1. Neuroscience
Download icon

Two single-point mutations shift the ligand selectivity of a pheromone receptor between two closely related moth species

  1. Ke Yang
  2. Ling-Qiao Huang
  3. Chao Ning
  4. Chen-Zhu Wang Is a corresponding author
  1. Institute of Zoology, Chinese Academy of Sciences, China
  2. University of Chinese Academy of Sciences, China
Research Article
Cited
0
Views
726
Comments
0
Cite as: eLife 2017;6:e29100 doi: 10.7554/eLife.29100

Abstract

Male moths possess highly sensitive and selective olfactory systems that detect sex pheromones produced by their females. Pheromone receptors (PRs) play a key role in this process. The PR HassOr14b is found to be tuned to (Z)−9-hexadecenal, the major sex-pheromone component, in Helicoverpa assulta. HassOr14b is co-localized with HassOr6 or HassOr16 in two olfactory sensory neurons within the same sensilla. As HarmOr14b, the ortholog of HassOr14b in the closely related species Helicoverpa armigera, is tuned to another chemical (Z)−9-tetradecenal, we study the amino acid residues that determine their ligand selectivity. Two amino acids located in the intracellular domains F232I and T355I together determine the functional difference between the two orthologs. We conclude that species-specific changes in the tuning specificity of the PRs in the two Helicoverpa moth species could be achieved with just a few amino acid substitutions, which provides new insights into the evolution of closely related moth species.

https://doi.org/10.7554/eLife.29100.001

Introduction

Almost all animals detect and react to pheromones and the other chemical cues that indicate food, shelter or predators, and their olfactory systems are mainly involved in the processes (Wyatt, 2003). As powerful chemical signals, pheromones are enormously varied in different animal species. How the animal olfaction has evolved at the molecular level to adapt to the changing pheromones is a forefront research subject in life sciences.

Moths are good model systems for pheromone communication study. Male moths fly upwind to find conspecific females releasing a plume of sex pheromone (Cardé and Haynes, 2004). Most moth sex pheromones have multiple components present in specific ratios that play significant roles in intraspecific sexual communication and in interspecific reproductive isolation (Cardé et al., 1977). Male moths possess highly sensitive and selective olfactory sensory neurons (OSNs) located in antennal sensilla that detect the pheromone molecules (Schneider, 1964; Hansson and Stensmyr, 2011). Pheromone receptors (PRs) located in the dendritic membrane of OSNs play a pivotal role in peripheral coding of sex pheromones (Leal, 2013; Sakurai et al., 2004).

Unlike general odorant receptors (ORs) that typically bind more than one ligand (de Fouchier et al., 2017), PRs are in general narrowly tuned to specific pheromone components (Grosse-Wilde et al., 2007; Miura et al., 2010; Zhang and Löfstedt, 2015). The ligands of some PRs in lepidopteran species have been successfully identified using heterologous expression systems, including Xenopus oocytes (Wetzel et al., 2001), the HEK293 cell line (Grosse-Wilde et al., 2006), the Sf9 cell line (Kiely et al., 2007), Drosophila melanogaster delta-halo mutants with an empty ab3A neuron (Dobritsa et al., 2003), and Or67D-GAL4 mutants (Kurtovic et al., 2007). However, the functions of many PRs in moth species are still unknown, thus hindering our understanding of pheromone detection at the molecular level in this group of insects.

Closely related moth species often use combinations of the same or similar pheromone components, which is a reflection of their common evolutionary history (Cardé and Haynes, 2004). They also possess homologous PRs with very similar sequences, but clearly differentiated in their ligand specificity. How do changes in amino acid sequences alter the ligand selectivity of PRs? The single study to date that has addressed this question showed that a single-point mutation in a PR is responsible for its different specificities in Ostrinia furnacalis and Ostrinia nubilalis (Leary et al., 2012). Studies of a number of ORs in D. melanogaster and Anopheles gambiae indicated that the determinant amino acids are located mainly in transmembrane domains and extracellular loops (Guo and Kim, 2010; Hughes et al., 2014; Nichols and Luetje, 2010; Pellegrino et al., 2011), but the molecular mechanisms that determine the ligand selectivity of ORs are still unclear.

The closely related moth species, Helicoverpa assulta and Helicoverpa armigera, are sympatric pests in Asia. The former is a specialist mainly feeding on solanaceous plants, including tobacco and hot pepper, whereas the latter is a polyphagous species and is one of the most devastating pests in the world. H. assulta and H. armigera share two compounds, (Z)−9-hexadecenal (Z9-16:Ald) and (Z)−11-hexadecenal (Z11-16:Ald) as their principal sex-pheromone components, but in inverse ratios, 93:7 and 3:97, respectively (Piccardi et al., 1977; Wang et al., 2005). (Z)−9-tetradecenal (Z9-14:Ald) acts as an antagonist in the pheromone communication of H. assulta (Boo et al., 1995; Wu et al., 2015). In that of H. armigera, Z9-14:Ald acts as an agonist in small amounts (0.3%) (Rothschild, 1978; Wu et al., 2015; Zhang et al., 2012) but an antagonist in higher amounts (1% and above) (Gothilf et al., 1978; Kehat and Dunkelblum, 1990; Wu et al., 2015). Three functional types of pheromone-sensitive sensilla, A, B and C, can be distinguished in the male antennae of the two species (Baker et al., 2004). Sensilla type A specifically respond to Z11-16:Ald, type B respond to Z9-14:Ald, and type C respond to Z9-16:Ald, Z9-14:Ald and some other structurally related compounds. B-type and C-type sensilla are classified into subtypes according to their response spectra (Xu et al., 2016). The population of A-type sensilla predominates in males of H. armigera, while C-type sensilla are most numerous in males of H. assulta (Wu et al., 2013; Xu et al., 2016). The two species share almost the same set of orthologous PRs. Previous functional studies of the PRs showed that HarmOr13 and HassOr13 are specifically tuned to Z11-16:Ald (Jiang et al., 2014; Liu et al., 2013), HarmOr14b and HassOr16 are tuned to Z9-14:Ald (Jiang et al., 2014; Liu et al., 2013), HarmOr16 is tuned to both Z9-14:Ald and (Z)−11-hexadecenol (Z11-16:OH) (Liu et al., 2013), while HarmOr6 and HassOr6 are mainly tuned to (Z)−9-hexadecenol (Z9-16:OH) (Jiang et al., 2014). However, it is still unclear which PR is specific for Z9-16:Ald, the major component of H. assulta sex pheromone.

In this study, we first identified the PR tuned to Z9-16:Ald in H. assulta. Because C-type sensilla responding to Z9-16:Ald are densely distributed in the male antennae of H. assulta, we predicted that this PR should be highly expressed in male antennae. Therefore, we used qPCR to analyze the expression level of all candidate PRs in male antennae in H. assulta, and then used the Xenopus oocyte expression system and two-electrode voltage-clamp recording to examine the function of highly expressed PRs. We surprisingly found that the PR tuned to Z9-16:Ald is HassOr14b, while its ortholog HarmOr14b is tuned to Z9-14:Ald in the closely related species H. armigera. Next, focusing on the two orthologous receptors, we identified the amino acid residues determining this functional shift. We used a series of regional replacements and single-point mutations, coupled with functional analyses, to demonstrate that two single-point mutations located in the intracellular regions of the molecule together determine their ligand selectivity. Our results suggest that a change in the tuning selectivity of PRs during the speciation of some moths could result from just a few mutations.

Results

Phylogenetic analysis of candidate PRs

The reported transcriptome data and full-length cloning of the PRs made it possible to analyze all candidate PRs in the two closely related species and in other species of Noctuidae. The amino acid sequences of seven PRs from Helicoverpa species (Jiang et al., 2014; Liu et al., 2014; Xu et al., 2015) and 32 PRs from other noctuids were used to construct a phylogenetic tree, where the Orco sequences represented an outgroup (Krieger et al., 2004; Liu et al., 2013; Mitsuno et al., 2008; Montagné et al., 2012; Zhang and Löfstedt, 2013; Zhang et al., 2015, Zhang et al., 2014Zhang et al., 2014Zhang et al., 2014) (Figure 1). In general, the tree was clustered into seven lineages, Or16, Or6, Or14b, Or14, Or15, Or11, and Or13. Each lineage contains the PR(s) from Helicoverpa and other noctuids except for Or14b, suggesting that the Or14b cluster specifically occurs in H. assulta and H. armigera. To investigate the evolutionary pressures acting on the coding regions of each cluster, we estimated the ratios of nonsynonymous (dN) to synonymous (dS) nucleotide substitution (ω = dN/dS) in the PR gene lineages and the Orco lineage using DnaSP version 5.10 (Librado and Rozas, 2009). The ω values < 1 were observed in all PR clusters and the Orco cluster (cluster Or16: ω <0.21; cluster Or6: ω <0.19; cluster Or14b: ω = 0.17; cluster Or14: ω <0.15; cluster Or15: ω = 0.13; cluster Or11: ω <0.13; cluster Or13, ω <0.17; cluster Orco: ω <0.03); this indicates that all the PRs and Orcos analyzed in this study are subjected to purifying selection, which is consistent with the previous studies (Zhang and Löfstedt, 2013; Zhang et al., 2014).

The phylogenetic tree of the PRs in Noctuidae.

The amino acid sequences are based on the reported transcriptome data of functionally identified PRs. The Orco lineage is defined as an outgroup. Bootstrap values are based on 1000 replicates, and values over 50 are shown at corresponding nodes. The bar indicates the phylogenetic distance value. The nonsynonymous (dN) to synonymous (dS) substitution ratio (ω) is labeled in the tree. Cluster Or14b and Or15 have a uniform ω value for all branches, whereas Cluster Or6, 11, 13, 14, and 16 have varying ω values for all branches within the lineage. The ω values of all clusters were less than 1, suggesting that all PRs were subjected to purifying selection. Abbreviations: Aseg, Agrotis segetum; Harm, H. armigera; Hass, H. assulta; Hvir, Heliothis virescens; Msep, Mythimna separata; Sexi, Spodoptera exigua; Slitt, Spodoptera littoralis; Slitu, Spodoptera litura; Sinf, Sesamia inferens. The PRs of H. assulta are indicated by red dots ‘’. The GenBank accession numbers of genes used in this analysis are listed in Supplementary file 2.

https://doi.org/10.7554/eLife.29100.002

Expression level of candidate PRs in antennae of H. assulta and H. armigera

The antennal expression levels in males and females were compared using quantitative real-time PCR (qPCR). All the candidate PRs were male-specific except for Or11, which was highly expressed in both male and female antennae (Figure 2 and Figure 2—figure supplement 2). In the male antennae of H. assulta, HassOr14b had the highest expression level, nearly twofold higher than the levels of HassOr6 and HassOr16, and five- to sixfold greater than that of HassOr13 (Figure 2). The values of fragments per kilobase of transcript per million reads (FPKM) in different tissues of H. assulta further demonstrated that HassOr14b is specifically expressed in the male antennae and that its expression level is the highest among the PRs (Figure 2—figure supplement 1). In the male antennae of H. armigera, HarmOr13 and HarmOr11 showed the highest expression level, which was about five- to sixfold higher than HarmOr16 and HarmOr14; HarmOr14b had a low expression level, even lower than HarmOr16 and HarmOr6 (Figure 2—figure supplement 2). Since the C-type sensilla responding to Z9-16:Ald were the most abundant type in the male antennae of H. assulta, we speculated that HassOr14b would be the PR tuned to Z9-16:Ald, different from HarmOr14b which tuned to Z9-14:Ald.

Figure 2 with 2 supplements see all
Relative mRNA expression levels of PRs by quantitative real-time PCR analysis in male and female antennae of H.

assulta. Hass♂, male antennae; Hass♀, female antennae. n = 3 replicates of 40–60 antennae each. Data are presented as mean ± SEM.

https://doi.org/10.7554/eLife.29100.003

PR specifically tuned to Z9-16:Ald in H. assulta

We re-cloned the sequence of HassOr14b and verified it by Sanger sequencing and the transcriptome data of H. assulta. We used the Xenopus laevis oocyte expression system and two-electrode voltage-clamp recording to study the function of HassOr14b although it has already been shown that its ortholog, HarmOr14b, is tuned to Z9-14:Ald (Jiang et al., 2014). Oocytes expressing HassOr14b/HassOrco responded robustly to Z9-16:Ald, and to a much lesser extent to Z9-16:OH at a concentration of 10–4 M (Figure 3A). Z9-16:Ald induced currents increasing from the lowest threshold concentration of 10−6 M to 3.3 × 10–3 M in a dose-dependent manner with an EC50 value of 8.65 × 10–5 M (Figure 4). We also verified the function of the ortholog of HassOr14b, HarmOr14b and the next most highly expressed PRs in male H. assulta, HassOr6 and HassOr16. As previously reported (Jiang et al., 2014), we verified that HarmOr14b is specifically tuned to Z9-14:Ald, and also weakly responds to Z9-16:Ald (Figure 3B), while HassOr6 is mainly tuned to Z9-16:OH (Figure 3—figure supplement 1A), and HassOr16 is specific for Z9-14:Ald (Figure 3—figure supplement 1B). Water-injected oocytes fail to respond to any of the pheromone component stimuli as negative controls (Figure 3—figure supplement 2 and Figure 4—figure supplement 1).

Figure 3 with 3 supplements see all
Two-electrode voltage-clamp recordings of Xenopus oocytes with co-expressed HassOr14b/HassOrco, and HarmOr14b/HarmOrco, stimulated with pheromone components and analogs.

(A) Inward current responses (left) and response profiles (right) of Xenopus oocytes expressing HassOr14b/HassOrco in response to 10−4 M concentrations of pheromone components and analogs. n = 7 replicates of cells, F = 31.75, p<0.001, one-way ANOVA, Tukey HSD test. (B) Inward current responses (left) and response profiles (right) of Xenopus oocytes expressing HarmOr14b/HarmOrco in response to 10−4 M concentrations of pheromone components and analogs. n = 7 replicates of cells, F = 17.67, p<0.001, one-way ANOVA, Tukey HSD test. Data are presented as mean ± SEM.

https://doi.org/10.7554/eLife.29100.006
Figure 4 with 1 supplement see all
Dose responses of Xenopus oocytes with co-expressed HassOr14b/HassOrco stimulated with a range of Z9-16:Ald concentrations.

Inward current responses (left) and response profiles (right) of Xenopus oocytes expressing HassOr14b/HassOrco in response to Z9-16:Ald at serial concentrations. The EC50 value for Z9-16:Ald was 8.65 × 10–5 M. n = 7–9 replicates of cells, F = 54.57, p<0.001, one-way ANOVA, Tukey HSD test.

https://doi.org/10.7554/eLife.29100.010

Co-localization of HassOr14b with other PRs in type C sensilla

Based on the above results and previous reports (Jiang et al., 2014; Xu et al., 2016), we considered that HassOr14b, HassOr6 and HassOr16 were most likely to be the PRs expressed in C-type sensilla of H. assulta. By two-color in situ hybridization, we further analyzed the co-localization of these three PRs. We found that HassOr14b and HassOr6 were co-localized in some sensilla (arrows, Figure 5A1–4), while in other sensilla only HassOr14b was detected (arrows, Figure 5B1–4). A similar situation was observed for HassOr14b and HassOr16. They were co-localized in some sensilla (arrows, Figure 5C1–4), but only HassOr14b was detected in other sensilla (arrows, Figure 5D1–4). However, HassOr6 and HassOr16 were always expressed in different sensilla (arrows, Figure 5E1–4). These results indicate that HassOr14b is co-localized with HassOr6 or HassOr16 in different C type sensilla.

Two-colour in situ hybridization visualizing the combinations of HassOr14b/HassOr6, HassOr14b/HassOr16, and HassOr6/HassOr16 in male antennae of H.assulta.

(A, B) The localization of HassOr14b and HassOr6. (C, D) The localization of HassOr14b and HassOr16. (E) The localization of HassOr6 and HassOr16. Signals were visualized by red (digoxin-labeled probes) (A1, B1, C1, D1), green (biotin-labeled probes) (A2, B2, C2, D2), and both red and green (A3, B3, C3, D3) fluorescence. Bright-field images are presented as references (A4, B4, C4, D4). Arrows indicate the cell location. Scale bars: 20 μm.

https://doi.org/10.7554/eLife.29100.012

Regional mutations of HassOr14b and functional analysis

HassOr14b and HarmOr14b exhibit 91% amino acid identity (402 out of 440, Figure 6—figure supplement 1), but their ligand selectivity is different. This provided an opportunity to examine the relationship between structure and function in the two orthologous PRs. From the sequence alignment and secondary structural analysis (Figure 6; TOPCONS, topcons.net), we found that the 38 differing amino acids were distributed fairly uniformly in the two proteins. Therefore, we separated the whole sequence into eight regions (RI–VIII) (Figure 6). Then we conducted a series of mutagenesis experiments by replacing each of the eight regions of HassOr14b with the corresponding segment of HarmOr14b, while maintaining the rest of the sequence unchanged. After successfully constructing the modified sequences, we analyzed their functions as for the wild type (Figure 7). Interestingly, comparing with the ligand selectivity of the wild type (Figure 7—figure supplement 2), we observed that the ligand selectivity of HassOr14b was changed remarkably by replacement of the region VI or VIII. HassOr14b after replacing the region VI had a significantly higher response to Z9-14:Ald than to Z9-16:Ald (Figure 7F), while after replacing the region VIII had strong responses to both Z9-16:Ald and Z9-14:Ald (Figure 7H). However, most of the region replacements in HassOr14b showed selectivities similar to that of the wild type, with Z9-16:Ald being the most effective ligand. In particular, replacement of the regions I or III produced significantly lower responses to Z9-16:Ald (Figure 7A and C), while replacement of the region VII resulted in a significantly stronger response to Z9-16:Ald (Figure 7G). Replacement of the regions II, IV or V did not affect the selectivity of the receptor with reference to the wild-type (Figure 7B,D and E).

Figure 6 with 1 supplement see all
Eight mutation regions in the predicted secondary structure of HassOr14b.

Each circle indicates an amino acid residue that differs between HassOr14b and HarmOr14b and is highlighted in orange. Black indicates mutation regions RI and RV; green indicates RII and RVI; grey indicates RIII and RVIII; blue indicates RIV and RVII. The image was constructed by TOPO2 software (http://www.sacs.ucsf.edu/TOPO2/) based on the secondary structure predicted by TOPCONS (topcons.net) models (Tsirigos et al., 2015). The structures of both HassOr14b and HarmOr14b were predicted and the model with a reliable 7-transmembrane structure was adopted.

https://doi.org/10.7554/eLife.29100.013
Figure 7 with 2 supplements see all
Two-electrode voltage-clamp recordings of Xenopus oocytes with co-expressed regional mutations and HassOrco stimulated with pheromone components and analogs.

(A) Inward current responses (left) and response profiles (right) of Xenopus oocytes expressing RI mutant/HassOrco in response to 10−4 M concentrations of pheromone components and analogs. n = 7 replicates of cells, F = 85.28, p<0.001. (B) Inward current responses (left) and response profiles (right) of Xenopus oocytes expressing RII mutant/HassOrco in response to 10−4 M concentrations of pheromone components and analogs. n = 6 replicates of cells, F = 29.11, p<0.001. (C) Inward current responses (left) and response profiles (right) of Xenopus oocytes expressing RIII mutant/HassOrco in response to 10−4 M concentrations of pheromone components and analogs. n = 5 replicates of cells, F = 24.94, p<0.001. (D) Inward current responses (left) and response profiles (right) of Xenopus oocytes expressing RIV mutant/HassOrco in response to 10−4 M concentrations of pheromone components and analogs. n = 6 replicates of cells, F = 31.12, p<0.001. (E) Inward current responses (left) and response profiles (right) of Xenopus oocytes expressing RV mutant/HassOrco in response to 10−4 M concentrations of pheromone components and analogs. n = 6 replicates of cells, F = 23.80, p<0.001. (F) Inward current responses (left) and response profiles (right) of Xenopus oocytes expressing RVI mutant/HassOrco in response to 10−4 M concentrations of pheromone components and analogs. n = 7 replicates of cells, F = 14.77, p<0.001. (G) Inward current responses (left) and response profiles (right) of Xenopus oocytes expressing RVII mutant/HassOrco in response to 10−4 M concentrations of pheromone components and analogs. n = 6 replicates of cells, F = 39.40, p<0.001. (H) Inward current responses (left) and response profiles (right) of Xenopus oocytes expressing RVIII mutant/HassOrco in response to 10−4 M concentrations of pheromone components and analogs. n = 10 replicates of cells, F = 41.57, p<0.001. Data are presented as mean ± SEM. One-way ANOVA, Tukey HSD test are used. Mutation regions are highlighted in red.

https://doi.org/10.7554/eLife.29100.015

Site-specific mutations of HassOr14b and functional analysis

Based on the observation that the ligand selectivity of HassOr14b was changed only by replacement of the region VI or VIII, we chose these two segments of the receptor for further single-site mutations and functional analysis. Five amino acids (E188G, E196D, F232I, R262K, and R270K) in the region VI, and three (T355I, R395K, and A425K) in the region VIII were different between HassOr14b and HarmOr14b (Figures 6 and 8). We replaced each amino acid in turn by mutating each of the eight residues. Comparing with the wild type (Figure 7—figure supplement 2), we found that the mutant F232I was activated more by Z9-14:Ald than by Z9-16:Ald (Figure 8C), while the mutant T355I showed very strong responses to both Z9-16:Ald and Z9-14:Ald (Figure 8F). The other mutations showed largely the same selectivity as the wild type although with different values of the currents. Compared to the wild type, E196D and R262K still responded to Z9-16:Ald but showed a decrease in current (Figure 8B and D), E188G and R270K also responded to Z9-16:Ald but showed an increase in current (Figure 8A and E), while R395K and A425K exhibited the same response level to Z9-16:Ald and, to a minor extent, to Z9-14:Ald (Figure 8G and H).

Two-electrode voltage-clamp recordings of Xenopus oocytes with co-expressed site mutations and HassOrco, stimulated with pheromone components and analogs.

(A) Inward current responses (left) and response profiles (right) of Xenopus oocytes expressing E188G/HassOrco in response to 10−4 M concentrations of pheromone components and analogs. n = 6 replicates of cells, F = 52.32, p<0.001. (B) Inward current responses (left) and response profiles (right) of Xenopus oocytes expressing E196D/HassOrco in response to 10−4 M concentrations of pheromone components and analogs. n = 7 replicates of cells, F = 65.87, p<0.001. (C) Inward current responses (left) and response profiles (right) of Xenopus oocytes expressing F232I/HassOrco in response to 10−4 M concentrations of pheromone components and analogs. n = 8 replicates of cells, F = 90.65, p<0.001. (D) Inward current responses (left) and response profiles (right) of Xenopus oocytes expressing R262K/HassOrco in response to 10−4 M concentrations of pheromone components and analogs. n = 8 replicates of cells, F = 31.96, p<0.001. (E) Inward current responses (left) and response profiles (right) of Xenopus oocytes expressing R270K/HassOrco in response to 10−4 M concentrations of pheromone components and analogs. n = 8 replicates of cells, F = 56.13, p<0.001. (F) Inward current responses (left) and response profiles (right) of Xenopus oocytes expressing T355I/HassOrco in response to 10−4 M concentrations of pheromone components and analogs. n = 9 replicates of cells, F = 85.70, p<0.001. (G) Inward current responses (left) and response profiles (right) of Xenopus oocytes expressing R395K/HassOrco in response to 10−4 M concentrations of pheromone components and analogs. n = 7 replicates of cells, F = 40.19, p<0.001. (H) Inward current responses (left) and response profiles (right) of Xenopus oocytes expressing A425K/HassOrco in response to 10−4 M concentrations of pheromone components and analogs. n = 5 replicates of cells, F = 24.24, p<0.001. (I) Inward current responses (left) and response profiles (right) of Xenopus oocytes expressing (F232I + T355I)/HassOrco in response to 10−4 M concentrations of pheromone components and analogs. n = 9 replicates of cells, F = 79.03, p<0.001. Data are presented as mean ± SEM. One-way ANOVA, Tukey HSD test are used. Mutation sites are highlighted by red dots ‘’.

https://doi.org/10.7554/eLife.29100.018

We next constructed a mutant bearing the two substitutions (F232I and T355I) that affected the ligand selectivity of HassOr14b. This two-site mutant showed a robust response to Z9-14:Ald and a minor response to Z9-16:Ald, reproducing the characteristic selectivity of HarmOr14b (Figure 8I).

Discussion

Characteristics of chemosensory receptor binding sites are emerging for vertebrate ORs, which are seven transmembrane-spanning G-protein-coupled receptors, but less is known about insect ORs (Kato and Touhara, 2009; Ramdya and Benton, 2010). In this study, we identify HassOr14b as the PR tuned to Z9-16:Ald, the major sex-pheromone component of H. assulta. Its ortholog HarmOr14b is specific for Z9-14:Ald in H. armigera and we further demonstrate that two single-point mutations, F232I and T355I, located in the intracellular domains of the receptor, together determine the functional shift between orthologs in the two closely related species.

Novel identified PR and the different combinations with other PRs

As the number of different kinds of OSNs in three types of sensilla is related to the expression level of the corresponding PRs, the characteristics and abundance of the sensilla thus can provide reliable information for identifying PRs’ function. The previous studies clarified that the C type sensilla responding to Z9-16:Ald are predominant in male antennae of H. assulta (Wu et al., 2013; Wu et al., 2015; Xu et al., 2016), the PR tuning to Z9-16:Ald must be highly expressed in male antennae. We compare the expression level of all candidate PR genes in the male antennae of H. assulta and H. armigera. We found that HassOr14b is the most highly expressed in H. assulta, while HarmOr14b has relatively low expression level in H. armigera. The functional study confirms that HassOr14b is specifically tuned to Z9-16:Ald, while its ortholog HarmOr14b is specifically tuned to Z9-14:Ald. This suggests that the two closely related species not only changed Or14b’s expressing level, but also altered its tuning selectivity. It is worth noting that inward currents of the oocytes expressing HassOr14b induced by Z9-16:Ald were distinct but relatively low. It is common that the oocytes expressing some PRs are relatively weaker than others in responding to their ligands. However, their responding patterns in the oocyte system are generally representative of those in native OSNs. A clear dose-response curve to the most effective ligand is always helpful to confirm the receptor’s function.

The previous functional studies of HassOr14b did not find its activity by using the Xenopus system (Chang et al., 2016; Jiang et al., 2014). In this study, we re-cloned the sequence of HassOr14b by use of Q5 High-Fidelity DNA Polymerase (New England Biolabs, Ipswich, MA) and repeated again to verify the sequence by Sanger sequencing for 10 samples, and also compared with the sequence in the transcriptome data of H. assulta. Finally, we got the correct sequence, in which there are three amino acids different from the sequence in Jiang et al. (2014) (Figure 3—figure supplement 3). Moreover, we further analyzed the transcriptome data in H. assulta and confirmed that in the three different amino acids positions, there is no sequence polymorphism. We used the accurate sequence this time and characterized the function of HassOr14b, which is specifically tuned to Z9-16:Ald, the major sex pheromone component in H. assulta. Chang et al., 2016 also used the LA-Taq polymerase (TaKaRa, Shiga, Japan) Jiang et al. (2014) used before when they cloned the sequence, and there is one amino acid different in the 5’ ends from ours (Figure 3—figure supplement 3). By analyzing the transcriptome data in H. assulta, we confirmed that this amino acid position has no sequence polymorphism. Another difference between the two studies is the vector used in the expression system. Chang et al., 2016 used the pT7Ts vector, while we use the pCS2+ vector in the expression system. We suggest that the accuracy and integrity of the sequence is crucial to identify the function of the receptors. Moreover, the selection of the appropriate expression vector could be also important.

As the ligands of HassOr14b and the second abundant PRs, HassOr6 and HassOR16 are all included in the responding spectrum of the C type sensilla, we further investigated the expressing sites of the three PRs in the sensilla. HassOr14b is co-localized with HassOR6 or HassOR16 in the neighboring neurons in the same sensilla, while HassOr6 and HassOr16 are always expressed in different sensilla, which is different from the previous study (Chang et al., 2016). Our results indicate that there are different combinations of the PRs in the C type sensilla, which is consistent with the previous single sensillum recording results that there are subtypes in the type C sensilla (Xu et al., 2016). To the best of our knowledge, this is the first study that shows the various combinations of PRs were the molecular basis for the different sensilla subtypes in moth species.

Two amino acids located in the intracellular domains together determine the OR selectivity

Insects use olfactory receptors to discriminate amongst thousands of volatiles or pheromones (Kaupp, 2010). Insect ORs require the co-expression of a ligand-selective OR and a universal odorant co-receptor (Orco) to form ligand-gated ion channels (Missbach et al., 2014; Vosshall and Hansson, 2011). In the absence of data on the crystalline structure of insect ORs, the relationship between structure and function in these molecules is elusive. By amino acid covariation across insect Orcos and ORs, Hopf et al. constructed the first 3D models of D. melanogaster ORs (Hopf et al., 2015). However, this provided only an indirect insight into protein structure (Carraher et al., 2015). Previous site-directed mutagenesis studies performed to probe OR specificity, mainly focused on the transmembrane domains (TMDs) and extracellular loops (ECLs), based on the assumption that the TMDs and ECLs of the OR form the ligand-binding pocket (Guo and Kim, 2010). Leary et al. reported that a single amino acid mutation located in the predicted third TMD could change the ligand specificity of a PR between that of the Asian corn borer and that of the European corn borer (Leary et al., 2012). Pellegrino et al. showed that a single natural polymorphism of D. melanogaster Or59B in the third transmembrane domain altered DEET sensitivity (Pellegrino et al., 2011). In A. gambiae, a single mutation of AgOr15 at the interface between ECL2 and TMD4, produced large changes in responses to odors (Hughes et al., 2014). To address the relationship between OR-Orco structure and function, several recent studies showed that some amino acid residues in the OR or Orco were essential for channel activity of the heteromeric insect OR-Orco complex (Kumar et al., 2013; Nakagawa et al., 2012; Turner et al., 2014).

The different ligand selectivities of HassOr14b and HarmOr14b provide a convenient system in which to study structure-function relationships of PRs. Comparing the whole amino acid sequences of the two orthologous receptors, we identified two regions that were responsible for their selectivity. This new method is convenient and efficient, particularly for functional comparisons between orthologous or paralogous genes with many differing amino acids. By further replacing single amino acids in the two regions, we finally detected two single-point mutations, T355I and F232I responsible for the different ligand selectivities of HassOr14b and HarmOr14b. It is for the first time to find that the two mutation sites in the intracellular domains (ICDs) rather than in the TMDs and ECLs were involved in determination of ligand selectivity. We suggest two possible explanations for the role of ICDs. First, the binding site of ligand-specific ORs, such as PRs, may have a complex structure, which involves TMDs (Leary et al., 2012), ECLs (Hughes et al., 2014) and ICDs. Alternatively, ICDs may be involved in the specific interactions of the PR with the related G proteins. To relay the signal into the cell interior, binding of an extracellular molecule to an OR is tightly followed by binding of the receptor to a trimeric G protein inside the cell (Ignatious Raja et al., 2014; Wicher et al., 2008). Elucidation of the details of the structural and functional mechanisms of ORs must await further study.

Implications for the modulation and evolution of OR selectivity

Animal nervous systems are shaped by shifting environmental selection pressures to perceive and respond to new sensory cues (Prieto-Godino et al., 2017). The olfactory systems found in all animals have nearly the same design features, which give olfaction a considerable flexibility for signaling to evolve. Since the central odor processing is relatively conserved, new olfactory pathways tend to evolve from the peripheral changes (Galizia and Rössler, 2010; Prieto-Godino et al., 2017). How mutations in olfactory receptors change the olfactory responses of animals and eventually impact on the evolution of animal behavior is crucial but remains unclear.

In moth species, the co-evolution of pheromones produced by females and their detection by males present a paradox. Under stabilizing selection, variation of the female pheromone blend is limited, and the males typically prefer the most common pheromone blends (Groot et al., 2016; Roelofs et al., 2002). The ω value for all clusters of PRs analyzed in this study are less than 1, suggesting that PRs would be subjected to purifying selection. However, at the same time, the male moths need to have a degree of plasticity to adapt to changes in signal structures associated with speciation. Site-directed mutagenesis and functional analyses could validate how many amino acid substitutions are required to alter a PR’s selectivity.

Based on previous studies, suggesting that H. assulta is ancestral to H. armigera (Cho et al., 2008; Fang et al., 1997), we tried to reproduce the assumed evolutionary process that mutated HassOr14b into HarmOr14b. Most of the regional replacements and site mutations did not change the ligand selectivity of HassOr14b, indicating the functional stability of this PR. Only F232I and T355I substitutions produced a large change of the ligand selectivity of Or14b, from Z9-16:Ald in H. assulta to Z9-14:Ald in H. armigera. The former site mutation produced a small shift from Z9-16:Ald to Z9-14:Ald in the response spectrum, the latter extended and strengthened the responses to both chemicals, while the two mutations together generated a complete functional shift from Z9-16:Ald to Z9-14:Ald. These results indicate that the substitutions of a few key amino acids are able to greatly change PR selectivity, laying the molecular foundations for PR plasticity. Moreover, it seems that at least two steps, involving in the functional extension and shift, are required for a major functional change of HassOr14b, each step with a single point mutation. In the course of speciation, the functional change of ORs could be a process with multiple amino acid mutations, a few making drastic changes and many making small modifications or even no change in function.

The two closely related species H. assulta and H. armigera are one of the ideal study systems for pheromone communication. They share two chemicals, Z9-16:Ald and Z11-16:Ald, as their principal sex pheromone components but with reverse ratios. Males possess sensitive olfactory systems to detect conspecific sex pheromone blends. Peripheral coding of the binary blends with reversed ratios is mainly attributed to two group of specific OSNs in separate antennal sensilla with reverse population sizes, which reflect different expression levels of related PRs (Wu et al., 2013). In this study, we discover that HassOr14b, the highest expressed PR in male antennae of H. assulta is tuned to Z9-16:Ald, the major component of the sex pheromone of H. assulta, while its ortholog HarmOr14b is tuned to Z9-14:Ald in H. armigera, which provides us an ideal model to study the determinants of OR selectivity. We systematically identify two single-point mutations, F232I and T355I, located in the intracellular regions of HassOr14b that together determine the functional shift to its ortholog, HarmOr14b, in H. armigera. The peripheral modifications of the two closely related species took place in both PR expression level and PR tuning selectivity. These findings not only help us specifically understand the evolution of the two Helicoverpa species, but also provide new insights into the structure and function of cell membrane receptors.

Materials and methods

Insects

H. assulta and H. armigera were originally collected as larvae in tobacco fields in Zhengzhou, Henan province of China, and were reared at the Institute of Zoology, Chinese Academy of Sciences, Beijing. The larvae were fed with an artificial diet, mainly composed of wheat germ, yeast and chili for H. assulta, wheat germ, yeast and tomato paste for H. armigera. Rearing took place at a temperature of 26 ± 1°C with a photoperiod of 16L:8D and 55–65% relative humidity. Male and female pupae were placed in separate cages for eclosion. A 10% honey solution was used as the diet for adults. Virgin adults at 1–3 days old were used in all experiments.

Xenopus laevis

All procedures were approved by the Animal Care and Use Committee of the Institute of Zoology, Chinese Academy of Sciences for the care and use of laboratory animals. Female X. laevis were provided by Prof. Zhan-Fen Qin from Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, and reared with pig liver as food in our laboratory. A total of nine healthy naive X. laevis with 18–24 months of age were at the time of the experiment. They were group housed in the box with purified water in 20 ± 1°C. The surgery was performed following the reported protocols (Nakagawa and Touhara, 2013). X. laevis were anesthetized by bathed in the mixture of ice and water in 30 min, and the oocytes were surgically collected before experiments.

Sequencing and PR genes of H. assulta expression analysis

Total RNA was extracted using the TRIzol reagent (Invitrogen, Carlsbad, CA) and treated with RNase-free DNase I. Poly(A) mRNA was isolated using oligo dT beads. First-strand complementary DNA was generated using random hexamer-primed reverse transcription, followed by synthesis of the second-strand cDNA using RNaseH and DNA polymerase I. Paired-end RNA-seq libraries were prepared following Illumina’s protocols and sequenced on the Illumina HiSeq 2000 platform (San Diego, CA). The RNA-seq reads were mapped using Bowtie2 (Langmead and Salzberg, 2012). Gene expression levels were measured using the reads per kb per million mapped reads criterion (FPKM). FPKM values were calculated by custom python script (https://github.com/ningchaozky/fpkm-calculate-from-bam-or-sam-.git [Ning, 2017]; copy archived at https://github.com/elifesciences-publications/fpkm-calculate-from-bam-or-sam-.git)Ning, 2017. Only genes with a FPKM >1 and coverage more than 0.6-fold of transcripts were used for further analysis. Differentially expressed genes were detected using the DEGseq (RRID: SCR_008480) (Wang et al., 2010), which was constructed based on the Poisson distribution and eliminated the influences of sequencing depth and gene length. Annotation of PR genes was performed by NCBI blastx against a pooled insect PR database and then the expression was extracted from the DEGseq result.

Phylogenetic analysis

Phylogenetic analysis of PRs was performed based on amino acid sequences contained in reports of PRs of Noctuidae. The phylogenetic tree was constructed using the MEGA6.0 program (RRID: SCR_000667) with neighbor-joining phylogeny using the p-distances model (Tamura et al., 2013). Node support was assessed using a bootstrap procedure based on 1000 replicates. The ratios of nonsynonymous to synonymous substitutions (dN/dS) were computed using DnaSP version 5.10 (RRID: SCR_003067) (Librado and Rozas, 2009).

RNA isolation and cDNA synthesis

The antennae from three-day-old virgin adults were dissected and immediately collected into a 1.5 mL Eppendorf tube, containing liquid nitrogen, and stored at −80°C until use. Total RNA was extracted by QIAzol Lysis Reagent following the manufacturer’s protocol (including DNase I treatment). RNA quality was checked with a spectrophotometer (NanoDrop 2000, Wilmington, DE). The single-stranded cDNA templates were synthesized using 2 μg total RNAs from various samples with 0.5 μg oligo (dT) 15 primer (Promega, Madison, WI), heated to 70°C for 5 min to melt the secondary structure within the template, then using M-MLV reverse transcriptase (Promega) at 42°C for 1 hr, and stored at −20°C.

Quantitative real-time PCR

qPCR was performed on an Mx3005P qPCR System (Agilent Technologies, Palo Alto, CA) with SYBR Premix Ex Taq (TaKaRa, Shiga, Japan). The gene-specific primers to amplify an 80–150 bp product were designed by Primer-BLAST (http://www.ncbi.nlm.nih.gov/tools/primer-blast/), and are listed in Supplementary file 1. The qPCR reaction was: 10 s at 95°C, followed by 40 cycles of 95°C for 5 s and 60°C for 31 s, followed by the measurement of fluorescence during a 55°C to 95°C melting curve to detect a single gene-specific peak, and to check the absence of primer dimer peaks. The product was verified by nucleotide sequencing. 18S ribosomal RNA (GenBank number: EU057177.1) was used as the control gene. Each reaction was run in triplicate (technical replicates) and the means and standard errors were obtained from three independent biological replicates. The relative copy numbers of PR genes were calculated according to the 2–ΔΔCt method (Livak and Schmittgen, 2001).

Cloning of the candidate pheromone receptor of H. assulta and H. armigera

Based on the full-length nucleotide sequences of PRs in H. assulta or H. armigera (GenBank numbers are listed in Supplementary file 2), specific primers were designed and are reported in Supplementary file 1. All amplification reactions were performed using Q5 High-Fidelity DNA Polymerase (New England Biolabs). The PCR conditions for the PRs were: 98°C for 30 s, followed by 30 cycles of 98°C for 10 s, 50°C for 30 s and 72°C for 1 min, and extension at 72°C for 2 min. Templates were obtained from male or female antennae of H. assulta or H. armigera. The sequences were verified by both the Sanger sequencing for 10 samples, and the transcriptome data.

In situ hybridization

Two-color double in situ hybridizations were performed following protocols reported previously (Krieger et al., 2002; Ning et al., 2016). The sense and antisense primers were used to synthesize the gene-specific probes from the open-reading frames (Supplementary file 1). Both digoxin (Dig)-labeled and biotin (Bio)-labeled probes were synthesized by DIG RNA labeling Kit version 12 (SP6/T7) (Roche, Mannheim, Germany), with Dig-NTP or Bio-NTP (Roche, Mannheim, Germany) labeling mixture, respectively. RNA probes were subsequently fragmented to 300 nt by incubation in carbonate buffer. Antennae were dissected from 2- to 4-day-old male moths, embedded in JUNG tissue freezing medium (Leica, Nussloch, Germany) and frozen at −80°C until use. Sections (12 µm) were prepared with a Leica CM1950 microtome at −22°C, then mounted on SuperFrost Plus slides (Thermo Scientific, Waltham, MA). After a series of fixing and washing procedures, 100 μL hybridization solution (Boster, Wuhan, China) containing both Dig and Bio probes was placed onto the tissue sections. A coverslip was added and slides were incubated in a humid box at 55°C overnight. After hybridization, slides were washed twice for 30 min in 0.1 × saline sodium citrate (SSC) at 60°C, treated with 1% blocking reagent (Roche, Mannheim, Germany) in TBST (100 mM Tris, pH = 7.5, 150 mM NaCl with 0.03% Triton X-100) for 30 min at room temperature, and then incubated for 60 min with anti-digoxigen (Roche, Mannheim, Germany) and Strepavidin-HRP (PerkinElmer, Boston, MA).

Hybridization signals were visualized by incubating the sections for 30 min with HNPP/Fast Red (Roche, Mannheim, Germany), followed by three 5 min washes in TBS with 0.05% Tween-20 (Tianma, Beijing, China) at room temperature, with shaking. The sections were incubated with Biotinyl Tyramide Working Solution for 8 min at room temperature followed by the tyramide-signal amplification (TSA) kit protocols (PerkinElmer, Boston, MA). After three additional washings for 5 min in TBS with 0.05% Tween-20 at room temperature with shaking, sections were finally mounted in Antifade Mounting Medium (Beyotime, Beijing, China). All the sections were analyzed under a Zeiss LSM710 Meta laser scanning microscope (Zeiss, Oberkochen, Germany). Adobe Illustrator CS6 (RRID: SCR_014198) (Adobe systems, San Jose, CA) was used to arrange figures only to adjust brightness and contrast.

Construction of the mutation sequences

To generate the regional mutations, we first cloned each fragment of the sequences. The mutation fragment was cloned using the primer of Mutant-F and Mutant-R, with the cDNA of H. armigera. The other parts were cloned using the primer of HassOr14b-F/HassOr14b-Fragment1-R, which generated the first fragment, and HassOr14b-Fragment2-F/HassOr14b-R, which generated the second fragment, with the cDNA of H. assulta. The mutation fragment had 25–60 bp overlap sequences with the other two fragments. The conditions were: 98°C for 30 s, followed by 25 cycles of 98°C for 10 s, 52°C for 30 s and 72°C for 30 s, and extension at 72°C for 2 min. Then we used the primers of HassOr14b-F/HassOr14b-R, with the mixture of purified fragment products as the template, to generate the regional mutation sequences. The conditions were: 98°C for 30 s, followed by 20 cycles of 98°C for 10 s, 52°C for 30 s and 72°C for 90 s, and extension at 72°C for 2 min. The construction diagram was presented in Figure 7—figure supplement 1. For the site mutations, we used the primer of HassOr14b-F/mutation1 R to generate the first fragment, and the mutation2-F/HassOr14 b-R to generate the second fragment. Then we used the primer of HassOr14b-F/HassOr14b-R, with the mixture of purified fragment products as the template, to generate the site mutation sequences. The conditions were the same as for the construction of regional mutation sequences. The primers are listed in Supplementary file 1.

Receptor functional analysis

The full-length coding sequences of PRs and mutations were first cloned into pGEM-T vector (Promega) and then subcloned into pCS2+ vector. cRNAs were synthesized from linearized modified pCS2+ vectors with mMESSAGE mMACHINE SP6 (Ambion, Austin, TX). Mature healthy oocytes were treated with 2 mg mL−1 of collagenase type I (Sigma-Aldrich, St. Louis, MO) in Ca2+-free saline solution (82.5 mM NaCl, 2 mM KCl, 1 mM MgCl2, 5 mM HEPES, pH = 7.5) for 20 min at room temperature. Oocytes were later microinjected with 27.6 ng PR cRNA and 27.6 ng Orco cRNA. Distilled water was microinjected into oocytes as a negative control. Injected oocytes were incubated for 3–5 days at 16°C in bath solution (96 mM NaCl, 2 mM KCl, 1 mM MgCl2, 1.8 mM CaCl2, 5 mM HEPES, pH = 7.5) supplemented with 100 mg mL−1 gentamycin and 550 mg mL−1 sodium pyruvate. Whole-cell currents were recorded with a two-electrode voltage clamp. Intracellular glass electrodes were filled with 3 M KCl and had resistances of 0.2–2.0 MΩ. Signals were amplified with an OC-725C amplifier (Warner Instruments, Hamden, CT) at a holding potential of −80 mV, low-pass filtered at 50 Hz and digitized at 1 kHz. Data acquisition and analysis were carried out with Digidata 1322A and pCLAMP software (RRID: SCR_011323) (Axon Instruments Inc., Foster City, CA). Dose-response data were analyzed using GraphPad Prism (RRID: SCR_002798 6) (GraphPad Software Inc., San Diego, CA).

Data analysis

Response values are indicated as mean ± SEM. Data were square-root transformed and differences were considered significant when p<0.05. n represents number of sections in all cases. One-way ANOVA and Tukey HSD tests with two distribution tails were performed using the Statistical Program for Social Sciences 22.0 (RRID: SCR_002865) (IBM Inc., Armonk, NY).

References

  1. 1
  2. 2
  3. 3
  4. 4
    Structure of the pheromone communication channel in moths
    1. RT Cardé
    2. KF Haynes
    (2004)
    In: RT Cardé, JG Millar, editors. Advances in Insect Chemical Ecology. Cambridge, UK: Cambridge University Press. pp. 283–332.
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10
  11. 11
  12. 12
  13. 13
  14. 14
  15. 15
  16. 16
  17. 17
  18. 18
  19. 19
  20. 20
  21. 21
  22. 22
  23. 23
  24. 24
  25. 25
  26. 26
  27. 27
  28. 28
  29. 29
  30. 30
  31. 31
  32. 32
  33. 33
  34. 34
  35. 35
  36. 36
  37. 37
  38. 38
  39. 39
  40. 40
  41. 41
  42. 42
  43. 43
  44. 44
  45. 45
  46. 46
  47. 47
  48. 48
  49. 49
  50. 50
  51. 51
  52. 52
  53. 53
  54. 54
  55. 55
  56. 56
  57. 57
  58. 58
  59. 59
  60. 60
  61. 61
  62. 62
  63. 63
  64. 64
  65. 65
    Pheromones and Animal Behaviour: Communication by Smell and Taste
    1. TD Wyatt
    (2003)
    Cambridge, UK: Cambridge Unvierstiy Press.
  66. 66
  67. 67
  68. 68
  69. 69
  70. 70
  71. 71
  72. 72

Decision letter

  1. Marcel Dicke
    Reviewing Editor; Wageningen University, Netherlands

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

[Editors’ note: the authors were asked to provide a plan for revisions before the editors issued a final decision. What follows is the editors’ letter requesting such plan.]

Thank you for submitting your article "Two single-point mutations shift the ligand selectivity of a pheromone receptor between two sister moth species" for consideration by eLife. Your article has been reviewed by Fred Gould (Reviewer #1), Astrid Groot (Reviewer #2), and Christer Loefstedt (Reviewer #3), and the evaluation has been overseen by a Reviewing Editor and Ian Baldwin as the Senior Editor.

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this letter to prompt a response from you concerning the serious concerns of the reviewers. Please address these concerns in a letter, which we will have the Board and reviewers consider before a binding recommendation is made.

Your manuscript has been evaluated by three experts in the field. Although they found your results interesting, they have also identified important concerns related to the experimental results in connection to previous studies by your and other groups. As a consequence, the story that you tell in the manuscript is flawed. For the manuscript to be acceptable for publication in eLife, you need to connect better to the literature and include caveats such as the suitability of the Xenopus system that does not seem to work well for HR14b in Hass and the fact that there are other receptors for Z9-16Ald in Hass and that Z9-16Ald may be antagonistic in both species. How does this affect the conclusions that you can draw from your experiments? Also the in situ hybridization images are not convincing and do not support previous publications. You will find the detailed comments by the reviewers below. In the light of these evaluations, the manuscript cannot be accepted at this moment. If you feel that you can effectively address the most critical concerns of the reviewers, please send us your responses for further evaluation but the Board and reviewers.

Reviewer #1:

General: Overall this is a very interesting paper that is one of only a couple of papers that have identified amino acid changes if a moth pheromone receptor that change specificity. This is a very important finding because it could show that the evolution of new sexual communication systems in moths could evolve based on very simple genetic changes. That said, I do have some concerns about specific issues.

Specific:

1) As far as I can tell only one peer reviewed paper indicates the Z9-14Ald is an active positive part of the Harm pheromone. In another paper this compound is found to be an antagonist in both Hass and Harm. This should be made clear to the reader because unlike some other systems where there is a clear change from using one compound to using another in two closely related species, in this case the selective pressure to evolve the receptor in Harm for Z9-14Ald is not very clear.

2) Earlier research papers that have searched for activity of HR14b in Hass have failed to find any activity. This includes a recent paper by some of the authors of the current paper. It would be useful for the authors to explain why they did not find activity in their other recent paper, but that they now find it.

3) Looking at the plots in the figures that show "current (nA)" on the Y-axis, it becomes clear that at least with the Xenopus system, the level of activity of the Hass HR14b receptor is much lower than for the Harm HR14b receptor. Whereas the response to Z9-14Ald in Harm is over 400 and for Z9-16Ald is around 50, the response of the Hass HR14b receptor to even Z9-16Ald (to which it is specific) is only about 60. Clearly, this oocyte system is not very efficient for Hass HR14b. This may be why previous studies didn't find any activity. The authors should acknowledge this.

4) This lack of sensitivity of the HR14b of Hass is problematic for interpreting the results of single mutations to the Hass HR14b receptor. Clearly, the T3551 mutation dramatically increases the overall response of HR14b while negating specificity. This makes it hard to interpret the interaction of the two mutations.

5) There is overlap between the paper Olfactory perception and behavioral effects of sex pheromone gland components in Helicoverpa armigera and Helicoverpa assulta, Meng Xu et al. 2016 and the current paper. This overlap should be made more clear.

Reviewer #2:

This manuscript describes the functional characterization of an olfactory receptor in the noctuid moth Helicoverpa assulta, HassOR14b, that is tuned to the major sex pheromone of this species, Z9-16:Ald, while its ortholog HarmOR14b in H. armigera is tuned to Z9-14:Ald, the secondary sex pheromone component of this species (the main sex pheromone component Z11-16:Ald, which is perceived through HarmOR13). Through a series of mutagenesis experiments, the authors convincingly show that two single point mutations, F232I and T355I, in HassOR14b changes the sensitivity of this receptor from Z9-16:Ald to Z9-14:Ald. These mutations correspond to the HarmOR14b amino acid sequences at these sites. In additions, the authors show with in-situ hybridizations that HassOR14b is co-localized with HassOR6 and HassOR16.

I only have a few minor, but essential comments on the text that I think can easily be addressed:

1. I miss citation to the recent article by De Fouchier et al. in Nature Communications (DOI: 10.1038/ncomms1570910)

2. Since HarmOR14b is tuned to Z9-14:Ald, I think it's important that the authors write 1-2 sentences on how this component is a minor (but essential) sex pheromone component in H. armigera. In the current text it seems that the sex pheromone system of these two species is a two-component system with similar reverse ratios as in the two pheromone strains of Ostrinia nubilalis, while the pheromone blend of H. armigera (and also H. assulta) is a bit more complex than just a two-component blend.

3. The authors give dN/dS ratios (ω) for the different ORs (Results, subsection “Phylogenetic analysis of candidate PRs”), reasoning that ω < 1 indicates puryfying selection, while ω > 1 indicates positive selection. As they found a ω = 0.17 for cluster OR14b, this thus indicates puryfying selection. However, in the discussion the authors do not come back to this result and instead write "The female moth produces a pheromone blend of several components, stabilized by strong selection pressure against any change in such blends (Roelofs et al., 2002). This requires an equivalent stability from the male moths to detect the same species-specific pheromones, but at the same time should allow for a degree of plasticity to adapt to changes in pheromone structures associated with speciation." This part needs to be revised, as 'stabilized by strong selection pressure' and 'equivalent stability' are strangely used. Also, the latter part "should allow for a degree of plasticity" comes across as hand waving. Similarly, "In the course of speciation, the functional change of ORs is a gradual process with multiple amino acid mutations, a few making drastic changes and many making small modifications or even no change in function" need to be revised, as 'gradual changes' contradicts 'a few making drastic changes' and as a whole this sentence doesn't make sense.

4. Discussion, subsection “Novel identified PR and the different combinations with other PRs”, first paragraph: HassOR14b should be "HassOR16 in the same sensilla". Do the authors know where the orthologous ORs are localized in H. armigera? Is HarmOR14b also co-localized with HarmOR6 or HarmOR16? Can I deduce from the intro information that these sensilla are the type C sensilla?

5. Discussion, subsection “Novel identified PR and the different combinations with other PRs”, second paragraph:: Where are these amino acid residues located? I think it's important to specify this, especially because this is the first study (right?) where amino acid sequence changes in the intracellular domains (ICDs). I would also like to read a bit more on how the authors think these changes may alter the function. They give two very short explanations in the next paragraph, but what do they mean with 'complex deep structure'?

Reviewer #3:

Yang et al. report that two amino acid substitutions in intracellular domains may account for the difference in ligand specificity between HarmOr14b and HassOr14b. The conclusion is based on a series of mutagenesis experiments. The results are interesting but "the story" is not ready for publication. A number of major issues are listed below. In addition, I think that the discussion of how mutations in the intracellular domains may influence ligand specificity (subsection “Two amino acids located in the intracellular domains together determine the OR selectivity”, last paragraph) is speculative and lacks both references and a possible mechanism.

The study is mainly based on the authors´ finding that HassOR14b is responsive to Z9-16:Ald whereas HarmOR14b is responsive to Z9-14:Ald. However, according to a previous study from the same laboratory (Jiang et al., 2014) and the study by Chang et al. (2015), HassOR14b did not respond to any tested compounds including Z9-16:Ald. This needs to be clarified. The authors do not mention about this discrepancy and do not even cite the Chang et al. paper:

Jiang, X.-J., Guo, H., Di, C., Yu, S., Zhu, L., Huang, L.-Q., Wang, C.-Z. (2014). Sequence similarity and functional comparisons of pheromone receptor orthologs in two closely related Helicoverpa species. Insect Biochemistry and Molecular Biology 48, 63-74.

Chang, H., Guo, M., Wang, B., Liu, Y., Dong, S., and Wang, G. (2016). Sensillar expression and responses of olfactory receptors reveal different peripheral coding in two Helicoverpa species using the same pheromone components. Scientific reports, 6, 18742.

In the phylogenetic tree, the authors only included Heliothine species in OR13 cluster, but not the orthologues from other species, thus the calculation of dN/dS value is biased. Moreover, the authors say the dN/dS value of OR11 cluster is larger than 1.8, this is inconsistent with the previous findings that dN/dS value of this cluster is low (around 0.1):

Zhang, D. D., and Löfstedt, C. (2013). Functional evolution of a multigene family: orthologous and paralogous pheromone receptor genes in the turnip moth, Agrotis segetum. PLoS One, 8(10), e77345.

Zhang, Y. N., Zhang, J., Yan, S. W., Chang, H. T., Liu, Y., Wang, G. R., and Dong, S. L. (2014). Functional characterization of sex pheromone receptors in the purple stem borer, Sesamia inferens (Walker). Insect molecular biology, 23(5), 611-620.

The in situ hybridization images are not convincing and this is problematic as the results do not correspond to what was reported in Chang et al. (2015). These authors claimed that HassOR6 and HassOR16 are localise in the same sensillum (Chang et al., 2015).

Discussion subsection “Novel identified PR and the different combinations with other PRs”, first paragraph: 'This suggests that the ancestor of the two sister species not only changed OR14b's expressing level, but also altered its tuning selectivity to meet the species specific demands': The authors did not compare the expression levels of HassOR14b and HarmOR14b so the statement is not supported. In addition, it is not clear to me how the authors can conclude anything about the expression levels in the ancestor. Based on analysis of the contemporary species (which at some point had a common ancestor) we can at the best conclude that expression levels are different in these two species. The reasoning further involves a teleological argument, i.e. that the species that evolved from the common ancestor had some "specific demands". This is not how evolution works.

In the subsection “Quantitative real-time PCR”: the authors write that the reference gene is 18s rRNA, but the GenBank number provided is actually the actin gene. This is confusing.

In the subsection “Construction of the mutation sequences”: The description of the construction strategy would benefit from a diagram visualizing the different steps.

[Editors’ note: formal revisions were requested, following approval of the authors’ plan. After the authors submitted their revised paper, further revisions were requested prior to acceptance, as described below]

Thank you for submitting your article "Two single-point mutations shift the ligand selectivity of a pheromone receptor between two closely related moth species" for consideration by eLife. Your article has been reviewed by Fred Gould (Reviewer #1), Astrid Groot (Reviewer #2), and Christer Loefstedt (Reviewer #3), and the evaluation has been overseen by a Reviewing Editor and Ian Baldwin as the Senior Editor.

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

The reviewers have appreciated your effective revision of the manuscript and the extensive explanation in the rebuttal. They conclude that your study provides very interesting results on how small differences in pheromone receptors may influence ligand selectivity. Still some important issues remain and I invite you to address these comments by the reviewers and prepare a second revision of the manuscript. Specific attention should be paid to sentences where you refer to the literature on odour reception and genetic mechanisms and evolutionary consequences. Some of these sentences are highlighted in the reviews, while others are not. For example, what is meant with 'a few steps' in the last sentence of the Abstract?

Reviewer #1:

The eLife editor summarized my concerns well as "For the manuscript to be acceptable for publication in eLife, you need to connect better to the literature and include caveats such as the suitability of the Xenopus system that does not seem to work well for HR14b in Hass and the fact that there are other receptors for Z9-16Ald in Hass and that Z9-16Ald may be antagonistic in both species. How does this affect the conclusions that you can draw from your experiments? "

The authors have made useful changes but I think a little more discussion of some of the issues with the Xenopus system in this specific case is warranted because the same issue is likely to arise in future studies.

I realize that in the current Abstract and in most of the manuscript the authors avoid discussing the relevance of their findings to the evolution of the two species of Helicoverpa. For example, in the Abstract they state that "We conclude that species-specific changes in the tuning specificity of the PRs of male moths could be achieved with just a few steps". As long as the authors don't indicate that their findings help us to specifically understand the evolution of the two Helicoverpa moths, I think they are on solid ground. The two mutations they study certainly change the tuning specificity.

It still would be good if the authors could elaborate on the fact their findings are of most interest in terms of neurophysiology and are not a direct commentary on the evolution of these specific moths.

Reviewer #2:

Overall, I'm very impressed with all the work the authors have done and also with their revision. I have a few remaining questions for the authors:

The numbering of the figures was a bit confusing, I'm not sure which figures are now supposed to be supplementary figures, so I'll go with the complete names of the figures.

1) In Figure 3 the authors show that HasOR14b responds to Z9-16:Ald, but when comparing the different graphs, the y-axis in Figure 3 is in a different (smaller) scale than the y-axis in Figure 3—figure supplement 1, which shows that HassOr6 actually responds more to Z9-16:Ald (300 nA) than HassOr14b (70 nA). Such a difference in scaling is also present in Figure 8; HassOr14b shows a current response to Z9-16:Ald of 70 nA again, and/but HarmOr14b shows a current response of 30-40 (?) nA. I do understand that HarmOr14b mostly responds to Z9-14:Ald, but the response to Z9-16:Ald may also be important, no? (In this respect it is indeed very impressive that the authors found an almost 10x stronger response when mutating T335I (see Figure 7—figure supplement 2). Which relates to my next question:

2) As Z9-16:Ald is also an important (albeit minor) sex pheromone component of H. armigera, with what receptor(s) do the authors think H. armigera is perceiving this component?

3) In Figure 2 the authors show that HassOR6 is highly expressed in males, and a little in females. Which makes sense to me, as females probably smell their own pheromone as well. I don't think that sex-specific expression is a necessary criterium for an OR to be a sex pheromone OR.

4) The authors have responded to previous comments on why previous functional assays with HassOr14b didn't work by writing that the sequence had not been correct in previous work and that this is now corrected. In Figure 3——figure supplement 3 they specify which 3 amino acids have been corrected (Yang et al. is the current sequence used). In checking these 3 amino acids in Figure 6, I found them in RIII, in RVII and in RVIII. I find it hard to determine whether one of these 3 SNPs were among the ones tested in Figure 7—figure supplement 2. Still, I think the authors make a convincing case that the right sequence together with the selection of the appropriate expression vector is crucial to identify the function of receptors.

5) Hopefully Figure 7—figure supplement 2 is not a supplementary figure, as I think this figure is a great way of showing the results of the single mutations.

6) As for the in situ hybridization pictures, these are clear to me, but I leave this to the more experienced reviewer(s) to decide whether this is sufficiently clear.

Reviewer #3:

In summary, I think that the revision represents a considerable improvement of the manuscript and the authors have to a large extent explained many of the mistakes and the missing information in the original submission. However, some of my previous comments/questions still apply.

In the rebuttal letter the authors write:

'Basically there are two major concerns: (Baker et al., 2004) the suitability of the Xenopus system […] (Boo et al., 1995) Contradiction of our in situ hybridization results'.

In my opinion this is not correct. At least not when it comes to my concerns. My major concern was that the authors avoided mentioning the inconsistency in the response of wild type HassOR14b compared to previous publications. In the current version, the authors explain that the gene sequences were not correct in previous publications. However, the explanation is not straight forward:

- The sequences in Chang et al., 2016 were based on the transcriptome data from the previous paper (Zhang et al., 2015) from the same group. There's a single amino acid difference of HassOR14b sequence between the current manuscript and the sequence in Chang's paper. Could it simply be explained by the use of different polymerases? How could the authors exclude the possibility of polymorphism? Although possible, it's not very likely that this one amino acid difference would totally abolish the response to Z9-16:Ald (No response to Z9-16:Ald or any other ligand in Chang et al. 2016) but significant response to Z9-16:Ald in the current study (70 nA – Figure 8). The authors of the present manuscript can of course not take the responsibility of any possible flaws in a study published by another group but the discrepancy requires an explicit comment in the Discussion.

Zhang, J., Wang, B., Dong, S., Cao, D., Dong, J., Walker, W. B.,.… & Wang, G. (2015). Antennal transcriptome analysis and comparison of chemosensory gene families in two closely related noctuidae moths, Helicoverpa armigera and H. assulta. PloS one, 10(2), e0117054.

I do not understand how different expression vectors could cause the difference in sequence. The authors have to explain how. Otherwise it is just an ad hoc explanation that needs to be tested. It's the cRNAs that are injected and expressed in the oocytes. The resulting receptor proteins should have the same sequences despite of what the original expression vectors were. In addition, the pT7TS vectors have been widely and successfully used in oocyte recordings and OR studies. Why should this vector not be suitable for HassOR14b expression and oocyte recordings?

The authors write in the rebuttal and mention in the Discussion that they re-sequenced HassOr14b and came up with the "right sequence". This is a new (and important) result and as such it should be mentioned in the Results section and the relevant methods and materials should be described in Materials and methods before the discrepancy with previous studies is discussed in the Discussion section.

As far as I understand reviewer #1 was not really questioning oocytes as a suitable system for OR de-orphanization in general, but said that the system may not be efficient for testing HassOR14b. In my opinion, however, a current of 60 nA is not too small to be noticed. Hence this is not the reason why Chang et al. did not find a response, but there seems to be a real difference in activity between the two studies.

It remains mysterious to me why the wild type HassOR14b responds so differently in this and previous studies and it's problematic that the authors did not mention this inconsistency in the first version. In addition, a number of other mistakes/errors were pointed out by me and the other reviewers. The authors have now corrected these errors but can we be sure that additional errors did not go unnoticed by the reviewers? Careful double checking appears necessary.

Some other points for your consideration.

1) 'Unlike general odorant receptors (ORs) that typically bind more than one ligand (de Fouchier et al., 2017), PRs are narrowly tuned to specific pheromone components (Grosse-Wilde et al., 2007).' This statement is not correct. There are both narrowly tuned PRs and broadly tuned PRs reported in the literature (Miura et al., 2010; Wanner et al. 2010; Zhang and Löfstedt 2015). Even in Grosse-Wilde et al., 2007 cited by the authors, it was proposed that 'there are two different designs of pheromone receptors'. The authors have to be more precise in their statements and more rigorous when citing the literature:

Miura, N., Nakagawa, T., Touhara, K., and Ishikawa, Y. (2010). Broadly and narrowly tuned odorant receptors are involved in female sex pheromone reception in Ostrinia moths. Insect biochemistry and molecular biology, 40(1), 64-73.

Wanner, K. W., Nichols, A. S., Allen, J. E., Bunger, P. L., Garczynski, S. F., Linn Jr, C. E., […] and Luetje, C. W. (2010). Sex pheromone receptor specificity in the European corn borer moth, Ostrinia nubilalis. PLoS One, 5(1), e8685.

Zhang, D. D., and Löfstedt, C. (2015). Moth pheromone receptors: gene sequences, function, and evolution. Frontiers in Ecology and Evolution, 3, 105.

2) 'For example, BmOr1 of.[…] to 1,3Z,6Z,9Z-19:H (Zhang et al., 2016).' As we suggested before, there's no point of listing these examples so this sentence can be removed.

3) Results and Discussion sections: ‘all the PRs and Orcos are subjected to purifying selection which is consistent with the previous studies (Zhang and Löfstedt, 2013; Zhang et al., 2014b)'. This is not a correct reference. Zhang and Löfstedt (2013) as well as other researchers (e.g. Leary et al., 2012) found that some PR clusters are under purifying selection while some are under relatively relaxed selective pressure. Since the authors used a limited number of moth species in the phylogenetic tree in the current study, it's too daring to draw the conclusion that 'all the PRs and Orcos are subjected to purifying selection'.

Leary, G. P., Allen, J. E., Bunger, P. L., Luginbill, J. B., Linn, C. E., Macallister, I. E., […] and Wanner, K. W. (2012). Single mutation to a sex pheromone receptor provides adaptive specificity between closely related moth species. Proceedings of the National Academy of Sciences, 109(35), 14081-14086.

4) In situ results: 'in some sensilla only HassOr14b was detected'. For the Discussion: Is this consistent with previous SSR data? Which gene is supposed to be expressed in the neighboring neuron?

5) The authors refer to the transcriptome a couple of times in the manuscript. Have the authors deposited the transcriptome data in a public database?

6) Discussion subsection “Implications for the modulation and evolution of OR selectivity”: 'The olfactory systems found in all animals have nearly the same design feathers, which give olfaction a considerably flexibility for signaling to evolve.' I don't understand this sentence. What does 'design feathers' mean? Should it be "features"?

7) The English still needs some attention throughout the manuscript before its scientific merits can be definitely evaluated.

https://doi.org/10.7554/eLife.29100.022

Author response

[Editors’ note: what follows is the authors’ plan to address the revisions.]

Your manuscript has been evaluated by three experts in the field. Although they found your results interesting, they have also identified important concerns related to the experimental results in connection to previous studies by your and other groups. As a consequence, the story that you tell in the manuscript is flawed. For the manuscript to be acceptable for publication in eLife, you need to connect better to the literature and include caveats such as the suitability of the Xenopus system that does not seem to work well for HR14b in Hass and the fact that there are other receptors for Z9-16Ald in Hass and that Z9-16Ald may be antagonistic in both species. How does this affect the conclusions that you can draw from your experiments? Also the in situ hybridization images are not convincing and do not support previous publications. You will find the detailed comments by the reviewers below. In the light of these evaluations, the manuscript cannot be accepted at this moment. If you feel that you can effectively address the most critical concerns of the reviewers, please send us your responses for further evaluation but the Board and reviewers.

Basically there are two major concerns: (Baker et al., 2004) the suitability of the Xenopus system for functional analysis of pheromone receptors like HassOr14b; (Boo et al., 1995) Contradiction of our in situhybridization results with Chang et al. (2016) in Scientific Reports. “Z9-16:Ald” you mentioned above could be “Z9-14:Ald”?

For the first major concern, the Xenopus laevis oocyte expression system is a valuable tool for the study of function of insect ORs, up to now, the function of many insect ORs have been characterized successfully. We think this system is also suitable for HassOr14b based on the following reasons:

1) The responding current to 10-4 M of Z9-16:Ald is about 70 nA, which is comparable to the current (80 nA) of the oocyte expressing HarmOr13 tuning to the major component, Z11-16:Ald in H. armigera. We have done many technical and biological replications, and the results are repeatable.

2) The oocytes expressing HassOr14b to Z9-16:Ald are retested within a dose–response paradigm, revealing a clear concentration dependency.

3) Other evidences also support the HassOr14b is the receptor of Z9-16:Ald in H. assulta: The expression level of HassOr14b is the highest among all the PRs in male antennae of H. assulta, and the OSNs responding to Z9-16:Ald is most abundant in male antennae of H. assulta; the localization patterns of HassOr14b with other PRs in male antennae of H. assulta are consistent with the single sensilla recording results.

Please also find a more detailed response to this question raised by the reviewer #1 later.

For the second major concern, especially raised by the reviewer #3, we provide three additional in situ hybridization images from our experiments to support our point that HassOr14b is co-localized with HassOr6 or HassOr16, and HassOR6 and HassOr16 are not co-localized.

We also carefully examined the related in situ hybridization images (Figure 8E) in Chang et al. (2016) (Scientific Reports 6, 18742), which indicated that HassOR6 and HassOr16 are co-localized. From their in situ hybridization images, it is difficult to judge if OR6 and Or16 are co-localized.

For the other questions we answer in the following one by one.

In this study we characterize the function of all the candidate PRs for Z9-16Ald, HassOr14b, HassOr6, HassOr16 in H. assulta, and find HassOr14b is the only receptor tuned to Z9-16Ald. We also prove the results of Jiang et al. (2014) that HassOr16 is the receptors in H. assulta for Z9-14:Ald, which is an antagonist for this species. For H. armigera, Z9-14:Ald acts as an agonist in small amounts (0.3%) but an antagonist in higher amounts (1% and above). We have discussed these previous results, which is helpful to understand the conclusions that we draw from our experiments.

Reviewer #1:

[…] 1) As far as I can tell only one peer reviewed paper indicates the Z9-14Ald is an active positive part of the Harm pheromone. In another paper this compound is found to be an antagonist in both Hass and Harm. This should be made clear to the reader because unlike some other systems where there is a clear change from using one compound to using another in two closely related species, in this case the selective pressure to evolve the receptor in Harm for Z9-14Ald is not very clear.

Yes, for H. armigera, Z9-14:Ald acts as an agonist in small amounts (0.3%) (Rothschild, 1978; Wu et al., 2015; Zhang et al., 2012) but an antagonist in higher amounts (1% and above) (Gothilf et al., 1978; Kehat and Dunkelblum, 1990; Wu et al., 2015); for H. assulta, Z9-14:Ald acts as an antagonist (Boo et al., 1995; Wu et al., 2015). We accept your suggestion and add this information in the revised manuscript (Introduction, fifth paragraph).

The selective pressure to evolve the receptor in H. armigera for Z9-14:Ald is intriguing. The SSR data prove that male antennae of H. armigera have olfactory sensory neurons (OSNs) responding to Z9-14:Ald (Wu et al. 2015; Xu et al., 2016). Two reasons may explain its evolution:

1) Z9-14:Ald acts as an agonist at the low dosage and an antagonist in the pheromone communication system of H. armigera, so it is necessary to have such a pheromone receptor. Up to now, HarmOr14b is the only receptor specifically responding to Z9-14:Ald in H. armigera.

2) The pheromone ratio of Z9-16:Ald to Z11-16:Ald in H. assulta is 93:7, and that in H. armigera is 3:97. Correspondingly, the population ratio of OSNs responding to Z9-16:Ald and Z11-16:Ald in H. assulta is 81:19, while that in H. armigera is 22:78. Assuming that H. assulta is more primitive species than H. armigera (Cho et al., 2008; Fang et al., 1997; Wang et al., 2005), shifting a given receptor’s selectivity could be the simplest and most efficient way to change the population ratio of pheromone OSNs in the evolution.

2) Earlier research papers that have searched for activity of HR14b in Hass have failed to find any activity. This includes a recent paper by some of the authors of the current paper. It would be useful for the authors to explain why they did not find activity in their other recent paper, but that they now find it.

It has long been a mystery about the receptor to Z9-16:Ald, the major sex pheromone component of H. assulta. In 2014, our group first cloned the full-length sequences of HassOr14b and HarmOr14b by the use of LA-Taq polymerase (Takara shuzo, Shiga, Japan), but failed to identify its function when using Xenopus system (Jiang et al., 2014). Along with identifying the function of more and more Ors by using Xenopus system, we found that the accuracy and integrity of the OR sequence is very important in functional analysis. Therefore, in this study we re-cloned the sequence of HassOr14b by use of Q5® High-Fidelity DNA Polymerase (New England Biolabs, USA) and repeated again to verify the sequence by Sanger sequencing for 10 samples, and also compared with the sequence in the antennal transcriptome data of H. assulta. Finally, we got the correct sequence, in which there are 3 amino acids different comparing with the old sequences (Figure 3—figure supplement 3). Moreover, we further analyzed the transcriptome data in H. assulta and confirmed that in the 3 different amino acids positions, there is no sequence polymorphism. We use the accurate sequence this time and finally get the function of HassOr14b, which is specifically tuned to Z9-16:Ald, the major sex pheromone component in H. assulta.

In the study of Chang et al. (2016), they also try to identify the function of HassOr14b but failed when using Xenopus system (Chang et al., 2016). There are two differences between the two studies. One is that they also used the LA-Taq polymerase we used before when they cloned the sequence, and there is 1 amino acid different in the 5’ ends with ours (Figure 3—figure supplement 3). By analyzing the transcriptome data in H. assulta, we confirmed that this amino acid position has no sequence polymorphism. Another difference is the vector used in the expression system. They used the pT7Ts vector, while we use the pCS2+ vector in the expression system.

In short, we find that the accuracy and integrity of the sequence is crucial to identify the function of the receptors. We use the right sequence so we get the activity of HassOr14b. Moreover, using the appropriate expression vector could be also important. We add this information in the second paragraph of the subsection “Novel identified PR and the different combinations with other PRs”.

3) Looking at the plots in the figures that show "current (nA)" on the Y-axis, it becomes clear that at least with the Xenopus system, the level of activity of the Hass HR14b receptor is much lower than for the Harm HR14b receptor. Whereas the response to Z9-14Ald in Harm is over 400 and for Z9-16Ald is around 50, the response of the Hass HR14b receptor to even Z9-16Ald (to which it is specific) is only about 60. Clearly, this oocyte system is not very efficient for Hass HR14b. This may be why previous studies didn't find any activity. The authors should acknowledge this.

The Xenopus laevis oocyte expression system with two electrode voltage clamping is an efficient system for identifying insect odorant (pheromone) receptors (Di et al., 2017; Lu et al., 2007; Mitsuno et al., 2008; Nakagawa and Touhara, 2013; Sakurai et al., 2004; Tanaka et al., 2009; Wanner et al., 2007; Xia et al., 2008). One major advantage of oocytes is that these cells only express a few ion channels and receptors, which are quite different from those of insects, so that the insect receptors can be studied without contamination from endogenous channels and receptors (Goldin, 2006). The oocyte system is particularly well suited for the study of many different mutations because injection and two-electrode voltage-clamping can be carried out rapidly and in a semi-automated fashion (Goldin, 2006; Hill et al., 2015; Hughes et al., 2014; Leal, 2013; Leary et al., 2012; Nakagawa et al., 2012).

We think this system is also suitable for HassOr14b based on the following reasons:

1) The responding current of the oocytes expressing HassOr14b to 10-4 M of Z9-16:Ald is about 70 nA, which is comparable to the current (80 nA) of the oocyte expressing HarmOr13 tuning to the major component, Z11-16:Ald in H. armigera (Liu et al., 2013; our unpublished data). We have done many technical and biological replications, and the results are repeatable.

2) The oocytes expressing HassOr14b to Z9-16:Ald are retested within a dose–response paradigm, revealing a clear concentration dependency.

3) Other evidences also support the HassOr14b is the receptor of Z9-16:Ald in H. assulta: The expression level of HassOr14b is highest among all the PRs in male antennae of H. assulta, and the OSNs responding to Z9-16:Ald is the most abundant in male antennae of H. assulta; the localization patterns of HassOr14b with other PRs in male antennae of H. assulta are consistent with the single sensilla recording results.

4) This lack of sensitivity of the HR14b of Hass is problematic for interpreting the results of single mutations to the Hass HR14b receptor. Clearly, the T3551 mutation dramatically increases the overall response of HR14b while negating specificity. This makes it hard to interpret the interaction of the two mutations.

The sensitivity of HassOr14b is well retested by the dose response paradigm of HassOr14b stimulated with a range of Z9-16:Ald concentrations (Figure 4), with the EC50 value for Z9-16:Ald was 8.65 × 10–5 M. T355I mutation increases the response of HassOr14b to the related compounds but not overall. We think this does not negate the specificity (not responding to all the tested ligands) but extend the tuning spectrum from Z9-16:Ald to both Z9-16:Ald and Z9-14:Ald.

In the mutagenesis study, the change of the tuning selectivity is what we focus on. We find that only T355I and F232I change the selectivity. When we made the combination of the two site mutations, the PR with two mutations shifted the ligand selectivity. The experimental work is systematic and the results are convincing.

5) There is overlap between the paper Olfactory perception and behavioral effects of sex pheromone gland components in Helicoverpa armigera and Helicoverpa assulta Meng Xu et al. 2016 and the current paper. This overlap should be made more clear.

The content of this study is closely related with Xu et al. (2016), the previous study in our group, but there is no data overlap between them. The only data similar in the two studies is the expression level of PRs. Xu et al. (2016) used the in situ hybridization to detect the numbers of the cells expressing different PRs. In this study, we use qPCR to analyze the expression level of all the PRs. The present results from the molecular level are consistent with the previous results from the cell level.

Reviewer #2:

[…] 1. I miss citation to the recent article by De Fouchier et al. in Nature Communications (DOI: 10.1038/ncomms1570910).

We accept the suggestion, and add this citation.

2. Since HarmOR14b is tuned to Z9-14:Ald, I think it's important that the authors write 1-2 sentences on how this component is a minor (but essential) sex pheromone component in H. armigera. In the current text it seems that the sex pheromone system of these two species is a two-component system with similar reverse ratios as in the two pheromone strains of Ostrinia nubilalis, while the pheromone blend of H. armigera (and also H. assulta) is a bit more complex than just a two-component blend.

We accept the suggestion, and add “(Z)-9-tetradecenal (Z9-14:Ald) acts as an antagonist in the pheromone communication of H. assulta (Boo et al., 1995; Wu et al., 2015). In that of H. armigera, Z9-14:Ald acts as an agonist in small amounts (0.3%) (Rothschild, 1978; Wu et al., 2015; Zhang et al., 2012) but an antagonist in higher amounts (1% and above) (Gothilf et al., 1978; Kehat and Dunkelblum, 1990; Wu et al., 2015).” in the revised manuscript’s Introduction.

3. The authors give dN/dS ratios (ω) for the different ORs (Results, subsection “Phylogenetic analysis of candidate PRs”), reasoning that ω < 1 indicates puryfying selection, while ω > 1 indicates positive selection. As they found a ω = 0.17 for cluster OR14b, this thus indicates puryfying selection. However, in the discussion the authors do not come back to this result and instead write "The female moth produces a pheromone blend of several components, stabilized by strong selection pressure against any change in such blends (Roelofs et al., 2002). This requires an equivalent stability from the male moths to detect the same species-specific pheromones, but at the same time should allow for a degree of plasticity to adapt to changes in pheromone structures associated with speciation." This part needs to be revised, as 'stabilized by strong selection pressure' and 'equivalent stability' are strangely used. Also, the latter part "should allow for a degree of plasticity" comes across as hand waving. Similarly, "In the course of speciation, the functional change of ORs is a gradual process with multiple amino acid mutations, a few making drastic changes and many making small modifications or even no change in function" need to be revised, as 'gradual changes' contradicts 'a few making drastic changes' and as a whole this sentence doesn't make sense.

Thanks for all the suggestions. We accept all of them and revise the manuscript as follows:

We add “The ω value for all clusters of PRs are less than 1, suggesting that all the PRs are subjected to purifying selection.” in the. In addition, we recalculate the dN/dS value of Or11 cluster, and find the correct value of Or11 cluster is less than 0.13. So the ω values of all clusters of PRs are less than 1. We have corrected the related sentences in the new version.

We revise the statement in the new version as “Under stabilizing selection, variation of the female pheromone blend is limited, and the males typically prefer the most common pheromone blends (Groot et al., 2016; Roelofs et al., 2002).” in the Discussion.

Also in the Discussion, we revise the statement in the new version as: “However, at the same time the male moths need to have a degree of plasticity to adapt to changes in signal structures associated with speciation. How mutations in PRs change their response to odorants remain unclear.”

We also revise this statement as “In the course of speciation, the functional change of ORs could be a process with multiple amino acid mutations, a few making drastic changes and many making small modifications or even no change in function.” in the Discussion.

4. Discussion, subsection “Novel identified PR and the different combinations with other PRs”, first paragraph: HassOR14b should be "HassOR16 in the same sensilla". Do the authors know where the orthologous ORs are localized in H. armigera? Is HarmOR14b also co-localized with HarmOR6 or HarmOR16? Can I deduce from the intro information that these sensilla are the type C sensilla?

To make this question clear, we add the new data about the expression pattern of all the PRs in H. armigera in Figure 2—figure supplement. It is clear that HassOR14b is co-localized with HassOR16 or HassOR6 in the C type sensilla in H. assulta. However, HarmOR14b had a very low expression level in the male antennae of H. armigera, even lower than HarmOr16 and HarmOr6. We speculate that the localization of PRs is not exactly same in the type C sensilla between H. armigera and H. assulta. It is still unclear about the localization of HarmOr14b, HarmOr6, and HarmOr16 in H. armigera. We will clarify this in future research.

5. Discussion, subsection “Novel identified PR and the different combinations with other PRs”, second paragraph:: Where are these amino acid residues located? I think it's important to specify this, especially because this is the first study (right?) where amino acid sequence changes in the intracellular domains (ICDs). I would also like to read a bit more on how the authors think these changes may alter the function. They give two very short explanations in the next paragraph, but what do they mean with 'complex deep structure'?

We used the TOPCONS (topcons.net) to predict the secondary structure of Or14b, and find that the two mutation sites were located in the intracellular domains (ICDs). This is the first study of ORs where amino acid sequence changes in the intracellular domains (ICDs) affect the tuning selectivity. We specify this in the revised Discussion. We also cite the new references for the amino acid changes may alter the function. We revised this statement as “the binding site of ligand-specific ORs, such as PRs, may have a complex structure, which involves TMDs (Leary et al., 2012), ECLs (Hughes et al., 2014) and ICDs.”

Reviewer #3:

Yang et al. report that two amino acid substitutions in intracellular domains may account for the difference in ligand specificity between HarmOr14b and HassOr14b. The conclusion is based on a series of mutagenesis experiments. The results are interesting but "the story" is not ready for publication. A number of major issues are listed below. In addition, I think that the discussion of how mutations in the intracellular domains may influence ligand specificity (subsection “Two amino acids located in the intracellular domains together determine the OR selectivity”, last paragraph) is speculative and lacks both references and a possible mechanism.

Many thanks for the comments. We revised the discussion of how mutations in the intracellular domains may influence ligand specificity. We added the references in the last paragraph of the subsection “Two amino acids located in the intracellular domains together determine the OR selectivity”.

The study is mainly based on the authors´ finding that HassOR14b is responsive to Z9-16:Ald whereas HarmOR14b is responsive to Z9-14:Ald. However, according to a previous study from the same laboratory (Jiang et al., 2014) and the study by Chang et al. (2015), HassOR14b did not respond to any tested compounds including Z9-16:Ald. This needs to be clarified. The authors do not mention about this discrepancy and do not even cite the Chang et al. paper:

Jiang, X.-J., Guo, H., Di, C., Yu, S., Zhu, L., Huang, L.-Q., Wang, C.-Z. (2014). Sequence similarity and functional comparisons of pheromone receptor orthologs in two closely related Helicoverpa species. Insect Biochemistry and Molecular Biology 48, 63-74.

Chang, H., Guo, M., Wang, B., Liu, Y., Dong, S., and Wang, G. (2016). Sensillar expression and responses of olfactory receptors reveal different peripheral coding in two Helicoverpa species using the same pheromone components. Scientific reports, 6, 18742.

The first question is the same with the reviewer #1’s question 2. Please find our response to reviewer #1.

We discussed this discrepancy and cite the Chang et al.’s paper in the second paragraph of the subsection “Novel identified PR and the different combinations with other PRs”.

In the phylogenetic tree, the authors only included Heliothine species in OR13 cluster, but not the orthologues from other species, thus the calculation of dN/dS value is biased. Moreover, the authors say the dN/dS value of OR11 cluster is larger than 1.8, this is inconsistent with the previous findings that dN/dS value of this cluster is low (around 0.1):

Zhang, D. D., and Löfstedt, C. (2013). Functional evolution of a multigene family: orthologous and paralogous pheromone receptor genes in the turnip moth, Agrotis segetum. PLoS One, 8(Fang et al., 1997), e77345.

Zhang, Y. N., Zhang, J., Yan, S. W., Chang, H. T., Liu, Y., Wang, G. R., and Dong, S. L. (2014). Functional characterization of sex pheromone receptors in the purple stem borer, Sesamia inferens (Walker). Insect molecular biology, 23(Carraher et al., 2015), 611-620.

We accepted the suggestions and have recalculated the dN/dS value in Or13 cluster including the orthologues from other species, and the ω value now is less than 0.17. We also rebuilt the Figure 1 in the revised version.

Thank you for this important comment. We recalculated the dN/dS value of Or11 cluster, and found the value we previously calculated about Or11 is not correct. In the sequence alignment analysis, we wrongly used “Align by ClustalW” (It should be “Align by ClustalW (Codons)”). We are very sorry for this mistake. Now, the value of Or11 cluster corrected is less than 0.13, which is consistent with Zhang and Löfstedt (2013) and Zhang et al. (2014). We have corrected this mistake and double checked all the calculated ω values in the revised manuscript in Figure 1 and subsection “Phylogenetic analysis of candidate PRs”.

The in situ hybridization images are not convincing and this is problematic as the results do not correspond to what was reported in Chang et al. (2015). These authors claimed that HassOR6 and HassOR16 are localised in the same sensillum (Chang et al., 2015).

We provide more in situ hybridization images of our experiments in Author response images 1, 2 and 3). For each treatment, we present three more sets of images besides the images we present in the manuscript (Figure 5). The results are clear and repeatable.

Author response image 1 – Part 1, HassOr14b and HassOr6 are co-localized in some sensilla (arrows); Part 2: Only HassOr14b is detected in other sensilla (arrows).

Author response image 2 – Part 1, HassOr14b and HassOr16 are co-localized in some sensilla (arrows); Part 2, only HassOr14b is detected in other sensilla (arrows).

Author response image 3 – HassOr6 and HassOr16 are always expressed in different sensilla. Part 1, HassOr16 is only expressed in some sensilla (arrows); Part 2, HassOr6 is only expressed in some sensilla (arrows).

All these indicate that HassOr14b was co-localized with HassOr6 or HassOr16 in the sensilla, and the results match with the functions of the PRs, the expression patterns of PRs, and the electrophysiology of related OSNs in H. assulta (Chang et al., 2016; Xu et al., 2016).

Author response image 1
Two-colour in situ hybridization visualizing the combinations of HassOr14b and HassOr6 in male antennae of H. assulta.

Arrows indicate the cell location. Scale bars: 20 μm.

Author response image 2
Two-colour in situ hybridization visualizing the combinations of HassOr14b and HassOr16 in male antennae of H. assulta.

Arrows indicate the cell location. Scale bars: 20 μm.

Author response image 3
Two-colour in situ hybridization visualizing the combinations of HassOr16 and HassOr6 in male antennae of H. assulta.

Arrows indicate the cell location. Scale bars: 20 μm.

In the paper of Chang et al.(2016), the only set of images about the localization of HassOr6 and HassOr16 in H. assulta in Figure 8E of Chang et al.is as the following (Author response image 4). From the images, it is very difficult to judge OR6 and Or16 are co-localized.

Author response image 4
Two-colour in situ hybridization visualizing the combinations of HassOr6 and HassOr16 in male antennae of H. assulta.

Arrows indicate the cell location. Scale bar: 20 μm; the boxed area: 10 μm (adapted from Figure 8E Chang et al., 2016).

Chang et al.(2016) suggested that HassOr6 is the receptor of Z9-16:Ald although they confirmed that the most effective ligand of HassOr6 is Z9-16:OH in Xenopus system (Supplementary Figure 1). From the Xenopus result (Z9-16:Ald is not the most effective ligand of HassOr6.) to the expression pattern of PRs (HassOr6 is not the highest expressed PR in male antennae of H. assulta.) and the electrophysiological data (the OSNs responding to Z9-16:Ald is the most abundant in male antennae.), all the evidences do not support that HassOr6 is the receptor of Z9-16:Ald.

Discussion subsection “Novel identified PR and the different combinations with other PRs”, first paragraph: 'This suggests that the ancestor of the two sister species not only changed OR14b's expressing level, but also altered its tuning selectivity to meet the species specific demands': The authors did not compare the expression levels of HassOR14b and HarmOR14b so the statement is not supported. In addition, it is not clear to me how the authors can conclude anything about the expression levels in the ancestor. Based on analysis of the contemporary species (which at some point had a common ancestor) we can at the best conclude that expression levels are different in these two species. The reasoning further involves a teleological argument, i.e. that the species that evolved from the common ancestor had some "specific demands". This is not how evolution works.

Thank you for these critical comments and sorry for our poor expression in English. We add the data about the expression pattern of all the PRs in H. armigera in Figure 2—figure supplement 2. It is clear that HarmOR14b had a very low expression level in the male antennae, even lower than HarmOr16 and HarmOr6. As HassOr14b is most highly expressed in H. assulta, and the functional study confirms that HassOR14b is specifically tuned to Z9-16:Ald, while its ortholog HarmOr14b is specifically tuned to Z9-14:Ald, we suggest that the Or14b orthologs of the two sister species not only changed Or14b’s expressing level, but also altered its tuning selectivity in evolution.

We accept the suggestion and revised the sentence as “This suggests that the two sister species not only changed Or14b’s expressing level, but also altered its tuning selectivity in speciation.”.

In the subsection “Quantitative real-time PCR”: the authors write that the reference gene is 18s rRNA, but the GenBank number provided is actually the actin gene. This is confusing.

We are sorry for this mistake and thank you for your carefulness. We used the 18s rRNA as the reference gene, and its GenBank number is changed into EU057177.1 accordingly. We have corrected it in the subsection “Quantitative real-time PCR”.

In the subsection “Construction of the mutation sequences”: The description of the construction strategy would benefit from a diagram visualizing the different steps.

It is a very nice suggestion, we have added the diagram visualizing the construction strategy in the revised manuscript (Figure 7—figure supplement 1).

[Editors' note: further revisions were requested prior to acceptance, as described below.]

The reviewers have appreciated your effective revision of the manuscript and the extensive explanation in the rebuttal. They conclude that your study provides very interesting results on how small differences in pheromone receptors may influence ligand selectivity. Still some important issues remain and I invite you to address these comments by the reviewers and prepare a second revision of the manuscript. Specific attention should be paid to sentences where you refer to the literature on odour reception and genetic mechanisms and evolutionary consequences. Some of these sentences are highlighted in the reviews, while others are not. For example, what is meant with 'a few steps' in the last sentence of the Abstract.

Thank you so much for your comments to the last revision and giving us a chance to revise the manuscript again. Although we cannot identify the highlights in the reviews, we have paid attention to the related sentences and rephrased them. The last sentence of the Abstract is revised as follows: “We conclude that species-specific changes in the tuning specificity of the PRs in the two Helicoverpa moth species could be achieved with just a few amino acid substitutions, which provides new insights into the evolution of closely related moth species.”.

Reviewer #1:

[…] The authors have made useful changes but I think a little more discussion of some of the issues with the Xenopus system in this specific case is warranted because the same issue is likely to arise in future studies.

Thank you so much for giving your recognition to our revision. We added some sentences in Discussion to discuss the issues with the Xenopus system: “It is worth noting that inward currents of the oocytes expressing HassOr14b induced by Z9-16:Ald were distinct but relatively low. […] A clear dose response curve to the most effective ligand is always helpful to confirm the receptor’s function.”.

I realize that in the current Abstract and in most of the manuscript the authors avoid discussing the relevance of their findings to the evolution of the two species of Helicoverpa. For example, in the Abstract they state that "We conclude that species-specific changes in the tuning specificity of the PRs of male moths could be achieved with just a few steps". As long as the authors don't indicate that their findings help us to specifically understand the evolution of the two Helicoverpa moths, I think they are on solid ground. The two mutations they study certainly change the tuning specificity.

Thank you so much for your suggestions. We added the statement in the revised Abstract as “We conclude that species-specific changes in the tuning specificity of the PRs in the two Helicoverpa moth species could be achieved with just a few amino acid substitutions, which provides new insights into the evolution of closely related moth species.”. We added the statement in the revised Discussion as “The peripheral modifications of the two closely related species took place in both PR expression level and PR tuning selectivity. These findings not only help us specifically understand the evolution of the two Helicoverpa species, but also provide new insights into the structure and function of cell membrane receptors.”.

It still would be good if the authors could elaborate on the fact their findings are of most interest in terms of neurophysiology and are not a direct commentary on the evolution of these specific moths.

Thank you for your suggestions. This is a good point. We added one paragraph in Discussion as follows: “Animal nervous systems are shaped by shifting environmental selection pressures to perceive and respond to new sensory cues (Prieto-Godino et al., 2017). […] How mutations in olfactory receptors change the olfactory responses of animals and eventually impact on the evolution of animal behavior is crucial but remains unclear.”

Reviewer #2:

.[…] 1) In Figure 3 the authors show that HasOR14b responds to Z9-16:Ald, but when comparing the different graphs, the y-axis in Figure 3 is in a different (smaller) scale than the y-axis in Figure 3—figure supplement 1, which shows that HassOr6 actually responds more to Z9-16:Ald (300 nA) than HassOr14b (70 nA). Such a difference in scaling is also present in Figure 8; HassOr14b shows a current response to Z9-16:Ald of 70 nA again, and/but HarmOr14b shows a current response of 30-40 (?) nA. I do understand that HarmOr14b mostly responds to Z9-14:Ald, but the response to Z9-16:Ald may also be important, no? (In this respect it is indeed very impressive that the authors found an almost 10x stronger response when mutating T335I (see Figure 7—figure supplement 2). Which relates to my next question:

The Xenopus system is a heterologous expression system. Its main disadvantage is that the oocytes are not the native cells in which the PRs are normally expressed. Although the functional properties that are observed may not be identical to those characterized in native olfactory sensory neurons, the most effective ligand obtained by using the Xenopus system is convincing. However, the response range is often wider than the natural situation (Wang et al., 2016). It is hard to judge whether the compounds inducing lower currents could actually activate the related neurons in the sensilla. For example, the neurons expressing HarmOr13 and HassOr13 in Type A sensilla are only tuning to Z11-16:Ald specifically (Wu et al., 2013; Chang et al., 2016; Xu et al., 2016). When HarmOr13 and HassOr13 are expressed in Xenopus oocytes, the most effective ligand is still Z11-16:Ald, but they also show a relatively lower response to Z9-14:Ald (Jiang et al., 2014; Liu et al., 2013). So we would better pay more attention to the response selectivity rather than the absolute current value among different ORs. In the Results part, we revise some statements in the paragraphs of “Regional mutations of HassOr14b and functional analysis” and “Site-specific mutations of HassOr14b and functional analysis” (first paragraph).

Thank you for your comment on the results. We think “Figure 7—figure supplement 2” you mentioned is “Figure 8” because the result of T355I is in Figure 8-F.

2) As Z9-16:Ald is also an important (albeit minor) sex pheromone component of H. armigera, with what receptor(s) do the authors think H. armigera is perceiving this component?

This is a very good question. The receptor(s) tuning to Z9-16:Ald of H. armigera is still a mystery.

In the PR clade in phylogenetic analysis, 7 PRs have been identified, they are HarmOr6, HarmOr11, HarmOr13, HarmOr14, HarmOr14b, HarmOr15, and HarmOr16. Up to now, 4 of them are functional characterized: HarmOr6, HarmOr13, HarmOr14b, HarmOr16. The most effective ligands are Z9-16:OH for HarmOr6 (Jiang et al., 2014), Z11-16:Ald for HarmOr13 (Liu et al., 2013), Z9-14:Ald for HarmOr14b (Jiang et al., 2014, and this study), Z11-16:OH (Chang et al., 2017; Liu et al., 2013) and Z9-14:Ald (Liu et al., 2013) for HarmOr16. The function of other 3 PRs (HarmOr11, HarmOr14, HarmOr15) are not yet identified. HarmOr11 is co-localized with HarmOr13 in Type A sensilla (Chang et al., 2016), it has no possibility tuning to Z9-16:Ald; HarmOr15 has a much lower expression in male H. armigera (this study, Figure 2—figure supplement 2), so could not be the candidate receptor. Therefore, it could be HarmOr14 as the potential receptor tuning to Z9-16:Ald, although all the studies up to now failed to identify its function (Jiang et al., 2014; Liu et al., 2013). In the recent ESITO meeting, a new PR clade was reported in Spodoptera, it would be also interesting to explore the new PRs in Helicoverpa armigera in the further study.

3) In Figure 2 the authors show that HassOR6 is highly expressed in males, and a little in females. Which makes sense to me, as females probably smell their own pheromone as well. I don't think that sex-specific expression is a necessary criterium for an OR to be a sex pheromone OR.

We totally agree with this. A recent study on Heliothis virescens (Zielonka et al., 2016) find that HR6 is expressed in both female and male antennae, which indicates that the females can detect pheromone components released by themselves or by conspecifics. We also agree that the sex-specific expression is not a necessary criterium for an OR to be a PR, functional analysis is the only criterium to judge whether it is a PR. However, the male biased expression pattern could narrow the scope of the candidate PRs in moths, which is the efficient way to identify most of the PRs.

4) The authors have responded to previous comments on why previous functional assays with HassOr14b didn't work by writing that the sequence had not been correct in previous work and that this is now corrected. In Figure 3—figure supplement 3 they specify which 3 amino acids have been corrected (Yang et al. is the current sequence used). In checking these 3 amino acids in Figure 6, I found them in RIII, in RVII and in RVIII. I find it hard to determine whether one of these 3 SNPs were among the ones tested in Figure 7—figure supplement 2. Still, I think the authors make a convincing case that the right sequence together with the selection of the appropriate expression vector is crucial to identify the function of receptors.

In the site mutation analysis, we just focused on the different amino acids between orthologs of H. assulta to H. armigera. Among the three revised amino acids of HassOr14b, two of them located in RIII and RVIII respectively are the same with those of HarmOr14b. The third amino acid located in RVII of HassOr14b is different from that of HarmOr14b, when we replaced the whole region VII of HassOr14b with that of HarmOr14b, we did not find the selectivity change. Therefore, we did not test these three amino acids specifically. We think “Figure 7—figure supplement 2” you mentioned should be “Figure 8”. Thank you for your comments to our results.

5) Hopefully Figure 7—figure supplement 2 is not a supplementary figure, as I think this figure is a great way of showing the results of the single mutations.

We think that “Figure 7—figure supplement 2” you mentioned is “Figure 8”, which is the result of the two-electrode voltage-clamp recordings of site mutations with pheromone components and analogs. This result is not a supplementary figure. Thank you for your comments to our results.

6) As for the in situ hybridization pictures, these are clear to me, but I leave this to the more experienced reviewer(s) to decide whether this is sufficiently clear.

Thank you for your comments to our results.

Reviewer #3:

[…] In the rebuttal letter the authors write: 'Basically there are two major concerns: (Baker et al., 2004) the suitability of the Xenopus system.… (Boo et al., 1995) Contradiction of our in situ hybridization results'

In my opinion this is not correct. At least not when it comes to my concerns. My major concern was that the authors avoided mentioning the inconsistency in the response of wild type HassOR14b compared to previous publications. In the current version, the authors explain that the gene sequences were not correct in previous publications. However, the explanation is not straight forward:

The full-length sequences of HassOr14b and HarmOr14b were first cloned by Jiang et al. in 2014, and Jiang et al. identified HarmOr14b was tuned to Z9-14:Ald, but failed to identify the function of HassOr14b (Jiang et al., 2014). In 2016, Chang et al. proved the function of HarmOr14b, and tried to identify the function of HassOr14b again but also failed. Because Jiang et al. (2014) is the first study of HassOr14b and there is no new progress of this gene in functional analysis in the following study, we just cited Jiang et al.’s work in the very first manuscript version. From the last revision, we accepted the reviewer’s suggestion and cited both of them.

- The sequences in Chang et al., 2016 were based on the transcriptome data from the previous paper (Zhang et al., 2015) from the same group. There's a single amino acid difference of HassOR14b sequence between the current manuscript and the sequence in Chang's paper. Could it simply be explained by the use of different polymerases? How could the authors exclude the possibility of polymorphism? Although possible, it's not very likely that this one amino acid difference would totally abolish the response to Z9-16:Ald (No response to Z9-16:Ald or any other ligand in Chang et al. 2016) but significant response to Z9-16:Ald in the current study (70 nA – Figure 8). The authors of the present manuscript can of course not take the responsibility of any possible flaws in a study published by another group but the discrepancy requires an explicit comment in the Discussion.

Zhang, J., Wang, B., Dong, S., Cao, D., Dong, J., Walker, W. B.,.… & Wang, G. (2015). Antennal transcriptome analysis and comparison of chemosensory gene families in two closely related noctuidae moths, Helicoverpa armigera and H. assulta. PloS one, 10(Boo et al., 1995), e0117054.

Many factors are possible to result in the sequence cloning error. We think DNA polymerase is the key factor in PCR systems. Among the DNA polymerases supplied by many companies, the fidelity of them is quite different. By our experiences, in the purpose of cloning the sequences for functional analysis, the high fidelity enzyme should be the first choice. The DNA polymerase used in Chang et al.’s study is LA-Taq (Chang et al., 2016), which is not a high fidelity enzyme.

Concerning the possibility of polymorphism, we excluded it using the following strategy: The transcriptome data represents many separate individuals, and the results of sequencing include abundant separate “reads”. For each gene, there are many “reads” mapping to it, but each “reads” is not exactly the same in all the positions, which may come from the polymorphism in such position. If all of the “reads” in one position is the same, we can exclude the possibility of polymorphism in that position. We analyzed all the related transcriptome data of H. assulta carefully, and if we found that all of the “reads” in one position is the same, we exclude the possibility of polymorphism in that position.

Concerning the amino acid difference from that in Chang et al. (2016), this amino acid is located in RI of HassOr14b, which is the same with that of HarmOr14b, so we do not test this amino acid specifically because in the site mutation experiments we just focused on the different amino acids between orthologs of H. assulta to H. armigera. Although we do not test whether the change of this amino acid would totally abolish the response of HassOr14b, we have experience that one amino acid change in HarmOr13 could totally abolish its function (our unpublished data). Therefore, we suggest that the accuracy and integrity of the sequence is crucial to identify the function of ORs. We have explained this in the second paragraph of the Discussion subsection “Novel identified PR and the different combinations with other PRs”.

I do not understand how different expression vectors could cause the difference in sequence. The authors have to explain how. Otherwise it is just an ad hoc explanation that needs to be tested. It's the cRNAs that are injected and expressed in the oocytes. The resulting receptor proteins should have the same sequences despite of what the original expression vectors were. In addition, the pT7TS vectors have been widely and successfully used in oocyte recordings and OR studies. Why should this vector not be suitable for HassOR14b expression and oocyte recordings?

It is true that the vector would not cause the difference in sequence, but the vector itself used in the Xenopus system could affect the identification of the ORs. pT7TS vectors have been successfully used in oocyte recordings in many OR studies, but Chang et al. used it and failed to identify HassOr14b’s function. We use the pCS2+ vector and successfully identify HassOr14b. Comparing with the Chang et al.’s study, we notice that there are two major differences: one is on the sequence of HassOr14b, another is on the expression vector, so we speculate that these two differences could be the possible reasons.

The authors write in the rebuttal and mention in the Discussion that they re-sequenced HassOr14b and came up with the "right sequence". This is a new (and important) result and as such it should be mentioned in the Results section and the relevant methods and materials should be described in Materials and methods before the discrepancy with previous studies is discussed in the Discussion section.

Thank you and we accept the suggestion, and added this information in the Results subsection “PR specifically tuned to Z9-16:Ald in H. assulta”, and the Materials and methods subsection “Cloning of the candidate pheromone receptor of H. assulta and H. armigera”.

As far as I understand reviewer #1 was not really questioning oocytes as a suitable system for OR de-orphanization in general, but said that the system may not be efficient for testing HassOR14b. In my opinion, however, a current of 60 nA is not too small to be noticed. Hence this is not the reason why Chang et al. did not find a response, but there seems to be a real difference in activity between the two studies.

It remains mysterious to me why the wild type HassOR14b responds so differently in this and previous studies and it's problematic that the authors did not mention this inconsistency in the first version. In addition, a number of other mistakes/errors were pointed out by me and the other reviewers. The authors have now corrected these errors but can we be sure that additional errors did not go unnoticed by the reviewers? Careful doublechecking appears necessary.

Thank you for explaining the results in the Xenopus oocyte system. Concerning the difference in activity, the fact is not that the previous studies found one function and we find another function of HassOr14b, but is that the previous studies all failed to identify its function and we find it in this study. We suggest that it would be the use of the right sequence and the selection of the expression vector that result in our success in this study. As both Jiang et al. and Chang et al. failed to identify the function of HassOr14b, and Jiang et al. is the first study identifying HassOr14b (Jiang et al., 2014, Chang et al., 2016), we just cited Jiang et al.’s work in the very first manuscript version. In the revised manuscript, we accept your suggestion and cite both of them. Thanks for your suggestion and we had double-checked the manuscript carefully.

Some other points for your consideration.

1) 'Unlike general odorant receptors (ORs) that typically bind more than one ligand (de Fouchier et al., 2017), PRs are narrowly tuned to specific pheromone components (Grosse-Wilde et al., 2007).' This statement is not correct. There are both narrowly tuned PRs and broadly tuned PRs reported in the literature (Miura et al., 2010; Wanner et al. 2010; Zhang and Löfstedt 2015). Even in Grosse-Wilde et al., 2007 cited by the authors, it was proposed that 'there are two different designs of pheromone receptors'. The authors have to be more precise in their statements and more rigorous when citing the literature:

Miura, N., Nakagawa, T., Touhara, K., and Ishikawa, Y. (2010). Broadly and narrowly tuned odorant receptors are involved in female sex pheromone reception in Ostrinia moths. Insect biochemistry and molecular biology, 40(Baker et al., 2004), 64-73.

Wanner, K. W., Nichols, A. S., Allen, J. E., Bunger, P. L., Garczynski, S. F., Linn Jr, C. E., […] and Luetje, C. W. (2010). Sex pheromone receptor specificity in the European corn borer moth, Ostrinia nubilalis. PLoS One, 5(Baker et al., 2004), e8685.

Zhang, D. D., and Löfstedt, C. (2015). Moth pheromone receptors: gene sequences, function, and evolution. Frontiers in Ecology and Evolution, 3, 105.

We accept the suggestion, and revised the statement as “PRs are in general narrowly tuned to specific pheromone components (Grosse-Wilde et al., 2007; Miura et al., 2010; Zhang and Löfstedt, 2015).”

2) 'For example, BmOr1 of.[…] to 1,3Z,6Z,9Z-19:H (Zhang et al., 2016).' As we suggested before, there's no point of listing these examples so this sentence can be removed.

We accept the suggestion and removed this sentence in the revised manuscript.

3) Results and Discussion sections: 'all the PRs and Orcos are subjected to purifying selection which is consistent with the previous studies (Zhang and Löfstedt, 2013; Zhang et al., 2014b)'. This is not a correct reference. Zhang and Löfstedt (2013) as well as other researchers (e.g. Leary et al., 2012) found that some PR clusters are under purifying selection while some are under relatively relaxed selective pressure. Since the authors used a limited number of moth species in the phylogenetic tree in the current study, it's too daring to draw the conclusion that 'all the PRs and Orcos are subjected to purifying selection'. Leary, G. P., Allen, J. E., Bunger, P. L., Luginbill, J. B., Linn, C. E., Macallister, I. E.,.[…] and Wanner, K. W. (2012). Single mutation to a sex pheromone receptor provides adaptive specificity between closely related moth species. Proceedings of the National Academy of Sciences, 109(Liu et al., 2014), 14081-14086.

We accept the suggestion and revised them as “this indicates that all the PRs and Orcos analyzed in this study are subjected to purifying selection, which is consistent with the previous studies (Zhang and Löfstedt, 2013; Zhang et al., 2014b).”; and “The ω value for all clusters of PRs analyzed in this study are less than 1, suggesting that PRs would be subjected to purifying selection.”

4) In situ results: 'in some sensilla only HassOr14b was detected'. For the Discussion: Is this consistent with previous SSR data? Which gene is supposed to be expressed in the neighboring neuron?

We found that HassOr14b and HassOr6 were co-localized in some sensilla, while in other sensilla only HassOr14b was detected. A similar situation was observed for HassOr14b and HassOr16. They were co-localized in some sensilla, but only HassOr14b was detected in other sensilla. HassOr6 and HassOr16 were always expressed in different sensilla. These results indicate that HassOr6 or HassOr16 were co-localized with HassOr14b in the neighboring neurons in the Type C sensilla, which is consistent with the previous SSR data that there are subtypes in the type C sensilla. We revised the statement in the Discussion as “HassOr14b is co-localized with HassOR6 or HassOR16 in the neighboring neurons in the same sensilla,”, and added the statement as “Our results indicate that there are different combinations of the PRs in the C type sensilla, which is consistent with the previous single sensillum recording results that there are subtypes in the type C sensilla (Xu et al., 2016).”

5) The authors refer to the transcriptome a couple of times in the manuscript. Have the authors deposited the transcriptome data in a public database?

The transcriptome data of H. assulta and H. armigera submitted by MPICE is deposited in the NCBI-Sequence Read Archive (SRA) (https://www.ncbi.nlm.nih.gov/sra). The transcriptome data of H. assulta and H. armigera acquired by our group is still under analysis (our unpublished work), and we will deposit it in a public database when we finish the analysis.

6) Discussion subsection “Implications for the modulation and evolution of OR selectivity”: 'The olfactory systems found in all animals have nearly the same design feathers, which give olfaction a considerably flexibility for signaling to evolve.' I don't understand this sentence. What does 'design feathers' mean? Should it be "features"?

We are sorry for this mistake and thank you for your carefulness. It should be “features” and we have corrected it in the first paragraph of the subsection “Implications for the modulation and evolution of OR selectivity”.

7) The English still needs some attention throughout the manuscript before its scientific merits can be definitely evaluated.

We accept the suggestion, and have revised the language throughout the manuscript.

https://doi.org/10.7554/eLife.29100.023

Article and author information

Author details

  1. Ke Yang

    1. State Key Laboratory of Integrated Management of Pest Insects and Rodents, Institute of Zoology, Chinese Academy of Sciences, Beijing, China
    2. College of Life Sciences, University of Chinese Academy of Sciences, Beijing, China
    Contribution
    Data curation, Software, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing—original draft, Writing—review and editing, Experiment design
    Competing interests
    No competing interests declared
    ORCID icon 0000-0002-4138-3373
  2. Ling-Qiao Huang

    State Key Laboratory of Integrated Management of Pest Insects and Rodents, Institute of Zoology, Chinese Academy of Sciences, Beijing, China
    Contribution
    Resources, Funding acquisition, Validation, Writing—original draft, Writing—review and editing
    Competing interests
    No competing interests declared
  3. Chao Ning

    1. State Key Laboratory of Integrated Management of Pest Insects and Rodents, Institute of Zoology, Chinese Academy of Sciences, Beijing, China
    2. College of Life Sciences, University of Chinese Academy of Sciences, Beijing, China
    Contribution
    Formal analysis, Validation
    Competing interests
    No competing interests declared
  4. Chen-Zhu Wang

    1. State Key Laboratory of Integrated Management of Pest Insects and Rodents, Institute of Zoology, Chinese Academy of Sciences, Beijing, China
    2. College of Life Sciences, University of Chinese Academy of Sciences, Beijing, China
    Contribution
    Conceptualization, Resources, Data curation, Software, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing—original draft, Project administration, Writing—review and editing, Experiment design
    For correspondence
    czwang@ioz.ac.cn
    Competing interests
    No competing interests declared
    ORCID icon 0000-0003-0418-8621

Funding

Strategic Priority Research Program of the Chinese Academy of Sciences (Grant number XDB11010300)

  • Chen-Zhu Wang

National Natural Science Foundation of China (Grant number 31130050)

  • Chen-Zhu Wang

National Basic Research Program of China (Grant number 2013CB127600)

  • Chen-Zhu Wang

National Key R&D Program of China (Grantnumber 2017YFD0200400)

  • Chen-Zhu Wang

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

We thank our colleagues Hao Guo, Lin Yang, Ya-Lan Sun, Rui Tang for their kind assistances in in situ hybridization, confocal microscopy, the construction of mutation sequences, and data analysis, respectively. We thank Dr. Ya-Nan Zhang from Huaibei Normal University and Dr. Da-Song Chen from Guangdong Institute of Applied Biological Resources for their kind assistances in phylogenetic analysis. We thank Prof. Zhan-Fen Qin from Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences for providing Xenopus laevis frogs. We thank Prof. Paolo Pelosi from University of Pisa, Italy and Prof. Bill Hansson from Max Planck Institute for Chemical Ecology, Germany for valuable comments. This work is supported by the Strategic Priority Research Program of the Chinese Academy of Sciences (grant number XDB11010300), the National Natural Science Foundation of China (grant number 31130050), the National Key R and D Program of China (grant number 2017YFD0200400), and the National Basic Research Program of China (grant number 2013CB127600).

Ethics

Animal experimentation: All procedures in this study were approved by the Animal Care and Use Committee of the Institute of Zoology, Chinese Academy of Sciences for the care and use of laboratory animals (protocol number IOZ17090-A). The surgery was performed following the protocols reported by Nakagawa and Touhara (2013). The Xenopus laevis was anesthetized by bathed in the mixture of ice and water in 30 min, the wounds were carefully treated to avoid infection. Every effort was made to minimize suffering.

Reviewing Editor

  1. Marcel Dicke, Reviewing Editor, Wageningen University, Netherlands

Publication history

  1. Received: May 30, 2017
  2. Accepted: October 22, 2017
  3. Accepted Manuscript published: October 24, 2017 (version 1)
  4. Version of Record published: November 6, 2017 (version 2)

Copyright

© 2017, Yang et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 726
    Page views
  • 199
    Downloads
  • 0
    Citations

Article citation count generated by polling the highest count across the following sources: PubMed Central, Scopus, Crossref.

Comments

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)