Genetic redundancy caused by gene duplications and its evolution in networks of transcriptional regulators

Abstract

In various organisms loss-of-function mutations of individual genes with unexpectedly weak or no phenotypic effects in the homozygous state have been observed. In several of these case, independent evidence shows that the respective gene products do have essential biological functions. An explanation emerging from detailed biochemical and genetic studies on such genes is that two or more genetically redundant genes contribute to that function, i.e., a group of genes that is able to substitute partially for a loss of function in one member of that group. The often-observed sequence similarity among redundant genes suggests gene duplications as a frequent source of genetic redundancy. Aside from this observation, the evolution of genetic redundancy is poorly understood. Genetic redundancy is potentially of great relevance to organismal evolution, since it may (i) 'protect' organisms from potentially harmful mutations, and (ii) maintain pools of functionally similar, yet diverse gene products, and thus represent a source of evolutionary novelty at the biochemical level. The question of how genetic redundancy evolves should ideally be answered by experimentation. However, the large time scales involved and insufficient quantitative understanding of the underlying regulatory pathways are likely to preclude such an approach in the foreseeable future. Preliminary answers are sought here by using a biochemically motivated model of a small but central part of a developmental pathway. Sets of transcription regulators are modeled that mutually regulate each other's expression and thereby form stable gene expression patterns. It is then studied how genetic redundancy caused by gene duplications might evolve in such networks. The results obtained suggest that redundancy may, at least in some cases, be a global property of gene interactions within a regulatory pathway, rather than a local property of genes in that pathway. They also raise the possibility that duplications of a whole regulatory gene network, as may have taken place during the evolution of HOM/Hox genes in chordates, are less likely to be reversible (by gene deletions) than duplications of individual network genes. These findings are discussed with reference to experimental evidence on the evolution of HOM/Hox genes.