Genome‐wide identification of olfactory receptor and odorant‐binding protein gene families and their roles in Heliothine chemosensory evolution Authors: Rong Guo, Boyang Ni, Megan L. Fritz Published: 2026 ABSTRACT AbstractChemosensory systems play key roles in the survival and reproductive success of insects. Two large and diverse chemosensory gene families, odorant receptors (ORs) and odorant‐binding proteins (OBPs), play critical roles in insect chemosensation and mediate odour‐guided behaviours. In the process of insect chemosensation, odorants from the environment pass through pores in the antennal sensilla and become soluble in the sensillar lymph, either directly on contact or by binding to an OBP. Solubilized odour molecules diffuse through the lymph until they reach and activate their cognate ORs, sending electrophysiological signals to the insect brain. To better understand the evolutionary roles of OR and OBP gene families among members of the Heliothinae, we systematically characterized these two gene families in Chloridea virescens (Lepidoptera: Noctuidae). A total of 81 ORs and 49 OBPs were identified genome‐wide. Based on the number and positions of conserved cysteine residues, the OBPs were classified into three types: 34 Classic OBPs, 8 Minus‐C OBPs and 7 Plus‐C OBPs. Phylogenetic analyses identified potential gene duplications and losses within OR and OBP gene families among members of the Heliothinae, which may be associated with differences in their volatile sensation and olfactory behaviours. Further motif and structural analyses identified a conserved region that was unique among pheromone receptors and predicted as key residues of the binding pocket, implying its critical role in pheromone detection. Future work should focus on experimentally validating its function. Overall, our findings provide important insights into how chemosensory gene evolution contributes to ecological adaptation and reproductive isolation in the Heliothine moths. INTRODUCTION The chemosensory system plays a crucial role in many insect behaviours, such as feeding, mating, finding oviposition sites and avoiding harmful factors such as predators (Leal, 2013). Accurate sensation and recognition of odorants are critical to fitness, particularly for pheromone components and host plant odorants (Pelosi et al., 2018; Zhang, Walker, & Wang, 2015). Generally, odorants from the environment are detected by olfactory sensory neurons (OSNs) housed in multiporous cuticular hairs on insect antennae. Hydrophobic odorants are first adsorbed to the waxy cuticular surface and then pass through pores of the olfactory sensillum. The molecules become water‐soluble through binding to odorant‐binding proteins (OBPs) in the lymph fluid, and they are carried to OSN dendrites, finally binding with odorant receptors (ORs) on the dendritic membrane (Sakurai et al., 2014). The binding triggers the activation of ion channels and converts chemical signals into electrical signals. Insect ORs have seven transmembrane domains but a topology that is distinct from mammalian ORs, which are considered to be G protein‐coupled receptors (GPCRs) (Carraher et al., 2015; Zhang et al., 2025). Membrane topology analysis of the insect ORs indicates that they have an intracellular N‐terminus and an extracellular C‐terminus, while other GPCRs have the C‐termini inside the cell (Carraher et al., 2015; Smart et al., 2008; Tsitoura et al., 2010). Insect ORs function as a hetero‐tetramer in OSNs: a co‐receptor Orco, previously known as OR83b (Vosshall & Hansson, 2011), which is highly conserved between species (Krieger et al., 2003), and an odorant‐binding subunit (i.e., a tuning OR). The numbers of odorant‐binding subunits vary widely between species. For example, head lice contain as few as 10 OR genes while ants contain as many as 300 OR genes (Kirkness et al., 2010; Smith et al., 2011). Pheromone receptors (PRs) are highly specialized ORs that are involved in pheromone‐based sexual communications, and the molecular mechanisms of pheromone recognition have been characterized in several lepidopteran families, such as Bombycidae (Sakurai et al., 2014), Crambidae (Wanner et al., 2010) and Noctuidae (Jiang et al., 2014; Liu et al., 2013). Female pheromones are a complex blend of volatiles, where the exact combinations of compounds and their ratios make up a species‐specific blend, to which the conspecific male's PRs respond. The species specificity of ligand‐receptor recognition enables precise mating and maintains reproductive isolation and species boundaries (Roelofs & Comeau, 1969), so even modest changes in PRs can lead to shifts in mating preferences or speciation (Cao et al., 2021). OBPs are small (ca. 150 amino acids), water‐soluble proteins secreted into the lymph fluid surrounding the OSNs. The first OBPs were identified as pheromone‐binding proteins (PBPs) in moths (Vogt & Riddiford, 1981). They are widely believed to bind with different odorants and solubilize and transport them across the lymph fluid to ORs expressed on the dendritic membrane (Pelosi & Maida, 1995; Vogt & Riddiford, 1981). In addition to this major function, OBPs have been proposed to protect odorants from degradation or deactivation by odorant‐degrading enzymes (Leal, 2013). A study by Larter et al. (2016) also revealed a novel role for OBPs in buffering sudden changes in the odorant levels of the environment, suggesting that OBPs have a broader functional role than odorant transport alone. Generally, OBPs can be classified into different sub‐families according to their putative functions: PBP, having the ability to bind to pheromone components (Vogt & Riddiford, 1981); general odorant‐binding protein (GOBP) that were considered to respond to general odorants, especially volatiles from host plants (Khuhro et al., 2017); and antennal binding protein (ABP; Gong et al., 2009). The 3D structural features of insect OBPs have attracted interest as a potential target for pest control strategies (Venthur & Zhou, 2018; Zhou et al., 2010). Structurally, insect OBPs can also be categorized into classes (‘Classic’, ‘C‐minus’, ‘C‐plus’ and ‘Duplex or Dimer’) based on the number and pattern of cysteines. The six conserved cysteines of Classic OBPs form three disulphide bonds that stabilize their tertiary structure and help define a hydrophobic binding cavity (Gong et al., 2009; Hekmat‐Scafe et al., 2002; Manoharan et al., 2013; Vieira et al., 2007; Vieira & Rozas, 2011; Zhou et al., 2004). Heliothine moths (Lepidoptera: Heliothinae) include some of the world's most devastating pest species. Chloridea virescens (previously Heliothis virescens), also known as tobacco budworm, is an important agricultural pest of many crops in the United States, attacking cotton, flax, soybean and tobacco in its larval stages (Blanco et al., 2007; Cunningham & Zalucki, 2014; Kogan et al., 1989). Two other closely related Heliothine species, Helicoverpa zea (Lepidoptera: Noctuidae) and Helicoverpa armigera (Lepidoptera: Noctuidae), are also significant agricultural pests in different regions of the world: H. zea in the Americas, while H. armigera in temperate Asia, Africa, South America and Australia (Pearce et al., 2017; Sharma, 2005). Although C. virescens, H. zea and H. armigera are all polyphagous, with overlapping host plant use, C. virescens (tobacco budworm) shows a preference for laying eggs on Solanaceae (e.g., tobacco and tomato), while the other two Helicoverpa do not (Cunningham & Zalucki, 2014). As their chemosensory systems play a critical role in mate finding and host plant recognition, a better understanding of the chemosensory receptors could contribute to studying the evolutionary process of speciation and exploring environmentally friendly approaches for pest control. The insect chemosensory gene repertoire evolves through an extremely rapid birth‐and‐death process, with expansions driven by gene duplication and neofunctionalization, and contractions caused by pseudogenization and deletion (Sánchez‐Gracia et al., 2009). Comparative genomic studies have shown that insect chemosensory genes appear in tandem arrays and likely originate by unequal crossing over, then evolve independently from each other (Eyun et al., 2017; Nozawa & Nei, 2007; Sánchez‐Gracia et al., 2009; Vieira et al., 2007). To study the evolutionary divergence of OR and OBP gene families among members of the Heliothinae, we manually curated OR and OBP genes in the publicly available reference genome of C. virescens (GenBank assembly accession: GCA_002382865.1; Fritz et al., 2018). The number, positions and sequences of ORs and OBPs were identified. OBP genes were then classified into three different types (Classic, Minus‐C and Plus‐C), according to the number and position of conserved cysteine residues. Phylogenetic relationships among chemosensory genes from the Heliothinae were examined to identify gene duplication, loss and sequence evolution and infer the gene family evolutionary histories. Additional structural analysis revealed that sequence motif may be structurally or functionally important to sensation and detection of odorants critical to maintenance of Heliothine species boundaries. EXPERIMENTAL PROCEDURES Genome‐wide identification of OR and OBP genes The webapollo platform hosted by the insect 5000 genomes initiative (i5k; Poelchau et al., 2014) was used for manual curation of OR and OBP genes in the reference genome of C. virescens (GenBank assembly accession: GCA_002382865.1; Fritz et al., 2018). Publicly available OR and OBP proteins from closely related species, H. zea, H. armigera and Bombyx mori (Lepidoptera: Bombycidae), and some chemosensory receptor proteins of C. virescens (in NCBI) were used to identify specific protein motifs or BLAST against the genome assemblies. Existing gene models from a structural annotation were either accepted or modified in light of both these tblastn results and guidance from spliced RNA‐seq reads from adult male and female antennae (BioProject accession number: PRJNA49697; USDA‐ARS‐BCIRL & USDA‐ARS‐SIMRU, 2010; GenBank accession numbers: SRR6417884 and SRR6417880). The genes were named according to the best‐‐matched ORs/OBPs of H. zea and H. armigera, which were identified and manually curated by Pearce et al. (2017). The predicted C. virescens OBP protein sequences were aligned to one another with ClustalW (Thompson et al., 2003) and were classified as ‘Classic’, ‘Minus‐C’ and ‘Plus‐C’ OBPs according to the conserved number and spacing of cysteine residues in the peptide sequences (Gong et al., 2009; Manoharan et al., 2013; Vieira & Rozas, 2011). Sequence alignment and phylogenetic tree construction To study the evolutionary divergence of OR and OBP gene families among members of the Heliothinae, we compared the OR and OBP protein sequences from C. virescens to closely related Heliothine species: Chloridea subflexa (previously Heliothis subflexa) (Lepidoptera: Noctuidae), H. zea, H. armigera and Helicoverpa assulta (Lepidoptera: Noctuidae). As a lepidopteran species with one of the best‐characterized genomes (International Silkworm Genome Consortium, 2008), B. mori was chosen as an outgroup to identify lineage‐specific gene duplications, losses or functional innovations. In Pearce et al. (2017), 83 ORs, one Orco and 40 OBPs, and 81 ORs, one Orco, and 40 OBPs had been manually annotated in the reference genomes of H. armigera and H. zea (GenBank assembly accessions: GCA_002156985.1, GCA_002150865.1), respectively. For the phylogenetic analysis, four and eight ORs of H. armigera and H. zea, respectively, were removed as the protein translations did not start with methionine (M). In addition to the 80 annotated C. virescens ORs, the protein sequence of HvirOR11 (GenBank accession number: QLF97429.1), which could not be found in the assembly due to its quality, was also included in the phylogenetic analysis. Six PRs (HsubOR6, HsubOR11, HsubOR13‐16) and HsubOrco from C. subflexa were added (Cao et al., 2021). The identified full‐length ORs from B. mori (Liu et al., 2014; Tanaka et al., 2009) and H. assulta (Jiang et al., 2014; Wu et al., 2019; Xu et al., 2015) were also added. Four fragmented OBP protein sequences were also removed from each of H. zea and C. virescens prior to phylogenetic analysis of the OBP gene family, as the protein translations did not start with methionine (M). The identified full‐length OBPs from B. mori (Gong et al., 2009; Vieira & Rozas, 2011) and H. assulta (Chang et al., 2017; Li, Yang, et al., 2009; Sun et al., 2012) were used for analysis. Because no PBPs were annotated from the reference genomes of H. armigera and H. zea, publicly available protein sequences of H. armigera PBP 1–3 (GenBank accession numbers: CAC08212.1, ACD01993.1, AAO16091.1) and H. zea PBP (GenBank accession number: AAC36315.1) were identified and downloaded from GenBank for the phylogenetic analysis. The full list of OR/OBP proteins used for phylogenetic analysis can be found in the Supporting Information. The protein sequences of the selected ORs and OBPs were aligned using the MUSCLE algorithm (Edgar, 2004), and a maximum likelihood phylogenetic tree was constructed by IQ‐TREE version 3.0.1 (Nguyen et al., 2015) using VT amino acid substitution model. To infer node support, 1000 bootstrap replicates were calculated. The resultant tree was plotted using iTOL (https://itol.embl.de/; Letunic & Bork, 2021). Potential gene duplication and loss events were identified from the trees according to the number of corresponding orthologues, under the assumption that every species should have one orthologue for each gene. Gene tree‐species tree reconciliation To reconstruct the evolutionary history of OR and OBP gene families and distinguish true evolutionary events from phylogenetic noise, we employed the probabilistic reconciliation tool GeneRax v2.0 (Morel et al., 2020). Chemosensory genes from C. subflexa and H. assulta were excluded from the analysis, as no systematic, genome‐wide manual curation was performed for these species. The Heliothinae species phylogeny described in Cho et al. (2008) was used as the species tree. We utilized the UndatedDTL probabilistic model, which accounts for Duplication, Transfer and Loss (DTL) events while correcting gene tree topology using the Subtree Pruning and Regrafting (SPR) strategy. This approach reconciled the gene trees with the species tree to minimize the joint likelihood of sequence evolution and reconciliation costs. The resulting reconciled trees were visualized using iTOL (https://itol.embl.de/; Letunic & Bork, 2021) and analysed to quantify the number of gene duplications, losses and transfers along each branch of the species phylogeny. Read depth estimation of OR and OBP genes To estimate the read depth of OR and OBP genes, the paired‐end libraries used in de novo assembly were aligned to the reference genome of C. virescens (GenBank assembly accession: GCA_002382865.1; Fritz et al., 2018) using Bowtie2 version 2.5.1 (Langmead & Salzberg, 2012) with default parameters. The resulting SAM files were converted to BAM files and sorted using samtools version 1.10 (Li, Handsaker, et al., 2009). The read depth of each gene was measured by samtools depth, and both the average and standard deviation were calculated across the length of each gene. Identification of conserved domains and motif composition analysis For each OR and OBP gene, the amino acid sequence was used to identify conserved domains on the conserved domain database (CDD) of NCBI (https://www.ncbi.nlm.nih.gov/Structure/bwrpsb/bwrpsb.cgi; accessed on 1 December 2024). The protein motifs were discovered using the MEME (version 5.5.7; Bailey et al., 2009) online server with classic mode (https://meme-suite.org/meme/tools/meme). A zoops (Zero or One Occurrence per Sequence) distribution pattern was employed, and the five most significant motifs were obtained with a minimum width of six and a maximum width of 50. The search was performed using the default expectation‐maximization settings with a maximum of 1000 iterations. Background amino acid frequencies were estimated from the input dataset. The results were visualized using TBtools‐II version 2.154 (Chen et al., 2023). Homology modelling of 3D structures AlphaFold3 (Abramson et al., 2024) was used to generate three‐dimensional (3D) structural models for the OR and OBP proteins with default parameters (https://alphafoldserver.com/; accessed on 7 January 2025). For each protein, the model with the highest predicted confidence was selected to ensure reliable structural inference. The pLDDT score (predicted Local Distance Difference Test) was produced by AlphaFold3 to indicate how reliable each residue of a predicted protein structure is. Next, we identified putative ligand‐binding pockets using P2RANK (Krivák & Hoksza, 2018) via the PRANK server (https://prankweb.cz/; Jendele et al., 2019; accessed on 10 January 2025), retaining only the top‐ranked pockets with the highest probability scores for the subsequent analysis. For each protein family, multiple sequence alignment (MSA) was performed using Clustal Omega (ClustalO; Sievers & Higgins, 2014) on the EMBL‐EBI server (https://www.ebi.ac.uk/jdispatcher/msa/clustalo; accessed on 12 January 2025). Amino acid residues of the predicted binding pockets were mapped onto the MSA. To quantify both pocket‐specific and overall protein conservation, we calculated the Shannon entropy (H) per alignment column, where H = −∑pi*log2(pi), ignoring gap positions. Lower entropy values indicate higher conservation, while higher values reflect greater sequence variability. Because the residues directly involved in ligand binding are often more conserved to maintain functional specificity, especially when the proteins bind to similar compounds, mean entropy values were then compared between pocket and non‐pocket columns, and statistical significance was evaluated by t‐test to determine whether the binding pockets exhibited significantly higher or lower variability relative to the rest of the protein. TABLE 1The olfactory receptor (OR) gene family annotated from the reference genome of Chloridea virescens.HvirORHzeaORHz_annotation_idHa_annotation_idHarmORScaffoldStartEndDirectionOR1HzeaOR1HzOG200691HaOG200691HarmOR1NWSH01000176.1234,098238,032−OR2NWSH01002961.1362912,151−MissingHzeaOR3HzOG200973HaOG200973HarmOR3OR4HzeaOR4HzOG200977HaOG200977HarmOR4NWSH01000145.1114,618118,386+OR5HzeaOR5HzOG200938HaOG200938HarmOR5NWSH01002259.175,48484,104−OR6HzeaOR6HzOG200931HaOG200931HarmOR6NWSH01000007.126,26030,918+OR6bHzeaOR6HzOG200931HaOG200931HarmOR6NWSH01000007.134,79453,163+OR7HzeaOR7HzOG200856HaOG200856HarmOR7NWSH01000040.133,00536,729+OR8HzeaOR8HzOG200985HaOG200985HarmOR8NWSH01006510.1625313,277−OR9HzeaOR9HzOG200747HaOG200747HarmOR9NWSH01002968.14172775−OR10HzeaOR10HzOG200920HaOG200920HarmOR10NWSH01007896.137,22039,263+MissingHzeaOR11HzOG201020HaOG201020HarmOR11OR12HzeaOR12HzOG200855HaOG200855HarmOR12NWSH01000040.121,71424,836+OR13NWSH01001866.145,14249,299−MissingHzeaOR13‐likeHzOG215628HaOG215628HarmOR13‐likeOR14HzeaOR14HzOG200928HaOG200928HarmOR14NWSH01000007.178,50683,663−OR15HzeaOR15HzOG200929HaOG200929HarmOR15NWSH01000007.168,29177,211+OR16HzeaOR16HzOG200930HaOG200930HarmOR16NWSH01000007.158,30764,441+OR17HzeaOR17HzOG200971HaOG200971HarmOR17NWSH01002272.111,81124,656+OR18HzeaOR18HzOG200915HaOG200915HarmOR18NWSH01003453.136216057+OR19HzeaOR19HzOG200978HaOG200978HarmOR19NWSH01004225.112,71618,929−OR20HzeaOR20HzOG200982HaOG200982HarmOR20NWSH01000786.117,34624,481+OR21HzeaOR21HzOG200944HaOG200944HarmOR21NWSH01002810.113,23315,497+OR22HzeaOR22HzOG200939HaOG200939HarmOR22NWSH01002259.185,09290,794−OR23HzeaOR23HzOG200904HaOG200904HarmOR23NWSH01005468.136,87442,304−OR24HzeaOR24HzOG200972HaOG200972HarmOR24NWSH01002098.110,22511,881+MissingHzeaOR25HzOG200988HaOG200988HarmOR25OR26HzeaOR26HzOG200850HaOG200850HarmOR26NWSH01001847.138,53441,144+OR27HzeaOR27HzOG200861HaOG200861HarmOR27NWSH01001630.131,21035,701+OR28HzeaOR28HzOG200822HaOG200822HarmOR28NWSH01002174.117,78020,561+OR29HzeaOR29HzOG200984HaOG200984HarmOR29NWSH01005276.114443487+OR30HzeaOR30HzOG200919HaOG200919HarmOR30NWSH01007896.130,88233,394+MissingHzeaOR31HzOG200779HaOG200779HarmOR31OR32HzeaOR32HzOG201024HaOG201024HarmOR32NWSH01003434.124,71833,623+OR33HzeaOR33HzOG200821HaOG200821HarmOR33NWSH01002174.1472614,297+MissingHzeaOR34HzOG200793HaOG200793HarmOR34OR35HzeaOR35HzOG200880HaOG200880HarmOR35NWSH01003963.115,96320,481+OR36HzeaOR36HzOG200946HaOG200946HarmOR36NWSH01002232.154129570−OR37HzeaOR37HzOG200979HaOG200979HarmOR37NWSH01000780.1173,644179,639−OR38HzeaOR38HzOG200917HaOG200917HarmOR38NWSH01001579.1123,876125,677+MissingHzeaOR39HzOG200916HaOG200916HarmOR39OR40HzeaOR40HzOG200839HaOG200839HarmOR40NWSH01004938.1988218,813+OR41HzeaOR41HzOG200827HaOG200827HarmOR41NWSH01001711.136,17042,596+OR41‐likeHzeaOR41‐likeHzOG212112HaOG212112HarmOR41‐likeNWSH01003130.135,53238,975+OR42HzeaOR42HzOG200780HaOG200780HarmOR42NWSH01002573.144,93745,844+MissingHzeaOR43HzOG200797HaOG200797HarmOR43OR44HzeaOR44HzOG201018HaOG201018HarmOR44NWSH01002690.120,04827,274+OR45HzeaOR45HzOG201021HaOG201021HarmOR45NWSH01000683.139,92961,566+OR46HzeaOR46HzOG200864HaOG200864HarmOR46NWSH01000975.118,82124,356+MissingHzeaOR47HzOG200772HaOG200772HarmOR47MissingHzeaOR48HzOG200783HaOG200783HarmOR48MissingHzeaOR49HzOG200943HaOG200943HarmOR49OR50HzeaOR50HzOG200941HaOG200941HarmOR50NWSH01002259.194,789101,740+MissingHzeaOR51HzOG200927HaOG200927HarmOR51OR52HzeaOR52HzOG200820HaOG200820HarmOR52NWSH01002174.111,84814,297+OR53HzeaOR53HzOG200932HaOG200932HarmOR53NWSH01006314.128,98238,256+OR54HzeaOR54HzOG200914HaOG200914HarmOR54NWSH01003519.143,34245,897−OR55HzeaOR55HzOG200840HaOG200840HarmOR55NWSH01001309.113,47814,547−OR56HzeaOR56HzOG201026HaOG201026HarmOR56NWSH01000442.1154,283158,788−OR57HzeaOR57HzOG201025HaOG201025HarmOR57NWSH01003434.135,52141,175+OR58HzeaOR58HzOG200947HaOG200947HarmOR58NWSH01002695.110,36113,284−OR59HzeaOR59HzOG201019HaOG201019HarmOR59NWSH01002690.1846111,151+OR60HzeaOR60HzOG200799HaOG200799HarmOR60NWSH01007359.1102473−MissingHzeaOR61HzOG200940HaOG200940HarmOR61OR62HzeaOR62HzOG200942HaOG200942HarmOR62NWSH01002259.192,11393,819+OR63HzeaOR63HzOG200798HaOG200798HarmOR63NWSH01002961.12561751−OR64HzeaOR64HzOG200918HaOG200918HarmOR64NWSH01001579.1109,385110,552+MissingHzeaOR65HzOG208187HaOG208187HarmOR65OR66HzeaOR66HzOG208188HaOG208188HarmOR66NWSH01002174.1974097+OR67HzeaOR67HzOG204092HaOG204092HarmOR67NWSH01002968.141266253−OR68HzeaOR68HzOG201254HaOG201254HarmOR68NWSH01000058.145,66650,843+OR69HzeaOR69HzOG201234HaOG201234HarmOR69NWSH01000302.158,05463,219+OR70HzeaOR70HzOG214106HaOG214106HarmOR70NWSH01001752.112,00726,950+OR71HzeaOR71HzOG204587HaOG204587HarmOR71NWSH01001446.1153510,573+OR72HzeaOR72HzOG204580HaOG204580HarmOR72NWSH01001574.121,57927,879+MissingHzeaOR73HzOG206392HaOG206392HarmOR73MissingHzeaOR74HzOG200987HaOG200987HarmOR74MissingHzeaOR75HzOG200989HaOG200989HarmOR75MissingHzeaOR76HzOG200990HaOG200990HarmOR76MissingHzeaOR77HzOG201793HaOG201793HarmOR77OR78HzeaOR78HzOG206527HaOG206527HarmOR78NWSH01006065.143,84946,374+MissingHzeaOR79HzOG207186HaOG207186HarmOR79OR80NWSH01002690.112,74814,293+OR81NWSH01002634.12452165+OR82NWSH01000442.1142,581147,694−OR83NWSH01000229.1602323,022+OR84NWSH01003458.1738710,221−OR85NWSH01001625.177,35583,869+OR86NWSH01002690.12113206+OR87NWSH01003458.114,48817,423−OR88NWSH01007896.133,45036,270+OR89NWSH01004177.110,49413,504−OR90NWSH01006537.111,30113,303−OR91NWSH01002810.140278937+OR92NWSH01002810.112002602+OR93NWSH01002810.164438937+OR94NWSH01003581.116,31217,839+OR95NWSH01000145.1290,063301,383−OR96NWSH01003458.12802178−OR97NWSH01005768.122805895+ The olfactory receptor (OR) gene family annotated from the reference genome of Chloridea virescens. RESULTS Genome‐wide identification of OR and OBP genes In total, we manually annotated 81 ORs (including one Orco; Table 1) and 49 OBPs (Table 2) from the C. virescens reference genome (GenBank assembly accession: GCA_002382865.1; Fritz et al., 2018) and named them according to the best matched ORs/OBPs of H. zea and H. armigera (Pearce et al., 2017). Our results indicated that the OBP genes were clustered in the reference genome of C. virescens, and at multi‐OBP loci, OBPs were frequently arrayed sequentially and often oriented in the same direction. For example, the 49 OBPs were distributed across 15 scaffolds (of 8826 scaffolds), and almost half of them (24 out of 49) were clustered within a 100 kb region of scaffold NWSH01000044.1 (Figure S1). OBP classification Insect OBPs are divergent, and the overall pairwise sequence similarity is modest (Hekmat‐Scafe et al., 2002). The spacing pattern of the six conserved cysteines is the most typical feature of classical insect OBPs and has been described in Drosophila and B. mori (Gong et al., 2009; Hekmat‐Scafe et al., 2002; Vieira et al., 2007; Zhou et al., 2004). Our alignment of the C. virescens full‐length classical OBP proteins showed the expected pattern and positions of the six conserved cysteines (Figure 1). Following the naming proposed by Hekmat‐Scafe et al. (2002), we identified three types of OBPs in C. virescens: 34 Classic OBPs that have the canonical six‐cysteine pattern, eight Minus‐C and seven Plus‐C OBPs that have fewer or more than six conserved cysteine residues, respectively (Table 2). Plus‐C OBPs are atypical OBPs or ‘dimer’ OBPs, which usually contain two domains that are homologous to the domain of Classic OBPs (Manoharan et al., 2013; Vieira & Rozas, 2011). Alignment of the full‐length classical odorant‐binding proteins (OBPs) of Chloridea virescens. The protein sequences were aligned by ClustalW. Conserved residues are shown in red, and the six highly conserved cysteines are highlighted as red and marked as C1–C6. TABLE 2The OBP gene family annotated from the reference genome of Chloridea virescens and their classifications.HvirOBPHzeaOBPHz_annotation_idHa_annotation_idHarmOBPScaffoldStartEndDirectionLength (aa)TypeGOBP1HzeaGOBP1HzOG200773HaOG200773HarmGOBP1NWSH01003619.172687915−166ClassicGOBP2HzeaGOBP2HzOG200774HaOG200774HarmGOBP2NWSH01000361.1112,098114,160+164ClassicOBP1HzeaOBP1HzOG200804HaOG200804HarmOBP1NWSH01000044.1115,144117,327+149ClassicOBP2HzeaOBP2HzOG200790HaOG200790HarmOBP2NWSH01000044.174,18075,575+144ClassicOBP3HzeaOBP3HzOG200802HaOG200802HarmOBP3NWSH01000044.196,59497,721+149ClassicOBP4HzeaOBP4HzOG200806HaOG200806HarmOBP4NWSH01000044.1140,536141,903+116Minus‐COBP5HzeaOBP5HzOG200805HaOG200805HarmOBP5NWSH01000044.1130,415138,664+149ClassicOBP6HzeaOBP6HzOG200803HaOG200803HarmOBP6NWSH01000044.1100,824105,689+149ClassicOBP7HzeaOBP7HzOG200791HaOG200791HarmOBP7NWSH01000044.179,09680,676+150ClassicOBP8HzeaOBP8HzOG200787HaOG200787HarmOBP8NWSH01000044.161,44662,694−123ClassicOBP10HzeaOBP10HzOG200809HaOG200809HarmOBP10NWSH01000044.1157,750160,797+138Minus‐COBP12HzeaOBP12HzOG200644HaOG200644HarmOBP12NWSH01004598.121,61523,142−153PlusCOBP13HzeaOBP13HzOG200969HaOG200969HarmOBP13NWSH01000208.1182,287185,013+143ClassicOBP14HzeaOBP14HzOG200841HaOG200841HarmOBP14NWSH01008731.132033811+121Minus‐COBP15HzeaOBP15HzOG200853HaOG200853HarmOBP15NWSH01001428.167,00768,231+170ClassicMissingHzeaOBP16HzOG200643HaOG200643HarmOBP16OBP17HzeaOBP17HzOG200863HaOG200863HarmOBP17NWSH01001889.136404149+139ClassicOBP18HzeaOBP18HzOG200788HaOG200788HarmOBP18NWSH01000044.167,70571,824+148ClassicOBP19HzeaOBP19HzOG200786HaOG200786HarmOBP19NWSH01000044.152,14953,682−151ClassicOBP22HzeaOBP22HzOG200789HaOG200789HarmOBP22NWSH01000044.172,17073,555+148ClassicOBP25HzeaOBP25HzOG200808HaOG200808HarmOBP25NWSH01000044.1154,235157,424+149ClassicOBP26HzeaOBP26HzOG200807HaOG200807HarmOBP26NWSH01000044.1147,121153,027−156ClassicOBP27HzeaOBP27HzOG200800HaOG200800HarmOBP27NWSH01000044.185,99088,126+143ClassicOBP28HzeaOBP28HzOG200801HaOG200801HarmOBP28NWSH01000044.190,12492,889−154ClassicMissingHzeaOBP31HzOG200796HaOG200796HarmOBP31OBP32HzeaOBP32HzOG200781HaOG200781HarmOBP32NWSH01003491.113,89717,386+228PlusCOBP33HzeaOBP33HzOG200967HaOG200967HarmOBP33NWSH01000549.189,63691,519−145Minus‐COBP34HzeaOBP34HzOG200966HaOG200966HarmOBP34NWSH01000549.193,72296,167−137Minus‐COBP35HzeaOBP35HzOG211753HaOG211753HarmOBP35NWSH01000509.196,40097,035+214ClassicOBP36HzeaOBP36HzOG200881HaOG200881HarmOBP36NWSH01000018.1120,612122,013−174ClassicOBP37HzeaOBP37HzOG200925HaOG200925HarmOBP37NWSH01006551.116,54722,213+194ClassicMissingHzeaOBP38HzOG200818HaOG200818HarmOBP38OBP39HzeaOBP39HzOG201016HaOG201016HarmOBP39NWSH01000610.140,22674,140−187ClassicMissingHzeaOBP41HzOG200642HaOG200642HarmOBP41OBP42HzeaOBP42HzOG200636HaOG200636HarmOBP42NWSH01006041.120064853−203PlusCOBP43HzeaOBP43HzOG200638HaOG200638HarmOBP43NWSH01006041.110,87712,402−139ClassicMissingHzeaOBP44HzOG200640HaOG200640HarmOBP44OBP45HzeaOBP45HzOG201656HaOG201656HarmOBP45NWSH01002197.1107,188108,457−151PlusCMissingHzeaOBP46HzOG201657HaOG201657HarmOBP46OBP47HzeaOBP47HzOG206450HaOG206450HarmOBP47NWSH01000044.149,75850,650−133ClassicOBP48NWSH01000044.183,83785,196+150ClassicOBP49NWSH01000044.1109,391110,854+149ClassicOBP50NWSH01000044.1137,478138,664+90Minus‐COBP51NWSH01000044.1145,015146,916+152ClassicOBP52NWSH01000361.1130,050131,132−167ClassicOBP53NWSH01000361.1121,965123,080+123Minus‐COBP54NWSH01000361.1125,279126,082+166ClassicOBP55NWSH01000361.1119,429120,561+172ClassicOBP56NWSH01000044.170,16471,824+144ClassicOBP57NWSH01006551.120372509+105Minus‐COBP58NWSH01000044.1111,807113,195+149ClassicOBP59NWSH01000044.1118,033120,613+144ClassicOBP60NWSH01004598.154437092−164PlusCOBP61NWSH01004598.116854806−179PlusCOBP62NWSH01004598.111,86613,459−180PlusC The OBP gene family annotated from the reference genome of Chloridea virescens and their classifications. Genetic divergence from phylogenetic analysis The phylogenetic trees of ORs and OBPs were constructed using the maximum likelihood method with 1000 bootstraps. In the phylogenetic tree of ORs, the Orco proteins were clustered as expected (Figure 2a, highlighted in light red). Orco is a special OR (also known as OR83b) that is highly conserved in most insect species. In addition, PRs from the Heliothine species were clustered with, but were distantly related to, B. mori ORs (Figure 2a, highlighted in light purple and Figure 2b). This suggests that distinct gene expansion and divergence events occurred in the Heliothine species and B. mori. A recent gene duplication of OR6 was identified in C. virescens, with the two tandem copies having very similar protein sequence identities (86.48%). Unrooted maximum likelihood phylogenetic tree of Chloridea virescens olfactory receptors (ORs) with other Lepidoptera ORs. Tree scales represent the number of amino acid substitutions per site. (a) Phylogenetic tree based on OR protein sequences from C. virescens, Chloridea subflexa, Helicoverpa zea, Helicoverpa armigera, Helicoverpa assulta and Bombyx mori. The C. virescens (red), C. subflexa (green), H. zea (pink), H. armigera (blue), H. assulta (yellow) members are shown. The pheromone receptor (PR) and Orco clades are highlighted in light purple and light red, respectively. (b) Phylogenetic tree based on PR protein sequences from C. virescens, C. subflexa, H. zea, H. armigera, H. assulta and B. mori. Red: C. virescens. The clades of OR6, OR14a and OR14b are highlighted as light green, light red and light yellow, respectively. The bootstrap values of the branches are indicated on the nodes. Hvir: C. virescens, Hsub: C. subflexa, Hzea: H. zea, Harm: H. armigera, Hass: H. assulta, Bmor: B. mori. In the phylogenetic tree of OBPs, PBPs and GOBPs were placed in the same clade (Figure 3, highlighted in light red), which was also observed in B. mori (Gong et al., 2009). Our results also indicated a gene expansion in the ‘Minus‐C’ group of B. mori (OBP 25–27) (Figure 3, highlighted in light blue). All of them were clustered in a small region of chromosome 5 in the genome of B. mori (Gong et al., 2009), indicating a recent rapid gene expansion and diversification. Another rapid gene expansion was observed in the Heliothine species, next to OBP11 of B. mori (Figure 3, highlighted in light blue). The OBPs from Heliothine species were more closely related than B. mori's OBP, which suggests a gene expansion and diversification in the Heliothinae. Among them, HvirOBP50 (a unique OBP in C. virescens) and HvirOBP4 were next to each other in the genome, indicating a recent gene duplication in C. virescens. Unrooted maximum likelihood phylogenetic tree based on odorant‐binding protein (OBP) protein sequences from Chloridea virescens (red), Helicoverpa zea (pink), Helicoverpa armigera (blue), Helicoverpa assulta (yellow) and Bombyx mori. Tree scale represents the number of amino acid substitutions per site. The pheromone‐binding protein (PBP)/general odorant‐binding protein (GOBP) clade is highlighted in light red, and two gene diversification events are highlighted in light blue. The bootstrap values of the branches are indicated on the nodes. Hvir: C. virescens, Hzea: H. zea, Harm: H. armigera, Hass: H. assulta, Bmor: B. mori. Phylogenetic reconciliation using GeneRax (Morel et al., 2020) revealed a high rate of gene duplication in C. virescens, including 17 duplications in ORs and 6 in OBPs (Tables S2 and S3). Notably, 7 of the 17 OR duplications and 5 of the 6 OBP duplications involved genes located on the same scaffold and in adjacent positions, indicative of recent tandem duplications (highlighted in yellow in Figures S2 and S3). This lineage‐specific expansion coincides with the expansion of host plant use to Solanaceae in C. virescens, suggesting a potential association between chemosensory gene expansion and adaptation to diverse host plants. Read depth estimation of OR and OBP genes We estimated read depth of the identified OR and OBP genes to test whether putative gene duplications represent true gene duplications or instead arise from haplotype‐specific alternative contigs (Tables S4 and S5). The average read depth of the whole genome was 54.75. The overall read depth of OR genes was 55.91 ± 18.57, and that of OBP genes was 58.65 ± 11.03, indicating that most of the identified chemosensory genes were true gene duplications. However, five ORs (HvirOR30, HvirOR33, HvirOR70, HvirOR88 and HvirOR90) and two OBPs (HvirOBP6 and HvirOBP50) only had approximately half of the expected coverage (average read depth < 30×), suggesting that they may come from the alternative contigs. Identification of conserved domains and motifs Analysis for the presence of conserved domains via NCBI‐CDD indicated the presence of a ~360 amino acid 7tm_6 superfamily domain in almost all (80 of 82) OR protein sequences (Figure 4). The domain with Pfam ID 7tm_6 is present in seven transmembrane receptors that are most likely Drosophila ORs (Paysan‐Lafosse et al., 2025). Furthermore, five significantly enriched ungapped motifs were searched in the OR protein sequences (Figure 4). Our results showed that three of the five motifs appeared more frequently and were mostly located towards the end of the protein sequences. The numbers of sites contributing to the construction of the three motifs were 68, 56 and 74 for motifs 1, 2 and 3, respectively, with a width of 21, 29 and 21 amino acids, respectively (Figure S4). We also found two ‘clade‐specific’ motifs: motif 4 only showed up in the PR clade, and motif 5 was unique in the clade HvirOR58‐HvirOR50; these two motifs may be associated with different types of odorants. Conserved domains and motifs in Chloridea virescens olfactory receptors (ORs) via NCBI‐CDD and MEME. Five significant enriched ungapped motifs were searched in the protein sequences. For the consensus sequences of the motifs, please see Figure S4. The top five most significant motifs were also identified from the OBP protein sequences (Figure 5, Figure S5). Our results indicated that OBPs in the same clade tended to have the same motifs, and a possible reason is that these OBPs are involved in binding molecules with similar structures. For example, the ‘GOBP clade’ had motifs 3 and 5, and the ‘Plus‐C clade’, including HvirOBP12, HvirOBP60‐62, had motifs 3 and 4. A PBP_GOBP superfamily domain was commonly found in OBPs (40 of 49), except for the Plus‐C OBPs (also found in HvirOBP32), implying their different functions in odorant detection. Conserved domains and motifs in Chloridea virescens odorant‐binding proteins (OBPs) via NCBI‐CDD and MEME. Five significant enriched ungapped motifs were searched in the protein sequences. For the consensus sequences of the motifs, please see Figure S5. Homology modelling and binding site prediction AlphaFold3 (Abramson et al., 2024) was applied to predict three‐dimensional (3D) structures of the two protein families, and only the highest‐confidence models were selected for binding site predictions using P2RANK (Jendele et al., 2019; Krivák & Hoksza, 2018). The predictions showed that some closely related members share relatively similar binding pockets, implying their potential roles in the detection of molecules with similar structures. Our results were consistent with the phylogenetic analysis. For example, HvirOR1 shares over 50% identity of the binding pocket with its neighbour HvirOR70; two closely related ORs, HvirOR30 and HvirOR88, also have more than 50% identity in their predicted binding pockets. In Figure 6, we showed the predicted structures and binding pockets of a few OR proteins that are involved in pheromone detection in C. virescens, where the colour corresponds to different confident scores (dark blue: pLDDT > 90; blue: 90 > pLDDT > 70; yellow: 70 > pLDDT > 50; orange: pLDDT < 50). These ORs shared four conserved motifs (motifs 1–4), and motif 4, which was a motif that only appeared in the PR clade, was predicted as key residues of the binding sites (Table S1). In particular, the neutral residues isoleucine (I), leucine (L) and glutamine (Q) of motif 4 were key residues of the binding pocket, mostly with a motif: LXXILXXQ. Homology modelling and binding site prediction of pheromone receptors (a) HvirOR6, (b) HvirOR11, (c) HvirOR13, (d) HvirOR14, (e) HvirOR15, (f) HvirOR16 in Chloridea virescens, where the colours correspond to different confidence scores from AlphaFold3 (dark blue: pLDDT > 90; blue: 90 > pLDDT > 70; yellow: 70 > pLDDT > 50; orange: pLDDT < 50). The binding pockets are coloured in red. For each gene family, amino acids of the predicted binding pockets were mapped to the MSA to quantify pocket‐specific and whole‐protein conservation, both assessed by Shannon entropy. For ORs, the mean Shannon entropy of the pocket region was 2.537, compared to 0.987 for the non‐pocket region (p‐value = 8.309 × 10−36). The mean Shannon entropy of the pocket region in OBPs was 2.517, and that of the non‐pocket region was 1.408, with a p‐value of 9.489 × 10−54 (Figure S6). The statistical significance demonstrated the diversification of the predicted binding sites rather than a rigid, highly conserved one, which is not surprising as these chemosensory proteins are involved in various volatile detection. DISCUSSION Antennal transcriptome data have been widely used to identify chemosensory genes due to its low cost and relatively easy data generation (Dong et al., 2023; Hu et al., 2022; Jiang et al., 2021; Yang et al., 2024; Zhang, Wang, et al., 2015). However, transcriptome‐based annotations may underestimate the total number of genes compared to genome‐wide identification, and one reason is that lowly expressed genes and pseudogenes are often not captured. In the study, we leveraged the structural annotations and existing evidence from transcriptome and protein analyses, to annotate 81 ORs, including one Orco, as well as 49 OBPs from the C. virescens reference genome. Our results showed that OBP genes were highly clustered in the reference genome of C. virescens, but OR genes appeared more scattered throughout the genome, having only a few clusters. The chromosomal clustering of OBPs was first observed in Drosophila melanogaster (Hekmat‐Scafe et al., 2002; Vogt et al., 2002) and is believed to be associated with the expression and regulation of OBPs (Vogt et al., 2002). Across the genomes of 12 Drosophila species, 69% of OBP genes are arranged in 10 clusters (Vieira et al., 2007). In B. mori, the OBP genes are not randomly distributed in the genome, and most of them appear to be in close physical proximity within a specific region, for example, 20 OBPs are clustered on five chromosomes (Gong et al., 2009). The same clustering distribution of OBPs was also observed in mosquitos: Anopheles gambiae (Xu et al., 2003), Anopheles sinensis (He et al., 2016), Culex quinquefasciatus (Ryazansky et al., 2024). Previous research showed that OBP genes in the same cluster can be co‐expressed in adult olfactory sensilla, suggesting that the clustering of these genes is linked to their co‐regulation (Vogt et al., 2002). An explanation for the multiple, widely dispersed gene clusters of C. virescens ORs is that inter‐chromosomal translocation events are more frequent in ORs than OBPs (Conceição & Aguadé, 2008). Chloridea virescens has more identified OBPs than other annotated Heliothine species: H. zea and H. armigera both have 40 OBPs in their reference genomes (Pearce et al., 2017). A model species in Lepidoptera, B. mori, has 44 identified OBPs and 43 of them are full‐length (Gong et al., 2009; Vieira & Rozas, 2011). One possible reason for the increase in C. virescens is the different qualities of their reference genomes. The reference genome of C. virescens is more fragmented than the other two Helicoverpa species, and considering the estimated read depth of OBP genes, HvirOBP6 and HvirOBP50 may not reflect true gene duplication events but rather come from alternative haplotype‐specific contigs. The highly fragmented assembly raises the possibility that haplotypic duplications may inflate gene counts. Although we performed read depth analyses to help distinguish true gene duplications from assembly artefacts, this limitation should still be considered in future analysis. Phylogenetic analyses revealed expected clustering patterns. Orco genes are clustered in the OR tree as expected, due to their high similarity and conservation across insect species. PRs from different Heliothine species also cluster together. This was expected because females of different Heliothine species create their pheromone blends with some of the same volatile compounds used in different ratios. Males can respond to the pheromone compounds, but the strength of the response may vary according to the corresponding physiological or functional assays (Cao et al., 2023; Jiang et al., 2014; Mo et al., 2023; Wang et al., 2024). Because these PRs can respond to some of the same ligands, there is also some sequence similarity at the protein level. Notably, within the PR clade (Figure 2b), OR14 has two copies in H. armigera, H. zea and H. assulta, but only one copy in the C. virescens genome. This pattern may reflect a potential gene loss in C. virescens and C. subflexa, or a gene duplication in the Helicoverpa species. Our results also showed a recent gene duplication of OR6 in C. virescens, and the two copies are located adjacent to each other in the genome. However, considering the fragmented reference genome and limited structural annotation in the region, further validation will be required to confirm the duplication. The PBP sub‐family of OBPs contains proteins that bind and solubilize the hydrophobic pheromone components and transfer them to the specific PRs. Surprisingly, no PBPs were annotated from H. armigera and H. zea by Pearce et al. (2017), although previous studies have identified several PBPs in Helicoverpa species using RT‐PCR (Callahan et al., 2000; Dong et al., 2017; Guo et al., 2012; Wang et al., 2004; Zhang et al., 2012). Blast analyses confirmed that these PBPs are present in the genomes of H. armigera and H. zea (GenBank assembly accessions: GCA_002156985.1, GCA_002150865.1) but were not annotated during manual curation; therefore, we manually added HarmPBP 1–3 and HzeaPBP to the phylogenetic analysis. An explanation for the missing PBPs by Pearce et al. (2017) is the quality limitation of the first reference genomes of H. armigera and H. zea, which were used here because some chemosensory receptor gene families had been manually curated. With a fragmented short‐read reference genome, these genes could be fragmented or missing from the identified genes. Phylogenetic reconciliation using GeneRax (Morel et al., 2020) revealed a high rate of chemosensory gene expansion in C. virescens after speciation, consistent with an adaptive response to the heterogeneous olfactory environments of polyphagous species. Although this approach provides a powerful framework for jointly inferring gene duplication and loss while accounting for the species tree, the limitations of our results need to be considered. In particular, chemosensory gene annotation is highly sensitive to reference genome quality, raising the possibility that some inferred gene duplications and losses may reflect undetected, fragmented or misannotated genes rather than true evolutionary events. Considering the quality of the C. virescens genome, we manually examined the identified duplication events within the species, and genes located on the same scaffold and in adjacent positions are more likely to reflect true duplication events. Gene duplication followed by neo‐ or sub‐functionalization is a well‐recognized mechanism generating molecular and functional diversity in sensory systems (Sánchez‐Gracia et al., 2009). In Drosophila, approximately two‐thirds of gene duplications undergo neofunctionalization (Assis & Bachtrog, 2013). Chemosensory gene duplications may be followed by positive selection on the ligand‐binding site (Saad et al., 2018) to detect new host volatiles or pheromones (neofunctionalization), or the duplicated genes may become more specialized in detecting specific odours instead of one gene that responds to several similar odours (sub‐functionalization). In Heliothine species, lineage‐specific expansions of chemosensory gene families may provide a molecular basis for ecological diversification and host adaptation. However, it remains challenging to connect these molecular changes to evolutionary, behavioural and ecological changes, as the functions of most chemosensory genes are unknown. Due to their large size (>40 kDa) and variability, experimentally resolving the three‐dimensional structures of insect ORs has been extremely challenging (Carraher et al., 2015), and most previous studies focused on understanding OR's structure via transmembrane topology prediction (Cao et al., 2021; Liu et al., 2018; Zhang et al., 2014). In our structural analyses, we used AlphaFold3, which leverages advanced AI to predict high‐resolution, three‐dimensional protein structures and identified key residues involved in ligand‐receptor interactions. We found a ‘PR‐specific’ motif (motif 4) within the binding sites of all PRs, and the motif appears to be under purifying selection, suggesting its functional importance and evolutionary constraint. Our results supported a model that the shared structural element within the binding pockets of different PRs plays an important role in the interactions with diverse pheromone components, and several possible explanations are as below:The motif may play a role in OBP‐OR recognition. While no OBP‐OR recognition models have been validated in invertebrates, Matarazzo et al. (2002) proposed an OBP‐OR direct recognition model in humans. In this model, high‐affinity interaction between OBP and OR suggested that under physiological conditions, ORs may be specifically paired with an OBP partner in the absence of odorant (Matarazzo et al., 2002). However, the recognition model may be unique to humans or vertebrates and may not be the case in insects, as they are evolutionarily divergent.Pheromone recognition across different PRs appears to rely on a conserved binding site (Del Mármol et al., 2021), and the shared motif plays a critical role in maintaining the structural integrity of the PR binding pocket. Recent research has identified diverse motifs among different groups of human ORs, and the motifs can interact with other regions of transmembrane domains, contributing to the stabilization of the receptor's 3D structure (Billesbølle et al., 2023; Ryu et al., 2024).The shared motif may directly contribute to ligand binding by forming hydrogen bonds, polar interactions or other contacts that anchor ligands in the binding pockets. Structural studies have shown that some conserved residues in the binding pocket can generate additional anchoring points for various odorant structures, in line with the receptor's wide range of recognition capabilities (Choi et al., 2023; Comte et al., 2025; Del Mármol et al., 2021).The pocket might have a set of residues that provide a baseline for binding, while amino acid diversification at key positions enables precise recognition of specific pheromone molecules. A recent study in B. mori demonstrated that males can distinguish two nearly identical compounds—bombykol and bombykal—through the formation of a reversible covalent bond (Jang et al., 2025). This work proposed a binding mode between PRs and pheromone compounds and showed that a key covalent bond allows moths to discriminate between molecules that differ only in a single hydrogen atom (Jang et al., 2025). The motif may play a role in OBP‐OR recognition. While no OBP‐OR recognition models have been validated in invertebrates, Matarazzo et al. (2002) proposed an OBP‐OR direct recognition model in humans. In this model, high‐affinity interaction between OBP and OR suggested that under physiological conditions, ORs may be specifically paired with an OBP partner in the absence of odorant (Matarazzo et al., 2002). However, the recognition model may be unique to humans or vertebrates and may not be the case in insects, as they are evolutionarily divergent. The motif may play a role in OBP‐OR recognition. While no OBP‐OR recognition models have been validated in invertebrates, Matarazzo et al. (2002) proposed an OBP‐OR direct recognition model in humans. In this model, high‐affinity interaction between OBP and OR suggested that under physiological conditions, ORs may be specifically paired with an OBP partner in the absence of odorant (Matarazzo et al., 2002). However, the recognition model may be unique to humans or vertebrates and may not be the case in insects, as they are evolutionarily divergent. Pheromone recognition across different PRs appears to rely on a conserved binding site (Del Mármol et al., 2021), and the shared motif plays a critical role in maintaining the structural integrity of the PR binding pocket. Recent research has identified diverse motifs among different groups of human ORs, and the motifs can interact with other regions of transmembrane domains, contributing to the stabilization of the receptor's 3D structure (Billesbølle et al., 2023; Ryu et al., 2024). Pheromone recognition across different PRs appears to rely on a conserved binding site (Del Mármol et al., 2021), and the shared motif plays a critical role in maintaining the structural integrity of the PR binding pocket. Recent research has identified diverse motifs among different groups of human ORs, and the motifs can interact with other regions of transmembrane domains, contributing to the stabilization of the receptor's 3D structure (Billesbølle et al., 2023; Ryu et al., 2024). The shared motif may directly contribute to ligand binding by forming hydrogen bonds, polar interactions or other contacts that anchor ligands in the binding pockets. Structural studies have shown that some conserved residues in the binding pocket can generate additional anchoring points for various odorant structures, in line with the receptor's wide range of recognition capabilities (Choi et al., 2023; Comte et al., 2025; Del Mármol et al., 2021). The shared motif may directly contribute to ligand binding by forming hydrogen bonds, polar interactions or other contacts that anchor ligands in the binding pockets. Structural studies have shown that some conserved residues in the binding pocket can generate additional anchoring points for various odorant structures, in line with the receptor's wide range of recognition capabilities (Choi et al., 2023; Comte et al., 2025; Del Mármol et al., 2021). The pocket might have a set of residues that provide a baseline for binding, while amino acid diversification at key positions enables precise recognition of specific pheromone molecules. A recent study in B. mori demonstrated that males can distinguish two nearly identical compounds—bombykol and bombykal—through the formation of a reversible covalent bond (Jang et al., 2025). This work proposed a binding mode between PRs and pheromone compounds and showed that a key covalent bond allows moths to discriminate between molecules that differ only in a single hydrogen atom (Jang et al., 2025). The pocket might have a set of residues that provide a baseline for binding, while amino acid diversification at key positions enables precise recognition of specific pheromone molecules. A recent study in B. mori demonstrated that males can distinguish two nearly identical compounds—bombykol and bombykal—through the formation of a reversible covalent bond (Jang et al., 2025). This work proposed a binding mode between PRs and pheromone compounds and showed that a key covalent bond allows moths to discriminate between molecules that differ only in a single hydrogen atom (Jang et al., 2025). Pheromone‐based control strategies represent a promising approach for pest management through disruption of moth mating behaviour via PR inhibition (Venthur & Zhou, 2018). Although C. virescens is effectively controlled by Bt crops in many regions of the world, it remains an important pest in tobacco, where no transgenic variety is available (Blanco, 2012). Chloridea virescens serves as a valuable model species for studying related agriculture pests within the Heliothinae sub‐family. Insights gained from C. virescens PR structures can be applied to design pheromone disruptors for other economically significant pests, paving the way for next‐generation, precision‐targeted pest management strategies. CONCLUSION In conclusion, we identified 81 ORs (including one Orco) and 49 OBPs from the C. virescens reference genome. Three types of OBPs were classified according to the pattern of six conserved cysteines: 34 classic OBPs, 8 Minus‐C OBPs and 7 Plus‐C OBPs. In addition, a phylogenetic analysis was performed to identify potential gene duplication and gene loss of OR and OBP gene families among members of the Heliothinae. Two novel gene expansion events were observed in the ‘Minus‐C’ group of B. mori and the Heliothine moths. The gene diversification events could be associated with the differences of odorant detection and olfactory‐mediated behaviours between the Heliothine moths. We also identified a ‘PR‐specific’ motif localized in the predicted binding pocket of PRs, which appears to be under purifying selection and likely plays a critical role in ligand binding and pheromone specificity. Our study provides important insights into the critical features of these PRs to understand pheromone specificity and enables further investigation into how mutations or structural variations in the motif affect ligand binding and, consequently, species‐specific pheromone perception and behaviour. AUTHOR CONTRIBUTIONS Rong Guo: Conceptualization; methodology; formal analysis; writing – original draft; writing – review and editing; visualization; software; data curation. Boyang Ni: Formal analysis; visualization; software. Megan L. Fritz: Conceptualization; methodology; supervision; funding acquisition; writing – review and editing; resources. CONFLICT OF INTEREST STATEMENT The authors declare that they have no conflicts of interest related to the research, authorship or publication of this manuscript. Supporting information Figure S1. OBP gene cluster presents in scaffold NWSH01000044.1 (scaffold length = 319,136 bp). Genes are depicted as triangles: red indicates forward transcription (+) and blue indicates reverse transcription (−). Figure S2. The inferred OR gene tree reconciled with the species tree. Nodes are labelled as D (duplication), S (speciation) or T (horizontal gene transfer). Species are coloured as follows: Chloridea virescens (pink), Helicoverpa zea (green), Helicoverpa armigera (purple) and Bombyx mori (black). Duplications involving genes located on the same scaffold and in adjacent positions are highlighted in light yellow. Figure S3. The inferred OBP gene tree reconciled with the species tree. Nodes are labelled as D (duplication), S (speciation) or T (horizontal gene transfer). Species are coloured as follows: Chloridea virescens (pink), Helicoverpa zea (green), Helicoverpa armigera (purple) and Bombyx mori (black). Duplications involving genes located on the same scaffold and in adjacent positions are highlighted in light yellow. Figure S4. Identification of conserved motifs in HvirORs via MEME. Figure S5. Identification of conserved motifs in HvirOBPs via MEME. Figure S6. The Shannon entropy distribution of the pocket and non‐pocket regions in (A) HvirORs and (B) HvirOBPs. Data S1. HvirORs. Data S2. HvirOR_OBP_cDNA. Data S3. HvOBPs. Data S4. ModelSelection_OBP. Data S5. ModelSelection_OR. Data S6. OBP_phylogenetic. Data S7. OR_phylogenetic. Table S1. Predicted binding site residues of PRs in Chloridea virescens. Table S2. The number of diversification events in OR gene family per species, estimated by Generax (Morel et al., 2020). Table S3. The number of diversification events in OBP gene family per species, estimated by Generax (Morel et al., 2020). Table S4. Read depth of OR genes in Chloridea virescens. Table S5. Read depth of OBP genes in Chloridea virescens.