Unravelling the Antibacterial Activity of Root Bark through a Metabolomic Approach.

Burch. ex. DC. (Combretaceae) is a popular remedy for the treatment of infectious diseases. It is widely prescribed by traditional healers and sold at informal markets and may be a good candidate for commercialisation. For this to be realised, a thorough phytochemical and bioactivity profile is required to identify constituents that may be associated with the antibacterial activity and hence the quality of raw materials and consumer products. The aim of this study was to explore the phytochemistry and identify the antibacterial constituents of root bark, using a metabolomic approach. The chemical profiles and antibacterial activities of 42 root bark samples collected from three districts in the Limpopo Province, South Africa, were evaluated. Dichloromethane:methanol (1:1) extracts were analysed using ultraperformance liquid chromatography (UPLC)-mass spectrometry (MS), and chemometric models were constructed from the aligned data. The extracts were tested against (ATCC 11778), (ATCC 12223), (ATCC 25923), (ATCC 8739), (ATCC 13883), (ATCC 27853), (ATCC 9292) and (ATCC 14028), using the minimum inhibition microdilution assay. Nine compounds; sericic acid, sericoside, resveratrol-3---rutinoside, ellagic acid, flavogallonic acid dilactone, methyl-flavogallonate, quercetin-3-(2''-galloylrhamnoside), resveratrol-3-(6''-galloyl)---d-glucopyranoside and arjunetin, were isolated from the root bark. All the compounds, with the exception of sericic acid, sericoside and resveratrol-3---rutinoside, were isolated for the first time from the root bark of Chemometric analysis revealed clustering that was not population specific, and the presence of three groupings within the samples, characterised by sericic acid, sericoside and an unidentified compound ( 682/4.66 min), respectively. The crude extracts from different populations displayed varied antibacterial activities against (minimum inhibitory concentrations (MICs) 0.25-1.0 mg/mL), but similar activity towards (1.0 mg/mL). Several compounds present in the root bark were highly active towards all or most of the pathogens tested, but this activity was not reflected by the chemical profiles of extracts prepared from the individual samples. Among the pure compounds tested, only flavogallonic acid dilactone and methyl-flavogallonate exhibited broad-spectrum activity. A biochemometric analysis indicated that there was no consistent association between the levels of phytochemicals and the activity of the active or non-active extracts. Although it was deduced that the major constituents of root bark contributed to the chemotypic variation, further investigation of the interactions of compounds present in the root bark may provide antibacterial efficacies not evident when examining compounds singularly. The data reported herein will provide information that is fundamentally important for the development of quality control protocols.