Lack of population differentiation patterns of previously identified putatively adaptive transposable element insertions at microgeographic scales

Genome-wide patterns of local adaptation in Western European Drosophila melanogaster natural populations

Signatures of spatially varying selection have been investigated both at the genomic and transcriptomic level in several organisms. In Drosophila melanogaster, the majority of these studies have analyzed North American and Australian populations, leading to the identification of several loci and traits under selection. However, several studies based mainly in North American populations showed evidence of admixture that likely contributed to the observed population differentiation patterns. Thus, disentangling demography from selection might be challenging when analyzing these populations. European populations could help identify loci under spatially varying selection provided that no recent admixture from African populations would have occurred. In this work, we individually sequence the genome of 42 European strains collected in populations from contrasting environments: Stockholm (Sweden) and Castellana Grotte (Southern Italy). We found low levels of population structure and no evidence of recent African admixture in these two populations. We thus look for patterns of spatially varying selection affecting individual genes and gene sets. Besides single nucleotide polymorphisms, we also investigated the role of transposable elements in local adaptation. We concluded that European populations are a good dataset to identify candidate loci under spatially varying selection. The analysis of the two populations sequenced in this work in the context of all the available D. melanogaster data allowed us to pinpoint genes and biological processes likely to be relevant for local adaptation. Identifying and analyzing populations with low levels of population structure and admixture should help to disentangle selective from non-selective forces underlying patterns of population differentiation in other species as well.

Putative adaptive inter-slope divergence of transposon frequency in fruit flies (Drosophila melanogaster) at "Evolution Canyon", Mount Carmel, Israel

Background

The current analysis of transposon elements (TE) in Drosophila melanogaster at Evolution Canyon, (EC), Israel, is based on data and analysis done by our collaborators (Drs. J. Gonzalez, J. Martinez and W. Makalowski, this issue). They estimated the frequencies of 28 TEs (transposon elements) in fruit flies (D. melanogaster) from the ecologically tropic, hot, and dry south-facing slope (SFS) or “African” slope (AS) of EC and compared it with the TE frequencies on the temperate-cool and humid north-facing slope (NFS) or “European” slope (ES), separated, on average, by 250 m. The flies were sampled from two stations on each slope. We received their results, including the frequencies of each TE on each slope, and the probabilities of the statistical analyses (G-tests) of each TE separately. We continued the analysis of the inter-slope differences of the frequencies of the TEs, and based our different conclusions on that analysis and on the difference between micro (=EC) and macro (2000 km.) comparisons [Gonzalez et al. 2015 doi:10.1186/s13062-015-0075-4].

Results

Our collaborators based all their conclusions on the non-significant results of each of the individual tests of the 28 TEs. We analysed also the distribution of the TE differences between the slopes, based on their results. Thirteen TEs were more frequent on the SFS, 11 were more frequent on the NFS, and four had equal frequencies. Because of the equalizing effect of the ongoing migration, only small and temporary differences between the slopes (0 – 0.06) were regarded by us as random fluctuations (drift). Three TEs were intermediate (0.08-0.09) and await additional research. The 11 TEs with large frequency differences (0.12 – 0.22) were regarded by us as putative adaptive TEs, because the equalizing power of ongoing migration will eliminate random large differences. Five of them were higher on the SFS and six were higher on the NFS. Gaps in the distribution of the differences distinguished between the large and small differences. The large gap among the 11 TEs favored on the NFS was significant and supports our rejection of drift as the only explanation of the distribution of the slope differences. The gaps in the distribution of the differences separated the putative TEs with strong enough selection from those TEs that couldn't overrule the migration.

The results are compared and contrasted with the directional effect of the frequencies of the same TEs in the study of global climatic comparisons across thousands of kilometers. From the 11 putative adaptive TEs in the local “Evolution Canyon,” six differentiate in the same direction as in the continental comparisons and four in the opposite direction. One TE, FBti0019144, differentiated in EC in the same direction as in Australia and in the opposite direction to that of North America.

Conclusions

We presume that the major divergent evolutionary driving force at the local EC microsite is natural selection overruling gene flow. Therefore, after we rejected drift as an explanation of all the large slope differences, we regarded them as putatively adaptive. In order to substantiate the individual TE adaptation, we need to increase the sample sizes and reveal the significant adaptive TEs.

The comparison of local and global studies show only partial similarity in the adaptation of the TEs, because of the dryness of the ecologically tropical climate in EC, in contrast to the wet tropical climate in the global compared climates. Moreover, adaptation of a TE may be expressed only in part of the time and specific localities.

Reviewers

Reviewed by Eugene Koonin, Limsoon Wong and Fyodor Kondrashov. For the full reviews, please go to the Reviewers’ comments section.

Electronic supplementary material

The online version of this article (doi:10.1186/s13062-015-0074-5) contains supplementary material, which is available to authorized users.