Natural variation in odorant recognition among odorant-binding proteins in Drosophila melanogaster

Abstract

Chemical recognition is essential for survival and reproduction. Adaptive evolution has resulted in diverse chemoreceptor families, in which polymorphisms contribute to individual variation in chemosensation. To gain insights into the genetic determinants of individual variation in odorant recognition, we measured olfactory responses to two structurally similar odorants in a population of wild-derived inbred lines of Drosophila melanogaster. Odorant-binding proteins (OBPs) are the first components of the insect olfactory system to encounter odorants. Previously four single-nucleotide polymorphisms (SNPs) in the Obp99 group were associated with variation in olfactory responses to benzaldehyde. Here, we identify six different SNPs that are associated with variation in responses to a structurally similar odorant, acetophenone, in the same Obp genes. Five SNPs are in coding regions of Obp99b and Obp99d and one SNP is in the 3'-untranslated region of Obp99a (A610G). We found that the 610G allele is associated with higher response scores to acetophenone than the 610A allele, but with lower expression of Obp99a, suggesting that binding of acetophenone to Opb99a might limit rather than facilitate access to odorant receptors. Our results show that overlapping sets of OBPs contribute to odorant recognition for structurally similar odorants, but that different SNPs are associated with odorant-specific individual variation. Thus, dual olfactory recognition where OBPs regulate odorant access to receptors may enhance olfactory discrimination.

Figure 1.—

Olfactory responses of 297 wild-derived inbred lines. (a) Distribution of mean olfactory response scores for male (solid bars) and female (shaded bars) flies to 3.5% (v/v) acetophenone. The inset in a shows the structural similarity between benzaldehyde and acetophenone. (b) Correlations between olfactory response scores to acetophenone and benzaldehyde for females (left graph) and males (right graph).

Figure 2.—

Associations of polymorphisms in the Obp99 gene group with variation in behavioral responses to acetophenone. (a) Relative locations of the four Obp99 genes on the right arm of the third chromosome (3R) with arrows showing the direction of transcription. The distances between the genes are indicated. (b) Schematic representations of the Obp99a, Obp99d, Obp99c, and Obp99b genes. Blue boxes, exons; red boxes, 5′-untranslated regions; white boxes, 3′-untranslated regions; and intervening black lines, introns. The purple horizontal line in each graph indicates the significance threshold for association determined by Bonferroni correction for multiple testing. Blue arrowheads indicate the locations of SNPs with significant associations with variation in olfactory responses to acetophenone. For comparison, orange bars indicate polymorphic markers in the same genes previously associated with olfactory response to benzaldehyde in a subset of the same lines (Wang et al. 2007).

Figure 3.—

Haplotype analysis for polymorphic markers associated with variation in response to acetophenone in the Obp99d gene. The bar graphs show significant differences in olfactory behavior to acetophenone of the haplotypes GTGG, GGGA, and GTAA from AGAA. The haplotypes analyzed are composed of the four SNPs associated with variation in response to acetophenone in Obp99d. Additional haplotypes were not observed or present at low frequency. Data were analyzed by ANOVA and haplotypes that differ significantly in olfactory behavior in response to acetophenone from one another were identified by post hoc Tukey's test and are indicated with different letters above the bars.

Figure 4.—

Linkage disequilibrium analysis of SNPs associated with variation in olfactory response to acetophenone. Linkage disequilibrium analysis for all possible pairwise combinations of SNPs in Obp99a, Obp99d, and Obp99b was based on complete sequence data for these genes from 297 wild-derived inbred lines. Color-coded boxes above the diagonal indicate corresponding P-values for the marker combinations, whereas R² values are indicated below the diagonal. Note the overall large extent of historical recombination. SNPs T192G, G293A, and G363A in Obp99d, which are associated with variation in olfactory response to acetophenone, are in linkage disequilibrium with each other and with C384T of Obp99b, which is located adjacent on the chromosome (Figure 2). Linkage disequilibrium among these SNPs is indicated on the right.

Figure 5.—

Correspondence between behavioral response to acetophenone and differences in expression of Obp99a correlated with the A and G alleles of the A610G SNP. Five high-responding lines with the G allele and five low-responding lines with the A allele of the polymorphic marker A610G in the 3′-untranslated region of Obp99a were selected (a) and expression levels of Obp99a were assessed (b). The dashed lines in a indicate the mean response scores for the five high- and low-responding lines, respectively (***, P < 0.0001; two-tailed Student's t-test). The y-axis in b shows the mean expression level of the five high-responding lines standardized to the mean expression level of the five low-responding lines, set at 1.0. Mean expression levels across lines with high response scores to acetophenone corresponding to the 610G allele showed ∼48% lower expression levels of the Obp99a transcript than lines with low response scores to acetophenone and the 610A allele (***, P < 0.0001; two-tailed Student's t-test).