Silkworm model for Bacillus anthracis infection and virulence determination

Abstract

Bacillus anthracis is an obligate pathogen and a causative agent of anthrax. Its major virulence factors are plasmid-coded; however, recent studies have revealed chromosome-encoded virulence factors, indicating that the current understanding of its virulence mechanism is elusive and needs further investigation. In this study, we established a silkworm (Bombyx mori) infection model of B. anthracis. We showed that silkworms were killed by B. anthracis Sterne and cured of the infection when administered with antibiotics. We quantitatively determined the lethal dose of the bacteria that kills 50% larvae and effective doses of antibiotics that cure 50% infected larvae. Furthermore, we demonstrated that B. anthracis mutants with disruption in virulence genes such as pagA, lef, and atxA had attenuated silkworm-killing ability and reduced colonization in silkworm hemolymph. The silkworm infection model established in this study can be utilized in large-scale infection experiments to identify novel virulence determinants and develop novel therapeutic options against B. anthracis infections.

Keywords: Bacillus anthracis; animal model; bombyx mori; host-pathogen interaction; silkworm; virulence.

Figure 1.

Bacillus anthracis kills silkworms. a. Dose-dependent killing of silkworms by B. anthracis. Data is representative of three independent experiments. b. Survival of silkworms 16 h post-infection. Data are presented as a combined result of three independent experiments. LD₅₀ is calculated by logistic regression analysis using the logit link function. c. Dead silkworms infected with B. anthracis (8.1 x 10² CFU/larva) 19 h post-infection. The dead silkworms turn black due to melanization. d. Alive silkworms injected with saline 19 h post-injection

Figure 2.

Live bacteria is required for silkworm killing. a. Survival of silkworms after injection of heat-killed B. anthracis. Live B. anthracis (2.6 x 10³ CFU/larva) and heat-killed B. anthracis (1.5 x 10⁶ CFU/larva) were injected to silkworms (n = 8), and survival was observed. Representative data of two independent experiments are shown. b. Survival of silkworms after infection with wild-type and BYF10124. Wild-type B. anthracis (5 x 10² CFU/larva) and BYF10124 (8.5 x 10² CFU/larva) were injected to silkworms (n = 7), and survival was observed. Representative data of two independent experiments are shown c. In vitro grown BYF10124 under the microscope. d, e. BYF10124 in silkworm hemolymph 3 h (d) and 6 h (e) post-infection under the microscope. White arrow-heads show representative hemocytes of silkworm hemolymph. Scale bars, 20 µm

Figure 3.

Treatment of B. anthracis infection by antibiotics. a. Survival of silkworms (n = 10) with and without antibiotics treatment. Representative data from three independent experiments are shown. b, c. Survival of silkworms treated with various concentrations of doxycycline (b) or ampicillin (c) 16 h post-infection. Data are mean ± SEM of three independent experiments. ED₅₀ value was calculated by logistic regression analysis using the logit link function. d. Bacterial burden after 6 h and 9 h post-infection with and without antibiotics treatment (1 mg/kg). Statistical analysis was performed by one-way analysis of variance (ANOVA) with Dunnett’s multiple comparison test compared with the wild-type. The dotted line shows the limit of detection

Figure 4.

Clearance of B. anthracis from silkworm hemolymph upon antibiotic treatment. Time course of B. anthracis clearance by antibiotic treatment at 3 h (a, b) and 6 h (c, d) post-infection are shown. Silkworms were infected with BYF10124 and injected with either vehicle (a, c) or 1 mg/kg ampicillin (b, d) at the specified time, hemolymph was obtained and visualized under the microscope. Scale bars, 20 µm

Figure 5.

Assessment of virulence of B. anthracis mutants using silkworms. a. Survival of silkworms after infection with wild-type and virulence gene-disrupted mutants. Exponentially growing bacteria were injected into the hemolymph of silkworms, and survival was observed. Data are shown as a combined result from two independent experiments (n = 15). Injected CFU/larva: WT (wild-type) = 5 x 10² and 2.6 x 10³; ΔatxA = 4 x 10² and 3.4 x 10³; Δlef = 7.5 x 10² and 2 x 10³; ΔpagA = 4.4 x 10³ and 1.0 × 10⁴. Statistical analysis was performed by Mantel-Cox log-rank test. b. Microbial burden of B. anthracis wild-type and mutants in silkworms 6 h post-infection. Exponentially growing bacteria were injected into the hemolymph of silkworms, hemolymph was recovered 6 h post-infection, and CFU was determined. Data are shown as a combined result from two independent experiments (n = 27). Injected CFU/larva: WT (wild-type) = 2 x 10² and 3 x 10²; ΔatxA = 1.8 x 10² and 4.4 x 10²; Δlef = 2.2 x 10² and 2.4 x 10²; ΔpagA = 1.1 x 10² and 2.8 × 10². Statistical analysis was performed by one-way analysis of variance (ANOVA) with Dunnett’s multiple comparison test compared with the WT (* p < 0.05, **** p < 0.0001). The dotted line represents the limit of detection. c. Quantitative evaluation of virulence using silkworms. Exponentially growing bacteria were injected into the hemolymph of silkworms, and survival was observed. Data are mean ± SEM of three independent experiments for ΔatxA and Δlef and six independent experiments for wildtype and ΔpagA. LD₅₀ value was calculated 16 h post-infection by logistic regression analysis using the logit link function. Statistical analysis was performed by one-way analysis of variance (ANOVA) with Dunnett’s multiple comparison test compared with the WT (* p < 0.05, **** p < 0.0001)