通过工程化改造Oct4研究体细胞重编程过程的基因组调控机制

Oct4 (octamer binding transcription factor 4) belongs to the Pit-Oct-Unc (POU)transcription factor (TF) family. Only Oct4, but no other POU protein, can support the induction of pluripotency from somatic cells. If Oct4 is replaced by, for example, the PouIII factor Brn4, reprogrammed cells are committing to a neural cell fate. Efforts to study the dynamic changes during the reprogramming process suggested that Oct4 plays different, albeit essential, roles during the early and late stages of reprogramming. To reveal the basis for the unique roles of Oct4, we conducted quantitative DNA binding assays and detected structural elements that allow Oct4 to discriminate between composite DNA binding sites in vitro. We swapped those elements between Oct4 and Oct6 to engineer synthetic proteins with altered preferences for DNA sequences. As a consequence, these synthetic Oct factors exhibited profound differences in reprogramming assays as compared to their wild-type counterparts. For example, a doubly mutated Oct6 protein is now gains the ability pluripotency induction. Moreover, a mild point mutation of Oct4 termed Oct4M impairs Oct4's ability to support pluripotency maturation. Rather, Oct4M pauses at D40 of reprogramming and re-differentiates. Apparently, Oct4M can support initial phases of reprogramming but lost the function to support pluripotency maturation. Here we propose to expand these studies to resolve fundamental questions in the field of cellular reprogramming. What is the molecular function of Oct4 at various reprogramming stages? How do the subtly different DNA motif preferences between Oct4, Oct6 and synthetic factors change the inductions of cell fate decisions? Can we rationally engineer synthetic TF proteins that execute gene expression programs and direct cell fates in a controlled manner and thereby exclude harmful oncogenic properties of the resulting cells? To answer these questions, we will set up a secondary inducible system that allows comparing the reprogramming of mouse and human fibroblasts induced by Oct4, PouIII members and at least of four reengineered versions thereof. Next, we will perform chromatin immunoprecipitation coupled with high-throughput sequencing (ChIP-seq) to carry out a time resolved analysis at 3 time points to compare of their genomic binding profiles. This work will clarify and separate the roles of Oct4 at different stages of the reprogramming process, explain its uniqueness compared to other POU factors and facilitate the rational design of synthetic proteins to produce safe tissues suitable for cell therapies.

Oct4（也称POU5F1）是POU转录因子家族中唯一已知能介导体细胞重编程建立多能性的成员。Oct4对于体细胞重编程是必须的，但其作用机制及在重编程不同时期的功能差异尚不清楚。本课题组利用DNA结合定量分析技术，检测到Oct因子识别DNA序列的不同结构域，并通过工程化互换Oct4和Oct6的这些结构域，发现改造型Oct因子与其野生型对重编程有着显著不同的作用：改造后，Oct6能诱导重编程而Oct4影响后期多能性的维持并趋向分化。本研究在前期工作基础上，通过建立二次诱导系统，对Oct因子的调控模式进行深入研究：利用二次诱导系统在重编程细胞中分别表达野生及改造型Oct4，结合染色质免疫共沉淀-高通量测序技术（ChIP-seq），获得Oct因子在重编程起始、成熟、稳定期的全基因组调控信息，从而阐明Oct4在重编程过程中决定性的作用机制，进一步研发出安全有效的工程化重编程因子。

细胞重编程为研究细胞类型转变中染色质的变化提供了强大的实验系统。关于重编程机制的几个关键点，以前的研究的得出了部分矛盾的结论。我们决定专注于用不能诱导多能性的Pit-Oct-Unc（POU）转录因子来研究体细胞重编程。Oct4可以在体细胞中诱导多能性，但是与其具有高度同源性的旁系同源因子，诸如Oct6，却不具有此功能。在这个研究项目中，我们使用RNA测序和全基因组结合以及染色质开放性分析来揭示为何Oct4具有激活多能性诱导调控程序的独特性。我们将Oct4与Oct6以及一个重编程功能缺失的Oct4突变体（Oct4defSox2）进行对比。这个突变体在与Sox2的DNA依赖性二聚化中存在缺陷。我们发现重编程过程中细胞的初始应答不需要Oct4。 Oct4defSox2的早期染色质结合位点分布类似于Oct4，而Oct6则结合至一组不相关的结合位点。然而，在Oct6条件下，许多原先结合Oct4的多能性增强子由闭合状态转变为开放状态，而Oct4defSox2不能打开染色质。我们得出的结论是Oct4启动驱染色质开放的直接作用是有限的，并且这种功能可以被不能重编程的因子代替。然而，Sox2/Oct4复合物的形成对完成多能性诱导是必不可少的。但多能性的维持并不依赖Sox2/Oct4异源二聚体。利用对Oct4和Sox因子的序列、结构和功能的深刻理解，我们合理地重新设计了Oct6，将其转化为可以诱导多能性的因子。此外，我们将合理设计和突变文库筛选结合，在多能性重编程实验中鉴定出几种功能大幅度地优于野生型Sox2的高性能Sox因子。总的来说，该研究项目的结果表明，结合结构生物学、生物化学和基因组学的工具对细胞状态重编程进行详细剖析，我们能够重新设计内源重编程因子的骨架为提高体外和体内转录因子介导的细胞状态转化速度，效率和质量提供可能。