Functional and translational studies of RUNX1 and CBFB in hematopoiesis

RUNX1和CBFB不仅对白血病的发生很重要，而且也是正常造血的关键调节因子。这两个基因在造血干细胞形成的最早阶段和许多血液谱系的许多后续阶段都是必需的。多项研究表明，由RUNX1和CBFB控制的正常转录程序的失调可能是白血病发生的重要机制。因此，更好地理解RUNX1/CBFB转录程序和RUNX1和CBFB在正常造血中的作用将导致更好地理解白血病发生的机制。 在上一个财政年度，我们一直在追求这个项目的两个具体目标。在第一个具体目标中，我们一直在确定CBFB和RUNX1在斑马鱼造血干细胞（HSC）形成中的作用。特别是，我们试图发现HSC是否可以在缺少RUNX1的情况下形成。以前的研究表明，RUNX1是胚胎发育过程中造血内皮细胞中永久造血干细胞（HSC）出现所必需的。例如，Runx1基因敲除小鼠胚胎缺乏所有确定的血液谱系，不能存活超过胚胎第13天。然而，我们最近发现，在runx1（runx1W84X/W84X）的ENU诱导的无义突变的纯合子斑马鱼能够从幼虫无血阶段恢复，并发展到具有多谱系造血功能的可育成体，这表明形成了runx1独立的成体HSC。然而，我们的发现是基于一个单一的斑马鱼突变系，这需要在独立的突变体中进行验证。 为了进一步研究RUNX1独立途径是否存在于成体HSC的形成中，我们使用转录激活因子样效应核酸酶（TALEN）技术产生了两种新的runx1突变体。这些突变导致移码和过早终止，导致runx1功能丧失（runx1-/-）。与先前描述的RUNX1敲除突变体类似，这些新的runx1-/-突变体胚胎未能发育出确定的造血。用共聚焦显微镜进行的延时记录证实了在runx1-/-胚胎中没有HSC的出现。runx1-/-幼虫在受精后（dpf）8至14天失去循环原始血细胞并变得无血。然而，一周后他们逐渐恢复了循环血细胞。最终，大约40%的runx1-/-突变体发育成具有多谱系循环血细胞的可生育成年人。综上所述，我们的数据与先前描述的runx1W84X/W84X表型一致，并支持HSC形成和永久性造血的runx1独立机制的存在。在下一个财政年度，我们将进行研究，以发现非RUNX1依赖性HSC形成的途径。 在第二个目标中，我们一直使用人类诱导多能干细胞（iPSC）来研究RUNX1在人类造血中的功能。 我们已经从患有家族性血小板疾病的患者中产生了iPSC，这是一种常染色体显性疾病，血小板数量减少，对白血病的易感性增加。我们还一直在与国家转化科学推进中心的科学家合作，研究iPSC分化为造血细胞的培养条件。最后，我们一直在使用基因组技术来确定iPSC的基因组完整性。近年来，人们对iPSC的基因组完整性进行了广泛的研究，仍然存在的一个核心问题是iPSC中基因组变异的起源。 先前的研究报告了关于基因组变异何时发生在iPSC中的相互矛盾的结果，无论它们是从亲本体细胞遗传而来，还是在重编程过程中产生。 我们已经采取了一种独特的方法来确定基因组变异的历史，通过从相同的亲本成纤维细胞获得单细胞克隆成纤维细胞系和克隆iPSC系，并应用下一代测序技术来检测DNA序列变异。 此外，我们通过SNP阵列询问较大的DNA改变，以确定每个样品中的结构变异。 全外显子组测序（WES）随后进行靶向深度再测序，结果显示，最初在WES子细胞中被鉴定为从头变异的大多数单核苷酸变异和小插入缺失是罕见的预先存在的变异，等位基因频率非常低。 深度重测序显示，从WES发现的总变体中只有不到0.1%在亲本成纤维细胞中完全不存在。拷贝数变异（CNV）倾向于在iPSC和成纤维细胞亚克隆中从头出现，而不是从亲本成纤维细胞遗传，但每个iPSC/成纤维细胞亚克隆仅携带1 - 3个CNV。总体而言，我们的数据显示，iPSC中存在的超过99.9%的基因组变异是从亲本体细胞遗传而来的，而不是重编程的结果。

RUNX1 and CBFB are not only important for leukemogenesis but they are also key regulators of normal hematopoiesis. These two genes are required during the earliest steps of hematopoietic stem cell formation and in many subsequent stages of many blood lineages. Multiple studies suggest that dysregulation of the normal transcriptional program controlled by RUNX1 and CBFB is likely to be an important mechanism for leukemogenesis. Therefore, better understanding of the RUNX1/CBFB transcriptional program and the roles of RUNX1 and CBFB in normal hematopoiesis will lead to better understanding of the mechanisms for leukemogenesis. We have been pursuing two specific aims in this project in the last fiscal year. In the first specific aim, we have been determining the roles of CBFB and RUNX1 in the formation of hematopoietic stem cells (HSCs) in zebrafish. In particular, we tried to discover if HSCs could form in the absence of RUNX1. Previous studies suggest that RUNX1 is required for the emergence of definitive hematopoietic stem cells (HSCs) from the hemogenic endothelium during embryo development. For example, Runx1 knockout mouse embryos lack all definitive blood lineages and cannot survive past embryonic day 13. However, we recently showed that zebrafish homozygous for an ENU-induced nonsense mutation in runx1 (runx1W84X/W84X) were able to recover from a larval bloodless phase and develop to fertile adults with multi-lineage hematopoiesis, suggesting the formation of runx1-independent adult HSCs. However, our finding was based on a single zebrafish mutant line, which requires verification in independent mutants. In order to further investigate if a RUNX1-independent pathway exists for the formation of adult HSCs, we generated two new runx1 mutants using the transcription activator-like effector nuclease (TALEN) technology. These mutations cause frameshifts and premature terminations, resulting in loss of function of runx1 (runx1-/-). Similar to previously described RUNX1 knockout mutants, these new runx1-/- mutant embryos failed to develop definitive hematopoiesis. Time-lapse recordings with confocal microscopy confirmed the absence of HSC emergence in the runx1-/- embryos. The runx1-/- larvae lost circulating primitive blood cells and became bloodless between 8 and 14 days post fertilization (dpf). However they gradually regained circulating blood cells one week later. Eventually, about 40% of runx1-/- mutants developed to fertile adults with circulating blood cells of multi-lineages. Taken together, our data is consistent with the previously described runx1W84X/W84X phenotype and supports the existence of a runx1-independent mechanism for HSC formation and definitive hematopoiesis. In the next fiscal year we will conduct research to discover the pathway for the RUNX1-independent HSC formation. In the second aim we have been using human induced pluripotent stem cells (iPSCs) to study the function of RUNX1 in human hematopoiesis. We have generated iPSCs from patients with familial platelet disorder, an autosomal dominant disease with reduced platelet numbers and increased susceptibility to leukemia. We have also been working on the culturing conditions for the differentiation of iPSCs to hematopoietic cells, in collaboration with scientists at the National Center for Advancing Translational Sciences. Finally, we have been using genomic technology to determine the genomic integrity of iPSCs. Genomic integrity of iPSCs has been extensively studied in recent years and one central question that still remains is the origin of genomic variations in iPSCs. Previous studies have reported conflicting results regarding when genomic variations occur in iPSCs, whether they are inherited from the parental somatic cells, or generated during the reprogramming process. We have undertaken a unique approach to determine the history of genomic variations by deriving single cell clonal fibroblast lines and clonal iPSC lines from the same parental fibroblasts and applied next generation sequencing technologies to detect the DNA sequence variations. Moreover, we interrogated larger DNA alterations by SNP arrays to identify structural variations in each samples. Whole exome sequencing (WES) followed by targeted deep re-sequencing revealed that most single nucleotide variants and small indels that were initially identified as de novo variants in the daughter cells from WES were rare pre-existing variants with very low allele frequency. Deep re-sequencing showed that only less than 0.1% of total variants discovered from WES were completely absent in the parental fibroblasts. Copy number variations (CNVs) tend to have emerged de novo in the iPSCs and fibroblast subclones than inherited from the parental fibroblasts, but each iPSC/fibroblast subclone carried only 1-3 CNVs. Overall our data showed that more than 99.9% of genomic variations present in iPSCs are inherited from parental somatic cells rather than as consequences of reprogramming.