Back

Co-occurring mutations in the POLE exonuclease and non-exonuclease domains define a unique subset of highly mutagenic tumors.

Shah, S. S.; Demidova, E. V.; Ringenbach, S.; Faezov, B.; Andrake, M.; Mur, P.; Viana-Errasti, J.; Xiu, J.; Swensen, J.; Valle, L.; Dunbrack, R.; Hall, M. J.; Arora, S.

2022-12-16 cancer biology
10.1101/2022.12.15.520593 bioRxiv
Show abstract

Somatic POLE mutations in the exonuclease domain (ExoD) are prevalent in colorectal cancer (CRC), endometrial cancer (EC), and others and typically lead to dramatically increased tumor mutation burden (TMB). To understand whether non-ExoD mutations also play a role in mutagenesis, we assessed TMB in 447/14541 POLE-mutated CRCs, ECs, and ovarian cancers (OC) based on classification TMB-High (TMB-H) or TMB-Low (TMB-L). TMB-H tumors were segregated as POLE ExoD driver, POLE ExoD driver plus POLE Variant, and POLE Variant TMB-H. Intriguingly, TMB was highest in tumors bearing POLE ExoD driver plus POLE Variant (p<0.001 in CRC and EC, p<0.05 in OC). Integrated analysis of AlphaFold2-modeled POLE models and quantitative estimate of stability indicated that multiple variants had significant impact on functionality. These data indicate that co-occurring POLE variants categorize a unique subset of POLE-driven tumors defined by ultra-high TMB, which has implications for abundance of tumor neoantigens, therapeutic response, and patient outcomes. SignificanceSomatic POLE ExoD driver mutations cause proofreading deficiency that induces high tumor mutation burden (TMB). This study defines a novel modifier role for non-ExoD mutations in POLE ExoD-driven tumors, associated with ultra-high TMB. These data may inform acquisition of tumor neoantigens, tumor classification, therapeutic response, and patient outcomes.

Matching journals

The top 9 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.