Tatsunari Sakurai

I n our original manuscript, the parameter values were incorrectly described in the caption of Figure 5. The parameters should be “au/a = 1 and av/a = 0.1” instead of “αu/α = αv/α = 0.1”. In addition, “au = av = 3.7” was incorrect, and it should be “au = 37 and av = 3.7”. The errors are typographical, and thus, these corrections do not affect the conclusions of this work.

Collective cell migration is seen in many developmental and pathological processes, such as morphogenesis, wound closure, and cancer metastasis. When a fish scale is detached and adhered to a substrate, epithelial keratocyte sheets crawl out from it, building a semicircular pattern. All the keratocytes at the leading edge of the sheet have a single lamellipodium, and are interconnected with each other via actomyosin cables. The leading edge of the sheet becomes gradually longer as it crawls out from the scale, regardless of the cell-to-cell connections. In this study, we found leading-edge elongation to be realized by the interruption of follower cells into the leading edge. The follower cell and the two adjacent leader cells are first connected by newly emerging actomyosin cables. Then, the contractile forces along the cables bring the follower cell forward to make it a leader cell. Finally, the original cables between the two leader cells are stretched to tear by the interruption and the lamellipodium extension from the new leader cell. This unique actomyosin-cable reconnection between a follower cell and adjacent leaders offers insights into the mechanisms of collective cell migration.

A concentric pulse by motile cells of Escherichia coli (E. coli) propagates and the cells aggregate to form self-organized patterns.We summarize experimental and numerical results on the self-organized pattern formation of E. coli to elucidate some aspects of its mechanism. Our presentation includes experiments on E. coli patterns, as well as numerical simulations on the basis of a reaction-diffusionchemotaxis model.We find good agreement for one-dimensional propagating fronts in observation and simulation. However, corresponding results for two-dimensional circular bacterial clusters have still not been obtained.

We investigated the phase-response curve of a coupled system of density oscillators with an analytical approach. The behaviors of two-, three-, and four-coupled systems seen in the experiments were reproduced by the model considering the phase-response curve. Especially in a four-coupled system, the clustering state and its incidence rate as functions of the coupling strength are well reproduced with this approach. Moreover, we confirmed that the shape of the phase-response curve we obtained analytically was close to that observed in the experiment where a perturbation is added to a single-density oscillator. We expect that this approach to obtaining the phase-response curve is general in the sense that it could be applied to coupled systems of other oscillators such as electrical-circuit oscillators, metronomes, and so on.