Simultaneous quantification of endogenous biomarkers and detection of testosterone esters in human serum using gas chromatography-high-resolution mass spectrometry.

Abstract

High-resolution mass spectrometry (HRMS) has been demonstrated to be an alternative platform for quantitative analyses, identifying unknown compounds and gathering information for the elucidation of chemical structures. This work describes a method to detect 13 esters of testosterone (T) and 5 biomarkers in 0.1mL of human serum using gas chromatography (GC) coupled to HRMS. Analytes were extracted from serum after deproteinization and liquid-liquid extraction. The trimethylsilyl derivatives were analyzed using a gas chromatograph coupled to HRMS at low electron energy to minimize molecule fragmentation. The acquisition in profiling full-scan mode was applied with a resolving power of 30 000 at m/z 400. Linearity, lower limit of quantitation, and measurement uncertainty were assessed. Precision and accuracy were assessed at 0.5 and 2ng/mL, respectively. Mass accuracy (MA) and mass extraction window (MEW) were also evaluated. T esters showed a linear response between 0.25 and 10ng/mL (except for undecanoate, enanthate, and propionate that showed lineal responses between 0.5 and 10ng/mL and isocaproate between 2 and 10ng/mL); detection limits remained between 0.1 and 0.5ng/mL and accuracy between 81% and 119%. The MA (MEW = 10ppm) was maintained between -2.4 and 4.8ppm. The biomarkers (T, androstenedione, dehydroepiandrosterone [DHEA], estradiol, and 17-OH-progesterone) showed a linear response within the evaluated range; quantification limits remained between 0.1 and 0.5ng/mL (except for DHEA), the accuracy between 88% and 99%, and precision between 3.5% and 10.8%. Measurement uncertainties were found between 5.6% and 17.2%. MA (MEW = 3ppm) was maintained between -0.47 and 0.12 ppm. The method to detect T esters and five endogenous biomarkers in serum using GC coupled to HRMS showed linear responses up to 10ng/mL with adequate precision, accuracy, and uncertainties. It was possible to distinguish cholesterol from T-isocaproate based on the MEW of 10ppm, preventing false positives. In addition, this method allows searching for other biomarkers and/or unknown metabolites and other ester forms not included here but at a later stage if necessary.