Segmentation and identity-domain patterning in body plans

Ten Tusscher and Hogeweg (2011) modelled body plan patterning by selecting independently for segmentation and for differentiation. They chose this fitness criterion because it invites modular networks to evolve.¹ Unlike the SS network, the SF minimum networks could generate either segments or domains autonomously and independently.² The correspondence was imperfect. The minimum segment network made 11 segments rather than 12, and its first segment was too wide.³ Some identity domains also depended on the segmentation process itself.²

In vertebrates, segment pattern is linked to oscillating gene expression. Grodstein, McMillen and Levin (2023) describe the clock as having originated in the clock-and-wavefront model. They note that it typically requires a global clock, though some versions avoid this.⁵ They also say that many details, including the reasons for its robustness, remain unknown.⁶ Pai and colleagues (2012) noted that the transcriptional waves driving the somitogenesis clock differ consistently in timing between the two sides of the body.⁴

McMillen and Levin (2024) treat the segmentation clock as an example of collective intelligence. They cite grafting experiments in which wild type cells placed in non-cycling mutant fish express her1 but do not cycle.⁷ On this reading, oscillation depends on the cells acting together and not on each cell alone. Grodstein and colleagues offer a contrast. Their own model uses waves for measurement, within a negative feedback system that needs no clock.⁶

Mondia and colleagues (2011) looked at how damage position along the head-to-tail axis affects tail regeneration. They measured shape differences by Procrustes distance. The difference was proportional to how close the insult was to the head.⁸ Their conclusion was that long-distance signals matter for the shape of the regenerate. Damage closer to the head changed the shape of the regenerate more.