T cell/B cell collaboration in autoimmunity

Regulatory T cells prevent the emergence of autoantibodies and excessive IgE but the precise mechanisms are unclear. Here we show that BCL6-expressing Tregs, known as follicular regulatory T (Tfr) cells produce abundant amounts of a neuropeptide that targets B cells. Mice lacking Tfr cells or the Tfr-derived neuropeptide in Foxp3-expressing cells accumulated early plasma cells in germinal centers (GCs) and developed autoantibodies against histones and tissue-specific self-antigens. Upon immunization, these mice also produced increased serum IgE and IgG1. We show that the Tfr-derived neuropeptide is taken up by B cells, causes phosphorylation of numerous proteins and dampens IgE class switching. The Tfr-derived neuropeptide reduced differentiation of mouse and human GC B cells into plasma cells, downregulated BLIMP-1 and upregulated BCL6. Administration of the neuropeptide to Tfr-deficient mice prevented the accumulation of early plasma cells in GCs. Production of this neuropeptide by Tfr cells emerges as a central mechanism to suppress B cell-driven autoimmunity and IgE-mediated allergies. 

Antibody affinity maturation depends on germinal center (GC) positive selection of rare clones that acquire higher-affinity B cell receptors via somatic hypermutation, present more antigen to follicular helper T (Tfh) cells and, consequently, receive more contact-dependent T-cell help. As Tfh and GC B cells do not maintain long-lasting contacts in the chaotic GC environment, it is unclear how sufficient T-cell help is cumulatively focused onto those rare clones. Here we show that, upon receipt of CD40 stimulation, GC B cells upregulate chemokine CCL22 and to a lesser extent CCL17. By engaging CCR4 on Tfh cells, CCL22 and CCL17 can attract attention to a B cell by multiple helper cells from a distance, thus increasing the chance of productive help. In a GC response, B cells that acquire higher antigen-binding affinities express higher levels of CCL22, and higher levels of CCL22 expression highlight GC B cells of higher affinities. Acute increase or blockade of Tfh help rapidly increase or decrease CCL22 expression by GC B cells. Therefore, a chemokine-based intercellular reaction circuit translates the amount of T-cell help individual B cells recently receive to their subsequent ability to attract more help. When CCL22 and CCL17 are ablated in B cells, GCs form but do not efficiently mature in affinity. When competing with wildtype B cells in the same reaction, those lacking CCL22 and CCL17 are disadvantaged in acquiring T-cell help to maintain GC participation or develop into bone-marrow plasma cells. Therefore, a chemokine-mediated mechanism highlights affinity-improved B cells for Tfh cells to focuses attention from a distance. This ensures efficient positive selection despite the chaotic crowding of competitors.

Follicular helper T cells (Tfh) cells shape the germinal center (GC) response by delivery of contact-dependent signals and cytokines, including IFN-γ and IL-21. The contribution of these discrete extrinsic signals to the dysregulation of GC B cells in lupus is unclear. In lupus, by contrast to acute viral challenge, GCs persist as disease progresses, and transcriptional analysis of lupus-Tfh cells at different stages of disease revealed an evolving gene expression profile that corresponded to disease progression. However, Tfh cells from early and chronic GCs continued to robustly co-produce both IL-21 and IFN-γ with ongoing genesis of IgG2 autoantibody production, necessary for immune complex-derived glomerulonephritis. Additionally, CD11c- and Tbet-expressing age-associated B cells (ABCs) comprise a distinct B-cell subset that contributes to autoantibody production. Unlike Tfh and GC B cells, we show that ABCs accumulate as disease progresses in lupus-prone mice. However, their developmental reliance upon T cell help and relationship to GC B cells are unclear. To this end, we examined ABCs in a murine acute LCMV infection model, finding that Tfh, not T helper 1, cells drove their differentiation prior to and independent of the GC response. ABCs rarely overlap with GC B cells in phenotype, transcriptional profile, and clonality, with their development often diverging from that of GC B cells. These data indicate that ABCs in response to viral infection are predominantly generated independent of the GC response, yet their development is dependent upon help delivered by Tfh cells.

T cell- B cell interactions often take place within target tissues in chronic autoimmune conditions, yet T cell populations that drive these peripheral T-B cell interactions remain difficult to define. Rheumatoid arthritis synovium provides a valuable test case, where the synovium is often studded with T cell-B cell aggregates. We found that rheumatoid synovium contains a highly expanded population of PD-1hi CXCR5- CD4+ T cells that express high levels of IL-21 and CXCL13 and resemble Tfh cells cytometrically and transcriptomically. However, this PD-1hi CXCR5- cell population differed from Tfh cells in particular in migratory programs, expressing a distinct set of chemokine receptors including CCR2 and CCR5. We called these cells T peripheral helper (Tph) cells to capture the idea that they would provide help to B cells in inflamed peripheral tissues rather than within lymphoid follicles. We found that Tph cells are also highly expanded in the circulation of patients with lupus and that they depend on expression of MAF to produce IL-21 and stimulate B cell responses in vitro. Current ongoing work aims to identify Tph cells in murine models of lupus and to evaluate their contribution to pathologic B cell responses in vivo.   

Type 1 diabetes (T1D) is an autoimmune disease that typically starts in childhood and results in the destruction of insulin-producing beta cells in the pancreas. Although type 1 diabetes is primarily considered to be a T-cell mediated disease, B cells clearly participate in the autoimmune process. For example, autoantibodies recognizing pancreatic islet antigens commonly appear in circulation before the onset of the disease. Using blood samples from a longitudinal follow-up cohort, we have analyzed the frequency of circulating T follicular helper (Tfh) and T peripheral helper (Tph) cells in children at different stages of T1D development. We have observed that both activated circulating Tfh cells (defined as CD4+CXCR5+PD-1+ICOS+) as well as Tph cells (CD4+CXCR5-PD1hi) are increased in the blood of children with newly diagnosed T1D, especially in those individuals positive for multiple autoantibodies. These cell subsets were also shown to be increased in frequency in autoantibody-positive at-risk individuals who later progressed to T1D. Our results suggest that an increase in circulating Tfh and Tph cells is associated with T1D onset and they could therefore potentially be used as biomarkers of disease progression. Moreover, our results lend support to the idea that Tfh/Tph activation pathways can be potential targets for the immunotherapy of T1D.

The lung is one of the largest barrier organs and not only involved in pulmonary infections but also different chronic inflammatory interstitial lung diseases. T and B cells are frequently found in inflammatory infiltrates. We recently showed in a mouse lung inflammation model that active T cell / B cell cooperation can take place directly in the lung tissue. A special population of T follicular helper- (Tfh-) like cells drives the differentiation of lung-infiltrating B cells into antibody-producing cells. Tfh-like cells (also known as peripheral T helper cells (Tph)) lack the classical Tfh markers CXCR5 and Bcl-6 but provide potent B cell help by production of IL-21 and high expression of CD40L. T/B cooperation in the lung takes places in unstructured, non-ectopic peribronchial infiltrates but nevertheless results in efficient B cell receptor hypermutation. 

We now identified a similar population of Tfh-like cells in inflamed lungs from sarcoidosis patients. This systemic inflammatory disease was previously thought to be a primarily Th1 and macrophage-driven disease. However, lung-infiltrating T cells also produce large amounts of IL-21 and provide potent B cell help in vitro. Large clusters of T and B cells can be found in the lung tissue and these infiltrates are typically surrounded by IgA-producing plasma blasts, indicating their local generation in the lung. These findings highlight the role of T cell / B cell cooperation as a novel pathomechanism for sarcoidosis.

While the spectrum of coronavirus disease 2019 (COVID-19) ranges from asymptomatic infection to severe and sometimes fatal, it remains unknown what determines divergent clinical trajectories among patients. COVID-19 patients quickly develop antibodies to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), but these antibodies may be protective or pathogenic. Protective antibodies neutralize virus or contribute to viral clearance by activating Fc receptor-expressing cells, while pathogenic antibodies can contribute to hyperinflammatory responses or even enhance infection by SARS-CoV-2. Studies of antibody responses in COVID-19 patients demonstrate that severe disease is associated with impaired T follicular helper (Tfh) cell and germinal center responses, suggesting that Tfh-independent pathways of antibody induction could contribute to COVID-19 severity. Using a murine model of SARS-CoV-2 infection, we defined the features of Tfh cell-dependent and Tfh cell-independent antibody responses. Tfh cell help was required for long-lived antibody responses to SARS-CoV-2. Tfh cell-dependent antibodies also demonstrated increased recognition of spike S2 fusion peptide-derived epitopes. Furthermore, antibodies that developed in the absence of Tfh cell help exhibited altered Fc isotype profiles. Ultimately, we aim to identify the cellular pathways and antibody features that mediate SARS-CoV-2-induced disease severity as well as define a new paradigm in which different CD4+ T cell subsets guide the production of protective versus pathogenic antibodies to viruses.