Comparative evaluation of pathways and gene expression profile similarity in differentiated stem cells versus normal adult cells in seven human tissues

Abstract

Pluripotent and multipotent stem cells (SCs) are the main utilized sources in differentiation studies. Even though multipotent SCs typically have a higher differentiation capability than pluripotent SCs, the differentiation potential of multipotent SCs is controversial due to the lack of a unanimous description. The aim of the current research is to use a bioinformatics approach to find the differentiated stem cells with the most similarity with corresponding normal cells. This approach can be used as a method to select the best source of differentiation. We compared the similarity among 27 gene expression profiles of human differentiated SCs toward bone, fat, cartilage, heart, muscle, neuron, and pancreas against their NACs. Pearson's correlations showed a transcriptome similarity between differentiated SCs and NACs. We compared the correlations to the experimental studies in the literature. In addition, Gene Set Enrichment Analysis (GSEA) of KEGG pathways revealed the differences between differentiated SCs and NACs by determining the higher and lower expression pathways in differentiated SCs against NACs. We demonstrated that differentiation of umbilical cord blood-derived mesenchymal SC (UC-MSC) toward fat and hurt tissues had the highest correlation with normal adipocyte and cardiomyocyte cells, respectively. Furthermore, in cardiogenesis, neurogenesis, and pancreatogenesis classes, the differentiated embryonic SCs had the lowest correlation against NACs, while in adipogenesis and osteogenesis classes, differentiated ESC-MSC had a high correlation. We observed a lower expression of cell cycle and DNA replication pathways in 16 out of 20 differentiated stem cells compared with their NACs. In this bioinformatics study, we compared the homology between 20 differentiated SCs and corresponding normal cells using Pearson's correlation and GSEA assay.