Mobile DNA and genetic variation in Drosophila melanogaster

Whilst it is commonly accepted that transposable elements can generate genetic variation, the significance of this for the maintenance and dissemination of such elements is controversial. Here long-term laboratory populations of Drosophila melanogaster, maintained at two discrete temperatures, are screened by Southern blotting for the patterns of insertion of several transposable elements (copia, mdg-2, mdg-4 and P). Consistent with a temperature-specific adaptive role for some insertions, several are apparently found at higher frequency in lines at one temperature.;Further characterisation of these putatively temperature-selected insertions was attempted. Of three distinct approaches taken towards cloning these insertions, a single-stranded DNA-ligation for PCR-amplification technique, not thought to have been previously exploited for isolating transposable element insertion sites, generated the best results. One copia insertion was successfully sequenced, although single fly PCR experiments suggested that the frequency of this in caged populations was not related to temperature.;A major collection of D. melanogaster from the French and Spanish Pyrenees was undertaken along four discrete altitudinal clines, with a view to screening for specific transposable element insertions. A novel strategy was developed for correlation of altitude with mean seasonal temperature of each collection site. Altitudinal variation in the frequency of Thr-Gly length polymorphism at the period locus was found to be consistent with predictions based on known latitudinal clines. This is the first known example of putatively adaptive clinal genetic variation in European populations of Drosophila collected along altitudinal transects.;Finally, a fundamental re-examination of the theoretical issues surrounding the possible adaptive significance of transposable elements is developed. It is demonstrated that he extension of ideas on the 'units of selection' to transposable elements has led to confusion. A model of transposable element evolution which presents a coherent alternative to the selfish DNA approach is presented.