PARKER, S.C.J.; daSILVA, J.; STELLWAG, E.J.*: Computational Method for Identification of Functionally Significant Non-Coding Sequences: Studies of Hox Cluster Organization

A question of considerable interest to evolutionary developmental biologists is the linkage between gene architecture and spatio-temporal expression patterns during embryogenesis. The prevailing hypothesis is that this linkage results from a network of interacting cis-regulatory elements that compete for and share the transcriptional regulatory apparatus of clustered genes. There are two important corollaries of this hypothesis that we are addressing. 1) If the clustered architecture of the Hox complex is dependent on the sharing and competition between cis-regulatory elements, then the elements should be functionally significant and undergoing purifying selection. 2) The network of regulatory elements within each cluster should be sufficiently extensive to account for the evolutionary maintenance of linkage between the clustered architecture and spatio-temporal expression patterns. We have developed a computational method to identify functionally significant non-coding seqeunces; i.e. sequences under purifying selection and used it to examine non-coding sequences both upstream and downstream of the Hoxb7a pseudogene in teleosts. Our results demonstrate the occurrence of an extensive network of intergenic sequence domains undergoing purifying selection in the neighborhood of the Hoxb7a pseudogene. The significance of the sequence domains identified is underscored by the fact that one of the domains identified in our analysis has previously been shown to be present on all Hox clusters in the same relative position. We believe our method provides an opportunity to use comparative genomic analyses to reveal functionally significant non-coding sequences.