Combining Cloud-Based Free-Energy Calculations, Synthetically Aware Enumerations, and Goal-Directed Generative Machine Learning for Rapid Large-Scale Chemical Exploration and Optimization.

The hit identification process usually involves the profiling of millions to more recently billions of compounds either via traditional experimental high-throughput screens (HTS) or computational virtual high-throughput screens (vHTS). We have previously demonstrated that, by coupling reaction-based enumeration, active learning, and free energy calculations, a similarly large-scale exploration of chemical space can be extended to the hit-to-lead process. In this work, we augment that approach by coupling large scale enumeration and cloud-based free energy perturbation (FEP) profiling with goal-directed generative machine learning, which results in a higher enrichment of potent ideas compared to large scale enumeration alone, while simultaneously staying within the bounds of predefined drug-like property space. We can achieve this by building the molecular distribution for generative machine learning from the PathFinder rules-based enumeration and optimizing for a weighted sum QSAR-based multiparameter optimization function. We examine the utility of this combined approach by designing potent inhibitors of cyclin-dependent kinase 2 (CDK2) and demonstrate a coupled workflow that can (1) provide a 6.4-fold enrichment improvement in identifying <10 nM compounds over random selection and a 1.5-fold enrichment in identifying <10 nM compounds over our previous method, (2) rapidly explore relevant chemical space outside the bounds of commercial reagents, (3) use generative ML approaches to "learn" the SAR from large scale in silico enumerations and generate novel idea molecules for a flexible receptor site that are both potent and within relevant physicochemical space, and (4) produce over 3 000 000 idea molecules and run 1935 FEP simulations, identifying 69 ideas with a predicted IC50 < 10 nM and 358 ideas with a predicted IC50 < 100 nM. The reported data suggest combining both reaction-based and generative machine learning for ideation results in a higher enrichment of potent compounds over previously described approaches and has the potential to rapidly accelerate the discovery of novel chemical matter within a predefined potency and property space.

Reaction-Based Enumeration, Active Learning, and Free Energy Calculations To Rapidly Explore Synthetically Tractable Chemical Space and Optimize Potency of Cyclin-Dependent Kinase 2 Inhibitors.

The hit-to-lead and lead optimization processes usually involve the design, synthesis, and profiling of thousands of analogs prior to clinical candidate nomination. A hit finding campaign may begin with a virtual screen that explores millions of compounds, if not more. However, this scale of computational profiling is not frequently performed in the hit-to-lead or lead optimization phases of drug discovery. This is likely due to the lack of appropriate computational tools to generate synthetically tractable lead-like compounds in silico, and a lack of computational methods to accurately profile compounds prospectively on a large scale. Recent advances in computational power and methods provide the ability to profile much larger libraries of ligands than previously possible. Herein, we report a new computational technique, referred to as "PathFinder", that uses retrosynthetic analysis followed by combinatorial synthesis to generate novel compounds in synthetically accessible chemical space. In this work, the integration of PathFinder-driven compound generation, cloud-based FEP simulations, and active learning are used to rapidly optimize R-groups, and generate new cores for inhibitors of cyclin-dependent kinase 2 (CDK2). Using this approach, we explored >300 000 ideas, performed >5000 FEP simulations, and identified >100 ligands with a predicted IC50 < 100 nM, including four unique cores. To our knowledge, this is the largest set of FEP calculations disclosed in the literature to date. The rapid turnaround time, and scale of chemical exploration, suggests that this is a useful approach to accelerate the discovery of novel chemical matter in drug discovery campaigns.

Multi-omic Dissection of Oncogenically Active Epiproteomes Identifies Drivers of Proliferative and Invasive Breast Tumors.

Proliferative and invasive breast tumors evolve heterogeneously in individual patients, posing significant challenges in identifying new druggable targets for precision, effective therapy. Here we present a functional multi-omics method, interaction-Correlated Multi-omic Aberration Patterning (iC-MAP), which dissects intra-tumor heterogeneity and identifies in situ the oncogenic consequences of multi-omics aberrations that drive proliferative and invasive tumors. First, we perform chromatin activity-based chemoproteomics (ChaC) experiments on breast cancer (BC) patient tissues to identify genetic/transcriptomic alterations that manifest as oncogenically active proteins. ChaC employs a biotinylated small molecule probe that specifically binds to the oncogenically active histone methyltransferase G9a, enabling sorting/enrichment of a G9a-interacting protein complex that represents the predominant BC subtype in a tissue. Second, using patient transcriptomic/genomic data, we retrospectively identified some G9a interactor-encoding genes that showed individualized iC-MAP. Our iC-MAP findings represent both new diagnostic/prognostic markers to identify patient subsets with incurable metastatic disease and targets to create individualized therapeutic strategies.

Structure-Activity Relationship Studies for Enhancer of Zeste Homologue 2 (EZH2) and Enhancer of Zeste Homologue 1 (EZH1) Inhibitors.

EZH2 or EZH1 (enhancer of zeste homologue 2 or 1) is the catalytic subunit of polycomb repressive complex 2 (PRC2) that catalyzes methylation of histone H3 lysine 27 (H3K27). PRC2 hyperactivity and/or hypertrimethylation of H3K27 are associated with numerous human cancers, therefore inhibition of PRC2 complex has emerged as a promising therapeutic approach. Recent studies have shown that EZH2 and EZH1 are not functionally redundant and inhibition of both EZH2 and EZH1 is necessary to block the progression of certain cancers such as mixed-lineage leukemia (MLL)-rearranged leukemias. Despite the significant advances in discovery of EZH2 inhibitors, there has not been a systematic structure-activity relationship (SAR) study to investigate the selectivity between EZH2 and EZH1 inhibition. Here, we report our SAR studies that focus on modifications to various regions of the EZH2/1 inhibitor UNC1999 (5) to investigate the impact of the structural changes on EZH2 and EZH1 inhibition and selectivity.

Targeting EZH2 and PRC2 dependence as novel anticancer therapy.

Distinctive patterns of chromatin modification control gene expression and define cellular identity during development and cell differentiation. Polycomb repressive complex 2 (PRC2), the sole mammalian enzymatic complex capable of establishing gene-repressive high-degree methylation of histone H3 at lysine 27 (H3K27), plays crucial roles in regulation of normal and malignant hematopoiesis. Recently, increasing evidence has indicated that recurrent gain-of-function mutation and overexpression of EZH2, the catalytic subunit of PRC2, drive and promote malignant transformation such as B-cell lymphomagenesis, providing a rationale for PRC2 inhibition as a novel anticancer strategy. Here, we summarize the recently developed strategies for inhibition of PRC2, which include a series of highly specific, highly potent, small-molecule inhibitors of EZH2 and EZH1, an EZH2-related methyltransferase. PRC2 establishes functional crosstalk with numerous epigenetic machineries during dynamic regulation of gene transcription. Perturbation of such functional crosstalk caused by genetic events observed in various hematologic cancers, such as inactivation of SNF5 and somatic mutation of UTX, confers PRC2 dependence, thus rendering an increased sensitivity to PRC2 inhibition. We discuss our current understanding of EZH2 somatic mutations frequently found in B-cell lymphomas and recurrent mutations in various other epigenetic regulators as novel molecular predictors and determinants of PRC2 sensitivity. As recent advances have indicated a critical developmental or tumor-suppressive role for PRC2 and EZH2 in various tissue types, we discuss concerns over potentially toxic or even adverse effects associated with EZH2/1 inhibition in certain biological contexts or on cancer genetic background. Collectively, inhibition of PRC2 catalytic activity has emerged as a promising therapeutic intervention for the precise treatment of a range of genetically defined hematologic malignancies and can be potentially applied to a broader spectrum of human cancers that bear similar genetic and epigenetic characteristics.

Polycomb genes, miRNA, and their deregulation in B-cell malignancies.

Posttranslational modifications of histone proteins represent a fundamental means to define distinctive epigenetic states and regulate gene expression during development and differentiation. Aberrations in various chromatin-modulation pathways are commonly used by tumors to initiate and maintain oncogenesis, including lymphomagenesis. Recently, increasing evidence has demonstrated that polycomb group (PcG) proteins, a subset of histone-modifying enzymes known to be crucial for B-cell maturation and differentiation, play a central role in malignant transformation of B cells. PcG hyperactivity in B-cell lymphomas is caused by overexpression or recurrent mutations of PcG genes and deregulation of microRNAs (miRNAs) or transcription factors such as c-MYC, which regulate PcG expression. Interplays of PcG and miRNA deregulations often establish a vicious signal-amplification loop in lymphoma associated with adverse clinical outcomes. Importantly, aberrant enzymatic activities associated with polycomb deregulation, notably those caused by EZH2 gain-of-function mutations, have provided a rationale for developing small-molecule inhibitors as novel therapies. In this review, we summarize our current understanding of PcG-mediated gene silencing, interplays of PcG with other epigenetic regulators such as miRNAs during B-cell differentiation and lymphomagenesis, and recent advancements in targeted strategies against PcG as promising therapeutics for B-cell malignancies.

Selective inhibition of EZH2 and EZH1 enzymatic activity by a small molecule suppresses MLL-rearranged leukemia.

Enhancer of zeste homolog 2 (EZH2) and related EZH1 control gene expression and promote tumorigenesis via methylating histone H3 at lysine 27 (H3K27). These methyltransferases are ideal therapeutic targets due to their frequent hyperactive mutations and overexpression found in cancer, including hematopoietic malignancies. Here, we characterized a set of small molecules that allow pharmacologic manipulation of EZH2 and EZH1, which include UNC1999, a selective inhibitor of both enzymes, and UNC2400, an inactive analog compound useful for assessment of off-target effect. UNC1999 suppresses global H3K27 trimethylation/dimethylation (H3K27me3/2) and inhibits growth of mixed lineage leukemia (MLL)-rearranged leukemia cells. UNC1999-induced transcriptome alterations overlap those following knockdown of embryonic ectoderm development, a common cofactor of EZH2 and EZH1, demonstrating UNC1999's on-target inhibition. Mechanistically, UNC1999 preferentially affects distal regulatory elements such as enhancers, leading to derepression of polycomb targets including Cdkn2a. Gene derepression correlates with a decrease in H3K27me3 and concurrent gain in H3K27 acetylation. UNC2400 does not induce such effects. Oral administration of UNC1999 prolongs survival of a well-defined murine leukemia model bearing MLL-AF9. Collectively, our study provides the detailed profiling for a set of chemicals to manipulate EZH2 and EZH1 and establishes specific enzymatic inhibition of polycomb repressive complex 2 (PRC2)-EZH2 and PRC2-EZH1 by small-molecule compounds as a novel therapeutics for MLL-rearranged leukemia.

Selective inhibitors of protein methyltransferases.

Mounting evidence suggests that protein methyltransferases (PMTs), which catalyze methylation of histone and nonhistone proteins, play a crucial role in diverse biological processes and human diseases. In particular, PMTs have been recognized as major players in regulating gene expression and chromatin state. PMTs are divided into two categories: protein lysine methyltransferases (PKMTs) and protein arginine methyltransferases (PRMTs). There has been a steadily growing interest in these enzymes as potential therapeutic targets and therefore discovery of PMT inhibitors has also been pursued increasingly over the past decade. Here, we present a perspective on selective, small-molecule inhibitors of PMTs with an emphasis on their discovery, characterization, and applicability as chemical tools for deciphering the target PMTs' physiological functions and involvement in human diseases. We highlight the current state of PMT inhibitors and discuss future directions and opportunities for PMT inhibitor discovery.

A chemical tool for in vitro and in vivo precipitation of lysine methyltransferase G9a.

Here we report the design, synthesis, and biochemical characterization of a new chemical tool, UNC0965. UNC0965 is a biotinylated version of our previously reported G9a chemical probe, UNC0638. Importantly, UNC0965 maintains high in vitro potency and is cell penetrant. The biotinylated tag of UNC0965 enables "chemiprecipitation" of G9a from whole cell lysates. Further, the cell penetrance of UNC0965 allowed us to explore the localization of G9a on chromatin both in vitro and in vivo through chemical inhibitor-based chromatin immunoprecipitation (chem-ChIP).

An orally bioavailable chemical probe of the Lysine Methyltransferases EZH2 and EZH1.

EZH2 or EZH1 is the catalytic subunit of the polycomb repressive complex 2 that catalyzes methylation of histone H3 lysine 27 (H3K27). The trimethylation of H3K27 (H3K27me3) is a transcriptionally repressive post-translational modification. Overexpression of EZH2 and hypertrimethylation of H3K27 have been implicated in a number of cancers. Several selective inhibitors of EZH2 have been reported recently. Herein we disclose UNC1999, the first orally bioavailable inhibitor that has high in vitro potency for wild-type and mutant EZH2 as well as EZH1, a closely related H3K27 methyltransferase that shares 96% sequence identity with EZH2 in their respective catalytic domains. UNC1999 was highly selective for EZH2 and EZH1 over a broad range of epigenetic and non-epigenetic targets, competitive with the cofactor SAM and non-competitive with the peptide substrate. This inhibitor potently reduced H3K27me3 levels in cells and selectively killed diffused large B cell lymphoma cell lines harboring the EZH2(Y641N) mutant. Importantly, UNC1999 was orally bioavailable in mice, making this inhibitor a valuable tool for investigating the role of EZH2 and EZH1 in chronic animal studies. We also designed and synthesized UNC2400, a close analogue of UNC1999 with potency >1,000-fold lower than that of UNC1999 as a negative control for cell-based studies. Finally, we created a biotin-tagged UNC1999 (UNC2399), which enriched EZH2 in pull-down studies, and a UNC1999-dye conjugate (UNC2239) for co-localization studies with EZH2 in live cells. Taken together, these compounds represent a set of useful tools for the biomedical community to investigate the role of EZH2 and EZH1 in health and disease.