Seed Priming with Rhizospheric Bacillus subtilis: A Smart Strategy for Reducing Fumonisin Contamination in Pre-Harvest Maize

Maize, one of the most important cereal crops in Bangladesh, is severely contaminated by fumonisin, a carcinogenic secondary metabolite produced by Fusarium including Fusarium proliferatum. Biocontrol with Bacillus strains is an effective approach to controlling this F. proliferatum as Bacillus has proven antagonistic properties against this fungus. Therefore, the present study aimed to determine how native Bacillus strains can reduce fumonisin in maize cultivated in Bangladesh, where BDISO76MR (Bacillus subtilis) strains showed the highest efficacy both in vitro in detached cob and in planta under field conditions. The BDISO76MR strain could reduce the fumonisin concentration in detached cob at 98.52% over untreated control, by inhibiting the conidia germination and spore formation of F. proliferatum at 61.56% and 77.01%, respectively in vitro. On the other hand, seed treatment with formulated BDISO76MR showed higher efficacy with a reduction of 97.27% fumonisin contamination compared to the in planta cob inoculation (95.45%) over untreated control. This implies that Bacillus-based formulation might be a potential approach in mitigating fumonisin contamination in maize to ensure safe food and feed.


Introduction
Maize or corn (Zea mays L.), a crop of the Poaceae family, is an emerging cereal whose popularity has been increasing for diverse uses, i.e., food, fodder, and edible oil [1].In Bangladesh, maize is the most productive cereal followed by wheat and rice [2].However, maize cultivation has been severely hampered by Fusarium spp., which produces a harmful secondary metabolite known as fumonisins [3].Fusarium spp. that produces fumonisin can infect many crops, foods, and feedstuffs, and its contamination can easily be detectable in the field [4].The presence of fumonisin in maize grains was found in 15 major maizeproducing districts of Bangladesh, where 10 districts exceeded the European Union's permissible limit for fumonisin [5].The most prevalent fumonisins are FB1, FB2, and FB3, with FB1 being the most poisonous and widely distributed [6].Animals and humans exposed to FB1 can experience acute toxicity, immunotoxicity, organ toxicity, and reproductive toxicity [7].Fumonisin exposure has been linked to esophageal and liver cancers in South Africa and Asia; however, the data are inconclusive and it is categorized as possibly carcinogenic to humans [8,9].Fumonisin is non-genotoxic and disrupts sphingolipid metabolism, affecting various cellular activities.This may contribute to the toxicity and carcinogenicity of fumonisin, as it has structural similarities with the long-chain (sphingoid) base backbones of sphingolipids [6].Mothers who consumed a lot of fumonisin-contaminated maize during the first trimester of pregnancy had a higher risk of developing neural tube defects in their unborn children [10].According to studies performed on Tanzanian youth, fumonisin exposure has been linked to stunted growth in children [11,12].
To control the fumonisin contamination in maize, different methods are used such as synthetic fungicides [13], natural resources, plant materials, microbial cultures, genetic materials [3], clay minerals [14], and enzymes [15].However, it is not easy to find effective and environmentally friendly approaches against the Fusarium and mycotoxin production in maize [3].Researchers are still searching for effective biologically safe alternatives to stop this contaminant from getting into the food chain [3].However, several rhizobacteria have shown notable antifungal properties to prevent fumonisin contamination in maize plants [16,17].The Bacillus is a rhizobacteria that can survive in various growth conditions and produce aerobically dormant endospores [18].The different genetic characteristics of Bacillus make them anti-pathogenic because they can actively reproduce and withstand adverse conditions [19].Numerous Bacillus strains were found to be efficient biocontrol agents that offered rhizosphere colonization and root disease control, as well as fumonisin contamination reduction in maize grains [20][21][22].
Fumonisin contamination is a significant risk to Bangladesh [23][24][25] and its contamination in maize needs to be monitored to ensure safe food and feed [5].A recent study showed that maize rhizospheric Bacillus can promote plant growth by reducing fumonisin contamination in stored maize grain [26].Therefore, five Bacillus strains (BDISO01RR: Bacillus amyloliquefaciens, BDISO76MR: Bacillus subtilis, BDISO36PR: Bacillus subtilis, BDISO45PR: Bacillus subtilis and BDISO49PR: Bacillus subtilis) were tested and identified in in vitro conditions, whereas all of them were effective against fumonisin production by colonizing the roots and inhibiting the mycelial growth of Fusarium proliferatum on maize grain [27].So, the present study was carried out to evaluate the potentiality of native Bacillus isolates in minimizing fumonisin production under field conditions.

Effects of Bacillus in Reducing Fumonisin Contamination In Vitro
Influence of Different Bacillus Isolates on Reducing Fumonisin Accumulation in Co-Inoculated Detached Cobs under In Vitro Condition An in vitro experiment was conducted to determine how different Bacillus isolates can reduce fumonisin concentrations in co-inoculated detached maize cobs.The lowest amount of fumonisin (3.63 ppm) was quantified when maize kernels were treated with BDISO76MR (B.subtilis), which was able to reduce the fumonisin concentration by 98.52% over control followed by BDISO01RR (5.06 ppm), BDISO49PR (9.55 ppm), BDISO45PR (33.10 ppm) and BDISO36PR (40.1 ppm) with reduction percentage 97.94,96.11, 86.52 and 83.67%, respectively as compared to untreated control (Table 1).However, maize kernels treated only with distilled water remained healthy (Figure 1).

In Vitro Ultrastructural Changes in F. proliferatum Hyphae by Bacillus Isolates
At 7 DAI, the PDA plates of Bacillus viz.BDISO76MR (B.subtilis) and BDISO01RR (B.amyloliquefaciens) showed mycelial deformation of F. proliferatum among the five isolates.The mycelial structure of F. proliferatum was destructed and deformed, including curling, shrinkage, and breakdown by BDISO76MR (B.subtilis) and BDISO01RR (B.amyloliquefaciens) isolates (Figure 2B,C) as compared to the control.The F. proliferatum hyphae in the untreated control were also found to be dense, regular, plump, and undamaged through microscopic examination (Figure 2A).
lates.The mycelial structure of F. proliferatum was destructed and deformed, including curling, shrinkage, and breakdown by BDISO76MR (B.subtilis) and BDISO01RR (B.amyloliquefaciens) isolates (Figure 2B,C) as compared to the control.The F. proliferatum hyphae in the untreated control were also found to be dense, regular, plump, and undamaged through microscopic examination.(Figure 2A).

Discussion
Bacillus, one of the largest genera of bacteria, is a potential biological control agent because of its antibiotic metabolite production and different genetic characteristics [18,19].In our laboratory, five Bacillus isolates viz.BDISO76MR (B.subtilis) from maize rhizosphere, BDISO01RR (B.amyloliquefaciens) from rice rhizosphere [28], BDISO36PR (B.subtilis), BDISO45PR (B.subtilis) and BDISO49PR (B.subtilis) from potato rhizosphere [29] were identified and examined their potentiality as a biocontrol agent against Fusarium proliferatum of maize in vitro condition [27].This study revealed the efficacy of these five Bacillus isolates to inhibit F. proliferatum of maize both in vitro and in planta, where BDISO76MR (B.subtilis) showed the highest efficacy in both conditions.
To assess the reduction in fumonisin accumulation, different Bacillus strains and F. proliferatum were co-inoculated in detached maize cobs in vitro, where BDISO76MR (B.subtilis) and BDISO01RR (B.amyloliquefaciens) showed the highest reduction in fumonisin accumulation over control by 98.52% and 97.94%, respectively.The reduction in fumonisin concentration might be a result of mycelial growth reduction in F. proliferatum by Bacillus strains.A similar result was found when combining F. udum with Bacillus and Pseudomonas species resulting in a considerable reduction in mycelial growth and mycotoxin accumulation in vitro [21].
Therefore, an in vitro dual-culture assay was conducted to reveal the effects of Bacillus strains on the reduction in radial mycelial growth of the F. proliferatum, where BDISO76MR (B.subtillis) and BDISO01RR (B.amyloliquefacience) were found to be the most effective.This reduction in mycelial growth might be because Bacillus spp.can produce antimicrobial substances like subtilin, bacilysin, mycobacillin, bacillomycin, mycosubtilin, iturins, fengycins, and surfactins [30].Although the mechanisms of all the antimicrobial substances are still unrevealed, some researchers have proven how they induce anti-pathogenic properties.Zalila-Kolsi et al. [31] found that the FZB42 strain of B. amyloliquefaciens can produce Bacillomycin D, which affects the plasma membrane morphology and cell wall of F. graminearum by inducing reactive oxygen species (ROS) accumulation.The inhibitory activity of surfactins was proven against F. verticillioides, and F. oxysporum because they can produce specific cationic channels in the membrane phospholipid bilayer, which might possess the antibiotic properties [32].Fengycin, another antimicrobial component produced by Bacillus against the filamentous fungi, inhibits the enzyme phospholipase A2 and aromatase functions and it could alter the structure of fungal cell membranes, increasing permeability, and produce permanent lesions that compromise the integrity of fungal cells on F. graminearum [33].However, a different study discovered the inhibitory power of fengycin against different Fusarium spp.[34].On the other hand, iturins demonstrate potent fungi toxic qualities by creating ion-conducting pores when they come into contact with fungal membranes and it is effective against F. oxysporum [35][36][37], and F. graminearum [31,38].
The inhibition of germination and formation of F. proliferatum conidia by Bacillus strains are an insightful outcome of the experiment.This was observed under a ZEISS Premio Star microscope, whereas the best view of fungal hyphae could be achieved by SEM and TEM analysis [39].The highest conidiation reduction was observed in BDISO76MR (77.01%), followed by BDISO01RR (71.09%) at 72 HPI over the untreated control.A similar study was conducted in vitro, where B. subtilis SG6 exhibited a high antifungal effect on the germination of F. graminearum with an inhibition rate of 95.6% over untreated control [40].Other studies discovered that the Bacillus species can induce their antagonistic activity in reducing the mycelial growth of several Fusarium species [20], while Baard et al. [41] found B. subtilis can suppress mycelial growth of F. proliferatum as same as this present study.Gong et al. [38] examined the antagonism of B. amyloliquefaciens S76-3 in wheat inoculated with F. graminearum and found that iturin A can kill the conidia of F. graminearum at minimal inhibitory concentration.Therefore, further research is necessary to find out which substances are responsible for the inhibitory power of BDISO76MR (B.subtilis) against F. proliferatum.
In addition to the biochemical analysis of the antagonistic substances, Adeniji et al. [42] examined seven Bacillus isolates with bio-suppressive effects on F. graminearum and discovered that those isolates had important gene clusters encoding biocontrol agents.Similarly, Chen et al. [11] characterized the genome of the Bacillus velezensis LM2303 strain, which has a strong biocontrol potential against F. graminearum.So, the characterization of BDISO76MR (B.subtilis) might add more precise information in controlling F. proliferatum.
In another study, fumonisin accumulation was evaluated in planta through seed treatment and co-inoculation of maize cob by different Bacillus strains.When seeds were treated with five individual Bacillus strains, BDISO76MR (B.subtilis) was found to be more effective in the case of reduction in fumonisin accumulation in planta maize cob.Again, similar findings were found when maize cobs were co-inoculated with Bacillus strains in planta, where the co-inoculation with BDISO76MR (B.subtilis) had the highest efficacy, with a reduction in fumonisin accumulation by 95.45% against F. proliferatum infection over untreated control.But seed treatment is better than the co-inoculation of Bacillus in planta, which might be because of the volatile metabolite produced by Bacillus that can trigger the induced systemic resistance in plants and so activate the plant defense mechanism [43].The host defense response is induced by the metabolites also [31].Additionally, Bacillus species work against fungal pathogens by generating fungi toxic compounds, competing together with them for nutrients, and resulting in systemic acquired resistance in plants [44][45][46].Our speculation is it would be interesting to see whether these Bacillus isolates can induce the expression of some Fumonisin biosynthesized gene in planta.So, the transcriptomic analysis from RNA seq would be the next step to unravel the complex network related to this induced resistance.It is also revealed that fengycin secreted by Bacillus could result in structural deformation of the hyphae of F. graminearum and reduce proliferation and mycotoxin production in planta, as well as hyphal permeabilization, and determine ear rot development in maize [40,47].

Conclusions
BDISO76MR (Bacillus subtilis) performed best against fumonisin production both in vitro and in planta among the five Bacillus isolates.The BDISO76MR (B.subtilis) strains demonstrated a powerful inhibitory effect on reducing the fumonisin concentration in inoculated maize cob and the mycelial growth of Fusarium proliferatum in vitro.In field conditions, the BDISO76MR (B.subtilis) strains showed the highest efficacy for seed treatment and co-inoculation, whereas seed treatment is better than co-inoculation.Therefore, Bacillus-based formulation could be a possible strategy to reduce fumonisin contamination in maize farming.

Conidial Suspension Preparation for In Vitro Co-Inoculation of Maize Kernels
A total of 1-2 loops from each bacterial isolate were taken in 20 mL distilled water in a falcon tube and the cell density was fixed at 1 × 10 8 CFU/mL.Subsequently, in another falcon tube, a 3-5 mm mycelial block of F. proliferatum from pure culture was added in 20 mL distilled water and the cell density was set at 1 × 10 6 spores/mL.The conidial suspensions were filtered through double-layer cheese cloth.The cell density was measured with a hemacytometer.5.1.3.In Vitro Co-Inoculation of Maize Kernels (Detached Cob) with F. proliferatum and Different Bacillus Isolates An in vitro experiment was conducted to detect the effect of different Bacillus isolates on F. proliferatum in maize kernels.Healthy maize cobs (cv.Everest hybrid maize seed; Company: Bayer Crop Science Ltd.Dhaka, Bangladesh) were harvested at their maturity stage and brought to the laboratory immediately.Then, the cobs were de-husked and cob surfaces were sterilized with 75% ethanol followed by washing with ddH 2 O.The sterilized kernels were scratched using sterilized toothpicks.The maize kernels were then inoculated with 50 µL (10 8 CFU/mL) of each Bacillus suspension (mentioned above) and 50 µL (10 6 spores/mL) of F. proliferatum conidial suspension (mentioned above).However, maize cobs treated with 25% methanol and 100 µL distilled water were considered as controls.The inoculated maize cobs were incubated at 25 • C and 100% humidity for 7 days with 12 h of daylight per day.The experiment was repeated in three replications.

Conidial Suspension Preparation for In Vitro Mycelial Growth Inhibition of F. proliferatum by Bacillus Isolates
From the bacterial pure culture, 2-3 loops of each isolate were transferred to 250 mL Erlenmeyer flasks in 100 mL sterile water, and cell densities were adjusted to 1 × 10 8 CFU/mL with the help of a hemacytometer.Fungal conidia suspension was prepared by inoculating 50 mg of fresh mycelia of F. proliferatum in 20 mL CMC medium (1 g NH 4 NO 3 , 1 g KH 2 PO 3 , 0.5 g MgSO 4 •7H 2 O, 1 g yeast extract, 15 g carboxymethyl cellulose, and 1 L ddH 2 O) and cultured at 25 • C, 180 rpm for 4 days [39].The number of conidia in the suspension was determined by 10 6 spores/mL using a hemacytometer.

In Vitro Mycelial Growth Inhibition of F. proliferatum by Bacillus Isolates
A dual-culture assay was performed to observe the morphological alterations of the F. proliferatum hyphae by different Bacillus isolates.The experiment was conducted three times with three replications.Each Bacillus isolate was then streaked on a potato dextrose agar (PDA) plate making a triangular loop with a sterile toothpick.Then, a 5 mm F. proliferatum mycelial block was placed at the center of the PDA plates.The plates were incubated at 25 • C for 7 days.PDA plates with F. proliferatum mycelial block were used as control.After 7 days the changes in hyphal morphology were observed at the growth junction between the edges of the bacterial growth and the fungal colony using a ZEISS Premio Star microscope, and the hyphal morphological structures were captured by ZEISS AxiocamERc 5s lens.

The Conidia Formation Assay
To determine the conidia formation rate, 5 mm fresh mycelial block of potential F. proliferatum and 500 µL Bacillus suspensions (10 8 CFU/mL) were added to 20 mL CMC media and incubated at 25 • C with 180 rpm for 3 days [39].Here, distilled water was used as control.The number of conidia formed by F. proliferatum was counted by a hemacytometer at 12, 24 and 72 h post-inoculation (HPI) and expressed in percentage over control.The experiment was repeated thrice with three replicates of each experiment.

The Conidia Germination Assay
The conidia formation assay was performed as described by Yu et al. [39].Briefly, 1 mL of F. proliferatum conidial suspension (10 6 spores/mL) and 500 (10 8 CFU/mL) of five different Bacillus isolates were inoculated in CMC media and cultured at 25 • C and 180 rpm.The 25% (v/v) methanol was used as a control.When the spores reached at least half the length of the germ tube, it was considered germinated spores.The germination rates were recorded at 12 and 24 h intervals using a ZEISS Premio Star microscope, where a minimum of 200 conidia were counted for each replication.Each experiment was replicated three times.

Crop Husbandry Using Seed Treatment Formulation of Bacillus Strains
To test the potential of the seed, all of the Bacillus strains were formulated using talc powder as described previously by Islam et al. [48].Firstly, the bacterial strains from their pure cultures were grown on LB agar media for 24 h.The 2-3 loops (10 8 CFU/mL) from each bacterial isolate were transferred in LB broth for about 6 h.The bacteria cultures were then centrifuged and resuspended in 200 mL peptone broth with bactopeptone.The broth cultures were then kept in a shaker for 2 h to grow.After that, 5 mL sterile 100% glycerol was added to 200 mL broth cultures of Bacillus.
Subsequently, 500 g talcum powder was mixed with 5 g carboxy methyl cellulose and 7.5 g Calcium carbonate.Here, calcium carbonate was used to adjust the pH at 7 and carboxy methyl cellulose worked as an adjuvant.Then, the powder mixture was autoclaved for two consecutive days at 121 • C under 15 PSI pressure for 30 min.
Then, 200 mL of each bacterial culture (fortified with 1% peptone and 1% glycerol) was added to 500 g of prepared talcum powder.The formulations were then dried overnight in a laminar flow hood.Finally, the formulations were powdered by hand and packed in plastic bags.The formulated bacterial antagonists could be stored at both room and 4-8 • C temperatures in the refrigerator.

Maize Production with Treated Seeds and Management
Hybrid maize seeds (cv.Everest hybrid maize seed; Company: Bayer Crop Science Ltd., Dhaka, Bangladesh) were treated with the above-mentioned formulated Bacillus at 10 g/kg maize seeds.After treatment, half an hour was waited to allow the formulations to adhere to the seed surface.In addition, the shelf-life of the best-formulated Bacillus was also tested.Control seeds were treated in 200 mL of distilled water only.The treated seeds were then planted manually in a 315 m 2 (21 m × 15 m) plot of a farmer's field at Sutiakhali, Mymensingh, Bangladesh following a randomized complete block design (RCBD).The whole plot was divided into three subplots measuring 27 m 2 (9 m × 3 m) at every 3 m distance.Then, subplots were divided into 10 rows and each row was divided into 10 hills (row-to-row distance 0.75 m and hill-to-hill distance 0.25 m).A single row was used for a single treatment with each subsequent row containing a control.Here, three subplots were considered as three replicates.Two treated maize seeds were sown on each hill.After two weeks when the plants grew, one single plant was kept in every hill.The plants were nourished with proper fertilizer doses following the Fertilizer Recommendation Guide (2018) provided by BARC [49] and irrigated when needed.The cobs were harvested at their maturity and forceps were used to remove kernels from each cob.Five cobs were maintained in each treatment.Lastly, kernels were kept in separate envelopes and refrigerated at 4 • C until analysis.

Maize Production with Co-Inoculation of Maize Cobs with F. proliferatum and Different Bacillus Isolates in the Field Condition
Fresh hybrid maize seeds of Everest hybrid maize seed (Company: Bayer Crop Science Ltd., Dhaka, Bangladesh) were planted manually in another 315 m 2 plot of a farmer's field

Table 1 .
Reduction in fumonisins accumulation in inoculated maize grain with different Bacillus strains and F. proliferatum.Untreated control: seed treatment with only distilled water.Three replicates were used for each treatment.* %Reduction over control = [(Fumonisin concentration in uninoculated maize grains − Fumonisin concentration in inoculated maize grains)/Fumonisin concentration in uninoculated maize grains] × 100.

Table 1 .
Reduction in fumonisins accumulation in inoculated maize grain with different Bacillus strains and F. proliferatum.

Table 2 .
Effects of Bacillus isolates on conidial formation and germination of F. proliferatum in vitro." represents significance at a 0.1% level of probability.Values with the same letter are statistically similar.Data are the averages of three replications.Each value is the average of three tests for different types of Bacillus.Control: only F. proliferatum.* %Reduction in conidia formation = [(conidial formation in uninoculated maize grains −conidial formation in inoculated maize grains)/ conidial formation in uninoculated maize grains] × 100.** %Reduction in conidia germination = [(conidial germination in uninoculated maize grains − conidial germination in inoculated maize grains)/conidial germination in uninoculated maize grains] × 100.
2.4.Effects of Bacillus in Reducing Fumonisin Contamination in Planta 2.4.1.Fumonisin Concentration of Maize Grains Raised from Treated Seed with Bacillus Formulation

Table 3 .
Effects of different Bacillus strains on F. proliferatum in seed treatment and co-inoculated maize cob in the field.Fumonisins Concentration of Maize Grains Obtained from the Co-Inoculated Cob with F. proliferatum and Bacillus Isolates "***" Represents significance at a 0.1% level of probability.Values with the same letter are statistically similar.Data are the averages of three replications.Control: seed treatment with only distilled water.* %Reduction over control = [(Fumonisin concentration in uninoculated maize grains − Fumonisin concentration in inoculated maize grains)/Fumonisin concentration in uninoculated maize grains] × 100.