|
|
|
|
|
|
|
|
|||||
| Home Help Feedback Subscriptions Archive Search Table of Contents | |||||
First published online 22 February 2005
doi: 10.1242/jcs.01708
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Article |
3ß1 integrin regulates MMP-9 mRNA stability in immortalized keratinocytes: a novel mechanism of integrin-mediated MMP gene expression
Center for Cell Biology and Cancer Research, Albany Medical College, MC-165, 47 New Scotland Avenue, Albany, New York 12208, USA
* Author for correspondence (e-mail: dipersm{at}mail.amc.edu)
Accepted 6 January 2005
 |
Summary |
|---|
 Top Summary IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
3ß1 is expressed at high levels in squamous cell carcinomas and in normal keratinocytes during cutaneous wound healing. We showed previously that 3ß1 is required for MMP-9/gelatinase B secretion in immortalized mouse keratinocytes (MK cells) and that this regulation was acquired as part of the immortalized phenotype, suggesting a possible role for 3ß1 during malignant conversion. In the current study, we identify a novel mechanism whereby 3ß1 regulates the induction of MMP-9 expression that occurs in response to activation of a MAPK kinase (MEK)/extracellular signal-regulated kinase (ERK) pathway. Inhibition of MEK/ERK signaling in wild-type MK cells with a pharmacological inhibitor, U0126, showed that ERK activation was necessary for high levels of endogenous MMP-9 gene expression and activity of a transfected MMP-9 promoter. Furthermore, activation of MEK/ERK signaling in these cells with an oncogenic mutant of Ras, RasV12, increased both endogenous MMP-9 gene expression and MMP-9 promoter activity. Experiments with 3ß1-deficient MK cells revealed that 3ß1 was required for both baseline levels and RasV12-induced levels of MMP-9 mRNA expression. However, 3ß1 was not required for RasV12-mediated activation of ERK or for ERK-dependent MMP-9 promoter activity. Direct comparison of mRNA turnover in the wild type and 3-null MK cells identified a requirement for 3ß1 in stabilization of MMP-9 mRNA transcripts. These results identify a novel function for integrins in promoting mRNA stability as a mechanism to potentiate MAPK-mediated gene expression. They also suggest a role for 3ß1 in maintaining high levels of MMP-9 mRNA expression in response to oncogenic activation of MEK/ERK signaling pathways.
Key words: 3ß1 integrin, MMP-9, Keratinocytes, mRNA stability
|
Introduction |
|---|
|
TopSummaryIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). Matrix metalloproteinases (MMPs) are secreted into the extracellular space as zymogens that are functionally activated through self-proteolytic cleavage or through proteolysis by other MMPs or proteinases (Werb, 1997). Activated MMPs play important roles in promoting tumor growth, progression and invasion. They can be expressed either directly by tumor cells (Canete-Soler et al., 1994; Jiang and Muschel, 2002; Pyke et al., 1992) or by other cells present in the tumor microenvironment, such as stromal cells or infiltrating inflammatory cells (Coussens et al., 2000; Werb, 1997).
MMP-9/gelatinase B belongs to the gelatinase subgroup of the MMP family and its expression and activity has been correlated with different stages of carcinoma progression. For example, MMP-9 has been shown to induce the angiogenic switch in a mouse model of pancreatic carcinogenesis (Bergers et al., 2000). Other studies have implicated MMP-9 gene activation in promoting tumor invasion and metastasis (Bernhard et al., 1994; Kupferman et al., 2000). Many potential substrates have been identified for MMP-9 that could contribute to its tumor-promoting functions, including numerous ECM proteins, other MMPs and proteinase inhibitors (Liu et al., 2000; Werb, 1997).
MMP-9 expression can be regulated at several levels. Although most published studies have focused on transcriptional control of MMP-9 (reviewed by Westermarck and Kahari, 1999), there is increasing evidence that its expression can also be regulated at the steps of mRNA stability, translation and protein secretion (Akool et al., 2003; Eberhardt et al., 2002; Jiang and Muschel, 2002; Morini et al., 2000; Sehgal and Thompson, 1999; Thant et al., 2000). The ability to modulate MMP-9 expression at multiple steps through distinct signaling pathways may be particularly important during malignant conversion and metastasis, when tumor cells need to induce or maintain MMP-9 levels in response to changing environmental cues. A number of distinct extracellular stimuli, including growth factors, cytokines and extracellular calcium, have been shown to regulate MMP-9 expression in normal or malignant keratinocytes via the mitogen-activated protein kinases (MAPKs) p38 and/or extracellular signal-regulated kinase (ERK) (Johansson et al., 2000; McCawley et al., 1998; Mukhopadhyay et al., 2004; Westermarck and Kahari, 1999; Zeigler et al., 1999). In addition, several studies have demonstrated the importance of cell adhesion in regulating the expression of MMP-9 or other MMPs in immortalized epithelial cells or carcinoma-derived cells (DiPersio et al., 2000; Lochter et al., 1999; Morini et al., 2000; Thomas et al., 2001).
Integrins are the major receptors for cell adhesion to the ECM and many integrins initiate `outside-in' signal transduction events that modulate cell functions such as gene expression, cell migration and invasion (Hynes, 2002). 3ß1 integrin is expressed at high levels on epithelial cells where it is a major receptor for laminin-5 (LN-5) and other basement membrane laminins (Kreidberg, 2000). Knockout studies in mice have revealed important roles for 3ß1 in maintaining basement membrane integrity during embryonic development of the epidermis and other tissues (DiPersio et al., 1997; Kreidberg et al., 1996). 3ß1 is also expressed at high levels in most primary and metastatic tumors (Bartolazzi et al., 1994; Natali et al., 1993; Patriarca et al., 1998) and several recent studies in immortalized epithelial cells or carcinoma-derived cell lines suggest that this integrin has important roles in regulating functions that promote malignant tumor growth and progression, including cell survival, migration, invasion and metastasis (Morini et al., 2000; Choma et al., 2004; Wang et al., 2004; Manohar et al., 2004). Some of these 3ß1-mediated cell functions may involve MMP-9, as 3ß1 regulates MMP-9 secretion in some immortalized and malignantly transformed epithelial cells (Morini et al., 2000; DiPersio et al., 2000). We showed previously that 3ß1-dependent secretion of MMP-9 was acquired as part of the immortalized phenotype in keratinocytes (DiPersio et al., 2000), consistent with a role during malignant conversion of squamous cell carcinoma (SCC). Many invasive carcinoma cells also express high levels of LN-5 (Pyke et al., 1995) and MMP-9 (Juarez et al., 1993; Morini et al., 2000; Pyke et al., 1992; Westermarck and Kahari, 1999), reinforcing a possible link between 3ß1-mediated adhesion and MMP-9 induction during tumor progression and invasion.
Despite the established importance of integrins in regulating MMP gene expression in immortalized or transformed epithelial cells, the mechanisms of this regulation are still unclear. In the current study, we determined the mechanism whereby 3ß1 integrin regulates the induction of MMP-9 gene expression in immortalized keratinocytes cultured on LN-5. We showed that ERK activation was necessary for high levels of MMP-9 expression and that 3ß1 was necessary for ERK-mediated induction of MMP-9 in response to oncogenic RasV12. Importantly, although 3ß1 was required for high levels of MMP-9 mRNA, it was not required for RasV12-mediated activation of ERK, or for ERK-mediated activation of a transfected MMP-9 promoter construct. However, 3ß1 was required post-transcriptionally to maintain high levels of MMP-9 mRNA, as mRNA decay experiments indicated that MMP-9 transcripts turn over more rapidly in 3-null cells than in 3ß1-expressing cells. These results identify a novel function for 3ß1 integrin in promoting mRNA stability as a mechanism to potentiate MAPK-mediated gene expression. In addition, our findings suggest an important role for 3ß1 integrin in augmenting the response of some immortalized or transformed cells to growth factors and other environmental cues, or to activated oncogenes, that stimulate MMP-9 gene expression through MAPK signaling pathways.
|
Materials and Methods |
|---|
|
TopSummaryIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
3 integrin3 integrin knockout mice, respectively (DiPersio et al., 2000). MK growth medium consisted of Eagle's minimum essential medium (EMEM; BioWhittaker, Walkersville, MD) supplemented with 4% FBS (BioWhittaker) from which Ca2+ had been chelated, 0.05 mM CaCl2, 0.4 µg/ml hydrocortisone (Sigma, St Louis, MO), 5 µ g/ml insulin (Sigma), 10 ng/ml EGF (Invitrogen Corporation, Carlsbad, CA), 2x10–9 M T3 (Sigma), 10 U/ml interferon 
(INF; Sigma), 100 U/ml penicillin and 100 µg/ml streptomycin (Invitrogen) and L-glutamine (Invitrogen). MK cell lines were maintained at 33°C, 8% CO2, on tissue culture plates coated with 30 µg/ml denatured rat tail collagen (Cohesion, Palo Alto, CA). For experiments, MK cells were sub-cultured on laminin-5-rich extracellular matrix (LN-5 ECM) prepared from the human squamous cell carcinoma line SCC-25 (Rheinwald and Beckett, 1981), as described previously (DiPersio et al., 2000). For experiments in Fig. 4, untreated or LN-5 ECM coated tissue culture plates were coated with fibronectin (20 µg/ml; a gift from Livingston Van De Water, Albany Medical College), vitronectin (15 µg/ml; a gift from Paula McKeown-Longo, Albany Medical College), or both fibronectin and vitronectin (20 µg/ml and 15 µg/ml, respectively), then blocked with 10 mg/ml heat-denatured BSA in PBS. For assessment of cell spreading and F-actin, parallel cultures were grown under serum-free conditions for 24 hours on glass coverslips coated with ECM proteins as described above, fixed in 4% paraformaldehyde and stained with TRITC-conjugated phalloidin (Sigma). Images were acquired on an Olympus BX60 microscope using a cooled CCD camera (Sensicam) and the software Slidebook 3.0.
|
3ß1 integrin expression was restored in MK–/– cells by stable transfection with the plasmid pCMVzeo-h3A, which contains the full-length human 3 cDNA (a generous gift from Martin Hemler, Dana-Farber Cancer Institute, Boston, MA) cloned into the XbaI site proximal to the CMV promoter in pcDNA3.1/Zeo(+) (Invitrogen). MK–/– cells were transfected using lipofectamine reagent (Invitrogen). Pools of zeocin-resistant cells were expanded and assayed for 3ß1 surface expression by FACS analysis with the anti-human 3 monoclonal antibody P1B5 (Gibco/BRL, Gaithersburg, MD), as described previously (DiPersio et al., 2000).
Western blotting
MK cell lysates were prepared in cell lysis buffer (Cell Signaling Technology, Beverly, MA) and equal amounts of protein from each lysate (either 10 µg or 20 µg) was subject to reducing 10% SDS-PAGE. Proteins were then transferred to nitrocellulose membranes and assayed by western blot. Primary antibodies were used at the following concentrations: rabbit anti-phosphorylated ERK1/2 (Cell Signaling Technology), 1:1000; rabbit anti-ERK1/2 (Promega, Madison, WI), 1:5000; mouse monoclonal anti-HA-tag (Covance Inc., Princeton, NJ), 1:1000. Peroxidase (HRP)-conjugated secondary antibodies were used at the following concentrations: goat anti-rabbit IgG (Cell Signaling Technology), 1:2000; goat anti-rabbit IgG (Pierce, Rockford, IL), 1:15,000; goat anti-mouse IgG (Pierce), 1:15,000. Chemiluminescence was performed with the SuperSignal Kit (Pierce).
Analysis of MMP-9 expression by gelatin zymography
MMP-9 protein expression was assayed essentially as described previously (DiPersio et al., 2000). Briefly, MK cells were plated onto 12-well tissue culture plates coated with LN-5 ECM at 1.4x105 cells per well and allowed to attach overnight in MK growth medium. Cultures were then rinsed and maintained in serum-free medium for an additional 36 or 48 hours, as indicated in figure legends, and culture media were collected for zymography. In some experiments, cells were treated with the MEK inhibitor U0126 (10 µm), as described in the figure legends. All zymography experiments were performed in the absence of INF, since INF has been reported to inhibit MMP-9 expression (Hujanen et al., 1994). The same zymography results were obtained whether cells were grown in serum-free medium containing or lacking additional hormonal supplements that are normally present in MK growth medium (see above). MMPs were concentrated from culture media by binding overnight at 4°C to gelatin-agarose beads (Sigma). Agarose beads were recovered by centrifugation, and bound MMPs were eluted in zymography sample buffer (2.25% SDS, 9% glycerol, 45 mM Tris-HCl, pH 6.8, Bromophenol Blue) and resolved by non-reducing SDS-PAGE on 10% polyacrylamide gels impregnated with 1 mg/ml gelatin (Sigma). Gels were processed for zymography as described previously (DiPersio et al., 2000).
Analysis of MMP-9 mRNA
For northern blots, MK cells were cultured on LN-5 ECM for 4 days in serum-containing growth medium, and total cellular RNA was isolated using the Purescript RNA isolation kit (Gentra Systems, Minneapolis, MN). 10 µg total RNA was denatured at 55°C for 15 minutes in 1x MOPS, 6.5% formaldehyde and 50% formamide prior to electrophoresis on agarose-formaldehyde gels (1.2% agarose, 1.1% formaldehyde, 1x MOPS). RNA was then transferred to Nytran membranes in 10x SSC (3 M NaCl, 0.3 M sodium citrate, pH 7.0), UV cross-linked and incubated for 3 hours at 42°C in pre-hybridization buffer (50% formamide, 5x Denhardt's solution, 1% SDS, 150 µg/ml salmon sperm DNA, 5x SSC). Blots were then incubated overnight at 42°C in hybridization buffer (pre-hybridization buffer with 2.5x Denhardt's solution and 10% dextran sulfate) containing a 32P-labeled cDNA probe against mouse MMP-9 (a generous gift from Karl Tryggvason, University of Oulu, Finland), washed and exposed to autoradiographic film. Blots were stripped and reprobed with a cDNA probe against A50 ribosomal protein as a control, as its expression does not change in response to cell adhesion (Samarakoon and Higgins, 2002).
For RT-PCR, total RNA was isolated as described above and reverse transcribed to produce cDNA template using the first-strand cDNA synthesis kit (Promega). PCR reactions were carried out in 12.5 µl PCR REDTaq ReadyMix (Sigma) with 5 µl cDNA and 0.4 µM of each primer. PCR conditions were 94°C for 60 seconds, 58°C for 90 seconds and 72°C for 90 seconds. Samples were subjected to 28 cycles for MMP-9 and 15 cycles for ß-actin, with the last cycle for each at 72°C for 7 minutes. PCR primers for MMP-9 were as described (Bouloumie et al., 2001) and generate a 371 bp product: forward primer, 5'-TGTACCGCTATGGTTACAC-3'; reverse primer, 5'-CCGCGACACCAAACTGGAT-3'. An alternative set of MMP-9 primers was used for certain experiments, as described (Kim et al., 2000) and generates a 754 bp product: forward primer, 5'-AGTTTGGTGTCGCGGAGCAC-3'; reverse primer, 5'-TACATGAGCGCTTCCGGCAC-3'. PCR conditions for these primers were similar to those described above, except that the annealing temperature was increased to 64°C for 60 seconds. PCR primers for ß-actin were as described (Li et al., 2002) and generate a 450 bp product: forward primer, 5'-AGGGAAATCGTGCGTGACAT-3'; reverse primer, 5'-CATCTGCTGGAAGGTGGACA-3'. PCR products were run on a 1.8% agarose gel, stained with ethidium bromide and visualized using a Bio-Rad Gel-Doc 2000. Signals were quantified using Quantity One software (Bio-Rad, Hercules, CA). The number of PCR cycles was optimized so that signals were within the linear range of detection for both MMP-9 and ß-actin.
For assaying mRNA turnover, cells cultured on LN-5 ECM in serum-free MK growth medium were pre-treated with 10 µg/ml cycloheximide for 16 hours to promote accumulation of MMP-9 mRNA, as described (Huang et al., 2000). Cycloheximide was then removed and 4 hours later cells were switched to serum-free MK growth medium containing 10 µg/ml actinomycin D to inhibit transcription. Control experiments demonstrated that this concentration of actinomycin D effectively inhibited transcription from the MMP-9 promoter in MK cells (data not shown). The initial time point was collected 30 minutes after actinomycin D treatment. Subsequent time points were collected as indicated in the figure legend and assayed for MMP-9 mRNA and ß-actin mRNA levels by RT-PCR.
Adenoviral infection of MK cells
MK cells were sub-cultured on LN-5 ECM in growth medium overnight and then infected for 24 hours with adenovirus expressing either HA-tagged RasV12 or ß-galactosidase as a control. RasV12 and ß-galactosidase were cloned into pAdTrack, as described (Meadows et al., 2001). Multiplicity of infection (MOI) for each experiment is provided in the figure legends. In some experiments, cells were pre-treated for 3 hours with 10 µM U0126 (Calbiochem, San Diego, CA), or with DMSO as a control, then cultured in serum-free medium for an additional 24-48 hours in the presence or absence of U0126, as indicated; for zymography experiments, U0126 was replenished in the medium every 12 hours. Culture media were assayed by zymography and cell lysates were assayed by western blotting, as described above.
Analysis of the transfected MMP-9 promoter
An MMP-9 promoter/firefly luciferase reporter plasmid was a generous gift from Y. Sasaguri (University of Occupational and Environmental Health, Kitakyushu, Japan) and contained a 1868 bp DNA fragment of the MMP-9 promoter region (–1879 to –12 from the transcription start site) cloned upstream of the luciferase gene in the pGL3 vector (Promega), as described (Shimajiri et al., 1999). Uninfected MK cells, or MK cells infected with RasV12 adenovirus (see above), were co-transfected with the MMP-9 promoter/luciferase reporter plasmid and a TK promoter/Renilla luciferase control plasmid (pRLTK, Promega) at a 50:1 ratio using lipofectamine. Following a 5 hour transfection, cells were cultured in MK growth medium without serum (unless otherwise indicated), in the presence or absence of 10 µM U0126 for an additional 24 hours. Cell lysates were then collected and assayed for luciferase expression using the Dual-Luciferase Reporter Assay Kit (Promega). Luciferase expression was measured in a TD-20/20 luminometer (Turner Designs) and expression from the MMP-9 promoter/luciferase plasmid was normalized to that from the control pRLTK plasmid for each sample. Normalized luciferase signals were then plotted as described in the legend to Fig. 5.
|
|
Results |
|---|
|
TopSummaryIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
3ß1 integrin in immortalized mouse keratinocytes (MK cells) cultured on laminin 5-rich extracellular matrix (LN-5 ECM) and this mode of regulation was acquired as part of the immortalized phenotype (DiPersio et al., 2000). ERKs are phosphorylated and activated by MEKs, and MEK/ERK pathways have been shown to regulate MMP-9 gene expression in a number of cell types (Gum et al., 1997; McCawley et al., 1999; Zeigler et al., 1999; Genersch et al., 2000). To determine whether MEK/ERK signaling is required for MMP-9 expression in MK cells, wild-type MK+/+ cells cultured on LN-5 ECM were treated with the MEK-specific inhibitor, U0126, and gelatin zymography was used to assess levels of MMP-9 secreted into the culture medium. Inhibition of MEK caused decreased levels of MMP-9 production by MK+/+ cells (Fig. 1). We showed previously that expression of MMP-9 protein was absent from 3ß1-deficient MK–/– cells, but was restored in these cells by transfection with the human 3 subunit (DiPersio et al., 2000). This 3ß1-mediated restoration of MMP-9 expression was also blocked with the MEK inhibitor (data not shown). To confirm that MEK-dependent ERK activation was inhibited effectively by U0126 in these experiments, lysates from the same cells were immunoblotted for the activated forms of ERK1/2 with an antibody specific for phosphorylation on residues Thr202 and Tyr204. Levels of activated ERK were reduced in U0126-treated cells (Fig. 1, pERK blot), whereas levels of total ERK protein were unchanged (Fig. 1, ERK blot). These results demonstrate that signaling through MEK/ERK is necessary for MMP-9 expression in MK cells.
|
Activation of MEK/ERK signaling by oncogenic RasV12 induces MMP-9 protein levels in an 3ß1-dependent manner
As activation of ERK is necessary for MMP-9 expression in the MK+/+ cells, we next tested whether ERK activation is sufficient to induce MMP-9 expression in the 3ß1-deficient MK–/– cells. Ras is a small GTPase that activates the MEK/ERK signaling pathway. Oncogenic mutations in the c-rasHa gene occur frequently in skin tumors (Yuspa, 1994) and Ras activation has been shown to induce expression of MMP-9 (Genersch et al., 2000; Gum et al., 1996). We expressed a constitutively activated mutant of Ras, RasV12, in the MK cells and determined the effects on ERK activation and MMP-9 expression. MK+/+ and MK–/– cells were each infected with an adenovirus expressing either HA-tagged RasV12 or ß-galactosidase as a control. For both MK+/+ cells and MK–/– cells, we did not observe any obvious differences in morphology between RasV12-infected and control-infected cells (data not shown). Gelatin zymography showed that expression of RasV12 in MK+/+ cells induced levels of secreted MMP-9 considerably above basal levels detected in control-infected cells (Fig. 2, mMMP-9, compare lanes 1 and 3). The induction of MMP-9 by RasV12 was MEK/ERK dependent as treatment with U0126 abrogated this response (Fig. 2, mMMP-9, compare lanes 3 and 4). Immunoblots of lysates from the corresponding cell layers confirmed the presence of HA-tagged RasV12 in infected cells (Fig. 2, RasV12, lanes 3 and 4). Immunoblotting for phosphorylated ERK confirmed that RasV12 caused an increase in ERK activation and that this activation was MEK-dependent (Fig. 2, pERK).
|
In stark contrast with the results in MK+/+ cells, RasV12 was unable to induce MMP-9 levels in MK–/– cells (Fig. 2, mMMP-9, compare lanes 5 and 7) suggesting that 3ß1 is required for RasV12-mediated induction of MMP-9. The inability of RasV12 to induce MMP-9 was not due to a defect in ERK activation, since ERK was phosphorylated robustly in RasV12-infected MK–/– cells (Fig. 2, pERK, lane 7). Thus, oncogenic RasV12 was able to activate ERK in both MK–/– cells and MK+/+ cells, but it did not induce MMP-9 secretion in the MK–/– cells, demonstrating that 3ß1 is required for MMP-9 expression at a point downstream of ERK activation.
3ß1 is required for MMP-9 mRNA accumulation
Regulation of MMP-9 expression has been reported to occur at the levels of gene transcription, mRNA stability and protein production/secretion (Akool et al., 2003; Eberhardt et al., 2002; Jiang and Muschel, 2002; Morini et al., 2000; Sehgal and Thompson, 1999; Thant et al., 2000; Westermarck and Kahari, 1999). In order to determine whether 3ß1 regulates MMP-9 mRNA levels, we performed northern blot analysis on total RNA isolated from MK cells cultured on LN-5 ECM in the presence of serum for 4 days, conditions which support similarly high survival of both MK+/+ cells and MK–/– cells (Manohar et al., 2004). Two distinct MMP-9 mRNA transcripts were detected in MK+/+ cells, a major transcript of 
3.5 kb and a minor transcript of 2.7 kb (Fig. 3A, +/+ lane), as described for other cell types (Graubert et al., 1993; Tanaka et al., 1993). However, MMP-9 transcripts were undetectable in MK–/– cells (Fig. 3A, –/– lane) suggesting that 3ß1 is required for MMP-9 mRNA expression. Indeed, restoration of 3ß1 integrin expression in MK–/– cells through stable transfection with the human 3 subunit (see Fig. 3C) restored MMP-9 mRNA expression (Fig. 3A, –/–, 3 lane), whereas transfection with a control vector did not (Fig. 3A, –/–, V lane). Using the more sensitive method of RT-PCR, we confirmed that MMP-9 mRNA was detected in MK+/+ cells and in 3-transfected MK–/– cells, but was barely detectable in 3ß1-deficient MK–/– cells (Fig. 3B).
|
We showed previously that 3-null MK–/– cells cultured on LN-5 ECM fail to spread properly and display reduced actin stress fiber formation (DiPersio et al., 2000; Choma et al., 2004), suggesting that reduced MMP-9 expression in these cells could be caused by a general spreading defect. Indeed, previous studies have shown that MMP expression can be regulated by changes in cell spreading and actin cytoskeletal organization (Kheradmand et al., 1998; Yan et al., 2000). To test whether restoring cell spreading to MK–/– cells can induce MMP-9 expression, MK–/– cells were plated either on a mixture of LN-5 ECM plus fibronectin or on purified fibronectin. MK–/– cells were able to spread on both fibronectin alone (Fig. 4B, c and d) and LN-5 ECM plus fibronectin (data not shown). However, neither of these substrates induced MMP-9 expression above the low basal levels seen on LN-5 ECM alone (Fig. 4A). By contrast, MK+/+ cells showed high levels of MMP-9 mRNA expression on either substrate (Fig. 4A); MMP-9 expression on fibronectin alone was probably due to deposition of endogenous LN-5 (DiPersio et al., 2000).
Although MK–/– cells were able to spread on fibronectin, the degree of spreading was detectably lower in MK–/– cells (Fig. 4B, c and d) than in MK+/+ cells (Fig. 4B, a and b), with a higher proportion of MK–/– cells showing a spindle-shaped appearance (Fig. 4d, arrow) and/or fewer stress fibers. The expression of other endogenous integrins is similar in the MK+/+ cells and MK–/– cells (DiPersio et al., 2000), suggesting that the ability of MK+/+ cells to spread better on fibronectin is most likely due to their ability to adhere to endogenously deposited LN-5 through 3ß1 (Frank et al., 2004; DiPersio et al., 2000). In an attempt to enhance MK–/– cell spreading, we also tested a mixed substrate of fibronectin plus vitronectin. Under these conditions, MK–/– cells showed markedly improved spreading and stress fiber formation that approached that seen in MK+/+ cells (Fig. 4B, e-h), although subtle differences in the number of spindle-shaped cells remained (Fig. 4B, h, arrow). Despite improved spreading, MK–/– cells still showed considerably lower MMP-9 mRNA levels than did MK+/+ cells on fibronectin plus vitronectin (Fig. 4A, F+V). 3ß1-dependent MMP-9 expression was also seen in MK cells cultured on vitronectin alone (Fig. 4A, V). Taken together, our results indicate that 3ß1 is required for MMP-9 mRNA expression and that engagement of other endogenous integrins/ECM receptors and subsequent spreading is not sufficient to restore MMP-9 expression in MK–/– cells. Nevertheless, we cannot rule out a requirement for cell spreading, or a role for subtle aspects of cell spreading or cytoskeletal organization, in the 3ß1-dependent regulation of MMP-9.
3ß1 is not required for MEK/ERK-dependent MMP-9 promoter activity
MEK/ERK signaling has been shown to regulate transcriptional activation of the MMP-9 promoter (Westermarck and Kahari, 1999). Given that 3ß1 was required for MEK/ERK-mediated induction of MMP-9 expression, we next tested whether 3ß1 is also required for MEK/ERK-mediated activation of the MMP-9 promoter. For these experiments, we used an MMP-9 promoter-driven luciferase assay that was shown previously to provide a sensitive measure of transcriptional responsiveness (Shimajiri et al., 1999). MK cells cultured on LN-5 ECM were transfected with a reporter plasmid containing the firefly luciferase gene under transcriptional control of a 1.9 kb MMP-9 promoter fragment that includes the major regulatory elements for responsiveness to growth factors and cytokines (Shimajiri et al., 1999; Westermarck and Kahari, 1999). Interestingly, MMP-9 promoter activity was reduced only slightly in MK–/– cells compared with MK+/+ cells and restoring 3ß1 expression in the former cells did not increase promoter activity (Fig. 5A), indicating that activity of the transfected promoter was independent of 3ß1. Treatment with U0126 reduced MMP-9 promoter activity below basal levels in MK+/+ cells, MK–/– cells and 3-transfected MK–/– cells (Fig. 5B), indicating that MEK/ERK signaling is necessary for baseline MMP-9 promoter function regardless of 3ß1 expression. We have consistently observed reduced levels of phosphorylated ERK in MK–/– cells compared with MK+/+ cells or 3-transfected MK–/– cells, indicating that 3ß1 is required for full ERK activation in these cells (Manohar et al., 2004). Nevertheless, the results presented here show that basal ERK activity in MK–/– cells is both necessary (Fig. 5B) and sufficient (Fig. 5A) for MMP-9 promoter activity.
To determine whether oncogenic Ras can induce MEK/ERK-dependent activation of the MMP-9 promoter, MK cells that had been infected with the RasV12 adenovirus were transfected with the MMP-9 promoter/luciferase reporter plasmid and cultured in the presence or absence of U0126. RasV12 induced the MMP-9 promoter in MK+/+ cells more than threefold over basal levels seen in uninfected cells (Fig. 5B, MK+/+) consistent with the induction of endogenous MMP-9 that was detected in these cells by zymography (Fig. 2). RasV12 also induced the MMP-9 promoter in 3-transfected MK–/– cells (Fig. 5B, 3). MMP-9 promoter activity in RasV12-infected cells was reduced below basal levels upon treatment with U0126 (Fig. 5B), showing that RasV12-mediated induction of the MMP-9 promoter is MEK-dependent. These results suggest that RasV12-mediated induction of MMP-9 secretion in MK+/+ cells is due, at least partly, to activation of a MEK/ERK signaling pathway that induces the MMP-9 promoter.
Importantly, even though RasV12 also induced the MMP-9 promoter in 3 1-deficient MK–/– ß cells (Fig. 5B, MK–/–), there was not a concomitant induction of endogenous MMP-9 secretion by RasV12 in these cells (Fig. 2). These results were confirmed in a separate set of experiments in which transfected MK–/– cells were assayed simultaneously for luciferase expression and for endogenous MMP-9 protein expression (data not shown). Thus, 3ß1 was not required for induction of the transfected MMP-9 promoter by Ras/MEK/ERK signaling, but it was required for endogenous MMP-9 expression in the same cells. As promoter-driven luciferase assays are an indirect measure of a gene's transcriptional activity, our results do not rule out a role for 3ß1 in transcriptional regulation of the endogenous MMP-9 gene. Nevertheless, they prompted us to explore the possibility that 3ß1 has a post-transcriptional role in MMP-9 mRNA accumulation (see below).
3ß1 is required for induction of endogenous MMP-9 mRNA by oncogenic RasV12
As RasV12 was able to induce the MMP-9 promoter independently of 3ß1 (Fig. 5B), but accumulation of endogenous MMP-9 mRNA required 3ß1 (Fig. 3), we next determined whether the ability of RasV12 to induce steady-state levels of endogenous MMP-9 mRNA was dependent on 3ß1. MK+/+ cells and MK–/– cells were cultured on LN-5 ECM and infected with adenovirus expressing either HA-tagged RasV12 or ß-galactosidase as a control. MMP-9 and ß-actin mRNA levels were then assayed by RT-PCR (Fig. 6A) and signals for MMP-9 were normalized to those for ß-actin (Fig. 6B). Although RasV12 had a slight inductive effect on MMP-9 mRNA levels in MK–/– cells, substantial induction by RasV12 was dependent on expression of 3ß1 (Fig. 6B). Taken together, results from Figs 5 and 6 suggest that 3ß1 may potentiate RasV12-mediated induction of MMP-9 mRNA through a mechanism that is downstream from ERK-dependent promoter activation. One possibility is that 3ß1 promotes post-transcriptional mRNA stability, and that increased mRNA turnover in 3-null cells prevents the accumulation of MMP-9 transcripts following transcriptional stimulation by RasV12.
|
3ß1 promotes MMP-9 mRNA stability
To directly compare MMP-9 mRNA stability between 3-null and 3-expressing MK cells, we utilized an established approach to inhibit new gene transcription and measure MMP-9 mRNA decay over time (Sehgal and Thompson, 1999). Owing to the very low basal level of MMP-9 mRNA in the MK–/– cells, it was necessary to first induce MMP-9 mRNA to detectable levels in these cells so that we could monitor the rate of mRNA turnover. Previous studies of mRNA stability have shown that treatment with the translation inhibitor cycloheximide leads to accumulation of unstable mRNAs in some cells, possibly by inhibiting new synthesis of labile destabilization factors that promote mRNA turnover (reviewed in Guhaniyogi and Brewer, 2001). Indeed, in preliminary experiments we observed that MMP-9 mRNA expression in 3-null MK–/– cells was induced tenfold to easily detectable levels by treatment with 10 µg/ml cycloheximide (data not shown). Therefore, cells were pre-treated with cycloheximide to induce accumulation of MMP-9 mRNA in the MK–/– cells. Cells were then cultured in the presence of the transcriptional inhibitor actinomycin D to prevent new mRNA synthesis, allowing us to monitor MMP-9 mRNA decay over a time course as described previously (Sehgal and Thompson, 1999). RT-PCR was used to assess mRNA levels for MMP-9, or for ß-actin as a control. MMP-9 mRNA levels remained high for up to 24 hours in MK+/+ cells (Fig. 7A, diamonds). By contrast, by 24 hours MMP-9 mRNA levels dropped to 37% of original levels in MK–/– cells cultured under identical conditions (Fig. 7A, squares), indicating dramatically decreased mRNA stability in these cells. Restoring expression of 3ß1 in MK–/– cells resulted in increased MMP-9 mRNA levels at each time point, compared with untransfected MK–/– cells (Fig. 7A, open circles). In contrast with MMP-9 mRNA, ß-actin mRNA levels were comparable between 3ß1-expressing MK cells and 3-null MK cells at each time point (Fig. 7B), demonstrating that 3ß1-mediated effects on transcript stability were specific to MMP-9 mRNA. These results identify a novel role for 3ß1 in promoting MMP-9 mRNA stability and they provide a mechanism whereby the integrin can potentiate the induction of MMP-9 expression in response to oncogenic Ras or other activators of MEK/ERK-mediated gene transcription.
|
|
Discussion |
|---|
|
TopSummaryIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
; Morini et al., 2000; Thomas et al., 2001). However, the mechanisms whereby integrins regulate MMP-9 gene expression have been unclear. In the current study, we demonstrated that 3ß1 integrin is required for accumulation of MMP-9 mRNA following MEK/ERK-mediated induction of the MMP-9 gene in immortalized keratinocytes. We propose a model in which 3ß1 potentiates MEK/ERK-induced MMP-9 synthesis by regulating mRNA stability downstream of ERK-mediated promoter activation (Fig. 8). Through this mechanism, 3ß1 potentiates induction of MMP-9 mRNA and protein by activated Ras. Thus, our data identify a novel mechanism whereby integrins can regulate MMP gene induction following the activation of MAPK pathways that stimulate transcription.
|
The acquisition of 3ß1-dependent MMP-9 expression by immortalized keratinocytes may reflect an important role for 3ß1 during malignant progression of SCCs or other tumors, as discussed below. Our findings may also be relevant to other tissue remodeling events in which 3ß1 has been implicated, such as embryonic tissue development or wound healing (Kreidberg, 2000). It seems likely that 3ß1-mediated regulation of MMP-9 expression is controlled by 3ß1 binding to LN-5, as it is well established that 3ß1 is a strong receptor for keratinocyte adhesion to LN-5 (Kreidberg, 2000). Alternatively, or in addition, MMP-9 expression may be controlled by functions of 3ß1 at sites of cell-cell adhesion, as 3ß1 was recently shown to be a functional component of E-cadherin-mediated adherens junctions in kidney collecting duct epithelial cells (Chattopadhyay et al., 2003). Indeed, we recently demonstrated that monolayer cultures of 3ß1-deficient MK–/– cells show reduced cell-cell interactions (Choma et al., 2004) and we have determined that 3ß1 is physically and functionally associated with E-cadherin at MK cell-cell junctions (N. Myneni and M.D., unpublished). Future experiments will address the relative importance of 3ß1 functions at cell-ECM adhesions and cell-cell adhesions in the regulation of MMP-9 expression.
Although most previous studies of MMP-9 gene expression have focused on transcriptional control, it has become increasingly evident that post-transcriptional mechanisms play critical roles in the regulation of MMP-9 mRNA levels in response to growth factors and cytokines (Akool et al., 2003; Sehgal and Thompson, 1999). In the current study we have identified mRNA stabilization as a primary mechanism of 3ß1-mediated MMP-9 induction in immortalized keratinocytes. Although regulation of mRNA stability is clearly an important post-transcriptional mechanism for the control of gene expression (Sachs, 1993), there have been few published studies regarding the roles of integrins and ECM in the modulation of mRNA stability (Feng et al., 1999; Retta et al., 2001; Xu and Clark, 1996) and this mechanism of integrin-mediated gene regulation remains largely unexplored. Regulation of mRNA half-life occurs primarily through conserved U-rich or AU-rich elements, collectively termed AREs, that are usually present in multiple, non-tandem copies within the 3'-untranslated region (3'-UTR) of mRNAs with short or variable half-lives (Brennan and Steitz, 2001). AREs interact with specific RNA-binding proteins (RBPs) that control the rate of mRNA decay. Two of the most extensively studied RBPs are HuR, which promotes mRNA stability (Brennan and Steitz, 2001) and AUF1/hnRNP D, several isoforms of which promote mRNA decay (Loflin et al., 1999). The 3'-UTR of the mouse and rat MMP-9 transcripts each contain several ARE consensus sequences, some of which have been shown to bind HuR and regulate MMP-9 mRNA stability (Graubert et al., 1993; Tanaka et al., 1993; Akool et al., 2003). Future studies should determine whether 3ß1 integrin promotes MMP-9 mRNA stability by regulating the functions of specific RBPs that bind to these AREs.
Although a number of specific signal transduction pathways have been implicated in the regulation of mRNA stability (Brennan and Steitz, 2001), several studies have reported a particularly prominent role for MAPK p38 signaling pathways in ARE-mediated mRNA stability (Montero and Nagamine, 1999; Dean et al., 1999; Reunanen et al., 2002; Winzen et al., 1999; Tran et al., 2003). Although we have not yet identified the signaling pathways whereby 3ß1 controls MMP-9 mRNA stability, we recently showed that several signaling proteins are activated by 3ß1 in MK cells, including focal adhesion kinase (FAK) and the Rho family GTPase Rac1 (Choma et al., 2004; Manohar et al., 2004). Rac1 is a particularly good candidate for mediating MMP-9 mRNA stability, as a Rac1/MKK3/p38 pathway has been shown to promote stability of the uPA mRNA in invasive breast epithelial cells (Han et al., 2002). It remains to be determined whether 3ß1-mediated activation of Rac1 and/or p38 regulates ARE-mediated stability of MMP-9 mRNA.
Previous studies have demonstrated cooperativity between integrins and growth factor receptors in the regulation of MAPK signaling pathways (Giancotti and Ruoslahti, 1999). For example, growth factor-dependent induction of ERK signaling in NIH 3T3 cells is strongly dependent on integrin-mediated cell adhesion (Aplin and Juliano, 1999; Renshaw et al., 1997). In the latter studies, activation of MEK/ERK signaling was identified as an adhesion-dependent event. Our model in Fig. 8 suggests a distinct, novel mechanism whereby integrins can cooperate with growth factors or other stimuli to induce MEK/ERK-dependent gene expression. Specifically, our results in MK cells using RasV12 as an activator of MEK/ERK showed that 3ß1 was not required for Ras-mediated ERK activation, but that it was required for MEK/ERK-mediated induction of MMP-9 expression at a point downstream from ERK activation (i.e. mRNA stabilization). Thus, 3ß1 may regulate the ability of some cells to synthesize and secrete MMP-9 in response to growth factors or cytokines that stimulate MMP-9 gene transcription through MEK/ERK pathways. In addition, our results suggest that 3ß1 expressed in malignant tumors may potentiate MMP-9 induction in response to MAPK pathways that are activated constitutively by oncogenic Ras or other oncogenes. Indeed, activating mutations in the c-rasHa gene are among the most common initiating mutations in epidermal tumors (Yuspa, 1994).
Carcinogenesis is generally accompanied by cellular changes that facilitate tumor growth and cell invasion, including altered ECM synthesis, altered integrin function and increased MMP expression (Johnsen et al., 1998; Werb, 1997; Westermarck and Kahari, 1999). Tumor cells also undergo changes in their capacity to respond to extracellular signals generated from growth factors or ECM, and they are likely to acquire certain signaling pathways to induce or maintain MMP expression. As discussed above, the ability of 3ß1 integrin to potentiate MEK/ERK-mediated induction of MMP-9 through mRNA stabilization may reflect an important role for this integrin in regulating the ability of some tumor cells to respond to environmental cues or oncogenes that stimulate gene transcription through MEK/ERK signaling pathways. This ability of 3ß1 to induce MMP-9 expression is acquired as part of the immortalized keratinocyte phenotype (DiPersio et al., 2000). Similarly, the ability of TGFß to induce MMP-9 expression in prostate cancer cells is acquired during cellular transformation (Sehgal et al., 1996; Sehgal and Thompson, 1999). In addition, changes in calcium-mediated regulation of MMP-9 expression occur during malignant transformation of oral keratinocytes (Mukhopadhyay et al., 2004). Thus, it is clear that certain signaling pathways that regulate MMP-9 expression are activated or altered during cellular immortalization or malignant progression. Many invasive tumors show increased expression levels of 3ß1 (Bartolazzi et al., 1994; Natali et al., 1993; Patriarca et al., 1998) and its ECM ligand laminin-5 (Pyke et al., 1995) and recent studies in immortalized keratinocytes and tumor-derived cell lines have demonstrated important roles for 3ß1 in several cell functions that contribute to tumor growth and malignant progression, including migration, invasion, metastasis and survival (Choma et al., 2004; Manohar et al., 2004; Morini et al., 2000; Tsuji et al., 2002; Wang et al., 2004). As MMP-9 has been implicated in some of these cell functions, our current data suggest that part of the role of 3ß1 may be to regulate the stability and degradation of MMP-9 mRNA. The acquisition by immortalized cells of 3ß1-dependent mRNA stability may extend to other mRNA transcripts as well and could represent an important selection step during tumor progression. Therefore, further study of the signal transduction pathways involved in this regulation may lead to the identification of therapeutic targets for inhibiting carcinoma progression and/or metastasis that are specific to cancer cells.
|
Acknowledgments |
|---|
|
References |
|---|
|
TopSummaryIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Akool, el-S., Kleinert, H., Hamada, F. M., Abdelwahab, M. H., Forstermann, U., Pfeilschifter, J. and Eberhardt, W. (2003). Nitric oxide increases the decay of matrix metalloproteinase 9 mRNA by inhibiting the expression of mRNA-stabilizing factor HuR. Mol. Cell Biol. 23, 4901-4916.
Aplin, A. E. and Juliano, R. L. (1999). Integrin and cytoskeletal regulation of growth factor signaling to the MAP kinase pathway. J. Cell Sci. 112, 695-706.[Abstract]
Bartolazzi, A., Cerboni, C., Nicotra, M. R., Mottolese, M., Bigotti, A. and Natali, P. G. (1994). Transformation and tumor progression are frequently associated with expression of the 3ß1 heterodimer in solid tumors. Int. J. Cancer 58, 488-491.[Medline]
Bergers, G., Brekken, R., McMahon, G., Vu, T. H., Itoh, T., Tamaki, K., Tanzawa, K., Thorpe, P., Itohara, S., Werb, Z. et al. (2000). Matrix metalloproteinase-9 triggers the angiogenic switch during carcinogenesis. Nat. Cell Biol. 2, 737-744.[CrossRef][Medline]
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. Natl. Acad. Sci. USA 91, 4293-4297.
Bouloumie, A., Sengenes, C., Portolan, G., Galitzky, J. and Lafontan, M. (2001). Adipocyte produces matrix metalloproteinases 2 and 9: involvement in adipose differentiation. Diabetes 50, 2080-2086.
Brennan, C. M. and Steitz, J. A. (2001). HuR and mRNA stability. Cell. Mol. Life Sci. 58, 266-277.[CrossRef][Medline]
Canete-Soler, R., Litzky, L., Lubensky, I. and Muschel, R. J. (1994). Localization of the 92 kd gelatinase mRNA in squamous cell and adenocarcinomas of the lung using in situ hybridization. Am. J. Pathol. 144, 518-527.[Abstract]
Chattopadhyay, N., Wang, Z., Ashman, L. K., Brady-Kalnay, S. M. and Kreidberg, J. A. (2003). alpha3beta1 integrin-CD151, a component of the cadherin-catenin complex, regulates PTPmu expression and cell-cell adhesion. J. Cell Biol. 163, 1351-1362.
Choma, D. P., Pumiglia, K. and DiPersio, C. M. (2004). Integrin 3ß1 directs the stabilization of a polarized lamellipodium in epithelial cells through activation of Rac1. J. Cell Sci. 117, 3947-3959.
Coussens, L. M., Tinkle, C. L., Hanahan, D. and Werb, Z. (2000). MMP-9 supplied by bone marrow-derived cells contributes to skin carcinogenesis. Cell 103, 481-490.[CrossRef][Medline]
Dean, J. L., Brook, M., Clark, A. R. and Saklatvala, J. (1999). p38 mitogen-activated protein kinase regulates cyclooxygenase-2 mRNA stability and transcription in lipopolysaccharide-treated human monocytes. J. Biol. Chem. 274, 264-269.
DiPersio, C. M., Hodivala-Dilke, K. M., Jaenisch, R., Kreidberg, J. A. and Hynes, R. O. (1997). 3ß1 integrin is required for normal development of the epidermal basement membrane. J. Cell Biol. 137, 729-742.
DiPersio, C. M., Shao, M., di Costanzo, L., Kreidberg, J. A. and Hynes, R. O. (2000). Mouse keratinocytes immortalized with large T antigen acquire 3ß1 integrin-dependent secretion of MMP-9/gelatinase B. J. Cell Sci. 113, 2909-2921.[Abstract]
Eberhardt, W., Akool el, S., Rebhan, J., Frank, S., Beck, K. F., Franzen, R., Hamada, F. M. and Pfeilschifter, J. (2002). Inhibition of cytokine-induced matrix metalloproteinase 9 expression by peroxisome proliferator-activated receptor alpha agonists is indirect and due to a NO-mediated reduction of mRNA stability. J. Biol. Chem. 277, 33518-33528.
Feng, X., Clark, R. A., Galanakis, D. and Tonnesen, M. G. (1999). Fibrin and collagen differentially regulate human dermal microvascular endothelial cell integrins: stabilization of alphav/beta3 mRNA by fibrin1. J. Invest. Dermatol. 113, 913-919.[CrossRef][Medline]
Frank, D. E. and Carter, W. G. (2004). Laminin 5 deposition regulates keratinocyte polarization and persistent migration. J. Cell Sci. 117, 1351-1363.
Genersch, E., Hayess, K., Neuenfeld, Y. and Haller, H. (2000). Sustained ERK phosphorylation is necessary but not sufficient for MMP-9 regulation in endothelial cells: involvement of Ras-dependent and -independent pathways. J. Cell Sci. 113, 4319-4330.[Abstract]
Giancotti, F. G. and Ruoslahti, E. (1999). Integrin signaling. Science 285, 1028-1032.
Graubert, T., Johnston, J. and Berliner, N. (1993). Cloning and expression of the cDNA encoding mouse neutrophil gelatinase: demonstration of coordinate secondary granule protein gene expression during terminal neutrophil maturation. Bloo
