Asymmetrical distribution of non-conserved regulatory sequences at PHOX2B is reflected at the ENCODE loci and illuminates a possible genome-wide trend

BACKGROUND: Transcriptional regulatory elements are central to development and interspecific phenotypic variation. Current regulatory element prediction tools rely heavily upon conservation for prediction of putative elements. Recent in vitro observations from the ENCODE project combined with in vivo analyses at the zebrafish phox2b locus suggests that a significant fraction of regulatory elements may fall below commonly applied metrics of conservation. We propose to explore these observations in vivo at the human PHOX2B locus, and also evaluate the potential evidence for genome-wide applicability of these observations through a novel analysis of extant data. RESULTS: Transposon-based transgenic analysis utilizing a tiling path proximal to human PHOX2B in zebrafish recapitulates the observations at the zebrafish phox2b locus of both conserved and non-conserved regulatory elements. Analysis of human sequences conserved with previously identified zebrafish phox2b regulatory elements demonstrates that the orthologous sequences exhibit overlapping regulatory control. Additionally, analysis of non-conserved sequences scattered over 135 kb 5' to PHOX2B, provides evidence of non-conserved regulatory elements positively biased with close proximity to the gene. Furthermore, we provide a novel analysis of data from the ENCODE project, finding a non-uniform distribution of regulatory elements consistent with our in vivo observations at PHOX2B. These observations remain largely unchanged when one accounts for the sequence repeat content of the assayed intervals, when the intervals are sub-classified by biological role (developmental versus non-developmental), or by gene density (gene desert versus non-gene desert). CONCLUSION: While regulatory elements frequently display evidence of evolutionary conservation, a fraction appears to be undetected by current metrics of conservation. In vivo observations at the PHOX2B locus, supported by our analyses of in vitro data from the ENCODE project, suggest that the risk of excluding non-conserved sequences in a search for regulatory elements may decrease as distance from the gene increases. Our data combined with the ENCODE data suggests that this may represent a genome wide trend.