Suckling CD1 mice as an animal model for studies of Legionella pneumophila virulence.

On the assumption that specific host defences are lower in newborn and infant animals, the susceptibility of CD1 suckling mice to Legionella pneumophila was studied with the hypothesis that this model could detect consistent differences in virulence among Legionella isolates from various clinical and environmental sources. Mice 3-14 days old were indeed markedly susceptible to intraperitoneal challenge with fresh clinical isolates, but not to serially subcultured or type collection strains of L. pneumophila. For example, intraperitoneal inoculation of 10(7) cells of a fresh clinical isolate of L. pneumophila (strain Monza 3) caused 60% mortality of suckling mice in 1 day whereas the same number of cells of a culture-attenuated derivative (strain Monza 3p50) caused <10% mortality in >15 days. Lethal infection by the 'virulent' Monza 3 strain was strictly dependent on mouse age (no death was observed in mice >26 days old), required the inoculation of viable cells and was not related to endotoxin production. The 'virulent' L. pneumophila strain was cleared from mouse lungs less rapidly, while adhering to, and being internalised into the peritoneal exudate cells (PEC) of suckling mice to a greater extent, than the avirulent derivative, as shown by immunofluorescence and confocal microscopy. The Monza 3 strain also induced the production by PEC in vivo of 5-to-10 times more tumour necrosis factor-alpha (TNF-alpha) mRNA than the Monza 3p50 strain. Overall, suckling CD1 mice appear to provide a promising, easily handled, highly reproducible and relatively inexpensive animal model for studies of the virulence of L. pneumophila, and possibly, of the role of pro-inflammatory cytokine production in this phenomenon.