Aquaculture 513 (2019) 734396

Contents lists available at ScienceDirect

Aquaculture
journal homepage: www.elsevier.com/locate/aquaculture

Yeast as a protein source during smoltification of Atlantic salmon (Salmo T

salar L.), enhances performance and modulates health
Christian Sahlmanna,1, Brankica Djordjevica,1, Leidy Lagosa, Liv Torunn Mydlanda,
Byron Morales-Langeb, Jon Øvrum Hansena, Ragnhild Ånestada, Luis Mercadob,
⁎
Milena Bjelanovica, Charles McLean Pressc, Margareth Øverlanda,
a
Department of Animal and Aquaculture Sciences, Faculty of Biosciences, Norwegian University of Life Sciences, Aas, Norway
b
Grupo de Marcadores Inmunológicos, Instituto de Biología, Facultad de Ciencias, Pontificia Universidad Católica de Valparaíso, Valparaíso, Chile
c
Department of Basic Sciences and Aquatic Medicine, Faculty of Veterinary Medicine, Norwegian University of Life Sciences, Oslo, Norway

A R T I C LE I N FO A B S T R A C T

Keywords: Yeast produced from lignocellulosic biomass has the potential to serve as a high-quality protein source with
Gene expression health benefits, especially during critical stages of the Atlantic salmon life cycle, such as during seawater transfer
Nutrition (SWT). In this study, we evaluated the effect of adding 25% Candida utilis yeast to salmon feed on growth
Growth performance and overall health by using morphometry, immunohistochemistry, cytokine enzyme-linked im-
Candida utilis
munosorbent assays (ELISA) and gene expression analysis during and after SWT. There were four dietary
Immune response
Histology
treatments: 1) control diet in freshwater (FW) and seawater (SW) (Control); 2) control diet in FW and a yeast-
based diet in SW (Control/Yeast); 3) yeast-based diet in FW and SW (Yeast); 4) yeast-based diet in FW and a
control diet in SW (Yeast/Control). Our results showed that fish fed the yeast diet throughout the FW and SW
period achieved higher feed intake and higher growth rate than fish fed the control diet. Morphometric and
immunochemical analyses of the distal intestine (DI) revealed decreased length and number of CD3 labeled cells
in the simple folds of fish fed control diet, while no changes were observed in fish fed the yeast diet.
Furthermore, yeast significantly decreased the secretion of IFNγ, TNFα, IL-1β, IL-8, and modulated the gene
expression of aquaporin 8 (aqp8ab), superoxide dismutase (sod1) and major histocompatibility complex 1 (mhc1)
in DI, suggesting reduced inflammatory processes in yeast fed fish. These findings indicate that Candida utilis
yeast is a promising alternative protein source with functional properties in diets for smolting Atlantic salmon
before and after SWT.

1. Introduction functional ingredients in fish feeds and adapted feeding protocols are a
possible solution to these challenges.
Atlantic salmon (Salmo salar L.) is the main species in Norwegian The rapid growth in the aquaculture industry, but stable production
aquaculture, and as an anadromous fish undergoes a series of physio- of the main protein source, fishmeal (FM), has led to the increasing use
logical, structural and functional changes in the transition from fresh- of alternatives, resulting in significant changes in the composition of
water (FW) to seawater (SW). During this period, fish are more sus- aquafeeds. Microbial ingredients such as bacterial meal (Aas et al.,
ceptible to stress, physical damage and infectious diseases (Stefansson 2006; Romarheim et al., 2013; Romarheim et al., 2010; Vasanth et al.,
et al., 2008). High mortality during seawater transfer (SWT) and the 2015) and yeast (Couture et al., 2019; Grammes et al., 2013; Meena
first months at sea represents a significant economic loss and a major et al., 2013; Micallef et al., 2017; Øverland et al., 2013), represent
challenge for the aquaculture industry (Hjeltnes et al., 2018). Current potential sustainable alternative protein sources in aquafeeds due to
disease control strategies in aquaculture include the use of che- their high nutritional value, in addition to their content of a wide range
motherapeutics, vaccination and selective breeding programs. These of bioactive components with potential as functional ingredients. Feed
measures may decrease smoltification stress, but additional action is ingredients are considered functional when they have health beneficial
required to increase salmon smolt robustness during SWT. The use of properties beyond their nutritional value (Montalban-Arques et al.,

⁎
Corresponding author.
E-mail address: margareth.overland@nmbu.no (M. Øverland).
1
These authors contributed equally to this work.

https://doi.org/10.1016/j.aquaculture.2019.734396
Received 29 January 2019; Received in revised form 21 May 2019; Accepted 10 August 2019
Available online 12 August 2019
0044-8486/ © 2019 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
C. Sahlmann, et al. Aquaculture 513 (2019) 734396

2015). The use of functional feeds to modulate the immune system has Table 1
received increasing attention in fish farming (Herman and Schmidt, Diet composition.
2016; Kiron, 2012). The positive effects of yeast or its components have Ingredient Control Yeast
been documented in several studies for salmonids (Øverland and
Skrede, 2017). However, yeast strain, processing technology, and in- Formulation %
Fish meala 15 10.65
clusion level can result in varying effects on the immune system
Wheat glutenb 20 14.24
(Hauptman et al., 2014; Øverland and Skrede, 2017). Yeast produced Soy protein concentratec 20 14.24
from lignocellulosic biomass is a high-quality protein source with Pre-gel potato starchd 8.9 5.3
beneficial effects on fish health (Grammes et al., 2013). Yeast contains a Candida utilise 0 25
wide range of components such as α-glucan, β-glucan, α-mannan, nu- Field beansf 3 2.14
Gelating 8 8
cleic acids (known as microbe-associated molecular patterns; MAMPs)
Celluloseh 4 0
and antioxidants (damage-associated molecular patterns; DAMPs) Corn gluteni 3 2.13
(Navarrete and Tovar-Ramírez, 2014). The immune modulation func- Rapeseed oilj 8 8.25
tions of glucans have been demonstrated at transcriptional and protein Fish oilk 8 8.25
Threoninel 0.2 0
level in salmonids, mainly the expression of pro-inflammatory cyto-
Lysinem 0.2 0
kines such as TNFα, IL-8 or IL-1β (Morales-Lange et al., 2015; Salah Rhodimetn 0.13 0.2
et al., 2017). The latter compounds can strengthen the gut barrier of MCPo 0.8 0.8
fish, improve immune status and increase resistance to infectious pa- Choline chloridep 0.15 0.15
thogens (Meena et al., 2013). When fish is fed a potential challenging Yttrium oxideq 0.01 0.01
Vitamins/mineral premixr 0.6 0.65
plant-based diets combined with exposure to other stressors such as
-1
sub-optimal environmental condition (Mosberian-Tanha et al., 2018) or Chemical composition g kg
Crude protein 491 495
SWT, this could lead to adverse effects on intestinal function and an
Crude lipids 161 176
intensified immune response. Starch 80 62
In the present study, we evaluated immunomodulatory effect of Ash 47 55
yeast under two challenging conditions (SWT and plant-based diet). We Gross energy, MJ 22 22
propose that the inclusion of Candida utilis yeast in fish diets has the a
LT fishmeal, Norsildmel, Egersund, Norway.
potential to enhance immunity and alleviate stress related to smoltifi- b
Wheat gluten, Amilina AB, Panevezys, Lithuania.
cation and SWT. The objective of the present study was to investigate c
SPC, Lyckeby Culinar, Fjälkinge, Sweden.
potential functional feed effects of moderate dietary levels of Candida d
Lygel F 60, Lyckeby Culinar, Fjälkinge, Sweden.
utilis on Atlantic salmon growth rate, intestinal morphology, cytokine e
Candida utilis, LYCC 7549, Lallemand Yeast Culture Collection.
secretion, and transcriptional levels of genes involved in intestinal f
Dehulled field bean, Copenhagen Merchants, Denmark.
homeostasis and immune responses during the stressful SWT period. g
Rousselot® 250 PS, Rousselot SAS, Courbevoie, France.
h
Alpha-Cel™ C100, International Fibre Europe NV, Belgium.
i
2. Materials and methods Heinz & Co. AG, Zurich, Switzerland.
j
AAK, Karlshamn, Sweeden.
k
2.1. Experimental design, diets and feeding NorSalmOil, Norsildmel, Egersund, Norway.
l
L-Threonine, CJ Biotech CO., Shenyang, China.
m
L-Lysine CJ Biotech CO., Shenyang, China.
Atlantic salmon (Salmo salar L.) pre-smolts were fed a commercial n
Rhodimet NP99, Adisseo ASA, Antony, France.
salmon feed (Skretting, Stavanger, Norway) prior to the start of the o
Monocalsium phosphate, Bolifor® MCP-F, Oslo, Norway.
experiment, at the fish facility at the Norwegian University of Life p
Choline chloride, 70% Vegetable, Indukern s.a., Spain.
Sciences, Aas, Norway. There were two diets, a plant-based control diet q
Yttrium, Metal Rare Earth Limited, Shenzhen, China.
and an experimental diet with 25% Candida utilis (Table 1). The diets r
Premix fish, Norsk Mineralnæring AS, Hønefoss, Norway. Per kg feed;
were produced by cold pelleting with a pasta machine (P35A Carasco, Retinol 3150.0 IU, Cholecalciferol 1890.0 IU, α-tocopherol SD 250 mg,
Genova, Italy) at the Norwegian University of Life Sciences, Aas, Menadione 12.6 mg, Thiamin 18.9 mg, Riboflavin 31.5 mg, d-Ca-Pantothenate
Norway. All dry ingredients (except gelatine), were mixed by a Morette 37.8 mg, Niacin 94.5 mg, Biotin 0.315 mg, Cyanocobalamin 0.025 mg, Folic
Foreni kneading machine (Spiry 25, Mondofolo, Italy). The gelatine was acid 6.3 mg, Pyridoxine 37.8 mg, Ascorbate monophosphate 157.5 g, Cu:
dissolved in cold water and subsequently heated to 60 °C in a micro- CuSulfate 5H2O 6.3 mg, Zn: ZnSulfate 151.2 mg, Mn: Mn(II)Sulfate 18.9 mg, I:
K-Iodide 3.78 mg, Ca 1.4 g.
wave oven and thereafter mixed with fish oil. Liquid and dry in-
gredients were mixed into a dough, which was processed in the pasta
machine to form 3 mm pellets. These pellets were dried to a moisture FW were either continued on the yeast-based diet or were switched to
content of 7–9%. the control diet. This resulted in four dietary treatments, where fish
The experiment lasted for 56 days and consisted of two periods: FW were either fed: 1) control diet in FW and SW (Control), 2) control diet
(0–28 days) and SW (28–56 days). During the FW period, fish (n = 600) in FW and yeast-based diet in SW (Control/Yeast), 3) yeast-based diet in
were placed in two 2800 l tanks with 300 fish per tank. One tank re- FW and SW (Yeast) and 4) yeast-based diet in FW and control diet in SW
ceived the control diet and the other, the yeast-based diet. Continuous (Yeast/Control) (Fig. 1). Seawater was supplied directly from a depth of
24 h light was provided during FW and SW periods. Average water 60 m (salinity: 35 ppt; average temperature 7.7 °C). Intake water was
temperature during the FW period was 12 °C, water flow approximately passed through an UV filter and a drum filter to remove larger particles.
40 l min−1 and the oxygen saturation between 85 and 90%. Automatic Average water flow in each tank was adjusted to 7 l min−1 and oxygen
belt feeders distributed feed continuously with a feeding level of 2% of saturation between 85 and 90%. In the SW period, fish were fed with
the body weight plus 20% in excess per day. During SWT on day 28, electric belt feeders, once per day for 120 min in the morning. The
480 fish were transferred to the SW facility at the Norwegian Institute amount of feed was based on 0.5–1% of the body weight, adjusted by
for Water Research (NIVA). Fish were distributed into 16 tanks of 300 l the average feed consumption in each tank over the last seven days with
water capacity with 30 fish per tank. Fish fed the control diet during the 10% excess. Uneaten feed was collected for two weeks during the SW
FW period were randomly assigned to one of eight tanks. Four tanks period. The recovery values for each feed (pellets) were measured per
were continued on the control diet, and the other four were switched to tank to ensure correct calculations of uneaten/eaten feed (dry matter
the yeast-based diet. Accordingly, fish fed the yeast-based diet during (DM), g) according to Helland et al. (1996). The experiment was

2
C. Sahlmann, et al. Aquaculture 513 (2019) 734396

measurements, 12 fish per dietary treatment, sampled in FW on d0, d14

and d28, and in SW on d42 and d56 were analyzed. Digital images were
captured with a Zeiss AxioCam ERc5s camera connected to a light
microscope (Zeiss Axio Lab.A1, Carl Zeiss, Germany). For the mea-
surement of simple fold length, only fish in the Control or the Yeast diet
treatment were examined. Morphometric measurements of the length
of simple folds in the DI were performed by one individual (CP) on
digital images of longitudinally orientated HE sections at 2.5× mag-
nification using ImageJ software, version v1.51r (Schneider et al.,
2012). The fold height was measured from stratum compactum to the
tip of the epithelium lining the fold. Three simple folds were measured
from each fish using the segmented line tool. A selected fold was the
first full-length simple fold with an intact epithelium adjacent to a
complex fold. Only one simple fold was selected for measurement be-
tween a pair of complex folds. Means were calculated based on three
measurements.

Fig. 1. Experimental design. The experiment consisted of two periods: the ex-
2.5. Immunohistochemistry (IHC)
perimental freshwater period (FW) lasted for 28 days and included two diets:
plant-based control diet (Control) and a plant-based diet with 25% yeast
(Candida utilis) inclusion (Yeast). At seawater transfer (SWT) on day 28 (end of CD3 T lymphocytes were identified in DI tissue sections by IHC
FW period), the Control group was divided into two groups, each in four tanks, using a monoclonal anti-CD3ɛ antibody (dilution 1:600) (Boardman
receiving either control diet (Control) or yeast diet (Control/Yeast). Similarly, et al., 2012). Twelve fish from all dietary treatment collected at FW
the Yeast group was divided into two groups, receiving either control diet (d28) and SW (d42) were subjected to IHC analysis. Briefly, formalin-
(Yeast/Control) or yeast diet (Yeast). The SW period lasted for 28 days. fixed, paraffin-embedded sections (4 μm) were mounted on poly-L-ly-
sine-coated glass slides (Superfrost Plus, Thermo Scientific, Braunsch-
performed according to the guidelines established by the Norwegian weig, Germany) and left to dry at 37 °C. The slides were incubated at
Animal Research Authority. All animals were cared for, according to 58 °C for 30 min, deparaffinized in xylene, and rehydrated in graded
laws and regulations for experiments on live animals in EU (Directive alcohols to distilled water. Antigen retrieval was performed by using
2010/637EU) and Norway (FOR-2015-06-18-761). hydrated autoclaving at 121 °C for 15 min in 0.01 M citrate buffer, pH 6.
Unless otherwise stated, sections were washed between steps in phos-
phate-buffered saline (PBS). The sections were incubated in phenyl
2.2. Yeast production
hydrazine at 37 °C for 40 min to quench endogenous peroxidase ac-
tivity. An automated immunostainer (Lab Vision™ Autostainer 360) was
Carbohydrate substrate, containing monosaccharides (C5 and C6
used for the IHC procedure. Non-specific antibody binding was blocked
sugars), was produced at Borregaard pilot plant, Sarpsborg, Norway),
by incubating in normal horse serum. The sections were incubated at
using wood chips from Norwegian spruce trees in a biorefinery process
room temperature with the primary antibody for one hour. The tissue
(BALI process; Sjöde et al., 2011; Sjöde et al., 2013; Sjöde et al., 2015).
sections were incubated with the biotinylated secondary antibody and
The BALI-sugars were mixed with beet molasses sugars in the ratio 1:1
then with ABC/PO reagent (ABC Vectastain Elite kit). Immunolabelling
in a 42,000-l fermenter (Lallemand plant, Salutaguse, Estonia), and
was revealed using the 3-amino-9-ethylcarbazole solutions (ImmPACT®
were used as the main carbon source to grow yeast (Candida utilis; LYCC
AEC, Vector Laboratories). Sections were counterstained in Alcian blue
7549; Lallemand Yeast Culture Collection). After fermentation, the
for 1 min and coverslips were mounted using a mounting medium
yeast cells were washed, centrifuged and heat-inactivated before drum
(Aquatex). For negative control, the primary antibody was replaced
drying.
with an irrelevant antibody.
For the calculation of the percentage of simple fold area occupied by
2.3. Sampling CD3 T cells and goblet cells, images of immunohistochemically labeled
sections at 10× magnification were analyzed by one individual (CP)
Fish were sampled in FW on day 0, 14 and 28 (d0, d14 and d28) and using ImageJ. Three simple folds were selected as described for HE
in SW on day 42 and 56 (d42 and d56). Total weight of fish biomass sections above. The freehand selection tool was used to trace the simple
was recorded per tank on d0, d28 and d56. During the FW period, 12 fold area from stratum compactum to the tip of the epithelium lining
fish per tank were randomly sampled, anesthetized (80 mg l−1 MS222) the fold, and a region of interest (ROI) defined. The fold area was
and weighed individually. Then fish were dissected for tissue samples. measured from stratum compactum, including the middle of the fold
For histology and immunohistochemistry, samples of distal intestine base on each side, and the whole simple fold. The colour deconvolution
(DI) were placed in 10% neutral buffered formalin for 48 h at room plugin was used to extract the red (Colour 2) and blue (Colour 1) grey
temperature and further processed according to routine histological scale images. For the measurement of the area labeled for CD3, the red
procedures. For gene and protein expression DI, spleen and head kidney image was segmented, and threshold settings and particle size settings
of approximately 5x5x5 mm size were placed in cryotubes containing for detection of immunolabelling were then determined, and area
1.5 ml RNAlater, placed at 4 °C for 24 h and stored at −80 °C until measurements within the ROIs of each image were performed. The
further analysis. During the SW period, three fish per tank (12 fish per measurement of threshold particles was confined to the ROI (simple
treatment) were sampled using the procedure described above. fold) and the percentage of area calculated. A mean for each individual
was calculated based on the three measurements. The same procedure
2.4. Histology and Morphometry was performed on the blue image to determine the mean percentage of
simple fold area occupied by Alcian blue stained goblet cells.
Tissues were embedded in paraffin with an orientation to provide
longitudinal sections. Sections (2 μm) were mounted on glass slides 2.6. RNA extraction, cDNA synthesis, quantitative real time PCR
(Menzel Gläser, Thermo Scientific, Braunschweig, Germany) and
stained with hematoxylin and eosin (HE). For morphometric Total RNA was extracted using RNeasy® Plus Universal protocol

3
C. Sahlmann, et al. Aquaculture 513 (2019) 734396

Table 2
Primer used in qPCR analyses.
Gene name Genes Fwd/Rwd Primer sequence GenBank accession.no Ref.

Hypoxanthine phospho-ribosyltransferase 1 HPRT 1 Fwd CCGCCTCAAGAGCTACTGTAAT XM_014212855.1 (Kortner et al., 2013; Kortner et al., 2012)
Rev GTCTGGAACCTCAAACCCTATG
Glyceraldehyde-3-phosphate dehydrogenase GAPDH Fwd AAGTGAAGCAGGAGGGTGGAA XM_014141819.1 (Kortner et al., 2013; Kortner et al., 2012)
Rev CAGCCTCACCCCATTTGATG
Superoxide dismutase SOD1 Fwd CCACGTCCATGCCTTTGG NM_001123587.1 This study
Rev TCAGCTGCTGCAGTCACGTT
Aquaporin-8ab AQP8ab Fwd GTTGGCATAGTTCTCCTTTGATG XM_014179685.1 (Kortner et al., 2013; Kortner et al., 2012)
Rev TTTCAACCCTCCCTTCACC
Mucin 2 MUC2 Fwd TCTGTCCTGATGGGATGAAAC XM_014183074.1 (Sahlmann et al., 2013)
Rev GGACTCCAAACAAACGCAAT
Interleukin 1 beta IL1b Fwd TCAGGGTCTGGATCTGGAGG XM_014170479.1 This study
Rev CACAGCACTCTCCAGCAAGA
Heat shock protein 70 HSP70 Fwd CCCCTGTCCCTGGGTATTG XM_014137172.1 This study
Rev CACCAGGCTGGTTGTCTGAGT
Glutathione S-transferase 3 GSTA3 Fwd AACGCCCAGAAATAGCCTCT NM_001140755.1 This study
Rev GACACGATTCATCCTCAGCA
Matrix metallopeptidase 13 MMP13 Fwd CCTTCCAAGTCCGAGGCTTT NM_001140524.1 This study
Rev GGCTCATGAGGGTCGATGTT
Major histocompatibility complex 1 MHC1 Fwd TGCACTCTGTTCAGCAAACC XM_014177344.1 This study
Rev ATACGTCCAACCAGCCTCAC
Interleukin 8 IL8 Fwd GGAAAGCGGCTCTCTTCTCAT NM_001140710.2 This study
Rev AGTCTGTTGTTATCTCGCTGGT

(Qiagen). The concentration of total RNA was determined using a PBS for 2 h at 37 °C, the plates were incubated for 90 min at 37 °C with
NanoDrop TM 8000 spectrophotometer (Thermo Fisher Scientific, the anti-epitope primary antibodies (IFNγ, IL1β, TNFα and IL8) at
Wilmington, USA). The RNA quality was checked using an Agilent 2100 1:200 dilution. Then, the second antibody-HRP (ThermoFisher Scien-
Bioanalyzer (Agilent Technologies, Waldbronn, Germany). Only sam- tific) was incubated for 1 h at 37 °C at 1:7000 dilution. Finally, chro-
ples of high quality (RIN ≥ 7) were included in the analyses. Purified magen substrate 3,3′,5,5′-tetramethylbenzidine single solution (Thermo
total RNA was stored at −80 °C until further analysis. All samples were Fischer Scientific, USA) was added (100 μl) and incubated for 30 min in
normalized to the same concentration (364 ng μl−1 for DI and the dark, at room temperature. The reaction was stopped with 50 μl of
396 ng μl−1 for spleen), prior to cDNA synthesis. The cDNA synthesis 1 N sulphuric acid and read at 450 nm on a Spectramax microplate
was performed using an AffinityScript QPCR cDNA Synthesis kit reader (Spectra Max M2, Molecular Devices).
(Agilent Technologies). In a total volume of 20 μl, 6 μl of RNA was
added together with 3 μl random primers, 1 μl AffinityScript RT / RNase
2.8. Statistical analysis
Block Enzyme Mixture and 10 μl cDNA Synthesis Master Mix. The
conditions for the cDNA synthesis were: 25 °C for 5 min, 42 °C for
The effects of dietary treatment on weight gain, feed intake and
45 min, 95 °C for 5 min, 4 °C. A set of genes involved in intestinal
relative weight gain in SW period were determined using one-way
homeostasis and epithelial integrity, oxidative stress responses and
ANOVA. Tukey post hoc comparisons were used for pairwise compar-
immune responses were analyzed by quantitative real-time PCR
isons for differences of means between groups (dietary treatments). The
(qPCR), using LightCycler® 480 System (Roche). Two reference genes
described analyses were performed using Minitab 18. For group ana-
(hprti and gapdh) met the qualifications of transcriptional stability in DI,
lysis and graphical presentation of the results for qPCR, histology,
while in spleen, only hprti met the qualifications of transcriptional
morphometry, IHC and cytokine levels, the statistical software
stability. The sequences of all primers used in this study are provided in
GraphPad Prism® 7.08 was used for calculation of statistical measures,
Table 2. The qPCR reactions were conducted in a total volume of 20 μl
including mean values, standard deviation, D'Agnostino Pearson om-
using 10 μl of LightCycler 480 SYBR Green I Master, 2 μl of gene specific
nibus normality test, t-tests and ANOVA. In addition, two-way ANOVA
primers, 3 μl of Milli-Q® water and 5 μl of cDNA previously diluted 1:10.
was run for interaction by two main factors, factor 1 (diet composition
The qPCR condition were: 95 °C for 5 min, 95 °C for 10 s, 60–64 °C for
before SWT) and factor 2 (diet composition after SWT). Statistical sig-
10–15 s depending on the primers, 72 °C for 15 s, in total 40–45 cycles.
nificance was declared at P-value < .05.
At the end of the program, a melting curve analysis was performed to
confirm the absence of primers dimers or unspecific products.
3. Results
2.7. Cytokine assay by indirect ELISA
3.1. Growth rate and feed intake
For the detection of IFNγ, TNFα, IL-1β and IL-8 samples of head
kidney, spleen and DI were homogenized in an automatic homogenizer During the eight weeks experimental period, no mortalities were
using metal beads and lysis buffer (Tris 20 mM, NaCl 100 mM, Triton X- observed, and fish appeared to be in good health in both FW and SW
100 0.05%, EDTA 5 mM, and protease inhibitor cocktail 1×). Then, the period. Growth rate and feed intake estimates during FW and SW period
homogenate was centrifuged at 12000 xg for 25 min at 4 °C. The su- are presented in Tables 3 and 4, respectively. Generally, the feed intake
pernatant, containing soluble proteins, was stored at −20 °C until use. after SWT was low for both dietary groups. One way ANOVA (Table 4)
All protein samples were quantified by a Pierce BCA Protein Assay Kit revealed that fish fed yeast-based diet during the entire experiment had
(ThermoFisher Scientific) following the manufacturer's instructions. higher weight gain and feed intake (average eaten DM (g) per g fish)
Each sample was diluted in carbonate buffer (NaHCO3 60 mM, pH 9.6) per treatment, for a two weeks period compared with fish fed the
and seeded (in duplicate) in a 96-well plate (Maxisorp, Thermo) at control diet during the same period. Two-way ANOVA revealed no in-
45 ng μl−1 (100 μl) for overnight incubation at 4 °C (Boltana et al., teraction between diets in FW and SW on growth parameters; however,
2018). After blocking with 5% Blotting-Grade Block (BioRad) diluted in a significant interaction was found for feed intake. (Table 4).

4
C. Sahlmann, et al. Aquaculture 513 (2019) 734396

Table 3 A Control Yeast

Mean initial and final fish weight for fish fed control or yeast-based diets during
the FW phase. Calculated weight gain is presented as mean weight gain per fish
and relative weight gain (RWG) is presented in percentage terms.
Freshwater Control Yeast

Initial fish weight (g) 80 85

Final fish weight (g) 105 115
Weight gain (g) 25 30
RWG % 31 35 FW(d28) FW(d28)

3.2. Histology and Morphometry

DI of fish examined during the FW and SW period showed normal

morphology, as described previously by others (Løkka et al., 2013).
Simple folds were slender with a thin lamina propria. Intestinal epi-
thelial cells were tall with the nucleus located basally, large vacuoles
SW(d42) SW(d42)
located apically and many intraepithelial lymphocytes. Goblet cells and
rodlet cells were scattered among the epithelial cells. The lamina pro-
pria adjacent to the stratum compactum was thin and numerous eosi- B Control Yeast
nophilic granule cells were present (Fig. 2A). We did not observe dif-
1500
ferences between diets. The length of simple folds in the DI was *

Simple fold height (µm)

measured in fish that received either the control diet (Control) or the
yeast-based diet (Yeast) throughout the experiment, i.e., from day 0 to
day 56 in both FW and SW periods. There was no difference in the 1000
length of simple folds between dietary treatments at the start of the
experiment (Fig. 2B). Two weeks after transfer to SW (Fig. 2B; SW
(d42)) there was a significant decrease in length of simple folds in the
DI of fish on the control diet compared with the fish under the same diet 500
before SWT (Fig. 2B; FW (d28)) (P < .05). There was no change in the
length of the simple folds in fish fed the yeast-based diet in any of the
sampling points (Fig. 2B).
0
3.2.1. Immunohistochemistry
)

)
)

)
)
)
)

)
)
FW ( d 0

SW 42
56

FW ( d 0

SW 42
56
FW 1 4
SW 28

FW 1 4
SW 28
The change in simple fold morphology that was detected with
(d
(d

(d
(d
(d
(d

(d
(d
FW

FW
transfer to SW was investigated further by the morphometric ex-
amination of immunohistochemical sections labeled to identify CD3 T
Fig. 2. Histological evaluation of distal intestine of Atlantic salmon. (A)
cells and counterstained with Alcian blue to identify goblet cells. At FW
Representative histological images of distal intestine from plant-based control
(d28), CD3 positive cells in Control and Yeast fed fish, showed an
diet (Control; images to the left) and plant-based diet with 25% yeast inclusion
abundant presence at the base of the epithelium, and along the entire (Yeast; images to the right) at four weeks in freshwater FW (d28) and two
length of simple folds (Fig. 3A). Only a few CD3 labeled cells were weeks in seawater SW (d42). (B) Morphometric measurements of simple fold
observed in the lamina propria adjacent to the stratum compactum. height of the distal intestine in Control and Yeast group at FW and SW. Data are
CD3 labeled cells were rare in the lamina propria of the simple fold. expressed as mean and standard deviation, n = 12 for all groups. The black
Further some rodlet cells showed labeling for CD3. There was a diffuse point outside the whisker range of SW (d42) is displayed as an outlier.
labeling of smooth muscle, which was interpreted as background la- Significant differences are denoted with * (P < .05).
beling. The percentage of simple fold area occupied by CD3 labeled
cells and Alcian blue-stained cells were estimated for all dietary groups difference (P < .05) in CD3 labeled cells between the fish that changed
at FW (d28) and SW (d42) (Fig. 3B). There was no difference between diet from control to yeast (Control/Yeast) and the fish that changed
the percentage of area occupied by CD3 labeled cells in the simple folds from yeast to control (Yeast/Control) was observed (Fig. 3B). Two-way
of fish fed the control and yeast diets in FW. Two weeks after transfer to ANOVA revealed an interaction between diets given in FW and SW on
SW (d42), there was a significant decrease in the presence of CD3 la- CD3 labeled cells. There were no differences detected between the fish
beled cells in the simple folds of fish fed the control diet (P < .05) groups in the presence of Alcian blue labeled goblet cells (Fig. 3C).
(Fig. 3B). However, there was no change in CD3 labeled cells in fish fed
yeast diet. Two weeks after SWT coupled with a shift in diet, a

Table 4
Fish weight at the start and the end of the SW period (g/fish), feed intake (g/fish/2 weeks period), and relative weight gain percentage (RWG %) for the four dietary
treatments in the SW phase. Data are presented as means ± standard error per fish per dietary treatment (three fish per tank; four tanks per treatment). Means with
different superscript letters in a row are significantly different (P < .05).
Seawater Control Control/yeast Yeast/control Yeast P-value

a a b b
Initial weight (g) 117 ± 0.91 115 ± 0.8 128 ± 0.41 129 ± 1.2 < 0.0001
Final weight (g) 120 ± 2.7a 122 ± 3.5a 140 ± 0.51b 142 ± 1.5b < 0.0001
Weight gain (g) 3 ± 2.6a 6 ± 3.8ab 12 ± 0.42ab 14 ± 1.2b 0.017
Feed intake (DM, g) 1.54 ± 0.05a 1.96 ± 0.06b 1.85 ± 0.06b 2.22 ± 0.03c < 0.0001
RWG % 3 ± 2.2a 6 ± 2.7ab 9 ± 0.34ab 11 ± 0.95b 0.041

5
C. Sahlmann, et al. Aquaculture 513 (2019) 734396

Fig. 3. Immunohistochemical evaluation of distal

A Control Yeast intestine of Atlantic salmon. (A) Representative im-
munochemistry images of DI tissue from plant-based
control diet (Control; images to the left) and plant-
based diet with 25% yeast inclusion (Yeast; images
to the right) at four weeks in freshwater FW (d28)
and two weeks in seawater SW (d42). Red colour
indicates the area of simple folds stained for CD3-T
cells and blue colour indicates the area of simple
folds stained for Goblet cells. (B) CD3-T cell % area
within the simple fold. (C) Goblet cell percentage
area within the simple fold. Data are expressed as
mean and standard deviation, n = 12 for all groups.
FW(d28) FW(d28) Significant differences are denoted with *
(P < .05). (For interpretation of the references to
colour in this figure legend, the reader is referred to
the web version of this article.)

SW(d42) SW(d42)

B C

3.3. Gene expression in DI and spleen (d56) (Fig. 4D). On the other hand, the results revealed a significant
upregulation (P < .05) of major histocompatibility complex 1 (mhc1)
The expression of several genes representing various regulatory and gene in DI in fish fed the Yeast diet in FW (d28) compared to fish fed the
immune functions were analyzed. The results in DI revealed increases Control (Fig. 4E), whereas the expression level of mhc1 was lower in the
(P < .05) in the expression levels of aquaporin 8 (aqp8ab) in Yeast SW period as compared to the FW period for both Control and Yeast.
treatment compared with the Yeast/Control treatment at the final Mucin 2 (muc2) gene expression in DI did not reveal significant dif-
sampling (d56) (Fig. 4A). The comparison between FW and SW for ferences between the dietary treatments (Fig. 4F).
Yeast and Control, showed increases in the expression of aqp8ab in SW
(d56) compared to FW (d28) for both dietary treatments, with sig- 3.4. Protein levels of cytokines
nificant interaction between diets given before and after SWT. At the
same time point (d56), expression of superoxide dismutase (sod1) in the Indirect ELISA was performed on DI, spleen and head kidney from
DI was lower in the Yeast treatment compared with the Control/Yeast fish in SW (d56). There was a significantly decreased protein level of
and Yeast/Control treatments, and a trend for decreased expression IFNγ, TNFα, IL-1β, IL-8 and in DI of fish fed Yeast compared to Control
compared with Control with significant interaction between diets given (Fig. 5A). No significant changes were observed in the protein level of
before and after SWT (Fig. 4B). Expression levels of the pro-in- any of the cytokines evaluated in head kidney or spleen (Fig. 5B-C).
flammatory marker interleukin 1 beta (il1b) showed a trend for de-
creased expression in Yeast and Control/Yeast compared with Control 4. Discussion
treatment in SW (d56) (Fig. 4C). No differences were observed in the
expression levels of interleukin 8 (il8) at any time point or in any In the present study, we investigated the potential of using 25%
treatments for DI, although, there was a trend for decreased expression Candida utilis yeast as a functional feed ingredient to improve feed in-
of il8 in Yeast and Control/Yeast compared with Control in DI in SW take, growth rate and reduce stress during the critical period of SWT on

6
C. Sahlmann, et al. Aquaculture 513 (2019) 734396

Fig. 4. Quantitative real-time PCR data of selected

genes in distal intestine: (A) Aquaporin-8ab (aqp8)
(B) Superoxide dismutase (sod1) (C) Interleukin 1
beta (il1b) (D) Interleukin 8 (il18) (E) Major histo-
compatibility complex 1 (mhc1) (F) Mucin 2 (muc2).
Data are presented as mean expression ratio with
box and whisker plots. Boxes mark the interquartile
range, the line dividing the boxes denotes the
median, and whiskers extend to 1.5 x interquartile
range. A significant difference between dietary
treatments during the whole experiment is denoted
with * (P < .05). A significant difference between
FW and SW period for Control and Yeast treatment
are denoted with different letters (P < .05).

Atlantic salmon. The rationale of including 25% of yeast was to eval- FW and SW period, grew faster and were more robust during the first
uate yeast as a protein source with functional properties and not only as weeks in SW. Moreover, fish fed yeast through the entire period also
an additive. Lower levels of yeast inclusion have also demonstrated had a higher feed intake and the results demonstrated a significant
beneficial health effects on Atlantic salmon (Hansen et al., 2019). Many interaction between feeding yeast in FW on the feed intake in SW. This
yeast strains contain several bioactive components such as β-glucans, result could be due to increased palatability of the diet containing yeast.
mannan, chitin and nucleotides, which have been shown to induce a Yeast possesses feed enhancing properties, including nucleotides and
general immune response, protect against bacterial infections, improve glutamate, which has proven to be taste enhancers in fish (Kasumyan
gastrointestinal barrier function and disease resistance (Mohan et al., and Døving, 2003). These properties could be an advantage when used
2018). The transition from FW to SW exposes salmon to large en- in combination with plant-based ingredients which can lead to reduced
vironmental changes that cause stress and can have detrimental effects feed intake due to their content of a wide range of antinutritional fac-
on growth performance and survival. Some of the challenges of SWT tors (Krogdahl et al., 2010). Immediately after SWT, fish usually do not
includes osmotic changes and changes in naturally occurring bacteria in exhibit typical feeding behavior and it can take several weeks until fish
water. To the best of our knowledge, this is the first investigation of the start eating normally. In our study, the fish presented this characteristic
effect of yeast inclusion during SWT and its effect on smolting Atlantic behavior of low feed intake after SWT. Fish fed yeast-based diet through
salmon. the entire experimental period (Yeast) or yeast in the FW period and
In the present study, fish fed diets containing 25% yeast during the then the plant-based control diet in SW (Yeast/Control) had a higher

7
C. Sahlmann, et al. Aquaculture 513 (2019) 734396

Fig. 5. The protein level of IFNγ, TNFα, IL-1β and IL-8 measured by indirect ELISA in (A) distal intestine (B) spleen and (C) head kidney. Data are presented in pg/μl,
as mean and standard deviation. Black points outside whisker range are displayed as outliers. Significant differences between treatment are denoted with *
(P < .05).

weight gain than fish in Control or the Control/Yeast treatment. No Yeast treatment indicating beneficial yeast effect on the regulation of
significant interaction between feeding yeast in FW and fish growth in water and ion transport during smoltification. In some studies, the in-
SW was observed. Long term feeding trial might be an option to assess clusion of new microbial ingredients into fish diets, greatly altered
the effect of yeast on growth. Our histology results demonstrated sig- mucus production and epithelial morphology (Ouwehand et al., 2005;
nificantly reduced length of simple folds after SWT in Control but not in Torrecillas et al., 2014). Herein, however, we did not detect differences
Yeast treatment. This reduction has also been reported in fish fed diets in mucin 2 (muc2) expression at the selected sampling points. This
with high levels of soybean meal (Baeverfjord and Krogdahl, 1996; finding correlates with histological observations where no differences
Booman et al., 2018), pea protein concentrate (Kortner et al., 2012; between dietary treatments regarding the abundance of goblet cells
Penn et al., 2011) and in fish fed diets where fish oil was substituted were seen. Several metabolic processes produce reactive oxygen species
with plant oils (Moldal et al., 2014). In our study, we used a plant-based (ROS) that may cause oxidative stress when the balance between ROS
control diet, which could have induced mild inflammation when fish and antioxidant is disturbed. Therefore, the expression of superoxide
were transferred to SW. Starvation could also be a possible cause of dismutase (sod1) plays a key role in controlling oxidative stress. In the
reduced simple fold length in SW (d42), as reported in earlier studies present study, sod1 was significantly lower in the Yeast treatment
(Baeverfjord and Krogdahl, 1996). The decreased levels of CD3 positive compared with the Control/Yeast and Yeast/Control in SW. As dis-
T cells at the same time point in DI of the Control treatment could cussed earlier, immune-modulating abilities of feed are widely re-
indicate an immunosuppression after SWT as observed by Johansson cognized and dietary changes can trigger an immune response that can
et al. (2016). In contrast, fish in the Yeast group did not show a de- be either beneficial or detrimental. Cytokines are secreted by activated
crease in CD3 positive T cells. Furthermore, fish that were switched immune-related cells and play pivotal roles as immune modulators. In
from yeast to control diet in SW presented higher levels of CD3 positive our study, we show that the inclusion of yeast to diets triggered an
T cells in the DI epithelium (Yeast/Control), compared with the Control intestinal immune modulation decreasing the secretion of IFNγ, TNF-α,
in SW, suggesting that yeast can reduce changes in immune cells in- IL-1β, and IL-8 locally in DI, but apparently without inducing a systemic
duced by SWT or plant-based diets. Earlier studies showed that the DI of response (i.e., no difference was observed in spleen or head kidney).
salmon reacts uniquely when challenged with a new diet. Soybean Cytokine autocrine and paracrine modes of action have been previously
meal-induced enteritis is probably the best described condition for in- documented in fish (Wang et al., 2011). Similarly, gene expression of
testinal inflammation in salmon, causing inflammation only in the DI, the pro-inflammatory cytokines Il1b and Il8 in DI were decreased in
while other intestinal regions and organs are unresponsive (Baeverfjord Yeast compared with the Control group in SW (d56). Il1b facilitates the
and Krogdahl, 1996; Chikwati et al., 2013). Our findings also confirmed organism to react quickly to infection by inducing the cascade of re-
this, having the effects of a change in the diet most prominently ob- sponses leading to inflammation (Zou and Secombes, 2016). Besides its
served in the DI. The increased expression of aquaporin (aqp8ab) in the role in inflammatory regulations, it has other diverse physiological
SW period compared to FW period in DI was in accordance with pre- functions, as affecting muscle metabolism and growth (Pooley et al.,
viously reported findings and as the expected response to increased 2013; Valenzuela et al., 2017). Interestingly, a significant up-regulation
salinity. Engelund et al. (2013) indicated the physiological role of of mhc1 in the DI of yeast fed fish was observed right before SWT. In
aqp8ab in transcellular water uptake across intestinal enterocytes upon mammals, major histocompatibility complex class I (MHC-1) molecules
SWT. We observed the trend of higher aqp8ab expression in SW (d56) in are key players in discriminating self from non-self. If foreign elements,

8
C. Sahlmann, et al. Aquaculture 513 (2019) 734396

such as viruses, are present in the cytoplasm, they are prone to de- Camilla Skagen-Sandvik for their support during the seawater stage of
gradation and presentation by MHC-1 molecules. It is assumed that the experiment. Felipe Reveco is thanked for regular critical feedback
presenting a larger repertoire of pathogen epitopes activates a wider and discussions about the feeding experiment and study design. Thanks
range of T cell clones, a response needed for protection against pa- also go to Boregaard and Lallemand for providing the yeast.
thogens (Tregaskes et al., 2016). Atlantic salmon possesses unique This study was funded by Foods of Norway, a Centre for Research-
genome duplication, turning duplicated MHC-1 region into a remark- based Innovation, the Research Council of Norway (RCN) 237841/030;
able target to study functional advantages of different immune activa- BIOFEED – Novel salmon feed by integrated bioprocessing of non-food
tion in salmonids (Grimholt, 2016). The upregulation of mhc1 induced biomass, RCN 239003/O30; Nutrition and Mucosal Immunity: how to
by yeast in this study may be of advantage in boosting the immune feed the immune system on fish, RCN 281052, and partially by Wood
system before SWT. However, the upregulation of mhc1 was followed prebiotics RCN 244259.
by a slight downregulation on d56 in SW in the Yeast group, suggesting
an immune system depression post SWT that is consistent with the References
findings by Johansson et al. (2016) and Jarungsriapisit et al. (2018).
Nevertheless, the immune suppression responses at the transcriptome Aas, T.S., Grisdale-Helland, B., Terjesen, B.F., Helland, S.J., 2006. Improved growth and
level were not as pronounced as reported previously. It may be that the nutrient utilisation in Atlantic salmon (Salmo salar) fed diets containing a bacterial
protein meal. Aquaculture 259, 365–376. https://doi.org/10.1016/j.aquaculture.
peak of immune suppression is occurring immediately after SWT and 2006.05.032.
within the first week post-SWT and, therefore, the gene expression le- Baeverfjord, G., Krogdahl, Å., 1996. Development and regression of soybean meal in-
vels may already have normalized. duced enteritis in Atlantic salmon, Salmo salar L., distal intestine: a comparison with
the intestines of fasted fish. J. Fish Dis. 19, 375–387. https://doi.org/10.1046/j.
1365-2761.1996.d01-92.x.
5. Concluding remarks Boardman, T., Warner, C., Ramirez-Gomez, F., Matrisciano, J., Bromage, E., 2012.
Characterization of an anti-rainbow trout (Oncorhynchus mykiss) CD3ɛ monoclonal
antibody. Vet. Immunol. Immunopathol. 145, 511–515. https://doi.org/10.1016/j.
This study described the effects of 25% Candida utilis yeast inclusion vetimm.2011.11.017.
in diets for Atlantic salmon pre and post SWT. Candida utilis inclusion Boltana, S., Sanhueza, N., Donoso, A., Aguilar, A., Crespo, D., Vergara, D., Arriagada, G.,
improved weight gain and feed intake, triggered a lower secretion of Morales-Lange, B., Mercado, L., Rey, S., Tort, L., Mackenzie, S., 2018. The expression
of TRPV channels, prostaglandin E2 and pro-inflammatory cytokines during beha-
cytokines in DI (IFNγ, TNF-α, IL-1β and IL-8), as well as decreased the
vioural fever in fish. Brain Behav. Immun. 71, 169–181. https://doi.org/10.1016/j.
expression of its genes. Yeast successfully modified immunosuppressive bbi.2018.03.023.
responses related to SW acclimation. In contrast to the control fish, the Booman, M., Forster, I., Vederas, J.C., Groman, D.B., Jones, S.R.M., 2018. Soybean meal-
yeast fed fish did not get reduced length of simple folds or decreased induced enteritis in Atlantic salmon (Salmo salar) and Chinook salmon
(Oncorhynchus tshawytscha) but not in pink salmon (O. gorbuscha). Aquaculture
levels of CD3 cells after SWT. The potential beneficial effects of yeast 483, 238–243. https://doi.org/10.1016/j.aquaculture.2017.10.025.
merit further investigation towards specific nutritional programming Chikwati, E.M., Sahlmann, C., Holm, H., Penn, M.H., Krogdahl, Å., Bakke, A.M., 2013.
and lower inclusion levels of yeast to improve health and robustness Alterations in digestive enzyme activities during the development of diet-induced
enteritis in Atlantic salmon, Salmo salar L. Aquaculture 402, 28–37. https://doi.org/
during SWT. Overall, the study suggests that yeast is a high-quality 10.1016/j.aquaculture.2013.03.023.
protein source with beneficial health properties in diets for smoltifying Couture, J., Geyer, R., Øvrum Hansen, J., Kuczenski, B., Øverland, M., Palazzo, J.,
Atlantic salmon. Sahlmann, C., Lenihan, H.S., 2019. Environmental benefits of novel non-human food
inputs to salmon feeds. Environ. Sci. Technol. 53, 1967–1975.
Engelund, M.B., Chauvigné, F., Christensen, B.M., Finn, R.N., Cerdà, J., Madsen, S.S.,
Conflict of interest 2013. Differential expression and novel permeability properties of three aquaporin 8
paralogs from seawater-challenged Atlantic salmon smolts. J. Exp. Biol. 2016,
3873–3885. https://doi.org/10.1242/jeb.087890.
The authors declare that they have no conflict of interest. Grammes, F., Reveco, F.E., Romarheim, O.H., Landsverk, T., Mydland, L.T., Øverland, M.,
2013. Candida utilis and Chlorella vulgaris counteract intestinal inflammation in
Submission declaration Atlantic salmon (Salmo salar L.). PLoS One 8, e83213. https://doi.org/10.1371/
journal.pone.0083213.
Grimholt, U., 2016. MHC and evolution in Teleosts. Biology 5, 6. https://doi.org/10.
All authors read and approved the final manuscript for submission. 3390/biology5010006.
The content of the manuscript has not been published, or submitted for Hansen, J.Ø., Hofossæter, M., Sahlmann, C., Ånestad, R., Reveco-Urzua, F.E., Press, C.M.,
Mydland, L.T., Øverland, M., 2019. Effect of Candida utilis on growth and intestinal
publication elsewhere.
health of Atlantic salmon (Salmo salar) parr. Aquaculture 511, 734239. https://doi.
org/10.1016/j.aquaculture.2019.734239.
Authors' contributions Hauptman, B.S., Barrows, F.T., Block, S.S., Gaylord, T.G., Paterson, J.A., Rawles, S.D.,
Sealey, W.M., 2014. Evaluation of grain distillers dried yeast as a fish meal substitute
in practical-type diets of juvenile rainbow trout, Oncorhynchus mykiss. Aquaculture
CS, MØ, LTM and JØH designed the experiment. CS coordinated the 432, 7–14. https://doi.org/10.1016/j.aquaculture.2014.03.026.
execution of the experiment as well as samplings and coordinated qPCR Helland, S.J., Grisdale-Helland, B., Nerland, S., 1996. A simple method for the mea-
and did the primer design and quality check. CS also performed data surement of daily feed intake of groups of fish in tanks. Aquaculture 139, 157–163.
https://doi.org/10.1016/0044-8486(95)01145-5.
analysis and statistics of the qPCR. JH and CS designed feed and per- Herman, E.M., Schmidt, M.A., 2016. The potential for engineering enhanced functional-
formed feed production. MB was helping during the practical part of the feed soybeans for sustainable aquaculture feed. Front. Plant Sci. 7, 440. https://doi.
feeding experiment and laboratory analysis. LM designed antibodies org/10.3389/fpls.2016.00440.
Hjeltnes, B., Bang-Jensen, B., Bornø, G., Haukaas, A., Walde, C.S., 2018. The Health
and BML performed Elisa assay. RÅ performed the qPCR analysis. CMP
Situation in Norwegian Aquaculture 2017. Norwegian Veterinary Institute. https://
performed histology and immunohistochemistry. BD was involved in www.vetinst.no/rapporter-og-publikasjoner/rapporter/2018/fish-health-report-
manuscript writing and reviewing, data analysis, producing figures and 2017.
Jarungsriapisit, J., Moore, L.J., Fiksdal, I.U., Bjørgen, H., Tangerås, A., Köllner, B.,
tables for the manuscript as well as statistical analysis and quality
Koppang, E.O., Patel, S., 2018. Time after seawater transfer influences immune cell
checking. MØ was involved in writing and reviewing the manuscript. LL abundance and responses to SAV 3 infection in Atlantic salmon. J. Fish Dis. 41,
data analysis for ELISA, manuscript reviewing. LTM was involved in 1269–1282. https://doi.org/10.1111/jfd.12820.
manuscript reviewing. Johansson, L.-H., Timmerhaus, G., Afanasyev, S., Jørgensen, S.M., Krasnov, A., 2016.
Smoltification and seawater transfer of Atlantic salmon (Salmo salar L.) is associated
with systemic repression of the immune transcriptome. Fish Shellfish Immunol. 58,
Acknowledgments 33–41. https://doi.org/10.1016/j.fsi.2016.09.026.
Kasumyan, A.O., Døving, K.B., 2003. Taste preferences in fishes. Fish Fish. 4, 289–347.
https://doi.org/10.1046/j.1467-2979.2003.00121.x.
The authors would like to thank Ricardo Tavares Benicio for his Kiron, V., 2012. Fish immune system and its nutritional modulation for preventive health
skillful help during the feeding experiment, to the staff of the Fish la- care. Anim. Feed Sci. Technol. 173, 111–133. https://doi.org/10.1016/j.anifeedsci.
boratory at NMBU and Solbergstrand, and to Prof Henning Sørum and 2011.12.015.

9
C. Sahlmann, et al. Aquaculture 513 (2019) 734396

Kortner, T.M., Skugor, S., Penn, M.H., Mydland, L.T., Djordjevic, B., Hillestad, M., protein sources in diets for Atlantic salmon (Salmo salar). Aquaculture 402, 1–7.
Krasnov, A., Krogdahl, Å., 2012. Dietary soyasaponin supplementation to pea protein https://doi.org/10.1016/j.aquaculture.2013.03.016.
concentrate reveals nutrigenomic interactions underlying enteropathy in Atlantic Penn, M.H., Bendiksen, E.Å., Campbell, P., Krogdahl, Å., 2011. High level of dietary pea
salmon (Salmo salar). BMC Vet. Res. 8, 101. https://doi.org/10.1186/1746-6148-8- protein concentrate induces enteropathy in Atlantic salmon (Salmo salar L.).
101. Aquaculture 310, 267–273. https://doi.org/10.1016/j.aquaculture.2010.10.040.
Kortner, T.M., Gu, J., Krogdahl, Å., Bakke, A.M., 2013. Transcriptional regulation of Pooley, N.J., Tacchi, L., Secombes, C.J., Martin, S.A.M., 2013. Inflammatory responses in
cholesterol and bile acid metabolism after dietary soyabean meal treatment in primary muscle cell cultures in Atlantic salmon (Salmo salar). BMC Genomics 14,
Atlantic salmon (Salmo salar L.). Br. J. Nutr. 109, 593–604. https://doi.org/10.1017/ 747. https://doi.org/10.1186/1471-2164-14-747.
S0007114512002024. Romarheim, O.H., Øverland, M., Mydland, L.T., Skrede, A., Landsverk, T., 2010. Bacteria
Krogdahl, Å., Penn, M., Thorsen, J., Refstie, S., Bakke, A.M., 2010. Important anti- grown on natural gas prevent soybean meal-induced enteritis in Atlantic salmon. J.
nutrients in plant feedstuffs for aquaculture: an update on recent findings regarding Ntr. 141, 124–130. https://doi.org/10.3945/jn.110.128900.
responses in salmonids. Aquac. Res. 41, 333–344. https://doi.org/10.1111/j.1365- Romarheim, O.H., Landsverk, T., Mydland, L.T., Skrede, A., Øverland, M., 2013. Cell wall
2109.2009.02426.x. fractions from Methylococcus capsulatus prevent soybean meal-induced enteritis in
Løkka, G., Austbø, L., Falk, K., Bjerkås, I., Koppang, E.O., 2013. Intestinal morphology of Atlantic salmon (Salmo salar). Aquaculture 402-403, 13–18. https://doi.org/10.
the wild Atlantic salmon (Salmo salar). J. Morphol. 274, 859–876. https://doi.org/ 1016/j.aquaculture.2013.03.011.
10.1002/jmor.20142. Sahlmann, C., Sutherland, B.J.G., Kortner, T.M., Koop, B.F., Krogdahl, Å., Bakke, A.M.,
Meena, D.K., Das, P., Kumar, S., Mandal, S.C., Prusty, A.K., Singh, S.K., Akhtar, M.S., 2013. Early response of gene expression in the distal intestine of Atlantic salmon
Behera, B.K., Kumar, K., Pal, A.K., Mukherjee, S.C., 2013. Beta-glucan: an ideal im- (Salmo salar L.) during the development of soybean meal induced enteritis. Fish
munostimulant in aquaculture (a review). Fish Physiol. Biochem. 39, 431–457. Shellfish Immunol. 34, 599–609. https://doi.org/10.1016/j.fsi.2012.11.031.
Micallef, G., Cash, P., Fernandes, J.M.O., Rajan, B., Tinsley, J.W., Bickerdike, R., Martin, Salah, A.S., El Nahas, A.F., Mahmoud, S., 2017. Modulatory effect of different doses of β-
S.A.M., Bowman, A.S., 2017. Dietary yeast cell wall extract alters the proteome of the 1, 3/1, 6-glucan on the expression of antioxidant, inflammatory, stress and immune-
skin mucous barrier in Atlantic Salmon (Salmo salar): increased abundance and ex- related genes of Oreochromis niloticus challenged with Streptococcus iniae. Fish
pression of a calreticulin-like protein. PLoS One 12, e0169075. https://doi.org/10. Shellfish Immunol. 70, 204–213. https://doi.org/10.1016/j.fsi.2017.09.008.
1371/journal.pone.0169075. Schneider, C.A., Rasband, W.S., Eliceiri, K.W., 2012. NIH image to ImageJ: 25 years of
Mohan, K., Ravichandran, S., Muralisankar, T., Uthayakumar, V., Chandirasekar, R., image analysis. Nat. Methods 9, 671.
Seedevi, P., Rajan, D.K., 2018. Potential uses of fungal polysaccharides as im- Sjöde, A., Frölander, A., Lersch, M., Rødsrud, G., 2011. Lignocellulosic Biomass
munostimulants in fish and shrimp aquaculture: a review. Aquaculture 500, 250–263. Conversion, Patent US 20110250638.
https://doi.org/10.1016/j.aquaculture.2018.10.023. Sjöde, A., Frölander, A., Lersch, M., Rødsrud, G., 2013. Lignocellulosic Biomass
Moldal, T., Løkka, G., Wiik-Nielsen, J., Austbø, L., Torstensen, B.E., Rosenlund, G., Dale, Conversion by Sulfite Pretreatment. (Patent EP 2376642).
O.B., Kaldhusdal, M., Koppang, E.O., 2014. Substitution of dietary fish oil with plant Sjöde, A., Frölander, A., Lersch, M., Rødsrud, G., Hals, K., Kløften, A.M., Delin, L.,
oils is associated with shortened mid intestinal folds in Atlantic salmon (Salmo salar). Johansson, M.H., 2015. Enzymatic Hydrolysis of Cellulose. (Patent US 9193982).
BMC Vet. Res. 10, 60. https://doi.org/10.1186/1746-6148-10-60. Stefansson, S.O., Björnsson, B.T., Ebbesson, L.O.E., McCormick, S.D., 2008.
Montalban-Arques, A., De Schryver, P., Bossier, P., Gorkiewicz, G., Mulero, V., Gatlin, Smoltification. In: Finn, R.N., Kapoor, B.G. (Eds.), Fish Larval Physiology. Science
D.M., Galindo-Villegas, J., 2015. Selective manipulation of the gut microbiota im- Publishers, Enfield, pp. 639–681.
proves immune status in vertebrates. Front. Immunol. 6, 512. https://doi.org/10. Torrecillas, S., Montero, D., Izquierdo, M., 2014. Improved health and growth of fish fed
3389/fimmu.2015.00512. mannan oligosaccharides: potential mode of action. Fish Shellfish Immunol. 36,
Morales-Lange, B., Bethke, J., Schmitt, P., Mercado, L., 2015. Phenotypical parameters as 525–544. https://doi.org/10.1016/j.fsi.2013.12.029.
a tool to evaluate the immunostimulatory effects of laminarin in Oncorhynchus Tregaskes, C.A., Harrison, M., Sowa, A.K., van Hateren, A., Hunt, L.G., Vainio, O.,
mykiss. Aquac. Res. 46, 2707–2715. https://doi.org/10.1111/are.12426. Kaufman, J., 2016. Surface expression, peptide repertoire, and thermostability of
Mosberian-Tanha, P., Schrama, J.W., Landsverk, T., Mydland, L.T., Øverland, M., 2018. chicken class I molecules correlate with peptide transporter specificity. Proc. Natl.
The effect of plant-based diet and suboptimal environmental conditions on digestive Acad. Sci. U. S. A. 113, 692–697. https://doi.org/10.1073/pnas.1511859113.
function and diet-induced enteropathy in rainbow trout (Oncorhynchus mykiss). Valenzuela, C.A., Zuloaga, R., Mercado, L., Einarsdottir, I.E., Björnsson, B.T., Valdés, J.A.,
Aquac. Nutr. 24, 112–122. https://doi.org/10.1111/anu.12539. Molina, A., 2017. Chronic stress inhibits growth and induces proteolytic mechanisms
Navarrete, P., Tovar-Ramírez, D., 2014. Use of yeasts as probiotics in fish aquaculture. In: through two different nonoverlapping pathways in the skeletal muscle of a teleost
Hernandez-Vergara, M., Perez-Rostro, C. (Eds.), Sustainable Aquaculture Techniques. fish. Am. J. Physiol. Regul. Integr. Comp. Physiol. 314, R102–R113. https://doi.org/
InTechOpen, pp. 135–172. https://doi.org/10.5772/57196. 10.1152/ajpregu.00009.2017.
Ouwehand, A.C., Derrien, M., de Vos, W., Tiihonen, K., Rautonen, N., 2005. Prebiotics Vasanth, G., Kiron, V., Kulkarni, A., Dahle, D., Lokesh, J., Kitani, Y., 2015. A microbial
and other microbial substrates for gut functionality. Curr. Opin. Biotechnol. 16, feed additive abates intestinal inflammation in Atlantic Salmon. Front. Immunol. 6,
212–217. https://doi.org/10.1016/j.copbio.2005.01.007. 409. https://doi.org/10.3389/fimmu.2015.00409.
Øverland, M., Skrede, A., 2017. Yeast derived from lignocellulosic biomass as a sustain- Wang, T., Huang, W., Costa, M.M., Secombes, C.J., 2011. The gamma-chain cytokine/
able feed resource for use in aquaculture. J. Sci. Food Agric. 97, 733–742. https:// receptor system in fish: more ligands and receptors. Fish Shellfish Immunol. 31,
doi.org/10.1002/jsfa.8007. 673–687. https://doi.org/10.1016/j.fsi.2011.05.016.
Øverland, M., Karlsson, A., Mydland, L.T., Romarheim, O.H., Skrede, A., 2013. Evaluation Zou, J., Secombes, C.J., 2016. The function of fish cytokines. Biology 5, 23. https://doi.
of Candida utilis, Kluyveromyces marxianus and Saccharomyces cerevisiae yeasts as org/10.3390/biology5020023.

10