A pooled mutational analysis identifies ionizing radiation-associated mutational signatures conserved between mouse...

Area of Science:

• Cancer Genomics
• Mutational Signatures Analysis
• Comparative Oncology

Background:

• Single nucleotide variants (SNVs) in cancer genomes can reveal mutational signatures linked to specific carcinogens.
• Non-negative matrix factorization (NMF) is a computational method used to de-convolute SNVs into mutational signatures.
• The accuracy of NMF-derived signatures is dependent on the quantity of available sequencing data, which is often limited for certain cancer types, particularly those induced by ionizing radiation (IR).

Purpose of the Study:

• To evaluate if data from mouse models can supplement scarce human data for cancer mutational signature analysis.
• To define the mutational processes associated with ionizing radiation (IR) in vivo.
• To determine if IR-induced mutational signatures can be distinguished from other mutagenic processes.

Main Methods:

• Pooled sequencing data from mouse models and human subjects exposed to ionizing radiation (IR).
• Inclusion of data from ultraviolet radiation (UV)-induced human skin cancer and urethane-induced mouse cancer for comparative analysis.
• Application of non-negative matrix factorization (NMF) to de-convolute SNVs and identify trinucleotide-based mutational signatures.

Main Results:

• NMF successfully distinguished mutational signatures from three distinct mutagenic agents: ionizing radiation (IR), ultraviolet radiation (UV), and urethane.
• In a pooled analysis of mouse and human data, IR exposure was associated with conserved mutational signatures across both species.
• The study demonstrated that combining data from mouse models and human subjects enhances the identification and characterization of mutational signatures.

Conclusions:

• Pooled analysis of mouse and human sequencing data is a valuable strategy to overcome data scarcity in cancer genomics.
• This approach aids in defining and differentiating mutational signatures, particularly for poorly characterized mutagenic processes like IR-induced mutagenesis.
• The findings highlight the utility of integrating animal model data with human data for robust cancer research.