Abstract 6441: Discovery of ARID1A loss as a patient biomarker for NXP800 - A developmental activator of the integrated stress response (ISR) and inhibitor of the HSF1 pathway in ovarian cancer
Robert H. te Poele,Marissa Powers,Swee Sharp,Emmanuel de Billy,Maria Taskinen,Loredana Pellegrino,Sharon Gowan,Asadh Miah,Angela Hayes,Matthew Cheeseman,Keith Jones,Suzanne Eccles,Florence Raynaud,Paul Clarke,Paul Workman
DOI: https://doi.org/10.1158/1538-7445.am2024-6441
IF: 11.2
2024-03-22
Cancer Research
Abstract:Abstract Background. HSF1 helps cancer cells cope with multiple stresses caused by oncogene activation. Methods. See details below. Results. We discovered the bisamide NXP800 as an inhibitor of HSF1-mediated transcription through phenotypic screening and med chem optimization. Gene expression microarray analysis of human cancer cell lines and tumor xenografts treated with bisamide inhibitors indicated activation of the ATF4 axis of the ISR as a key mechanism of action. In SK-OV-3 ovarian cancer xenografts, NXP800 showed clear PK/PD relationships with increased expression of ATF4 transcriptional target genes alongside decreased expression of HSF1 transcriptional targets. NXP800 also caused tumor regressions in this model. Expanding in additional human ovarian cancer xenograft models, we observed efficacy in five and a complete lack of response in three others. The sensitive models all had homozygous deleterious mutations in the ARID1A gene, whereas the non-responding xenografts were all wild type (WT). ARID1A is a component of the SWI/SNF chromatin remodelling complex involved in repression and activation of target genes. Subsequent screening of the large Sanger human cancer cell line panel confirmed ARID1A as the most significant common disease-related alteration predicting sensitivity to NXP800 in ovarian cancer cell lines. This predictive relationship was confirmed in an ARID1A isogenic HCT-116 cell line pair; sensitivity was greater in the homozygous ARID1A mutant cells compared to WT, resulting in PARP cleavage in the mutant cells only. In vivo there was no effect in WT HCT-116 xenografts whereas growth inhibition was observed in the homozygous mutant cells, resulting in significantly smaller tumors. In addition, the induction of ATF4 target genes was stronger and more prolonged in the mutant cells. It is known that whereas short term ATF4 activation is adaptive, persistent activation can promote the induction of apoptosis. CHIP-seq analysis confirmed clear relationships between ARID1A status, ATF4 and HSF1 promoter occupancy, and the distribution of BRG1 and RNA pol II at target sites, although these were often gene-specific and complex. A relatively simple example is the regulation of the INHBE gene (expression of which is a PD biomarker). In untreated samples there is no binding of ATF4, BRG1 and RNA pol II at the INHBE promoter. NXP800 treatment results in the recruitment of BRG1, ATF4 and RNA pol II in ARID1A mutant TOV-21G cells but not in the RMGI WT cells, leading to increased expression in the mutant cells only. Conclusions. We propose that ARID1A loss alters the binding and recruitment of ATF4 and HSF1 leading to the altered and prolonged expression of ATF4 target genes and increased sensitivity to NXP800. NXP800 is currently in phase Ib for the treatment of ARID1A mutant platinum resistant ovarian cancer (NCT05226507). Citation Format: Robert H. te Poele, Marissa Powers, Swee Sharp, Emmanuel de Billy, Maria Taskinen, Loredana Pellegrino, Sharon Gowan, Asadh Miah, Angela Hayes, Matthew Cheeseman, Keith Jones, Suzanne Eccles, Florence Raynaud, Paul Clarke, Paul Workman. Discovery of ARID1A loss as a patient biomarker for NXP800 - A developmental activator of the integrated stress response (ISR) and inhibitor of the HSF1 pathway in ovarian cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 6441.
oncology