The one-carbon metabolic enzyme MTHFD2 promotes resection and homologous recombination after ionizing radiation

The one-carbon metabolism enzyme bifunctional methylenetetrahydrofolate dehydrogenase/cyclohydrolase 2 (MTHFD2) is among the most overexpressed proteins across tumors and is widely recognized as a promising anticancer target. While MTHFD2 is mainly described as a mitochondrial protein, a new nuclear function is emerging. Here, we observe that nuclear MTHFD2 protein levels and association with chromatin increase following ionizing radiation (IR) in an ataxia telangiectasia mutated (ATM)- and DNA-dependent protein kinase (DNA-PK)-dependent manner. Furthermore, repair of IR-induced DNA double-strand breaks (DSBs) is delayed upon MTHFD2 knockdown, suggesting a role for MTHFD2 in DSB repair. In support of this, we observe impaired recruitment of replication protein A (RPA), reduced resection, decreased IR-induced DNA repair protein RAD51 homolog 1 (RAD51) levels and impaired homologous recombination (HR) activity in MTHFD2-depleted cells following IR. In conclusion, we identify a key role for MTHFD2 in HR repair and describe an interdependency between MTHFD2 and HR proficiency that could potentially be exploited for cancer therapy.

Abstract A020: Structural variants in the pathogenesis of colorectal cancer: The elephant in the room

Background: Cancer is caused by somatic DNA alterations, comprising single/small nucleotide variants (SNVs), somatic copy number alterations (SCNAs) and chromosomal rearrangement structural variants (SVs). We previously demonstrated that SVs are recurrently identified in hundreds of genes and are highly prevalent in common fragile site genes, e.g., in MACROD2 in >40% of colorectal cancers (CRCs). However, computational methods that discriminate SV-driver from SV-passenger events are lacking and laboratory methods to detect SVs at nucleotide resolution from routinely obtained formalin-fixed paraffin-embedded (FFPE) tumor tissue material are underdeveloped. Therefore, despite the abundant presence of SVs, knowledge about their biological and clinical impact is limited. Aim: The aim of our studies is to identify genes of which the function is frequently affected by SV, to understand how these genes contribute to CRC pathogenesis, and to translate these SVs into clinically relevant biomarkers. Methods: We made use of publicly available deep whole genome DNA sequencing data and tumor-matched RNA sequencing data from the Hartwig Medical Foundation to develop the algorithm ‘CoBRA’: Computation of Biologically Relevant Alterations. Adenoma-derived organoids were used for CRISPR/Cas9-mediated gene modulation for functional analysis of SV-driver events. Cergentis’ targeted locus capture (FFPE-TLC) technology was used to detect SVs at nucleotide resolution from FFPE material, which were translated into droplet digital PCR (ddPCR) assays for the detection of SVs in cell-free circulating tumor DNA (ctDNA) in liquid biopsies. Results: The CoBRA algorithm associated the presence of SV-events in frequently affected genes to the extent in which genome-wide RNA sequencing data were altered. In this way, CoBRA ranked SV-events in genes according to their putative impact on tumor biology. SVs in MACROD2 ranked among those with the highest impact on tumor biology. Therefore, we generated focal deletions in MACROD2 in adenoma-derived organoids for functional analyses. Moreover, using FFPE tumor tissue material we detected SVs at nucleotide resolution in MACROD2 and three other genes in 21 out of 29 patients. SVs were verified by PCR on tumor tissue and subsequently translated into ddPCR biomarker assays for detection of SVs in ctDNA in blood from the same patients. Conclusions: We developed the computational method CoBRA and succeeded to detect SVs with high impact on tumor biology. These SVs are prioritized for functional analysis in pre-malignant adenoma-derived organoids; for targeted detection in routinely obtained FFPE tumor tissue material; and for translation into liquid biopsy ctDNA assays. Proof of concept was delivered for MACROD2. Our novel computational and laboratory methodologies provide valuable tools to effectively explore the biological and clinical impact of SVs, which will contribute to our understanding of these common recurrent somatic alterations in CRC and their translation into clinically relevant biomarker applications.

Citation Format: Elise van Bree, Carmen Rubio Alarcón, Soufyan Lakbir, Ellen Stelloo, Caterina Buranelli, Amber Hondema, Iris van 't Erve, Daan Vessies, Pien Delis-van Diemen, Marianne Tijssen, Anne Bolijn, Mirthe Lanfermeijer, Dorothe Linders, Joost Swennenhuis, Daan van den Broek, Jaap Heringa, Gerrit Meijer, Beatriz Carvalho, Harma Feitsma, Sanne Abeln, Remond J. A. Fijneman. Structural variants in the pathogenesis of colorectal cancer: The elephant in the room [abstract]. In: Proceedings of the AACR Special Conference on Colorectal Cancer; 2022 Oct 1-4; Portland, OR. Philadelphia (PA): AACR; Cancer Res 2022;82(23 Suppl_1):Abstract nr A020.