Understanding and Exploiting the Evolution of Drosophila Regulatory Sequences

摘要

My lab is examining variation in Drosophila regulatory sequences over short (within species),intermediate (within genus) and long (between family) evolutionary time scales and evaluating the effects of this variation on transcription factor binding and gene regulation. Our goals are to dissect the molecular basis for variation in gene expression and to better understand the constraints on regulatory sequence evolution imposed by selection to maintain expression output. My presentation will draw from several ongoing projects.Extensive rearrangement, but functional conservation, of enhancers between Drosophila,Sepsidae and Tephritidae: We have been comparing the sequences and activities of earlyembryonic enhancers from D. melanogaster to their orthologs from distantly related fly species (families Sepsidae and Tephritidae) that share basic patterning mechanisms with Drosophila. Although there is almost no detectable sequence similarity between Drosophila and sepsid and tephritid non-coding DNA, their enhancers (which we identify computationally)recapitulate endogenous gene expression patterns in transgenic D. melanogaster embryos. This extreme sequence variability without functional variation highlights the extreme flexibility of the molecular machinery that regulates gene expression. Our results suggest that enhancers are organized into functional submodules with tight constraints on their structure,but that these submodules can be extensively rearranged without compromising enhancer activity.Comparison of tephritid genomes to identify Drosophila enhancers: The tephritid sequence data turned out to be useful in a very unexpected way. The genomes of the species we are analyzing (including the medfly Ceratitis capitata, and species in the genera Bactrocera and Rhogoletis) are significantly bigger (600 -900 Mb) than those of most Drosophila species.Unlike in Drosophila, where there is near ubiquitous conservation of non-coding DNA, comparison of tephritid genomes reveals many blocks of conserved non-coding sequence flanked by large stretches of poorly conserved sequence. This landscape of non-coding conservation is similar to that observed in vertebrates, and immediately suggested to us that comparisons of tephritid genomes could be used to identify regulatory sequences with the same effectiveness as in vertebrates (and which, because of the lack of non-conserved sequences, have been much less effective in Drosophila). We tested the regulatory activity of ~20 conserved noncoding sequences from the tephritid orthologs of the well-characterized D. melanogaster genes, and found that a substantial fraction have enhancer activity. We conclude that the relative ineffectiveness of comparative genomics methods to identify invertebrate regulatory sequences was an artifact of the relatively small size of sequenced invertebrate genomes, and suggest -somewhat paradoxically -that comparison of tephritid genomes may be a more effective way of identifying Drosophila enhancers than comparison of Drosophila genomes.Experimental characterization of transcription factor binding in D. melanogaster and D. yakuba: To better understand the biochemical basis for enhancer, we have been using ChiP-chip and ChlP-seq along with high-resolution imaging to examine how DNA binding and gene expression vary with sequence variation between and within Drosophila species. The most striking observation from our study of the binding of 20 transcription factors involved in early embryonic patterning, is the large number of sites bound by each factor in developing embryos(often well in excess of 10,000). This observation and several accompanying computational analyses suggest that a large fraction of this binding does not affect transcription or contribute to organismal fitness. The leaves us with the challenge of developing methods to identify which binding events are likely to be functional, something we are addressing by examining binding variation with D. melanogaster and between D. melanogaster and D. yakuba. Early data from D. yakuba show that, against the backdrop of strong conservation of binding in many well-characterized loci, there is extensive variation in binding between the two species -often -though not always -coupled to the gain/loss of transcription factor binding sites. We are now using this data to develop a better understanding of the sequence determinants of DNA binding, and the relationship between transcription factor binding and regulatory activity.