High dietary consumption of anthocyanins has been associated with reduced risk of cardiovascular disease (CVD) and improved vascular function (Jennings, Welch et al. 2012, McCullough, Peterson et al. 2012, Cassidy, Mukamal et al. 2013). The beneficial effects of anthocyanins have also been supported by randomised controlled trials, for example, in a trial conducted by Zhu et al(Zhu, Xia et al. 2011, Zhu, Ling et al.), hypercholesterolaemic volunteers ingested anthocyanins (320 mg/day) daily in short term (4 hrs) and long term (24 weeks) interventions, which resulted in improved flow-mediated dilatation and lipid profile in peripheral blood serum. In addition, randomised controlled trials investigating anti- inflammatory activities of bilberry anthocyanins reported reduced inflammatory markers [interleukin-6 (IL-6)] in subjects with metabolic syndrome and at elevated risk of CVD (Karlsen, Paur et al. 2010, Kolehmainen, Mykkänen et al. 2012). The bioactivity of anthocyanins has also been studied extensively in vitro to understand underlying mechanisms of action, and anthocyanins have been shown to up-regulate endothelial nitric oxide synthase (eNOS) (Xu, Ikeda et al. 2004a, Xu, Ikeda et al. 2004b) and attenuate expression of key inflammatory markers such as IL-6 and vascular cell adhesion molecule-1 (VCAM-1) in HUVECs(Xia, Ling et al. 2007, Xia, Ling et al. 2009). Though anthocyanins have been reported to possess bioactivity, their poor bioavailability and instability at physiological pH suggest that their reported bioactivity may come, at least in part, from their degradants and subsequent metabolites (Del Rio, Borges et al. 2010, Williamson and Clifford 2010). In fact, recently published human bioavailability studies report that degradants and metabolites of anthocyanins are present in much greater concentrations than their parent structures in the circulation (Azzini, Vitaglione et al. 2010, Czank, Cassidy et al. 2013). However, the bioactivity of these phenolic metabolites of anthocyanins remains relatively unknown. The current thesis sought to address this discrepancy in scientific literature by investigating recently identified metabolites of cyanidin-3-glucoside (C3G) for their in vitro vascular and anti-inflammatory activity. Moreover, the activity of metabolites was
investigated at physiologically relevant concentrations, namely 0.1, 1 and10 µM(Czank, Cassidy et al. 2013).
A total of 12 compounds were screened for bioactivity (Figure 7.1), specifically the parent anthocyanin – C3G - and 11 of its metabolites. The selection of metabolites was targeted to allow basic structure activity relationship examination. For example, both degradants of C3G, protocatechuic acid (PCA) and phloroglucinaldehyde (PGA), were chosen to examine the effect of degradation on bioactivity. The products of methyl, glucuronide and sulfatate conjugation of PCA were also included to investigate the effect of B-ring catechol modification via phase II metabolism on bioactivity. Finally, ferulic acid (FA) was one of the common metabolites identified in 13C-C3G bioavailability study (Czank, Cassidy et al. 2013) and therefore included in the present investigation.
Nitric oxide (NO) is a key mediator in maintaining endothelial homeostasis and its loss leads to endothelial dysfunction (Bian, Doursout et al. 2008). The loss of NO could result from down regulation of eNOS, or over production of superoxide which reacts with NO to produce the extremely reactive species peroxynitrite. Therefore, the vascular activity of C3G and selected metabolites was examined by investigating effects on basal eNOS up-regulation [Figure 7.1(1)].
In these experiments three of the 12 compounds significantly up-regulated basal levels of eNOS [Chapter 2, p<0.05, C3G, PGA and vanillic acid (VA)], three compounds significantly reduced eNOS [p<0.05, PCA, PCA-3-sulfate (PCA-3-sulf) and PCA-4-sulfate (PCA- 4-sulf), Chapter 2], while 7 had no activity. These experiments showed that degradation and subsequent metabolism of C3G has variable effects on eNOS regulation as some metabolites retained the activity of the parent structure, while other showed reduced activity.
Figure 7.1 Experimental scheme for assessment of vascular and anti-inflammatory activity in vitro Glucose O O H O+ OH OH OH OH O H OH O O OH OH O CH3 OH OH O O H OH OH O O H Experiment Treatment compounds Vanillic Acid (VA) (1)Basal eNOS and ET-1 Bioactivity Screening (3) NF- κB
(oxLDL, CD40L & IL-1β stimulated) (2) IL-6 (oxLDL& CD40L- stimulated) (1)Ang II stimulated O2.-(cytochrome
cand EPR probe reduction) (2)VCAM-1 (oxLDL& CD40L- stimulated) Cyanidin-3-glucoside (C3G) Ferulic acid (FA) Phloroglucinaldehyde (PGA) Protocatechuic acid (PCA) OH O O O H O H O H OH O OH O OH OH O H O O OH O O H O O H CH3 O OH O H CH3 OH O O H PCA-3-Glucuronide
(PCA-3-Gluc) PCA-4-Glucuronide (PCA-4-Gluc)
Vanillic acid (VA) Isovanillic acid (IVA) Protocatechuic acid (PCA) CH3 O OH O H
Scheme for investigation of vascular (1) & anti-inflammatory (2) activity in vitro, in primary human endothelial cell model, of selected treatment compounds, and exploration of mechanisms potentially underlying observed bioactivity with identified lead compounds (3). ‘X’ denotes possible bioactivity mechanism not selected for further investigation owing to lack of activity of anthocyanin degradants in previous experiments. CD40L, cluster of differentiation 40 ligand; EPR, electro paramagnetic resonance; ET-1, endothelin–1; eNOS, endothelial nitric oxide synthase; IL-1β, interleukin-1 beta; IL-6, interleukin-6; oxLDL, oxidised low density lipoprotein; O2˙-,
superoxide; NF-κB, nuclear factor kappa B; VCAM-1, vascular cell adhesion
Anti- inflammatory Vascular activity O OH O H S O H O O OH O O H S O O OH O O O H S O O OH C H3 O O O H S O H O O CH3
In addition, there was no effect observed for any of the compounds screened on basal endothelin–1 (ET-1) expression in endothelial cells [Figure 7.1(1)], though perhaps a cell model of stimulated ET-1 expression may provide a better insight into the vasodilatory bioactivity of the selected metabolites. Moreover, no effect was observed for any of the compounds tested on stimulated endothelial superoxide production [Figure 7.1(1)], owing to the lack of effect of the stimulus (angiotensin II) in the cell model utilised (HUVEC) as measured by reduction of cytochrome c (measured by spectrophotometry) and 1-hydroxy-3- carboxy- 2,2,5,5-tetramethylpyrrolidine (CPH) probe [measured by electro paramagnetic resonance (EPR)]. Hence, a more potent stimulus such as oxidised low density lipoprotein (oxLDL) (Heinloth, Heermeier et al. 2000) may be required to stimulate superoxide production in HUVECs, or a different cell type such as human coronary arterial endothelial cells (HCAECs) or monocytes. In addition to the use of a more potent stimulus to stimulate superoxide production, a shear stress effect (laminar vs oscillatory) might also be employed to modulate endothelial superoxide production and eNOS expression (Boo and Jo 2003, Hwang, Ing et al. 2003, Hsiai, Hwang et al. 2007).
Key inflammatory mediators such as VCAM-1 and IL-6 propagate the formation of atherosclerotic plaques which eventually result in significant clinical events (Cybulsky, Iiyama et al. 2001, Schuett, Luchtefeld et al. 2009). Therefore, all 12 compounds were also screened for their anti-inflammatory activity against VCAM-1 and IL-6 expression [Figure 7.1(2)] in response to two physiologically relevant stimuli, namely, oxLDL and cluster of differentiation 40 – ligand (CD40L) (Ishigaki, Oka et al. 2009, Pamukcu, Lip et al. 2011). In this case the majority of the compounds tested reduced expression of VCAM-1 (Chapter 3) and IL-6 (Chapter 4) under both stimulation conditions. Of the 12 compounds tested, seven reduced CD40L-stimulated VCAM-1 production (Chapter 3), eight reduced CD40L-stimulated IL-6 expression (Chapter 4), and nine compounds reduced oxLDL-stimulated IL-6 production in HUVECs (Chapter 4). Alternatively, three compounds, increased CD40L-induced VCAM-1 protein production in HUVECs, namely PCA-4-sulf, PCA-3-sulf and VA-4-sulf. In the present study soluble VCAM-1 was quantified, however proteolytic cleavage is required to produce the soluble form and an assay comparing levels of the soluble form of VCAM-1 and membrane bound VCAM-1 may provide more insight into the mechanistic activity of the phenolic metabolites. The bioactive compounds in the current study showed greater effects onanti-inflammatory activity, as opposed to vascular activity, it is possible that anthocyanin
inflammatory activities. Therefore, an understanding of the underlying mechanisms of action is crucial, and as such activation of a key inflammatory transcription factor, nuclear factor kappa-B (NF-κB), which up-regulates VCAM-1 and IL-6 expression, was investigated (Chapter 5). As the aim was to investigate the underlying mechanisms of action for the most bioactive metabolites of C3G tested, PCA and VA were chosen as they attenuated both VCAM-1 and IL-6 expression under both stimulation conditions [Figure 7.1(3)].
NF-κB activation was investigated by measuring phosphorylation of the p65 subunit of NF-κB (using flow cytometry), which results in translocation of NF-κB p65 to the nucleus and up-regulation of pro-inflammatory cytokines (Huang, Yang et al. 2010). However, no increase in phosphorylation of NF-κB p65 was observed following oxLDL and CD40L stimulation of HUVECs. Therefore, to explore the activity of PCA and VA on NF-κB p65 phosphorylation, HUVEC stimulation was performed using a more potent stimulus of p65 physphorylation, namely, IL-1β (Nelson, Paraoan et al. 2002). The results from this investigation indicated that both PCA and VA were capable of reducing the phosphorylation of NF-κB p65, and therefore, it can be postulated that this is one mechanism by which anthocyanin phenolic metabolites, exert their activity. However, these observations should also be confirmed using methodologies such as an NF-κB reporter assay (luciferase assay) or an NF-κB DNA binding activity assay (Xia, Ling et al. 2007).