Sex pheromone recognition and characterization of three pheromone-binding proteins in the legume pod borer, Maruca vitrata Fabricius (Lepidoptera: Crambidae)

Pheromone-binding proteins (PBPs) are essential for the filtering, binding and transporting of sex pheromones across sensillum lymph to membrane-associated pheromone receptors of moths. In this study, three novel PBP genes were expressed in Escherichia coli to examine their involvement in the sex pheromone perception of Maruca vitrata. Fluorescence binding experiments indicated that MvitPBP1-3 had strong binding affinities with four sex pheromones. Moreover, molecular docking results demonstrated that six amino acid residues of three MvitPBPs were involved in the binding of the sex pheromones. These results suggested that MvitPBP1-3 might play critical roles in the perception of female sex pheromones. Additionally, the binding capacity of MvitPBP3 with the host-plant floral volatiles was high and was similar to that of MvitGOBP2. Furthermore, sequence alignment and docking analysis showed that both MvitGOBP2 and MvitPBP3 possessed an identical key binding site (arginine, R130/R140) and a similar protein pocket structure around the binding cavity. Therefore, we hypothesized that MvitPBP3 and MvitGOBP2 might have synergistic roles in binding different volatile ligands. In combination, the use of synthetic sex pheromones and floral volatiles from host-plant may be used in the exploration for more efficient monitoring and integrated management strategies for the legume pod borer in the field.

RNA extraction, cloning and sequencing. We extracted total RNA from antennae and other tissues of male and female moths using an OMEGA E.Z.N.A TM Total RNA Kit (Omega, USA). A Prime Script first-strand cDNA synthesis kit (Invitrogen, USA) was used to synthesize the first-strand cDNA, following the manufacturer's protocols. MvitPBP1 (Genebank: KU517652), MvitPBP3 (Genebank: KU517653) and MvitPBP2 (Genebank: KU517654) genes were obtained from the National Center for Biotechnology Information (NCBI), and their open reading frames (ORF) were amplified by PCR with gene specific primers. The total PCR reaction mixture of 25 μ L contained 9.5 μ L of ddH 2 O, 1 μ L of sample cDNA, 1 μ L of forward primer (10 μ M), 1 μ L of reverse primer (10 μ M), and 12.5 μ L of rTaq mix DNA polymerase (Takara, Dalian, Liaoning, China). PCR reaction conditions were as follow: 94 °C for 3 min; 30 cycles at 94 °C for 30 s, at 55-60 °C for 30 s, and at 72 °C for 30 s; and 72 °C for 10 min. PCR products were inserted into T1 vectors (TransGen Biotech., China), following sequencing by Shanghai Sunny Biotechnology Co., Ltd. Sequence analysis. The cDNA sequences and deduced amino acid sequences of MvitPBP1-3 were analyzed using the online program BLAST (http://blast.ncbi.nlm.nih.gov/Blast.cgi) and the Expert Protein Analysis System (http://www.au.expasy.org/). Calculated molecular weights and predicted isoelectric points were obtained through ExPASy (http://web.expasy.org/compute_pi/). N-terminal signal peptides and most likely cleavage sites were predicted by the SignalP4.1 Server (http://www.cbs.dtu.dk/services/SignalP/). MEGA 6 and DNAMAN were used for multiple alignments and construction of the phylogenetic tree for MvitPBP1-3 with similar PBPs of other insect species.
Expression profiles of MvitPBPs. Tissue expression patterns of the three PBPs were assessed by real-time PCR with cDNA templates from different tissues of male and female moths. Total RNA was prepared in triplicate using TRIzol (Omega, USA), and the genomic DNA was digested with RNA-free DNase. Six primers (PBP1YF, PBP1YR, PBP2YF, PBP2YR, PBP3YF and PBP3YR, shown in Table 1) were used to determine the relative abundance of mRNA of the three PBP genes, with the actin gene used as the reference. Real-time PCR was performed  on a Bio-Rad CFX 96 real-time PCR system with SYBR Green I fluorescent dye. To check reproducibility, each real-time PCR reaction for each sample was conducted in three technical replicates and three biological replicates. Real-time PCR was conducted in 20 μ L reactions that contained 10 μ L of 2×TransStart Top Green qPCR SuperMix, 0.3 μ L of each primer, 2 μ L of sample cDNA and 7.4 μ L of ddH 2 O. The cycling conditions were as follow: 95 °C for 3 min; 40 cycles at 95 °C for 10 s and at 50 °C for 30 s; and melt curve at 65 °C to 95 °C for 5 s. The data were analyzed by the 2 −ΔCt method, and SigmaPlot 10.0 was used to draw the histogram.
Recombinant expression of MvitPBPs. The entire coding regions, without the signal peptide sequence, of MvitPBP1, MvitPBP2 and MvitPBP3 were subcloned into the NcoI/XhoI and EcoRI/XhoI sites of a PET32a (+ ) expression vector. BL21 (DE3) E. coli competent cells were transformed by heat shock and colonies were grown on LB ampicillin agar plates. A single positive clone was first identified and then grown in 6 mL of liquid LB with ampicillin overnight at 37 °C. The culture was diluted to 1:100 in fresh medium and cultured for 4 h at 37 °C until the OD value reached 0.6. IPTG was added to the culture with a final concentration of 0.3 mM, and then the culture was incubated at 30 °C for 6 h. After the incubation, the cells were collected by centrifugation (5000 rpm, 3 min) and dissolved in 1 × PBS buffer. The suspension was crushed by sonication and then separated into supernatant and sediment by centrifugation (5000 rpm, 3 min). Then, MvitPBPs were purified from the supernatant using Ni ion affinity chromatography (Thermo, USA), and enterokinase was used to remove the His-tag. The size and purity of MvitPBPs were verified by SDS-PAGE analysis, which were then stored at − 80 °C.
Fluorescence binding assays. Fluorescence binding activity was determined according to the method of Sun et al. 33 . Emission fluorescence spectra were recorded on a Hitachi F-4500 at 25 °C in a right angle configuration, with a 1 cm light path quartz cuvette and 5 nm slits for both excitation and emission. The protein was dissolved in 20 mM Tris-HCl buffer (pH 7.4) and ligands were added as 1 mM methanol solutions. To measure the affinity of the fluorescent ligand 1-NPN to recombinant target proteins, a 2 μ M protein solution in 20 mM Structural modeling and molecular docking. The presumed tertiary structures of MvitPBP1-3 and MvitGOBP2 were established using the SWISS-MODEL prediction algorithm (http://swissmodel.expasy.org/) and were displayed by PyMOL Viewer (http://www.pymol.org/). A templates search was performed on the RCSB Protein Data Bank (http://www.rcsb.org/pdb/home/home.do) using Position-Specific Iterated BLAST. After evaluating the fit between the sequences and each of the alternative 3-D models, the model with the highest score was chosen. The crystal structures of BmorGOBP2 (PDB: 2wck) and BmorPBP1 (PDB: 2p70) were used as templates to construct the three-dimensional structures of MvitPBP1-3 and MvitGOBP2. Alignment was conducted with DNAMAN and ESPript (http://espript.ibcp.fr/ESPript/ESPript/). Based on the established homology model,  the docking program AutoDock Vina was used to find the potential binding mode between MvitPBP1-3 and MvitGOBP2 and the ligand. The 3-D structure of the ligand was obtained from ZINC (http://zinc.docking.org/).

Results
Cloning and sequence analysis of MvitPBPs. Full  These results demonstrated that PBPs were highly conserved among diverse species of Lepidoptera. Sequences from MvitPBP1-3 and PBPs from other insects were used to construct the phylogenetic tree to assess the evolutionary relationships among the proteins. As shown in Fig. 2, MvitPBP1 was first clustered with CsupPBP1 in the phylogenetic tree, which was consistent with the highest sequence similarity between them. Then, MvitPBP2 was clustered with MsexPBP2 (AF117588.1), AipsPBP2 (JQ822241.1), HarmPBP2 (HQ436360.1), and HassPBP2 (EU316186.2), and MvitPBP3 was clustered with DplePBP3 (EHJ71308.1), BmorPBP3 (AM403101.1), MsexPBP3 (AF117580.1), HarmPBP3 (AF527054.1), HassPBP3 (DQ286414.1), AipsPBP3 (JQ822242.1) and SinfPBP3 (AEQ30020.1); thus, two large branches were formed with some of the PBP2 and PBP3 proteins from subfamilies of Lepidoptera. The three MvitPBPs were clearly separated from one another and were assigned to different subgroups, which illustrated that the three genes were highly conserved in their evolution within this family.
Expression patterns of MvitPBPs. Quantitative real-time PCR was used to measure the transcript abundance of the three MvitPBP genes. Transcript abundance for each PBP was determined for multiple tissues (including antenna, head, thorax, abdomen, leg and wing) from M. vitrata. As shown in Fig. 3, MvitPBP1-3 transcripts were more highly expressed in antennae than in other tissues, but the expression was not specific to  antennae, with low expression also detected in wing and head tissues (without antennae). Moreover, the expression of the three MvitPBP genes was sex-biased, and MvitPBP1 was specifically expressed in male antennae with   soluble, and after expression was induced by 0.3 mM IPTG, the proteins were purified with Ni-NTA resin after ultrasonication (Fig. 4). The target proteins underwent two rounds of purification: the first round purified the recombinant protein from the total protein, and then the His-tag of the recombinant MvitPBP1-3 was removed by enterokinase. The SDS-PAGE results were consistent with the expected sizes of the MvitPBPs. Recombinant MvitPBPs were stored at −80 °C until used in the binding experiment.
Fluorescence binding affinities. The purified target proteins were used to illustrate the binding specificity of MvitPBPs using 1-NPN as the fluorescence probe in competitive binding assays, which displayed a strong blue shift in fluorescence intensity when bound to MvitPBP1-3 (Fig. S1). The binding curves and Scatchard plots indicated that the binding of the fluorescent ligand to each of the three PBPs increased with increasing concentrations of the 1-NPN (Fig. 5A).  Table 2). MvitPBP1 and    (Fig. 5D). Additionally, MvitPBP1-3 was tested in competitive binding assays with seventeen synthetic ligands from floral volatile chemicals that elicited obvious electroantennogram responses. Shown in Table 3 are the IC 50 values and the calculated inhibition constants (Ki), when possible, for each MvitPBP/ligand combination. The binding abilities of most of the tested volatiles to MvitPBP1, MvitPBP2 and MvitPBP3 were different (Fig. 6A-L). MvitPBP3 had the highest binding affinities with butanoic acid octyl ester and 2-methyl-3-phenylpropanal of the floral volatile components, with Ki values of 7.64 and 8.35 μ M, respectively (Table 3). Therefore, through binding with butanoic acid octyl ester and 2-methyl-3-phenylpropanal from the floral volatiles, MvitPBP3 might have a significant role in enhancing odorant signal transduction for host plant recognition in M. vitrata, which was similar to the role of MvitGOBP2.

Structural modeling and molecular docking of MvitPBPs and MvitGOBPs with different ligands.
To further investigate the binding mode and potency of MvitPBPs and MvitGOBP2 with the different tested ligands, structural modeling and molecular docking analyses were used to calculate the optimized conformation and potential key binding sites. All sex pheromone and important floral odor molecules that exhibited high binding capacities were docked into the binding cavity of MvitPBPs and MvitGOBP2. Based on the crystal structures of BmorPBP1, the structures of MvitPBP1, MvitPBP2 and MvitPBP3 were modeled. The constructed 3-D  structures contained six α -helices and an additional α -helix (helix α 7) (Fig. 7), which may be involved in binding and release of sex pheromones. The framework of helices was stabilized by the three disulfide bridges: for MvitPBP1, Cys40-Cys75, connecting helices α 1 and α 3, Cys71-Cys129, between α 3 and α 6, and Cys118-Cys138, between α 5 and α 6; for MvitPBP2, Cys42-Cys77, connecting helices α 1 and α 3, Cys73-Cys132, between α 3 and α 6, and Cys120-Cys140, between α 5 and α 6; and for MvitPBP3, Cys49-Cys84, connecting helices α 1 and α 3, Cys80-Cys138, between α 3 and α 6, and Cys127-Cys147, between α 5 and α 6 (Fig. 8). Additionally, some key residues were observed on the 3-D models of the three MvitPBPs, including a histidine (His) involved in pH-dependent conformational change 3 .
As shown in Fig. 9A, the amino acid sequence of MvitGOBP2 was compared with the templates of BmorGOBP2, and based on alignment analysis, the similarity of their sequences was high. Based on the crystal structures of BmorGOBP2 34 , the 3-D structure of MvitGOBP2 was also constructed by SwissModel. Some key hydrophobic and hydrophilic residues were in the amino acid sequence of MvitGOBP2, which might be involved in the binding and release of various floral volatile ligands. MvitGOBP2 was composed of six α -helices and an additional α -helix structure formed by residues 2-15 (α 1a), 16-24 (α 2), 46-58 (α 3), 70-79 (α 4), 83-101 (α 5), 107-122 (α 6), and 131-138 (α 7) (Fig. 9B). Moreover, three disulfide bridges for stabilizing the framework of helices and the histidine (His) involved in pH-dependent conformational change are marked in the 3-D structure of MvitGOBP2 in Fig. 9B.
Additionally  Table 4). The key binding sites of MvitPBP3 and MvitGOBP2 with butanoic acid octyl ester were alanine 130 (A130) and arginine 130 (R130), respectively. As shown in Figs 11 and 12 and Table 4, MvitPBP3 and MvitGOBP2 had analogous special protein structures while sharing the identical binding sites (arginine, R140/R130) for 2-methyl-3-phenylpropanal from floral volatiles of the host plant.

Discussion
Odorant-binding proteins (OBPs), chemosensory proteins (CSPs), chemosensory receptors (odorant receptors, ORs; ionotropic receptors, IRs), odorant-degrading enzymes (ODEs) and sensory neuron membrane proteins (SNMPs) are primarily involved in the transduction process of insect olfactory chemical signals [35][36] . The initial molecular interactions for chemical signals (semiochemicals) such as sex pheromones and host odors are with odorant-binding proteins (OBPs), which likely ferry the semiochemical molecules across the antennal sensillum lymph to the olfactory receptors 34 . Subsequently, the odorant molecules are rapidly degraded by ODEs, and the chemical signals are converted into electrophysiological signals to complete pheromone conduction 37 . In Lepidoptera, the odorant-binding proteins are classified into pheromone-binding proteins (PBPs) and general odorant-binding proteins (GOBPs) based on primary sequence homology 38,39 . Moreover, PBPs and GOBPs are highly expressed in various types of olfactory sensilla on insect antennae and play different roles in recognizing sex pheromones and volatile odorants of the host. Therefore, with the identification and functional analysis of pheromone-binding proteins in M. vitrata in this study, new methods can be developed for controlling this pest by interfering with olfactory perception and subsequent mating behaviors.
qRT-PCR analysis indicated that MvitPBP1-3 genes were involved in odorant (including sex pheromones) detection because these genes were primarily expressed in the antennae of both sexes, and the level of expression was very low in other tissues such as the head, thorax, abdomen, leg and wing. MvitPBP1 gene was more abundantly expressed in male antennae than in female antennae, and a similar pattern of expression is reported in many other lepidopterans, including B. mori, Agrotis segetum, H. armigera, Heliothis virescens and Spodoptera exigua [40][41][42][43][44][45][46] . Notably, the expression level of MvitPBP2 and MvitPBP3 genes was much lower in male antennae than that in female moth antennae, which is similar to the expression of AipsPBP2-3 8 . Moreover, based on the higher expression of MvitPBP1 gene in male M. vitrata than that of the other PBP genes, MvitPBP1 gene might have a major role in male-female recognition. Furthermore, female-biased expression of both of MvitPBP2 and MvitPBP3 genes might indicate involvement in the autodetection of sex pheromone compounds, which has been demonstrated in other lepidopterans [47][48][49][50] .
The legume pod borer M. vitrata is a serious pantropical insect pest of grain legumes such as cowpea (V. unguiculata), pigeon pea (Cajanus cajan) and common bean (Phaseolus vulgaris) [51][52][53][54] . Because of the economic damage caused by this pest, insecticides and sex pheromone components have been tested as different control strategies in the fields of southern China. Lu et al. reported that a blend of E10, E12-16: Ald; E10-16: Ald; and E10 E12-16: OH (ratio = 100:5:5) attracted significantly more males than any other bait in a field test, including the primary component alone, a two component blend or virgin females 31 . In our ligand binding experiments, the four sex pheromone components of M. vitrata strongly bonded with MvitPBP1-3, with different levels of sensitivity. MvitPBP1 and MvitPBP3 were the most sensitive to E10E12-16: Ald, whereas MvitPBP2 was the most sensitive to E10-16: Ald. Both of these organic aldehyde compounds are reported to be the primary sex pheromone components of M. vitrata 28,29 , and MvitPBPs showed excellent binding affinities for these two components. Moreover, the binding of E10E12-16: Ald with MvitPBP1 and MvitPBP3 was significantly stronger than that with E10-16: Ald. To explain the differences in binding capacity with MvitPBP proteins, E10, E12-16: Ald has one more unsaturated bond than E10-16: Ald, which is consistent with the binding of PxylPBPs 46 .
PBPs play major roles in sex pheromone perception by binding and transporting hydrophobic pheromone molecules across the aqueous sensillar lymph to the olfactory receptors and by discriminating different semiochemicals including plant volatiles 55 . In our previous study, 17 electroantennogram-active compounds were identified from floral volatiles of V. unguiculata by GC-MS and GC-EAD 25 . Based on the fluorescence binding experiments, MvitPBP1-2 had very weak ligand binding capacities with all floral volatiles from the host plant V. unguiculata, which were much lower than those for the sex pheromone. However, MvitPBP3 displayed higher binding capacities with partial floral volatile components than those of MvitPBP1-2. Among the 17 tested compounds, the binding activity of butanoic acid octyl ester to MvitPBP3 was the strongest and displaced half the 1-NPN from the MvitPBP3/1-NPN complex at a ligand concentration of 20 mM (Table 3), although the abundance of this ester was low and the EAG response was weaker than that of other plant volatiles 25 . Additionally, 2-methyl-3-phenylpropanal bonded strongly with MvitPBP3, which was the most abundant compound in the floral volatiles of the host plant and elicited a high electrophysiological response from antennae of M. vitrata 26 . MvitGOBP2 also had high binding affinities with butanoic acid octyl ester and 2-methyl-3-phenylpropanal among the floral volatile components. When the concentration of butanoic acid octyl ester and 2-methyl-3-phenylpropanal reached 8.5 and 3.81 μ M, respectively, the fluorescence intensity of the MvitGOBP2/1-NPN complex rapidly decreased to approximately 50% 26 . Based on these results, MvitPBP3 and MvitGOBP2 might be derived from the same olfactory protein family because they shared similar amino acid binding sites with the identical volatile ligands.
In this study, we found that the three MvitPBPs had different energy values and key residues that interacted with ligands, with six amino acid residues of MvitPBP1-3 involved in binding a sex pheromone within a cavity. The different binding affinities of the three MvitPBPs toward the tested sex pheromone ligands was an indication of their sequence and structural differences. These results provide further support for the results from a previous field application of sex pheromones for pest population monitoring of M. vitrata [29][30][31] . Notably, based on the docking analysis and fluorescence binding assays, MvitPBP3 and MvitGOBP2 had similar binding energies and excellent binding capacities with the identical volatile ligand (2-methyl-3-phenylpropanal). Additionally, based on sequences alignment analysis, these two olfactory proteins had the identical amino acid residues (R140/R130) and similar protein structures around the binding cavity (Table 4, Fig. 12). These results indicated that arginine (R140/R130) might be a key binding site involved in the initial recognition of volatile ligands. Many previous studies report that PBPs and general odorant-binding proteins (GOBPs) are in different subfamilies of OBPs that are involved in the recognition and transport of pheromones and host odors through the lymph of chemosensilla 3,13 . Therefore, we speculate that MvitPBP3 and MvitGOBP2 may play a synergistic role in binding different types of floral volatile ligands with a high affinity, which highlights that MvitPBP3 and MvitGOBP2 are likely involved in the functional differentiation of odorant-binding protein family of M. vitrata.
Scientific RepoRts | 6:34484 | DOI: 10.1038/srep34484 Conclusion Pheromone-binding proteins are important components of insect olfactory systems, and sensitive olfaction is vital for recognition of hosts, mating and oviposition in insects 21 . In this study, we provided more evidence that MvitPBPs had excellent binding affinities with the sex pheromones and that partial floral volatiles from the host plant were key ligands of MvitPBP3. The identification and functional analysis of pheromone-binding proteins in M. vitrata will lead to new methods for controlling this pest by interfering with their olfactory perception. Moreover, molecular docking results for MvitPBPs and MvitGOBPs will contribute to the understanding of the multiple roles and synergistic effects of these proteins in the host seeking, oviposition and mating behavior of adult moths. Thus, synthetic sex pheromones of M. vitrata and two types of key floral volatiles from the host plant V. unguiculata may be used in the exploration for more efficient monitoring and integrated management strategies for the legume pod borer in the field.