Content uploaded by Jonathan B Weitzman
Author content
All content in this area was uploaded by Jonathan B Weitzman
Content may be subject to copyright.
INTRODUCTION
Angiogenesis plays a fundamental role in many physiological
and pathological processes including wound healing and tumor
growth. The formation of new vessels, which involves the
migration of stimulated endothelial cells and subsequent tube
formation, depends on a tightly controlled proteolysis of the
components of the extracellular matrix. The secretion of matrix
metalloproteinase (MMP) activity was suggested to constitute
an essential step in the angiogenic process since the MMP
inhibitors TIMP-1 and TIMP-2 have been shown to inhibit
angiogenesis in both in vitro and in vivo model systems
(Johnson et al., 1994; Murphy et al., 1993), as have synthetic
inhibitors of metalloproteinases activity (Galardy et al., 1994;
Taraboletti et al., 1995). However, since these inhibitors are not
specific for any particular MMP and can inhibit all members
of the MMP family, these studies could not determine which
MMP may be involved.
The two mammalian gelatinases (gelatinase A/MMP-2 and
gelatinase B/MMP-9) are members of the MMP family that
specifically degrade gelatin, type-IV, type-V and type-XI
collagen (Matrisian, 1990; Murphy et al., 1991). It is assumed
that the secretion of gelatinases having specificity for type IV
collagen would endow endothelial cells with an advantage for
degradation of the extracellular matrix (ECM) and subsequent
migration across the basement membrane. The gelatinases can
also cleave a variety of non-ECM molecules such as the basic
fibroblast growth factor (bFGF) receptor, galectin-3, IL-1β,
substance P, myelin basic protein and amyloid β peptide (Levi
et al., 1996; Ochieng et al., 1994; Vu and Werb, 1998).
Although MMP-9 can be induced in most cells in culture by
specific cytokines and by tumor promoters such as PMA, its
expression in vivo is mainly restricted to inflammatory cells
and to pathological states such as inflammatory arthritis, tumor
invasion, corneal ulcers and Alzheimer’s disease (Vu and Werb,
1998). MMP-9 is implicated in the invasive behaviour of
metastatic tumor cells and trophoblasts (Alexander et al., 1996;
Bernhard et al., 1994; Bischof et al., 1995; Himelstein et al.,
1994; Stetler-Stevenson, 1990). Furthermore, a role for the
gelatinases in angiogenesis has recently been suggested by
studies of genetically modified knockout mice (Itoh et al.,
1998; Vu et al., 1998). The delayed skeletal growth plate
vascularization phenotype of mutant animals implicated MMP-
9 in the release of angiogenic activators (Vu et al., 1998) and
host MMP-2 in tumor progression (Itoh et al., 1998).
Tumor progression and the angiogenic switch are regulated
1283
Journal of Cell Science 112, 1283-1290 (1999)
Printed in Great Britain © The Company of Biologists Limited 1999
JCS0219
Angiogenesis and the formation of new blood vessels
requires coordinated regulation of matrix proteolysis and
endothelial cell migration. Cellular proteolytic capacity is
the balance between secreted matrix metalloproteinases
(MMP) and their inhibitors (TIMPs). We have examined
the regulation of the gelatinase/TIMP balance by
transforming growth factor-β1 (TGF-β1) and phorbol
myristate acetate (PMA) in bovine endothelial cells. The
low constitutive expression of gelatinase A/MMP-2 was
upregulated by TGF-β1 in a dose-dependent manner.
Gelatinase B/MMP-9 was only detected upon treatment
with either PMA or TGF-β1. However, addition of both
factors together revealed a striking synergistic effect
causing upregulation of MMP-9 and downregulation of
TIMPs, thereby increasing the net MMP-9/TIMP balance
and the gelatinolytic capacity. These effects were observed
at both the protein and mRNA levels. We demonstrate that
changes in different members of the Jun oncogene family
with distinct transactivation properties may account for
this synergistic effect. We investigated the contribution of
these changes in gelatinolytic balance to endothelial cell
migration and invasion. The endothelial cells showed
increased cell motility in response to PMA, but the addition
of TGF-β1 had an inhibitory effect. Hence, regulation of
the MMP-9/TIMP balance failed to correlate with the
migratory or invasive capacity. These results question a
direct role for MMP-9 in endothelial cell motility and
suggest that gelatinases may contribute in alternative ways
to the angiogenic process.
Key words: Metalloproteinase, Endothelial cell, Angiogenesis,
Migration
SUMMARY
Examining the relationship between the gelatinolytic balance and the invasive
capacity of endothelial cells
Alain Puyraimond
1,
*, Jonathan B. Weitzman
2,
*, Emeline Babiole
1
and Suzanne Menashi
1,‡
1
U353 INSERM, Hôpital Saint Louis, 75010 Paris, France
2
Unité des Virus Oncogènes, URA1644 CNRS, Institut Pasteur, 75724 Paris, France
*These authors contributed equally to this work
‡
Author for correspondence (e-mail: s.menashi@chu-stlouis.fr)
Accepted 19 February; published on WWW 8 April 1999
1284
by a large range of tumor-secreted factors. Many of these
molecules have been shown to have diverse effects on cell
proliferation, migration and proteinase secretion. Transforming
growth factor-β1 (TGF-β1) is a multi-functional cytokine
which has both positive and negative effects on endothelial cell
functions (Madri et al., 1988; Pepper et al., 1990, 1993; Yang
and Moses, 1990). TGF-β1 can promote angiogenesis in vivo,
but is a potent inhibitor of endothelial cell proliferation and
migration in vitro. Furthermore, TGF-β1 is known to regulate
ECM by stimulating the synthesis of its components
(collagens, fibronectin, tenascin and proteoglycans), by
inhibiting matrix degradation through the down-regulation of
proteinases such as plasminogen activators, collagenase-1 and
stromelysin-1, and by up-regulating proteinase inhibitors such
as PAI-1 and TIMP-1 (Kerr et al., 1990; Matrisian et al., 1992;
Mauviel et al., 1993). However, TGF-β1 has been shown to
upregulate both the 72 kDa and 92 kDa gelatinases (MMP-2
and MMP-9) in cultured fibroblasts (Overall et al., 1991),
keratinocytes (Salo et al., 1991), and in several tumor cell lines
(Shimizu et al., 1996; Welch et al., 1990). The tumor promoter
phorbol myristate acetate (PMA) has been shown to induce
MMP secretion by a large number of mesenchymal cell and
invasive tumor cell lines (Crawford and Matrisian, 1996). PMA
is also capable of inducing cell invasion and tube formation of
both microvascular and large vessel endothelial cells in vitro
(Montesano and Orci, 1985; Montesano and Orci, 1987).
Although the association between MMPs and tumor
progression is well established, few studies have addressed
their involvement in the migration of endothelial cells which
is essential for neo-vascularization. Furthermore, the effects of
several angiogenic regulators on MMP induction has been
studied in tumor cell lines, but the mechanisms of MMP
regulation in endothelial cells and their role in angiogenesis are
still unclear. To address these issues, we have studied the
regulation of MMP-2, MMP-9, TIMP-1 and TIMP-2 secretion
in bovine endothelial cells treated with TGF-β1 and PMA,
alone or in combination. The consequences of the regulation
of the gelatinolytic potential on the migratory and invasive
capacity of endothelial cells were also investigated.
MATERIALS AND METHODS
Cell culture
Calf pulmonary artery endothelial cells (CPAE) were kindly provided
by Dr J. Badet (Laboratoire de Biotechnologie des Cellules
Eucaryotes, University de Créteil, France) and were grown in
minimum essential medium (MEM) containing 20% fetal calf serum
(FCS) and used between passages 12 and 20. When cells reached
approximately 80% confluency, they were washed and incubated in
the presence of TGF-β1 (R&D, Abingdon, UK) or PMA (Sigma) for
24 hours in serum-free medium. Medium was collected, centrifuged
at 10000 g, for 2 minutes and assayed for gelatinase or TIMP activity
by zymography.
Gelatin zymography
The presence of gelatinases MMP-2 and MMP-9 in endothelial cell
conditioned medium was analysed by zymography in 10%
polyacrylamide gels containing 1 mg/ml gelatin (Sigma) as previously
described (Fridman et al., 1995). Briefly, samples were mixed with
Laemmli sample buffer without reducing agents or heating and were
subjected to SDS-PAGE. The gels were incubated for 30 minutes at
22°C in renaturating buffer (2.5% Triton X-100), rinsed in distilled
H
2
O, and then incubated in developing buffer (50 mM Tris buffer pH
7.5, 200 mM NaCl, 5 mM CaCl
2
) for 18 hours at 37°C. The gels were
stained with 0.2% Coomassie Blue R250 in a solution of 20%
isopropanol and 10% acetic acid and then destained in 20%
isopropanol, 10% acetic acid. Reverse zymography for the analysis of
TIMP activity was performed as previously described (Oliver et al.,
1997), by incorporating 150 ng/ml recombinant MMP-9 in addition
to 2.5 mg/ml gelatin in 12% polyacrylamide gels. Otherwise the
procedure is identical to that of direct zymography.
Cell migration assay
Endothelial cell motility was assessed using a monolayer wound assay
(Goodman et al., 1985). The cells were seeded into 12-well culture
dishes at a concentration of 5×10
4
cells/well, and cultured in medium
containing 10% FCS for 24 hours. The nearly confluent monolayers
were wounded with a plastic pipette tip and any cellular debris was
removed by washing with PBS. The wounded monolayers were then
incubated for 24 hours in serum-free medium in the presence or
absence of added factors (1 ng/ml TGF-β1, 50 ng/ml PMA), alone or
in combination, and then photographed. In some experiments, tissue
kallikrein which is an efficient activator of pro-MMP-9 (Desrivieres
et al., 1993) was added at 50 ng/ml in the presence or absence of the
factors. Additional experiments were performed in the presence of
exogenous activated or latent human recombinant MMP-9 (0.5 µg/ml)
which was added immediately after wounding. Recombinant pro-
MMP-9 was a kind gift from Dr R Fridman (Wayne State University,
MI) and was prepared as described previously (Olson et al., 1997).
Activation of MMP-9 was accomplished by incubating with 1 mM 4-
aminophenylmercuric acetate (APMA) in buffer containing 50 mM
Tris-HCl pH 7.5, 150 mM NaCl, 5 mM CaCl
2
, for one hour at 22°C,
after which the APMA was removed by Bio-Spin chromatography
columns (Bio-Rad, CA). Both latent and activated MMP-9 (5 µl) were
diluted in 1 ml of serum-free medium and sterilised by filtration.
Cell invasion assay
Endothelial cell invasiveness was studied in modified Boyden
chambers (Frandsen et al., 1992) containing chemotaxis membranes
of 13 mm diameter with 12 µm pore size (Nucleopore) which were
coated with 60 µg of the reconstituted basement membrane Matrigel
(Beckton Dickinson). Cells were detached with 1 mM EDTA, washed
and resuspended in serum-free medium containing 0.2 mg/ml BSA.
Then 2×10
5
cells were added to the upper compartment of the Boyden
chamber with or without additional factors (50 ng/ml PMA or 1 ng/ml
TGF-β1). Medium containing 10 ng/ml bFGF as a chemoattractant
and 2 mg/ml BSA was placed in the lower chamber. After incubation
at 37°C for 24 hours, filters were fixed and stained with Diff-Quik
(Dade AG, Switzerland) and the cells attached to the bottom side of
the membrane were counted visually under the microscope. The data
are expressed as the total number of cells counted per ten microscopic
fields.
RT-PCR analysis
Endothelial cells (2×10
6
) were plated in 60 mm dishes and were
treated the next day with either TGF-β1 (1 ng/ml) or PMA (50 ng/ml)
or both, for 24 hours. Then cell monolayers were washed with PBS
and total mRNA was prepared using the TRIzol reagent (Gibco/BRL)
following the manufacturer’s instructions. Approx. 2 µg of mRNA
were used for cDNA synthesis using 200 ng random hexamer primer,
and 200 U MMTV-RT (Gibco/BRL) in a 20 µl final reaction volume
at 40°C for 60 minutes. The samples were then diluted 5-fold to be
within the linear range for PCR amplification. Aliquots of cDNA (2
µl) were amplified in a final reaction volume of 25 µl containing 2
µM oligonucleotide primers, 1.5 mM MgCl
2
, 10× reaction buffer, and
Taq polymerase (Promega). Amplification consisted of 25 cycles of
94°C 1 minute, 54°C 1 minute, 72°C 1 minute. 10 µl of the PCR
reaction were analysed on a 1.5% agarose gel containing ethidium
bromide and photographed. The sequence of oligonucleotide primers
A. Puyraimond and others
1285Gelatinases and endothelial cell motility
(puchased from Eurogentec, Belgium) and the size of amplified DNA
fragments were as follows; MMP-9 5′-TCC TGG TGC TGG CTT
GCT GC-3′ (sense) and 5′-CAA TGT CAG CTT CGG GGC CG-3′
(antisense) 467 bp fragment; MMP-2 5′-TTT GGA CTG CCC CAG
ACA GG-3′ (sense) and 5′-GCT GCG GCC AGT ATC AGT GC-3′
(antisense) 518 bp fragment; TIMP-1 5′-TTG CAC CCA TGG CCT
CTG GC-3′ (sense) and 5′-GCA GGA CTC AGG CCA TCT GG-3′
(antisense) 620 bp fragment; TIMP-2 5′-ACC CGC AAC AGG CGT
TTT GC-3′ (sense) and 5′-GCC GTC GCT TCT CTT GAT GC-3′
(antisense) 499 bp fragment. Primers for the amplification of β-actin
or glyceraldehyde-3-phosphate dehydrogenase (GPDH) genes were
used as controls: β-ACTIN 5′-CCA GAC AGC ACT GTG TTG GC-
3′ (sense) and 5′-GAG AAG CTG TGC TAC GTC GC-3′ (antisense)
270 bp fragment; GPDH 5′-GGC AAA GTG GAC ATC GTC GC-3′
(sense) and 5′-CCT CTG ACG CCT GCT TCA CC-3′ (antisense).
The design of primers was based on the bovine sequences available
in the GenBank database, with the exception of MMP-2 for which
sequences from the human gene were used as no bovine sequences
were available. Primers were designed to be of approximately equal
length, GC content and annealing temperature.
Immunoblot analysis
Immunoblot analysis to detect c-Jun, JunB and JunD proteins was
carried out using specific antibodies as previously described (Pfarr et
al., 1994). Whole cell protein extracts of CPAE cultures treated with
either TGF-β1 (1 ng/ml) or PMA (50 ng/ml) or both, for 2 hours were
scraped directly into Laemmli loading buffer and 20 µg of extracts
were separated by 10% SDS-PAGE and transferred onto nitrocellulose
filters prior to immunoblotting with anti-Jun antibodies followed by
enhanced chemiluminescence detection (Pfarr et al., 1994).
RESULTS
Effect of TGF-β1 and PMA on gelatinase secretion
by endothelial cells
We studied gelatinase secretion of primary cultures of calf
pulmonary artery endothelial (CPAE) cells by gelatin
zymography. Endothelial cells grown in serum-free medium
secreted low levels of gelatinaseA/MMP-2, while gelatinaseB/
MMP-9 was not detected (Fig. 1). To examine a role for MMP-
2 and MMP-9 in angiogenesis, we investigated the effect of a
number of angiogenic stimulators on gelatinase secretion by
endothelial cells. The greatest effect was observed upon
addition of TGF-β1 or PMA, whereas several other angiogenic
factors (including vascular endothelial growth factor, tumor
necrosis factor-α, and bFGF) had negligible effect (data not
shown). TGF-β1 stimulated the secretion of MMP-2 in a
concentration-dependent manner (Fig. 1). At higher
concentrations (3 ng/ml), stimulation of MMP-9 secretion was
also detected. The most striking effect was the synergy
observed when TGF-β1 was added in combination with PMA.
When TGF-β1 was added together with 50 ng/ml PMA to
cultures of CPAE a dramatic increase in MMP-9 secretion was
observed. Similar synergy was observed with zymogram
analysis of whole cell lysates (not shown). This effect was
shown to be concentration-dependent (Fig. 1). Hence, both
TGF-β1 and PMA were able to alter the gelatinase profiles of
endothelial cells and exhibited a synergistic effect on MMP-9
induction.
Effect of TGF-β1 and PMA on TIMP secretion by
endothelial cells
Proteolytic potential is determined by the balance between
secreted MMP proteases and their inhibitors, the TIMPs. We
examined the expression of TIMPs secreted by CPAE cells by
reverse zymography. There was a constitutive expression of
both TIMP-1 and TIMP-2 in these cells. The addition of either
PMA or TGF-β1 increased the expression of TIMP-1 in CPAE
cells (Fig. 2). However, treatment with a combination of TGF-
β1 plus PMA together had a marked negative effect on TIMP-
1 (Fig. 2). The combination of the two factors also had a similar
negative effect on TIMP-2 expression. Thus, both TGF-β1 and
PMA were able to upregulate TIMP-1 expression when added
alone, but the combination of both factors had an inhibitory
effect. Interestingly, these downregulation effects were
reciprocal with the synergistic upregulation of MMP-9 by these
factors. Hence, in endothelial cells TGF-β1 plus PMA induced
a strong increase in the net balance between gelatinases and
TIMPs.
The effects of TGF-β1 and PMA are transcriptionally
mediated
In order to determine whether the observed effects of TGF-β1
and PMA on MMP and TIMP production were
Fig. 1. Effect of TGF-β1 and PMA on gelatinase secretion by
endothelial cells. Gelatinase secretion was studied in CPAE cells by
gelatin zymography. Endothelial cells were grown in medium
containing FCS until almost confluent. They were then switched to
serum-free medium containing increasing concentrations of TGF-β1,
in the presence or absence of 50 ng/ml PMA. Conditioned medium
was collected 24 hours later and analysed by gelatin zymography. We
observed a synergistic upregulation of MMP-9 on addition of both
factors together.
Fig. 2. Effect of TGF-β1 and PMA on TIMP secretion by endothelial
cells. The expression of TIMPs secreted by endothelial cells was
examined by reverse zymography on polyacrylamide gels containing
recombinant MMP-9 and gelatin. The addition of either TGF-β1 (1
ng/ml) or PMA (50 ng/ml) increased the expression of TIMP-1.
However, the combination of TGF-β1 plus PMA had a negative
effect on TIMP expression in CPAE cells.
1286
transcriptionally regulated, we established a sensitive RT-PCR
assay to detect mRNA transcripts for the MMP and TIMP
genes. Sets of oligonucleotide primers were designed of similar
length, GC content and annealing temperatures. Primers for the
amplification of β-actin or GPDH genes were used as controls.
The results of such analyses are shown in Fig. 3. These results
correlate closely with the protein expression levels observed by
zymography. Hence, we observed a synergistic induction of
MMP-9 mRNA levels upon combination of the two factors in
CPAE cells. Induction of MMP-9 mRNA was detectable as
early as two hours after addition of TGF-β1 plus PMA (data
not shown). Also, inhibition of TIMP-1 and TIMP-2 mRNA
transcripts upon treatment with TGF-β1 plus PMA was
observed (Fig. 3). Neither factor had any effect on the levels
of β-actin or GPDH. Hence, RT-PCR analysis showed that
measured mRNA levels correlated well with the secreted
protein levels, suggesting that regulation by TGF-β1 and PMA
is at the transcriptional level.
The upregulation of MMP-9 by PMA in tumor cells has been
shown to be mediated by the AP-1 transcription factor which
is composed of members of the Fos and Jun proto-oncogene
families (Fini et al., 1994; Sato and Seiki, 1993). Therefore,
we examined whether the response of the CPAE cells to TGF-
β1 and PMA treatment could be due to specific regulation of
Jun family members. There are three mammalian Jun proteins
(c-Jun, JunB and JunD) which are able to form homodimers or
Jun/Fos heterodimers. The three proteins differ in their
induction by signalling pathways and in their transactivation
properties (Angel and Karin, 1991). We investigated the levels
of the three Jun members in whole cell extracts of CPAE cells
by western blot analysis with specific antibodies (Pfarr et al.,
1994). TGF-β1 and PMA both induced expression of the c-Jun
protein, whereas treatment with both factors together lead to
an increased induction of c-Jun and a striking downregulation
of JunB levels (Fig. 4). There was a high basal level of JunD
expression which was relatively unresponsive to treatment with
either factor. JunB is known to be a much poorer transcriptional
activator than c-Jun and to antagonise c-Jun transactivation
(Chiu et al., 1989; Schutte et al., 1989). Hence, the dramatic
switch in the c-Jun/JunB ratio upon combined TGF-β1 plus
PMA treatment correlated with the synergistic upregulation of
MMP-9 transcription and down-regulation of TIMP
expression.
Effect of TGF-β1 and PMA on endothelial cell
invasion and migration
The response of the CPAE endothelial cells to the combinations
of TGF-β1 and PMA presented us with a model system to
evaluate the importance of MMP-9 expression and the MMP-
9/TIMP balance in endothelial cell invasion and migration.
The invasive capacity of the endothelial cells in response to
PMA and TGF-β1 was investigated by measuring the invasion
of a Matrigel layer in a Boyden chamber assay. The effect of
incubating CPAE cells with TGF-β1 or PMA on bFGF-induced
cell invasion is shown in Fig. 5. PMA had a dramatic effect on
cell invasiveness, but this was abolished by co-incubation with
PMA plus TGF-β1. The migratory response of the endothelial
cells was evaluated by observing their ability to fill in a
denuded area of the culture dish following wounding of
monolayers. Results of typical experiment are shown in Fig. 6.
PMA increased the migration potential of the CPAE cells,
whereas TGF-β1 had limited effects on migration and
abolished the stimulatory effect of PMA on migration when
added together. This inhibitory effect of TGF-β1 is consistent
with previous reports of the inhibition of bFGF-induced
locomotion by TGF-β1 in aortic and capillary endothelial cells
(Heimark et al., 1986; Muller et al., 1987).
Thus, the significant increase in the MMP-9/TIMP ratio in
CPAE cells upon treatment with PMA plus TGF-β1 was
without any effect on either cell migration or invasion. In
order to check whether this is because MMP-9 is present in
its latent form, the migration experiment was repeated in the
A. Puyraimond and others
Fig. 3. Effect of TGF-β1 and
PMA on MMP and TIMP
mRNA levels. The expression
of mRNA transcripts for the
MMP-2, MMP-9, TIMP-1 and
TIMP-2 genes was detected by
RT-PCR analysis. CPAE
cultures were grown under the
same conditions as
zymography assays for 24
hours. Primers for the
amplification of β-actin (β-
ACT) or GPDH genes were
used as controls. Cells were
either untreated (lane 1),
incubated with 1 ng/ml TGF-
β1 (lane 2), 50 ng/ml PMA
(lane 3), or both factors
together (lane 4). The effects
of TGF-β1 and PMA on MMP
and TIMP mRNA levels
correlated very closely with
the results of zymography
analysis.
Fig. 4. Effect of TGF-
β1 and PMA on levels
of the Jun family
proteins. Levels of the
three Jun proteins (c-
Jun, JunB and JunD)
were determined in
whole cell extracts of
endothelial cells by
immunoblot analysis
with specific
antibodies. Cells were
either untreated (lane
1), incubated with 1
ng/ml TGF-β1 (lane 2),
50 ng/ml PMA (lane
3), or both factors
together (lane 4).
Treatment for two
hours with TGF-β1 or
PMA induced marked
differences in the c-Jun
and JunB profiles.
1287Gelatinases and endothelial cell motility
presence of 50 ng/ml tissue kallikrein which causes MMP-9
activation (Desrivieres et al., 1993). Tissue kallikrein was
added to cells in the presence or absence of TGF-β1 and
PMA. However, the activation of the secreted MMP-9 by
tissue kallikrein, which was confirmed by gelatin
zymography, had no effect on cell migration (results not
shown). Finally, an additional experiment was performed in
which migration assays were performed following the
addition of exogenous APMA-activated or latent MMP-9 (see
Materials and Methods). This too had no effect on endothelial
cell migration. These results suggest that the regulation of
MMP-9 by TGFβ and PMA does not play a direct role in the
migration of endothelial cells. Plating cells at varying cell
densities did not affect MMP-9 induction by PMA and TGFβ,
indicating that motile and non-motile cells did not differ in
their response to these stimuli.
The effect of PMA and TGF-β1 on cell proliferation was
also studied in order to determine whether the inhibitory effect
of TGF-β1 on cell migration could be explained by an
inhibitory effect on cell proliferation. These effects appear to
be independent of growth regulation, as PMA caused only a
very modest increase in CPAE cell proliferation and TGF-β1
had a negligable effect (data not shown).
DISCUSSION
Defining the regulation of specific endothelial cell proteases by
cytokines and their contribution to the migratory steps of
angiogenesis is critical to our ability to intervene and develop
new therapeutic strategies. Here we have shown that two
factors associated with tumor progression are able to cooperate
to induce MMP-9 expression and decrease TIMP expression.
Furthermore, we have exploited this response to investigate the
relationship between the MMP/TIMP gelatinolytic balance and
ednothelial cell migration and invasion. Our results suggest
that MMP-9 proteolytic potential does not correlate with
motility, raising the possibility that MMPs contribute in
complex ways to the angiogenic process.
Regulation of MMPs and TIMPs by TGF-β1 and PMA
in endothelial cells
We have shown that the addition of TGF-β1 and PMA, either
alone or in combination, can regulate the expression of MMP-
2, MMP-9, TIMP-1 and TIMP-2 in endothelial cells. Few
studies have investigated the regulation of both gelatinases and
both TIMPs by cytokines in endothelial cells and none have
looked at the effects of TGF-β1. Our results show that MMP-
2 and MMP-9 genes are regulated differently in response to
these factors (see Table 1).
TGF-β1 induced MMP-2 expression in CPAE cells in a
dose-dependent manner. This is in contrast to suggestions that
some of the biological effects of TGF-β1 are mediated by
inhibition of MMP expression (Kerr et al., 1990; Matrisian et
al., 1992; Mauviel et al., 1993). This contrast could reflect the
fact that different MMP genes respond differently to TGF-β
and that MMP regulation may depend on the cell type. For
control TGFβ1 PMA TGFβ1+PMA
0
200
400
600
800
1000
number of invasive cells
Fig. 5. Effects of TGF-β1 and PMA on endothelial cell invasion. The
invasive capacity of CPAE cells in response to PMA (50 ng/ml) and
TGF-β1 (1 ng/ml) was investigated by measuring the invasion of a
Matrigel layer in a Boyden chamber assay. Cells migrating through
the gel to the membrane filter after 24 hours were counted. The
results shown represent the mean ± s.d. of triplicate samples. Similar
results were obtained in three independent experiments.
Fig. 6. Effects of TGF-β1 and PMA on endothelial cell migration.
The migration of the CPAE cells was evaluated by their ability to fill
in a denuded area of the culture dish following wounding of
monolayers. PMA (50 ng/ml) increased the cell motility, whereas
TGF-β1 (1 ng/ml) had an inhibitory effect on migration abolishing
the stimulatory effect of PMA when added together.
Table 1. Summary of the effects of TGF-β1 and PMA on
gelatinase and TIMP expression, Jun levels, cell migration
and invasion
Control TGF-β1 PMA TGF-β1+PMA
MMP-2 + +++ +++ ++
MMP-9 − + + +++
TIMP-1 + +++ +++ +/−
TIMP-2 + + + +/−
c-Jun +/− ++ + +++
JunB +++ + + +/−
Migration − +/− +++ −
Invasion + + +++ +
The table summarizes our results of treatment with TGF-β1 and/or PMA.
The comparative levels are represented by −, undetectable; +/−, very low; +,
low; ++, medium; +++, high levels.
1288
example, while TGF-β inhibits MMP-1 expression in
fibroblasts it was shown to induce MMP-13 in these cells (Uria
et al., 1998). Furthermore, TGF-β strongly upregulated MMP-
1 in cultured keratinocytes (Mauviel et al., 1996). Upregulation
of MMP-2 by TGF-β appears not to be restricted to endothelial
cells, as it has also been reported in fibroblasts and
keratinocytes (Overall et al., 1991; Salo et al., 1991). MMP-9
also responded to TGF-β1 stimulation. The most striking
observation was the synergistic upregulation of MMP-9 upon
treatment with both TGF-β1 and PMA together. This result
clearly indicates that the effects of cytokines depend very much
on the concentration of other extracellular regulators. This
point is important when considering the combinations of
cellular factors that contribute to the progression of the
angiogenic switch (Hanahan and Folkman, 1996). This
cooperation between TGF-β1 and PMA was also observed in
the reciprocal downregulation of TIMP-1 and TIMP-2 levels
in CPAE cells. This resulted in dramatic effects on the net
gelatinolytic potential in response to stimulation.
Many reports have examined the signal transduction
pathways responsible for MMP regulation in fibroblasts and
keratinocytes (Crawford and Matrisian, 1996; Matrisian et al.,
1992; Mauviel et al., 1996; Overall et al., 1991; Salo et al.,
1991; Uria et al., 1998). However, far less is known about the
signal transduction pathways and transcription factors that
regulate endothelial MMP gene regulation. Our results offer an
attractive system to investigate the molecular mechanisms
underlying the regulation of proteolytic potential in endothelial
cells. We used RT-PCR analysis to verify that the regulation
that we observed reflected changes at the mRNA level.
Although it is possible that there may be some regulation of
mRNA stability, it is likely that the upregulation of MMP and
TIMP genes is transcriptional.
Many of the MMP genes are known to be regulated by the
AP-1 transcription factor binding to TRE sites present in their
promoters (Crawford and Matrisian, 1996) and MMP-9 has
been reported to be regulated by AP-1 in fibroblasts (Sato and
Seiki, 1993). AP-1 is a dimer composed of members of the Jun
and Fos family of proto-oncogenes (Angel and Karin, 1991).
We investigated whether the effects we observed could be
explained by regulation of the Jun members in response to
extracellular stimuli. It has been shown that the three Jun
proteins (c-Jun, JunB and JunD) differ in their regulation and
transactivation potential (Angel and Karin, 1991; Chiu et al.,
1989; Pfarr et al., 1994; Schutte et al., 1989). Specifically,
while c-Jun is a very efficient transcriptional activator, JunB
and JunD are less effective. Furthermore, JunB and JunD may
antagonise that effects of c-Jun, and it appears that the net
balance of different Jun dimers determines the AP-1 activity
and tissue-specific cellular outcome (Chiu et al., 1989; Pfarr et
al., 1994; Schutte et al., 1989). Our immunoblot analysis
revealed that the addition of TGF-β1 plus PMA lead to changes
in Jun levels. Neither factor had a significant effect on JunD
levels, consistent with previous reports of JunD regulation
(Pfarr et al., 1994). However, TGF-β1 plus PMA treatment
caused a net increase in the c-Jun:JunB ratio. These
observations likely account for the cooperative upregulation of
MMP-9 in CPAE cells. A similar shift in the c-Jun:JunB ratio
has been reported to account for the inhibitory effect of TGF-
β on MMP-1 expression (Mauviel et al., 1993). and differential
MMP regulation in fibroblasts and keratinocytes (Mauviel et
al., 1996). Changes in AP-1 composition could also play a role
in TIMP-1 regulation in endothelial cells, as its promoter has
been shown to be under complex AP-1-mediated control
(Logan et al., 1996).
TGF-β has also been reported to activate the c-Jun N-
terminal kinase (JNK) (Wang et al., 1997). Immunoblot
analysis with an antibody that specifically recognised the
phosphorylated form of c-Jun, showed that TGF-β1 (and PMA)
was also able to activate JNK activity in CPAE cells (data not
shown). Recent studies of the mechanism of TGF-β signalling
have shown that SMAD proteins play critical roles in signalling
from the surface receptor to the nucleus (Kretzschmar and
Massague, 1998). It will be interesting to examine further a
potential role for endothelial SMADs in the regulation of
gelatinase and TIMP genes in response to TGF-β1. The results
we have described offer a model system to explore AP-
1/SMAD cooperation and their contribution to the endothelial
response to TGF-β. We suggest that future study of the MMP-
9 and TIMP promoters in CPAE cells will be informative to
our understanding of the convergence of signal transduction
pathways and different transcriptional outcomes in endothelial
cells.
Gelatinolytic balance and endothelial cell invasive
capacity
The observed effect of TGF-β1 plus PMA on the overall
gelatinolytic balance offered us an attractive system to examine
the relationship between gelatinolytic potential and the
migratory capacity of endothelial cells. MMP-9 has been
shown to play a role in the migratory and metastatic behaviour
of tumorigenic cells (Bernhard et al., 1994; Himelstein et al.,
1994). However, the enhanced proteolytic potential that we
observed in CPAE cells treated with PMA and TGF-β1,
appeared to be without consequences on endothelial cell
motility. CPAE cultures treated with PMA plus TGF-β1
showed similar or even reduced locomotion relative to
untreated cells. The effects of TGF-β1 on cell migration appear
to be unrelated to the effects that this factor has on cell
proliferation. This is consistent with the idea that the
pleiotropic effects of TGF-β1 are unrelated to each other. For
example, although TGF-β1 blocks cell motility, when capillary
endothelial cells are grown in three-dimensional collagen gels
TGF-β1 induces the formation of a complex branching and
tube-like structures, without influencing their growth (Madri et
al., 1988). The effects of the combination of TGFβ and PMA
on tube formation will be the focus of further research in our
laboratory. Our results suggest that the striking effect of TGF-
β1 on MMP-9 regulation is also independent of effects on cell
proliferation or cell migration.
Our results question a direct role for MMP-9 in endothelial
cell motility in vitro. It is possible that in vivo abundant cellular
interactions and complex cooperation between different
proteases play important roles. It is also possible that other
MMPs, such as MMP-1 and MMP-3, contribute significantly
to the endothelial cell migratory capacity. Indeed, a recent
study implicated MT1-MMP as the critical MMP in
angiogenesis (Hiraoka et al., 1998). However, our observations
are consistent with the recent suggestions concerning complex
roles for MMPs in processes such as angiogenesis and
metastasis. For example, a recent study showed that
homozygous mice with a null mutation in the MMP-9 gene
A. Puyraimond and others
1289Gelatinases and endothelial cell motility
exhibited delayed skeletal growth plate vascularization and
ossification (Vu et al., 1998). On the basis of their evidence the
authors suggested that the role of MMP-9 in vascularization is
indirect; possibly by regulating chondrocytes apoptosis, by
generating angiogenic signals from the surrounding tissues or
by inactivating angiogenic inhibitors (Vu et al., 1998). Indeed,
it is possible that physiological substrates for MMP-9 are not
restricted to basement membrane ECM components. While the
prevailing view has been that the role of MMPs in metastasis
was to promote migration and invasion of cancer cells in and
out of the blood or lymphatic vessels, recent evidence using
intravital video microscopy, genetic and pharmacological
manipulation of MMPs and TIMPs suggests a broader, more
complex role for these proteases than previously believed
(Chambers and Matrisian, 1997). Furthermore, MMPs and
their inhibitors now appear to be important regulators of the
tumor growth, possibly by the regulation of growth
environment, access to growth factors or the regulation of
growth factor themselves (Chambers and Matrisian, 1997;
Fowlkes et al., 1994). Hence, MMP-9 activity may generate
angiogenic regulatory peptides (Patterson and Sang, 1997) or
release factors from the surrounding tissue. We propose that
the role of gelatinases in angiogenesis is more complex than
merely increasing the motility of endothelial cells. Finally,
further understanding of the regulation of proteolytic capacity
and its contribution to endothelial cell migration and
angiogenesis will be important to the development of future
clinical applications of anti-angiogenic therapeutic strategies.
We thank Dr D. Lallemand for kindly providing Jun-specific
antibodies, Dr J. Badet for CPAE cells, and Dr R. Fridman for
recombinant MMP-9. We thank Dr M. Yaniv for his support and for
his insightful comments on the manuscript. J.B.W. was supported by
a TMR Marie Curie postdoctoral fellowship from the EC.
REFERENCES
Alexander, C. M., Hansell, E. J., Behrendtsen, O., Flannery, M. L.,
Kishnani, N. S., Hawkes, S. P. and Werb, Z. (1996). Expression and
function of matrix metalloproteinases and their inhibitors at the maternal-
embryonic boundary during mouse embryo implantation. Development 122,
1723-1736.
Angel, P. and Karin, M. (1991). The role of Jun, Fos and the AP-1 complex
in cell-proliferation and transformation. Biochim. Biophys. Acta 1072, 129-
157.
Bernhard, E. J., Gruber, S. B. and Muschel, R. J. (1994). Direct evidence
linking expression of matrix metalloproteinase 9 (92-kDa
gelatinase/collagenase) to the metastatic phenotype in transformed rat
embryo cells. Proc. Nat. Acad. Sci. USA 91, 4293-4297.
Bischof, P., Martelli, M., Campana, A., Itoh, Y., Ogata, Y. and Nagase, H.
(1995). Importance of matrix metalloproteinases in human trophoblast
invasion. Early Pregnancy 1, 263-269.
Chambers, A. F. and Matrisian, L. M. (1997). Changing views of the role
of matrix matalloproteinases in metastasis. J. Nat. Canc. Inst. 89, 1260-
1270.
Chiu, R., Angel, P. and Karin, M. (1989). Jun-B differs in its biological
properties from, and is a negative regulator of, c-Jun. Cell 59, 979-
986.
Crawford, H. C. and Matrisian, L. M. (1996). Mechanisms controlling the
transcription of matrix metalloproteinase genes in normal and neoplastic
cells. Enzyme Protein 49, 20-37.
Desrivieres, S., Lu, H., Peyri, N., Soria, C., Legrand, Y. and Menashi, S.
(1993). Activation of the 92 kDa type IV collagenase by tissue kallikrein.
J. Cell. Physiol. 157, 587-593.
Fini, M. E., Bartlett, J. D., Matsubara, M., Rinehart, W. B., Mody, M. K.,
Girard, M. T. and Rainville, M. (1994). The rabbit gene for 92-kDa matrix
metalloproteinase. Role of AP1 and AP2 in cell type-specific transcription.
J. Biol. Chem. 269, 28620-28628.
Fowlkes, J. L., Enghild, J. J., Suzuki, K. and Nagase, H. (1994). Matrix
metalloproteinases degrade insulin-like growth factor-binding protein-3 in
dermal fibroblast cultures. J. Biol. Chem. 269, 25742-25746.
Frandsen, T. L., Boysen, B. E., Jirus, S., Zwiebel, J., Spang-Thomas, M.,
Thompson, E. W. and Brunner, N. (1992). Assays for the study of human
cancer cell invasion and metastasis. Fibrinolysis 6, 71-76.
Fridman, R., Toth, M., Pena, D. and Mobashery, S. (1995). Activation of
progelatinase B (MMP-9) by gelatinase A (MMP-2). Cancer Res. 55, 2548-
2555.
Galardy, R. E., Grobelny, D., Foellmer, H. G. and Fernandez, L. A. (1994).
Inhibition of angiogenesis by the matrix metalloprotease inhibitor N-[2R-2-
(hydroxamidocarbonymethyl)-4-methylpentanoyl)]-L-tryptophan
methylamide. Cancer Res. 54, 4715-4718.
Goodman, S. L., Vollmers, H. P. and Birchmeier, W. (1985). Control of cell
locomotion: perturbation with an antibody directed against specific
glycoproteins. Cell 41, 1029-1038.
Hanahan, D. and Folkman, J. (1996). Patterns and emerging mechanisms of
the angiogenic switch during tumorigenesis. Cell 86, 353-364.
Heimark, R. L., Twardzik, D. R. and Schwartz, S. M. (1986). Inhibition of
endothelial regeneration by type-beta transforming growth factor from
platelets. Science 233, 1078-1080.
Himelstein, B. P., Canete, S. R., Bernhard, E. J., Dilks, D. W. and Muschel,
R. J. (1994). Metalloproteinases in tumor progression: the contribution of
MMP-9. Inv. Metast. 14, 246-258.
Hiraoka, N., Allen, E., Apel, I. J., Gyetko, M. R. and Weiss, S. J. (1998).
Matrix metalloproteinases regulate neovascularisation by acting as
pericellular fibrinolysins. Cell 95, 365-377.
Itoh, T., Tanioka, M., Yoshida, H., Yoshioka, T., Nishimoto, H. and
Itohara, S. (1998). Reduced angiogenesis and tumor progression in
gelatinase A-deficient mice. Cancer Res. 58, 1048-1051.
Johnson, M. D., Kim, H. R., Chesler, L., Tsao, W. G., Bouck, N. and
Polverini, P. J. (1994). Inhibition of angiogenesis by tissue inhibitor of
metalloproteinase. J. Cell. Physiol. 160, 194-202.
Kerr, L. D., Miller, D. B. and Matrisian, L. M. (1990). TGF-β1 inhibition
of transin/stromelysin gene expression is mediated through a Fos binding
sequence. Cell 61, 267-278.
Kretzschmar, M. and Massague, J. (1998). SMADs: mediators and
regulators of TGF-β signaling. Curr. Opin. Genet. Dev. 8, 103-111.
Levi, E., Fridman, R., Miao, H. Q., Ma, Y. S., Yayon, A. and Vlodavsky,
I. (1996). Matrix metalloproteinase 2 releases active soluble ectodomain of
fibroblast growth factor receptor 1. Proc. Nat. Acad. Sci. USA 93, 7069-
7074.
Logan, S. K., Garabedian, M. J., Campbell, C. E. and Werb, Z. (1996).
Synergistic transcriptional activation of the tissue inhibitor of
metalloproteinases-1 promoter via functional interaction of AP-1 and Ets-1
transcription factors. J. Biol. Chem. 271, 774-782.
Madri, J. A., Pratt, B. M. and Tucker, A. M. (1988). Phenotypic modulation
of endothelial cells by transforming growth factor-β depends upon the
composition and organization of the extracellular matrix. J. Cell Biol. 106,
1375-1384.
Matrisian, L. M. (1990). Metalloproteinases and their inhibitors in matrix
remodeling. Trends Genet. 6, 121-125.
Matrisian, L. M., Ganser, G. L., Kerr, L. D., Pelton, R. W. and Wood, L.
D. (1992). Negative regulation of gene expression by TGF-β. Mol. Reprod.
Dev. 32, 111-120.
Mauviel, A., Chen, Y. Q., Dong, W., Evans, C. H. and Uitto, J. (1993).
Transcriptional interactions of transforming growth factor-β with pro-
inflammatory cytokines. Curr. Biol. 3, 822-831.
Mauviel, A., Chung, K. Y., Agarwal, A., Tamai, K. and Uitto, J. (1996).
Cell-specific induction of distinct oncogenes of the Jun family is responsible
for differential regulation of collagenase gene expression by transforming
growth factor-β in fibroblasts and keratinocytes. J. Biol. Chem. 271, 10917-
10923.
Montesano, R. and Orci, L. (1985). Tumor-promoting phorbol esters induce
angiogenesis in vitro. Cell 42, 469-477.
Montesano, R. and Orci, L. (1987). Phorbol esters induce angiogenesis
in vitro from large-vessel endothelial cells. J. Cell. Physiol. 130, 284-
291.
Muller, G., Behrens, J., Nussbaumer, U., Bohlen, P. and Birchmeier, W.
(1987). Inhibitory action of transforming growth factor β on endothelial
cells. Proc. Nat. Acad. Sci. USA 84, 5600-5604.
1290
Murphy, A. N., Unsworth, E. J. and Stetler-Stevenson, W. G. (1993). Tissue
inhibitor of metalloproteinases-2 inhibits bFGF-induced human
microvascular endothelial cell proliferation. J. Cell. Physiol. 157, 351-358.
Murphy, G., Cockett, M. I., Ward, R. V. and Docherty, A. J. (1991). Matrix
metalloproteinase degradation of elastin, type IV collagen and proteoglycan.
A quantitative comparison of the activities of 95 kDa and 72 kDa
gelatinases, stromelysins-1 and -2 and punctuated metalloproteinase
(PUMP). Biochem. J. 277, 277-279.
Ochieng, J., Fridman, R., Nangia-Makker, P., Kleiner, D. E., Liotta, L. A.,
Stetler-Stevenson, W. G. and Raz, A. (1994). Galectin-3 is a novel
substrate for human matrix metalloproteinases-2 and -9. Biochemistry 33,
14109-14114.
Oliver, G. W., Leferson, J. D., Stetler-Stevenson, W. G. and Kleiner, D. E.
(1997). Quantitative reverse zymography: analysis of picogram amounts of
metalloproteinase inhibitors using gelatinase A and B reverse zymograms.
Anal. Biochem. 244, 161-166.
Olson, M. W., Gervasi, D. C., Mobashery, S. and Fridman, R. (1997).
Kinetic analysis of the binding of human matrix metalloproteinase-2 and -
9 to tissue inhibitor of metalloproteinase (TIMP)-1 and TIMP-2. J. Biol.
Chem. 272, 29975-29983.
Overall, C. M., Wrana, J. L. and Sodek, J. (1991). Transcriptional and post-
transcriptional regulation of 72-kDa gelatinase/type IV collagenase by
transforming growth factor-β1 in human fibroblasts. Comparisons with
collagenase and tissue inhibitor of matrix metalloproteinase gene
expression. J. Biol. Chem. 266, 14064-14071.
Patterson, B. C. and Sang, Q. A. (1997). Angiostatin-converting enzyme
activities of human matrilysin (MMP-7) and gelatinase B/type IV
collagenase (MMP-9). J. Biol. Chem. 272, 28823-28825.
Pepper, M. S., Belin, D., Montesano, R., Orci, L. and Vassalli, J. D. (1990).
Transforming growth factor-β1 modulates basic fibroblast growth factor-
induced proteolytic and angiogenic properties of endothelial cells in vitro.
J. Cell Biol. 111, 743-755.
Pepper, M. S., Vassalli, J. D., Orci, L. and Montesano, R. (1993). Biphasic
effect of transforming growth factor-β1 on in vitro angiogenesis. Exp. Cell
Res. 204, 356-363.
Pfarr, C. M., Mechta, F., Spyrou, G., Lallemand, D., Carillo, S. and Yaniv,
M. (1994). Mouse JunD negatively regulates fibroblast growth and
antagonizes transformation by ras. Cell 76, 747-755.
Salo, T., Lyons, J. G., Rahemtulla, F., Birkedal-Hansen, H. and Larjava,
H. (1991). Transforming growth factor-β1 up-regulates type IV collagenase
expression in cultured human keratinocytes. J. Biol. Chem. 266, 11436-
11441.
Sato, H. and Seiki, M. (1993). Regulatory mechanism of 92 kDa type IV
collagenase gene expression which is associated with invasiveness of tumor
cells. Oncogene 8, 395-405.
Schutte, J., Viallet, J., Nau, M., Segal, S., Fedorko, J. and Minna, J. (1989).
jun-B inhibits and c-fos stimulates the transforming and trans-activating
activities of c-jun. Cell 59, 987-997.
Shimizu, S., Nishikawa, Y., Kuroda, K., Takagi, S., Kozaki, K., Hyuga, S.,
Saga, S. and Matsuyama, M. (1996). Involvement of transforming growth
factor β1 in autocrine enhancement of gelatinase B secretion by murine
metastatic colon carcinoma cells. Cancer Res. 56, 3366-3367.
Stetler-Stevenson, W. G. (1990). Type IV collagenases in tumor invasion and
metastasis. Cancer Metast. Rev. 9, 289-303.
Taraboletti, G., Garofalo, A., Belotti, D., Drudis, T., Borsotti, P., Scanziani,
E., Brown, P. D. and Giavazzi, R. (1995). Inhibition of angiogenesis and
murine hemangioma growth by batimastat, a synthetic inhibitor of matrix
metalloproteinases. J. Nat. Cancer Inst. 87, 293-298.
Uria, J. A., Jimenez, M. G., Balbin, M., Freije, J. M. P. and Lopez-Otin,
C. (1998). Differential effects of transforming growth factor-β on the
expression of collagenase-1 and collagenase-3 in human fibroblasts. J. Biol.
Chem. 273, 9769-9777.
Vu, T. H. and Werb, Z. (1998). Gelatinase B: structure, regulation and
function. In Matrix Metalloproteinases (ed. W. C. Parks and R. P. Mecham),
pp. 115-148. San Diego: Academic Press.
Vu, T. H., Shipley, J. M., Bergers, G., Berger, J. E., Helms, J. A., Hanahan,
D., Shapiro, S. D., Senior, R. M. and Werb, Z. (1998). MMP-9/gelatinase
B is a key regulator of growth plate angiogenesis and apoptosis of
hypertrophic chondrocytes. Cell 93, 411-422.
Wang, W., Zhou, G., Hu, M. C. T., Yao, Z. and Tan, T. H. (1997). Activation
of the hematopoietic progenitor kinase-1 (HPK1)-dependent, stress-
activated c-Jun N-terminal kinase (JNK) pathway by transforming growth
factor β (TGF-β)-activated kinase (TAK1), a kinase mediator of TGFβ
signal transduction. J. Biol. Chem. 272, 22771-22775.
Welch, D. R., Fabra, A. and Nakajima, M. (1990). Transforming growth
factor β stimulates mammary adenocarcinoma cell invasion and metastatic
potential. Proc. Nat. Acad. Sci. USA 87, 7678-7682.
Yang, E. Y. and Moses, H. L. (1990). Transforming growth factor β1-induced
changes in cell migration, proliferation, and angiogenesis in the chicken
chorioallantoic membrane. J. Cell Biol. 111, 731-741.
A. Puyraimond and others
