Six macroalgae samples, identified as Callithamnion tetragonum, Plocamium cartilagineum, Pachymenia orbicularis, Desmarestia Antarctica, Phaeurus antarcticus, and Melanthalia abscissa, were collected from the seashore of King George Island, Antarctica (Gui et al., 2020). After washing with sterilized seawater, the macroalgae samples were brought back to the laboratory to collect the surface-associated bacteria.

The macroalgae were again rinsed with sterilized seawater, and then cut into fragments. The pieces were placed in sterilized tubes, and then 20 ml sterilized seawater was added. The mixture was shaken for 30 s, three times, with a vortex mixer. The liquid was filtered to trap the associated bacteria on a filter membrane (47 mm × 0.22 μm, Merck Millipore, Darmstadt, Germany).

The total genomic DNA of enriched macroalgae-associated bacteria was extracted using the FastDNA Spin Kit For Soil (MP Biomedicals, Santa Ana, CA, United States) following the manufacturer’s instructions. The quantified DNA samples were sequenced and analyzed on the Illumina HiSeq platform (Illumina, San Diego, CA, United States) by Novogene Bioinformatics Technology Co., Ltd (Beijing, China).

The vector pUC118, restriction endonuclease and Eschericia coli DH5α cells were purchased from TaKaRa (Dalian, China), the expression vector pET-30a (+) and E. coli BL21 (DE3) cells were purchased from Novagen (Merck, United States). His-tag-specific purification kit was purchased from GE Healthcare (Pittsburgh, PA, United States). All the chemical reagents are chemically pure.

The compositions of the media and Lugol’s dye solution were as follows: Luria–Bertani (LB) medium: tryptone 10.0 g/L, yeast powder 5.0 g/L, NaCl 10.0 g/L; Screening plates: tryptone 10.0 g/L, yeast powder 5.0 g/L, carrageenan 15 g/L, and NaCl 10.0 g/L. The LB and screening plate media were sterilized by autoclaving (1.21 × 10⁵ Pa, 20 min). Lugol’s solution consisted of 5 g I₂ and 20 g KI dissolved in 100 ml sterilized distilled water.

A presumptive carrageenase gene, designated as car1383, was obtained from the predicted functional genes in the metagenomic data obtained from Antarctic macroalgae-associated bacteria. Multiple sequence alignment was conducted using DNAMAN software.¹ BlastP and BlastN² and Motif Search³ were used for identity and motif analyses, respectively. A phylogenetic tree was conducted using the neighbor-joining (NJ) method by MEGA 5.0. The sequence of car1383 with specific restriction sites of SalI/XhoI was synthesized by the Beijing Genomics Corp (Beijing, China). This sequence was ligated to the pET30a vector, which was digested with same restriction endonucleases, to obtain the recombinant plasmid pET30a + car1383. The recombinant plasmid was used to transform to E. coli BL21, and the transformed cells were cultured on a screening plate on medium containing 50 μg/ml kanamycin. After culture for 24 h at 37°C, the bacteria were dyed with Lugol’s solution. The positively transformed strain showed a distinct clear circle and was assumed to have carrageenase activity.

To obtain purified recombinant carrageenase Car1383, the positive strain was inoculated into LB liquid medium containing 50 μg/ml kanamycin. When the OD₆₀₀ of the culture medium reached 0.6, adding 0.5 mM isopropyl-β-D-thiogalactopyranoside (IPTG) to induce the expression of the fusion protein Car1383. The culture medium was centrifuged at 10,000 g at 4°C, then the pellet was dissolved in PBS buffer (pH 7.0). The harvested cells were disrupted using a high-pressure cell cracker (Constant Systems, Co., Ltd., Daventry, United Kingdom). After centrifugation, the supernatant was purified using a His Bind Purification kit (GE Healthcare, Piscataway, NJ, United States) according to the manufacturer’s instructions. The purified recombinant protein was identified by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE).

The enzyme activity of Car1383 was determined using the 3,5-dinitrosalicylic acid method (Miller, 1959). In brief, purified Car1383 was pre-incubated with 0.1% κ-carrageenan solution (0.2 mol/L Na₂HPO₄-NaH₂PO₄ buffer, pH 7.0) at 50°C for 30 min, and then the reaction was terminated by heating 100°C for 10 min. One unit of enzyme activity was defined as the amount of enzyme that catalyzes the conversion of 1 μmol of reducing sugars per minute under these conditions.

To identify the active site responsible for the activity of Car1383, point mutations were introduced at two putative active sites and four conserved amino acids (Figure 1; Table 1). Mutated genes (M1–M18) corresponding to mutations listed in Table 1 were synthesized by the Beijing Genomics Corp (Beijing, China). These synthetic genes were first digested by restriction endonucleases SalI/XhoI, then were purified by the Gel Extraction Kit. The purified gene was ligated into vector pET30a, which had been digested with the same two restriction endonucleases. These recombinant plasmid were transformed into E. coli BL21 (DE3) competent cells and confirmed by DNA sequencing. The transformants harboring recombinant plasmid were cultured in LB medium with 50 μg/ml kanamycin at 37°C to OD₆₀₀ of 0.6 and induced with 0.5 mM isopropyl-β-D-thiogalactopyranoside (IPTG). Eighteen recombinant mutants were firstly checked using plate-based assays, then the enzymatic activity of purified recombinant proteins was determined using the standard DNS method described above.

The significant differences of enzyme activity between selected mutants were determined using one-way ANOVA with the SPSS 19.0 statistical software (^*p < 0.05, ^**p < 0.01).

The car1383 gene, encoding a mature form of the protein with 22 amino acids signal peptide sequence, was ligated to the pET30a vector and correctly expressed in E. coli BL21 (DE3). The SDS-PAGE result confirmed that 0.5 mM IPTG induced its expression. After purification using a Ni-NTA His Tag Kit, a single band corresponding to 40 kDa protein marker was obtained, as determined by SDS-PAGE. This size was consistent with the theoretical molecular mass of Car1383 (Figure 3).

The reaction temperature and thermostability are very important biochemical characteristics of an enzyme for industrial applications. Unexpectedly, the optimal temperature for Car1383 activity was 50°C (Figure 4A), higher than those of CgkX, CgkS, CgkHC4, and CgkZ (Sun et al., 2010; Li et al., 2013; Liu et al., 2013; Wang et al., 2015; Table 2). Even though Car1383 originates from the Antarctic marine ecosystem, it maintained more than 70% of its initial activity at 60°C (Figure 4B). Thus, Car1383 has high thermostability, with half-lives of 9 h at 60°C and 24 h at 50°C. Carrageenan is highly viscous at room temperature and is an inhibitor of carrageenase degradation activity. Usually, increasing the dissolution temperature is used to reduce the viscosity of carrageenan solution. However, most identified carrageenases are unstable above 40°C (Chauhan and Saxena, 2016). Therefore, a thermostable carrageenase such as Car1383 is of considerable interest as a new biocatalyst for producing enzymatic-degradation products of carrageenan on a commercial scale.

The optimal pH for Car1383 activity was 6.0, lower than that reported for most other κ-carrageenases (Ma et al., 2010; Sun et al., 2010; Kobayashi et al., 2012; Li et al., 2013; Wang et al., 2015; Table 2). In our experiments, Car1383 retained almost 50% of its original activity at pH 10.0 and 40% of its original activity at pH 11.0 (Figure 4C).

Metal ions had different effects on the activity of recombinant Car1383. Its activity was partially inhibited by Sr²⁺, Ca²⁺, Ni²⁺, Ba²⁺, and Mn²⁺, and completely inhibited by Cu²⁺, Fe³⁺, and Mg²⁺. The activity of many enzymes is negatively affected by heavy metal ions like mercury (Hg²⁺), cobalt (Co²⁺), zinc (Zn²⁺), Cu²⁺, aluminum (A1³⁺), and lead (Pb²⁺), indicating that enzyme activity can be cation-independent and that these metals can alter the conformation of enzymes (Araki et al., 1999; Ma et al., 2010; Li et al., 2013). However, the carrageenase activity of recombinant Car1383 was enhanced by Na⁺, K⁺, and especially Fe²⁺, which more than doubled its initial activity. In other studies, cations such as sodium (Na⁺) and potassium (K⁺) were found to increase carrageenase activity (Ma et al., 2010; Li et al., 2013, 2014), especially high concentrations Na⁺ (Ma et al., 2013), while reagents such as iodoacetic acid, Tween-80, and EDTA had little effect on carrageenase activity (Sarwar et al., 1983). In our research, Car1383 activity was completely inhibited by EDTA (Figure 4D), indicating that metal ions play an important role in the activity of this enzyme.

In this study, the V_max and K_m of Car1383 were 55.77 mg/ml·min and 6.51 mg/ml, respectively (Figure 5). The K_m of Car1383 was lower than that of some κ-carrageenases, such as CgkAJ5, CgkHC4, and CgkS (Ma et al., 2010; Sun et al., 2010; Wang et al., 2015), but higher than that of CgkZ (0.84 mg/ml; Liu et al., 2013; Table 2). The lower the K_m value for an enzyme, the higher its affinity for the given substrate. Therefore, the low K_m of Car1383 is indicative of a high affinity toward its substrate.

Thin layer chromatography is a convenient and effective method to detect different polymerization degrees oligosaccharides (Dhalwal et al., 2008). From the TLC results, we observed that Car1383 degraded κ-carrageenan into different polymerization degrees oligosaccharides. According to the migration rate of degraded oligosaccharides, the main hydrolysis products were disaccharides (Dp2) and tetrasaccharides (Dp4). This range of products is the same as that produced by CgkS (Wang et al., 2015), but different from those produced by CgkAJ5, CgkX, CgkHC4, and CgkZ (Ma et al., 2010; Sun et al., 2010; Li et al., 2013; Liu et al., 2013; Table 2). Hexasaccharides (Dp6) and those with higher degrees of polymerization were not observed on the TLC plate (Figure 6). From 30 min to 12 h of hydrolysis, the Dp4 products became less abundant, while the Dp2 products became more abundant. This may mean that κ-carrageenan was firstly degraded into Dp6, then Dp6 was degraded into Dp2 and Dp4.

However, the presence of sulfate groups in sulfate oligosaccharides has an effect on migration and resolution of these oligosaccharides on TLC plates (Zhang et al., 2007), so, the degradation products were further analyzed by electrospray mass spectrometry (ESI-MS). The mass spectra of products after 30 min of hydrolysis had signals of κ-carrageenan oligosaccharides, mainly Dp2, Dp4, and Dp6. The peak at m/z 403.11 was assigned to a disaccharide [(G4S-An)]⁻ carrying one sulfate group. The peak at m/z 811.31 was assigned to a sodium salt of tetrasaccharide [(G4S-An)₂Na]⁻ carrying two sulfate groups. And the peak at m/z 1219.25 was likely corresponding to a hexasaccharide with three sulfate groups carrying two sodium ions [(G4S-An)₃2Na]⁻ (Figure 7A). In the mass spectra of the reaction mixture after 12 h of hydrolysis, only signals for Dp2 (m/z 403.23) and Dp4 (m/z 811.31) were observed, and there were no signals corresponding to Dp6 (Figure 7B). Together with the TLC results, the ESI-MS data indicated that κ-carrageenan was firstly degraded into Dp6, then Dp6 was degraded into Dp2 and Dp4. These results showed enzymatic properties of Car1383, for which the cleavage group was An-G4S unit, resulting in a series of neocarrabiose oligosaccharides with the code of An-G4S. Previous studies have shown that carrageenases cleave the internal β-1, 4 linkage in κ-, ι-, and λ-carrageenan (Shen et al., 2017), similar with reported carrageenases, Car1383 is an endohydrolases and belonged to the GH family.

Protein engineering, including directed evolution and rational design, is an effective technology for improving enzyme properties (Liu et al., 2017). To understand the potential active sites responsible for activity, the residues E₁₉₀ and E₁₉₅ were replaced with Ala or Lys. To explore the effect of conserved sites on activity, four amino acid residues were replaced with alkaline amino acids, Lys, Arg, or His. The results showed that the activity of those mutants was lower than that of wild-type Car1383, whether E₁₉₀ and E₁₉₅ were replaced by Ala or Lys, or both of them were replaced with Ala or Lys. These findings suggested that E₁₉₀ and E₁₉₅ play important roles in the enzymatic activity of Car1383. In contrast, the activities of the mutants W183K and G255K were almost 2-fold higher than that of wild-type Car1383 (Figures 8, 9). This may be because Lys has a positively charged side chain, which leads to enhanced binding to negatively charged carboxyl groups of the substrate (Xu et al., 2021). However, the activity of the mutant D169K was a little lower than that of wild-type Car1383, maybe because the site was located near the active sites. The substitutions had little effect on the pH adaptability of the mutants W183K and G255K, which showed an optimal pH of 6.0 (Figure 8). Xu et al. (2021) successfully increased the catalytic efficiency of alginate lyase AlgL-CD through site-directed mutagenesis, and the catalytic activity of XynB (Wu et al., 2017) and alginate lyase AlgL (Jang et al., 2016) have been enhanced by rational design. Wu et al. (2017) reported that the substrate affinity of polysaccharide lyases could be improved by tight binding to Ca²⁺ via one amino acid substitution. Our previous site-directed mutagenesis study indicated that two Ca-binding sites are responsible for the Ca²⁺ activation and thermostability of agarase Aga3463 (Li et al., 2019).

In summary, we obtained a carrageenase Car1383, belonging to the GH16 family, derived from the metagenome of bacteria associated with the surface of Antarctic macroalgae. Our results showed that recombinant Car1383 has a higher optimal reaction temperature and better thermostability than most reported carrageenases, even though it is derived from the Antarctic environment. This study provides a new candidate for industrial preparation of bioactive algal oligosaccharides and paves the way for efficient utilization of macroalgae.