アイテム詳細

要旨

Understanding the network of transcription factors, controlling pluripotency in human embryonic stem cells (ESCs) and human embryonal cancer cells (ECs),is essential for possible future therapies in medicine. Connecting the expression levels after ablation of OCT4 with potential binding sites allows a higher predictability of motif specific driven expression modules important for selfrenewal and differentiation. In this study several peak analysis programs have been used to access a refined list of OCT4 targets in human EC cells and this data was connected to ES cell specific OCT4 binding and expression. A highly enriched POU-motif could be verified, discovered by a de novo approach, thus enabling connections to the distribution of OCT4 connected motifs like for the dimerisation factor SOX2. Selected targets have been validated, containing an OCT4-SOX2 binding site in their proximal promoter, and targets not connected to the classical HMG motif. Of those USP44 was further examined, containing a highly conserved POU-motif and GADD45G, having an impact on cell cycle regulation.The overexpression of GADD45G in EC cells resulted in an enrichment for upregulated genes, connected to differentiation pathways. Additionally preferred distances for the HMG and the POU motif could be observed, giving cause for additional binding modes than the classical HMG-POU consensus sequence. New OCT4 connected targets were discovered, and their importance in ESC differentiation and pluripotency was highlighted. Through a highly connected database, everyone can test now simple hypotheses based on their target genes. The use of NCCIT cells as a model to test pluripotency associated pathways in terms of potential functional binding sites has been demonstrated.Furthermore array based comparisons of gene expression levels between ES and EC cells have been conducted and new links have been established for further functional characterisation of these cells. Finally a ChIP-seq study revealed an unbiased genome wide view on putative OCT4 bound regions and suggested a genome wide binding pattern for OCT4 which is not centered for five prime proximal promoters.