Bcl6-Mediated Transcriptional Regulation of Follicular Helper T cells (T_FH)

Abstract

Follicular helper T cells (T_FH) are essential B cell-help providers in the formation of germinal centers (GC), affinity maturation of GC B cells, differentiation of high-affinity antibody-producing plasma cells, and production of memory B cells. Transcription factor (TF) Bcl6 is at the center of gene regulation in T_FH biology, including differentiation and function, but how Bcl6 does this, and what additional TFs contribute, remain complex questions. This review focuses on advances in our understanding of Bcl6-mediated gene regulation of T_FH functions, and the modulation of T_FH by other TFs. These advances may have important implications in deciphering how repressor TFs can regulate many immunological cell types. An improved understanding of T_FH biology will likely provide insights into biomedically relevant diseases.

Follicular helper T cells and Bcl6

Vaccines are among the most cost-effective treatments in modern times [1], and most current human vaccines function by eliciting protective antibody (Ab) responses. Indeed, a key aspect of adaptive immunity to many pathogens and vaccines is the B cell help provided by T cells. As such, mammalian follicular helper CD4⁺ T cells (T_FH) are specialized B cell help facilitators [2]. In humans and mouse models, there is compelling evidence that T_FH cells frequently limit the magnitude of germinal center (GC; see Glossary) responses and, hence, limit the production of GC-derived high-affinity Abs, memory B cells, and long-lived plasma cells, which constitute the basis of long-lived protective humoral immunity [3]. Therefore, an improved understanding of T_FH cell biology can enhance our understanding of anti-pathogen immune responses and vaccine-elicited humoral immunity, as well as enable strategies to mitigate autoimmunity in numerous contexts [3].

B cell lymphoma 6 (Bcl6) -- initially identified as the master regulator of GC B cells differentiation [4] -- is recognized as the lineage-defining transcription factor (TF) for T_FH differentiation [5–7]. Work by several laboratories, including ours, established that the differentiation and function of T_FH cells depend on the expression of the transcriptional repressor Bcl6 [5–7]. The generation of T_FH is completely impaired in mouse Bcl6^−/− CD4⁺ T cells [5–7], and enforced Bcl6 expression by retroviral transduction in CD4⁺ T cells elicits robust T_FH differentiation in mice [5]. Recent work showed that tamoxifen-mediated temporal Bcl6 ablation in CD4⁺ T cells (Bcl6^fl/fl CD4-Cre^ERT2) results in the transition of already established T_FH cells into T helper type 1 (T_H1) cells during acute lymphocytic choriomeningitis virus (LCMV) infection (Armstrong strain) in mice [8]; this suggested that Bcl6 can be essential not only for the development of T_FH cells, but also for the maintenance of T_FH integrity [8]. This is important because continuous help signals from T_FH are crucial for the maintenance of GCs [3]. While T_FH cells have been extensively studied, how Bcl6 accomplishes its functional task as a lineage-defining TF has remained a longstanding knowledge gap. Elucidating immunobiology underlying the differentiation of T_FH and the process of generating protective antiviral antibody responses may facilitate future vaccine development. This review discusses recent advances concerning Bcl6 mediated transcriptional regulation of T_FH cells, including the demonstration that genes positively associated with a cell type can be upregulated downstream of a lineage-defining TF located at the apex of a repressor-of-repressors network. The up-to-date TF network in T_FH differentiation is also summarized, aiming to increase our understanding of the gene regulatory circuitry of T_FH cells.

T_FH and GC-T_FH Differentiation in Mice

T_FH differentiation is a multi-factorial and multi-stage process [3,9]. Our group, and others, have shown that T_FH differentiation begins at the time of dendritic cell (DC) priming in mice [3]. It initiates with activation of a naïve CD4⁺ T cell receiving signals in the form of peptide:MHC, costimulatory molecules, and cytokines provided by an antigen-presenting cell (APC) such as a DC (Fig. 1). Inducible T-cell Costimulator (ICOS) and interleukin-6 receptor (IL-6R) provide signals for early T_FH differentiation in mice, both of which contribute to Bcl6 expression [3,9]. Helper T cell fate decision in vivo is determined as early as the 2^nd cell division: Bcl6⁺ CD4⁺ T cells (further differentiate to T_FH) and Blimp1⁺ CD4⁺ T cells (further differentiate into a non-T_FH fate) [10–13]. There is a reciprocal antagonistic relationship between Bcl6 and Blimp1 (encoded by Prdm1); Bcl6 represses Prdm1, and Blimp1 represses Bcl6 [5,14]. T_FH cells accomplish the inhibition of Prdm1 by Bcl6 in coordination with Tcf1 (transcription factor 1; encoded by Tcf7) and additional TFs [3,15–17], as described below.

Figure 1. The differentiation process of T_FH and GC-T_FH cells in mice.

T_FH differentiation is a multi-factorial, multi-stage process. Bcl6 is the lineage-defining TF for T_FH differentiation. T_FH differentiation is initiated by dendritic cell (DC) priming. CD4⁺ T cells migrate to the T-B border of secondary lymphoid organs where T_FH cells interact with cognate B cells. B cells provide additional signals to T_FH cells to drive GC-T_FH differentiation. This figure was created using BioRender (https://biorender.com/).

Cognate B cells interact with T_FH cells through peptide:MHC, CD40L, CD80/CD86, and SLAM receptors, and provide additional signals to T_FH cells to drive GC-T_FH differentiation [3]. GC-T_FH cells can be identified by a number of markers, C-X-C motif chemokine receptor 5 (CXCR5)^hi Bcl6^hi Programmed cell death protein-1 (PD-1)^hi SLAM-associated protein (SAP)^hi ICOS⁺ B- and T-Lymphocyte-Associated Protein (BTLA)^hi CD200^hi in humans and mice (Fig. 1) [2]. Almost all of these core signature markers of GC-T_FH cells are conserved across species [18–20]. In GCs, fully polarized GC-T_FH cells provide signals for the generation and maintenance of GC B cells [21,22]. GC-T_FH cells are commonly limiting for GCs [5,23–25], indicating that focusing on T_FH biology may be an effective way to enhance vaccine-elicited humoral immunity. In addition, TFs are central controllers of T cell differentiation and function, as described below.

Bcl6 Control of T_FH Differentiation and Function: Direct repression

Multiple models of T_FH differentiation have been suggested [2]. One proposed model suggested that T_FH fate is the default differentiation pathway of activated CD4⁺ T cells [2]. In this model, Blimp1⁺ CD4⁺ T cells potentially differentiate into non-T_FH effector cells, T_H1, T_H2, T_H17, or regulatory T cells (T_REG), through repression of the default T_FH pathway by Blimp1. The primary role of Bcl6 in that model would be to repress Blimp1, blocking non-T_FH differentiation pathways and thus, releasing the default T_FH differentiation pathway [26]. We recently tested this hypothesis using multiple genetically modified mice [26]. If T_FH were the default pathway of CD4⁺ T cell differentiation, CD4⁺ T cells might be presumed to differentiate into T_FH cells in the absence of both Bcl6 and Prdm1[26]. However, T_FH differentiation did not occur when Bcl6/Prdm1 double-deficient CD4⁺ T cells (Bcl6^fl/fl Prdm1^fl/fl CD4-Cre) were primed in mouse LCMV infection and KLH-gp₆₁ immunization models [26]. Furthermore, GCs did not develop in CD4-specific Bcl6-deficient mice (Bcl6^fl/fl CD4-Cre, devoid of endogenous T_FH development) when receiving Bcl6^fl/fl Prdm1^fl/fl CD4-Cre CD4⁺ T cells, indicating that Bcl6/Prdm1 double-deficient CD4⁺ T cells cannot differentiate into functional T_FH cells. Similar observations were made using Bcl6^fl/fl Prdm1^fl/fl CD4-Cre mice in a sheep red blood cells (SRBC) injection model [27]. These results demonstrated that T_FH was not the default fate of activated CD4⁺ T cells [26]. This was also relevant as it indicated that Bcl6 had important activities beyond inhibiting Prdm1 to drive T_FH differentiation [26].

How does Bcl6 regulate T_FH differentiation and function? Bcl6, as an obligate repressor in mouse B cells [28–31], might control non-T_FH and T_FH genes, presumably by at least two repressor modes-of-action: (i) direct repression and (ii) repression-of-repressor mechanisms. Genes downregulated in T_FH cells (and genes upregulated in non-T_FH cells) are likely controlled by the direct repression mechanism as Bcl6 is highly expressed in T_FH cells. This category of genes could include lineage-defining TFs for alternative cell fates, cytokines and cytokine receptors expressed by non-T_FH cells, as well as migration-associated genes that are downregulated in T_FH cells. Genes upregulated in T_FH cells (and genes downregulated in non-T_FH cells) might be controlled by the indirect repression-of-repressor mechanism as their expression amounts positively correlate with Bcl6 expression in T_FH cells. This category of genes might include molecules positively involved in T_FH differentiation and function (Fig. 2A) [26].

Figure 2. Model of control of T_FH differentiation and function by Bcl6.

(A) Bcl6 can control non-T_FH and T_FH genes by at least two mode-of-action: (i) direct repression and (ii) repression-of-repressor mechanisms. Bcl6 directly binds and represses a set of genes for alternative cell fates, cytokines, receptors, and migratory genes. Bcl6 indirectly upregulates important functional molecules by repression-of-repressor mechanisms via a set of Bcl6 target TFs (Bcl6-r TFs), including Prdm1, Id2, Runx2, Runx3, and Klf2. Illustration concept based on [43]. (B) Runx2 and Runx3 repress important T_FH molecules, including Cxcr5, Icos, and Cd200, downstream of Bcl6. (C) Klf2 represses important T_FH molecules, including Cd200, Pdcd1, Icos, Il6ra, Il21, and Il4, downstream of Bcl6 [26].

Bcl6 directly binds and represses a set of TFs important for alternative cell fates (Fig. 2A). Prdm1, Tbx21, Gata3, Rora, Rorc, and Stat5, are all directly repressed by Bcl6 in human GC-T_FH cells and mouse CD4⁺ T cells [6,7,32–35]. Moreover, Bcl6 inhibits non-T_FH differentiation fates by repressing Prdm1 in mice [5,9,11]. In human T_FH, BCL6 inhibits Runt-related transcription factor 3 (RUNX3) [26,34]. Enforced expression of Runx3 by retroviral transduction has also been shown to disrupt T_FH differentiation and increase the formation of a T_H1 population in mouse LCMV infection [26]. From another angle, Bcl6 repression of T-bet, GATA-3, and RORγt in the context of T_FH differentiation has been somewhat controversial, given that some TFs are co-expressed with Bcl6 [36,37]. For example, T-bet co-expression with Bcl6 is observed at early T_FH/T_H1 differentiation in mouse LCMV infection [38,39]. Another example at steady state, is GATA-3, which is co-expressed with Bcl6 in T_FH cells in Alternaria extract-immunized mice [36]. Simultaneous expression of GATA-3 and Bcl6 is also observed in T_FH cells in Ets1-deficient (Ets1^fl/fl CD4-Cre) mice harboring systemic lupus erythematosus (SLE)-like disease [37]. It seems that alternative fate TFs might be expressed in the presence of Bcl6 under particular pathologic conditions and drive different gene expression programs, including signature genes, such as those upregulated in T_FH and T_H2 cells. However, direct evidence of Bcl6 inhibition via T-bet, GATA-3, and RORγt will require experiments assessing the selective disruption of Bcl6-mediated inhibition under normal versus pathologic conditions.

Bcl6 can also directly bind and repress key cytokines, receptors, and migration genes that are expressed by non-T_FH cells (Fig. 2A). For instance, Bcl6 binds the IL17A and IL17F loci in human GC-T_FH cells [34] and represses the expression of Il17a in mouse CD4⁺ T cells [6]. Moreover, cytokine receptors essential for signals supporting T_H1 (IL-12R, and IL-23R), T_H2 (IL-4R), T_H17 (IL-23R, TGFβR), and T_REG (TGFβR) are direct targets of Bcl6 [34]. The activation of Blimp1 and Signal transducers and activators of transcription 5 (STAT5) by IL-2/IL-2R signaling is a potent inhibitory mechanism of T_FH differentiation [11,40,41]. IL-7 mediated activation of STAT5 can also contribute to repression of T_FH associated genes, including Bcl6 [35,42]. Bcl6 also binds the Il2r and Il7r genes and represses their expression [34,35]. Furthermore, migration of T_FH cells to the T-B border and further localization to GC is accomplished by upregulation of CXCR5 and downregulation of C-C motif chemokine receptor 7 (CCR7), P-selectin glycoprotein ligand-1 (PSGL1; encoded by Selplg), and EBV-induced G-protein-coupled receptor 2 (EBI2; encoded by Gpr183) [43,44]. EBI2 is required for positioning of T_FH cells at the T:B border, and later, EBI2 expression is downregulated to complete further migration of T_FH cells to GCs [44,45]. Indeed, Bcl6 binds to Ccr7, Selplg, and Gpr183 genes and downregulates their expression, thereby preventing chemokine receptor-dependent migration to the T cell zone and enabling proper localization of GC-T_FH cells to GCs [26,34,44]. In sum, Bcl6 directly binds and represses genes downregulated in T_FH cells, including important TFs and functional molecules for migration and for acquiring alternative CD4⁺ T cell fates (Fig. 2A).

Bcl6 gene regulatory circuits: A Repressor-of-repressors

Bcl6 is an example of a TF acting as a lineage-defining TF via direct repression [34]. It also upregulates important functional molecules, such as CXCR5, in T_FH cells by repression-of-repressors, i.e. indirect induction, in mice (Fig. 2A) [26]. One general proposal of differentiation of lymphocytes has been that a main function of lineage-determining TFs might be to restrict the number of genes that are widely induced by T cell receptor (TCR) and cytokine signaling via repression (or other mechanisms) [46]. For example, Foxp3 is a lineage-defining TF that acts primarily as a repressor. A recent study showed that Foxp3 downregulates Tcf1 expression to indirectly establish T_REG-specific chromatin accessibility changes for the Tcf1-bound region in mouse T_REG cells [47]. Accordingly, work from our laboratory demonstrated how a lineage-defining TF, located at the apex of a repressor-of-repressors network, might upregulate downstream genes that are positively associated with a specific cell type (see below). Specifically, this study provided a thorough analysis of Bcl6 target TFs (called “Bcl6-r”) involved in positive regulation of T_FH genes, and thus, specific Bcl6-r TFs, including Runx2, Runx3, and Krüppel-like factor2 (Klf2) were identified [26] (Fig. 2B, C), as discussed in detail below.

Prdm1 and Id genes

Targeted by Bcl6, Blimp1 accomplishes positive expression of T_FH-associated proteins via a repression-of-repressor circuit. The upregulation of CXCR5 and PD-1 in T_FH cells is controlled by the Bcl6-directed repression of Blimp1. Enforced expression of Blimp1 by retroviral transduction represses the expression of Cxcr5 in mouse CD4⁺ T cells [5]. Moreover, Blimp1 directly binds to the 5’ upstream enhancer and intron 1 of the Cxcr5 gene and represses the expression of Cxcr5 in follicular regulatory T cells (T_FR) and follicular cytotoxic T cells in mice [48,49]. Blimp1 directly represses PD-1 by repressing the expression of NFATc1, an activator of PD-1 expression in mouse CD8⁺ T cells [50]. In addition to Blimp1, Bcl6 represses Id2, and Id2 inhibits Cxcr5 expression by complexing with E proteins, particularly E2A and HEB, early in T_FH differentiation [51], and Achaete-scute homolog 2 (Ascl2), late in T_FH differentiation in mice [52]. Regarding PD-1, Bcl6 also represses T-bet, which is known to repress PD-1 expression, resulting in the upregulation of PD-1 expression via Bcl6 in mouse T_FH cells [27] (Fig. 2A).

Runx family

Runx family members are known to act as activators or repressors of transcription at different genetic loci [53,54]. Runx proteins are required for the silencing of the Cd4 promoter in thymic development and for silencing Il4 in T_H1 cells [55–57]. In naïve CD8⁺ T cells, Runx3 is well known to play a pioneering role during TCR activation to both promote chromatin accessibility and direct chromatin condensation, and to cooperate with T-box transcription factors to program regions of stable chromatin accessibility that establish transcription in effector and memory CD8⁺ T cells in mice [58,59].

Runx2 and Runx3 repress important T_FH molecules, including CXCR5, ICOS, and CD200, in mice [26] (Fig. 2B). Although the effect on T_FH gene expression by disruption of Runx2 or Runx3 has been similar, Runx3 can suppress T_FH differentiation more strongly than Runx2 when overexpressed by retroviral transduction [26], suggesting that Runx proteins might have distinct regulatory mechanisms in CD4⁺ T cells, including DNA binding affinity [60]. Runx3-deficient (Runx3^fl/fl hCD2-Cre) CD8⁺ T cells aberrantly upregulate T_FH-associated genes, including Cxcr5 [61], and possess some characteristics of follicular CD8⁺ T cells [49,62,63], suggesting that Runx3 may repress a follicular program in both CD4⁺ and CD8⁺ T cells. Given the extensive overlap in TF circuitry between effector CD4⁺ and CD8⁺ T cell differentiation [13], there may be additional interplay between Bcl6, Runx2/Runx3, and Blimp1 during early differentiation of effector CD4⁺ T cells, although this remains to be further assessed. This is relevant, as such interplay of highly connected factors reciprocally controlling both T_FH and non-T_FH genes might represent a general feature of lymphocyte gene networks that provide tight gene regulation, thus reinforcing specific differentiation outcomes.

Klf2

Klf family members are well-known regulators of naïve T cell migration to T cell zone in secondary lymphoid organs. Klf2 is essential for the maintenance of CCR7, CD62L, PSGL1, and Sphingosine-1-phosphate receptor 1 (S1PR1) expression, all involved in T cell migration in mice [64,65]. Klf2 is also involved in T_FH differentiation downstream of ICOS signaling, involving the migration genes above, but that process has been considered Bcl6-independent and primarily focused on genes upregulated in non-T_FH cells in mouse Listeria monocytogenes infection and NP-OVA immunization models (downregulated in T_FH cells; i.e., Klf2 functioning as an activator) [66,67]. In contrast, our group revealed that Klf2 is an important Bcl6-r TF repressor of T_FH genes [26] (Fig. 2C) (see below). Klf proteins contain both transactivating and transrepression domains [68], and while Cxcr5 can be bound by Klf2 [67], Cxcr5 regulation is not a primary target of Klf2 activity in T_FH cells in mouse LCMV infection [26]. This was concluded from head-to-head studies in which Blimp1 and Id2/E proteins exhibited more dominant regulation of Cxcr5 expression than Klf2 [26,51]. Instead, Klf2 was characterized as a potent repressor of CD200, PD1, IL-6Rα, IL-4 [69], and IL-21, as well as the TF Tcf1 in mouse LCMV infection and KLH-gp₆₁ immunization models [26]. While IL-4 and IL-21 are the canonical cytokines of T_FH cells, and T_FH cells require Bcl6, no clear connection between gene regulation of these cytokines and Bcl6 had been previously described [9,43,70]. Evidence showed that Klf2 could negatively regulate IL-4 and IL-21 of CD4⁺ T cells and was itself negatively regulated by Bcl6 in mouse LCMV infection and KLH-gp₆₁ immunization models [26], suggesting a potential regulatory circuit of Bcl6 inhibiting Klf2 inhibiting Il21 and Il4 (Fig. 2C). However, this gene circuitry has substantial additional complexity and entanglements. Klf2 is repressed via Foxo1 inactivation by ICOS signaling [67] or by Itch [71], while Foxo1 inactivation also releases Bcl6 expression in mice [71,72]. Additionally, Klf2 represses Tcf7 in mice [26]. Tcf1 is an important TF for T_FH differentiation and function, directly mediating induction of Bcl6, maintaining expression of the IL-6 receptor chains, and inducing ICOS, among other functions in mice [15–17]. Thus, a positive feedback loop appears to exist with loss of Klf2 inducing ICOS expression via Tcf1 and Bcl6 blocking Runx2/Runx3/Klf2 pathways. In turn, ICOS expression can repress Klf2 via Foxo1 inactivation and thus induce Bcl6 expression in mouse CD4⁺ T cells [26,67] (Fig. 3). In sum, Bcl6 indirectly regulates a set of genes upregulated in T_FH cells, including important genes for T_FH differentiation and function; this occurs via Bcl6 repression of TFs, including Blimp1, Id2, Runx2, Runx3, and Klf2 (Fig. 2A–C).

Figure 3. The transcription factor network of T_FH cell differentiation in mice.

The transcriptional regulatory network of T_FH cell differentiation is shown. Transcription factors activating (arrows) or repressing (closed lines) the genes associated with T_FH cell differentiation and function are described. The major Bcl6-target transcription factors (Bcl6-r TFs) are indicated in dark blue boxes. Illustration inspired by [9]. This figure was created using BioRender (https://biorender.com/).

TF Networks in T_FH Cell Differentiation

Tcf1, Maf, Lef1, and Batf

In addition to Bcl6, T_FH cells express a number of additional TFs, some of which modulate T_FH cell functions [2,9,43] (Fig. 3, Box 1). Among other TFs, Tcf1, Maf, and Batf are considered key regulators of early T_FH differentiation. Tcf1 and Lymphoid enhancer-binding factor 1 (Lef1) are key regulators of early T_FH differentiation and are involved in the induction of Bcl6, Icos, and Il6r in mice [15–17]. Tcf1 directly binds to the promoter of Bcl6 and recruits the histone methyltransferase Ezh2 to activate Bcl6 transcription in mice [73]. Moreover, Tcf1 represses Blimp1 and IL-2R expression by directly binding to intron 3 of the Prdm1 gene and intron 1 and upstream regulatory region of the Il2ra gene in mice [17,74]. Tcf1 also suppress Ctla4 expression in T_FH cells by its intrinsic histone deacetylase activity in KLH-gp immunized mice [75]. Enforced expression of Lef1 by retroviral transduction augments T_FH differentiation and enhances expression of IL-6 receptors and ICOS, while genetic ablation of Lef1 (Lef1^fl/fl CD4-Cre) impairs GC-T_FH differentiation in mice [15]. Tcf7 and Lef1 double deficiency (Tcf7^fl/fl Lef1^fl/fl CD4-Cre) results in more severe impairment of T_FH generation than Tcf7 single deficiency (Tcf7^fl/fl CD4-Cre), suggesting that Tcf1 and Lef1 might synergistically regulate T_FH genes [15]. Batf is a positive regulator of Bcl6 and Maf and is considered as an early activation-associated TF in mice [76]. Batf directly binds to upstream enhancers of Bcl6 and Maf in mice [76]. And, Maf instructs early T_FH differentiation in cooperation with Bcl6 in both humans and mice [70,77,78].

Box 1. Other TFs Involved in T_FH Differentiation in Mice.

In addition to the TFs discussed in the main text, Thpok, early growth response gene 2/3 (Egr2/3), Bob1, NF-κB1, Ets1, and peroxisome proliferator-activated receptor γ (PPARγ) have recently been reported to positively or negatively regulate T_FH differentiation in mice (see Figure 3 in the main text). For instance, Thpok, Egr2/3, and Bob1 have all been suggested to act upstream of Bcl6 [78, 124, 125]. Thpok (encoded by Zbtb7b) is an upstream regulator of Bcl6 by binding to intron 1 of Bcl6, and genetic ablation of Zbtb7b (Zbtb7b^fl/fl Ox40-Cre) in CD4⁺ T cells interfered with Bcl6 and Maf expression and T_FH differentiation [78]. Egr2 directly regulates Bcl6 expression and induces T_FH differentiation [124]. Specifically, T_FH differentiation is defective in Egr2/3 double-deficient (CD2-Egr2^−/−Egr3^−/−) CD4⁺ T cells. Enforced expression of Bcl6 by retroviral transduction in Egr2/3-deficient CD4⁺ T cells rescued T_FH differentiation and GC formation, suggesting that Egr2/3 are potential regulators of Bcl6 in mice [124]. Transcriptional coactivator Bob1 (encoded by Pou2af1) can directly bind to Bcl6 and Btla promoters and induces their expression in cooperation with octamer transcription factors 1/2 (Oct1/2) [125]. Moreover, CD4-specific deletion of Nfkb1 selectively impairs Cxcr5 expression, while other T_FH genes, including Bcl6, ll21, and Pdcd1, remain intact [126]. Also, Ets1 suppresses T_FH- and T_H2-associated genes, including Cxcr5, Bcl6, and ll4ra. Ets1-deficient CD4⁺ T cells (Ets1^fl/fl CD4-Cre) increase GATA-3⁺Bcl6⁺lL-4⁺ CD4⁺ T cell populations and induce SLE-like autoimmunity compared with WT cells [37]. Moreover, PPARγ can negatively regulate broad T cell activation [127]. Specifically, PPARγ deficiency in CD4⁺ T cells (PPARγ^fl/fl CD4-Cre) resulted in increased Bcl6 and IL-21 expression, as well as moderate autoimmunity in old mice, compared with controls [127].

Tcf1, Maf, and Batf also appear to cooperatively generate a positive regulatory network controlling cytokine production in T_FH cells (Fig. 3). Maf induces IL-21 expression in mouse CD4⁺ T cells by binding to the Il21 promoter and the CNS2 enhancer [79], and MAF can regulate IL-21 expression in human CD4⁺ T cells [80]. Maf induces IL-4 expression together with JunB and NFAT in a GATA-3 independent manner [81,82]. Batf is important for IL-4 expression in T_FH cells by binding to the CNS2 region in the Il4 locus in a complex with Interferon response factor 4 (IRF4) in mice [83]. Tcf1 may be involved in the expression of both IL-21 and IL-4 in mouse T_FH cells [26]. Also, the Il21 promoter has a Tcf1 binding site. And, in Klf2-deficient CD4⁺ T cells (achieved via CRISPR-Cas9-mediated gene knockdown), the expression of Tcf1, IL-21, and IL-4 were increased in mice relative to wildtype (WT) controls [26]. Thus, T_FH cells might induce IL-21 and IL-4 expression through Maf, Batf, and Tcf1 [26,79,80,82,83]. In contrast, T_H2 cell expression of IL-4 appears to be more GATA-3-dependent than T_FH [26,32]. The induction of IL-21 by Maf in T_FH cells is ICOS dependent [84], and Tcf1 reinforces Icos expression in T_FH cells [15], suggesting a Tcf1–ICOS-Maf–IL-21 positive feedback loop. However, Klf2-deficient CD4⁺ T cells (CRISPR-Cas9-mediated gene knockdown) increased IL-21 expression while not changing Maf expression; instead, increased Tcf1 expression was associated with increased IL-21 expression in mice [26], adding a more complex regulatory mechanism of IL-21 expression in T_FH cells. Further studies assessing temporal kinetics of TF expression and binding at Il21 and Il4 gene loci are necessary to dissect the mechanism of timely expression of IL-21 and IL-4 proteins in T_FH cells [85].

Irf4

IRF4 is also considered an early activation-associated TF that regulates the expression of downstream lineage-defining TF for multiple CD4⁺ T cell subsets in mice [46]. IRF4 can bind to different cis-elements of target genes by interacting with multiple different TFs (e.g., Jun, Batf, STAT3, or NFATc2) [86–88]. IRF4 is essential for T_FH differentiation, as Irf4-deficient mice (Irf4^−/−) have reduced numbers of T_FH cells relative to WT mice [88,89]. IRF4 facilitates IL-21 signaling by complexing with STAT3 and binding to genes, including Prdm1 in mice [88]. Recent mouse studies show that IRF4 expression correlates with TCR signaling strength [90,91]. TCR signal strength has been associated with T_FH differentiation in context-dependent ways [41,91–94]. IRF4 can complex with different binding partner proteins and different complexes, and has different affinities to distinct cis-elements [90]. Greater IRF4 protein abundance frequently results in IRF4 binding to low-affinity DNA binding motifs, including IRF4 binding sites in the Prdm1 gene locus, which can promote non-T_FH fate decisions [91,95].

E proteins

E proteins are recognized as positive regulators of T_FH differentiation, with particular importance for CXCR5 expression [51,52]. The Cxcr5 locus has E-box protein binding sites [26,52,96]. In addition, E2A is expressed in resting and early-activated CD4⁺ T cells, and enforced expression of E2A isoforms augments Cxcr5 expression [51]. E protein activity is inhibited by Id2, and the expression of Id2 is repressed by Bcl6 [26,51]. Although Id2 is an important repressor of CXCR5 expression in mouse T_FH cells, combined genetic ablation of Bcl6, Prdm1, and Id2 (Bcl6 ^fl/fl Prdm1 ^fl/fl Id2^fl/fl CD4-Cre) does not rescue GC-T_FH cells [26], suggesting that additional factors are required to drive the GC-T_FH program beyond Blimp1 and Id2. Ascl2 is a stronger Cxcr5 inducer than other E proteins in mice [51,52]. Enforced expression of Ascl2 by retroviral transduction induces Cxcr5 expression but not Bcl6 expression in mouse CD4⁺ T cells [52]. Ascl2 is generally not expressed early after T cell activation [51], and Ascl2-deficient CD4⁺ T cells (Ascl2^fl/fl CD4-Cre) show moderately reduced T_FH frequency relative to that of WT controls in an influenza virus mouse infection model [52]; this has suggested that multiple E proteins, in addition to Ascl2, might augment Cxcr5 expression by binding enhancer regions, although this remains to be further examined. It is reasonable to speculate that common E proteins such E2A may be important early in T_FH differentiation, while Ascl2 is important in GC-T_FH cells, as Ascl2 expression is highest later in GC-T_FH cells, also awaiting further assessment [51].

Tox proteins

Thymocyte selection-associated high mobility group box protein (Tox) and Tox2 are recently recognized contributors to T_FH differentiation in mice [97,98]. Tox and Tox2 are highly expressed in T_FH cells in a Bcl6-dependent manner [97,98]. IL-6 or IL-21 induce Tox2 mRNA expression through Bcl6 or STAT3 pathway in in vitro cultured CD4⁺ T cells [97]. Enforced expression of Tox2 or Tox by retroviral transduction drives differentiation of T_FH cells in vivo, while combined deletion of Tox2 and Tox (shTox-LMP-GFP_Tox2^−/−) impairs the generation of T_FH in mice [97]. In addition, Tox has been reported to be post-transcriptionally regulated by miR-23~27~24 miRNA clusters in mouse T_FH cells [98]. Tox2 directly binds to and induces T_FH-associated genes, including Bcl6, and promotes chromatin accessibility, suggesting that Tox and Bcl6 proteins might participate in a feed-forward loop, although this warrants further investigation [97]. Bcl6 binds to the Tox2 gene locus, and thus, it is proposed that Tox2 might be a direct positive target of Bcl6 [97]. In contrast, Tox2 and Tox are also defined as genes regulated by Bcl6-r TFs [26], suggesting that Tox2 and Tox might be indirect targets of Bcl6. Current observations from different studies [26,97] support both models: Tox genes are potentially direct targets or indirect targets of Bcl6. Further mechanistic studies are evidently necessary to examine how Tox2 and Tox genes are regulated, and which Bcl6-r TFs might be involved in this process.

STATs

STATs are important signaling molecules that mediate cytokine receptor signaling. Among other STATs, STAT3 is the most potent inducer of Bcl6 expression and T_FH differentiation in humans and mice [99–102]. IL-6 derived STAT3 activation is important for repressing Il2ra and suppressing the T_H1 fate [102] and contributes to IL-4 expression in T_FH cells by binding to the Il4 CNS-2 enhancer in mice [83]. STAT3 physically interacts with Ikaros zinc finger family TF Aiolos (encoded by Ikzf3) to form a TF complex that induces expression of Bcl6 and Il6ra in mice [103]. STAT1 binds to Bcl6, Cxcr5, and Pdcd1 loci and is also involved in early T_FH differentiation in mice [102,104]. STAT4 can also be involved in modulating the expression of Bcl6 and Il21 in mice [39]. By contrast, STAT5 is a strong inhibitor of T_FH differentiation and function in mice [11]. STAT5 is an important mediator of IL-2 receptor signaling [11]. Constitutive expression of STAT5 impairs T_FH differentiation and GC formation in a Blimp1-dependent manner [11]. Stat5 deficiency (Stat5^fl/− by Cre-RV mice) has upregulated Bcl6 expression and T_FH differentiation [105]. Furthermore, IL-6-mediated prevention of STAT5 binding to the Il2rb locus appears to be required for GC-T_FH differentiation in mice [106]. Altogether, the IL-2-STAT5-Blimp1 axis potently negatively regulates T_FH differentiation [41,106].

Foxo1 and Bach2

Foxo1 (forkhead box protein O1) and Foxp1 are examples of negative regulators of T_FH differentiation. Specifically, Genetic ablation of Foxo1 or Foxp1 (Foxo1^fl/fl CD4-Cre or Foxp1^fl/fl Cre^ERT2) [72,107,108] or ubiquitination-mediated degradation of Foxo1 by Itch [71] induce T_FH differentiation in mice. Moreover, ICOS-mTORC2 signaling represses Foxo1 activity in mouse CD4⁺ T cells [109]. As described above, Foxo1-mediated gene regulation appears to be a part of an ICOS—Foxo1—Bcl6—Klf2—Tcf1/ICOS complex feedback mechanism to regulate the T_FH genetic program [26,67]. It is tempting to speculate that Bcl6 might be a central regulator of this positive feedback loop to enhance the upregulation of T_FH genes, which certainly merits further investigation.

BTB and CNC homology 2 (Bach2) is another negative regulator of T_FH differentiation [110,111]. Bach2 maintains the naïve status of CD4⁺ T cells by repressing effector or memory programs in mice [112,113]. Bach2-deficient (Bach2^fl/fl Lck-Cre) naïve CD4⁺ T cells strongly express Blimp1 and IL-4 upon TCR stimulation [113]. Enforced expression of Bach2 by retroviral transduction represses T_FH-associated genes, including Bcl6, Il21, and Tigit, while deletion of Bach2 in CD4⁺ T cells induces mouse T_FH differentiation [110]. Bach2 directly binds to the Bcl6 promoter and competes with IRF4/Batf complexes to bind the Bcl6 promoter [110]. Bach2 also directly binds to the Cxcr5 promoter and negatively regulates the expression of Cxcr5 [111]. In sum, many TFs have been revealed as either positive or negative regulators of T_FH differentiation. Even after more than 10-years of research, there is still a knowledge gap on the mechanistic function of TFs in T_FH biology, either under homeostatic or disease conditions. Further studies on the functionality and underlying mechanisms of T_FH differentiation are thus required to draw a blueprint of the entire TF network in T_FH biology.

Concluding Remarks

Helper T cell differentiation and function are important processes in many diseases. There have been significant advances in T_FH biology in the past decade, including aspects that are relevant in molecular and cellular biology, as well as in disease. Despite these successes, how lineage-defining TFs of CD4⁺ T cell types mechanistically regulate differentiation and function has remained a challenging topic. A parsimonious model of Bcl6 serving as the apex of a repressor-of-repressors network has provided further insights into T_FH regulation, and is relevant for considering how other lymphocyte lineage-defining TFs might function. The detailed study of TFs, including Bcl6, can provide a better understanding of T_FH biology and facilitate relevant biomedical research.

Further studies are needed to gain a better understanding of T_FH gene regulation by Bcl6, including (i) Bcl6 pan-genome binding sites in mouse T_FH cells by chromatin immunoprecipitation with high throughput sequencing (ChIP-seq) or related technologies, (ii) robust and consistent in vitro T_FH differentiation protocols, and (iii) better defined connections between mouse and human T_FH biology. The future directions on TF and T_FH biology are each discussed below (see Outstanding Questions).

Outstanding Questions.

• Bcl6 represses multiple TFs that have different target genes throughout T_FH differentiation and memory development. How are T_FH genes regulated in a timely fashion? How does Bcl6 control each TF at each specific stage? What are the activities of Bcl6-target TFs in T_FH differentiation and function?

• Does the Bcl6 repression-of-repressors network apply to germinal center B cells (B_GC) during differentiation and proliferation? Bcl6 is a key TF in the generation of B_GC [4]. Some Bcl6 target genes overlap between T_FH and B_GC cells, but others do not; therefore, it is possible that T_FH and B_GC cells have different repression-of-repressors networks.

• Can Bcl6 and Bcl6-target TFs be successfully targeted to potentiate putative drug development for T_FH-related diseases, such as atherosclerosis, antibody-mediated autoimmune pathologies, and cancers?

• Are there any functional differences in T_FH cells across species, e.g., humans and mice? Humans and mice share many surface markers of GC-T_FH cells, suggesting that mammalian T_FH cells are evolutionary conserved. However, some induction signals are different between humans and mice, including IL-6 and IL-12 requirements for T_FH in mice and humans, respectively. Thus, hypothetically, there may be potential T_FH functional differences across species.

• Does Bcl6 instruct CD4⁺ T cells for memory development through its repressor activity? It is unclear how Bcl6 informs CD4⁺ T cell memory, and Bcl6 expression is reduced in memory T_FH cells. Given its interactions with Id2, Runx3, and Blimp1, Bcl6 can undoubtedly impact CD4⁺ T cell memory development.

High-resolution mouse Bcl6 ChIP-seq data are needed. High-resolution BCL6 ChIP-seqs from primary human T_FH and GC-T_FH cells are available [18,34], but the relatively low signal-to-noise ratio and minimal peaks mapped by unique NGS reads of Bcl6 ChIP-Seq makes it difficult to accurately define Bcl6 target genes [35]. Murine Bcl6 ChIP-seq of T_FH cells is particularly challenging because of the general requirement for large cell numbers to obtain robust and reproducible peaks. The advent of new techniques such as ChIP-exo 5.0, CUT&RUN, or CUT&TAG can allow for high-resolution ChIP-seq with a lower than usual amount of starting material [114–116].

A consistent in vitro mouse T_FH differentiation protocol is needed. Simple in vitro assay are powerful tools to facilitate research on molecular mechanisms. In contrast to the other helper T cell subsets, a consistent and convincing in vitro mouse T_FH differentiation assay (i.e., stable differentiation of CXCR5⁺Bcl6⁺PD-1⁺IL-21⁺ CD4⁺ T cells over multiple days) that works across multiple laboratories is not yet established. Individual studies have reported in vitro mouse T_FH differentiation with different combinations of cytokines and costimulatory molecules, but a protocol for the generation of mouse T_FH cells that works in multiple laboratories is still lacking [38,42,117–121]. In contrast to mouse, in vitro human T_FH differentiation is well established [101,122,123], suggesting differences in the biology of T_FH between humans and mice. While T_FH differentiation and function are substantially conserved between species, important differences in the biology of T_FH also exist between mice and humans. For example, IL-6 is an inducer of T_FH differentiation in mice, but the role of IL-6 in human T_FH cells is unclear [3]. Using gene-editing technology, such as CRISPR-Cas9 mediated gene deletion/editing, and traditional lentivirus-mediated gene delivery, the roles of TFs in T_FH differentiation can be tested in human CD4⁺ T cells. These might also provide future viewpoints regarding the function of T_FH or T_FH-like cells, with implications in vaccine design, as well as in various relevant diseases, including autoantibody-mediated autoimmune pathologies and cancers.

Highlights.

• B cell help, afforded by T cells, is an essential process of adaptive immunity.

• Transcription factor Bcl6, by serving as an apex of a repressor-of-repressors network, can provide further insights into T_FH regulation in mice and humans.

• Bcl6 is essential for T_FH differentiation and harbors various significant activities, including inhibition of Prdm1 expression.

• We posit that Bcl6 can control non-T_FH and T_FH genes by at least two modes-of-action: direct repression and repression-of-repressor mechanisms.

Glossary

Antigen-presenting cell (APC)

Specialized cells that can process antigens and display the processed peptide fragments on the cell surface and which are required for activating naïve T cells. Dendritic cells, macrophages, and B cells are primary APCs for naïve T cells

Germinal center (GC)

Specialized anatomic sites in lymphoid follicles for B-cell proliferation, differentiation, somatic hypermutation, and class switch responses during an adaptive immune response

Humoral immunity

Immunity mediated by proteins or other macromolecules circulating in extracellular fluids (e.g., blood), such as antibodies, complements, and antimicrobial peptides

Regulatory T cells (T_REG)

Specific CD4⁺ helper T cell subset that modulates or suppresses the immune response. T_REG cells are derived from thymic development and are also induced in the periphery from naïve CD4⁺ T cells. Foxp3 transcription factor is a signature molecule of T_REG. Induced T_REG cells can be induced by TGF-β. T_REG cells express a high amount of IL-2Rα and TGF-β, and IL-10

T-cell receptor (TCR)

Specialized protein complex and cell-surface receptor expressed on T lymphocytes; responsible for the recognition of antigen peptides presented on MHC molecules of APC. It consists of two different protein chains, α:β (95% in humans) or γ:δ chains (5% in humans). α:β or γ:δ chains complexes with the invariant CD3 and ζ proteins, which have a signaling function

T helper type 1 cells (T_H1)

Specific CD4⁺ helper T cell subset that induces cell-mediated responses against intracellular bacteria and protozoa; induced by IL-12 through STAT4 and T-bet transcription factors. T_H1 cells produce IFNγ and IL-2 as effector cytokines

T helper type 2 cells (T_H2)

Specific CD4⁺ helper T cell subset that induces humoral immune responses, generally against extracellular parasites, such as helminths; induced by IL-4 and IL-2 through STAT6 and GATA-3 transcription factors. T_H2 cells produce IL-4, IL-5, IL-9, IL-10, IL-13, and IL-25 as effector cytokines

T helper type 17 cells (T_H17)

Specific CD4⁺ helper T cell subset to induce immune responses generally against extracellular pathogens and fungi; induced by TGF-β and IL-6/IL-21 through STAT3 and RORγt transcription factors. T_H17 cells produce IL-17A, IL-17F, IL-21, and IL-22 as effector cytokines

Transcription factor (TF)

Protein that regulates transcription of DNA to mRNA through binding to a specific DNA sequence. TF function as activators, repressors, or both, by promoting or blocking the recruitment of RNA polymerases to genes. TF work alone or in complex with other proteins such as coactivators or corepressors