Towards understanding substrate recruitment by PP1 holophosphatases

Protein phosphorylation is a vital regulatory mechanism within the cell which is commonly dysregulated in diseases such as cancer and neurodegeneration. The addition of a phosphoryl group to a protein is catalysed by a group of enzymes known as kinases and the removal is catalysed by phosphatases. Over 98% of phosphorylation sites on proteins occur on serine and threonine residues and a major serine/threonine phosphatase within the human cell is protein phosphatase 1 (PP1). This enzyme functions as a holoenzyme, with a core catalytic subunit (PP1c) bound to one or two non-catalytic subunits. These non-catalytic subunits have various proposed functions including inhibiting PP1c, aiding PP1 biogenesis, providing substrate specificity, and localising PP1c. Over 200 non-catalytic subunits for PP1 have been predicted; however the vast majority of these are not characterised. Of the few characterised non-catalytic subunits providing substrate specificity to PP1c, the knowledge of substrate recruitment is still limited. This study aimed to improve the current understanding of substrate recruitment to PP1 holophosphatases. Initial work focused on two of the better understood PP1 holoenzymes known as the eukaryotic initiation factor 2 alpha (eIF2α) phosphatases. These holoenzymes consist of PP1c bound to either the constitutively expressed PPP1R15B (R15B) or stress inducible PPP1R15A (R15A). These holoenzymes catalyse the dephosphorylation of eIF2α at serine 51, an important signalling event in the integrated stress response pathway, enabling cells to deal with various stresses. The substrate recruitment mechanism of these R15-PP1c holophosphatases was unclear. Previous work with recombinant proteins proposed conflicting mechanisms. One mechanism proposed the carboxy region of R15 recruits eIF2α with G-actin providing specificity while the other proposed that the central region of the R15 proteins alone was sufficient to recruit eIF2α. Here, a cellular overexpression system was utilised to show that both R15A and R15B recruit eIF2α via the central region, independently of G-actin. Furthermore, a small subset of residues in the central region of R15B were identified as important for substrate recruitment, therefore, advancing our knowledge of how R15B recruits the substrate eIF2α. During work with the R15 proteins, phospho-substrate traps domains were identified. I then aimed at expanding on the notion that fragments of non-catalytic subunits of phosphatases could be used to trap substrates. This phospho-substrate trapping concept was applied to a novel PP1 holoenzymes to try tackle the difficult challenge of identifying new substrates of PP1 holoenzymes. In summary, this work expands our understanding of eIF2α phosphatase substrate recruitment, providing concepts which can be applied to novel PP1 holophosphatases.