| Summary: | The evolution and development of multicellular organisms requires cells to differentiate, interact and "collaborate". Our understanding of the molecular mechanisms is still hazy. In this dissertation mathematical modelling is used to integrate available experimental data and to make testable predictions about such mechanisms. The thesis is split into three parts, each of which addresses one of the three challenges: differentiation, adhesion and collaboration. In the first part, a mathematical model is developed to explain how, in the absence of polarizing cues from the environment, sister cells with identical genomes can follow distinct routes of differentiation. It is shown that difference in cell size, resulting from asymmetric cell division, is sufficient to induce differential cell fate in Bacillus subtilis. The model predicts that this effect depends on the allosteric behaviour of a kinase and the low catalytic rate of the corresponding phosphatase; both properties were subsequently confirmed in experiments. During the development of multicellular organisms, differentiation can arise in response to gradients. By example of dorso-ventral patterning it is shown how a shallow maternal gradient can be converted into a sharp pattern. In the second part, a model for cell adhesion via integrins is developed, and it is shown that, for physiological parameters, binding of a ligand and of a stabilizing factor such as talin are insufficient for ligand-dependent integrin activation, and that a positive signaling feedback is required. In the final part, antibody affinity maturation is studied as an example for division of labour between collaborating cells. A novel B cell selection mechanism, based on competition for T cell help rather than for antigen, is proposed and shown to reconcile heretofore inexplicable experimental observations. Such a mechanism requires B cells to discriminate among different affinities of binding, and it is further shown that this can be achieved if B cell signaling is initiated by antigen-dependent receptor-inhibitor segregation. The predictions of the model match experimental measurements quantitatively.
|
|---|
