Of clocks and waves, stripes and shapes: evo-devo models of sequential segmentation

Abstract

Segments are repeated elements along the main body axis of many animals, such as the rings of earthworms and myriapods or the vertebrae of mammals. These structures are produced very early in development. Typically, they are produced one by one from a posterior growth zone that starts just behind the head and then retreats to form the body and tail – this is called sequential segmentation. Before segments take shape, the boundaries between segments (and between the different parts of each segment) are defined by a striped pattern of different segmentation genes. In insects and vertebrates, these stripes are layed down via the interaction between a molecular clock and a morphogen gradient in the tissue that is to be segmented. The clock consists of multiple interacting genes, whose expression alternates between high and low: the period of these oscillations determines the pace of segment formation. The morphogen has a high concentration in the growth zone, but decays in the rest of the tissue so that its concentration is lower near the already-formed segments. When the concentration of the morphogen is low enough, the clock oscillations are transformed into the striped segment pattern. The mechanism that causes this transformation is still debated, and also the evolutionary origin is still unclear. In this thesis, we used computer models to study the evolution and mechanism of sequential segmentation. First, we focused on specific properties of segmentation in vertebrates. We studied what caused their peculiar clock dynamics, where gene expression oscillations are faster in the growth zone than in the tissue closer to the formed segments: this is called a frequency gradient. We found that a far-reaching, shallow morphogen gradient in combination with noisy gene expression more often causes the in silico evolution of frequency gradients, and may therefore have played a major role in vertebrate segment evolution. In another study, we investigated differences in the segmentation mechanism between vertebrates; for instance the role of the frequency gradient differs. We showed that these differences, which normally don’t result in qualitatively different phenotypes, are responsible for the different phenotypes when one of the governing morphogens is removed. We conclude that the mechanistic differences between species are not completely neutral. In the second part of the thesis we focused on the interaction between tissue growth and the segment pattern. We showed that in order to evolve sequential segmentation at all (i.e. a posterior growth zone which produces segments), a posterior morphogen gradient already needs to be present to give directionality to the tissue. Furthermore, strong selection for a finite tissue size reduces the likelihood of evolving sequential segmentation. Finally we studied how the formation of the main body axis interacts with a pre-existing segment pattern. We show that the cellular movement associated with tissue shaping may disrupt the segment pattern. This can be solved by segment-specific adhesion, in which cells adhere more strongly to cells of the same segment. Strikingly, this type of differential adhesion can also drive tissue shaping itself. In short, our results give new insights into the evolution and the mechanism of sequential segmentation, and provides new directions for future research.