Phage-Based Control of Methicillin Resistant Staphylococcus aureus in a Galleria mellonella Model of Implant-Associated Infection

Abstract

Staphylococcus aureus implant-associated infections are difficult to treat because of the ability of bacteria to form biofilm on medical devices. Here, the efficacy of Sb-1 to control or prevent S. aureus colonization on medical foreign bodies was investigated in a Galleria mellonella larval infection model. For colonization control assays, sterile K-wires were implanted into larva prolegs. After 2 days, larvae were infected with methicillin-resistant S. aureus ATCC 43300 and incubated at 37 °C for a further 2 days, when treatments with either daptomycin (4 mg/kg), Sb-1 (10⁷ PFUs) or a combination of them (3 x/day) were started. For biofilm prevention assays, larvae were pre-treated with either vancomycin (10 mg/kg) or Sb-1 (10⁷ PFUs) before the S. aureus infection. In both experimental settings, K-wires were explanted for colony counting two days after treatment. In comparison to the untreated control, more than a 4 log¹⁰ CFU and 1 log¹⁰ CFU reduction was observed on K-wires recovered from larvae treated with the Sb-1/daptomycin combination and with their singular administration, respectively. Moreover, pre-infection treatment with Sb-1 was found to prevent K-wire colonization, similarly to vancomycin. Taken together, the obtained results demonstrated the strong potential of the Sb-1 antibiotic combinatory administration or the Sb-1 pretreatment to control or prevent S. aureus-associated implant infections.

Keywords: Galleria mellonella; K-wire; Staphylococcus aureus; antibiotic resistance; biofilm; phage therapy; prosthetic infections; staphylococcal bacteriophages.

Figure 1

Evaluation of the Sb-1 bactericidal activity against biofilm-embedded MRSA ATCC 43300. The histograms represent the mean CFU number ± SEM of biofilm dislodged MRSA treated/untreated with different titers (PFU/mL) of Sb-1. Dashed line indicates 3 log¹⁰ CFU reduction in comparison to the untreated control (0). Three independent experiments were performed.

Figure 2

Evaluation of the optimum infectious dose for K-wire colonization in G. mellonella infection. Different inocula (0, 10⁴, 10⁵ and 10⁶ CFUs) of methicillin-resistant S. aureus ATCC 43300 were tested to find the proper infectious dose. Fifty CFUs were the detection limit. Data from a representative experiment are reported here.

Figure 3

Sb-1 stability in G. mellonella larvae over time. Larvae were infected with Sb-1 (10⁸ PFU/mL) and sacrificed at different time points (up to 24 h) to collect hemolymph and perform plaque assays for phage counting. Data from a representative experiment are reported here.

Figure 4

Treatment of MRSA ATCC 43300 K-wire-associated infection with different Sb-1 and daptomycin formulations in G. mellonella larvae. After infection with MRSA ATCC 43300 (10⁵ CFUs) larvae were treated with (i) Sb-1 (10⁷ PFUs) alone, (ii) daptomycin (4 mg/kg) alone and staggered administration of (iii) Sb-1 (10⁷ PFUs)/PBS, (iv) PBS/daptomycin (4 mg/kg), and (v) Sb-1 (10⁷ PFUs)/daptomycin (4 mg/kg). Statistical differences in bacterial count per K-wire obtained from different treatment groups were determined by the Kruskal–Wallis test using Dunn correction (* p ≤ 0.05; ** p ≤ 0.01; *** p ≤ 0.001; **** p < 0.0001). Data obtained from three independent experiments are reported.

Figure 5

Efficacy of Sb-1 in preventing MRSA ATCC 43300 K-wire colonization in a G. mellonella model of implant-associated infection. Sb-1 (10⁷ PFUs) and vancomycin (10 mg/kg) were injected into larvae before the infection with MRSA ATCC 43300 (10⁵ CFUs). The graphs show (a) CFU number of K-wires explanted 2 days post-infection from larvae pretreated with either Sb-1 or vancomycin; (b) CFU/mL number from hemolymph collected 2 days post-infection from larvae pretreated with either Sb-1 or vancomycin; (c) CFU number of K-wires explanted 5 days post-infection from larvae pretreated with either Sb-1 or vancomycin; (d) CFU number from hemolymph collected 5 days post-infection from larvae pretreated with either Sb-1 or vancomycin. Statistical differences in bacterial count per K-wire and bacterial count per mL of hemolymph obtained from different pre-treatment groups were determined by the Kruskal–Wallis test using Dunn correction (ns p > 0.05; ** p ≤ 0.01; *** p ≤ 0.001; **** p < 0.0001, ns = not significant). Data obtained from two independent experiments are reported.

Figure 6

Schematic representation of the experimental model for the establishment of implant-associated infection in G. mellonella larvae. (A) Evaluation of the optimum infectious dose of MRSA ATCC 43300 able to colonize K-wires implanted in larvae. (B) Sb-1 stability in hemolymph of G. mellonella larvae and kinetic of phage inactivation.

Figure 7

Schematic representation of the procedure used in the experimental model for the evaluation of the Sb-1 in vivo activity in treating (A) and preventing (B) implanted K-wire colonization by methicillin-resistant S. aureus ATCC 43300.