Edited by: Guixiu Shi, Xiamen University, China
Reviewed by: Shanzhong Gong, University of Texas at Austin, United States; Valerio Chiurchiù, Università Campus Bio-Medico, Italy
Specialty section: This article was submitted to Inflammation, a section of the journal Frontiers in Immunology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Macrophages are crucially involved in the pathogenesis of rheumatoid arthritis (RA). Macrophages of the M1 phenotype act as pro-inflammatory mediators in synovium, whereas those of the M2 phenotype suppress inflammation and promote tissue repair. SIRT1 is a class 3 histone deacetylase with anti-inflammatory characteristics. However, the role played by SIRT1 in macrophage polarization has not been defined in RA. We investigated whether SIRT1 exerts anti-inflammatory effects by modulating M1/M2 polarization in macrophages from RA patients. In this study, SIRT1 activation promoted the phosphorylation of an adenosine monophosphate-activated protein kinase (AMPK) α/acetyl-CoA carboxylase in macrophages exposed to interleukin (IL)-4, and that this resulted in the expressions of M2 genes, including MDC, FcεRII, MrC1, and IL-10, at high levels. Furthermore, these expressions were inhibited by sirtinol (an inhibitor of SIRT1) and compound C (an inhibitor of AMPK). Moreover, SIRT1 activation downregulated LPS/interferon γ-mediated NF-κB activity by inhibiting p65 acetylation and the expression of M1 genes, such as CCL2, iNOS, IL-12 p35, and IL-12 p40. Macrophages from SIRT1 transgenic (Tg)-mice exhibited enhanced polarization of M2 phenotype macrophages and reduced polarization of M1 phenotype macrophages. In line with these observations, SIRT1-Tg mice showed less histological signs of arthritis, that is, lower TNFα and IL-1β expressions and less severe arthritis in the knee joints, compared to wild-type mice. Taken together, the study shows activation of SIRT1/AMPKα signaling exerts anti-inflammatory activities by regulating M1/M2 polarization, and thereby reduces inflammatory responses in RA. Furthermore, it suggests that SIRT1 signaling be viewed as a therapeutic target in RA.
Rheumatoid arthritis (RA) is a chronic inflammatory disease that can activate the immune system
Histone/protein deacetylase SIRT1 and adenosine monophosphate-activated protein kinase (AMPK) work in tandem to control cell metabolism, transcriptional gene expression, neuroprotection, and inflammation (
Nevertheless, the function of SIRT1 in RA remains controversial. The activity and expression of SIRT1 have been shown to be diminished in the peripheral blood mononuclear cells (PBMCs) of RA patients (
Given SIRT1/AMPKα signaling has an important role in regulating macrophage polarization to an anti-inflammatory M2 phenotype, and that activation of this signaling may attenuate joint inflammation in RA, we sought to establish whether SIRT1 induces AMPKα phosphorylation and subsequent NF-κB downregulation, leading to the upregulations of M2-associated cytokines and to the downregulations of the expressions of M1-associated pro-inflammatory mediators in macrophages obtained from RA patients or SIRT1-overexpressing mice. In addition, we assessed a prevention of synovial inflammation and joint destruction in SIRT1 transgenic (Tg) mice with collagen-induced arthritis (CIA).
Resveratrol (RSV), LPS, compound C (CC), and Ficoll-Paque were purchased from Sigma-Aldrich (St. Louis, MO, USA). Recombinant human IL-4, recombinant human macrophage-colony-stimulating factor (M-CSF), and interferon (INF) γ were from Pepro Tech (Rocky Hill, NJ, USA). Sirtinol was obtained from Calbiochem (La Jolla, CA, USA). Antibodies specific for AMPKα, phosphor-AMPKα, phosphor-NF-κB p65 (Lys310), acetyl-NF-κB p65 (Lys310), acetyl-CoA carboxylase (ACC), and phosphor-ACC were from Cell Signaling (Danvers, MA, USA). Antibodies specific for SIRT1, NF-κB p65, ariginase-1, and Histone H1 were from Santa Cruz Biotechnology Inc. (Santa Cruz, CA, USA). Antibodies specific for TNFα and IL-1β were from Abcam (Cambridge, UK). Antibody specific for TRANCE/TNFSF11 was from R&D Systems (Minneapolis, MN, USA).
Synovial fluid was obtained during the therapeutic arthrocentesis from the affected knees of patients with RA who fulfilled the American Rheumatism Association classification criteria (1987). All patients provided informed consent, and the study was approved the Medical Ethics Committees of the Academic Medical Center (Dong-A University Hospital, Busan, South Korea).
Mononuclear cells from synovial fluid were isolated using density gradient centrifugation methods (Ficoll-Paque). Cells were incubated in RPMI with 10% FBS for 24 h, and then adherent cells were incubated in culture medium plus M-CSF (100 ng/ml) for 7 days. M1 polarization was stimulated by LPS (100 ng/ml) or INFγ (20 ng/ml), and M2 polarization was achieved by treating cells with IL-4 (20 ng/ml).
For measurement of mRNA levels, total RNA isolation and RT-PCR were performed as previously described (
Oligonucleotide sequences used for qPCR.
Human | Mouse | ||
---|---|---|---|
Oligonucleotides, 5′–3′ | Oligonucleotides, 5′–3′ | ||
FcεRII | Forward: GGGAGAATCCAAGCAGGAC Reverse: GGAAGCTCCTCGATCTCTGA |
Arginase-1 | Forward: CCAGAAGAATGGAAGAGTCAGTGT Reverse: GCAGATATGCAGGGAGTCACC |
Interleukin (IL)-10 | Forward: AAGACCCAGACATCAAGGCG Reverse: AGGCATTCTTCACCTGCTCC |
Fizz | Forward: TCCCAGTGAATACTGATGAGA Reverse: CCACTCTGGATCTCCCAAGA |
IL-12 p35 | Forward: GCCACAGGTCTGCATCCA Reverse: GACCTGGCGGGCTGAGTA |
FcεRII | Forward: GGGACACAGCTCATGTTGGT Reverse: GCAGTGTCTCCCAGCTGTTT |
IL-12 p40 | Forward: AGCCTCCTCCTTGTGGCTA Reverse: TGGTTTTATCTTTTGTG TGTGC |
IL-1β | Forward: GGTGTGTGACGTTCCCATTAG Reverse: TCGTTGCTTGGTTCTCCTTGT |
iNOS | Forward: GTTCTCAAGGCACACCAGGTCTG Reverse: GCAGGTCACTTATGTCACTTATC |
IL-12 p35 | Forward: CTTAGCCAGTCCCGAAACCT Reverse: TTGGTCCCGTGTGATGTCT |
MCP-1 | Forward: GGAGCATCCACGTGTTGGC Reverse: ACAGCTTCTTTGGGACACC |
IL-12 p40 | Forward: GTTCAACATCAAGAGCAGTAGCA Reverse: CTGCAGACAGAGACGCCATT |
MDC | Forward: GGTTGTCCTCGTCCTCCTTG Reverse: GAAGGTTAGCAACACCACGC |
MCP-1 | Forward: TCCCACTCACCTGCTGCTACTCA Reverse: GCTTCTTTGGGACACCTGCTG |
MrC1 | Forward: TCAATGGCATGAAGCGGAGA Reverse: TACTGTTCAGGGCGATCCAC |
TNFα | Forward: ATGAGAAGTTCCCAAATGGC Reverse: CTCCACTTGGTGGTTTGCTA |
18S | Forward: GGCCCTGTAATTGGAATGAGTC Reverse: CCAAGATCCAACTACGAGCTT |
Yim-1 | Forward: GGGCATACCTTTATCCTGAG Reverse: CCACTGAAGTCATCCATGTC |
Actin | Forward: GCCCTGAGGCTCTTTTCCAG Reverse: TGCCACAGGATTCCATACCC |
Proteins were loaded into 10% polyacrylamide gels. Protein transferred to nitrocellulose membranes, which were immunoblotted with antibodies against AMPKα, p-AMPKα, ACC, p-ACC, NF-κB p65, p-NF-κB p65, ac-NF-κB p65, and SIRT1. Protein bands were visualized using the Supersignal West Dura Chemiluminescent Substrate (Thermo Fisher Scientific Inc., Rockford, IL, USA). Signals from bands were quantified using a UN-SCAN-IT gel™ software (Silk Scientific, Orem, UT, USA).
For SIRT1 gene knockdown, cells were transfected with SIRT1 siRNA oligonucleotide (GenBank accession no. NM_019812.1: Daejeon, South Korea) using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) following the manufacturer’s protocol.
Cells were transiently transfected with NF-κB luciferase reporter vector using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s instruction. The activity of firefly and luciferase was measured using the dual luciferase reporter assay system (Promega, Madison, WI, USA).
SIRT1-Tg mice (C57BL/6N) were a generous gift from Jong-Wan Park (Seoul National University, South Korea). C57BL/6N mice were from Japan SLC (Shizuoka, Japan). All experimental procedures were approved by the Animal Experimental Committee of the College of Medicine, Pusan National University (PNU-2016-1107) and done in accordance with guidelines for animal research. Murine bone marrow cells were isolated from mouse femoral and tibial bone marrow. Bone marrow-derived macrophages (BMDMs) were prepared from bone marrow cells and cultured in RPMI with 10% FBS and M-CSF (100 ng/ml).
To trigger CIA, mice were sensitized by injecting 100 µg of chicken type II collagen (CII) supplemented with complete Freund’s adjuvant (Sigma) intradermally at the tail base, and received a booster injection of CII supplemented with incomplete Freund’s adjuvant in the same manner 14 days later by Inglis et al. (
Arthritis severity scoring system.
0 | Normal |
1 | Slight erythema or swelling |
2 | Distinct erythematous swelling |
3 | Joint distortion |
4 | Ankylosis of the joint |
0 | Normal |
1 | Synovial inflammation—mild |
2 | Synovial inflammation—moderate |
3 | Synovial inflammation—moderate |
Tissue sections were obtained from paraffin blocks and rehydrated, and then incubated with anti-TNFα, anti-IL-1β, and anti-TRANCE/TNFSF11 antibodies. Immunoreaction products were visualized using a broad-spectrum immunohistochemistry kit (Diaminobenzidine substrate kit, Vector Laboratories, Inc., Burlingame, CA, USA).
Statistical analyses were performed using GraphPad Software (San Diego, CA, USA). Means and SDs were calculated. The parametric Student’s
Macrophages were generated from synovial monocytes cultured in the presence of M-CSF for 7 days (
To investigate the effect of SIRT1 on the M2 polarization of macrophages induced by IL-4 (20 ng/ml), we detected the mRNA levels of M2 macrophage markers, that is, MDC (macrophage-derived chemokine), FcεRII (low-affinity IgE receptors), MrC1 (C-type mannose receptor 1), and IL-10. As shown in Figures
Effects of SIRT1 activation by resveratrol (RSV) on M1 and M2 markers in rheumatoid arthritis macrophages.
M1 phenotype macrophages are promoted by macrophage-activating factors, such as LPS and IFNγ (
These results show that SIRT1 activation by RSV augments the expression of anti-inflammatory M2 macrophages and attenuates the expression of pro-inflammatory M1 macrophages.
To confirm the role of SIRT1 in RA macrophage polarization into the M2 phenotype, SIRT1 expression in macrophages was inhibited by using of siRNA against SIRT1 gene. When cells were subjected to SIRT1 gene knockdown, they showed ~60% reduction in SIRT1 protein expression. These cells did not show increase in SIRT1 protein expression in response to RSV as contrasted to the cells transfected with negative siRNA (Figure
Expression of M1/M2 polarization markers in SIRT1 siRNA-transfected rheumatoid arthritis macrophages.
The mRNA levels of MDC and MrC1 were not elevated by RSV in cells transfected with SIRT1 siRNA, whereas in negative control cells MDC and MrC1 mRNA levels were significantly increased (Figures
Based on a report that SIRT1 regulates AMPK signaling in attenuation of pro-inflammatory activity (
Enhanced macrophage M2 polarization by resveratrol (RSV) by SIRT1-mediated p-adenosine monophosphate-activated protein kinase α (AMPKα) upregulation.
We also investigated the involvement of the SIRT1/AMPKα signaling pathway in M2 macrophage polarization. As shown in Figures
NF-κB activation has been reported to play a key role in LPS-induced M1 macrophage polarization, and M1 macrophages are known to have κB sites in their promoter regions, including those of CCL2, iNOS, and TNFα (
To investigate the mechanism whereby SIRT inhibits polarization of macrophage to the M1 phenotype, we examined whether SIRT1 could inhibit NF-κB signaling. RA macrophages were pretreated with RSV (50 µM) for 24 h, and then further stimulated with LPS (1 µg/ml) plus IFNγ (20 ng/ml) for 1 h. As shown in Figure
Suppression of NF-κB signaling by resveratrol (RSV) in the LPS/interferon (IFN)γ-induced M1 macrophage polarization.
Furthermore, to define whether the prevention of NF-κB activity by RSV was mediated by SIRT1 activation, RA macrophages were subjected to SIRT1 knockdown by SIRT1 siRNA transfection. In the SIRT1 knockdown macrophages, the LPS plus IFNγ increased acetylation at K310 and phosphorylation of p65 at Ser536 were not suppressed by RSV, whereas they were significantly suppressed by RSV in negative control cells (Figure
To determine the phenotypic characters of macrophages, we isolated BMDMs from SIRT1 Tg-mice and determined SIRT1 protein and AMPKα phosphorylation levels in these cells.
Levels of SIRT1 protein and p-AMPKα were significantly higher in BMDMs from SIRT1 Tg-mice when compared with those from wild type (WT) mice (Figure S3 in Supplementary Material). BMDMs from SIRT1 Tg-mice were treated with IL-4 (20 ng/ml), and the mRNA levels of M2 macrophage markers, including ariginase-1, Fizz, Yim-1, and FcεRII were assessed. As shown in Figures
Upregulation of the M2 phenotype and downregulation of the M1 phenotype in bone marrow-derived macrophage (BMDMs) obtained from SIRT1 transgenic (Tg)-mice.
By contrast, when BMDMs from SIRT1 Tg-mice were incubated with LPS (1 µg/ml) plus IFNγ (20 ng/ml), mRNA levels of the M1 phenotype markers: TNF-α, IL-1β, MCP-1, IL-12 p35, and IL-12 p40 were significantly suppressed (Figures
The mouse CIA model was employed because it is similar immunologically and pathologically to RA patients. SIRT1-Tg CIA mice developed less severe arthritis than WT CIA mice, as evidenced by lower scores of disease activity from days 26 to 38 (Figure
Analysis of the pathological severity of the joints of C57BL/6N and SIRT1 transgenic (Tg) mice (
On day 38 after first injection, knee joints from SIRT1-Tg CIA and WT CIA mice were obtained for microscopic analysis. They were blindly scored for histological signs of arthritis, that is, cartilage destruction, bone erosion, and cell infiltration. Histological articular damage was significantly less severe in the knee joints of in SIRT1 Tg-CIA mice than in those of WT CIA mice (Figure
In addition, we assessed the expression of M2 marker arginase-1, M1 marker TNFα, and IL-1β in the knee joints by immunohistochemistry. Ariginase-1 marker in SIRT1-Tg mice was detected relatively higher in SIRT1-Tg mice than WT mice, and SIRT1 Tg-CIA mice showed less decrease in ariginase-1 expression. In contrast to ariginase-1, TNFα and IL-1β were rarely detected in SIRT1 Tg-CIA mice, compared to WT CIA mice. A similar result was obtained for RANKL expression (Figure
The current study shows when RA macrophages were treated with IL-4, SIRT1 enhances macrophage polarization into the M2 phenotype by upregulating the phosphorylations of AMPKα and ACC. In line with these findings, SIRT1 significantly suppressed the M1 phenotype polarization of macrophages by inhibiting NF-κB activation. In addition, we observed that expressions of TNFα and IL-1β, the pro-inflammatory cytokines, were significantly lower in SIRT1 Tg-CIA mice than in WT CIA mice, and these reductions were associated with the suppression of synovial inflammation and bone destruction. These findings support the notion that the anti-inflammatory responses in RA are induced by activation of SIRT1/AMPKα signaling pathways.
Kennedy et al. (
SIRT1 regulates the functions of several important transcription factors with anti-inflammatory effects (
It has been argued SIRT1 aggravates inflammation by upregulating the expressions of pro-inflammatory cytokines in synovial fibroblasts (
We found that SIRT1 strongly suppressed p65 acetylation at K310 and phosphorylation at Ser536, and that SIRT1 increased IκBα expression in cytosol, and decreased the nuclear translocation NF-κB p65, and consequently decreased NF-κB to κB binding, which agrees with previous reports (
In this study, the findings obtained from SIRT1-Tg CIA mice model supported the implication of SIRT1 in RA inflammation
All experimental procedures were approved by the Animal Experimental Committee of the College of Medicine, Pusan National University (PNU-2016-1107) and done in accordance with guidelines for animal research.
SYP designed and did the major experiments, analyzed the experimental data, and contributed to the writing. KWH and CDK designed and contributed to the writing. SWL and SYL designed the experiments and analyzed the experimental data. SSB and KK did the critical revision of the manuscript.
The authors declare no potential conflicts of interest.
The authors would like to thank Dr. J. W. Park (Seoul National University, Korea) for providing SIRT1-Tg mice.
The Supplementary Material for this article can be found online at