Quantifying cell behaviour in morphogenesis

在动物发育过程中，胚胎细胞经历巨大的运动和重组，将自己塑造成身体的组织和器官。这一过程如此复杂，以至于可能会出现错误的阶段很多，其结果可以通过出生缺陷的形式看到。在许多情况下，例如神经管缺陷，这些错误的后果可能是严重的。要了解这些错误是如何产生的，研究取决于我们是否能够观察和测量细胞在重新排列期间的运动，以询问它们的行为如何偏离正常发育。显微镜的进步现在使我们能够看到细胞的三维轮廓，并随着这些运动的发生而跟踪活胚胎的发育。在我们的工作中，我们使用了一种小鱼，斑马鱼，对于斑马鱼来说，有许多突变株具有类似于人类发育障碍的缺陷，特别是在中枢神经系统的发育方面。通过对这种模型动物的研究，希望能够发现理解可以转化为对人类状况的洞察。我们在这里提出的工作是开发新的、更先进的计算方法，使我们能够跟踪和测量可视化的细胞的运动和重组的细节。细胞和组织具有复杂多样的三维形状，单个3D图像包含数百个细胞。其中只有很小一部分可以手动分析，但我们检测到的许多发育错误都是正常情况下的细微变化，因此需要许多准确的估计。大脑等复杂器官的发育涉及许多不同的运动。一个例子是两只眼睛的形成。在发育早期，只存在一层扁平的细胞，它会分裂并重塑成两个眼球。通过使用这些新的成像方法来跟踪活胚胎的延时电影中数百个细胞的运动，人们现在才能理解这所需要的折叠和重塑。通过比较正常胚胎和突变胚胎中眼睛无法正常发育的运动，我们可以确定导致眼睛形成的不同机制。通过这种方式，我们现在可以开始询问有关这些机制中的错误是如何导致出生缺陷的详细问题。为了进一步推进这些研究，我们现在需要仔细和全面地测量眼睛细胞的运动，同时操纵与产生这些缺陷有关的基因的活动。我们在这里提出的方法将允许我们以前所未有的方式研究这个问题和类似的问题。我们在这里提出的方法解决三个相关的问题。第一个是使我们能够在三维中测量我们收集的发育中组织的3D电影中所有细胞的形状和运动。第二是开发必要的数学方法，通过表征整个组织(如大脑)的形状变化和重排，来衡量这种结构内的重组如何随着时间的推移而变化。研究的第三条途径是开发模拟实际胚胎行为的发育组织的计算机模型。数百个细胞的运动非常复杂，但通过建立数值模拟，我们能够询问在实现在真实动物中看到的运动中，哪些特征是重要的。该模型的所有特征都可以与实验观察进行比较，然后这些模型被用来测试关于胚胎内的力是如何实际导致大脑形成的假说，以及为什么它们有时会出错。

During animal development, the cells of the embryo undergo enormous movements and reorganisations to shape themselves into the tissues and organs of the body. The complexity of the process is such that there are many stages at which errors might occur, the results of which can be seen in the form of birth defects. In many cases, such as neural tube defects, the consequences of there errors can be severe. Research into understanding how these errors arise depends upon us being able to view and measure the movements of cells during the time at which they are rearranging to ask how their behaviours diverge from normal development. Advances in microscopy now allow us to see the outlines of cells in three dimensions and to follow the development of living embryos as these movements take place. In our work we use a small fish, the zebrafish, for which there are many mutant strains with defects analogous to human developmental disorders, particularly for the development of the central nervous system. By studying this model animal it is hoped that understanding can be found that can be translated into insight into the human condition.The work that we propose here is to develop new, more advanced computational methods that will allow us to follow and measure the movements and reorganisations of cells visualised in great detail. Cells and tissues have complex and varied three dimensional shapes and individual 3D images contain many hundreds of cells. Only a very small proportion of these can be analysed manually, but many of the developmental errors that we detect are subtle variations from the normal path, so many precise estimates are needed. The development of a complex organ such as the brain involves many different movements. One example case is the formation of the two eyes. Early during development there exists just one flat sheet of cells that will split and reshape to form two eyeballs. The folding and reshaping that this entails is only now being understood by using these new imaging methods to follow the movements of many hundreds of cells in time-lapse movies of living embryos. Comparing movements seen in normal embryos with mutant embryos, in which the eyes fails to develop correctly, has allowed us to identify the distinct mechanisms that lead to the formation of the eye. In this way we can now begin to ask detailed questions about how errors arise in these mechanisms to cause birth defects. To progress further with these studies we now need to carefully and comprehensively measure the movements of cells of the eye while manipulating the activity of the genes involved in generating these defects. The methods that we propose here will permit us to study this and similar problems in ways that have never before been possible.The methods that we propose to develop here address three related problems. The first is to enable us to measure in three dimensions the shapes and movements of all of the cells within the 3D movies we collect of developing tissues. The second is to develop the mathematical methods needed to measure how reorganisations within this structure changes over time by characterising shape changes and rearrangements of the cells of whole tissues, such as the brain. The third avenue of research is to develop computer models of developing tissues that mimic the behaviour of the actual embryo. The movements of many hundreds of cells is very complicated to understand, but by building numerical simulations we are able to ask what features are important in achieving the movements seen in real animals. All characteristics of the model can be compared to experimental observations and the models then used to test hypotheses about how forces are applied within the embryo to actually cause the brain to form, and why they sometimes go wrong.