.; Scaros, M. G.; Goodmonson, O. J.; Getman,
D. P.; Periana, R. A.; Beck, G. A. Synthesis 1992, 741.
9
J. AM. CHEM. SOC. VOL. 125, NO. 14, 2003 4093

A R T I C L E S
Szczepankiewicz et al.
(100 MHz, DMSO-d6) (mixture of rotamers) δ (170.6 and 170.4), (169.2
and 169.1), (167.3 and 166.6), (163.5 and 163.1), 162.3, (138.6 and
138.5), (136.1 and 135.8), 137.2, (135.3 and 135.2), (133.2, 132.9, and
132.5), (132.0 and 131.9), 131.4, (130.3 and 130.2), (128.9 and 128.5),
127.9, (127.4 and 127.4), (126.7 and 126.6), 126.4, (125.2 and 124.9),
124.5, (124.4 and 124.2), (53.4 and 53.3), (38.5 and 38.4), (35.4 and
35.0), 28.5, (28.4 and 28.4), (22.5 and 22.4), 21.6, 13.8; MS (ESI) m/z
534 ([M + H]+).
3-((5-Aminopentyl)oxy)naphthalene-2-carboxylic Acid Methyl
Ester Hydrochloride (19). To an ice-cooled solution of 1.44 g (71.3
mmol) of 3-hydroxy-2-naphthoic acid methyl ester (18), 1.45 g (7.13
mmol) of 5-((tert-butoxycarbonyl)amino)-1-pentanol, and 1.87 g (7.13
mmol) of triphenylphosphine in 15 mL of THF was added 1.2 mL
(7.62 mmol) of diethylazodicarboxylate. The reaction was stirred and
allowed to warm to ambient temperature over 15 h. The solvent was
removed in vacuo, and then the thick oily residue was stirred with 15
mL of 30% ethyl acetate/hexanes to precipitate triphenylphosphine
oxide. The precipitate was filtered away and washed with 5 mL of
30% ethyl acetate/hexanes, and then the combined filtrate and washings
were concentrated to an oil. This was purified by silica gel chroma-
tography, eluting with 30% ethyl acetate/hexanes to give 1.54 g (56%)
of a pale yellow oil: 1H NMR (400 MHz, CDCl3) δ 8.28 (s, 1H), 7.80
(d, 1H, J ) 8.0 Hz), 7.70 (d, 1H, J ) 8.0 Hz), 7.49 (ddd, 1H, J ) 7.1,
6.8, 1.2 Hz), 7.35 (ddd, 1H, J ) 7.1, 7.1, 0.9 Hz), 7.17 (s, 1H), 4.66
(bs, 1H), 4.13 (t, 2H, J ) 6.3 Hz), 3.94 (s, 3 Hz), 3.17 (brq, 2H, J )
5.2 Hz), 1.90 (m, 2H), 1.59 (m, 4H); 13C NMR (100 MHz, CDCl3) δ
166.8 (C), 156.0 (C), 155.0 (C), 136.1 (C), 132.6 (CH), 128.6 (CH),
128.3 (CH), 127.5 (C), 126.4 (CH), 124.3 (CH), 122.1 (C), 107.7 (CH),
79.0 (C), 68.4 (CH2), 52.1 (CH3), 40.4 (CH2), 29.6 (CH2), 28.6 (CH2),
28.4 (CH3), 23.2 (CH2); MS (ESI) m/z 288 ([M - Boc + 2H]+), 388
([M + H]+), 410 ([M + Na]+).
The oil was dissolved in 10 mL of 4 M HCl in dioxane and stirred
at ambient temperature for 1.75 h. The solvent was removed in vacuo
to give 1.23 g (95%) of 3-((5-aminopentyl)oxy)naphthalene-2-carboxy-
lic acid methyl ester hydrochloride as a light yellow solid: 1H NMR
(400 MHz, DMSO-d6) δ 8.23 (s, 1H), 8.01 (bs, 3H), 7.93 (d, 1H, J )
8.0 Hz), 7.84 (d, 1H, J ) 8.3 Hz), 7.54 (ddd, 1H, J ) 7.1, 6.8, 1.2
Hz), 7.45 (s, 1), 7.39 (ddd, 1H, J ) 7.1, 6.8, 1.1 Hz), 4.12 (t, 2H, J )
6.3 Hz), 3.85 (s, 3H), 2.79 (t, 2H, J ) 7.7 Hz), 1.79 (m, 2H), 1.65 (m,
2H), 1.52 (m, 2H); 13C NMR (100 MHz, DMSO-d6) δ 166.2 (C), 154.0
(C), 135.5 (C), 131.2 (CH), 128.4 (CH), 128.2 (CH), 127.0 (C), 126.4
(CH), 124.3 (CH), 122.3 (C), 107.7 (CH), 67.8 (CH2), 52.0 (CH3), 38.6
(CH2), 27.9 (CH2), 26.5 (CH2), 22.4 (CH2); MS (ESI) m/z 288 ([M +
H]+ ).
S-3-(4-Aminonaphthalen-1-yl)-2-(tert-butoxycarbonylamino)pro-
pionic Acid Methyl Ester (16). To a mixture of 3.00 g (9.11 mmol)
of N-(tert-butoxycarbonyl)-3-iodo-L-alanine methyl ester and 3.57 g
(54.6 mmol) of zinc dust was added 10 mL of DMF.41 The mixture
was purged with N2 and then heated at 65 °C for 15 min. The
disappearance of the starting iodide was monitored by quenching an
aliquot in 1 M HCl, then extracting the acid solution with ethyl acetate,
and TLC analysis of the organic layer, eluting with 20% EtOAc/
hexanes. After the iodide had been consumed, the excess zinc was
allowed to settle, and the organozinc solution was added via syringe
to a stirred and N2-purged solution of 2.02 g (9.11 mmol) of 4-bromo-
1-naphthylamine, 556 mg (1.82 mmol) of tris(o-tolyl)phosphine, and
100 mg (0.450 mmol) of palladium(II) acetate in 5 mL of DMF. The
reaction was stirred at 65 °C for 1.5 h and then poured into 50 mL of
water and extracted with diethyl ether (3 × 15 mL). During the first
extraction, the mixture was filtered through diatomaceous earth to
remove the insoluble zinc salts and precipitated palladium. The
combined ether layers were back-extracted with water (1 × 15 mL)
and brine (1 × 15 mL), dried over MgSO4, filtered, and concentrated
to a gummy residue. This was purified via silica gel chromatography,
eluting with 40% ethyl acetate/hexanes to give 1.45 g (46%) of 3-(4-
aminonaphthalen-1-yl)-2-(tert-butoxycarbonylamino)propionic acid meth-
yl ester as a yellow foam. The product contained a minor higher Rf
impurity that could be removed during the next step: 1H NMR (400
MHz, CDCl3) δ 8.01 (d, 1H, J ) 8.3 Hz), 7.84 (d, 1H, J ) 8.3 Hz),
7.52 (t, 1H, J ) 7.2 Hz), 7.46 (t, 1H, J ) 7.4 Hz), 7.07 (d, 1H, J ) 7.7
Hz), 6.69 (d, 1H, J ) 7.7 Hz), 5.02 (bd, 1, J ) 8.0 Hz), 4.64 (bq, 1H,
J ) 6.1 Hz), 4.20 (bs, 2H), 3.61 (s, 3H), 3.47 (dd, 1H, J ) 14.2, 6.0
Hz), 3.36 (dd, 1H, J ) 14.0, 6.0 Hz), 1.40 (s, 9H); 13C NMR (100
MHz, CDCl3) δ 172.8 (C), 155.0 (C), 141.7 (C), 132.9 (C), 128.0 (CH),
126.2 (CH), 124.7 (CH), 124.2 (CH), 124.1 (C), 122.7 (C), 121.5 (CH),
109.2 (CH), 79.8 (C), 54.5 (CH or CH3), 52.0 (CH3 or CH), 35.2 (CH2),
28.3 (CH3); MS (ESI) m/z 245 ([M - Boc + 2H]+), 345 ([M + H]+),
367 ([M + Na]+).
S-3-(4-Aminonaphthalen-1-yl)-2-(tert-butoxycarbonylamino)pro-
pionic Acid (17). To a solution of 1.26 g (3.66 mmol) of 3-(4-
aminonaphthalen-1-yl)-2-(tert-butoxycarbonylamino)propionic acid meth-
yl ester (16) in 5 mL of CH3OH was added 1.5 mL of 4 M NaOH(aq).
The reaction was stirred at ambient temperature for 2.5 h and then
concentrated in vacuo to an oil. This was diluted with 20 mL of water
and extracted with diethyl ether (3 × 5 mL). The aqueous layer was
cooled with an ice bath and stirred vigorously, 10 mL of ethyl acetate
was added, and then 6 mL of 1 M HCl was added immediately. The
layers were separated, and the aqueous layer was extracted with
additional ethyl acetate (2 × 10 mL). The combined ethyl acetate layers
were back-extracted with brine (1 × 10 mL), dried over MgSO4, filtered,
and concentrated in vacuo to 1.10 g (91%) of a tan solid: 1H NMR
(300 MHz, DMSO-d6) δ 8.08 (d, 1H, J ) 8.1 Hz), 7.94 (d, 1H, J )
8.5 Hz), 7.48 (ddd, 1H, J ) 7.1, 6.8, 1.4 Hz), 7.38 (ddd, 1H, J ) 7.5,
6.8, 0.9 Hz), 7.13 (d, 1H, J ) 7.8 Hz), 7.10 (d, 1H, J ) 9.5 Hz), 6.58
(d, 1H, J ) 7.5 Hz), 4.11 (m, 1H), 3.39 (dd, IH, J ) 14.4, 4.4 Hz),
3.02 (dd, 1H, J ) 14.2, 10.2 Hz), 1.30 (s, 9H); 13C NMR (100 MHz,
CDCl3) δ 174.2 (C), 141.4 (C), 135.9 (C), 132.9 (C), 128.0 (CH), 126.1
(CH), 124.6 (CH), 124.3 (CH), 124.2 (C), 123.2 (C), 121.4 (CH), 109.5
(CH), 79.8 (C), 54.3 (CH), 34.8 (CH2), 28.2 (CH3); MS (ESI) m/z 231
([M - Boc + 2H]+), 331 ([M + H]+), 353 ([M + Na]+).
S-3-{[5-[(3-(4-Aminonaphthalen-1-yl)-2-(tert-butoxycarbonyl-
amino)propionyl)amino]pentyl]oxy}naphthalene-2-carboxylic Acid
Methyl Ester (20). To 510 mg (1.54 mmol) of acid 17, 413 mg (2.16
mmol) of 1-(3-(dimethylamino)propyl)-3-ethylcarbodiimide hydrochlo-
ride (EDCI), and 208 mg (1.54 mmol) of 1-hydroxybenzotriazole was
added 12 mL of CH2Cl2 and then 350 µL (3.18 mmol) of N-
methylmorpholine. The solution was stirred at ambient temperature for
10 min, and then 500 mg (1.54 mmol) of amine-HCl 19 was added.
The solution was stirred at ambient temperature for 19 h and then
concentrated in vacuo. The residue was taken up in 20 mL of ethyl
acetate and then extracted with water (2 × 10 mL), saturated NaHCO3-
(aq) (2 × 10 mL), and brine (1 × 10 mL), dried over MgSO4, filtered,
and concentrated to a thick oil. The product was purified via silica gel
chromatography, eluting with 60% ethyl acetate/hexanes to give an
oil. Trituration with a small amount of diethyl ether followed by rapid
evaporation in vacuo gave 480 mg (52%) of the amide as a light brown
foam: 1H NMR (400 MHz, CDCl3) δ 8.27 (s, 1H), 8.14 (d, 1H, J )
8.3 Hz), 7.81 (d, 1H, J ) 8.3 Hz), 7.78 (d, 1H, J ) 8.6 Hz), 7.72 (d,
1H, J ) 8.0 Hz), 7.51 (t, 2H, J ) 7.5 Hz), 7.40 (t, 1H, J ) 7.4 Hz),
7.37 (dt, 1H, J ) 7.1, 0.9 Hz), 7.15 (s, 1H), 7.11 (d, 1H, J ) 7.4 Hz),
6.64 (d, 1H, J ) 7.7 Hz), 5.47 (bs, 1H), 5.36 (bs, 1H), 4.39 (bq, 1H,
J ) 6.5 Hz), 4.01 (t, 2H, J ) 6.3 Hz), 3.90 (s, 3H), 3.47 (m, 1H), 3.27
(m, 1H), 3.05 (m, 2H), 1.74 (m, 2H), 1.41 (bs, 9H), 1.24 (m, 2H); 13
C
NMR (100 MHz, CDCl3) δ 171.2 (C), 166.6 (C), 155.3 (C), 155.0
(C), 141.6 (C), 136.1 (C), 132.6 (CH), 128.7 (CH), 128.4 (CH), 128.3
(CH), 127.5 (C), 126.3 (CH), 124.7 (CH), 124.4 (CH), 124.3 (CH),
124.0 (C), 123.4 (C), 122.0 (C), 121.4 (CH), 109.3 (CH), 107.6 (CH),
79.8 (C), 68.3 (CH2), 55.8 (CH), 52.2 (CH3), 39.1 (CH2), 36.4 (CH2),
28.5 (CH2), 28.3 (CH2), 28.3 (CH3), 23.0 (CH2); MS (ESI) m/z 500
([M - Boc + 2H]+), 600 ([M + H]+), 622 ([M + Na]+).
(41) Jackson, R. F. W.; Moore, R. J.; Dexter, C. J. Org. Chem. 1998, 63, 7875.
9
4094 J. AM. CHEM. SOC. VOL. 125, NO. 14, 2003

Diaryloxamic Acid-Based PTP1B Inhibitors
A R T I C L E S
Chart 1.
3-{{5-[[2-(tert-Butoxycarbonylamino)-3-[4-((2-carboxyphenyl)-
amino)naphthalen-1-yl]propionyl]amino]pentyl}oxy}naphthalene-
2-carboxylic acid (21). To 382 mg (0.637 mmol) of aminonaphthalene
20, 262 mg (0.766 mmol) of diphenyliodonium-2-carboxylate mono-
hydrate (2), and 5 mg (0.028 mmol) of copper(II) acetate was added 3
mL of 2-propanol. The mixture was stirred at reflux under N2 for 17 h
and then concentrated in vacuo to a foam. This was taken up in 10 mL
of ethyl acetate, then extracted with water (2 × 5 mL) and brine (1 ×
5 mL), dried over MgSO4, filtered, and concentrated to 481 mg
(>100%) of a brown foam. The crude product contained some
iodobenzene and a small amount of unreacted iodonium salt, but these
did not interfere with the next reaction. Pure acid 21 could be obtained
by silica gel chromatography, eluting with 50% ethyl acetate/hexanes
to remove nonpolar impurities and then stepping to 75% ethyl acetate/
hexanes to recover the product. From 100 mg of crude product and 30
mL of silica gel, 16 mg of pure acid 21 was obtained: 1H NMR (400
MHz, DMSO-d6) δ 10.03 (bs, 1H), 8.23 (d, 1H, J ) 5.8 Hz), 8.22 (s,
1H), 7.99 (d, 1H, J ) 8.6 Hz), 7.93 (m, 2H), 7.85 (m, 1H), 7.83 (d,
1H, J ) 8.3 Hz), 7.61 (dt, 1H, J ) 7.1, 1.2 Hz), 7.54 (m, 2H), 7.41
(m, 2H), 7.24 (t, 1H, J ) 7.4 Hz), 6.91 (d, 1H, J ) 8.6 Hz), 6.76 (d,
1H, J ) 8.3 Hz), 6.71 (t, 1H, J ) 7.7 Hz), 4.30 (m, 1H), 4.09 (t, 2H,
J ) 6.3 Hz), 3.83 (s, 3H), 3.49 (dd, 1H, J ) 13.8, 5.2 Hz), 3.14 (m,
3H), 1.75 (m, 2H), 1.43 (m, 4H), 1.27 (bs, <9H), 0.98 (bs, minor
rotamer); 13C NMR (100 MHz, DMSO-d6) δ 171.0 (C), 170.3 (C), 166.2
(C), 155.0 (C), 154.0 (C), 148.9 (C), 135.5 (CH), 135.0 (C), 133.8
(C), 132.8 (C), 131.7 (CH), 131.1 (CH), 131.0 (C), 129.2 (C), 128.4
(CH), 128.2 (CH), 127.6 (CH), 126.9 (C), 126.4 (CH), 126.3 (CH),
125.8 (CH), 124.5 (CH), 124.3 (CH), 122.5 (C), 122.4 (CH), 120.0
(CH), 116.6 (CH), 113.5 (CH), 112.1 (C), 107.6 (CH), 77.9 (C), 68.0
(CH2), 55.3 (CH), 51.9 (CH3), 38.4 (CH2), 35.1 (CH2), 28.6 (CH2),
28.1 (CH2), 28.0 (CH3), 27.5 (CH3, minor), 22.7 (CH2); MS (ESI) m/z
620 ([M - Boc + 2H]+), 720 ([M + H]+), 742 ([M + Na]+).
3-({5-[(N-Acetyl-3-{4-[(carboxycarbonyl)(2-carboxyphenyl)amino]-
1-naphthyl}-L-alanyl)amino]pentyl}oxy)-2-naphthoic Acid (23). To
an ice-cooled solution of 100 mg (0.139 mmol) of amino acid 21 in 1
mL of DMF was added 200 µL (1.44 mmol) of triethylamine and then
100 µL (0.90 mmol) of ethyl chlorooxoacetate. The ice bath was
removed, and the reaction was stirred at ambient temperature for 2 h.
Next, 2 mL of 2 M NaOH was added, along with 5 mL of water. After
being stirred for another 2 h, the reaction was diluted with 10 mL of
water, and stirred again for 2 h. The reaction was monitored for
disappearance of the less polar naphthoic ester by reverse-phase
analytical HPLC. Additional 2 M NaOH could be added if hydrolysis
was incomplete, though a total of 3 mL of 2 M NaOH was sufficient
to complete the hydrolysis. The reaction was cooled with an ice bath,
and then 3 mL of 6 M HCl was added. The mixture was extracted
with ethyl acetate (3 × 5 mL) and then the combined ethyl acetate
layers were back-extracted with water (2 × 5 mL) and brine (1 × 5
mL), dried over MgSO4, filtered, and concentrated to a foam containing
crude oxamic acid 22. This was dissolved in 1 mL of trifluoroacetic
acid and allowed to stand at ambient temperature for 10 min. The
trifluoroacetic acid was removed in vacuo, and the resulting foam was
taken up in 3 mL of 2 M NaOH. To the basic solution was added 300
µL of acetic anhydride, and the mixture was stirred at ambient
temperature for 30 min, after which the pH ) 7. A small amount of
insoluble material was filtered away, and then the product was isolated
by reverse phase HPLC, eluting with a 0-70% acetonitrile in 0.1%
(v/v) aqueous trifluoroacetic acid gradient to give 28 mg (28%) of
compound 23 as a light yellow solid. The enantiomeric purity of
compound 23 was determined to be 100% by chiral HPLC. The
synthesis of ent-23 was accomplished according to the same sequence
but starting with N-(tert-butoxycarbonyl)-3-iodo-D-alanine methyl ester
to prepare ent-16. Both enantiomers were treated with diazomethane
and then evaluated using a ChiralPak AS column from Chiral
Technologies (4.6 × 250 mm, 10 µm), eluting with 30% 2-propanol/
1
hexanes. tR for 23 ) 9.9 min, and tR for ent-23 ) 15.8 min. H NMR
(500 MHz, DMSO-d6): mixture of rotamers, δ 8.45-8.42 (m, 1H),
8.35-8.30 (m, 1H), 8.26-8.21 (m, 1H), 8.18 (s, 1H), 8.06-7.95 (m,
2H), 7.92 (d, 1H, J ) 8.2 Hz), 7.84 (m, 2H), 7.67-7.47 (m, 5H), 7.17
(m, 1H), 6.83 (t, 1H, J ) 6.4 Hz), 4.70-4.58 (m, 1H), 4.07 (t, 1H, J
) 6.4 Hz), 4.03 (t, 1H, J ) 6.4 Hz), 3.59-2.99 (m, 4H), 2.07 (s, 3H),
1.80-1.63 (m, 4H), 1.39-1.16 (m, 5H). MS (ESI): m/z 720 ([M +
H]+), 742 ([M + Na]+).
Protein NMR Spectroscopy. Human PTP1B (residues 1-288) was
cloned into the pGEX-5X vector (Amersham Pharmacia) and expressed
in Escherichia coli BL21(DE3) cells. The final protein construct used
for the NMR studies was 292 residues comprised of the following amino
acid sequence shown in Chart 1. Two point mutations exist in the
construct (E252D and D265E, underlined), along with four additional
N-terminal residues (GFSH) that remain after cleavage of the GST-
fusion.
Uniformly 15N-labeled and 13CH3(IVLM)-labeled PTP-1B was
prepared by growing the bacteria on a minimal medium containing 1
g/L of 15NH4Cl (Cambridge Isotopes) or by adding 100/100/50 mg/L
of 13C-γ-methionine/3,3′-13C-R-ketovalerate/3-13C-R-ketobutyrate24 1 h
before induction with 1 mM IPTG, respectively. Carbenicillin (0.1 mg/
mL) was used as the selection agent. In addition to the standard salts,
metals, and vitamins, the media contained the following additives (g/L
of culture): Ala (0.5); Arg (0.4); Asp (0.4); Cys (0.05); Glu (0.65);
Gln (0.4); Gly (0.55); His (0.1); Lys (0.42); Phe (0.13); Pro (0.1); Ser
(2.1); Thr (0.23); Tyr (0.17); adenine (0.5); guanosine (0.65); thymine
(0.2); uracil (0.5); cytosine (0.2); sodium acetate (1.5); succinic acid
(1.5). The cell pellets were resuspended in a buffer containing 10 mM
sodium phosphate (pH 7.4) and 150 mM NaCl (PBS buffer). The cells
were then lysed using a microfluidizer (Microfluidics International)
and the resulting lystate clarified via centrifugation at 76 000g for 10
min. The clarified cell lysate was loaded onto a Glutathione Sepharose
4B column (Pharmacia), and the resin was washed with 20 column
volumes of PBS buffer. The protein was eluted with 5 column volumes
of a buffer containing 50 mM TRIS (pH 8.0) and 10 mM reduced
glutathione. The N-terminal GST fusion was cleaved with thrombin (1
U thrombin/mg of PTP-1B) at 4 °C for 72 h. Cleavage was monitored
by SDS-PAGE analysis. The resulting solution was then concentrated
and dialyzed to remove the glutathione (final dilution factor of 10-10
)
and loaded onto a Glutathione Sepharose 4B column to remove the
GST. The flow-through (containing PTP-1B) was collected and dialyzed
against NMR buffer (see below). Typical protein yields were 10-20
mg of PTP-1B/L of culture.
13
The NMR samples were composed of uniformly 15N-labeled or
-
CH3(IVLM) PTP-1B in an H2O/D2O (9/1) TRIS buffered solution (25
mM, pH ) 7.5) containing DTT (10 mM). Ligand binding was detected
by acquiring sensitivity-enhanced 1H/15N- or 1H/13C-HSQC spectra42,43
(42) Kay, L. E.; Keifer, P.; Saarinen, T. J. Am. Chem. Soc. 1992, 114, 10663.
(43) Stonehouse, J.; Shaw, G. L.; Keeler, J.; Laue, E. D. J. Magn. Reson. 1994,
107, 178.
9
J. AM. CHEM. SOC. VOL. 125, NO. 14, 2003 4095

A R T I C L E S
Szczepankiewicz et al.
on 400 µL of PTP-1B in the presence and absence of added compound.
Protein concentrations were 300 and 25 µM for the 15N- and 13CH3-
(IVLM)-labeled samples, respectively. A Bruker sample changer was
used on a Bruker AMX500 spectrometer equipped with a cryoprobe.44
Compounds were individually tested at concentrations of 0.1-15.0 mM,
and binding was determined by monitoring changes in the HSQC
spectra. Dissociation constants were obtained for selected compounds
by monitoring the chemical shift changes of the protein resonances as
a function of ligand concentration. Data were fit using a single binding
site model. A least-squares grid search was performed by varying the
values of KD and the chemical shift of the fully saturated protein.
X-ray Crystallography. The final concentration of the purified
protein sample was 3-4 mg/mL and contained various amounts of
reducing agents, normally 2-4 mM DTT. Crystals of PTP1B (residues
1-321) belonging to space group P3121 (a ) b ) 88.4 Å, c ) 104.5
Å) were grown routinely using the protocol described by Puius31 with
the following changes. The buffers used in protein crystallization were
degassed for approximately 30 min and then purged with argon gas
prior to their use for crystallization. This simple precaution extended
the lifetime of the reactive cysteine (Cys215) in the active site and
provided a more stable and structurally homogeneous conformation of
the active site. This protocol allowed the soaking of all the PTP1B
inhibitors directly into the crystals. Soaking times ranged from one to
5 h and were normally no more that 2 h. Soaking was performed using
50 µL of the reservoir in the crystallization well and mixed with a
2-4 µL of stock (DMSO) solution of the compound at 100 mM
concentration. The soaking protocol permitted the rapid structure
determination of PTP1B:inhibitor complexes.
After the soaking period, crystals were transferred to a cryoprotectant
solution for 5-10 s with rayon loops and exposed to the X-rays. The
cryoprotectant solution consisted of 100 mM HEPES and 0.2 M
magnesium acetate adjusted to pH 7.1 with NaOH at a 16% w/v
concentration of PEG8000 with a 35% v/v concentration of glycerol.
Crystallographic data were collected both in a conventional, in-house,
protein crystallography laboratory and at sector 17 at the Advanced
Photon Source (ID and BM lines of IMCA-CAT). Conventional sources
were used routinely. The data collection hardware were a Mar Research
Image Plate (180 mm), a Mar Research 345 mm image plate, or a Mar
Research 165 mm CCD detector, all driven by the manufacturer’s
software. For a complete data set, 80-100 1° frames were collected at
a crystal to detector distance of 80-90 mm. Exposure times were
typically 10 min/frame for in-house data collection and 10 s for the
synchrotron source. This strategy resulted in >96% completeness in
the vast majority of the data sets with reducing R-factors between 5
and 9%. Data were processed with HKL2000 and scaled with
SCALEPACK or within HKL2000. Phase and map calculations were
performed using XPLOR or CNX. The modeling and electron-density
fitting software QUANTA was used to manipulate the models. Refined
crystallographic coordinates for the structure of hPTP1B complexed
with compounds 5, 12, and 23 have been deposited with the Protein
Data Bank (www.rcsb.org) with entry codes 1NO6, 1NL9, and 1NNY,
respectively.
Biological Assay. Protein tyrosine phosphatase 1B (PTP1B) activity
was determined by measuring the rate of hydrolysis of a surrogate
substrate, p-nitrophenyl phosphate (pNPP, C1907 Sigma, St. Louis,
MO). The assay was carried out at room temperature in 96 well
polypropylene or polyethylene plates in a total volume of 100 µL/well.
Appropriate dilutions of the compounds were made in DMSO and then
diluted 10-fold into assay buffer (25 mM HEPES (pH ) 7.5), 150 mM
NaCl, and 0.1 mg/mL BSA). A 10 µL volume of 5 different
concentrations of the test compound (inhibitor) or 10% DMSO in assay
buffer was added to individual wells containing 40 µL of 3.2, 8, 20,
and 50 mM pNPP in water. The reaction was initiated by adding 50
µL of PTP1B diluted in 2× assay buffer (50 mM HEPES, 300 mM
NaCl, and 0.2 mg/mL BSA). The phosphatase activity resulted in the
formation of the colored product p-nitrophenol (pNP) which was
continuously monitored at 405 nm every 30 s for 15 min using an
appropriate plate reader. The absorbance at 405 nm was converted to
nanomoles of pNP using a standard curve, and the initial rate of pNP
formation was calculated. For each concentration of test compound
(inhibitor) or DMSO control, the initial rates were used to fit the
rectangular hyperbola of Michaelis-Menten by nonlinear regression
analysis (GraphPad Software Prism 3.0). The ratio of the apparent Km/
Vmax vs inhibitor concentration was plotted, and the competitive Ki was
calculated by linear regression to be the negative x-intercept. The
uncompetitve Ki was similarly calculated from the x-intercept of the
plot of the reciprocal of the apparent Vmax versus the inhibitor
concentration.27
Acknowledgment. The authors thank Jamey Mack for
preparation of the 13C methyl-labeled PTP1B for NMR-based
screening and Mark Mullally for determining the enantiomeric
purity of inhibitor 23.
Supporting Information Available: Kinetic data of inhibition
for naphthyloxamic acid 5. This material is available free of
charge via the Internet at http://pubs.acs.org.
(44) Hajduk, P. J.; Gerfin, T.; Boehlen, J.-M.; Ha¨berli, M.; Marek, D.; Fesik,
S. W. J. Med. Chem. 1999, 42, 2315.
JA0296733
9
4096 J. AM. CHEM. SOC. VOL. 125, NO. 14, 2003