| Rice blast disease, caused by the fungus Magnaporthe grisea, is one of the most serious diseases of cultivated rice (Oryza sativa L.) worldwide. It is estimated that 10-30% of the annual rice harvest is lostdue to rice blast disease, which is enough to feed 60 million people. Following the classical gene-for-gene system, protein-protein interactions between different species, rice and fungus, might be the key factor for the causing of blast disease. In this paper, firstly, a protein-protein interaction data between inter-species was collected from the database and references. Secondly, Interlog and domain-domain interaction methods were used to predict the protein-protein interactions between rice and fungus. Finally, the reliability and the function of the interaction network were also analyzed. The following results were reached.1) Prediction of protein-protein interaction between rice and fungus. Interlog was applied to predict the protein-protein interaction between rice and fungus, and the results were confirmed by domain-domain interacting information. After that, the predicting interactions were further verified by the fungus protein function following the hypothesis that only the secreted and membrane proteins of fungus could interacted with the proteins from rice. Totally,532 protein-protein interactions between rice and fungus were yielded, including 27 fungus proteins and 236 rice proteins. The average degree of fungus proteins in the interaction network was 19.37, higher than that of rice proteins. This indicated that these 27 fungus proteins played key role in the network.2) Verification of the protein-protein interactions between rice and fungus. Two methods were used to confirm the above protein-protein interaction. Compared to the 1,000 random interactions, the pathogenesis genes of blast fungus were enriched in the above protein-protein interaction network (P value< 0.0001). Support Vector Machine (SVM) was used to classify the proteins in the interacting network,10-fold cross validation showed that the classifying accuracy was 93.85%, while independent test implied that the accuracy was 84.67%. This results indicated that the predicting protein-protein interaction networks between rice and fungus were reliable.3) GO analysis of the protein-protein interactions between rice and fungus. The regulatory network was drawn based on the protein-protein interaction and partitioned into 228 sub-networks. The largest sub-network had 3684 nodes. The result of Go enrichment analysis for these sub-networks suggested these sub-network functioned in the key biological process of rice plants. Furthermore, we identified 10 targeted rice genes which connected to multiple sub-networks in the regulatory network.4) The function analysis of the regulatory networks in rice. The result from the microarray data of rice plants in response to three pathogen infection showed that the rice plants might have propensity in immune response when infected by different pathogen. There were 46 master regulators which were the interactant of the fungus proteins. Among these master regulators,34 of them were up-regulated master regulators and 12 are down-regulated master regulators. PARTHER analysis showed that ubiquitin proteasome pathway were regulated by all the 3 pathogen, while apoptosis signaling pathway was induced by fungus and bacteria infection. This result explored the common response of rice plants under the different pathogen infection.In conclusion, the result in this paper provide insight into the process of rice blast disease caused by blast fungus, and would help researchers to understand the immune-related pathway which was activated by fungus. Meanwhile, the common regulatory pathway induced by different pathogen would be useful in the breeding of new rice cultivars with pathogen-resistant ability. |
PDF Full Text Request