List of identified sterols, bile acids, and acylcarnitines

Version 1

Objective: The aim of this study was to present and validate a sensitive and simple liquid chromatography–tandem mass spectrometry (LC–MS/MS) method to detect the presence and distribution of NAFLD-related metabolites (sterols, bile acids, and acylcarnitines) in biological samples. These compound classes were considered as they are implicated in the development and progression of NAFLD; their detection by MS-based analytics is challenging due to the occurrence of many isomers and stereoisomers; in-source fragmentation patterns, and their relatively low concentrations in biological samples.
Method: The approach was based on concurrently analyzing the “Bile acid, Carnitine, and Sterol Library of Standards” (“BACSMLS”, IROA technologies) along with the biological samples in the same analytical run. Liquid chromatography was performed on a 1290 Infinity Binary UPLC (Agilent Technologies). Reversed-phase separation was employed using a Zorbax Eclipse XDB-C18 column (dimensions 2.1 × 100 mm, particle size 1.8 µm) as the stationary phase. The mobile phase consisted of ultra-pure water (solution A) and methanol (solution B), both containing 0.1% v/v formic acid. The flowrate of the eluent was 0.6 mL/min. The elution gradient profile was as follows (t [min], %B): (0, 2), (10, 100), (14.5, 100), (14.51, 2), (16.5, 2). A measure of 1 µL of each standard or sample was injected for analysis.
Mass spectrometry was performed on an Agilent 6540 Q-TOF with a Jet Stream ESI ion source. The fragmentor voltage used was 100 eV and scan range 20–1600 m/z. From every precursor scan cycle (400 milliseconds), four most abundant precursor ions were automatically selected for MS/MS fragmentation, excluded after two acquired MS/MS spectra, and released again from the exclusion list after 0.25 min. The collision energies used for the MS/MS analysis were (±)10, 20, and 40 eV, for compatibility with online databases such as METLIN and HMDB database. Both negative and positive ionization modes were used in the study.
After collection of the raw instrumental data, MS-DIAL (Version 4.18) was employed for automated peak picking and alignment from the raw data. The minimum peak height was set to 2000, MS/MS tolerance was 0.05, and RT tolerance was set to 0.1 min. Default settings were used for the other parameters. After peak picking, the alignment result as peak areas were exported into Microsoft Excel. The exported excel file contained a data matrix comprising the LC–MS characteristics along with the peak areas of the different standard compounds and the biological samples. The raw spectral data from the pure compound standards and biological samples were manually examined and the unique peaks corresponding to each of the individual standard compounds were located, followed by annotation in the data matrix. Thereafter, the spectral base peak corresponding to each standard was annotated as the representative signal for that standard. Moreover, the typical adduct and neutral loss peaks were annotated including [M+H]+, [M+H–H2O]+, [M+H–2H2O]+, [2M+H]+, [M+NH4]+, and [M+Na]+ for the positive ionization mode and [M–H]−, [M–H–H2O]−, [2M–H]−, and [M+FA–H]− for the negative ionization mode. Further, MS/MS data from METLIN database and HMDB were used as additional examination for the fragmentation of some of the standards. The vendor software—Agilent MassHunter Qualitative Analysis B.07.00—was used to explore raw data extracted ion chromatograms (EICs) and MS/MS fragmentation spectra of the pure standards and biological samples. The limit of detection of the metabolites was set by following two criteria: signal-to-noise ratio >5 and raw peak area >10,000. All those signals with a peak area <10,000 were denoted as traces.