Abstract
Many Australian native honeys possess significant antimicrobial properties due to the production of hydrogen peroxide (H(2)O(2)) by glucose oxidase, an enzyme derived from the honeybee. The level of H(2)O(2) produced in different honey samples is highly variable, and factors governing its production and stability are not well understood. In this study, highly active Australian honeys that had been stored for >10 years lost up to 54 % of their antibacterial activity, although almost all retained sufficient activity to be considered potentially therapeutically useful. We used a simple colourimetric assay to quantify H(2)O(2) production. Although we found a significant correlation between H(2)O(2) production and antibacterial activity across diverse honey samples, variation in H(2)O(2) only explained 47 % of the variation observed in activity, limiting the assay as a screening tool and highlighting the complexity of the relationship between H(2)O(2) and the killing power of honey. To further examine this, we tested whether H(2)O(2) detection in honey was being inhibited by pigmented compounds and if H(2)O(2) might be directly degraded in some honey samples. We found no correlation between H(2)O(2) detection and honey colour. Some honey samples rapidly lost endogenous and spiked H(2)O(2), suggesting that components in honey, such as catalase or antioxidant polyphenols, may degrade or quench H(2)O(2). Despite this rapid loss of H(2)O(2), these honeys had significant peroxide-based antibacterial activity, indicating a complex relationship between H(2)O(2) and other honey components that may act synergistically to augment activity.