Follicular helper T cells (TFH cells) are a specialized subset of T cells that provide signals necessary for antigen-specific B cells to generate the germinal center (GC). This structure is required for class-switch recombination, somatic hypermutation and the selection of B cells that produce high-affinity antibodies, as well as for the generation of long-lived antibody-secreting plasma cells and memory B cells1. A distinguishing marker of TFH cells is the chemokine receptor CXCR5, which is required for their entry into B cell follicles. TFH cells access the follicle by upregulating CXCR5 and by downregulating CCR7 and P-selectin glycoprotein ligand 1 (CD162). The TFH cells with the highest expression of CXCR5 and another marker, PD-1, seem to 'preferentially' accumulate in GCs by sensing the CXCR5 ligand CXCL13 (refs. 2,3) (Fig. 1). TFH cells also express several costimulatory molecules, including CD40L, ICOS, OX40 and members of the SLAM family, all of which have important roles in T cell–dependent B cell responses driven by cognate T cell–B cell interactions2,3. TFH cells are also the source of interleukin 21 (IL-21) and IL-4, cytokines that are necessary for class-switch recombination. In this issue of Nature Immunology, Lifan Xu et al.4 and Youn Soo Choi et al.5 identify the transcriptional regulators LEF1-1 (encoded by Lef1) and TCF-1 (encoded by Tcf7) as important participants early in the TFH developmental process (Fig. 1).

Figure 1: LEF-1 and TCF-1 control early steps in the TFH developmental process.
figure 1

Marina Corral Spence/Nature Publishing Group

LEF-1 and TCF-1 are positive regulators that control the expression of Bcl6, Il6ra, Il6st and Icos, which encode products known to be required for TFH differentiation. IgG1, immunoglobulin G1; CCL19 and CCL21, chemokines; TCR, T cell antigen receptor; DC, dendritic cell.

The lineage commitment of TFH cells begins with the initial priming of naive CD4+ T cells by dendritic cells or by other myeloid cell–derived antigen-presenting cells in the T cell zone (Fig. 1). After a few rounds of cell division, these T cells express Bcl-6, a transcriptional repressor that belongs to the BTB-POZ family, whose function is critical for TFH differentiation, as indicated by the complete absence of TFH cells among Bcl6−/− CD4+ T cells6,7,8. Moreover, ectopic Bcl6 expression in CD4+ T cells reconstitutes the generation TFH cells. Bcl-6 expression in B cells is also critical for TFH development, because TFH cells are again abrogated in mice with conditional deficiency of Bcl-6 in B cells8. Those observations indicate that the transcriptional program of Bcl-6 seems to be controlled via priming by antigen-presenting cells.

It is important both for basic knowledge and for vaccine development to understand how the TFH program, including regulation of Bcl6 expression, is controlled. Cytokine pathways, such as IL-6–STAT1, IL-12–STAT4 and IL-21–STAT3, have been linked to the induction of TFH cells and Bcl6 expression2. Batf and Ascl2 are critical transcription factors that support TFH differentiation by controlling the expression of Bcl6 and Cxcr5, respectively, which suggests that Bcl-6 and CXCR5 are regulated by independent molecular mechanisms. Despite such advances, the gap between the regulation of Bcl6 expression and the TFH program is still puzzling, because there are no obvious mechanisms by which Bcl-6 controls transcriptional programs for TFH differentiation, and this makes it likely that other unknown factors are involved.

Choi et al.5 and Xu et al.4 identify two additional participants in the TFH program: LEF1-1 and TCF-1. These are members of the TCF-LEF subfamily and belong to a family of high-mobility-group proteins that work as downstream repressors of the canonical Wnt signaling pathway. TCF-1 controls mainly the proliferation and population expansion of thymocytes; thus, lack of TCF-1 causes a substantial reduction in thymic cellularity and a partial block in thymocyte differentiation at the transition from the CD8+ immature single-positive stage to the CD4+CD8+ double-positive stage. Although LEF1-1-deficient mice have normal T cell development in the thymus, deficiency in both LEF1-1 and TCF-1 leads to a complete block in differentiation9,10. TCF-1 and LEF-1 are also involved in the function and formation of memory CD8+ T cells11. In CD4+ T cells, TCF-1 is reported to be a positive regulator of the expression of genes encoding the transcription factor GATA-3 and IL-17A and a negative regulator of expression of the gene encoding interferon-γ12.

Choi et al. perform an unbiased transcriptome analysis of early TFH cells and the TH1 subset of helper T cells to identify transcription factors that control the TFH program5. They identify Lef1 because it satisfies two further criteria: its 'preferential' expression in early TFH cells in vivo, and the large effect on differentiation into fully committed GC TFH cells and the generation of GC B cells when it is ablated. These authors further demonstrate high expression of Tcf7 by early TFH cells and that synergistic regulation by LEF-1 and TCF-1 is essential for the full TFH cell program. Studies using a green fluorescent protein (GFP) reporter system (TCF-1–GFP) further support the proposal of the functional importance of TCF-1 in TFH cells. The expression of TCF-1–GFP is highest in naive T cells but is lower in antigen-primed T cells and effector T cells. After infection with lymphocytic choriomeningitis virus, TCF-1–GFP expression is greatly diminished in TH1 cells, while high expression is maintained in TFH cells. Mature T cells with conditional deletion of Lef1 and Tcf7, which lack both Lef-1 and TCF-1, generate considerably fewer fully committed GC TFH cells than do their wild-type counterparts, which suggests that coordination of LEF-1 and TCF-1 is required for TFH differentiation through the regulation of circuits upstream of Bcl6. The binding of LEF-1 and TCF-1 to several different genes consequently results in the promotion of Bcl6 expression, sustained expression of the cytokine receptor subunits IL-6Rα and gp130 (IL-6st), and enhanced expression of ICOS to accomplish TFH differentiation.

Xu et al. also find higher expression of TCF-1 in early committed TFH cells after infection with lymphocytic choriomeningitis virus, in contrast to its decreased expression in TH1 cells4. They also show diminished development of fully committed TFH cells resulting from TCF-1 deficiency in two independent conditional deletion systems (T cell–specific deletion (Cd4-Cre) and tamoxifen-induced deletion (ERT2-Cre)), in system of chimeras reconstituted with a mixture of Tcf7−/− bone marrow and wild-type bone marrow, and in a system of short hairpin RNA–mediated knockdown. However, unlike Choi et al.5, they find that TCF-1 needs no cooperation with LEF-1 in the TFH program. Moreover, TFH cells lacking TCF-1 show lower expression than TCF-1-sufficient cells of genes encoding several TFH markers (including Bcl6, Icos, Ascl2, Cxcr5, Il6ra, Il6st, Il21 and Il4) and instead show increased expression of non-TFH cell signature genes (Tbx21, Gzmb, Gata3, Rorc and Foxp3); this suggests that TCF-1 is a cell-fate regulator that functions by suppressing the differentiation programs of non-TFH effector cells.

Xu et al. further demonstrate that expression of TCF-1 is critical for the axis of Bcl-6 and the transcription factor Blimp-1 in TH1-versus-TFH cell-fate determination4. Studies of an in vitro 293T human embryonic kidney cell overexpression system indicate that co-expression of the p33 isoform of TCF-1, Bcl-6 and TLE3 (the dominant member of the TLE corepressor complex) in CD4+ T cells enhances formation of the p33–Bcl-6 complex. These in vitro data indicate that TCF-1 binds both directly and indirectly to the Bcl6 promoter to upregulate its expression, while indirect binding of the p33–Bcl-6 complex to 5′ regulatory regions of the gene encoding Blimp-1 represses Blimp-1 expression. Given these results, Xu et al. propose that TCF-1 is an upstream regulator of the Bcl-6–Blimp-1 axis required for TFH differentiation4.

These two studies raise some interesting questions about how the activity of TCF-1 and/or LEF-1 is able to influence the differentiation program of TFH cells and other helper T cell subsets. Both studies agree on the point that TCF-1 is a positive regulator of Bcl-6 and a negative regulator of Blimp-1, but there is clear discrepancy in the proposed activity of LEF-1. Choi et al. suggest that LEF-1 is needed in a synergistic way with TCF-1 to regulate commitment to the TFH lineage5, while Xu et al. raise the conflicting possibility that LEF-1 may not necessarily lead to full commitment4. TCF-1 directly binds to the Bcl6 promoter to control its expression regardless of LEF-1's binding, whereas TCF-1 can repress Lef1 expression. TCF-1 is needed to inhibit Blimp-1 activity to stop differentiation into other helper T cell subsets. In the same context, TCF-1 is known to be positive regulator for the functions of TH2 cells and TH17 cells11, which indicates that TCF-1 must have a complicated role to be able to provide positive differentiation signals for both TFH cell subsets and non-TFH cell subsets. Indeed, a published study has indicated that TFH cells can differentiate from TH2 cells able to express several TFH cell signature genes13. Since both TH2 cells and TFH cells need TCF-1 activity for their differentiation, TCF-1 may provide the mechanism by which TH2 cells have the potential to differentiate into TFH cells. Choi et al.5 and Xu et al.4 propose several interesting molecular mechanisms downstream of TCF-1 and LEF-1, but it remains unclear whether the TFH program is regulated by a single mainstream pathway or by the coordination of multiple pathways. Further research should focus on determining how molecules downstream of TCF-1 and LEF-1 control the TFH-differentiation program.