1/11 -6 - -4

advertisement
Bachelors of Science in Bioresource Research Thesis of Sonny S.
Simonian presented on June 15, 2000.
APPROVED:
Major Professor, representing Food Science
Co-Advising Professor, representing Food Science
-6
-4
1/11
-
Program Director, Bioresource
search
I understand that my thesis will become part of the permanent collection of the
Bioresource Research Library. My signature below authorizes release of my thesis
to any reader upon request.
Sonny S. Simn' an
i
Anthocyanin and Antioxidant Analysis of Sweet and Tart cherry
varieties of the Pacific Northwest
By
Sonny Simonian
A THESIS
submitted to
Oregon State University
Bioresource Research Major
In partial fulfillment of
the requirements for the
degree of
Bachelor of Science in Bioresource Research
June 2000
i
ABSTRACT
Northwest cherry cultivars including Prunus avium `Bing', Prunus avium
`Rainier', Prunus avium `Royal Anne' and Prunus cerasus `Montmorency' were
separated into epidermal tissue, flesh and pit and cryogenically milled with liquid
nitrogen. These samples as well as whole samples from each variety were
extracted with acetone followed by a chloroform partition. The monomeric
anthocyanin composition for each epidermal tissue sample was quantified using
pH-differential methodology. Calculated (cyanidin-3-glucoside basis) levels were
130.0 ± 21 mg/100g of epidermal tissue for `Bing' cherries, 33.8 mg/100g for
`Montmorency', 3.0 mg/100g ± 0.1 for `Rainier' and 10.5 ± 7 mg/100g for `Royal
Anne'. HPLC, Mass Spectroscopy and spectral analyses, of Prunus avium L.
indicated that >95% of the anthocyanin composition was comprised of cyanidin-3-
glucoside, cyanidin-3-rutinoside and peonidin-3-rutinoside. Analyses for Prunus
cerasus `Montmorency' found that >94% of its composition was cyanidin-3-
glucosylrutinoside, cyanidin-3-rutinoside and peonidin-3-rutinoside.
The antioxidant absorbing capacity of all varieties as well as separated
samples of `Bing' and `Rainier' were measured using Oxygen free-Radical
Absorbing Capacity (ORAL) and Ferric Reducing Antioxidant Power (FRAP)
assays. Samples were measured for their free-radical binding capacity and
compared using Trolox equivalents. `Bing' sample results were 30.1 ± 0.51mM for
ii
ORAC and 23.69 ± 0.44mM for FRAP, 23.07 ± 0.13mM and 26.09 ± 0.47mM
FRAP for 'Montmorency', 18.28 ± 0.45mM ORAC and 12.69 ± 0.39mM FRAP for
`Royal Anne' and 7.37 ± 0.08mM ORAC and 4.17 ± 0.43mM FRAP for `Rainier'.
iii
TABLE OF CONTENTS
Page
Abstract............................................................................ iii
List of Equations ................................................................vii
List of Figures .................................. ..................................viii
List of Tables .....................................................................ix
Introduction ........................................................................1
Anthocyanins ................................................................1
Antioxidants ..................................................
.............2
Material & Methods ................................................................4
Sample Material .............................................................4
Sample Extraction ...........................................................5
Monomeric Anthocyanin Content ........................................7
Polymeric Anthocyanin Content ..........................................9
Titratable Acidity & pH ...................................................12
HPLC Analysis of Anthocyanins ..........................................12
Mass Spectroscopy Analysis of Anthocyanins .........................13
Antioxidant Analysis (ORAC & FRAP) ................................ 14
Results & Discussion ..............................................................16
Physicochemical Analysis of Cherries ..................................16
Monomeric and Polymeric Anthocyanin Analysis ....................18
iv
Semi-quantitative HPLC Anthocyanin Analysis ....................... 20
HPLC Peak Identification .................................................22
Antioxidant Capacity Analysis ...........................................24
Conclusions ..........................................................................25
Recommendations ..................................................................26
Literature Cited ....................................................................27
Appendices ..........................................................................31
Appendix 1 ..................................................................31
Appendix 2 ..................................................................32
LIST OF EQUATIONS
Lambert-Beer's equation .....................................................................
7
Absorbance adjustment for sample haze ...................................................7
Total monomeric anthocyanin ............................................................... 9
Color density .................................................................................10
Polymeric color ..............................................................................10
Percent polymeric color .....................................................................10
Titratable acidity ..............................................................
...........12
vi
LIST OF FIGURES
Sample extraction procedure and methodology ..........................................6
pH differential structural transformation ..................................................8
Bisulfite bleaching of anthocyanins .......................................................11
Monomeric anthocyanin level comparison ...............................................19
HPLC chromatograms .......................................................................21
ES/MS results .................................................................................23
Antioxidant assay results ...................................................................25
vii
LIST OF TABLES
Comparative weights of separated cherry samples ....................................17
Standard physicochemical attributes of cherry samples ...............................18
Levels of polymerized anthocyanins ......................................................20
Viii
INTRODUCTION
Scientists and parents alike have frequently implied the beneficial properties
of fruits and vegetables. In the early 20th century, Willstatter and Zolinger (1916)
characterized phenolic pigments within the Prunus avium L. variety of cherries as
anthocyanins. Early identification of anthocyanins within sweet cherries was
performed using paper chromatography. Some of the anthocyanin pigments
identified included cyanidin and peonidin derivatives. In 1979 peonidin-3rutinoside was identified as the major pigment found in varieties of Bigarreau and
Napoleon cherries (Lynn and Luh, 1979). Recently High-Performance Liquid
Chromatography (HPLC) has been used in the quantification of pigments within
sweet cherry varieties. Pitted samples of highly pigmented cherries yield cyanidin3-rutinoside and cyanidin-3-glucoside as major peaks and peonidin-3-rutinoside,
peonidin-3-glucoside and pelargonidin-3-rutinoside as minor peaks (Gao and
Mazza, 1995).
Similar evaluations of the Prunus cerasus L. varieties identified cyanidin-3glucosylrutinoside, cyanidin-3-rutinoside, cyanidin-3-glucoside, peonidin-3-
rutinoside and cyanidin-3-sophoroside (Schaller and Von Elbe, 1968). Pigments of
tart cherries were later quantified through the use of HPLC to reveal cyanidin-3sophoroside and cyanidin-3-rutinoside as being the major anthocyanins in
Monmorency tart cherries (Chandra et al. 1992). Anthocyanins within
1
Monmorency and Balton varieties of tart cherries reported by Wang et al. 1997
included cyandin-3-glucosylrutinoside, cyanidin-3-rutinoside and cyanidin-3glucoside.
Utilization of anthocyanin-containing extracts such as grape skin and grape
color has application as synthetic colorant alternatives (Hong and Wrolstad, 1990).
Anthocyanin content in fruits such as sweet and tart varieties of cherries has
provided an index of fruit quality (Mazza and Miniati, 1993). Alternatively,
anthocyanins and other compounds found in tart cherries have exhibited
antioxidant properties (Hong et al., 1997). The theory that antioxidants present in
natural food systems provide protection against atherosclerosis and low density
lipoprotein accumulation is key in the growth of the health conscious targeted
nutraceutical industry (Robards et al., 1999).
The distribution of anthocyanins and other phenolic compounds within the
higher plant species is quite diverse with respect to each individual (Robards et al.,
1999). Despite this diversity, a commonality is that all posses the ability to
scavenge active oxygen species, inhibit nitrosation (Helser and Hotchkiss, 1994)
and prevent autoxidation by chelating metal ions (Robards et al., 1999). This holds
true for sweet and tart varieties of cherries. As reported by Michigan State
scientists, the addition of ground tart cherries as a fat substitute improved "warmedover flavor," limited rancidity and provided antioxidant properties greater than that
2
of vitamin E, vitamin C and selected synthetic antioxidants (Crackel et al., 1988
and Britt et al., 1998).
Recent correlation has been proposed between oxygen free radical
absorbing capacity and fruit pigmentation (Kalt et al., 1999). A cross-commodity
study of strawberries, raspberries and blueberries strongly correlated the
anthocyanin content of these fruits with the ability to scavenge free radicals and
prevent oxidative damage. In addition to the antioxidant properties of
anthocyanins, it has been reported that other phenolic compounds also contribute to
the antioxidant power of natural fruit systems (Prior et al., 1998). Anthocyanins
that are abundant in cherry varieties such as cyanidin-3-glucoside and cyanidin-3rhamnoglucoside have up to 3.5 times the antioxidant capacity of Trolox standards
(Wang et al., 1997). Acylation, substitution and glycosylation of the
anthocyanidins has also been linked with affecting the potency of anthocyanins
(Rice, Evans et al., 1996 and Wang et al., 1997).
The ability of anthocyanins to act as antioxidants has only increased the
need to further study natural systems and potential compounds associated with
current epidemiological studies. With over 250 naturally occurring anthocyanins
(Strack and Wray, 1993), there is an intense interest to categorize and quantify the
antioxidant capacities of these valuable flavonoid colorants. The daily intake of
anthocyanins has been estimated to be as much as 180-215 mg/day (Kuhnau, 1976)
3
and even though the physiological effect of them are not fully understood, they
have been attributed to a variety of therapeutic activities. Scharrer and Ober
attributed anthocyanins to the treatment of diabetic retinopathy (1981) as well as
vision improvement (Hong et al., 1997).
MATERIALS AND METHODS
Sample material
Hood River Farms (Hood River, OR) samples of cherry varieties included
Prunus avium 'Bing' and Prunus avium `Rainier'. Material was donated in three
boxes of separately harvested (about a day apart) cold packed fruit. Prunus avium
`Royal Anne' and Prunus cerasus 'Montmorency' cherries (one gallon buckets)
were obtained from Louis Brown Farms (Corvallis, OR). Leaves and stems were
removed from the cherries and the fruit was washed with cold water. Cherry
samples were then stored at 1°C for no more than three days. After removing
moldy or rotten cherries, a seemingly representative sample was obtained for
analysis. Each variety was hand peeled and separated into epidermal tissue, seeds
and flesh and then frozen under liquid nitrogen. Using the separated samples,
weight percentages of skin, flesh and pit material were obtained for each variety.
Separated and whole samples were then stored at -21°C. These methods are
described in more detail in Wrolstad et al. 1990 (Figure 1).
4
Soluble solid concentrations
Soluble solid concentrations were measured for each whole cherry sample
using a Reichert-JungTM temperature controlled refractometer, prior to individual
sample extraction.
Sample extraction
Extraction for each sample followed methods described by Wrolstad et al.
(1990) and Hong and Wrolstad (1990). Separated samples (skin, pit and flesh)
were cryogenically milled using liquid nitrogen in a Waring stainless steel blender.
Powdered material of approximately 25g was initially extracted using 100%
acetone (1:1 sample weight:solvent w/v) and filtered on a Buchner funnel using
Whatman #1TM qualitative filter. The filter cake and sample residue were then
sonicated for five to ten minutes and then re-extracted with 30% aqueous acetone
(30:70 water:acetone v/v) until a clear runoff was obtained. The filtrates from each
sample was collected and then combined in a separatory funnel with chloroform
(1:1 acetone: chloroform v/v). The contents was shaken and stored overnight at
1°C. After phase separation, the aqueous top portion, which contained
anthocyanins, phenolics, sugars, organic acids and other water soluble compounds,
was collected and residual acetone evaporated on a Buchi roto-evaporator, at 40°C
5
(5-10 min). The bottom layer or bulk phase, which was made up of immiscible
organic solvents, dissolved lipids, carotenoids, chlorophyll pigments and other non-
polar molecules, was then discarded. Each sample was then diluted to a known
volume using distilled, deionized water (Figure 1).
Figure 1. Outline of preliminary sample preparation and extraction. Fresh fruit
stored at 1°C. Extracted samples diluted to known volume and stored at -21°C.
fresh fruit
% soluble solids
separated parts
into liquid nitrogen
composition %
wt. (skin, flesh, pit)
cryogenically
milled (liquid nitrogen)
-25g sample
extraction
100% acetone
1:2 (w/v)
presscake
re-extracted
30% aq. acetone
partitioned
1:1 (v/v)
chloroform: acetone
evaporated
6
Monomeric anthocyanin content
Equation 1. Lambert-Beer's Law equation: C = A
EL
C
: molar concentration
A
:absorbance
E
: molar extinction coefficient
L
: pathlength (cm)
Using pH differential methods described in Wrolstad, 1976; Fuleki and
Francis, 1968, in accordance with Lambert-Beer's Law (Equation 1), the total
anthocyanin content of each extracted sample was obtained. Since the oxonium
form of the anthocyanin chomophore (colored form) predominate at pH 1.0,
samples were diluted using pH 1.0 buffer to put samples within the linear
absorbance of the spectrophotometer (approximately 1.2). Due to structural
transformations within the anthocyanin molecule, the same dilution factor was used
to obtain the colorless hemiketal (Figure 2) form of the sample using a buffer of pH
4.5. Measurements were made using ShimadzuTM 300 UV/visible
spectrophotometer and lcm path-length disposable cells. Spectral data was
recorded at 512nm (maximum absorbance) and 700nm (for calculation of sample
haze). Absorbance correction for sample haze is described in Equation 2.
7
Equation 2. Absorbance adjustment for haze:
A = (A(512nm) - A(700nm)) pH 1.0 - (A(512nm) - A(700nm)) pH 4.5
The absorbance of the sample was calculated using (Equation 2) and then this
value was used to calculate the monomeric anthocyanin pigment content (Equation
3). Pigment content was calculated as cyanidin-3-glucoside (MW = 449.2) with an
extinction coefficient of 26,900.
8
Figure 2. Anthocyanin structural transformation for pH differential color
measurements.
Ri
Ri
HO
ady
Q.i ncrd dd bcse(d ue)
pH = 7
aH y
FIaiiIiL ncdicn(aKariu.rnfcrm)
R2
Chd acre: od crI Ess
pH =4.5
Cab nd pse .
bcse(hari kdd form) : od a'I Ess
pH = 4.5
9
Equation 3. Total monomeric anthocyanin (mg/L) = (A x MW x DF x 1000)
A
(EL)
absorbance of the sample corrected for haze
MW
molecular weight (cyanidin-3-glucoside = 449.2)
DF
: dilution factor
L
: pathlength (cm)
$
molar extinction coefficient (26,900)
Polymeric anthocyanin content
Using pH differential methods described in Wrolstad, 1976; Somers and
Evans, 1974, in accordance with Lambert-Beer's Law (Equation 1), the effects of
browning and the formation of polymeric color within each sample was obtained.
Dilution values obtained in monomeric measurements are made with distilled water
in this assay. For comparison of polymeric color, 0.2m1 of distilled water was
added to 3m1 of diluted sample in a lcm disposable cell and 0.2m1 of bisulfite
(potassium metabisulfite (K2S205)) was added to the other. The bisulfite diluted
sample is used to produce a colorless sulfonic acid adduct (Figure 3). Monomeric
anthocyanins are readily "bleached" using the bisulfite solution, however
polymeric forms of anthocyanin-tannin interactions and melanoidin pigments that
may be formed, are resistant to this reaction of bisulfite bleaching. Absorbance
readings were taken at 420nm (polymeric color), 512nm (maximum) and 700nm
10
(haze). Color density for the control sample was calculated and used in the
determination of polymeric color (Equation 4).
Equation 4. Color density of control sample:
color density = [(A(420nm) - A(700nm)) - (A(512nm) - A(700nm))1 x DF
The anthocyanin/tannin complexes, which are not affected by the bisulfite
"bleaching," therefore enabled the spectrophotometric measurement of browning
or polymeric color (Equation 5). The percent polymeric color was then calculated
(Equation 6).
Equation 5. Polymeric color of the bisulfite bleached sample:
polymeric color = [(A(420nm) - A(700nm)) - (A(512nm) - A(700nm))l x DF
Equation 6. Percent polymeric color = polymeric color x 100
color density
11
Figure 3. Bisulfite bleaching and sulfonic acid adduct.
HO
strcng
aid
tir
12
Titratable acidity & pH
A MetrohmTM auto-titrating system, allowed for the rapid analysis of total
acidity and pH in levels within the extracted samples. Initial pH readings were
recorded for each of the prepared samples. Samples were then titrated to a pH of
8.1, using O.1N NaOH, and the volume used was recorded. Calculations for total
acidity were made using (Equation 7). All values were calculated and reported as g
malic acid/100ml.
Equation 7. Total acidity = NaOH (ml) x Normality x meq wt. x DF x 100
Sample (ml)
meq wt.
: acid MW (malic acid)
proton number
DF
: dilution factor
HPLC analysis of anthocyanins
To ensure clear samples and equipment integrity, diluted extractions were
filtered using a 0.4µm filter. Filtered samples were then transferred to (Altec) auto
sampler glass vials and capped using rubber center injection caps.
Chromatographic data was obtained using a Perkin-ElmerTM Series 400
liquid chromatograph, with a Hewlett-PackardTM 1040A photodiode detector.
13
Computer analysis of data was collected on a Hewlett-Packard 9000 computer
system. Chem Station TM software was used in the analysis and collection of peak
data. Spectra were recorded at 280nm, 320nm and 520nm simultaneously.
A gradient mobile phase (10% HOAc, 5%CH3CN, 1% H3PO4 by volume)
was used as follows: Time 0 (minutes): 0% CH3CN, 100% mobile phase. Time 5:
0% CH3CN, 100% mobile phase. Time 20: 20% CH3CN, 80% mobile phase.
Time 25: 40% CH3CN, 60% mobile phase. Time 30: 0% CH3CN, 100% mobile
phase. Separation procedures were performed using a reverse phase C-18 column,
with a flow rate of lml/minute. Specifically, a Prodigy ODS-3TM column (250 x
4.6mm ID) by Phenomenex, fitted with a Prodigy ODS-3TM guard column.
Mass Spectroscopy analysis of anthocyanins
Sample analysis was performed using a Perkin-ElmerTM SCIEX API IH+
Mass Spectrometer, with loop-injection and equipped with an Ion Spray source
(ISV = 4700, orifice voltage of 80).
Isolation of anthocyanin compounds for mass spectroscopy was performed
using a high-load C-18 mini column (AlltechTM Inc.). Column was initially
activated using methanol, then 0.01% HCl (Wrolstad et al., 1990). Extracts were
passed through the column resulting in retention of anthocyanins (and other
14
phenolics). Sugars, acids and other water soluble interfering compounds eluded out
with 2 aliquots of 0.01% aqueous HCI. Anthocyanins adsorbed to C-18 column
could then be collected using a methanol/0.01% HCl solution (v/v). The purified
anthocyanins remained in methanol for ES/MS analysis. Samples were partially
purified using a 0.45µm pore filter, removing interfering compounds prior to C-18
mini-column separation. Purified samples were then analyzed using low-resolution
electro-spray mass spectroscopy.
Antioxidant assays
Photo-Chemiluminescence analysis of Oxygen Radical Absorbance
Capacity (ORAL) and Ferric Reducing Ability of Plasma (FRAP). Assays were
carried out in conjunction with The Linus Pauling Institute.
ORAC. Original source for methodology: Cao, et al. (1993).
For ORAC analysis, a microplate fluorometer Cytoflour 4000 (PerSeptive
Biosystems, Framingham, MA) was used in conjunction with an excitation filter of
485nm and an emission filter of 585nm.
Initial samples were diluted with sodium phosphate buffer (PBS) and then
30µ1 aliquots were added to the 96 well plate, in triplicate, followed by addition of
15
200µl/well of pre-warmed stock solution of (3-phycoerythrin (0-PE), Sigma # P-
1286, 417 pM solution (1 mg dry powder/900µl PBS) in PBS. Trolox standards of
((+/-)-6-Hydroxy-2,5,7,8-tetramethylchromane-2-carboxylic acid, MW: 250.32,
Fluka #565 10) in 10% McOH/water) were prepared at 40, 20, 10, 5 and 0µM
concentrations. Wells around the outside of the tray are filled with 37°C prewarmed water. Initiation of the reaction occurs with the addition of 70µl 2,2'Azobis (2-amidino-propane) dihydrochloride dissolved in water (1/10 final
volume), MW: 271.17, Wako #992-11062 (AAHP reagent). The plate was then
placed into a pre-warmed (37°C) fluorometer and the kinetic change in (3- PE
fluorescence (excitation filter 485 nm, emission filter 585 nm) was measured every
two minutes for two hours (61 cycles).
Protein-independent antioxidant activity was measured by treating samples
with equal volume of 0.5M perchloric acid followed by five minutes of incubation
at room temperature and 15 minutes of centrifugation at 16,000 rpm using an
Eppendorf centrifuge. The supernatant was collected and diluted with PBS to fit
into calibration curve. Trolox standards were prepared in corresponding
concentrations of perchloric acid.
FRAP antioxidant assay. Original source for methodology: The Ferric
Reducing Ability of Plasma (FRAP) as a measure of antioxidant power: the FRAP
assay. Analytical Biochem 1996 239:70-76
16
FRAP analysis equipment included a microplate reader ThermoMaxTM, by
Molecular Devices, Forster City, CA; using a filter 550nm.
40µL samples and Trolox standards at concentrations of 500, 250, 125, 62.5
and 0 µM were initially diluted with water and added in duplicate to a 96 well plate
(flat-bottom). 300µl/well of pre-warmed FRAP reagent was added to each sample
and standard and the plate was then incubated for 15 minutes at 37°C (inside prewarmed Microplate reader). Absorbance was read at 550nm and antioxidant levels
are calculated against a standard (Trolox) calibration curve using linear
approximation. Values are expressed in Trolox concentration (M) with
corresponding activity (for biological fluids) or in mole amount of Trolox
necessary to match an activity of 1 mole of sample (for purified compounds) using
SoftMaxTM software.
RESULTS & DISCUSSION
Physicochemical analysis of cherry cultivars
Prunus avium 'Bing' and Prunus avium `Rainier' samples were harvested at
three similar times from the Hood River (OR) orchards. Prunus avium `Royal
Anne' and Prunus cerasus `Monmorency' varieties were harvested two separate
times from the Louis Brown Farms in Corvallis (OR). Average weights for each
17
cultivar was assessed as well as the constituent weights (skin, flesh and seed) for
each variety (Table 1). The `Rainier' cherries yielded the highest average whole
fruit size of 14.7g. In sub-sampling trials, whole fruit size ranged from 7.6g
('Monmorency') to 15.3g ('Rainier') (Table 1) and after separation, average
constituent weight percentages for skin, flesh and pit were 19.5%, 68.4% and
12.1% respectively.
Table 1. Comparative and percent weights of separated cherry samples (mean ±
standard deviation).
Cherry variety
Bing°`
average wt.
Rainiera
Royal Annex
Montmorency
14.7
13.8
12.8
9.7
% skin
20.5 ± 3.7
16.4 ± 3.0
17.1 ± 0.5
24.2 n/a
O 'Replication, n = 3. x Replication, n = 2.
8
% flesh
66.8 ± 3.9
70.8 ± 4.3
72.0 ± 2.1
64.1 n/a
% seed
9.0 ± 1.2
8.4 ± 1.3
10.9 ± 2.5
11.7 n/a
Replication, n = 1.
Soluble solid concentrations (SSC) measurements were made on the
original the fresh cherries (Table 2). Each whole and separated cherry sample was
analyzed for pH and total acidity (TA) (Table 2). Calculations were made using
malic acid as the primary acid (Girard and Kopp, 1998). Calculations for
SSC:acid ratios of each variety are reported in Table 2 as well. Girard and Kopp
previously reported correlation coefficient values for pH and titratable acidity (r = -
0.78), results from this study found a lower correlation (r = -0.57). The correlation
between TA and SSC were also reported (r = 0.50), this study found a much higher
18
correlation (r = 0.96). Soluble solid concentrations appeared highest in the `Royal
Anne' and `Bing' samples. The `Montmorency' variety provided the highest acidity
levels, however, due to the limited sample size, an accurate total acidity evaluation
could not be assessed. Colorimetric assays (L*, a*, b*) were not used for
measurement of fruit maturity (Girard and Kopp, 1998).
Table 2. Standard physicochemical attributes of cherry samples (mean ± standard
deviation). Analyses included soluble solid concentration (% soluble solids), pH,
titratable acidity and % solid to acid ratios.
Cherry variety
Bing°`
SSC
19.9±2.1
pH
3.7±0.1
TA (% malic acid)
0.7±0.1
Rainier a
Royal Anne X
Montmorency S
17.5 ± 2.6
4.1 ± 0.1
0.4 ±
0.1
46.1 ± 1.9
22.5 ± 4.2
18.5 n/a
4.0 ± 0.3
3.2 n/a
0.7 ± 0.1
32.4 ± 3.4
1.1 n/a
16.8 n/a
O 'Replication, n = 3. x Replication, n = 2.
S
SSC/TA
32.8±2.4
Replication, n = 1.
Monomeric and polymeric anthocyanin analysis
Epidermal tissue separated from each sample comprised the majority of
monomeric anthocyanins within that commodity. The highly pigmented `Bing'
variety yielded 132mg/100g within epidermal tissue and 68mg/lOOg in the flesh of
the fruit (Figure 4). Epidermal tissue isolated from the tart `Montmorency' variety
provided considerably lower values of 34mg/100g. Trace amounts of pigments
19
were present in most of the other samples the highest concentrations existing within
the epidermal and fleshy tissue.
Figure 4. Comparison of constituent (skin, flesh and pit) total monomeric
anthocyanin levels obtained through pH differential methods.
140.00
120.00
100.00
0 80.00
0
0)
E 60.00
40.00
20.00
0.00
skin
flesh
pit
Bing ® Montmorency ED Royal Anne 0 Rainier
20
Bisulfite bleaching and UV/visible analysis at 700nm (haze), maximum
wavelength (520nm) and 420nm (polymeric color) revealed that cryogenic milling
and solvent extraction methods did not fully inhibit browning and pigment
polymerization. Varieties such as `Rainier' with low initial pigmentation, obtained
about 90% of its color from polymerization. The lowest percentage of
polymerization occurred in the `Bing' flesh samples (Table 3).
Table 3. Monomeric anthocyanin levels reported (520nm). Results from sulfonic
acid (H2SO3) bleaching of the monomeric anthocyanins, polymerized anthocyanins
are measured at 420nm. Measurements correspond to all constituents of separated
samples.
Sample
Bing' Skin
Bing' Flesh
Bing' Pit
Rainier' Skin
Rainier' Flesh
Rainier' Pit
Royal Anne' Skin
Royal Anne' Flesh
Royal Anne' Pit
Montmorency' Skin
Montmorency' Flesh
Montmorency' Pit
Anthocyanin Conc.
mg/100g
Color
Density
Polymeric
Color
130.0
71.3
23.4
16.0
6.7
5.6
1.9
0.7
34.3
16.0
34.9
3.0
3.0
1.3
46.4
0.0
1.3
0.8
57.3
0.0
0.4
0.3
93.9
10.5
3.3
1.2
36.4
0.1
1.8
0.7
39.0
0.0
0.6
0.3
50.0
33.8
12.2
1.7
14.0
9.3
4.7
0.8
17.1
0.5
1.3
0.4
30.8
1.1
%
Polymeric
21
HPLC semi-quantitative analysis of anthocyanins
Anthocyanin concentrations of the extracted samples were analyzed using
HPLC. Data was collected at 280, 320 and 520nm and the primary visible region
of 520nm was used for anthocyanin analysis (Figure 5). Approximately 90% of
anthocyanin levels in the sweet cherry varieties (Figure 5 - II) were present in peaks
4 and 6. About 7% of the total area was contained in peaks 3 and 5.
Anthocyanin data collected at 520nm for the `Monmorency' variety
determined that the primary peaks (2 and 5) comprised about 90% of the area.
Peaks 1, 3 and 7 made up about 6% of the total area (Figure 5 - I). The
'Montmorency' sample indicated a more polar primary peak with a retention time
of 7.5 minutes.
22
Figure 5. HPLC separation (detection at 520nm) of I. Prunus cerasus
'Montmorency' and H. Prunus avium 'Bing'. Milli-absorbtion units (mAU) vs.
time (min).
l
1.
12
2
100
A
80
4
6
1
201
3
0
5
20
15
10
Elution time (min)
4
6
5
10
15
20
Elution time (min)
Peak assignments
1. cyanidin-3-sophoroside, 2. cyanidin-3-glucosylrutinoside, 3. cyanidin-3glucoside, 4. cyanidin-3-rutinoside, 5. peonidin-3-glucoside and 6. peonidin-3rutinoside.
23
Peak identification
With the overall charge of anthocyanins being one, mass/charge ratios
obtained through the use of ES/MS included molecular weights of 449 and 595 for
sweet varieties corresponding to cyanidin-3-glucoside and cyanidin-3-rutinoside.
Molecular weights of 358, 391, 538, 611 and 633 also appeared but did not
correspond to known values (Figure 6). Molecular weights of 595 and 757
confirmed peak assignments for cyanidin-3-rutinoside and cyanidin-3-
glucosylrutinoside for the `Monmorency' variety. In addition to the known
anthocyanin peak assignments, molecular weights of 291, 325, 355, 458, 597, 611
and 748 were detected (Figure 6).
24
Figure 6. Mass spectrometry of I. Prunus avium 'Bing' and H. Prunus ceracus
'Montmorency'.
I.
Molecular weights of 449 and 595 for sweet varieties correspond to cyd-3glucoside and cyd-3-rutinoside. Molecular weights of 358, 391, 538, 611
and 633 also appeared but did not correspond to known values
II.
Molecular weights of 595 and 757 confirmed peak assignments for cyd-3rutinoside and cyd-3-glucosylrutinoside for the `Monmorency' variety. In
addition to the known anthocyanin peak assignments, molecular weights of
291, 325, 355, 458, 597, 611 and 748 were detected
1.
6;
391,0
595.2
a)
311
449.0
343 0
0
II.
633.2
450.0
25
... n....._5l».._......d,....,.__.. t,.lx........1 I.I'u\...li... 4.,k._..n........`w. _....iL.u__J1....Ml_.J... I.
400
300
500
600
700
800
Mass to charge ratio
91.0
3251.0
3
8
8
25
579 2
595.2
I
757;2
WJ4JjLq.,U.-4WJYPJ
300
400
500
600
700
800
Mass to charge ratio
25
Antioxidant capacity analysis
The free-radical binding capacities of all whole fruit samples as well as
separated samples (skin, flesh and pit) of `Bing' and `Rainier' varieties were
performed at The Linus Pauling Institute. All samples were analyzed in triplicate
with both the oxygen free-radical absorbing capacity (ORAL) and the ferric
reducing antioxidant power (FRAP) assays. Results for both the ORAC and FRAP
assays are reported in Trolox equivalent units in conjunction with a Trolox
standard.
In addition to the measurement of antioxidant capacity, ascorbic acid values
for all of the samples were analyzed. Results of ascorbic acid assays were below
the detection limit of 1µM and were therefore not reported in this study.
Results obtained by the FRAP assay were generally lower than those
obtained using ORAC methods. The FRAP assay as with the ORAC assay
indicated that separated flesh samples from the `Bing' variety had the greatest
antioxidant power (ORAC: 87.60mM and FRAP 87.94mM Trolox). This assay
also indicated that the whole sample of the sour variety, 'Montmorency' had a
higher reducing ability than that of the sweet `Bing' whole sample (Figure 7).
26
As with the FRAP assay, the ORAC assay indicated that the separated flesh
sample of the `Bing' cherries had the greatest free-radical absorbing capacity of
87.94mM Trolox units and 87.60mM Trolox units respectively (Figure 7).
However, ORAC measurement for the whole `Bing' sample was greater than that
of the `Montmorency' sample (30.10mM and 23.07mM respectively), yet slightly
less in the FRAP assay (23.69mM and 26.09mM respectively) (Figure 7).
Contradictory to reported results of weak correlation between these two assays,
data obtained provided a correlation coefficient of 0.903 (Prior and Cao, 1999).
Figure 7 ORAC and FRAP analysis of all whole cherry samples and separated
Prunus avium `Bing' and `Rainier' samples (Trolox units (mM).
ORAC mM Trolox units U FRAP mM Trolox units
Bing pit
Bing flesh
Bing skin
Rainier pit
Wl
Rainier flesh
Rainier skin
Bing
Royal Anne
Rainier
Montmorency
10
20
30
40
50
60
70
80
90
10(
27
CONCLUSIONS
Anthocyanin identification using HPLC, UV-visible spectral analyses and
ES/MS results, previous results for sweet and sour cherry varieties. Major
anthocyanin identities were confirmed, using HPLC. Peaks included those of
cyanidin and peonidin. However, through the use of mass spectrometry, previously
reported pelargonidin moieties were not found in any of the sweet varieties.
Mass/charge ratios identified a molecular weight corresponding to pelargonidin-3-
rutinoside within the sour variety of 'Montmorency.' Of the three major peaks
detected in the sweet cultivars, ca. 95% of the total anthocyanins present included
cyanidin-3-rutinoside and peonidin-3-rutinoside. And, of the seven anthocyanin
peaks indicated within the sour variety, ca. 94% was comprised of cyanidin-3-
glucosylrutinoside, cyanidin-3-rutinoside and peonidin-3-rutinoside. Anthocyanins
were highly concentrated in the epidermal tissue of all samples.
Strong correlation, for all samples, was not found between cherry
pigmentation and individual antioxidant absorbing capacity. Cross-commodity
studies have indicated a strong positive correlation between anthocyanin levels and
antioxidant power r = 0.91) (Kalt et al., 1999). Most likely the weak correlation in
this study was due to the high levels detected in the 'Bing' flesh samples and the
inclusion of the 'Montmorency' variety and the contribution of colorless phenolics
28
reported in recent work (Wang, 1997). However, comparison of the whole sweet
varieties, a much closer correlation was noted (r = 0.901).
RECOMMENDATIONS
Other polyphenolics, sugars and organic acids were not assayed in this
study and will be a part of continuing research of cherry phytochemicals. Further
compound identification through HPLC analysis, NMR and MS/MS, is required for
a complete compositional analysis of cherries. Of the compounds identified, in this
study using ES/MS, many of the unknown peaks would be better understood after
fractionating the molecules into their substituent parts using MS/MS separation
techniques. NMR may also be a useful tool in determining the phytochemical
composition of various cherry cultivars.
Further analysis is also required for evaluation of post-processing material
and the losses involved in processing and brining. Analysis of pit as well as presscake material may reveal valuable nutraceutical, natural colorant and antioxidant
sources.
29
LITERATURE CITED
Benzie, I.F.F. and Strain, J.J. 1996. The ferric reducing ability of plasma (FRAP)
as a measure of "antioxidant power": the FRAP assay. Anal. Biochem. 239: 7076.
Britt, C., Gomaa, E.A. Gray, J.I. Booren, A.M. 1998. Influence of cherry tissue on
lipid oxidation and heterocyclic aromatic amine formation in ground beef Patties.
J. Agric. Food Chem. 46: 4891-4897.
Cao, G., Alessio, H.M. and Cutler, R.G. 1993. Oxygin-radical absorbance
capacity assay for antioxidants. Free Radic. Biol. Med. 14: 303-311.
Chandra, A. and Nair, M.G. 1992. Evaluation and characterization of the
anthocyanin pigment in tart cherries (Prunus cerasus L.). J. Agric. Food Chem.
40: 967-969.
Cheng, G.W. and Crisosto, C.H. 1995. Browning potential, phenolic composition
and polyphenoloxidase activity of buffer extracts of peach and nectarine skin tissue.
Amer. Soc. Hort. Sci. 120: 835-838.
Crackel, R. L., Gray, J. I., Booren, A. M., Pearson, A. M. and Buckley, D. J. 1988.
Effect of antioxidants on lipid stability in restructured beef sticks. J. Food Sci. 53;
656-657.
Friedrich, J.E., Lee, C.Y. 1998. Phenolic compounds in sweet and sour cherries.
polyphenols communications 98 (pp. 527, 528). XIXth International Conference
on Polyphenols, Lille, 1-4 September.
Fuleki, T. and Francis, F.J. 1968. Quantitative methods for anthocyanins.
Extraction and determination of total anthocyanin in cranberries. J. Food Sci. 33:
78-83.
Gao, L., Mazza G. 1995. Characterization, quantitation and distribution of
anthocyanins and colorless phenolics in sweet cherries. J. Agric. Food Chem. 43:
343-346.
Girard, B. and Kopp, T.G. 1998. Physicochemical characteristics of selected
sweet cherry cultivars. J. Agric. Food Chem. 46: 471-476.
Helser, M.A. and Hotchkiss, J.H. 1994. Comparison of tomato phenolic and
ascorbate fractions on the inhibition of N-nitroso compound formation. J. Agric.
Food Chem. 42: 129-132.
30
Hong, V. and Wrolstad, R.E. 1990. Use of HPLC separation/photodiode array
detection for characterization of anthocyanins. J. Agric. Food Chem. 38: 708-715.
Hong, W., Guohua, C. and Prior, R.L. 1997. Oxygen radical absorbing capacity of
anthocyanins. J. Agric. Food Chem. 45: 304-309.
Kalt, W., Forney, C.F., Martin, A. and Prior, R.L. 1999. Antioxidant capacity,
vitamin C, phenolics, and anthocyanins after fresh storage of small fruits. 47:
4638-4644.
Kuhnau, J. 1976. The flavonoids. A class of semi-essential: Their role in human
nutrition. Wld. Rev. Nut. Diet. 24: 117-191.
Lynn, D.Y.C. and Luh, B.S. 1964. Anthocyanin pigments in Bing cherries. J.
Food Sci. 29: 735.
Macheix, J., Fleuriet, A., Billot, J., Harborne, J. 1990. Fruit Phenolics. CRS Press,
Inc., Boca Raton, FL.
Mazza, G. and Miniati, E. 1993. Anthocyanins in fruits, vegetables, and grains.
CRC Press: Bocca Raton, Florida. 57-63.
Prior, R. L. and Cao, G. 1999. In vivo total antioxidant capacity: comparison of
different analytical methods. Free Rad. Bio. Med. 27(11/12): 1173-1181.
Rice-Evans, C.A., Miller, N.J. and Paganaga, G. 1996. Structure-antioxidant
activity relationships of flavonoids and phenolic acids. Free Rad. Bio. Med. 20(7):
933-956.
Robards, K., Prenzler, P., Tucker, G., Swatsitang, P., Glover, W. 1999. Phenolic
compounds and their role in oxidative processes in fruits. Food Chem. 66: 401-46.
Rodriguez-Saona, L.E., Wrolstad, R.E. 2000. Current Protocols in Analytical
Chemistry, Part F-Colors, Chapter F2: Anthocyanins, Unit F2.1: Extraction,
Isolation and Purification of Anthocyanins. In Press.
Schaller, D.R. and Von Elbe, J.H. 1968. The minor pigment component of
Monmorency cherries. J. Food Sci. 3: 442-443.
Scharrer, E., Geary, N. and Grothschel, H. 1981. Meal Patterns and body weight
changes during insulin hyperphagia and postinsulin hypophagia. Behav. Neural
Bio. 31(4): 435-442.
31
Shahrzad, S., and Bitsch, I. 1996. Determination of some pharmacologically
active phenolic acids in juices by high-performance liquid chromatography. J.
Chroma. 741: 223-231.
Slinkard, K.W., Singleton, V.L. 1984. Phenol content of grape skins and the loss
of ability to make anthocyanins by mutation. Vitis. 46: 4592-4597.
Somers, T.C. and Evans, M.E. 1974. Wine quality: correlation's with color
density and anthocyanin equilibri in a group of young red wines. J. Sci. Fd. Agric.
25: 1369-1379.
Wang, H., Nair, M., Stasburg, G., Booren, A. Gray, J. 1999. Antioxidant
polyphenols from tart cherries (Prunus cerasus). J. Nat.-Prod. 62: 1, 86-88.
Willstatter, R. and Zollinger, E.H. 1916. Investigation on anthocyanin. XIV.
Pigments of cherries and wild plums. Justus Liebigs Ann. Chem. 412: 1654.
Wrolstad, R.E. Color and Pigment Analyses in Fruit Products. 1978. Agricultural
Experiment Station, Oregon State University Station Bulletin 624.
32
APPENDIX 1
Anthocyanin general chemical structure with molecule substitutions.
R1=R2=H
Pelargonidin
R1 = OH, R2 = H
R1 = OCH3, R2 = H
Cyanidin
R1=R2=OH
R1 = OCH3, R2 = OH
R1 = R2 = OCH3
Peonidin
Delphinidin
Petunidin
Malvidin
R1
33
APPENDIX 2
Color and distribution of anthocyanidins in common fruits.
Anthocyanidins
Color
Example
Delphinidin
Bluish-red
Black currant
Cyanidin
Orange-red
Cherry
Petunidin
Bluish-red
Blueberry
Pelargonidin
Orange
Strawberry
Malvidin
Bluish-red
Grape
Peonidin
Red
Cranberry
34
Download