Holistic Lipidomics of the Human Gut Phenotype Using Validated Ultra-High-Performance Liquid Chromatography Coupled to Hybrid Orbitrap Mass Spectrometry

Anal Chem. 2017 Nov 21;89(22):12502-12510. doi: 10.1021/acs.analchem.7b03606. Epub 2017 Oct 30.

Abstract

As lipids are assigned a plethora of biological functions, it is evident that dysregulated lipid metabolism signifies a key element in many pathological conditions. With this rationale, this study presents a validated lipidomics platform to map the fecal lipidome, which integrates unique information about host-gut microbiome interactions, gastrointestinal functionality, and dietary patterns. This particular method accomplished coverage across all eight lipid categories: fatty acyls, glycerolipids, phosphoglycerolipids, polyketides, prenols, saccharolipids, sphingolipids, and sterols. Generic extraction of freeze-dried feces was achieved by solid-liquid extraction using methanol and methyl tert-butyl ether. Extracted components were separated by liquid chromatography, whereby the selected ethylene-bridged hybrid phenyl ultra-high-performance liquid chromatography stationary phase allowed fast separation of both individual lipid species and categories. Detection was achieved by high-resolution full-scan Q-Exactive Orbitrap mass spectrometry and covered a broad m/z scan range (67-2300 Da). Method validation was performed in a targeted fashion to evaluate the analytical performance across all lipid categories, revealing excellent linearity (R2 ≥ 0.9921), acceptable repeatability (coefficients of variance ≤15.6%), and stable recovery (coefficients of variance ≤11.9%). Method suitability for untargeted fingerprinting was verified, demonstrating adequate linearity (R2 ≥ 0.90) for 75.3% and acceptable repeatability (coefficients of variance ≤30%) for 84.5% of about 9000 endogenous fecal compounds. Eventually, the potential of fecal lipidomics was exemplified within a clinical context of type 2 diabetes, thereby revealing significant perturbations [orthogonal partial least-squares discriminant analysis Q2(Y) of 0.728] in the fecal lipidome between participants with normal blood glucose levels (n = 26) and those with type 2 diabetes (n = 17).

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Chromatography, High Pressure Liquid
  • Feces / chemistry*
  • Humans
  • Lipids / analysis*
  • Lipids / genetics
  • Mass Spectrometry
  • Phenotype

Substances

  • Lipids