Antennal transcriptome analysis and expression profiling of odorant binding proteins and odorant receptors in the teak defoliator, Hyblaea puera (Lepidoptera: Hyblaeidae)

Abstract

Background

Insects utilize a sophisticated olfactory system for host and mate location. This process depends on antennal olfactory proteins that facilitate chemical recognition and subsequent signal transduction. The moth Hyblaea puera is a major polyphagous defoliator, yet its underlying olfactory mechanisms remain poorly understood. To elucidate this process, we profiled the complete antennal transcriptomes of adults of both sexes, identified key chemosensory gene families, and reconstructed phylogenetic trees. The tissue-specific expression profiles of key OBP and OR genes were analyzed using quantitative real-time PCR (qPCR).

Results

A repertoire of chemosensory genes was identified through antennal transcriptome analysis of H. puera, comprising 22 odorant-binding proteins (OBPs), 7 chemosensory proteins (CSPs), 46 odorant receptors (ORs), 20 ionotropic receptors (IRs), 9 gustatory receptors (GRs) and 4 sensory neuron membrane proteins (SNMPs). qPCR-based expression profiling revealed a striking female antennae-biased expression pattern for HpueOBP6, HpueOR2, HpueOR7, HpueOR11, HpueOR12, HpueOR13, HpueOR21, HpueOR29 and HpueOR47. On the other hand, a cluster of genes, including HpueOBP3, HpueOBP10, HpueOBP16, HpueOBP19, HpueOBP20, HpueOBP21, HpueORco, HpueOR5, HpueOR33, HpueOR35 and HpueOR43 were predominantly expressed in male antennae. In addition to its significantly high expression in male antennae, HpueOR43 also showed moderate expression in male legs. Notably, HpueOBP12 showed exclusive and high expression in male abdominal tips, and HpueOBP14 was notably abundant in the legs of both sexes.

Conclusion

This study identified 108 key chemosensory genes in H. puera. Phylogenetic analysis enabled their classification into distinct chemosensory gene families. Analysis by qPCR also showed a localized expression pattern in specific tissues for important HpueOBPs and HpueORs, suggesting that several may be putatively involved in recognizing sex pheromones and host plant volatile organic compounds (VOCs); however, these inferred roles require functional validation. Together, these results offer a platform for characterizing the functions of these key proteins and for designing novel control measures that disrupt the pest’s olfactory sensing.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12864-026-13043-9.

Background

Driven by a highly refined sense of smell, insects can identify and differentiate among diverse environmental odors. This olfactory acuity is vital for numerous survival and reproductive activities. The antennae constitute the core of this sensory apparatus, serving as the primary site for olfaction that enables both the detection of external chemical cues and intraspecific signaling [1]. This olfactory process relies on the coordinated action of multiple specialized proteins. The transport of chemical ligands is mediated by odorant-binding proteins (OBPs), which include pheromone-binding proteins (PBPs) and general odorant-binding proteins (GOBPs), as well as chemosensory proteins (CSPs). Subsequent signal recognition and transduction involve odorant receptors (ORs), ionotropic receptors (IRs), gustatory receptors (GRs), and sensory neuron membrane proteins (SNMPs), while odor-degrading enzymes (ODEs) ensure the final clearance of odorant molecules. Chemical stimuli are thereby converted into neuronal signals that are then transmitted to the brain for processing. Collectively, this network forms a sophisticated transduction system that governs critical insect behaviors. Among these, OBPs and ORs are the most extensively studied. They play pivotal roles in specifically recognizing sex pheromones and host volatiles and in initiating downstream signal transduction cascades [2]. For example, it has been found that SfruOBPs (encoded by Sfru10933 and Sfru21187) are crucial for sex pheromone detection in male Spodoptera frugiperda; specifically, Sfru10933 has been shown to be involved in both electroantennogram (EAG) responses and mating behavior, while Sfru21187 affects mating behavior, with the two proteins acting synergistically [3]. For the oriental fruit moth (Grapholita molesta), the receptor Gmol-OR2 is crucial in the male-specific detection of the major sex pheromone component [4]. Similarly, in the guava fruit fly (Bactrocera correcta), BcorOBP19d-2 can bind to multiple host plant volatiles and is involved in the recognition process that enables females to locate host plants for oviposition [5]. Therefore, elucidating the functions of insect olfactory-related genes is essential for revealing the underlying molecular mechanisms of olfactory recognition.

With advances in insect transcriptomics, RNA sequencing has become a key tool for functional genomics research, including gene mining and expression profiling [6]. In recent years, this technology has been extensively applied to major agricultural pests, including the diamondback moth (Plutella xylostella), cotton bollworm (Helicoverpa armigera), and fall armyworm (S. frugiperda). These studies have clarified the molecular mechanisms by which olfactory proteins detect sex pheromones and host plant volatiles. These insights underscore the value of transcriptomics, both for pinpointing regulators of insect behavior and to guide the design of sustainable pest control strategies that disrupt olfaction [7–9].

The teak defoliator, Hyblaea puera Cramer (Lepidoptera: Hyblaeidae) is a polyphagous foliar pest native to Southeast Asia, characterized by its migratory ability. It has a broad host range, infesting at least 46 plant species across 5 families. Facilitated by international trade and the widespread planting of teak, it has now spread to over 30 countries and regions across Oceania, Central America, and Africa [10, 11]. Since its first detection in Hainan and Yunnan provinces of China in 1984 [12], its distribution and damage have continued to expand. Notably, after 2010, frequent outbreaks have occurred in mangrove areas along the southern coast of China, posing a serious threat to the ecological security of Guangxi, Guangdong, and Hainan [13–16]. In China, this pest primarily damages key tree species, including teak (Tectona grandis), grey mangrove (Avicennia marina), Hainan gmelina (Gmelina hainanensis), and greyish-white beautyberry (Callicarpa cana), with teak and grey mangrove being the most preferred hosts [12, 15]. Within mangrove ecosystems, H. puera exhibits a specialized feeding preference for Avicennia marina communities. It can complete up to 11 generations per year, frequently causing extensive canopy defoliation [17]. Long-term monitoring indicates that this sustained damage significantly compromises the structural and functional integrity of mangrove ecosystems, thereby posing a systemic threat to coastal ecological security [18, 19]. The synergistic effects of its high reproductive rate, strong migratory capacity, and the availability of suitable habitats have limited the efficacy of current control measures [17, 20]. Consequently, there is an urgent need to develop innovative technologies for the sustainable management of its populations.

Currently, the molecular mechanism underlying olfactory recognition in H. puera remains unreported. Against this backdrop, this study aimed to characterize the chemosensory gene families of H. puera. Utilizing antennal transcriptome sequencing and bioinformatic analyses, we systematically identified putative chemosensory genes and performed a comparative phylogenetic analysis with related species. To this end, we utilized quantitative real-time PCR (qPCR) to delineate the expression patterns of key OBP and OR genes in various tissues of male and female H. puera adults, for the first time. These results lay the groundwork for subsequent functional investigations into the pest’s chemosensory genes and inform the development of sustainable management strategies that target olfactory pathways.

Results

Transcriptomic analysis of antennae and gene set assembly

Sequencing of antennal transcriptome libraries from both sexes of H. puera on an Illumina HiSeq platform generated 37.81 Gb of high-quality filtered reads. Each sample contributed no less than 6.01 Gb of data, with a Q30 base percentage surpassing 95.54% across all samples. Trinity was used for de novo assembly of all clean reads. After optimization, the final assembly comprised 54,892 unigenes (67,477 transcripts), with an average length of 937.86 bp and an N50 of 1,800 bp. Read mapping evaluation showed that 86.28% to 89.12% of reads from each sample could be mapped back to this assembly. The longest transcript per gene was selected to represent the unigene, as it typically provides the most complete coding sequence representation (Table 1).

Table 1.

Assembly statistics of H. puera transcriptome

Type	Unigene	Transcript
Total number	54,892	67,477
Total base	51,481,187	75,244,089
Largest length (bp)	15,839	15,839
Smallest length (bp)	201	201
Average length (bp)	937.86	1115.11
N50 length (bp)	1800	2243
E90N50 length (bp)	3530	3052
Fragment mapped percent(%)	80.5	90.531
GC percent (%)	41.75	41.36
TransRate score	0.38351	0.42457
BUSCO score	C:96.0%[S:63.2%;D:32.8%]	C:97.1%[S:50.9%;D:46.2%]

BUSCO scores: “C” = complete (S: single-copy; D: duplicated)

Analysis of the antennae transcriptome sequencing data

BUSCO (Benchmarking Universal Single-Copy Orthologs) analysis revealed high completeness of the assembled transcriptome, with the transcript-level assembly achieving 97.1% complete BUSCOs and the unigene-level assembly achieving 96.0% complete BUSCOs (Table 1). These high completeness metrics provide a reliable foundation for subsequent identification of chemosensory gene families.

After expression quantification using RSEM, transcripts with zero expression across all samples were filtered out, yielding 53,851 expressed unigenes (66,174 expressed transcripts) for subsequent gene identification and expression profiling. Among these, 32,056 unigenes (59.53% of the total) were successfully annotated through BLASTx or BLASTn searches against public databases using an E-value threshold of 10^− 5. The distribution of successfully annotated unigenes across the six major functional databases is presented in Fig. 1A. As shown, the number of annotated unigenes ranged from 21,360 (Gene Ontology (GO), 39.67%) to 29,781 (EggNOG Orthologous Groups (EggNOG), 55.30%).

Fig. 1.

Functional annotation results for unigenes derived from the H. puera antennal transcriptome. A Summary of functional annotation. B Gene Ontology (GO) classification. The histogram displays the distribution of unigenes across secondary-level GO terms. C KEGG pathway enrichment analysis. The scatter plot shows the top 20 significantly enriched pathways (p-adjusted < 0.5), colored by their main pathway categories. D Clusters of Orthologous Groups (COG) classification. The histogram summarizes the functional categories of the predicted unigenes

Based on the GO classification system, a comprehensive annotation of the antennal transcriptome was conducted. The annotated unigenes were grouped into three main functional categories: Molecular Function, Biological Process, and Cellular Component. Within Molecular Function, the dominant annotations were “binding” (11,561 unigenes) and “catalytic activity” (8,990 unigenes). Within Biological Process, “cellular process” and “metabolic process” predominated, comprising 9,095 and 7,745 unigenes, respectively. For Cellular Component, “cell part” and “membrane part” contained the highest numbers of annotated unigenes, totaling 10,357 and 6,023, respectively (Fig. 1B).

To systematically decipher the biological functions of genes in the antennal transcriptome of H. puera, we performed KEGG pathway annotation for all unigenes. KEGG annotation revealed that the unigenes were assigned to all six standard Level A pathways, confirming a broad functional coverage. Among all B-level pathways, Signal transduction was the most highly represented, with 2,789 annotated unigenes, substantially outnumbering all other pathways. Additionally, pathways for Translation, Folding, sorting and degradation, and Carbohydrate metabolism were also among the most annotated. It is noteworthy that although the Human Diseases category contained a considerable number of gene annotations, these were predominantly enriched in sub-pathways such as infectious diseases, cancers, and neurodegenerative disorders. This enrichment likely reflects the evolutionary conservation of molecular mechanisms underpinning insect immune defense, cell cycle regulation, and basic neuronal functions (Fig. 1C).

Functional annotation based on the EggNOG database successfully classified the unigenes from the teak trunk borer antennal transcriptome into 22 functional categories. The category for “Posttranslational modification, protein turnover, chaperones” represented the most abundant group of genes. The second most prevalent category was “Intracellular trafficking, secretion, and vesicular transport.” Furthermore, substantial gene numbers were also annotated to categories including “Signal transduction mechanisms,” “Translation, ribosomal structure and biogenesis,” among others (Fig. 1D).

Phylogeny of OBPs

Transcriptomic analysis of H. puera antennae revealed 22 candidate OBPs (Additional file 1: Table S1). Among these, six genes (HpueOBP1, HpueOBP6, HpueOBP12, HpueOBP14, HpueOBP16, and HpueOBP21) possessed complete open reading frames (ORFs) and encoded amino acid sequences longer than 120 residues.

A set of amino acid sequences was employed to infer evolutionary relationships through phylogenetic tree construction, including the OBPs from H. puera and homologous sequences from Diptera (Drosophila melanogaster), Coleoptera (Tribolium castaneum), Hymenoptera (Apis mellifera), and other lepidopteran species (Fig. 2). Based on multiple sequence alignment and phylogenetic tree analysis, the identified OBPs in H. puera were classified into 1 Minus-C OBP, 17 Classic OBPs, and 1 Plus-C OBP.

Fig. 2.

Maximum likelihood phylogenetic tree of H. puera OBPs. The tree includes OBP sequences from Hymenoptera, Coleoptera, Diptera, and Lepidoptera. HpueOBPs identified in this study are highlighted in red. The clades corresponding to classic, minus-C, plus-C OBPs are indicated by shaded backgrounds. Bootstrap support values exceeding 50% are shown at the nodes

Classic OBPs were further classified based on sequence homology and phylogenetic clustering with functionally characterized references revealed 5 putative pheromone-binding proteins (PBPs), including HpueOBP11, HpueOBP16, HpueOBP20, HpueOBP21, and HpueOBP22, and 2 putative general odorant-binding proteins (GOBPs), including HpueOBP9 and HpueOBP19, with the remaining 10 Classic OBPs classified as unclassified Classic OBPs. The phylogeny showed that the five HpueOBPs formed a monophyletic clade within the broader PBP clade, suggesting their potential roles in sex pheromone recognition. While HpueOBP9 and HpueOBP19 clustered within the clade containing lepidopteran GOBP2, confirming their classification as GOBP2 orthologs, multiple sequence alignment using Clustal Omega further revealed that they share 79.52% and 72.60% sequence identity, respectively, with GOBP2 from Galleria mellonella (GenBank accession: QEI46781.1). The phylogenetic tree was well-supported, with the majority of key nodes exhibiting high bootstrap values.

Phylogeny of CSPs

Transcriptomic analysis of H. puera antennae revealed 7 candidate CSPs (Additional file 1: Table S2). Compared to OBPs, CSPs generally exhibit more conserved sequence architectures, typically containing four conserved cysteine residues.

To elucidate the evolutionary relationships of CSPs in H. puera, a phylogenetic tree was constructed using the amino acid sequences of the HpueCSPs, along with homologous sequences from representatives of five insect orders (Fig. 3). Multiple sequence alignment and phylogenetic analysis revealed that HpueCSP1 clustered with CresCSP5 from Clostera restitura (98% bootstrap support), HpueCSP2 with OfurCSP11 from Ostrinia furnacalis (95%), HpueCSP3 with MsepCSP14 from Mythimna separata (93%), HpueCSP6 with OfurCSP15 (74%), and HpueCSP7 with CresCSP6 (98%).

Fig. 3.

Maximum likelihood phylogenetic tree of H. puera CSPs. The tree includes CSP sequences from Hymenoptera, Coleoptera, Diptera, and Lepidoptera. HpueCSPs identified are shown in red. Bootstrap support values exceeding 50% are shown at the nodes

Phylogeny of ORs

Transcriptomic analysis of H. puera antennae revealed 46 candidate ORs (Additional file 1: Table S3). Among the 46 candidate ORs, 44 contained full-length ORFs. To further validate the candidate ORs, we predicted their transmembrane domains (TMDs) using TMHMM v2.0. Among the 46 ORs, 33 (71.7%) were predicted to contain 4–8 TMDs, including 5 with 4 TMDs, 8 with 5 TMDs, 13 with 6 TMDs, 4 with 7 TMDs, and 2 with 8 TMDs. The remaining 13 ORs were predicted to have fewer than 4 TMDs.

A phylogenetic tree was built from an alignment of amino acid sequences, which included the ORs from H. puera along with homologous sequences from Coleoptera, Hymenoptera, Diptera, and other Lepidoptera (Fig. 4). Phylogenetic analysis placed HpueORco within the conserved ORco (odorant receptor co-receptor) clade. Notably, HpueOR43 clustered within the pheromone receptor (PR) clade, suggesting a putative role as a pheromone receptor. Furthermore, most HpueORs were grouped into distinct clades alongside their respective orthologs from closely related species, with these groupings well-supported by high bootstrap values.

Fig. 4.

Maximum likelihood phylogenetic tree of H. puera ORs. The tree includes OR sequences from Hymenoptera, Coleoptera, Diptera, and Lepidoptera. HpueORs are shown in red. OR, PR and ORco clades are indicated by shaded backgrounds. Bootstrap support values exceeding 50% are shown at the nodes

Phylogeny of IRs

Transcriptomic analysis of H. puera antennae revealed 20 candidate IRs (Additional file 1: Table S4). Among them, ten genes (HpueIR1, IR3, IR4, IR5, IR8, IR10, IR13, IR15, IR16 and IR17) contained complete ORFs, encoding proteins with lengths no less than 600 amino acids.

The phylogenetic analysis provided further resolution of the classification of these IRs, which included the 20 IRs from H. puera along with homologous sequences from representatives of Coleoptera, Hymenoptera, Diptera, and other Lepidoptera. Based on phylogenetic analysis following the classification scheme of Benton et al. [21], these IRs were categorized into three groups: 18 antennal IRs, 1 divergent IR (HpueIR9), and 1 non-NMDA iGluR (HpueIR12) (Fig. 5). Within the antennal IRs clade, putative orthologs of the broadly expressed co-receptors IR8a (HpueIR13) and IR25a (HpueIR3) were identified, both with strong bootstrap support (100%). Sequence alignment further revealed that HpueIR13 shares 87.55% sequence identity with CfagIR8a, and HpueIR3 shares 93.35% sequence identity with CfagIR25a, supporting their conserved roles as co-receptor orthologs. Other putative antennal IR subfamilies are also indicated. Detailed classification is provided in Fig. 5.

Fig. 5.

Maximum likelihood phylogenetic tree of IRs from H. puera. The tree includes IR sequences from Hymenoptera, Coleoptera, Diptera, and Lepidoptera. HpueIRs are shown in red. Major clades are labeled, including NMDA iGluR, non-NMDA iGluR (AMPA and Kainate), antennal IRs, and divergent IRs. Conserved lineages within antennal IRs are also indicated. Bootstrap support values exceeding 50% are shown at the nodes

Phylogeny of GRs

We annotated a total of 9 GRs from the H. puera antennal transcriptome (Additional file 1: Table S5).

To examine the phylogenetic relationships of these GRs, we constructed a phylogenetic tree comprising the GRs from H. puera together with homologous sequences from Coleoptera, Hymenoptera, Diptera, and other Lepidoptera (Fig. 6). Multiple sequence alignment and phylogenetic analysis demonstrated that HpueGR1, HpueGR2, and HpueGR9 clustered within a clade containing known sugar receptors from multiple lepidopteran species, confirming their assignment to the sugar receptor lineage. Similarly, HpueGR3, HpueGR7, and HpueGR8 grouped within the fructose receptor lineage, while HpueGR4 and HpueGR6 were placed within the bitter receptor clade, and HpueGR5 clustered with carbon dioxide receptors. These phylogenetic classifications were designated according to previously established GR subfamily assignments in insects [22, 23] and are supported by bootstrap values indicated in Fig. 6.

Fig. 6.

Maximum likelihood phylogenetic tree of GRs from H. puera. The tree includes GR sequences from Hymenoptera, Coleoptera, Diptera, and Lepidoptera. HpueGRs are shown in red. Major clades are labeled, including Sugar, Fructose, Bitter, and Carbon dioxide receptors. Bootstrap support values exceeding 50% are shown at the nodes

Phylogeny of SNMPs

We annotated a total of 4 SNMPs from the H. puera antennal transcriptome (Additional file 1: Table S6). Two distinct subfamilies, SNMP1 and SNMP2, were identified among these SNMPs by phylogenetic analysis (Fig. 7). Among them, HpueSNMP1 clustered within the SNMP1 clade with strong bootstrap support (100%), confirming its classification as an SNMP1 ortholog. The remaining three SNMPs (HpueSNMP2, HpueSNMP3, and HpueSNMP4) were placed within the SNMP2 clade, forming a well-supported group (bootstrap = 100%) together with SNMP2 orthologs from other lepidopteran species.

Fig. 7.

Maximum likelihood phylogenetic tree of SNMPs from H. puera. The tree includes SNMP sequences from Hymenoptera, Coleoptera, Diptera, and Lepidoptera. HpueSNMPs are shown in red. Major clades are labeled, including SNMP1 and SNMP2. Bootstrap support values exceeding 50% are shown at the nodes

All raw transcriptome data have been deposited in the NCBI SRA database under BioProject accession number PRJNA1449938. All sequences identified in this study have been submitted to NCBI and are available under accession numbers PX423514–PX423601 and PZ255782–PZ255795.

Antennal expression profiles of olfactory genes in H. puera adults

Expression levels of chemosensory genes in H. puera antennae were analyzed using TPM normalization, with heatmaps generated separately for each gene family (Fig. 8). Within the OBP family, most genes exhibited moderate to high expression, and most of them showed male-biased expression. Notably, HpueOBP15, HpueOBP19, HpueOBP20, and HpueOBP21 were among the most highly expressed (Fig. 8A). CSPs showed variable expression, with HpueCSP3 and HpueCSP7 highly expressed in both sexes (Fig. 8B). For ORs, most genes were expressed at low abundance. Notably, HpueORco showed the highest expression level among all ORs, with TPM values exceeding 360 in both sexes (Fig. 8C). Within the IR family, expression levels varied considerably; HpueIR3 (IR25a ortholog), HpueIR13 (IR8a ortholog), HpueIR5, HpueIR10 and HpueIR16 showed relatively high expression (Fig. 8D). While all GR genes were found at low abundance in antennae, HpueGR2 constituted an exception as the most highly expressed member, reaching a TPM of 12.5 (Fig. 8E). Among SNMPs, HpueSNMP1 and HpueSNMP3 exhibited high expression, whereas HpueSNMP2 and HpueSNMP4 showed very low expression (Fig. 8F).

Fig. 8.

Expression heatmaps of chemosensory gene families in H. puera antennae. Heatmaps were generated separately for each gene family (OBPs (A), CSPs (B), ORs (C), IRs (D), GRs (E), and SNMPs (F)) using ChiPlot. Expression levels were normalized as transcripts per million (TPM) and subsequently transformed using log₁₀ (TPM + 0.01). Hierarchical clustering was performed using the complete linkage method with Euclidean distance. Red indicates high expression, blue indicates low expression

Selection of reference genes for qPCR

To ensure the reliability of qPCR data, the expression stability of six candidate reference genes (GAPDH, rsp27, RPS3, α-tubulin, β-tubulin, and EF1-α) was evaluated across the four tissues (antennae, thorax, legs, and abdominal tip) from both sexes. Box plots illustrate the variability in Ct values for each reference gene among different tissue types and sexes. Among them, GAPDH showed the widest range of Ct values, indicating poor expression stability, whereas rsp27 and RPS3 exhibited tightly clustered Ct values with minimal variation, suggesting more stable expression (Additional file 3: Fig. S1). Subsequent stability analysis using four algorithms yielded the following results: geNorm, which assesses stability based on the average M-value (lower M-value indicates higher stability), identified rsp27 and RPS3 as the most stable gene pair (M = 0.547) (Additional file 3: Fig. S2). Based on the geNorm analysis of paired variation (Vn/n + 1) values, an assessment was conducted to identify the ideal number of reference genes. The calculated V2/3 value (0.449) exceeded the default threshold of 0.15 (Additional file 3: Fig. S2), indicating that the inclusion of a second reference gene is necessary for accurate normalization. NormFinder analysis revealed that across different tissues and sexes, the stability ranking of the six candidate reference genes was as follows: rsp27 (0.93) > RPS3 (1.02) > β-tubulin (1.89) > α-tubulin (1.93) > EF1-α (2.7) > GAPDH (5.77) (Additional file 3: Fig. S4). Using the standard deviation (SD) and coefficient of variation (CV) of Ct values as criteria, BestKeeper analysis ranked rsp27 as the gene with the smallest CV (4.25) and SD (0.84) (Additional file 3: Table S1). Finally, RefFinder was used to integrate the rankings from the four methods and compute a geometric mean, generating an overall stability ranking of the candidate reference genes (Additional file 3: Fig. S5). Accordingly, the geometric mean of rsp27 and RPS3 was used as the normalization factor for all subsequent qPCR analyses.

Tissue expression analysis of selected OBP and OR genes

We selected 12 OBP and 15 OR genes for qPCR analysis based on high antennal TPM values or complete ORFs, to validate their antennal expression and investigate their expression profiles across four chemosensory tissues (antennae, thorax, legs, and abdominal tip) in H. puera.

The qPCR analysis revealed distinct tissue-specific expression patterns for the examined olfactory genes, which were consistent with the TPM expression trends (Figs. 9 and 10). Specifically, HpueOBP6, HpueOR2, HpueOR7, HpueOR11, HpueOR12, HpueOR13, HpueOR21, HpueOR29 and HpueOR47 showed female-biased antennal expression, with transcript levels significantly higher in females than in males and other tissues. In contrast, HpueOBP3, HpueOBP10, HpueOBP16, HpueOBP19, HpueOBP20, HpueOBP21, HpueORco, HpueOR5, HpueOR33, HpueOR35 and HpueOR43 were markedly enriched in male antennae. Among these, HpueOBP20, HpueOBP21, and HpueOR43 exhibited particularly pronounced male-biased expression. Furthermore, HpueOBP12 showed relatively high expression in male abdominal tip, while HpueOBP14 was notably upregulated in the legs of both sexes.

Fig. 9.

Relative expression levels of candidate OBPs in different tissues of H. puera. Expression levels were measured by qPCR and calculated using the 2^−ΔΔCt method. Values represent relative expression normalized to the geometric mean of two reference genes (rsp27 and RPS3) and calibrated to female antennal expression (set to 1). Tissues are denoted as: An, antennae; T, thorax; L, legs; Ab, abdominal tip. One-way ANOVA followed by Tukey’s post hoc test (P < 0.05) was used to assess statistical significance among tissues, as indicated by superscript letters. Non-detectable expression is reported as “N/A”. Error bars represent SD from three biological replicates (each with three technical replicates)

Fig. 10.

Relative expression levels of candidate ORs in different tissues of H. puera. Expression levels were measured by qPCR and calculated using the 2^−ΔΔCt method. Values represent relative expression normalized to the geometric mean of two reference genes (rsp27 and RPS3) and calibrated to female antennal expression (set to 1). Tissues are denoted as: An, antennae; T, thorax; L, legs; Ab, abdominal tip. One-way ANOVA followed by Tukey’s post hoc test (P < 0.05) was used to assess statistical significance among tissues, as indicated by superscript letters. Non-detectable expression is reported as “N/A”. Error bars represent SD from three biological replicates (each with three technical replicates)

Discussion

In lepidopteran insects, the antenna functions as the primary olfactory organ, containing a diverse array of olfactory proteins that enable the detection of environmental chemical cues [24]. Although the molecular basis of chemoreception has been extensively characterized in many lepidopteran species, it remains poorly understood in H. puera. To address this knowledge gap, we conducted comprehensive analysis of the antennal transcriptome in adult H. puera, including both sexes. Our study identified and characterized key gene families implicated in chemosensation, including OBPs, PBPs, CSPs, ORs, IRs, and GRs. Overall, our findings lay the groundwork for subsequent studies on the olfactory system and its underlying molecular mechanisms in H. puera, as has been demonstrated in other lepidopteran species where antennal transcriptome analyses have facilitated the functional characterization of chemosensory genes [25, 26].

The high BUSCO completeness (unigene: 96.0%; transcript: 97.1%) and substantial functional annotation rates confirm the reliability of our transcriptome dataset for gene discovery and downstream functional studies in H. puera [27]. GO classification provides a systematic framework for functional annotation. Among the 21,360 unigenes annotated in the GO database, enrichment analysis revealed a significant overrepresentation of terms associated with molecular recognition and enzymatic reactions, indicating active cellular and metabolic processes in the antennae of H. puera. Within the cellular component classification, the enrichment of genes associated with membrane structures aligns with the known ultrastructure of olfactory sensilla, in which the dendritic membranes of olfactory neurons are densely populated with chemosensory proteins [2, 28]. KEGG enrichment analysis highlighted Signal transduction as the most prominent pathway, indicating active intracellular signaling that aligns with the antenna’s core olfactory function [29]. Furthermore, the enrichment of pathways for Translation, Protein folding, and Degradation points to vigorous protein synthesis and turnover, essential for maintaining functional olfactory proteins. The enrichment of terms within the Human Diseases category largely mirrors evolutionarily conserved processes like immunity and neuronal function, highlighting the molecular foundations of cellular defense and neural activity in the antennae. Together, these results collectively portray the antenna of H. puera as an olfactory organ defined by active signal transduction and high protein turnover. In the COG classification, “Posttranslational modification, protein turnover, and chaperones” represented the largest category, suggesting highly active protein processing and quality control in the antenna. This is critical for efficient olfaction, which relies on the rapid synthesis, folding, and turnover of olfactory proteins [30]. In summary, the functional profile of the H. puera antenna underscores its role as a specialized chemosensory organ. These molecular features form an efficient olfactory system that enables sensitive detection and rapid response to environmental chemical signals.

In H. puera adults of both sexes, we detected 108 candidate chemosensory genes within the antennal transcriptomes. These genes were annotated as OBPs, CSPs, ORs, GRs, IRs and SNMPs. Notably, 22 OBPs were discovered, with 5 of them further classified as putative PBPs based on phylogenetic analysis. This number of OBPs is comparable to that in Spodoptera litura (21 OBPs [31]) and Chilo suppressalis (26 OBPs [32]), but lower than in H. armigera (34 OBPs [7, 26]) and S. frugiperda (54 OBPs [33]). In summary, compared with previously reported lepidopteran species, H. puera possesses a moderate number of OBPs, which is comparable to that of the oligophagous C. suppressalis, but significantly lower than those of the highly polyphagous H. armigera and S. frugiperda. This difference suggests that the expansion of the OBP gene family does not simply scale linearly with host range. A recent comparative genomic analysis of herbivorous insects indicated that specialist species generally have fewer chemosensory receptors and detoxification enzymes than generalist species. However, this trend appears weaker in Lepidoptera than in Coleoptera and Diptera. Notably, with respect to OBP gene numbers, the difference between specialist (oligophagous/monophagous) and generalist species within Lepidoptera seems less pronounced [34]. For example, the polyphagous H. armigera (34 OBPs) and its oligophagous congener H. assulta (29 OBPs) possess similar numbers of OBPs, showing no expansion that matches their marked difference in host breadth [26]. Furthermore, among highly polyphagous species, the OBP numbers of S. frugiperda (54) and S. litura (21) are markedly different [31, 33]. These results suggest that different polyphagous species may rely on fundamentally different chemosensory strategies. On one hand, as seen in S. frugiperda and H. armigera, polyphagy is associated with an expanded OBP repertoire that enables recognition of a broader range of plant volatiles. On the other hand, in species such as H. puera and S. litura, despite their relatively wide host ranges, we infer that they may depend on a moderately sized OBP repertoire with broad ligand-binding capabilities, combined with the recognition of conserved volatile signatures released by their host plants, thereby achieving efficient host location. Thus, interspecific variation in OBP numbers may reflect alternative pathways of chemosensory adaptation. The antennal transcriptome yielded a reduced repertoire of only seven annotated CSP genes, which is fewer than the numbers found in E. hippophaecolus [35] or S. litura [31]. and H. armigera [7, 26] (each with 18 CSPs), but comparable to dipterans such as D. melanogaster [28]. This low number may reflect three scenarios: genuine lineage-specific contraction, incomplete transcriptome assembly leading to omission of low-abundance paralogues, or predominant expression of these genes in non-antennal tissues. It should be noted that the gene counts reported in cross-species comparisons originate from either transcriptomic or genomic data, and technical differences among studies may introduce certain comparative biases. The present study is based on antennal transcriptomes, and the conclusions therefore await further validation using genome-level data and functional assays.

As soluble proteins secreted into the aqueous sensillum lymph, Both CSPs and OBPs exhibit high affinity for hydrophobic odorants and mediate their transport from the environment to olfactory receptors, thereby enabling the solubilization and circulation of these volatiles in the aqueous sensillum lymph [2, 36]. Phylogenetic analysis revealed that the five putative PBPs of H. puera (HpueOBP11, HpueOBP16, HpueOBP20, HpueOBP21 and HpueOBP22) formed a distinct clade and share high sequence similarity with known lepidopteran PBPs, which putatively supports their involvement in sex pheromone detection. Specifically, HpueOBP22 and HpueOBP16 clustered with EhipPBP1 from Eogystia hippophaecolus, which exhibits strong binding to Z7-14:Ac and E3-14:Ac [34]. Likewise, HpueOBP21 and HpueOBP11 grouped with HarmPBP2 from H. armigera, a protein known to recognize the key pheromone component Z11-16:Ald [37]. These conserved clustering patterns suggest a putative functional conservation among lepidopteran PBPs [38] and tentatively implicate the putative HpuePBPs in sex pheromone recognition ; however, direct binding assays are required to confirm their ligand specificities. In contrast, HpueOBP9 and HpueOBP19 clustered within a well-defined branch containing GOBPs from other lepidopterans, separate from the PBP clade. This phylogenetic placement suggests that they belong to the GOBP subfamily and are putatively involved in recognizing common environmental odorants [39]. Notably, their high sequence identity with a functionally validated GOBP2 from G. mellonella, which binds sex pheromone components and their analogues [40], raises the possibility that these H. puera GOBP2 orthologs might also participate in pheromone-related communication, a hypothesis that awaits experimental testing.

Phylogenetic analysis placed HpueCSP1 in a clade with CresCSP5 from C. restitura (98% bootstrap support), which is highly expressed in 3- to 4-day-old adults and implicated in mating behavior, particularly sex pheromone secretion and recognition [41]. Given the frequent association of CSPs with reproductive behaviors in insects [42, 43], we hypothesize putatively that HpueCSP1 fulfills a comparable role in H. puera, but this requires functional validation. HpueCSP3 clustered with MsepCSP14 from M. separata (93% support). In M. separata, MsepCSP14 shows predominant expression in female antennae and participates in the recognition of host plant volatiles [44]. In contrast, HpueCSP3 exhibited high antennal abundance in both sexes with a pronounced male bias, possibly indicating a sex-specific role. Nevertheless, its presence in female antennae suggests it might also contribute to chemosensory processes common to both sexes, albeit to a lesser extent. HpueCSP2 and HpueCSP6 clustered with OfurCSP11 and OfurCSP15 from O. furnacalis, respectively (95% and 74% support), both of which exhibit mating status-dependent expression patterns [45], hinting at a possible but unconfirmed role in reproductive biology. HpueCSP7 grouped with CresCSP6 from C. restitura (98% support), which is exclusively highly expressed in 1-day-old females [41], raising the possibility that HpueCSP7 putatively contributes to early adult female physiological processes.

In Lepidoptera, ORs play crucial roles in recognizing both plant volatiles and sex pheromones [46]. The number of HpueORs (45) similar to that reported in Spodoptera littoralis (47 ORs [47]) and C. suppressalis (47 ORs [32]). Transmembrane domain prediction revealed that the majority of candidate ORs in H. puera (71.7%) possessed 4–8 TMDs, which is consistent with the structural characteristics reported for insect ORs [48, 49]. The remaining ORs were predicted to have fewer than 4 TMDs, which likely reflects incomplete transcript assembly (missing 5’ or 3’ regions) or atypical receptor structures. Given the established role of Orco as an obligate co-receptor for ORs, which is essential for mediating chemosensory behaviors, including reproduction, host location, and oviposition, in a wide range of insect species [50, 51], we propose that HpueORco serves a similarly essential co-receptor function in H. puera chemosensation. Furthermore, HpueOR43 clustered within a clade of known pheromone receptors. This phylogenetic placement suggests its putative role in sex pheromone detection in H. puera, as this clade includes HarmOR13 (tuned to Z11‑16:Ald [52]), SexiOR13(responsive to Z9,E12‑14:OAc [25]), and SfruOR13 (sensitive to Z9‑14:Ac [53]). However, the sex pheromone composition of H. puera remains unknown, as no chemical identification has been reported to date. Future studies focusing on pheromone component analysis and heterologous expression of HpueOR43 will be necessary to confirm their specific ligand specificities and functional roles. In a separate lineage, HpueOR3 formed a highly supported monophyletic clade with HarmOR42, which is characterized by high specificity for phenylacetaldehyde [54], putatively indicating a potential role for HpueOR3 in detecting plant volatiles such as phenylacetaldehyde in H. puera. However, these functional assignments are based solely on phylogenetic inference and await experimental validation (e.g., heterologous expression or EAG).

IRs are evolutionarily ancient compared to ORs and exhibit diverse sequence divergence. Their expression is not restricted to the antennae, and they function as receptors for a variety of external stimuli, including chemical cues, humidity, and temperature [55–57]. In the present study, the number of HpueIRs (20) is consistent with that of C. suppressalis (20 IRs [32]), higher than those of H. armigera (12 IRs [7]) and Manduca sexta (18 IRs), but lower than that of Bombyx mori (31 IRs [58]). This variation suggests that the size of the IR gene family differs considerably across Lepidoptera, potentially reflecting differences in ecological niche or olfactory complexity among species. Phylogenetic analysis revealed that the majority (18 IRs) belong to the antennal IR clade, including putative orthologs of the broadly expressed co-receptors IR8a (HpueIR13) and IR25a (HpueIR3), both with strong bootstrap support (100%), suggesting their potential roles as co-receptorss [59, 60]. Notably, HpueIR2 belongs to the IR21a family, members of which have been implicated in thermosensation in other insects [61], suggesting a putative conserved role in temperature perception. Other antennal IRs were assigned to diverse subfamilies (IR40a, IR41a, IR60a, IR64a, IR75, and IR93a) implicated in acid, amine, temperature, and humidity detection [56, 62], indicating possible functional conservation. In addition, one divergent IR (HpueIR9) was identified with extremely low expression levels. A non-NMDA ionotropic glutamate receptor (iGluR), designated HpueIR12, was also detected in the antennal transcriptome. In D. melanogaster, neither divergent IRs nor iGluRs are typically expressed in the antenna [21, 59, 63], raising the possibility that these genes may not play a direct role in olfactory reception. Their presence in the antennal transcriptome could be attributed to residual contamination from non-olfactory tissues like hemolymph or low-level transcriptional leakage. Alternatively, this could reflect genuine species-specific expression differences between Lepidoptera and Diptera, although this possibility requires further validation.

The ability to detect sugars and bitter compounds, a function underpinned by GRs, is crucial for insect chemoreception and directly shapes behaviors such as host selection and oviposition [55, 64]. 9 GRs were annotated from the H. puera antennal transcriptome, a count similar to those reported for S. exigua (7 GRs [65]) and P. xylostella (7 GRs [8]). Among these, HpueGR5 was identified as a member of the GR21a family. In D. melanogaster, GR21a functions as a heterodimer with GR63a to mediate CO₂ sensing in antennal neurons [66]. However, no GR63a ortholog was detected in the antennal transcriptome of H. puera. This is not entirely unexpected, as the expression patterns of CO₂ receptor genes vary among insect orders. In Lepidoptera, for example, CO₂ receptor genes are not exclusively expressed in the antennae. In H. armigera, the GR63a ortholog (HarmGR1) is specifically expressed in the labial palps rather than the antennae [67]. Therefore, the absence of a GR63a ortholog in our antennal dataset does not necessarily indicate its genomic absence, it may instead be expressed in other chemosensory tissues such as the labial palps. To fully elucidate the CO₂ sensing mechanism in H. puera, expression profiling of the labial palps and other potential chemosensory tissues would be necessary. Furthermore, HpueGR1, HpueGR2 and HpueGR9 clustered within a clade containing multiple known sugar receptors, putatively ndicating their potential role in recognizing sugar compounds. Similarly, HpueGR3, HpueGR7 and HpueGR8 grouped in a branch with identified fructose receptors, suggesting a putative role in fructose perception. In some female lepidopterans, sugar recognition serves as a key factor informing both host plant choice and oviposition behavior [68, 69]. The high bootstrap values observed at most nodes of the phylogenetic tree support the reliability of the inferred orthologous relationships, which may offer insights for future functional studies.

In H. puera, four SNMPs were identified, including one SNMP1 (HpueSNMP1) and three SNMP2 paralogs (HpueSNMP2, HpueSNMP3, and HpueSNMP4). In Lepidoptera, SNMP1 is abundantly expressed in pheromone-sensitive olfactory sensory neurons and functions as a co-receptor involved in sex pheromone detection [70]. The abundant antennal expression of HpueSNMP1 suggests a putative role in pheromone detection. In contrast, SNMP2 is localized in antennal support cells and participates in the clearance of lipophilic waste products, such as fatty acids resulting from pheromone inactivation [71, 72]. Among the three SNMP2 paralogs, HpueSNMP3 exhibited high expression levels, whereas HpueSNMP2 and HpueSNMP4 showed low expression, suggesting possible subfunctionalization. Determining the exact cellular distribution and biological functions of these genes will necessitate additional experiments.

We analyzed the antennal expression profiles of these olfactory genes using TPM data (Additional file1: Table S1-6) to pinpoint candidate families involved in odor perception. Notable differences in expression abundance were observed among families. Specifically, OR and GR families showed markedly low transcript levels, consistent with their roles as membrane-bound receptors. ORs are typically expressed at low copy numbers on olfactory neuron membranes. Similarly, low GR expression in antennae aligns with their primary function in contact gustation via mouthparts, indicating this tissue is not their major site of action [73]. In contrast, OBP genes exhibited high expression, implying their abundance in antennal lymph is crucial for solubilizing and transporting hydrophobic odorants to initiate recognition.

To validate the above transcriptomic trends, we further examined the spatial expression profiles of selected chemosensory genes. Specifically, the expression of 12 OBP and 15 OR genes in H. puera was analyzed across tissues using qPCR. To ensure reliable qPCR data, the expression stability of six candidate reference genes was evaluated across four tissues (antennae, thorax, legs, and abdominal tip) from both sexes using geNorm, NormFinder, BestKeeper, and RefFinder. geNorm analysis showed that pairwise variation (Vn/n + 1) values for all tissue samples exceeded the default threshold of 0.15, indicating that adding more reference genes would not reduce variation below this threshold. Similar findings have been reported in other insects, such as C. suppressalis [74] and Endoclita signifier [75]. Previous studies suggest that while multiple reference genes are advantageous over a single one, using too many complicates data processing; thus, two reference genes are typically optimal [76, 77]. Accordingly, two reference genes (rsp27 and RPS3) were selected, and their geometric mean was used for normalization in all subsequent qPCR analyses. Corroborating the antenna-specific expression pattern typical of insect olfactory genes, 10 HpueOBPs and all 15 HpueORs were found to be predominantly expressed in the antennae [78]. Notably, HpueOBP6, HpueOR2, HpueOR7, HpueOR11, HpueOR12, HpueOR13, HpueOR21, HpueOR29 and HpueOR47 displayed female antennae-specific enrichment, suggesting their involvement in female-specific behaviors, particularly the detection of host plant volatiles for oviposition site selection [79]. In contrast, a clear male-biased expression pattern was observed for HpueOBP20, HpueOBP21, and HpueOBP16, which were specifically enriched in male antennae. This pattern aligns with the established role of PBPs in binding and transporting female sex pheromones to facilitate mate detection [80]. Similarly, HpueOR5, HpueOR33, HpueOR35, and HpueOR43 also exhibited pronounced male-biased expression. Given that PRs in Lepidoptera are typically male-specific and dedicated to recognizing female-emitted sex pheromones [81], as exemplified by BmorOR1 in B. mori [82] and OnubOR6 in Ostrinia nubilalis [83], we tentatively propose that these four HpueORs may be putatively involved in pheromone recognition in H. puera. HpueORco exhibited strong antennal expression in both sexes, showing a significant male bias (Fig. 10). Notably, its expression level was substantially higher than that of most other ORs, consistent with its conserved role as an obligate co-receptor for odorant receptors. Unlike specific ORs, ORco does not bind odorants directly but forms heteromeric complexes with them to constitute odorant-gated ion channels, which are indispensable for olfactory signal transduction [84]. This abundant antennal expression underscores the central importance of the olfactory pathway in H. puera. Furthermore, two OBP genes showed significant enrichment in non-antennal tissues, suggesting potential functional divergence. For instance, the female ovipositor of Helicoverpa assulta shows higher expression of HassOR31, a receptor implicated in detecting host plant volatiles and mediating oviposition choices [85]. Similarly, implicated in close-range chemical communication, HpueOBP12 was specifically enriched in the male abdominal tip of H. puera, suggesting a function in pheromone-mediated interactions such as residue detection. Additionally, in the legs of both sexes, HpueOBP14 showed higher expression, suggesting a potential role in contact chemosensation.

This study systematically identifies and characterizes the major chemosensory protein families in the teak moth, H. puera. Combining phylogenetic analysis with qPCR expression profiling, we define putative roles for specific OBP and OR genes in mediating sex-specific behaviors, thereby suggesting potential molecular mechanisms that may facilitate host plant location and mate finding. These inferred functions require experimental validation.

Methods

Insect collection and rearing

A laboratory colony of H. puera was established from pupae supplied by the Research Group in Insect Pathology and Insect Pathogenic Microorganisms, Chinese Academy of Forestry Sciences. Following eclosion, adults were maintained at 25 ± 1 °C and 50 ± 5% relative humidity under a 16:8 h (L: D) photoperiod in a controlled insectary. Adults were housed in 2000-mL transparent containers provisioned with a 10% (w/v) honey solution ad libitum as an adult diet. Oviposition sites were provided by securing a layer of surgical gauze over the container opening. Upon hatching, larvae were placed on artificial diet in rearing boxes and kept under identical conditions through to the pupal stage.

Workflow from antennal RNA extraction to illumina sequencing

Antennae from H. puera adults were processed separately. For each of the three biological replicates per sex, 100 pairs of antennae were dissected, pooled in RNAlater, and stored at -80 °C. Following the kit protocol, total RNA was isolated with the Omega RNA Kit (Feybio Biotechnology Co, Guangzhou, China). RNA quality and concentration were assessed using a Nanodrop 8000 spectrophotometer and an Agilent 2100 Bioanalyzer with the RNA Nano Chip (Thermo, Waltham, MA, USA; Agilent, Santa Clara, CA, USA). Antennal transcriptome sequencing of H. puera was conducted by Shanghai Majorbio Bio-pharm Technology Co., Ltd. Six antennal samples (each a pool of 100 pairs of mixed-sex antennae) were subjected to library construction and de novo sequencing on the Illumina HiSeq 4000 platform. Briefly, for library preparation, enriched mRNA was fragmented into inserts of approximately 300 bp. Subsequently, using random hexamer primers, the first cDNA strand was generated via reverse transcription of the fragmented mRNA. Finally, second-strand synthesis was performed using the first-strand cDNA as a template to generate double-stranded cDNA.

The cDNA fragments underwent end repair and 3’ adenylation, with subsequent ligation of Illumina paired-end adapters. The resulting libraries were amplified by PCR for 15 cycles, and fragments of the desired size were purified by excision from a 2% agarose gel. The purified libraries were quantified using a TBS380 (PicoGreen) fluorometric system, clustered on the cBot system via bridge amplification, and subjected to paired-end sequencing (2 × 150 bp) on the Illumina HiSeq platform.

Transcriptome assembly and annotation

The high-quality clean reads obtained from all samples were processed for de novo assembly via the Trinity assembler [86]. The initial transcript set was refined by removing low-confidence transcripts with TransRate (http://hibberdlab.com/transrate/) and CD-HIT (http://weizhongli-lab.org/cd-hit/) [87, 88], and its completeness was evaluated using BUSCO (http://busco.ezlab.org) [27]. Clean reads were then mapped back to the optimized assembly, and the longest transcript from each locus was selected as a unigene to represent the coding sequence for subsequent analyses.

Functional annotation was performed through BLAST homology searches with a stringent E-value threshold of < 1 × 10⁻⁵ [89–91] against several databases, including NR, Swiss-Prot, Pfam, GO, KEGG, and EggNOG. GO terms were further assigned via the Blast2GO pipeline [92–94]. Gene expression levels were quantified using RSEM [95] based on the cleaned reads mapped to the assembled transcriptome. Transcript abundance was normalized as transcripts per million (TPM). Transcripts with zero TPM values across all samples were considered non-expressed and removed from subsequent analyses.

Identification of chemosensory genes in H. puera

To identify putative chemosensory genes, we screened the H. puera antennal transcriptome using tBLASTn searches with known insect (Insecta) protein sequences of OBPs, CSPs, ORs, IRs GRs and SNMPs as queries. Candidate unigenes were initially selected based on NCBI BLASTx alignments using an E-value cutoff of 1e⁻⁵. These candidate sequences for all six gene families were then manually curated to assess sequence integrity, alignment coverage, and similarity. Final candidate gene ORFs were predicted using NCBI ORF Finder. For the candidate ORs, TMDs were predicted using TMHMM v2.0 with default parameters.

Phylogenetic analysis of chemosensory proteins in H. puera

The SignalP-5.0 server was employed to predict signal peptides and their cleavage sites within the candidate OBP and CSP protein sequences [96]. Multiple sequence alignments of chemosensory proteins (including OBPs, CSPs, ORs, IRs, GRs and SNMPs) from H. puera and related insect species were generated with ClustalX [97]. We then generated phylogenetic trees in MEGA X using the Maximum Likelihood method. Node support was evaluated with 1000 bootstrap replicates [98]. The Jones-Taylor-Thornton (JTT) model [99] was applied to estimate the evolutionary distances. For comparative phylogenetic analysis, protein sequences of homologs were sourced from the NCBI database (Additional File 2). The final phylogenetic trees were annotated and formatted using the web-based tool chiplot based on Newick format tree files [100]. Multiple sequence alignments were performed using Clustal Omega with default parameters. Pairwise sequence identities were calculated to further support orthology assignments. The heatmap was constructed using the same platform, with transcript expression levels normalized as TPM and subsequently transformed using log₁₀ (TPM + 0.01).

Quantitative real-time PCR (qPCR)

We collected antennae (An), thorax (T), legs (L), and abdominal tip (Ab) tissues from H. puera, with samples from each sex processed separately. For qPCR analysis, three biological replicates were collected for each tissue and sex, with each biological replicate consisting of pooled tissues from 50 to 150 individuals depending on tissue size and availability. Three technical replicates were performed for each biological replicate. RNA quality control was performed on a Nanodrop 8000 spectrophotometer. Only samples meeting the required purity criteria (A₂₆₀/A₂₈₀ = 1.8-2.0) were used for subsequent experiments. We synthesized first-strand cDNA using the PrimeScript RT Reagent Kit (Takara, Cat. No. RR047A, Japan) with total RNA as template. We designed target-specific primers for OBPs, ORs, and candidate reference genes were designed via NCBI Primer-BLAST with the following parameters: 20 bp length, 40–60% GC content, and amplicon size of 80–150 bp.

To ensure the reliability of qPCR data, the expression stability of six candidate reference genes (GAPDH, rsp27, RPS3, α-tubulin, β-tubulin, and EF1-α) was assessed across the four tissues from both sexes. Box plots were used to present the distribution of Ct values for six candidate reference gene, allowing visualization of expression variability among tissue types. Subsequently, six candidate reference genes were subjected to a comprehensive stability assessment through four software tools, including geNorm, NormFinder, BestKeeper, and RefFinder. geNorm evaluated gene expression stability via the M-value and derived the required number of reference genes from the Vn/Vn + 1 values [101]. According to NormFinder, the most stable gene was identified as the one with the minimal stability value derived from variance analysis [102]. Stability was assessed by BestKeeper through computation of the standard deviation (SD) and coefficient of variation (CV) between gene pairs, following the principle that lower values correlate with higher stability [103]. Finally, RefFinder was employed to comprehensively rank the reference genes by calculating the geometric mean (GM).

We screened the transcriptome data for OBP and OR genes that exhibited high antennal expression (TPM) or contained full-length ORFs. Genes fulfilling these criteria were selected for further investigation of their tissue-specific expression patterns. The expression levels were quantified using SYBR^® Green chemistry on a CFX Connect Real-Time PCR System (Bio-Rad, USA). Each 20 µL reaction contained cDNA template, gene-specific primers, and SYBR^® Green Pro Taq HS Premix. The thermal cycling protocol included an initial denaturation, 40 cycles of amplification (with annealing at 60 °C), followed by melt curve analysis to confirm amplicon specificity. For data normalization in the RT-qPCR analysis, all primer sequences for reference genes and target genes are listed in Additional File 3: Table S1.

Statistical analysis

We analyzed the tissue-specific expression of candidate OBP and OR genes in both sexes using the 2^(–ΔΔCt) method to calculate relative expression levels [104]. Relative expression levels were normalized to the expression in female antennae, which was set to 1. A one-way ANOVA was conducted to determine significant expression differences, coupled with Tukey’s post hoc comparison (P < 0.05) using the SAS OnDemand platform. GraphPad Prism version 8.0.0 was utilized for the construction of bar graphs to present the experimental results.

Authors’ contributions

Q.D. and Q.Z. contributed equally to this work. Q.D. undertook the experimental and computational work (molecular genetics, sequence alignment, and statistics) and drafted the manuscript. Q.Z. assisted in the molecular genetic research on select chemosensory genes (ORs). H.L. provided supervision and experimental guidance. Y.X.C. assisted in insect rearing. P.F.L. conceived and designed the study, and was primarily responsible for writing and revising the manuscript. The final version was read and approved by all authors.

Funding

This work was supported by the National Key R&D Program of China (2023YFC2604802). The funders were not involved in the conception of the study, data acquisition and interpretation, the decision to publish, or the writing of the manuscript.

Data availability

All data supporting the findings of this study are encompassed within the main article and its integral supplementary files. All raw transcriptome data have been deposited in the NCBI SRA database under BioProject accession number PRJNA1449938 (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1449938/). All sequences identified in this study have been submitted to NCBI and are available under accession numbers PX423514–PX423601 and PZ255782–PZ255795.

Declarations

Ethics approval and consent to participate

H. puera (Lepidoptera: Hyblaeidae) is a well-documented invasive forestry pest. We used insects from a laboratory-maintained colony for this study. This species is not listed on the “List of Protected Animals in China” or other international endangered species lists. All procedures complied with accepted ethical guidelines for insect research to minimize suffering.

Consent for publication

Not applicable. This manuscript does not contain any individual person’s data in any form.

Competing interests

The authors declare no competing interests.