Abstract
Artifacts due to metabolite extraction, derivatization, and detection techniques can result in aberrant observations that are not accurate representations of actual cell metabolism. Here, we show that α-ketoglutarate (α-KG) is reductively aminated to glutamate in methanol:water metabolite extracts, which introduces an artifact into metabolomics studies. We also identify pyridoxamine and urea as amine donors for α-KG to produce glutamate in methanol:water buffer in vitro, and we demonstrate that the addition of ninhydrin to the methanol:water buffer suppresses the reductive amination of α-KG to glutamate in vitro and in metabolite extracts. Finally, we calculate that glutamate levels have been overestimated by 10-50%, depending on cell line, due to α-KG reductive amination. These findings suggest that precautions to account for α-KG reductive amination should be taken for the accurate quantification of glutamate in metabolomics studies.