Integrated transcriptome and hormone profiling highlight the role of multiple phytohormone pathways in wheat resistance against fusarium head blight

Abstract

Fusarium head blight (FHB or scab) caused by Fusarium spp. is a destructive disease of wheat. Since the most effective sources of FHB resistance are typically associated with unfavorable agronomic traits, breeding commercial cultivars that combine desired agronomic traits and a high level of FHB resistance remains a considerable challenge. A better understanding of the molecular mechanisms governing FHB resistance will help to design more efficient and precise breeding strategies. Here, multiple molecular tools and assays were deployed to compare the resistant variety Sumai3 with three regionally adapted Canadian cultivars. Macroscopic and microscopic disease evaluation established the relative level of Type II FHB resistance of the four varieties and revealed that the F. graminearum infection process displayed substantial temporal differences among organs. The rachis was found to play a critical role in preventing F. graminearum spread within spikes. Large-scale, organ-specific RNA-seq at different times after F. graminearum infection demonstrated that diverse defense mechanisms were expressed faster and more intensely in the spikelet of resistant varieties. The roles of plant hormones during the interaction of wheat with F. graminearum was inferred based on the transcriptomic data obtained and the quantification of the major plant hormones. Salicylic acid and jasmonic acid were found to play predominantly positive roles in FHB resistance, whereas auxin and ABA were associated with susceptibility, and ethylene appeared to play a dual role during the interaction with F graminearum.

Fig 1. Disease assessment of four wheat varieties.

(A) Macroscopic disease symptoms are indicated as the percentage of bleached or brown rachis internodes (grey bar) and spikelets (white bar) above or below the inoculation site over a three-week period. Values represent means ± standard error. For each variety, the Student’s t test was performed between the percentage of infected rachis internodes and spikelets. * indicates P < 0.05. Among the four varieties, a one-way ANOVA of data based on infected rachis internodes was performed at each time point at α = 0.05 to determine significance. Histograms with different letters are statistically different. SU, Sumai3; FL, FL62R1; ST, Stettler; MM, Muchmore. (B) Microscopic observation of the F. graminearum infection process in the rachis of the four varieties at 2 wpi. The cross-section of rachis internodes and rachillas below the inoculated site were dissected individually and stained with Wheat Germ Agglutinin (WGA). Each sample was photographed separately under fluorescence and light microscope. The scale bar indicates 200 μm. A schematic illustration of a wheat spike is shown on the left-hand side. Red lines indicate the approximate position of hand cross sections in sequential rachises. Each row of photographs represents the same cross section level of each of the four varieties. For the mock treatment, only the adjacent rachilla (Ra0) and rachis internode (Rn0) of mock-treated spikelet were examined and shown in the bottom two rows. Ra, Rachilla; Rn, Rachis internode.

Fig 2. Principal component analysis of the transcriptomes of the two organs from each variety at each time point.

Principal component analysis (PCA) was performed using the prcomp function from the R base package. Ellipses (confidence level = 0.9) are graphical representations of samples at different dpi. Light yellow areas indicate spikelet-based transcriptomes; grey areas indicate rachis-based transcriptomes. Coloured ovals indicate different time points. Ra, Rachis; Sp, Spikelet; D, dpi.

Fig 3

GO enrichment for DEGs activated (A) and repressed (B) following F. gramineaerum inoculation. Heatmap represents the strenghts of the p values of GO term overrepresentation in the F. graminearum-responsive DEG sets that become significantly changed for the first time at the given time points. Colour index represents level of significance (p values).

Fig 4. Gene Ontology (GO) enrichment analysis for phytohormone related pathways.

Gene ratio equals the number of differentially expressed genes against the number of genes associated with a GO term in wheat genome. The Fisher test was performed to indicate the significance of GO enrichment (FDR < 0.05).

Fig 5

Hormone and transcriptome profiling of defense related phytohormone pathways: salicylic acid (A), jasmonic acid (B) and ethylene (C) after F. graminearum infection. Left panels: hormone content, Values = means ± standard error (n = 3). In each organ, a two-way ANOVA of data was performed at α = 0.05 to determine significance. Histograms with different letters are statistically different. Right panels: extent of differential expression (Log2 fold change). Asterisks (*) denote significant (p ≤ 0.01). A full list of DEGs is provided in S7 Dataset.

Fig 6. Hormone and transcriptome profiling of auxin (IAA) and two of its amino acid conjugates.

(A) extent of differential expression patterns after F. graminearum infection (Log2 fold change). Asterisks (*) denote significant (p ≤ 0.01). A full list of DEGs is provided in S7 Dataset. (B) IAA, (C) IAA-Asp and (D) IAA-Glu contents in the spikelet and rachis. Values = means ± standard error (n = 3). In each organ, a two-way ANOVA of data was performed at α = 0.05 to determine significance. Histograms with different letters are statistically different.

Fig 7. Contents of ABA and its metabolites after F. graminearum inoculation.

ABA (A), PA (B) and DPA (C) contents in the spikelet and rachis. Values = means ± standard error (n = 3). In each organ, a two-way ANOVA of data was performed at α = 0.05 to determine significance. Histograms with different letters are statistically different.