Select Stabilization of a Tumor-Suppressive PP2A Heterotrimer

Abstract

In cancer, suppression of protein phosphatases, such as protein phosphatase 2A (PP2A), that normally counteract kinases, contributes to aberrant signaling. Leonard et al. recently demonstrated that a novel small-molecule activator of PP2A, DT-061, selectively stabilizes a specific PP2A holoenzyme responsible for dephosphorylating critical oncogenic targets, including MYC. The 3.6-Å cryo-electron microscopy map of the heterotrimer assembly provides insight into the druggable structure of PP2A, guiding future phosphatase therapeutics.

Following the landmark discovery of imatinib (a tyrosine kinase inhibitor) for chronic myeloid leukemia, recent decades of cancer research have focused on kinase inhibition for cancer treatment [1]. Although the most common pathological mechanisms in cancer involve activated kinase signaling, they also frequently involve the loss of specific protein phosphatases, such as PP2A, that normally counteract kinase signaling [2]. PP2A represents a major family of serene/threonine protein phosphatases responsible for the regulation of ~60% of all serene/threonine phosphorylation events, thus making it a key player in phosphorylation signaling cascades and the regulation of numerous processes, including cell growth, cell cycle, mitosis, differentiation, and apoptosis [3]. Overall, PP2A has been established to function as a tumor suppressor, diminishing the activation of master oncogenic regulators, including MYC, extracellular signal-regulated kinases (ERK), protein kinase B (AKT), and B cell lymphoma 2 (BCL2) [4]. Given the ubiquitous role of PP2A, it is not surprising that its dysregulation contributes to disease, making it an attractive target in many clinical settings [5]. Despite the significant role of PP2A in multiple disorders, mechanistic insight into targeting it with small molecules remains obscure.

The varied role of PP2A in cellular regulation comes from its unique ability to form >90 distinct heterotrimeric holoenzymes, each containing a common heterodimeric core comprising the scaffold A subunit and catalytic C subunit bound to one of many regulatory B subunits (Figure 1). The diversity of holoenzyme formation is made possible by the inherent flexibility of the A subunit and the large range of B subunits from four families (B, B′, B′′, and B′′′), which confer broad substrate specificity, cellular location, and tissue distribution [3]. This gives rise to a multitude of activities, which could be influenced by the stoichiometry of the different B subunit family members.

Figure 1. Schematic Depicting Effects of DT-061 on Protein Phosphatase 2A (PP2A) and Its Regulation of c-MYC.

PP2A is a diverse family of heterotrimeric serine/threonine phosphatases containing distinct B subunits from one of four families, where a specific holoenzyme, PP2A-B56α, has an important role in regulating the MYC oncoprotein family. Depicted here, c-MYC is phosphorylated by enzymes, such as extracellular signal-regulated kinase (ERK), at the serine 62 position (pS62; black sphere), which in turn increases MYC protein stability and target gene activation. In cancer, downregulation of PP2A supports increased MYC activity (A). Leonard et al. [11] now show that DT-061, a small-molecule activator of PP2A (SMAP), selectively binds and stabilizes the PP2A-B56α holoenzyme (B). With DT-061, PP2A-B56α dephosphorylates S62-phosphorylated c-MYC to facilitate its inactivation via both immediate S62 dephosphorylation and S62 dephosphorylation subsequent to glycogen synthase kinase 3β (GSK3β)-mediated Threonine 58 phosphorylation (pT58; orange sphere). These regulated phosphorylation events trigger ubiquitin (pink spheres)-mediated degradation of MYC, promoting cellular homeostasis. Dotted lines denote that pathways are suppressed or inactivated, whereas thicker lines indicate increased activity. Figure created using BioRender (https://biorender.com/). Abbreviation: c-MYC, cellular MYC.

Unlike other tumor suppressors [e.g., tumor protein TP53 (p53), retinoblastoma protein (Rb), or phosphatase and tensin homolog (PTEN)], PP2A is most frequently inactivated by nongenetic mechanisms, such as phosphorylation and/or demethylation of the C-terminal tail of the catalytic C subunit or increased expression of endogenous PP2A inhibitors [3,4]. These endogenous regulators of PP2A demonstrate oncogenic properties, likely through their targeting of the tumor-suppressive PP2A holoenzymes, and have been the main focus for PP2A-reactivating targeting strategies for the past decade [4]. One such example is inhibitor 2 of PP2A (SET), which binds to, and inactivates, the catalytic C subunit of PP2A, and is frequently overexpressed in cancers [3,4]. Efforts to target this negative regulator have resulted in effective antagonists of SET, such as the peptide OP449, which causes the intracellular release of SET from PP2A and subsequently increases PP2A activity [6]. Groups studying other endogenous inhibitors, such as cancerous inhibitor of protein phosphatase 2A (CIP2A), and bioactive sphingolipids, such as ceramides or their derivatives, have also shown success in activating PP2A tumor suppressor function [4,7].

Although these approaches prove promising, they lack the ability to address complexity in the field, where different PP2A holoenzymes can have varied activities in regulating cellular processes due to the diversity of the PP2A family. Selectively targeting specific holoenzymes to direct PP2A activity against specific pathogenic drivers has not been possible previously. Recent studies have shown efficacy of re-engineering FDA-approved tricyclic neuroleptics, including small-molecule activators of PP2A (SMAPs) [8]. These compounds have been shown to induce apoptosis, promote dephosphorylation of PP2A targets, such as MYC and AKT, and suppress tumor growth in numerous mouse models [9,10]. However, until now, which isoform and subunit are stabilized by SMAPs has been unclear.

In a recent study, Leonard et al. [11] present the structural and biological basis of direct activation of PP2A by DT-061, one of the lead engineered compounds from tricyclic neuroleptics. In describing the 3D structure of the DT-061-bound PP2A trimeric complex, Leonard et al. revealed a unique intersubunit pocket that DT-061 occupies, providing insight into its mechanism of PP2A activation, namely direct, specific binding and stabilization of the PP2A-B56α holoenzyme, which drives an important antitumor function of PP2A (Figure 1). This critical information not only opens the field of cancer therapeutics to a new class of drug targets (tumor-suppressive phosphatases), but also provides a path for the structural design and development of novel and improved drugs for the biased targeting of selected PP2A holoenzymes.

Leonard et al. designed a cell-based split luciferase assay to study how DT-061 affects PP2A. They observed a structural constriction in the AC complex, indicative of an increase in B-AC complex binding between 1 and 2 h post DT-061 exposure. Complex binding then reverted back to baseline within 6 h after drug washout. This was also shown in vivo using lung adenocarcinoma xenograft experiments, where tumors treated with DT-061 showed a significant increase in PP2A B56α-containing trimeric holoenzymes, while other B subunit-containing holoenzymes remained unchanged compared with vehicle.

To further investigate the specific binding of DT-061 to the PP2A-B56α holoenzyme, a second split luciferase assay separately tagging the N terminus of the scaffold A subunit and the C terminus of B56α was utilized. This, combined with co-immunoprecipitation (co-IP) experimental results, demonstrated that DT-061 selectively enhances the B56α-containing holoenzyme, increasing specific AB56αC complex assembly. The authors also showed a concentration-dependent association of the AB56αC complex by size exclusion chromatography coupled with UV absorbance. Importantly, fluorescence polarization and surface plasmon resonance confirmed that DT-061 increased the binding affinity of the B56α-subunit to the AC dimer and stabilized the complex.

The authors identified the binding site of DT-061 at the trimeric interface of the AB56αC complex, elucidating the molecular basis of the biased complex stabilization. The N terminus of the A subunit anchors the B subunit, whereas its C terminus anchors the C subunit. Here, sandwiched between the A and B subunits, a terminally methylated L-309 residue at the C terminus of the C subunit is adjacent to the binding site of DT-061. Of note, L-309 methylation is enriched in the DT-061-stabilized AB56αC holoenzyme, demonstrating nucleation of the active, methylated enzyme by DT-061. Leonard et al. showed that the molecular arrangement at the trimeric junction is unique to the B56α heterotrimer, which explains both the specificity of DT-061 and the increased stability of the assembled complex. Electron microscopy grids prepared with recombinant AB56αC trimer were unable to maintain the heterotrimeric composition in the absence of DT-061, resulting in monomeric and dimeric 2D class averages, while inclusion of DT-061 yielded a homogeneous distribution of AB56αC trimers. Thus, the unique stabilizing property of DT-061 allowed the elusive structure of a PP2A heterotrimer to be solved [11].

A key substrate of PP2A-B56α, MYC, is involved in transcriptional regulation of a multitude of genes that are essential for cellular programs required for both normal and neoplastic cells (Figure 1) [12]. MYC has been intensively studied for its role in driving nearly all types of cancer development, metastasis, and therapeutic resistance [6,10,12,13]. Yet, after decades of research, a way to therapeutically target MYC remains intangible. Importantly, MYC oncogenic activity is post-translationally controlled, providing a potential new route for therapeutic intervention [12]. The best-known MYC-activating phosphorylation event on Serine 62 is dephosphorylated by PP2A-B56α, making DT-061 an exciting approach for targeting active MYC through stabilization of PP2A-B56α (Figure 1). The authors characterized the kinetic loss and recovery of MYC following DT-061 treatment of xenograft tumors, and showed that it had an inverse relationship with L-309 methylation and a positive correlation with the induction of apoptosis [11].

While animal studies have demonstrated the potent elimination of tumors upon genetic inactivation of MYC in tumor cells, tumor clearance is dependent upon an intact immune system [13,14]. PP2A has been shown to control functions of many immune cells, including T cells, natural killer (NK) cells, and myeloid-derived suppressor cells (MDSCs) through modulation of upstream signaling and chemokine production [5,15]. Furthermore, inhibition of specific PP2A holoenzymes, such as PP2A-B55δ, can increase antitumor immunity [15]. This complexity of PP2A function emphasizes the importance of the study by Leonard et al., where DT-061, a systemically administered drug, can selectively increase the activity of a specific tumor-suppressive PP2A holoenzyme (PP2A-B56α). Given that DT-061 shows clear in vivo efficacy in murine tumor models [9,10], future studies will be needed to determine its effects on the complexity of immune function in disease.

Ultimately, Leonard et al. [11] utilized multiple technologies to structurally identify and validate the biased stabilizing effects of DT-061, a small-molecule activator of PP2A, on PP2A-B56α, overcoming any concern related to indiscriminate enzyme activation and providing a platform for new SMAP development for other clinical presentations with dysregulated levels of PP2A.