MMP inhibition by barbiturate homodimers

Article abstract of DOI:10.1016/j.bmcl.2012.11.063

We have studied some homodimeric compounds derived from 5-homopiperazine substituted pyrimidine triones (barbiturates) with linkers in the range 2-20 carbon atoms. The compounds were designed to be capable of resisting absorption, to be stable in the gut and to maintain inhibitory potency against gelatinases and related function. The compounds were then assessed for inhibitory potency against a panel of MMPs (1, 2, 8, 9 and 13). The dimer compounds had similar potency and selectivity to the homopiperazine barbiturate monomer class. At 100 nM, selected dimers significantly inhibited cancer cell invasion in a matrigel assay using Caco-2 cells stimulated by hepatic growth factor. Finally, selected dimers showed adequate stability in simulated intestinal fluid to suggest the capacity to transit to the colon.

Full text of DOI:10.1016/j.bmcl.2012.11.063

Bioorganic & Medicinal Chemistry Letters 23 (2013) 444–447
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
MMP inhibition by barbiturate homodimers
⇑
Jun Wang, Marek W. Radomski, Carlos Medina, John F. Gilmer
School of Pharmacy and Pharmaceutical Sciences, Trinity College, Dublin 2, Ireland
a r t i c l e i n f o
a b s t r a c t
Article history:
We have studied some homodimeric compounds derived from 5-homopiperazine substituted pyrimidine
triones (barbiturates) with linkers in the range 2–20 carbon atoms. The compounds were designed to be
capable of resisting absorption, to be stable in the gut and to maintain inhibitory potency against gela-
tinases and related function.
The compounds were then assessed for inhibitory potency against a panel of MMPs (1, 2, 8, 9 and 13).
The dimer compounds had similar potency and selectivity to the homopiperazine barbiturate monomer
class. At 100 nM, selected dimers significantly inhibited cancer cell invasion in a matrigel assay using
Caco-2 cells stimulated by hepatic growth factor. Finally, selected dimers showed adequate stability in
simulated intestinal fluid to suggest the capacity to transit to the colon.
Received 20 August 2012
Revised 15 November 2012
Accepted 18 November 2012
Available online 29 November 2012
Keywords:
Matrix metalloproteinase inhibitor
Intestinal disease
Dimer
Cancer
Ó 2012 Elsevier Ltd. All rights reserved.
Matrix metalloproteinases (MMPs) are a group of essential zinc
dependent proteolytic enzymes that become dysregulated in sev-
eral disease processes such as inflammatory states and cancer.1–3
The MMPs are long-standing drug development targets with noto-
riously low levels of clinical translation of inhibitors due both to
lack of efficacy and side effects in clinical trials. The failure of suc-
cessive MMP inhibitors is frequently explained by the inadequate
selectivity of test compounds.4,5 Selectivity in this homologous
family is indeed a great challenge, but it is not the only one. There
is increasing evidence that MMP inhibition has the capacity to pro-
voke homeostatic responses within local proteolytic networks that
can oppose the therapeutic goal (this is very well reviewed in Ref.
6). The failure of genetic ablation studies to establish unambiguous
pathophysiological roles for individual MMPs is at least consistent
with this. Specific MMPs may play disease promoting roles in spe-
cific tissue and possibly only at specific times within the disease
life-cycle.6,7
Two MMPs, the so-called gelatinases, MMP-2 and MMP-9, are
associated with inflammatory and neoplastic diseases of the intes-
tinal tract. Thus, levels of MMP-2 and MMP-9 are increased in colo-
rectal tumours compared to normal mucosa.8,9 The latent MMP-2,
but little or no active MMP-2, has been found in healthy mucosa,
but the active MMP-2 is significantly increased in flat-depressed
adenomas with invasiveness capacity and colorectal cancer speci-
mens.10,11 By contrast, up-regulation of pro-MMP-9 is an early
event in the colorectal adenoma-carcinoma sequence.11,12
We reasoned that the disposition of the gelatinases in the
intestinal tract and their putative involvement in intestinal disease
processes suggests that non-absorbable inhibitors might have a
disease modifying effect. The properties of a compound that
promote or inhibit its enteral absorption are well known and a
non-absorbable strategy has been adopted successfully in several
other therapeutic areas where confinement within the GIT is
considered desirable.13
We have recently reported on N-acyl homo-piperazine substi-
tuted pyrmidine triones that have excellent in vitro potency for
the gelatinases.14 Since the barbiturates, along with other Zn
binding inhibitors of the MMPs, bind on the surface of the enzyme,
we hypothesized that the barbiturate homo-dimers might retain
inhibitory properties of the monomers but have reduced cell
penetration due to their mass. In this paper we present preliminary
results showing that the dimeric barbiturates are endowed with a
number of properties important for drug development, that is,
exert good MMP inhibitory activity, they significantly inhibit
colorectal cancer cell invasion (CACO-2) in a model of cancer cell
invasion while they retain adequate stability in intestinal condi-
tions to transit the intestinal tract.
The dimer compounds 2–9 were synthesized by coupling of 1
(available from our previous work) with the appropriate acid
dichlorides at low temperature (ꢀ70 °C)15 (Scheme 1). A further
three dimers (10–12) were prepared by adding dicarboxylic acids
to a mixture of 1, DCC, and DMAP in THF at room temperature
for 24 h. The reaction mixtures were dried in vacuo to remove
the solvent. Test articles were chromatographically pure (HPLC)
and characterized by NMR and MS. In order to examine the effects
of the linker chemistry on inhibitory potency and selectivity a fur-
ther two N-acyl compounds were prepared but without the second
inhibitory unit. The butyrate and myristate monomers 13, 14 were
prepared by treating the amine with the relevant carboxylic acid in
the presence of DCC/DMAP.
⇑
Corresponding author. Tel.: +353 1 896 2795; fax: +353 1 896 2793.
E-mail address: gilmerjf@tcd.ie (J.F. Gilmer).
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2012.11.063

J. Wang et al. / Bioorg. Med. Chem. Lett. 23 (2013) 444–447
445
O
O
n
O
O
O
N
N
N
O
N
O
O
O
2-9
NH
HN
HN
NH
n = 2 2; n = 3 3;
n = 4 4; n = 5 5;
n = 6 6; n = 7 7;
n = 8 8; n = 10 9;
O
i
O
O
O
O
O
O
O
N
N
n
NH
ii
N
O
N
O
N
O
O
O
O
NH
HN
HN
NH
HN
NH
10-12
O
n = 12 10; n = 14 11;
n = 20 12;
iii
iv
1
O
O
O
O
O
O
N
N
N
n
N
O
N
O
O
N
O
O
O
NH
HN
HN
NH
HN
NH
14-15
n = 2 14; n = 12 15;
O
O
13
O
Scheme 1. (i) Acid dichlorides, triethylamine, THF, ꢀ70 °C, 8 h, 20–31%; (ii) DCC, DMAP, diacids, THF, rt, 24 h, 17–27%; (iii) terephthaloyl dichloride triethylamine, THF,
ꢀ70 °C, 8 h, 24%; (iv) butyryic or myrstic acid, DCC, DMAP, THF, 24 h, 25–45%.
The inhibitory effects of the dimer series on MMP-2, -9, -8 and -
13 activity were measured using a fluorogenic assay with human
recombinant enzymes. The compounds were tested at 0.01–
10,000 nM following incubation for 45 min. The compounds did
not exhibit significant activity when screened against MMP-1 at
The most MMP-2 selective compound was 11 (IC50: MMP-
2 = 52 nM, MMP-9 = 222 nM), whereas 10 showed the best MMP-
9 selectivity. The butyric acid monomeric amide 14 was the most
potent compound tested (IC50 21, 13 nM, MMP2/9). Gelatinase
inhibitory potency in the monomeric N-acyl substituted homopip-
erazine class is influenced by the N-acyl substituent.14 The result
for amide 14 is consistent with earlier observations. The gelatinas-
es are intolerant of binding modes requiring placement of rigid
barbiturate substituents at the S20 pocket but they are tolerant of
a variety of flexible side chain groups-highest potency has been
observed in compounds with relatively small P20 substituents. On
the other hand, the longer chain monomer compound 15, a
myristic acid amide was the least potent compound in this study
(IC50 600 nM), being 5–10 fold lower in potency than the dimers
of comparable chain length (9–12).
10 l
M, an expected consequence of their phenoxyphenyl P10
substitution.
All dimers exhibited IC50 values in the nanomolar range. Com-
pounds 2, 6, 7, and 13, resulted in IC50 values <100 nM (Table 1).
Table 1
IC50 values (nM) for inhibition of MMPs (n = 3 over five concentration levels)
Cpd
MMP-2
MMP-9
MMP-8
MMP-13
2
3
4
5
6
7
8
9
10
11
12
13
14
15
57 (48–67)
54 (42–71)
149 (122–182)
82 (66–101)
67 (58–79)
85 (66–108)
188 (153–232)
115 (91–146)
121 (98–149)
80 (70–92)
44 (30–64)
85 (62–117)
60 (42–85)
11 (8.0–16)
164 (117–229)
23 (18–28)
Figure
2 shows a representative conformation following
docking experiments (AutoDock 4) with 3 using the X-ray crystal
structure of a mutant MMP-9 bound to a 5-piperazinyl barbiturate
(PDB code 2OVX). Figure 2 shows the barbiturate in the expected
Zn binding mode with the phenoxyphenyl group penetrating the
deep hydrophobic P10 pocket, the homopiperazine directed into
the P20 pocket, away from the enzyme and the second Zn binding
unit bound on the enzyme surface.
The gelatinases are found in various monomeric and oligomeric
states. MMP-2 binds to TIMP-2, -3, -4 and interacts with glycos-
aminoglycans. The activation of MMP-2 is unique among MMPs
117 (90–151)
88 (62–123)
98 (81–118)
91 (75–111)
125 (80–196)
217 (142–333)
282 (221–361)
162 (129–203)
46 (34–64)
648 (53–78)
155 (121– 200)
44 (36–54)
106 (86–132)
104 (82–130)
70 (38–127)
134(82–129)
225 (159–318)
23 (19–27)
56 (48–65)
69 (54–88)
71 (62–81)
131 (106–162)
120 (82–175)
81 (159–323)
222 (150–330)
144 (103–201)
50 (40–64)
145 (91–231)
227 (159–323)
52 (35–78)
71 (44–115)
54 (40–74)
21 (16–27)
598 (446–805)
13 (9.5–18.6)
608 (408–905)

446
J. Wang et al. / Bioorg. Med. Chem. Lett. 23 (2013) 444–447
80
70
60
50
40
30
20
10
**
**
**
**
**
**
**
0
Figure 1. Inhibition of Caco-2 cell invasion by barbiturate dimer analogs (⁄⁄signif-
icantly different from control P <0.01). Data were analyzed using one-way analysis
of variance (GraphPad Prism 3, San Diego, CA, USA). The results are expressed as
mean SEM of at least three independent experiments.
Figure 2. A picture of 3 docking with a ‘Connolly’ surfaces on MMP-9. The surfaces
were generated by the programme PyMOLÒ based on the crystal structures of MMP-
9 (PDB code 2ovx). Ligand orientations were used which had the lowest binding
energy of the conformations of 50 runs with 25,000,000 energy evaluations in
Autodock4 and displayed by PyMOLÒ. The active site zinc atoms are shown as the
red spheres.
as it needs the formation of the trimeric activator complex proM-
MP-2/TIMP2/MT1-MMP).16 MMP-9 is secreted in monomeric and
stable homodimeric forms. It is the only secreted MMP to form
homodimers but it is relatively promiscuous, forming complexes
with MMP-1, haptoglobulin, CD44, CD91, Fetuin A and proteogly-
can.17 The interaction with CD44 may be especially relevant to cell
mobility.18 The MMP-9 homodimer is formed through interactions
between blade IV of the hemopexin-like domain whereas heterodi-
meric interactions tend to be mediated by blade 1.19 Peptides mim-
icking these regions of MMP-9 suppress cancer cell mobility
suggesting that the oligomeric forms of MMP-9 may be useful to
target in preventing cancer cell invasion.20 Selectivity between
the MMP-9 homodimer form and monomer form could be antici-
pated if an inhibitory dimer could occupy two adjacent MMP-9
sites thereby increasing affinity relative to monomeric forms. The
inhibitor groups would need to be spaced at the correct distance
and; therefore, inhibitor potency/selectivity might emerge as a
function of spacer length. For example, potency and selectivity of
opioid agonist homodimers was found to be a function of spacer
length, leading to new insights into the disposition of G-protein
coupled receptors.21 Homodimerization may also cause inhibitory
groups to bind at exosites in a non-specific fashion that could im-
part favourable binding or selectivity properties. The effect of
homodimerization has since been investigated in numerous areas
of medicinal chemistry including HIV protease inhibitors, cholines-
terase inhibitors, kinase inhibitors and MMP-inhibitors.22–24 Re-
cent studies have reported on the dimer approach to inhibition
of MMP-1, -2, -9, and -14, based on the hydroxamic acid moiety.25
These compounds exhibited useful potency but disappointing
selectivity between MMPs. We also observed little selectivity of
tested inhibitors on the set of MMPs suggesting that the homodi-
mers were not suitably arranged to make connections with sec-
ondary active or exo-sites, which could impart selectivity.6,26
However, purified recombinant enzymes may not be ideal for
assessing homodimer properties and functional potential in this
context because they cannot heterodimerize. Indeed, polyglycosy-
lation, which is absent in recombinant human MMP-9, is required
for its homo-dimerization.17 Therefore, selected dimer compounds
with various spacer lengths were evaluated in a functional assay in
which MMP-9 homo- and heterdimerization is possible and where
the homodimer is believed to play a functional role. The inhibitors
and invasiveness of Caco-2 cells.27 Compound 3 at 10
lM inhibited
70% Caco-2 cell invasion (Fig. 1). Even at 100 nM, compounds 2, 3,
4, 5, 8, and 13 also significantly reduced Caco-2 cell invasion. How-
ever, there was no evidence of a relationship between linker size
and inhibitory effect suggesting that the dimeric compounds acted
at a single site to inhibit invasion.
Finally, in order to assess the suitability of the compounds to
transit the gastrointestinal tract, 3 was incubated at 37 °C in simu-
lated human duodenal fluid and the amount of compound remain-
ing assessed by RP HPLC as a function of time. Greater than 50% of
3 remained at 6 h, the typical oro-caecal transit time.
It has proven extremely difficult to produce clinically useful
compounds that target MMP active sites because of enzyme simi-
larity and polyfunctionality. Simultaneous targeting of active site
and exosites or secondary active sites in oligomeric MMPs is a
promising means of improving selectivity and therapeutic index
since it is more closely aligned with function.6 Due to their mass,
polyvalent inhibitors may also be confined within the intestinal
system from the oral route, another promising means of targeting
specific MMP subpopulations. The distribution, function and acces-
sibility of intestinal MMPs will be important in determining effi-
cacy in this case. The compounds described are suitable for
investigating this question because of their size and stability. They
are being assessed currently using animal models of IBD. Dimeric
compounds have additional potential to selectively inhibit the
functionally important dimeric form of MMP-9, provided that dual
occupancy increases affinity over physiological forms of this and
related enzymes. The present study will hopefully advance efforts
in this area. Homo- and hetero-dimeric compounds with greater
linker chain length and hydrophilicity need to be investigated.
Acknowledgments
This work was supported in part by a Science Foundation of Ire-
land (SFI) Grant (RFP/BMT2781). C.M. is SFI Stokes lecturer.
References and notes
1. Close, D. Ann. Rheum. Dis. 2001, 60, 62.
2. Visse, R.; Nagase, H. Circ. Res. 2003, 92, 827.
3. Verma, R. P.; Hansch, C. Bioorg. Med. Chem. 2007, 15, 2223.
4. Corbitt, C. A.; Lin, J.; Lindsey, M. L. Recent Pat. Anticancer Drug Discov. 2007, 2,
135.
were evaluated in a cancer cell invasion assay at 10
100 nM. Matrigel membranes with 8.0 m pore were used, while
hepatocyte growth factor (HGF) was used to stimulate growth
lM and/or
l

J. Wang et al. / Bioorg. Med. Chem. Lett. 23 (2013) 444–447
447
5. Agrawal, A.; Romero-Perez, D.; Jacobsen, J. A.; Villarreal, F. J.; Cohen, S. M.
ChemMedChem 2008, 3, 812.
6. Sela-Passwell, N.; Rosenblum, G.; Shoham, T.; Sagi, I. Biochim. Biophys. Acta
1803, 29.
7. Markoulli, M.; Papas, E.; Cole, N.; Holden, B. A. Invest Ophthalmol. Vis. Sci. 2012,
53, 1479.
8. Emmert-Buck, M. R.; Roth, M. J.; Zhuang, Z.; Campo, E.; Rozhin, J.; Sloane, B. F.;
Liotta, L. A.; Stetler-Stevenson, W. G. Am. J. Pathol. 1994, 145, 1285.
9. Baker, E. A.; Leaper, D. J. Eur. J. Surg. Oncol. 2002, 28, 24.
10. Parsons, S. L.; Watson, S. A.; Collins, H. M.; Griffin, N. R.; Clarke, P. A.; Steele, R. J.
Br. J. Cancer 1998, 78, 1495.
16. Itoh, Y.; Ito, N.; Nagase, H.; Seiki, M. J. Biol. Chem. 2008, 283, 13053.
17. Malla, N.; Sjoli, S.; Winberg, J. O.; Hadler-Olsen, E.; Uhlin-Hansen, L. Connect.
Tissue Res. 2008, 49, 180.
18. Chetty, C.; Vanamala, S. K.; Gondi, C. S.; Dinh, D. H.; Gujrati, M.; Rao, J. S. Cell
Signal. 2012, 24, 549.
19. Olson, M. W.; Bernardo, M. M.; Pietila, M.; Gervasi, D. C.; Toth, M.; Kotra, L. P.;
Massova, I.; Mobashery, S.; Fridman, R. J. Biol. Chem. 2000, 275, 2661.
20. Dufour, A.; Zucker, S.; Sampson, N. S.; Kuscu, C.; Cao, J. J. Biol. Chem., 285, 35944.
21. Portoghese, P. S. J. Med. Chem. 2001, 44, 2259.
22. Accetta, A.; Corradini, R.; Sforza, S.; Tedeschi, T.; Brognara, E.; Borgatti, M.;
Gambari, R.; Marchelli, R. J. Med. Chem. 2009, 52, 87.
11. Gimeno-García, A. Z.; Santana-Rodríguez, A.; Jiménez, A.; Parra-Blanco, A.;
Nicolás-Pérez, D.; Paz-Cabrera, C.; Díaz-González, F.; Medina, C.; Díaz-Flores,
L.; Quintero, E. Eur. J. Cancer 2006, 42, 3246.
23. Decker, M. J. Med. Chem. 2006, 49, 5411.
24. Zhang, Y.; Gilliam, A.; Maitra, R.; Damaj, M. I.; Tajuba, J. M.; Seltzman, H. H.;
Thomas, B. F. J. Med. Chem. 2010, 53, 7048.
12. Jeziorska, M.; Haboubi, N. Y.; Schofield, P. F.; Ogata, Y.; Nagase, H.; Woolley, D.
E. Int. J. Colorectal. Dis. 1994, 9, 141.
25. Rossello, A.; Nuti, E.; Catalani, M. P.; Carelli, P., et al Bioorg. Med. Chem. Lett.
2005, 15, 2311.
13. Charmot, D. Curr. Pharm. Des. 2012, 18, 1434.
14. Wang, J.; O’Sullivan, S.; Harmon, S.; Keaveny, R.; Radomski, M. W.; Medina, C.;
Gilmer, J. F. J. Med. Chem. 2012, 55, 2154.
15. Wang, J.; Medina, C.; Radomski, M. W.; Gilmer, J. F. Bioorg. Med. Chem. 2011, 19,
4985.
26. Lauer-Fields, J. L.; Whitehead, J. K.; Li, S.; Hammer, R. P.; Brew, K.; Fields, G. B. J.
Biol. Chem. 2008, 283, 20087.
27. Kermorgant, S.; Aparicio, T.; Dessirier, V.; Lewin, M. J.; Lehy, T. Carcinogenesis
2001, 22, 1035.