Molecular population genetics of the OBP83 genomic region in Drosophila subobscura and D. guanche: contrasting the effects of natural selection and gene arrangement expansion in the patterns of nucleotide variation

Abstract

Chromosomal inversion polymorphism play a major role in the evolutionary dynamics of populations and species because of their effects on the patterns of genetic variability in the genomic regions within inversions. Though there is compelling evidence for the adaptive character of chromosomal polymorphisms, the mechanisms responsible for their maintenance in natural populations is not fully understood. For this type of analysis, Drosophila subobscura is a good model species as it has a rich and extensively studied chromosomal inversion polymorphism system. Here, we examine the patterns of DNA variation in two natural populations segregating for chromosomal arrangements that differentially affect the surveyed genomic region; in particular, we analyse both nucleotide substitutions and insertion/deletion variations in the genomic region encompassing the odorant-binding protein genes Obp83a and Obp83b (Obp83 region). We show that the two main gene arrangements are genetically differentiated, but are consistent with a monophyletic origin of inversions. Nevertheless, these arrangements interchange some genetic information, likely by gene conversion. We also find that the frequency spectrum-based tests indicate that the pattern of nucleotide variation is not at equilibrium; this feature probably reflects the rapid increase in the frequency of the new gene arrangement promoted by positive selection (that is an adaptive change). Furthermore, a comparative analysis of polymorphism and divergence patterns reveals a relaxation of the functional constraints at the Obp83b gene, which might be associated with particular ecological or demographic features of the Canary island endemic species D. guanche.

Figure 1

In situ hybridization on an O₃₊₄ polytene chromosome of D. subobscura using the complete Obp83 region as a biotinylated probe. The arrow indicates the hybridization signal.

Figure 2

Scheme of the location of the Obp83 and rp49 regions (Rozas and Aguadé, 1994) in different chromosomal arrangements of D. subobscura. Shaded bars indicate the regions affected by inversions. O₃ refers to a gene arrangement not present in extant populations of D. subobscura.

Figure 3

Sliding window of silent polymorphism in D. subobscura (black line) and silent divergence between D. subobscura and D. guanche (grey line) in the Obp83 region. (a) Total D. subobscura data; (b) O_[3+4] chromosomal arrangement and (c) O₃₊₄₊₂₃ chromosomal arrangement.

Figure 4

(a) ML tree of the Obp83 region. Italic numbers indicate the percentage of bootstrap replicates (1000 replicates) supporting the main nodes (only bootstrap values higher than 70% are shown). (b) Neighbour joining tree of the Obp83 region. Trees were built using either nucleotide substitutions (a) or InDel events (b) information, and rooted with the D. guanche sequence. The scale bars in (a, b) indicate the number of substitutions per site and the number of InDel events, respectively. Lines from Bizerte and El Pedroso populations are depicted in white and black, respectively, whereas those of O_[3+4] and O₃₊₄₊₂₃ arrangements are represented by a square and a circle, respectively.

Figure 5

LD r² statistic (Hill and Robertson, 1968) between pairs of polymorphic sites across the Obp83 genomic region in the two chromosomal arrangements. (a) O_[3+4] chromosomal arrangement and (b) O₃₊₄₊₂₃ chromosomal arrangement. Black line represents the straight-line fit to the plot using least squares regression.