Recently Synthesized Class of Vinylphosphonates as Potent Matrix Metalloproteinase MMP-2 Inhibitors.

76 Srebnik et al.
Abed Al-Aziz Al-Quntara,
Ofir Bauma,
Reuven Reichb,*
Morris Srebnika*
Arch. Pharm. Pharm. Med. Chem. 2004, 337, 76−80
Recently Synthesized Class of
Vinylphosphonates as Potent Matrix
Metalloproteinase (MMP-2) Inhibitors
a
Department of Medicinal
Chemistry and Natural
Products, Hebrew University
in Jerusalem, Jerusalem,
Israel
b
Department of Pharmacology,
School of Pharmacy, Hebrew
University in Jerusalem,
Jerusalem, Israel
An in vitro evaluation of MMP-2 inhibitors for a series of novel vinylphosphonic
acids and phosphonic esters that contain various functional groups, shows that
various types exhibit excellent efficiency, and points towards potent, promising
compounds. Other types displayed relatively weak activity.
Keywords: Vinylphosphonates; Phosphonates; Matrix Metallo-Proteinase;
Structure⫺activity relationships; Cancer
Received: July 18, 2003; Accepted: September 9, 2003 [FP828]
DOI 10.1002/ardp.200300828
Introduction
Matrix metalloproteinases (MMPs) are a family of
structurally related enzymes. In addition to the presence of zinc at their active site, they contain calcium
for enzyme stability. MMPs, mainly MMP-2 which is a
type IV collagenase (gelatinase) play an important role
in the degradation of extracellular matrix proteins that
constitute cellular connective tissue and are strongly
involved in both normal and pathological tissue remodeling [1⫺5]. The degradative activity of MMPs is
tightly controlled both by the latency of the secreted
enzymes as well as by the presence of naturally occurring inhibitors including general plasma proteinase inhibitors and tissue inhibitors of metalloproteinases. Imbalance between the levels of activated enzymes and
their inhibitors causes a breakdown of the extracellular
matrix [6, 7]. It has been shown that MMPs play a role
in primary tumor growth [8]. In addition, they mediate
invasion and metastasis, which are the processes that
lethally spread cancer cells through the body [9, 10].
First, cancer cells detach from the primary site (which
is often in an epithelial tissue) and breach the basement membrane separating them from other tissue
layers [6,11]. Some of these invasive cells can penetrate the basement membrane surrounding a blood
vessel, as well as the layer of endothelial cells lining
it. The cells are then free to circulate via the bloodstream. Eventually, a cancer cell may lodge in a capillary. If it then adheres to, and penetrates the capillary
wall again, it can create a secondary tumor which may
survive to colonize another tissue.
The mechanism for the degradation of the proteins is
based on the scissile amide carbonyl coordination to
Correspondence: Morris Srebnik, Department of Medicinal
Chemistry and Natural Products, Hebrew University in
Jerusalem, POB 12065, Jerusalem 91120, Israel. Phone:
+972 2 675-7301; Fax: +972 2 675-8201; e-mail:
msrebni@md2.huji.ac.il
the active site zinc(II) ion. Simultaneously, it is attacked by a water molecule that is both hydrogen
bonded to glutamic acid and coordinated to the zinc(II)
ion. Then the water donates a proton to the scissile
amide. Finally, the peptide bond is cleaved after the
remaining proton from the water molecule is shuttled
to the nitrogen of the scissile amide [12]. Enhanced
gelatinase activity is associated with a higher tumor
grade in breast [13], bladder [14], and gastric cancers
[15]. It also correlates with many types of inflammations [16], including arthritis [17], restenosis [18],
multiple sclerosis [19], and others [20, 21]. In order to
maintain the balance of MMPs in these pathological
processes, native inhibitors, such as TIMP-1 [22] and
TIMP-2 [23], have been considered for therapeutic
aims. Synthetic inhibitors have been developed, such
as succinyl hydroxamates; these compounds are accompanied by various problems, such as low water
solubility and toxicity [24], carboxylic acids [15], thiols
[25], and more favorable phosphorous containing
compounds [26].
The requirement for a molecule to be an effective inhibitor of MMPs is a functional group capable of chelating the active-site zinc(II) ion. In addition, functional
groups which provide a hydrogen bond interaction with
the enzyme, and side chains which undergo effective
van der Waals-interactions, are required [12].
In this investigation, new classes of vinylphosphonic
acids that have been recently prepared in our lab have
been tested for the first time as MMP-2 inhibitors.
Chemistry
cis-Vinylphosphonate 1 was prepared by addition of 1hexynylphosphonate to the Negishi reagent, followed
* M. S.a and R. R.b are affiliated with the Bloom Center for
Pharmacy at the Hebrew University.
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Arch. Pharm. Pharm. Med. Chem. 2004, 337, 76−80
Vinylphosphonates as MMP-2 inhibitors 77
Scheme 3.
Scheme 1.
The 3-aminovinylphosphonates, compounds 24⫺29,
were synthesized by addition of imines to the alkynylphosphonate titanium(II) complexes (Scheme 3) [32].
by hydrolysis of the ester with Me3SiBr in CH2Cl2 [27].
Compounds 2⫺7 2-(hydroxymethyl)vinylphosphonates were synthesized by the insertion of cyclic ketones into three membered ring zirconacycles
(Scheme 1). The free phosphonic acid compounds
5⫺7 were obtained by the hydrolysis of the phosphonate esters by Me3SiBr in CH2Cl2. The other types of
2-(hydroxymethyl)vinylphosphonic acids 8⫺14 were
obtained by insertion of aldehydes or ketones into the
zirconacycle (Scheme 1) and, similarly, were hydrolyzed [28, 29]. The 3-oxovinylphosphonic acids
15⫺18 were prepared by insertion of acid chlorides
into zirconacycles (Scheme 1), followed by similar hydrolysis with Me3SiBr [30].
The preparation of di- and tri-substituted vinylphosphonic acids 19⫺23 was accomplished by the tuning
of the addition of acid chlorides and Grignards to the
three membered ring titanacycle (Scheme 2) followed
by the usual hydrolysis to yield the free acid [31].
Biology
Cell culture
Human HT-1080 fibro sarcoma cells from ATCC were
maintained in Minimal Essential Medium, supplemented with calf serum, 10%. Glutamine, pyruvate,
nonessential amino acids, vitamins, and antibiotics
(Biological Industries, Kibbutz Beth HaEmek, Israel)
are added as additional supplements.
Basement membrane invasion
Boyden chamber chemo invasion assays were performed [10]. This assay is completely dependent on
the expression of MMP by the cells. Matrigel (25 µg)
was dried on a polycarbonated filter (PVP free, Nucleopore, Whatman). Fibroblast conditioned medium (obtained from confluent NIH-3T3 cells cultured in serum
free DMEM) was used as the chemo attractant. Cells
were harvested by brief exposure to 1 mM EDTA,
washed with DMEM containing 0.1% bovine serum
albumin, and added to the Boyden chamber (200,000
cells). The chamber was incubated in a humidified incubator at 37 sC in 5% CO2/95% air atmosphere for
6 h. The cells which traversed the Matrigel layer and
attached to the lower surface of the filter were stained
with Diff Quick (American Scientific Products, AHS del
Caribe, Inc., Aguada, Puerto Rico) and counted.
Chemotaxis
Scheme 2.
Chemotaxis evaluations were performed in a similar
way to basement membrane invasion, with the exception that the filters were coated with 5 µg collagen IV
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
78 Srebnik et al.
Arch. Pharm. Pharm. Med. Chem. 2004, 337, 76−80
led us to a classification of these compounds into different categories. First, cis-vinylphosphonic acid 1
showed relatively weak activity as a MMP-2 inhibitor
(Figure 1).
Second, unlike compounds 2, 3, 4 which did not
posses any satisfactory activity as the esters, compounds 5⫺7 showed very promising activity. Specifically, compound 5 was the most active in this series
(Figure 2).
Figure 1. MMP-2 inhibitors of compounds 1⫺4. Error
bars represent the standard deviation; the experiment
was repeated three times.
5 showed enhanced activity even at low micromolar
range. This is supported by results from the literature
which showed that the introduction of a cyclohexyl
substituent in the carboxylic acid compounds improved the inhibition significantly [33]. The activity of
compounds 8⫺14 varied from moderate to interesting.
It can also be observed that the presence of phenyl
groups in compounds 8, 13, 14 does not improve inhibition (Figure 3) [34].
Figure 2. MMP-2 inhibition of compounds 5⫺7. Error
bars represent the standard deviation; the experiment
was repeated three times.
instead of Matrigel. This amount of collagen does not
form a barrier to the migrating cells, but rather an attachment substratum.
Results and discussion
Various new types of vinylphosphonates and vinylphosphonic acids, compounds 1⫺29, were tested in vitro
for inhibition of MMP-2 protease. A study of the results
Figure 3. MMP-2 inhibitors of compounds 8⫺14. Error
bars represent the standard deviation; the experiment
was repeated three times.
On the other hand, increasing the chain length slightly
increased the inhibition, as seen for compound 12
[35]. The introduction of a carbonyl group together with
the phosphonic acid dramatically enhanced the inhi-
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Arch. Pharm. Pharm. Med. Chem. 2004, 337, 76−80
Vinylphosphonates as MMP-2 inhibitors 79
olar range, particularly compounds 27, 28, and 29
(Figure 6).
Figure 4. MMP-2 inhibitors of compounds 15⫺18.
Error bars represent the standard deviation; the
experiment was repeated three times.
Figure 6. MMP-2 inhibition of compounds 24⫺29.
Error bars represent the standard deviation; the
experiment was repeated three times.
Conclusions
Figure 5. MMP-2 inhibition of compounds 19⫺23,
(error bars represent the std. deviation, and the experiment was repeated three times).
bition of MMP-2 protease, 15⫺18, particularly compounds 15, 16, (Figure 4).
This is consistent with the mechanism stating that the
active site of zinc(II) first coordinates to the carbonyl
group of the scissile amide [12].
The di- and tri-substituted phosphonic acid compounds, 19⫺23, are very potent as MMP-2 protease
inhibitors (Figure 5).
As phosphonic esters, the 3-aminovinylphosphonates,
24⫺29, are among the most potent and effective
MMP-2 protease inhibitors, even at a very low microm-
Various new classes of substituted vinylphosphonates
were recently synthesized in our lab and have been
investigated as MMP-2 inhibitors. Cyclic hydroxy-vinylphosphonates as the free phosphonic acids have
shown very potent activity compared to the (cyclic hydroxyl)vinylphosphonate esters which exhibited relatively weak activity. (Hydroxymethyl)vinylphosphonic
acids showed moderate inhibiton. Vinylphosphonic
acid containing carbonyl groups and various other diand tri-substituted vinylphosphonic acids were very
potent compounds. The 3-aminovinylphosphonates
are among the most potent and effective compounds
as MMP-2 protease inhibitors as phosphonic esters,
even at low micromolar concentrations. Hopefully,
these results will revive the field of MMP-2 inhibition.
Acknowledgments
We thank the Israeli Science Foundation, the Middle
East Cancer Consortium, and the Horowitz Foundation for support of this work.
References
[1] A. Powell, L. Matrisian, Curr. Top. Microbiol. Immunol.
1996, 213, 1.
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
80 Srebnik et al.
[2] P. Brown,Med. Oncol. 1997, 14, 1.
[3] A. Yu, R. Hewitt, W. Stevenson, Drugs Aging 1997, 11,
229.
[4] I. Collier, S. Wilhem, A. Eisen, B. Marmer, G. Goldberg,
J. Biol. Chem. 1988, 263, 6579.
[5] S. Wilhem, I. Collier, B. Marmer, A. Eisen, G. Grant, G.
Goldberg, J. Biol. Chem. 1989, 264, 17213.
[6] L. Denis, J. Verweij, Invest. New Drugs 1997, 15, 175.
[7] I. Pergament, R. Reich, M. Srebnik, Bio. Med. Chem.
Lett. 2002, 12, 1215.
[8] A. Chambers, L. Matrisan, J. Nat. Cancer Inst. 1997,
89, 1260.
[9] H. Kawamata, K. Kameyama, Y. Tanaka, L. Nan, W.
Barch, W. Stetler-Stevenson, R. Oyasu, Int. J. Cancer
1995, 63, 568.
[10] R. Reich, E. Thompson, Y. Iwamoto, G. Martin, J. Deason, G. Fuller, R. Miskin, Cancer Res. 1988, 48, 33073.
[11] B. Summers, S. Davidson, Ann. Rep. Med. Chem. 1998,
33, 131.
[12] M. Whittaker, F. Floyd, A. Gearing, Chem. Rev. 1999,
99, 2735.
[13] B. Davies, D. Miles, L. Happerfield, M. Naylor, L. Bobroe,
R. Rubens, F. Balkwill, J. Cancer 1993, 67, 1126.
[14] B. Davis, J. Waxman, J. Williams, A. Thomas, F. Balkwill,
Cancer Res. 1993, 53, 5365.
[15] A. Errico, S. Garbisa, L. Liotta, W. Castronovo, W.
Stetler-Stevenson, W. Grgioni, Mod. Pathol. 1991, 4,
239.
[16] K. Chapman, I. Kupka, P. Durette, C. Esser, T. Lansza,
M. Izquierdo-Martin, L. Niedzwiecki, B. Chang, R. Harrison, W. Hagman, Med. Chem. 1993, 36, 4293.
[17] T. Cawston, Pharm. Ther. 1996, 70, 163.
[18] M. Bendeck, C. Irvin, M. Reidy, Circ. Res. 1996, 78, 38.
[19] K. Gijbels, R. Galary, L. Stienman, Clin. Invest. 1994,
94, 2177.
Arch. Pharm. Pharm. Med. Chem. 2004, 337, 76−80
[20] G. Wallace, M. Stansford, R. Whiston, J. Immunology
1996, 89, 53.
[21] Y. Morimoto, K. Nishkawa, M. Ohashi, Life Sci. 1997,
61, 795.
[22] A. Docherty, A. Lyons, B. Smith, E. Wright, P. Stephens,
T. Harris, Nature, 1985, 318, 66.
[23] W. Stetler-Stevenson, H. Krurzch, L. Liotta, J. Biol.
Chem. 1989b, 264, 17374.
[24] W. Johnson, N. Roberts, N. Borkakoti,J. Enz. Inhib.
1987, 2, 1.
[25] K. Daralak, R. Miller, M. Spak, A. Spatola, R. Gray, J.
Biol. Chem. 1990, 265, 5199.
[26] K. Mookhtiar, C. Marlowe, P. Bartlett, Van Wart, H. Biochemistry, 1987, 26, 962.
[27] A. Quntar, M. Srebnik, Organic Lett. 2001, 3, 1379.
[28] A. Quntar, M. Srebnik, J. Org. Chem. 2001, 66,
6650⫺6653.
[29] A. Quntar, A. Melman, M. Srebnik, Synlett, 2002, 1,
61⫺64.
[30] A. Quntar, A. Melman, M. Srebnik, J. Org. Chem. 2002,
67, 3769⫺3772.
[31] A. Quntar, M. Srebnik Chem. Comm. 2003, 59, 58.
[32] A. Quntar, V. Dembitsky, M. Srebnik, Organic Lett. 2003,
5, 357.
[33] C. Fink, J. Carlson, C. Boehm, P. McTaggart, Y. Qiao, J.
Doughy, V. Ganu, R. Melton, R. Goldberg, Bioorg. Med.
Chem. Lett. 1999, 9, 195.
[34] S. Sahoo, C. Caldwell, K. Chapman, P. Durette, C.
Esser, I. Kopka, S. Polo, K. Sperow, L. Niedzwiecki, B.
Chang, R. Harrison, R. Stien, M. MacCoss, W. Hagmann, Bioorg. Med. Chem. Lett. 1995, 5, 2441. (They
found a similar effect in the carboxylic acid compounds.
When the phenyl groups are replaced by linear alkyl
chains, the MMP-2 inhibition improves.)
[35] This is consistent with the other studies. When the chain
length of phenyl substituents increases, the inhibition improves. (see reference 33)
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim