MétaCan
Menu
Back to cohort

Abstract A012: AI-assisted Design of Novel NRF2Mut Inhibitors

2025· article· en· W4412163905 on OpenAlexaboutno aff
Zhaohui Xiong, Zachary Ladd, Candice Bui-Linh, Francis Spitz, Hai‐Ning Wang, Xin Chen

Bibliographic record

VenueClinical Cancer Research · 2025
Typearticle
Languageen
FieldChemistry
TopicSynthesis and Biological Evaluation
Canadian institutionsnot available
Fundersnot available
KeywordsMedicineComputational biologyPharmacologyBiology

Abstract

fetched live from OpenAlex

Abstract Nuclear factor erythroid 2-related factor 2 (NRF2) is a master transcription factor that plays a central role in orchestrating cellular responses to oxidative and electrophilic stress. NRF2 is regulated by Kelch-like ECH-associated protein 1 (KEAP1), which promotes NRF2 ubiquitination and proteasomal degradation. However, in approximately 2% of human cancers, NRF2 becomes hyperactive due to mutations, leading to uncontrolled NRF2 signaling. This aberrant activation drives chemo- and radioresistance, metabolic reprogramming, and tumor progression, and is associated with poor clinical outcomes. In our previous work, we discovered pyrimethamine (PYR), an FDA-approved antimalarial drug, as a potent NRF2 inhibitor. We validated PYR’s efficacy in NRF2Mut esophageal squamous cell carcinoma (ESCC) cells in vitro and a Nrf2E79Q-driven esophageal phenotype mouse model in vivo. Mechanistically, PYR acts as a dual-function compound: it serves as a molecular glue that facilitates the interaction between NRF2Mut and KEAP1, resulting in enhanced proteasomal degradation, and also functions as a dihydrofolate reductase (DHFR) inhibitor. Encouraged by these findings, PYR has advanced into a Phase I “window of opportunity” clinical trial (NCT05678348). In the current study, we aimed to further understand the molecular glue mechanism of PYR and explore additional NRF2-targeting compounds. We identified triamterene (TRM), a potassium-sparing diuretic, as a second NRF2 inhibitor that similarly promotes the degradation of NRF2W24C by enhancing KEAP1 binding and increasing ubiquitination. Both PYR and TRM significantly reduced the half-life of NRF2W24C. Computational modeling showed that both compounds dock within KEAP1’s Kelch domain, increasing protein-protein interactions and lowering the binding energy between KEAP1 and NRF2Mut. A structure-activity relationship analysis using seven TRM analogs further supported the importance of the Kelch pocket in TRM-induced NRF2 degradation. To develop novel NRF2 inhibitors, we utilized Chemistry42, an AI-driven drug discovery platform, to design compounds targeting the Kelch pocket. Of the four compounds synthesized, one—designated X30—showed strong activity in NRF2Mut ESCC cells but not in NRF2WT cells. Western blot assays confirmed X30-induced proteasomal degradation of NRF2W24C via KEAP1 interaction. Importantly, this effect was lost in KEAP1KO ESCC cells, validating the KEAP1 dependency of X30 and TRM. Notably, unlike PYR, X30 does not act through DHFR inhibition, suggesting it operates via a distinct and KEAP1-dependent mode of action. In summary, X30 represents an entirely new chemical scaffold with no prior reports in the literature and very limited structural similarity to known bioactive compounds. It was exclusively generated using the AI-driven platform Chemistry42, showcasing the creative potential of AI in developing selective NRF2Mut inhibitors. Further evaluation of additional AI-designed compounds is underway to advance therapeutic strategies against NRF2Mut cancers. Citation Format: Zhaohui Xiong, Zachary Ladd, Candice Bui-Linh, Francis Spitz, Haining Wang, Xiaoxin Luke. Chen. AI-assisted Design of Novel NRF2Mut Inhibitors [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Artificial Intelligence and Machine Learning; 2025 Jul 10-12; Montreal, QC, Canada. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(13_Suppl):Abstract nr A012.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.005
metaresearch head score (Gemma)0.006
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesInsufficient payload (model declined to judge)
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Bench or experimental · Consensus signal: Bench or experimental
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.357
Threshold uncertainty score0.998

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0050.006
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0000.000
Science and technology studies0.0000.000
Scholarly communication0.0000.000
Open science0.0000.000
Research integrity0.0000.001
Insufficient payload (model declined to judge)0.0030.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.576
GPT teacher head0.589
Teacher spread0.013 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one teacher head, not a consensus.

Study designBench or experimental
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations0
Published2025
Admission routes1
Has abstractyes

Explore more

Same venueClinical Cancer ResearchSame topicSynthesis and Biological EvaluationFrench-language works237,207