2 MATERIALS AND METHODS
2.1 Study system
The Palouse region of eastern Washington and northern Idaho, USA, is home to many legumes including P. sativum (Black et al., 1998). In P. sativum fields, S. lineatus , a chewing herbivore, co-occurs with A. pisum , a phloem-feeding herbivore that can transmit pathogens such as PEMV (Chisholm et al., 2019). PEMV is one of several viruses that infects P. sativum , and this pathogen is obligately transmitted by aphids in a persistent manner (Chisholm et al., 2019).
Sitona lineatus adults overwinter outside of P. sativumfields and migrate into fields in late spring before A. pisumarrive (Cárcamo et al., 2018). After S. lineatus eggs hatch, larvae burrow into the soil to feed and pupate before emerging as adults in the summer (Cárcamo et al., 2018); these second-generation adults often occur on plants under attack from A. pisum and PEMV (Chisholm et al., 2019). Thus, S. lineatus attacks plants in the field both before and after A. pisum and PEMV. However, it is unknown if responses of P. sativum differ based on the number of stressors, and their order of attack. Moreover, molecular mechanisms that mediate interactions among these stressors are largely unknown (Chisholm et al., 2019; Bera et al., 2020).
To address these questions, we conducted greenhouse assays to assess interactions between S. lineatus , A. pisum, and PEMV onP. sativum plants, and molecular mechanisms affecting these interactions. First-generation adult S. lineatus for experiments were collected from commercial P. sativum fields, or wild patches of Vicia villosa , immediately prior to experiments. Colonies of infectious A. pisum with PEMV, and uninfectious A. pisum , were started from Palouse field-collected individuals (Chisholm et al., 2019) and were maintained on P. sativum plants in a greenhouse (21-24°C during day cycle, 16-18°C during dark cycle, 16:8 h light:dark).
2.2 Experimental design
We conducted a 3 × 2 greenhouse (21-24°C day cycle, 16-18°C dark cycle, 16:8 h light:dark) experiment that varied S. lineatus , A. pisum , and PEMV (Fig. 1). There were three S. lineatustreatments: (i) control: no adults prior to A. pisum treatments (none), (ii) two adults that fed for 48 h prior to A. pisumtreatments (first), and (iii) two adults that fed for 48 h afterA. pisum treatments (second). The two A. pisum treatments were: (i) sham: 10 5-d old uninfectious adults that fed for 48 h and (ii) PEMV: 10 5-d old PEMV-infectious adults that fed for 48 h. For treatments with S. lineatus first, they were removed by hand prior to A. pisum treatments; for treatments with A. pisumfirst, they were removed by aspirator prior to S. lineatustreatments. Treatments were conducted on individual P. sativumplants in mesh ‘bug dorms’ (0.6 × 0.6 × 0.6 m), with six replicates randomly assigned to each treatment in a factorial design (3 S. lineatus treatments × 2 A. pisum treatments). After insects were removed, plants were allowed to develop for 7 d before we harvested tissue to assess viral titer, gene expression, and nutrients. Tissue samples from the whole aboveground portion plants were collected and flash frozen in liquid nitrogen and stored in a -80°C freezer until processing. Viral titer samples confirmed that 100% of plants in the PEMV treatments became infected over the course of the experiment.
2.3 Analysis of plant defense and biosynthetic genes
Plant tissue was processed using liquid nitrogen in sterilized mortars and pestles. Powdered tissue samples (50 to 100 mg) were used for RNA extraction with Promega SV total RNA isolation kits (Promega, Madison, WI). The quantity and quality of RNA was estimated on a NanoDrop1000 and agarose gel electrophoresis, respectively and 1 µg of total RNA from each sample was used for cDNA synthesis (Bio-Rad iScript cDNA Synthesis kits). Gene specific primers (Table S3) for qRT-PCR were designed using the IDT Primer Quest Tool. Each qRT-PCR reaction (10 µl) was set up containing 3 µl of ddH2O, 5 µl of iTaq Univer SYBR Green Supermix (Bio-Rad), 1 µl of specific primer mix (forward and reverse [concentration 10µM]), and 1 µl of diluted (1: 25) cDNA template. Reactions were set up in triplicates for each sample and ran on a CFX96 qRT-PCR machine (Biorad). The qRT-PCR program included an initial denaturation for 3 min at 95°C, followed by 40 denaturation cycles for 15 s at 95°C, annealing for 30 s at 60°C, and extension for 30 s at 72°C. For melting curve analysis, a dissociation step cycle (55°C for 10 s and then 0.5°C for 10 s until 95°C) was added. The comparative 2−ΔΔCt method (Livak & Schmittgen, 2001; Kozera & Rapacz, 2013) was used to calculate the relative expression level of each gene, with β-tubulin as an endogenous control.
We assessed expression of seven genes associated with defense in peas. Gene sequences were obtained using accession numbers (available genes) or using Pea Marker Database (Kulaeva et al., 2017) and blast searching through the reference genome (Kreplak et al., 2019). Four genes were associated with plant hormone biosynthesis: (i) Isochorismate synthase1 (ICS1 ) (salicylic acid), (ii) Lipoxygenase 2(LOX2 ) (jasmonic acid), (iii) Aldehyde oxidase 3(AO3 ) (abscisic acid), and (iv) Gibberellin 2-oxidase(GA2ox ) (gibberellic acid). ICS1 converts chorismate to isochorismate, a precursor of salicylic acid biosynthesis (Seguel et al., 2018), while LOX2 is a precursor to jasmonic acid biosynthesis (Wasternack & Hause, 2013). AO3 catalyzes abscisic acid biosynthesis by oxidizing abscisic aldehyde, and GA2OXcatalyzes bioactive giberrelic acids or their immediate precursors to inactive forms (Zdunek-Zastocka & Sobczak, 2013; Serova, Tsyganova, Tikhonovich, & Tsyganov, 2019; He et al., 2019). All of these gene transcripts can affect plant defense and plant-microbe interactions (Lee et al., 2012; Yergaliyev et al., 2016).
The three additional genes examined were associated with defense response transcripts that occur downstream from hormone induction. One of these genes, Pathogenesis-related protein 1 (PR1 ) affects systemic acquired resistance-mediated defense signaling and occurs downstream in the salicylic acid pathway (Fondevilla, Küster, Krajinski, Cubero, & Rubiales, 2011; Miranda et al., 2017). The second defense response transcript was an antimicrobial defensin peptide called Disease resistance response gene (DRR230), which has been reported to provide resistance in peas against various pathogens (Lacerda et al., 2014; Selim, Sanssené, Rossard, & Courtois, 2017). The third defense response transcript assessed was Lectin (PsLectin) . Plant lectins are a group of carbohydrate binding proteins, and Lectin genes can be induced by salicylic acid, jasmonic acid, and herbivores to stimulate phytoalexin and pistatin production in peas (Fondevilla et al., 2011; Armijo et al., 2013; Macedo, Oliveira, & Oliveira, 2015).
2.4 Measurement of plant phytohormones
Plant tissue samples were assessed for three phytohormones: jasmonic acid, salicylic acid, and abscisic acid following procedures of Patton, Bak, Sayre, Heck, & Casteel (2019). Briefly, tissue samples were first flash frozen in liquid nitrogen before being lyophilized and weighed. Hormones were extracted in iso-propanol:H2O:HCL1MOL (2:1:0.005) with 100 μl of internal standard solution (1000 pg of each). Samples were evaporated to dryness, resuspended in 100 μl of MeOH, filtered, and 10 µl of each sample was injected into an Agilent Technologies 6420 triple quad liquid chromatography-tandem mass spectrometry instrument (Agilent, Santa Clara, CA). A Zorbax Extend-C18 column 3.0 × 150mm (Agilent, Santa Clara, CA) was used with 0.1% formic acid in water (A) and 0.1% (v/v) formic acid in acetonitrile (B) at a flow rate of 600 mL min–1. The gradient used was 0–1 min, 20% B; 1–10 min, linear gradient to 100% B; 10-13 min, 100% A. Retention times were: jasmonic acid (D5) standard (5.740 min), jasmonic acid (5.744 min), salicylic acid D4 standard (4.677 min), salicylic acid (4.720 min).
2.5 Analysis of plant nutritional components
For amino acid analysis, leaf tissue was lyophilized, weighed, and extracted with 20mM of HCL (Patton et al., 2019). Derivation was done using AccQTag reagents following the manufacturer’s instructions (Waters, Shinagawa-ku, Tokyo), and derivatised samples (10 μl) were then injected. Ground tissue was extracted with 100 μl of 20 mM HCl, centrifuged, and the supernatant was saved. Amino acids were derivatized using AccQ-Fluor reagent kits (Waters, Milford, MA), with L-Norleucine as an internal standard. 10 μl from each sample were injected with an Agilent 1260 Infinity pump with a Nova-Pak C18 column and fluorescence detector, and Agilent Chemstation software for data recording. Amino acid derivatives were detected with an excitation wavelength of 250 nm and an emission wavelength of 395 nm. Peak areas were compared to a standard curve made from a serial dilution of amino acid standards (Sigma-Aldrich, St. Louis, MO). injected into a Agilent 1260 Infinity HPLC (Agilent, Santa Clara, CA) with a Nova-Pak C18 column (Casteel et al., 2014). Solvent A, AccQ‐Tag Eluent A, was premixed from Waters; Solvent B was acetonitrile:water (60:40). The gradient used was 0–0.01 min, 100% A; 0.01–0.5 min, linear gradient to 3% B; 0.5–12 min, linear gradient to 5% B; 12–15 min, linear gradient to 8% B; 15–45 min, 35% B; 45–49 min, linear gradient to 35% B; 50–60 min, 100% B. The flow rate was 1.0 ml min−1. Amino acid derivatives were measured with an Agilent fluorescence detector with an excitation wavelength of 250 nm and an emission wavelength of 395 nm. For concentration calculations, standard curves were generated for each amino acid using dilutions of the standard.
2.6 Data analysis
To evaluate effects of our treatments on host-plant defenses and host-plant quality, we ran a series of multivariate models using R ver. 3.5.2 (R Working Group, 2018). First, gene expression was evaluated withICS1 , LOX2 , GA2ox , AO3 , PR1 ,DRR230 , and PsLectin as the responses, with MANOVA to assess treatment effects on relative gene expression (2–∆∆Ct ) based on cycle threshold values for each observed gene transcript. Estimated marginal mean of Ct values, and standard error of the mean, were generated using the emmeans package in R (Lenth, 2016). The methodology for 2–∆∆Ct followed modified recommendations from Rao, Huang, Zhou, & Lin (2013) and Kozera & Rapacz (2013), using housekeeping gene β-tubulin to normalize expression and a sham aphid (non-infective Pea aphid and no weevil addition) treatment as a control.
Hormone levels in plants were evaluated using MANOVA, with salicylic acid, jasmonic acid, and abscisic acid as responses (3 variables). Total amino acid content was evaluated using a generalized linear model (GLM) with total concentration among all amino acids as the response. All models assessed treatment effects, using S. lineatus addition,A. pisum infection status, and their interaction as predictors. Finally, changes in the amino acid profile was evaluated using non-metric multidimensional scaling (NMDS) with the vegan package (Oksanen et al., 2019) following Ceulemans, Hulsmans, Ende, & Honnay (2017).