The influence of phase II conjugation on the biological activity of flavonoids

Flavonoid consumption is often correlated with a wide range of health effects, such as the prevention of cardiovascular diseases, neurodegenerative diseases, and diabetes. These effects are usually ascribed to the activity of the parent flavonoid aglycones, even though these forms of the flavonoids generally have a low systemic bioavailability. During uptake, flavonoids undergo phase II metabolism and are present in the systemic circulation nearly exclusively as conjugated metabolites. The aim of this thesis was to study the effect of conjugation on the biological activity of selected flavonoids towards different endpoints relevant for human health. To this end, conjugation with glucuronic acid was taken as the model type of conjugation because this modification is generally observed to be the most important metabolic conjugation reaction for flavonoids in man. A review of scientific literature published until early 2012 reveals that metabolic conjugation can affect the biological activity of flavonoids in different ways. Conjugation can increase, decrease, inverse or not affect the biological activity, depending on the flavonoid, the type and position of conjugation, the endpoint studied, and the assay system used. Based on the literature reviewed it is concluded that the effect of conjugation has to be studied on a case-by-case basis. As the research on the biological activity of biologically relevant flavonoid conjugates is often hampered by the generally low commercial availability and high prices of these conjugates, a simple and versatile method for the biosynthesis of metabolically relevant flavonoid conjugates is described. Using this method, relevant conjugates can be prepared from different flavonoid substrates in sufficient quantities for in vitro bioassays. Further, an efficient strategy for the identification of these flavonoid conjugates by LC-MS and 1H-NMR using MetIDB (Metabolite Identification Database), a publicly accessible database of predicted and experimental 1H-NMR spectra of flavonoids, is presented. To study the effect of conjugation on the biological activities of flavonoids, several different assay systems and endpoints were used to study the activity of different flavonoids and their conjugates. The effects of quercetin, kaempferol, and their main plasma conjugates quercetin-3-O-glucuronide and kaempferol-3-O-glucuronide (K-3G) on different endpoints related to peroxisome proliferator-activated receptor (PPAR)-γ were studied. PPAR-γ activation is reported to have positive health effects related to adipogenesis, insulin resistance and inflammation. The presented results show that the flavonoid aglycones increased PPAR-γ mediated gene expression in a stably transfected reporter gene cell line, and that glucuronidation diminished this effect. These observed increases in reporter gene expression were accompanied by increased PPAR-γ receptor-mRNA expression upon exposure to kaempferol, an effect that was also reduced by glucuronidation. Using the cell-free Microarray Assay for Real-time Coregulator-Nuclear receptor Interaction (MARCoNI) it was demonstrated that, unlike the known PPAR-γ agonist rosiglitazone, neither the flavonoid aglycones nor the conjugates are agonistic ligands of the PPAR-γ receptor. Supporting the hypothesis that the tested compounds have a different mode of action from normal LBD agonism, quercetin appeared to synergistically increase the effect of rosiglitazone in the reporter gene assay. The modes of action behind the observed effects remain to be elucidated and might include effects on protein kinase activities affecting expression of the PPAR-γ receptor, or posttranscriptional modifications of PPAR-γ. Another type of nuclear receptor known to be targeted by certain flavonoids are the estrogen receptor (ER)α- and ERβ. ERs are the main targets of estrogenic compounds, and upon their activation different transcriptional responses with opposite effects on cell proliferation and apoptosis are elicited; ERα activation stimulates cell proliferation, while ERβ activation causes apoptosis and reduces ERα mediated induction of cell proliferation. Using the MARCoNI assay, the intrinsic estrogenic effects of the two main dietary isoflavones daidzein and genistein, and their plasma conjugates daidzein-7-O-glucuronide and genistein-7-O-glucuronide on the ligand induced coregulator binding of ERα- and ERβ-LBD were studied and compared to the effect of the positive control 17β-estradiol (E2). The results show that the tested isoflavone compounds are less potent agonists of ERα- and ERβ-LBD than E2, although they modulate the LBD-coregulator interactions in a manner similar to E2. Genistein is shown to be a more potent agonist than daidzein for both receptor subtypes. While in the MARCoNI assay genistein had a strong preference for ERβ-LBD activation over ERα-LBD activation, daidzein had a slight preference for ERα-LBD activation over ERβ-LBD activation. Glucuronidation reduced the intrinsic agonistic activities of both daidzein and genistein to induce ERα-LBD and ERβ-LBD - coregulator interactions and increased their average half maximal effective concentrations (EC50s) by 8 to 4,400 times. The results presented further show that glucuronidation changed the preferential activation of genistein from ERβ-LBD to ERα-LBD and increased the preferential activation of daidzein for ERα-LBD; this is of special interest given that ERβ activation, which is counteracting the possible adverse effects of ERα activation, is considered one of the supposedly beneficial modes of action of isoflavones. Many flavonoids are reported to be inhibitors of protein kinases. To study the effect of conjugation on the inhibition of serine/threonine protein kinases by flavonoids, kaempferol and its main plasma conjugate K-3G were selected as model compounds. Protein kinases are involved in a wide range of physiological processes by controlling signaling cascades and regulating protein functions; modulation of their activities can have a wide range of biological effects. The inhibitory effects of kaempferol, K-3G, and the broad-specificity protein kinase inhibitor staurosporine on the phosphorylation activity of recombinant protein kinase A (PKA) and of a lysate prepared from the hepatocellular carcinoma cell line HepG2 were studied using a microarray platform that determines the phosphorylation of 141 putative serine/threonine phosphorylation sites derived from human proteins. The results reveal that glucuronidation reduces the intrinsic potency of kaempferol to inhibit the phosphorylation activity of PKA and HepG2 lysate on average about 16 and 3.5 times, respectively. It is shown that the inhibitory activity of K-3G in the experiments conducted was not caused by deconjugation to the aglycone. Furthermore, the results show that kaempferol and K-3G, unlike the broad-specificity protein kinase inhibitor staurosporine, did not appear to inhibit all protein kinases present in the HepG2 lysate to a similar extent, indicating that kaempferol selectively targets protein kinases, a characteristic that appeared not to be affected by glucuronidation. The fact that K-3G appeared to be only a few times less potent than kaempferol implies that K-3G does not necessarily need to be deconjugated to the aglycone to exert potential inhibitory effects on protein kinases. The results obtained in the present thesis support the conclusion that glucuronidation of flavonoids does not necessarily abolish their activity and that flavonoid glucuronides may be biologically active themselves, albeit at higher concentrations than the parent aglycones. In line with the conclusions from the earlier literature review, an updated literature review on the effect of conjugation on the biological activity of flavonoids concludes that that the effect of conjugation on the biological activity of flavonoids depends on the type and position of conjugation, the endpoint studied and the assay system used. Based on the results described and the literature reviewed in this thesis, several recommendations and perspectives for future research are formulated. Several methodological considerations are formulated that need to be taken into account when studying the biological activity of flavonoids and their conjugates to avoid confounding results. Further, the relevance of the gut microbiome for flavonoid bioactivity is highlighted, and considerations regarding the pharmacokinetics and pharmacodynamics of flavonoids in vivo are formulated. Altogether, it can be concluded that circulating flavonoid conjugates may exert biological activities themselves, and that understanding these is a prerequisite to successfully elucidate the mechanisms of action behind the biological activities linked to flavonoid consumption.