Characterization and Simultaneous LC-MS Quantification of Phenolic Compounds from Blueberries

Characterization and simultaneous LC-MS quantification of phenolic compounds from blueberries with widely divergent profiles

Abstract

A simple and sensitive LC-IT-TOF-MS method was validated for the profiling and simultaneous quantification of  anthocyanins, flavan-3-ols, flavonols, phenolic acids, and resveratrol in  blueberry genotypes with fruit color ranging from deep purple (Vaccinium angustifolium) to various shades of pink (crosses of V. corymbosum, V. darrowii, and V. ashei). Separation was performed on a C₁₈ column with a pore size of 12 nm with a gradient of water (formic acid 0.5% v/v) and methanol at a flow rate of 0.6 mL/min, and a column temperature of 50 ºC. Standard calibration curves were linear for all analytes with all correlation coefficients >0.99. The relative standard deviation for intra- and inter-day precision was lower than 10%. In vitro cellular bioassays indicated that the anthocyanins have the highest contribution to the antioxidant and anti-inflammatory activities. The method allowed an easy and selective identification and quantification of phenolics in blueberry samples with divergent profiles.

Key Words: Pink-fruited; blueberries; LC-IT-TOF-MS; quantification; antioxidant; anti-inflammatory

1. Introduction

Phenolic compounds comprise a wide and diverse range of secondary metabolites that are naturally present in fruits and vegetables. These compounds are part of an everyday diet and are used as medicines or supplements. Accumulated scientific evidence strongly suggests that long term consumption of diets rich in plant polyphenols offers protection against development of cancers, cardiovascular diseases, diabetes, osteoporosis and neurodegenerative diseases (Scalbert, Manach, Morand, Remesy, & Jimenez, 2005; Yao et al., 2004). Phenolic compounds are the major group of phytochemicals in berry fruits including flavonoids (anthocyanins, flavonols, flavones, flavanols, flavanones, and isoflavonoids), stilbenes, tannins, and phenolic acids.

Blueberry fruit contains several classes of bioactive phenolic phytochemicals including phenolic acids, anthocyanins, proanthocyanidins, stilbenes, and organic acids (Hakkinen et al., 1999; Kalt et al., 2001; Prior, Lazarus, Cao, Muccitelli, & Hammerstone, 2001). Since both the qualitative and quantitative aspects of the phenolic content contribute to the human health value, there is a need for a rapid, streamlined, sensitive, and simple method to investigate all groups of phenolics in blueberries.

In this study, a new LC/MS method was developed and validated to enable simultaneous qualitative and quantitative determination of anthocyanins, flavan-3-ols, flavonol glycosides, phenolic acids and stilbenes over a broad range of genotypes.  Six pink-fruited blueberry clones were cross-compared to lowbush wild blueberry to develop a streamlined method applicable for analysis of blueberries with markedly different flavonoid profiles and concentrations. Moreover, the antioxidant and in vitro anti-inflammatory activities of the blueberry extracts were investigated, and correlated to phenolic composition.

2. Materials and methods

2.1 Chemicals

Folin-Ciocalteu reagent, sodium carbonate, Trolox (6-hydroxy-2,5,7,8-tetramethyl chromane-2-carboxylic acid), ABTS (2,2′- azinobis-(3-ethylbenzothiazoline-6-sulfonic acid), dexamethasone (DEX), 2′,7′-dichlorodihydrofluorescein diacetate (H₂DCFDA) and lipopolysaccharide (LPS) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Dulbecco’s modified eagle’s medium (DMEM), TrypLE^TM, fetal bovine serum (FBS), Trizol, and cDNA Reverse Transcription kit were obtained from Life Technologies, (NY, USA). The mouse macrophage cell line RAW 264.7 (ATCC TIB-71) was obtained from American Type Culture Collection (Livingstone, MT, USA). All other solvents and chemicals used in this investigation were obtained from VWR International (Suwanee, GA, USA).

The following compounds were used as reference standards in the LC-MS/MS analysis: procyanidins B1 and B2, catechin, epicatechin, cyanidin-3-galactoside, cyanidin-3-glucoside, and malvidin-3-galactoside were obtained from Chromadex (Irvine, CA, USA). Delphinidin-3-glucoside was purchased from Cayman Chemicals (Ann Arbor, MI, USA). Delphinidin-3-galactoside, malvidin-3-glucoside, petunidin-3-glucoside, myricetin-3-glucoside, kaempferol-3-glucoside, and syringetin-3-glucoside were obtained from Extrasynthese (Genay Cedex, France). Cyanidin-3-arabinoside, and peonidin-3-glucoside were obtained from Polyphenols (Sandnes, Norway). Gallic acid, caffeic acid, chlorogenic acid, 2,4-dihydroxybenzoic acid, quercetin, quercetin glucoside and galactoside, quercetin arabinoside were purchased from Sigma-Aldrich (St. Louis, MO, USA). Phlorizin and daidzin were used as internal standards, and were also purchased from Sigma-Aldrich. The actual concentration of all these standards was calculated based on their purity.

2.2 Plant material

Pink-fruited blueberry clones (Vaccinium x hybrids) used in this study were grown at USDA-ARS plots at the Marucci Center for Blueberry and Cranberry Research and Extension at Chatsworth, NJ, USA. Six blueberry hybrids, derived from V. corymbosum, V. virgatum and V. darrowii crosses were investigated in this study. Table S1 (Supplementary material) lists the six clones; Pink Lemonade (PLE), Pink Champagne (PCH), Florida Rose (FLR), US 2117, US 2211 and US 2235. The ripe berries were harvested in late June-early July, and immediately frozen and stored at -70 °C until they were shipped to North Carolina on dry ice. A uniform composite individually quick frozen (IQF) sample (from harvested sites in Maine [USA] and Quebec, Nova Scotia, Prince Edward Island and New Brunswick [Canada]) of commercially grown lowbush blueberries (V. angustifolium, Aiton), was obtained from the Wild Blueberry Association of North America (Old Town, ME, USA) and stored as above. All blueberry samples were freeze-dried prior to extraction.

2.3 Extraction of blueberries

Freeze-dried berries (0.75 g freeze-dried blueberries × 3 replicates) were homogenized in 8 mL of 70% aqueous methanol (0.5% acetic acid) using a Pro 250 homogenizer (Pro Scientific Inc. Oxford, CT, USA) for 4 minutes. The obtained slurry was centrifuged (Sorvall RC-6 plus, Asheville, NC, USA) for 10 minutes at 4000 rpm. The supernatant was collected in a 25-mL volumetric flask. The residue was then extracted two more times, each with 8 mL solvent, supernatants were collected all together and a brought to final volume of 25 mL. The extract was used for investigations using chemical radical scavenging assays.  An additional aliquot was evaporated to a dry residue and used for in vitro cell culture assays.

2.4 Total phenolics, anthocyanins, proanthocyanidins, and radical scavenging assays

The amount of total phenolics was determined spectrophotometrically according to the Folin-Ciocalteu procedure (Singleton et al, 1999). Concentrations were expressed as mg gallic acid equivalent per g sample (DW) based on a standard curve created with gallic acid.  Total anthocyanins were determined by HPLC and quantified as mg cyanidin-3-glucoside equivalent/g sample. Total proanthocyanidins were determined by DMAC assay against a standard curve created with procyanidin B1 reference compound. Results were expressed as procyanidin B1 equivalent (Grace, Esposito, Dunlap, & Lila, 2014; Prior et al., 2001).

The radical scavenging activity was measured using the stable ABTS radical cation assay (Re et al., 1999) and Trolox as reference antioxidant. Results were reported as Trolox equivalents (µmol TE/g DW). The ferric reducing power of berry extracts was performed in a 96-well microplate using the FRAP assay (Grace et al., 2014). The reducing capacity of the extracts was calculated with reference to the reaction signal given by a FeSO₄ solution; FRAP values were expressed as µmoles of Fe²⁺/g of dried berry.

2.5 Preparation of samples and standard solutions for LC-MS analysis

Each of the twenty-four phenolic reference compounds was accurately weighed and dissolved individually in a solvent mix (methanol-water-formic acid, 65:35:5%) at concentration of 2.5 mg/mL. The individual standard solutions were stored at -80 °C freezer. Prior to LC-MS analysis, equal volumes from each standard solution were mixed together and diluted with the solvent mix to prepare standard stock mix solution (50 µg/mL). The standard working solutions, used for calibration curves, were prepared by appropriate dilution of the stock solution (0.02-40 µg/mL) in methanol-water-formic acid (65:35:5%). A stock solution mix of internal standards phlorizin and daidzin was prepared at a concentration of 0.2 mg/mL; it was added to all samples at a final concentration of 5 µg/mL.

LC-MS samples of pink fruited blueberries were prepared at 95% concentration by adding 950  µL sample extract + 25 µL internal standard mix + 25 µL solvent. Lowbush blueberry LCMS solution was prepared at 70% concentration by mixing 700 µL extract + 25 µL internal standard mix + 275 µL solvent. The dilutions were considered in the calculations for final concentrations.

2.6 LC-MS/MS instrument and conditions

Analysis was performed on a hybrid IT-TOF mass spectrometer (Shimadzu LC-MS-IT-TOF, Kyoto, Japan) equipped with two LC-20AD pumps, a SIL-20AC autosampler, a CTO-20A column oven, an SPD-M20A PDA detector, a CBM-20A system controller coupled to an IT-TOF-MS through an ESI interface. All data were processed by Shimadzu software, specifically, LCMS Solution Version 3.60, Formula Predictor Version 1.2, and Accurate Mass Calculator (Shimadzu).

The chromatography separations were performed on a Shim-pack XR-ODS column (50 mm × 3.0 mm × 2.2 μm) (Shimadzu, Japan) protected with a Phenomenex Security Guard column (4 mm × 3.0 mm, 5.0 m) (Phenomenex, Torrance, CA, USA), and the temperature of the column oven was maintained at 50 ºC. The following solutions were utilized as eluents: water (formic acid 0.5%, v/v) (A) and methanol (B). The percentage of solvent B was increased in a multistep mode: from 5–20% (0-30 min), 20-35% (30-35 min), 35-40% (35-37 min), 40-90% (37-38 min), 90-5% (38-40 min), then the column was re-equilibrated for 3 min at 5% B, at the flow rate of 0.6 mL/min.

The mass spectrometer was programmed to carry out a full scan over m/z 70–700 (MS1) and m/z 70–500 (MS2) in both positive and negative ionization modes. The heat block and curved desolvation line (CDL) temperature was maintained at 200 ºC; nitrogen was used as the nebulizing gas at a flow rate 1.5 L/min, and as the drying gas at 75 kPa; the interface voltage was (+), 4.5 kV; (−), −3.5 kV; the detector voltage was 1.61 kV, and sodium trifluoroacetate solution (2.5 mM) was used to calibrate the mass range from 50 to 2000 Da, and mass error of <5 mDa. Ultrahigh purity argon was used as the collision gas for CID experiments, and the collision energy was set at 50% for MS2.   In the MS1 mode, a 10 ms ion accumulation time was used, while in MS2 mode, instead, a 20 ms was used, and the window used for precursor ion isolation corresponds to a width of 3 amu. Shimadzu’s Formula Predictor software was used to predict the molecular formulae of the compounds. For quantitative analysis, multiple stage fragmental energy was canceled (CID was disabled), and ion accumulation was set at 80 ms. In the method for quantitative analysis, a table of selected ions (preferred ions) was set up based on accurate mass and retention time of each phenolic compound. Anthocyanin and flavonol glycosides were quantified as their extracted-ion chromatograms (EIC) in the positive ion mode with daidzin as an internal standard. Flavan-3-ols, phenolic acids, and resveratrol were quantified in the negative ion mode using phlorizin as an internal standard. Anthocyanin compounds that have no available reference standards were quantified using the standard curves of the anthocyanin having the same aglycone.

2.7 Method validation

To evaluate linearity, phenolic compound calibration curves consisted of 10 concentration levels (0.01-40 µg/mL) and each concentration was prepared and assayed in three runs on three separate days. The ratios of peak area of the analytes to that of internal standard were plotted against nominal concentrations of the analytes, and standard curves were in the form of Y = aX + b. The linearity ranges of the compounds, regression coefficients (r²), and the linear regression equations are given separately in Table 2

. The lowest limit of detection (LOD) and the lowest limit of quantitation (LOQ) were set at a signal to noise ratio (s/n) of 3:1 and 9:1, respectively (Table 2).

Intra- and inter-day variations were used to evaluate the precision of the developed method. Variations were expressed as the relative standard deviation (RSD) of the replicates.

To determine repeatability, three independently prepared solutions of lowbush blueberry were analyzed in triplicate. The RSD value was taken to be the measure of repeatability.

2.8 In vitro antioxidant and anti-inflammatory assays

Murine macrophage RAW264.7 cells (ATCC®, Rockville, MD, WI, USA) were subcultured in DMEM (10% FBS) according to the protocol previously (Esposito, Chen, Grace, Komarnytsky, & Lila, 2014).

2.8.1 Cell viability assay and dose range determination

RAW 264.7 cells were seeded in a 96 well plate for the viability assay. Cell viability was measured by the MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide] assay (Zheng & Wang, 2003) in triplicate and quantified spectrophotometrically at 550 nm using a microplate reader SynergyH1 (BioTek, Winooski, VT, USA). The concentrations of test samples that showed no changes in cell viability compared with that vehicle (80% ethanol) were selected for further studies.

2.8.2 Reactive oxygen species production in RAW264.7 macrophages

In vitro reactive oxygen species (ROS) were determined using a fluorescent dye protocol (Choi et al., 2007). The known antioxidant dexamethasone (DEX) was used as a positive control at 20 μM. The experiments were performed with two independent replications, each replication assayed at least in duplicate.

2.8.3 Nitric oxide radical inhibition in RAW264.7 macrophages.

The ability of test samples to inhibit nitric oxide (NO) radical formation was determined by a colorimetric assay using Griess reagent system (Promega Corporation, WI, USA) according to manufacturer protocol. The absorbance was recorded at 540 nm. NO production levels for each treatment were normalized to lipopolysaccharide (LPS).

2.8.4 Biomarkers of inflammation by gene expression analysis

The RAW 264.7 cells treated with all blueberry extracts were harvested in TRIzol reagent for total RNA extraction and purification, according to manufacturer protocol. cDNA synthesis and quantitative PCR were conducted, adopting a previously reported method (Esposito et al., 2014). All analyses were performed at least 3 times.

3.  Statistical analysis

Statistical analyses were performed using Prism 6.0 (GraphPad Software, San Diego, CA, USA). The results are expressed as means ± SEM. Data were analyzed by one-way ANOVA with treatment as a factor. Post hoc analyses of differences between individual experimental groups were made using the Dunnett’s multiple comparison tests.

4.  Results and discussion

Six pink-fruited blueberry clones, PLE, PCH, FLR, US 2117, US 2211, US 2235, and a uniform composite sample of lowbush wild blueberry were investigated in this study.  The information about the genotypes, pedigree, origin, and approximate color are listed in Table S1. The dry matter content ranged between 15 to 20.7% (Table 1).

4.1 Total phenolics, anthocyanins, and proanthocyanidins

All blueberry genotypes were analyzed for their content of phenolic constituents, and final concentrations were expressed as mg/g dry weight (DW). Total phenolics (TP) measured using Folin Ciocalteu assay and expressed as gallic acid equivalent showed that lowbush blueberries had 24.5 ± 0.69 mg/g TP. The pink-fruited blueberry samples contained TP concentrations that varied from 7.53 ± 0.19 mg/g for US 2211 to 15.0 ± 0.35 mg/g for PLE. Total anthocyanin content, based on measurements of HPLC peak areas recorded at 520 nm against a cyanidin-3-glucoside standard curve, indicated that anthocyanin levels ranged from 0.20 ± 0.02 mg/g for US 2211 to 2.12 ± 0.07 mg/g for US 2235 for pink-fruited blueberries. Lowbush blueberry had 12.6 ± 0.15 mg/g anthocyanin (Table 1). Total proanthocyanidins, measured by DMAC spectrophotometric assay and quantified as procyanidin B2 standard reference, indicated that FLR contained the highest proanthocyanidin concentration (3.92 ± 0.10 mg/g). PLE (3.51 ± 0.04 mg/g) and lowbush blueberry (3.47 ± 0.06 mg/g) were significantly lower than FLR. Other blueberries contained proanthocyanidin concentrations ≤ 1.13 mg/g (Table 1). There was high correlation between total phenolic content and total anthocyanin content at p˂0.001, but no significant correlation was observed between total phenolic content and total proanthocyanidins.

4.2 HPLC profiles for anthocyanins and proanthocyanidins

The HPLC-DAD profile of lowbush blueberry showed 17 peaks corresponding to 22 anthocyanins, which represented the spectrum of mono-glycosidic and acylated conjugates of the anthocyanindins delphinidin, cyanidin, petunidin, peonidin, and malvidin.  The lowbush blueberries has a relatively high anthocyanin content in contrast to the pink-fruited blueberry samples, thus afforded blueberries from both ends of the spectrum of anthocyanin content and flavonoid profile by which to validate this streamlined analytical protocol (see Fig. S1, Supplementary Material for HPLC-DAD profiles) (Wu & Prior, 2005).

Some of the pink-fruited samples (PLE and FLR) showed no detectable levels of acylated anthocyanin (30-37 min), while others showed traces. The monomeric anthocyanin peaks (#1-15) varied widely in relative intensities among the seven blueberry genotypes investigated.

Normal phase HPLC with fluorescence detection was able to separate proanthocyanidin components in the berry extracts according to their degree of polymerization (DP). FLR showed a large broad peak around 40 min indicating high concentration of polymeric proanthocyanidins (DP>12). The monomeric and oligomeric proanthocyanidins (DP1-DP10) showed a similar profile in all investigated berries but differed significantly in intensities (see Fig. S2, Supplementary Material).

4.3 Antioxidant capacity

The ABTS method measures the ability of the antioxidant to quench ABTS^•+ radical, while the FRAP assay measures the ability of the antioxidant to reduce the yellow ferric-TPTZ complex to the blue ferrous-TPTZ complex. Results showed that lowbush blueberry had the highest antioxidant capacity level, as measured by ABTS radical scavenging activity, and FRAP reducing power (127 ± 5.3 µmol Trolox equivalent/g and 389 ± 19.4 FeSO₄ µmol equivalent/g, respectively). All of the other blueberry genotypes showed lower ranges of antioxidant activities, 32.5 – 58.5 µmol/g for ABTS, and 149 – 347 µmol/g for FRAP (Table 1). The relative antioxidant capacity measured by ABTS assay strongly correlated with total phenolic contents (r=0.965, p<0.001) and total anthocyanins (r=0.962 for both HPLC-UV and LC-MS, p<0.001) (Prior et al., 1998), while antioxidant capacities measured by FRAP assay significantly correlated with total proanthocyanidins (r=0.764, p<0.05).

4.4 Qualitative and quantitative LC-MS analysis

4.4.1 Optimization of chromatographic and mass spectrometric parameters

To achieve best separation and strong ion signals, the chromatographic conditions, including mobile phase, column temperature and flow rate were optimized. Different linear gradients of mobile phase A (water with formic acid (0.1-0.5%) and mobile phase B (methanol with formic acid (0 or 0.5%), different column temperatures (30, 40, 50 ºC), and different flow rates (0.35, 0.45, and 0.6 mL/min) of solvent gradient were compared. The chromatographic conditions which led to the best base-line separation of anthocyanins from closely related structures was water with formic acid (0.5% v/v) (A) and methanol (B) as the eluents, with a flow rate of 0.6 mL/min, and a column temperature of 50 ºC. Due to complexity of the blueberry extracts, a detailed gradient program was employed (Section 2.6). Typical separation showing the extracted ion chromatograms (EIC) of standard reference mixture and the lowbush blueberry sample is shown in Fig.1.

4.4.2 Characterization of phenolic constituents by LC-IT-TOF-MS

In this study, some anthocyanin compounds were identified by comparison with reference standards, and others were tentatively characterized based on their retention times, UV/Vis, MS and MS/MS spectra referring to the literature (Gavrilova, Kajdzanoska, Gjamovski, & Stefova, 2011; Wu & Prior, 2005), and to our previous work (Grace et al., 2014). 37 phenolic compounds including 22 anthocyanins, 4 flavan-3-ols, 6 flavonols, 4 phenolic acids, and resveratrol were identified in the blueberry samples. The identification data are shown in Table S2 in the Supplementary Material.

4.4.3 Quantitative analysis of phenolic components by LC-IT-TOF-MS

The aim of this method was to simultaneously quantify several classes of phenolic constituents with a wide range of concentrations in different blueberry genotypes. HPLC-UV analysis has been the common method for quantification of anthocyanins by measuring peak areas recorded at 520 nm against a standard curve constructed with a single anthocyanin reference standard (Grace et al., 2014; Lohachoompol, Mulholland, Srzednicki, & Craske, 2008; Yousef et al., 2013) At this wavelength, anthocyanins can be selectively detected in the presence of other flavonoids, which have maximum absorbance at characteristic wavelengths other than 520 nm. The majority of the HPLC methods involve the use of high percentages of acids (1.0 to 15% v/v) as mobile phases to maintain a low pH (≤ 2) as a requirement for maintaining the stability of anthocyanins in solution in the form of flavylium cations (Merken & Beecher, 2000). HPLC-UV methods, however, require a long run time and consequently large volumes of solvent to achieve optimal resolution and to avoid co-elution of peaks. In addition, they have limited sensitivity and are inadequate for quantification of flavonoids at very low concentrations. Analytical techniques utilizing HPLC-MS have been used extensively for qualitative or semiquantitative analyses providing information on identification of anthocyanins. There are some reports on application of LC-tandem mass spectrometry for quantification of anthocyanin in food products, plant extracts and biological specimens (Ling et al., 2009; Montoro et al., 2006; Tian, Aziz, Stoner, & Schwartz, 2005; Wang, Kalt, & Sporns, 2000). However, the simultaneous quantification of anthocyanins and other classes of phenolics in food products by LC-MS is seldom reported in the literature (Nagy, Redeuil, Bertholet, Steiling, & Kussmann, 2009).   Mullen et al. used a full scan high resolution mass spectrometry with an Orbitrap analyzer for quantification of anthocyanin (as their precursor ions) in berries and berry-fed greenfinch brain tissue. Their results indicated that the Orbitrap analyzer had ca. 200-fold more sensitivity than traditional tandem MS in selected reaction monitoring (SRM) mode, and enabled both targeted and non-targeted compounds to be detected at much lower detection limits than HPLC-tandem mass (Mullen, Larcombe, Arnold, Welchman, & Crozier, 2010). The reason for this is that anthocyanins ionize approximately 10-100 times more efficiently compared to their aglycone counterparts due to the presence of the sugar moieties. This phenomenon was quantitatively reflected by lower limits of detection, quantification, and lower calibrated range for glycosylated analytes compared to their aglycones (Nagy et al., 2009). The use of LC-ESI-IT-TOF mass spectrometer for quantification purposes based on EIC of their precursor ion proved to be successful and very sensitive for quantification and quality control of plant metabolites. Examples include determinations of anthocyanin in commercial red and blue fruit juices (Fanali et al., 2011), phenolic acids and flavonoids in Apocynum ventum (An, Wang, Lan, Hashi, & Chen, 2013), multiple constituents in Chinese medicine (Liu et al., 2016), and monoterpenoids in peony root (Shi et al., 2016).

The method developed here used IT-TOF-MS for quantification of phenolics as their precursor ions, [M]⁺ for anthocyanin, [M+H]⁺ for flavonols, [M-H]^– for flavan-3-ols, phenolic acids and resveratrol. Separation was performed using relatively low volumes of formic acid (0.5%) which was not only enough to maintain the optimum pH (2.12) for anthocyanin stability and separation, but it was also more gentle on the instrumentation for long term use.

4.4.4 Method validation

The specificity, linear range and sensitivity, precision, stability, reproducibility and accuracy of the developed method were validated. The specificity of the method showed a high resolution of the extracted ion chromatograms for reference standards, and in lowbush blueberry sample as shown in Fig. 1. All calibration curves had good linearity with regression coefficient (r²) ≥ 0.99 within the test ranges. The limits of detection (LOD) and quantitation (LOQ) of the reference compounds were determined (Table 2). Results indicated good precision for all analytes with relative standard deviation (RSD) in intra- and inter-day of less than 10%, respectively. The reproducibility of all analytes was ˂ 5% (Table 2). These method validation results indicated that the newly developed LC-IT-TOF-MS method was acceptable for quantitative analysis of blueberry genotypes even when the selections featured large variability in flavonoid profiles and concentrations.

4.5 Method application for quantification of phenolics in blueberries

Sample solutions of six pink-fruited blueberry clones and lowbush blueberry were run using the LC-IT-TOF-MS validated method for simultaneous quantification of 22 anthocyanins, 4 flavan-3-ols, 6 flavonols, and 4 phenolic acids and resveratrol.  The results of this quantification were listed in Table 3.

The anthocyanin content in pink-fruited blueberries was relatively low compared to the dark blue colored lowbush blueberry. The anthocyanin content, as a sum of individual components, was 3323, 1339, 1195, 999.3, 327.4 and 169.7 µg/g for US 2235, PCH, US 2117, PLE, FLR, and US 2211, respectively, and 14529 µg/g for the lowbush blueberry (Table 3). These values were very well correlated (r=1, p<0.001) with results obtained by HPLC-UV measured by a calibration curve created with cyanidin-3-glucoside. The specificity of the LC-MS in selecting the ions in both positive and negative modes makes this method superior to HPLC-UV in quantifying very closely eluted or overlapping compounds. On the other hand, comparison between the results from each method indicated that the HPLC-UV is satisfactory for routine analysis of anthocyanins.

Four flavan-3-ol compounds were quantified in the blueberry samples. Catechin levels were higher than epicatechin in all samples (1.6 to 5.4 fold). Similarly, procyanidin B1 dimer levels exceeded the B2 procyanidin dimer in all samples (1.3 to 3.9 fold). The sum of monomeric and the sum of dimeric procyanidin were in agreement with the published data for lowbush blueberry (Gu et al., 2004). The sum of quantified of flavan-3-ol showed its highest level in PLE (589 µg/g), and the lowest was in US 2235 (289 µg/g).

Six flavonol compounds were quantified in this study, quercetin and its 3-O-glucoside + galactoside and arabinoside, and the 3-O-glucosides of myricetin, kampferol and syringetin. Quercetin-galactoside and –glucoside eluted as one peak and were quantified together as the quercetin glucoside equivalent. The tested blueberries showed variation in the concentration of individual flavonols. Kampferol-glucoside was not detected in PCH and lowbush blueberry. Syringetin-glucoside was not detected in PLE. Quercetin arabinoside was absent in PCH. The sum of the quantified flavonols showed highest levels in US 2235 and US 2117 (894 and 879 µg/g, respectively). FLR, PLE and lowbush blueberry showed comparable levels of flavonols (623, 547 and 568 µg/g, respectively). PCH contained the lowest concentration of total flavonols (162 µg/g).

Chlorogenic acid represented the major phenolic acid in all blueberries. The highest level was found in lowbush blueberry at 4150 µg/g. Pink-fruited blueberries showed ranges from 1492 µg/g for PCH to 3311 µg/g for PLE. Caffeic acid was present at comparative levels (27.1-35.8 µg/g) in PLE, FLR, US 2211, and lowbush blueberry, but was not detected in PCH, US 2235 and US 2117. 2,4-Dihydroxybenzoic acid showed a concertation range from 52.2 µg/g for US 2211 to 66.3 µg/g for PLE. Traces of gallic acid were detected in all samples. Traces of resveratrol were also detected in all samples, but at lower levels than the quantification limit.

4.6 In vitro antioxidant and anti-inflammatory activities

4.6.1 Effect of blueberry extracts on ROS and NO production

During the inflammatory reaction, effectively controlling the cellular nitric oxide (NO) production and reactive oxygen species (ROS) levels are the essential tools for inhibiting LPS induced macrophage hyper-activation and macrophage mediated acute inflammation responses (Esposito et al., 2014).To gauge the inflammatory activity, blueberry extracts were incubated with LPS-activated Raw 264.7 macrophages. When LPS was administered to macrophages, NO and ROS production levels almost evoked a nearly 2.0-fold induction versus the naive control (Fig. 2). When testing the seven blueberry genotypes at 50 µg/mL, only lowbush blueberry inhibited the production of ROS and NO (data not shown). Therefore, a concentration of 250 µg/mL was executed to allow detection of the activity of pink-fruited blueberries.  Results showed that all pink-fruited berry extracts significantly reduced the generation of reactive ROS (p<0.001) and NO production (p<0.05) with US 2211, FLR, PCH, and PLE extracts reducing the ROS levels back to non-stimulated levels at concentration 250 μg/mL. At a concentration of 50 μg/mL, ROS and NO induction were inhibited by lowbush blueberry, to baseline levels (p<0.001) (Fig. 2). Phytochemical analysis of the blueberry genotypes (Table 1 and Table 3) indicated that lowbush blueberry has between 4- to 85-fold higher anthocyanin than the pink-fruited berries, which was not the case for the concentrations of other phenolic components. The results obtained here suggested that anthocyanins play a major role in inhibiting the production of ROS and NO in LPS induced Raw 264.7 macrophages.

Examination of cytotoxicity of the extracts on macrophages by the MTT assay indicated that up to 250 μg/mL, none of the tested samples affected the viability of RAW 264.7 (data not shown). Therefore, the inhibitory effect of the blueberry extracts on the LPS induced oxidative stress (via ROS or NO) was not a result of cytotoxicity.

4.6.2 Effect of blueberry extracts on inflammatory markers

Exposure of mammalian cells to LPS leads to release of pro-inflammatory cytokines and in turn activates inflammatory cascades including cytokines, lipid mediators and adhesion molecules such as interleukin-1β (IL-1β), cyclooxygenase-2 (COX-2), inducible nitric oxide synthase (iNOS), and interleukin-6 (IL-6), which are common genetic biomarkers involved in the LPS-stimulated murine RAW 264.7 macrophage model inflammatory response (Esposito et al., 2014).

Pink-fruited berry extracts, at a concentration of 250 μg/mL, effectively suppressed  IL-1β and IL-6 (p<0.05) genes by inhibiting their expression, especially for IL-6 genes (p<0.001) based on ≥ 50% suppression relative to the LPS-stimulated controls. Among pink-fruited blueberry extracts, only PCH and US 2235 showed inhibitory effects on the expression of all genes. The latter two blueberry genotypes contained the highest levels of anthocyanin among pink-fruited clones (Table 1 and 3). This indicates that anthocyanins contribute largely to the inhibition of iNOS and IL-6. Meanwhile, lowbush blueberry extract was able to suppress significantly all the four tested pro-inflammatory markers, at a much lower concentration of only 50 μg/mL (Fig. 2).

Conclusion

In the present study, a validated HPLC-IT-TOF/MS method was employed for rapid simultaneous quantitative determination of multiple classes of phenolic compounds in blueberry samples. Thirty-seven phenolic compounds were successfully separated, identified and quantified including: anthocyanins, flavan-3-ols, flavonols, phenolic acids and resveratrol. The method enabled efficient separation of anthocyanins isomers, unique selectivity based on ESI mass ions, and sensitivity for quantifying 37 analytes belonging to 5 classes of phenolics at a wide range of concentrations. Furthermore, the method was applied successfully for the investigation of six pink-fruited blueberry clones for the first time. In vitro cell bioassays for antioxidant and anti-inflammatory activities indicated that the anthocyanin group of phenolics are responsible for the bioactivity.

References

An, H., Wang, H., Lan, Y., Hashi, Y., & Chen, S. (2013). Simultaneous qualitative and quantitative analysis of phenolic acids and flavonoids for the quality control of apocynum venetum L. leaves by HPLC-DAD-ESI-IT-TOF-MS and HPLC-DAD. Journal of Pharmaceutical and Biomedical Analysis, 85, 295-304.

Choi, S. Y., Ko, H. C., Ko, S. Y., Hwang, J. H., Park, J. G., Kang, S. H., et al. (2007). Correlation between flavonoid content and the NO production inhibitory activity of peel extracts from various citrus fruits. Biological and Pharmaceutical Bulletin, 30, 772-778.

Esposito, D., Chen, A., Grace, M., Komarnytsky, S., & Lila, M. (2014). Inhibitory effects of wild blueberry anthocyanins and other flavonoids on biomarkers of acute and chronic inflammation in vitro. Journal of Agricultural and Food Chemistry, 62, 7022-7028.

Fanali, C., Dugo, L., D’Orazio, G., Lirangi, M., Dacha, M., Dugo, P., et al. (2011). Analysis of anthocyanins in commercial fruit juices by using nano-liquid chromatography-electrospray-mass spectrometry and high-performance liquid chromatography with UV-vis detector. Journal of Separation Science, 34, 150-159.

Gavrilova, V., Kajdzanoska, M., Gjamovski, V., & Stefova, M. (2011). Separation, characterization and quantification of phenolic compounds in blueberries and red and black currants by HPLC-DAD-ESI-MSn. Journal of Agriculture and Food Chemistry, 59, 4009-4018.

Grace, M., Esposito, D., Dunlap, K., & Lila, M. (2014). Comparative analysis of phenolic content and profile, antioxidant capacity, and anti-inflammatory bioactivity in wild alaskan and commercial vaccinium berries. Journal of Agriculture and Food Chemistry, 62, 4007-4017.

Gu, L., Kelm, M. A., Hammerstone, J. F., Beecher, G., Holden, J., Haytowitz, D., et al. (2004). Concentrations of proanthocyanidins in common foods and estimations of normal consumption. Journal of Nutrition, 134, 613-617.

Hakkinen, S., Heinonen, M., Karenlampi, S., Mykkanen, H., Ruuskanen, J., & Torronen, R. (1999). Screening of selected flavonoids and phenolic acids in 19 berries. Food Research International, 32, 345-353.

Kalt, W., Ryan, D. A., Duy, J. C., Prior, R. L., Ehlenfeldt, M. K., & Vander Kloet, S. P. (2001). Interspecific variation in anthocyanins, phenolics, and antioxidant capacity among genotypes of highbush and lowbush blueberries (vaccinium section cyanococcus spp.). Journal of Agricultural and Food Chemistry, 49, 4761-4767.

Ling, Y., Ren, C., Mallery, S. R., Ugalde, C. M., Pei, P., Saradhi, U. V., et al. (2009). A rapid and sensitive LC-MS/MS method for quantification of four anthocyanins and its application in a clinical pharmacology study of a bioadhesive black raspberry gel. Journal of Chromatography B, 877, 4027-4034.

Liu, C., Ju, A., Zhou, D., Li, D., Kou, J., Yu, B., et al. (2016). Simultaneous qualitative and quantitative analysis of multiple chemical constituents in YiQiFuMai injection by ultra-fast liquid chromatography coupled with ion trap time-of-flight mass spectrometry. Molecules, 21, 640. doi: 10.3390/molecules21050640.

Lohachoompol, V., Mulholland, M., Srzednicki, G., & Craske, J. (2008). Determination of anthocyanins in various cultivars of highbush and rabbiteye blueberries. Food Chemistry, 111, 249-254.

Merken, H. M., & Beecher, G. R. (2000). Measurement of food flavonoids by high-performance liquid chromatography: A review. Journal of Agricultural and Food Chemistry, 48, 577-599.

Montoro, P., Tuberoso, C. I., Perrone, A., Piacente, S., Cabras, P., & Pizza, C. (2006). Characterisation by liquid chromatography-electrospray tandem mass spectrometry of anthocyanins in extracts of myrtus communis L. berries used for the preparation of myrtle liqueur. Journal of Chromatography A, 1112, 232-240.

Mullen, W., Larcombe, S., Arnold, K., Welchman, H., & Crozier, A. (2010). Use of accurate mass full scan mass spectrometry for the analysis of anthocyanins in berries and berry-fed tissues. Journal of Agricultural and Food Chemistry, 58, 3910-3915.

Nagy, K., Redeuil, K., Bertholet, R., Steiling, H., & Kussmann, M. (2009). Quantification of anthocyanins and flavonols in milk-based food products by ultra performance liquid chromatography-tandem mass spectrometry. Analytical Chemistry, 81, 6347-6356.

Prior, R. L., Lazarus, S. A., Cao, G., Muccitelli, H., & Hammerstone, J. F. (2001). Identification of procyanidins and anthocyanins in blueberries and cranberries (vaccinium spp.) using high-performance liquid chromatography/mass spectrometry. Journal of Agricultural and Food Chemistry, 49, 1270-1276.

Prior, R. L., Cao, G., Martin, A., Sofic, A., McEwen, J., O’Brien, C., et al. (1998). Antioxidant Capacity As influenced by total phenolic and Anthocyanin Content, maturity, and variety of vaccinium species. Journal of Agricultural and Food Chemistry, 46, 2686–2693.

Re, R., Pellegrini, N., Proteggente, A., Pannala, A., Yang, M., & Rice-Evans, C. (1999). Antioxidant activity applying an improved ABTS. radical cation decolorization assay. Free Radical Biology & Medicine, 26, 1231-1237.

Scalbert, A., Manach, C., Morand, C., Remesy, C., & Jimenez, L. (2005). Dietary polyphenols and the prevention of diseases. Critical Reviews in Food Science and Nutrition, 45, 287-306.

Shi, Y. H., Zhu, S., Ge, Y. W., Toume, K., Wang, Z., Batkhuu, J., et al. (2016). Characterization and quantification of monoterpenoids in different types of peony root and the related paeonia species by liquid chromatography coupled with ion trap and time-of-flight mass spectrometry. Journal of Pharmaceutical and Biomedical Analysis, 129, 581-592.

Tian, Q., Aziz, R. M., Stoner, G. D., & Schwartz, S. J. (2005). Anthocyanin determination in bBlack raspberry (rubus occidentalis) and biological specimens using liquid chromatography-electrospray ionization tandem mass spectrometry. Journal of Food Science, 70, C43-C47.

Wang, J., Kalt, W., & Sporns, P. (2000). Comparison between HPLC and MALDI-TOF MS analysis of anthocyanins in highbush blueberries. Journal of Agricultural and Food Chemistry, 48, 3330-3335.

Wu, X., & Prior, R. L. (2005). Systematic identification and characterization of anthocyanins by HPLC-ESI-MS/MS in common foods in the united states: Fruits and berries. Journal of Agricultural and Food Chemistry, 53, 2589-2599.

Yao, L. H., Jiang, Y. M., Shi, J., Tomas-Barberan, F. A., Datta, N., Singanusong, R., et al. (2004). Flavonoids in food and their health benefits. Plant Foods for Human Nutrition, 59, 113-122.

Yousef, G. G., Brown, A. F., Funakoshi, Y., Mbeunkui, F., Grace, M. H., Ballington, J. R., et al. (2013). Efficient quantification of the health-relevant anthocyanin and phenolic acid profiles in commercial cultivars and breeding selections of blueberries (vaccinium spp.). Journal of Agricultural and Food Chemistry, 61, 4806-4815.

Zheng, W., & Wang, S. Y. (2003). Oxygen radical absorbing capacity of phenolics in blueberries, cranberries, chokeberries, and lingonberries. Journal of Agricultural and Food Chemistry, 51, 502-509.

Table 1

Polyphenol content, anthocyanins and antioxidant capacities of blueberries

Blueberry genotype	Dry matter %	Total phenolics1	Total anthocyanins2		Total proanthocyanidins3	ABTS4	FRAP5
			HPLC-UV	LC-MS
Pink Lemonade (PLE)	19.9	15.0 ± 0.35b	0.80 ± 0.06d	1.00	3.51 ± 0.04b	50.3 ± 2.7bc	308 ± 13.5bc
Pink Champagne (PCH)	20.7	8.43 ± 0.16e	1.17 ± 0.01c	1.34	0.87 ± 0.005d	39.6 ± 1.9cd	217 ± 9.5d
Florida Rose (FLR)	17.0	14.1 ± 0.18b	0.27 ± 0.00e	0.35	3.92 ± 0.10a	51.9 ± 1.5bc	347 ± 12.9ab
US 2117	19.8	11.0 ± 0.43d	0.83 ± 0.02d	1.20	1.13 ± 0.01c	58.5 ± 2.1b	276 ± 9.4c
US 2211	18.5	7.53 ± 0.19e	0.20 ± 0.02f	0.17	0.73 ± 0.005d	32.5 ± 1.7d	149 ± 4.8e
US 2235	15.3	12.3 ± 0.22c	2.12 ± 0.07b	3.32	0.76 ± 0.02d	45.9 ± 5.2c	299 ± 9.4bc
Lowbush	15.0	24.5 ± 0.69a	12.6 ± 0.15a	14.5	3.47 ± 0.06b	127 ± 5.3a	389 ± 19.4a

¹Total phenolics quantified by Folin Ciocalteu assay as mg gallic acid equivalent/g; ²total anthocyanins measured by HPLC quantified as mg cyanidin-3-glucoside equivalent/g, LC-MS quantification (mg/g) refers to Table 3; ³total proanthocyanidins quantified by DMAC assay as mg procyanidin B2 equivalent/g; ⁴radical scavenging activity by ABTS as µmol Trolox equivalent/g; ⁵antioxidant capacity by FRAP assay as µmol FeSO₄ equivalent/g. Results were expressed as mean ± SD (n=3). All concentrations were calculated based on dry weight. Means with different superscript letters within the same column are significantly different (p<0.05).

Table 2

Method validation for simultaneous quantification of phenolic compounds in blueberries

Phenolic standards*	Regression equations	r2	Linear range (ppm)	LLOQ/ LLOD (ppm)	Precision RSD (%)		Repeatability RSD (%)
					Intra-day	Inter-day
Anthocyanins
Dp-3-gal	Y = 1.1111X – 0.0038	0.9997	0.4-40	1.20/0.40	0.60	3.89	0.34
Dp-3-glc	Y = 0.6237X – 0.0554	0.9998	0.2-40	0.60/0.20	1.15	5.94	3.10
Cyn-3-gal	Y = 1.0376X + 0.0311	0.9998	0.06-40	0.18/0.06	2.35	4.74	2.26
Cyn-3-glc	Y = 1.9267X + 0.0190	1.0000	0.04-40	0.12/0.04	1.22	8.73	0.53
Cyn-3-ara	Y = 1.1997X – 0.0159	0.9996	0.06-40	0.18/0.06	1.81	2.79	5.53
Pet-3-glc	Y = 1.8492X + 0.0353	0.9998	0.06-40	0.18/0.06	0.21	5.54	1.10
Peo-3-glc	Y = 2.5897X + 0.0744	0.9979	0.04-10	0.12/0.04	4.39	5.72	4.29
Mv-3-gal	Y = 1.2713X + 0.0149	0.9998	0.04-20	0.12/0.04	0.80	2.15	0.83
Mv-3-glc	Y = 1.7519X + 0.0420	0.9995	0.04-20	0.12/0.04	2.09	1.50	2.85
Flavan-3-ols
Procyanidin B1	Y = 0.7880X – 0.0087	0.9995	0.06-10	0.12/0.06	0.50	6.59	3.67
Catechin	Y = 1.7000X + 0.0253	0.9991	0.02-10	0.06/0.0.2	5.21	5.95	0.87
Procyanidin B2	Y = 0.7561X – 0.0050	1.0000	0.04-10	0.12/0.04	0.59	8.90	4.07
Epicatechin	Y = 1.8823X + 0.0374	0.9999	0.04-20	0.12/0.04	4.73	6.26	0.99
Flavonoid
Myricetin-3-glc	Y = 0.2363X – 0.0065	0.9983	0.2-10	0.60/0.20	2.43	3.14	4.97
Quercetin-3-glc	Y = 0.3534X – 0.0057	0.9982	0.04-20	0.12/0.04	5.32	3.32	0.36
Quercetin-3-ara	Y = 0.4187X + 0.0040	0.9997	0.01-20	0.03/0.01	0.39	4.43	2.64
Kaempferol-3-glc	Y = 0.3589X + 0.0096	0.9985	0.04-10	0.12/0.04	1.97	3.02	2.66
Syringetin-3-glc	Y = 0.2655X + 0.0100	0.9959	0.02-10	0.06/0.02	0.35	3.04	5.23
Quercetin	Y = 0.8367X – 0.0269	0.9993	0.01-40	0.03/0.01	1.28	1.67	5.31
Phenolic acid
Gallic	Y = 0.5952X + 0.0057	0.9999	0.04-40	0.12/0.04	2.17	2.26	3.62
2,4 dihydroxybenzoic	Y = 1.1111X – 0.0038	0.9999	0.02-10	0.06/0.02	1.88	5.86	5.23
Caffeic	Y = 1.5040X + 0.1045	0.9981	0.08-10	0.24/0.08	5.41	6.29	6.42
Chlorogenic	Y = 2.0099X + 0.0391	0.9989	0.01-10	0.03/0.01	5.31	2.26	1.10
Resveratrol	Y = 0.1976X + 0.0003	0.9972	0.01-10	0.03/0.01	0.64	5.52	1.53

*Dp = delphinidin; Cyn = cyanidin; Pet = petunidin; Peo = peonidin; Mv = malvidin; gal = galactoside; glc = glucoside; ara = arabinoside, RSD = relative standard deviation

Table 3

Quantitative analytical results of anthocyanins and other phenolic metabolites in six-pink fruited blueberries and lowbush blueberry by LC-MS (µg/g DW)

	Phenolic compound	Pink-Lemonade (PLE)	Pink-Champagne (PCH)	Florida Rose (FLR)	US 2117	US 2211	US 2235	Lowbush
	Anthocyanins
1	Dp-3-gal*	265.1	163.3	63.6	194.1	28.4	223.8	1017
2	Dp-3-glc*	52.3	182.2	48.6	28.5	(BQL)	465.9	3134
3	Cyn-3-gal*	133.5	60.8	44.6	67.6	(BQL)	75.7	858.2
4	Dp-3-ara	418.7	381.5	120.8	397.2	73.7	458.1	1266
5	Cyn-3-glc*	(BQL)	21.3	(BQL)	(BQL)	—–	33.2	476.0
6	Cyn-3-ara*	129.7	74.9	36.7	65.9	(BQL)	978.0	648.8
7	Pet-3-gal	(BQL)	(BQL)	(BQL)	88.7	(BQL)	109.7	415.7
8	Pet-3-glc*	(BQL)	55.8	(BQL)	(BQL)	…	151.4	820.2
9	Peo-3-gal	(BQL)	(BQL)	(BQL)	(BQL)	(BQL)	(BQL)	136.1
10	Pet-3-ara	(BQL)	(BQL)	(BQL)	92.7	(BQL)	136.8	332.3
11	Peo-3-glc*	—–	(BQL)	(BQL)	—–	—–	29.9	330.3
12	Mv-3-gal*	(BQL)	135.7	13.1	117.9	67.6	228.8	1030
13	Mv-3-glc*	—–	54.7	(BQL)	(BQL)	—–	179.6	1028
14	Mv-3-ara	(BQL)	208.7	(BQL)	142.0	(BQL)	252.5	895.1
15	Dp-3-(ac-glc)	—–	(BQL)	—–	—–	—–	(BQL)	772.5
16	Pet-3-(ac-gal)	—–	—–	—–	—–	—–	—–	(BQL)
17	Peo-3-(ac-gal)	—–	—–	—–	—–	—–	—–	(BQL)
18	Cyn-3-(ac-glc)	—–	(BQL)	—–	—–	—–	(BQL)	242.0
19	Mv-3-(ac-gal)	—–	(BQL)	—–	(BQL)	(BQL)	—–	268.7
20	Pet-3-(ac-glc)	—–	(BQL)	—–	—–	—–	(BQL)	268.1
21	Peo-3-(ac-glc)	—–	(BQL)	—–	—–	—–	..	(BQL)
22	Mv-3-(ac-glc)	—–	(BQL)	—–	(BQL)	—–	(BQL)	589.6
	Total	999.3	1339	327.4	1195	169.7	3323	14529
	Flavan-3-ols
23	Procyanidin B1*	207.2	82.6	125.2	177.9	102.2	98.2	125.1
24	Catechin*	259.9	131.1	202.9	203.9	179.1	139.3	117.5
25	Procyanidin B2*	65.9	54.3	98.7	51.0	30.0	25.1	72.0
26	Epi-catechin*	55.6	53.5	109.9	57.1	40.3	26.0	74.6
	Total	588.6	321.5	536.7	489.9	351.6	288.6	389.2
	Flavonols
27	Myrcetin-glc*	121.8	71.9	65.2	172.2	29.1	26.5	130.5
28	Quercetin-glc/gal*	27.1	74.0	15.9	248.9	75.4	107.1	189.8
29	Quercetin-ara*	140.4	—–	82.6	167.4	60.2	275.4	53.3
30	Kampferol-glc*	242.5	—–	441.0	271.6	79.5	475.3	—–
31	Syringetin-glc*	—–	10.0	10.9	5.65	7.9	4.17	181.8
32	Quercetin*	15.4	5.6	7.52	13.1	7.8	5.72	12.2
	Total	547.2	161.5	623.12	878.85	259.9	894.19	567.6
	Phenolic acids
33	Gallic*	5.4	3.3	2.80	2.60	0.05	1.47	3.51
34	2.4 Hyroxybenzoic*	66.3	36.7	62.6	50.7	52.2	39.4	72.7
35	Caffeic*	35.8	—–	29.0	—–	27.1	—–	23.7
36	Chlorogenic*	3311	1492	2698	2586	2088	3020	4150
37	Resveratrol*	0.4	0.1	0.47	0.33	0.12	0.68	1.15
	Total phenolics	5554	3354	4280	5203	2949	7568	19737

* Reference standards available. Dp = delphinidin; Cyn = cyanidin; Pet = petunidin; Peo = peonidin; Mv = malvidin; gal = galactoside; glc = glucoside; ara = arabinoside; Anthocyanin compounds, with no reference standard available, were quantified as the closest anthocyanin compound having the same aglycone. BQL= detected but below quantification limit

Figure Captions

Fig. 1 The extracted ion chromatograms (EIC) of phenolic metabolites by LC-IT-TOF-MS showing a mixture of 24 standard references (a), and of all compounds quantified in lowbush wild blueberry (b).

Fig. 2 Effect of blueberry extracts on reactiveoxygen species (ROS) and nitric oxide (NO) production (A and B) and on proinflammatory gene expression (C-F) in the LPS-stimulated RAW264.7 macrophage cells. Cells were treated with pink blueberry extracts (250 µg/mL) and lowbush blueberry (50 µg/mL) for 1 h, and stimulated with lipopolysaccharide (LPS, 1 μg/mL), then incubated for 18 h. Changes in gene expression were measured by comparing mRNA quantity relative to LPS. Vehicle values were obtained in the absence of LPS or test samples; dexamethasone (DEX) was used as positive control at a concentration of 10 μM. Results were expressed as means ± SEM, n = 2 experiments. *p < 0.05, **p < 0.01, ***p < 0.001 ****p < 0.0001 vs. the LPS treated group. One-way ANOVA, Dunnett’s post hoc test. PLE = Pink Lemonade; PCH = Pink Champagne; FLR = Florida Rose.