Identification and Molecular Characterization of Superoxide Dismutases Isolated From A Scuticociliate Parasite: Physiological Role in Oxidative Stress

Philasterides dicentrarchi is a free-living microaerophilic scuticociliate that can become a facultative parasite and cause a serious parasitic disease in farmed fish. Both the free-living and parasitic forms of this scuticociliate are exposed to oxidative stress associated with environmental factors and the host immune system. The reactive oxygen species (ROS) generated by the host are neutralized by the ciliate by means of antioxidant defences. In this study we aimed to identify metalloenzymes with superoxide dismutase (SOD) activity capable of inactivating the superoxide anion (•O2−) generated during induction of oxidative stress. P. dicentrarchi possesses the three characteristic types of SOD isoenzymes in eukaryotes: copper/zinc-SOD, manganese-SOD and iron-SOD. The Cu/Zn-SOD isoenzymes comprise three types of homodimeric proteins (CSD1-3) of molecular weight (MW) 34–44 kDa and with very different AA sequences. All Cu/Zn-SODs are sensitive to NaCN, located in the cytosol and in the alveolar sacs, and one of them (CSD2) is extracellular. Mn- and Fe-SOD transcripts encode homodimeric proteins (MSD and FSD, respectively) in their native state: a) MSD (MW 50 kDa) is insensitive to H2O2 and NaN3 and is located in the mitochondria; and b) FSD (MW 60 kDa) is sensitive to H2O2, NaN3 and the polyphenol trans-resveratrol and is located extracellularly. Expression of SOD isoenzymes increases when •O2− is induced by ultraviolet (UV) irradiation, and the increase is proportional to the dose of energy applied, indicating that these enzymes are actively involved in cellular protection against oxidative stress.


Experimental animals.
Turbot of approximately 50 g body weight were obtained from a local fish farm. The fish were held in 250-L tanks with aerated recirculating sea water maintained at 14 °C. They were subjected to a photoperiod of 12 L:12D and fed daily with commercial pellets (Skretting, Burgos, Spain). The fish were acclimatized to laboratory conditions for 2 weeks before the start of the experiments.
Eight to 10-week-old Institute for Cancer Research (ICR) (Swiss) CD-1 mice, initially supplied by Charles River Laboratories (USA), were bred and maintained in the Central Animal Facility of the University of Santiago de Compostela (Spain). All experimental protocols carried out in the present study followed the European legislation (Directive 2010/63/EU) and the Spanish legislative requirements related to the use of animals for experimentation (RD 53/2013) and were approved by the Institutional Animal Care and Use Committee of the University of Santiago de Compostela (Spain).
Purification of SODs by anion exchange chromatography. Ciliates were collected by centrifugation at 700 g for 5 min and resuspended in saline phosphate buffer (PBS) containing 1x protease inhibitor cocktail (Sigma-Aldrich). The ciliates present in the solution were then lysed by ultrasonic treatment (W-250 sonifier, Branson Ultrasonic Corporation, USA) and centrifuged at 15000 g for 20 min at 4 °C 29 . The supernatant thus obtained was dialyzed against a start buffer containing 20 mM Tris-HCl pH 8.0, before being filtered (0.45 μm) Samples of 1 mL of lysed extract of the ciliate (SE) were subjected to anion exchange chromatography (AEC). For this purpose, an AEC HiTrapQ column and an automatic protocol were integrated into the Äktaprime plus system (GE Healthcare, Sweden), and the sample was eluted using a buffer containing 20 mM Tris-HCl pH 8.0 and 1.0 M NaCl. The eluted sample was collected in 2 mL fractions. Those fractions associated with peaks determined by absorbance at 280 nm were pooled, dialyzed against distilled water, lyophilized and stored at −20 °C until analysis by native polyacrylamide gel electrophoresis, as described in detail below.
Determination of SOD activity in native polyacrylamide gels. The SOD activity was determined on polyacrylamide gels (PAGE) following the method of Weydert and Cullen 30 . The ciliates were cultured at a concentration of 5 × 10 5 trophozoites/mL in 24-well culture plates (Corning, USA) and were maintained under conditions of normoxia, with or without treatment with inhibitors: H 2 O 2 , KCN, NaN 3 and trans resveratrol (RESV). After incubation for 30 min without or with the inhibitors (100 μM), the ciliate samples were collected by centrifugation at 700 g for 5 min and washed twice in incomplete L-15 medium (medium without bovine serum). The pellet containing the ciliates was then resuspended in a loading buffer containing 1.5 M Tris-HCl pH 6.8, 50% glycerol and 5% bromophenol blue, which lyses the ciliates by osmotic shock. In some experiments, lyophilized samples were separated by anion exchange chromatography and resuspended in loading buffer. The enzymatic activity of the samples was determined on native PAGE, formed by a 5% concentrating gel polyacrylamide in 1.5 M Tris-HCl buffer pH 6.8 and a 12.5% separating gel in Tris-HCl buffer pH 8.8. Gel polymerization was carried out by the addition of 0.04% ammonium persulphate (APS) and 0.0005% tetramethylethylenediamine (TEMED). After gel polymerization, a pre-electrophoresis step was carried out for 1 h at 20 mA in electrophoresis buffer containing 200 mM Tris-HCl pH 8.8 and 0.7 mM Na 2 EDTA at 4 °C to remove the remains of any APS, which can inactivate the enzyme. All of the initial buffer was then removed, and the samples were prepared in loading buffer and placed in the concentrator gel. Electrophoresis was finally carried out in an electrophoresis buffer containing 50 mM Tris-HCl pH 8.3, 1.5 mM Na 2 EDTA and 0.3 M glycine, for 1.5 h at 50 mA. The electrophoresed gels were washed twice in distilled water and stained with a solution of nitroblue tetrazolium chloride (NBT) (2.43 M NBT, 28 mM TEMED, 0.14 M riboflavin-5′-phosphate) for 20 min, with agitation at room temperature. The gels were then washed twice with distilled water and exposed, while still in the distilled water, to light for 12 h. The presence of enzymatic activity was observed by the appearance of destained bands on the violet-stained gel.
Transmission electron microscopy (TEM). For TEM analysis we followed the technique described by Paramá et al. 32 . Briefly, the cultured ciliates were collected by centrifugation at 1000 g for 5 min. Cells were fixed in 2.5% (v/v) glutaraldehyde in 0.1 M cacodylate buffer at pH 7.2. They were then washed several times with 0.1 M cacodylate buffer and post-fixed in 1% (w/v) OsO 4 , pre-stained in saturated aqueous uranyl acetate, dehydrated through a graded acetone series and embedded in Spurr's resin. Semi-thin sections were then cut with an ultratome (Leica Ultracut UCT, Leica microsystems, Germany) and stained with 1% toluidine blue for examination by light microscopy. Ultrathin sections were stained in alcoholic uranyl acetate and lead citrate and viewed in a Jeol JEM-1011 transmission electron microscope (Jeol, Japan) at an accelerating voltage of 100 kV.
Exposure of the ciliates to ultraviolet radiation. Ciliates (5 × 10 5 ciliates mL −1 ) were cultured in 12-well culture plates (Corning, USA) in a final volume of 2 mL of complete L-15 medium. The plates were inserted in a UV crosslinker (UVC500, Hoefer, USA) until reaching energies of 1, 2 and 3 Joules/cm 2 (J/cm 2 ). At the end of the exposure period, the viability of the ciliates was checked by observing their morphology and motility under an inverted microscope. The ciliates were centrifuged at 700 g for 5 min, the supernatant was removed, and the pellet was frozen at −20 °C until use.
Immunizations and serum collection. A group of five ICR (Swiss) CD-1 mice were immunized by ip injection with 200 μL per mouse of a solution of 200 μg of recombinant proteins (rCSD2,3, rFSD and rMSOD) in 1% chitosan hydrogel (CH), prepared according to the method of Barua and Das 33 . The mice were injected ip with the same dose of purified recombinant proteins in CH, 15 and 30 days after first immunization. Each mouse was bled via retrobulbar venous plexus 7 days after the final injection, and if the antibody level was satisfactory, the mouse was completely bled by decapitation. The blood was left to coagulate overnight at 4 °C, and the serum was separated by centrifugation at 2000 × g for 10 min, mixed 1:1 with glycerol and stored at −20 °C until use.

Sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) and Western-blot.
SDS-PAGE of the recombinant SODs (rCSD3, rFSD and rMSD) and peak 2 (P2) obtained by AEC (see Fig. 1) from P. dicentrarchi not exposed or exposed to different levels of UV radiation was performed on linear 12.5% polyacrylamide mini gels in a Mini-Protean ® Tetra cell system (BioRad, USA), as described by Iglesias et al. 34 .
Samples were included in a loading buffer with 62 mM Tris-HCl buffer, pH 6.8, containing 2% SDS and 10% glycerol. The gels consisted of 4% stacking gel and 12.5% linear separating gel. Samples were dissolved in 62 mM Tris-HCl buffer pH 6.8 with 2% SDS, 10% glycerol and 0.004% bromophenol blue, and they were then heated for 5 min in a boiling water bath. Electrophoresis was carried out at a constant 200 V in Tris-glycine electrode buffer (25 mM Tris, 190 mM glycine, pH 8.3).
For Western-blot analysis, samples separated by electrophoresis were immunoblotted at 15 V for 35 min to Immobilon-P transfer membranes (0.45 μm; Millipore, USA) in a trans-blot SD transfer cell (Bio-Rad, USA) with electrode buffer containing 48 mM Tris, 29 mM glycine, 0.037% SDS and 20% methanol, pH 9.2. Membranes were washed with Tris buffer saline (TBS; 50 mM Tris, 0.15 M NaCl, pH 7.4) and stained with Ponceau S to verify transfer, blocked for 2 h at room temperature with TBS containing 0.2% Tween 20 and 5% non-fat dry milk, before being washed in TBS and incubated for 1 h with sera of immunized mice anti-rCSD2-3, anti-rFSD and anti-rMSD (1:100 dilution). All samples were then incubated with peroxidase-conjugated rabbit anti-mouse Ig (Dakopatts; dilution 1:800) and finally with 0.003% H 2 O 2 and 0.06% 3,30-diaminobenzidine tetrahydrochloride containing 0.03% NiCl 2 (DAB/NiCl 2 , Sigma, USA). The reaction was stopped after approximately 3 min by exhaustive washing with TBS. Membranes were scanned, and the bands obtained were quantified on the basis of their signal intensity in the images by using Image J software (https://imagej.nih.gov/ij/).

Immunofluorescence and confocal microscopy. For immunolocalization of Pd-Cu/Zn-SOD3 and
Pd-Mn-SOD in the trophonts, an immunofluorescence assay was performed, as previously described 35 . Briefly, ciliates (5 × 10 6 ) were centrifuged at 700 × g for 5 min, washed twice with PBS pH 7.0 and fixed for 15 min in a solution of 4% formaldehyde in PBS at room temperature. The ciliates were then washed twice with PBS, resuspended in a solution containing 0.3% Triton X-100 in PBS for 3 min, washed twice with PBS and incubated with 1% BSA for 30 min. After this blocking step, the ciliates were washed in PBS and incubated at room temperature with stirring at 750 rpm for one hour with a 1:100 dilution in PBS of mouse sera anti-rCSD3 and anti-rMSD. The samples were washed 3 times with PBS, before fluorescein isothiocyanate (FITC) conjugated rabbit/anti-mouse Ig (DAKO, Denmark) (dilution, 1:1000) was added, and the samples were incubated for 1 h at room temperature and in darkness. After another three washes in PBS, the samples were mounted in PBS-glycerol (1:1) and visualized by confocal microscopy (Leica TCS-SP2, Leica Microsystems, Germany).
Phenazine methosulfonate (PMS)-nitroblue tetrazolium (NBT) assay. SOD activity was measured spectrophotometrically in the PMS-NBT assay 36 . In this assay, reduction of NBT occurs by •O 2 − generated by a mixture of nicotinamide adenine dinucleotide (NADH) and PMS at non-acidic pH. This was done by adding 100 μL of 1 mg/mL of the P2 of P. dicentrarchi, or 100 μL of P2 from ciliates treated with UV at 3 J/cm 2 , containing 50 μM of NBT and 78 μM NADH in 100 mM sodium phosphate buffer (PB) at pH 7.4, to each well of 96-well microplates (Corning, USA). The reaction was started by adding 100 μL of PMS (5 μM PMS in 100 mM PB pH 7.4). Assay mixtures were incubated at 25 °C for 60 min. The blue formazan resulting from the reduction of NBT by •O 2 − generated from autooxidation of PMS was measured spectrophotometrically at 560 nm. As the enzymatic activation proceeds, a reduction in the blue colour generated is produced. The enzymatic activity was quantified as the decrease in absorbance at 560 nm/min. Enzyme-linked immunosorbent assay (ELISA). The CSD2, CSD3, FSD and MFSD proteins in the culture medium and in the trophonts after exposure to ultraviolet radiation at 3 J/cm 2 were detected and quantified www.nature.com/scientificreports www.nature.com/scientificreports/ by ELISA. Briefly, 1 µg of P2 peak from AEC of ciliates exposed or not exposed to UV radiation or purified recombinant SODs (rCSD2, rCSD3, rFSD and rMSD) in 100 µl of carbonate-bicarbonate buffer pH 9.6 (coupling buffer), or 90 μL of incomplete L-15 medium from ciliates exposed or not exposed to UV radiation and 10 μL of coupling buffer 10×, was added to 96-well hydrophilic, protein-binding plates (ThermoFisher Scientific, USA) and incubated overnight at 4 °C. The plates were then washed 3 times with TBS (50 mM Tris, 0.15 M NaCl, pH 7.4), blocked for 1 h with TBS containing 0.2% Tween 20 (TBS-T 1 ), 5% non-fat dry milk, incubated for 15 min at 37 °C and at 750 rpm in a microplate shaker with 100 µl of a 1:100 dilution (in TBS-T 1 containing 1% non-fat dry milk) of immunized mice serum (anti-rCSD2, anti-rCSD3, anti-rFSD and anti-rMSD, serum), and washed 5 times with TBS containing 0.05% Tween 20. Bound mouse antibodies were detected with peroxidase-conjugated rabbit anti-mouse Ig (Dako) diluted 1:1000 in TBS-T 1 and incubated for 15 min at 37 °C and 750 rpm. The plates were then washed 5 times in TBS, and 100 µl of 0.04% o-phenylendiamine (OPD; Sigma) prepared in phosphate-citrate buffer, pH 5.0, containing 0.001% H 2 O 2 was added to each well. The reaction was stopped after 20 min, with 3 N H 2 SO 4 , and the optical density (OD) was measured at 492 nm in an ELISA reader (Titertek Multiscan, Flow Laboratories).
Reverse transcriptase-quantitative polymerase chain reaction (RT-qPCR). Total RNA of 10 6 trophozoites/mL of P. dicentrarchi not exposed or exposed to ultraviolet radiation was isolated with a NucleoSpin RNA isolation kit (Macherey-Nagel), following the manufacturer's instructions. After purification of the RNA, the quality, purity and concentration were measured in a NanoDrop ND-1000 Spectrophotometer (NanoDrop Technologies, USA). The reaction mixture (25 μL Bioinformatic and statistical analysis. InterPro software 38 was used for functional analysis of proteins and classification into different families predicting the domains and important sites. The Phobius 39 , SignalP 40 and Signal-3L ver. 2.0 41 programs were used to predict the topology of transmembrane and location of signal peptide cleavage sites in AA (AA) sequences. The MitoProt II-v1.101 program 42 was used to analyze the N-terminal region of the protein that might contain a mitochondrial signal sequence and its cleavage site. The MotiFinder tool of the Japanese GenomeNet network, accessible online at https://www.genome.jp/tools/motif/MOTIF.html was used to search for protein sequence motifs. The ProtParam tool was used to predict the physicochemical parameters for a given protein 43 . The SWISS-MODEL Protein server was used for modelling 40 . The bioinformatic tools LocTree3 42 and PredictProtein 43 were used to predict sub-cellular localization. The Clustal Omega multiple sequence alignment program was used to align the aa sequences of Pd-Cu/ZnSOD3 and Pd-Fe-SOD 44 . The Maximum Likelihood (ML) trees were constructed with a method based on a JTT model 45 with the Mega7 program 46 , and the reliability of internal branches was assessed using a nonparametric bootstrap method with 1,000 replicates. The Bayesian inference (BI) analysis was performed with MrBayes 3.2.6 47 .
The values shown in the text and figures are means ± SEM. One-way analysis of variance (ANOVA) was used for comparison of more than two samples, and the Tukey-Kramer test was used for pairwise comparisons. The Student's t-test was used for comparison of two samples. In both cases, differences were considered significant at P < 0.05.

Results
Isolation of the SODs and evaluation of enzymatic activities. For identification and purification of P. dicentrarchi SODs, a soluble extract of the ciliate (SE) was separated by anion exchange chromatography (AEC). The fractions thus obtained were analyzed by native electrophoresis, to detect enzymatic activity.
The results presented in Fig. 1A show the chromatographic profile obtained after separation of the SE fractions obtained by AEC by elution of the sample with the elution buffer. Three peaks of maximum absorption at 280 nm were observed. Analysis of the SOD activity of the three fractions, by non-denaturing PAGE, shows that the enzymatic activity occurs exclusively in the second elution peak (P2), detecting the appearance of three major bands with relative molecular weights of approximately 35, 50 and 60 kDa respectively (Fig. 1B).

Molecular characterization of the SOD enzymes of P. dicentrarchi.
After annotation of the transcripts obtained in an RNAseq experiment, we found that P. dicentrachi expressed three transcripts that encode proteins homologous to the Cu/Zn-SOD enzymes of ciliates ( Fig. 2A), a transcript that encodes a protein homologous to Mn/Fe-SOD and a transcript that encodes a protein homologous to Fe-SOD. In homology modelling of the structure of the proteins encoded by the transcripts, the DeepView program (Swiss Pbd-Viewer) predicted that the oligomeric state of the proteins is dimeric. Figure 2A shows the aa sequences of three transcripts obtained from an RNAseq assay. Analysis by the InterPro bioinformatics tool, designed for the analysis and classification of proteins, predicted that the aa sequences belong to the Cu/Zn-SOD superfamily. Regarding the biochemical parameters, the transcript corre-  (Fig. 2A); however, the global identity of the aa sequences is low (43-51%) (Fig. 2B). The gene ontology annotations of these sequences indicate that their molecular function is related to metal ion binding and biological process of oxidation-reduction and superoxide metabolic process. According to Signal-3L 2.0, a bioinformatic program predictor of signal peptides, CSD2, has a signal peptide between AA 1-40 . However, the Phobius bioinformatic predictor of signal peptides and transmembrane topology, predicted a signal peptide in the CSD2 protein between AA 1-20 . The topology of the CSD3 protein predicted by the Phobius bioinformatics tool indicates a cytoplasmic region between the AA 1-157 , a transmembrane region located between the AA 158-176 and a non-cytoplasmic region between the AA 177-196 .
We also found that P. dicentrarchi has two SODs dependent on Fe and Mn/Fe, as a result of the analysis of P. dicentrarchi sequences with the Blastx program and using the Tetrahymena thermophila (AA) database for the search. The Mn-dependent enzyme (Pd-Mn-SOD) has a 663 bp ORF (GenBank accession number MK348948) that encodes a protein (MSD) of 220 AA, of MW of about 25074.43 Da and with an estimated pI of 7.06. This protein contains an export signal to the mitochondria between AA 1-12 with a cleavage site al AA 13 . The Pd-Fe-SOD enzyme has a 750 bp ORF (GenBank accession MK348949) that encodes a protein (FSD) of 249 AA with an estimated MW of 29154.81 Da and an estimated pI of 5.79. The Phobius tool predicts for the FSD protein a signal peptide between the AA 1-21 . The homology modelling of the structure of the proteins encoded by the transcripts corresponding to the Pd-Mn-and Fe-SOD forms revealed the existence of an oligomeric state of the protein forming a homodimer.
The molecular evolution of Pd-Cu/Zn-SODs and Pd-Mn/Fe-SODs enzymes was studied within the framework of known SODs sequences of ciliate species and other species belonging to the most representative www.nature.com/scientificreports www.nature.com/scientificreports/ eukaryote groups (Fig. 3). The phylogenetic tree obtained by maximum likelihood (ML) and Bayesian inference (BI) methods indicate that Cu/Zn-SODs of P. dicentrarchi are closely related to Cu/Zn-SODs of ciliates and more separated from the SODs of other species of eukaryotes (Fig. 3A). More specifically, the Pd-Cu/Zn-SOD1-3 are closely related to a Cu/Zn-SOD of the ciliate Ichthyophthirius multifiliis, while the Pd-Cu/Zn-SOD2 is more closely related to a SOD of the scuticociliate Pseudocohnilembus persalinus (Fig. 3A). With respect to the Pd-Mn/Fe-SODs isoenzymes, the phylogenetic analysis confirms that both enzymes are related to the SODs of the Mn-SOD and Fe-SOD types of eukaryotes (Fig. 3B). On the other hand, the phylogenetic tree shows that Pd-Mn-SOD is grouped with Fe/Mn-SODs of ciliates and closely related to a Fe/Mn-SOD of the scuticociliate P. persalinus, while the Pd-Fe-SOD is grouped with a Fe/Mn-SOD from the ciliate Paramecium and a Fe-SOD from the plant Arabidopsis (Fig. 3B).

Cellular localization of SODs in P. dicentrarchi.
We used two approaches to determine the subcellular localization of the Pd-SOD isoenzymes: (1) bioinformatic prediction and (2) analysis by indirect immunofluorescence assay. In the first case, we used several bioinformatics tools such as PredictProtein and LocTree3 and, in the second case, we generated antibodies against CSD3 and MSD proteins (Fig. 4). Bioinformatic predictions indicate that the enzymes Pd-Cu/Zn-SOD1 and Pd-Cu/Zn-SOD2 would be located in the cytosol, while the Pd-Cu/ Zn-SOD3 isoenzyme would be located in structures related to plant chloroplasts or in the periplasmic space of bacteria. The subcellular localization of Pd-Mn-SOD is predicted to be in the mitochondria and the Pd-Fe-SOD in the cytoplasm. In the second case, in order to determine the cellular localization of the different Pd-SODs by immunofluorescence analysis, we generated antibodies against the recombinant proteins corresponding to Pd-Cu/Zn-SOD3 (rCDS3) and Pd-Mn-SOD (rMSD) (Fig. 4A). The size of the recombinant monomeric proteins corresponds to the original ciliated proteins from which they were derived: 21 kDa for the CSD3 protein monomer, 25 kDa for the MSD protein monomer and between 29-30 kDa for the FSD protein monomer (Fig. 4A). In a Western blot, the antibodies generated against the recombinant monomeric proteins recognize native proteins of the ciliate bands of twice the molecular weight of the monomers (Fig. 4B). After mice were immunized with the different recombinant proteins, the antibodies obtained were tested against P. dicentrarchi trophozoites by immunofluorescence analysis, which revealed that the anti-rCSD3 antibodies recognised the alveolar sacs (Fig. 4C). The antibodies generated by injection with the rMSD recombinant protein strongly recognised the mitochondria (Fig. 4D) aligned just below the plasma membrane of the trophozoite (Fig. 4E).
Expression of SODs in trophozoites of P. dicentrarchi exposed to ultraviolet radiation and oxidative stress. Ciliates exposed to ultraviolet light generated a higher level of SOD isoenzymes than non-exposed ciliates (Fig. 5). The protein levels of CSD2-3, MSD and FSD increased significantly after exposure to ultraviolet radiation with an energy of 3 J/cm 2 . In this experiment, we also analyzed the possible existence of extracellular SODs secreted by the ciliate. To investigate this phenomenon, we use an ELISA to determine the SOD levels in the culture medium of irradiated and non-irradiated ciliates. The assay showed an increase in the www.nature.com/scientificreports www.nature.com/scientificreports/ amount of Pd-Cu/Zn-SOD2, suggesting that this isoenzyme is released extracellularly, and also in the amount of Pd-Cu/Zn-SOD3 (Fig. 5A,B). On the other hand, the Pd-Mn-SOD isoenzyme was not detected extracellularly (Fig. 5C), and the Pd-Fe-SOD isoenzyme was detected in the culture medium, although the levels of this enzyme decreased after irradiation of trophozoites (Fig. 5D). Expression of SOD enzymes, at both the protein level (Fig. 6A) and transcriptomic level (Fig. 6B), was dependent on the dose of irradiation administered. Finally, exposure of the ciliates to oxidative stress through the chemical generation of •O 2 − in the culture medium caused a significant increase in SOD activity in the exposed trophozoites (Fig. 6C).

Discussion
Eukaryotes possess three types of SOD families characterized by the presence of metal cofactors (Mn 2+ , Fe 3+ , Cu 2+ or Zn 2+ ) in the active sites, location in different organelles and cellular compartments, and different sensitivities to cyanide, azide and hydrogen peroxide 48,49 . In the staining of antioxidant activity on native polyacrylamide gels, three bands of activity corresponding to three SOD isoenzymes were observed in a soluble extract of P. dicentrarchi purified by anion exchange chromatography. Identification of the enzymatic activity bands of the SOD isoenzymes was initially based on sensitivity to several inhibitors. Thus, the activity of the Cu/Zn-SODs was identified by the sensitivity to NaCN 50 , while the Fe-SOD activity was identified by sensitivity to H 2 O 2 and NaN 3 51,52 . The Mn-SOD activity was identified because it was not inhibited by NaCN or by H 2 O 2 53 . In this study, we also analyzed the effect of the phytoalexin trans-resveratrol (RESV), which has been shown to inhibit SOD in plants with an apparent K i of 10 μM 54 . The inhibitory capacity of RESV on the SOD activity of P. dicentrarchi was also confirmed in a previous study, in which we found that a concentration of 100 μM RESV caused inhibition of SOD activity 55 . In the present study, we showed that RESV partly inhibited the activity of the Pd-Mn-SOD isoenzyme and completely inhibited the activity of the Pd-Fe-SOD isoenzyme, while the Pd-Cu/Zn-SODs isoenzymes were insensitive to this polyphenol.
With the purpose of identifying the protein sequences of the Pd-SOD family of enzymes, we carried out a transcriptomic analysis with an RNAseq assay to locate homologous sequences and to identify the molecules involved. Although it was initially believed that protozoa lacked genes that encode the Cu/Zn-SOD isoenzyme 56,57 , several studies have shown its existence and activity in these unicellular organisms, including, for example, the amphizoic ciliates and amoebas 24,58 . In this study, we detected the presence in P. dicentrarchi of three transcripts with different sequences and with showing some similarity to Cu/Zn-SOD enzymes. The existence of several Cu/ZnSOD forms is frequently observed among ciliates. Thus, in Tetrahymena thermophila three Cu/Zn-SOD genes have been identified that encode enzymes of MW 17.9-21.4 kDa; in Euplotes focardii two genes that encode proteins of MW 16.8-20 kDa; and in Oxytrichia trifallax two genes that encode proteins of MW 17.5-17.6 kDa 24 . In addition, two types of Cu/Zn-SOD (Ec-Cu/Zn-SOD1 and EC-Cu/Zn-SOD2) identified in E. crassus express proteins of sizes between 17.3 and 19.9 kDa and pI of 4.98 and 6.65, respectively; the enzyme corresponding to type 1 has a signal peptide for extracellular activity 25 . In this study, the CSD2 protein encoded by the Pd-Cu/Zn-SOD2 transcript also has a signal peptide that indicates a relationship with a function in extracellular activity, while the CSD3 protein encoded by the Pd-Cu/ZnSOD3 transcript has a transmembrane region, indicating that it is attached to membranes. In the nematode Caenorhabditis elegans, the presence of Cu/ Zn-SOD isoforms with a consensus signal peptide at the N-terminus and similar to the extracellular-types of Cu/ Zn-SODs in mammals, has also been detected and associated with membranes with a presumed transmembrane domain at the C-terminal region generated through an alternative splicing process 59 . Other parasites such as the helminth Schistosoma mansoni also have extracellular Cu/Zn-SODs forms that contain a signal peptide and membrane-associated transmembrane regions 60 . Likewise, the presence of four variants in the isoelectric point (pI) has also been observed in adult forms of the same parasite, although in all cases the pI of the Cu/Zn-SODs was slightly lower than 7 60 . In the Pd-Cu/Zn-SODs transcripts, the pI of a transcript that encodes the protein CSD1 is below 7 and in two transcripts that encode proteins CSD2 and CSD3, the value of pI is greater than 7, indicating that pI is lower than 7 in cytoplasmic forms, while it is greater than 7 in extracellular forms or form associated with membranes. The existence of extracellular forms with pI > 7 has also been observed in plants, and the extracellular forms of the Cu/Zn-SODs have a higher pI than cytosolic forms 61,62 . Modelling of the proteins encoded by the three transcripts associated with Pd-Cu/Zn-SODs indicates that these are oligomeric proteins of the homodimer type, which may indicate that the MW of the active enzymes in their native form will be between 34 and 44 kDa. In most eukaryote organisms, Cu/Zn-SODs are presented as homodimer enzymes 49,63 . In the marine ciliate E. focardii and the amoeba Acanthamoeba castellani, the enzymes Cu/ZnSODs are homodimeric 26,58 . The low homology between the AA sequences of the Cu/Zn-SOD in P. dicentrarchi seems to indicate that these transcripts are derived from the expression of paralogous genes, as occurs in some archaebacteria 64 .
The Fe-and Mn-SOD enzymes are the oldest SOD group and have probably evolved from orthologous genes 20 . Fe-SOD enzymes are found in prokaryotes and in chloroplasts, while Mn-SODs occur both in prokaryotes and in the mitochondrial matrix of eukaryotes 26 . It has been suggested that iron may have been the first metal used as a cofactor associated with the active site of the first SOD due to its abundance at that time in the form of soluble Fe 2+65 . In protozoa, the Mn-SOD isoenzyme has been detected in Euglena gracilis 66 and in the ciliates Euplotes focardii and E. crassus 25,26 . Fe-SOD was originally considered a bacterial cytosolic enzyme; however, it . The level of recognition was evaluated against the SODs present in peak 2 (P2) purified by ion exchange chromatography (see Fig. 1), using the proteins recombinants (rCSD2, rCSD3, rMSD and rFSD) as response controls and a sample of L5 culture medium to determine the presence of extracellular forms of SOD. The asterisks indicate statistically significant differences (P < 0.01), relative to the non-irradiated controls (determined by Student t test). www.nature.com/scientificreports www.nature.com/scientificreports/ has also been detected in archaea, in plant chloroplasts, as well as in the cytosol, glycosomes and mitochondria of protists 67 . Fe-SODs are common in Protozoa, e.g. in the amitochondriate Entamoeba histolytica 68 , as well as in T. pyriformis 69 , Plasmodium falciparum 70 , Leishmania chagasi 71 , Trypanosoma cruzi 72 , Perkinsus marinus 73,74 and Trichomonas vaginalis 75 . The Fe-SOD form seems to predominate in anaerobic or microaerophilic organisms, while the Mn-SOD form predominates in aerobes 53 . The presence of an Fe-SOD form in P. dicentrarchi may be an adaptation to its microaerophilic nature and sensitivity to high concentrations of oxygen because it is a benthic organism 76 . The Pd-Mn-SOD isoenzyme possesses four Mn binding sites, as do other Mn-SODs of eukaryotes 77 . In bacteria and eukaryotes, Fe-and Mn-SOD exist in both the homodimeric and homotetrameric forms with 22 kDa subunits 67,77 ; however, in P. dicentrarchi, as in the symbiotic dinoflagellate protozoon Symbiodinium or the apicomplex P. falciparum, both Mn-SOD and Fe-SOD represent dimeric forms 78,79 . By contrast, the Fe-SOD isolated from the ciliate T. pyriformis is reported to be tetrameric 69 . In parasitic protists such as T. vaginalis, Fe-SOD is a dimeric protein that displays high structural similarity to Fe-SODs of prokaryotes, possibly indicating that its presence in eukaryotes may be due to an endosymbiotic process 80 . The CSD3 protein encoded by Pd-Cu/ Zn-SOD3 transcript and the FSD protein encoded by Pd-FeSOD transcript are both phylogenetically related to ciliated SODs and these proteins show greater identity with SODs of the scuticociliate P. persalinus 81 .
The bimetallic enzymes copper-and zinc-containing SODs are a family of isoenzymes found in both intracellular and extracellular locations. This family comprises ubiquitous enzymes that appear primarily in the cytosol but can also occur in the mitochondrial intermembrane space, the secretory pathway and even the nucleus 82 . In the present study, the bioinformatic prediction and the immunofluorescence findings indicate that the Pd-Cu/ Zn-SOD3 form is in the alveolar sacs, in a similar way that the Cu/Zn-SODs of bacteria occur in the periplasm 83 . The other two Pd-Cu/Zn-SODs appear to occur in the cytosol (Pd-Cu/Zn-SOD1) and extracellularly (Pd-Cu/ Zn-SOD2). In eukaryotes, the presence of cytosolic and extracellular Cu/Zn-SODs is very common 84,85 . The presence of a signal peptide in Pd-Cu/ZnSOD2 seems to indicate that this enzyme is extracellular. As •O 2 − is generally unable to cross the cell membrane, the substrate for this Cu/Zn-SOD in eukaryotes should be produced outside the cell 82 .
The identity analysis of the 220 AA transcript, determined using the Blastp tool, indicates that it possesses the highest identity with a SOD [Fe] protein in T. thermophila SB210 and with an Mn/Fe-SOD in the scuticociliate P. persalinus 81 . Fe-and Mn-SODs are found in a wide variety of species and may be located either in the cytosol Figure 6. (A) Expression of Cu/Zn-SOD3 (CSD3) in P. dicentrarchi trophonts exposed to different levels of UV radiation, as determined in a Western blot (WB) assay with a polyclonal antibody against the recombinant protein rCSD3 (anti-rCSD3) and proteins in peak 2 (P2) purified by ion exchange chromatography. The graph corresponds to the densitometric analysis of the WB (n = 3). A representative WB (not cropped) is shown in the figure. (B) Expression of Cu/Zn-SOD3 transcripts by ciliates exposed to different doses of UV radiation and quantified by RT-qPCR. (C) SOD activity in the P2 fraction from ciliates treated by UV irradiation (3 J/cm 2 ) and untreated ciliates (control), as determined by the PMS-NBT (phenazine methosulfonatenitroblue tetrazolium) assay. The enzymatic activity was quantified as the decrease in absorbance at 560 nm/min (−Δabsorbance/min). The mean values and standard error are represented in the bar graph, and the asterisks indicate statistically significant differences (*P < 0.05; **P < 0.01) relative to the non-irradiated control. www.nature.com/scientificreports www.nature.com/scientificreports/ or in the mitochondria or in both; however, in animals Mn-SOD is usually present in the mitochondria 57 . The Mn-SOD enzyme in eukaryotes is synthesized in the cytosol and exported post-translationally to the mitochondrial matrix where 90% of cellular O 2 is consumed 86 . The results of the immunofluorescence assay indicate that the protein encoded by the 220 AA transcript occurs in the mitochondria. In P. dicentrarchi, the protein encoded by the 220 AA transcripts also sends an export signal to the mitochondria and potentially has a homodimeric structure. Together with its predicted molecular size and its insensitivity to inhibition by H 2 O 2 and NaN 3 , the above findings seem to indicate that the protein is an Mn-SOD with mitochondrial localization. With respect to the transcript generated by the 249 AA protein, Blastp analysis of the NCBI indicates that the maximum identity of this sequence occurs with bacterial SOD sequences. However, use of this tool to perform the search in the database of Tetrahymena yielded maximum identity with an Fe-SOD of T. borealis and with an Mn/Fe-SOD of Oxitrichia. These results are also consistent with the prediction of the InterPro bioinformatics program, which indicates that this protein has a signal peptide, lacks transmembrane regions and possesses domains of the Mn/ Fe-SODs family. In addition, considering the molecular size of this enzyme in its homodimeric form, the inability of cyanide to inhibit its enzymatic activity and its sensitivity to azide together indicate that this enzyme contains iron as a metallic cofactor in the active site 80 .
In the marine environment where the free-living cilia live, various environmental changes can lead to the generation of a high level of oxidative stress 18,26 . During infection, the ciliates are also subjected to high levels of ROS generated by the cells of the host's innate immune system 87 . Among the factors that can intervene the following are environmentally most important in generating high levels of ROS: temperature, excess oxygen in water, solar UV radiation and the presence of various types of pollutants 18,88 . Likewise, when fish are cultured in open-circuit farms, the water is commonly oxygenated by the direct supply of O 2 or by aeration, and disinfection is carried out by UV radiation 89,90 . In order to survive, marine organisms use SOD enzymes that allow them to eliminate not only the endogenous •O 2 − produced during metabolic processes, including mitochondrial respiration, but also the exogenous •O 2 − present in environments with high level of oxidative stress 57 . In the present study, we observed that the levels of expression of all SODs increase significantly after exposure of trophonts to UV radiation. Likewise, we also observed that both Pd-Cu/Zn-SOD2 and Pd-Fe-SOD are excreted into the culture medium after UV irradiation. These findings indicate the existence of extracellular forms of Pd-Cu/Zn-SOD and of Pd-Fe-SOD. In both cases, although the cellular localization was initially predicted to be cytosolic, the existence of signal peptides in both isoenzymes seems to confirm that these proteins can be secreted and participate in the neutralization of extracellular •O 2 − generated by UV radiation. Mammals also possess a Cu/Zn-SOD and an extracellular Fe-SOD induced under oxidative stress 91 . Cytosolic and extracellular Cu/Zn-SODs have also been identified in several parasitic organisms [92][93][94] . In the marine ciliate E. focardii, one of the Cu/Zn-SODs is extracellular 26 . An excreted Fe-SOD has also been detected in the Tripanosomatid belonging to the genus Phytomonas, and which also has an immunogenic capacity 22 . Expression of the Pd-Cu/Zn-SODs, at both the protein and transcription levels increase in proportion to the dose of UV radiation administered, and the enzymatic activity is also significantly increased at the highest dose of non-lethal UV radiation.
In conclusion, P. dicentrarchi possesses the three characteristic types of SODs present in eukaryotes, i.e. Cu/ Zn-SOD, Mn-SOD and Fe-SOD, the functional forms of which are oligomeric enzymes of the homodimeric type. Three types of Pd-Cu/Zn-SOD enzymes are sensitive to NaCN: one is cytosolic, another is present in the alveolar sacs of the trophonts and a third is extracellular. All three have very different aa sequences. Another two other isoenzymes have also been identified: a mitochondrial Mn-SOD insensitive to H 2 O 2 and NaN 3 , and an extracellular cytosolic Fe-SOD sensitive to H 2 O 2 and NaN 3 . The activity of all SODs isoenzymes increases under conditions of oxidative stress induced by UV radiation. This study highlights the role of SODs as enzymes that protect the ciliate from the toxic action of •O 2 − generated both in the marine environment, during its free life phase, and in the host, during the parasitic phase. The extracellular forms of these enzymes, with very different aa sequences from the host, may yield potential diagnostic targets, as well as potential antigens, in order to produce vaccines in the future 95,96 .

Data Availability
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.