Can follicular helper T cells be targeted to improve vaccine efficacy?

Introduction

Vaccination is one of the most successful, cost-effective interventions for combating infectious disease, thereby reducing infection-related disease, disability and death worldwide¹. Despite this enormous success, there are still multiple infections that require a vaccination solution, including vaccines that protect against HIV and malaria^2,3, and a way to improve vaccine efficacy in older persons⁴. The majority of current vaccines have been developed empirically rather than rationally, suggesting that a change in approach to vaccine development may enable breakthroughs in vaccine design⁵. All routine human vaccinations, with the exception of the Bacillus Calmette-Guérin tuberculosis vaccine, provide protection by generating antibodies that block the ability of a pathogen to establish an infection and that target it for destruction. Vaccine-induced antibody responses are supported by T follicular helper (Tfh) cells; here, we discuss how advances in the knowledge of Tfh cell biology could be used to improve vaccine efficacy.

The production of vaccine-specific antibodies can occur via two cellular routes: the extrafollicular or germinal centre (GC) responses. The extrafollicular response produces an initial burst of antibodies early after immunisation and can occur with or without T cell help⁶. These extrafollicular plasma cells are short-lived and because of this are not able to provide a long-term source of protective antibodies⁷. The GC is a specialised microenvironment that forms in secondary lymphoid tissues after immunisation when antigen-activated B cells migrate to the B cell follicle, begin to proliferate, and undergo somatic hypermutation (SHM) of their immunoglobulin genes⁸. The cellular products of the GC are long-lived plasma and memory B cells that can provide protection for decades after initial exposure⁹. Because of the longevity of these cells, the GC represents an exciting target to improve vaccine responses in situations in which there is an unmet clinical need.

The GC is a microenvironment of intense cellular collaboration: GC B cells, Tfh cells, T follicular regulatory (Tfr) cells, tingible body macrophages and follicular dendritic cells (FDCs) act together to generate a robust response (Figure 1). The “multi-player” nature of the GC means that there are a number of cellular targets that can be manipulated in the GC during vaccination in an attempt to modulate its output. Here, we will discuss whether targeting Tfh and Tfr cells may be a successful strategy for improving the GC in response to vaccination.

Figure 1. Pathways to antibody production.

The germinal centre (GC) is initiated when activated B cells migrate into the B cell follicle, begin to divide, and mutate their B-cell receptor (BCR) genes. These proliferating centroblasts then exit the cell cycle, becoming centrocytes that use their mutated BCR to collect antigen from the surface of follicular dendritic cells (FDCs) and present processed peptide on major histocompatibility complex class II to T follicular helper (Tfh) cells. Tfh and T follicular regulatory (Tfr) cell differentiation initiates during T cell priming when Bcl-6, the transcription factor required for Tfh and Tfr cell differentiation, is upregulated. Subsequent interactions with B cells stabilise Bcl-6 expression in pre-Tfh and pre-Tfr cells, allowing them to migrate into the GC where Tfh cells provide help to centrocytes and Tfr cells act as GC suppressors. As an alternative to entering the GC, antigen-activated B cells can also receive T cell help and differentiate into short-lived extrafollicular plasma cells that produce the first wave of protective antibodies. DC, dendritic cell; FDC, follicular dendritic cell; mB, memory B cell; SHM, somatic hypermutation; TBM, tingible body macrophage; Th, T helper cell; Treg, regulatory T cell.

Follicular T cells and the GC response

In physiological conditions, the GC is absolutely dependent on T cell help for its formation. Experimentally, immunisation with superantigens can initiate GC formation in the absence of T cells, however, the response collapses after five days, demonstrating the absolute requirement of T cells for the GC to produce plasma cells¹⁰. The specific requirement for T cell help is fulfilled by a specialised subset of CD4⁺ helper T cells: Tfh cells. Tfh cells differentiate in response to immunisation by using a specific differentiation pathway that requires various cytokine signals and multiple rounds of antigen presentation that culminates in the expression of the transcriptional repressor Bcl-6 and localisation to the GC¹¹. The signals that are required for Tfh cell development have been well characterised in both humans and mice; these are summarised in Table 1^12–42. These pathways may represent viable targets to manipulate the number of Tfh cells that form after vaccination.

After establishment of the GC, it divides into two distinct zones: the GC B cells rapidly divide in the dark zone and undergo SHM of their B-cell receptor (BCR) genes and then exit the cell cycle and migrate to the light zone of the GC, where the fully differentiated Tfh cells are located. The random nature of SHM requires that mutated B cells undergo a selection process before they exit the GC as plasma cells or memory B cells. The GC B cells test their newly mutated BCR by collecting antigen-containing immune complexes from the surface of FDCs and then present processed antigen to Tfh cells on major histocompatibility complex (MHC) class II⁴³. The B cells that are able to present the most antigen to Tfh cells are the recipients of T cell help⁴⁴, which results in the upregulation of c-myc and subsequent return to the dark zone to undergo further rounds of proliferation and mutation^45–47. By this mechanism, Tfh cells act as a limiting factor in the selection of high-affinity GC B cells. This could simply be numerical, as B cells outnumber Tfh cells in the GC. Alternatively, there may be an interaction threshold that needs to be overcome before Tfh and B cells form productive conjugates in the GC. During Tfh and B cell interactions in the GC, there is a bidirectional exchange of signals: Tfh cells provide help in the form of CD40L, interleukin-21 (IL-21) and IL-4 to GC B cells, which supports proliferation and survival, while B cells provide inducible T cell co-stimulator ligand (ICOSL) to Tfh cells^48,49. Thus, Tfh cells facilitate the preferential expansion and mutation of high-affinity GC B cell clones and are key regulators of the size and quality of the GC response.

In addition to Tfh cells, there is a second specialised subset of CD4⁺ T cells present in the GC: Tfr cells. Tfr cells derive from Foxp3⁺ regulatory T (Treg) cells and act as suppressors of the GC response. Tfr cells limit the size of the GC response, thereby acting as a counterbalance to Tfh cells^50–52. Although there are some transcriptional and phenotypic similarities between Tfh and Tfr cells, functionally Tfr cells are suppressive and thus resemble conventional Treg cells⁵³.

Together, Tfh and Tfr cells are key regulators of the GC response; Tfh cells positively control the size and output of GC, whereas Tfr cells act as negative regulators of the response. This suggests that strategies to enhance Tfh number or function (or both) or reduce Tfr cells may enhance GC responses and promote a more potent response to vaccination.

Circulating peripheral blood Tfh as biomarkers of GC Tfh cells

The majority of advances in Tfh and Tfr cell biology have occurred through studying secondary lymphoid tissues in mice. This is largely because of the impressive range of tools (e.g., genetically modified mice) that allow precise dissection of GC biology in mice and because access to healthy human lymphoid tissue can be difficult, particularly for the purpose of studying GC responses to a defined antigenic stimulus. In an attempt to circumvent this issue, a circulating cellular biomarker of GC Tfh cells has been used to further investigate Tfh cell biology in humans. These cells have been coined circulating Tfh-like (cTfh) cells and were first identified in sanroque mice and in patients with systemic lupus erythematous and were defined by CXCR5, programmed cell death protein 1 (PD-1), and ICOS expression⁵⁴. Subsequently, it has been shown that an increased frequency of cTfh cells coincides with the peak GC response in mice and the plasmablast response to influenza vaccination in humans^55,56. These studies suggest that cTfh cells may be a key tool for studying the role of Tfh cells in human vaccine responses. However, the use of cTfh cells as a surrogate of GC Tfh cell responses in humans requires a robust assessment of the strengths and limitations of this approach.

Studies in both humans and mice support a link between the GC Tfh and cTfh cells. Human cTfh cells can provide help to B cells in vitro and upon stimulation display several features consistent with GC Tfh cells, including ICOS expression and expression of IL-21 and CXCL13^57–59. Although cTfh cells do not express BCL6, they have low levels of BLIMP1 and express cMAF, and this indicates that they share features of transcriptional control with GC Tfh cells^57–59. Several human immunodeficiency syndromes that are associated with severely impaired GC responses due to loss of functional CD40L⁶⁰, ICOS^15,61, STAT3⁶² or IL-12βR1³⁶ display corresponding reductions in blood cTfh cells, suggesting that cTfh cells can be a biomarker for an active GC response. Conversely, mice deficient for Sh2d1a have impaired GC reactions but unchanged cTfh frequencies⁵⁵. Consistent with this, patients with X-linked lymphoproliferative disease (XLP) caused by defects in SH2D1A, or healthy XLP carriers, did not display alterations in cTfh frequencies⁵⁵. These data suggest that although cTfh cells resemble GC Tfh cells, a GC reaction is not required for cTfh cell development, which parallels the development of extrafollicular Tfh cells⁶³. One possibility is that cTfh cells are memory cells that are induced upon vaccination to enable fast GC Tfh responses following subsequent infection. Consistent with this idea, tetanus- and smallpox-specific cTfh cells can be identified in humans years after vaccination^58,64. In mice, cTfh cells have the capacity to become GC Tfh cells and support the GC response^55,65, suggesting that cTfh cells may be an important component of secondary immune responses and therefore a biologically relevant cell population in successful vaccination. Despite the recent surge in correlative studies assessing cTfh cells in a multitude of disease settings, unsupervised comparisons of gene expression in GC Tfh cells have not been performed in blood and lymphoid tissue samples from the same individual⁵⁸, and antigen-specific responses have not been determined. Addressing these issues will help to clarify the relationship between circulating and GC Tfh cells.

An interesting feature of GC Tfh cells is their well-described heterogeneity⁶⁶, and cTfh cells are not an exception. Analysis of blood CD4⁺CXCR5⁺ cells for expression of PD1, CCR7, CXCR3, CCR6 and ICOS has been proposed to define nine populations of cTfh cells⁶⁷. However, across the range of studies, robust B cell helper function in vitro has consistently been demonstrated for CD4⁺CXCR5⁺ cells that express high levels of PD-1 or ICOS or both⁶⁷. CXCR3 and CCR6 expression on cTfh enables identification of cTfh cells with Th1-like (cTfh1, CXCR3⁺CCR6⁻), Th2-like (cTfh2, CXCR3⁻CCR6⁻) and Th17-like (cTfh17, CXCR3⁻CCR6⁺) properties, including the expression of transcription factors and cytokines that define these T helper subsets⁵⁷. cTfh2 and cTfh17 can support naïve and memory B cells to produce antibodies in vitro, whereas cTfh1 cells have limited in vitro helper function^57,58, although following influenza vaccination a population of ICOS⁺ cTfh1 cells were able to help memory B cells make antibodies⁵⁶. One limitation of these studies is that it remains unclear to what extent in vitro B cell helper function reflects effective GC Tfh help in vivo. Although these cTfh cell subtypes have been identified in blood, characterisation of GC Tfh cell populations by using these markers has been limited, calling into question the relevance of these subsets to GC biology. However, tonsillar Tfh can co-express BCL6 and RORγt⁶⁷ and a proportion of human lymph node Tfh cells express CXCR3 (D.L. Hill, unpublished), and this suggests that comparable heterogeneity exists within in the GC Tfh cell population. But whether there is a specialised role for Th1/Th2/Th17 polarised GC Tfh cells in the GC has yet to be elucidated.

The polarisation of GC Tfh cells depends on the stimuli provided during differentiation. In mice, Th2-biased infections produce IL-4-secreting GC Tfh cells, whereas Th1-biased infections support interferon-gamma-positive (IFNγ⁺) GC Tfh cells^68–71. In humans, cTfh2 cell frequency increases in people with Th2-polarised Schistosoma japonicum infection⁷², whereas cTfh1 cells are preferentially expanded during Th1-biased acute Plasmodium falciparum infection and after seasonal influenza vaccination^56,73. Thus, different cytokine environments induced by specific infections or immunisations appear to drive Tfh cell polarisation and may enable Tfh cells to appropriately support B cell production of the antibody isotype required to clear the infection. For example, in mice, IFNγ⁺ Tfh cells could be found in conjugates with Ig2a⁺ B cells, whereas IL-4⁺ Tfh cells were more likely to be paired with IgG1⁺ B cells⁷⁴. Immunity against pathogens relies upon production of specific antibody isotypes that ultimately play an important role in clearing infections. For example, inappropriate production of Th1-supported isotypes to the parasitic roundworm Wuchereria bancrofti⁷⁵ and Th2-supported isotypes in P. falciparum malaria⁷⁶ correlates with poor disease outcomes. Therefore, cTfh cell heterogeneity may reflect the ability of Tfh cells to be shaped by the environmental signals present during differentiation, which enables them to guide an appropriate B cell response to infection or vaccination, to facilitate pathogen clearance.

It has been proposed that the limited efficacy of seasonal influenza vaccination results from the preferential induction of cTfh1 cells⁵⁸. As such, skewing Tfh cells away from Tfh1-like and toward Tfh2/17-like may represent a potential target to enhance antibody titres following influenza vaccination (Figure 2A). Interestingly, blocking the Th1 cytokines IL-2 and tumour necrosis factor (TNF) improved Tfh-mediated B cell help in vitro^77,78. However, this approach may not be effective for generating protective responses to vaccination in vivo. Passive transfer of broadly neutralising antibodies to hemagglutinin can protect mice from succumbing to experimental influenza infection. Importantly, for some clones, this protection is conferred only by a Th1 polarised IgG2a antibody and not Th2 polarised IgG1, despite having the same ability to bind hemagglutinin⁷⁹. This suggests that production of Th1-supported isotypes and the selective induction of Th1-like Tfh cells are likely important for generating protective influenza vaccine responses. Yet as current influenza vaccine formulations fail to generate a protective immune response in up to 30% of vaccine recipients⁸⁰, further enhancing Tfh cell responses may improve vaccine efficacy.

Figure 2. Strategies for manipulating follicular T cells to enhance the output of the germinal centre (GC) response.

(A) Altering the balance of different subsets of T follicular helper (Tfh) cells toward Tfh2 and Tfh17 cells may enhance the B cell helper capacity of the Tfh cells within the GC and the production of the necessary antibody isotypes. (B) Increasing the number of Tfh cells may help increase the size and output of the GC response. (C) Reducing the frequency of suppressive T follicular regulatory (Tfr) cells may enhance the output of the GC response. FDC, follicular dendritic cell; IFNγ, interferon-gamma; IL-17A, interleukin-17A; mB, memory B cell.

The potential use of cTfh cell subsets as a biomarker of vaccine efficacy is an attractive possibility that would be easily amenable to clinical trials. Although further study is needed, mounting evidence supports cTfh cells as a relevant population for the study of vaccine responses in humans. However, it is clear that there are limitations for using cTfh cells to study the GC Tfh cell response, as although they appear to be a biomarker for the GC response, they differ from bona fide GC Tfh cells in a number of aspects, including BCL6 expression and dependence on SAP for their differentiation. Because of this, analysing cTfh cells will likely not become a substitute for studies aimed at understanding GC Tfh cell biology, but rather represents an additional tool to interrogate the human response to vaccination. Research that combines assessment of cTfh cells after vaccination, GC Tfh cells from human secondary lymphoid tissues and mouse vaccination models will enable thorough interrogation of strategies that target Tfh cells to improve vaccine efficacy.

Potential avenues for modifying follicular T cells to enhance vaccination efficacy

Within the GC, Tfh cells support the GC response and Tfr cells negatively regulate the magnitude of the GC. As such, manipulating the frequency of Tfh cells or enhancing their function may improve the GC response. Conversely, reducing the number of Tfr cells or their functional capacity may also increase antibody responses to vaccination. Here, we explore the potential strategies for manipulating these T cell subsets with the view to increase GC output following vaccination.

Potential dangers of enhancing Tfh responses

Although it is clear that Tfh cells are essential for a productive response to vaccination, they have also been implicated in a number of autoimmune conditions as key drivers of disease. Sanroque mice have a point mutation in the Roquin1 gene (Rc3h1) that causes a lupus-like phenotype that is driven by Tfh cells that support GCs in the absence of exogenous antigen^116,117. As B cells can acquire self-reactivity during somatic hyper-mutation, findings from the sanroque mice suggest that increases in Tfh cell number may lead to a break in GC tolerance, enabling self-reactive B cells to be selected in the GC. Correspondingly, several autoimmune conditions have been associated with an increase in cTfh frequencies^54,118–120 or have a skewed cTfh population away from cTfh1 and toward cTfh2 or cTfh17 or both^{57,118,121,122}. This is an important consideration in the context of enhancing Tfh responses to vaccination, particularly in older persons, as the occurrence of autoantibodies increases with age¹²³, suggesting that the B cell pool may contain a higher frequency of autoreactive B cells able to enter the GC. Taken together, these studies demonstrate a potential hazard of increasing Tfh cell frequencies or function in vaccination and indicate that autoantibody production would need to be accounted for in future study design as a possible outcome.

Concluding remarks

The GC is critical for the production of long-lived antibody-secreting plasma cells after vaccination, making it a promising cellular response to improve vaccine efficacy. There are many players in the GC response; Tfh and Tfr cells tightly control its size and output and thus make them key targets to manipulate in vaccine design. Altering vaccines in a way that increases Tfh cell formation or function (or both) or reduces the suppression exerted on the GC by Tfr cells may be a rational strategy to improve vaccine responses. As vaccines need to have an extremely high safety profile, any perturbations to vaccines must be very low-risk. Because of this, the most logical way to manipulate follicular T cells is to use antigen doses or adjuvants that favour differentiation of Tfh cells with excellent B cell helper capacity, and suppress Tfr cell development. Recent research into next-generation adjuvants demonstrates that adjuvants that support enhanced antibody production also associate with increased numbers of Tfh cell in experimental animals. To determine whether this can translate into enhanced Tfh responses in humans, cTfh cells will be a useful biomarker of GC Tfh responses in preliminary clinical trials. Further research in both humans and animal models into precisely how to manipulate Tfh and Tfr cells to improve vaccine responses may enable us to address some of our current unmet clinical requirements for improved vaccines.