Genome-Guided Identification of an OTA-Degrading Amidohydrolase AMH2102 from Acinetobacter kookii AK4 with Enhanced Soluble Expression in Escherichia coli

TL;DR

Researchers identified a highly efficient OTA-degrading enzyme from Acinetobacter kookii and improved its expression in E. coli for rapid toxin removal.

Contribution

A novel OTA-degrading amidohydrolase AMH2102 was identified and its soluble expression was significantly enhanced in E. coli.

Findings

Acinetobacter kookii AK4 degraded 95.44% of OTA within 6 hours.

Codon optimization and SUMO tagging increased AMH2102 soluble expression 14.81-fold.

AMH2102 achieved 100% OTA degradation within 3 minutes.

Abstract

Ochratoxin A (OTA) is a globally distributed mycotoxin that poses serious threats to food safety and human health due to its nephrotoxic, hepatotoxic, and carcinogenic properties. Previous enzymatic detoxification strategies for OTA have been constrained by low degradation efficiency or poor soluble expression of highly active enzymes. In this study, a bacterial strain with strong OTA-degrading activity was isolated and identified as Acinetobacter kookii AK4, which degraded 95.44% of 1 μg/mL OTA within 6 h. The predominant OTA-degrading activity was derived from intracellular enzymes. Through genome mining and experimental validation, gene2102 was identified as encoding an amidohydrolase. The enzyme was designated AMH2102 and was heterologously expressed in Escherichia coli. Codon optimization combined with fusion of an N-terminal SUMO tag increased the soluble expression of AMH2102 by 14.81-fold, enabling complete (100%) OTA degradation within 3 min. Overall, this study achieved the identification of an efficient OTA-degrading strain and enzyme and explored strategies for improving enzyme expression, yielding effective outcomes that provide useful references for future studies on strain mining and enzyme engineering.

Figures

Full text

1. Introduction

Ochratoxin A (OTA) is a Group 2B carcinogen with pronounced nephrotoxicity and has also been reported to exert hepatotoxic [1], teratogenic [2], neurotoxic [3], and immunosuppressive effects [4]. Owing to its efficient absorption and slow elimination, OTA can persist in the body and cause prolonged systemic exposure in animals [5], thereby triggering oxidative stress and multi-organ pathological damage, particularly in the liver and kidneys [6]. In addition, recent studies have shown that OTA can impair muscle growth by inhibiting myogenesis and inducing ferroptosis, further highlighting its broad spectrum of health hazards in animals [7]. OTA is widely detected in diverse food and feed matrices, including cereal products, grape-derived products, coffee, cocoa, dairy products, and meat products, as well as compound feeds and feed ingredients [8]. Its dispersed contamination sources and high stability under conventional food- and feed-processing conditions make OTA difficult to eliminate once it enters the production chain, and its resistance to thermal processing and transformation into multiple analogs further complicate effective control [9]. Although physical and chemical control strategies can reduce OTA levels under specific conditions, their application is often constrained by concerns related to product quality preservation, process compatibility, and environmental sustainability [10]. In contrast, biological detoxification approaches based on microorganisms or enzymatic reactions, which operate under mild conditions and exhibit high specificity, are increasingly regarded as more compatible with the requirements of modern food and feed safety control.

Recent studies have demonstrated that microorganisms from diverse taxonomic groups, including bacteria, yeasts, and molds, are involved in OTA detoxification, with reported genera such as Lysobacter [11], Bacillus [12], Acinetobacter [13], Brevundimonas [14], Stenotrophomonas [15], Saccharomyces [16], Cryptococcus [17], Aspergillus [18], and Metarhizium [19]. OTA degradation by Acinetobacter species has been documented since 1994 [20], and strains with confirmed OTA-degrading capability include A. calcoaceticus [20], A. baumannii [13], A. pittii [21], Acinetobacter sp. neg1 [22], and A. tandoii [23]. Numerous OTA-degrading enzyme resources have been identified from Acinetobacter species. In A. pittii AP19, transcriptomic analysis identified the carboxypeptidase encoded by dacC as a major contributor to OTA degradation [21]. An amidohydrolase, Amse, mined from A. baumannii, was reported to achieve 93% OTA degradation within 5 min [13]. An α/β-hydrolase ABH from A. tandoii exhibited activity toward OTA, its derivative ochratoxin B (OTB), and other carboxypeptidase substrates [23].

At the enzymatic level, representative hydrolases such as carboxypeptidases and amidohydrolases have been characterized to catalyze the cleavage of the amide bond linking the phenylalanine moiety to the isocoumarin structure. While enzymatic OTA degradation has been demonstrated to be feasible, many reported enzymes exhibit only moderate catalytic efficiency. For instance, the D-Ala-D-Ala carboxypeptidase DacA from Bacillus subtilis CW14 degraded only 71.3% of OTA after 24 h [24], and a recombinant ochratoxinase from Aspergillus niger achieved 50% degradation of 50 μg/kg OTA within 1 h, whereas carboxypeptidase A degraded merely 8% under the same conditions [25]. In addition to limited catalytic efficiency, some highly active OTA-degrading enzymes face substantial challenges in heterologous expression and soluble production. A representative example is the high-efficiency amidohydrolase ADH3, which can completely degrade 0.05 μg/mL OTA within 90 s [15] yet exhibits a soluble expression level of less than 5% when expressed in Escherichia coli [26]. Soluble expression is a major bottleneck in recombinant enzyme production, particularly for OTA-degrading enzymes, as excessive expression in E. coli often leads to inactive inclusion bodies, making continuous optimization strategies essential. Collectively, these limitations highlight the urgent need to develop feasible strategies to improve soluble heterologous expression, thereby facilitating the practical application of enzymatic OTA detoxification.

In this study, an OTA-degrading Acinetobacter kookii strain was investigated, and a highly efficient amidohydrolase, AMH2102, was identified through genome-based screening and experimental validation. Furthermore, the soluble expression of AMH2102 was enhanced by codon optimization and the introduction of solubility tags. Collectively, this work presents an initial attempt to improve soluble expression in the context of OTA-degrading enzyme exploration, providing a basis for further optimization.

2. Results

2.1. Screening of AK4 Strains

Soil samples suspended in sterile saline were inoculated into LB medium for enrichment. After 24 h of incubation, the cultures were centrifuged, resuspended in saline, and incubated with OTA, revealing detectable OTA-degrading activity. The enriched cultures were then serially diluted and plated on LB agar, followed by repeated streaking to obtain pure isolates. The OTA-degrading ability of each isolate was subsequently evaluated using cell suspensions incubated with 1 μg/mL OTA. In total, thirteen OTA-degrading bacterial strains were obtained, among which strain AK4 exhibited the highest degradation efficiency and was therefore selected for further characterization.

After 24 h of cultivation on LB solid plates, strain AK4 formed pale yellow circular colonies with a smooth and shiny surface and neat and smooth edges; the colonies were opaque and easy to pick (Figure 1A). Scanning electron microscopy (SEM) observations revealed that the cells were overall oval or short ellipsoidal in shape, approximately 0.5 × 1 μm in size, and exhibited typical binary fission characteristics (Figure 1C,D). Meanwhile, AK4 is a Gram-negative (Figure 1B), coccobacillary, obligate aerobic bacterium.

BLAST analysis revealed that the 16S rRNA gene sequence (1474 bp) of AK4 shares 99.33% similarity with that of Acinetobacter kookii 11-0202 (NR135727.1). The phylogenetic tree constructed based on the 16S rRNA sequence placed AK4 within the genus Acinetobacter, clustering it with Acinetobacter kookii, indicating a close phylogenetic relationship between the two (Figure 1E). To further clarify its species-level taxonomic status, a multi-gene phylogenetic tree was constructed using the housekeeping genes infC, tsf, pgk, and rpoB. The infC, tsf, pgk, and rpoB genes encode translation initiation factor IF-3, elongation factor Ts, phosphoglycerate kinase, and the β subunit of RNA polymerase, respectively. These genes are highly conserved in bacteria and are involved in essential cellular processes, including protein synthesis, energy metabolism, and transcription. The results similarly showed AK4 clustering with A. kookii (Figure 1E). Through TCS of whole-genome data, the results revealed a similarity of 99.88% between AK4 and Acinetobacter kookii (Table S1). Therefore, strain AK4 was identified as Acinetobacter kookii.

2.2. Determination of the Degradation Ability of the AK4 Strain

The degradative capacity of AK4 toward OTA was evaluated with regard to different OTA concentrations, incubation times and treatments.

According to GB 2761-2017 Maximum Levels of Mycotoxins in Foods [27] and GB 13078-2017 Hygienic Standards for Feeds [28], the regulatory limits for OTA in food and feed are 2, 5, 10 and 100 μg/kg (Table S2). To simulate higher contamination scenarios, these limits were scaled up, and OTA concentrations of 200, 500, and 1000 μg/L were tested. Within 6 h, AK4 live cell suspensions achieved degradation ratios of 99.89% and 98.97% at 200 and 500 μg/L, respectively (Figure 2A). Although the degradation efficiency declined slightly at higher OTA concentrations, AK4 retained strong activity, indicating that its degradation system remains effective across a broad concentration range.

The degradation ability of live cell suspensions toward 1 μg/mL OTA was examined using TLC at 6 h and 12 h, showing that OTA was below the detection limit after 6 h. (Figure S1) Subsequently, the time-dependent degradation of OTA by live cell suspensions was analyzed using HPLC from 0 to 6 h, showing 95.44% removal after 6 h (Figure 2B). Meanwhile, OTα accumulated with time but to a lesser extent than the reduction of OTA (Figure 2C), implying further transformation or partial adsorption.

The degradation of OTA (1 μg/mL) by different treatments of AK4 was evaluated (Figure 2D–F). The cell-free supernatant exhibited only weak activity, indicating that extracellular components played a negligible role. Heat-inactivated cells removed merely 2.89% of OTA (Figure 2D), suggesting slight adsorption by the cell wall. In contrast, the cell lysate and its soluble and insoluble fractions were able to completely degrade OTA. However, the activity was abolished after treatment at 100 °C, and most of the degradation capacity was retained in the >10 kDa soluble fraction (Figure 2E), indicating that the active component is a heat-sensitive protein. Inhibitor assays showed that SDS markedly suppressed the degradation activity, while treatments with proteinase K, PMSF, and EDTA had minimal effects (Figure 2F). Overall, it can be inferred that the degradation ability of AK4 is mainly attributed to a non-metal-dependent intracellular protein.

2.3. Mining and Analysis of Potential OTA-Degrading Enzyme Genes

The complete genome of AK4 comprises 3,281,503 bp, including a circular chromosome (3,216,507 bp) and a plasmid (64,996 bp), with GC contents of 42.96% and 37.25%, respectively. A total of 3037 coding sequences (average length 916.76 bp) were identified, representing 84.85% of the genome (Table 1). In the COG classification, AK4 showed distinctive gene distribution, with a high proportion of translation and ribosome-related genes (J), indicating strong protein synthesis potential. Enrichment of cell wall/membrane/envelope biogenesis (M) genes suggests structural stability and environmental tolerance, while amino acid (E) and lipid metabolism (I) genes highlight flexibility in substrate utilization and energy conversion. Notably, numerous genes of unknown function (S) imply potential novel pathways (Figure 3B). GO annotation supported these findings, showing enrichment in DNA binding, ATP binding, and catalytic activities such as hydrolases, transferases, and oxidoreductases, along with significant representation of cytoplasm, plasma membrane, and ribosome components, emphasizing its metabolic activity and protein synthesis capacity (Figure 3C). KEGG pathway analysis further revealed that metabolic pathways dominated, particularly amino acid, carbohydrate, cofactor, and energy metabolism, reflecting strong energy transformation capacity. In addition, pathways related to membrane transport, signal transduction, replication and repair, and protein folding were abundant, suggesting both environmental adaptability and robust genetic information processing (Figure 3D). Overall, these multi-dimensional annotations consistently demonstrate that AK4 combines efficient metabolism and protein synthesis with adaptive regulatory mechanisms and potential novel metabolic routes, providing a solid genetic foundation for its detoxification functions.

Based on the whole-genome data of strain AK4, annotation and screening against the NR database identified 18 proteases, 34 peptidases, 6 amidases, 3 amidohydrolases, 1 β-lactamase, 4 carboxypeptidases and 83 other hydrolases, resulting in 149 candidate enzyme genes. BLASTP was then used to compare the proteins encoded by these candidate genes with previously reported OTA-degrading enzymes (Table S3). Entries with query coverage greater than 50% were retained, yielding 10 potential candidate enzyme genes. Functional annotations of these proteins in the NR database included 2 carboxypeptidases, 1 amidohydrolase, 1 endopeptidase, 1 metallopeptidase, 4 members of the α/β-hydrolase family, and 1 NUDIX hydrolase; the sequence identities between the candidate genes and the reference enzymes ranged from approximately 22% to 89%. (Table S4). Among them, the proteins encoded by gene2018 and gene2102 showed amino acid sequence identities of 89% and 74.49%, respectively, to known OTA-degrading enzymes, both of which were derived from Acinetobacter species [13,21,22].

2.4. Heterologous Expression of Proteins Encoded by 10 Candidate Enzyme Genes in Escherichia coli

Ten candidate OTA-degrading enzyme genes were amplified from the genome of strain AK4 and individually cloned into the Escherichia coli expression vector pET-28a(+) using either double-enzyme digestion or homologous recombination. Restriction analysis and DNA sequencing confirmed the successful construction of all recombinant plasmids (Figure 4A). These expression plasmids were subsequently transformed into E. coli BL21(DE3), and positive clones were selected and cultured in shake flasks to induce protein expression. After ultrasonic disruption of the harvested cells, crude enzyme preparations were obtained from the lysates and evaluated for their ability to degrade OTA within 24 h. The results showed that among all candidates, only the protein encoded by gene2102 exhibited OTA-degrading activity (Figure 4B), and this protein was designated AMH2102. Further enzymatic analysis revealed that the crude enzyme of AMH2102 could degrade 1 μg/mL OTA within 15 min, with a degradation ratio of 99.96% (Figure 4C), demonstrating remarkably high catalytic efficiency.

2.5. Optimize the Soluble Expression Level of AMH2102 in Escherichia coli

In preliminary expression experiments, AMH2102 was expressed at a relatively low level in Escherichia coli, with only weak target protein signals detected by Western blot analysis (Figure 4B). This limited expression prompted us to further optimize the expression strategy to improve the effective soluble production of AMH2102. We employed a series of optimization strategies, including codon optimization, fusion of a SUMO solubility tag, and selection of different expression hosts. Based on the codon preference of E. coli, the codon adaptation index (CAI) value of gene2102 was improved from 0.73 to 0.82, and the GC content was increased from 46.98% to 53.81%. The CAI value reflects the degree of codon usage compatibility between a target gene and the host organism. The original gene2102 sequence was designated as A, while the optimized version was referred to as B, which was subsequently cloned into the pET-28a(+) vector. A SUMO tag was further fused to the N-terminus of the target gene to promote protein solubility. Two expression strains, BL21(DE3) and Rosetta(DE3), were used to evaluate expression performance, and the overall optimization strategy is illustrated in Figure 5A. The results showed that both codon optimization alone and the combined strategy of codon optimization with SUMO tagging markedly increased the soluble expression of AMH2102 (Figure 5B), yielding 2.66-fold and 14.81-fold improvements, respectively, compared with A-BL21(DE3). No significant difference in soluble expression was observed between BL21(DE3) and Rosetta(DE3) (Figure 5C). Notably, among all tested constructs, the crude enzyme obtained from BL21(DE3) expressing SUMO-tagged, codon-optimized AMH2102 (gene B) achieved complete (100%) degradation of OTA within 3 min (Figure 5D), demonstrating excellent catalytic efficiency.

3. Discussion

Multiple studies have shown that Acinetobacter species are capable of transforming not only OTA but also other major mycotoxins, including zearalenone [29,30,31], aflatoxin B1 [29], and deoxynivalenol [32]. Beyond their role in mycotoxin removal, Acinetobacter species have been widely reported to participate in the biodegradation and bioremediation of diverse environmental contaminants, including petroleum hydrocarbons, aromatic compounds, phenolic pollutants, and pharmaceutical residues in aquatic and terrestrial ecosystems [33,34,35,36,37], highlighting the remarkable metabolic adaptability of this genus and its promise as a microbial resource for bioremediation and detoxification applications. To our knowledge, this study provides the first report of efficient OTA-degrading activity in A. kookii. Notably, strain AK4 exhibited markedly superior OTA-degrading efficiency, achieving 95.44% degradation within 6 h, whereas previously reported Acinetobacter strains typically require one day to several days to achieve partial or complete OTA degradation at comparable concentrations [13,21,22].

A. kookii was first described in 2013 following its isolation from soil [38] and was subsequently reported in a clinical case involving a giraffe calf with severe polyarthritis, leading to speculation that this species may represent an emerging opportunistic pathogen [39], although its clinical relevance remains poorly defined. In the present study, strain AK4 did not display apparent antibiotic resistance under the tested conditions (Figure S2), indicating that antibiotic susceptibility may vary among A. kookii strains and that clinical risk cannot be inferred solely from species-level classification. Notably, several members of the genus Acinetobacter, including A. baumannii and A. calcoaceticus, are well-recognized opportunistic pathogens, underscoring the necessity for a cautious evaluation of biosafety when considering applications related to food or feed. From this perspective, the direct use of live bacterial strains may be less appropriate, whereas the identification and heterologous production of functional OTA-degrading enzymes, followed by rigorous safety assessment, represent a more feasible and responsible strategy for practical detoxification applications.

Traditionally, OTA-degrading enzymes have often been identified through activity-guided purification strategies involving sequential steps such as salting out, organic solvent precipitation, ion-exchange chromatography, and size-exclusion chromatography. This classical approach has proven effective in directly linking enzymatic activity to specific proteins and has contributed substantially to the identification of several OTA-degrading enzymes [25,40,41,42]. However, such purification-based strategies are inherently dependent on the abundance and stability of target enzymes in their native hosts and are often labor-intensive and time-consuming [43]. In addition, the requirement for relatively high enzymatic activity during each purification step may limit their applicability for systematically exploring multiple potential candidates. In recent years, increasing efforts have been devoted to the discovery of mycotoxin-degrading enzymes by integrating genome analysis with transcriptomic profiling, proteomic identification, and other omics-based approaches, enabling a more systematic exploration of enzymatic resources [19,21,24,44]. Compared with traditional activity-guided purification strategies, genome-based enzyme mining offers a complementary route for identifying potential OTA-degrading enzymes at the genetic level, independent of their native expression abundance. When coupled with heterologous expression and functional validation, this strategy allows the efficient screening of multiple candidate enzymes and broadens the accessible pool of OTA-degrading resources. Notably, recent studies have also reported the application of machine learning-assisted approaches for mining OTA-degrading enzymes from sequence databases, further highlighting the diversification of enzyme discovery strategies [45].

Although many OTA-degrading enzymes have been identified, only a few have been translated into practical applications. Existing commercial OTA-degrading enzyme preparations, such as the amidase ADH3 from Stenotrophomonas acidaminiphila CW117, demonstrate the feasibility of enzymatic detoxification [15]. The successful application of these products indicates that factors including soluble expression, enzyme yield, and production feasibility are critical for industrial application [46]. Accordingly, improving soluble heterologous expression, as achieved in this study, represents an important step toward bridging enzyme discovery and practical utilization.

Soluble expression is a major bottleneck in recombinant protein production, especially for enzymes, as overexpression in Escherichia coli often leads to inactive inclusion bodies. Optimization strategies, such as codon optimization and fusion with solubility-enhancing tags, are therefore commonly employed. Similar approaches have successfully improved soluble expression of other mycotoxin-degrading enzymes. For example, codon optimization enhanced the lactonohydrolase ZHD101 [47], and codon optimization combined with cold-shock-induced expression facilitated soluble expression of PsMnp while reducing inclusion bodies [48]. Building on these precedents, the present study employed a combination of codon optimization, fusion of a solubility tag, and modification of the expression host to systematically improve the originally weak expression of the target enzyme in E. coli. Codon optimization alone increased soluble expression by 2.66-fold, whereas a dual optimization strategy combining codon optimization with solubility tag fusion resulted in a marked 14.81-fold enhancement in soluble protein yield. These results were confirmed by Western blot analysis (Figure 5C) and activity assays of the crude enzyme (Figure 5D). To our knowledge, this is the first report in the OTA-degrading enzyme field demonstrating improved soluble expression through fusion tag application. Nevertheless, further optimization is required to achieve efficient soluble production. Future efforts will explore alternative expression systems, such as yeast and vector engineering, to further improve expression and support potential applications.

The present study has several limitations that warrant further investigation. This work focused on improving the expression level of soluble proteins during expression in E. coli, and therefore the evaluation was mainly conducted at the crude enzyme level. Subsequent enzymatic characterization based on purified proteins will be necessary. Moreover, considering the ultimate application of soluble production, further evaluation of its safety and detoxification performance in real food or feed matrices will be an important direction for future research. Furthermore, it remains to be determined whether auxiliary cofactors, coenzymes, or intrinsic regulatory mechanisms within the AK4 strain contribute to enzyme activity. Comprehensive investigation of these factors will be essential for a deeper understanding of the enzyme’s functional mechanism and for guiding further optimization.

4. Conclusions

In this study, an OTA-degrading strain, Acinetobacter kookii AK4, and its key amidohydrolase gene, gene2102, were identified and characterized. The encoded enzyme AMH2102 was successfully heterologously expressed, and its soluble expression was markedly improved through codon optimization and N-terminal SUMO tagging, resulting in rapid OTA degradation. These results demonstrate an effective strategy for enhancing the soluble expression of OTA-degrading enzymes and support their potential application in enzymatic OTA detoxification.

5. Materials and Methods

5.1. Chemicals and Materials

Ochratoxin A (purity ≥ 99%) was purchased from Pribolab Pte., Ltd. (Qingdao, China). Yeast extract, tryptone and skimmed milk powder were supplied by Oxoid. (London, UK). Ethylenediaminetetraacetic acid (EDTA) and agar were purchased from Biofroxx GmbH (Einhausen, Germany). Gram stain sets were purchased from Beijing Land Bridge Technology Co., Ltd. (Beijing, China). Methanol, ethanol, ethyl acetate, formic acid, acetonitrile, sodium chloride and glycine were purchased from Macklin Inc. (Shanghai, China). Toluene, trichloromethane and hydrochloric acid were supplied by Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Phosphate-buffered saline (PBS) and SDS-PAGE Sample Loading Buffer were purchased from Beyotime Biotechnology Co., Ltd. (Shanghai, China). Phenylmethanesulfonyl fluoride (PMSF), kanamycin, glutaraldehyde, proteinase K and Isopropyl-β-D-thiogalactopyranoside (IPTG) were purchased from Solarbio Science & Technology Co., Ltd. (Beijing, China). Tris was purchased from Biotopped Technology Co., Ltd. (Beijing, China). Sodium dodecyl sulfate (SDS) was supplied by VWR Corporation. (Radnor, PA, USA). SuperKine™ West Pico PLUS Chemiluminescent Substrate (ECL) was purchased from Abbkine Scientific Co., Ltd. (Wuhan, China). TIANamp Bacteria DNA Kit and EndoFree Mini Plasmid Kit were purchased from Tiangen Biochemical Technology Co., Ltd. (Beijing, China). FastPure Gel DNA Extraction Mini Kit, 2 × Rapid Taq Master Mix, 2 × Phanta Flash Master Mix, SwiftCut BamH I, SwiftCut Xho I, SwiftCut EcoR I, SwiftCut Sal I, ClonExpress Ultra One Step Cloning Kit V3 and T4 DNA Ligase were purchased from Vazyme Biotech Co., Ltd. (Nanjing, China). Plasmid pET-28a(+) was purchased from Miaoling Biotech Co., Ltd. (Wuhan, China). Anti-His Tag Monoclonal Antibody (MS IgG1) and HRP-Goat Anti-Mouse IgG (H+L) were purchased from Biorigin Biotech Co., Ltd. (Beijing, China). Silica gel plates for thin-layer chromatography (G type) were provided by Dingkang Silica Co., Ltd. (Qingdao, China).

5.2. Culture Conditions of Strains

The AK4 strain was isolated from wheat farmland soil in Shijiazhuang City, Hebei Province, China, and selected from multiple isolates based on its OTA-degrading capacity by our laboratory. The Escherichia coli DH5α strain and BL21(DE3) strain were provided by Vazyme Biotech Co., Ltd. (Nanjing, China). The E. coli Rosetta (DE3) Chemically Competent Cell was purchased from Coolaber Technology Co., Ltd. (Beijing, China). Bacterial cultivation was carried out using Luria–Bertani (LB) medium, composed of tryptone (10 g/L), yeast extract (5 g/L), and NaCl (10 g/L). For plate preparation, agar was incorporated into the LB formulation at a final concentration of 15 g/L.

5.3. Mining of OTA-Degrading Strains

The methodology for mining OTA-degrading strains was referenced from Peng’s work with minor modifications [14]. Specifically, soil samples were collected from approximately 5 cm below the surface of a wheat field and suspended in sterile saline. After shaking at 37 °C and 200 rpm for 1 h to release microbial cells or dormant forms, the suspension was allowed to settle. The supernatant was inoculated into LB broth and incubated at 37 °C with shaking for 24 h for enrichment. Following enrichment, the culture was centrifuged at 6000 rpm for 10 min, and the pellet was resuspended in phosphate-buffered saline (PBS). The cell suspension was incubated with OTA at a final concentration of 1 μg/mL at 37 °C for 24 h. The OTA-degrading activity of the culture was assessed using thin-layer chromatography (TLC). To obtain pure OTA-degrading isolates, the active culture was subjected to serial dilution and plated on LB agar. Colonies were screened for OTA-degrading activity by TLC, and those showing activity were repeatedly streaked on fresh LB plates until morphologically uniform single colonies were obtained. These purified isolates were then used for subsequent experiments.

5.4. Identification of AK4 Strains

The genus and species of OTA-degrading strains were identified through colony morphology observation, microscopic examination after Gram staining, scanning electron microscope (SEM) observation, 16S rRNA and housekeeping gene sequencing, and whole-genome comparison.

Colony morphology observation: The purified strain was spread on LB solid plates and incubated upside down at 37 °C for 24 h. The colony morphology and color were then observed.

Microscopic examination after Gram staining: The bacteria fixed on the glass slide were stained with crystal violet, followed by iodine solution, decolorized with 95% ethanol, counterstained with safranin, and observed using a 10× eyepiece and a 100× oil immersion lens.

Scanning Electron Microscope (SEM) observation: Centrifuge to collect bacterial cells, fix with 2.5% glutaraldehyde at 4 °C for 10 h, prepare slides, and examine under a scanning electron microscope. Sample preparation and observation were performed by Zhongke Baice Company.

16S rRNA and Housekeeping Gene-Based Phylogenetic Analysis: Genomic DNA was extracted from the target strain using a genomic DNA extraction kit. The 16S rRNA gene was amplified with primers 27F (5′-AGAGTTTGATCMTGGCTCAG-3′) and 1492R (5′-GGTTACCTTGTTACGACTT-3′), sequenced by Sangon Biotech (Shanghai), and analyzed by BLAST in the NCBI database. Sequence alignment of the 16S rRNA gene was performed using Muscle in MEGA 11.0, and a phylogenetic tree was constructed with the neighbor-joining method. Based on the whole-genome sequencing data, the sequences of four housekeeping genes (infC, tsf, pgk and rpoB) were obtained. After BLAST alignment, the four genes were concatenated using ACOPTools, and a multi-gene phylogenetic tree was constructed with MEGA 11.

Whole-genome comparison: A Tetra Correlation Search (TCS) was performed using JSpeciesWS (https://jspecies.ribohost.com/jspeciesws/#home, accessed on 19 September 2025) to compare the whole-genome sequence of strain AK4 with the reference genomes in the GenomesDB database [49].

5.5. OTA Degradation Capability Validation

The OTA-degrading activity of strains was initially screened by TLC [14]. After incubation with OTA, reaction mixtures were acidified and extracted with chloroform, and the organic phase was analyzed on silica gel TLC plates, where OTA appeared as a blue-green fluorescent spot (Rf ≈ 0.65) under UV light (365 nm). Quantitative analysis was subsequently performed by HPLC [14]. Acidified samples were extracted with chloroform, dried, redissolved in methanol, and analyzed using a C18 reversed-phase column (250 × 4.6 mm, 5 μm) under isocratic conditions with fluorescence detection (excitation = 333 nm, emission = 460 nm). The mobile phase consisted of acetonitrile containing 0.1% (v/v) formic acid (A) and water containing 0.1% (v/v) formic acid (B) with a flow rate of 1 mL/min (A: 0.7 mL/min; B: 0.3 mL/min). The OTA content of samples was determined based on peak area, and the degradation ratio was calculated by comparing the OTA content with that of the control group.

The OTA-degrading ability of strain AK4 was evaluated under different OTA concentrations, incubation times, and treatment conditions. Bacteria were cultured in LB medium (1%, v/v) at 37 °C, 180 rpm for 13 h, then centrifuged at 6000 rpm and resuspended in PBS to obtain a live cell suspension. OTA was added at final concentrations of 200, 500, and 1000 μg/L and incubated at 37 °C for 6 h. For time-dependent analysis, the suspension was incubated with 1 μg/mL OTA at 37 °C for 1–6 h. The cell-free supernatant was obtained by 0.22 μm filtration, and heat-inactivated cells were prepared by autoclaving at 121 °C for 20 min. Cell lysates were produced by ultrasonication, followed by centrifugation at 12,000 rpm (4 °C, 10 min) to obtain soluble and insoluble fractions, which were further heat-treated at 100 °C for 10 min. The soluble fraction was also separated using a 10 kDa ultrafiltration tube. Cell lysates were additionally treated with proteinase K (1 mg/mL), PMSF (1 mM), SDS (1%), and EDTA (1 mM). All samples were reacted with 1 μg/mL OTA at 37 °C for 6 h, and OTA degradation was analyzed by HPLC.

5.6. Complete Genome Analysis

Genomic DNA was extracted from the A. kooki strain, and whole-genome sequencing was performed using the PacBio Sequel IIe and Illumina sequencer (model: NovaSeq6000). The PacBio Sequel IIe raw data were HiFi reads. After quality control, the Illumina data and HiFi reads were mixed for hybrid de novo assembly to construct the genome assembly. This work was performed by Shanghai Meiji Biotechnology Co., Ltd.

The coding sequences (CDs) of chromosome and plasmid were predicted using Prodigal v2.6.3 [50] and GeneMarkS v4.3 [51], respectively. The tRNA-scan-SE v 2.0 software [52] (http://trna.ucsc.edu/software/, accessed on 15 March 2025) was used for tRNA prediction, and Barrnapv0.9 software (https://github.com/tseemann/barrnap, accessed on 15 March 2025) was used for rRNA prediction. The predicted CDs were annotated from NR, Swiss-Prot, Pfam, Gene Ontology (GO, http://www.geneontology.org/, accessed on 15 March 2025), Clusters of Orthologous Groups (COG, https://www.ncbi.nlm.nih.gov/COG/, accessed on 15 March 2025), and the Kyoto Encyclopedia of Genes and Genomes (KEGG, https://www.kegg.jp/, accessed on 15 March 2025) database using sequence alignment tools such as BLAST v2.3.0, Diamond v0.8.35 and HMMER v3.1b2. Briefly, each set of query proteins was aligned with the databases, and annotations of best-matched subjects (e-value < 10^−5^) were obtained for gene annotation. All of the analyses were performed using the online platform of Majorbio Cloud Platform (http://cloud.majorbio.com, accessed on 1 April 2025) from Shanghai Majorbio Bio-pharm Technology Co., Ltd.

5.7. Mining of Potential Degradative Enzyme Genes

Based on the whole-genome information, the AK4 genome was compared against the NR database to identify genes encoding hydrolases, proteases, peptidases, amidases, amidohydrolases, β-lactamases, and carboxypeptidases. The proteins encoded by the candidate genes were compared with reported OTA-degrading enzymes using BLASTP implemented in the NCBI online BLAST tool (https://blast.ncbi.nlm.nih.gov/Blast.cgi; accessed on 3 June 2025) with default NCBI parameters (BLOSUM62 matrix, E-value cutoff of 0.05), and proteins exhibiting a sequence coverage greater than 50% were selected as potential OTA-degrading genes. The reference enzymes used for BLASTP comparison are listed in Table S3 and comprise 20 enzymes, including 7 carboxypeptidases, 2 peptidases, and 11 hydrolases, which were previously reported in related studies [11,13,15,19,21,22,23,24,25,26,40,53,54,55,56,57]. The average length of the reference enzymes was 454.35 amino acids, whereas the average length of the candidate genes was 933.18 bp.

5.8. Construction of Escherichia coli Expression Vectors

Potential OTA-degrading genes were amplified from the genomic DNA of AK4 using specific primers (Table S5). PCR amplification was performed under the following conditions: initial denaturation at 95 °C for 5 min; 30 cycles of 94 °C for 30 s, 57 °C for 30 s, and 72 °C for 90 s; and a final extension at 72 °C for 10 min. The amplified products were verified by sequencing. Subsequently, the amplified DNA fragments and the pET-28a(+) vector were digested with the restriction endonucleases BamH I, Xho I, EcoR I, and Sal I, followed by ligation using T4 DNA ligase (16 °C, overnight) or 2× CE Mix V3 (50 °C, 15 min). The gene2102 sequence was codon-optimized, with the original and optimized sequences designated as A and B, respectively. The optimized gene B, as well as an N-terminal SUMO-tagged construct, was cloned into expression vectors by Sangon Biotech (Shanghai, China).

5.9. Expression and Verification of the Recombinant Enzyme

The recombinant plasmids were introduced into Escherichia coli DH5α competent cells by heat-shock transformation and selected on LB agar plates supplemented with kanamycin (50 μg/mL). Positive transformants were identified by colony PCR, and the inserted fragments were confirmed by sequencing. Validated plasmids were extracted and transformed into E. coli BL21(DE3) or E. coli Rosetta(DE3) for heterologous expression. The transformed cells were cultured in LB medium at 37 °C and 180 rpm until OD595 reached 0.6–0.8, followed by induction with 0.1 mM IPTG at 16 °C for 12 h. The induced cells were collected by centrifugation (12,000 rpm, 4 °C, 10 min), washed twice with Tris-HCl buffer (pH 7.4), and lysed by ultrasonication on ice. The lysate was centrifuged at 12,000 rpm for 20 min at 4 °C, and the supernatant was collected as the soluble fraction (crude enzyme, the crude cell lysate containing unpurified enzymes).

Protein expression and solubility were analyzed by SDS-PAGE and Western blotting. The OTA-degrading activity of the crude enzyme was evaluated by TLC and HPLC analyses. SDS-PAGE and Western blot analyses were performed with slight modifications based on the method described by Ming et al [58]. Briefly, protein samples were mixed with 5×SDS-PAGE sample loading buffer and boiled at 95 °C for 10 min. Electrophoresis was carried out in 1× SDS-PAGE electrophoresis buffer with Tris-Gly. The gel was subsequently stained with Coomassie Brilliant Blue R-250 for 1 h and destained until the background was clear. For Western blotting, proteins separated by electrophoresis were transferred onto an activated PVDF membrane under ice-cold conditions. The membrane was sealed with 5% (w/v) skim milk powder at room temperature for 1 h, followed by incubation with the primary antibody (Anti-His Tag Monoclonal Antibody, 1:2000) at 4 °C overnight. After washing 3 times with TBST, the membrane was incubated with the secondary antibody (HRP-labeled Goat Anti-Mouse IgG (H+L), 1:100,000) at room temperature for 1 h. The membrane was then washed 3 times with TBST. After incubating the membrane with ECL in a darkroom, the protein bands were visualized using a chemiluminescence imaging system. The expression levels of the target proteins were analyzed by densitometric quantification of Western blot band intensities using ImageJ (1.54p) software within the expected molecular weight regions.

5.10. Statistical Analysis

Data are expressed as mean ± SD. Differences among groups were evaluated by one-way ANOVA followed by Dunnett’s test, with p < 0.001 considered statistically significant. Statistical analyses were performed using GraphPad Prism 9.0 (GraphPad Software, San Diego, CA, USA).