Identification of Candidate Olfactory Genes in Scolytus schevyrewi Based on Transcriptomic Analysis

The bark beetle, Scolytus schevyrewi (S. schevyrewi), is an economically important pest in China that causes serious damage to the fruit industry, particularly, in Xinjiang Province. Chemical signals play an important role in the behavior of most insects, accordingly, ecofriendly traps can be used to monitor and control the target pests in agriculture. In order to lay a foundation for future research on chemical communication mechanisms at the molecular level, we generate antennal transcriptome databases for male and female S. schevyrewi using RNA sequencing (RNA-seq) analysis. By assembling and analyzing the adult male and female antennal transcriptomes, we identified 47 odorant receptors (ORs), 22 ionotropic receptors (IRs), 22 odorant-binding proteins (OBPs), and 11 chemosensory proteins (CSPs). Furthermore, expression levels of all the candidate OBPs and CSPs were validated in different tissues of male and female adults by semiquantitative reverse transcription PCR (RT-PCR). ScosOBP2 and ScosOBP18 were highly expressed in female antennae. ScosCSP2, ScosCSP3, and ScosCSP5 were specifically expressed in the antennae of both males and females. These results provide new potential molecular targets to inform and improve future management strategies of S. schevyrewi.


INTRODUCTION
Olfaction serves to detect environmental chemical information necessary for insect behavior such as finding food sources, mates, and oviposition sites (Hanson, 1999;Clyne et al., 2000). Insects have a sophisticated olfactory system that begins with the reception of odorants at the peripheral chemosensory system. Insect olfaction is dependent on olfactory receptor neurons (ORNs) in sensilla (Leal, 2013) distributed mainly in antennae and also in maxillary palps or labial palps (Stoker et al., 1990). The research of molecular mechanisms of olfactory reception in insects has predominantly been in the model organism Drosophila melanogaster. These studies have shown diverse olfactory genes encoding proteins, such as odorant receptors (ORs), ionotropic receptors (IRs), odorant-binding proteins (OBPs), and chemosensory proteins (CSPs), involved in different chemical signal transduction processes (Benton et al., 2009;Wilson, 2013;Xiao et al., 2019).
Odorant receptors play a critical role in recognizing thousands of odorant molecules in the insect olfactory system. Insect ORs were first identified in Drosophila which has the characteristic feature of a seven-transmembrane domain (TMD) structure that is unrelated to the ORs in vertebrates (Clyne et al., 1999;Benton et al., 2006). Every ORN can express a single or two OR genes (Vosshall and Hanson, 2011). Specificity of OR relies on the ligand-banding ORs (Dobritsa et al., 2003;Elmore et al., 2003;Hallem et al., 2004), while Orco functions as an obligatory chaperon for the Orco-OR complex (Larsson et al., 2004;Benton et al., 2006;Stengl, 2017).
Evolved from the ionotropic glutamate receptor superfamily, IRs have been shown to be involved in odor reception. They are expressed in the sensory neurons that respond to many distinct odors, such as acids, amines, and other chemicals that cannot be recognized by ORs (Benton et al., 2006). Aside from olfaction, IRs serve various functions, such as cool sensation (Ni et al., 2015), hygrosensation (Knecht et al., 2016), circadian clock (Chen et al., 2015), and detection of carbon dioxide (CO 2 ) (Breugel et al., 2018).
In addition to ORs and IRs, other multigene families encode proteins that also play critical roles in olfaction. OBPs are small soluble proteins secreted in the sensillar lymph. They are characterized by an N-terminal signal peptide sequence and a set of six conserved cysteine residues that form three disulfide bridges (Pelosi et al., 2005(Pelosi et al., , 2006. Studies of defective mutants and wild-type counterparts of OBP76a (also known as LUSH) in Drosophila have shown that this protein has a key role in the perception of alcohol and 11-cis vaccenyl acetate (Kim et al., 1998;Xu et al., 2005;Gomez-Diaz et al., 2013). OBPs have also been reported as a pheromone-binding protein in Lepidoptera (Jing et al., 2019). Some OBPs operate similar to LUSH in response to pheromones. In vivo studies have shown that OBPs significantly affect pheromone perception in moths. Knocking out these OBPs significantly reduced electrophysiological responses to pheromones in several species, such as Helicoverpa armigera , Spodoptera litura (Liu et al., 2012;Zhu et al., 2016a), and Chilio suppressalis .
Chemosensory proteins are also small soluble proteins but are shorter in amino acid sequence length than that of OBPs, and CSPs share the same structure of having four conserved cysteines forming two disulfide bridges (Pelosi et al., 2005(Pelosi et al., , 2006Honson et al., 2015). As semiochemical carriers, some CSPs are involved in chemodetection (Pelosi et al., 2018;Li et al., 2021) because CSPs are abundant in the lymph of chemosensory hairs (Angeli et al., 1999;Jacquin-Joly et al., 2001;Monteforti et al., 2002;Sun et al., 2014). Some of them already have been reported to function such as OBPs, e.g., CSP3 of the honeybee, which specially binds some components of brood pheromone (Briand et al., 2002).
Bark beetles (Coleoptera; Curculionidae; Scolytinae) feed on woods and several of them pose serious threats to forestry, e.g., Ips typographus (Wermelinger, 2004), Dendroctonus ponderosae (Andersson et al., 2013). Since their host-finding relies on chemical communication, e.g., aggregation behavior based on male-produced pheromone (Schlyter et al., 1987), pheromonebased technique could be used for the detection and control of this pest. In order to develop this technique efficiently, one way is to exploit olfactory genes that are critical for successful mate and host finding. Transcriptomic and genomics studies have been performed for searching olfactory genes in bark beetles (Andersson et al., 2013Mitchell et al., 2019), and functional studies were limited to only seven ORs Yuvaraj et al., 2021). Scolytus schevyrewi (S. schevyrewi) (Cleoptera: Scolylidae) is one of the most destructive insect pests of fruit trees in China. It has a wide host range and has been reported to attack several families of trees in Xinjiang province (Li et al., 1995). Several studies have focused on the identification and field bioassay of chemical attractants in the bark beetle (Fan et al., 2014). In order to provide a molecular basis for gene targets for putative chemical lures of this pest, we performed Illumina Hiseq 2000 sequencing of the transcriptome of adult male and female antennae samples.

Insect Rearing and Tissue Collection
Scolytus schevyrewi larvae were reared on the branches of their host plants (Armeniaca vulgaris) collected from Baren County, Xinjiang province, China (39.0 • N, 75.8 • E) and maintained in the lab under the following conditions of 26.5 • C, a cycle of 14-h light:10-h dark, and 65% relative humidity. Pupae were placed on a branch and the emerged adults were collected every day. Twoday-old adults were used to collect male and female antennae, heads (without antennae), thorax, abdomen, legs, and wings using the fine-tip forceps, immediately frozen in liquid nitrogen and stored at−80 • C until RNA isolation.

RNA Extraction
Total RNA from different tissues of S. schevyrewi was obtained using TRIzol reagent (Invitrogen, Carlsbad, California, USA) following the instruction of manufacturer. The total RNA from each pair of antennae, legs, and wings were separately obtained from each adult, totaling 300 males and 300 females. Heads (without antennae), thoraxes, and abdomens were separately collected from 20 to 30 adult males and 20 to 30 females. Total RNA was dissolved in RNase-free water, and RNA integrity was verified by gel electrophoresis. RNA concentration and purity were determined on the Nanodrop ND-2000 spectrophotometer (NanoDrop products, Wilmington, DE, USA).

cDNA Library Construction and Sequencing
A total of 1 µg of total RNA of each sample of male and female antennae were used to construct two separate cDNA libraries, one for each sex. Paired-end reads of 100 bp were sequenced using the Illumina HiSeq 2000 platform to obtain library-sequencing information at Beijing Genome Institute (Shenzhen, China). The detailed protocols for cDNA library construction and sequencing applied have been described in the previous studies Zhang et al., 2015). The raw data were uploaded to the NCBI SRA database (Accession: PRJNA732801, https://www. ncbi.nlm.nih.gov/sra).

Assembly
Low-quality reads were filtered out, low-quality nucleotides at each end were trimmed, and 3 ′ adaptors and poly-A/T were removed from the raw reads to generate the clean reads. Subsequently, the clean reads were used to form a de novo assembly using the Trinity platform (v2.1.0) with default parameters (Grabherr et al., 2011). The Trinity outputs were then clustered by TIGR gene indices clustering tools (TGICL) to generate the final unigene dataset (Pertea et al., 2003).

Identification of Olfactory Genes
Unigenes were annotated using blastx against NCBI nonredundant (nr) sequences with e <1e −5 . The blast results were then imported into the Blast2Go (version 3.1) with default parameters (Conesa et al., 2005). OR, IR, OBP, and CSP genes of the candidates were selected according to the nr sequence annotation results in the remote server from the lab. All candidate olfactory genes were manually checked using the blastx program against the nr sequence database. The open-reading frames (ORFs) of the putative olfactory genes were predicted using the ExPASy (Expert Protein Analysis System) translate tool (https://web.expasy.org/translate/). The TMDs of ORs and IRs were predicted using TMHMM server version 2.0 (http:// www.cbs.dtu.dk/services/TMHMM). Putative N-terminal signal peptides of OBPs and CSPs were predicted using the SignalP 4.0 server (http://www.cbs.dtu.dk/services/SignalP-4.0/) with default parameters.

Phylogenetic Analysis
Olfactory genes from S. schevyrewi, Ips typographus, Dendroctonus ponderosae (Andersson et al., 2013), and Holotrichia parallela (Yi et al., 2018) were selected for the phylogenetic analysis. Sequence information was listed in Supplementary Table 2. Amino acid sequences were aligned by MAFFT (https://www.ebi.ac.uk/Tools/msa/mafft/). Phylogenetic trees of olfactory genes were constructed using RAxML version 8 with the Jones-Taylor-Thornton amino acid substitution model. Node support was assessed using a bootstrap method based on   Figure 2).

Transcriptome Assembly
The transcriptomes of male and female S. schevyrewi antennae were separately sequenced by the Illumina HiSeq 2000 platform.

Gene Identification and Functional Annotation
The functional annotations of the unigenes were performed mainly based on the blastx results against the nr sequence database. We matched 22,815 (66.9%) unigenes to known proteins by blastx. Among those annotated genes, 16,725 (73.3%) unigenes showed strong homology (e-values lower than 1e −45 ), while 6,090 (26.7%) unigenes showed poor matches with e-values between 1e −15 and 1e −5 . The similarity analysis showed that 11,514 (50.5%) unigenes had more than 60% similarity with known proteins. Most of the annotated unigenes were matched to Tribolium castaneum (67.3%), followed by D. ponderosae (13.7%) and others species (19.0%). Gene ontology (GO) annotations of the entire set of unigenes were performed using the Blast2GO pipeline based on the blastx searches against nr sequences. A total of 12,720 unigenes were assigned various GO terms. In the molecular function category, genes involved in the binding activity and catalytic activity were most abundant. In the cellular component category, genes involved in cell, cell part, macromolecular, membrane, organelle, and organelle part were enriched. In the biological process category, genes involved in the cellular process, metabolic process and single-organism process were the most represented.

Identification of Candidate Odorant Receptors
The candidate ORs were identified by keyword search of the blastx annotations. We identified 47 putative OR genes. Thirteen of them were full-length putative OR genes ranging from 1,100 to 1,400 bp with complete ORFs and 5 to 7 TMDs, which are characteristics of typical insect ORs. This includes the full-length ScosOrco gene encoding 488 amino acids. Seven of the predicted incomplete ORs were shorter in length and contained a deduced protein longer than 300 amino acids. Four of the predicted incomplete ORs were even shorter than 200 amino acids. The blastx results indicated that the identities of the most predicted ORs shared with known insect ORs were very low, ranging from 24 to 49%. Nine predicted ORs (ScosOR1, ScosOR27, ScosOR7, ScosOR38, ScosOR39, ScosOR2, ScosOR8, ScosOR9, and ScosOR34) had greater identity (52-62%) with the OR from D. ponderosae. ScosOrco had 88% identity with the Orco from Rhynchophorus ferrugineus. Phylogenetic analysis was performed with ORs from D. ponderosae, I. typographus, H. parallela, and S. schevyrewi (Figure 1). A branch for Orco was identified in the phylogenetic tree. Two expanded branches in this species relative to others in the comparison were also identified. One branch consisted of ScosOR5, ScosOR6, ScosOR10, ScosOR11, ScosOR25, and ScosOR28 and the other consisted of ScosOR17, ScosOR18, Scos22, Scos31, Scos32, Scos37, Scos40, and Scos45. Most of the branches in the tree were supported by high-local support values and few branches were not reliable.
Information on unigene reference, length, and best blastx hit of all 47 ORs are listed in Table 1.

Identification of Candidate Ionotropic Receptors
Bioinformatics analysis identified 22 putative IRs in the S. schevyrewi transcriptome. Only ScosIR76b was a full-length sequence with 555 amino acids and five TMDs; the other IRs were incomplete due to the lack of the 5 ′ or 3 ′ terminus.
Information on unigene reference, length, and best blastx hit of all 22 OBPs are listed in Table 3.

Identification of Putative Chemosensory-Binding Proteins
A total of 11 putative CSPs were identified from the S. schevyrewi antennal transcriptome. Seven of them had full-length ORFs and nine of them had the predicted signal peptide. All of them shared the typical structure of a CSP except ScosCSP3 and ScosCSP10 because these two lacked the signal peptide. FIGURE 6 | Sequence alignment of candidate ScosCSPs. The four conserved cysteine residues are marked as C1-C4. Additional residues that conserved in this species were also marked.
All of the predicted CSPs shared relatively high identity (57-100%) with known insect CSPs. The phylogenetic analysis of the CSPs in different beetles showed that most of the ScosCSPs were clustered with orthologs of D. ponderosae, I. typographus, and H. parallela in a separate clade (Figure 5). Only ScosCSP2 and ScosCSP3 formed a small subgroup.
Information on sequence alignment, unigene reference, length, and best blastx hit of all 11 CSPs are shown in Figure 6 and Table 4.

Tissue-and Sex-Specific Expression of Candidate ScosOBP and ScosCSP Genes
The expression patterns of ScosOBPs and ScosCSPs were analyzed by RT-PCR and are shown in Figures 7, 8. ScosOBP1,2,3,7,9,10,16,17,18,20,and 22 were highly expressed or specifically expressed in the antennae and head tissues. Among them, ScosOBP2 and OBP18 expressed at higher levels in female antennae than in male antennae. ScosOBP4,5,6,11,12,13,15,16,and 19 were generally expressed in multiple tissues. Among them, ScosOBP12 and ScosOBP19 expressions were stronger in the female than in the male antennae. ScosOBP8 and ScosOBP21 were not detected by RT-PCR possibly because their expression levels were too low to detect. ScosCSP2, ScosCSP3, and ScosCSP5 were specifically expressed in the male and female antennae. Other ScosCSPs were expressed in multiple tissues. Among them, ScosCSP1 was not detected in male antennae and ScosCSP10 was not detected in the antennae of both the sexes. Potentially due to undetectable expression levels, ScosCSP6, ScosCSP8, and ScosCSP9 were not detected by RT-PCR.

DISCUSSION
The genes reported in our study provide additional knowledge on the pool of identified olfactory genes in coleopterans. Compared with a large number of studies on Lepidopteran species, the current understanding of olfactory genes in Coleoptera is mainly sourced from a few reported studies on T. castaneum (Engsontia et al., 2008), Megacyllene caryae (Mitchell et al., 2012), I. typographus, and D. ponderosae (Andersson et al., 2013), Leptinotarsa decemlineata (Liu et al., 2015), H. parallela (Yi et al., 2018), Rhynchophorus ferrugineus (Antony et al., 2016;Gonzalez et al., 2021), etc. S. schevyrewi belongs to the genus of bark beetles and shares similar biology with the related species that are destructive forest pests, such as I. typographus and D. ponderosae. Aggregation behaviors are critical for bark beetle survival and rely on chemical communication (Byers, 1989). The genes we identified might contribute to aggregation behavior and provide molecular targets for novel pest management techniques. We identified a total of 47 OR genes in the S. schevyrewi antennae transcriptome. In another coleopteran, 265 candidate OR genes were annotated in the T. castaneum genome (Richards, 2008), which is much more than the known number of OR genes reported by other beetles. The numbers of ORs in M. caryae (Mitchell et al., 2012), I. typographus (Andersson et al., 2013), D. ponderosae (Andersson et al., 2013), and H. parallela (Yi et al., 2018) range from 43 to 57. The number of ScosORs identified in this study is consistent with that identified in these reports. Most of the predicted ORs in S. schevyrewi share greater identity with ORs of D. ponderosae, another bark beetle, indicating that these two species may be able to share the same ecological environments and detect similar semiochemicals. Functional studies in OR from bark beetles were relatively limited to only seven ORs Yuvaraj et al., 2021). ItypOR46 and ItypOR49 were responsive to single enantiomers of the common bark beetle pheromone compounds ipsenol and ipsdienol, respectively . The other five ItypORs were responsive to monoterpenoids of different ecological origins . Future studies should be focused on deorphinized ScosORs with similar functions to provide potential molecular targets for detection and control methods.
We identified, in total, 22 IR genes in the S. schevyrewi antennae transcriptome. ScosIR8a and ScosIR25a were identified as coreceptors. The numbers of IR genes in I. typographus, D. ponderosae, and H. parallela (Yi et al., 2018) are 7, 15, and 27, respectively (Andersson et al., 2013;Yi et al., 2018). The number of ScosIRs identified in this study is considerable compared with the numbers reported in the previous studies. More than half of the predicted IRs shared relatively low identity with other coleopteran IRs. These IRs with low identity were probably not conserved in Coleoptera, and they might serve diverse functions in S. schevyrewi.
A total of 11 CSPs were identified in the S. schevyrewi antennae transcriptome. In the T. castaneum genome, a total of 40 CSPs were identified (Richards, 2008). Other coleopterans have fewer CSPs in their antennae transcriptomes; in D. ponderosae, I. typographus, and H. parallela, 11, 6, and 16 were annotated, respectively (Andersson et al., 2013;Ju et al., 2014). Our recorded number of ScosCSPs is comparable with these reports. The high level of sequence conservation (57-100%) indicates the function of CSPs is likely conserved among coleopterans.
Scolytinae beetles respond to volatiles that emanate from both the host and non-host plants (Zhang and Schlyter, 2004;Erbilgin et al., 2007;Andersson et al., 2010). However, most individuals locate target trees by relying on an important signal called an aggregation pheromone released by beetles that have already attacked a tree (Andersson et al., 2013). Thus, olfactory signals and proteins serve critical roles in insect behavior. In this study, ScosOBP1,2,3,7,9,10,16,17,18,20, and 22 might be important in odor perception because they were only expressed in the antennae and head, especially, ScosOBP2 and ScosOBP18. These two may be the key proteins in female olfactory behavior based on the specificity of protein expression we observed. ScosCSP2, ScosCSP3, and ScosCSP5 might also be important in olfaction due to their antennaespecific expression. Other ScosOBPs and ScosCSPs might not be involved in odor reception. Studies have shown a multitude of other roles that insect OBPs and CSPs have in Pelosi et al. (2018) releasing semiochemicals in pheromone glands (Benton, 2007), regeneration and development (Cheng et al., 2015), anti-inflammatory action (Isawa et al., 2002), nutrition (Zhu et al., 2016b), carrying visual pigments (Wang et al., 2007), and insecticide resistance (Bautista et al., 2015).

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found at: https://www.ncbi.nlm.nih. gov/bioproject/PRJNA732801.

AUTHOR CONTRIBUTIONS
XZ and WL designed the research, analyzed the data, and wrote the paper. YL gave a lot of adivces and help to revise the paper. AK, BS, and HC provided biological samples. XZ, BX, and ZQ performed the experiment. All authors contributed to the article and approved the submitted version.

ACKNOWLEDGMENTS
We thank Sai Zhang and Yilu Feng for rearing the experimental larvae for the study.