Cutting Edge: Circulating Plasmablasts Induce the Differentiation of Human T Follicular Helper Cells via IL-6 Production

B cells require CD4+ T follicular helper (Tfh) cells to progress through the germinal center and provide protective Ab responses. In this article, we reveal a reciprocal interaction whereby circulating human plasmablasts are potent inducers of the Tfh cell–differentiation program, including the expression of their key transcription factor Bcl-6. The markedly increased propensity of plasmablasts, compared with naive B cells, to induce Tfh cell differentiation was due to their increased production of IL-6. Specific targeting of IL-6 using tocilizumab therapy in patients with rheumatoid arthritis led to a significant reduction in circulating Tfh cell numbers and IL-21 production, which was correlated with reduced plasmablast formation. Our data uncover a positive-feedback loop between circulating plasmablasts and Tfh cells that could sustain autoimmunity and spread Ab-driven inflammation to unaffected sites; this represents an important therapeutic target, as well as reveals a novel mechanism of action for tocilizumab.

B cells require CD4 + T follicular helper (Tfh) cells to progress through the germinal center and provide protective Ab responses. In this article, we reveal a reciprocal interaction whereby circulating human plasmablasts are potent inducers of the Tfh cell-differentiation program, including the expression of their key transcription factor Bcl-6. The markedly increased propensity of plasmablasts, compared with naive B cells, to induce Tfh cell differentiation was due to their increased production of IL-6. Specific targeting of IL-6 using tocilizumab therapy in patients with rheumatoid arthritis led to a significant reduction in circulating Tfh cell numbers and IL-21 production, which was correlated with reduced plasmablast formation. Our data uncover a positive-feedback loop between circulating plasmablasts and Tfh cells that could sustain autoimmunity and spread Ab-driven inflammation to unaffected sites; this represents an important therapeutic target, as well as reveals a novel mechanism of action for tocilizumab. The Journal of Immunology, 2015, 194: 2482-2485. C D4 + T follicular helper (Tfh) cells are a specialized Th subset that provides signals to B cells and guides their development through the germinal center (GC) (1). This Tfh cell-B cell interaction lies at the heart of the GC and is key for efficient immune responses; however, it also can promote autoimmune disease. Tfh cells express the transcriptional repressor Bcl-6; produce IL-21, which is essential for robust high-affinity humoral responses; and express ICOS and CXCR5, the latter directing this cell lineage to the GC. B cells with the highest affinity for Ag present cognate peptides to Tfh cells and, in turn, receive critical signals allowing their survival and differentiation into memory B cells and plasma cells. The differentiation of Tfh cells is reliant upon multiple signals, including cytokines, such as IL-6 and IL-21 (2,3). Murine B cells can reciprocally modulate Tfh cell dynamics and promote their formation, such as by the pro-vision of IL-6. In contrast, little is known about the role of B cells in human Tfh cell differentiation. In this article, we investigate the role of B cells in the induction of human Tfh cells.

Healthy individuals and patients
Blood was obtained from healthy individual volunteers and patients with rheumatoid arthritis (RA) before and 6 mo after treatment with tocilizumab. The mean disease activity score (DAS28) in the paired samples before and after tocilizumab treatment was 7.2 and 3.5, respectively. These patients were not taking any other disease-modifying antirheumatic drugs and were on ,7.5 mg prednisolone/d. The University College London Hospital ethics committee approved the study. The n values in the figure legends refer to independent donors.

Cell culture
PBMCs were stimulated with 2 mg/ml soluble anti-CD3 (HIT3a) and anti-CD28 (CD28.2) (eBioscience). A total of 50,000 cells each was used for cocultures of T and B cells. For Tfh functional assays, CD4 + T cells were resorted after 4 d of culture, with or without plasmablasts, and cultured with autologous freshly sorted naive B cells in the presence of 2 mg/ml endotoxinreduced Staphylococcal enterotoxin B (Sigma-Aldrich). Naive T cells were cultured with 10 ng/ml of IL-21 or IL-6 or a combination of both (Pepro-Tech).

Statistical analysis
Data were analyzed for significance by the paired t test using Prism (GraphPad, La Jolla, CA). The Mann-Whitney U test was used to compare healthy individuals and patients with RA. Correlation coefficients and their significance were analyzed by the Pearson correlation.

Results and Discussion
Plasmablasts promote Tfh cell expansion Human Tfh cells coexpress CXCR5 and ICOS at high density (4). In our study, Tfh cells were defined as CD4 + CXCR5 + I-COS + . Only a small fraction of circulating CD4 + T cells in healthy donors possessed these characteristics (Fig. 1A), as well as expressed PD1 but were negative for CD45RA (Supplemental Fig. 1A). Bcl-6 was expressed at low levels in  circulating Tfh cells (Fig. 1B), consistent with previous findings (4). In contrast, when PBMCs were stimulated with anti-CD3/CD28, CXCR5 and ICOS expression greatly increased (Fig. 1A), as did Bcl-6 expression in Tfh cells (Fig.  1B). We sought to identify whether B cells promoted the formation of Tfh cells and found that removal of the former from PBMCs led to a significant reduction in Tfh cells (Fig.  1A), as well as Bcl-6 expression in the Tfh cell population, after stimulation with anti-CD3/CD28 (Fig. 1B), suggesting a role for B cells in the maintenance of this T cell lineage.
Given the close correlation between plasmablasts and Tfh cells, we hypothesized that this B cell subset was responsible for the increased frequency of Tfh cells. Therefore, plasmablasts were isolated from peripheral blood, according to their expression of CD38 and CD27 (5) (Supplemental Fig. 1D), and compared with naive B cells for their ability to expand Tfh cells. Plasmablasts, but not naive B cells, could drive the expansion of Tfh cells from Tresp cells (Fig. 1F). Our data reveal that not all circulating B cells are equally equipped to expand Tfh cells and identify the potent properties of plasmablasts in increasing Tfh cell numbers.

Plasmablasts induce the differentiation of Tfh cells
We next assessed whether circulating plasmablasts possessed the capacity to induce the differentiation of Tfh cells from naive T cells. Thus, plasmablasts, naive and memory B cells, and monocytes were isolated and added to naive T cells in the presence of anti-CD3/CD28. Plasmablasts were the most potent inducers of Tfh cell differentiation ( Fig. 2A). Tfh cells induced in the presence of plasmablasts expressed high levels of PD-1 (Supplemental Fig. 1E) and Bcl-6 ( Fig. 2B). In addition, plasmablast-induced Tfh cells secreted significantly more IL-21 compared with naive B cell-induced Tfh cells (Fig. 2C). Although naive B cells possessed some ability to induce Tfh cell differentiation, this could be explained, in part, by the formation of plasmablasts from naive B cells during the culture (Supplemental Fig. 1F). Indeed, the per- centage of Tfh cells correlated with the numbers of plasmablasts present in these "naive B cell" cultures (Supplemental Fig. 1G). This finding suggests that the induction of Tfh cell differentiation was due, in part, to newly formed plasmablasts from naive B cells, although other mechanisms may contribute, given that monocytes can increase the number of Tfh cells above that seen with naive T cells alone. To support this finding we cocultured different ratios of naive CD4 + T cells/ plasmablasts and assessed the induction of Tfh cell differentiation. The ratio of plasmablasts/naive CD4 + T cells in these cultures was directly proportional to Tfh cell differentiation (Supplemental Fig. 1H). To ascertain whether the differentiated Tfh cells were functional, T cells were repurified after 4 d of culture with plasmablasts and incubated with freshly isolated autologous naive B cells or memory B cells, as previously described (6). Plasmablast-induced Tfh cells provided enhanced B cell help and supported naive B cells and memory B cells to produce significantly more IgM (Fig. 2D) and IgG (Fig. 2E), respectively, compared with T cells that had not been exposed to plasmablasts. To our knowledge, our results show for the first time that human plasmablasts are potent inducers of Tfh cell differentiation. Of relevance, murine plasma cells were shown to inhibit, rather than promote, Tfh cell formation (7). This discrepancy may reflect differences between plasma cells and circulating plasmablasts, or it could be due to differential regulation of Tfh cells in mice and humans.

Plasmablast-derived IL-6 induces Tfh cells
In our efforts to elucidate the mechanism by which plasmablasts induce Tfh cell differentiation, we discovered that plasmablasts produce large amounts of IL-6, more than either naive or memory B cells (Fig. 3A, Supplemental Fig. 2A). Plasmablasts cultured with naive T cells yielded more IL-6 than did naive B cells and T cells or naive T cells alone (Fig. 3B). In addition, IL-6 detected in the coculture supernatants correlated with the percentage of Tfh cells (Supplemental Fig. 2B). Addition of IL-21 and IL-6 to naive T cells significantly induced Tfh cell differentiation (Fig. 3C,  Supplemental Fig. 2C). Tfh cell differentiation was greater when a combination of IL-21 and IL-6 was used, confirming previous observations in mice whereby loss of both IL-6 and IL-21 led to a significant reduction in Tfh cell numbers (2,3). Although IL-12 has been implicated in human Tfh cell differentiation (8), we could not detect IL-12 production by plasmablasts.
We next investigated whether the production of IL-6 and IL-21 mediated the differentiation of Tfh cells by plasmablasts. In vitro blockade of IL-21R or IL-6 in cocultures of naive T cells and plasmablasts significantly decreased the acquisition of a Tfh cell phenotype (Fig. 3D). Blockade of both cytokines further reduced the percentage of Tfh cells to the level found when naive B cells replaced plasmablasts in these cultures. Furthermore, IL-6 blockade in cultures of plasmablasts and naive T cells from healthy individuals also reduced IL-21 production by T cells (Fig. 3E), suggesting that IL-6 plays an important role in inducing Tfh cells through the stimulation of T cell IL-21 production.

Tocilizumab therapy in patients with RA reduces Tfh cells and plasmablasts
To glean some insight into the in vivo relevance of our findings in patients, we investigated the interrelationships between Tfh cells and plasmablasts in patients with RA before and after tocilizumab therapy. Tocilizumab targets IL-6R and is an effective treatment for patients with RA (9). In agreement with previous studies (10), RA patients showed significantly more circulating Tfh cells compared with healthy individuals (Figs. 1A, 3F, 3G). Consistent with our in vitro data, RA patients responding to tocilizumab therapy showed a marked reduction in circulating Tfh cells (Fig. 3G). Paralleling our in vivo observations, CD4 + T cell IL-21 production was decreased significantly after treatment with tocilizumab (Fig. 3H). Tocilizumab treatment also reduced the plasmablast frequency in patients with RA (Fig. 3I, 3J). Moreover, there was a significant correlation between the percentage of Tfh cells and plasmablasts in patients with RA after tocilizumab treatment (Fig. 3K). These results suggest that tocilizumab could ameliorate disease by reducing Tfh cells and Ab-secreting cells in RA.
Our data implicate a previously unrecognized function of human circulating plasmablasts as potent inducers of Tfh cell differentiation via IL-6. One can envisage a feedback mechanism by which circulating plasmablasts traffic from one inflammatory site and create further foci of inflammation through the differentiation of Tfh cells, which, in turn, induces plasmablast formation. This feedback mechanism is tightly controlled under normal conditions; however, in infections and autoimmune diseases it could rapidly result in large numbers of Tfh cells and plasmablasts in several locations. This amplification loop represents an important therapeutic target in autoimmune diseases, such as RA, and it also could be boosted to improve responses to infection and to enhance vaccination strategies.