Disrupting the epigenetic alliance: structural insights and therapeutic strategies targeting DNMT1–UHRF1

Abstract

Maintenance DNA methylation relies on a coordinated partnership between DNMT1 and its chromatin cofactor UHRF1. UHRF1’s SRA domain flips 5-methylcytosine out of hemimethylated DNA, and UHRF1-installed ubiquitin marks on histone H3 (H3K18/K23Ub; H3Ub₂) and PAF15 (PAF15Ub₂) are recognized by the DNMT1 RFTS domain to relieve autoinhibition and license copying of parental methylation during S phase. Tumors often upregulate this axis to enforce promoter hypermethylation programs, whereas approved azanucleosides act via DNMT1 trapping and are associated with DNA-damage–linked toxicities. Over ~ 15 years of structural work—from the 2008 SRA–DNA complexes to a 2022 cryo-EM structure of DNMT1 engaged with hemimethylated DNA and H3Ub₂—has mapped two tractable sites: the UHRF1-SRA aromatic cage and the ubiquitin-binding surface on DNMT1’s RFTS. These insights catalyzed small-molecule discovery. The anthraquinone UM63 validated SRA-pocket engagement but intercalates into DNA; newer non-intercalating SRA-directed inhibitors AMSA-2 (hydroxyanthracene/anthrarobin) and MPB-7 (imidazoquinoline) retain low-micromolar potency. In cells, AMSA-2 and MPB-7 disrupt UHRF1/DNMT1 colocalization at replication foci and induce replication-coupled global hypomethylation, with preferential cytotoxicity in UHRF1-high cancer lines relative to non-transformed cells. Beyond SRA antagonism, DNMT1 can be down-regulated pharmacologically: the non-nucleoside inhibitor GSK-3,484,862 triggers proteasome-dependent DNMT1 degradation alongside hypomethylation, and the first DNMT1-targeting PROTAC (KW0113) achieves selective DNMT1 degradation and growth inhibition in AML models. Remaining hurdles include potency ceilings, nuclear exposure/pharmacokinetics, and adaptive chromatin rewiring upon DNMT1 inhibition; nonetheless, structure-guided optimization and degrader strategies outline a credible path to precision epigenetic therapeutics that directly disrupt the DNMT1–UHRF1 maintenance machinery.