J. Microbiol. Biotechnol. (2017), 27(1), 57–66 https://doi.org/10.4014/jmb.1607.07064 Research Article Review jmb

Aflatoxin B1 Detoxification by Aspergillus oryzae from Meju, a Traditional Korean Fermented Soybean Starter S Kyu Ri Lee1†, Sun Min Yang1†, Sung Min Cho1, Myunghee Kim2, Sung-Yong Hong1*, and Soo Hyun Chung1*

1Department of Integrated Biomedical and Life Science, Korea University, Seoul 02841, Republic of Korea 2Department of Food Science and Technology, Yeungnam University, Gyeongsan 38541, Republic of Korea

Received: July 28, 2016 Revised: October 26, 2016 Aflatoxins are classified as Group 1 (carcinogenic to humans) by the International Agency for Accepted: October 28, 2016 Research on Cancer. In this study, a total of 134 fungal strains were isolated from 65 meju

First published online samples, and two fungal isolates were selected as potential aflatoxin B1 (AFB1)- November 4, 2016 biodetoxification fungi. These fungi were identified as Aspergillus oryzae MAO103 and

*Corresponding authors A. oryzae MAO104 by sequencing the beta-tubulin gene. The two A. oryzae strains were able to S.H.C. degrade more than 90% of AFB1 (initial concentration: 40 µg/l) in a culture broth in 14 days. Phone: +82-2-3290-5641; Fax: +82-2-928-7825; The mutagenic effects of AFB1 treated with A. oryzae MAO103 and MAO104 significantly E-mail: [email protected] decreased to 5.7% and 6.4%, respectively, in the frame-shift mutation of Ames tests using S.-Y.H. Salmonella typhimurium TA98. The base-substituting mutagenicity of AFB was also decreased Phone: +82-2-3290-5641; 1

Fax: +82-2-928-7825; by the two fungi. Moreover, AFB1 production by Aspergillus flavus was significantly decreased E-mail: [email protected] by the two A. oryzae strains on soybean-based agar plates. Our data suggest that the two AFB1-

† These authors contributed detoxifying A. oryzae strains have potential application to control AFB1 in foods and feeds. equally to this work.

S upplementary data for this Keywords: Aflatoxin B1, biodegradation, detoxification, soybean fermentation starter paper are available on-line only at http://jmb.or.kr.

pISSN 1017-7825, eISSN 1738-8872

Copyright© 2017 by The Korean Society for Microbiology and Biotechnology

Introduction by nearly all countries. Permissible levels of total aflatoxin content in foods and feeds have been set, varying from zero Aflatoxins are a group of mycotoxins that are produced tolerance to 50 μg/kg. For example, the European Union has mainly by toxigenic Aspergillus flavus and Aspergillus enacted considerably lower limits of the toxin: 2.0 μg/kg parasiticus [1, 2]. A. flavus produces only B types of for AFB1 and 4.0 μg/kg for total aflatoxins in foods. South aflatoxins, but A. parasiticus produces both B and G types. Korea has established a regulatory tolerance threshold of

Other Aspergillus strains, such as A. nomius, A. tamarii, and 10.0 μg/kg AFB1 and 15.0 μg/kg total aflatoxins for cereals, A. pseudotamarii, are also able to produce aflatoxins [3, 4]. It red pepper powders, and fermented soybean products for has been reported that aflatoxins, especially aflatoxin B1 human consumption, which are in accordance with the

(AFB1), are mutagenic, carcinogenic, genotoxic, hepatotoxic, AFB1 limits of most countries (10 to 20 μg/kg AFB1 in foods). and immunosuppressive to both humans and animals [5-7]. Three methods to detoxify aflatoxins in foods and feeds Aflatoxins are classified as Group 1 (carcinogenic to humans) have been used: physical, chemical, or biological. Physical by the International Agency for Research on Cancer (IARC, methods include autoclaving [8], pasteurization [9], solvents 2002). Considering the toxin’s potential hazard to human [10], adsorbent materials [11, 12], UV light [13], and health, monitoring programs and regulatory actions for processing [14, 15] such as roasting, cleaning, separation, aflatoxins in various commodities have been implemented and milling. Chemical detoxification methods include the

January 2017 ⎪ Vol. 27⎪ No. 1 58 Lee et al.

use of H2O2 [16], ozone [17], ammonia [18], urea [19], Na- was a Soytone-Czapek (SC) medium, which was composed of 3 g hypochlorite [20], and Na-bisulfite [21]. These two categories soytone, 1 g K2HPO4, 0.5 g KCl, 0.5 g MgSO4·7H2O, 0.01 g FeSO4·H2O, of aflatoxin detoxification methods, however, suffer from 30 g sucrose, and 1 L of distilled water. For identifying the problems such as reducing the nutritional value of the degradation ability, fungal isolates were grown in 50 ml of SC medium containing 40 μg/l AFB after 105 fungal spores were food, altering its organoleptic qualities, causing undesirable 1 inoculated. The culture was incubated at 25oC with shaking at health effects, and giving rise to high equipment costs, 130 rpm. limiting their application to food manufacturing systems [22-24]. Thus, the most desirable means for aflatoxin Reagents and Chemicals detoxification in foods are biological methods. Shantha [25] The standard AFB1 was supplied by Sigma-Aldrich (USA), and indicated that some fungi, such as Rhizopus sp., Trichoderma the standard aflatoxin mixture was obtained from Supelco sp., Phoma sp., Sporotrichum sp., and Alternaria sp., inhibit (Aflatoxin mix kit; Supelco, USA). HPLC-grade methanol and

AFB1 synthesis. Additionally, it has been reported that acetonitrile were purchased from J.T. Baker (Avantor Performance some bacteria, such as Mycobacterium sp. [26], Rhodococcus Materials, Inc., USA). Ethyl acetate was obtained from Daejung sp. [24], Bacillus sp. [27], and Myxococcus [28], have Chemicals & Metals Co. (Korea). Deionized water was purified detoxification capabilities. Recently, Samuel et al. [29] using Direct-Q 3 (Merck Millipore, USA). All solvents for HPLC analysis were filtered through a membrane filter (47 mm × reported that Pseudomonas putida efficiently degraded AFB1 by about 90% in a salt glucose medium containing 40 ng/ml 0.45 μm, GHP; Pall Corporation, USA). For sample filtration before HPLC analysis, a syringe filter (13 mm × 0.2 μm, GHP) was of the toxin. also obtained from Pall Corporation. Meju, a starter for traditional Korean fermented soybean products such as doenjang, ganjang, and kochujang, can AFB1 Biodegradation Analysis by TLC be occasionally contaminated with aflatoxins through For qualitative analysis of AFB1 biodegradation, TLC analysis aflatoxigenic fungi [30, 31]. However, it has been documented was conducted. AFB1 in culture media, which were cultured as that doenjang and ganjang contain much lower levels of described above, was extracted by liquid-liquid extraction (LLE). aflatoxins than meju, a raw material for the products (Jeong Briefly, the fungal culture broth was filtered using Whatman GF/A et al. 2009. Abstr. Annu. Meet. Kor. Soc. Food Sci. Nutri. glass filter paper (Whatman Inc., USA) after incubation for 7 and Changwon, Korea, p. 252). This suggested that there was a 14 days. Then, 1 ml of filtrate and 3 ml of ethyl acetate were mixed possibility of aflatoxin detoxification by some fungi in meju by vortexing for 30 sec. After the mixture was placed at room during soybean fermentation and that the aflatoxin temperature for 30 min, 2 ml of the upper layer was transferred to detoxification ability of these fungi could provide benefits a new glass vial. Two milliliter of ethyl acetate was added to the lower layer, mixed, and then the upper layer was combined with for the safety of fermented products because AFB would 1 the first extract. The ethyl acetate extracts were evaporated to be decreased during fermentation in situ. Moreover, these dryness under a gentle stream of N . The dried extracts were fungi could possibly inhibit the growth of toxigenic fungi 2 dissolved in 50 μl of benzene-acetonitrile (98:2 (v/v)). For the TLC through competition. As a result, the total amount of toxins analysis, 30 μl of the solution was spotted on a pre-coated TLC in the products would be reduced efficiently during the plate (TLC Silica gel 60 F254; Merck Millipore, USA). Development fermentation process. was performed at room temperature in a glass TLC chamber

The aim of the present study was to investigate the AFB1- previously saturated with benzene-methanol-acetic acid (180:10:10 detoxifying activity of two A. oryzae strains isolated from (v/v/v)) as a developing solvent. After development, the TLC meju. This is the first report showing that A. oryzae is plate was removed from the tank, dried at room temperature, and illuminated under UV light at λ = 365 nm in a dark chamber capable of detoxifying AFB1. (Chambres Noires darkroom; Vilber Lourmat, France) [32-34].

Materials and Methods Identification of Fungal Isolates For morphological identification, fungi were inoculated on the Isolation of AFB -Degrading Fungi 1 PDA and incubated at 25oC for 5 days. To identify the fungal For the fungal isolation, dichloran rose-bengal chloramphenicol isolates at the genus level, colony and microscopic morphologies agar (BD Difco, USA) was used. Each isolation was transferred were observed. For molecular identification, fungal spores (105) onto potato dextrose agar (PDA; MB Cell, Korea) slants, and spore were inoculated in a potato dextrose broth (PDB) and incubated at suspensions were prepared with 0.05% Tween 80 solution. Fungal 25oC for 48 h. After the fungal mycelia were harvested from the spore suspensions were stored in a 20% glycerol solution at -20oC media, total chromosomal DNA was extracted using a HiGene until use. The medium for determining AFB -degrading capability 1 Genomic DNA Prep Kit (BIOFACT, Korea). The beta-tubulin

J. Microbiol. Biotechnol. Aflatoxin B1-Detoxifying Aspergillus oryzae 59

region was selected for molecular identification, and the PCR was were treated with 1 ml of TFA-10% acetonitrile (10:90 (v/v)) and performed with two specific primers, Bt2a and Bt2b (5’-GGT AAC mixed by a vortex mixer (Fisher Scientific, USA) for 30 sec. The CAA ATC GGT GCT TCC-3’ and 5’-ACC CTC AGT GTA GTG mixture was placed in darkness for 3 h, and then filtered through

ACC CTT GGC-3’), using a PC708 thermal cycler (Astec, Japan) a syringe filter (13 mm × 0.2 μm, GHP). The AFB 1 standard [35]. The PCR mixture contained 5 μl of 10× PCR buffer (BIOFACT), mixture was also derivatized with TFA using the same procedure.

3 μl of 2.5 mM deoxyribonucleotide triphosphate (BIOFACT), The samples and AFB1 standard mixture were programmed to 0.4 μl of 100 pmol of each primer (Neoprobe, Korea), 0.3 μl of Taq inject 50 μl and to run for 20 min through a CAPCELL PAK C18 polymerase (BIOFACT), 39.9 μl of sterile deionized water, and column (4.6 mm i.d. × 250 mm, 5 μm; Shiseido, Japan). The mobile 1 μl of DNA template. The PCR was performed as follows: pre- phase was acetonitrile-methanol-water (15:15:70 (v/v/v)) pumped o o denaturation for 4 min at 95 C, denaturation for 1 min at 95 C, at a constant flow rate of 1 ml/min. The determination of AFB1 annealing for 1 min at 65oC, extension for 2 min at 72oC, and a was carried out using a fluorescence detector with 360 nm and post-extension for 7 min at 72oC. The PCR was repeated for 35 440 nm for excitation and emission, respectively [30, 40]. cycles. The PCR products were sequenced by BIOFACT [36]. The Basic Ames Test

Local Alignment Search Tool was used to find regions of similarity AFB1-degrading fungi were inoculated in 50 ml of SC medium o between sequences from the fungal isolates and the strains of supplemented with 800 μg/l AFB1 and then incubated at 25 C GenBank in the National Center for Biotechnology Information with shaking at 100 rpm for 14 days. The culture was filtered with (NCBI) nucleotide database. Each sequence of beta-tubulin gene a Whatman GF/A and then a sterile syringe filter (25 mm × from the two A. oryzae strains was used to determine phylogenetic 0.45 μm, GHP) before use. A control sample without the fungal relationships with those of the A. oryzae type strain and of Aspergillus culture also contained 800 μg/l AFB1 in SC medium. The pre- section Flavi strains. A phylogenetic tree was constructed by a incubation assay was conducted according to the method described neighbor-joining tree using MEGA5 [37, 38]. DNA data were by Sugimura et al. [41]. Salmonella typhimurium strains TA98 processed using the Tamura-Nei parameter distance calculation (CCARM 0272) and TA100 (CCARM 0273) were used as tester model with gamma-distributed substitution rates in order to strains. The metabolic activation was performed in a reaction analyze the taxonomic positions of the two A. oryzae strains. To mixture containing 0.1 ml of test sample, 0.1 ml of culture medium, determine the support for each clade, bootstrap analysis was and 0.5 ml of S9 liver enzyme fraction (Moltox, USA) for 30 min at performed with 1,000 replications. 37oC. After the activation step, the top agar with the activated fraction was poured on a minimal glucose agar plate made with Test for Aflatoxin Production by Fungal Isolates 50 ml of 40% glucose, 20 ml of 50× VB salts, 15 g agar powder In order to determine whether aflatoxin is produced by the (Daejung Chemicals & Metals Co., Korea), and 930 ml of distilled A. oryzae isolates or not, molecular analysis at the DNA level and water. It was incubated for 48 h at 37oC. Each sample was tested in production analysis by fungal cultures were performed. For the triplicates. molecular analysis of aflatoxigenicity, genetic analysis of 17 genes in the aflatoxin biosynthetic pathway was conducted according to AFB1 Production by A. flavus M 2034 Co-Cultured with AFB1- Tao and Chung [36]. For the aflatoxin production test, yeast Biodegrading A. oryzae in Soybean-Based Media extract-sucrose media (YES media: 2% yeast extract, 20% sucrose) Each AFB1-biodegarding A. oryzae isolate or A. oryzae KACC [39] and SC media were used. One milliliter of each fungal spore 46464 was co-cultured with aflatoxigenic A. flavus M 2034 (isolated suspension (105/ml) was inoculated into 100 ml of media and from meju) [42] on soybean-based media. The soybean-based incubated for 3 weeks at 25oC with shaking at 130rpm. The media were prepared as follows: 500 g of white soybean was amount of aflatoxins in the culture media was analyzed by HPLC boiled with 1 L of water and then soybean water was extracted by after LLE. squeezing it through a cheesecloth. A soybean-based agar plate was made by combining the soybean water extract with 2% agar

AFB1 Biodegradation Analysis by HPLC powder (Daejung Chemicals & Metals Co.). Each A. oryzae strain

For quantitative analysis of AFB1 biodegradation, fungal culture was inoculated on the right side of the plate with A. flavus M 2034 media were taken at 0, 1, 3, 5, 7, 10, and 14 days in triplicates. on the left side and incubated for 7 days. The amount of AFB1 After the culture medium was filtrated through a glass microfiber produced by A. flavus M 2034 on the soybean-based media was filter (GF/A; Whatman Inc.), AFB1 was extracted from 1 ml of analyzed by HPLC after extraction [43]. Briefly, fungal cell mass filtrate with 3 ml of ethyl acetate by LLE as described above. For plus agar on the plate was homogenized, and aflatoxin was the AFB1 mycelial binding test, the mycelia of A. oryzae MAO 103 extracted with methanol-chloroform (1:2). Each experiment was and MAO 104 from 7-day and 14-day cultures were collected and performed in triplicates. dried, and the toxin was extracted with 10 ml of ethyl acetate. For derivatization of AFB1, trifluoroacetic acid (TFA; Sigma-Aldrich, Statistics USA) was used as a labeling agent. The dried culture samples Data were statistically evaluated using one-way ANOVA. The

January 2017 ⎪ Vol. 27⎪ No. 1 60 Lee et al.

Table 1. Number and frequency of fungal isolates from meju samples. Fungal species Number Frequency (%) Aspergillus section Flavi 47 35.1 Aspergillus section Nigri 64.5 Other Aspergillus sp. 25 18.7 Penicillium sp. 20 14.9 Mucor sp. 27 20.1 Unknown 9 6.7 Total 134 100 The fungal strains were obtained from DRBC agar plates, transferred onto PDA plates, and incubated for 5 days at 25oC. results were considered significant when p values were <0.05. Tukey’s multiple comparison tests were used when a significant difference was encountered. The statistical analyses were performed with SPSS software (SPSS, USA).

Results and Discussion

Screening of AFB1-Biodegrading Fungi One hundred thirty-four fungi were isolated from 65 lumps of meju that were collected from the middle and southern regions of South Korea (Table 1). The fungal isolates were divided into five genera by morphological examination: Aspergillus section Flavi, Aspergillus section Nigri, other Aspergillus sp., Penicillium sp., and Mucor sp. The predominant isolates belonged to Aspergillus section

Flavi (35.1%). The fungal isolates were tested for AFB1- biodegrading activity by using TLC. Two fungal isolates

(MAO 103 and MAO 104) showed AFB 1 biodegradation activities; Fig. 1A shows the TLC results of the 7-day and 14-day cultures of the two fungal isolates. The brightness of the blue spots (AFB1) from the 7-day culture samples was faded compared with the control sample, and furthermore, the spots of AFB1 from the 14-day culture samples were Fig. 1. Screening of AFB1-biodegrading fungi. only dimly visible on the TLC plate. The AFB1 biodegradation activity by the two fungal isolates (MAO 103 and MAO (A) TLC analysis of culture media extracted by using ethyl acetate. Each strain was grown in SC broth with 40 µg/l AFB added. STD 104) was also analyzed by HPLC (Figs. 1B and 1C). Similar 1 indicates AFB standard (50 µg/l). (B) The HPLC chromatogram of to the result of TLC, HPLC chromatograms showed 1 AFB1 in a culture broth of MAO 103. The retention time of the AFB1 decreased amounts of AFB1 in the 7-day and 14-day culture peak was 8.4 min. 1. Medium of control (0-day culture). 2. MAO 103 samples. (7-day culture). 3. MAO 103 (14-day culture). (C) The HPLC

chromatogram of AFB1 in a culture broth of MAO 104. The retention

Identification of AFB1-Biodegrading Fungi time of the AFB1 peak was 8.4 min. 1. Control (0-day culture). 2. MAO 104 (7-day culture). 3. MAO 104 (14-day culture). The two AFB1-biodegrading fungal strains were identified by morphological examination and DNA sequencing of the beta-tubulin region. The color of the 7-day-old MAO 103 and later changed to a light brown (Figs. 2A and 2B). and MAO 104 colonies was a pale olive on the PDA plates Under microscopic examination, they showed the typical

J. Microbiol. Biotechnol. Aflatoxin B1-Detoxifying Aspergillus oryzae 61

characteristics of Aspergillus sp., such as conidiophores on an unbranched stipe with a swollen vesicle, phialides borne directly on the vesicle, and conidia radiated on the phialide from the vesicle (Figs. 2C and 2D). The beta- tubulin region was selected for molecular identification as the two isolates belonged to Aspergillus sp. on the basis of the fungal morphology. The sizes of the PCR product of the beta-tubulin gene of Aspergillus MAO 103 and MAO 104 were 497 and 496 bp, respectively. The Aspergillus MAO 103 and MAO 104 isolates were identified as Aspergillus oryzae by comparison of nucleotide sequences in the PCR products from beta-tubulin with sequences in the same region from strains selected from the NCBI database (NCBI Accession ID: KY020433 for A. oryzae MAO 103, KY020434 for A. oryzae MAO 104). The taxonomic positions of the two

AFB1-biodegrading A. oryzae strains with representative strains of Aspergillus section Flavi based on the beta-tubulin gene are shown in Fig. 3. As we expected, the two A. oryzae strains were clusted with the A. oryzae type strain (A. oryzae T CBS100925 EF203138) and A. flavus type strain (A. flavus Fig. 2. Morphological examination of two isolates. T CBS 100927 EF203132) into one group. (A) The colony of A. oryzae MAO 103 (7-day culture) on PDA, and As it has been reported that A. oryzae strains have a microscopic examination of A. oryzae MAO 103. (B) The colony of genetic similarity with Aspergillus flavus, which produces A. oryzae MAO 104 (7-day culture) on PDA, and microscopic examination of A. oryzae MAO 104. AFB1 and AFB2 [44], a genetic analysis of 17 genes in the aflatoxin biosynthetic pathway was conducted to verify non-production of the toxin by the two A. oryzae isolates aflatoxin biosynthetic genes (aflG, aflI, aflK, aflM, aflO, aflP, [36]. The two A. oryzae strains possessed only seven and aflQ), and not 10 aflatoxin biosynthetic genes (Table 2).

Fig. 3. Phylogenetic tree depicting taxonomic positions of A. oryzae MAO 103 and A. oryzae MAO 104. The DNA sequences of the beta-tubulin genes of each fungus were obtained from the NCBI and compared. The phylogenetic tree was constructed with MEGA5. The “T” after the collection number indicates the type strain of the species.

January 2017 ⎪ Vol. 27⎪ No. 1 62 Lee et al.

From this result, A. oryzae MAO 103 and MAO 104 could be genetically discriminated from aflatoxigenic A. flavus. In addition, neither A. oryzae MAO 103 nor MAO 104 produced aflatoxins during 3-week cultures in SC media and YES media, whereas A. flavus M 2034 produced considerable amounts of AFB1 (≥15 mg/l).

Degradation of AFB1 during the Cultivation of A. oryzae MAO 103 and MAO 104

Quantitative AFB1 biodegradation activity during the cultivation of A. oryzae MAO 103 and MAO 104 was investigated by HPLC analysis. The two A. oryzae strains showed time-dependent AFB1 biodegradation activity and were able to degrade more than 90% of the AFB1 in culture media over 14 days (Fig. 4). A. oryzae MAO 103 showed more rapid biodegradation activity than MAO 104 after Fig. 4. Time course of the degradation of AFB1 in a culture 7 days. In the A. oryzae MAO 103 culture, a sharp decline of broth of A. oryzae MAO 103 and MAO 104. the toxin was detected between 3 and 7 days, with an Each strain was grown in SC broth. The toxin concentrations are almost 70% reduction of AFB measured at 7 days. On the shown as the mean of three replicates. Error bars correspond to the 1 standard deviation. other hand, A. oryzae MAO 104 degraded AFB 1 gradually over 14 days. The toxin degradations in the 14-day culture were 97.4% and 96.1% (p < 0.001) in A. oryzae MAO 103 and amounts of AFB1 removed from the total cultures of A. oryzae MAO 104, respectively. Among other aflatoxin derivatives, MAO 103 and MAO 104 were 92.4% and 78.4% of the initial aflatoxin B2 (AFB2) was significantly decreased 69.8% and toxin amount, respectively, at 14 days. Considering the 67.7% (p <0.001) by A. oryzae MAO 103 and MAO 104, total reduction of toxin in the culture broth, the adsorption respectively, in 14-day culture media (initial AFB2 of AFB1 to fungal mycelia was relatively low in both concentration: 40 μg/l). However, there was no significant A. oryzae MAO 103 and MAO 104. This indicates that the decrease in aflatoxin G1 (AFG1) or G2 (AFG2) (Fig. S1). main AFB1 biodegradation activity by A. oryzae MAO 103 and MAO 104 was not due to fungal adsorption.

Testing for AFB1 Mycelial Binding

It has been reported that yeast and lactic acid bacteria Loss of Mutagenicity of AFB1 by A. oryzae MAO 103 and can reduce aflatoxins by adhesion to their cell wall MAO 104 components [45-47]; thus we tested if the fungal mycelia The Ames assay has been used to measure the mutagenicity bind AFB1 using the 7-day and 14-day cultures of A. oryzae of chemicals and complex mixtures [48-52]. It is known

MAO 103 and MAO 104. The amounts of AFB1 detected in that AFB1 is mutagenic in the Ames system and that the both culture broth and mycelia are shown in Table 3. The mutagenicity is generally effective in preincubation systems

Table 2. PCR amplification of aflatoxin biosynthetic genes in the gene cluster of A. oryzae MAO 103, A. oryzae MAO 104, and A. flavus M 2034. PCR amplification Aflatoxin aflR aflJ aflA aflC aflD aflE aflF-aflU aflG aflH aflI aflK aflL aflM aflO aflP aflQ aflT production A. oryzae ------+-++-++++- - MAO 103 A. oryzae ------+-++-++++- - MAO 104 A. flavus ++++++ + ++++++++++ + M 2034 In PCR amplification: +, amplification signal present; -, amplification signal absent. In aflatoxin production: +, aflatoxin production; -, no aflatoxin production. For aflatoxin production testing, YES media and SC media were used, and each fungus was cultured for 3 weeks.

J. Microbiol. Biotechnol. Aflatoxin B1-Detoxifying Aspergillus oryzae 63

Table 3. The amounts of residual AFB1 in culture broth filtrate and mycelia and the total AFB1 degradation by A. oryzae MAO 103 and MAO 104. Residual AFB (ng) Degraded AFB (ng) Strains 1 1 Culture broth filtrate Mycelia (Degradation rate, %) 0 days 1,213.8 ± 13.60.0 0.0 ( 0.0) MAO 103 7 days 387.4 ± 26.1 218.5 ± 18.3 607.8 ± 23.0 (50.1) 14 days 40.1 ± 10.8 52.0 ± 7.4 1,121.6 ± 12.1 (92.4) MAO 104 7 days 505.3 ± 133.7 115.0 ± 48.2 593.4 ± 127.8 (48.9) 14 days 118.0 ± 8.6 144.4 ± 20.1 951.4 ± 25.1 (78.4)

[53, 54]. Results of Ames tests using two histidine-requiring base-substituting mutagenicity tested with the TA100 strain mutants of S. typhimurium (TA98 and TA100) are shown in also decreased in samples treated with A. oryzae strains

Fig. 5. The mutagenic effect of AFB1 was decreased by (49.5% in A. oryzae MAO 103 and 50.8% in MAO 104, treatment with A. oryzae MAO 103 and MAO 104. The p < 0.001) to the same levels as that in a blank sample untreated AFB1 control sample, after incubation for 72 h at (52.3%). These results indicate that the mutagenic effects of

30°C, still had strong mutagenic responses in both Salmonella AFB1 were decreased by the two AFB1-degrading fungi. It has strains, whereas the results from the samples treated with been reported that the mutagenic effects of AFB1 were caused A. oryzae MAO 103 and MAO 104 showed that the mutagenic by both the lactone portion and the dihydrofuranofuran effects of AFB1 decreased to 5.7% and 6.4% (p < 0.001) in moiety of the molecule [55-57]. Our Ames test data suggest frame-shift mutations of the TA98 strain, respectively. The that loss of the mutagenic structures of AFB1 may have occurred through the cultures of A. oryzae MAO 103 and MAO 104.

Analysis of AFB1 Biodegradation Products When the 14-day culture broth of each fungal strain was examined to detect degradation products by using HPLC, none were observed. These results are similar to those from previous researchers [58-60]. Nakazato et al. [61] reported that several fungi, including Aspergillus niger and A. flavus,

converted AFB1 to aflatoxicol and vice versa. This interconversion was observed in the fungal culture broth and cell-free systems obtained from disrupted mycelia, but not in culture filtrates from intact mycelia of the same fungal strains. In our results, the fungal culture filtrates did

not show AFB1-degradation activity, which was in agreement with their results. On the other hand, there was no aflatoxicol in the fungal cultures of A. oryzae MAO 103 and

MAO 104. This suggests that AFB1 is likely to be metabolized

Fig. 5. Reduction of AFB1 mutagenic effects by A. oryzae into biodegradation products with different structural and MAO 103 and MAO 104. chemical properties. S. typhimurium TA98 and TA100 were used as tester strains, which Recently, Iram et al. [62] proposed several degraded were incubated with a 14-day culture broth of A. oryzae MAO 103 and products from the detoxification of AFB1 by Corymbia citrodora MAO 104. SC media spiked with 800 µg/l AFB were used for the 1 plant extracts when using LC/MS/MS. In their study, the culture of A. oryzae strains and the control (without fungal toxicity of these degraded products was significantly inoculation). Blank samples were SC media with neither toxin nor reduced because of the removal of the double bond on the fungal inoculation. The tests were conducted in triplicates for each test sample. Asterisks indicate statistically significant differences terminal furan ring. Our results showed a significant between control and treated samples (***p < 0.001); error bars reduction of the mutagenic effect in the results of the Ames correspond to the standard deviation. test. We will investigate possible degraded products of

January 2017 ⎪ Vol. 27⎪ No. 1 64 Lee et al.

with A. flavus M 2034. When considering the differences of

AFB1 production by cultures of A. flavus M 2034, A. oryzae KACC 46464, and the two A. oryzae isolates, the reduced

AFB1 production in the co-culture of A. oryzae MAO 103 or

MAO 104 was due to their AFB1-biodegrading ability rather than fungal competition. These results suggest that A. oryzae MAO 103 and MAO 104 can be used together as an effective

tool for control of AFB1 in food systems.

Liu et al. [63] found an AFB1-degrading mushroom. In

other studies, AFB1 biodegradation microorganisms were isolated from soil [26, 58, 64], contaminated crops [61, 65, 66], or animal feces [28]. Those biodegradation methods for aflatoxin detoxification may be limited in their potential application to food fermentation processes. However, the The AFB biodegradation by A. oryzae MAO 103 and Fig. 6. 1 two fungi from this study, A. oryzae MAO 103 and MAO MAO 104 on soybean-based agar plates. 104, could be directly applied to soybean fermentation Af, AFB production with A. flavus M 2034; Af+Ao 46464, AFB 1 1 because they were isolated from a food source, meju. production from A. flavus M 2034 co-cultured with A. oryzae KACC Moreover, A. oryzae is listed as a generally recognized as 46464; Af+Ao 103, AFB production from A. flavus M 2034 co-cultured 1 safe microorganism by the FDA [67, 68]. with A. oryzae MAO 103; and Af+Ao 104, AFB1 production from A. flavus M 2034 co-cultured with A. oryzae MAO 104. The results In conclusion, the AFB1 degradation activity of A. oryzae were obtained from triplicate tests. Asterisks indicate statistically MAO 103 and MAO 104 was 97.4% and 96.1%, respectively, significant differences between AFB1 production on Af plates and in 14-day culture media. The loss of AFB1 mutagenicity via those on Af+Ao 103 or Af+Ao 104 plates (**p < 0.01); error bars the strains was implicated as a potential method for correspond to the standard deviation. detoxification of AFB1 in food systems. A. oryzae MAO 103 and MAO 104 could be applied to fermented food systems

AFB1 by A. oryzae MAO 103 and MAO 104 as well as AFB 1 for controlling AFB1 contamination. This is the first report degradation mechanisms in a future study. that has identified the ability of AFB1 detoxification by A. oryzae.

Reduction of AFB1 Production by AFB1-Biodegrading A. oryzae MAO 103 and MAO 104 Acknowledgments We used soybean-based media to provide an artificial condition similar to that of Korean fermented soybean This research was supported by the Basic Science Research products such as doenjang. After 7 days of incubation, Program through the National Research Foundation of fungal growth on the soybean-based agar plates was Korea (NRF) funded by the Ministry of Education (NRF- observed, and the amounts of AFB1 in the plates were 2013R1A1A2013284). determined by HPLC. There seemed to be no fungal competition in growth between the A. flavus and A. oryzae References strains, A. oryzae KACC 46464, A. oryzae MAO 103, and A. oryzae MAO 104, except at the contact border lines of two 1.Hesseltine C, Shotwell OL, Ellis J, Stubblefield R. 1966. Aflatoxin formation by Aspergillus flavus. Bacteriol. Rev. 30: 795. fungal. A. flavus M 2034 produced 105.1 ng/g AFB1 in a 7-day culture, and A. flavus M 2034 co-cultured with A. oryzae 2.Weckbach L, Marth E. 1977. Aflatoxin production Aspergillus by KACC 46464 produced 93.1 ng/g, which was not significantly parasiticus in a competitive environment. Mycopathologia different in AFB production. However, A. flavus M 2034 62: 39-45. 1 3. Ito Y, Peterson SW, Wicklow DT, Goto T. 2001. Aspergillus co-cultured with A. oryzae MAO 103 and MAO 104 produced pseudotamarii, a new aflatoxin producing species in Aspergillus 34.3 and 48.0 ng/g AFB , respectively (Fig. 6). This indicates 1 section Flavi. Mycol. Res. 105: 233-239. that the AFB production in co-cultured soybean-based agar 1 4. Kurtzman C, Horn B, Hesseltine C. 1987. Aspergillus nomius, plates with the two AFB1-detoxifying isolates A. oryzae a new aflatoxin-producing species related to Aspergillus flavus MAO 103 and MAO 104 was significantly reduced to 32.6% and Aspergillus tamarii. Antonie Van Leeuwenhoek 53: 147-158. and to 45.7% (p < 0.01), respectively, compared with that 5. Corrier D. 1991. Mycotoxicosis: mechanisms of immuno-

J. Microbiol. Biotechnol. Aflatoxin B1-Detoxifying Aspergillus oryzae 65

suppression. Vet. Immunol. Immunopathol. 30: 73-87. 23.Samarajeewa U, Sen A, Cohen M, Wei C. 1990. Detoxification 6. Shimada T, Guengerich FP. 1989. Evidence for cytochrome of aflatoxins in foods and feeds by physical and chemical P-450NF, the nifedipine oxidase, being the principal enzyme methods. J. Food Prot. 53: 489-501. involved in the bioactivation of aflatoxins in human liver. 24.Teniola OD, Addo PA, Brost IM, Farber P, Jany KD,, Alberts JF Proc. Natl. Acad. Sci. USA 86: 462-465. et al. 2005. Degradation of aflatoxin B1 by cell-free extracts 7. Wong JJ, Hsieh D. 1976. Mutagenicity of aflatoxins related of Rhodococcus erythropolis and Mycobacterium fluoranthenivorans to their metabolism and carcinogenic potential. Proc. Natl. sp. nov. DSM44556(T). Int. J. Food Microbiol. 105: 111-117. Acad. Sci. USA 73: 2241-2244. 25. Shantha T. 1999. Fungal degradation of aflatoxin B1. Nat. 8. Coomes TJ, Crowther P, Feuell A, Francis B. 1966. Experimental Toxins 7: 175-178. detoxification of groundnut meals containing aflatoxin. 26.Hormisch D, Brost I, Kohring G-W, Giffhorn F, Kroppenstedt Nature 209: 406-407. R, Stackebradt E, et al. 2004. Mycobacterium fluoranthenivorans 9. Purchase I, Steyn M, Rinsma R, Tustin R. 1972. Reduction of sp. nov., a fluoranthene and aflatoxin B1 degrading bacterium the aflatoxin M content of milk by processing. Food Cosmet. from contaminated soil of a former coal gas plant. Syst. Toxicol. 10: 383-387. Appl. Microbiol. 27: 653-660. 10. Aibara K, Yano N. 1977. New approach to aflatoxin 27.Petchkongkaew A, Taillandier P, Gasaluck P, Lebrihi A. removal, pp. 151-161. In Rodricks JV, Hesseltine CW, 2008. Isolation of Bacillus spp. from Thai fermented soybean Mehlamn MA (eds.). Mycotoxins in Human and Animal (Thua-nao): screening for aflatoxin B1 and ochratoxin A Health. Pathotox Publishers Inc., Park Forest South, IL. detoxification. J. Appl. Microbiol. 104: 1495-1502. 11.Diaz DE, Hagler Jr WM, Blackwelder JT, Eve JA, Hopkins28.Zhao L, Guan S, Gao X, Ma Q, Lei Y, Bai X, Ji C. 2011. BA, Anderson KL, et al. 2004. Aflatoxin binders II: reduction Preparation, purification and characteristics of an aflatoxin of aflatoxin M1 in milk by sequestering agents of cows degradation enzyme from Myxococcus fulvus ANSM068. J. consuming aflatoxin in feed. Mycopathologia 157: 233-241. Appl. Microbiol. 110: 147-155. 12. Grant PG, Phillips TD. 1998. Isothermal adsorption of 29.Samuel MS, Sivaramakrishna A, Mehta A. 2014. Degradation aflatoxin B1 on HSCAS clay. J. Agric. Food Chem. 46: 599-605. and detoxification of aflatoxin B1 by Pseudomonas putida. Int. 13. Mishra H, Das C. 2003. A review on biological control and Biodeterior. Biodegradation 86: 202-209. metabolism of aflatoxin. Crit. Rev. Food Sci. Nutr. 43: 245. 30.Jung YJ, Chung SH, Lee HK, Chun HS, Hong SB. 2012. 14. Park DL. 2002. Effect of processing on aflatoxin, pp. 173- Isolation and identification of fungi from a meju contaminated 179. Mycotoxins and Food Safety. Springer, Berlin. with aflatoxins. J. Microbiol. Biotechnol. 22: 1740-1748. 15.Yazdanpanah H, Mohammadi T, Abouhossain G, Cheraghali31.Yang Z-Y, Shim W-B, Kim J-H, Park S-J, Kang S-J, Nam B-S, AM. 2005. Effect of roasting on degradation of aflatoxins in Chung D-H. 2004. Detection of aflatoxin-producing molds in contaminated pistachio nuts. Food Chem. Toxicol. 43: 1135-1139. Korean fermented foods and grains by multiplex PCR. J. 16.Sreenivasamurthy V, Parpia H, Srikanta S, Murti AS. 1967. Food Prot. 67: 2622-2626. Detoxification of aflatoxin in peanut meal by hydrogen 32.Móricz ÁM, Fatér Z, Otta KH, Tyihák E, Mincsovics E. 2007. peroxide. J. Assoc. Off. Anal. Chem. 50: 350-354. Overpressured layer chromatographic determination of 17.Dwarakanath C, Rayner E, Mann G, Dollear F. 1968. aflatoxin B1, B2, G1 and G2 in red paprika. Microchem. J. Reduction of aflatoxin levels in cottonseed and peanut 85: 140-144. meals by ozonization. J. Am. Oil Chem. Soc. 45: 93-95. 33. Stroka J, Van Otterdijk R, Anklam E. 2000. Immunoaffinity 18. Park DL, Lee L, Koltun S. 1984. Distribution of ammonia- column clean-up prior to thin-layer chromatography for the related aflatoxin reaction products in cottonseed meal. J. determination of aflatoxins in various food matrices. J. Am. Oil Chem. Soc. 61: 1071-1074. Chromatogr. A 904: 251-256. 19. Shantha T, Sreenivasa M, Rati E, Prema V. 1986. Detoxification 34.Var I, Kabak B, Gök F. 2007. Survey of aflatoxin B1 in helva, of groundnut seeds by urea and sunlight. J. Food Saf. 7: 225-231. a traditional Turkish food, by TLC. Food Control 18: 59-62. 20.Sen A, Wei C, Fernando S, Toth J, Ahmed E, Dunaif35. Glass G. NL, Donaldson GC. 1995. Development of primer sets 1988. Reduction of mutagenicity and toxicity of aflatoxin B1 designed for use with the PCR to amplify conserved genes by chlorine gas treatment. Food Chem. Toxicol. 26: 745-752. from filamentous ascomycetes. Appl. Environ. Microbiol. 61: 21. Doyle M, Marth E. 1978. Bisulfite degrades aflatoxin: effect 1323-1330. of temperature and concentration of bisulfite. J. Food Prot. 36. Tao L, Chung SH. 2014. Non-aflatoxigenicity of commercial 41: 774-780. Aspergillus oryzae strains due to genetic defects compared to 22. Phillips TD, Clement BA, Park DL. 1994. Approaches to aflatoxigenic Aspergillus flavus. J. Microbiol. Biotechnol. 24: Reduction of aflatoxins in foods and feeds, pp. 383-406. In 1081-1087. Eaton LD, Groopman JD (eds.). The Toxicology of Aflatoxins: 37. Hong S-B, Kim D-H, Samson RA. 2015. Aspergillus associated Human Health, Veterinary, and Agricultural Significance. with meju, a fermented soybean starting material for Academic Press, New York. traditional soy sauce and soybean paste in Korea. Mycobiology

January 2017 ⎪ Vol. 27⎪ No. 1 66 Lee et al.

43: 218-224. dietary factors on the mutagenicity of aflatoxin B1: in vitro 38.Tamura K, Peterson D, Peterson N, Stecher G, Nei M,effect of plant flavonoids. Mutat. Res. 222: 393-401. Kumar S. 2011. MEGA5: molecular evolutionary genetics 54.Mandal S, Ahuja A, Shivapurkar NM, Cheng S-J, Groopman analysis using maximum likelihood, evolutionary distance, and JD, Stoner GD. 1987. Inhibition of aflatoxin B1 mutagenesis maximum parsimony methods. Mol. Biol. Evol. 28: 2731-2739. in Salmonella typhimurium and DNA damage in cultured rat and 39. Davis ND, Diener U, Eldridge D. 1966. Production of human tracheobronchial tissues by ellagic acid. Carcinogenesis aflatoxins B1 and G1 by Aspergillus flavus in a semisynthetic 8: 1651-1656. medium. Appl. Microbiol. 14: 378-380. 55. Lee L, Dunn J, DeLucca A, Ciegler A. 1981. Role of lactone 40.Abbas HK, Zablotowicz RM, Bruns HA, Abel CA. 2006. ring of aflatoxin B1 in toxicity and mutagenicity. Experientia Biocontrol of aflatoxin in corn by inoculation with non- 37: 16-17. aflatoxigenic Aspergillus flavus isolates. Biocontrol Sci. Technol. 56.Wehner F, Thiel P, Van Rensburg S, Demasius IP. 1978. 16: 437-449. Mutagenicity to Salmonella typhimurium of some Aspergillus 41.Sugimura T, Kawachi T, Nagao M, Yahagi T, Seino Y,and Penicillium mycotoxins. Mutat. Res. 58: 193-203. Okamoto T, et al. 1977. Mutagenic principle(s) in tryptophan and 57. Wogan G, Edwards G, Newberne P. 1971. Structure-activity phenylalanine pyrolysis products. Proc. Jpn. Acad. 53: 58-61. relationships in toxicity and carcinogenicity of aflatoxins 42.Choi S, Jun H, Bang J, Chung S-H, Kim Y,, etKim al. B-S and analogs. Cancer Res. 31: 1936-1942. 2015. Behaviour of Aspergillus flavus and Fusarium graminearum 58.Alberts JF, Engelbrecht Y, Steyn PS, Holzapfel WH, van Zyl on rice as affected by degree of milling, temperature, and WH. 2006. Biological degradation of aflatoxin B1 by Rhodococcus relative humidity during storage. Food Microbiol. 46: 307-313. erythropolis cultures. Int. J. Food Microbiol. 109: 121-126. 43.Filtenborg O, Frisvad JC, Svendsen JA. 1983. Simple screening59.Alberts JF, Gelderblom WC, Botha A, van Zyl WH. 2009. method for molds producing intracellular mycotoxins in Degradation of aflatoxin B1 by fungal laccase enzymes. Int. pure cultures. Appl. Environ. Microbiol. 45: 581-585. J. Food Microbiol. 135: 47-52. 44. Scheidegger K, Payne G. 2003. Unlocking the secrets behind 60.Sangare L, Zhao Y, Folly YME, Chang J, Li J, ,Selvaraj et JN secondary metabolism: a review of Aspergillus flavus from al. 2014. Aflatoxin B1 degradation by a Pseudomonas strain. pathogenicity to functional genomics. J. Toxicol. Toxin Rev. Toxins 6: 3028-3040. 22: 423-459. 61.Nakazato M, Morozumi S, Saito K, Fujinuma K, Nishima T, 45. Dalié D, Deschamps A, Richard-Forget F. 2010. Lactic acid Kasai N. 1990. Interconversion of aflatoxin B1 and aflatoxicol bacteria – potential for control of mould growth and by several fungi. Appl. Environ. Microbiol. 56: 1465-1470. mycotoxins: a review. Food Control 21: 370-380. 62. Iram W, Anjum T, Iqbal M, Ghaffar A, Abbas M. 2015. Mass 46.Haskard CA, El-Nezami HS, Kankaanpää PE, Salminen S, spectrometric identification and toxicity assessment of Ahokas JT. 2001. Surface binding of aflatoxin B1 by lactic degraded products of aflatoxin B1 and B2 by Corymbia acid bacteria. Appl. Environ. Microbiol. 67: 3086-3091. citriodora aqueous extracts. Sci. Rep. 5: 14672. 47. Shetty PH, Hald B, Jespersen L. 2007. Surface binding of 63.Liu D-L, Yao D-S, Liang R, Ma L, Cheng W-Q, Gu L-Q. aflatoxin B1 by Saccharomyces cerevisiae strains with potential 1998. Detoxification of aflatoxin B1 by enzymes isolated decontaminating abilities in indigenous fermented foods. from Armillariella tabescens. Food Chem. Toxicol. 36: 563-574. Int. J. Food Microbiol. 113: 41-46. 64.Abbas HK, Williams WP, Windham GL, Pringle HC, Xie W, 48.Czeczot H, Tudek B, Kusztelak J, Szymczyk T, Dobrowolska Shier WT. 2002. Aflatoxin and fumonisin contamination of B, Glinkowska G, et al. 1990. Isolation and studies of the commercial corn (Zea mays) hybrids in Mississippi. J. Agric. mutagenic activity in the Ames test of flavonoids naturally Food Chem. 50: 5246-5254. occurring in medical herbs. Mutat. Res. Genet. Toxicol. 240: 65.Dorner JW, Cole RJ, Blankenship PD. 1998. Effect of inoculum 209-216. rate of biological control agents on preharvest aflatoxin 49. Florin I, Rutberg L, Curvall M, Enzell CR. 1980. Screening contamination of peanuts. Biol. Control 12: 171-176. of tobacco smoke constituents for mutagenicity using the 66.Farzaneh M, Shi Z-Q, Ghassempour A, Sedaghat N, Ames’ test. Toxicology. 15: 219-232. Ahmadzadeh M, Mirabolfathy M, Javan-Nikkhah M. 2012. 50. Maron DM, Ames BN. 1983. Revised methods for the Aflatoxin B1 degradation by Bacillus subtilis UTBSP1 isolated Salmonella mutagenicity test. Mutat. Res. 113: 173-215. from pistachio nuts of Iran. Food Control 23: 100-106. 51. Reifferscheid G, Heil J. 1996. Validation of the SOS/umu 67. Abe K, Gomi K, Hasegawa F, Machida M. 2006. Impact of test using test results of 486 chemicals and comparison with Aspergillus oryzae genomics on industrial production of the Ames test and carcinogenicity data. Mutat. Res. 369: 129-145. metabolites. Mycopathologia 162: 143-153. 52. Vargas VMF, Motta V, Henriques JAP. 1993. Mutagenic 68. Taylor MJ, Richardson T. 1979. Applications of microbial activity detected by the Ames test in river water under the enzymes in food systems and in biotechnology. Adv. Appl. influence of petrochemical industries. Mutat. Res. 319: 31-45. Microbiol. 25: 7. 53. Francis A, Shetty T, Bhattacharya R. 1989. Modifying role of

J. Microbiol. Biotechnol.