Function and regulation of the reversible methylation of Protein Phosphatase 2A

Reversible phosphorylation of serine/threonine residues is one of the most frequent post-translational modifications of proteins and is an important regulatory mechanism in many cellular processes. The steady-state phosphorylation status of a protein is controlled by the opposing actions of protein kinases and phosphatases. Protein phosphatase 2A (PP2A) is a major family of multimeric serine/threonine phosphatases and belongs together with protein phosphatase 4 (PP4) and 6 (PP6) to the family of the PP2A-like phosphatases. PP2A is implicated in the regulation of a plethora of cellular processes such as apoptosis, transcription, translation, DNA replication, signal transduction, cell division and tumourigenesis. The participation of PP2A in these pleiotropic functions implies that its activity must be tightly controlled in vivo. The PP2A core enzyme consists of a catalytic subunit (C subunit/PP2AC) and a scaffolding A subunit (PR65/A subunit) through which one of several regulatory B subunits (PR55/B, PR61/B’ or PR72/B”) can bind. Proper functioning and regulation of PP2A is achieved by the association of the regulatory B subunits with the PP2A core enzyme, resulting in the formation of heterotrimeric PP2A holoenzymes with specific catalytic properties, subcellular locations and substrate specificities. Therefore, the assembly of the complex with the appropriate B-type subunit is the key to specificity and regulation of PP2A. In this process, methylation and phosphorylation of the conserved C-terminal PP2AC tail seem to play a crucial role. Methylation occurs on the carboxyl group of the carboxyterminal leucine 309 residue of PP2AC and is catalyzed by an S-adenosylmethionine-dependent leucine carboxyl methyltransferase (LCMT1), whereas demethylation is catalyzed by a specific phosphatase methylesterase (PME-1). However, neither the exact consequences nor the regulation of PP2AC methylation and phosphorylation are very clear. In this work, I showed that the diverse regulatory B subunits require distinct modifications or structural features of the PP2AC C-terminal tail for heterotrimer formation and that PP2AC methylation is essential for cell viability. More specifically, carboxylmethylation of the C-terminal leucine 309 residue of PP2AC is absolutely necessary for formation of PP2A trimers with the PR55/B subunits (PP2AT55), whereas it is not absolutely required for PP2AT61,72 formation. PP2AT61a,b,e assembly is specifically inhibited by phosphorylation-mimicking mutations of tyrosine 307, whereas introduction of a negative charge at position threonine 304 inhibits only PP2AT55 formation. These results have important implications since they show that the selectivity for the B-class subunits can be based on the methylation and the phosphorylation state of PP2AC. This opens the possibility for dynamic subunit interchanges partially regulated by covalent modifications of the PP2AC C-terminal tail, thereby explaining differential trimeric assembly of PP2A. Further, I demonstrated that PME-1, in contrast to LCMT1, is mainly nuclear correlating with the predominant occurrence of demethylated PP2AC in the nucleus. The nuclear targeting of PME-1 is mediated by a functional nuclear localisation signal (NLS) and overexpression of a cytoplasmic NLS mutant of PME-1 resulted in demethylation of PP2AC at the ‘wrong’ place, i.e. in the cytoplasm. Therefore, targeting of PME-1 to the nucleus can provide a mechanism for spatial regulation of PP2A methylation as it restricts demethylation to nuclear PP2AC. I also established an additional function for PME-1 in the regulation of PP2A, since PME-1 demethylates not only PP2AC, but also stabilizes an inactivated PP2A pool. This inactive form of PP2A is not due to demethylation of the phosphatase, since LCMT1 could not reactivate this inactive PP2A pool. In contrast, reactivation of the Ser/Thr phosphatase activity of the inactive PP2A-PME-1 complex is achieved by addition of the phosphatase two A phosphatase activator (PTPA), leading to a dissociation of PME-1 from PP2A and a stable activation of PP2A. Since PME-1 could not inactivate active PP2A and an inactive PME-1 S156A mutant also bound inactive PP2A, it seems that the catalytic activity of PME-1 itself does not cause PP2A inactivation. Instead, PME-1 stabilizes an inactivated PP2A pool which is - given the nuclear localisation of PME-1 - presumably present in the nucleus. However, the function and precise nature of this inactive nuclear PP2A pool still remains to be determined. Finally, I have investigated the substrate specificity of LCMT1 and PME-1, so far known to be specific for PP2A. I demonstrated by an in vitro assay that both PP4C and PP6C can be reversibly methylated on their C-terminal leucine residue by the same LCMT1 and PME-1. These findings show that the PP2A-like phosphatase family can share, at least in vitro, the same (de)methylating enzymes and establish reversible methylation as a common regulatory mechanism for all PP2A-like phosphatases.status: publishe