High-throughput discovery of mycobacterium tuberculosis protein tyrosine phosphatase B (MptpB) inhibitors using click chemistry

Article abstract of DOI:10.1021/ol9023419

A ～500-member library of bidentate inhibitors against protein tyrosine phosphatases (PTPs) was rapidly assembled using click chemistry. Subsequent high-throughput screening had led to the discovery of highly potent (K _i as low as 150 nM) and se

Full text of DOI:10.1021/ol9023419

ORGANIC
LETTERS
2009
Vol. 11, No. 22
5102-5105
High-Throughput Discovery of
Mycobacterium tuberculosis Protein
Tyrosine Phosphatase B (MptpB)
Inhibitors Using Click Chemistry
Lay Pheng Tan,† Hao Wu,‡ Peng-Yu Yang,‡ Karunakaran A. Kalesh,‡
Xiaohua Zhang,‡ Mingyu Hu,‡ Rajavel Srinivasan,‡ and Shao Q. Yao*,†,‡
Departments of Biological Sciences and Chemistry, NUS MedChem Program of the
Life Sciences Institute, National UniVersity of Singapore, Singapore 117543
chmyaosq@nus.edu.sg
Received August 5, 2009
ABSTRACT
A ∼3500-member library of bidentate inhibitors against protein tyrosine phosphatases (PTPs) was rapidly assembled using click chemistry.
Subsequent high-throughput screening had led to the discovery of highly potent (Ki as low as 150 nM) and selective MptpB inhibitors, some
of which represent the most potent MptpB inhibitors developed to date.
Protein tyrosine phosphatases (PTPs) are important signaling
enzymes in the cell. Dysregulation of these enzymes has been
linked to various human diseases.1 For example, PTP1B is
one of the best-known PTPs and has been identified as a
key player against diabetes, obesity, and cancer. MptpB,
another well-known protein tyrosine phosphatase encoded
by Mycobacterium tuberculosis, is a promising target for new
tuberculosis (TB) drugs.2 Progress in the development of
potent and selective PTP inhibitors, however, has been slow
because of the highly conserved active site shared by most
members of the PTP family. A significant advance came from
the seminal discovery that there exists a unique secondary
site near the active site of PTP1B (and possibly in many
other PTPs).3 As a result, major drug discovery efforts have
focused on the use of fragment-based approaches for the
identification of the so-called bidentate PTP1B inhibitors
capable of binding to both the active site and the secondary
site of the enzyme simultaneously.4 We recently demon-
strated the first example of using click chemistry followed
by direct in situ screening for the rapid assembly and
† Department of Biological Sciences.
(3) Puius, Y. A.; Sullivan, M.; Lawrence, D. S.; Almo, S. C.; Zhang,
Z.-Y. Proc. Natl. Acad. Sci. U.S.A 1997, 94, 13420–13425.
‡ Department of Chemistry.
(1) (a) Montalibet, J.; Kennedy, B. P. Drug DiscoVery Today: Ther.
Strategies 2005, 2, 129–135. (b) Blume-Jensen, P.; Hunter, T. Nature 2001,
411, 355–365. (c) Cook, W. S.; Unger, R. H. DeV. Cell 2002, 2, 385–387.
(d) Zhang, Z.-Y. Curr. Opin. Chem. Biol. 2001, 5, 416–423. (e) Hunter, T.
Cell 2000, 100, 113–127.
(4) (a) Liu, G.; Xin, Z.; Pei, Z.; Hajduk, P. J.; Abad-Zapatero, C.;
Hutchins, C. W.; Zhao, H.; Lubben, T. H.; Ballaron, S. J.; Haasch, D. L.;
Kaszubska, W.; Ronodinone, C. M.; Trevillyan, J. M.; Jirousek, M. R.
J. Med. Chem. 2003, 46, 4232–4235. (b) Liu, G.; Xin, Z.; Liang, H.; Abad-
Zapatero, C.; Hajduk, P. J.; Janowick, D. A.; Szczepankiewicz, B. G.; Pei,
Z.; Hutchins, C. W.; Ballaron, S. J.; Stashko, M. A.; Lubben, T. H.; Berg,
C. E.; Rondinone, C. M.; Trevillyan, J. M.; Jirousek, M. R. J. Med. Chem.
2003, 46, 3437–3440.
(2) Greenstein, A. E.; Grundner, C.; Echols, N.; Gay, L. M.; Lombana,
T. N.; Miecskowski, C. A.; Pullen, K. E.; Sung, P. Y.; Alber, T. J. Mol.
Microbiol. Biotechnol. 2005, 9, 167–181.
10.1021/ol9023419 CCC: $40.75
Published on Web 10/23/2009
 2009 American Chemical Society

identification of potent bidentate inhibitors against PTP1B.5
This approach takes advantage of the highly efficient and
modular Cu(I)-catalyzed 1,3-dipolar cycloaddition reaction
between an azide and an alkyne6 and has thus far been
successfully applied to the discovery of small molecule
inhibitors against other classes of enzymes.7 In our previous
study, we were able to identify a “click-based” inhibitor that
possesses inhibitory properties against PTP1B similar to
those identified from conventional fragment-based ap-
proaches.5 Herein, we report a ∼3500-member PTP bidentate
inhibitor library synthesized using our collection of alkynes
and our previously synthesized azide library8 and, from the
in situ screening of this library, the discovery of the first
click-based small molecule inhibitor against MptpB (Ki )
150 nM), which not only possesses a high specificity (11-
to 43-fold) over other PTPs but also represents the most
potent inhibitor of MptpB known in the literature.
Our bidentate inhibitors have three components: (i) the
warheads, alkyne-containing N-phenyloxamic acids that
are cell-permeable, potent bioisosteric phosphotyrosine
mimics;4,5 (ii) a variety of different types of building blocks
that act as the secondary-site binders; and (iii) azide-
containing linkers of different lengths joining the warhead
and the building blocks (Figure 1). We anticipated that
interested in finding inhibitors for MptpB as it is a relatively
new target. Recent X-ray structural studies have also revealed
the presence of a unique secondary binding site near the
enzyme active site, a feature analogous to PTP1B.9
Of the 11 warheads used in this library, W2, W3, W5,
W6, W7, and W9 were synthesized as previously reported.5,8
W1, W4, W8, and W10 were newly designed N-phenylox-
imic acid analogues guided by our previous results and
computational modeling.5 The final warhead, W11, was
inspired by the recently discovered MptpB inhibitor from
Ellman and co-workers.10 Detailed chemical syntheses of the
warheads are described in Supporting Information. The
representative synthesis of one of the warheards, W11, is
shown in Scheme 1. The synthesis of the 325-member
Scheme 1. Synthesis of Warhead W11
secondary-site binders, which consist of diverse aromatic and
aliphatic building blocks linked to various azide-containing
linkers, was previously reported.8 With both the alkyne and
the azide libraries in hand, we carried out high-throughput
click assembly of the 11 × 325 compound combinations in
384-well microtiter plates with the aid of an automatic liquid
handling system as previously reported.11 Upon optimiza-
tions, we found click chemistry conditions carried out with
CuSO4 and sodium ascorbate as catalysts and H2O/t-BuOH
as cosolvents gave the best results. Subsequently, ap-
proximately 10% of the ∼3500 compounds obtained were
further characterized by LC-MS and confirmed to be of
sufficient purity (>90% in most cases, see Supporting
(5) Srinivasan, R.; Uttamchandani, M.; Yao, S. Q. Org. Lett. 2006, 8,
713–716.
(6) For reviews, see: (a) Kolb, H. C.; Sharpless, K. B. Drug DiscoVery
Today 2003, 8, 1128–1137. (b) Meldal, M.; Tornoe, C. W. Chem. ReV.
2008, 108, 2952–3015.
Figure 1. Overall strategy of the click-based high-throughput
(7) (a) Kalesh, K. A.; Yang, P.-Y.; Srinivasan, R.; Yao, S. Q. QSAR
Comb. Sci. 2007, 26, 1135–1144. (b) Birk, A.; Wu, C. Y.; Wong, C. H.
Org. Biomol. Chem. 2006, 4, 1446–1457. (c) Krasin˜ski, A.; Radik, Z.;
Manetsch, R.; Raushel, J.; Taylor, P.; Sharpless, K. B.; Kolb, H. C. J. Am.
Chem. Soc. 2005, 127, 6686–6692.
discovery of MptpB inhibitors. A total of 11 alkyne warheads and
325 azides were assembled to generate the ∼3500-member bidentate
library that, upon in situ screening, gave rise to two highly potent
MptpB inhibitors (boxed).
(8) Srinivasan, R.; Tan, L. P.; Wu, H.; Yang, P.-Y.; Kalesh, K. A.; Yao,
S. Q. Org. Biol. Chem. 2009, 7, 1821–1828.
(9) Grundner, C.; Perrin, D.; van Huijsduijnen, R. H.; Swinnen, D.;
Gonzalez, J.; Gee, C. L.; Wells, T. N.; Alber, T. Structure 2007, 15, 499–
509.
combining these components would result in a sizable
bidentate inhibitor library that might be used to target
different PTPs with little structural knowledge of the nature
of their active and secondary sites. We were particularly
(10) Soellner, M. B.; Rawls, K. A.; Grundner, C.; Alber, T.; Ellman,
J. A. J. Am. Chem. Soc. 2007, 129, 9613–9615.
(11) Srinivasan, R.; Li, J.; Ng, S. L.; Kalesh, K. A.; Yao, S. Q. Nat.
Protoc. 2007, 2, 2655–2664.
Org. Lett., Vol. 11, No. 22, 2009
5103

Figure 2. Inhibition profiles of the click library against MptpB. (a) 4-D plots of the inhibition fingerprints from the 2 largest sublibraries
that are assembled using the 11 warheads and the Type I and II azides. Each inhibitor is represented by spheres positioned spatially across
the warheads, linkers, and azides. Relative potency is given by both color spectrum (blue ) least potent, red ) most potent) and by size
of the sphere (big ) more potent, small ) less potent) (b) Bar graph representation of averaged inhibitor potencies across sublibraries
derived from each of the 1 warheads (red bar ) sublibrary derived from W11). An arbitrary threshold of “8%” (dashed line) was set to
distinguish the seven most potent warheads.
Information). They were subsequently used for direct in situ
screening against recombinant MptpB enzyme. It should be
highlighted that the key attribute of our strategy lies in the
ability to rapidly assemble high-quality bidentate inhibitors
in a matter of weeks or even days, thereby alleviating one
of the major challenges in the traditional drug discovery
process. The 3500-member library shown herein represents
the largest click-based library assembled to date.
to 10% of the highest averaged inhibition potency (∼80%)
exhibited by inhibitors derived from W11) was set to select
for the seven most potent warheads, W11, W1, W10, W5,
W8, W2, and W3 (Figure 2b). Next, the contribution of the
linker units and the building blocks were further analyzed
(Figure 3); the Type I and Type II azide-containing subli-
braries were chosen as they gave the most comprehensive
representation of different inhibitor components. Two inter-
esting observations were made from this analysis. First, for
the Type I azides, there was a marked preference for the
4-carbon alkyl linker, indicating an optimal length for fitting
the bidentate inhibitor into the active and secondary sites of
MptpB (Figure 3, Type I and Figure S3 in Supporting
Information). In contrast, inhibitors assembled using the Type
II azides did not displayed a distinct preference for any of
the alkyl linkers (Figure 3, Type II and Figure S3 in
Supporting Information). Second, there were recurring build-
ing blocks that consistently showed high potency regardless
of the linker length among the Type I and II azides (Figure
3, in red). The top-50 inhibitors from each of the seven
sublibraries were next identified, listed in Table S4 in
Supporting Information and analyzed using Venn diagram
(Figure S4 in Supporting Information) to obtain inhibitors
that were potent across all the warheads. A total of 10
individual azides were found to be particularly strong
secondary-site binders. Most of these azides contain hydro-
phobic building blocks (Table S5 in Supporting Information),
with the highest-ranked azide being the recurring one as
shown in Figure 3. Taken together, these observations
indicate the potential of our small molecule library for
evaluating individual components in a bidentate PTP inhibi-
tor.
Upon screening against MptpB, the relative potency of
each inhibitor was obtained and presented as a colored heat
map (Figure S2 in Supporting Information). The inhibition
fingerprints from the largest sublibraries consisting of Type
I and II azides are also presented in 4-D cube plots in Figure
2a. A quick survey of the heat map and the cube plots
revealed that almost all of the inhibitors derived from W11
were more potent than those from other warheads, implying
that W11 plays a dominant role in contributing to inhibitor
potency against MptpB. Interestingly, most inhibitors, es-
pecially those assembled from W1, showed varying degrees
of inhibition potencies across different azides (see Supporting
Information), highlighting the cooperative binding effect from
both the warhead and the azide component in the bidentate
inhibitor. Next, quantitative data analysis was carried out
by averaging the inhibition potency across sublibraries
derived from each warhead, and the results are graphically
presented in Figure 2b. Inhibitors derived from W11 were
found to be 24-fold more potent than those from W6, which
lacks the cyclohexyl group. This is presumably due to the
considerable plasticity of the active site in MptpB;10 the
additional cyclohexyl group may be accommodated to gain
favorable hydrophobic interactions. W7, which has a pyra-
zole ring in place of the isoxazole, was the least potent among
the eleven warheads, indicating that the oxygen atom in the
isoxazole ring is crucial as a hydrogen bond acceptor with
the enzyme, since replacement with a hydrogen bond donor
abolished most of the inhibitory activity. For further subli-
brary analysis, an arbitrary threshold of “8%” (equivalent
Finally, to obtain a better quantitative representation of
the inhibitor potency, IC50 determination was carried out with
∼350 selected compounds (Figure S5 in Supporting Infor-
mation), of which 16 of the most potent inhibitors (structures
shown in Figure S1 in Supporting Information) were scaled-
5104
Org. Lett., Vol. 11, No. 22, 2009

Table 1. IC50 (in µM) and Ki (in µM) of Selected Purified
Inhibitors against MptpB and Other PTPs
selectivitya permeability
PTP/MptpB Papp (nm s-1
)
inhibitor
MptpB IC50 (Ki)
H1-6C-W11 0.55 ( 0.03 (0.17 ( 0.03)
F1S-6C-W11 0.64 ( 0.09 (0.15 ( 0.01)
G4-6C-W11 0.84 ( 0.09 (0.77 ( 0.12)
F7S-6C-W11 0.95 ( 0.08 (0.45 ( 0.01)
E4S-4C-W11 1.3 ( 0.18 (1.1 ( 0.11)
F5S-4C-W11 1.4 ( 0.12 (1.3 ( 0.04)
F7S-4C-W11 1.5 ( 0.28 (1.3 ( 0.15)
E2S-4C-W11 1.6 ( 0.10 (ND)
23-52
11-43
9-43
7-26
6-24
2-19
6-21
2-15
5-18
ND
ND
ND
545
ND
ND
ND
ND
ND
ND
ND
F6S-4C-W11 1.8 ( 0.03 (ND)
W11
3.56 ( 0.04 (1.62 ( 0.04)
a Ratio of IC50 for other PTPs to that for MptpB (see Supporting
Information for details). ND ) not determined.
screened individually against MptpB. While W11 alone was
10-fold less potent than H1-6C-W11 and F1S-6C-W11, none
of the azide components showed any significant inhibition
against MptpB. Finally, the two most potent inhibitors were
computationally docked against the active site of MptpB
using the Sybyl software, on the FlexX suite (Figure S6 in
Supporting Information); results indicate that in both cases
the N-phenyloxamic acid containing warhead binds in the
active site pocket of MptpB, with its azide-containing
secondary binder projecting into the secondary pocket, thus
indicating the possible bidentate nature of our inhibitors.
In conclusion, a ∼3500-member library of bidentate
inhibitors against protein tyrosine phosphatases (PTPs) has
been synthesized, in high throughput, using click chemistry.
This, to the best of our knowledge, constitutes the largest
click chemistry library ever assembled. Subsequent direct
in situ screening led to the identification of highly potent
and selective MptpB inhibitors. The systematic analysis of
the structure-activity relationship made possible by our
inhibitor library against PTPs may shed light on the future
development of other bidentate inhibitors against different
PTPs.
Figure 3. Bar graph representing the averaged inhibition potencies
from inhibitors assembled using the seven warheads and the Type
I and Type II azides. The two recurring building blocks (structures
shown below the bar graph) from each set of sublibraries are
indicated in red.
up, purified, and fully characterized (by LCMS and NMR)
before full biochemical evaluations were carried out against
MptpB and other PTPs. The IC50, Ki values, and cell
permeability of selected potent inhibitors are shown in Table
1 (see Tables S6, S7 and S8 in Supporting Information for
full data). All compounds tested were in general cell-
permeable. The most potent and selective inhibitors, H1-
6C-W11 and F1S-6C-W11 (see structures in Figure 1),
showed Ki values of 170 and 150 nM, respectively, against
MptpB, which are even more potent than the best MptpB
inhibitor known in the literature (Ki ) 220 nM; reported by
Ellman et al.10). More significantly, it showed 11- to 52-
fold increase in selectivity over other PTPs tested. The azide
and warhead components of the purified inhibitors were also
Acknowledgment. Funding was provided by MOE (R143-
000-394-112), BMRC (R143-000-391-305) and CRP (R143-
000-218-281). We thank T. Alber (Berkeley) for the recom-
binant construct of MptpB.
Supporting Information Available: Experimental pro-
cedures, characterization of new compounds, click chemisry,
and biological screening. This material is available free of
charge via the Internet at http://pubs.acs.org.
OL9023419
Org. Lett., Vol. 11, No. 22, 2009
5105