The catalase-peroxidase sequences were obtained from PeroxiBase¹ (Passardi et al., 2007), where they have been verified and annotated. Multiple sequence alignment was performed using Clustal Omega², while structural alignment was conducted with the ESPript program suite³. The phylogenetic tree was generated using MEGA7, employing the maximum likelihood method as described by Kumar et al. (2016) and the Jones-Taylor-Thornton (JTT) model, with 1,000 bootstrap replicates (Jones et al., 1992). The structure of StKatG1 and StKatG2 was predicted using the machine learning algorithm AlphaFold3 (v3.0)⁴. PyMOL software⁵ was utilized for the structural alignment of the homology model with the AlphaFold3 structure, as well as for calculating and visualizing root mean square deviations.

The Salinicola tamaricis F01 strain was cultured on LB5 medium (peptone 10 g/L, yeast powder 5 g/L, NaCl 50 g/L, agar 15 g/L) at 30°C for 48 h. Genomic DNA of strain F01 was isolated using the Qiagen genomic DNA isolation kit. The stkatG2 gene was amplified from the genomic DNA using the primers 5’-GGAATTCCATATGATGAGT GAAGAGATCAAGATCGGTGG-3′ (with the NdeI restriction site underlined) and 5’-CGCGGATCCGAGACGTCGAAGCGGTC GTG-3′ (with the BamHI restriction site underlined) for cloning into the pET-22b (+) expression vector, which includes a C-terminal hexahistidine tag. The recombinant plasmid was constructed as described and subsequently transformed into Escherichia coli BL21 competent cells (Vazyme, Nanjing, China). Transformants were selected on LB24 medium (peptone 10 g/L, yeast powder 24 g/L, NaCl 50 g/L, agar 15 g/L) with 50 μg/mL ampicillin.

The correct transformants were cultured to an OD_600 nm of 0.6 in LB24 medium containing 50 μg/mL ampicillin. To induce the expression of recombinant StKatG2, 0.05 mmol/L isopropyl-β-d-thiogalactopyranoside (IPTG) was added. The culture was incubated at 20°C for 10 h. Cells were lysed by suspending them in a 50 mmol/L solution of Tris–HCl (pH 7.4) and sonicated on ice to facilitate further disruption (20 kHz power for 12 min with 5 s burst sonication cycles at 5 s intervals). At 4°C, the crude enzyme solution was purified using a nickel column (Yeasen, Shanghai, China), with 4 mL of HisSep Ni-NTP Agarose Resin loaded into the purification column. Lysis Buffer (50 mmol/L NaH₂PO₄, 300 mmol/L NaCl, 10 mmol/L imidazole, pH 8.0) was employed for equilibration, followed by the introduction of 20 mL of crude enzyme solution into the nickel column resin. After washing away impurities with Wash Buffer (50 mmol/L NaH₂PO₄, 300 mmol/L NaCl, 20 mmol/L imidazole, pH 8.0), StKatG2 was eluted using Elution Buffer (50 mmol/L NaH₂PO₄, 300 mmol/L NaCl, 250 mmol/L imidazole, pH 8.0) and 7.5 mL was collected, and subsequently dialyzed for 12 h in a 50 mmol/L Tris–HCl solution (pH 7.4). The purified protein was analyzed by SDS-PAGE, and the StKatG2 protein concentration was determined using a BCA kit (Yeasen, Shanghai, China).

To evaluate the molecular weight of the proteins, purified protein samples were subjected to chromatographic separation and molecular weight determination using High-Performance Liquid Chromatography-Mass spectrometry (HPLC-MS) technology (Thermo Fisher Scientific, United States, Ultimate 3,000 Ultra High Performance Liquid Chromatography System, Q Exactiv™ Hybrid Quadrupole-Orbitrap™ Mass Spectrometer High Resolution Mass Spectrometer).

Manganese chloride (MnCl₂) was employed as the substrate to evaluate the Mn(II) oxidation activity of the recombinant StKatG2 enzyme in HEPES buffer at pH 8.0. The buffer composition included 59 mmol/L NaCl, 10 mmol/L CaCl₂, 0.4 mmol/L NADḤNa₂ (Sigma), 1  mmol/L H₂O₂, 1 μmol/L heme (Sigma) in 50 mmol/L, and 10 μmol/L pyrroloquinoline quinone (Sigma) (Kono and Fridovich, 1983). The reaction mixture was incubated for 10 h at 30°C. A 300 μL aliquot of each sample was reacted for 2 h at 30°C with 60 μL of LBB (Sigma, 0.04%, w/v) (Jones et al., 2019) and 900 μL of acetic acid (45 mmol/L). The activity was assessed by monitoring the absorbance at 620 nm and the equilibrated potassium permanganate concentration.

To determine the kinetic constant of recombinant StKatG2, various concentrations of MnCl₂ (ranging from 0–80 mmol/L) were incubated with 0.4 mg/mL of StKatG2 for 10 h at 30°C in 50 mmol/L Tris–HCl (pH 7.4) (Jang et al., 2005). The values for K_m and V_max were derived using non-linear regression fitting.

The effect of temperature on the enzymatic activity of StKatG2 was investigated by incubating the reaction mixture at temperatures ranging from 25 to 80°C under standard assay conditions at pH 7.0. The residual enzyme activity was evaluated at 50°C, with the maximum enzyme activity defined as 100%. Additionally, the optimal pH was determined by varying pH values from 4.0 to 8.0 while maintaining a constant temperature of 50°C.

To assess thermostability, StKatG2 samples were incubated at different temperatures (50, 60, 70, and 80°C) for various time intervals. Subsequently, residual activity was measured at 50°C and compared with protein samples stored on ice, which was considered to represent 100% activity.

EDTA and several metal ions, including Ca²⁺, Co²⁺, Cu²⁺, Fe³⁺, Mg²⁺, Zn²⁺, and Ni²⁺, were incubated with StKatG2 at final concentrations ranging from 0.1 mmol/L to 10 mmol/L at 50°C for a duration of 10 h. The effects of these ions on the activity of recombinant StKatG2 were then evaluated.

The MnCl₂ (50 mmol/L) and various auxiliary components were reacted with recombinant StKatG2 (0.5 mg/mL) at 55°C for 10 h. The fresh Mn oxides (BioMnOx) pellets were collected by centrifugation (10,000 × g, 10 min), and washed three times with deionized water before drying for 12 h at 55°C. Surface topography observations were performed using a scanning electron microscope (Thermo Fisher Scientific, United States, Apreo 2C) at 5.00 kV, equipped with an Oxford X-Max 80T EDX system (Oxford Instruments Ultim Max 65, Oxford, United Kingdom) for energy dispersive analysis of BioMnOx.

The phases were determined by scanning the dry form of the dioxide using an X-ray diffractometer (Beijing Purkinje General Instrument Co, China, X-D6) in the 2θ range of 5–85°. The valence state and composition of Mn were identified using the X-ray photoelectron spectroscopy (XPS; ESCALAB XI+, Thermo Fischer, United States).

The MG decolorization experiments were conducted in 2 mL tubes using a reciprocating shaker set at 30°C and 120 rpm. The reaction mixture comprised a 50 mmol/L Tris–HCl buffer containing 20 mg/L, 50 mg/L, and 80 mg/L of MG, along with 500 μg/mL of purified recombinant StKatG1 or StKatG2. Control groups were established that excluded the StKatG1 or StKatG2 enzymes. The absorbance of MG at 620 nm was measured using a UV spectrophotometer. The formula for calculating the decolorization rate is as follows:

All experiments were performed in triplicate.

High-Performance Liquid Chromatography-Mass Spectrometry (HPLC-MS) was employed to detect malachite green (MG) and its degradation products (Daneshvar et al., 2007). The malachite green or its metabolites were dried, dissolved in 1 mL of methanol, and subsequently injected into an Agilent 1,290 Infinity II HPLC system (Agilent, Santa Clara, United States) at a volume of 3 μL. This system was equipped with a reversed-phase C-18 analytical column and coupled to a quadrupole mass spectrometer. Methanol served as the mobile phase, with a flow rate of 0.6 mL min⁻¹. The initial methanol proportion was set at 5%, which was gradually increased to 95% over 10 min, maintained at 95% for an additional 2 min, and then reduced back to 5% over 2 min. The entire analytical process lasted for a total of 20 min (Han et al., 2020). The analysis utilized electrospray ionization (ESI) in positive ion mode, and the data were processed using Qualitative Analysis B.05.00 software (Agilent, United States).

The protonation state of all compounds was set at pH 7.4, and the compounds were converted to three-dimensional structures using Open Babel (O'Boyle et al., 2011). AutoDock Tools (ADT3) were utilized to prepare and parametrize the receptor protein and ligands. Docking grid files were generated using the AutoGrid feature of Sitemap, and docking simulations were performed with AutoDock Vina (1.2.0) (Trott and Olson, 2010; Eberhardt et al., 2021). The optimal pose was selected for interaction analysis. Finally, the protein-ligand interaction figure was generated using PyMOL.

Two catalase-peroxidases, StKatG1 (~80 kDa) and StKatG2 (~85 kDa), were identified from S. tamaricis F01. While StKatG1 has been confirmed to exhibit Mn(II) oxidation activities in previous research (Zhao et al., 2017; Zhao et al., 2023), the function of StKatG2 remains uncharacterized. StKatG2 shares 53.6% amino acid sequence identity with StKatG1 (Figure 1A and Supplementary Figure S1). Phylogenetic analysis revealed that StKatG1 clusters with catalase-peroxidases from Burkholderia pseudomallei, Salmonella enterica, and Escherichia coli, among others, whereas StKatG2 is distantly related (Figure 1B). Alphafold3 (v3.0) analysis indicated that StKatG2 possesses a typical catalase-peroxidase structure, comprising 23 α-helices, 22 short 3 10 helices, and 13 β-strands, with the critical histidine-arginine pair (His-112 and Arg-108) located in α-helix 4, essential for H₂O₂ breakdown (Poulos and Kraut, 1980). Additionally, StKatG2 also contains the triad His295-Asp406-Trp347 as the proximal heme iron ligand, along with Asp-130, Arg-444, and Tyr-248, which act as a molecular switch regulating catalase activity (Regelsberger et al., 2000; Jakopitsch et al., 2003; Yu et al., 2003; Singh et al., 2004; Carpena et al., 2005). A comparison of the structures of StKatG1 and StKatG2 revealed significant similarities, with a root mean square deviation (RMSD) of 0.98 Å. Both StKatG1 and StKatG2 exhibit three types of secondary structures: α-helices, short 3_10 helices, and β-strands. StKatG1 contains 24 α-helices, whereas StKatG2 has 23. In contrast, StKatG1 features 17 3_10 helices, while StKatG2 possesses 22. Regarding β-strands, StKatG1 has 10, compared to 13 in StKatG2 (Figure 1C). In conclusion, despite being derived from the same organism, StKatG1 and StKatG2 demonstrate notable distinctions in both sequence and structure.

To investigate the function of the StKatG2 protein, the full-length gene (2,297 bp) was cloned and inserted into the pET-22b (+) expression vector with a His6-tag at the C-terminal (Supplementary Figures S2, S3). The resulting recombinant plasmid was confirmed through double digestion with NdeI and BamHI restriction enzymes, yielding a correct 2.3 kb DNA fragment. Subsequently, the recombinant plasmid was transformed into E. coli BL21, and transformants on LB medium supplemented with ampicillin (50 μg/mL) were selected via colony PCR. The purified recombinant StKatG2 protein was obtained using an immobilized Ni²⁺-affinity column. A total of 20 mL of crude proteins (1792.71 μg/mL) was subjected to an immobilized Ni²⁺-affinity column, resulting in 7.5 mL of eluted proteins with a concentration of 229 μg/mL. Analysis by SDS-PAGE (10%) revealed a band at approximately about 85 kDa (Supplementary Figures S4, S5), consistent with the expected mass of StKatG2. The HPLC-MS result demonstrated that the actual molecular weight of StKatG2 was 86.97 kDa (Supplementary Figure S6B and Supplementary Tables S1, S2).

The manganese-oxidizing activity of recombinant StKatG2 was assessed using the LBB method (Jones et al., 2019). Upon reaction of StKatG2 with MnCl₂ in 50 mmol/L Tris–HCl (pH 7.4) buffer, insoluble brown particles were observed. Subsequent mixing of the reaction solution with the LBB solution resulted in a blue color change (Figure 2A). These reactions confirmed the formation of Mn oxides, demonstrating the ability of recombinant StKatG2 to oxidize Mn(II).

The oxidation ability of StKatG2 toward Mn(II) was further analyzed by calculating kinetic values at different concentrations of MnCl₂. A Michaelis–Menten plot was generated through non-linear regression curve fitting (Figure 2B). Table 1 provides a comparison of the kinetic characteristics of various manganese-catalases. It is evident that the K_m value of StKatG2 is 2.529 mmol/L, higher than that of StKatG1 (Zhao et al., 2023). Moreover, the K_m value of StKatG2 was found to be lower than those of most reported multicopper oxidases, suggesting that StKatG2 has a higher affinity for the substrate and greater manganese oxidation capacity. The V_max value for StKatG2 was determined to be 10.07 μmol/L·min⁻¹, slightly lower than StKatG1 (Zhao et al., 2023), which exhibited a V_max of 10.3 μmol/L·min⁻¹. This implies that StKatG1 has a slightly higher maximum velocity in the catalytic reaction compared to StKatG2. The K_cat of StKatG2 is 2.82 min⁻¹, slightly higher than that of StKatG1 with a K_cat value of 2.78 min⁻¹, indicating a slightly superior catalytic performance in terms of substrate turnover for StKatG2. These differences in kinetic values between StKatG1 and StKatG2 may be attributed to variances in enzyme structure, active site residues, or the microenvironment surrounding the active site, which could impact the efficiency of the catalytic reaction.

In order to further characterize the biogenic manganese oxides (BioMnOx) catalyzed by StKatG2, scanning electron microscopy (SEM) combined with energy dispersive spectroscopy (EDS) was conducted. Unlike layered sheet-like structure of BioMnOx catalyzed by StKatG1 (Zhao et al., 2023), the BioMnOx nanoparticles produced by StKatG2 appeared somewhat spherical or irregular, arranged closely with void spaces between them (Figures 3A,B). EDS analysis revealed the elemental composition on the surface of the BioMnOx particles, with mass fractions of 21.97% C, 13.64% O, and 64.39% Mn, and atomic fractions of 47.46% C, 22.12% O, and 30.41% Mn (Figure 3C).

To analyze the phase of the product, X-ray diffraction (XRD) was utilized for qualitative analysis of the phase composition of bio-manganese oxide (Figure 3D). The results revealed that the primary phase composition of bio-manganese oxide consisted of MnO₂ (JCPDS 39–0735), Mn₂O₃ (JCPDS 24–0734), and C₂MnO₄ (JCPDS 32–0646). These findings suggest that the Mn compounds catalyzed by StKatG2 exhibit multiple oxidation states. Furthermore, notable peaks of NaCl (JCPDS 99–0059) were observed, likely attributed to the incorporation of sodium chloride during the preparation of BioMnOx.

Valence states compositions in the BioMnOx were identified through photoelectron spectroscopy (XPS). The XPS analysis showed two peaks in Mn2p_1/2 and five in Mn2p_3/2 (Figure 3E), consistently indicating that the BioMnOx produced by StKatG2 is a mixed-valence manganese compound. Specifically, XPS measurements revealed the presence of four valence states: Mn(II), Mn(III), Mn(IV), and Mn(VII) in the BioMnOx (Figure 3F). The multiple oxidation states in the BioMnOx product catalyzed by StKatG2 are similar to those catalyzed by StKatG1, which also exhibits mixed valences (Zhao et al., 2023).

The impact of pH conditions on the Mn(II)-oxidizing activity of StKatG2 was investigated. The optimal pH for StKatG2 to exhibit Mn(II)-oxidizing activity was found to be 7.5 (Figure 4A). The relative enzyme activity of StKatG2 is only 8.3% at pH 4.0, and the highest enzyme activity was observed around 7.5, which indicated that neutral or slightly alkaline environments were beneficial to the Mn(II)-oxidizing activity of StKatG2. This trend is consistent with other studies on manganese oxidases, such as CopA with an optimal pH of 8.0. CueO, another manganese oxidase, shows peak activity between pH 7.6 and 8.2, with minimal oxidation activity below pH 7.0 (Su et al., 2014). These findings suggest that most manganese oxidases involved in biomanganese oxidation exhibit maximal enzyme activity in neutral to alkaline pH ranges, while chemical oxidation of manganese becomes predominant at pH levels above 8.0, as demonstrated by the control sample without StKatG2 (Supplementary Figure S7).

To determine the optimal temperature for Mn(II) oxidizing activity, StKatG2 was evaluated across a temperature range of 25–80°C. The study revealed that the recombinant StKatG2 exhibited its peak enzymatic activity at 55°C (Figure 4B). Notably, the enzyme retained 40% activity at lower temperatures (25°C) and maintained over 80% activity within the 45–55°C range. However, enzyme activity significantly decreased beyond 55°C, with a sharp drop to 38.3% at 80°C. These findings indicate that StKatG2 is a mesophilic enzyme. Moreover, the optimal temperature for manganese oxidase activity in recombinant StKatG2 was higher than that observed for StKatG1 (50°C) (Zhao et al., 2023).

Thermostability of StKatG2 was evaluated by subjecting the protein to different temperatures (50, 60, 70, and 80°C) for a specific duration, followed by measuring residual activity at 55°C (Figure 4C). After an 8-h exposure to 50°C, the enzyme retained 82.3% of its activity, surpassing StKatG1, which only maintained 73.6% activity under similar conditions (Zhao et al., 2023). Notably, StKatG2 demonstrated superior thermal stability, retaining 45.1% activity after 8-h exposure to 80°C. In contrast, StKatG1 lost 92% of its activity when exposed to 80°C for 8 h (Zhao et al., 2023). These results collectively indicate that StKatG2 exhibits robust thermal stability in high-temperature environments and shows promise for broader applications in such environments.

The impact of metal ions and the chelating agent EDTA on the enzymatic activity of StKatG2 was assessed (Figure 5). EDTA significantly suppressed the Mn(II)-oxidizing activity of StKatG2, suggesting a reliance on divalent metal ions. At a concentration of 0.1 mmol/L, Fe³⁺ inhibited the enzyme activity (Figure 5A), likely because of its structural and chemical similarity to Mn, leading to potential competition for binding sites on the enzyme and consequent disruption of its structure. Moreover, the addition of 0.1 mmol/L Cu²⁺, Mg²⁺, and Zn²⁺ enhanced the enzyme activity, possibly by forming a metalloenzyme complex that boosts enzymatic function. However, a higher concentration (10 mmol/L) of Cu²⁺ and Zn²⁺ almost completely inhibited the enzymatic activity, while Fe³⁺ at the same concentration affected the final detection outcome due to its inherent color, which holds no analytical value (Figure 5C).

To evaluate the performance of recombinant StKatG2, toxic malachite green was utilized as the research material. Upon mixing with the enzyme solution, the decolorization ability was monitored. StKatG2 exhibited the fastest decolorization efficiency within the initial 15 min for both 20 mg/L and 50 mg/L of malachite green, peaking around 30 min, with a decolorization rate of up to 30% after 4 h of action (Figure 6). Although the BioMnOx structures of StKatG1 and StKatG2 are similar and the kinetic values of Mn(II) are also comparable (Supplementary Table S3), StKatG2 exhibits excellent decolorization efficiency for malachite green. After 4 h of treatments, the maximum decolorization rates of StKatG2 reached 73.38% for 20 mg/L MG and 60.08% for 50 mg/L MG (Figure 6A), while StKatG1 for 20 mg/L mg and 50 mg/L mg were 63.91 and 51.78%, respectively (Figure 6B). When the concentration of MG reaches 80 mg/L, the decolorization rates of StKatG1 and StKatG2 were 34.63 and 18.37%, respectively (Supplementary Figure S8). These findings highlight StKatG2 as a promising biological approach for the efficient degradation of malachite green within a short period.

To further investigate the decolorization mechanism of MG by treatment with StKatG2, the MG metabolites were analyzed using HPLC-MS. Following treatment with StKatG2, the MG peak, which exhibited a retention time of 6.845 min, was found to be reduced (Figure 7A). In contrast, several peaks with retention times ranging from 5 to 8 min increased (Figure 7B) when compared to the control setup. Three distinct peaks were identified as MG metabolites in the StKatG2 treatment setup (Figures 7C–E): malachite green (7.153 min, m/z = 329), leucomalachite green (6.845 min, m/z = 315), and 4-(dimethylamino) benzophenone (7.791 min, m/z = 226). The latter is recognized as a degradation metabolite of MG (Yang et al., 2015; Abu-Hussien et al., 2022). The process of MG degradation catalyzed by StKatG2 is illustrated in Figure 7F.

Furthermore, We conducted an analysis of the interactions between the proteins and MG (Supplementary Figure S9). Utilizing these interaction forces, the binding energy of the StKatG1-MG and StKatG2-MG complexes were determined to be −7.5 kcal/mol and −8.0 kcal/mol, respectively (Supplementary Figure S9). Molecular docking results indicate that MG interacts with both StKatG1 and StKatG2, with a position away from the heme binding sites, Mn²⁺ binding sites^, and the H₂O₂ catalytic pocket. Moreover, the docking positions of StKatG1 and StKatG2 with MG are inconsistent, which may contribute to the differing decolorization efficiencies observed for MG. When heme and Mn²⁺ combine with StKatG2 for manganese oxidation, this interaction may influence the molecular conformation of the proteins and release redox forces, ultimately leading to MG decolorization.