Transcriptional noise as a driver of gene evolution

As novel genes emerge in the evolution of species, pre-existing genes expand their expression patterns to diversify their functions and the expression patterns of gene duplicates diverge to pursue functional specialization. All these processes require genes to be expressed, however, the level and specificity of gene expression at the early stages of these processes are unclear. In this study, I propose that transcriptional noise is a mechanism to test genes for new functions, and I hypothesize the 'in-service' mechanism of gene evolution. In contrast to other hypotheses that suggest that there are specialized sites for gene evolution, such as tumors (Kozlov, 2010) or the testis (Kaessmann, 2010) this hypothesis proposes that emerging genes are expressed nonspecifically in many normal tissues, due to transcriptional noise. New genes are continuously 'tested' in various cells and under various conditions, thereby allowing the genes to evolve functions at the sites of their future work. The hypothesis of 'in-service' gene evolution also proposes that pre-existing genes are continuously tested under extrinsic conditions, due to transcriptional noise; this testing facilitates the emergence of alternative promoters and the diversification of the genes' expression patterns and functions.

► Two current hypotheses on evolution of novel genes are briefly reviewed. ► Origin and function of transcriptional noise are briefly reviewed. ► The 'in-service' evolution of genes (noise-driven evolution) is hypothesized. ► It is discussed in relation to new and pre-existing genes.