An integrated platform for directly widely-targeted quantitative analysis of feces part II: An application for steroids, eicosanoids, and porphyrins profiling

• An online stepped optimization strategy was proposed to optimize parameters.
• 71 analytes were detected, identified, and quantitatively analyzed in feces.
• Significant differences were observed among feces from four groups.
• A feasible solution for metabolomics of solid-state biological samples was developed.

Steroids, especially bile acids, along with eicosanoids and porphyrins in feces play pivotal roles for the clinical diagnosis of various diseases. However, their reliable measurement is extensively obstructed by poor stability, structural diversity, broad content ranges, and tedious sample preparation protocols that account for a majority of the measurement errors. In current study, in-depth component screening was initially carried out by flexibly integrating diverse modes, such as predefined multiple reaction monitoring, stepped multiple ion monitoring, neutral loss scan, and precursor ion scan on a hybrid triple quadrupole-linear ion trap mass spectrometer, which also provided MS2 spectra via enhanced product ion experiments. Meanwhile, a hybrid ion trap-time of flight mass spectrometer served as a complementary tool by providing accurate mass spectral information. Afterwards, because authentic compounds were unavailable for most analytes, an online optimization strategy was then proposed to optimize parameters, including precursor-to-product ion transitions and spectrometric parameters, notably collision energy. Finally, direct analysis of all detected components in feces was carried out by employing a platform integrating online pressurized liquid extraction, turbulent flow chromatography, and LC–MS/MS, and applying those optimized parameters. Seventy-one compounds, including 52 steroids and 13 eicosanoids, together with 6 porphyrins, were found and annotated in a fecal pool, and then relatively quantified in various fecal matrices. The quantitative dataset was subjected for multivariate statistical analysis and significant differences were observed among the quantitative chemome profiles of the fecal matrices from different groups. The findings obtained in the two parts demonstrated that the analytical platform in combination with the work-flow is qualified for not only directly simultaneous measurement of diverse endogenous substances, but widely targeted metabolomics of fecal matrices.