Discrimination of Oviposition Deterrent Volatile β-Ionone by Odorant-Binding Proteins 1 and 4 in the Whitefly Bemisia tabaci

The whitefly, Bemisia tabaci, is an important invasive economic pest of agricultural crops worldwide. β-ionone has a significant oviposition repellent effect against B. tabaci, but the olfactory molecular mechanism of this insect for recognizing β-ionone is unclear. To clarify the binding properties of odorant-binding proteins (OBPs) with β-ionone, we performed gene cloning, evolution analysis, bacterial expression, fluorescence competitive binding assay, and molecular docking to study the binding function of OBP1 and OBP4 on β-ionone. The results showed that after the OBP1 and OBP4 proteins were recombined, the compound β-ionone exhibited a reduction in the fluorescence binding affinity to <50%, with a dissociation constant of 5.15 and 3.62 μM for OBP1 and OBP4, respectively. Our data indicate that β-ionone has high affinity for OBP1 and OBP4, which play a crucial role in the identification of oviposition sites in B. tabaci. The findings of this study suggest that whiteflies employ β-ionone compound in the selection of the suitable egg-laying sites on host plants during the oviposition behavior.


Introduction
Insects have many chemosensory organs, including antennae and other parts of the body, that play a key role in communication, such as in host-plant recognition. These chemosensory organs enable insects to successfully complete a series of important functions such as foraging, mating, oviposition, and avoidance. These behaviors contribute to insect survival in a complex environment [1,2]. Various types of proteins are produced by insect chemosensory organs, which include odorant-binding proteins (OBPs), chemosensory proteins (CSPs), olfactory receptors [3], odorant-degrading enzymes (ODES), and sensory neuron membrane proteins (SNMPs). OBPs are the most extensively studied chemosensory proteins. The first OBP was discovered in the 1980s in the male moth antenna of Antheraea polyphemus [3], which then subsequently prompted entomologists to study OBP function.
It will provide a better understanding of host preferences and repellent effects against B. tabaci, as well as potential schemes in controlling further spread of this invasive pest.

Insect Rearing
The B. tabaci Q type was established in the greenhouse of the Beijing Academy of Agricultural and Forestry Sciences (Beijing, China). The host plant was the tomato variety Xianke 8. The feeding conditions include a photoperiod light:dark = 16:8 h, temperature 25-28 • C, and relative humidity ranging from 60 to 80%.

Cloning and Sequence Analysis of OBP1 and OBP4
Total RNA was extracted from about 200 heads of whitefly using TRIzol (ThermoFisher Scientific-Life Technologies, Carlsbad, CA, USA). The first strand cDNA was synthesized using a PrimeScript RT reagent kit (TaKaRa, Dalian, China) with a gDNA Eraser, 2 µL of DNA eraser buffer, 1 µL of gDNA eraser, 1 µg of RNA, RNase-free dH 2 O to 10 µL, 42 • C for 2 min, then placed on ice, plus 1 µL of PrimeScript RT enzyme mix, 1 µL of RT primer mix, 4 µL of PrimerScript buffer, and 4 µL RNase-free dH 2 O. The reaction conditions were as follows: 37 • C for 15 min and 85 • C for 5 s. The synthesized cDNA was stored at -20 • C until use.
Each PCR reaction consisted of the following: 10 µL of the 5× PrimeSTAR Buffer, 4 µL of the dNTP mixture, 1 µL of each of the forward and reverse primers, 0.5 µL of cDNA, 0.5 µL of PrimeSTAR HS DNA polymerase (TaKaRa), and 33 µL of ddH 2 O. The PCR program was an initial denaturation at 94 • C for 3 min; followed by 30 cycles of 94 • C for 15 s, 60 • C for 5 s, and 72 • C for 1 min; then 72 • C for 10 min. The PCR products were detected by 1% agarose gel electrophoresis, and the target fragment was recovered and ligated into a pMD18-T cloning vector (TaKaRa, Dalian, China) and identified by PCR.

Bacterial Expression and Protein Purification of Recombinant OBP1 and OBP4
The positive clones of OBP1 and OBP4 genes were digested with NdeI and XhoI simultaneously with a pET-30a (+) expression vector, ligated with T4 DNA ligase, and transformed into Trans 5α competent cells. After PCR and double restriction enzyme digestion, the positive plasmids were transformed into Escherichia coli BL21 (DE3) competent cells. The positive clones were selected for culture to an OD 600 value of 0.4−0.6, and isopropyl-β-D-thiogalactopyranoside (IPTG) was added at a final concentration of 1 mmol·L −1 . Protein expression was induced at 28 • C and 180 rpm for 10 h. The cells were collected by centrifugation, and the bacterial lysate was added for ultrasonication (sonicated for 30 min, worked for 3 s, stopped for 5 s). After crushing, the supernatant and the precipitate were collected by centrifugation, and protein expression was detected by SDS-PAGE.
The protein supernatant was added to a HisTrap affinity chromatography column (TaKaRa, Dalian, China) and purified by gradient imidazole buffer (0, 20, 50, 100, 250, 500 mmol·L −1 ), and the target protein eluate was eluted at 50 mM Tris-HCl (pH 7.4) and dialyzed for 24 h. Protein detection was performed by SDS-PAGE, and protein concentrations were measured according to instructions of the bicinchoninic acid (BCA) Protein Assay Kit (ThermoFisher Scientific-Life Technologies, Carlsbad, CA, USA).

Ligand Binding Experiments of OBP1 and OBP4
N-Phenyl-1-naphthylamine (1-NPN), 4,4 -Dianilino-1,1 -binaphthyl-5,5 -disulfonic acid dipotassium salt (bis-ANS), β-ionone, α-ionone, and hydroxy-β-ionone were diluted with chromatographic-grade methanol to a final concentration of 1 mmol·L −1 , and the protein was diluted with 50 mM Tris-HCl (pH 7.4) to a final concentration of 2 µmol·L −1 . When 1-NPN was used as fluorescent probe, the excitation wavelength of the fluorescence spectrophotometer was set to 337 nm, and the scanning range was 350−500 nm. When bis-ANS was used as fluorescent probe [33,34], the excitation wavelength of the fluorescence spectrophotometer was set to 295 nm, and the scanning range was 300−550 nm. The fluorescent probe was gradually added to the protein-containing quartz tube to make it close to saturation. According to the Scatchard equation [35], the binding dissociation constants K 1-NPN of OBP1 and the fluorescent probe 1-NPN and K bis-ANS of OBP4 and fluorescent probe bis-ANS were determined. Protein and fluorescent probes were added to the quartz dish (the final concentration was 2 µM), and ionone was gradually added to determine the binding ability of the protein and ionone, and the dissociation constant K i was calculated using the formula:

Molecular Docking
After searching the current Protein Data Bank (PDB) (http://www.rcsb.org) using the amino acid sequence of OBP1 and OBP4 as a probe, the crystalline structures of a vetch aphid Megoura viciae OBP3 (PDB ID: 4Z39, Chain A, resolution 1.3 Å) [36] and currant-lettuce aphid Nasonovia ribisnigri OBP3 (PDB ID: 4Z45, Chain A, resolution 2.02 Å) [36] were selected as templates. Sequence alignment of the target and template were then conducted, and three initial models were built using MODELLER [37]. Ramachandran plot and Profile-3D [23] was used to estimate the structural alignment of the models. Ramachandran plot was generated by ProCheck [38]. Protein preparation were conducted using the Clean Protein module to correct problems such as incomplete residues, alternate conformations and modified hydrogens. In addition, the small molecule was prepared using the minimize ligand module. CDOCKER was used to analyze the interaction energy between OBP and β-ionone based on CHARMm [37]. All analyses were conducted in Discovery Studio (Accelrys, San Diego, CA, USA) version 4.0.

Gene Cloning and Sequence Analysis
After total RNA extraction, cDNA synthesis, PCR amplification, and sequencing, the OBP1 and OBP4 gene sequences of B. tabaci were obtained, which were consistent with the OBP sequences in our previous reports. The open reading frame (ORF) is 429 bp in length, with predicted signal peptides consisting of 19 and 24 amino acids starting from the N-terminus. B. tabaci OBP1 and OBP4 have six conserved cysteine sites ( Figure 2). Sequence alignment and phylogenetic tree construction of B. tabaci OBP1, OBP4, and the OBPs of other insects show that Q-type B. tabaci OBP1 is closely related to B-type B. tabaci OBP1. The protein sequence identity of OBP1 between B and Q species was as high as 99%, with a score of 292, and an E-value of 2 × 10 −85 . OBP4 has a sequence identity of 96% between the two species, a score of 275, and an E-value of 7 × 10 −79 . Based on the phylogenetic tree, Q-type B. tabaci OBP4 and B-type B. tabaci OBP4 showed the closest phylogenetic relationship (Figure 3). B. tabaci OBP1 and LstiOBP2 from Loxostege sticticalis, HoblOBP2 from Holotrichia oblita in a large clan, and LstiOBP2, HoblOBP2 have been confirmed to bind β-ionone. In total, eight conservative motifs were found after MEME motif analysis ( Figure 4). Among 13 OBPs binding β-ionone, only motif1 exists in all OBP genes, and HoblOBP2, LstiOBP2, AcerOBP2 and AlinOBP1 only have motif1. This result indicates that motif1 may play an important role in binding to β-ionone.   Table S1.

Binding Characteristics of OBP1 and OBP4
The recombinant plasmids of pET30a-OBP1 and pET30a-OBP4 were constructed and transformed into BL21 (DE3) competent cells, and protein expression was induced at 1 mmol·L −1 IPTG, 28 • C, and 180 rpm. The target protein band was obtained after passing through a Ni-NTA column ( Figure S1), and protein concentration was measured using the BCA method [21] after dialysis for 24 h.
The binding characteristics of OBP1 and fluorescent probe 1-NPN were detected by molecular fluorescence spectrometry ( Figure 5A). β-ionone was added to a mixture of OBP1 and 1-NPN at a concentration of 2 µM, and a fluorescence competition binding experiment was performed. The fluorescence intensity was changed with the concentration of the ligand and the IC 50 was calculated, and the dissociation constant K i was calculated according to the formula (Table S2). The β-ionone was tested to compete for fluorescence intensity below half of the initial value ( Figure 5C), with IC 50 of 7.16 µM and a dissociation constant K i of 5.15 µM. In the same manner, β-ionone was added to a mixture of OBP4 and bis-ANS at a concentration of 2 µM to perform a fluorescence competition binding experiment, changes in fluorescence intensity based on ligand concentration were recorded, IC 50 was calculated, and the dissociation constant K i was calculated according to the formula. After testing, β-ionone was able to compete for fluorescence intensity below half of the initial value ( Figure 5D), with IC 50 of 5.14 µM and a dissociation constant K i of 3.62 µM. We also tested the binding ability of two analogs of β-ionone, including α-ionone and dihydro-β-ionone, to OBP1 and OBP4. α-Ionone showed weaker binding to both OBP1 and OBP4. Dihydro-β-ionone hardly displaced the fluorescent probe from both complexes. Though the linear regressions of bis-ANS with OBP1 and 1-NPN with OBP4 were far from ideal, we used them as fluorescent probes, respectively, to test the binding ability of ionone to OBP1 and OBP4. β-ionone bonded well with OBP1 and OBP4, while α-ionone and dihydro-β-ionone did not ( Figure S2C,D).

Homology Modeling and Molecular Docking
Using homology modeling, OBP1 shared 35% and 34% sequence identity with templates 4Z39 and 4Z45 ( Figure S3). However, for OBP4, no template shared a sequence similarity of >30%, and thus, we only used OBP1 for homology modeling and molecular docking analysis. Based on the results of the Ramachandran plot ( Figure S4), 96.3% of all residues were in the favored regions, with no residues in disallowed regions. In addition, the verified score of the OBP1 model using 3D-Profile was 43.12, with verified expected high and low scores of 53.29 and 23.92, respectively. CDOCKER analysis indicated an interaction energy of 19.2217 kcal/mol, and the 2D ligand interaction diagram is shown in Figure 6.  Figure 6 shows that the β-ionone molecule is surrounded by ALA1, MET7, TRP27, ARG28, VAL29, MET39, TYR102, MET105, LYS106, and ASN109, which are potential active amino acids for the OBP1-β-ionone interaction. The amino acid TYR102 was suitable for the formation of alkyl interactions, and nine other amino acids form internal energy using van der Waals potentials.

Discussion
During evolution, insects have adapted a sensitive chemosensory system. In the past 30 years, entomologists have conducted in-depth research studies on the molecular mechanisms of insect sensing of host plant volatiles and sex pheromones, and determined that sensitive chemical systems play an important role in insect searches for hosts, mating, and oviposition [39][40][41]. Thus, elucidating how whitefly recognizes oviposition repellent volatiles may provide a theoretical basis for the ecological control of pests.
Plant volatiles play a very important role in the chemical communication of insects, and these volatiles play a key role in influencing the selection of insects for feeding and ovipositioning on hosts. In this study, we analyzed the binding properties of B. tabaci odorant-binding proteins OBP1 and OBP4 to the ovipositioning repellent volatile β-ionone. The respective dissociation constants of β-ionone and OBP1 and OBP4 were 5.15 µmol/L and 3.62 µmol/L; both were <10 µmol/L, indicating strong binding ability. β-ionone is a floral odorant of lipids, and insects use some floral volatiles to locate plants [42]. The results of this study are consistent with the findings of previous studies [24][25][26][27][28][29][30]. Some insect OBPs strongly bind to β-ionone in test candidate substances, indicating that β-ionone has high affinity for insect OBPs. Because β-ionone is a floral odor chemical and is widely found in flowering plants, it is possible that OBP1 and OBP4 are involved in insect ovipositioning on host flowering plants.
Motif analysis showed that motif1 existed in all the OBPs, and this conserved domain may play an essential role for the OBPs to bind β-ionone. In subsequent studies, the experiment should be further confirmed to confirm the role of this motif.
In addition, previous studies involving bioinformatic analysis of the genome and transcriptome of B. tabaci have identified eight OBP-encoding genes in B. tabaci [9,31], among which OBP8 has strong binding affinity to β-ionone (k i value was 13.32 µmol/L). This indicates that many insect olfactory proteins, including OBP in whitefly, function in recognizing β-ionone. Our past expression profiling studies have also shown that OBP1 and OBP4 are highly expressed in heads of female adults [20]. Accordingly, we can infer that β-ionone may play a crucial role in the selection of oviposition hosts for B. tabaci. Therefore, multiple olfactory proteins of B. tabaci retain the function of binding this compound. In addition, OBP2 and OBP6 were also highly expressed in the heads of female adults [20], and the binding characteristics of these two OBPs to β-ionone need to be further confirmed by experiments in the future.
In addition, in this study, 1-NPN was found to be a suitable fluorescent reporter for OBP1, and bis-ANS is a suitable fluorescent reporter for B. tabaci OBP4. We found that the commonly used 1-NPN probe is not suitable as a fluorescence reporter for OBP4. These results indicate that in the functional study of the olfactory proteins of B. tabaci, researchers should exercise more caution in selecting fluorescent reporters, and different olfactory proteins may be required for various fluorescent reporters such as 1-NPN or bis-ANS.
In the current study, α-ionone showed a weak binding to OBP1 and OBP4; according to previous study, a-ionone also showed a weak ovipositioning repellent to whitefly at the doses of 0.05 ng/µL and 0.5 ng/µL [10]. Thus, this binding ability is consistent with ovipositioning repellent activity, further emphasizing the importance of OBP1 and OBP4 in the binding of ovipositioning repellent compounds.
Protein homology and evolutionary analyses revealed that OBP1 and OBP4 have extremely high homology (99% and 96%, respectively) between B. tabaci B and Q, thereby indicating orthology, prompting us to hypothesize that OBP1 and OBP4 in B-type whitefly also function in β-ionone recognition. In addition, OBP3 of Daktulosphaira vitifoliae is on the same branch of the phylogenetic tree and, thus, we can infer that the gene also functions in β-ionone recognition. The ability of these candidate OBP genes to bind to β-ionone thus needs to be confirmed experimentally in future studies.
In this study, modeling of OBP1 structures and their docking relationship with β-ionone molecules can be used to predict the oviposition repellent compounds of B. tabaci by molecular docking to other small molecules. In addition, it can also be used to promote modifications of β-ionone by chemical modification to develop the analog of β-ionone. Similar studies have been successfully conducted in the development and application of aphid alarm pheromone (E)-β-farnesene analogs, which provides a good case for the further field application of β-ionone in whitefly management. For OBP4, because the homology with the PDB database template is <30%, it is not suitable for structural modeling research studies [38,43]. The availability of additional crystal structures of insect OBPs may improve our understanding of the interaction between OBP4 and β-ionone.
β-Ionone is a common plant aroma chemical, and previous studies have found that the compound can attract bees to pollinate and increase crop yield. Caceres et al. [10] found that β-ionone has a strong repellent effect toward both the flea beetle and the spider mite, and significant oviposition deterrence to whiteflies. Therefore, the compound has potential applications to agro-ecosystems. For example, in the tomato production system, the compound can be formulated as a molecule agent for spraying on tomato, thereby improving bee pollination and affecting the oviposition of whiteflies, and reaching the goals of high-quality, high-yield, and safe tomato production.
The current study explored the olfactory molecular mechanism of β-ionone in B. tabaci using OBP. In the future, it is necessary to comprehensively identify the odorant receptor and neurons in whitefly that contribute to β-ionone recognition to fully reveal the underlying olfactory molecular mechanism and provide valuable clues for the prevention and control of this insect pest. The study on the function of OBPs of B. tabaci may help in elucidating the underlying host preference mechanism of B. tabaci, provide a theoretical basis for its ecological control, and promote the application of "push-pull strategy" in the field or greenhouse.

Conclusions
We tested the binding function of OBP1 and OBP4 proteins with β-ionone, which is an oviposition repellent volatile against B. tabaci. Both proteins showed high affinity for β-ionone, with dissociation constants with OBP1 and OBP4 of 5.15 and 3.62 µM, respectively. These proteins play an important role in the identification of oviposition volatiles in B. tabaci.

Conflicts of Interest:
The authors declare no conflict of interest.