The application of next-generation sequencing technology in the field of entomology has greatly promoted the efficiency and quantity of gene annotation [19]. Meantime, a lot of antennal transcriptomes olfactory-related genes were identified [20-22]. In this research, we identified 26 OBP genes, 19 CSP genes, 55 OR genes and 20 IR genes from the C. pinicolalis antennal transcriptome, and all the genes are reported for the first time. C. pinicolalis is a sibling species of C. punctiferlis, and has ever been recognized as the same species [10]. In C. punctiferlis, totally 25 OBPs, 15 CSPs, 62 ORs and 10 IRs were identified from antennae transcriptome [23], and the numbers of OBPs, CSPs and ORs are similar with C. pinicolalis, whereas more IRs were identified from the C. pinicolalis antennal transcriptome dataset. The sequence similarity of olfactory-related genes was analyzed and shown in the evolution tree (Fig. 3, Table 3), OBP, CSP, OR and IR genes sequences showed high similarity with C. punctiferlis. Most of the identities are more than 90%, 4 OBP, 5 OR, 2 IR and 2 CSP genes had 99% sequence similarity with the C. punctiferlis (Table 3). These two pests were first identified by Koizumi et al. [7] and classified into pinaceae-feeding type (PFT) and fruit-feeding type (FFT) based on their feeding habits and morphological characters. They were later named as C. pinicolalis and C. punctiferalis [10]. Further investigation revealed their behaviors, morphologies, and feeding patterns, and indicated reproductive isolation between these two types [9,16,18]. Recently, using mitochondrial cytochrome c oxidase subunits I, II and cytochrome b gene sequencing, Wang et al. proved that the C. pinicolalis was different from C. punctiferalis [15]. The phylogenetic tree also revealed an evolutionary relationship with other Lepidopteran species. The GOBP/PBP genes sequences include six subgroups (GOBP1 and 2, PBP1-4) formed a conserved order (Fig. 3). ORs and IRs genes indicated the Ostrinia furnacalis is the next close neighbor in the same clade (Additional file 2: Figure. S3). On the other hand, OBPs and CSPs genes showed Cnaphalocrocis medinalisin in the same clade as a close neighbor after C. punctiferlis. Olfactory-related genes in Bombyx mori showed gene divergence when compared with these two sibling species.
Menken et al [25] suggested the two major transitions in the evolution of larval (Lepidoptera) feeding, switching from litter-feeding to herbivory. Larvae feeding on leaf-litter from a single dominant tree species would have been the main precursor for evolving from litter-feeding to leaf-mining type. In the course of evolution, leaf-mining type gained the new type of enzymatic system to digest the nutritious freshly fallen leaves. Once this evolved niche had been acquired the ability of leaf-mining and with the special digestive system could apparently exploit the diversity more and larval feeding mode had evolved in searching of new host-plants [26]. Insects olfaction system allows them to recognize and track the volatile cues from host-plant, mating and evade from their predators. The polyphagous insects significantly adapted to recognize, digest and detoxify a large variety of host-plants. Polyphagous insects must handle the defensive toxic molecules (secondary metabolites) produced by the host-plant. Genes from the moth pheromone glands could have evolved and altered the normal fatty acid metabolism [27]. In a previous study, experiments proved the major change in the pheromone blend in various moth species, the existence of different desaturase from mRNA in the moth pheromone gland [28]. In Spodoptera frugiperda, due to tandem duplications within a single region of the genome 10 OBP genes expansion was observed when compared with B. mori. In the same study, the author showed a difference in IRs gene count between the strains, S. frugiperda corn strain had 42 IRs and rice strain had 43 IRs [29]. Similarly, in our study C. pinicolalis had 10 more IRs when compared with C. punctiferlis. Evidently, the selection of host plant is also a reason that leads to gene duplications, insertions or deletions when there is a need to adapt to an environment.
As in other insects [30-32] OBPs and CSPs were highly detected in the antennae of both male and female (Additional file 3: Table S1). Among these genes, many of them were sexual biased genes (Fig. 1). PBPs were widely thought to be sex pheromone binding function, normally insects have 3-5 PBP genes. Many research reported at least one of the PBPs could bind well with the sex pheromones [33-35]. In our analysis, PBP2 showed significantly male biased expression, and PBP1, PBP3 and PBP4 showed significantly female biased expression. In male moth, the main assignment is to trail the sex pheromones to find a female moth for mating. We speculated the PBP2 might play a critical role in pheromone binding. Females are often selective when deciding to mate they find a healthy counterpart. GOBP1 and GOBP2 genes, as well as OBP6, OBP7 and OBP9, were also highly expressed in female, this may play some important roles and need for further study. GOBPs are proposed to detect host plants volatiles, food and oviposition sites and PBPs play a key role in detecting sex pheromones [36-38]. However, some studies have demonstrated that GOBPs were also strongly bound with sex pheromones and possibly responsible for conducting the function [39]. Sex pheromones interaction with GOBPs provides us with an important vision, adaptive evolution (gene expansion) might occur in C. pinicolalis. Although the transcriptome of C. pinicolalis and C. punctiferlis possess higher similarity, the C. pinicolalis adult rely on fresh masson pine branches for laying eggs, which the case is very different in C. punctiferlis adult, they have a wide variety of host plants selection. Therefore, both GOBPs and PBPs from C. pinicolalis and C. punctiferlis might have a greater interest in future research.
CSPs were found in insect contact lymph and olfactory sensilla, but members showed peculiar functions. Binding activity of volatile compounds has been described in a similar way as OBPs. In C. pinicolalis antennae transcriptome, we totally identified 19 putative CSPs, and found the transcript per kilobase million (TPM) values of five CSPs (CSP4, CSP5, CSP11, CSP14, and CSP17) were significantly higher in female antennae (Fig. 1B). MsepCSP8 of Mythimna separate was specially expressed in female antennae and showed less sensitive to plant volatiles after RNAi [40]. Also in Locusta migratoria, nearly 17 CSPs abundantly expressed in the female reproductive organs [41]. The above results provide a valuable understanding that CSPs may have an important role in female, especially when it comes to tracking the volatile cues from host-plants and oviposit.
OR or IR genes are responsible for receiving and detecting odor molecules sent by OBP during the recognition process. Totally there were 55 OR genes identified from male and female antennal transcriptome dataset, among them 22 ORs showed a significant difference in TPM ratio (Table S1). In Lepidoptera, OR1 and OR3-8 were identified as pheromone receptors. Our result obviously showed OR1, OR3 and OR6 were specially expressed in male antennae, this may suggest OR1, OR3 and OR6 genes focus on sex pheromones recognition. PRs in other Lepidoptera were reported to bind with sex pheromones [42]. OR34 also performed biased expression in male antennae, but till now, the function is unknown. More numbers of ORs were highly expressed in female antennae (Fig. 2), this is also discovered in mosquitos [43], female need more receptors for host seeking, especially during oviposition. In Bombyx mori, more female biased ORs suggested having function of oviposition cues or male-produced courtship pheromones [44]. This indicated more OR bias in female C. Pinicolalis might provide more receptors for the detection of correct host plants and sex pheromones as well.
We have identified 20 IRs in pinicolalis that is one-fold higher than the IRs reported in punctiferlis. Indeed, the expression of IR may have a certain specificity. For example, there are some IRs were expressed exclusively in Spodoptera littoralis and Helicoverpa armigera [45, 46]. Also, different IR genes were detected in gustatory organs in Drosophila melanogaster [47]. However, in this study the IR gene family from transcriptome data analyzed only from the C. pinicolalis antennae and compared with C. punctiferlis antennal dataset. Based on the transcriptome data analysis, we cannot conclude that there are only 20 (C. pinicolalis) and 11 (C. punctiferlis) [23] IR isoforms in C. pinicolalis and C. punctiferlis antenna. The identified IR isoforms in C. pinicolalis could help to study gene expansion/deletion and existence of other possible IR isoforms in the C. punctiferlis antenna and evolutionary relationship between these two species. NormFinder and geNorm programs are commonly used to screen and optimize the number of internal reference genes for qRT-PCR analysis [48, 49]. At the same time, the difference between reference genes can be compared, but only one optimal gene can be screened when using the NormFinder [50]. In this research, we used both methods to screen the reference gene. The GeNorm result showed Actin and GAPDH were more stable during different development stages of the C. pinicolalis, and NormFinder showed the RP49 as a stable reference gene. This variation may be due to different algorithms coded in this software. Different software were used for calculating the reference gene stability at different developmental stages in the yellow peach moth, RP49 and GAPDH was found to be more stable [51]. Since the expression of the reference gene differs for different developmental stage and tissue, therefore the selection of two or more reference genes is useful to calibrate the expression level of the gene of interest. Cardoso et al. [52] reported three different reference genes (Actin, 60S ribosomal protein L3, RPL13; and peptidylprolyl isomerase, PPI) for different developmental stages in Aphidius gifuensis [53]. Also, Actin, GAPDH and RP49 reported being the most stable reference gene in the Calliphoridae family [54]. According to our results, it is recommended to use GAPDH or RP49 at different developmental stages of the C. pinicolalis. On another hand, ribosomal proteins are involved in translation and protein synthesis, this recommended us to use RP49 and RPL13 for different tissues in yellow peach moth [51]. Similarly, our findings indicate that both RP49 and RPL13 are the best reference genes for the different body part of the adult.
Furthermore, the female bias genes expression level of OBPs (PBP1, PBP3, PBP4, GOBP1, GOBP2, OBP6, OBP7 and OBP9) were verified by RT-qPCR and extremely consistent with the TMP values obtained from the transcriptome dataset. In addition, the fold change expression results of CSPs, ORs and IRs are consistent with the TMP values (Additional file 2: Figure. S2). Therefore, we compared these olfactory-related gene expression levels of C. pinicolalis with C. punctiferalis, reported by Ge xing et al., 2016 [23]. Astonishingly, gene expression pattern reported from C. punctiferalis mostly differs from our study. Noteworthy, most of the ORs (OR2, OR3, OR5, OR6, OR13 and OR15) were significantly expressed in male antenna, whereas in C. punctiferalis the ORs were highly expressed in female antenna. At this point, we suggest these ORs might be functionally active in male moths when comparing with C. punctiferalis males. On the other hand, OBPs (OBP2, 5 and 6) and GOBPs (GOBP1 and 2) expression patterns were similar to that of C. punctiferalis. Exclusively, PBP (PBP1, 2, 3 and 4) genes expression was highly recorded in the C. punctiferalis male antenna [23]. In contrast, PBP1, 3 and 4 genes were significantly expressed in C. pinicolalis female antenna, only PBP3 had a similar expression pattern. However, most of the gene expression patterns of these olfactory-related proteins were different when compared with C. punctiferalis dataset [23], since C. pinicolalis is a monophagous pest that mainly feeds on Masson pines.