R. pedestris were fed with bean sprouts and maintained at a temperature of 26 ± 1°C under a 14:10 h light:dark photoperiod. Female and male adults, as well as larvae, were placed in insect cages for reproduction. Fifth instar larvae were collected and raised separately to obtain three-day-old virgin adults. The heads, thoraxes, abdomens, legs, wings, and antennae of virgin adults were collected. All collected samples were immediately frozen in liquid nitrogen and stored at ˗80°C for future use.

Genome data, gene, protein, RNA and annotation files of R. pedestris were obtained from the Genome Warehouse (https://ngdc.cncb.ac.cn/gwh/Assembly/18849/show). We used the protein data of R. pedestris and blasted with different database of Nr (Non-Redundant Protein Sequence), Nt (nucleotide sequence database), UniProt (The Universal Protein Resource), KOG (Clusters of orthologous groups for eukaryotic complete genomes), eggNOG (evolutionary genealogy of genes: No-supervised Orthologous Groups), Interpro (the integrative protein signature), GO (Gene Ontology), and KEGG (Kyoto Encyclopedia of Genes and Genomes) databases (Huang et al., 2021), so R. pedestris proteins were annotated based on homology (These data were derived from a previous genome paper). We acquired the genes and ORF sequences of OBPs and CSPs by corresponding protein ID using R. pedestris genomic database. To ensure the accuracy of gene sequences, we used OBP and CSP genes to blast with Nr (Non-Redundant Protein Sequence) database in NCBI BLAST (http://blast.ncbi.nlm.nih.gov/). We also selected some genes (RpedOBP8, RpedOBP37, RpedCSP5) to clone and sent to sequencing to verify the correctness of the sequences. A total of 49 OBPs and 25 CSPs were obtained based on the similarity analysis. Putative N-terminal signal peptides of all OBPs and CSPs were predicted using SignalP 4.1 (http://www.cbs.dtu.dk/services/SignalP/), (Petersen et al., 2011).

Total RNA was extracted using a FastPure^® Cell/Tissue Total RNA Isolation Kit (Vazyme, Nanjing, China) following the manufacturer’s instructions, and RNA quality was checked using a spectrophotometre (NanoDrop^TM 2000; Thermo Fisher Scientific, United States). Single-stranded cDNA templates were synthesised from 1 μg of total RNA from various tissue samples using the MonScript™ RTlll Super Mix with dsDNase (Two-Step) (Monad, Shanghai, China).

The qRT-PCR primers for 49 OBPs and 25 CSPs (Supplementary Table S1) were designed using Beacon Designer 7.9 (PREMIER Biosoft International, CA, United States). Expression profilings of RpedOBPs and RpedCSPs were performed using qRT-PCR in a LightCycler^® 96 (Roche, Switzerland) with a mixture of 5 μL MonAmp™ ChemoHS qPCR Mix (Monad, Shanghai, China), 0.2 μL of each primer (10 μM), 2.5 ng of sample cDNA, and 3.6 μL of sterilised ultrapure H₂O. The reaction program was as follows: 10 min at 95°C, 40 cycles at 95°C for 10 s, 60°C for 10 s, and 72°C for 30 s. The results were analysed using LightCycler® 96 SW 1.1. Then, fluorescence was measured over a 55–95°C melting curve to detect a single gene-specific peak and to check the absence of primer dimer peaks; a single and discrete peak was detected for all primers tested. Negative controls consisted of non-template reactions in which the cDNA was replaced with H₂O.

The expression levels of RpedOBPs and RpedCSPs were calculated relative to the reference genes RpedGAPDH (R. pedestris glyceraldehyde-3-phosphate dehydrogenase) and RpedEF (R. pedestris elongation factor) using the Q-Gene method in Microsoft Excel-based software Visual Basic (Muller et al., 2002; Simon, 2003). For each sample, three biological replicates were performed with three technical replicates per biological replicate.

Based on sequence alignments, all phylogenetic trees in this study were constructed using the MEGA7 software (Kumar et al., 2016) using the neighbour-joining method, and each tree was tested by bootstrapping with 1,000 replicates. A phylogenetic tree was constructed based on the alignment results of cytochrome oxidase subunit I (COI) genes from different species (Aphis gossypii: KR017753.1, Nilaparvata lugens: AB325705.1, Sogatella furcifera: LC005703.1, Adelphocoris lineolatus: MZ608737.1, Adelphocoris suturalis: KY367052.1, Nysius ericae: KM022105.1, and R. pedestris: MG838422.1). The amino acid sequences of the RpedOBPs, RpedCSPs, and other hemipteran OBPs and CSPs were listed in Supplementary Table S2. The totals numbers of OBPs and CSPs in other insects have been reported in previous studies (Gu et al., 2011; Gu et al., 2013; Cao et al., 2014; He and He, 2014; Xue et al., 2014; Yang et al., 2014; Zhou et al., 2014; He et al., 2015; Zhou et al., 2015; Cui et al., 2017). Gene structure and exon/intron structure maps were generated using TBTools (version 1.098728) (Chen et al., 2020) based on R. pedestris annotated file (Gene Location Visualisation from GTF/GFF and Visualisation of Gene Structure). Pairwise similarity of sequences was also generated by TBTools based protein sequences (Protein Pairwise Similarity Matrix). Expression heat maps and bars were drawn using TBtools and GraphPad Prism 9.0, respectively, based on the qRT-PCR results.

The qRT-PCR data (mean ± SE) of RpedOBPs and RpedCSPs from various samples were subjected to one-way nested analysis of variance (ANOVA) followed by a least significant difference test (LSD) to compare means using the SPSS Statistics software (version 22.0; SPSS Inc., Chicago, IL, United States).

Recent progress in the whole-genome sequencing provides insights into the molecular mechanisms of olfaction in insects (Cheng et al., 2017; Wan et al., 2019). Second- and third-generation sequencing methods have also been successfully used for R. pedestris genome assembly. Based on BUSCO completeness and contig length, the Wtdbg2 assembly was used for the draft assembly of the R. pedestris genome (Huang et al., 2021). We downloaded the R. pedestris genome, protein, RNA and annotation files using Genome Warehouse and further annotated them using the Nr, Nt, SwissProt, KOG, eggNOG, Interpro, GO, and KEGG databases. A total of 53 OBPs and 25 CSPs of R. pedestris were identified and corrected using the following correction through BLAST. Finally, 49 OBPs and 25 CSPs were identified and named RpedOBP1-49 (43 were new genes), RpedCSP1-25 (17 were new genes) (Table 1). The predicted results of the sequences analysis revealed that all OBPs and CSPs had full-length open reading frames (ORF), and 39 OBPs and 23 CSPs had a signal peptide, respectively. The 49 OBPs without signal peptides share 22.5–99.19% amino acid identities with each other, while the 25 CSPs share 22.5–99.07% amino acid identities with each other (Supplementary Table S3). Full-length sequences of the 49 RpedOBPs and 25 RpedCSPs were presented in the supplementary files.

The number of RpedOBP and RpedCSP genes identified for R. pedestris is larger than those in other hemipterans (Figure 1). For example, 11 OBPs and 17 CSPs were identified in N. lugens (Xue et al., 2014; Yang et al., 2014; Zhang Y.-N. et al., 2016), 14 OBPs and 8 CSPs in A. lineolatus (Gu et al., 2011; Zhang Y.-N. et al., 2016), 16 OBPs and eight CSPs in A. suturalis (Cui et al., 2017), 28 OBPs and 16 CSPs in N. ericae (Zhang Y.-N. et al., 2016), nine OBPs and nine CSPs in A. gossypii (Gu et al., 2013; Zhang Y.-N. et al., 2016), and 12 OBPs and nine CSPs were identified in S. furcifera (He and He, 2014; Zhou et al., 2015). The differences in gene numbers may be explained by: 1) the different behaviours of different insects requiring distinct molecular mechanisms that have been developed over evolutionary time, and 2) the genomic data of R. pedestris that is more conducive to the comprehensive mining of OBP and CSP genes than other hemipteran species.

To clarify the position of OBPs and CSPs in chromosomes, we carried out chromosome location analysis of all genes, and the results showed that 48 OBP genes were distributed across four chromosomes (Figure 2A); only RpedOBP49 was located on scaffold056, which could not be mapped to a chromosome based on the current genome assembly. Thirty OBPs clustered together on chromosome 2, followed by chromosome 5 (11 OBPs), 3 (four OBPs), and 1 (three OBPs). Twenty-five CSP genes were distributed across four chromosomes, with 21 CSPs clustered together on chromosome 3 and the others on chromosome X (two CSPs), 1 (one CSP), and 5 (one CSP) (Figure 2B). More than 60% OBP and 80% CSP genes are located within clusters, as seen in other insects (Gong et al., 2009; Gu et al., 2013), indicating a relatively recent expansion of the OBP and CSP families of R. pedestris and the diverse functions of genes have evolved in response to different odorants in the environment.

To further clarify the genomic structural characteristics of OBPs and CSPs, we obtained the gene lengths and intron numbers of OBPs and CSPs based on the genome annotation file of R. pedestris (Figure 3). The lengths of the OBP genes ranged from 3.065 to 46.888 kb, with 33 OBPs having six introns and the other 16 OBPs having four, five, seven, eight, nine, and 12 introns, respectively (Figure 3A). The CSP genes were much shorter, ranging from 2.114 to 34.628 kb, with one, two, or three introns (Figure 3B). The phylogenetic trees of OBPs and CSPs in R. pedestris showed that the genes clustered together tended to have similar genomic structures, which also implies that they may have similar functions. The sequences of RpedOBPs were longer and had more introns than those of RpedCSPs, indicating that they may have complex features of functional differentiation.

Two phylogenetic trees were constructed for the OBPs and CSPs using protein sequences from R. pedestris, A. lineolatus, Apolygus lucorum, Aphis gossypii, and other hemipteran species (Gu et al., 2011; Gu et al., 2013; Cao et al., 2014; He and He, 2014; Xue et al., 2014; Yang et al., 2014; Zhou et al., 2014; He et al., 2015; Zhou et al., 2015; Cui et al., 2017). Similar to that in other studies (Gu et al., 2011; Zhang Y.-N. et al., 2016; Cui et al., 2017), the OBP tree in this study showed that eight RpedOBPs (OBP1, 5–9, 13, and 42) could be divided into the Plus-C OBP subfamily, and the other 41 RpedOBPs clustered into the classic OBP subfamily (Figure 4). In the constructed CSP tree, our results indicated that all 25 RpedCSPs were distributed along various branches, and each clustered with at least one other moth orthologue (Figure 5). The diversity of the RpedOBPs and RpedCSPs families suggests a role for positive selection in the rapid evolution and functional diversification of these genes. We speculate that both RpedOBP and RpedCSP genes had some gene expansions, such as OBP1/5/6/7/8/13, OBP23/10/48/16/33/32/31/49/46/45/44/47/43, OBP25/26/27/28, OBP40/39/41/34/15/36/35/11, CSP13/8/19/22, and CSP12/15/14/16/17/19/18/6/7, indicating that these genes may be involved in the recognition of important odorants related to R. pedestris behaviour (Pelosi et al., 2005; Matsuo et al., 2007; Gu et al., 2012; Poivet et al., 2013; Martin-Blazquez et al., 2017; He et al., 2019).

We used qRT-PCR to assess expression profiles of all R. pedestris OBPs and CSPs in the heads, thoraxes, abdomens, legs, wings, and antennae of the adults. The results showed that all OBPs and CSPs were expressed in the adult antennae of R. pedestris. Among the 49 RpedOBPs, 33 (approximately 67%) were significantly highly expressed in the antennae, including three male-biased (RpedOBP19, RpedOBP21, and RpedOBP32) and nine female-biased (RpedOBP2, RpedOBP6, RpedOBP9, RpedOBP17, RpedOBP24, RpedOBP34, RpedOBP36, RpedOBP48, and RpedOBP49). Among the 49 RpedOBPs, RpedOBP37 exhibited the highest expression level in male and female antennae (Figure 6). Compared to RpedOBPs, RpedCSPs were highly expressed in adult antennae as well as in non-antennal tissues. Of the 25 identified RpedCSP genes, only 11 RpedCSPs (approximately 44%) displayed antennal-biased expression; four RpedCSPs (RpedCSP3, RpedCSP12, RpedCSP20, and RpedCSP21) were male-biased and five RpedCSPs (RpedCSP4, RpedCSP9, RpedCSP11, RpedCSP13, and RpedCSP24) were female-biased in their expression (Figure 7). Several studies have shown that OBPs and CSPs are required for the correct recognition of some odorants from the external environment (Zhang et al., 2014; Liu et al., 2015; Chen G.-L. et al., 2018; Pelosi et al., 2018; Zhang et al., 2020a; Zhang et al., 2020b), therefore, we infer that the 33 RpedOBPs and 11 RpedCSPs highly expressed in adult antennae are likely to be involved in the crucial odorant reorganisation of R. pedestris (Krieger et al., 1996; Bohbot and Vogt, 2005; Zhang et al., 2014; Missbach et al., 2015; Chen G.-L. et al., 2018). The sex-biased RpedOBPs and RpedCSPs may be involved in the reorganisation of plant volatiles from oviposition sites or other sex-related odorants (He et al., 2010; Zhou et al., 2013; Zhang et al., 2019). Further analysis is needed to explore their exact roles, such as through fluorescence competitive binding assays (Liu et al., 2015; Ingham et al., 2020; Li et al., 2022), CRISPR/Cas9 mediated genome editing (Zhu G.-H. et al., 2016; Han et al., 2022), and gene mutations (Stowers and Logan, 2008; Zhang et al., 2020b).

Similar to the findings of previous studies (Zhang et al., 2013; McKenzie et al., 2014; Gu et al., 2015; Zhang L.-W. et al., 2016), we also found that there were 12 RpedOBP and six RpedCSP genes highly expressed in non-antennal tissues, including four leg-biased genes (RpedOBP14, RpedOBP35, RpedOBP44, and RpedCSP6), six head-biased genes (RpedOBP16, RpedOBP29, RpedOBP30, RpedOBP31, RpedOBP45, and RpedOBP46), one thorax-biased gene (RpedOBP39), two abdomen-biased genes (RpedOBP26 and RpedOBP28), and five wing-biased genes (RpedCSP1, RpedCSP2, RpedCSP8, RpedCSP10, and RpedCSP25), indicating that these genes may have other non-olfactory functions.