Robust scaling and self-organisation of the Drosophila anteroposterior axis

一个成年的有机体包含许多不同类型的细胞，组织成一个复杂但有序的排列。胚胎模式化是发育生物学的一个领域，它关注这种复杂的排列是如何形成的。胚胎模式化通常是稳健的，尽管输入和条件不同，但仍能产生可靠的输出。它可以扩展（适应胚胎大小成比例），例如在双胞胎的生产过程中。最后，它需要大量的自组织（从低层次过程中出现高层次模式），因为成年生物比最初的受精卵复杂得多。胚胎模式化已经研究了几十年，使用“模型生物”实验和数学理论的混合。然而，主流理论很难解释实验操作中观察到的鲁棒性、可扩展性和自组织性，这表明与生物学现实存在严重的不匹配。这种不匹配在早期果蝇胚胎中尤其明显，这是一个重要的模型系统，比大多数其他物种更简单，更广泛地研究，更有利于遗传实验。果蝇的前后运动的早期阶段（头到尾）模式比我们能解释的更强大，即使我们知道涉及的15个基因，胚胎母亲发出的4个初始信号，假设我们相信，果蝇模式的健壮性来自于整个早期前后模式的结构网络（15个基因和它们的4个输入之间的全套调控相互作用），结合这些基因表达的mRNA和蛋白质分子在附近的核之间扩散的事实。虽然在早期的果蝇胚胎中有数千个细胞核，但直到最初的前后模式形成之后，细胞膜才在它们之间形成。我们假设，图案网络利用核之间的空间相互作用，在整个组织的水平上产生图案调节;这一观点与目前应用于果蝇胚胎的数学模型形成了鲜明的对比，后者假设对模式化的输入将被解释（读出）局部。连续性我们的目标是解析图案化网络的结构，并解释为什么它如此可靠地产生接近野生型（正常）胚胎模式的输出，即使胚胎中的起始条件受到相当强烈的干扰。然后，我们将使用这种新的理解模式在果蝇胚胎提取新的一般原则，可用于了解其他动物胚胎的发育模式，或在合成的胚胎样结构，现在可以从干细胞产生。我们将使用最先进的成像方法来揭示模式如何在野生型和突变型胚胎中展开，然后使用计算机模拟来了解这些行为是如何由潜在的基因网络产生的。潜在的应用和好处这项工作将通过解决一个长期存在的基本问题来推进我们对胚胎发育的基本理解。我们的发现将直接关系到发展生物学家（包括果蝇研究人员和研究其他动物系统的人），以及从理论角度研究模式的数学生物学家。我们发现的原则也将在合成和干细胞生物学中有实际应用，有助于在发育疾病，再生和生物工程中的长期转化应用。

BACKGROUNDAn adult organism contains many different types of cells, organised into a complicated but orderly arrangement. Embryo patterning is the field of developmental biology concerned with how this complicated arrangement is created.Embryo patterning is typically robust, producing reliable outputs despite variable inputs and conditions. It can scale (adapt proportionally to embryo size), for example during the production of twins. Finally, it requires a great deal of self-organisation (emergence of high level pattern from lower level processes), since the adult organism is much more complicated than the initial fertilised egg.Embryo patterning has been studied for decades, using a mixture of "model organism" experiments and mathematical theory. However, prevailing theories struggle to account for the robustness, scalability and self-organisation observed during experimental manipulations, indicating a serious mismatch with biological reality.This mismatch is particularly apparent in the early Drosophila (fruit fly) embryo, an important model system that is simpler, more extensively studied, and more conducive to genetic experiments than most other species. The early stages of Drosophila's anteroposterior (head-to-tail) patterning are more robust than we can account for, even though we know the 15 genes involved, the 4 initial signals laid down by the embryo's mother that they respond to, and some of the regulatory interactions between these components.HYPOTHESISWe believe that the robustness of Drosophila patterning emerges from the structure of the whole early anteroposterior patterning network (the full set of regulatory interactions between the 15 genes and their 4 inputs), combined with the fact that the mRNA and protein molecules expressed from these genes diffuse between nearby nuclei. While there are thousands of nuclei in the early Drosophila embryo, cell membranes do not form between them until after the initial anteroposterior pattern is laid down. We hypothesise that the patterning network exploits the spatial interactions between nuclei to generate pattern regulation at the level of the whole tissue; this idea contrasts with the mathematical models currently applied to the Drosophila embryo, which assume that the inputs to patterning will be interpreted (read-out) locally.OBJECTIVESWe aim to resolve the structure of the patterning network, and explain why it so reliably produces an output close to the wild-type (normal) embryo pattern, even if the starting conditions in the embryo are quite strongly perturbed. We will then use this new understanding of patterning in the Drosophila embryo to extract new general principles that can be used to understand developmental patterning in other animal embryos, or in the synthetic embryo-like structures that can now be generated from stem cells.APPROACHThis is an interdisciplinary proposal, which combines microscopy, genetics, and mathematical modelling. We will use cutting-edge imaging approaches to reveal how patterning unfolds within wild-type and mutant embryos, then use computational simulations to understand how these behaviours are produced by the underlying gene network.POTENTIAL APPLICATIONS AND BENEFITSThis work will advance our basic understanding of embryonic development, by solving a long-standing and fundamental problem. Our findings will be directly relevant to developmental biologists (both Drosophila researchers and those studying other animal systems), plus mathematical biologists studying patterning from a theoretical perspective. The principles we uncover will also have practical applications in synthetic and stem cell biology, contributing to long-term translational applications in developmental disease, regeneration, and bioengineering.