(MMP)-1 and MMP-3 Induce Macrophage MMP-9: Evidence for the Role of TNF-α and Cyclooxygenase-2 This information is current as of September 29, 2021. Michel Steenport, K. M. Faisal Khan, Baoheng Du, Sarah E. Barnhard, Andrew J. Dannenberg and Domenick J. Falcone J Immunol 2009; 183:8119-8127; Prepublished online 18 November 2009; doi: 10.4049/jimmunol.0901925 Downloaded from http://www.jimmunol.org/content/183/12/8119

References This article cites 53 articles, 22 of which you can access for free at: http://www.jimmunol.org/ http://www.jimmunol.org/content/183/12/8119.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists by guest on September 29, 2021 • Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2009 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Matrix Metalloproteinase (MMP)-1 and MMP-3 Induce Macrophage MMP-9: Evidence for the Role of TNF-␣ and Cyclooxygenase-21

Michel Steenport,*† K. M. Faisal Khan,*† Baoheng Du,‡ Sarah E. Barnhard,*† Andrew J. Dannenberg,‡ and Domenick J. Falcone2*†§

Matrix metalloproteinase (MMP)-9 ( B) participates in a variety of diverse physiologic and pathologic processes. We 3 3 recently characterized a cyclooxygenase-2 (COX-2) PGE2 EP4 receptor axis that regulates macrophage MMP-9 expression. In the present studies, we determined whether MMPs, commonly found in inflamed and neoplastic tissues, regulate this prosta- noid-EP receptor axis leading to enhanced MMP-9 expression. Results demonstrate that exposure of murine peritoneal macro- phages and RAW264.7 macrophages to MMP-1 (-1) or MMP-3 (stromelysin-1) lead to a marked increase in COX-2 Downloaded from expression, PGE2 secretion, and subsequent induction of MMP-9 expression. Proteinase-induced MMP-9 expression was blocked in macrophages preincubated with the selective COX-2 inhibitor celecoxib or transfected with COX-2 small interfering RNA (siRNA). Likewise, proteinase-induced MMP-9 was blocked in macrophages preincubated with the EP4 antagonist ONO-AE3-208 or transfected with EP4 siRNA. Exposure of macrophages to MMP-1 and MMP-3 triggered the rapid release of TNF-␣, which was blocked by MMP inhibitors. Furthermore, both COX-2 and MMP-9 expression were inhibited in macrophages preincubated with anti-TNF-␣ IgG or transfected with TNF-␣ siRNA. Thus, proteinase-induced MMP-9 expression by macrophages is dependent on http://www.jimmunol.org/ ␣ the release of TNF- , induction of COX-2 expression, and PGE2 engagement of EP4. The ability of MMP-1 and MMP-3 to regulate macrophage secretion of PGE2 and expression of MMP-9 defines a nexus between MMPs and prostanoids that is likely to play a role in the pathogenesis of chronic inflammatory diseases and cancer. These data also suggest that this nexus is targetable utilizing anti-TNF-␣ therapies and/or selective EP4 antagonists. The Journal of Immunology, 2009, 183: 8119–8127.

3 atrix metalloproteinase (MMP)-9, a neutral endopep- clooxygenase-2 (COX-2) expression, increased PGE2 synthesis,

tidase, participates in diverse physiologic and patho- and PGE2 binding to the EP4 prostanoid receptor (4, 5). MMP-9 logic processes (1–3). The degradation of extracellular expression was markedly reduced in macrophages isolated from M by guest on September 29, 2021 matrix (ECM) components is commonly thought to be the princi- COX-2-deficient mice or in wild-type macrophages treated with pal role of MMP-9 in these diverse biological processes. However, selective COX-2 inhibitors (4, 5). Likewise, induction of macro- in addition to degrading ECM, the biological actions of MMP-9 phage MMP-9 expression was blocked by selective EP4 antago- result from its ability to degrade and modify the activities of cy- nists or EP4 silencing (5). Taken together, these data point to the tokines and , growth factors, and proteinase inhibitors 3 3 important role the COX-2 PGE2 EP4 receptor axis plays in (1, 2). regulating macrophage MMP-9 expression. MMP-9 expression is low or absent in most normal tissues, and MMP-9 is secreted as a zymogen (pro-MMP-9) and is activated it is markedly elevated during inflammation, wound healing, and through the proteolytic removal of its pro-domain by MMP-3 and neoplasia (1). Major inducers include inflammatory cytokines, indirectly by plasmin via the activation of pro-MMP-3 (6, 7). growth factors, LPS, and ECM components (1). In this regard, we However, the role of MMP-3 and other MMP family members in previously demonstrated that macrophage MMP-9 expression in- the regulation of macrophage MMP-9 expression has not been ex- ␣ duced by TNF- , LPS, and ECM is dependent on enhanced cy- plored. In studies reported herein, we determined whether MMPs commonly found in both inflamed and neoplastic tissues regulate *Department of Pathology and Laboratory Medicine, †Vascular Biology Center, ‡De- macrophage MMP-9 expression by stimulating the COX- partment of Medicine, and §Department of Cell and Developmental Biology, Weill 23PGE 3EP4 receptor axis. We found that macrophage expo- Cornell Medical College, New York, NY 10065 2 sure to MMP-1 or MMP-3 led to a marked increase in COX-2 Received for publication June 18, 2009. Accepted for publication October 7, 2009. expression and PGE2 secretion, and subsequent induction of The costs of publication of this article were defrayed in part by the payment of page MMP-9. Proteinase-induced MMP-9 expression was blocked in charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. macrophages preincubated with celecoxib or transfected with 1 These studies were supported by National Institutes of Health Grant HL073375 (to COX-2 small interfering RNA (siRNA). Likewise, proteinase-in- D.J.F.) and by grants from the New York Crohn’s Foundation (to A.J.D.) and the duced MMP-9 was blocked in macrophages preincubated with the Center for Cancer Prevention Research (to D.J.F.). EP4 antagonist ONO-AE3-208 or transfected with EP4 siRNA. 2 Address correspondence and reprint requests to Dr. Domenick J. Falcone, Depart- ment of Pathology and Laboratory Medicine, Weill Cornell Medical College, 1300 Exposure of macrophages to MMP-1 and MMP-3 triggered the York Avenue, New York, NY 10065. E-mail addresses: [email protected] rapid release of TNF-␣, which was blocked by MMP inhibitors. 3 Abbreviations used in this paper: MMP, matrix metalloproteinase; ECM, extracel- Furthermore, both COX-2 and MMP-9 expression were effectively lular matrix; COX-2, cyclooxygenase-2; siRNA, small interfering RNA; APMA, blocked in macrophages preincubated with anti-TNF-␣ IgG or 4-aminophenylmercuric acetate; LE, low endotoxin. transfected with TNF-␣ siRNA. Thus, proteinase-induced MMP-9 Copyright © 2009 by The American Association of Immunologists, Inc. 0022-1767/09/$2.00 expression by macrophages is dependent on the release of TNF-␣, www.jimmunol.org/cgi/doi/10.4049/jimmunol.0901925 8120 MMPs INDUCE COX-2 EXPRESSION

␣ induction of COX-2 expression, and PGE2 engagement of EP4. COX-2 and TNF- siRNA transfection Taken together, these data identify a novel regulatory mechanism RAW264.7 macrophages were transfected with either COX-2, TNF-␣,or whereby MMP-1 and MMP-3 contribute to the inflammatory pro- nonspecific siRNA (Santa Cruz Biotechnology) according to the manufac- ␣ cess by liberating TNF- and stimulating PGE2 secretion via the turer’s protocol, which was modified as follows. Macrophages (T-75 flask) induction of COX-2, which leads, in turn, to increased MMP-9 were washed three times with Dulbecco’s PBS to remove serum, and me- expression. The ability of select MMPs to regulate macrophage dium was replaced with antibiotic-free DMEM. The cells were mechani- cally harvested, recovered by centrifugation, and resuspended in antibiotic- secretion of PGE2 defines a nexus between MMPs and prostanoids free DMEM supplemented with 10% FBS. The cells (2 ϫ 106/well) were that is likely to play a role in the pathogenesis of chronic inflam- aliquoted into a 6-well plate and incubated overnight. The next morning, matory diseases and cancer. 9.1 ␮lof10␮M siRNA was added to 152 ␮l of transfection medium (Santa Cruz Biotechnology) and incubated 5 min at room temperature. In another Materials and Methods tube, 9.1 ␮l of the transfection reagent (Santa Cruz Biotechnology) was added to the transfection medium and incubated 5 min. The two tubes were Macrophages combined and incubated 20 min to form the siRNA transfection reagent Thioglycollate-elicited peritoneal macrophages were obtained from Swiss complex. Immediately before transfection, macrophage medium was re- Webster mice by the method of Edelson and Cohn (8) as described pre- moved and replaced with 1.5 ml of antibiotic-free DMEM supplemented viously (9). Mice were injected i.p. (3 ml/mouse) with 3% Brewer thio- with 10% FBS. The plate was placed on a rocker plate in the laminar flow glycollate medium (Difco). Four days later, cells were harvested by lavage hood, and the transfection reagent complex was added dropwise. Cells with cold Dulbecco’s PBS. Peritoneal cells were recovered by centrifuga- were incubated 72 h at 37°C. tion and resuspended in DMEM supplemented with 10% FBS, penicillin EP4 siRNA transfection (100 U/ml), streptomycin (100 ␮g/ml), and 4 mM glutamine, and plated into tissue culture flasks or multiwell plates. Cells were allowed to adhere RAW264.7 macrophages were transfected with EP4 or nonspecific siRNA Downloaded from for 4 h and then washed free of nonadherent cells. All tissue culture sup- (Thermo Scientific/Dharmacon) utilizing the Accell siRNA delivery pro- plies were obtained from Invitrogen. The murine macrophage cell line tocol. Macrophages were aliquoted into a 6-well plate (0.5 ϫ 105/well) in RAW264.7 (10) was obtained from American Type Culture Collection and DMEM-10% FBS. The next day media were replaced with 1.5 ml of Accell maintained as adherent cultures in DMEM-10% FBS. All animal studies delivery medium containing 1 ␮M siRNA, and cells were incubated for described in this report have been reviewed and approved by the Weill 72 h at 37°C. Cornell Medical College Institutional Animal Care and Use Committee. RT-PCR Matrix metalloproteinases http://www.jimmunol.org/ RNA was prepared using TRIzol reagent kits (Invitrogen). RNA (2 ␮g) Human recombinant pro-MMP-1, active MMP-2, an active fragment of was reverse transcribed using Moloney murine leukemia virus reverse tran- MMP-3, and active MMP-7 were obtained from Calbiochem. Human re- scriptase (Roche Applied Science) and oligo(dT)16 primer. The resulting combinant active MMP-9 was obtained from Oncogene. Before use, pro- cDNA was then used for amplification. The primers and conditions for MMP-1 was activated by incubation (2 h at 37°C) with 1 mM 4-amino- PCR analysis of murine COX-2, MMP-9, EP4, and actin have been pre- phenylmercuric acetate (APMA) in 50 mM Tris (pH 7.5), containing viously described (5). Primers for murine TNF-␣ were obtained from R&D 0.05% Triton X-100 and 5 mM CaCl2 (11). The conversion of pro-MMP-1 Systems. PCR products were electrophoresed in a 1% agarose gel with 0.5 ϳ to its active form is associated with a loss of the pro-domain ( 10 kDa), ␮g/ml ethidium bromide and photographed under UV light. PCR condi- which was verified by SDS-PAGE and Western blot (data not shown). tions for the target genes were optimized to ensure that experimental sam- ples, excluding exuberant positive controls, are in the linear range when Preparation of cell lysates by guest on September 29, 2021 scanned. Densitometric analysis of the scanned target mRNAs were deter- Cell lysates utilized in the analysis of COX-2 were prepared by adding 1ϫ mined utilizing Scion Image and normalized for levels of actin mRNA, SDS sample buffer directly to washed macrophage monolayers. In exper- which served as a loading control. erk1/2 iments designed to monitor levels of phosphorylated and total MAPK , ␣ macrophages were lysed in TBS containing 2 mM EDTA, 1% Triton Determination of PGE2 and TNF- levels in macrophage X-100, 25 mM ␤-glycerophosphate, 1 mM sodium vanadate, 2 mM sodium conditioned media pyrophosphate, 1 mM PMSF, and 10 ␮g/ml aprotinin. Lysates were cen- trifuged (14,000 ϫ g) for 20 min at 4°C. The supernatants were recovered, The concentrations of PGE2 in conditioned media were determined utiliz- normalized for protein, and mixed with SDS sample buffer with 2-ME and ing the STAT-PGE2 ELISA kit (Cayman Chemical). Concentrations of TNF-␣ were determined utilizing the Legend Max ELISA kit for mouse boiled 5 min. Equal amounts of cell lysates were applied to gels based on ␣ protein content. TNF- (BioLegend). Western blots Cytokine Ab array Cell lysates were electrophoresed in 4–15% polyacrylamide gels, and pro- The presence of 18 cytokines released into conditioned media from mac- teins were transferred to a polyvinylidene difluoride membrane. The mem- rophages incubated 1 h with 50 nM MMP-3 was monitored utilizing an Ab brane was blocked in 5% defatted milk in TBST, washed in PBS, and array according to the manufacturer’s instructions (Panomics). The assay is incubated1hinblocking buffer containing rabbit IgG raised against a based on the sandwich ELISA method. ␮ peptide containing amino acids 584–598 of murine COX-2 (0.5 g/ml; Evaluation of cellular toxicity Cayman Chemical) or rabbit anti-phosphospecific p44/p42 MAPK (i.e., erk1/2 MAPK ) IgG (75 ng/ml; Cell Signaling Technology). Membranes were To rule out toxic effects of the inhibitors utilized in these studies, cellular washed twice in TBST and incubated1hinblocking buffer containing goat morphology and/or protein content following incubation are regularly anti-rabbit IgG conjugated to HRP (0.3 ␮g/ml; Transduction Laboratories). monitored. Additionally, we test for potential cytotoxicity by monitoring The membranes were washed three times in TBST, and bound HRP was mitochondrial dehydrogenase activity utilizing a MTT assay kit (Sigma- visualized utilizing chemiluminescence (HyGlo). In the case of membranes Aldrich). Levels of mitochondrial dehydrogenase activity were not signif- erk1/2 probed with anti-phosphospecific MAPK , following visualization of icantly reduced following exposure to any of the inhibitors over the time bound HRP, the membranes were stripped in 62.5 mM Tris buffer (pH 6.7) periods described herein. containing 100 mM 2-ME and 2% SDS for 30 min at 50°C, washed, and probed for total MAPKerk1/2 (Cell Signaling Technology). Cytokines and chemokines Macrophage conditioned media were electrophoresed in gradient gels ␣ ␣ ␥ and proteins were transferred to a polyvinylidene difluoride membrane. Recombinant murine TNF- , MIG (CXCL9), MIP- (CCL3), INF- , and The membrane was placed in blocking buffer for 1 h, washed in PBS, and RANTES (CCL5) were obtained from R&D Systems. incubated2hinblocking buffer containing rabbit anti-mouse MMP-9 IgG (0.2 ␮g/ml; Chemicon International) or incubated at 4°C overnight in Results blocking buffer containing rabbit anti-murine TNF-␣ (2 ␮g/ml; Thermo MMP-1 and MMP-3 induce macrophage COX-2 expression Scientific). Membranes were washed twice in TBST and incubated1hin 3 3 blocking buffer containing goat anti-rabbit IgG conjugated to HRP or goat In previous studies, we characterized a COX-2 PGE2 EP4 re- anti-mouse IgG conjugated to HRP. ceptor axis that regulates macrophage MMP-9 expression (5). In The Journal of Immunology 8121 Downloaded from

FIGURE 2. MMP catalytic activity is required for COX-2 induction. RAW264.7 macrophages (48-well plate; 2 ϫ 105 cells/well) were incu- bated in DMEM-0.1% LE-BSA containing 50 nM pro-MMP-1 or APMA- activated MMP-1. In other wells, macrophages were preincubated for 15

min with the broad spectrum MMP inhibitor GM6001 (50 nM) or MMP-3 http://www.jimmunol.org/ inhibitor I (20 ␮M) before the addition of 50 nM MMP-3. The next day lysates were prepared and levels of COX-2 and GAPDH were monitored by Western blot. The data are representative of four separate experiments.

FIGURE 1. MMP-1 and MMP-3 induce macrophage expression of COX-2. A, RAW264.7 macrophages or thioglycollate-elicted peritoneal To determine whether MMP-induced COX-2 expression was macrophages were suspended in DMEM containing 10% FCS and dis- dependent on their enzymatic activity, we compared the abilities of persed into a 48-well plate (2 ϫ 105 cells/well). Following adherence, cells pro-MMP-1- and APMA-activated MMP-1 to induce COX-2 ex- were washed three times with Dulbecco’s PBS and media were replaced pression in RAW264.7 macrophages. As seen in Fig. 2 (top panel), by guest on September 29, 2021 with DMEM-0.1% low endotoxin (LE) BSA alone (Ctrl) or DMEM-0.1% in contrast to active MMP-1, incubation with pro-MMP-1 failed to LE-BSA containing 50 nM MMP-1, MMP-2, MMP-3, MMP-7, MMP-9, or induce COX-2 expression. Next, we determined whether two un- 10 ng/ml LPS. The next day cell lysates were prepared and levels of related MMP inhibitors could block MMP-3-induced COX-2 ex- COX-2 and GAPDH were determined utilizing Western blot. B, Macro- pression. For this purpose, RAW264.7 macrophages were incu- ϫ 6 phages were dispersed into a 6-well plate (2 10 cells/well). Following bated with either a hydroxamic acid-based broad spectrum MMP adherence, cells were washed. Media were replaced with DMEM-0.1% inhibitor (GM6001) or an inhibitory peptide corresponding to the LE-BSA alone (Ctrl) or DMEM-0.1% LE-BSA containing 50 nM MMP-1, MMP-3, or 10 ng/ml LPS. After overnight incubation, total RNA was pro-domain of MMP-3 (MMP-3 inhibitor 1). As seen in Fig. 2 isolated, and mRNA levels for COX-2 and actin were determined utilizing (bottom panel), MMP-3-induced COX-2 expression was com- RT-PCR. The data are representative of three separate experiments. pletely blocked by either GM6001 or MMP-3 inhibitor 1. Taken together, these data demonstrate that the induction of COX-2 ex- pression by MMP-1 and MMP-3 is dependent on their enzymatic activities. the present study, we have determined whether MMP family mem- bers commonly found in inflamed and neoplastic tissues regulate MMP-1 and MMP-3 induce macrophage MMP-9 expression expression of this prostanoid-EP receptor axis. For this purpose, Macrophage MMP-9 expression is regulated by COX-2-dependent

RAW264.7 macrophages (10) and thioglycollate-elicited macro- synthesis of PGE2 (4, 5). Therefore, we next determined whether phages were incubated overnight with 50 nM active MMP-1, MMP-1 and MMP-3 induced macrophage MMP-9 expression in a MMP-2 (gelatinase A), MMP-3, MMP-7 (matrilysin), or MMP-9. COX-2-dependent manner. As seen in Fig. 3A, the levels of Incubation with 10 ng/ml LPS, a potent inducer of COX-2 expres- COX-2 and MMP-9 Ags in RAW264.7 lysates and conditioned sion by macrophages, served as a positive control (12, 13). Levels media, respectively, were elevated following an overnight incuba- of COX-2 in cell lysates were determined via Western blot. Over- tion with MMP-1, MMP-3, or LPS (positive control). The MMP-9 night incubation with MMP-1 or MMP-3 strongly stimulated in conditioned media from MMP-1- and LPS-treated macrophages COX-2 expression in RAW264.7 macrophages and elicited mac- was exclusively in the pro-form (105 kDa), which is slightly larger rophages; in contrast, MMP-2, MMP-7, and MMP-9 had little or than the human pro-form (92 kDa). In contrast, both pro (P; 105 no effect (Fig. 1A). MMP-induced changes in COX-2 gene expres- kDa) and active (A; 95 kDa) MMP-9 were identified in cultures sion were monitored utilizing semiquantitative RT-PCR. When incubated with MMP-3. MMP-3-dependent activation of pro- normalized for actin mRNA levels (i.e., loading controls), levels of MMP-9 was confirmed utilizing zymography (not shown) as pre- COX-2 mRNA in MMP-1- and MMP-3-treated macrophages were viously described (4), and this is consistent with earlier reports elevated 17- and 28-fold, respectively, in RAW264.7 macro- indicating that pro-MMP-9 is activated extracellularly by MMP-3 phages, and 12- and 8-fold in elicited macrophages (Fig. 1B). (6, 7, 14, 15). 8122 MMPs INDUCE COX-2 EXPRESSION Downloaded from

FIGURE 4. MMP-induced macrophage expression of MMP-9 is MAPKerk1/2- and COX-2-dependent. A, RAW264.7 macrophages (48-well plate; 2.5 ϫ 105/well) were cultured in DMEM-0.1% LE-BSA for 24 h.

Cells were then incubated with 50 nM MMP-1 or MMP-3 for 5–60 min, http://www.jimmunol.org/ and lysates were prepared. Levels of phosphorylated and total MAPKerk1/2 were determined utilizing Western blot. B, RAW264.7 macrophages (48- well plate; 2.5 ϫ 105/well) were cultured in DMEM-0.1% LE-BSA for 24 h. Cells were then incubated with media alone (Ctrl) or media contain- ing 50 nM MMP-1, MMP-2, MMP-3, MMP-7, or MMP-9 for 30 min, and lysates were probed for phosphorylated and total MAPKerk1/2. C, RAW264.7 macrophages (48 well plate; 2.5 ϫ 105/well) were preincu- bated with U0126 (10 ␮M; MEK-1 inhibitor) or celecoxib (5 ␮M; COX-2 inhibitor). Following an overnight incubation with 50 nM MMP-1, MMP-3, or 10 ng/ml LPS, levels of MMP-9 in conditioned media and by guest on September 29, 2021 COX-2 and GAPDH in lysates were determined by Western blot. The entire experiment was repeated twice with similar results.

dia containing 50 nM MMP-1 or MMP-3 were compared. Media FIGURE 3. MMP-1 and MMP-3 stimulate macrophage expression of ϫ 5 derived from control cells contained 107 pg/mg cell protein (av- PGE2 and MMP-9. A, RAW264.7 macrophages (48-well plate; 2 10 cells/well) were incubated in DMEM-0.1% LE-BSA alone (Ctrl) or erage of duplicate wells). Following incubation with MMP-1 or DMEM-0.1% LE-BSA containing 50 nM active MMP-1, MMP-3, or 10 MMP-3, PGE2 levels increased to 661 and 905 pg/mg cell protein, ng/ml LPS. Cells were incubated overnight, conditioned media collected, respectively. and lysates were prepared. Levels of MMP-9 in conditioned media and erk1/2 COX-2 and GAPDH in lysates were determined by Western blot. B, Activation of MAPK plays a causal role in MMP-induced RAW264.7 macrophages were incubated with 50 nM MMP-3 for 0–24 h, COX-2 expression and MMP-9 synthesis following which levels of MMP-9 in conditioned media and COX-2 and We previously have shown that the activation of MAPKerk1/2 is a GAPDH in lysates were determined. C, RAW264.7 macrophages were determinant of ECM-induced COX-2 expression and MMP-9 syn- incubated in DMEM-0.1% LE-BSA alone or media containing 50 nM MMP-3 for 0–24 h. Levels of PGE in conditioned media were determined thesis (4, 5). Therefore, we determined whether exposure of cells 2 erk1/2 utilizing ELISA. The data in each panel are representative of three separate to MMP-1 or MMP-3 led to the activation of MAPK . As seen erk1/2 experiments. in Fig. 4A, the levels of phosphorylated MAPK in macro- phages were elevated 20–30 min following the addition of active MMP-1 or MMP-3. In contrast, incubation with MMP-2, MMP-7, In a subsequent experiment, we examined the temporal relation- and MMP-9 did not result in the activation of MAPKerk1/2 (Fig. ship between MMP-3-induced COX-2, PGE2 and MMP-9 expres- 4B). Furthermore, preincubation of cells with the MEK-1 inhibitor sion (Fig. 3, B and C). Levels of COX-2 Ag were increased 4–6 U0126 (10 ␮M) blocked MMP-1- and MMP-3-induced phosphor- h following the addition of MMP-3, and they were strongly ele- ylation/activation of MAPKerk1/2 (data not shown), as well as vated at 18 and 24 h. Levels of PGE2 in media recovered from MMP-induced expression of COX-2 and MMP-9 (Fig. 4C). Thus, MMP-3-treated macrophages were markedly elevated at 18 and the activation of MAPKerk1/2 plays a causal role in MMP-induced

24 h, as compared with PGE2 levels in media recovered from COX-2 expression and MMP-9 synthesis. control cells. Following the increase in PGE2 levels, MMP-9 Ag in conditioned media of MMP-3-treated cells was strongly elevated MMP-induced MMP-9 synthesis is dependent on COX-2 at 24 h. In a separate experiment, levels of PGE2 release by To examine the role of COX-2 in proteinase-induced expression of RAW264.7 macrophages incubated 18 h with media alone or me- MMP-9, macrophages were preincubated with the selective The Journal of Immunology 8123 Downloaded from

FIGURE 5. COX-2 silencing, EP4 antagonist, or EP4 silencing block MMP-induced MMP-9 expression. A, RAW264.7 macrophages transfected with nonspecific (NS) or COX-2 siRNA were suspended in DMEM-0.1% FIGURE 6. ␣ A LE-BSA and aliquoted into a 24-well plate. Following adherence, cells MMP-1 and MMP-3 release TNF- from macrophages. , RAW264.7 macrophages (24-well plate; 2.5 ϫ 105/well) were incubated in

received media alone (Ctrl) or media containing 50 nM MMP-1, MMP-3, http://www.jimmunol.org/ ␮ or LPS (10 ng/ml) and were incubated overnight. Levels of MMP-9 in DMEM-0.1% LE-BSA containing 50 nM MMP-3 or MMP-3 and 20 M conditioned media and COX-2 and GAPDH in lysates were determined by MMP-3 inhibitor 1 overnight. The next day conditioned media from MMP- Western blot. B, RAW264.7 macrophages were incubated with 50 nM 3-treated macrophages were recovered and lysates were prepared. Condi- MMP-3 alone or in the presence of 25–100 nM EP4 antagonist ONO-AE3- tioned media alone or media to which MMP-3 inhibitor 1 was added (to 208 overnight. Levels of MMP-9 in conditioned media and GAPDH in block residual MMP-3 activity) were added to naive macrophages and lysates were determined by Western blot. C, Macrophages transfected with incubated overnight. Levels of COX-2 and GAPDH were monitored by B NS or EP4 siRNA were incubated overnight in media containing 50 nM Western blot. The data are representative of three separate experiments. , MMP-1, MMP-3, or LPS (10 ng/ml). Levels of MMP-9 in conditioned Conditioned media recovered from RAW264.7 macrophages (6-well plate; ϫ 6 media and COX-2 and GAPDH in lysates were determined by Western 1.15 10 /well), incubated with MMP-3 for 1 h, were collected and an- blot. Levels of EP4 mRNA in transfected cells were determined utilizing alyzed for the presence of the indicated cytokines and growth factors uti- by guest on September 29, 2021 C ϫ 5 RT-PCR. The entire experiment was repeated twice with similar results. lizing Ab array. , RAW264.7 macrophages (24-well plate; 2.5 10 / well) were incubated in DMEM-0.1% LE-BSA containing 50 nM of the indicated MMPs alone or in the presence of the broad-spectrum MMP inhibitor GM6001 (50 nM) overnight. The next day conditioned media COX-2 inhibitor celecoxib. Pretreatment with 5 ␮M celecoxib were recovered and levels of TNF-␣ were determined by Western blot. The blocked MMP-induced MMP-9 expression (Fig. 4C), which was data are representative of two separate experiments.

restored with the addition of PGE2 (not shown). The observation that celecoxib also attenuates COX-2 expression is consistent with

the previously reported positive feedback loop whereby PGE2 stimulates COX-2 expression by engaging the EP4 receptor and mRNA in cells transfected with EP4 siRNA was reduced ϳ95% triggering a Ras-MAPK signaling cascade (5, 16). when compared with cells transfected with nonspecific siRNA As an alternative strategy to examine the role of COX-2 in (Fig. 5C, left panel). When transfected cells were exposed to pro- MMP-induced expression of MMP-9, RAW264.7 macrophages teinases, MMP-1, MMP-3, and LPS (positive control) failed to were transfected with COX-2 siRNA. As observed in Fig. 5A, significantly induce COX-2 or MMP-9 expression by macrophages levels of COX-2 Ag in cells transfected with nonspecific siRNA transfected with EP4 siRNA (Fig. 5C, right panel). Thus, the in- were increased following incubation with MMP-1, MMP-3, or duction of MMP-9 expression by selected MMPs is causally linked LPS, whereas reduced levels of COX-2 were observed in MMP- or to MAPKerk1/2-dependent stimulation in COX-2 expression and

LPS-treated macrophages transfected with COX-2 siRNA. When subsequent PGE2 engagement of EP4. conditioned media were examined for MMP-9 Ag, MMP-1, ␣ MMP-3, and LPS treatment led to a large increase in MMP-9 MMP-dependent release of TNF- stimulates macrophage expression, which was markedly attenuated in macrophages in COX-2 expression which COX-2 expression was knocked down. Proteinases can affect cellular function by releasing tethered

Finally, the diverse biologic actions of PGE2 are mediated by growth factors (20–22), cytokines (23, 24), or fragments of extra- the EP1–4 family of prostanoid receptors (17, 18). Earlier studies cellular matrix (25–27) or by activating -activated receptor

have demonstrated that PGE2-dependent MMP-9 expression was family members (28, 29). To determine whether MMP-induced mediated via engagement of EP4 (5, 19). Thus, we examined the COX-2 expression resulted from the release or activation of a sol- ability of a selective EP4 antagonist ONO-AE3-208 to block uble factor, conditioned media from macrophages exposed to MMP-induced MMP-9 expression. Preincubation of macrophages MMP-3 overnight were recovered, treated with MMP-3 inhibitor with 50 nM ONO-AE3-208 blocked MMP-3-induced MMP-9 ex- 1, and transferred to naive cells. As seen in Fig. 6A (left panel), pression (Fig. 5B). To corroborate these findings, RAW264.7 mac- levels of COX-2 expression were elevated in macrophages incu- rophages were transfected with EP4 siRNA. The level of EP4 bated overnight with MMP-3. The ability of MMP-3 to induce 8124 MMPs INDUCE COX-2 EXPRESSION

COX-2 expression in macrophages coincubated with MMP-3 in- hibitor 1 was noticeably reduced. Similar results are seen in Fig. 2 (upper panel). Likewise, conditioned media recovered from cells following an overnight incubation with MMP-3 stimulated COX-2 expression by naive macrophages (Fig. 6A, right panel). The abil- ity of conditioned media to up-regulate COX-2 expression in naive cells was relatively unaffected by the addition of MMP-3 inhibitor 1. These data indicate that exposure of cells to MMP-3 released or activated a soluble factor that was responsible for MMP-induced COX-2 expression. In an effort to identify the factors released by macrophages fol- lowing exposure to MMP-3, the presence of 18 cytokines was monitored utilizing Ab array. As seen in Fig. 6B,1hoftreatment with MMP-3 resulted in a substantial increase in levels of TNF-␣ previously reported to be a potent stimulator of COX-2 (5, 30) and MMP-9 expression by macrophages (5, 31). To corroborate and extend the Ab array data, conditioned media were recovered from RAW264.7 macrophages treated with MMP-1 and MMP-3 for 18 h, and probed for levels of TNF-␣ utilizing Western blot. Re- Downloaded from combinant murine TNF-␣ (25 ng/lane) was utilized as a standard. When compared with the standard TNF-␣, media samples recov- ered from MMP-treated macrophages contained relatively high levels of TNF-␣ (Fig. 6C, upper panel). In contrast, TNF-␣ was not detected in media recovered from cells incubated overnight with MMP-1 or MMP-3 in the presence of the MMP inhibitor http://www.jimmunol.org/ GM6001 (Fig. 6C, lower panel). Finally, to obtain a quantitative measure of MMP-induced TNF-␣ release, media were recovered FIGURE 7. Neutralizing anti-TNF-␣ IgG blocks MMP-induced from RAW264.7 macrophages (48-well plate; 2.5 ϫ 105/well) COX-2 expression. Conditioned media recovered from RAW264.7 ϫ 5 treated with 50 nM MMP-1, 50 nM MMP-3, or 10 ng/ml LPS macrophages (24-well plate; 2.5 10 /well) exposed (1 h) to 50 nM ␣ MMP-1 were subsequently incubated 2 h with nonimmune IgG1 (20 (positive control) for 18 h and probed for levels of TNF- utilizing ␮ ␣ ␮ Ϯ g/ml) or rat monoclonal anti-mouse TNF- IgG1 (20 g/ml). Condi- ELISA. Media from control cells contained 698 109 pg/ml tioned media and Ab-treated conditioned media were then added to Ϯ ϭ (mean SEM; n 3). Following exposure to MMP-1 or MMP-3, naive macrophages in duplicate wells. In a similar experiment, condi- Ϯ Ϯ levels increased 25-fold (17,555 6,544) and 34-fold (23,767 tioned media recovered from cells incubated with MMP-3 were treated by guest on September 29, 2021 1,574), respectively. TNF-␣ levels in media from cells incubated with MMP-3 inhibitor 1 (20 ␮M) and/or anti-TNF-␣ (20 ␮g/ml) before with LPS increased 160-fold (114,333 Ϯ 1,828). adding to naive macrophages. Macrophages were incubated with con- Next, we utilized a neutralizing rat monoclonal anti-murine ditioned media overnight and lysates were prepared. Levels of COX-2 TNF-␣ IgG to determine whether MMP-induced release of TNF-␣ and GAPDH in lysates were determined by Western blot. The data are was responsible for inducing COX-2 and MMP-9 expression in representative of three separate experiments. RAW264.7 macrophages. For this purpose, conditioned media re- covered from cells incubated with MMP-1 or MMP-3 were pre- ␮ ␣ incubated 2 h with 20 g/ml anti-TNF- IgG or normal IgG before In these knockdown studies, LPS was included as a control to transferring to naive cells. As shown in Fig. 7, conditioned media monitor nonspecific effects of transfection with TNF-␣ siRNA. derived from MMP-1-treated cells stimulated COX-2 and MMP-9 Although LPS potently stimulates macrophage expression of expression in naive macrophages. Preincubation of the conditioned the inflammatory cytokines TNF-␣ and IL-1␤, LPS-induced ␣ media with anti-TNF- IgG blocked its ability to induce COX-2 COX-2 expression in human macrophages is independent of and MMP-9 expression; in contrast, normal IgG had no effect. either inflammatory cytokine (32). Consistent with these obser- Likewise, the addition of MMP-3 conditioned media to naive mac- vations, TNF-␣ knockdown had relatively little effect on LPS- rophages stimulated their COX-2 and MMP-9 expression. The ad- induced COX-2 and MMP-9 expression by murine RAW264.7 dition of MMP-3 inhibitor I or normal IgG to the conditioned macrophages (Fig. 8, right panel). Moreover, the presence of media recovered from MMP-3-treated macrophages had no effect neutralizing anti-TNF-␣ IgG (20 ␮g/ml) failed to block LPS- on its ability to induce COX-2 or MMP-9 in naive cells; in con- induced COX-2 and MMP-9 expression in murine RAW264.7 trast, preincubation with anti-TNF-␣ IgG suppressed its ability to macrophages (data not show). stimulate COX-2 and MMP-9 by naive macrophages. Finally, the Ab array in Fig. 6 revealed that in addition to TNF-␣ To directly test the role of TNF-␣ in proteinase-induced MMP-9 several other cytokines and chemokines were induced by stimula- expression, RAW264.7 macrophages were transfected with either tion with MMP-3, although to a minor degree. These data suggest TNF-␣ siRNA or nonspecific siRNA. The level of TNF-␣ mRNA that other factors may contribute to the observed proteinase induc- in cells transfected with TNF-␣ siRNA was reduced ϳ87% when tion of COX-2 and MMP-9 expression. However, an overnight compared with cells transfected with nonspecific siRNA (Fig. 8, incubation of RAW264.7 macrophages with 0.1–1.0 ␮g/ml MIG left panel). When TNF-␣ siRNA-transfected cells were exposed to (CXCL9), 0.5–50 ng/ml MIP-␣ (CCL3), 0.1–10 ng/ml INF-␥,or proteinases, MMP-1, MMP-3 failed to significantly induce COX-2 1–100 ng/ml RANTES (CCL5) failed to induce COX-2 or MMP-9 or MMP-9 expression. Importantly, COX-2 and MMP-9 expres- (data not shown). Taken together, these data indicate that MMP-1 sion were restored in these TNF-␣ siRNA-transfected cells by the and MMP-3 stimulate the release of TNF-␣, which induces mac- addition of exogenous TNF-␣ (Fig. 8, right panel). rophage COX-2-dependent MMP-9 expression. The Journal of Immunology 8125

FIGURE 8. TNF-␣ silencing blocks MMP-induced MMP-9 expression. RAW264.7 macrophages transfected with nonspecific (NS) or TNF-␣ siRNA were suspended in DMEM-0.1% LE-BSA and aliquoted into a 24-well plate. Following adherence, cells received media alone (Ctrl) or media containing 50 nM MMP-1, MMP-3, or LPS (10 ng/ml) and were incubated overnight. Additionally, cells transfected with TNF-␣ siRNA were incubated with MMP-1, MMP-3, or LPS in the presence of exoge- nous TNF-␣ (50 ng/ml). Levels of MMP-9 in conditioned media and FIGURE 9. MMP-1 and MMP-3 induce macrophage MMP-9 expres- ␣ COX-2 and GAPDH in lysates were determined by Western blot. Levels of sion: role of TNF- and COX-2. Our data suggest that active MMP-1 and MMP-3 trigger the release TNF-␣, which induces COX-2 expression and TNF-␣ mRNA were determined utilizing RT-PCR. The data are represen- Downloaded from erk1/2 tative of two separate experiments. PGE2 secretion. PGE2 binding to EP4 stimulates MAPK -dependent increase in MMP-9 expression.

Discussion induced COX-2 expression. Likewise, the failure of MMPs to Genetic evidence indicates that MMP-9 participates in the re- induce COX-2 expression in macrophages pretreated with cele-

cruitment of stem cells and tissue revascularization, skeletal coxib or transfected with EP4 siRNA is consistent with the http://www.jimmunol.org/ growth plate and ossification, choroidal neovas- hypothesis that PGE2-dependent positive feedback loops regu- cularization following injury, tumor angiogenesis, dermal-epi- late/amplify COX-2 expression. dermal separation in bullous pemphigoid, resolution of allergic The release of TNF-␣ is a well-defined mechanism by which inflammation, degradation of fibrin, ventricular enlargement proteinases regulate cell function. Pro-TNF-␣, tethered to the and accumulation following myocardial infarction, cell membrane by its NH2-terminal domain, is efficiently pro- and the pathogenesis of occlusive and aneurysmal vascular dis- cessed to its mature form and released by membrane-associated eases (2, 33, 34). The activity of this pleiotropic proteinase is TNF-␣ converting (i.e., TACE) (50, 51), which is a regulated at three levels: gene expression and pro-enzyme ac- member of the ADAMs (a disintegrin and metalloproteinase) tivation and inhibition (1, 2). The proteolytic activation of pro- family of metalloproteinases (i.e., ADAM17) (52). In addition by guest on September 29, 2021 MMP-9 is currently understood to take place in the fluid phase to TACE/ADAM17, several members of the MMP family are and is catalyzed by MMP-3 (6, 7, 35). However, the ability of reported to process pro-TNF-␣, although less efficiently (23, MMP-3 and/or other MMP family members to regulate MMP-9 24). Studies reported herein confirm the observation that MMP expression has not been explored. Results of experiments re- family members can process pro-TNF-␣. We demonstrate that ported herein define a complex mechanism by which extracel- the levels of TNF-␣ in media recovered from macrophages lular MMP-1 and MMP-3 stimulate macrophage MMP-9 ex- incubated with active MMP-1 or MMP-3 were elevated as com- pression via the liberation of TNF-␣ and subsequent induction pared with control cells. The observed release of TNF-␣ pro- of COX-2, as well as PGE2 engagement of EP4 receptor. vides a mechanism whereby these proteinases induce COX-2- PGE2, a member of the prostanoid family of autacoids, is an dependent MMP-9 expression. We and others have reported important mediator of the inflammatory response (36). PGE2 that macrophage COX-2 (5, 30) and MMP-9 expression (5, 31) can stimulate vasodilation and increase permeability, pain, fe- are stimulated by TNF-␣, and knockdown of COX-2 expression ver (37, 38), and tissue destruction/remodeling via enhanced blocks TNF-␣-induced MMP-9 expression (5). Moreover, as proteinase expression (4, 39–42). Synthesis of PGE2 requires reported herein, the ability of conditioned media recovered the liberation of arachidonic acid from membrane phospholip- from macrophages incubated with MMP-1 or MMP-3 to induce ␣ ids, COX-1- and/or COX-2-dependent generation of PGH2, and COX-2 and MMP-9 was blocked by neutralizing anti-TNF- its isomerization to PGE2 by members of the PGE synthase IgG. Furthermore, proteinases failed to induce either COX-2 or family (36). In most cells, Cox-1 expression is constitutive, MMP-9 in macrophages in which TNF-␣ expression was whereas COX-2 expression is induced in response to a variety knocked down utilizing siRNA. Thus, MMP-1-and MMP-3-in- of inflammatory mediators (43, 44). As reported here, levels of duced COX-2 and MMP-9 expression by macrophages is de- COX-2 mRNA and Ag in macrophages were selectively in- pendent on the release of TNF-␣ (Fig. 9). creased following exposure to catalytically active MMP-1 and MMP family members are potent proinflammatory and immu- MMP-3. Two mechanisms proposed for regulating COX-2 ex- noregulatory molecules that are implicated in the pathogenesis of pression may be relevant in MMP-induced COX-2 expression: chronic inflammatory diseases and cancer (2, 53, 54). Results of activation of MAPK superfamily pathways (45–48), and a experiments reported here define a complex mechanism by which

PGE2-dependent positive feedback loop (5, 16, 49). Results of collagenase-1 and stromelysin-1, commonly found in both in- experiments reported herein suggest that both pathways are im- flamed and neoplastic tissues, stimulate macrophage MMP-9 ex- portant. A causal role of activated MAPKerk1/2 in MMP-induced pression via the liberation of TNF-␣ and subsequent modulation 3 3 COX-2 expression is supported by the observation that MMP-1 the COX-2 PGE2 EP4 receptor axis (Fig. 9). The ability of and MMP-3 trigger MAPKerk1/2 activation in macrophages, and MMPs to induce COX-2-dependent MMP-9 expression by mac- preincubation of cells with a MEK-1 inhibitor blocks MMP- rophages defines a nexus between MMPs and prostanoids that is 8126 MMPs INDUCE COX-2 EXPRESSION likely to play a role in the pathogenesis of chronic inflammatory 20. Saksela, O., and D. B. Rifkin. 1990. Release of basic fibroblast growth factor- diseases and cancer. These data also suggest that this nexus is heparan sulfate complexes from endothelial cells by plasminogen activator-me- diated proteolytic activity. J. Cell Biol. 110: 767–775. targetable utilizing anti-TNF-␣ therapies and/or selective EP4 21. Sahin, U., G. Weskamp, K. Kelly, H. M. Zhou, S. Higashiyama, J. Peschon, antagonists. D. Hartmann, P. Saftig, and C. P. Blobel. 2004. Distinct roles for ADAM10 and ADAM17 in ectodomain shedding of six EGFR ligands. J. Cell Biol. 164: 769–779. 22. Falcone, D. J., T. A. McCaffrey, A. Haimovitz-Friedman, J.-A. Vergilio, and Acknowledgments A. C. Nicholson. 1993. Macrophage and foam cell release of matrix bound The authors thank Dr. Takayuki Maruyama (ONO Pharmaceutical) for growth factors: role of plasminogen activation. J. Biol. Chem. 268: 11951–11958. supplying the EP4 antagonist ONO-AE3-208. 23. Gearing, A. J., P. Beckett, M. Christodoulou, M. Churchill, J. Clements, A. H. Davidson, A. H. Drummond, W. A. Galloway, R. Gilbert, J. L. Gordon, et al. 1994. Processing of tumour necrosis factor-␣ precursor by metalloprotein- ases. Nature 370: 555–557. Disclosures 24. Mohan, M. J., T. Seaton, J. Mitchell, A. Howe, K. Blackburn, W. Burkhart, The authors have no financial conflicts of interest. M. Moyer, I. Patel, G. M. Waitt, J. D. Becherer, et al. 2002. The tumor necrosis factor-␣ converting enzyme (TACE): a unique metalloproteinase with highly defined substrate selectivity. Biochemistry 41: 9462–9469. 25. Giannelli, G., J. Falk-Marzillier, O. Schiraldi, W. G. Stetler-Stevenson, and References V. Quaranta. 1997. Induction of cell migration by matrix metalloprotease-2 1. Opdenakker, G., P. E. Van den Steen, and J. Van Damme. 2001. Gelatinase B: a cleavage of laminin-5. Science 277: 225–228. tuner and amplifier of immune functions. Trends Immunol. 22: 571–579. 26. Khan, K. M. F., G. W. Laurie, T. A. McCaffrey, and D. J. Falcone. 2002. Ex- 2. Parks, W. C., C. L. Wilson, and Y. S. Lopez-Boado. 2004. Matrix metallopro- posure of cryptic domains in the ␣1-chain of laminin-1 by stimulates teinases as modulators of inflammation and innate immunity. Nat. Rev. Immunol. macrophages urokinase and matrix metalloproteinase-9 expression. J. Biol.

4: 617–629. Chem. 277: 13778–13786. Downloaded from 3. Deryugina, E. I., and J. P. Quigley. 2006. Matrix metalloproteinases and tumor 27. Adair-Kirk, T. L., J. J. Atkinson, D. G. Kelley, R. H. Arch, J. H. Miner, and metastasis. Cancer Metastasis Rev. 25: 9–34. R. M. Senior. 2005. A chemotactic peptide from laminin ␣5 functions as a reg- 4. Khan, K. M. F., L. R. Howe, and D. J. Falcone. 2004. - ulator of inflammatory immune responses via TNF␣-mediated signaling. J. Im- induced cyclooxygenase-2 regulates macrophage proteinase expression. J. Biol. munol. 174: 1621–1629. Chem. 279: 22039–22046. 28. Coughlin, S. R. 2005. Protease-activated receptors in hemostasis, thrombosis and 5. Pavlovic, S., B. Du, K. Sakamoto, K. M. F. Khan, C. Natarajan, R. M. Breyer, vascular biology. J. Throm. Haemost. 3: 1800–1814. A. J. Dannenberg, and D. J. Falcone. 2006. Targeting prostaglandin E2 receptors 29. Boire, A., L. Covic, A. Agarwal, S. Jacques, S. Sherifl, and A. Kuliopulos. 2005.

as an alternative strategy to block cyclooxygenase-2-dependent extracellular ma- PAR1 is a matrix metalloprotease-1 receptor that promotes invasion and tumor- http://www.jimmunol.org/ trix-induced matrix metalloproteinase-9 expression by macrophages. J. Biol. igenesis of breast cancer cells. Cell 120: 303–313. Chem. 281: 3321–3328. 30. Wu, M. H., H. S. Sun, C. C. Lin, K. Y. Hsiao, P. C. Chuang, H. A. Pan, and 6. Dreier, R., S. Wallace, S. Fuchs, P. Bruckner, and S. Grassel. 2001. Paracrine S. J. Tsai. 2002. Distinct mechanisms regulate cyclooxygenase-1 and -2 in peri- interactions of chondrocytes and macrophages in cartilage degradation: articular toneal macrophages of women with and without endometriosis. Mol. Hum. Re- chondrocytes provide factors that activate macrophage-derived pro-gelatinase B prod. 8: 1103–1110. (pro-MMP-9). J. Cell Sci. 114: 3812–3822. 31. Heidinger, M., H. Kolb, H. W. Krell, M. Jochum, and C. Ries. 2006. Modulation 7. Ramos-DeSimone, N., E. Hahn-Dantona, J. Sipley, H. Nagase, D. L. French, and of autocrine TNF-␣-stimulated matrix metalloproteinase 9 (MMP-9) expression J. P. Quigley. 1999. Activation of matrix metalloproteinase-9 (MMP-9) via a by mitogen-activated protein kinases in THP-1 monocytic cells. Biol. Chem. 387: converging plasmin/stromelysin-1 cascade enhances tumor cell invasion. J. Biol. 69–78. Chem. 274: 13066–13076. 32. Barrios-Rodiles, M., G. Tiraloche, and K. Chadee. 1999. Lipopolysaccharide 8. Edelson, P. J., and Z. A. Cohn. 1976. Purification and cultivation of monocytes modulates cyclooxygenase-2 transcriptionally and posttranscriptionally in human and macrophages. In In Vitro Methods in Cell Mediated Tumor Immunity. macrophages independently from endogenous IL-1␤ and TNF-␣. J. Immunol. by guest on September 29, 2021 B. R. Bloom and J. R. David, eds. Academic, New York, pp. 333–340. 163: 963–969. 9. Falcone, D. J., and M. J. Ferenc. 1988. Acetyl-LDL stimulates macrophage- 33. Heissig, B., K. Hattori, S. Dias, M. Friedrich, B. Ferris, N. R. Hackett, dependent plasminogen activation and degradation of extracellular matrix. R. G. Crystal, P. Besmer, D. Lyden, M. A. S. Moore, Z. Werb, and S. Rafii. 2002. J. Cell. Physiol. 135: 387–396. Recruitment of stem and progenitor cells from the bone marrow niche requires 10. Raschke, W. C., S. Baird, P. Ralph, and I. Nakoinz. 1978. Functional macrophage MMP-9 mediated release of Kit-ligand. Cell 109: 625–637. cell lines transformed by Abelson leukemia virus. Cell 15: 261–267. 34. Ducharme, A., S. Frantz, M. Aikawa, E. Rabkin, M. Lindsey, L. E. Rohde, 11. Visse, R., and H. Nagase. 2003. Matrix metalloproteinases and tissue inhibitors F. J. Schoen, R. A. Kelly, Z. Werb, P. Libby, and R. T. Lee. 2000. Targeted of metalloproteinases: structure, function, and biochemistry. Circ. Res. 92: deletion of matrix metalloproteinase-9 attenuates left ventricular enlargement and 827–839. collagen accumulation after experimental myocardial infarction. J. Clin. Invest. 12. Matsumoto, H., H. Naraba, M. Murakami, I. Kudo, K. Yamaki, A. Ueno, and 106: 55–62. S. Ohishi. 1997. Concordant induction of prostaglandin E2 synthase with cyclo- 35. Okada, Y., Y. Gonoji, K. Naka, K. Tomita, I. Nakanishi, K. Iwata, K. Yamashita, oxygenase-2 leads to preferred production of prostaglandin E2 over thromboxane and T. Hayakawa. 1992. Matrix metalloproteinase 9 (92-kDa gelatinase/type IV and prostaglandin D-2 in lipopolysaccharide-stimulated rat peritoneal macro- collagenase) from HT 1080 human fibrosarcoma cells: purification and activation phages. Biochem. Biophys. Res. Commun. 230: 110–114. of the precursor and enzymic properties. J. Biol. Chem. 267: 21712–21719.

13. Vogel, S., M. J. Hirschfeld, and P. Y. Perera. 2001. Signal integration in lipo- 36. Park, J. Y., M. H. Pillinger, and S. B. Abramson. 2006. Prostaglandin E2 syn- polysaccharide (LPS)-stimulated murine macrophages. J. Endotoxin Res. 7: thesis and secretion: the role of PGE2 synthases. Clin. Immunol. 119: 229–240. 237–241. 37. Trebino, C. E., J. L. Stock, C. P. Gibbons, B. M. Naiman, T. S. Wachtmann, 14. Ogata, Y., Y. Itoh, and H. Nagase. 1995. Steps involved in activation of the J. P. Umland, K. Pandher, J. M. Lapointe, S. Saha, M. L. Roach, et al. 2003. pro-matrix metalloproteinase 9 (progelatinase B)-tissue inhibitor of metallopro- Impaired inflammatory and pain responses in mice lacking an inducible prosta- teinases-1 complex by 4-aminophenylmercuric acetate and proteinases. J. Biol. glandin E synthase. Proc. Natl. Acad. Sci. USA 100: 9044–9049. Chem. 270: 18506–18511. 38. Ivanov, A. I., and A. A. Romanovsky. 2004. Prostaglandin E2 as a mediator of 15. Goldberg, G. I., A. Strongin, I. E. Collier, L. T. Genrich, and B. L. Marmer. 1992. fever: synthesis and catabolism. Front. Biosci. 9: 1977–1993. Interaction of 92-kDa type IV collagenase with the tissue inhibitor of metallo- 39. Corcoran, M. L., W. G. Stetler-Stevenson, D. L. DeWitt, and L. M. Wahl. 1994. proteinases prevents dimerization, complex formation with interstitial collage- Effect of cholera toxin and pertussis toxin on prostaglandin H synthase-2, pros- nase, and activation of the proenzyme with stromelysin. J. Biol. Chem. 267: taglandin E2, and matrix metalloproteinase production by human monocytes. 4583–4591. Archives Biochem. Biophys. 310: 481–488. 16. Wang, D., F. G. Buchanan, H. Wang, S. K. Dey, and R. N. DuBois. 2005. 40. Shankavaram, U. T., D. L. DeWitt, S. E. Funk, E. H. Sage, and L. M. Wahl. 1997. Prostaglandin E2 enhances intestinal adenoma growth via activation of the Ras- Regulation of human monocyte matrix metalloproteinases by SPARC. J. Cell. mitogen-activated protein kinase cascade. Cancer Res. 65: 1822–1829. Physiol. 173: 327–334. 17. Breyer, R. M., C. K. Bagdassarian, S. A. Myers, and M. D. Breyer. 2001. Pro- 41. Ottino, P., and H. E. P. Bazan. 2001. Corneal stimulation of MMP-1, -9 and uPA stanoid receptors: subtypes and signaling. Annu. Rev. Pharmacol. Toxicol. 41: by platelet-activating factor is mediated by cyclooxygenase-2 metabolites. Curr. 661–690. Eye Res. 23: 77–85. 18. Hata, A. N., and R. M. Breyer. 2004. Pharmacology and signaling of prostaglan- 42. Dohadwala, M., R. K. Batra, J. Luo, Y. Lin, K. Kostyantyn, M. Pold, S. Sharma, din receptors: multiple roles in inflammation and immune modulation. Pharma- and S. M. Dubinett. 2002. Autocrine/paracrine prostaglandin E2 production by col. Ther. 103: 147–166. non-small cell lung cancer cells regulates matrix metalloproteinase-2 and CD44 19. Cipollone, F., M. L. Fazia, A. Iezzi, C. Cuccurullo, D. De Cesare, S. Ucchino, in cyclooxygenase-2-dependent invasion. J. Biol. Chem. 277: 50828–50833. F. Spigonardo, A. Marchetti, F. Buttitta, L. Paloscia, et al. 2005. Association 43. Hla, T., D. Bishop-Bailey, C. H. Liu, H. L. Schaefers, and O. C. Trifan. 1999. between prostaglandin E receptor subtype EP4 overexpression and unstable phe- Cyclooxygenase-1 and -2 isoenzymes. Int. J. Biochem. Cell Biol. 31: 551–557. notype in atherosclerotic plaques in human. Arterioscler. Thromb. Vasc. Biol. 25: 44. Smith, W. L., and R. Langenbach. 2001. Why there are two cyclooxygenase 1925–1931. isozymes. J. Clin. Invest. 107: 1491–1495. The Journal of Immunology 8127

45. Hendrickx, N., C. Volanti, U. Moens, O. M. Seternes, P. De Witte, activated protein kinase in interleukin-1␤-treated human synovial fibroblasts. J. R. Vandenheede, J. Piette, and P. Agnostinis. 2003. Up-regulation of cyclo- J. Biol. Chem. 276: 31720–31731. oxygenase-2 and apoptosis resistance by p38 MAPK in hypericin-mediated pho- 50. Moss, M. L., S. L. Jin, M. E. Milla, D. M. Bickett, W. Burkhart, H. L. Carter, todynamic therapy of human cancer cells. J. Biol. Chem. 278: 52231–52239. W. J. Chen, W. C. Clay, J. R. Didsbury, D. Hassler, et al. 1997. Cloning of 46. Pratt, P. F., D. Bokemeyer, M. Foschi, A. Sorokin, and M. J. Dunn. 2003. Alterations in a disintegrin metalloproteinase that processes precursor tumour-necrosis fac- subcellular localization of p38 MAPK potentiate endothelin-stimulated COX-2 expres- tor-␣. Nature 385: 733–736. sion in glomerular mesangial cells. J. Biol. Chem. 278: 51928–51936. 51. Black, R. A., C. T. Rauch, C. J. Kozlosky, J. J. Peschon, J. L. Slack, 47. Mizumura, K., S. Hashimoto, S. Maruoka, Y. Gon, N. Kitamura, K. Matsumoto, M. F. Wolfson, B. J. Castner, K. L. Stocking, P. Reddy, S. Srinivasan, et al. S. Hayashi, K. Shimizu, and T. Horie. 2003. Role of mitogen-activated protein 1997. A metalloproteinase disintegrin that releases tumour-necrosis factor-␣ kinases in influenza virus induction of prostaglandin E2 from arachidonic acid in from cells. Nature 385: 729–733. bronchial epithelial cells. Clin. Exp. Allergy 33: 1244–1251. 52. Edwards, D. R., M. M. Handsley, and C. J. Pennington. 2008. The ADAM 48. Subbaramaiah, K., T. P. Marmo, D. A. Dixon, and A. J. Dannenberg. 2003. metalloproteinases. Mol. Aspects Med. 29: 258–289. Regulation of cyclooxgenase-2 mRNA stability by taxanes: evidence for involve- 53. Van Lint, P., and C. Libert. 2007. and cytokine processing by ment of p38, MAPKAPK-2, and HuR. J. Biol. Chem. 278: 37637–37647. matrix metalloproteinases and its effect on leukocyte migration and inflam- 49. Faour, W. H., Y. L. He, Q. W. He, M. de Ladurantaye, M. Quintero, mation. J. Leukocyte Biol. 82: 1375–1381. A. Mancini, and J. A. Di Battista. 2001. Prostaglandin E2 regulates the level 54. Fingleton, B. 2006. Matrix metalloproteinases: roles in cancer and metastasis. and stability of cyclooxygenase-2 mRNA through activation of p38 mitogen- Front. Biosci. 11: 479–491. Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021