Novel Protac-Based EZH2 Degraders Effectively Target EZH2-Driven Lymphoid Malignancies By Abolishing Both Methylation-Dependent and -Independent Functions

Across various cancers, overexpression or mutation of the epigenetic silencer EZH2 correlates with invasive cancer growth, resistance and poor survival outcomes. While the tumorigenic role of EZH2 is largely attributed to the H3K27me3-mediated repression of tumor suppressor genes, reports have also established a methylation-independent role for EZH2 in the transcriptional activation of oncogenes, particularly in aggressive cancers like Natural Killer T-cell Lymphoma (NKTL). Here, traditional EZH2 inhibitors have been ineffective. We hypothesized that targeted depletion of the EZH2 protein will more effectively disrupt oncogenic signals arising from enzyme-independent activities of EZH2. Thus, we embarked on the development of EZH2 Proteolysis Targeting Chimeras (PROTACs) to trigger targeted ubiquitination and proteasome-mediated degradation of EZH2. No clinical EZH2 PROTAC has been reported yet.

Our initial efforts focused on the iterative synthesis of a series of PROTAC candidates considering factors such as permeability, solubility and stable ternary complex formation. We developed a library of 53 compounds with various permutations of EZH2- and E3 ligase-binding ligands as well as linker lengths. Through extensive screening and testing of the EZH2 degradation activity via the Nano-Glo HiBiT lytic detection system and immunoblotting, we identified three lead Protac candidates. Compared to the clinical EZH2 inhibitor Tazemetostat (Taz), EZH2 PROTACs, C2911 and C1311 demonstrated higher cytotoxic activity in diffuse large B-cell lymphoma (DLBCL), a cancer in which Taz was evaluated during clinical trial. Importantly, C2911 and C1311 also successfully triggered EZH2 degradation in a dose-dependent manner thereby qualifying for further evaluation in NKTL.

Previous work has highlighted a critical role for JAK3-mediated phosphorylation of EZH2 in promoting NKTL proliferation. Treatment of NKTL with C2911 and C1311 leads to a dose and time-dependent EZH2 degradation, loss of cell viability and increased apoptosis. These effects were absent with Taz and Pomalidomide treatments. The concomitant loss of H3K27me3 levels as well as PRC2 complex subunits, SUZ12 and EED suggest effective degradation of the EZH2 protein. We also observed that induction of NKTL cell death required a longer treatment period as compared to DLBCL. This may be potentially ascribed to the longer half-life of the EZH2 protein in NKTL as determined through cycloheximide treatment.

In order to validate the mechanism of action of the EZH2 PROTACs, cells were either pretreated with pomalidomide or siRNA targeting cereblon. Indeed, this mitigated PROTAC-induced degradation of the EZH2 protein and rescued cancer cell death. Similarly, inhibiting the activity of the 26S proteasome through bortezomib also prevented EZH2 degradation. Characterization of EZH2 PROTAC-induced protein expression changes through immunoblotting highlighted that there was specific loss of H3K27me3 and H3K27me2 but not H3K27me1. Of note, we detected a concomitant decrease of Jak3 protein expression and STAT5 phosphorylation levels, more pronounced than PRC2 complex subunits. This suggests that the PROTAC may act by disrupting the non-canonical binding of EZH2 to Jak3 in NKTL.

Furthermore, gene expression analyses revealed that EZH2 PROTAC treatment more effectively reactivated canonical EZH2 target genes such as MYT1 and ANPEP as compared to Taz. As the JAK-STAT-MYC axis is dysregulated in NK leukemia, we also evaluated the gene expression of c-Myc and c-Myc target gene, E2F1. EZH2 PROTAC treatment reduces the transcription levels of these oncogenes corroborating with loss of c-Myc protein expression. Studying the pharmacokinetics of the lead PROTACs in a mouse model revealed good plasma stability, half-life (> 4 h), and low clearance.

Collectively, our data supports a critical role for EZH2 PROTACs in abolishing both the enzyme-dependent and independent activities of EZH2 through protein degradation, resulting in enhanced anti-cancer activity against DLBCL and NKTL. This may potentially benefit other EZH2-driven cancers. Further evaluation of the EZH2 PROTACs for clinical application is in progress. EZH2 PROTACs will next be tested in vivo. An in-depth investigation of the mechanisms stabilizing EZH2 protein levels and underlying PROTAC induced-cell death will be made to identify potential mechanisms of resistance.