1,2,3,4-Tetrahydroisoquinolinyl sulfamic acids as phosphatase PTP1B inhibitors

Article abstract of DOI:10.1016/j.bmcl.2005.12.051

High-throughput screening of the P&GP corporate repository against several protein tyrosine phosphatases identified the sulfamic acid moiety as potential phosphotyrosine mimetic. Incorporation of the sulfamic acid onto a 1,2,3,4-tetrahydroisoquinoline sca

Full text of DOI:10.1016/j.bmcl.2005.12.051

Bioorganic & Medicinal Chemistry Letters 16 (2006) 1574–1578
1,2,3,4-Tetrahydroisoquinolinyl sulfamic acids as phosphatase
PTP1B inhibitors
Sean R. Klopfenstein,* Artem G. Evdokimov, Anny-Odile Colson, Neil T. Fairweather,
Jeffrey J. Neuman, Matthew B. Maier, Jeffrey L. Gray, Gina S. Gerwe, George E. Stake,
Brian W. Howard, Julie A. Farmer, Matthew E. Pokross, Thomas R. Downs,
Bhavani Kasibhatla and Kevin G. Peters
Procter & Gamble Pharmaceuticals, Health Care Research Center, 8700 Mason-Montgomery Rd., Mason, OH 45040, USA
Received 2 November 2005; revised 8 December 2005; accepted 8 December 2005
Available online 4 January 2006
Abstract—High-throughput screening of the P&GP corporate repository against several protein tyrosine phosphatases identified the
sulfamic acid moiety as potential phosphotyrosine mimetic. Incorporation of the sulfamic acid onto a 1,2,3,4-tetrahydroisoquinoline
scaffold provided a promising starting point for PTP1B inhibitor design.
Ó 2006 Elsevier Ltd. All rights reserved.
The PTPase PTP1B has been shown to be involved in
the insulin signaling cascade as a negative regulator of
insulin signaling.1 There is a growing body of evidence
to suggest that inhibition of PTP1B can enhance insulin
signaling primarily through prolonged activation of the
insulin receptor. Recent studies have shown that PTP1B
knock-out mice exhibit increased insulin sensitivity.2,3 In
addition, treatment of animal models of type-2 diabetes
with antisense oligonucleotides has been shown to nor-
malize blood glucose levels.4 These results have spurred
the research community to develop small molecule
inhibitors of PTP1B for the treatment of type-2 diabetes
and several groups have reported compounds to this
end.5–7
High-throughput screening of the P&GP corporate
repository against multiple phosphatase enzymes identi-
fied a number of sulfamic acid derivatives, typified by 1,
which were potent inhibitors of several soluble phospha-
tases, but showed only weak inhibition of PTP1B. Based
on these structures we hypothesized that a sulfamic acid
moiety linked directly to an aryl group is likely to be a
better mimic of the PTP1B substrate—phosphotyrosine.
The assumption of the aryl sulfamic acid moiety as a
potential phosphotyrosine mimetic allowed us to launch
PTP1B inhibitor discovery efforts.8 As a starting point
toward this goal, simple aryl sulfamic acids were
prepared which demonstrated varying levels of potency
against several PTPases (Fig. 1). One of the initial
Figure 1. Starting points for PTP1B inhibitor design.
Keywords: Phosphatase PTP1B; Sulfamic acid; Diabetes; HTS screening; Tetrahydroisoquinoline.
*
Corresponding author. Tel.: +1 513 622 3692; fax: +1 513 622 1195; e-mail: klopfenstein.sr@pg.com
Present address: Computational Chemistry Section, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of
Health, 12 Center Dr, Bethesda, MD 20892, USA.
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.12.051

S. R. Klopfenstein et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1574–1578
1575
Scheme 1. Reagents and conditions: (a) KNO3, H2SO4, 5 °C to rt; (b) Boc2O, Et3N, H2O/MeOH; (c) EDCI, HOBt, DiPEA, DMF, MeNH2; (d) 4 M
HCl/1,4-dioxane; (e) R1CO2H, EDCI, HOBt, DMF, DiPEA, 0 °C or R1COCl, DiPEA, DCM or R1NCO, CH2Cl2, DiPEA; (f) H2, Pd/C, MeOH; (g)
Pyr–SO3, Pyr then NH4OH.
scaffolds which showed promise for further PTP1B
inhibitor design was based on the 7-substituted-1,2,3,4-
tetrahydroisoquinoline (TIQ) structure.9 The initial
TIQ derivative 3 provided comparable potency to the
PTP1B substrate-like phenylalanine analog 2. Also, we
observed a modest 4-fold selectivity of the TIQ analog
for PTP1B versus T cell protein tyrosine phosphatase
(TC-PTP).10 In addition to the in vitro screening data
we were excited about the prospect of developing a con-
strained ring system which could provide two points of
diversity for rapid SAR development. Herein, we report
a new class of PTP1B inhibitors based on a sulfamic acid
pTy mimetic and the 1,2,3,4-tetrahydroisoquinoline
system.11,12
Nitration of (R) or (S)-1,2,3,4-tetrahydroisoquinoline-
3-carboxylic acid (4) with KNO3/H2SO4 provided the
7-nitro derivative in high yield which was converted
directly to the Boc derivative 5. Bulk synthesis of the
3-methylcarboxamide derivative 6 was done using a
mixed anhydride approach13 followed by Boc deprotec-
tion with 4 M HCl/1,4-dioxane. This sequence has been
routinely carried out on a multi-gram scale. Diversity at
the 2-position was introduced using routine chemistry.
Introduction of the sulfamic acid functionality was car-
ried out in two steps. The 7-nitro group was reduced
either by catalytic hydrogenation with 10% palladium
on carbon or with SnCl2 (e.g., 8k and m). The resulting
primary anilines were treated with pyridine–SO3 com-
plex in pyridine to give the desired products. This pro-
cess generally provided the final products in 25–50%
yield following RP-HPLC purification.14
Synthesis of the described TIQ sulfamic acids was
accomplished in seven steps from readily available start-
ing materials (Scheme 1).
Additional diversity was introduced at the 3-position of
the TIQ ring system (Table 2) from intermediate 5 using
Table 1. PTP1B inhibition data for compounds 8a–m
Compound
R1
PTP1B IC50 (lM)
Table 2. PTP1B inhibition data for compounds 9a–f
8a
8b
L-Phe(Boc)
>500
285.7
8c
271.2
8d
164
Compound
R2
PTP1B IC50 (lM)
8e
153.5
148
9a
>500
8f
8g (R)
134.8
9b
9c
9d
9e
9f
120.2
51.4
19.0
16.7
8.23
8h
8i
95.3
83.6
64.4
37
8j
8k
8l
36.7
24.9
8m

1576
S. R. Klopfenstein et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1574–1578
standard conditions. Following introduction of the R2
substituent the synthesis was completed in a similar
manner as that outlined in Scheme 1.
of 3, the tert-butyl carbamate side chain was oriented in
the general direction of the previously reported second
aryl phosphate binding site (P + 1 pocket), while 8g
adopted a flipped conformation and oriented the Boc side
chain in the general direction of Lys41 (P À 1 pocket).17
With this information, we decided to target the P + 1
pocket for potency enhancements by diversifying the
(S)-TIQ scaffold. As a starting point, we chose to investi-
gate the 2-position of the TIQ ring keeping the 3-position
fixed as the methylamide similar to 3 and replacing the
Boc group with simple substituents.
The initial TIQ sulfamic acid analog, 3, which showed
moderate PTP1B inhibition was the basis for further
exploration around a core 1,2,3,4-tetrahydroisoquinoline
structure. Compound 3 was shown to be a time-indepen-
dent, competitive, reversible inhibitor of PTP1B with a
Ki = 24 lM.15 Inversion of the stereochemistry to provide
(R)-3 (8g) resulted in a 3-fold loss in potency. The X-ray
structures of both 3 and 8g were obtained by soaking
methods and showed opposing orientations of the side
chains within the active site of the enzyme.16 In the case
Incorporation of an amino acid residue, as in 8a, result-
ed in a complete loss of potency. This is in contrast to
Table 3. SAR survey of P + 1 aryl substitution for TIQ sulfamic acids
Compound
X
Ar
PTP1B IC50 (lM)
10a
CH2
242
10b
10c
CH2
CH2
224.8
198
10d
CH2
183.7
10e
10f
CH2
CH2
146
101.1
10g
10h
10i
10j
O
93.4
82.1
4.8
CH2
CH2
CH2
2.5

S. R. Klopfenstein et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1574–1578
1577
what has been observed in other less rigid peptide-based
inhibitors.18–20 Initial experiments summarized in Table
1 were designed to gain insight into the ideal linker
length between the TIQ scaffold and the aryl group
which was to interact with residues in the P + 1 pocket.
The data seem to indicate a preference for a three-atom
linker between the TIQ system and the aryl ring (i.e.,
8j–m) and that a urea linkage (8d) is not well tolerated.
Table 2 provides a representative example of the SAR
developed for the 3-positionof the TIQ ring system. Intro-
duction of a benzyl amide side chain in this position
proved to be uniformly poor (e.g., 9b) for all analogs
regardless of the R1 substituent. A phenethyl amide side
chain (9c) returned compounds of roughly equal potency
to that of 3. Incorporation of an ester functionality in the
3-position provided a potency enhancement over 3 and
this seemed to be somewhat independent of the size of
the ester (9d and e). Reduction of the methyl ester 9d to
the primary alcohol 9a resulted in a complete loss of activ-
ity, the reasons for which are unclear. Incorporation of an
oxadiazole 9f provided a 5-fold gain in potency and one of
the most potent compounds from this survey.
Figure 3. Crystal structure of 10i bound to PTP1B. Carbon atoms of
the compound are shown in green, those of the key residues in the P0
and P1 sites are colored dark blue and light brown, respectively.
Table 4. Selectivity of 10i for a panel of protein phosphatases
Phosphatases
IC50 (lM)
PTP1B
HCPTPA
HPTPb
SHP-1
SHP-2
HPTPl
HPTPg
LAR
4.8
3.1
The second phase of the investigation centered on iden-
tifying R1 side chains which would be able to capture the
positive interactions within the P + 1 pocket which have
been previously reported.21 These interactions center on
key ionic interactions with Arg24 and Arg54 of the pro-
tein. We were fortunate to have X-ray structural data
for most of the three-atom-linked aryl systems. Most
notably the X-ray structure of 8k showed the benzyl car-
bamate side chain to be positioned in the P + 1 pocket
and to be ideally situated to interact with these residues
if functionalized with the appropriate groups. Due to
ease of synthesis and comparable potency to 8k, entry
8j was chosen as the starting point for optimization of
interactions in the P + 1 pocket. We investigated substi-
tution patterns on the aryl ring to identify elements
capable of picking up the key interactions and to further
probe the electronic and steric requirements of the
pocket (Table 3).
20.7
203
>500
>500
150
>500
>500
>500
8.1
CD45
PTPc
STP-1
VHR
TC-PTP38
HPTPe
4.0
9.9
>500
sulfamic acid bound tightly in the P + 1 pocket. The
second sulfamic acid moiety picks up hydrogen bonding
interactions with Gln262, Arg24, and Arg254 of the
protein. The charged sulfamic acid pulls Arg24 into
the P + 1 and out of its native conformation to form a
salt bridge. A similar result was obtained for the
3-carboxylate analog 10j which provided a 16-fold in-
crease in potency. Once again, X-ray crystallography
showed the aryl carboxylate bound tightly in the P + 1
pocket with similar interactions to the bis-sulfamic
acid 10i. Analog 10i was screened against a panel of
14 PTPases (Table 4). High selectivity was observed
against several other therapeutically useful phosphatases
Electron-withdrawing or releasing groups in the 3-posi-
tion (10e, f, and h) did not provide gains in potency
relative to 8j. Similarly, substitution in the 4-position
(10a–c) was detrimental to potency.
Incorporation of a second sulfamic acid residue in the
3-position of the side chain (10i) did result in a 13-fold
increase in potency. X-ray crystal data of 10i bound in
the active site of PTP-1B (Fig. 3) show the second aryl
Figure 2. Simplification of 10i and the effect on potency and selectivity.

1578
S. R. Klopfenstein et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1574–1578
2. Elchebly, M. et al. Science 1999, 283, 1544.
3. Klaman, L. D. et al. Mol. Cell. Biol. 2000, 20, 5479.
4. Zinker, B. A. et al. Proc. Natl. Acad. Sci. U.S.A. 2002, 99,
11357.
(i.e., CD45, PTPc, SHP2, etc.). Unfortunately, selectivi-
ty over the closely homologous TC-PTP was poor (2-
fold).
5. Bialy, L.; Waldmann, H. Angew. Chem., Int. Ed. 2005, 44,
3814.
6. Liu, G. Curr. Med. Chem. 2003, 10, 1407.
7. Blaskovich, M. A.; Kim, H.-O. Expert Opin. Ther. Patents
2002, 12, 871.
8. During the course of this work a report appeared
describing sulfamic acids as phosphatase inhibitors. San-
kar et al.; Preparation of sodium sulfamic acid salts as
inhibitors of human cytoplasmic protein tyrosine phos-
phatases for the treatment of wounds and of damaged
tissues; US 2002-368901.
9. Tetrahydroisoquinolines have been reported to have
cdc25B inhibitory activity Fritzen, E. L. et al. Bioorg.
Med. Chem. Lett. 2000, 10, 649.
Complete removal of the 3-position substituent from 10i
provided 12 which showed a 5-fold loss in potency
(Fig. 2). Additional simplification of 10i by removal of
the constraining TIQ ring gave the benzyl amine deriva-
tive 11 which also suffered from a loss in potency (6-fold).
Interestingly, the X-ray structure of 11 showed a reversal
in the binding mode of the compound. In the case of 11,
the sulfamic acid residue of the propionic acid side chain
was shown to bind in the P0 pocket, while the benzyl
amine unit was extended into the P + 1 pocket.
High-throughput screening of the P&GP corporate
repository identified the sulfamic acid functionality as
a potential pTyr mimetic. Incorporation of a sulfamic
acid moiety at the 7-position of the 1,2,3,4-tetrahydro-
isoquinoline scaffold provided compounds which
showed promising initial PTP1B inhibitory properties.
X-ray crystallography provided a guide to inhibitor de-
sign by determining that side chains extending from the
2-position of the natural enantiomer pointed toward the
second aryl phosphate binding region (P + 1 pocket).
An initial survey of side chains extending from the 2-po-
sition of the TIQ scaffold indicated that a three-atom
linker between the TIQ ring system and the distal aryl
ring was preferred. A second investigation provided an
order of magnitude increase in potency by incorporating
an acidic functionality in the 3-position of the aryl ring.
10. Asante-Appiah, E. et al. J. Biol. Chem. 2001, 276, 26036.
11. Klopfenstein, S. R. et al. US 2004-0167183 A1.
12. Klopfenstein, S. R. et al. WO 2004-074256 A1.
13. Chaudhary, A.; Girgis, M.; Parshad, M.; Hu, B.; Har, D.;
ˇ
Repic, O.; Blacklock, T. J. Tetrahedron Lett. 2003, 44,
5543.
14. All compounds were fully characterized by 1H and 13C
NMR, HPLC, MS, and combustion analysis.
15. Phosphatase activity and kinetics: recombinant phospha-
tases (nM) were diluted in assay buffer (150 mM NaCl, 0–
0.1% BSA, 5 mM DTT, and 50 mM Tris–HCl) (pH 7) or
10 mM Na/Tris acetate (pH 6) and incubated with 10 lM
DiFMUP (Molecular Probes). After 15 min, the fluores-
cence increase was measured on a Victor V plate reader
(Wallac). Inhibitors were pre-incubated with enzyme for
10 min before substrate addition. Kinetic measurements
were measured using 3–5 concentrations of inhibitor
around the IC50 over a range of substrate concentrations
(3–500 lM). Data were used to determine Km/Ki/compet-
itive inhibition using GraphPad Prism software.
16. Crystal structures: preparation of recombinant PTP1B
catalytic domain (1–322) was essentially as described
earlier (see Ref. 21). Crystallization of apo-PTP1B for
soaking experiments, co-crystallization of PTP1B bound
to various inhibitors, X-ray data collection, and structure
solution are described in the Supplementary material.
17. All structures referred to in this article were deposited with
the RCSB (www.rcsb.org/pdb) under PDB-IDs of 2F6T
(3), 2F6Y (8g), 2F6V (8k), 2F6Z (10i), 2F71 (10j), 2F70
(11), and 2F6W (13-Supplementary material).
Simplification of the TIQ ring system by removal of the
3-carboxamide functionality reduced potency as did
removal of the constraining saturated 6-membered ring.
X-ray crystallography of these improved compounds re-
vealed the nature of the ligand–protein interactions and
should provide a stepping stone for further potency
enhancements in this series of compounds.
Supplementary material
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmcl.2005.12.051.
18. Sarmiento, M.; Puius, Y. A.; Vetter, S. W.; Keng, Y.-F.;
Wu, L.; Zhao, Y.; Lawrence, D. S.; Almo, S. C.; Zhang,
Z.-Y. Biochemistry 2000, 39, 8171.
19. Shen, K.; Keng, Y.-F.; Wu, L.; Guo, X.-L.; Lawrence, D.
S.; Zhang, Z.-Y. J. Biol. Chem. 2001, 276, 47311.
20. Manuscript in preparation.
References and notes
21. 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.
1. Cheng, A.; Dube, N.; Gu, F.; Tremblay, M. L. Eur. J.
Biochem. 2002, 269, 1050.