Attractive internuclear force drives the collective behavior of nuclear arrays in Drosophila embryos

Abstract

The collective behavior of the nuclear array in Drosophila embryos during nuclear cycle (NC) 11 to NC14 is crucial in controlling cell size, establishing developmental patterns, and coordinating morphogenesis. After live imaging on Drosophila embryos with light sheet microscopy, we extract the nuclear trajectory, speed, and internuclear distance with an automatic nuclear tracing method. We find that the nuclear speed shows a period of standing waves along the anterior-posterior (AP) axis after each metaphase as the nuclei collectively migrate towards the embryo poles and partially move back. And the maximum nuclear speed dampens by 28-45% in the second half of the standing wave. Moreover, the nuclear density is 22–42% lower in the pole region than the middle of the embryo during the interphase of NC12-14. To find mechanical rules controlling the collective motion and packing patterns of the nuclear array, we use a deep neural network (DNN) to learn the underlying force field from data. We apply the learned spatiotemporal attractive force field in the simulations with a particle-based model. And the simulations recapitulate nearly all the observed characteristic collective behaviors of nuclear arrays in Drosophila embryos.

MorphoSim: an efficient and scalable phase-field framework for accurately simulating multicellular morphologies

“…Despite that MorphoSim can simulate over one hundred interacting cells, there is still a great need from the field to keep increasing the cell number and decreasing the computational cost. For example, the early Drosophila embryogenesis involves thousands of cells and was usually simulated using the coarse-grained model or vertex model 67 , 68 . On the one hand, the limited GPU memory hampers mesh enlargement/densification when considering a large number of cells.…”