Novel 1-hydroxypiperazine-2,6-diones as new leads in the inhibition of metalloproteinases

Article abstract of DOI:10.1021/jm200593b

New compounds containing a novel zinc-binding group (1-hydroxypiperazine-2, 6-dione, HPD) have been identified as effective inhibitors of matrix metalloproteinases (MMPs), with activities in the nanomolar concentration range. That moiety seemed to bind th

Full text of DOI:10.1021/jm200593b

Article
pubs.acs.org/jmc
Novel 1-Hydroxypiperazine-2,6-diones as New Leads in the Inhibition
of Metalloproteinases
Sergio M. Marques,† Tiziano Tuccinardi,*,‡ Elisa Nuti,‡ Salvatore Santamaria,‡,§ Vania Andre,†
́
̂
́
Armando Rossello,‡ Adriano Martinelli,‡ and M. Amelia Santos*,†
́
†
́
Centro de Quımica Estrutural, Instituto Superior Tec
‡Dipartimento di Scienze Farmaceutiche, Universita
§Kennedy Institute of Rheumatology Division, Faculty of Medicine, Imperial College London, 65 Aspenlea Road, London W6 8LH,
United Kingdom
́
nico, Av. Rovisco Pais 1, 1049-001 Lisboa, Portugal
̀
degli Studi di Pisa, Via Bonanno 6, 56126 Pisa, Italy
S
* Supporting Information
ABSTRACT: New compounds containing a novel zinc-binding
group (1-hydroxypiperazine-2,6-dione, HPD) have been iden-
tified as effective inhibitors of matrix metalloproteinases (MMPs),
with activities in the nanomolar concentration range. That
moiety seemed to bind the catalytic zinc ion of MMPs, revealing
itself as a new potential substitute for the hydroxamate group in
the next generation of metalloproteinase inhibitors. The X-ray
crystal structure of 1b elucidated its 3D conformation and
supramolecular packing in solid state. Theoretical procedures
were used to investigate the binding mode of this class of
compounds, within the active site of MMP13. A computational
method involving docking and hybrid quantum mechanical
and molecular mechanical (QM/MM) dynamic simulations was developed and applied. This study suggested that the HPD
moiety binds bidentately to the catalytic zinc through its oxygen atoms. The final structure obtained will allow straightforward
drug design approaches in view of further optimization and development of new MMP inhibitors bearing the HPD moiety.
of hydroxamate-based inhibitors is due to its poor bioavailability
and the toxicity arising from its metabolic stability limitations (the
amide bond can be easily hydrolyzed with formation of hydro-
xylamine and the corresponding carboxylic acids).8,9 On the
other hand, most of the inhibitors tested were broad-spectrum
MMP inhibitors (e.g. CGS 27023A,10 Chart 1) that make no
distinction between these enzymes.
1. INTRODUCTION
Matrix metalloproteinases (MMPs) are a class of zinc-dependent
endopeptidases that degrade most components of the extracellular
matrix such as collagens and gelatins. These enzymes are normally
expressed only in physiological processes of tissue remodeling and
repairing, homeostatic regulation, and control of innate immunity.
However, in cases of overexpression, they are responsible for
several pathological situations involving degradation of the
connective tissue, such as rheumatoid arthritis, osteoarthritis,
neuroinflammatory diseases, aneurisms, angiogenesis, and tumor
invasion.1−3 Hence, there has been a demand for synthetic
inhibitors that are able to control their activity for the last three
decades. Most of the inhibitors reported bear a hydroxamic acid as
a zinc-binding group (ZBG) because it provides the strongest
inhibitory properties by coordinating the catalytic zinc ion of the
MMPs.1,4 Except for a few ligands, such as AZD1236, under
development by AstraZeneca, and CTS-1027, from Conatus
Pharmaceuticals, there are currently no synthetic or biologic MMP
inhibitors (MPIs) in clinical trials or in use. This is primarily due
to the failure of early studies, which were mainly focused on
compounds containing hydroxamate as the zinc-chelating group.5,6
A tetracycline derivative, doxycycline (Periostat; CollaGenex
Pharmaceuticals Inc., Newtown, PA, USA), is currently the only
MPI approved by the U.S. FDA and is used as an adjunct therapy
in adult periodontitis.7 One of the main reasons for the limitations
Chart 1
The implication of certain members of the MMP family in
cancer and tumor progression is widely accepted, but the difficulty
lies in finding the true targets because their roles in the “protease web”
Received: May 12, 2011
Published: October 21, 2011
© 2011 American Chemical Society
8289
dx.doi.org/10.1021/jm200593b|J. Med. Chem. 2011, 54, 8289−8298

Journal of Medicinal Chemistry
Article
a
Scheme 1
a
Reagents and conditions: (i) ArSO2Cl, TEA, 1:1 dioxane/H2O, rt; (ii) BnBr, TEA, CH3CN, reflux; (iii) BrCH2COOtBu, K2CO3, KI, DMF, rt;
(iv) H2, Pd/C, MeOH, rt; (v) ECF, NMM, THF, 0 °C; (vi) NH2OBn, MeOH, 0 °C; (vii) TFA, CH2Cl2, rt.
are complex. In fact, some MMP members may act as activators
of the other enzymes and play important roles in various cascade
reactions, which may turn out to be protective against the
disease.3,8,11,12 Furthermore, concerning the composition and 3D
structure of the active site, MMPs are very conservative, which
makes it more difficult to develop selective inhibitors. At this point,
the type of ZBG in use may be of great importance. Some authors
claim that functional groups that chelate the metal ion too strongly
may imbalance the importance of the whole of the molecule in
binding to the protein, making it difficult to find selectivity for a
specific metalloprotein. This may be the case in the hydroxamic
group, which forms strong complexes with zinc and other metal
ions3,4,9,13 and may not only bind indiscriminately to all members
of the MMP family but also to other metalloproteinases.14
The number of drawbacks related to the hydroxamate-based
drugs prompted us to investigate new types of ZBGs suitable
for use in MMP inhibitors in a tendency followed by others.15−17
Such efforts led us to the discovery of a novel ZBG, the
1-hydroxypiperazine-2,6-dione (HPD), which can be success-
fully included in new inhibitors of MMPs or other metal-
loproteinases (Chart 1, compounds 1a−c). This group is
related to the hydroxamic acids (i.e., compound 2,18 Chart 1),
however, due to its six-membered ring structure, it is expected
to be more stable and more resistant to hydrolysis, with
additional favorable pharmacokinetic properties. On the other
hand, it has previously shown lower binding strength with a
series of metal ions, including zinc(II), with respect to the
hydroxamic moiety.19 This discovery drove us to investigate the
binding mode of these compounds within MMPs in order to
develop new MPIs bearing the HPD moiety as ZBG.
This group is tractable for structure optimization at three
points (3- and 5-positions of the ring and the arylsulfonyl group
linked to the 4-position; see Chart 1). This strategy envisages
obtaining important enhancements in the activity and
selectivity profiles of new inhibitors and thus lead to new
effective MMP targeting drugs.
compound 2 (Chart 1).18 Starting from D-valine, 2a−c were
prepared after a series of coupling reactions that involved a
protection/deprotection strategy with orthogonal activating
groups (see Scheme 1, steps i−vii).18 The reactions of those
carboxylic derivatives (2a−c) with a carboxylic-activating agent,
through an intramolecular nucleophilic attack of the N-
hydroxylamide on the activated carbonyl group, resulted in
the 6-ring closure and the formation of compounds 3a−c. The
reaction always yielded the same products independently of
the activating agent (ECF or TBTU) and the temperature
(−40 or 0 °C). This fact clearly indicates that 3a−c are the
most thermodynamically stable products of this reaction. After
catalytic hydrogenolysis for benzyl deprotection, the target
compounds 1a−c were obtained.
To evaluate the aqueous stability of the new compounds, the
1H NMR spectra were recorded for 1a in D2O for several days.
At pD ca. 2, almost no decomposition was observed for the
first 24 h (<1%), only ca. 2% decomposition was observed after
48 h, ca. 5% after 5 days, and ca. 8% after 10 days. At pD ca. 7,
almost no decomposition was observed for the first 24 h
(<1%), ca. 3% after 3 days, ca. 6% after 5 days, and ca. 7%
after 20 days. These results suggest considerable stability of
the HPD moiety to hydrolysis both in acidic and neutral
conditions.
3. RESULTS AND DISCUSSION
3.1. X-ray Crystallographic Structure of Compound
1b. Single crystals of compound 1b were obtained from the
slow evaporation of a solution of pure compound in
dichloromethane. The crystalline displayed a monoclinic
symmetry, and the acentric P21 space group supported the
chirality of the molecule. All the distances and angles were
within the expected values for related compounds.
The molecules interacted with each other via the (N−)O−
H···O(−S) (0.82 Å; 1.90 Å; 2.697(3) Å; 163°) H-bond.
Although this was not necessarily the active species, it allowed
us to draw conclusions relatively to the correct 3D
conformation of the ligand in the molecular modeling
structures with MMP13 and to the selection of the proposed
binding hypothesis (vide infra).
2. CHEMISTRY
The compounds 1a−c appeared as cyclic analogues of
previously reported hydroxamic-based compounds such as
8290
dx.doi.org/10.1021/jm200593b|J. Med. Chem. 2011, 54, 8289−8298

Journal of Medicinal Chemistry
Article
The supramolecular arrangement of 1b is characterized by C(8)
chains along the a axis, which are formed by the above-mentioned
interaction (Figure 1B). Furthermore, C−H···π interactions between
phenoxyphenyl-sulfonyl group, displayed the highest activities
against all the tested enzymes, whereas in general, compound
1a, containing a p-methoxyphenyl-sulfonyl group, demonstra-
ted the lowest activities. Interestingly, this inhibitory profile
parallels previous reports for hydroxamate-based inhibitors
containing arylsulfonyl P1′ groups.18
An initial attempt to rationalize these results, in terms of
inhibitor−enzyme interactions, leads us to conclude that, most
probably, the arylsulfonyl moiety, present in these inhibitors
and in other known analogues, interacts with the cavity S1′ of
the enzymes in a similar manner. Compared to hydroxamic-
type related inhibitors, such as compound 2 and CGS 27023A
(Chart 1), the new compounds seem to be weaker MPIs. While
2 displayed subnanomolar inhibitory activities against several
enzymes (IC50 values between 0.33−0.77 nM over MMP2, 8, 9,
and 13), 1c exhibited IC50 values between 9.5−30 nM, which
shows a decrease in activity from 25- to 60-fold among the
tested MMPs. This fact indicates that the interactions
established between this ZBG and the enzymes are weaker
than those formed by the hydroxamic acid moiety.
On the other hand, although 1c is the best inhibitor of the
series, 1a and 1b showed more interesting selectivity profiles (see
Figure S2, Supporting Information). In fact, these two compounds
showed higher affinity for MMP12 over the other tested MMPs
and TACE. On the other hand, 1c showed selectivity for MMP13
over all the other MMPs, but in general with lower selectivity
values than those displayed by 1a and 1b for MMP12.
Overall, these results suggest that the HPD moiety is an
effective chelating group that may be included in the scaffold of
a new class of metalloproteins inhibitors. To selectively inhibit
very similar enzymes, it is necessary to take advantage of the
small differences between their protein parts. Therefore, a new
approach has been hypothesized based on reducing the
enormous contribution of certain binding groups, namely the
ZBG, to the ligand−enzyme binding strength, at the expenses
of increasing the binding interactions with other inhibitor
moieties.13 HPD derivatives are reported to chelate metal ions
with less strength than hydroxamates (with stability constants
of their zinc(II) complexes ca. 1 order of magnitude lower),19
which may reveal a positive property for that strategy of
increasing selectivity on metalloproteinase inhibition.
3.3. Molecular Modeling. A deep knowledge of the binding
mode of these compounds within MMPs is of recognized
importance for the rationalization of their activity, namely in
terms of their established interactions and improving the design of
new inhibitor drugs. To achieve that goal, the existence of an
inhibitor−enzyme complex model may greatly ease computer-
aided structure optimization by using a scaffold-constrain docking
procedure, as previously reported.18 Compounds 1a−c proved to
be effective MMP inhibitors, in which the HPD moiety was
expected to play the role of zinc binding group. Our research group
previously conducted a complexation study of a compound
Figure 1. (a) ORTEP diagram for compound 1b with ellipsoids set at
50%. (b) Supramolecular arrangements of compound 1b showing the
chain formed along a via (N−)O−H···O(−S) hydrogen bonds.
the CH2 of the HPD moiety and the π system terminal ring of the
biphenyl (3.686(4) Å) play an important role in the 3-dimensional
packing of this structure, as they connect two consecutive antiparallel
chains (see Figure S1 in Supporting Information).
3.2. Enzyme Inhibition. The inhibitory activities of the
compounds were evaluated against a wide panel of enzymes
comprising 10 MMPs (MMP1, 2, 3, 7, 8, 9, 12, 13, 14, 16) and
TACE (TNF-α converting enzyme). The results thus obtained
(see Table 1) show that the inhibitory activities against
these enzymes range from micromolar to low nanomolar
values. The highest inhibitory activity was found for 1c against
MMP13 (IC50 = 9.5 nM). Overall, this inhibitor, containing a
a
Table 1. Enzyme Inhibitory Activities (IC50 Values, nM) of the 1-Hydroxypiperazine-2,6-dione Derivatives 1a−c and
Reference Hydroxamate-Based Inhibitors
compd
MMP1
MMP2
MMP3
MMP7
MMP8
MMP9
MMP12
MMP13
MMP14
MMP16
TACE
1a
1b
1c
2
4900
1300
1600
59
1200
240
30
3400
3900
220
6.7
39000
nd
300
67
690
660
30
66
19
23
nd
nd
300
52
2200
3900
83
1100
2200
39
33000
16000
10000
2290
nd
21
9.5
0.33
5.7
0.50
25
nd
0.40
7.7
0.77
5.0
3.9
1.6
CGS 27023A
56
16
380
23
6.8
160
a
Enzymatic data are the mean values for three independent experiments. SD are within ±10%; nd: not determined.
8291
dx.doi.org/10.1021/jm200593b|J. Med. Chem. 2011, 54, 8289−8298

Journal of Medicinal Chemistry
Article
containing this moiety with Zn(II), in an aqueous inorganic
medium, where it showed a preference for binding to those ions via
a (O,O)-bis-chelating mode.19 However, studies on the interaction
of this functional group with the active site of metalloenzymes have
not been reported. The most accurate way for getting information
on the inhibitor−enzyme binding mode is undoubtedly via X-ray
crystallographic studies of the corresponding complexes. However,
in its absence, theoretical approaches have been accepted as
secondary tools to produce reliable results. With this aim, we
focused our attention on the binding geometry of this new ZBG
(in particular, compound 1c with MMP13) and developed a
computational approach based on docking and hybrid quantum
mechanical/molecular mechanical (QM/MM) calculations.
Although in many cases docking calculations have been successfully
applied with MMP inhibitors, the coordination of the reported
ZBG with MMPs has never been predicted. Therefore, to support
the reliability of the docking calculations and to have a deeper
analysis of the ZBG interactions, a QM/MM approach was applied
to the docking results.
The docking of compound 1c into MMP13 was conducted
to provide possible binding hypotheses of the ZBG. The ASP
fitness-scoring function of GOLD software was used, as it was
proven a good method for docking MMP inhibitors.18 As
explained in detail in the Experimental Section, two docking
runs were carried out, considering the catalytic zinc ion either
with a tetrahedral or a trigonal bipyramidal coordination
geometry. The two docking runs converged in a unique ligand
binding mode (see Figure 2) with the ligand chelating the zinc
To confirm the reliability of the docking results and to
further refine the structure of the 1c−MMP13 complex, a high-
level computational modeling study was performed and is
presented below.
3.4. QM/MM Dynamics Simulations. The 1c−MMP13
complex resulting from the docking calculation was subjected to
QM/MM molecular dynamics (MD) simulations. Many authors
in the field of drug design have used the QM/MM approaches,
with good success in finding the correct interactions in biological
systems.22,23 In these calculations, the molecular segment with
the highest interest, in terms of binding interaction, is quantum
mechanically treated, while the rest of the system is treated by
means of classical mechanics, thus saving considerable computa-
tional expenses. We used the QM/MM module recently
implemented in AMBER10,24 applying in the QM system the
density functional theory-based tight-binding (DFTB) Hamiltonian
model.25 Because the zone of highest interest includes a
metal ion (Zn), we thought this method would be the most
suitable for obtaining an accurate prediction rather than a purely
MM dynamics simulation. There are two ways to model the
force field of this zinc ion in the MM dynamics simulations: the
bonded model and the nonbonded model. In the bonded model,
the commonly used bonded terms describe the coordinates
between zinc and ligand/MMP, including bond stretching. In the
nonbonded approach, van der Waals and nonbonded electro-
static terms are used to model the zinc−ligand/MMP
interactions. The bonded model is quite efficient but requires
the parametrization of the interactions between the zinc ion and
the ligand/MMP.26 On the other hand, the nonbonded method
is highly sensitive to the choice of the electrostatic model and can
suffer from an inability to retain a low coordination number.
Furthermore, with the AMBER force field, the nonbonded
approach generally fails to give the correct coordination number,
even when the long-range electrostatic interactions are correctly
accounted for using an infinite cutoff.27,28
For these reasons, we preferred to adopt a hybrid QM/MM
approach that would be able to avoid the problems correlated with
the MM dynamic simulations. As already mentioned, the QM/MM
module was recently reported and, up to now, has never been used
to analyze inhibitor−MMP complexes. Therefore, prior to
evaluating our 1c−MMP13 binding hypothesis, we tested the
reliability of the method by launching QM/MM simulations on two
MMP13 crystal structures, one containing a hydroxamate inhibitor
(4-[4-(4-chloro-phenoxy)-benzenesulfonylmethyl]-tetrahydro-
pyran-4-carboxylic acid hydroxyamide, PDB entry 830C29) and
another containing a barbiturate (5-(2-ethoxyethyl)-5-[4-(4-
fluorophenoxy)phenoxy]pyrimidine-2,4,6(1H,3H,5H)-trione, PDB
entry 1YOU30). These were chosen based on the similarity between
our ZBG and both hydroxamate and barbiturate groups. Each
structure, in an explicit solvent environment, was subjected to two
minimization and three MD steps, the last one consisting of 4 ns of
QM/MM MD simulation (see the Experimental Section for details).
As shown in Figure 3, both complexes seemed to be stable
during the simulations. By analyzing the root-mean-square
deviation (rmsd) of all the heavy atoms from the X-ray structures,
we observed an initial increase due to the equilibration of the
system, followed, after 500 ps, by a stabilization of the rmsd
value around 1.1 Å. Regarding the geometry of the ligand, we
analyzed the rmsd of the position of the ligands with respect to
the X-ray structures during the simulation. Figure 3 shows that
both ligands demonstrated rmsd values between 0.3 and 0.7 Å.
From the X-ray structures of the complexes, we observed that
the two ligands formed four H-bonds with MMP13. As shown
Figure 2. Superimposition of the 1c best docking pose into MMP13,
considering the catalytic zinc ion with a tetrahedral (1c colored light-
blue) and a trigonal bipyramidal (1c colored magenta) coordination
geometry. The crystal structures of MMP13 complexed with a
pyrimidinetrione-based inhibitor (yellow, PDB entry 1YOU) is
displayed as a reference structure.
ion through the 6-ketone oxygen atom and the 1-hydroxyl
group of the piperazine ring, which also formed an H-bond
with Glu223. The remain part of compound 1c was placed in
the binding site, similarly to many compounds containing
arylsulfonyl moieties,20,21 with the sulfonyl O-atoms forming H-
bonds with the backbone N-atoms of Leu185 and Ala186, at
the entrance of the S1′ cavity, while the aromatic rings were
well inserted into this hydrophobic cavity.
8292
dx.doi.org/10.1021/jm200593b|J. Med. Chem. 2011, 54, 8289−8298

Journal of Medicinal Chemistry
Article
the QM/MM MD), showed that all four H-bonds were present
and the ligand geometries were quite similar to the X-ray starting
structures (see Figure S3 in the Supporting Information).
Once we tested and validated the QM/MM simulation
approach for modeling ligands−MMP complexes, we subjected
the 1c−MMP13 complex obtained from the docking studies to
the same protocol used above. Figure 4 shows that the 1c-complex
Figure 4. Analysis of the QM/MM simulations for the 1c−MMP13
complex. The plot shows the rmsd during the simulations of the heavy
atoms of the whole system (upper curve, HAS) and of the heavy atoms
of 1c (downward curves, HAL) from the starting model structures.
seemed to be stable during the simulations. The rmsd of the
heavy atoms from the X-ray structures was very similar to those
reported for the hydroxamate and barbiturate derivatives. The
rmsd presented a small initial increase, but after 500 ps,
remained between 0.9 and 1.1 Å for the three systems.
Concerning the geometry of the ligand, after an initial increase,
it showed a rmsd value between 0.4 and 0.7 Å in the last 4 ns.
Figure 5 shows the minimized average structure of the last 4
ns of the MD simulation of the 1c−MMP13 complex. The zinc
Figure 3. Analysis of the QM/MM simulations for the hydroxamic-
(A) and barbiturate-based compounds (B) complexed with MMP13.
The plots show the rmsd during the simulations of the heavy atoms of
the whole system (upper curve, HAS) and of the heavy atoms of the
ligands (downer curves, HAL) from the starting model structures.
in Table 2, both compounds interacted with Leu185, Ala186,
and Glu223, and these interactions were maintained during the
Table 2. H-Bond Analysis during the 5 ns MD Simulation for
the Hydroxamate (a) and Barbiturate (b) Based Ligands
distance
(Å)
%
distance
(Å)
%
H-bond
occupied
H-bond
occupied
HB1
HB2
HB3
HB4
2.7
3.0
3.3
3.0
99.6
68.4
43.0
93.8
HB1
HB2
HB3
HB4
2.6
2.8
3.2
2.9
87.4
100.0
71.3
Figure 5. Minimized average structures resulting from the MD
simulation of the 1c−MMP13 complex (orange) superimposed with
the crystal structures of MMP13 complexed with a hydroxamate
inhibitor (green, PDB entry 830C).
92.9
MD simulation. Finally, the complexes obtained by minimizing
the average structure of the last 4 ns of MD (corresponding to
ion displayed a trigonal bipyramidal geometry, chelated by the
ligand through the 6-ketone O-atom and the 1-hydroxyl group
8293
dx.doi.org/10.1021/jm200593b|J. Med. Chem. 2011, 54, 8289−8298

Journal of Medicinal Chemistry
Article
of the piperazine ring, which also formed an H-bond with the
carboxylate of Glu223. With regards to the sulfonyl O-atom, it
formed H-bonds with the backbone N-atoms of Leu185 and
Ala186 at the entrance of the S1′ cavity, while the aromatic
rings were inserted into this hydrophobic cavity. As shown in
Table S1 in the Supporting Information, the H-bond analysis of
the simulation confirmed the interactions described above, as
the three H-bonds were maintained during the MD trajectory.
The comparison between the docking and the QM/MM
results for the 1c−MMP13 complex highlighted the existence
of considerable differences on the ZBG disposition. As shown
in Figure S4A in the Supporting Information, the HPD resulting
from the QM/MM calculations was rotated about 90° with
respect to the ZBG docking disposition. A further analysis of
the GOLD docking poses for 1c revealed that, considering a
trigonal bipyramidal coordination geometry for the zinc, the
pose ranked as tenth showed a disposition very similar to the
QM/MM results (see Supporting Information Figure S4B).
These data support the hypothesis that GOLD is able to find
the correct binding disposition of the HPD moiety but is
unable to rank it as the preferred disposition and is therefore in
agreement with the use of QM/MM calculations for refining
the starting docking results.
optimizations converged into solutions similar to the structure
obtained from the QM/MM calculation. In Figure 6, the final
Figure 6. Superimposition of the minimized average structures
resulting from the 1c−MMP13 simulation (orange) and the simplified
model optimized with the QM method (green).
The final average structure of the 1c−MMP13 complex,
arising from the hybrid QM/MM MD simulation, gave extra
evidence of the potential of the HPD moiety as an effective
ZBG to be used in the inhibition of MMPs and other
metalloproteinases. Beyond the modification of the 4-
arylsulfonyl substituent that is known to modulate activity
against the different MMP subtypes,31,32 a structure optimiza-
tion can still be performed through substitution at the 3- and
5-positions of the HPD ring. As shown in Figure S5A in the
Supporting Information, the superimposition of the 1c−
MMP13 complex with the crystal structure of MMP13
complexed with a potent pyrimidinetrione-based inhibitor
highlights the fact that the 3-position of HPD corresponds to
the C-5 position of the pyrimidinetrione ring. Previous studies
on the C-5 substitution of the pyrimidinetrione-based inhibitor
showed that this portion of the ligands interacts with the
solvent-exposed S1 region of the MMPs, which is useful for
modulating their activity.30,33 Furthermore, an opportune
substitution at the 3-position of HPD could also mimic the
5-position of hydantoin-based MMP inhibitors (see Figure S5B
in the Supporting Information), which can bear large
substituents that are able to drastically change the MMP
inhibition activity.34,35 Finally, the superimposition between the
1c−MMP13 complex and MMP3 complexed with a potent
heterocycle-based inhibitor (see Figure S5C in the Supporting
Information), highlights the fact that the 5-position of HPD
could be profitably used to explore the S2′ region, which is
considered a useful feature for modulating the activity and
selectivity of MPIs.36
structures of the QM optimization (6-31++G(d,p) basis set) and
the average of the QM/MM MD simulation of 1c−MMP13 are
superimposed. These two structures are very close to each other,
also when considering the ZBG, with a rmsd of 0.8 Å for the heavy
atoms of the HPD ring. Such facts emphasize the accuracy of our
computational method when investigating the binding mode of
the new ZBG type.
4. CONCLUSIONS
A new set of compounds containing a novel zinc-binding moiety
was identified and proved to be effective MMP inhibitors,
with the highest inhibition observed for compound 1c toward
MMP13 (IC50 value of 9.5 nM). The 1-hydroxypiperazine-2,6-
dione (HPD) moiety presented itself as a valid and promising
ZBG to be used in a new class of metalloproteinase inhibitors. A
theoretical approach was employed to analyze the binding mode
of new types of inhibitors with metalloproteins and, in
particular, to disclose the binding geometry of compound 1c
with MMP13. The resulting structure was suitable for fast
docking calculations using scaffold constraints, which can be
applied to the design optimization of new derivatives with
improved inhibitory profiles toward the target enzymes. The
general structure of the new inhibitors (Chart 1) has three main
feature points available for optimization: the aromatic group
“Ar”, interacting within the hydrophobic pocket S1′, the
substituent group at 3-position of the HPD ring interacting
with the S1 region, and the substituent group at 5-position of
the ZBG ring pointing toward the S2′ region.
3.5. QM Optimization. To confirm the results obtained by
the docking-QM/MM approach, QM optimization of the
interaction between the ZBG and MMP13 was carried out. The
catalytic zinc binding region was retrieved from the minimized
average structure of the last 4 ns of 1c−MMP13 MD simulation
and submitted to purely quantum mechanics optimization, using
Gaussian software, with the B3LYP chemical model and the
LANL2DZ and the 6-31++G(d,p) basis set. Simplifications on the
binding site were made, and only the main atoms around the ZBG
were included, while the phenoxyphenyl group of 1c was reduced
to a simpler methoxyphenyl (see Experimental Section). The QM
5. EXPERIMENTAL SECTION
5.1. Synthesis of the Compounds. General Methods and
Materials. Analytical grade reagents were purchased from Sigma-
Aldrich, Fluka and Acros and were used as supplied. Solvents were
dried according to standard methods.37 The chemical reactions were
monitored by TLC using alumina plates coated with silica gel 60
F254 (Merck). The purity of the compounds was determined by
HPLC, and it was found to be higher than 95% for all target
compounds (see Supporting Information for details). The melting
points were measured with a Leica Galen III hot stage apparatus and
1
are uncorrected. The H NMR spectra were recorded on a Bruker
8294
dx.doi.org/10.1021/jm200593b|J. Med. Chem. 2011, 54, 8289−8298

Journal of Medicinal Chemistry
Article
AVANCE III 300 MHz spectrometer, and the 13C NMR spectra were
measured on a Bruker AVANCE III 400 MHz spectrometer, all at
room temperature. Chemical shifts (δ) are reported in ppm from the
standard internal reference tetramethylsilane (TMS) for the organic
solvents, sodium 3-(trimethylsilyl)-[2,2,3,3-d4]-propionate (DSS) for
D2O solutions, or the solvent peak for the 13C NMR. The following
abbreviations are used: s = singlet, d = doublet, t = triplet, m =
multiplet. For the NMR-monitored stability studies, solutions of 1a
(ca. 15 mM) were prepared in D2O, and the pD (pD = −log [D+])
was adjusted by addition of DCl and KOD solutions in D2O. The mass
spectra (FAB) were performed in a VG TRIO-2000 GC/MS
instrument. The elemental analyses were performed for the target
compounds (1a−c) on a Fisons EA1108 CHNF/O instrument and
were within the limit of ±0.4%.
Preparation of Compounds 3a−3c. (R)-1-(Benzyloxy)-3-iso-
propyl-4-(4-methoxyphenylsulfonyl)piperazine-2,6-dione (3a). To
a solution of (R)-2-(N-(1-(benzyloxyamino)-3-methyl-1-oxobutan-
2-yl)-4-methoxyphenylsulfonamido)-acetic acid (2a), obtained as
previously reported18 (0.200 g, 0.44 mmol), in dry THF (30 mL) at
0 °C was added ethylchloroformate (ECF, 0.050 mL, 0.53 mmol) and
dry N-methylmorpholine (NMM, 0.058 mL, 0.53 mmol); the mixture
was stirred for 1 h, and then it was evaporated. The residue was taken
into 1:1 ethyl ether/ethyl acetate (40 mL), and this solution was
washed with 0.1 M HCl (2 × 40 mL), 5% NaOH (2 × 40 mL), and
water (40 mL). After drying the organic phase over anhydrous
Na2SO4, solvent evaporation under vacuum afforded the pure product
as a white solid (0.112 g, 59% yield); mp 75−76 °C. 1H NMR
(CDCl3), δ: 7.74 (d, J = 8.7 Hz, 2H, ArH), 7.35 (s, 5H, PhH), 7.00 (d,
J = 8.4 Hz, 2H, ArH), 4.73 (d, J = 19.8 Hz, 1H, NCH2CON), 4.39 (d,
J = 9.0 Hz, 1H, CH2Ph), 4.32 (d, J = 9.0 Hz, 1H, CH2Ph), 4.22 (d, J =
10.2 Hz, 1H, NCH(iPr)CON), 4.05 (d, J = 19.5 Hz, 1H,
NCH2CON), 3.77 (s, 3H, OCH3), 1.99−1.90 (m, 1H, CH(CH3)2),
1.10 (d, J = 6.3 Hz, 3H, CHCH3), 1.03 (d, J = 6.9 Hz, 3H, CHCH3);
m/z (FAB): 433 (M + H)+, 455 (M + Na)+.
(R)-4-(Biphenyl-4-ylsulfonyl)-1-hydroxy-3-isopropylpiperazine-
1
2,6-dione (1b). White solid (86% yield); mp 164−165 °C. H NMR
(CDCl3), δ: 7.81 (d, J = 8.4 Hz, 2H, ArH), 7.74 (d, J = 9.0 Hz, 2H,
ArH), 7.59 (d, J = 6.6 Hz, 2H, ArH), 7.51−7.43 (m, 3H, ArH), 4.80
(d, J = 19.5 Hz, 1H, NCH2CON), 4.34 (d, J = 9.6, 1H,
NCH(iPr)CON), 4.15 (d, J = 19.5 Hz, 1H, NCH2CON), 2.17−2.03
(m, 1H, CH(CH3)2), 1.18−1.12 (m, 6H, CH(CH3)2). 13C NMR
(CDCl3), δ: 164.3, 161.6 (2 peaks, CO), 147.2, 138.8, 135.5, 129.0,
(4 peaks, para- and ipso-Ar-C), 129.3, 128.6, 127.55, 127.52 (4 peaks,
ortho- and meta-Ar-C), 64.0 (CCH(CH3)2), 45.6 (NCH2CO), 29.5
(CH(CH3)2), 20.2, 19.2 (2 peaks, H(CH3)2); m/z (FAB): 389 (M +
H)+, 411 (M + Na)+.
(R)-1-Hydroxy-3-isopropyl-4-(4-phenoxyphenylsulfonyl)-
1
piperazine-2,6-dione (1c). White hygroscopic solid (96% yield). H
NMR (CDCl3), δ: 7.68 (d, J = 9.0 Hz, 2H, ArH), 7.40 (t, 2H, ArH),
7.42 (t, J = 7.8 Hz, 1H, ArH), 7.06−7.03 (m, 4H, ArH), 4.75 (d, J =
19.8 Hz, 1H, NCH2CON), 4.28 (d, J = 9.6 Hz, 1H, NCH(iPr)CON),
4.12 (d, J = 19.2 Hz, 1H, NCH2CON), 2.14−2.03 (m, 1H,
CH(CH3)2), 1.16−1.11 (m, 6H, CH(CH3)2). 13C NMR (CDCl3),
δ: 164.5, 162.7 (2 peaks, CO), 162.1, 154.7, 130.0, 125.2, (4 peaks,
para- and ipso-Ar-C), 130.2, 129.2, 120.3, 118.3 (4 peaks, ortho- and
meta-Ar-C), 63.9 (CCH(CH3)2), 45.4 (NCH2CO), 29.2 (CH(CH3)2),
20.0, 19.0 (2 peaks, CH(CH3)2); m/z (FAB): 427 (M + Na)+, 405
(M + H)+.
5.2. X-ray Diffraction. The crystallographic data are displayed in
Table 3. Colorless crystals of compound 1b were obtained from slow
Table 3. Crystallographic Data for Compound 1b
1b
chemical formula
Mr
C19H20N2O5S
388.43
temperature/K
wavelength/Å
morphology, color
crystal size/mm
crystal system
space group
a/Å
150(2)
0.71069
plate, colorless
0.20 × 0.12 × 0.04
monoclinic
P21
(R)-1-(Benzyloxy)-4-(biphenyl-4-ylsulfonyl)-3-isopropylpipera-
1
zine-2,6-dione (3b). White solid (40% yield); mp 130−133 °C. H
NMR (CDCl3), δ: 7.88 (d, J = 8.1 Hz, 2H, ArH), 7.74 (d, J = 8.4 Hz,
2H, ArH), 7.46−7.38 (m, 5H, ArH), 7.25 (s, 5H, ArH), 4.79 (d, J =
19.8 Hz, 1H, NCH2CON), 4.32−4.22 (m, 3H, CH2Ph, NCH(iPr)-
CON), 4.10 (d, J = 19.5 Hz, 1H, NCH2CON), 2.6−1.05 (m, 1H,
CH(CH3)2), 1.14 (d, J = 6.6 Hz, 3H, CHCH3), 1.07 (d, J = 6.3 Hz,
3H, CHCH3); m/z (FAB): 479 (M + H)+, 501 (M + Na)+.
7.1064(5)
7.8402(6)
16.3344(12)
94.589(4)
907.16(12)
2
b/Å
c/Å
β/deg
V/Å3
(R)-1-(Benzyloxy)-3-isopropyl-4-(4-phenoxyphenylsulfonyl)-
1
piperazine-2,6-dione (3c). White hygroscopic solid (44% yield). H
NMR (CDCl3), δ: 7.73 (d, J = 9.0 Hz, 2H, ArH), 7.43−7.38 (m, 4H,
ArH), 7.25−7.15 (m, 4H, ArH), 7.03 (d, J = 9.0 Hz, 2H, ArH), 6.83
(d, J = 7.8 Hz, 2H, ArH), 4.73 (d, J = 18.0 Hz, 1H, NCH2CON), 4.44
(d, J = 8.7 Hz, 1H, CH2Ph), 4.39 (d, J = 8.4 Hz, 1H, CH2Ph), 4.23 (d,
J = 10.2 Hz, 1H, NCH(iPr)CON), 4.07 (d, J = 19.5 Hz, 1H,
NCH2CON), 2.01−1.92 (m, 1H, CH(CH3)2), 1.12 (d, J = 6.3 Hz, 3H,
CHCH3), 1.05 (d, J = 6.6 Hz, 3H, CHCH3); m/z (FAB): 495 (M +
H)+, 517 (M + Na)+.
Z
calcd density/mg·m−3
absorption coefficient/mm−1
θ min/deg
θ max/deg
reflns collected/unique
Rint
1.422
0.213
1.25
27.94
10231/4294
0.0506
GoF
0.988
Preparation of Compounds 1a−1c. (R)-1-Hydroxy-3-isopropyl-
4-(4-methoxyphenylsulfonyl)piperazine-2,6-dione (1a). The general
method for benzyl deprotection by catalytic hydrogenolysis was
followed: to a solution of 3a (0.100 g, 0.23 mmol) in methanol
(10 mL), 10% Pd/C (0.025 g) was added and the suspension was
stirred for 3 h under H2 (1.5 bar). After filtration and solvent removal,
final recrystallization from ethyl ether/petroleum ether afforded the
threshold expression
R1 (obsd)
I > 2σ(I)
0.0487
wR2 (all)
0.0967
evaporation of a dichloromethane solution of the compound. The
single crystals were mounted on a cryoloop using Fomblin as
protective oil. Single-crystal X-ray diffraction data was collected at
150 K on a Bruker AXS-KAPPA APEX II diffractometer with graphite-
monochromated radiation (Mo Kα, λ = 0.71069 Å). X-ray data
collection was monitored by SMART program (Bruker, 2003). All the
data were corrected for Lorentzian, polarization, and absorption effects
using SAINT38 program. SIR9739 was used for structure solution, and
SHELXL-9740 was used for full matrix least-squares refinement on F2.
All non-hydrogen atoms were refined anisotropically. All H atoms
were added in calculated positions and refined riding on their resident
1
pure product as a white hygroscopic solid (0.076 g, 96% yield). H
NMR (CDCl3), δ: 7.68 (d, J = 8.7 Hz, 2H, ArH), 6.98 (d, J = 8.7 Hz,
2H, ArH), 4.75 (d, J = 19.5 Hz, 1H, NCH2CON), 4.27 (d, J = 9.9 Hz,
1H, NCH(iPr)CON), 4.11 (d, J = 19.5 Hz, 1H, NCH2CON), 3.86
(s, 3H, OCH3), 2.13−2.02 (m, 1H, CH(CH3)2), 1.15−1.10 (m, 6H,
CH(CH3)2). 13C NMR (CDCl3), δ: 164.5, 164.1 (2 peaks, CO),
162.5, 128.4, (2 peaks, para- and ipso-Ar-C), 129.3, 115.2 (2 peaks,
ortho- and meta-Ar-C), 64.0 (CCH(CH3)2), 55.9 (OCH3), 45.5
(NCH2CO), 29.4 (CH(CH3)2), 20.1, 19.2 (2 peaks, CH(CH3)2); m/z
(FAB): 343 (M + H)+, 365 (M + Na)+.
8295
dx.doi.org/10.1021/jm200593b|J. Med. Chem. 2011, 54, 8289−8298

Journal of Medicinal Chemistry
Article
atoms. ORTEP 3.241 was used for molecular representations, and
MERCURY 2.342 was used for packing diagrams.
ions were added as counterions to neutralize the system. Prior to the
MD, two minimization steps were carried out. In the first stage, the
protein was fixed using a harmonic force constant of 500 kcal/mol·Å2
and only the positions of the water molecules were minimized. In the
second stage, the entire system was minimized by applying a harmonic
force constraint of 15 kcal/mol·Å2 on the α-carbons, the two zinc ions,
and the corresponding coordinating atoms (from the protein and the
ligand). The first minimization consisted of 5000 steps with a
combined algorithm, namely the sequential use of Steepest Descent
(SD) and Conjugate Gradient (CG) methods, for the first 1000 and
the last 4000 steps, respectively, while the second one consisted of
9000 steps with 1000 SD and 8000 CG steps.
5.3. Enzyme Inhibition.43 Recombinant human progelatinase
A (pro-MMP2) and B (pro-MMP9), pro-MMP8, MMP16, and MMP14
catalytic domains were a kind gift of Prof. Gillian Murphy (Department of
Oncology, University of Cambridge, UK). Pro-MMP1, pro-MMP3, pro-
MMP7, pro-MMP13, and TACE (ADAM-17) were purchased from
Calbiochem. Pro-MMP12 was purchased by R&D Systems.
Proenzymes were activated immediately prior to use with p-
aminophenylmercuric acetate (APMA, 2 mM for 1 h at 37 °C for
MMP2 and MMP8, 1 mM for 1 h at 37 °C for MMP9 and MMP13, 2
mM for 2 h at 37 °C for MMP1, 3.3 mM for 2 h at 37 °C for MMP7).
Pro-MMP3 was activated with trypsin, 5 μg/mL for 30 min at 37 °C,
followed by soybean trypsin inhibitor (SBTI, 62 μg/mL). Pro-MMP12
was autoactivated by incubating in the fluorimetric assay buffer (FAB:
Tris 50 mM, pH = 7.5, NaCl 150 mM, CaCl2 10 mM, Brij 35 0.05%,
and DMSO 1%) for 30 h at 37 °C.
For the assay measurements, the inhibitor stock solutions (DMSO,
10 mM) were diluted at seven different concentrations for each MMP
in FAB. The activated enzyme (final concentration 0.5 nM for MMP2,
1.3 nM for MMP9, 1.5 nM for MMP8, 0.3 nM for MMP13, 5 nM for
MMP3, 1.3 nM for MMP7, 1.0 nM for MMP14 cd, 15 nM for
MMP16 cd, 2.0 nM for MMP1, 1.0 nM for MMP12, and 7.5 nM for
TACE) and inhibitor solutions were incubated in the assay buffer for
4 h at 25 °C. After the addition of 200 μM solution of the fluorogenic
substrate Mca-Arg-Pro-Lys-Pro-Val-Glu-Nva-Trp-Arg-Lys(Dnp)-NH2
(Sigma) for MMP3 and Mca-Lys-Pro-Leu-Gly-Leu-Dap(Dnp)-Ala-
Arg-NH2 (Bachem) for all the other enzymes in DMSO (final
concentration 2 μM), the hydrolysis was monitored every 15 s for
Particle mesh Ewald50 (PME) electrostatics and periodic boundary
conditions were used in the simulation. The MD trajectory was
conducted using the minimized structure as the starting conformation.
The time step of the simulations was 2.0 fs with a cutoff of 10 Å for the
nonbonded interaction, and SHAKE was employed to keep rigid all
bonds involving hydrogen atoms. Constant volume was carried out for
200 ps, during which the temperature was raised from 0 to 300 K
(using the Langevin dynamics method). Next, under constant-
pressure, an 800 ps MD simulation was carried out at 300 K. During
these two MD steps, the same constraints were applied as described
for the second minimization step. General Amber force field (GAFF)
parameters were assigned to the ligands, while partial charges were
calculated using the AM1-BCC method as implemented in the
Antechamber suite of AMBER 10. A 4 ns MD simulation was then
performed using the hybrid quantum mechanical/molecular mechan-
ical (QM/MM) method of AMBER 10. The quantum mechanics
(QM) region was described by the DFTB theory25 and contained the
catalytic zinc ion, the imidazole rings of the three His residues
surrounding it, the Glu223 carboxylic group, and the ZBG
(hydroxamic acid and barbiturate moiety). Aside from those atoms
within the QM region, the same constraints of the previous MD were
applied. In the molecular mechanics (MM) region, the parameters
were the same as in the previous MD, but in the QM region, the PME
algorithm was deactivated; and this region’s charge was defined as 1.
This MD simulation protocol was also applied to the MMP13−
compound 1c complex that resulted from the docking study.
5.6. QM Optimization. Geometry optimization was performed by
means of quantum mechanical calculations derived from the Gaussian
03W software.51 The minimized average structure of the last 4 ns of
the MD simulation of the MMP13−compound 1c complex was used
as the starting structure. Only the most important residues of the
binding site region were taken into account, i.e. the catalytic zinc ion,
the imidazole rings of the three His residues surrounding it, the
Glu223 carboxylic group, Leu185, Ala186, Val219, and the ligand 1c,
with its portion 4-phenoxyphenylsulfonyl substituted by a smaller 4-
methoxyphenylsulfonyl. The QM calculation was carried out using the
B3LYP chemical model and two different basis sets (i.e., the
LANL2DZ and the 6-31++G(d,p) basis set). A direct self-consistent
field (SCF) method with a SCF convergence criterion of 10−5 was
used. The backbone atoms of the residues were kept fixed.
20 min, recording the increase in fluorescence (λ ex = 325 nm, λem
=
395 nm) using a Molecular Devices SpectraMax Gemini XS plate
reader. The assays were performed in a total volume of 200 μL per
well in 96-well microtiter plates (Corning, black, NBS). Control wells
lack inhibitor. The MMP inhibition activity was expressed in relative
fluorescent units (RFU). Percent of inhibition was calculated from
control reactions without the inhibitor. IC50 was determined using the
formula: Vi/Vo = 1/(1 + [I]/ IC50), where Vi is the initial velocity of
substrate cleavage in the presence of the inhibitor at concentration [I]
and Vo is the initial velocity in the absence of the inhibitor. Results
were analyzed using SoftMax Pro software44 and GraFit software.45
5.4. Docking of Compound 1c. The ligand structure was built
using Maestro 9.046 and it was minimized with Macromodel 9.7.47 The
conjugated gradient method was applied, until a convergence value of
0.05 kJ/Å·mol was reached, using the MMFFs force field and a water
environment model (generalized-Born/surface-area model), with a
distance-dependent dielectric constant of 1.0. The minimized ligand
was then subjected to a conformational search of 100 steps, using an
algorithm based on the Monte Carlo method, with the same force field
and parameters used in the minimization. The structure of the
MMP13 was extracted from the RCSB Protein Data Bank48 (PDB
code 830C29). Hydrogen atoms were added by means of Maestro, and
the region of interest used by the docking program GOLD version
4.049 was defined in order to contain the residues within 15 Å from the
original position of the ligand in the X-ray structure. In the docking
calculations, the catalytic zinc ion was set to have either a tetrahedral
or a trigonal bipyramidal coordination geometry. The “allow early
termination” option was deactivated. The clusterization was set for an
rmsd limit of 0.75 Å between the different docking solutions. The
remaining GOLD default parameters were used, and the ligand was
submitted to 100 genetic algorithm runs by applying the ASP fitness
scoring function. The clustered structures obtained from the two
docking runs were then compared.
ASSOCIATED CONTENT
■
S
* Supporting Information
Extra data on the crystallographic structure, enzyme inhibition,
QM/MM dynamic structure, and compounds characterization.
This material is available free of charge via the Internet at
http://pubs.acs.org.
5.5. QM/MM Simulations. All simulations were performed using
AMBER 10.24 Molecular dynamics (MD) simulations were carried out
using the parm03 force field at 300 K. Two MMP13 structures
AUTHOR INFORMATION
■
Corresponding Authors
complexed with inhibitors (a hydroxamate, PDB entry 830C;29
a
barbiturate, PDB entry 1YOU30) were taken from the RCSB Protein
Data Bank.48 The complexes were placed in a rectangular
parallelepiped water box, an explicit solvent model for water, TIP3P,
was used, and the complex was solvated with a 10 Å water cap. Sodium
*For M.A.S.: phone, 00351-218419273; fax, 00351-218464455;
e-mail, masantos@ist.utl.pt. For T.T.: phone, 0039-
0502219595; fax, 0039-0502219605; e-mail, tuccinardi@farm.
unipi.it.
8296
dx.doi.org/10.1021/jm200593b|J. Med. Chem. 2011, 54, 8289−8298

Journal of Medicinal Chemistry
Article
(13) Overall, C. M.; Kleifeld, O. Towards third generation matrix
metalloproteinase inhibitors for cancer therapy. Br. J. Cancer 2006, 94,
941−946.
ACKNOWLEDGMENTS
■
We acknowledge the Portuguese Fundaca
̧
o para a Cien
̂
cia e
̃
Tecnologia (FCT) for financial support, with the postdoc grant
SFRH/BPD/29874/2006 (S.M.M.) and the Ph.D. grant
SFRH/BD/40474/2007 (V.A.). We also thank the Portuguese
NMR Network (IST-UTL Center) for providing access to the
NMR facility. The mass spectra were obtained at the IST Node,
which is part of the National Mass Spectrometry Network
(RNEM) created by the FCT. We are also thankful to Prof.
Teresa Duarte for the helpful discussions on the X-ray analysis.
Molecular graphics images were produced using the UCSF
Chimera package from the Resource for Biocomputing,
Visualization, and Informatics at the University of California,
San Francisco (supported by NIH grant P41 RR01081).
(14) Saghatelian, A.; Jessani, N.; Joseph, A.; Humphrey, M.; Cravatt,
B. F. Activity-based probes for the proteomic profiling of metal-
loproteases. Proc. Natl. Acad. Sci. U.S.A 2004, 101, 10000−10005.
(15) Puerta, D. T.; Griffin, M. O.; Lewis, J. A.; Romero-Perez, D.;
Garcia, R.; Villarreal, F. J.; Cohen, S. M. Heterocyclic zinc-binding
groups for use in next-generation matrix metalloproteinase inhibitors:
potency, toxicity, and reactivity. J. Biol. Inorg. Chem. 2006, 11,
131−138.
(16) Ledour, G.; Moroy, G.; Rouffet, M.; Bourguet, E.; Guillaume,
D.; Decarme, M.; Elmourabit, H.; Auge, F.; Alix, A. J.; Laronze, J. Y.;
Bellon, G.; Hornebeck, W.; Sapi, J. Introduction of the 4-(4-
bromophenyl)benzenesulfonyl group to hydrazide analogs of Ilomastat
leads to potent gelatinase B (MMP-9) inhibitors with improved
selectivity. Bioorg. Med. Chem. 2008, 16, 8745−8759.
(17) Li, D. L.; Zheng, Q. C.; Fang, X. X.; Ji, H. T.; Yang, J. G.; Zhang,
H. X. Theoretical study on potency and selectivity of novel non-
peptide inhibitors of matrix metalloproteinases MMP-1 and MMP-3.
Polymer 2008, 49, 3346−3351.
(18) Marques, S. M.; Nuti, E.; Rossello, A.; Supuran, C. T.;
Tuccinardi, T.; Martinelli, A.; Santos, M. A. Dual inhibitors of matrix
metalloproteinases and carbonic anhydrases: iminodiacetyl-based
hydroxamate−benzenesulfonamide conjugates. J. Med. Chem. 2008,
51, 7968−7979.
ABBREVIATIONS USED
■
HPD, 1-hydroxypiperazine-2,6-dione; MMP, matrix metallopro-
teinase; QM, quantum mechanics; MM, molecular mechanics;
ZBG, zinc-binding group; MPI, MMP inhibitor; ECF, ethyl-
chloroformate; TBTU, O-(benzotriazol-1-yl)-N,N,N′,N′-tetrame-
thyluronium tetrafluoroborate; MD, molecular dynamics; DFTB,
density functional theory-based-tight-binding; rmsd, root-mean-
square deviation
(19) Chaves, S.; Marques, S.; Santos, M. A. Iminodiacetyl-
hydroxamate derivatives as metalloproteinase inhibitors: equilibrium
complexation studies with Cu(II), Zn(II) and Ni(II). J. Inorg. Biochem.
2003, 97, 345−353.
(20) Nuti, E.; Panelli, L.; Casalini, F.; Avramova, S. I.; Orlandini, E.;
Santamaria, S.; Nencetti, S.; Tuccinardi, T.; Martinelli, A.; Cercignani,
G.; D’Amelio, N.; Maiocchi, A.; Uggeri, F.; Rossello, A. Design,
synthesis, biological evaluation, and NMR studies of a new series of
arylsulfones as selective and potent matrix metalloproteinase-12
inhibitors. J. Med. Chem. 2009, 52, 6347−6361.
(21) Nuti, E.; Casalini, F.; Avramova, S. I.; Santamaria, S.;
Cercignani, G.; Marinelli, L.; La Pietra, V.; Novellino, E.; Orlandini,
E.; Nencetti, S.; Tuccinardi, T.; Martinelli, A.; Lim, N. H.; Visse, R.;
Nagase, H.; Rossello, A. N-O-Isopropyl sulfonamido-based hydrox-
amates: design, synthesis and biological evaluation of selective matrix
metalloproteinase-13 inhibitors as potential therapeutic agents for
osteoarthritis. J. Med. Chem. 2009, 52, 4757−4773.
(22) Topf, M.; Varnai, P.; Richards, W. G. Ab initio QM/MM
dynamics simulation of the tetrahedral intermediate of serine
proteases: insights into the active site hydrogen-bonding network.
J. Am. Chem. Soc. 2002, 124, 14780−14788.
REFERENCES
■
(1) Whittaker, M.; Floyd, C. D.; Brown, P.; Gearing, A. J. Design and
therapeutic application of matrix metalloproteinase inhibitors. Chem.
Rev. 1999, 99, 2735−2776.
(2) Egeblad, M.; Werb, Z. New functions for the matrix
metalloproteinases in cancer progression. Nature Rev. Cancer 2002, 2,
161−174.
(3) Dorman
́
, G.; Cseh, S.; Hajdu, I.; Barna, L.; Kon
́
ya, D.; Kupai, K.;
́
Kovac
́
s, L.; Ferdinandy, P. Matrix metalloproteinase inhibitors: a
critical appraisal of design principles and proposed therapeutic utility.
Drugs 2010, 70, 949−964.
(4) 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.
(5) Coussens, L. M.; Fingleton, B.; Matrisian, L. M. Matrix
metalloproteinase inhibitors and cancer: trials and tribulations. Science
2002, 295 (5564), 2387−2392.
(6) Overall, C. M.; Kleifeld, O. Tumour microenvironment
opinion: validating matrix metalloproteinases as drug targets and anti-
targets for cancer therapy. Nature Rev. Cancer 2006, 6, 227−239.
(7) Pirard, B. Insight into the structural determinants for selective
inhibition of matrix metalloproteinases. Drug. Discovery Today 2007,
12, 640−646.
(8) Flipo, M.; Charton, J.; Hocine, A.; Dassonneville, S.; Deprez, B.;
Deprez-Poulain, R. Hydroxamates: relationships between structure and
plasma stability. J. Med. Chem. 2009, 52, 6790−6802.
(9) 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.
(10) MacPherson, L. J.; Bayburt, E. K.; Capparelli, M. P.; Carroll,
B. J.; Goldstein, R.; Justice, M. R.; Zhu, L.; Hu, S.; Melton, R. A.;
Fryer, L.; Goldberg, R. L.; Doughty, J. R.; Spirito, S.; Blancuzzi, V.;
Wilson, D.; O’Byrne, E. M.; Ganu, V.; Parker, D. T. Discovery of CGS
27023A, a non-peptidic, potent, and orally active stromelysin inhibitor
that blocks cartilage degradation in rabbits. J. Med. Chem. 1997, 40,
2525−2532.
(23) Yonezawa, Y.; Nakata, K.; Sakakura, K.; Takada, T.; Nakamura,
H. Intra- and Intermolecular Interaction Inducing Pyramidalization on
Both Sides of a Proline Dipeptide during Isomerization: An ab Initio
QM/MM Molecular Dynamics Simulation Study in Explicit Water.
J. Am. Chem. Soc. 2009, 131, 4535−4540.
(24) Case, D. A.; Darden, T. A.; Cheatham, T. E., III; Simmerling,
C. L.; Wang, J.; Duke, R. E.; Luo, R.; Crowley, M.; Walker, R. C.;
Zhang, W.; Merz, K. M.; Wang, B.; Hayik, S.; Roitberg, A.; Seabra, G.;
́
Kolossvary, I.; Wong, K. F.; Paesani, F.; Vanicek, J.; Wu, X.; Brozell,
S. R.; Steinbrecher, T.; Gohlke, H.; Yang, L.; Tan, C.; Mongan, J.;
Hornak, V.; Cui, G.; Mathews, D. H.; Seetin, M. G.; Sagui, C.; Babin,
V.; Kollman, P. A. AMBER, version 10; University of California: San
Francisco, CA, 2008.
(25) Seifert, G.; Porezag, D.; Frauenheim, T. Calculations of
molecules, clusters, and solids with a simplified LCAO-DFT-LDA
scheme. Int. J. Quantum Chem. 1996, 58, 185−192.
(26) Tuccinardi, T.; Martinelli, A.; Nuti, E.; Carelli, P.; Balzano, F.;
Uccello-Barretta, G.; Murphy, G.; Rossello, A. Amber force field
implementation, molecular modelling study, synthesis and MMP-1/
MMP-2 inhibition profile of (R)- and (S)-N-hydroxy-2-(N-isopropox-
ybiphenyl-4-ylsulfonamido)-3-methylbutanamides. Bioorg. Med. Chem.
2006, 14, 4260−4276.
(11) Morrison, C. J.; Butler, G. S.; Rodriguez, D.; Overall, C. M.
Matrix metalloproteinase proteomics: substrates, targets, and therapy.
Curr. Opin. Cell Biol. 2009, 21, 645−653.
(12) Strongin, A. Y. Proteolytic and non-proteolytic roles of
membrane type-1 matrix metalloproteinase in malignancy. Biochim.
Biophys. Acta 2010, 1803, 133−141.
8297
dx.doi.org/10.1021/jm200593b|J. Med. Chem. 2011, 54, 8289−8298

Journal of Medicinal Chemistry
Article
(27) Toba, S.; Damodaran, K. V.; Merz, K. M. Jr. Binding
preferences of hydroxamate inhibitors of the matrix metalloproteinase
human fibroblast collagenase. J. Med. Chem. 1999, 42, 1225−1234.
(28) Donini, O. A.; Kollman, P. A. Calculation and prediction of
binding free energies for the matrix metalloproteinases. J. Med. Chem.
2000, 43, 4180−4188.
(29) Lovejoy, B.; Welch, A. R.; Carr, S.; Luong, C.; Broka, C.;
Hendricks, R. T.; Campbell, J. A.; Walker, K. A.; Martin, R.; Van Wart,
H.; Browner, M. F. Crystal structures of MMP-1 and -13 reveal the
structural basis for selectivity of collagenase inhibitors. Nature Struct.
Biol. 1999, 6, 217−221.
(30) Blagg, J. A.; Noe, M. C.; Wolf-Gouveia, L. A.; Reiter, L. A.;
Laird, E. R.; Chang, S. P.; Danley, D. E.; Downs, J. T.; Elliott, N. C.;
Eskra, J. D.; Griffiths, R. J.; Hardink, J. R.; Haugeto, A. I.; Jones, C. S.;
Liras, J. L.; Lopresti-Morrow, L. L.; Mitchell, P. G.; Pandit, J.;
Robinson, R. P.; Subramanyam, C.; Vaughn-Bowser, M. L.; Yocum,
S. A. Potent pyrimidinetrione-based inhibitors of MMP-13 with
enhanced selectivity over MMP-14. Bioorg. Med. Chem. Lett. 2005, 15,
1807−1810.
(31) 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.
(32) Santos, M. A.; Marques, S. M.; Tuccinardi, T.; Carelli, P.;
Panelli, L.; Rossello, A. Design, synthesis and molecular modeling
study of iminodiacetyl monohydroxamic acid derivatives as MMP
inhibitors. Bioorg. Med. Chem. 2006, 14, 7539−7550.
(33) Breyholz, H. J.; Wagner, S.; Faust, A.; Riemann, B.; Holtke, C.;
Hermann, S.; Schober, O.; Schafers, M.; Kopka, K. Radiofluorinated
pyrimidine-2,4,6-triones as molecular probes for noninvasive MMP-
targeted imaging. ChemMedChem 2010, 5, 777−789.
(34) Yu, W.; Guo, Z.; Orth, P.; Madison, V.; Chen, L.; Dai, C.; Feltz,
R. J.; Girijavallabhan, V. M.; Kim, S. H.; Kozlowski, J. A.; Lavey, B. J.;
Li, D.; Lundell, D.; Niu, X.; Piwinski, J. J.; Popovici-Muller, J.; Rizvi,
R.; Rosner, K. E.; Shankar, B. B.; Shih, N. Y.; Siddiqui, M. A.; Sun, J.;
Tong, L.; Umland, S.; Wong, M. K.; Yang, D. Y.; Zhou, G. Discovery
and SAR of hydantoin TACE inhibitors. Bioorg. Med. Chem. Lett. 2010,
20, 1877−1880.
(44) Voigt, J. H.; Elkin, C.; Madison, V. S.; Duca, J. S. Cross-docking
of inhibitors into CDK2 structures. 2. J. Chem. Inf. Model. 2008, 48,
669−678.
(45) Gleeson, M. P.; Gleeson, D. QM/MM as a tool in fragment
based drug discovery. A cross-docking, rescoring study of kinase
inhibitors. J. Chem. Inf. Model. 2009, 49, 1437−1448.
(46) Maestro, version 9.0; Schrodinger Inc.: Portland, OR, 2009.
̈
(47) Macromodel, version 9.7; Schrodinger Inc.: Portland, OR, 2009.
̈
(48) Berman, H. M.; Westbrook, J.; Feng, Z.; Gilliland, G.; Bhat, T.
N.; Weissig, H.; Shindyalov, I. N.; Bourne, P. E. The Protein Data
Bank. Nucleic Acids Res. 2000, 28, 235−242.
(49) Verdonk, M. L.; Cole, J. C.; Hartshorn, M. J.; Murray, C. W.;
Taylor, R. D. Improved protein−ligand docking using GOLD. Proteins
2003, 52, 609−623.
(50) Essmann, U.; Perera, L.; Berkowitz, M. L.; Darden, T.; Lee, H.;
Pedersen, L. G. A smooth particle mesh Ewald method. J. Chem. Phys.
1995, 103, 8577−8593.
(51) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Montgomery, J., J. A.; Vreven, T.;
Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.;
Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.;
Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.;
Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao,
O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross,
J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann,
R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.;
Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg,
J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.;
Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J.
B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.;
Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen,
W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, revision C.02;
Gaussian, Inc.: Wallingford, CT, 2004.
(35) Yu, W.; Tong, L.; Kim, S. H.; Wong, M. K.; Chen, L.; Yang,
D. Y.; Shankar, B. B.; Lavey, B. J.; Zhou, G.; Kosinski, A.; Rizvi, R.; Li,
D.; Feltz, R. J.; Piwinski, J. J.; Rosner, K. E.; Shih, N. Y.; Siddiqui,
M. A.; Guo, Z.; Orth, P.; Shah, H.; Sun, J.; Umland, S.; Lundell, D. J.;
Niu, X.; Kozlowski, J. A. Biaryl substituted hydantoin compounds as
TACE inhibitors. Bioorg. Med. Chem. Lett. 2010, 20, 5286−5289.
(36) Pikul, S.; Dunham, K. M.; Almstead, N. G.; De, B.; Natchus,
M. G.; Taiwo, Y. O.; Williams, L. E.; Hynd, B. A.; Hsieh, L. C.; Janusz,
M. J.; Gu, F.; Mieling, G. E. Heterocycle-based MMP inhibitors with
P2′ substituents. Bioorg. Med. Chem. Lett. 2001, 11, 1009−1013.
(37) Armarego, W. L. F.; Perrin, D. D., Purification of Laboratory
Chemicals, 4th ed.; Butterworth-Heinemann: Oxford, 1999.
(38) Bruker AXS: SAINT+, release 6.22; Bruker Analytical Systems:
Madison, WI, 2005.
(39) Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, G. L.;
Giacovazzo, C.; Guagliardi, A.; Moliterni, A. G. G.; Polidori, G.;
Spagna, R. SIR97: a new tool for crystal structure determination and
refinement. J. Appl. Crystallogr. 1999, 32, 115−119.
(40) Sheldrick, G. M. A short history of SHELX. Acta Crystallogr.,
Sect. A: Found. Crystallogr. 2008, 64 (Pt 1), 112−122.
(41) Barnes, C. L. ORTEP-3 for Windowsa version of ORTEP-III
with a Graphical User Interface (GUI). J. Appl. Crystallogr. 1997, 30,
568−568.
(42) Macrae, C. F.; Bruno, I. J.; Chisholm, J. A.; Edgington, P. R.;
McCabe, P.; Pidcock, E.; Rodriguez-Monge, L.; Taylor, R.; van de
Streek, J.; Wood, P. A. Mercury CSD 2.0new features for the
visualization and investigation of crystal structures. J. Appl. Crystallogr.
2008, 41, 466−470.
(43) Knight, C. G.; Willenbrock, F.; Murphy, G. A novel coumarin-
labelled peptide for sensitive continuous assays of the matrix
metalloproteinases. FEBS Lett. 1992, 296, 263−266.
8298
dx.doi.org/10.1021/jm200593b|J. Med. Chem. 2011, 54, 8289−8298