Work

Integrative computational evaluation of tazemetostat for ARID1A-deficient anaplastic thyroid cancer via predicted polypharmacology and epigenetic target engagement

Year: 2026

Type: article

Abstract: Anaplastic thyroid carcinoma (ATC) is a rare, highly aggressive malignancy with limited therapeutic options, due largely to rapid metastasis and treatment resistance. A potential vulnerability in a subset of ATC cases is the loss of ARID1A, a core subunit of the SWI/SNF chromatin remodeling complex, which induces a synthetic lethal dependency on EZH2, the catalytic subunit of the opposing PRC2 complex. Although tazemetostat was previously approved for selected cancers, recent safety concerns and market withdrawal have limited its direct clinical applicability; nevertheless, it remains a clinically characterized EZH2 inhibitor useful for mechanistic interrogation of ARID1A-EZH2 dependency. This study used tazemetostat as a mechanistic computational probe to evaluate EZH2-centered vulnerability and predicted protein interactions in ARID1A-deficient thyroid cancer. Through network pharmacology, we identified five hub targets central to ARID1A-deficient thyroid cancer progression: EZH2, EGFR, MAPK1, GSK3B, and CYP3A4. Molecular docking predicted favorable binding of tazemetostat to all targets, with docking scores ranging from ΔG = − 9.04 kcal/mol (EZH2) to − 13.27 kcal/mol (EGFR). Triplicate molecular dynamics simulations supported overall pose stability across the five complexes, with target-dependent inter-replica variability. Statistical analysis indicated the strongest inter-replica consistency for EZH2 and MAPK1, whereas CYP3A4, EGFR, and GSK3B showed significant replica-level differences despite maintaining stable overall trajectory behavior. Our findings support tazemetostat as a mechanistic computational probe for investigating EZH2-centered vulnerability in ARID1A-deficient anaplastic thyroid cancer, with predicted binding to EZH2 and additional cancer-relevant proteins. These results suggest potential polypharmacology; however, docking and MD simulations demonstrate binding stability rather than confirmed functional inhibition or pathway modulation. Experimental validation will be required to establish the biological significance of these predicted interactions.

Source: Clinical Epigenetics

DOI: 10.1186/s13148-026-02222-w

Authors: Zhang Haoyong, Qian Cheng

Institution: Shaoxing People's Hospital

Language: English

Topic: Computational Drug Discovery Methods

Subfield: Computational Theory and Mathematics

Field: Computer Science

Domain: Physical Sciences

SDG: Good health and well-being

Open Access status: gold

Estimated APC paid: $3,890