Application of Deep Learning Models on Single-Cell RNA Sequencing Analysis Uncovers Novel Markers of Double Negative T Cells

Abstract

Abstract Double negative T (DNT) cells, a unique subset of CD3+TCRαβ+ T lymphocytes lacking CD4, CD8, or NK1.1 expression, constitute 3-5% of the total T cell population in C57BL/6 mice. Traditional machine learning models such as principal component analysis in single-cell RNA sequencing (scRNA-seq) analysis have been utilized to characterize this subset, but only deep learning models such as Single Cell Variational Inference (SCVI) can capture nonlinear gene expression in the sequencing data. In this study, using the deep learning approach, we have uncovered novel markers of splenic DNT cells in C57BL/6 mice. We classified DNT cells into two subgroups, naïve DNT (nDNT) cells and activated DNT (aDNT) cells, which could be differentiated by the expression of Ly6C and MHC-II, respectively. In addition, our data identified CD153 and CD153 as unique markers for aDNT cells. SCVI analysis suggested, and flow cytometry analysis confirmed, that Ly49G2 and CCR2 are respective markers for two distinct subsets within nDNT and aDNT populations. In addition to cell surface markers, we further discovered that nDNT cells expressed higher Rgs2 that is required for T cell activation and that aDNT cells expressed higher Rbpj involved in Notch signaling. Together, our comprehensive analysis has uncovered and validated novel markers for different subpopulations of DNT cells that can be used in the phenotypic and/or functional characterization of these relatively rare cells in health and disease.