Structure-based design and synthesis of a potent matrix metalloproteinase-13 inhibitor based on a pyrrolidinone scaffold [1]

Full text of DOI:10.1021/jm0001368

© Copyright 2000 by the American Chemical Society
Volume 43, Number 12
J une 15, 2000
Communications to the Editor
Str u ctu r e-Ba sed Design a n d Syn th esis of
a P oten t Ma tr ix Meta llop r otein a se-13
In h ibitor Ba sed on a P yr r olid in on e
Sca ffold
Ralph P. Robinson,* Ellen R. Laird, J ames F. Blake,
J on Bordner, Kathleen M. Donahue,
Lori L. Lopresti-Morrow, Peter G. Mitchell,
Matthew R. Reese, Lisa M. Reeves,
Ethan J . Stam, and Sue A. Yocum
Central Research Division, Pfizer Inc.,
Eastern Point Road, Groton, Connecticut 06340
Received March 23, 2000
observed involves hydrogen bonding between an accep-
tor on the inhibitor and main chain NHs of amino acid
residues flanking the catalytic site, specifically Leu-185
and Ala-186 (MMP-13 numbering). A few inhibitors also
possess a proton donor that interacts with Ala-186.8,10
The close proximity of the latter two interaction sites
led us to consider lactams that could make both hydro-
gen bond-donating and -accepting contacts and, being
cyclic, would possess the entropic advantage of ring
constraint. Using a homology model of the catalytic
domain of MMP-13 based on the X-ray structure of
MMP-8,15,16 five- and six-membered ring lactams were
examined to identify the position most favorable for
attaching a ligand for the active site Zn atom. Among
the various possibilities, the pyrrolidin-5-one 5 bearing
hydroxamate at the 2-position was especially attractive
(Chart 1). As determined using a Monte Carlo fragment
positioning algorithm,17 this structure is able to adopt
an energetically favorable orientation capable of direct-
In tr od u ction . The inhibition of matrix metallopro-
teinases (MMPs) is an attractive approach toward the
treatment of a variety of diseases.1-3 For example,
inhibitors of MMP-2 and MMP-9 are sought for preven-
tion of cancer tumor growth,4 while inhibitors of MMP-
13 show potential for reducing cartilage loss in osteoar-
thritis.5 Several MMP inhibitors are now in advanced
clinical trials. However, there remains considerable
room for optimization of oral absorption and duration
of action while maintaining potency and selectivity. The
vast majority of reported MMP inhibitors can be viewed
as being derivatives of either the substrate-like peptidic
motif 1 (e.g., marimistat, 2) or the aryl sulfonamide-
based hydroxamic acids 3 (e.g., CGS 27023A, 4).6
Although substantial modification of 1 and 3 is often
tolerated, the discovery of new structural classes should
prove valuable in the search for clinically useful MMP
inhibitors. We now describe the structure-based design
and synthesis of a novel, potent, and selective inhibitor
of MMP-13 that utilizes a pyrrolidinone ring as a
scaffold.7
Ch a r t 1
Resu lts a n d Discu ssion . X-ray crystal structures of
MMPs (including MMP-13) with bound inhibitors reveal
that inhibitor binding interactions typically include
coordination of the active site Zn atom by a ligand (e.g.,
hydroxamic acid) and occupancy of the S1′ pocket with
a hydrophobic group.8-14 Another interaction often
* To whom correspondence should be addressed. Tel: 860-441-4923.
Fax: 860-715-2469. E-mail: ralph_p_robinson@groton.pfizer.com.
10.1021/jm0001368 CCC: $19.00 © 2000 American Chemical Society
Published on Web 05/27/2000

2294 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 12
Communications to the Editor
Sch em e 1a
a
Reagents: (a) I2, pyridine, CCl4 (31%); (b) 4-MeO(C6H4)B(OH)2, cat. Pd(PPh3)4, aq Na2CO3, toluene, 4 (77%); (c) H2 (3 atm), 20%
Pd(OH)2/C, EtOH (100%); (d) aq 0.5 M HCl (1 equiv), THF (48%); (e) H5IO6, cat. CrO3, wet CH3CN/0 °C (98%); (f) PhCH2ONH2‚HCl, BOP,
i-Pr2NEt, CH2Cl2 (73%); (g) HCl(g)/CH2Cl2 (80%); (h) H2 (3 atm), 5% Pd/BaSO4, MeOH (96%).
ing a group (R) from the cis 4-position into the S1′ pocket
as in 6 (Chart 1; see also Figure 1).
well as docking studies using the MMP-13 homology
model, the 4-(4-fluorophenoxy)phenyl compound 6b was
of particular interest (Figure 1). Indeed, this compound
exhibited high potency against MMP-13 (IC50 ) 7 nM).
As for 6a , inhibition of MMP-1 by 6b was weak, a
potentially desirable result in consideration of side
effects attributed to MMP-1 inhibition by “broad-
spectrum” MMP inhibitors.4 Additionally, 6b showed
modest to high selectivity for MMP-13 versus MMP-2,
MMP-3, and MMP-9 (Table 2).
Because the pyrrolidinones 6 are conformationally
restricted, it was expected that the potency of a given
analogue against various MMPs would be particularly
dependent on the degree to which the enzyme must
reorganize (or its capacity to do so) upon ligand binding.
Insofar as the S1′ site involves a large area of interac-
tion, the involvement of amino acid residues defining
this pocket is particularly relevant. The observation that
6b is a very potent inhibitor of MMP-13 suggests that
the residues comprising the MMP-13 S1′ pocket are pre-
organized to accept this inhibitor without requiring
energetically demanding movement. For the other MMPs,
residues that may affect access to (or the flexibility of)
the S1′ pocket can be identified to explain the decreases
in the activity of 6b. In MMP-1, an arginine side chain
(Arg-214) residing within S1′ normally blocks large
groups from entry resulting in a preference by MMP-1
for small P1′ substituents. Although in some cases large
hydrophobic groups such as a diphenyl ether can be
accommodated by displacement of Arg-214, binding
potency is often reduced.14 Similarly, in MMP-3, a
tyrosine side chain (Tyr-223) has been observed to
restrict access to S1′.28 In MMP-2 and MMP-9 there are
fewer residues in the loop linking helices B and C that
forms part of S1′.29 Therefore, the loop may be less
flexible and so may be more restricted in accommodating
rigid ligands.
F igu r e 1. Model of binding mode for compound 6b in the
active site of MMP-13. The protein is rendered as a solvent-
accessible surface and is clipped to reveal the S1′ pocket. The
catalytic zinc is portrayed as a space-filled violet sphere.
On the basis of the reported MMP binding mode of 4
and related aryl sulfone analogues (in which an aryl
group fills the S1′ pocket),14,18 as well as the availability
of starting materials, compound 6a (R ) 4-methoxyphe-
nyl) was first prepared. The synthesis began with the
R,â-unsaturated γ-lactam 719 which was R-iodinated20
to obtain 8 (Scheme 1). Suzuki coupling21 introduced the
4-methoxyphenyl group, providing 9,22 and hydrogena-
tion, giving 10, established the cis relationship required
between the substituents on the pyrrolidinone ring.23
Selective removal of the tert-butyldimethylsilyl protect-
ing group24 and oxidation of the intermediate alcohol25
gave the carboxylic acid 11. Final conversion to 6a was
carried out in three steps involving coupling to O-benzyl-
hydroxylamine and removal of the BOC and benzyl26
protecting groups.
Testing of 6a against MMP-13 showed the compound
to have modest activity (IC50 ) 1480 nM, Table 1). This
result suggested that the pyrrolidinone scaffold could
be functioning as proposed and that appropriate sub-
stitution of the phenyl ring with larger hydrophobic
groups might improve potency by filling the S1′ pocket
more completely. Accordingly, compounds 6b -f were
prepared by routes paralleling that in Scheme 1.27 On
the basis of the reported SAR among analogues of 3, as
Compounds 6c-f were significantly less potent (at
least ∼25-fold) than 6b against MMP-13 (Table 1). The
tight SAR is consistent with occupation of a well-defined
pocket such as the S1′ site. An examination of 6c and
6d within the homology model shows that although the
aryl substituents can be accommodated, they are less
complementary than the naturally bent 4-(4-fluorophe-
noxy)phenyl group of 6b (Figure 1). The 4-benzyloxy-
phenyl group of 6e exhibits a similar geometry to 6b,
and therefore, the modest potency of 6e against MMP-
13 is surprising. The low potency of 6f is consistent with

Communications to the Editor
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 12 2295
Ta ble 1. Inhibition of MMP-1 and MMP-13 by 6a -f, 12, and 13
a
Values are (SD of 3 determinations unless otherwise noted (n).
Ta ble 2. Inhibition of Various MMPs by 6b
be possible by optimization of the P1′ (R) group. Fur-
thermore, the conformational rigidity of the series may
be advantageous as a means to obtain selectivity. Thus
6b may be regarded as a novel, nonpeptidic, and low-
molecular-weight lead for continued pursuit of MMP
inhibitors as drugs.
enzyme
IC50 (nM)a
MMP-1
MMP-2
MMP-3
MMP-9
MMP-13
1560 ( 120
39 ( 17
703 ( 170
82 ( 16
7 ( 1
a
Values are (SD of 3 determinations.
Ack n ow led gm en t. We thank Drs. J ulian Blagg,
Allen J . Duplantier, Mark C. Noe, and Lawrence A.
Reiter for helpful comments and reviews of the manu-
script. We also thank Melissa Rockwell for enantiomeric
excess determinations.
the model since a poor fit of the 3-(4-fluorophenoxy)-
phenyl group within S1′ is predicted.
As expected, the trans (2R,4R) compound 1230 was
considerably less potent versus MMP-13 (IC50 ) 3390
nM) than its cis stereoisomer 6b. The carboxylate
analogue of 6b (13)31 was devoid of MMP-13 activity in
keeping with the SAR in other series where the car-
boxylates possess markedly less MMP inhibitory activity
than the corresponding hydroxamates.
Refer en ces
(1) Michaelides, M. R.; Curtin, M. L. Recent Advances in Matrix
Metalloproteinase Inhibitor Research. Curr. Pharm. Des. 1999,
5, 787-819.
(2) Whittaker, M.; Floyd, C. D.; Brown, P.; Gearing, A. J . H. Design
and Therapeutic Application of Matrix Metalloproteinase Inhibi-
tors. Chem. Rev. 1999, 2735-2776.
(3) Beckett, R P.; Whittaker, M. Matrix Metalloproteinase Inhibitors
1998. Exp. Opin. Ther. Patents 1998, 8, 259-282.
(4) Rothenberg, M. L.; Nelson, A. R.; Hande, K. R. New Drugs on
the Horizon: Matrix Metalloproteinase Inhibitors. Oncologist
1998, 3, 271-274.
(5) Bottomley, K. M.; J ohnson, W. H.; Walter, D. S. Matrix Metal-
loproteinase Inhibitors in Arthritis. J . Enzyme Inhib. 1998, 13,
79-101.
(6) MacPherson, L. J .; Bayburt, E. K.; Capparelli, M. P.; Carroll,
B. J .; Goldstein, R.; J ustice, M. R.; Zhu, L.; Hu, S.; Melton, R.
A.; Fryer, L.; Goldberg, R. L.; Doughty, J . R.; Spirito, S.;
Blancuzzi, V.; Wilson, D.; O′Byrne, E. M.; Ganu, V.; Parker, D.
T. Discovery of CGS 27023A, A Non-Peptidic, Potent and Orally
Active Stromelysin Inhibitor That Blocks Cartilage Degradation
in Rabbits. J . Med. Chem. 1997, 40, 2525-2532.
(7) During the course of this work, patent applications disclosing
unrelated pyrrolidinone-based MMP inhibitors were published:
(a) J acobsen, E. J . PCT Int. Appl. WO 97/32846. (b) Duan, J .;
Decicco, C. P.; Wasserman, Z. R.; Maduskuie, T. P., J r. PCT Int.
Appl. WO 99/18074.
Con clu sion . In summary, we have discovered a
potent inhibitor of MMP-13 by a structure-based ap-
proach using a homology model of the enzyme derived
from MMP-8. A pyrrolidinone template was selected
based on an analysis of hydrogen-bonding interactions
made by known MMP inhibitors and the vectors re-
quired for correct projection of established MMP binding
groups (hydroxamate for Zn chelation, substituted aryl
for occupation of the S1′ site). Despite the availability
of several crystal structures of MMPs, few reports of
their explicit use in the design of scaffolds exist.32 The
novel structure of 6b, a 4-arylpyrrolidin-5-one-2-car-
boxylic acid hydroxyamide, represents a significant
departure from those of the majority of reported MMP
inhibitors. Although our work focused on inhibition of
MMP-13, it seems likely that improved inhibition of
other MMPs by analogues in this structural class may

2296 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 12
Communications to the Editor
(8) Lovejoy, B.; Cleasby, A.; Hassell, A. M.; Longley, K.; Luther, M.
A.; Weigl, D.; McGeehan, G.; McElroy, A. B.; Drewry, D.;
Lambert, M. H.; J ordan, S. R. Structure of the Catalytic Domain
of Fibroblast Collagenase Complexed with an Inhibitor. Science
1994, 263, 375-377.
(9) Borkakoti, N.; Winkler, F. K.; Williams, D. H.; D’Arcy, A.;
Broadhurst, M. J .; Brown, P. A.; J ohnson, W. H.; Murray, E. J .
Structure of the Catalytic Domain of Human Fibroblast Colla-
genase Complexed with an Inhibitor. Nat. Struct. Biol. 1994, 1,
106-110.
(10) Becker, J . W.; Marcy, A. I.; Rokosz, L. L.; Axel, M. G.; Burbaum,
J . J .; Fitzgerald, P. M. D.; Cameron, P. M.; Esser, C. K.;
Hagmann, W. K.; Hermes, J . D.; Springer, J . P. Stromelysin-1:
Three-dimensional Structure of the Inhibited Catalytic Domain
and of the C-Truncated Proenzyme. Protein Sci. 1995, 4, 1966-
1976.
(11) Browner, M. F.; Smith, W. W.; Castelhano, A. L.; Matrilysin-
Inhibitor Complexes: Common Themes among Metalloproteases.
Biochemistry 1995, 34, 6602-6610.
(22) Under the Suzuki coupling conditions (aq Na2CO3/toluene/
reflux), long reaction times (>12 h) gave appreciable racemiza-
tion. The use of 8 instead of the corresponding bromo lactam
was preferred because the higher coupling rate allowed reaction
times to be shortened to about 1 h, thus minimizing racemization
as determined by chiral HPLC (ee’s > 98%).
(23) Proof of the assigned stereochemistry was obtained from single-
crystal X-ray structure analysis of racemic 6f (prepared via
Suzuki coupling of 8 with 3-(4-fluorophenoxy)phenylboronic acid
in refluxing toluene for 16 h). Comparison of the 1H NMR spectra
of 6a -e to the spectrum of 6f showed a common pattern of
chemical shifts and coupling constants for the pyrrolidinone ring
protons. The pattern was clearly distinguishable from that
exhibited by 12.
(24) Retention of the N-BOC group was essential for the success of
the subsequent oxidation.
(25) Zhao, M.; Li, J .; Song, Z.; Desmond, R.; Tschaen, D. M.;
Grabowski, E. J . J .; Reider, P. J . A Novel Chromium Trioxide
Catalyzed Oxidation of Primary Alcohols to the Carboxylic Acids.
Tetrahedron Lett. 1998, 39, 5323-5326.
(26) Nikam, S. S.; Kornberg, B. E.; J ohnson, D. R.; Doherty, A. M.
Synthesis of Hydroxamic Acids: Pd/BaSO4 as a New Catalyst
for the Deprotection of O-Benzyl Hydroxamates. Tetrahedron
Lett. 1995, 36, 197-200.
(27) The synthesis of the 4-benzyloxyphenyl analogue 6e differed
slightly by incorporation of minor changes allowing preservation
of the O-benzyl ether: (1) use of EtOAc as solvent in the
hydrogenation of the double bond (the benzyl ether is not
hydrogenolyzed); (2) use of 2-(trimethylsilyl)ethyl (removed in
the last step using BF3‚OEt2 in CH2Cl2) instead of benzyl as the
hydroxamate protecting group; (3) complete removal of the BOC
group prior to coupling to the hydroxylamine derivative, TMS-
(CH2)2ONH2‚HCl, due to partial loss of the BOC group during
the H5IO6/cat. CrO3 oxidation.
(28) Chen, L.; Rydel, T. J .; Gu, F.; Dunaway, C. M.; Pikul, S.;
Dunham, K. M.; Barnett, B. L. Crystal Structure of the Stromel-
ysin Catalytic Domain at 2.0 A¨ Resolution: Inhibitor-Induced
Conformational Changes. J . Mol. Biol. 1999, 293, 545-557.
(29) Kiyama, R.; Tamura, Y.; Watanabe, F.; Tsuzuki, H.; Ohtani, M.;
Yodo, M. Homology Modeling of Gelatinase Catalytic Domains
and Docking Simulations of Novel Sulfonamide Inhibitors. J .
Med. Chem. 1999, 42, 1723-1738.
(12) Stams, T.; Spurlino, J . C.; Smith, D. L.; Wahl, R. C.; Ho, T. F.;
Qoronfleh, M. W.; Banks, T. M.; Rubin, B. Structure of Human
Neutrophil Collagenase Reveals Large S1′ Specificity Pocket.
Nat. Struct. Biol. 1994, 1, 119-123.
(13) Grams, F.; Crimmin, M.; Hinnes, L.; Huxley, P.; Pieper, M.;
Tschesche, H.; Bode, W. Structure Determination and Analysis
of Human Neutrophil Collagenase Complexed with a Hydrox-
amate Inhibitor. Biochemistry 1995, 34, 14012-14020.
(14) Lovejoy, B.; Welch, A. R.; Carr, S.; Luong, C.; Broka, C.;
Hendricks, R. T.; Campbell, J . A.; Walker, K. A. M.; Martin, R.;
Van Wart, H.; Browner, M. F. Crystal Structures of MMP-1 and
-13 Reveal the Structural Basis for Selectivity of Collagenase
Inhibitors. Nat. Struct. Biol. 1999, 6, 217-221.
(15) Bode, W.; Reinemer, P.; Huber, R.; Kleine, T.; Schnierer,
Tschesche, H. The X-ray Crystal Structure of the Catalytic
Domain of Human Neutrophil Collagenase Inhibited by
a
Substrate Analogue Reveals the Essentials for Catalysis and
Specificity. EMBO J . 1994, 13, 1263-1269.
(16) We did not have access to an X-ray crystal structure of MMP-
13. The model, based on MMP-8, was built using the program
COMPOSER and minimized using the AMBER all-atom force
field as implemented in the Sybyl program. (Sybyl 6.3, Rev 3A;
Tripos Inc., 1699 Hanley Rd, Suite 303, St. Louis, MO 63144.)
(17) Blake, J . F. Unpublished information.
(18) Li, Y.-C.; Zhang, X.; Melton, R.; Ganu, V.; Gonnella, N. C.
Solution Structure of the Catalytic Domain of Human Stromel-
ysin-1 Complexed to a Potent, Nonpeptidic Inhibitor. Biochem-
istry 1998, 37, 14048-14056.
(19) The (S) enantiomer is known: (a) Woo, K.-C.; J ones, K. Asym-
metric Synthesis from R-Amino Acids; Some Reactions of (S)-
Pyroglutamate. Tetrahedron Lett. 1991, 6949-6952. (b) Shima-
moto, K.; Ishida, M.; Shinozaki, H.; Ohfune, Y. Synthesis of Four
Diastereomeric L-2-(Carboxycyclopropyl)glycines. Conformation-
ally Constrained L-Glutamate Analogues. J . Org. Chem. 1991,
56, 4167-4176.
(20) J ohnson, C. R.; Adams, J . P.; Braun, M. P.; Senanayake, C. B.
W.; Wovkulich, P. M.; Uskokovic, M. R. Direct R-Iodination of
Cycloalkenones. Tetrahedron Lett. 1992, 33, 917-918.
(21) In the Suzuki couplings of 8, the arylboronic acids were prefer-
ably generated from the corresponding diethanolamine com-
plexes; see: Caron, S.; Hawkins, J . M. Directed Ortho Metalation
of Neopentyl Benzoates with LDA: Preparation of Arylboronic
Acids. J . Org. Chem. 1998, 63, 2054-2055.
(30) The isomer 12 was obtained via treatment of the intermediate
corresponding to 10 with DBU (3.5 equiv) in toluene at room
temperature. The epimeric product was then advanced to 12 in
the usual way.
(31) The acid 13 was prepared by treatment of the corresponding
N-BOC compound with HCl gas in CH2Cl2.
(32) Szardenings, A. K.; Harris, D.; Lam, S.; Shi, L.; Tien, D.; Wang,
Y.; Patel, D. V.; Navre, M.; Campbell, D. A. Rational Design and
Combinatorial Evaluation of Enzyme Inhibitor Scaffolds: Iden-
tification of Novel Inhibitors of Matrix Metalloproteinases. J .
Med. Chem. 1998, 41, 2194-2200.
J M0001368