Abstract
TLR have emerged as important primary sensors for diverse stimuli and are increasingly implicated in various diseases. However, the molecular mechanisms underlying the regulation of the TLR system remain poorly understood. In this study, we report that some PGs may control TLR-mediated inflammatory events through modulation of TLR2 expression in brain immune cells. We first found that 15-deoxy-Δ12,14-PG J2 (15d-PGJ2) markedly altered the expression of TLR2 but not TLR4, TLR1, and TLR9 at the message and protein levels in activated glia. Down-regulation of TLR2 expression and downstream events of TLR2 activation, including phagocytosis by 15d-PGJ2, were also observed in cells treated with representative TLR2 ligands such as lipoteichoic acid and Pam3CSK4. We further revealed that certain 15d-PGJ2-related PGs such as 15d-PGD2 and PGD2 also suppressed the ligand-stimulated increase of TLR2 expression, whereas PGE2 and arachidonic acids did not. Interestingly, TLR2 expression was down-regulated even when such PGs were added at several hours after stimulator treatment. These findings appear to be independent of peroxisome proliferator-activated receptor γ and D prostanoid receptors (DPs) because potent synthetic peroxisome proliferator-activated receptor γ agonists, selective DP1 agonist, or DP2 agonist did not mimic the effects of such PGs on TLR2 expression. Taken together, our results suggest that 15d-PGJ2, 15d-PGD2, and PGD2 may play notable roles as modulators of the TLR2-mediated inflammatory events, and provide new insight into the resolution of inflammation in the brain.
Toll-like receptors, the mammalian homologs of the Drosophila Toll receptor, are key components of the immune system against pathogen attack. The TLRs are mainly expressed on the cells at the initial line of host defense, such as neutrophils, macrophages, and dendritic cells, where they act as primary sensors by recognizing a diverse range of endogenous and exogenous stimuli (1, 2, 3, 4, 5). To date, 13 mammalian TLRs have been identified and functionally characterized in mammals. Each TLR recognizes a distinct ligand and elicits different, but sometimes overlapping, immune responses (5). They use distinct combinations of adapter molecules and intracellular signaling molecules, thereby resulting in robust production or induction of inflammation-associated molecules that could influence both innate and adaptive immunity (6, 7).
Although the TLR-mediated inflammation is a critical response of the host to defend itself against pathological stimuli, excessive inflammatory signaling can be detrimental to the host, as is the case in endotoxin shock (8). Recent studies have shown that multiple mechanisms tightly regulate TLR-mediated responses to ensure appropriate response to pathogens and protect the host from a fatal response to pathogenic environments (5, 8). It has been reported that TLRs are negatively regulated at several levels, including down-regulation of TLR expression, sequestration of signaling molecules, blockade of their recruitment, degradation of target proteins, and inhibition of transcription (7). Several molecules, including A20, IL-1R-associated kinase-M, Toll-interacting protein, suppressor of cytokine signaling (SOCS)3 1, and PDZ-LIM domain protein 2, have been proposed to negatively regulate TLR signaling (9, 10, 11, 12, 13, 14). However, the precise mechanisms controlling the balance of TLR-mediated responses to pathogenic conditions are not yet fully understood.
Microglia and astrocytes are the resident immunoeffector cells of the CNS. Although these cells are quiescent under normal conditions, they are rapidly activated in response to pathological stimuli. Upon activation, astrocytes and microglia change their morphology, immunophenotype, and expression pattern of inflammatory mediators, which leads to immune and inflammatory responses (15, 16). These inflammatory mediators include PGs, a family of bioactive lipid mediators that are produced via the metabolism of arachidonic acid (17, 18). Arachidonic acid is converted to PGH2, which is subsequently converted to PGD2, PGE2, PGF2α, prostacyclin, or thromboxane A2. They bind to specific G protein-coupled receptors, designated D prostanoid receptors (DPs; for D PG receptor), EP, FP, IP, and TP receptor, respectively (19, 20).
PGs have long been considered to be potent proinflammatory regulators, and blockade of PG synthesis and activity has been a widely accepted treatment for inflammation-associated diseases. However, recent increasing reports have revealed that PGs can exert both toxic and paradoxically protective roles during inflammation (21). In particular, some PGs are shown to be increased during the resolution phase of inflammation and to alleviate inflammation in animal models (22, 23, 24). PGD2 and its metabolites reduced inflammatory events in several inflammation-associated disease models including asthma, allergic, and lung inflammation. Moreover, deficient production of PGs by genetic deficiency or by drugs such as cyclooxygenase (COX) inhibitors even worsened the inflammatory conditions (25, 26, 27, 28). In the CNS, PGD2 and 15-deoxy-Δ12,14-PG J2 (15d-PGJ2), a cyclopentenone PG derived from PGD2, showed protective effects on neuronal injury and suppressed the clinical features of several neuronal diseases, thus presenting as a novel therapeutic target for neurological diseases (29, 30, 31, 32, 33, 34). These opposing properties of some PGs have recently received much attention as targets for resolving inflammation.
In this report, we provide new insight into the anti-inflammatory actions of 15d-PGJ2, PGD2, and 15d-PGD2 based on their regulatory effects on TLR2 signaling in brain resident immune cells. We find that these PGs may regulate TLR2-mediated inflammatory responses by modulating TLR2 expression, but not TLR4, TLR1, and TLR9, independent of peroxisome proliferator-activated receptor γ (PPARγ) and DPs in activated glia. Our findings suggest not only a novel molecular explanation for the anti-inflammatory actions of these PGs but also suggest a possible new therapeutic intervention for TLR2-associated inflammatory diseases.
Materials and Methods
Reagents
Salmonella typhimurium LPS was purchased from Sigma-Aldrich. Purified lipoteichoic acid (LTA)-streptavidin and CpG oligodeoxynucleotide (ODN2216) were purchased from InvivoGen. Pam3Cys-Ser-(Lys)4 (Pam3CSK4) was purchased from Calbiochem. 15d-PGJ2, 15d-PGD2, PGD2, PGE2, DK-PGD2
Cell culture
Primary microglia and astrocytes were cultured from the cerebral cortices of 1- to 3-day-old ICR mice or Sprague-Dawley rats (Samtaco). Briefly, cortices were triturated into single cells in MEM (Sigma-Aldrich) containing 10% FBS (HyClone), plated in 75-cm2 T flasks (1 hemisphere/flask for mice, 0.5 hemisphere/flask for rat) and incubated for 2 wk. The microglia were detached from the flasks by mild shaking and applied to a nylon mesh to remove astrocytes and cell clumps. Cells were plated in 6-well plates (5 × 105 cells/well), 60-mm2 dishes (8 × 105 cells/dish), or 100-mm2 dishes (2 × 106 cells/dish). One hour later, the cells were washed to remove unattached cells before being used in experiments. Following removal of the microglia, primary astrocytes were prepared using trypsinization. The BV2 murine microglia was cultured in DMEM (Invitrogen Life Technologies) supplemented with 5% FBS (HyClone). Mouse peritoneal macrophages were obtained from sterile lavage of the peritoneal cavity with cold PBS (pH 7.4). The lavage fluid was centrifuged at 1000 × g for 10 min. The cell pellet was resuspended in DMEM containing 10% FBS, followed by plating in a 60-mm2 dishes, and then incubation at 37°C. Nonadherent cells were removed by repeated washing after 2 h. The adherent cells (macrophages) were cultured overnight.
RT-PCR analysis
Total RNA was extracted using RNAzol B (Tel-Test) and cDNA was prepared using reverse transcriptase that originated from avian myeloblastosis virus (Takara), according to the manufacturer’s instructions. PCR was performed with 25 cycles of sequential reactions: 94°C for 30 s, 55°C for 30 s, and 72°C for 30 s. Oligonucleotide primers were purchased from Bioneer. The sequences of PCR primers were as follows: reverse (R): 5′-GTA GCC CAC GTC GTA GCA AA-3′, forward (F): 5′-TCC CTC AAG ATT GTC AGC AA-3′ for TLR2; (R) 5′-AAA CGC AAA CCT TAC CAG-3′, (F) 5′-CCG AGA ACC GCT CAA C-3′ for TLR1; and (R) 5′-GCC ATC TCC TGC TCG AAG TCT AG-3′, (F) 5′-CAT GTT TGA GAC CTT CAA CAC CCC-3′ for actin. PCR products were separated by electrophoresis in a 1.5% agarose gel and detected under UV light.
Western blot analysis
Cells were washed twice with cold PBS, and then lysed in ice-cold modified RIPA buffer (50 mM Tris-HCl (pH 7.4), 1% Nonidet P-40, 0.25% sodium deoxycholate, 150 mM NaCl, 1 mM Na3VO4 and 1 mM NaF) containing protease inhibitors (2 mM PMSF, 100 μg/ml leupeptin, 10 μg/ml pepstatin, 1 μg/ml aprotinin, and 2 mM EDTA). The lysates were centrifuged at 12,000 × g for 10 min at 4°C and the supernatant was collected. Proteins were separated by SDS-PAGE and transferred to a nitrocellulose membrane, which was incubated with primary Abs and peroxidase-conjugated secondary Abs (Vector Laboratories). The results were visualized using an ECL system (Sigma-Aldrich).
Flow cytometric analysis
Cells were washed twice with PBS containing 1% FBS, collected, and stained with FITC-conjugated anti-mouse TLR2, PE-conjugated anti-mouse TLR4/MD2 Ab, or isotype control Ab (eBioscience) for 1 h at 4°C. After washing, the cells were analyzed with a FACSVantage (BD Biosciences). To stain intracellular TLR9, IFN-γ (5 U/ml)-primed cells were treated with CpG (5 μM) in the presence or absence of 15d-PGJ2, fixed in 2% paraformaldehyde, and washed with ice-cold FACS staining buffer containing 2.5% saponin on ice for 20 min. Then, the cells were stained with FITC-conjugated anti-mouse TLR9 (eBioscience) for 30 min at 4°C. The data analyses were done with CellQuest software (BD Bioscience) and Win-MDI software (The Scripps Research Institute, La Jolla, CA).
Phagocytosis assay
BV2 cells were plated in 60-mm2 dishes (2.5 × 104 cells/dish) or were left untreated with LPS, purified LTA, or Pam3CSK4 for 18 h. Phagocytic capacity was measured by incubating cells with FITC-conjugated phagocytic beads (8 × 106 beads/ml; FluoSpheres polystyrene microspheres; Molecular Probes) at 37°C. Subsequently, cells were washed three times with PBS and then gently removed from the wells using cell scrapers. To visualize the effect of PGs on phagocytosis, cells containing FITC-conjugated phagocytic beads were analyzed by fluorescence microscopy (Zeiss).
Immunofluorescence and confocal microscopy
BV2 cells were plated onto 12-mm round coverslips (Fisher Scientific), and then untreated or treated with LPS, LTA, or Pam3CSK4 in the presence or absence of 15d-PGJ2, and fixed with ice-cold 100% methanol. The cells were permeabilized with PBS containing 0.1% Triton X-100 and blocked with 1% BSA (Sigma-Aldrich). The samples were then washed with PBS containing 0.1% Triton X-100, and incubated overnight at 4°C with TLR2 Ab and then secondary Ab. 4′,6′-Diamidino-2-phenylindole hydrochloride (DAPI) was used to stain nuclei. Images were obtained with a Carl Zeiss LSM510 confocal microscope, and were analyzed by LSM software (release 3.2).
Statistical analysis
All data were expressed as the mean ± SD and analyzed by one-way ANOVA followed by post-hoc comparisons (Student-Newman-Keuls test) using Statistical Package for Social Sciences 8.0 (SPSS).
Results
Ligand-dependent change of TLR2 expression is notably modulated by 15d-PGJ2 in activated microglia
In an effort to explore the regulatory mechanism underlying TLR-mediated inflammation, we investigated the effect of 15d-PGJ2 on TLR-associated inflammatory events in brain immune cells. First, we examined the action of 15d-PGJ2 on cell surface expression of TLR2 and 4 because their expression is strictly regulated in response to endogenous and exogenous stimuli (35, 36). BV2 murine microglial cells were treated with Salmonella typhimurium LPS in the presence or absence of 2.5 μM 15d-PGJ2, and the cell surface expression levels of TLR2 and TLR4 were determined by flow cytometric analysis. When cells were stimulated with LPS, the cell surface expression levels of both TLR2 and TLR4 were markedly altered in BV2 microglial cells, with LPS treatment stimulating the up- and down-regulation of TLR2 and TLR4, respectively. Interestingly, LPS-stimulated enhancement of TLR2 expression was significantly attenuated by pretreatment with 15d-PGJ2. 15d-PGJ2 considerably reduced the LPS-stimulated increase of TLR2 expression even at concentrations <0.5 μM, with further suppression seen in response to increasing concentrations of 15d-PGJ2 (Fig. 1⇓A). In contrast, no significant changes of TLR4 expression were observed under these conditions. The LPS-dependent decrease of TLR4 expression was not affected by exposure to 15d-PGJ2 at any of the tested concentrations. Moreover, we also observed that 15d-PGJ2 markedly altered the change of TLR2 expression, whereas it did not affect TLR4 expression in cells treated with ganglioside, an endogenous ligand for TLR2 and TLR4 (Refs. 37 and 38 ; Fig. 1⇓C). These results further support the suppressive effects of 15d-PGJ2 on the ligand-stimulated induction of TLR2 expression.
15d-PGJ2 effectively attenuates the ligand-dependent enhancement of TLR2 cell surface expression in microglia. BV2 microglial cells were pretreated with 15d-PGJ2 or control medium for 1 h, and then left untreated (control) or treated with 100 ng/ml LPS (A and B), 25 μg/ml brain ganglioside mixture (Gmix, C), 1 μg/ml CpG (D), 10 μg/ml LTA (E), or 100 ng/ml Pam3CSK4 (E) for 12 h. TLR2, TLR4, or TLR9 expression was analyzed by flow cytometric analysis using FITC-conjugated anti-mouse TLR2, PE-conjugated anti-mouse TLR4, or FITC-conjugated anti-mouse TLR9 mAbs, respectively. Data shown are representative of at least four independent experiments.
The enhancement of TLR2 expression by other representative TLR2 agonists were also suppressed in the presence of 15d-PGJ2
To further explore the action of 15d-PGJ2 on TLR systems, we next examined whether 15d-PGJ2 could modulate TLR9 and TLR1 expressions. As shown in Fig. 1⇑D, TLR9 expression was increased after exposure to CpG, a ligand for TLR9. However, 15d-PGJ2 did not change the enhancement of TLR9 expression by CpG at any of the concentrations tested. Furthermore, we did not observe any effects of 15d-PGJ2 on TLR1 expression (data not shown). These results raised the possibility that 15d-PGJ2 may modulate the expression level of TLR2, but not TLR1, 4, and 9, in activated glia, thus regulating TLR2-mediated inflammatory responses.
To address this issue, we examined whether 15d-PGJ2 could also influence the enhancement of TLR2 expression induced by other representative TLR2 ligands such as LTA and Pam3CSK4. Consistent with the above results, the enhancement of TLR2 expression triggered by both LTA and Pam3CSK4 was significantly lower in 15d-PGJ2-treated cells compared with the untreated control (Fig. 1⇑E). In all cells treated with LTA, Pam3CSK4, or LPS, we observed the 15d-PGJ2-dependent reduction of TLR2 cell surface expression within 3 h after stimulator treatment (data not shown). These results provide strong evidence for the modulation of TLR2 cell surface expression by 15d-PGJ2 in TLR2 ligand-stimulated microglia, suggesting that 15d-PGJ2 may act to regulate TLR2-mediated inflammatory events through modulation of TLR2 expression level in brain microglia.
15d-PGJ2 also attenuates the expression of TLR2 at message and protein levels as well as cell surface level
TLR2 plays a critical role as a signal-transducing molecule for a large repertoire of exogenous and endogenous ligands, and it might contribute to the initiation and progression of CNS pathology (39, 40, 41, 42). To visualize the effect of 15d-PGJ2 on TLR2 expression, we performed immunofluorescence microscopy using TLR2 Ab. As compared with untreated control cells, TLR2 expression was considerably increased in cells within 3 h of LTA treatment, but this enhancement was apparently attenuated by treatment with 15d-PGJ2 (Fig. 2⇓A). Similar results were also observed in cells with Pam3CSK4 or LPS (data not shown).
Message and protein levels of TLR2 are also affected by 15d-PGJ2. A, BV2 cells were left untreated (control) or treated with 10 μg/ml LTA in the absence or presence of 2.5 μM 15d-PGJ2, and then analyzed by immunofluorescence with anti-TLR2. DAPI, Nuclear staining. Original magnification, ×400. All images are representative of at least three independent experiments. B, Cells were left untreated or treated with 100 ng/ml LPS or 100 ng/ml Pam3CSK4 for 6 h following pretreatment with the indicated concentrations of 15d-PGJ2 for 1 h. Total RNA was then extracted for RT-PCR analysis. C, Cells were treated with 100 ng/ml LPS, 10 μg/ml LTA, or 100 ng/ml Pam3CSK4 for 12 h in the absence or presence of 2.5 μM 15d-PGJ2, followed by Western blot analysis using TLR2 and α-tubulin Abs. Data shown are representative of at least three independent experiments.
To further investigate the action of 15d-PGJ2 on TLR2 expression, we examined the effects of 15d-PGJ2 treatment on the message levels of TLR2 by RT-PCR analysis. As shown in Fig. 2⇑B, TLR2 message levels were markedly and concentration-dependently reduced by 15d-PGJ2. However, LPS-stimulated induction of TLR1 transcript was not influenced by 15d-PGJ2. TLR2 protein levels were also influenced by treatment with 15d-PGJ2. 15d-PGJ2 apparently suppressed the induction of TLR2 protein levels in all cells treated with LPS, LTA, or Pam3CSK4 (Fig. 2⇑C). Taken together, these results consistently indicated that 15d-PGJ2 modulates the expression of TLR2 at the message and protein levels as well as cell surface levels following treatment of cells with TLR2 ligands, further supporting the modulation of TLR2 expression by 15d-PGJ2.
Downstream events of TLR2 activation are significantly affected by treatment of glia with 15d-PGJ2
Activation of TLR2 cause triggering complex signaling cascades, thereby resulting in robust production of inflammation-associated molecules including cytokines and NO (40, 43). Having determined that 15d-PGJ2 regulates the expression level of TLR2, we next carefully investigated the functional importance of this effect by testing the action of 15d-PGJ2 on TLR2 ligand-stimulated phagocytosis. Phagocytosis is accompanied by secretion of inflammatory mediators, and TLR2 has been closely implicated in phagocytic events (5). The phagocytic response of BV2 cells was measured using FITC-conjugated fluorescent beads. Cells were treated with or without LPS, LTA, or Pam3CSK4 for 15 h, followed by incubation with FITC-conjugated beads for 1 h. The uptake of fluorescent beads by microglia was then determined by flow cytometry. Treatment of microglia with LPS, LTA, or Pam3CSK4 considerably increased the cellular uptake of fluorescent beads, and these enhancements were apparently decreased in the presence of 15d-PGJ2 (Fig. 3⇓, A and B, and data not shown). Using fluorescence microscopy, we also observed the effects of 15d-PGJ2 on TLR2 ligand-dependent phagocytosis. Compared with untreated control cells, LPS- or Pam3-treated cells were somewhat flattened and aggregated, where the uptake of fluorescent beads was increased. However, in the presence of 15d-PGJ2, the uptake of fluorescent beads by microglia was significantly attenuated (Fig. 3⇓C). Furthermore, we found that 15d-PGJ2 significantly suppressed the transcription of MCP-1 and TNF-α, iNOS expression, and NO release in cells treated with LTA or Pam3CSK4 (data not shown). Overall, these results show that not only TLR2 expression but also TLR2-mediated inflammatory events are suppressed by treatment of 15d-PGJ2, and convincingly suggest that 15d-PGJ2 may regulate the activation of microglia, at least in part, by down-regulating TLR2 expression.
15d-PGJ2 does affect TLR2 ligand-stimulated phagocytosis. BV2 microglial cells were incubated with 100 ng/ml LPS (A), 10 μg/ml LTA (B), or 100 ng/ml Pam3CSK4 (C) for 15 h in the absence or presence of 2.5 μM 15d-PGJ2. The cells were then incubated with FITC-conjugated fluorescent beads for 1 h, and the cellular uptake of the fluorescent bead was determined by flow cytometric analysis and fluorescence microscopy as described in Materials and Methods. The mean fluorescence intensity (MFI) values are the mean ± SD of three independent experiments. ∗, p < 0.05; ∗∗, p < 0.01 when compared with ligand-treated cells, respectively.
15d-PGD2 and PGD2, but not PGE2 or arachidonic acid, modulate the expression of TLR2
PGs are inflammatory mediators that are produced via the metabolism of arachidonic acid by COX and a chain of PG synthase enzymes (18, 19). Because 15d-PGJ2 is derived from PGD2, a metabolite of PGH2, and PGD2 is now recognized to have not only proinflammatory but also anti-inflammatory properties during inflammation, we investigated whether PGD2 could also alter the ligand-dependent change of TLR2 expression. Like 15d-PGJ2, PGD2 significantly reduced the enhancement of TLR2 expression in cells treated with LPS, LTA, or Pam3CSK4 (Fig. 4⇓A and data not shown). In addition, 15d-PGD2, another metabolite of PGD2, also suppressed the ligand-triggered increase of TLR2 expression. Under these conditions, TLR2 ligand-stimulated induction of iNOS expression and phagocytosis was significantly attenuated (Fig. 4⇓, B and C).
Both 15d-PGD2 and PGD2, but not PGE2 or arachidonic acid (AA), affect TLR2 agonist-stimulated enhancement of TLR2 expression, iNOS expression, and phagocytosis. A, BV2 microglial cells were pretreated with 20 μM 15d-PGD2, 5 μM PGD2, 20 μM PGE2, or 30 μM AA for 1 h, and stimulated with 100 ng/ml Pam3CSK4 for 12 h. TLR2 expression levels were then analyzed using flow cytometry. B, BV2 microglial cells were treated with or without 10 μg/ml LTA for 18 h in the absence or presence of 2.5 μM 15d-PGD2 or PGD2. Protein extracts were then prepared and subjected to Western blotting using Abs against iNOS and α-tubulin (loading control). C, The cellular uptake of fluorescent beads was analyzed as described in Materials and Methods. The MFI values are mean ± SD of three independent experiments. ∗, p < 0.05; ∗∗, p < 0.01 when compared with stimulator-treated cells. Data shown are representative of at least four independent experiments.
We then tested whether TLR2 expression could be altered by PGE2, another metabolite of PGH2. In contrast to the above results, no significant changes in the levels of TLR2 were detected in cells treated with PGE2. In addition, arachidonic acid, the precursor of PGD2, had no effect on ligand-dependent changes in TLR2 expression (Fig. 4⇑A). Taken together, these results indicate that some 15d-PGJ2-related PGs such as PGD2 and 15d-PGD2, but not PGE2 and arachidonic acid, markedly attenuate the ligand-dependent enhancement of TLR2 expression.
Similar effects are observed in primary glial cells and peritoneal macrophage
Because our results demonstrated that some PGs could modulate the ligand-dependent change of TLR2 expression in BV2 microglial cells, we next examined the contribution of these PGs to the regulation of TLR2 expression in mouse and rat primary glial cells. Consistent with the results from BV2 cells, 15d-PGJ2, 15d-PGD2, and PGD2 reduced the ligand-stimulated increase of TLR2 expression in primary microglia and astrocytes (Fig. 5⇓A and data not shown). Moreover, treatment with such PGs also apparently attenuated the response of TLR2 to its ligands in primary peritoneal macrophage (Fig. 5⇓B). Similar effects were observed in the mouse macrophage cell line, J774 (data not shown). Overall, our results indicate that PGs such as 15d-PGJ2, PGD2, and 15d-PGD2 influence TLR2 expression in both primary glia and macrophage, providing further evidence that these PGs control TLR activation by down-regulating TLR2 expression.
Similar effects are observed in primary glial cells and peritoneal macrophage. A, Primary astrocytes were treated with 100 ng/ml LPS or 25 μg/ml brain gangliosides for 12 h in the absence or presence of 15d-PGJ2, then Western blot analysis was performed using TLR2 and α-tubulin Abs. B, Mouse primary peritoneal macrophage were stimulated with 100 ng/ml LPS for 12 h following pretreatment with 2.5 μM 15d-PGJ2 for 1 h. TLR2 expression was analyzed at the cell surface level by flow cytometric analysis. C, BV2 cells were untreated or treated with 100 ng/ml Pam3CSK4 for 6 h, after which cells were further incubated for 18 h with or without 2.5 μM 15d-PGJ2. TLR2 expression was analyzed at the cell surface levels by flow cytometry.
15d-PGJ2 and PGD2 are also capable of suppressing a pre-existing increase in TLR2 expression
The above findings prompted us to consider whether these PGs could modulate a pre-existing activation of glia. To address this more clinically relevant question, we examined the expression level of TLR2 in cells treated with 15d-PGJ2 at various times after treatment with TLR2 ligands. BV2 cells were left untreated or stimulated with 100 ng/ml Pam3CSK4 for 6 h, whereupon the cells were further incubated for 18 h with or without 2.5 μM 15d-PGJ2. Interestingly, the suppressive effects of 15d-PGJ2 on cell surface expression of TLR2 were consistently observed in cells receiving ligand 6 h before the addition of 15d-PGJ2 (Fig. 5⇑C). Similar results were observed in cells treated with PGD2 and 15d-PGD2 (data not shown). These results indicate that—in addition to preventing ligand-dependent increases of TLR2—15d-PGJ2 and PGD2 are capable of attenuating an ongoing ligand-induced increase in TLR2 expression.
Thiazolidinediones, representative PPARγ agonists, do not mimic the actions of 15d-PGD2 and PGD2 on TLR2 expression
PPARγ is a ligand-dependent nuclear receptor whose ligands include several PGs and antidiabetic thiazolidinediones such as rosiglitazone (44). 15d-PGJ2 is shown to be a natural endogenous ligand for PPARγ and to exert some of its anti-inflammatory effects by binding to PPARγ. Therefore, to gain insight into the action mechanism by which these PGs suppress TLR2 expression, we tested the possibility that their inhibitory effects on TLR2 expression could be mediated by PPARγ. For this, we examined whether potent synthetic PPARγ agonists such as rosiglitazone and pioglitazone could affect the cell surface expression of TLR2 in cells stimulated with LPS, LTA, or Pam3CSK4. Neither rosiglitazone nor pioglitazone had any effects on the cell surface expression of TLR2 at any of tested concentrations and time points (Fig. 6⇓). In addition, we did not detect any changes of TLR2 message and protein levels following treatment with thiazolidinediones (data not shown). Under the same conditions, we observed the significant activation of a PPARγ-responsive element-luciferase reporter construct and the inhibitory effects of synthetic PPARγ agonists on expression of inflammatory mediators (data not shown). These results indicate that the inhibitory actions of 15d-PGJ2 and PGD2 on the ligand-dependent enhancement of TLR2 expression do not occur through PPARγ.
Synthetic PPARγ agonists do not affect the ligand-dependent changes in TLR2 expression. BV2 microglial cells were treated with 100 ng/ml LPS, 10 μg/ml LTA, or 100 ng/ml Pam3CSK4 for 12 h in the presence or absence of 20 μM rosiglitazone (A) or pioglitazone (B), after which TLR2 expression was analyzed at the cell surface protein level. Data shown are representative of at least four independent experiments.
Attenuation of ligand-induced TLR2 expression is not dependent of DP1 or DP2
PGD2 binds and activates two distinct G protein-coupled receptors, DP1 and chemoattractant receptor-like molecule expressed on Th2 (CRTH2; also named DP2). It has been shown that the physiological effects of PGD2 and 15d-PGD2 are mediated in part by DP1 or DP2 (21, 28, 30). Thus, we questioned whether 15d-PGJ2, PGD2, and 15d-PGD2 could exert their effects on TLR2 expression through DP1 or DP2. To address it, we examined the effects of the selective DP1 agonist BW245C or the selective DP2 agonist DK-PGD2 on ligand-dependent change in TLR2 expression. As shown in Fig. 7⇓A, treatment with BW245C or DK-PGD2 did not reduce the increased levels of TLR2 message in activated primary microglia. Moreover, neither BW245C nor DK-PGD2 had any effect on the LTA-stimulated increase of TLR2 protein expression at all tested concentrations and time points, whereas BW245C, but not DK-PGD2, suppressed the increase of COX2 under the same conditions (Fig. 7⇓B). Similar results were observed in primary microglia and astrocytes treated with LPS or Pam3CSK4 (data not shown). Taken together, these results suggest that neither DP1 nor DP2 mediate the regulatory effects of PGD2, 15d-PGJ2, and 15d-PGD2 on ligand-dependent increase of TLR2 expression.
Neither the DP1 agonist BW245C nor the DP2 agonist DK-PGD2 attenuates the ligand-stimulated increase of TLR2 expression. A, Rat primary microglia was treated with 10 μg/ml LTA or 100 ng/ml LPS for 6 h following pretreatment with 50 nM BW245C or 0.5 μM DK-PGD2 for 1 h. Total RNA was then extracted for RT-PCR analysis, and cell extract was prepared for Western blot analysis. B, Primary astrocytes were treated with 10 μg/ml LTA for 12 h in the indicated concentrations of BW245C or DK-PGD2, then Western blot analysis was performed using TLR2, COX2, and α-tubulin Abs. Data shown are representative of at least three independent experiments.
Discussion
The TLR system has been extensively characterized over the past several years, and considerable progress has been made in identifying the involved ligand specificities and downstream signaling cascades. However, relatively little is known regarding how the TLR system is regulated to maintain the balance between defense functions and harmful effects. Recent studies have revealed that dysregulation of TLR signaling may contribute significantly to a variety of diseases (4, 45), and many efforts are now being made to understand the regulation of the TLR system and develop therapeutic strategies to stimulate or inhibit TLR signaling (46, 47, 48). Here, we provide important new results showing that some PGs may contribute to the regulation of the TLR2-associated inflammatory responses through down-regulation of TLR2 expression in the brain.
Many CNS diseases including Alzheimer disease, Parkinson’s disease, multiple sclerosis, ischemia, and cancer are accompanied by robust activation of immune and inflammatory responses; these effects are closely associated with glial cells, which are the main immune cells of the brain (17, 18, 49). Recently, we and other researchers have shown that microglia and astrocytes express TLRs, and TLR-mediated inflammatory signaling may play a crucial role in glial activation and inflammation-associated neuronal diseases (37, 38, 50, 51). In particular, several reports have shown that TLR2 and TLR4 might contribute to the initiation and progression of CNS pathologies (39, 40, 41, 42, 52, 53). We have therefore focused on investigating the regulatory mechanisms that limit excessive TLR-mediated responses and bring about the resolution of brain inflammation. Interestingly, we found that 15d-PGJ2, a cyclopentenone PG, significantly modulates TLR2 message, protein, and cell surface levels in microglia and astrocytes activated by Pam3CSK4, LTA, as well as LPS and gangliosides. Notably, 15d-PGJ2 appears to have differential effects on the various TLR types in glia; whereas TLR2 expression was regulated by 15d-PGJ2, expression of TLR1, 4, and TLR9 was not affected by 15d-PGJ2 treatment of ligand-stimulated cells (Fig. 1⇑). TLR expression has been shown to be strictly regulated in a stimulus-dependent, cell type-, and state-specific manner (54, 55, 56, 57). Our results suggest the possibility that 15d-PGJ2 may exert a role in the modulation of TLR2 expression, thus contributing to the resolution of TLR2-stimulated inflammation in the brain.
Some PGs have been reported to be produced during inflammatory episodes and have an anti-inflammatory role in the resolution of inflammation, in addition to well-known proinflammatory abilities in the pathogenic state (21, 22, 23, 24, 25, 58, 59). Therefore, we questioned whether not only 15d-PGJ2 but also its related PG metabolites could influence the ligand-stimulated increase of TLR2 expression. Interestingly, we found that both PGD2 and 15d-PGD2 significantly reduced the ligand-dependent enhancement of TLR2 expression and the events downstream of TLR2 activation, but neither PGE2 nor arachidonic acid had any effect on TLR2 expression (Fig. 4⇑). Consistent with this notion, we observed similar effects in primary microglia and astrocytes, and obtained the concordant data from experiments in primary macrophage and macrophage cell lines (Fig. 5⇑).
Our above results raised the question of how these PGs could influence the ligand-stimulated enhancement of TLR2 expression in glial cells. 15d-PGJ2 and PGD2 can activate PPARγ, which has been shown to exert roles in suppression of inflammation as well as adipocyte differentiation (21). In vivo and in vitro studies have provided that 15d-PGJ2 effectively suppresses inflammation, which is either dependent on or independent of PPARγ (60, 61, 62, 63, 64). Based on these reports, we examined whether PPARγ could mediate the inhibitory action of 15d-PGJ2 on TLR2 expression using synthetic PPARγ agonists such as rosiglitazone and pioglitazone. However, even though rosiglitazone and pioglitazone are more potent PPARγ agonists than 15d-PGJ2, they did not have any detectable effects on the ligand-stimulated increase of TLR2 cell surface expression (Fig. 6⇑). In addition, the TLR2 message and protein levels in our system were also unaffected by treatment with rosiglitazone or pioglitazone, whereas PPARγ-responsive element-luciferase reporter activity was significantly increased and the expression levels of proinflammatory mediators were attenuated under the same conditions. Thus, we conclude that the inhibitory actions of 15d-PGJ2 and PGD2 on the ligand-dependent enhancement of TLR2 expression occur independently of PPARγ activation in the brain.
How, then, might they regulate the ligand-stimulated enhancement of TLR2 expression independent of PPARγ? We next considered the possibility that the DP1 or DP2 receptors could mediate these effects because numerous physiological activities of PGD2 and its metabolites have been demonstrated to be dependent on these two G protein-coupled receptors (27, 28, 29, 30). Activation of the DP1 by a selective agonist has been shown to suppress asthma by modulation of lung dendritic cell function and induction of regulatory T cells (27, 28). In the CNS, Liang et al. (29) showed that PGD2 and the DP1 selective agonist mediated neuronal protection via the DP1 receptor in hippocampal neurons. Taniguchi et al. (30) also reported that PGD2 protected the neonatal brain from hypoxic ischemic injury mainly via DP1 receptors. However, we did not detect any effect of the DP1 selective agonist BW245C or the DP2 agonist DK-PGD2 in ligand-stimulated increase of TLR2 expression.
We have previously reported that JAK-STAT signaling contributes to activation of glia in response to several stimuli including LPS and gangliosides, and that inhibition of JAK-STAT signaling by SOCS, Src homology domain-containing tyrosine kinase, or pharmacological inhibitors significantly reduces the abnormal activation of glia (60, 65, 66). Thus, we next examined the possibility that the JAK-STAT-signaling pathway could be linked with the action of some PGs on TLR2 expression using pharmacological inhibitors. However, we did not observe any meaningful effects of JSI-124, an inhibitor for STAT, or AG490, an inhibitor for JAK, on the suppression of TLR2 up-regulation by PG (data not shown). In addition, we did not obtain results supporting an association between induction of SOCS 1–3 and regulation of TLR2 expression by PGs (data not shown). Although further research will be required to definitely elucidate it, it seems unlikely that STAT signaling may be directly linked with the regulation of TLR2 expression by PGs.
Another possibility is that they may act by modification of specific cellular molecules. Some PGD2 metabolites have been shown to interact directly with intracellular proteins and regulate signaling pathway through modifying protein function. For instance, 15d-PGJ2 inhibited NF-κB signaling through binding to IκB kinase and NF-κB (67, 68), and also inhibited AP-1 activity by interaction with c-Jun (69). In addition to inflammation-associated transcription factors, there are reports that 15d-PGJ2 also regulates the activities of several proteins, including H-Ras and estrogen receptor-α (70, 71). Based on these reports and our current findings, it is possible that such PGs may functionally associate with certain signaling molecules in the TLR2-specific signaling pathway. In this regard, we are now investigating this possibility.
In the CNS, the levels of PGs are low under normal conditions but increase during inflammatory processes. In vivo and in vitro studies have shown that some PGs including PGD2 and its related cyclopentenone PGs play a crucial role in the resolution of inflammation by acting as endogenous anti-inflammatory mediators although other reports dispute these properties (22, 23, 24, 25, 26, 27, 28, 64, 72, 73). Our present data indicate that some PGs can attenuate TLR2 expression even when administered following exposure to TLR2 ligands. This observation suggests that they may participate in balancing appropriate inflammatory responses by acting as feedback inhibitors of TLR2-dependent inflammatory responses. The capacity of such PGs to reduce ongoing inflammatory responses is notable in the development of novel therapeutic approaches for TLR2-associated diseases. Further study on the precise molecular mechanism underlying these effects should contribute to a better understanding of the inflammation system, and may lead to the development of new therapeutic interventions for existing and emerging inflammation-associated neuronal diseases.
Acknowledgments
We thank all the anonymous reviewers for their thoughtful comments and all the members of the Division of Fusion Technology at the National Cancer Center for their helpful discussions and suggestions.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
↵1 This work was supported by the National Cancer Center (0710320-1) and by KRF grant funded by the Korean Government (KRF-2005-041-E00103).
↵2 Address correspondence and reprint requests to Dr. Eun Jung Park, Immune and Cell Therapy Branch, National Cancer Center, Goyang, 410-351, Korea. E-mail address: ejpark{at}ncc.re.kr
↵3 Abbreviations used in this paper: SOCS, suppressor of cytokine signaling; DP, D prostanoid receptor; COX, cyclooxygenase; 15d-PGJ2, 15-deoxy-Δ12,14-PG J2; PPAR, peroxisome proliferator-activated receptor; LTA, lipoteichoic acid; iNOS, inducible NO synthase; DAPI, 4′,6′-diamidino-2-phenylindole hydrochloride.
- Received August 24, 2007.
- Accepted April 10, 2008.
- Copyright © 2008 by The American Association of Immunologists
References
In this issue
