Bias caused by incomplete metabolite extraction and matrix effect: Evaluation of critical factors for plasma sample preparation prior to metabolomics
Journal
Journal of Pharmaceutical and Biomedical Analysis
Journal Volume
219
Pages
114930
Date Issued
2022-09-20
Author(s)
Abstract
Metabolomics is an omics strategy to study the metabolite alteration in the biological system. Unbiased observation of the metabolite level is essential for targeted metabolite quantification and untargeted metabolic profiling. State-of-the-art instruments and versatile tools have been developed for accurate observation of metabolic alterations in various studies. Several analytical pitfalls, such as sample overloading and signal-saturation-induced bias, have been revealed and addressed. In this study, we proposed incomplete-metabolite-extraction-caused bias is also an important issue that results in biased observation when performing metabolomics. In the demonstration example, numerous metabolites exhibited no significant difference between extracted plasma samples with different plasma contents, which is attributed to incomplete-metabolite-extraction-caused bias and matrix effect. Matrix effect is a well-known factor that result in biased observation, it can be reduced by sample dilution and compensated by using stable isotope labelled internal standards. The detection of metabolite signals in the following consecutive extractions provided further evidence of incomplete metabolite extraction. The completeness of metabolite extraction is crucial for unbiased observation of metabolic profile changes. To address this issue, we optimized the extraction time and methanol volume to reduce the incomplete-metabolite-extraction-caused bias and evaluated the metabolite signals in consecutive extractions. Methanol extraction performed with a plasma-to-methanol ratio of 1:14 resulted in metabolite responses of less than 18.1 % in the second extractions observed by metabolomic profiling. Finally, the optimized sample preparation procedure and untargeted profiling platform were applied to detect metabolite alterations associated with patients with cerebrovascular diseases and several features with significant difference were successfully identified. This study revealed and evaluated the bias caused by incomplete metabolite extraction and matrix effect in the commonly used methanol extraction method for human plasma sample preparation for metabolomics. We anticipate the proposed metabolite extraction evaluation method could benefit more clinical and biological metabolomics studies.
SDGs
Type
journal article