Screening and Characterization of a Chryseobacterium timonianum Strain with Aflatoxin B1 Removal Ability Open Access

Aflatoxins belong to the family of structurally related mycotoxins secreted by Aspergillus flavus and A. Parasiticus. The four main aflatoxins are B1, B2, G1, and G2. The letters represent blue or green ultraviolet (UV) fluorescence, while the numbers represent their migration distance across a thin-layer chromatographic plate [1]. These secondary metabolites are difuranocoumarin derivatives produced by a common polyketide pathway involving seventeen genes encoding twelve enzymatic conversions [2, 3].

AFB1 has a core structure comprising dihydrofuro [2,3-b] furan and a coumarin ring, with the key toxic component being the furan ring’s double bond. In the liver, this double bond (C-8/C-9) gets converted to AFB1-8,9-epoxide, forming highly toxic adducts with DNA, glutathione S-transferase, or N7 guanine [4]. The International Agency for Research on Cancer (IARC) has announced aflatoxins as Class 1 carcinogens, which are highly poisonous toxic substances. Among the aflatoxins, aflatoxin B1 (AFB1) is the most potent. The Food and Agriculture Organization (FAO) of the United Nations has estimated that about 25% or more of the world’s food crops are affected and destroyed by AFB1 contamination each year.

Aspergillus species thrive in humid, warm tropical and subtropical areas where aflatoxin destroys crops. During preharvest to storage, aflatoxin can contaminate peanuts, corn, wheat, and soybeans [3]. AFB1 metabolism and toxicity in livestock depend on species, age, sex, feed mix, dosage, and exposure time. Resulting conditions include induced immunosuppression with poor cellular and humoral immunity, liver bleeding, renal necrosis, pancreatic cancer, and decreased body weight gain, milk and egg production [5]. Thus, AFB1-mediated livestock toxicity and further AFB1 carryover can be restricted by physical, chemical, or biological means. AFB1 can be physically adsorbed by bentonite or yeast cell walls [6] or chemically degraded by ozone, hydrogen peroxide, acidic electrolyzed water solution, and gamma and UV radiation [7‒10]. Existing physical and chemical methods do not fully meet commercial productivity, safety, and cost standards [11]. Research indicates that AFB1-metabolizing bacteria or their derived metabolites effectively eliminate AFB1 [12‒17]. However, this approach, while relatively safe, may yield toxic bio-transformed metabolites, necessitating assessment of AFB1 biodegradation and metabolite profiles.

Coumarin has been used as a carbon source in selective media for screening AFB1-removing bacteria [18]. In this study, we employed a coumarin-selective medium to isolate AFB1-removing bacteria. Bacterial strains capable of growing in a coumarin-selective medium were selected for further analysis. Our laboratory-isolated bacterium, the c4a strain, was included in the screening because its genome contains important AFB1-degrading enzyme genes, including multicopper oxidase (NCBI accession number: WP_228452318) and DyP-peroxidase (NCBI accession number: WP_160139647). Given its high growth rate in the coumarin medium, we investigated whether the c4a strain can metabolize AFB1. The c4a strain demonstrated the highest AFB1 degradation among the selected coumarin-degrading strains. The c4a strain was identified as C. timonianum through physiologically, biochemically, and phylogenetically. The AFB1 removal activity of the c4a strain was optimized by adjusting the medium composition, reaction temperature, pH, and metal ion supplementation. The final transformed metabolites were subjected to LC-MS/MS analysis to determine possible structural changes in the AFB1 molecule, followed by a comparative toxicity study. Here, we report a novel, highly thermostable, extracellular AFB1 removal phenomenon by the C. timonianum c4a strain, resulting in structural modification of the furan ring of AFB1, which is linked to its toxicity.

Luria-Bertani (LB) broth, nutrient broth, tryptone-yeast extract (TYE) broth, and M9 medium were purchased from Acumedia (Lansing, MI, USA), while coumarin and AFB1 were purchased from Sigma-Aldrich, Inc. (St. Louis, MO, USA). A coumarin medium, which contained coumarin as the sole carbon source, was prepared according to the method described by Guan et al. [18]. Each liter of coumarin medium contained 10.0 g coumarin, 0.25 g KH₂PO₄, 1.0 g NH₄NO₃, 1.0 g CaCl₂, 0.25 g MgSO₄, 1.0 mg FeSO₄, and 15.0 g agar, while the pH was adjusted to 7.0.

Forty Arbor Acres broilers (1 day old) were randomly divided into two groups: group A birds were not fed mycotoxin, and group B birds were given feed containing 200 ppb of AFB1. On day 35, the cecal digesta samples of one bird from group A and two from group B were subjected to bacterial screening. A soil sample enriched with chicken feathers obtained from the experimental farm of National Taiwan University (Taipei, Taiwan) was also used for bacterial screening.

The cecal digesta and soil samples (1 g each) were homogenized with 9 mL of sterile phosphate-buffered saline (PBS buffer, 0.1 m, pH 7.0) and serially diluted from 10⁻¹–10⁻⁴ fold before plating onto coumarin agar. The plates were incubated at 37°C aerobically and anaerobically for 4–7 days until colonies appeared. The colonies were transferred to new coumarin plates three consecutive times, following which single colonies were inoculated in coumarin broth and cultured at 37°C and 220 rpm, with bacterial growth monitored using a spectrophotometer at 600 nm. All isolated bacterial strains were further determined for their AFB1 removal activity.

Overnight cultures of the isolated bacterial strains were inoculated into 10 mL sterile LB broth containing 1 ppm of AFB1 and incubated at 37°C at 220 rpm agitation for 72 h. A 450 μL aliquot of culture media was centrifuged at 10,000 g for 10 min at 4°C, and the amount of residual AFB1 in the LB broth was analyzed by high-performance liquid chromatography (HPLC). The bacterial strain, c4a, showing the maximum AFB1 biodegradation, was subjected to further characterization studies.

A single colony of the c4a strain was inoculated into 3 mL LB broth and incubated at 30°C with 220 rpm agitation. After 7.5 h, the exponentially growing culture was added to 10 mL of sterile LB broth to obtain a final bacterial concentration of 1 × 10⁹ CFU/mL. After 24 h under the same incubation regime, extracellular supernatant and cell pellets were separately obtained by centrifugation at 13,000 g for 10 min at 4°C. The cell pellet was washed twice with sterile PBS (0.1 m, pH 7.0), resuspended in 5 mL of sterile PBS, and sonicated on ice for 10 min with an ultrasonicator (Model XL, Misonix, Farmingdale, NY, USA). The intracellular extracts and cell wall fractions were obtained by centrifuging the sonicated cells at 13,000 g for 10 min at 4°C. The cell wall fraction was redissolved in sterile PBS with 1 ppm AFB1, while the extracellular supernatant and intracellular extract were mixed with 1 ppm AFB1, and all mixtures were incubated at 30°C and 220 rpm agitation for 96 h. During the incubation, 450 μL of aliquots were collected every 24 h for HPLC analysis of AFB1 concentration.

The c4a strain showed the highest AVB1 removal activity among the isolated strains and thus was subjected to further morphological and biochemical characterization. The c4a strain was streaked onto a sterile LB agar plate and incubated overnight at 30°C to observe colony morphology. The c4a cells were Gram stained and then observed under a microscope. The biochemical characterization and carbohydrate metabolism profiling of the c4a strain was performed using API 20 NE and API 50 CH kits (bioMerieux Inc., Marcy l’Etoile, France) according to the manufacturer’s instructions.

Phylogenetic trees were constructed using single genes (16S rRNA and gyrB) and concatenated gene sequences (rpoB, gyrB, and rpoD). The 16S rRNA, gyrB, rpoB, and rpoD gene sequences of the c4a strain (GenBank assembly accession number GCA_009900745.1) and selected type strains, including Chryseobacterium balustinum DSM 16775, C. culicis DSM 23031, C.glaciei IHBB 10212, C. gleum ATCC 35910, C. indologenes NBRC 14944, C.nakagawai G0041 (GCF_900637665.1), C. piscium CCUG 51923 (GCA_003385415.1), C. scophthalmum DSM 16779 (GCA_900143185.1), C.timonianum G972 (GCA_900078205.2), and Riemerella anatipestifer ATCC 11845 (GCA_000252855.1), were obtained from the NCBI whole genome sequence data. For the type strain, Chryseobacterium lecithinasegens PAGU 2197, the NCBI database contained the housekeeping genes used in this study; but the whole genome sequence was unavailable. Upon sequence alignment with BioEdit, MEGA 10.0.4 software was used to construct phylogenetic trees with 1,000 bootstrap replications using neighbor-joining and maximum likelihood algorithms.

The extracellular supernatant of the c4a strain grown in nutrient broth, LB broth, TYE broth, and M9 medium was tested for AFB1 removal. 1% overnight c4a culture was inoculated into 100 mL growth media and incubated for 72 h in an orbital shaker at 30°C and 220 rpm. The extracellular supernatant was collected by centrifuging the bacterial culture at 10,000 g for 10 min at 4°C. The extracellular supernatant was mixed with 1 ppm AFB1 stock solution and incubated at 30°C for 72 h. The extracellular supernatant of the c4a strain grown in nutrient broth with gelatin peptone and beef extract in a 5:3 ratio had maximum AFB1 removal activity, so the ratio was optimized. The modified nutrient broth contained gelatin peptone and beef extract in a ratio of 1:9, 3:7, 1:1, 5:3, 7:3, 4:1, or 9:1, and the final total concentration of gelatin peptone and beef extract in the modified nutrient broth was 8 g/L. Based on AFB1 removal activity, the modified nutrient broth (modified NB) with gelatin peptone and beef extract in a 4:1 ratio was chosen as the final growth medium.

The extracellular supernatant of the c4a strain was prepared as described above. To study the effects of reaction temperature on AFB1 removal activity, the extracellular supernatant of the c4a strain was mixed with AFB1 stock solution to achieve a final concentration of 1 ppm AFB1 and then incubated at 15, 20, 30, 40, 50, or 60°C for 72 h on an orbital shaker at 220 rpm. To study the effects of reaction pH on AFB1 removal activity, the extracellular supernatant of the c4a strain was mixed with AFB1 stock solution to achieve a final concentration of 1 ppm AFB1, and then the pH values of the mixtures were adjusted to 6, 7, 8, or 9 using 1 N HCl or 1 N NaOH and then incubated at 30°C for 72 h on an orbital shaker at 220 rpm. To study the effects of mineral ions on AFB1 removal activity, the extracellular supernatant of the c4a strain was mixed with MgCl₂, ZnSO₄, CuSO₄, MnCl₂, LiCl, or FeCl₃ stock solution to achieve a final concentration of 10 mm of each mineral ion. Then, the mixture was mixed with AFB1 stock solution to achieve a final concentration of 1 ppm AFB1 and incubated at 30°C for 72 h on an orbital shaker at 220 rpm. To study the effects of heat treatment on AFB1 removal activity, the extracellular supernatant of the c4a strain was autoclaved at 121°C for 15 min and then mixed with AFB1 stock solution to achieve a final concentration of 1 ppm AFB1 and incubated at 30°C for 72 h on an orbital shaker at 220 rpm. To study the effects of protease inhibitors and SDS on AFB1 removal activity, the extracellular supernatant of the c4a strain was mixed with an EDTA-free protease inhibitor (cOmplete ULTRA Tablets, Mini, Roche Diagnostics, Basel, Switzerland) or SDS (5%). Then, the mixture was mixed with AFB1 stock solution to achieve a final concentration of 1 ppm AFB1 and incubated at 30°C for 72 h on an orbital shaker at 220 rpm. At the end of the experiment, aliquots of 450 μL from each sample were taken to analyze AFB1 concentration.

The high-resolution LC-MS and LC-MS/MS data were collected by a Thermo UPLC-ESI Orbitrap Elite mass spectrometer equipped with an ACQUITY UPLC system (0–6 min 5–99.5% ACN, 6–8 min 99.5% ACN, 8–8.2 min 99.5–5% ACN, 8.2–10 min 5% ACN, both ACN and H2O contain 0.1% formic acid; ACQUITY UPLC BEH C18, 1.7 μm, 2.1 × 100 mm; flow rate: 0.4 mL/min). The extracellular fraction of the c4a strain and sterile growth media were dosed with 2 ppm AFB1 under optimum reaction conditions. The sample injection volume was 10 μL. Positive ion mode mass data were acquired between m/z 100–1,500 at 15,000 resolution. The top four intense ions from each full mass scan were selected for collision-induced dissociation. The selected ions were fragmented with normalized collision energy 35.0, activation Q 0.250, activation time 10.0, and 15,000 resolution for collision-induced dissociation. The isolation width was 2.5 Da [20]. MS/MS data were obtained in duplicates utilizing data-dependent acquisition. MS scan was performed using Thermo Xcalibur 2.2 SP 1.48 software over mass-to-charge (m/z) range 100–1,500.

All AFB1 removal experiments were performed in triplicate, and the generated data were analyzed via one-way analysis of variance (ANOVA) using RStudio version 1.2.5. Duncan’s multiple range test was performed to detect statistical differences, and a p value <0.05 was considered statistically significant.

Bacterial strains with AFB1 removal potential were isolated from chicken digesta and soil samples using the coumarin-selective medium. Three bacterial strains, 3PB1, BN5, and 4CW, were isolated from mycotoxin-fed chicken digesta; one bacterial strain, IL1, was isolated from non-mycotoxin-fed chicken digesta; and one bacterial strain, c4a, was isolated from the soil. As shown in Figure 1, the c4a strain was the most effective AFB1 remover of the five isolated strains, so it was chosen for further investigation. Figure 2 represents the HPLC profile of AFB1 removal by extracellular supernatant of strain c4a, demonstrating AFB1 removal activity. The AFB1 removal activities of the intracellular extract and cell wall fraction of the c4a strain were also tested. As shown in Figure 3, the intracellular extract and cell wall fraction of the c4a strain had negligible AFB1 removal activity, implying that c4a-mediated AFB1 removal occurs exclusively extracellularly.

When grown on LB or nutrient agar, the c4a strain produced bright yellow colonies (1–3 mm). After 18–24 h of incubation at 30°C, the isolated colonies become elevated and smooth, with a buttery consistency (Fig. 4a). The yellow-orange convex colonies indicated the presence of flexirubin pigment, which turns red when 20% KOH is added. When grown on LB agar, these colonies also emit a fruity odor. Microscopic examination of the Gram-stained c4a cells revealed red-colored rods (Fig. 4b), which were Gram-negative and exhibited no motility or endospore formation.

The biochemical properties of the c4a strain were investigated using the API 20 NE system. The substrate assimilation profile of strain c4a revealed that it produced indole (tryptophan positive), urease, oxidase, hydrolyzed esculin, gelatin, assimilated glucose, mannose, maltose, and citrate. The c4a strain could not reduce nitrate to nitrogen, ferment glucose, produce arginine dihydrolase, and assimilate arabinose, n-acetylglucosamine, potassium gluconate, capric acid, adipic acid, malic acid, and phenylacetic acid. Based on these findings, the c4a strain was 99.4% identical to C. indologenes. However, the API database listed C. indologenes as the only species of the Chryseobacterium genus. The API 50 CH system was used to investigate further carbohydrate substrate utilization in the c4a strain. The results revealed that the c4a strain metabolized glycerol, amygdalin, esculin, d-trehalose, amidon, glycogen, gentiobiose, and xylitol (Table 1).

Figure 5 shows phylogenetic trees based on 16S rRNA and gyrB gene sequences. In the 16S rRNA phylogenetic tree, the c4a strain is located on a distinct branch with closely related neighbors, C. timonianum and C. lecithinasegens (Fig. 5a). In the gyrB phylogenetic tree, the c4a strain shared a phylogenetic branch with C. nakagawai and C. lecithinasegens (Fig. 5b). These phylogenetic trees could not determine the phylogenetic placement of the c4a strain. To improve phylogenetic resolution, multilocal sequence alignment (MLSA) was performed with a concatenated sequence of three housekeeping genes (rpoB-gyrB-rpoD) (Fig. 6). The c4a strain formed a direct branch with its closest neighbor, C. timonianum, suggesting that it might belong to C. timonianum.

To improve the AFB1 removal activity, the c4a strain was cultured in different growth media, and the AFB1 removal activity of the extracellular supernatants of the c4a strain was compared. As shown in Figure 7a, the extracellular supernatant of the c4a strain cultured in nutrient broth showed the highest AFB1 removal activity, followed by those cultured in LB broth and TYE broth. The extracellular supernatant of the c4a strain cultured in M9 minimal salt (M9) medium showed the lowest AFB1 removal activity. Figure 7b shows the effects of varied combinations of gelatin peptone and beef extract on the AFB1 removal activity of the extracellular supernatant of the c4a strain. The extracellular supernatant of the c4a strain cultured in the modified nutrient broth containing gelatin peptone and beef extract in a ratio 4:1 showed the highest AFB1 removal activity.

The effect of reaction temperature on AFB1 removal activity of the extracellular supernatant of the c4a strain was determined after 72 h of incubation of the extracellular supernatant of c4a with 1 ppm of AFB1 at 15, 20, 30, 40, 50, or 60°C. As shown in Figure 8a, the AFB1 removal activity of the extracellular supernatant of the c4a strain was favored at higher temperatures, reaching a maximum at 60°C.

The effect of reaction pH on AFB1 removal activity of the extracellular supernatant of the c4a strain was determined after 72 h of incubation of the CFF of c4a with 1 ppm of AFB1 at pH 6, 7, 8, or 9. As shown in Figure 8b, the AFB1 removal activity of the extracellular supernatant of the c4a strain was more effective in the alkaline range and maximal at pH 8.

The addition of metal ions in the extracellular supernatant of the c4a strain altered the AFB1 removal activity. As shown in Figure 8c, the addition of Mn²⁺ and Mg²⁺ ions significantly increased the AFB1 removal activity of the extracellular supernatant of the c4a strain while adding Fe³⁺ ions in the extracellular supernatant of the c4a strain decreased the AFB1 removal activity.

The addition of the protease inhibitor cocktails in the extracellular supernatant of the c4a strain significantly reduced the AFB1 removal activity, and the addition of SDS completely deactivated the AFB1 removal activity of the extracellular supernatant of the c4a strain, thus suggesting that the molecule(s) in the extracellular supernatant of the c4a strain-mediated the AFB1 removal is proteinic. The AFB1 removal activity of the autoclaved extracellular supernatant of the c4a strain was retained, indicating that heat treatment did not affect the AFB1 removal activity of the extracellular supernatant of the c4a strain (Fig. 8d).

The AFB1 degradation pattern (m/z values) by the c4a strain was found to be 417.21 → 412.25 → 331.08 → 313.07 → 285.08, with loss of H₂O in the second last step, followed by loss of CO in the final step. The final degradation product had the molecular formula C₁₇H₁₄O₇. Figure 9a shows the MS-1 profile of strain c4a-mediated AFB1 degradation. It showed the predominant presence of an ion fragment of m/z 313.07. Figure 9b shows the MS-2 profile at m/z 331. 08, which shows the dominant presence of ion fragment m/z 313.07. Figure 9c shows the MS-2 profile at m/z 313.07 having an ion fragment of m/z 285.08.

Coumarin is a phenolic compound derived from cis-O-hydroxycinnamic acid with a structure similar to that of AFB1, and so has been used in the selection media to screen for AFB1-degrading microbial strains [18, 21, 22]. Coumarin media was thus employed to selectively isolate bacterial strains with the potential to remove AFB1 in the present study. Five bacterial strains, 3PB1, BN5, 4CW, IL1, and c4a, were obtained following coumarin-selective screening. The c4a strain, isolated from the chicken feather enriched soil sample, showed the highest AVB1 removal activity among the isolated strains and thus was subjected to further study.

The primary morphological and biochemical characteristics of the c4a strain were similar to the other members of the genus Chryseobacterium. A comparative analysis of the biochemical profiles to understand the unique metabolic capabilities of the c4a strain and its close neighbors, C. timonianum G972 [23] and C. lecithinasegens CCUG 17150 [24], was carried out. All three strains tested positive for acid production using d-glucose, d-mannose, esculin, d-maltose, d-trehalose, and gentibiose. C. lecithinasegens and the c4a strain utilized amygdalin and produced indole, whereas C. timonianum did not. These comparisons revealed that the c4a strain and C. timonianum have a more varied metabolic profile than C. lecithinasegens. Unlike C. lecithinasegens and C. timonianum, the c4a strain could not metabolize glycogen or glycerol.

The 16S rRNA-based phylogenetic tree showed that the c4a strain formed a monophyletic branch with C. timonianum (99.21%) and C. lecithinasegens (98.92%). To understand the differentiation of strain c4a within this genus, another phylogenetic tree was created using gyrB gene sequences. The most similar sequences to strain c4a were C. lecithinasegens (92.92%), followed by C. nakagawai (92.87%) and C. timonianum (91.58%). The c4a strain’s gyrB sequence was less than 89% similar to other Chryseobacterium strains. We created an additional phylogenetic tree using a rpoB-gyrB-rpoD concatenated sequence of conserved housekeeping genes and found that the c4a strain and C. timonianum shared a close phylogenetic relationship, implying that it may belong to C. timonianum.

The Chryseobacterium genus is classified within the Flavobacteriaceae family, Flavobacteriales order, and in the Flavobacteriia class. The majority of Chrsyeobacterium species were found in soil, freshwater or saltwater, insects, biofilms, and industrial sources like beer bottling factories. Although some Chryseobacterium species have been isolated from human clinical samples, it is difficult to determine whether they are direct or opportunistic pathogens or merely contaminants [25]. Similarly, the c4a strain was isolated from farm soil that had not been fertilized or treated with chemicals, whereas C. timonianum was isolated from sputum.

Characterization studies showed that the c4a strain-mediated AFB1 removal was constitutive and extracellular. There is abundant evidence of microbial degradation across multiple genera, and physical AFB1 binding to bacterial and yeast cell walls has been widely reported [12]. This surface binding may not be specific, leading to binding minerals and trace nutrients like vitamins [26, 27]. Enzymatic degradation is a more effective and targeted biocontrol strategy for AFB1 contamination. After 72 h, the c4a strain’s extracellular supernatant removed the highest AFB1 in this study. The growth media components affected the c4a strain’s AFB1 removal activity, with gelatin peptone being a critical media component. AFB1 removal activity of strain c4a spanned a wide temperature range and was found to be highly thermotolerant. We observed that autoclaving did not affect this phenomenon. A microbial consortium with dominant Geobacillus and Tepidimicrobium species [28] and single bacterial strains like Escherichia coli CG1061 [15], Bacillus velezensis DY3108 [13], and Stenostrophomonas sp. CW117 [16] have shown thermostable AFB1 removal activity at 80°C–100°C.

The pH dependence and metal cofactor specificity for the AFB1 removal activity of the c4a strain were observed, with acidic pH significantly reducing the AFB1 removal activity of the c4a strain, similar to B. velezensis AD8 [29]. The highest AFB1 removal activity of the c4a strain was found at pH 8, similar to that observed for E. coli CG1061 [15]. The addition of a protease inhibitor caused a reduction in AFB1 removal activity of the c4a strain, which implied that the presence of protease positively affected AFB1 removal activity. The proteinic AFB1 removal activity was completely inhibited after SDS addition, as SDS causes the structural disintegration of proteins in the extracellular cell-free fraction. The AFB1 removal activity of the c4a strain was dependent on metal ions supplementation, with Fe³⁺ ion addition reducing AFB1 removal activity, whereas Mn²⁺ and Mg²⁺ ions significantly improved AFB1 removal activity. Fe³⁺ induced inhibition of AFB1 removal activity was found to be similar to the case of the bacterium Myroides odoratimimus [14]. AFB1 removal improvement effect in the case of Mn²⁺ and Mg²⁺ ion additions was found to be identical to that reported for Stenotrophomonas sp. CW117 [16]. These characterization results indicate that the AFB1 removal activity of the c4a strain resulted from complex protein interactions with excellent thermotolerance and a specific preference for pH and metal ions.

UPLC-ESI MS/MS analysis gave insight into strain c4a-mediated AFB1 degradation. At 2.55 s, a metabolite peak of various ionic fragments emerged. The MS-1 and MS-2 analyses of this peak showed that ion 313.07 m/z disintegrated into a final transformed product of 285.08. Recently, an interesting work documented all AFB1 breakdown products after UV irradiation using UPLC-Q-TOF-MS/MS and NMR [4]. This report revealed the same degradation product with chemical formula C17H14O7 as the c4a strain (285.08 m/z). The transformation pattern suggested water molecule addition on the double bond (C-8/C-9) of AFB1. The article assayed degradation product toxicity on LO-2, Hep-G2, and MCF-7 cell lines [4]. The cytotoxic activity (µM) of the transformed product (m/z 285.08) was much lower (>100 μm) than pure AFB1(22.47, 29.08, 36.57 μm) against LO-2, HepG2, and MCF-7 cell lines. These findings showed that strain c4a extracellular supernatant may target the C8-C9 double bond of AFB1’s furan ring, lowering its cytotoxicity.

Based on the relevant literature review, AFB1 removal activity of any Chrsyeobacterium species has not been reported yet. To the best of our knowledge, this is the first report of the AFB1 removal potential of a bacterial strain belonging to C. timonianum. This evidence makes the current study a promising reference for future studies concerning the environmental biocontrol of AFB1 by C. timonianum.

We are grateful to Dr. Yu Liang Yang for guidance in LC-MS/MS experimentation and analysis. This study was conducted in the metabolomics core facility, ABRC, Academica Sinica, Taipei, Taiwan.

The animal procedures were approved by the Institutional Animal Care and Use Committee of National Taiwan University, Approval No. NTU105-EL-00027.

The authors have no competing interests to declare that are relevant to the content of this article.

The work was supported by a grant from the Ministry of Science and Technology (Grant No. MOST 110-2313-B-002-052-MY2).

All authors contributed to the study’s conception and design. Material preparation, data collection, and analysis were performed by Aniket Limaye. The first draft of the manuscript was written by Aniket Limaye and Dr. Je Ruei Liu. All authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.