Resveratrol decreases CD45+CD206− subtype macrophages in LPS‐induced murine acute lung injury by SOCS3 signalling pathway

Abstract Acute lung injury/acute respiratory distress syndrome (ALI/ARDS) are life‐threatening condition in critically ill patients. Resveratrol (Res), a natural polyphenol, has therapeutic effect in animal model with ALI; however, whether Res attenuates ALI through modulation of macrophage phenotypes in the animal model remains unknown. We in this study treated LPS‐induced murine ALI with 30 mg/kg Res and observed significantly reduced severity of ALI in the Res‐treated mice 48 hours after Res treatment. Neutrophil infiltrates were significantly reduced, accompanied with lower infiltration of CD45+Siglec F− phenotype macrophages, but higher population of CD45+Siglec F+ and CD45+CD206+ alternatively activated macrophages (M2 cells) in the Res‐treated mice with ALI. In addition, the expression of IL‐1beta and CXCL15 cytokines was suppressed in the treated mice. However, Res treatment in mice with myeloid cell‐restricted SOCS3 deficiency did not significantly attenuate ALI severity and failed to increase population of both CD45+Siglec F+ and CD45+CD206+ M2 subtype macrophages in the murine ALI. Further studies in wild‐type macrophages revealed that Res treatment effectively reduced the expression of IL‐6 and CXCL15, and increased the expression of arginase‐1, SIRT1 and SOCS3. However, macrophages’ lack of SOCS3 expression were resistant to the Res‐induced suppression of IL‐6 and CXCL15 in vitro. Thus, we conclude that Res suppressed CD45+Siglec F− and CD45+CD206− M1 subtype macrophages through SOCS3 signalling in the LPS‐induced murine ALI.

lung tissues and blood circulation, sharing different cell subtypes.
The different macrophage subtypes exclusively release high amount of pro-inflammatory, anti-inflammatory cytokines and chemokines, participating in the pathogenesis of ALI/ARDS, asthma and other inflammatory lung diseases. 2,3 There are no effective therapeutics for ARDS so far. Patients are usually treated with low tidal volume ventilation and fluid restriction, 4 high dose of corticosteroid 5 and antioxidants, such as N-acetylcysteine. 6 However, the therapeutic approaches produced elusive therapeutic effects and did not improve patient survival rate. Thus, it is essential to explore novel and more effective therapeutics for ARDS.
Resveratrol (Res), a phytoalexin antioxidant from red grapes, has anti-inflammatory properties. 7,8 Studies in vivo have confirmed that Res can attenuate the severity of ALI in animal models. 9,10 Res or Res-curcumin hybrids can significantly attenuate disease severity of ALI, accompanied with lower production of pro-inflammatory cytokines and chemokines, such as TNF-alpha, IL-1beta, IL-6, IL-12, IL-33, MIP-2 and IL-18. In contrast, the anti-inflammatory cytokines and molecular mediators, such as IL-10, heme oxygenase-1 (HO-1),Nrf2 (nuclear factor-erythroid 2 related factor) and SOD (superoxide dismutase), were increased in the Res-treated animal models. 10,11 The beneficial effects may be associated with up-regulation and activation of SIRT1, a nicotinamide adenine dinucleotide (NAD+dependent deacetylase sirtuin 1). 12 It was reported that SIRT1 has multiple biological functions on variable cells, by suppressing cell activation, preventing cell senescence, etc. A variable signalling pathways are involved in the beneficial effects, such as NF-kappa B and JAK/STAT signalling pathway 7 and PI3K/Nrf2/HO-1 signalling pathway. 10 In the Res-treated cells, the levels of pro-inflammatory cytokines (TNF-alpha, IL-1beta, IL-6) and chemokines (CCL2/MCP-1, CCL4/MIP-1beta, CCL5/RANTES, CXCL10/IP-10) were significantly suppressed. 13 However, knockdown of SIRT1 diminished the antiinflammatory effects of Res, indicating the critical role of SIRT1 in the Res-mediated suppression of inflammation. 14 Furthermore, Res may suppress inflammation through inducing macrophages pyroptosis and apoptosis. 15,16 Therefore, Res has variable biological function on cells.
IL-6 and IL-22 induce activation of signal transducer and activator of transcription-3 (STAT3) through acetylation in lysine (Lys)685 and phosphorylation in tyrosine (Tyr)705 residues, but activation of STAT3 is suppressed by SIRT1, a class III deacetylase. 17,18 More activation of STAT3 induces more expression of cytokine expression. Blocking STAT3 activation by small-molecule STAT3 inhibitor (LLL12) can suppress the expression of IL-1beta, IL-6, TNF-alpha, iNOS, CCL2 and MHC class II in macrophages and inflammatory cells from LPS-induced ALI mouse model, indicating the pro-inflammatory function of STAT3. 19 It is known that SOCS3 is a negative regulator of STAT3. Activation of STAT3 induces transcriptional up-regulation of SOCS3 and subsequently suppresses STAT3 activation in the inflammatory condition. The negative feedback loop of STAT3/SOCS3 signalling has important implication in overcoming excessive tissue inflammation and preventing tissue damage during inflammatory immune responses. 17,20 Our previous study and report from other group 21 indicated that STAT3/SOCS3 signalling suppressed macrophage polarization and activation in the murine ALI mouse model, whereas disruption of STAT3/SOCS3 signalling in SOCS3 myeloid cell-restricted conditional knock-out mice (cKO) significantly enhanced the severity of LPS-induced ALI and sepsis, accompanied   with high expression of M1 cell genes (IL-1beta, IL-6, IL-12, IL-

| Animal procedure
Eight-to ten-week-old male wild-type (WT) C57BL/6 mice were obtained from the Shanghai Biomodel Organism Science & Technology.
Animal protocol was reviewed and approved by the laboratory animal care and use committee of Zhongshan Hospital, Fudan University. All animal experiments were performed in the Zhongshan Hospital, Fudan University. SOCS3 cKO were obtained by breeding SOCS3fl/fl mice with Lyz2-Cre transgenic mice under the control of myeloid cell-restricted lysozyme 2 (Lyz2) promoter, and the mice with SOCS3 −/− phenotype were used for this study. 2

| Isolation and culture of bone marrow-derived macrophages (BMDMs)
To isolate murine BMDMs, bone marrow cells were flushed from the femurs and tibiae of WT or cKO mice. These cells werecultured in RPMI 1640 medium containing 10% FBS and 20 ng/mL of murine M-CSF (PeproTech) for 7 days to obtain BMDMs. Cell purity was determined by FACS analysis for CD11b and F4/80 (>96%). The cells were pre-treated with indicated concentration of Res 1 hour prior to LPS 500 ng/mL treatment for 24 hours. The untreated cells and the cells treated Res and LPS alone were used as controls. Gene and protein expression in the treated cells were analysed by qRT-PCR, Western blot analysis.

| ELISA assay for cytokines
IL-1beta and CXCL15 protein concentration in BAL and lung digests were measured by ELISA kit according to the manufacturer's instructions (R&D systems Inc).

| Western blot analysis
The concentration of acetylated and phosphorylated STAT3, total STAT3 and SIRT1 protein in lung tissue extracts and macrophage lysates were analysed by Western blot analysis. The procedure was previously described. 2 Primary antibodies for Western blot analysis include rabbit anti-total STAT3 (Clone: 79D7), rabbit anti-phosphorylated STAT3 at residue Tyr-705, rabbit anti-acetylated STAT3 at residue Lys-685, rabbit anti-SIRT1 (Hangzhou HuaAn Biotech). The antimouse GAPDH antibody was used as a loading control. The blots were washed with TBST buffer and incubated with horseradish peroxidase (HRP)-conjugated anti-rabbit immunoglobulin (Ig) and then developed with ECL substrate solution (Amersham Biosciences). Protein expression was quantitatively analysed on ImageJ software by the ratio of densitometric intensity of target protein to internal control GAPDH.

| Immunostaining assay
Acetylated STAT3 expression and SIRT1 expression in the lung tissues of treated mice and cells were analysed by immunostaining assay. Briefly, lung tissue sections and cells were fixed with 4% paraformaldehyde, followed by incubation with 0.05% Triton X-100 and blocking buffer (10% goat serum in PBS). The sections and cells were incubated with first antibody rabbit anti-acetylated STAT3 and SIRT1 (1:200 dilution) for 3 hours and followed by secondary antibody HRP-conjugated anti-rabbit IgG or Cy3-conjugated anti-rabbit IgG (1:500 dilution) for 2 hours. The positive cells were developed by 3,3'-diaminobenzidine (DAB) and visualized under contrast microscope or fluorescence microscope.
The relative gene expression was analysed using the 2 −ΔΔCT method.

| Statistical analysis
Results are presented as mean ± standard error (SE) for each group.
Student's t test was used to determine statistical significance between two groups. One-way analysis of variance (ANOVA) followed by Tukey's multiple comparisons test was performed for parametric multivariable analysis on GraphPad Prism 7 software. The data were considered statistically significant for P values < .05.

| Res attenuated acute lung inflammation in mice with LPS-induced ALI
To investigate the effects of Res on murine ALI and underlying immunological mechanisms, 30 mg/kg Res (Res/LPS group) was i.p.  Figure 1A,B). Consistent with the results above, the total cell counts ( Figure 1C), absolute number of neutrophils ( Figure 1D) and percentage of CD3 + CD4 + T cells ( Figure 2A) were significantly increased in BAL of mice with ALI, but they were markedly reduced in the mice co-treated with Res and LPS.
To further investigate the changes of lung macrophages and subtypes associated with ALI suppression by Res, we analysed macrophages and their phenotypes in the treated mice.

| Res attenuated pro-inflammatory cytokine expression and activation of STAT3
CXCL15 attracts neutrophils into the inflamed sites, favouring pro-inflammatory M1 macrophage polarization. 22 To investigate whether these mediators are modulated by Res, we measured IL-1beta and CXCL5 expression in the lung tissues by ELISA ( Figure 4A-B) and qRT-PCR analysis ( Figure 4C).

| Res attenuated ALI by STAT3/SOCS3 signalling pathway
Our previous study indicated that STAT3/SOCS3 signalling path- Data were presented as mean ± standard error, n = 5, *P < .05 vs LPS group. E, M1 and M2 subtype macrophages in lung digests and CXCL15 protein expression in BAL were, respectively, analysed by flow cytometry and ELISA analysis. Data were presented as mean ratio of M1/M2 and expression of CXCL15 ± standard error, n = 5, *P < .05 vs LPS group were consistent with the results of total cell counts ( Figure 5C) and absolute number of neutrophils ( Figure 5D) in BAL. Flow cytometry analysis showed that the percentage of F4/80(low)Ly6G + phenotype neutrophils in BAL and lung tissues in WT mice with ALI were significantly suppressed, but not in those of cKO mice ( Figure 5E). More studies on macrophage subtypes by flow cytometry showed that the population of CD45 + Siglec F − subtype infiltrating macrophages was significantly reduced in the lung of Res-treated WT mice with ALI (P < .05), but not in the lung of cKO mice with ALI ( Figure 5F). Furthermore, we observed the heightened ratio of M2/M1 subtype macrophages in the Res-treated WT mice, but not altered in the Res-treated cKO mice ( Figure 5G).
To further investigate whether pro-inflammatory cytokines and other mediators were affected by Res treatment in cKO mice, we measured CXCL15 and IL-1beta expression by ELISA assay.
The results showed a significantly reduced expression of CXCL15 and IL-1beta expression at protein ( Figure 6A

| Res suppressed M1 cell-biased polarization of BMDMs from WT mice, but not from cKO mice
To further confirm whether the suppressive effects of Res in murine ALI are mediated by STAT3/SOCS3 signalling in macrophages, we treated BMDMs with different concentrations of Res. The results indicated that LPS activated STAT3, inducing phosphorylation and acetylation of STAT3, but the STAT3 activation was moderately suppressed by Res treatment in BMDMs ( Figure 7A-C). Because SOCS3 is a negative feedback regulator of STAT3, our further analysis by F I G U R E 4 Res attenuated cytokine expression and STAT3 activation in WT mice with ALI. IL-1beta in lung tissues (A) and CXCL15 (B) protein in BAL were measured by ELISA assay. IL-1beta mRNA transcripts in lung tissues (C) was quantitatively analysed by qRT-PCR. Data were presented as mean ΔΔCt relative to GAPDH internal control. n = 6. * P < .05 vs LPS group. D, Acetylated STAT3 at Lys-685, phosphorylated STAT3 at Tyr-705 and SIRT1 protein expression in the lung tissues were analysed by Western blot analysis. GAPDH was used as loading control. One representative blot is shown. The acetylated (E) and phosphorylated STAT3 (F) were quantitatively analysed by Image J software. Data were presented as ratio of target protein densitometric density to GAPDH. n = 6, * P < .05 vs LPS group; #P < .05 vs PBS group. Data were statistically analysed by one-way analysis of variance (ANOVA) followed by Tukey's multiple comparisons test. G, Immunostaining for acetylated-STAT3 expression in lung tissues.   Res alone or in combination with 500 ng/mL LPS. GAPDH was used as internal loading control. One representative blot is shown. B, Quantitative analysis of phosphorylated-STAT3 and acetylated STAT3 expression on Western blots by ImageJ software. Data were presented as ratio of target protein densitometric density to loading control GAPDH. *P < .05 vs the Res-untreated LPS control. Data were statistically analysed by one-way analysis of variance (ANOVA) followed by Tukey's multiple comparisons test. C, Immunostaining for acetylated STAT3 and SIRT1 expression in the treated BMDMs. The cells were incubated with rabbit anti-acetylated STAT3 antibody or anti-SIRT1 antibody and followed by Cy3-conjugated anti-rabbit IgG (Red). Nuclei were stained with DAPI (Blue). 50 μmol/L Res suppressed acetylation of STAT3 and increased expression of SIRT1, but induced nuclei damage and cell apoptosis (white arrow). One representative photograph is shown. D, qRT-PCR analysis for SOCS3, SIRT1, arginase-1 (Arg-1) and iNOS mRNA transcripts. The cells were treated with the indicated concentration of Res for 24 h in triplicate wells. Data were presented as mean of ΔΔCt relative to GAPDH ± standard error. E, qRT-PCR analysis for IL-6 and CXCL15 mRNA transcripts in WT or SOCS3 cKO mice-derived BMDMs 24 h after treatment with 500 ng/mL LPS with or without 50 µmol/L Res pre-treatment. Data were presented as mean ΔΔCt relative to GAPDH ± standard error, n = 3. *P < .05 vs Res-untreated control. It is documented that SIRT1 is critically involved in variable cell biological function. [22][23][24] Loss of SIRT1 in SIRT1 knock-out mice delayed locomotor recovery, in association with more expression of pro-inflammatory cytokines and population of M1 type macrophages in a mouse model with spinal cord injury. 25 Consistent with the previous report, 22  In conclusion, this study indicated that Res suppressed lung inflammation of mice with LPS-induced ALI, in association with significantly reduced neutrophil infiltration and production of F I G U R E 8 Schematic diagram of resveratrol (Res)-mediated suppression of ALI through STAT3/SOCS3 signalling. LPS induces oxidative stress and activates STAT3, subsequently induces IL-6, IL-1beta and CXCL15 expression, ultimately causes ALI/ARDS. Activation of STAT3 up-regulates expression of SOCS3 that negatively suppresses STAT3 signalling-mediated pro-inflammatory cytokine expression and maintains a balance between pro-inflammatory and anti-inflammatory responses. Res increases the expression and activation of anti-inflammatory SIRT1 and SOCS3, subsequently lead to lower expression of IL-6, IL-1beta and CXCL15 pro-inflammatory cytokines. Res significantly modulated macrophage activation and polarization, with enhanced polarization of anti-inflammatory M2 and CD45 + Siglec F + subtype macrophages by Res treatment. STAT3/SOCS3 signalling critically participated in macrophage polarization and mediated the suppressive effects of Res in mice with ALI ( Figure 8). Thereby, this study gained insights into a new role and underlying molecular mechanisms of Res in the therapy of murine ALI. The results provided a solid molecular basis for application of Res to clinical trial in the future.

| CLINI C AL PER S PEC TIVE S
• Res has anti-inflammatory and antioxidant effects. Previous studies showed that Res attenuated the severity of ALI in animal models, associated with lower expression of pro-inflammatory cytokines.
However, whether SOCS3 signalling pathway in macrophages is involved in the attenuation of ALI remains unknown.
• In this study, we compared the effects of Res on lung inflammation of murine ALI and macrophage phenotypes in WT and SOCS3 cKO mice. The results revealed that Res suppressed lung inflammation in murine ALI, associated with the suppressed CD45 + Siglec F − and CD45 + CD206 − M1 subtype macrophages, but increased CD45 + Siglec F + and CD45 + CD206 + M2 subtype macrophages in murine ALI.
However, the effects were not observed in SOCS3 cKO mice.
• The results gained insights into the important role of STAT3/ SOCS3 signalling in polarization of CD45 + Siglec F − subtype and CD45 + CD206 − M1 subtype macrophages, as well as resveratrol-mediated attenuation of lung inflammation in murine ALI.
Therefore, modulation of macrophage subtypes by targeting SOCS3 signalling is a promising therapeutic approach in the treatment of murine ALI and patients with ARDS in the future.

CO N FLI C T O F I NTE R E S T
The authors confirm that there are no conflicts of interest.