Insects recognize a large number of odor volatile compounds in the environment primarily via their chemosensory systems that induce corresponding behavioral responses. Compound perception is mediated by a series of proteins, including odorant-binding proteins (OBPs), chemosensory proteins (CSPs), odorant receptors (ORs), gustatory receptors (GRs), ionotropic receptors (IRs), sensory neuron membrane proteins, and odorant-degrading enzymes (ODEs) (Field et al., Reference Field, Pickett and Wadhams2000; Leal Reference Leal2013; He et al., Reference He, Mang, Wang, Wang, Ma, Wang, Chen, Zhang and He2020).
In insects, odor perception is initiated in their olfactory organs, and the initial step of odor detection is the binding of odor molecules to OBPs or CSPs (Getahun et al., Reference Getahun, Thoma, Lavista-Llanos, Keesey, Fandino, Knaden and Hansson2016). OBPs and CSPs, produced in the lymph of the chemosensilla, bind to chemical cues in the external environment and transmit them to ORs via the sensillum lymph, thereby activating signal transduction (Jacquin-Joly and Merlin, Reference Jacquin-Joly and Merlin2004; Leal Reference Leal2013; Chen et al., Reference Chen, Pan, Ma, Wang, He and He2018; Pelosi et al., Reference Pelosi, Iovinella, Zhu, Wang and Dani2018; He et al., Reference He, Durand and Dong2019a, Reference He, Chen, Li, Wang, Ma, Pan and He2019b). Electrostatic interactions between the ligand and transmembrane segments of the olfactory receptor and Van Der Waals interactions between the ligand and hydrophobic pocket of the receptor are responsible for receptor activation (Vogt, Reference Vogt, Blomquist and Vogt2003; Clark and Ray, Reference Clark and Ray2016). Eventually, odor molecules are inactivated by ODEs to maintain receptor activity by preventing persistent binding to the receptor (Durand et al., Reference Durand, Carot-Sans, Bozzolan, Rosell, Siaussat, Debernard, Chertemps and Maibeche-Coisne2011; He et al., Reference He, Zhang, Li, Yang, Zhu, Liu and Dong2014, Reference He, Zhang, Yang, Li and Dong2015).
OBPs are water-soluble small-molecule proteins and are present at high concentrations in the aqueous sensillum lymph (Zhou, Reference Zhou2010; Pelosi et al., Reference Pelosi, Iovinella, Zhu, Wang and Dani2018). They can be divided into different types in Lepidoptera, including general odorant-binding proteins (GOBPs), which are used by all insects to recognize plant volatile substances or environmental odors and are distributed within the bascionic sensillum, and pheromone-binding proteins (PBPs), which primarily exist in the antennae of adult moths and are distributed in the trichoid sensillum (Picimbon and Gadenne, Reference Picimbon and Gadenne2002; Guo et al., Reference Guo, Huang, Pelosi and Wang2012; Liu et al., Reference Liu, Yang, Yang, He, Niu, Xu, Anderson and Dong2015b). CSPs are widely distributed in various chemoreceptor lymphatic fluids of insects. They are acidic water-soluble proteins with molecular weights of 12–14 kDa, transport odor molecules to corresponding receptors, and serve as carriers of chemical irritants (Maleszka et al., Reference Maleszka, Forêt, Saint and Maleszka2007). Genome and transcriptome sequencing have led to the identification of CSPs in several insect species (Liu et al., Reference Liu, Guo, Huang, Pelosi and Wang2014b; Zhang et al., Reference Zhang, Ye, Yang and Dong2014; Cheng et al., Reference Cheng, Lu, Zeng, Liang and He2015; Ingham et al., Reference Ingham, Anthousi, Douris, Harding, Lycett, Morris, Vontas and Ranson2020).
Insect ORs, with a seven-transmembrane domain (Hopf et al., Reference Hopf, Morinaga, Ihara, Touhara, Marks and Benton2015) are located on the dendritic membrane of odorant receptor neurons (ORNs) and are generally composed of 400 amino acids. In addition to variable odorant specificity subunits, the odorant receptor co-receptor (Orco) subunit required for conventional ORs in every insect is broadly present in ORNs and plays a role in the detection of odorant substances by ORs (Larsson et al., Reference Larsson, Domingos, Jones, Chiappe, Amrein and Vosshall2004). For example, the deletion of Orco results in a loss of response to sex pheromones in Bombyx mori and greatly influences mating behavior (Liu et al., Reference Liu, Liu, Zeng, Wang, Hao and Huang2017); knockdown of Orco resulted in the inability of rice planthoppers to seek or locate host plants (He et al., Reference He, Engsontia, Chen, Yin, Wang, Lu, Zhang, Li and He2018). Furthermore, Orco is a highly conserved gene and it is important to examine its evolution and functional flexibility (Soffan et al., Reference Soffan, Subandiyah, Makino, Watanabe and Horiike2018). Similar to ORs, GRs have a seven-transmembrane domain and are also a key part of the chemosensory system in Lepidoptera (Agnihotri et al., Reference Agnihotri, Roy and Joshi2016). They are often co-expressed in single ORNs and detect different environmental stimuli (Mang et al., Reference Mang, Shu, Tanaka, Nagata, Takada, Endo, Kikuta, Tabunoki, Iwabuchi and Sato2016). The first insect GRs were identified in Drosophila melanogaster and since then, many GRs have been identified in other insects such as Anopheles gambiae, B. mori, and Heliothis virescens (Hallem et al., Reference Hallem, Dahanukar and Carlson2006). Insect GRs can be sugar-, bitter-, fructose-, and carbon dioxide-related (Xu et al., Reference Xu, Zhang and Anderson2012; Raad et al., Reference Raad, Ferveur, Ledger, Capovilla and Robichon2016). Moreover, insect GRs have additional functions, such as BmGR8 in B. mori, a sugar receptor subfamily that has a specific response to inositol and could prolong the larval duration (Zhang et al., Reference Zhang, Anderson, Trowell, Luo, Xiang and Xia2011), and DmGR5a in D. melanogaster that serves as a trehalose-specific receptor (Chyb et al., Reference Chyb, Dahanukar, Wickens and Carlson2003; Dahanukar et al., Reference Dahanukar, Lei, Kwon and Carlson2007). These findings suggest that ORs and GRs are ubiquitous in insects and the chemosensory and non-chemosensory functions they mediate are complex and diverse. However, there is a lack of studies on the exact role of ORs and GRs in the larval feeding process of serious agricultural pests such as Spodoptera litura.
S. litura (Lepidoptera: Noctuidae) is a polyphagous insect in tropical and subtropical regions worldwide (Dinesh-Kumar et al., Reference Dinesh-Kumar, Srimaan, Chellappandian, Vasantha-Srinivasan, Karthi, Thanigaivel, Ponsankar, Muthu-Pandian Chanthini, Shyam-Sundar, Annamalai, Kalaivani, Hunter and Senthil-Nathan2018; Li et al., Reference Li, Chen, Jin, Li, Wang and Shu2018). It can harm more than 389 economically important plants such as corn, peanut, and cabbage, and has become one of the most destructive agricultural pests in Asia (Ahmad et al., Reference Ahmad, Ghaffar and Rafiq2013; CABI 2018). This pest has a high capacity for reproduction and development (Wu et al., Reference Wu, Liu, Di and Yang2018). S. litura larvae feed on a wide range of plants; they eat the whole leaf, buds, fruit, and flowers (Smith et al., Reference Smith, McNamara, Scott and Holderness1997). These findings indicate that S. litura larvae have a highly selective and sensitive chemosensory system, which is crucial for finding food sources, development, and avoiding dangerous or unsuitable habitats and hosts. There is an urgent need for safe management strategies for the control of S. litura. One novel alternative to chemical pesticides, which lead to serious environmental problems, is based on the disruption of the insect chemosensory system (Xu et al., Reference Xu, Cao and Dong2010; Ingham et al., Reference Ingham, Anthousi, Douris, Harding, Lycett, Morris, Vontas and Ranson2020). In this study, we successfully identified 158 putative chemosensory genes, including 45 OBPs (nine were new), 23 CSPs, 60 ORs (three were new), and 30 GRs (three were new), in S. litura based on genomic and transcriptomic data. We then constructed phylogenetic trees based on these genes in moths and analyzed the dynamic expression levels of various genes in larval head capsules during development using quantitative real-time polymerase chain reaction (qRT-PCR). The chemosensory genes identified in this study are potential targets for future functional studies aimed at the development of novel behavioral disturbance agents to control S. litura populations.
Materials and methods
Insects and tissue sample collection
S. litura individuals were raised under laboratory conditions as follows: temperature of 27 ± 1°C, relative humidity of 75 ± 5%, and photoperiod of 14L:10D. The antennae were collected separately from 3-day-old virgin female and male adults and sent to Novogene Bioinformatics Technology Co., Ltd. (Beijing, China) for transcriptome sequencing. For gene expression analyses, new fresh eggs (Egg) (50 for each replicate), heads of 1- to 6-day-old larvae (LH) (the 1- to 2-day-old larvae, 50 for each replicate; and the 3- to 6-day-old larvae, 20 for each replicate), and antennae of 3-day-old virgin male (MA) and female (FA) adults (25 pairs for each replicate) were collected and placed in nuclease-free centrifuge tubes. Each sample type included three biological replicates, and all samples were preserved at −80°C until RNA extraction.
In accordance with our previously described methods (Zhang et al., Reference Zhang, Zhu, Fang, He, Wang, Chen, Sun, Ye, Deng and Li2015b; Zhang et al., Reference Zhang, Li, He, Sun, Li, Fang, Ye, Deng and Zhu2016), total RNAs of all samples were extracted, and high-quality RNA was used to synthesize corresponding cDNAs. Subsequently, the cDNA was used as templates for qRT-PCR analyses. The qRT-PCR primers are listed in Table S1 and amplification was executed following the manufacturer's instructions, on a LightCycler 96 Multiwell Plate. SlitGAPDH and SlitEF were used as internal reference genes (Zhang et al., Reference Zhang, Li, He, Sun, Li, Fang, Ye, Deng and Zhu2016) to calculate relative expression levels. The data were analyzed using Q-Gene (Muller et al., Reference Muller, Janovjak and Miserez2002; Simon, Reference Simon2003). Each biological sample had three technical replicates.
The purity of total RNA was determined using the NanoDrop instrument (based on the OD260/280 ratio). Subsequently, cDNA synthesis, library construction, sequencing, de novo assembly, and annotation of unigenes were performed following our previously described procedures (Zhang et al., Reference Zhang, Zhu, Fang, He, Wang, Chen, Sun, Ye, Deng and Li2015b; Zhang et al., Reference Zhang, Li, He, Sun, Li, Fang, Ye, Deng and Zhu2016).
Reference sequence alignment
HISAT (2.0.4) was used to analyze the genomic location of the filtered sequences. HISAT effectively compares spliced reads with RNA sequencing data (RNA-seq). It is currently the most accurate alignment software. In this study, default parameters were used. If an appropriate reference genome is selected and the data are not contaminated, the percentage of mapped reads or fragments generated by a laboratory is typically greater than 70%. The percentage of multiple mapped reads or fragments assigned to multiple locations does not typically exceed 10% (Mortazavi et al., Reference Mortazavi, Williams, McCue, Schaeffer and Wold2008).
Analysis of expression abundance
The expression abundance of each unigene was calculated following the Fragments Per Kilobase per Million (FPKM) method. This method can eliminate the influence of differences in length and number of sequences on the expression abundance. In this study, HTSeq (v0.6.1) was used to evaluate the expression levels. FPKM values of 0.1 and 1 were used as thresholds to determine whether a gene is expressed.
Phylogenetic trees based on amino acid sequences were constructed using OBPs, CSPs, ORs, and GRs of S. litura and other Lepidoptera species including B. mori, Manduca sexta, Helicoverpa armigera, Dendrolimus punctatus, Epiphyas postvittana, Spodoptera littoralis, and Danaus plexippus. This method is more convenient when analyzing the evolutionary relationship of chemosensory genes between S. litura and other Lepidoptera species and also helps to infer the possible functional characteristics of some chemosensory genes of S. litura. Signal peptide analysis, sequence alignment, and phylogenetic tree construction were performed following our previously reported procedures (Zhang et al., Reference Zhang, Zhu, Fang, He, Wang, Chen, Sun, Ye, Deng and Li2015b; Zhang et al., Reference Zhang, Zhu, Ma, Dong, Xu, Kang and Zhang2017).
Cycle threshold (Ct) values were obtained using qRT-PCR and used to determine the mean normalized expression value for each gene using Q-Gene 96 method. The expression levels of chemosensory genes were compared using one-way analysis of variance in SPSS 21.0 (SPSS Inc., Chicago, IL, USA); values of P < 0.05 indicated a significant difference.
Results and discussion
Overview of transcriptomic data
We assembled and analyzed the transcriptomes of four tissues of S. litura, based on the Illumina sequencing data, as follows: eggs (referred to as SlitEgg; reads from National Center for Biotechnology Information [NCBI] server ID: DRX080341), larval heads (referred to as SlitLH; reads from NCBI server ID: DRX080340), female antennae (referred to as SlitFA), and male antennae (referred to as SlitMA) (fig. 1). In total, we obtained 3,766,1368 raw reads from SlitEgg, 39,848,092 from SlitLH, 70,642,568 from SlitFA, and 73,142,860 from SlitMA. After removing low-quality and adaptor-containing reads, 38,709,312, 37,661,368, 70,642,568, and 73,142,860 clean reads were obtained from the four tissues, respectively. The GC contents and Q30 values were over 40 and 90%, respectively (table 1), suggesting that the transcriptomic data were of good quality and could be used for subsequent analyses.
The proportion of total reads mapped to the reference genome was greater than 70%, with less than 2.5% of reads mapped to multiple locations, and spliced reads accounted for less than 30% of the total reads (Table S2), indicating the absence of contamination and appropriate reference genome selection. In an RNA-seq analysis, gene expression levels could be estimated by obtaining counts in genome regions or reads in gene exons. The read count is positively correlated with gene length and sequencing depth in addition to the true gene expression level. FPKM values are usually used to compare the expression levels of different genes. We found that the distributions of FPKM values in the four tissue types of S. litura at each interval value were similar, and the highest proportion of reads had an FPKM value of 3–60 (Table S3). This indicates that the differentially expressed genes are common to the different tissues of S. litura, providing a basis for the identification of key differentially expressed genes for further functional analyses.
Characteristics of chemosensory genes in S. litura
Some chemosensory genes belonging to S. litura have been obtained using RNA-seq analyses exclusively on adult antennae or pheromone glands. Gu et al. (Reference Gu, Zhou, Gao, Wang, Li, Guo and Zhang2015) reported 38 OBPs obtained from adult antennae, and Feng et al. (Reference Feng, Lin, Zheng, Qian, Chang and Du2015) reported 74 (26 ORs, 21 OBPs, 18 CSPs) obtained from adult antennae. In addition, we previously obtained 25 OBPs and 14 CSPs from pheromone glands of female moths (Zhang et al., Reference Zhang, Zhu, Fang, He, Wang, Chen, Sun, Ye, Deng and Li2015b). However, the type and number of chemosensory genes identified in these studies are less comprehensive than those based on genome analyses (Cheng et al., Reference Cheng, Wu, Wu, Chilukuri, Huang, Yamamoto, Feng, Li, Chen, Guo, Liu, Li, Wang, Peng, Liu, Guo, Fu, Li, Liu, Chen, Tomar, Hilliou, Montagne, Jacquin-Joly, d'Alencon, Seth, Bhatnagar, Jouraku, Shiotsuki, Kadono-Okuda, Promboon, Smagghe, Arunkumar, Kishino, Goldsmith, Feng, Xia and Mita2017). A total of 45 IRs have been identified and analyzed based on the genomic and transcriptomic data by Zhu et al. (Reference Zhu, Xu, Zhang and Liu2018). Using phylogenetic analysis and extensive-expression profiles, they demonstrated that SlitIRs have diverse functional roles in olfaction, taste, and reproduction. Therefore, in this study, we conducted a more in-depth analysis based on the results of genome analyses. According to the sequence similarity and characteristics of insect chemosensory genes, 158 putative genes (45 OBPs, 23 CSPs, 60 ORs, and 30 GRs) were identified from the transcriptomic and genomic data (Tables S4 and S5), which included 15 new genes (nine OBPs, three ORs, and three GRs), a number higher than that of genes identified in other moth species such as H. armigera (122 genes) (Liu et al., Reference Liu, Xu, Papanicolaou, Dong and Anderson2014a), H. assulta (147 genes) (Xu et al., Reference Xu, Papanicolaou, Liu, Dong and Anderson2015), and Plutella xylostella (116 genes) (Yang et al., Reference Yang, Cao, Wang and Liu2017). The large amount of transcriptome data suggests that S. litura is a useful taxon for understanding chemosensory genes in insects.
We obtained 45 different unigenes encoding putative OBPs in S. litura, including 15 that have not been reported in previous genomic studies of the species (Cheng et al., Reference Cheng, Wu, Wu, Chilukuri, Huang, Yamamoto, Feng, Li, Chen, Guo, Liu, Li, Wang, Peng, Liu, Guo, Fu, Li, Liu, Chen, Tomar, Hilliou, Montagne, Jacquin-Joly, d'Alencon, Seth, Bhatnagar, Jouraku, Shiotsuki, Kadono-Okuda, Promboon, Smagghe, Arunkumar, Kishino, Goldsmith, Feng, Xia and Mita2017) and nine that were newly identified. Sequence analysis showed that all sequences had full-length open reading frames (ORFs) and ten SlitOBPs did not have signal peptide regions. Furthermore, based on the FPKM results, 43, 39, 36, and 37 SlitOBPs were found in the Egg, LH, FA, and MA samples, respectively, and SlitOBPs were most frequent in the LH (14 genes, including three new SlitOBPs), followed by MA (ten genes, including two new SlitOBPs), FA (eight genes, including three new SlitOBPs), and Egg (five genes, including one new SlitOBP) samples (fig. 2), indicating that these genes of larval and adult moths play important roles in the perception of external odorants involving in development and hatching or host searching by newly hatched larvae.
The phylogenetic analysis showed that 36 SlitOBPs and nine new SlitOBPs were widely distributed along various branches. SlitPBP1, SlitPBP2, SlitPBP3, SlitGOBP1, and SlitGOBP2 formed a group including the PBP/GOBP subfamily (fig. 2c), suggesting that SlitOBPs have functions similar to those of other moth OBPs (Poivet et al., Reference Poivet, Rharrabe, Monsempes, Glaser, Rochat, Renou, Marion-Poll and Jacquin-Joly2012; Jeong et al., Reference Jeong, Shim, Oh, Yoon, Kim, Moon and Montell2013; Pelosi et al., Reference Pelosi, Mastrogiacomo, Iovinella, Tuccori and Persaud2014; Liu et al., Reference Liu, Yang, Yang, He, Niu, Xu, Anderson and Dong2015b). Notably, seven SlitOBPs (SlitOBP6, SlitOBP8, SlitOBP9, SlitOBP11, and SlitOBP31–33, SlitOBP32 identified as new OBP) formed a cluster, indicating that these genes have species-specific functions, which should be evaluated in the future.
Twenty-three putative genes encoding CSPs were obtained in S. litura based on the transcriptomic data, consistent with previous counts based on the S. litura genome (Cheng et al., Reference Cheng, Wu, Wu, Chilukuri, Huang, Yamamoto, Feng, Li, Chen, Guo, Liu, Li, Wang, Peng, Liu, Guo, Fu, Li, Liu, Chen, Tomar, Hilliou, Montagne, Jacquin-Joly, d'Alencon, Seth, Bhatnagar, Jouraku, Shiotsuki, Kadono-Okuda, Promboon, Smagghe, Arunkumar, Kishino, Goldsmith, Feng, Xia and Mita2017). All SlitCSPs had full-length ORFs with four conserved cysteines and 21 SlitCSPs were predicted to have signal peptides at their N termini. Based on the FPKM results, 21, 22, 23, and 23 SlitCSPs were found in the Egg, LH, FA, and MA tissues, respectively, and eight genes were abundantly expressed in FA, four in MA, and seven in LH. Similarly, the number of SlitCSPs enriched in the LH was higher than that in the Egg (fig. 3). SlitCSPs might play a key role in physiological and behavioral activities in male adults and larvae.
A phylogenetic tree of moth CSPs was constructed using various CSPs of S. litura, B. mori, M. sexta, and H. armigera. Most branches were highly supported (bootstrap values greater than 70%), and most SlitCSPs formed a cluster on the same branch with BmorCSPs, MsexCSPs, and HarmCSPs (fig. 3c). Based on these findings, these SlitCSPs may have chemosensory or other physiological functions similar to those of CSPs in other moths (Pelosi et al., Reference Pelosi, Calvello and Ban2005; Celorio-Mancera Mde et al., Reference Celorio-Mancera Mde, Sundmalm, Vogel, Rutishauser, Ytterberg, Zubarev and Janz2012; Zhang et al., Reference Zhang, Ye, Yang and Dong2014).
Sixty unigenes encoding putative ORs were acquired by analyzing the transcriptome data for S. litura, of which four were not found in previous S. litura genomic studies (Cheng et al., Reference Cheng, Wu, Wu, Chilukuri, Huang, Yamamoto, Feng, Li, Chen, Guo, Liu, Li, Wang, Peng, Liu, Guo, Fu, Li, Liu, Chen, Tomar, Hilliou, Montagne, Jacquin-Joly, d'Alencon, Seth, Bhatnagar, Jouraku, Shiotsuki, Kadono-Okuda, Promboon, Smagghe, Arunkumar, Kishino, Goldsmith, Feng, Xia and Mita2017) and three were novel SlitORs (SlitOR74, 75, 77). Except for SlitOR74 and SlitOR75, other SlitORs were predicted to have full-length ORFs that encode 385–473 amino acids with 3–8 transmembrane domains (TMDs). The FPKM analyses indicated 31, 22, 59, and 58 SlitORs in the Egg, LH, FA, and MA tissues, respectively, and SlitORs were most frequently expressed in MA (30 genes, including a new gene SlitOR75), followed by FA (24 genes, including the new gene SlitOR74), Egg (three genes, including the new gene SlitOR77), and LH (2 genes) (fig. 4). These results indicate that SlitORs are involved in the chemosensory process of adult moths and have different functions similar to ORs of other insects (Carraher et al. Reference Carraher, Dalziel, Jordan, Christie, Newcomb and Kralicek2015, Yan et al. Reference Yan, Jafari, Pask, Zhou, Reinberg and Desplan2020), and the OR genes expressed in eggs may help the newly hatched S. litura larvae to search for the host in time.
A phylogenetic tree was constructed based on the amino acid sequences of moth ORs from S. litura, B. mori, H. armigera, and D. punctatus, E. postvittana, S. littoralis, and D. plexippus, and each cluster had at least one orthologue in another moth taxon (fig. 4c). All the three new SlitOR genes were clustered with the OR genes of other moths (SlitOR74-BmorOR67, SlitOR75-HarmOR35, and SlitOR77-DpleOR9/DpleOR10/DpleOR11/DpleOR42/BmorOR32), indicating that these genes may have similar physiological functions. A total of six SlitORs (SlitOR1, 3, 5, 6, 7, and 57) clustered with the moth PR subfamily, the number is more than that in previous reported (Zhang et al., Reference Zhang, Yan, Liu, Jacquin-Joly, Dong and Wang2015a). According to the recent study of a novel candidate pheromone receptors in moths (Bastin-Heline et al., Reference Bastin-Heline, de Fouchier, Cao, Koutroumpa, Caballero-Vidal, Robakiewicz, Monsempes, Francois, Ribeyre, Maria, Chertemps, de Cian, Walker, Wang, Jacquin-Joly and Montagne2019), four SlitORs (SlitOR11, SlitOR13, SlitOR14, and SlitOR53) in S. litura were clustered in novel lineage of candidate PRs. In the future, we plan to conduct in-depth functional analysis of these traditional and novel candidate PRs to help better reveal the exact molecular mechanism of sex pheromone perception in S. litura. As expected, SlitOR2clustered with the Orco genes of other moths, indicating that the gene belong to the conserved Orco subfamily.
We obtained 30 different unigenes encoding putative SlitGRs in S. litura, three of which were previously unreported (Cheng et al., Reference Cheng, Wu, Wu, Chilukuri, Huang, Yamamoto, Feng, Li, Chen, Guo, Liu, Li, Wang, Peng, Liu, Guo, Fu, Li, Liu, Chen, Tomar, Hilliou, Montagne, Jacquin-Joly, d'Alencon, Seth, Bhatnagar, Jouraku, Shiotsuki, Kadono-Okuda, Promboon, Smagghe, Arunkumar, Kishino, Goldsmith, Feng, Xia and Mita2017). However, this number is far less than that identified based on the S. litura genome (Cheng et al., Reference Cheng, Wu, Wu, Chilukuri, Huang, Yamamoto, Feng, Li, Chen, Guo, Liu, Li, Wang, Peng, Liu, Guo, Fu, Li, Liu, Chen, Tomar, Hilliou, Montagne, Jacquin-Joly, d'Alencon, Seth, Bhatnagar, Jouraku, Shiotsuki, Kadono-Okuda, Promboon, Smagghe, Arunkumar, Kishino, Goldsmith, Feng, Xia and Mita2017). This may be explained by the presence of GRs in other chemosensory tissues of S. litura, such as the larval mouthparts or adult labial palps, which should be evaluated further. Sequence analysis showed that 27 sequences had full-length ORFs that encode 339–482 amino acids with 5–8 TMDs. Based on the FPKM results, 18, 13, 23, and 26 SlitGRs were found in the Egg, LH, FA, and MA tissues, respectively. Similar to SlitORs, SlitGRs were most frequent in MA (14 genes), followed by FA (seven genes, including the new gene SlitGR240), Egg (nine genes, including two new genes SlitGR241 and 242), and LH (0 genes) (fig. 5), indicating that these SlitGRs may play key roles in the chemosensory process of adult moths, based on the antennae.
A phylogenetic tree was constructed based on GRs of S. litura, B. mori, H. armigera, and D. plexippus. Three SlitGRs (SlitGR1, 2, and 3) belong to the carbon dioxide receptor subfamily, 2 SlitGRs (SlitGR8 and 9) belong to the fructose receptor subfamily, and 4 SlitGRs (SlitGR4, 5, 6, and 7) belong to the sugar receptor subfamily (fig. 5c), indicating that these SlitGRs participate in the recognition of carbon dioxide (Jones et al., Reference Jones, Cayirlioglu, Kadow and Vosshall2007; Kwon et al., Reference Kwon, Dahanukar, Weiss and Carlson2007), fructose (Jiang et al., Reference Jiang, Ning, Guo, Jia, Huang, Qu and Wang2015), and sugar (Sato et al., Reference Sato, Tanaka and Touhara2011).
Dynamic expression of chemosensory genes in the heads of S. litura larvae
We randomly selected eight chemosensory genes (SlitOBP6, SlitOBP8, SlitOBP9, SlitOBP25, SlitCSP10, SlitCSP18, SlitOR77, and SlitGR22) for validation using qRT-PCR. The qRT-PCR results were consistent with the FPKM data (fig. 6).
To estimate the function of chemosensory genes during larval feeding in S. litura, we measured the dynamic expression levels of genes with LH-biased expression using qRT-PCR. Prior to these experiments, it was necessary to establish the criteria for determining distinct stages of larval development. According to previously established criteria, the larvae were assigned to six instars according to the size of the larval head capsule (Sannino and Espinosa, Reference Sannino and Espinosa1999). In total, 2 (SlitCSP1 and SlitCSP7) and 7 (SlitOBP8, SlitOBP9, SlitOBP25, SlitCSP18, SlitOR34, SlitGR240, and SlitGR242) genes had higher expression in the heads of 3- to 5-day-old larvae than at other stages (P < 0.05), respectively (fig. 7). Usually, 3- to 5-day-old larvae of S. litura and other noctuid moths are crucial for crop feeding (Jacquin-Joly et al., Reference Jacquin-Joly, Vogt, Francois and Nagnan-Le Meillour2001; Huang et al., Reference Huang, Huang, Zheng and Shen2009; Zou et al., Reference Zou, Xu, Zou, Liu, Chen, Feng and Zheng2016). SexiOBP13 of S. exigua is highly expressed in the heads of 3-day-old larvae and can bind to two kinds of host volatiles (nerolidol and farnesol) (Jin et al., Reference Jin, Liu, Liu and Dong2015). SexiOBP2 of S. exigua displays LH-biased expression and functional studies have indicated that it may participate in larval food searching (Liu et al., Reference Liu, Zhang, Ye, Li and Dong2015a). De Fouchier et al. (Reference De Fouchier, Sun, Caballero-Vidal, Travaillard, Jacquin-Joly and Montagné2018) found nine ORs in S. littoralis larvae that are involved in host searching. BmorGR66 of B. mori shows larval maxilla-biased expression and is a major factor affecting the larvae feeding preference (Zhang et al., Reference Zhang, Zhang, Niu, Ji, Liu, Li, Bai, Palli, Wang and Tan2019). Based on these findings, we predicted that the nine chemosensory genes of S. litura may play key roles in larval feeding (Jacquin-Joly et al., Reference Jacquin-Joly, Vogt, Francois and Nagnan-Le Meillour2001). However, the other three genes (SlitCSP6, SlitOR77, and SlitGR13) of S. litura did not show similar expression characteristics, suggesting that they may be involved in olfactory and other physiological behaviors of larvae; however, experimental confirmation is needed.
The supplementary material for this article can be found at https://doi.org/10.1017/S0007485321000109
The authors thank Miss Yong-Jie Zhang for insect photography, Dr Tingcai Cheng for kindly proving the genomic data of S. litura, and Professor Kazuei Mita for kindly proving the sequences of some S. litura chemosensory genes.
This work was supported by the National Natural Science Foundation of China (31970456), Anhui Provincial Natural Science Foundation (2008085MC63), and Natural Science Fund of Education Department of Anhui Province, China (KJ2019A0586 and KJ2019B09).
Conflict of interest
The authors declare that they have no conflict of interest.
This article does not include any study on human participants or animals performed by any of the authors.