Enriching protein-protein and functional interaction networks in human embryonic stem cells

Human Embryonic Stem Cells (hESCs) have a great therapeutic potential in regenerative medicine, but the precise molecular mechanisms by which hESCs maintain or regulate their characteristics remain largely unknown. Since protein-protein interaction is vitally important in regulating hESCs, we utilized a network-based bioinformatics analysis in order to learn what and how specific proteins interact with each other. By combining protein-protein interaction data and a collection of genes over-expressed in hESCs, we constructed a protein interaction network using a breadth-first search algorithm. This scale-free network which is significantly larger than networks generated by random samplings, illustrates how these hESC-enriched proteins might interact with each other in hESCs. Of the top 5% highly connected nodes (corresponding to 21 proteins including MYC, H2AFX, RUVBL1, DDX18, CDC2, HDAC2 and HIST1H4C) presumably critical for determining the fate of hESCs, nearly half are known to be regulated by NANOG/SOX2/MYC. This underscores importance of these transcription factors in hESCs. In addition, in silico cis-element analysis suggests that NF-Y may be an important transcription factor regulating many of these hub proteins (high connected nodes) in hESCs. To further abstract the functional significance, directly connected proteins were matched to and grouped by gene ontology (GO) terms in molecular function category. Sixty- six interacting GO-GO terms paired through protein interactions were found over-represented in hESCs. This functional enrichment may be essential for understanding molecular characteristics in hESCs. Collectively, we analyzed hESC-enriched genes based on protein-protein interaction data, from which an hESC-enriched protein interaction network was constructed and a network of molecular functional terms was also identified. The results of this analysis, on the systems level, may shed new light to further our understanding of hESCs.