Control of Aflatoxigenic Molds by Antagonistic Microorganisms: Inhibitory Behaviors, Bioactive Compounds, Related Mechanisms, and Influencing Factors

Aflatoxin contamination has been causing great concern worldwide due to the major economic impact on crop production and their toxicological effects to human and animals. Contamination can occur in the field, during transportation, and also in storage. Post-harvest contamination usually derives from the pre-harvest infection of aflatoxigenic molds, especially aflatoxin-producing Aspergilli such as Aspergillus flavus and A. parasiticus. Many strategies preventing aflatoxigenic molds from entering food and feed chains have been reported, among which biological control is becoming one of the most praised strategies. The objective of this article is to review the biocontrol strategy for inhibiting the growth of and aflatoxin production by aflatoxigenic fungi. This review focuses on comparing inhibitory behaviors of different antagonistic microorganisms including various bacteria, fungi and yeasts. We also reviewed the bioactive compounds produced by microorganisms and the mechanisms leading to inhibition. The key factors influencing antifungal activities of antagonists are also discussed in this review.


Introduction
Aflatoxins are the most common contaminants occurring widely in oilseeds and grains. Aflatoxins B 1 , B 2 , G 1, and G 2 are a group of potent hepatotoxic and carcinogenic secondary metabolites produced mainly by Aspergillus section Flavi spp. like A. flavus and A. parasiticus [1][2][3]. Aflatoxigenic molds can cause a decrease in production, a loss of nutritional value, and a diminution of market value of agricultural products, and also cause serious diseases like allergic reactions in humans and animals. Aflatoxin B 1 (AFB 1 ), the most toxic and commonly occurring one, has been classified as group I

Antagonistic Microbes against Aflatoxigenic Strains
Various microorganisms including bacteria, fungi such as nontoxigenic Aspergillus, Trichoderma and penicillium spp., and yeast strains have been investigated as potential biocontrol agents against aflatoxigenic strains. As shown in Figure 1, the articles reporting bacterial antagonists were dominant (61%) compared with the articles reporting antagonistic fungi (27%) or yeasts (12%). Additionally, a comprehensive list of all microorganisms (approximately 50 different species) that have been well documented for their anti-aflatoxigenic potential is given in Table 1. Main characteristics and inhibitory behaviors of these antagonists are described as well.
Toxins 2020, 12, 24 2 of 22 group Ι human carcinogen by the International Agency for Research on Cancer [4]. Aflatoxins M1 and M2, which are by-products of the above aflatoxins, may be found in dairy products from animals fed with contaminated feed and are closely related to the safety of dairy food. Physical strategies such as field managements, physical separations, and moisture controls, and chemical strategies (e.g., using fungicides and chemical absorbents) have been applied to control aflatoxin and its producing molds. In most cases, the physical and chemical methods were inefficient, due to a nutritional loss of the processed foods, a difficulty in removing residues of the toxic compounds, or a development of resistant biotypes of pathogens. The biological control has been regarded as a more environmentally friendly and safer method [5,6], which was carried out generally at pre-and/or post-harvest. The post-harvest strategies focus mainly on the removal of aflatoxin [7][8][9][10][11][12]. However, once the agro-foods and feeds are contaminated, the contaminants such as aflatoxins can never be completely removed. Therefore, preventing aflatoxin production and fungal infection is the most efficient strategy.
In the past decades, many research studies have historically focused on the biocontrol of aflatoxigenic molds [13,14]. Three main modes of inhibitory actions are involved: antagonists grow rapidly to occupy ecological niche and compete for nutrients and/or living places, which leads to a displacement of pathogens; another involves inhibiting fungal growth, which leads to a reduction of fungal infection and colonization; the third is based on inhibiting aflatoxin biosynthesis. This review explores inhibitory behaviors, bioactive compounds, mechanisms of inhibitory actions, and factors influencing biological activities.

Antagonistic Microbes against Aflatoxigenic Strains
Various microorganisms including bacteria, fungi such as nontoxigenic Aspergillus, Trichoderma and penicillium spp., and yeast strains have been investigated as potential biocontrol agents against aflatoxigenic strains. As shown in Figure 1, the articles reporting bacterial antagonists were dominant (61%) compared with the articles reporting antagonistic fungi (27%) or yeasts (12%). Additionally, a comprehensive list of all microorganisms (approximately 50 different species) that have been well documented for their anti-aflatoxigenic potential is given in Table 1. Main characteristics and inhibitory behaviors of these antagonists are described as well.  We searched for research articles on the topic of "biocontrol of aflatoxigenic fungi" on Web of Science (http://www.webofknowledge.com). Related research articles account for approximately 150, and each slice of the pie represents a percentage of the articles reporting each sort of microorganisms. Bacillus spp. are a multifunctional group of bacteria. As shown in Figure 1, 21% of research articles reported Bacillus spp., which were most widely assessed in controlling aflatoxigenic strains. Aflatoxin accumulation in potato dextrose broth was almost totally inhibited by B. megaterium [16]. B. subtilis was also able to inhibit A. parasiticus growth and aflatoxin production by a percentage up to 92% and 100%, respectively [15]. Thus, B. megaterium and B. subtilis showed the highest biocontrol activity, inhibiting the growth of as well as aflatoxin production by aflatoxigenic strains, while B. amyloliquefaciens was also able to reduce A. parasiticus growth as well as degrade aflatoxins B 1 , B 2 , G 1, and G 2 after several days of co-cultivation [53,54]. González et al. [18] demonstrated that B. mojavensis, B. cereus, and B. mycoides isolated from soil had ability to significantly inhibit A. parasiticus growth. Isolates of B. pumilus were also demonstrated with ability to inhibit aflatoxin production [17]. As reviewed by Schallmey et al. [55], Bacillus spp. were intensively assessed as biological agents, probably because they grew rapidly, produced a wide range of antimicrobial compounds, and generally were recognized as safe species.
It was found that P. fluorescens could reduce AFB 1 production by A. flavus in peanut medium at a rate of 99.4% [20], as well as inhibit conidial germination of A. flavus by up to nearly 20% [56]. A known fact is that Pseudomonas is one of the most prevalent genera isolated from soil (plants rhizosphere or nonrhizosphere). Palumbo et al. [19] demonstrated that the chitinolytic P. chlororaphis strains isolated from maize fields and maize rhizospheres could completely inhibit A. flavus growth. Mannaa et al. [21] found that P. protegens strain AS15 isolated from rice grains also significantly inhibited aflatoxin production by and mycelial growth of A. flavus at rates of 82.9% and 68.3%, respectively. Several other Pseudomonas strains were also demonstrated with an ability to completely inhibit the growth of A. flavus in different media [34].

Lactobacillus spp.
Lactic acid bacteria (LAB) are bacteria producing organic acids-mainly lactic acid-by carbohydrate fermentation. In food production, these bacteria are traditionally used to prevent spoilage and increase shelf life of foods. As shown in Table 1, L. plantarum, L. rhamnosus, L. casei, L. fermentum, L. pentosus, L. paraplantarum, and L. delbrueckii subsp. Lactis have been identified as biocontrol agents against aflatoxigenic fungi. Ahlberg et al. [57] demonstrated that LAB strains showed an ability to physically bind aflatoxins. In another study of Ahlberg et al. [26], 171 LAB strains were tested against A. flavus, and the species with the highest antifungal ability was identified as L. plantarum. The genus Lactobacillus, mainly the species L. plantarum, has been widely found to inhibit aflatoxigenic strains in various living environments [22,[58][59][60][61]. LAB strains are mainly divided into four genera: Lactobacillus, Lactococcus, Pediococcus, and Leuconostoc. As was reported by Sangmanee and Hongpattarakere [22], the supernatant obtained from L. plantarum culturing broth could inhibit the mycelial growth and aflatoxin production of A. flavus by 100%. L. casei, L. fermentum, L. reuteri, and L. acidophilus were also proved to have an inhibitory effect higher than 80% on Aspergillus niger, Penicillium sp., and Fusarium graminearum [25]. Ahlberg et al. [26] demonstrated that Lactobacillus spp. with high or moderate anti-mycotoxigenic activities were identified as L. pentosus, L. paraplantarum, and L. plantarum. Species of L. delbrueckii subsp. Lactis were also found to completely inhibit aflatoxin G 2 production and significantly control A. parasiticus growth [27].

Streptomyces spp.
Streptomyces spp. are gram (+) filamentous bacteria that widely grow in soils and on plants.
A Streptomyces strain isolated from peanuts was found to completely inhibit, directly or via secondary metabolites, mycelial growth and conidial germination of A. flavus [62]. Streptomyces strain ASBV-1 was found to be able to reduce the viability of A. parasiticus spores and subsequently, inhibit aflatoxin accumulation in peanut grains [63]. Verheecke et al. [64] reported that several soil-born Streptomyces isolates had a strong bioactivity against aflatoxin B 1 and B 2 production by A. flavus. Several other Streptomyces species (Table 1) have been evaluated as bioactive agents providing an antagonistic activity against aflatoxigenic isolates. Shakeel et al. [28] demonstrated that culture filtrates and crude extracts of S. yanglinensis could completely inhibit mycelial growth of A. flavus. Studies demonstrated that S. anulatus [29], S. alboflavus [30], and S. roseolus [31] also exerted an effective antifungal activity toward aflatoxigenic strains and other common agricultural crops pathogens.

. Other Bacteria Species
Serratia marcescens strain JPP1 isolated from peanut hulls is an endophytic bacterium which lives inside the plant tissue and does not cause visible morphological changes. Strain JPP1 exhibited remarkable inhibitory effects on aflatoxin production (rate >98%) and mycelial growth (rate >95%) of A. parasiticus [32]. Stenotrophomonas sp., a soil bacterium, could produce inhibitors against aflatoxin production, but without affecting fungal growth [33]. Nannocystis exedens, a myxobacterium commonly found in soil, had a potential to control the growth of A. flavus and A. parasiticus by lysing pathogens' colony [35]. Palumbo et al. [34] isolated 171 bacteria from California almond orchard samples; apart from the familiar genera Bacillus and Pseudomonas, Burkholderia cepacia, B. pyrrocinia, Delftia acidovorans, D. acidovorans, and Ralstonia paucula were also demonstrated with potential activity against A. flavus growth. Achromobacter xylosoxidans, a gram-negative and catalase-positive bacterium, is already known to have wide biological control abilities [65]. Yan et al. [36] demonstrated that A. xylosoxidans could produce inhibitory substances remarkably inhibiting A. flavus and A. parasiticus growth.
According to the current studies, antagonistic bacteria were actually highly effective on aflatoxigenic strains in vitro. However, their colonization in soil and on crops has not been evaluated under field conditions. Due to genetic and environmental differences, it is not easy to bring the bacterial cells to the Aspergilli infection sites. This may be the reason why most of the anti-aflatoxigenic studies are performed only in vitro, and no bacterial agents are already commercialized.

Nontoxigenic Aspergillus spp.
A. flavus are variable in respect to aflatoxin-producing ability, and were described into S (small) and L (large) strains on the basis of sclerotial morphological types [66]. On average, S strains produce higher levels of AFB 1 with less variation in aflatoxin production [67,68]; L strains are more variable in aflatoxin production and even include nonproducers entirely lacking the ability to produce aflatoxins [69]. Currently, introduction of nontoxigenic A. flavus into fields is the most promising strategy for preventing pre-harvest aflatoxin contamination. The use of nontoxigenic A. flavus to competitively exclude aflatoxigenic strains was first introduced by Cotty and Bayman [70]. As shown in Figure 1, there have been many studies subsequently focusing on nontoxigenic Aspergillus. Prevention of aflatoxin accumulation by inoculation with nontoxigenic A. flavus CT3 and K49 was assessed in a 4year field study [37], with results indicating that the reduction percentages of aflatoxin on southern US corns were 65-94%. Alaniz Zanon et al. [38] also found that the nontoxigenic A. flavus had a higher biocontrol efficacy against aflatoxin accumulation (inhibition rate = 78-90%) in a two-year study in northern Argentina. In addition, nontoxigenic A. flavus isolates were demonstrated with an ability to reduce aflatoxin contamination of maize by a rate higher than 80% in Kenya [71]. Importantly, nontoxingenic A. flavus strains, AF36 (NRRL 18543) and Afla-Guard ® (NRRL 21882), have been commercialized for use in groundnut and maize production, respectively, in USA. This biocontrol approach has also been proved to be effective on peanut [72,73], cottonseed [69], and corn [37,74] under field conditions. An application of nontoxigenic A. parasiticus in the field was also able to reduce aflatoxin contamination in storage [39]. Nontoxigenic A. niger strain FS10, isolated from fermented soybean, could not only significantly inhibit A. flavus growth, but also inhibit AFB 1 production (rate = 94.5%) [40,75]. A. oryzae, the nontoxigenic domesticated ecotype of A. flavus, is used as a "Generally Recognized As Safe (GRAS)" microorganism for food fermentation [76,77]. Alshannaq et al. demonstrated that co-inoculation with A. oryzae and A. flavus on peanuts with a ratio of 1:100 could effectively inhibit AFB 1 production [41]. The species of A. clavatus could secrete ribonuclease [78], while Skouri-Gargouri and Gargouri revealed that A. clavatus could inhibit the growth of several plant pathogens such as Fusariuym oxysporum and Aspergillus niger due to the secretion of an antifungal peptide [42].

Trichoderma spp.
Trichoderma spp. comprise a large number of rhizocompetent filamentous strains in soils and root ecosystems. Their potential as fungal biocontrol agents against plant pathogenic fungi has been known for a long time [79]. The majority of Trichoderma isolates used industrially for biological control belong to the species T. harzianum, including strains T22 and T39 [80,81]. Trichoderma species can not only control crop diseases, but also exert beneficial effects on root growth and enhance crop productivity [82]. There have been several Trichoderma species reported showing varying degrees of control of aflatoxigenic strains since last century [79]. T. harzianum and T. viride were proved to be highly antagonistic and inhibit mycelial growth and aflatoxin production of A. flavus by a rate higher than 80% [43]. Evaluation of Trichoderma spp. for biocontrol of pre-harvest seed infection by A. flavus in groundnut was performed by Anjaiah et al. [44], with results indicating that in greenhouse and field experiments, the treatment of seeds with Trichoderma spp. including T. harzianum, T. longibrachiatum, T. viride, and T. auroviride reduced A. flavus populations (as cfu) by a percentage higher than 50%. A known fact is that Trichoderma species are historically a group of the most studied beneficial filamentous fungi. Sarrocco and Vannacci [14] gave a list of commercial bio-pesticides containing 14 different Trichoderma strains that belong to T. harzianum, T. atroviride, T. virens, T. asperellum, T. gamsii, and T. polysporum; however, these species have not been investigated as commercialized products to biocontrol aflatoxigenic molds.

Penicillium spp.
Several species of Penicillium are able to grow rapidly in the presence of toxigenic strains [83]. The strain RP42C of P. chrysogenum was reported as antifungal-protein producer with a biological activity against the growth of aflatoxigenic strains on dry-cured ham [45,84]. P. chrysogenum, a fungal starter culture for mold-fermented foods production, is related to P. nalgiovense. Nielsen et al. [85] demonstrated that P. nalgiovense showed a higher inhibitory effect on the growth of the common fungal pathogens. Additionally, Geisen [46] demonstrated that P. nalgiovense had a greater inhibition on the secondary metabolites production of fungal strains. As fungal starter cultures, the antifungal activity of Penicillium species would play an important role in the safety of mold-fermented food.

Yeast Strains
Due to the ability to consume lactic acid in the presence of oxygen, yeast strains have been regarded as deteriorating agents for a long time. Yeast strains are also popular in household because of the ability of leavening dough. Marine yeast Debaryomyces hansenii BCS003 strain can decrease mycelial growth by almost 98% in a radial inhibition assay against Aspergillus strains [47], while native D. hansenii strains were also demonstrated with a significantly antagonistic activity on the growth rate and aflatoxin production of A. parasiticus in meat products [48]. Saccharomyces cerevisiae RC008 and RC016 are strains demonstrated with the ability of inhibiting the growth of and AFB 1 production by A. parasiticus under different regimes of water activities, pH values, temperatures, and oxygen availabilities [49]. As shown in Table 1, Kluyveromyces, Pichiaanomala, and Candida maltosa isolates were also demonstrated to have an impact on mycelial growth, conidial germination, or aflatoxin production when interacting with aflatoxigenic Aspergillus strains [48,50,52,[86][87][88].

A Conclusion of Antagonistic Microbes
Many bacteria agents have been demonstrated with an ability to inhibit aflatoxigenic molds; however, none of bacterial agents has been commercialized. At current research status, only nontoxigenic A. flavus NRRL 18543 and NRRL 21882 have been commercialized and applied in fields [72], and Trichoderma species are just showing a high potential to be commercialized for the future use [14]. For yeast strains, however, we need further studies to look for strains with high efficacy.

Inhibitory Compounds Produced by Different Antagonistic Microbes
Secondary metabolites produced by various microorganisms are high-value natural products, many of which exhibit significant pharmacological properties. The inhibitory compounds discussed here are secondary metabolites with powerful bioactive properties in biological control of aflatoxin-producing fungi. Based on the results obtained in vitro experiments, inhibitory compounds produced by various antagonistic microorganisms and their bioactivities against aflatoxigenic molds are listed in Table 2. These compounds are divided into four different types of substances, including micromolecular organics, organic acids, antibiotics, and enzymes ( Figure 2). The following paragraphs describe the producers, anti-aflatoxigenic activities, and main characteristics of these compounds in more detail. Show strong antifungal ability at the lowest concentration of 1 mg/mL [92] Indole lactic acid (ILA) About 1 mg/mL was sufficient to inhibit aflatoxins production by 90% 2-butyl-4-hexyloctahydro-1Hindene, Oleic acid, palmitic acid, linoleic acid and 2,4-ditertbutylphenol In cell-free supernatant; resistant to sterilization and proteolytic enzymes [22,24] Peptides Completely inhibit A. flavus growth on corn [59] Streptomyces spp.

Antibiotics and Proteases Produced by Bacillus spp.
Bacillus species generally have characteristics to produce antimicrobial substances, mainly including lipopeptides, protease antibiotics, and bacteriocins [100]. These structurally diverse compounds exhibit a wide range of antimicrobial activity [101], especially the lipopeptides secreted by Bacillus presenting antifungal activity [102].
Bacillus strains isolated from aquatic environments were evaluated for their antifungal effect on A. flavus and A. carbonarius, producers of AFB 1 and ochratoxin A, respectively [89]. Results showed that the lipopeptides (iturin A and surfactin isomers in extracts) produced by Bacillus sp. P1 strain exhibited high anti-Aspergillus activities on mycelial growth, conidial germination, and AFB 1 and ochratoxin A production. Veras et al. [89] also analyzed the extracts from supernatants and cell pellets, and results indicated that lipopeptides were extracted mainly from cell-free supernatants. González Pereyra et al. [18] also demonstrated that lipopeptides, the extracellular compounds produced by soil Bacillus strains, were able to almost completely inhibit A. parasiticus growth and AFB 1 production. In another report [103], mutants of B. subtilis obtained after varying doses of gamma irradiation could significantly inhibit A. flavus growth and aflatoxin production in pistachio nuts compared with the parental strain, because lipopeptides production of mutants increased. Additionally, Farzaneh et al. [104] reported that cell-free supernatants from B. subtilis had a significant effect on A. flavus spores viability, and the mass spectrometric analysis revealed that surfactin and fengycin were responsible for the biocontrol activity. These studies indicated that fengycin, surfactin, and iturin families of lipopeptides produced by Bacillus species were the dominant compounds potentially reducing Aspergillus spp. growth or aflatoxins production and, generally, these compounds were obtained from cell-free supernatants. Bacillomycin D, a lipopeptide substance produced by B. subtilis, was also demonstrated with abilities of significantly affecting mycelial growth, sporulation, and destabilizing the cell wall and cell membrane of A. flavus [90].
Proteases, especially alkaline proteases, are the well-known products of Bacillus strains. B. subtilis and B. amyloliquefaciens were able to inhibit A. parasiticus growth and showed a good proteolytic activity [15]. Additionally, three peptides of L-Asp-L-Orn (D1O), L-Asp-L-Asn (D1N), and L-Asp-L-Asp-L-Asn (D2N) produced by B. megaterium could significantly inhibit the growth of A. flavus [91]. Another study reported that unknown volatiles produced by B. megaterium could inhibit aflatoxin production, mycelial growth, and conidial germination of A. flavus in rice grains [105].
Overall, we can conclude from these studies that extracellular compounds of Bacillus species were able to inhibit aflatoxigenic molds. The compounds, especially lipopeptides and proteases may be the main effective antifungal factors inhibiting aflatoxin production, sporulation, and conidial germination and reducing mycelial growth.

Chitinolytic Enzyme Produced by Pseudomonas spp.
Akocak et al. [56] demonstrated that the chitinolytic enzyme produced by P. fluorescens could reduce the growth of A. flavus by inducing the morphological changes on conidial germination and mycelial growth. As reviewed by D'Aes et al. [106], biosurfactants such as cyclic lipopeptide and rhamnolipid produced by Pseudomonas spp. were involved in important functions of biocontrol. Phenazines produced by Pseudomonas strains were also major determinants controlling several plant pathogens [107]. However, biological activities of bio-surfactants and phenazines against aflatoxigenic strains have not been investigated. Therefore, speeding up the identification of bioactive compounds could potentially enhance application values of Pseudomonas species.

Organic Acids and Peptides Produced by Lactobacillus spp.
As revealed by Russo et al. [61], Lactobacillus spp. have broad antifungal activities because of the high production of lactic acid. Apart from lactic acid, phenyllactic acid (PLA), hydroxyphenyllactic acid (OH-PLA), and indole lactic acid (ILA) were also found to strongly inhibit aflatoxin-producing fungi [25,92]. Additionally, the antifungal compounds secreted by L. plantarum were investigated against the growth of and aflatoxin production by A. flavus and A. parasiticus, with results indicating that the antifungal compounds obtained from the cell-free supernatant, apart from lactic acid, majorly were 2-butyl-4-hexyloctahydro-1H-indene, oleic acid, palmitic acid, linoleic acid, and 2,4-di-tertbutylphenol [22].
Apart from organic acids, inhibitory peptides produced by L. plantarum were also demonstrated to be effective against A. flavus and A. parasiticus [59,60], while organic acids were dominant, probably associated with their low pH values [61].

Micromolecular Organics, Organic Acids, and Enzymes Produced by Streptomyces spp.
Streptomyces spp., known to produce over 7500 bioactive compounds including anticancer agents, vitamins, and antibiotic compounds, have a better tolerance to water stress [28]. They usually do not secrete toxic residues that may contaminate environments because of their natural origin. 2-methylisoborneol, the volatile organic compound generated by S. alboflavus, was proved to have an ability of inhibiting A. flavus, Fusarium moniliforme, and Penicillum citrinum in vitro [30]. Aflastatin A, extracted from mycelial cake of Streptomyces sp., was a strong inhibitor of aflatoxin production [93]. Dimethyl trisulfide and Benzenamine, the small molecular organic compounds generated by S. alboflavus, played an important role in controlling aflatoxin production and A. flavus growth [95]. Dimethyl disulfide, the micromolecular volatile organic identified from the volatiles of S. alboflavus, was proved to act as an antagonistic substance against some plant pathogens in vitro [30]. Dioctatin A, an organic acid, produced by Streptomyces spp., was found to strongly inhibit aflatoxin production and conidiation of A. parasiticus [94]. The thermostable endochitinase purified from Streptomyces sp. [96] and the chitinase (Chi242) obtained from the culture supernatant of S. anulatus [29] have been found to inhibit the mycelial growth of A. parasiticus and A. niger, respectively. From these studies, we are able to see out that inhibitory compounds produced by Streptomyces spp. were highly species-specific. As Manivasagan et al. [108] reviewed, Actinomycetes, especially Streptomyces spp., have a tremendous potential to produce various secondary bioactive metabolites. In this case, Streptomyces species definitely have a great potential to be used for the biocontrol of aflatoxigenic fungi.

Micromolecular Organics and Enzymes Produced by Yeast Strains
Yeast strains are increasingly targeted for the production of bioactive substances, especially the budding yeast species Saccharomyces cerevisiae, which has been proven to be a powerful microorganism for heterologous expression of biosynthetic pathways [109]. The biocontrol activity of Pichia anomala WRL-076 was attributed to the production of 2-phenylethanol, which was the major volatile compound affecting the growth, aflatoxin production, and gene expression of A. flavus [51]. Studies also demonstrated that isoamyl acetate and isoamyl alcohol produced by Candida maltosa were able to inhibit the conidial germination of Aspergillus brasiliensis [52]. 4-Hydroxyphenethyl alcohol, 4,4-Dimethyloxazole, and 1,2-Benzenedicarboxylic acid dioctyl ester in the supernatant extracts of Saccharomyces cerevisiae provided the antifungal activity against aflatoxigenic growth and aflatoxins biosynthesis [97]. Tayel et al. [98] demonstrated that Pichia anomala was able to produce β-1,3-glucanase and exo-chitinase, which were suggested as a mode of antifungal action leading to cause hyphal lysis of A. flavus.

Protease and Extracellular Enzymes Produced by Trichoderma spp.
Regarding to Trichoderma species, only a few inhibitory compounds that play roles in their antagonistic interactions with aflatoxingenic fungi were reported. Deng et al. [99] demonstrated that the aspartic protease P6281 secreted by T. harzianum could efficiently inhibit the conidial germination and the growth of A. flavus. Mostafaet al. [43] demonstrated that T. harzianum and T. viride showed a high antagonism and inhibited aflatoxins production of A. flavus by 90%, which were explained partially by the liberation of extracellular enzymes and the production of inhibitory volatile compounds.

Inhibitory Compounds Produced by the Other Microorganisms
Apart from the inhibitory compounds described in the above sections, chitinase produced by Serratia marcescens was able to efficiently degrade fungal cell walls [32]. The antifungal protein PgAFP produced by Penicillium chrysogenum could inhibit the growth of toxigenic molds [84]. Antifungal peptide produced by Aspergillus clavatus was thermostable and exhibited a strong inhibitory activity against mycelial growth of several plant pathogenic fungi [42]. Cyclo (L-Leucyl-L-Prolyl) produced by Achromobacter xylosoxidans was able to inhibit the growth of A. parasiticus, and it also remarkably repressed the transcription of the aflatoxin-biosynthesis related gene aflR [36].

A Conclusion of Inhibitory Compounds
Approximately 30 different compounds have been found to be bioactive against aflatoxigenic fungi. According to these studies, we identified three deficiencies in the research field that need improvement: (1) the variety of inhibitory compounds is still limited; and (2) all of the inhibitory compounds were tested only in vitro, in which case, it is difficult to relate with the real antagonistic efficacy in vivo because of the diversity of microbes in soils, differences of soil temperature, humidity, and pH, and the genetic and metabolic complexity of biocontrol antagonists; and (3) most, even all of the studies focused only on inhibitory efficiency, however, studies such as the resistance in Aspergillus and interactions among inhibitory compound, pathogen, antagonist, and environment were scarce. These deficiencies could be mirrored by the example of Trichoderma spp. The antagonistic Trichoderma strains have the ability to produce various compounds with antibiotic activity [81]. However, few antibiotic compounds have been identified from Trichoderma spp. for the biocontrol of aflatoxigenic molds. Although Trichoderma species play an important role in biocontrol of plant diseases, frequently enhance root growth, and induce systemic resistance responses of plants [82], the interaction among aflatoxigenic fungi, Trichoderma, soil, and plants has not been elucidated yet.

Inhibitory Mechanisms by Antagonistic Bacteria
For antagonistic bacteria, their bioactive metabolites play a major role in controlling Aspergillus spp. growth and subsequent aflatoxin production. Inhibitory mechanisms by antagonistic bacteria mainly include (1) lysis of hyphae or spores by destablizing structure and composition of cell wall; (2) probably affecting intracellular activities of mitochondria, cytoplasmic membrane, and nucleus; and (3) down-regulating expression of aflatoxin-synthesis related genes. Illustrations were made as follows: chitinolytic enzymes produced by P. fluorescens reduced the growth of A. flavus by altering the germination pattern of spores [56]; the cell-free supernatant of L. plantarum caused morphological changes in seven-day-old A. flavus and A. parasiticus, because of severe damage to the mitochondria and nucleus, formation of the membrane-bound vesicles, and degeneration of the cytoplasmic membrane [22]; and dioctatin A produced by Streptomyces decreased expression of aflR and brlA (encoding a condition-specific transcription factor) and significantly inhibited the production of norsolorinic acid and sterigmatocystin that were precursors for aflatoxin synthesis [94].

Inhibitory Mechanisms by Nontoxigenic Aspergillus spp.
Fungal invasion, colonization, and competition between aflatoxigenic and atoxigenic strains of A. flavus have been studied [70,110]. Regarding nontoxigenic Aspergillus spp. as antagonist, two mechanisms are dominant: (1) toxigenic strains are physically excluded by the displacement of nontoxigenic strains during infection; and (2) nontoxigenic strains competed for nutrients that were required for aflatoxin biosynthesis. However, as Ehrlich [111] reviewed, there were a lot of challenges to using nontoxigenic Aspergillus species. Primarily, due in part to inherent diversity of Aspergillus species and genetic complexity, genetic mutations may happen in nontoxigenic Aspergillus spp., which potentially leads atoxigenic strains to mutate to aflatoxigenic strains; therefore, from a long-term security, nontoxigenic Aspergillus strains were also suggested to be cautiously used [112][113][114].

Inhibitory Mechanisms by Antagonistic Yeasts
How did the antagonistic yeasts act as biological agents to control aflatoxigenic growth and aflatoxin production? That the yeast strain Pichia anomala could efficiently inhibit the growth of and aflatoxin production by A. flavus can be attributed to the production of 2-phenylethanol, which led to remarkable effects on conidial germination and expression of genes necessary for aflatoxin biosynthesis [51], and the production of chitinase and glucanase, which led to hyphal lysis and deterioration [98]. That Debaryomyces hansenii was able to control A. flavus growth was attributed to the production of extracellular compounds and the competition for nutrients and spaces [47]. For Saccharomyces cerevisiae, the production of exochitinase and extracellular secondary metabolites could explain its mode of action for antifungal activity on the growth of A. flavus [97]. Therefore, for yeast strains, possible mechanisms of the inhibitory actions may involve two: (1) inhibiting aflatoxigenic growth by the production of extracellular enzymes and metabolites which lead to spores and hyphal deterioration, and (2) inhibiting aflatoxin production by down-regulating expression of aflatoxin biosynthesis genes.

Inhibitory Mechanisms by Antagonistic Trichoderma Strains
The antagonistic properties of Trichoderma strains are based on the activation of multiple physical and chemical mechanisms. The physical mechanisms included faster growth speed to compete for nutrients and living space, and mycoparasitism mediated by physical contact. Common interactions between antagonistic fungi and pathogens were divided into the following types [79]: 1 = antagonist overgrowing pathogen and pathogen stopped; 1/2 = antagonist overgrowing pathogen but pathogen still growing; 2/1 = pathogen overgrowing antagonist but antagonist still growing; 2 = pathogen overgrowing antagonist and antagonist stopped; 3 = mutual inhibition ≤2mm distance; 4 = extremely mutual inhibition >4mm distance.
Calistru et al. [115] discovered only three interaction types between Trichoderma and A. flavus, namely antagonist overgrowing pathogen with growth inhibition of pathogen, pathogen overgrowing antagonist with growth inhibition of antagonist, and mutual inhibition. By a scanning electron microscopical investigation, Calistru et al. [115] revealed that mycoparasitism is not the mechanism of the inhibitory interaction between A. flavus and Trichoderma spp. (T. harzianum and T. viride). Conversely, Mostafa et al. [43] drew a conclusion that the aggressive behavior towards A. flavus by T. harzianum was explained by mycoparasitism.
The chemical mechanisms were also involved in producing cell walllytic enzymes and inducing the plant's defense system to resist pathogens [116]. The production of extracellular enzymes was responsible for the inhibitory effect of T. viride on toxigenic A. flavus [43]. T. harzianum actively attached to the toxigenic Aspergillus species followed by enzymatic lysis of the mycelial filaments [117]. Such, mechanisms of the inhibitory actions against the growth of A. flavus by T. harzianum are strains-specific and mainly include (1) faster growth speed to compete for nutrients and living space, (2) mycoparasitism, and (3) the production of extracellular enzymes, deteriorating aflatoxigenic mycelia. However, the research on the mechanism of inhibitory effects on aflatoxin production is still at initial stage.

A Conclusion of Mechanisms
According to all of the above studies, we listed four main mechanisms of inhibitory actions ( (3) affecting intracellular activities of mitochondria, nucleus, and cytoplasmic membrane, and (4) down-regulating expression of aflatoxin-synthesis related genes. Importantly, inhibitory actions are most likely determined by a combination of different mechanisms, not by only one.We also listed some genes that have been analyzed under the treatment of different biocontrol agents (Figure 4). at initial stage.

A Conclusion of Mechanisms
According to all of the above studies, we listed four main mechanisms of inhibitory actions ( Figure 3): (1) Physically competing for living spaces and nutrients, (2) destabilizing cell wall structure, (3) affecting intracellular activities of mitochondria, nucleus, and cytoplasmic membrane, and (4) down-regulating expression of aflatoxin-synthesis related genes. Importantly, inhibitory actions are most likely determined by a combination of different mechanisms, not by only one.We also listed some genes that have been analyzed under the treatment of different biocontrol agents ( Figure 4).

A Conclusion of Mechanisms
According to all of the above studies, we listed four main mechanisms of inhibitory actions ( Figure 3): (1) Physically competing for living spaces and nutrients, (2) destabilizing cell wall structure, (3) affecting intracellular activities of mitochondria, nucleus, and cytoplasmic membrane, and (4) down-regulating expression of aflatoxin-synthesis related genes. Importantly, inhibitory actions are most likely determined by a combination of different mechanisms, not by only one.We also listed some genes that have been analyzed under the treatment of different biocontrol agents ( Figure 4).

Factors Influencing Antifungal Activities
It is a well-known fact that the growth rate of and aflatoxin production by aflatoxigenic strains were strongly influenced by environments, cultural conditions, and nutritional factors. The combined effects of incubation time, temperature, water activity (aw), and CO 2 on the growth of and aflatoxin production by A. flavus were studied [119]. Nutritional sources were also demonstrated to have a significant influence on fungal growth and mycotoxin production [120,121]. Additionally, the expression of aflatoxin-synthesis related genes were demonstrated to be highly in relation to changes in water activity and temperature levels [122][123][124]. Similarly, antifungal activities of various biocontrol microbes were also related to these biotic and abiotic factors. Examples are described here below.

pH Value
Studies showed that the bioactivity of L. plantarum was pH-dependent. The low pH was responsible for the highlighted bioactivity of L. plantarum against aflatoxin-producing strains [22,61]. Gerez et al. [25] demonstrated that the antifungal activity of some Lactobacillus strains was lost after the neutralization treatment because the acidic nature of the antifungal metabolites was destroyed. In addition, Saccharomyces cerevisiae RC008 and RC016 showed a great antagonistic activity at pH 4, where strains can highly decrease the growth rate of A. parasiticus [49]. Conversely, the bacterium Bacillus pumilus grew very slightly at pH 4, where it showed the lowest anti-aflatoxigenic activity only with 38% inhibition of aflatoxin production [17]. These studies indicated that the best pH value for different antagonists against aflatoxigenic molds is remarkably species-dependent.

Temperature and Water Activity
Culturing temperature and water activity (aw) are also key factors. The maximum activity of protease P6281 produced by T. harzianum was observed at 40 • C [99]. The appropriate conditions for the growth of Kluyveromyces spp. were 60 min of incubation at 45 • C and 0.95 aw [125], while Penna and Etcheverry [86] demonstrated that Kluyveromyces isolates could impact both A. flavus growth and AFB 1 accumulation at a wide range of water activities (0.93-0.99). La Penna et al. [50] found that several Kluyveromyces isolates showed anti-aflatoxigenic activity and inhibitory activity on aflatoxin production at all water activities tested. A notable finding was that the yeast strains of Debaryomyces hansenii could stimulate aflatoxins production by A. parasiticus at water activity of 0.99, whereas significantly reduce aflatoxins production at 0.92 aw [48]. Therefore, temperature and water activity are also important factors influencing antifungal efficiency of antagonists.

Other Factors such as Incubation Time, Culturing Medium, and Mutagenesis
Furthermore, incubation time is also a key factor affecting the production of anti-aflatoxigenic metabolites. Munimbazi and Bullerman [17] gave an evident proving that the greatest inhibitory activity arose up after 3 and 4 days incubation of Bacillus pumilus, and aflatoxin production was completely inhibited in supernatant obtained only from 3 and 4 day old bacterium. Whipps [79] demonstrated that different media appeared to be related to antifungal behaviors. Afsharmanesh et al. [103] found that a random mutagenesis of Bacillus subtilis could significantly inhibit A. flavus growth and aflatoxin production compared with the parental strain. This shows that mutant study can potentially improve biocontrol activity in inhibiting aflatoxigenic strains.
As shown in Figure 5, incubation conditions such as growing period, temperature, water activity, pH values, and nutritional sources could not only influence pathogens', but also antagonists' growth and/or metabolism. Therefore, dynamic growing conditions should be taken into account in performing strategies to biocontrol aflatoxigenic molds and eliminate aflatoxin risk by aflatoxigenic fungi.
Toxins 2020, 12, 24 15 of 22 growth and/or metabolism. Therefore, dynamic growing conditions should be taken into account in performing strategies to biocontrol aflatoxigenic molds and eliminate aflatoxin risk by aflatoxigenic fungi. Figure 5. Key factors influencing antifungal activities against aflatoxigenic fungi. These key factors have influences on both baflatoxigenic and antagonists' growth and metabolisms. As a result, the combination of these factors is playing an important role in biocontrol efficacy.

Perspective and Conclusions
The biocontrol strategy for preventing aflatoxigenic fungi has been discussed in this review. It is clear that some microbes, including various bacteria, nontoxigenic Aspergillus, Trichoderma, and yeasts have shown potentials to biocontrol aflatoxigenic molds. The inhibitory compounds that have potential biocontrol effects on aflatoxigenic strains, together with mechanisms and influencing factors of the bioactive actions are also reviewed. The current research status is still not very optimistic, because there are still many aspects needing urgent improvements. The above reviewed research works do, however, suggest that deeper practical works must be conducted to identify effective and environmental biocontrol agents, substantially to reach an advanced stage of application and commercialization. Additionally, a comprehensive and systematic study, covering inhibitory behaviors, mechanisms, factors, and pathogen-antagonist-plant interactions, is also urgently needed.

Perspective and Conclusions
The biocontrol strategy for preventing aflatoxigenic fungi has been discussed in this review. It is clear that some microbes, including various bacteria, nontoxigenic Aspergillus, Trichoderma, and yeasts have shown potentials to biocontrol aflatoxigenic molds. The inhibitory compounds that have potential biocontrol effects on aflatoxigenic strains, together with mechanisms and influencing factors of the bioactive actions are also reviewed. The current research status is still not very optimistic, because there are still many aspects needing urgent improvements. The above reviewed research works do, however, suggest that deeper practical works must be conducted to identify effective and environmental biocontrol agents, substantially to reach an advanced stage of application and commercialization. Additionally, a comprehensive and systematic study, covering inhibitory behaviors, mechanisms, factors, and pathogen-antagonist-plant interactions, is also urgently needed.

Conflicts of Interest:
The authors declare no conflicts of interest.