Comprehensive characterization and molecular insights into the salt tolerance of a Cu, Zn-superoxide dismutase from an Indian Mangrove, Avicennia marina

Superoxide dismutases are important group of antioxidant metallozyme and play important role in ROS homeostasis in salinity stress. The present study reports the biochemical properties of a salt-tolerant Cu, Zn-superoxide from Avicennia marina (Am_SOD). Am_SOD was purified from the leaf and identified by mass-spectrometry. Recombinant Am_SOD cDNA was bacterially expressed as a homodimeric protein. Enzyme kinetics revealed a high substrate affinity and specific activity of Am_SOD as compared to many earlier reported SODs. An electronic transition in 360–400 nm spectra of Am_SOD is indicative of Cu2+-binding. Am_SOD activity was potentially inhibited by diethyldithiocarbamate and H2O2, a characteristic of Cu, Zn-SOD. Am_SOD exhibited conformational and functional stability at high NaCl concentration as well in alkaline pH. Introgression of Am_SOD in E. coli conferred tolerance to oxidative stress under highly saline condition. Am_SOD was moderately thermostable and retained functional activity at ~ 60 °C. In-silico analyses revealed 5 solvent-accessible N-terminal residues of Am_SOD that were less hydrophobic than those at similar positions of non-halophilic SODs. Substituting these 5 residues with non-halophilic counterparts resulted in > 50% reduction in salt-tolerance of Am_SOD. This indicates a cumulative role of these residues in maintaining low surface hydrophobicity of Am_SOD and consequently high salt tolerance. The molecular information on antioxidant activity and salt-tolerance of Am_SOD may have potential application in biotechnology research. To our knowledge, this is the first report on salt-tolerant SOD from mangrove.


Results
Avicennia marina showed the highest superoxide dismutase activity. The present study started with screening out the particular Avicennia species with maximum SOD activity. The free radical scavenging activity of 3 different species of Avicennia from the Indian mangrove forest (Figure 1a) was compared in terms of the SOD activity of the leaf. The comparative SOD activity in the crude extract (in unit per minute) prepared from each gram of leaf tissue of these 3 species is shown in Figure 1b in which A. marina displayed the highest activity among all the 3 species. For each species, leaf tissues were collected from 6 different populations grown at 6 different locations of the mangrove forest. The intra-species variation in SOD activity was very insignificant as evident from the standard deviation of the data that ruled out the possibilities of experimental error while performing the enzyme assay as well as the impact of variation in environmental factors during sample collection.
A 16 kDa protein of A. marina displayed SOD activity. Next, by employing a three-step purification strategy, the native SOD enzyme was purified from A. marina leaf into partial homogeneity. In step-1, proteins present in the pellet fraction after 60% ammonium sulfate cut were obtained and were subjected to strong anion exchange chromatography in step-2. Five eluted fractions obtained from step-2 ( Figure 2a) were screened by SOD enzyme assay and fraction 5 displayed the highest SOD activity (Figure 2b). In the SDS-PAGE profile of this fraction, a 16 kDa protein was found to have been enriched along with some other proteins ( Figure 2c). Hence, fraction 5 was re-fractionated in a gel filtration column in step-3. In gel-filtration, fraction 5 was separated into 3 sub-fractions (Fr 5A to 5C in Figure 2d). Fraction 5B was found to have the highest SOD activity (Figure 2e) and contain a 16 kDa protein with >90% purity in SDS-PAGE ( Figure 2f). This protein band was excised from the gel, trypsin-digested, and analyzed by LC-MS/MS. As shown in supplementary Table S1, a total of 6 unique peptides were identified from this 16 kDa protein and all of them showed a significant match with a Cu, Znsuperoxide dismutase of A. marina in the UniProt database (Acc. no. Q9AXH2). Together, these 6 peptides account for about 48% sequence coverage to the intact protein. This identified protein is designated as Am_SOD throughout the entire study.
Recombinant Am_SOD was homodimeric. The mass-spectrometry deduced sequence of the purified protein with SOD activity was used to identify the gene from A. marina genome (NCBI genome ID: 16351) 25 . tBLASTn search with Am_SOD amino acid sequence revealed a 768 bp long mRNA transcript (GeneBank Acc. AF328859.1). The transcript was found to contain a 459 bp long ORF coding for the full-length Am_SOD protein. The full-length mRNA transcript and the ORF were separately PCR-amplified from the first-strand cDNA prepared from the total RNA of A. marina leaf. The full-length Am_SOD gene (with exons and introns) was also PCR amplified from the genomic DNA. The 768 bp long mRNA transcript and the full-length gene were separately cloned in the pCR TM 2.1 vector by the TA-cloning method. The ORF was cloned in Nde1 and Xho1 sites of the pET22b+ vector under the control of the T7 promoter. Sequence analysis and comparison of these 3 cloned inserts revealed that the 2027 bp long gene consists of 7 exons with 6 introns. Upon splicing, a 768 bp long mRNA transcript is generated which consists of 459 bp long ORF with a 46 bp long 5′ UTR and a 263 bp long 3′ UTR. The domain architecture and nucleotide sequence of the full-length Am_SOD gene are shown in supplementary Fig. S1a, b. A positive clone with the Am_SOD cDNA insert positioned in an accurate reading frame (supplementary Fig. S2) was selected for recombinant expression. The expression of N-terminal 6xHis tagged recombinant Am_SOD was induced in E. coli cells with IPTG and the recombinant protein was found to be in a soluble form. The recombinant Am_SOD was then purified using Ni-NTA affinity column under native condition ( Figure 3a) followed by the second round of purification in size exclusion column to remove undesired aggregates and non-specific E. coli proteins. The yield of recombinant Am_SOD was ~8-10 mg L −1 of culture. The oligomerization status of the purified Am_SOD was checked in SDS-PAGE shown in Figure 3b. In non- Am_SOD displayed characteristic Cu, Zn-superoxide dismutase activity. Sequence analysis of Am_SOD revealed the presence of a conserved SOD catalytic domain with multiple copper and zinc ion binding sites. The spectral pattern of Am_SOD in the visible region (300-800 nm) showed the evidence of electronic transitions at a region between 380-400 nm indicative of Cu 2+ interaction with the imidazole ring of Histidine-62 ( Figure 3c). Therefore, the enzymatic activity of the purified Am_SOD was studied by performing an assay using the riboflavin-NBT system. The principle of this assay is based on the fact that illumination of riboflavin generates free superoxide radicals which can convert NBT into blue-colored formazan which is spectrophotometrically measured at 560 nm. However, in presence of SOD, these radicals are scavenged and hence, the NBT conversion is reduced. Therefore, the high the A 560 value the less the amount of SOD enzyme present in the system and vice versa. As shown in Figure 3d, Am_SOD followed a typical Michaelis-Menten kinetics with an increasing concentration of riboflavin as substrate. The kinetic data were then plotted in a double reciprocal Lineweaver-Burk plot shown in Figure 3e. The V max and K m of recombinant Am_SOD were obtained to be 1557.14 unit/mg and 0.15738 µM respectively in 50 mM Tris buffer at pH 7.5. In addition to spectrophotometric assay, an in-gel activity assay was also performed in native PAGE shown in Figure 3f in which the Am_SOD Figure 1. A. marina displayed highest SOD activity among 3 species. (a) Photograph of twigs with inflorescence of 3 species of Avicennia collected from Sundarban mangrove. Horizontal yellow bar represents 10 cm. (b) SOD activity assay from leaf extract of these 3 species. SOD activity is expressed as unit of enzyme present in each gram of leaf tissue converting the substrate into product in a minute plotted in y-axis. Each bar graph represents the mean of 6 biological replicates collected from 6 different locations (n = 6) and error bar as SD. www.nature.com/scientificreports/ appeared as a hyaline zone while the rest of the gel turned blue due to the oxidation of NBT. The enzymatic activity of Cu, Zn-SOD is specifically inhibited by diethyldithiocarbamate (DDC) and H 2 O 2 . In this study, the activity of Am_SOD was specifically inhibited in a dose-dependent manner by sodium diethyldithiocarbamate trihydrate, and the IC 50 was obtained at 1.5 mM ( Figure 3g). As compared to DDC, H 2 O 2 was found to have a less inhibitory effect on Am_SOD as the IC 50 value was obtained at 8 mM ( Figure 3h). A similar pattern of Am_SOD inhibition was observed in zymography assay performed with DDC and H 2 O 2 (Figure 3i, j). The activity of Am_SOD was fully inhibited by both inhibitors at high concentrations.
Am_SOD showed halotolerance. Having a mangrove origin, Am_SOD was expected to be a salt-tolerant protein. The salt tolerance was investigated and established through a couple of experiments. First, the tyrosine (Tyr)-fluorescence spectra of Am_SOD were investigated at various NaCl concentrations (Figure 4a). In absence of salt, the wavelength of maximum emission for Am_SOD was obtained at 310 nm, which is typical of a tyrosine residue. No significant change in the Tyr-fluorescence of Am_SOD was observed in presence of NaCl at a concentration as high as 700 mM. Next, to understand further the salt-tolerant feature of Am_SOD we performed the Tyr-fluorescence quenching experiment using acrylamide and potassium iodide (KI) as 2 complementary sets of water-soluble quenchers. Acrylamide is a neutral quencher and can enter the interior of a protein. On the other hand, iodide is a negatively charged and bulky quencher that can quench the fluorescence of the surface residues. Am_SOD has no tryptophan residue but only a single tyrosine residue in its sequence. Therefore, in this study, the quenching data were analyzed by Stern-Volmer plot considering a homogenous emission from a single tyrosine. The quenching constant of this single tyrosine is reported here as effective Stern-Volmer constant (K SV ) eff . The acrylamide and KI quenching data of Am_SOD under control and high NaCl stress are represented in Figure 4b, c respectively. The values of (K SV ) eff and f α (quenchable fraction) are displayed in the tables adjacent to each corresponding plot. Considering the presence of only one tyrosine in Am_SOD, 100% quenching of fluorophore was observed in both experiments. Hence, this tyrosine residue is presumably located on the surface of Am_SOD. In acrylamide and KI quenching, insignificant change in the (K SV ) eff of Am_SOD was observed both in absence of NaCl as well as in presence of 500 mM NaCl. The data indicated that there was a marginal conformational change in Am_SOD in presence of a high concentration of NaCl as compared to no salt control. The conformational behavior of Am_SOD in presence of NaCl was further investigated by Bis-ANS  www.nature.com/scientificreports/ fluorescence assay (Figure 4d) that exploits the surface hydrophobicity of a protein. Bis-ANS is a conformationsensitive hydrophobic probe with a low quantum yield. However, it becomes highly fluorescent when binds to the hydrophobic pockets exposed on the protein surface. Unlike salt-sensitive proteins where hydrophobic pockets get buried under salt stress, Am_SOD displayed a significant increase (~50%) in surface hydrophobicity. Next, the salt-induced aggregation pattern of Am_SOD was studied by a single light scattering experiment shown in Figure 5a-i. A previously reported salt-sensitive and allergenic profilin Sola m 1 (a gift from Dr. Swati Gupta Bhattacharya of Bose Institute, Kolkata, India) isolated from eggplant 26 was used as a control to compare www.nature.com/scientificreports/ the results. Am_SOD did not show any aggregation even in the presence of 500 mM NaCl as evident from very insignificant/no increase in the absorbance at 360 nm. On the contrary, Sola m 1 started forming aggregates in the presence of 400 mM NaCl (Figure 5a-ii). All the above experiments are focused on studying the salt-induced conformational and structural changes in Am_SOD. In addition to these, the impact of salt concentration on the catalytic activity of Am_SOD was investigated as shown in Figure 5b. Am_SOD exhibited catalytic activity in presence of a wide range of NaCl concentrations. Maximum activity was observed at 25 mM NaCl and a further increase in salt concentration resulted in a gradual decrease in the specific activity. However, the enzymatic activity of Am_SOD was not drastically altered (<25% reduction) in presence of NaCl as high as 250 mM as compared to no salt control. Altogether, it was found that the biological function of Am_SOD was not considerably affected by high salt stress.
Am_SOD displayed enzyme activity at alkaline pH. The effect of pH on recombinant Am_SOD activity was studied using buffer systems of 4 different pH values. As illustrated in Figure 5c, the SOD activity was almost diminished at acidic pH of 3.6. However, the protein displayed enzyme activity in mildly acidic pH and the pH optimum was obtained at pH 7, which is a physiological pH. Interestingly, considerable retention of enzyme activity of Am_SOD was observed at a strongly alkaline pH of 10.

Functional complementation of salt tolerant Am_SOD in E. coli. For functional identification of
Am_SOD gene, a genetic complementation test was performed in a double SOD deficient mutant strain (ΔsodA and ΔsodB) of E. coli named QC774. The cells were transformed with Am_SOD cloned in pET22b+ vector. For control, QC774 cells and wild type E. coli K12 strain transformed with pET22b+ vector without any insert were used. Transformed cells were first selected on LB-agar plates containing ampicilin. An individual transformed colony was then streaked on LB-agar plate supplemented with ampicilin, methyl viologen dichloride for induc-  www.nature.com/scientificreports/ ing oxidative stress, and 500 mM NaCl for inducing salt stress. For untreated control, LB-agar plate was used without NaCl but with ampicilin and methyl viologen. Protein expression was induced by adding IPTG in all the  www.nature.com/scientificreports/ plates. As shown in Fig. 5d, only QC774 cells harboring Am_SOD constructs were able to grow under oxidative stress as well as salinity stress. On contrary, K12 cells with functional sod genes were able to survive only under the oxidative stress but couldn't grow in presence of high NaCl concentration. QC774 cells harboring empty vector were unable to survive under oxidative as well as salinity stress. This observation suggests the salt-tolerant feature of Am_SOD in addition to its potential role in combating oxidative stress.
Am_SOD displayed a certain degree of heat tolerance. Deconvolution of CD spectra of Am_SOD ( Figure 6a) at 25 °C revealed a correctly folded protein with predominantly β-sheets as evident from the minimum obtained at 215 nm. Also, a characteristic shoulder at 222 nm indicated the presence of a certain degree of α-helices. In step-wise thermal scanning, Am_SOD did not exhibit temperature-dependent denaturation since an inconspicuous change in the CD signal was observed at 90 °C as compared to what was observed at 25 °C (Figure 6b). A melting curve of Am_SOD shown in Figure 6c represents the ratio between α-helical fraction and β-sheeted fractions present in this protein at various temperatures. No significant decline in this melting curve of Am_SOD was observed when the temperature was gradually raised from 25 to 90 °C indicating no heatinduced conformational change in the protein. For comparison, a previously reported heat-sensitive pectate lyase of sunflower designated as Hel a 6 (a gift from Dr. Nandini Ghosh of Vidyasagar University, West Bengal, India) was used as a control. Hel a 6 protein was reported to show reversible heat denaturation 27 . Hence, the Hel a 6 melting curve exhibited a sharp decline with increasing temperature (AS or ascending scan) and the native folds were gradually lost. However, Hel a 6 partially refolded from a fully denatured state when the CD-scanning temperature was set back to 25 °C. To substantiate this observation, the effect of temperature on the catalytic activity of Am_SOD was investigated as shown in Figure 6d. Unlike the conformation-dependent melting curve  www.nature.com/scientificreports/ in Figure 6c, the enzymatic activity of Am_SOD remarkably declined at temperatures as high as 70 °C and 80 °C. However, Am_SOD was able to retain up to 70% of its catalytic activity at 60 °C.

Reduction in surface hydrophobicity is linked to halotolerance of Am_SOD.
A rational mutagenesis approach was undertaken to understand the role of a few selected residues in conferring salt tolerance to Am_SOD. Previous studies have shown that increased salt tolerance of a halophilic protein is associated with a noticeable increase in surface-exposed charge residues (negatively charged in particular) and reduction in surface hydrophobicity [28][29][30] . In this study, a comparison of Am_SOD with 3 non-halophilic Cu, Zn-SODs (Pa_SOD from Potentilla, Nt_SOD from tobacco, and Sl_SOD from tomato) by multiple sequence alignment (Figure 7a) revealed the presence of 8 less-hydrophobic residues in the N-terminus as compared to more hydrophobic residues on the corresponding positions of non-halophilic SODs. However, no significant change in surfaceexposed charged residues was observed between Am_SOD and non-halophilic SODs. Hence, we anticipated the involvement of these 8 residues in the salt tolerance of Am_SOD. Out of 8, 5 residues were found to be sufficiently surface exposed on the tertiary structural model of Am_SOD (Figure 7b) and were estimated to have high SASA values as listed in supplementary Table S2. Residues of non-halophilic SODs corresponding to these 5 residues were also found to be solvent accessible. Each of these 5 residues on Am_SOD was found to have the lowest hydropathy index value (i.e. lowest hydrophobicity) as compared to the corresponding residues on 3 non-halophilic SODs (Figure 7c). Here, we decided to replace each of these 5 residues on Am_SOD with the residue having the highest hydropathy index value on the corresponding position among the 3 non-halophilic SODs. The strategy of amino acid substitution is illustrated in supplementary Table S2. In this way, 5 single-point mutants were generated by site-directed mutagenesis. A sixth mutant carrying all the 5 substitutions in the same protein was also generated by gene synthesis. The recombinant versions of all these 6 mutants were expressed in soluble forms and were found to remain in dimer as shown in non-reducing SDS-PAGE (data not shown). Now, the superoxide dismutase activity of these mutants was compared to that of the WT Am_SOD in gradually increasing NaCl concentrations. As shown in Fig 8A, the SOD activity of all the 6 mutants was nearly similar to that of the WT enzyme when assayed in presence of 25, and 100 mM NaCl. However, a significant reduction (p<0.05) in SOD activity of the 6 mutants was noticed when the NaCl concentration was increased up to 500 mM. Among the 6 mutants, the multiple-point mutant displayed maximum reduction (>50%) in SOD activity indicating a cumulative impact of these 5 substitutions on increasing the surface hydrophobicity and subsequently perturbing the halotolerance of the protein.

Discussion
The present study presents a comprehensive characterization of a novel SOD enzyme isolated from a mangrove species of Indian origin using biochemical and biophysical methods. Mangroves are adapted to survive in high salinity environments. The generation of free radicals in the form of reactive oxygen species is a major manifestation of salt stress. To combat this challenge, mangroves are equipped with strong antioxidant systems that can function in a highly saline microenvironment. SOD enzymes are crucial members of the enzymatic antioxidant system. In this study, a high SOD activity of A. marina among 3 different Avicennia species was found to be associated with a 16 kDa protein designated as Am_SOD. The purity level and yield of natural Am_SOD protein purified from A. marina leaf were found to be considerably low. Hence, the full-length gene coding for this protein was isolated and purified in recombinant form. The analysis of the Am_SOD sequence revealed the presence of a conserved domain along with 6 conserved histidine residues responsible for metal ion (Cu 2+ and Zn 2+ ) binding which are characteristic of a Cu, Zn-SOD. Any organelle-specific putative signal peptide was not found in Am_SOD and its sequence showed homology mostly with cytosolic SOD enzymes. Interestingly,   www.nature.com/scientificreports/ be a functional enzyme since it retained all the native folds as well as the catalytic activity. The kinetic data of Am_SOD represents a high substrate affinity and strong superoxide dismutation activity as compared to many previously reported Cu, Zn-SODs of eukaryotic origin. Such a robust activity of Am_SOD is thought to be the key for homeostasis of the exceptionally high level of ROS resulting from salinity stress and thereby protecting the cellular components from oxidative damage. To perform the biological activity, Am_SOD is thought to remain functional in a stressful microenvironment with extreme physiological conditions like high osmolarity and ionic strength. Here, we established the halotolerant feature of Am_SOD in terms of conformational stability and resistance to aggregation under high salt stress. The conformational stability of Am_SOD as observed in its tyrosine fluorescence quenching pattern was similar to the tryptophan fluorescence quenching reported for a halophytic rice protein PINO1 33 . In Bis-ANS spectrofluorometric assay, a considerable increase in surfaceexposed hydrophobic pockets in presence of high NaCl concentration was also noticed in another salt-tolerant protein DNA Pol-λ from Arabidopsis 34 . This structural stability of Am_SOD can be linked to the retention of its catalytic activity under highly saline conditions. The ability of Am_SOD to exert antioxidant activity under highly saline microenvironment was further confirmed by a functional complementation test where introgression of Am_SOD within a sod double mutant of E. coli conferred tolerance to salt as well as oxidative stress. In addition to salt tolerance, Am_SOD also displayed a certain degree of heat resistance. The CD spectra-based melting curve of Am_SOD indicates retention of > 85% of its native structural folds at 95 °C. However, in temperaturedependent enzyme assay, Am_SOD exhibited a sharp decline in functional activity at 70 °C and onwards. This can be interpreted as even a small fraction of heat-induced conformational change has somehow perturbed the catalytic domain of Am_SOD. Thermostable SOD enzymes are predominantly found in peroxisomes. Assuming cytosolic origin, Am_SOD is probably an exceptional non-peroxisomal SOD that is resistant to heat denaturation. Am_SOD was also found to well tolerate the alkaline pH, which is not very common among the Cu, Zn-SODs. Hence, Am_SOD is less resistant to pH-induced conformational changes and metal-ligand leaching. Similar to halophilic Am_SOD, some SOD enzymes tolerant to alkaline pH were reported from marine organisms living in a saline environment 35,36 . Altogether, Am_SOD can be claimed as a stress-tolerant enzyme with strong free radical scavenging properties. The remarkably high salt tolerance of Am_SOD intrigued us to investigate its molecular basis at the residue level. For this purpose, a combinatorial approach consisting of comparative in silico sequence analysis with non-halophilic SODs followed by a mutational study was undertaken. The non-halophilic SODs were selected based on the availability of atomic details of their crystal structures. Many previous reports on extremophilic enzymes claimed that enhanced salt tolerance of a protein is linked to increased accumulation of negatively charged residues (such as aspartate and glutamate) and a decrease in hydrophobic residues on the surface of the protein [28][29][30] . Such a surface pattern is supposed to facilitate increased hydration even in presence of high salt in the protein microenvironment. Here, we reported the role of 5 residues located in the N-terminal portion of Am_SOD that are critical for its halotolerance. These residues are typically located outside of the conserved catalytic domain and were relatively less hydrophobic as compared to their non-halophilic counterparts. This observation was further experimentally corroborated by mutational analysis. It was also noted that not a single residue but the cumulative effect of all the 5 residue substitutions resulted in a drastic fall in salt tolerance. Hence, the salt tolerance of Am_SOD can be attributed to a synergistic impact of 5 N-terminal residues that together brought about a decrease in hydrophobic surface area for molecular adaptation. Taken together, the present study presents detailed molecular information on a least characterized stress-tolerant SOD enzyme from an Indian mangrove plant. It also provides a deeper molecular insight at the residue level to understand its mechanism to withstand high salt concentration. It is tempting to speculate that such information will help in crop engineering with better performance in a stressful environment. Such an antioxidant enzyme with noticeably high-stress tolerance will also help to formulate anti-toxicity and anti-aging products of pharmaceutical and cosmetic importance respectively. Mass spectrometry. The desired band was gel-excised, trypsin-digested as described in 38  UV-Vis spectrophotometry. The absorbance spectra of 0.8 mg/ml of Am_SOD were taken at wavelength from 300 to 800 nm at 25 °C in a double beam Hitachi U-2900 spectrophotometer (Japan).

Methods
Enzyme kinetics. The specific activity and kinetic parameters (V max and K m ) of recombinant Am_SOD were determined by the riboflavin-NBT method as described in 39 . Each 200 µl reaction mixture consisting of 50 mM Tris-Cl pH 7.5, 9.9 mM L-Methionine, 0.57 µM NBT, 1 µg Am_SOD, 0.025% Triton-X, and serially increasing concentration of riboflavin (0-0.75 µM) was prepared. Enzyme blanks and non-illuminated sets were prepared for each riboflavin concentration. Absorbance was taken at 560 nm. The specific activity of Am_SOD for each riboflavin concentration was calculated by considering 1 unit of SOD enzyme equivalent to a 50% reduction in NBT conversion.

SOD inhibition assay.
The reaction mixtures were prepared as described in 'Enzyme kinetics' but with increasing concentrations of either sodium diethyldithiocarbamate trihydrate (0-3 mM) or H 2 O 2 (0-10 mM) for 30 min. The riboflavin concentration was kept constant at 1.17 µM and specific activity was calculated.
Zymography. Purified Am_SOD protein was run in 10% non-reducing native PAGE. The gel was incubated in 1.26 mM NBT with gentle shaking for 20 min in dark followed by riboflavin buffer (10 mM potassium phosphate pH 8, 126 µl TEMED, and 34 µM riboflavin) with continuous illumination.
SOD assay under various physicochemical parameters. The reaction mixtures were prepared as described in 'Enzyme kinetics' but either with buffers of various pH values or various NaCl concentrations or various temperatures. In each assay condition, the rest of all the physicochemical parameters were kept constant except only the variable one. Comparative enzyme assay with NaCl-treated mutants was performed following the same method for WT Am_SOD. The riboflavin concentration was kept constant at 1.17 µM and specific activity was calculated.
Static light scattering. 0.5 mg/ml of Am_SOD or the mutant was mixed with various concentrations of NaCl (0-500 mM) and the absorbance at 360 nm was recorded in a UV-Vis spectrophotometer starting from 20 to 180 min at 25 °C.

Functional complementation test. Escherichia coli strain QC774 was transformed with either pET22b-
Am_SOD construct or empty pET22b+. For control, WT E. coli K12 strain with functional sod genes was transformed with used. Cells were spread on LB-agar plates supplemented with 100 µg/ml ampicilin. An individual colony from each plate was streaked on LB-agar plate supplemented with 100 µg/ml ampicilin, 0.025 mM methyl viologen dichloride, 0.5 mM IPTG, and either 500 mM NaCl or without salt.
Circular dichroism spectrometry. CD spectra of 5 µM of either Am_SOD or Hel a 6 protein were recorded at 25 °C and 50 nm min -1 scan speed in Jasco J-810 spectropolarimeter (Jasco, Tokyo, Japan). The raw CD data was converted in molar ellipticity in CAPITO server 40 . In a step-scan, the CD spectra were recorded by gradually increasing the system's temperature from 20 to 90 °C at an interval of 10 °C. In a descending scan, the system was cooled down to 20 °C and the spectra were recorded once again.
Fluorescence spectroscopy. 0.05 mg/ml of Am_SOD in 25 mM Tris-Cl pH 7.8 containing 5% glycerol was separately incubated with 0-700 mM NaCl for 2 h. 2 ml of each sample was taken in a quartz cuvette (4 × 4 mm) and tyrosine autofluorescence was recorded in Hitachi F-7100 spectrofluorimeter (Japan www.nature.com/scientificreports/ wavelength was set at 276 nm and, the emission was scanned from 290 to 400 nm at 30 nm/min speed with 5 nm slit lengths. An average of 3 scans was taken and corrected for control buffer spectra. The maximum emission wavelength was determined by the instrument software with an in-built derivative analysis. Fluorescence quenching assay. Am_SOD (0.05 mg/ml) was incubated either with or without 500 mM NaCl for 3 h. Excitation was set at 276 nm. Emission of each sample was scanned at 310 nm, first without quencher, and then freshly prepared 5 M of either KI or acrylamide was added in 2 µl increment 10 times. After each addition, the solution was gently pipetted and left for 2 min to attain equilibrium. Quencher concentrations were corrected for 'dilution effect' . Correction of 'inner filter effect' was done using Eq. (1).
F and F corr represent the uncorrected and corrected fluorescence respectively. A ex and A em indicate the absorbance at excitation and emission wavelengths, respectively. The quenching data were analyzed according to the modified Stern-Volmer Eq. (2), where F is the difference between F0 (I 304 without quencher) and F (I 304 with quencher); [Q] indicates molar concentration of quencher; fα is accessible fraction of Tyrosine; effective Stern-Volmer quenching constants (K SV ) eff is equal to fα. K SV values were obtained from the slope and intercept of the linear plot.
Bis-ANS fluorescence assay. NaCl treated or untreated Am_SOD (0.02 mg/ml) was taken in a 3 ml quartz cuvette. A freshly prepared aqueous solution of 300 µM Bis-ANS was added in a 2 µl increment 10 times. After each addition, the solution was gently pipetted and left for 2 min to attain equilibrium. Emission and excitation were set at 490 nm and 390 nm respectively.
Bioinformatics studies. tBLASTn against NCBInr and nBLAST against the A. marina genome were performed to identify the transcript and the full-length gene respectively. SOD sequences of Potentilla atrosanguinea (UniProt, B2CP37), Solanum lycopersicum (UniProt, Q43779), and Nicotiana tabacum (UniProt, A0A1S3ZTX1) were retrieved. Multiple sequence alignments were done in ClastalOmega server 41 . Homology modeling of Am_SOD was performed in SWISS-MODEL server 42 using PDB:2Q2L 43 as template followed by stereochemical quality checking in PROCHECK server 44 . The hydropathy index values of selected amino acids were recorded from 45 . The SASA value of each residue was calculated in GETAREA server 46 .
Generation of mutants. Mutant constructs in pET22b+ vector were generated by outsourcing from Bio-Bharati LifeScience Pvt. Ltd. (Kolkata, India) as illustrated in supplementary Table S2 and sequenced from Xcelris TM Genomics Labs Ltd., India. The mutant proteins were purified following the same method described for wild-type Am_SOD.
Statistical analysis. Comparison of SOD activity was performed by students t-test in GraphPad prism software V6.1 and significance value was set as p < 0.05.