The identification of β-hydroxy carboxylic acids as selective MMP-12 inhibitors

Article abstract of DOI:10.1016/j.bmcl.2009.07.155

A new class of selective MMP-12 inhibitors have been identified via high throughput screening. Crystallization with MMP-12 confirmed the mode of binding and allowed initial optimization to be carried out using classical structure based design.

Full text of DOI:10.1016/j.bmcl.2009.07.155

Bioorganic & Medicinal Chemistry Letters 19 (2009) 5760–5763
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
The identification of b-hydroxy carboxylic acids as selective MMP-12 inhibitors
Ian P. Holmes, Simon Gaines, Steve P. Watson, Olivier Lorthioir, Adam Walker, Suzanne J. Baddeley,
Shane Herbert, Danielle Egan, Máire A. Convery, Onkar M. P. Singh, Jeffrey W. Gross, John M. Strelow,
*
Robert H. Smith, Augustin J. Amour, David Brown, Stephen L. Martin
GSK Medicines Research Centre, Gunnels Wood Road, Stevenage, Hertfordshire SG1 2NY, United Kingdom
a r t i c l e i n f o
a b s t r a c t
Article history:
A new class of selective MMP-12 inhibitors have been identified via high throughput screening. Crystal-
lization with MMP-12 confirmed the mode of binding and allowed initial optimization to be carried out
using classical structure based design.
Received 5 June 2009
Revised 30 July 2009
Accepted 31 July 2009
Available online 6 August 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
Inhibitor
MMP-12
Macrophage metalloelastase
Metalloproteinase
COPD
Rheumatoid arthritis
Angiogenesis
The matrix metalloproteinases (MMPs) are a family of zinc con-
taining endoproteinases catalyzing degradation of the extracellular
matrix (ECM). As such, they represent potential therapeutic targets
for the treatment of a diverse range of pathological conditions
including rheumatoid arthritis, atherosclerosis and chronic
obstructive pulmonary disorder (COPD).1 A growing body of evi-
dence has specifically linked MMP-12 (macrophage metalloelas-
tase) with COPD2 making the development of small molecule
MMP-12 ligands of high interest.3 Historically, MMP inhibitors
have been typically non-selective as they have relied on potent
metal binding motifs, for example, hydroxamates and reverse
hydroxamates, for the majority of their potency. As MMP activity,
and indeed activity at other related metallo-enzymes, is required
for numerous essential processes within the body, broad spectrum
metallo-enzyme inhibition is considered undesirable. As a conse-
quence we sought to identify potent and selective inhibitors of
MMP-12.
group, for example, a carboxylic acid, would result in compounds
with better selectivity profiles than those in which the binding
interaction was dominated by high affinity metal chelation.
An initial measure of selectivity was carried out by determining
potency at the closely homologous enzyme MMP-13 (Collagenase-
3). This information allowed the identification of a number of clas-
ses of MMP-12 selective compounds. Finally, a detailed kinetic pro-
file of exemplars of each of these classes was obtained and analysis
of the data using a Lineweaver–Burk plot of 1/v against 1/[S] al-
lowed classification of the modes of action (competitive, noncom-
petitive, mixed etc.).
Following these analyses a number of structures were identified
as possible synthetic starting points—detailed here is the initial
optimization of 1 (Fig. 1).
OH
O
A high throughput screen of the GSK compound collection iden-
tified a number of MMP-12 active compounds. These were refined
by removal of molecules with undesirable physiochemical proper-
ties (high Mw, clog P etc.) and compounds carrying high affinity
metal binding motifs (hydroxamates, reverse hydroxamates and
thiols). It was anticipated that the use of a weaker zinc binding
Mw=270
HBA=2
HBD=3
OH
clogP=3.19
Solubility=0.96 mg/mL at pH 7.4
Solubility=0.45 mg/ml at pH 6.2
logD=0.55 at pH 7.4
logD=1.43 at pH 6.2
MOA: Competitive/Reversible
1
MMP12 IC50=0.52 uM
MMP13 IC50=12.0 uM
* Corresponding author. Tel.: +44 01438 768015.
E-mail address: steve.l.martin@gsk.com (S.L. Martin).
Figure 1. Initial profile of the HTS hit 1.
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.07.155

I. P. Holmes et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5760–5763
5761
This encouraging initial data led us to obtain a full selectivity
OH
O
profile for racemic 1 and both of its enantiomers against the GSK
panel of metallo-enzymes and other proteinases and to initiate
crystallographic studies in order to understand the mode of bind-
ing of 1 to MMP-12. Cross screening of 1 and its constitutive enan-
tiomers revealed an encouraging broader selectivity profile. All
were inactive against a diverse range of serine, aspartyl and cys-
teine proteinases. Furthermore, high selectivity was observed
against a range of metalloproteinases (Table 1).4
OH
Figure 3. Regions of 1 chosen for modification.
Initially it was speculated that 1 was binding with MMP-12 in a
mode where both the alcohol and carboxylic acid of the b-hydroxy
acid functionality were coordinated to the catalytic zinc, however,
this proved not to be the case. Crystallographic investigations re-
vealed that only the carboxylic acid coordinated to zinc and that
the hydroxyl group is forming a hydrogen bonded interaction to
the Leu181 residue (Fig. 2).4
Though Leu181 and its flanking residues are highly conserved in
human MMPs5 it was reasoned that this hydrogen bonding interac-
tion was the key to the selectivity of 1. This interaction positions
the biaryl fragment deep in the S10 pocket of the MMP-12 where
substantial differences in homology between the MMPs exist and
previous studies have suggested selectivity can be obtained.6
On the basis of these observations two regions of the molecule
were identified for preliminary investigation via analogue synthe-
ses, occupancy of the S10 pocket and the H-bonding interaction to
Leu181 (Fig. 3).
Two types of variation in P10 were initially investigated, substi-
tution of the terminal aryl ring and complete replacement of the
biaryl fragment. Variation in the terminal aryl ring was achieved
by the use of Suzuki chemistry on iodide 5, which was prepared
from commercially available benzyl bromide 2 (Scheme 1).7
It was found that either thermal or microwave conditions could
be successfully employed to couple a diverse range of boronic acids
and esters to 5 in the presence of cesium carbonate and a palla-
dium source in DMF. The palladium source of choice was identified
as FibrecatÒ 1001 which improved the ease of reaction workup and
purification (Scheme 2).
Cross screening of a range of the compounds prepared revealed
clear SAR for substitution of the terminal aromatic ring (Table 2).
Introduction of a para substituent gave clear gains in potency
and by suitable choice of Ar it was demonstrated that selectivity
could also be varied. Crystallographic analysis4 of 6c suggested
that the source of the observed potency gain was water mediated
hydrogen bonding interactions to residues at the bottom of the
S10 pocket, specifically main chain atoms of Lys233 and Arg249
(Fig. 4).
Table 1
Variation of the complete biaryl fragment was also investigated
using a similar reaction sequence to that described in Scheme 1.
Starting from a diverse range of commercially available benzylic
halides or alcohols a number of analogues of 1 were prepared
(7a–d) and subsequently profiled (Table 4).
Inhibition of sample metalloproteinases by 1 and its enantiomers
Metalloproteinase
Inhibition IC50
Ent-1
(lM)
1
Ent-2
MMP-1
MMP-2
MMP-3
MMP-7
MMP-9
MMP-12
gp-MMP12
MMP-13
MMP-14
TACE
>98
4.52
>98
>98
NT
>98
NT
Crystallographic analysis4 of 7a and comparison of its binding
mode with that observed for 6c revealed that introduction of a con-
formational constraint prevented the biaryl system from adopting
>98
>98
10.5
0.34
1.14
9.99
36.6
>98
>250
>98
>98
19.1
0.79
4.67
36.8
>9
>98
16.5
0.52
0.91
12.0
43.5
>98
O
O
a
>98
>250
Br
OtBu
ACE
>250
I
I
NT = Not tested; all human enyzymes except gp-MMP12 = guinea pig.
2
3
b
OH
O
OH
O
c
OtBu
OH
I
I
5
4
Scheme 1. Preparation of aryl iodide 5. Reagents and conditions: (a) t-butyl
acetoacetate, NaH, n-BuLi, 0 °C to rt; (b) NaBH4, MeOH, 0 °C; (c) silica gel, toluene,
reflux.8
OH
O
OH
O
a
OH
OH
I
Ar
6
5
Scheme 2. Suzuki coupling of boronic acids/esters to iodide 5. Reagents and
conditions: (a) boronic acid (1.2 equiv), cesium carbonate (2.5 equiv), Fibrecat 1001
(10 mol %), DMF, microwave, 100 °C, 1 h.
Figure 2. The binding mode of 1 in hMMP-12 determined at 2.0 Å resolution by
X-ray crystallographic analysis. H-bond interactions are indicated by dashed lines.

5762
I. P. Holmes et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5760–5763
Table 2
Sample cross screening data for compounds 1 and 6a–f
Ar
Compound
Inhibition IC50
MMP-13
(lM)
MMP-12
0.52
MMP-1
>98
1
12.0
52.0
NC
NC
6a
3.39
>98
>98
>98
>98
6b
6c
6d
0.086
0.062
0.058
2.45
0.97
1.13
O
Figure 5. A comparison of the binding modes of 6c and 7a (determined at 2.3 Å by
resolution X-ray crystallographic analysis).in the S10 pocket of hMMP-12—6c wire,
7a rods. Note the twist in 6c compared to 7a.
S
phenyl boronic acid and t-butyl ester cleavage gave methyl ether
8. An analogous sequence acylating 4 with acetic anhydride gave
ester 9. The corresponding ketone 10 was prepared from ketone
3 by Suzuki coupling with phenyl boronic acid and subsequent t-
butyl ester deprotection. A number of corresponding sulfur ana-
logues were also prepared (Scheme 3).
6e
6f
0.055
0.181
3.76
50.2
NT
NT
N
O
NT = Not tested.
OMs
CO2H
a, b
11
12
c
O
O
S
S
OH
OEt
d
14
13
Scheme 3. Preparation of thioether 11. Reagents and conditions: (a) LiAlH4, THF,
0 °C; (b) MsCl, Et3N, CH2Cl2; (c) HSCH2CO2Et, NaOEt, EtOH; (d) LiOH, THF, H2O.
Table 3
Sample cross screening data for compounds 1, 8–10 and 14–16
O
X
OH
Figure 4. The binding mode of 6c at the bottom of S10 pocket of hMMP-12
determined at 1.7 Å resolution by X-ray crystallographic analysis.
X
Compound
Inhibition IC50
MMP-13
(lM)
MMP-12
MMP-1
CH2OH
CH2OMe
CH2OAc
C@O
S
S@O
1
8
9
10
14
15
16
0.52
3.53
54.4
7.84
11.8
4.5
12.0
>98
>98
>98
95.2
94.9
11.3
>98
>98
>98
>98
>98
>98
>98
a twisted pose along the aryl–aryl axis (Fig. 5). Though the addition
of an extra atom is sterically tolerated the restriction in rotation
leads to a suboptimal occupation of the S10 pocket.
Finally, variations of the hydroxyl group, involved in the hydro-
gen bonding interaction with Leu181, were investigated. Methyla-
tion of 4 with methyl iodide followed by Suzuki coupling with
SO2
2.49

I. P. Holmes et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5760–5763
5763
Table 4
selectivity to be obtained over a range of closely related MMPs as
the molecules binding interactions are not dominated by metal
chelation. On the basis of initial synthetic studies and crystal struc-
tures obtained this series appears to be an excellent starting point
for further optimization.
Sample cross screening data for compounds 1 and 7a–d
OH
O
Ar
OH
Ar
Compound
Inhibition IC50
(l
M)
MMP-1
MMP-12
0.52
MMP-13
Acknowledgements
We thank Pat McDonough for solubility and log D measure-
ments, Eric Hortense and Steve Jackson for analytical and prepara-
tive chiral HPLC support and Martin Read for plate based array
purifications.
1
12.0
>98
>98
>98
>98
>98
7a
7b
7c
7d
1.15
20.1
19.9
3.75
26.1
33.5
41.9
>98
References and notes
1. (a) Skiles, J. W.; Gonnella, N. C.; Jeng, A. Y. Curr. Med. Chem. 2001, 8, 425; (b)
Matter, H.; Schudok, M. Curr. Opin. Drug Disc. Dev. 2004, 7, 513.
2. (a) Shapiro, S. D. Curr. Opin. Cell Biol. 1998, 10, 602; (b) Molet, S.; Belleguic, C.;
Lena, H.; Germain, N.; Bertrand, C. P.; Shapiro, S. D.; Planquois, J.-M.; Delaval, P.;
Lagente, V. Inflamm. Res. 2005, 54, 31; (c) Demedts, I. K.; Brusselle, G. G.; Bracke,
K. R.; Vermaelen, K. Y.; Pauwels, R. A. Curr. Opin. Pharmacol. 2005, 5, 257; (d)
Daheshia, M. Curr. Med. Res. Opin. 2005, 21, 587.
3. (a) Anon, Expert Opin. Ther. Patents 2004, 14, 163.; (b) Dublanchet, A.-C.; Ducrot,
P.; Andrianjara, C.; O’Gara, M.; Morales, R.; Compère, D.; Denis, A.; Blais, S.;
Cluzeau, P.; Courté, K.; Hamon, J.; Moreau, F.; Prunet, M-L.; Tertre, A. Bioorg. Med.
Chem. Lett. 2005, 15, 3787; (c) Li, W.; Li, J.; Wu, Y.; Wu, J.; Hotchandani, R.;
Cunningham, K.; McFayden, I.; Morgan, P.; Schlerman, F.; Xu, X.; Tam, S.;
Goldman, S. J.; Williams, C.; Sypek, J.; Mansour, T. S. J. Med. Chem. 2009, 52, 1799.
4. (a) IC50 values are given as the average of at least two replicates (n P 2) with
values deviating by less than two fold across the range. Assays were conducted
using recombinant metalloproteinases from commercial sources, or generated
in-house, with fluorescently tagged peptide substrates. Details available on
request. (b) Protein for crystallization was obtained by overexpression of
MMP12 residues 106–268 in E. coli followed by refolding and standard
O
O
O
purification procedures in the presence of
a hydroxamate tool compound.
Crystals were grown with the tool compound from 100 mM Mes pH 5.2–5.6, 16–
20% PEG 5KMME and 200 mM ammonium sulfate. The crystals are P1 with
approximate cell dimensions of a = 46 Å, b = 65 Å, c = 68 Å, a = 66°, b = 84°,
g = 71°. Complexes of 1, 6c and 7a were obtained by replacement soaking of the
crystals. For all three complexes, data was collected at Daresbury synchrotron
and density for the ligand was clear and unambiguous. Final Rfactors are 0.184,
0.172, 0.152 and Rfree 0.246, 0.217 and 0.219, respectively, with pdb deposition
codes of 2wo8, 2wo9 and 2woa.
It was possible to specifically oxidize 14 using 1 equiv of mCPBA
to give the corresponding racemic sulfoxide 15 and then by treat-
ing 15 with a further equivalent of mCPBA to obtain sulfone 16.
Cross screening of these compounds was then carried out (Table 3).
Whilst the free alcohol was the most potent of the compounds
prepared it was demonstrated that a number of viable alternative
groups were tolerated.
In summary, we have identified a series of potent, selective
MMP-12 inhibitors based around a b-hydroxy acid template. The
lower affinity zinc binding group (carboxylic acid) allows excellent
5. Massova, I.; Kotra, L. P.; Fridman, R.; Mobashery, S. FASEB J. 1998, 12, 1075.
6. Terp, G. E.; Cruciani, G.; Christensen, I. T.; Jørgensen, F. S. J. Med. Chem. 2002, 45,
2675.
7. Gaines, S.; Holmes, I. P.; Watson, S. P. WO 2004/110974.
8. Jackson, R. W. Tetrahedron Lett. 2001, 42, 5163.