Optimization of Extraction Methods for NMR and LC-MS Metabolite Fingerprint Profiling of Botanical Ingredients in Food and Natural Health Products (NHPs)

Metabolite fingerprint profiling is a robust tool for verifying suppliers of authentic botanical ingredients. While comparative studies exist, few apply identical conditions across multiple species; this study utilized a cross-species comparison to identify versatile solvents despite biochemical variability. Multiple solvents were used for sample extraction prior to analysis by proton NMR and liquid chromatography-mass spectrometry (LC-MS) for multiple botanicals including <i>Camellia sinensis</i>, <i>Cannabis sativa</i>, <i>Myrciaria dubia</i>, <i>Sambucus nigra</i>, <i>Zingiber officinale</i>, <i>Curcuma longa</i>, <i>Silybum marianum</i>, <i>Vaccinium macrocarpon</i>, and <i>Prunus cerasus</i>. Comparisons were normalized by total intensity; deuterated methanol aids NMR lock but is not required for LC-MS. Hierarchical clustering analysis (HCA) evaluated solvent efficacy. Methanol-deuterium oxide (1:1) was the most effective extraction method, yielding 155 NMR spectral metabolite variables for <i>Camellia sinensis</i>, while methanol (90% CH₃OH + 10% CD₃OD) produced 198 for <i>Cannabis sativa</i> and 167 for <i>Myrciaria dubia</i>, with 11, 9, and 28 assigned metabolites, respectively. LC-MS detected 121 metabolites in <i>Myrciaria dubia</i> in methanol as the most effective extraction method. Methanol (10% deuterated) is the most effective extraction method for comprehensive metabolite fingerprinting using NMR and LC-MS protocols because it provides the broadest metabolite coverage. This study advances fit-for-purpose methods to qualify suppliers of botanical ingredients in food and NHP quality control programs.