A Superefficient Ochratoxin A Hydrolase with Promising Potential for Industrial Applications

In a previous study, aflatoxin B1 degradation by strain CW117 was characterized (25), but the difunctional strain equally degrades OTA efficiently. As illustrated in Fig. 2A, the strain CW117 completely degraded 50 μg/L OTA within 60 h. Further investigation found that cells suspension and cell lysate in phosphate-buffered saline (PBS, pH 7.2) showed efficient degrading activity on OTA, but the cell-free supernatant from the culture of strain CW117 had no degradation ability (Fig. 2B). At the same cell density, fresh cells completely degraded 50 μg/L OTA in about 6 h while the cell lysate took 12 h (Fig. 2C). The results indicated that the degradation agent from CW117 is intracellular.

The degradation process of OTA by strain CW117 and the degradation product were monitored by high-performance liquid chromatography (HPLC) analysis. The treatment of OTA by cell suspension or cell lysate produced a new chromatographic peak with the same retention time of OTα (Fig. 2D). With the incubation time extension, the chromatographic peak of OTA was decreased gradually, which was accompanied with the increase of the new product. After the incubation with cell suspension of strain CW117 for 6 h, OTA peak was completely disappeared. By liquid chromatography-tandem mass spectrometry (MS/MS), the new product derived from OTA degradation by strain CW117 produced [M+H⁺] at m/z 257 as precursor ion in MS spectrum, and product ions at m/z 167 and 210.9 in MS/MS spectrum, which are consistent with OTα mass spectrometry (Fig. S1) (10). The degradation dynamics analysis showed that at each time point of degradation process, the molar equivalent of OTα production was equal to the OTA degradation, indicating that the degraded OTA was completely transformed into OTα (Fig. 1E). The product OTα was not decreased, indicating that OTα cannot be degraded further by the strain.

Effects of environmental factors (or as enzymatic characterization) on the ability of the strain to degrade OTA were investigated (Fig. 3A to D). The optimal temperature for OTA degradation was 20°C to 40°C. The function to degrade OTA was destroyed and decreased to 5% of original activity by the temperature up to 60°C. However, low temperature showed a little effect on the ability to degrade OTA. Even the incubation temperature down to 0°C, the function was decreased only by 17%. The optimal pH for the OTA degradation of strain CW117 was 6 to 8, while pH 5 or 10 significantly inhibited the function. Cu²⁺ or Zn²⁺ ions or 1% SDS significantly destroyed the degrading ability of strain CW117 on OTA. Other factors including Mg²⁺, Ca²⁺, Fe³⁺, metal-chelators (EDTA and EGTA), and proteinase K showed moderate effects to inhibit the degrading function. The enzymatic characterization indicated that the OTA degradation agent should be allocated to active proteins.

The results above indicated that the degradation agent of CW117 as intracellular enzyme(s) to hydrolyze the amide bond of the chemical and produce non-toxic chemical OTα (10, 12, 26). Degradation of OTA to OTα is attributed to amide bond cleavage in OTA structure (10). A complete genome analysis of CW117 was performed to prospect potential candidates including peptidase, amidase, amidohydrolase and carboxypeptidase that possible to hydrolyze amido bond, and thus to degrade OTA. By the CW117 genome (GenBank accession no. CP062156.1) analyses, 53 hydrolase genes (Table S4) were screened and they are overexpressed in Escherichia coli for testing OTA degradation. As a result, only hydrolase of ADH3 (Table S4, H7691_12935) showed efficient OTA degradation activity, and thus selected for further analysis. Based on the result of multiple sequences alignments (Fig. S5, Fig. 4A) and OTA degradation tests, ADH3 which encoded by the gene adh3 was identified as the amido hydrolase in strain CW117 for OTA degradation. ADH3 was characterized as amidohydrolase and composed of 427 amino acid residues with a predicted molecular weight of 45.6 kDa and isoelectronic point of 6.9. The recombinant protein rADH3 (Fig. S6) showed strong ability to degrade OTA. With the concentration of 1.2 μg/mL protein, rADH3 completely degraded OTA of 0.1 μM in 90 s. The degraded product was OTα as produced by strain CW117 (the MS/MS spectrum was identical to Fig. S1).

As reported above in cell suspension or cell lysate prepared from strain CW117, rADH3 was susceptible for some environmental factors including protein denaturants (1% SDS) and proteinase K (Fig. 4B to F). However, ADH3 showed higher temperature adaptability to exert the degradation function than its host strain CW117 and the commercial detoxification enzyme CPA, it showed the considerable activity even the temperature down to 0°C or up to 70°C (Fig. 4E). Several previously identified OTA detoxification enzymes including OTase, CP, PJ_1540, and AfOTase (24, 26–28) were overexpressed in E. coli (Table 1), and the recombinant proteins were purified in this study to compare the abilities of OTA detoxification with ADH3 (Fig. 5, Fig. 6). The kinetic study showed that Km value of rADH3 was 0.000039 mM, and the Kcat/Km value was 303937.85 s⁻¹.mM⁻¹, while the Kcat/Km values of other OTA detoxification enzymes were from 0.55 to 5363.25 s⁻¹.mM⁻¹ (Table 2). Compared with the mainstream commercial detoxification enzyme CPA identified from bovine pancreas with a Kcat/Km value of 8.7561 s⁻¹.mM^−1, the current rADH3 increased the hydrolytic efficiency of OTA by almost 35,000 times, showing an extremely high detoxification ability.

Based on the homology model of ADH3 (Fig. 7A), we selected 10 candidate amino acid residues located at the catalytic active center of ADH3 for point mutation analysis. Those ADH3 mutants were recombinantly expressed in E. coli BL21 to compare OTA degradation abilities. Activity validation results suggested that four residues (83H, 85H, 271H, and 344D) on ADH3, which form quaternary catalytic center, were important for OTA degradation (Fig. 7B). Point mutations on those four residues resulted in complete function loss (Fig. 7A and B). As listed in Table 2, the OTA detoxification activity of ADH3 is ∼210 and ∼56 times stronger than that of OTase and AfOTase, respectively. Solvent accessibility (SAS) assay revealed that the catalytic center on ADH3 has wider hydrophobic areas and larger open areas than that on OTase, and thus facilitates substrate OTA to reach catalytic center (Fig. 7C). Compared with ADH3, the catalytic center of OTase was located deep in the pore area, and it sank ∼5Å. Furthermore, we analyzed the root mean squared deviation (RMSD) and radius of gyration (R_g) of quaternary catalytic sites in those three OTA detoxification enzymes using molecular dynamics (MD) simulation. Among them, ADH3 quaternary catalytic site has the highest RMSD (Fig. 7D) and R_g (Fig. 7E), suggesting the most actively catalytic site.

In this study, an amidohydrolase ADH3 with a strong ability to transform OTA into nontoxic product ochratoxin-α was discovered. The rADH3 catalytic efficiency (Kcat/Km) was 56.7 to 35,000 times higher than those of previous published detoxifying enzymes rAfOTase, rOTase, and commercial enzyme CPA. The catalytic mechanism prediction suggested that ADH3 forms a larger hydrophobic area than other detoxifying enzymes at the cavity of the catalytic sites, and this structure gives the OTA easier access to catalytic sites. In addition, ADH3 shows considerable temperature adaptability (0 to 70°C) to exert the hydrolytic function. Collectively, detoxifying enzyme ADH3 showed promising potential for industrial applications.

Fresh culture of strain CW117 (1.0 mL) was inoculated into 100 mL nutrient broth (NB, Difco) containing about 50 μg/L OTA, and agitated at 160 revolutions per minute (rpm) and 37°C. The E. coli DH5α was used as the negative control. During degradation tests, 1.0 mL of the culture was taken at time points of 0, 1st, 3rd, 6th, 9th, 12th, 24th, 36th, 48th, and 60th h, and OTA residues were cleaned by OchraTest™ column (OchraTest™, Vicam, Milford, MA, USA) and analyzed by HPLC following a previously described method (10). During HPLC analysis, OTA was separated by an XBridge^TM C₁₈ column (4.6 × 250 mm, 5 μm; Waters, USA), and detected by the fluorescence detector 2475 (Waters, USA) with excitation wavelength of 340 nm and emission wavelength of 460 nm. The mobile phase consisted of acetonitrile, water, and glacial acetic acid by the volume ratio of 48:51:10 and the flow rate was 1.0 mL/min, with the injection volume of 10 μL.

The degradation agents (i.e., culture supernatant, cell lysate, and cell suspension) of strain CW117 were prepared by procedures similar to those described by Cai et al. (25). Strain CW117 was grown in NB at 37°C for 48 h. Bacterial cells were collected by centrifugation at 8,000 × g at 4°C for 5 min; by centrifugation, the pellet and culture supernatant were separated. Then, the culture supernatant was passed through a sterile filter (0.22 μm, Millipore, Billerica, MA, USA) to produce the cell-free culture supernatant. The cell suspension was prepared by resuspending bacterial cells in PBS (pH 7.2) after the cell pellet washed twice with distilled water. The cell lysate was prepared by disintegrating the cell suspension with an ultrasonicator (Qsonica Q700, Newtown, CT, USA), then centrifuged at 10,000 × g for 10 min, and passed through the supernatant with a sterile filter (0.22 μm, Millipore, Billerica, MA, USA). After degradation agents prepared, OTA degradation activity was tested. Briefly, 2.0 mL degradation agent (culture supernatant, cell suspension, or cell lysate) was mixed with the same volume of OTA standard (diluted with PBS, pH 7.2) at a final concentration of about 50 μg/L, and incubated at 37°C for 12 h, and the OTA residues were analyzed by HPLC as above. Degradation dynamics of cell suspension (i.e., cells) or cell lysate was determined by 10.0 mL degradation agent mixed with 10 mL of OTA standard as above, and the OTA residues were analyzed at time points 0, 1st, 3rd, 6th, 9th, 12th, and 24th h. The sterile NB was selected as the control for degradation tests.

During the sample analysis on OTA residues in degradation tests, one degradation product that affined to OchraTest column (Vicam, Milford, MA, USA) was observed and collected by the immunoaffinity (i.e., OchraTest) column according to the manufacturer’s instructions. The degradation product was further analyzed with OTA and OTα standards by HPLC and LC-MS/MS according to the method of Wei et al. (10). The dynamics of OTA degradation and OTα production by cell suspension (i.e., cells) and cell lysate were performed in 20 mL degradation mixture as above, and the degradation samples were collected at 0, 1st, 3rd, 6th, 9th, and 12th h for OTA and OTα determination by HPLC. The sterile NB was selected as the control for the degradation tests.

Unless otherwise indicated, OTA solution in the following tests was prepared by diluting OTA standard with PBS (pH 7.2). The 1.0 mL cell suspension (PBS, pH 7.2) was mixed with the same volume of OTA solution (final concentration 40 μg/L), and incubated at 0, 20, 30, 40, 50, and 60°C for 24 h, respectively, and then OTA residues were analyzed. For pH evaluation, the collected cell pellet were resuspended in buffers with different pH values (pH 2.0 to 5.0, glycine-HCl buffer; pH 6.0 to 7.0, phosphate buffer; pH 8.0 to 9.0, Tris-HCl buffer; pH 10.0 to 12.0, glycine-NaOH buffer). After that, 1.0 mL prepared cell suspension was mixed with 1.0 mL OTA solution that prepared by the same pH buffer (final concentration 40 μg/L), and incubated at 37°C for 24 h, then the OTA residues were analyzed. For metal ions test, the collected cell pellet was suspended in PBS (pH 7.2) containing 0.02 mol/L each metal ion. The 1.0 mL prepared cells suspension mixed with the same volume of OTA solution (final concentration 40 μg/L) and incubated at 37°C for 6 h, and the OTA residues were analyzed. For metal-chelator and protein denaturant tests, the collected cell pellet was resuspended in PBS (pH 7.2) which contained 0.2 mol/L metal-chelator (EDTA or EGTA), 1.0 mg/mL proteinase K, 1% SDS, or 1.0 mg/mL proteinase K plus 1% SDS, and the prepared cell suspensions were incubated at 37°C for 6 h before the OTA degradation tests. After pretreatment, 1.0 mL cell suspension with different treatments was mixed with the same volume of OTA solution (final concentration 40 μg/L) and incubated at 37°C for 12 h, and then the OTA residues were analyzed. The PBS with corresponding metal-chelator or denaturant was selected as negative control and untreated bacterial cells were selected as positive control.

S. acidaminiphila CW117 was cultivated in NB with 160 rpm shaking for 48 h. Genomic DNA was extracted by using the DNA isolation kit (Qiagen, Shenzhen, China) and purified by phenol:chloroform:isoamyl alcohol (25:24:1, vol/vol/vol) extraction. Genome sequencing was performed by Guangdong MAGIGENE Biotechnology Co., Ltd (Guangzhou, China) using the PacBio Sequel II and Novaseq 6000 sequencer. The reads from PacBio Sequel were assembled using SMRT Link v5.1.0, and the reads contained data from Novaseq and PacBio Sequel were assembled using Unicycle v0.4.8. Putative genes were identified using Glimmer 3.02 (35) or Prodigal (36). Noncoding RNAs of tRNA and rRNA were identified using tRNAscan-SE (version 1.3.1), rRNAmmer (version 1.2), and the sRNA was identified using Rfam database and screened by cmsearch. Interspersed repeats and tandem repeats were analyzed by RepeatMasker (version open-4.0.5) and Tandem repeats finder (version 4.07b). Prophage was identified by PHAST, and Genomic Islands (GIs) were predicted by IslandPath-DIOMB. Annotation was performed with BLAST search against databases (37), including the National Center for Biotechnology Information (NCBI) nonredundant proteins (NR) (38), clusters of orthologous groups of proteins (COG) (39), the Kyoto Encyclopedia of Genes and Genomes (KEGG) (40), Gene Ontology (GO) (41), Swiss-prot (42), Pfam (43), Carbohydrate-Active Enzymes Database (CAZy) (44), pathogen host interactions (PHI) (45), virulence factors of pathogenic bacteria (VFDB) (46), and Antibiotic Resistance Genes Database (ARDB) (47). After gene annotation, amido bond hydrolases, which are considered as the possible OTA detoxification genes, were screened for following protein expression and OTA degradation assay in vitro.

The candidate genes were amplified by using CW117 genomic DNA and each gene primer pair (Text S1, Table S4) with Primerstar max DNA polymerase (TaKaRa, Dalian, China). The purified PCR product (e.g., adh3 fragment) and expression plasmid pGEX-4T-1 were digested by BamHI (or other corresponding endonuclease) and XhoI (or other corresponding endonuclease) at 37°C overnight, respectively. The digest mixture contained 1.0 μL each endonuclease (e.g., XhoI and BamHI), 1 μg of purified PCR fragment (or 1 μg pGEX-4T-1), and 5 μL of 10× H buffer (TaKaRa, Dalian, China), and supplemented distilled water up to 50 μL. The purified digested fragment and digested expression plasmid were ligated by solution I DNA ligase for 1 h at 16°C. The recombinant plasmid (e.g., pGEX-4T-1/adh3) was constructed by ligation solution consisting of 2 μL digested pGEX-4T-1, 4 μL digested PCR fragment, and 6 μL ligase. Thereafter, recombinant plasmid (5 μL) was transformed to E. coli Trans1-T1 and identified by PCR sequencing. The positive clone was enriched by Luria-Bertani (LB) broth contained 50 μg/mL ampicillin (AmpR), and the recombinant plasmid was extracted by Plasmid MiniPrep Kit (Axygen, NY, USA) according to manufacturer's instructions. The recombinant plasmid (e.g., pGEX-4T-1/adh3) was transformed to E. coli BL21 by heated shock and the transformant was incubated in AmpR LB broth at appropriate temperature (from 16°C to 37°C for different genes) with agitation of 160 rpm. When the culture optical density (OD₆₀₀) reached about 0.6, 0.1 mM Isopropyl β-D-1-thiogalactopyranoside (IPTG) was added for induction, and followed by additional 4 h to 6 h incubation (48).

Additionally, the genes of OTA detoxification enzymes OTase, CP, PJ_1540, and AfOTase identified from other microbial strains with known polypeptide sequences were commercially synthesized by Sangon Biotech (Shanghai, China). Protein expressions of the synthesized genes were performed using the vector pGEX-4T-1 (for gene afotase), pET32a (for genes cp and otase), and pET28a (for gene pj_1540), the restriction endonucleases EcoRI and XhoI, and the primer pairs as Table 1, under previously described conditions (24, 26–28).

After protein expression, bacterial cells were disintegrated by an ultrasonicator and centrifuged at 10,000 × g for 10 min, and the crude enzyme for the expressed gene was produced from the supernatant by passing through a sterile filter (0.22 μm, Millipore, Billerica, MA, USA). OTA degradation activity of each recombinant protein was firstly evaluated by incubating the crude enzyme with OTA solution (PBS, pH 7.2), and the crude enzyme with OTA degradation activity was selected for further purification. The active recombinant proteins (i.e., ADH3, OTase, CP, PJ_1540, and AfOTase) were purified by affinity chromatography with GSTrap FF (GE Healthcare, Piscataway, NJ, USA) or nickel-nitrilotriacetic acid agarose resin (Ni–NTA, Qiagen, CA) according to manufacturer's instructions. The purity of recombinant protein was examined by SDS-PAGE (Bio-Rad, Hercules, CA, USA), and the protein concentration was evaluated by spectrophotometrically method with bovine serum albumin (BSA) as a standard (49).

OTA degradation by rADH3 was performed in 1 mL degradation mixture that contained 1.2 μg rADH3 protein and 50 ng OTA in PBS (pH 7.2), at 37°C for 90 s, and then the catalytic reaction was stopped by 3.0 mL acetonitrile. OTA residue was determined by HPLC. As a parallel test, 1.2 μg denatured rADH3 in PBS degradation mixture (pH 7.2) was used as the negative control. Moreover, the OTA degradation product by rADH3 was collected using the same procedures of degradation product from the strain CW117, and the analytical conditions of LC-MS/MS identification were consistent as above. The dynamics of OTA degradation and OTα production by rADH3 were performed in 5 mL PBS degradation mixture (pH 7.2) which contained 4.0 μg rADH3 protein (0.8 μg/mL) and 0.5 nM OTA (i.e., 0.1 μmol/L), and the degradation samples were collected at 0 s to 140 s with the interval of 20 s for OTA and OTα determination by HPLC. Here, 4.0 μg denatured rADH3 was selected for the negative control as parallel tests. Comparatively, the dynamics of OTA degradation and OTα production by rOTase and commercial enzymes of CPA (Catalogue no. C9268) and CPY (Catalogue no. C3888) from Sigma–Aldrich were evaluated by the same procedure, but the protein content for each commercial enzyme or rOTase was 75 μg (i.e., 15 μg/mL) and incubation times were 160 min for rOTase, 24 h for CPA, and 8 days for CPY.

The pH range and optimal pH of rADH3 were evaluated by adding 1.0 μg rADH3 protein to 1 mL OTA solution (final concentration 50 μg/L) that prepared by different pH buffers (pH 2.0 to 5.0, glycine-HCl buffer; pH 6.0 to 7.0, phosphate buffer; pH 8.0 to 9.0, Tris-HCl buffer; pH 10.0 to 11.0, glycine-NaOH buffer), and incubated at 37°C for 2 min. The temperature range and optimal temperature of rADH3 were evaluated by adding 1.0 μg rADH3 protein to 1 mL OTA solution (final concentration 50 μg/L, PBS, pH 7.2), and incubated at 0 to 100°C with 10°C as the interval for 2 min. In SDS and proteinase tests, the rADH3 was firstly pretreated by final concentration of 1% SDS, 1 mg/mL proteinase K, or 1% SDS plus 1 mg/mL proteinase K, at 37°C for 6 h, respectively. After that, the degradation activity of the pretreated rADH3 was determined as above at 37°C for 2 min. The PBS with corresponding denaturant or untreated rADH3 was selected as negative and positive control. In metal-chelator tests, ADH3 was first pretreated by 0.1 mol/L EDTA (EDTA) or ethylene glycol-bis(beta-aminoethyl ether)-N-tetraacetic acid (EGTA) at 37°C for 6 h. After that, the degradation activity of the pretreated rADH3 was determined as above at 37°C for 2 min. The PBS with corresponding metal-chelator or untreated rADH3 was selected as negative and positive control. The metal ions effect on ADH3 was evaluated by adding 1.0 μg rADH3 protein to 1.0 mL OTA solution (final concentration 50 μg/L) that contained 0.01 mol/L metal ion (i.e., Li⁺, Ca²⁺, Mg²⁺, Fe²⁺, Zn²⁺, or Cu²⁺), and the degradation tests were performed at 37°C for 2 min. The OTA degradation mixture without any metal ion was selected the control. The characterization on mainstream commercial enzyme CPA was evaluated by the same procedure, but the protein content for CPA was 15 μg/mL and incubation time was 48 h.

During the kinetic constants determination, enzymatic reactions were carried out under the optimal conditions of 1 min for ADH3, 20 min for OTase and AfOTase, 1 h for CPA, 36 h for CP, 80 h for PJ_1540, and 136 h for CPY at appropriate substrate concentrations (i.e., 30-to 90 μg/L OTA), respectively. The degradation product OTα was cleaned using the OchraTest immunoaffinity column (Vicam, Milford, MA, USA) and determined by HPLC following the method as above. The reaction rate versus the substrate concentration was plotted, then the kinetic constants of Kcat, Km (including Kcat/Km), and V_max were calculated as Stander et al. by a nonlinear regression of the Michaelis-Menten equation with GraphPad PRISM version 5.0 (GraphPad Software, La Jolla, CA) (23).

The homology models of ADH3, AfOTase, and OTase were modeled from known crystal structures 2QS8 (29), 6SLF (30), and 4C5Y (24), respectively. A short-time (10 ps) MD simulation was performed to obtain the optimized structure model for SAS assay. MD simulation of enzymes was performed to evaluate the stability and conformational changes of these enzymes. All MD simulations were performed using Amber99SB-ILDN force field (50) using GROMACS 5.1 package (51), and running on high-performance Linux system (National Supercomputing Center in Shenzhen). During MD simulations, all the systems were solvated using TIP3P (52) water model in a periodic box, followed by addition Na⁺ or Cl^– ions to neutralize the systems. Before MD simulations, energy minimization, NVT (canonical ensemble), and NPT (isothermal-isobaric ensemble) equilibration were performed. Finally, MD simulations were run for 20 ns time scale under constant temperature (300 K) and pressure (1 atm). The catalytic sites predicted from MD simulations were substituted with glycine using the point mutation method (53), respectively. Each gene mutant was expressed using the vector pGEX-4T-1 and E. coli BL21, and OTA degradation activity of mutated protein was determined as ADH3.