Development of an untargeted DNA adductomics method by ultra-high performance liquid chromatography coupled to high-resolution mass spectrometry

Genotoxicants originating from inflammation, diet, and environment can covalently modify DNA, possibly initiating the process of carcinogenesis. DNA adducts have been known for long, but the old methods allowed to target only a few known DNA adducts at a time, not providing a global picture of the "DNA adductome". DNA adductomics is a new research field, aiming to screen for unknown DNA adducts by high resolution mass spectrometry (HRMS). However, DNA adductomics presents several analytical challenges such as the need for high sensitivity and for the development of effective screening approaches to identify novel DNA adducts. In this work, a sensitive untargeted DNA adductomics method was developed by using ultra-high performance liquid chromatography (UHPLC) coupled via an ESI source to a quadrupole-time of flight mass spectrometric instrumentation. Mobile phases with ammonium bicarbonate gave the best signal enhancement. The MS capillary voltage, cone voltage, and detector voltage had most effect on the response of the DNA adducts. A low adsorption vial was selected for reducing analyte loss. Hybrid surface-coated analytical columns were tested for reducing adsorption of the DNA adducts. The optimized method was applied to analyse DNA adducts in calf thymus, cat colon, and human colon DNA by performing a MS^E acquisition (all-ion fragmentation acquisition) and screening for the loss of deoxyribose and the nucleobase fragment ions. Fifty-four DNA adducts were tentatively identified, hereof 38 never reported before. This is the first untargeted DNA adductomics study on human colon tissue, and one of the few untargeted DNA adductomics studies in the literature reporting the identification of such a high number of unknowns. This demonstrates promising results for the application of this sensitive method in future human studies for investigating novel potential cancer-causing factors.