The transcriptomes of female antennae (FA) and male antennae (MA) of E. grisescens were sequenced with two independent biological replicates. About 45.72 (FA-1), 46.51 (FA-2), 42.15 (MA-1), and 49.97 (MA-2) million raw reads were obtained for each transcriptome. The datasets of transcriptomes during the current study are available in the NCBI SRA database (http://trace.ncbi.nlm.nih.gov/Traces/sra/, accession numbers: SRR6004297–SRR6004301). After filtering and assembling, 114,595 transcripts were generated with an N₅₀ length of 1,715 nt (Table 1). For annotation, we combined the female- and male-antennal transcriptomes of E. grisescens and searched against the non-redundant (NR) database by BLASTX with a cut-off e-value of 10⁻⁵. The best match percentage (40.2%) was with Bombyx mori sequences, followed by sequences from Plutella xylostella (16.7%), Danaus plexippus (16.6%), Acyrthosiphon pisum (1.5%), and Papilio xuthus (1.4%) (Figure 1A). Gene ontology (GO) annotation was used to classify the transcripts into functional categories.

Among the distinct transcripts, 32,710 (28.54%) corresponded to at least one GO term.

Within the molecular function category, most transcripts were involved in binding (e.g., nucleotide-, ion-, and odorant-binding) and catalytic activity (e.g., hydrolase and oxidoreductase) (Figure 1B).

Sequence annotation led to the identification of 40 candidate EgriOBPs in the E. grisescens antennae transcriptome (File S1). Sequence analyses showed that 34 of them were full-length EgriOBP genes, and 31 had a predicted signal peptide (Figure 2A). EgriOBP13, 14, 15, 32, 37, and 39 contained four conserved cysteines but lacked the conserved cysteines C2 and C5. EgriOBP4, 5, 6, 7, 28, 35, and 40 contained more than six conserved cysteines. EgriOBP8 contained five conserved cysteines but lacked the conserved cysteine C2. All of the E. grisescens classic OBPs present the C-pattern common to Lepidoptera, C1-X₂₅₋₃₀-C2-X3-C3-X₃₆₋₄₂-C4-X₈₋₁₄-C5-X₈-C6. A total of 30 CSP genes were identified in E. grisescens antennae, 26 of which contained a full-length open reading frame (ORF), a signal peptide, and four conserved cysteine residues (Figure 2B). All of the E. grisescens CSPs have the C-pattern of Lepidoptera, C1-X₆₋₈-C2-X₁₈-C3-X₂-C4. By homology analysis, we identified 59 candidate EgriORs in E. grisescens antennae. Sequence analyses revealed that 45 (EgriOR1–44 and EgriORco) of the 59 sequences had an intact ORF with characteristics typical of insect OR genes (full-length ORFs of about 1,200 bp). In total, we identified 24 EgriIRs from E. grisescens antennal transcriptomes, and 16 of them contained intact ORFs.

To further investigate the function of E. grisescens OBP/CSP/OR/ IR genes, phylogenetic trees were constructed using sequences of typical OBP/CSP/OR/IRs from other insect species for which the whole genome or transcriptome data were available. In the resulting phylogenetic tree, we observed four well-supported groups; PBP, GOBP, Plus-C OBP, and Minus-C OBP (Figure 3). Three EgriOBPs (EgriOBP2, 3, and 25) were grouped in the PBP clade with another Lepidoptera PBP. The orthologs EgriOBP1 and 29 were in the GOBP group. EgriOBP4, 5, 6, 7, 28, 35, and 40 were distributed in the Plus-C OBP group, and EgriOBP8, 13, 14, 15, 32, 37, and 39 were distributed in the Minus-C OBP group. In the CSP phylogenetic tree, EgriCSP8 was grouped into the same clade as HarmCSP6 (Figure 4).

The EgriORs were distributed among the orthologous groups in the OR phylogenetic tree (Figure 5). EgriORco was strongly associated with ObruORco, HvirORco, and BmorOR2 with high bootstrap support. EgriOR25, 28, and ObruOR1 were grouped with B. mori, H. armigera, Helicoverpa assulta, and Heliothis virescens PRs, which are known to be receptors for Type-I pheromones. EgriOR24, 31, 37, and 44 were independently grouped without any orthologs of other Lepidoptera insects. In the IR phylogenetic tree (Figure 6), 11 EgriIRs (Egri3, 6, 8, 11, 12, 13, 14, 16, 18, and 24) were clustered with the presumed “antennal” orthologs IR64a, IR21a, IR31a, IR68a, IR75d, IR76b, IR93a, IR60a, and IR40a. Two EgriIRs, EgriIR10, and EgriIR21, were respectively distributed in the IR8a and IR25a groups, which are co-receptors. EgriIR1 was grouped with NMDA iGluRs (N-methyl-d-aspartate ionotropic receptors).

We further characterized the tissue expression pattern and abundance of E. grisescens OBP/CSP/OR/IR genes in the antennae by qPCR and by evaluating the RPKM (reads per kilobase per million mapped reads) values. The qPCR results indicated that 24 EgriOBPs were uniquely or more strongly expressed in female and male antennae, except for EgriOBP4, 7, 8, 15, 21, 22, 23, 24, 27, 28, 35, 37, 39, and 40 (Figure 7A). Of these 24 EgriOBPs showing antenna-biased expression, EgriOBP2, 3, 9, 12, and 25 showed significantly higher transcript levels in male antennae than in female antennae (9.8-, 10.3-, 9.0-, 7.8-, and 12.8-fold higher RPKM values, respectively, in male antennae than in female ones). Of the five male antenna-biased EgriOBPs, EgriOBP2 and 3 were much more abundant in the antennae transcriptome. EgriOBP7, 13, 21, and 33 transcripts were abundant mainly in female and male proboscises. Compared with EgriOBPs, EoblCSPs showed wider and more diverse expression patterns (Figure 7B). The transcript levels of EgriCSP5, 8, 13, 14, 15, 16, and 17 were markedly higher in the E. grisescens antennae transcriptomes than in the transcriptomes of other tissues. Of these seven EgriCSPs that were abundant in the antennae, EgriCSP8 was expressed at higher levels in male antennae than in female ones. The abundance of EgriCSP21 and 25 transcripts was markedly higher in female antennae than in male ones. The abundance levels of EgriCSP5, 15, and 17 were also higher in female antennae than in male ones, but the differences were not significant.

Analyses of the expression profile of EgriORs showed that these EgriOR genes were uniquely or more strongly expressed in antennae than in other tissues (Figure 8A). Among the ORs, including EgriORco, EgriOR28 and 37 showed the highest expression levels in antennae. Five EgriORs (EgriOR24, 28, 37, 44) were predominantly expressed in male antennae, with RPKM values in male antennae being 51.6-, 29.6-, 20.9-, and 72.7-fold that of their respective counterparts in female antennae. EgriOR31 was uniquely expressed in male antennae. Analyses of the expression patterns of EgriIR genes indicated that EgriIR8, 10, 11, 12, 16, 21, and 24 were highly expressed in the antennae (Figure 8B). Of these eight EgriIRs, EgriIR10, 21, and 24 showed relatively high RPKM values in female and male transcriptomes.

The Xenopus oocytes and two-electrode voltage clamping recording system were used to characterize the function of the EgriOR1 and 31, by co-expressing with the corresponding receptor EgriORco. The results showed that oocytes co-expressing EgriOR31 and EgriORco robustly responded to the triene Z3Z9-6,7-epo-18:Hy, but less so to Z3Z6Z9-18:Hy. The oocytes co-expressing EgriOR1 and EgriORco showed no responses (Figure 9).

Individuals of E. grisescens were originally collected from the Experimental Tea Plantation of the Tea Research Institute, Chinese Academy of Agricultural Sciences (Hangzhou, Zhejiang, China). Experimental insects were reared on fresh tea shoots in enclosed nylon mesh cages (70 × 70 × 70 cm) and kept in a climate-controlled room at 25 ± 1°C and 70 ± 5% relative humidity under a 14-h light:10-h dark photoperiod. After pupation, male and female pupae were kept separately in cages for eclosion. After emergence, adult moths were supplied with 10% honey solution. For transcriptome sequencing, 100 female and 100 male antennae were collected from 2-day-old virgin insects, with two replicates. Female and male antennae, heads, thoraxes, thoraxes, legs, wings, proboscises, and pheromone glands were collected from 3-day-old virgin insects for qRT-PCR analyses. These tissues were immediately frozen and stored at −80°C until RNA isolation.

Total RNA was extracted from female and male antennae using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). Quality of the RNA were assessed by agarose gel electrophoresis, Nanodrop (Thermo), and Agilent 2100. The sampling quality satisfy the requirements of cDNA libraries construction. The cDNA library construction and Illumina sequencing were performed at Novogene Bioinformatics Technology Co., Ltd. (Beijing, China). Briefly, mRNAs were isolated from 5 μg pooled total RNA using oligo (dT) magnetic beads and fragmented into short fragments in the presence of divalent cations in fragmentation buffer at 94°C for 5 min. Using these short fragments as templates, random hexamer primers were used to synthesize first-strand cDNA. Second-strand cDNA was generated using RNase H, and DNA polymerase I. After end repair and adaptor ligation, short sequences were amplified by PCR and purified with a QIAquick® PCR purification kit (Qiagen, Venlo, The Netherlands), and sequenced on the HiSeq™ 2500 platform (San Diego, CA, USA).

Transcriptome de novo assembly was carried out with the short-read assembly program Trinity (r20140413p1) (Grabherr et al., 2011) based on the paired-end reads with default settings. Transcripts longer than 150 bp were first aligned by BLASTX to protein databases (NR, Swiss-Prot, KEGG, and COG; e-value < 10⁻⁵) to retrieve proteins with the highest sequence similarity to the unigenes along with their protein functional annotations. We then used the Blast2GO (Conesa et al., 2005) for GO annotation of the transcripts and WEGO software (Ye et al., 2006) to plot the GO annotation results.

Transcript abundance was calculated by the RPKM (reads per kilobase per million mapped reads) method (Mortazavi et al., 2008), which can eliminate the influence of different transcript lengths and sequencing discrepancies when calculating abundance (Mortazavi et al., 2008). We used the following equation: RPKM(A)=C×10N×L1036 where RPKM (A) is the RPKM value of transcript A; C is the number of reads uniquely aligned to transcript A; N is the total number of fragments uniquely aligned to all transcripts; and L is the number of bases in transcript A.

Genes showing differential expression between two conditions/groups were detected using the DESeq R package (1.10.1), which provides statistical routines to determine differential expression from digital gene expression data using a model based on negative binomial distribution. The resulting P values were adjusted using Benjamini and Hochberg's approach to control the false discovery rate. Genes with an adjusted P < 0.05 found by DESeq were designated as being differentially expressed.

Amino acid sequence alignments of the candidate 40 EgriOBPs and 30 EgriCSPs were performed using ClustalX 2.0 (Larkin et al., 2007), and visualized using Jalview 2.4.0 b2 (Waterhouse et al., 2009). The signal peptides of EgriOBPs and EgriCSps were predicted by SignalP 4.1 (http://www.cbs.dtu.dk/services/SignalP/). To investigate the phylogenetic relationships of OBPs, CSPs, ORs, and IRs between E. grisescens and other typical insect species, we compared them using MAFFT (E-INS-I parameter) (Katoh and Standley, 2013). Phylogenetic trees were constructed using PhyML 3.1 with LG substitution model was used to construct a maximum likelihood phylogenetic tree using Bayesian analysis (Guindon et al., 2010).

The tissue expression patterns of 40 OBPs, 30 CSPs, 59 ORs, and 24 IRs in different tissues were measured by a qPCR method conducted according to the minimum information for publication of quantitative Real-Time PCR Experiments (Bustin et al., 2009). Total RNA was isolated using the SV Total Isolation System (Promega, Madison, WI, USA) according to the manufacturer's instructions. Quality of the RNA were assessed by agarose gel electrophoresis, Nanodrop (Thermo). Single-stranded cDNA templates were synthesized using the Reverse Transcription System (Promega) as per the manufacturer's instructions. The qRT-PCRs were performed on a Bio-Rad CFX96 touch real-time PCR detection system (Bio-Rad, Hercules, CA, USA). The primers were designed using Beacon Designer 7.7 based on the E. grisescens nucleotide sequences obtained from the transcriptome data (File S2). Templates diluted into five-fold series were used to construct a relative standard curve to determine the PCR efficiencies and for quantification analysis. Each reaction was run in triplicate (technical repeat). Two reference genes, guanine nucleotide-binding protein G(q) subunit alpha and glyceraldehyde-3-phosphate dehydrogenase (sequences were provided in File S3), were selected in qPCR analysis. A blank control without template cDNA (replacing cDNA with H₂O) served as the negative control. Each reaction included three independent biological replicates and was repeated three times (technical replicates). Relative transcript levels were calculated using the comparative 2^−ΔΔCq method. The level of each tested mRNA was determined using SYBR® Premix Ex Taq™ II (TaKaRa, Dalian, China) according to the manufacturer's instructions.

A Xenopus oocytes expression system was used to express the EgriOR1 and 31. EgriOR1 and 31 were amplified using specific primers (File S2). The purified PCR products were ligated into pT7Ts vector using an In-Fusion® HD Cloning Kit (Clontech, USA) following manufacturer's instructions. The cRNAs of EgriORs were synthesized using mMESSAGE Mmachine T7 kit (Ambion, Austin, TX). Electrophysiological recording was performed according to previously reported protocols (Wang et al., 2010). Mature healthy oocytes (stage V–VII) were treated with 2 mg/ml collagenase I(GIBCO, Carlsbad, CA) in washing buffer [96 mM NaCl, 2 mM KCl, 5 mM MgCl₂, and 5 mM HEPES (pH = 7.6)] for about 1.5 h at room temperature. And then, oocytes were microinjected with 27.6 ng EgriOR cRNAs and 27.6 ng EgriORco. After injection, oocytes were incubated for 4–7 days at 18°C in 1 × Ringer's solution [96 mM NaCl, 2 mM KCl, 5 mM MgCl₂, 0.8 mM CaCl₂, and 5 mM HEPES (pH = 7.6)] supplemented with 5% dialyzed horse serum, 50 mg/ml tetracycline, 100 mg/ml streptomycin, and 550 mg/ml sodium pyruvate. Whole-cell currents were recorded from the injected Xenopus oocytes with a two-electrode voltage clamp and recorded with an OC-725C oocyte clamp (Warner Instruments, Hamden, CT, USA). Stock solutions of tested compounds were prepared using ethyl alcohol, which were diluted to the indicated concentrations by 1 × Ringer's buffer before use. Oocytes were exposed to1 × 10⁻⁵ M of sex pheromone compounds and ethyl alcohol. Oocytes without cRNA injection were set as negative control. Data acquisition and analyses were carried out with Digidata 1440A and pCLAMP 10.2 software (Axon Instruments Inc., Union City, CA).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.