Transforming Growth Factor-β1 Inhibits Cytokine-mediated Induction of Human Metalloelastase in Macrophages*

Matrix metalloproteinases (MMP) have been identified in vulnerable areas of atherosclerotic plaques and may contribute to plaque instability through extracellular matrix degradation. Human metalloelastase (MMP-12) is a macrophage-specific MMP with broad substrate specificity and is capable of degrading proteins found in the extracellular matrix of atheromas. Despite its potential importance, little is known about the regulation of MMP-12 expression in the context of atherosclerosis. In this study, we report that in human peripheral blood-derived macrophages, MMP-12 mRNA was markedly up-regulated by several pro-atherosclerotic cytokines and growth factors including interleukin-1β, tumor necrosis factor-α, macrophage colony-stimulating factor, vascular endothelial growth factor, and platelet-derived growth factor-BB. In contrast, the pleiotropic anti-inflammatory growth factor transforming growth factor-β1 (TGF-β1) inhibited cytokine-mediated induction of MMP-12 mRNA, protein, and enzymatic activity. Analyses of MMP-12 promoter through transient transfections and electrophoretic mobility shift assays indicated that both its induction by cytokines and its inhibition by TGF-β1 depended on signaling through an AP-1 site at −81 base pairs. Moreover, the inhibitory effect of TGF-β1 on MMP-12 was dependent on Smad3. Taken together, MMP-12 is induced by several factors implicated in atherosclerosis. The inhibition of MMP-12 expression by TGF-β1 suggests that TGF-β1, acting via Smad3, may promote plaque stability.

Atherosclerosis leads to myocardial infarction and stroke and is the principal cause of death in Western societies (1). Advanced atherosclerotic lesions consist of a fibrous cap and a lipid core (2,3). Lesions with a thin fibrous cap overlying a large lipid-rich core are more vulnerable to coronary plaque rupture, a key pathophysiologic event often resulting in myo-cardial infarction (4). The fibrous cap is composed of extracellular matrix (ECM) 1 that includes collagens, elastins, and proteoglycans (5). A number of proteolytic enzymes, known as matrix metalloproteinases (MMPs), are also present in atherosclerotic lesions (6,7). Degradation of ECM components by MMPs may promote plaque instability (2,3,7).
Initially identified as a 22-kDa protein with elastolytic activity (20), MMP-12 degrades a range of ECM proteins that includes not only elastin but also type IV collagen, fibronectin, laminin, vitronectin, and basement membrane proteins (21,22). The functional significance of MMP-12 is demonstrated by the failure of peritoneal macrophages from MMP-12 Ϫ/Ϫ mice to penetrate basement membranes (23). Human and mouse MMP-12 undergo self-activation through autolytic processing (16), and recombinant rabbit MMP-12 activates MMP-2 and MMP-3 (24,25). These observations suggest that once MMP-12 is expressed there is a cascade of MMP activation that leads to degradation of nearly all the extracellular components found in atherosclerotic plaques. Despite the importance of MMP-12 in conditions that range from atherosclerosis to aneurysm formation, arthritis, and emphysema (26,27), the molecular basis of MMP-12 gene expression in these conditions is incompletely understood.
The pleiotropic anti-inflammatory growth factor transforming growth factor-␤1 (TGF-␤1) belongs to a family of growth factors that has diverse effects on cellular differentiation, activation, and proliferation (28). The downstream effects of TGF-␤1 are mediated by a superfamily of proteins termed Smads. TGF-␤ signaling occurs through the formation of a heteromeric receptor complex of two serine/threonine kinase receptors, type I and type II. Upon TGF-␤1 binding, the type II receptor phosphorylates the type I receptor. The type I receptor, once active, may then phosphorylate and activate Smads, which act as intracellular effectors by translocating to the nucleus to direct transcriptional responses (29). Of the three general classes of Smads, the pathway-restricted (or receptorregulated) Smads, Smad2 and Smad3, become phosphorylated in the cytoplasm by the type I receptor in response to TGF-␤1. Smad2 and Smad3 can then partner with Smad4, the common Smad, and translocate to the nucleus where they can regulate transcriptional responses (29). The third type of Smads are the so-called inhibitory Smads that include Smad6 and Smad7. These Smads may exert their inhibitory function by binding to the TGF-␤1 type I receptor and preventing the phosphorylation of the receptor-regulated Smads, Smad2 and Smad3, thereby blocking downstream signaling (28).
In human alveolar macrophages, Shapiro et al. (24) found that MMP-12 mRNA is induced by lipopolysaccharide and inhibited by dexamethasone. To investigate mechanisms of MMP-12 expression that might apply in atherosclerosis, we studied the effect of cytokines and growth factors known to be present in atherosclerotic plaques on MMP-12 expression by human peripheral blood-derived macrophages. We identified several cytokines and growth factors that induced MMP-12 expression. This induction was prevented by treatment with TGF-␤1. Analysis of the MMP-12 promoter by reporter gene transfection experiments implicated an AP-1 site (bp Ϫ81 to Ϫ75) in mediating cytokine inducibility. Inhibition by TGF-␤1 occurred at least in part by preventing the increase in AP-1 protein binding. Furthermore, this TGF-␤1 inhibitory effect was dependent on signaling through Smad3.

EXPERIMENTAL PROCEDURES
Cell Culture and Reagents-Human peripheral blood monocytes were isolated from buffy coat (Blood Bank, The Children's Hospital, Boston) by the Ficoll-Paque centrifugation technique as described (30), with some modification. In brief, mononuclear cell preparations were washed twice in Hanks' balanced salt solution without calcium or magnesium and then resuspended in Iscove's modified Dulbecco's modified Eagle's medium (Life Technologies, Inc.) supplemented with 10% human serum type AB (Sigma), penicillin (100 units/ml), and streptomycin (100 g/ml). Monocytes were seeded at a density of 1 ϫ 10 7 cells per 150-mm dish and enriched by adherence during a 1-h incubation in a humidified incubator (37°C, 5% CO 2 ). Nonadherent cells were eliminated by three washes with Hanks' balanced salt solution. Cell viability was Ͼ95% as assessed by trypan blue exclusion, and monocyte purity was ϳ90% as determined by nonspecific esterase staining. RAW264.7 cells (American Type Culture Collection) were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, penicillin (100 units/ml), and streptomycin (100 g/ml). Cells were seeded at a density of 0.2 ϫ 10 6 cells/ml. All cytokines were stored at Ϫ80°C until use. Cytokines dissolved in phosphate-buffered saline (Life Technologies, Inc.), as described by the manufacturer, included recombinant human basic fibroblast growth factor (bFGF) (Collaborative Biomedical, Bedford, MA), platelet-derived growth factor-BB RNA Extraction and RNA Blot Analysis-Total RNA was isolated from cultured cells by guanidinium isothiocyanate extraction and centrifugation through cesium chloride (31). RNA was fractionated on a 1.3% formaldehyde-agarose gel and transferred to nitrocellulose filters. The filters were hybridized with 32 P-labeled, random-primed cDNA probes as described (32,33). Hybridized filters were washed in 30 mM sodium chloride, 3 mM sodium citrate, and 0.1% SDS at 55°C and autoradiographed on Kodak XAR film at Ϫ80°C. The blots were hybridized with an 18 S oligonucleotide probe as a qualitative indicator of loading. The filters were scanned, and radioactivity was measured on a PhosphorImager running the ImageQuant software (Molecular Dynamics, Sunnyvale, CA).
Western Blot Analysis-After conditioned medium had been centrifuged (500 ϫ g for 15 min at 4°C), concentrated (25 times) by lyophilization, and solubilized in sample buffer, electrophoresis was performed under reducing conditions according to the method of Laemmli (34). Samples were resolved through a 12% SDS-polyacrylamide gel and transferred to nitrocellulose membranes (Schleicher & Schuell) according to the method of Towbin et al. (35). Membranes were blocked in 0.025 M Tris, pH 8, 0.125 M NaCl, and 0.1% Tween 20 (TBST) containing 4% nonfat dry milk and hybridized with rabbit polyclonal anti-MMP-12 serum (1:1000 dilution, Triple Point Biologics). After three washes (20 min each) in TBST, the blots were incubated with horseradish peroxidase-conjugated donkey anti-rabbit secondary antibody (1:4000, Amersham Pharmacia Biotech) for 40 min. After three final washes (20 min each) in TBST, MMP-12 in the blots was visualized by an enhanced chemiluminescence method (Amersham Pharmacia Biotech).
Zymography-MMP-12 enzymatic activity was determined by ␣-casein zymography as described (36), with some modification. In brief, samples of concentrated conditioned medium were resolved through 12% SDS-polyacrylamide gels containing 2 mg/ml ␣-casein (Sigma) under nonreducing conditions. After electrophoresis, proteins were renatured by washing the gels twice (15 min each time) in 2.5% Triton X-100. The gels were then incubated in substrate buffer (50 mM Tris, pH 7.5, 5 mM CaCl 2 , and 0.02% NaN 3 ) at 37°C for 60 h with gentle shaking and were stained with 0.1% Coomassie Blue R-250. Casein-degrading enzymatic activity was indicated by the appearance of clear bands against a dark background. Prestained molecular weight markers (Amersham Pharmacia Biotech) were used for standardization.
The same reverse primer was used for all the above 5Ј deletion constructs at ϩ39 (5Ј-TAAACTTCTAAACGGATCAAT-3Ј). XhoI and HindIII sites were added to facilitate cloning. PCR cycling conditions for Pfu DNA polymerase were as the manufacturer (Stratagene, La Jolla, CA) suggested. The various DNA fragments were then digested and ligated into the XhoI and HindIII sites of pGL3-basic (Promega, Madison, WI) to produce (Ϫ1046/ϩ39), (Ϫ362/ϩ39), (Ϫ95/ϩ39), and (Ϫ75/ϩ39) pGL3-MMP12-luciferase. Reporter constructs containing fragments of the human MMP-12 5Ј-flanking sequence were named according to the location of the fragment from the transcription start site in the 5Ј direction.
The AP-1 site (Ϫ81 to Ϫ75 bp) within the MMP-12-luciferase reporter construct (Ϫ1046 to ϩ39) was mutated by overlapping PCR mutagenesis. First, two fragments were generated. Fragment A from Ϫ1046 to Ϫ73 was created by PCR with the same forward primer at Ϫ1046 and an internal primer (antisense) containing the AP-1 mutation (underlined) 5Ј-GAGTGCCTCGTAGTTGATATCATCCC-3Ј. Fragment B (from Ϫ84 to ϩ39) was created by using an internal primer (sense) containing the AP-1 mutation 5Ј-CTACGAGGCACTCATAG-GATTC-3Ј and the same reverse primer at ϩ39. The two PCR products were then used as templates for PCR with the primers at Ϫ1046 and ϩ39, which contained XhoI and HindIII sites, respectively, to facilitate cloning. The resulting PCR product was then digested with XhoI and HindIII and subcloned into the same sites of pGL3-basic to generate pGL3-MMP-12-(AP-1mut)-luciferase.
MMP-12 cDNA probes of ϳ1.2 kilobase pairs were generated by reverse transcription-PCR from human macrophage or murine RAW264.7 cell RNA with a forward primer (5Ј-ATGTGCTATTTGGT-GAGAGA-3Ј) and a reverse primer (5Ј-GGAAGTTCTTGGTAATCAGT-3Ј). Smad expressions were all generated in cytomegalovirus-promoter expression constructs. The c-Jun expression construct was obtained from Michael Greenberg (Boston). The dominant-negative Smad3 expression construct (Smad⌬C) was obtained from Rik Derynck (San Francisco). The PAI-1 promoter was obtained from Doug Vaughn (Nashville). Analysis of potential binding sites for sequence-specific transcription factors in the MMP-12 promoter was performed using the MatInspector Program (37). All PCR products were sequenced from both orientations by the dideoxy chain termination method, with Thermosequenase (Amersham Pharmacia Biotech), or on an automated DNA sequencer (Li-Cor, Lincoln, NE) according to the manufacturer's instructions.
Transient Transfections-RAW264.7 cells were transfected with Fugene TM 6 Transfection Reagent (Roche Molecular Biochemicals) on 6-well plates as described by the manufacturer. The total amount of plasmid DNA was kept constant within each experiment. Treatment with TPA (100 nM), TPA plus 30 min pretreatment with TGF-␤1 (10 ng/ml), or control vehicle (ethanol) occurred 12 h after transfection. Cells were harvested for assays of luciferase and ␤-galactosidase activity 24 h later. Luciferase activity was normalized to ␤-galactosidase activity (to correct for differences in transfection efficiency) by cotransfecting pCMV-␤gal plasmid (300 ng) (CLONTECH) in all experiments. Luciferase and ␤-galactosidase assays were performed on transfected cell lysates as described (38). The ratio of luciferase to ␤-galactosidase activity for each transfection was normalized to pGL3-basic and expressed as relative luciferase activity. All transfections were performed in triplicate from at least three independent experiments.

MMP-12 mRNA Is Induced by Cytokines and Growth
Factors-Northern blot analysis was performed with human peripheral blood-derived macrophages that had been differentiated and stimulated with human recombinant IL-1␤, TNF-␣, M-CSF, VEGF, bFGF, PDGF-BB, or TPA for 72 h. MMP-12 mRNA was up-regulated by all of these cytokines and growth factors except bFGF (Fig. 1). These results show that several cytokines and growth factors, commonly found in advanced atherosclerotic plaques, induce MMP-12 expression in human macrophages.
TGF-␤1 Inhibits Cytokine-mediated Induction of MMP-12-Although TGF-␤1 antagonizes the effects of many inflammatory cytokines, its effect on MMP induction by cytokines in human macrophages has not been previously characterized. Northern blot analysis showed that induction of MMP-12 mRNA by IL-1␤ was inhibited by TGF-␤1 ( Fig. 2A) and that this inhibition occurred in a dose-dependent manner (Fig. 2B). Moreover, TGF-␤1(10 ng/ml) inhibited induction of MMP-12 mRNA when human macrophages were stimulated with a mixture of IL-1␤, TNF-␣, and M-CSF or with the phorbol ester TPA ( Fig. 2A). To assess whether the inhibitory effect of TGF-␤1 on MMP-12 mRNA correlated with a change in protein and enzymatic levels, we performed Western blot and zymographic analyses. Conditioned medium was obtained from human macrophages grown under serum-free conditions with vehicle (control), IL-1␤ (20 ng/ml), or IL-1␤ (20 ng/ml) plus TGF-␤1 (10 ng/ml) and analyzed by immunoblotting or casein zymography. By both Western analysis and casein zymography, IL-1␤ induced the active form of MMP-12 (22 kDa), and TGF-␤1 inhibited this induction (Fig. 2C). Thus, TGF-␤1 treatment inhibited cytokine-mediated induction of MMP-12 mRNA and led also to a reduction in the level of MMP-12 protein and enzymatic activity.
Cytokine Induction of MMP-12 and TGF-␤1 Inhibition Occur at the Level of Transcription-MMP-12 mRNA is induced by TPA and inhibited by TGF-␤1 in the RAW264.7 macrophage cell line (Fig. 3A). To elucidate the mechanism underlying MMP-12 mRNA induction, we performed experiments with the transcriptional inhibitor actinomycin D. RAW cells were stimulated for up to 6 h with TPA (100 nM), and actinomycin D (10 g/ml) or vehicle control were added 30 min before stimulation. By Northern analysis (Fig. 3B), actinomycin D blocked induction of MMP-12 mRNA, which suggests that induction occurred at the level of transcription. We have also measured the half-life of MMP-12 stimulated by TPA and found that TGF-␤1 has no effect on its half-life (data not shown). To confirm further that cytokines induce MMP-12 and that inhibition by TGF-␤1 occurs at the level of transcription, we performed reporter gene transfection experiments using a 1-kilobase pair fragment of the human MMP-12 promoter. This promoter has been shown to be active in the macrophage cell line P388D1 (19). Because of the difficulty of transfecting primary human macrophages, we used the macrophage cell line RAW264.7. Treatment of RAW cells with TPA increased activity of the (Ϫ1046/ϩ39) MMP-12 promoter by 5-fold (Fig. 4A,  top). When cells were treated with TGF-␤1 (10 ng/ml), promoter activity was repressed by 50% (Fig. 4A, top). Correcting for basal promoter activity, this correlated with a 67% reduction in cytokine inducibility. Cytokines and growth factors used in Fig. 1 are known to activate AP-1 proteins through a variety of cell signaling pathways. These pathways often involve members of the mitogen-activated protein kinases including Jun N-terminal, extracellular stimulus responsive, and Fos-regulating kinases (57, 58). Indeed, a similar level of induction occurred in response to M-CSF (Fig. 4B). These results suggest that induction of the MMP-12 mRNA and inhibition of induction by TGF-␤1 occurred, at least in part, at the level of transcription.
The AP-1 Site Is Required for Basal and Inducible Activity of the MMP-12 Promoter-We identified multiple potential binding sites for sequence-specific transcription factors by computer-assisted analysis. In addition to one putative binding site for AP-1 identified at position Ϫ81 (from the transcription initiation site), putative binding sites were also identified for a CCAAT box (Ϫ887), three PEA3 sites (Ϫ875, Ϫ690, and Ϫ355), and a Pu.1/Spi.1 site (Ϫ17). To determine which cis-acting elements contribute to MMP-12 promoter activity, we transiently transfected RAW cells with 5Ј deletion constructs of the MMP-12 promoter upstream to a luciferase reporter gene. Be- FIG. 1. MMP-12 mRNA is up-regulated by cytokines and growth factors found in atherosclerotic plaques. Human peripheral blood-derived macrophages were stimulated for 72 h with IL-1␤ (20 ng/ml), TNF-␣ (20 ng/ml), M-CSF (750 units/ml), VEGF (50 ng/ml), bFGF (20 ng/ml), PDGF-BB (20 ng/ml), TPA (75 nM), or vehicle control (Ctrl). Total RNA was isolated, and Northern blot analysis was performed with 10 g of total RNA per lane. After electrophoresis, RNA was transferred to nitrocellulose filters and hybridized to a 32 P-labeled human MMP-12 cDNA probe. Blots were also hybridized with a 32 Plabeled 18 S oligonucleotide probe as a qualitative indicator of loading.
cause deletion of bp Ϫ1046 to Ϫ95 did not significantly affect promoter activity or responsiveness to the presence of TPA or that of TPA plus TGF-␤1 (Fig. 4A), we generated a 5Ј deletion construct removing the fragment from bp Ϫ95 to Ϫ75. The (Ϫ75/ϩ39) construct, which excludes the AP-1 site (bp Ϫ81 to Ϫ75), nearly abolished basal promoter activity (Fig. 4A). In addition, induction of luciferase activity, as obtained with the Ϫ1046, Ϫ362, and Ϫ95 bp MMP-12 promoter constructs, was entirely inhibited. On the basis of these results, we introduced a 2-bp mutation into the AP-1 site of the (Ϫ1046/ϩ39) MMP-12 promoter construct by site-directed mutagenesis. A similar mutation of the AP-1 motif in the human CD44 promoter (TTAGTCA to CTAGGCA) abolishes the function of this site (39). Mutation of the AP-1 site in the MMP-12 promoter abolished activity, even in the presence of TPA (Fig. 4A). These experiments establish the importance of this AP-1 site for basal and inducible MMP-12 promoter activity.
TGF-␤1 Decreases Protein Binding to the MMP-12 AP-1 Site-We assessed nuclear protein interaction with this AP-1 site by electrophoretic mobility shift assay. An oligonucleotide probe containing the putative AP-1 element was incubated with 2 g of nuclear protein from RAW cells that had been treated with TPA for 24 h. Competition studies were performed to confirm that the dominant complex was specific for the AP-1 site. An unlabeled probe containing the putative MMP-12 AP-1 element and a probe with a consensus AP-1 site both competed successfully for binding, whereas a mutated AP-1 oligonucleotide and a nonspecific oligonucleotide probe could not compete for binding (Fig. 5A). To determine whether TGF-␤1 affected AP-1 DNA-protein binding, we incubated the same oligonucleotide probe containing the MMP-12 AP-1 site with 2 g of nuclear protein from RAW cells that had been treated with vehicle, TPA, or TPA plus TGF-␤1. A dominant complex was induced by almost 7-fold by TPA in comparison with control, and TGF-␤1 inhibited this binding by 66% (Fig. 5B). Taken together, these results suggest that the inhibitory effect of TGF-␤1 on MMP-12 promoter activity is mediated, at least in part, through a decrease in nuclear protein binding to the AP-1 site.
Smad3 Inhibits MMP-12 Promoter Activity-To determine which Smad(s) may mediate the inhibitory effect of TGF-␤1 on the MMP-12 promoter, we performed reporter gene transfection experiments using Smads 1-4, 6, and 7. As shown in Fig.  6, only Smad3 repressed the (Ϫ1046/ϩ39) MMP-12 promoter in a similar manner as TGF-␤1. As a positive control we found that Smad3 can transactivate the PAI-1 promoter in RAW cells. To determine which cis-acting elements of the (Ϫ1046/ ϩ39) MMP-12 promoter might be important for Smad3 repres- A and B, total RNA was extracted, and Northern blot analysis was performed with 10 g of total RNA per lane. After electrophoresis, the RNA was transferred to nitrocellulose filters, which were hybridized with 32 P-labeled human MMP-12 cDNA. An 18 S oligonucleotide probe was hybridized as a qualitative indicator of loading. C, Western blot and casein zymographic analyses of MMP-12. Conditioned medium was obtained from human macrophages grown under serum-free conditions for 72 h exposed to no stimulus (ctrl), IL-1␤ (20 ng/ml), or IL-1␤ (20 ng/ml) after being treated for 30 min with TGF-␤1 (10 ng/ml) and subjected to Western blot and casein zymography as described under "Experimental Procedures. "   FIG. 3. MMP-12 mRNA expression in RAW246.7 macrophages. A, RAW246.7 macrophages were treated for 6 h with vehicle control, TPA (100 nM), or TPA (100 nM) plus pretreatment (30 min) with TGF-␤1 (10 ng/ml). B, RAW264.7 macrophages were pretreated with vehicle control or actinomycin D (10 g/ml) and then stimulated with TPA (100 nM) for 6 h. A and B, total RNA was extracted, and Northern blot analysis was performed with 10 g of total RNA per lane. After electrophoresis, the RNA was transferred to nitrocellulose filters, which were hybridized with 32 P-labeled mouse MMP-12 cDNA. An 18 S oligonucleotide probe was hybridized as a qualitative indicator of loading. sion, we transiently transfected RAW cells with the same series of 5Ј-deletion constructs as described above, but cotransfected Smad3 instead of administering TGF-␤1. Smad3 inhibited MMP-12 promoter constructs from Ϫ1046 bp through Ϫ95 bp in a manner analogous to TGF-␤1 (Fig. 7). We also found that the AP-1 transcriptional activator c-Jun potently transactivated the (Ϫ1046/ϩ39) MMP-12 promoter up to 20-fold and Smad3 repressed inducibility Ͼ50%. This effect was retained in all tested MMP-12 promoter constructs containing the AP1 site ranging in size from Ϫ1046/ϩ39 to Ϫ95/ϩ39 MMP-12 (data not shown). These results indicate that Smad3 can inhibit AP-1mediated induction of the MMP-12 promoter.
Dominant-negative Smad3 Blocks TGF-␤1 Inhibition of the MMP-12 Promoter-To assess whether Smad3 is responsible for the inhibitory effect of TGF-␤1, we transfected a Smad3 dominant-negative (Smad3 ⌬C) construct that lacks 39 amino acids on the C-terminal end. This domain contains three phosphorylation sites that are necessary for Smad3 nuclear translocation. The inhibitory effect of TGF-␤1 on the (Ϫ1046/ϩ39) MMP-12 promoter was completely blocked by the addition of Smad3 ⌬C (Fig. 8). These data indicate an essential role for Smad3 in mediating inhibition of TGF-␤1 of the MMP-12 promoter. DISCUSSION MMP-12 is a macrophage-specific matrix metalloproteinase with broad substrate specificity. In the experiments presented here, we identified several cytokines and growth factors that affect the expression of MMP-12 and are known to be present in atherosclerotic plaques. Potential regulators of MMP-12 in the atherosclerotic environment include inflammatory cytokines and growth factors. Indeed, we observed that several cytokines, IL-1␤, TNF-␣, and M-CSF, up-regulated MMP-12 mRNA by 3-8-fold (Fig. 1). A similar induction by the growth factors PDGF-BB and VEGF was also observed. These factors are released by a number of cells in the atherosclerotic milieu, such as macrophages, vascular smooth muscle cells, and endothelial cells (40,41). Since recombinant MMP-12 activates MMP-2 and MMP-3 (24,25), expression of MMP-12 could lead, in turn, to activation of these MMPs, which would increase further degradation of the fibrous cap and promote plaque rupture.
Our findings indicate a transcriptional mechanism underlying the induction of MMP-12. To evaluate the inducibility of MMP-12 transcription directly, we analyzed the effect of TPA on the MMP-12 promoter. Deletion of sequences from Ϫ1046 to FIG. 4. Transcriptional regulation of the MMP-12 promoter. 5Ј deletion constructs (1.5 g) of the MMP-12 promoter from Ϫ1046 to Ϫ75 bp were transiently transfected into RAW cells as described under "Experimental Procedures." A, after transfection, cells were tested for responsiveness to vehicle control, TPA (100 nM), or TPA (100 nM) plus pretreatment (30 min) with TGF-␤1 (10 ng/ml). B, cells were tested for responsiveness to vehicle control, M-CSF (750 units/ml), or M-CSF (750 units/ml) plus pretreatment (30 min) with TGF-␤1 (10 ng/ml). Cells were harvested for luciferase and ␤-galactosidase activity after 24 h of incubation. A representative of three independent experiments is shown. Data (mean Ϯ S.D.) were subjected to one-way analysis of variance (ANOVA). * denotes statistical significance when compared with vehicle control (p Ͻ 0.05). ** denotes statistical significance when compared with TPA or M-CSF-treated constructs (p Ͻ 0.05).
Ϫ95 bp did not significantly affect inducibility by TPA, so the three PEA3 sites and the CCAAT site in this region probably play little or no role in the responsiveness of the MMP-12 promoter or in its basal activity. Point mutations in the MMP-12 AP-1 site not only reduced basal activity but also blocked inducibility (Fig. 4A). Nuclear protein binding to the MMP-12 AP-1 site increased by 7-fold in response to TPA (Fig. 5B). In contrast, the promoters for MMP-1 and MMP-3 both have two AP-1 sites (at approximately Ϫ190 and Ϫ75 bp) and maintain inducibility by TPA with point mutations at FIG. 5. Effect of TGF-␤1 on protein binding to the MMP-12 promoter AP-1 consensus site. A, nuclear extract (2 g) from RAW cells that had been treated with TPA (100 nM) for 24 h was incubated with a radioactively labeled oligonucleotide probe containing the putative AP-1 site of the MMP-12 promoter. The binding specificities of the protein-DNA complexes were assayed with the indicated unlabeled competitor probes added at a 50-M excess. B, nuclear extracts (2 g) from RAW cells that had been treated for 24 h with vehicle control, TPA (100 nM), or TPA (100 nM) plus 30 min pretreatment with TGF-␤1 (10 ng/ ml) were incubated with a radioactively labeled oligonucleotide probe containing the putative AP-1 site of the MMP-12 promoter. mut, mutation; NE, nuclear extract; wt, wild type.
FIG. 6. Smad3 inhibits MMP-12 promoter activity. A, the (Ϫ1046/ϩ39) MMP-12 promoter (1.5 g) was transiently transfected into RAW cells along with Smad expression constructs 1-4, 6, or 7 (0.75 g) as described under "Experimental Procedures." Equal amount of total DNA was used per transfection by adding an appropriate amount of control vector. Cells were then treated with vehicle control, TPA (100 nM), or TPA (100 nM) plus pretreatment (30 min) with TGF-␤1 (10 ng/ml) as indicated. B, the PAI-1 promoter (1.5 g) was transiently transfected into RAW cells along with Smad3 (0.75 g) or an equal amount of pcDNA3 control. A representative of three independent experiments is shown. Data (mean Ϯ S.D.) were subjected to ANOVA. * denotes statistical significance when compared with TPA-treated alone (p Ͻ 0.05). ** denotes statistical significance when compared with control vector (p Ͻ 0.05). either site (42)(43)(44). Sequence analysis of the MMP-12 5Јflanking sequences shows only one AP-1 site at Ϫ81 bp, and our functional studies indicate that this site is necessary for both basal promoter activity and TPA responsiveness.
We also examined factors that may inhibit MMP-12. Because TGF-␤1 can function as an anti-inflammatory growth factor and is present in atherosclerotic plaques (45), we determined its effect on MMP-12 expression. Our experiments indicate that TGF-␤1 inhibits cytokine-mediated induction of MMP-12 mRNA, protein, and enzymatic activity in human macrophages (Fig. 4). The mechanism by which TGF-␤1 inhibits MMP-12 expression also occurs at the level of transcription. TGF-␤1 inhibited all 5Ј MMP-12 promoter deletion constructs bearing an intact AP-1 site and decreased nuclear protein binding to the AP-1 site (Fig. 5B). Thus, the inhibitory effect of TGF-␤1 on MMP-12 expression is mediated, at least in part, through the AP-1 site. The inhibition of MMP-12 expression by TGF-␤1 requires the presence of Smad3. A dominant-negative form of Smad3 (Smad3⌬C) fails to translocate to the nucleus in response to TGF-␤1 ligand (46). In our reporter gene transfection experiments, Smad3⌬C blocked the inhibition of MMP-12 promoter by TGF-␤1 (Fig. 8). Because Smad3 has been shown to bind with Jun family members, including c-Jun (47,48), we hypothesize that Smad3 may decrease nuclear binding to the AP-1 site by sequestering c-Jun, thereby decreasing AP-1-dependent transactivation of the MMP-12 promoter. Alternatively, since both AP-1 proteins and Smad3 can bind to the transcriptional coactivators CREB-binding protein and p300 (46,49,50), it is also possible that Smad3 and c-Jun compete for limiting amounts of these coactivators, resulting in inhibition of AP-1mediated MMP-12 activity by TGF-␤1.
The role of TGF-␤1 in atherosclerosis and plaque rupture remains controversial. Low levels of active TGF-␤1 have been correlated with advanced atherosclerosis in patients (45). In contrast, adenovirus-mediated overexpression of TGF-␤1 in rat arterial endothelium promoted vascular wall remodeling, including intimal hyperplasia, apoptosis, and cartilaginous metaplasia (51). Of the many effects that TGF-␤1 has on cellular function, its pro-fibrotic effect may be the most critical on plaque stability. TGF-␤1, acting via Smad3, may inhibit expression of MMPs, such as MMP-12, thereby preventing extracellular matrix degradation. TGF-␤1 may also inhibit macrophage adhesion, transendothelial migration, thrombogenicity, FIG. 7. Smad3 inhibits TPA-induced MMP-12 promoter deletion constructs. 5Ј deletion constructs (1.5 g) of the MMP-12 promoter from Ϫ1046 bp to Ϫ75 bp were cotransfected with Smad3 (0.75 g) as indicated into RAW cells as described under "Experimental Procedures." An equal amount of total DNA was used per transfection by adding an appropriate amount of pcDNA3. After transfection, cells were tested for responsiveness to vehicle control or TPA (100 nM). Cells were harvested for luciferase and ␤-galactosidase activity after 24 h of incubation. A representative of three independent experiments is shown. Data (mean Ϯ S.D.) were subjected to ANOVA. * denotes statistical significance when compared with TPA-treated alone (p Ͻ 0.05).
FIG. 8. Dominant-negative Smad3 (Smad3⌬C) blocks TGF-␤1 inhibition of the MMP-12 promoter. The (Ϫ1046/ϩ39) MMP-12 promoter (1.5 g) was cotransfected with 0.75 g of Smad3 or Smad3⌬C as indicated. An equal amount of total DNA was used per transfection by adding an appropriate amount of pcDNA3. After transfection, cells were tested for responsiveness to vehicle control, TPA (100 nM), or TPA (100 nM) plus pretreatment (30 min) with TGF-␤1 (10 ng/ml). Cells were harvested for luciferase and ␤-galactosidase activity after 24 h of incubation. A representative of three independent experiments is shown. Data (mean Ϯ S.D.) were subjected to ANOVA. * denotes statistical significance when compared with TPA ϩ TGF-␤1 (p Ͻ 0.05). and lipid uptake, since it has been shown to down-regulate the expression of ICAM-1 (52), VCAM-1 (53), tissue factor (54), and the scavenger receptor (55), respectively. Indeed, mice that lack TGF-␤1 suffer from spontaneously uncontrolled inflammatory responses with macrophage-infiltrated organs (56). An understanding of the molecular mechanisms that regulate the antiinflammatory effect of TGF-␤1 in macrophages may offer novel therapeutic approaches to stabilizing vulnerable plaques.