Deciphering human cell type enriched transcriptomes across tissue types and the functional study of the endothelial enriched protein KANK3

Abstract

The identification of cell type-specific genes and their modification under certain conditions is essential for our understanding of the human body. Such knowledge also allows us to study the function of uncharacterised genes in the correct cell type. Single-cell RNA sequencing (scRNAseq) can be used to profile different cell types, but cell removal and processing can introduce artifacts. In the presented thesis, we used unfractionated bulk RNA sequencing (RNAseq) data to identify cell type-specific transcriptomes, and furthermore, used the generated data to select a candidate gene for functional analysis in endothelial cells (EC). We used an integrative correlation-based method to analyze RNAseq data from the Genotype Tissue Expression Project (GTEx). The gastrointestinal tract organs stomach (paper I) and colon (paper II) were analyzed. In both cases, we profiled cell-enriched transcriptomes of cell types that are absent from the major scRNAseq databases and identified several non-coding genes with cell type enriched profiles. A sex-based subset analysis revealed that cell profiles were broadly similar in males and females. We extended our analysis to include 15 human organs (paper III). A cross-tissue analysis revealed common cell type enriched gene signatures between related cell types, or the same cell types found in different tissues, such as EC. We subsequently selected one of the most consistently EC enriched genes, which had no described function in this cell type, for further study, KANK3 (paper IV). We found that KANK3 was regulated by shear stress. EC KANK3 depletion also revealed a role in EC migration, and in regulation of tissue factor (TF) expression, which is involved in coagulation. These studies provide a roadmap for cell type enrichment profiles and demonstrate how such data can be used to select genes for functional investigation. Additionally, highlighting the value of using alternative methods to extract information from previously published data.