Salinity stress, enhancing basal and induced immune responses in striped catfish Pangasianodon hypophthalmus (Sauvage)

Journal of Proteomics 167 (2017) 12–24

Contents lists available at ScienceDirect

Journal of Proteomics journal homepage: www.elsevier.com/locate/jprot

Salinity stress, enhancing basal and induced immune responses in striped catfish Pangasianodon hypophthalmus (Sauvage) Mélodie Schmitz a,⁎, Tamar Ziv b, Arie Admon b, Sébastien Baekelandt a, Syaghalirwa N.M. Mandiki a, Maëlenn L'Hoir a, Patrick Kestemont a a b

Research Unit in Environmental and Evolutionary Biology, University of Namur, rue de Bruxelles 61, B-5000 Namur, Belgium The Smoler Proteomics Center, Emerson Life Science Building, Technion – Israel Institute of Technology, 32000 Haifa, Israel

a r t i c l e

i n f o

Article history: Received 23 February 2017 Received in revised form 2 August 2017 Accepted 3 August 2017 Available online 7 August 2017 Keywords: Catfish Infectious disease Immunity Osmoregulation Label-free proteomics

a b s t r a c t In the Mekong Delta, striped catfish are faced with chronic salinity stress related to saltwater intrusion induced by global climatic changes. In this study, striped catfish juveniles were submitted to a prolonged salinity stress (up to 10 ppt) over three weeks followed by infection with a virulent bacterial strain, Edwardsiella ictaluri. Osmoregulatory parameters were investigated. In addition, a label free quantitative proteomics workflow was performed on kidneys. The workflow consisted of an initial global profiling of relative peptide abundances (by LC/MS, peak area quantification based on extracted ion currents), followed by identification (by MS/MS). The aim of the study was to highlight specific functional pathways modified during realistic salinity stress, particularly those involved in immunity. In kidney proteome, 2483 proteins were identified, of which 400 proteins were differentially expressed between the freshwater and the saline water conditions. Several pathways and functional categories were highlighted, mostly related to energy metabolism, protein metabolism, actin cytoskeleton, signaling, immunity, and detoxification. In particular, the responsiveness of proteins involved in small GTPases and Mitogen Activated Protein Kinase p38 signaling, phagolysosome maturation, and T-cells regulation is discussed. Statement of significance: In the Mekong River Delta (Vietnam), striped catfish production is threatened by extensive sea water intrusion exacerbated by sea level rise. In fish, the effect of chronic exposure to salinity stress on immune capacities and response to disease has been poorly investigated. This study aims to highlight the main molecular changes occurring in the kidney during acclimation to salinity stress, particularly those involved in the immune defences of fish. © 2017 Elsevier B.V. All rights reserved.

1. Introduction The striped catfish (Pangasianodon hypophthalmus, Sauvage) industry has become by far the biggest inland aquaculture production industry in South East Asia, particularly in the Mekong Basin in Vietnam. In 2014, striped catfish production reached 1.2 million tons and its products were exported to over 151 countries, principally the United States and in the European Union [1]. Since 2004, several studies have raised great concerns regarding the effects of sea level rise in the Mekong Basin on agricultural development in the region [2,3]. In South East Asia, long-term projections predict an annual rise in sea level of 2–4 mm per year [4]. In 2016, salinity intrusion inland reached 90 km and peaked at 9 ppt in many aquaculture provinces [3]. In catfish farms located in the lower Mekong River Delta, plasma osmolality has been positively correlated with

Abbreviations: BHI, brain heart infusion; MAPK, mitogen-activated protein kinase. ⁎ Corresponding author. E-mail address: [email protected] (M. Schmitz).

http://dx.doi.org/10.1016/j.jprot.2017.08.005 1874-3919/© 2017 Elsevier B.V. All rights reserved.

salinity [3]. In addition, key innate immune parameters have tended to increase in farms located in tidal areas [3]. In aquatic organisms, activation of several major evolutionally conserved processes, such as energetic metabolism, protein metabolism, lipid metabolism, redox homeostasis, ionic homeostasis, and stress response have been pointed to by several authors in response to acute or chronic salinity stressors [5–9]. In the trunk kidney of juvenile ayu Plecoglossus altivelis, differentially expressed proteins involved in energy metabolism, cytoskeleton, iron trafficking, protein metabolism were differentially expressed three weeks after seawater transfer [10]. In kidney papilla of mice, proteins that were significantly affected by the diuresis state were associated with cytoskeleton structure and signaling (e.g. Rho-GTPases), chaperones functioning and antioxidant functions [11]. However, acclimation mechanisms from hyposmotic to hyperosmotic environments have been investigated mainly in terms of changes in osmoregulation, while the effects of such osmotic stressors on the immune system remain largely unexplored. In teleost fish, chronic salinity stress has been correlated with increases in abundance and activity of immune cells, lysozyme activity, complement activity, total IgM levels, and higher

M. Schmitz et al. / Journal of Proteomics 167 (2017) 12–24

mucosal antibody response [12–16]. In mammals, a growing body of evidence suggests that chronic hyperosmotic stressors function as inflammatory mediators, triggering release of pro-inflammatory cytokines and activating innate immune cells [17,18]. Similarly, in fish, several studies have demonstrated the implications for inflammatory proteins during salinity stress. Gill proteome of Mozambique tilapia (Oreochromis mossambicus) upregulated heat shock proteins (HSP) and T-complex protein-1 members during short-term exposure at 34 ppt [8]. Gill proteome of seawater acclimated Japanese eels (Anguilla japonica) differentially expressed several immune proteins involved in the signaling of IL-6, IL-8, IL-9, HMGB-1, and iNOS [19]. In addition, broad-nosed pipefish infected with Vibrio sp. lowered the expression of two key anti-inflammatory molecules after three days exposure to salinity stress (6, 18 or 30 ppt) [20]. In zebrafish, it has been shown that disruption of local constitutive osmotic gradients is perceived as a danger signal and is necessary for the development of the inflammatory response after a wound [21]. Functional proteomics and transcriptomics technologies are becoming powerful tools to identify the sets of genes and proteins involved in salinity adaptation. In particular, the development of nano-flow liquid chromatography techniques associated with high sensitivity spectrometry has enabled large-scale quantitation of proteins. The suitability of label free quantitation to characterise the phenotype of an animal organ at a molecular level has already been debated in several independent studies [22,23]. Quantitation of ion current intensity was used for relative quantitation of kidney proteins, comparing fish faced with chronic salinity stress and fish raised in freshwater, combined or not with a bacterial infection. In this study, striped catfish juveniles (Pangasianodon hypophthalmus) were submitted to a gradual salinity stressor for 20 days and up to 10 ppt. Salinity stress was followed by infection with a virulent bacterial strain of Edwardsiella ictaluri, responsible for Enteric Septicemia in catfish [24] (the major cause of mortality in the catfish industry [25]). The main osmoregulatory parameters—plasma osmolality and gill Na+ K+ ATPase activity—were measured. Using spectrometry-based label free proteomics, we investigated the kidney proteome of salinity exposed fish upon infection or not. The resulting differentially expressed proteins were clustered depending on their biological meaning and visualised on metabolic pathway maps, with special emphasis on immunity. We hypothesised that prolonged exposure to salinity stressors may trigger the inflammatory response, leading to activation of innate immunity, and thereby confer protection against bacterial disease.

13

increased daily by 0.5 ppt per day for 20 days up to 10 ppt by adding natural marine salt (Instant Ocean, Belgium) mixed with tap water. On day 20, twelve fish from each tank were caught with a net, anaesthetised in 150 mg l−1 of MS 222, and intraperitoneally injected with 0.025 ml g−1 fish of a bacterial solution suspended in Hank's Balanced Salt Solution (105 Colony Forming Unit (CFU) ml−1). Mortality rate was monitored over ten days. Alongside the experiment, six fish per tank were randomly collected with nets on day 0, day 10, day 20 and day 23 (i.e. 72 h after bacterial infection), and anaesthetised. Blood was collected from the caudal vein using a sterile heparinised syringe within 5 min after capture. Fish were euthanized by cervical dislocation. Plasma was collected after blood centrifugation (4 °C, 7000g, 10 min) and kept at −80 °C, pending analysis. Branchial left arches 1 and 2, and the kidney, were sampled and immediately frozen in liquid nitrogen. The oxygen level (5.7 ± 0.5 mg l−1; pH 8.4 ± 0.24; temperature 28.2 ± 0.1 °C) −1 and nitrogen level (N-NO− ; N-NO− 3 3.55 ± 2.27 mg l 2 0.019 ± −1 0.005 mg l−1; N-NH+ 0.19 ± 0.25 mg l ) were monitored daily in 3 the outlet pipe. Measured salinities were close to expected values (±0.15 ppt) (Fig. 1). 2.2. Bacterial challenge Bacterial challenges were performed according to Schmitz et al. [7]. Briefly, a virulent strain of Edwardsiella ictaluri (TNA 015, Can Tho University, Vietnam) was cultured on Brain Heart Infusion (BHI) Agar (Sigma) at 28 °C. A reference bacterial solution (optical density of 0.1 at 590 nm, 109 CFU ml−1 in BHI) was numbered by serial counts of CFU on agar plates (dilution 100 to 10 [10]) as well as on black filters in fluorescent microscopy after 4′-6′-diamidino-2-phenylindole staining. A lethal dose of 50% (72 h), corresponding to an intraperitoneal injection of 0.025 ml g−1 fish of a bacterial solution containing 106 CFU ml−1, was determined in a preliminary experiment; hence, 105 CFU g−1 fish was injected. Infection was confirmed on the kidney by the BRLFD-CER group (Belgian Reference Laboratory of Fish Diseases, Centre d'Economie Rurale, Aye, Belgium). Kidneys of infected fish were crushed with a sterile mortar and pestle, and inoculated on Sheep Blood Agar. Agar plates were examined after two days at 26 °C and demonstrated growth of apparent homogenous gram negative bacteria (after gram stain). A 96wells Gram Negative Biolog identification system MicroLog, version 4.0, was used for speciation. The test confirmed with 100% probability for E. ictaluri and 0.783 of similitude with the reference strain. Additional confirmation of bacterial infection was performed through the detection of bacterial 16SrRNA of Edwardsiella ictaluri in quantitative PCR [26].

2. Materials and methods 2.3. Osmoregulatory parameters 2.1. Experimental design and statistical rationale Investigations were conducted in UNamur following the guidelines for animal care and use, in compliance with European regulations on animal welfare (Protocol No. KE 12/189). Minimum sample size was determined as followed on statistical calculation n = (Zα + Zβ)2. (Sd2/Δ2), with Zα = 1.64 (α = 5%), Zβ = 0.84 (power consideration), Sd: standard deviation, Δ: difference of the means in freshwater and brackish water. Several blood immune parameters (e.g. lysozyme, complement, blood cells) were tested in a dozen of animals in a preliminary experiment and the resulting data allows us to determine the minimum sample size of the present experiment. One week old striped catfish (2000 fish) were imported to the University of Namur from the Nam Sai fish farm in Ban Sang, Thailand. Fish were maintained in a recirculating aquaculture system under constant photoperiod (12L:12D) at 28 °C and fed daily ad libitum with commercial dry pellets (Troco Supreme 16, Coppens, The Netherlands). At three months (body weight 42 ± 11 g), fish were transferred into two experimental circuits, each including four aerated 100 l tanks, and acclimated over ten days to these new housing conditions. The control group was maintained in fresh water (0 ppt) during the whole experiment. In the experimental group, salinity was

Gill filaments were homogenised in sterile Potter for 2 × 30 s, in cold SEI buffer (Sucrose 0.25 M, EDTA 1 mM, Imidazole 50 mM, pH 7.4)

Fig. 1. Striped catfish salinity experiment regimens. Striped catfish were exposed (or not) to increasing salinity during 20 days in two independent recirculating systems. FW: Freshwater; LSW: Low saline water (0–10 ppt). The arrows represent bacteria inoculation at day 20.

14

M. Schmitz et al. / Journal of Proteomics 167 (2017) 12–24

containing a protease inhibitor cocktail (Sigma). Homogenates were centrifuged twice to remove main debris (10,000g, 5 min, 4 °C). Na+ K+ ATPase activity was measured in duplicates following Mc Cormick [27]. One unit of Na+ K+ ATPase activity represents the consumption of 1 μmol NADH min−1 ml−1. Plasma osmolality was measured in duplicates with a micro-osmometer (Type 6, Löser Messtechnik, Germany). 2.4. Statistical analysis Regarding osmoregulatory parameters, heterogeneity of variances was tested by Levene test, and normality was checked by Shapiro– Wilk test. Data were analysed by two way analysis of variance ANOVA 2 (two factors, ‘salinity’ and ‘day’, fixed parameters), followed by pairwise multiple comparisons procedures by Scheffe test (p b 0.05) in SigmaPlot version 12. Owing to the absence of non infected fish on day 23, the effect of infection is not disentangle from the effect of “day”. Data are represented as the mean ± SEM, with n = 4 (4 tanks). In order to decrease the residual variability, “tank” was used as the experimental unit and represented the mean of the fish caught in the tank at each sampling day. Mortalities per tank were monitored each day over ten days. Daily mortality rates in freshwater and brackish water were statistically compared using Student t-test. Data are represented as the mean ± SEM (n = 4). 2.5. Quantitative label free proteomic analysis 2.5.1. Proteolysis Proteomics analysis was performed on samples on days 20 and 23. We implemented a cost-efficient strategy to study proteome responses and omitted individual protein expression variation. Fish from the same tank were pooled (n = 3, pool of six fish) and the tank was considered as the experimental unit. Whole kidney was ground in 9 M urea and sonicated. Then, 10 μg of the resulting proteins were reduced with DTT 2.8 mM (60 °C, 30 min), modified with 8.8 mM iodoacetamide in 100 mM ammonium bicarbonate (dark, room temperature, 30 min), and digested in 2 M Urea, 25 mM ammonium bicarbonate with modified trypsin (Promega) at a 1:50 enzyme-to-substrate ratio (overnight, 37 °C). An additional second digestion was performed over 4 h. Then, the tryptic peptides were desalted using C18 stage-tip (Harvards), dried, and suspended in 0.1% formic acid. 2.5.2. Mass spectrometry The resulting peptides were loaded onto a C18 trap column (0.3 × 5 mm, LC-Packings), connected to a homemade capillary column (25 cm, 75 μm ID) packed with Reprosil C18-Aqua (Dr Maisch GmbH, Germany) in 0.1% formic acid in water. They were analysed by LC-MS/ MS using a Q Exactive plus mass spectrometer (Thermo) fitted with a capillary HPLC (easy nLC 1000, Thermo).The peptide mixture was resolved with a linear gradient from 5 to 28% of 95% acetonitrile with 0.1% formic acid for 120 min, followed by a gradient of 28–95% for 5 min and then at 95% acetonitrile with 0.1% formic acid in water for 25 min at flow rates of 0.15 μl min−1. Mass spectrometry was done in a positive mode (m/z: 350–1800), using repetitively full MS scan, followed by high collision induces dissociation (HCD, at 35 normalized collision energy) of the ten most dominant ions (N 1 charges) selected from the first MS scan. A dynamic exclusion list was enabled with exclusion duration of 20 s. 2.5.3. Data analysis The MS raw data were analysed using the MaxQuant software (version 1.5.2.8) for peak picking and quantitation, followed by identification using the Andromeda search engine [28], searching against the Characiphysae section of the NCBI-NR database (Jan 2016, 54,767 proteins) with mass tolerance of 20 ppm for the precursor masses and fragment ions. Since there was almost no protein information for

Pangasianodon hypophthalmus, the Characiphysae section that contains several fish types (mainly Ictalurus punctatus and Astyanax mexicanus) was used. Methionine oxidation was set as variable post-translational modifications, and carbamidomethyl on cysteine as a static one. Minimal peptide length was set to six amino acids, and a maximum of two miscleavages was allowed. Razor and unique peptides are available in the Supplementary table. To eliminate identifications from the reverse database and common contaminants, peptide- and protein-level false discovery rates (FDRs) were filtered to 1% using the target–decoy strategy. Data were quantified by label free analysis using MaxQuant, based on extracted ion currents (XICs) of peptides enabling quantitation from each LC/MS run for each peptide identified in any of experiments. Briefly, for every peptide, corresponding total signals from multiple runs were compared to determine peptide ratios. Pairwise peptide ratios were only determined when the corresponding peak was detected in both LC-MS runs. A robust estimation of the protein ratio is calculated as the median of pairwise peptide ratios interpolated with the square root of the ratio of summed-up intensities. The analysis was done after first recalibrate of the retention times. Data were transformed to log2 intensities in order to get a normal distribution. Label free quantitation was performed according to Cox et al. [29]. Pearson correlation was done between the samples and as the so correlation between the nonnormalized intensities and the LFQ normalized intensities were more the 0.97, we decided to use the raw intensities and thus eliminate errors due to over normalization. Since several organisms were included in the database, homolog proteins were clustered as one protein group, and only proteins that were identified with at least two peptides are listed [29]. Missing values were replaced by the minimum intensity detected in the experiment (=18 in log2 intensity). T-test with Permutationbased FDR (with 250 randomization, threshold value = 0.05) was done using Perseus 1.5 between the freshwater and saline groups. The dataset has been deposited into ProteomeXChange Contorsium via the Pride partner repository with the dataset identifier PXD004571, username [email protected], password Oov046pi. Peptide and protein tables are available in the Supplementary table. Volcano plots were performed in R. 2.5.4. Clustering and functional analysis Hierarchical clustering and functional analysis were performed with the differentially expressed proteins between the freshwater (infected or not) condition and the brackish water (infected or not) condition. For these analysis, proteins which contain more than one missing values were discarded from the dataset. Hierarchical clustering of log2 intensities using Euclidian distances was performed in Permutmatrix [30]. A homogeneous annotation of the dataset was performed by homology with the zebrafish, using GI accession numbers. Functional Analysis were performed in The Database for Annotation, Visualisation and Integrated Discovery (DAVID) version 6.831 and Ingenuity Pathway Analysis (IPA) version 01-10 (Quiagen, Bioinformatics). In total, DAVID processed 352 proteins whereas IPA processed 336 proteins. 2.5.5. Validation Using a Speed Mill Vac Bound Homogeniser, kidney lysates were obtained by homogenising kidney tissue for 2 × 30 s in the following buffer (1:3): Tris-HCl 50 mM, NaCl 150 mM, SDS 0.1%, Triton X-100 0.1%, apopritin 0.001 mg ml−1, pH 8. Lysates were then sonicated 3 × 10 s at 45 kHz and 5 × 1 s at 65 kHz, both on ice, and centrifuged at 10000g for 10 min to remove main debris. Total protein abundance in the samples was measured using Pierce 660 nm Protein Assay Reagent (22,660, ThermoScientific). Western blot analyses were performed to validate the specificity of antibodies following Schmitz et al. [15] using anti-Cdc42 610,929 (BD Biosciences), anti-vinculin V9131 (Sigma), and anti-Raf1 antibodies 610,151 (BD Biosciences) 1/2000. Western Blots of each tested antibody resulted in a unique band detected around 20 kda (Cdc42), 70 kda (Raf1), and 110 kda (Vinculin). Quantification was performed by dot blotting following Schmitz et al. [15]. Briefly, 1 μl of

M. Schmitz et al. / Journal of Proteomics 167 (2017) 12–24

proteins extract was spotted in triplicate onto the prewetted PVDF membrane and allowed to dry out for 30 min at room temperature, blocked in PBS-Tween containing 5% skimmed milk for 1 h at room temperature, and probed in the same manner as for western blotting. After immunodetection, a protein loading control was performed using Coomassie Brilliant Blue R-250 (Biorad) [32]. Dot quantification was done using Image J software. Data were analysed by two-way analysis of variance ANOVA 2, followed by pairwise multiple comparisons procedures by Scheffe test (p b 0.05) in SigmaPlot version 12. Data are represented as the mean ± SEM (n = 3). 3. Results 3.1. Osmoregulatory response of catfish to chronic saline stress The osmoregulatory capacity of striped catfish was investigated through plasma osmolality (Fig. 2A) and gill Na+ K+ ATPase (Fig. 2B). Plasma osmolality of fish kept in freshwater averaged 265 ± 2.7 mosm l−1 and significantly increased in brackish water to reach up to 292 ± 1.8 mosm l−1 at 10 ppt (p b 0.001). On day 23 (infected fish), no significant changes were observed in plasma osmolality. Na+ K+ ATPase activity did not significantly vary between fish in freshwater and brackish water on days 0, 10 and 20, and was comprised between 0.28 and 0.64 U mg−1 gill min−1. On day 23 (infected fish), Na+ K+ ATPase significantly decreased up to 0.45 ± 0.06 U mg−1 gill min−1 in brackish water, compared to freshwater values (0.69 ± 0.076 U mg−1gill min−1) (p b 0.01). 3.2. Susceptibility to enteric septicaemia of catfish Fig. 2C shows cumulative mortalities over ten days after bacterial inoculation in the freshwater and saline water groups. Regarding daily mortality rate monitoring, no significant difference was observed

15

between the freshwater and brackish water groups. Mortalities began two days and three days after bacterial inoculation in brackish water and freshwater respectively. Cumulative mortalities after ten days reached 79 ± 8% in freshwater and 67 ± 14% in brackish water. 3.3. Proteomic study of catfish kidney proteome in response to chronic saline stress 3.3.1. Volcano plots and hierarchical clustering Complete analysis resulted in identification of 2483 proteins, among which 400 proteins show significant changes in abundance, with elevated salinity. Volcano plots are provided in Fig. 3 for the 400 differentially expressed proteins, in non-infected and infected fish respectively. In non-infected fish, 366 proteins were differentially expressed with salinity (in red, p b 0.05), whereas in infected fish, 67 proteins were differentially expressed with salinity (in red, p b 0.05). Hierarchical clustering was performed on differentially expressed proteins in non-infected fish (366 proteins) and infected fish (67 proteins) (Fig. 3). The experimental groups were divided into two clusters (freshwater and brackish water) and biological replicates were clustered together. In non-infected fish, salinity stress upregulated 353 proteins and downregulated 13 proteins (p b 0.05). In infected fish, salinity stress upregulated 51 proteins and downregulated 16 proteins (p b 0.05). 3.3.2. Functional pathway analysis of kidney proteome: an overview A pathway analysis of the whole dataset was performed to isolate the major specific functional pathways in DAVID and IPA using Danio rerio GI accession number. Table 1 summarises the major pathways that were significantly enriched in this study (p b 0.05), and which are involved in energy metabolism, carbohydrate metabolism, amino acid metabolism, protein processing, cell structure, signaling, immunity, and detoxification. Proteins related to higher metabolism were strongly represented. Regarding energy production, salinity upregulated enzymes involved

Fig. 2. Osmoregulatory response and mortality curve of non-infected and infected striped catfish exposed to freshwater and brackish water. A, B: Plasma osmolality and gill Na+ K+ ATPase activity of catfish exposed to freshwater (black) and brackish water (grey) over 20 days and three days after bacteria inoculation (day 23); C: Mean cumulative mortality for ten days after inoculation of Edwardsiella ictaluri into catfish exposed to freshwater (black) and brackish water (grey). The values are presented as the mean ± SEM (n = 4), analysed by Student's t-test. **p b 0.01 compared to freshwater values; ***p b 0.001 compared to freshwater values.

16

M. Schmitz et al. / Journal of Proteomics 167 (2017) 12–24

Fig. 3. Volcano plots data representation and hierarchical clustering. Volcano plots represent the differentially expressed proteins with salinity selected in our analysis. The red dots are the differentially expressed proteins with salinity (p b 0.05) in non-infected and infected fish, respectively.

in fatty acid alpha and beta oxidation, glycolysis, citrate cycle, pentose phosphate pathway, and ketone bodies synthesis in both non-infected and infected fish, on average 1.6-fold. Regarding protein metabolism, several clusters implicated in the biosynthesis of amino acids, protein localisation and transport, proteases (mainly proteasome subunits), and protein modifications were highlighted. Enzymes involved in amino acid metabolism were upregulated by on average 1.6-fold—particularly those involved in amino acid biosynthesis and branched amino acid degradation. The cysteine acid decarboxylase, involved in hypotaurine synthesis, a key osmolyte during osmotic shock, was upregulated by N43-fold with salinity. Regarding protein processing, salinity upregulated by 1.8-fold several proteins involved in COP II vesicle transport via Sec-dependent pathway, as well as folding proteins (e.g. HSP and ubiquitin ligase complex). This pathway also includes phospholipase A2, which is involved in inflammatory processes, and which upregulated by 3.7-fold with salinity in infected fish, but downregulated by 2.9-fold with salinity in non-infected fish. Regarding protein modifications, salinity tended to increase the abundance of proteins involved in pyridoxal phosphate dependent modifications and ubiquitination.

The next pathway groups gather proteins involved in carbohydrate metabolism, including amino sugar and nucleotide sugar metabolism, N-glycan, and fructose/mannose metabolism, which were upregulated by salinity mainly in non-infected fish. In particular, the Nacetylglucosamine kinase was upregulated by 5.4-fold with salinity and by 15.5-fold on day 23 (infected fish). Proteins related to detoxification involved components of CYP450 (1A and 20A), Heat Shock Protein 90, cullin3, sulfotransferase SUT1A1, aldehyde dehydrogenases and enzymes related to glutathione metabolism (glutathione peroxidases and transferases). These proteins were all upregulated by on average 2-fold in non-infected and infected fish in brackish water. Strongly represented proteins, particularly in non-infected fish, were members of the small GTPase superfamily including G beta gamma, Rho GTPases (i.e. members of Rac, Cdc42 and RhoA superfamily) and Rabtype GTPases. With the exception of Rab-11, GTPases were all upregulated in saline water. Fig. 4 shows the Rac signaling pathway (Rho GTPases superfamily) which contain many representatives of signaling proteins that might be in the hyperosmotic transduction signal in our experiment. Representatives of the Rac superfamily were upregulated in brackish

M. Schmitz et al. / Journal of Proteomics 167 (2017) 12–24 Table 1 Summary of the functional analysis of differentially expressed proteins in striped catfish (Pangasianodon hypophthalmus, S.) following saline stress (0 to 10 ppt) derived from IPA and DAVID. Count: proteins involved in the pathway; p-value: to examine the significance of protein-term enrichment with a Fisher's exact test; FC: Mean fold change with salinity in non-infected fish; FCinf: Mean fold change with salinity in infected fish. NS: non significant. Pathways Energy production Glycolysis/gluconeogenesis Citrate cycle Mitochondrial fatty acid beta oxidation Peroxisomal fatty acid alpha oxidation Pentose phosphate pathway Synthesis and degradation of ketone bodies Amino acid metabolism Biosynthesis of amino acid Tryptophane metabolism Glycine, serine, threonine metabolism Histidine metabolism Valine, leucine, isoleucine degradation Alanine, aspartate, glutamate metabolism Arginine, proline metabolism Protein metabolism Protein processing in endoplasmic reticulum Protein localization and transport Protease Pyridoxal phosphate dependent transferase Protein ubiquitination pathway Carbohydrate metabolism Amino sugar and nucleotide sugar metabolism N-glycan biosynthesis Fructose and mannose metabolism Immunity and detoxification Metabolism of xenobiotics Phagocytosis Leukocyte extravasion Acute phase signaling Production of NOS and ROS by macrophages Signaling pathways G beta gamma signaling Paxillin signaling (focal adhesion) Integrin signaling Signaling by RhoGTPases family Rac family Ephrin signaling Calcium signaling Cell remodelling and adhesion Tight junction Remodelling of epithelial adherens junction Actin cytoskeleton signaling Regulation of actin-based motility by Rho Ion transport

Count

p-Value

FC

FCinf

20 13 16 5 7 7

3.3E−11 7.8E−8 6.8E−8 4.6E−7 1.3E−3 1.8E−3

1.7 1.5 2.1 1.8 1.6 2.3

1.2 1.3 2.0 2.0 1.0 1.3

21 13 11 8 10 8 8

7.3E−11 2.0E−7 1.0E−5 2 .2E−5 1.9E−4 9.3E−4 6.7E−3

1.9 2.0 1.8 2.1 1.8 1.8 1.8

1.7 1.8 1.3 2.2 1.4 1.7 1.8

15 23 11 7 12

1.3E−2 2.2E−2 3.0E−2 2.0E−4 2.8E−4

1.7 2.2 1.7 3.3 1.8

1.6 1.3 1.5 7.4 1.5

9 6 6

3.3E−3 4.9E−2 2.3E−2

2.3 1.9 1.6

1.0 1.3 1.1

18 13 18 8 7

1.5E−7 2.3E−2 1.8E−6 2.8E−3 2.0E−2

1.9 3.0 1.7 2.0 1.6

1.8 1.6 1.2 NS 1.4

6 12 16 17 8 11 10

1.5E−3 6.7E−8 9.3E−8 8.8E−8 2.5E−4 3.4E−5 2.2E−4

1.5 1.6 1.8 1.7 1.7 1.9 2.2

1.2 NS 1.4 1.2 1.5 NS NS

9 12 18 7 5

6.2E−4 1.1E−6 4.3E−9 3.0E−4 2.5E−3

1.8 1.9 2.1 1.9 1.5

1.2 NS 1.3 NS 1.3

water (regardless the infections status), as shown by the red shading in Fig. 4. Particularly, the NCK-associated protein-1, member of the WAVE complex, required for actin-filament reorganization and endosome trafficking, increased by 3.8 fold in brackish water in infected fish. Lastly, the pathway related to cell morphology and adhesion includes proteins involved in cytoskeleton organisation (actin polymerization and nucleation, focal adhesion assembly, actin stabilization, formation of actin-stress fibres). Few proteins specific to cells junctions such as plakoglobin (adherens junction) were also upregulated in brackish water compared to freshwater conditions. Ultimately, several ion transporters were upregulated in brackish water (at the exception of Na/HCO3 transporter) and these proteins include ATP synthase subunits, calcium transporters, V-type proton ATPases, transferrin, and chloride channel protein 4 by 1.4-fold on average. 3.3.3. Functional analysis of proteins involved in the immune response Functional analysis of immune proteins was conducted on IPA by selecting “Cellular Immune Response”, “Cellular Stress and Injury”, and “Cytokine signaling” in the canonical pathways to display. Only proteins

17

involved in phagocytosis were observed using KEGG pathway maps on DAVID interface because DAVID displays a more general overview of the phagocytic pathway. Fig. 5 shows the response of immune pathways to experimental salinity stress in kidney of striped catfish. Significantly enriched immune pathways according to Fischer's Exact test calculations in non-infected fish in this study are illustrated in Fig. 5A (Z score N 1, p-value b 0.05). Similar tendency was observed in infected fish. In total, 23 immune pathways were significantly enriched in the dataset and were related to phagocytosis, leukocyte's extravasion, cytokine signaling (chemokines and interleukins) and oxidative stress response. The color of the bars represents the calculated Z-score which is the statistical measure of correlation between relationship direction and protein abundance. Fig. 5A shows 11 immune pathways that were expected to be increased (orange bars, Z-score N 2), 9 immune pathways for which IPA failed to predict the activity pattern (grey bars) and 3 immune pathways for which the Zscore was not significant (white bars). Regarding leukocytes adhesion, the pathway gathered several non-specific proteins including MAPK proteins, members of the focal adhesion complex and actin-binding proteins involved in cell to cell adhesion. Particularly, the leukocyte adhesion molecules very-late antigen 4 (VLA-4) of the beta 1 integrin family, normally expressed at leukocytes' plasma membrane, is upregulated by 1.8 fold in brackish water in non-infected fish. The canonical pathways related to cytokines signaling, iNOS signaling and HMGB1 (High Mobility Group Blot 1) signaling gathered common proteins which consist in small GTPases and members of the MAPK signaling. Downstream effect analysis were performed in order to identify key biological processes involved in immunity and influenced by experimental salinity stress. As the major parts of the immune proteins were involved in the inflammatory response, downstream effects analysis were performed on the inflammatory process. Fig. 5B shows the downstream effects on the inflammatory process by differentially expressed proteins with salinity. Several steps of the inflammatory process, including platelets aggregation, extravasion of leukocytes cell lines (i.e. binding and adhesion), degranulation and phagocytosis, were predicted to be increased during experimental salinity stress (p-value b 0.05, Z-score N 2) (Fig. 5B). Strongly represented proteins that were significantly differentially regulated in brackish water were proteins involved in the phagocytic process. Fig. 6A details the different steps of the phagocytic process and the NAPDH oxidase complex which is activated in the phagosome during respiratory burst (derived from DAVID). Differentially expressed proteins in brackish water are denoted with red stars. Salinity upregulated proteins involved in the formation of phagocytic cup and internalization (F-actin, coronin-1A and 1C, Rac-1), phagosome acidification (V-type proton ATPases), endosomal maturation (syntaxin-7, dynein, complement protein 3, Rab-7A), activation mechanisms of NAPDH oxidase (Rac-1, cytochrome b-245), phagolysosome transporters (the Sec61 complex, TAP1), and phagocytic promotor receptors (complement receptor 3, integrin αMβ2). These proteins were upregulated in brackish water by on average 3.0-fold in non-infected fish and 1.6-fold in infected fish. In particular, Rab-7A and syntaxin-7 increased by, respectively, 7.2fold and 15.2-fold in saline water in non-infected fish, and by 7.3-fold and 28.9-fold in brackish water in infected fish. Fig. 6B displays phagocytosis network derived from IPA analysis. The response of complement protein 3, Rac-1, Rab-7A, cytochrome b-245, MAPK and catalase were already discussed above. In addition, the abundance of downstream effectors, including myosin, Rab-35, lumican, caveolin-1, GRK-6, responsible for plasma membrane remodelling that allows particle internalization in a membrane-enclosed phagosome, was upregulated by 2.3-fold in non-infected fish during salinity stress. Whereas during infection, salinity stress induced upregulation of lumican (1.7-fold) and caveolin-1 (5.2fold), downregulation of GRK-6 (1.6-fold) and Rab-35 (3.5-fold) and had no effect on the abundance of myosin. Moreover, the chloride channel-4 protein, involved in phagosomal acidification, increased by 2.1 fold in non-infected fish and 1.7-fold in infected fish. Last of all, two proteins

18

M. Schmitz et al. / Journal of Proteomics 167 (2017) 12–24

Fig. 4. Rac signaling pathway response to experimental salinity stress in kidney of striped catfish (Pangasianodon hypophthalmus, S.). Red shading denotes upregulation in expression in brackish water relative to fresh water. The darker the color, the higher the fold change. The pathway was built using Ingenuity Pathway Analysis software.

implicated in β-1 integrin signaling and initial antigen recognition were also upregulated by on average 1.6-fold in saline condition. Acute phase proteins includes complement3, fibrinogen, plasminogen, and MAPKp38 components signaling, upregulated by 1.5- to 2-fold and the AMBP (alpha macroglobulin/bikunin protein), upregulated by 3.6-fold in brackish water in non-infected fish. In infected fish, complement3 increased by 1.6-fold, plasminogen by 1.3 fold whereas neither fibrinogen nor the AMBP were differentially regulated in brackish water. The final representatives of immune proteins gather those involved in T-cells activation and regulation (e.g. plastin 2, nitrilase homolog 1, proto oncogene cRel, galectin-9, IL-2 enhancer, and cleft lip and palate transmembrane protein). These proteins were upregulated with salinity by 2.2-fold in non-infected fish and 1.4-fold in infected fish. In particular, proto-oncogene cRel, involved in T-cell differentiation, lymphopoiesis and activation of NF-κB signalisation, was upregulated by 4.6-fold in brackish water. 3.3.4. Upstream regulator analysis Upstream analysis was conducted on IPA software independently of the infection status. Top 10 regulators (transcription and nuclear factors) are summarized in Table 2. Among them, 4 regulators belong to the Peroxisomal proliferator-activated receptors family (subfamily α, γ, δ).

PPARs are predicted to be activated in brackish water and target proteins mainly involved proteins implicated in lipid metabolism in the dataset. In addition, 4 transcription regulators (Hepatocyte nuclear-factor 4A, Tumor protein p53, Myc proto-oncogene protein, and mitochondrial transcription factor A), one growth factor (Transforming growth factor β) and one endogenous molecule (D-glucose) were also predicted to increase in brackish water. 3.3.5. Validation Vinculin, Cdc42, and Raf1 were chosen for their key implication in three significant major pathways highlighted in our proteomic analysis: cytoskeleton/focal adhesion, MAPK signaling pathway, and small GTPases superfamily. Results from the validation experiment are shown in Fig. 7. Regarding Cdc42 abundance, salinity and infection induced significant upregulation, by 1.2- to 2.3-fold in dot blotting, and by 1.3- to 1.4-fold in proteomics. According to dot blot analysis of Raf1, salinity alone did not induce significant changes, while infection induced upregulation by 1.6-fold, and the combined effect of salinity and infection induced upregulation by 2.1-fold. In proteomics, salinity, infection, and their combination induced upregulation by 1.4-fold, 1.3fold and 1.8-fold, respectively. Lastly, in dot blotting, neither salinity alone nor infection alone induced a significant change in vinculin

M. Schmitz et al. / Journal of Proteomics 167 (2017) 12–24

19

Fig. 5. Immune pathways response to experimental salinity stress in kidney of striped catfish (Pangasianodon hypophthalmus, S.). A. Immune canonical pathways expression (Z score N 1, p value b 0.05). The calculated z-score indicates a pathway with proteins exhibiting overall increased abundance level (orange bars) or decreased abundance level (blue bars). The p value measures the likelihood that association between the differentially expressed proteins in the data set and the pathway is due to random chance. The threshold indicates the minimum significance level [scored as p-value from Fisher's exact test, set here to 0.05]. The ratio refers to the number of proteins from the dataset that map to the pathway listed divided by the total number of proteins that map to the canonical pathway from within the Ingenuity Pathway Analysis knowledgebase. B. Heat map derived from proteins involved in the inflammatory response (Z score N 1, p value b 0.05) derived from Ingenuity Pathway Analysis. The color scale represents the Z-score.

abundance, while the combination of both infection and salinity induced significant increase in vinculin abundance, by 1.28-fold. In proteomics, salinity alone induced upregulation by 1.3-fold, infection alone induced no significant change, while the combination of both infection and salinity upregulated the vinculin by 1.3-fold. 4. Discussion In this study, salinity was responsible for upregulation of various metabolic pathways in the kidney of striped catfish, which are mainly involved in osmoregulation, energy production, protein biosynthesis and transport, detoxification, and immunity. Moreover, those pathways tend to respond to salinity stress equivalently in non-infected and infected fish. However, the number of proteins significantly affected by salinity decreases on day 23 (infected fish) and fold changes were generally lower, which suggests that both stressors may act in synergy on protein abundances, thereby sharing major metabolic routes. 4.1. Acclimation to salinity stress of kidneys' excretory functions Plasma osmolality in freshwater were close to those in catfish reported by other authors [33,34], differing by b10%. At 10 ppt, plasma osmolality significantly increased whereas gill Na+ K+ ATPase activity remained stable. In channel catfish, long-term acclimation to brackish water increased plasma electrolytes content and fish produced smaller amounts of concentrate urines [35]. Therefore, salinity stress requires significant adaption of renal excretory mechanisms in order to maintain body volume and ensure clearance of metabolite wastes. In this study, salinity stress downregulated Na/HCO3 transporter, but upregulated chloride channel 4, as well as proteins involved in the biosynthesis of

osmolytes (particularly glycerol, taurine, proline, sarcosine, betaine, and spermidine), independently of the infection status. In eukaryotes, osmolytes can be safely up/downregulated in order to conform to extracellular fluids' osmolarity, with little impact on cellular function [36]. In addition, osmolytes may act as antioxidants, stabilise macromolecules, and counteract perturbants in non-interchangeable ways [9]. In Mozambique tilapia, exposure to salinity stress elevated the abundance of myo-inositol, creatin, and glycine in kidney, brain, and muscles [37]. In cultured lens of Atlantic salmon (Salmo salar), varying media osmolality correlated with N-acetylhistidine content [38]. Therefore, it is likely that the intracellular osmolytes content increased in catfish kidney in order to conform kidney tubules to progressive elevation of urine tonicity. In addition, salinity stress upregulated various proteins related to actin network organisation in non-infected and infected fish. Upregulation of cytoskeleton proteins/transcripts during salinity stress in fish proteome/ transcriptome has already been mentioned by other authors [5,7,39]. Following osmotic stressor, eukaryotic cells are able to rearrange their cytoskeleton in order to increase cell rigidity and oppose cell shape changes. In vitro, increase in media osmolarity is responsible for filamentous actin polymerization and migration to periphery [40,41]. Changes in the actin cytoskeleton structure may play a key role in resistance to bacterial infection, because they may indirectly generate a specialized niche for bacteria replication and interfere with the positioning of bacteria-containing vacuoles [42]. In particular, a number of enriched cytoskeleton proteins in this study are connected to focal adhesion assembly, which is known to be involved in various intracellular signaling pathways—such as Arp2/3 complex and Rho-family GTPases—responsible for filamentous actin assembly into stress fibres and protrusion forces [43]. In dogfish sharks (Squalus acanthias), proteome analysis revealed that cytoskeleton-associated Rho-GTPases were overrepresented in osmoregulatory

20

M. Schmitz et al. / Journal of Proteomics 167 (2017) 12–24

organs including kidney [44]. In mammals, exposure to hypertonic fluids induced focal adhesion assembly and activated focal adhesion kinases [45]. In fish, Marshall et al. suggest that focal adhesion kinases might be key scaffolding proteins in the regulation of the Na/K/Cl transporter in the chloride cells of the seawater-acclimated killifish (Fundulus heteroclitus) [46]. Remodelling of actin cytoskeleton is of key importance during the immune response and is the basis of several immune pathways such as phagocytosis, cell migration and chemotaxis. 4.2. General stress response Multiple pathways of energy production, lipid metabolism, carbohydrate, and amino acid metabolism were strongly represented, consistent with a higher metabolism and elevated need for energy. Upregulation of these pathways in fish faced with multiple of environmental stressors has been largely discussed by other authors, including in striped catfish [5–7,47]. In this study, salinity also induced upregulation of mechanisms involved in detoxification, including antioxidant, glutathione metabolism, and CYP450 1A/20A. Similarly, seawater-acclimated gilthead seabream, Sparus aurata, overexpressed glutathione peroxidase and transferase [5]. In addition, upregulation of CYP450 1A has also been observed during salinity stress in seabass (Dicentrachus labrax), rainbow trout (Oncorhynchus mykiss), and gilthead seabream [5,48,49]. Regarding protein metabolism, two co- or post-translational protein modifications were enriched in this study, N-glycosylation, and pyridoxal phosphate modification. In particular, proteins involved in pyridoxal phosphate modifications were upregulated by 3.3-fold in non-infected fish and 7.4-fold in infected fish. While involvement of N-glycosylation or pyridoxal phosphate modification in salt tolerance has never been

investigated, a single study does reveal that prevention of N-glycosylation in the kidney of frogs (Xenopus laevis) impairs ionic transporter insertion into plasma membrane and reduces its functional expression [50]. 4.3. Effect of salinity on immune-related pathways In this study, salinity enhanced the abundance of proteins involved in several pathways involved in innate immune defences, especially inflammation (Fig. 5 A, B). Many differentially expressed proteins in the dataset were related to leukocyte's diapedesis, phagocytosis and respiratory burst activity. Phagocytosis relies on a network of endocytic vesicles to deliver cargo from new formed phagosome to lysosome for degradation [51,52]. In this study, salinity upregulated proteins involved in the regulation of the phagocytic process, formation of the phagocytic cup (including opsonins), maturation and acidification of the phagosome, activation of respiratory burst (i.e. NADPH oxidase complex), production of iNOS, and phagosome localisation (Fig. 6 A,B). In particular, Rab7 and syntaxin 7 were upregulated with salinity by 7-fold and 29-fold respectively. On the one hand, Rab7 is known to mediate the redistribution of late phagosome from cell periphery to the juxtanuclear region, together with microtubule-based motor dynein, in order to render it accessible to the host immune system [53]. On the other hand, syntaxin 7 specifically induces the fusion between late endosome and lysosome [54]. Stimulation of the phagocytic process during salinity stress has also been described in other fish species. In brown trout (Salmo trutta), and Mozambique tilapia, the phagocytic activity of renal leukocytes increased after seawater acclimation [14,55]. Moreover, respiratory burst is

Fig. 6. Phagocytosis pathway response to experimental salinity stress in kidney of striped catfish (Pangasianodon hypophthalmus, S.). A. Canonical pathway of phagocytosis. Red stars denote measured proteins with upregulation in expression in brackish water relative to fresh water. The pathway was built using the Database for Annotation, Visualisation and Integrated Discovery Beta. B. Pathway network. C3: complement protein-3; CAT: catalase; CAV1:caveolin-1; CLIC4:chloride intracellular channel-4; CORO1A/C:coronin-1A/C; CYBB: cytochrome b245; GRK6:G-protein coupled receptor kinase 6; ITGB1:integrin-β1; LUM: lumican; MAPK1/14:mitogen-activated protein kinase 1/14; MYH9:myosin heavy chain 9; RAB7A/35:Rasrelated protein RAB7A/35; RAC1: Ras-related C3 botulinum toxin substrate 1; RACK1: receptor of activated protein C kinase1. The pathway was built using Ingenuity Pathway Analysis.

M. Schmitz et al. / Journal of Proteomics 167 (2017) 12–24

21

Fig. 6 (continued).

enhanced after acute seawater transfer in Mozambique tilapia and rainbow trout [4,12]. Enhancement of the phagocytic process may be beneficial in fish disease resistance. In particular, the survival and replication of some intracellular bacteria, such as Edwarsiella sp., depend on their capacity to inhibit the last steps of the host's phagocytic process, mainly the fusion between bacteria-containing phagosomes and lysosome [56]. A key immunological consequence of phagosome maturation is the presentation of foreign antigens by Major Histocompatibility Complex II, which leads to the activation of antigen-specific CD4+ T-cell [57]. In this study, several proteins thought to be involved in T-cell activation, such as plastin 2, proto-oncogene cRel, galectin-9, IL-2 enhancer, and cleft lip and palate transmembrane protein-1 were upregulated with salinity. In particular, salinity upregulated up to 5-fold the proto-oncogene cRel, a Rel/NF-kB family transcription factor that plays a crucial function in lymphoid cells' development and interleukin production via CD28mediated signaling [58]. In mammals, elevated plasma osmolarities (i.e. by hypertonic saline treatment or KCl injection) increased T-cells proliferation and activity via MAPK-p38 stimulation and ATP release [59,60]. López-Rodrı́guez et al. show that the NFAT5, which links the Rel NF-kB and NFAT families, regulates the production of specific cytokines in Tcells during osmotic stress [61]. In fish, the effect of hyperosmotic stressors on T-cell abundance/function has been poorly investigated.

However, acute and chronic salinity stressors stimulate the proliferation and antimicrobial functions of innate immune cells, as well as the release of pro-inflammatory cytokines, in several euryhaline and stenohaline fish species [13–16]. In Mozambique tilapia, overexpression of genres related to IgG and IgA following abrupt seawater transfer has also been described [47]. Therefore, it is likely that, similar to mammals, salinity stress stimulates lymphocytes abundance and functions in fish. 4.4. Signaling in innate immunity and inflammation In this study, most of the proteins involved in immunity belong to signaling pathways. In eukaryotes, inflammation is triggered when innate immune cells detect infection or tissue injury. Innate immune cells residing in tissues recognize cell damage and pathogen invasion with intracellular or surface expressed pattern recognition receptors (PRR). Activated PRRs then assemble to multi-subunit complexes that initiate signaling cascades and trigger the release of cytokines, chemokines, adhesion molecules, regulators of the extracellular matrix which in turn promote the inflammatory response and the recruitment of leukocytes. Lipid-based inflammatory mediators involved in the acute phase response after stress or injury such as complement3, platelet-activating factors and vasoactive amines were upregulated in brackish water.

Table 2 Top 10 regulators of differentially expressed proteins in striped catfish (Pangasianodon hypophthalmus, S.) following saline stress (0 to 10 ppt) derived from IPA. State: predicted activation state; p-value: to examine the significance of upstream regulator enrichment with a Fisher's exact test; Count: number of target molecules in the dataset. Upstream regulator

Molecule type

State

Z-score

p-Value

Count

PPARA HNF4A

Nuclear receptor Transcription regulator Endogenous mammalian

Activated Activated –

2079 2891 1317

6,55E−21 1,32E−17 2,93E−15

40 80 37

Transcription regulator Transcription regulator Transcription regulator Growth factor Nuclear receptor Transcription regulator Nuclear receptor

Activated Activated Activated Activated Activated Activated Activated

2862 2403 3793 2312 3580 3477 2158

7,22E−13 1,68E−12 3,87E−09 1,79E−08 6,67E−07 8,37E−07 1,32E−06

13 48 24 54 22 15 14

D-Glucose TP53 MYC NFE2L2 TGFB1 PPARG PPARGC1A PPARD

22

M. Schmitz et al. / Journal of Proteomics 167 (2017) 12–24

Fig. 7. Validation experiments. Dot blots and relative abundances of Cdc42, Raf1 and Vinculin in striped catfish (Pangasianodon hypophthalmus, S.) raised in fresh water and brackish water (0–10 ppt) over 20 days, infected or not. The values are presented as the mean abundance relative to fresh water values, non-infected ± SEM (n = 3).

These proteins may act as chemoattractant and trigger signaling in inflammation. Similarly, a kidney proteome analysis performed in seawater-acclimated Mozambique tilapia revealed upregulation of complement fragments and suggest that C3 acts as acute phase response proteins during hyperosmotic stresses [62]. Moreover, small GTPases including G proteinsβγ, Rho GTPases (especially Rac) as well as the related MAPK-p38 signaling cascade. In mammals, it has been shown that lipidbased inflammatory mediators activate G-proteins αβγ [63]. GTP-bound Ras, activated by a mechanism that is unclear, and G-proteins βγ dimer activates Rac signaling, leading to superoxide production by activation of the NADPH oxidase complex [63]. In addition, Rac regulates leukocytes' chemotaxis by actin cytoskeleton reorganization via the WAVE complex and the ARP2/3 complex [63]. Indeed, asymmetrical polymerization of F-actin to the leading edge of the cell facing the chemoattractant drive the protrusion force in the direction of the migration [63]. Although the activation of G-protein-coupled receptors has not yet been described during hyperosmotic stress, upregulation of these signaling proteins in our study may indicate the involvement of similar signal transduction pathways. On the other hand, Rac GTPases are known to have a key regulatory role on the control of MAPK p38 activity during osmotic stress via the osmosensing OSM-MEKK3 complex [64]. In mammals, it is well established that MAPK p38 is critical for long term exposure to prolonged hyperosmotic stressor by regulating gene transcription and post-translational modification of osmoprotective proteins [64]. In rat inner medullary renal cells, increase in extracellular osmolality stimulates the ERK1/MAPK-p38 pathway, and inhibition of p38-MAPK blocks the osmotic induction of the Natriuretic Peptide Receptor, which plays a major role in determining urinary sodium content [65]. Moreover, MAPK p38 mediates various cellular responses including apoptosis, cytoskeletal organisation, immune responses (e.g. inflammatory response and cytokine storm) and growth in response to diverse extracellular signals [66]. In fish, the involvement of small GTPases and MAPK p38 in osmoregulation and immunity has also been suggested. In gills of Japanese eel (Anguilla japonica), seawater acclimation induces upregulation of small GTPases signal transduction pathway [39]. In gill transcriptome of Mozambique tilapia, seawater transfer induced overexpression of MAPKKK7, the master regulator of the MAPKs signaling pathway [49]. In killifish, hypertonic shock induces phosphorylation of MAPK-p38, and the use of MAPK-p38 blockers inhibits Cl− secretion in chloride cells [46]. In Atlantic salmon, the use of MAPK-p38 blocker in head kidney leukocytes induced downregulation of key immune genes (e.g. IL-1, TNF-α, HSPs, CD83, COX2) involved in bacterial disease resistance [67]. 4.5. Upstream regulator analysis Top 10 regulators are summarized in Table 2. The most represented regulators belong to the family of Peroxisome Proliferator-activated

Receptor. In vertebrates, PPARs consist in three isotypes, namely PPARs α, β, γ and are members of the superfamily of nuclear hormone receptors which act as transcription factors mainly involved in the regulation of energy stores, carbohydrate and lipid metabolism. Moreover, PPARs are specialized receptors in the detection of fatty acid-derived signal molecules and are key candidates for being the receptors that transduce a fraction of the lipid-mediated inflammatory events [68]. In mammals, PPARs have been related to inflammatory diseases induced by chronic hypertonic exposure. For example, inactivation of PPARγ in corneal epithelial cells decreased the production of inflammatory cytokines induced after hypertonic stress [69]. Similarly, inhibition of PPARγ activity of kidney medullary cells inhibits COX2-mediated prostacyclin release and contribute to renal necrosis associated with the use of non-steroidal anti-inflammatory drugs [70]. In aquatic organisms, the role of PPARs in osmoregulation and immunity has been poorly investigated. In fish, the involvement of PPARs during acclimation to hyper-/hypo-osmotic conditions in relation with the presence of lipid-based molecules has been recently addressed but the underlying mechanisms remain to be investigated. In rabbitfish Siganus canaliculatus, the expression of PPARs α, β, γ increased at low ambient salinity and in fish fed with vegetable diets in comparison with fish fed with fish oil [71]. In Nile tilapia, increasing salinities induced downregulation of PPARα and associated RXR-receptor [72]. 4.6. Limitations of the study The aim of the infection was to stimulate the immune system in order to generate an induced immune response in fish. The absence of fish that were not injected by the bacteria on day 23 does not allow to disentangle the effect of infection from other residual factors such as the stress induced by the manipulation, the short water emersion, the injection alone or the eventual developmental differences between day 20 and day 23. Therefore, up- or downregulation of proteins on day 23 (infected fish) in comparison to days 0, 10 and 20 (non-infected fish) should be interpreted with caution. The second key limitation of the study resides in the use of heterologous databases in quantitative proteomics. Proteomics studies in nonmodel organisms are hampered by the lack of fully annotated, detailed, and high quality proteome, and thereby limit the value of a proteomic approach for protein identification and quantitation [73]. Protein identification is based on cross-species matching, and thus success is reliant on the rate of divergences of protein sequences and the taxonomic proximity from higher quality proteome [74]. Therefore, cross-species comparisons may weaken the confidence of the identification, and decrease the number of proteins considered in our study [74]. This explains, for instance, the relatively poor sequence coverage compared to similar analysis in model species. In addition, the requirements for precursor ion mass matching and product ion alignments induce that only conserved

M. Schmitz et al. / Journal of Proteomics 167 (2017) 12–24

regions of the protein can elicit matching [73,74]. Consequently, rapidly evolved proteins may be less abundant than slowly evolving proteins, which matched better with their phylogenetic neighbours [74]. Moreover, the functional classification was clarified by an analysis of the proteome using the DAVID and IPA resources. Because accurate pathway maps were only available for zebrafish, and the annotation of the Andromeda search engine was heterogeneous, we needed to unify the proteome annotation by homology with the zebrafish. Due to the lack of correspondences, this step constituted a significant unspecific filter, and the number of proteins considered in our analysis was limited. 5. Conclusion This study provides novel insight into the molecular mechanisms that regulate the response to salinity stress in fish kidneys. In addition, a multi-stress approach has been applied in order to study the cross-responses between osmoregulation and immunity. Our results show that moderate salinity stress may enhance the immune defences of fish. Transparency document The Transparency document associated with this article can be found, in the online version. Acknowledgements We are grateful to FRIA (Fonds de la Recherche dans l'Industrie et l'Agriculture, Wallonia-Brussels Federation) for providing a PhD grant to Mélodie Schmitz. Authors also thank the BELSPO funded project IAP Aquastress (P7/31). We also thank the associate editor and the two reviewers for their very instructive comments. The authors have declared no conflict of interest. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.jprot.2017.08.005. References [1] FAO – Food and Agriculture Organization of the United Nation, The state of World Fisheries and Aquaculture. Food and Agriculture Organization of the United Nations. Rome, 2014 243. [2] A.L. Nguyen, V.H. Dang, R.H. Bosma, J.A. Verreth, R. Leemans, S.S. De Silva, Simulated impacts of climate change on current farming locations of striped catfish (Pangasianodon hypophthalmus, Sauvage) in the Mekong Delta, Vietnam, Ambio 43 (2014) 1509–1568. [3] M. Schmitz, S. Baekelandt, L.K. Tran Thi, S. Mandiki, J. Douxfils, N. Thinh, D.T.T. Huong, Osmoregulatory and immunological status of the pond-raised striped catfish (Pangasianodon hypophthalmus S.) as affected by seasonal runoff and salinity changes in the Mekong Delta, Vietnam, Fish Physiol. Biochem. (2016)http://dx. doi.org/10.1007/s10695-016-0266-7. [4] Intergovernmental Panel on Climate Changes, The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, IPPC, 2013. [5] I. Boutet, C.L. Long Ky, F. Bonhomme, A transcriptomic approach of salinity response in the euryhaline teleost, Dicentrachus labrax, Gene 379 (2006) 40–50. [6] S. Kalujnaia, S. McWilliam, V.A. Zaguinaiko, A.L. Feilen, J. Nicholson, N. Hazon, C. Cutler, G. Cramb, Transcriptomic approach to the study of osmoregulation in the European eel Anguilla anguilla, Physiol. Genomics 31 (2007) 385–401. [7] M. Schmitz, S. Mandiki, J. Douxfils, T. Ziv, A. Admon, P. Kestemont, Synergic stress in striped catfish (Pangasianodon hypophthalmus, S.) exposed to chronic salinity and bacterial infection: effects on kidney protein expression profile, J. Proteome 142 (2016) 91–101. [8] D. Kültz, J. Li, A. Gardell, R. Sacchi, Quantitative molecular phenotyping of gill remodelling of gill in a cichlid fish responding to salinity stress, Mol. Cell. Proteomics 12 (2013) 3962–3975. [9] A. Kumari, P. Das, A.K. Parida, P.K. Agarwal, Proteomics, metabolomics and ionomics perspectives of salinity tolerance in halophytes, Front. Plant Sci. 6 (2015) 1–20. [10] J. Chen, H. Wu, Y. Shi, C. Li, M. Li, The effects of environmental salinity on trunk kidney proteome of juvenile ayu (Plecoglossus altivelis), Comp. Biochem. Physiol. D4 (2009) 263–267.

23

[11] J. Gabert, D. Kültz, Osmoprotective proteome adjustments in mouse kidney papilla, Biochim. Biophys. Acta 2011 (1814) 435–448. [12] T. Yada, T. Azuma, Y. Takaji, Stimulation of non-specific immune functions in seawater-acclimated rainbow trout Oncorhynchus mykiss, with reference to the role of growth hormone, Comp. Biochem. Physiol. B 129 (2001) 695–701. [13] A. Cuesta, R. Laiz-Carrion, M. Martin, J. Meseguer, J. Mancera, M. Esteban, The salinity influences humoral immune parameters of gilthead seabream (Sparus aurata), Fish Shellfish Immunol. 18 (2005) 255–261. [14] I.-F. Jiang, V.B. Khumar, D.N. Lee, C.F. Weng, Acute osmotic stress affects tilapia (Oreochromis mossambicus) innate immune responses, Fish Shellfish Immunol. 25 (2008) 41–846. [15] M. Schmitz, J. Douxfils, S. Mandiki, C. Morana, S. Baekelandt, P. Kestemont, Chronic hyperosmotic stress interferes with immune homeostasis in striped catfish (Pangasianodon hypophthalmus, S.) and leads to excessive inflammatory response during bacterial infection, Fish Shellfish Immunol. 55 (2016) 550–558. [16] J. Delamare-Deboutteville, D. Wood, A.C. Barnes, Response and function of cutaneous mucosal and serum antibodies in barramundi (Lates calcarifer) acclimated in seawater and freshwater, Fish Shellfish Immunol. 21 (2006) 92–101. [17] L. Schwartz, A. Guais, M. Pooya, M. Abolhassani, Is inflammation a consequence of extracellular hyperosmolarity? J. Inflamm. 6 (21) (2009). [18] C. Brocker, D. Thompson, V. Vasiliou, The role of hyperosmotic stress in immunity and disease, Biomol. Concepts 3 (4) (2012) 345–364. [19] K.P. Lai, J.W. Li, J. Gu, T.F. Chan, W.K. Tse, C.K. Wong, Transcriptomic analysis reveals specific osmoregulatory adaptive responses in gill mitochondria-rich cells and pavement cells of the Japanese eel, BMC Genomics 16 (2015) 1072. [20] S.C. Birrer, T. Reusch, O. Roth, Salinity changes impair pipefish immune defence, Fish Shellfish Immunol. 33 (2012) 1238–1248. [21] B. Enyedi, S. Kala, T. Nikolich-Zugich, P. Niethammer, Tissue damage detection by osmotic surveillance, Nat. Cell Biol. 15 (2013) 1123–1130. [22] P.R. Cutillas, B. Vanhaesebroeck, Quantitative profile of five murine core proteomes using label-free functional proteomics, Mol. Cell. Proteomics 6 (2007) 1560–1573. [23] R. Wang, A. Fabregat, D. Rios, D. Ovelleiro, J.M. Foster, R.G. Cote, J. Griss, A. Csordas, Y. Perez-Riverol, F. Reisinger, H. Hermjakob, L. Martens, J.A. Vizcaino, PRIDE inspector, a tool to visualize and validate MS proteomics data, Nat. Biotechnol. 30 (2012) 135–137. [24] J.-P. Hawke, A.C. McWhorther, A.G. Steigerwalt, D.J. Brenner, Edwardsiella ictaluri sp. nov., the causative agent of enteric septicaemia of catfish, Int. J. Syst. Bacteriol. 31 (1981) 396–400. [25] United States Department of Agriculture, Losses caused by enteric septicaemia of catfish (ESC) 2002–09, Animal and Health Inspection Service, Veterinary Services, 2011. [26] M. Schmitz, S. Baekelandt, S. Bequet, P. Kestemont, Chronic hyperosmotic stress inhibits renal toll-like receptors expression in striped catfish (Pangasianodon hypophthalmus, Sauvage) exposed or not to bacterial infection, Dev. Comp. Immunol. 73 (2017) 139–143. [27] S.D. Mc Cormick, Methods for nonlethal gill biopsy and measurement of NaKATPase activity, Can. J. Fish. Aquat. Sci. 50 (1993) 656–658. [28] J. Cox, N. Neuhauser, A. Michalski, R.A. Scheltema, J.V. Olsen, M. Mann, Andromeda — a peptide search engine integrated into the MaxQuant environment, J. Proteome Res. 10 (2011) 1794–1805. [29] J. Cox, M.Y. Hein, C.A. Luber, I. Paron, N. Nagaraj, M. Mann, Accurate proteome-wide label-free quantification by delayed normalization and maximal peptide ratio extraction, Termed MaxLFQ. Mol. Cell. Proteomics 13 (2014) 2513–2526. [30] G. Caraux, S. Pinloche, Permutmatrix, a graphical environment to arrange gene expression profiles in optimal linear order, Bioinformatics 21 (2005) 1280–1281. [31] D.W.H. Huang, B.T. Sherman, R.A. Lempicki, Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources, Nat. Protoc. 4 (2009) 44–57. [32] C. Welinder, L. Ekblad, Coomassie staining as loading control in western blot analysis, J. Proteome Res. 10 (2011) 1416–1419. [33] S.M. Eckert, T. Yada, B.S. Shepherd, M. Stetson, T. Hirano, E.G. Grau, Hormonal control of osmoregulation in the channel catfish Ictalurus punctatus, Gen. Comp. Endocrinol. 122 (2001) 270–286. [34] T.H.N. Phuc, D.T.T. Huong, P.B. Mather, D.A. Hurwood, Experimental assessment of the effects of sublethal salinities on growth performance and stress in cultured tra catfish (Pangasianodon hypophthalmus), Fish Physiol. Biochem. 40 (2014) 1839–1848. [35] V. Norton, K. Davis, Effect of abrupt change in the salinity of the environmental on plasma electrolytes, urine, volume, volume and electrolytic excretion in channel catfish, Ictalurus punctatus, Comp. Bioch. Physiol. A 56 (1977) 425–431. [36] D. Kultz, Physiological mechanisms used by fish to cope with salinity stress, J. Exp. Biol. 218 (2015) 1907–1914. [37] J.C. Fiess, A. Kunkel-Patterson, L. Mathias, L. Riley, P. Yancey, T. Hirano, G. Grau, Effects of environmental salinity and temperature on osmoregulatory ability, organic osmolytes, and plasma hormone profiles in the Mozambique tilapia (Oreochromis mossambicus), Comp. Biochem. Physiol. 146 (2) (2007) 252–264. [38] J.D. Rhodes, O. Breck, R. Waagbo, E. Bjerkas, J. Sanderson, N-acetylhistidine, a novel osmolyte in the lens of Atlantic salmon (Salmo salar L.), Am. J. Phys. 299 (2010) 1075–1081. [39] W. Tse, J. Sun, H. Zhang, K. Lai, J. Gu, J. Qiu, C. Wong, iTRAQ-based quantitative proteomic analysis reveals acute hypo-osmotic responsive proteins in the gills of Japanese eel (Anguilla japonica), J. Proteome 105 (2014) 133–143. [40] H. Aizawa, M. Katadae, M. Maruya, M. Sameshima, K. Murakami-Murofushi, I. Yahara, Hyperosmotic stress-induced reorganization of actin bundles in Dictyostelium cells over-expressing cofilin, Genes Cells 4 (1999) 311–324. [41] C. Di Ciano-Oliveira, A.C.P. Thirone, K. Szászi, A. Kapus, Osmotic stress and the cytoskeleton: the R(h)ole of Rho GTPases, Acta Physiol. 187 (2006) 257–272. [42] S. Méresse, K. Unsworth, A. Habermann, G. Griffiths, F. Fang, M. Lorenzo, S. Waterman, J. Gorven, D. Holden, Remodelling of the actin cytoskeleton is essential for replication of intravacuolar Salmonella, Cell. Microbiol. 3 (2001) 567–577.

24

M. Schmitz et al. / Journal of Proteomics 167 (2017) 12–24

[43] S. Mitra, D. Hanson, D. Schlaepfer, Focal adhesion kinase integrates signals to promote cell migration, Nat. Cell Biol. 6 (2005) 56–68. [44] J. Lee, N. Valkova, M. White, D. Kültz, Proteomic identification of processes and pathways characteristic of osmoregulatory tissues in spiny dogfish shark (Squalus acanthias), Comp. Biochem. Physiol. D 3 (2006) 328–343. [45] S.K. Quadri, M. Bhattacharjee, K. Parthasarathi, T. Tanita, J. Bhattacharya, Endothelial barrier strengthening by activation of focal adhesion kinase, J. Biol. Biochem. 278 (2003) 13342–13349. [46] W.S. Marshall, C.G. Ossum, E.K. Hoffmann, Hypotonic shock mediation by p38 MAPK, JNK, PKC, FAK, OSR1 and SPAK in osmosensing chloride secreting cells of killifish opercular epithelium, J. Exp. Biol. 208 (2005) 1063–1077. [47] D.F. Fiol, S.Y. Chan, D. Kültz, Identification and pathway analysis of immediate hyperosmotic stress responsive molecular mechanisms in tilapia (Oreochromis mossambicus) gill, Comp. Biochem. Physiol. D1 (2006) 244–356. [48] C.L. Ky, J. De Lorgeril, C. Hirtz, N. Sommerer, M. Rossignol, F. Bonhomme, Effect of environmental salinity on the proteome of the sea bass (Dicentrachus labrax L.), Anim. Genet. 38 (2007) 601–608. [49] I. Leguen, N. Odjo, Y. Le Bras, B. Luthringer, D. Baron, G. Monod, P. Prunet, Effect of seawater transfer on CYP1A gene expression in rainbow trout gills, Comp. Biochem. Physiol. A 156 (2010) 211–217. [50] A. Paredes, C. Plata, M. Rivera, E. Moreno, N. Vazquez, R. Munoz-Clares, S. Hebert, G. Gamba, Activity of the renal Na K 2Cl cotransporter is reduced by mutagenesis of N glycosylation sites: role for protein surface charge in Cl-transport, APS Journals 290 (2006) 1094–1102. [51] T. Nishi, M. Forgac, The vacuolar (H +)-ATPases—nature's most versatile proton pumps, Nat. Rev. Mol. Cell Biol. 3 (2002) 94–103. [52] O.V. Vieira, R.J. Boltelho, S. Grinstein, Phagosome maturation: aging gracefully, Biochem. J. 306 (2002) 689–704. [53] I. Jordens, M. Borja, M. Marsman, S. Dusseljee, L. Janssen, J. Calafat, H. Janssen, J. Neefjes, The Rab7 effector protein RILP controls lysosomal transport by inducing the recruitment of dynein dynactin motors, Curr. Biol. 11 (2001) 1680–1685. [54] B.M. Mullock, C.W. Smith, G. Ihrke, N.A. Bright, M. Lindsay, E.J. Parkinson, D.A. Brooks, R.G. Parton, D.E. James, J.P. Luzio, R.C. Piper, Syntaxin 7 is localized to late endosomal compartments, associates with vamp 8, and is required for late endosome-lysosome fusion, Mol. Biol. Cell 11 (2000) 3137–3159. [55] A. Marc, C. Quentel, A. Severe, P. Le Bail, G. Boeuf, Changes in endocrinal and nonspecific immunological parameters during seawater exposure of brown trout, J. Fish Biol. 46 (1995) 1065–1081. [56] O. Steele-Mortimer, The Salmonella containing vacuole — moving with the times, Curr. Opin. Microbiol. 11 (2008) 38–45. [57] M. Krogsgaard, Q.J. Li, C. Sumen, J. Huppa, M. Huse, M. Davis, Agonist/endogenous peptide-MHC heterodimers drive T cell activation and sensitivity, Nature 434 (2005) 238–243. [58] D. Huang, Y. Chen, M. Ruetsche, C. Phelps, G. Ghosh, X-ray crystal structure of protooncogene c-Rel bound to the CD28 response element of IL-2, Structure 9 (2001) 669–678. [59] W.G. Junger, F.C. Liu, W.H. Loomis, Hypertonic saline enhances cellular immune function, Circ. Shock. 42 (1994) 190–196.

[60] T. Woerhle, L. Yip, M. Manohar, Y. Sumi, Y. Yao, Y. Chen, W. Junger, Hypertonic stress regulates T cell function via pannexin-1 hemichannels and P2X receptors, J. Leukoc. Biol. 88 (2010) 1181–1189. [61] C. López-Rodríguez, J. Aramburu, L. Jin, A.S. Rakeman, M. Michino, A. Rao, Bridging the NFAT and NF-κB families: NFAT5 dimerization regulates cytokine gene transcription in response to osmotic stress, Immunity 15 (2005) 47–58. [62] V.B. Kumar, F. Jiang, H.H. Yang, C.F. Weng, Effects of serum on phagocytic activity and proteomic analysis of tilapia (Oreochromis mossambicus) serum after acute osmotic stress, Fish Shellfish Immunol. 26 (2009) 760–767. [63] K. Newton, V.M. Dixit, Signaling in innate immunity and inflammation, Cold Spring Harb. Perspect. 4 (2012) a006049. [64] M. Uhlik, A. Abell, N. Johnson, W. Sun, B. Cuevas, K. Lobel-Rice, E. Horne, M. Dell'Acqua, G. Johnson, Rac–MEKK3–MKK3 scaffolding for p38 MAPK activation during hyperosmotic shock, Nat. Cell Biol. 5 (2003) 1104–1110. [65] S. Chen, D.G. Gardner, Osmoregulation of natriuretic peptide receptor signaling in inner medullary collecting duct. A requirement for p38MAPK, J. Biol. Chem. 277 (8) (2002) 6037–6043. [66] K.J. Cowan, K.B. Storey, Mitogen-activated protein kinases: new signaling pathways functioning in cellular responses to environmental stress, J. Exp. Biol. 206 (2003) 1107–1115. [67] E. Hollen, S. Winterthun, Z. Du, A. Krovel, Inhibition of p38MAPK during cellular activation modulate gene expression of head kidney leukocytes isolated from Atlantic salmon (Salmo salar) fed soy bean oil or fish oil based diets, Fish Shellfish Immunol. 30 (2011) 397–405. [68] T. Varga, Z. Czimmerer, L. Nagy, PPARs are a unique set of fatty acid regulated transcription factors controlling both lipid metabolism and inflammation, Biochim. Biophys. Acta 2011 (1812) 1007–1022. [69] J. Yuan, F. Zhang, H. Yang, P. Reinach, I. Polikarpov, PPAR mediates hypertonicity-induced increases in proinflammatory release in human corneal epithelial cells, Invest. Ophthalmol. Vis. Sci. 51 (2010) 5895. [70] C. Hao, R. Redha, J. Morrow, M. Breyer, Peroxisome proliferator-activated receptor δ activation promotes cell survival following hypertonic stress, J. Biol. Chem. 277 (2002) 21341–21345. [71] C. You, D. Jiang, Q. Zhang, D. Xie, S. Wang, Y. Dong, Y. Li, Cloning and expression characterization of peroxisome proliferator-activated receptors (PPARs) with their agonists, dietary lipids, and ambient salinity in rabbitfish Siganus canaliculatus, Comp. Biochem. Physiol. B 206 (2017) 54–64. [72] Z. Xu, L. Gan, C. Xu, K. Chen, X. Wang, J. Qin, L. Chen, E. Li, Transcriptome profiling and molecular pathway analysis of genes in association with salinity adaptation in Nile tilapia Oreochromis niloticus, PLoS One 10 (2015), e0136506. [73] H.L. Bayram, A.J. Claydon, P.J. Brownridge, J.L. Hurst, A. Mileham, P. Stockley, R. Beynon, D.E. Hammond, Cross-species proteomics in analysis of mammalian sperm proteins, J. Proteome 135 (2016) 38–50. [74] G. Arnold, T. Frohlich, Dynamic proteome signatures in gametes, embryos and their maternal environment, Reprod. Fertil. Dev. 23 (2011) 81–93.