PP2A/B55α Substrate Recruitment As Defined By The Retinoblastoma-Related Protein p107

Abstract

Protein phosphorylation is a reversible post-translation modification that is essential in cell signaling. It is estimated that a third of all cellular proteins are phosphorylated (reviewed in Ficarro et al., 2002), with more than 98% of those phosphorylation events occurring on serine and threonine residues (Olsen et al., 2006). Kinases are the necessary enzymes for phosphorylation and protein phosphatases dynamically reverse this action. While the mechanisms of substrate recognition for kinases have been well-characterized to date, the same is not true for phosphatases that play an equally important role in opposing kinase function and determining global phosphorylation levels in cells. This dichotomy has also translated into the clinic, where there has been a persistently narrow research focus on the development of small-molecule kinase inhibitors for use as chemotherapeutic agents, without an equal effort being placed into the generation of the analogous phosphatase activators (reviewed in Westermarck, 2018). Members of the phosphoprotein phosphatase (PPP) family of serine/threonine phosphatases are responsible for the majority of dephosphorylation in eukaryotic cells, with protein phosphatase 1 (PP1) and protein phosphatase 2A (PP2A) accounting for more than 90% of the total phosphatase activity (Moorhead et al., 2007; Virshup and Shenolikar, 2009). Structurally, PP2A is a trimeric holoenzyme consisting of a scaffold (A) subunit, a regulatory (B) subunit, and a catalytic (C) subunit. B55α is a ubiquitous regulatory subunit that is reported to target many substrates with critical functions in processes including cell division. A long-standing question that has persisted in the field of cellular signaling is as to how the most abundant serine/threonine PP2A holoenzyme, PP2A/B55α, specifically recognizes substrates and presents them to the enzyme active site for subsequent dephosphorylation. Such critical data have only recently become well understood for the B56 family of ‘B’ regulatory subunits, where an LxxIxE short linear motif (or SLiM) has been identified in a subset of protein targets and shown via crystal structure analysis to dock into a 100% conserved binding pocket on the B56 surface (Hertz et al., 2016; Wang et al., 2016a; Wang et al., 2016b; Wu et al., 2017). Here, we show how B55α recruits p107, a pRB-related tumor suppressor and B55α substrate. Using molecular and cellular approaches, we identified a conserved region 1 (R1, residues 615-626) encompassing the strongest p107 binding site. This enabled us to identify an “HxRVxxV619-625” SLiM in p107 as necessary for B55α binding and dephosphorylation of the proximal pSer-615 in vitro and in cells. Numerous additional PP2A/B55α substrates, including TAU, contain a related SLiM C-terminal from a proximal phosphosite, allowing us to propose a consensus SLiM sequence, “p[ST]-P-x(5-10)-[RK]-V-x-x-[VI]-R”. In support of this, mutation of conserved SLiM residues in TAU dramatically inhibits dephosphorylation by PP2A/B55α, validating its generality. Moreover, a data-guided computational model details the interaction of residues from the conserved p107 SLiM, the B55α groove, and phosphosite presentation to the PP2A/C active site. Altogether, these data provide key insights into PP2A/B55α mechanisms of substrate recruitment and active site engagement, and also facilitate identification and validation of new substrates, a key step towards understanding the role of PP2A/B55α in many key cellular processes. As a parallel continuation of our efforts to identify novel B55α substrates/regulators, we generated mutant B55α constructs that occlude PP2A/A-C dimer engagement but retain substrate binding to the β-propeller structure (allowing us to interrogate direct interactors). Our preliminary AP-MS data led to the identification of several proteins that bound better to our “monomeric B55α” mutant compared to wild-type B55α in the context of the PP2A/B55α heterotrimer, including the centrosomal proteins HAUS6 and CEP170 (two substrates previously validated in a phosphoproteomic screen by our lab), suggesting that these mutants trap substrates as they cannot be dephosphorylated by PP2A/C. These analyses also identified an enrichment of T-complex protein 1 subunits in the “monomeric B55α” mutant elutions, further supporting the notion that these mutants may function as dominant negatives. Several additional proteins of interest were identified in the two independent rounds of mass spectrometry, including subunits of the DNA-directed RNA polymerases I, II, and IV, as well as the double-strand break repair protein MRE11, which can be followed up as potential novel B55α substrates. These studies can contribute to significant advances in our understanding of the network of proteins that B55α interacts with, and thus the signaling pathways that can be modulated by PP2A/B55α complexes in cells. Moreover, these advances can also provide translational benefits as has been demonstrated through the study of PP2A activators termed SMAPs, which demonstrate selective stabilization of PP2A/B56α complexes in cells that result in selective dephosphorylation of substrates including the oncogenic target c-MYC.