Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)
Article
LC-MS untargeted metabolomics to explain the signal metabolites inducing browning in fresh-cut lettuce Carlos J. Garcia, Rocio Garcia-Villalba, Maria I. Gil, and Francisco A. Tomas-Barberan J. Agric. Food Chem., Just Accepted Manuscript • Publication Date (Web): 16 May 2017 Downloaded from http://pubs.acs.org on May 16, 2017
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Agricultural and Food Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 35
Journal of Agricultural and Food Chemistry
1
LC-MS untargeted metabolomics to explain the signal metabolites inducing
2
browning in fresh-cut lettuce
3 4
Carlos J. García, Rocío García-Villalba, María I. Gil, Francisco A. Tomas-Barberan*
5 6 7
Affiliation: Research group on Quality, Safety, and Bioactivity of Plant Foods. CEBAS
8
(CSIC), P.O. Box 164, Espinardo, Murcia 30100, Spain
9 10
Revista: Journal of Agricultural and Food Chemistry.
11
Section: Omics applied to Food and Agriculture.
12
*Corresponding author:
[email protected] 13
1 ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
Page 2 of 35
14
ABSTRACT: Enzymatic browning is one of the main causes of quality loss in lettuce
15
as a prepared and ready to eat cut salad. An untargeted metabolomics approach using
16
UPLC-ESI-QTOF-MS was performed to explain the wound response of lettuce after
17
cutting and to identify the metabolites responsible of browning. Two cultivars of
18
Romaine lettuce with different browning susceptibility were studied at short time
19
intervals after cutting. From the total 5975 entities obtained from the raw data after
20
alignment, filtration reduced the number of features to 2959, and the statistical analysis
21
found that only 1132 entities were significantly different. Principal component analysis
22
(PCA) clearly showed that these samples grouped according to cultivar and time after
23
cutting. From those, only fifteen metabolites belonging to lysophospholipids,
24
oxylipin/jasmonate metabolites, and phenolic compounds were able to explain the
25
browning process. These selected metabolites showed different trends after cutting;
26
some decreased rapidly, others increased but decreased then after, whereas others
27
increased during the whole period of storage. In general, the fast-browning cultivar
28
showed a faster wound response and a higher raw intensity of some key metabolites
29
than the slow-browning one. Just after cutting the fast-browning cultivar contained
30
eleven out of the fifteen browning-associated metabolites, while the slow-browning
31
cultivar only had five of them. These metabolites could be used as biomarkers in
32
breeding programs for the selection of lettuce cultivars with lower browning potential
33
for fresh-cut.
34 35
KEYWORDS: wound-signal, UPLC-ESI-QTOF-MS, PCA, phenolic metabolism,
36
minimal processing, lettuce
37 38
2 ACS Paragon Plus Environment
Page 3 of 35
Journal of Agricultural and Food Chemistry
39
3 ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
Page 4 of 35
40
INTRODUCTION
41
Lettuce (Lactuca sativa) is a major fresh vegetable crop of great economic importance
42
for the fresh-cut processing industry which demands products with longer shelf life.
43
Enzymatic browning is an important biochemical disorder that leads to significant
44
quality loss after lettuce is cut and is promoted during storage, causing product
45
rejection. Polyphenol-oxidase (PPO) is the main enzyme involved in the oxidative
46
browning reaction of cut lettuce. PPO is found in young plants in the thylakoid
47
membrane of chloroplasts and is released through membrane disruption .1 Wounding
48
induces activation from latent to fully active PPO, although the total lettuce PPO levels
49
(latent plus active) are not affected .2,3 Wound stress of plant tissues triggers an increase
50
in the levels of phenolic compounds, particularly phenylpropanoids. This wound-
51
induced phenolic biosynthesis is mediated through an increase in phenylalanine
52
ammonia-lyase (PAL) activity. PAL, the first enzyme in the biosynthesis of phenolic
53
PPO substrates via the phenylpropanoid pathway, increases immediately after lettuce
54
cutting and has been found to be closely related to browning development.4.5 The fresh-
55
cut process induces an abiotic stress response that triggers phenylpropanoid
56
metabolism.6 To date, the most investigated and widely accepted hypothesis of
57
browning development in lettuce is related to the presence of PAL and the constitutive
58
composition of phenolic compounds .7 Several authors have correlated the extent of
59
browning in cut lettuce with the content of phenolic compounds.8,9 Other secondary
60
metabolites, such as sesquiterpene lactones, have also been associated with lettuce
61
discoloration, although following a different pathway.10 There is a consensus in the
62
mechanism of wound-induced tissue browning that starts with a signal produced from
63
the damaged membranes that induces phenolic biosynthesis enzymes, particularly PAL,
64
mediated by jasmonic acid and ethylene.11,12 However, the changes at a metabolic level
4 ACS Paragon Plus Environment
Page 5 of 35
Journal of Agricultural and Food Chemistry
65
during the browning process have not been studied systematically using a metabolomics
66
approach.
67
In this context, metabolomics offers a powerful tool to study global changes in
68
the entire metabolite set using advanced analytical technologies combined with
69
multivariate statistical analysis13,14 and has been applied to understand the composition
70
and origin of plant-derived foods.15,16 The potential of metabolomics has been applied
71
recently to better characterize the lettuce metabolome through NMR17, GC-TOF MS18,
72
and UPLC-QTOF.19 These techniques have already been used in several studies
73
associated with the exposure of lettuce to arsenic-contaminated irrigation water20, and
74
pesticides21, and the effect of lettuce transgenic modification.22A protocol for studying
75
the metabolome of cut lettuce has recently been reported.23 However, to the best of our
76
knowledge, untargeted metabolomics has not been explored to understand the browning
77
process so far.
78
The aim of this study was to get insight into the global metabolic changes
79
occurring after lettuce cutting to identify the responsible metabolites related to the
80
wound response and the subsequent browning process. To this end, an untargeted
81
metabolomics approach using UPLC-ESI-QTOF was applied to evaluate two Romaine
82
lettuce cultivars with different browning susceptibility after cutting.
83 84
MATERIALS AND METHODS
85
Chemicals and Reagents. Methanol and acetonitrile were from J.T Baker
86
(Deventer, The Netherlands) and formic acid was from Panreac (Barcelona, Spain). All
87
other chemicals and reagents were of analytical grade. Milli-Q system (Millipore Corp.,
88
Bedford, MA, USA) ultrapure water was used throughout the study. Authentic
5 ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
Page 6 of 35
89
standards of phenolic metabolites were obtained from the CEBAS-CSIC phenolic
90
standards collection.
91
Plant material. Two Romaine lettuce cultivars were selected, FB (fast
92
browning) and SB (slow browning), after an intensive screening of lettuce genotypes
93
with different browning susceptibility.24 Lettuces were harvested in Torrepacheco
94
(Murcia, Spain) in March 2014 when heads reached the commercial maturity stage
95
based on head weight and core length.25
96
Processing and storage conditions. Lettuces were harvested the day before
97
processing and stored for 24 h at 7 ºC and 80% RH. Lettuce processing was as
98
previously described.23 Six leaves from the middle part of the head were selected.
99
Midribs were carefully excised and cut in pieces of 1.5 cm length in less than 20 min to
100
reduce as much as possible the induction of the wound response. Samples of 30 g were
101
packaged in plastic bags (120mm x 200 mm) without sealing to maintain an air
102
atmosphere and high humidity avoiding water loss. Midrib samples were stored for 2, 6,
103
24, 48, 72, 96 and 120 h after cutting at 7 ºC. Plant material was either frozen
104
immediately using liquid nitrogen (samples at 0 h) or after the different time intervals,
105
then freeze-dried and finally milled to a fine powder. Four replicates (corresponding to
106
four heads) per cultivar and time after cutting were processed.
107
Digital image analysis. Images of midribs were captured using a digital camera
108
(Nikon D7100) in a darkness room. The set up conditions for the camera were: lens
109
(Nikon Advanced Multi-CAM 3500DX), light provided by a flash (Metz Mecablitz)
110
with a speed of 1/125s, color temperature of 5600 K, focal opening 22 (f) and length 35
111
mm. In each image, 20 midrib pieces were placed together facing up the cut edge. At
112
each sampling time, an image of a grey card was captured, in the same conditions for
113
the balance of the image colors. Four images, one per replicate, were captured per
6 ACS Paragon Plus Environment
Page 7 of 35
Journal of Agricultural and Food Chemistry
114
cultivar at 0 h and 120 h after storage and saved as JPEG file. Images were analyzed
115
using the software ImageJ version 1.48v (NIH Image, National Institute of Health,
116
Bethesda, USA) as reported previously.23 The total cut edge area was selected by
117
contrast and considered as the study area. The extension of tissue browning was
118
evaluated as: % Browning =
Browned pixels Total pixels area
119
The RGB color model was used to identify the browned pixels in the range of
120
hue (20-35), saturation (40-195) and brightness (0-225). Values of hue, saturation, and
121
brightness were measured following the HSB color model. One-way analyses of
122
variance (ANOVA) with a significant level of p < 0.05 were performed. Bilateral
123
correlations were determined by Pearson’s correlation with a confidence interval
124
established at 95% using PASW Statistics 20 for Windows (SPSS Inc., Chicago, IL,
125
USA).
126
Untargeted metabolomics analysis. Compounds were extracted according to
127
the method described by Llorach et al (2008)26 with minor modifications. Freeze dried
128
samples (0.3 g) were mixed with 10 mL of methanol:water (80:20). The extraction
129
mixture was sonicated at 30 °C for 30 min, centrifuged at 5,000 g for 15 min and
130
directly filtered through 0.22 µm PVDF filter to be analyzed by UPLC-ESI-QTOF.
131
Chromatographic separation was performed on an UPLC system (Agilent 1290
132
Infinity, Agilent Technologies, Waldbronn, Germany) coupled to Mass QTOF (6550
133
Accurate) with electrospray ionization via Jet Stream Technology and a C18 column
134
(Poroshell 120, 3 x 100 mm, 2.7 µm pore size). Chromatographic and mass
135
spectrometry conditions were the same as described previously.23 Briefly, 3 µL of
136
methanolic extracts were injected and separated on the column with a flow rate of 0.4
137
mL/min and mobile phases of water + 0.1% formic acid (A) and acetonitrile (ACN) + 7 ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
Page 8 of 35
138
0.1% formic acid (B). Nitrogen was used as nebulizer (35 psi, 9 L/min) and drying gas
139
(280 ºC, 9 L/min, sheath gas temperature 400 ºC and sheath gas flow 12 L/min). Spectra
140
were acquired in the range 100-1100 m/z in negative mode, with a fragmentor voltage of
141
100 V and acquisition rate 1.5 spectra/s. Samples were also analyzed in the positive
142
mode, but in this case the differences between cultivars were less significant and
143
therefore were not further studied. Targeted MS/MS analyses were developed to add
144
confidence to compound identification. MS/MS product ion spectra were collected at an
145
m/z range of 50–800 using a retention time window of min, an isolation window of 4.0
146
amu, a collision energy of 30 V and an acquisition rate of 4 spectra/s.
147
The main parameters for data extraction and pre-processing were previously
148
optimized23 (Figure 1). Briefly, Mass Hunter Qualitative Analysis software (version
149
B.06.00 Agilent Technologies) was used for feature (ion) extraction applying molecular
150
feature extraction algorithm (MFE). Parameters were set as minimal intensity 5000
151
counts, m/z range 100-1100, charge state and the lowest number of ions in the isotopic
152
distribution 2 with a peak spacing tolerance of 0.0025 m/z plus 7 ppm. Once ions were
153
extracted, files were transformed to .cef format and exported to Mass Profiler
154
Professional (MPP) software (version B.12.01, Agilent Technologies, Santa Clara, CA,
155
USA). Samples were classified into different experimental groups (named conditions)
156
considering two variables: time after cutting (0, 2, 6, 24, 48, 72, 96 and 120 h) and
157
cultivars (FB and SB). Compounds from different samples were aligned with a retention
158
time window of 0.15 min and a mass window of 5 ppm + 2 mDa. Then filter by flag
159
was applied considering only the molecular features that were present in at least 100%
160
of replicates in one condition. Data were log transformed and mean centered before
161
statistical analysis.
8 ACS Paragon Plus Environment
Page 9 of 35
Journal of Agricultural and Food Chemistry
162
Two-way ANOVA analysis with variables time after cutting and cultivar was
163
applied to the pretreated samples considering p