Evolutionarily entrenched odorant receptors for (E)-β-farnesene signaling in Aphidinae aphids

TL;DR

Aphidinae aphids use specific odorant receptors to detect an alarm pheromone, and these receptors are evolutionarily conserved.

Contribution

The study identifies three conserved odorant receptors that selectively detect the alarm pheromone (E)-β-farnesene in Aphidinae aphids.

Findings

Three odorant receptors (OR5, OR40, OR43) are selective detectors for the alarm pheromone (E)-β-farnesene in Aphidinae aphids.

Knockdown of these receptors suppressed EBF-induced repellency in Acyrthosiphon pisum.

OR gene loss occurred only in lineages that no longer use EBF as an alarm signal.

Abstract

Alarm pheromones are crucial for the survival of social insects, enabling coordinated escape from predators. Their evolution has co-evolved with the development of olfactory recognition systems, suggesting specific molecular mechanisms underlie this adaptive relationship. We analyzed alarm pheromone compositions across 36 aphid species and found that EBF serves as the sole or primary alarm signal in the Aphidinae subfamily. Genomic annotation of 13 aphid species identified eight conserved, Aphidinae-specific odorant receptor (OR) clades under strong purifying selection. Three receptors—OR5, OR40, and OR43—were EBF-selective in Aphidinae species. Their individual or collective knockdown suppressed EBF-induced repellency in Acyrthosiphon pisum, indicating non-redundant roles in receptor combinational coding. Phylogenetic analyses demonstrated variation in gene age among these receptors, with losses confined to species that do not use EBF as an alarm signal. This study demonstrates a multi-receptor system for EBF detection in Aphidinae aphids and advances the understanding of olfactory system evolution.

•Identified three conserved ORs as selective detectors for the alarm pheromone EBF•Silencing OR5, OR40, or OR43, singly or together, blocked EBF-induced repellency in aphids•Phylogenetics reveals that OR gene loss occurred exclusively in lineages losing EBF usage

Identified three conserved ORs as selective detectors for the alarm pheromone EBF

Silencing OR5, OR40, or OR43, singly or together, blocked EBF-induced repellency in aphids

Phylogenetics reveals that OR gene loss occurred exclusively in lineages losing EBF usage

Evolutionary Biology; Molecular Biology and Genetics; Neurobiology and Sensory Systems; Chemical Ecology; Entomology and Insect Physiology

Figures

Annotation and evolutionary dynamics of the OR gene family in 13 aphid species(A) Quality screening of genome assemblies represented as the length of contig N50 versus the percentage of complete single-copy BUSCO genes. Abbreviations are defined at the end of the caption.(B) The phylogenetic tree at the left shows inferred species relatedness; different colors indicate aphid subfamilies. The corresponding bar plot at right shows the number of OR genes annotated from each genome assembly: pink, yellow, and gray bars represent intact genes, pseudogenes, and partial genes, respectively. Binomial

Eight conserved OR clades in Aphidinae species(A) Gene tree constructed from protein sequences of 533 intact aphid ORs based on the maximum likelihood algorithm. Branches indicating ORs of Aphidinae and non-Aphidinae species are colored in blue and red, respectively. The eight single-copy ortholog OR clades are indicated with arrowheads in different colors.(B) Phylogenetic tree and pairwise amino acid sequence similarities of ORs from eight conserved OR clades. Color codes in the heatmap indicate pairwise amino acid sequence similarities, which range from dark blue (65.9% identity) to yellow (

Functional characterization of OR40 and OR43 from Aphidinae species(A) The representative inward current response of ApisOR40 and ApisOR43 to (*E*)-β-farnesene (EBF) and geranyl acetate (GA).(B) Inward current values of ApisOR40 and ApisOR43 to EBF and GA. Student’s *t* test was used to compare responses. Data are plotted as mean ± SEM (*n* = 7 for ApisOR40; *n* = 6 for ApisOR43).(C) Dose-response trace of ApisOR40 and ApisOR43 to EBF.(D) Dose-response curves of ApisOR40 and ApisOR43 to EBF. EC_50_ value of ApisOR40 and ApisOR43 is 1.58 × 10^−5^ mol/L (*n* = 8) and 2.91 × 10^−6^ mol/L (*n* = 6

Function of ApisOR5, ApisOR40, and ApisOR43 in the mediation of EBF repellency in *Acyrthosiphon pisum*(A–C) Relative expression levels of *ApisOR5, ApisOR40*, and *ApisOR43* transcripts after respective dsRNA infiltration in *A. pisum* (mean ± SEM, *n* = 3, GLM followed by Duncan’s multiple range test).(D) EAG responses of dsRNA-infiltrated *A. pisum* to EBF (*n* = 15–22, GLM followed by Duncan’s multiple range test). Bars labeled with different letters are significantly different. Plotted data are mean ± SEM.(E) Diagram of the experimental setup of the T-tube. Hexane was used as a control.(F

Evolution of the EBF receptors(A) Evolutionary age inferred for each aphid OR. The scaled estimated age of each OR gene is categorized as 1, 2, 3, 4, or 5, with 1 indicating the oldest evolutionary age and 5 indicating the youngest.(B) Evolutionary pattern of EBF receptors in the genomes of 12 species with or without EBF in their alarm pheromones. The occurrence of a single-copy ortholog of OR5, OR40, and OR43 is indicated with a solid red circle; multi-copy orthologs are indicated by a solid teal circle, and the absence of each ortholog is indicated with a hollow gray circle. Species with ide

Full text

Introduction

Predation poses a significant threat to insect survival, driving the evolution of alarm pheromones that alert conspecifics to danger.1^,^2 Aphids, sap-feeding insects in the order Hemiptera, are exemplary models for studying alarm pheromones and their ecological roles.3^,^4^,^5 When threatened, aphids release waxy secretions from their cornicles—specialized structures on the fifth or sixth abdominal segment—containing alarm pheromones. These secretions elicit behavioral responses such as walking away or dropping from the host plant, enhancing colony survival.6

The sesquiterpene (E)-β-farnesene (EBF) was the first alarm pheromone identified in aphids7 and remains the only or the primary component in many species, including economically significant pests.8^,^9 However, alternative compounds, such as germacrene A in Therioaphis,10 and monoterpenes α-pinene and β-pinene in Megoura viciae,11 also act as alarm pheromones in certain species. Additionally, some aphids release blends of EBF with other monoterpenes,12 highlighting diversity in alarm pheromone composition and potential differences in their reception mechanisms.

Aphid olfactory perception relies heavily on odorant receptors (ORs), expressed on the dendritic membranes of olfactory sensory neurons, which translate chemical signals into neural responses that drive behavioral outcomes.13^,^14 Despite this, the molecular basis of aphid alarm pheromone recognition remains poorly understood. Although ApisOR5 and SmisOR5 have been implicated in EBF detection in A. pisum and Sitobion miscanthi,15 the presence of additional receptors and their evolutionary relationships in EBF-dependent and independent species remain unresolved.

Advances in genomic resources provide critical tools for studying gene family evolution.16 While the number of aphid genomes largely increased,17^,^18 comprehensive OR gene annotation remains limited.19 Here, we summarized the alarm pheromone profiles of multiple aphid species, identifying EBF as the dominant component in the Aphidinae subfamily. We annotated and analyzed OR genes from diverse aphid subfamilies to identify those specific to Aphidinae that mediate EBF detection. Using two-electrode voltage-clamp assays, RNAi, and behavioral experiments, we dissected the molecular mechanisms underlying EBF detection. Evolutionary analyses further revealed the co-evolution of EBF and its receptors, with receptor losses occurring exclusively in species that do not utilize EBF. Our findings highlight a specialized multi-receptor system for EBF detection and provide important insights into the evolution of alarm pheromone recognition in aphids.

Results

(E)-β-farnesene is an important alarm pheromone compound in Aphidinae species

We summarized the alarm pheromone profiles of 36 aphid species based on published data (Table S1). These pheromones predominantly consist of terpenoids, either as single compounds or complex mixtures. The profiles indicated that EBF was identified in all the tested Aphidinae species (Table S1). Among the 31 species for which composition data were available (Table 1), EBF was identified as the sole alarm pheromone in 21 of the 28 Aphidinae species, as the dominant component (>70%) in three others, and was present at low levels in the remaining four species. Thus, EBF is the primary or exclusive alarm pheromone in all Aphidinae species examined. In contrast, EBF was not observed in the alarm pheromone compositions of the tested species from non-Aphidinae subfamilies, such as Calaphidinae and Drepanosiphinae. However, a minor EBF presence was detected in Chaitophorinae species (Table 1). These findings suggest that the olfactory reception of EBF might be a conserved trait within the Aphidinae subfamily, distinguishing it from other aphid lineages.Table 1. Proportion of alarm pheromone components across 31 aphid speciesAphid speciesEBF (%)α-Pinene (%)β-Pinene (%)Limonene (%)Germacrene (%)Other (%)SubfamilyAphis craccivoraa10000000AphidinaeAphis fabaeb10000000AphidinaeAphis glycinesc10000000AphidinaeAphis gossypiid10000000AphidinaeAcyrthosiphon pisumb10000000AphidinaeAphis sambucib10000000AphidinaeAphis spiraecolae10000000AphidinaeAphis urticatab10000000AphidinaeBrachycaudus carduib10000000AphidinaeBrachycaudus persicaee10000000AphidinaeBrachycaudus schwartzib10000000AphidinaeHyalopterus prunib10000000AphidinaeHyperomyzus lactucaeb10000000AphidinaeMacrosiphoniella abrotanib10000000AphidinaeMetopolophium dirhodumb10000000AphidinaeMyzus cerasib10000000AphidinaeMyzus persicaea10000000AphidinaeRhopalosiphum padif10000000AphidinaeRhopalosiphum maidise10000000AphidinaeSitobion avenaeb10000000AphidinaeSchizaphis graminumg10000000AphidinaeAulacorthum solanib91.800005.2AphidinaeMyzus lythrib83.3000012.8AphidinaeCapitophorus elaeagnib73.1000.7026.2AphidinaeMegoura viciaeh32.649.4249.745.2403.14AphidinaeDysaphis plantagineab15.9000084.1AphidinaeBrevicoryne brassicaeb5026.2086.8AphidinaeAphis idaeib1.3627.70071AphidinaeChaitophorus populetib553.316.60.7024.4ChaitophorinaeDrepanosiphum platanoidisb012.841.332.8013.1DrepanosiphinaeEuceraphis punctipennisb00004258CalaphidinaeaBayendi Loudit et al.9bFrancis et al.12cEichele et al.20dByers21eVerheggen et al.22fFan et al.23gBowers et al.7hSong et al.24

Annotation of OR genes in eight Aphidinae and five non-Aphidinae species

To characterize the OR gene repertoires across aphid species, we analyzed 15 published genomes (10 Aphidinae and 5 non-Aphidinae), sourced from NCBI and AphidBase.25 Genome completeness was assessed using contig N50 values and BUSCO scores. Genomes failing to meet the criteria of >90% complete BUSCO genes and a contig N50 > 1,000 Kb were excluded, leading to the rejection of two Aphidinae species (Diuraphis noxia and Myzus cerasi) (Figure 1A and Table S2). The final dataset comprised 13 genome assemblies representing six subfamilies: Lachninae, Erioasomatinae, Hormaphidinae, Chaitophorinae, Calaphidinae, and Aphidinae (Figure 1B).Figure 1. Annotation and evolutionary dynamics of the OR gene family in 13 aphid species(A) Quality screening of genome assemblies represented as the length of contig N50 versus the percentage of complete single-copy BUSCO genes. Abbreviations are defined at the end of the caption.(B) The phylogenetic tree at the left shows inferred species relatedness; different colors indicate aphid subfamilies. The corresponding bar plot at right shows the number of OR genes annotated from each genome assembly: pink, yellow, and gray bars represent intact genes, pseudogenes, and partial genes, respectively. Binomial names at the end of the caption.(C) Correlation between several tandemly organized OR genes and the total number of OR genes.(D) Correlation between the number of OR pseudogenes and the total number of OR genes.(E) A cladogram of 13 aphid species shows the inferred number of OR gene gain and loss events along branches and the number of intact OR genes in the genomes. Aphis gossypii (Agos), Aphis glycines (Agly), Acyrthosiphon pisum (Apis), Cinara cedri (Cced), Diuraphis noxia (Dnox), Eriosoma lanigerum (Elan), Hormaphis cornu (Hcor), Myzus cerasi (Mcer), Myzus persicae (Mper), Pentalonia nigronervosa (Pnig), Rhopalosiphum padi (Rpad), Rhopalosiphum maidis (Rmai), Sipha flava (Sfla), Sitobion miscanthi (Smis), and Therioaphis trifolii (Ttri).

From these genomes, we manually annotated 634 OR genes, including 83 pseudogenes and 19 partial genes. Among these, 533 were classified as intact and potentially functional, as their protein-coding sequences exceeded 350 amino acids, a size typical of insect OR genes. The average protein length ranged from 410 to 419 amino acids (Table S3), consistent with reported insect ORs (Robertson, 2019). To confirm the integrity of these genes, we predicted the number of transmembrane domains (TMDs) for each OR protein. Using TOPCONS and DeepTMHMM, we identified an average of 6.25 ± 0.02 (mean ± SE) and 6.03 ± 0.01 TMDs, respectively, comparable to the predicted TMD count in Drosophila melanogaster (Table S4).

Table S4. The number of predicted OR transmembrane (TM) domains in aphids, related to STAR Methods

OR gene numbers varied significantly across species, with the highest count (86 genes) in Acyrthosiphon pisum and the lowest (23 genes) in Sipha flava (Figure 1B and Table S3). Even within the Aphidinae subfamily, variability was evident: for instance, Pentalonia nigronervosa had only 33 OR genes, far fewer than A. pisum.

Tandemly clustered OR genes were prevalent in many genomes, constituting 71.43% of OR genes in Therioaphis trifolii, 69.39% in Rhopalosiphum maidis, and 67.12% in Eriosoma lanigerum (Table S3). A significant positive correlation was observed between the number of tandemly clustered OR genes and the total OR gene count (R^2^ = 0.8612, p = 2.955 × 10^−6^) (Figure 1C). Additionally, pseudogene numbers correlated positively with total OR gene counts (R^2^ = 0.5023, p = 0.004) (Figure 1D), indicating that aphid OR gene evolution follows a birth-and-death model. To investigate the evolutionary history of aphid OR genes, we analyzed gene gain and loss events across taxonomic lineages. Substantial OR gene gain events were identified in Aphidinae (Figure 1E), highlighting significant olfactory adaptations following the divergence of Aphidinae from other subfamilies.

Eight conserved Aphidinae-specific single-copy odorant receptor clades identified by phylogenetic and sequence analyses

Considering that EBF is commonly released and detected by Aphidinae aphids, we hypothesized that its recognition mechanism might be conserved. A phylogenetic tree of 533 intact ORs from 13 aphid species was constructed (Figure S1), defining Aphidinae-conserved clades as those showing no gene loss or copy number variation among the eight Aphidinae species analyzed. Eight such clades were identified and named based on their orthologs in A. pisum: OR2, OR3, OR4, OR5, OR20, OR39, OR40, and OR43 (Figure 2A).Figure 2. Eight conserved OR clades in Aphidinae species(A) Gene tree constructed from protein sequences of 533 intact aphid ORs based on the maximum likelihood algorithm. Branches indicating ORs of Aphidinae and non-Aphidinae species are colored in blue and red, respectively. The eight single-copy ortholog OR clades are indicated with arrowheads in different colors.(B) Phylogenetic tree and pairwise amino acid sequence similarities of ORs from eight conserved OR clades. Color codes in the heatmap indicate pairwise amino acid sequence similarities, which range from dark blue (65.9% identity) to yellow (100% identity).(C) Percentage of pairwise amino acid sequence similarities in each OR clade. The violin plots show the distribution of pairwise sequence similarities of each clade. The black bar within the box found in each violin plot indicates the median values of pairwise sequence identities in each clade. The red line shows the mean value of pairwise sequence similarities in each clade.(D) The nonsynonymous (dN) to synonymous (dS) substitution rate (ω) of each OR clade.(E) Genomic position and synteny of orthologous ORs of the eight conserved OR clades in five chromosome-level genome assemblies.

Pairwise amino acid sequence alignments revealed high similarities among orthologs within each clade, particularly between closely related species (Figures 2B and S2). For instance, OR5 and OR43 orthologs of A. glycines and A. gossypii (Agly vs. Agos) showed 100% and 99.5% sequence similarities, respectively, and those of R. padi and R. maidis (Rpad vs. Rmai) exhibited 97.0% and 95.5% sequence similarity, respectively. On average, pairwise sequence similarity exceeded 80% across all clades (Figure 2C). The OR2 clade exhibited the highest average similarity (89.46%), ranging from 84.75% to 99.27%. In contrast, the OR4 clade showed greater sequence diversity, with similarities ranging from 65.89% to 98.38%, averaging 80.01% (Figure 2C). Selective pressure analysis revealed that all eight clades underwent purifying selection, with ω ratios (dN/dS) consistently below 1 (Figure 2D).

Genomic location analyses across five chromosome-level assemblies indicated that these ORs were located exclusively on autosomes. No tandem clusters were observed between any of the eight Aphidinae-conserved ORs. While orthologs OR3 and OR20 were consistently adjacent, their separation (>100,000 bp) excluded them from tandem duplication classification (Figure 2E). This suggests these conserved ORs originated independently rather than being derived from gene duplication events of any of these ORs.

Odorant receptor 5, odorant receptor 40, and odorant receptor 43 function as (E)-β-farnesene receptors mediating repellent behavior

To test their assumed roles of the eight Aphidinae-conserved ORs in EBF perception, we expressed A. pisum OR genes in Xenopus oocytes and analyzed them using a two-electrode voltage clamp technique. Oocytes co-expressing ApisOR2/Orco, ApisOR3/Orco, ApisOR4/Orco, ApisOR20/Orco, or ApisOR39/Orco did not respond to EBF (Figure S3). Surprisingly, alongside the previously characterized ApisOR5,15 both ApisOR40/Orco and ApisOR43/Orco exhibited responses to EBF and geranyl acetate (GA), a structural analog of EBF (Figure 3A). Among these, ApisOR40 showed the strongest response to GA, while ApisOR43 displayed the highest sensitivity to EBF (Figure 3B). Dose-response analyses revealed EC_50_ values for ApisOR40 and ApisOR43 on EBF of 1.58 × 10^−5^ mol/L and 2.91 × 10^−6^ mol/L, respectively, underscoring the larger responses of ApisOR43/Orco to EBF (Figures 3C and 3D).Figure 3. Functional characterization of OR40 and OR43 from Aphidinae species(A) The representative inward current response of ApisOR40 and ApisOR43 to (E)-β-farnesene (EBF) and geranyl acetate (GA).(B) Inward current values of ApisOR40 and ApisOR43 to EBF and GA. Student’s t test was used to compare responses. Data are plotted as mean ± SEM (n = 7 for ApisOR40; n = 6 for ApisOR43).(C) Dose-response trace of ApisOR40 and ApisOR43 to EBF.(D) Dose-response curves of ApisOR40 and ApisOR43 to EBF. EC_50_ value of ApisOR40 and ApisOR43 is 1.58 × 10^−5^ mol/L (n = 8) and 2.91 × 10^−6^ mol/L (n = 6), respectively. Data are reported as mean ± SEM.(E) Phylogenetic tree of OR40 and OR43 homologs for A. pisum, M. persicae, and S. miscanthi. SmisOR28 and MperOR45, hereafter indicated as SmisOR43homo and MperOR43homo. MperOR39 and SmisOR26, hereafter indicated as MperOR40homo and SmisOR40homo.(F) Representative inward current response of MperOR40homo and SmisOR40homo to EBF and GA.(G) Inward current values of ApisOR40, MperOR40homo, and SmisOR40homo to GA and EBF. A significant difference was observed between the responses to GA and EBF in ApisOR40 (∗p < 0.05), MperOR40homo (∗∗p < 0.01), and SmisOR40homo (∗p < 0.05). Data are plotted as mean ± SEM (n = 7 for ApisOR40; n = 7 for MperOR40homo; n = 7 for SmisOR40homo).(H) Dose-response curves of ApisOR40, MperOR40homo, and SmisOR40homo to EBF, with EC_50_ value of 1.46 × 10^−5^ mol/L (n = 6), 6.31 × 10^−6^ mol/L (n = 5) and 5.29 × 10^−6^ mol/L (n = 6), respectively.(I) Representative inward current response of MperOR43homo and SmisOR43homo to EBF and GA.(J) Inward current values of ApisOR43, MperOR43homo, and SmisOR43homo to GA and EBF. A significant difference was observed between the responses to GA and EBF in ApisOR43 (∗∗∗p < 0.001), MperOR43homo (∗∗∗p < 0.001), and SmisOR43homo (∗p < 0.05). Data are plotted as mean ± SEM (n = 5 for ApisOR43; n = 5 for MperOR43homo; n = 5 for SmisOR43homo).(K) Dose-response curves of ApisOR43, MperOR43homo, and SmisOR43homo to EBF, with EC_50_ value of 2.91 × 10^−6^ mol/L (n = 6), 2.92 × 10^−6^ mol/L (n = 3) and 7.12 × 10^−6^ mol/L (n = 4), respectively.

To assess the conservation of EBF receptor functionality across Aphidinae species, homologs of ApisOR40 (OR40) and ApisOR43 (OR43) from M. persicae (MperOR39 and MperOR45) and S. miscanthi (SmisOR26 and SmisOR28) were evaluated (Figure 3E). Both OR40 homologs responded to GA and EBF, with stronger responses to GA (Figures 3F–3H). Conversely, OR43 homologs showed significantly stronger responses to EBF than GA while maintaining comparable sensitivity to EBF (Figures 3I–3K). These results confirm the functional conservation of OR40 and OR43 across Aphidinae species, consistent with previous findings for OR5.15 Notably, OR43 demonstrated the highest affinity for EBF, solidifying its role as the most potent EBF receptor.

RNA interference (RNAi) experiments further demonstrated the roles of ApisOR5, ApisOR40, and ApisOR43 in EBF-mediated repellency. Silencing each OR individually reduced the electrophysiological responses of A. pisum antennae to EBF at concentrations of 1 μg, 10 μg, and 100 μg compared to wild-type (WT) and dsGFP controls (Figures 4A–4D). At the low doses of 0.01 μg and 0.1 μg, individual or combined silencing of these three ORs did not cause a significant reduction in the EAG response. This is likely because the low doses themselves only elicit weak responses, making any silencing effect difficult to detect above this low baseline (Figure 4D). Behavioral assays using a T-tube olfactometer revealed that dsRNA-treated aphids suppressed EBF-repellent behavior (Figures 4E and 4F). Combined knockdown of ApisOR5, ApisOR40, and ApisOR43 resulted in a further reduction in antennal responses to 100 μg EBF compared to single knockdowns (Figure 4D) and entirely abolished EBF avoidance behavior (Figure 4F). These findings highlight potential synergistic interactions among these ORs and demonstrate their important roles in mediating EBF repellency in A. pisum.Figure 4. Function of ApisOR5, ApisOR40, and ApisOR43 in the mediation of EBF repellency in Acyrthosiphon pisum(A–C) Relative expression levels of ApisOR5, ApisOR40, and ApisOR43 transcripts after respective dsRNA infiltration in A. pisum (mean ± SEM, n = 3, GLM followed by Duncan’s multiple range test).(D) EAG responses of dsRNA-infiltrated A. pisum to EBF (n = 15–22, GLM followed by Duncan’s multiple range test). Bars labeled with different letters are significantly different. Plotted data are mean ± SEM.(E) Diagram of the experimental setup of the T-tube. Hexane was used as a control.(F) Behavioral choices of wild-type and dsRNA-infiltrated A. pisum to EBF (10 μg/μL) in a T tube olfactometer. EBF is significantly repellent to wild-type and dsGFP-infiltrated aphids (∗∗p < 0.01; chi-square test*,* χ^2^ = 18.375–21.496, df = 1, n = 96–121), while no significant differences of behavioral choice are found in dsApisOR5, dsApisOR40, dsApisOR43, and dsApisOR5&40&43 infiltrated aphids to EBF versus hexane (n.s., p > 0.05, no significant difference; chi-square test*,* χ^2^ = 0.269–2.600, df = 1, n = 65–93).

Evolution of (E)-β-farnesene receptors in aphids

To test whether the three EBF receptors were co-evolved, we analyzed the gene ages of each OR gene within the aphid OR repertoire. Gene ages were categorized from 1 (oldest) to 5 (youngest). The mean ages for the OR5, OR40, and OR43 clades were 2.00, 3.00, and 2.75, respectively (Figure 5A and Table S5), suggesting that OR40 and OR43 evolved more recently than OR5.Figure 5. Evolution of the EBF receptors(A) Evolutionary age inferred for each aphid OR. The scaled estimated age of each OR gene is categorized as 1, 2, 3, 4, or 5, with 1 indicating the oldest evolutionary age and 5 indicating the youngest.(B) Evolutionary pattern of EBF receptors in the genomes of 12 species with or without EBF in their alarm pheromones. The occurrence of a single-copy ortholog of OR5, OR40, and OR43 is indicated with a solid red circle; multi-copy orthologs are indicated by a solid teal circle, and the absence of each ortholog is indicated with a hollow gray circle. Species with identified alarm pheromone compositions were indicated with red triangle (with EBF), a black triangle (without EBF), or an empty gray triangle (unknown composition).

Table S5. Estimated evolutionary age of each OR, related to Figure 5A

Among non-Aphidinae aphids, alarm pheromone composition varies. Studies have shown that EBF is absent from the alarm pheromones of D. platanoidis, T. trifolii,10^,^12 E. lanigerum, and C. cedri (this study; Figure S4). To test whether the three EBF receptors were conserved among species with different alarm pheromone, the existence of EBF receptor genes was compared among seven Aphidinae species and five non-Aphidinae species with known alarm pheromone profiles (Figure 5B).

The results reveal distinct evolutionary trajectories between EBF receptors in Aphidinae and non-Aphidinae species. In Aphidinae, OR5, OR40, and OR43 are single-copy genes, while in non-Aphidinae, these genes exhibit multiple loss events and occasional duplications. For example, OR43 is absent in all non-Aphidinae species, OR5 is absent except in D. platanoidis, and OR40 is missing in T. trifolii but has two copies in D. platanoidis.

Discussion

Aphid ORs are pivotal in sensitively detecting alarm pheromones, such as EBF. This study provides a comprehensive annotation of OR genes across a diverse range of aphid taxa, including basal lineages (e.g., Hormaphidinae) and, more recently diverged lineages (e.g., Aphidinae). The resulting OR repertoire is valuable for elucidating the evolutionary mechanisms that have shaped the diversification of aphid odorant receptors in response to key ecological cues such as alarm pheromone detection. Although this study focuses on EBF detection, the annotated OR gene set could also inform studies on other key biological processes in aphids.

Our findings highlight significant diversity in the size of the OR gene family across aphid species, reflecting rapid turnover in gene gain and loss. Nevertheless, eight ORs are highly conserved within the Aphidinae subfamily, suggesting critical roles in the aphid life cycle. Among these, several ORs have known functions: for example, ApisOR20 detects cis-jasmone,26 a plant-derived compound with repellent effects on aphids.27^,^28^,^29 Additionally, ApisOR4 responds to multiple host plant volatiles.30 Despite these insights, the roles of OR2, OR3, and OR39, in addition to the well-characterized EBF receptors OR5, OR40, and OR43, remain unclear, warranting future research.

The analysis of alarm pheromones across 31 aphid species underscores EBF’s prevalence as the dominant or sole component in the alarm pheromones of Aphidinae species. Notably, EBF is absent in species from basal lineages such as Lachninae, Drepanosiphinae, and Calaphidinae. While OR5, a previously identified EBF receptor, exhibits higher responsiveness to GA (a structural analog of EBF) than to EBF itself,15 this study identifies OR40 and OR43 as additional EBF receptors. Unlike OR40, which shows a similar GA preference as OR5, OR43 demonstrates high specificity for EBF, likely due to distinct ligand-binding properties. This specificity may reflect adaptive selection for heightened sensitivity to EBF within the Aphidinae subfamily.

Variability in OR-ligand interactions may arise from subtle differences in their molecular binding mechanisms. Recent structural studies resolved the ApisOR5-Orco heterocomplex from the pea aphid A. pisum, both unbound and bound to its activating ligand GA.31 These findings suggest that ApisOR40 may share similar binding modes to GA and EBF as ApisOR5, reflecting comparable molecular recognition mechanisms. By contrast, ApisOR43 likely possesses a distinct ligand-binding pocket, enabling specific recognition of EBF. Detailed studies on the docking mechanisms of ApisOR40 and ApisOR43 with their respective ligands, supported by molecular dynamics simulations, will provide deeper insights into the olfactory mechanisms aphids use to detect and respond to alarm pheromones across different OR subfamilies.

The EBF-mediated repellency mechanism in aphids aligns with a “smell-and-avoid” model, wherein EBF is detected by ORs linked to neural circuits that drive avoidance responses.32 Notably, each EBF receptor plays an essential role in repellency, suggesting a multi-receptor system underpins this behavior. A comparable multi-receptor mechanism has been observed in Bactrocera dorsalis, where geosmin—a fungal volatile repellent—activates OR7, OR21, and OR25 in males and non-gravid females. Interestingly, geosmin serves a dual role as an attractant for ovipositing females.33 Such multi-receptor systems demonstrate the advantages of synergistic receptor activity in odor discrimination and behavior modulation.34^,^35^,^36 However, the neural pathways mediating EBF-triggered avoidance behavior in aphids remain to be elucidated.

Functional data demonstrate that the aphid perception of the alarm pheromone EBF is mediated by a combinatorial coding mechanism involving OR5, OR40, and OR43 (Figure 4). The foundation of this mechanism lies in the distinct response thresholds and kinetic profiles exhibited by these three receptors toward EBF—a pattern analogous to the strategy in Drosophila, where receptors of varying sensitivities extend the dynamic range of detection.37 We propose that such a functionally non-redundant receptor combination enables more nuanced decoding of EBF concentration, facilitating accurate initiation of adaptive behaviors such as escape in fluctuating environments. This finding reinforces the evolutionary principle in insect olfaction that combinatorial coding predominates over single-receptor dominance, as exemplified by the involvement of multiple ORs in benzaldehyde perception in adult Drosophila.38 Even for a single ecologically critical compound such as EBF, recognition still relies on cooperative input from multiple receptors. Together with studies on ethyl acetate evaluation in Drosophila larvae and benzaldehyde response in adults, our results illustrate that integrating signals through multiple receptors to enhance olfactory resolution represents a conserved feature of insect olfactory coding.39 Importantly, this coding strategy may be closely linked to ecological adaptation; the refined decoding capacity for EBF in aphids likely underlies their survival strategy against predation threats.

The evolutionary trajectories of EBF receptors reveal marked differences between aphid species that release EBF and those that do not. For example, the EBF receptor genes are entirely absent in T. trifolii, likely due to its adoption of germacrene A as a primary alarm pheromone.10 The use of germacrene A as an alarm pheromone may have likely relaxed the selective pressure to maintain EBF receptors, ultimately resulting in their loss. This divergence highlights parallel evolutionary processes in the recognition mechanisms for alarm pheromones among aphid species.

The origin of alarm pheromones remains a pivotal question in chemical ecology.40 Our findings support the “sensory exploitation hypothesis,” which posits that a chemical compound becomes a communication signal because the receptor predates the ability of conspecifics to produce the compound.41 Evolutionary analysis indicates that EBF receptors predate the emergence of the Aphidinae subfamily. For instance, OR40 orthologs are present in more ancient species such as E. lanigerum and C. cedri. This suggests that early aphids relied on external sources of EBF, such as plant emissions, to detect threats rather than producing the compound themselves.

Plants also emit EBF when attacked by aphids,42 emphasizing the ecological significance of plant-derived EBF in shaping aphid-plant interactions. The ability to detect plant-released EBF likely conferred an early survival advantage, even before aphid-produced EBF evolved as an alarm pheromone. In aphid species that do not produce EBF, OR40 may have diversified its function to detect other volatiles, yet the function tests remain to be performed. Further studies on EBF receptor homologs in these species could illuminate how aphids modify their sensory systems to adapt to environmental changes.

Limitations of the study

The study found that individual or collective knockdown of EBF-selective ORs suppressed EBF-induced repellency in A. pisum. However, the expression patterns of these ORs on odorant receptor neurons and the activation mechanisms of EBF signals in the central nervous system remain unclear. Additionally, since the alarm pheromone components of most aphid species are still unknown, the evolutionary relationship between alarm pheromone receptors and pheromone components requires further in-depth investigation.

Resource availability

Lead contact

Requests for further information and resources should be directed to and will be fulfilled by the lead contact, Guirong Wang (wangguirong@caas.cn).

Materials availability

This study did not generate new unique reagents.

Data and code availability

• •Data: The data supporting the findings of this study are available within the article and the supplemental information.
• •Code: This study does not generate original code.
• •Other items: Any additional information required is available upon reasonable request to the lead contact.

Acknowledgments

This work was supported by the 10.13039/501100001809National Natural Science Foundation of China (32472553), Major special projects for Green Pest Control (110202201017(LS-01)), Postdoctoral Fellowship Program (Grade C) of China Postdoctoral Science Foundation (GZC20233066) and Fellowship from the China Postdoctoral Science Foundation (2023M743837), the Shenzhen Science and Technology Program (KQTD20180411143628272), and was supported by the 10.13039/501100012421Agricultural Science and Technology Innovation Program (CAAS-BRC-CB-2025-02).

Author contributions

Conceptualization, T.H., B.W., and G.W.; methodology, T.H.; investigation, T.H., L.Y., Y.T., BoW., W.L., and P.B.; writing – original draft, T.H.; writing – review and editing, T.H., L.Y., B.W., and G.W.; funding acquisition, T.H. B.W., and G.W.; resources, Y.L., B.W., and G.W.; supervision, F.F., X.C., B.W., and G.W.

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCESOURCEIDENTIFIERChemicalsGeranyl acetateSigmaCAS: 105-87-3trans-β-FarneseneSigmaCAS: 18794-84-8Deposited dataA protein database of coding sequences from 46 insect speciesFigsharehttps://doi.org/10.6084/m9.figshare.28021820Software and algorithmsTOPCONShttps://topcons.cbr.su.se/version 2.0DeepTMHMMhttps://dtu.biolib.com/DeepTMHMMversion 1.0.13MAFFThttps://mafft.cbrc.jp/alignment/server/index.htmlversion 7.305trimAlhttps://www.megasoftware.net/version 1.2IQ-TREEhttps://iqtree.github.io/version 2.1.3OrthoFinderhttps://github.com/davidemms/OrthoFinderversion 2.5.2r8shttps://sourceforge.net/projects/r8s/version 1.81EVOLVIEWhttps://www.evolgenius.info/evolview/version 3.0NOTUNGhttps://www.cs.cmu.edu/∼durand/notung/version 3.0HyPhyhttps://github.com/veg/hyphyversion 2.5.25

Experimental model and study participant details

Animal procedures

The protocol for the animal experiments was approved by the Experimental Animal Welfare and Ethical of Institute of Plant Protection, Chinese Academy of Agricultural Sciences (IPP2025-09). The experiments were conducted strictly with governmental and international guidelines on animal experimentation. Mature female Xenopus laevis were purchased from the Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences. According to Biosafety and Animal Ethics requirements, all efforts were made to minimize the number of animals used and reduce their suffering.

Method details

Genome-wide OR gene annotation

Acyrthosiphon pisum OR sequences were used as queries due to the finely corrected gene models supported by RNAseq data.19 OR genes were annotated from genome assemblies of 11 aphid species: Aphis gossypii,43 R. padi,44 R. maidis,45 S. miscanthi,46 M. persicae,47 P. nigronervosa,48 S. flava (NCBI database: GCF_003268045.1), T. trifolii,49 C. cedri,50 E. lanigerum,51 H. cornu.52 OR gene models for A. pisum and Aphis glycines were re-annotated using updated genome assembly versions.47^,^53

Amino acid sequences of ApisORs were searched against aphid genomes using TBLASTN with an e-value cutoff of 1e-5. Matched high-scoring regions were extracted to infer potential OR coding regions. Coding regions were chained based on criteria,54 considering closely organized regions as a single entity if their upstream query was N-terminal to the downstream query. Exon-intron boundaries were manually validated using homologous CDS sequence evidence. Annotated ORs were added to the query set for iterative TBLASTN searches until no additional OR gene models were detected. ORs were named sequentially based on genome order, except for A. pisum and Aphis glycines, where names followed existing conventions. OR genes with frameshift mutations or premature stop codons were designated pseudogenes, while incomplete gene models lacking such mutations were considered partial genes.

Transmembrane domains were predicted using TOPCONS (v.2.0)55 and DeepTMHMM (v.1.0.13).56 Genes and pseudogenes within 10 kb of each other were identified as tandem arrays.

Phylogenetic and evolutionary analyses

Amino acid sequences were aligned using MAFFT (v.7.305)57 with poorly matched regions trimmed by trimAl (v.1.2).58 Phylogenetic trees were constructed using IQ-TREE (v.2.1.3),59 with branch support estimated using UFBoot260 with 5,000 replicates.

For species phylogeny, the longest predicted protein sequences of aphid genomes and Apolygus lucorum61 (outgroup) were analyzed using OrthoFinder (v.2.5.2).62 Divergence times were estimated using r8s (v.1.81)63 with calibration from TimeTree: A. pisum vs. M. persicae divergence at 42.5–48.0 mya. Phylogenies and gene orthology results were visualized using EVOLVIEW (v.3.0).64 Gene gain/loss events were inferred by reconciling the OR gene tree with the species tree using NOTUNG (v.3.0 BETA).65

Nonsynonymous (dN) to synonymous (dS) substitution rates (ω) were calculated for conserved branches using FitMG94.bf in HyPhy (v.2.5.25).66 To estimate the gene age of aphid ORs, we used the Phylostratigraphy pipeline (https://github.com/AlexGa/Phylostratigraphy) with a protein database of coding sequences from 46 insect species (https://doi.org/10.6084/m9.figshare.28021820). This approach employs a statistical framework to trace the evolutionary origins of protein-coding genes by grouping them phylogenetically, thereby uncovering footprints of significant adaptive events. BLASTP employed an e-value cutoff of 1e-8, and age predictions were simplified for clarity, namely 16, 17, 19, 20, and 28 were respectively replaced by 1, 2, 3, 4, and 5.

Homologs of OR5, OR40, and OR43 were annotated in the Drepanosiphum platanoidis genome67 using the aforementioned pipeline. Homologs clustering phylogenetically with Aphidinae ortholog lineages (UFBoot2 support >95%) were classified as orthologs.

Vector construction, cRNA synthesis, and two-electrode voltage-clamp recordings

Full-length coding sequences of ApisOR2, 3, 4, 5, 20, 39, 40, 43, and Orco were cloned from A. pisum antennal cDNA.68 Coding sequences were amplified by PCR and subcloned into pT7TS expression vectors.69 Gene cloning primers and vector details are listed in Table S6. cRNAs were synthesized using the mMESSAGE mMACHINE T7 Ultra Kit (Ambion, Austin, Texas, USA) following the manufacturer’s protocol.

Mature Xenopus oocytes were injected with 27.6 ng of OR cRNA and 27.6 ng of Orco cRNA using a Nanoliter 2010 injector (WPI Inc., Sarasota, FL) and incubated at 18 °C for 4–5 days. Odorant response curves were tested using a two-electrode voltage clamp (OC-725C, Warner Instruments, Hamden, CT). Data were acquired and analyzed using a Digidata 1440A and pCLAMP 10.2 software (Axon Instruments Inc., Union City, CA).

Stock odorant solutions (1 M) were prepared in DMSO and diluted in Ringer’s buffer (96 mM NaCl, 2 mM KCl, 5 mM MgCl_2_, 0.8 mM CaCl_2_, 5 mM HEPES, adjusted to pH 7.6 with NaOH) for recordings at a final concentration of 10^−4^ M. Dose-response curves were generated using a concentration gradient: 1 × 10^−8^, 1 × 10^−7^, 3 × 10^−7^, 1 × 10^−6^, 3 × 10^−6^, 1 × 10^−5^, 3 × 10^−5^, 1 × 10^−4^, and 3 × 10^−4^ M, with Ringer’s buffer containing 0.1% DMSO as the negative control.

Synthesis of dsApisORs and RNAi using nanocarrier/dsApisOR complex

The primers for dsRNA synthesis and quantitative real-time PCR (qRT-PCR) are listed in Table S6. Amplified sequences were purified via gel extraction. Subsequently, 1 μg of purified sequences was inserted into a pEASY-Blunt Cloning Vector (TransGen, China) and transfected into Trans1-T1 Phage-Resistant Chemically Competent Cells (TransGen, China). The extracted plasmid served as a template for synthesizing dsApisOR5, dsApisOR40, and dsApisOR43 using T7 RNA polymerase (Thermo, USA). Negative control dsRNA (dsGFP) was synthesized using the same method.

To prepare the RNAi complex, 1 μL of dsApisOR (1 μg/μL) was mixed with 1 μL of the nanocarrier at room temperature, following the procedure described in previously published work.70 The nanocarrier consists of a cationic dendrimer featuring a fluorescent perylene diimide core and four peripheral arms functionalized with amino acids. A nanocarrier/dsApisOR complex was obtained by combining detergent with the nanocarrier/dsApisOR solution in a 1:10 ratio. The nanocarrier/dsGFP complex was prepared similarly for use as a control. A 100 nL droplet of the nanocarrier/dsApisOR complex was administered onto the abdomen of adult aphids via microinjection. Aphids treated with either nanocarrier/dsApisOR or nanocarrier/dsGFP for 48 h were subjected to gene expression analysis and behavioral assays.

RT-qPCR

Total RNA from dsApisOR-treated aphids was extracted using TRIzol reagent (Invitrogen, USA). cDNA synthesis was performed using RevertAid First Strand cDNA Synthesis Kits (Thermo, USA). qPCR was conducted using GoTaq qPCR Master Mix (Promega, USA) on an Applied Biosystems 7500 Fast Real-Time PCR System (ABI, USA). The relative expression of ApisOR genes was normalized against ApisNADH and ApisSDHB genes using the 2^−ΔΔCT^ method.71^,^72 Primer sequences for RT-qPCR are provided in Table S6.

EAG recording

The base of the antennae was excised, and the tips were trimmed before insertion into glass electrodes containing 0.1 M KCl solution. For testing, a 10 μL test solution (containing odorant or solvent) was applied to filter paper, which was then inserted into a Pasteur pipette. For testing, odorant stock solutions (1 M) were diluted with paraffin oil. A stimulus controller (CS-55, Syntech, Kirchzarten, Germany) delivered humidified airflow at 30 mL/s. Stimuli were applied as 0.2 s pulses at 10 mL/s airflow with a 30 s interval between applications. Signals were amplified using a 10× AC/DC headstage preamplifier (Syntech) and recorded with an IDAC-4-USB Intelligent Data Acquisition Controller (Syntech). Data were analyzed using Syntech EAG software. EAG responses to stimuli were recorded for RNAi-treated aphids exposed to EBF.

Behavioral experiments

Behavioral assays were performed using a glass T-tube olfactometer (1.5 cm diameter, 1 cm trunk length, 20 cm branch length). Filter paper strips loaded with EBF (10 μg/μL) and hexane were placed randomly at the ends of the two branch arms. Aphids were released onto the wire, and their movements were observed. The aphid entered one arm of the T-tube 5 cm in length and remained there for 30 s, indicating a choice. A total of 96–121 replications (one aphid per replication) were conducted per group.

Chemical information collection

The components of alarm pheromones in Eriosoma lanigerum and Cavariella cedri were analyzed. Adults (10 C. cedri and 20 E. lanigerum) were submerged in 45 mL hexane containing 5 mL of a 0.75 ng/mL heptyl acetate solution (standard) in brown sample bottles. The mixture was swirled for 30 s and left at room temperature for 30 min before being stored at −20 °C. Samples were analyzed using a GC-MS system (QP2020, Shimadzu) with an Rtx-5MS column (30 m × 0.25 mm × 0.25 μm, Shimadzu). The GC oven temperature was programmed from 50 °C (held for 3 min) to 190 °C at 15 °C/min, then to 240 °C at 15 °C/min. Mass spectra were acquired at 70 eV, with mass scans between 35 and 500 amu at 0.2 scans/s. Chemical structures were confirmed by comparing known standards under identical conditions.

Quantification and statistical analysis

The distribution of aphids between olfactometer arms was analyzed using χ^2^ goodness-of-fit tests. For comparisons across multiple groups, one-way ANOVA followed by Duncan’s multiple-range test was applied. Two-sample comparisons were conducted using Student’s t-tests.