OSTMAN, Bjorn*; OAKLEY, Todd H.; University of California, Santa Barbara: Evolutionary rates of loss and gain of gene expression

The importance of gene regulatory elements for evolutionary events has been largely overshadowed by a focus on the evolution of protein-coding sequences. Nevertheless, regulatory elements should be important for evolution, in part due to their short length (5-20 base pairs), which often could allow mutations to change gene expression in a meaningful way (non-neutral, non-deleterious). As examples of how to study gene expression evolution using genomic data we will present results of two hypothesis tests. First, since the loss of transcription is mechanistically easier than gain within an organism, we tested the hypothesis that such a mechanistic bias results in rates of expression loss that are greater than gain during evolution. Secondly, we tested whether the evolutionary rate of gain of expression within the same germ layer is higher than the rate of gain from one germ-layer to another. For these tests, we used a database of non-redundant Drosophila ESTs associated with annotated in situ hybridization images from Drosophila embryos of stage 13-16. The ESTs were grouped into gene families, and phylogenies were established for each gene family. Evolutionary rate parameters were then estimated using maximum likelihood, while treating gene expression patterns as evolving phenotypes of genes. As predicted, preliminary results indicate that the relative rate of loss of an expression domain is higher than the rate of gain, and that relative rates of domain-gain within a germ-layer (ecto-, meso-, or endoderm) are higher than the rate of gain in new germ-layers.