Abstract NG10: First-in-class dual AR/AR-V7 molecular glue degraders targeting the undruggable N-terminal domain to overcome therapeutic resistance in prostate cancer

Abstract Metastatic castration-resistant prostate cancer (mCRPC) remains a uniformly lethal malignancy, driven by persistent dependence on the androgen receptor (AR) signaling pathway. Over 90% of prostate cancer (PC) therapies target the AR ligand-binding domain (LBD). However, the selective pressure imposed by these agents inevitably promotes therapeutic resistance and disease progression, ultimately driving mortality. Expression of AR splice variants, most notably AR-V7, that lack the LBD yet remain constitutively active, is a key mechanism of treatment resistance. Clinically, AR-V7 expression is strongly associated with resistance to all AR pathway inhibitors (ARPIs), including enzalutamide and abiraterone. Alarmingly, AR-V7 is detected in 75% of men with mCRPC, yet there are no approved inhibitors—underscoring a profound unmet clinical need. Thus, directly targeting AR-V7 represents one of the most urgent and unsolved challenges in prostate cancer drug development, with clear implications for survival and quality of life. To address this challenge, we conducted a high-throughput phenotypic screen of 170, 000 small molecules and identified a novel hit compound—the first-in-class dual degrader of AR-V7 and full-length AR, defined by a previously unreported chemotype. Medicinal chemistry/SAR optimization yielded potent lead compounds (IC50 10 nM) with improved selectivity and favorable drug-like properties. This systematic optimization effort combined iterative analog synthesis, structure-activity analysis, and physicochemical profiling to deliver nanomolar potency alongside favorable solubility and stability profiles, enabling downstream in vivo evaluation. In vivo, pharmacokinetic (PK) -optimized compounds significantly suppressed tumor growth in the enzalutamide-resistant 22Rv1 xenograft model, reversed resistance, and outperformed ARV-110, the only full-length AR-PROTAC degrader in clinical trials, by 10-fold in efficacy, with no effect on mouse body weight—establishing both robust antitumor activity and excellent tolerability. Pharmacodynamic (PD) studies revealed near-complete degradation of AR/AR-V7 proteins within 24 hours of final dose, tightly correlating with rapid tumor growth inhibition. Importantly, toxicity studies across multiple dosing regimens showed no histopathological abnormalities in major organs, further supporting the translational potential of this new therapeutic class. Mechanistic studies confirmed rapid degradation of AR-V7/AR via the ubiquitin-proteasome pathway, dependent on Cullin-RING E3 ligase activity. A CRISPR-based E3 ligase screen implicated the DDB1-RBX1-Cullin-RING complex as the mediator of degradation, consistent with a molecular glue mechanism. This represents one of the first demonstrations of a molecular glue-mediated engagement of this E3 ligase complex in prostate cancer, further underscoring the novelty of our approach. Global proteomics demonstrated deep proteome coverage of the entire proteome of ∼10, 000 proteins and exceptional selectivity: AR/AR-V7 ranked among the top 1% of significantly downregulated proteins, while sparing all other major human nuclear receptors, including ER, PR, and GR, and avoiding liabilities of prior PROTAC-based AR degraders—highlighting strong specificity and favorable translational potential for clinical development. Using cellular thermal shift assays, AR truncation mutants, and Cryo-EM studies, we demonstrated direct binding of our lead compounds to the intrinsically disordered N-terminal domain (NTD), shared by both AR and AR-V7. We further mapped the binding site to the TAU1/AF-1 subdomain— the first demonstration of small-molecule engagement of this notoriously “undruggable” region. This provides long-sought proof-of-concept that the unstructured AR-NTD, which has defied decades of drug targeting efforts, can indeed be targeted with small molecules. Our findings therefore expand the druggable proteome to include intrinsically disordered regions, with broad implications beyond prostate cancer. Importantly, our lead compounds were able to overcome enzalutamide resistance and inhibit all clinically validated ARPIs-resistant isoforms, including LBD-domain mutations. These results suggest that our compounds deliver not only a solution for AR-V7-mediated resistance but also a generalized “resistance-proof” strategy by simultaneously targeting multiple clinically relevant AR variants. Taken together, these data establish our lead compounds as first-in-class dual AR-V7/AR molecular glue degraders with a novel NTD-targeted mechanism and strong preclinical activity. To advance toward clinical translation, the next phase of our research will systematically define the structural basis of compound-NTD interaction, evaluate drug metabolism and safety in higher-order animal models, and explore rational combination strategies with standard-of-care agents. The objective here is to define the precise molecular mechanism of action and assess therapeutic potential in clinically relevant models. Collectively, these findings support a paradigm-shifting therapeutic strategy for targeting previously “undruggable” drivers like AR-V7. Unlike all existing AR-directed therapies that target the LBD, this approach simultaneously degrades AR and AR-V7 in a single treatment, directly addressing the central clinical challenge of therapeutic resistance. Moreover, by establishing proof-of-concept for disordered protein domain-targeted glue degraders, this work opens a broadly applicable platform to drug other transcription factors and intrinsically disordered proteins implicated across multiple cancers and diseases. By enabling dual degradation of AR (the initiating oncogenic driver in PC) and AR-V7 (the lethal, treatment-resistant variant), our drug candidate has the potential to transform the treatment landscape—benefiting patients with both hormone-sensitive and advanced castration-resistant disease, while delaying or preventing AR-V7-mediated progression. Ultimately, this work charts a new course for prostate cancer therapy by providing a first-in-class approach capable of addressing the major cause of therapeutic failure, with the potential to meaningfully extend survival and improve quality of life. Citation Format: CheukMan Cherie Au, Michelle Naidoo, Catrina Estrella, Kiran Kumari Sahu, Charles D. Warren, Prerna Vatsa, Ying Wang, Zhao Wang, David M. Nanus, Urko del Castillo, Jacob B. Geri, Paraskevi Giannakakou. First-in-class dual AR/AR-V7 molecular glue degraders targeting the undruggable N-terminal domain to overcome therapeutic resistance in prostate cancer abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts) ; 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86 (8Suppl): Abstract nr NG10.