Rapid assembly of matrix metalloprotease inhibitors using click chemistry

Article abstract of DOI:10.1021/ol061431a

A panel of 96 metalloprotease inhibitors was assembled using "click chemistry" by reacting eight zinc-binding hydroxamate warheads with 12 azide building blocks. Screens of the bidentate compounds against representative metalloproteases provided discerning inhibition fingerprints, revealing compounds with low micromolar potency against MMP-7. The relative ease and convenience of the strategy in constructing focused chemical libraries for rapid in situ screening of MMPs is thereby demonstrated.

Full text of DOI:10.1021/ol061431a

ORGANIC
LETTERS
2006
Vol. 8, No. 17
3821-3824
Rapid Assembly of Matrix
Metalloprotease Inhibitors Using Click
Chemistry
Jun Wang,† Mahesh Uttamchandani,‡ Junqi Li,† Mingyu Hu,† and Shao Q. Yao*,†,‡
Departments of Chemistry and Biological Sciences,
Medicinal Chemistry Program of the Office of Life Sciences,
National UniVersity of Singapore, 3 Science DriVe 3, Singapore 117543,
Republic of Singapore
chmyaosq@nus.edu.sg
Received June 11, 2006
ABSTRACT
A panel of 96 metalloprotease inhibitors was assembled using “click chemistry” by reacting eight zinc-binding hydroxamate warheads with
12 azide building blocks. Screens of the bidentate compounds against representative metalloproteases provided discerning inhibition fingerprints,
revealing compounds with low micromolar potency against MMP-7. The relative ease and convenience of the strategy in constructing focused
chemical libraries for rapid in situ screening of MMPs is thereby demonstrated.
Matrix metalloproteases (MMPs) are a family of zinc-
containing metalloproteases which play critical roles in a
variety of physiological processes. Tight regulation of these
enzymes is essential in maintaining normal cellular function
and development. As one of the most important classes of
therapeutic targets, MMPs are responsible for a variety of
human diseases, including arthritis, Alzheimer’s disease,
cancer, and heart diseases.1 One of the most widely exploited
scaffolds of MMP inhibitors is the peptide-based succinyl
hydroxamate.2 Inhibitors containing this kind of zinc-binding
groups normally exhibit broad-spectrum inhibition toward
most metalloproteases, rather than MMPs alone. This is
largely due to the structural similarity in the active sites of
metalloproteases which possess highly conserved zinc-
binding residues essential for enzyme catalysis. Therefore,
it has been an ongoing challenge to develop highly efficient
synthetic strategies that allow rapid generation and screening
of small molecule inhibitors possessing not only high potency
but more importantly good selectivity toward MMPs.1
Fragment-based assembly, which allows medicinal chem-
ists to explore N2 possibilities with N+N combinations, is
widely employed for high-throughput drug discovery.3 It
typically involves a two-step process including fragment
identification and fragment linkage. A number of methods
have been developed to assist in this process,4 including
NMR/X-ray approaches,4a the MS-based tethering strategy,4b
and “click chemistry” coupled with in situ screening.4c
Among them, approaches based on “click chemistry” 5 were
(3) Arkin, M. R.; Wells, J. A. Nat. ReV. Drug Disc. 2004, 3, 301-317.
(4) (a) Szczepankiewicz, B. G. et al. J. Am. Chem. Soc. 2003, 125, 4087-
4096 and references therein. (b) Erlanson, D. A.; Braisted, A. C.; Raphael,
D. R.; Randal, M.; Stroud, R. M.; Gordon, E. M.; Wells, J. Proc. Natl.
Acad. Sci. U.S.A. 2000, 97, 9367-9372. (c) Birk, A.; Wu, C.-Y.; Wong,
C.-H. Org. Biomol. Chem. 2006, 4, 1446-1457.
† Department of Chemistry.
‡ Department of Biological Sciences.
(1) Overall, C. M.; Kleifeld, O. Nat. ReV. Cancer 2006, 6, 227-239.
(2) Whittaker, M.; Floyd, C. D.; Brown, P.; Gearing, A. J. H. Chem.
ReV. 1999, 99, 2735-2776.
(5) For a recent review on “click chemistry”, see: Kolb, H. C.; Sharpless,
K. B. Drug Disc. Today 2003, 8, 1128-1137.
10.1021/ol061431a CCC: $33.50
© 2006 American Chemical Society
Published on Web 07/27/2006

shown to be a highly versatile and effective choice for rapid
synthesis and identification of inhibitors against a number
of biological targets, including HIV protease,6a,b sulfotrans-
ferases,6c fucosyltransferases,6d protein tyrosine phosphatases,6e
acetylcholinesterase,6f and others.6g However, it remains to
be seen whether this approach is applicable against other
important therapeutic targets, e.g., matrix metalloproteases.
Herein, we describe, for the first time, a “click chemistry”
approach for the rapid synthesis/assembly of a small mole-
cule library based on different succinyl hydroxamates, and
subsequent in situ screening for identification of candidate
hits which possess moderate inhibitory activity against MMPs
over other metalloproteases. Our approach thus lays the
foundation for future exploration of more potent and selective
MMP inhibitors, in high throughput, using “click chemistry”.
Scheme 1. Synthesis of Eight Alkyne-Containing Warheads
Our library design was based on the general structure of
hydroxamate-based MMP inhibitors (Figure 1). With this
′
tates the projection of the hydrophobic moiety into the P4
binding pocket of a targeted MMP, thereby improving both
potency and specificity of the resulting inhibitors.
When devising a suitable route for the warhead synthesis,
two key criteria were considered: (1) the strategy has to be
general, enabling the facile incorporation of a wide variety of
P1′ side chains and, when necessary, with precise stereospe-
cific control of the chiral center located in the warhead; (2) the
resulting warhead must be compatible with standard solid-
phase chemistry, allowing future large-scale library synthesis.
Both criteria were fulfilled by the method shown in Scheme
1, recently developed by us for the synthesis of 18B (a
hydroxamate warhead bearing an isobutyl side chain).7 The
strategy is based on the well-established enolate chemistry
with Evan’s oxazolidinone auxiliary.8 The hydroxamate was
protected with a trityl group, ensuring its compatibility with
standard Fmoc peptide chemistry/TFA cleavage procedures.
Figure 1. Structure of (top) general hydroxamate inhibitors and
(bottom) “click chemistry” inhibitors reported herein against MMPs.
class of inhibitors, it was previously shown that (1) hydro-
phobic P1′ residues are in general preferred, (2) a variety of
substitutions are tolerated at P2′ and P3′ positions, and (3)
hydrophobic P4′ residues are preferred and sometimes could
confer a good degree of specificity among different MMPs.2
As such, a total of eight succinyl hydroxamates bearing a
variety of alkyl, cycloalkyl, and aromatic side chains were
synthesized as shown in Scheme 1. A common alkyne handle
was tethered to each warhead, facilitating the subsequent
assembly with twelve different azides (Scheme 2) using
“click chemistry”. The azides bear a hydrophobic moiety
connected via a linker with a varying alkyl length (n ) 2-4).
By changing the linker length of the azides, our design facili-
Scheme 2. Synthesis of 12 Azides
(6) (a) Birk, A.; Lin, Y.-C.; Elder, J.; Wong, C.-H. Chem. Biol. 2002, 9,
891-896. (b) Whiting, M.; Muldoon, J.; Lin, Y.-C.; Silverman, S. M.;
Lindstrom, W.; Olson, A. J.; Kolb, H. C.; Finn, M. G.; Sharpless, K. B.;
Elder, J. H.; Fokin, V. V. Angew. Chem., Intl. Ed. 2006, 45, 1435-1439.
(c) Best, M. D.; Birk, A.; Chapman, E.; Lee, L. V.; Cheng, W.-C.; Wong,
C.-H. ChemBioChem 2004, 5, 811-819. (d) Lee, L. V.; Mitchell, M. L.;
Huang, S.-J.; Fokin, V. V.; Sharpless, K. B.; Wong, C.-H. J. Am. Chem.
Soc. 2003, 125, 9588-9589. (e) Srinivasan, R.; Uttamchandani, M.; Yao,
S. Q. Org. Lett. 2006, 8, 713-716. (f) Manetsch, R.; Krasinski, A.; Radic,
Z.; Raushel, J.; Taylor, P.; Sharpless, K. B.; Kolb, H. C. J. Am. Chem. Soc.
2004, 126, 12809-12818. (g) Pagliai, F.; Pirali, T.; Del Grosso, E.; Di
Brisco, R.; Tron, G. C.; Sorba, G.; Genazzani, A. A. J. Med. Chem. 2006,
49, 467-470.
3822
Org. Lett., Vol. 8, No. 17, 2006

Overkleeft and co-workers have independently reported a
similar method based on the synthesis of N-Boc-O-TBS-hy-
droxamates.9 Two additional steps were performed to intro-
duce an alkyne handle in the warhead for use in the subse-
quent “click chemistry”. As shown in Scheme 1, a total of
eight different hydrophobic warheads were synthesized (A-H).
Briefly, a suitable carboxylic acid, 13, was converted to
14 with oxazolidinone, followed by enolate chemistry to
introduce the succinyl template 15. In our case, a racemic
mixture of 15 was synthesized in order to save cost. If
desired, the same chemistry could be directly applied for
the synthesis of chiral warheads using the corresponding
chiral oxazolidinone. Next, selective removal of the tert-
butyl group under acidic conditions (to give 16), followed
by coupling of trityl-protected hydroxylamine (to give 17)
and base hydrolysis with H2O2/LiOH, gave 18A-H. Sub-
sequently, 18A-H were coupled to propagylamine using the
HATU/DIEA coupling method, giving 19A-H in excellent
yields. Final deprotection of the trityl group with TFA
furnished the final warheads, A-G, which were unambigu-
ously characterized by LC-MS and 1H and 13C NMR
(Supporting Information). All steps shown in Scheme 1 gave
good to excellent yields, except the last step, where TFA
was used to remove the trityl group. Upon closer examina-
tion, a prominent side product was consistently observed.
Three of these side products were isolated, further character-
ized by NMR and ESI-MS, and confirmed to be the cyclic
adducts 20-22 (Figure 2). They are presumably generated
2-bromoacetyl chloride, 3-bromopropionyl chloride, or 5-bro-
mopentanoyl chloride, followed by an SN2 substitution
reaction with sodium azide in DMF to generate the corre-
sponding azides in excellent yields.
Next, a 96-member inhibitor library was assembled using
“click chemistry” in a 96-deep well block. Each of the eight
alkyne warheads was mixed with each of the twelve azides
(in slight excess; see the Supporting Information for details)
in a t-BuOH/H2O solution, followed by addition of catalytic
amounts of sodium ascorbate and CuSO4. The “click
chemistry” proceeded with high efficiency at room temper-
ature for >12 h. LC-MS confirmed, in almost all cases,
the complete consumption of the alkynes and quantitative
formation of the desired triazole products, thus ensuring that
they may be used directly for subsequent in situ enzymatic
screening without any further purification.
MMP-7 is one of the few MMPs that is secreted by cancer
cells and contributes to proliferation of intestinal adenomas
as well as pancreatic cancer.1 It was chosen for this study,
together with collagenase and thermolysin that have roles
in the progression bacterial corneal keratitis and the me-
tabolism of Bacillus sp., respectively.10 All three enzymes
were screened in a high-throughput, automated fashion
against the 96-member library panel using standard fluores-
cence assays in microplates.6 The inhibitor potency was
evaluated from the reduction in fluorescence output when
introduced in standard enzymatic assays with quenched
substrates. The resulting inhibition fingerprints obtained are
displayed in Figure 3a, demonstrating how such enzymes
Figure 2. Proposed mechanism of the side reaction during TFA
deprotection of warheads.
Figure 3. (a) Inhibitor fingerprints of (I) MMP-7, (II) thermolysin,
and (III) collagenase, represented as “barcodes”. Black: minimum
inhibition; Red: maximum inhibition. (b) Screening of “clicked”
inhibitors against MMP-7. Heat map obtained using TreeView
displays the inhibition fingerprint, with most potent inhibitors
indicated in bright red.
from the acid-catalyzed cyclization reaction of the warheads
as proposed in Figure 2. Nevertheless, moderate yields (40-
55%) were routinely obtained in this step.
The 12-member azide library was synthesized via a highly
efficient two-step procedures (Scheme 2). Twelve different
amines, each bearing a different hydrophobic moiety with
varied functional groups, were first acylated with either
may be easily discerned through their unique inhibition
“barcodes”, which differ from previous fingerprint profiles
generated by other methods11 in that our current method is
able to directly reflect an enzyme’s inhibition profiles (poten-
cy and selectivity). By taking advantage of such inhibition fin-
gerprints, it would become possible to characterize and group
(7) Leeuwenburgh, M. A.; Geurink, P. P.; Klein, T.; Kauffman, H. F.;
van der Marel, G. A.; Bischoff, R.; Overkleeft, H. S. Org. Lett. 2006, 8,
1705-1708.
(8) Wang, J.; Uttamchandani, M.; Sun, L. P.; Yao, S. Q. Chem. Commun.
2006, 717-719.
(10) (a) Eijsink, V. G.; Veltman, O. R.; Aukema, W.; Vriend, G.;
Venema, G. Nat. Struct. Biol. 1995, 2, 374-379. (b) Scozzafava, A.;
Supuran, C. T. Bioorg. Med. Chem. 2000, 8, 637-645.
(9) Evans, D. A.; Bartoli, J.; Shih, T. L. J. Am. Chem. Soc. 1981, 103,
2127-2129.
Org. Lett., Vol. 8, No. 17, 2006
3823

proteins in an activity-dependent manner. Furthermore, such
analysis may prove useful in distinguishing not only inhibi-
tors that are potent against specific enzyme classes, but also
those that provide good discriminatory potential, thereby
minimizing off-target effects of potential drug candidates.
A broader evaluation of the inhibition profile obtained
against MMP-7 (as shown in Figure 3b) revealed a unique and
consistent trend. Hydroxamate warheads containing alkyl and
cycloalkyl side chains at the P1′ position (i.e., isobutyl,
cyclohexyl and cyclopentyl in B, F, and G, respectively) con-
tributed highly to potency of inhibitors against MMP-7. The
strongest inhibition was observed for scaffolds containing the
latter two unnatural cyclic analogues. This correlates well with
generic inhibitors against metalloproteases such as GM6001
and batimastat that are designed with a small hydrophobic
residue (Leu), i.e., isobutyl in the P1′ pocket.12 Furthermore,
both thermolysin and collagenase (profiles provided in the
Supporting Information) do not display such an exclusive and
distinctive preference for these small cyclic residues. Our re-
sults thus highlight the potential of such pharmacophores in
the design of novel inhibitors against MMP-7 and possibly
other MMPs. Another observation is that, azides having a
4-carbon linker (n ) 4; F8-11, G8-11 in Figure 3b) appear-
ed to contribute negatively to the inhibitor potency, which in-
dicates the linker might be too long to properly project the
hydrophobic moiety of the azides into P4′ pocket of MMP-7.
To unambiguosly confirm the potency of these scaffolds
from the preliminary screen, two compounds,F5 and G6,
representing each of the cyclohexyl and cyclopentyl warheads
were selected for detailed evaluations. These molecules were
purified and fully characterized by NMR and LC-MS,
before being evaluated against the panel of metalloproteases
to elucidate the relevant inhibition constants. As shown in
Table 1, F5 and G6 indeed inhibited MMP-7 strongly with
We docked G6, the most potent inhibitor identified from
our screening, against the MMP-7 active site using the Sybyl
software, on the FlexX suite.13 As would be expected, the
optimized docking configuration of the inhibitor/enzyme
complex shows the inhibitor adopts an extended conforma-
tion, fitting comfortably in the enzyme active site (Figure
4). The hydroxamate group from the inhibitor was shown to
Figure 4. In silico docking displays the possible binding mode of
G6/MMP-7 complex. The hydroxamic group in the inhibitor
chelates with the zinc atom (green sphere) in the enzyme active
site, with the inhibitor projecting into the S′ pockets of the enzyme.
Docking was performed using Sybyl v7.2 (Tripos, MO) with
electrostatic surface images generated using WebLab ViewerLite
(Accelrys, San Diego, CA).
chelate to the target zinc atom. The cyclopentyl P1′ side chain
also fits nicely into the S1′ pocket. As we were hoping in
the original design, the phenolic ring from the azide
component appeared to sit in the S4′ pocket as well. In
addition, a number of favorable hydrogen bonds were also
evident in the complex (Supporting Information).
In conclusion, we have developed a high-throughput
strategy for assembling and screening MMP inhibitors. The
panel of compounds we have synthesized show promise as
a valuable resource for screening various MMPs and obtain-
ing unique inhibition fingerprints. Moreover, we have
identified specific scaffolds that show good potency and
moderate selectivity for MMP-7 over other metalloproteases.
We anticipate that the pharmacophore we have uncovered
in this study may lead to future development of more potent
and selective inhibitors against other MMPs. We are thus
extending our experiments against a wider panel of MMPs
to confirm the selectivity patterns obtained against MMP-7
and these results will be reported in due course.
Table 1. IC50 (in µM) and Ki (in µM) of Selected Inhibitors
Acknowledgment. Funding support was provided by
National University of Singapore (NUS) and the Agency for
Science, Technology and Research (A*STAR) of Singapore.
Supporting Information Available: Experimental details
and characterizations of compounds. This material is avail-
able free of charge via the Internet at http://pubs.acs.org.
MMP-7
IC50
thermolysin
IC50
collagenase
IC50
Ki
F5
G6
3.8
1.4
12.5
6.5
>50
>50
>50
>50
OL061431A
(11) (a) Reddy, M. M.; Kodadek, T. Proc. Natl. Acad. Sci. U.S.A. 2005,
102, 12672-12677. (b) Takahashi, M.; Nokihara, K.; Mihara, H. Chem.
Biol. 2003, 10, 53-60. (c) Chan, E. W. S.; Chattopadhaya, S.; Panicker,
R. C.; Huang, X.; Yao, S. Q. J. Am. Chem. Soc. 2004, 126, 14435-14446.
(d) Srinivasan, R.; Huang, X.; Ng, S. L.; Yao, S. Q. ChemBioChem 2006,
7, 32-36.
(12) Rao, B. G. Curr. Pharm. Des. 2005, 11, 295-322.
(13) Kramer, B.; Rarey, M.; Lengauer, T.; Klebe, G. J. Mol. Biol. 1996,
261, 470-489.
a Ki of 1.4 and 3.8 µM, respectively. More importantly, these
inhibitors were 10-35 times more potent toward MMP-7
than the bacterial metalloproteases tested, demonstrating the
potential of the strategy in elucidating inhibitors with both
good potency and high selectivity against MMPs.
3824
Org. Lett., Vol. 8, No. 17, 2006