Functional Characterization of Chemosensory Proteins in the Scarab Beetle, Holotrichia oblita Faldermann (Coleoptera: Scarabaeida)

Chemosensory proteins (CSPs) play important roles in chemical communication by insects, as they recognize and transport environmental chemical signals to receptors within sensilla. In this study, we identified HoblCSP1 and HoblCSP2 from a cDNA library of Holotrichia oblita antennae, successfully expressed them in E. coli and purified them by Ni ion affinity chromatography. We then measured the ligand-binding specificities of HoblCSP1 and HoblCSP2 to 50 selected ligands in a competitive binding assay. These results demonstrated that HoblCSP1 and HoblCSP2 have similar ligand-binding spectra. Both proteins displayed the highest affinity for β-ionone, α-ionone and cinnamaldehyde, indicating that they prefer binding to odorants other than sex pheromones. Additionally, immuno-localization revealed that HoblCSP1 is highly concentrated in sensilla basiconica, while HoblCSP2 is specifically localized to sensilla placodea. In conclusion, HoblCSP1 and HoblCSP2 are responsible for binding to general odorants with slightly different specificities due to their different in vivo environments.


Introduction
The scarab beetle, Holotrichia oblita Faldermann (Coleoptera: Scarabaeida), is a prominent underground pest that causes great economic loss throughout its life cycle [1]. This insect is a polyphagous pest that feeds on a range of plants, such as peanut, soybean, wheat, and potato [2]. Studies have demonstrated that H. oblita responds differentially to different plant leaf odors, indicating that H. oblita can find hosts using plant volatiles as cues [3][4][5].
Insects develop an elaborate, sensitive, and specific olfactory system to perceive chemical signals in the environment. This system confers the capacity to communicate with mates, locate foods, oviposit, and avoid natural enemies [3,6,7]. The insect olfactory system is primarily composed of an antennae lobe in the brain and morphologically distinct sensilla [8]. Most sensilla locate in antennae and/or maxillary palps, which are rich in olfactory receptor neurons (ORN). Odorants pass through a specific channel in the cuticle to the ORN lymph, where they stimulate the odorant receptors (ORs). After the ORs are activated, the odorants are degraded by odorant-degrading enzymes (ODEs) [9].
Chemosensory proteins (CSPs), also known as olfactory specific-D like (OS-D like) proteins or sensory appendage proteins (SAPs), make up one of the most important sensor protein groups in insect chemoreceptors [10]. Drosophila melanogaster OS-D protein [11] and A-10 [12] were the first two reported CSPs. CSPs are generally acidic, soluble proteins that are approximately 13 kDa with 100-115 amino acids. All CSPs contain 4 conserved cysteine residues, which form 2 disulfide bonds (S-S) with their neighboring sulfur side chains. Each of these S-S bonds forms a ring, one by linking 8 surrounding amino acids, the other by linking 4 amino acids [13,14]. Phylogenetic analysis of 180 CSPs from seven different insect orders demonstrated that CSPs are highly conserved with a N-terminal signature sequence, ''YTTKYDN [-VI][ND][LV]DEIL'' [15,16] and several a-helix domains in the secondary structures [17][18][19][20]. For example, Bombyx mori pheromone binding protein contains six a-helices, forming a cavity for binding pheromones [21]. The primary and secondary structures of CSPs are highly conserved across all insects [17,19,20]. This specific structure allows CSPs to interact with linear-chain compounds such as oleamide, which is an endogenous ligand of locust CSPs [22]. However, only a few three-dimensional CSP structures have been reported, including only Mamestra brassicae CSP-A6 (MbraCSP-A6) [23], Schistocerca gregaria CSP4 (SgrCSP4) [24], and B. mori CSP1 (BmorCSP1) [25].
CSPs exist in both male and female organisms. They are not only found in sensilla within antennae [13,26,27], proboscises [28], maxillary palps [29], labial palps [30], and tarsus [31], but they are also observed in the abdomen, truncus, cuticle [32], legs [33], wings [22], and pheromone glands [34]. This non-confined localization would allow them to function in a wide spectrum of signaling events [35]. Furthermore, the expression of CSPs in insect pheromone glands indicates that these molecules are involved in formation and transportation of sex pheromones, such as lipocalin in the uterus and/or saliva of mammals [19]. The expression of SgreCSP1-5 and MbraCSP6 in the sensillum lymph indicates that they function in odorant sensing in the same manner as odorant binding proteins (OBPs) [36]. Additionally, differential expression of CSPs throughout development indicates that CSPs functions are also regulated temporally [37]. For example, Apis mellifera CSP5 (AmelCSP5) was found in the embryonic ectoderm of the Italian honeybee, a result was thought to be relevant to the embryo casings' formation [38].
CSPs present in insects across different orders, including Diptera (4 CSPs in D. melanogaster, 8 CSPs in Anopheles gambiae [39]), Lepidoptera (10 CSPs in M. brassicae [28] and 16 CSPs in B. mori) and so on. Combining the recent discovery of 20 CSPs in Tribolium castaneum, these indicated the importance of CSPs and their potential as targets for pest control. Here, we constructed a cDNA library from antennae of H. oblita and found two CSPs, HoblCSP1 and HoblCSP2. Using a competitive binding assay with the fluorescent probe 1-NPN (N-phenyl-1-naphthylamine), we characterized the ligand-binding specificities and inspected the spatial localizations of these HoblCSPs as a means to improve our understanding of CSPs roles in insects.

Ethics statement
All animal experiments in this study were performed in strict accordance with the guidelines developed by the State Key Laboratory for Biology of Plant Diseases and Insect Pests, Institute of Plant Protection, and the Chinese Academy of Agricultural Science (IPP, CAAS). The protocol was approved by the committee on the Ethics of Animal Experiments of the IPP, CAAS. The Approval ID is SYXK (Beijing) and the Permit Number is 2008-008.

Insects
Adult scarab beetles (H. oblita) were collected from the Hefei Experimental Base of the Institute of Plant Protection, Anhui Academy of Agriculture Science, Hefei, Anhui Province, China, and maintained in the lab at 28uC under a standard photoperiod (L/D: 16 h/8 h). Antennae from adult females were excised and immediately frozen in liquid nitrogen.

Construction of an H. oblita antennae cDNA library
Total RNA from 100 female antennae was extracted with Trizol (Invitrogen, USA). Full-length double-stranded cDNA (ds cDNA) with blunt cDNA ends was synthesized and amplified using the Creator TM SMART TM cDNA Library Construction Kit (Clontech, USA). Synthesized ds cDNA was then incubated with 0.08 mg/ml proteinase K at 45uC for 20 min to inactivate the DNA polymerase. After size fractionation using CHROMA SPIN TM columns, the cDNA was incorporated into SfiI-digested lTripIE vector. The recombinant phage vector was transduced into E. coli XL1-Blue (TaKaRa Co., China). The plaques were counted to calculate the phage titer (pfu/ml), and the recombination efficiency was estimated by calculating the ratio of white (recombinant) to blue (non-recombinant) plaques. Fragments . 350 bp were sequenced.

Identification and sequence analysis of HoblCSP1 and HoblCSP2
Using contig alignment coupled with NCBI BLAST, HoblCSP1 and HoblCSP2 were identified from the antennae cDNA library. The full-length sequences of HoblCSP1 and HoblCSP2 were cloned and verified by fishing with sequence-specific primers. The primers for HoblCSP1 were: forward 59-GAAAAGAAAAACGA-TAACGAA-39 and reverse 59-CACAATTTTACGTTGGAA-GAT-39, while the primers for HoblCSP2 were: forward 59-AGATATACAACAAAATATGATAA-39 and reverse 59-CAATGTATGCAACAGTGTCCAAG-39. The signal peptides were predicted through SignalP 3.0 [40], and the molecular weights were calculated using the SWISS-PROT (ExPASy server) program ''Compute pI/Mw.'' The hydrophobicity of each predicted protein was analyzed at http://us.expasy.org/cgi-bin/ protscale.pl. And, BLAST and Mult-Alin were used for homology searches and the alignment of nucleotide and/or amino acid sequences.
The evolutionary history of insect CSPs was inferred by using the Maximum Likelihood method based on the JTT matrix-based model [41] in MEGA 6.06 [42]. Briefly, the CSP cDNAs were aligned in ClustalW2 [43] and the alignment was improved by removing the spurious sequences and poorly aligned regions in TrimAl [44] by setting the gap threshold to 0.25. The tree with the highest log likelihood (211948.7261) is shown. The analysis involved 109 amino acid sequences. All positions with less than 95% site coverage were eliminated. That is, fewer than 5% alignment gaps, missing data, and ambiguous bases were allowed at any position.

Prokaryotic expressions of HoblCSP1 and HoblCSP2
Recombinant pET30a(+)/CSP1 and pET30a(+)/CSP2 were generated by ligating the sticky ends of the designed HoblCSP1 and HoblCSP2 constructs into the expression vector pET30a (Novagen, Germany). pET30a(+)/CSP1 and pET30a(+)/CSP2 were then transformed into E. coli BL21 (DE3) pLysS cells. The expressions of recombinant HoblCSP1 and HoblCSP2 were induced for 4 h by 0.5 mM IPTG following a 3 h pre-incubation. The cells were harvested by centrifugation and then homogenized in phosphate-buffered saline (PBS, 0.04 M, pH 7.0). After centrifugation at 12,0006g for 20 min at 4uC, the supernatants were purified by Ni ion affinity chromatography (GE-Healthcare). Recombinant HoblCSP1 and HoblCSP2 were identified by western blot analysis with antibodies designed against the Histag (Abcam, USA). To prevent confounding effects in the subsequent experiments, the His-tag was removed by recombinant enterokinase (rEK) (Bio Basic Inc.) and NaCl was removed by dissolving the proteins in dH 2 O and filtering with a 10 kDa Amicon Ultra-0.5 Device (Millipore, USA). Finally the recombinant proteins were stored at 280uC, until required.

Fluorescence competition assays
Fluorescence binding assay was based on method previously described by Yin et al. [45]. Briefly, fifty compounds (Sigma-Aldrich, Germany) with chemical purities $97% were tested in a Lengguang 970 CRT spectrofluorimeter (Shanghai Jingmi, China). Assuming the proteins were 100% active, the binding affinities for N-phenyl-1-naphthylamine (1-NPN) were determined by adding aliquots of 1-NPN into a 2 mM protein solution for final concentrations of 2,24 mM. The dissociation constants of the binding competitors were calculated from IC 50 [46]. Binding data for each ligand was collected from 3 measurements.

Preparation of anti-HoblCSP1 and anti-HoblCSP2 antibodies
Purified full-length HoblCSP1 and HoblCSP2 were injected into New Zealand white rabbits following a standard immunization protocol for antibody production. Briefly, 100 mg of recombinant CSP was injected with an equal volume of Freund's complete adjuvant, followed by three additional injections of 500 mg, with one each on the 21st, 35th, and 49th day. The antiserum was then tested using an enzyme-linked immunosorbent assay (ELISA) and used without further purification. The pre-injected rabbit serum was used as a negative control. After incubation, the sections were washed in 2 times in PBGT then incubated with anti-rabbit IgG secondary antibody (1:20), coupled to 10-nm colloidal gold, for 90 min at room temperature. Gold granules were size-increased by silver intensification for 15 min in the dark, followed by incubation with 2% uranyl acetate for 15 min to increase the contrast. The samples were imaged by transmission electron microscopy (HitachiH-7500).

Characteristics of the HoblCSP1 and HoblCSP2 sequences
From the antennae cDNA library, full-length HoblCSP1 (GenBank: HQ683720) and HoblCSP2 (GenBank: HQ688991) genes were obtained and verified. The open reading frame (ORF) of HoblCSP1 contained 399 nucleotides, encoding 132 amino acids. The predicted molecular weight of HoblCSP1 was 15.55 kDa. The ORF of HoblCSP2 was comprised of 390 nucleotides, encoding 129 amino acids, and the predicted molecular weight was 14.78 kDa. At their N-termini, HoblCSP1 and HoblCSP2 contain signal peptides of 16 and 18 residues, respectively. Both proteins contained 4 conserved cystine residues (Fig. 1A, 1B), consistent with the model Cys-X 6-8 -Cys-X 16-21 -CysX 2-4 -Cys, and also contained a hydrophobic domain. The isoelectric points (PI) of HoblCSP1 and HoblCSP2 were dramatically different at 4.93 and 8.14, respectively. Phylogenetic analysis based on the deduced amino acid sequence also revealed that HoblCSP1 and HoblCSP2 are highly divergent, with HoblCSP1 and HoblCSP2 in two separated Coleopteran mono-phylogenetic groups ( Figure 1C).

Procaryotic expression of HoblCSP1 and HoblCSP2
The recombinant proteins pET30a(+)/CSP1 and pET30a(+)/ CSP2 were successfully expressed in BL21(DE3) PlysS cells. For both CSPs, a specific band less than 24 kDa (including His-tag) was observed by western blot analysis (Figure 2), which was consistent with the molecular weight deduced from their predicted amino acid sequences. HoblCSP1 and HoblCSP2 were purified at concentrations of 1.1 mg/ml and 1.2 mg/ml, respectively.
Binding specificities of HoblCSP1 and HoblCSP2 largely overlapped Based on the dissociation constants of CSP1/1-NPN (2.53 mM) and CSP2/1-NPN (2.93 mM) calculated from the binding curves ( Figure 3A), fifty potential odorant compounds were selected for a fluorescence competition assay with 1-NPN. These molecules included Ricinus communis leaf volatiles that attract H. oblita [47]; volatiles isolated from H. oblita host plant, Ulmus pumila Linnaeus [48,49]; putative sex pheromones from closely related beetles; and previously reported compounds ( Table 1). The inhibition constants Ki (for each CSP/ligand combination) are summarized in Table 1. The binding curves of a few representative fluorescence competition assays are presented in figure 3B. These binding curves, coupled with the Ki values, demonstrated that HoblCSP1 and HoblCSP2 displayed similar spectra of binding activity. Of the 50 selected compounds, HoblCSP1 preferred 22, while HoblCSP2 preferred 18. Within these groups, 15 compounds were covered in the binding spectra of both HoblCSP1 and HoblCSP2 (Table 1). Furthermore, both HoblCSP1 and HoblCSP2 bound most strongly to b-ionone, followed by a-ionone and cinnamaldehyde. Other ligands, however, were unique to each of the proteins. For example, HoblCSP1 was able to bind camphene, albeit with a high K i value, while HoblCSP2 could not bind camphene at all ( Figure 3B).

Spatial localizations of HoblCSP1 and HoblCSP2 in antennae of H. oblita
Generally, both HoblCSP1 and HoblCSP2 were found in the antenna of both male and female adult H. oblita. These two proteins were primarily distributed in the outer sensillum lymph, with different concentrations in different types of sensilla. HoblCSP1 was concentrated in sensilla placodea of both males and females ( Figure 4B, 4B', 4D and 4D'), and more highly concentrated in male sensilla basiconica ( Figure 4A, 4C). However, HoblCSP2 was found to be sensilla-specific, as it was highly expressed in sensilla placodea but rarely found in sensilla basiconica ( Figure 5).

Discussion
In this study, we characterized two CSPs, HoblCSP1 and HoblCSP2, from a H. oblita antennal cDNA library. Similar numbers of CSPs were also found in other beetles, like 3 CSPs in Batocera horsfieldi [50], 2 CSPs in Leptinotarsa decemlineata, and 1 CSP in Diaprepes abbreviatus [51]. In contrast, 12, 11, and 20 CSPs were found in Monchamus alternatus [52], Dendroctonus ponderosae [53], and T. casteneum [54] respectively. Different numbers of CSPs may be required to distinguish different host plants due to the varieties of their ligand-binding specificities. However, further investigations are required to verify this. HoblCSP1 and HoblCSP2 are highly divergent in deduced amino acid sequences. Additionally, phylogenetic analysis demonstrated that HoblCSP1 and HoblCSP2 are not in a recent monophylogeny group (Figure 1C), indicating that they diverged from each other a long time ago. Coupled with the markedly different PIs of HoblCSP1 (4.93) and HoblCSP2 (8.14), we predicted that HoblCSP1 and HoblCSP2 function differently. Surprisingly, we found that the binding spectra of HoblCSP1 and HoblCSP2 largely overlapped. Among the 50 selected compounds, b-ionone and its isoform a-ionone and cinnamaldehyde displayed the highest affinities to both HoblCSP1 and HoblCSP2. This is consistent with the result from the electroantennogram (EAG) examinations that showed a-ionone and cinnamaldehyde can elicit strong electrophysiological responses in H.oblita antennae [1]. ponderosae [53], 6 from Ips typographus [53], 8 from Diabrotica virgifera [51], 1 from D. abbreviates [51], 2 from L. decemlineata [51], 7 from M. alternatus [52], 7 from Dastarcus helophoroides [52], 3 from B. horsfieldi [50], 16 from B. mori [69], 3 from Heliothis virescens [70], 6 from Apis mellifera [16], 2 from Locusta migratoria [22], 5 from A. gambiae [57], 2 from D. melanogaster [15], 10 from Acyrthosiphon pisum [71]. Also, a previous study showed cinnamaldehyde displays a great attraction to H.oblita [55]. b-ionone and a-ionone are general plant volatiles [56] that are found in H.oblita's host plant, potato [2], while cinnamaldehyde is a main volatile from H.oblita's host plant, castor [2,55]. These three chemicals showed the best affinities to HoblCSPs, illustrating the underlying mechanism via which H.oblita recognizes their host plants. These chemicals would be the best candidates to distract H.oblita from recognizing their host plants, which in turn can be applied to successfully control them in the field. Also, HoblCSP1 and HoblCSP2 exhibited medium affinities to trans-2-hexenal, geraniol, myrcene, and benzaldehyde, and limited affinities to sex pheromones such as L-proline methyl ester, Glycine ethyl ester, and L-isoleucine methyl ester. Both proteins bound to very few alkanes, alcohols, and aldehydes.
These data demonstrated the discriminatory power of insect olfactory systems, with the ability to distinguish different isomers of the same compound [45]. Particularly, HoblCSP1 and HoblCSP2 preferred benzene rings in a ligand structure ( Figure 3B). These preferential binding affinities of HoblCSP1 and HoblCSP2 indicated that they play important roles in odorant binding beyond simply the sex pheromone response, although HoblCSP1 and HoblCSP2 are also found in locations other than the antennae (unpublished data).
However, HoblCSP1 and HoblCSP2 reserved unique binding affinities to other compounds. HoblCSP1 displayed higher affinities to aromatic compounds, including dimererhyl phthalate, eugenol, and methyl salicylate, whereas HoblCSP2 showed higher affinities to green leaf volatiles, such as cis-3-hexen-1-ol, cis-3hexen-1-ol, and 6-methyl-5-hepten-2-one. Since the binding affinities of HoblCSP1 and HoblCSP2 were tested under the same condition, it is therefore possible that they may display different binding activities due to their different in vivo environments. Interestingly, the homologous CSPs in different species could also have divergent affinities. In our experiments, the best ligand of HoblCSP1 is b-ionone, whereas the best ligands for its homologous AgamCSP3 are 2-pentylcinnamaldehyde, retinal, citronellal, and nonanal [57]. This could be due to the adaptation of insect olfactory systems to the specific odorant of their hosts [58][59][60][61].
H. oblita antennae are sexually dimorphic. The sensilla placodea and sensilla basiconica are the most common sensilla in the antennae of beetles [62]. The numbers of sensilla placodea and sensilla basiconica are approximately equal in females, whereas in males there are significantly more (9 times) sensilla basiconica than sensilla placodea. However, their functions remain unknown. The sensilla placodea, rather than sensilla basiconica, was proposed as the organ responsible for responding to sex pheromones in Popillia japonica and Anomala osakana [63]. In Anomala corpulenta, the sensilla diverged for different functions,   Table 1). All proteins used were diluted to a fixed concentration of 2 mM, while the concentration of 1-NPN varied with the respective dissociation constant of HoblCSPs/1-NPN. The mixed solution was then titrated with 1 mM of each competing ligand to final concentrations of 0-28 mM. Fluorescence intensities are plotted as percent of the initial fluorescence in the absence of ligands. The molecular formulas of representative ligands are shown here as well. The calculated dissociation constants for all of the ligands are listed in Table 1 with the umbilicate sensilla placodea responding to green leaf volatiles and the hidden sensilla placodea primarily interacting with sex pheromones [64]. In H. oblita, scientists have proposed that the sensilla basiconica is the sensillum used for sex pheromone responses [65] because females synthesize sex pheromones and males respond to them, most likely through CSPs in the antennae [66,67]. Here, we found that HoblCSP1 was highly concentrated in the sensilla basiconica, and HoblCSP2 was densely localized in the sensilla placodea. This result indicated that, due to its higher affinity to odorants, HoblCSP1 may confer sensilla basiconica the ability to respond to sex pheromones. Additionally, the localizations of HoblCSPs excluded the possibility that they function as homo or heterodimers, which is consistent with a previously published fluorescence competition assay [48]. CSPs closely relate to insect behavioral plasticity. Our work here discovered that HoblCSP1 and HoblCSP2 have specialized characteristics and unique localization patterns. These will assist in devising strategies to disrupt the aggregation of H. oblita. Also, 924 OBPs and 300 CSPs had been identified (UniProt) to date [68], especially up to 20 CSPs in a single species [54]. The ligand  binding overlap between different CSPs, and between CSPs and OBPs, point to intriguing questions regarding the evolution of insect olfactory systems and the underlying mechanisms of olfactory recognition.