Effect of Clonally Expanded PD‐1highCXCR5–CD4+ Peripheral T Helper Cells on B Cell Differentiation in the Joints of Patients With Antinuclear Antibody–Positive Juvenile Idiopathic Arthritis

INTRODUCTION

Juvenile idiopathic arthritis (JIA) is the most common childhood rheumatic disease and is characterized by synovial lymphocyte infiltration and progressive joint destruction (1). Findings from functional and genetic analyses indicate that CD4+ T helper cells may play a central role in JIA pathogenesis (2). Additionally, autoantibodies (e.g., antinuclear antibodies [ANAs]) can be detected in ~50% of patients with JIA, and the presence of ANAs is correlated with synovial lymphoid neogenesis and B cell hyperactivity (3). However, the mechanisms that promote the aberrant activation of autoreactive B cells in JIA are still poorly understood.

In JIA patients, the synovial B cell compartment mainly consists of activated memory B cells and, among these, CD21low/–CD11c+CD27–IgM– double-negative (DN) B cells seem to be preferentially expanded in the joints of ANA-positive patients (3-7). CD21low/–CD11c+ B cells constitute a discrete B cell population that expresses transcription factor T-bet, accumulates in inflamed tissue, and is expanded in many autoimmune diseases (8, 9). The subset of CD21low/–CD11c+ B cells in humans resembles age-associated B cells in mice, which have been shown to play a central role in the pathogenesis of murine lupus (10). The human CD21low/–CD11c+ B cell population often overlaps with CD27–IgD– DN B cells or atypical memory B cells that are expanded in settings of chronic autoantigen/antigen exposure (11, 12). CD21low/–CD11c+ DN B cell differentiation seems to depend on the interaction of these cells with CD4+ T helper cells and their secreted cytokines (7, 10, 12, 13). However, which T helper cell subsets may promote pathogenic B cell responses at the site of inflammation in JIA patients is not yet known.

T follicular helper (Tfh) cells have been implicated in the pathogenesis of many autoimmune diseases (14, 15). They are characterized by the lineage-defining transcription factor Bcl-6, high expression of programmed death 1 (PD-1), and the chemokine receptor CXCR5 that recruits them into the follicles of secondary lymphoid organs (16). These cells exert their specific “B helper” function through the secretion of cytokines (e.g., interleukin-21 [IL-21]) and by production of costimulatory molecules (e.g., inducible T cell costimulator, OX40, CD154) (17). In contrast to classic Tfh cells, which act in the germinal centers of secondary lymphoid organs, extrafollicular or peripheral T helper (Tph) cell subsets have been detected in the inflamed joint tissue of patients with autoimmune diseases, and also have been observed in various murine models of autoimmune disease (18, 19). In particular, Tph cells that accumulated in the joints of patients with seropositive rheumatoid arthritis (RA) also secreted IL-21 but did not express Bcl-6 and CXCR5, and therefore Tph cells in RA seem to differ from classic Tfh cells (20). Interestingly, IL-21–expressing CD4+ T helper cells could also be detected in the inflamed joints of ANA-positive JIA patients; however, neither the phenotype nor the function of this T helper cell subset has been investigated in detail to date (21). Therefore, in this study, we investigated the occurrence and phenotype of CD4+ T cells that exhibit a “B helper” function at the site of inflammation in JIA patients and examined the functional impact of these cells on B cell differentiation.

DISCUSSION

Synovial inflammation in ANA-positive JIA is characterized by B cell hyperactivity; however, the mechanisms driving aberrant B cell activation at the site of inflammation are not yet understood (3, 4, 6, 7). We have previously demonstrated that the SF CD4+ T helper cell pool in ANA-positive JIA patients is particularly enriched in IL-21–secreting cells (21). We have also shown that CD21low/–CD11c+ DN B cells accumulate in the joints of these patients (5). In the present study, we demonstrate that the expansion of IL-21–secreting cells is attributable to a distinct subset of PD-1highCXCR5–HLA–DR+ CD4+ T cells. This CD4+ T helper cell subset exerts a potent B helper function and particularly skews B cell differentiation toward a CD21low/–CD11c+ phenotype in vitro, suggesting a functional relationship between both cell subsets.

High expression of activation and proliferation markers, as well as contraction of the TCRVB repertoire within SF PD-1highCXCR5–HLA–DR+CD4+ T cells, indicates clonal expansion of this T cell subset within the joints of JIA patients. Furthermore, the increased clonal overlap of PD-1highCXCR5–HLA–DR+CD4+ T cells between different joint sites in individual patients and between different JIA patients, as well as the preferential accumulation in the joints of ANA-positive JIA patients, may also suggest that clonal expansion has been elicited after encountering currently unknown autoantigens/antigens that are present at the site of inflammation. Therefore, PD-1highCXCR5–HLA–DR+CD4+ T cells appear to represent a pathogenic T cell subset that is expanded in a subgroup of JIA patients, particularly in ANA-positive patients. Dysregulation of T helper cells has been observed in the PB of JIA patients (29, 30). A circulating subset of “pathogenic-like” CD4+ T helper cells has been detected in the PB of JIA patients that partially resembles the PD-1highCXCR5–HLA–DR+CD4+ T cell subset described in the present report (29). However, whereas these circulating CD4+ T helper cells displayed an inflammatory cytokine pattern, including increased expression of IFNγ, TNF, and IL-17, the PD-1highCXCR5–HLA–DR+CD4+ T cell subset in the SF of JIA patients in the present study also showed increased expression of IFNγ and TNF, but did not show an increase in IL-17 levels. Furthermore, high PD-1 expression on CD4+ T cells in the SF of JIA patients was particularly correlated with increased expression of IL-21.

Hence, high PD-1 expression on SF CD4+ T helper cells seems to represent a surrogate marker of a clonally expanded T helper cell subset that is characterized by IL-21 expression, thereby suggesting a B helper function. This T helper cell subset appears to be functionally distinct from other proinflammatory T helper cell subsets that have previously been characterized in the joints of JIA patients, e.g., Th1, Th17, and Th1/Th17 cells (24, 26). However, CD161 expression on a majority of these IL-21+IFNγ+IL-17–PD-1high CD4+ T cells suggests that these cells may have originated from the Th17 subset of T helper cells, as has been described previously as the potential origin of nonclassic Th1 cells (24-26, 31-33). IL-12, which, in an environment of inflammation, has been shown to induce a shift from Th17 to Th1/Th17 or nonclassic Th1 cells and has been shown to induce IL-21 and IFNγ–coexpressing T helper cells in vitro, might be one of the factors involved in the conversion of Th17 cells to IL-21+IFNγ+IL-17– PD-1high CD4+ T cells (7, 24-26).

The PD-1highCXCR5–HLA–DR+CD4+ T cells in the joints of JIA patients shared a distinct phenotype with other IL-21–secreting CD4+ T helper cell subsets detected within the inflamed tissue in patients with different autoimmune diseases (20, 34). In particular, expression of activation markers (PD-1, OX40, HLA–DR) and a high BLIMP-1:Bcl-6 ratio in conjunction with c-MAF expression, as well as an extrafollicular/inflammatory chemokine receptor pattern (CCR2+CXCR5–), also characterize the transcriptional program of synovial Tph cells first described in patients with seropositive RA (20). Since then, using PD-1highCXCR5– as the common denominator in the phenotype of Tph cells, a potential circulating counterpart of this subset has been detected in the PB of patients with systemic lupus erythematosus and type 1 diabetes and is thought to contribute to disease pathogenesis by inducing autoantibody/antibody-secreting plasma cells (35, 36). Indeed, synovial Tph cells in seropositive RA displayed a potent B helper function, as assessed by the capacity to induce plasma cell differentiation and Ig secretion in vitro (20).

Consistent with those studies, we also provide the first evidence of a potent B helper cell function of PD-1highCXCR5–HLA–DR+ T helper cells derived from the SF of ANA-positive JIA patients. Extending the previous work on Tph cells, our current analysis establishes a more differentiated functional impact of Tph cells on B cell differentiation. Indeed, the results we have presented from in vitro experiments and ex vivo studies highly suggest that SF PD-1highCXCR5–HLA–DR+ CD4+ T cells not only induce plasma cell differentiation and Ig secretion but also particularly skew B cell differentiation toward CD21low/–CD11c+DN B cells. We assume that the cellular pattern observed in the SF of JIA patients mirrors that in the synovia. However, we were not able to extend our analysis to include synovial tissue from JIA patients, since synovectomy is rarely performed in JIA and ethical considerations prevented collecting synovial biopsy specimens from the children included in the analysis.

CD4+ T helper cells coexpressing IL-21 and IFNγ have been detected in several infections and autoimmune diseases, according to the corresponding setting and cellular phenotype: “Th1-Tfh cells,” Tfh-like cells or, as discussed above, Tph cells (18, 19, 37-44). Concomitantly, differentiation of CD21low/–CD11c+ DN B cells appeared to depend on IFNγ and IL-21, and expansion of Th1-Tfh cells correlates with the presence of CD21low/–CD11c+ DN B cells (10, 41, 45).

Consistent with the results from these studies, several findings from our experiments also suggest that Tph cells secreting IL-21 in conjunction with IFNγ support the differentiation of CD21low/–CD11c+ DN B cells in the joints of JIA patients. These findings include 1) the correlation between numbers of CD21low/–CD11c+ DN B cells and numbers of IL-21 and IFNγ–coexpressing PD-1highCD4+ T cells in the SF of JIA patients, 2) the need for co–cytokine blockade with blocking of IFNγ together with IL-21 to impair SF PD-1highCD4+ T cell–induced B cell differentiation toward CD21low/–CD11c+ B cells in vitro, 3) the lack of in vitro induction of CD21low/–CD11c+ B cells by tonsil Tfh cells that predominantly express IL-21 and IL-4 but do not show increased expression of IFNγ, and 4) the synergistic effect of IFNγ and IL-21 on T-bet induction in CD40L-stimulated control B cells in vitro. However, whereas plasmablast/plasma cell differentiation in vitro was essentially dependent on JAK/STAT signaling, and in particular was dependent on the expression of IL-21, the mechanisms underlying the induction of CD21low/–CD11c+ B cells by SF Tph cells seem to be more complex, since blocking JAK/STAT signaling (thereby blocking the effects of IFNγ and IL-21) only partially impaired differentiation in vitro.

Of note, another essential prerequisite for the formation of CD21low/–CD11c+DN B cells is the engagement of B cell–intrinsic Toll-like receptors (TLRs), including TLR-7 and TLR-9, which sense components of nuclear antigens that are delivered via antinuclear B cell receptors (BCRs) (10, 12, 13, 46). Although this process could not be addressed in the present study, it is tempting to speculate that expansion of CD21low/–CD11c+ DN B cells in the joints of ANA-positive JIA patients might have been induced after concomitant signaling through antinuclear BCRs and TLR-7/TLR-9 in the presence of IL-21 and IFNγ secreted by Tph cells (10, 47).

In summary, we have characterized a distinct subset of PD-1highCXCR5–HLA–DR+ CD4+ T cells that is clonally expanded in the joints of ANA-positive JIA patients and that displays a phenotype similar to that of Tph cells. These Tph cells functionally differ from classic Tfh cells and particularly promote differentiation of CD21low/–CD11c+ DN B cells that are concomitantly expanded in the joints of ANA-positive JIA patients. These cells are potentially triggered by autoantigens/antigens present at the site of inflammation. Hence, the characteristic expansion of Tph cells and CD21low/–CD11c+ DN B cells in the joints of ANA-positive JIA patients might reflect the autoimmune response at the site of inflammation.

ACKNOWLEDGMENT

We would like to thank the FACS Core Unit of the IZKF Würzburg for supporting this study.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Morbach had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Fischer, Dirks, Girschick, Morbach.

Acquisition of data. Fischer, Dirks, Klaussner, Haase, Holl-Wieden, Hofmann, Hackenberg, Girschick, Morbach.

Analysis and interpretation of data. Fischer, Dirks, Klaussner, Haase, Hofmann, Hackenberg, Girschick, Morbach.

留言 (0)

沒有登入
gif