Interaction of Apoptotic Cells with Macrophages Upregulates COX-2/PGE2 and HGF Expression via a Positive Feedback Loop

Recognition of apoptotic cells by macrophages is crucial for resolution of inflammation, immune tolerance, and tissue repair. Cyclooxygenase-2 (COX-2)/prostaglandin E2 (PGE2) and hepatocyte growth factor (HGF) play important roles in the tissue repair process. We investigated the characteristics of macrophage COX-2 and PGE2 expression mediated by apoptotic cells and then determined how macrophages exposed to apoptotic cells in vitro and in vivo orchestrate the interaction between COX-2/PGE2 and HGF signaling pathways. Exposure of RAW 264.7 cells and primary peritoneal macrophages to apoptotic cells resulted in induction of COX-2 and PGE2. The COX-2 inhibitor NS-398 suppressed apoptotic cell-induced PGE2 production. Both NS-398 and COX-2-siRNA, as well as the PGE2 receptor EP2 antagonist, blocked HGF expression in response to apoptotic cells. In addition, the HGF receptor antagonist suppressed increases in COX-2 and PGE2 induction. The in vivo relevance of the interaction between the COX-2/PGE2 and HGF pathways through a positive feedback loop was shown in cultured alveolar macrophages following in vivo exposure of bleomycin-stimulated lungs to apoptotic cells. Our results demonstrate that upregulation of the COX-2/PGE2 and HGF in macrophages following exposure to apoptotic cells represents a mechanism for mediating the anti-inflammatory and antifibrotic consequences of apoptotic cell recognition.


Introduction
The clearance of apoptotic cells by tissue macrophages and nonprofessional phagocytes is an essential process in tissue homeostasis, immunity, and resolution of inflammation. Apoptotic cell recognition actively leads to the production of anti-inflammatory mediators such as TGF-, IL-10, and PGE 2 [1,2]. Interactions between apoptotic and phagocytic cells play important roles in the regeneration and repair of damaged tissues by induction of growth-maintenance factors, such as VEGF, HGF, and PGE 2 , which can reconstitute the damaged tissue leading to decrease in fibroproliferative sequelae [3,4].
The COX-2 enzymatic product, PGE 2 , is a lipid mediator that, similar to TGF-, has been shown to have pro-or anti-inflammatory properties under differing circumstances. In the lung, PGE 2 plays roles in tissue repair processes and in limiting immune-inflammatory responses [5]. PGE 2 , which is generated via the conversion of arachidonic acid to PGH 2 via the COX-1 or COX-2 enzymes, is the major eicosanoid produced by lung fibroblasts and many other lung cells, including alveolar macrophages. Through Eprostanoid receptor 2 (EP2)-mediated increases in intracellular cyclic AMP, PGE 2 directly inhibits several major pathobiologic functions of fibroblasts, including chemotaxis, 2 Mediators of Inflammation proliferation [6], collagen synthesis [7], and differentiation into myofibroblasts [8]. Diminished PGE 2 production and/or signaling can be observed in human and animal lung fibrosis [9,10].
In studies by Fadok and colleagues [1], the exogenous addition of PGE 2 to LPS-stimulated macrophages decreased their production of proinflammatory cytokines, such as TNF-, IL-1 , and IL-8. Addition of indomethacin restored suppression of proinflammatory cytokine production caused by apoptotic cells in macrophages stimulated with LPS. Recently, we demonstrated in vitro that apoptotic cell-induced HGF reduces inflammatory cytokine expression in macrophages [11]. Moreover, we found that in vivo exposure to apoptotic cells induces anti-inflammatory effects through induction of COX-2/PGE 2 , and HGF signaling in bleomycin-stimulated lungs, using pharmacologic approaches [11,12]. Our previous studies also demonstrated that in vivo exposure to apoptotic cells resulted in enhanced expression of HGF [11] and COX-2 and secretion of PGE 2 [12] until the late fibrotic phase in bleomycin-induced lung injury. These data indicate that the anti-inflammatory and antifibrotic effects in the lung following apoptotic cell instillation are correlated with coordinated increases in HGF and COX-2/PGE 2 signaling. However, the mechanism underlying the prolonged induction of HGF and COX-2 by apoptotic cells is not clearly understood at the cellular model in vitro. Therefore, a key issue is whether recognition of apoptotic cells by macrophages can direct the signaling pathways toward coordinated induction of these anti-inflammatory and growth-maintenance factors. In the present study, we first characterized induction of COX-2 and PGE 2 by in vitro exposure of RAW 264.7 cells and murine primary peritoneal macrophages to apoptotic cells. We then determined how macrophages programmed by apoptotic cells orchestrate the interaction between COX-2/PGE 2 and HGF signaling.

Isolation and
Culture of Primary Cells. Resident peritoneal macrophages were isolated from 6-to 7-week-old pathogen-free male C57BL/6 mice (Orient Bio, Sungnam, Korea) weighing 20-22 g. The Animal Care Committee of the Ewha Medical Research Institute (Seoul, Republic of Korea) approved the experimental protocol (number 11-0171). Resident peritoneal macrophages were isolated by lavage with 5 mL of ice-cold sterile HBSS after mice were euthanized with CO 2 . The lavage fluid was centrifuged and resident peritoneal cells were plated at 1 × 10 6 cells/well and cultured in a humidified 5% CO 2 atmosphere at 37 ∘ C in DMEM supplemented with 10% heat-inactivated FBS, 2 mM l-glutamine, 100 g/mL of streptomycin, and 100 U/mL of penicillin. The isolated macrophages were stimulated with apoptotic Jurkat T cells (3 × 10 6 cells/mL). Suspended peritoneal macrophages were over 95% viable, as determined by trypan blue dye exclusion. Individual thymocytes were isolated from 3-to 4week-old mice by mincing the thymus through a 70 m pore size cell strainer (BD Biosciences, Bedford, MA). Human neutrophils were obtained from normal, healthy donors in accordance with a protocol reviewed and approved by the Institutional Review Board. Using endotoxin-free reagents and plastic ware, human neutrophils were isolated by the plasma Percoll method previously [13].

Induction of Apoptosis.
Human Jurkat T lymphocytes, HeLa epithelial cells, and murine thymocytes were exposed to UV irradiation at 254 nm for 10 min followed by incubation in RPMI-1640 with 10% fetal bovine serum for 2 h at 37 ∘ C and 5% CO 2 . Human neutrophils (>95% purity) were either cultured overnight at 9 × 10 5 /mL in RPMI 1640 at 37 ∘ C in 5% CO 2 or UV-irradiated for 10 min followed by incubation for 2 h before addition to macrophages (3 × 10 5 /mL). Evaluation of nuclear morphology using light microscopy on Wright-Giemsa stained samples indicated that the irradiated cells were approximately 70-80% apoptotic [14]. Apoptosis was confirmed by annexin V-FITC/propidium iodide (BD Biosciences, San Jose, CA) staining followed by flow cytometry analysis on a FACSCalibur system (BD Biosciences) [15]. Human aged neutrophils were shown to be typically 60-70% apoptotic by assessment of nuclear condensation on Wright-Giemsa stained samples and necrosis was less than 2% by trypan blue exclusion [16].

ELISA and EIA.
Culture supernatants were collected 2-24 h after stimulation. The levels of PGE 2 and 15d-PGJ 2 in the supernatants were determined using EIA kits. HGF and TGF-1 concentrations were measured by ELISA, according to the manufacturer's instructions.

Immunoblot
Analysis. Cells were lysed in 0.5% Triton X-100 lysis buffer and proteins were resolved on 10% SDS-PAGE gels and then electrophoretically transferred onto nitrocellulose membranes. The membranes were blocked for 1 h at room temperature with Tris-buffered saline (100 mM Tris-Cl, pH 7.5, 150 mM NaCl, and 0.1% Tween-20) containing 5% skim milk and then incubated with various primary antibodies at 4 ∘ C overnight and probed with a mouse antimouse HRP-conjugated secondary antibody. Membranes were developed using an enhanced chemiluminescence system (GE Healthcare, Buckinghamshire, UK).
2.9. Immunocytochemistry. COX-2 protein expression was evaluated in RAW 264.7 cells by immunocytochemistry. Cells were fixed with 4% paraformaldehyde, permeabilized with Triton X-100, and stained overnight at 4 ∘ C with rabbit polyclonal anti-COX-2 antibody (1 : 400; Abcam, Cambridge, UK). Subsequently, cells were washed with PBS three times and incubated with fluorescent isothiocyanate-conjugated donkey anti-rabbit IgG (1 : 500; Jackson ImmunoResearch). The slides were mounted with Vectashield mounting medium with DAPI (Vector Laboratories, Inc.) and examined using a confocal microscope (LSM5 PASCAL; Carl Zeiss) equipped with a filter set with excitation at 488 and 543 nm.

In Vitro Exposure of Macrophages to Apoptotic Cells
Induces mRNA and Protein Expression of COX-2. Before evaluation of the interaction between the COX-2/PGE 2 and HGF signaling pathways in macrophages following in vitro exposure to apoptotic cells, we determined the characteristics of COX-2 expression and PGE 2 production in macrophages. First, to evaluate COX-1 and COX-2 mRNA expression, semiquantitative RT-PCR was performed using total RNA extracted from RAW 264.7 cells. COX-2 mRNA expression was distinct at 2 h after in vitro exposure to apoptotic Jurkat T cells and increased gradually up to 6 h, and slightly declined at 12 h, but at 24 h the level of COX-2 mRNA declined ( Figure 1(a)). In contrast, viable Jurkat cells did not affect COX-2 mRNA expression over this time period (Figure 1(b)). There was no change in COX-1 mRNA expression within 24 h of exposure to apoptotic or viable Jurkat cells (Figure 1(a)). In addition, COX-2 mRNA expression was also measured following exposure to various cell types. Exposure to apoptotic neutrophils, apoptotic HeLa cells, and apoptotic thymocytes also induced COX-2 mRNA expression, but the timing of peak expression differed (Figures 1(c)-1(e)). The peak increase in COX-2 mRNA expression was observed at 1, 2, and 8 h after exposure to apoptotic HeLa cells, neutrophils, and thymocytes, respectively. Why the kinetics of COX-2 mRNA expression are different is not clearly explained in this experimental setting, but different cell types may cause that. We analyzed the levels of COX-2 mRNA expression following exposure to UV-irradiated apoptotic and aged apoptotic human neutrophils after standardization of the amount of these apoptotic cells, since the proportion of apoptotic cells in aged neutrophils is lower compared to UV-irradiated neutrophils (61% positive for apoptotic aged neutrophils as detected by Annexin V staining versus 80% positive for UV-irradiated apoptotic neutrophils) [16].
In addition to UV-irradiated apoptotic cells, aged apoptotic human neutrophils induced significantly COX-2 mRNA expression ( Figure 1(c)). These findings suggest that COX-2 mRNA expression induced by exposure to apoptotic cells in macrophages is a global phenomenon independent of cell type and apoptotic process. Expression of COX-1 and COX-2 protein was evaluated by immunoblot analysis of lysates of cultured RAW 264.7 cells. COX-2 expression increased progressively up to 24 h after addition of apoptotic Jurkat cells, but COX-1 expression did not change over this time period (Figure 1(f)). Exposure to viable cells had no effect on either COX-2 or COX-1 expression over the period examined ( Figure 1(g)). Pretreatment of RAW 264.7 cells with either actinomycin D or cycloheximide for 1 h before stimulation with apoptotic cells completely inhibited COX-2 expression, indicating that COX-2 mRNA de novo synthesis is required for its protein expression (Figure 1(h)). Confocal microscopy also demonstrated progressive increase in COX-2 protein in RAW 264.7 cells from 2 to 24 h after in vitro exposure to apoptotic cells ( Figure 1(i)).

In Vitro Exposure of Macrophages to Apoptotic Cells
Induces COX-2-Dependent PGE 2 Production. PGE 2 secretion, as measured by EIA, increased significantly in RAW 264.7 cells following exposure to apoptotic Jurkat cells (Figure 2(a)). A significant increase in PGE 2 production was observed 2 h after in vitro exposure to apoptotic cells, and PGE 2 production continued to increase up to 24 h. To confirm that COX-2 induction by exposure to apoptotic cells mediates the enhanced PGE 2 production in macrophages, RAW 264.7 cells were pretreated with the highly selective COX-2 inhibitor NS-398 (1∼50 M) or the nonselective COX inhibitor indomethacin (10 M) and incubated with apoptotic Jurkat T cells for 2 or 24 h. NS-398 reduced apoptotic cell-induced PGE 2 secretion in a dose-dependent manner (Figure 2(b)). Indomethacin also inhibited apoptotic cell-induced PGE 2 production. These data suggest that the apoptotic cell-induced increase in PGE 2 production in RAW 264.7 cells derives predominantly from induction of COX-2 expression. The expression of another product of the COX-2 pathway, 15d-PGJ 2 , was similarly enhanced (Figure 2(c)). Apoptotic cell-induced 15d-PGJ 2 secretion was also reduced by 1 M NS-398 or 10 M indomethacin (Figure 2(d)). These data suggest that the apoptotic cell-induced increase in PGE 2 and 15d-PGJ 2 production in RAW 264.7 cells derives predominantly from induction of COX-2 expression.

Enhancement of COX-2/PGE 2 Signaling by Interaction with Apoptotic Cells Mediates the Upregulation of HGF
Production. Park et al. [4] reported that enhanced HGF mRNA expression in RAW 264.7 cells following apoptotic cell exposure peaks at 2 h and that secretion of HGF protein is increased 24 h after exposure. In the present study, the role of COX-2 in apoptotic cell-induced HGF expression was also evaluated at these time points. Experiments were performed using pharmacologic inhibitors, such as NS-398, indomethacin, and siRNAs targeting COX-2 or COX-1. RAW 264.7 cells were pretreated with NS-398 or indomethacin and were then cultured with apoptotic Jurkat cells for 2 h to assess the effect on HGF mRNA expression and 24 h to assess the effect on HGF protein production. NS-398 (1, 10, and 50 M) and indomethacin (10 M) completely inhibited apoptotic cell-induced HGF mRNA (Figures 3(a) and 3(b)). Complete inhibition of HGF protein expression was also 6 Mediators of Inflammation To further examine the contribution of COX-2 to apoptotic cell-induced HGF expression in RAW 264.7 cells, experiments were performed using COX-2-specific siRNA. The negative control siRNA did not alter the COX-2 protein level in cells stimulated with apoptotic cells. COX-2 protein expression was completely inhibited at 2 and 24 h after apoptotic cell exposure in cells transfected with COX-2 siRNA by 67 and 100%, respectively, (Figures 3(d) and 3(e)) but COX-1 protein levels were unchanged, as determined by immunoblot analysis. Knockdown of the COX-2 gene prevented apoptotic cell-induced HGF mRNA and protein expression without affecting the mRNA and protein expression of the endogenous control, -actin (Figures 3(f) and 3(g)). In contrast, when COX-1 expression was silenced by transfection with COX-1-specific siRNA, HGF protein secretion was unaffected (Figures 3(h) and 3(i)). These data strongly suggest that only COX-2 induction is required for the induction of HGF mRNA and protein expression in RAW cells exposed to apoptotic cells in vitro.
Treatment of RAW 264.7 cells with 1 or 10 nM PGE 2 resulted in increases in the level of HGF protein in the culture medium (Figure 4(a)). We then examined the involvement of PGE 2 in mediating the effects of COX-2 on HGF induction. RAW 264.7 cells were treated with PGE 2 together with apoptotic cells in the presence of 10 M NS-398. Addition of 1 or 10 nM PGE 2 completely restored HGF mRNA expression suppressed by COX-2 inhibition (Figure 4(b)). Similarly, the reduced HGF secretion by cells treated with NS-398 or transfected with COX-2-specific siRNA was completely restored by addition of PGE 2 (Figures 4(c) and 4(d)).
A number of PGE 2 -specific receptors have been identified, including EP1, EP2, EP3, and EP4 [17,18], and it has been reported that macrophages can express both EP2 and EP4 [19]. To determine which receptors are involved in the apoptotic cell-induced PGE 2 signaling pathway, RAW 264.7 cells were pretreated with the EP2 receptor antagonist AH-6809 or the EP4 receptor antagonist GW-627368X for 1 h before apoptotic cells were added. The EP2 receptor antagonist, but not the EP4 receptor antagonist, blocked apoptotic cell-induced HGF secretion (Figure 4(e)). When RAW 264.7 cells were incubated with apoptotic Jurkat cells over a 24 h period, kinetic analysis showed that the rise in PGE 2 preceded the increase in HGF production (Figure 4(f)).
Regarding the effects of COX-2 inhibition using pharmacologic and genetic approaches, these results demonstrate that the COX-2/PGE 2 /EP2 axis induced by apoptotic cell exposure is instrumental in upregulating HGF production.

COX-2/PGE 2 Signaling Is Required for Upregulation of HGF Expression in Murine Peritoneal Macrophages in
Response to Apoptotic Cells. In addition to RAW 264.7 macrophages, we also examined COX-2/PGE 2 signaling in a primary cell model by isolating resident peritoneal macrophages by lavage from naive mice and then incubating these cells with apoptotic or viable Jurkat T cells. Apoptotic cell stimulation resulted in increased COX-2 mRNA and protein expression by the peritoneal macrophages, whereas exposure to viable cells did not (Figures 5(a) and 5(b)). There were no changes in COX-1 mRNA and protein expression. HGF mRNA expression induced by apoptotic cell exposure was inhibited by pretreatment with 10 M NS-398 ( Figure 5(c)). Furthermore, inhibition of COX-2 mRNA expression with NS-398 (10 M) and inhibition of the EP2 receptor with AH-6809 (10 M) significantly reduced HGF secretion ( Figure 5(d)). Immunoblot analysis of peritoneal macrophage lysates using anti-HGF -chain antibody indicated that intracellular HGF protein expression was also reduced by treatment with NS-398 or AH-6809 ( Figure 5(e)).
To confirm that COX-2 is required for apoptotic cellinduced HGF expression, peritoneal macrophages were transfected with COX-2-specific siRNA or a negative control siRNA, before addition of apoptotic Jurkat cells. Apoptotic cell-induced COX-2 protein expression decreased by approximately 60% in cells transfected with COX-2-specific siRNA, but COX-1 protein expression did not change ( Figure 5(f)). Upon exposure of primary peritoneal macrophages to apoptotic cells, the siRNA-mediated silencing of COX-2 mRNA expression significantly inhibited HGF production ( Figure 5(g)), as was observed in RAW 264.7 cells.

HGF Activation Mediates Upregulation of COX-2/PGE 2 in RAW 264.7 and Primary Peritoneal Macrophages in Response
to Apoptotic Cell Exposure. A previous report suggested that COX-2 and PGE 2 are downstream mediators of HGF expression in fibroblasts [9], which encouraged us to investigate the effect of HGF in mediating in vitro apoptotic cellinduced COX-2 and PGE 2 expression by macrophages. We evaluated the effect of PHA-665752, a selective inhibitor of the HGF receptor c-Met, on COX-2 mRNA and protein expression. At 2 and 6 h after in vitro exposure to apoptotic Jurkat cells, COX-2 mRNA expression was significantly inhibited by 10 M PHA-665752 but not by 1 M PHA-665752 ( Figure 6(a)). COX-2 protein expression was significantly inhibited in a dose-dependent manner by 1 and 10 M PHA-665752 at 6 and 24 h after in vitro exposure to apoptotic cells ( Figure 6(b)). Similarly, the pattern of PGE 2 production after inhibition of HGF signaling paralleled that of COX-2 protein expression. At 6 and 24 h after in vitro exposure to apoptotic cells, PGE 2 secretion was suppressed by both 1 and 10 M PHA-665752 in a dose-dependent manner (Figure 6(c)). To confirm the effect of HGF on PGE 2 production in RAW 264.7 cells, c-Met blocking antibody was also used. This antibody suppressed PGE 2 secretion by approximately 40% at 24 h after exposure to apoptotic cells (Figure 6(d)). Furthermore, 10 M PHA-665752 decreased partially HGF production at 24 h after in vitro exposure to apoptotic Jurkat cells, implicating the possibility of a positive feedback loop between COX-2/PGE 2 and HGF signaling pathways (Figure 6(e)).
The involvement of HGF signaling in enhancing induction of COX-2 and PGE 2 expression in response to apoptotic cell exposure was also shown in primary murine peritoneal macrophages, in which 1 or 10 M PHA-665752 inhibited COX-2 mRNA and protein expression as well as PGE 2 production (Figures 7(a)-7(c)). The suppressive effects of PHA-665752 on COX-2 mRNA and protein expression as well as PGE 2 production were consistent and dose-dependent, indicating that HGF partially mediates COX-2 expression and PGE 2 production induced by apoptotic cells.

COX-2/PGE 2 Signaling Is Not Required for Upregulation of TGF-in Unstimulated Macrophages in Response to
Apoptotic Cell Exposure. TGF-expression is also induced by apoptotic cells [1] but is reciprocally balanced with HGF [20]. Different regulatory systems are thought to be involved in the induction of HGF and TGF-expression. Thus, we examined whether COX-2/PGE 2 signaling functions differently in TGF-1 production following apoptotic cell exposure. Neither NS-398 nor indomethacin reduced apoptotic cell-induced TGF-1 mRNA expression in RAW 264.7 cells (Figure 8(a)). Similarly, TGF-1 protein secretion did not decrease significantly following COX-2 inhibition (Figure 8(b)). COX-2-specific siRNA resulted in limited inhibition of TGF-1 secretion (24% reduction) (Figure 8(c)). Compared to the effect on HGF expression, which was suppressed to the basal level following COX-2 inhibition, the reduction in TGF-1 secretion was not remarkable. In resident peritoneal macrophages from naive mice, apoptotic cell-induced TGF-1 mRNA and protein expression were unaffected or only minimally decreased by pharmacologic inhibition of COX-2 and PGE 2 signaling using 10 M NS-398 and 10 M AH-6809 (Figures 8(d) and 8(e)). Similarly, COX-2-specific siRNA did not inhibit induction of TGF-1 mRNA and protein expression in peritoneal macrophages (Figures 8(f) and 8(g)). These findings demonstrate that COX-2/PGE 2 signaling is not required for apoptotic cellinduced upregulation of TGF-1 mRNA and protein expression in macrophages.
Data from previous and present studies provide evidence that COX-2/PGE 2 and HGF activation are not involved in TGF-production in unstimulated RAW 264.7 cells in response to apoptotic cells [11]. Therefore, we wondered whether COX-2/PGE 2 and HGF activation mediate TGFproduction in stimulated RAW 264.7 cells in response to apoptotic cells. Macrophages were treated with 1.0 g/mL LPS; at the same time, apoptotic cells were added. Similar to the findings of Fadok and colleagues [1], the levels of TGF-by LPS-stimulated macrophages were enhanced significantly in response to apoptotic cells when compared with LPS-treated macrophages (Figure 8(h)). By adding the effect of pharmacologic inhibitors, including NS-398, AH-6809, or PHA-665752, TGF-production in LPS-stimulated macrophages was significantly inhibited at 24 h after exposure to apoptotic cells. These data indicate that anti-inflammatory response after apoptotic cell recognition is mediated, at least in part, via TGF-derived from the positive cross talk between COX-2/PGE 2 /EP2 and HGF/c-Met signaling pathways.

Discussion
The present study adds to the emerging view that macrophages recognizing apoptotic cells themselves can reinforce signaling pathways toward greater production of anti-inflammatory and antifibrotic mediators in a feedforward manner. We first demonstrated that expression of COX-2 (but not COX-1) mRNA and protein increased in RAW 264.7 cells as well as primary peritoneal macrophages exposed to apoptotic Jurkat T cells. The COX-2 mRNA induction was also observed when RAW 264.7 cells were exposed to other types of apoptotic cells, such as those of human neutrophils, HeLa epithelial cells, murine thymocytes, indicating the universality of the phenomenon independent of cell type. In addition to UV-irradiated 12  apoptotic cells, aged apoptotic human neutrophils also induced COX-2 mRNA expression. These data suggest that macrophages have switched their functional program in response to dying cells and triggered the COX-2-dependent pathways for production of immunosuppressive and/or regenerative mediators. Various PGs are autocrine mediators derived from metabolism of arachidonate through the action of COX. In this study we focused on PGE 2 synthesis because of its physiologic effects, which include limitation of the immuneinflammatory response and control of tissue repair processes. In vitro exposure of RAW 264.7 cells to apoptotic Jurkat cells increased PGE 2 production. Moreover, our time course study showed rapid induction and a continuous increase in PGE 2 production up to 24 h. Previous reports suggested that increased PGE 2 production along with decreased expression of proinflammatory eicosanoids, thromboxane B 2 , and leukotriene C 4 in human monocyte-derived macrophages in response to apoptotic cell exposure is indicative of a selective effect of apoptotic cell uptake on macrophage eicosanoid generation [1]. In the present study, using the highly selective COX-2 inhibitor NS-398, we provide evidence that PGE 2 production in macrophages is enhanced by exposure to apoptotic cells, predominantly via induction of COX-2 expression.

Mediators of Inflammation
Pharmacologic inhibition of COX-2 activity and siRNAmediated knockdown of COX-2 expression reduced apoptotic cell-induced HGF mRNA and protein expression in RAW 264.7 cells and primary peritoneal macrophages. However, knockdown of COX-1 expression did not affect HGF production in response to apoptotic cell exposure. These findings suggest that COX-2, but not COX-1, is specifically involved in the production of HGF. Addition of PGE 2 restored HGF mRNA and protein expression, which were suppressed by pharmacologic inhibition or genetic knockdown of COX-2 activity or expression. These data provide evidence that the effects of COX-2 on HGF induction in response to apoptotic cells are mediated by a direct effect of COX-2-derived PGE 2 on macrophages. Moreover, inhibition of PGE 2 signaling by an EP2 receptor antagonist (but not an EP4 antagonist) suppressed apoptotic cell-induced HGF production. Transcription of the HGF gene and HGF production are well known to be stimulated by substances that increase cyclic AMP, including PGE 2 [21,22]. It is therefore likely that COX-2-derived PGE 2 induces transcriptional HGF production by macrophages in response to apoptotic cells via the cyclic AMP pathway after interaction with an EP2 receptor. However, the COX-2/PGE 2 /EP2 axis is not the only signaling pathway involved in apoptotic cell-induced HGF production. A series of experiments in a previous study emphasized the importance of the RhoA/Rho kinase/PI3K/Akt/MAPK (including p38 MAPK, ERK, and JNK) axis, which is required for the upregulation of HGF mRNA and protein expression in RAW 264.7 cells in response to apoptotic cell exposure [8]. The PI3K/Akt and MAPK pathways also reportedly play a role in the regulation of COX-2 expression in response to a variety of extracellular stimuli [23]. Following MAPK signaling, the activation of transcriptional factors such as E-26 like protein 1 (ELK-1), activating transcription factor 2 (ATF-2), STAT, c-fos, c-Jun, and AP-1 in the COX-2 promoter region can increase COX-2 expression [24]. Our data indicate that RhoA is not involved in apoptotic cell-induced COX-2 expression because neither knockdown of RhoA nor inhibition of RhoA with the specific inhibitor C3 transferase suppressed COX-2 protein production (data not shown). However, the role of the PI3K/Akt and MAPK pathways in COX-2 induction by apoptotic cells requires further investigation. Likewise, production of another COX-2-derived PG, 15d-PGJ 2 , was also enhanced by in vitro exposure of macrophages to apoptotic Jurkat cells. Indeed, in vitro exposure of macrophages to apoptotic cells has been shown to increase the intracellular levels of PGE synthase1 as well as PGD synthase [2]. Similar to PGE 2 , 15d-PGJ 2 secretion in response to apoptotic cells was derived predominantly via induction of COX-2 expression. In addition, 15d-PGJ 2 induces expression of HGF in mesangial cells via a peroxisome proliferator response element in the HGF promoter through activation of peroxisome proliferator-activated receptors (PPAR)-. Here, we demonstrated 15d-PGJ 2 -induced HGF production in macrophages (data not shown). Thus, 15d-PGJ 2 may also participate in the upregulation of HGF expression by macrophages exposed to apoptotic cells, although this effect was not examined in our current study.
A number of prosurvival factors, including HGF and COX-2/PGE 2 , normally promote survival of epithelial and endothelial cells, fibroblast quiescence, and normal regulation of the extracellular matrix [25]. HGF signaling via the Met receptor upregulates COX-2 expression in different cell types, including fibroblasts and epithelial cells [26], and animal models have revealed a role for HGF as a mediator of COX-2/PGE 2 signaling-driven antifibrosis in vivo [9]. Thus, we investigated the possibility of cross talk between the HGF and COX-2/PGE 2 signaling pathways by exposing RAW 264.7 cells and peritoneal macrophages to apoptotic cells in vitro. The c-Met inhibitor PHA-665752 suppressed expression of COX-2 mRNA and protein in macrophages in response to apoptotic cells. Similarly, this inhibitor reduced PGE 2 production in a dose-dependent manner. Based on data from the time course study, decrease in PGE 2 production in RAW 264.7 cells was minimal at 24 h (∼35% reduction) and maximum at 6 h (∼60% reduction) after exposure to apoptotic cells. Using another c-Met inhibitor, a blocking antibody, we found a similar effect on PGE 2 production at 24 h (∼40% reduction). These data suggest that HGF/c-Met signaling alone does not fully exert positive effect on synthesis of PGE 2 by exposure to apoptotic cells. Moreover, PHA-665752 also reduced partially HGF production after in vitro exposure to apoptotic Jurkat cells. Collectively, our data suggest that the HGF/c-Met and COX-2/PGE 2 signaling pathways are interrelated through positive cross talk in macrophages stimulated by apoptotic cells over a 24 h period. This positive feedback loop may provide, at least in part, greater PGE 2 and HGF production in macrophages, which contribute to anti-inflammatory and antifibrotic activities induced by the interaction with apoptotic cells [12]. However, other mediators are likely involved and might participate in COX-2 and PGE 2 expression through different mechanisms since the c-Met inhibitor inhibited in part the PGE 2 as well as HGF production. In studies by Freire-de-Lima and colleagues [2], arachidonic acid release, COX-2 protein expression, and PGE 2 production are significantly dependent on the TGF-production in response to apoptotic cells. Indeed, macrophages expressing the truncated TGF-receptor did not show upregulation of COX-2 and PGE 2 in response to apoptotic cells. Thus, although not explored directly herein, it seems reasonable to assume that released TGF-contributes to greater COX-2 and PGE 2 expression, leading to HGF production.
Even though TGF-has been shown to play pivotal roles in the anti-inflammatory and anti-immunogenic responses to apoptotic cell clearance, it often acts as an antiproliferative or profibrotic agent. In contrast, HGF is an important tissue repair molecule. Administration of exogenous HGF accelerates tissue repair in several organs after acute injury or ischemia [27,28]. TGF-is a potent negative regulator of HGF expression [29], and HGF seems to have biological activities that oppose those of TGF-through diverse mechanisms [26]. Notably, our data indicate that COX-2/PGE 2 signaling is involved in transcription of HGF but not TGF-1 in unstimulated macrophages in response to apoptotic cells. In addition, data from previous studies using anti-c-Met blocking antibody showed that HGF activation is not involved in TGF-production in unstimulated RAW 264.7 cells in response to apoptotic cells [11]. However, in LPSstimulated macrophages, our in vitro studies using pharmacological inhibitors provide evidence that COX-2/PGE 2 and HGF signaling pathways are involved in enhancement of TGF-production in response to apoptotic cells. Thus, in this limited context, we see the possibility that anti-inflammatory response after apoptotic cell recognition is mediated, at least in part, via TGF-derived from the positive cross talk between COX-2/PGE 2 /EP2 and HGF/c-Met signaling pathways in the inflammatory lesion. On the other hand, prolonged inhibition of the COX-2/PGE 2 or HGF signaling pathway reversed the reduction of TGF-1 production and the hydroxyproline content in lung tissue following in vivo exposure to apoptotic cells at the late fibrotic phase [12]. Thus, the COX-2/PGE 2 and HGF signaling pathways may provide an important additional control of the balance between HGF and TGF-, favoring antifibrotic effects in the efferocytosis system.
In conclusion, the findings of this study reveal a novel function of COX-2/PGE 2 /EP2 signaling in macrophages exposed to apoptotic cells in vitro that results in enhanced HGF expression. Furthermore, the COX-2/PGE 2 pathway showed different regulatory effects for TGF-1, and HGF activation was shown to mediate COX-2 and PGE 2 expression induced by exposure to apoptotic cells in vitro. Therefore, it is likely that the ability of COX-2/PGE 2 /EP2 signaling to promote HGF synthesis is part of a positive feedback loop that results in amplification of macrophage COX-2 and PGE 2 expression (Figure 9). This positive cross talk mechanism between COX-2/PGE 2 /EP2 and HGF/c-Met signaling pathways may contribute to the anti-inflammatory and antifibrotic consequences of apoptotic cell recognition.