Multidimensional, integrative profiling identifies BCL2L1 methylation as a predictor of MCL1 dependency in pediatric malignancies.

Abstract

Pediatric high-grade gliomas (pHGGs) are the most aggressive brain tumors in children, necessitating innovative therapies to improve outcomes. Unlike adult gliomas, recent research reveals that childhood gliomas have distinct biological features, requiring specific treatment strategies. Here, we focused on deciphering unique genetic dependencies specific to childhood gliomas. Using a pooled CRISPR/Cas9 knockout screening approach on 65 pediatric and 10 adult high-grade glioma (HGG) cell lines, myeloid cell leukemia 1 (MCL1) emerged as a key antiapoptotic gene essential in pediatric but not adult gliomas. We demonstrated that MCL1 is targetable using current small molecule inhibitors, and its inhibition leads to potent anticancer activity across pediatric HGG cell lines irrespective of genotype. Employing predictive modeling approaches on a large set of childhood cancer cell lines with multiomics data features, we identified a potentially previously unreported cluster of CpG sites in the antiapoptotic BCL-xL/BCL2L1 gene, which predicted MCL1 inhibitor response. We extended these data across multiple pediatric tumor types, showing that BCL2L1 methylation is a broad predictor of MCL1 dependency in vitro and in vivo. Overall, our multidimensional, integrated genomic approach identified MCL1 as a promising therapeutic target in several BCL2L1-methylated pediatric cancers, offering a translational strategy to identify patients most likely to benefit from MCL1 inhibitor therapy.