Article Text

Extended report
Resveratrol modulates murine collagen-induced arthritis by inhibiting Th17 and B-cell function
  1. Gao Xuzhu1,
  2. Mousa Komai-Koma1,
  3. Bernard P Leung2,
  4. Hwee Siew Howe2,
  5. Charles McSharry1,
  6. Iain B McInnes1,
  7. Damo Xu1
  1. 1Institute of Infection, Immunity and Inflammation, College of Medical, Veterinary and Life Sciences, University of Glasgow, Glasgow, UK
  2. 2Department of Rheumatology, Allergy and Immunology, Tan Tock Seng Hospital, Novena, Singapore
  1. Correspondence to Dr Damo Xu, Institute of Infection, Immunity and Inflammation, College of Medical, Veterinary and Life Sciences, University of Glasgow, B421, 120 University Place, Glasgow G12 8TA, UK; damo.xu{at}


Background Alcohol intake is inversely related to rheumatoid arthritis (RA) disease incidence and severity. Resveratrol, a safe, well-described plant-derived compound, possesses anti-inflammation and immune-regulatory properties and is present in red wine. As such, it could mediate anti-inflammatory properties of the latter and offer novel therapeutic utility in is own right.

Objective To evaluate the therapeutic effect of resveratrol on collagen-induced arthritis (CIA) and its putative immune modulation in mice.

Methods CIA was induced in DBA1 mice by immunisation with collagen II. Different doses of resveratrol were administered before or after the development of CIA. The levels of antibody and cytokines in serum or in draining lymph node (DLN) lymphocyte culture supernatants were measured by ELISA and Th17 cell development in DLN was monitored by flow cytometry.

Results Either prophylactic or therapeutic administration of resveratrol attenuated clinical parameters and bone erosion in CIA mice. The arthritis-protective effects were associated with markedly reduced serum levels of pro-inflammatory cytokines and collagen-specific, but not total, IgG, and with reduced numbers of Th17 cells and the production of IL-17 in DLN.

Conclusion Resveratrol modulates inflammatory arthritis in rodents by selectively suppressing key cellular and humoral responses necessary for disease development. This may partly explain the protective effects of red wine but importantly may offer a novel, effective and safe pathway whereby novel agents could be developed to treat RA.

View Full Text

Statistics from

Rheumatoid arthritis (RA) is a common inflammatory arthropathy of unknown aetiology.1,,3 Genetic factors, operating together with environmental triggers including smoking, obesity and periodontal disease, appear to drive the evolution of disease.4 5 T and B lymphocytes, and particularly T-helper (Th) type 17-dependent pathways are thought to play a pivotal role in the initiation and persistence of disease.6,,8 B cells may contribute by producing autoantibodies and by acting as antigen-presenting cells.9,,11 The advent of effective treatment with B-cell-depleting agents (rituximab) and costimulatory blockade (abatacept) confirms a critical role for adaptive immune processes in disease.12 13 T cells may activate a wide range of innate effector cells including macrophages in the synovium by cell contact and cytokine release.3 7 14 These activated innate cells then release a spectrum of pro-inflammatory mediators including tumour necrosis factor alpha (TNFα), interleukin (IL) 1β, IL-6 and IL-17, which are responsible for joint inflammation and destruction.3

Current therapeutics of RA are directed primarily at reducing inflammation. Conventional disease-modifying anti-inflammatory drugs such as methotrexate, sulphasalazine and glucocorticoids, and the newer generation of biological agents comprise current standard of care. However, a substantial proportion of RA patients respond poorly to these new generation drugs and long-term unmet clinical need remains.15 16 Several new lines of recombinant immune-modulating drugs targeting a variety of immune modulatory targets are under development.17 However, the long-term beneficial and detrimental effects of such agents are difficult to predict and they will probably be expensive. More effective, safe and affordable drugs are needed to control RA. Many natural compounds exhibit anti-inflammatory properties and have the potential for treating inflammatory disorders such as RA. We focus our attention here in particular on resveratrol (3,5,4′-trihydroxystilbene), a plant phytoalexin.18 19

Resveratrol is a natural antimicrobial compound found in various plants and fruits.20 It is particularly rich in the skin of red grapes and therefore also in red wine.21 22 Resveratrol has attracted great attention following the discovery of its cardioprotective properties associated with drinking red wine.21 22 This has been used partly to explain the ‘French paradox’, which is the observation that there is a low incidence of coronary heart disease.21 22 Mechanistic studies demonstrate that resveratrol also possesses anti-inflammatory, antioxidant, anticancer and anti-aging properties.19 20 22,,25 It has been reported that resveratrol can inhibit several experimental autoimmune diseases including diabetes, encephalomyelitis and colitis.26,,28 Whereas the detailed mechanisms are not fully understood, resveratrol may modulate the function of key cellular receptors, inflammatory genes or signalling transcription factors, including nuclear factor kappa B, STAT3 and cyclo-oxygenase 2 and these also play critical roles in RA pathogenesis.24,,26 29,,31 In this context, resveratrol is a cyclo-oxygenase 2 inhibitor and protects chondrocytes against oxidant injury and apoptosis in vitro.29 30 In vivo, resveratrol can ameliorate the onset of lipopolysaccharide-induced acute joint inflammation in the rabbit.31 Intriguingly, alcohol intake including red wine consumption may also be protective in RA.32 33 However, until now the effect and mechanisms of resveratrol on experimental arthritis and RA remain unknown. In this study we tested the hypothesis that resveratrol could attenuate experimental arthritis and that it can do so by the modulation of autoreactive immune responses.

Materials and methods


Male DBA1 mice were obtained from Harlan Olac (Bicester, UK). All mice were used at 8–10 weeks old and maintained at the Biological Services Unit, University of Glasgow, in accordance with the Home Office UK guidelines.

Induction and treatment of collagen-induced arthritis

Collagen-induced arthritis (CIA) was induced in mice as previously described.34 Briefly, DBA mice were immunised by intradermal injection of 100 μg of bovine type II collagen (Chondrex, Washington, USA) emulsified in complete Freund's adjuvant (CFA; MD Biosciences, Zürich, Switzerland). Trans-resveratrol (Sigma, Gillingham, Dorset, UK) was dissolved in dimethyl sulphoxide as described previously.35 CIA mice were injected intraperitoneally daily for 10 days with resveratrol either from day 10 (prevention) or from day 23 (treatment) after the initial immunisation. Control procedures used the same volume of dimethyl sulphoxide. Paw thickness was measured with a dial-calliper (Schnelltaster; Kroeplin, Surrey, UK). Clinical scores were assigned as described.34

Cell culture

Draining popliteal lymph nodes (DLN) were removed on day 42 after primary immunisation. Single cells (2×106 cells/ml) were activated with plate-bound anti-CD3 (5 μg/ml) with or without resveratrol (30 μm) for 72 h.

Th17 cells were polarised as described before.36 Briefly, CD4 T cells (0.5×106 cells/ml) from naive mice were isolated (Miltenyi Biotec GmbH Bergisch Gladbach, Germany)and polarised with anti-CD3 (2.5 μg/ml), anti-CD28 (1.5 µg/ml), transforming growth factor beta (1 ng/ml), IL-1β (10 ng/ml) and IL-6 (20 ng/ml) in the presence of anti-IL-4 (10 μg/ml), anti-IL-12 (10 µg/ml) and anti-interferon gamma (IFNγ) (10 μg/ml) antibodies for 72 h with or without 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) (160 nM; SUPELCO Analytical, Bellefonte, Pennsylvania, USA)or resveratrol (30 μM) alone or together. The IL-17 levels in the culture or in the cells were analysed by ELISA or flow cytometry.

Flow cytometry

Freshly harvested DLN cells (3×105 cells/tube) were activated with phorbol myristate acetate and ionomycine. Cells were stained with anti-CD4 and then permeabilised and stained with anti-IL-17A-PE or anti-IFNγ-FITC and isotype-matched antibodies as controls (all from Becton Dickinson, Abingdon, UK).Cells were analysed by FACSCalibur using Cell Quest (Becton Dickinson BD Bioscience, Oxford, UK).

Cytokines and antibody measurement

All cytokine concentrations were detected by a 20-plex mouse cytokine assay (Invitrogen, Paisley, UK) according to the manufacturer's instructions using a luminex platform. The serum titres of total and anti-type II collagen antibodies were detected by enzyme immunoassay as previously described.34


The hind limbs were fixed in 10% neutral-buffered formalin and decalcified. Sections (4 µm) were stained with H&E. The joint pathology was examined and scored as described before.34

Statistical analysis

Clinical scores were analysed using the non-parametric Mann–Whitney U test and histological scores by the one-way analysis of variance test. Differences between cumulative incidences at a given time point were analysed by the χ2 contingency analysis. Cytokine and IgG levels were compared using the Student's t test.


Dose-dependent preventive effect of resveratrol on the development of CIA

We first evaluated whether resveratrol can prevent the development of CIA. DBA1 mice were induced with CIA on day 0 and injected intraperitoneally with different doses of resveratrol (15 or 20 mg/kg per mouse/day) daily from day 10 for 10 consecutive days. The development of arthritis was monitored until day 38. Whereas control mice developed severe CIA observed from day 21, mice receiving 20 mg/kg of resveratrol showed significantly reduced disease incidence, number of involved paws, footpad thickness and clinical index compared with the wild-type mice (figure 1A). Importantly, 15 mg/kg of resveratrol had no significant effect on the development and severity of CIA (see supplementary figure S1A, available online only). Histological analysis demonstrated that resveratrol also markedly reduced mononuclear and polymorphonuclear cell infiltration into the joint, synovial hyperplasia and adjacent cartilage and bone erosion in treated CIA mice compared with controls. These results suggested that resveratrol can prevent the development of CIA.

Figure 1

Resveratrol (REV) can prevent the development of collagen-induced arthritis (CIA). DBA mice were induced with CIA and injected with REV daily from day 10 for 10 days. (A) The clinical parameters were determined regularly. (B) The paw sections were stained with H&E. The pathogenic score was determined. Data are mean±SD (*p<0.05 vs control group by Mann–Whitney U test (n=10)) and are representative of two experiments. DMSO, dimethyl sulphoxide.

Therapeutic effect of resveratrol on established CIA

We next evaluated whether resveratrol can suppress ongoing CIA. For this purpose, CIA mice were treated with or without resveratrol (20 mg/kg per mouse) intraperitoneally daily from days 23 to 33 when arthritis had already commenced and was developing (figure 2A). Resveratrol-treated mice rapidly reduced all clinical parameters of CIA from the second dose of treatment compared with the controls (figure 2A). A similar therapeutic effect with 30 mg/kg of resveratrol was also observed in an independent experiment (see supplementary figure S1B,C, available online only). Histological examination of joints at day 36 revealed a marked reduction of cell infiltration, synovial hyperplasia and bone erosion in resveratrol-treated mice when compared with control mice (figure 2B). The therapeutic effect of resveratrol was maintained for at least 1 week (until day 40) without further resveratrol treatment (data not shown). Together, these data clearly indicate that resveratrol has a therapeutic effect on ongoing inflammatory arthritis.

Figure 2

Therapeutic effect of resveratrol on established collagen-induced arthritis (CIA). DBA mice were induced with CIA and treated with resveratrol (REV) daily from day 23 for 10 days. (A) The clinical parameters were determined. (B) The paw sections were stained with H&E. The pathogenic score was determined. Data are mean±SD (*p<0.05 and **p<0.01 vs control group by Mann–Whitney U test (n=10)) and are representative of two experiments. DMSO, dimethyl sulphoxide.

Resveratrol selectively inhibits autoantibody and inflammatory cytokine production in CIA mice

We next assessed the immune mechanisms by which resveratrol prevented and suppressed ongoing CIA. We determined its role in the regulation of humoral immunity by measuring collagen-specific and total IgG levels in serum. Consistent with the attenuated clinical parameters (figures 1 and 2), resveratrol at 20 mg/kg significantly prevented the development of serum collagen-specific IgG2a and IgG1 compared with control mice (figure 3A). This reduction was also observed after resveratrol treatment of established CIA (figure 3B). However, resveratrol (20 mg/kg) had no significant influence on the levels of total IgG1 or IgG2a, suggesting that resveratrol selectively controls the collagen-specific B-cell response in CIA rather than acting as a general B-cell-depleting/suppressing factor.

Figure 3

Resveratrol (REV) selectively inhibits autoantibody production in collagen-induced arthritis (CIA) mice. The serum antibody levels from different treated CIA mice were determined by ELISA. (A) REV treatment suppressed collagen-specific IgG1 and IgG2a production in CIA mice. (B) Preventive injection of REV inhibited specific IgG production but not total IgG production in CIA mice (C). Data are mean±SEM (*p<0.05 and **p<0.01 vs control group by Student's t test (n=10)). Data are from pooled mouse serum (n=10/group) and are representative of three experiments. DMSO, dimethyl sulphoxide.

We also assessed the effect of resveratrol on serum levels of cytokines in CIA mice at day 23. CIA mice were treated with resveratrol as described in figure 1 and serum cytokine levels were measured by luminex. Among the cytokines tested, IFNγ, TNFα, IL-6 IL-1 and IL-4 levels were significantly reduced, and particularly IL-17 expression was almost completely abrogated. There were no significant differences in serum levels of IL-2 and IL-10 (figure 4A). Furthermore, an identical resveratrol injection regime had no effect on the cytokine profile in naive DBA1 mice (data not shown). Together, these data suggest that resveratrol does modulate the creation of the cytokine milieu that is present during the onset of and perpetuation of CIA.

Figure 4

Resveratrol (REV) selectively inhibits inflammatory cytokine production in collagen-induced arthritis (CIA) mice. CIA mice were treated with or without REV as in Figure 1. The serum cytokine levels were determined by luminex. (A) REV inhibited inflammatory cytokine production in CIA mice. (B) REV had no effect on the production of interleukin (IL) 2 and IL-10 in CIA mice. Data are mean±SD (*p<0.01 and **p<0.01 compared with controls by Student's t test), from pooled mouse serum (n=8/group) and are representative of three experiments. DMSO, dimethyl sulphoxide; IFN, interferon; TNF, tumour necrosis factor.

Resveratrol suppresses T-cell expansion and pro-inflammatory cytokine production in vivo and in vitro

We next sought to determine whether resveratrol was able to inhibit T lymphocyte, in particular Th17, cell expansion and cytokine production in CIA mice ex vivo and in vitro. The CIA mice injected with or without resveratrol as in figure 1A had DLN harvested on day 23. DLN lymphocytes were cultured with or without anti-CD3 antibody for 3 days. The activated DLN cells from control CIA mice showed significantly higher levels of IL-17, IFNγ, TNFα, IL-1β, granulocyte macrophage colony-stimulating factor, macrophage inflammatory protein 1α, IL-5 and IL-13 than those from the resveratrol-treated mice (figure 5A). In some experiments, the DLN lymphocytes were cultured with anti-CD3 plus resveratrol in order to examine its direct effects on T-cell function in vitro. The cytokine and inflammatory mediator expression in the cultured DLN cells from untreated mice could be inhibited or from resveratrol-treated mice could be further inhibited by adding resveratrol to the culture (figure 5A). In contrast, resveratrol treatment had no significant influence on the levels of IL-4, IL-10 or IL-2 in anti-CD3 activated DLN T lymphocytes in vitro (see supplementary figure S2, available online only, and data not shown).

Figure 5

Resveratrol (REV) suppresses T-cell expansion and proinflammtory cytokine production in vivo and in vitro. Collagen-induced arthritis mice were injected with or without REV (Figure 1) and draining lymph nodes (DLN) were collected and analysed. (A) The DLN cells (2×106/ml) were cultured with or without anti-CD3 (5 μg/ml) with or without REV (40 μM) for 72 h. Cytokine concentrations in the culture were measured by luminex. (B) DLN cells were activated with phorbol myristate acetate and ionomycine and then stained with anti-CD4 and anti-interleukin (IL) 17A-PE or anti-interferon (IFN) γ-FITC and analysed by FACS. (C) The number of total, CD4, T-helper (Th) 1 and Th17 cells was determined by cell counting and FACS. Data are mean±SD (*p<0.05 and **p<0.01 by Student's t test) and are representative of two experiments. DMSO, dimethyl sulphoxide; GM–CSF, granulocyte macrophage colony-stimulating factor; TNF, tumour necrosis factor.

Furthermore, compared with controls, freshly isolated DLN CD4 T cells from resveratrol-protected mice showed markedly reduced expression of IL-17 (from 9.62% to 1.42%) and to a lesser extent, IFNγ (from 3.19% to 2.22%) assessed by flow cytometry (figure 5B). Resveratrol injection also decreased the numbers of total cells, CD4 IL-17+ Th17 cells and or CD4 IFNγ + Th1 cells in DLN of CIA mice compared with the cells from control mice (figure 5C).

These results demonstrated that resveratrol can suppress the expansion and cytokine production in Th17 and Th1 cells in CIA mice.

It was reported that aryl hydrocarbon receptor (AhR) signals are required for optimal Th17 cell polarisation and that resveratrol is an antagonist of AhR.36 37 We therefore investigated the role of AhR in resveratrol-mediated Th17 cell inhibition under standard Th17 polarisation conditions in vitro. As shown in figure 6A,B, resveratrol markedly suppressed Th17 cell proliferation and IL-17 production. The effect of resveratrol was not due to the induction of T-cell apoptosis or necrosis (data not shown). Furthermore, TCDD is a well-defined exogenous AhR agonist.36 We found that TCDD significantly enhanced IL-17 production under Th17 cell polarising conditions (figure 6C). Importantly, the effect of TCDD was completely abolished by the preincubation of the cells with resveratrol for 30 min before adding TCDD (figure 6C). Our results therefore suggest that resveratrol may suppress Th17 cell polarisation and proliferation, at least partly, by AhR.

Figure 6

Resveratrol (REV) suppresses T-helper (Th) 17 cell polarisation by aryl hydrocarbon receptor in vitro. Th17 cells were polarised as described in the Materials and methods section in the presence or absence of different doses of REV. (A) T-cell proliferation was determined by [3H] thymidine incorporation, interleukin (IL) 17 production by ELISA and (B) intracellular cytokine levels by flow cytometry. (C) Th17 cells were also polarised in the presence of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) (160 nM) and REV (30 μM) alone or together for 3 days and the cytokine levels were determined by ELISA. Data are mean±SD (*p<0.05 and **p<0.01 by Student's t test) and are representative of two experiments. IFN, interferon.


Data presented here demonstrate for the first time that resveratrol, a molecule produced by plants in response to stress or infection can effectively prevent and treat experimental inflammatory arthritis. Its joint-protective properties appear to be attributed to its inhibitory effect on the proinflammatory cytokine production of T lymphocytes, in particular the Th17 subset, and autoantibody production. This suppressive effect was antigen specific and arthritis related. Resveratrol may thus represent a novel anti-arthritis agent or point to novel pathways that are tractable in the management of the same.

Th17 cells are critically associated with CIA and RA.7 8 Our results suggest that resveratrol has a profound inhibitory effect on Th17 cell expansion and IL-17 production (figures 5 and 6). While the detailed mechanism is still unknown, resveratrol may directly inhibit Th17 cell development by blocking the function of endogenous AhR agonists (figure 6) or indirectly by suppressing the production of the key Th17 polarisation cytokines, IL-6 and IL-1, in CIA mice and in human mononcytes (figures 4 and 5 and unpublished data).38 Alternatively, because resveratrol can affect dendritic cell maturation and antigen-presentation capacity in vitro,39 it is possible that resveratrol may control arthritic T-cell responses by attenuating dendritic cell function in CIA mice. IL-10 is a key anti-inflammatory cytokine and can be induced by resveratrol in the encephalomyelitis model.27 However, IL-10 levels were not changed in our studies (figure 4B and supplementary figure S2, available online only), thus it is unlikely that resveratrol suppresses arthritis by means of IL-10 in the CIA context. Nevertheless, resveratrol may be a natural immune modulator for Th17 cell development and effector function in arthritic disease.

How resveratrol suppresses collagen-specific IgG production in this study is not fully understood. As IgG1 is mainly induced by type II cytokines (IL-4, IL-13) and IgG2a by type I cytokine (IFNγ)40 and we found that resveratrol can inhibit the production of IL-4, IL-13 and IFNγ in CIA mice (figures 4 and 5A), resveratrol may suppress production of these antibodies by downregulating both arms of the cytokine response. However, the mechanism by which resveratrol suppresses cytokine production in vivo is less well understood and needs to be investigated further. Nevertheless, this is consistent with previous reports showing that resveratrol can regulate type I and type II cytokine production in vitro and in vivo41 42 and suppress IgG production in vitro.43 However, again, we can not rule out a regulatory role of resveratrol in dendritic cells, helper T and B-cell function in the antigen-specific humoral response in our study.

The clinical relevance of our observations concerning resveratrol in RA remains unknown. A well-designed clinical trial is needed to address this issue. Resveratrol is currently undergoing clinical trial in human cancer, and phase I result suggest that it is well tolerated in humans in that context.44 It is part of the normal diet. Dietary constituents are known to be associated with disease incidence, and alcohol including red wine consumption is beneficial for RA.32 33 Our results may provide a molecular explanation at least partly for these observations.

RA is a disease with a complex pathogenesis that in some cases responds poorly to current therapy. Resveratrol, given its potential beneficial effect against cardiovascular diseases, may also represent a novel approach to the management of RA and related syndromes.


The authors wish to thank Mrs Helen Arthur for critical reading of the manuscript.


View Abstract

Supplementary materials


  • The first two authors contributed equally to the work.

  • Funding This work was supported by Arthritis Research UK and the Medical Research Council UK.

  • Competing interests None.

  • Patient consent Obtained.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.