Discovery of selective activators of PRC2 mutant EED-I363M.

Many common disease-causing mutations result in loss-of-function (LOF) of the proteins in which they occur. LOF mutations have proven recalcitrant to pharmacologic intervention, presenting a challenge for the development of targeted therapeutics. Polycomb repressive complex 2 (PRC2), which contains core subunits (EZH2, EED, and SUZ12), regulates gene activity by trimethylation of histone 3 lysine 27. The dysregulation of PRC2 catalytic activity by mutations has been implicated in cancer and other diseases. Among the mutations that cause PRC2 malfunction, an I363M LOF mutation of EED has been identified in myeloid disorders, where it prevents allosteric activation of EZH2 catalysis. We describe structure-based design and computational simulations of ligands created to ameliorate this LOF. Notably, these compounds selectively stimulate the catalytic activity of PRC2-EED-I363M over wildtype-PRC2. Overall, this work demonstrates the feasibility of developing targeted therapeutics for PRC2-EED-I363M that act as allosteric agonists, potentially correcting this LOF mutant phenotype.

Discovery of Peptidomimetic Ligands of EED as Allosteric Inhibitors of PRC2.

The function of EED within polycomb repressive complex 2 (PRC2) is mediated by a complex network of protein-protein interactions. Allosteric activation of PRC2 by binding of methylated proteins to the embryonic ectoderm development (EED) aromatic cage is essential for full catalytic activity, but details of this regulation are not fully understood. EED's recognition of the product of PRC2 activity, histone H3 lysine 27 trimethylation (H3K27me3), stimulates PRC2 methyltransferase activity at adjacent nucleosomes leading to H3K27me3 propagation and, ultimately, gene repression. By coupling combinatorial chemistry and structure-based design, we optimized a low-affinity methylated jumonji, AT-rich interactive domain 2 (Jarid2) peptide to a smaller, more potent peptidomimetic ligand (Kd = 1.14 ± 0.14 µM) of the aromatic cage of EED. Our strategy illustrates the effectiveness of applying combinatorial chemistry to achieve both ligand potency and property optimization. Furthermore, the resulting ligands, UNC5114 and UNC5115, demonstrate that targeted disruption of EED's reader function can lead to allosteric inhibition of PRC2 catalytic activity.

The EED protein-protein interaction inhibitor A-395 inactivates the PRC2 complex.

Polycomb repressive complex 2 (PRC2) is a regulator of epigenetic states required for development and homeostasis. PRC2 trimethylates histone H3 at lysine 27 (H3K27me3), which leads to gene silencing, and is dysregulated in many cancers. The embryonic ectoderm development (EED) protein is an essential subunit of PRC2 that has both a scaffolding function and an H3K27me3-binding function. Here we report the identification of A-395, a potent antagonist of the H3K27me3 binding functions of EED. Structural studies demonstrate that A-395 binds to EED in the H3K27me3-binding pocket, thereby preventing allosteric activation of the catalytic activity of PRC2. Phenotypic effects observed in vitro and in vivo are similar to those of known PRC2 enzymatic inhibitors; however, A-395 retains potent activity against cell lines resistant to the catalytic inhibitors. A-395 represents a first-in-class antagonist of PRC2 protein-protein interactions (PPI) for use as a chemical probe to investigate the roles of EED-containing protein complexes.

YC-1 binding to the ß subunit of soluble guanylyl cyclase overcomes allosteric inhibition by the α subunit.

Soluble guanylate cyclase (sGC) is a heterodimeric heme protein and the primary nitric oxide receptor. NO binding stimulates cyclase activity, leading to regulation of cardiovascular physiology and making sGC an attractive target for drug discovery. YC-1 and related compounds stimulate sGC both independently and synergistically with NO and CO binding; however, where the compounds bind and how they work remain unknown. Using linked equilibrium binding measurements, surface plasmon resonance, and domain truncations in Manduca sexta and bovine sGC, we demonstrate that YC-1 binds near or directly to the heme-containing domain of the ß subunit. In the absence of CO, YC-1 binds with a Kd of 9-21 µM, depending on the construct. In the presence of CO, these values decrease to 0.6-1.1 µM. Pfizer compound 25 bound ∼10-fold weaker than YC-1 in the absence of CO, whereas compound BAY 41-2272 bound particularly tightly in the presence of CO (Kd = 30-90 nM). Additionally, we found that CO binds much more weakly to heterodimeric sGC proteins (Kd = 50-100 µM) than to the isolated heme domain (Kd = 0.2 µM for Manduca ß H-NOX/PAS). YC-1 greatly enhanced binding of CO to heterodimeric sGC, as expected (Kd ∼ 1 µM). These data indicate the α subunit induces a heme pocket conformation with a lower affinity for CO and NO. YC-1 family compounds bind near the heme domain, overcoming the α subunit effect and inducing a heme pocket conformation with high affinity. We propose this high-affinity conformation is required for the full-length protein to achieve high catalytic activity.

An extracellular disulfide bond forming protein (DsbF) from Mycobacterium tuberculosis: structural, biochemical, and gene expression analysis.

Disulfide bond forming (Dsb) proteins ensure correct folding and disulfide bond formation of secreted proteins. Previously, we showed that Mycobacterium tuberculosis DsbE (Mtb DsbE, Rv2878c) aids in vitro oxidative folding of proteins. Here, we present structural, biochemical, and gene expression analyses of another putative Mtb secreted disulfide bond isomerase protein homologous to Mtb DsbE, Mtb DsbF (Rv1677). The X-ray crystal structure of Mtb DsbF reveals a conserved thioredoxin fold although the active-site cysteines may be modeled in both oxidized and reduced forms, in contrast to the solely reduced form in Mtb DsbE. Furthermore, the shorter loop region in Mtb DsbF results in a more solvent-exposed active site. Biochemical analyses show that, similar to Mtb DsbE, Mtb DsbF can oxidatively refold reduced, unfolded hirudin and has a comparable pK(a) for the active-site solvent-exposed cysteine. However, contrary to Mtb DsbE, the Mtb DsbF redox potential is more oxidizing and its reduced state is more stable. From computational genomics analysis of the M. tuberculosis genome, we identified a potential Mtb DsbF interaction partner, Rv1676, a predicted peroxiredoxin. Complex formation is supported by protein coexpression studies and inferred by gene expression profiles, whereby Mtb DsbF and Rv1676 are upregulated under similar environments. Additionally, comparison of Mtb DsbF and Mtb DsbE gene expression data indicates anticorrelated gene expression patterns, suggesting that these two proteins and their functionally linked partners constitute analogous pathways that may function under different conditions.