A gatekeeper residue for inhibitor sensitization of protein tyrosine phosphatases

Article abstract of DOI:10.1016/j.bmcl.2006.05.011

Allele-specific enzyme inhibitors are powerful tools in chemical biology. However, few general approaches for the discovery of such inhibitors have been described. Herein is reported a method for the sensitization of protein tyrosine phosphatases (PTPs) to small-molecule inhibition. It is shown that mutation of an active-site isoleucine to alanine (I219A) sensitizes PTP1B to inhibition by a class of thiophene-based inhibitors. This sensitization strategy succeeds for both 'orthogonal' inhibitors, designed to be incompatible with wild-type PTP active sites, and previously optimized wild-type PTP inhibitors. The finding that the I219A mutation sensitizes phosphatase domains to a variety of compounds suggests that isoleucine 219 may act as a 'gatekeeper' residue that can be widely exploited for the chemical-genetic analysis of PTP function.

Full text of DOI:10.1016/j.bmcl.2006.05.011

Bioorganic & Medicinal Chemistry Letters 16 (2006) 4002–4006
A gatekeeper residue for inhibitor sensitization
of protein tyrosine phosphatases
Anthony C. Bishop* and Elizabeth R. Blair
Department of Chemistry, Amherst College, Amherst, MA 01002, USA
Received 11 April 2006; revised 2 May 2006; accepted 3 May 2006
Available online 23 May 2006
Abstract—Allele-specific enzyme inhibitors are powerful tools in chemical biology. However, few general approaches for the discov-
ery of such inhibitors have been described. Herein is reported a method for the sensitization of protein tyrosine phosphatases (PTPs)
to small-molecule inhibition. It is shown that mutation of an active-site isoleucine to alanine (I219A) sensitizes PTP1B to inhibition
by a class of thiophene-based inhibitors. This sensitization strategy succeeds for both ‘orthogonal’ inhibitors, designed to be incom-
patible with wild-type PTP active sites, and previously optimized wild-type PTP inhibitors. The finding that the I219A mutation
sensitizes phosphatase domains to a variety of compounds suggests that isoleucine 219 may act as a ‘gatekeeper’ residue that
can be widely exploited for the chemical–genetic analysis of PTP function.
Ó 2006 Elsevier Ltd. All rights reserved.
Large gene families that encode homologous proteins
represent particularly challenging cases for chemical
biology. This is, in large part, due to the ‘degeneracy’
problem; it is difficult to chemically differentiate between
active (or allosteric) sites that bear a high degree of
structural similarity with one another. One, now well
established, method for circumventing the degeneracy
problem is through the engineering of protein/small-
molecule interfaces—that is, through modification of a
protein receptor’s binding site, in addition to the com-
plementary modification of a potentially selective
small-molecule ligand.1,2
that can readily be identified through protein sequence
alignments.6 Targeted mutation of the gatekeeper resi-
due (to alanine or glycine) has led to the creation of a
general strategy, termed chemical–genetic analysis, for
the study of protein kinase function. Recently, this
method has been used to elucidate the in vivo function
of an impressive array of yeast kinases (e.g., Cdc28,6,7
Ime2,7 Pho85,8 Fus3,6 and Apg19), as well as mammali-
an kinases (e.g., CamKII,10 v-erbB,11 and GRK2.12).
We have recently described the first example of inhibi-
tor-sensitization for another large and important class
of signaling enzymes: the protein tyrosine phosphatases
(PTPs).13,14 The engineering of a prototype phospha-
tase, the type II diabetes drug target PTP1B, was guided
by the following criteria: an amino acid chosen for
mutagenesis must be large enough that substitution by
a small amino acid creates a novel binding pocket; the
mutant PTP must retain catalytic activity; and the resi-
due identified for PTP1B-sensitization must be present
in other PTPs, eliminating the need to redesign the
PTP/inhibitor interface for each target. Valine 49 and
isoleucine 219 (V49 and I219, human PTP1B number-
ing) both meet these criteria.13
One powerful application of protein/small-molecule
engineering is the generation of allele-specific enzyme
inhibitors. Several critical enzyme families in the eukary-
otic proteome, including protein kinases,3 protein meth-
yltransferases,4 and phosphoinositide 3-kinases,5 have
been engineered to possess novel inhibitor sensitivity.
This engineered sensitivity, not present in related wild-
type enzymes, allows for the identification of selective
inhibitors from relatively small panels of putative inhib-
itors. The enzyme-sensitization approach has proven to
be particularly effectual for the protein kinases, as mem-
bers of the kinase family possess a ‘gatekeeper’ residue
In a previous report, we utilized the indolic nitrogen of
an oxalylaminoindole PTP inhibitor (compound 1,
Fig. 1a) as a ‘hook’ to which we attached groups that
were designed to be incompatible with wild-type PTP ac-
tive sites.13 While this approach did yield a 10-fold selec-
Keywords: Protein engineering; Inhibitor sensitization; Protein tyrosine
phosphatases; PTP1B.
*
Corresponding author. Tel.: +1 413 542 8316; fax: +1 413 542
2735; e-mail: acbishop@amherst.edu
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.05.011

A. C. Bishop, E. R. Blair / Bioorg. Med. Chem. Lett. 16 (2006) 4002–4006
4003
Figure 1. (a) Chemical structures of compounds 1–4. (b) Chemical structures and synthetic schemes for compounds 9a and 9b. Reagents and
conditions: (i) H2, AcOH, Pd/C, cat. H2SO4; (ii) ethyl cyanoacetate, sulfur, morpholine, EtOH, reflux; (iii) ethyl oxalyl chloride, THF; (iv) NaOH,
H2O, EtOH.
tive inhibitor of a mutated PTP1B (I219A), it suffered
from the inherent lack of potency of oxalylaminoindole
PTP inhibitors at physiological pH.13 Moreover, the in-
dole-derivative inhibitors showed little to no selectivity
for mutants of valine 49, a position that is particularly
attractive for a general PTP inhibitor-sensitization strat-
egy, as it is occupied by valine or isoleucine in 35 of the
37 classical PTP catalytic domains in humans. (Position
219 is occupied by either valine or isoleucine in 28 clas-
sical PTP domains.15)
hydrobenzo[b]thiophene-3-carboxylic acid (compound
2, Fig. 1a).16 Previous reports have shown that com-
pound 2 effectively inhibits several PTPs, including
PTP1B and its closest homolog, TCPTP, as well as
phosphatases that are not especially homologous to
PTP1B (e.g., PTPH1).17–20 The three-dimensional struc-
ture of compound 2 bound to a PTP has not been
reported; however, a crystal structure of a slightly less
potent PTP inhibitor that differs from 2 only by the
replacement of a methylene group with an oxygen atom
(3, Fig. 1a) is known.16 Surmising that compounds 2 and
3 bind to PTP active sites in identical orientations, we
inspected the PTP1B/3 structure for positions that could
As a scaffold for a second generation of allele-specific
PTP inhibitors, we selected 2-oxalylamino-4,5,6,7-tetra-

4004
A. C. Bishop, E. R. Blair / Bioorg. Med. Chem. Lett. 16 (2006) 4002–4006
be derivatized to target PTPs sensitized by mutations at
valine 49 or isoleucine 219 (Fig. 2). Carbons 5 and 7 (C-
˚
5 and C-7) lie less than 5 A away from the V49 and I219
side chains, respectively (Fig. 2). Based on this analysis,
we hypothesized that analogs of 2, derivatized at C-5
and/or C-7, may selectively target V49A and/or I219A
PTP1B. In addition, extension of chemical groups ‘up’
and ‘down’ from the sp3-hybridized carbons on the
puckered six-membered ring of 2 appeared to hold
promise for targeting the novel binding pocket of a
V49A mutant, since V49 lies beneath the fused ring sys-
tem. This analysis is consistent with previous findings
that V49 mutants are not effectively targeted by flat in-
dole-based inhibitors; in these inhibitors the chemical
appendages extend ‘sideways’ from an sp2-hybridized
nitrogen atom, missing the engineered pocket of V49
mutants.13
Figure 3. Selective inhibition of engineered PTP1B mutants by
compounds 2, 9a, and 9b. The indicated compounds (20 lM) were
incubated with wild-type (background), V49A (middle), or I219A
(foreground) PTP1B, and para-nitrophenyl phosphate (concentration
corresponding to the KM for the particular enzyme at pH 7.0: 2.08,
6.57, and 3.09 mM, respectively). Percent PTP1B activities in the
presence of the inhibitors, normalized to a no-inhibitor (vehicle)
control, are shown as bars. The bars represent the average value from
three independent experiments, with relative standard deviations of
less than 4%.
We synthesized two ‘bumped’ inhibitors, compounds 9a
and 9b, which were designed to exploit the novel binding
pockets of the I219A and/or V49A mutants (Fig. 1b, see
Supplementary data for synthetic details). These com-
pounds contain dimethyl moieties at either C-5 (9a) or
both the C-5 and C-7 positions (9b). (Due to synthetic
difficulties, analogs with bumps exclusively at C-7 were
not prepared.)
parent ligand, reduced potency with respect to the wild-
type enzyme and comparable (V49A) or increased
(I219A) potency with the sensitized enzymes (Fig. 3).
To more accurately determine the level of selectivity
for the sensitized PTP1B mutants, the inhibitory con-
stant (KI) values for 9b with the wild-type and mutant
enzymes were determined (Table 1). Compound 9b is
truly orthogonal with respect to wild-type PTP1B
(KI > 300 lM), and it demonstrates >10-fold selectivity
for V49A PTP1B (KI = 27 lM) and >30-fold selectivity
for I219A PTP1B (KI = 10 lM).
Using the artificial PTP substrate, para-nitrophenyl
phosphate (p-NPP), compounds 9a and 9b, along with
the parent molecule 2, were assayed for selective inhibi-
tion of PTP1B activity at pH 7.0. Figure 3 shows the re-
sults of these inhibition screens for the GST-fusion
proteins of wild-type PTP1B, V49A PTP1B, and
I219A PTP1B. The C-5-derivatized analog, 9a, shows
an inhibitory profile that is essentially unchanged from
the parent molecule (Fig. 3). Presumably the dimethyl
moiety at C-5 simply extends out of the active site. Thus,
9a was not investigated further. By contrast, the tetra-
methyl derivative, 9b, shows the classical inhibitory pro-
file for an ‘orthogonal’ design strategy: compared to the
Compound 9b is the first known compound to demon-
strate substantial allele-specific character with a V49
PTP mutant. Since compounds 2 and 9a show no al-
lele-specific affinity for V49A PTP1B, it is clear that
derivatization at C-7 is necessary for successful V49A
targeting. These data may represent a useful starting
place for future panels of more V49-directed inhibitors.
However, the tetrahydrobenzothiophene scaffold will
probably not be ideal for such an approach: V49A
PTP1B is less potently inhibited by the parent molecule,
2, than is wild-type PTP1B, suggesting that the V49 side
chain makes important contacts with the six-membered
ring of 2 that are lost upon mutation of V49 to alanine.
Surprisingly, strong I219A PTP1B-specific character
was also observed for the parent molecule, 2 (Fig. 3
Table 1. Inhibition constants for compounds 2, 9b, and 4 on wild-type
and sensitized PTP1B enzymes
Figure 2. Surface representation of compound 3 bound to PTP1B
(PDB: 1C87).16 The PTP1B protein surface is shown in magenta, with
the portions of the surface comprising Val49 (green), Ile219 (cyan)
highlighted. Compound 3 is colored by element: gray for carbon, red
for oxygen, blue for nitrogen, and yellow for sulfur. The arrows
indicate the presumptive positions of C-5 and C-7 in the complex of
compound 2 with PTP1B.
Compound Wild-type
V49A
I219A
PTP1B KI (lM) PTP1B KI (lM) PTP1B KI (lM)
2
34 1.3
>300
6.0 0.34
ND
27 1.8
ND
1.0 0.089
10 0.64
0.23 0.0046
9b
4

A. C. Bishop, E. R. Blair / Bioorg. Med. Chem. Lett. 16 (2006) 4002–4006
4005
and Table 1). Compound 2 is, approximately, a 30-fold
more potent inhibitor of I219A PTP1B than wild-type
PTP1B (wild-type PTP1B: KI = 34 lM; I219 PTP1B:
KI = 1.0 lM). Apparently, the I219 side chain of wild-
type PTP1B precludes the optimal binding orientation
of the fused thiophene-based inhibitors—regardless of
whether or not the six-membered rings contain bumps
that are added in an orthogonal approach. Moreover,
it appears that the 219 position may act as a gatekeeper
in PTPs: I219 controls access of inhibitors to the active
site, whether or not the compounds are concertedly de-
signed as ‘allele-specific.’
To further test the idea that I219 may serve as a gate-
keeper for many inhibitors, even potent wild-type inhib-
itors, compound 4 (Fig. 1a) was synthesized. It has been
shown previously that the substitution of 2’s C-6 meth-
ylene with nitrogen leads to increased potency for
PTP1B and TCPTP.16 A salt-bridge interaction between
the introduced amino group and aspartate 48 (D48)
increases the potency of 4 with respect to the closely
related 2 and 3, giving rise to substantial selectivity for
D48-containing PTPs (e.g., PTP1B and TCPTP) over
PTPs that contain other amino acids at this position
(Fig. 4a).16 We hypothesized that the potency and selec-
tivity of this ‘optimized’ PTP inhibitor could be in-
creased even further by sensitization of a target PTP.
Indeed, compound 4 inhibits I219A PTP1B approxi-
mately 25 times more potently (KI = 0.23 lM) than it
does wild-type PTP1B (Table 1). The 4/I219A PTP1B
KI value of 230 nM is remarkably low for an active-
site-directed PTP inhibitor with a molecular weight of
less than 300 g/mol.21 Moreover, at neutral pH, 4 is
roughly 200-fold more potent than previously identified
allele-specific PTP inhibitors.13 Importantly, this height-
ened potency also gives rise to substantial selectivity not
observed with wild-type PTPs. For example, 4 is only 2-
fold selective for PTP1B with respect to PTP1B’s closest
homolog, TCPTP (KI = 10 0.74 lM)—a common and
vexing problem among PTP1B inhibitors.21 By contrast,
4 is 45-fold selective for I219A PTP1B over TCPTP, sug-
gesting that 4 could be used in signaling studies with
I219A PTP1B-expressing cells (or lysates) to specifically
target PTP1B activity. Additionally, the gatekeeper
functionality of position 219 is not limited to PTP1B.
Introduction of the corresponding mutation in TCPTP
switches the selectivity, rendering I220A TCPTP
(KI = 0.71 0.030 lM) 8-fold more susceptible to inhi-
bition by 4 than wild-type PTP1B (KI = 6.0 lM).
Figure 4. (a) Interaction of compound 4 with aspartate 48 in PTP1B
(PDB: 1C88).16 The PTP1B backbone is shown as a ribbon diagram in
magenta. The Asp48 and Ile219 side chains and compound 4 are
colored by element as in Figure 2. (b and c) Selective inhibition of
I219A PTP1B in a complex proteome mixture. Crude cell lysate from
wild-type PTP1B- (b) or I219A PTP1B-overexpressing (c) Escherichia
coli was incubated with DMSO (vehicle, filled circles) or 1 lM 4 (open
circles), and 0.5 mM p-NPP. Phosphatase activity was measured by
monitoring the time-dependent increase in absorbance at 405 nm.
To test whether the enhanced potency of 4 could be used
to selectively target sensitized PTPs in the context of
complex proteomes, the activities of wild-type PTP1B-
and I219A PTP1B-containing crude Escherichia coli ly-
sates were assayed. Indeed, the selectivity of 4 in assays
with purified enzyme translates directly to the lysate-cat-
alyzed reaction. At 1 lM 4, only slight inhibition of
wild-type PTP1B is observed (Fig. 4b). By contrast, al-
most complete suppression of activity is observed in
reactions containing I219A PTP1B lysate (Fig. 4c).
Since wild-type PTP1B is (slightly) more susceptible to
inhibition by 4 than is TCPTP, these experiments also
suggest that the activity of a sensitized PTP1B could
be ablated in a mammalian proteome without signifi-
cantly disrupting TCPTP activity. Moreover, since 4
was previously optimized for TCPTP/PTP1B inhibition,
the affinity of 4 for I219A PTP1B should be far greater
than for any other PTP tested to date.16
The clear advantage of a successful allele-specific inhibi-
tor approach, over medicinal chemistry, is in its generali-
ty: specificity for a given target can be straightforwardly
engineered instead of ‘hunted’ for. However, the identifi-

4006
A. C. Bishop, E. R. Blair / Bioorg. Med. Chem. Lett. 16 (2006) 4002–4006
3. Bishop, A. C.; Shah, K.; Liu, Y.; Witucki, L.; Kung, C.;
Shokat, K. M. Curr. Biol. 1998, 8, 257.
4. Lin, Q.; Jiang, F.; Schultz, P. G.; Gray, N. S. J. Am.
Chem. Soc. 2001, 123, 11608.
5. Alaimo, P.; Knight, Z. A.; Shokat, K. M. Bioorg. Med.
Chem. 2005, 13, 2825.
6. Bishop, A. C.; Ubersax, J. A.; Petsch, D. T.; Matheos, D.
P.; Gray, N. S.; Blethrow, J.; Shimizu, E.; Tsien, J. Z.;
Schultz, P. G.; Rose, M. D.; Wood, J. L.; Morgan, D. O.;
Shokat, K. M. Nature 2000, 407, 395.
cation of gatekeeper residues in an enzyme family is a key
prerequisite for a widely applicable allele-specific ap-
proach. Our data show that I219 is an important gate-
keeper for PTP inhibition. While the degree of
sensitization varies from compound to compound,
I219A mutant PTPs have proven to contain novel inhibi-
tor sensitivity for many members of the two inhibitor fam-
ilies tested to date.13,14 The current findings show that
engineering of the gatekeeper residue can improve the
potency and selectivity even of known compounds,
heightening the utility of this approach, and possibly ren-
dering synthesis of new compounds unnecessary for al-
lele-specific PTP-inhibitor discovery. Our data,
combined with recent findings of mutations that confer
novel inhibitor resistance to PTPs,22 will allow for the
sculpting of PTPs with tunable inhibitor sensitivities. This
combination of genetic manipulation and small-molecule
control of PTP activity will facilitate the chemical–genetic
analysis of PTP function in cells and model organisms.
7. Benjamin, K. R.; Zhang, C.; Shokat, K. M.; Herskowitz,
I. Genes Dev. 2003, 17, 1524.
8. Carroll, A. S.; Bishop, A. C.; DeRisi, J. L.; Shokat, K. M.;
O’Shea, E. K. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 12578.
9. Abeliovich, H.; Zhang, C.; Dunn, W. A.; Shokat, K. M.;
Klionsky, D. J. Mol. Biol. Cell 2003, 14, 477.
10. Wang, H.; Shimizu, E.; Tang, Y. P.; Cho, M.; Kyin, M.;
Zuo, W.; Robinson, D. A.; Alaimo, P. J.; Zhang, C.;
Morimoto, H.; Zhuo, M.; Feng, R.; Shokat, K. M.; Tsien,
J. Z. Proc. Natl. Acad. Sci. U.S.A. 2003, 4287.
11. Fan, Q. W.; Zhang, C.; Shokat, K. M.; Weiss, W. A. Curr.
Biol. 2002, 12, 1386.
12. Kenski, D. M.; Zhang, C.; von Zastrow, M.; Shokat, K.
M. J. Biol. Chem. 2005, 280, 35051.
Acknowledgments
13. Hoffman, H. E.; Blair, E. R.; Johndrow, J. E.; Bishop, A.
C. J. Am. Chem. Soc. 2005, 127, 2824.
14. Blair, E. R.; Hoffman, H. E.; Bishop, A. C. Bioorg. Med.
Chem. 2006, 14, 464.
15. Andersen, J. N.; Mortensen, O. H.; Peters, G. H.; Drake,
P. G.; Iversen, L. F.; Olsen, O. H.; Jansen, P. G.;
Andersen, H. S.; Tonks, N. K.; Moller, N. P. Mol. Cell.
Biol. 2001, 21, 7117.
16. Iversen, L. F.; Andersen, H. S.; Branner, S.; Mortensen, S.
B.; Peters, G. H.; Norris, K.; Olsen, O. H.; Jeppesen, C.
B.; Lundt, B. F.; Ripka, W.; Moller, K. B.; Moller, N. P.
J. Biol. Chem. 2000, 275, 10300.
17. Iversen, L. F.; Andersen, H. S.; Moller, 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.;
Hundahl Moller, N. P. Biochemistry 2001, 40, 14812.
18. Andersen, H. S.; Olsen, O. H.; Iversen, L. F.; Sorensen, A.
L.; Mortensen, S. B.; Christensen, M. S.; Branner, S.;
Hansen, T. K.; Lau, J. F.; Jeppesen, L.; Moran, E. J.; Su,
J.; Bakir, F.; Judge, L.; Shahbaz, M.; Collins, T.; Vo, T.;
Newman, M.; Ripka, W. C.; Moller, N. P. J. Med. Chem.
2002, 45, 4443.
This research was supported by awards from the Ca-
mille and Henry Dreyfus Foundation (SU-02-001) and
from the National Institutes of Health (1 R15
GM071388-01A1). The molecular-graphics images were
produced using the UCSF Chimera package from the
Computer Graphics Laboratory, University of Califor-
nia, San Francisco (supported by NIH P41 RR-
01081). Mass spectral data were obtained at the Univer-
sity of Massachusetts Mass Spectrometry Facility,
which is supported, in part, by the National Science
Foundation.
Supplementary data
Complete description of experimental protocols, syn-
thetic procedures, and new-compound characterization.
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.bmcl.2006.
05.011.
19. Pedersen, A. K.; Guo, X. L.; Moller, K. B.; Peters, G. H.;
Andersen, H. S.; Kastrup, J. S.; Mortensen, S. B.; Iversen,
L. F.; Zhang, Z. Y.; Moller, N. P. Biochem. J. 2004, 378,
421.
References and notes
20. Drake, P. G.; Peters, G. H.; Andersen, H. S.; Hendriks,
W.; Moller, N. P. Biochem. J. 2003, 373, 393.
21. Bialy, L.; Waldmann, H. Angew. Chem., Int. Ed. 2005, 44,
3814.
22. Montalibet, J.; Skorey, K.; McKay, D.; Scapin, G.;
Asante-Appiah, E.; Kennedy, B. P. J. Biol. Chem. 2006,
281, 5258.
1. Bishop, A.; Buzko, O.; Heyeck-Dumas, S.; Jung, I.;
Kraybill, B.; Liu, Y.; Shah, K.; Ulrich, S.; Witucki, L.;
Yang, F.; Zhang, C.; Shokat, K. M. Annu. Rev. Biophys.
Biomol. Struct. 2000, 29, 577.
2. Koh, J. T.; Biggins, J. B. Curr. Top. Med. Chem. 2005, 5,
413.