Supplementary MaterialsSupp Table 8. and nominating testable therapeutic targets3. Its application to longitudinal samples could afford an opportunity to uncover cellular factors associated with the evolution of disease progression without potentially confounding inter-individual variability4. Here, we present an experimental and computational methodology that uses scRNA-Seq to characterize dynamic cellular programs and their molecular drivers, and apply it to HIV contamination. By performing scRNA-seq on blood from four untreated individuals prior SBI-115 to and longitudinally during acute contamination5, we are powered within each to discover gene response modules that vary by time and cell subset. Beyond previously-unappreciated individual- and cell-type-specific interferon stimulated gene (ISG) upregulation, we describe temporally-aligned gene expression responses obscured in bulk analyses, including those involved in pro-inflammatory T cell differentiation, prolonged monocyte MHC-II upregulation, and persistent NK cytolytic killing. We further identify response features arising in the first weeks of infectione.g. proliferating NK cellswhich, potentially, may associate with future viral control. Overall, our approach provides a unified framework for characterizing multiple dynamic cellular responses and their coordination. Despite advances in pre-exposure prophylaxis, there were 1.7 million new cases of HIV contamination in 20186, highlighting the need for effective HIV vaccines. A better understanding of key immune responses during the earliest stages of infectionespecially Fiebig Stage I & II, prior to and at peak viral loadcould help identify future prophylactic and therapeutic targets7. Using historical samples, collected before standard-of-care included treatment during acute contamination, from the Females Rising through Education, Support and Health (FRESH) study5, we assayed evolving immune responses during hyper-acute (1C2 weeks post-detection) SBI-115 and acute (3 weeks – 6 months) HIV contamination. We performed Seq-Well-based massively-parallel scRNA-Seq on peripheral blood mononuclear cells (PBMCs) from four FRESH participants who became infected with HIV during study. We analyzed multiple timepoints from pre-infection through one year following viral detection (Fig. 1a; Supplementary Table 1; Methods) over which all four demonstrated a rapid rise in plasma viremia and drop in CD4+ T cell counts8 (Fig. 1b; Extended Data Fig. 1a). Altogether, we captured 59,162 cells after performing quality controls, with an average of 1,976 cells per participant per timepoint (Extended Data Fig. 1b; Supplementary Table 2). Open in a separate window Physique 1: Longitudinal profiling of peripheral immune cells in hyper-acute and acute HIV-infection by single-cell RNA-Sequencing.(a) Depiction of the typical trajectory of HIV viral load in the plasma during hyper-acute and acute HIV infection adapted from Fiebig et al.8, and the timepoints sampled in this study. Since participants are tested twice weekly, there is an uncertainty of up to 3 days in where around the viral load curve the first detectable viremia occurs (error bar is usually representative). The exact days sampled are available in Supplementary Table 1. (b) Viral load and CD4+ T cell count for the four participants assayed in this study. Dotted lines indicate a missing data point for the metric. (c) tSNE analysis of PBMCs from all participants and timepoints sampled (n=59,162). Cells are annotated based on differential Rabbit Polyclonal to IRF-3 (phospho-Ser386) expression analysis on orthogonally discovered clusters. (d) tSNE in c annotated by timepoint (left) and participant (right). (e) Scatter plot depicting the correlation between cell frequencies of CD4+ and CD8+ T cells measured by Seq-Well (n = 2 array replicates) and FACS (n = 1 flow replicate). R-squared values reflect variance described by an F-test for linear regression. To SBI-115 assign cellular identity, we analyzed the combined data from all participants and timepoints (Methods). These analyses yielded few participant-specific features, suggesting disease biology, rather than technical artifact, is the main driver of variation (Fig. 1d; Extended Data Fig. 1c,?,d).d). We annotated clusters by comparing differentially expressed genes (DEGs) defining each to known lineage markers and previously published.
