Software-Assisted LC-HRMS Annotation of Urinary Modified Nucleosides: Proof-of-Concept in Prostate Cancer

Abstract High-resolution mass spectrometry (HRMS) has greatly expanded the untargeted annotation of RNA modifications and the investigation of epitranscriptomic profiles at the single-nucleoside level. However, the comprehensive characterization of endogenous modified nucleosides in biological matrices remains challenging due to their low abundance, structural diversity, and isomerism, despite their emerging role as biomarkers of RNA turnover, cellular stress, and disease-related metabolic alterations. We developed an integrated LC-HRMS suspect screening workflow for the structured annotation of modified nucleosides in clinically relevant samples. The approach combines a curated in-house mass list of biologically relevant nucleoside modifications with a dedicated data-processing pipeline, enabling efficient feature prioritization and the annotation of 60 nucleoside-related compounds with different levels of identification confidence. Where authentic analytical standards were available, selected annotations were further confirmed by agreement in accurate mass, retention time, and MS/MS fragmentation. Chromatographic conditions were optimized to resolve isomeric species, while stepped normalized collision energy acquisition provided complementary fragmentation data to support structural annotation. The workflow was further applied to a small pilot set of urine samples from patients with prostate cancer (PCa) and benign prostatic hyperplasia (BPH) to illustrate its applicability in a clinically relevant matrix. Exploratory statistical analysis was used only to visualize preliminary abundance trends between the two groups. The results highlight the ability of the workflow to prioritize modified-nucleoside features for further investigation while requiring minimal sample preparation. Although confirmation of the remaining low-abundance and isomeric annotations and assessment of their biological relevance will require further validation, and larger independent cohorts, the proposed workflow provides a practical framework for systematic modified-nucleoside annotation and hypothesis generation in future metabolomics studies.