Positive correlation between in vivo and in vitro assays for the evaluation of Pseudomonas virulence.

Some bacterial phenotypes measured in vitro can be used to access bacterial virulence, on the premise that they are positively correlated with data from in vivo experiments. We show here that in vitro assessment of bacterial phenotypes, such as adherence and cytotoxicity, are positively correlated with data from in vivo experiments in Drosophila and can be used to assess bacterial virulence in vivo. Manipulation of environmental parameters, such as iron availability, induced changes in the phenotypes measured in vitro that correlated with changes in vivo virulence of all strains tested. Applying these assays, we demonstrate the pathogenic potential of a Pseudomonas fluorescens strain, initially isolated as a non-pathogenic milk contaminant. This strain displayed adherence and cytotoxicity comparable to those of the Pseudomonas aeruginosa pathogenic strain PAK, and colonized the infected flies as rapidly as the PAK strain. These results indicate that this "a priori" non-pathogenic bacterium is capable of escaping the host immune response, supporting the use of in vitro tests for screening of potential pathogens.