The relationship between antiglycation activity and ...

167

Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Université de Montréal on 01/15/15 For personal use only.

The relationship between antiglycation activity and procyanidin and phenolic content in commercial grape seed products Cathy Sun, Kristina McIntyre, Ammar Saleem, Pierre Selim Haddad, and John Thor Arnason

Abstract: Eight commercial grape seed products (GSPs) were assessed for their inhibition of the formation of advanced glycation end-products in vitro. All 8 commercial GSPs included in this study were potent inhibitors of advanced glycation end-product formation with IC50 values ranging from 2.93 to 20.0 µg/mL. Total procyanidin content ranged from 60% to 73%. HPLC–DAD–ELSD results indicate that (+)-catechin, (–)-epicatechin, procyanidin B1, and procyanidin B2 were predominant and ubiquitously present in all the products under study, while gallic acid and procyanidin B4 were present in relatively minor amounts. The IC50 values correlated with total phenolic content, and multiple regression analysis indicated that IC50 is a linear function of the concentration of gallic acid and procyanidins B1, B2, and B4. Based on this study, GSPs have the potential to complement conventional diabetes medication toward disease management and prevention. Key words: advanced glycation end-products, grape seed extract, procyanidins, HPLC–DAD–ELSD, Vitis vinifera, antiglycation. Résumé : Nous avons examiné huit produits commerciaux à base de pépins de raisins (PPR) pour d’évaluer leur capacité à inhiber la formation des produits terminaux de glycation avancée (PTGA) in vitro. Les huit produits ont inhibé significativement la formation des PTGA avec des valeurs de CI50 allant de 2,93 à 20,0 µg/mL. La teneur en procyanidine totale a varié entre 60 et 73 %. Les résultats de la CLHP–DBD–DDL montrent la prédominance et la présence ubiquitaire de la (+)-catéchine, de la (+)-épocatéchine, de la procyanidine B1 et de la procyanidine B2 dans ces produits, alors que l’acide gallique et la procyanidine B4 n’y sont relevées qu’en assez petites quantités. Les valeurs de CI50 ont été corrélées avec le contenu phénolique total, et l’analyse de régression multiple a indiqué que la IC50 est une fonction linéaire de la concentration d’acide gallique et des procyanidines B1, B2 et B4. D’après nos résultats, les PPR peuvent s’avérer un complément utile aux médicaments classiques utilisés pour prévenir et traiter le diabète. [Traduit par la Rédaction]

Introduction Diabetes mellitus is a chronic disease that affects 246 million people worldwide, and present trends suggest that 380 million people will have diabetes by 2025 (International Diabetes Federation 2007). Diabetes is characterized by hyperglycemia, which favors the formation of glycated proteins and is responsible for many complications related to diabetes (Smit and Lutgers 2004; Ahmed 2005). Advanced glycation end-products (AGEs) are formed from protein glycation, which is a nonenzymatic addition of reducing sugars to proteins. AGEs are a consequence of high blood glucose in diabetes, and subsequently a cause of other degenerative diabetic conditions. Control of blood glucose is the most important treatment method in diabetes, but is rarely fully achieved, so the use of antiglycation agents to prevent formation of AGEs may be helpful in preventing damage, and progression of the disease symptoms.

The formation of glycated proteins begins when the carbonyl functional groups on the open chains of reducing sugars undergo a nucleophilic addition reaction with the amino group of proteins to form a Schiff base (Ahmed 2005). It takes a few days for the Schiff base to undergo the chemical rearrangements necessary to form a stable and usually irreversible Amadori product (Ahmed 2005), which eventually forms AGEs. AGEs are important contributors to diabetic complications (Ahmed 2005) and age-related health conditions (Ulrich and Cerami 2001). For example, glycation of the crystalline lens contributes to cataracts, glycation of myelin inhibits nerve signal transduction, and glycation of low-density lipoprotein (LDL) and collagen contribute to atherosclerosis (Smit and Lutgers 2004). Inhibition of AGE formation could thus alleviate complications in diabetic patients, and decrease the symptoms of AGE-associated conditions in the human aging process. Synthetic antiglycation drugs such as

Received 3 January 2011. Accepted 21 October 2011. Published at www.nrcresearchpress.com/cjpp on 9 February 2012. C. Sun, K. McIntyre, A. Saleem, and J.T. Arnason. Laboratory for the Analysis of Synthetic and Environmental Toxins (LANSET), Centre for Research in Biopharmaceuticals and Biotechnology, Department of Biology, University of Ottawa, 20 Marie Curie Private, Ottawa, ON K1N 6N5, Canada. P.S. Haddad. Natural Health Products and Metabolic Diseases Laboratory, Department of Pharmacology, Université de Montréal, Montréal, QC H3C 3J7, Canada; Institute of Nutraceuticals and Functional Foods, Université Laval, Québec City, QC G1K 7P4, Canada. Corresponding author: John Thor Arnason (email: [email protected]). Can. J. Physiol. Pharmacol. 90: 167–174 (2012)

doi:10.1139/Y11-121

Published by NRC Research Press

168

Can. J. Physiol. Pharmacol. Vol. 90, 2012


Fig. 1. Structures of the compounds assessed in grape seed product extracts.

Fig. 2. Phytochemical characterization of representative commercial grape seed products. Upper panel, high-performance liquid chromatography (HPLC) – evaporative light scattering detector chromatograms of grapeseed product D. Lower panel, HPLC – diode array detector chromatograms of grapeseed product G.

aminoguanidine have had some limited success in human clinical trials (Peyroux and Sternberg 2006). In addition, many natural plant extracts such as guava (Wu et al. 2009), corn silk (Farsi et al. 2008), blueberry (McIntyre et al. 2009), and sorghum bran (Farrar et al. 2008) have been shown to be effective inhibitors of protein glycation in vitro, and thus may provide an opportunity to inhibit glycation with dietary supplements. In vivo work has shown that green tea (Babu et al. 2006) and cucurmin (Jain et al. 2006) significantly reduces AGEs in diabetic rats.

Reactive oxygen species are elevated in diabetes and promote glycation reactions. Previously, we found that the anitoxidant capacity of plant extracts and compounds (especially phenolics) is positively correlated with antiglycation activity (Harris et al. 2011). Therefore, we conducted the present study on natural health products formulated with the phenol-rich seed extracts of grape (Vitis vinifera L.). Grape seed extract (GSE) has been shown to offer protection against cardiovascular disease, including atherosclerosis, which is one of the most common diseases associated with Published by NRC Research Press

Sun et al.

169

Table 1. Content of individual compounds in each grape seed product as determined by HPLC–DAD–ELSD analysis.


Content (mg compound/g product) Product code A B C D E F G H

Gallic acid 0.24±0 0.04±0 0 0 0.11±0 0 0.06±0 0.22±0

Procyanidin B1 5.20±0.01 0.60±0.01 2.76±0 2.28±0 3.31±0 0.31±0.06 1.22±0 10.25±0.06

Procyanidin B3 0 0 0 0 0 0 0 1.42±0.02

Catechin 8.44±0 0.88±0 1.80±0 2.52±0.01 4.24±0 0.37±0 1.74±0 11.04±0.53

Procyanidin B4 1.00±0.01 0.15±0 0.79±0.01 0 0 0 0.31±0 0

Procyanidin B2 5.18±0.02 0.45±0 11.21±0.02 2.61±0.01 7.32±0 0.81±0 3.32±0.01 5.25±0.35

Epicatechin 8.18±0.01 1.22±0 1.69±0.01 1.47±0.01 1.21±0 0.38±0 2.40±0.01 9.63±0.34

Note: HPLC–DAD–ELSD, high-performance liquid chromatography – diode array detector – evaporative light scattering detector.

Fig. 3. Concentration-dependent inhibition of advanced glycation end-product (AGE) formation by grape seed product E showing mean percent inhibition from 4 replicate microplate wells. The percent inhibition of the formation of AGEs by grape seed product E is tightly related to its concentration. The relationship is described by the following linear equation: y ¼ 27:2logðxÞ þ 15:4; r2 = 0.99 (p < 0.05).

Fig. 5. Relationship between antiglycation activity and total phenolic content. IC50 (presented as µg/mL) is plotted against total phenolic content (expressed as µg quercetin equivalents per milligram extract). Values represent the mean ±SE of 3 replicates, except product D for which 2 replicates are included. The relationship is described by the following linear equation: y ¼ 1:98x þ 21:5; r2 = 0.73, p < 0.05.

Fig. 4. Comparison of advanced glycation end-product (AGE) inhibition activity for commercial grape seed products. The mean IC50 (±SE) of the 8 commercial grape seed products were obtained as detailed in Materials and methods for the inhibition of AGEs. Results are based on 3 replicates, except for product D for which 2 replicates are included.

Table 2. IC50 values of standard compounds in the inhibition of advanced glycation end-product formation. Standard compound Catechin Epicatechin Gallic acid Quercetin Procyanidin B1 Procyanidin B2

IC50 (µg/mL) ± SE 5.58±0.79 4.65a 0.56±0.08 1.76±0.71 2.18±0.23 1.41±0.17

Note: IC50 values for all standard compounds except epicatechin are based on 3 replicates. a IC50 values for epicatechin are based on 2 replicates.

aging and also one of the most life-threatening complications of diabetes (Kar et al. 2006). Grape seeds represent 0%–5% of the weight of the grape, but contain two-thirds of the grape’s extractable phenolics (Kar et al. 2006). Such grape phenolics have been linked to the French paradox, which

suggests that the French population has a lower risk of cardiovascular disease despite a diet that is relatively high in fats because of their consumption of alcoholic grape derived products such as wine (Renaud and De Lorgeril 1992). More specifically, the cardiovascular benefits of red wine have been attributed to the phenolics found in the skin and seeds of red grapes (Kar et al. 2006; Leifert and Abeywardena 2008a). The mechanism of the cardioprotective actions likely implicates the antioxidant properties of grape phenolics Published by NRC Research Press

170



Fig. 6. Relationship between antiglycation activity and grape seed product (GSP) content in specific phenolics. IC50 (µg/mL) for each product is plotted against its content in (a) gallic acid, (b) procyanidin B1, (c) procyanidin B5, and (d) procyanidin B2 (mg/g product). No single compound succeeds in explaining the major part of the variation in antiglycation activity. However, multiple regression analysis demonstrated that a distinct combination of these 4 compounds can account for most of the IC50 variation. The equation that describes this relationship is the following: IC50 = 8.91 – 0.09[gallic acid] + 0.002[procyanidin B1] – 0.002[procyanidin B2] + 0.01[procyanidin B4].

(Zhang et al. 2006; Leifert and Abeywardena 2008a). Recently, a GSE was shown to decrease levels of glycated hemoglobin in diabetic mice (Hwang et al. 2009). However, the antiglycation effects of a variety of commercial grape seed products (GSPs) have not been studied. The present study is based on GSPs, which are different from GSEs. GSEs contain only extracted grape seeds, whereas GSPs are more complex natural health products, also available over the counter. Each brand of GSPs represents a distinct formulation of a GSE component and other additives. GSEs are commonly used in studies (Yilmaz and Toledo 2004; Zhang et al. 2006; Hwang et al. 2009), whereas GSPs are rarely used (Leifert and Abeywardena 2008b), even though they represent a greater proportion of what is actually consumed by the public. GSPs are considered to be safe dietary supplements in Europe and the USA (Kar et al. 2006) and are licensed natural health products in Canada.

The purpose of this study was to determine the effectiveness of 8 commercially available GSPs in the inhibition of AGE formation and to correlate this activity to their phenolic content. Owing to the predominance of procyanidins, we also assessed 4 B-type procyanidins and their 2 precursors (+)-catechin and (–)-epicatechin.

Materials and methods Materials The HPLC-grade mobile phase for chromatographic analysis was purchased from Fisher Scientific (Ottawa, Ontario, Canada). Gallic acid, (+)-catechin, (–)-epicatechin, and quercetin were from Sigma (St. Louis, Missouri, USA) and resveratrol was from Extrasynthèse (Genay, France), whereas proanthocyanidin B1, B2, B3, and B4 were provided courtesy of D. Ferreira (University of Mississippi, USA). Published by NRC Research Press


Sun et al.

Eight commercial GSPs, in the form of gelatin or softgel capsules and caplets, were purchased as over-the-counter natural health products from local stores in Ottawa, Ont. The brands and manufacturers for GSP gelatin or softgel capsules were Health Balance (NPN 80003787; Inverness Medical Nutritionals Group, Freehold, New Jersey, USA), Natural Factors (NPN 80003899; Natural Factors, Coquitlam, British Columbia, Canada), Now (Now Foods, Bloomingdale, Illinois, USA), Organika (Organika Health Products Inc., Richmond, B.C.), Swiss (Swiss Herbal Remedies Ltd., Richmond Hill, Ont.), and Webber Naturals (NPN 80009680; WN Pharmaceuticals Ltd., Coquitlam, B.C.). The brands and manufacturers for GSP caplets were Equate (NPN 80003870; Vita Health Products Inc., Winnipeg, Manitoba, Canada) and Jamieson (NPN 80006061; Jamieson Laboratories, Toronto, Ont.). Each GSP was randomly assigned a letter from A to H to protect manufacturers’ interests and identities. Sample preparation Commercial GSPs were removed from the capsule and the formulations extracted using an orbital shaker at 30 rpm for 2 h with 80% ethanol (10 mL/capsule), according to a protocol previously optimized in our laboratory to allow the effective extraction of plant phenolics. The extract was centrifuged at 1000g and the supernatant removed. The residue was re-extracted with 5 mL of 80% ethanol and centrifuged again. The supernatants were combined and reduced to dryness in a rotary evaporator. Stock solutions of the extracts and standards for the bioassay were prepared in 80% ethanol (40 mg/mL); they were sonicated for 20 min to completely dissolve the extracts and filtered through 0.2 µm PTFE filters (Chromatographic Specialties Inc., Brockville, Ont.) to ensure they were sterile. For high-performance liquid chromatography – diode array detector – evaporative light scattering detector (HPLC–DAD–ELSD) analysis, the extracts were dissolved in 80% ethanol + 10% of 0.1% trifluoroacetic acid (TFA) at 4 mg/mL, stored overnight at 4 °C, then raised to room temperature, vortexed for 30 s, sonicated for 5 min, and filtered through 0.2 µm PTFE filters. HPLC–DAD–ELSD analysis An HPLC–DAD–ELSD method was developed using the following authentic standards available in our phytochemical library: gallic acid (1), procyanidin B1 (2), procyanidin B3 (3), (+)-catechin (4), procyanidin B4 (5), procyanidin B2 (6), and (–)-epicatechin (7) (Fig. 1). Thirty microlitres of each extract at a final concentration of 4 mg/mL were injected into an 1100 series HPLC–DAD–ELSD system (Agilent Technologies, Inc., Santa Clara, Calif.), consisting of an autosampler with a 100 µL built in loop, a quaternary pump, a column thermostat, a photodiode array detector, and an ELSD and Chemstation software (version B.03.02). Separations of 8 phytochemicals were achieved on a Zorbax XDBC18 column (4.6 mm × 150 mm; particle size 5 µm) (Agilent Technologies, Inc.). The column thermostat temperature was maintained at 47 °C. The solvent gradient was based on the following mobile phases: (A) 0.1% TFA in HPLC-grade water, (B) 0.1% TFA in HPLC-grade acetonitrile (ACN), and (C) HPLC-grade methanol. The gradient was delivered at a flow rate of 1.5 mL/min. The separation and quantification of target compounds was achieved by delivering 95% A and

171

5% C in isocratic mode from 0–5 min, 90% A, 5% B, and 5% C from 5–10 min, and 50% A, 25% B, and 25% C from 10–30 min. The column was washed with 100% B for 5 min followed by equilibration in 95% A and 5% C for 8 min before the next injection. ELSD conditions were as follows: temperature, 40 °C; gain, 1; offset, 0; sampling time, 10 Hz; and filter, 1 s. The calibration curves were generated by injecting 3 known concentrations of pure standards that covered the range of concentrations in the extracts. The identification of the compounds present in the extracts was confirmed by matching the UV spectra of the unknowns with the UV spectra of authentic standards (available in the Chemstation software library of the laboratory’s phenolic standards) injected under similar chromatographic conditions and eluting at the same retention time. Area under the peak quantification was carried out at the monitoring wavelength of 280 nm (bandwidth 4, reference off). Peak summing function was used in calibration settings to calculate the total procyanidin contents and reported as % in the extracts after including all the peaks that matched procyanidins spectra and excluding compounds 1, 4, and 7. Products A, B, E, F, and G were analysed by HPLC–DAD analysis and products C and D by HPLC–ELSD analysis to achieve the best resolution of peaks. Total phenolics assay The total phenol content of the GSP samples were estimated using the Folin–Ciocalteau method described by Singleton and Rossi (1965) as modified by Harris et al. (2007). Quercetin prepared in 80% ethanol was used to generate a standard curve. For the assay, 160 µL of GSE or quercetin was added to 800 µL of Folin reagent (BDH Inc., Toronto, Ont.). The mixture was briefly vortexed at 200 rpm and left at room temperature for 5 min. A 540 µL volume of 7.5% NaHCO3 was added and the mixture gently stirred. Three replicates of 200 µL/well of sample were transferred to a nonsterile clear-bottom 96-well plate (Nalgene International, Rochester, New York, USA.). The covered plate was incubated in the dark for 2 h at room temperature. Absorbance at 725 nm was read on a microplate reader (SpectraMax M5, Molecular Devices, Sunnyvale, California, USA), and total phenolic content was determined using data from 6 assay replicates and expressed as micrograms of quercetin equivalents per milligram of extract. Fluorescence-based inhibition of AGE formation assay AGE formation was assessed according to the method of Farsi et al. (2008) with some modifications (McIntyre et al. 2009). Briefly, a 200 mmol/L glucose – 200 mmol/L fructose mixture was mixed in equal amounts with 2 mg/mL BSA (both prepared in 100 mmol/L sodium phosphate buffer) and individual GSP extracts dissolved in ethanol at various concentrations were added. Several controls were also incorporated. A buffer blank (buffer only), a vehicle blank (buffer and vehicle), a BSA blank (buffer, BSA, and vehicle) served to control for fluorescence inherent in BSA. Extract blanks (buffer, BSA, glucose–fructose, and different sample concentrations) were used to control for any fluorescence inherent in the sample and negative (buffer, BSA, and glucose–fructose) and positive (buffer, BSA, glucose–fructose, and 10 µg/mL quercetin) controls were also included in each assay. For Published by NRC Research Press


172

every experimental condition, 4 replicates of 200 µL each were transferred to the wells of sterile polystyrene 96-well clear bottom plates (Corning, Inc., Corning, N.Y.). The covered plates were sealed and incubated in darkness at 37 °C for 7 d with shaking at 200 rpm. Following the incubation, fluorescence was measured using a microplate reader at excitation and emission wavelengths of 355 nm and 460 nm, respectively, and the percent inhibition of AGE formation was calculated as previously described (Farsi et al. 2008). The IC50, which is the plant extract concentration necessary to achieve 50% inhibition of the formation of AGEs, was determined using data from 3 assay replicates, unless indicated otherwise. Statistical analysis Linear and multiple regression analyses were conducted using S-Plus version 8.0 statistical software.

Results and discussion Grape seed product formulation The total mass of formulation contained in each capsule/ caplet ranged from 400 to 500 mg yielding 50 to 100 mg of 80% ethanol extract. Depending on the brand, the rest of the mass was made up of various components such as the capsule, starch, cellulose, magnesium stearate, etc. In this study, the sample extracts prepared reflect the different mixture composition of each GSP manufacturer and shows the variety in the brands available for consumption. HPLC–DAD–ELSD analysis All products showed a unique HPLC–DAD–ELSD profile, but with similar phytochemical markers (Fig. 1) upon analysis of the extract. Products A, B, E, F, and G were best resolved by HPLC–DAD, whereas C and D were best resolved by HPLC–ELSD. Representative chromatograms for each method are shown (Figs. 2a and 2b). The procyanidins eluted within 7.5–12.1 min in each chromatogram. Total procyanidin content, obtained by summing all the peak areas whose spectra matched with the selected procyanidins, ranged from 60% in product H to 73% in product D. The majority of the extracts showed a similar amount of total procyanidin of ~61%. HPLC–DAD–ELSD analysis showed that 2, 4, and 7 are the major phytochemicals in the samples, while 1, 3, and 5 were present in relatively minor amounts. Product B was the only extract that exhibited the presence of myricetin and an unknown flavonoid eluting at 14.6 min and 15.1 min, respectively. Products E and F showed unknown late eluting peaks of high intensity, compared with procyanidins. The composition of these late peaks was not determined. Table 1 presents a summary of the content for each selected compound present in each commercial product. Antiglycation activity and correlation with phytochemical content All 8 GSPs effectively inhibited the formation of AGEs in a concentration-dependent fashion, as illustrated in Fig. 3 for product E. Figure 4 shows the significant variation in the observed IC50 values of the commercial products. Product E was the most potent inhibitor of protein glycation with an IC50 of 2.93 ± 0.62 µg/mL, followed by products H and A.


Product D was the least effective inhibitor of protein glycation, with an IC50 of 20.03 ± 1.66 µg/mL. Product E’s IC50 was lower than that of the purified standard compounds catechin and epicatechin, which are major grape seed phenolics (Yilmaz and Toledo 2004). Interestingly, the IC50 values of all 8 GSPs tested were lower than values reported for aminoguanidine (44.4 µg/mL (Farsi et al. 2008) and 71.1 µg/mL (Jang et al. 2009)), a known synthetic antiglycation drug. This indicates that GSPs may be stronger AGE inhibitors than this reference compound. The novel antiglycation activity identified herein for GSE products adds a new dimension to other known health benefits such as antioxidant, antiinflammatory, and antiatherosclerosis (Kar et al. 2006; Leifert and Abeywardena 2008a). The assay for total phenolics revealed a variation in the phenolic content of the 8 GSPs. The total phenolic content in GSP’s correlated with their IC50values in the inhibition of the AGE assay (Fig. 5). The higher the GSP’s total phenolic content, the lower the IC50 in the antiglycation assay, which indicates more effective inhibition of AGE formation. Major grape phenolics (catechin, epicatechin, gallic acid, etc.) were all shown to be highly effective antiglycation agents when pure compounds were assessed individually (Table 2). Phenols act both by quenching reactive oxygen species and by preventing their formation via chelating the metal ions that generate them (Pietta 2000). Figure 6 shows the relationship of IC50 values with the content of each of these 4 compounds. No single biomarker could efficiently and significantly predict antiglycation activity. However, multiple regression analysis showed that a distinct combination of gallic acid, procyanidin B1, procyanidin B2, and procyanidin B4 content is an excellent predictor of the IC50 value (r2 = 0.96). The following equation describes the relationship: IC50 ¼ 8:91 0:09½gallic acid þ 0:002½procyanidin B1 0:002½procyanidin B2 þ 0:01½procyanidin B4 All independent variables entered the model significantly (p < 0.05), and the overall p-value was