Distinct phenotypes of rare autoreactive and virus-reactive CD8+ T cells revealed through novel clustering algorithm

• Published in: The Journal of immunology (1950) Vol. 210; no. 1_Supplement; pp. 78 - 78.07
• Main Authors: Kus, Anna, Wiedeman, Alice, Long, Alice
• Format: Journal Article
• Language: English
• Published: 01.05.2023
• ISSN: 0022-1767; 1550-6606
• DOI: 10.4049/jimmunol.210.Supp.78.07

Antigen-specific cells are implicated in the development and progression of disease in autoimmunity, infectious disease, and cancer, and are an attractive target for therapies. While robust detection of antigen-specific cells is possible with tetramer and activation-induced marker assays, extensive characterization of these cells remains challenging due to their rarity. Here we present a novel computational tool, DISCOV-R, that enables deep phenotyping of rare antigen-specific cells through unbiased clustering of a parent population into subsets defined by expression of phenotyping markers and overlaying the antigen-specific population. Using this tool with 35-parameter mass cytometry (CyTOF) along with class I MHC-tetramer assays, we identified phenotypes enriched among rare islet autoantigen-reactive and chronic virus-reactive CD8+ T cells when compared to total CD8+ T cells from 46 type 1 diabetics. CD8+ T cells were clustered based on expression of 24 phenotypic markers, revealing 12 CD8+ T cell phenotypes shared across subjects. Overlaying antigen-specific tetramer-positive cells on this phenotypic landscape revealed that both islet- and virus-reactive cells were enriched for an exhausted-like memory phenotype (p=1.7E-07 and 1.6E-05, respectively) consistent with chronic antigen exposure, while islet-reactive cells were enriched for two transitional memory phenotypes (p=0.01 and 0.03). The characterization of these rare antigen-specific cells using DISCOV-R reveals unique phenotypes associated with their specificity that may indicate different function, and enables comparison over time and between individuals to identify biomarkers and potential pathways to disease development and progression.