A single cell sequencing approach to determine the heterogeneity, dynamics and cell fate decisions of retinal progenitor cells in vivo and in vitro

被告知患有无法治疗的视力障碍可能很难接受，但这是全世界2.85亿人必须承受的负担。26%的全球失明是由视网膜功能障碍引起的，视网膜是最内层的光敏组织，排列在眼睛的后部，对光传感和图像处理至关重要。视网膜功能障碍和随后的视力丧失可以通过我们从父母那里继承的错误基因的影响以及损伤的积累和我们一生中各种疾病的影响而发生。我们预防和治疗视力丧失的能力与我们对“视网膜如何形成”以及何时以及什么可能出错的知识密切相关。我们的视网膜主要在出生前发育;因此，从这一时期开始研究的组织的可用性非常有限。我的团队在弥合这一差距方面处于独特的地位，通过与人类发育生物学资源密切和良好的合作，可以获得人类视网膜，该资源在母亲同意的情况下收集流产胚胎和胎儿的样本。我们还具有在实验室中创建称为“视网膜类器官”的三维结构的优势，这些结构类似于人类视网膜在发育过程中的形成，并包含关键的视网膜细胞类型。我们的目标是利用这两种独特的资源来了解视网膜是如何形成的，以及何时形成的，以及当出现故障时导致视力丧失的基因的作用。视网膜是一种复杂的组织，由七种细胞类型组成：这些细胞在我们发育过程中的不同时间点从祖细胞库中出现，其本身是异质的，具有各种亚群，这些亚群被认为在时间的简明进程中产生不同的细胞类型。由于这个原因，很难用依赖于细胞混合物研究的传统研究方法来确定产生构成人类视网膜的所有细胞类型的祖细胞。在这里，我们建议使用一种重要的新技术，称为单细胞分析，它使我们能够看到哪些基因在群体中的每个细胞中被打开。单细胞水平的基因表达是精确分类细胞的一个非常可靠的工具，使我们能够识别在显微镜下观察其形状或位置时不明显的细胞类型。我们将以此为第一步，探索发育中的视网膜和我们实验室产生的视网膜类器官中单个视网膜细胞的分子差异。然后，使用先进的数据分析技术将使我们能够建立细胞类型及其特征基因的目录，将祖细胞与发育过程中的各种细胞类型进行匹配，预测它们的最终命运，并评估实验室产生的视网膜器官的接近程度。类器官模仿人类视网膜的发育。其次，我们将使用单细胞测序数据，使用生物信息学工具重建谱系树。这种方法在“伪时间”中组织细胞，预测细胞命运决定的顺序和模式，使我们能够预测占据树的分支点周围特殊位置的基因。第三，我们将应用一种新的方法，使我们能够将单个细胞的基因表达谱与它们在视网膜中的位置相关联，从而创建我们视网膜发育的空间图。这个空间图将使我们能够验证在分支点附近发现的关键基因的表达，这对于理解祖细胞在最终成为视网膜细胞的过程中做出的决定可能很重要。最后，我们将评估基因表达周围的分支点发挥积极作用，通过操纵他们的表达控制细胞命运的决定。这些信息将提供给所有科学家和临床医生，以帮助他们了解视网膜发育和疾病。

Being told that you have a visual impairment that can't be treated can be difficult to accept but this is the burden that 285 million people worldwide must bear. 26% of global blindness is caused by dysfunction of the retina, which is the innermost, light sensitive tissue that lines the back of the eye and is vital for light sensing and image processing. Dysfunction of retina and subsequent vision loss can occur through the effect of faulty genes we inherit from our parents as well as the accumulation of damage and the effect of various diseases throughout our lives. Our ability to prevent and treat vision loss is closely linked to our knowledge of "how our retinas form" and when and what is likely to go wrong. Our retinas develop mostly before birth; hence the availability of tissue to study from this time period is very limited. My group is in a unique position to bridge this gap, having access to human retinas through close and well established collaborations with the Human Developmental Biology Resource, which collects samples from aborted embryos and fetuses with the mother's consent. We also have the advantage of creating in the lab three-dimensional structures called "retinal organoids", which resemble the formation of human retina during development and contain the key retinal cell types. Our aim is to use both of these unique resources to understand how and when the retina forms and the role of genes that cause loss of vision when faulty.Retina is a complex tissue and is composed of seven cell types: these emerge at different points during our development from a pool of progenitor cells, which in itself is heterogeneous with various subsets suggested to give rise to the different cell types in a concise progression through time. For this reason, it has been difficult to pinpoint the progenitor cells which give rise to all the cell types that make up the human retina with the traditional research methods that rely on studies of cell mixtures. Here we propose to use an important new technology called single cell analysis which allows us to look at which genes are turned on in each cell in the population. Gene expression at the single cell level is a very reliable tool for the precise categorisation of cells and allows us to identify types of cells that are not noticeable when looking under the microscope at their shape or position. We will use this as a first step to explore the molecular differences of individual retinal cells in both developing retinas and the retinal organoids generated in our lab. The use of advanced data analysis techniques will then allow us to build a catalogue of cell types and the genes that characterise them, to match the progenitors to the various cell types across development, to predict their ultimate fate and to assess how closely the lab generated retinal organoids mimic the development of human retina. Second, we will use the single cell sequencing data to reconstruct a lineage tree using bioinformatics tools. This approach organises cells in 'pseudo-time', predicting the order and mode in which cell fate decisions are made, enabling us to predict genes that occupy special positions around branch points of the tree. Third, we will apply a new approach, which allows us to correlate the gene expression profile of individual cells with their location in the retina, thus creating a spatial map of our retinas as they develop. This spatial map will allow us to validate the expression of key genes that are found near the branch points which may be important to understand the decision that progenitor cells make towards their final trip to become retinal cells. Finally, we will assess whether genes expressed around branch points play an active role in controlling cell fate decisions by manipulating their expression. The information will be available to all scientists and clinicians to help their understanding of retinal development and disease.