Conservation of indole responsive odorant receptors in mosquitoes reveals an ancient olfactory trait

Abstract

Aedes aegypti and Anopheles gambiae are among the best-characterized mosquito species within the Culicinae and Anophelinae mosquito clades which diverged ∼150 million years ago. Despite this evolutionary distance, the olfactory systems of these mosquitoes exhibit similar morphological and physiological adaptations. Paradoxically, mosquito odorant receptors, which lie at the heart of chemosensory signal transduction pathways, belong to a large and highly divergent gene family. We have used 2 heterologous expression systems to investigate the functional characteristics of a highly conserved subset of Ors between Ae. aegypti and An. gambiae to investigate whether protein homology correlates with odorant-induced activation. We find that these receptors share similar odorant response profiles and that indole, a common and ecologically relevant olfactory cue, elicits strong responses from these homologous receptors. The identification of other highly conserved members of this Or clade from mosquito species of varying phylogenetic relatedness supports a model in which high sensitivity to indole represents an ancient ecological adaptation that has been preserved as a result of its life cycle importance. These results provide an understanding of how similarities and disparities among homologous OR proteins relate to olfactory function, which can lead to greater insights into the design of successful strategies for the control of mosquito-borne diseases.

Figure 1

The OR2/9/10 clade predates the Anophelinae/Culicinae split. (A) Phylogenetic relationships of the mosquito OR2 (light gray shaded area), OR9 (medium gray shaded area), and OR2 (dark gray shaded area) clade of mosquitoes. Aa: Aedes aegypti; Cx: Culex quinquefasciatus; Ag: An. gambiae; Aq: An. quadriannulatus; As: An. stephensi. The P-distance tree was generated using MEGA 3.1 using a Neighbor-joining model. Branch lengths are proportional to the scale of sequence distance indicated by the bar below the tree. Bootstrap values (%) are based on 10 000 replicates. Gene structures are indicated by black and whites boxes. Intron positions and protein lengths are indicated above the gene structure. Intron phases are indicated in bold below the gene structure. Complete cDNA sequence was characterized for AqOr2, whereas an incomplete genomic DNA sequence was obtained thus providing only the position of the first 2 introns. (B) Microsynteny and gene structure of Ae. aegypti Or2/10, C. quinquefasciatus Or2/9/10, and An. gambiae Or2/10 genes. Each filled black and white squares represent an exon. Distance between genes is indicated below individual contig.

Figure 2

Members of the OR2/10 clade exhibit overlapping sensitivities. Tuning curves of AaOR2, AgOR2, AaOR9, AaOR10, AgOR10, and AgOR8. The 30 odorants are ordered along the x axis, with those eliciting the strongest responses for the OR2/10 clade near the center. Benzaldehyde and 1-octen-3-ol (chemical structures shown) elicited the highest response for the OR2/10 clade and OR8, respectively. Arrows indicate odorants for which AaOR9 responses are more similar to OR2s than to OR10s.

Figure 3

Indole activation correlate with the primary structure of the OR2/9/10 protein. Normalized CRCs of OR2s, OR9, and OR10s expressed in Xenopus oocytes (A) and in HEK cells (B) in response to indole (n = 4–7). Odorant concentrations were plotted on a logarithmic scale. (C) Scatter chart displaying 3 functional OR groups: the OR2, AaOR9/AgOR10, and AaOR10 groups (n = 4–8) expressed in Xenopus oocytes. (D) Scatter chart displaying the sensitivity of the OR2/10 clade expressed in HEK cells in response to indole, 4-MP, and benzaldehyde (n = 3–8). Three asterisks, P < 0.001 (analysis of variance test with Tukey post test). The mean EC₅₀ values and standard error of the mean of their scatter were determined using Prism5.