1-Hydroxy-2-pyridinone-based MMP inhibitors: Synthesis and biological evaluation for the treatment of ischemic stroke

Article abstract of DOI:10.1016/j.bmcl.2007.10.045

Matrix metalloproteinase-9 (MMP-9) has been implicated in the breakdown of the blood-brain barrier during cerebral ischemia. As a result, inhibition of MMP-9 may have utility as a therapeutic intervention in stroke. Towards this end, we have synthesized a series of 1-hydroxy-2-pyridinones that have excellent in vitro potency in inhibiting MMP-9 in addition to MMP-2. Representative compounds also demonstrate good efficacy in the mouse transient mid-cerebral artery occlusion (tMCAO) model of cerebral ischemia.

Full text of DOI:10.1016/j.bmcl.2007.10.045

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 409–413
1-Hydroxy-2-pyridinone-based MMP inhibitors: Synthesis
and biological evaluation for the treatment of ischemic stroke
Yue-Mei Zhang,a,* Xiaodong Fan,a Devraj Chakaravarty,a Bangping Xiang,b
Robert H. Scannevin,a Zhihong Huang,a Jianya Ma,a Sharon L. Burke,a
Prabha Karnachi,c Kenneth J. Rhodesa and Paul F. Jacksona
aEast Coast Research and Early Development, Johnson & Johnson Pharmaceutical Research and Development,
Welsh & McKean Roads, Spring House, PA 19477, USA
bMerck, 126 E. Lincoln Avenue, Rahway, NJ 07065, USA
cHoffmann-La Roche, 340 Kingsland Street, Nutley, NJ 07110, USA
Received 10 August 2007; revised 30 September 2007; accepted 5 October 2007
Available online 18 October 2007
Abstract—Matrix metalloproteinase-9 (MMP-9) has been implicated in the breakdown of the blood–brain barrier during cerebral
ischemia. As a result, inhibition of MMP-9 may have utility as a therapeutic intervention in stroke. Towards this end, we have syn-
thesized a series of 1-hydroxy-2-pyridinones that have excellent in vitro potency in inhibiting MMP-9 in addition to MMP-2. Rep-
resentative compounds also demonstrate good efficacy in the mouse transient mid-cerebral artery occlusion (tMCAO) model of
cerebral ischemia.
ꢀ 2007 Elsevier Ltd. All rights reserved.
Matrix metalloproteinases (MMPs) are a family of zinc-
dependent endopeptidases responsible for the digestion
and turnover of extracellular matrix components. Recent
studies have shown that MMPs are involved in several
neurological diseases, notably in the pathophysiology of
cerebral damage after ischemic stroke (IS).1 The up-regu-
lation of MMP-9 in a rat transient middle cerebral artery
occlusion (tMCAO) model correlates with increased
blood–brain barrier (BBB) permeability. This breakdown
in the BBB then leads to edema, hemorrhage, and infiltra-
tion of inflammatory agents, all of which cause significant
brain damage. Therefore, MMP-9 is believed to be a po-
tential therapeutic target for acute IS. The goal of our
MMP program is to discover and develop novel MMP-
2/-9 (gelatinase) inhibitors, which demonstrate in vivo
efficacy in animal models of stroke.
O
O
O
O
HO
O
O
N
H
N
H
CN
OH
1b
MMP-2,-3,-9: 34%@100uM
IC50(nM): MMP-2: 610
MMP-3: 10
1a
Zn2+
O
R2
N
O
O
linker
recognition motif
HO
P1'
N
S
O2
R
H
ZBG
R1
1-HO-2-pyridinone
6- to 8-membered
heterocycles
3
2
Figure 1. MMPIs with heterocycles as ZBGs.
ability, potential metabolic liabilities (hydrolysis and glu-
curonidation), and side effects of hydroxamates,2 there is
considerable effort to discover non-hydroxamate MMPIs
to improve ADME properties.3 MMP inhibitors typically
consist of a Zn-binding group (ZBG) and a recognition
motif that binds to subsites within the MMP active site.
Recently, hydroxypyridinones (HOPOs) and pyrones
have been investigated as novel ZBGs by Cohen’s group
and a series of full length MMP-3 inhibitors containing
these heterocyclic ZBGs were synthesized.4,5 By far, pyr-
one-based compounds showed the most promising inhib-
To date, the majority of MMP inhibitors (MMPIs) are
hydroxamate-based compounds due to their high
in vitro potency. However, because of low oral bioavail-
Keywords: Matrix metalloproteinase inhibitor; MMP-9; 1-Hydroxy-2-
pyridone; Stroke.
*
Corresponding author. Tel.: +1 609 409 3498; fax: +1 609 655
6930; e-mail: yzhang5@prdus.jnj.com
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.10.045

410
Y.-M. Zhang et al. / Bioorg. Med. Chem. Lett. 18 (2008) 409–413
itoryactivity with IC50 values ranging from 10 nM against
MMP-3 to 600 nM against MMP-2 (Fig. 1, 1a).6 Interest-
ingly, the regio isomer 1b was basically inactive agains0t
MMP-1, -2, -3, and -9,7 suggesting the position of the P1
attachment to ZBGs is critical for inhibitory activity
against MMPs.
phenoxides to give 5. Intermediate 5 can further be con-
verted into thiopyridinone 4 by treatment with NaSH.
To better define the scope of HOPOs as zinc chelators, a
‘‘fully reduced version’’ compound 6, containing 1-hy-
droxy-2-piperidinone as a ZBG, was also prepared
(Scheme 2). Sulfonylation of benzyl-protected (R)-3-
amino-1-hydroxy-2-piperidinone followed by benzyl
cleavage with neat MeSO3H gave 6.10
It has been reported that increasing the steric bulk at the
a-carbon to the ZBG slows down metabolism of the
hydroxamic acid (HA) functionality and improves the
PK properties of HA-based MMPIs.8 We hypothesized
that by linking the nitrogen of the HA to the a-carbon
to form a ring the steric bulk in the region of the HA as
well as the rigidity of the ZBG would be increased, and
thus the PK properties of the resulting MMPIs would
be improved. Based on this hypothesis and inspired by
the reported work from Cohen’s research group, we have
expanded the scope of the previously reported HOPOs
and discovered several classes of potent MMPIs utilizing
a variety of 6- to 8-membered heterocycle-derived ZBGs
(Fig. 1, 2). Here we wish to report the first series of 1-hy-
droxy-2-pyridinone-based gelatinase inhibitors contain-
ing a sulfonamide scaffold (Fig. 1, 3).
The inhibitory activity of compounds 3a–3r, 4, and 6
against MMP-2 and MMP-9 was evaluated and the re-
sults are shown in Table 1. The biaryl P01 group (3a) does
not demonstrate good activity, which is supported by
docking experiments to our MMP-9 homology model
(Fig. 2).11 Substitution on the pyridinone ring at any po-
sition does not seem favorable. IC50 values for com-
pounds 3d, 3e, and 3f increased significantly compared
to the unsubstituted 3c. In contrast, R2 substitution
plays an important role in the inhibitory activity. A sim-
ple methyl substitution on 3c improved the enzyme po-
tency 50- to 100-fold compared to 3b. The MMP-9
homology model suggests that the binding modes of
3b and 3c are considerably different and this may con-
tribute to dramatic changes in affinity for the enzyme.
Bulky and hydrophobic groups (3j and 3k) at the R2 po-
sition reduced potency. However, longer side chains (at
least a two-carbon extension) containing water-solubi-
lizing groups (3l–3o) were well tolerated. Figure 2 shows
the binding mode of compounds 3c (orange) and 3l
(white) based on docking of the molecules into the
homology model of MMP-9. The modeling indicates
that the sulfonamide linker forms a hydrogen-bonding
interaction with Leu 188 on the backbone. R2 substitu-
tion extends to the solvent exposed surface of the en-
zyme. Though SAR indicates that biaryl ether groups
fit well into the S01 subsite, a highly electron withdrawing
para-substituent (3p) on the terminal phenyl was found
to be detrimental to potency. Interestingly, comparison
of 3b and 6 suggests that the 1-hydroxy-2-piperidinone
group is still an effective ZBG. This has never been re-
ported in the literature, though 1-hydroxy-2-piperidi-
none and 1-hydroxy-azepan-2-one have been used in
the syntheses of vasopeptidase inhibitors and sidero-
phores.12 On the other hand, thiopyridinone 4 showed
very poor inhibitory activity against MMP-2 and
MMP-9. This result is in contrast to the literature re-
port,4 which found that thio-HOPOs themselves are
more potent than their oxygen analogs due to the thio-
philicity of zinc. We hypothesized that the larger thio-
pyridone ZBG hindered the optimal binding to the S10
site.
Synthesis of 1-hydroxy-2-pyridinone-based sulfonamides
3 is outlined in Scheme 1. The reaction of 3-amino-2-bro-
mopyridine with various commercially available aryl sul-
fonyl chlorides, followed by alkylation of the sulfonamide
nitrogen, gave compound 1. Oxidation of pyridine 1 in the
presence of urea–H2O2 complex (UHP) and trifluoroace-
tic anhydride gave the pyridine N-oxide in high yield.9
When R2 is bromoethyl, the substituent can be further
derivatized into a nitrogen-containing side chain to yield
2. Conversion of compound 2 into 2-methoxypyridine
N-oxide and subsequent acidic hydrolysis gave 3. Alter-
natively, the 4-bromo- or 4-fluoroaryl analog of 1 can
be transformed into the desired P01 groups under Suzuki
coupling conditions or by simple displacement with basic
R
N
Br
Br
R2
N
R
c, d
NH2
Br
a,b
N
N
S
O2
-O
N
N+
R1
Br
S
O2
R2=
R1
R1
c, g
R2
1
when R=Br or F
2
e,f
X
Br
-O
N
e,f
N+
R1
S
O2
R3
3
X = bond or O
5
h
Ph
O
S
N
S
HO
N
The selectivity against several MMPs of representative
compounds is summarized in Table 2. As expected, all
O2
4
Scheme 1. Reagents and conditions: (a) ClSO2Ph(4-R), Pry/CH2Cl2,
0 ꢁC—rt, 81%; (b) R2 halides, Cs2CO3, DMF, 70 ꢁC, 70–95%; (c)
UHP/TFAA (2.1/2), CH2Cl2, 0 ꢁC—rt, 94%; (d) 1.1 equiv Amines,
TEA, MeCN, 65 ꢁC, 58–84%; (e) 1.5 equiv NaOMe, MeOH, reflux,
54%; (f) 2 N HCl/MeOH (1/1.5), reflux, 95%; (g) when R = Br,
(OH)2BPh(4-R3), 5 mol% Pd(PPh3)4, Na2CO3 (2 M), toluene, 79%;
when R = F, HOPh(4-R3), Cs2CO3, DMF, 90 ꢁC, 60%; (h) NaSH,
DMSO/H2O, 88%.
O
O
O
H
N
a, b
NH2
.HBr
HO
BnO
S
O2
N
N
Cl
6
Scheme 2. Reagents and conditions: (a) ClSO2Ph(4-(40-Cl)PhO), TEA,
CH2Cl2, rt, 88%; (b) MeSO3H, rt, 80%.

Y.-M. Zhang et al. / Bioorg. Med. Chem. Lett. 18 (2008) 409–413
411
a
Table 1. IC50 (nM) values and SAR of compounds 3, 4, and 6
O
R2
N
O
S
O
HO
N
R3
R1c
X
R1a
R1b
3
No.
R1a
R1b
R1c
X
R2
R3
MMP-2 (nM)
MMP-9 (nM)
3a
3b
3c
3d
3e
3f
–H
–H
–H
–H
–H
–CH3
–H
–H
–H
–H
–H
–H
–H
–H
–H
–Br
–H
–H
–H
–H
–H
–H
–H
–H
–H
–CH3
–H
–H
–H
–H
–H
–H
–H
bond
O
–H
–H
–Cl
–Cl
–Cl
–Cl
–Cl
–Cl
–CH3
–H
>1000
242
5.0
>1000
315
2.4
O
–CH3
–CH3
–CH3
–CH3
–CH3
–CH3
O
66.2
662
66.4
6.1
158.5
1387
34.9
4.4
O
O
3g
3h
3i
O
O
8.3
6.5
5.0
3.6
O
–CH2CH3
Isopropyl
Benzyl
–H
–H
3j
3k
O
41.0
22.4
57.5
33.4
O
–H
N
O
3l
–H
–H
–H
–H
–H
–H
–H
–H
–H
–H
–H
–H
O
O
O
O
–Cl
–Cl
–Cl
–Cl
5.0
6.1
1.7
4.9
4.2
4.9
1.3
4.8
N
S
N
3m
3n
3o
N
O
N
O
3p
3q
3r
4
–H
–H
–H
—
–H
–H
–H
—
–H
–H
–H
—
O
–CH3
–CH3
–CH3
–CF3
–OCH3
–OCF3
—
104.5
5.6
5.7
32.6
0.87
1.8
O
O
—
—
See Scheme 1
See Scheme 2
4591
108.4
693
57.9
6
—
—
—
—
aIC50 (nM) values are means of at least two experiments.
a
Table 2. IC50 values against selected MMPs of 3c, 3l, and 3q
compounds showed high selectivity over MMP-1, which
has a shallow S01 subsite, and moderate selectivity for
Compound MMP-1 MMP-2 MMP-3 MMP-9 MMP-13
3c
3l
>1000
>1000
>1000
5.0
5.0
5.6
56.5
62.9
79.1
2.4
4.2
2.5
4.0
6.8
3q
0.87
a IC50 (nM) values are means of at least two determinations.
Table 3. Pharmacokinetics parametera of 3c in rats
AUCb
(lg-h/mL) (L/kg) mL/min/Kg
Vss
CLb
b
b
b
Compound C0
T1/2
(lM) (h)
3c
8.4 47
179 1.75 0.5
a Mean value of four animals (rat) at 2 mg/kg iv dose.
b C0, concentration at T = 0; T1/2, apparent elimination half-life; AUC,
area under the concentration–time curve; Vss, volume of distribution
at steady state; CL, systemic clearance.
MMP-9 vs MMP-3 (20- to 40-fold). Compounds 3c,
3l, and 3q are also potent MMP-13 inhibitors.
The pharmacokinetics of 3c was studied in rats following
intravenous dosing (Table 3). Compound 3c demon-
strates an excellent PK profile with a long half-life. This
compound also exhibited a high plasma concentration
in mice at an oral dose of 10 mg/kg dose (34 lM at the
1 h time point).
Figure 2. Compounds 3c (orange) and 3l (white) docked in homology
model of MMP-9.

412
Y.-M. Zhang et al. / Bioorg. Med. Chem. Lett. 18 (2008) 409–413
treated group. However, this enhancement in brain
85
80
75
70
a
vehicle
water content was partially prevented after drug treat-
ment, while no difference was observed in the non-ische-
mic hemisphere between the drug-treated and vehicle-
treated groups. Furthermore, even after normalizing
for the water content of the non-ischemic hemisphere,
drug treatment was effective in reducing the percentage
increase of brain water content, as shown in Figure
3b. These results indicate that 3c could reduce brain ede-
ma induced by focal ischemia in mice.
Compd 3c
*
In summary, a series of novel 1-hydroxy-2-pyridinone-
based sulfonamide inhibitors of MMP-2/-9 were de-
signed and synthesized. The optimized compounds pos-
sess nanomolar potency in the enzyme assay and
demonstrate excellent pharmacokinetic profiles in rats.
Compound 3c reduced early brain edema after transient
ischemia in mice and is also orally bioavailable. MMP
inhibitors with other heterocyclic 6- to 8-membered ring
ZBGs will be reported in due course.
Ischemic
Hemisphere
Non-ischemic
Hemisphere
b
5
4
3
2
1
0
Acknowledgments
The authors thank the Chem/Bio support group for
the ADME tests and Dr. Mark J. Macielag for the kind
review and helpful comments.
*
Supplementary data
Protocols for MMP enzyme assays, in vivo tMCAO
studies, and the synthetic details. Supplementary data
associated with this article can be found, in the online
version, at doi:10.1016/j.bmcl.2007.10.045.
Vehicle
6mg/Kg
Figure 3. 3c treatment reduced brain edema after focal ischemia in C57
Black/6 mice (20–25 g, Charles River). Animals were subjected to
transient MCA occlusion for 2 h. 3c (n = 9) or vehicle (5 ml/kg, n = 10)
was given intravenously as a bolus immediately after ischemia and at
2 h following ischemia. Brain edema was evaluated by wet/dry method
at 5 h after MCA occlusion. Data are expressed as means SEM.
*p < 0.05 as compared to vehicle-treated group by Student’s t-test.
References and notes
1. Rosenberg, G. A.; Estrada, E. Y.; Dencoff, J. E. Stroke
1998, 29, 2189.
2. (a) Aranapakam, V.; Davis, J. M.; Grosu, G. T.; Baker, J.;
Ellingboe, J.; Zask, A.; Levin, J. I.; Sandanayaka, V. P.;
Du, M.; Skotnicki, J. S.; DiJoseph, J. F. J. Med. Chem.
2003, 46, 2376; (b) Coussens, L. M.; Fingleton, B.;
Martrisian, L. M. Science 2002, 295, 2387.
3. For a recent review, see: Jacobsen, F. E.; Lewis, J. A.;
Cohen, S. M. ChemMedChem 2007, 2, 152; Breuer, E.;
Frant, J.; Reich, R. Expert Opin. Ther. Pat. 2005, 15, 253;
For related references, see: Eriksson, A.; Lepisto, M.;
Lundkvist, M.; Munck af Rosenschold, M.; Stenvall, K.;
Zlatoidsky, P. WO 03/040098 A1; Ma, D.-W.; Wu, W.-G.;
Qian, J.; Lang, S.-H.; Ye, Q.-Z. Chin. J. Chem. 2001, 19,
286.
4. Puerta, D. T.; Cohen, S. M. Inorg. Chem. 2003, 42, 3423;
Puerta, D. T.; Lewis, J. A.; Cohen, S. M. J. Am. Chem.
Soc. 2004, 126, 8388.
5. Puerta, D. T.; Cohen, S. M.; Lewis, J. A. US 2007/0117848
A1.
6. Puerta, D. T.; Mongan, J.; Tran, B. L.; McCammon, J. A.;
Cohen, S. M. J. Am. Chem. Soc. 2005, 127.
7. Yan, Y.-L.; Cohen, S. M. Org. Lett. 2007, 9, 2517.
8. Matter, H.; Schudok, M. Curr. Opin. Drug Discovery Dev.
2004, 7, 513.
The efficacy of compound 3c on the attenuation of brain
injury after IS was evaluated in the mouse tMCAO
model.13 Focal cerebral ischemia of male C57 Black/6
mice was induced by the occlusion of the middle cerebral
artery (MCA) and then subjected to reperfusion after
2 h. Compound 3c (6 mg/kg) or an equal volume of
the vehicle (20% PEG400 + 80% of 20% solutol in
dH2O, 5 ml/kg) was administered intravenously as a bo-
lus immediately after ischemia and then again immedi-
ately after reperfusion. Five hours after ischemia,
animals were euthanized and the brain quickly removed.
Brain edema was evaluated by the wet/dry method.14
The water content in the two hemispheres of the brain
tissue was calculated as follows: 100 · [(wet weight ꢀ
dry weight)/wet weight] (%). The percentage increase
of the brain water content after ischemia was calculated
by the difference in water content between the ipsilateral
and contralateral hemispheres. Figure 3a shows, that
after focal ischemia, the brain water content increased
in the ischemic hemisphere as compared to that of the
corresponding non-ischemic hemisphere in the vehicle-
9. Caron, S.; Do, N. M.; Sieser, J. E. Tetrahedron Lett. 2000,
41, 2299.

Y.-M. Zhang et al. / Bioorg. Med. Chem. Lett. 18 (2008) 409–413
413
10. Dong, L.; Miller, M. J. Org. Chem. 2002, 67, 4759.
11. The homology model of MMP9 was build using a crystal
structure of MMP2 and using the software Prime and
refined using Macromodel. The compounds were docked
into the active site using the software GLIDE (Glide 3.5,
Prime 1.4, Macromodel from Schrodinger, New York.
2005). No scaling factors were applied to the van der
Waals radii. Default settings were used for all the
remaining parameters. The top 20 poses based on Glide-
score were energy-minimized with Macromodel using the
OPLS-AA force field, and the refined poses reranked using
Glidescore.
12. Walz, A. J.; Miller, M. J. Org. Lett. 2002, 4, 2047; Xu, Y.;
Miller, M. J. J. Org. Chem. 1998, 63, 4314; Hu, J.; Miller,
M. J. J. Am. Chem. Soc 1997, 119, 3462.
13. For the protocol for tMCAO model, see supplemental
material and the following: (a) Zea-Longa, E.; Weinstein,
P. R.; Carlson, S.; Cummin, R. W. Stroke 1989, 2084; (b)
Huang, Z.; Huang, P. L.; Panahian, N.; Dalkara, T.;
Fishman, M. C.; Moskowitz, M. A. Science 1994, 265,
1883.
14. Hara, H.; Huang, P. L.; Panahian, N.; Fishman, M. C.;
Moskowitz, M. A. J. Cereb. Blood Flow Metab. 1996, 16,
605.