Carbonyl Compounds in Manuka Honey:: Antibacterial Activity, Reactions and Metabolic Transit

New Zealand is the world’s third-largest honey exporter by value behind China and Argentina and honey accounts for up to 80 % of New Zealand’s exports. However, it is only the 16th biggest global supplier by volume. Manuka honey from New Zealand is sold for premium prices and merchandised for its health benefits. Because of its exceptional antibacterial effect, there is a strong market demand and the price for a kilogram of manuka honey has tripled in recent years (Ministry for Primary Industries 2015). When consumers are willing to pay prices up to 200 €/kg manuka honey, the risk of misleading advertisement and intended fraud increases. This thesis aims to further characterize manuka honey and contribute to the development of a manuka honey definition. The first part deals with the antibacterial activity of manuka honey. The effect of manuka honey is mainly due to methylglyoxal, whereas the effect of non-manuka honeys is primarily caused by hydrogen peroxide. The objective is to develop a method to quantify the effect solely due to one of the respective chemical compounds and compare their effectiveness. Finally, an evaluation of the contribution of methylglyoxal and hydrogen peroxide to the inhibitory effect of honey should be given. The second part deals with chemical reactions of carbonyl compounds in honey. Because of the reactive nature of carbonyl compounds, the formation of specific glycation compounds in honey is assumed. Since the carbonyl profile of manuka honey differs remarkably from non-manuka honeys, the reaction products are expected to vary widely. Specific compounds, solely present in manuka honey, could serve as quality control parameters to ensure manuka honey authenticity. The final part deals with the metabolism of food-derived carbonyl compounds. Carbonyl compounds, like methylglyoxal or 3-deoxyglucosone are discussed to be potentially toxic to human tissues. Until now, only little is known about the impact of the diet on the physiological carbonyl-load and the metabolism of carbonyl compounds. With the help of nutrition studies and the analysis of body fluids, the question of metabolic transit of carbonyl compounds shall be addressed. The antibacterial studies showed that bacterial species are affected differently by bioactive compounds present in honey. Methylglyoxal (MGO), which is solely present in manuka honeys and hydrogen peroxide, which is formed in most conventional honeys by glucose oxidase, are strong inhibitors of the growth of S. aureus and E. coli. The strain of P. aeruginosa used for this work was not inhibited by MGO, whereas B. subtilis was not inhibited by hydrogen peroxide. To compare and quantify the effect of MGO and hydrogen peroxide, a mathematic model was created. By comparing the slopes of the linearized dose-response curves, it was found that S. aureus, E. coli and P. aeruginosa were more sensitive to hydrogen peroxide than to MGO. However, the natural amounts of MGO in honey are higher than the formation of hydrogen peroxide. Although most bacteria are more sensitive to hydrogen peroxide, MGO is the predominantly antibacterial compound in honey, because of its higher concentrations compared to hydrogen peroxide formation. The inclusion of manuka honey in α-cyclodextrin had only minor consequences on bioavailability and antibacterial activity. The commercial product “Cyclopower” (α-cyclodextrin with manuka honey) does not enhance the antibacterial activity of manuka honey on S. aureus, E. coli and P. aeruginosa. With the help of the newly developed quantitative model, it was shown that the growth of B. subtilis is synergistically inhibited with cyclopower compared to manuka honey and α-cyclodextrin alone. The study of bacterial enzymes as possible targets for bacterial inhibition with manuka honey revealed that MGO and DHA inhibited jack bean urease, which was used as a model for Helicobacter pylori urease. The concentration of MGO and DHA in manuka honey positively correlated with its urease inhibition. Conventional honeys, which lack MGO and DHA, showed significantly less urease inhibition. Based on the unique presence of MGO, manuka honey has extraordinary effects on bacteria, which might lead to further application to fight the emerging crisis of antibacterial resistance to antibiotics. Until now, there is no consistent definition for the term “genuine manuka honey”. In the present work, an approach based on unique chemical reactions in manuka honey was followed. It was shown that the exceptional high amounts of MGO induced the formation of 2-acetyl-1-pyrroline (2-AP). In manuka honey containing ≥ 250 mg/kg MGO, the 2-AP concentration was significantly increased compared to conventional honey. Moreover, honey proteins form MGO-derived reactions products, which were studied by measuring the molecular size of honey proteins. Manuka honey proteins significantly shifted to high molecular weights (HMW) with a size above 510 kDa. The amount of HMW protein in non-manuka honey was significantly lower. The cleavage of disulphide bonds led to a decrease of HMW fraction of conventional honeys but not of manuka honeys. It is hypothesized that MGO cross-linking of proteins is mainly responsible for the formation of HMW adducts in manuka honey. The formation of HMW adducts was also shown with fluorescence analysis, whereby manuka honey proteins had higher fluorescence intensities at λex=350 nm and λem=450 nm compared to non-manuka honeys. The artificial addition of MGO and its precursor dihydroxyacetone (DHA) to a non-manuka honey did not lead to an increased fluorescence up to the level of commercial manuka honeys. The MGO-derived modifications of proteins were further studied by quantifying the protein-bound Maillard reaction products N-ε-carboxyethyllysine (CEL) and methylglyoxal-derived hydroimidazolone 1 (MG-H1) after enzymatic hydrolysis of honey proteins and LC-MS/MS analysis. Their amount was significantly higher in manuka compared to conventional honeys and correlated with the MGO content of the honey. Most of the MGO-derived reactions could be simulated by spiking a conventional honey or a low MGO manuka honey with artificial MGO and subsequent storage at elevated temperatures. Higher storage temperatures were associated with a quick increase of 5-hydroxymethylfurfuraldehyd (HMF). The HMF level in honey is used as a quality parameter and should not exceed 40 mg/kg (Codex Alimentarius Commission, 2001). High concentrations of HMF may point to a fraudulent addition of MGO and the production of artificial high-price manuka honey products. Taken together, the Maillard reaction in honey could be used to control the natural origin of MGO and DHA. The consumption of honey and especially manuka honey exposes humans to high levels of dietary dicarbonyl compounds like MGO and 3-deoxyglucosone (3-DG). Both compounds were discussed as potential risk factors for the development of age-related diseases. The simulated digestion of manuka honey in the presence of gastric and ileal fluids showed that only 9 % of the initial concentration can be recovered after 8 h. The honey matrix had no stabilising effect on MGO compared to a synthetic MGO solution. In contrast to MGO, the manuka honey compound DHA was stable during all simulated digestion steps. The complexation of MGO with α-cyclodextrin did not enhance the stability of MGO. The metabolic transit of dietary MGO and 3-DG was further studied with an intervention study with healthy volunteers, who collected their daily urine. It was shown that urinary concentrations of 3-DG and its less reactive metabolites 3-deoxyfructose (3-DF) and 2-keto-3-deoxygluconic acid (3-DGA), but not MGO, were influenced by the diet. During the intervention studies, up to 40 % of dietary 3-DG was recovered as the sum of 3-DG, 3-DF and 3-DGA. The metabolite 3-DGA only played a minor role in the metabolism of dietary 3-DG in comparison to 3-DF. The concentrations 3-DF and 3-DGA in plasma only increased after the consumption of dietary 3-DG and not after the uptake of carbohydrate rich meals in general. This led to the conclusion that dietary 3-DG is effectively metabolized to 3-DF extracellularly on the apical site of the intestinal epithelium and is resorbed slowly into the circulation. In contrast, 3-DG, which is formed (intracellularly) postprandial from glucose, bypasses this metabolic system and cannot be metabolized as rapidly to 3-DF. Preliminary results obtained with saliva instead of urine as a bio fluid to study the dietary influence of dicarbonyl compounds, confirmed the hypothesis. Based on the present results, dietary dicarbonyl compounds are effectively metabolized during digestion