Comprehensive targeted profiling of multiple steroid classes in rodent plasma using liquid chromatography-mass spectrometry

Background

Reliable quantification of multiple steroid classes in biological fluids within a single method remains an analytical challenge despite many previously published methods. Crosstalk of positional isomers, overlap of stereoisomer fragmentation patterns, differing proton affinities, in-source fragmentation, varying stability of protonated ions in the gas phase across steroid classes, and non-existence of steroid-free matrix are the main challenges limiting the number of simultaneously profiled steroids.

Results

In this study, we focused on the development of a derivatization-free, achiral, high-throughput, and cost-effective UHPLC-MS/MS approach that allows simultaneous profiling of a spectrum of 38 steroids covering progestogens, androgens, corticosteroids, and estrogens, while properly addressing the hurdles of steroid analysis. Within a 20-min method, 16 stereoisomers and 15 positional isomers were fully resolved within a single run while separated from 7 additional non-interfering steroids and matrix interferences in rodent plasma. Protein precipitation (PP) and supported liquid extraction (SLE) methods using only 40 μL of sample were developed to achieve the lowest possible limits of quantification. Nevertheless, 5α-dihydroprogesterone and 3α,5α-THDOC could be only qualitatively assessed when using PP. In contrast, DHEA-S could not be quantified or identified when using SLE. A novel surrogate matrix-background subtraction approach, using rat plasma after the animal's adrenalectomy, has been implemented into the optimized PP-UHPLC-MS/MS workflow, successfully validated according to the unified ICH/EMA M10 guidelines, and compared to the traditional quantification strategies. Moreover, the validity of the newly adopted approach has been verified by the targeted profiling of multiple biologically active endogenous steroids in more than 500 samples of mouse plasma in total.

Significance

Underestimation of hurdles associated with steroid analysis often compromises the accurate steroid quantification. Our comprehensive, fully validated UHPLC-MS/MS method targeting a wide spectrum of endogenous steroids, mitigating steroid crosstalk and using a minimal sample volume together with a novel surrogate matrix-background subtraction approach significantly advances steroid analysis for research and clinical applications covering multiple biological scopes.