Identification of cis-regulatory elements by large-scale Comparative Sequencing. B.D. Pfeiffer , R. Hoskins , T. Agbayani , K. Wan , B. Kronmiller , G. Rubin , S.E. Celniker. Berkeley Drosophila Genome Project, University of California, Berkeley, Lawrence Berkeley National Laboratory, Berkeley, CA 94720.

   We have initiated a comparative sequencing project to identify conserved cis-regulatory elements in the Drosophilagenome. Previous studies have established the utility of comparative sequence analysis for the discovery of enhancers. Introns, 3 UTRs, and 5 flanking regions, diverge more rapidly than coding sequences.We are addressing two questions. 1) Which fly species is a suitable evolutionary distance from D. melanogastersuch that comparative sequence analysis will allow us to distinguish functional cis-regulatory elements from background? and 2) Does more than one additional species need to be sequenced to identify such cis-elements? To answer these questions, four species,D. erecta, D. pseudoobscura, D. willistoniand D. littoralis(a close relative of D. virilis), that have diverged from D. melanogasterby 15, 25-30, 25-30, and 40-60 million years, respectively, were chosen for sequence comparison. The identity of each species was verified by David Grimaldi at the Museum of Natural History, NY. Fosmid libraries (37 kb average insert size) were constructed by A. Gnirke at Exelixis. Clones were gridded and filters representing 5-genome equivalents were produced at BACPAC Resources (Oakland, CA). Ten genes whose regulatory sequences have been well defined in Drosophila melanogasterwere chosen for a pilot study; even-skipped, fushi tarazu, twist, rhomboid, apterous, Abdominal-B,and four rhodopsins. Degenerate primers were designed from a coding exon for each gene and used for genomic PCR. The genomic PCR product was sequenced and species-specific short double-stranded oligonucleotide probes (40 bp overgos) were designed and used to screen each library. Fosmids containing each gene are being sequenced. We will present preliminary sequence comparisons and conclusions on the selection of a second species for whole genome shotgun sequencing.