Abstract
We have sequenced the hsp82 heat shock gene and 5'-flanking DNA of four species of Drosophila in order to identify conserved features of possible regulatory significance and to determine the nature and rate of evolutionary change in various domains of the gene. All conserved sequences identified in the 5' non-transcribed region of the hsp82 gene of melanogaster, simulans, pseudoobscura and virilis lie within 150 base-pairs of the RNA initiation site. These include a 34 base-pair imperfect but highly conserved dyad made up of three overlapping copies of the consensus heat shock regulatory sequence dyad made up of three overlapping copies of the consensus heat shock regulatory sequence C-T-N-G-A-A-N-N-T-T-C-N-A-G. Several other highly conserved features are also seen. In pseudoobscura only, the region from -21 to -133 is almost exactly repeated 698 base-pairs upstream. The upstream repeat retains all of the sequences common to the four species in the -21 to -133 interval except for the T-A-T-A motif. One possibility is that this upstream sequence is involved in dosage compensation of the hsp82 gene, which is pseudoobscura is on the X chromosome. In melanogaster, simulans and virilis there is an oppositely oriented, non-heat shock gene 0.6 to 0.8 kb (kb = 10(3) base-pairs) upstream from the hsp82 transcription initiation site. Comparisons of the hsp82 transcription unit from the distantly related species show little or no sequence conservation in the intron and non-translated exon I sequences. In contrast, the coding regions of these species are 90% homologous at the DNA level and 97 to 99% identical at the amino acid level. Half of the amino acid changes have occurred within a 15-amino acid region that lies within an unusually polar domain of hsp82. From the sequence comparisons, we estimate the rate of silent (synonymous) site substitution in Drosophila as 1 X 10(-8)/nucleotide site per year, similar to that for mammals.
