Extracellular Matrix-induced Cyclooxygenase-2 Regulates Macrophage Proteinase Expression*

Chronic inflammatory diseases are characterized by the persistent presence of macrophages and other mononuclear cells, tissue destruction, cell proliferation, and the deposition of extracellular matrix (ECM). The tissue degradation is mediated, in part, by enhanced proteinase expression by macrophages. It has been demonstrated recently that macrophage proteinase expression can be stimulated or inhibited by purified ECM components. However, in an intact ECM the biologically active domains of matrix components may be masked either by tertiary conformation or by complex association with other matrix molecules. In an effort to determine whether a complex ECM produced by vascular smooth muscle cells (SMC) regulates macrophage degradative phenotype, we prepared insoluble SMC matrices and examined their ability to regulate proteinase expression by RAW264.7 and thioglycollate-elicited peritoneal macrophages. Here we demonstrate that macrophage engagement of SMC-ECM triggers PKC-dependent activation of MAPKerk1/2 leading to increased expression of cyclooxygenase (COX)-2 and prostaglandin (PG) E2 synthesis. The addition of PGE2 to macrophage cultures stimulates their expression of both urokinase-type plasminogen activator and MMP-9, and the selective COX-2 inhibitor NS-398 blocks ECM-induced proteinase expression. Moreover, ECM-induced PGE2 and MMP-9 expression by elicited COX-2–/– macrophages is markedly reduced when compared with the response of either COX-2+/– or COX-2+/+ macrophages. These data clearly demonstrate that SMC-ECM exerts a regulatory role on the degradative phenotype of macrophages via enhanced urokinase-type plasminogen activator and MMP-9 expression, and identify COX-2 as a targetable component of the signaling pathway leading to increased proteinase expression.

In addition to the uPA/plasminogen system, macrophages express MMPs, which are associated with tissue remodeling via their ability to degrade ECM components (21,22). Macrophages, depending on their origin and state of activation, express several members of the MMP family including MMP-9 (23,24). As observed for the uPA/plasminogen system, several studies have suggested a role for MMP-9 in cell migration, and leukocyte infiltration and tissue remodeling are reduced in MMP-9 null mice (20,(25)(26)(27)(28)(29).
Because of their role in migration and/or tissue remodeling, the regulation of uPA and MMP-9 expression in macrophages has received much attention. The expression of both proteinases, as well as their specific inhibitors, is regulated by cytokines and growth factors (30 -32). In addition, it has been reported that macrophage proteinase expression is either stimulated (33)(34)(35)(36)(37)(38)(39) or inhibited (40,41) by purified ECM components. Moreover, in some cases, the regions of ECM molecules that regulate macrophage proteinase expression are cryptic and require proteolytic processing to be exposed (39,42,43). Because the regulatory domains of individual matrix components may be masked by tertiary conformation or by their association with other matrix molecules in a complex ECM, it is difficult to predict the effect that intact cell-derived ECM would have on macrophage proteinase expression.
Here we have examined the effect of a complex ECM deposited in situ by vascular smooth muscle cells (SMC) on proteinase expression by RAW264.7 macrophages and peritoneal macrophages. We demonstrate that uPA and MMP-9 expression by RAW264.7 macrophages and thioglycollate-elicited peritoneal macrophages is triggered by contact with SMC-ECM matrix. Induction of macrophage proteinase expression was dependent on a PKC-dependent activation of MAPK erk1/2 lead-ing to enhanced COX-2 expression and COX-dependent synthesis of PGE 2 .

MATERIALS AND METHODS
Cell Culture-Murine RAW264.7 macrophages (44) were obtained from American Type Culture Collection. Cells were maintained as adherent cultures in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% Cellect Gold fetal bovine serum (FBS), penicillin (100 units/ml), streptomycin (100 g/ml), and 4 mM glutamine (Invitrogen). Experiments to determine the effect of matrix on macrophage proteinase expression were carried out in DMEM supplemented with 0.1% low endotoxin bovine serum albumin.
Preparation of Extracellular Matrix-coated Dishes-Smooth muscle cell-derived matrices were prepared as described previously (12). Rat aortic smooth muscle cells (provided by Dr. T. A. McCaffrey, The George Washington University Medical Center) were plated into 6-, 12-, or 24-well plates in DMEM supplemented with 10% FBS. Three to 4 days after reaching confluence, the cell layer was removed by sequential exposure to 0.5% Triton X-100 in DPBS (10 min at room temperature) and 0.2 mM NH 4 OH in DPBS (3 min at room temperature). The remaining matrices were washed three times with DPBS and stored at 4°C.
Isolation of Peritoneal Macrophages-Thioglycollate-elicited peritoneal macrophages were obtained from Swiss Webster and COX-2 wild type, COX-2 heterozygote, and COX-2 null mice by the method of Edelson and Cohn (47) as described previously (9). Mice were injected intraperitoneally (3 ml/mouse) with 3% Brewer Thioglycollate Medium containing 0.3 mM thioglycollate (Difco). Four days later cells were harvested by lavage with cold DPBS. Peritoneal cells were recovered by centrifugation and resuspended in DMEM, 10% FBS, and plated into appropriate wells. Cells were allowed to adhere for 2 h and then washed free of nonadherent cells.
Determination of Plasminogen Activator Activity-Plasminogen activator activity was quantitated by utilizing a sensitive functional assay for plasmin (13). Aliquots of conditioned media were added to microtiter wells containing 82 l of DPBS, 0.05% Tween 20, 13 g of the plasmin substrate D-Val-Leu-Lys-aminomethylcoumarin (Enzyme Systems Products), and 0.5 g of bovine plasminogen (American Diagnostica). Samples were mixed and incubated at 37°C for 2.5 h. Cleavage of the substrate was monitored by measuring the increase in fluorescence in a Fluoroscan microplate reader (excitation, 330 -380 nm; emission, 430 -530 nm). Concentrations of uPA in the test samples were extrapolated from a standard curve by utilizing high molecular weight uPA (American Diagnostica). Plasminogen activator activity in macrophage-conditioned media was completely inhibited when preincubated with a polyclonal anti-human uPA IgG (American Diagnostica).
Determination of Metalloproteinase Activity-The presence of metalloproteinase activity in cellular conditioned media was determined by utilizing enzyme zymography as described previously (36). Conditioned media were mixed with SDS sample buffer (without mercaptoethanol) and incubated for 30 min at 37°C. Samples and molecular weight markers were electrophoresed in a 10% polyacrylamide gel containing 0.1% gelatin. The gel was then washed (two times) in 2.5% Triton X-100 to remove SDS. The gel was incubated at 37°C for 48 h in 200 mM NaCl containing 40 mM Tris-HCl and 10 mM CaCl 2 , pH 7.5, and stained with Coomassie Blue. The presence of gelatinolytic activity was identified as clear bands on a uniform blue background following destaining.
Western Blot Identification of Phosphorylated and Total MAPK erk1/2 -Cell lysates were electrophoresed in 4 -15% polyacrylamide gradient gels. Proteins were transferred to a PVDF membrane, following which the membrane was placed in blocking buffer for 1 h. Following one wash in 25 mM Tris buffer, pH 8.00, containing 137 mM NaCl, 2.7 mM KCl, and 0.5% Tween (TTBS), the membrane was incubated for 1 h in blocking buffer containing 75 ng/ml rabbit anti-phosphospecific p44/p42 MAPK IgG (Cell Signaling Technology). The membrane was washed (two times; TTBS) and incubated for 1 h in blocking buffer containing 0.3 g/ml goat anti-rabbit IgG conjugated to HRP (Transduction Laboratories). The membrane was washed in TTBS (three times). Bound HRP was visualized utilizing enhanced chemiluminescence. Membranes were stripped in 62.5 mM Tris buffer, pH 6.7, containing 100 mM ␤-mercaptoethanol and 2% SDS for 30 min at 50°C, washed, and probed for total MAPK erk1/2 with 0.2 g/ml rabbit anti-p44/p42 MAPK IgG (Cell Signaling Technology).
Determination of PGE 2 Levels in Macrophage-conditioned Media-The concentrations of PGE 2 in conditioned media were determined by utilizing competitive enzyme immunoassay (STAT-Prostaglandin E 2 EIA Kit; Cayman Chemical).
Western Blot Identification of COX-1 and COX-2-Cell lysates were electrophoresed in gradient gels, and proteins were transferred to a PVDF membrane. Following transfer, the membrane was placed in blocking buffer for 1 h, washed in PBS (one time), and incubated for 1 h in blocking buffer containing 0.5 g/ml rabbit IgG raised against a peptide containing amino acids 584 -598 of murine COX-2 or 1.0 g/ml rabbit IgG raised against a peptide containing amino acids 274 -288 of murine COX-1 (Cayman Chemical). The membrane was washed (two times; TTBS) and incubated for 1 h in blocking buffer containing 0.3 g/ml goat anti-rabbit IgG conjugated to HRP (Transduction Laboratories). The membrane was washed in TTBS (three times). Bound HRP was visualized utilizing enhanced chemiluminescence.
Western Blot Identification of Glyceraldehyde-3-phosphate Dehydrogenase (GAPDH)-Cell lysates were electrophoresed in gradient gels, and proteins were transferred to a PVDF membrane. Following transfer, the membrane was placed in blocking buffer for 1 h, washed in PBS (one time), and incubated for 1 h in blocking buffer containing 8 g/ml monoclonal anti-GAPDH IgG (Biodesign). The membrane was washed (two times; TTBS) and incubated for 1 h in blocking buffer containing 80 ng/ml rabbit anti-mouse IgG conjugated to HRP (Pierce). The membrane was washed in TTBS (three times). Bound HRP was visualized utilizing enhanced chemiluminescence.
Western Blot Identification of MMP-9 -Macrophage-conditioned media were electrophoresed in gradient gels, and proteins were transferred to a PVDF membrane. Following transfer, the membrane was placed in blocking buffer for 1 h, washed in PBS (one time), and incubated for 2 h in blocking buffer containing rabbit anti-mouse MMP-9 IgG (Chemicon). The membranes were washed (two times; TTBS) and incubated for 1 h in blocking buffer containing 0.3 g/ml goat anti-rabbit IgG conjugated to HRP (Bio-Rad). The membrane was washed in TTBS (three times). Bound HRP was visualized by utilizing enhanced chemiluminescence.

SMC-ECM Stimulates uPA and MMP-9 Expression by
Macrophages-Individual ECM components or their fragments can either stimulate or inhibit macrophage proteinase expression (33)(34)(35)(36)(37)(38)(39)(40)(41). Consequently, the effect of a complex ECM on proteinase expression remains unclear. Therefore, we cultured macrophages on an ECM deposited in situ by cultured aortic SMC, and we determined its effect on macrophage uPA and metalloproteinase expression. As shown in Fig. 1A, uPA activity in media recovered from murine RAW264.7 macrophages cultured on ECM was 15-fold greater than that observed in media from cells cultured on plastic. Likewise, the expression of MMP-9 by RAW264.7 macrophages cultured on ECM was markedly elevated as compared with that in cells cultured on plastic. Virtually all the MMP-9 secreted by murine RAW264.7 macrophages was in the pro-form (105 kDa), which is slightly larger than that expressed by human cells (92 kDa). Neither uPA nor MMP-9 activities were stimulated when macrophages were suspended above the ECM on porous inserts or incubated with ECM-conditioned media (data not shown).
To determine whether ECM regulated proteinase expression by primary macrophages, we isolated resident and thioglycollate-elicited peritoneal macrophages from Swiss Webster mice, and we plated them on either plastic or SMC-ECM. In contrast to RAW264.7 macrophages, an overnight incubation on ECM did not stimulate uPA expression by resident macrophages, although MMP-9 expression was strongly stimulated (Fig. 1B). As expected, conditioned media derived from thioglycollateelicited macrophages contained increased levels of uPA, MMP-9, and MMP-2 relative to resident cells (9,43,48). When elicited macrophages were cultured on ECM, their expression of uPA and MMP-9 activities was further increased. Thus, proteinase expression by macrophages is stimulated by their engagement of a complex ECM deposited in situ.
ECM-induced Proteinase Expression by Macrophages Requires Protein Kinase C-dependent Activation of MAPK erk1/2 -Monocyte/macrophage adherence to purified ECM components triggers tyrosine phosphorylation of several proteins including MAPK erk1/2 and subsequent induction of gene expression (49 -52). In previous studies (42,43), we demonstrated that the induction of macrophage proteinase expression by synthetic laminin ␣1-chain peptides was mediated by protein kinase C-dependent activation of MAPK erk1/2 . Therefore, we determined whether the stimulation of uPA and MMP-9 expression by SMC-ECM was dependent on a similar pathway. As shown in Fig. 2A, the adhesion of RAW264.7 macrophages to ECM triggered a rapid and persistent phosphorylation of MAPK erk1/2 . Equal protein loading was confirmed by stripping and probing for total MAPK erk1/2 . The observed activation of MAPK erk1/2 was not simply because of adhesion, because levels of phosphorylated MAPK erk1/2 are elevated in cells cultured on ECM versus plastic (Fig. 2B). Pre-incubation of cells with an inhibitor of either MEK-1 (U0126) or protein kinase C (calphostin C) blocked ECM-induced phosphorylation of MAPK erk1/2 without affecting levels of total MAPK erk1/2 in cell lysates (Fig.  2B). Moreover, ECM-induced uPA and MMP-9 expressions were blocked in RAW264.7 cells treated with either U0126 or calphostin C (Fig. 2, C and D). Taken together, these data demonstrate that engagement of SMC-ECM up-regulates macrophage uPA and MMP-9 expression and confirms the causal role for protein kinase C-dependent activation of MAPK erk1/2 in ECM-induced proteinase expression.
ECM-induced Proteinase Expression by Macrophages Is Cyclooxygenase-dependent-Activated MAPK erk1/2 triggers a diverse set of cellular responses by phosphorylating (and activating) transcription factors and other serine threonine kinases (53). In addition, MAPK erk1/2 phosphorylates phospholipase A 2 leading to an increase in prostaglandin synthesis. Prostaglandin (PG) E 2 , an important mediator of the inflammatory response, stimulates proteinase expression by a variety of cell types including monocytes (35, 54 -56). Therefore, we determined whether macrophage engagement of ECM stimulated their expression of PGE 2 . As seen in Fig. 3, the concentration of PGE 2 in media derived from elicited macrophages grown on ECM was increased Ͼ2-fold. Moreover, the addition of exogenous PGE 2 to elicited macrophages grown on plastic stimulated their expression of uPA and MMP-9 in a dose-dependent manner (Fig. 4).
To test the hypothesis that ECM-induced PGE 2 expression by macrophages is dependent on the activation of MAPK erk1/2 , RAW264.7 macrophages (5 ϫ 10 5 /well) were cultured on plastic, ECM, or ECM in the presence of U0126. After 18 h, media were recovered and assayed for PGE 2 . Media derived from macrophages cultured on plastic contained 65 Ϯ 14 pg/ml PGE 2 (mean Ϯ S.E.; n ϭ 4). Levels of PGE 2 increased 70-fold (4565 Ϯ 1741) when cells were cultured on ECM. In contrast, ECMinduced PGE 2 expression was markedly attenuated (126 Ϯ 28) when cells were cultured on ECM in the presence of the MEK-1 inhibitor.
Because prostaglandin synthesis is dependent on COX activity, we probed blots of RAW264.7 lysates, prepared from cells FIG. 1. SMC-ECM stimulates macrophage proteinase expression. A, RAW264.7 macrophages were suspended in DMEM containing 0.1% low endotoxin BSA (LE-BSA) and aliquoted into 24-well control plates or plates coated with SMC-ECM (2.5 ϫ 10 5 cells/well). Conditioned media were collected 18 h later and assayed for uPA and MMP-9 as described under "Materials and Methods." B, resident and thioglycollate-elicited macrophages were suspended in DMEM, 0.1% LE-BSA and aliquoted into 12-well control or ECM-coated plates (7.5 ϫ 10 5 cells/ well). Conditioned media were collected and assayed for proteinase activity. The uPA data are the mean activity (ϮS.E.) for three individual wells. The zymographic data illustrate MMP activity in conditioned media from one well for each group and are representative of the other wells. cultured either on plastic or ECM, with anti-murine COX-1 or COX-2 IgG. As seen in Fig. 5, the predominant COX isoform expressed by RAW264.7 macrophages is COX-2, and levels of COX-2 were up-regulated when cells were plated on ECM. We next examined the ability of the selective COX-2 inhibitor NS-398 to block proteinase expression induced by ECM. Macrophages were pre-incubated with NS-398 for 30 min and then plated on ECM in media containing the COX-2 inhibitor. Following adherence, cells received 0.1, 1, or 10 M PGE 2 and were incubated 18 h. MMP-9 activity and protein levels in macrophage-conditioned media were determined utilizing enzyme zymography and Western blot, respectively. Control cells expressed low levels of MMP-9 activity and protein (Fig. 6). When macrophages were cultured on ECM, the levels of MMP-9 activity and protein in their conditioned media were markedly increased. ECM-induced MMP-9 expression by RAW264.7 macrophages was blocked by NS-398 and partially restored by exogenous PGE 2 . Likewise, ECM-induced uPA expression was inhibited by NS-398 and partially restored by the addition of PGE 2 (data not shown). Thus, we conclude that COX-2-derived FIG. 2. The causal role of PKC-dependent activation of MAPK erk1/2 in ECM-induced proteinase expression. A, RAW264.7 macrophages were suspended in DMEM containing 0.1% LE-BSA and aliquoted into 6-well control plates or plates coated with SMC-ECM (2.0 ϫ 10 6 cells/well). At the indicated time points, cell lysates were prepared and phosphorylated, and total MAPK erk1/2 was identified by Western blot as described under "Materials and Methods." B, suspended macrophages (2 ϫ 10 6 cells/well) were pre-incubated for 30 min with 10 M MEK-1 inhibitor U0126 or 50 nM protein kinase C inhibitor (calphostin C) prior to plating on ECM. Lysates were prepared 30 min later, and phosphorylated and total MAPKs erk1/2 were identified by Western blot. C and D, suspended macrophages (2.5 ϫ 10 5 cells/well) were pre-incubated for 30 min with 10 M MEK-1 inhibitor U0126 or 50 nM protein kinase C inhibitor calphostin (Calph) C prior to plating on ECM in 24-well plates. Conditioned media were collected the next day and assayed for uPA and MMP-9 activities. The uPA data are the mean activity (ϮS.E.) from three individual wells. The zymographic data illustrate MMP activity in conditioned media from one well for each group and are representative of the other wells. Similarly, lysates derived from resident and thioglycollateelicited peritoneal macrophages were probed with anti-murine COX-1 or COX-2 IgG (Fig. 7). Western blots revealed that resident macrophages express relatively high levels of COX-1, which was unaffected by culture on ECM. In contrast to resident macrophages, elicited macrophages express both COX-1 and COX-2, and an overnight incubation with ECM markedly stimulated COX-2 expression. Consequently, COX-2 emerges as the major isoform expressed by inflammatory macrophages when plated on ECM. We next compared the ability of aspirin, a nonselective COX inhibitor, and NS-398 to block ECM-induced MMP-9 expression by resident and elicited macrophages. As seen in Fig. 8, ECM-induced MMP-9 expression by resident macrophages, which were shown to express COX-1, was markedly inhibited by aspirin (50 M), whereas the selective COX-2 inhibitor, NS-398, had little effect. In contrast to resident macrophages, elicited cells cultured on vascular ECM primarily express COX-2 (Fig. 7). Pre-incubation of elicited cells with either aspirin or NS-398 inhibited ECM-induced MMP-9 expression (Fig. 8). Taken together, these data demonstrate that ECM-induced proteinase expression by resident and elicited macrophages is COX-1 and COX-2-dependent, respectively.

ECM-induced COX-2 Expression by Macrophages Is MAPK erk1/2 -dependent and PGE 2 -independent-
We have demonstrated that ECM-induced proteinase expression is dependent on the synthesis of PGE 2 . To determine whether ECM-induced COX-2 expression is also dependent on PGE 2 , RAW264.7 cells were incubated with aspirin or MEK-1 inhibitor (U0126) prior to plating on ECM. Following an overnight incubation, cell lysates were prepared and probed for immunoreactive COX-2 and GAPDH. Levels of COX-2 were elevated in cells cultured on ECM (Fig. 9) and were unaffected by aspirin. In contrast, COX-2 was either undetectable or markedly reduced in U0126-treated macrophages cultured on plastic or ECM. The reduction in COX-2 expression appears to be specific, because levels of GAPDH were unchanged. Based on these data, we conclude that COX-2 expression is dependent on activated MAPK erk1/2 but is independent of PGE 2 .

ECM-induced PGE 2 and MMP-9 Expression by Macrophages
Derived from COX-2 Null Mice Is Markedly Reduced-To test directly the role of COX-2 in ECM-induced MMP-9 expression by inflammatory macrophages, COX-2 ϩ/ϩ , COX-2 ϩ/Ϫ , and COX-2 Ϫ/Ϫ mice were injected (intraperitoneally) with thioglycollate, and elicited macrophages recovered several days later. The cells were incubated on plastic or SMC-ECM overnight; Cells were then aliquoted into 6-well control plates or plates coated with SMC-ECM (2.0 ϫ 10 6 cells/well). Following an overnight incubation, cell lysates were prepared, and levels of COX-2 and GAPDH were determined by Western blot as described under "Materials and Methods." conditioned media were collected and assayed for PGE 2 and MMP-9. Western blots of cell lysates confirmed the absence of COX-2 in macrophages isolated from the null mice and demonstrated that ECM strongly up-regulated COX-2 expression by macrophages recovered from wild type and heterozygous mice (Fig. 10). ECM-induced PGE 2 expression by macrophages isolated from wild type and heterozygotes ranged from 2000 to 8000 pg/mg cell protein. In marked contrast, ECM-induced PGE 2 expression by null macrophages was 750 pg/mg. When conditioned media were examined for MMP-9 activity, ECM stimulated MMP-9 expression by COX-2 wild type and heterozygous macrophages but not null macrophages. Thus, although either COX-1 or COX-2 can mediate ECM-dependent proteinase expression by macrophages, the selective up-regulation of COX-2 expression in inflammatory macrophages by vascular ECM identifies COX-2 as a key regulator of tissue degradation at inflammatory sites. DISCUSSION The hallmarks of chronic inflammatory diseases such as rheumatoid arthritis and atherosclerosis are the persistent presence of macrophages and other mononuclear cells, tissue destruction, cell proliferation, and the deposition of ECM (57). The tissue degradation observed in these diseases is mediated, in part, by enhanced proteinase expression by macrophages. Results of experiments reported here demonstrate that macrophage engagement of SMC-ECM triggers PKC-dependent activation of MAPK erk1/2 leading to increased expression of COX-2 and PGE 2 synthesis. The addition of PGE 2 to elicited macrophages stimulates their expression of both uPA and MMP-9, and the selective COX-2 inhibitor NS-398 blocks ECM-induced proteinase expression. Moreover, PGE 2 and MMP-9 expression by COX-2 Ϫ/Ϫ macrophages was markedly reduced when compared with either COX-2 ϩ/Ϫ or COX-2 ϩ/ϩ macrophages. These data clearly demonstrate that vascular ECM exerts a regulatory role on uPA and MMP-9 expression of macrophages and identify COX-2 as a targetable component of the signaling pathway leading to increased proteinase expression.
The accumulation of macrophages at sites of injury and inflammation requires that monocytes leave the circulation, engage and migrate through the ECM, and differentiate. In addition to providing scaffolding for cells to attach, engagement of the ECM profoundly regulates monocyte/macrophage gene expression (49,50,58). Cellular engagement of ECM components, which is largely integrin-mediated, triggers tyrosine phosphorylation of several proteins including MAPK erk1/2 (42,49,51,52,59). The MAPK superfamily is a major signaling pathway by which extracellular signals regulate gene expression (53,60). In studies reported here, the activation of MAP-K erk1/2 was shown to be required for the up-regulation of the proteinase expression of macrophages induced by adhesion to SMC-ECM. These data confirm our earlier studies examining the role of MAPK erk1/2 in the induction of proteinase expression by synthetic ␣1-chain laminin peptides and laminin-1 fragments (42,43).
Activated MAPK erk1/2 triggers a diverse set of cellular responses by phosphorylating and activating several substrates including transcription factors, other serine/threonine kinases, and phospholipase A 2 . We have concluded that ECM-induced proteinase expression is regulated by MAPK erk1/2 activation of phospholipase A 2 and subsequent increase in PGE 2 synthesis based on the following observations. PGE 2 levels in conditioned media from macrophages cultured on ECM were significantly elevated as compared with cells attached to plastic controls. ECM-induced PGE 2 expression was blocked by a MEK-1 inhibitor. The addition of exogenous PGE 2 to peritoneal macrophages increased their expression of uPA and MMP-9.
The rate-limiting step in prostaglandin synthesis is catalyzed by the enzymes COX-1 and COX-2 (61). In most cells, including macrophages, COX-1 expression is constitutive, whereas COX-2 expression is induced in response to a variety of inflammatory mediators. Both isoenzymes can play a role in ECM-induced proteinase expression. Resident (resting) macrophages express COX-1 only, and aspirin blocks ECMinduced proteinase expression by these cells. Elicited (inflammatory) macrophages express both COX-1 and COX-2. However, the expression of COX-2 by elicited macrophages was markedly up-regulated when cultured on ECM. ECMinduced proteinase expression by inflammatory macrophages was blocked by selective COX-2 inhibition and was significantly blunted in macrophages isolated from COX-2 Ϫ/Ϫ mice. Thus, the selective up-regulation of COX-2 expression in inflammatory macrophages by vascular ECM identifies COX-2 as a key regulator of tissue degradation at inflammatory sites.
Two mechanisms have been proposed for regulating COX-2 expression, which may be relevant in ECM-induced COX-2 expression: activation of MAPK superfamily pathways (62)(63)(64)(65) and PGE 2 -dependent stabilization of COX-2 mRNA (66). In studies reported here, ECM-induced COX-2 and PGE 2 expression by elicited macrophages was markedly attenuated when FIG. 10. ECM induction of proteinase expression is blunted in COX-2 null macrophages. Top panel, thioglycollate-elicited macrophages isolated from COX-2 ϩ/ϩ , COX-2 ϩ/Ϫ , or COX-2 Ϫ/Ϫ mice were suspended in DMEM, 0.1% LE-BSA and aliquoted into 12-well control or ECM-coated plates (4.0 ϫ 10 5 cells/well). Following overnight incubation, conditioned media were collected and assayed for PGE 2 levels. Middle panel, elicited macrophages were suspended in DMEM, 0.1% LE-BSA and aliquoted into 12-well control or ECM-coated plates (5.6 ϫ 10 5 cells/well). Following an overnight incubation, cell lysates were prepared, and COX-2 was identified by Western blot utilizing anti-COX-2 IgG as described under "Materials and Methods." Lower panel, elicited macrophages were suspended in DMEM, 0.1% LE-BSA and aliquoted into 12-well control or ECM-coated plates (4.0 ϫ 10 5 cells/ well). Following an overnight incubation conditioned media were collected and assayed for MMP activity as described under "Materials and Methods." cells were incubated with an inhibitor of MAPK erk1/2 activation. However, it is important to emphasize that ECM-induced COX-2 expression by elicited cells was unaffected by aspirin, suggesting that PGE 2 does not play a major role in regulating COX-2 expression in this setting. Moreover, the engagement of ECM by resident macrophages leads to MAPK activation (data not shown) and MMP-9 expression without the induction of COX-2. These data suggest that activation of MAPK is required but not sufficient for COX-2 induction.
Evidence that COX-2 regulates ECM degradation and tissue remodeling observed in chronic inflammatory diseases is emerging. Macrophage proteinase expression contributes to the rupture of atherosclerotic plaques, which leads to thrombosis, myocardial infarction, and stroke. Several studies (67)(68)(69) have demonstrated increased COX-2 expression in atherosclerotic lesions. Moreover, COX-2, prostaglandin E synthase, and MMP activities were elevated on the shoulders of symptomatic plaques (70). Finally, treatment of low density lipoprotein receptor-deficient mice with a specific COX-2 inhibitor leads to reduced aortic atherosclerosis (71), and statin-dependent plaque stabilization in human carotid arteries was associated with decreased COX-2, prostaglandin E synthase, and MMP activities (72). Results of experiments reported here demonstrate, for the first time, the important regulatory role that SMC-ECM may have on COX-dependent tissue remodeling. Macrophage engagement of SMC-ECM initiates a signaling pathway, which leads to enhanced COX-2 expression, increased PGE 2 synthesis, and enhanced proteinase expression by inflammatory macrophages.