Identification and Behavioral Assays of Alarm Pheromone in the Vetch Aphid Megoura viciae

Aphids are destructive pests, and alarm pheromones play a key role in their chemical ecology. Here, we conducted a detailed analysis of terpenoids in the vetch aphid, Megoura viciae, and its host plant Pisum sativum using gas chromatography/mass spectrometry. Four major components, (-)-β-pinene (49.74%), (E)-β-farnesene (32.64%), (-)-α-pinene (9.42%) and ( +)-limonene (5.24%), along with trace amounts of ( +)-sabinene, camphene and α-terpineol) (3.14%) were found in the aphid. In contrast, few terpenoids were found in the host plant, consisting mainly of squalene (66.13%) and its analog 2,3-epoxysqualene (31.59%). Quantitative analysis of the four major terpenes in different developmental stages of the aphid showed that amounts of the monoterpenes increased with increasing stage, while the sesquiterpene amount peaked in the 3rd instar. (-)-β-Pinene was the most abundant terpene at all developmental stages. Behavioral assays using a three-compartment olfactometer revealed that the repellency of single compounds varied in a concentration-dependent manner, but two mixtures [(-)-α-pinene: (-)-β-pinene: (E)-β-farnesene: ( +)-limonene = 1:44.4:6.5:2.2 or 1:18.4:1.3:0.8], were repellent at all concentrations tested. Our results suggest that (-)-α-pinene and (-)-β-pinene are the major active components of the alarm pheromone of M. viciae, but that mixtures play a key role in the alarm response. Our study contributes to the understanding of the chemical ecology of aphids and may help design new control strategies against this aphid pest.

Megoura viciae feed exclusively on Fabaceae, causing serious damage to the broad bean Vicia faba and the pea Pisum sativum (Kunert et al. 2008;Leroy et al. 2011). Its unique composition of terpene alarm pheromone components differentiates it from most other aphid species. It has been shown, by rearing aphids on artificial diets and antibiotics, that the cotton aphid, Aphis gossypii, synthesizes alarm pheromone itself (Sun and Li 2017). However, it is unclear whether other aphids also synthesize alarm pheromone de novo or sequester it from host plants or symbionts. Our group has been working on the biosynthesis of aphid alarm pheromone in species that utilize (E)-β-farnesene, including the green peach aphid Myzus persicae (Cheng and Li 2018;Zhang and Li 2008;, A. gossypii (Ma et al. 2010;Sun and Li 2017;Sun and Li 2018) and the bird cherry-oat aphid Rhopalosiphum padi (Zhang and Li 2012;. Thus, M. viciae provides an opportunity for a comparative study on species that use other compounds. In this study, we first analyzed the composition of terpenoids in M. viciae and its host plant Pisum sativum using gas chromatography/mass spectrometry (GC/MS). Next, we investigated the quantities of the major terpenes in different developmental stages of M. viciae. Finally, we conducted a series of behavioral assays in a three-compartment olfactometer testing responses to single compounds and mixtures at different concentrations. Our study identified a novel set of alarm pheromone components in M. viciae.

Culture of Aphids
Aphids for our colony were provided by the Laboratory of Biological Control (Dr. Tinghui Liu, Hebei Agricultural University). They were maintained on P. sativum in the Laboratory of Insect Molecular Ecology at China Agricultural University and reared in a climate incubator (RXZ-300B, Ningbo, China) at 19 ± 1 ℃, 70 ± 5% RH with a photoperiod of 16L:8D.

Collection of Terpenoids from M. Viciae and its Host Plant P. Sativum
Aphids (overlapping developmental stages, n = 200) were collected in a 1.5 ml centrifuge tube, on ice, containing 500 μl of n-hexane. They were then homogenized and centrifuged at 4 ℃ for 30 min, and the supernatant transferred to a 2 ml vial for GC/MS analysis. A similar procedure was used to collect terpenoids from P. sativum seedlings (2.0 g).

Identification of Terpenoids from M. Viciae and P. Sativum by GC/MS
Samples were analyzed on an Agilent 6890-5973 (Agilent Technologies Inc., California, USA), equipped with a HP-5 capillary column (300 mm × 0.25 mm × 0.25 μm, Agilent, Santa Clara, USA). The column oven program was 40 °C for 1 min, followed by an increase to 130 °C at 4 °C.min −1 , maintained for 5 min, and then increased at 10 °C.min −1 to 250 °C. The injector and ion source temperatures were set to 250 °C. The MS was operated with electron impact ionization (El, 70 eV) and a scan range of m/z 35-650. Terpenes were identified by comparing retention times and mass spectra with standards (SigmaAldrich, Oakville, Canada). The quantity of each component was estimated by the peak area ratio of the sample to external standard of (-)-β-pinene (Magdalena and Henryk 2017). Three replicates were analyzed for each treatment. The proportions of single components were calculated as percentages of total terpenoids.

Quantitative Analysis of Terpenes In Aphid Developmental Stages
Aphids of the same developmental stage (n = 30), including 1 st , 2 nd , 3 rd , 4 th instars or adults, were homogenized in a 1.5 ml centrifuge tube containing 100 μl hexane. The supernatant was transferred to a vial for GC/MS analysis as described above. (-)-β-Pinene (purity > 99%; Sigma-Aldrich, Oakville, Canada) was used as the external standard. Three replicates were analyzed for each stage. The amount of each terpene was calculated based on the peak area ratio of the sample to standard.

Behavioral Assays
Behavioral assays were carried out in a three-compartment plexiglass olfactometer modified from a previous design (Khashaveh et al. 2020;Satyajeet et al. 2021;Yu et al. 2019) (Fig. 1). The olfactometer comprised three compartments (each 7 cm × 13 cm × 5 cm) connected by a door (3 cm × 3 cm) between two adjacent compartments. The samples, (-)-α-pinene, (-)-β-pinene, ( +)-limonene and (E)-β-farnesene were each diluted to three concentrations (0.1, 1 and 10 μg/μl) with light mineral oil. Two mixtures (Mix I and Mix II) were prepared in ratios of 1:44.4:6.5:2.2 and 1:18.4:1.3:0.8 of the four terpenes, respectively. These mixtures corresponded to the ratios of the four terpenes in the 3 rd and 4 th instars, respectively. Light mineral oil was used as a negative control. Sample (B) and control (C) were placed in a petri dish (diam. 3.0 cm) near the door of the side compartments, with five host plants fixed in smaller petri dishes (diam. 1.0 cm; covered with 10% agar) placed in the far side of the lateral compartments. Twenty wingless 3 rd -and 4 th instars were introduced into the petri dish in the middle (A) and allowed to move freely for 30 min. A total of 100 nymphs was used to test preference to each sample. All behavioral assays were performed in the dark to avoid light interference. At the end of a test, the numbers of aphids crawling close to the host plants in the two lateral compartments were counted and a behavioral index value (BIV) calculated according to the formula BIV = [(C − T)/ (C + T)] × 100, where C and T are the numbers of aphids in the control and sample compartments, respectively. Behavioral responses were categorized into four types: NR (no response, BIV < 20%), W (weak, 20 < BIV < 40%), M (moderate, 40% < BIV < 60%) and S (strong, BIV > 60%) (Hieu et al. 2014;Khashaveh et al. 2020).

Quantitative Analysis of Terpene Compounds in Aphid Developmental Stages
The contents of the four major terpenes in different developmental stages (1 st , 2 nd , 3 rd , and 4 th instar, and adult) of  the aphid were investigated. The amount of (-)-β-pinene increased from 1 st to 2 nd instar, the amount of (E)-βfarnesene increased from 2 nd to 3 rd instar, while the amounts of ( +)-limonene and (-)-α-pinene increased from 3 rd to 4 th instars (Fig. 3 top). All components were at a high level in 4 th instars. As a general trend, the amounts of the monoterpenes increased with development, while the amount of (E)β-farnesene increased from 1 st to 3 rd instar, and decreased slightly by the adult. The percentages of terpenes were calculated for different developmental stages (Fig. 3 bottom); the most abundant component throughout was (-)-β-pinene (> 81%). The ratios of the four major terpenes for 3 rd and 4 th instars were used for preparing the two mixtures used in the behavioral assays.

Behavioral Responses of m. Viciae to Single Terpenes and Mixtures
The responses of M. viciae to single terpenes and mixtures were recorded in the olfactometer (Table 3 and    repellency was observed for all single compounds, but the two mixtures had moderate repellency (F 17, 72 = 5.748, P < 0.05).

Discussion
GC/MS analysis of M. viciae identified four major terpenes, namely (-)-β-pinene (49.74%), (E)-β-farnesene (32.64%), (-)-α-pinene (9.42%) and ( +)-limonene (5.24%), in addition to some minor components (3.14%). Compared with previously reported data (Francis et al. 2005), the proportion of (E)-β-farnesene in our study was much higher (32.64% vs 14.2%), while the proportion of (-)-β-pinene much lower (49.74% vs 74.0%). In our study, the samples used for GC/MS analysis contained both winged and wingless aphids at different developmental stages. Thus, differences in aphid developmental composition might explain the difference in chemical proportions. Alternatively, it is possible the difference may be explained by the use of different populations of aphids. Indeed, the terpenoid composition of two geographical populations (Herault and East Pyrenees) of the aphid Pinus nigra from two different locations, Herault and the East Pyrenees, were found to be different (Bojovic et al. 2005). The ecological significance of our results needs further investigation. Quantitative analysis of the four major terpenes in M. viciae across different developmental stages revealed changes, with the relative amounts of monoterpenes increasing with development, while the sesquiterpene amount peaked at 3 rd instar. All the terpenes remained at a high level in the 4 th instar, with (-)-β-pinene being most abundant at all stages. This is the first report of developmental changes in the composition of terpenoids in an aphid species.
Behavioral assays revealed that the repellency of the individual components was concentration-dependent but not so for the mixtures. All single components were repellent to M. viciae at 1.0 μg/μl or above but had no or very weak repellency at 0.1 μg/μl. By contrast, the two mixtures were moderately repellent to M. viciae at all concentrations tested. Overall, our results suggest that (-)-α-pinene and (-)-β-pinene are the major active components of the alarm pheromone of M. viciae, but that a specific blend of terpenes can play a key role in the alarm response. In a previous study testing five terpenes, (-)-α-pinene, ( ±)-α-pinene, β-pinene, ( +)-limonene and (E)-β-farnesene, the authors found that ( ±)-α-pinene, β-pinene and (E)-β-farnesene individually did not repel M. viciae, although did so when combined (Bruno et al. 2018).
Lastly, our results revealed that the host plant P. sativum did not contain any of the aphid alarm pheromone components. This suggests that alarm pheromone is not sequestered directly from the plant but is synthesized de novo in M. viciae (Sun and Li 2017).
In summary, we identified four terpenes in M. viciae that exhibited changes in amount across the developmental stage. Behavioral assays revealed that the single compounds repelled aphids in a concentration-dependent manner, while specific mixtures of the four compounds were repellent at all concentrations tested. In addition to increasing our understanding of the chemical ecology of aphids, our work should help the design of alternative control strategies for M. viciae.