The two most fundamental processes describing change in biology – development and evolution – occur over drastically different timescales, difficult to reconcile within a unified framework. Development involves a temporal sequence of cell states controlled by a hierarchy of regulatory structures. It occurs over the lifetime of a single individual, and is associated to the gene expression level change of a given genotype. Evolution, by contrast entails genotypic change through the acquisition or loss of genes, and involves the emergence of new, environmentally selected phenotypes over the lifetimes of many individuals. Here we present a model of regulatory network evolution that accounts for both timescales. We extend the framework of boolean models of gene regulatory network (GRN) – currently only applicable to describing development – to include evolutionary processes. As opposed to one-to-one maps to specific attractors, we identify the phenotypes of the cells as the relevant macrostates of the GRN. A phenotype may now correspond to multiple attractors, and its formal definition no longer require a fixed size for the genotype. This opens the possibility for a quantitative study of the phenotypic change of a genotype, which is itself changing over evolutionary timescales. We show how the realization of specific phenotypes can be controlled by gene duplication events, and how successive events of gene duplication lead to new regulatory structures via selection. It is these structures that enable control of macroscale patterning, as in development. The proposed framework therefore provides a mechanistic explanation for the emergence of regulatory structures controlling development over evolutionary time.
