Background

Insect chemosensory system mainly includes the olfactory system and gustatory system, which are involved in complex behaviors, such as feeding, mating, ovipositing, and avoiding dangers and enemies [1,2,3,4]. The major chemosensory proteins include odorant-binding proteins (OBPs), chemosensory proteins (CSPs), odorant receptors (ORs), gustatory receptors (GRs), ionotropic receptors (IRs), and sensory membrane proteins (SNMPs) [5,6,7,8,9,10,11]. In the proposed process of odorant and taste detection, molecules are first bound and transported by OBPs or/and CSPs within the sensillum lymph, then detected by ORs, GRs or/and IRs expressed at the membrane of olfactory sensory neurons. After that, electrical signals are transmitted to the central nervous system through nerve axons inducing an action potential [12,13,14,15]. Other protein families, such as SNMPs, are also critical for pheromone detection [16, 17].

The first insect OBP was a pheromone-binding protein that was discovered from an antenna extract of the giant moth Antheraea polyphemus [5]. The structure of OBPs is further stabilized by three interlocked disulfide bridges between conserved cysteines [18,19,20]. The classical OBP fold possesses a core of six α-helices, a member of the C-plus class of OBPs, which has a longer sequence than classical OBPs, and C-minus OBPs have a shorter conserved cysteine [21,22,23,24].

The first identified member of the CSP family, called p10, was first identified in the American cockroach Periplaneta americana [25]. Later, a similar protein, OS-D (or A10) proteins, were identified by subtractive hybridization in the antennae of Drosophila melanogaster [18]. OBPs and CSPs, belong to a class of small water-soluble proteins and are impregnated in the sensilla lymph [26, 27]. CSPs, instead, seem to form a more homogeneous group of proteins, some with only four instead of six α-helical domains [28]. A large number of OBPs and CSPs have been identified in different insect species, particularly in recent years, due to genome projects and transcriptome sequencing [29,30,31,32,33].

ORs and GRs are ligand-gated cation channels, and encode seven transmembrane domain proteins with an intracellular N terminus and an extracellular C terminus [34, 35]. Insect ORs are composed of two related heptahelical subunits: a divergent conventional ligand-binding OR subunit that confers odor specificity [36, 37] and a highly conserved co-receptor (Orco) subunit [38]. Unlike ORs, GRs can function independently as monomers, such as the sugar receptor [39, 40] or form obligatory heteromers of two receptors as the CO2 receptor [41]. Insect IRs act as ligand-based ion channels that are related to ionotropic glutamate receptors [42]. Similar to ORs, the antennal IRs form heteromeric complexes with one or more co-receptors (IR8a, IR25a or IR76b) to perform their functions [43, 44].

The willow leaf beetle Plagiodera versicolora (Coleoptera: Chrysomelidae), is a leaf-eating forest pest, which mainly damages salicaceous trees, such as willows (Salix) and poplars (Populus) [45, 46]. In our previous study, we obtained 111 chemosensory genes from antennae [47] and forelegs transcriptomes [48], including 41 ORs, 28 OBPs, 11 CSPs, 17 GRs, 10 IRs, and 4 SNMPs. Antennae and maxillae are the paired principal chemosensory organs located on the head of the larvae [49]. Considering that the head is the tissue where most olfactory and gustatory sensilla are located, we identified more chemosensory genes from the larvae head transcriptomes of P. versicolora. Phylogenetic analysis was conducted among these chemosensory genes and other insect species. The quantitative real-time PCR was utilized in our study to screen the expression patterns. This work could provide a basis to facilitate functional clarification of these chemosensory genes at the molecular level.

Methods

Insect rearing and tissue collection

The larvae and adults of P. versicolora were collected from Sha Lake Park in Wuhan, Hubei Province, and maintained at the laboratory following conditions at 28 ± 1 °C, 70% ± 5% relative humidity, and a 12 h: 12 h light: dark (L: D) photoperiod. Larvae and adults were fed with a fresh leaf of willows. A total of 150 third instar larvae heads were collected and the tissues were immediately frozen in liquid nitrogen and stored at −80 °C for RNA extraction.

RNA extraction, cDNA Library Construction and Sequencing

The total RNA of the third instar larvae heads was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer’s instruction. RNA integrity was verified by gel electrophoresis. RNA concentration was measured on a Nanodrop ND-2000 spectrophotometer (NanoDrop products, Wilmington, DE, USA). Three biological replicates were used in the study. A total RNA of about 2 μg of each sample was used to construct the cDNA libraries. The libraries were sequenced using the Illumina Novaseq 6000 and performed at Shanghai Majorbio Bio-pharm Technology Co., Ltd. following the detailed protocol.

Transcriptome assembly and annotation

Datasets of clean reads were obtained by removing adaptor reads, ploy-N and low-quality reads, from the raw data and were further assembled into transcripts using Trinity v2.5.1 with default parameters [50]. Functional annotation of the longest transcript sequence (unigenes) was performed against six databases, non-redundant database (Nr), Pfam, Clusters of Orthologous Groups (COG), Swiss-Prot, Kyoto Encyclopedia of Genes and Genomes (KEGG), and Gene Ontology (GO), respectively.

Chemosensory Gene Identification and phylogenetic tree construction

To identify putative chemosensory genes (ORs, GRs, IRs, OBPs, CSPs, and SNMPs), we search the results of non-redundant protein (Nr) annotation from our transcriptome dataset. Annotations of all retrieved sequences were confirmed by blastx on the National Center for Biotechnology Information (NCBI). The Open reading frames (ORFs) of all putative chemosensory genes were predicted by ORF finder (https://www.ncbi.nlm.nih.gov/orffinder/). The transmembrane domains of ORs, IRs and GRs were predicted by TMHMM2.0 (https://services.healthtech.dtu.dk/service.php? TMHMM-2.0). Putative N-terminal signal peptides of OBPs and CSPs were predicted using SignalP 5.0 (https://services.healthtech.dtu.dk/service.php? SignalP-5.0).

Phylogenetic analyses were performed based on amino acid sequences from candidate chemosensory genes in P. versicolora and other insects. For the OBPs and CSPs tree sets, we selected the protein sequences including four Coleoptera from different families (Table S1), Colaphellus bowringi [51], Monochamus alternatus [52], Dastarcus helophoroides [52] and Ophraella communa [53]. For the GRs tree, the protein sequences from two Coleoptera (Chrysomelidae) species (C. bowringi and O. communa), the sequences that have been functionally identified, including three Lepidoptera (Plutella xylostella [54], Helicoverpa armigera [51, 55, 56] and Bombyx mori [35, 39, 57]) and four Diptera (D. melanogaster [40, 58,59,60], Anopheles gambiae [61], Aedes aegypti [41] and Anopheles coluzzii [62]) were used. ClustalW (gap opening penalty: 10, gap extension penalty: 0.2) was employed for amino acid sequence alignment and the neighbor-joining method was constructed using MEGA6, the gap/missing data treatment was set as partial deletion with a 95% site coverage cutoff [63, 64]. Branch support was assessed with 1000 bootstrap replicates. Phylogenetic trees were visualized and edited by FigTree 1.4.3(http://tree.bio.ed.ac.uk/software/figtree/).

Expression analysis by quantitative real-time PCR

Quantitative real-time PCR (qRT-PCR) was performed to verify the expression patterns of candidate chemosensory genes by using CFX Connect Real-Time System (Bio-Rad, CA, USA). Different stages, including eggs, first instar larvae, second instar larvae heads, third instar larvae heads and pupae, were collected. Total RNA was isolated using the methods described above and reverse transcribed into cDNA using HiScript® III RT SuperMix for qPCR (+gDNA wiper) (Vazyme, Nanjing, China). The qRT-PCR was conducted in a 20 μL reaction system, containing 10 μL ChamQ Universal SYBR qPCR Master Mix (2×), 0.4 μL each of the forward and reverse primers (10 μM), 1 μL cDNA template, and 8.2 μL of nuclease-free water. The reaction conditions were as follows: 95 °C for 30 s, 40 cycles of 95 °C for 5 s and 60 °C for 30 s. Gene expression profiles were analyzed using the 2 −ΔΔCT method [65]. The means and variances of the treatments were analyzed in a one-way ANOVA using SPSS 26.0 software. Each reaction for each tissue was performed in three biological replicates. The RPS18 gene (accession number: OM885975) was used as the reference gene for normalizing the expression of various samples [66]. And RPS18 gene has been proven stable in different development stages, female and male adults, different tissues, different temperatures, and pathogen treatment [67]. The gene-specific primers were designed using Beacon Designer 8.0 (amplicon length: 100 ± 20 bp; Tm: 60 ± 2 °C; GC%: 45–55%) (PRIMER Biosoft International, CA, USA), and were listed in Supplementary Table S2.

Results

Transcriptome sequencing, assembly and analysis

To identify the chemosensory genes of P. versicolora, the transcriptome sequencing of third instar larvae heads was sequenced with three independent biological replicates. Approximately 75.14 million, 68.24 million and 77.56 million raw reads, and a total of 74.50 million, 67.73 million and 76.97 million clean reads were generated. Clean reads were assembled into 28,268 unigenes with an average length of 1128 bp, and an N50 of 2500 bp. Length distribution analysis showed that 45.35% of all unigenes were longer than 500 bp in size (Fig. S1).

After function annotations, a total of 10,030 unigenes (35.74%) and 13,310 unigenes (47.43%) were successfully annotated through GO and Nr datasets, respectively. In the GO analysis, the unigenes were assigned to three main functions: biological process (13,448), cellular component (16,640), and molecular function (12,036). Within three classes, the subcategories “binding” (5321), “cell part” (4857) and “cellular process” (4475) contained the majority of the unigenes, respectively (Fig. S2).

Identification of candidate OBPs

In this study, a total of 29 predicted PverOBP transcripts were identified from third instar larvae heads transcriptomes, 12 PverOBPs were new compared to antennae and forelegs. The complete ORF was detected in 22 PverOBPs, with lengths ranging from 131 to 241 amino acids. Among them, 22 transcripts with the complete sequence have a signal peptide sequence (Table S3). Sequence analysis showed that 6 PverOBPs belonged to the Classical OBP subclass with the typical six conserved cysteine residues. Fifteen PverOBPs belonged to the minus-C OBP subgroup with 4 conserved cysteine residues (Fig. 1). A phylogenetic analysis showed that PverOBP4 and OBP12 were on the same branch as MaltOBP3 and OBP10, which functions were identified in M. alternatus, respectively (Fig. 2).

Fig. 1
figure 1

Multiple amino acid alignments of the predicted Classic OBPs and Minus-C OBPs. Conserved cysteine residues are marked with a red box (C2 and C5), black (C3, C4 and C6) and pink (C1) background. The number of 61 and 107 amino acids were deleted at the begin of OBP14 and end of OBP38, respectively

Full size image
Fig. 2
figure 2

Phylogenetic analysis of the OBPs (odorant-binding proteins) from four insect species: P. versicolora (Pver), C. bowringi (Cbow), M. alternatus (Malt), O. communa (Ocom). The P. versicolora genes are shown in blue. The values at the branch nodes represent bootstrap values based on 1000 replicates

Full size image

Identification of candidate CSPs

Six transcripts encoding CSPs were identified from the P. versicolora transcriptomes and PverCSP12 and CSP13 were newly compared to antennae and forelegs transcriptomes. All but one transcript (PverCSP13) included full-length ORFs, with lengths from 122 to 137 amino acids and a predicted signal peptide. Five CSP proteins have four highly conserved cysteine residues, which are characteristic of typical insect CSPs (Fig. 3). Phylogenetic analyses of CSPs from P. versicolora and other four Coleoptera species showed that PverCSP9 was clustered with MaltCSP5 in M. alternatus that functional characteristics have been performed (Fig. 4).

Fig. 3
figure 3

Multiple amino alignments of the predicted CSPs. Conserved cysteine (C1-C4) residues are marked with a blue box

Full size image
Fig. 4
figure 4

Phylogenetic analysis of the CSPs (chemosensory proteins) from four insect species: P. versicolora (Pver), C. bowringi (Cbow), M. alternatus (Malt), O. communa (Ocom) and D. helophoroides (Dhel). The P. versicolora genes are shown in blue. The values at the branch nodes represent bootstrap values based on 1000 replicates

Full size image

Identification of candidate GRs

Thirteen PverGR transcripts were obtained from P. versicolora third instar larvae heads and 7 PverGRs were newly identified. Among 13 PverGRs, only PverGR15 has a full-length ORF, with 7 transmembrane domains (TMDs), which are characteristic of typical insect GRs. Based on the phylogenetic tree analysis that PverGR15 and GR18 (newly identified) were clustered in carbon dioxide (CO2) receptors, PverGR3 and GR23 (newly identified) were clustered in sugar receptors (except fructose). In addition, PverGR9 and GR12 were clustered with the reported fructose receptors (Fig. 5).

Fig. 5
figure 5

Phylogenetic analysis of the GRs (gustatory receptors) from ten insect species: P. versicolora (Pver), C. bowringi (Cbow), O. communa (Ocom), A. gambiae (Agam), A. aegypti (Aage), A. coluzzii (Acol), D. melanogaster (Dmel), H. armigera (Harm), P. xylostella (Pxyl) and B. mori (Bmor). The P. versicolora genes are shown in blue. The values at the branch nodes represent bootstrap values based on 1000 replicates

Full size image

Identification of candidate ORs and IRs

From the assembled transcripts, 14 of them were annotated encoding ORs, and 5 PverORs were newly identified from the transcriptomes of third instar larvae heads. The Orco gene was also obtained in the transcriptomes (Table S4). None of these OR unigenes have full-length ORFs, encoding lengths from 45 to 289 amino acids. We identified 8 transcripts encoding candidate IRs in the P. versicolora transcriptome and 4 PverIRs were newly identified, compared to antennae and forelegs transcriptomes. These IRs were confirmed belong to IR family by the blast and phylogenetic analysis (Fig. S3). Of these, 3 PverIRs contained a putative full-length ORF, with three to four TMDs.

Specific expression profiles of candidate genes

The qRT-PCR was used to measure the specific expression levels of candidate chemosensory genes during the eggs, larvae and pupae stage (Fig. 6). The PCR efficiencies of primers ranged from 95–105% with high correlation coefficient (R2) values (0.991–0.999). PverOBP31, OBP34, OBP35, OBP38 and OBP40 were significantly higher expressed in 1st to 3rd instar larvae, while PverOBP33 and OBP37 were significantly higher expressed in pupae. PverCSP12 and CSP13 have significantly higher expression in 3rd instar larvae and eggs, respectively. PverGR22 has significantly higher expression in eggs, 3rd instar larvae and pupae, while PverGR24 has no significant difference in the tested tissues. PverOR42 was significantly higher expressed in pupae, moderately in 2nd and 3rd instar larvae, while the expression level of PverOR41 was no significant difference among eggs, larvae and pupae. Both PverIR12 and IR14 were no significant difference among eggs, larvae and pupae.

Fig. 6
figure 6

Expression levels of chemosensory genes in different stages performed by qRT-PCR. E, eggs; 1st, first instar larvae; 2nd, second instar larvae heads; 3rd, third instar larvae heads; P, pupae. Error bars indicate standard error of three biological replicates

Full size image

Discussion

Compared to Dipterans and Lepidopterans, the molecular basis of chemoreception in Coleopterans is relatively poorly understood. In the current study, we sequenced and analyzed the transcriptome of the third instar larvae heads from P. versicolora. In total, 74 chemosensory genes were identified, including 29 OBPs, 6 CSPs, 14 ORs, 13 GRs, 8 IRs and 4 SNMPs. Compared to our previous antennae and forelegs transcriptome data in adults, 12 OBPs, 2 CSPs, 5 ORs, 4 IRs and 7 GRs were newly identified (Fig. S4). Moreover, the expression profile of 16 chemosensory genes in different tissues was validated by qRT-PCR, which would facilitate the exploration of the function of these genes. Systematic research on chemosensory genes in larvae may provide valuable information on understanding the molecular mechanisms of insect chemosensory systems.

The phylogenetic tree of OBPs and CSPs of P. versicolora was constructed using various Coleoptera species of C. bowringi, M. alternatus and O. communa. PverOBP4, PverOBP12 and PverCSP9 were clustered with MaltOBP3 (bound the beetle- or host-plant-related compounds) [68], MaltOBP10 (bound the volatiles from pine bark) [69] and MaltCSP5 (bound the odor molecules of pine volatiles) [70], respectively, suggesting that PverOBPs and PverCSPs may have functions (detect the host-plant-related volatiles) similar to these of OBPs and CSPs. Notably, PverOBP26 (identified in forelegs transcriptome) and five PverOBPs (PverOBP29, 30, 32, 35 and 39) formed a cluster with CbowOBP24, indicating that these genes of the clade may have suffered gene duplication.

The phylogenetic tree of GRs of P. versicolora was constructed using different species, including Coleoptera, Lepidoptera and Diptera. Insect GRs usually have three CO2 receptors (GR1, GR2 and GR3) family [61, 62, 71], and a new CO2 receptor (PverGR18), belonging to GR2 family, was identified from our larvae transcriptome. Considering three CO2 receptors have been identified, the function of these putative CO2 receptors genes could elucidate in the future. BmGr10 plays an important role in the myo-inositol recognition required for B. mori larval feeding behavior [57]. PverGR9 and GR12 were homologues to BmGr10, indicating three PverGRs were expressed in larval heads may guide several chemosensory behaviors. OR41 and OR52 were obtained from larval antennae and maxillae transcriptome in H. armigera, and responded to a four-component blend that strongly attracted larvae [72]. It meant that some ORs were identified in larval heads play role in chemosensory-guided behaviors. However, no PverORs were homologues to HarmOR41/OR52 (Fig. S5), as the function of 5 PverORs need to identify in the future.

OBPs are commonly postulated to play an important role in the perception of environmental volatiles, in both adult and larval stages [73, 74]. PverOBP31 and OBP34 have a highly expression level in the third instars larval heads, and PverOBP35 showed increased expression levels from the first to third instars. Usually, higher instar larvae of agriculture and forestry pests are crucial for crop and tree feeding [45, 75]. The studies have indicated that PverOBP31, PverOBP34 and PverOBP35 may participate in larval food searching.

Some pieces of evidence inciting that a relatively simple system, OBPs, abundantly expressed in the larval antennae, also involves the detection of pheromones in larvae [76, 77]. In vitro expression and ligand-binding of SexiOBP13 (highly expressed in the larval heads, and showed increased expression levels from the first- to third-instars), the result showed that it displayed a high binding affinity to Z9, E12–14:Ac, the major sex pheromone component of S. exigua, while low affinities to the tested host plant volatiles [78]. In general, PBPs have significantly higher expression in the antennae of adults. This report on S. exigua provides a new idea for identifying the function of PBPs in P. versicolora that may have a significantly higher expression level (PverOBP31, 34, 35, 38 and 40) in larvae. In addition, these five PverOBPs may detection of any other compounds, such as host or non-host volatiles.

Two special PverOBPs (OBP33 and 37) were highly expressed in pupae which were also consistent with the results obtained in other species. For example, five OBPs in Procecidochares utilis, OBP44a in Bactrocera dorsalis, and three OBPs (OBP15, 17, and 25) in Galeruca daurica were found to be abundantly expressed in the pupal stage, respectively [79,80,81]. In addition, PverOR42 was highly expressed in pupae, the same as three ORs in P. utilis and eight ORs in Chlorops oryzae [79, 82]. These results suggest that these OBPs and ORs in P. versicolora may be related to the beginning of the development of chemosensory tissue during pupation. On the other side, maybe it is also likely that these ORs and OBPs are a consequence of development.

Different expression profiles during development indicate that these CSP genes might have different functions. PverCSP12 has a high expression level in the third instar larva stage, which might suggest a role of these proteins in food searching and feeding. Particularly interestingly, PverCSP13 was highly expressed in the eggs. It has been demonstrated that CSP5 in the eggs of honeybees is required for the correct development of the embryo [83, 84]. It suggested that PverCSP13 may involve in the development of the eggs.