In vivo functional characterisation of pheromone binding protein-1 in the silkmoth, Bombyx mori

Male moths detect sex pheromones emitted by conspecific females with high sensitivity and specificity by the olfactory sensilla on their antennae. Pheromone binding proteins (PBPs) are highly enriched in the sensillum lymph of pheromone sensitive olfactory sensilla and are supposed to contribute to the sensitivity and selectivity of pheromone detection in moths. However, the functional role of PBPs in moth sex pheromone detection in vivo remains obscure. In the silkmoth, Bombyx mori, female moths emit bombykol as a single attractive sex pheromone component along with a small amount of bombykal that negatively modulates the behavioural responses to bombykol. A pair of olfactory receptor neurons, specifically tuned to bombykol or bombykal, co-localise in the trichodeum sensilla, the sensillum lymph of which contains a single PBP, namely, BmPBP1. We analysed the roles of BmPBP1 using BmPBP1-knockout silkmoth lines generated by transcription activator-like effector nuclease-mediated gene targeting. Electroantennogram analysis revealed that the peak response amplitudes of BmPBP1-knockout male antennae to bombykol and bombykal were significantly reduced by a similar percentage when compared with those of the wild-type males. Our results indicate that BmPBP1 plays a crucial role in enhancing the sensitivity, but not the selectivity, of sex pheromone detection in silkmoths.

The silkmoth Bombyx mori is one of the model insects used in sex pheromone communication research. Peripheral pheromone detection system of this species, including relationships among the sensillum types, ORNs, pheromone receptors, and PBPs, is well characterised 12 . Female silkmoths emit bombykol [(E,Z)-10,12-hexadecadien-1-ol] and bombykal [(E,Z)-10,12-hexadecadienal] from their sex pheromone gland at a typical ratio of 11:1 13 . Of these two compounds, only the major component, bombykol, is sufficient to induce pheromone source orientation behaviour in male moths 13,14 , whereas the minor component, bombykal, negatively modulates the initiation of this behaviour 13,15 . Male silkmoths detect these pheromones using long sensillum trichodea on their antennae 16 . This type of sensilla contain a pair of bombykol-and bombykal-sensitive ORNs that express sex pheromone receptors specific to bombykol (BmOR1) and bombykal (BmOR3), respectively [14][15][16][17][18][19] . In the silkmoth, one PBP gene (BmPBP1) and two PBP-like genes (BmPBP2, 3) have been reported 20,21 . However, only BmPBP1 has been shown to be expressed in the sensillum lymph of pheromone sensitive sensilla and its associated accessory cells at mRNA and protein levels 21,22 , whereas BmPBP2 and BmPBP3 are expressed in accessory cells that are not associated with the pheromone-sensitive sensilla 21 . In addition, immunohistochemical analysis has shown that other OBPs that have been examined (GOBP1, GOBP2, and ABPX) are rarely or not expressed in the long sensilla trichodea of male antennae 22 . These findings suggest that BmPBP1 is involved in the detection of bombykol and bombykal whereas BmPBP2 and BmPBP3 play roles in the detection of compounds other than sex pheromones. However, the possible functions of BmPBP1 have yet to be conclusively established, particularly with respect to whether BmPBP1 plays a crucial role in the discrimination between bombykol and bombykal: BmPBP1 has been reported to be required for the response selectivity of BmOR1 expressed in HEK293T cells to bombykol 23 , whereas BmOR1 expressed in Xenopus oocytes and Sf21 cells is able to respond specifically to bombykol without BmPBP1. However, these results have yet to be verified by conducting in vivo functional analyses using BmPBP1 gene knockout moths.
In the present study, we generated BmPBP1-knockout silkmoth lines by using transcription activator-like effector nuclease (TALEN)-mediated gene targeting in order to elucidate the in vivo functions of BmPBP1 in the sex pheromone detection of male silkmoths. By electroantennogram (EAG) recordings, we show that the response of BmPBP1-knockout male antennae to both bombykol and bombykal was significantly reduced compared with those in the wild-type moths. We also established that the sensitivity of the initiation of pheromone source orientation behaviour is reduced in BmPBP1-knockout male moths. On the basis of our results, we discuss whether BmPBP1 contributes to the sensitive and selective detection of bombykol and bombykal in male silkmoths.

Results
Establishment of BmPBP1-knockout silkmoths. To characterise the functional role of PBPs in the detection of sex pheromones in vivo, we generated BmPBP1-knockout silkmoths using TALENs that were designed to target the second exon of the BmPBP1 gene, which is located on the 19th chromosome (Fig. 1a). By genomic PCR screening of G 1 individuals, followed by sequence analyses, we obtained two genetic lines with 4-and 12-bp deletions in the target region of the TALENs, respectively (Figs 1b and S1). From these two genetic lines, we selected a line that had the 4-base deletion. This deletion caused a frame shift at the 55th amino acid residue of the BmPBP1 protein and the introduction of a stop codon at the 73rd amino acid residue. Given that the first 22 amino acids of this protein constitute a signal peptide, this premature translational termination would result in a truncated protein of only 51 amino acids, compared with the 142 amino acids of the wild-type protein 20 . The mutant BmPBP1 protein contained only three out of the nine residues that form the bombykol binding pocket and only one out of the six cysteine residues required for proper folding 24,25 , resulting in the synthesis of a loss-of-function BmPBP1 protein (Fig. 1c).
To confirm the 4-bp deletion at the transcript level, we amplified full-length coding sequences of BmPBP1 from the antennal cDNA of homozygous mutant (BmPBP1−/BmPBP1−) male moths using RT-PCR. After cloning the PCR products into a sequence plasmid vector, we sequenced 11 clones and confirmed that the DNA insert in all the clones had the same deletion as that in the genomic sequence, further supporting our observation that this allele encodes a substantially truncated BmPBP1 protein. We noted that both BmPBP1−/BmPBP1− female and male moths were fertile, and that their offspring grew normally to adulthood.

EAG analysis of BmPBP1-knockout male antennae.
To determine the effects of the loss of function of BmPBP1 on the olfactory response in male moths, we recorded the EAG responses of male antennae to sex pheromone components. The peak EAG amplitudes of the antennae of BmPBP1-knockout males to bombykol and bombykal were significantly reduced compared with those of the antennae of wild-type males (Fig. 2a,b). Notably, the degree of reduction in EAG amplitude was similar between the responses to bombykol and bombykal ( Table 1), suggesting that BmPBP1 contributes to the detection of these two components to a similar extent, and thus is not likely to have selectivity for either of the two compounds. Although the peak amplitudes were significantly lower than those of the wild-type males, antennae from the BmPBP1-knockout males showed clear dose-dependent responses to bombykol and bombykal (Fig. 2a,b), suggesting that male moths can detect bombykol and bombykal, albeit with low sensitivity, in the absence of functional BmPBP1. the deletion regions (left) or PCR primers with the 3′ primer designed to anneal to the deleted sequence (the black arrow in (b))(right) were used. (d) Deduced amino acid sequences of wild-type (top) and 4-bp deletion BmPBP1 moths (bottom). The black arrowhead on the sequences indicates a signal peptide cleavage site, and the red arrow indicates the position of a frame shift caused by the deletions. Red boxes indicate amino acids that are identical between the two sequences. To exclude the possibility that the loss of BmPBP1 affected the olfactory response of antennae as a whole, we analysed the EAG response of male antennae to general odorants, linalool and citral 26 . The peak EAG amplitudes for the response to both linalool and citral were not significantly different between the antennae of the BmPBP1-knockout and wild-type males (Figs 2c and S2). Although linalool reportedly inhibits most of bombykol receptor neurons and activates some of them 27 , our results that BmPBP1-knockout did not affect EAG responses to general odorants indicate the lowered response of BmPBP1-knockout male moths to pheromones was not due to a general effect on olfactory detection.
Behavioural analysis of BmPBP1-knockout male moths. Finally, using a closed-box assay 28,29 , we investigated whether the reduced response of antennae to bombykol caused by BmPBP1 mutation affected the behavioural responsiveness to bombykol. In this assay, we used wing flapping behaviour as a criterion for initiation of the pheromone source orientation behaviour. Dose-response analysis revealed that the percentage of BmPBP1-knockout males that initiated the pheromone source orientation behaviour was lower than that of the wild-type males (Fig. 3a). As in wild-type males, BmPBP1-knockout males did not show any behavioural Error bars represent ± SEM. The asterisks indicate significant differences between the groups (**p < 0.01), as determined using Student's t-test for comparing pairs of data. (c) Comparison of linalool and citral-induced peak EAG amplitudes of BmPBP1−/BmPBP1− (red; n = 5 for linalool, n = 6 for citral) and wild-type (blue; n = 5 for linalool, n = 7 for citral) male moths. Error bars represent ± SEM. No significant difference was detected between the two groups (Student's t-test; p = 0.700 for linalool, p = 0.529 for citral).  Fig. 3b). These results indicate that the behavioural responsiveness to bombykol was reduced in BmPBP1-KO males.

Discussion
In this study, we demonstrated that the loss of functional BmPBP1 led to the lowering of EAG responses to both bombykol and bombykal to a similar extent. These observations indicate two important aspects of BmPBP1 function: (1) BmPBP1 is necessary for the sensitive detection of bombykol and bombykal and (2) BmPBP1 is not involved in the discrimination between bombykol and bombykal. Our results are consistent with those of a recent study by Ye et al., wherein the authors observed significant reductions in the EAG responses of the antennae of PBP1-knockout Helicoverpa armigera males to three sex pheromone components in this species 30 . Previous studies suggested two different molecular mechanisms underlying the specific response of male silkmoths to bombykol. On the basis of an in vivo opened sensillum tip analysis, Pophof reported that BmPBP1 mediates the response of BmOR1 to bombykol but not to bombykal, and thus proposed that response selectivity to pheromones is intermediated by PBPs specialized in the recognition of particular ligands 31 . Using HEK293T cell expression system, Große-Wilde et al. showed that BmOR1 was able to respond to both bombykol and bombykal when these compounds were solubilised by dimethyl sulfoxide (DMSO), but responded only to bombykol when BmPBP1 was used as a solubiliser instead of DMSO 23 , suggesting that BmPBP1 selectively solubilised bombykol and that the interplay between bombykol and BmPBP1 is important for the selective response of BmOR1 to bombykol. Accordingly, these authors proposed that other PBPs that bind to bombykal and mediate the bombykal response should be present in the sensilla trichodea of male moths. In contrast, it has been demonstrated that Xenopus oocytes or Sf21 cells, co-expressing BmOR1 and the co-receptor BmOrco, responded specifically to bombykol that was dissolved in DMSO, suggesting that specific interaction between bombykol and BmOR1 defines the specificity of the response 18,32 . Our results, demonstrating that BmPBP1 mediates the response to both bombykol and bombykal in vivo, are consistent with the latter mechanism because a BmOR1-expressing bombykol-sensitive ORN and a BmOR3-expressing bombykal-sensitive ORN co-localise in the same sensilla in the silkmoth, and are thus bathed in the same sensillum lymph containing BmPBP1 22 . Further, our results are consistent with the docking simulation and in vitro binding assay reported by Gräter et al., which showed that BmPBP1 bound to both bombykol and bombykal with nearly the same affinity 33 .
Although the responsiveness of the antennae of BmPBP1-knockout males to sex pheromones was considerably lower than that of the antennae of the wild-type males, BmPBP1-knockout male antennae still showed dose-dependent EAG responses to bombykol and bombykal. Therefore, even in the absence of a functional BmPBP1, subsets of pheromone molecules can reach the dendritic membranes of ORNs and activate pheromone receptor proteins expressed on the membranes. The residual EAG response of the antennae of BmPBP1-knockout male moths may be explained by the non-specific binding and solubilisation of the pheromone molecules with some water soluble proteins in the sensillum lymph. In this regard, we cannot exclude the possibility that other OBPs are co-expressed with BmPBP1 in the sensillum trichodea. Although we cannot completely exclude the possibility that the mutant BmPBP1 still has some binding capacity to the pheromones, it's highly unlikely because mutant BmPBP1 protein doesn't contain amino acid residues required for proper folding and formation of the binding pocket for bombykol (see Results). Alternatively, there may be another pathway from the olfactory pore to the pheromone receptors, and in this regard, it has been reported that at least some of the olfactory tubules in the sensillum trichodea come into contact with the dendritic membrane of ORNs in the giant silkmoth Antheraea polyphemus, thereby raising the possibility of a direct pathway from the olfactory pores to the dendritic membranes 34,35 . Further studies are accordingly required to clarify how pheromone molecules reach the pheromone receptors in the absence of solubilisation by BmPBP1.
We also demonstrated the effects of BmPBP1-knockout at the behavioural level, showing that the percentage of male moths that initiate pheromone source orientation behaviour was significantly reduced in BmPBP1-knockout males. To gain a more precise understanding of the functional role of BmPBP1 in the sex pheromone communication system, it would be informative to examine not only the sensitivity of behavioural response initiation but also the efficiency of male orientation to female moths and successful copulation.
In summary, by using BmPBP1-knockout moths, we showed that BmPBP1 contributes to the sensitivity of pheromone detection, but does not play a significant role in the discrimination of bombykol and bombykal. Apart from the solubilisation and transportation of pheromones, PBPs have also been proposed to play additional roles 36 , including the protection of pheromone molecule from enzymatic degradation 37 and rapid inactivation of pheromones 38 . Detailed physiological analyses of BmPBP1-knockout moths will shed further light on the underlying mechanisms and modes of action of BmPBP1 in pheromone detection in vivo.

Methods
Animals and chemicals. The w1-pnd strain of Bombyx mori, which is non-diapausing and characterised by non-pigmented eggs and eyes, was used for the generation of BmPBP1-knockout moths. Larvae were reared on an artificial diet (Nihon Nosan Kogyo, Yokohama, Japan) at 25 °C under a 16:8 h (light/dark) photoperiod. The purity (>99.5%) of synthetic bombykol and bombykal were verified by gas chromatography under previously described conditions 39,40 .

Construction of TALEN vectors and synthesis of RNA for injection. TALEN expression vectors
were constructed as described previously 41,42 . The RNA used for injection was prepared as described previously 28 using a Qiagen Hispeed plasmid midi kit (Qiagen, Hilden, Germany) and an mMESSAGE mMACHINE T7 Ultra Transcription kit (Ambion, Austin, USA). Extracted RNA was purified by LiCl precipitation, followed by four washes with 70% ethanol. The RNA of left and right TALENs was dissolved at a concentration of 0.2 μg/μL in 0.5 mM phosphate buffer (pH 7.0) containing 5 mM KCl, and this RNA solution was injected into preblastoderm-stage embryos, as described previously 43 .
Screening of mutagenised moths. G 1 eggs were obtained by the sibling mating of G 0 adults. Genomic DNA of G 1 eggs from different broods was extracted separately using a DNeasy Blood & Tissue Kit (Qiagen, Hilden, Germany). The region surrounding the target site was PCR amplified using genomic DNA as the template and specific primers (sense: 5′-CGACCTTCGCAAGGTATGAT-3′ and antisense: 5′-AGGCACATTATAGCGCATCC-3′). The PCR products were sequenced directly using an ABI3700 DNA analyser (Applied Biosystems, Foster City, CA, USA). G 1 broods that showed overlapping sequencing patterns with the target sequence were reared to the adult stage. After the G 1 moths had been crossed with wild-type adults, their genomic DNA was extracted, PCR amplified, and sequenced, as described above, to identify the mutagenised individuals. BmPBP1−/BmPBP1− males were obtained by crossing BmPBP1−/BmPBP1+ females with BmPBP1−/BmPBP1+ males.

Reverse-transcription-PCR.
Total RNA was extracted from male moth antennae using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) as described previously 29 . The extracted RNA was reverse transcribed using an oligo(dT) adaptor primer (Takara-Bio, Otsu, Japan) and AMV reverse transcriptase (Takara-Bio, Otsu, Japan) at 42 °C for 30 min. The cDNA of BmPBP1 was amplified using Ex Taq DNA polymerase (Takara-Bio, Otsu, Japan) and a primer pair for BmPBP1 (5′-ATGTCTATCCAAGGACAGATCGC-3′ and 5′-TCAAACTTCAGCTAAAATTTCTCCC-3′), with thermal cycling at 95 °C for 5 min, followed by 30 cycles at 95 °C for 30 s, 53 °C for 30 s, and 72 °C for 30 s, and a final extension at 72 °C for 5 min. Equal amounts of PCR products were separated by electrophoresis on 2.0% agarose gels. The cDNA was cloned into a pGEM-T Easy vector (Promega, Madison, USA), and the PCR products were sequenced using an ABI3700 DNA analyser (Applied Biosystems, Foster City, CA, USA).

Electroantennogram (EAG) recordings.
The antennae of male moths were excised at the base, and a few segments at the tip were cut off. The antennae were then mounted on the EAG probe using electrode gel (SPECTRA 360; Parker Laboratories, Fairfield, NJ, USA). A glass cartridge (inner diameter, 5 mm) was prepared for stimulation by inserting a piece of filter paper (1.5 × 1.5 cm), and 5 μL of pheromone solution in n-hexane or neat n-hexane (control) was administered. For general odorant stimulation, 5 μL of 10% (v/v) linalool or citral (Wako, Osaka, Japan) in paraffin oil (Sigma-Aldrich, St. Louis, USA) was used as the stimulant. A charcoal-purified airstream (1 L/min) was passed through a glass pipette and directed onto the antenna. The EAG responses were amplified using a custom-made amplifier (Minegishi and Kanzaki, unpublished), low-pass filtered at 300 Hz, and digitised at 1 kHz (USB-6210; National Instruments, Austin, TX, USA). The data were analysed using a custom-written programme (MATLAB; Mathworks, Natick, MA, USA). Following EAG measurements, the genotype of all males was confirmed by PCR as described in the section 'Screening of mutagenised moths' . The percentage reduction in peak EAG amplitude (Table 1) was calculated by dividing the peak EAG amplitude of BmPBP1−/BmPBP1− male antennae by that of wild-type males.
Behavioural experiments. The behavioural responses to bombykol were examined as described previously 29 , with the exception that the male silkmoths were used within 1 to 5 days after eclosion. The moths were exposed to increasing doses of bombykol or bombykal (0.01, 0.1, 1, and 10 ng) at 1-min intervals. Wing flapping within 10 s of the stimulation that lasted for more than 10 s was counted as a response. Subsequent to these behavioural experiments, the genotype of all the males was determined by PCR.

Statistical analysis.
To assess the statistical significance of differences in the EAG and behavioural response data for wild-type and BmPBP1-knockout moths, we used Student's t-test and Fisher's exact probability test for comparing pairs of data, respectively, by using Microsoft Excel 2010 and a commercial macroprogramme (Statcel version 3; Seiun-sya, Japan). The error bars shown in figures represent SEMs.