Comparative analysis of the root transcriptomes of cultivated and wild rice varieties in response to Magnaporthe oryzae infection revealed both common and species-specific pathogen responses

Abstract

Background: Magnaporthe oryzae, the causal fungus of rice blast disease, negatively impacts global rice production. Wild rice (Oryza rufipogon), a relative of cultivated rice (O. sativa), possesses unique attributes that enable it to resist pathogen invasion. Although wild rice represents a major resource for disease resistance, relative to current cultivated rice varieties, no prior studies have compared the immune and transcriptional responses in the roots of wild and cultivated rice to M. oryzae.

Results: In this study, we showed that M. oryzae could act as a typical root-infecting pathogen in rice, in addition to its common infection of leaves, and wild rice roots were more resistant to M. oryzae than cultivated rice roots. Next, we compared the differential responses of wild and cultivated rice roots to M. oryzae using RNA-sequencing (RNA-seq) to unravel the molecular mechanisms underlying the enhanced resistance of the wild rice roots. Results indicated that both common and genotype-specific mechanisms exist in both wild and cultivated rice that are associated with resistance to M. oryzae. In wild rice, resistance mechanisms were associated with lipid metabolism, WRKY transcription factors, chitinase activities, jasmonic acid, ethylene, lignin, and phenylpropanoid and diterpenoid metabolism; while the pathogen responses in cultivated rice were mainly associated with phenylpropanoid, flavone and wax metabolism. Although modulations in primary metabolism and phenylpropanoid synthesis were common to both cultivated and wild rice, the modulation of secondary metabolism related to phenylpropanoid synthesis was associated with lignin synthesis in wild rice and flavone synthesis in cultivated rice. Interestingly, while the expression of fatty acid and starch metabolism-related genes was altered in both wild and cultivated rice in response to the pathogen, changes in lipid acid synthesis and lipid acid degradation were dominant in cultivated and wild rice, respectively.

Conclusions: The response mechanisms to M. oryzae were more complex in wild rice than what was observed in cultivated rice. Therefore, this study may have practical implications for controlling M. oryzae in rice plantings and will provide useful information for incorporating and assessing disease resistance to M. oryzae in rice breeding programs.

Keywords: Cultivated rice; Magnaporthe oryzae; RNA-sequencing; Transcriptome analysis; Wild rice.

Fig. 1

Phenotype of non-inoculated and inoculated roots of cultivated and wild rice varieties. a Visible phenotype of roots of C, C + F, W and W + F groups. b, c, d and e Microscopic observation of safranin-stained roots of C (b), C + F (c), W (d) and W + F (e), respectively. The four treatments were non-inoculated cultivated rice (C), cultivated rice inoculated with Magnaporthe oryzae (C + F), non-inoculated wild rice (W), and wild rice inoculated with M. oryzae (W + F). Black lines are the scale bars, which represent 2 cm in (a) and 5 μm in (b, c, d, and e). Pink arrows indicate the representative parts of infected roots. Blue arrows indicate the plant epidermal regions, and orange arrows indicate the cortex regions

Fig. 2

Chitinase activity, and contents of soluble sugars and proline in wild and cultivated rice roots with and without Magnaporthe oryzae infection. a Chitinase activity. b Soluble sugar content. c Proline content. The four treatments were non-inoculated cultivated rice (C), cultivated rice inoculated with M. oryzae (C + F), non-inoculated wild rice (W), and wild rice inoculated with M. oryzae (W + F). The error bars represent standard deviations of the means. Different letters above the bars indicate significant differences among samples at P < 0.05

Fig. 3

Numerical values of differentially expressed gene (DEG) analysis. (a) Diagrams showing the experimental design and comparisons. (b) Diagram showing the total, up- and down-regulated gene sets in four comparisons. (c, d, and e) Venn analysis of total, up- and down-regulated DEGs derived from W + F vs W and C + F vs C comparisons. The four treatments were non-inoculated cultivated rice (C), cultivated rice inoculated with Magnaporthe oryzae (C + F), non-inoculated wild rice (W), and wild rice inoculated with M. oryzae (W + F)

Fig. 4

Validation of the RNA-sequencing data by quantitative reverse transcription PCR (qRT-PCR). Seven genes were selected from the RNA-sequencing data for qRT-PCR. a Expression patterns in C and C + F groups. b Expression patterns in W and W + F groups. The four treatments were non-inoculated cultivated rice (C), cultivated rice inoculated with Magnaporthe oryzae (C + F), non-inoculated wild rice (W), and wild rice inoculated with M. oryzae (W + F). The error bars represent standard deviations of the means. The asterisks above the bars indicate significant differences among the samples at P < 0.05

Fig. 5

Classification of up- or down-regulated genes by gene ontology (GO) terms. a GO-term classification of up-regulated genes in W + F vs W comparison. b GO-term classification of up-regulated genes in C + F vs C comparison. c GO-term classification of down-regulated genes in the W + F vs W comparison. d GO term classification of down-regulated genes in the C + F vs C comparison. The four treatments were non-inoculated cultivated rice (C), cultivated rice inoculated with Magnaporthe oryzae (C + F), non-inoculated wild rice (W), and wild rice inoculated with M. oryzae (W + F). Numbers shown next to the terms indicate the number of up- or down-regulated genes

Fig. 6

Analysis of (a) jasmonic acid (JA)-, (b) ethylene (ET)-, (c) chitinase-, and (d) WRKY- related genes that were differentially expressed in W + F vs W and C + F vs C comparisons using the fold-change values. The four treatments were non-inoculated cultivated rice (C), cultivated rice inoculated Magnaporthe oryzae (C + F), non-inoculated wild rice (W), and wild rice inoculated with M. oryzae (W + F). Color intensity indicates the fold-change values as designated by the colored bar. The white color represents unchanged gene expression

Fig. 7

Hypothetical model of wild and cultivated rice varieties in response to pathogen attack. a Defense responses of wild rice roots in response to Magnaporthe oryzae. In the roots of wild rice, fatty acids were degraded and served as precursors for diterpenoid synthesis. Fatty acids were desaturated by an ω-3 fatty acid desaturase to produce unsaturated fatty acids that were then used to promote linolenic acid synthesis. The linolenic acid subsequently promoted jasmonic acid (JA) synthesis which then induced systemic resistance and could promote chitinase activity. Starch was metabolized to produce shikimic acid for phenylpropanoid synthesis. The phenylpropanoid was then used to produce lignin that was subsequently infused into the cell walls of roots in order to increase resistance to M. oryzae. b Defense responses of cultivated rice roots in response to M. oryzae. In response to M. oryzae, roots of cultivated rice induced genes related to elongation of fatty acids. Elongated fatty acid is then promoted synthesis of wax and cutin which are infused into cell walls of roots to promote resistance to M. oryzae. Phenylpropanoid metabolism was elevated in response to the pathogen, and was directed to flavone synthesis rather than lignin synthesis. Solid arrows indicate the identified pathway. Dotted arrows indicate the supposed pathways. 1. Pathogen elicits WRKY TFs (Fig. 6d); 2. WRKY TFs elicit amylase (Table 4); 3. Starch metabolite occurs (Table 4); 4. Shikimic acid pathway assumably occurs; 5. Phenylpropanoid synthesis (Table 4); 6. Lignin synthesis (Table 4, Additional file 7: Figure S5); 7. WRKY induces β-oxidase activity (Table 4); 8. Fatty acid is degraded (Table 4); 9. NADPH and acyl-CoA promote diterpenoid synthesis; 10. WRKY TFs induce ω-3 fatty acid desaturase activity (Table 4); 11. ω-3 fatty acid desaturase promotes linolenic acid synthesis [Table 4 and assumed, Simopoulos (2016)]; 12. JA synthesis occurs (Fig. 6a); 13. JA promotes chitinase activity (Fig. 6a); 14. ET synthesis is induced under stress; 15. JA and ET promote production of WRKY TFs [Fig. 6b, Schluttenhofer and Yuan 2015]; ① Fatty acid synthesis is promoted (Table 4); ② Fatty acid is accumulated (Table 4); ③ Wax and cutin synthesis [Table 4, Lattanzio et al. 2006]; ④ Peroxisome is produced (Table 4); ⑤ Phenylpropanoid metabolism occurs (Table 4); ⑥ Flavone synthesis [Table 4, Zhao et al. 2016]. The dashed arrows and boxes represent the putative pathway in accordance to published literature, while the solid lines represent the findings of the present study. Steps 4 and 11 occur in mitochondria. The germinating spores shown represent the pathogen Magnaporthe oryzae. ET, ethylene; WRKY TFs, WRKY transcription factors