Identification of Drosophila Mitotic Genes by Combining Co-Expression Analysis and RNA Interference

• DOI: 10.1371/journal.pgen.1000126
• Publication Date: 2008-07-17T22:35:49Z
• Authors: Maria Patrizia Somma; Francesca Ceprani; Elisabetta Buccia...; Valeria Naim; Valeria De Arcang...; Roberto Piergentili; Antonella Palena; Laura Ciapponi; Maria Grazia Gian...; Claudia Pellacani; Romano Petrucci; Giovanni Cenci; Fiammetta Vernì; Barbara Fasulo; Michael L. Goldberg; Ferdinando Di Cunto; Maurizio Gatti
• Keywords: 0301 basic medicine; Animals, Chromosome Segregation, Cytokinesis, Drosophila; genetics/metabolism, Gene Expression, Genes; Insect, Mitosis; genetics, Mitotic Spindle Apparatus; genetics/metabolism, RNA Interference, RNA; Double-Stranded; metabolism; Gene Expression; Mitosis; Genes, Insect; Spindle Apparatus; QH426-470; [SDV] Life Sciences [q-bio]; 03 medical and health sciences; Chromosome Segregation; Genetics; Animals; Drosophila; RNA Interference; Research Article; Cytokinesis; RNA, Double-Stranded

RNAi screens have, to date, identified many genes required for mitotic divisions of Drosophila tissue culture cells. However, the inventory such remains incomplete. We have combined powers bioinformatics and technology detect novel genes. found that involved in mitosis tend be transcriptionally co-expressed. thus constructed a co-expression-based list 1,000 are highly enriched functions, we performed each these By limiting number examined, were able perform very detailed phenotypic analysis examined dsRNA-treated cells possible abnormalities both chromosome structure spindle organization. This allowed identification 142 genes, which subdivided into 18 phenoclusters. Seventy not previously been associated with defects; 30 them assembly and/or segregation, 40 prevent spontaneous breakage. note latter type has never detected previous any system. Finally, against encoding kinetochore components or conserved splicing factors results identical defects highlighting an unanticipated role centromere function. These findings indicate our method detection functions works remarkably well. can foresee elaboration co-expression lists using same phenocluster will provide candidate small-scale aimed at completing proteins.