Molecular Characterization of a New Tetrodotoxin-Binding Protein, Peroxiredoxin-1, from Takifugu bimaculatus

Pufferfish are considered a culinary delicacy but require careful preparation to avoid ingestion of the highly toxic tetrodotoxin (TTX), which accumulates in certain tissues. In this study, the tissue distribution of peroxiredoxin-1 from Takifugu bimaculatus was investigated. The peroxiredoxin-1 protein was obtained by in vitro recombinant expression and purification. The recombinant protein had a strong ability to scavenge hydroxyl radicals, protect superhelical DNA plasmids from oxidative damage, and protect L929 cells from H2O2 toxicity through in vitro antioxidant activity. In addition, we verified its ability to bind to tetrodotoxin using surface plasmon resonance techniques. Further, recombinant proteins were found to facilitate the entry of tetrodotoxin into cells. Through these analyses, we identified, for the first time, peroxiredoxin-1 protein from Takifugu bimaculatus as a potential novel tetrodotoxin-binding protein. Our findings provide a basis for further exploration of the application of peroxiredoxin-1 protein and the molecular mechanisms of tetrodotoxin enrichment in pufferfish.


Introduction
Takifugu bimaculatus is a carnivorous groundfish that belongs to the Tetraodontiformes and is widely distributed from the South Yellow Sea to the South China Sea. Its delicious, chewy taste and high economic value make this fish popular in Southeast Asian markets [1,2]. T. bimaculatus can survive at lower latitudes and can adapt to higher temperatures and salinities as compared to the other three cultural pufferfish species in China; thus, farming production of T. bimaculatus has attracted much interest in recent years [3]. Accordingly, research on T. bimaculatus has mainly focused on genetic breeding and cultivation techniques. However, pufferfish contain highly toxic tetrodotoxin (TTX), which poses a significant risk to food safety.
TTX is a highly toxic non-protein natural neurotoxin that selectively blocks Na + channels, alters Na + membrane permeability, has a highly specific effect on nerve cell membranes, and possesses unique nerve blocking properties that block excitation transmission between nerves and muscles [4]. The median lethal dose (LD 50 ) for TTX given intraperitoneally to mice is 10.7 µg/kg [5]. To date, TTX has been identified in or on the body surface of many vertebrates and invertebrates, as well as in marine microorganisms [6]. Levels of TTX are significantly correlated with species, tissues, and seasons. The tissue distribution

Tissue Distribution of Tbprx-1 mRNA
A quantitative polymerase chain reaction (qPCR) method was used to investigate the distribution of prx-1 transcripts among the tissues of immature and mature T. bimaculatus ( Figure 1). In immature samples, the spleen had the highest expression; the gill, intestine, and kidney had moderate expression levels; and the liver, muscle, testis, heart, and skin had the lowest expression levels. In contrast, in mature samples, expression was highest in the gill and kidney; moderate in the heart and skin; and lowest in the liver, muscle, testis, spleen, and intestine.

Characterization and Phylogenetic Analysis of the Tbprx-1 Gene
Structure prediction showed that the Tbprx-1 protein sequence contains two highly conserved cysteines (Cys51 and Cys172; Figure 2a). Tbprx-1 encodes 198 amino acids with a molecular weight of 22.08 kDa and a predicted isoelectric point of 5.47. The signal peptide was not detected in the present gene sequence. The results of amino acid sequence comparison with other known species are shown in Figure 2b. To elucidate the molecular evolutionary relationships, a phylogenetic tree was constructed by comparing T. bimaculatus with the Tbprx-1 gene from different species. The tree ( Figure 2c) depicts evolutionary relationships based on the similarity of Tbprx-1 proteins from different species; the results show that the Tbprx-1 protein is 99.5% homologous to the prx-1 protein of T. flavidus, which is closer than that of other species. The results were analyzed using an independent-samples t-test. Different letters (a,b,c,d) indicate significant differences in the same group (p < 0.05); the same letter indicates no significant difference (p > 0.05).

Characterization and Phylogenetic Analysis of the Tbprx-1 Gene
Structure prediction showed that the Tbprx-1 protein sequence contains two highly conserved cysteines (Cys51 and Cys172; Figure 2a). Tbprx-1 encodes 198 amino acids with a molecular weight of 22.08 kDa and a predicted isoelectric point of 5.47. The signal peptide was not detected in the present gene sequence. The results of amino acid sequence comparison with other known species are shown in Figure 2b. To elucidate the molecular evolutionary relationships, a phylogenetic tree was constructed by comparing T. bimaculatus with the Tbprx-1 gene from different species. The tree ( Figure 2c) depicts evolutionary relationships based on the similarity of Tbprx-1 proteins from different species; the results show that the Tbprx-1 protein is 99.5% homologous to the prx-1 protein of T. flavidus, which is closer than that of other species.

Recombinant Expression Vector Construction and Recombinant Tbprx-1 Protein Purification
The expression vector used in the experiment contained a T7 lac promoter that induced the secretion of the target gene. The six histidine residues fused at the C-terminus enabled the target protein to be purified using 6× His-Tag affinity chromatography. Figure 3a shows the construction of the pET-28a(+)-prx-1 expression vector. The amplified Tbprx-1 target fragment was double digested with the restriction endonucleases NdeI and NotI to obtain a gene fragment that could be ligated with pET-28a(+) containing the same sticky ends. The digested pET-28a(+) plasmid was loaded into a 0.8% (w/v) agarose gel and subjected to electrophoresis in TAE to measure the cleavage efficiency (Figure 3b). ing. The recombinant E. coli BL21 strain without induction with IPTG shows no band. A fusion protein Tbprx-1 was highly soluble as shown in Figure 3c. Tbprx-1 protein wa purified by affinity chromatography using a HisTrap FF Crude column and AKTA Prim system (both GE Healthcare). Unspecific binding proteins and a little quantity of prxwere washed with 70 mM imidazole of the binding buffer. Thereafter pure prx-1 wa eluted with 500 mM imidazole, which appeared as a single band of 22.08 kDa (Figure 3c   Tbprx-1 was induced with 0.2 mM isopropyl-beta-D-thiogalactopyranoside (IPTG) and sonication. Fractions were assessed for recombinant Tbprx-1 purity by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Coomassie blue staining. The recombinant E. coli BL21 strain without induction with IPTG shows no band. A fusion protein Tbprx-1 was highly soluble as shown in Figure 3c. Tbprx-1 protein was purified by affinity chromatography using a HisTrap FF Crude column and AKTA Prime system (both GE Healthcare). Unspecific binding proteins and a little quantity of prx-1 were washed with 70 mM imidazole of the binding buffer. Thereafter pure prx-1 was eluted with 500 mM imidazole, which appeared as a single band of 22.08 kDa (Figure 3c).

Radical Scavenging Ability of rTbprx-1
rTbprx-1 had good hydroxyl free radical scavenging ability ( Figure 4). The 50% hydroxyl free radical clearance rate for rTbprx-1 was 0.12 mg/mL. When glutathione (GSH) was used as a positive control, the GSH concentration associated with a 50% hydroxyl free radical clearance rate was 0.95 mg/mL. These findings showed that the rTbprx-1 protein possessed a strong hydroxyl free radical scavenging ability.

In Vitro Supercoiled DNA Protection Activity of rTbprx-1
Next, we evaluated the ability of rTbprx-1 to protect in vitro supercoiled DNA. The supercoiled DNA PUC18 is 2686 bp in size in its normal form; following attack by hydroxyl radicals produced by the metal-catalyzed oxidation (MCO) system, the DNA is converted from the superhelical form to the nickel form. The degree of DNA damage is indicated by a pattern of bands with different mobility changes resulting from the electrophoretic breakdown of the two forms of DNA in the gel. In the absence of rTbprx-1, the DNA showed the most damage, whereas adding rTbprx-1 to the MCO system reduced the smear produced by hydroxyl radical-dependent DNA damage in a concentration-dependent manner. The strongest DNA protection was achieved when the protein concentration was 0.9 mg/mL ( Figure 5).

Radical Scavenging Ability of rTbprx-1
rTbprx-1 had good hydroxyl free radical scavenging ability ( Figure 4). The 50% hydroxyl free radical clearance rate for rTbprx-1 was 0.12 mg/mL. When glutathione (GSH) was used as a positive control, the GSH concentration associated with a 50% hydroxyl free radical clearance rate was 0.95 mg/mL. These findings showed that the rTbprx-1 protein possessed a strong hydroxyl free radical scavenging ability.

In Vitro Supercoiled DNA Protection Activity of rTbprx-1
Next, we evaluated the ability of rTbprx-1 to protect in vitro supercoiled DNA. The supercoiled DNA PUC18 is 2686 bp in size in its normal form; following attack by hydroxyl radicals produced by the metal-catalyzed oxidation (MCO) system, the DNA is converted from the superhelical form to the nickel form. The degree of DNA damage is indicated by a pattern of bands with different mobility changes resulting from the electrophoretic breakdown of the two forms of DNA in the gel. In the absence of rTbprx-1, the DNA showed the most damage, whereas adding rTbprx-1 to the MCO system reduced the smear produced by hydroxyl radical-dependent DNA damage in a concentration-dependent manner. The strongest DNA protection was achieved when the protein concentration was 0.9 mg/mL ( Figure 5).

Protective Effects of rTbprx-1 against H2O2-Induced Cellular Damage
To determine whether rTbprx-1 protein reduced cell viability, L929 cells were in bated with purified rTbprx-1 protein for 24 h, and 3-

Protective Effects of rTbprx-1 against H 2 O 2 -Induced Cellular Damage
To determine whether rTbprx-1 protein reduced cell viability, L929 cells were incubated with purified rTbprx-1 protein for 24 h, and 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxy phenyl)-2-(4-sulfonyl)-2 H-tetrazolium/phenazine methosulfate (MTS/PMS) assays were performed. As shown in Figure 6, the protein had no significant effect on cell viability in the concentration range of 0-112 µg/mL as compared with that in untreated cells (Figure 6a). We further evaluated the effects of rTbprx-1 on protection against H 2 O 2 -induced cell damage according to the method described by Qiao [26]. Our results showed that L929 cell viability increased as the protein concentration increased. The viability of cells treated with 56 and 112 µg/mL rTbprx-1 was significantly higher than that in untreated cells (Figure 6b). Thus, the rTbprx-1 protein protected L929 cells from peroxidative damage induced by

Protein Modeling and Molecular Docking
The homology modeling structure is shown in Figure 7a, and the Ramachandran plot of the best modeling structure is shown in Figure 7b. We selected 2z9s as the template with 87.94% sequence identity; 89.9% of amino acid residues were located in the favored region and 9.5% in allowed regions, with only 0.6% being outliers, indicating the feasibility of the modeled 3D structure. The docking (binding energy) score of TTX to the ab initio predicted protein was

Protein Modeling and Molecular Docking
The homology modeling structure is shown in Figure 7a, and the Ramachandran plot of the best modeling structure is shown in Figure 7b. We selected 2z9s as the template with 87.94% sequence identity; 89.9% of amino acid residues were located in the favored region and 9.5% in allowed regions, with only 0.6% being outliers, indicating the feasibility of the modeled 3D structure. The docking (binding energy) score of TTX to the ab initio predicted protein was −5.

SPR Measurement of the Binding Activity of TTX and rTbprx-1
Next, we used SPR to evaluate binding of rTbprx-1 with TTX. The results showed the response value for rTbprx-1 and TTX binding was 10,935.6 RU, and the equilibrium dissociation constant (KD) for rTbprx-1 and TTX was calculated using 4.10 × 10 −5 M kinetics. (Figure 8).

SPR Measurement of the Binding Activity of TTX and rTbprx-1
Next, we used SPR to evaluate binding of rTbprx-1 with TTX. The results showed the response value for rTbprx-1 and TTX binding was 10,935.6 RU, and the equilibrium dissociation constant (KD) for rTbprx-1 and TTX was calculated using 4.10 × 10 −5 M kinetics. (Figure 8).

Effects of rTbprx-1 on TTX Entry into Epithelioma Papulosum Cyprini (EPC) Cells
The effect of rTbprx-1 on the entry of TTX into EPC cells was examined using flow cytometry as shown in Figure 9. The EPC cells supplemented with rTbprx-1 showed significantly more TTX compared with the control group ( Figure 9).

Effects of rTbprx-1 on TTX Entry into Epithelioma Papulosum Cyprini (EPC) Cells
The effect of rTbprx-1 on the entry of TTX into EPC cells was examined using flow cytometry as shown in Figure 9. The EPC cells supplemented with rTbprx-1 showed significantly more TTX compared with the control group ( Figure 9).

Effects of rTbprx-1 on TTX Entry into Epithelioma Papulosum Cyprini (EPC) Cells
The effect of rTbprx-1 on the entry of TTX into EPC cells was examined using flow cytometry as shown in Figure 9. The EPC cells supplemented with rTbprx-1 showed significantly more TTX compared with the control group ( Figure 9).

Discussion
In this study, we investigated the expression characteristics of the Tbprx-1 gene and the active function of the protein in T. bimaculatus. In addition to its good antioxidant activity, a new function of prx-1, which can bind to TTX, was identified by molecular docking and SPR techniques.
The distribution of mRNA expression and protein content of Tbprx-1 was evaluated. As differential expression of genes in different tissues suggests a direct or indirect relationship between these genes' functions and these tissues, the relatively high expression level of immune-related tissues suggests that Tbprx-1 is involved in the immune system in this species. Wang et al. [27] showed that prx-1 mRNA is most highly expressed in the spleen, intestines, kidney, and gills of the golden pompano. This is consistent with our results and suggests that Tbprx-1 could play diverse roles in T. bimaculatus. The intestine is constantly exposed to ROS generated by the luminal contents (i.e., oxidized food, salivary oxidants) because of its vulnerability to oxidative stress [8]. Multiple sequence alignment results show that the protein was conserved in the evolution of various species. Prx-1 is a typical member of the 2-Cys prx subfamily and is found primarily in the cytoplasm, endoplasmic reticulum, mitochondria, and peroxisomes of cells [11]. Functional domain prediction shows that typical 2-cys prx interacts through beta strands at one edge of the monomer (B-type interface) to form the functional homodimer, and uses an A-type interface (similar to the dimeric interface in atypical 2-cys prx and prx5) at the opposite end of the monomer to form the stable decameric (pentamer of dimers) structure. Furthermore, the N-terminal Cys51 (peroxidatic cysteine = C P ) in prx-1 forms an intermolecular disulfide bond with another conserved C-terminal Cys172 (resolving cysteine = C R ) residue and is oxidized to cysteine-sulfonic acid (Cys52-SOH) in the thioredoxin-reducing protein system, which is equivalent to detoxifying peroxides [28]. Thus, prx-1 has been shown to have the highest enzymatic activity among all peroxisomal proteins [29]. These two conservative sites have been found in prx-1 subunits from other species, such as Lateolabrax japonicus [30], Trachinotus ovatus [27], and Anoplopoma fimbria [31]. Interestingly, based on the analysis of amino acid sequence differences of the Tbprx-1 in different species and docking with TTX, GLU193 in the prx-1 was unique in fish and distinct from other species (Figure 2b). This finding suggests that GLU193 may be the key amino acid mediating Tbprx-1 binding to the toxin.
In this study, we obtained recombinant prx-1 from T. bimaculatus, which showed a single band at 22 kDa. To obtain the soluble expressed recombinant product, an expression vector pET-28a(+) was used in this study, wherein the rTbprx-1 protein proved to be stable and a highly soluble partner to fuse with the target protein. As expected, the resulting rTbprx-1 was highly expressed in E. coli and the soluble component was obtained as shown in Figure 3. The result showed that the rTbprx-1 is highly soluble, most of the rTbprx-1 existed in the sonicated supernatant rather than in inclusion bodies. The recombinant product can be easily purified through immobilized metal ion affinity chromatography. Approximately 20 mg pure rTbprx-1 was obtained from 500 mL of culture medium and 97% purity was achieved by washing and elution with a gradient of imidazole. To assess the ROS-scavenging ability of rTbprx-1, we detected the ability of rTbprx-1 in scavenging hydroxyl free radicals and performed MCO DNA protection assays. The MCO system generates hydroxyl radicals that cause the sugar-phosphate backbone of supercoiled DNA to be cleaved into a nicked form, whereas the reducing agent, DTT, acts as an electron donor to facilitate DNA cleavage. Upon comparison with GSH (which is typically used as a scavenger positive control [32]), the purified rTbprx-1 was found to possess a strong ability to scavenge hydroxyl free radicals. The protection of superhelical DNA by rTbprx-1 is further indication of its antioxidant capacity, and our electropherogram shows slightly less protection compared with the previous research, probably owing to the different activity of rTbprx-1 in different species. The dragging bands that appear in the electropherogram may be caused by interference from proteins. In this evaluation of cell viability of the protein in L929 mouse fibroblast cells, the highest concentration has cytotoxic effects and inhibits cell growth. H 2 O 2 , a stable ROS, has inhibited the growth of L929 cells and results in their apoptosis [33]. Our results showed that rTbprx-1 has a protective effect against H 2 O 2 -induced cell damage. Similarly, Madusanka et al. [34] conducted WST-1 assays to examine FHM cell viability following oxidative cell damage. FHM cells were transfected with pcDNA3.1(+)/SsPrdx1 or pcDNA3.1 (+); compared with untransfected cells, pcDNA3.1 (+)/SsPrdx1 transfected cells displayed significantly higher cell viability than those transfected with pcDNA3.1 (+) or untransfected controls. H 2 O 2 can cause excessive ROS accumulation. During catalysis, Prxs conserves C P as the site of oxidation to sulfenic acid (C P -SOH) by H 2 O 2 , which then forms an inter-or intramolecular disulfide bond with a C R . Finally, the disulfides are reduced by cellular thiodisulfide exchange mechanisms to recover C P and C R [35].
Biomolecular interaction analysis using SPR technology was performed on a Biacore T200 system. According to the experimental data, Ka (1/M*S) is 35.74, Kd (1/S) is 0.001467, and the binding kinetics of rTbprx-1 and TTX is characterized by rapid binding/fast dissociation. The affinity shows the binding strength between the two molecules, which is represented by KD. The KD value of rTbprx-1 bind TTX is 4.10 × 10 −5 M, and the rTfVWD protein has been reported to bind TTX with a KD value of 2.92 × 10 −3 M, whose affinity is lower than rTbprx-1 [25]. The protein-small molecule binding affinity constant or KD value is usually between 10 −3 to 10 −6 M, indicating there is an interaction between rTbprx-1 and TTX. To further study the effect of rTbprx-1 on TTX transport, we co-EPC cells with rTbprx-1 and TTX, and found that rTbprx-1 can promote the entry of TTX into cells. Based on the kinetic diagram of SPR, our experimental sensorgram is of the "fast up and fast down" type (fast binding and fast dissociation), suggesting that the combination of TTX and rTbprx-1 is instantaneous, and dissociates under specific conditions, which is conducive to the transport of TTX. Further investigation is needed to elucidate how rTbprx-1 transports TTX in vivo and which transcription factors are involved in regulation.

Animals and Tissue Collection
T. bimaculatus was purchased from the Takifugu Breeding Farm in Fujian Province. Three immature and three mature puffer fish were collected and histologically observed. The immature ones were one-year-old fish (16.3 ± 0.90 cm in length and 108.0 ± 12.2 g in weight); the mature ones were two-year-old fish (19.1 ± 0.74 cm in length and 187.0 ± 12.3 g in weight). After induction of anesthesia with 0.01% Eugenol cement(Medical Equipment Factory of Shanghai Medical Instruments Co., Ltd., Shanghai, China), fish were dissected, and nine tissue types (kidney, intestine, spleen, liver, heart, gill, skin, muscle, and testis) were collected from each fish. Approximately 100 mg of each tissue was cut into small pieces (<0.5 cm), and the samples were immediately placed in RNAstore Reagent (Tiangen, Beijing, China) and stored at −20 • C overnight or at 4 • C for subsequent transcriptional expression studies.

qPCR Analysis of rTbprx-1 Expression in T. bimaculatus
Total RNA was extracted using an RNAprep Pure Tissue Kit (cat. no. DP431; Tiangen, Beijing, China). After determining the total RNA concentration using an Infinite200 Pro microplate reader (TECAN, Männedorf, Switzerland), 4 µg total RNA was extracted from each tissue, and cDNA was synthesized by reverse transcription using Evo M-MLV Reverse Transcription Reagent Premix (for qPCR; Accurate Biotechnology, Hunan, China) according to the manufacturer's instructions. The qPCR was conducted in a final volume of 20 µL, containing 10 ng total RNA, 10 pmol specific primers (prx-1 forward-[F]: 5 -GAGGTCATTGGCTGCTCCAT-3 and prx-1 reverse-[R]: 5 -CTTCCTTCAGGACGCCGTAG-3 ), and 10 µL FastStart Universal SYBR Green Master (Roche). The reaction conditions were as follows: pre-incubation at 95 • C for 600 s; followed by 45 cycles of denaturation at 95 • C for 10 s, annealing at 60 • C for 10 s, and extension at 72 • C for 10 s; melting (95 • C for 10 s, 60 • C for 10 s, and 97 • C for 1 s). Primers for 18S (q18S-F: 5 -TCCGATAACGAACGAGACTCCG-3 ; q18S-R: 5 -CATCTAAGGGCATCACAGACCTGT-3 ) were designed to amplify the internal reference gene. The qPCR data were calculated using the 2 −∆∆CT method. Melt-curve analysis was performed to analyze the specificity of PCR products.

Construction of the rTbprx-1 Vector
The Tbprx-1 gene was synthesized, and the tag sequence was added at the N-terminal end. The target gene was amplified, double digested, and ligated between NdeI and XhoI of the pET-28a(+) vector. The product was transformed into E. coli BL21 (DE3) pLysS (Vazyme, Nanjing, China), inoculated onto LB (Luria-Bertani) agar plates containing 250 µL/mL kanamycin, and incubated at 37 • C. Positive clones were obtained, and the plasmid was confirmed by DNA sequencing through Sangon Biotech Co., Ltd. (Shanghai, China).

Prokaryotic Expression and Purification of the pET28a-Tbprx-1 Vector
A single colony was grown in 10 mL LB medium (10 g/L tryptone, 5 g/L yeast extract, 10 g/L NaCl, pH 7.4) and cultured at 37 • C with shaking at 180 rpm until an OD 600 value of 2.0-6.0 was obtained. This bacterial broth was inoculated into the LB liquid medium containing kanamycin at a ratio of 1:50, and the cells were then incubated at 37 • C with shaking at 180 rpm until an OD 600 value of 0.5 was obtained. Next, 0.2 mM IPTG was added, and the cultures were incubated at 37 • C with shaking at 180 rpm for 5 h to induce expression. The culture was centrifuged at 6000× g and 4 • C for 15 min, washed twice with phosphate-buffered saline, resuspended, sonicated, and lysed until the liquid was brightly colored. The cultures were then centrifuged at 12,000× g and 4 • C for 10 min, and the supernatant was collected. SDS-PAGE was then used to analyze the expression level of the target protein. The supernatant was filtered through a 0.45-µm filter membrane before purification.
Purification was conducted using an AKTA Purifier 100 (GE Healthcare Life Sciences, Waukesha, WI, USA) and immobilized metal affinity chromatography. Three to five column volumes of MilliQ water were used to wash the prepacked column (HisTrapTM FF Crude 5 mL; GE Healthcare Life Sciences, Uppsala, Sweden), and 8-10 column volumes of equilibration buffer (20 mM phosphate buffer, 50 mM NaCl, 70 mM imidazole, pH 7.4) were used to equilibrate the column. The filtered supernatant was loaded onto the column at 1 mL/min, and the eluate was collected. Next, 5-8 column volumes of equilibration buffer were used to wash away unbound protein from the column, after which elution buffer (20 mM phosphate buffer, 50 mM NaCl, 500 mM imidazole, pH 7.4) was used to elute the protein. The elution peaks of the protein were collected, and small amounts were validated using SDS-PAGE by the PAGE Gel Fast Preparation kit (EpiZyme, Shanghai, China).

Radical Scavenging Ability of rTbprx-1
The hydroxyl radical scavenging activity of rTbprx-1 was detected using theHydroxyl Radical Assay Kit(A018; Nanjing Jiancheng Bioengineering Institute, Nanjing, China). Glutathione was used as the positive control experimental group; blank tubes, control tubes, and test tubes were set up to form three groups each for each reaction. After adding the reagent solution in a 37 • C water bath for 1 min, 2 mL of color developer prepared in advance was added and the mixture was protected from light for 10 min. The OD 550 value was measured using an enzyme marker. The hydroxyl radical inhibition rate was calculated according to the instructions.

MCO Protection Assay
rTbprx-1 can protect superhelical DNA from DTT/Fe 3+ /O 2 mediated damage, as assessed by in vitro MCO assays [36,37]. Briefly, the 50-µL reaction system contained 0.2 mMFecl3, 2 mM DTT, and different concentrations of rTbprx-1. Samples were incubated for 30 min at 37 • C, and 500 ng pUC18 superhelical DNA plasmid was then added for another 2 h. DNA degradation was evaluated using 8% agarose gel electrophoresis. Reactions with 0.5 mg/mL bovine serum albumin under the same conditions were used as controls.

Measurement of the Antioxidant Activity of rTbprx-1
The effects of rTbprx-1 on the proliferation of L929 mouse fibroblasts were detected using MTS/PMS assays. Briefly, L929 cells were treated with different concentrations of rTbprx-1 (0, 7, 14, 28, 56, 112 µg/mL) and then cultured at 37 • C in an atmosphere containing 5% CO 2 for 24 h. MTS/PMS cell proliferation working solution was added, and the cells were incubated in the dark for 0.75-2 h. Absorbance was measured at 490 nm using a microplate reader.
In the cell oxidative damage experiment, L929 cells were treated with 150 µM H 2 O 2 and different concentrations of rTbprx-1 protein. Untreated cells were used as a blank control. The cells were incubated at 37 • C in an atmosphere containing 5% CO 2 for 3 h before the MTS/PMS assay, and assays were performed as described above.

Protein Homologous Modeling and Molecular Docking
The SWISS-MODEL, 2z9s was selected as a template for homology modeling and protein structural plausibility assessment SAVES v5.0 server (University of California, Los Angeles, CA, USA)was used to assess the structure.
Molecular docking operations were performed using the AutoDock program, Version 4.02 (Scripps Research Insititute, La Jolla, CA, USA) in two steps: the first was to initially identify the tetrodotoxin-binding sites by blind docking; the second was to predict the mode of action of tetrodotoxin at each protein binding site by exact docking.

SPR Measurement of the Binding Activity of TTX and rTbprx-1
We diluted the rTbprx-1 to 80 µg/mL in immobilization buffer 10 mM sodium acetate (pH 4.0). The activator was prepared by mixing 400 mM EDC and 100 mM NHS (GE) immediately before injection. The CM5 sensor chip was activated for 420 s with the mixture at a flow rate of 10 µL/min. Then, 80 µg/mL of rTbprx-1 in immobilization buffer of 10 mM sodium acetate (pH 4.0) was injected into the sample channel (Fc4); the immobilization level was 10,935.6 RU. The chip was deactivated by 1 M ethanolamine hydrochloride (GE).
We diluted TTX with the same running buffer to six concentrations. TTX was injected to Fc3-Fc4 of the channel at a flow rate of 30 µL/min for an association phase of 120 s, followed by 300 s of dissociation. The association and dissociation processes were all performed in the running buffer. We then repeated the six cycles of interaction analysis according to analyte concentrations in ascending order. After each cycle of interaction analysis, the sensor chip surface regenerated completely; we used 10 mM Glycine-HCl as an injection buffer at a flow rate of 30 µL/min for 30 s to remove the ligand and any bound analyte.

Effects of rTbprx-1 on TTX Entry into EPC Cells
Epithelioma Papulosum Cyprini (EPC) cells were seeded in 6-well plates at a density of 3 × 10 5 cells and incubated in MEM containing 10% FBS and 1% penicillin-streptomycin at 28 • C with 5% CO 2 . rTbprx-1 (112 µg/mL) in culture medium was co-incubated with EPC cells for 24 h. Then, 0.07 µg of TTX labelled with cy5 fluorescent dye marker was added and the plates were wrapped with tin foil to prevent fluorescence burst. The wrapped plates were further incubated at 28 • C for 12 h. The cells were then washed twice with DPBS, scraped and collected into EP tubes, centrifuged, and the supernatant was discarded. Finally, the cells were resuspended in 500 µL DPBS, and the fluorescence signal of cy5-TTX in the cells was detected on a BD FACSCalibur flow cytometer (BD Biosciences). The parameters were adjusted with those of negative cells (cells without any treatment) to detect the number of cells in the FL4 channel for cy5-TTX added alone and the number of cells in the FL4 channel for rTbprx-1 and cy5-TTX added; data were analyzed using Flowjo v.10 sofware (FlowJo LLC, Ashland, OR, USA).

Statistical Analysis
One-way analysis of variance (ANOVA) followed by Tukey's tests and independentsamples t-tests were performed to compare differences using the SPSS software, version 24.0 (IBM, Armonk, NY, USA). All data are expressed as means ± standard errors, and results at p < 0.05 were considered significant.

Conclusions
We expressed a recombinant prx-1 protein from Takifugu bimaculatus, a peroxidase, and evaluated its antioxidant capacity. Our results show it was effective against ROS. Molecular docking and in vivo SPR techniques were used to evaluate the potential binding properties of TTX to rTbprx-1, which were found to facilitate the entry of tetrodotoxin into cells. We found that Tbprx-1 is a novel tetrodotoxin-binding protein for the first time, and it is highly expressed in immune organs.