Download View Full PDF - Biochemical Society Transactions

Matrix Turnover: Mechanisms and Common Denominators
Matrix metalloproteinases: useful and deleterious
E. Ganea*1 , M. Trifan*, A.C. Laslo*, G. Putina† and C. Cristescu†
*Institute of Biochemistry, Romanian Academy, Bucharest, Romania, and †National Institute for Chemical Pharmaceutical Research and Development,
Bucharest, Romania
Abstract
MMPs (matrix metalloproteinases) are zinc-dependent endopeptidases that degrade both matrix and
non-matrix proteins. They play an important role in morphogenesis, and in a wide range of processes including tissue repair and remodelling. Their abnormal expression contributes to pathological processes including
arthritis, cancer, and cardiac and central nervous system diseases, which explains the large interest in
finding specific MMP inhibitors for therapeutic use. In this review we describe the structural features of
MMPs, with special emphasis on their interaction with specific inhibitors. The effect of new, hydroxamatebased inhibitors on MMP isolated from bovine brain is evaluated.
Introduction
MMPs (matrix metalloproteinases) are zinc-dependent endopeptidases that play a key role in the degradation of ECM
(extracellular matrix). They are extracellular proteins, but
recently it has been reported the presence of some MMPs
also intracellularly [1]; in addition, a number of non-matrix
proteins have been identified as possible MMP substrates [2].
Therefore it has been accepted that they have many other
functions besides ECM degradation, and they are associated
with a variety of normal and pathological conditions. MMPs
play a role in intercellular communication, and can influence
cell migration and tumour progression. Thus studies on the
newest member of the MMP family, epilysin (MMP-28),
demonstrated a role for this enzyme in the regulation of
epithelial cell function and in the induction of carcinogenesis
[3]. According to their structure and substrate specificities
(although the true physiological substrates are not entirely
known yet), the MMP family members are classified into
subgroups, such as: collagenases, gelatinases, stromelysins,
matrilysins, MT-MMPs (membrane-type MMPs) and
‘others’, including seven MMPs not classified in the five
mentioned subgroups [4].
Structural aspects of MMPS
From a structural point of view, most MMPs are organized
into basic, well-conserved domains: an N-terminal signal sequence (‘pre’ domain), followed by an N-terminal propeptide, a catalytic domain and a C-terminal Hpx (haemopexinlike) domain [1,5,6].
The N-terminal propeptide contains a conserved cysteine,
in the ‘cysteine switch’ motif PRCGXPD, which chelates
the catalytic Zn2+ ion, keeping proMMP inactive. All the
MMPs are synthesized as zymogens, and the propeptide has
Key words: bovine brain, endogenous proteinase inhibitor, extracellular matrix, hydroxamate,
matrix metalloproteinase, synthetic inhibitor.
Abbreviations used: CNS, central nervous system; ECM, extracellular matrix; Hpx, haemopexinlike; MMP, matrix metalloproteinase; MMPI, MMP inhibitor; MT-MMP, membrane-type MMP;
TIMP, tissue inhibitor of metalloproteinase; ZBG, zinc-binding group.
1
To whom correspondence should be addressed (email [email protected]).
to be removed during the activation, either by other MMPs
or other proteases.
The catalytic domain contains the zinc-binding motif
HEXXHXXGXXH, where three histidine residues coordinate a zinc ion, and also a conserved methionine residue,
forming a ‘Met-turn’, which contributes to protect the catalytic zinc [1]. Unlike other MMPs, MMP-2 and MMP-9 contain fibronectin type II domains inserted in the middle of the
catalytic domain, which may enhance the substrate binding.
The Hpx domain, present in all MMPs, but MMP7 and
MMP26, is a regulatory subunit, supposed to control MMPs
substrate specificity; a variable hinge region separates this
domain from the catalytic domain. The hinge region also
contributes to the specificity of MMPs, either by direct
binding of the substrate, or by influencing the orientation of
haemopexin and the catalytic domain [7].
This basic structure evolved into different subgroups by
the addition or deletion of structural or functional domains.
For example, incorporation of a hydrophobic peptide of approx. 25 amino acids at the C-terminal end of the propeptide
introduces a furin cleavage site, which is a characteristic
of MT-MMPs, except that MMP-14, -15, -16 and -24 have
transmembrane and cytosolic domains, whereas MMP-17 and
-25 have C-terminal hydrophobic extensions that act as a
glycosylphosphatidylinositol-anchoring signal [6].
The three-dimensional structure of a number of MMPs
has been determined by X-ray crystallography and NMR;
although the domains’ primary structure showed little homology, the polypeptide folds of MMP catalytic domains are
almost superimposable. The catalytic domain consists of
5-stranded β-pleated sheets, three α-helices and connecting
loops, two zinc ions (structural and catalytic) and three
calcium ions which stabilize the structure. The substratebinding site includes the hydrophobic ‘S1 pocket’, variable
in depth, contributing to MMPs substrate specificity [1]. The
propeptide domain consists of three α-chains and connecting
loops; the ‘cysteine switch’ lies in the substrate-binding
pocket, the cysteine sulfhydryl group interacting with the
catalytic zinc ion. In the case of crystallized recombinant
C The
C 2007 Biochemical Society
Authors Journal compilation 689
690
Biochemical Society Transactions (2007) Volume 35, part 4
human proMMP-1 it has been found that the prodomain
interacts with the Hpx domain, resulting in a ‘closed’
configuration, in contrast with the ‘open’ configuration
of the active MMP-1, which suggests the mechanism of
collagen–collagenase interaction [8].
Figure 1 Schematic illustration of hydroxamate inhibitors (a) and
their interaction with MMP (b)
MMPs, friends and enemies
MMPs are proteolytic enzymes that play a key role in the
degradation of ECM and basement membrane, and are
implicated in both physiological and pathological processes.
They are very important for morphogenesis, wound
healing, tissue repair and remodelling in response to injury,
e.g. after myocardial infarction, but also in progression of
diseases such as rheumatoid arthritis, and in tumour growth,
invasion and metastasis. MMPs can enhance leucocyte invasion and regulate the inflammatory process, but enzymes
such as MMP-2 and -9 can favour either anti- or pro-inflammatory action [9].
MMPs are associated with the blood–brain barrier opening
after brain injury, oedema formation and the demyelinating
process, and are generally considered detrimental to the
CNS (central nervous system). However, they may play an
important role in neural plasticity, including remodelling of
synaptic connections. Moreover, it has been demonstrated
recently that at a certain disease phase they may have an
important regenerative role [10].
MMPIs (MMP inhibitors)
Endogenous inhibitors
The in vivo activity of MMPs is controlled both by activation
of the latent pro-enzymes, and by inhibition induced by endogenous inhibitors, such as α 2 -macroglobulin or other protease inhibitors, and by specific TIMPs (tissue inhibitors of
metalloproteinases); currently four TIMPs are known in humans [1]. Unbalanced activities of MMPs and TIMPs are associated with pathological conditions. Based on the multitude of
data resulting from the studies on structures and functions of
MMPs as well as of their inhibitors, a great number of inhibitors have been designed and synthesized, and some of them
were clinically tested. Unfortunately, the therapeutic use of
TIMPs is limited; they are proteins of 21–29 kDa, and besides
their inhibitory activity they possess other biological capacities as well [11]. As a consequence, the therapeutic application
of TIMPs in cancer or cardiovascular disease through gene
therapy or directly, as a protein, is still in an early phase.
Synthetic inhibitors
There is an increasing interest nowadays in designing and
synthesizing MMPIs. Most MMPIs have been identified
by a structure-based approach of inhibitor design, namely
the amino acid sequence of the cleavage site in a collagen
molecule; the presence of a ZBG (zinc-binding group) in
the MMPI molecule is essential for chelating the catalytic
Zn(II) ion. Peptide or peptide-like compounds, most of them
C The
C 2007 Biochemical Society
Authors Journal compilation containing hydroxamic acid derivatives as zinc chelators [12],
as well as other classes of compounds, such as carboxylate,
organoborate and dithiolate, have been identified as MMPIs
[13]. Hydroxamic acid (R-CO-NH-OH) compounds proved
to be stronger inhibitors than corresponding carboxylates
(R-CO-OH), although many carboxylates are effective
inhibitors. In a group of MMP-1 identical inhibitors,
differing only by ZBG, the inhibitory activity decreased from
hydroxamate to carboxylate, with formyl hydroxylamine,
thiol, phosphinate, aminocarboxylate as intermediate [14].
Binding of a hydroxamate group to MMP involves the coordination of the two hydroxamate oxygens to the catalytic
zinc ion and stable hydrogen bonds between hydroxamate
nitrogen and the carbonyl of the enzyme backbone (Figure 1);
van der Waals interactions and hydrophobic contacts stabilize
the inhibitor–enzyme complex.
N-Sulfonyl amino acid hydroxamates have been identified
as MMPIs and some of them are orally available, broadspectrum inhibitors. Recently identified MMPIs are aryl
hydroxamates and the corresponding heterocyclic analogues.
Unfortunately, the progress of these compounds through
the clinical trial is uncertain, because of their mutagenic
properties. Another promising group of inhibitors are
tetracycline derivatives, which not only inhibit MMPs, but
also decrease their production, inhibit activation and increase
their degradation.
Matrix Turnover: Mechanisms and Common Denominators
Figure 2 Gelatin zymogram showing the inhibitory effect of the
hydroxamate compounds C1, C2 and C3 on the gelatinolytic
activity of MMPs from bovine brain separated by Sephadex G-100
chromatography
p1 = peak 1, native enzymes; p1 + C1 (or C2 or C3) = enzymes incubated
with C1 (or C2 or C3).
Conclusions
Considerable progress has been made in understanding the
structure and function of MMPs, but the biological role of
some members of this family is not clear yet.
The role of MMPs in pathology has been extensively
studied and these enzymes are generally found to be
detrimental to the CNS. However, their ‘dual’ nature should
be carefully considered, due to some of their positive effects,
such as degradation of amyloid fibrils by MMP-9.
The development of synthetic inhibitors for therapeutic
use should continue, in spite of the lack of success so far,
provided very complex preclinical studies are carried out
before the clinical trial.
References
In spite of the great number of MMPIs synthesized and
pursued as clinical candidates especially for cancer, arthritis
and cardiovascular disease, most of them have been discontinued (82%), and from the seven compounds remaining in trial,
only one has been approved (Periostat ). The lack of success
could be explained by their low specificity, high toxicity,
or the stage of the disease. The MMPs’ ability to adapt
the binding pocket to the inhibitor shape [13], besides their
intrinsic similarity, could contribute to the lack of specificity.
This is why the identification of new inhibitory compounds is necessary. Our own research programme led to
discovery of heterocyclic hydroxamate-based compounds,
which are benzimidazole derivatives substituted at CS
with various groups, as broad-spectrum MMPIs. These
compounds, denoted C1, C2 and C3, had an inhibitory effect
on all MMP fractions isolated from bovine brain (Figure 2),
the amount of inhibition decreasing in the following order:
C1 > C2 > C3. The selective inhibition by C1 of MMP-2,
the enzyme implicated in CNS pathology [15], suggests the
possible therapeutic application of this compound. Future
studies should prove these findings by testing the inhibitor
effect on both extracellular and cytoplasmic zinc proteinases.
1 Nagase, H., Visse, R. and Murphy, G. (2006) Cardiovasc. Res. 69,
562–573
2 McCawley, L.J. and Matrisian, L.M. (2001) Curr. Opin. Cell Biol. 13,
534–540
3 Illman, S.A., Lehti, K. and Keski-Oja, J. (2006) J. Cell Sci. 119,
3856–3865
4 Visse, R. and Nagase, H. (2003) Circ. Res. 92, 827–839
5 Sternlicht, M.D. and Werb, Z. (2001) Annu. Rev. Cell Dev. Biol. 17,
463–516
6 Massova, I., Kotra, L.P., Fridman, R. and Mobashery, S. (1998) FASEB J.
12, 1075–1095
7 Roeb, E., Schleinhofer, K., Kernebeck, T., Pötsch, S., Jansen, B.,
Behrmann, I., Matern, S. and Grötzinger, J. (2002) J. Biol. Chem. 277,
50326–50332
8 Jozic, D., Bourenkov, G., Lim, N.-H., Visse, R., Nagase, H., Bode, W. and
Maskos, K. (2005) J. Biol. Chem. 280, 9578–9585
9 Le, N.T., Xue, M., Castelnoble, L.A. and Jackson, C.J. (2007) Front. Biosci.
12, 1475–1487
10 Larsen, P.H., Da Silva, A.G., Conant, K. and Yong, V.W. (2006) J. Neurosci.
26, 2207–2214
11 Baker, A.H., Edwards, D.R. and Murphy, G. (2002) J. Cell Sci. 115,
3719–3727
12 Hoekstra, R., Eskens, F.A. and Verweij, J. (2001) The Oncologist 6,
415–427
13 Rosenblum, G., Meroueh, S.O., Kleifeld, O., Brown, S., Singson, P.,
Fridman, R., Mobashery, S. and Sagi, I. (2003) J. Biol. Chem. 278,
27009–27015
14 Whittaker, M. and Ayscough, A. (2001) Celltransmissions 17, 3–14
15 Aoki, T., Kataoka, H., Morimoto, M., Nozaki, K. and Hashimoto, N. (2007)
Stroke 38, 162–169
Received 26 April 2007
C The
C 2007 Biochemical Society
Authors Journal compilation 691