From a Long List to Causal Variants: High-Throughput Gene Regulatory Assays in Developing Tissues

摘要 我的实验室正在探索最近进化的人类特征的遗传基础，这些特征没有在 黑猩猩或其他类人猿。我们正在为实现这一目标而努力，确定并在功能上 以人类基因组中进化最快的区域为特征。虽然这一系列的研究已经 之前一直在进行研究，研究总是集中在基因组的5%显示出强大的 非人类哺乳动物中的跨物种序列限制，因此明显具有功能。 然而，我们现在知道，有许多功能区并不表现出很强的跨物种 在不同的哺乳动物中受到限制。这也是真的，许多表型转换在 人类的分支，如脑扩张和直立姿势，也见于其他哺乳动物 血统。例如，海豚和大象的大脑比人类的大。狐猴和 大型足目动物(如袋鼠)的脊椎通常与地面垂直。为 这些原因，我们假设许多人类适应和适应的基因组元素 疾病风险将不会局限于基因组中表现出强烈跨物种的5% 在其他哺乳动物中的保护。因此，我的实验室正在冒险进入另外95%的人类 基因组，寻找在人类中迅速变化但也可能在其他区域中变化的区域 物种。这使得我们面临着最初将研究人员限制在 5%的哺乳动物基因组是高度受限的。我们如何确定我们2089个地区中的哪些地区 是有功能的，我们如何知道人类谱系上的序列变化是否会改变他们的 功能呢？这是基因组学研究中常见的脱节：计算屏幕已经产生了 数以千计有趣的突变，无论是人类快速进化的区域，还是来自 全基因组关联研究，但为深入研究这些突变建立小鼠模型 只能对一到十个突变进行。这会导致两个数量级的断开 在我们可以识别用于进一步分析的突变和我们可以进一步分析的突变之间。浩瀚无边 人类基因组中进化最快的2089个区域中，大部分位于 编码蛋白质的外显子，并可能调节基因的时空表达 发展。虽然有高通量的方法来测试细胞系中的增强子活性，但 细胞的活动类型通常是事先不知道的，细胞系也不会捕捉到 细胞类型在发育过程中是暂时存在的。为了在我们自己的工作中解决这个问题，并与 为了被其他小组采用，我们正在开发一种方法来测试这种基因 数千个DNA片段在发育中的组织和单细胞分辨率下的调节潜力。

Abstract My laboratory is pursuing the genetic basis of recently evolved human traits that are not observed in chimpanzees, or other great apes. We are working towards this goal by identifying and functionally characterizing the fastest evolving regions in the human genome. While this line of research has been pursued before, the studies have always focused on the 5% of the genome that shows strong cross-species sequence constraint in non-human mammals, and is therefore clearly functional. However, we now know there are many functional regions that do not show strong cross-species constraint across diverse mammals. It is also true that many phenotypic transitions seen on the branch to humans, such as brain expansion and upright posture, are also seen in other mammalian lineages. For example, dolphins and elephants have larger brains than humans. Lemurs and macropods (e.g. kangaroos) also have a spine that is generally perpendicular to the ground. For these reasons, we hypothesize that many genomic elements underlying human adaptations and disease risks will not be restricted to the 5% of the genome showing strong cross-species conservation in other mammals. My laboratory is therefore venturing into the other 95% of the human genome, in search of regions that have rapidly changed in humans, but may also vary among other species. This has caused us to be faced with the problems that originally restricted researchers to the 5% of mammalian genomes that is highly constrained. How do we identify which of our 2089 regions are functional, and how do we know if the sequence changes on the human lineage change their function? This is a common disconnect in genomics research: a computational screen has generated thousands of interesting mutations, whether it be fast evolving regions in humans, or haplotypes from genome-wide association studies, but making mouse models for in depth study of these mutations can only be done for one to ten mutations. This results in a disconnect of two orders of magnitude between the mutations we can identify for further analysis and those we can further analyze. The vast majority of the 2089 fastest evolving regions in the human genome are located outside of protein-coding exons and are likely regulating the spatial and temporal expression of genes during development. While there are high-throughput methods for testing enhancer activity in cell lines, the cell type of activity is not usually known beforehand, and cell lines do not capture the complexity of cell types transiently present during development. To address this problem for our own work, and with the intention of it being adopted by other groups, we are developing a method to assay the gene regulatory potential of thousands of DNA segments in developing tissue and at single cell resolution.