Extent, Heritability, and Functional Relevance of Single Cell Expression Variability in Highly Homogeneous Populations of Human Cells

Abstract

Because of recent technological developments, single-cell assays such as single-cell RNA sequencing (scRNA-seq) have become much more widely available and have achieved unprecedented resolution in revealing cell heterogeneity. The extent of intrinsic cell-to-cell variability in gene expression, or single cell expression variability (scEV), has thus been increasingly appreciated. However, it remains unclear whether this variability is functionally important and, if so, what its implications are for multi-cellular organisms. We therefore analyzed multiple scRNA-seq data sets from lymphoblastoid cell lines (LCLs), lung airway epithelial cells (LAECs), and dermal fibroblasts (DFs). For each of the three cell types, we estimated scEV in homogeneous populations of cells; we identified 465, 466, and 291 highly variable genes (HVGs), respectively. These HVGs were enriched with specific functions precisely relevant to the cell types, from which the scRNA-seq data used to identify HVGs were generated — e.g., HVGs identified in lymphoblastoid cells were enriched in cytokine signaling pathways, LAECs collagen formation, and DFs keratinization. HVGs were deeply embedded in gene regulatory networks specific to corresponding cell types. We also found that scEV is a heritable trait, partially determined by cell donors’ genetic makeups. Furthermore, across genes, especially immune genes, levels of scEV and between-individual variability in gene expression were positively correlated, suggesting a potential link between the two variabilities measured at different organizational levels. Taken together, our results support the “variation is function” hypothesis, which postulates that scEV is required for higher-level system function. Thus, we argue that quantifying and characterizing scEV in relevant cell types may deepen our understating of normal as well as pathological cellular processes.