Résumé
Moniliophthora perniciosa, the fungus responsible for the witches' broom disease, brought serious problems to cacao (Theobroma cacao L.) cultivation in the infected areas such as in Southern Bahia, Brazil. For this reason, several molecular studies have been recently developed from T. cacao and/or M. perniciosa, and subsequent computational analyses were developed. Among them, the systems biology provides a framework for assembling models of biological systems from systematic measurements obtained by experimental analysis. In the case of M. perniciosa, which is still little studied, the first step for protein-protein interaction (PPI) network analysis by systems biology consists in the identification of an adequate organism for ortholog search. Neurospora crassa is a well-known filamentous fungus, considered as a model organism that has been used for more than 90 years to study genetics, biochemistry and fungal biology. Moreover, lots of genomics and molecular data are available for N. crassa, including those related to architecture and hyphae development, as well as cell wall degradation apparatus. Here, the objective was to evaluate the possible use of N. crassa as model for determination of M. perniciosa PPI networks, and as an example to test our hypothesis, two polygacturonases from M. perniciosa (MpPG1 and MpPG2) were analyzed. First, a reciprocal BLASTp of MpPG1 and MpPG2 was performed on N. crassa database using stringent conditions (10-10). The respective N. crassa orthologs (NCU06961 and NCU02369) were used to build the PPI network in STRING 9.05. This network was analyzed in the Cytoscape 2.8.2 software with the Molecular Complex Detection, Biological Network Gene Ontology and CentiScaPe plugins. The PPI network contained 892 nodes (proteins), 43.035 connectors, and was organized in 11 modules corresponding to biological processes, such as ribosome biogenesis, regulation of gene expression, processes related to carbohydrate metabolism, among others. Twenty-two proteins interacted directly or indirectly with MpPG2- NCU02369 and were associated to post-translational protein modification or pectin degradation. From the 892 and 22 specific proteins, 544 (61%) and 14 (63.6%) were found by reciprocal BLASTp when we looked back on the M. perniciosa database. Considering that the M. perniciosa published sequence only represent 69% of the total genome, the results we obtained confirmed that N. crassa could be used as a good model for M. perniciosa in systems biology analysis of hyphae proteins. Work supported by BNB, CNPq, FAPESB, Cirad. (Texte intégral)