Abstract NG07: Discovery of TRIM21 molecular glues that drive potent anti-tumor efficacy via nuclear pore complex degradation

Abstract Cancer cells are acutely dependent on nuclear transport due to elevated transcriptional activity, suggesting a therapeutic opportunity for inhibition of the nuclear pore complex (NPC). Through a combination of large-scale phenotypic profiling, genome-scale functional genomics, and mass spectrometry-based proteomics, we discovered that the clinical drug PRLX-93936 (PRLX) is a molecular glue that binds and reprograms the TRIM21 E3 ubiquitin ligase to degrade the NPC, resulting in potent inhibition of pancreatic ductal adenocarcinoma (PDAC) cells. PRLX, an optimized derivative of erastin, was previously tested in phase 1 trials with modest activity. Importantly, unlike erastin, PRLX induced apoptosis rather than ferroptosis via an unknown molecular target. To identify sensitive cancer models, we profiled PRLX against ∼900 cell lines using the PRISM barcoded screening approach, which demonstrated lineage-specific activity against PDAC and head and neck cancers. This activity strongly correlated with high mRNA expression of the TRIM21 E3 ligase. Next, we conducted genome-scale CRISPR/Cas9 knockout and activation screens, which revealed TRIM21 as the top functional mediator of PRLX response across the entire genome. TRIM21 knockout conferred complete resistance to PRLX, while TRIM21 overexpression was sufficient to sensitize cells to the drug. Given the strong functional evidence, we hypothesized that TRIM21 is the direct molecular target of PRLX. TRIM21 is a cytosolic antibody receptor known for its role in degrading intracellular pathogens. We first performed a tiled mutagenesis screen of the TRIM21 gene, which identified multiple mutations clustering at a putative binding pocket in the PRYSPRY substrate recognition domain. We then confirmed direct compound binding to the recombinant TRIM21 PRYSPRY domain via surface plasmon resonance (SPR). Since TRIM21 is a ubiquitin ligase, we sought to identify its potential neo-substrates upon compound treatment. We performed extensive proteomic profiling to evaluate changes in protein abundance, ubiquitination, and TRIM21 proximity labeling. All three methods unexpectedly identified nuclear pore complex proteins as the top hits. Re-analysis of our CRISPR knockout screen data revealed that a single guide RNA (out of four) against NUP98 conferred strong resistance. This active guide uniquely targeted the NUP98 autoproteolytic domain (APD) and yielded complete resistance to PRLX. Using multiple orthogonal assays (NanoBiT, TR-FRET, and SPR), we demonstrated that the compound acts as a molecular glue to induce physical proximity between TRIM21 and NUP98. This molecular glue mechanism is highly effective against cancer because, as prior work shows, cancer cells undergo apoptosis upon NPC disruption while normal cells can recover. We confirmed this selective vulnerability. Compound treatment led to a loss of short-lived pro-survival mRNAs in cancer cells (by qPCR and RNAscope), and live-cell imaging showed that while the proliferation of non-malignant cells was slowed, PDAC cells underwent rapid and irreversible cell death. These findings support a model where temporary, TRIM21-mediated NPC degradation creates a therapeutic window by preferentially inducing cancer cell death. Using PRLX as a starting point, phenotype-guided medicinal chemistry optimization yielded novel compounds with 10-fold greater potency, improved drug-like properties, and robust oral bioavailability. An improved derivative, JWZ-8-103, was successfully co-crystallized with the TRIM21 PRYSPRY domain, structurally confirming the binding mode and enabling further optimization. To establish translational potential, we employed patient-derived PDAC organoids and in vivo models. Prospectively selecting organoids based on TRIM21 expression confirmed its potential as a predictive biomarker of compound efficacy. Furthermore, optimized compounds demonstrated robust anti-PDAC efficacy in vivo against multiple KRAS-mutant xenograft models. In summary, our work demonstrates the power of unbiased phenotypic screening to uncover novel cancer biology and develop targeted therapeutics. We have discovered potent, orally bioavailable TRIM21 molecular glues that degrade the NPC, validated a predictive biomarker, and established a strong therapeutic rationale and preclinical efficacy for their use in solid tumors. Ongoing work is focused on developing compounds with enhanced potency, evaluating drug combinations, and performing preclinical toxicology studies. The discovery of TRIM21-mediated NPC degraders presents an unexpected and promising new strategy for the treatment of PDAC and other challenging solid tumors. Citation Format: Steven M. Corsello. Discovery of TRIM21 molecular glues that drive potent anti-tumor efficacy via nuclear pore complex degradation 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 NG07.