Extra binding region induced by non-zinc chelating inhibitors into the S1 subsite of matrix metalloproteinase 8 (MMP-8)

Article abstract of DOI:10.1021/jm801166j

The mode of binding and the activity of the first two non-zinc chelating, potent, and selective inhibitors of human neutrophil collagenase are reported. The crystal structures of the catalytic domain of MMP-8, respectively complexed with each inhibitor, r

Full text of DOI:10.1021/jm801166j

1040
J. Med. Chem. 2009, 52, 1040–1049
Extra Binding Region Induced by Non-Zinc Chelating Inhibitors into the S1′ Subsite of Matrix
Metalloproteinase 8 (MMP-8)†
Giorgio Pochetti,*,‡ Roberta Montanari,‡ Christian Gege,*,§,3 Carine Chevrier,§ Arthur G. Taveras,|,# and Fernando Mazza‡,
Istituto di CristallografiasCNR, Area della Ricerca Roma 1, Via Salaria Km.29, 300, I-00016 Monterotondo Stazione, Roma, Italy, Alantos
Pharmaceuticals AG, Im Neuenheimer Feld 584, D-69120 Heidelberg, Germany, Alantos Pharmaceuticals Inc., RiVerside Technology Center,
840 Memorial DriVe, Cambridge Massachusetts 02139, Dipartimento di Scienze della Salute, UniVersita` di L’Aquila, I-67010, L’Aquila, Italia
ReceiVed September 17, 2008
The mode of binding and the activity of the first two non-zinc chelating, potent, and selective inhibitors of
human neutrophil collagenase are reported. The crystal structures of the catalytic domain of MMP-8,
respectively complexed with each inhibitor, reveals that both ligands are deeply inserted into the primary
specificity subsite S1′, where they induce a similar conformational change of the surrounding loop that is
endowed with the main specificity determinants of MMPs. Accord to this rearrangement, both inhibitors
remove the floor of the pocket formed by the Y227 side-chain, rendering available an extra binding region
never explored before. The present data show that potent and more selective inhibitors can be obtained by
developing ligands able to interact with the selectivity regions of the enzyme rather than with the catalytic
zinc ion, which is the common feature of all MMP members.
Introduction
substrate-like fragment designed to fit the S1′ primary specificity
subsite and adjacent subsites.21
Matrix metalloproteinases (MMPsa) belong to a family of
zinc-endopeptidases that use the electrophilic zinc ion to cleave
the constituents of the extracellular matrix.1 These enzymes are
involved in many physiological processes such as ovulation,
embryogenesis, angiogenesis, cellular differentiation, and wound
healing.2,3
Under normal physiological conditions, endogenous tissue
inhibitors of MMPs (TIMPs) control their activity.4 Overex-
pression of MMP activity, or inadequate control by TIMPs, have
been observed and associated with a variety of pathological
conditions such as psoriasis,5 multiple sclerosis,6,7 osteoarthritis,8
rheumatoid arthritis,9,10 osteoporosis,11,12 Alzheimer’s disease,13
tumor growth and metastasis,4,14,15 and unwanted degradation
of extracellular proteins.
The development of low molecular weight synthetic inhibitors
of MMPs is one approach to the therapeutic treatment of these
pathologies.16-18 Many inhibitors have been synthesized and
their mode of binding, in the active site of MMPs, have been
determinedbyX-raycrystallographyandNMRspectroscopy.17,19,20
These inhibitors generally include a zinc binding group (ZBG),
capable to efficiently chelate the catalytic zinc ion, bound to a
Hydroxamate is considered the most effective ZBG because
it forms five-membered chelates and two additional H-bonds
with the enzyme.22-24 Hydroxamates, however are affected by
lack of selectivity,25 not only toward the other members of the
same family but even to other physiologically important
metalloenzymes. Moreover, they show poor pharmacokinetic
properties26 and may cause toxicity resulting from metabolic
activation of hydroxylamine.27 Even though alternative ZBGs
have been investigated,28-34 no MMP inhibitor has emerged
on the market, with the sole exception of doxycycline (Periostat),
an antibiotic which nonselectively inhibits MMPs.35
To overcome the nonselective toxicity, novel MMP inhibitors,
which do not bind the catalytic inc ion, have been designed.36,37
Chen et al.38 have reported micromolar inhibitors of MMP-13,
occupying the S1′ subsite without interacting with the catalytic
zinc ion. Engel et al.39 optimized, on the structural basis, highly
selective pyrimidine dicarboxamide inhibitors of MMP-13
characterized by the absence of interactions with the catalytic
zinc ion. Morales et al.40 published the crystal structures of two
nonzinc chelating selective inhibitors, respectively complexed
with the catalytic domain of MMP-12. Very recently, Johnson
et al.41 and Li et al.42 showed that a MMP-13 selective inhibitor
is absent of musculoskeletal syndrome toxicity and reduces
cartilage damage in vivo. Some of these inhibitors are located
halfway down the S1′ subsite but the most potent ones39,41,42
extend deeply into the pocket without removing the floor formed
by the F252 side-chain (corresponding to Y227 in MMP-8).
†
The refined atomic coordinates and processed structure factors ampli-
tudes have been deposited in the Brookhaven Protein Data Bank (PDB)
with the following codes: 3DNG, MMP-8:(S)-2; 3DPE, MMP-8:1 (mono-
clinic form); 3DPF, MMP-8:1 (orthorhombic form).
* To whom correspondence should be addressed. For G.P.: phone,
0039.06.90672627; fax, 0039.06.90672630; E-mail: giorgio.pochetti@ic.cnr.it.
For C.G.: phone, 0049.6221.7390181; fax, 0049.6221.7390139; E-mail,
christian.gege@web.de.
‡
At present, no selective inhibitors of MMP-8 have been
discovered. MMP-8 is the most effective collagenase for type
1 collagen and the most active on neutrophils, suggesting a
central role in the infiltration of neutrophils. Although MMP-8
is considered as an antitarget for the treatment of cancer,43
potential indications for the use of selective MMP-8 inhibitors
could be in the field of acute liver failure, where MMP-8
deficient mice are resistant to TNF-R induced lethal hepatitis44
or multiple sclerosis,45 and other areas including inflammation
or cancer progression.46 The use of such inhibitors could help
Istituto di CristallografiasCNR.
Alantos Pharmaceuticals AG.
Alantos Pharmaceuticals Inc.
§
|
Dipartimento di Scienze della Salute, Universita` di L’Aquila.
Present address: Biogen Idec Inc., 14 Cambridge Center, Cambridge
#
Massachusetts 02142.
3
Present address: X2-Pharma GmbH, Im Neuenheimer Feld 584,
D-69120 Heidelberg, Germany.
a Abbreviations: MMP, matrix metalloproteinase; TIMP, tissue inhibitor
of MMP; ZBG, zinc binding group; AH, acetohydroxamate; EDCI, 1-(3-
dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride; HOAt, 1-hy-
droxy-7-azabenzotriazole; NMM, N-methylmorpholine.
10.1021/jm801166j CCC: $40.75
2009 American Chemical Society
Published on Web 01/27/2009

MMP-8 Non-Zinc-Chelating SelectiVe Inhibitors
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 4 1041
Scheme 1. Synthesis of Compounds 1 and 2a
Figure 1. Chemical structure of 1 and 2.
to determine the relevance of the MMP-8 enzyme as a novel
therapeutic target for these diseases.
a Reagents and conditions: (a) CNCO2Et, 4 N HCl ·dioxane, 50 °C; (b)
1 N LiOH; (c) 6, EDCI, HOAt, NMM, DMF; (d) CNCH2COOEt, AcOH,
NH4OAc, reflux; (e) sulphur, HNEt2, MeOH, 50 °C; (f) CNCO2Et, 4 N
HCl·dioxane, 50 °C; (g) 1 N LiOH (2 equiv); (h) 12, EDCI, HOAt, NMM,
DMF; (i) 1 N LiOH; (j) NH4Cl, EDCI, HOAt, NMM, DMF.
Here we report the activity and mode of binding of the first
two nonzinc chelating inhibitors of human neutrophil collage-
nase.47 Both ligands inhibit MMP-8 with activity in the low
nM range and are highly selective versus several other MMPs.
The crystal structures of the complexes show that both ligands
deeply insert into the primary specificity pocket S1′ induce a
dramatic conformational change of the loop delimiting the
pocket, removing the floor formed by the aromatic side-chain
of Y227, and disclose an extra binding region never explored
in other MMP-8 complexes. Their high selectivity profile can
be attributed to the conformational change induced to the S1′
loop, which is known to present a considerable variation in the
size, shape and composition among the MMPs.
unchanged. In all cases, we have observed a distorted tetrahedral
coordination, with the largest angle involving the water and
H207. The zinc ion does not interact with the inhibitors, the
shortest distances between the Zn ion and the inhibitor atoms
being 4.5 Å (Figure 2).
Moreover, the electron density found for the water oxygen
in the various complexes is of spherical or elipsoidal shape.
This observation can be related to the catalytic role played by
the water, requiring its possible mobility. In fact, the water
should be activated by the catalytic glutamate prior the
nucleophilic attack to the substrate scissile peptide bond.
Inhibitor Mode of Binding. The chemical formulas of the
inhibitors are shown in Figure 1. Their central scaffold, formed
by two fused-ring systems separated by a methylencarboxamide
group, presents similar size and shape. Main differences are
given by the squaramide moiety bound to one terminal end of
1 and by the carbamoylmethyl substituent, attached to the
asymmetric carbon atom at the opposite end of 2.
Two crystal forms, one monoclinic and the other orthorhom-
bic, respectively containing one and two enzyme-inhibitor
complexes per asymmetric unit, have been found when we
cocrystallized MMP-8 with 1 (MMP-8:1 complex). A mono-
clinic form, containing two complexes per asymmetric unit, was
obtained for the S enantiomer of 2 (MMP-8:(S)-2 complex).
The cocrystallization trials for obtaining the MMP-8:(S)-2
complex were performed by adding the racemic mixture of 2
to the enzyme. In the first attempt to fit the ligand in the Fourier
difference map, electron density was clearly evident for the
carbamoylmethyl group as substituent of the asymmetric carbon
with S configuration.
Results and Discussion
Chemistry. Compounds 1 and 2 (Figure 1) were prepared
according to the synthetic sequence outlined in Scheme 1. Thus,
cyclization of commercially available ethyl 2-amino-4,7-dihy-
dro-5H-thieno[2,3-c]thiopyran-3-carboxylate 6,6-dioxide 3 with
ethyl cyanoformate in hydrochloric acid afforded ester 4, which
was deprotected with excess of aqueous lithium hydroxide to
give acid 5, which was finally coupled with amine 650 using
1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride
(EDCI), 1-hydroxy-7-azabenzotriazole (HOAt) and N-methyl-
morpholine (NMM) to yield 1 after aqueous workup. Knoev-
enagel condensation of (2-oxo-cyclohexyl)-acetic acid ethyl ester
7 with ethylcyanoacetate afforded intermediate 8, which was
transformed via Gewald reaction with sulfur to the thiophene
derivative 9. Cyclization as described above furnished ester 10,
which was selectively saponified with 2 equiv of aqueous lithium
hydroxide to afford the monoacid 11. Coupling with 6-ami-
nomethyl-4H-benzo[1,4]oxazin-3-one building block 1250 fur-
nished derivative 13, which was saponified to acid 14 and finally
coupled with ammonium chloride using EDCI, HOAt, and
NMM to afford inhibitor 2 as racemate.
Zinc Coordination. The enzyme essential catalytic machinery
is formed by the Zn ion, coordinated by the three H residues
belonging to the binding motif HEXXHXXGXXH, and the
water molecule bridged with the glutamate of the same motif.
A comparison of the active site of our complexes with that
of the uninhibited enzyme48 shows that the Zn coordination is
The mode of binding of each inhibitor in the independent
units is similar, and the central scaffold of the two inhibitors
gives rise to similar interactions with the enzyme in all
complexes.
Both inhibitors are deeply inserted into the primary specificity
pocket S1′, where they adopt an extended conformation with

1042 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 4
Pochetti et al.
Figure 2. Zinc coordination: (A) in the complex MMP-8:1, and (B) in the complex MMP-8:(S)-2.
Figure 3. H-bonds (dashed lines) formed by (A) 1 and (B) (S)-2 in the active site of MMP-8.
the mean planes of the two fused-ring systems almost perpen-
dicular to each other. The binding modes found in MMP-8:1
and MMP-8:(S)-2 crystals are reported in parts A and B of
Figure 3, respectively, and their similar interactions are the
following. The aromatic cloud of the H197 imidazole ring gives
rise to a π-stacking interaction, respectively with that of the
isoquinoline of 1 and the benzoxazinone of (S)-2. The NH and
CO of the central carboxamide group of both inhibitors form
H-bonds with the L214CO and A220NH, respectively. The NH
and CO groups of the 3H-pyridin-4-one ring of both ligands
engage H-bonds with the A220CO and R222NH groups,
respectively. The charged guanidinium group of R222 stacks
onto the aromatic cloud of the 3H-thieno[2,3-d]pyrimidin-4-
one system of both inhibitors, giving rise to cation-π interac-
tions. The extended conformation of the R222 side-chain is also
favored by the H-bond between its guanidinium group and the
G212CO belonging to the facing loop.
The different binding interactions involving the terminal parts
of the two inhibitors are as follows. The NH groups of L160
and A161 residues, both belonging to the antiparallel ꢀ-strand
forming the upper rim of the active site, give rise to a bifurcated
H-bond with one carbonyl of the squaramide ring of 1, whereas
they engage at the same way the carbonyl of the benzoxazinone
ring of (S)-2 through a bridged water molecule. It is worth noting
that the substrate-enzyme H-bond with L160NH is always
conserved in the crystal complexes of MMP-8 available on PDB.
The CO group of P217, a residue located at the entrance of the
S1′ pocket, is H-bond acceptor from the amino group of the
squaramide ring of 1, and from the NH of the benzoxazinone
ring of (S)-2. The NH of S228, a residue at the end of the S1′
loop, engages both sulfonyl oxygens of 1 through a bifurcated
H-bond. The terminal carbamoyl group of (S)-2, pointing toward
the central part of the S1′ loop, forms a complex H-bonding
network involving R222CO and Y227OH. The inhibitor NH2
is H-bond donor to the R222CO. The Y227OH is H-bond donor
to the T224CO and acceptor of a bifurcated H-bond from both
the NH and CO groups of the inhibitor carbamoyl moiety; each
branch of the bifurcated H-bond is bridged by one water
molecule. The similar relative position of the central scaffold
of the two inhibitors can be appreciated by the superposition
of their crystal complexes shown in Figure 4.
Inhibitor Induced Conformational Changes at the S1′
loop. A comparison between the crystal form of the uninhibited
MMP-8 and the inhibitor-bound complexes reported here reveals
important structural differences regarding essentially the S1′
specificity loop. The first part of this loop, containing the third
Zn-coordinated H207 and two consecutive ꢀ-turns (210-216)
with the strictly invariant M215, remains practically unaltered.
Large conformational changes are induced by both inhibitors
on the segment 219-229, separating the tube-like crevice from
bulk water. The largest CR displacements induced by both
inhibitors occur for the sequence R222-N226, as reported in
Table S1 (Supporting Information). As a consequence, the Y227
side-chain of both complexes has been pushed from the position
occupied in the uninhibited enzyme. At this regard, we have
examined the position of Y227 in the crystals of all the other
inhibited MMP-8, available on the PDB. In all cases, this residue
practically occupies the same position as that of the uninhibited
MMP-8, forming the floor of the S1′ pocket, and the S1′ loop
maintains essentially the same conformation (Figure 5). It should

MMP-8 Non-Zinc-Chelating SelectiVe Inhibitors
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 4 1043
Conformational changes of such extent have never been
observed in the previous inhibited crystal complexes of MMP-8
available on the PDB. Our ligands constitute the first example
of the most profound occupation of the MMP-8 S1′ specificity
pocket.
The inhibitors disclosed here show activities in the nanomolar
range against MMP-8, as reported in Table 1, although their
binding is limited to the S1′ hydrophobic pocket, without
interacting with the catalytic zinc ion.
The 8× potency difference between 1 and 2 could be of 16×
considering that the IC50 reported for 2 represents that of the
racemic mixture. This difference can be rationalized as follows.
As shown in Figure 4, both inhibitors induce a similar
conformational change of the specificity loop, with the rear-
rangement of the Y227 side-chain toward a “floor-out” confor-
mation, but, at variance with 1, the carbamoyl group of 2
protrudes even further toward the extra binding region. There,
the S enantiomer stabilizes better than 1 the opening to the extra
binding region through the H-bonding network with Y227OH,
as previously described. The new conformation of the specificity
loop is further on stabilized by a H-bond between the Y227OH
and the T224CO (see Figure 3B and Figure 7).
Figure 4. Superposition between MMP-8:1 (inhibitor in yellow color)
and MMP-8:(S)-2 (inhibitor in cyan color). The protein trace (gray color
for MMP-8:1) of both complexes is shown.
Selectivity against Other MMPs. The selectivity profile of
our inhibitors can be deduced from inspection of Table 1, where
their activities against other MMPs are reported. To investigate
the role of the S1′ specificity loop for MMP-8 selectivity, we
aligned the amino acid sequence of this loop with that of other
MMP members of known 3D structure, as reported in Table 2.
Conformational restrictions appear obvious for those specific-
ity loops shorter than that of MMP-8. In particular, MMP-1,
MMP-2, and MMP-9 present two consecutive deletions in
correspondence of positions 224 and 225 of MMP-8, where the
largest CR displacements occur in the reported complexes. For
MMP-7, although there is one deletion only for the above-
mentioned position 225, the conformationally restricted Pro,
occurring just two residues before that deletion, causes a kink
and consequently a shrinking of the S1′ loop.
Figure 5. Side view of the S1′ specificity loop obtained by superposi-
tion of the uninhibited MMP-8 (PDB code 2OY4) and 17 inhibited
complexes of MMP-8 available on PDB (PDB codes: 1JAN, 1JAP,
1I73, 1I76, 1ZVX, 1ZS0, 1MNC, 1MMB, 1JJ9, 1JAQ, 1JAO, 1A86,
1KBC, 1BZS, 1JH1, 1ZP5, 2OY2).
Conformational restrictions for MMPs with similar or longer
loops should be based on the equivalent positions, which may
not be able to adopt the complementary geometry required to
form favorable interactions between the specificity loop and the
inhibitors.
be noted that none of the previous known inhibitors was long
enough to approach the side-chain of Y227. Figure 6A
represents a superposition between the MMP-8 complexed with
a phosphonate inhibitor (PDB accession code 1ZVX) represent-
ing the deepest P1′ group (cyan color), as known before the
present finding, and the MMP-8:(S)-2 complex (yellow color).
The comparison with MMP-8:1 is very similar, and it has been
omitted from Figure 6A for clarity. There, it can be noted that
the conformational change of Y227 side-chain is determined
by the tricyclic ring system in order to avoid steric overlapping
between their rings. The switching of the Y227 side-chain is
realized with a concerted rearrangement of the S1′ loop involving
the region R222-Y227. In this way, an extra-binding region at
the bottom of the primary specificity subsite S1′ is rendered
accessible, as shown in Figure 6B. There, the van der Waals
surface of (S)-2 (blue color) and that of the phosphonate inhibitor
mentioned above are reported (yellow color). The new position
of Y227 side-chain shows that the extra-binding region becomes
available whenever the switching of this side-chain occurs.
Therefore, the Y227 residue plays the role of “selective
gatekeeper”, rendering accessible the extra binding region of
the S1′ specificity subsite.
Comparison among MMP-8, MMP-13, and MMP-3. The
ligands here described are also active against MMP-13, for
which selective inhibitors, deeply inserted into the S1′ pocket
and not interacting with the zinc ion, have been described.39,41
In this regard, we have checked the putative occupation of our
inhibitors into the S1′ subsite of MMP-13. We deduced that it
could take place if the F252, equivalent to the Y227 in MMP-
8, and the loop region 247-252 would adopt a conformational
change analogous to that found in our MMP-8 complexes.
Parts A and B of Figure 8 show a CR superposition of MMP-
8:1, MMP-8:(S)-2 and MMP-13 bound to nonzinc chelating
inhibitors (PDB codes 1XUR and 2OW9). While the central
scaffold of the MMP-13 inhibitors leaves unperturbed the “floor-
in” conformation of F252, the bulkier tricylic ring system of
MMP-8 inhibitors causes the switching of Y227 toward the
“floor-out” conformation. Moreover, a cation-π interaction is
realized by the aromatic scaffold of 1 and (S)-2 with the R222
of MMP-8. It should be noted that in the sequence alignment
among MMPs reported in Table 2, R222 is the only positively
charged residue in this position. The shorter and uncharged side-
chain of T226 in MMP-13, equivalent to R222 in MMP-8, is
unable to give the same interaction. The lack of the cation-π

1044 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 4
Pochetti et al.
Figure 6. (A) superimposed crystal structures of MMP-8 complexed, respectively with the phosphonate inhibitor (PDB code 1ZVX, cyan color)
and the compound (S)-2 (yellow color); (B) extra binding region deducible as difference between the van der Waals surfaces of the inhibitors
shown in (A). The vdW surface is blue colored around (S)-2 and yellow colored around the phosphonate inhibitor. The Y227 side-chain in the two
complexes is also presented.
Table 1. IC50 Values (nM) of 1 and 2 against MMPsa
MMP-1
MMP-2
MMP-3
MMP-7
MMP-8
MMP-9
MMP-12
MMP-13
MMP-14
1
2
>10000
>10000
>2500
>10000
<100
>2500
>10000
>10000
57
7.4
>10000
>10000
>5000
>10000
<25
<25
>10000
>10000
a Errors are in the range of 5-10% of the reported value. IC50 values of 2 have been determined for the racemic mixture.
inspection of the inhibited complexes available on PDB shows
that the S1′ loop of MMP-3 can adopt two different arrange-
ments. In one case, an additional ꢀ-turn (PDB code: 2D1O), or
a γ-turn (PDB code: 1B32) can be formed in the region
224-227. These secondary structure elements lack in the
equivalent region (220-222) of MMP-8, where A220 and R222
form three key H-bonds with our inhibitors, as shown in the
superposition given in Figure 9A. In the other case (PDB code:
1CAQ, 1CIZ), the region 224-227 of MMP-3 has a conforma-
tion similar to that equivalent of MMP-8 and therefore capable
to form the key H-bonds with our ligands. However, the
additional ꢀ-turn is shifted forward at the region 228-231,
equivalent to that 224-227 of MMP-8, where the largest CR
displacements have been induced by our inhibitors (see Figure
9B). This ꢀ-turn and the additional residue at this position could
preclude the optimal geometry of the loop required to give
complementary interactions with (S)-2, particularly, the complex
H-bonding network formed by the carbamoyl group of (S)-2,
which points toward the central part of S1′ loop (see Figure
3B). This could explain the activity of 2 significantly lower than
1 against MMP-3.
Figure 7. Details of the MMP-8 specificity loop: CR superposition of
the complexes of MMP-8 (gray) with 1 (yellow color) and (S)-2 (cyan
color). MMP-8 trace is shown in magenta for the complex with (S)-2.
Table 2. Structure-Based Sequence Alignment of Known Crystal
Structure MMPsa
219 220
s
221 222 223 224 225 226 227 228 229
In conclusion, the longer S1′ loop of MMP-3, rather than
increase its internal flexibility, gives rise to a structural
organization that can hinder the optimal geometry required by
productive enzyme-ligand interactions.
MMP-8
MMP-1
MMP-2
MMP-3
MMP-7
MMP-9
MMP-12
MMP-13
MMP-14
Y
Y
Y
Y
Y
Y
Y
Y
Y
A
T
T
H
G
R
K
T
s
s
s
S
N
s
s
s
s
F
F
Y
L
G
F
Y
Y
W
R
S
T
T
D
T
V
T
M
E
G
K
D
P
T
s
s
L
Q
s
I
S
s
s
T
s
s
N
S
N
D
N
R
N
G
T
Y
V
F
F
F
P
F
F
F
S
Q
R
R
K
P
R
M
V
L
L
L
L
L
L
L
L
L
E
D
G
D
Conclusions
K
T
H
N
Q
E
Inhibitors without the ZBG typically risk losing potency while
gaining selectivity. In fact, zinc binding, the common feature
of all MMP members, follows a rather universal recognition
motif. Although lacking the ZBG, the inhibitors disclosed here
show activity against MMP-8 in the nanomolar range. Moreover,
their high selectivity profile can be attributed to the conforma-
tional change induced to the S1′ loop. In fact, it is known that
this loop presents a considerable variation in the composition,
size, and shape among the MMP family members. Therefore,
the present data show that potent and more selective inhibitors
a The region shown corresponds to the S1′ loop and the numbering refers
to the MMP-8.
interaction could be compensated by the presence of the charged
residue K228 of MMP-13 (T224 in MMP-8) capable to give
electrostatic interactions, as observed in the structures of the
complexes of MMP-13 with selective inihibitors (PDB accession
codes 2OW9 and 2OZR).
Our inhibitors show a reduced activity toward MMP-3, whose
S1′ loop is one residue longer than that of MMP-8. The

MMP-8 Non-Zinc-Chelating SelectiVe Inhibitors
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 4 1045
Figure 8. Comparison of MMP-8/MMP-13 complex structures: CR superposition of the complexes of MMP-8 (gray) with 1 and (S)-2 (cyan) and
MMP-13 (green) with (A) a pyrimidine dicarboxamide inhibitor (yellow) (PDB code: 1XUR), and (B) a selective inhibitor (yellow) (PDB code:
2OW9). In both figures, Y227 of MMP-8 is shown in pink, the equivalent F252 of MMP-13 in green.
Figure 9. Comparison of MMP-8/MMP-3 complex structures: side view of the S1′ specificity loop obtained by superposition of MMP-8:(S)-2
(protein trace in green color, (S)-2 in yellow) and (A) MMP-3 (PDB code 2D1O, cyan color), (B) MMP-3 (PDB code 1CAQ, purple color).
can be obtained by targeting selective regions of the enzyme,
rather than the catalytic zinc ion. In this way, the inhibitor can
select the appropriate complementary sequence from the con-
formational ensemble of the apo enzyme.
In light of the present data, it could be argued that inhibitor
binding of MMP-8 is not a simple diffusion process. On the
contrary, inhibition takes place through an induced fit mecha-
nism operating on the loop surrounding the S1′ subsite, which
defines the shape and the size of the pocket.49
Finally, the sequence alignment of S1′ specificity loop among
MMPs shows that MMP-8 differs from the other members for
two residues: the unique positively charged R222 and the Y227
forming the floor. A phenylalanine residue, unable to form a
H-bond, replaces the tyrosine residue in the other MMP
members. The key position in the loop of these two residues
offers interesting hints for designing new and more selective
inhibitors of MMP-8.
Experimental Section
This induced fit mechanism operating on the S1′ loop opens
an extra binding region of the pocket by the switching of the
Y227 side-chain, which plays the role of a selective gatekeeper.
This extra binding region has never been explored in the
previously examined complexes of MMP-8 reported on the
PDB. Moreover, CR displacements of the S1′ loop residues as
large as those reported have never been evidenced in any
MMP-8 complex with known inhibitors and the removal of the
floor of the S1′ pocket has never been observed in any MMP-
complex. Further molecular modeling studies, aimed to increase
ligand-enzyme interactions, could consider the displacement
of the two water molecules, bridged in the bifurcated H-bond
between the terminal carbamoyl group of (S)-2 and the hydroxyl
of Y227 in order to form a direct H-bond with the Y227 side-
chain.
Chemical Methods. 1H NMR spectra were recorded on a Bruker
Avance (250 MHz) instrument at 300 K in CDCl3, or DMSO-d6.
Chemical shifts are reported in δ values (ppm); the hydrogenated
residues of deuterated solvent were used as internal standard
(CDCl3: δ 7.26 and DMSO: δ 2.50). Signals are described as s, d,
t, dd, m, and b for singlet, doublet, triplet, double-doublet,
multiplet, and broad, respectively. Mass spectra (LC-MS) were
measured on a Ion Trap Esquire 3000+ instrument. Microanalyses
of target compounds were carried out with an Eurovector Euro 3000
model analyzer; the analytical results are within (0.4% of
theoretical values. Chemical names follow IUPAC nomenclature.
Starting materials were purchased from Aldrich, Acros, Lancaster,
Fluka, or Enamine (for compound 3) and were used without
purification. Column chromatography was performed using silica
gel (40-63 µm), and the reaction progress was determined by thin

1046 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 4
Pochetti et al.
layer chromatography analyses on Merck silica gel plastic plates
60F254. The following compounds were prepared according to
previously described procedures: 3-amino-4-(7-aminomethyl-3,4-
dihydro-1H-isoquinolin-2-yl)-cyclobut-3-ene-1,2-dione hydrochlo-
ride (6),50 and 6-aminomethyl-4H-benzo[1,4]oxazin-3-one hydro-
chloride (12).50
4,7,7-Trioxo-3,4,5,6,7,8-hexahydro-7,9-dithia-1,3-diaza-fluorene-
2-carboxylic Acid Ethyl Ester (4). To a solution of ethyl 2-amino-
4,7-dihydro-5H-thieno[2,3-c]thiopyran-3-carboxylate 6,6-dioxide
(500 mg, 1.81 mmol) in hydrochloric acid (4 M in dioxane; 10
mL) was added ethyl cyanoformate (270 µL, 2.72 mmol). The
mixture was stirred at 50 °C for 4 h and concentrated under reduced
pressure. The residue was dissolved in ethyl acetate, washed with
water and then brine, and dried over MgSO4. After evaporation
the crude product was obtained as orange solid (524 mg, 88%). 1H
NMR (DMSO-d6): δ 8.06 (d, 1H, NH), 4.71 (s, 2H, CH2SO2), 4.36
(q, 2H, CH2Me), 3.50 (s, 4H, CH2CH2SO2), 1.34 (t, 3H, CH3).
C12H12N2O5S2; MW: 328; MS 329 [M + H]+.
5-Ethoxycarbonylmethyl-4-oxo-3,4,5,6,7,8-hexahydro-
benzo[4,5]thieno[2,3-d]pyrimidine-2-carboxylic Acid Ethyl Ester
(10). To a solution of compound 9 (515 mg, 1.65 mmol) in
hydrochloric acid (4 M in dioxane; 10 mL) was added ethyl
cyanoformate (250 µL, 2.48 mmol). The mixture was stirred at 50
°C for 2 h and concentrated under reduced pressure. The residue
was dissolved in ethyl acetate, washed with water and then brine,
and dried over MgSO4. After evaporation the crude product was
purified by flash chromatography using dichloromethane/methanol
98/2 as eluent to afford the desired product as yellow solid (165
1
mg, 27%). H NMR (CDCl3): δ 9.87 (s, 1H, NH), 4.54 (q, 2H,
CH2Me), 4.16 (q, 2H, CH2Me), 3.86 (bd, 1H, CH), 2.98 (dd, 1H,
CHHCO), 2.86-2.77 (m, 2H, CH2), 2.47 (dd, 1H, CHHCO),
1.97-1.87 (m, 4H, 2 CH2), 1.47 (t, 3H, CH3), 1.27 (t, 3H, CH3).
C17H20N2O5S; MW: 364; MS 365 [M + H]+.
5-Ethoxycarbonylmethyl-4-oxo-3,4,5,6,7,8-hexahydro-
benzo[4,5]thieno[2,3-d]pyrimidine-2-carboxylic Acid (11). To a
solution of ester 10 (160 mg, 0.44 mmol) in THF (10 mL) was
added 1 M LiOH (0.88 mL, 0.88 mmol). The mixture was stirred
at room temperature for 30 min and acidified by addition of 1N
HCl. The precipitate was filtered and dried under high vacuum to
afford the product as a white solid (130 mg, 88%). 1H NMR
(DMSO-d6): δ 4.08 (q, 2H, CH2Me), 3.65 (bd, 1H, CH), 2.85 (dd,
2H, CH2CO), 2.48-2.35 (m, 2H, CH2), 1.85-1.75 (m, 4H, 2 CH2),
1.18 (t, 3H, CH3). C15H16N2O5S; MW: 336; MS 337 [M + H]+.
{4-oxo-2[(3-oxo-3,4-Dihydro-2H-benzo[1,4]oxazin-6-ylmethyl)-
carbamoyl]-3,4,5,6,7,8-hexahydrobenzo[4,5]thieno[2,3-d]pyrimidin-
5-yl}-acetic Acid Ethyl Ester (13). To a solution of acid 11 (120
mg, 0.36 mmol), EDCI (137 mg, 0.71 mmol), and HOAt (48 mg,
0.36 mmol) in DMF (3 mL) were added NMM (300 µL) and
6-aminomethyl-4H-benzo[1,4]oxazin-3-one hydrochloride 12 (92
mg, 0.43 mmol). The mixture was stirred overnight at room
temperature and then concentrated. The remaining residue was
suspended in 10% aqueous citric acid, and the residue was filtered
and dried to afford the product (175 mg, 99%) as an off-white solid.
1H NMR (DMSO-d6): δ 12.34 (bs, 1H, NH), 10.68 (s, 1H, NH),
9.63 (t, 1H, NH), 6.89 (m, 3H, aromatic), 4.52 (s, 2H, CH2O), 4.32
(d, 2H, CH2N), 4.08 (q, 2H, CH2Me), 3.65 (bd, 1H, CH), 2.89-2.73
(m, 3H, CH2 + CHH), 2.46-2.35 (m, 1H, CHH), 1.88-1.68 (m,
4H, 2 CH2), 1.18 (t, 3H, CH3). C24H24N4O6S; MW: 496; MS 497
[M + H]+.
{4-oxo-2[(3-oxo-3,4-Dihydro-2H-benzo[1,4]oxazin-6-ylmethyl)-
carbamoyl]-3,4,5,6,7,8-hexahydrobenzo[4,5]thieno[2,3-d]pyrimidin-
5-yl}-acetic Acid (14). To a solution of ester 13 (200 mg, 0.40 mmol)
in THF (10 mL) was added 1 M LiOH (1.2 mL, 1.2 mmol). The
mixture was stirred at room temperature for 4 h and acidified by
addition of 1N HCl. The precipitate was filtered and dried under
high vacuum to afford the product as a white solid (150 mg, 80%).
1H NMR (DMSO-d6): δ 12.35 (s, 1H, NH), 10.68 (s, 1H, NH),
9.63 (t, 1H, NH), 6.89 (m, 3H, aromatic), 4.52 (s, 2H, CH2CO),
4.33 (d, 2H, CH2N), 3.62 (bd, 1H, CH), 2.84-2.74 (m, 3H, CH2
+ CHH), 2.32 (dd, 1H, CHH), 1.87-1.75 (m, 4H, 2 CH2).
C22H20N4O6S; MW: 468; MS 469 [M + H]+.
4,7,7-Trioxo-3,4,5,6,7,8-hexahydro-7,9-dithia-1,3-diaza-fluorene-
2-carboxylic Acid (5). To a solution of ester 4 (300 mg, 0.91 mmol)
in THF (10 mL) was added 1 M LiOH (2.75 mL, 2.75 mmol). The
mixture was stirred at room temperature for 90 min and then the
solvent removed by evaporation at 30 °C. The mixture was acidified
by addition of 1N HCl and the precipitate filtered and dried under
high vacuum to afford the product as a pale-brown solid (85 mg,
1
31%). H NMR (DMSO-d6): δ 4.66 (s, 2H, CH2SO2), 3.42 (bs,
4H, CH2CH2SO2). C10H8N2O5S2; MW: 300; MS 301 [M + H]+.
4,7,7-Trioxo-3,4,5,6,7,8-hexahydro-7,9-dithia-1,3-diaza-fluorene-
2-carboxylic Acid [2-(2-Amino-3,4-dioxo-cyclobut-1-enyl)-1,2,3,4-
tetrahydro-isoquinolin-7-ylmethyl]-amide (1). To a solution of acid
5 (40 mg, 0.13 mmol), EDCI (51 mg, 0.26 mmol), and HOAt (18
mg, 0.13 mmol) in DMF (3 mL) were added NMM (100 µL) and
3-amino-4-(7-aminometyl-3,4-dihydro-1H-isoquinolin-2-yl)-cyclobut-
3-ene-1,2-dione hydrochloride 6 (47 mg, 0.16 mmol). The mixture
was stirred overnight at room temperature and then concentrated.
The remaining residue was suspended in 10% aqueous citric acid
and the residue was filtered and dried. After recrystallization in
MeOH the desired product was obtained (32 mg, 44%) as an off-
1
white solid. H NMR (DMSO-d6): δ 12.47 (bs, 1H, NH), 9.68 (t,
1H, NH), 7.80 (s, 2H, NH2), 7.17-7.10 (m, 3H, aromatic), 4.87
(bs, 2H, CH2SO2), 4.68 (s, 2H, CH2N), 4.39 (d, 2H, CH2NH), 3.93
(bs, 2H, CH2CH2SO2), 3.50 (s, 4H, NCH2CH2 + CH2CH2SO2),
2.93-2.84 (m, 2H, NCH2CH2). Anal. (C24H21N5O6S2) C, H, N, S.
MW: 539; MS 540 [M + H]+.
Cyano-(2-ethoxycarbonylmethyl-cyclohexylidene)-acetic Acid
Ethyl Ester (8). A solution of (2-oxo-cyclohexyl)-acetic acid ethyl
ester (1.00 g, 5.43 mmol), ethyl cyanoacetate (0.80 g, 7.05 mmol),
acetic acid (100 µL), and ammonium acetate (40 mg) in toluene
was refluxed overnight with a Dean-Stark apparatus. After the
solution was cooled to room temperature, the solvent was removed
and the residue purified by flash chromatography on silica gel using
cyclohexane/ethyl acetate 8/2 as eluent. The crude product (1.5 g)
was obtained as a yellow oil and used without further purification
in the next step. 1H NMR (CDCl3): δ 4.29-4.24 (m, 2H, CH2Me),
4.17-4.08 (m, 2H, CH2Me), 2.88-1.61 (m, 11H, CH, 5 CH2),
1.37-1.22 (m, 6H, 2 CH3). C15H21NO4; MW: 279; MS 280 [M +
H]+.
5-Carbamoylmethyl-4-oxo-3,4,5,6,7,8-hexahydro-benzo[4,5]thi-
eno[2,3-d]pyrimidine-2-carboxylic Acid (3-oxo-3,4-dihydro-2H-
benzo[1,4]oxazin-6-yl-methyl)amide (2). To a solution of acid 14
(20 mg, 43 µmol), EDCI (16 mg, 71 µmol) and HOAt (6 mg, 43
µmol) in DMF (3 mL) were added NMM (100 µL) and ammonium
chloride (3 mg, 51 µmol). The mixture was stirred overnight at
room temperature and then concentrated. The remaining residue
was suspended in 10% aqueous citric acid and the residue was
filtered and dried to afford the title compound (13.7 mg, 69%) as
2-Amino-4-ethoxycarbonylmethyl-4,5,6,7-tetrahydro-benz[b]-
thiophene-3-carboxylic Acid Ethyl Ester (9). A solution of com-
pound 8 (980 mg, 3.5 mmol) and sulfur (125 mg, 3.86 mmol) in
MeOH was heated at 50 °C and then diethylamine (180 µL, 1.75
mmol) was added slowly. After 4 h at 50 °C, the solvent was
removed by evaporation. The residue was purified by flash
chromatography using cyclohexane/ethyl acetate 8/2 as eluent to
give the thiophene derivative (517 mg) as yellow oil. The presence
of starting material was still detectable but the product was used
without further purification. 1H NMR (CDCl3): δ 6.08 (s, 2H, NH2),
4.27 (q, 2H, CH2Me), 4.14 (q, 2H, CH2Me), 3.62 (bd, 1H, CH),
2.90-1.62 (m, 8H, 4 CH2), 1.41-1.22 (m, 6H, 2 CH3).
C15H21NO4S; MW: 311; MS 312 [M + H]+.
1
an off-white solid. H NMR (DMSO-d6): δ 12.36 (bs, 1H, NH),
10.68 (s, 1H, NH), 9.63 (t, 1H, NH), 7.14 (s, 1H; NHH), 6.89 (m,
3H, aromatic), 6.81 (s, 1H, NHH), 4.52 (s, 2H, CH2O), 4.33 (d,
2H, CH2NH), 3.60 (bd, 1H, CH), 2.89-2.61 (m, 4H, CH2CON +
CH2), 2.18-1.68 (m, 4H, 2 CH2). Anal. (C22H21N5O5S) C, H, N,
S. MW: 467; MS 468 [M + H]+.
Enzyme Assay for Determination of the Inhibition Constants
IC50. MMP-8 activity and the selectivity assays were performed
using the catalytic domains of recombinant human MMPs (MMP-

MMP-8 Non-Zinc-Chelating SelectiVe Inhibitors
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 4 1047
Table 3. Statistics of Crystallographic Data and Refinement
MMP-8:(S)-2 (without AH)
0.8726
MMP-8:1 (with AH)
0.8726
MMP-8:1 (with AH)
0.8726
wavelength (Å)
temperature (K)
100
100
100
space group
no. of mol in the AU
cell axes (Å)
beta angle (deg)
resolution range (Å)
Rmerge (%)
multiplicity
I/σ (I)
completeness (%)
Rfactor (%)
P21 monoclinic
2
42.34; 69.38; 52.71
92.35
P21 monoclinic
1
32.29; 68.95; 32.75
105.41
P21212 orthorhombic
2
68.30; 69.21; 81.23
25.00-2.00 (2.12-2.00)a
12.4 (29.7)a
3.1 (3.1)a
3.6 (2.2)a
99.6 (100)a
22.0
25.00-1.60 (1.69-1.60)a
10.3 (30.9)a
4.1 (3.9)a
5.7 (2.4)a
99.2 (96.1)a
21.0
25.00-2.10 (2.20-2.10)a
11.9 (30.9)a
4.1 (4.0)a
4.0 (2.0)a
96.9 (80.6)a
23.8
Rfree (%)
28.0
24.2
27.0
no. of reflections (work)
no. of reflections (test)
rmsd bond lengths/bond angles
Ramachandran plot favored/
allowed/generously allowed (%)
protein nonhydrogen atoms
solvent molecules
nonhydrogen inhibitors atoms
ions
19025
1400
0.006/1.293
86.6/12.3/1.1
16280
1780
0.005/1.439
84.8/13.0/1.4
20614
1519
0.006/1.270
89.5/9.8/0.7
2568
361
66
1284
216
37
4
8.8
21.9
13.7
4.2
2568
426
74
8
8
protein average B factors (Å2)
water ave. B factors (Å2)
inhibitor ave. B factors (Å2)
ions ave. B factors (Å2)
17.0
35.8
16.2
10.8
18.8
37.6
24.8
13.0
a The values in parentheses refer to the outer shell.
1, -2, -3, -7, -9, -12: Biomol, Hamburg, Germany; MMP-8:
Calbiochem, Schwalbach, Germany; MMP-13 and MMP-14: In-
vitek, Berlin, Germany) and the appropriate fluorogenic peptide
substrate. MMP-13 activity was tested using the specific MMP-13
substrate MCA-Pro-Cha-Gly-Nva-His-Ala-Dpa-NH2 (Calbiochem,
Schwalbach, Germany). MMP-3 was tested using the NFF-3
substrate MCA-Arg-Pro-Lys-Pro-Val-Glu-Nva-Trp-Arg-Lys(DNP)-
NH2 (Calbiochem, Schwalbach, Germany) and all remaining MMPs
were tested using OmniMMP substrate MCA-Pro-Leu-Gly-Leu-
Dpa-Ala-Arg-NH2 ·AcOH (Biomol, Hamburg, Germany).51
The MMP-3 assay was performed in 50 mM MES buffer, pH
6.0, 10 mM CaCl2 and 0.05% Brij-35. All other MMP assays were
performed in 50 mM Tris-HCl buffer, pH 7.5, 150 mM NaCl, 5
mM CaCl2, and 0.05% Brij-35. Concentrations of enzymes and
substrates were optimized for each assay and varied between 1 and
10 nM enzyme and 4-10 µM substrate. The enzyme activity was
measured after 10 min preincubation of the enzyme with varying
concentrations of the inhibitor. Each assay was run at least in
duplicate and the IC50 values were calculated using Life Science
Workbench (LSW) Data Analysis Plugin for Microsoft Excel. The
data were fitted into the formula y ) Vmax/(1 + ([I]/IC50).
Expression and Purification of the Protein. The truncated form
M80-G242 of the catalytic domain of MMP-8 has been supplied
by Prof. H. Tschesche, University of Bielefeld (Germany). The
protocol of expression and purification has been previously de-
scribed.52
Protein Crystallization. Acetohydroxamate (AH) was added to
45 mL of protein solution (MMP-8 2.4 µM, in 5 mM Tris HCl,
pH 7.0, 100 mM NaCl and 5 mM CaCl2) to a final concentration
of 50 mM. The weak inhibitor AH was added to the protein
solution to aid protein stability by preventing autoproteolysis
during concentration. However, 15 mL of protein solution have
been used for crystallization trials without AH in order to check
whether the binding mode of the inhibitors could be perturbed
by the presence of AH. The MMP-8 protein with AH was
concentrated with Amicon-Ultra 15, dialyzing with the crystal-
lization solution (5 mM CaCl2, 100 mM NaCl, 0.5 mM ZnCl2,
3 mM MES-NaOH, 0.02% NaN3, pH 6.0) to a final concentration
of 5.8 mg/mL. The protein solution was equilibrated with the
inhibitor stock solutions (100 mM in DMSO) at 4 °C overnight,
so that the inhibitor were 5-fold in excess and the final
concentration of DMSO 1.5%. Crystallizations were performed
by hanging-drop vapor diffusion at 18 °C. Hanging droplets were
made by mixing 1.5-3.0 µL of protein/inhibitor solution with
5 µL of PEG solution (10% (m/v) PEG6000, 0.2 M Mes-NaOH,
0.02% NaN3, pH 6.0). Droplets were concentrated against a
reservoir buffer containing 1.0-2.0 M sodium phosphate, 0.02%
NaN3, pH 6.0. Crystals appeared in 3-4 days.
The inhibitor from the stock solution was added to the protein
solution without AH in large excess and complexation was allowed
to continue overnight at 4 °C. The complex was then concentrated,
dialysing with the crystallization solution, up to 2-3 mg/mL. More
than half of the protein has been lost during the concentration,
cleaved, and absorbed to the filter. Only 10 drops of crystallization
could be made, by mixing 2.5-3 µL of protein solution with 5 µL
of PEG solution. The reservoir was made of 1.7-2.0 M sodium
phosphate, pH 6.0. Crystals appeared in 3-7 days.
Data Collection and Processing. X-ray data were collected
under cryogenic conditions (100 K) at the ID23-2 microfocus
beamline of ESRF, Grenoble, using a wavelength of 0.873 Å
and a MAR-Mosaic225 CCD detector. The crystals were flash-
frozen in the nitrogen stream after transferring them for few
seconds into the mother solution containing 35% PEG400. Data
were integrated and scaled using the programs MOSFLM and
SCALA.53 The statistics of collection and refinement is given
in Table 3.
Structure Solution and Refinement. The orientation and trans-
lation of the protein molecules within the crystallographic unit
cell was determined using the program AmoRe54 with the
coordinates of MMP-8 (PDB accession code 1I7655) as a search
model. Using data in the 12-3 Å resolution range an unambiguous
solution was obtained for each complex. All the structures were
refined with CNS.56 The Rfree validation was based on a subset of
7% (10% for the monoclinic form of MMP-8:1) of the reflections
omitted during the refinement. The inhibitors were fitted manually
into the Fo - Fc electron density and unambiguously modeled. After
several rounds of refinement water molecules were added.
Acknowledgment. We acknowledge G. Becher, M. Burger,
H. Kheil, A. Timmermann, and J. Weik from Alantos Pharma-
ceuticals AG for the determination of the IC50 values and Fulvio
Loiodice from Dipartimento Farmaco-Chimico, Universita` degli
Studi di Bari, for the elemental analysis.

1048 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 4
Pochetti et al.
Supporting Information Available: CR distances at the S1′ loop
region between the uninhibited MMP-8 and the inhibited complexes
MMP-8:1 and MMP-8:(S)-2, elemental analysis data, and NMR
spectra of compounds 1 and 2. This material is available free of
charge via the Internet at http://pubs.acs.org.
(25) Freskos, J. N.; McDonald, J. J.; Mischke, B. V.; Mullins, P. B.; Shieh,
H. S.; Stegeman, R. A.; Stevens, A. M. Synthesis and identification
of conformationally constrained selective MMP inhibitors. Bioorg.
Med. Chem. Lett. 1999, 9, 1757–1760.
(26) Tamura, Y.; Watanabe, F.; Nakatani, T.; Yasui, K.; Fuji, M.;
Komurasaki, T.; Tsuzuki, H.; Maekawa, R.; Yoshioka, T.; Kawada,
K.; Sugita, K.; Ohtani, M. Highly selective and orally active inhibitors
of type IV collagenase (MMP-9 and MMP-2): N-sulfonylamino acid
derivatives. J. Med. Chem. 1998, 41, 640–649.
(27) Hodgson, J. Remodeling MMPIs. Biotechnology (N.Y.) 1995, 13, 554–
557.
(28) Puerta, D. T.; Lewis, J. A.; Cohen, S. M. New beginnings for matrix
metalloproteinase inhibitors: identification of high-affinity zinc-binding
groups. J. Am. Chem. Soc. 2004, 126, 8388–8389.
(29) Puerta, D. T.; Mongan, J.; Tran, B. L.; McCammon, J. A.; Cohen,
S. M. Potent, selective pyrone-based inhibitors of stromelysin-1. J. Am.
Chem. Soc. 2005, 127, 14148–14149.
(30) Jacobsen, F. E.; Lewis, J. A.; Cohen, S. M. A new role for old ligands:
discerning chelators for zinc metalloproteinases. J. Am. Chem. Soc.
2006, 128, 3156–3157.
(31) Li, X.; Li, Y.; Xu, W. Design, synthesis, and evaluation of novel galloyl
pyrrolidine derivatives as potential antitumor agents. Bioorg. Med.
Chem. 2006, 14, 1287–1293.
(32) Zhang, L.; Zhang, J.; Fang, H.; Wang, Q.; Xu, W. Design, synthesis
and preliminary evaluation of new cinnamoyl pyrrolidine derivatives
as potent gelatinase inhibitors. Bioorg. Med. Chem. 2006, 14, 8286–
8294.
References
(1) Sternlicht, M. D.; Werb, Z. How matrix metalloproteinases regulate
cell behavior. Annu. ReV. Cell DeV. Biol. 2001, 17, 463–516.
(2) Shapiro, S. D. Matrix metalloproteinase degradation of extracellular
matrix: biological consequences. Curr. Opin. Cell Biol. 1998, 10, 602–
608.
(3) Whittaker, M.; Floyd, C. D.; Brown, P.; Gearing, A. J. Design and
therapeutic application of matrix metalloproteinase inhibitors. Chem.
ReV. 1999, 99, 2735–2776.
(4) Overall, C. M.; Lo´pez-Otin, C. Strategies for MMP inhibition in cancer:
innovations for the post-trial era. Nat. ReV. Cancer 2002, 2, 657–672.
(5) Holleran, W. M.; Galardy, R. E.; Gao, W. N.; Levy, D.; Tang, P. C.;
Elias, P. M. Matrix metalloproteinase inhibitors reduce phorbol ester-
induced cutaneous inflammation and hyperplasia. Arch. Dermatol. Res.
1997, 289, 138–144.
(6) Hewson, A. K.; Smith, T.; Leonard, J. P.; Cuzner, M. L. Suppression
of experimental allergic encephalomyelitis in the Lewis rat by the
matrix metalloproteinase inhibitor Ro31-9790. Inflamm. Res. 1995,
44, 345–349.
(7) Chandler, S.; Coates, R.; Gearing, A.; Lury, J.; Wells, G.; Bone, E.
Matrix metalloproteinases degrade myelin basic protein. Neurosci. Lett.
1995, 201, 223–226.
(33) Campestre, C.; Agamennone, M.; Tortorella, P.; Preziuso, S.; Biasone,
A.; Gavuzzo, E.; Pochetti, G.; Mazza, F.; Hiller, O.; Tschesche, H.;
Consalvi, V.; Gallina, C. N-Hydroxyurea as zinc binding group in
matrix metalloproteinase inhibition: mode of binding in a complex
with MMP-8. Bioorg. Med. Chem. Lett. 2006, 16, 20–24.
(34) Mannino, C.; Nievo, M.; Machetti, F.; Papakyriakou, A.; Calderone,
V.; Fragai, M.; Guarna, A. Synthesis of bicyclic molecular scaffolds
(BTAa): an investigation towards new selective MMP-12 inhibitors.
Bioorg. Med. Chem. 2006, 14, 7392–7403.
(8) Lohmander, L. S.; Hoerrner, L. A.; Lark, M. W. Metalloproteinases,
tissue inhibitor, and proteoglycan fragments in knee synovial fluid in
human osteoarthritis. Arthritis Rheum. 1993, 36, 181–189.
(9) Ahrens, D.; Koch, A. E.; Pope, R. M.; Stein-Picarella, M.; Niedbala,
M. J. Expression of matrix metalloproteinase 9 (96-kd gelatinase B)
in human rheumatoid arthritis. Arthritis Rheum. 1996, 39, 1576–1587.
(10) Blaser, J.; Triebel, S.; Maasjosthusmann, U.; Romisch, J.; Krahl-
Mateblowski, U.; Freudenberg, W.; Fricke, R.; Tschesche, H. Deter-
mination of metalloproteinases, plasminogen-activators and their
inhibitors in the synovial fluids of patients with rheumatoid arthritis
during chemical synoviorthesis. Clin. Chim. Acta 1996, 244, 17–33.
(11) Ohishi, K.; Fujita, N.; Morinaga, Y.; Tsuruo, T. H-31 human breast
cancer cells stimulate type I collagenase production in osteoblast-like
cells and induce bone resorption. Clin. Exp. Metastasis 1995, 13, 287–
295.
(12) Witty, J. P.; Foster, S. A.; Stricklin, G. P.; Matrisian, L. M.; Stern,
P. H. Parathyroid hormone-induced resorption in fetal rat limb bones
is associated with production of the metalloproteinases collagenase
and gelatinase B. J. Bone Miner. Res. 1996, 11, 72–78.
(13) Peress, N.; Perillo, E.; Zucker, S. Localization of tissue inhibitor of
matrix metalloproteinases in Alzheimer’s disease and normal brain.
J. Neuropathol. Exp. Neurol. 1995, 54, 16–22.
(14) Chang, C.; Werb, Z. The many faces of metalloproteases: cell growth,
invasion, angiogenesis and metastasis. Trends Cell. Biol. 2001, 11,
S37-S43.
(15) Coussens, L. M.; Fingleton, B.; Matrisian, L. M. Matrix metallopro-
teinase inhibitors and cancer: trials and tribulations. Science 2002, 295,
2387–2392.
(16) Puerta, D. T.; Cohen, S. M. A bioinorganic perspective on matrix
metalloproteinase inhibition. Curr. Top. Med. Chem. 2004, 4, 1551–
1573.
(17) Matter, H.; Schudok, M. Recent advances in the design of matrix
metalloprotease inhibitors. Curr. Opin. Drug DiscoVery DeV. 2004,
7, 513–535.
(18) Fingleton, B. Matrix metalloproteinases as valid clinical targets. Curr.
Pharm. Des. 2007, 13, 333–346.
(19) Skiles, J. W.; Gonnella, N. C.; Jeng, A. Y. The design, structure, and
clinical update of small molecular weight matrix metalloproteinase
inhibitors. Curr. Med. Chem. 2004, 11, 2911–2977.
(20) Rao, B. G. Recent developments in the design of specific matrix
metalloproteinase inhibitors aided by structural and computational
studies. Curr. Pharm. Des. 2005, 11, 295–322.
(21) Babine, R. E.; Bender, S. L. Molecular recognition of protein-ligand
complexes: applications to drug design. Chem. ReV. 1997, 97, 1359–
1472.
(22) 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 hydroxamate inhibitor.
Biochemistry 1995, 34, 14012–14020.
(23) Muri, E. M.; Nieto, M. J.; Sindelar, R. D.; Williamson, J. S.
Hydroxamic acids as pharmacological agents. Curr. Med. Chem. 2002,
9, 1631–1653.
(35) Nuti, E.; Tuccinardi, T.; Rossello, A. Matrix metalloproteinase
inhibitors: new challenges in the era of post broad-spectrum inhibitors.
Curr. Pharm. Des. 2007, 13, 2087–2100.
(36) Pirard, P. Insight into the structural determinants for selective inhibition
of metalloproteinases. Drug DiscoVery Today 2007, 12, 640–646.
(37) Georgiadis, D.; Yiotakis, A. Specific targeting of metzincin family
members with small-molecule inhibitors: progress toward a multifari-
ous challenge. Bioorg. Med. Chem. 2008, 16, 8781–8794.
(38) Chen, J. M.; Nelson, F. C.; Levin, J. I.; Mobilio, D.; Moy, F. J.;
Nilakantan, R.; Zask, A.; Powers, R. Structure-based design of a novel,
potent, and selective inhibitor for MMP-13 utilizing NMR spectroscopy
and computer-aided molecular design. J. Am. Chem. Soc. 2000, 122,
9648–9654.
(39) Engel, C. K.; Pirard, B.; Schimanski, S.; Kirsch, R.; Habermann, J.;
Klingler, O.; Schlotte, V.; Weithmann, K. U.; Wendt, K. U. Structural
basis for the highly selective inhibition of MMP-13. Chem. Biol. 2005,
12, 181–189.
(40) Morales, R.; Perrier, S.; Florent, J. M.; Beltra, J.; Dufour, S.; De
Mendez, I.; Manceau, P.; Tertre, A.; Moreau, F.; Compere, D.;
Dublanchet, A. C.; O’Gara, M. Crystal structures of novel non-peptidic,
non-zinc chelating inhibitors bound to MMP-12. J. Mol. Biol. 2004,
341, 1063–1076.
(41) Johnson, A. R.; Pavlovsky, A. G.; Ortwine, D. F.; Prior, F.; Man,
C. F.; Bornemeier, D. A.; Banotai, C. A.; Mueller, W. T.; McConnell,
P.; Yan, C.; Baragi, V.; Lesch, C.; Roark, W. H.; Wilson, M.; Datta,
K.; Guzman, R.; Han, H. K.; Dyer, R. D. Discovery and characteriza-
tion of a novel inhibitor of matrix metalloprotease-13 that reduces
cartilage damage in vivo without joint fibroplasia side effects. J. Biol.
Chem. 2007, 282, 27781–27791.
(42) Li, J. J.; Nahra, J.; Johnson, A. R.; Bunker, A.; O‘Brien, P.; Yue,
W. S.; Ortwine, D. F.; Man, C. F.; Baragi, V.; Kilgore, K.; Dyer,
R. D.; Han, H. K. Quinazolinones and pyrido[3,4-d]pyrimidin-4-
ones as orally active and specific matrix metalloproteinase-13
inhibitors for the treatment of osteoarthritis. J. Med. Chem. 2008,
51, 835–841.
(43) Overall, C. M.; Kleifeld, O. Tumour microenvironmentsopinion:
validating matrix metalloproteinases as drug targets and antitargets
for cancer therapy. Nat. ReV. Cancer 2006, 6, 227–239.
(44) Van Lint, P.; Wielockx, B.; Puime`ge, L.; Noe¨l, A.; Lo´pez-Otin, C.;
Libert, C. Resistance of collagenase-2 (matrix metalloproteinase-8)-
deficient mice to TNF-induced lethal hepatitis. J. Immunol. 2005, 175,
7642–7649.
(45) Folgueras, A. R.; Fueyo, A.; Garcia-Sua´rez, O.; Cox, J.; Astudillo,
A.; Tortorella, P.; Campestre, C.; Gutie´rrez-Ferna´ndez, A.; Fanjul-
Ferna´ndez, M.; Pennington, C. J.; Edwards, D. R.; Overall, C. M.;
Lo´pez-Otin, C. Collagenase-2 deficiency or inhibition impair experi-
(24) Levin, J. I. The design and synthesis of aryl hydroxamic acid inhibitors
of MMPs and TACE. Curr. Top. Med. Chem. 2004, 4, 1289–1310.

MMP-8 Non-Zinc-Chelating SelectiVe Inhibitors
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 4 1049
mental autoimmune encephalomyelitis in mice. J. Biol. Chem. 2008,
283, 9465–9474.
(52) Schnierer, S.; Kleine, T.; Gote, T.; Hillemann, A.; Knauper, V.;
Tschesche, H. The recombinant catalytic domain of human neutrophil
collagenase lacks type I collagen substrate specificity. Biochem.
Biophys. Res. Commun. 1993, 191, 319–326.
(53) Leslie, A. G. W. Joint CCP4 and ESF-EACMB Newsletters on Protein
Crystallography; SERC Daresbury Laboratory: Warrington, UK, 1992;
Vol. 26.
(54) Navaza, J. AMoRe: an automated package for molecular replacement.
Acta Crystallogr., Sect. A: Found. Crystallogr. 1994, 50, 157–163.
(55) Gavuzzo, E.; Pochetti, G.; Mazza, F.; Gallina, C.; Gorini, B.;
D’Alessio, S.; Pieper, M.; Tschesche, H.; Tucker, P. A. Two crystal
structures of human neutrophil collagenase, one complexed with a
primed- and the other with an unprimed-side inhibitor: implications
for drug design. J. Med. Chem. 2000, 43, 3377–3385.
(56) Brunger, A. T.; Adams, P. D.; Clore, G. M.; DeLano, W. L.; Gros,
P.; Grosse-Kunstleve, R. W.; Jiang, J. S.; Kuszewski, J.; Nilges, M.;
Pannu, N. S.; Read, R. J.; Rice, L. M.; Simonson, T.; Warren, G. L.
Crystallography & NMR system: a new software suite for macromo-
lecular structure determination. Acta Crystallogr., Sect. D: Biol.
Crystallogr. 1998, 54 (Pt 5), 905–921.
(46) Van Lint, P.; Libert, C. Matrix metalloproteinase-8: cleavage can be
decisive. Cytokine Growth Factor ReV. 2006, 17, 217–223.
(47) Gege, C.; Cevrier, C.; Schneider, M.; Bluhm, H.; Hochgu¨rtel, M.;
Deng, H.; Gallagher, B.; Sucholeiki, I.; Taveras, A. G. Heterotri-
cyclic metalloprotease inhibitors. PCT Int. Appl. WO2008063667,
2008.
(48) Bertini, I.; Calderone, V.; Fragai, M.; Luchinat, C.; Maletta, M.; Yeo,
K. J. Snapshots of the reaction mechanism of matrix metalloprotein-
ases. Angew. Chem., Int. Ed. 2006, 45, 7952–7955.
(49) Teague, S. J. Implications of protein flexibility for drug discovery.
Nat. ReV. Drug DiscoVery 2003, 2, 527–541.
(50) Steeneck, C.; Gege, C.; Richter, F.; Hochgu¨rtel, M.; Feuerstein, T.;
Bluhm, H.; Sucholeiki, I.; Boer, J.; Wu, X.; Schneider, M.; Nolte, B.;
Gallagher, B.; van Veldhuizen, J.; Deng, H.; Essers, M.; Kroth, H.;
Kiely, A.; Powers, T.; Taveras, A. G. Pyrimidine or triazine fused
bicyclic metalloprotease inhibitors. PCT Int. Appl. WO2006128184,
2006.
(51) Knight, C. G.; Willenbrock, F.; Murphy, G. A novel coumarin-labelled
peptide for sensitive continuous assays of the matrix metalloprotein-
ases. FEBS Lett. 1992, 296, 263–266.
JM801166J