Francisella tularensis is a highly infective, intracellular bacterium. It is capable of infecting a wide range of mammals and causes the disease tularemia in humans. As a result of its high infectivity there have been a lot of efforts made to create a generally available vaccine against this pathogen. One potential vaccine candidate is the FSC043 strain, a spontaneous mutant that has acquired mutations making it attenuated for replication both in vitro and in the experimental mouse model. However, it was noted that it afforded protection against challenge with a highly virulent F. tularensis strain. The aim of this thesis has been to delineate the mechanisms of its attenuation to better understand F. tularensis pathogenesis and to obtain a better knowledge about the prerequisites of protective immunity against this potent pathogen. Microarray and whole-genome sequencing revealed four mutations in the attenuated FSC043 strain that were not present in the virulent SCHU S4 isolate. One of these mutations has been described earlier as it results in a fusion protein also found in other attenuated strains. Among the other differences, two mutations were identical nonsense mutations in a duplicated gene region known as the Francisella pathogenicity island (FPI). The affected gene, pdpC, is coding for PdpC (pathogenicity determinant protein C). We found that these mutations resulted in a truncated form of PdpC, and also that the downstream gene was severely downregulated due to these mutations. Further, our studies revealed that the intracellular phenotype of the FSC043 strain differed from other tested strains in that a small portion of the intracellular bacteria were able to escape the phagosome and multiply within the host, while the majority of intracellular bacteria stayed confined to the phagosome. We wanted to study the specific function of pdpC and therefore deleted both copies of it in the virulent SCHU S4strain as well as the Live Vaccine Strain, an empirically attenuated strain often used as a model for the virulent strains of F. tularensis. The resulting mutants showed an attenuated phenotype; no intracellular growth in murine cells, and no virulence in mice. When studying the intracellular localization of the LVS Δpdpc mutant, we found that it was uniformly located adjacent to phagosomal membrane-like structures but that the membrane was markedly disrupted. Further, this mutant induced an MOI-dependent cytotoxicity, measured by LDH release, and also the release of IL-1β, an inflammatory cytokine not induced by phagosomally contained mutants. Studies on markers for host cell death revealed that the LVS ΔpdpC mutant induced mitochondrial instability, phosphatidylserine (PS) presentation, and TUNEL-specific DNA fragmentation in infected cells, rather similar to the wild-type strain, despite its lack of replication. This study reveals that the pdpC gene is an important gene required for F. tularensis virulence. We also show that non-replicating intracellular bacteria can induce host cell death, hypothesizing that release of bacterial components in the host cell cytosol is required for this induction. The FSC043 mutant showed a unique phenotype where a small subset of bacteria was able to escape the phagosome and replicate in the host cell. This was also seen in the pdpC deletion mutant of SCHU S4, but not with the LVS ΔpdpC. However, regardless of genetic background, the ΔpdpC mutant had an effect on phagosomal escape; either by affecting the phagosomal membranes in a unique way or by allowing phagosomal escape of a small proportion of the bacteria.