UCSF

Systems immunology characterizes how immune cell types interact within different contexts to lead to immunological outcomes. Several of the tools of systems immunology are drawn from the single-cell -omics world, fittingly since the first single-cell technology, flow cytometry, was and is primarily used in immunology to phenotype diverse immune cell types. Here, we present the use of single-cell transcriptomics and proteomics to take a systems immunology approach to study how immune set points, created by genetic and environmental exposures, affect individual immune responses in three major classes of disease, autoimmune (rheumatoid arthritis), infection (Zika virus), and cancer (metastatic prostate cancer). How pathogenic CD4 T cells develop to cause autoimmunity remains unknown. In the SKG autoimmune arthritis mouse model, we profiled arthritogenic naïve CD4 T cells by bulk and single cell RNA and T cell antigen receptor (TCR) sequencing prior to arthritis onset. Our analyses reveal that despite impaired proximal TCR signaling, a subset of SKG naïve CD4 T cells that have most recently encountered antigen more highly express gene programs associated with positive regulation of T cell activation and cytokine signaling compared to wild type cells. These cells also induce genes associated with negative regulation of T cell activation but for a subset of tolerogenic markers (e.g., Izumo1r, Tnfrsf9, Bach2, Eomes, Tigit, Tox, Tox2) do so at lesser amounts than wild type cells. Furthermore, their TCR sequences exhibit a previously unrecognized bias towards Vbs that recognize superantigen from endogenous retrovirus (ERV) mouse mammary tumor virus (MMTV). In arthritic joints, these biased Vbs are further expanded and ERVs are readily detected. Inhibition of viral reverse transcription significantly reduced SKG arthritis development. Together, our results suggest that endogenous viral products promote autoreactive naïve CD4 T cells which recognize endogenous viral superantigens to break tolerance via changes to their transcriptome and activation state. Although generating high neutralizing antibody levels is a key component of protective immunity after acute viral infection or vaccination, little is known about why some individuals generate high versus low neutralizing antibody titers. Here, we leverage the high-dimensional single-cell profiling capacity of mass cytometry to characterize the longitudinal cellular immune response to Zika virus (ZIKV) infection in viremic human blood donors in Puerto Rico. During acute ZIKV infection, we identify widely coordinated responses across innate and adaptive immune cell lineages. High frequencies of multiple activated cell types during acute infection are associated with high titers of ZIKV neutralizing antibodies 6 months post-infection, while stable immune features suggesting a cytotoxic-skewed immune set point are associated with low titers. Our study offers insight into the coordination of immune responses and identifies candidate cellular biomarkers that may offer predictive value in vaccine efficacy trials aimed at inducing high levels of antiviral neutralizing antibodies. Cancer immunotherapy has been a revolutionary anti-tumor treatment, but in prostate cancers, and other solid tumors, the response has been limited. Sipuleucel-T, an autologous antigen-presenting cell vaccine involving ex vivo peptide stimulation, is the only approved immunotherapy for advanced prostate cancer. Understanding the immunogenic and tolerogenic myeloid cell states in prostate cancer could improve the limited immunotherapy response. We used a genetic multiplexing strategy to simultaneously profile gene and protein expression on single cells from ~400,000 peripheral blood mononuclear cells (PBMCs) from longitudinal sampling of a metastatic castration-resistant prostate cancer (mCRPC) human cohort receiving combined immunotherapy (sipuleucel-T and ipilimumab). We identified co-expressed chronic interferon and complement gene signatures in the peripheral CD14+ myeloid compartment which predicted immunotherapy resistance. In contrast, the responders had higher frequency of progenitor exhausted CD8+ T cells (Tpex) suggesting a permissive myeloid environment requires a productive T cell response for anti-tumor activity. Future trials could use low co-expression of the CD14+ myeloid chronic interferon and complement gene signatures as a biomarker to select participants who are more likely to respond to therapy and could specifically target this myeloid state to potentially improve response to immunotherapy.