The Inhibitory Effect of Pseudomonas stutzeri YM6 on Aspergillus flavus Growth and Aflatoxins Production by the Production of Volatile Dimethyl Trisulfide

Aspergillus flavus and the produced aflatoxins cause great hazards to food security and human health across all countries. The control of A. flavus and aflatoxins in grains during storage is of great significance to humans. In the current study, bacteria strain YM6 isolated from sea sediment was demonstrated effective in controlling A. flavus by the production of anti-fungal volatiles. According to morphological characteristics and phylogenetic analysis, strain YM6 was identified as Pseudomonas stutzeri. YM6 can produce abundant volatile compounds which could inhibit mycelial growth and conidial germination of A. flavus. Moreover, it greatly prevented fungal infection and aflatoxin production on maize and peanuts during storage. The inhibition rate was 100%. Scanning electron microscopy further supported that the volatiles could destroy the cell structure of A. flavus and prevent conidia germination on the grain surface. Gas chromatography/mass spectrometry revealed that dimethyl trisulfide (DMTS) with a relative abundance of 13% is the most abundant fraction in the volatiles from strain YM6. The minimal inhibitory concentration of DMTS to A. flavus conidia is 200 µL/L (compound volume/airspace volume). Thus, we concluded that Pseudomonas stutzeri YM6 and the produced DMTS showed great inhibition to A. flavus, which could be considered as effective biocontrol agents in further application.


Introduction
A. flavus, as a harmful phytopathogen, can infect peanuts, maize, and soybeans as well as their products, and produce highly toxic aflatoxins in the field and storage conditions [1]. Aflatoxins are considered unavoidable contaminants of human food and animal feed by the US Food and Drug Administration (FDA), which contaminate over 25% of the world's crops annually [1] and cause great hazards to food security and human health [2]. Aflatoxin B1 (AFB1) is the most toxic member of the mycotoxin group, and its toxicity is 10 times that of potassium cyanide and 68 times that of arsenic [3]. The AFB1 is classified as a group I carcinogen by the International Agency for Research on Cancer (IARC) for its high degree of toxicity, and it may cause liver and lung carcinogen, and even acute death in humans [4].
It is estimated that more than 5 billion people are exposed to the harm of aflatoxins annually [5]. In the 1960s, hundreds of thousands of turkeys died from the aflatoxincontaminated grain diet, which was known as the proverbial "Turkey X disease" [6].

Identification of Strain YM6
The colony of strain YM6 was opaque and pale yellow with a smooth surface and entire edge on the NA medium. YM6 was identified as a gram-negative, short-rod bacterium under the microscope. The 16S rDNA sequence of YM6 was homologous to three species (P. stutzeri, P. putida, and P. xanthomarina) of Pseudomonas spp. Eight strains of three species with great similarity to YM6 were selected to construct the phylogenetic tree. Strain YM6 and P. stutzeri 28a42 (AJ312165.1) showed the highest level of homology in terms of clade-level. Thus, the YM6 was identified as a strain of P. stutzeri (Figure 1). Its sequence was submitted to the NCBI database (accession number KF135442).

Inhibition of Pseudomonas stutzeri YM6 against A. flavus
Strain YM6 can greatly inhibit the growth of A. flavus in face-to-face dual culture without contact. The growth diameter of A. flavus inoculated on PDA medium was 4.5 cm at 5 dpi (days post inoculation). In the treatment group, YM6 inoculated on the NA medium produced abundant volatiles and greatly inhibited the mycelia growth of A. flavus in the PDA medium. The mycelia showed no signs of growth with YM6 treatment (Figure 2A). The inhibitory rate of volatiles from YM6 on mycelia growth of A. flavus was 100% at 5 dpi.
Similar results were also observed in the inhibitory tests of YM6 on conidia germination of A. flavus. In the control group, the A. flavus conidia inoculated on PDA medium germinated quickly, producing germ tubes that grow to form hyphae in 12 h ( Figure 2B). In the subsequent 12 h, the hyphae branched outward, forming an extensive network that resembles the branches of a tree, and covered the surface of the PDA medium ( Figure 2B). In the treatment group, the volatiles from YM6 inhibited the germination of A. flavus conidia with no formation of germ tubes during 24 h ( Figure 2B). Thus, the inhibitory rate of volatiles from YM6 on conidia germination of A. flavus was 100% in dual culture.

Inhibition of Pseudomonas stutzeri YM6 against A. flavus
Strain YM6 can greatly inhibit the growth of A. flavus in face-to-face dual culture without contact. The growth diameter of A. flavus inoculated on PDA medium was 4.5 cm at 5 dpi (days post inoculation). In the treatment group, YM6 inoculated on the NA medium produced abundant volatiles and greatly inhibited the mycelia growth of A. flavus in the PDA medium. The mycelia showed no signs of growth with YM6 treatment (Figure 2A). The inhibitory rate of volatiles from YM6 on mycelia growth of A. flavus was 100% at 5 dpi.

Biocontrol Activity of YM6 against A. flavus and Aflatoxins in Peanut and Maize during Storage
In the control groups, the peanut and maize kernels were severely infected by A. flavus. There were a large number of green mycelia of A. flavus on the surface of peanut and maize seeds. Additionally, the number of infected kernels at a higher level of water activity (aw) outnumbered that at a lower level of aw. At aw of 0.9, the disease incidence was 100% in the control groups of maize and peanuts. At aw of 0.8 and 0.7, the disease incidence of the peanuts group was 82%, and of the maize control group was 32% ( Figure  3A,B). In comparison, the infection of A. flavus was greatly inhibited in the presence of Similar results were also observed in the inhibitory tests of YM6 on conidia germination of A. flavus. In the control group, the A. flavus conidia inoculated on PDA medium germinated quickly, producing germ tubes that grow to form hyphae in 12 h ( Figure 2B). In the subsequent 12 h, the hyphae branched outward, forming an extensive network that resembles the branches of a tree, and covered the surface of the PDA medium ( Figure 2B). In the treatment group, the volatiles from YM6 inhibited the germination of A. flavus conidia with no formation of germ tubes during 24 h ( Figure 2B). Thus, the inhibitory rate of volatiles from YM6 on conidia germination of A. flavus was 100% in dual culture.

Biocontrol Activity of YM6 against A. flavus and Aflatoxins in Peanut and Maize during Storage
In the control groups, the peanut and maize kernels were severely infected by A. flavus. There were a large number of green mycelia of A. flavus on the surface of peanut and maize seeds. Additionally, the number of infected kernels at a higher level of water activity (a w ) outnumbered that at a lower level of a w . At a w of 0.9, the disease incidence was 100% in the control groups of maize and peanuts. At a w of 0.8 and 0.7, the disease incidence of the peanuts group was 82%, and of the maize control group was 32% ( Figure 3A,B). In comparison, the infection of A. flavus was greatly inhibited in the presence of YM6. At higher a w (0.9), peanut kernels were slightly infected by A. flavus with a disease incidence of 32%, while no infection was observed in the maize group. At lower a w (0.7 and 0.8), no disease symptom was observed in maize and peanut kernels ( Figure 3). The results clearly showed that the disease incidence rate in the control group was significantly higher than that in the YM6 treatment group. Based on these observations, we conclude that the volatiles from YM6 can greatly inhibit the infection of A. flavus conidia, and significantly prevent the disease development in peanut and maize kernels.
The aflatoxins in peanut and maize samples were also determined through quantitative analysis. In the control group, the total amounts of aflatoxins in peanut samples were 99.49, 330.17, and 1767.61 ppb at a w of 0.740, 0.859, and 0.923, respectively. Similar phenomenon was also observed in maize samples that the content of aflatoxins was 27.09, 178.39, and 466.13 ppb at a w of 0.785, 0.866, and 0.934, respectively (Table 1). On the one hand, these results showed that A. flavus produced more aflatoxins (with maximum AFB1 of over 80%) in crop seeds at higher a w in the control group. On the other hand, no aflatoxin was detected in maize samples under three a w, and peanut samples under a w 0.740, 0.859 with the treatment of YM6. A little amount of aflatoxin (3.74 ppb) was detected in peanuts at a w of 0.923 (Table 1). Therefore, the volatiles from YM6 are able to inhibit the production of aflatoxins by A. flavus in peanut and maize samples at higher a w under storage conditions.  The aflatoxins in peanut and maize samples were also determined through quantitative analysis. In the control group, the total amounts of aflatoxins in peanut samples were 99.49, 330.17, and 1767.61 ppb at aw of 0.740, 0.859, and 0.923, respectively. Similar phenomenon was also observed in maize samples that the content of aflatoxins was 27.09, 178.39, and 466.13 ppb at aw of 0.785, 0.866, and 0.934, respectively (Table 1). On the one hand, these results showed that A. flavus produced more aflatoxins (with maximum AFB1 of over 80%) in crop seeds at higher aw in the control group. On the other hand, no aflatoxin was detected in maize samples under three aw, and peanut samples under aw 0.740, 0.859 with the treatment of YM6. A little amount of aflatoxin (3.74 ppb) was detected in peanuts at aw of 0.923 (Table 1). Therefore, the volatiles from YM6 are able to

Analysis of A. flavus Cell Structure Affected by YM6
The experiment inoculated peanuts with conidia of A. flavus at a w of 0.923 and cultured them at 28 • C for 5 days. The phenotype of conidia was analyzed through a scanning electron microscope and the results indicated conidial germination and transition from conidia to hyphae in the control group. The peanut seed coat was found to be covered by abundant mycelia. The hyphae also formed conidiophores and produced large amounts of conidia which induced secondary infection. The conidia were uniform in shape with fertile spherical proliferations on the surface. In contrast, only a few severely dehydrated conidia were found on the surface of peanuts in the YM6 group. The conidia were too deformed to germinate into hyphae ( Figure 4). The observation suggested that under non-contact

Chemical Identification of Volatiles Produced by the YM6 Strain
GC-MS analysis showed that the YM6 strain could produce abun stances ( Figure 5). These volatiles were dimethyl trisulfide, 1-(trimeth oxalic acid, 1-methyl-2-pentyl-cyclohexane, isobutyl pentadactyl est 3,5-dimethyl-Isoxazole (Table 2). These substances with a molecular we 97 to 356 Dalton (D) can easily volatilize. Only one type of substance based on comparison with the library NIST 08, with a similarity hig large relative abundance (over 1%, peak area/sum area of all peaks), ered as a candidate volatile from YM6 and the key to the anti-fungal ac was further identified to be DMTS based on the comparison of retent ment ions with the commercial standard.

Chemical Identification of Volatiles Produced by the YM6 Strain
GC-MS analysis showed that the YM6 strain could produce abundant volatile substances ( Figure 5). These volatiles were dimethyl trisulfide, 1-(trimethylsilyl)-1-propyne oxalic acid, 1-methyl-2-pentyl-cyclohexane, isobutyl pentadactyl ester, undecane, and 3,5dimethyl-Isoxazole (Table 2). These substances with a molecular weight ranging from 97 to 356 Dalton (D) can easily volatilize. Only one type of substance was characterized based on comparison with the library NIST 08, with a similarity higher than 90% and large relative abundance (over 1%, peak area/sum area of all peaks), which was considered as a candidate volatile from YM6 and the key to the anti-fungal activity. The volatile was further identified to be DMTS based on the comparison of retention time and fragment ions with the commercial standard.
Toxins 2022, 14, 788 7 of 13 97 to 356 Dalton (D) can easily volatilize. Only one type of substance was characterized based on comparison with the library NIST 08, with a similarity higher than 90% and large relative abundance (over 1%, peak area/sum area of all peaks), which was considered as a candidate volatile from YM6 and the key to the anti-fungal activity. The volatile was further identified to be DMTS based on the comparison of retention time and fragment ions with the commercial standard.    a The peak area of the identified compound to the total area of all peaks. b the mass spectrum of identified compound compared to the spectrum of standard substance in NIST 08 database, respectively. Mass means mass to charge ratio.

Minimal Inhibitory Concentration of DMTS against A. flavus
DMTS purchased from Sigma was used to test the inhibitory effect against A. flavus. The result revealed that DMTS could significantly inhibit the growth of A. flavus in confined spaces. The mycelia of A. flavus spread quickly to the edge of the Petri dishes and produced abundant green conidia at 5 dpi in the control group. Additionally, DMTS showed increased antagonistic effects against A. flavus with elevated concentrations. At a lower concentration of 50 µL/L (compound volume/airspace volume), the inhibition rate was only 7.5%. At 100 µL/L, the inhibition rate was 20.93%. DMTS completely inhibited the growth of A. flavus at 200 µL/L. Thus, the minimal inhibitory concentration (MIC) for DMTS against A. flavus was 200 µL/L ( Figure 6).

Discussion
A. flavus is a globally distributed saprophytic fungus that infects many important crops, such as maize, peanuts, and soybeans during storage [20]. Additionally, it pro-

Discussion
A. flavus is a globally distributed saprophytic fungus that infects many important crops, such as maize, peanuts, and soybeans during storage [20]. Additionally, it produces highly toxic and carcinogenic aflatoxins which threaten food security and human health. Hence, identifying safe and efficient agents to control A. flavus and aflatoxins in crops during storage is of significance and worth further study.
Microbes as important organisms are widely distributed in the natural environment. They can produce abundant secondary metabolites which have been extensively used in control plant pathogens such as Fusarium graminearum [21], Penicillium digitatum [22], Rhizoctonia solani [23], Penicillium digitatum [24], and A. flavus [25]. It is difficult to prevent A. flavus infection and aflatoxins production in grains during storage as A. flavus possesses great spore production, discharge, and dispersal capacities. Many chemical fungicides, although effective in controlling A. flavus and aflatoxins in storage, exhibit detrimental effects on food security and human health due to food contamination caused by pesticide residue. Compared to stable pesticides, compounds with smaller molecular sizes evaporate faster and can be evenly distributed in the airspace of storage, which is crucial to the control of A. flavus and aflatoxins. In 2015, our lab first demonstrated that volatile organic compounds (VOCs) produced by Shewanella algae strain YM8 [26] could greatly inhibit A. flavus growth and aflatoxin production in grains during storage. Some other microbes also have been proved with the capacity to prevent A. flavus infection and aflatoxin production in grains such as Streptomyces alboflavus [27], Streptomyces yanglinensis [28], Bacillus megaterium [29], Pseudomonas protegens [29], Staphylococcus saprophyticus [18], and Serratia marcescens [19]. However, there was no report of volatiles from Pseudomonas stutzeri regarding their effects against A. flavus and aflatoxins.
In the current study, YM6, which is a strain of Pseudomonas stutzeri, was isolated from marine sediment. It can produce volatiles that can greatly inhibit A. flavus growth and aflatoxins production in storage. The volatiles from strain YM6 exhibited a significant antifungal effect as they could inhibit the mycelial growth and spore germination of A. flavus by 100%, and severely damage the cell structure of A. flavus. The key antimicrobial substance produced by YM6 was identified as DMTS, with a minimal inhibitory concentration against A. flavus being 200 µL/L.
We also found that DMTS could deform the conidia of A. flavus and prevent the cell germination, but the inhibitory mechanism of DMTS for A. flavus remains unknown. However, on the basis of the study by Tang et al. [30], DMTS could cause deterioration of subcellular structures of Colletotrichum gloeosporioides, such as cell walls, plasma membranes, Golgi bodies, and mitochondria, as well as contribute to the leakage of protoplasm and cell death. Moreover, Zuo et al. [31] found that, in Fusarium oxysporum cells, DMTS could affect the glycosylation and ROS accumulation, inhibit steroid biosynthesis and glycerophospholipids metabolism, disrupt the cell membrane integrity, and finally result in the cell death. Taken together, DMTS could enter into fungal cells, disrupt cell structures, and inhibit cell growth. These results provided important evidence for understanding the inhibitory mechanism of DMTS for A. flavus, whereas the molecular target as well as interacted proteins of DMTS have not been investigated in fungal cells, which is worth further investigation.
Many volatiles produced by microbes have been demonstrated effective in controlling A. flavus and aflatoxins during storage, such as DMTS, dimethyl disulfide, methyl isovalerate, 1-Pentanol, Phenylethyl Alcohol and 3, 3-Dimethyl-1,2-epoxybutane, et al. [5,18,20,26]. Among these, sulfide exhibits have been demonstrated as excellent biocontrol effects in controlling soil-borne diseases by effectively inhibiting several pathogens and nematodes in soil [32,33]. DMTS, as an active substance, is a colorless or pale yellow substance with a strong mint odor, which exists in fresh onions, Chinese chive [34], and produced by some microbes [26]. However, this paper is the first to report that DMTS was produced by Pseudomonas stutzeri. DMTS is commonly used as permissible food additives in seasoning, gravy, and soup in many countries. DMTS can be applied in the storage of grain and oil products for its fast evaporation and dispersal capacities, and even distribution in confinement spaces to control plant pathogens and mycotoxins.
In conclusion, we were the first to prove that strain YM6 of Pseudomonas stutzeri showed great inhibition on A. flavus and aflatoxins in grains during storage by producing volatile DMTS. As a result, we deduced that the bacterial strain of YM6 and one of its VOCs (DMTS) could be used predominantly as a bio-control agent for crop protection in postharvest stage. In addition, the microbe-associated volatiles showed great antifungal activity to A. flavus in laboratory investigation, but it is still a long distance from lab to commercial application in grains during storage. Some important questions should be answered before commercial application. For example, different to the ventilated requirement in large-scale granaries and small barns nowadays, the application of volatiles developed here needed to be kept in airtight storage condition. Apart from that, some other queries should also be solved before commercial application such as the mass production and enrichment of volatiles, usage cost limitation, usage dosage and persistence, volatile residues, as well as biosecurity to livestock and humans. These limits presented here provide us with further research directions.

Microbes and Plants
Bacteria YM6 was isolated from sea sediment in the Yellow Sea of China. YM6 was cultured in nutrient agar (NA) medium at 37 • C in the dark for 24 h. Aspergillus flavus strain 535 isolated from diseased peanuts was stored in our lab and inoculated onto potato dextrose agar (PDA) medium and cultured at 28 • C in the dark for anti-fungal tests [26].
Peanuts (cultivar Silihong) were purchased from supermarkets and used for inoculation tests. Dimethyl trisulfide (DMTS, CAS: 03658-80-8) was purchased from Sigma (Sigma-Aldrich, St. Louis, MO, USA) and used as the chemical standard for volatile identification.
The experiments were carried out in Molecular Biotechnology Laboratory of Triticeae Crops, Huazhong Agricultural University, Wuhan, China.

DNA Extraction and Phylogenic Analysis
The strain YM6 was inoculated onto NB medium and cultured in a flask at 28 • C at 200 rpm in the dark for 24 h. The suspension was centrifuged at 12,000 rpm for 10 min. The collected cells were used for genomic DNA extraction [26]. Then, 16S rDNA was amplified by PCR and amplified fragments were sequenced by Shanghai Sangon Biological Technology Company with the following sequencing primers: 27F (AGAGTTTGATCCTG-GCTCAG) and 1541R (AAGGAGGTGATCCAG CCGC). The PCR was performed under the following conditions: initial denaturation at 94 • C for 5 min; followed by 30 cycles of 94 • C for 30 s, 55 • C for 30 s, and 72 • C for 40 s; and then 72 • C for 10 min. The 16S rDNA sequences of YM6 were submitted and aligned in the GenBank database. The 16S rDNA of bacteria with similarity over 95% to the sequences of YM6 were used for the construction of a phylogenic tree. The phylogenic tree of homologous strains was constructed by MEGA software with the neighbor-joining method.

Inhibitory Effect of YM6 on Mycelial Growth of A. flavus
A single YM6 colony was streaked over the NA medium surface and cultured at 37 • C in the dark for 24 h. The sterile water was added in NA medium to wash off the bacteria cells on the NA surface with an adjusted concentration of 10 8 CFU/mL for a further anti-fungal test.
A. flavus was cultured on PDA medium for 4 days and produced abundant conidia. The conidia were then inoculated in PDB medium and cultured at 28 • C at 200 rpm for 3 days to produce mycelial pellets. Fresh A. flavus conidia and mycelial pellets were used in the face-to-face dual cultural test to analyze the antagonistic activity of YM6.
The regular size of A. flavus pellets were inoculated in the center of PDA plates, respectively. Fresh YM6 cells (100 µL, 10 8 CFU·mL −1 ) were spread on the surface of the NA plate. Then, the PDA plate inoculated with A. flavus pellet was placed above the NA plate containing YM6. The two plates were placed face-to-face and sealed with two-layer tapes. The PDA plate inoculated with A. flavus pellet and challenged with NA medium was used as a control. Each treatment was repeated three times. All plates were cultured at 28 • C in the dark for 5 days. The mycelium diameters of A. flavus and the inhibition rate of YM6 in these treatments were calculated as follows: Inhibition rate (%) = [(mycelium diameter in control group − mycelium diameter in YM6 group)/mycelium diameter in control group] × 100.

Inhibitory Effect of YM6 on the Conidia Germination of A. flavus
FTF dual cultural test was used to analyze the inhibitory effect of YM6 on conidia germination of A. flavus. A total of 100 µL of A. flavus spores (5 × 10 5 CFU/mL) was evenly coated onto cellophane covering the surface of the PDA plate, and the YM6 (100 µL, 10 8 CFU/mL) was coated onto the surface of the NA plate. The two plates were placed FTF, and the PDA plate was inoculated with A. flavus conidia on top. The PDA plates inoculated with A. flavus conidia challenged with NA medium were used as controls. Each pair of the plates was sealed with two-layer tapes and cultured at 28 • C in the dark for 24 h. The character characteristics of conidia on cellophane were detected on a microscope (Hitech Instruments Co. Ltd., Shanghai, China), and the germination rate of conidia in the control and YM6 group was determined. The inhibition rate was calculated as follows: Inhibition rate (%) = [(conidia germination rate in control − conidia germination rate in YM6 treatment)/control conidia germination rate] × 100.

Biocontrol Activity of YM6 against A. flavus and Aflatoxins in Peanut and Maize
Maize and peanut kernels (100 g each) were grouped into 3 sets respectively and transferred into six flasks (250 mL) to be sterilized at an environment of 121 • C and 1.01 MPa for 20 min. The kernels were then cooled to room temperature. Fresh A. flavus conidia (1 mL, 5 × 10 5 CFU·mL −1 ) was added into each flask and mixed for 10 min. Three flasks containing peanut kernels were filled with sterilized water; the water activity was determined at 0.785, 0.866, and 0.934 through an electronic dewpoint water activity meter, Aqualab Series 3 model TE (Decagon Devices, Pullman, Washington, DC, USA), respectively [26]. The values of water activity in flasks containing maize were measured at 0.740, 0.859, and 0.923, respectively. The maize and peanut kernels were used to test the biocontrol activity of YM6. The kernels in each flask were equally grouped into two sets, with one set placed in one compartment of the Petri dish (two in total). The other compartment containing NA medium was spread with YM6 strain (50 µL, 10 9 CFU/mL). The kernels in one compartment challenged with NA medium (in the other compartment) were used as controls. All Petri dishes were sealed and cultured at 28 • C in darkness. The disease incidence of peanut and maize was calculated at 7 dpi.
Aflatoxins in maize and peanuts were extracted with acetonitrile/water and analyzed through ultra-performance liquid chromatography and mass spectrometry (UPLC-MS, Thermo Scientific, New York, NY, USA) [26].

Structural analysis of A. flavus Treated by Volatile from YM6
A. flavus spores were inoculated to the peanut surface with a water activity of 0.934 and challenged with YM6 through FTF dual culture for 5 days. The A. flavus inoculated on peanuts surface without the presence of YM6 was used as a control. The peanut seeds were fumigated with osmic acid (0.1%, v/v) for 1 h, and then placed at room temperature for 3 h [35]. A small piece of the peanut coat (0.5 cm × 0.5 cm) was removed with a dissecting knife, clamped with stubs, and coated with a layer of gold for examination of SEM (JSM-6390, Hitachi Corporation, Tokyo, Japan) to observe the ultra-structure of A. flavus.

Identification of Volatiles from the Strain YM6
The strain YM6 was inoculated onto the surface of the NA medium in a 100 mL flask. The flask containing NA medium without YM6 inoculation was used as a control. All flasks were sealed with double-layer plastic film and cultured at 28 • C in the dark for 48 h. All the experiments were repeated twice. Then, the flasks were incubated in a water bath of 40 • C for 30 min. The volatile substance was extracted by solid phase micro-extraction (SPME, divinylbenzene/carboxy/polydimethylsiloxane) and analyzed by gas chromatographymass spectrometry (GC-MS) (Agilent Technologies, Santa Clara, CA, USA). For volatile enrichment, the metal head of SPME was inserted through the plastic film into the flask. The coated fiber in the metal head was pushed out and placed in the center position above the NA medium to absorb the volatiles for 50 min. The coated fiber was retracted and transferred into the GC-MS system for volatile analysis. The detection parameter was used as below [26].
The splitless injection of GC was used in the GC-MS analysis. The carrier gas was helium; the inlet temperature was 250 • C. The oven was set for procedures as follows: started at 40 • C for 3 min; heated up to 160 • C at the rate of 3 • C/min for 2 min; and then raised to 220 • C at the rate of 8 • C/min for 3 min. The MS analysis was performed with an EI source under a temperature of 230 • C. The quadrupole temperature was set at 150 • C. The collision energy was 70 eV. The mass range was set from 50 to 500 amu. The detected volatiles were verified by aligning in the National Institution of Standards and Technology (NIST 08) database, and the retention time was compared with authentic compound purchased from the company.

Inhibition Effect of DMTS on A. flavus
The volatile substance produced by YM6 showed a significant antagonistic effect on A. flavus growth. DMTS (liquid state, purity ≥ 98%), the major component emitted by YM6, was purchased from Sigma and tested as the anti-fungal agent against A. flavus. The anti-fungal analysis was conducted through the FTF dual cultural test. A piece of circular filter paper (5 cm diameter), moistened by DTMS, was placed in a Petri dish. The final concentration of DMTS in each treatment was adjusted to 200 µL/L, 100 µL/L, and 50 µL/L, respectively. The other Petri dish containing PDA medium was inoculated with A. flavus mycelia. The dish containing A. flavus was placed FTF above the dish containing DMTS and sealed with two layers of tape. The dish of A. flavus co-cultured with the same volume of water was used as a control. All plates were cultured at 28 • C in darkness for 5 days. The mycelium diameter of A. flavus was measured and the inhibition rate was calculated.

Data Analysis
All experiments were conducted in duplicate. The amount of aflatoxin was evaluated and shown as mean ± SE. The significant differences between mean values were determined using Duncan's multiple range test (p < 0.05) following one-way analysis of variance (ANOVA). The statistical analysis was performed using SPSS 16.0 software (SPSS Inc., Chicago, IL, USA).