Accepted Manuscript Title: Liquid chromatography–mass spectrometry coupled with multivariate analysis for the characterization and discrimination of extractable and nonextractable polyphenols and glucosinolates from red cabbage and Brussels sprout waste streams Author: Gerard Bryan Gonzales Katleen Raes Hanne Vanhoutte Sofie Coelus Guy Smagghe John Van Camp PII: DOI: Reference:
S0021-9673(15)00683-4 http://dx.doi.org/doi:10.1016/j.chroma.2015.05.009 CHROMA 356490
To appear in:
Journal of Chromatography A
Received date: Revised date: Accepted date:
13-2-2015 16-4-2015 6-5-2015
Please cite this article as: G.B. Gonzales, K. Raes, H. Vanhoutte, S. Coelus, G. Smagghe, J. Van Camp, Liquid chromatographyndashmass spectrometry coupled with multivariate analysis for the characterization and discrimination of extractable and nonextractable polyphenols and glucosinolates from red cabbage and Brussels sprout waste streams, Journal of Chromatography A (2015), http://dx.doi.org/10.1016/j.chroma.2015.05.009 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.
Highlights Red cabbage and Brussels sprouts waste contain high total polyphenol content Nonextractable fraction contains more total phenolics than the extractable fraction Polyphenols and glucosinolates from Brassica waste were identified using LC-MS PCA and OPLS-DA were used to distinguish EP from NEP fractions
te
d
M
an
us
cr
ip t
Ac ce p
1 2 3 4 5 6 7
1 Page 1 of 37
7
Liquid chromatography – mass spectrometry coupled with multivariate analysis for the
8
characterization and discrimination of extractable and nonextractable polyphenols and
9
glucosinolates from red cabbage and Brussels sprout waste streams
ip t
10
Gerard Bryan Gonzalesabc, Katleen Raesb, Hanne Vanhouttea, Sofie Coelusa, Guy Smagghec,
12
John Van Campa*
cr
11
13 14
a
15
University, Ghent, Belgium, bDepartment of Industrial Biological Science, Faculty of
16
Bioscience Engineering, Ghent University, Kortrijk, Belgium,
17
Protection, Faculty of Bioscience Engineering, Ghent University, Ghent, Belgium
c
Department of Crop
M
an
us
Department of Food Safety and Food Quality, Faculty of Bioscience Engineering, Ghent
18
*Corresponding author:
21
Prof. dr. John Van Camp
22
[email protected] 23
telephone:
+32-(0)9
Ac ce p
20
te
d
19
264
62
08
2 Page 2 of 37
24 25
Abstract
26
Nonextractable polyphenol (NEP) fractions are usually ignored because conventional
28
extraction methods do not release them from the plant matrix. In this study, we optimized the
29
conditions for sonicated alkaline hydrolysis to the residues left after conventional polyphenol
30
extraction of Brussels sprouts top (80°C, 4M NaOH, 30 mins) and stalks (60°C, 4M NaOH,
31
30 mins), and red cabbage waste streams (80°C, 4M NaOH, 45 mins)
32
characterize the NEP fraction. The NEP fractions of Brussels sprouts top (4.8±1.2 mg gallic
33
acid equivalents [GAE]/g dry waste) and stalks (3.3±0.2 mg GAE/g dry waste), and red
34
cabbage (11.5 mg GAE/g dry waste) waste have significantly higher total polyphenol contents
35
compared to their respective extractable polyphenol (EP) fractions (1.5±0.0, 2.0±0.0 and
36
3.7±0.0 mg GAE/g dry waste, respectively). An LC-MS method combined with principal
37
components analysis (PCA) and orthogonal partial least squares – discriminant analysis
38
(OPLS-DA) was used to tentatively identify and discriminate the polyphenol and
39
glucosinolate composition of the EP and NEP fractions. Results revealed that phenolic
40
profiles of the EP and NEP fractions are different and some compounds are only found in
41
either fraction in all of the plant matrices. This suggests the need to account both fractions
42
when analyzing the polyphenol and glucosinolate profiles of plant matrices to attain a global
43
view of their composition. This is the first report on the discrimination of the phenolic and
44
glucosinolate profiles of the EP and NEP fractions using metabolomics techniques..
to release and
Ac ce p
te
d
M
an
us
cr
ip t
27
45 46
Keywords: Principal components analysis (PCA), orthogonal partial least squares –
47
discriminant analysis (OPLS-DA), polyphenols, Brassica waste, liquid chromatography, mass
48
spectrometry
3 Page 3 of 37
49 50
1. Introduction
51
Phenolic compounds comprise a diverse group of bioactive compounds found in nature and
53
are widely distributed as secondary metabolites in plants and hence part of the human diet.
54
Currently, the phenolic structure of about 10.000 compounds has been described in literature
55
[1]. Renewed interest in the study of phenolic compounds arose when they were discovered to
56
be powerful antioxidants, followed by numerous studies focusing on their biological and
57
bioactive characteristics [2,3]. So far, most of the studies only refer to extractable phenolics
58
(EP) and do not consider the nonextractable phenolic fraction (NEP) and thus are often
59
overlooked by current literature [4,5]. The NEP fraction comprises of phenolic compounds
60
that are bound or trapped in the plant matrix and consequently remain in the residue after
61
extraction with aqueous-organic solvents.
cr
us
an
M
d te
62
ip t
52
Increasing food waste has been a growing concern in modern society. Efforts to reduce food
64
waste has been the subject of many academic and non-academic fora. Valorization of
65
agricultural wastes is thus a major step in alleviating this problem. We have previously shown
66
that agricultural wastes also possess high amounts of polyphenols, which could be harnessed
67
for use in food applications, such as functional ingredients, antioxidants, etc [6,7]. If these
68
bioactive components are recovered from the waste stream, they could be used as additives to
69
food and/or cosmetics to create high-value products. It has earlier been reported that higher
70
amounts of phenolics are found in the NEP fraction of agricultural waste streams compared to
71
their EP fractions [8-10]. Exploiting this fraction therefore results to better valorization of the
72
waste streams. However, the differences in the phenolic profiles of the extractable and
73
nonextractable fractions have not been deeply studied. In this paper, we investigate the EP
Ac ce p
63
4 Page 4 of 37
and NEP fractions of two different waste streams belonging to the Brassica family, Brussels
75
sprouts and red cabbage. Initially, the EP fraction was obtained by conventional solvent
76
extraction and the phenolic composition was characterized. Thereafter, the residue left after
77
solvent extraction was collected and the parameters for NEP extraction were optimized for
78
each waste stream. This is the first report about the EP and NEP characterization of Brussels
79
sprouts and red cabbage waste streams. Also in this study, we show the use ultrahigh-pressure
80
liquid chromatography – mass spectrometry combined with metabolomics-based analysis
81
tools, principal components analysis (PCA) and orthogonal partial least squares –
82
discriminant analysis (OPLS-DA), to discriminate the phenolic profile of the EP from NEP
83
fractions and to determine which compounds cause their distinction. This provides a rapid and
84
convenient analytical method for screening and characterizing EP and NEP without the need
85
for quantification of the individual components or manual integration of each
86
chromatographic peak.
d
M
an
us
cr
ip t
74
89 90 91
2. Materials and Methods
Ac ce p
88
te
87
2.1.
Reagents and plant material
92
U(H)PLC–MS grade methanol and formic acid were acquired from Biosolve (Valkenswaard,
93
the Netherlands) whereas analytical grade methanol, HCl and NaOH were purchased from
94
VWR International (Leuven, Belgium).
95 96
Red cabbage (Brassica oleracea var. capital f. rubra) and Brussels sprouts (Brassica oleracea
97
var. gemmifera) waste were harvested in November 2013. For Brussels sprouts, the top leafy
98
part was separated from the stalks and analyzed separately due to their big structural
5 Page 5 of 37
difference, while the sample material of red cabbage consisted only of the external leaves.
100
The plant materials were cut, blended into smaller pieces and immediately stored in a freezer
101
at -20°C. Approximately 250 grams of each plant material were freeze-dried and ground into
102
a fine powder with an IKA-M20 Werke Grinder and stored at -20°C until further analysis.
ip t
99
103
2.2.
Solvent extraction of EP and collection of residues containing NEP
cr
104 105
The solvent extraction protocol was based on the method by Olsen et al [11]. Initially, 2
107
grams of the freeze-dried plant powder was weighed and placed in 50 mL centrifuge tubes
108
with 15 mL of 100% MeOH and homogenized using an IKA T25 digital Ultraturrax at 10.000
109
rpm for 45 seconds. The tubes were then placed on ice for 15 minutes and centrifuged at
110
13000 g for 10 minutes at 4°C. The supernatant (1) was collected in a 50 mL volumetric flask
111
while the residue left in the tube was re-extracted with 10 mL 80% MeOH and re-
112
homogenized at 10.000 rpm for 45 seconds, placed on ice for 15 minutes and then centrifuged
113
at 13.000 g for 10 minutes at 4°C. The supernatant was added to supernatant (1) and the
114
volume was adjusted to 50 mL using 100% MeOH. Subsequently, the residue left in the
115
centrifuge tubes after extraction was dried under reduced pressure and was used to extract
116
NEP.
118 119
an
M
d
te
Ac ce p
117
us
106
2.3.
Measurement of total phenolic content (TPC)
120
The total phenolic content (TPC) of the extracts was determined with the colorimetric Folin-
121
Ciocalteau assay previously optimized [7]. Briefly, 1.5 mL cuvettes were filled with 1200 µL
122
distilled water, 50 µL of the plant extract and 100 µL Folin-Ciocalteau phenol reagent (diluted
123
10 times in distilled water). For making the calibration curves, 50µL gallic acid was placed
6 Page 6 of 37
instead of the sample with concentrations of 0 to 250 µg gallic acid mL-1. Thereafter, the
125
cuvettes were incubated in the dark for 5 minutes and the solution was then mixed with 150
126
µL of sodium carbonate (Na2CO3). Finally, the cuvettes were incubated for 2 hours in the
127
dark at room temperature, immediately followed by measuring the resulting chromophores
128
with a Varian Cary 50 Series spectrophotometer at a wavelength of 760 nm. Total phenolic
129
content was then expressed as mg gallic acid equivalents (GAE) per gram dried plant
130
material.
cr
ip t
124
133
2.4.
Sonicated alkaline hydrolysis of the residue left after solvent
an
132
us
131
extraction
M
134
The optimization of the parameters for the sonicated alkaline hydrolysis is summarized in
136
Table 1, comprising of 27 combinations, which were analyzed in triplicates.
te
137
d
135
Briefly, 0.1 gram of the residue was placed in a tube and mixed with 2 mL of NaOH (1M, 2M
139
or 4M). The tubes were flushed with nitrogen for 30 seconds and sealed to prevent the
140
oxidation of the phenolic compounds. Furthermore, the samples were incubated in a
141
temperature-controlled ultrasonic water bath in an Elmasonic S60H unit with a frequency of
142
37kHz and a nominal power of 180W. The temperatures (40°C, 60°C and 80°C) and
143
incubation times (15, 30 and 45 min) were varied depending on the set-up as earlier described
144
in Table 1. Due to heating during sonication, the temperature in the water bath was checked
145
every five minutes and adjusted if necessary.
Ac ce p
138
146 147
After hydrolysis, the samples were neutralized by adding HCl. The liberated NEP was then
148
extracted using 4 mL of MeOH (0.1% formic acid) followed by vortexing for 2 minutes. The
7 Page 7 of 37
149
tubes were centrifuged at 10.000 g for 10 minutes at 4°C. Extraction was performed twice and
150
the final volume was adjusted to 20 mL using 100% methanol.
151
2.5.
Solid phase extraction (SPE)
ip t
152 153
Aliquots (1mL) of the polyphenol fractions were diluted in 20 mL of 0.1% (vol) formic acid
155
(in ultrapure water) and loaded into a preconditioned C18 solid phase extraction (SPE)
156
cartridge (500 mg per 4 mL) (Davison Discover Science, Deerfield, IL, USA). Columns were
157
preconditioned by loading 2 × 3 mL methanol and 2 × 3 mL water, wherein each solvent was
158
allowed to stand for 2 min prior to use. After loading the sample, the columns were washed
159
with 5 mL of MilliQ water (0.1% (vol) formic acid). The polyphenols were recovered using 3
160
mL MeOH (0.1% (vol) formic acid). The samples were then dried using light stream of
161
nitrogen and re-dissolved in 1 mL of 10% dimethyl sulfoxide (DMSO) in acidified water prior
162
to LC-MS analysis.
165
2.6.
Identification of compounds using U(H)PLC-ESI-MS
Ac ce p
164
te
163
d
M
an
us
cr
154
166
LC-MS analysis was performed with a Waters Acquity UPLC system (Waters Corp., Milford,
167
MA, USA) connected to a Synapt HDMS-TOF-MS (Waters Corp., Milford, MA, USA). Plant
168
extract components were separated using a Waters Acquity BEH C18 column (2.1 mm × 150
169
mm, 1.7 µm particle size) attached to a Waters VanGuard Pre-column (2.1 mm × 5 mm)
170
during gradient elution with a flow rate of 250 µL min-1, as earlier reported [6]. The mobile
171
phase was composed of (A) water containing 0.1% (vol) formic acid and (B) methanol
172
containing 0.1% (vol) formic acid at a controlled temperature of 40 °C. The elution program
173
was as follows: 0 min, 10% B; 0–6 min, 0–20% B linear; 6–12 min, 20% B isocratic; 12–13
8 Page 8 of 37
min, 20–30% B linear; 13–23 min, 30–50% B linear; 23–30 min, 50–90% B linear; 30–35
175
min, 90% isocratic; 35–40 min, 90–10% B linear; and 40–45 min 10% B isocratic. Ionization
176
was done with an electrospray ionization (ESI) source. Data were acquired in continuum
177
negative ionization in V-mode and, additionally, in positive mode for the red cabbage
178
samples. Source parameters for the MS were set as follows: capillary voltage, 2 kV; sampling
179
cone voltage, 40 V; extraction cone voltage, 4 V; source temperature 150 °C; desolvation
180
temperature, 350 °C; cone gas flow rate, 50 L h-1; desolvation gas flow rate, 550 L h-1.
181
Collision energies were set at 6 V for the low energy and 45 V for high energy. Mass range
182
was set at 100–1500 Da with a scan speed of 0.2 s per scan using the MassLynx software 4.1
183
(Waters Corp., Milford, MA,USA). Structural characterization of the flavonoids were based
184
on an earlier published method [6] while other components were tentatively identified by their
185
exact mass and fragmentation patterns which was viewed using the MSE data viewer software.
M
an
us
cr
ip t
174
2.7.
Data analysis
te
187
d
186
188
Data for the optimization of NEP yield (based on TPC) were analyzed by ANOVA using
190
SPSS v.21 (IBM, Chicago, USA) while comparison between means (post-hoc) was achieved
191
using Tukey test. Differences were considered significant when p