Validation of metabolites in resistant vs. susceptible maize lines
Using 29 of the most resistant and 28 of the most susceptible maize lines in this study, we sought correlations between the presence of zeaxanthin, ß-carotene, lutein, chlorophyll a, chlorophyll b, and pheophytin b, with 7- and 14-day damage rating. The carotenoids were measured because of the multiple carotenoid-related pathways identified by PAST, as were the chlorophylls and pheophytins. Pheophytins are identical in structure to chlorophylls but lack magnesium(II). Chlorophylls can be degraded to pheophytins under low pH or enzymatically. Along with chlorophylls and ß-carotene pheophytins are also involved in photosynthesis (photosystem II). We were unable to find measurable levels of zeaxanthin in the leaf samples, which is unsurprising as this compound tends to be produced in maize grains. How the biosynthesis pathway is then associated with reduced leaf feeding by FAW is thus unknown. The four other metabolites measured showed phenotypic variation among the maize lines tested, although about 80% of the lines showed low levels, with less than 300 µg/g of lutein, 2,000 µg/g of ß-carotene, 3E6 counts/g of chlorophyll a, 2E6 counts/g of chlorophyll b and 2E6 counts/g of pheophytin b (Supplemental Table 5).
Levels of these four metabolites were correlated with each other, and some with damage ratings (Fig. 2A). Lutein and pheophytin b were correlated with an increase in FAW damage, and chlorophyll a was correlated with a decrease in FAW damage, in agreement with pathway PWY-6927 (chlorophyll a degradation II) being associated with increased FAW damage ratings at 7 days (and less significantly so at 14 days, along with pathway PWY-5098, chlorophyll a degradation I), (Supplemental Table 4). Chlorophyll b is positively correlated to ß-carotene, lutein and pheophytin but negatively correlated with chlorophyll a. The chlorophyll cycle converts a to b forms of chlorophyll via a step through the intermediary 3,8-divinyl-chlorophyllide a; thus, it makes sense that a and b are negatively correlated and also that pathway CHLOROPHYLL-SYN, which produces 3,8-divinyl-chlorophyllide a, is associated with increased damage in the PAST analysis. ß-carotene was positively correlated to pheophytin and lutein, which were correlated to FAW damage, but ß-carotene itself was not significantly correlated with FAW damage scores. These metabolites appear to agree with the metabolic pathway analysis that found carotenoids to be important in FAW resistance but are not conclusive.
The wide distribution of phenotype in relation to metabolite content made straightforward conclusions difficult. The cluster analysis (Fig. 2B) and PCA (Fig. 2C) of metabolites and FAW ratings showed six groups, formed based on FAW rankings and also metabolite content. For example, two of the clusters showed high levels of lutein content, one of which were FAW-resistant and the other FAW-susceptible. For a clearer breakdown of metabolite and resistance phenotype, Supplemental Fig. 2 shows the levels of each metabolite within each of the six clusters; it is apparent that some of the clusters are highly significantly different than the others for levels of that metabolite. Although the numbers of lines within each cluster reduce the statistical power of the analysis, it is likely that within the resistant lines, there is more than one mechanism leading to resistance, but that levels of lutein, pheophytin, and chlorophyll a in particular are almost certainly contributing to resistance (or susceptibility). Although the effect of each metabolite on resistance may not be enough to create resistant lines, they do help to validate the results of the PAST metabolic pathway analysis.
Figure 2 : Targeted metabolite analysis for lutein, β-carotene, chlorophyll a and b and pheophytin b. A) Correlation matrix, B) dendrogram, and C) principal component analysis (PCA) of metabolite and Fall Armyworm rating analysis. In the dendrogram, maize lines are colored salmon (BEST) and blue (WORST) representing resistance phenotype. In the dendrogram and PCA, clusters are represented with the following colors, 1 = black, 2 = blue, 3 = yellow, 4 = green, 5 = purple and 6 = red. In the correlation matrix graph, blue represents positive correlation, red negative correlation and black “X” indicates that the comparison was not significantly correlated with a p-value cut-off of 0.05. In the PCA, shape of points represents resistant phenotype, circle (BEST) and triangle (WORST).