Development of Thioaryl-Based Matrix Metalloproteinase-12 Inhibitors with Alternative Zinc-Binding Groups: Synthesis, Potentiometric, NMR, and Crystallographic Studies

Article abstract of DOI:10.1021/acs.jmedchem.8b00096

Matrix metalloproteinase-12 (MMP-12) selective inhibitors could play a role in the treatment of lung inflammatory and cardiovascular diseases. In the present study, the previously reported 4-methoxybiphenylsulfonyl hydroxamate and carboxylate based inhibi

Full text of DOI:10.1021/acs.jmedchem.8b00096

Subscriber access provided by NATIONAL UNIV OF SINGAPORE
Article
Development of Thioaryl-based Matrix Metalloproteinase-12
Inhibitors with Alternative Zinc-Binding Groups: Synthesis,
Potentiometric, NMR and Crystallographic Studies
Elisa Nuti, Doretta Cuffaro, Elisa Bernardini, Caterina Camodeca, Laura Panelli,
Silvia Chaves, Lidia Ciccone, Livia Tepshi, Laura Vera, Elisabetta Orlandini, Susanna
Nencetti, Enrico A. Stura, M. Amelia Santos, Vincent Dive, and Armando Rossello
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b00096 • Publication Date (Web): 04 May 2018
Downloaded from http://pubs.acs.org on May 4, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Development of Thioaryl-based Matrix Metalloproteinase-12 Inhibitors
with Alternative Zinc-Binding Groups: Synthesis, Potentiometric, NMR
and Crystallographic Studies
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
a
a,b
a
a
b
Elisa Nuti, Doretta Cuffaro, Elisa Bernardini, Caterina Camodeca, Laura Panelli, Sílvia Chaves,
c,d,
c
c,e¶
f
a
Lidia Ciccone, Livia Tepshi, Laura Vera, Elisabetta Orlandini, Susanna Nencetti, Enrico A.
c
b
c
,a
Stura, M. Amélia Santos, Vincent Dive, and Armando Rossello*
a
Dipartimento di Farmacia, Università di Pisa, Via Bonanno 6, 56126 Pisa, Italy
Centro de Química Estrutural, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais
, 1049ꢀ001 Lisboa, Portugal
b
1
c
CEA, Institut des Sciences du Vivant Frédéric Joliot, Service d’Ingénierie Moléculaire des
Protéines (SIMOPRO), Université ParisꢀSaclay, GifꢀsurꢀYvette, 91190
d
Synchrotron SOLEIL, L’Orme des Merisiers, SaintꢀAubin, BP 48, 91192 GifꢀsurꢀYvette, France
Laboratory of Biomolecular Research, Division of Biology and Chemistry, Paul Scherrer Institute,
232 Villigen, Switzerland
e
5
f
Dipartimento di Scienze della Terra, Università di Pisa, via Santa Maria 53, 56126 Pisa, Italy
*Corresponding Author: Armando Rossello, phone: +39 0502219562; eꢀmail:
armando.rossello@farm.unipi.it
Abstract
Matrix metalloproteinaseꢀ12 (MMPꢀ12) selective inhibitors could play a role in the treatment of lung
inflammatory and cardiovascular diseases. In the present study, the previously reported 4ꢀ
methoxybiphenylsulfonyl hydroxamate and carboxylateꢀbased inhibitors (1b and 2b) were modified to
enhance their selectivity for MMPꢀ12. In the newly synthesized thioaryl derivatives, the nature of the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
zinc binding group (ZBG) and the sulfur oxidation state were changed. Biological assays carried out in
vitro on human MMPs with the resulting compounds led to identify a sulfide, 4a, bearing a Nꢀ1ꢀ
hydroxypiperidineꢀ2,6ꢀdione (HPD) group as new ZBG. 4a is a promising hit compound since it
displayed a nanomolar affinity for MMPꢀ12 with a marked selectivity over MMPꢀ9, MMPꢀ1 and
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2+
MMPꢀ14. Solution complexation studies with Zn were performed to characterize the chelating
abilities of the new compounds and confirmed the bidentate binding mode of HPD derivatives. Xꢀray
crystallography studies using MMPꢀ12 and MMPꢀ9 catalytic domains were carried out to rationalize
the biological results.
1
. Introduction
Matrix metalloproteinases (MMPs), also designated matrixins, constitute a family of more than 20
structurally and functionally related metalloenzymes. Through the hydrolysis of different components
of the extracellular matrix (ECM), these proteinases play a pivotal role in many biological processes,
such as embryonic development, tissue remodeling, wound healing, and angiogenesis. Their expression
is finely regulated at many levels (transcription, activation, inhibition) while unregulated profiles were
found to be associated with pathological conditions. It is for this reason that these enzymes represent a
real interest for researchers as therapeutic targets. In fact, they are involved in the pathogenesis of
several serious and chronic diseases such as osteoarthritis, rheumatoid arthritis, vascular diseases and
1
cancer metastasis. MMPs are zincꢀdependent metalloenzymes. The conserved catalytic domain
2
+
2+
2+
contains a catalytic Zn ion, a structural Zn ion, and one to three structural Ca ions necessary for
2
enzyme stability. In general, a small molecule MMP inhibitor (MMPI) consists of a backbone and a
zincꢀbinding group (ZBG). The backbone is a classic drugꢀlike structure that interacts with the protein
through nonꢀcovalent bonds while the ZBG is a group able to chelate the catalytic zinc ion. Widely
3
utilized ZBGs include: hydroxamic acids, carboxylic acids, pyrimidinetriones, thiols and phosphinates.
ACS Paragon Plus Environment

Page 3 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
Among them, the most used is the hydroxamate group due to its excellent chelating properties. It
2+
coordinates the catalytic Zn in a bidentate fashion to form a complex with distorted trigonal
2
+
bipyramidal geometry. Strong affinity for Zn is often accompanied by poor selectivity. In addition,
the hydroxamate group is characterized by a wellꢀknown hydrolytic susceptibility which often
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
precludes the clinical use of hydroxamateꢀbased MMPIs. Carboxylateꢀbased MMPIs also present some
6
drawbacks related to a lower potency and higher plasma protein binding than their hydroxamate
7
analogues which often cause a reduced bioavailability and in vivo efficacy.
Among all MMPs, gelatinases (MMPꢀ2 and ꢀ9) and metalloelastase (MMPꢀ12) have a fundamental role
in the development of different and relevant pathologies. MMPꢀ2 and MMPꢀ9 are particularly involved
8
in cancer pathogenesis and progression. MMPꢀ12 is mainly produced by macrophages and it is
involved in acute and chronic pulmonary inflammatory diseases associated with intense airway
9
remodeling, such as chronic obstructive pulmonary disease (COPD) and emphysema. Moreover,
1
0
MMPꢀ12 plays an important role in the development of aneurysms and atherosclerosis. Thus, the
development of MMPꢀ12 selective inhibitors could be useful for the treatment of lung inflammatory
and cardiovascular diseases. So far, only a few inhibitors highly selective for MMPꢀ12 have been
1
1
described. We previously reported the design and synthesis of a series of arylsulfone hydroxamates
1
2
and carboxylates as potent MMPꢀ12 inhibitors. Among these, in the present study we have selected to
optimize the 4ꢀmethoxybiphenylsulfonyl derivatives (1b and 2b, Figure 1) with the aim of improving
their selectivity profiles. The compound’s backbone was left unchanged, given its good fit into the
channelꢀlike S1’ pocket of MMPꢀ12, but the ZBG was varied. This strategy is supported by recent
studies that have shown that also the ZBG is important in determining the selectivity of inhibition
13,14
toward a specific MMP.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. Structure of the previously reported MMPꢀ12 inhibitors.
We started with the synthesis of new diaryl sulfone derivatives with ZBGs other than hydroxamic acids
(
compounds 3-8b, Chart 1) to evaluate how the potency and selectivity for MMPꢀ12 is affected by the
ZBG. We then synthesized their sulfide analogues (compounds 1-8a, Chart 1) to understand how the
oxidation state of the sulfur atom impacts on the binding of the inhibitor to MMPꢀ12. The absence of
the oxygen atoms would reduce the ability to establish Hꢀbonds with the enzyme in proximity of the
catalytic zinc, but also remove steric constraints that might prevent the spatial optimization of the
compound in the S1’ pocket. The inhibitory activities of the new thioarylꢀbased MMPꢀ12 inhibitors
were tested in vitro by using recombinant MMPs. Solution equilibrium Zn(II)ꢀcomplexation studies
were performed by potentiometric and spectroscopic (NMR) techniques to characterize the chelating
abilities of the new compounds and Xꢀray crystallography studies were used to rationalize the
biological results.
Compd
ZBG
1
a,b
a,b
2
ACS Paragon Plus Environment

Page 5 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
O
OH
3a,b
COOH
4a,b
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
a,b
a,b
6
7
a,b
a,b
8
Chart 1
2
. Results and discussion.
.1. Design of MMP-12 inhibitors with new ZBGs.
2
With the aim of developing new thioarylꢀbased MMPꢀ12 inhibitors with improved selectivity and
hydrolytic stability with respect to 1b, we selected different functional groups to be used as ZBGs
alternative to the hydroxamic acid. Firstly, a Nꢀ1ꢀhydroxypiperidineꢀ2,6ꢀdione (HPD) moiety was
chosen for derivatives 4a,b based on previous results with compounds containing a similar nonꢀ
classical ZBG (1ꢀhydroxypiperazineꢀ2,6ꢀdione) which were effective inhibitors of MMPs, with
1
5
activities in the nanomolar range. Novel ZBGs, namely hydroxypyridinones (HOPOs) and pyrones
1
6,17
have also been shown to be effective for the design of MMP inhibitors.
Due to its sixꢀmembered
ring closure, the HPD moiety is likely to be more stable and better resistant to hydrolysis than
hydroxamates. In derivatives 3a,b, a dicarboxylic acid was taken in consideration as an alternative
bidentate ZBG, given its promising results in previously described MMPIs that showed a good activity
18
and selectivity for MMPꢀ2 and ꢀ9 over MMPꢀ1. Then, three functional groups considered to be “soft
binders” for the catalytic zinc ion were selected: a Nꢀacylmethanesulfonamide (derivatives 6a,b), a Nꢀ
acyl(trifluoromethane)sulfonamide (derivatives 7a,b) and a methylsulfonylethanone group (derivatives
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
8
a,b). Finally, hydroxamic acid (1a,b), carboxylic acid (2a,b) and phosphonate (5a,b) derivatives were
also prepared for comparison.
.2. Chemistry.
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Diarylsulfide derivatives 1a, 2a, 6a and 7a were prepared as reported in Scheme 1. The previously
12
described, 2ꢀ(2ꢀ(4ꢀbromophenylthio)phenyl)acetic acid 9 , was protected as ethyl ester by treatment
with thionyl chloride (SOCl ) in ethanol. This step was necessary to improve the subsequent crossꢀ
2
coupling reaction, which exhibited poor yield if conducted on carboxylic acid. A Palladiumꢀcatalyzed
crossꢀcoupling reaction (Suzuki conditions) of protected arylbromide 10 with 4ꢀmethoxyphenylboronic
acid afforded biphenyl derivative 11. The key intermediate 11 was converted to its corresponding
hydroxamic acid 1a by direct treatment with hydroxylamine hydrochloride (NH OH HCl) and
2
potassium hydroxide (KOH) in a mixture of methanol/dioxane. Basic hydrolysis of ester 11 by NaOH
in dioxane afforded carboxylic acid 2a, in quantitative yields. Subsequently, 2a was converted into
sulfonamides 6a and 7a by condensation with methanesulfonamide (CH SO NH ) or
3
2
2
trifluoromethansulfonamide
(CF SO NH ),
respectively,
in
the
presence
of
Nꢀ(3ꢀ
3
2
2
dimethylaminopropyl)ꢀN’ꢀethylcarbodiimide (EDC) and 4ꢀdimethylaminopyridine (DMAP) using
dichloromethane as solvent.
ACS Paragon Plus Environment

Page 7 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 1. Synthesis of compounds 1a, 2a, 6a and 7a. Reagents and conditions: i) SOCl , EtOH, 60
2
°C, 4 h, 85%; ii) 4ꢀmethoxyphenylboronic acid, Pd(PPh ) , Na CO , toluene/EtOH, 100 °C, 2 h, 72%;
3
4
2
3
iii) NH OH HCl, KOH, MeOH, 16 h, 88%; iv) NaOH 1N, dioxane, 16 h, quant.; v) CH SO NH , EDC,
2
3
2
2
DMAP, CH Cl , 5 h, 55%; vi) CF SO NH , EDC, DMAP, CH Cl , 5 h, 82%.
2
2
3
2
2
2
2
The synthesis of dicarboxylic acid 3a and Nꢀhydroxypiperidinꢀdione 4a is described in Scheme 2. The
biphenylꢀthioaryl ester 11 was alkylated by Michael addition, using ethyl acrylate and sodium hydride
(NaH) in tꢀBuOH/DMF 4:1, to give diester 12. The dicarboxylic acid 3a was obtained by
saponification with KOH solution of 12 and it was subsequently converted into the corresponding Nꢀ
hydroxypiperidinꢀdione 4a. This step was performed in THF, through a oneꢀpot reaction, where the
carboxylic groups have previously been activated by reacting with ethyl chloroformate, followed by
1
9
treatment with hydroxylamine hydrochloride in basic methanol and THF.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 2. Synthesis of compounds 3a and 4a. Reagents and conditions: i) ethyl acrylate, NaH,
tBuOH/DMF, 4 h, 58%; ii) KOH, H O, 130 °C, 16 h, 88%; iii) ethyl chloroformate, NMM, NH OH
2
2
HCl, KOH, THF/MeOH, 4 h, 40%.
A similar sequence of reactions was followed to synthesize the diarylsulfone derivatives 1-4b, 6b and
7b as detailed in Scheme 3. The hydroxamate 1b and carboxylate 2b were prepared following a route
12
different from the one previously described, which gave improved yields. The ethyl ester 11 was
oxidized to the corresponding sulfone 13, using potassium peroxymonosulfate (Oxone) in a solution of
THF/MeOH/H O. Subsequently, the ethyl ester 13 was hydrolyzed (NaOH 1N/dioxane) to the
2
carboxylic acid 2b which was transformed into the corresponding Nꢀacylmethanesulfonamide 6b and
Nꢀacyl(trifluoromethane)sulfonamide 6b as described above for compounds 6a and 7a (Scheme 1). The
ethyl ester 13 was also converted into the diethyl ester 14 through a Michael addition, using similar
conditions to those previously described for diester 12 (Scheme 2). The diethyl ester 14 was hydrolyzed
(
KOH/H O), to give the dicarboxylic acid 3b. The last compound was firstly transformed into an Oꢀ
2
benzylic cyclic intermediate by reaction with Oꢀbenzylhydroxylamine hydrochloride in the presence of
2
0
triethylamine (TEA), EDC and 1ꢀhydroxybenzotriazole (HOBT). Then, the Oꢀbenzyl protecting
group was removed by treatment with boron trichloride (BCl ) in controlled conditions (30 min at 0 °C
3
ACS Paragon Plus Environment

Page 9 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
under nitrogen atmosphere) to give the Nꢀhydroxypiperidinꢀdione 4b. Finally, hydroxamic acid 1b was
obtained directly from ethyl ester 13 by reaction with hydroxylamine hydrochloride and KOH in a
mixture of methanol/dioxane.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The synthesis of phosphonic acids 5a and 5b is described in the Scheme 4. The known 2ꢀ(4ꢀ
12
bromophenylthio)benzoic acid 15, was protected as ethyl ester 16 and submitted to a palladiumꢀ
catalysed crossꢀcoupling (Suzuki conditions) with 4ꢀmethoxyphenylboronic acid to afford biphenyl
derivative 17. A basic hydrolysis of 17 with NaOH 1N in dioxane provided the carboxylic acid 18,
which was reduced with boraneꢀtetrahydrofuran (BH ꢀTHF 1M) to give benzylic alcohol 19. The
3
chloride derivative 20, obtained by treatment of 19 with SOCl in CH Cl , was transformed into the
2
2
2
phosphonate derivative 21 by direct reaction with triethyl phosphite (P(OEt) ) at 130 °C (Arbuzov
3
reaction). The phosphonic ester 21 was deprotected using trimethylsilyl bromide (TMSBr) in
acetonitrile at 0 °C to give the phosphonic acid 5a. Moreover, the oxidation of 21 with Oxone afforded
the corresponding sulfone derivative 22 which, after acid hydrolysis with HCl 12N under reflux, gave
the sulfone phosphonic acid 5b.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 3. Synthesis of compounds 1-4b and 6b, 7b. Reagents and conditions: i) Oxone, THF/MeOH,
H O, 48 h, quant.; ii) NH OH HCl, KOH, MeOH, 16 h, 50%; iii) NaOH 1N, dioxane, 16 h, 93%; iv)
2
2
ethyl acrylate, NaH, tBuOH/DMF, 4 h, 97%; v) KOH, H O, 130 °C, 16 h, 68%; vi) (1) NH OBn HCl,
2
2
TEA, EDC, HOBT, CH Cl , 72 h, 51%; (2) BCl , CH Cl , 0 °C, 1 h, 73%; vii) CH SO NH , EDC,
2
2
3
2
2
3
2
2
DMAP, CH Cl , 5 h, 56%; viii) CF SO NH , EDC, DMAP, CH Cl , 5 h, 81%.
2
2
3
2
2
2
2
ACS Paragon Plus Environment

Page 11 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 4. Synthesis of compounds 5a and 5b. Reagents and conditions: i) EDC, DMAP, EtOH,
CH Cl , RT, 18h, 82%; ii) 4ꢀmethoxyphenylboronic acid, Pd(PPh ) , Na CO , toluene/EtOH, 100 °C, 5
2
2
3 4
2
3
h, 87%; iii) NaOH 1N, dioxane, RT, 48 h, quant.; iv) BH THF 1M, RT, 26 h, 33%; v) SOCl , CH Cl ,
3
2
2
2
RT, 1 h, 82%; vi) P(OEt) , 130 °C, 6h, 53%; vii) TMSBr, MeCN, 0°C to RT, 12 h, 69%; viii) Oxone,
3
THF/MeOH, H O, RT, 16h, 89%; ix) HCl 12N, reflux, 20 h, 60%.
2
Methylsulfonylethanone derivatives 8a,b were synthesized as reported in Scheme 5. Ethyl ester 17 was
oxidized by treatment with Oxone to the corresponding sulfone 23. Both esters 17 and 23 were
respectively converted into final compounds 8a and 8b by reaction with dimethyl sulfone and
potassium tertꢀbutoxide in DMSO.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
COOEt
COOEt
S
O
2
S
i
SO2
ii
SO2
O
1
7
23
8b
O
O
O
ii
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
S
O
2
S
O
8a
O
Scheme 5. Synthesis of compounds 8a and 8b. Reagents and conditions: i) Oxone, THF/MeOH, H O,
2
RT, 48 h, 78%; ii) dimethyl sulfone, tꢀBuOK, DMSO, 60 °C, 1.5 h, 18ꢀ20%.
2.3. MMPs Inhibition.
The newly synthesized compounds 1-8a and 3-8b were tested on human recombinant MMPꢀ12 by
fluorometric assay to evaluate their inhibitory activity in vitro, in comparison with the previously
reported sulfones 1b and 2b. Initially their selectivity over the other members of the MMP family was
examined by testing them on MMPꢀ2 and MMPꢀ9 (gelatinases), two important proteolytic enzymes
respectively involved in cancer and vascular pathologies. Gelatinases differ from MMPꢀ12 by having
2
1
an “intermediate S1’ pocket”, whereas the latter is classified as a “deep S1’ pocket MMP”. Improved
selectivity for MMPꢀ12 over gelatinases could be indicative of the ability of the new compounds to
22
discriminate among the various MMP family members. However, a recent study conducted on a
common COPD animal model, using longꢀterm cigarette smoke (CS) exposed mice, showed changes in
the expression levels of MMPs in the lungs confirmed by western blot analysis. Expression of both
MMPꢀ12 and MMPꢀ9 were significantly increased in the CS group. Therefore they concluded that
while MMPꢀ12 is an essential mediator of the pathophysiology of COPD, MMPꢀ9 can be considered as
an indicator of the clinical COPD stage.
ACS Paragon Plus Environment

Page 13 of 50
Journal of Medicinal Chemistry
1
2
3
Inhibitory activity data are reported in Table 1. In general, diaryl sulfide derivatives possessed a higher
affinity for MMPꢀ12 than the corresponding sulfones. Moreover, in accordance with previously
reported data, both activity and selectivity for this class of compounds with a common backbone
depend on which ZBG is used. As expected, hydroxamate derivatives possessed the highest activity
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
IC = 4 nM on MMPꢀ12) in both series of compounds but a poor selectivity over the other MMPs
50
tested. The best results in terms of activity and selectivity for MMPꢀ12 were obtained with Nꢀ1ꢀ
hydroxypiperidineꢀ2,6ꢀdione derivatives 4a,b, and in particular with sulfide 4a which presented an IC50
=
33 nM on MMPꢀ12 with a 176ꢀfold selectivity over MMPꢀ9 and 20ꢀfold selectivity over MMPꢀ2.
Diaryl sulfide 3a bearing a dicarboxylic acid as ZBG displayed a similar activity, with an IC = 26 nM
5
0
on MMPꢀ12 and a good selectivity over MMPꢀ9 and MMPꢀ2. All derivatives of both series (S and SO2)
with “softꢀbinders” as the ZBG (6-8a,b) showed reduced activity on all the MMPs tested (IC > 3 µM)
5
0
and the methylsulfonylethanone derivatives 8a,b were totally inactive (IC > 100 µM). Phosphonate
5
0
(
5a,b) and carboxylate (2a,b) derivatives showed a moderate activity (40 < IC < 600 nM) towards
50
MMPꢀ12 with sulfide compounds endowed with a higher potency than sulfone analogues.
a
Table 1. In vitro inhibitory activity (IC , nM) of diaryl sulfides 1-8a and diaryl sulfones
50
1-8b on MMPs.
Compd MMPꢀ2 MMPꢀ9 MMPꢀ12
1
a
6.8
9.7
120
94
4.0
4.8
40
1
b
2
a
114
47
830
2
b
730
140
26
3a
660
1350
670
414
2960
14300
5800
1430
3
b
510
33
4
a
4
b
40
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5
a
250
380
1120
2500
250
600
5b
6
a
5300
3600
46000
2500
18000
3700
3300
3500
2100
7500
6
b
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
a
50000
24500
7
b
8a
230000 332000 230000
>10000 >10000 >10000
8
b
a
Enzymatic data are mean values for three
independent experiments performed in
duplicate. SD were generally within ±10%.
At this point, to better characterize the selectivity profile of the new HPD derivatives 4a,b with respect
to the corresponding hydroxamates 1a,b, their activity was tested also on MMPꢀ1 and MMPꢀ14 (Table
2). The selectivity over these two MMPs is particularly important because their inhibition is considered
responsible for the musculoskeletal side effects observed with broadꢀspectrum MMPI in clinical
23
trials.
Table 2. Selectivity profile of Nꢀ1ꢀhydroxypiperidineꢀ2,6ꢀdiones 4a and 4b in comparison with
hydroxamates 1a and 1b.
a
IC50 (nM)
Compd
MMPꢀ1 MMPꢀ2 MMPꢀ9 MMPꢀ12 MMPꢀ14
4a
260000
>10000
250000
8300
670
414
6.8
9.7
5800
1430
120
94
33
40
40000
4000
2000
710
4
b
1
a
4.0
4.8
1
b
ACS Paragon Plus Environment

Page 15 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
The IC values are the average of three determinations with a
5
0
standard deviation of <10%.
HPD derivatives 4a,b showed a higher selectivity for MMPꢀ12 over MMPꢀ14 with respect to the
corresponding hydroxamates 1a,b. The activity towards MMPꢀ1, already poor for hydroxamates, was
not affected by the change of ZBG. Moreover, in both classes of compounds, sulfide derivatives
presented a better selectivity than sulfones.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
These data suggest that the combined effects of sulfur oxidation state and the nature of the ZBG can
determine the selectivity profile of the tested compounds. In fact, the combination of a sulfide group in
the diaryl moiety and a Nꢀ1ꢀhydroxypiperidineꢀ2,6ꢀdione as new ZBG, led to the identification of a
derivative, 4a, with a greatly improved selectivity for MMPꢀ12 with respect to the hit compound 1b.
2.4. Solution equilibrium studies.
2
+
Since the biological role of MMP inhibitors is largely determined by their interaction with Zn that is
present at the enzyme active site, the chelating ability of the developed compounds towards this metal
ion was evaluated through solution studies. These studies were conducted to determine the stability
constants of the corresponding metal complexes and to establish the coordination modes of the
corresponding metal complexes, as well as their speciation at different pH. For that purpose, twelve
compounds – 1-4 (a,b) and 6-7 (a,b) – were studied through potentiometric titrations and nuclear
1
magnetic resonance of proton spectroscopy ( H NMR). Compound 1b was not studied since it was
completely hydrolyzed to compound 2b, at the chosen working medium, while the phosphonic
derivatives 5a,b were not considered herein since metal phosphonates are already well studied in the
literature and stability constant values are expected to be similar to the ones already obtained for
2
4
phosphoric acid (log βZnHL = 14.14).
.4.1 Acid-base properties
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ꢀ
ꢀ
All the ligands, except 2a, 7a e 7b (L ) or 3b (HL ), were obtained in the neutral form, respectively H L
2
(
3a) and HL (1a, 2b, 4a,b, 6a,b). Due to poor water solubility of some ligands, especially 1a and 4a, a
mixed DMSO/water (60% w/w) medium was used in these studies. In fact, DMSO is a quite common
and well tolerated solvent used in biological and cellular studies (with Cligand < 7 ꢀM), and so the herein
used medium would correspond to a quite low DMSO concentration (< 1%) with concomitant no
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
5
alterations in tissues.
A brief analysis of the protonation constants (Table 1S, Supporting Information) allows to conclude
that the sulfide and sulfonyl groups have negligible effect on log K, because they are too far from the
labile protons. The log K values for the cyclic HPD derivatives 4a,b (9.03ꢀ9.35) are lower than for the
simple hydroxamate derivative 1a (10.29), due to electron delocalization (resonance and inductive
effects due to the extra carbonyl group). Regarding the sulfonamide derivatives, the protonation
constants drop drastically ꢀ 6a,b (log K ~ 5.5) and 7a,b (log K < 2) ꢀ due to electron withdrawing effect
of trifluoromethyl group.
The profiles of the species distribution curves (Figure 1S, Supporting Information) show that under
ꢀ
2ꢀ
physiological conditions (pH 7.4) the predominant species is L or L for all the compounds, except for
the hydroxamate derivatives (1a, 4a and 4b) for which only the neutral HL species is present. The
predominance of anionic species at physiological conditions may be expected to convey some
hydrophilic character.
2.4.2. Studies on Zn(II) complexation
The chelating ability of the compounds towards Zn(II) was evaluated by determining the global
formation constants of the complexes, by pHꢀpotentiometric titrations at 1:1 and 1:2 metalꢀtoꢀligand
molar ratios, and curve fitting with Hyperquad 2008 program. Analysis of the titration curves showed a
drop in the deprotonation profile of the ligand curve due to the presence of the metal ion, such as in
Figure 2 for ligand 7a) and 0 < a < 1, therefore evidencing the formation of metal complexes. Fitting
ACS Paragon Plus Environment

Page 17 of 50
Journal of Medicinal Chemistry
1
2
3
analysis of the overall potentiometric titration data shows that the Zn(II) complex formation occurs
predominantly for pH between 3 and 8, with the presence mainly of 1:1 and 1:2 Zn/L species for all the
compounds except 2a, 6a and 7a,b, for which 1:2 species were not found (Table 1S, Figure 3 and
Figure 2S, Supporting Information).
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1
1
1
3
2
1
0
9
8
7
6
5
4
3
pH
L
Zn/L 1:2
Zn/L 1:1
-
1.5
-1
-0.5
0
0.5
1
1.5
2
2.5
3
a
2
+
Figure 2. Potentiometric titration curves for 7a alone and in the presence of Zn at 1:1 and 1:2 metalꢀ
ꢀ4
toꢀligand molar ratios (C = 6.7 × 10 M in 60% w/w DMSO/water).
L
1
00
100
80
60
40
20
0
8
0
ZnL
Zn
ZnHL
Zn
60
ZnL
ZnH L
-
1
4
0
ZnL2
ZnL2
ZnH-3
ZnH-1
20
0
ZnH-1
3
5
7
2
4
6
pH
pH
a)
b)
2
+
Figure 3. Species distribution curves of two ligandꢀZn systems at 1:2 metalꢀtoꢀligand molar ratios
ꢀ4
(
Zn:L): a) 2b, b) 3b (C = 6.7 × 10 M in 60% w/w DMSO/water).
L
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
Aimed to get some insight on the coordinating groups for each compound, H NMR spectra were
performed for compounds (L) as well as for the respective 1:1 and 1:2 Zn/L systems. The results
obtained for 1a (Figure 3Sa, Supporting Information) and 4a,b predict a (O,O)ꢀhydroxamate
2
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
coordination mode, as found for acetohydroxamic acid (AHA) in M(AHA)2 and for NꢀbenzylꢀN’ꢀ
27
hydroxypiperazineꢀ2,6ꢀdione (BHPD) in M(BHPD) . Regarding the sulfonamides (6a,b and 7a,b)
2
mainly 1:1 ZnL complexes are formed, involving a N,O or most probably a N,O,O coordination to the
metal ion, namely with the amidine nitrogen and a carbonyl and a sulfonyl oxygen in the 6ꢀmember
ring metal coordination mode.
Comparison of the zinc binding affinity of the compounds was based on the pZn values (pH = 6, to
avoid metal hydrolysis) instead of the global stability constants, due to differences in equilibrium
models and acidꢀbase properties. The pZn value is usually defined as the negative logarithm of the
concentration of the free metal ion in solution, for total concentration of metal 1 ꢁM and 10ꢀfold excess
2
8
of ligand, at physiological pH 7.4, but it can be used under other pH conditions.
a
º
Table 3. pZn values obtained for the compounds (T = 25.0 ± 0.1 C, I = 0.1 M KCl, 60% w/w
DMSO/water).
Comp.
1
a
2a
2b
3a
3b
4a
4b
6a
6b
7a
7b
pH
6
.0
6.04 6.04 6.1
6.04 6.14 6.05 6.13 6.10 6.56 6.75 6.91
a
−6
pZn = −log[Zn(II)] with C /C = 10 and C = 10 M
L
Zn
Zn
The obtained pZn values are quite similar (pZn ~ 6), but a more detailed analysis of the data evidenced
some differences (Table 3). For the stoichiometric conditions used in this study, the introduction of the
second carboxylic group (compare 2a, 3a and 2b, 3b) does not improve the stability of the zinc
ACS Paragon Plus Environment

Page 19 of 50
Journal of Medicinal Chemistry
1
2
3
complexes, because the size of its ethylene linker is too long to allow bischelating coordination (7ꢀ
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
member ring). Regarding the sulfonamide derivatives, 6a,b and 7a,b, the respective pZn values (6.1ꢀ
6
.9) suggest an increase of complex stability for trifluoromethyl derivatives.
1
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Overall, based on equilibrium solution and H NMR studies for the zinc complexation with this set of
ligands, equilibrium models and respective coordination modes were proposed, allowing a comparative
analysis of their relative affinity for zinc. In spite of the diversity of functional groups, the pZn values
appear quite similar (pZn ~ 6), except for the trifluoromethane sulfonamide derivatives (pZn ~ 7).
However, it must be stated that the inhibitory capacity of the compounds does not depend only on the
zinc chelating capacity of their binding group, measured in solution, but hangs on many factors, such as
the conformation of the inhibitor and its accommodation and interactions within the enzyme active site
and subsites.
2
.5. Crystallographic studies.
The crystal structures for inhibitors 1b, 2b, 3a,b (Figure 4) were obtained in three different
polymorphs with 1ꢀ4 molecules in the asymmetric unit in complex with MMPꢀ12 and for 3a in
complex with MMPꢀ9 (Supporting Information, Table 3S). The packing is different between the two
polymorphs in space groups P2 2 2 and C2 (respectively, PDB codes: 6ENM and 6EDX), although
1
1 1
the protein construct is the same. In the 6EDX structure, grown in space group C2, which diffracts to
.3 Ǻ resolution, the Nꢀterminal residues from Gly106 onwards are visible. But, for the P2 2 2
1 1
1
1
polymorph, which diffracts to lower resolution (1.59 Ǻ, 6ENM), the Nꢀterminal density is visible only
from His112 onwards. In the C2 polymorph, this sixꢀresidue segment from Gly106 to Lys111 is wellꢀ
ordered and packs against the loops and turns between Pro187 and Tyr262, while in the P2 2 2
1
1 1
polymorph crystal contacts between monomers A and B occupy the same zone. Nonetheless, the
differences in crystal packing between these two polymorphs are far from the zinc sites and do not alter
significantly the modes of ligand binding. All the crystals diffract to near atomic resolution, from 1.6 Å
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
9
to 1.1 Å, as defined by Diederichs et al. . The FoꢀFc Fourier maps contoured at 3σ reveal the ligand
positions in the electron density for all the structures (Figure 4) and no significant changes can be
detected regarding the protein conformation. The 2FoꢀFc Fourier maps contoured at 1σ are also shown
in Figure 4S, Supporting Information. Even if the FoꢀFc density at 3σ is not complete for compound
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
b, our attempts to fit the ligand into other binding modes did not reveal suitable 2FoꢀFc density, and in
each case strong FoꢀFc density appeared in the S1’ pocket. Thus the presence of compound 2b in the
3
0
S1’ binding site is confirmed, even though its LLDF value of 3.31 is slightly higher than the
recommended threshold value 2.0. However, recent discussions point out the unreliability of LLDF as a
3
1
metric and, in our experience, even LLDF values as high as 5, do not exclude the presence of the
ligand.
HPD derivatives 4a and 4b did not crystallize in complex with the MMPꢀ12 catalytic domain at a pH
under which these two compounds are stable. The only crystal structures obtained with 4a and 4b, at
1
4
high pH (9.5ꢀ10), showed degraded ligands.
ACS Paragon Plus Environment

Page 21 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. Fo-Fc maps contoured at 3σ for the inhibitors in complex with MMP-12. The PDB codes
for the structures are shown in red. (A) The electron density for the hydroxamic inhibitor 1b (PDB ID:
6
ENM). (B) The electron density for the carboxylic analog 2b (PDB ID: 6EOX) is much weaker than
3
0
for 1b coherent with an LLDF = 3.13. (C) The electron density for inhibitors 3a (PDB ID: 6ELA) and
(D) 3b are similar (PDB ID: 6EKN). All the inhibitors, in both structures have negative LLDF values.
All of the inhibitors with a sulfonyl functional group superimpose well on each other (Figure 5A),
although for the inhibitor 3b the phenyl ring is slightly moved probably in order to place the ethyleneꢀ
carboxylate chain (a chain of the pentanedioic acid group, where the other carboxylate is the ZBG). A
small shift in the position of the sulfonyl group is dictated by the change of the ZBG from a carboxylic
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
to a hydroxamic chelator (Figure 5B). The superposition between 1b and 3b also displays slightly a
different orientation of the phenyl group in the hydroxamic derivative.
The comparison between the sulfonyl and sulfide derivatives (3b and 3a, Figure 5C) shows that despite
the fact that the biphenyls sit in the S1’ pocket and that the chelating groups are unmoved, the sulfonyl
linker slightly reꢀorients the phenyl moiety and the position of the ethyleneꢀcarboxylate chain. The
sulfonyl linker forms two extra strong Hꢀbonds with the amide nitrogen of Leu181 and the amide
nitrogen of Ala182 of the MMPꢀ12 backbone. As a consequence, the biphenyl group is partially
displaced out of the S1’ pocket, and the phenyl moiety is reꢀorientated so that the ethyleneꢀcarboxylate
group interacts differently with the sidechain of His183 (Figure 5G). For the sulfide derivative this
interaction is mediated by a water molecule (Figure 5E), while for the sulfonyl derivative a direct Hꢀ
bond is made to ND1 of His183. These differences in binding may explain the reduction in potency of
the sulfonyl derivatives (3b) compared to the sulfide derivatives (3a).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The binding mode of the sulfide derivative 3a appears to be common to both MMPꢀ12 and MMPꢀ9
(Figure 5D). The almost perfect superposition of the inhibitor on the two enzymes makes it difficult to
understand which elements might determine the selectivity. In the top portion of the binding site, one
residue distinguishes MMPꢀ12 from MMPꢀ9: at MMPꢀ12 position 180, an Ile residue is present, while
at the equivalent position there is a Leu in MMPꢀ9. The Ile in the MMPꢀ12ꢀ3a complex (PDB: 6ELA)
is clearly defined by a single conformation while the same residue in the MMPꢀ12ꢀ3b crystal structure
(PDB: 6EKN) is found in two alternative conformations. A double conformation for the equivalent Leu
residue is also observed in MMPꢀ9 (PDB: 6ESM). However, the most likely contribution to the
selectivity between MMPꢀ12 and MMPꢀ9 may lie elsewhere. The root mean square derivation (RMSD)
calculated on C of the 152 residues between monomer A of MMPꢀ9 and the four chains of MMPꢀ12
α
gave very similar values ranging from 0.65 Ǻ to 0.75 Ǻ. The comparison between the binding of 3a to
MMPꢀ9 and MMPꢀ12 shows also a difference in positioning of the ligand within the S1’ pocket, with a
ACS Paragon Plus Environment

Page 23 of 50
Journal of Medicinal Chemistry
1
2
3
RMSD value of 0.90 Ǻ between the two ligands (Figure 5D). The most striking difference between the
structures of the two enzymes lies at Arg249 in MMPꢀ9, which forms a direct Hꢀbond to an opposing
main chain carbonyl oxygen of Glu241 (3.2 Ǻ), in the S1’ pocket. In MMPꢀ12, the corresponding
residues are Ala241 and Lys233, respectively (Figure 5S, Supporting Information), and the Ala241 side
chain cannot form a direct Hꢀbond to the carbonyl of Lys233. Other more subtle differences occur in
the stretch of residues between Glu241 and Pro255 (MMPꢀ9), which wraps around the biphenyl
moieties of the ligand and where there are a number of sequence variations (Figure 5S). In the tertiary
coordination sphere of the ligand 3a, there is an insertion of one residue in the MMPꢀ9 segment
between Leu209 and Tyr218, which shifts the loop relative to MMPꢀ12 (Figure 5S). Even though the
S1’ pockets are rather similar between the two enzymes, it is expected that their subtle differences in
sequence and position are likely to give a major contribution to selectivity. Not only is the positioning
of 3a almost identical in MMPꢀ9 and MMPꢀ12, but also the water molecule that mediates the
interaction between the 3a carboxylate and His 183 (MMPꢀ12 numbering) is strictly conserved (Figure
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
E). Given the high resolution of the structures, in particular for MMPꢀ9 where the individual atoms
are clearly defined (Figure 5F), the accuracy in the ligand positioning is irrefutable.
As a result of the different interactions established by the sulfide and sulfonyl derivatives, the two
classes of MMP inhibitors should be treated separately. Comparison of sulfonyl derivatives bound to
32
MMPꢀ12 and MMPꢀ9 (for example 4H82, 5I12 or 4H3X) shows that the binding mode of these
derivatives is conserved. The best superposition between ligands in all these structures is obtained at
the biphenyl moiety rather than at the sulfonyl position, coherent with the role of the ZBG that can alter
the sulfonyl position.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. Comparison of ligand binding. The Hꢀbonds are displayed as red dashed lines. (A)
Superposition of inhibitors 2b and 3b bound to MMPꢀ12. (B) Superposition of inhibitors 1b and 3b
bound to MMPꢀ12. (C) Superposition of inhibitors 3a and 3b bound to MMPꢀ12. (D) Comparison of
the binding of 3a (MMPꢀ12: gold; MMPꢀ9: cyan) to MMPꢀ12 (gray) and MMPꢀ9 (rose). RMSD value
is 0.90 Ǻ between the atoms of the two ligands. (E) Hꢀbonded interactions between 3a and MMPꢀ12.
ACS Paragon Plus Environment

Page 25 of 50
Journal of Medicinal Chemistry
1
2
3
3
3
(
F) Electron density from COOT showing individual atoms around the MMPꢀ9 catalytic zinc for the
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
complex with 3a to highlight the high resolution obtained from the trigonal polymorph of MMPꢀ9. The
FoꢀFc Fourier map is contoured at 3 σ in green and the 2FoꢀFc Fourier map is contoured at 1 σ in blue
mesh. (G) Hꢀbonded interactions between 3b and MMPꢀ12.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
. Conclusions
In the present study, a new series of thioarylꢀbased MMPꢀ12 inhibitors bearing alternative ZBGs was
developed and evaluated. Starting from the promising results previously obtained with two 4ꢀ
methoxybiphenylsulfonyl hydroxamate and carboxylateꢀbased (1b and 2b) inhibitors, we modified the
nature of the ZBG and the sulfur oxidation state with the aim of improving the selectivity for MMPꢀ12.
The new compounds were tested on human MMPs by using a fluorometric assay. In general, diaryl
sulfide derivatives possessed a higher affinity for MMPꢀ12 than the corresponding sulfones. In
particular, the best results in terms of activity and selectivity for MMPꢀ12 were obtained with sulfide
4a, presenting a Nꢀ1ꢀhydroxypiperidineꢀ2,6ꢀdione (HPD) group as new ZBG. This cyclic hydroxamate
showed an IC = 33 nM on MMPꢀ12 with a 1200ꢀfold selectivity over MMPꢀ14, 176ꢀfold selectivity
5
0
over MMPꢀ9 and 20ꢀfold selectivity over MMPꢀ2, thus resulting more selective than the starting
hydroxamate 1b. Diaryl sulfide 3a, bearing a dicarboxylic acid as ZBG, displayed a similar activity,
with an IC50 = 26 nM on MMPꢀ12 and a good selectivity over MMPꢀ9 and MMPꢀ2. Solution
2+
complexation studies with Zn were performed to characterize the chelating abilities of the new
compounds. HPD derivatives 4a,b were confirmed to be bidentate and quite strong Zn(II) chelators,
comparable to the nonꢀcyclic hydroxamate 1a, while 3a coordinates as a monocarboxylic ligand.
Moreover, Xꢀray crystallographic studies performed on the complexes of 3a and 3b with MMPꢀ12 and
3
a with MMPꢀ9 clarified the binding mode in the catalytic domain, but did not fully elucidate the
reasons for the selectivity of these new dicarboxylic derivatives.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
This preliminary study allowed us to identify a hit compound with a new ZBG endowed with a good
selectivity for MMPꢀ12. Further optimization efforts are ongoing with the aim of improving the
hydrophilicity and bioavailability of the new HPD sulfide 4a before testing it in cellꢀbased and in vivo
models of cardiovascular or pulmonary pathologies.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
. Experimental section.
4.1. Chemistry. Melting points were determined on a Kofler hotstage apparatus and are uncorrected.
1
13
19
H, C and F NMR spectra were determined with a Varian Gemini 200 MHz spectrometer or a
Bruker Avance III HD 400 MHz spectrometer. Chemical shifts (δ) are reported in parts per million and
coupling constants (J) are given in hertz (Hz). The following abbreviations are used: singlet (s), doublet
(d), triplet (t), doubleꢀdoublet (dd), doubleꢀtriplet (dt), broad (br) and multiplet (m). Chromatographic
separations were performed on silica gel columns by flash column chromatography (Kieselgel 40,
0
.040–0.063 mm; Merck) or using Isolute Flash Si II cartridge (Biotage). Reactions were followed by
thinꢀlayer chromatography (TLC) on Merck aluminum silica gel (60 F254) sheets that were visualized
under a UV lamp and hydroxamic acids were visualized with FeCl aqueous solution. Evaporation was
3
performed in vacuo (rotating evaporator). Sodium sulfate was always used as the drying agent.
Commercially available chemicals were purchased from SigmaꢀAldrich (Milan, Italy). Elemental
analysis has been used to determine purity of the described compounds, which is >95%. Analytical
results are within ± 0.40% of the theoretical values (Table 2S in Supporting Information).
2
-(2-((4'-Methoxy-[1,1'-biphenyl]-4-yl)thio)phenyl)acetic hydroxamic acid (1a). To a solution of Nꢀ
hydroxylamine hydrochloride (1.882 g, 27.08 mmol) in MeOH (13.21 mL), powdered KOH (1.518 g,
7.08 mmol) was added portionwise at 0 °C. After the addition, the solution was stirred for 1h at RT.
2
The precipitate of KCl was filtered off and the filtrate containing the appropriate free hydroxylamine
was added dropwise to an ice cooled solution of the ethyl ester 11 (50 mg, 0.132 mmol) dissolved in
ACS Paragon Plus Environment

Page 27 of 50
Journal of Medicinal Chemistry
1
2
3
MeOH (2 mL). Subsequently, an additional amount of powered KOH was added (until pH = 9) and the
reaction was stirred at RT overnight. The solvent was evaporated in vacuo, the residue was treated with
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
a saturated solution of NH Cl and extracted with EtOAc (3 x 25mL). The combined organic phases
4
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
were washed with brine, dried over Na SO , filtered and evaporated under reduced pressure.
2
4
Compound 1a was obtained without further purification as a white powder (42 mg; 88% yield). M.p.
1
1
60ꢀ162 °C. H NMR (400 MHz, DMSOꢀd ) δ: 3.55 (s, 2H); 3.79 (s, 3H); 7.00ꢀ7.03 (m, AB system,
6
1
3
2H); 7.22ꢀ7.42 (m, Ar, 6H); 7.58ꢀ7.63 (m, Ar, 4H); 8.85 (s, 1H); 10.64 (s, 1H). C NMR (100 MHz
DMSOꢀd ) δ: 37.28; 55.18; 114.41; 127.09; 127.63; 127.91; 128.05; 130.23; 130.83; 131.55; 133.08;
6
1
34.06; 137.61; 138.36; 159.03; 166.14; 166.55.
1
2
2-(2-((4'-Methoxy-[1,1'-biphenyl]-4-yl)sulfonyl)phenyl)acetic hydroxamic acid (1b) . The title
compound was synthesized as previously reported for the preparation of compound 1a, starting from
ethyl ester 13 (100 mg, 0.2436 mmol). The product was purified by a trituration in Et O collecting 48
2
1
mg of compound 1b (50% yield). M.p. 158 °C. H NMR (400 MHz, DMSOꢀd ) δ: 3.81 (s, 3H); 3.94 (s,
6
2
4
5
H); 7.05ꢀ7.07 (m, AB system, 2H); 7.44ꢀ7.46 (m, Ar, 1H); 7.56ꢀ7.70 (m, Ar, 4H); 7.84ꢀ7.96 (m, Ar,
13
H); 8.10ꢀ8.12 (m, Ar, 1H); 8.87 (s, 1H); 10.62 (s, 1H). C NMR (100 MHz, DMSOꢀd ) δ: 31.16;
6
5.74; 115.06; 127.51; 128.28; 128.55; 128.91; 129.75; 130.80; 134.19; 134.64; 139.33; 139.40;
145.16; 160.38; 171.94.
2
-(2-((4'-Methoxy-[1,1'-biphenyl]-4-yl)thio)phenyl)acetic acid (2a). The ethyl ester 11 (500 mg,
1
.321 mmol) was dissolved in a minimum amount of dioxane. Subsequently, 2 equiv. of NaOH 1N (2.7
mL) were added to the stirred solution. The reaction mixture was stirred at RT overnight. After the
evaporation of dioxane, water was added and the mixture was acidified with HCl 1N until pH = 1. The
white precipitate was filtrated off and washed with water. Compound 2a was obtained as a white
1
powder without any further purification (459 mg; quantitative yield). M.p. 185ꢀ187 °C. H NMR (400
MHz, CD OD) δ: 3.79 (s, 2H); 3.82 (s, 3H); 6.96ꢀ6.99 (m, AB system, 2H); 7.20ꢀ7.24 (m, Ar, 3H); 7.3
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
3
(
dt, J =7.6 Hz, J =1.4 Hz, Ar, 1H); 7.35ꢀ7.39 (m, Ar, 2H); 7.48ꢀ7.54 (m, 2 AB system, 4H). C NMR
1 2
(
100 MHz, DMSOꢀd ) δ: 40.81; 55.19; 114.41; 127.03; 127.64; 128.05; 129.95; 131.43; 131.62;
6
1
33.28; 133.86; 134.50; 138.13; 138.96; 159.00; 172.5.
-(2-((4'-Methoxy-[1,1'-biphenyl]-4-yl)sulfonyl)phenyl)acetic acid (2b) . The title compound was
1
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
synthesized following the procedure described above for the preparation of compound 2a, starting from
ethyl ester 13 (130 mg, 0.316 mmol). After workꢀup, compound 2b was obtained as a white powder
1
without any further purification (92 mg; 93% yield). M.p. 189 °C. H NMR (400 MHz, DMSOꢀd ) δ:
6
3
.81 (s, 3H); 3.95 (s, 2H); 7.06ꢀ7.08 (m, AB system, 2H); 7.45 (dd, Ar, J =7.6 Hz, J =1.2 Hz, 1H); 7.59
1 2
(
dt, J =1.4 Hz, J = 7.8 Hz, Ar, 1H); 7.66ꢀ7.72 (m, Ar, 3H); 7.85ꢀ7.90 (m, Ar, 4H); 8.11 (dd, J = 7.9 Hz,
1
2
1
1
3
J =1.3 Hz, Ar, 1H); 12.34 (s, 1H). C NMR (100 MHz, DMSOꢀd ) δ: 35.71; 55.23; 114.50; 126.91;
2
6
1
1
2
27.66; 127.90; 128.33; 129.17; 130.26; 132.85; 133.47; 134.94; 138.77; 138.93; 144.58; 159.78;
65.73.
-(2-((4'-methoxy-[1,1'-biphenyl]-4-yl)thio)phenyl)pentanedioic acid (3a). A mixture of diethyl
ester 12 (219 mg, 0.486 mmol) and aqueous solution of KOH (272 mg, 4.861 mmol in 7.84 mL water)
was stirred overnight at 130 °C. After being cooled to RT the reaction mixture was acidified with HCl
1
N until pH strongly acidic. The resulting white precipitate was filtered off and washed with water. The
crude product was purified by trituration with nꢀhexane/CH Cl 1:1 to give the final compound 3a as a
2
2
1
white solid (180 mg, 88% yield). M.p. 190ꢀ193 °C. H NMR (400 MHz, DMSOꢀd ) δ: 1.86ꢀ1.90 (m,
6
1
H); 2.1ꢀ2.24 (m, 3H); 3.8 (s, 3H); 4.26 (t, J =7.3 Hz, 1H); 7.00ꢀ7.03 (m, AB system, 2H); 7.28ꢀ7.39
1
1
3
(m, Ar, 6H); 7.58ꢀ7.61 (m, Ar, 4H); 12.27 (s, 2H). C NMR (100 MHz DMSOꢀd ) δ: 28.40; 32.02;
6
4
1
2
7.20; 55.66; 114.88; 127,53; 128.12; 128.55; 128.63; 129.05; 130.96; 132.01; 134.12; 134.52; 134.59;
38.96; 141.48; 159.51; 174.15; 174.53.
-(2-((4'-Methoxy-[1,1'-biphenyl]-4-yl)sulfonyl)phenyl)pentanedioic acid (3b). The title compound
was synthesized following the procedure described for the preparation of compound 3a, starting from
ACS Paragon Plus Environment

Page 29 of 50
Journal of Medicinal Chemistry
1
2
3
diethylester 14 (411 mg, 0.804 mmol). The crude product was purified by flash chromatography
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
petroleum ether 40ꢀ60 °C/ EtOAc in gradient from 3:1 to EtOAc) using an Isolute Flash Si II cartridge
1
to give the final compound 3b as a paleꢀyellow powder (247 mg; 68% yield). M.p. 170°Cꢀ173°C. H
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
NMR (400 MHz, DMSOꢀd ) δ: 1.60ꢀ1.63 (m, 1H); 1.86ꢀ1.94 (m, 1H); 2.01ꢀ2.11 (m, 2H); 3.80 (s, 3H);
6
4
7
.47 (t, J = 7 Hz, 1H); 7.06ꢀ7.06 (m, AB system, 2H); 7.50ꢀ 7.52 (m, Ar, 1H); 7.60ꢀ7.62 (m, Ar, 1H);
1
1
3
.66ꢀ7.72 (m, Ar, 3H); 7.83ꢀ7.90 (m, Ar, 4H); 8.18ꢀ8.20 (m, Ar, 1H); 12.28 (s, 2H). C NMR (DMSOꢀ
d6) δ: 29.38; 32.39; 45.23; 55.76; 115.04; 127.53; 128.39; 128.90; 129.66; 130.47; 130.92; 134.64;
1
39.38; 139.49; 139.65; 145.24; 160.37; 173.72; 173.96.
1
-Hydroxy-3-(2-((4'-methoxy-[1,1'-biphenyl]-4-yl)thio)phenyl)piperidine-2,6-dione (4a). To an iceꢀ
chilled solution of 3a (100 mg, 0.236 mmol) in THF (1 mL) under nitrogen atmosphere, was added
dropwise a solution of ethyl chloroformate (45 ꢁL, 0.473 mmol) and Nꢀmethylmorpholine (0.05 mL,
0
.473 mmol) in dry THF (1 mL). After stirring for 30 min at 0 °C, a precipitate of Nꢀ
methylmorpholinium chloride was formed. In the meanwhile, a solution of hydroxylamine was
prepared by adding of KOH (26.55 mg, 0.473 mmol) to a solution of NH OH HCl (32.88 mg, 0.473
2
mmol) in dry methanol (0.52 mL). The mixture was stirred for 30 min at 0°C, the solid KCl was filtred
off and 0.52 mL of the solution were added dropwise to the first solution containing the activated
dicarboxylic acid. The final mixture was stirred for 4h at 0°C. The residue obtained from solvent
evaporation was dissolved in water and extracted with EtOAc (2x25mL). The organic phase was dried
over Na SO , filtered and brought to dryness. The crude product was purified by reverse phase flash
2
4
chromatography using an Isolute Flash C18 cartridge (in gradient from CH CN/H O 2:1 to 1:1) and
3
2
triturated with Et O and MeOH obtaining 40.3 mg of the final compound 4a as a white solid (40%
2
1
yield). M.p. 108°Cꢀ110°C. H NMR (400 MHz, DMSOꢀd ) δ: 1.90ꢀ1.95 (m, 1H); 2.27ꢀ2.34 (m, 1H);
6
2
.67ꢀ2.74 (m, 1H); 2.90ꢀ2.94 (m, 1H); 3.79 (s, 3H); 4.62 (dd, J =12.2 Hz, J = 5 Hz, 1H); 7.02ꢀ7.03 (m,
1 2
1
3
AB system, 2H); 7.27ꢀ7.44 (m, Ar, 6H); 7.58ꢀ7.61 (m, Ar, 4H); 10.04 (s, 1H). C NMR (100 MHz,
ACS Paragon Plus Environment