Mathematical model-assisted HPLC-MS/MS analysis on global, pseudo-targeted ceramide profiling and quantitation in serum.

Ceramides (Cers) play a crucial role in sphingolipid metabolism with multiple biological activities and functions. Due to the high regularity and variability of their structures, there exist thousands of possible Cers. The structural diversity endows them with various biological functions but also poses significant challenges for qualitative and quantitative analysis. The lack of in-depth characterization methods for such lipids resulted in only a small fraction of Cers being reported, severely hindering the exploration of their biological functions and activities. This work presented a lipid analysis method based on a liquid chromatography-mass spectrometry platform, enabling the accurate quantification of 337 Cers simultaneously. Supported by a mathematical model, this work succeeded in generating a quadratic equation relationship between retention time and Cers carbon number. Subsequently, this method was applied to the large-scale quantitative detection of Cers in serum samples from Alzheimer's disease (AD) patients, identifying and characterizing 62 differential Cers. These could potentially serve as serum biomarkers for AD diagnosis. This study demonstrates a strategy for the large-scale in-depth characterization of complex endogenous lipid molecules with highly variable and regular structures in the absence of sufficient commercial standard materials. This work provides a novel analysis method and reference for exploring and developing the functions of such endogenous bioactive molecules.