Patterns of crossover distribution in <i>Drosophila mauritiana</i> necessitate a re-thinking of the centromere effect on crossing over

We present an SNP-based crossover map for Drosophila mauritiana. Using females derived by crossing 2 different strains of D. mauritiana, we analyzed crossing over on all 5 major chromosome arms. Analysis of 105 male progeny allowed us to identify 327 crossover chromatids bearing single, double, or triple crossover events, representing 398 crossover events. We mapped the crossovers along these 5 chromosome arms using a genome sequence map that includes the euchromatin-heterochromatin boundary. Confirming previous studies, we show that the overall crossover frequency in D. mauritiana is higher than is seen in Drosophila melanogaster. Much of the increase in exchange frequency in D. mauritiana is due to a greatly diminished centromere effect. Using larval neuroblast metaphases from D. mauritiana-D. melanogaster hybrids we show that the lengths of the pericentromeric heterochromatin do not differ substantially between the species, and thus cannot explain the observed differences in crossover distribution. Using a new and robust maximum likelihood estimation tool for obtaining Weinstein tetrad distributions, we observed an increase in bivalents with 2 or more crossovers when compared with D. melanogaster. This increase in crossing over along the arms of D. mauritiana likely reflects an expansion of the crossover-available euchromatin caused by a difference in the strength of the centromere effect. The crossover pattern in D. mauritiana conflicts with the commonly accepted view of centromeres as strong polar suppressors of exchange (whose intensity is buffered by sequence nonspecific heterochromatin) and demonstrates the importance of expanding such studies into other species of Drosophila.