TIGIT-positive circulating follicular helper T cells and sickle cell alloimmunization

Transfusion therapy remains an important treatment modality for patients with sickle cell disease (SCD). Despite its therapeutic benefits, red blood cell transfusion results in the development of antibodies against disparate antigens on transfused red cells (alloimmunization) in approximately 20–50% of patients with SCD. Alloantibodies can cause delayed hemolytic transfusion reactions, which in SCD patients can trigger hyperhemolysis, a life-threatening poorly understood phenomenon in which the transfused and the patient’s own red blood cells are destroyed.1 In addition, finding compatible units for patients with alloantibodies can be difficult, causing potentially life-threatening transfusion delays. Matching for key immunogenic antigens disparate between patients (of African ancestry) and blood donors (mostly white), namely Rh-C, -E and K, has been a strategy used to reduce alloimmunization in patients with SCD; however, patients transfused with C,E,K-matched donor RBC (from mostly black donors), can still develop pathogenic Rh antibodies due to the genetic diversity of the RH locus in individuals of African ancestry.2,3 This highlights the need for better characterization of alloimmunization triggers and risk factors/biomarkers in this highly vulnerable population.

The etiology of alloimmunization in patients with SCD remains unclear. Our group has recently studied antigen-specific CD4 helper T-cell responses in peripheral blood of transfused patients with SCD.4 We found that in alloantibody responder patients, antigen-specific CD4 T-cell responses were predominantly T_H17, and that interleukin (IL)-21, the prototypic follicular helper T-cell (T_FH)-associated cytokine driving B-cell help, was only expressed by antigen-specific T_FH derived from responders.4 These data are consistent with the role of heightened CD4 T-cell effector functions driving alloantibody production in alloimmunized SCD patients.

T_FH cells have emerged as the main effector CD4 T cells specialized in providing help to support responses of antigen-engaged B cells to generate the initial wave of antibody response as well as in supporting B-cell differentiation into high affinity antibody-producing B cells and long-lasting IgG antibodies.5 T_FH cells are universally characterized by: (i) the expression of the chemokine (C-X-C motif) receptor 5 (CXCR5) on memory (CD45RA) cells which directs them to the B-cell follicles in response to the specific ligand CXCL13;6–8 (ii) expression of IL-21, a potent cytokine promoting growth, differentiation, and class switching of B cells although they also secrete other B-cell help cytokines including IL-4;9,10 and (iii) co-stimulatory molecules specialized in providing B-cell help including inducible T-cell co-stimulatory (ICOS) and CD40L.5,11 Given the key role of T_FH cells in providing help to B cells to support their activation, expansion and differentiation, these cells have attracted particular attention as potential targets for developing therapeutic strategies not only to enhance vaccination responses but also to prevent antibody-mediated autoimmunity.5 Our studies of antigen-specific T_FH responses in alloimmunized SCD patients support the notion that T_FH cells are also likely to be critical in alloimmunization biology.4 More recently, T_FH-related cells with ability to promote antibody responses were identified in the peripheral circulation of humans and mice,12–19 opening up the possibility of studying T_FH cell responses in blood samples.

Figure 1.Hypothesized model of TIGIT⁺ circulating follicular helper T-cell differentiation following chronic stimulation. T_H1 or T_FH-type 1 cells express high levels of type 1 cytokines such as interferon-gamma but no or low levels of IL-21. Upon chronic stimulation such as following regular transfusions, these T cells become exhausted, losing the ability to produce interferon-gamma. At the same time, immune checkpoint molecules, including TIGIT and PD-1, are up-regulated and the cells express the prototypic B-cell help cytokine IL-21, which subsequently drives antibody production by B cells.

In the report from Yazdanbakhsh’s group published in this issue of the journal,20 Godefroy et al. examine the role of two immune checkpoints21 expressed by CD4/CD45RA/CXCR5 circulating follicular helper T (cT_FH) cells: the programmed cell death-1 (PD-1) and the T-cell immunoreceptor with Ig and immunoreceptor tyrosine-based inhibitory domains (TIGIT) surface molecules. The function of PD-1 on T_FH cells remains controversial and has been described as both promoting15,16 and reducing17,22,23 humoral responses in different models. The recent discovery of regulatory follicular T cells (T_FR) expressing high levels of PD-117 may help to reconcile these divergent findings. Although TIGIT expression on T cells has been studied previously,24–29 the function of TIGIT on T_FH cells is being addressed for the first time in the study by Godefroy et al.20

The authors show that TIGIT-expressing cells represent almost half of the cT_FH population, while PD-1 cells, all co-expressing TIGIT, are significantly less frequent, accounting for about 9% of the cT_FH cells.20 Several subsets of cT_FH have previously been characterized. Three main subsets, named T_FH1, T_FH2 and T_FH17, can be identified not only by their ability to produce type 1, 2 and 17 cytokines or express their respective transcription factors (T-bet, GATA3, RORc), but also by their expression of chemokine receptors CXCR3 and CCR6.13 T_FH1 cells express CXCR3, but not CCR6; T_FH2 cells express neither; T_FH17 cells express CCR6, but not CXCR3.13 Based on chemokine receptor expression, the authors show that TIGIT cT_FH represent a subset phenotypically different from any of the three populations previously described.20 Interestingly, TIGIT cT_FH can produce IL-4 and to some extent interferon-gamma. Accordingly, both T_FH1 and T_FH2 can be found in TIGIT cT_FH independently of the cells’ PD-1 expression.20 IL-17 secretion was not tested. Nonetheless, these findings establish that TIGIT cells, whether or not they express PD-1, represent a novel subset of cT_FH.

Importantly, TIGIT cT_FH cells exerted strong B-cell help functions.20 Compared to TIGIT cT_FH, they were able to better promote differentiation of B cells into plasmablasts and IgG-producing cells.20 Furthermore, the study of sorted cT_FH subsets showed that TIGIT/PD-1 cells provide similar help to B cells as TIGIT/PD-1 cells,20 suggesting that PD-1 plays a limited role in promoting humoral responses in this model. Despite the fact that it still needs to be proven, TIGIT cT_FH cells heightened help to B cells probably lie in their enhanced expression of CD40L and ICOS co-stimulation molecules and/or their capacity to produce large amounts of IL-21. Alternatively, B-cell help could be provided directly by signaling through TIGIT’s ligand(s). Whether TIGIT cT_FH cells influence isotype switching of Ig production, as shown for T_FH2 and T_FH17, remains to be determined.

Using blocking antibodies, Godefroy et al. also demonstrated that whereas PD-1 did not seem to play a major role in this system, TIGIT was directly implicated in the function of these cells.20 Indeed, blocking TIGIT not only decreased the expression levels of ICOS, CD40L and IL-21, but also prevented B cells from producing IgG.10 Other authors proposed a direct role for PD-1 in T_FH function.15–17,22,23 In the study by Godefroy et al., the authors did not observe a significant effect after blocking PD-1.20 This could be due to the use of different experimental systems or a mixture of opposite effects mediated by PD-1 cancelling each other. Instead, the fact that all PD-1 cT_FH co-express TIGIT20 could suggest that PD-1 represents a non-functional marker on cT_FH and that TIGIT/PD-1 cT_FH B-cell help functions are due, at least in part, to TIGIT.

The study suggests a role for TIGIT-expressing cT_FH cells in alloimmunization in the context of SCD.20 In alloimmunized patients, TIGIT cT_FH cells offered more potent help to B cells than in non-alloimmunized patients: IL-21 and CD40L expression was higher in alloimmunized patients than in non-alloimmunized patients.20 This could be explained either by faulty TIGIT signaling in non-alloimmunized patients or by an exacerbated response to TIGIT triggering in alloimmunized patients. Collectively, these data suggest involvement of various subsets of cT_FH cells displaying different functions in the alloimmunization process of SCD patients.

Chronic antigen stimulation can result in T-cell exhaustion associated with lack of cytotoxic activity and failure to produce type 1 cytokines, interferon-gamma and IL-2 as well as up-regulation of several immune checkpoint molecules including TIGIT on the exhausted T cells.21 Godefroy et al. propose an interesting model in which repeated exposure to allogeneic red blood cells in patients with SCD could lead to exhaustion of red blood cell-specific T cells, which would end up triggering the expression of immune checkpoints such as TIGIT. This brings into question whether exhausted type 1 CD4 T cells could be skewed towards a T_FH phenotype by expressing TIGIT (or other “exhaustion” markers) after repeated stimulation caused by chronic transfusions. Considering the latter question from a non-transfusion perspective, the authors propose an interesting hypothesis: if chronic stimulation induces T_H1 or T_FH1 cells to differentiate into T_FH cells, could this process reflect an evolutionary mechanism by which TIGIT would be involved in reducing type 1 chronic inflammation as well as shifting immune responses from cellular to potentially more “suitable”, humoral responses?

Overall, these studies provide new insight into T_FH polarization, and help give rise to the development of therapeutic strategies, not only regarding alloimmunization, but also in vaccination as well as in a large array of disorders, including inflammatory diseases, autoimmunity and immune deficiencies.

Footnotes

• Financial and other disclosures provided by the author using the ICMJE (www.icmje.org) Uniform Format for Disclosure of Competing Interests are available with the full text of this paper at www.haematologica.org.