Insects rely on olfaction to guide a wide range of adaptive behaviors, including mate- and food-localization, mate choice, oviposition site selection, kin recognition and predator avoidance. In nocturnal insects, such as moths and cockroaches, mate finding is stimulated predominantly by long-range species-specific sex pheromones, typically emitted by females. During courtship, at close-range, males in most moth species emit a blend of pheromone compounds from an everted, often large, pheromone gland. While long-distance communication with sex pheromones has been remarkably well characterized in thousands of moth species, close-range chemosensory sexual communication remains poorly understood. We reveal that in the moth Chloridea virescens, the male pheromone consists of three distinct classes of compounds: de novo biosynthesized alcohols, aldehydes, acetates and carboxylic acids that resemble the female’s emissions, and newly identified compounds that are unique to the male pheromone: aliphatic polyunsaturated hydrocarbons and sequestered plant secondary compounds and hormone derivatives, including methyl salicylate (MeSA). Thus, males employ a mosaic pheromone blend of disparate origins that may serve multiple functions during courtship. We show that two olfactory receptors in female antennae are tuned to MeSA, which facilitates female acceptance of the male. Because MeSA is emitted by plants attacked by pathogens and herbivores, the chemosensory system of female moths was likely already tuned to this plant volatile, and males appear to exploit the female’s preadapted sensory bias. Interestingly, while female moths (largely nocturnal) and butterflies (diurnal) diverged in their use of sensory modalities in sexual communication, MeSA is used by males of both lineages.

Chloridea rearing

We used the standard laboratory YDK strain of C. virescens for our experiments, which was originally collected in 1988 from seven tobacco fields in Yadkin County, and one field each in neighboring Stokes and Forsythe Counties, North Carolina, USA.⁴⁴ Eggs were briefly rinsed in a 1.3% bleach solution, and after drying eggs were placed individually in 30 ml diet cups (DART, Mason, MI, USA) each containing 7.5 ml of artificial diet45 and maintained at 26 ± 1°C with a photoperiod of 14:10 (L: D). Pupae were separated by sex and placed in separate plastic cups. Emerged adults were checked daily and provided with a 10% sucrose solution. Male and female pupae and adults were kept in separate incubators at 26 ± 1°C and on a reversed photocycle (14 L:10 D) with L-off at 09:00.

GC-EAD analysis of male HP components

Gas chromatography (GC) coupled to electroantennogram detection (EAD) was performed as described previously.46 A Hewlett-Packard 5890 Series II GC (Agilent, Folsom, CA, USA), equipped with either a non-polar capillary column (EC-5, 30 m × 0.25 mm ID, 0.25 μm film thickness; Alltech, Deerfield, IL, USA) or a polar capillary column (EC-Wax, 30 m × 0.25 mm ID, 0.25 μm film thickness; Alltech) was used for GC-EAD analyses. The oven temperature was held at 50°C for 2 min, then increased by 15°C/min to 250°C. Injector and detector temperatures were both set to 250°C. Splitless injection was used with hydrogen as carrier gas at a flow of 1.2 ml/min. The column effluent was split 1:1 to a flame ionization detector (FID) and through a heated transfer line (250°C) to the EAD.

Moth antennae were excised with micro-scissors from 1-3-day-old virgin females and positioned between two gold wire electrodes within saline-filled glass capillaries. The physiological saline solution was adjusted specifically for Chloridea based on Hayashi and Hildebrand47: 8.77 mg/ml NaCl, 0.3 mg/ml KCl, 0.67 mg/ml CaCl₂, 2.4 mg/ml HEPES, and 0.9 mg/ml D-(+)-glucose were dissolved in diH₂O, and the pH adjusted to ~7 with 1 N NaOH. Aliquots of the saline solution were stored at –30°C. The electrodes were held in a custom-made acrylic holder that was placed inside a humidified cooling condenser maintained at ~15°C. Pure humidified air at ambient temperature was continuously blown over the antenna preparation in the cooled condenser. The output signal from the antennae was amplified 10× by a custom high-input impedance DC amplifier and filtered by a high-pass filter with a cutoff frequency of 0.5 Hz.⁴⁶ The amplifier output was routed through a signal acquisition board within the GC and displayed along with the FID signal in Agilent OpenLab CDS (version A.01.07).

GC-MS analysis of male HP components

For compound identification we used a 6890N GC coupled with a 5975 mass-selective detector (MS, Agilent, Folsom, CA, USA) operating in electron impact ionization mode (MS ion source set to 230°C, MS quadruple set to 150°C). Samples were injected in splitless mode with an initial temperature of 280°C. Separation of compounds was performed on a non-polar column (HP-5ms, 30 m x 0.25 mm x 0.25 μm (5%-phenyl)-methyl-siloxane stationary phase, Agilent). The temperature program started at 50°C for 1 min and then increased by 10°C per min to a final temperature of 320°C. Compounds were identified according to their Kovats indices, diagnostic ions, and their respective mass spectra. The Wiley 7/NIST 05 database was used for reference, and the identity of novel compounds was confirmed with synthetic standards (Sigma-Aldrich, St. Louis, MO, USA) processed in the GC-MS as well as on the GC-EAD-FID; both non-polar and polar columns were used.

Behavioral assays

Observations of moth mating behavior were conducted in a dark room at 26 ± 1°C., illuminated with dim red lights. Two-day old adult moths, which are best suited for behavioral assays,48 were used in all mating behavior experiments. The moths were taken out of the incubators and allowed to acclimate in the dark room 1 hr before L-off. Thirty min before L-off, one male and one female (2-days old) were placed in an observation arena (cylinder, 9 cm diameter x 7.5 cm high) (Fabri-Kal Corp, Kalamazoo, MI, USA) and moths were observed for mating every 30 min for 5 hrs. Moths mate for several hours, so pairs that remained in copula for at least 60 min were considered mated.²¹

Effects of HP ablation on mating

Application of gentle pressure to the abdomen caused the hairpencil (HP) complex to extend. HP structures were ablated by removing only the hairs associated with structure-2 (Figure S1A–D) with fine forceps. Sham-operated males were squeezed to expose the HPs, but no hairs were removed. All sham-operations and ablations were conducted before the onset of the photophase, about 14 hrs before the behavioral experiments, to give adequate time for the males to recover after the surgery.

Two groups of male moths were set up: (1) HP⁺: Sham-operated C. virescens males; (2) HP^–: HP-ablated C. virescens males. The mating rate of these two groups of males with 2-day-old virgin female moths was monitored in single pair assays.

Extraction of HP chemicals after HP-ablation

The hairs removed during HP ablation are largely storage and emission sites, whereas their underlying cells presumably biosynthesize the male sex pheromone. To determine whether HP ablation in fact reduced the amount of male sex pheromone, the HPs were extracted, and the quantity of male sex pheromone was determined by GC-FID analysis. In separate groups of males, the structure-2 hairs of the HPs (Figure S1A–D) were excised, and either 1 hr or 14 hrs later the whole HP complex was removed and extracted. Two extractions were performed. Short (10 sec) extractions were used to represent the chemicals on the surface of the HP gland. Long (20 min) extractions were designed to extract the chemicals from both the surface and interior of the HP glands. Each HP was extracted in 100 μl hexane containing 40 ng pentadecyl acetate (15:OAc) as internal standard. All extracts were kept at −30°C prior to chemical analysis and analyzed individually on an Agilent 7890A GC equipped with a 30 m × 0.25 mm ID DB-5 capillary column (Agilent). The amount of 16:OAc, the main component of the HPs of C. virescens, was quantified in each extract.

Effects of HP extract and MeSA on mating

A large pool of HP extracts was produced by excising the structure-2 hairs of 2–4-day-old C. virescens males in 100 μl hexane per male. For behavioral assays, one male-equivalent (1 ME; 100 μl of extract) was applied to a 42.5 mm diameter filter paper disc (Whatman #1, Sigma, St. Louis, MO, USA). The hexane was allowed to evaporate, and the filter paper was added to the behavioral assay arena.

The effect of HP extract and MeSA on mating of C. virescens was tested using HP-ablated males and 2-day old virgin females. Hair pencil extract and a range of MeSA doses were loaded onto filter papers and the mating rates in different groups were assessed. Four concentrations of MeSA were tested (100 ng, 10 ng, 1 ng and 0.1 ng) along with a hexane control for a total of six treatment groups: (1) Hexane: 100 μl hexane; (2) HP extract: 100 μl (1 ME) of C. virescens HP extract; (3) MeSA 100 (100 ng MeSA: 100 μl of 1 ng/μl); (4) MeSA 10 (10 ng MeSA: 100 μl of 0.1 ng/μl); (5) MeSA 1 (1 ng MeSA: 100 μl of 0.01 ng/μl); (6) MeSA 0.1 (0.1 ng MeSA: 100 μl of 0.001 ng/μl).

Cloning and sequence analysis of MeSA ORs

The full ORF sequences of CvirOR27 and CvirOR43 were obtained in the C. virescens antennal transcriptomes based on the basic local alignment search tool (BLAST) using the HarmOR27 and HarmOR43 sequences as queries. Full-length coding sequences of CvirOR27 and CvirOR43 were obtained from the antennal cDNA of C. virescens adults by PCR with specific primer pairs for each gene. Phylogenetic analysis was performed using 20 ORs from C. virescens (including 18 previously reported ORs and the two ORs we identified in this study) and 65 ORs from H. armigera. Amino acid sequences were aligned using ClustalW and neighbor-joining tree was constructed using Poisson model as implemented in MEGA X software.

Functional characterization of ORs

The Functional characterization of individual ORs was performed by heterologous expression in Xenopus oocytes combined with a two-electrode voltage-clamp system.⁴⁹ Briefly, the full-length gene of each OR and Orco were subcloned into the eukaryotic expression vector pT7TS. cRNAs were generated from linearized expression vectors using the mMESSAGE mMACHINE T7 kit (Ambion, Austin, TX, USA). Then, the cRNA mixture of ORx and Orco (27.6 ng each) was injected (Nanoliter 2010, WPI Inc., Sarasota, FL) into Xenopus oocytes. After incubation in nutrient solution at 18°C for 3–5 days, the response profile of each oocyte to multiple plant odorants was recorded via a two-electrode voltage clamp (OC-725C oocyte clamp, Warner Instruments, Hamden, CT, USA) at a holding potential of -80 mV. Data were acquired by using a Digidata 1440 A and were analyzed by pCLAMP 10.2 software (Molecular Devices, San Jose, CA, USA).

Stock solutions of each odorant were prepared at 1 M using DMSO as a solvent, and each odorant was diluted in 1× Ringer’s solution (96 mM NaCl, 2 mM KCl, 5 mM MgCl₂, 0.8 mM CaCl₂, and 5 mM HEPES, pH 7.6) to the indicated concentrations for electrophysiological recording. Information on the 66 odorants used in this study is listed in Table S5.

QUANTIFICATION AND STATISTICAL ANALYSIS

All quantification and statistical analysis methods are described in the Figure legends and Table legends. Quantification of compounds in HP extracts was done using Agilent OpenLab CDS software (version A.01.07). The distribution of 16:OAc and MeSA in various structures of the HPs and the effects of extraction duration were assessed using ANOVA followed by Tukey’s Honestly Significant Difference test (JMP). The time-course of cumulative mating success was analyzed using survival analysis. A Cox Proportional Hazards model was fitted and the Wald test was used to compare treatments. Dunnett’s multiple comparison test was used to compare supplementation with various doses of MeSA to the no-supplementation control (0-ng treatment). For all statistical analyses α = 0.05.