Subscriber access provided by UNIV OF DURHAM
Omics Technologies Applied to Agriculture and Food
Reduced dietary levels of EPA and DHA have a major impact on the composition of skin membrane lipids in Atlantic salmon (Salmo salar L.) ken cheng, Marta Bou, Bente Buyter, Jana Pickova, Emad Ehtesham, Liang Du, Claudia Venegas, and Ali A. Moazzami J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.8b02886 • Publication Date (Web): 25 Jul 2018 Downloaded from http://pubs.acs.org on July 30, 2018
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 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 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.
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 31
Journal of Agricultural and Food Chemistry
1
Reduced dietary levels of EPA and DHA have a major impact on the composition of
2
skin membrane lipids in Atlantic salmon (Salmo salar L.)
3
Ken Chenga, *, Marta Boub, Bente Ruyterb, Jana Pickovaa, Emad Ehteshama, Liang Dub,
4
Claudia Venegasc, Ali A. Moazzamia
5
a
6
Sciences, Uppsala, Sweden
7
b
8
c
9
*Correspondent at Department of Molecular Sciences, Uppsala BioCenter, Swedish
10
University of Agricultural Sciences, P.O. Box 7015, 75007 Uppsala, Sweden; Tel: +46 18
11
6720 11; E-mail address: cheng-ken@live.cn (K. Cheng)
Department of Molecular Sciences, Uppsala BioCenter, Swedish University of Agricultural
Nofima (Norwegian Institute of Food, Fisheries and Aquaculture Research), Ås, Norway
AVS Chile, Puerto Varas, Chile
ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
Page 2 of 31
12
Abstract
13
Membrane lipids, including sphingolipids and glycerol-phospholipids, are essential in
14
maintaining skin barrier function in mammals, but their composition in fish skin and their
15
response to diets have not been evaluated. This study investigated the impacts of reducing
16
dietary EPA and DHA on membrane lipids in skin of Atlantic salmon, through a 26-week
17
feeding regime supplying different levels (0%‒2.0% of drymass) of EPA/DHA. Ceramide,
18
glucosylceramide, sphingomyelin, sphingosine, and sphinganine in salmon skin were
19
analyzed for the first time. Higher concentrations of glucosylceramide and sphingomyelin,
20
and higher ratios of glucosylceramide:ceramide and sphingomyelin:ceramide, were detected
21
in the deficient group, indicating interruptions in sphingolipidomics. Changes in the glycerol-
22
phospholipid profile in fish skin caused by reducing dietary EPA and DHA were observed.
23
There were no dietary impacts on epidermal thickness and mucus cell density, but the
24
changes in phospholipid profile suggest that low dietary EPA and DHA may interrupt the
25
barrier function of fish skin.
26
Keywords:
27
sphingolipidomics
ceramide,
DHA,
EPA,
glycerol-phospholipids,
ACS Paragon Plus Environment
fish
skin
health,
2
Page 3 of 31
Journal of Agricultural and Food Chemistry
28
1 Introduction
29
The n-3 long-chain polyunsaturated fatty acids (LC-PUFA), mainly eicosapentaenoic acid
30
(EPA; 20:5n-3) and docosahexaenoic acid (DHA, 22:6n-3), have been identified as essential
31
fatty acids (EFA) in the diet of Atlantic salmon (Salmo salar L.) for good growth
32
performance, health, and final product quality
33
used to date in salmon diets, but due to relatively stable fish oil production and growing
34
global demand for farmed fish, the aquaculture industry is facing a challenge in meeting the
35
demand for fish oil in Atlantic salmon production
36
such as algae, krill, and genetically modified plant oils, have been the subject of extensive
37
research, but so far this has not yielded an economically and ecologically sustainable solution
38
for salmon farming
39
currently inevitable. It is important to know the possible impacts of reduced dietary EPA and
40
DHA on salmon growth and health. There is still a knowledge gap on Atlantic salmon
41
requirements for dietary EPA and DHA under different environmental conditions. Under
42
controlled environment in tanks on land, 10g/kg EPA and DHA (1% of feed dry mass) is in
43
general considered to be sufficient
44
require above 10 g/kg to maintain fish robustness and good health under demanding
45
environmental conditions in sea cages.
46
The importance of EPA and DHA on fish performance has been studied previously, mostly
47
focusing on the impacts on fish growth, survival and early development, and fatty acid (FA)
48
composition in fish liver and muscle
49
effects of dietary EPA and DHA on fish skin health 11. As with terrestrial vertebrate skin, fish
50
skin acts as the main barrier to the external environment, maintaining homeostasis in the
51
organism and protecting against potential physical damage and environmental pathogens 16.
52
However, unlike human skin, the fish epidermis lacks a keratinised layer (stratum corneum)
1–3
. Fish oil rich in n-3 LC-PUFA has been
4,5
. Alternative sources of n-3 LC-PUFA,
6–9
. As a result, a reduction in n-3 LC-PUFA levels in salmon feed is
1,10–12
. However, a recent study
11
showed that salmon
1,2,13–15
. Very few experiments have investigated the
ACS Paragon Plus Environment
3
Journal of Agricultural and Food Chemistry
Page 4 of 31
17
53
and hairs, but it contains a mucus layer and bone tissue-related scales
. The mucus layer
54
contains anti-microbial and anti-infection enzymes such as lysozyme, protease, and
55
immunoglobulin, which are important for fish skin health 17–19. Their immunological enzyme
56
activities have been found to be implicated in fish epidermis histological parameters, such as
57
epidermal thickness and mucus cell density
58
period, the density of mucus cells, mainly goblet cells, was found to be positively correlated
59
with epidermal layer thickness, but negatively correlated with parasite density 20,22.
60
The permeability barrier of skin is primarily localized at the stratum corneum in terrestrial
61
vertebrates. Ceramide (Cer), composed of a sphingosine (So) and a fatty acid (FA), is the
62
main lipid (>50% of total lipid mass) in the stratum corneum
63
EFA deficiency results in impaired sphingolipid metabolism, such as conversion of
64
sphingomyelin (Sph) and glucosyl-ceramide (GlcCer) into Cer, leading to abnormal
65
permeability barrier function in the epidermis of mammals
66
the skin of terrestrial vertebrates in structure, many essential functions are shared, such as
67
mechanical and chemical barrier formation to maintain osmotic homeostasis 17. To the best of
68
our knowledge, only one previous publication has determined the total content of
69
sphingolipids, including Sph and GlcCer, in fish skin, in a study on Pacific saury (Cololabis
70
saira) using high-performance liquid chromatography (HPLC) coupled with evaporative light
71
scattering detection (ELSD) 26. The composition and function of Cer and related sphingolipid
72
metabolites, such as Sph, GlcCer, So, and sphinganine (Sa), in fish skin, and their responses
73
to dietary treatments, are still unknown.
74
Glycerol-phospholipids (GPL), phosphatidylcholine (PC), phosphatidylethanolamine (PE),
75
phosphatidylserine (PS), and phosphatidylinositol (PI) are other important types of membrane
76
lipids in the epidermis
77
fluidity of cell membranes, which is important for signal transduction and substance
20,21
. During a six-week experimental infection
23
. It has been reported that
24,25
. Although fish skin is unlike
27
. The PUFA in GPL are essential components in maintaining the
ACS Paragon Plus Environment
4
Page 5 of 31
Journal of Agricultural and Food Chemistry
28,29
78
transportation
. Lowered levels of n-6 PUFA, especially arachidonic acid (20:4n-6), and
79
elevated levels of monounsaturated fatty acids (MUFA) have been observed in epidermal PC
80
and PE in patients with atopic dermatitis
81
(Oncorhynchus mykiss) showed that an EFA-deficient diet containing 93.4% saturated FA
82
strongly influenced the GPL composition in fish skin, while no changes were detected in the
83
permeability to water 29. The function and biosynthesis of GPL in fish skin is still not clear,
84
which makes it interesting to determine the FA composition in GPL subclasses in skin when
85
feeding diets deficient in EPA and DHA.
86
The aim of the present study was thus to investigate the impacts of lowering dietary EPA and
87
DHA levels on phospholipids in skin of Atlantic salmon. The composition of sphingolipids
88
and FA composition in GPL subclasses (PC, PE, PS, and PI) in skin and epidermal
89
histological parameters (epidermal thickness and goblet cell density) were examined. The
90
effects of feeding duration were also evaluated.
91
2 Materials and methods
92
2.1 Fish feed formulation
93
Thirteen experimental diets with different levels of EPA and DHA were formulated in the
94
study. The feed ingredients are thoroughly described in another paper
95
experimental diets were iso-proteic (46.6-47.0%), iso-lipidic (24.6-25.9%) and iso-energetic
96
(22.1-22.6 MJ/kg), but contain 0%, 0.5%, 1.0%, 1.5% and 2.0% (of feed dry weight) of either
97
only EPA, only DHA, or 1:1 mixture of EPA and DHA (EPA+DHA, Table 1). Among these,
98
a diet completely depleted in EPA and DHA (0% EPA+DHA) was used as a negative control
99
diet. The experimental diets were fishmeal and fish oil free, but carefully designed to meet
100
fish nutritional requirements. Blended poultry oil and rapeseed oil (1:1) which are naturally
101
lacking in EPA and DHA, were used as basic lipid sources in the experimental feeds. EPA
102
and DHA oil concentrates in the form of triacylglycerol (Croda Chemicals Europe Ltd., East
30
. Moreover, a study on rainbow trout
ACS Paragon Plus Environment
12
. Briefly, the
5
Journal of Agricultural and Food Chemistry
Page 6 of 31
103
Yorkshire, UK) were used to control dietary levels of EPA and DHA. All experimental diets
104
were produced by Nofima feed technology center (Bergen, Norway). The measured chemical
105
composition and gross energy in the experimental fish feeds are provided in Table S1.
106
A diet resembling a commercial diet with 2.2% 1:1 mixture of EPA and DHA (BioMar,
107
Trondheim, Norway) was included as a commercial-type control (CC), in which 26%
108
fishmeal and 9.8% fish oil were used. The main purpose of using the CC was to set a
109
benchmark for growth.
110
The FA composition in all diets were described in Bou et al.
111
18:3n-3, the precursor of EPA and DHA in biosynthetic pathway was kept at the same level
112
(about 4.7% of total FA) in all diets. The EPA and/or DHA dietary groups contained
113
increased content of EPA and DHA as it was designed, and 0% EPA+DHA diet had little
114
EPA (0.05% of total FA) and DHA (0.08% of total FA).
115
2.2 Experimental design
116
The feeding trial conditions are described in detail in Bou et al. 12. In brief, Atlantic salmon
117
with mean initial body weight of 52.8 g were randomly distributed into 33 tanks, and 70
118
fish/tank (2 tanks/diet for the 0.5, 1.0, and 1.5 % dietary groups, 3 tanks/diet for the CC, 0%
119
EPA+DHA, and 2.0% dietary groups; Table 1), and reared at Nofima Institute in
120
Sunndalsøra, Norway, for 26 weeks. All tanks (1 m2 surface area, 0.6 cm water depth) were
121
supplied with 15 L/min seawater (33 g/L salinity) at ambient temperature. The water
122
temperature varied between 6.3 °C and 13.8 °C and the oxygen saturation level was kept
123
above 85%. Prior to the experiment, the fish were fed a commercial diet (Skretting,
124
Stavanger, Norway) and treated with light to induce smoltification. Feed ration was 15−20%
125
higher than assessed feed intake and was supplied by automatic belt feeders.
126
Skin samples for lipid analysis were collected twice, following the same sampling
127
procedures, when fish reached a body weight of 182.9 ± 69.3 g (referred to as 200 g), after 19
ACS Paragon Plus Environment
12
. Importantly, the content of
6
Page 7 of 31
Journal of Agricultural and Food Chemistry
128
weeks of feeding, and when they reached a body weight of 379.7 ± 96.5 g (referred to as 400
129
g), after 26 weeks of feeding. Five fish were randomly selected from each tank and killed
130
using overdose MS 222 (0.05−0.08 g/L). Skin samples with mucus and scales from the right
131
fillet were dissected from the dorsal fin to the caudal fin, pooled by tank, and homogenized in
132
dry ice. The homogenate was kept at -40 °C, with the bags left open until the dry ice
133
evaporated, and thereafter stored at -80 °C until analysis. Skin covering the white muscle
134
from the Norwegian Quality Cut from the left fillet of fish were used for histology analysis.
135
Samples were randomly taken at termination of the experiment (at 400 g, after feeding for 26
136
weeks, n=5 fish per tank), cut into sizes (approximately 0.5 cm2) suitable for histological
137
analysis, and fixed in 10% buffered formalin. The experimental procedure was in accordance
138
with National Guidelines for Animal Care and Welfare published by the Norwegian Ministry
139
of Education and Research.
140
2.3 Sphingolipidomics analysis using LC-QToF MS
141
Sample preparation
142
Fish skin samples from eight groups (CC, 0% EPA+DHA, 0.5% EPA, 0.5% DHA, 0.5%
143
EPA+DHA, 2.0% EPA, 2.0% DHA, and 2.0% EPA+DHA) at stage 200 g and 400 g were
144
subjected to sphingolipidomics analysis, using methods described elsewhere 31,32. In brief, the
145
homogenized, pooled skin samples made out of five fish from each tank were analyzed three
146
times. The homogenate containing an internal standard cocktail (0.15 nmol of C17
147
sphingosine, C17 sphinganine, C17 sphingosine-1-phospate, C17 sphinganine-1-phosphate,
148
C12 sphingomyelin, C12 ceramide, C12 glucosyl(β)-ceramide, C12 lactosyl(β)-ceramide and
149
C12 ceramide-1-phosphate; sphingolipid mix II, LM-6005, Avanti Polar lipids, Alabaster,
150
AL, US) were extracted twice, using 3 mL chloroform:methanol (1:2, v/v) each time, under
151
sonication in a water-bath for 30 min at room temperature. The extract was centrifuged (1800
152
g, 20 min) at room temperature and the supernatant was collected.
ACS Paragon Plus Environment
7
Journal of Agricultural and Food Chemistry
Page 8 of 31
153
Since the amount of Sph in the skin samples was much higher than that of the other
154
sphingolipids measured (Cer, So, Sa, and GlcCer), the content of Sph was determined
155
separately. Skin extract (0.25 mL×2) was transferred to two tubes, one with C12:0 Sph
156
internal standard (0.17 nmol; Avanti Polar lipids, Alabaster, AL, US) and one without.
157
Sample solvent was evaporated under nitrogen and re-dissolved in 0.5 mL ethanol. The
158
remaining skin extract (5.2 mL) was used for quantification of the other sphingolipids. After
159
evaporation, samples were re-dissolved in 1 mL ethanol. All samples were centrifuged at 12
160
000 g for 20 min at 4 °C before analysis.
161
Liquid chromatography-mass spectrometry analysis
162
Liquid chromatography-mass spectrometry (LC-MS) was carried out on an HP1100 LC
163
system (Hewlett-Packard, Palo Alto, CA) coupled to an electrospray ionization-
164
quadropole/time of flight mass spectrometer (ESI-QTOF MS; Bruker maXis Impact; Bruker
165
Daltonik GmbH, Bremen, Germany). System integrity was controlled by HystarTM software
166
(Bruker Daltonik GmbH). A sodium formate solution (4 µL formic acid, 20 µL 1 M NaOH,
167
100 mL H2O, and 100 mL 2-propanol) was used as the MS calibrant to correct for any mass
168
drift in the analyte. The spectra were acquired in positive ionization mode scanning within a
169
50-1500 m/z range.
170
Analyte separation was performed on a hydrophilic interaction chromatograph (Atlantis silica
171
HILIC column, particle size 3 µm, 2.1 mm×150 mm, Waters, Wexford, Ireland). The
172
injection volume was 10 µL and the column temperature was maintained electronically at 30
173
°C. The mobile phase consisted of: eluent A, with 1% v/v formic acid, and 10 mM
174
ammonium formate in MS grade water; and eluent B, with 0.1% v/v formic acid in
175
acetonitrile, at a constant flow rate of 0.25 mL/min. The programmed eluent gradient started
176
initially with 95% A to 5% A for 0.5 min, ramping to 60% A for 10 min, which was held for
177
4.5 min by further ramping to 5% A for 2 min and held for 15 min before the next run. A
ACS Paragon Plus Environment
8
Page 9 of 31
Journal of Agricultural and Food Chemistry
178
plasma reference and a sphingolipid standard mixture (sphingolipid mix II, LM-6005, Avanti
179
Polar lipids, Alabaster, AL, US) were run three times throughout the analysis as a quality
180
control to check the stability of the instruments. The MS raw data were calibrated
181
automatically and converted to mzXML files using Compass DataAnalysis software (Bruker
182
Daltonik GmbH). Peak height gave good linearity when comparing QToF responses to a
183
standard Cer C17:0 (Larodan AB, Solna, Sweden) at different concentrations (0.1-1 µg/mL).
184
Therefore, peak height for the compounds of interests was calculated by Mzmine software
185
(version 2.15) based on their assigned m/z and retention time. The concentrations of
186
sphingolipids were determined against known amounts of internal standards and expressed in
187
nmol/g tissue. The contribution from overlapping signals from the
188
compounds was accounted for when relevant.
189
2.4 Fatty acid analysis of glycerol-phospholipids using TLC and GC-FID
190
Sample preparation
191
Total lipid in fish skin samples (2 g, from five fish per tank) was extracted with 50 mL
192
chloroform:methanol (2:1, v/v) containing 0.07% (w/v) butylated hydroxytoluene as
193
antioxidant and 6 mL NaCl (0.9%), according to the method described by Folch et al. 33. The
194
organic phase was collected and dried under a stream of nitrogen. The GPL fraction was
195
separated from the other lipid classes, such as triacylglycerol, diacylglycerol, and free FA, by
196
thin layer chromatography (TLC; silica gel 20×20 cm plates, Merck, Darmstadt, Germany)
197
using a mixture of petroleum ether, diethyl ether, and acetic acid (113:20:1, v/v/v) as mobile
198
phase, using the method described by Bou et al.
199
plates were sprayed with 2% 2,7-dichlorofluorecin in 96% ethanol. Lipid classes were
200
identified under ultraviolet (UV) light at 366 nm. The GPL bands were scraped off the plates
201
and soaked in a mixture of chloroform, methanol, acetic acid, and water (50:39:1:10, v/v/v/v)
202
for 4 hours at -40 °C, to elute GPL from silica gel. The GPL fractions were collected after
13
C isotope of other
12
and Thomassen et al.13. After drying, the
ACS Paragon Plus Environment
9
Journal of Agricultural and Food Chemistry
Page 10 of 31
203
adding 0.5 mL NaCl (0.9%), centrifuged twice at 700 g for 10 min, and dried under a stream
204
of nitrogen 12.
205
The different types of GPL (PC, PE, PS, and PI) were isolated by the second TLC procedure
206
using chloroform:methanol:acetic acid:water (100:75:6:2, v/v/v/v) 12,34. The GPL classes were
207
revealed by spraying with 2% 2,7-dichlorofluorecin in 96% ethanol and detected under UV
208
light at 366 nm by comparing with an external standard (Nu-chek Prep, Elysian, MN, USA).
209
The GPL bands were then separately scraped off the TLC plates and trans-methylated to FA
210
methyl esters (FAME) with benzene, methanolic HCl, and 2,2-dimethoxypropane (10:10:1,
211
v/v/v) overnight at room temperature
212
methylation. Tricosylic acid (C23:0, Nu-chek Prep, Elysian, MN, USA) was used as an
213
internal standard.
214
Gas chromatography-flame ionization analysis
215
The FAME were analyzed using a gas chromatograph (Hewlett Packard 6890, Palo Alto, CA,
216
USA) equipped with an auto-injector in split mode (HP 7683, Agilent, Avondale, PA, USA),
217
a BPX70 capillary column (SGE Victoria, Australia, 60 m length, 0.25 mm internal diameter,
218
0.25 µm thickness film), and a flame ionization detector (Hewlett Packard 6890) 13. Helium
219
was the carrier gas, with a constant flow of 20 mL/min. Both the injector and the detector
220
temperature were set at 270 °C. The oven temperature was initially held at 50 °C for 1.2 min,
221
then ramped at 4 °C/min to 170 °C, at 0.5 °C/min to 200 °C, and to the final temperature of
222
240 °C at a rate of 10 °C/min. The individual FA were identified by comparing the retention
223
time with external standards (Nu-chek Prep, Elysian, MN, USA). Peak areas were integrated
224
using HP ChemStation to calculate the relative FA content.
225
2.5 Skin histological analysis
226
Histopathological evaluation was performed on the skin of fish from the eight treatments
227
(CC, 0% EPA+DHA, 1.0% EPA, 1.0% DHA, 1.0% EPA+DHA, 2.0% EPA, 2.0% DHA, and
35
. Samples were neutralized with 6% NaHCO3 after
ACS Paragon Plus Environment
10
Page 11 of 31
Journal of Agricultural and Food Chemistry
228
2.0% EPA+DHA; n=10 per dietary group). Paraplast-embedded skin samples were
229
microtome-cut (5 µm) and stained with standard hematoxylin and eosin (Merck KGaA,
230
Darmstadt, Germany). Stained slides were examined using a standard light microscope
231
(Nikon Optiphot, Japan). Images were captured by means of a Micropublisher camera and
232
QCapture software using a 40X objective. Samples were firstly subjected to a blinded
233
evaluation on histopathology which means that the identity of samples were unrevealed,
234
followed by a second evaluation after decoding the samples which provide a description per
235
dietary group, to ensure the observation are unbiased. Epidermal thickness and goblet cell
236
numbers/100 µm were evaluated using Image J (NIH, Bethesda, MD, USA).
237
2.6 Data analysis
238
Statistical Analysis System (SAS 9.3, SAS Institute, Cary, NC, USA) was used for univariate
239
data analysis within the experimental groups. The FA data in percentages were square root-
240
arcsine transformed before test. Data normality distribution (Anderson-Darling test) and
241
homoscedasticity (Bartlett’s test or Levene’s test) were checked. If the tests were failed, the
242
initial data were log-transformed and re-tested. General Linear Model was used for statistical
243
comparisons. For comparison of sphingolipid concentrations, two-way ANOVA was used
244
with diet and sampling time as fixed factors. For comparison of FA composition in GPL
245
fractions, data from different sampling times (at 200 g and 400 g, after feeding for 19 and 26
246
weeks, respectively) were analyzed separately using one-way ANOVA. For evaluation of the
247
histological parameters, one-way ANOVA was conducted. If the data did not satisfy the test
248
of distribution or homoscedasticity, Mann-Whitney test was applied as a non-parametric test.
249
Furthermore, Tukey’s test was employed as a post hoc test against a predefined significance
250
level (P