Journal of Applied Biosciences 224: 24966 – 24974
ISSN 1997-5902
Toxicity of delta endotoxins from native Bacillus thuringiensis isolates against the larger grain borer Prostephanus truncatus (Horn) under different maize-based toxin delivery forms
Otieno, S.A.¹²* and Onyango, P.A.³
¹ Centre for Biotechnology and Bioinformatics, University of Nairobi, P.O. Box 30197–00100, Nairobi, Kenya
² Department of Crops, Horticulture and Soils, Egerton University, P.O. Box 536–20115, Egerton, Njoro, Nakuru, Kenya
³ Health Services Department, Jomo Kenyatta University of Agriculture and Technology (JKUAT), P.O. Box 62000–00200, Nairobi, Kenya
*Corresponding author: akellosandy@gmail.com
Submitted 12/02/2026, Published online on 30/09/2026 in the https://www.m.elewa.org/journals/journal-of-applied-biosciences-about-jab/ https://doi.org/10.35759/JABs.224.3
ABSTRACT
Objectives: To evaluate the toxicity of delta endotoxins from native Bacillus thuringiensis isolates against adult Prostephanus truncatus and determine whether maize-based toxin delivery form and toxin solubility influence mortality under laboratory conditions.
Methodology and Results: Adult P. truncatus were exposed separately to soluble toxin fractions or insoluble spore–crystal preparations from eight B. thuringiensis isolates delivered on whole maize grains, coarsely ground maize particles or maize flour. Re-entry of the surviving mortality worksheet at the ten-assessment structure documented in the thesis appendix gave 539 observations. Three-way ANOVA detected effects of isolate (F8,485 = 4.268, p < 0.001) and maize form (F2,485 = 3.676, p = 0.026), but not solubility (F1,485 = 0.249, p = 0.618); all interactions were non-significant.
Conclusions and application of findings: Toxicity was isolate-dependent, with KG 20 producing the highest pooled mortality. Dunnett-adjusted comparisons showed greater mortality than the untreated control for KG 12, KG 20, KG 411, TEN, KUR and KM 31, whereas KM 24 and ISR were not significant after multiplicity adjustment. Maize form affected mortality, but the absence of a solubility effect indicates that a less processed spore–crystal formulation may be a practical development route. The findings support confirmatory dose–response, persistence, food-safety and whole-grain storage trials.
Keywords: Bacillus thuringiensis; delta endotoxins; Prostephanus truncatus; stored maize; biological control
INTRODUCTION
Post-harvest losses caused by insect pests remain a major constraint to food security in sub-Saharan Africa. The larger grain borer, Prostephanus truncatus (Horn), is a destructive pest of stored maize. Its feeding damages kernels and produces frass, increasing quantitative and quality losses (Muatinte et al., 2019; Panagiotakis et al., 2023). Reliance on synthetic insecticides also raises concerns about resistance, residues, environmental contamination and human exposure (Pathak et al., 2022). Bacillus thuringiensis (Bt) is used as a microbial insecticide because its pesticidal proteins can act selectively after ingestion. Toxicity nevertheless varies among Bt strains and insects because toxin composition, solubilization, proteolytic activation and midgut receptor interactions differ (Jurat-Fuentes et al., 2021; Endo, 2022; Aswathi et al., 2024). Formulation can also affect stability, adherence and ingestion of Bt products (de Oliveira et al., 2021). This study evaluated delta endotoxins from five archived Kenyan Bt isolates and three reference varieties against adult P. truncatus. It tested whether Bt isolate, maize-based delivery form and toxin solubility affected mortality, with the aim of identifying candidates for further development for stored-maize protection.
MATERIALS AND METHODS
The activity of Bacillus thuringiensis (Bt) delta endotoxins was determined following confirmation of isolate identity and subsequent purification and extraction of the toxins using the procedures described below.
Bacillus thuringiensis isolates: Eight Bacillus thuringiensis preparations were evaluated: Kenyan isolates KG 411, KG 12-0 and KG 20 from Gazi in coastal Kenya, and KM 31 and KM 24 from Machakos in eastern Kenya, together with B. thuringiensis var. tenebrionis, var. israelensis and var. kurstaki used as reference comparators. The cultures were obtained from the Centre for Biotechnology and Bioinformatics, University of Nairobi, where they had been maintained at −20 °C for approximately three years before this study.
Culture and toxin preparation: The Bt isolates were cultured following Wang’ondu et al. (2003), with study-specific modifications. Cultures were grown in nutrient broth at 28 ± 2 °C and 200 rpm until approximately 90% cell lysis after 72 h. Spores and endotoxins were harvested by centrifugation at 5,000 rpm and 5 °C for 15 min, washed with phosphate buffer (0.1 M NaH2PO4·H2O and 0.1 M NaCl, pH 6.0), and prepared for the assays described below.
Microscopy: Purity of crystal preparations was assessed using Gram staining (Poinar, 2012). Gram-positive cells appeared dark violet with visible refractile spores, while Gram-negative cells appeared pink following counterstaining.
Protein Estimation and Solubilization of Bt Toxins : Protein concentration of each Bt isolate was estimated using bovine serum albumin (BSA) standards spanning 0–2.0 mg mL⁻¹. Absorbance was measured at 280 nm, and the original standard readings were fitted by ordinary least-squares linear regression. Stock concentrations were back-calculated from the fitted equation before preparation of the working bioassay suspensions. The archived enzymatic-activity readings at 750 nm were not used for protein quantification. Purified Bt crystals (0.5 mg) were solubilized in 50 mM Na₂CO₃–HCl buffer (pH 9.5) containing 10 mM dithiothreitol (DTT). The mixture was incubated in a water bath at 37 °C for 14 h following a modification of the protocol by Haider et al. (1986). Insoluble material was removed by centrifugation at 4,000 × g for 10 min at 4 °C. The insoluble protoxin fraction was washed in phosphate-buffered saline (PBS) by centrifugation to remove residual Na₂CO₃, HCl and DTT. Protein estimation was carried out for both soluble and insoluble protoxin fractions as described above, and samples were stored at −20 °C.
Protease Treatment of Bt Protoxins : Solubilized protoxin fractions were dialysed in 0.1 M Tris-HCl buffer (pH 8.0) at 4 °C for 24 h, while insoluble protoxin fractions were washed three times in the same buffer by centrifugation at 4,000 × g for 10 min at 4 °C. Samples were treated with bovine pancreatic trypsin and α-chymotrypsin (Serva, Heidelberg, Germany), as well as partially purified trypsin-like protease from Glossina spp. midgut. Crude gut homogenate was prepared from adult P. truncatus by dissecting the guts and homogenising them for 1 min in ice-cold 0.1 M Tris-HCl (pH 8.0). After centrifugation at 12,000 × g for 10 min at 4 °C, trypsin activity in the supernatant was measured using α-N-benzoyl-DL-arginine-p-nitroanilide (BApNA) (Wamunyokoli, 1992). Protoxin was also treated separately with bovine pancreatic trypsin, α-chymotrypsin, or partially purified trypsin-like protease from Glossina midgut; enzyme–protoxin mixtures were incubated at 37 °C for 60 min and then held at 4 °C.
Screening of Bacillus thuringiensis Toxicity against Prostephanus truncatus: Adult P. truncatus (at least 30 days old) were obtained from KARI-NARL (now KALRO). The laboratory colony was maintained on autoclaved maize grains at 27 ± 2 °C and 65 ± 5% relative humidity. Unsexed adults were used. Soluble toxin fractions and insoluble spore–crystal preparations were evaluated separately for each isolate; toxins from different isolates were not mixed. For each treatment, 2 mL of preparation was mixed with 10 g of whole maize grain (MG), coarsely ground maize particles (MP), or finely ground maize flour (MF). Treated diets were air-dried for 24 h at 30 °C. Ten adults were placed in each Petri dish, with three replicate dishes per treatment. Mortality was recorded over 21 days, and the experiment was conducted three times. Reported descriptive means are numbers of dead adults on a 0–10 scale for the experimental unit, not totals out of 30 adults.
Data Analysis: The analysis was carried out by having dead-adult counts were entered from the ten numbered assessment blocks corresponding to the N = 10 cell structure using SPSS version 21 and in Python 3.12 using SciPy 1.17. Available replicate-dish entries were averaged within each assessment × isolate × maize form × solubility cell. One KUR × maize-particle × insoluble assessment was missing, giving N = 539 and error df = 485. Mortality was analysed by three-way fixed-effects ANOVA using sum-to-zero contrasts. Partial η² was calculated for each effect. Unadjusted least significant difference (LSD) comparisons were used to derive compact-letter groupings, while two-sided Dunnett simultaneous tests compared each Bt isolate with the untreated control. Statistical significance was assessed at α = 0.05.
RESULTS
Growth Characteristics and Microscopy of Bacillus thuringiensis: Growth and microscopic characteristics confirmed all isolates as B. thuringiensis. Reanalysis of the original BSA standards showed a strong positive linear relationship between absorbance and protein concentration (r = 0.9637; R² = 0.9286; p < 0.001). All B. thuringiensis isolates formed pale cream colonies on nutrient agar (Plate 1a). Microscopic examination revealed actively multiplying vegetative cells and the presence of spores in sporulating cultures (Plate 1b). Gram staining showed that the isolates were Gram-positive rods, occurring singly or in chains, with refractile endospores visible either within vegetative cells or as free spores.
Plate 1. (a) Colony morphology of Bacillus thuringiensis on nutrient agar; (b) Gram-stained cells showing vegetative cells and refractile spores (×1000). Black arrows indicate intracellular spores; blue arrows indicate free spores.
Protein Estimation :Protein concentrations of Bt delta endotoxins were estimated from the author-supplied BSA standard series (Fig. 1). The fitted equation was absorbance = 0.2720 × concentration + 3.0751 (r = 0.9637; R² = 0.9286; p < 0.001). Back-calculated stock estimates ranged from 0.389 mg mL⁻¹ for KG 12-0 to 1.492 mg mL⁻¹ for KUR (Supplementary Table S4).
Figure 1: Original bovine serum albumin standard curve used to estimate Bt protein concentrations. Points are the supplied standards; the line is the ordinary least-squares fit (n = 11).
Mortality of P. truncatus differed among B. thuringiensis isolates (F₈,₄₈₅ = 4.268, p < 0.001, partial η² = 0.066) and maize delivery forms (F₂,₄₈₅ = 3.676, p = 0.026, partial η² = 0.015), but not between soluble and insoluble preparations (F₁,₄₈₅ = 0.249, p = 0.618). No interaction was significant (Table 1).
Table 1: Effects of Bacillus thuringiensis isolate, maize form and toxin solubility on mortality of Prostephanus truncatus
| Source of variation | df | F-value | p-value | Partial η² |
| Bt isolate | 8 | 4.268 | < 0.001* | 0.066 |
| Maize form | 2 | 3.676 | 0.026* | 0.015 |
| Solubility | 1 | 0.249 | 0.618 ns | 0.001 |
| Bt × maize form | 16 | 1.436 | 0.120 ns | 0.045 |
| Bt × solubility | 8 | 0.991 | 0.442 ns | 0.016 |
| Maize × solubility | 2 | 1.250 | 0.288 ns | 0.005 |
| Bt × maize × solubility | 16 | 0.955 | 0.505 ns | 0.031 |
Significant at p ≤ 0.05; ns = not significant. Three-way fixed-effects ANOVA; N = 539, error df = 485. Partial η² is reported as the effect-size estimate.
Pooled estimated marginal mortality was highest for KG 20 (3.519 adults per dish). LSD comparisons showed that KG 20 exceeded KM 24 (p = 0.006), ISR (p = 0.004) and TEN (p = 0.037); no other Bt-to-Bt LSD comparison was significant. Dunnett-adjusted comparisons with the untreated control were significant for KG 12 (p = 0.001), KG 20 (p < 0.001), KG 411 (p < 0.001), TEN (p = 0.009), KUR (p = 0.003) and KM 31 (p = 0.004), but not KM 24 (p = 0.059) or ISR (p = 0.085). Compact-letter groups were KG 20 (a); KG 411, KG 12, KUR and KM 31 (ab); TEN, KM 24 and ISR (b); and control (c) (Fig. 4; Supplementary Table S3).
Figure 2: Mean mortality (± SD) of Prostephanus truncatus exposed to soluble Bacillus thuringiensis preparations delivered on maize grains (MG), maize particles (MP) or maize flour (MF). Values are assessment-level mean numbers of dead adults per dish on a 0–10 scale (n = 10 assessments per cell).
Figure 3:Mean mortality (± SD) of Prostephanus truncatus exposed to insoluble Bacillus thuringiensis spore–crystal preparations delivered on maize grains (MG), maize particles (MP) or maize flour (MF). Values are assessment-level mean numbers of dead adults per dish on a 0–10 scale; n = 10 assessments per cell except KUR–MP (n = 9).
Figure 4: Estimated marginal mean mortality (± SE) pooled across maize form and solubility. Isolates sharing a letter are not different by unadjusted LSD at p ≤ 0.05. An asterisk marks a significant two-sided Dunnett comparison with the untreated control after simultaneous adjustment.
DISCUSSION
The isolate main effect confirms that Bt activity against P. truncatus was strain-dependent. KG 20 had the highest pooled estimated marginal mortality, while KG 12, KG 20, KG 411, TEN, KUR and KM 31 each differed from the untreated control after Dunnett adjustment. KM 24 and ISR did not meet the adjusted 0.05 threshold. Strain-dependent activity is biologically plausible because pesticidal-protein composition, activation and receptor binding determine potency (Jurat-Fuentes et al., 2021; Endo, 2022; Aswathi et al., 2024). KG 20 is therefore the clearest lead from the pooled raw-data reanalysis, with KG 12, KG 411, KUR and KM 31 sharing an overlapping intermediate LSD group. These isolates are research leads rather than finished products. Performance should be confirmed through concentration–response assays, appropriate positive controls, independent culture batches and time-to-mortality analysis; characterization of pesticidal-protein profiles would help explain the isolate effect. Maize was both food and toxin carrier, so substrate structure could alter contact, adherence and ingestion. The significant maize-form main effect indicates that the delivery matrix mattered. However, treatment means were heterogeneous: particles did not produce the highest mortality for every isolate and solubility combination. The result should therefore be interpreted as evidence that formulation and substrate require joint optimization, rather than as a universal particle-size effect. Whole grain is the relevant storage commodity for most farmers. Grinding before storage would increase exposed surface area and may worsen handling and contamination risks; this study does not recommend that farmers store maize as particles. The practical route is to develop a formulation that adheres to intact kernels and remains available during P. truncatus feeding. Candidate formulations should be evaluated for persistence, grain quality, mould and mycotoxin risk, non-target effects and acceptability before recommendation. Solubility had no detectable main effect (p = 0.618), and none of the tested interactions was significant at α = 0.05. The additional solubilization step therefore did not measurably improve mortality under these assay conditions. A simpler spore–crystal formulation could reduce processing requirements, but that remains a product-development hypothesis requiring dose standardization, shelf-life testing and validation on whole grain (de Oliveira et al., 2021). Direct efficacy comparisons with entomopathogenic fungi or contact insecticides are inappropriate because exposure routes, doses, endpoints and mechanisms differ. Bt preparations should be compared with Bt-based benchmarks under matched bioassay conditions. Recent work nevertheless confirms the continuing need for safer, effective P. truncatus management in stored maize (Muatinte et al., 2019; Panagiotakis et al., 2023; Mlambo et al., 2026). Together, the isolate, solubility and delivery-form results identify KG 20 as the primary candidate for further formulation work and support confirmatory evaluation of KG 12, KG 411, KUR and KM 31. Whole-grain performance should be strengthened through improved adherence and persistence while retaining the low-processing advantage suggested by comparable soluble and insoluble preparations.
CONCLUSION AND APPLICATION OF RESULTS
Raw-data reanalysis identified KG 20 as the strongest pooled Bacillus thuringiensis candidate against Prostephanus truncatus; KG 12, KG 411, TEN, KUR and KM 31 also produced greater mortality than the untreated control after Dunnett adjustment. Maize delivery form influenced mortality, whereas toxin solubility did not. Further work should standardize dose, characterize pesticidal proteins, test grain-adherent formulations on intact maize, and assess persistence, food safety, mould and mycotoxin risk under realistic storage conditions before any farmer recommendation or product claim.
ACKNOWLEDGEMENTS
The authors acknowledge the School of Biological Sciences and the Centre for Biotechnology and Bioinformatics, University of Nairobi, for research support. Additional support from the Health Services Department, Jomo Kenyatta University of Agriculture and Technology, is also gratefully acknowledged.
REFERENCES
Aswathi N, Balakrishnan N, Srinivasan T, Kokiladevi E, Raghu R, 2024. Diversity of Bt toxins and their utility in pest management. Egyptian Journal of Biological Pest Control 34: 40. https://doi.org/10.1186/s41938-024-00803-6
de Oliveira JL, Fraceto LF, Bravo A, Polanczyk RA, 2021. Encapsulation strategies for Bacillus thuringiensis: from now to the future. Journal of Agricultural and Food Chemistry 69: 4564–4577. https://doi.org/10.1021/acs.jafc.0c07118
Endo H, 2022. Molecular and kinetic models for pore formation of Bacillus thuringiensis Cry toxin. Toxins 14: 433. https://doi.org/10.3390/toxins14070433
Haider MZ, Knowles BH, Ellar DJ, 1986. Specificity of Bacillus thuringiensis δ-endotoxins for insect species. European Journal of Biochemistry 156: 531–540.
Jurat-Fuentes JL, Heckel DG, Ferré J, 2021. Mechanisms of resistance to insecticidal proteins from Bacillus thuringiensis. Annual Review of Entomology 66: 121–140.
Mlambo S, Mvumi BM, Machekano H, Nyamukondiwa C, 2026. Prostephanus truncatus outcompetes Sitophilus zeamais under varying temperatures and introduction sequence on stored maize grain. Journal of Stored Products Research 115: 102880.
Muatinte BL, Kavallieratos NG, Boukouvala MC, García-Lara S, López-Castillo LM, Mvumi BM, 2019. The threat of the larger grain borer, Prostephanus truncatus, and practical control options for the pest. CABI Reviews 2019: 1–25.
Otieno SA, 2013. Determination of native Bacillus thuringiensis delta-endotoxin potency against Prostephanus truncatus and fumonisin-producing Fusarium verticillioides isolated from maize grains. MSc thesis, University of Nairobi, Nairobi, Kenya.
Panagiotakis A, Baliota GV, Rumbos CI, Athanassiou CG, 2023. Efficacy of contact insecticides for the control of the larger grain borer, Prostephanus truncatus, on stored maize. Agriculture 13: 1502. https://doi.org/10.3390/agriculture13081502
Pathak VM, Verma VK, Rawat BS, Kaur B, Babu N, Sharma A, Dewali S, Yadav M, Kumari R, Singh S, Mohapatra A, 2022. Current status of pesticide effects on environment, human health and eco-friendly management as bioremediation: a comprehensive review. Frontiers in Microbiology 13: 962619. https://doi.org/10.3389/fmicb.2022.962619
Poinar GO, 2012. Diagnostic Manual for the Identification of Insect Pathogens. Springer, Dordrecht.
Wamunyokoli FA, 1992. Isolation, purification and characterization of Bacillus thuringiensis endotoxin active against mosquito larvae. MSc thesis, University of Nairobi, Nairobi, Kenya.
Wang’ondu VW, Kahindi JHP, Olembo NK, Ochanda JO, 2003. Isolation and characterization of Bacillus thuringiensis strains from soil samples from Kakamega and Machakos Districts in Kenya. Journal of Tropical Microbiology and Biotechnology 2: 3–10. https://doi.org/10.4314/jtmb.v2i1.35418