F. tularensis is capable of rapid escape from the phagosome, which is followed by efficient growth within the cytosol of monocytic cells. The molecular mechanisms behind the intracellular life style of the bacterium are not well understood, but have been shown to be dependent on many FPI-encoded genes, of which the most well-studied are the members of the iglABCD operon [16, 28, 37]. Evidence indicates that many of the FPI proteins collectively constitute a T6SS, however, while such systems have been identified in nearly 100 different bacterial species to date, their homologies to the FPI system are weak, indicating that the latter constitutes an evolutionarily distinct group [1, 14, 22]. While the FPI proteins IglA, IglB, PdpB, VgrG, and DotU show modest similarities to common components of T6SSs, the remaining FPI proteins appear to be unique and this makes it laborious and tedious to understand their roles and functions. The accumulating evidence indicates that many of them are essential core components and as such critically required and, thereby, their absence leads to a null mutant phenotype characterized by lack of phagosomal escape, no intracellular replication, and avirulence . A majority of the investigated FPI mutants appears to belong to this group but, in contrast, the ΔpdpE mutant exhibits full virulence . Not all of the investigated mutants so far fit into this polarized pattern, however, since the ΔiglI and ΔiglG mutants of LVS show delayed cytopathogenicity and a lack of virulence, although intact intracellular replication in some macrophage types .
In a recent study, we characterized the markedly attenuated FSC043 strain, a spontaneous mutant of the highly virulent strain SCHU S4, belonging to subspecies tularensis. Whole-genome sequencing revealed that only one deletion event and three point mutations discriminated the strains, two of which were identical single nucleotide deletions in each of the two copies of pdpC. Although one of the other mutated genes was fupA, which confers the most important contribution to the attenuation of LVS, we observed other features of the FSC043 strain that were distinct from those observed for a ΔfupA mutant and this led to our interest in understanding the role of PdpC . The present investigation reveals that the ΔpdpC mutant of LVS is another example of an FPI mutant with a very distinct and paradoxical phenotype, since it in some aspects mimics that of the LVS strain, whereas it in other aspects is very different since it does not fully escape into the cytosol, lacks intramacrophage replication, and is highly attenuated in the mouse model.
F. novicida strain U112 has been widely used to study the functions of the FPI, presumably since it harbors only one copy of the FPI and, thus, is more amenable to genetic manipulation and, moreover, does not require BSL3 containment. However, the results are not always in agreement when FPI mutants of F. tularensis and F. novicida are studied, as exemplified by our recent finding that iglI mutants of F. novicida and LVS show distinct phenotypes . Moreover, a recent study of F. novicida FPI mutants revealed that a ΔpdpC mutant showed normal intracellular replication in murine cells and also in insect cells and Drosophila melanogaster[39–41]. Our only explanation for the disparate results on the ΔpdpC mutants is that the functions of PdpC are distinct between the U112 strain of F. novicida and the LVS strain. In support of this hypothesis, there are 72 amino acids that discriminate the two proteins.
In view of the paradoxical phenotypes of ΔpdpC; lack of intracellular replication, but much more distinct cytopathogenic effects than the ΔiglC mutant, to some extent resembling those of the so called hypercytotoxic mutants that were recently identified by Peng et al. , we found an in-depth analysis of the physical properties of the mutant warranted. An additional rationale was that our bacterial fractionation assay revealed that PdpC predominantly is an inner membrane protein and the hypercytotoxic phenotype has been suggested to be caused by physical instability of mutants that, not surprisingly, are defective for important membrane proteins, or components of the LPS or O-antigens [25, 42]. This instability leads to bacterial lysis in the cytosol, which normally does not occur for the LVS or U112 strains. However, regardless of methodology used, we did not observe any similarity between the ΔpdpC mutant and the hypercytotoxic mutants and we therefore believe that the phenotype of the former depends directly or indirectly on the functions of PdpC and not on any physical instability of the mutant. The distinct expression of FPI proteins in the mutant was of interest in this regard, since the IglA, IglB, IglC, IglD, IglH, and VgrG proteins showed markedly lower expression and this was also reflected in lower transcription of the iglABCD operon. As most of these proteins play key roles for the virulence of the bacterium, their reduced expression may be important for the distinct phenotype of the mutant and, thereby, the contribution of PdpC to this phenotype may be indirect. One possible mechanism whereby such effects on protein levels could be mediated is via direct protein-protein interactions, however, our two-hybrid analysis revealed no such interaction for PdpC to any other FPI protein nor to any of the FPI regulatory proteins MglA, SspA, FevR, and PmrA. This indicates that one of the roles of PdpC is likely regulatory, but distinct from the MglA/SspA/FevR regulatory complex since this complex affects expression of all FPI proteins.
The findings on the ΔpdpC mutant illustrate certain caveats concerning methods to discern the intracellular localization of bacteria. A very widely used assay is based on the late endosomal and phagosomal marker LAMP-1, however, in the case of the ΔpdpC mutant, we conclude that the 75% co-localization we observed is not indicative of normal phagosomal entrapment, since the TEM analysis clearly indicated that almost all bacteria were surrounded by slightly or highly damaged membranes, thereby explaining the high degree of LAMP-1 colocalization. This phenotype was very distinct compared to the ΔiglC mutant, which was associated almost exclusively with intact membranes at similar time points.
The lack of intramacrophage replication was, not surprisingly, also reflected in a much attenuated phenotype in the mouse model, though the mutant was capable of limited systemic spread. However, the most paradoxical phenotype was that, despite its lack of intracellular replication, the mutant modulated the inflammatory response of the host cells in a way that was different from that of the ΔiglC mutant. An assay that clearly illustrates this distinction is secretion of IL-1β. We and others have shown that phagosomally contained mutants, e.g., ΔiglC, do not induce release of this cytokine [17, 19, 20, 22, 38], however, the ΔpdpC mutant showed much higher levels than ΔiglC. This indicates that the damage of the phagosomal membrane is a major trigger for the inflammasome activation. In view of the hypothesis by Peng et al., that the phenotype of the hypercytotoxic mutants is dependent on bacterial lysis in the cytosol , which does not occur for wild-type strains, our present data suggest that the lysis of the physically intact bacteria occurs in the phagosome and that the DNA that activates the AIM2 inflammasome is released when the phagosomal membrane is damaged, as is the case for the ΔpdpC mutant, but not for the ΔiglC mutant. This would explain the intermediate levels of IL-1β secretion induced by the ΔpdpC mutant. Another example of the potent immunomodulating effect of the ΔpdpC mutant was suppression of the E. coli LPS-induced TNF-α secretion, an inflammasome-independent event. We have previously concluded that there is a close relationship between the mitigation of the LPS-induced inflammatory response and the subcellular localization of F. tularensis. The ΔpdpC mutant adds to the understanding of this mechanism, since it, as the LVS strain, completely abrogated the TNF-α secretion. Thus, this phenotype is not related to intracellular replication, but only to the ability to disrupt the phagosomal membrane.
The findings reported herein demonstrate that the relationship between bacterial intracellular location and infection-mediated effects on host cell is not always straightforward and indicate that a key event in mediating the latter is the disruption of the phagosomal membrane and presumably the concomitant release of bacterial DNA and effector proteins of the T6SS and possibly other secretion systems. This situation is to some degree analogous to recently published data on mycobacteria. Although Mycobacterium tuberculosis and other mycobacteria are primarily considered to be vacuolar pathogens, it has become evident that the ESX-1 secretion system effectuates limited perforation of the phagosomal membrane, although the bacterium still remains within the phagosome. Recent publications demonstrate that this perforation results in mixing of phagosomal and cytoplasmic contents and induces a cytosolic host response triggered by bacterial DNA [43–45]. Thus, although the ultrastructural findings on the ΔpdpC mutant are distinct from those on mycobacteria, the bacteria-induced effects on the host cells are in both cases critically dependent on the permeabilization of the phagosomal membranes and leakage of DNA and, possibly, bacterial effectors into the cytosol.
Collectively, our data show that the ΔpdpC mutant distinctly modulates the interaction between F. tularensis and the phagocytic cell, since it shows incomplete phagosomal escape, lack of intramacrophage growth, intermediate cytopathogenic effects, and marked attenuation in vivo, but almost intact modulation of the macrophage inflammatory response. The unique phenotype of the mutant provides novel information, since it demonstrates that some of the cytopathogenic effects and modulation of host cell signaling is not dependent on bacterial replication, but only requires disruption of the phagosomal membrane. Therefore, further elucidation of the exact functions of PdpC will be important in order to understand the enigmatic mechanisms behind the intracellular life style of F. tularensis.