Dynamic changes in transcriptome in rice compatible and incompatible interactions with M. oryzae are different
To gain molecular insight into the mechanisms involved in the resistance to blast disease in Hui1586, in-depth comprehensive transcriptome profiles compared with susceptible Nipponbare were achieved through RNA-seq experiments. The two-week-old rice plants were spray inoculated with spores of M. oryzae Guy11 or H2O (mock treatment), and then leaf samples with 3 biological replicates were collected at 12, 24, 36 and 48 hpi for RNA-seq. The mock for each time point was processed to ensure appropriate comparisons. In total, 48 samples were subjected to RNA-seq, resulting in 6.8-gigabyte 100-bp paired-end reads per sample on average.
To visualize the variation as well as the similarity for all samples, we performed a principal component analysis (PCA) on the normalized FPKM (fragments per kilobase of transcript per million mapped reads) values of all the detected genes. The PCA plot showed that the data for three biological replicates were clustered closely and were separated by the time point, treatments and genotypes (Fig. 2A). Among the data for the mock treatment, the 12, 24, 36, and 48 h samples were clustered away from each other, indicating that the circadian clock is a major factor affecting gene expression. At 12 h, the M. oryzae treatments were clustered away from the mock treatment in both susceptible Nipponbare and resistant Hui1586. However, at the late time points of 24, 36, and 48 h, the M. oryzae and mock treatments were clustered closely on Hui1586, while they were clustered away on Nipponbare, indicatingM. oryzae induced large transcriptome changes in the susceptible Nipponbare at all time points but high-amplitude transcriptome changes only at the early infection stage in resistant Hui1586.
A hierarchical clustering analysis was conducted on the FPKM values of all detected genes. The samples were clustered into 4 larger groups according to the time point (Fig. 2B), again showing that the circadian clock is a major regulator in the rice transcriptome. In each large group, the H2O treatments were clustered into the same subgroup, except the 36 hpi samples of H2O-treated Hui1586, which were clustered with M. oryzae- treated Hui1586, suggesting the occurrence of small transcriptional changes in Hui1586 at 36 hpi with M. oryzae treatment. The 24, 36, and 48-hpi samples of M. oryzae -treated Nipponbare were clustered away from the mock samples (Fig. 2B), indicating that M. oryzae persistently induced strong transcriptional changes in the susceptible Nipponbare.
Then, we identified differentially expressed genes (DEGs; fold change >2; P < 0.001) between the fungal and mock treatment in Nipponbare or Hui1586 at each time point. A large number of DEGs in Nipponbare were identified, especially at 12 h (4680 upregulated and 2045 downregulated) and 36 h (3347 upregulated and 3653 downregulated). While Guy11 induced more dramatic transcriptional changes (6808 upregulated and 2895 downregulated genes) in the resistant Hui1586 at 12 h, the numbers of DEGs markedly decreased at 24, 36 and 48 h (Fig. 2C). In total, 12333 and 12147 DEGs, accounting for approximately 25% of the rice genes, were identified respectively in Nipponbare and Hui1586 in the 48 h time-series of the fungal infection. Among these DEGs, 8211 were commonly regulated in both cultivars (Fig. 2D and Supplemental dataset 1). Together, the data indicated thatM. oryzae infection induced dramatic and dynamic transcriptional reprogramming in rice. Remarkably, the resistant Hui1586 mounted faster and stronger transcriptional reprogramming at the early infection stage than the susceptible Nipponbare, which likely restricted the fungal infection more efficiently and resulted in smaller transcriptional changes at the later time points. In contrast, Guy11 could successfully infect and constitutively multiply on the susceptible Nipponbare, inducing continuous dramatic transcriptional changes within all infection stages.