From Subtle to Severe: Multiple Mechanisms Impair POLE and POLD1 Function leading to Distinct Effects on Mutation Rates

Abstract

Background POLE and POLD1 are DNA polymerase proteins crucial for maintaining genome integrity during replication. While several drivers leading to POLE and POLD1 proofreading deficiency (PPD) have been identified, the complete set of variants causing PPD is not fully established. This study aims to identify new variants that contribute to PPD by analyzing a large dataset of sequenced tumors. Methods We analyzed 204,453 tumors to identify variants associated with POLE and POLD1 proofreading deficiency after extensive filtering and normalizing of the different datasets. Using the support vector machine algorithm, we classified variants as either drivers or non-drivers of PPD based on the number of mutations and mutational signature activity. A saturation analysis was performed to determine the completeness of the identified variants. Results We identified 46 somatic driver variants, compared to only 17 known driver variants, thus more than doubling the current catalog of mutations linked to PPD. This extensive dataset enabled us to characterize the clinical features of PPD tumors, noting a male bias and younger age at diagnosis compared to non-PPD tumors. Different driver variants led to distinct phenotypes in mutational patterns and subsequent mutated genes.