The chemicals (Supplementary Table 1) used for the deorphanization of AaegOR9 were obtained from Sigma-Aldrich (Milwaukee, WI, USA), ChemCruz (Dallas, TX, USA), Glentham Life Sciences (Corsham, UK), FluoroChem (Hadfield, UK), SL Moran (Jerusalem, Israel), Holland Moran (Yehun, Israel), Alfa Aesar (Ward Hill, MA, USA) and from the generous contribution of the Dr. Kolodkin-Gal Lab (Weizmann Institute of Science, Israel).

The methodologies and protocols have been described in details elsewhere (Bohbot and Dickens, 2009). AaegOr9 and Aaeg-ORco cRNAs (Bohbot et al., 2011) were synthesized from linearized pSP64DV expression vectors using the mMESSAGE mMACHINE^® SP6 kit (Life Technologies). Stage V-VII oocytes were harvested from Xenopus laevis females, mechanically separated, treated with collagenase (8 mg/mL, 30 min, 18°C) and rinsed in washing solution (96 mM NaCl, 2 mM KCl, 5 mM MgCl₂ and 5 mM HEPES, pH 7.6). Oocytes were microinjected with 27.6 ng AaegOr9 and AaegORco cRNAs, incubated at 18°C for 3–4 days in ND96 solution (96 mM NaCl, 2 mM KCl, 5 mM MgCl₂, 0.8 mM CaCl₂ and 5 mM HEPES, pH 7.6), supplemented with 5% dialyzed horse serum, 50 μg/mL tetracycline, 100 μg/mL streptomycin and 550 μg/mL sodium pyruvate. Whole-cell currents were recorded using the two-microelectrode voltage-clamp technique. During recording sessions, the holding potential was maintained at −80 mV using an OC-725C oocyte clamp (Warner Instruments, LLC, Hamden, CT, USA). Oocytes placed in a RC-3Z oocyte recording chamber (Warner Instruments, LLC, Hamden, CT, USA) were exposed to odorants for 8 s. Current was allowed to return to baseline between odorant applications. Data acquisition and concentration-response analyses were carried out with a Digidata 1550A and the pCLAMP10 software (Molecular Devices, Sunnyvale, CA, USA), and analyzed using GraphPad Prism 7 (GraphPad Software Inc., La Jolla, CA, USA). Stock concentration of odorants (10⁻² M) were dissolved in ringer solution containing 2% dimethyl sulfoxide (DMSO) in order to solubilize the hydrophobic indolic compounds.

The response profile was established using multiple sessions, each including six compounds at a time and indole as an internal reference. The order in which these compounds were administered was reversed within a session to mitigate against any potential sequence effects between compounds (none were observed). All the response values were normalized to the indole reference in each recording session (Supplementary Figure 1).

For the establishment of the concentration-response curves, oocytes were exposed to increasing concentrations of indole, skatole and indole-3-carboxaldehyde (I3C; Supplementary Figure 2). Quantitative characterization of OR sensitivity was estimated using the averaged effective concentration at 50% of the maximal response (EC₅₀) over the sample population. The data to establish the concentration response curve and EC₅₀ of AaegOR10-skatole was extracted from a previous study (Bohbot and Dickens, 2012).

All the sequences used in our phylogenic analysis were obtained from the VectorBase and NCBI databases using AaegOr2/9/10 as query (for accession numbers, see Supplementary Table 3). DNA sequences for Toxorhynchites Or2 and Or10 can be accessed here: http://dx.doi.org/10.6084/m9.figshare.1092617. MAFFT version 7 (Nakamura et al., 2018) was used for multiple amino-acid sequence alignment. The phylogenic software IQ-TREE (Nguyen et al., 2015; Kalyaanamoorthy et al., 2017; Hoang et al., 2018) and the FigTree editor¹ were used for building the mosquito indolergic receptor phylogenic tree based on the maximum likelihood method (Model: JC, UFbootstrap: 5,000). Using MAFFT (default parameters) and the Vectorbase database, we located the intron positions on the indolergic receptor genes.

Based on its larval expression (Bohbot et al., 2007) and functional characterization (Bohbot et al., 2011), we initially surmised that AaegOR9 would be narrowly tuned to a water-soluble indolic ligand. To test this hypothesis, we screened AaegOR9 with a panel of 31 indolic derivatives (Supplementary Table 1) exhibiting some degree of water solubility using the two electrodes voltage clamp of Xenopus laevis oocytes. We also included indole and skatole as references (Bohbot et al., 2011). The screen was carried out using a low odorant concentration (500 nM) in order to mitigate the caveats associated with high ligand concentrations, including receptor adaptation, antagonist effects and technical artifacts such as broad molecular receptive ranges (Bohbot and Pitts, 2015).

At this concentration, skatole was decisively the most potent ligand, eliciting responses five times higher than indole (Figure 1) confirming the hypothesis that the potential cognate ligand of AaegOR9 is an indole derivative. However, this result contradicts our water-soluble ligand hypothesis. Indeed, skatole is about seven times less water soluble (0.5 mg/mL, ChemIDplus) than indole (3.56 mg/mL, ChemIDplus; Figure 1) and is considered rather insoluble in water as a result.

I3C, indole, 3-indole acetonitrile and methyl indole-3-carboxylate were the next most potent ligands suggesting that indole derivatives with a short side chain on position C3 can fit into the binding pocket of AaegOR9. However, comparable small indolic compounds such as indole-3-carbinole or gramine were among the least potent ligands (Figure 1). Overall, the AaegOR9 response profile was narrow (kurtosis value of 17.75), especially considering that our odorant panel was restricted to indole derivatives.

To characterize the pharmacological sensitivity of AaegOR9, we measured the amplitudes of the current responses of this receptor when exposed to increasing concentrations (100 pM to 10 μM) of skatole, indole and I3C (Figure 2A). Compound sensitivities were determined using the extrapolated EC₅₀ values. This analysis confirmed that AaegOR9 is 176 times more sensitive to skatole than to indole. AaegOR9 is a significantly more sensitive skatole receptor (EC₅₀ ≃ 5 nM) than AaegOR10 (EC₅₀ = 100 nM; Figure 2B). Although I3C and indole elicited comparable responses in our tuning curve experiment (Figure 1), their EC₅₀ values were significantly different (Figure 2A), underscoring the caveat of inferring receptor sensitivity based on tuning curves.

Plotting the EC₅₀ values of AaegOR9 against previously characterized OR-cognate odorant pairs (Supplementary Table 2) using our pharmacological platform (Bohbot and Pitts, 2015) reveals that the AaegOR9-skatole pair is the most sensitive indolergic OR (Figure 2C) and the most sensitive OR-ligand pair identified so far, outperforming the most sensitive pheromone receptor.

The conserved Or2, Or9, Or10 genes were initially identified from the An. gambiae and Ae. aegypti genomes (Hill et al., 2002; Bohbot et al., 2007, 2011). These genes encode proteins with high amino-acid sequence identity considering the sequence divergence characteristic of this family (Figure 3A). We have extended our previous analysis (Bohbot et al., 2011) by including additional Anopheles species, Ae. albopictus and Toxorhynchites amboinensis. We confirm that the indolergic receptor clade is divided into the OR2 and OR10 subgroups indicating their ancestral origin (Figure 3A). Or9 emerged in the Culicinae lineage 52–54 mya (Arensburger et al., 2010), which includes Ae. albopictus and Cu. quinquefasciatus. There is no information regarding Or expression in To. amboinensis. We failed to identify any signature of other paralogs in An. gambiae and Ae. aegypti consistent with the hypothesis that the Or9-Or10 split occurred in Culicinae.

The greater similarity between AaegOR9 and AaegOR10 not only includes sequence conservation and amino-acid substitution rates but is also reflected by the patterns of conserved introns (Figure 3A). Both Culicinae Or10 and Or9 genes are missing introns 3 and 5. The absence of intron 5 is also a conserved feature within the Anophelinae Or10 clade. Or2, Or9, Or10 are clustered together in region 44 of the q arm of chromosome 2 in Ae. aegypti (Figure 3B). This region corresponds to region 34B of the orthologous chromosome 3R in An. gambiae, reflecting the extensive paracentric inversion events that occurred within chromosome segments between these two species (Severson et al., 2004; Arensburger et al., 2010).