Functional analyses of the versicolorin B synthase gene in Aspergillus flavus

Abstract Aflatoxin is a toxic, carcinogenic mycotoxin primarily produced by Aspergillus parasiticus and Aspergillus flavus. Previous studies have predicted the existence of more than 20 genes in the gene cluster involved in aflatoxin biosynthesis. Among these genes, aflK encodes versicolorin B synthase, which converts versiconal to versicolorin B. Past research has investigated aflK in A. parasiticus, but few studies have characterized aflK in the animal, plant, and human pathogen A. flavus. To understand the potential role of aflK in A. flavus, its function was investigated here for the first time using gene replacement and gene complementation strategies. The aflK deletion‐mutant ΔaflK exhibited a significant decrease in sclerotial production and aflatoxin biosynthesis compared with wild‐type and the complementation strain ΔaflK::aflK. ΔaflK did not affect the ability of A. flavus to infect seeds, but downregulated aflatoxin production after seed infection. This is the first report of a relationship between aflK and sclerotial production in A. flavus, and our findings indicate that aflK regulates aflatoxin formation.


| INTRODUCTION
The genus Aspergillus is a family of filamentous fungi with worldwide distribution, which is well-studied for its important secondary metabolites, both beneficial and harmful. Within this genus, Aspergillus flavus infects a wide range of plants, animals, and humans as a diseasecausing pathogen (Amaike & Keller, 2011;Shieh et al., 1997). It can cause serious agricultural problems by contaminating important crops and producing extremely carcinogenic and highly toxic secondary metabolites known as aflatoxins (Amaike & Keller, 2011). Crops contaminated with aflatoxins pose a serious or even fatal threat to animals and humans. The gene cluster for aflatoxin biosynthesis in both A. flavus and Aspergillus parasiticus has been reported to encode at least 27 enzymes and regulatory factors within a 70 kb region (Yabe & Nakajima, 2004).
The functions of many genes within this cluster have been previously explored. For example, aflX is required for the conversion of versicolorin A, similar to the function of aflN, aflM, and aflY (Cary, Ehrlich, Bland, & Montalbano, 2006). The expression of aflR and aflQ of the aflatoxins biosynthesis cluster was analyzed Sweeney, Pamies, & Dobson (2000) using reverse transcription PCR, while Mayer, Bagnara, Farber, & Geisen (2003)  also been compared (Ehrlich & Mack, 2014). Furthermore, aflatoxin *These authors contributed to this work equally. biosynthesis gene expression was determined relative to water activity and temperature (Schmidt-Heydt, Abdel-Hadi, Magan, & Geisen, 2009), and we previously conducted a transcriptome analysis of A. flavus in response to water activity (Zhang et al., 2014).
Of the aflatoxin biosynthesis genes, vbs, also known as aflK, codes versicolorin B synthase (VBS) which is responsible for the conversion of versiconal to versicolorin B (Yu, Bhatnagar, & Ehrlich, 2002). vbs has been heterologous expressed, purified, isolated, and characterized in A. parasiticus (McGuire, Silva, Casillas, & Townsend, 1996;Silva, Minto, Barry, Holland, & Townsend, 1996;Silva & Townsend, 1997), and the distribution and subcellular localization of VBS was shown to change with respect to culture time in A. parasiticus (Chiou et al., 2004). Linz, Wee, & Roze (2014) hypothesized that VBS transport is tightly regulated by the timing and synthesis level of versicol- Almost all studies on aflK were performed in A. parasiticus, while few have investigated aflK in A. flavus. Therefore, to understand the potential role of aflK in A. flavus, we constructed the first known aflK deletion mutant and its complementation strain of A. flavus. We then analyzed the function of aflK, including its effect on growth, the formation of conidia and sclerotia, aflatoxin biosynthesis, and host colonization of A. flavus.

| Phylogenetic tree and domain architecture
The protein sequence of AflK from A. flavus was identified and used in a BLAST search on the National Center for Biotechnology Information (NCBI) website (http://www.ncbi.nlm.nih.gov/). All available homogenous sequences from different organisms were downloaded and used to construct a phylogenetic tree with MEGA6.0 software and the neighbor-joining method (Tamura, Stecher, Peterson, Filipski, & Kumar, 2013). Bootstrap analysis was set as 1,000 replicates. To visualize the AflK domain, information from the SMART database (http:// smart.embl-heidelberg.de/) was subjected to DOG2.0 .

| Fungal strains and media
Aspergillus flavus CA14PTSΔpyrG, a uracil auxotrophic, was obtained from Prof. Chang P. K. (Chang, Scharfenstein, Wei, & Bhatnagar, 2010 Academy of Agricultural Sciences, Beijing, China). Aspergillus flavus was cultured on yeast extract sucrose (YES) agar plates for fungal growth, and YGTUU media was applied for the preparation of protoplasts. Yeast extract glucose (YGT) agar was used to screen gene deletion strains and to prepare protoplasts of deletion strains. Czapek's agar with pyrithiamine was used to screen complementation strains.
Transformation was performed according to previously published techniques with minor modifications (Cary et al., 2006;He et al., 2007;Szewczyk et al., 2006). Protoplasts, the gel extraction product of fu- The plates were cultured at 37°C for about 3 days.

| Construction of deletion and complementation mutants
A previous publication was used as a reference to construct the aflK deletion strain ΔaflK and complementation strain ΔaflK::aflK (Ren et al., 2016). All gene sequences, including the open reading frame of aflK, the upstream and downstream sequence of aflK, and the pyrG sequence of A. fumigatus Af293, were searched for and downloaded from the NCBI website and Aspergillus Comparative Database (http:// www.broadinstitute.org/annotation/genome/aspergillus_group/ MultiHome.html). The PCR fusion program was performed as described by Szewczyk et al. (2006), and the PCR-amplified complementary fragment was inserted into chromosomal integrating shuttle vector pPTR I (containing a pyrithiamine resistance gene as the selection marker, TAKARA, Japan) using Kpn I and Hind III (Thermo, USA) digestion and the ClonExpress II One Step Cloning Kit (C112-01, Vazyme Biotech Co., Ltd, Nanjing, China). After bacterial transformation and PCR verification, the vectors were transformed into protoplasts of ΔaflK to construct the complementary strain ΔaflK::aflK. pyrG was amplified from A. fumigatus Af293 and transformed into A. flavus CA14PTSΔpyrG protoplasts, which were then used as wild-type (WT).

| RNA extraction and reverse transcription (RT)-PCR
The mycelia of WT, ΔaflK, and ΔaflK::aflK were harvested after cultured on YES agar for 48 hr. After grinding in liquid nitrogen, the result powder was suspended with TRIzol reagent (Biomarker Technologies, Beijing, China) for RNA extract. RNA molecules were purified with the DNA-free kit (TransGen Biotech, Beijing, China) to remove genomic DNA. First-strand cDNA was synthesized with the TransScript ® Allin-One First-Strand cDNA Synthesis SuperMix.

| Analysis of colony growth, conidial production, and sclerotial formation
To record the colony growth, 2 μl of 10 6 spores/ml suspension of WT, ΔaflK, and ΔaflK::aflK were inoculated onto YES media. The diameters were measured until colonies filled the plate. For the enumeration of conidia, 2 μl spores suspension was inoculated onto PDA media, and cultured for 6 days at 37°C. Three pieces of 1-cm diameter were harvested from the edge to the centre of each colony, and homogenized in 3 ml of distilled water. Then, spore number was counted manually with a hemocytometer. Conidiophores were observed according to a previous report . To analyze sclerotial production, 2 μl spores suspension of the three strains was cultured at 37°C for 10 days, then 75% ethanol was used to wash away conidia for the enumeration of sclerotia . The sclerotia were also observed microscopically. Each experiment was performed three times using four replicates. One-way analysis of variance (ANOVA) was used for statistical testing.

| Aflatoxin analysis
For aflatoxin biosynthesis analysis, 10 6 spores of the three strains were incubated into YES liquid media, and cultured at 28°C for 6 days in the dark with shaking at 180 rpm. Aflatoxin was extracted with chloroform, then 10 ml of chloroform was transferred to a new tube and evaporated to dryness. Thin-layer chromatography (TLC) was used to analyze aflatoxin production, and the outcome observed under ultraviolet light at 365 nm. The JD-801 Computer-aided Image Analysis System (JEDA Co., Nanjing, China) was used for the quantitative analysis of aflatoxin biosynthesis . Each experiment was performed three times using four replicates and analyzed by one-way ANOVA. High-performance liquid chromatography

| Seed infection
The ability of the three strains to infect peanut and maize was determined as described previously (Tsitsigiannis & Keller, 2006). After asepsis with 0.05% sodium hypochlorite and 75% ethanol, peanut cotyledons and maize kernels were inoculated with 10 5 spores/ml suspension for 30 min, with shaking at 80 rpm. Then 20 cotyledons and 10 maize kernels were placed in culture dishes lined with three pieces of moist sterile filter paper to maintain humidity. A blank control was included in which the cotyledons within sterile water. Peanuts and maize were incubated at 28°C for 4 days in the dark, and humidity was maintained. Peanut seeds and maize kernels were then harvested in 50 ml tubes, and vortexed to release the spores into 15 ml of 0.05% Tween 20 (v/v in water); conidia were counted hemocytometrically.
An equal amount of chloroform was used to extract aflatoxins. Each experiment was performed three times using four replicates and analyzed by one-way ANOVA.  to 635th amino acid) and D-amino acid oxidase (DAO) domain (from 76th to 382nd amino acid) or the N-terminal of GMC oxidoreductase (GMC_oxred_N, from 75th to 394th amino acid) (Figure 1b). These results suggested that AflK was conserved but relatively unique.  Table 1, and a schematic of this method is given in Figure 2a.  Figure 3c shows that the conidiophores of these strains cultured on YES agar were identical. These results indicated that aflK does not affect the growth and conidia production of A. flavus.

| Reduction of aflatoxin biosynthesis in ΔaflK
aflK is located in the aflatoxin synthesis gene cluster of A. flavus, so we next investigated aflatoxin biosynthesis after the deletion of aflK.
To examine the role played by aflK in aflatoxin synthesis, we tested aflatoxin production in WT, ∆aflK, and ∆aflK::aflK by TLC at 6 th day.
The aflatoxin B 1 (AFB 1 ) produced by ΔaflK was reduced compared with that of WT and ΔaflK::aflK (Figure 4a), and Figure 4b shows the quantitative analysis of AFB 1 production. HPLC was also used to confirm these results, which indicated that AFB 1 production of ΔaflK was lower than that of WT and ΔaflK::aflK strains (Figure 4c). The original maps of TLC and HPLC are shown as the representative result. These findings suggested that aflK plays a vital role in aflatoxin biosynthesis in A. flavus.

| Negative role of aflK in sclerotial production
Previous studies have shown that sclerotial production may be associated with aflatoxin biosynthesis (Calvo et al., 2004;Duran, Cary, & Calvo, 2007). Therefore, we determined the effect of aflK on sclerotial production by culturing the strains on Wickerham agar at 37°C for 7 days in the dark. The conidia were washed away with 75% ethanol to visualize the sclerotial phenotypes, as shown in Figure 5a and b. Surprisingly,

| Seed infections of WT, ΔaflK, and ΔaflK::aflK
A. flavus has a high number of diverse virulence factors and is a biological hazard to crops (Amaike & Keller, 2011). Based on the observed changes in aflatoxin production and sclerotial production caused by aflK deletion, we determined whether the aflK deletion also affected the ability of the strain to infect seeds. Peanut and maize were treated with conidia from WT, ΔaflK, and ΔaflK::aflK strains, but no obvious infection ability of A. flavus was observed (Figure 6a), and there was no significant difference in the number of conidia after infection of seeds by WT, ΔaflK, and ΔaflK::aflK (Figure 6b). Aflatoxin was extracted
Our research focused not only on the role of aflK in aflatoxin biosynthesis, but also in development and infection of A. flavus.
Our results showed that the deletion of aflK had no effect on growth and conidiation of A. flavus. However, aflK appeared to play an important role in aflatoxin production, which is consistent with previous studies about aflatoxin biosynthesis (Yabe & Nakajima, 2004;Yu et al., 2002Yu et al., , 2004. Moreover, the aflK deletion did not completely inhibit the production of aflatoxin, as previously seen following the deletion of A. parasiticus aflE (Trail, Chang, Cary, & Linz, 1994 found that the inactivation of aflK decreased sclerotial production in A. flavus. The synchronous variation between sclerotia and aflatoxin was comparable with previous studies of veA and laeA (Amaike & Keller, 2009;Calvo et al., 2004;Duran et al., 2007;Kale et al., 2008).
Aflatoxin biosynthesis in A. flavus and A. parasiticus was previously shown to be regulated by veA and laeA, which are also necessary for sclerotial formation (Calvo et al., 2004;Kale et al., 2008). We therefore speculate that sclerotial formation was decreased in response to the deletion of aflK, under the effects of these regulatory genes. But this is the first time to reveal the effect of aflatoxin biosynthesis gene aflK on sclerotial formation. Our results therefore indicate that aflK has an important role in aflatoxin biosynthesis and sclerotial production in A. flavus.
We found that an aflK deletion did not affect the infection of A. flavus, but caused a notable decrease in aflatoxin production after infection of peanut and maize. The aflK deletion also had no effect on conidia production after infection, consistent with our findings that aflK did not affect growth or conidia production, but regulated aflatoxin biosynthesis of A. flavus. This could reflect the fact that A. flavus virulence is a multifactorial process closely connected with secondary metabolism, developmental linkage of sporulation, and other aspects of A. flavus (Amaike & Keller, 2011). Hence, because aflK is only one part of the aflatoxin biosynthesis gene cluster, it did not influence the infection of A. flavus, but decreased aflatoxin production after infection.
In conclusion, AflK appears to be a conservative and unique oxidoreductase of Aspergillus sp. This is the first report of the influence of aflK on both sclerotial formation and aflatoxin biosynthesis in A. flavus.
A more comprehensive investigation of the role of aflK in the development and pathogenicity of A. flavus will be conducted in future studies.