Salmonella enterica comes in many different forms and causes a wide range of diseases in different animal hosts. In addition to gastrointestinal infections invasive infections are very common with serious consequences. The progression of an invasive infection is often measured by counting the viable salmonellae present in an organ. This gives an overall picture of the dynamic nature of these infections and the host-pathogen relationship, but underlying the apparently simple growth kinetics often observed there might well be hidden complexity. Often this complexity is broken down for study by using live cells in vitro, infecting them with salmonellae and measuring the fate of both the bacteria and the cells. We have been taking this information and testing whether it is valid in true in vivo infection systems. We have used Salmonella Typhimurium infections of mice, a natural host-pathogen relationship, rather than other large animal systems for reasons of availability and cost.
We have discovered that, contrary to what might be expected from in vitro studies, salmonellae are present inside macrophages in infected mice in low numbers. A substantial increase in viable bacterial counts per organ over time, associated with a lethal infection, is not reflected in increased intracellular counts, but rather by increased numbers of infected cells. Mathematical models of this process can capture our data and the dynamic processes that can be inferred. Using this framework for analysis, we have also shown that the Salmonella Pathogenicity Island 2 Type Three Secretory System is not an absolute requirement for intra-macrophage growth of salmonellae, but appears instead to be required for bacterial escape from infected cells and spread to infect new cells. Attenuation by knocking out these genes, the basis for new live vaccines, is not therefore dependent on simply preventing intracellular growth of the bacteria.