Invaluable progress has been made last decades in the molecular, genetic and cellular characterization of morphogenetic processes. Yet, the precise physical processes governing the shape and dynamics of cells in early embryos remain poorly characterized. The laboratory is developing theoretical models of morphogenesis, combining physics, mechanics and advanced numerical simulations.
To understand how morphology controls biological functions, and, ultimately, how early embryos self-organize, we aim at integrating molecular, cellular and multicellular levels of description (see Figure 1) into a new and versatile simulation framework for multicellular morphogenesis: Virtual Embryo.
As elegantly illustrated by D’Arcy Thompson in the early 20th century, cells in early embryos adopt spatial configurations remarkably similar to soap bubbles, an analogy which can be drawn from the concept of surface tension. In animal cells, the surface tension is mainly provided by the contractile forces generated by myosin molecular motors within the actomyosin cortex, a thin layer of polymeric actin filaments beneath the plasma membrane. In contrast to passive objects like bubbles, the tension of the cortex is actively regulated in space & time by several biochemical pathways, such as the RhoA signaling cascade. The cortical cell surface forms hence the main force-organizing center that controls cells shape in early embryos. Yet, its complex rheological features remain poorly understood, and in particular in relation to the microscopic properties of its constituents. From a biological perspective, the tools available to perturb the cortex operate directly at the molecular level (drugs, genetic engineering etc...). A first and critical step to better relate experiments to quantitative models of cell morphogenesis is therefore to characterize how molecular players in the cortex (actin filaments, myosin motors, crosslinkers etc...) control its coarse-grained physical properties.
At the mesoscopic level, the recent active gel hydrodynamic theories have proven their efficiency in capturing the essential physics of various actomyosin-based cell processes. To study the drastic deformations undergone by a cell over cytokinesis, the last step of cell division, we developed pioneering numerical simulations, in axisymmetric non-linear geometry, where the cortex is represented as a thin layer of active gel, generating visco-active surface tensions (see Video 1 and Figure 3). The mechanics of the cortical surface is itself regulated by various signaling molecules, which exhibit a complex surface dynamics, involving adsorption, desorption, diffusion, advection by cortical flows, and dilution/accumulation upon surface deformation. These mechano-chemical couplings make the surface of cells complex self-organizing platforms, where mechanics not only translates biochemical information into morphology, but also feedbacks onto molecular expression to actively control cell's function. Such feedbacks are particularly important in processes such as cell division & polarization, that we aim to better understand through the development of powerful simulation tools in 3D.
At the multicellular level, morphogenesis is furthermore regulated by mechanical interaction & biochemical communication between cells, and by external mechanical constraints. In particular, the interplay between cell contractility, cell-cell adhesion, molecular expression and fate specification remains poorly understood in early embryos and small tissues. To identify and understand the minimal self-organization principles driving multicellular morphogenetic processes, it is essential to develop realistic 4D models of interacting cells, offering general & accurate description of cell surface mechanics but also complemented by versatile options to model surface signaling dynamics and simple gene networks regulation (see Figure 4).
Such a multiscale approach is particularly essential to understand the self-organization and remarkable regulative capacities of early mammalian blastocysts. Combining numerical simulations with physical modeling, we expect to unveil, in only a few years from now, pioneering simulations of pre-implantation mammalian embryos in toto, from the zygote to the blastocyst.
Dr. Jean-Léon Maître, Institut Curie.