Candidate odorant binding proteins and chemosensory proteins in the larval chemosensory tissues of two closely related noctuidae moths, Helicoverpa armigera and H. assulta

In order to acquire enough nutrients and energy for further development, larvae need to invest a large portion of their sensory equipments to identify food sources. Yet, the molecular basis of odor-driven behavior in larvae has been poorly investigated. Information on olfactory genes, particularly odorant binding proteins (OBPs) and chemosensory proteins (CSPs) which are involved in the initial steps of olfaction is very scarce. In this study, we have identified 26 OBP and 21 CSP genes from the transcriptomes of Helicoverpa armigera larval antennae and mouthparts. A comparison with the 34 OBP and 18 CSP genes of the adult antenna, revealed four novel OBPs and seven novel CSPs. Similarly, 27 OBPs (six novel OBPs) and 20 CSPs (6 novel CSPs) were identified in the transcriptomes of Helicoverpa assulta larval antennae and mouthparts. Tissue-specific profiles of these soluble proteins in H. armigera showed that 6 OBP and 4 CSP genes are larval tissue-specific, 15 OBPs and 13 CSPs are expressed in both larvae and adult, while the rest are adult- specific. Our data provide useful information for functional studies of genes involved in larval foraging.


Introduction
For survival, insects need a specialised sensory system to monitor environmental odors. Olfactory stimuli in Lepidoptera can be divided into intra-specific pheromones, mainly mediating communication between sexes, and plant volatiles used as cues for larval foraging and oviposition [1][2][3]. Odor detection is achieved by ten thousand chemosensilla on the two main sensory organs, antenna and mouthparts, housing olfactory sensory neurons (OSNs) that respond to volatiles and send electrical impulses to antennal lobes. From these organs cognate project PLOS  performed a transcriptome analysis to identify OBP and CSP genes in larval chemosensory organs of H. armigera and H. assulta. Moreover, we conducted RT-PCR assays on H. armigera adult and larval olfactory organs to find OBP and CSP genes with specific expression in larval antennae or mouthparts.

Insect rearing
H. armigera were reared at the Institute of Plant Protection, Chinese Academy of Agricultural Sciences, Beijing, China. The H. assulta larvae were collected from the tobacco fields with the permission of the Experiment Station of Henan University of Science and Technology in Xuchang, Henan Province, China. Larvae were reared on an artificial diet and placed on a 16:8 h (light: dark) photoperiod at 27 ± 1˚C, 55-65% RH. Pupae were sexed and male and female individuals were placed in separate cages for eclosion. The adults were fed on 10% honey solution. In expression profile studies, all adult tissues were collected from 3-day-old male and female moths, all larval tissues were collected from fifth instar larvae.

RNA extraction
Fresh larval antennae and mouthparts were grinded in a liquid nitrogen cooled homogenizer, later adding 1mL of TriZol reagent (Invitrogen, Carlsbad, CA, USA) and the total RNA extraction were performed following the manufacturer's instructions. The RNA sediment was dissolved in 20μL RNase-free water, RNA integrity was verified by gel electrophoresis. RNA quantity were measured on a Nanodrop ND-2000 spectrophotometer (NanoDrop products, Wilmington, DE, USA) and purity was verified by gel electrophoresis.

cDNA library construction and sequencing
Five micrograms total RNA of each samples (H. armigera larval antennae, H. armigera larval mouthparts, H. assulta larval antennae, and H. assulta larval mouthparts,) was used to construct the cDNA library respectively. cDNA library construction and Illumina HiSeq 2000 (Illumina, San Diego, CA, USA) sequencing of the samples were performed at Beijing Genomics Institute (BGI, Shenzhen, China). The length of insert sequence was around 200 bp. The libraries were paired-end sequenced using PE90 strategy. The detailed procedures have been described in previous work from our laboratory [33,34].

Assembly and functional annotation
After removing low quality reads, trimming low quality nucleotides of both ends, trimming 3 adaptors and poly-A/T tails, the remainder raw-reads were considered as clean-reads. De novo assembly in each sample was conducted using Trinity (version 20120608). Then the unigenes derived from the Trinity outputs were clustered by TGICL [46,47]. The consensus cluster sequences and singletons make up the unigenes dataset. The annotation of unigenes were performed via a NCBI blastx against non-redundant (nr) and SwissProt database. Candidate unigenes encoding putative OBPs and CSPs, were identified according to nr and SwissProt annotation results.
Alignments of amino acid sequences (without signal peptides) were performed by ClustalX 2.0. The phylogenetic trees of OBPs and CSPs were constructed using MEGA5 software by the neighbor-joining method with Jones-Taylor-Thornton (JTT) model and the node support was assessed using a bootstrap procedure of 1000 replicates.  Insect OBPs are generally grouped into three main subfamilies: "Classic" OBPs with six conserved cysteines, "Minus-C" with only four cysteines, and "Plus-C" with more cysteines in addition to those of the conserved motif [45,48,49]. Among the larval OBPs, 14 of H. armigera and 17 of H. assulta were assigned to the Classic OBP group, while 3 can be classified as Minus-C OBPs in both species. 7 OBPs in both species belong to the Plus-C group, while others could not be assigned due to incomplete sequences (Fig 1).

Illumina sequencing and functional annotation
A phylogenetic tree was constructed using OBP sequences from H. armigera, H. assulta, H. virescens, M sexta and B. mori (Fig 2). Accordingly, the OBPs can be grouped into ABPI (antennal binding protein I), ABPII (antennal binding protein II), CRLBP (classic OBP), Minus-C, Plus-C, and PBP/GOBP (general odorant binding protein/pheromone binding protein) clusters based on the classification of OBPs from B. mori [45]. At the same time, most OBPs of H. armigera and H. assulta defined as Minus-C and Plus-C clustered with B. mori proteins of the same groups. However, among "classic" OBPs, only two sequences were found in the CRLBP branch, the others in the ABPX branches. Based on the bootstrap values on the tree, for all novel HarmOBPs we could find orthologous genes in H. assulta with more than 90% sequence identity. Only for HassOBP38 we could not identify an orthologue in H. armigera.

Identification of candidate chemosensory proteins
In our transcription sets, a total of 21 sequences in H. armigera and 20 sequences in H. assulta can be matched with sequences of known CSPs in other Lepidoptera species. Of these, 17 HarmCSPs and HassCSPs had full-length ORFs and predicted signal peptides. Their lengths range from 107 to 292 amino acids (Table 3). A comparison with CSPs previously reported for Candidate OBPs and CSPs in the larval chemosensory tissues of Helicoverpa armigera and H. assulta   (Fig 3). These sequences were used to build a neighbor-joining tree with the CSPs of C. suppressalis, B. mori and H. virescens. In the tree we could recognize four groups of genes clustered together with a 99% bootstrap value, while the remaining sequences could not be grouped. Based on this homology analysis, we named the novel CSPs as HarmCSP20/HassCSP20, HarmCSP21/ HassCSP21, HarmCSP22/HassOBP22, HarmCSP23/HassCSP23, HarmCSP24, HarmCSP25, HarmCSP26, HassCSP24 and HassCSP25 following the numbers assigned to previously reported CSPs (Fig 4). All these novel genes were deposited in the GeneBank: HarmCSP20-26 (GeneBank accession numbers: KY810184, KY810185, KY810186, KY810187, KY810188, KY815026, KY815027), HassCSP20-25 (GeneBank accession numbers: KY810189, KY810190, KY810191, KY810192, KY810193, KY810194).

Expression of the OBPs and CSPs in larva and adult H. armigera
To better understand the functional role of OBPs and CSPs in larval olfactory systems, we investigated the expression patterns of all candidate HarmOBPs and HarmCSPs via semiquantitative reverse transcription PCR. The tissues used were larval antenna, larval mouthpart, adult antenna and adult abdomen. The results reported in Fig 5 show   The conserved cysteine residues were marked with "5". All these OBPs were assignment into CRLBP with six conserved cysteine residues, Minus-C with four conserved cysteine residues and Plus-C with more than more than six conserved cysteine residues. https://doi.org/10.1371/journal.pone.0179243.g001 larval and adult tissues. Compared to OBPs, CSPs were more expressed in non-olfactory tissues suggesting diverse functions. Eight of them showed similar expression levels in all tissues, while the others were specifically detected in olfactory organs. In particular, four genes (HarmCSP20, 22, 23 and 24) were specific of larval olfactory tissues, one (HarmCSP14) was detected only in adult antenna, and three (HarmCSP7, HarmCSP15 and HarmCSP25) were found in both larval and adult olfactory organs with no significant differences.

Discussion
In Lepidoptera, the main tasks of adults are reproduction and species dispersal. To accomplish them they use a sophisticated olfactory system for correct mating and oviposition on the suitable host plant [50,51]. Compared to adults, larvae show limited activity, their major tasks being feeding, growing and accumulating energy [52,53]. Therefore larvae are expected to harbor a simpler olfactory system than adults. One of the characteristics of monophagous insects is the strict specificity to their host, a typical example being the specialization of M. sexta for Solanaceous plants [32]. In this case, the mother choses the host plant while ovipositing and larvae may not need to move away through their life [54,55]. In contrast, larvae of polyphagous speciesoften ignore their mother's choices, disperse actively, and often move between different host plants for feeding [56,57]. For example, sometimes larvae need to abandon their prior host and select another one, because the plant resources are exhausted, or because of competition with other herbivores, or else because the plant has become infected [58][59][60][61]. Such differences in foraging behaviors are genetically determined [56]. H. armigera and H. assulta are two closely related species both representing serious pests in China and other countries. H. armigera is a polyphagous insect which attacks about 180 species Candidate OBPs and CSPs in the larval chemosensory tissues of Helicoverpa armigera and H. assulta of plants [62], while H. assulta is oligophagous, mainly feeding on tobacco [63]. In both species antennae and mouthparts are the main chemosensory structures guiding the larvae to their host plants. Thus, a study of larval antennae and mouthparts at the molecular level can provide useful information for larva-based pest control.
In this work, we focused on two families of soluble protein OBPs and CSPs which play some roles in the interactions between odorant molecules and olfactory receptors. We identified a total of 26 OBPs and 21 CSPs in the larval chemosensory tissues of H. armigera as well as 27 OBPs and 20 CSPs in H. assulta. Combined with the data available for in adult antennae, the total number of OBP genes identified in H. armigera and H. assulta are 38 and 35 respectively. These numbers are lower, although in the same order, than those reported for other species (46 in B. mori) [45]. The total number of CSP genes identified in H. armigera (25) and H. assulta (23) are also in the same order of magnitude as in other species such as B. mori (21), and S. littoralis (23) [39,45].
For most of HarmOBPs and HarmCSPs we could find homologue genes in H. assulta. The high similarities in sequence between pairs of orthologous genes suggest that H. armigera and H. assulta larvae detect similar volatile substances. This idea is supported by the observation that often mixed populations of the two species are present on tobacco and some solanaceous plants [63]. However, for some genes we could not find orthologs in the sister species. This fact, if confirmed, could suggest that during evolution, the two species can have developed some unique characteristics in their chemosensory systems to become adapted to different ecosystems. For nearly half of the HarmCSPs, we detected expression in non-olfactory organ, such as adult abdomen, suggesting roles different from chemosensing. Similarly, in other species, some CSPs were found to be expressed in non-olfactory tissues, such as the pheromone glands, where they likely assist delivery of semiochemicals in the environment [64][65][66][67], or in reproductive organs, with putative roles in egg and embryo development [68,69]. Most of OBPs and CSPs are expressed both in adults and in larvae chemosensory organs, suggesting some common olfactory related behaviors. In particular, the gene encoding GOBP2 is expressed in larval antenna, where it might bind pheromone cues. Such hypothesis was originated from what was observed in Plutella xylostella [70]. However, for all PBP genes we could not find their expression in H. armigera larval tissues. This case, although being inconsistent with what was observed in S. littoralis [53], was common in other species. We also found three OBPs and six CSPs presenting larva-specific expression, suggesting that they may be involved in larval-foraging behaviors. Three OBPs and ten CSPs were found to be expressed more in larval antennae than in mouthparts, whereas the other proteins were only detected in larval mouthparts, suggesting that these genes may be involved in taste. Our results contribute to a better understanding of the chemoreception mechanisms of larvae at the molecular level and might help the development of larva-targeted strategies for population control in these two important agricultural pests.