Correlation of the antibacterial activity of commercial manuka and Leptospermum honeys from Australia and New Zealand with methylglyoxal content and other physicochemical characteristics.

Variation in the antibacterial potency of manuka honey has been reported in several published studies. However, many of these studies examine only a few honey samples, or test activity against only a few bacterial isolates. To address this deficit, a collection of 29 manuka/Leptospermum honeys was obtained, comprising commercial manuka honeys from Australia and New Zealand and several Western Australian Leptospermum honeys obtained directly from beekeepers. The antibacterial activity of honeys was quantified using several methods, including the broth microdilution method to determine minimum inhibitory concentrations (MICs) against four species of test bacteria, the phenol equivalence method, determination of antibacterial activity values from optical density, and time kill assays. Several physicochemical parameters or components were also quantified, including methylglyoxal (MGO), dihydroxyacetone (DHA), hydroxymethylfurfural (HMF) and total phenolics content as well as pH, colour and refractive index. Total antioxidant activity was also determined using the DPPH* (2,2-diphenyl-1-picrylhydrazyl) and FRAP (ferric reducing-antioxidant power) assays. Levels of MGO quantified in each honey were compared to the levels stated on the product labels, which revealed mostly minor differences. Antibacterial activity studies showed that MICs varied between different honey samples and between bacterial species. Correlation of the MGO content of honey with antibacterial activity showed differing relationships for each test organism, with Pseudomonas aeruginosa showing no relationship, Staphylococcus aureus showing a moderate relationship and both Enterococcus faecalis and Escherichia coli showing strong positive correlations. The association between MGO content and antibacterial activity was further investigated by adding known concentrations of MGO to a multifloral honey and quantifying activity, and by also conducting checkerboard assays. These investigations showed that interactions were largely additive in nature, and that synergistic interactions between MGO and the honey matrix did not occur.