Citrullinated Fibrinogen Promotes Bone Marrow Mesenchymal Stem Cells To Assume A Pro-inflammatory Phenotype

Background:Mesenchymal stem cells (MSC) are not only immune-inhibitory cells, and may act as pro-inflammatory cells due to lack of a proper "licensing". The role of MSC in autoimmune settings of rheumatoid arthritis (RA) has remained to be elucidated. Citrullinated fibrinogen (cfb) has been considered as one of the important specific autoantigens in RA and contributes to perpetuate the disease. cfb could act as damage-associated molecule pattern (DAMP) activing toll-like receptor (TLR)-associated downstream pathways, and contribute to the pro-inflammatory effect in some cells such as macrophages or fibroblast-like synoviocytes (FLS). However, it is still not clear whether this autoantigen could also function as DAMP to MSC, dampening their immunomodulatory effect or promoting their pro-inflammatory action, and thus actuating the development of RA.Objectives:To further confirm the pathogenic role of cfb, we comapared the level of autoantibody to cfb in RA patients to the healthy controls, and did the correlated analysis between the autoantibody level and disease activity. We analyzed the responses of bone marrow derived MSC (BMSC) to cfb. In the presence of cfb or not, the production of pro-inflammatory cytokines and immunomodulatory functions of BMSC were both evaluated. We also analyzed the effect of cfb on another type of MSC-umbilical cord derived MSC (UC-MSC). Through these experiments we aimed to clarify whether cfb could act as a dangerous "licensing" to MSC and the related molecular mechanism in this process was also analyzed. The effect of cfb primed BMSC on collagen induced arthritis (CIA) was compared to BMSC, in order to examine whether cfb compromised the modulatory function of BMSC in vivo.Methods:Serum was collected from RA patients and helthy controls, and the serum levels of autoantibody to native fibrinogen (fb) or cfb were determined by ELISA. The correlated analyses between the autoantibody level and disease activity score in 28 joints (DAS28) or inflammatory cytokines were further evaluated. BMSC were isolated from RA and osteoarthritis (OA) patients and exposed to cfb, and fb was used as a control. Pro-inflammatory cytokines stimulated by cfb or fb were determined by Real-time quantitative polymerase chain reaction (PCR) and ELISA. To evaluate the role of the TLR4-NFκB pathway in the induction of pro-inflammatory cytokines by cfb, parallel experiments were performed using inhibitors of TLR4 or NFκB pathway. The expression of TLR4 was detected by PCR and flow cytometry (FCM). The expression of NFκB pathway associated signaling proteins was determined by western blot (WB). To investigate whether BMSC retain their immunosuppressive properties after exposure to cfb, we detected the effects of cfb primed BMSC on proliferation or activation of peripheral blood mononuclear cells (PBMC), and also on the differentiation of Th17 cells from naive CD4+ T cells. We also investigated whether the expression of immunomodulatory molecules such as Jagged1, TGF-β1 and indoleamine dioxygenase (IDO) was altered in the presence of cfb. CIA was induced with the immunization of type Ⅱ collagen (CⅡ) in DBA/1 mice. Human BMSC (1×106) and cfb primed human BMSC (1×106) were injected i.v. into mice on day 28 after the immunization, respectively. All mice were sacrificed 3 weeks later and arthritis severity was assessed by clinical and histological scoring. The frequency of CD4+ T cell subsets in spleen was analyzed by flow cytometry. Serum levels of autoantibody to mouse CⅡ and inflammatory cytokines were also determined.Results:The serum level of autoantibody to cfb was significantly higher in RA patients compared to healthy controls. The level of this autoantibody in RA patients was positively correlated to DAS28 or the level of anti-CCP antibody or the levels of pro-inflammatory cytokines such as CCL2. These results suggested that cfb is a potential major target of pathogenic autoimmunity in RA. The autoantigen cfb increased the gene expression and protein production of IL-6, IL-8 and CCL2 in BMSC from both RA and OA patients. However, the productions of IL-1β and TNFa were unaffected by cfb. Compared with fb, cfb induced significantly higher expression of the pro-inflammatory cytokines (IL-6, IL-8 and CCL2) in these cells. cfb did not change the expression of TLR4 but activated NFκB pathway through modulating the expression of signaling proteins such as p-p65 and p-IκBα. Blocking experiments showed the participation of TLR4-NFκB pathway in the induction of these cytokines. The autoantigen cfb also compromised the ability of BMSC to inhibit PBMC proliferation and Th17 cell differentiation, as well as, impaired the potency of BMSC to produce the key immunomodulatory molecule IDO. In CIA, cfb primed BMSC showed impaired ability in down-regulating Th17 cells and modulating the serum levels of inflammatory cytokines when compared to BMSC. These results demonstrated that cfb could also cripple the function of BMSC in vivo. However, cfb did not alter the function of UC-MSC as that in BMSC, and UC-MSC appeared to be resistant to the autoantigen cfb. The low expression of TLR4 in UC-MSC may give some explanations to their resistance to cfb.Conclusion:Our results suggest that cfb is an important autoantigen of pathogenic autoimmunity in RA. This autoantigen promotes pro-inflammatory responses in BMSC and negatively affects their immunomodulatory functions both in vitro and in vivo. Citrullinated fibrinogen may act as a dangerous "licensing" for BMSC and thereby contributing to the propagation of inflammation in RA.