Evaluation of different stool extraction methods for metabolomics measurements in human fecal samples

Abstract

AbstractBackgroundMeasurement of metabolomics in human stool samples is of great interest for a broad range of applications in biomedical research including early detection of colorectal neoplasms. However, due to the complexity of metabolites there is no consensus on how to process samples for stool metabolomics measurements to obtain a broad coverage of hydrophilic and hydrophobic substances.MethodsWe used frozen stool samples (50mg) from healthy study participants. Stool samples were processed after thawing using 8 different processing protocols and different solvents. Metabolites were measured afterwards using the MxP® Quant 500 kit (Biocrates). The best performing protocol was subsequently applied to compare stool samples of participants with different dietary habits.ResultsIn this study, we were able to determine up to 340 metabolites of various chemical classes extracted from stool samples of healthy study participants with 8 different protocols. Polar metabolites such as amino acids could be measured with each method while other metabolite classes, particular lipid species, are more dependent on the solvent or combination of solvents used. Only a small number of triglycerides or acylcarnitines were detected in human feces. Extraction efficiency was higher for protocols using isopropanol or those using ethanol or methanol and MTBE including an evaporation and concentration step than for other protocols. We detected significant fecal metabolite differences between vegetarians, semi-vegetarians and non-vegetarians.ConclusionFor the evaluation of metabolites in fecal samples we found protocols using solvents like isopropanol and those using ethanol or methanol and MTBE including an evaporation and concentration step to be superior over others tested in this study.