Differential Expression of Genes Related to Growth and Aflatoxin Synthesis in Aspergillus flavus When Inhibited by Bacillus velezensis Strain B2

Aspergillus flavus is a saprophytic soil fungus that infects and contaminates seed crops with the highly carcinogenic aflatoxin, which brings health hazards to animals and humans. In this study, bacterial strains B1 and B2 isolated from the rhizosphere soil of camellia sinensis had significant antagonistic activities against A. flavus. Based on the phylogenetic analysis of 16SrDNA gene sequence, bacterial strains B1 and B2 were identified as Bacillus tequilensis and Bacillus velezensis, respectively. In addition, the transcriptome analysis showed that some genes related to A. flavus growth and aflatoxin synthesis were differential expressed and 16 genes in the aflatoxin synthesis gene cluster showed down-regulation trends when inhibited by Bacillus velezensis strain B2. We guessed that the Bacillus velezensis strain B2 may secrete some secondary metabolites, which regulate the related gene transcription of A. flavus to inhibit growth and aflatoxin production. In summary, this work provided the foundation for the more effective biocontrol of A. flavus infection and aflatoxin contamination by the determination of differential expression of genes related to growth and aflatoxin synthesis in A. flavus when inhibited by B. velezensis strain B2.


Introduction
Aspergillus flavus is a saprophytic soil fungus that infects and contaminates seed crops and it is in the spotlight for the production of aflatoxins that contaminate oil-rich seeds such as maize, peanuts, cotton seeds and treenuts before and after harvest. A. flavus, as the second Aspergillus fumigatus only pathogen, produces secondary metabolites containing polyketones, which are strong carcinogens and can seriously affect human and animal health [1]. Aflatoxins are one of the most harmful mycotoxins, which are produced by A. flavus, A. parasiticus, and other fungi that are commonly found in the production and preservation of grain and feed. Aflatoxins can cause harm to animal and human health due to their toxic (carcinogenic, teratogenic, and mutagenic) effects [2]. Aflatoxins were first discovered and characterized in the early 1960s following a severe livestock poisoning incident in England involving turkeys; they are kinds of polyketide-derived carcinogenic and mutagenic secondary metabolites [3][4][5]. Aflatoxins belong to a class of secondary metabolites of A. flavus and include varieties and the four major types are B1, B2, G1 and G2. Aflatoxin B1 (AFB1) exerts strong mutagenicity and carcinogenicity [6], which is classified as a Group 1 human carcinogen by the International Agency for Research on Cancer [7,8]. It is converted to AFB1-8 and 9-epoxide in the liver, catalyzed by cytochrome P450 associated enzymes after being ingested by animals, forming adducts with the guanine base of DNA, thus resulting in acute and chronic diseases called aflatoxicoses in both human and household animals feeding high levels of aflatoxins-contaminated food [9]. AFB1 biosynthesis was catalyzed by the coordinated cascade of enzymes that are encoded by approximately 30 different genes that group into a cluster located near the telomeric region of chromosome three of the aflatoxinogenic species [10]. The initial stage of aflatoxin biosynthesis is similar to fatty acid biosynthesis, with acetyl-CoA as the starting unit and malonyl-CoA as the elongating unit, which is catalyzed by polyketide synthase (PKSA) to form the aflatoxin of the polyketone backbone [11].
Aflatoxin not only harms the health of human and animals, but also causes economic losses. Human and animals themselves cannot degrade aflatoxin through their own metabolism and so avoid contact with or ingestion of aflatoxin [12]. The pollution of aflatoxin to grain occurs in all stages before harvest, during harvest, storage and processing. Therefore, at each stage, from planting to harvest, there are different methods for the prevention and control of aflatoxin. The prevention and control of aflatoxin can be divided into prenatal and postpartum prevention and control [13]. At present, the main method to control the pollution of A. flavus before harvest is to use the non-toxigenic-strain of A. flavus to compete with the toxigenic-strain of A. flavus for the nutrient substrate to achieve inhibition [14,15]. COTTYPJ et al. [16] applied the non-toxigenic-strain of A. flavus to cotton fields to effectively reduce aflatoxin pollution in cotton seeds. The effect of the field biological control of non-toxin-producing A. flavus strains can last until harvest and storage [17]. At present, large-scale research has led to the rapid development of non-toxigenic-strains of A. flavus in commercial application [18,19]. The control measures of A. flavus and its toxin in post-harvest crops mainly include physical control, chemical control and biological control [20]. Physical methods mainly include ozone, irradiation, high temperature heating and so on, while chemical methods include ammoniation, alkali, oxidation and so on. However, these two methods have some shortcomings, such as an unstable effect, large loss of nutrients and difficulty in large-scale production [21]. The biological control of A. flavus is a method to control A. flavus and its toxin using microorganisms and their secondary metabolites or enzymes is secreted. A large number of studies have shown that many microorganisms can inhibit the growth and toxin production of A. flavus, including bacteria, actinomycetes, yeast, algae and so on [22]. During their growth, they usually produce a series of secondary metabolites, including alcohols, aldehydes, acids, esters and ketones [23,24]. At present, scholars at home and abroad mainly study the volatile substances produced by Bacillus, Pseudomonas and Streptomyces. The volatile compounds produced by plant rhizosphere bacteria can not only inhibit the growth and toxin production of A. flavus, but also have a certain effect on plant growth [25]. For example, volatile substances produced by Bacillus amylolyticus can effectively stimulate the growth and development of plants [26]. Cutierrez found that volatile organic compounds produced by Bacillus could promote root elongation and change the root structure of Arabidopsis thaliana [27]. In addition, the use of microorganisms to detoxify aflatoxins is a promising new technology with broad application prospects [2]. For example, the excellent adsorption capacity and natural fermentation function make the use of Lactobacillaceae and Saccharomyces essential in the process of detoxifying food. Lactobacillus rhamnosus strain is an excellent biosorption species. The combination of heat-treatment and anaerobic solid fermentation can remove 100% of AFB1 [28]. Research proved the capacity of L.acidophilus to bind AB1 and AM1 in cow's milk [29], which can reduce AFM1 (aflatoxin M1) and potentially decrease toxins in yogurt to a safe concentration for consumption (below 0.05 µg/kg) [30].
Bacillus velezensis is a plant growth-promoting bacterium that can inhibit plant pathogens [31], but can also produce a variety of antibacterial proteins and lipopeptide antibacterial substances with broad-spectrum antibacterial activities. Li et al. demonstrated that lipopolysaccharide produced by B. velezensis can not only inhibit spore of A. flavus germination and cause abnormal mycelium expansion and cell rupture, but can also significantly down-regulate the genes (aflK, aflR, Vea, and omtA) of aflatoxin synthesis pathways [32]. KONGQ et al. carried out in vivo and in vitro experiments on a marine B. velezensis strain isolated from the Yellow Sea in eastern China to test its effect on post-harvest corruption of peanuts caused by A. flavus. The results showed that the B. velezensis had significant biological control effects in vivo, and could reduce the biosynthesis of aflatoxin by inhibiting the transcription of the aflR gene and aflS gene [33]. B. velezensis DY3108, isolated from the soil, could detoxify fungal contamination in food by degrading 90% of aflatoxins B1 [34]. Vahidinasab et al. [35] reported that the B. velezensis UTB96 strain showed the highest antifungal activity against A. flavus and was capable of degrading aflatoxin. B. velezensis as a probiotic can compete with aflatoxigenic strains, either for nutrients and space, or for the degradation of aflatoxins by enzyme production [36].
Thus far, the molecular mechanism of Bacillus against A. flavus is not clear. The aim of this study is to find potential biocontrol agents against A. flavus growth and aflatoxin synthesis and understand the antagonistic mechanism. We found two bacterial strains, B1 and B2, isolated from the rhizosphere soil of camellia sinensis had significant antagonistic activities against A. flavus and identified them as the Bacillus tequilensis strain B1 and Bacillus velezensis strain B2. Transcriptome analysis showed that some genes related to A. flavus growth and aflatoxin synthesis were differentially expressed and 16 genes in the aflatoxin synthesis gene cluster showed down-regulation trends when inhibited by Bacillus velezensis strain B2. This provided the foundation for the more effective biological control of A. flavus and aflatoxin synthesis.

Fungal Strain
The fungal strain Aspergillus flavus was obtained from Prof. Xiufang Hu at the microbiology lab of Zhejiang Sci-Tech University. A. flavus was cultured on a PDA (Potato Dextrose Agar) medium and incubated at 28 • C [32,37,38] for 10 days.

Isolation of Bacterial Isolates from Camellia sinensis
The rhizosphere soil of Camellia sinensis was collected from tea garden in Meijiawu, Hangzhou, Zhejiang province. Samples were collected aseptically into Ziplock bags and marked labels, respectively, which were then transported to the laboratory for analysis. We weighed 10 g of rhizosphere soil and added a conical flask containing 90 mL sterile water to make a 10 −1 soil suspension, and then we made 10 −2~1 0 −9 soil suspensions by serial dilution. 100 µL of 10 −7 , 10 −8 , and 10 −9 soil suspensions were, respectively, spread on the LB (Luria Bertani) medium and then incubated at 37 • C for 4 days. Finally, 10 bacterial strains were obtained by isolation and purification.

Screen for Potential Biocontrol Agents against A. flavus
Using a sterile cork-bore, mycelial disc (7 mm in diameter) of A. flavus from a 10-days-old culture on a PDA plate was placed in the center of a freshly prepared PDA plate (9 cm in diameter). Additionally, bacterial isolates were inoculated in an LB liquid culture and incubated at 37 • C for 16 h with shaking at 200 rpm, and 20 µL of the bacterial suspensions were, respectively, added to the three aseptic filter paper pieces at a distance of 3 cm from the mycelial discs. Control tests were carried out by only inoculating A. flavus. All the plates were placed in a 28 • C incubator, and the diameters of A. flavus were measured everyday from 2 d to 7 d. The relative inhibition rates were calculated as (C-T)/C × 100% [39], where C denotes the radius of A. flavus in the control and T denotes the radius of A. flavus in the treatment. All the experiments were done in triplicate.

Identification of Bacterial Isolates That Show Antagonism against A. flavus
The antagonistic bacterial isolates were identified both morphologically and genetically. For morphology, the isolates were cultured on plates at 37 • C for 4 days for colony morphology observation, and the pure colonies were stained with gram and observed using light microscopes. At the same time, the pure cultures were used to extract genomic DNA using Wizard Genomic DNA Purification Kit (Solario) according to the manufacturer's instructions. 16SrDNA sequences of bacterial isolates were amplified by using the forward primer (TCCGTAGGTGAACCTGCGG) and reverse primer (TCCTCCGCTTATTGATATGC) and sequenced by the Sangon Biotech Company (Shanghai, China). Phylogenetic trees were constructed using the neighbor-joining method (MEGA 7) based on 16SrDNA gene sequences. Bootstrap resampling analysis was performed to estimate the confidence of the tree topologies.

Conidiation Assays of A. flavus Antagonized by B1 and B2
Conidial yields were determined, as previously described [40]. Briefly, 10 µL of the conidial suspension (1 × 10 7 conidia mL −1 ) of A. flavus were inoculated on the center of PDA plates (90 mm in diameter). Additionally, three asptic filter paper pieces equidistantly spaced around 2.5 cm with A. flavus were then inoculated with 10 µL bacterial suspension of B1 and B2, which were cultured in LB at 37 • C for 16 h (200 rpm), while treatments inoculated with 10 µL of sterile LB were used as negative controls. All the PDA plates were placed in a 28 • C incubator. After 7 days culture, three agar discs (5 mm) of A. flavus were cut using a sterile cork-bore and respectively put into 1 mL 0.01%TitonX-100 (v/v). Conidial yields of A. flavus were determined after ultrasonication (30 Hz for 10 min) of the agar discs (JXFSTPRP-24 Shanghai Jingxin, Shanghai, China).

Transcriptomic Analysis of A. flavus Inhibited by B2 Using RNA-Seq
RNA-seq was used to profile transcriptomes of A. flavus inhibited by the B. velezensis strain B2. Briefly, B2 was inoculated in LB liquid culture and incubated at 37 • C for 16 h with shaking 200 rpm, and 20 µL of the bacterial suspensions were inoculated in the center of a freshly prepared PDA plate (9 cm in diameter), mycelial discs (7 mm in diameter) of A. flavus from a 10-days-old culture on a PDA plate and were respectively placed at a distance of 3 cm from the B2. Treatment of A. flavus without B2 inhibition was used as a negative control. After 4 days incubation at 28 • C, the fungal biomass of A. flavus were, respectively, subjected to RNA extraction with TRIzol reagent (Life Technologies, Carlsbad, CA, USA).
Construction of cDNA libraries and sequencing with the Illumina HiSeq 2000 platform were performed by Frasergen (Shanghai, China). After paired-end sequencing, clean reads were used for gene expression quantification using the RPKM method. Differential expression analysis was performed with edgeR v 3.24 software using the cut-offs of the adjusted p-value 0.05 and a greater than or equal to twofold change [41].

Screen for Potential Biocontrol Agents against A. flavus
To screen for potential biocontrol agents, we isolated 10 bacterial strains from the oilseed plant of Camellia sinensis, which is very susceptible to be infected and contaminated by A. flavus and its production of aflatoxins ( Figure S1). All strains (B1~B10) were tested to determine if they could inhibit the growth of A. flavus using the plate confrontation antagonism test. Only strains of B1 and B2 showed significant inhibiting effects on A. flavus ( Figure 1A,B), and their relative inhibition rates were 34.22% and 35.72%, respectively (Table S1).

Identification of Strains of B1 and B2
The resulting colony of B1 was 2.0-3.0 mm in diameter, smooth, circular, flat and white after being grown on LB media at 37 • C for four days, and B2 was 2.0-3.0 mm in diameter, smooth, circular but with an irregular edge, convex and white. B1 and B2 were both Grampositive (Figure 2A). BLASTn searches based on the 16SrDNA gene sequences of B1 and B2 appeared that they matched Bacillus tequilensis and Bacillus velezensis with 99% similarity, respectively. Phylogenetic trees constructed by the neighbor-joining method showed that B1 and B2, respectively, clustered with the Bacillus tequilensis strain 10b (NR 118290.1) and Bacillus velezensis strain CBMB205 (NR 116240.1), indicating that B1 and B2 are likely isolates belonging to Bacillus (Figure 2B,C).

Conidiation of A. flavus Antagonized by B1 and B2
In A. flavus, conidiation is accompanied by aflatoxin synthesis [32,42]. Compared to the control, conidial yields of A. flavus antagonized by B1 and B2 significantly decreased (p < 0.05) (Figure 3). Therefore, it was speculated that B1 and B2 may secrete some compounds that inhibit the conidiation of A. flavus, and then reduce the synthesis of aflatoxin.

Overview of RNA-Seq Analysis and DEGs Identification of A. flavus Antagonized by B2
To find the genes that may be related to the synthesize of aflatoxins and regulated by bacterial isolate B2, the transcriptomes of A. flavus, which was inhibited by B2, were profiled by Illumina Hiseq 2000 RNA-Seq (A. flavus without antagonism as a negative control), three biological replicates and eight datasets were established. In this section, we compared a set of differentially expressed genes (DEGs) between A. flavus antagonized by B2 and A. flavus alone. Additionally, 'up-regulated genes' are genes with higher expression levels when A. flavus antagonized by B2, 'down-regulated genes', are those with lower expression levels under the same conditions. When A. flavus was antagonized by B2, there were a total of 185 DEGs, including 151 up-regulated and 34 down-regulated genes ( Figure 4).

GO Terms Analysis of DEGs
Among the 185 genes that were differentially expressed in A. flavus inhibited by B2, 30 GO (Gene Ontology) terms were enriched, including ten in biological process, ten in cellular component and ten in molecular function ( Figure 5). A GO term for aflatoxin biosynthesis was enriched for two up-regulated genes of aldo-keto-reductase family and cytochrome P450 related genes (AFLA_3422 and AFLA_860) ( Table 1). Aldehyde-ketoreductase family genes synthesize NAD(P)-linked oxidoreductases that catalyze the reduction of NAD(P)H to NAD(P) + . Cytochrome P450 gene synthesizes cytochrome P450 monooxygenase. In the synthesis of aflatoxin, the role of this enzyme is to catalyze the conversion of O-methyl variegated toxin and dihydro-O-methyl variegated toxin into aflatoxins B1, G1, B2, and G2, but the expression of their activities requires NADPH as a cofactor, and the level of intracellular NADPH/NADP + affects the expression of aflatoxin biosynthesis regulator gene aflR [11]. Therefore, it is speculated that the compound produced by B2 can promote the production of NAD(P)-linked oxidoreductase and catalyze the reduction of intracellular NAD(P)H to NAD(P) + , thereby inhibiting the production of aflatoxin.

KEGG Enrichment Analysis of DEGs
The results of the KEGG enrichment of 185 DEGs showed that they were involved in 45 metabolic pathways. Among them, genes in the biosynthetic pathway of amino acids and the metabolic pathway of carbohydrates were mainly up-regulated, and they were consistent with the results of the research reported previously [4]. Ribosomes are the cellular factories responsible for making proteins. Additionally, according to the results of transcriptome sequencing, genes related to ribosome synthesis were mainly up-regulated in A. flavus (Table 1). Therefore, it was speculated that the active components secreted by B2 may lead to protein damage and membrane structure destruction in A. flavus, and A. flavus synthesizes proteins from other pathways to maintain its own basic growth. The pathways enriched for down-regulated genes mainly include the biosynthetic process, energy metabolism, transcription, folding, transport, and catabolism (Table 1); these genes were down-regulated may affect ribosome assembly and the protein synthesis of A. flavus, thus resulting in growth that was inhibited and the biosynthesis of aflatoxins was reduced.

Discussion
Aspergillus flavus is an universal pathogen of crops and it produces aflatoxin in the seeds of a variety of crops before and after harvest, which poisons humans and animals through the food chains. Currently, many strategies have been commonly used to control aflatoxin production. Preharvest control has relied on identifying resistant crop lines, planting regionally adapted cultivars and planting at appropriate seed densities, thus limiting insect damage and so on. However, these measures are not always efficient to avoid aflatoxin formation [10,43]. More and more biocontrols have been developed. As one strategy, aflatoxin non-producing strains of A. flavus excluded competitive aflatoxinproducing Aspergillus species to decrease crop aflatoxin contamination [9]. Further, aflatoxin biocontrol by beneficial microorganisms, including bacteria, actinomycetes and so on, is considered to be one of the most promising practices [44].
Camellia is a kind of oil plant and seed vulnerable to A. flavus infection and aflatoxin contamination. To screen for potential biocontrol agents that inhibit A. flavus growth and reduce aflatoxin biosynthesis, we isolated 10 bacterial strains from the leaves, flowers, seeds and rhizosphere soil of camellia. We found that only bacterial strains of B1 and B2 can inhibit the growth of A. flavus among the 10 bacterial strains. Some studies showed that volatile organic compounds, such as 2, 3-butanediol produced by soil-borne endophytic bacteria, increase plant pathogen resistance and promote plant growth [45,46]. However, B1 and B2 did not inhibit A. flavus by producing volatile compounds ( Figure S2), and they may have other antagonistic mechanisms.
BLASTn searches and phylogentic trees based on the 16SrDNA gene sequences showed that the bacterial strains B1 and B2 clustered well with the Bacillus tequilensis and Bacillus velezensis, respectively. In addition, B1 and B2 had typical morphological characteristics that were similar to those of Bacillus. Therefore, B1 and B2 are both identified as the Bacillus genus. Members of the Bacillus genus produce abundant biologically active molecules, such as lipopeptides (LPs), polyketides, peptides, phosphatides, polyenes and so on, and they play important roles in controlling plant pathogens [47]. In a previous study, an antagonistic strain of Bacillus velezensis, with obvious anti-Aspergillus flavus fungi activity, was isolated from the surface of healthy rice, and the main components of LPs produced by this strain were identified by HPLC-MS analysis as fengycin and iturins, and they had shown that LPs can inhibit spore germination and even cause abnormal hyphal expansion and cell rupture [44]. Therefore, we guessed that B1 and B2 may produce similar antimicrobial compounds to inhibit A. flavus growth.
The study found that the level of intracellular NADPH/NADP + has a certain effect on the expression of aflR, and NADPH is a cofactor for the expression of cytochrome P450 monooxygenase activity. In A. flavus, cytochrome P450 monooxidase catalyzes the conversion of O-methyl versicolor and dihydro-O-methyl versicolor to aflatoxin B1, G1, B2, and G2, respectively [11]. We found that there were 185 DEGs, including 152 up-regulated genes and 34 down-regulated genes in A flavus antagonized by the Bacillus velezensis strain by the transcriptome analysis. GO term and KEGG enrichment analysis were performed on these DEGs, and we found that the GO term of aflatoxin biosynthesis was enriched for two up-regulated genes. They were namely aldosterone reductase family genes and cytochrome P450 related genes, which synthesize NAD(P)-linked oxidoreductase and cytochrome P450 monooxygenase, respectively. NAD(P)-linked oxidoreductase catalyzes the reduction of NAD(P)H to NAD(P) + . Therefore, it is speculated that the compound produced by B2 can promote the production of NAD(P)-linked oxidoreductase and catalyze the reduction of intracellular NAD(P)H to NAD(P) + , thus inhibiting the monooxygenase activity of cytochrome P450 and even the aflatoxin production. The KEGG enrichment analysis found that related genes in amino acid biosynthesis pathway, carbohydrate metabolism pathway and ribosome synthesis pathway were all up-regulated, and genes in the energy metabolism pathway, transcription, folding, transport pathway, biosynthesis and catabolism pathway were down-regulated. It was speculated that the active components secreted by B2 lead to protein damage and membrane structure disruption in A. flavus and A. flavus synthesizes proteins through other pathways to maintain its own basic growth. However, the downregulated genes may affect ribosome assembly and protein synthesis, thus resulting in growth that was inhibited and biosynthesis of aflatoxins that was reduced.
Aflatoxins are PKS-derived mycotoxins synthesized from a large cluster, which is composed of approximately 30 different genes. The expression level of each gene in this pathway directly affects the amount of aflatoxins synthesis. AflA, aflB and aflC are required to synthesize norsolorinic acid, which is the first stable aflatoxin precursor [10]. VeA, as the core protein of the velvet complex, regulates the growth and development of A. flavus, including participating in conidiogenesis and the publication of sclerotia [48]. AflR and aflS are the regulatory genes of aflatoxin biosynthesis, which play a key role in the regulation of aflatoxin biosynthesis [49,50]. AflK Paticipates in the conversion of VAL to VERB, which is a key step in the formation of aflatoxin and blocks the bifuran ring of aflatoxin [51]. AflP (omtA) gene encoding protein is a key enzyme in the late stage of aflatoxin synthesis [52]. Transcriptome analysis of A. flavus, when inhibited with B2, indicated that 16 genes in the gene cluster of aflatoxin synthesis show down-regulation trends (Table S2). This further indicates that the secondary metabolites produced by B2 can down-regulate the transcription of related genes and reduce the production of aflatoxin.
At present, non-toxigenic-strain of A. flavus or non-toxingenic-strain of Parasitic Aspergillus are often applied to the soil to competitively inhibit the growth of aflatoxinproducing strains, so as to effectively prevent aflatoxin pollution before and after harvest [53]. Jia chang Li uses starch, sodium alginate and thermosensitive gel as a matrix, and non-toxigenic-strain of A. flavus suspension as the source of biocontrol bacteria and prepares temperature-responsive immobilized microspheres by immobilized microtechnology, which can continuously release biocontrol bacteria at a suitable temperature in the field and effectively prevent aflatoxin contamination [54]. However, in practical application, the requirement of non-toxigenic-strain of A. flavus is very high. It can not only produce toxin but must also have strong vitality and competitiveness. Such strains are not easy to obtain in the environment [55]. Bacillus is a kind of aerobic or facultative anaerobic bacteria. It is not only an important microbial population in soil and plant surface rhizosphere, but also a common endophytic bacterium in plants. It has the characteristics of a wide distribution and is easy to obtain [56]. Secondary metabolites produced by Bacillus in plant rhizosphere can not only inhibit the growth and toxin production of A. flavus, but can also promote plant growth [25]. Previous studies have reported the usefulness of B. velezensis as a probiotic in the aquaculture industry [57][58][59]. Due to its wide distribution in nature and enrichment in its own metabolites [60], B. velezensis may have the potential to be used as a probiotic in animal feed. Compared to non-toxigenic-strains of A. flavus, Bacillus strains are easier to culture and obtain and have stronger environmental adaptability. We can prepare liquid bacterial agent B2 and add a large amount of it to rhizosphere soil of tea to inhibit A. flavus, which is a saprophytic soil fungus, and has a better inhibitory effect on the actual crop groups in the field. Bacillus can also produce a variety of antibacterial proteins and lipopeptide antibacterial substances, which have broad-spectrum antibacterial activity [28]. The lipopeptide produced by B. velezensis B2 can be isolated and purified by the fermentation process and made into a preparation to inhibit the growth of A. flavus and aflatoxin production during grain storage [56]. To sum up, B. velezensis B2 isolated from camellia has great potential to become a biocontrol bacterium for the prevention of aflatoxin pollution.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/foods11223620/s1, Figure S1: Morphology of colonies of bacterial isolates obtained from the Camellia sinensis; Figure S2: Inhibition effect of volatile compounds produced by B1 and B2 on the growth of A. flavus; Table S1: Relative inhibition rates of A. flavus inhibited by bacterial isolates B1~B10; Table S2: Expression profiling genes involved in aflatoxin biosynthesis in A. flavus.