| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
FJ
EXPRESS SUMMARY ARTICLE The Full-length version of this article is also available, published online July 24, 2000 as doi:10.1096/fj.99-0947fje. |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Medizinische Klinik III,
* Institut für Biochemie, 52057 Aachen, Germany
1Correspondence: Med. Klinik III, RWTH Aachen, Pauwelsstrasse 30, 52057 Aachen, Germany. E-mail: Elke.Roeb{at}post.klinikum.rwth-aachen.de
SPECIFIC AIMS
The aim of our study was a specific neutralization of TIMP-1 known to play a crucial role in the pathogenesis of fibrosis.
PRINCIPAL FINDINGS
1. The histidineresidues (H401, H405, H411) in the catalytic domain of MMP-9 were subsequently replaced by alanine. HepG2 cells were transiently transfected with the expression vectors coding for MMP-9, MMP-9-H401A, MMP-9-H405A, MMP-9-H411A, and a double mutant (MMP-9-H401/405A) and serum-free supernatants were analyzed with a polyclonal antibody against MMP-9 by Western blotting. All MMP-9-mutants were expressed and secreted, albeit to different degrees.
2. To answer the question of whether all MMP-9 mutants exhibit gelatinolytic activity, we analyzed supernatants of transfected HepG2 cells by zymography. In contrast to wild-type MMP-9, the mutants (H401A, H405A, and H411A) were inactive toward gelatin substrate.
3. We tested whether catalytically inactive MMP-9 mutants were still
able to bind TIMP-1. To measure complex formation of MMP-9 mutants with
TIMP-1, we used metabolically labeled 35S MMP-9 mutants,
metabolically labeled 35S TIMP-1 from transiently
transfected COS-7 cells, and rat polyclonal antibodies against TIMP-1.
The complexes were precipitated with protein A-Sepharose. No
coprecipitation of mutant MMP-9-H411A was observed (Fig. 1
). Densitometric analysis revealed a binding capacity threefold lower
for MMP-9-H401A and sixfold lower for MMP-9-H405A compared to wild-type
MMP-9 (Fig. 1)
. MMP-9 mutants were present in excess.
|
4. For generation of HepG2, cell lines secreting constitutively large amounts of the MMP-9 mutant H401A HepG2 cells were transfected with the expression vector pcDNA3.1 containing the neomycin resistance gene and the entire coding region for the mutant MMP-9-H401A. Cells were selected in neomycin-containing medium and multiple clones were analyzed by Northern blotting.
5. To demonstrate MMP-9 production by positive clones, we carried out Western blots using polyclonal antibodies against MMP-9. H401A was running with an electrophoretic mobility comparable to recombinant human MMP-9 used as a positive control. Small differences in mobility are attributed to the difference in molecular weight between human and mouse MMP-9. No measurable amounts of MMP-9 could be detected in wild-type HepG2 cells.
6. Recently we showed that HepG2 cells stably transfected with cDNA coding for TIMP-1 do not grow as a monolayer. They grow in nests and lie on top of each other. To test whether MMP-9-H401A is not only able to bind to TIMP-1 but also to neutralize TIMP-1 effects in vivo we investigated the influence of mutant MMP-9 on the growth behavior of HepG2 cells secreting constitutively large amounts of TIMP-1. HepG2 cells stably transfected with mutant MMP-9 grow as a monolayer. Cocultures of HepG2-TIMP-1 cells with HepG2-MMP-9-H401A cells completely abrogate the TIMP-1-associated phenotype. For comparison, cocultures of HepG2-TIMP-1 cells with wild-type HepG2 cells do not affect the typical TIMP-1-associated migration pattern.
7. To determine whether MMP-9-H401A is able to inhibit TIMP-1 in
vitro , a modified MMP-9 activity assay was carried out using
purified components and the serum-free supernatants of
HepG2-MMP-9-H401A cells. To exclude potential side effects of HepG2
cell supernatant on MMP-9 activity, MMP-9 and MMP-9 plus TIMP-1 were
added to serum-free wild-type HepG2 cell supernatant. Incubation of
MMP-9 with TIMP-1 in equimolar amounts reduced total gelatinase
activity by almost 50% (Fig. 2
). In the presence of cell supernatant containing MMP-9-H401A, this loss
of activity was almost completely abrogated, thus demonstrating that
the mutant MMP-9 is able to neutralize the ability of TIMP-1 to inhibit
MMP-9 activity. In addition, reverse zymography revealed that
HepG2-MMP-9-H401A cells do not secrete any detectable amounts of TIMP-1
activity. Wild-type HepG2 cells, however, showed low measurable levels
of TIMP-1. Thus, one could envisage that MMP-9-H401A may be able to
neutralize the endogenously expressed TIMP-1. These data indicate that
MMP-9-H401A, which has no gelatinolytic activity, is still able to bind
TIMP-1. Whether the zinc binding region of MMP-9 is part of a binding
site to TIMP-1 or is necessary for stabilizing other epitopes
responsible for TIMP-1 interaction remains open.
|
CONCLUSIONS AND SIGNIFICANCE
Over the last decade there have been major advances in understanding the cellular and molecular events involved in the pathogenesis of liver fibrosis. Potential therapies in chronic liver diseases would aim at control of hepatic stellate cell activation, neutralization of proliferative and fibrogenic mediators, inhibition of matrix synthesis, and stimulation of matrix degradation. If chronic liver disease has already caused an excessive matrix accumulation, a stimulation of matrix degradation would be of significant importance. Up to now there was no way to applicate proteases exogenously. Stimulation of endogenous proteases was successful in vitro only.
Studies with cultured primary rat hepatocytes, hepatic stellate cells, rat models of liver fibrosis, and plasma levels in patients with chronic liver disease revealed that progressive fibrosis is associated with an increased TIMP-1 expression.
There are many stimulators of TIMP-1 expression, but little is known about the down-regulation of TIMP-1. Since cloning of TIMP-1 in 1986, a reduction of TIMP-1 expression could be demonstrated only by dexamethasone and concanavalin A. Histological findings from patients with liver fibrosis and from animal models of fibrosis indicate that recovery of liver fibrosis with diminution of excess extracellular matrix proteins is possible. An increased collagenase activity during recovery could be detected as a key mechanism in spontaneous resolution of fibrosis in rat liver homogenates. Since fibrotic liver already has an increased proteolytic activity (which, however, is inhibited by high TIMP-1 activity), our therapeutical approach aims at the inactivation of TIMP-1.
We chose MMP-9 for construction of TIMP-1 antagonists due to its high
affinity (Ki values <50 pM) toward TIMP-1. The
exact mechanisms of complex formation between TIMP-1 and MMP-9 are not
yet clear. Concerning binding of TIMP-1 to MMP-1, detailed studies
favor a model of a noncompetitive two-step mechanism. The first step is
thought to be a fast, reversible complex formation, followed by a slow
transformation into the final stable complex. The carboxy terminus of
MMPs plays an important role in the initial fast binding step. The
carboxy-terminal domain of progelatinase B is not involved in autolytic
or cellular activation and does not affect the catalytic activity of
the enzyme. However, carboxy-terminal domain interactions between
active MMP-9 and TIMP-1 significantly enhance the rate of complex
formation (Fig. 3
). After shortening the carboxy terminus of
MMP-9 by 125 amino acids, binding of TIMP-1 is reduced and completely
abolished by truncating the carboxy terminus by 142 residues (Roeb et
al., unpublished results). Moreover this mutant does not show any MMP
activity. Substrate binding is also reduced by that deletion similar as
already described for MMP-2. From these experiments, we conclude that
an effective binding of TIMP-1 is only possible by retaining most of
the hemopexin domain.
|
Antiproliferative and antifibrotic effects on cardiac fibrosiswere associated with elevation of MMP-9 after treatment of cardiac fibroblasts with mimosine, a prolyl 4-hydroxylase inhibitor. In liver cirrhosis in rats, an OK-432 (a biological response modifier) induced increased production of MMP-9 and improved fibrotic processes. Our aim was to mutate MMP-9 in order to delete the aggressive proteolytic activity without disturbing the possibility of tight binding to TIMP-1. All MMP-9-mutants were expressed and secreted after transient transfection of HepG2 cells, as shown. None of the MMP-9-mutants had any proteolytic activity. Thus, the replacement of one amino acid (histidine) by alanine is sufficient to inactivate MMP-9.
We demonstrated that inactivated MMP-9 is still able to bind to TIMP-1
(Fig. 1)
, thus proving a partial independence of catalytic activity and
TIMP-1 binding capacity of MMP-9. However, TIMP-1 binding capacity of
wild-type MMP-9 is about threefold higher than that of MMP-9-H401A. The
structure of MMP-2 that shows the highest homology to MMP-9 within the
family of the MMPs was solved by X-ray crystallography recently.
Comparison of this structure with the structure of MMP-3 in complex
with TIMP-1 revealed that the MMP-2 propeptide binds in a manner very
similar to the active site of MMP-2 as TIMP-1 to MMP-3. In the latter
complex, the histidine that corresponds to H411 in MMP-9 is part of the
MMP-3/TIMP-1 interface and therefore crucial not only for Zn binding,
but also for the MMP/TIMP interaction. Since MMP-9-H411A is not able to
bind TIMP-1 at all, we propose a similar function for H411 in MMP-9.
This structural comparison also rationalizes the graduate effects in
TIMP-1 binding seen with the other two MMP-9 mutants. Whereas the
histidine corresponding to H401 in MMP-9 is mostly buried in the
catalytic domain and has no direct contact to TIMP-1, the histidine
corresponding to H405 is at least partially accessible and is located
at the edge of the MMP-3/TIMP interface.
To study the effects of the mutant MMP-9-H401A in vivo , we transfected HepG2 cells with an expression plasmid coding for this protein. To minimize differences in transfection efficiency, we established cell lines stably transfected with that MMP-9-mutant. Three cell lines were characterized and behaved similarly concerning the secretion of MMP-9-H401A and the growth pattern. These cells showed a similar growth pattern compared with wild-type cells, but their cell–cell contact seems to be reduced. They grow in a monolayer and are widespread, in contrast to almost cubic HepG2 wild-type cells.
In previous reports we have demonstrated an altered phenotype of stably transfected cells engineered to express mouse TIMP-1, a phenomenon that has been shown to be independent of the cell type. In cocultures of stably transfected HepG2-TIMP-1 cells with stably transfected HepG2-MMP-9-H401A cells, the TIMP-1-associated growth behavior is completely abrogated. Neutralization of the TIMP-1-associated phenotype was seen by the HepG2-MMP-9-H401A cells but not by wild-type HepG2 cells. This is the first in vivo effect of the MMP-9 mutant H401A and demonstrates that this mutant is not only able to bind TIMP-1, but also to neutralize TIMP-1 in cell culture. By a modified gelatinase activity assay, we could demonstrate that the mutant MMP-9-H401A is also able to neutralize the ability of TIMP-1 to inhibit MMP activity in an in vitro system. Further experiments are necessary to test whether this mutant is able to inhibit other TIMP-1-associated phenotypes, i.e., fibrotic processes.
In conclusion, we demonstrated a partial independence of catalytic activity and TIMP-1 binding capacity of MMP-9 and showed that the TIMP-1-associated phenotype of liver cells was completely neutralized by an inactivated MMP-9. We anticipate that our findings will influence the design of future antifibrotic strategies in progressive liver fibrosis.
FOOTNOTES
To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.99-0947fje
This article has been cited by other articles:
|
|
 |
|
  D. Cosgrove, D. T. Meehan, D. Delimont, A. Pozzi, X. Chen, K. D. Rodgers, R. M. Tempero, M. Zallocchi, and V. H. Rao Integrin {alpha}1{beta}1 Regulates Matrix Metalloproteinases via P38 Mitogen-Activated Protein Kinase in Mesangial Cells: Implications for Alport Syndrome Am. J. Pathol., March 1, 2008; 172(3): 761 - 773. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Roderfeld, R. Weiskirchen, S. Wagner, M.-L. Berres, C. Henkel, J. Grotzinger, A. M. Gressner, S. Matern, and E. Roeb Inhibition of hepatic fibrogenesis by matrix metalloproteinase-9 mutants in mice FASEB J, March 1, 2006; 20(3): 444 - 454. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Roeb, K. Schleinkofer, T. Kernebeck, S. Potsch, B. Jansen, I. Behrmann, S. Matern, and J. Grotzinger The Matrix Metalloproteinase 9 (MMP-9) Hemopexin Domain Is a Novel Gelatin Binding Domain and Acts as an Antagonist J. Biol. Chem., December 20, 2002; 277(52): 50326 - 50332. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Vaillant, M. G. Chiaramonte, A. W. Cheever, P. D. Soloway, and T. A. Wynn Regulation of Hepatic Fibrosis and Extracellular Matrix Genes by the Th Response: New Insight into the Role of Tissue Inhibitors of Matrix Metalloproteinases J. Immunol., December 15, 2001; 167(12): 7017 - 7026. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
