ADVANCED CHARACTERIZATION OF THE YEAST KEOPS/EKC COMPLEX

During my PhD I have studied the properties of two yeast proteins, the protein kinase Bud32 and the putative protease Kae1, that take part in a nuclear complex named KEOPS/EKC. Actually, while Kae1 is associated uniquely to the proteins of the complex, Bud32 has many other partners in the cell; in fact, I have also studied its strong relationship with the Grx4 glutaredoxin. The yeast KEOPS/EKC complex has been isolated in 2006 by two different groups and has been shown to be involved in both telomere homeostasis and transcription regulation. The complex is evolutionarily conserved and is composed of five proteins: the protein kinase Bud32, the hypothetical metallo-protease Kae1, the still uncharacterized Cgi121, and two other proteins of small size, Pcc1 and Pcc2/Gon7. For many years, our attention has been focused on the atypical protein kinase Bud32, which interacts with many other yeast proteins, suggesting that it may play several roles in the cell. Among these Bud32 partners, we demonstrated that the glutaredoxin Grx4 is a substrate of the protein kinase, being readily phosphorylated by Bud32 mainly at Ser 134. Also, this modification is upregulated by the previous phosphorylation of Bud32 at its Ser258 residue by the Sch9 protein kinase (the yeast homologue of mammalian Akt/PKB). During the first part of my PhD I deepened the study of the physiological significance of this new phosphorylation cascade. By the phenotypic analysis of yeast strains expressing mutagenized forms of Grx4, I demonstrated that the phosphorylation of Grx4 by Bud32 is important for Grx4 functionality in vivo. However I could not identify a specific effect of the Bud32-mediated phosphorylation of Grx4 on the known activities of the glutaredoxin, wich is involved in iron cellular homeostasis and in the survival under oxidative stress conditions. This result suggests that the Bud32-mediated phosphorylation of Grx4 play a role in different, uncharacterized functions of the glutaredoxin. I also checked if the phosphorylation of Bud32 by Sch9 could modulate the activity of the whole KEOPS complex, but the analysis of telomeres length and of the activation rate of the galactose-inducible GAL1 gene (one of the main transcriptional targets of KEOPS) showed that these functions are unaffected in a Bud32 unphosphorylatable mutant (S258A). These results suggest that the phosphorylation of Bud32 at Ser258 is unrelated to its function within the KEOPS complex. I then addressed my attention to the Kae1 subunit of the complex. By using two strains expressing mutagenized forms of Kae1, I could demonstrate that the activity of this protein is essential for the complex, both at the telomere and at the transcriptional level. The biochemical function of Kae1 is however still unknown. It was initially classified as a protease, and, in effect, in 2006 an endopeptidase activity was indirectly demonstrated for the human homologue of Kae1, OSGEP. On the contrary, in 2007 Hecker et al. demonstrated that an archaeal orthologue of Kae1 is an AP-endonuclease . During my PhD I tried to define the activity of yeast Kae1, but the results obtained are not sufficient to clarify this point. Finally, I decided to verify the hypothesis, coming from a recent work that describes the atomic structure of an archaeal-derived KEOPS complex, that Kae1 could be a substrate of Bud32. Using the yeast Bud32 and Kae1 proteins, co-expressed and purified from E.coli, I observed that the also the recombinant proteins are tightly associated, forming a kind of catalytic KEOPS subcomplex. Using several mutagenized forms of these proteins I demonstrated, by in vitro phosphotranspherase assays, that Bud32 is able to phosporylate Kae1 and that the binding of Kae1 has an inhibitory effect on the catalytic activity of the kinase. An important confirmation comes from the MS analysis of phosphorylated Kae1, that identified Ser 367 as a target of Bud32. However this might not be the only phosphorylated residue. Altogether these results indicate that, at least in vitro, a regulatory relationship exists between Bud32 and Kae1. This is interesting as the two proteins are liable to carefully modulate the functions of the entire KEOPS complex.