Comparative Analysis of Transcriptome and sRNAs Expression Patterns in the Brachypodium distachyon- Magnaporthe oryzae Pathosystems

Abstract

The hemibiotrophic fungus Magnaporthe oryzae (Mo) is the causative agent of rice blast and can infect aerial and root tissues of a variety of Poaceae, including the model Brachypodium distachyon (Bd). To gain insight in gene regulation processes occurring at early disease stages, we comparatively analyzed fungal and plant mRNA and sRNA expression in leaves and roots. A total of 310 Mo genes were detected consistently and differentially expressed in both leaves and roots. Contrary to Mo, only minor overlaps were observed in plant differentially expressed genes (DEGs), with 233 Bd-DEGs in infected leaves at 2 days post inoculation (DPI), compared to 4978 at 4 DPI, and 138 in infected roots. sRNA sequencing revealed a broad spectrum of Mo-sRNAs that accumulated in infected tissues, including candidates predicted to target Bd mRNAs. Conversely, we identified a subset of potential Bd-sRNAs directed against fungal cell wall components, virulence genes and transcription factors. We also show a requirement of operable RNAi genes from the DICER-like (DCL) and ARGONAUTE (AGO) families for fungal virulence. Overall, our work elucidates the extensive reprogramming of transcriptomes and sRNAs in both plant host (Bd) and fungal pathogen (Mo), further corroborating the critical role played by sRNA species in the establishment of the interaction and its outcome.

Keywords: RNAi; argonaute; dicer; gene expression; plant disease; small RNA; virulence.

Figure 1

Predicted 3D structures of (A) M. oryzae MoAGO1 and (B) MoAGO3. Structures were modeled with SWISS-MODEL and domains were highlighted with PyMOL: blue = N-domain, red = DUF1785, yellow = PAZ (Piwi Argonaut and Zwille), green = L2, purple = PIWI (P-element Induced WImpy testis), the rest of the sequence with no domain predicted is colored in grey.

Figure 2

Infection assays of M. oryzae RNAi mutants. (A) Detached second youngest leaves of three-week-old Bd seedlings were drop inoculated with 10 μL Mo conidia suspension 50 × 10³ spores mL⁻¹ in 0.002% Tween20 and kept under high humidity at 16 h light/8 h dark cycle at 22 °C/18 °C. Fungal pathogenicity was assayed via ImageJ software at 5 DPI. The experiment was conducted two times (n = 8 plants per experimental group) with similar results. Comparisons between groups was performed via ANOVA and Tukey’s range test for multiple comparisons. (B) Detached Bd leaves were spray inoculated with a total of 250 μL Mo conidia suspension 50 × 10³ spores mL⁻¹ in 0.002% Tween20 and kept and evaluated as in (A). The experiment was conducted two times (n = 8 plants per experimental group) with similar results. Comparisons between groups was performed via Welch one-way test coupled with pairwise t-tests with Benjamini–Hochberg p-value adjustment. (C) Three-week-old Bd seedlings were spray inoculated with Mo conidia suspension 120 × 10³ spores mL⁻¹ in 0.002% Tween20 and kept and evaluated as in (A). The experiment was conducted two times (n = 18 plants per experimental group) with similar results. Comparisons between groups was performed via Kruskal–Wallis test and Dunn’s test of multiple comparisons. (D) Roots of seven-day-old Bd seedlings were inoculated with 1 mL of Mo conidia suspension 125 × 10³ spores mL⁻¹ in 0.002% Tween20 for 3 h. Afterwards, seedlings were transplanted in small pots (3 cm Ø) and grown for an additional 5 days before harvesting. Fungal amount was calculated by qPCR based on the ratio of fungal actin (MoActin). The experiment was conducted two times (n = 6 roots per experimental group) with similar results. Comparisons between groups was performed via ANOVA and Tukey’s range test for multiple comparisons. (A–D) Letters represent statistical difference among all groups’ means (α < 0.05). Asterisks represent statistical difference of each group against wildtype (wt) (* p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001).

Figure 3

Brachypodium blast caused by M. oryzae Mo 70-15 on leaves and roots of B. distachyon Bd21-3. (A,D) Detached 21-day-old Bd leaves were drop-inoculated with 10 μL of Mo suspension (50,000 conidia/mL) and kept for (A) 2 and (C) 4 days at high humidity, or mock-inoculated (B,D) with Tween water. Roots of 7-day-old Bd seedlings were (E) drop-inoculated with 1 mL of Mo suspension (250,000 conidia/mL) or (F) mock-inoculated or and kept for 4 days under high humidity at 16 h light/8 h dark cycle at 22 °C/18 °C.

Figure 4

Volcano plots of DESeq2 results based on mRNAseq analysis of M. oryzae-infected leaves and roots of B. distachyon vs. control. Differentially expressed genes (DEGs) are highlighted in red with significant adjusted p-values (padj < 0.05).

Figure 5

Venn diagram of differentially expressed B. distachyon and M. oryzae genes. Significantly (A) downregulated (fold change (FC) < 0 padj < 0.05) and (B) upregulated (FC > 0 padj < 0.05) Bd genes shared between setups. Significantly (C) downregulated (FC < 0 padj < 0.05) and (D) upregulated (FC > 0 padj < 0.05) Mo genes shared between setups (“Leaf 2 DPI”, “Leaf 4 DPI” and “Root”). Transcript downregulation was calculated from mRNAseq data with DESeq2.

Figure 6

Expression of predicted ARGONAUTE (AGO) and DICER-like (DCL) during the interaction of B. distachyon and M. oryzae from mRNAseq results. Normalized read counts of each RNAi component were retrieved from DESeq2 analyses of infected datasets: leaf 2 DPI (white), leaf 4 DPI (grey) and root (black).

Figure 7

Size distribution of unique sRNA reads in the interaction of M. oryzae and B. distachyon. Relative size distribution (in percentage) of unique filtered sRNA reads assigned to (A) Mo or (B) Bd. Reads were assigned to either Mo or Bd only if aligning 100% to the organism of origin genome and had at least two mismatches to the interacting organism genome. (C,D) Relative size distribution of unique filtered sRNA reads assigned to (C) Mo or (D) Bd and induced or increased in infected samples compared to controls (i.e., axenic fungal cultures and non-inoculated plants, respectively). Samples for sRNA sequencing by Illumina HiSeq1500 were taken from different setups: leaf biotrophic phase (“2 DPI”), leaf necrotrophic phase (“4 DPI”) and root (“root”).

Figure 8

Venn diagrams of unique filtered sRNAs reads from M. oryzae and B. distachyon. (A) Venn diagram of Mo sRNA reads (18–32 nt) from axenic culture and Mo-infected Bd leaves (2 DPI, 4 DPI) and roots (4 DPI). (B) Venn diagram of Bd sRNA reads (18–32 nt) in non-infected and Mo-infected Bd leaves.

Figure 9

Size composition of unique candidate sRNA effectors from M. oryzae and B. distachyon. The number of unique reads is reported on the y-axis.

Figure 10

Venn diagram of downregulated predicted RNAi targets in M. oryzae and B. distachyon. Significantly downregulated (FC < 0 padj < 0.05) (A) Bd mRNA targets with complementarity to putative Mo sRNAs effectors shared between setups. (B) Mo mRNA targets with complementarity to putative Bd sRNA effectors. Setups: Leaf biotrophic phase (2 DPI) Leaf necrotrophic phase (4 DPI), and root. Transcript downregulation was assessed from mRNAseq data with DESeq2.