Psychrobacter sp. ANT206 (P.sp. ANT206) was collected from the Antarctic sea ice (GenBank accession number: MK968312) and kept frozen at −80°C until use. The Escherichia. coli (E.coli) BL21 strain and the pET-28a(+) plasmid were kept in our laboratory. The PCR reagents were obtained from TaKaRa Bio (Dalian, China). 2-Hydroxyethyl disulfide (HED), GR, nicotinamide adenine dinucleotide phosphate (NADPH), GSH, diethyldithiocarbamate (DDC), sodium azide (NaN₃), dithiothreitol (DTT), hydrogen peroxide (H₂O₂), tert-butyl hydroperoxide (tBHP) and cumene hydroperoxide (CHP) were purchased from Sigma-Aldrich (Shanghai, China). All the other reagents were acquired from Sinopharm (Beijing, China), and were of analytical grade or higher.

Based on the sequence and annotation of the genome of Psychrobacter. sp. ANT206, the grx3 gene was amplified by PCR using the forward primers: 5′-GACGGATCCATGACTGTATCTGTTA-3′, which contained a unique BamHI restriction site; and the reverse primer 5′-CATAAGCTTTGCTCGATAATCGCCC-3′ containing a unique HindIII restriction site. The PCR mix contained 0.2 μL (250 U) of Taq DNA polymerase, 1 μL of dNTPs (2.5 mM), 0.5 μM each primer, 25 ng of genomic DNA, 2 μL of 10 × PCR buffer (with Mg²⁺), and PCR-grade water to a final volume of 20 μL. PCR was performed with an initial denaturation at 94°C for 5 min, followed by 30 cycles at 94°C for 1 min, 57.1°C for 1 min and 72°C for 90 s, with a final extension at 72°C for 10 min. The phylogenetic tree was established using Mega 5.2 and sequences were obtained from NCBI. Multiple sequence alignments were performed using the BioEdit and ESPript programs.

The 3D structure of Ps-Grx3 was obtained based on its amino acid sequence¹ (Kelley et al., 2015), and the quality of the model was evaluated by Ramachandran plot generated by PROCHECK². The software PyMOL was employed for homology modelings and to analyze the 3D model of the target proteins. Structural adaptation features were predicted based on previously described methods (Tina et al., 2007). The GSH 3D structure was obtained from the Zinc database and Open Babel was used to change the sdf file to a pdb file format. GSH was chosen for studying molecular docking (protein-ligand) interactions. Interactions between Ps-Grx3 and its ligand and between EcGrx3 (PDB ID: 1FOV) and its ligand were analyzed by AutoDock Vina (Trott and Olson, 2010). AutoDock Vina was run several times to obtain various docking configurations. The parameters were used for Ps-Grx3 and EcGrx3 were as follows: center x = 7.280; y = 0.479, z = 15.903, size x = 1.147; y = −0.037; and z = 0.525.

Plasmids containing recombinant Ps-Grx3 and its mutant forms were expressed in E. coli and purified as previously described (Wang et al., 2019). The recombinant plasmids were transformed into competent E. coli BL21 for Ps-Grx3 expression. E. coli BL21 harboring the recombinant plasmids were inoculated into LB liquid medium supplemented with kanamycin (100 mg/L) and cultured at 37°C to an OD₆₀₀ of 0.5-0.6. Then, IPTG was added to a final concentration of 1.0 mM, and the mixture was incubated for 12 h at 25°C. Next, the induced cells were treated with 8 M urea overnight to generate inclusion bodies. Finally, the protein solutions were centrifuged, and the supernatants, which contained crude Ps-Grx3 extract or those of its mutant forms, were collected. The supernatants were then diluted 30 times in PBS (pH 8.0) and kept at 25°C for 2 h for protein refolding. Ps-Grx3 and its mutant forms were purified by Ni-NTA affinity chromatography. The samples were eluted with 50 mM imidazole buffer (20 mM Tris–HCl, 500 mM NaCl, pH 8.0) at a flow rate of 1 mL/min and stored at −20°C. The purification fold was obtained by the ratio of the specific activity of purified Ps-Grx3 to that of the crude enzyme. The molecular mass of Ps-Grx3 was estimated by SDS-PAGE (12.5% polyacrylamide gels).

The standard HED assay was used to measure enzyme activity as previously described (Zhang et al., 2017) with some modifications. The reaction mixture consisted of 100 mM Tris–HCl (pH 8.0), 0.6 μg/mL GR, 0.5 mM HED, 0.2 mM NADPH, 1.0 mM GSH, and 0.12 μg/mL purified protein in a total volume of 200 μL. The changes in absorbance were determined at 340 nm and 25°C. One unit of Ps-Grx3 activity was defined as the amount of Ps-Grx3 needed to oxidize 1 μmol of NADPH at 25°C every minute. Protein concentrations were determined by the Bradford method (Bradford, 1976).

The Cys13, Pro14, Tyr15, Cys16, Arg43, Thr55, Val56, and Pro57 residues in the active sites and binding sites were mutated to evaluate the effects of these residues on enzyme activity. To generate mutant forms of Ps-Grx3, the pET-28a(+) vector was used as a template for the site-directed mutagenesis of the Cys13, Pro14, Tyr15, Cys16, Arg43, Thr55, Val56, and Pro57 residues to alanine using the QuikChange Site-directed Mutagenesis Kit (Stratagene) following the manufacturer’s instructions. The primers designed for site-directed mutagenesis are shown in Table 1. The expression and purification of the mutant proteins were performed as described for wild-type Ps-Grx3.

For all Circular Dichroism (CD) measurements, five repeated scans were carried out to establish a baseline. Far-UV CD measurements were performed in a 0.1-cm quartz cuvette (Hellma, Germany) between 260 and 180 nm for 12 s every 1.0 nm using a bandwidth of 1.5 nm in a Jasco J-107 spectropolarimeter (Jasco, Japan). Ps-Grx3 and its mutant forms were separately diluted in a solution of 100 mM Tris–HCl (pH 8.0). Changes in secondary structure were presented as molar ellipticity [θ], according to the following equation: to the following equation:

in which θ is the ellipticity measured at a given wavelength λ (deg), n is the number of amino acids, l is the cell path length (cm), and c is the protein concentration (mol/L).

The optimum temperature for Ps-Grx3 activity was evaluated at different temperatures (0–60°C). To understand how temperatures influences Ps-Grx3 stability, the enzyme was incubated in a 100 mM Tris–HCl buffer (pH 8.0) solution for 1 h at 30, 40, and 50°C. Residual activity was measured as mentioned above.

The purified enzyme was incubated in 100 mM Tris–HCl buffer for 30 min to detect the effects of different ions and chemicals on its activity. After a 30-min pre-incubation, residual activity was compared to that of the enzyme incubated without any reagent. To measure the effects of chemicals such as DDC, NaN₃, DTT, H₂O₂, tBHP, and CHP, purified Ps-Grx3 was preincubated with 0.5 mM of each chemical for 30 min. Enzyme activity was measured as described above. Based on the results, we further explored the effect of different concentrations of H₂O₂ (0–1.0 mM) on the activity of purified Ps-Grx3.

The reaction mixture contained 0.6 μg/mL GR, 0.2 mM NADPH, and 1.0 mM GSH in 100 mM Tris–HCl buffer (pH 8.0), and various amounts of substrates (HED, bovine serum albumin BSA, dehydroascorbate DHA, S-sulfocysteine or L-cystine) at different concentrations (0.5 mM to 2.0 mM) in a total volume of 200 μL. Changes in absorbance at 340 nm were measured at 25 °C. The same amount of purified protein was used for all the assays. The K_m, V_m, k_cat, and k_cat/K_m values were obtained at 25°C based on previously described methods (Lonhienne et al., 2000). The above experiments were performed three times.

To evaluate the ability of Ps-Grx3 to protect supercoiled DNA, an Metal-Catalyzed Oxidation (MCO) DNA cleavage protection assay was performed as previously described (Revathy et al., 2012). The reaction mixture contained 30 μM FeCl₃, 3 mM DTT, and 6.25–50 μg/mL purified protein/BSA in a total volume of 100 μL, and the assay was performed at 25°C for 2.5 h. Next, 1 μg of pUC19 supercoiled DNA was added to the mixture, and incubation was continued for 1.5 or 2.5 h at 15°C and 25°C. Finally, the reaction mixture was purified and the sample was evaluated on a 1% agarose gel. The same reaction conditions were used throughout the assay.

The phylogenetic tree showed that Grx from Psychrobacter sp. ANT206 was indeed a class I Grx3 (Supplementary Figure 1). The open reading frame (ORF) of Ps-Grx3 (GenBank accession number: MW245063) contained 264 bp and encoded an 87-residue protein, while Grx from Taenia solium consisted of 315 bp and coded for a 105-amino acid protein (Nava et al., 2019). Sequence alignment (Figure 1) indicated that Cys13 and Cys16 were the catalytic sites and that Ps-Grx3 contained a redox-active site (CPYC motif), indicating that Ps-Grx3 is a dithiol Grx. The amino acid sequence of Grx3 from E. coli K-12 (EcGrx3, PDB ID: 1FOV) consists of 82 residues and contains a redox-active motif, Cys-Pro-Tyr-Cys, typical of the glutaredoxin family (Aslund et al., 1996). Additionally, Ps-Grx3 exhibited, respectively, 94.25 and 90.80% sequence similarity with Psychrobacter sp. KH172YL61 Grx3 and P.fozii Grx3.

The crystal structure of EcGrx3 from E. coli K-12 (PDB ID: 1FOV) was used as a template to elucidate the structure of Ps-Grx3 (Figure 2). The 3D model was verified by Ramachandran plot analysis, and the parameters indicated that the structural model of Ps-Grx3 was valid. The 3D structures of Ps-Grx3 and EcGrx3 were compared and showed good overlap (Figure 2).

To explore the cold-adaptive characteristics of Ps-Grx3, we compared it with its homolog EcGrx3. Compared with EcGrx3 (61), Ps-Grx3 has fewer hydrophobic interactions (50) (Table 2), which are known to be related to rigidity. Fewer hydrophobic bonds might lessen the rigidity of Ps-Grx3, thereby reducing its structural stability and greatly increasing its flexibility (Li et al., 2018). Further, Ps-Grx3 and EcGrx3 formed 70 and 1 versus 84 and 3 hydrogen bonds and salt bridges, respectively. The stability of cold-adaptive proteins is related to the number of both hydrogen bonds and salt bridges, and cold-adaptive enzymes usually have fewer of both than their mesophilic counterparts (Zheng et al., 2016), which was also consistent with the results of this study. Furthermore, as reported for other cold-adapted enzymes (Fu et al., 2013; Wang et al., 2020), the Arg/(Arg + Lys) ratio of Ps-Grx3 (0.33) was lower than that of EcGrx3 (0.45), which might be due to the increased conformational flexibility that would be expected to improve its catalytic ability under low temperatures (Feller and Gerday, 2003).

Molecular docking analysis is commonly used to predict ligand-protein binding modes (Hossain et al., 2016). Here, we used AutoDock Vina to compare the affinity of Ps-Grx3 and EcGrx3 (PDB ID: 1FOV) for GSH and the interaction between Ps-Grx3 or EcGrx3 and GSH. As shown in Table 3, the highest and lowest binding energy values for the interaction between GSH and Ps-Grx3 or EcGrx3 were determined to be −4.9 and −4.4 kcal/mol, respectively, for Ps-Grx3; and −5.0 and −4.4 kcal/mol, respectively, for EcGrx3. This indicated that Ps-Grx3 and EcGrx3 had similar affinities for GSH.

The recombinant plasmid pET-28a(+) was transfected into E. coli BL21. As shown in Figure 3, the SDS-PAGE profile of the purified Ps-Grx3 enzyme demonstrated a single major band (∼13.8 kDa), The purification factor for Ps-Grx3 was 6.14-fold and the recovery yield was 67.5% (Table 4). Additionally, the specific activity of Ps-Grx3 was 217.88 U/mg, while that of E. coli K-12 Grx3 was 410 U/mg (Nordstrand et al., 1999a).

The activity of Ps-Grx3 and that of its mutant forms was examined using HED as a substrate. Among all the mutants, the R43A, T55A, and P57A forms exhibited the largest reduction in enzyme activity (65, 80, and 70%, respectively), indicating that these residues might be involved in the catalytic activity of the enzyme (Figure 4). Importantly, the C13A, P14A, and C16A forms lost all enzyme activity, which demonstrated that these residues are likely to be crucial for catalysis. In E. coli, the C14S, C14A, and H15V class I Grx3 mutants exhibit reduced activity, while the Cys-14 to Ala mutation results in a complete loss of activity (Nordstrand et al., 1999b). Moreover, the E. coli Grx3 C15S mutant retains near-total activity, whereas the Grx3 C12S mutant shows no activity (Shi et al., 1999), Additionally, the Grx3 C65Y mutation does not affect the in vivo activity of the enzyme (Elgan et al., 2010). Importantly, the C13, P14, and C16 residues are conserved in Ps-Grx3, and Grx from Taenia solium has a conserved dithiol C34PYC37 active site (Nava et al., 2019). Furthermore, two catalytic sites—Cys25 and Cys28—are also present in Chlorella sorokiniana T-89 Grx (Chuang et al., 2014).

CD-spectroscopy was used to confirm that the secondary structure of the mutant forms of Ps-Grx3 had not changed and to ensure that the loss or weakening of the activity of the mutant enzyme was indeed due to point mutations in key amino acid residues rather than changes in secondary structure. The mutations used throughout this study do not appear to have affected the conformation of the respective proteins, as judged by CD spectrometry results (Figure 5), indicating that C13, P14, and C16 of Ps-Grx3 were active sites and played an important role in its activity.

The effect of temperature on Ps-Grx3 enzyme activity is shown in Figure 6A. Analysis of the temperature dependence of Ps-Grx3 showed that the enzyme exhibited optimal activity against HED at 25°C. The optimal temperature for Chlorella sorokiniana Grx is 50°C (Chuang et al., 2014). In the thermal inactivation assay, Ps-Grx3 retained almost 40% residual activity at incubations between 0 and 10°C, was highly stable from 0 to 30°C, and retained a greater than 90% residual activity after 90 min at 30°C (Figure 6B). After 30 min at 40°C, Ps-Grx3 lost approximately 60% of its activity, irreversible denaturation occurred after 30 min at 50°C, and its half-life was 18 min at 50°C. In comparison, Grx3 from Pseudoalteromonas sp. ANT178 lost almost all of its activity after incubation at 50°C for 30 min (Wang et al., 2014), while Chlorella sorokiniana Grx retained at least 80% of maximal activity when incubated at 50°C for 1 h (Chuang et al., 2014). This difference suggests that Ps-Grx3 is a thermostable enzyme, as observed for other cold-adaptive enzymes.

The results of the effects of different compounds on Ps-Grx3 activity are shown in Figure 7. In our study, Ps-Grx3 was deactivated by exposure to 5 mM concentrations of Sn²⁺ and Cr²⁺. In the presence of 5 mM Hg²⁺ and Zn²⁺, Ps-Grx3 retained less than 20% activity. Zn²⁺ has also been reported to inhibit the activity of Grx3 from Pseudoalteromonas sp. ANT178 (Wang et al., 2014). The addition of other metals, such as Na²⁺ and Mn²⁺, partially inhibited Ps-Grx3 activity, whereas the opposite occurred with the addition of urea and SDS. Notably, exposure to Fe³⁺ also increased the activity of Ps-Grx3.

Ps-Grx3 exhibited differential responses to different chemicals (Figure 8A). The activity of Ps-Grx3 was inhibited by NaN₃ and DTT, but Ps-Grx3 remained 66.8% active after treatment with 0.5 mM DTT. In contrast, Grx from Chlorella sorokiniana remained 94.0% active when treated with 1 mM DTT (Chuang et al., 2014). This demonstrated that Ps-Grx3 was more sensitive to DTT than Chlorella sorokiniana Grx. DTT is a reducing agent with the capacity to reduce Grx isoforms and proteins are active in their reduced state. Unexpectedly, the activity of Ps-Grx3 was reduced following preincubation with 0.5 mM DTT. This result suggested that DTT might affect either the activity of either GR or GSH (Rendon et al., 2004; Ates et al., 2009). Furthermore, Ps-Grx3 retained 15.3 and 9.8% activity following treatment with tBHP and CHP, respectively, which represented a significant reduction in activity when compared with H₂O₂ treatment. This result was in line with that of a previous study (Fernandes and Holmgren, 2004). Of all the chemicals tested, Ps-Grx3 was least sensitive to H₂O_2. Ps-Grx3 lost 10 and 25% of its activity in the presence of 0.2 mM and 0.6 mM H₂O₂, respectively, but then retained at least 40% activity even when the H₂O₂ concentration was raised to 1 mM (Figure 8B).

The kinetic parameters (K_m, V_m, and k_cat) of Ps-Grx3 are shown in Figure 9 and the kinetic parameter values for the various forms of Ps-Grx3 are summarized in Table 5. The K_m values for L-cystine, HED, and S-sulfocysteine were 0.52, 0.69, and 0.79 mM, respectively. The K_m values for HED and S-sulfocysteine were reported to be 1.68 and 1.77 mM for human Grx2 (Gladyshev et al., 2001). The K_m values of Ps-Grx3 for S-sulfocysteine and BSA were higher, whereas those for L-cystine, HED, and DHA were relatively lower. Ps-Grx3 has a relatively low K_m value for HED (0.69 mM), suggesting that Ps-Grx3 can efficiently reduce the disulfide bond. Additionally, the k_cat/K_m values for DHA and BSA were 68.40 and 25.55 1/s/mm However, the highest V_m and k_cat values at the same enzyme concentration were observed with S-sulfocysteine. The k_cat of Ps-Grx3 was different with the five substrates. Ps-Grx3 displayed the highest efficiency with S-sulfocysteine, followed by HED, and then DHA.

Using HED (0.5–2.0 mM) as the substrate, the kinetic parameters of Ps-Grx3 were measured at 25°C, and the results were compared with those of Grxs from different sources (Supplementary Table 1). The results showed that the apparent V_m at 25°C was 6.04 μmol/min/mg for Ps-Grx3 and 188.68 μmol/min/mg for Taiwanofungus camphorata Grx. The k_cat represents the maximum number of substrate molecules converted per second for a single catalytic site at a given enzyme concentration. The k_cat and physiological efficiency or specific constant (k_cat/K_m) of Ps-Grx3 were 44.57 1/s and 64.59 1/s/mM, respectively, under optimum conditions (25°C, pH 8.0). For Taenia solium Grx using HED as a substrate, the k_cat and k_cat/K_m values were reported to be 5.6 1/s and 7.8 1/s/mM, respectively. A lower K_m is suggestive of a higher affinity toward for a substrate. A K_m value of 0.1 mM has been reported for E. coli Grx using HED as substrate, which was lower than that for Ps-Grx3 (0.69 mM), indicating that Ps-Grx3 has a lower affinity for HED when compared with E. coli Grx.

The ability of Ps-Grx3 to protect DNA from nicking was assessed using the MCO DNA cleavage protection assay. This assay involves ROS-mediated DNA disruption, which results in a single-strand break in supercoiled plasmid DNA (Revathy et al., 2015). Here, the pUC19 plasmid was used to investigate the DNA-protective activity of Ps-Grx3 (Figure 10). When the MCO assay components were not all added (lanes 2 and 3), the supercoiled DNA of pUC19 was not damaged, as indicated by the equal ratios of supercoiled: nicked DNA in the control sample (lane 1). However, after the simultaneous addition of DTT and FeCl₃, the pUC19 DNA showed extensive shredding (lane 4). When BSA was added with the MCO reagents (lane 5), the supercoiled DNA was completely nicked, indicating that BSA had no DNA-protective ability. Conversely, as the concentration of Ps-Grx3 increased (lanes 9–12), the proportion of nicked DNA decreased. Interestingly, at the same, Ps-Grx3 concentration, but at a lower temperature (15°C) and with a shorter incubation time (1.5 h), Ps-Grx3 was also capable of reducing the percentage of nicked DNA (lanes 6 and 9). This indicated that supercoiled pUC19 DNA was sheared by the addition of MCO assay reagents, but that this effect could be reversed by the addition of Ps-Grx3, which reflected its antioxidant activity. The results further implied that the effectiveness of the antioxidant activity of Ps-Grx3 might be closely linked to temperature and duration of incubation, as previously observed for peroxidase (Godahewa et al., 2016). Combined, these results indicated that the newly identified Ps-Grx3 is capable of protecting supercoiled DNA from oxidation-induced damage at low temperatures.