Isolation and production of newly oligosporogenic polyvalent Bacillus thuringiensis mutants for agricultural pest control

Abstract

Background

To develop a new generation of eco-friendly biopesticides, this study used an optimized random mutagenesis approach to generates oligosporogenic mutants from B. thuringiensis var. kurstaki BLB1. Although random mutagenesis is not a novel approach, it remains a practical and efficient strategy providing a simple and accessible alternative.

Results

Screening of over than 600 mutants showed that only 14 clones harbour less rate of spores ranging from 1 ± 0.5 10^–6 to 126 ± 1 10^–6 spores ml⁻¹. Among them, a number of 6 mutants showed an improvement of delta-endotoxins until 222.92%. Protease production showed that S6 mutant exhibited the most increased rate of about 4136.36 U ml⁻¹ ± 89.65 in addition to its enhancement of 34,96% in delta-endotoxins production. Captively, S3 and S4 mutants were largely active against A. tumefaciens and three fungi revealing values of MIC and MFC ranging from 0.07 to 0.31 µg ml⁻¹ and from 0.05 to 0.225 µg ml⁻¹ against A. niger, F. culmorum and B. cinerea, respectively. Oligosporogenic S6 mutant was considered as a polyvalent and eco-friendly biopesticide since it is the most toxic against semolina larvae E. kuehniella (LC₅₀ of 39.52 µg g⁻¹). To reduce its production cost, we explored the use of date waste as a cost-effective culture medium. An improvement of 43.15% of delta-endotoxins production of S6 mutant was obtained via Taguchi design while valorising agricultural by-products.

Conclusion

The study successfully generated oligosporogenic mutants of B. thuringiensis BLB1 with enhanced delta-endotoxin and enzyme production while reducing spore formation. The S6 mutant showed the highest bioactivity, and its production was further increased in a cost-effective way using date waste as a culture medium, highlighting its potential as a useful and eco-friendly biopesticide.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12866-026-04728-y.

Introduction

Insect damage and plant diseases cause increased losses in yield and quality of growing crops. The control of phytopathogens and insect pests by using chemical pesticides leads to crucial harmfulness to health and environment. In fact, chemical residues remain frequently on food products causing numerous health problems like cancers, hormone disruption and neurological problems [1]. Besides, they are harmful to non-target organisms and contribute to the development of resistant pests, making future pest control more difficult and less effective. Moreover, their accumulation in the environment disrupts biodiversity and can lead to the change of aquatic life by water pollution [2, 3]. To address these environmental concerns many farmers and scientists are turning to biopesticides. These natural alternatives are derived from plants, bacteria and other organic sources.

Up to now, Bacillus thuringiensis (B. thuringiensis) is the most widely commercialized biocontrol agent worldwide. Its benefit as a biological control agent extends beyond its eco-friendly profile [4]. In fact, B. thuringiensis is a safe alternative to conventional chemical pesticides, promoting a healthier and more sustainable agricultural ecosystem [5] as well as its minimized collateral damage to non-target organisms [6]. B. thuringiensis is a Gram-positive soil bacterium characterized by the formation of spores and parasporal crystals during the sporulation phase [7]. Insecticidal activity of B. thuringiensis towards wide range of insect orders is based on the production of delta-endotoxins proteins accumulated as a crystalline parasporal inclusions such as Lepidoptera, Coleoptera and Diptera [8, 9]. In addition, B. thuringiensis produces other insecticidal factors such as Cyt proteins, vegetative insecticidal proteins (Vip), secreted proteins (Sip), and exotoxins, including thuringiensin (β-exotoxin), which can contribute to toxicity and broaden the insecticidal spectrum [10, 11]. In addition to δ-endotoxin production, B. thuringiensis produces hydrolytic enzymes such as chitinases, β-glucosidases, and proteases, which contribute to the elimination of phytopathogenic fungi, as well as insect pests and bacteria, by degrading invertebrate exoskeletons and fungal cell walls, thereby disrupting cell integrity [12]. Otherwise, B. thuringiensis spores are known to be ubiquitous in the environment due to their resistance against extreme conditions, including radiation, heat, toxic chemicals, etc. [13]. Previous studies have reported that spores can persist on the leaf surface for more than 72 h after application [14]. These traits present environmental concerns for the biopesticide industry, because the release of large quantities of B. thuringiensis spores can alter soil microbial communities and disrupt the balance of native bacterial populations. Recent studies have shown that B. thuringiensis can persist in insect hosts in a non-sporulated vegetative state, and that these cells may be more resistant than those produced under standard laboratory conditions [15]. This indicates that sporulation is not the only persistence strategy in natural environments. However, such naturally occurring non-sporulating cells are often lack the stability, reproducibility, and consistent biological activity required for agricultural applications. Therefore, random mutagenesis represents a practical strategy to generate stable oligosporogenic strains that retain high biological activity while limiting environmental spore dissemination, offering advantages over naturally occurring phenotypes [15, 16]. In fact, strains with delayed sporulation are important for industrial biopesticide production because they allow prolonged vegetative growth, increasing biomass and toxin yield while reducing production costs [17, 18]. Oligosporogenic and asporogenic B. thuringiensis strains, which form fewer spores, also limit environmental spore dissemination and tend to disappear more quickly from the environment compared to fully sporulating strains, reducing ecological risks and long-term contamination [16, 19, 20]. To address these challenges, random mutagenesis has been used to create mutants with reduced sporulation that still express insecticidal crystals, balancing effective pest control with lower environmental persistence and enhancing both industrial practicality and ecological safety [16, 18]. To get around this problem, creating mutants defective in sporulation and still expressing toxin crystals was investigated by means of random mutagenesis. In fact, asporogenic and oligosporogenic mutants of B. thuringiensis strain can be carried out by treating the cell culture with a variety of both physical and chemical mutagenic agents that usually induce single-nucleotide changes or small deletions in the genome that might encompass multiple genes or regulatory sequences [21]. Almost mutagenic agents act on the genome differently and affect numerous pathways of the cell machinery [22]. Recent advances in molecular biology, particularly targeted genome-editing approaches such as CRISPR systems, gene cloning, and other molecular techniques, have greatly facilitated the study of gene function by enabling precise and efficient genetic modifications [23, 24]. Nevertheless, the application of these approaches often requires advanced technical expertise, detailed genomic information, and specialized infrastructure, which may limit their use in certain experimental contexts [16, 19]. Although random mutagenesis is not a novel approach, it remains a practical and efficient strategy, offering a simple and accessible alternative to advanced molecular biology tools such as CRISPR-based systems and gene cloning, which frequently demand sophisticated infrastructure and precise genomic information [16, 23, 24].

Several studies have sought to establish relationships between the production of hydrolytic enzymes, sporulation, and δ-endotoxin synthesis in Bacillus thuringiensis. These investigations aim to elucidate the interactions among these biological processes, thereby providing insight into the regulatory mechanisms underlying the biological activity of B. thuringiensis [25]. The objective of the present study was to generate oligosporogenic mutants of B. thuringiensis strain BLB1 through optimized random mutagenesis of the wild-type strain. This modification was characterized by enhanced δ-endotoxin and lytic enzyme production, along with reduced sporulation to minimize ecological impact, followed by evaluation of their efficacy against phytopathogenic microorganisms and insect larvae, as well as assessment of their production in a low-cost medium.

Materials and methods

Bacterial strains and growth conditions

B. thuringiensis strain BLB1 isolated from Tunisian soil sample was used as a reference strain during this study [17]. A number of 14 oligosporogenic mutants of B. thuringiensis were isolated through BLB1 mutagenesis using 3 mutant agents.

Luria–Bertani (LB) culture medium was used for growth of B. thuringiensis strains and solid T3 medium was used to produce the crystal inclusion bodies during the sporulation growth phase [26]. Additionally, Glucose-Based Medium described by [6] was used in order to produce the parasporal crystals during growth; flasks containing 50 ml of this culture medium were incubated 72 h at 30 °C and 200 rpm in a rotary shaker.

Six fungal strains including Aspergillus niger (A. niger) (CTM1099), Alternaria alternata (A. alternata) (CTM10230), Fusarium culmorum (F. culmorum) (ISPAVE21w), Fusarium graminarium (F. graminarium) (ISPAVE271), Fusarium oxysporum (F. oxysporum) (CTM10402), and Botrytis cinerea (B. cinerea) (food isolate) were used to evaluate the antifungal activity of the wild strain BLB1 and the obtained mutants. All fungi were grown on Potato Dextrose Agar (PDA) for 5–7 days at 25 °C. Spore number was determined using a Thomas counter [27]. The phytopathogenic bacteria A. tumefaciens strain C58 was used in this study. It was routinely grown in Luria–Bertani (LB) agar at 30 °C and maintained at 4 °C before use [28].

Optimization of random mutagenesis using Taguchi method

The Taguchi experimental design was applied to optimize mutagenesis conditions using three mutagenic agents. The tested parameters included nitrous acid concentration (30 mg ml⁻¹, 70 mg ml⁻¹), its incubation time (30 min, 60 min), acridine orange concentration (15 µg ml⁻¹, 30 µg ml⁻¹), its incubation time (15 min, 60 min), and UV exposure time (2 min, 45 min). Each factor was tested at two levels to assess its impact on cell viability and maximize arbitrary DNA damage. Statistical analysis, including the signal-to-noise ratio, was conducted to identify the optimal conditions for minimizing cell viability. All cultures were carried out in 1000 ml Erlenmeyer flasks. Each flask contains 50 ml of liquid Glucose-Based Medium. After sterilization, the flasks were inoculated and maintained under different conditions using various levels of the chosen factors of the model [29]. ANOVA was used to determine the significance of factors in the Taguchi design.

Induction of randomly mutagenesis

An isolated colony of the wild strain BLB1 was dispensed into 3 ml of LB medium and incubated overnight at 30 °C with an agitation of 200 rpm in a rotary shaker. A volume of this culture was diluted into 250 ml flasks containing 50 ml of LB medium with O.D.₆₀₀ adjusted at 0.15 and incubated at 30 °C until the O.D.₆₀₀ reached almost 0.5 after nearly 4 h of incubation [26].

For treatment of vegetative cells with nitrous acid, a volume of 500 µl of the culture was diluted with LB medium and centrifuged for 15 min at 12,000 rpm. Then, the pellet was suspended in 250 µl phosphate buffer (100 mM; pH 7.0) and added to 50 µl of nitrous acid at a final concentration of 30 mg ml⁻¹ and 70 mg ml⁻¹. After incubation at 30 °C with shaking at 200 rpm in a rotary shaker for 30 and 60 min, the recuperated cells were washed twice with 1.950 ml µl of physiological water 0,9% and 1.950 ml of sterile bidistilled water and resuspended in 2 ml of LB medium. Serial dilutions from 10^–1 to 10^–6 of the treated cells were transferred on T3 medium and incubated 72 h at 30 °C. After 24 h, colony forming units (CFU) of surviving mutants were counted in three replicates [30].

For treatment of vegetative cells with acridine orange, the mutagenesis was performed by acridine orange treatment of the obtained pellets described above. Acridine Orange was filtered and added into pellets at the concentrations of 15 µg ml⁻¹ and 30 µg ml⁻¹. The reaction mixture was incubated for 15 min and 60 min at 30 °C. Consecutive dilutions were used as described above in order to count the survival mutants [31].

For treatment of vegetative cells with UV rays, the mutagenesis was done with short wavelength UV irradiation using a Philips tube (15 W). The cell suspension (O.D.₆₀₀ of 0.1) was distributed in Petri plates (10–15 ml in each plate) and exposed to UV radiation for variable periods (2 min, 45 min) while keeping the distance of the UV source fixed at 20 cm. Serial dilutions were used as described above in order to count the survival mutants [32].

Biomass and delta-endotoxin determination

Spore and viable cell counts were determined from post-logarithmic phase cultures obtained after mutagenesis. Aliquots of each culture were serially diluted in sterile saline solution. Total viable cells were enumerated by plating appropriate dilutions on LB agar plates and incubating at 30 °C for 24 h. To specifically quantify spores, parallel aliquots were subjected to heat treatment at 80 °C for 10 min to eliminate vegetative cells. Heat-treated samples were then serially diluted, plated onto LB agar, and incubated at 30 °C for 24 h.

Delta-endotoxin concentration was determined from solubilized crystal protein preparations obtained from each culture medium, following the method described by Jamoussi et al. [33] with minor modifications. Briefly, after incubation, 1 mL of bacterial culture was centrifuged at 13,000 rpm for 10 min. The resulting crystal-spore pellet was washed twice with sterile 0.85% NaCl solution and once with distilled water to remove residual medium components. The washed pellets were then resuspended in 1 mL of 50 mM NaOH (pH 12.5) and incubated for 2 h at 30 °C to allow alkaline solubilization of the crystal proteins. Insoluble material was removed by centrifugation at 13,000 rpm for 15 min. The supernatant containing the solubilized insecticidal crystal proteins was collected and used for delta-endotoxin quantification. Protein concentration was determined using the Bradford assay, with bovine serum albumin (BSA) as standard. All the results related to determination of delta-endotoxin production and CFU counts were the average of three determinations of two separate experiments for each cultural condition performed using Microsoft Excel Software 2016.

Determination of hydrolytic enzymes activities

Proteolytic activity was assayed by incubating 1 ml of the diluted enzyme solution with 1 ml of casein substrate (1%) which is prepared in Tris–HCl buffer 100 mmol L⁻¹ at pH 7 according to Kembhavi [34]. The reaction mixture was incubated for 20 min at 60 °C and then the reaction was stopped with 3 ml of 5% trichloroacetic acid. Parallel blanks were prepared with an inactivated enzyme. At the end of the incubation period, samples and blanks were centrifuged at 12,000 rpm for 15 min. The absorbance of the supernatant was measured at 280 nm. One proteolytic activity unit was defined as the amount of enzyme required to liberate 1 µg of tyrosine from casein per minute, under the experimental conditions. The values presented are the average of three replicates of two separate experiments for each cultural condition [35]. β-glucosidase activity was determined by measuring the release of p-nitrophenol (p-NP) from the p-NPG substrate para-nitrophenyl-β-D-glucopyranoside (ACROS/ORGANICS). For the assay, an appropriate volume of culture supernatant from the mutants, diluted in 100 mM sodium acetate buffer, was mixed with 200 µl of 1 mM p-NPG solution and incubated for 15 min. Control reactions included all components except the enzyme preparation. The enzymatic reaction was terminated by adding 600 µl of 400 mM Glycine–NaOH buffer, pH 10.8, and the absorbance was measured at 400 nm. One unit of β-glucosidase activity was defined as the amount of enzyme that releases 1 µmol of pNP per minute under the assay conditions [36]. The chitinase activity was monitored according to the method of Barboza-Corona [37]. A single colony from B. thuringiensis mutants was inoculated by sticking solid medium supplemented with 2 g L⁻¹ casaminoacids and 0.6% colloidal chitin (w/v). Plates were incubated at 30 °C for 2 days and then transferred to 37 °C for a week until the appearance of clearing zones.

Antagonistic assays and evaluation of minimal inhibitory concentration (MIC) and minimal fungicidal concentration (MFC)

The 14 oligosporogenic B. thuringiensis mutants obtained through random mutagenesis were tested for their antifungal and antibacterial activities. Firstly, fungal spores or bacteria suspensions were plated onto medium surface and colonies of oligosporogenic mutants were placed on the surface of media. In addition, supernatants were also recovered by centrifugation for 10 min at 17,500 g and poured in wells obtained by punching solid medium. PDA and LB solid media were incubated at 25 °C and 30 °C for testing antifungal and antibacterial activities, respectively. Antibiosis activities were estimated by evaluation of the inhibition zones surrounding the colonies after 2–3 days of incubation for bacteria and 4–5 days for fungi. Amphotericin B and gentamicin were used as positive controls.

Using 96-well microplates, the MICs and MFCs of S3 and S4 B. thuringiensis oligosporogenic mutants were determined as described by Ellof [38] against A. niger, F. culmorum and B. cinerea. The culture was centrifuged for 10 min at 11,000 rpm and the corresponding supernatant was lyophilized. Briefly, the principle of this method is to perform in the microplate wells a serial dilution of the supernatant ranging from 700 µg ml⁻¹–2.5 µg ml⁻¹ and 500 µg ml⁻¹–1.9 µg ml⁻¹ for S3 and for S4, respectively. In fact, each microplate wells were filled with a final volume of 100 μl formed by 80 μl PDB (Potatoes dextrose broth), 10 μl of a dry supernatant dissolved in distilled water and filtered, and 10 μl of fungus spores’ solution (10⁵ spores ml⁻¹). Negative controls consisting of 90 μl PDB and 10 μl of distilled water. The plates were incubated at 25 °C for 3 days. The MIC was defined as the minimum concentration of the sample which generates the inhibition of any visible growth of the microorganism after incubation. Concerning MFC evaluation, 10 μl from each well that does not show any fungus proliferation was taken, plated on PDA and incubated at 25 °C for 3 days. The lowest concentration that yielded no growth was taken as the MFC, indicating that > 99.9% of the original inoculums were killed. Three replicates were maintained to confirm the antifungal activity.

Bioassays against E. kuhienella

Bioassays were carried out using the second instar E. kuehniella larvae. Ten E. kuehniella larvae were transferred to a sterile Petri dish containing 0.5 g flour mixed with different concentrations of spore-crystal mixtures from the 14 oligosporogenic mutants selected and incubated at room temperature (27° C). Ten insect larvae were fed with water mixed with flour used as a negative control. The wild strain BLB1 was used as a positive control. Assays were tested in three replicates for each of three experimental runs. Fifty percent lethal concentration values (LC₅₀) for each experiment were calculated from pooled raw data by probit analysis using programs written in the R. language [39]. For each strain, five concentrations (20, 60, 100, 140, 180 µg/g) were tested using 10 larvae per concentration. The relationship between mortality and concentration was linearized by log-transformation of the doses. Statistical parameters, including the 95% confidence intervals (CI) and the slope of the regression lines, were calculated using SPSS. The goodness-of-fit of the model was assessed using the Pearson chi-square X² test.”

Delta-endotoxins production using date waste

The S6 oligosporogenic mutant was selected for delta-endotoxin production in culture media based on date waste. This biowaste was collected from local farms, air-dried and stored at 4 °C. In preliminary study, effect of date waste concentration on delta-endotoxin production was tested at various concentrations (2.5%, 5%, 7.5%, 10%, 12.5%, 15%, 17.5%, 20%). The fermentation was carried out in 250 mL Erlenmeyer flasks containing 12.5 ml of distilled water at 30 °C with shaking at 200 rpm until complete sporulation. The concentration that produced the highest delta-endotoxin yield was selected for further optimization. Subsequently, fermentation parameters were optimized using a Taguchi experimental design to maximize delta-endotoxin production. The Taguchi design was applied to optimize critical fermentation parameters, including substrate concentration (12.5%, 17.5% and 22.5%), agitation speed (150, 200, and 250 rpm), initial optical density (0.1, 0.15, and 0.2), and aeration ratio, assessed using three Erlenmeyer flask sizes (250 mL, 500 mL, and 1000 mL). Each parameter was tested across three distinct levels to determine the optimal conditions for enhanced fermentation efficiency. Statistical analysis, including signal-to-noise ratio, were performed to determine the optimal conditions for maximum delta-endotoxin production. ANOVA was used to determine the significance of factors in the Taguchi design.

Results and discussion

Screening of B. thuringiensis oligosporogenic mutants

Modern molecular biology tools, including targeted genome-editing approaches such as CRISPR systems, gene cloning, and other molecular techniques, have significantly advanced our ability to manipulate genes and study their functions with high precision [23, 24, 40]. However, their application requires specific expertise, detailed genomic information, and advanced infrastructure, which can limit their use [16, 19]. In contrast, random mutagenesis provides a practical and accessible alternative for generating genetic and phenotypic diversity without prior knowledge of target genes. While it lacks the precision of CRISPR-based approaches, random mutagenesis remains simple, stable, and cost-effective, making it suitable for exploratory and applied studies. In the present work, this strategy enabled the efficient selection of oligosporogenic B. thuringiensis mutants with improved biological activities [16, 18]. To better understand the effects of random mutagenesis with nitrous acid, acridine orange and UV rays on B. thuringiensis metabolic pathways, we investigated their effects on B. thuringiensis cell viability, since the mutant agent treatment efficiency is related to acute cell lethality. For this purpose, five factors were selected to determine their impact on the viable cells production and to determine their optimal levels using L12 Taguchi experimental design. The experimental plan with their experimental responses is presented in Tables 1 and 2.

Table 1.

Definition of the factors and their corresponding levels and contributions studied by Taguchi’s L12 design

Code	Variables	Level 1	Level 2	Contribution (%)	F-value	P-value
X1	Nitrous acid concentration (mg ml−1)	30	70	37	15.21	0.027
X2	Nitrous acid incubation time (min)	30	60	17	8.64	0.042
X3	Acridine orange concentration (µg ml−1)	15	30	20	7.17	0.044
X4	Acridine orange incubation time (min)	15	60	5	0.76	0.569
X5	UV incubation time (min)	2	45	20	9.17	0.031

Level 1: the first answer. Level 2: the second level answer

Table 2.

Taguchi’s L12 experiment plan with their experimental responses

Runs	X1	X2	X3	X4	X5	Log CFU ml−1
1	1	1	1	1	1	4.82 ± 0.62
2	1	1	1	1	1	8.53 ± 1.12
3	1	1	2	2	2	6.83 ± 0.88
4	1	2	1	2	2	4.63 ± 0.54
5	1	2	2	1	2	4.52 ± 0.49
6	1	2	2	2	1	7.23 ± 0.57
7	2	1	2	2	1	4.52 ± 0.64
8	2	1	2	1	2	6.23 ± 0.72
9	2	1	1	2	2	2.52 ± 0.31
10	2	2	2	1	1	4.58 ± 0.59
11	2	2	1	2	1	4.27 ± 0.48
12	2	2	1	1	2	2.82 ± 0.36

According to Taguchi’s signal-to-noise ratio, the nitrous acid concentration had the highest effect at level 1, the incubation time was most significant at level 1, the acridine orange concentration showed the strongest impact at level 2, and its incubation time was more appropriate at level 2 (Fig. 1). Finally, UV exposure had the highest effect at level 1, in fact, the contribution of nitrous acid concentration was 37%, nitrous acid incubation time was 17%, acridine orange concentration was 20%, acridine orange incubation time was 5% and UV exposure time was 20% (Table 1). The analysis of variance (ANOVA) for the mean response to mutagenic agents demonstrate that the proposed model adequately explains the variability observed in the experimental data. The coefficient of determination (R²) is 73.6%, showing that nearly three-quarters of the variability in the mean response is explained by the studied factors. The analysis of variance indicates that nitrous acid concentration, nitrous acid incubation time, acridine orange concentration and UV incubation time are factors with a statistically significant effect on the mean response at the 5% significance level. Nitrous acid concentration had the higher F-value of 15.21, with a corresponding P-value of 0.027, confirming its significant highest contribution to the observed variability. To confirm the robustness of the ANOVA results and the relevance of the identified optimal conditions, a confirmation experiment was subsequently performed using the factor levels predicted by the model. The experiment conducted under these optimized conditions resulted in a marked decrease in cell viability compared with the initial experimental runs. Overall, the confirmation experiment demonstrates that the optimized factor combination effectively enhances viability, supporting the reliability and practical applicability of the ANOVA-based optimization.

Fig. 1.

Main plot effects for signal-to-noise ratios: influence of nitrous acid, UV radiations and acridine orange on cell viability

Under these predicted optimal conditions, BLB1 cell count number was validated experimentally as log CFU ml⁻¹ = 2.67. These conditions were adopted to perform all mutagenesis runs and to analyse the resulted oligosporogenic mutants. Since the mortality rate demonstrates the efficiency of the random mutagenesis process, these results showed that the optimization of this process by Taguchi design was very successful, underscoring the significant impact of the association of these three mutant agents on cell viability. These results validate the optimal conditions selected by Taguchi’s Design.

Table 3 showed that vegetative cells of the wild-type strain BLB1 (299 ± 1.1 10^–6 CFU ml⁻¹) were highly sensitive to the three mutagenic agents since the cell counts of mutant clones ranged from 84 ± 0 10⁶ CFU ml⁻¹ to 282 ± 1.5 10⁶ CFU ml⁻¹ revealing that B. thuringiensis cells continue to multiply despite the alteration in sporulation.

Table 3.

Characterization of B. thuringiensis oligosporogenic mutants derived from BLB1 strain, on the basis of their δ-endotoxin production and proteolytic activity using Glucose-Based Medium

Strains	Total cells (CFU 10–6 ml−1)	Spores (Spores 10–6 ml−1)	Spore Reduction (%)	δ -endotoxins (mg L−1)	δ-endotoxins Improvement (%)	Proteolytic Activities (U ml−1)	Proteases Improvement (%)
BLB1	299 ± 1,1	141 ± 9,5	-	1231,13 ± 70.10	-	3554,54 ± 128.01	-
S1	201 ± 1,5	1 ± 0,50	99	451,36 ± 17.11	-	190,90 ± 8.05	-
S2	119 ± 17	34 ± 0,50	75	3335,66 ± 105.25	170,94	1172,72 ± 99.64	-
S3	173 ± 2,08	126 ± 10	10	2667,69 ± 118.77	116,68	45,45 ± 3.38	-
S4	174 ± 1	69 ± 12,40	51	1233,92 ± 112.10	0,22	710 ± 12.58	-
S5	198 ± 1,5	114 ± 0,50	19	823,08 ± 16.45	-	981,81 ± 30.85	-
S6	268 ± 0,5	10 ± 6,08	92	1661,54 ± 61.14	34,96	4136,36 ± 89.65	16.36
S7	161 ± 1,1	84 ± 0,50	40	2181,38 ± 59.45	77,18	3390,90 ± 136.51	4.60
S8	84 ± 0	30 ± 12,01	78	3975,68 ± 125.29	222,92	1618,18 ± 96.46	-
S9	270 ± 0	78 ± 10	44	1110,95 ± 121.54	-	1145,45 ± 78.16	-
S10	146 ± 0,5	19 ± 5,70	86	2368,64 ± 107.05	92,39	-	-
S11	282 ± 1,5	25 ± 2,50	82	1189,21 ± 106.62	-	790,90 ± 53.33	-
S12	251 ± 1,7	34 ± 8,80	75	663,77 ± 57.64	-	27,27 ± 3.26	-
S13	134 ± 0	13 ± 0,50	90	498,88 ± 43.44	-	-	-
S14	118 ± 1,7	61 ± 0,50	56	915,31 ± 62.61	-	1272,72 ± 151.17	-

 + depicts a standard error of the mean value

Interestingly, analysis of 130 mutant clones showed that only 14 mutants were oligosporogenic with less rate of spores ranging from 1 ± 0.5 10^–6 spores ml⁻¹ to 126 ± 1 10^–6 spores ml⁻¹ comparing to the wild strain BLB1 which produced around 141 ± 9.5 10^–6 spores ml⁻¹. In particular, S1 mutant showed the most reduced spores’ number with 99% percentage of reduction. It has been reported that B. thuringiensis can persist in a non-sporulated form under certain conditions, such as during the late stages of infection in insect hosts [15]. This finding indicate that sporulation is not always the only strategy for persistence in the environment and that naturally occurring non-sporulating or low-sporulating strains can be observed under specific ecological conditions. However, sporulation-deficient cells observed in infection of insect hosts do not always possess the characteristics needed for agricultural use, such as strong biological activity, regular growth and stability under field or industrial conditions. For this reason, we investigated the generation of mutants with reduced sporulation but maintained active molecules expression using random mutagenesis. This approach improves both safety and effectiveness for biopesticides, beyond what natural strains can offer [15, 16]. Moreover, Strains with delayed sporulation are valuable for industrial biopesticide production because they allow prolonged vegetative growth, increasing biomass and toxin yield [16, 18]. Our oligosporogenic B. thuringiensis mutants maintained high viable cell counts while producing fewer spores, consistent with reports that such strains decline more rapidly in soil, reducing environmental persistence and ecological risks [16, 19, 41]. These results suggest that random mutagenesis can generate mutants combining industrial practicality with ecological safety [16, 18]. These results indicate that random mutagenesis can profoundly affect both the metabolic framework and the spore-forming ability of B. thuringiensis. In fact, it was necessary to reduce the dissemination of spores in the environment by the development of a new generation of eco-friendlier biopesticides that minimize spore release while maintaining effectiveness [14]. In this study, we succeeded to obtain oligosporogenic mutants by spontaneous mutation without affecting the capacity of producing viability. In the same purpose, [42] developed a safe sporeless bioinsecticide based on B. thuringiensis mutants which can overproduce insecticidal crystal protein by using only nitrous acid. Additionally, previous studies obtained B. thuringiensis mutants using UV radiation and nitrous acid separately [43]. However, in this study, we tried an advance approach using three mutant agents simultaneously as an innovative strategy that has not previously used for B. thuringiensis. This multi-agent method proved its efficiency to enhance the production of highly effective and safe bioinsecticides by reducing spore production.

Delta-endotoxin production using glucose-based medium

To elucidate the random mutagenesis effect on delta-endotoxin production of the 14 selected mutants, we studied their production using glucose-based medium. This medium has been previously shown to be efficient by detecting minor variations in B. thuringiensis delta-endotoxin production [44]. As shown in Table 3, 6 mutants showed an improvement in delta-endotoxin production compared to the wild strain BLB1, varying from 34,96 to 222.92%. Among of them, the S8 mutant has shown a particular increase in delta-endotoxins production with an improvement of 222,92% and reduced spores’ number. This significant improvement highlights the potential of random mutagenesis to produce more eco-friendlier strains with significantly increased delta-endotoxin production. The difference in delta-endotoxin production between B. thuringiensis oligosporogenic mutants could be explained by the arbitrary nature of the random mutagenesis effect, which varies from mutant to another. In fact, there could be perturbations at different levels of their metabolic pathways as well as in the mechanism of delta-endotoxin synthesis [21]. In fact, nitrous acid has been shown to act on nucleic acids by deaminating the amino groups of adenines and converting the AT base pair to the GC base pair [20, 45, 46]. In addition, UV radiation causes deamination of 5-methylcytosine to thymine, a common spontaneous reaction, resulting in a GC to AT base pair transition [47]. However, acridine orange is an aromatic compound that intercalates between base pairs in DNA, resulting in insertions and deletions of nucleotide bases during replication. This alteration causes a shift in the translated reading frame of the encoded information, resulting in an altered sequence of amino acids [48, 49]. Similar results were reported by Ghribi [21], who demonstrated that this type of classical mutagenesis randomly affected B. thuringiensis cells in the production of their delta-endotoxins, with improved clones and others adversely affected in the production of Cry toxins. On the other hand, it was reported that the overproduction of delta-endotoxins can be explained by the prolonged stationary phase in absence of sporulation, which increase the time for gene transcription, protein synthesis and crystal formation [50].

Proteases production

In this study, the screening of protease production of oligosporogenic mutants revealed different proteolytic activities among the oligosporogenic mutants. These results demonstrated that nitrous acid, orange acridine and UV rays randomly affects BLB1 clones in different genes and with different effects either positively or negatively influencing protease toxin production. Interestingly, S6 mutant exhibited the most increased proteolytic activity of about 4136.36 U ml⁻¹ ± 89.65 comparing to the wild strain BLB1 (3554,54 U ml⁻¹ ± 128.01 ml⁻¹) with an improvement percentage of 16.36% followed by S7 mutant (3390 ± 90 U ml⁻¹ ± 136.51) with an improvement percentage of 4.60% (Table 3).

Based on the improvement of proteases and delta-endotoxins, we classified the mutants to 3 classes. The first group of oligosporogenic mutants including S6 and S7 mutant showed important yields in both delta-endotoxin production and proteolytic activity. The persistence of improved delta-endotoxins yields in spite of the presence of increased proteases amounts can be explained either by the protection of higher quantities of delta-endotoxins which can restrict proteases action or by the abolition of the target proteases sites in delta-endotoxins. These conclusions are also in line with those found by Donovan [51] who demonstrated that the stability and production yield of crystalline proteins were improved by the elimination of extracellular proteases specific to B. thuringiensis and that the amounts of proteases could prolong or reduce proteolysis of the crystal. Besides, the second group include S2, S3, S8 and S10 mutants showing a significant improvement of delta-endotoxins and a weak proteolytic activity even a totally absence in S10 mutant. In this case, proteolytic activities were low and altered which explained that delta-endotoxins did not undergo any proteolysis. These results agree with the results of Ennouri [35]. who found an inverse correlation between the production of proteases and delta-endotoxins. Moreover, a third group including S1, S5, S9, S11, S12, S13 and S14 mutants showed a decrease in delta-endotoxin production coupled with a low proteolytic activity. The S4 mutant was considered among the third group since its improvement in delta-endotoxins is not significantly statistically. In this context, Ben Khedher [42] demonstrated that expression of proteases genes is induced by the sporulation process and given that these mutants are deficient in sporulation, it is expected that their protease activities is very low.

Bioassays against the semolina larvae E. kuehniella

Among the objectives of this research, in the frame of biological control, we tried to obtain eco-friendly toxic oligosporogenic mutants towards the Mediterranean flour moth E. kuehniella Zeller (Lepidoptera: Pyralidae). To this end, a toxicity test was conducted against E. kuehniella in the second larval stage (L2). All values are validated by the results of probit analysis summerized in Table 4. As shown in Fig. 2, S2 and S6 mutants exhibited the lowest LC50 values, indicating their highest toxicities. It is reasonable to expect such increased toxicity levels, given that S2 mutant produced raised yield of delta-endotoxins concentration (3335,66 ± 105.25 mg L⁻¹) and S6 mutant showed the most important protease activity (4136,36 ± 89.65 UI) in addition to the important delta-endotoxins production which is about 34.96%. This enhancement could be also explained by occurrence of different mutations in cry genes sequences that enhanced delta-endotoxins toxicity. In fact, endogenous proteases play a pivotal role in the protoxins activation, which occurs independently from protoxin activation inside insect larvae [52]. These results are consistent with those of Dammak [53] who demonstrated that the combination of both delta-endotoxins and proteases could improve the toxicity against Spodoptera Littoralis larvae. Such an increase in toxicity levels make these mutants optimal candidates as biological agents for the control of this pest. However, all remaining mutants exhibited lower toxicities levels comparing to BLB1 even S3, S4, S7, S8 and S10 mutants harbouring important concentrations of delta-endotoxins. This phenomenon could be explained by the fact that the crystals of these oligosporogenic mutants remained encapsulated in the mother cells since the sporulation process was altered, which made them inaccessible to the digestion and solubilisation by the insect larvae as demonstrated by Ben Khedher [42]. It could be also attributed to the occurrence of randomly mutation affecting the cry genes sequences causing the loss of delta-endotoxins activity in spite of their important amounts. These results highlight the importance of delta-endotoxin production and protease activity in the efficacy of biological agents against pests.

Table 4.

Toxicological parameters of Bacillus thuringiensis strains determined by Probit analysis

Strain	LC50​ (µg/g)	95% Confidence Interval	Slope (± SE)	X2
S1	122.45	[102.8–144.1]	2.06 ± 0.12	0.99
S2	40.26	[31.5–49.8]	2.08 ± 0.12	0.95
S3	120.8	[101.5–142.3]	2.07 ± 0.13	1.08
S4	129.64	[108.8–152.7]	2.03 ± 0.11	0.92
S5	160.58	[134.8–189.4]	1.96 ± 0.10	0.96
S6	32.41	[24.1–41.8]	2.15 ± 0.14	1.02
S7	90.35	[75.8–106.4]	2.21 ± 0.13	0.98
S8	114.78	[96.5–135.2]	2.10 ± 0.14	0.91
S9	132.85	[111.5–156.5]	2.01 ± 0.12	1.03
S10	95.36	[80.1–112.5]	2.18 ± 0.12	1.04
S11	127.8	[107.3–150.5]	2.04 ± 0.12	0.87
S12	164.25	[138.4–195.7]	1.95 ± 0.11	0.88
S13	122.54	[102.9–144.2]	2.06 ± 0.13	1.15
S14	155.25	[130.4–183.1]	1.97 ± 0.11	1.1
BLB1	70.56	[58.2–84.1]	2.30 ± 0.15	1.12

LC ₅₀ Lethal concentration required to kill 50% of the test population, SE Standard Error of the slope, X² Pearson chi-square goodness-of-fit

Fig. 2.

Graphs representation of Lethal concentration 50 (LC₅₀) (µg g⁻¹) of the oligosporogenic B. thuringiensis delta-endotoxins against E. kuehniella. Error bars depict a standard deviation of the mean values. Each data point is a mean of three experiments

Antifungal and antibacterial activities

Antifungal and antibacterial activities have been largely studied in B. thuringiensis strains, that’s why we aimed to investigate whether the random mutagenesis affects in a way or another the activity against six phytopathogenic fungal strains (A. niger, A. alternata, F. culmorum, F. graminarium, F. oxysporum, B. cinerea) and against the phytopathogenic bacteria A. tumefaciens. The screening of both antifungal and antibacterial activities among the 14 B. thuringiensis mutants was performed using supernatant after 24 h of growth and mutant pellets, respectively. Interestingly, S3 and S4 mutant clones showed an important antifungal activity against A. niger, F. culmorum and B. cinerea demonstrated by significant inhibition zones compared to the wild strain BLB1 lacking this activity (Table 5). The minimum inhibitory concentration (MIC) and minimum fungicidal concentration (MFC) tests were performed to evaluate this activity. Hence, the S3 and S4 clones showed MIC and MFC values ranging from 0.07 to 0.31 µg ml⁻¹ and from 0.05 to 0.22 µg ml⁻¹ against A. niger, F. culmorum and B. cinerea, respectively (Table 6).

Table 5.

Antifungal and antibacterial activities of S3 and S4 mutants

	Inhibitory diameter of S3 mutant (mm)	Inhibitory diameter of S4 mutant (mm)
B. cinerea	24,33 ± 1,52	17,33 ± 1,52
F. culmorum	28,66 ± 1,52	21,33 ± 2,08
A. niger	22 ± 1	15,33 ± 1,52
A. tumefaciens	15,66 ± 2,08	10,33 ± 1,52

Table 6.

MIC and MFC evaluation of S3 and S4 Bacillus thuringiensis mutants

Fungi	Strains
	S3 mutant		S4 mutant
	MIC	MFC	MIC	MFC
F. culmorum	70	70	56	22.5
B. cinerea	70	70	56	56
A. niger	31.5	31.5	22.5	22.5

MIC The minimum inhibitory concentrations (as µg ml⁻¹), MFC The minimum fungicidal concentration (as µg ml⁻¹)

On the other hand, we performed the screening of antibacterial activity against A. tumefaciens, the major causal agents of crown gall disease in tomato and vineyard crops via co-culture method since we noted the absence of activity with culture supernatants after 24 h and 48 h (data not shown). Surprisingly, the results showed that among the 14 B. thuringiensis mutants’ clones, the S3 and S4 mutants’ harbours also an important activity with inhibitory diameter of about 15,66 ± 2,08 mm and 10,33 ± 1,52 mm, respectively (Table 5) compared to the wild strain BLB1 showing a total absence of antibacterial activity.

All test fungi are among the highest fungal plant pathogens on the basis of their perceived importance, scientifically or economically essentially the three susceptible fungi A. niger, F. culmorum and B. cinerea. In fact, A. niger caused black mold disease on certain fruits and vegetables especially onions and ornamental plants [54]. Regarding, F. culmorum it causes foot and root rot on different small-grain cereals, in particular wheat and barley by the contamination of the grain with mycotoxins [55]. Moreover, B. cinerea known as grey mould, can infect more than 200 plant species as a typical necrotroph, which induce programmed cell death pathways in the host to achieve infection [56].

Bacillus strains have been reported to produce many hydrolytic enzymes that inhibit phytopathogenic fungi among of them β-glucanases, chitinases and some biosurfactant lipopeptides which can act in synergy [57, 58]. In order to explain the important antifungal activity detected in S3 and S4 B. thuringiensis mutants, we performed the assays to evaluate glucanases and chitinases activities. The obtained results showed the absence of these two enzymes in the supernatant of S3 and S4 B. thuringiensis mutants indicating that the increase of antifungal activity is not due to these two enzymes. This suggests that the high observed antifungal activity may be due to other antifungal molecules such as lipopeptides, biosurfactants, proteases or even due to synergistic activity between numerous hydrolytic enzymes [58, 59]. Rarely publications have studied the antibacterial activity against A. tumefaciens, the most important one is that published by Kamoun [60] which demonstrated that a bacteriocin of 11 kDa exhibited potent antimicrobial activity against Agrobacterium species.

In summary, enhancement of antimicrobial activities can be explained by occurrence of different mutations that altered the metabolic pathways of these mutants resulting a more efficient activity against the susceptible phytopathogenic fungi and bacteria. Further investigation into the specific mechanisms of action of oligosporogenic mutants implicated biomolecules is warranted given their potential role in this antifungal activity.

Use of date waste to produce delta-endotoxins

Biopesticide production cost including the expensive fermenters and fermentation media is the main challenge in the application of biocontrol technologies [25]. The use of cost-effective fermentation media mainly through studying the efficacy of locally available energy sources or wastes valorisation has been previously reported to reduce the cost of producing Bacillus sp. [61]. For example, tons of date palm are discarded by the date processing industry during the picking, storage and conditioning process [62]. Actually, this waste has limited uses such as animal feed. In Tunisia, the mean annual yield of date fruits can achieve 100,000 tons. That’s why it’s very important to develop an alternative use of this waste like biopesticide production. Therefore, we focused on the optimization of delta-endotoxins production using date waste, a very important agricultural by-product available abundantly in the Mediterranean area. For that, we selected the S6 mutant since it is the most toxic and effective clone against E. kuehniella larvae. Thus, the inoculation of S6 mutant clone into eight different concentrations varying from 2.5% to 20% of date waste medium was studied in order to select the most suitable concentration. A percentage of 17.5% giving the highest delta-endotoxins production was selected for further optimisation by using Taguchi design (Fig. 3). The results of nine experiments were analysed using the variance analysis (Tables 7 and 8). The experimental plan for the optimisation of delta endotoxin production using date waste was saturated, resulting in zero degrees of freedom for the error term and thus precluding direct significance testing. To address this limitation, we followed standard design-of-experiments practice for orthogonal arrays by identifying the factor with the lowest contribution to the total sum of squares (initial optical density) and excluding it from the model. The sum of squares associated with this factor was then pooled to estimate the experimental error variance. This pooling strategy allowed a valid ANOVA response, with R-squared value of 72.4%, demonstrating the contribution of these selected factors on the variation of delta-endotoxin production, allowing calculation of F-statistics and p-values which are now presented in Table 7. The results showed that the substrate concentration and agitation speed had a significant impact on the delta-endotoxin production. Account for substrate concentration, aeration ratio and agitation speed of the total variability, their effect is confirmed with F-values of 13.66, 7.01 and 9.14 and p values of 0.024, 0.047 and 0.035 respectively. To confirm the significance of these results, a new experiment was conducted with optimal conditions, which showed a significant increase in delta-endotoxins level with p value = 0.001 using Paired T-test. According to Taguchi’s signal–noise ratio, the substrate concentration has higher effect in level 3, the initial O.D.₆₀₀ was more significant in level 2, the aeration was better in level 3 and the agitation had the highest effect in level 3 (Fig. 4). Using optimized conditions of Taguchi design, we observed that the delta endotoxins production of the studied mutant increased to 1360 ± 25.49 mg L⁻¹ comparing to 710 ± 14.71 mg L⁻¹ assessed without using optimisation. This improvement representing 43.15% can be explained by the richness of date in carbohydrates and other nutrients such as minerals, proteins, fats, and vitamins essential of B. thuringiensis growth and delta-endotoxins formation [63].

Fig. 3.

Effects of different date waste percentages on delta endotoxins production

Table 7.

Definition of the factors and their corresponding levels and contributions studied by Taguchi’s L9 design

Code	Variables	Level 1	Level 2	Level 3	Contribution(%)	F- value	P- value
X1	Substrate concentration (%)	12.5	17.5	22.5	68.22	13.66	0.024
X2	Initial optical density	0.1	0.15	0.2	4.24	-	-
X3	Aeration	250	500	1000	6.94	7.01	0.047
X4	Agitation (rpm)	150	200	250	20.60	9.14	0.035

Level 1: the first answer. Level 2: the second level answer. Level 3: the third level answer

Table 8.

Taguchi’s L9 experiment plan with their experimental responses

Runs	X1	X2	X3	X4	Delta-endotoxins (mg L−1)
1	1	1	1	1	500 ± 14.14
2	1	2	2	2	320 ± 10.80
3	1	3	3	3	690 ± 16.32
4	2	1	2	3	1170 ± 66.83
5	2	2	3	2	1080 ± 35.59
6	2	3	1	1	480 ± 14.71
7	3	1	3	2	960 ± 10.80
8	3	2	1	3	780 ± 14.71
9	3	3	2	1	420 ± 10.80

Fig. 4.

Main plot effects of signal-to-noise ratios: evaluation of the impact of substrate concentration, initial concentration, aeration and agitation speeds on delta-endotoxins production

These results highlight the importance of the use of this waste in the manufacturing of microbial products.

Conclusion

These findings highlight the potential of the optimized random mutagenesis using three mutant agents as an efficient method affecting different genes either positively or negatively leading us to produce more eco-friendlier strains for biological control as a sustainable alternative to chemical pesticides.

Authors’ contributions

S. S. carried out the experiments, analysed the results and wrote the manuscript, F.A. and Z.G. assisted with the experiment work, M.K. performed additional data analysis, W.J. and S.R. supervised the study and provides critical feedback. All authors reviewed and approved the final manuscript.

Funding

This research received funding from the project “SAFWA-Alternative Biopesticides for Safe Integrated Pest and Water Management around the Mediterranean”, which is funded by the EU PRIMA program.

Data availability

“All data generated or analysed during this study are included in this published article.”

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.