Morphomic analysis of a simple chordate

描述(由申请者提供)：这个提议的合作项目将研究推动脊索胚胎发生的基本过程。该项目将结合两个研究小组的技能和专业知识：一个在发育生物学领域工作，另一个在图像分析和计算机视觉领域。该项目的目标是采用全胚胎方法在所有四个维度(x、y、z和t)研究活胚胎的形态发生。具体地说，我们将收集和分析共聚焦显微镜图像，以获得选定发育器官和整个胚胎中所有细胞的分裂、形状、体积和运动的定量数据。该项目在样本选择和标记、图像捕获以及图像分析和可视化方面提出了许多挑战。大多数模型生物从根本上不适合捕捉整个组织中每个细胞的形状和运动。那些适合的通常是脊索动物的远亲，因此缺乏与人类健康和疾病相关的器官和组织。然而，海鞘(海鞘)是脊椎动物的近亲，脊椎动物结合了脊索状的身体平面和适合于整体成像的小而简单的胚胎结构。在这样的观测中收集的数据，特别是在观察细胞行为所需的高分辨率下收集的数据将是巨大的。从时间流逝图像集中提取大规模的定量信息给图像分割和分析带来了巨大的挑战。我们已经在3D海鞘图像分割方面取得了重大进展，并提出了这些方法的进一步改进。分割的数据将允许对许多复杂的细胞行为进行定量分析，包括驱动原肠形成、神经形成和收敛延伸的细胞形状和运动的根本变化。这些分析将导致关于驱动形态发生的细胞和分子机制的具体假设。为了进一步研究这些细胞现象的分子基础，我们还将扩展我们的成像和分析努力，以研究在形态发生中起关键作用的几种蛋白质的亚细胞定位的动态模式。

DESCRIPTION (provided by applicant): This proposed collaborative project will investigate fundamental processes driving chordate embryogenesis. The project will combine the skills and expertise of two research groups: one that works in the area of developmental biology, and the other in the area of image analysis and computer vision. The goal of the project is take a whole-embryo embryo approach to investigating morphogenesis in live embryos in all 4 dimensions (x,y, z and t). Specifically, we will collect and analyze confocal microscopy images to derive quantitative data on the division, shape, volume and movements of all cells in both selected developing organs and in whole embryos. This project presents many challenges in sample selection and labeling, image capture, and image analysis and visualization. Most model organisms are fundamentally unsuited to the challenge of capturing the shape and movement of every cell in an entire tissue. Those that are suitable typically are distantly related to the chordates, and thus lack organs and tissues relevant to human health and disease. The ascidians (sea squirts), however, are close relatives of the vertebrates that combine a chordate body plan with a small and simple embryonic architecture suitable for imaging in toto. The data to be collected in such observations, particularly at the high resolution needed to observe cellular behavior, will be enormous. The extraction of large-scale quantitative information from timelapse image sets poses substantial challenges in terms of image segmentation and analysis. We have already made significant progress in segmenting 3D ascidian images and further refinements of these approaches are proposed here. The segmented data will allow the quantitative analysis of many complex cell behaviors, including the fundamental changes in cell shape and motility driving gastrulation, neurulation and convergent extension. These analyses will lead to specific hypotheses regarding the cellular and molecular machinery driving morphogenesis. To further investigate the molecular basis for these cellular phenomena, we will also extend our imaging and analysis efforts to investigate the dynamic patterns of subcellular localization of several proteins with key roles in morphogenesis.