During unresolved infections, some viruses escape immunological control and establish a persistant reservoir in certain cell types, such as human immunodeficiency virus (HIV), which persists in follicular helper T cells (TFH cells), and Epstein-Barr virus (EBV), which persists in B cells. Here we identified a specialized group of cytotoxic T cells (TC cells) that expressed the chemokine receptor CXCR5, selectively entered B cell follicles and eradicated infected TFH cells and B cells. The differentiation of these cells, which we have called 'follicular cytotoxic T cells' (TFC cells), required the transcription factors Bcl6, E2A and TCF-1 but was inhibited by the transcriptional regulators Blimp1, Id2 and Id3. Blimp1 and E2A directly regulated Cxcr5 expression and, together with Bcl6 and TCF-1, formed a transcriptional circuit that guided TFC cell development. The identification of TFC cells has far-reaching implications for the development of strategies to control infections that target B cells and TFH cells and to treat B cell-derived malignancies.

These findings demonstrate that C. concisus and other non-C. jejuni Campylobacter species may play a role in initiating gastrointestinal diseases.

Objectives To compare the potency, toxicity and mechanism of action of multiple histone deacetylase inhibitors (HDACi) in activating HIV production from latency. Design In-vitro analysis of HDACi in a primary T-cell model of HIV latency and latently infected cell lines. Methods Latently infected chemokine ligand 19 (CCL19)-treated CD4+ T cells and the latently infected cell lines ACH2 and J-Lat were treated with a panel of HDACi, including entinostat, vorinostat, panonbinostat and MCT3. Viral production and cell viability were compared. Expression of cellular HDACs was measured by western blot and PCR. Association of HDACs with the HIV long-terminal repeat (LTR) using latently infected CCL19-treated primary CD4+ T cells in the presence and absence of specific HDACi was determined by chromatin immunoprecipitation (ChIP). Results We demonstrated considerable variation in the potency and toxicity of HDACi in latently infected primary CD4+ T cells and cell lines. All HDACi tested activated HIV production in latently infected primary T cells with greatest potency demonstrated with entinostat and vorinostat and greatest toxicity with panobinostat. Following the addition of HDACi in vitro, there were no changes in markers of T-cell activation or expression of the HIV coreceptors chemokine (C-X-C motif) receptor 4 (CXCR4) or chemokine (C-C motif) receptor type 5 (CCR5). ChIP analysis of latently infected CCL19-treated primary CD4+ T cells showed binding by HDAC1, HDAC2 and HDAC3 to the LTR with removal of HDAC1 and HDAC2 following treatment with the HDACi vorinostat and HDAC1 only following treatment with entinostat. Conclusion The HDACi entinostat, selective for inhibition of class I HDACs, induced virus expression in latently infected primary CD4+ T cells making this compound an attractive novel option for future clinical trials.