Accepted Manuscript

Title: Characterization and Quantification of Flavonoids and

Organic Acids over Fruit Development in American Cranberry
(Vaccinium macrocarpon) Cultivars using HPLC and
APCI-MS/MS

Authors: Yifei Wang, Jennifer Johnson-Cicalese, Ajay P.

Singh, Nicholi Vorsa

PII: S0168-9452(17)30311-4
DOI: http://dx.doi.org/doi:10.1016/j.plantsci.2017.06.004
Reference: PSL 9617

To appear in: Plant Science

Received date: 11-4-2017

Revised date: 16-5-2017
Accepted date: 6-6-2017

Please cite this article as: Yifei Wang, Jennifer Johnson-Cicalese, Ajay P.Singh, Nicholi
Vorsa, Characterization and Quantification of Flavonoids and Organic Acids over Fruit
Development in American Cranberry (Vaccinium macrocarpon) Cultivars using HPLC
and APCI-MS/MS, Plant Sciencehttp://dx.doi.org/10.1016/j.plantsci.2017.06.004

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
Characterization and Quantification of Flavonoids and Organic Acids over Fruit

Development in American Cranberry (Vaccinium macrocarpon) Cultivars using HPLC and

APCI-MS/MS

Yifei Wang1, Jennifer Johnson-Cicalese2, Ajay P. Singh1 and Nicholi Vorsa1, 2

1
Department of Plant Biology and Pathology, Rutgers University, New Brunswick, NJ, USA

2
Philip E. Marucci Center for Blueberry and Cranberry Research and Extension, Rutgers

University, Chatsworth, NJ, USA

Corresponding author: Dr. Nicholi Vorsa, Tel: (609) 726-1590 ext. 11; Fax: (609) 726-1593

Email: vorsa@aesop.rutgers.edu

Graphical abstract
Highlights:

 Quercetin-3-galactoside is the most abundant flavonol in all cranberry cultivars.

 Levels of major flavonols remain consistent during fruit development and ripening.

 Levels of proanthocyanidins declined during cranberry fruit development.

 The effect of cultivar was evident in proanthocyanidin and anthocyanin levels.

Abstract

Cranberry flavonoids, including anthocyanins, flavonol glycosides and proanthocyanidins, and

organic acids were characterized and quantified by HPLC and LC-MS/MS during fruit

development and ripening in eight cranberry cultivars. Anthocyanin biosynthesis initiated at

early fruit development and reached highest level in mature fruit, with significant differences

between cultivars. Major flavonol glycosides, including the most abundant quercetin-3-

galactoside and myricetin-3-galactoside, showed consistent concentrations during the season

with moderate fluctuation, and were at similar levels in mature fruits of the eight cultivars.

Proanthocyanidins declined during fruit development and then increased slightly in later

maturation stages. Levels of various proanthocyanidins oligomers/polymers with different

degree-of-polymerization were highly correlated within a cultivar during fruit development.

Cultivars with coancestry exhibited similar levels (high/low) of anthocyanins or

proanthocyanidins, indicating genetic effects on biosynthesis of such flavonoids. All cultivars

showed similar levels of malic and citric acids, and declining levels of quinic acid during fruit

development. Benzoic acid was extremely low early in the season and increased sharply during

fruit ripening. Levels of quinic and citric acids were significantly different among cultivars in the
mature fruit. Concentrations of proanthocyanidins, anthocyanins, quinic acid and benzoic acid

have a strong developmental association in developing ovaries.

Keywords: Cranberry, Anthocyanins, Flavonols, Proanthocyanidins, Organic acids, Fruit

development

1. Introduction

American cranberry (Vaccinium macrocarpon) has long been recognized as one of the leading

plant sources of bioactive secondary metabolites with important implications to human health.

Among cranberry secondary metabolites, phenolic compounds have received most of the

research interest due to their high contents in cranberry and well-proven health benefits.

Cranberry phenolics are comprised of flavonoids, including flavonols, anthocyanins and flavan-

3-ols (proanthocyanidins), as well as phenolic acids [1]. Besides phenolics, cranberry also

contains various non-phenolic organic acids such as quinic, citric and malic acids, and small

amount of benzoic and glucuronic acids [2].

Cranberry anthocyanins and flavonols mainly appear as monomeric glycosides conjugated to a

number of different sugar moieties [3,4]. Cranberry proanthocyanidins (PACs), on the other hand,

are oligomers or polymers of flavan-3-ols. While the most common linkage between PAC

building blocks is C-C bond (B-type) [5], in cranberry and few other plant species, a C-O-C bond

also occurs and forms a double linkage (A-type) with the C-C bond [6], making them structurally

unique compared with B-type PACs found in blueberry, grape and cocoa. Both flavonols and
PACs have shown various bioactivities, such as anti-oxidant, anti-cancer and anti-inflammatory

activities, as well as cardiovascular health benefits [7-13].

Organic acids are important components in cranberry fruits and contribute to their characteristic

flavor. Besides being employed to determine fruit quality such as maturity and flavor [14,15],

they also have been widely used as food additives in juice and beverage products to improve

nutrition and preservation stability [16]. Compositions of flavonoids and organic acids in

cranberry fruits can be influenced by multiple factors such as cultivar, maturity and growth

conditions. Although early studies revealed composition patterns of different phytochemicals in

cranberry fruit and other species [15,17-19], diversity of plant materials and analyzed

compounds were limited. Comprehensive studies with extensive sampling and analytical

approaches are needed for a better understanding of how multiple factors, including genetic

background, growth stage, and biosynthetic pathway, can affect phytochemical profiles of

cranberry.

In the present study, we investigated the concentrations of four major classes of cranberry

phytochemicals - anthocyanins, flavonols, PACs and organic acids, during fruit development and

ripening in eight cranberry cultivars representing diverse genetic backgrounds. HPLC and LC-

MS/MS methods have been developed and optimized to identify individual cranberry

phytochemicals and provide accurate quantitative analysis. The objectives were to determine: (1)

whether flavonoid and organic acid constituents and their levels are a function of fruit

development stages, (2) if there is evidence for genetic variation, and (3) which cultivars have

higher levels of phytochemical constituents.

2. Materials and methods

2.1. Plant material

The eight cranberry cultivars used were: selections domesticated from endemic cranberry

populations - Early Black (EB) selected in 1852, Howes (HO) selected in 1843, and Ben Lear

(BL) selected in 1910; 1 st breeding cycle cultivars Stevens (ST) and #35, developed in 1940s;

and 2nd breeding cycle cultivars Crimson Queen (CQ), Demoranville (DM) and Mullica Queen

(MQ), released post 2000. Their genetic relationship and coancestry is summarized in Fig. 1. The

cultivar trial was established at Philip E. Marucci Center for Blueberry and Cranberry Research

and Extension in May 2010; plots were in a randomized complete block design (RCBD) with 3

replications (blocks). Fruit samples were harvested from each plot at 7-17 day intervals over 8

dates in 2014, initiating at mid to late fruit set on Jul 21 through fruit maturity on Oct 6.

Sampling dates were Jul 21 (Julian Day (JD) 202), 28 (JD 209); Aug 7 (JD 219), 18 (JD 230), 28

(JD 240); Sep 8 (JD 251), 19 (JD 262); and Oct 6 (JD 279). Samples were placed in a -20 ºC

freezer within 1 hour of collection and remained there until analysis. For phytochemical analysis,

samples of each replicate (block) were analyzed as a set (8 dates × 8 varieties = 64 samples/set);

Note: replicate variation includes laboratory experimental variation of the set.

2.2. Reagents

All solvents were purchased from EMD Millipore (Billercia, MA). Acetic acid was purchased

from Avantor Performance Materials (Center Valley, PA), formic acid was purchased from

Mallinckrodt Baker (Phillipsburg, NJ) and phosphoric acid was purchased from Amresco (Solon,

OH). Sephadex® LH-20 was obtained from GE Healthcare Bio-Science (Piscataway, NJ) and

BAKERBOUND® Diol was obtained from Avantor Performance Materials.

The flavonoid standards, including the anthocyanin cyanidin-3-galactoside, and flavonols

quercetin-3-galactoside, quercetin-3-glucoside, quercetin-3-rhamnoside and quercetin aglycone

were purchased from Indofine Chemical Company (Somerville, NJ). Other cranberry flavonol

standards were isolated using a semi-prep HPLC system described by Singh et al [20]. Citric acid,

malic acid, quinic acid and benzoic acid standards were obtained from Sigma (St. Louis, MO).

Individual PAC standards were isolated using Sephadex® LH-20 and BAKERBOUND® Diol

gravity column chromatography [20].

2.3. Extraction of cranberry flavonoids and organic acids

For flavonoid quantification, about 10 g of fruits were weighed and crushed in 40 ml 80%

aqueous acetone with 0.1% acetic acid in laboratory blender for 1 minute, followed by 30 min

sonication and filtration with filter paper. Liquid extracts were dried in rotary evaporator under

high vacuum and 35°C water bath and re-dissolved in methanol to a final volume of 6 ml.

Samples were filtered through 0.45 µm Spin-X® centrifuge filter tube before analysis. For

flavonoid characterization and isolation, acetone extracts were further partitioned with n-hexane

and ethyl acetate and pre-purified in Sephadex® LH-20 column as previously described [20].

For organic acid identification and quantification, about 3 g of fruit were weighed and crushed in

30 ml distilled water in a blender for 1 minutes and heated in 90 ºC water bath for 10 min. After

filtration through filter paper, aqueous extracts were collected for HPLC analysis.

2.4. HPLC Apparatus and Conditions

Cranberry flavonols were analyzed in Dionex UltiMate® 3000 LC system. A Gemini® 150 x 4.6

mm C18 110 Å, 5 µm LC column was used and flavonols were detected at 366 nm. Cranberry

PACs were analyzed in Waters Alliance® LC system. A Develosil® 250 x 4.6 mm 100Diol-5,
5µm LC column was used and PACs were detected in both PDA detector at 280 nm and

fluorescence detector with excitation/emission wavelengths at 280/308 nm. Organic acids were

analyzed in a Dionex® HPLC system with AS50 Autosampler, AS50 Thermal Compartment,

PDA-100 Detector and GP40 Gradient Pump. A Waters Atlantis® 250 x4.6 mm dC18, 5µm LC

column was used and organic acids were detected at 210 to 230 nm in PDA detector. All solvent

systems and elution gradients are summarized in Table 1.

2.5. MS Spectrometry

An Applied Biosystems API 3000TM triple-quad LC-MS/MS mass spectrometer coupled with the

Dionex UltiMate® 3000 LC system was used in LC/MS-MS analysis. MS data was obtained

under atmospheric pressure chemical ionization (APCI) in negative ion detection mode, with

following parameters: Curtain gas: 12 psi, Nebulizer gas: 7 psi, Nebulizer current: -2.0 mA,

Entrance potential: -10 V, Focusing potential: -300 V, Declustering potential: -60 V, Collision

energy: -50 V, Collision cell exit potential: -5.0 V, Source temperature: 500 ºC. The same LC

methods were used for flavonol and PAC separation.

2.6. Characterization and quantification of flavonols, proanthocyanidins and organic acid

Characterization of specific compounds was carried out by comparing their HPLC peak retention

time, UV-Vis absorbance spectra, fluorescence spectra and/or MS data with those of standards.

Due to the detection limit of APCI-MS, only PAC oligomers with degree-of-polymerization less

than 8 can be identified by mass spectra. PAC polymers greater than heptamers were identified

by comparing their HPLC peak retention time patterns with previously published data [21]. For

compound quantification, HPLC standard curves were first prepared using purchased or isolated
standards. Flavonols, PACs (oligomers and polymers) and organic acid were then quantified by

calculating peak area corresponding to relevant standard curves.

2.7. Quantification of total monomeric anthocyanins

Anthocyanins were quantified using a pH differential spectrophotometric method [22] with slight

modification. Cranberry flavonoid extracts (50 to 200 μl) were mixed with pH 1.0 (0.025M

potassium chloride) or pH 4.5 (0.4M sodium acetate) buffer (1800 to 1950 μl depending on the

volume of added flavonoid extract). Absorbance at 520 and 700 nm were measured by a Thermal

Scientific GenesysTM 10S UV-Vis spectrophotometer for both mixtures and the subtracted

absorbance A= (A520-A700)pH 1.0-(A520-A700)pH 4.5 was used for quantification. Cyanidin-3-

galactoside was used as the standard.

2.8. Statistical analysis

Statistical analysis was performed with SAS 8.0 (SAS Institute, Inc., Cary, NC) and SPSS

Statistics 19 (IBM Corporation, Armonk, NY). To evaluate differences between cultivars, SAS

procedure PROC GLM was used for analysis of variance (ANOVA) and LSD test for mean

separation. SAS procedure PROC COMP was used for principal component analysis (PCA) of

the flavonol and PAC concentration data on the last harvest date. Compounds analyzed in the

PCA include: (1) Flavonols: myricetin-3-galactoside, myricetin-3-arabinofuranoside, quercetin-

3-galactoside, quercetin-3-glucoside, quercetin-3-xylopyranoside, quercetin-3-arabinopyranoside,

quercetin-3-arabinofuranoside and quercetin-3-rhamnopyranoside; (2) PACs: Epicatechin

(monomer), dimers, trimers, tetramers, pentamers, hexamers, heptames to decamers and

polymers greater than decamer. Linear/quadratic regression analysis was performed in SPSS
Statistics 19 to evaluate the relationship between flavonoid levels and sampling date (Julian day

used in analysis).

3. Results

3.1. Total Monomeric Anthocyanins

Levels of total monomeric anthocyanins in eight cranberry cultivars throughout fruit

development are shown in Fig. 2A. Low concentrations (0.12-0.32 mg/100g fruit) of

anthocyanins were detected in all cultivars at fruit set (Jul 21). Levels of anthocyanins increased

slowly during fruit development and then increased substantially in ripening fruit after Sep 8.

Highest levels of anthocyanins were observed at the last sampling date in all cultivars.

Cultivars DM, EB, CQ and BL exhibited higher anthocyanin levels than MQ, ST, #35 and HO

during the entire sampling period, having significantly higher anthocyanin (54-86 mg/100g fruit)

on October 6, compared to the other 4 cultivars (12-38 mg/100g) (F=17.51; P≤0.0001). The

effect of pedigree in anthocyanin levels was evident: BL had high anthocyanin, as did its

offspring DM and CQ; while HO, and HO offspring #35 and MQ had low anthocyanin levels

(Fig. 1). Accumulation of total anthocyanins exhibited a strong relationship with harvest date

which accounted for the majority of anthocyanin variation for all cultivars (R2 > 90% in

quadratic fit; R2 > 62 in linear fit) (Supplementary Table 1).

In contrast to anthocyanin accumulation, the size and weight of cranberry fruit increased

substantially during fruit development between Jul and Aug, then slowly increased or remained

at the same weight during fruit ripening (Fig. 2B). Different cultivars also differed in cranberry

weight. On the last harvest date, CQ had largest mature fruits with average weight at 3.7 g. HO

and EB, in contrast, had smallest mature fruits with average weights at 1.6 and 1.5 g.
3.2. Flavonol Glycosides

HPLC-APCI-MS/MS analysis was used to identify individual flavonols in cranberry fruits. Fig. 3

and Table 2 show the HPLC chromatograph, retention times, UV spectra and mass fragmentation

of isolated cranberry flavonols. Peaks 1 and 2 had molecular ion peaks at m/z 479 and 449

respectively, and they both showed MS-MS fragment of m/z 317, which corresponds to

myricetin aglycone. They were identified as myricetin-3-galactoside (M-3-Gal) and myricetin-3-

arabinofuranoside (M-3-A) respectively, based on comparison with previously published data

[23]. Peaks 3 to 8 all showed MS-MS fragment at m/z 301, which corresponds to quercetin

aglycone. Their molecular ion peaks, at m/z 463, 433 or 449, identified them as quercetin-

hexoside (463), quercetin-pentoside (433) or quercetin-methyl-pentoside (449). By comparing

the retention time and MS-MS fragmentation pattern with reference standards or previously

published data [23], they were identified as quercetin-3-galactoside (Q-3-Gal), quercetin-3-

glucoside (Q-3-Glu), quercetin-3-xylopyranoside (Q-3-X), quercetin-3-arabinopyranoside (Q-3-

A-P), quercetin-arabinofuranoside (Q-3-A-F), and quercetin-3-rhamnopyranoside (Q-3-R). Peak

9 had molecular ion peak at m/z 301 and MS-MS fragment of m/z 179, indicating its structure as

quercetin aglycone, which was confirmed by authentic standard. The structures of identified

cranberry flavonols are shown in Fig. 4A-I. Despite being identified in Sephadex® LH-20

column purified flavonol fraction, quercetin aglycone appeared in extremely low levels in

cranberry fruit and was not quantified in the current study.

Q-3-Gal was the most abundant flavonol glycoside in all cultivars, accounting for 31%-46%

(w/w) of total quantified flavonols, followed by M-3-G at 19%-32%, Q-3-A-F at 7%-17% and

Q-3-R at 7%-14%. Overall, there were relatively minor fluctuations in the levels of individual

flavonol glycosides over the course of the season (Fig. 5A). All cultivars showed similar levels
of individual or total flavonols at the final harvest date, Oct 6. However, when the eight

cranberry cultivars were compared with each other over the fruit development and ripening

stages, significant differences were found among them for both total and individual flavonols

(Fig. 5B, Fig. 6A-H).

For Q-3-Gal, significant differences among cultivars were observed on three sampling dates (Fig.

6A): MQ showed significantly higher Q-3-Gal level than DM at fruit set, Jul 21 (F=4.05;

P≤0.0098); BL and ST exhibited higher Q-3-Gal concentrations than MQ, EB, CQ and DM on

August 7 (F=6.64; P≤0.001); and BL again showed higher Q-3-Gal concentrations than others on

September 19 (F=4.24; P≤0.0081). Except for those dates, all 8 cultivars had similar levels of Q-

3-Gal during sampling period, including the final harvest date. In all cultivars, no significant

linear/quadratic relationship with date was evident (Supplementary Table 1).

From Jul 21 to Sep 8, from fruit set to the beginning of the fruit ripening, significant differences

on M-3-Gal level were observed: HO and ST exhibited higher M-3-Gal levels than other

cultivars, and BL had less M-3-Gal (Fig. 6B). Similar to Q-3-Gal, M-3-Gal concentrations

exhibited no significant relationship with date in all cultivars.

Q-3-A-F, the third most abundant flavonol glycoside exhibited a significant linear (ST and #35)

or quadratic (MQ, DM, EB, and HO) relationship with date in 6 of the 8 cultivars (Fig. 6C), in

which sampling date accounted for 67%-93% of the variation (Supplementary Table 1). All

cultivars exhibited highest levels of Q-3-A-F between late Jul and early August, early fruit

development, and then declined until late September, when Q-3-A-F concentration started to

increase again until final harvest (Fig. 6C).

Concentrations of Q-3-R were significantly different among the 8 cultivars throughout the

sampling period, except on the final harvest date, October 6. ST had highest Q-3-R levels

throughout entire sampling period, with an average concentration of 6.03 mg/100g fruit. DM, on

the other hand, showed lowest concentrations of Q-3-R during most of the sampling period (Fig.

6D).

Significant differences among cultivars in M-3-A concentrations were observed from Jul to

August, fruit set to mid-fruit development stages. However, for the last two sampling dates,

representing fruit maturity, all cultivars exhibited similar levels of M-3-A. HO had the highest

M-3-A levels during entire sampling period (Fig. 6E).

Two cultivars, MQ and EB, showed a largely quadratic relationship between their Q-3-X

concentrations and date, whereas for ST the relationship was largely linear, while cultivars DM,

CQ, #35, BL and HO did not exhibit a significant relationship between Q-3-X and date

(Supplementary Table 1). Levels of Q-3-X increased during the last two harvest dates (Sep 19

and Oct 6), in all cultivars except for BL. BL had high Q-3-X on Sep 19, and then a sharp

decline Oct 6 (Fig. 6F).

Differences between cultivars in Q-3-A-P levels were seen on half of the sampling dates (Jul 21,

Aug 7, Aug 28, Sep 19). ST showed relatively higher concentrations of Q-3-A-P from late Jul to

early August, during early fruit development, and reduced Q-3-A-P levels during final fruit

development and early ripening stages, until late September. DM, on the other hand, had lowest

concentrations of Q-3-A-P during most of the sampling period, with Q-3-A-P concentrations

frequently less than 1 mg/100g fruit (Fig. 6G).

All 8 cultivars exhibited significant linear relationships between Q-3-Glu concentrations and date

(Supplementary Table 1). In general, concentrations of Q-3-Glu increased during fruit

development and maturation, from an average 0.8 mg/100g fruit at fruit set (Jul 21) to 1.8

mg/100g fruit on the last harvest date (Oct 6) (Fig. 6H). Except Aug 28 and Sep 8, all cultivars

exhibited similar levels of Q-3-Glu.

To evaluate if genetic background is associated with flavonol level and/or profiles, principal

component analysis (PCA) was performed on the flavonol concentration data from the Oct 6

harvest. The first two vectors have eigenvalues accounting for 68% and 19% of the total variance

for flavonol concentrations. Fig. 7A shows the plot of PC scores and eigenvectors of the first two

PCs on the 8 cranberry cultivars, each with 3 replications.

PC1, with eigenvectors between 0.29 and 0.40 for each flavonol glycoside, represents the overall

level of total flavonols. Although EB had lower amount of flavonols than others and

correspondingly lower PC1 scores in two replicates, replicates of the other cultivars were

generally scattered across the PC1 score, indicating that cultivar differences only account for a

small percentage of total flavonol variation and other factors such as environmental variation

may have a heavier impact. PC2, with considerable negative eigenvectors for M-3-Gal, M-3-A

and Q-3-Glu and positive eigenvectors for Q-3-Gal, Q-3-X ad Q-3-A-F, represents the variation

of those compounds. The correlation between different flavonols (Supplementary Table 2A) also

revealed a close connection of these flavonol glycosides. PC2 scores of different cultivars

suggest varietal variation on those specific flavonols. Three HO replicates had low, negative PC2

scores. Indeed, this cultivar had highest levels of M-3-Gal, M-3-A and Q-3-Glu and lowest levels

of Q-3-X and Q-A-F among eight cultivars at Oct. 6 (Fig. 6). On the other hand two BL

replicates showed high, positive PC2 scores, consistent with the observation that BL had lower
levels of M-3-Gal, M-3-A and Q-3-Glu (lowest), and higher levels of Q-3-Gal, Q-3-X and Q-3-

A-F (highest) among eight cultivars (Fig. 6). In addition, we did not observe clustered plots of

the cultivars based on genetic relationships, such as CQ or DM with BL, or MQ with 35 (Fig. 1).

This observation is consistent with the fact that no significant difference was observed on the

levels of individual or total flavonols among 8 cultivars harvested at Oct 6.

3.3. Proanthocyanidins

Cranberry PAC oligomers and polymers were analyzed by LC-MS and HPLC with fluorescence

detection. As shown in Fig. 8 and Table 3, all PACs exhibited fluorescence excitation/emission

at 280/308 nm. Peak 1 showed molecular ion at m/z 575.4, corresponding to a PAC dimer (DP-2)

which contains two building units (epicatechin/catechin, molecular weight: 290), and one double

inter-flavan linkage (A-type). Its structure is shown in Fig. 4N. Peaks 2 and 3 showed molecular

ions at m/z 864.0 and 864.1, indicating their identities as A-type PAC trimers (DP-3) with one

double inter-flavan linkage (Fig. 4O). Similarly, peaks 4 and 5 were identified as A-type PAC

tetramers (DP-4) with one or two double linkages by their molecular ions at m/z 1149.7 and

1151.7; peaks 6 and 7 were identified as A-type PAC pentamers (DP-5) with one or two double

linkages (m/z 1438.2 and 1440.4), peaks 8 and 9 as A-type hexamers (DP-6) with one double

linkage (m/z 1729.0 and 1729.4) and peak 10 as A-type PAC heptamer (DP-7) with two double

linkages (m/z 2014.2).

Individual cranberry PAC monomers and polymers were quantified by HPLC with fluorescence

detection, including single compounds from epicatechin to DP-6, and PAC combinations of DP-

7,8; DP-9,10 and DP-11+ (PACs with degree-of-polymerization greater than 10). Fig. 9A shows

the concentrations of total PACs over the season in eight cranberry cultivars. During entire
sampling period, HO exhibited highest levels of PACs, with average concentration of 594

mg/100g fruit. CQ and BL showed lowest amount of PACs, with average PAC concentrations of

331 and 337 mg/100g fruit, respectively. The concentrations of individual PACs are shown in

Fig. 9B-F. DP-11+ accounted for 45%-55% (w/w) of total PACs, followed by DP-4 (14%-18%),

DP-9,10 (10%-12%), DP-6 (7%-8%), DP-3 (6%-8%), DP-7,8 (3%-5%), DP-2 (2%-3%), DP-5

(1%-2%) and epicatechin (<1%). In all cultivars, levels of PACs declined during fruit

development, and increased after late September during fruits’ final ripening. As a result,

significant linear/quadratic relationships were observed between PAC concentrations and date in

all cultivars (Supplementary Table 1).

Individual PAC compounds tended to share similar seasonal concentration patterns within a

cultivar. This is in contrast to the individual flavonol glycosides, where individual compounds

exhibited very different patterns within a cultivar. However, also in contrast to the flavonols,

significant differences were found between cultivars, for total PAC levels and all individual

PAC’s, on all eight evaluation dates. For example, HO possessed the highest levels of total and

individual PACs among cultivars. Concentrations of PACs in HO continuously declined from

fruit set (Jul 21) to early fruit development stage (Aug 18), increased between Aug 18 and Aug

28, further declined during ripening (Sep 19), and then stayed at similar levels at the later fruit

ripening stage (Fig. 9B).

EB and #35 exhibited similar PAC concentrations over the season, which were higher than ST,

MQ, DM, BL and CQ. In EB, concentrations of most PACs declined by 34%-63% from fruit set

(Jul 21) to fruit development (Sep 8), and slightly increased until the last harvest date. DP-6,

however, first increased during early fruit development between July 21 and July 28 and then

behaved similarly to the other PACs (Fig. 9C). In #35, PACs declined gradually by 47%-57%
from fruit set (Jul 21) to the final fruit development stage (Sep 19), and then slightly increased in

the mature fruit (Fig. 9D).

MQ, ST and DM exhibited similar PAC concentrations and seasonal patterns. Levels of

individual PACs sharply declined during early fruit development between Jul 28 and Aug 7 in

the three cultivars, and increased during the last sampling period. Fig. 9E shows the PAC

concentration pattern in MQ during fruit development and ripening (ST and DM data not shown).

BL and its offspring CQ contained the lowest levels of total and individual PACs. In BL,

concentrations of individual PACs showed minor changes between the first two sampling periods

(Jul 21-Jul 28), declining levels until Sep 19, and then minor changes on the last sampling date

(Oct 6) (Fig. 9F). CQ showed consistent declining PAC concentrations from fruit set (Jul 21) to

final fruit development stage (Sep 19), followed by slight increase in the last sampling period

(data not shown).

The Principal Component Analysis on Oct 6 PAC measurements revealed variation in levels of

cranberry PACs due to cultivar (Fig. 7B). PC1 has eigenvectors around 0.3 for all individual

PACs, it represents the overall PAC content and due to the close correlation between individual

PAC’s (Supplementary Table 2B), it accounts for over 94% of the total variance of PAC

concentrations. We observed differentiated PC1 scores on different cultivar groups. Specifically,

BL and its offspring CQ and DM had all replicates with lower PC1 scores and clustered on the

left side of the plot (Figure 7B, left circle), while HO and its offspring #35 and MQ had high

PC1 scores and were clustered on the right side (Figure 7B, right circles). LSD test (data not

shown) revealed significant difference (α < 0.05) on PC1 scores between cultivars from the two

groups and no significant difference among cultivars in the same group. BL, CQ and DM had the
lowest levels of total PACs during entire season and HO, #35 and MQ exhibited higher levels of

PACs (Fig. 9A). PC2 with eigenvalue of 0.3 accounts for 3.5% of total variance. It mainly

represents the variation on epicatechin (eigenvector 0.84), as HO with highest epicatechin levels

during entire season has three replicates with high PC2 scores.

3.4. Organic Acids

Cranberry organic acids were identified by comparing peak retention time and UV absorbance

spectra with those of authentic standards. As shown in Table 4 and Fig. 10, peaks 1-4 were

identified as quinic, malic, citric and benzoic acids (Fig. 4J-M).

Fig. 11 shows the levels of four organic acids over the season in cranberry cultivars.

Concentrations of quinic acid gradually declined during fruit development and ripening, from an

average 27.7 mg/g at fruit set (Jul 21) to 16.0 mg/g at fully mature fruit (Oct 6). Cultivars

exhibited significantly different levels of quinic acid during the sampling period. HO and #35

had the highest quinic acid levels, with average quinic acid levels at 26.1 and 24.6 mg/g fruit.

MQ and ST showed average quinic acid levels at 22.0 and 21.2 mg/g fruit, followed by BL (18.4

mg/g), EB (18.0 mg/g) and DM (17.8 mg/g). CQ had lowest quinic acid levels with an average

of 17.0 mg/g fruit.

Malic acid concentration increased during cranberry fruit development and ripening, with

average concentrations increasing from 6.4 mg/g at fruit set (Jul 21) to 8.3 mg/g in fully mature

fruit (Oct 6) (Fig. 11B). All cultivars had similar levels of malic acid during the final fruit

ripening stage between Sep 19 and Oct 6. Before Sep 19, significant differences in malic acid

levels were observed: DM and CQ exhibited the highest malic acid concentrations, while ST

possessed lowest levels of malic acid during entire sampling period.

Concentrations of citric acid fluctuated moderately in most of the cultivars over the sampling

period (Fig. 11C). Cultivars exhibited significantly different levels of citric acid during most of

the sampling period, except on Aug 28. ST, EB and BL showed highest levels of citric acid

during sampling period, with average citric levels at 10.5, 10.1 and 9.3 mg/g fruit, respectively.

On Oct 6, these 3 cultivars had citric acid levels between 8.8 to 9.3 mg/g fruit, significantly

higher than the other 5 cultivars (6.1 - 6.4 mg/g).

Benzoic acid was detected in extremely low concentrations at the first sampling date and

increased slowly through Jul, with concentrations less than 0.01mg/g fruit in all cultivars. From

August onwards, benzoic acid accumulated in all cultivars, along with fruit development, and

continued to increase from middle September to October during fruits’ final ripening (Fig. 11D).

The average concentration of benzoic acid on Oct 6 was 0.08mg/g fruit, with no significant

differences found among the eight cultivars on this final harvest date.

4. Discussion

Flavonoids are secondary metabolites which occur widely across plant species. They provide UV

protection, play an important role in plant sexual reproduction, contribute to plant structure

formation, and have anti-herbivory effects [24,25]. Organic acids in plants are important

intermediates in photosynthesis, and they contribute to a plant’s reaction to nutrient deficiencies

and plant-microbe interaction [24]. Composition and concentration of both flavonoids and

organic acids vary depending on plant species, cultivar, tissues, and growth stages [18,26-28].

Citric and malic acids are important components to fruit flavor [29].

Cranberry fruit development can be visually determined by fruit size and color; the transition

from green (immature) to red (mature) results from anthocyanin accumulation in the epidermis.
Levels and accumulation rates of anthocyanins varied across different cranberry cultivars during

fruit development and ripening. Four early-ripening cultivars exhibited not only earlier

anthocyanin accumulation but also higher anthocyanin levels at the final harvest. This result is

consistent with our previous report comparing two cultivars ST and BL [18]. Anthocyanin level

is an important fruit quality parameter in commercial cranberry fruit production.

In contrast to proanthocyanidins and anthocyanins, all eight cranberry cultivars showed

fluctuating levels of total flavonol glycosides during fruit development and ripening. Six

cultivars had significantly higher flavonol glycoside levels between the last two sampling dates,

being the highest at fruit maturity. In other fruit species, flavonol levels also varied by fruit

maturity stages. Quercetin glycosides were reported to decrease by 50% during fruit

development in skin of two apple cultivars Granny Smith and Splendour, and increased during

ripening in Splendour [30]. Flavonol aglycones (based on a quantification method employing

hydrolysis) were at lower concentrations in mature fruits of red and white currants, plums,

blueberries and cherries [31]. Total and individual flavonol glycosides in lowbush blueberries

decreased significantly from red to blue fruit, and moderately decreased from blue to over

mature fruit [32]. Similar to cranberry, total flavonols in grapes were found to decline during

berry development but increase during ripening [33]. The declining level of flavonols in grape

and cranberry during berry development can be partially attribute to the decreased skin to flesh

ratio caused by enlarged fruit size. Previous studies reported that grape polyphenols, including

flavonols are mainly located in the fruit skin compared to the flesh [34,35]. An analysis

comparing cranberry epidermis and flesh is needed for a better understanding of cranberry

flavonol localization. Interestingly, although EB had smallest mature fruits (largest epidermis to
flesh ratio) among 8 cultivars, it contained the lowest levels of flavonols at the last harvest date,

suggesting that genotypic variation, e.g. variety, is a significant effect.

Q-3-Gal and M-3-Gal were the most abundant flavonol glycosides in all eight cranberry cultivars.

Q-3-Gal was determined to be the most abundant flavonol in highbush blueberries also, in all

maturity stages [32]. Among the eight cultivars, ST exhibited significantly higher concentrations

of Q-3-R and Q-3-A-P during entire or part of the sampling period; HO showed significantly

higher levels of M-3-Gal, M-3-A; and DM exhibited lowest concentrations of Q-3-R and Q-3-A-

P during most of the season. These results indicate genetic background influences flavonol

profiles among cranberry cultivars. As individual flavonol glycosides exhibit differential human

health-related bioactivities and bioavailabilities [36-38], understanding their distinct

concentration patterns in different cranberry cultivars could benefit health-related application of

cranberry products.

Most of the flavonol glycosides did not show concentration patterns which could be related to

cranberry fruit development and ripening. Levels of Q-3-A-F and Q-3-Glu (Fig. 6C, 6H),

however, did exhibit a significant correlation with date in most cultivars, which suggests a

temporal regulation of their biosynthetic pathways in cranberry. Although substrate specificity of

flavonoid glycosyltransferases (GTs) has not been thoroughly elucidated, in Arabidopsis

rhamnosyltransferase and glucosyltransferase were identified as possessing specific substrate

selectivity [39]. Thus, it’s possible that cranberry expresses specific GTs for biosynthesis of

different flavonol glycosides. The different concentration patterns between two quercetin-3-

arabinoside isomers (Fig. 6C and 6G) also indicate the influence of sugar moiety structure on

GTs’ selectivity.
PACs occur in many plant species and play important roles in plant protection against pathogens

and herbivores [40,41]. Individual PACs exhibited similar concentration patterns in all cranberry

cultivars: their concentrations first declined during fruit development, and then slightly increased

during fruit ripening. A similar pattern for total PACs was previously observed in ST and BL

[18]. PAC concentrations were reported to decline during fruit ripening in highbush blueberry,

strawberry, bilberry, grape and apple skin [32,42-45]. In cranberry, we observed an increase in

the PAC concentration during the final fruit ripening stage. Individual PAC polymers exhibited

similar concentration patterns within the same cultivar, suggesting a highly coordinated

regulation of their metabolism. Although enzymatic biosynthesis of major PAC building blocks

catechin and epicatechin has been extensively studied, mechanisms of PAC oligomer and

polymer assembly are yet to be elucidated [46].

In the current study, DP-11+ represented the largest portion of PACs, accounted for 45%-55%

(w/w) of total PACs in cranberry. Similarly, in Gu et al.’s study [47], DP-11+ (>decamers)

accounted for an average 55.8% (w/w) of total quantified PACs in cranberry fruit. Relative DP

of PACs was found to increase sharply during sorghum seed maturation [48]. It was also

speculated that loss of astringency during fruit maturation is associated with increased DP of

PACs [49]. However, in the current study we observed consistent compositions of individual

PACs during entire fruit development and ripening period for all cranberry cultivars. Therefore,

the reduction of astringency during cranberry fruit maturation is more likely associated with the

reduced level of overall PACs, not the change of PAC profile.

An apparent effect of pedigree on levels of anthocyanins and PACs was observed. HO and its

offspring #35 and MQ possessed the lowest level of anthocyanins across most of the season,

including the final harvest. These three cultivars also ranked 1 st, 2nd and 4th in terms of PAC
concentration in mature fruit. In contrast, BL and its offspring CQ and DM possessed lowest

level of PACs during entire sampling period, and were among the top cultivars for anthocyanin

production. The PCA result on PAC measurement also confirmed such pedigree effect, as

cultivars with close genetic background were found clustered in the PC score plot (Fig. 7B).

Interestingly, a negative correlation (R=-0.67, P≤0.002) between levels of anthocyanins and

PACs was observed among those 6 cultivars on Oct 6. Cultivars with high anthocyanin contents

in mature fruit tended to have low levels of PACs, and vice versa, suggesting that these cultivars

channeled resources towards specific flavonoid synthesis direction (anthocyanin or PAC) during

fruit maturation.

Similar to many other fruits [15,50-54], quinic, malic and citric acids are the major organic acids

found in cranberry. We observed a decline of quinic acid in all cranberry cultivars during fruit

development and ripening, with minor fluctuation between consecutive collection dates. On the

contrary, Forney et al. reported 15% increase of quinic acid in cranberry cultivar ST during fruit

maturation [55], potentially due to different cranberry growth conditions, fruit extraction and

analysis methods between the two studies. The inclusion of eight cranberry cultivars, high

sampling frequency (8 collection dates), and aqueous extraction of organic acids used in the

current study, ensured a comprehensive analysis of cranberry organic acids. In other fruits,

quinic acid has been found to peak at early development stages and then decline in orange, lime,

peach and kiwifruit [51,52,54], or decline continuously during fruit development and maturation

in blueberry and loquat [53,55]. As a side product of shikimic acid pathway, which primarily

contributes to aromatic amino acids biosynthesis, it has been suggested that quinic acid

metabolism is involved in aromatic biosynthesis regulation [56]. Thus, decline of quinic acid
concentrations in cranberry during development and maturation could result from plants’

redistribution of resources toward other shikimic acid pathway products.

We also observed a slight increase in malic acid concentration, and fluctuating but consistent

levels of citric acid in cranberry during fruit development and ripening. The increase of malic

acid is consistent with Forney et al.’s study on ST [55], however, they reported a significant

reduction of citric acid in ST over three sampling dates, during fruit transition from green to red.

In orange and lime, malic acid concentrations were consistent [52], while in peach, malic acid

declined during late fruit development and increased slightly during fruit ripening [51]. Citric

acid levels have been found to increase during fruit development in acidic citrus fruits and

kiwifruit [52,54]; or increase until late fruit development then decline during fruit maturation in

peach and pineapple [15,51]. Interestingly, in the current study, certain cranberry cultivars

showed opposite levels of malic and citric acids in mature fruits. ST and EB had the highest

levels of citric acid on Oct 6 and the lowest malic acid concentrations among the 8 cultivars. In

contrast, MQ and #35 had low levels of citric acid on Oct 6, but were the top two cultivars in

terms of malic acid concentration.

Levels of cranberry benzoic acid are extremely low during early fruit development and increase

sharply during late fruit development and ripening. Among the 8 cranberry cultivars, HO

possessed the highest level of benzoic acid and quinic acid in mature fruit (Oct 6). CQ, on the

other hand, exhibited the lowest levels of benzoic and quinic acids in the mature fruit. Benzoic

acid is an important building block of benzoyl group compounds which further contribute to

biosynthesis of numerous plant natural products [57]. Benzoic acid is also highly effective

against yeast [58], thus the temporal accumulation of benzoic acid in mature cranberry fruit

could possibly contribute to the plant’s defense against pathogens.

In summary, the current study revealed that fruit development and ripening process, as well as

genetic variation, can all affect cranberry flavonoid and organic acid biosynthesis and

composition. Further studies will focus on evaluating the effect of different environmental and

biological conditions, such as temperature, irrigation, soil pH, light exposure, as well as pathogen

infection, on the production and regulation of these phytochemicals in cranberry. Anthocyanins

and organic acids are major factors affecting the quality (color, flavor) of cranberry fruits and

related products. Flavonols and PACs are important secondary metabolites with great potential

for human health. Understanding their concentration patterns will add insights to the study of

their biosynthetic pathways and contribute to improving cranberry breeding and harvest

strategies.

Funding Source

This work was supported by National Institutes of Health [grant number NIDCR/NIH

1R01DE16139] and U.S. Department of Agriculture [grant number USDA-NIFA-AFRI 2013-

67013-21107].

Acknowledgement

The authors are grateful to Mr. Graham Gibson (Applied Biosystems) for his gift of the API-

3000TM LC-MS-MS system.

References

[1] S. Gregoire, A. Singh, N. Vorsa, H. Koo, Influence of cranberry phenolics on glucan

synthesis by glucosyltransferases and Streptococcus mutans acidogenicity, J. Appl. Microbiol.
103 (2007) 1960-1968.
[2] I. Borukh, V. Kirbaba, G. Senchuk, Antimicrobial properties of cranberry, Vopr. Pitan. 31
(1972) 82.
[3] S. Häkkinen, S. Auriola, High-performance liquid chromatography with electrospray
ionization mass spectrometry and diode array ultraviolet detection in the identification of
flavonol aglycones and glycosides in berries, J. Chromatogr. A 829 (1998) 91-100.

[4] R.L. Prior, S.A. Lazarus, G. Cao, H. Muccitelli, J.F. Hammerstone, Identification of
procyanidins and anthocyanins in blueberries and cranberries (Vaccinium spp.) using high-
performance liquid chromatography/mass spectrometry, J. Agric. Food Chem. 49 (2001) 1270-
1276.

[5] C. Santos‐Buelga, A. Scalbert, Proanthocyanidins and tannin‐like compounds - nature,

occurrence, dietary intake and effects on nutrition and health, J. Sci. Food Agr. 80 (2000) 1094-
1117.

[6] L.J. Porter, Flavans and Proanthocyanidins, in: J.B. Harborne (Ed.), The Flavonoids,
Chapman & Hall, London, pp. 21–62.
[7] J.A. Choi, J.Y. Kim, J.Y. Lee, C.M. Kang, H.J. Kwon, Y.D. Yoo, T.W. Kim, Y.S. Lee, S.J.
Lee, Induction of cell cycle arrest and apoptosis in human breast cancer cells by quercetin, Int. J.
Oncol. 19 (2001) 837-844.
[8] X. Yan, B.T. Murphy, G.B. Hammond, J.A. Vinson, C.C. Neto, Antioxidant activities and
antitumor screening of extracts from cranberry fruit (Vaccinium macrocarpon), J. Agric. Food
Chem. 50 (2002) 5844-5849.
[9] B.T. Murphy, S.L. Mackinnon, X. Yan, G.B. Hammond, A.J. Vaisberg, C.C. Neto,
Identification of triterpene hydroxycinnamates with in vitro antitumor activity from whole
cranberry fruit (Vaccinium macrocarpon), J. Agric. Food Chem. 51 (2003) 3541-3545.

[10] H.P. Kim, K.H. Son, H.W. Chang, S.S. Kang, Anti-inflammatory plant flavonoids and
cellular action mechanisms, J. Pharmacol. Sci. 96 (2004) 229-245.
[11] D.L. McKay, J.B. Blumberg, Cranberries (Vaccinium macrocarpon) and cardiovascular
disease risk factors, Nutr. Rev. 65 (2007) 490-502.
[12] C.C. Neto, Cranberry and its phytochemicals: a review of in vitro anticancer studies, J. Nutr.
137 (2007) 186S-193S.

[13] A.P. Singh, R.K. Singh, K.K. Kim, K. Satyan, R. Nussbaum, M. Torres, L. Brard, N. Vorsa,
Cranberry proanthocyanidins are cytotoxic to human cancer cells and sensitize platinum‐resistant
ovarian cancer cells to paraplatin, Phytother. Res. 23 (2009) 1066-1074.
[14] Y. Soyer, N. Koca, F. Karadeniz, Organic acid profile of Turkish white grapes and grape
juices, J. Food Compos. Anal. 16 (2003) 629-636.

[15] P. Saradhuldhat, R.E. Paull, Pineapple organic acid metabolism and accumulation during
fruit development, Sci. Hort. 112 (2007) 297-303.
[16] G. Shui, L.P. Leong, Separation and determination of organic acids and phenolic
compounds in fruit juices and drinks by high-performance liquid chromatography, J. Chromatogr.
A 977 (2002) 89-96.
[17] M.A. Awad, A. de Jager, L.H. van der Plas, A.R. van der Krol, Flavonoid and chlorogenic
acid changes in skin of ‘Elstar’and ‘Jonagold’apples during development and ripening, Sci. Hort.
90 (2001) 69-83.

[18] I.O. Vvedenskaya, N. Vorsa, Flavonoid composition over fruit development and maturation
in American cranberry, Vaccinium macrocarpon Ait, Plant Sci. 167 (2004) 1043-1054.
[19] A.D.R. Castrejón, I. Eichholz, S. Rohn, L.W. Kroh, S. Huyskens-Keil, Phenolic profile and
antioxidant activity of highbush blueberry (Vaccinium corymbosum L.) during fruit maturation
and ripening, Food Chem. 109 (2008) 564-572.
[20] A.P. Singh, T. Wilson, A.J. Kalk, J. Cheong, N. Vorsa, Isolation of specific cranberry
flavonoids for biological activity assessment, Food Chem. 116 (2009) 963-968.
[21] T. Wilson, A.P. Singh, N. Vorsa, C.D. Goettl, K.M. Kittleson, C.M. Roe, G.M. Kastello,
F.R. Ragsdale, Human glycemic response and phenolic content of unsweetened cranberry juice, J.
Med. Food 11 (2008) 46-54.
[22] J. Lee, R.W. Durst, R.E. Wrolstad, Determination of total monomeric anthocyanin pigment
content of fruit juices, beverages, natural colorants, and wines by the pH differential method:
collaborative study, J. AOAC Int. 88 (2005) 1269-1278.
[23] I.O. Vvedenskaya, R.T. Rosen, J.E. Guido, D.J. Russell, K.A. Mills, N. Vorsa,
Characterization of flavonols in cranberry (Vaccinium macrocarpon) powder, J. Agric. Food
Chem. 52 (2004) 188-195.

[24] R.E. Koes, F. Quattrocchio, J.N. Mol, The flavonoid biosynthetic pathway in plants:
function and evolution, BioEssays, 16 (1994) 123-132.
[25] ] K.S. Gould, C. Lister, Flavonoid functions in plants. in: Ø.M. Andersen, K.R. Markham
(Eds.), Flavonoids. Chemistry, Biochemistry, and Applications. CRC Taylor and Francis, Boca
Raton, FL, 2006, pp. 397-441.
[26] T. Iwashina, The structure and distribution of the flavonoids in plants, J. Plant Res. 113
(2000) 287-299.
[27] J. Lopez-Bucio, M.F. Nieto-Jacobo, V. Ramirez-Rodrıguez, L. Herrera-Estrella, Organic
acid metabolism in plants: from adaptive physiology to transgenic varieties for cultivation in
extreme soils, Plant Sci. 160 (2000) 1-13.
[28] Y. Nogata, K. Sakamoto, H. Shiratsuchi, T. Ishii, M. YANO, H. Ohta, Flavonoid
composition of fruit tissues of citrus species, Biosci. Biotechnol. Biochem. 70 (2006) 178-192.

[29] R.P. Bates, J.R. Morris, P.G. Crandall, Principles and practices of small and medium-scale
fruit juice processing. in Agricultural services bulletin (Vol. 146). Food & Agriculture Org.,
Rome, 2001.
[30] C.E. Lister, J.E. Lancaster, K.H. Sutton, J.R. Walker, Developmental changes in the
concentration and composition of flavonoids in skin of a red and a green apple cultivar, J. Sci.
Food Agr. 64 (1994) 155-161.
[31] J.J. Macheix, A. Fleuriet, J. Billot, Fruit Phenolics, CRC Press, Boca Raton, 1990.
[32] L. Gibson, H. Rupasinghe, C.F. Forney, L. Eaton, Characterization of changes in
polyphenols, antioxidant capacity and physico-chemical parameters during lowbush blueberry
fruit ripening, Antioxidants 2 (2013) 216-229.
[33] M.O. Downey, J.S. Harvey, S.P. Robinson, The effect of bunch shading on berry
development and flavonoid accumulation in Shiraz grapes, Aust. J. Grape Wine Res. 10 (2004)
55-73.
[34] M. Falchi, A. Bertelli, R. Lo Scalzo, M. Morassut, R. Morelli, S. Das, J. Cui, D.K. Das,
Comparison of cardioprotective abilities between the flesh and skin of grapes, J. Agric. Food
Chem. 54 (2006) 6613-6622.

[35] E.S. Lago-Vanzela, R. Da-Silva, E. Gomes, E. Garcia-Romero, I. Hermosin-Gutierrez,

Phenolic composition of the edible parts (flesh and skin) of Bordô grape (Vitis labrusca) using
HPLC–DAD–ESI-MS/MS, J. Agric. Food Chem. 59 (2011) 13136-13146.
[36] E.U. Graefe, J. Wittig, S. Mueller, A.K. Riethling, B. Uehleke, B. Drewelow, H. Pforte, G.
Jacobasch, H. Derendorf, M. Veit, Pharmacokinetics and bioavailability of quercetin glycosides
in humans, J. Clin. Pharmacol. 41 (2001) 492-499.
[37] Y. Wang, A. Han, E. Chen, R.K. Singh, C.O. Chichester, R.G. Moore, A.P. Singh, N. Vorsa,
The cranberry flavonoids PAC DP-9 and quercetin aglycone induce cytotoxicity and cell cycle
arrest and increase cisplatin sensitivity in ovarian cancer cells, Int. J. Oncol. 46 (2015) 1924-
1934.

[38] Y. Wang, A.P. Singh, H.N. Nelson, A.J. Kaiser, N.C. Reker, T.L. Hooks, T. Wilson, N.
Vorsa, Urinary Clearance of Cranberry Flavonol Glycosides in Humans, J. Agric. Food Chem.
64 (2016) 7931-7939.
[39] P. Jones, B. Messner, J.-I. Nakajima, A.R. Schäffner, K. Saito, UGT73C6 and UGT78D1,
glycosyltransferases involved in flavonol glycoside biosynthesis in Arabidopsis thaliana, J. Biol.
Chem. 278 (2003) 43910-43918.
[40] P. Feeny, H. Bostock, Seasonal changes in the tannin content of oak leaves, Phytochemistry
7 (1968) 871-880.
[41] A. Scalbert, Antimicrobial properties of tannins, Phytochemistry 30 (1991) 3875-3883.
[42] G.W. Cheng, P.J. Breen, Activity of phenylalanine ammonia-lyase (PAL) and
concentrations of anthocyanins and phenolics in developing strawberry fruit, J. Am. Soc. Hortic.
Sci. 116 (1991) 865-869.
[43] L. Jaakola, K. Määttä, A.M. Pirttilä, R. Törrönen, S. Kärenlampi, A. Hohtola, Expression of
genes involved in anthocyanin biosynthesis in relation to anthocyanin, proanthocyanidin, and
flavonol levels during bilberry fruit development, Plant Physiol. 130 (2002) 729-739.
[44] M.O. Downey, J.S. Harvey, S.P. Robinson, Analysis of tannins in seeds and skins of Shiraz
grapes throughout berry development, Aust. J. Grape Wine Res. 9 (2003) 15-27.
[45] A.M. Takos, B.E. Ubi, S.P. Robinson, A.R. Walker, Condensed tannin biosynthesis genes
are regulated separately from other flavonoid biosynthesis genes in apple fruit skin, Plant Sci.
170 (2006) 487-499.
[46] D.Y. Xie, R.A. Dixon, Proanthocyanidin biosynthesis–still more questions than answers?,
Phytochemistry 66 (2005) 2127-2144.
[47] L. Gu, M.A. Kelm, J.F. Hammerstone, G. Beecher, J. Holden, D. Haytowitz, S. Gebhardt,
R.L. Prior, Concentrations of proanthocyanidins in common foods and estimations of normal
consumption, J. Nutr. 134 (2004) 613-617.
[48] L.G. Butler, Relative degree of polymerization of sorghum tannin during seed development
and maturation, J. Agric. Food Chem. 30 (1982) 1090-1094.

[49] J.L. Goldstein, T. Swain, Changes in tannins in ripening fruits, Phytochemistry 2 (1963)
371-383.
[50] W. Kalt, J.E. McDonald, Chemical composition of lowbush blueberry cultivars, J. Am. Soc.
Hortic. Sci. 121 (1996) 142-146.
[51] A. Moing, L. Svanella, D. Rolin, M. Gaudillère, J.-P. Gaudillère, R. Monet, Compositional
changes during the fruit development of two peach cultivars differing in juice acidity, J. Am. Soc.
Hortic. Sci. 123 (1998) 770-775.
[52] M.-V. Albertini, E. Carcouet, O. Pailly, C. Gambotti, F. Luro, L. Berti, Changes in organic
acids and sugars during early stages of development of acidic and acidless citrus fruit, J. Agric.
Food Chem. 54 (2006) 8335-8339.
[53] F.X. Chen, X.H. Liu, L.S. Chen, Developmental changes in pulp organic acid concentration
and activities of acid-metabolising enzymes during the fruit development of two loquat
(Eriobotrya japonica Lindl.) cultivars differing in fruit acidity, Food Chem. 114 (2009) 657-664.
[54] K.B. Marsh, H.L. Boldingh, R.S. Shilton, W.A. Laing, Changes in quinic acid metabolism
during fruit development in three kiwifruit species, Funct. Plant Biol. 36 (2009) 463-470.
[55] C.F. Forney, W. Kalt, M.A. Jordan, M.R. Vinqvist-Tymchuk, S.A. Fillmore, Blueberry and
cranberry fruit composition during development, J. Berry Res. 2 (2012) 169-177.
[56] C. Leuschner, K.M. Herrmann, G. Schultz, The metabolism of quinate in pea roots
(purification and partial characterization of a quinate hydrolyase), Plant Physiol. 108 (1995) 319-
325.
[57] C. Hertweck, A.P. Jarvis, L. Xiang, B.S. Moore, N.J. Oldham, A mechanism of benzoic acid
biosynthesis in plants and bacteria that mirrors fatty acid β‐oxidation, ChemBioChem 2 (2001)
784-786.
[58] C. Verduyn, E. Postma, W.A. Scheffers, J.P. Van Dijken, Effect of benzoic acid on
metabolic fluxes in yeasts: a continuous‐culture study on the regulation of respiration and
alcoholic fermentation, Yeast 8 (1992) 501-517.
Figure captions

Figure 1. Pedigree chart of cranberry cultivars. Cultivars in bold font were analyzed.
Figure 2. Levels of total monomeric anthocyanins (A) and average fruit weight (B) of 8

cranberry cultivars during fruit development and ripening.

Figure 3. HPLC chromatograph of cranberry flavonols purified by Sephadex ® LH-20 column.

All peaks were detected at UV absorbance of 366 nm. Labels 1-9 on peaks correspond to peaks

1-9 in Table 2.
Figure 4. Chemical structures of cranberry flavonols (A-I), organic acids (J-M), and

proanthocyanidins (N and O).

Figure 5. Concentrations and composition of cranberry flavonols during fruit development and

ripening. (A) Individual flavonol glycosides, average of 8 cranberry cultivars. (B) Total

flavonols in 8 cranberry cultivars.

Figure 6. Concentrations of individual flavonol glycosides in 8 cranberry cultivars during fruit

development and ripening. In order of abundance: (A) quercetin-3-galactoside; (B) myricetin-3-

galactoside; (C) quercetin-3-arabinofuranoside; (D) quercetin-3-rhamnopyranoside; (E)

myricetin-3-arabinopyranoside; (F) quercetin-3-xylopyranoside; (G) quercetin-3-

arabinopyranoside; (H) quercetin-3-glucoside.

Figure 7. Principle component analysis (PCA) for eight cranberry cultivars on levels of

individual flavonols (A) and PACs (B), on Oct 6. Scores of principal components (PC) 1 and 2

were plotted.
Figure 8. HPLC chromatogram of cranberry PACs purified by Sephadex® LH-20 column. All

peaks were detected at 280/308 nm of excitation/emission in fluorescence. Labels 1-10 on peaks

correspond to peaks 1-10 in Table 3.

Figure 9. Concentrations of PACs during cranberry fruit development and ripening. (A)

Concentrations of total PACs in 8 cultivars. (B-F): Concentrations of individual PAC oligomers

and polymers in cultivars Howes (B); Early Black (C); #35 (D); Mullica Queen (E) and Ben Lear

(F).
Figure 10. HPLC chromatogram of cranberry organic acids at UV absorbance of 210 nm (peaks

1-3) or 230 nm (peak 4). Labels 1,2,3,4 on peaks correspond to peaks 1,2,3,4 in Table 4.
Figure 11. Concentrations of quinic acid (A), malic acid (B), citric acid (C) and benzoic acid (D)

in 8 cranberry cultivars during fruit development and ripening. (Benzoic acid occurs at much

lower concentrations).
Tables

Table 1. Solvent systems and elution gradients for HPLC analysis of flavonols, PACs and

organic acids in cranberry.

Flavonols PACs Organic Acids

Time Flow A B Time Flow A B Time Flow A B
(min) (ml/min) (%) (%) (min) (ml/min) (%) (%) (min) (ml/min) (%) (%)
0 1 100 0 0 1 100 0 0 0.6 100 0
1 1 85 15 5 1 90 10 11 0.6 100 0
5 1 84 16 7 1 88 12 13 0.6 80 20
10 1 84 16 8 1 88 12 18 0.6 40 60
25 1 83 17 10 1 87 13 20 0.6 20 80
28 1 83 17 15 1 80 20 25 0.6 20 80
30 1 70 30 35 1 60 40 28 0.6 80 20
38 1 55 45 45 1 100 0 30 0.6 100 0
40 1 20 80 50 1 100 0 40 0.6 100 0
43 1 100 0
50 1 100 0
Solvent
0.1% formic acid in water 0.1% acetic acid in acetonitrile 0.5% phosphoric acid in water
A
Solvent 0.1% formic acid in 0.1% acetic acid in 95% 0.5% phosphoric acid in
B acetonitrile methanol/water acetonitrile
Table 2. Retention time, UV λ max, m/z value of [M-H]- and fragment ions in LC-MS/MS and

peak identities of cranberry flavonols.

Retention [M-H]- and Fragment

Peak Time λmax (nm) Ions in APCI MS-MS Peak Identity
(min) (m/z)
1 10.58 257.2, 355.9 479, 317 myricetin-3-galactoside
2 15.39 260.8, 353.6 449, 317 myricetin-3-arabinofuranoside
3 16.72 254.9, 353.6 463, 301 quercetin-3-galactoside
4 17.93 253.7, 352.4 463, 301 quercetin-3-glucoside
5 20.91 254.9, 352.4 433, 301 quercetin-3-xylopyranoside
6 22.41 257.2, 353.6 433, 301 quercetin-3-arabinopyranoside
7 25.05 254.9, 352.4 433, 301 quercetin-3-arabinofuranoside
8 27.20 254.9, 348.8 449, 301 quercetin-3-rhamnopyranoside
9 36.28 254.9, 365.7 301, 179 quercetin (aglycone)
Table 3. Retention time, λexcitation/λemission, m/z value of [M-H]- in LC-MS, A-type linkage

numbers, and degree-of-polymerization of analyzed cranberry PACs.

Retention Excitation/ [M-H]- in No. of

Degree-of- Peak
Peak Time Emission APCI MS A-type
Polymerization Identity
(min) λmax (nm) (m/z) linkage
1 11.04 280/308 575.4 1 2 Dimer
2 15.48 280/308 864.0 1 3 Trimer
3 16.71 280/308 864.1 1 3 Trimer
4 20.26 280/308 1149.7 2 4 Tetramer
5 21.20 280/308 1151.7 1 4 Tetramer
6 22.22 280/308 1438.2 2 5 Pentamer
7 23.06 280/308 1440.4 1 5 Pentamer
8 24.23 280/308 1729.4 1 6 Hexamer
9 25.89 280/308 1729.0 1 6 Hexamer
10 27.29 280/308 2014.2 2 7 Heptamer
Table 4. Retention times, UV λmax and peak identities of analyzed cranberry organic acids.

Peak Number Retention Time (min) λmax (nm) Peak Identity

1 6.263 214 quinic acid
2 7.140 210 malic acid
3 11.407 210 citric acid
4 21.267 230, 276 benzoic acid