Botanical extracts
Plants of Synandrospadix vermitoxicus (Griseb.) Engl were collected in December 2011, at Coronel Moldes, Salta, Argentina (25°14´21.1´´ S, 65°25´41.7´´ W). A voucher sample was deposited at the Museum of Natural Science of Salta, Argentina (number 11904-MCNS) (Fig. 1).
Fresh tubers were cut into 2–3 cm cubes (total weight 6 kg) and were macerated with ethanol at room temperature for 24 h, then the solvent was evaporated at reduced pressure to yield 50 g of ethanolic extract (Sv-E). A 20 g fraction of Sv-E was suspended in ethanol-water (3:1) solution and extracted successively with hexane (3 x 300 ml) to yield 1.092 g of hexanic extract (Sv-H) (Fig. 2). Both extracts were kept at 4°C until use in biological assays and to test for presence of flavonoids, alkaloids, sugars. The Sv-E extract (200 mg) was suspended in 96% ethanol to assess the preliminary phytochemical analysis for flavonoids and alkaloids.
The presence of flavonoids was assessed by the Shinoda’s test. Metallic magnesium (0.1 g) was added to 5 mL of the Sv-E extract followed by a few drops of concentrated hydrochloric acid; a reddish, violet, or orange colour indicates the presence of flavonoid (Lock de Ugaz 1994; Carvajal et al. 2008).
The presence of alkaloids was evaluated by three tests: (1) Dragendorff’s reagent (Iodine in potassium bismuth iodide), (2) Meyer’s reagent (potassium mercuric iodide solution), and (3) Warner’s reagent (Iodine in potassium iodide). Briefly, a solution of 5% HCl is added to 5 mL of Sv-E extract. Then, four drops of each reagent are added to 1 mL of this solution. For the Dragendorff’s reaction a yellow precipitate indicates the presence of alkaloids, while for the two other reagents a white precipitate indicates its presence (Lock de Ugaz 1994; Wagner and Blandt 1996).
Purification of Sv-H extract
The hexanic extract (1.092 g) was fractionated by silica gel flash chromatography and eluted with a gradient of increasing polarity of CH2Cl2 and mixtures of CH2Cl2:CH3OH to reach 95 fractions which were combined according to its similar profiles in thin layer chromatography analyses (TLC), visualized with UV-light, vanillin solution and sulfuric acid:ethanol mixture (1:4). Fractions eluted with CH2Cl2:CH3OH (3:0.5) were purified on TLC with CH2Cl2:CH3OH mixtures of different polarities to yield compounds number 7 and a mixture of 8 and 9, selected for analysis by 1H NMR.
Analysis of plant extracts by spectrometry (NMR) and chromatography (HPLC)
NMR spectra were recorded on a Bruker AVANCE II AV-400 operating at 400.13 MHz for 1H. Chemical shifts are given in ppm (δ) downfield from the TMS internal standard. Qualitative and quantitative analyses of sugars were carried out on an Agilent 1100 System, with an Agilent Hi-Plex Ca (Dúo) column, 300 x 6.5 mm (PL1F70-6850) 80ºC. The injection volume for Sv-E extract (2 mg in 200 µL Milli-Q water) and the standard was 10 µL, and the separation was carried out at a flow rate of 0.4 mL min− 1, and eluent of 100% Milli-Q water. Sugars were detected by refractive index at 30°C (Ball and Lloyd, 2011) using standards of sucrose, glucose, and fructose (Sigma®).
Qualitative and quantitative analyses of flavonoids were carried out on an Agilent 1100 System Nº G1365, with an Eclipse XDB-C18 column. The sample was prepared as follows: 2 mg of dry Sv-E was dissolved on 200 µL of methanol/hexane, 1:0.5 mL, v/v, then, it was centrifuged 10 min at 13.000 rpm. The hexane layer was discarded, and the methanol fraction was dried completely by rotavap. Sample was then dissolved for analysis in 200 µL methanol (modified after Torres et al. 2005). The extract and standard injection volume were 5 µL, and separation was carried out at a flow rate of 1 mL min− 1, using a mobile phase of water/acetic acid of 99.9:0.1, v/v; and a linear gradient of 15%-36.5%, 30 min of acetonitrile/acetic acid (99.9:0.1, v/v). Flavonoids were detected at 270 nm using standards of galic acid, chlorogenic acid, catechin, rutin, ferulic acid, quercetin, naringenin and kaempferol (Sigma®). All eluents used were of HPLC grade (Sintorgan). Sugar and flavonoid contents were quantified on the basis of standard curves constructed using area ratio of respective standard. Least-squares linear regression was used for this purpose.
Insects and plants
Myzus persicae Sulzer colonies were reared on radish (Raphanus sativus L). Aphids used in the experiments came from a colony maintained at the Faculty of Natural Science, National University of Salta, Argentina. This colony was initiated from a single virginoparous apterous individual collected in field in 2009. The colony was reared in a climate chamber at 22 ± 2°C, 30%-40% RH, and 16L:8D h photoperiod to induce parthenogenesis. A new colony was started every week, and newly moulted apterae adult aphids were used for the experiments.
Potato plants of Solanum tuberosum (L) cultivar PO 97.11.9 were provided by the germplasm bank of INTA-EEA Balcarce, (Balcarce, Buenos Aires, Argentina). Plants were propagated in vitro on Murashige and Skoog with vitamins, 3% sucrose w/v, pH 5.8. After two weeks on agar, the plantlets with developed roots were transferred to soil in 500 g plastic pots and maintained in a chamber at 22 ± 2ºC and 16L:8D h photoperiod.
Choice test to evaluate antifeedant activity
Antifeedant activity of the extracts was evaluated by studying the aphid’s preference over potato leaves treated with Sv-E or, Sv-H at different concentrations vs. control leaves, with a choice test. Choice tests were carried out in plastic cages, 100 mm diameter and 50 mm high, with four 15 mm diameter lateral ventilation holes sealed with non-woven mesh. The cage was closed at the top and bottom by a removable polystyrene Petri dish. The top Petri dish had a layer base of water-agar 1.5% (to prevent the leaves from dehydration) over which four leaves previously treated on the abaxial side, were attached in cross by its adaxial side to the agar. Then, to support the leaves, the dish containing the leaf was covered with a plastic lid with four 15 mm diameter holes to expose the leaves and finally placed on top of the cage (Lucatti, 2014) (Fig. 3). The treatments were: (A) Treated potato leaves sprayed with Sv-E extract at concentrations of 2.5, 5 and 10 mg/mL in a solution of Tween-20 at 0.05% v/v in water (emulsifier solvent); (B) Treated potato leaves sprayed with Sv-H extract previously diluted with acetone 5% at concentrations of 2.5, 5 and 10 mg/mL in a solution of Tween-20 at 0.05% v/v in water. For treatments (A) and (B) the control were leaves treated with each of the solvents respectively; and (C) Blank leaves, which consisted of natural untreated leaves. Blank treatment was included to test the effect that the Tween-20, 0.05% solution used as emulsifier has on aphids. Treated leaves were sprayed with 100 µL of the corresponding solution and air dried for 60 minutes to allow the evaporation of the solvent. Then, on each cage, two equally treated leaves were placed opposite and on cross side to two control leaves (Fig. 3a and 3b). Forty recently molted apterae adults of M. persicae were gently placed with a paintbrush at the centre of a polystyrene lid (arena) that was immediately used to close the plastic cage from the bottom. The number of aphids settled on each treated leaf was recorded at 5, 10, 15, 20, 25, 30, 45, 60, 90, 120, 180 and 240 minutes after the start of the experiment. Each treatment was repeated (biological replicates) 12–14 times for Sv-E, and 8 times for Sv-H. The experiment was performed in room under controlled conditions at 22 ± 2°C, 30%-40% RH, and normal light.
A settling inhibition index (%SI) was calculated for all the treatments at each time scored as %SI = [1- (%T/%C)] x 100, where %T is the percentage of aphids on the treated surface, and %C is the percentage of aphids on the control surface (Gutiérrez et al. 1997). For those cases where the control had lower values than the extracts treatments (then the negative values represent attraction for the extracts treatments, and the index can range from − 1 to -100), the index was modified as %SI = {[1- (%C/%T)] x 100}x -1.
Toxicity trial to evaluate aphicidal activity
The aphicidal activity of Sv-E, Sv-H on M. persicae was assessed by measuring the toxic effect over adult aphids and their nymphs maintained on artificial diet containing the Sv-E or Sv-H extracts. Three concentrations of each extract, 50, 100 and 1000 µg/mL, were administrated through aphid’s artificial diets (150 mM amino acids, 500 mM sucrose, vitamins, and minerals (diet composition modified by Douglas after Prosser and Douglas 1992). The control treatment consisted of aphids fed on artificial diet without the extracts. Five recently moulted apterae adult aphids were confined on plastic cages (3 cm diameter and 2 cm high) sealed on top with a diet sachet containing 100 µL of diet solution between two layers of Parafilm, and sealed at the bottom with a mesh (Fig. 3c and 3d) (Prosser and Douglas 1991; Koga et al. 2007; Machado-Assefh et al. 2015). Aphids were placed on the diet cages (14 and 15 replicates per control and treatments respectively) and maintained during the experiment in a climate chamber at 22 ± 2°C, 30%-40% RH, in darkness. After 96 h of treatment on artificial diets, the adult survival and the total number of offspring produced were counted per cage, then the mean number of nymphs per adult were calculated.
EPG to evaluate the probing behaviour
The EPG technique (McLean and Kinsey 1964; Tjallingii, 1978, 1985a,b, 1987) was used to monitor the probing behaviour of adult M. persicae aphids on potato leaves treated with Sv-E or, Sv-H extracts. Two leaves per plant was sprayed with 1 mL each of Sv-E or Sv-H extract, previously diluted with acetone 5%, at a concentration of 10 mg/mL in a solution of Tween-20 at 0.05% v/v in water (emulsifier solvent). The control were leaves treated with the solvents and the blank leaves consisted of natural untreated leaves. Blank treatment was included to test the effect that the emulsifier solvent has on aphids. After sprayed on both sides, treated leaves were air dried for 60 minutes to allow the evaporation of the solvent. A Giga-8 EPG-DC amplifier (EPG Systems, Wageningen, The Netherlands) was used to monitor eight wingless wired aphids placed on the stem of the plant, one per plant was recorded simultaneously for 4 h. The number of replicates obtained was 10. Signals were acquired and analysed using Stylet + software (EPG Systems).
During probing (periods where the insect is with the stylets inside the plant tissue, calculated without distinction of activities), six activities (EPG waveforms) were distinguished, and related to: (1) stylet pathway (C), reflecting the first electrical stylet contact with the epidermis, intercellular sheath salivation in epidermis and mesophyll, and extracellular stylet penetration movements in all plant tissues; (2) potential drop (pd), reflecting brief intracellular stylet punctures during C; (3) derailed stylet mechanics (F), reflecting stylet penetration difficulties; and (4) xylem ingestion (G), reflecting active sap ingestion from xylem elements. The phloem phase appears after an abrupt voltage drop caused by the intracellular penetration of sieve elements (SE) by the stylet tips and includes (5) SE salivation (E1), and (6) phloem sap ingestion (E2) with concurrent salivation (Tjallingii 1978b). An event was considered as an uninterrupted period of a waveform. A number of 127 standardized EPG variables were calculated automatically using an EXCEL Data Workbook CSIC-UAL elaborated by Garzo E. (Institute of Agricultural Sciences, CSIC, Spain) and Alvarez A.J. (University of Almería, Spain). The variables related to non-probing, probing, pathway, and phloem fall into five categories: (1) ‘number of’, mean number of times that an activity occurred per insect, (2) ‘mean duration of’, mean duration of an activity per insect per event, (3) ‘total time on’, total time on an activity per insect, (4) ‘time to’, mean time to the first occurrence per insect of an activity from the start of the experiment, and (5) ‘proportion of’, number or percentage of aphids that showed a particular activity.
Statistical analysis
In the choice test results, the differences between extracts treatments and control, and control vs. blank were analysed by the Binomial test, assuming the null hypothesis of no preference for the treatments; the analyses were performed using SISA online statistical analysis (https://www.quantitativeskills.com/sisa/distributions/binomial.htm) (Uitenbroek 1997).
The %SI were analysed by Kruskal–Wallis nonparametric analysis of variance (ANOVA), P ≤ 0.05, at one way of classification, followed by multiple comparisons with Conover test between means of treatments with Bonferroni’s correction (P ≤ 0.008.) The toxicity of the extract was analysed by means of adults’ survival and their nymphs’ survival, using the Kruskal–Wallis ANOVA, P ≤ 0.05 followed by multiple comparisons with Conover test with Bonferroni’s correction (P ≤ 0.002) (Weisstein 1999).
The dose that kills 50% of the aphids in the population (LD50) of each extract was obtained using GraphPad Prism 6.01 (Swift 1997). To do so, an inhibition graph was made indicating the percentage mortality of M. persicae adults as a function of dose concentration.
The EPG variables were analysed with a Mann-Whitney U test, with P < 0.05. The Fisher’s exact test was used to evaluate the difference in proportions of individuals performing each type of activity. The statistical analyses were performed using Infostat Profesional v2015 (http://www.infostat.com.ar) (Di Rienzo et al. 2015).