Genome size as a mutation-selection-drift process

A novel method for estimating neutral rates and patterns of DNA evolution i n Drosophila takes advantage of the propensity of non-LTR retrotransposable elements to create nonfunctional, transpositionally inactive copies as a p roduct of transposition. For many LINE elements, most copies present in a g enome at any one time are nonfunctional "dead-on-arrival" (DOA) copies. Bec ause these are offshoots of active, transpositionally competent "master" li neages, in a gene tree of a LINE element from multiple samples from related species, the DOA lineages are expected to map to the terminal branches and the active lineages to the internal branches, the primary exceptions being when the sample includes DOA copies that are allelic or orthologous. Analy sis of nucleotide substitutions and other changes along the terminal branch es therefore allows estimation of the fixation process in the DOA copies, w hich are unconstrained with respect to protein coding; and under selective neutrality, the fixation process estimates the underlying mutational patter n. We have studied the retroelement Helena in Drosophila. An unexpectedly h igh rate of DNA loss was observed, yielding a half-life of unconstrained DN A sequences approximately 60-fold faster in Drosophila than in mammals. The high rate of DNA loss suggests a straightforward explanation of the seemin g paradox that Drosophila has many fewer pseudogenes than found in mammalia n species. Differential rates of deletion in different taxa might also cont ribute to the celebrated C-value paradox of why some closely related organi sms can have very different DNA contents, New data presented here rule out the possibility that the transposition process itself is highly mutagenic, hence the observed linear relation between number of deletions and number o f nucleotide substitutions is most easily explained by the hypothesis that both types of changes accumulate in unconstrained sequences over time.