Document resource
Background T cells are the primary mediators of cellular immunity and provide co-stimulation for humoral immunity as well as durable memory against encountered pathogens. Many immunotherapies, such as immune checkpoint inhibitors, CAR-T cells, TILs, cancer vaccines, and T cell engagers/redirectors, work by inducing a robust T cell response targeting tumor-associated antigens. Laboratory tests that provide detailed T cell activation profiles may, therefore, have predictive value as biomarkers of clinical response, and can reveal mechanistic information for this class of drugs. Here we demonstrate how high-plex proteomics alongside single-cell transcriptomics enables deep characterization of T cell response against target antigens.Methods PBMC from five COVID-19 infected subjects were cultured overnight with a pool of 315 peptides derived from Spike protein (JPT) or Cytostim polyclonal stimulus (Miltenyi Biotec). After stimulation, cell culture supernatants were collected and frozen for proteomic analysis by Olink Reveal. Cultured cells were stained with antibodies against activation markers (CD69, CD134, CD137) and sorted by FACS to isolate antigen-specific T cells. Single-cell transcriptomics was performed on the sorted cells using the 10x Genomics 5’ HT v2 workflow. Libraries were constructed using procedures established at CellCarta and sequenced on an Illumina NovaSeq X to the required depth. Sequencing files were processed using software developed by Olink and 10x Genomics, respectively, followed by custom pipelines developed at CellCarta.Results The 1,034-plex protein data from Olink Reveal showed that profiles of COVID-19 specific T cell response from peptide stimulation were distinguishable from polyclonal stimulation. Hallmarks of T cell activation, such as IL-2, TNF-alpha, IFN-gamma, Granzyme B, and various chemokines were present with both stimuli but to varying degrees. Peptide stimulation was biased towards production of IL-2 while polyclonal stimulation induced higher levels of most other factors. Many proteins showed differences across the five subjects evaluated here, indicating, not surprisingly, that the quality of T cell response varies across individuals. To understand the source of this variation, we used the single-cell transcriptomic data to quantify the cell frequency and magnitude of transcript expression corresponding to the proteins from the culture supernatant. We identified a set of proteins that were induced in culture but showed no expression in sorted activated T cells, presumably secreted by antigen-presenting cells during culture.Conclusions The combination of high-plex proteomics and single-cell RNA-seq provides a comprehensive picture of antigen-specific T cell activation. We predict that these type of deep, multi-omic approaches will provide valuable clinical biomarker data in immuno-oncology trials.Ethics Approval All participants gave informed consent before taking part in this study.