Quantitative RT-PCR (qRT-PCR)
For each biological sample, antennal pairs from 100 unmated male or unmated female C. pomonella adults were dissected into RNA-Later. Total RNA from all samples was extracted and purified with the RNeasy Plus Mini Kit (Qiagen, Venlo, Netherlands), and cDNA was synthesized with the QuantiTech Reverse Transcription Kit cDNA synthesis kit (Qiagen) according to the manufacturer's protocol. qRT-PCR experiments were carried out with Roche Light Cycler 480 II thermocycler (Roche, Basel, Switzerland). qRT-PCR primers for each gene were designed with the IDT RealTime PCR tool (Integrated DNA Technologies; https://www.idtdna.com/scitools/Applications/RealTimePCR/) (Supplementary Table 1). For each primer assay, amplification was performed on biological triplicates of unmated male and female antennal samples with technical duplicates. For all reactions, the following contents were added: 2.0 μL of cDNA sample, 12.5 μL of enzymatic mix, iTaq SYBR Green supermix with ROX (Bio-Rad Laboratories, Hercules, CA, USA), 5.5 μL water, and 5.0 μL of gene-specific primers (100 nM final concentration of each primer), for a final reaction volume of 25 μL. The qRT-PCR amplification protocol was run as follows: initiation phase: 3 min at 95 °C; amplification phase (40 cycles): 10 s at 95 °C, 40 s at 56 °C; melting curve phase: 40–95 °C gradient, with analysis every 1.0 °C; melting curves were analyzed to verify the specificity of amplification products. Primer efficiencies were calculated for all primer pairs using a Topo4 plasmid (Invitrogen) with an insert of cloned amplicon using the same primers for the qRT-PCR assay, with six serial dilutions ranging from 2.0 ng to 2e-5 ng and a dilution factor of 10. For OR22, relative expression was normalized to the expression of two reference genes, ActR2 and HSP40; these reference genes were selected from a pool of six candidate reference genes(Lü et al. 2018) based upon optimal stability of expression across all biological samples and primer efficiency values as close as possible to 100%. Candidate reference gene sequences were identified in our previously published transcriptome(Walker et al. 2016) or the genome database associated with the published C. pomonella genome (Wan et al. 2019). Relative gene expression differences across male and female samples were assessed by determination of relative quantities (primer efficiency^Delta(CT)) of the OR22 gene relative to the geometric mean of the relative quantities of the two reference genes (Vandesompele et al. 2002). Delta(CT) values for each biological replicate were calculated relative to the average CT value of the three male antennal samples (Supplementary Data File 1). To control for potential differences in olfactory gene expression in male versus female antennae, OR22 expression differences were further compared to the expression of the odorant receptor coreceptor, CpomOrco. Statistical differences were assessed across conditions on binary log-transformed relative quantities using an independent t-test with two-tailed distribution and two samples with equal variance; significance was assessed at (P < 0.05).
Cloning and heterologous expression of CpomOR22 in Drosophila empty neuron system
A synthetic construct of the OR22 ORF of C. pomonella was obtained (Eurofins Genomics, Ebersberg, Germany) as a plasmid insert in pCR2.1-Topo; the ORF was based upon its sequence identified in the antennal transcriptome (Walker et al. 2016)and was codon optimized for expression in Drosophila melanogaster (Supplementary Data File 2). The complete ORF encoding CpomOR22 was amplified by PCR combining specific codon-optimized CDS-primers (Fw: 5'-ATGAAGTTTGAAGAGGCCGAC-3'; Rv: 5´-TTACTCGATAGAGGATTTGAGCATGA-3') with the pCR2.1 plasmid as a template. Purified PCR products were then cloned into the PCR8/GW/TOPO plasmid (Invitrogen, Waltham, MA, USA). Insert integrity and orientation were confirmed by Sanger sequencing, with a 3730xl DNA Analyzer (Eurofins Genomics). Cassettes with inserts were then transferred from their PCR8/GW/TOPO plasmids to the destination vector (pUASg-HA.attB, constructed by E. Furger and J. Bischof, kindly provided by the Basler group, Zürich) using the Gateway LR Clonase II kit (Invitrogen). Insert integrity and orientation were again checked by Sanger sequencing.
Transformant D. melanogaster lines with pUAS-CpomOR22 were generated by Best Gene (Chino Hills, CA, USA), injecting into Best Gene Strain #24749 with genotype M{3xP3-RFP.attP}ZH-86Fb and insertion locus on the third chromosome. Crossings were performed with standard balancer lines and the Δhalo chromosomal background to drive the expression of CpomOR22 in the A neuron of ab3 basiconic sensilla (ab3A OSNs) according to procedures already established in our labs (Gonzalez et al. 2016). The final crossing was performed with w;Δhalo/CyO;pOr22a-Gal4 mutant line (Dobritsa et al. 2003, Hallem et al. 2004), and selection of ∆halo homozygotes was based on the straight wings phenotype. The final strain tested by SSR and GC-SSR had the following genotype: w;Δhalo;pUAS-CpomOR22/pOR22a-Gal4. Insects were reared in our facilities at room temperature (25 ± 2°C) on a sugar-yeast-cornmeal diet (https://bdsc.indiana.edu/information/recipes/bloomfood.html) at a relative humidity of 50 ±5 % and under 12:12 light:dark photoperiod.
Volatile collection from apple headspace
Volatile collections from Hoplocampa testudinea (Hymenoptera:Tenthredinidae) infested branches of apple trees (M. domestica v. 'Discovery', with 10-15 pieces of 1-3 cm diameter apples and 10-20 leaves, sample names abbreviated as Hoplomalus) and branches with non-infested fruits (sample names abbreviated as Malus) were used in the GC-SSR recordings. The H. testudinea insect-infested apples were selected to qualitatively broaden the chemical profile for screening purposes.The volatile collections were done in the laboratory using Porapak Q filled volatile collection traps. The biological samples were placed in 25 x 38 cm polyester oven roasting bags (Look, Teri- nex Ltd., England), and charcoal-filtered air was pumped into the bag at 410 mL/min to keep the bag inflated during sampling. The volatile collection traps were connected to the oven bag and fit with a Teflon tube, while the headspace was drawn through each trap at the rate of 200 mL/min for 2 hours 40 minutes, 6 hours, 14 hours, or 20 hours depending on the sample (2 hours 40 minutes: Malus 559, Hoplomalus 560, Malus 561; 6 hours: Hoplomalus 583, Malus 590;14 hours: Hoplomalus 566, Malus 603; 20 hours Hoplomalus 500). The samples were desorbed from the volatile traps with 200-400 uL hexane, and 1.0 ug heptyl acetate was added as an internal standard to the volatile extracts. The extracts were stored at -40°C in melted glass capillaries.
Single sensillum recordings
CpomOR22 expressed in the A neuron of ab3 basiconic sensilla was tested through single sensillum recordings (SSR), adapting the protocols we recently described (Cattaneo et al. 2023). In brief, three- to eight-day-old female flies were immobilized in 100 μL pipette tips with only the top half of the head protruding. The right antenna of each insect was gently pushed with a glass capillary against a piece of glass. This piece of glass and the pipette tip were fixed with dental wax on a microscope slide. Electrolytically sharpened tungsten electrodes (Harvard Apparatus Ltd, Edenbridge, United Kingdom) were used to penetrate the insect's body: the reference electrode was manually inserted in the right eye of the fly, while the recording electrode was maneuvered with a DC-3K micromanipulator equipped with a PM-10 piezo translator (Märzhäuser Wetzler GmbH, Wetzler, Germany) and inserted in ab3-sensilla. Signals coming from the olfactory sensory neurons were amplified ten times with a probe (INR-02, Syntech, Hilversum, the Netherlands), digitally converted through an IDAC-4-USB (Syntech) interface, and visualized and analyzed with the software Autospike v. 3.4 (Syntech). To carry the odorant stimulus, prevent antennal dryness, and minimize the influence of background odors from the environment, a constant humidified flow of 2.5 L/min charcoal-filtered air was delivered through a glass tube and directed to the preparation. To confirm the expression of CpomOR22-transgenes, basic spiking of ab3-neurons was compared with parental flies Δhalo-homozygous (w;Δhalo;pOr22a-Gal4 and w; Δhalo;+ mutants) (Cattaneo et al. 2023). A panel of 27 odorants (Table 1) was made of synthetic compounds from our collection (Cattaneo et al. 2023, Lebreton et al. 2017), including ligands that have been previously reported among codling moth pheromones, compounds emitted from fruit and yeast (Bengtsson et al. 2014, Bäckman et al. 2000, Bengtsson et al. 2001, Witzgall et al. 2012, Jennings et al. 1964), and a few novel ligands. Among these, we included some primary pheromones and kairomones of C. pomonella based on previous functional studies on its primary pheromone receptors (Bengtsson et al. 2014, Cattaneo et al. 2017). In the screening, we also added (Z, Z)-3,13-Octadecadien-1-yl acetate (CAS: 53120-27-7), a main pheromone compound from the red-belted clearwing moth Synanthedon myophaeformis, for which further investigation will be part of a separate study (Cattaneo et al. in preparation).
Based on the database of odorant responses (http://neuro.uni-konstanz.de/DoOR/content/DoOR.php(Münch and Galizia 2011, Galizia et al. 2010), the panel also included 2-heptanone (CAS 110-43-0) and 3-octanol (CAS: 589-98-0) as positive controls to validate recordings from ab3 sensilla by testing activation of D. melanogaster ab3B. To discriminate ab3 from ab2 sensilla, the ab2A activator ethyl acetate (CAS: 141-78-6) was included as a negative control. To test absence in the ab3A neuron of the wild-type expression of the D. melanogaster OR22a/b-subunits, ethyl hexanoate (CAS 123-66-0) was included as an additional negative control.
To screen the panel, all odorants were diluted in hexane (Sigma Aldrich, St. Louis, MO-USA) at 1.0 μg/μL. Stimuli were prepared by applying 10.0 μL of each dilution on grade 1 - 20 mm circles filter paper (GE Healthcare Life Science, Little Chalfont, United Kingdom), previously inserted into glass Pasteur pipettes (VWR, Milan, Italy), for a total amount of 10.0 μg of compound per stimulus. To minimize possible effects from the solvent, pipettes were left at least 10 minutes after preparation under the fume hood for solvent evaporation. Puffing provided an additional 2.5 mL of air through the pipette for 0.5 seconds by inserting the pipette within a side hole of the glass tube, directing the humidified air flow to the antennae. To characterize the intensity of the response, spike frequency was calculated as in Lebreton et al. (2017) by subtracting ab3A spikes counted for 0.5 seconds before the stimulus from the number of spikes counted for 0.5 seconds after the stimulus to calculate spike frequency in terms of ∆spikes/0.5sec. Responses to compounds of the panel were compared for five replicates, using a single insect as a replicate. Before validating significant differences in spike counting, tests of normality with the IBM SPSS Statistics software 29.0 (https://www.ibm.com/) unveiled that for some ligands, data were not normally distributed (Kolmogorov-Smirnova/Shapiro-Wilk test p<0.05, Supplementary Data File 3). Using the same software, spike frequencies of each compound were compared with respective values from the solvent (hexane) by the non-parametric Wilcoxon Signed Rank Test (p<0.05). For box-plot analysis, ∆spikes/0.5sec of each recording was normalized to the averaged ab3A firing rate for the specific insect replicate, as done in our previous studies(Cattaneo et al. 2023).
Gas chromatography coupled with single sensillum recordings (GC-SSR)
GC-SSR was performed with the same GC equipment in our labs that interfaced with the SSR rig we used in our previous investigation (Cattaneo et al. 2023). In brief, samples were injected on a 7890 GC-system (Agilent Technologies Inc., Santa Clara, CA, USA) provided with a 30 m x 0.32 mm fused silica capillary column (Agilent Technologies Inc.), coated with HP-5, df = 0.25 µm, programmed from 30°C (hold 3 min) at 8°C/min to 250°C (hold 5 min) (software: GC-SSR-1 - Agilent.OpenLab, Agilent Technologies). The outlet split from the GC column was a 1:1 ratio between the flame ionization detector and the mounted antenna, according to instrument settings. A humidified flow of 3.5-4.0 L/min charcoal-filtered air was directed into a 90-degree-angled glass tube with a hole on the angle where part of the column exiting from the transfer line was accessed. Glass-tubing was adjusted to a length of 17 cm, and ab3 sensilla was tested following the same optimization to 1.0 nanogram of active compound that we have adopted in Cattaneo et al (2023). The recording window was set to 35 min upon preliminary observation of retention times for the injected compounds. By GC-SSR we tested 1.0 ng and 10.0 ng aliquots of nonanal and (Z)-6-undecenal that we have chosen from our SSR-screening based on their effects (Table 1). Compounds were diluted in hexane between 0.001 and 0.010 μg/μL depending on the experiment condition, injecting 2.0 μL dilutions into the gas-chromatograph. Parallel experiments tested volatile collections from apple headspace (Hoplomalus and Malus) already available in our labs. To test headspace collections by GC-SSR, aliquots of 4.0 μL were injected into the gas chromatograph.
The headspace components that invoked responses were validated by recording in the GC-SSR using a DB-wax column. In brief, 2.0 μL of headspace was injected in the GC (7890) with an injector temperature of 225°C in splitless mode. The GC was fitted with a silica capillary column coated with DB-wax column (Agilent Technologies Inc., df = 0.25 μm) and temperature programmed from 30°C (hold time 3 min) until 225 °C (hold time 8 min) at 8 °C/min.
Hydrogen gas was used as a mobile phase at 7.5 mL/min. The GC effluent was carried onto the antennal preparation through a Gerstel ODP-2 transfer line connected to a glass tubing as described above. Retention times associated with neuronal activation were collected from the chromatograms exported from Chemstation B.03.02 (Agilent Technologies, Santa Clara, CA-USA) or from Autospike. To estimate neuronal activation, spikes were counted within 5 seconds from the emission of their respective GC peaks or the beginning of the ab3A effect, depending on the case. These counted spike numbers were subtracted from spikes five seconds antecedent to the effect and divided by 5 to calculate ∆spikes/second. Statistical analysis was performed as described above, including tests of normality (Supplementary Data File 3).
GC-MS analysis of active headspaces
The volatile samples were injected on a GC-MS (Agilent technologies, 7890B GC coupled with 5975 MSD) equipped with a DB-WAX capillary column (60 m x 250 μm x 0.25 μm). A volume of 2.0 μL samples were injected into the splitless injection port at 225°C in splitless mode. The carrier gas was helium, and the total column flow was 34.883 mL/min. The temperature program of the oven started at 30°C, which was held for 3 minutes and heated up at the rate of 8°C/min to 225°C, holding for 10 minutes the final temperature. To perform GC-MS, we used the mass spectrometer in electron ionization mode at 70 eV, and the detector scanned in the 29-400 mass-to-charge range.
The volatile samples were also injected on a GC-MS (Agilent technologies, 6890 GC coupled with 5977A MSD) equipped with an HP-5 capillary column (column: 60 m x 250 μm x 0.25 μm). The column, inlet, and mass spectrometer settings were the same as those described above. The oven temperature program started at 50°C, it was held for 2 minutes, and increased to 250°C at the rate of 8°C/min, holding the final temperature for 10 minutes.
GC-MS data were analyzed using Agilent Mass Hunter B.08.00. The area counts were calculated using manual integration. The volatile components were tentatively identified by comparing the experimental mass spectra to those found in MS Libraries (NIST11 and Wiley12) using the Nist MS Search v. 2.4 program. The spectrum-based identification was verified by calculating the Kovats retention indices (RI) of components and comparing those to Kovats retention indices found in the NIST WebBook (https://webbook.nist.gov/) or PubChem databases.
Cloning and heterologous expression of CpomOR22 in oocytes from Xenopus laevis
To further investigate the responsiveness of CpomOR22 to environmental volatiles, we expressed CpomOR22 in conjunction with the obligate coreceptor, C. pomonella Orco (CpomOrco), in unfertilized, defolliculated X. laevis oocytes. The main objectives of the Xenopus experiments were to validate positive findings from the Drosophila empty neuron system and expand the odor space/set of odorants used to test OR22, beyond those used in the SSR and GC-SSR experiments.
Using the two-electrode voltage clamp technique, we recorded responses of the odorant receptor complex, observed by the change in inward current, to 16 different odorant blends. These blends together contain 138 compounds, including 15 that were screened in Drosophila, grouped by chemical classes based on structure (Supplementary Table 2). The reason we have chosen this method was to make use of parallel heterologous systems, allowing us to test a broader range of ligands within blends. Among these ligands, we kept nonanal as a reference, based on our observed SSR responses from this ligand being one the most active, which was also validated in the headspace from apple. The individual compounds, each at a concentration of 10-4 M in the blends, were applied to the oocyte with buffer perfusion.
CpomOrco and CpomOR22 templates were synthesized by Twist Bioscience (South San Francisco, CA, USA) and delivered in pENTR vector, and then subcloned into the X. laevis compatible destination vector pSP64t (Gateway™ LR Clonase™ II Enzyme Mix, Invitrogen Corp., Carlsbad, CA, USA). Plasmids were purified using GeneJET Plasmid Miniprep Kit (ThermoFisher Scientific, Waltham, MA, USA) and verified using bidirectional Sanger sequencing (Psomagen, Rockville, MD, USA). 5,000 ng of each plasmid was linearized using XbaI (FastDigest XbaI, ThermoFisher Scientific, Waltham, MA, USA). The linearized plasmids were purified (GeneJET Gel Extraction Kit, ThermoFisher Scientific) and checked for concentration using the Nanodrop (NanoDrop™ One UV-Vis Spectrophotometer, Witec AG, Sursee, Switzerland). cRNA was synthesized from the linearized plasmids (mMESSAGE mMACHINE™ SP6 Transcription Kit, ThemoFisher Scientific, Waltham, MA, USA) and incubated at 37°C for 12 hours. 1.0 uL of RNase inhibitor was added to each reaction (RNaseOUT™ Recombinant Ribonuclease Inhibitor, ThermoFisher Scientific, Waltham, MA, USA). Synthesis was verified using gel electrophoresis, and concentration was checked using the Nanodrop (ThermoFisher Scientific).
Stage V-VII X. laevis defolliculated oocytes were ordered from Xenopus1 (Dexter, MI, USA) and incubated in ND96 incubation media (96 mM NaCl, 2 mM KCl, 5 mM HEPES, 1.8 mM CaCl2, 1 mM MgCl2, pH 7.6) augmented with 5% horse serum (ThermoFisher Scientific, Waltham, MA, USA), 50 μg/mL tetracycline, 100 μg/mL streptomycin, 100 μg/mL penicillin, and 550 μg/mL sodium pyruvate. Oocytes were injected with 30 nL of RNA (30 ng of each cRNA) using the Nanoliter 2010 injector (World Precision Instruments, Inc., Sarasota, FL, USA). The resting membrane potential and odorant-induced changes of oocytes expressing odorant receptor cRNAs were recorded 72 hours post-injection using the 2-microelectrode voltage-clamp technique (TEVC). The OC-725C oocyte clamp (Warner Instruments, LLC, Hamden, CT, USA) held a −80 mV holding potential. The effect of the blends (Supplementary Data File 4) on odorant receptor complexes was determined by perfusing 10−4 M concentration blends across individual oocytes. The membrane current was permitted to return to baseline between blend introductions. Data was recorded with the Digidata 1550 B digitizer and pCLAMP10 software (Molecular Devices, Sunnyvale, CA, USA). Raw data was collected and normalized according to the blend or compound producing the greatest response (Supplementary Data 3). Following initial screenings, unitary compounds comprising activating blends were perfused to establish tuning curves. Concentration responses for gamma-undecalactone and undecanal were determined by challenging oocytes with half-log concentrations between 10−7 M and 10−4 M for 10 s. The current was allowed to return to baseline between compound administrations. Blends and unitary compounds were perfused on 8-10 oocytes per trial. GraphPad Prism 8 (GraphPad Software, Inc., La Jolla, CA, USA) was utilized for data analyses.