Crossover rate is known to vary at broad scales in all species, typically being low in pericentromeric regions. In yeast and mammals, fine-scale crossover rate heterogeneity has also been thoroughly documented, with very fine "hotspots" of high recombination interspersed amidst regions of much lower recombination rate. Based primarily on research on intragenic crossing over within the rosy gene and patterns of linkage disequilibrium (rapidly decaying), many had assumed that Drosophila do not bear such fine-scale crossover rate heterogeneity. As such, investigations of recombination rate's association with polymorphism, divergence, or codon bias frequently used "smoothed" estimates of recombination rate.
In 2006, rotation student in our laboratory investigated a 2-megabase region on the XL chromosome of Drosophila pseudoobscura to look for evidence of fine-scale crossover rate heterogeneity. She found that there was significant heterogeneity (>30-fold between hottest and coldest regions), this heterogeneity was significantly associated with (CA/GT)n repeats, and this heterogeneity was significantly associated with differences in codon bias among genes within these intervals (see abstract1). Followup work by another rotation student documented similar patterns on the XR chromosome.
More recently, we examined fine-scale crossover rate heterogeneity across all of the Drosophila pseudoobscura second chromosome using 454 genome sequences (for SNP discovery) and microbead arrays (for genotyping)(see abstract2). We are currently examining this same chromosome in sister species D. persimilis to see if the same recombinational "hotspots" persist between species. We are also pursuing related work examining factors that affect recombination, including one gene that is also associated with cancer in humans. This work is ongoing.
Genomic patterns of nucleotide diversity and divergence in Drosophila
One of the most influential findings in molecular evolutionary genetics has been the repeated association of crossover rate with 
nucleotide diversity within species. An elegant report of this association in Drosophila was published in 1992 by Begun and Aquadro, and they explained this pattern as resulting from the ubiquitous action of natural selection. Part of this interpretation came from the lack of a similar association between crossover rate with divergence between species. However, studies of humans/ chimpanzee and other taxa have observed an association between crossover rate and divergence between species, suggesting possible mutational (rather than, or in addition to, selective) explanations.
Using our estimates of fine-scale crossover rate and analyses of 454 whole-genome sequences that we have generated, we are examining this association in the Drosophila pseudoobscura species group. In our first study of this (collaborative with Rob Kulathinal), we found that crossover rate is strongly correlated with both nucleotide diversity within species and divergence between species, but that this relationship is only detectable when using fine-scale crossover rate rather than estimates of crossover rate from more distantly spaced markers. This explains why such a relationship was not detected in this species group previously, and may also explain some of the differences between species.
We also proposed a two-phase hypothesis for the association between crossover rate and diversity, based in part on our results (see abstract). As suggested by Begun & Aquadro and many others, regions of very low recombination have almost certainly had diversity reduced through the action of natural selection (either via hitchhiking or background selection). However, a relationship between crossover rate and diversity persists even in regions of high recombination possibly resulting from a mutagenic nature of double-strand break repair (DSBR), the precursor to crossing over. If true, then both selective and mutational forces create the association of crossover rate and nucleotide diversity, both in Drosophila and in humans.
Related research using genomic approaches to understand these and other evolutionary forces are continuing in the lab in both Drosophila and Saccharomyces.
Fine-mapping the genetic architecture of hybrid male sterility
One of the most fundamental problems in evolutionary genetics is why certain genes function 
normally within species (producing fertility) but fail to function in a hybrid genetic background (resulting in hybrid sterility). Further, why should the XY sex consistently evolve hybrid sterility earlier in divergence than the XX sex, a pattern called Haldane's Rule? Drosophila pseudoobscura and D. persimilis were among the first species studied to understand the genetic basis of hybrid male sterility and its evolutionary underpinnings, with work dating back to the early 1930's.
Resolving these questions requires the isolation of genes contributing to hybrid sterility and examination of their individual effects as well as their interactions with other genes. We have mapped multiple factors contributing to hybrid male sterility between D. persimilis and D. pseudoobscura bogotana in a recent study. Currently, we are building on this work in several respects. First, we are examining the individual and combinatorial effects of these alleles, and we have found that the dominance of the effects of these alleles is affected by alleles at other loci. Second, we are evaluating whether there is variation within species for these alleles, such that certain alleles from D. persimilis at a particular locus may cause sterility in a D. pseudoobscura bogotana genetic background while other alleles from D. persimilis may not. Finally, we are fine-mapping via introgression with the ultimate goal of attempting to identify candidate genes within these regions causing hybrid male sterility.
Chromosomal inversions and the genetics of reproductive isolation in Drosophila pseudoobscura and D. persimilis
We have spent many years examining the genetic bases of all barriers to gene exchange (such as hybrid sterility and mating discrimination) in the Drosophila pseudoobscura and D. persimilis system. These species have served as a model system for the study of speciation since the 1930's. Using molecular linkage maps of these species (see Noor et al 2000 and Ortiz-Barrientos et al 2006) and the Drosophila pseudoobscura genome sequence we have mapped the genetic basis of all these known barriers to gene exchange (reproductive isolating mechanisms) between these species. We found that the inverted regions on the X and 2nd chromosomes of these species are very strongly associated with all forms of reproductive isolation between them (see abstract1 and abstract2). Complementary research by the Machado and Hey laboratories also noted that introgression between these species was more limited in regions inverted between them than in collinear regions. This finding suggests that the chromosomal inversions separating these species may have contributed to their persistence in the face of gene flow. More recent work in other systems suggest the antirecombinational effects of chromosomal rearrangements may be generally important in species persistence.
More recently, we tested a model whereby chromosomal inversions (and other factors that can reduce recombination) are special to the speciation process in HYBRIDIZING taxa. We predicted a weaker association of hybrid sterility and other barriers to inversions in taxa that do not hybridize. This can be tested by comparing the genetics of hybrid sterility in the allopatric pair D. pseudoobscura bogotana/D. persimilis to the sympatric pair D. pseudoobscura pseudoobscura/D. persimilis. Our results strongly supported our inversion model: all sterility is associated with inversions in the hybridizing pair, while factors outside the inversions are detectable in the allopatric species pair (see abstract3 and abstract4).
Currently, we are examining how far outside the inversions the reduction in introgression extends. This indicates the domain of influence of chromosomal inversions on stopping introgression, and therefore their extent of effect on "speciation." Based on patterns of sequence polymorphism within and between species, this domain may be 2 megabases or greater (see abstract5 and abstract6). We are currently using 454 genome sequence analyses to further test these and other related hypotheses.
Back to Noor lab homepage.