Modifications of transfer RNA (tRNA) are widespread and numerous but their function is not well understood. Particularly, three tRNAs (tRNALysAAA, tRNAGluGAA and tRNAGlnCAA) show universal modifications of the second and fifth carbons of the wobble base. The «Elongator» complex (Elp1-6) is proposed to modify carbon five while the Ctu1-Ctu2 complex is responsible for the thiolation of carbon two. In both cases, the modification is thought to be required to optimize the codon-anticodon interaction. In the fission yeast, the deletion of either complexes results in thermosensitivity associated with growth defects on specific compounds as well as cell cycle and cytokinesis defects. A key step in understanding how these phenotypes result from a tRNA modification defect is to compare the proteome from wild type and mutants. We determined the proteome-wide relative expression level of each S. pombe protein by combining an integrated ORFeome to reverse protein arrays. We identified subsets of proteins whose expression level was altered in the strains lacking either modification. The groups identified can be linked to the phenotypes observed. As a detailed example, the subset of proteins involved in the « mitotic cell cycle transition from G2 to M » has been thoroughly studied. Several genes coding for proteins belonging to this subset show genetic interaction with ctu1 or elp3. One of the more interesting of these genes is cdr2, coding for a protein kinase negatively regulating Wee1 to allow the entry into mitosis. The cell cycle phenotypes associated with the absence of tRNA modifications can be explained by the strong decrease in Cdr2 level observed in the corresponding mutants. Moreover, the ctu1 and elp3 mutant phenotypes mimic most of the reported cdr2 mutant phenotypes. Yet, these phenotypes are aggravated in the triple ctu1 elp3 cdr2 mutant, suggesting that Cdr2 is not the only protein responsible for the phenotypes observed in absence of modification. Why is the translation of the proteins in these subsets – and particularly Cdr2 – hyper sensitive to the absence of mcm5s2U tRNA modification? The answer is found in the codon content of the corresponding genes. Indeed, all the sensitive groups are enriched in AAA codons (by opposition to AAG), which is read by tRNALysUUU, one of the three mcm5s2U modified tRNAs. This link between codon content and sensitivity to the absence of tRNA modification has been demonstrated experimentally via the study of Cdr2. We suppressed the translation defect of Cdr2 by overproducing tRNALysUUU from a plasmid, or by replacing every AAA lysine codon by a AAG-lysine (read by the naturally unmodified tRNALysCUU). Taken all together, these experiments show that the translation of groups of genes enriched in AAA codon is less efficient in the absence of mcm5s2U modification. How is the selection pressure leading to an enrichment of these genes in AAA codons exerted? Why enriching a gene with the lysine AAA codon that is read with low efficiency? These issues are part of the remaining questions. We suggest the possibility of a regulation specific to these groups of genes thanks to their codon bias and the mcm5s2 double modification of tRNAs.
|Date of Award||22 Dec 2011|
|Supervisor||Damien Hermand (Supervisor), Alain Chariot (Jury), Anders Bystrom (Jury), Anabelle Decottignies (Jury), Patricia Renard (Jury) & JEAN-JACQUES LETESSON (President)|