Cellular evaluation of the antioxidant activity of U.S. Pecans [Carya illinoinensis (Wangenh.) K. Koch]

Food Chemistry 293 (2019) 511–519

Contents lists available at ScienceDirect

Food Chemistry journal homepage: www.elsevier.com/locate/foodchem

Cellular evaluation of the antioxidant activity of U.S. Pecans [Carya illinoinensis (Wangenh.) K. Koch] Mary E. Kelletta, Phillip Greenspanb, Yi Gonga, Ronald B. Pegga, a b

T

⁎

Department of Food Science & Technology, College of Agricultural and Environmental Sciences, The University of Georgia, 100 Cedar Street, Athens, GA 30602, USA Department of Pharmaceutical & Biomedical Sciences, College of Pharmacy, The University of Georgia, Athens, GA 30602, USA

A R T I C LE I N FO

A B S T R A C T

Keywords: Pecans Antioxidants Phenolics Proanthocyanidins Cell-based assay Caco-2 cell line

Clinical trials show an inverse relationship between the consumption of antioxidant-rich tree nuts and the development of chronic diseases. This study examined antioxidant efficacy of U.S. pecans using a modified cellular antioxidant activity (CAA) assay with comparisons to data from in vitro antioxidant assays (hydrophilic-oxygen radical absorbance capacity {H-ORACFL} and ferric reducing antioxidant power {FRAP}). Crude phenolic extracts from both raw and roasted pecans were analyzed. In the CAA assay, pecan phenolics were taken up by human colorectal adenocarcinoma (Caco-2) cells and bestowed CAA, determined by monitoring the fluorescence of 2′,7′-dichlorofluorescein. Phenolics (25–100 μg/mL) demonstrated a reduction in fluorescence by 37–69% for raw and 26–68% for roasted pecans. The primary active phenolic constituents were determined by high-performance liquid chromatography–electrospray ionization–mass spectrometry (HPLC–ESI–MS) to be epi(catechin) dimers and trimers. These oligomeric procyanidins, ranging in size from 560 to 840 g/mol appear to be small enough for cellular uptake, showing pecans are an effective antioxidant in biological systems, regardless of roasting.

1. Introduction In recent years, tree nuts have gathered positive attention due to their myriad of health benefits, which include fewer instances of cancer, longer lifespans, and better weight management (Martínez-González & Bes-Rastrollo, 2011; Bao et al., 2013). The healthfulness of tree nuts was asserted officially in July 2003 when the United States Food and Drug Administration (US FDA) granted tree nuts a category B qualified health claim, due to strong evidence that regular consumption reduced the risk of coronary heart disease; this point was reinforced in the 2010 Dietary Guidelines for Americans which state, “moderate evidence indicates that eating peanuts and certain tree nuts reduces risk factors for cardiovascular disease when consumed as part of a diet that is nutritionally adequate and within calorie needs” (Center for Food Safety and Applied Nutrition, 2003; U.S. Department of Agriculture & U.S. Department of Health and Human Services, 2010). Phenolic compounds in food possess antioxidant activity based on in vitro assays, and these exogenous sources are postulated to help the body’s endogenous antioxidant enzymes with protection against oxidative stress. This is especially important, as the damage caused by reactive oxygen species (ROS) and other free radicals is known to play a role in aging, as well as the development of several chronic diseases,

⁎

including cancer and heart disease (Ames, Shigenaga, & Hagen, 1993; do Prado, Aragão, Fett, & Block, 2009). ROS such as the superoxide anion (O2%−), alkoxyl radical (RO%), peroxyl radical (ROO%), hydrogen peroxide (H2O2), hydroxyl radical (HO%) and singlet oxygen (1O2) are produced naturally in the body and enhanced by environmental stresses. For this reason, there are endogenous antioxidant systems in place including antioxidant enzymes, metal-ion binding proteins, and DNA repair systems (Sies, 1993; Rizzo et al., 2010). However, when the ROS loads increase beyond levels that endogenous antioxidant systems can handle, extensive damage to cells, organelles and DNA can occur, leading to the development of chronic disease states (Halliwell, 1996a; Wolfe & Liu, 2007). With an increased awareness of chronic diseases and their causes, the importance of dietary antioxidants at purportedly ameliorating such conditions is gaining traction in the public sphere. Through consuming a diet rich in a variety of tocopherols, carotenoids, and phenolics, antioxidant balance can be maintained in the body (Halliwell, 1996b; Khan, Afaq, & Mukhtar, 2008). This was shown specifically for pecans in a clinical trial performed by Hudthagosol et al. (2011): they reported tocopherol levels in the blood were doubled, oxidized LDL cholesterol levels were reduced by nearly a third, and ORACFL values for lipophilic- and hydrophilic-antioxidants had increased significantly in the plasma in the hours following test meals

Corresponding author. E-mail address: [email protected] (R.B. Pegg).

https://doi.org/10.1016/j.foodchem.2019.04.103 Received 24 October 2017; Received in revised form 14 April 2019; Accepted 25 April 2019 Available online 03 May 2019 0308-8146/ © 2019 Elsevier Ltd. All rights reserved.

Food Chemistry 293 (2019) 511–519

M.E. Kellett, et al.

(AAPH), Trolox® (6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid), (+)-catechin hydrate (≥98% purity by HPLC), quercetin dehydrate (≥98% purity by HPLC), TPTZ (2,4,6-tripyridyl-S-triazine), iron (II) sulfate heptahydate, and iron(III) chloride hexahydrate were purchased from the Sigma-Aldrich Chemical Co. (St. Louis, MO, USA). Advanced DMEM (Dulbecco's Modified Eagle Medium), phosphatebuffered saline (PBS, pH 7.4), Hanks’ Balanced Salt Solution (HBSS), fetal bovine serum (FBS), penicillin-streptomycin and MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) were purchased from Life Technologies (Grand Island, NY, USA).

with pecans. In general, antioxidants function via two main mechanisms: hydrogen atom transfer (HAT) and single electron transfer (SET) (Craft, Kerrihard, Amarowicz, & Pegg, 2012). The HAT mechanism is considered more biologically relevant and involves the donation of a hydrogen atom to a radical species, effectively stopping its propagation. The antioxidant then undergoes resonance stabilization to create a stable, nonreactive radical. The SET mechanism is more like a classical redox reaction, by which an electron is transferred to quench a radical. As these are very different antioxidative reaction mechanisms and some phytochemicals use a combination of both pathways to quench radicals, it is important to use multiple assays to fully analyze a sample’s antioxidant capacity (Prior, Wu, & Schaich, 2005). A common HAT assay is the Oxygen Radical Absorbance Capacity (ORACFL) assay, while a typical SET assay is the ferric reducing antioxidant power (FRAP) assay. Though in vitro assays such as ORACFL and FRAP are in wide use, evidence has shown that these measurements do not strictly correlate to the performance in vivo (López-Alarcón & Denicola, 2013). Their biological relevance has been held in question, and the United States Department of Agriculture (USDA) currently no longer reports ORACFL values, citing that “there is no evidence that the beneficial effects of polyphenol-rich foods can be attributed to the antioxidant capacity of these foods” as seen in this chemical test (Bhagwat, Haytowitz, & Holden, 2007). Overall, these in vitro assays miss important aspects of biological antioxidant function such as bioavailability, uptake, and metabolism of antioxidant compounds (Liu & Finley, 2005; Wolfe & Liu, 2007). Effective in vivo antioxidants may even modulate gene expression and up-regulate endogenous antioxidant enzymes (Jones, 2006; López-Alarcón & Denicola, 2013). Nevertheless, these in vitro assays are not entirely without merit: they can be a good starting point to identify foodstuffs with antioxidant/radical-scavenging activity worthy of further study. For the above reasons, the popularity of cell-based assays is on the rise, as they are a good middle ground between chemical tests with questionable applicability in the body and expensive, time-consuming human clinical trials and animal feeding studies. Cell-based systems like the cellular antioxidant assay (CAA) have been used to evaluate the potential of food antioxidants (Wolfe & Liu, 2007). Though many such studies use the HepG2 cell line (hepatocellular carcinoma), it has been said that the human Caco-2 (colorectal adenocarcinoma) cell line is more representative, because it bears marked similarity to epithelial cells in the small intestine (Goya, Mateos, & Bravo, 2007; Wolfe & Liu, 2007; Song et al., 2010; Wan, Liu, Yu, Sun, & Li, 2015; Kellett, Greenspan, & Pegg, 2018). Though several studies have employed the CAA assay to evaluate a variety of foodstuffs (mainly fruits and vegetables) as well as pure phytochemical antioxidative compounds, there has been no research examining the antioxidant activity of pecans or other tree nuts using a cell-based system (Wolfe & Liu, 2007; Wolfe et al., 2008; Song et al., 2010; Wan et al., 2015). In the present study, phenolic compounds from raw and roasted Georgia pecans were extracted and subsequently fractionated in order to assess their antioxidant properties by a modified CAA assay with Caco-2 cells.

2.2. Sample preparation Raw in-shell pecans, ‘Desirables’, from three different orchards in Ocilla, GA, were shipped to the Department of Food Science and Technology in Athens, GA, where they were kept frozen at −80 °C until analyzed. On the day of analysis, pecans were shelled, the nutmeats packaged in polyethylene plastic bags and stored at −80 °C between trials. This temperature maintained pecan freshness as well as facilitated better grinding and particle size reduction. 2.3. Roasting A portion of the shelled pecans (500 g for each trial) was removed from the −80 °C freezer and roasted before subsequent lipid and phenolics extraction. A roasting schedule designed by Erickson, Santerre, and Malingre (1994) and modified by Robbins (2012) was employed. Briefly, an impingement oven (Model 1450, Lincoln Foodservice Products, Fort Wayne, IN, USA) was used to roast pecan halves at 175 ± 10 °C for 8 min. These parameters were optimized by Robbins (2012) to match commercially the color of roasted pecans using the Commission Internationale de l’Éclairage (CIE) L * C * h system. The roasted pecans were cooled and subsequently stored at −80 °C until further analysis. 2.4. Lipid extraction Lipid constituents were removed from both raw and roasted samples using a Soxhlet apparatus. Briefly, pecan nutmeats (raw and roasted) were taken from the −80 °C freezer and placed in liquid nitrogen. Roughly 60 g of nutmeat were ground in a commercial coffee mill (Grind Central Coffee Grinder, Cuisinart, East Windsor, NJ, USA) using an intermittent pulsing technique until a fine powder was achieved. The frozen temperature and pulsing action kept any oil from being expelled during grinding. Twenty grams of ground pecan nutmeat were placed in a cellulose extraction thimble (single thickness, 43 mm i.d. × 123 mm external length, Whatman International Ltd., Maidstone, England). A thin plug of glass wool was placed in the top of the thimble to ensure the entirety of the sample would remain in place throughout extraction. Lipids were extracted for 18 h with ca. 300 mL of hexanes. When complete, the thimbles were removed and allowed to dry overnight in a fumehood. Hexanes were removed from the lipid extract with a Büchi Rotavapor R-210 using a V-700 vacuum pump connected to a V-850 vacuum controller (Büchi Corporation, New Castle, DE, USA) and a water bath at 45 °C. The lipid residue was dried in a 103 °C oven for 1 h. After drying and cooling the glassware in a desiccator, the lipid portion was weighed for gravimetric analysis.

2. Materials and methods 2.1. Chemicals Glass wool, sodium carbonate, ACS-grade hexanes, methanol, ethanol (95%), and acetone, as well as HPLC-grade water, methanol, hexanes, 2-propanol and acetonitrile were purchased from the Fisher Scientific Co., LLC (Suwanee, GA, USA). Glacial acetic acid, hydrochloric acid and dimethyl sulfoxide (DMSO) were acquired from VWR International, LLC (Suwanee, GA, USA). Sephadex LH-20, Amberlite, XAD-16, Folin & Ciocalteu’s phenol reagent, 2′,7′-dichlorofluorescin diacetate (DCFH-DA), 2,2′-azobis[2-amidinopropane] dihydrochloride

2.5. Extraction of phenolic compounds Phenolic compounds were extracted from defatted nutmeat using ((CH3)2CO/H2O/CH3COOH, 70.0/29.5/0.5, v/v/v) and the resultant crude extract was lyophilized as described by Robbins, Gong, Wells, Greenspan, and Pegg (2015). 512

Food Chemistry 293 (2019) 511–519

M.E. Kellett, et al.

2.6. Desugaring and fraction separation

2.9. Cellular antioxidant activity (CAA) assay

In order to further analyze the pecan phenolic compounds and to isolate the effects of proanthocyanidins, the crude phenolic extracts were desugared and separated into low-molecular-weight (LMW) and high-molecular-weight (HMW) fractions. Following the method of Srivastava et al. (2010) with slight modification, ca. 2 g of phenolic extract were mixed in a small amount of 10% CH3OH and sonicated until dissolved. The sample was applied to the top of a chromatographic column packed with Amberlite XAD-16 (bead size: 20–60 mesh; column 25 × 450 mm [I.D. × length]) using a Pasteur pipette. The column was then washed with deionized water until 0.0% Brix registered on a PAL-1 digital hand-held refractometer (Atago USA, Inc., Bellevue, WA, USA). This occurred after ca. 500 mL of distilled water had run through the column. At this point, the eluent was changed over to anhydrous CH3OH to recover the phenolics from the polystyrene resin. The CH3OH in the resultant extract was then removed using the Rotovapor and the aqueous remainder was placed in a crystallization dish, frozen at −80 °C and then lyophilized as aforementioned. Next, samples were mixed in a small volume of 75% ethanol and sonicated until dissolved. Using a Pasteur pipette, the desugared phenolic sample was then applied to a chromatographic column packed with Sephadex LH-20 (bead size: 25–100 μm; Chromaflex column, 30 × 400 mm [i.d. × length], Kontes, Vineland, NJ, USA). Compounds comprising the LMW fraction were eluted using ca. 1.5 L of 95% ethanol. Once the LMW fraction was removed, the eluent was changed to 50% aqueous (CH3)2CO. With ca. 600 mL of mobile phase, the HMW fraction was eluted. Ethanol and (CH3)2CO were removed from the LMW and HMW fractions, respectively, using the Rotavapor. Aqueous residues were frozen and lyophilized as described above.

Cellular antioxidant measurements were made following the method of Kellett et al. (2018). Briefly after Caco-2 cells reached confluence in Corning 25-cm2 culture flasks, cells were washed with PBS twice, then dissociated from the surface using 0.05% trypsin-EDTA. Cells were plated (6.0 × 104) in 100 μL cell culture media/well in Corning Costar® 96-well, black, flat bottom tissue culture-treated dishes and incubated until confluent (24–48 h). Wells on the perimeter were left empty to reduce any variation due to plate location. Growth medium was removed after confluence was achieved, and the cells were then washed with PBS to remove any non-adherent cells. Next, 50 μL of 25 μM DCFH-DA were applied to each well, followed by 50 μL of antioxidant treatments (in triplicate wells). Final treatments ranged from 2.5 to 200 μM quercetin and (+)-catechin, as well as 25 to 100 μg pecan phenolics/mL. For a control, 50 μL of dye and 50 μL of cell culture media (no antioxidant included) were applied to triplicate wells. Once the dye and antioxidant treatments were added, cells were incubated for 1 h at 37 °C. After this period, cells were washed with PBS 3 × to ensure any antioxidant effect was due to compounds being taken up by the cells. One hundred μL of the free radical generator, AAPH (600 μM), were added. Cells were immediately placed in a BMG FLUOstar Omega (Ω) microplate reader (BMG LABTECH, Inc., Cary, NC, USA), where the fluorescence was read initially and every five minutes thereafter for 1 h (13 readings in total). Fluorescence readings were taken at excitation/ emission wavelengths of 485/538 nm and the triplicate readings were averaged. 2.10. Quantitation of the CAA assay Effectiveness of antioxidant treatments was quantitated by examining the percent reduction in fluorescence. Briefly, a curve was generated by the 13 fluorescence response readings of each treatment over the course of the 1 h assay. The area under each curve was calculated by integration with the MARS Data Analysis Software (BMG LABTECH). Control wells yielded the maximum fluorescence, as there was no inhibition of the AAPH and DCFH-DA reaction. Percent reduction was determined by subtracting the ratio of the sample well (with antioxidant) to the control well (without antioxidant) from 1 and multiplied by 100. In other words,

2.7. Cell culture protocol for maintaining the Caco-2 cells The Caco-2 cell line was obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). Following the method of Xie, Kosińska, Xu, and Andlauer (2013), cells were cultured in Advanced Dulbecco’s Modified Eagle’s Medium (Advanced DMEM) supplemented with 10% endotoxin-free, heat-inactivated fetal bovine serum (FBS), 1% L-glutamine, 1% penicillin (10,000 U/mL) and 1% streptomycin (10,000 μg/mL) and incubated in a humidified 5% (v/v) CO2 environment (NAPCO Series 8000 WJ, Thermo-Scientific, Waltham, MA, USA) at a temperature of 37 °C. Cells were maintained in Corning 25-cm2 canted-neck cell culture flasks and split using 1:2 ratios when confluent, as confirmed using a Zeiss Primo Vert inverted microscope (Carl Zeiss, Inc., Oberkochen, Germany).

area under sample curve ⎞ % reduction = ⎛1 − ∗ 100 area under control curve ⎠ ⎝

2.11. Cell cytotoxicity assay To test the viability of the Caco-2 cells after pecan treatments were applied, the colorimetric MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay was employed (Bedoya-Ramírez, Cilla, Contreras-Calderón, & Alegría-Torán, 2017). This method is based on viable cells with active metabolism reducing the yellow colored MTT to a water-soluble purple colored formazan chromophore; i.e., a dye. When cells die, they lose their capability to convert MTT to formazan; hence, color formation serves as a useful and convenient marker for only viable cells. In other words, the quantity of dye formed is linearly associated with cellular enzyme activity and indirectly the number of viable cells present. DCFH-DA and the pecan phenolics were applied in the same fashion as described for the CAA assay and allowed to incubate for 1 h. Then, the cells were thoroughly washed with PBS to remove all phenolic residues and traces of media, which can potentially interfere with absorbance readings of the cytotoxicity assay. Control wells contained MTT and media, but no phenolics. One hundred microliters of serumfree culture media were applied to the cells along with 10 μL of 12 mM MTT. This mixture was incubated at 37 °C with 5% (v/v) CO2 for 4 h.

2.8. Preparation of chemical and phenolic sample solutions Stock solutions of each of the phenolic extracts from pecans were prepared by dissolving lyophilized powders in DMSO at a concentration of 25 mg/mL. On the day of analysis, samples were then diluted to final concentrations ranging from 25 to 100 μg/mL in culture media; final treatment solutions contained < 2% DMSO. For standard curves, quercetin dihydrate and (+)-catechin hydrate were similarly dissolved in DMSO for 50 mM stock solutions. On the day of analysis, standards were diluted to concentrations ranging from 2.5 to 200 μM in cell culture media with final treatment solutions containing < 2% DMSO. A 12.5 mM stock solution of the DCFH-DA fluorescent probe in CH3OH was prepared monthly; before each experiment, working solutions of 25 μM DCFH-DA in media were used. A 60 mM stock solution of AAPH in HBSS was prepared weekly and further diluted to 600 μM before experimental use. All DCFH-DA and AAPH solutions were kept frozen at −20 °C between uses, while samples in DMSO were kept refrigerated at 4 °C. 513

Food Chemistry 293 (2019) 511–519

M.E. Kellett, et al.

After which, 85 μL of media were removed and 50 μL of DMSO were added to each well to dissolve the generated formazan chromagen within the cells, and then mixed thoroughly with a multi-channel pipette. The samples were incubated at 37 °C again for 10 min and remixed to ensure thorough dissolution. The absorbance of the purple colored formazan was measured at 540 nm using the microplate reader. The assay was quantitated by comparing the absorbance of samples against that of the control; treatments resulting in significantly (p < 0.05) lower absorbance readings are considered cytotoxic. All assays were performed in triplicate.

Further characterization of the proanthocyanidin sample was accomplished by diol-phase HPLC-ESI-MS/MS. Briefly, an Agilent 1100 HPLC system was coupled to a Waters® Q-ToF (time-of-flight) micro™ Mass Spectrometer equipped with an ESI interface (Waters Corporation, Milford, MA, USA) operating in both positive- and negative-ion modes using capillary voltages of +3.5 kV and −2.5 kV, respectively. The HPLC conditions of separation were those described above. The microchannel plate detector voltage was set at +2.35 kV. Nitrogen was employed as the desolvation gas at a temperature of 100 °C and flow rate of 150 L/h. Argon was used as the collision gas. For normal MS, the collision voltage was set at 5 V, but for MS/MS, the collision voltage was increased to 30 V. A full scan was performed over the mass range of 200–3000 Da at a rate of m/z 13,000 per second. The MS/MS analyses were acquired by automatic fragmentation where the three most intense mass peaks where fragmented. Each mass spectrum generated was based on an average of five scans. Proanthocyanidin identity was determined based on tR mapping, m/z comparison with those reported in the literature, and MS/MS fragmentations.

2.12. Antioxidant and phenolic assays All tests were conducted on the following six powdered extracts of pecan phenolics: Raw Crude, Raw LMW, Raw HMW, Roasted Crude, Roasted LMW and Roasted HMW. All antioxidant tests were carried out using the microplate reader, equipped with temperature control, in clear-bottomed, black 96-well microplates. Hydrophilic-ORACFL assay. To assess the antioxidant power of pecan phenolics in vitro by the HAT mechanism, the hydrophilic-ORACFL assay was performed as described by Robbins et al. (2015). Phenolic extract powders from pecans were diluted to 500 μg/mL in ethanol and then further diluted to 25 μg/mL in the phosphate buffer. Values were reported as mmol Trolox® eq. (TE)/100 g nutmeat fresh weight (f.w.). Samples were prepared in duplicate and analyzed in triplicate for a total of six replicates that were averaged. Total phenolics content (TPC) assay. The TPC of pecan samples was determined as described by Robbins et al. (2015). Phenolic extract powders were dissolved in CH3OH and further diluted to 0.2 mg/mL. TPC values were reported as mg (+)-catechin eq./100 g pecan nutmeat f.w. Samples were analyzed in quadruplicate and averaged. Ferric reducing antioxidant power (FRAP) assay. The FRAP assay was employed to determine the antioxidant activity of pecan extracts using the SET mechanism as described by Robbins et al. (2015). The freezedried phenolic extracts (crude, LMW and HMW) were dissolved in CH3OH at a concentration of 0.2 mg/mL. FRAP values were reported as mmol Fe2+ eq./100 g nutmeat f.w. Samples were prepared in duplicate and analyzed in triplicate for a total of six replicates that were averaged.

2.14. Statistical analysis To analyze the data from the CAA assay and the MTT assay, an analysis of variance was used for all treatments. CAA data were gathered from a minimum of 12 wells and reported as means ± standard error. Differences were analyzed among all raw and roasted samples, sample preparations (crude, LMW and HMW) and concentrations (25–100 μg phenolics/mL media). Differences in means were determined using the least squares means procedure and a Student–Newman–Keuls (SNK) means separation test with p < 0.05 using IBM SPSS Statistics 23 (IBM, Armonk, NY, USA). 3. Results and discussion 3.1. In vitro antioxidant assays of pecan phenolic extracts Three extracts were prepared from both raw and roasted pecans (i.e., crude phenolic, LMW, and HMW) for a total of six treatments, which were assessed using standard in vitro measurements of antioxidant potential and determination of total phenolics content; that is, H-ORACFL, FRAP, and TPC. These three tests gave a fairly comprehensive picture, as they accounted for both HAT and SET reaction mechanisms. The results from these assays can be found in Table 1. The TPC values were reported as mg (+)-catechin equivalents (CE)/ 100 g nutmeat f.w. Values ranged from 344 ± 16 to 2156 ± 27 for the raw LMW fraction and roasted crude extract, respectively. The LMW fractions possessed the lowest amount of total phenolics. The crude

2.13. Characterization and quantitation of phenolic components Proanthocyanidins in the HMW fractions were separated based on their degree of polymerization (DP) using diol-phase HPLC with fluorescence detection. An Agilent 1200 Series HPLC was utilized for the normal-phase analyses coupled with a Princeton SPHER DIOL column (4.6 × 250 mm, 5 μm particle size, 60 Å; Princeton Chromatography, Inc., Cranbury, NJ, USA), equipped with a guard cartridge/holder system, a thermostatic column compartment set at 30 °C, and a fluorescence detector. A gradient elution comprising mobile phase (A) CH3CN/CH3COOH (98:2, v/v) and (B) CH3OH/H2O/CH3COOH (95:3:2, v/v/v) was employed. The linear gradient was completed at 1.0 mL/ min as follows: 0–35 min, 0–40% B; held for 5 min; 40–45 min, 40–0% B; and then an additional 5-min hold in order to equilibrate the system. A volume of 20 μL was injected for each sample; extracts were first dissolved in anhydrous CH3OH (20 mg/mL) and then further diluted 1:1 (v/v) with mobile phase A to a final concentration of 10 mg/mL. Samples were passed through a 0.45 μm PTFE syringe filter prior to injection. Fluorescence detection was employed with excitation/emission at 276/316 nm, respectively. The photomultiplier tube gain was held constant at 10 for the duration of the run time. Tentative identification, based on the DP, was made by tR mapping of commercial cocoa proanthocyanidin standards with DPs ranging from 2 to 10 (Planta Analytica LLC, Danbury, CT, USA) prior to ESI-MS/MS analysis. Quantitation of samples was based on tR values of the proanthocyanidin B2 standard.

Table 1 TPC (n = 4) and two antioxidant capacity assays (n = 6) for assorted pecan phenolics.a Pecan Phenolic Sample

TPC (mg (+)-catechin eq./ 100 g fresh nutmeat)

H-ORACFL (mmol TE/100 g fresh nutmeat)

FRAP (mmol Fe2+ eq./100 g fresh nutmeat)

Raw Crude Raw LMW Raw HMW Roasted Crude Roasted LMW Roasted HMW

1804 ± 21b 344 ± 16e 1645 ± 5c 2156 ± 27a

17.5 ± 1.3b 8.9 ± 0.2d 9.5 ± 0.3d 20.7 ± 1.1a

11.7 ± 1.2b 4.1 ± 0.2d 5.7 ± 0.3c 19.5 ± 1.0a

414 ± 4d 1837 ± 68b

8.4 ± 0.3d 11.8 ± 0.3c

5.3 ± 0.7c 11.6 ± 0.9b

a Means ± standard deviations followed by the different letter in a column are significantly different (p < 0.05). Abbreviations are as follows: TPC, total phenolics content; H-ORACFL, hydrophilic-oxygen radical absorbance capacity; TE, Trolox equivalents; FRAP, ferric reducing antioxidant power; LMW, low-molecular-weight fraction; and HMW, high-molecular-weight fraction.

514

Food Chemistry 293 (2019) 511–519

M.E. Kellett, et al.

from Georgia, New Mexico and Texas (Robbins, 2012). Research by Blomhoff, Carlsen, Andersen, and Jacobs (2006) indicated that pecans from the U.S., Mexico, and Norway had FRAP values ranging from 6.3 to 11.1 mmol Fe2+ eq./100 g nutmeat f.w. In general, these were somewhat lower than the results reported here. This is likely due to different methods of sample preparation, as well as the concentration effect that occurred through the fractionation process. Additionally, as the FRAP assay is time dependent, slight differences in reaction time before readings were taken could have a large effect. As mentioned above, all of these antioxidant measures can be affected by crop year, growing location, cultivar and a variety of other factors (Robbins, 2012). Additionally, it is normal for the trends seen in different antioxidant determinations to vary, as they measure antioxidant activity based on completely different mechanisms. For example, H-ORACFL is a HAT assay while FRAP is a SET assay. As antioxidants in food frequently operate using a combination of both mechanisms, multiple measures, as seen here, are required for thorough analysis of any given sample.

phenolic extracts had a TPC value of 1804 ± 21 and 2156 ± 27 for raw and roasted, respectively; the difference can be attributed to the heat during roasting liberating some bound phenolics within the pecan as well as the possibility of Maillard reaction products (MRPs) formation, notably melanoidins (Bolling, Blumberg, & Chen, 2010). What is noteworthy is that the phenolics of pecans are not susceptible to thermal degradation during roasting. Significant differences (p < 0.05) between raw and roasted pecans were seen between nearly all extracts, showing that the TPC is highly dependent on sample preparation. These differences were expected, as the LMW and HMW fractions required additional preparation steps, and were considered as more refined samples. Comparison to previous research proved difficult, because most TPC values in literature are reported as mg gallic acid equivalents (GAE)/ 100 g f.w.; in many cases there isn’t any consideration relating the chemistry of the dominant phenolic compound in the samples analyzed. This study used a (+)-catechin standard, as it is a more accurate and representative standard of phenolics found in pecan kernels. Average values from this study approximated those reported by Wu et al. (2004) (2016 ± 103 mg GAE/100 g nutmeat f.w.). Bolling, Chen, McKay, and Blumberg (2011) performed a review of all available antioxidant literature and found pecans had an average TPC value of 1588 mg GAE/ 100 g nutmeat f.w. In this work, both raw and roasted crude phenolic extracts had TPC values greater than studies such as those performed by de la Rosa, Alvarez-Parrilla, and Shahidi (2011) and Yang, Liu, and Halim (2009), which were 1170–1250 and 1227 mg/100 g nutmeat f.w., respectively; this was likely due to differences in sampling and the choice of extraction solvent. Results from the H-ORACFL assay showed slightly less variability between raw and roasted pecans than those of the FRAP assay, especially if the LMW fractions are regarded separately. The values reported in Table 1 ranged from 8.4 ± 0.3 to 20.7 ± 1.1 mmol TE/100 g nutmeat f.w. for the roasted LMW fraction and roasted crude extract, respectively. Just like the TPC data, the raw and roasted crude extracts were significantly different (p < 0.05) from one another and possessed the greatest H-ORACFL values of all samples tested. The LMW fractions exhibited no significant differences from one another, and possessed the lowest antioxidant activity. The data in this study is comparable to the work of de la Rosa et al. (2011), who reported three Mexican pecan cultivars of having H-ORACFL values of 23.1 ± 1.5, 26.2 ± 3.8 and 22.7 ± 5.0 mmol TE/ 100 g nutmeat f.w. The average value for the roasted crude extract in the present study was 20.7 ± 1.1 mmol TE/100 g nutmeat f.w., which was slightly lower than the de la Rosa et al. (2011) study. Wu et al. (2004) reported data for H-ORACFL that was also similar; they reported an average antioxidant capacity of 17.5 ± 1.0 mmol TE/100 g nutmeat f.w. This test shows good precision but values clearly vary due to nut cultivars, growing conditions, harvest locations, and extraction schemes, making larger pooled samples a good idea for consistent pecan data. The FRAP results presented in Table 1 ranged from 4.1 ± 0.2 to 19.5 ± 1.0 mmol Fe2+ eq./100 g nutmeat f.w. Once again, the LMW fractions possessed the lowest antioxidant activity at 4.1 ± 0.2 and 5.3 ± 0.7 mmol Fe2+ eq./100 g nutmeat f.w. for raw and roasted pecans, respectively. The raw crude extract had a FRAP value of 11.7 ± 1.2 mmol Fe2+ eq./100 g nutmeat f.w., while the roasted crude extract was nearly double at 19.5 ± 1.0 mmol Fe2+ eq./100 g nutmeat f.w. Interestingly, the roasted HMW fraction was markedly different from its raw counterpart with a FRAP value of 11.6 ± 0.9 vs that of 5.7 ± 0.3 mmol Fe2+ eq./100 g nutmeat f.w. This again demonstrates that antioxidant constituents did not degrade because of roasting. In fact, roasting might have released bound phenolics or generated Maillard reaction products to account for the increased antioxidant activity (Bolling et al., 2010), as determined by the FRAP assay. Previous data from this laboratory showed a range of 14.2 ± 1.2 to 20.7 ± 0.7 mmol Fe2+ eq./100 g nutmeat f.w. for a variety of cultivars

3.2. Cellular antioxidant activity (CAA) assay of pecan phenolic extracts The six extracts prepared for the in vitro tests were also examined using the modified CAA assay (Kellett et al., 2018). These included crude phenolic extracts as well as LMW and HMW fractions from both raw and roasted pecans. CAA results were measured against (+)-catechin not quercetin, the standard recommended by Wolfe and Liu (2007), because it is devoid in pecans. Phenolic treatment of cells ranged from 0.025 to 0.10 mg phenolics/mL, whereas standards ranged in concentration from 2.5 to 200 μM. Effectiveness of the antioxidant constituents in Caco-2 cells was measured by monitoring changes in fluorescence as DCFH-DA, after deacetylation by cellular esterases in the cell, was oxidized to 2′,7′-dichlorofluorescein (DCF) over the course of 1 h. Results were obtained by determining the area under the curve (AUC) for control wells compared to the AUC for various sample treatments. The control wells contained the maximum possible fluorescence, as no phenolic antioxidants were present to stop azo-initiated free radicals from attacking the DCFH probe. The findings were recorded as average % reduction in fluorescence (when compared to the control) with standard error and can be seen in Fig. 1 and Fig. 2A and B for (+)-catechin standards and pecan phenolics, respectively. For the (+)-catechin standard preparations, fluorescence was reduced in a dose-dependent manner: it was diminished by 25.5% at the lowest concentration (2.5 μM) and 61.0% at the highest concentration (200 μM). These results differ markedly from the previous research of Wolfe and Liu (2007), as the development of the CAA assay for HepG2

% Reduction in Fluorescence

60

50

40

30

20

0

50

100

150

200

(+)-Catechin Concentration ( M) Fig. 1. Cellular antioxidant activity of (+)-catechin standards at different concentrations. 515

Food Chemistry 293 (2019) 511–519

M.E. Kellett, et al.

% Reduction in Fluorescence

A 60

*

*

*

the HMW fractions. Interestingly, there was substantial overlap between LMW and crude treatments. Only the HMW fractions were significantly different (p < 0.05) than the crude and LMW treatments when applied at concentrations of 0.025 to 0.10 mg extract/mL media. No significant differences existed (p > 0.05) between raw and roasted samples, with the exception of the lowest concentration of roasted phenolics. This observation is contrary to the findings of the in vitro antioxidant activity assays in Table 1, which show that roasted pecan phenolic preparations were superior to their raw counterparts. During roasting, MRPs are generated, and these possess antioxidant activity. LMW MRPs can undergo cyclizations, dehydrations, retroaldolizations, rearrangements, isomerizations, and condensations to form melanoidins with molecular masses typically > 10 kDa (Wang, Qian, & Yao, 2011). Although melanoidins might impart a noticeable increase in antioxidant activity when performing an in vitro chemical assay, their size would prohibit these compounds from being taken up and contributing to the activity observed in a cellular system. This likely explains why no marked differences in the cellular antioxidant activity were evident between raw and roasted phenolic preparations. As the CAA assay was developed 12 years ago, standardization has not yet occurred (Wolfe & Liu, 2007). This makes it difficult to compare results to other works in literature based on the difference in choice of cell line, food or food extract being tested, and ‘optional’ wash steps with PBS. Wolfe and Liu (2007) employed the assay in HepG2 cells for a variety of extracts from common fruits; their 2007 study standardized all results in relation to the reduction in fluorescence of quercetin and found that for fruits, blueberry extracts were the most effective cellular antioxidant. Different results were observed with and without this wash step. With no PBS wash, blueberry extracts had the equivalent CAA of 171 ± 12 μmol quercetin eq. (QE)/100 g fruit, but with the critical PBS wash, the CAA was drastically reduced to the equivalent of 47 ± 1.9 μmol QE/100 g fruit (Wolfe & Liu, 2007). The CAA assay has also been performed on a variety of vegetables using gallic acid standards: beets and red pepper were found to be the most effective with CAAs equivalent to 41.9 ± 6.2 and 41.4 ± 1.8 μM QE, respectively (Song et al., 2010). Much like Wolfe and Liu’s work, the vegetable study showed that the employment of a PBS wash dramatically decreased CAA values for beets and red pepper to 4.78 ± 0.38 and 4.64 ± 0.19 μM QE, respectively (Song et al., 2010). For the present study, we also found that the PBS wash was an essential step in the assay. Sessa, Tsao, Liu, Ferrari, and Donsì (2011) used Caco-2 cells to look at the CAA of nanoencapsulated resveratrol and found that antioxidant activity was very high (> 80% reduction in fluorescence), which validated the use of the human colorectal cell line for these analyses. As this is the first reporting of CAA results for tree nuts, little direct comparison is possible. As indicated above, employment of a PBS wash in this study helps validate the method, as all reductions in fluorescence can be attributed to antioxidative processes within the cell, rather than surface phenomena. Because this investigation carried out a PBS wash with all runs and continued to see high CAA values for pecan phenolic extracts, these results are on par with and sometimes higher than others reported in the literature, thereby confirming that pecans can be an effective antioxidant in cells. This complements clinical data that showed that the phytochemicals in pecans are not only absorbable but also contribute to postprandial antioxidant defenses (Hudthagosol et al., 2011). More research is needed, however, to determine the specific pathways of antioxidant defense and the transporters involved, but these results assert the significance of a cell-based mechanism.

*

40

20

0

0.025

0.050

0.075

0.100

mg Pecan Phenolics/mL

*

% Reduction in Fluorescence

B 60

*

* *

40

20

0 0.025

0.050

0.075

0.100

mg Pecan Phenolics/mL Fig. 2. Cellular antioxidant activity of phenolic extracts from (A) raw pecan , raw pecan low-molecular-weight fraction, , raw crude extract, ; and (B) roasted pecan crude expecan high-molecular-weight fraction, , roasted pecan low-molecular-weight fraction, , roasted pecan tract, . An * denotes that there is a significant high-molecular-weight fraction, difference (p < 0.05) compared with the crude and low-molecular-weight treatments at each respective phenolic concentration.

cells showed virtually no antioxidant activity (i.e., less than 3% reduction in fluorescence when compared to quercetin), citing the lack of a 2,3-double bond and 4-keto group in the B-ring of catechin as the reason (Wolfe & Liu, 2007, 2008). However, in Caco-2 cells this was not a problem at all, as even modest concentrations significantly reduced fluorescence. As active membrane transport systems play a large role in cell-based assays, these discrepancies are likely due to differences in uptake or efflux transporters between the two cell lines (Déprez, Mila, Huneau, Tome, & Scalbert, 2001). The substantial CAA of catechin standards in Caco-2 cells supports our approach, as (+)-catechin and other flavan-3-ols are the basis for most phenolic compounds in pecans (de la Rosa et al., 2011; Hudthagosol et al., 2011), and are typical of most other tree nut species. Raw and roasted pecan phenolics protected the DCFH probe, and fluorescence was markedly reduced. As seen in Fig. 2A and B, crude acetonic phenolic extracts reduced fluorescence by 37.1 ± 3.0% to 48.2 ± 1.7% for raw pecans and 25.9 ± 3.8% to 48.6 ± 1.3% for roasted pecans. The HMW fractions significantly (p < 0.05) outperformed all other samples and reduced fluorescence in the range of 50.1 ± 1.2 to 69.1 ± 1.2% in raw samples, and 52.3 ± 1.1 to 68.3 ± 1.5% in roasted samples. In general, phenolic compounds of the LMW fractions were less effective cellular antioxidants than those of

3.3. Cytotoxicity results of pecan phenolic extracts To test the cytotoxicity of the pecan extracts and to confirm that the reduced fluorescence was resultant from antioxidant defenses and not cell death, the MTT viability assay was performed. Absorbance readings from this cytotoxicity assay can be seen for all pecan treatments and the 516

Food Chemistry 293 (2019) 511–519

M.E. Kellett, et al.

fractions having greater quantities of proanthocyanidins overall. Varying DPs were seen, but all were comprised of (epi)catechin and (epi)gallocatechin units with B-type linkages. As seen in Table 2, there were some monomers detected in the HMW fractions, but the majority of these constituents (and some of the smaller polymers) was recovered from the Sephadex LH-20 column with the first mobile phase of 95% ethanol, as reported in similar research by Robbins, Ma, Wells, Greenspan, and Pegg (2014). It is noteworthy that the vast majority (∼80%) of the proanthocyanidins in the HMW pecan extracts (using 50% aqueous acetone) were comprised of dimers and trimers. Roasting the ‘Desirable’ pecans did not affect the extent of polymerization of the proanthocyanidin constituents. Studies by Robbins et al. (2014) as well as Gong and Pegg (2017) reported similar results with dimers and trimers comprising 81.9% of the HMW phenolic fraction (i.e., 56.6% dimers; 25.3% trimers). In this work, dimers accounted for 44.1% and 46.4% of the proanthocyanidins in raw and roasted pecans, respectively, while trimers comprised 34.3% in raw pecans and 33.8% in their roasted counterparts. Previous work has demonstrated the antioxidant efficacy of individual proanthocyanidins. For instance, Lotito et al. (2000) employed a model experimental system of AAPH-induced oxidation of phosphatidylcholine liposomes and found that catechin monomers, dimers, trimers, and tetramers were equally effective at inhibiting lipid oxidation. Zhu, Holt, Lazarus, Orozco, and Keen (2002) demonstrated that dimers, trimers, and tetramers from cocoa were more effective than monomers in protecting against AAPH-induced hemolysis of rat erythrocytes in vitro. Additionally, Zumdick, Deters, and Hensel (2012) incubated procyanidins with Caco-2 cells and detected, by HPLC analyses, measurable quantities of dimers, trimers and tetramers in cell lysates. Similar results were observed when Déprez et al. (2001) incubated radiolabeled proanthocyanidins with Caco-2 cells. These aforementioned studies all provide evidence that the dimers and trimers present in pecans, representing the vast majority of HMW phenolics, should be able to either bind to or be taken up by Caco-2 cells, and would be effective antioxidants in the presence of free-radical generators. Pecan HMW extracts also included tetramers and pentamers, but these were detected at lesser concentrations, making up 13.8 and 17.2% of raw and roasted proanthocyanidins, respectively. Because of the large proportion of dimers and trimers (with molecular weights ranging from 560 to 840 g/mol), rather than larger polymers, coupled with clinical data and cellular data, one can postulate that many of the proanthocyanidins are absorbable for possible use in the body’s antioxidant defenses. Zhang et al. (2016) reported that proanthocyanidins are not acid labile, instead, they are stable in gastric juice, which enables them to be absorbed as intact molecules once they enter the small intestine. Their permeability is affected, however, by the DP. Proanthocyanidin polymers and oligomers with DPs > 4 are not directly absorbed in vivo, but procyanidin monomers and dimers could be detected in plasma. Such findings are also supported by research from Déprez et al. (2001), who proved that colonic microflora can breakdown polymeric proanthocyanidins into smaller molecules for absorption and utilization in the body. In line with this, clinical data by Hudthagosol et al. (2011) showed that despite the fact that pecan phenolics are complex, phenolic bioactives were found in plasma after pecans were consumed. If the compounds were too large, they would not be as readily absorbed by the epithelial cells and therefore could not bestow antioxidant activity within biological systems. Yet, these studies suggest that pecan phenolics are in fact small enough for effective antioxidant activity in the body. This large proportion of oligomeric proanthocyanidins is somewhat unique to pecans. In contrast, other food sources of proanthocyanidins, such as peanut skins, have much larger DPs (> 6), which are likely less biologically available due to their size (Ma et al., 2014). When these results are combined with the strong reduction in fluorescence seen in the CAA experiments, it is clear that pecan phenolics are very promising biological antioxidants.

Absorbance (

540 nm)

0.4

0.3

0.2

0.1

0.0

Control

0.10 Raw

0.025 Raw

0.025 Roasted

0.10 Roasted

mg/mL Phenolics Fig. 3. Cytotoxicity results from the MTT assay. No significant differences (p > 0.05) between treatments and control were observed.

control wells in Fig. 3. Crude pecan extracts of varying concentrations were added to the cells for 1 h and no significant differences (p > 0.05) were seen among sample cells and control cells. From this, we conclude that the pecan phenolics were not toxic to the cells at any of the concentrations examined in the study. 3.4. HPLC characterization of pecan tannins The HMW tannin fractions of both raw and roasted pecans, isolated by Sephadex LH-20 column using 50% (v/v) aqueous acetone as the mobile phase, were analyzed via NP-HPLC to determine what compounds were bestowing the antioxidant activity. Quantitation was achieved through tR mapping and area values from the cocoa proanthocyanidin standards. The chromatograms are depicted in Fig. 4 and characterization and quantitation for each peak can be found in Table 2. The tannin fraction was made up of proanthocyanidins, not ellagitannins. As evident in the chromatogram, the HMW fractions from raw and roasted pecans had very similar characteristics, with roasted

2

40

4

30

mAU

3

20

Roasted Raw

5

7

1

8

6

10 4

6

8

10

12

14

16

18

20

Retention Time, (tR, min) Fig. 4. NP-HPLC chromatograms for the high-molecular-weight extracts of raw and roasted pecans. The peak numbers refer to the characterization of the proanthocyanidins as reported in Table 2. 517

Food Chemistry 293 (2019) 511–519

M.E. Kellett, et al.

Table 2 Characterization of the proanthocyanidins in the high-molecular-weight fraction from pecans via HPLC-ESI-MS/MS. Peak No.

tR (min)

DP

Unit Type

Linkage

[M–H]− (m/z)

Raw (mg/g fr.)

Roasted (mg/g fr.)

1 2 3 4 5 6 7 8

4.1 5.6 7.1 8.6 9.4 10.5 12.2 14.5

1 2 2 3 3 4 5 5

catechin 2 (epi)catechin (epi)catechin + (epi)gallocatechin 3 (epi)catechin 2 (epi)catechin + 1 (epi)gallocatechin 4 (epi)catehin 5 (epi)catechin 4 (epi)catechin + 1 (epi)gallocatechin

N/A B B B B B B B

289 577 593 865 881 1153 1441 1457

2.81 ± 0.33 12.83 ± 1.35 3.17 ± 0.21 7.02 ± 1.33 5.43 ± 0.35 2.52 ± 0.18 0.96 ± 0.08 1.52 ± 0.21

1.02 ± 0.10 14.81 ± 2.35 3.63 ± 0.71 7.31 ± 1.04 6.12 ± 0.72 2.25 ± 0.20 2.28 ± 0.13 2.32 ± 0.58

Abbreviations are as follows: tR, retention time; DP, degrees of polymerization; fr., fraction; N/A, not applicable.

4. Conclusions

Food Science and Food Safety, 11, 148–173. de la Rosa, L. A., Alvarez-Parrilla, E., & Shahidi, F. (2011). Phenolic compounds and antioxidant activity of kernels and shells of Mexican pecan (Carya illinoinensis). Journal of Agricultural and Food Chemistry, 59, 152–162. Déprez, S., Mila, I., Huneau, J.-F., Tome, D., & Scalbert, A. (2001). Transport of proanthocyanidin dimer, trimer, and polymer across monolayers of human intestinal epithelial Caco-2 cells. Antioxidants & Redox Signaling, 3, 957–967. do Prado, A. C. P., Aragão, A. M., Fett, R., & Block, J. M. (2009). Antioxidant properties of pecan nut [Carya illinoinensis (Wangenh.) C. Koch] shell infusion. Grasas y Aceites, 60, 330–335. Erickson, M. C., Santerre, C. R., & Malingre, M. E. (1994). Oxidative stability in raw and roasted pecans: Chemical, physical and sensory measurements. Journal of Food Science, 59, 1234–1238. Gong, Y., & Pegg, R. B. (2017). Separation of ellagitannin-rich phenolics from U.S. pecans and Chinese hickory nuts using fused-core HPLC columns and their characterization. Journal of Agricultural and Food Chemistry, 65, 5810–5820. Goya, L., Mateos, R., & Bravo, L. (2007). Effect of the olive oil phenol hydroxytyrosol on human hepatoma HepG2 cells. European Journal of Nutrition, 46, 70–78. Halliwell, B. (1996a). Antioxidants in human health and disease. Annual Review of Nutrition, 16, 33–50. Halliwell, B. (1996b). Commentary oxidative stress, nutrition and health. Experimental strategies for optimization of nutritional antioxidant intake in humans. Free Radical Research, 25, 57–74. Hudthagosol, C., Haddad, E. H., McCarthy, K., Wang, P., Oda, K., & Sabaté, J. (2011). Pecans acutely increase plasma postprandial antioxidant capacity and catechins and decrease LDL oxidation in humans. The Journal of Nutrition, 141, 56–62. Jones, D. P. (2006). Redefining oxidative stress. Antioxidants & Redox Signaling, 8, 1865–1879. Kellett, M. E., Greenspan, P., & Pegg, R. B. (2018). Modification of the cellular antioxidant activity (CAA) assay to study phenolic antioxidants in a Caco-2 cell line. Food Chemistry, 244, 359–363. Khan, N., Afaq, F., & Mukhtar, H. (2008). Cancer chemoprevention through dietary antioxidants: Progress and promise. Antioxidants & Redox Signaling, 10, 475–510. Liu, R. H., & Finley, J. (2005). Potential cell culture models for antioxidant research. Journal of Agricultural and Food Chemistry, 53, 4311–4314. López-Alarcón, C., & Denicola, A. (2013). Evaluating the antioxidant capacity of natural products: A review on chemical and cellular-based assays. Analytica Chimica Acta, 763, 1–10. Lotito, S. B., Actis-Goretta, L., Renart, M. L., Caligiuri, M., Rein, D., Schmitz, H. H., ... Fraga, C. G. (2000). Influence of oligomer chain length on the antioxidant activity of procyanidins. Biochemical and Biophysical Research Communications, 276, 945–951. Ma, Y., Kosińska-Cagnazzo, A., Kerr, W. L., Amarowicz, R., Swanson, R. B., & Pegg, R. B. (2014). Separation and characterization of phenolic compounds from dry-blanched peanut skins by liquid chromatography–electrospray ionization mass spectrometry. Journal of Chromatography A, 1356, 64–81. Martínez-González, M. A., & Bes-Rastrollo, M. (2011). Nut consumption, weight gain and obesity: Epidemiological evidence. Nutrition, Metabolism & Cardiovascular Diseases, 21, S40–S45. Prior, R. L., Xu, X., & Schaich, K. (2005). Standardized methods for the determination of antioxidant capacity and phenolics in foods and dietary supplements. Journal of Agricultural and Food Chemistry, 53, 4290–4302. Rizzo, A. M., Berselli, P., Zava, S., Montorfano, G., Negroni, M., Corsetto, P., & Berra, B. (2010). Endogenous antioxidants and radical scavengers. Advances in Experimental Medicine and Biology, 698, 52–67. Robbins, K. S. (2012). Bioactive constituents of pecans [Carya illinoinensis (Wangenh.) K. Koch]. Ph.D. DissertationAthens, GA: University of Georgia. Robbins, K. S., Gong, Y., Wells, M. L., Greenspan, P., & Pegg, R. B. (2015). Investigation of the antioxidant capacity and phenolic constituents of U.S. pecans. Journal of Functional Foods, 15, 11–22. Robbins, K. S., Ma, Y., Wells, M. L., Greenspan, P., & Pegg, R. B. (2014). Separation and characterization of phenolic compounds from U.S. pecans by liquid chromatography–tandem mass spectrometry. Journal of Agricultural and Food Chemistry, 62, 4332–4341. Sessa, M., Tsao, R., Liu, R., Ferrari, G., & Donsì, F. (2011). Evaluation of the stability and antioxidant activity of nanoencapsulated resveratrol during in vitro digestion. Journal of Agricultural and Food Chemistry, 59, 12352–12360. Sies, H. (1993). Strategies of antioxidant defense. European Journal of Biochemistry, 215, 213–219.

Pecans have garnered increased interest recently, as their health benefits and the phytochemicals believed to be responsible are being isolated and characterized. Many of these benefits are due to the nut’s phenolic contents, particularly the proanthocyanidins. By both in vitro and cellular assays, the antioxidant efficacies of raw and roasted pecan phenolics were examined, with the HMW fractions outperforming LMW fractions. No significant differences (p > 0.05) existed between the cellular antioxidant activity of raw and roasted pecans. Because HPLCESI-MS/MS revealed that the majority of constituents in HMW fractions are (epi)catechin dimers and trimers, it is inferred that these are the compounds most likely responsible for the antioxidant activity of both crude and HMW extracts. As pecan extracts displayed significant antioxidant potential in a Caco-2 cell line, phenolic compounds appear to be small enough to be taken up by cells and can therefore bestow antioxidant activity on the cellular level, positively affecting human health. Acknowledgements The authors would like to acknowledge the USDA-NIFA-SCRI Award No. 2011-51181-30674 and the Georgia Agricultural Commodity Commission for Pecans (GACCP) for funding this research. Conflict of interest There are no conflicts to declare. References Ames, B. N., Shigenaga, M. K., & Hagen, T. M. (1993). Oxidants, antioxidants, and the degenerative diseases of aging. Proceedings of the National Academy of Sciences, 90, 7915–7922. Bao, Y., Han, J., Hu, F. B., Giovannucci, E. L., Stampfer, M. J., Willett, W. C., & Fuchs, C. S. (2013). Association of nut consumption with total and cause-specific mortality. New England Journal of Medicine, 369, 2001–2011. Bedoya-Ramírez, D., Cilla, A., Contreras-Calderón, J., & Alegría-Torán, A. (2017). Evaluation of the antioxidant capacity, furan compounds and cytoprotective/cytotoxic effects upon Caco-2 cells of commercial Colombian coffee. Food Chemistry, 219, 364–372. Bhagwat, S., Haytowitz, D. B., & Holden, J. M. (2007). USDA database for the oxygen radical absorbance capacity (ORAC) of selected foods. Presented at the American Institute for Cancer Research Launch Conference, Washington, D.C.. Blomhoff, R., Carlsen, M. H., Andersen, L. F., & Jacobs, D. R., Jr. (2006). Health benefits of nuts: Potential role of antioxidants. British Journal of Nutrition, 96, S52–S60. Bolling, B. W., Blumberg, J. B., & Chen, C.-Y. (2010). The influence of roasting, pasterusation, and storage on the polyphenol content and antioxidant capacity of California almond skins. Food Chemistry, 123, 1040–1047. Bolling, B. W., Chen, C.-Y. O., McKay, D. L., & Blumberg, J. B. (2011). Tree nut phytochemicals: Composition, antioxidant capacity, bioactivity, impact factors. A systematic review of almonds, Brazils, cashews, hazelnuts, macadamias, pecans, pine nuts, pistachios and walnuts. Nutrition Research Reviews, 24, 244–275. Center for Food Safety and Applied Nutrition. (2003). Labeling & Nutrition – Summary of Qualified Health Claims Subject to Enforcement Discretion. Retrieved April 12, 2019, http://www.fda.gov/Food/IngredientsPackagingLabeling/LabelingNutrition/ ucm073992. htm#cardio. Craft, B. D., Kerrihard, A. L., Amarowicz, R., & Pegg, R. B. (2012). Phenol-based antioxidants and the in vitro methods used for their assessment. Comprehensive Reviews in

518

Food Chemistry 293 (2019) 511–519

M.E. Kellett, et al.

Wolfe, K. L., Kang, X., He, X., Dong, M., Zhang, Q., & Liu, R. H. (2008). Cellular antioxidant activity of common fruits. Journal of Agricultural and Food Chemistry, 56, 8418–8426. Wu, X., Beecher, G. R., Holden, J. M., Haytowitz, D. B., Gebhardt, S. E., & Prior, R. L. (2004). Lipophilic and hydrophilic antioxidant capacities of common foods in the United States. Journal of Agricultural and Food Chemistry, 52, 4026–4037. Xie, Y., Kosińska, A., Xu, H., & Andlauer, W. (2013). Milk enhances intestinal absorption of green tea catechins in in vitro digestion/Caco-2 cells model. Food Research International, 53, 793–800. Yang, J., Liu, R. H., & Halim, L. (2009). Antioxidant and antiproliferative activities of common edible nut seeds. LWT, 42, 1–8. Zhang, L., Wang, Y., Li, D., Ho, C.-T., Li, J., & Wan, X. (2016). The absorption, distribution, metabolism and excretion of procyanidins. Food & Function, 7, 1273–1281. Zhu, Q. Y., Holt, R. R., Lazarus, S. A., Orozco, T. J., & Keen, C. L. (2002). Inhibitory effects of cocoa flavanols and procyanidin oligomers on free radical-induced erythrocyte hemolysis. Experimental Biology and Medicine, 227, 321–329. Zumdick, S., Deters, A., & Hensel, A. (2012). In vitro intestinal transport of oligomeric procyanidins (DP 2 to 4) across monolayers of Caco-2 cells. Fitoterapia, 83, 1210–1217.

Song, W., Derito, C. M., Liu, M. K., He, X., Dong, M., & Liu, R. H. (2010). Cellular antioxidant activity of common vegetables. Journal of Agricultural and Food Chemistry, 58, 6621–6629. Srivastava, A., Greenspan, P., Hartle, D. K., Hargrove, J. L., Amarowicz, R., & Pegg, R. B. (2010). Antioxidant and anti-inflammatory activities of polyphenolics from southeastern U.S. range blackberry cultivars. Journal of Agricultural and Food Chemistry, 58, 6102–6109. U.S. Dept. of Agriculture & U.S. Dept. of Health and Human Services. Dietary Guidelines for Americans, 2010. Retrieved April 12, 2019, http://www.cnpp.usda.gov/ DietaryGuidelines. Wan, H., Liu, D., Yu, X., Sun, H., & Li, Y. (2015). A Caco-2 cell-based quantitative antioxidant activity assay for antioxidants. Food Chemistry, 175, 601–608. Wang, H.-Y., Qian, H., & Yao, W.-R. (2011). Melanoidins produced by the Maillard reaction: Structure and biological activity. Food Chemistry, 128, 573–584. Wolfe, K. L., & Liu, R. H. (2007). Cellular antioxidant activity (CAA) assay for assessing antioxidants, foods, and dietary supplements. Journal of Agricultural and Food Chemistry, 55, 8896–8907. Wolfe, K. L., & Liu, R. H. (2008). Structure−activity relationships of flavonoids in the cellular antioxidant activity assay. Journal of Agricultural and Food Chemistry, 56, 8404–8411.

519