Targeted metabolite confirmation analysis
To test differential production of metabolites identified via PAST, leaves of the 31 most resistant lines in the study and 28 of the most susceptible lines based on the 7-day visual ratings from 2019 and 2020 were collected in 2021. Best and worst lines are indicated in Supplemental Table 1 in green or red highlight, respectively. Leaf tissue was collected from 10 plants of two replications of each genotype after the 14-day damage rating scores had been collected. Leaves from each plot of each selected genotype (except Mp704 and Mp707, which had too few live plants in the field) were harvested and flash frozen in liquid nitrogen and then used in a confirmation study whereby carotenoids and chlorophylls were extracted and analyzed for preferential abundance between the selected lines. This was done as previously described (Wojdyło et al., 2021) with the modification that magnesium carbonate (MgCO3, 50 mg) was added to lyophilized leaf samples (500 mg) to prevent isomerization. The samples were then extracted with 5 ml hexane:acetone:methanol (2:1:1, v/v/v) containing 1% BHT for 1 h in the dark on an orbital shaker (5000 rpm). The extracts were filtered through cotton plugs and the filtrates were concentrated via speedvac (Savant, Thermo Scientific). Each extract was re-dissolved in methanol (1 ml) and particulates were removed via centrifuge.
For carotenoid and chlorophylls analysis, the re-dissolved, centrifuged extracts were analyzed on a Waters Acquity UPLC system coupled to a Waters Xevo G2 XS Quadrupole Time-of-Flight (QTOF) mass spectrometer (MS). The QTOF MS was equipped with a Z-spray ionization source running in ESI+ mode using MassLynx 4.2 software with the following settings: source temperature: 100 °C, desolvation temperature: 250 °C, desolvation gas flow: 600 L/h, cone gas flow: 50 L/h, capillary voltage: 3.0 kV, sampling cone voltage: 40 V. Analyses were performed in sensitivity and continuum mode, with a mass range of m/z 50–1200 and a scan time of 0.1 s. A data-independent acquisition method with elevated collision energy (MSE) was used with 6 eV low energy and a high energy ramp from 15−45 eV. Separation was performed on a Waters BEH C18 1.7 µm, 2.1 x 50 mm column with the following gradient solvent system: (0.5 ml/min, solvent A: 0.1% formic acid in water; solvent B: 7:3 acetonitrile:methanol): 75% B (0.0-0.6 min.), gradient to 95.1% B (0.6-6.5 min.), gradient to 100% B (6.5-13.6 min.), 100% B (13.6-14.6 min.), then column equilibration to 5% B (14.6-17.5 min.). Data were analyzed on Waters UNIFI 1.9.4 software using the “Quantify Assay Tof 2D” analysis method with lock mass (leucine-enkephalin) corrected by UNIFI. Analytical standards (lutein and zeaxanthin) were purchased from ChromaDex and used to confirm identity of analytes and generate standard curves. Lutein and zeaxanthin are reported in µg carotenoid/g leaf sample. Relative amounts of chlorophyll a, chlorophyll b, and pheophytin b are reported as intensity counts/g leaf sample.
Normality of distribution on the metabolite levels was tested using QQplots and by performing a Shapiro-wilk test in R (Team, 2015). To achieve normality, metabolite content levels were transformed using log base 10. Normalized metabolite data and FAW ratings for day 7 and 14 from the 2021 field season were the input variables used to perform a correlation analysis using hierarchical clustering with 0.095 confidence intervals and 0.05 cut-off p-value in R. The data was scaled then used to generate a dendrogram by applying the Euclidean distance method and clustered by using the complete agglomeration method in R. Principal component analysis was performed using the same input variables used for the dendrogram. The first two PCs were used to create a PCA plot and variance by variable and individual was extracted from the analysis. Transformed and scale variables were plotted in relation to the cluster identity.
Results and Discussion