Virulence of Shigatoxigenic and Enteropathogenic Escherichia coli O80:H2 in Galleria mellonella Larvae: Comparison of the Roles of the pS88 Plasmids and STX2d Phage

SIMPLE SUMMARY: Following the “Replacement/Reduction/Refinement” policy, insects are proposed to replace mammals and birds as experimental models to study the virulence of bacterial pathogens and to identify their virulence properties. The aim of this study was to assess in larvae of the Galleria mellonella moth the virulence of the Shigatoxigenic and enteropathogenic Escherichia coli O80:H2 and the respective roles of two virulence properties: the pS88 plasmid-encoded invasiveness properties and the phage-encoded Shiga toxin 2d. The objectives were to compare: (i) the virulence of bovine Shigatoxigenic and enteropathogenic E. coli O80:H2; (ii) the roles of the pS88 plasmid and Shiga toxin 2d-encoding phage; (iii) the virulence of E. coli O80:H2 and O80:non-H2. The results and the conclusions are: (i) E. coli O80:H2 and O80:non-H2 are lethal at log5 and log6 concentrations; (ii) the pS88 plasmids are partially responsible for the virulence of E. coli O80:H2; (iii) the phage-encoded Stx2d toxin is entirely responsible for the virulence of the Shigatoxigenic Escherichia coli O80:H2; (iv) the virulence properties of E. coli O80:non-H2 could not be identified. As a general conclusion, G. mellonella larvae represent a useful model to study the virulence of bacterial pathogens but are limited in identifying their virulence properties. ABSTRACT: The invasiveness properties of Shigatoxigenic and enteropathogenic Escherichia coli (STEC and EPEC) O80:H2 in humans and calves are encoded by genes located on a pS88-like ColV conjugative plasmid. The main objectives of this study in larvae of the Galleria mellonella moth were therefore to compare the virulence of eight bovine STEC and EPEC O80:H2, of two E. coli pS88 plasmid transconjugant and STX2d phage transductant K12 DH10B, of four E. coli O80:non-H2, and of the laboratory E. coli K12 DH10B strains. Thirty larvae per strain were inoculated in the last proleg with 10 μL of tenfold dilutions of each bacterial culture corresponding to 10 to 10(6) colony-forming units (CFUs). The larvae were kept at 37 °C and their mortality rate was followed daily for four days. The main results were that: (i) not only the STEC and EPEC O80:H2, but also different E. coli O80:non-H2 were lethal for the larvae at high concentrations (from 10(4) to 10(6) CFU) with some variation according to the strain; (ii) the Stx2d toxin and partially the pS88 plasmid were responsible for the lethality caused by the E. coli O80:H2; (iii) the virulence factors of E. coli O80:non-H2 were not identified. The general conclusions are that, although the Galleria mellonella larvae represent a useful first-line model to study the virulence of bacterial pathogens, they are more limited in identifying their actual virulence properties.