Transcriptional Regulation of Aflatoxin Biosynthesis and Conidiation in Aspergillus flavus by Wickerhamomyces anomalus WRL-076 for Reduction of Aflatoxin Contamination

Aspergillus flavus is a ubiquitous saprophytic fungus found in soils across the world. The fungus is the major producer of aflatoxin (AF) B1, which is toxic and a potent carcinogen to humans. Aflatoxin B1 (AFB1) is often detected in agricultural crops such as corn, peanut, almond, and pistachio. It is a serious and recurrent problem and causes substantial economic losses. Wickerhamomyces anomalus WRL-076 was identified as an effective biocontrol yeast against A. flavus. In this study, the associated molecular mechanisms of biocontrol were investigated. We found that the expression levels of eight genes, aflR, aflJ, norA, omtA, omtB, pksA, vbs, and ver-1 in the aflatoxin biosynthetic pathway cluster were suppressed. The decreases ranged from several to 10,000 fold in fungal samples co-cultured with W. anomalus. Expression levels of conidiation regulatory genes brlA, abaA, and wetA as well as sclerotial regulatory gene (sclR) were all down regulated. Consistent with the decreased gene expression levels, aflatoxin concentrations in cultural medium were reduced to barely detectable. Furthermore, fungal biomass and conidial number were significantly reduced by 60% and more than 95%, respectively. The results validate the biocontrol efficacy of W. anomalus WRL-076 observed in the field experiments.


Introduction
Aspergillus flavus is a saprophytic and pathogenic fungus. Many isolates of A. flavus produce the hepatocarcinogenic aflatoxin (AF) B 1 , which is often detected in agricultural crops including corn, cotton, peanuts, and tree nuts, and in many dried fruits and spices. AFB 1 contamination in food results in substantial economic losses worldwide [1][2][3][4][5][6][7]. Therefore, more than 100 countries have established specific regulation guidelines limiting allowable amounts of AFB 1 in foodstuffs [8,9]. Major importers of agricultural commodities have imposed threshold levels for AFB 1 below 10 µg/kg, and these restrictions have had a major negative impact on the exportability of a number of crops [10].
samples collected at 24 h. However, the repression of gene transcription was leveled off at 48 h and 72 h for CA90, but for M52, the repression was observed at 48 h and continued at 72 h. The Smorphotype strains CA28 and CA42 showed a moderate repression on pksA, norA, omtA, omtB, vbs, and ver-1 in the range under 100 fold compared to the L-morphotype strains. Apparently, the entire gene cluster of aflatoxin biosynthesis was affected by W. anomalus. The global regulatory gene, veA was repressed a few fold. The repression level was also strain and time dependent.

Effect of W. anomalus on Transcription of Genes of Conidiation and Sclerotial Formation
AF biosynthesis and Aspergillus development are closely associated processes [38]. We examined the expression levels of brlA, abaA, and wetA in four A. flavus S-and L-strains. The expression levels of three central regulatory genes of conidiation were decreased in co-cultures of A. flavus and yeast compared to those from A. flavus cultures without yeast. The expression of brlA level was several folds lower for the L-morphotype (CA90, M52) at 24 h and 48 h than the S-morphotype, CA28 and CA42. Both had the greatest decrease at 48 h. The wetA expression levels had the lowest decreases in comparison to brlA and abaA. Repression of these three genes in CA28, CA42, CA90, and M52 were variable depending on the strains. The four A. flavus strains showed a peak decrease either at 24 h, 48 h, or 72 h (see Figure 2).
SclR is a transcription factor for sclerotial formation. The veA gene positively regulates the production of aflatoxin and conidial and sclerotial formation. Transcriptional levels of the veA gene were repressed in both S and L strains of A. flavus (see Figure 2). comparison to brlA and abaA. Repression of these three genes in CA28, CA42, CA90, and M52 were variable depending on the strains. The four A. flavus strains showed a peak decrease either at 24 h, 48 h, or 72 h (see Figure 2).
SclR is a transcription factor for sclerotial formation. The veA gene positively regulates the production of aflatoxin and conidial and sclerotial formation. Transcriptional levels of the veA gene were repressed in both S and L strains of A. flavus (see Figure 2).

Inhibition of Aflatoxin Production in Yeast and A. flavus Dual Cultures
We examined the influence of W. anomalus on AFB1 production on the four toxigenic A. flavus strains including two L-and two S-morphotypes [50]. AFB1 concentrations of these strains grown in potato dextrose broth (PDB) ranged from 2.6, 3.5, 8.5, to 9.0 µg/culture for CA90, M52, CA42, and CA28, respectively. The two S-morphotypes, CA28 and CA42 produced a higher amount of AFB1 than the two L-morphotypes, CA90 and M52. No aflatoxin was detected when these strains were grown in the presence of W. anomalus (Table S1). The aflatoxin produced from dual cultures of toxigenic A. flavus and W. anomalus was significantly lower than that from the toxigenic A. flavus control at a p-value < 0.05 by ANOVA test. The results demonstrated the yeast biocontrol agent W. anomalus WRL-076 is effective in inhibiting aflatoxin biosynthesis.
The decrease of aflatoxin B1 concentrations produced by A. flavus strains was due to the down regulation of the expression of the entire clustered aflatoxin biosynthetic genes.

Inhibition of Aflatoxin Production in Yeast and A. flavus Dual Cultures
We examined the influence of W. anomalus on AFB 1 production on the four toxigenic A. flavus strains including two L-and two S-morphotypes [50]. AFB 1 concentrations of these strains grown in potato dextrose broth (PDB) ranged from 2.6, 3.5, 8.5, to 9.0 µg/culture for CA90, M52, CA42, and CA28, respectively. The two S-morphotypes, CA28 and CA42 produced a higher amount of AFB 1 than the two L-morphotypes, CA90 and M52. No aflatoxin was detected when these strains were grown in the presence of W. anomalus (Table S1). The aflatoxin produced from dual cultures of toxigenic A. flavus and W. anomalus was significantly lower than that from the toxigenic A. flavus control at a p-value < 0.05 by ANOVA test. The results demonstrated the yeast biocontrol agent W. anomalus WRL-076 is effective in inhibiting aflatoxin biosynthesis.
The decrease of aflatoxin B 1 concentrations produced by A. flavus strains was due to the down regulation of the expression of the entire clustered aflatoxin biosynthetic genes.

Reduction of Fungal Biomass and Number of Conidial Formation
The fungal mass of CA28, CA42, CA90, and M52 was reduced when co-cultured with W. anomalus. The percentage of reduction was 62, 60, 56, and 80%, respectively (Table S2). The percentage of reduction of fungal biomass samples of dual cultures of A. flavus with W. anomalus was significantly lower than toxigenic A. flavus alone at p-value < 0.05 by ANOVA Duncan's multiple range test.
The fungal conidia formed on fungal balls after two weeks of dual cultures of A. flavus with W. anomalus were significantly lower than the A. flavus control at a p-value < 0.05 by ANOVA test for CA42 and CA42+WRL-076, CA90 and CA90+WRL-076, and M52 and M52+WRL-076. However, the difference between CA28 and CA28+WRL-076 was not significant.
Aspergillus flavus primarily reproduces by forming asexual spores called conidia, whose formation and maturation are governed by the central genetic regulatory circuit consisting of BrlA, AbaA, and WetA. Genes encoding the regulators were repressed when co-cultured with W. anomalus, resulting in a significant reduction of A. flavus conidial production ( Figure 3, Table S3).
The veA gene positively regulates the production of aflatoxin, and conidia and sclerotial formation. Both veA and sclR were down regulated by W. anomalus and their transcriptional levels decreased several folds in both S-and L-strains. We did not detect any sclerotia in CA28 and CA42 (S-strains) and CA90 and M52 (L-strains).
The fungal conidia formed on fungal balls after two weeks of dual cultures of A. flavus with W. anomalus were significantly lower than the A. flavus control at a p-value < 0.05 by ANOVA test for CA42 and CA42+WRL-076, CA90 and CA90+WRL-076, and M52 and M52+WRL-076. However, the difference between CA28 and CA28+WRL-076 was not significant.
Aspergillus flavus primarily reproduces by forming asexual spores called conidia, whose formation and maturation are governed by the central genetic regulatory circuit consisting of BrlA, AbaA, and WetA. Genes encoding the regulators were repressed when co-cultured with W. anomalus, resulting in a significant reduction of A. flavus conidial production ( Figure 3, Table S3).
The veA gene positively regulates the production of aflatoxin, and conidia and sclerotial formation. Both veA and sclR were down regulated by W. anomalus and their transcriptional levels decreased several folds in both S-and L-strains. We did not detect any sclerotia in CA28 and CA42 (S-strains) and CA90 and M52 (L-strains).

Conclusion
Transcription of AF biosynthetic genes and conidial regulatory genes in A. flavus were both down regulated. Consistent with the decreased gene expression levels, the aflatoxin concentrations in cultural medium were greatly reduced to non-detectable levels. Fungal biomass and the number of conidia were significantly reduced by 60% and more than 95%, respectively. However, the biocontrol yeast cells from fungal ball of dual cultures grew and reached 1 to 2 × 10 8 CFU /mL (Table  S2 and Table S3). The data demonstrate that W. anomalus is a robust biocontrol agent.
The Food and Agriculture Organization (FAO) of the United Nations estimates that 25% of the world's food crops are affected by mycotoxins. Contamination by mycotoxins such as aflatoxin in tree

Conclusions
Transcription of AF biosynthetic genes and conidial regulatory genes in A. flavus were both down regulated. Consistent with the decreased gene expression levels, the aflatoxin concentrations in cultural medium were greatly reduced to non-detectable levels. Fungal biomass and the number of conidia were significantly reduced by 60% and more than 95%, respectively. However, the biocontrol yeast cells from fungal ball of dual cultures grew and reached 1 to 2 × 10 8 CFU /mL (Table S2 and Table S3). The data demonstrate that W. anomalus is a robust biocontrol agent.
The Food and Agriculture Organization (FAO) of the United Nations estimates that 25% of the world's food crops are affected by mycotoxins. Contamination by mycotoxins such as aflatoxin in tree nuts, peanut, corn, and cottonseed is a serious food safety hazard to both humans and animals. The results of this study demonstrate that W. anomalus is a promising biocontrol agent to reduce aflatoxin, conidia, and sclerotia of A. flavus in agricultural production of crops.

Microbial Strains and Media
W. anomalus WRL-076 and A. flavus strains, CA28, CA42, CA90, and M52, were maintained on potato dextrose agar (PDA, Becton Dickinson & Co., Sparks, MD) at 4 • C. The CA28 and CA42 strains produced small sclerotia (S-morphotype) and CA90 and M52 produced large sclerotia (L-morphotype) as classified by Cotty [50]. Suspensions of yeast and fungal spores were prepared in 0.05% Tween 80 solution and counted using a hemocytometer. Potato dextrose broth (PDB) was the medium used to grow yeast and fungus for investigating the biocontrol antagonistic activities.

Experimental Design
A. flavus spores were inoculated into 25 mL of PDB (to a final concentration of 10 5 /mL) and grown at 28 • C in triplicates on a rotary shaker at 150 rpm. For dual culture, yeast (W. anomalus) cells and fungal spores in a ratio of 1:1 were used. Fungal hyphae were collected at 24 h, 48 h, and 72 h after inoculation. Yeast cells from dual cultures were separated from the fungal hyphae by filtering through the Cellector tissue sieve with 38.1 µm pore size (VWR Scientific, Brisbane, CA, USA) [25]. The hyphae on the sieve was rinsed with DEPC (0.1% diethylpyrocarbonate) water and transferred to several layers of filter paper with suction, dried, and stored at −80 • C. The processed fungal hyphae were used for total RNA extraction.
Total fungal RNA isolation was carried out by using RNeasy ® Plant Mini Kit (Qiagen, Valencia, CA, USA). The RNA samples were treated with Ambion ® TURBO DNA-free™ DNase (Ambion, Austin, TX, USA). The purity and concentration of fungal RNA were examined by measuring the absorbance of samples at 260 nm and 280 nm using an ND-1000 Spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA). Samples were stored in a −80 • C freezer. GeneAmp ® RNA PCR Core Kit (Applied Biosystems) was used for reverse transcription to obtain cDNA according to the manufacturer's procedure. For negative control, the same reactions were performed in the absence of the enzyme.

Real Time RT-PCR Analysis of AF Biosynthetic Genes and Conidia Regulatory Genes
Primers were designed with ABI Primer Express 3.0 software (Applied Biosystems, Foster City, CA, USA). Primers for RT-PCR are listed in Table 3 [26,51]. Quantitative PCR reactions were carried out in an ABI 7300 Real Time PCR System. SYBR ® Green PCR Master Mix (Applied Biosystems, Foster City, CA USA), which increases fluorescence upon binding to double-stranded DNA product, was used as the amplification detector. Triplicates of each reaction were performed. The final primer concentration was 500 nM in the 25 µL reaction mixture. Input cDNA quantities in the reaction mixture were within the recommended 150 ng. The PCR cycles were programmed as follows-2 min at 50 • C for AmpErase ® UNG Activation, 10 min at 95 • C for AmpliTaq Gold ® DNA polymerase activation, followed by 40 cycles of 15 s at 95 • C and 1 min at 60 • C for both primer annealing and product extension. Melting curve analysis was performed using Dissociation Curves software (Applied Biosystems) to ensure only a single product was amplified. Amplification of A. flavus 18S ribosomal RNA was used as the endogenous control (reference gene) due to its relatively stable expression level. Plates and quantification assay documents were created in SDS ® Software 1.3.1 (Applied Biosystems). The relative quantification of gene expression changes was computed by using the 2 ∆∆Ct method [52][53][54]. AFB 1 was extracted from a 2 mL liquid fungal and yeast co-culture by adding 1 mL of acetonitrile into the conical tube, vortex for 10 min, and 0.5 mL of the supernatant was filtered through SINGLE StEP TM eXtreme/FV 0.45 mm Nylon (Thomson Instrument Company; Oceanside, CA). Filtered samples were analyzed by high performance liquid chromatography (HPLC) on an Agilent model 1260 Infinity ChemStation (Agilent, Palo Alto, California, USA). HPLC was performed on a Supelcosil LC-18 reversed-phase column (150 mm × 4.6 mm i.d., 5 µm particle size) at a flow rate of 1 mL/min.
The mobile phase was methanol/acetonitrile/H 2 O (20:20:60). Aflatoxins were quantified by a fluorescent detector with excitation at 365 nm and emission at 455 nm and quantified by peak areas relative to a standard curve of authentic AFB 1 [55]. Aflatoxin standards were purchased from Sigma-Aldrich (St. Louise, MO, USA).

Determination of Fungal Biomass and Conidia Numbers
A. flavus hyphae grown in PDB with shaking in triplicate flasks formed tiny fungal balls and increased in size over the time of incubation. Dual cultures of A. flavus and W. anomalus also formed fungal balls. After incubating the cultures for 72 h, fungal balls from each flask with and without W. anomalus were collected on a meshed screen rinsed with sterile water and transferred to an empty Petri Dish to induce conidiation. The harvested fungal balls were weighed and then incubated at 28 • C for two weeks [25]. Conidia (spore) were then extracted in 5 mL of 0.05% Tween 80 solution, and conidia and yeast cells were counted using a hemocytometer.

Statistical Analysis
Statistical analyses were performed with SAS Enterprise Guide (version 6.1, SAS Institute Inc., Cary, NC, USA). ANOVA (one-way analysis of variance) by Duncan's multiple range test at a 95% confidence level (p-value < 0.05) was performed on all the samples.