Use of P-element-mediated transformation to identify the molecular basis of naturally occurring variants affecting...

The purpose of the work reported here is to identify the molecular basis of the difference in level of expression between the polymorphic Slow and Fast alcohol dehydrogenase (Adh) alleles in Drosophila melanogaster. Previous studies have shown that Fast lines typically have a two- to threefold higher activity level than Slow lines and they also have a substantially higher level of ADH-protein (estimated immunologically). The results of a restriction fragment length polymorphism study in relation to ADH activity variation had previously suggested that the difference in Adh expression between allozymes might not be due entirely to the amino acid replacement substitution, but could be due in part to linkage disequilibrium with a regulatory site polymorphism. Here we describe an approach that makes use of P-element-mediated transformation in order to identify the nucleotide substitution(s) responsible for this difference in ADH level. This approach consists of generating recombinants in vitro between Adh region clones derived from a typical Slow/Fast pair of alleles and then testing for the effects of particular restriction fragments on expression in vivo by transformation. Using this approach, the effect on both ADH activity and ADH-protein level clearly maps to a 2.3-kb restriction fragment that includes all of the Adh coding sequence and some intron and 3' flanking sequence, but excludes all of the 5' flanking sequence of the distal (adult) transcriptional unit. Comparison of Kreitman's DNA sequences for this fragment from several Slow and Fast alleles showing the typical difference in activity level shows that only three nucleotide substitutions distinguish all Fast from all Slow alleles. Thus, it is likely that one or more of these substitutions causes the major difference in Adh expression between allozymic classes. One of these substitutions is, of course, the Slow/Fast amino acid replacement substitution (at 1490) while the other two are nearby third position silent substitutions (at 1443 and 1527). A quantitative analysis of variation among transformant stocks shows that the P-element transformation approach can be used to localize even relatively small effects on gene expression (on the order of 20%).