PAFR activation of NF-κB p65 or p105 precursor dictates pro- and anti-inflammatory responses during TLR activation in murine macrophages

Platelet-activating factor receptor (PAFR) is a G protein-coupled receptor (GPCR) implicated in many diseases. Toll-like receptors (TLRs) play a critical role in shaping innate and adaptive immune responses. In this study, we investigated whether PAFR signaling changes the macrophages responsiveness to agonists of TLR2 (Pam3Cys), TLR4 (LPS), and TLR3 agonist Poly(I:C). Exogenous PAF inhibited the production of pro-inflammatory cytokines (IL-12p40, IL-6, and TNF-α) and increased anti-inflammatory IL-10 in macrophages challenged with Pam3Cys and LPS, but not with Poly (I:C). PAF did not affect mRNA expression of MyD88, suggesting that PAF acts downstream the adaptor. PAF inhibited LPS-induced phosphorylation of NF-κB p65 and increased NF-κB p105 phosphorylation, which is processed in the proteasome to generate p50 subunit. The PAF potentiation of IL-10 production was dependent on proteasome processing but independent of NF-κB transactivation domain. Inhibition of p50 abolished the PAF-induced IL-10 production. These findings indicate that the impaired transcriptional activity of the p65 subunit and the enhanced p105 phosphorylation induced by PAF are responsible for down regulation of pro-inflammatory cytokines and up regulation of IL-10, respectively, in LPS-challenged macrophages. Together, our data unveil a heretofore unrecognized role for PAFR in modulating activation of NF-κB in macrophages.

MTT assay. A total of 2 × 10 6 macrophages were plated in 12-well flat bottom plates and then stimulated with cPAF and LPS. After 8 h, the supernatants were removed and 500 μ L of 5 mg/mL MTT solution in RPMI were added to each well for 4 h. After removal of the medium, 200 μ L of DMSO were added to each well to dissolve the formazan crystals. The absorbance at 540 nm was determined using a spectrophotometer.

Measurement of cytokines.
Production of IL-12p40, IL-6, TNF-α , and IL-10 in the supernatant of the macrophage culture was measured using OptEIA TM Mouse Set ELISA kits (BD Pharmingen, San Diego, CA) according to the manufacturer's instructions.
Scientific RepoRts | 6:32092 | DOI: 10.1038/srep32092 PGE 2 quantification. PGE 2 production was measured in the supernatants of macrophage cultures by competitive immunoassay using a PGE 2 EIA kit (Cayman Chemical, Ann Arbor, MI) according to the manufacturer's instructions.

RNA isolation and quantitative PCR (qPCR).
Total RNA was extracted from cultured cells using TRIzol reagent (Ambion Life Technologies, Carlsbad, CA) and the concentration and purity of the samples were determined by spectrophotometer readings at 260 nm and 280 nm. Single-stranded cDNA was synthesized using the RevertAid First Strand cDNA Synthesis kit (Thermo Scientific, Rockford, IL). The following PCR primers were used: Myd88 (F: 5′ -TGA TGA CCC CCT AGG ACA AA-3′ ; R: 5′ -TCA TCT CCT GCA CAA ACT CG-3′ ) and GAPDH (F: 5′ -AGG TCG GTG TGA ACG GAT TTG-3′ ; R: 5′ -TGT AGA CCA TGT AGT TGA GGT CA-3′ ). Real-time qPCR was performed in a SYBR Green PCR Master Mix (Applied Biosystems, Life Technologies, Warrington, UK) using the Stratagene Mx3005P qPCR System with the following cycling conditions: initial denaturation and enzyme activation for 10 min at 95 °C, followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min. Data were normalized to GAPDH expression and the relative abundance of transcripts was calculated by the comparative 2 −ΔΔC T method as described previously 20 .

Results
PAF modulates the production of inflammatory cytokines induced by TLR2 and TLR4 in macrophages. Initially, to determine the role of PAF in TLR-induced macrophages response, we challenged macrophages with PAF plus the TLR agonists LPS (TLR4), Pam3Cys (TLR2) and Poly(I:C) (TLR3). Our results show that PAF signaling inhibited both LPS and Pam3Cys-induced production of IL-12p40 after both 8 and 24 h of exposure to stimuli, whereas IL-6 and TNF-α levels were decreased after 8 h of stimulation by LPS. However, PAF potentiated the production of IL-10 induced by both agonists at both time points tested (Fig. 1A,B). However, PAF did not affect Poly(I:C) -induced cytokine production ( Fig. 1C), indicating that PAF actions on TLR activation are restricted to TLR2 and TLR4. In addition to changes in gene transcription, PAF could influence the activity of enzymes involved in macrophage activation, such as cyclooxygenase 2 (COX2), which generates different prostanoids, including prostaglandin E 2 (PGE 2 ). As demonstrated in Fig. 2A, concomitant treatment of macrophages with PAF plus LPS or Pam3Cys inhibited COX2 expression after 24 h of stimulation as compared with either agonist alone. Furthermore, these findings correlated with decreased PGE 2 accumulation in macrophage cultures stimulated as above (Fig. 2B).

PAF does not influence Myd88 expression in TLR-stimulated macrophages.
To further investigate the downstream effects of PAF on TLR activation, we studied whether PAF inhibits the expression of Myd88 mRNA, which could consequently affect cytokine production induced by LPS and Pam3Cys. Our results show that while both LPS and Pam3Cys enhance Myd88 expression, PAF did not influence TLR-enhanced Myd88 expression (Fig. 3A,B), suggesting that PAF influences signaling effectors downstream of MyD88 actions during TLR activation.

PAFR differentially regulates NF-κB subunits to modulate TLR-induced cytokine generation.
Next, we sought to study the signaling programs downstream of MyD88 that are involved in PAF effects on TLR activation. Since NF-κ B activation drives both pro-and anti-inflammatory networks, we speculated that PAF could differentially influence NF-κ B dimer formation. Initially we investigated the effect of PAF on inhibitory protein Iκ Bα expression and did not observe any effect of PAF, neither in LPS nor in Pam3Cys-mediated Iκ Bα activation (Fig. 4A), indicating that the effects of PAF could be downstream of changes in Iκ Bα degradation. Another regulatory level of NF-κ B activation is p65 phosphorylation, an event required for optimal NF-κ B transcriptional activity 21 . As seen in Fig. 4B, PAF decreased the phosphorylation of p65 induced by LPS in 15 min, indicating that PAF inhibits the production of pro-inflammatory cytokines by affecting p65 transcriptional activity.
In order to identify the molecular pathways involved in the PAF-mediated effects, we enquired whether other transcription factors and receptors involved in IL-10 production, such as CREB (cAMP response element binding protein) and PPAR-γ (peroxisome proliferator-activated receptor-γ ) 22,23 , could be involved in PAF-induced anti-inflammatory programs during TLR activation. Neither CREB (KG-501) nor PPAR-γ (GW9662) inhibitors prevented PAF-enhanced IL-10 production (Fig. 5A,B).
Furthermore, when we investigated whether PAF could change the phosphorylation of STAT3 (which drives IL-10 expression and action) we did not observe changes in STAT3 expression and activation during PAF + LPS co-incubation (Fig. 6A,B).   We next investigated whether NF-κ B activation was involved in PAF-induced IL-10 potentiation. We first tested the proteasome inhibitor Calpain inhibitor acetyl-L-leucyl-L-leucyl-L-norleucinal (ALLN), that prevents the proteolysis of Iκ Bα and Iκ Bβ by the ubiquitin-proteasome complex 24 .
The ALLN treatment did not affect macrophages viability at 1 and 10 μ M, measured by MTT assay (Fig. 7A). When proteasome activity was abrogated, indicated by IL-12 inhibition at 10 μ M (Fig. 7B), the IL-10 production was also inhibited (Fig. 7C), suggesting that IL-10 production is dependent on proteasome processing.  We next tested the pyrrolidinedithiocarbamate (PDTC) treatment, that inhibits the transactivation domain of NF-κ B subunits 25 . The PDTC treatment did not affect cell viability at all tested doses (Fig. 7D). At 50 and 100 μ M, PDTC inhibited IL-12 production (Fig. 7E) indicating that it effectively inhibited the transactivation domain, but did not affect PAF-induced IL-10 production (Fig. 7F), suggesting that the transactivation domain is not involved in PAF-induced IL-10 potentiation triggered by LPS.
The transactivation domain is present in the p65, c-Rel and Rel B subunits of NF-κ B, but not in p100/52 and p105/p50 17 . NF-κ B p105/p50 is known to enhance the IL-10 transcriptional machinery 26 . Therefore, we speculated that PAF acts on NF-κ B p105/p50, increasing TLR-induced IL-10. Our data show that PAF enhances LPS phosphorylation of the p105 subunit compared with LPS alone (Fig. 8A), but the same was not observed with Pam3Cys. To confirm that p105 is implicated in the IL-10 increase induced by PAF + LPS treatment, macrophages were pretreated for 1 h with a p105/p50 inhibitory peptide followed by LPS stimulation for 8 h. Our data show that p105/p50 inhibition, but not the peptide control, reduced PAF-mediated IL-10 increase (Fig. 8B).
Taken together, our results imply that PAF drives pro-and anti-inflammatory responses by controlling p65 and p105 NF-κ B subunits phosphorylation during TLR activation in macrophages.

Discussion
Here we investigated the effect of PAFR activation on macrophage responsiveness to TLR agonists. In summary, our results show that PAF reduces the generation of pro-inflammatory cytokines and PGE 2 and increases the generation of anti-inflammatory IL-10 in macrophages stimulated with LPS or Pam3cys. Similar results were described by Jeong et al. 27 in LPS-stimulated murine peritoneal macrophages where PAF was shown to potentiate IL-10 production and inhibit IL-12, IL-6, and TNF-α . IL-10 is the prototypic anti-inflammatory cytokine that acts by binding to the IL-10 receptor present on various cell types 28 . Mice genetically deficient in the IL-10 receptor develop an exacerbated inflammatory response and autoimmune diseases. Impaired IL-10 responses have been linked to human diseases, including inflammatory bowel diseases, arthritis, asthma, and psoriasis [29][30][31] . PAFR KO mice also exhibit (A) Peritoneal macrophages were exposed to LPS 100 ng/mL or Pam3Cys 100 ng/mL alone or combined with cPAF 100 nM. At the indicated times, total RNA was isolated and samples were probed for Stat3 mRNA levels by qPCR. (B) Protein samples were incubated with Ab to phosphorylated STAT3. Cropped blot is shown from one representative experiment. Full-length gels are included in the Supplementary information file. Densitometric analysis of STAT3 protein levels is represented as mean ± SD from 5 independent experiments. For qPCR, the data are representative of one independent experiment and presented as mean ± SD of n = 5 animals per group. *p < 0.05 compared with control. A. U., arbitrary units.
Scientific RepoRts | 6:32092 | DOI: 10.1038/srep32092 sterile inflammation and insulin resistance by increasing pro-inflammatory macrophages in the adipose tissue 32 . Thus, the PAF/IL-10 axis might have an important role in preserving homeostasis by down regulating inflammation.
In our study, PAF induced a regulatory phenotype (IL-10 high /IL-12 low ) in both LPS and Pam3Cys-stimulated macrophages. Regulatory macrophages are known to modulate inflammatory responses. Ziegler et al. 33 demonstrated that adoptive transfer of regulatory macrophages increases local and systemic IL-10 production while it attenuates allergic airway inflammation, decreasing allergen-specific IgE levels, eosinophil influx to the airways, Th2 cytokine production, and the production of mucus in the lungs. A regulatory phenotype was also observed in Fcγ R-activated macrophages that were able to reduce lethal endotoxemia by preventing pro-inflammatory cytokine responses 34 .
The PAF effects on the regulatory phenotype are not limited to macrophages, as similar effects were observed in LPS-stimulated murine dendritic cells, potentiating IL-10 and reducing pro-inflammatory cytokines and, in addition, downregulating antigen-presenting function 35 . We also recently found that transfer of dendritic cells that were treated with a PAFR antagonist increased antigen-specific lymphocyte proliferation 35 . Based on these and other line of evidence, we propose that activation of PAFR by endogenous ligands might be important to the "fine-tune" regulation of inflammatory and immune responses. As endogenous ligands, several oxidized phospholipids, including those expressed in apoptotic cell plasma membranes, bind to PAFR 36 . Also, UV radiation and several environmental hazards were shown to induce the production of molecules that bind to PAFR 37 . Chemo-and radiotherapy also induce PAFR ligands and this was shown to promote tumor growth, in part because they induce regulatory macrophages 38 .
The dual effect of PAF, inhibition of pro-inflammatory cytokines and potentiation of the anti-inflammatory IL10, was dependent on the inhibition of NF-κ B p65 phosphorylation and enhancement of NFκ B p105/50 phosphorylation, respectively. We have previously shown that the LTB 4 receptor BLT1, a receptor that is coupled to Gα i protein, potentiates MyD88-dependent TLR activation, but not TRIF-dependent activation 39,40 . In the present study, PAF did not affect MyD88 expression, indicating that the PAF effects on pro-inflammatory cytokine production are downstream of adaptors.
We found that PAF does not affect TLR-induced Iκ Bα phosphorylation, but inhibits NF-κ B p65 phosphorylation at serine 536. Different GPCR subunits are known to activate NF-κ B. Plasma membrane-associated PAFR is coupled to Gα q and activates NF-κ B. Bradykinin receptor 2 utilizes a signaling pathway that involves Gα q for NF-κ B activation by enhancing Iκ Bα phosphorylation 41 . The receptor for FMLP (N-formyl-Met-Leu-Phe), coupled to Gα I also activates NF-κ B in neutrophils 42 . Interestingly, PGE 2 also activates NF-κ B via the Gα s -coupled receptors, EP2 and EP4, in a PKA-dependent manner 43 . However, whether different G proteins coupled to PAF receptors influence different pathways that culminate in pro-or anti-inflammatory events remain to be investigated.
It is well documented that the IKK complex, which triggers Iκ Bα degradation, is involved in p65 phosphorylation 44 . Yang et al. 45 described an essential role of IKKβ in LPS-induced S536 p65 phosphorylation and, according to our results, this phosphorylation is reduced by PAF treatment. Thus, PAF could influence the actions of phosphatases involved in p65 activation in a similar manner as described for WIP-1, a p53-induced phosphatase 1 that targets the Ser536 p65 subunit, which acts as a negative regulator of NF-κ B signaling 46 .
IL-10 can be produced by macrophages as result of multiple signaling pathways and transcription factors after TLR activation 47,48 . Here we have demonstrated that PAF-induced potentiation of IL-10 production triggered by LPS is dependent on proteasome processing and NF-κ B p105/50, but independent of CREB, STAT-3, and PPAR-γ . Peritoneal macrophages were simultaneously exposed to LPS (100 ng/mL) and cPAF (100 nM). (A) After 15 min of stimulation, cell lysates were subjected to immunoblotting using specific Ab against phosphorylated NF-κ B p105/p50. Cropped blot is shown from one representative experiment. Full-length gels are included in the Supplementary information file. Densitometric analysis of p105 protein levels is represented as mean ± SD from 5 independent experiments. (B) Macrophages pretreated for 30 min with NF-κ B p105/p50 inhibitor or peptide control were stimulated with LPS 100 ng/mL and cPAF 100 nM. After 8 h, the levels of IL-10 in the culture supernatants were measured by ELISA. The data are representative of one independent experiment and presented as mean ± SD of n = 5 animals per group. *p < 0.05 compared with control. # p < 0.05 compared with cPAF + LPS. A.U., arbitrary units. It has been demonstrated that NF-κ B p50 homodimers are transcriptionally active and promote the production of IL-10 26,49 . The product of the Nfκb1 gene, p105, originates NF-κ B subunit p50 after proteasome processing, and functions as an Iκ B-like molecule by sequestering p50 in the cytoplasm 50 . After proteasome processing, p105 generates p50 48 . Additionally, macrophages from p105 KO mice fail to produce IL-10 when stimulated with LPS 51,52 . Hence, our results showing that PAF enhances LPS-induced p105 phosphorylation suggest, along with the fact that inhibition of p105/p50 prevented PAF-induced enhanced IL-10 production, that p105 has an essential role in mediating PAF-induced enhancement of IL- 10. It has been shown that the IKK complex activates the p65 subunit and regulates p105 phosphorylation 53 . More specifically, IKKα and IKKβ directly phosphorylate p105 on serine 927, which resides in a conserved motif homologous to the IKK target sequence in Iκ Bα 54,55 , suggesting that PAF-mediated p105 activation could be dependent on the IKK complex. However, this needs to be further investigated.
Together, our data show that PAF interferes with selected pathways induced by TLRs in macrophages, driving them to a regulatory phenotype via IL10. PAF affects NF-κ B activation by inhibiting p65 transcriptional activity and leading to a decrease in pro-inflammatory mediator formation and enhancement of IL-10 production in a manner dependent on the action of the NF-κ B precursor p105. The regulation of TLR-mediated responses by PAF may provide potential new ways to manipulate the innate immune response.