Improving potency and selectivity of a new class of non-Zn-chelating MMP-13 inhibitors

Article abstract of DOI:10.1016/j.bmcl.2009.07.151

Discovery and optimization of potency and selectivity of a non-Zn-chelating MMP-13 inhibitor with the aid of protein co-crystal structural information is reported. This inhibitor was observed to have a binding mode distinct from previously published MMP-1

Full text of DOI:10.1016/j.bmcl.2009.07.151

Bioorganic & Medicinal Chemistry Letters 19 (2009) 5321–5324
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Improving potency and selectivity of a new class of non-Zn-chelating MMP-13
inhibitors
a
a
a
a
Alexander Heim-Riether a, , Steven J. Taylor , Shuang Liang , Donghong Amy Gao , Zhaoming Xiong ,
E. Michael August a, Brandon K. Collins a, Bennett T. Farmer II a, Kathleen Haverty a, Melissa Hill-Drzewi a,
Hans-Dieter Junker c, S. Mariana Margarit a, Neil Moss a, Thomas Neumann c, John R. Proudfoot a,
Lana Smith Keenan a, Renate Sekul c, Qiang Zhang a, Jun Li b, Neil A. Farrow a
*
a Boehringer Ingelheim Pharmaceuticals, Department of Medicinal Chemistry, 900 Ridgebury Rd, Ridgefield, CT 06877, USA
b Boehringer Ingelheim Pharmaceuticals, Department of Immunology and Inflammation, 900 Ridgebury Rd, Ridgefield, CT 06877, USA
c Graffinity Pharmaceuticals GmbH, Im Neuenheimer Feld 518, 69120 Heidelberg, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Discovery and optimization of potency and selectivity of a non-Zn-chelating MMP-13 inhibitor with the
aid of protein co-crystal structural information is reported. This inhibitor was observed to have a binding
mode distinct from previously published MMP-13 inhibitors. Potency and selectivity were improved by
extending the hit structure out from the active site into the S10 pocket.
Received 3 July 2009
Revised 27 July 2009
Accepted 29 July 2009
Available online 3 August 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
MMP-13
Non-Zn-chelating inhibitor
SAR
Matrix metalloproteinases (MMPs) are zinc- and calcium
dependent endopeptidases involved in the degradation of extracel-
lular matrix and tissue remodeling.1 MMP-13, the most efficient
type II collagen-degrading MMP,2 is an attractive therapeutic tar-
get because evidence supports it playing a critical role in the path-
ogenesis of osteoarthritis (OA).3 Broad-spectrum MMP inhibitors
have failed in clinical trials at least in part due to a painful, joint-
stiffening side effect, termed musculoskeletal syndrome (MSS).4
It is believed that MSS is caused by the inhibition of normal extra-
cellular matrix turnover, likely due to the inhibition of MMPs other
than MMP-13.5 The extent to which other MMPs contribute to MSS
is not well understood. Nevertheless, we sought to develop MMP-
13 inhibitors with high isoform selectivity to mitigate unperceived
risk associated with inhibiting other isoforms.
MMPs have a conserved active site motif, where a tris(histi-
dine)-bound zinc(II) acts as the catalytic site for substrate hydroly-
sis. Historically, most MMP inhibitors gained affinity through
interacting with the catalytic zinc via a chelating moiety (e.g.,
hydroxamic acid) and by positioning hydrophobic functionality
in the S10 pocket.4a,6 The S10 pocket is in part formed by the selec-
tivity loop which varies in length and amino acid sequence for dif-
ferent MMP isoforms. These S10 differences between MMP family
members have been utilized to design MMP inhibitors with differ-
ent selectivity profiles.7 MMP-13 has an additional region for
inhibitor binding that has not been observed in other MMP iso-
forms. This region is referred to as the S10* pocket. At present,
the most potent and selective MMP-13 inhibitors occupy both
the S10 and S10* pockets.8 The potency provided by binding to the
S10 and S10* sites reduces the need for inhibitors to require a Zn
binding group.
Through a Graffinity SPR microarray screening campaign,9 we
identified hit structure 1 that was subsequently determined to
have an IC50 of 430 nM against MMP-13 (Scheme 1). Compound
1 attracted our attention since it did not contain an obvious Zn-
chelating group and showed selectivity over MMP-14 (Table 1)—
the one isoform postulated to play a role in MSS.5 We obtained a
co-structure of 1 with MMP-13 which revealed a binding mode
not previously observed for a compound lacking a Zn-chelating
group (Fig. 1).
O
O
NH
NH
O
H
N
HN
HN
Cl
O
O
O
O
O
1
2
* Corresponding author. Tel.: +1 203 798 5579; fax: +1 203 791 6072.
E-mail address: alexander.heim-riether@boehringer-ingelheim.com (A. Heim-Riether).
Scheme 1. Screening hit and first modification.
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.07.151

5322
A. Heim-Riether et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5321–5324
Figure 1. MMP13-co-crystal structure of compound 1. No interaction to the
Figure 2. Overlay of co-crystal structures of compound 1 (green) and 2 (gray)
within the S10 pocket of MMP-13. 2 reaches deeper into the S10 pocket (selectivity
loop highlighted in red).10
catalytic Zn is observed; entrance of S10 pocket is occupied by phenyl ring.10
Compound 1 binds in the active site region despite not interact-
ing with the catalytic zinc. The observed key interactions are three
hydrogen bonds from and to the amino acid portion of the mole-
O2N
O
O
O
NH
Ar
H2N
N
H
a, b, c
HN
HN
cule and a
p-stacking interaction of the phenyl ring with one of
+
O
O
the histidines.10 Since it is not precedented that potency and selec-
tivity can be easily gained through interactions in the active site,
we focused our attention on the observation that compound 1 does
not fill the S10 pocket. The phenyl ring occupies only the entrance
to the S10 pocket thus providing the opportunity to grow into the
S10 pocket to improve potency against MMP-13 and the overall
selectivity profile against additional MMP isoforms. This opportu-
nity and the distinctly different binding mode compared to other
known MMP-13 inhibitors made 1 an attractive starting point for
a discovery program.
O
OH
3
O
4
2 (Ar = benzofuran)
Scheme 2. Reagents: (a) EDC, HOBt, DIEA, DCM; (b) Zn, AcOH; (c) ArCO2H, HATU,
DIEA, DMF.
A co-structure of 2 with MMP-13 showed the benzofuran moi-
ety occupying the S10 pocket (Fig. 2, compound 2 in gray). No spe-
cific interactions of the benzofuran carboxamide with the protein
were observed,10 which may explain the lack of potency increase.
The selectivity gain against other isoforms was attributed to differ-
ences in shape and size of the S10 pockets which are defined in part
by the amino acid sequences and lengths of the respective selectiv-
ity loops of the various isoforms. Compared to MMP-13, this loop is
one to two amino acids shorter for some isoforms (MMP-1, -2, -9,
-11), presumably leading to a smaller S10 pocket that does not
Our initial attempt at expanding compound 1 to fill the unoccu-
pied S10 pocket involved replacing the chlorine with a benzofuran
carboxamide (compound 2, Scheme 1). This modification was
based on published SAR of a Zn-chelating, partially selective
MMP-13 inhibitor.6 Compound 2 was essentially equipotent to
compound 1 but provided an improved selectivity profile against
a panel of MMP isoforms (Table 1).
Table 1
Aniline amide SAR11
Compd
Ar
MMP-13 IC50 (nM)
MMP-1
MMP-2
MMP-3
MMP-7
MMP-8
MMP-9
MMP-10
MMP-12
MMP-14
Fold over MMP-13
1
2
—
430
620
>50
>30
6
28
>50
>30
2
9
9
4
5
>100
>100
O
>100
>30
>30
>30
>30
O
5
6
7
8
9
63
57
31
180
88
>100
>100
>100
>100
>100
>100
73
13
3
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
4
1
1
4
2
>100
51
>100
>100
>100
>100
>100
O
N
N
63
3
33
N
N
30
8
26
>100
7
>100
HN
N
10
17
>1000
>1000
37
>1000
>1000
>1000
21
>1000
>1000
N

A. Heim-Riether et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5321–5324
5323
accommodate the benzofuran. For MMP-3 and -10 the loop is one
O
O
NH
NH
= Ar
X
amino acid longer, presumably leading to a larger S10 pocket. This
may in part explain the low selectivity of compound 2 against
MMP-10.
a
HN
HN
R
O
O
Br
O
O
Although the gain in selectivity for compound 2 over 1 was
encouraging, affinity for MMP-13 and selectivity against MMP-10
needed improvement. Based on the co-crystal structure of 2 it
was apparent that extensions from the 4- and 5-position of the
benzofuran had the potential to improve potency and selectivity
by leveraging the S10* pocket. However, such modifications would
result in significant increases in molecular weight (2: MW = 499).
We focused on identifying groups that took better advantage of
the potentially unique shape of the S10 pocket of MMP-13 com-
pared to other isoforms. A lipophilic residue proximal to the spec-
ificity loop, proline 255, further distinguishes MMP-13 from most
MMPs in our panel which have a polar residue at this position
(MMP-2, -3, -7, -9, -10, -12: serine). It turned out that groups smal-
ler than the benzofuran provided interesting potency and selectiv-
ity profiles (Table 1). These compounds were prepared in a straight
forward manner as shown in Scheme 2.
Compound 5 led to a 10-fold increase in potency against MMP-
13 but this was accompanied by a decrease in selectivity against
MMP-3. MMP-3 is the other MMP in the panel possessing a selec-
tivity loop one amino acid longer than MMP-13. Therefore, further
reducing the size of the group occupying the S10 pocket might be
expected to cause an additional loss in selectivity against MMP-3
and -10. This was supported by removing one of the methyl groups
of 5 to provide compound 6 which, while equipotent against MMP-
13, lost selectivity against MMP-2, -3, -10 and -12. Similar trends in
potency and selectivity were observed when switching from the
furanyl to the pyrazole ring system (compounds 7–9). However,
the most significant advance came from incorporating a methyl
pyridyl group, leading to compound 10 which combines improved
potency with an overall improved selectivity profile compared to
compound 2.
12 (R = Cl, X = N)
11
Scheme 3. Reagents and conditions: (a) 4-R-ArB(OH)2 (Ar = Ph, pyridyl), Pd(PPh3)4,
2 M Na2CO3, DMF, 80–120 °C.
Table 2
Baseline compounds (13–20) for alternative functionalities to connect the terminal
aryl group to the core
Compds
R
X
MMP-13
IC50 (nM)
HT-Sol pH 4.5/7.4
g/mL)
(l
1
12
13
14
15
16
17
18
19
20
Cl
Cl
CH
N
430
245
1550
6000
460
480
290
>50,000
220
32
1.5/1.9
35/37
15/17
16/17
53/61
50/52
—
—
1.4/1.6
33/>49
MeC(@O)NH–
MeNHC(@O)–
MeSO2NH–
MeNHSO2–
H3C–
PhO–
MeO–
EtO–
CH
CH
CH
CH
CH
CH
CH
N
Interestingly, acetamide 13, the de-arylated analogue of the ter-
minal aryl SAR of Table 1, was together with its reversed amide iso-
mer 14 among the least potent analogs tested. Based on its
potency, small size, chemical feasibility and the potential to mimic
the 2-atom-linkage of the amide, the ether linkage contained in 19
and 20 was selected for subsequent terminal aryl SAR (Table 3).
Analogue 20, derived from the combination of the ether linkage
of 19 with the pyridyl ring of compound 12, promised not only
the potential for more potent but also relatively soluble inhibitors.
Combinations with the aryl ether functionality are highlighted in
Table 3 and they were synthesized using the synthetic sequence
as illustrated in Scheme 4.
While compound 10 demonstrated a more attractive balance of
potency and selectivity than 2, the terminal aryl amide linkage was
considered a liability since hydrolysis would lead to a potentially
mutagenic fragment. Alternative linkages without the terminal
aryl group were surveyed (Table 2). A straightforward synthesis
is shown in Scheme 3.
Adding our most selective aryl groups, the benzofuran of com-
pound 2 and the methylpyridine of 10, to the ether linkage re-
vealed different trends for potency and selectivity compared to
Table 3
Benzylether SAR (compounds are selective against MMP-1, -7, and -14)11
Compds
R
MMP-13 IC50 (nM)
MMP-2
MMP-3
MMP-8
MMP-9
MMP-10
MMP-12
HT-Sol pH 4.5/7.4 (lg/mL)
Fold over MMP-13
1
H3C
20
24
32
6
4
8
5
5
10
33/>49
>500
26
>500
>500
>100
>100
>1000
>500
>500
>100
4
11
>1000
<0.1/0.1
O
H3C
HN
25
26
27
28
29
N
40
6
21
6
4
21
7
6
14
6
68
7.4/3.4
39/43
—
>100
21
N
25
2
10
19
>100
>100
65
N
H3C
H3C
N
80
45
64
24
>500
>100
25/31
>57/>57
HN
N
H3C
4
>500
N
H3C
N

5324
A. Heim-Riether et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5321–5324
Geoghegan, K. F.; Hambor, J. E. J. Clin. Invest. 1996, 97, 761; (c) Reboul, P.;
BnO
BnO
a
b, c, d
Pelletier, J.-P.; Tardif, G.; Cloutier, J.-M.; Martel-Pelletier, J. J. Clin. Invest. 1996,
97, 2011; (d) Wernicke, D.; Seyfert, C.; Hinzmann, B.; Gromnica-Ihle, E. J.
Rheumatol. 1996, 23, 590; (e) Billinghurst, R. C.; Dahlberg, L.; Ionescu, M.;
Reiner, A.; Bourne, R.; Rorabeck, C.; Mitchell, P.; Hambor, J.; Diekmann, O.;
Tschesche, H.; Chen, J.; Van Wart, H.; Poole, A. R. J. Clin. Invest. 1997, 99, 1534;
(f) Freemont, A. J.; Byers, R. J.; Taiwo, Y. O.; Hoyland, J. A. Ann. Rheum. Dis. 1999,
58, 357.
H
O
N
N
Br
O
21
22
O
O
NH
NH
4. (a) Skiles, J. W.; Gonnella, N. C.; Jeng, A. Y. Curr. Med. Chem. 2001, 8, 425; (b)
Clark, I. M.; Parker, A. E. Exp. Opin. Ther. Targets 2003, 7, 19; (c) Renkiewicz, R.;
Qiu, L.; Lesch, C.; Sun, X.; Devalaraja, R.; Cody, T.; Kaldjian, E.; Welgus, H.;
Baragi, V. Arthritis Rheum. 2003, 48, 1742.
HN
e
O
HN
O
HO
R
O
O
N
O
N
24 (R = CH2-benzofuran)
23
5. Historically, MMP-1 and -14 were the postulated key players in causing MSS.
However, the non-selective MMP-13 inhibitor CP-544439, which is inactive
against MMP-1, has caused MSS in clinical trials: Reiter, L. A.; Freeman-Cook, K.
D.; Jones, C. S.; Martinelli, G. J.; Antipas, A. S.; Berliner, M. A.; Datta, K.; Downs,
J.; Eskra, J. D.; Forman, M. D.; Greer, E. M.; Guzman, R.; Hardink, J. R.; Janat, F.;
Keene, N. F.; Laird, E. R.; Liras, J.; Lopresti-Morrow, L. L.; Mitchell, P. G.; Pandit,
J.; Robertson, D.; Sperger, D.; Vaughn-Bowser, M. L.; Waller, D. M.; Yocum, S. A.
Bioorg. Med. Chem. Lett. 2006, 16, 5822. MMP-14 has KO mice data supporting it
being a strong candidate for MSS: see Ref. 3a.
6. (a) Wu, J.; Rush, T. S.; Hotchandani, R.; Du, X.; Geck, M.; Collins, E.; Xu, Z.-B.;
Skotnicki, J.; Levin, J. I.; Lovering, F. E. Biorg. Med. Chem. Lett. 2005, 15, 4105; (b)
Li, J.; Rush, T. S.; Li, W.; DeVincentis, D.; Du, X.; Hu, Y.; Thomason, J. R.; Xiang, J.
S.; Skotnicki, J. S.; Tam, S.; Cunningham, K. M.; Chockalingam, P. S.; Morris, E.
A.; Levin, J. I.; Lovering, F. E. Biorg. Med. Chem. Lett. 2005, 15, 4961.
7. Matter, H.; Schudok, M. Curr. Opin. Drug Discovery Dev. 2004, 7, 513.
8. (a) Skiles, J. W.; Gonnella, N. C.; Jeng, A. Y. Curr. Med. Chem. 2004, 11, 2911; (b)
Rao, B. G. Curr. Pharm. Des. 2005, 11, 295; (c) Engel, C. K.; Pirard, B.; Schimanski,
S.; Kirsch, R.; Habermann, J.; Klingler, O.; Schlotte, V.; Weithmann, K. U.;
Wendt, K. U. Chem. Biol. 2005, 12, 181; (d) Pirard, B. Drug Discovery Today 2007,
12, 640.
Scheme 4. Reagents and conditions: (a) 2-furaldehyde-4-boronic acid, Pd(PPh3)4,
2 M Na2CO3, DME, 120 °C; (b) NaClO2, NaH2PO4, HOSO2NH2, dioxane; (c) (S)-2-
amino-2-C6H11-N-methyl-acetamide, TBTU, DIEA, DMF; (d) 10% Pd/C, 1,4-cyclo-
hexadiene, EtOAc, MeOH; (e) ArCH2OH, DIAD, PPh3, THF.
the corresponding amide analogs. Compound 24, the combination
of the benzofuran with the ether linkage, was significantly more
potent than its amide counterpart 2 while maintaining a similar
selectivity profile. In contrast, the methylpyridine ether analogue
25 demonstrated comparable potency to the corresponding amide
linked compound 10, but surprisingly, lost selectivity against
MMP-2, -3, -10, and -12. However, it turned out that the more
soluble pyrazole ether combinations (26–29) depending on the
substitution pattern can best balance both potency and selectivity.
Dimethylpyrazole 29, the combination of N-and C-methylated
pyrazoles 27 and 28, respectively, became the most potent and
selective MMP-13 inhibitor of the series.
9. High-throughput SPR imaging by Graffinity against the company’s fragment/
leadlike 110,000 compound library (www.graffinity.com).
10. Key interactions of 1 and 2:
p-Stacking with His222; 3 hydrogen bonds:
Carbonyl oxygen of furanyl amide with the backbone amide of Leu185, methyl
amide carbonyl oxygen with the backbone amide nitrogen of Tyr244 of the
specificity loop, and the methyl amide with the backbone carbonyl of Gly183.
Additional interaction for 2: hydrogen bond between the carbonyl oxygen of
the benzofuran carboxylate and the backbone amide nitrogen of Thr245.
Thr245 is at the beginning of the specificity loop and conserved for some
MMPs. The observed HB interaction may effect the conformation of the
selectivity loop and therefore influence overall selectivity. However, since it is
a backbone HB we considered the lipophilic interaction of the benzofuran
within the S10 pocket to be more important.Coordinates for co-structures
deposited as 3I7G (compound 1) and 3I7I (compound 2) with the RCSB protein
data bank.
In summary, we have reported the optimization of a novel class
of non-Zn-chelating MMP-13 inhibitors with the aid of co-crystal
structural information. The hit structure was extended out from
the active site into the S10 pocket by adding an aryl group through
two different linking functionalities. Depending on the linkage dif-
ferent trends for potency and selectivity for the respective aryl
groups were observed. Compounds with excellent potency and
acceptable selectivity profiles were obtained.
References and notes
11. Biological activity of the compounds against the catalytic domain of human
MMP-1, -2, -3, -7, -8, -9, -10, -12, -13, and -14 (all MMPs purchased from
BioMol except for MMP-13 which was made in house and refolded from E-coli)
were assessed by using the EnzoLyteTM 520 Generic MMP Assay Kit (AnaSpec
1. (a)Matrix Metalloproteinases and TIMPs; Woessner, J. F., Nagase, H., Eds.; Oxford
University Press: New York, 2000; (b) Cawston, T. E.; Wilson, A. J. Best Pract.
Res., Clin. Rheumatol. 2006, 20, 983; (c) Huxley-Jones, J.; Foord, S. M.; Barnes, M.
R. Drug Discovery Today 2008, 13, 685.
2. Knäuper, V.; Will, H.; Lopez-Otin, C.; Smith, B.; Atkinson, S. J.; Stanton, H.;
Hembry, R. M.; Murphy, G. J. Biol. Chem. 1996, 271, 17124.
3. Recent review: (a) Rowan, A. D.; Litherland, G. J.; Hui, W.; Milner, J. M. Exp.
Opin. Ther. Targets 2008, 12, 1. and references therein; (b) Mitchell, P. G.;
Magna, H. A.; Reeves, L. M.; Lopresti-Morrow, L. L.; Yocum, S. A.; Rosner, P. J.;
Inc.). This kit uses
a
5-FAM/QXLTM520 fluorescence resonance 10 energy
transfer (FRET) peptide as an MMP substrate. In the intact FRET peptide, the
fluorescence of 5-FAM is quenched by QXLTM520. Upon cleavage into two
separate fragments by MMPs, the fluorescence of 5-FAM is recovered, and can
be monitored at excitation/emission wavelengths = 490 nm/520 nm. The
assays are performed in a 96-well or 384-well microplate format. Reported
IC50 values reflect an n of 2–5.