The Potential of Bacilli-Derived Biosurfactants as an Additive for Biocontrol against Alternaria alternata Plant Pathogenic Fungi

Fungal diseases caused by Alternaria alternata constitute a significant threat to the production and quality of a wide range of crops, including beans, fruits, vegetables, and grains. Traditional methods for controlling these diseases involve synthetic chemical pesticides, which can negatively impact the environment and human health. Biosurfactants are natural, biodegradable secondary metabolites of microorganisms that have also been shown to possibly have antifungal activity against plant pathogenic fungi, including A. alternata being sustainable alternatives to synthetic pesticides. In this study, we investigated the potential of biosurfactants of three bacilli (Bacillus licheniformis DSM13, Bacillus subtilis DSM10, and Geobacillus stearothermophilus DSM2313) as a biocontrol agent against A. alternata on beans as a model organism. For this fermentation, we describe using an in-line biomass sensor monitoring both permittivity and conductivity, which are expected to correlate with cell concentration and products, respectively. After the fermentation of biosurfactants, we first characterised the properties of the biosurfactant, including their product yield, surface tension decrement capability, and emulsification index. Then, we evaluated the antifungal properties of the crude biosurfactant extracts against A. alternata, both in vitro and in vivo, by analysing various plant growth and health parameters. Our results showed that bacterial biosurfactants effectively inhibited the growth and reproduction of A. alternata in vitro and in vivo. B. licheniformis manufactured the highest amount of biosurfactant (1.37 g/L) and demonstrated the fastest growth rate, while G. stearothermophilus produced the least amount (1.28 g/L). The correlation study showed a strong positive relationship between viable cell density VCD and OD600, as well as a similarly good positive relationship between conductivity and pH. The poisoned food approach in vitro demonstrated that all three strains suppressed mycelial development by 70–80% when applied with the highest tested dosage of 30%. Regarding in vivo investigations, B. subtilis post-infection treatment decreased the disease severity to 30%, whereas B. licheniformis and G. stearothermophilus post-infection treatment reduced disease severity by 25% and 5%, respectively. The study also revealed that the plant’s total height, root length, and stem length were unaffected by the treatment or the infection.


Introduction
In recent decades, the toxicity of chemically synthesised fungicides has become an increasing problem, so technologies that offer more sustainable, environmentally friendly solutions have come to the fore in all areas of industry [1]. Surfactants are molecules that can reduce surface tension and are used in various applications, including cleaning products, detergents, cosmetics, and agriculture. Synthetic surfactants have traditionally been used for these purposes, but there has been a growing interest in using surfactants produced by living organisms, particularly microorganisms, as a sustainable alternative. Biosurfactants have several advantages over synthetic surfactants, including being less

Isolation of Biosurfactant
First, cells were removed from the fermented broth by centrifuging at 4000 rpm for 20 min with Janetzki K23D (MLW, GDR) set to 4 • C. The biosurfactants were then recovered using acid precipitation from the clear supernatant. The pH of the supernatant was set to 2 using 6 M HCl, and the temperature was maintained at 4 • C overnight. This precipitated the biosurfactants out of the solution, which was then centrifuged at 4000 rpm for 20 min to separate the biosurfactant-containing precipitate from the supernatant. The supernatant was dumped out, and the residue was resuspended. Using NaOH, the pH of the resuspended material was adjusted to 7. After pipetting the suspension into preweighed vials, it was lyophilised (Alpha 2-4, Martin Christ, Germany). We can determine the dry mass of the sample by calculating the difference between these two values, which provides us with the mass of the biosurfactant left after lyophilization [26][27][28].

Antifungal Assay
Biosurfactant's antifungal activities were in vitro investigated utilising the poisoned food technique. Mixing 39 g of potato dextrose agar (PDA) powder with 1 litre of distilled water was followed by boiling, mixing, and autoclaving for 15 min until homogenous. After semi-cooling, various concentrations (0, 5, 10, 20, and 30%) of cell-free fermentation broth supernatant (1 mL/plate) containing biosurfactant were gently mixed with 25 mL PDA and put in a sterile petri dish before being further cooled. After solidification, a sterilised cork borer was transferred to a 6 mm agar disk from A. alternata culture grown for 7 days onto each culture plate. The cultures were incubated for an additional week at 25 • C. The diameter of the fungal colony was determined, and the inhibition percentage of the mycelial growth of the test fungus by the biosurfactant was calculated using the following formula [29].
Inhibition of mycelial growth (%) = dc − dt dc × 100 (1) where Microorganisms 2023, 11, 707 4 of 14 dc represents the diameter of the fungal colony in the control group (without biosurfactant), and dt represents the diameter of the fungal colony in the treatment group (with the biosurfactant).
2.6. Planting, Transplanting, Growing, and Treating Plants The planting of pinto beans (Phaseolus vulgaris) in 48 planting trays was performed with BIORG MIX potting soil. The soil was regularly watered and then covered with a transparent covering to minimise moisture loss. After one week, the seedlings were transferred into pots in triplicate, resulting in 10 groups of 3 plants each (30 plants in total). The plants were stored in an ozone-sterilised (UV-C lamp 2.5 W/2 m 2 , Anco Electrics Ltd.,Budapest, Hungary) 900 × 900 × 780 mm container. The plants were exposed to 3700 lux of light for 14 h daily at 25 degrees Celsius. The vegetation was watered three times each week. Using a spray bottle, infections and treatments were administered outside the box. The box was divided into two compartments by a transparent plexi sheet, one for control (not infected) and the other for fungal-infected plants. Plants without fungal infection consisted of nontreated (control) and bacterial broth treated ones (BS-0, GS-0, BL-0). A. alternata spores infected plants were divided into two groups, bacterial broth treatment before fungal infections (BS-A, GS-A, BL-A) and treatment after infection (A-BS, A-GS, A-BL) as indicated in Table 1. Thus, we could test the efficacy of both preventive (pre-infection treatment) and therapy (post-infection treatment) applications. Bacterial broth for treatment was made by fermenting bacilli for four days, centrifuging them, and the supernatant was sprayed onto plant leaves.

Disease Incidence and Severity
The disease severity index (DSI%) formula is a mathematical calculation that measures the severity of a disease in a particular population or sample. It is often used in plant pathology. The experiment was stopped at week eight after transplanting from the nursery. The incidence and severity of Alternaria infection on the leaves of the plants were determined according to the following classification shown in Table 2 [30]. Leaves with a value other than 0 were considered infected; the number of infected leaves divided by the total number of leaves was calculated as the incidence: Disease incidence (%) = number of infected leaves (−) total number of leaves on plant (−) × 100 (2) The severity of the infection was calculated using the following formula: Disease severy index (DSI%) : where n 0 -n 5 is the number of leaves in corresponding categories.

Chlorophyll Concentration of Leaves
The chlorophyll content of leaves was measured using a spectrophotometer. Each plant had a leaf of comparable size cut, weighed on an analytical scale, and then crushed with acetone to extract pigments. The pigments were then separated from the solvent by centrifuging the acetone solution in 15 mL Falcon tubes at 4000 rpm for 10 min. At two distinct wavelengths (=663, 645 nm) [31], the absorbance of the supernatant was determined using a spectrophotometer. Three repetitions of the measurement were performed using glass cuvettes.
Using the following formulas, the chlorophyll a and b concentrations in the leaf were calculated: total chlorophyll ( mg /g tissue) = chlorophyll a + chlorophyll b (6) where A663 is the absorbance of the solution measured at λ = 663 nm (-), A645 is the absorbance of the solution measured at λ = 645 nm (-), V is the volume of the solution (mL), and W is the weight of the leaf (g).

Data Analysis
In this study, the means of the three treatments were compared using the Tukey test on all measurements. The statistical analysis was conducted utilising the IBM SPSS Statistics 26.0 software package (IBM Corp., Chicago, IL, USA) for data evaluation and analysis. To determine statistical significance, the significance level for all statistical tests was set to p < 0.05. Figure 1 introduces the time course of the fermentation. The usefulness of the in-line sensor is evident since the significant cell growth occurred in samplings gaps, i.e., during the night. All bacteria reached an OD600 value between 2.5-3, while 1.1-1.3 g/L could be reached regarding product amounts. B. licheniformis was found as most productive since it reached a plateau in 10 h compared to the other two bacteria around 20 h. The time courses also demonstrate the correlations of permittivity-based VCD measurements to off-line determined OD600 values and conductivity to pH values. reached regarding product amounts. B. licheniformis was found as most productive since it reached a plateau in 10 h compared to the other two bacteria around 20 h. The time courses also demonstrate the correlations of permittivity-based VCD measurements to off-line determined OD600 values and conductivity to pH values. courses also demonstrate the correlations of permittivity-based VCD measurements to off-line determined OD600 values and conductivity to pH values. Verification of In-Line Sensor Usefulness

Fermentation
Correlations are more exactly presented in Figure 2, where VCD is plotted versus OD600, resulting in correlation coefficients between 0.6 and 0.92, where the lowest value belongs to B. subtilis. The lack of off-line determined middle-range data for those bacteria  Correlations are more exactly presented in Figure 2, where V OD600, resulting in correlation coefficients between 0.6 and 0.92, w belongs to B. subtilis. The lack of off-line determined middle-range because of the night period causes a bit higher uncertainty besides tion coefficient. Correlations are more exactly presented in Figure 2, where VCD i OD600, resulting in correlation coefficients between 0.6 and 0.92, where belongs to B. subtilis. The lack of off-line determined middle-range data f because of the night period causes a bit higher uncertainty besides an ac tion coefficient. Verification of In-Line Sensor Usefulness Correlations are more exactly presented in Figure 2, where VCD is plotted versus OD600, resulting in correlation coefficients between 0.6 and 0.92, where the lowest value belongs to B. subtilis. The lack of off-line determined middle-range data for those bacteria because of the night period causes a bit higher uncertainty besides an acceptable correla- Correlations are more exactly presented in Figure 2, where VCD is plotted versus OD600, resulting in correlation coefficients between 0.6 and 0.92, where the lowest value belongs to B. subtilis. The lack of off-line determined middle-range data for those bacteria because of the night period causes a bit higher uncertainty besides an acceptable correla- Verification of In-Line Sensor Usefulness Correlations are more exactly presented in Figure 2, where VCD is plotted versus OD600, resulting in correlation coefficients between 0.6 and 0.92, where the lowest value belongs to B. subtilis. The lack of off-line determined middle-range data for those bacteria because of the night period causes a bit higher uncertainty besides an acceptable correla-VCD (=10× permittivity [mS/cm] + 1.8), ♦ OD600 and product. Correlations are more exactly presented in Figure 2, where VCD is plotted versus OD600, resulting in correlation coefficients between 0.6 and 0.92, where the lowest value belongs to B. subtilis. The lack of off-line determined middle-range data for those bacteria because of the night period causes a bit higher uncertainty besides an acceptable correlation coefficient.

Verification of In-Line Sensor Usefulness
Correlations are more exactly presented in Figure 2, where VCD is plotted versus OD600, resulting in correlation coefficients between 0.6 and 0.92, where the lowest value belongs to B. subtilis. The lack of off-line determined middle-range data for those bacteria because of the night period causes a bit higher uncertainty besides an acceptable correlation coefficient.   2 depicts that the correlation between VCD and OD600 has a substantial positive link, especially for B. licheniformis and G. stearothermophilus, with an R 2 of 0.9262 and 0.9202, respectively. This indicated that as permittivity rises, so does OD600.
In Figure 3, the correlation between conductivity and pH with an R 2 value of 0.9043, 0.889, and 0.6363 is observed for B. licheniformis, B. subtilis, and G. stearothermophilus, respectively. The weak correlation for the latter is probably because of an observed time shift between pH and conductivity recording.
Microorganisms 2023, 11, x FOR PEER REVIEW 8 of 15 Figure 2 depicts that the correlation between VCD and OD600 has a substantial positive link, especially for B. licheniformis and G. stearothermophilus, with an R 2 of 0.9262 and 0.9202, respectively. This indicated that as permittivity rises, so does OD600.
In Figure 3, the correlation between conductivity and pH with an R 2 value of 0.9043, 0.889, and 0.6363 is observed for B. licheniformis, B. subtilis, and G. stearothermophilus, respectively. The weak correlation for the latter is probably because of an observed time shift between pH and conductivity recording. To summarise the usefulness of the in-line sensor, one can conclude that out of the tested 6 correlations, 4 were close to or over 0.9, indicating a strong correlation between both permittivity-based VCD and optical density as well as pH and conductivity. The outstanding of G. stearothermophilus can be because of the different behaviour of its biosurfactant. It was reported that this strain produces a polymer-type biosurfactant instead, while To summarise the usefulness of the in-line sensor, one can conclude that out of the tested 6 correlations, 4 were close to or over 0.9, indicating a strong correlation between both permittivity-based VCD and optical density as well as pH and conductivity. The outstanding of G. stearothermophilus can be because of the different behaviour of its biosurfactant. It was reported that this strain produces a polymer-type biosurfactant instead, while the other two produce lipopeptide type biosurfactants [31].

Disease Incidence and Severity
Fungal disease occurrence (Figure 4.) was significantly lower in the case of all three tested bacterial supernatants if applied before fungal infection. Post-infection treatments are inefficient since disease occurrences were similar to untreated but infected controls.

Disease Incidence and Severity
Fungal disease occurrence (Figure 4.) was significantly lower in the case of all three tested bacterial supernatants if applied before fungal infection. Post-infection treatments are inefficient since disease occurrences were similar to untreated but infected controls. However, Figure 5 depicts the impact of the three tested biosurfactants on the severity of the fungal infections, which is remarkably lower in the case of B. subtilis and B. licheniformis but is higher for G. stearothermophilus. At the same time, pre-infection treatment was observed to be more efficient than post-infection treatment. These indicate that emulsifying biosurfactants (polymeric type) are less effective against fungal infection than surface-tension-reducing (lipopeptide type).  However, Figure 5 depicts the impact of the three tested biosurfactants on the severity of the fungal infections, which is remarkably lower in the case of B. subtilis and B. licheniformis but is higher for G. stearothermophilus. At the same time, pre-infection treatment was observed to be more efficient than post-infection treatment. These indicate that emulsifying biosurfactants (polymeric type) are less effective against fungal infection than surface-tension-reducing (lipopeptide type).

Chlorophyll Concentration
The decrement in chlorophyll contents (Figure 6.) indicates a decrease in the photosynthetic activities of the tested plants. The nontreated and non-infected control group contained one of the highest concentrations of chlorophyll a and b, suggesting that its photosynthetic activity is the highest. Similar chlorophyll contents were found only for bacilli-treated groups (BS-0, GS-0, BL-0), which indicates that these bacterial treatments are not harmful to bean plants. The statistical analysis reveals significant differences in chlorophyll levels across the groups, as shown by the values of Pr > F. (all less than 0.001). The lowest chlorophyll contents could be obtained in the post-infection treatment group, which confirms our above-described findings, i.e., pre-infection bacterial supernatant treatment is more efficient, resulting in 90% chlorophyll content of uninfected plants.

Chlorophyll Concentration
The decrement in chlorophyll contents (Figure 6.) indicates a decrease in the photosynthetic activities of the tested plants. The nontreated and non-infected control group contained one of the highest concentrations of chlorophyll a and b, suggesting that its photosynthetic activity is the highest. Similar chlorophyll contents were found only for bacilli-treated groups (BS-0, GS-0, BL-0), which indicates that these bacterial treatments are not harmful to bean plants. The statistical analysis reveals significant differences in chlorophyll levels across the groups, as shown by the values of Pr > F. (all less than 0.001). The lowest chlorophyll contents could be obtained in the post-infection treatment group, which confirms our above-described findings, i.e., pre-infection bacterial supernatant treatment is more efficient, resulting in 90% chlorophyll content of uninfected plants.

Chlorophyll Concentration
The decrement in chlorophyll contents (Figure 6.) indicates a decrease in the photosynthetic activities of the tested plants. The nontreated and non-infected control group contained one of the highest concentrations of chlorophyll a and b, suggesting that its photosynthetic activity is the highest. Similar chlorophyll contents were found only for bacilli-treated groups (BS-0, GS-0, BL-0), which indicates that these bacterial treatments are not harmful to bean plants. The statistical analysis reveals significant differences in chlorophyll levels across the groups, as shown by the values of Pr > F. (all less than 0.001). The lowest chlorophyll contents could be obtained in the post-infection treatment group, which confirms our above-described findings, i.e., pre-infection bacterial supernatant treatment is more efficient, resulting in 90% chlorophyll content of uninfected plants.  Figure 7 demonstrates that changes in total plant height, root length, and shoot length are not significant; thus, this method is not sensitive enough to indicate the fungal disease and its bacterial treatment. The ability of A. alternata to display both endophytic and pathogenic activity exemplifies this fungus' adaptability and its ability to interact with a diverse array of host plants. This diversity may be attributable, in part, to the wide range of metabolites generated by A. alternata and the taxonomic group to which it belongs, which might support a variety of host interactions and nutritional strategies [32,33]. Somewhat higher lengths were obtained for B. subtilis treated but not infected plants. These might indicate that B. subtilis-based antifungal treatments may positively affect plant height, chlorophyll contents, disease severity and occurrences, and inhibition of fungal growth in vitro. While the Pr > F values are more than 0.05, statistical analysis reveals that none of the variations in plant height between the groups was statistically significant.  Figure 7 demonstrates that changes in total plant height, root length, and shoot length are not significant; thus, this method is not sensitive enough to indicate the fungal disease and its bacterial treatment. The ability of A. alternata to display both endophytic and pathogenic activity exemplifies this fungus' adaptability and its ability to interact with a diverse array of host plants. This diversity may be attributable, in part, to the wide range of metabolites generated by A. alternata and the taxonomic group to which it belongs, which might support a variety of host interactions and nutritional strategies [32,33]. Somewhat higher lengths were obtained for B. subtilis treated but not infected plants. These might indicate that B. subtilis-based antifungal treatments may positively affect plant height, chlorophyll contents, disease severity and occurrences, and inhibition of fungal growth in vitro. While the Pr > F values are more than 0.05, statistical analysis reveals that none of the variations in plant height between the groups was statistically significant.

Discussions
The study explored the fermentation of biosurfactants by B. licheniformis and discovered a good correlation between the conductivity and primary fermentation products and between permittivity and the optical density of the fermentation broth. Such in-line realtime and continuous measurements support a better understanding of biosurfactant formation and reach higher productivity via cutting fermentations in optimal time [34,35]. The comparison of the tested three potential bio-fungicide-producing bacteria from fermentation revealed that the fastest cell growth could be reached by B. licheniformis with the highest biosurfactant amount resulting in the highest productivity. Compared to B. subtilis and G. stearothermophilus, B. licheniformis generated a larger output of biosurfactant throughout the fermentation process, with 15% and 5.2% more crude product, respectively. Compared to earlier research, which has reported yields ranging from 0.5 g/L to

Discussion
The study explored the fermentation of biosurfactants by B. licheniformis and discovered a good correlation between the conductivity and primary fermentation products and between permittivity and the optical density of the fermentation broth. Such in-line real-time and continuous measurements support a better understanding of biosurfactant formation and reach higher productivity via cutting fermentations in optimal time [34,35]. The comparison of the tested three potential bio-fungicide-producing bacteria from fermentation revealed that the fastest cell growth could be reached by B. licheniformis with the highest biosurfactant amount resulting in the highest productivity. Compared to B. subtilis and G. stearothermophilus, B. licheniformis generated a larger output of biosurfactant throughout the fermentation process, with 15% and 5.2% more crude product, respectively. Compared to earlier research, which has reported yields ranging from 0.5 g/L to 1.8 g/L [3,36,37], the yield of the present study is relatively high. Our results indicate that our biosurfactant production method effectively achieved a high crude product yield.
The investigation into the potential of three bacterial strains to inhibiting the plant pathogenic fungus of A. alternata indicated that all of them were efficient under in vitro test conditions. However, in vivo plant experiments showed that the G. stearothermophylus broth supernatant was less successful in reducing disease incidence and severity. In contrast, the plants' height and chlorophyll content were similar to the other bacterial treatments. On the other hand, it was discovered that B. subtilis and B. licheniformis were equally successful in treating the plants regarding severity and incidence. It is hypothesised that variations in bacterial biosurfactants may account for the different degrees of efficacy found in in vivo experiments. Including lipopeptide-type biosurfactants, such as lichenysin by B. licheniformis and surfactin by B. subtilis, in the broth resulted in comparable efficacy; however, polymeric-type emulsifying biosurfactants in the supernatant of G. stearothermophilus [31] resulted in a decreased efficacy.
The biosurfactant-producing bacteria B. subtilis, G. stearothermophilus, and B. licheniformis reduced the severity of A. alternata disease by up to 30%, 5%, and 25%, respectively, when applied to plants (i.e., in vivo). Additionally, the chlorophyll a and b concentrations of the bacterial supernatant-treated plants ranged from moderate to high, suggesting enhanced photosynthetic activity.
Overall, the research indicates that B. licheniformis is the most efficient bacterium for biosurfactant generation and biosurfactant treatment of plants. This work adds to the expanding body of knowledge on the use of biosurfactants for sustainable agriculture and indicates the potential for more research and development in this area.

Conclusions
The correlation coefficients between conductivity and pH and the association between permittivity and OD600 imply that these variables might be utilised to monitor biosurfactant production on a wide scale.
In conclusion, we can observe that lipopeptide-type biosurfactants are more effective bio fungicides than polymeric-type emulsifiers. The reason can be that air-mycelia and conidiospores (having somewhat hydrophobic behaviour) of plant pathogenic filamentous fungi can be better attacked if surface tension is controlled by biosurfactant-based bio fungicide in comparison to emulsification. The tested biosurfactant-producing bacilli, namely B. licheniformis and B. subtilis, are frequently used in biotechnology for different purposes, where the bacterial biomass is a discarded by-product. Our study highlighted the application of these cells as bio-fungicide, which can be a sustainable solution in the future.
Additionally, regarding in vivo plant treatments, increasing the sample size and replicating could help reduce the variability in the results and increase the study's statistical power. Improved disease severity analyses using digital imagery and the connection between biosurfactant treatment and plant yield can be an interesting continuation.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.