LAUSR

Background: Bile’s potential to reflect the health of the biliary system has led to increased attention, with proteomic analysis offering deeper understanding of biliary diseases and potential biomarkers. With the emergence of normothermic machine perfusion (NMP), bile can be easily collected and analyzed. However, the composition of bile can make the application of proteomics challenging. This study systematically evaluated various trypsin digestion methods to optimize proteomics of bile from human NMP livers. Methods: Bile was collected from 12 human donor livers that were accepted for transplantation after the NMP viability assessment. We performed tryptic digestion using six different methods: in-gel, in-solution, S-Trap, SMART, EasyPep, and filter-aided sample purification, with or without additional precipitation before digestion. Proteins were analyzed using untargeted proteomics. Methods were assessed for total protein IDs, variation, and protein characteristics to determine the most optimal method. Results: Methods involving precipitation surpassed crude methods in protein identifications (4500 vs 3815) except for in-gel digestion. Filtered data (40%) resulted in 3192 versus 2469 for precipitated and crude methods, respectively. We found minimal differences in mass, cellular components, or hydrophobicity of proteins between methods. Intermethod variability was notably diverse, with in-gel, in-solution, and EasyPep outperforming others. Age-related biological comparisons revealed upregulation of metabolic-related processes in younger donors and immune response and cell cycle-related processes in older donors. Conclusions: Variability between methods emphasizes the importance of cross-validation across multiple analytical approaches to ensure robust analysis. We recommend the in-gel crude method for its simplicity and efficiency, avoiding additional precipitation steps. Sample processing speed, cost, cleanliness, and reproducibility should be considered when a digestion method is selected for bile proteomics.