Transcriptional Pause Release Of Pluripotency Genes Is Rate-limiting Step During Somatic Cell Reprogramming

iPSCs have two pivotal features of ESCs, namely the ability of self-renewal and differentiation into almost all cell types. Like ESCs, iPSCs can be applied to toxicology, drug screening and disease model study in a short term, and eventually to regenerative medicine. Besides, iPSCs can overcome ethical problems and immunological rejection that trouble the application of ESCs. iPSC technology has brought huge hope to scientists who want take ESCs into application. However, we have to admit that we are far away from the promising future. What mainly block iPSCs from its application are the efficiency of reprogramming and safety of iPSCs. The key to solve these two problems is to illustrate the underneath mechanism of somatic reprogramming. We aimed to explore reprogramming from gene transcriptional perspective. P-TEFb is a complex containing CDK9 and cyclinTl, cyclinT2 or cyclin K, among which CDK9 is the catalytic subunit, while cyclinTl, cyclinT 2 or cyclin K is the regulatory subnit. CDK9 can regulate the pause release of RNA pol II by phosphorylating the C-terminal domain of its largest subunit at serine 2(S2), thus make transcription proceed to productive elongation and produce mature mRNA. We were interested in elongation’s role in reprogramming. Firstly, we downregulated expression level of CDK9 by two different shRNA, then inhibited CDK9 activity by an inhibitor FP, and finally competitively inhibit its activity by a dominate-negative mutant CDK9 DN, and we found all the three measures that inhibit CDK9 block reprogramming. Interesting, they just inhibited formation of GFP+ colonies with no effect on colonies formation. By constructing a Dox-inducible CDK9 DN whose expression can be controlled at different time, we found CDK9 DN did not inhibit reprogramming until it was expressed at late stage during reprogramming. It has been well known that reprogramming can be divided into early stage and late stage roughly. Early stage involves a process defined as MET (mesencymal-to-epithelial transition) while a critical feature of late stage is the activation of endogenous pluripotency network, such as Oct4, Sox2 and Nanog. Hence, we hypothesed that CDK9 may regulate transcriptional pause release of pluripotency genes during the late stage of reprogramming. Consistently, our qPCR result showed that CDK9 DN over-expression inhibited the expression of pluripotency genes but with no effect on MET related genes. P-TEFb participates in forming two different complexes. The inactive P-TEFb is bound by HEXIM1 and 7SK RNP, while the active P-TEFb is bound by BRD4 (bromodomain-containing protein 4). CTD domain of BRD4 can bind CDK9 and cyclinT to release them from inactive complex. When we over-expressed BRD4 during reprogramming, efficiency of GFP+colonies formation was enhanced at about four folds compared to control. To further testify BRD4’ role in reprogramming, we also treated the cells with two different shRNAs for BRD4 and an inhibitor of BRD4, finding reprogramming was blocked. Considering BRD4 can activate P-TEFb, there was a possibility that BRD4 enhanced reprogramming by activating CDK9. To prove this, we constructed truncated mutants containing different functional domain of BRD4. Only BRD4 and truncated mutants covering CTD domain of BRD4 could promote reprogramming. Moreover, BRD4-CTD over-expression enhanced reprogramming more significantly than BRD4. It implies that BRD4 enhanced reprogramming by interacting with CDK9. Later, we constructed mutants that disrupted binding between BRD4-CTD and CDK9 and these mutants failed to promote reprogramming. By constructing Dox inducible BRD4-CTD, we found BRD4-CTD only enhanced reprogramming at late stage, which was consistent with CDK9. Besides, we over-expressed BRD4-CTD during reprogramming of human urine cells and found efficiency was also improved significantly. Our qPCR data showed over-expression of BRD4 and BRD4-CTD enhanced expression of pluripotency genes and did not affect expression of MET related genes. On the contrary, knocking down BRD4 by shRNA inhibited pluripotency genes’expression and ESCs with BRD4 knocked down could not maintain pluripotency of ESCs. Finally, we performed ChIP with MEF, intemediates at day 5 and day 8 and ESCs for RNA pol Ⅱ. We found that when BRD4-CTD was over-expressed, more RNA pol Ⅱ were bound to gene body of pluripotency genes, indicating higher elongation rate and expression level. Consistently, when we over-expressed HEXIM1 during reprogramming, expression of pluripotency genes was down-regulated and reprogramming was inhibited, knock-down of HEXIM1 by shRNA had opposite results. All these results demonstrated that pause release of RNA pol Ⅱ at pluripotency genes controlled by P-TEFb was a rate-limiting step during reprogramming. In summary, we found another critical checkpoint during reprogramming, namely transcriptional elongation of pluripotency genes. This is the first time we showed that general transcriptional elongation play a critical role in reprogramming.