A Multiscale Approach to Genes Growth and Geometry

许多复杂的系统，如生物有机体、互联网或经济，涉及在许多不同规模上运行的综合并行过程。例如，有机体表现出大规模的平行组织，其单位在从分子到生态系统的各个层面上相互作用。一个关键的挑战是理解这种平行的多尺度相互作用是如何工作的。就生物系统而言，近年来的一个主要焦点是分子到细胞的尺度。然而，发育生物学和成像的进步，特别是在3D中分离和研究影响生物体发育的基因的能力，提高了从亚细胞到生物体水平的更大规模系统方法的可能性。这些尺度之间的关键联系是生长，因为正是通过生长，分子变化导致细胞大小、形状和分裂的改变，最终导致多细胞组织和生物体的产生。生长结构可分为达到稳定状态的结构，如植物的生长尖端，和形状不断变化的结构，如发育中的昆虫翅膀或植物叶子。在后一种情况下，当结构在形状和大小上经历几个数量级的变化时，捕获多重相互作用是一个特别的挑战。研究这个问题需要在许多层面上对生长、形状、模式和基因作用进行分析和整合。此外，需要开发正式语言和建模框架，以允许捕获动态和交互。通过在许多尺度上并行地追求这些方法，应该出现关键的实验和概念联系，而单独研究每个尺度是不明显的。实验和计算科学的最新进展使这一设想成为可能。该项目的目的是通过研究树叶获得其特征形状和图案的机制，为这种方法提供一个平台。该项目将包括实验分析、图像处理、计算机建模和四个不同尺度的集成。(1)在亚细胞水平上，我们将在3D中跟踪细胞骨架组成部分的合成和随时间平行变化的方式。然后将构建模型来解释这种行为，并通过进一步的实验进行测试。这些模型也将与细胞特性相关。(2)在细胞水平上，通过实时三维成像来确定生长叶片多个区域的细胞生长和分裂模式。将开发计算机语言，根据生长细胞之间的局部相互作用来模拟这种行为。模型将通过在叶片发育的特定地点和时间修改基因活性来测试。这些模型也将与整个器官水平的特性有关。(3)在器官水平，将使用多种成像方法在3D中跟踪叶片的生长。将开发一个建模框架，允许观察到的细胞和组织特性与基因作用相结合。模型将通过分析局部基因活性的实验操作如何改变生长来测试。这将有助于开发一个系统，用于测量在不同发育阶段影响叶片形状的多种突变体的3D形状。(4)在整个植物层面，叶片生长将被纳入一个虚拟植物中，其中模块之间的局部相互作用解释了生长和结构的动态。通过将各种模型结合起来，一个综合的植物发育的多尺度视图应该出现。该项目还将培养一批熟悉从分子遗传学、发育生物学、生物成像、图像处理到计算机建模等学科的新科学家。

Many complex systems, such as biological organisms, the internet or the economy, involve integrated parallel processes operating at many different scales. Organisms for example, exhibit massively parallel organisation, with units interacting at levels ranging from molecules to ecosystems. A key challenge is to understand how such parallel multiscale interactions work. In the case of biological systems, a major focus in recent years has been on the molecular to cellular scale. However, advances in developmental biology and imaging, in particular the ability to isolate and study genes affecting the development of organisms in 3D, raises the possibility of a systems approach at a larger scale, from the subcellular to organism levels. The key link between these scales is growth, as it is through growth that molecular changes lead to modifications in cell size, shape and division, eventually leading to production of multicellular tissues and organisms. Growing structures can be divided into those that reach a steady state, such as the growing tips of plants, and those in which shape continually changes, such as developing insect wings or plant leaves. The latter cases present the particular challenge of capturing multiple interactions as a structure undergoes changes in shape and size over several orders of magnitude. Studying this problem requires growth, shape, patterning and gene action to be analysed and integrated at many levels. Moreover formal languages and modelling frameworks need to be developed to allow the dynamics and interactions to be captured. By pursuing these approaches at many scales in parallel, key experimental and conceptual links should emerge that would not be evident from studying each scale in isolation. Recent advances in both experimental and computational science now make this feasible. The aim of this project is to provide a platform for such an approach by studying the mechanism by which leaves acquire their characteristic shapes and patterns. The project will involve a combination experimental analysis, image-processing, computer modelling and integration at four different scales. (1) At the subcellular level, we will track the way components of the cytoskeleton are synthesised and change in parallel over time in 3D. Models will then be constructed to account for this behaviour and tested through further rounds of experimentation. These models will also be related to cellular properties. (2) At the cellular level, the pattern of cell growth and division will be determined for multiple regions of a growing leaf by live 3D imaging. Computer languages will be developed for modelling this behaviour based on local interactions between growing cells. Models will be tested by modifying gene activity at particular places and times during leaf development. These models will also be related to properties at the whole organ level. (3) At the organ level, the growth of the leaf in 3D will be tracked using a variety of imaging methods. A modelling framework will be developed that allows the observed cellular and tissue properties to be integrated with gene action. Models will be tested by analysing how experimental manipulation of local gene activity modifies growth. This will be aided by developing a system for measuring the 3D shapes of a diverse collection of mutants that affect leaf shape at various stages of development. (4) At the whole plant level, leaf growth will be incorporated into a virtual plant in which local interactions between modules account for the dynamics of growth and architecture. By interfacing the various models, an integrated multiscale view of plant development should emerge. The project will also train a new cohort of scientists familiar with disciplines ranging from molecular genetics, developmental biology, bio-imaging, image-processing to computer modelling.