IglC and PdpA are important for promoting Francisella invasion and intracellular growth in epithelial cells

Abstract

The highly infectious bacteria, Francisella tularensis, colonize a variety of organs and replicate within both phagocytic as well as non-phagocytic cells, to cause the disease tularemia. These microbes contain a conserved cluster of important virulence genes referred to as the Francisella Pathogenicity Island (FPI). Two of the most characterized FPI genes, iglC and pdpA, play a central role in bacterial survival and proliferation within phagocytes, but do not influence bacterial internalization. Yet, their involvement in non-phagocytic epithelial cell infections remains unexplored. To examine the functions of IglC and PdpA on bacterial invasion and replication during epithelial cell infections, we infected liver and lung epithelial cells with F. novicida and F. tularensis 'Type B' Live Vaccine Strain (LVS) deletion mutants (ΔiglC and ΔpdpA) as well as their respective gene complements. We found that deletion of either gene significantly reduced their ability to invade and replicate in epithelial cells. Gene complementation of iglC and pdpA partially rescued bacterial invasion and intracellular growth. Additionally, substantial LAMP1-association with both deletion mutants was observed up to 12 h suggesting that the absence of IglC and PdpA caused deficiencies in their ability to dissociate from LAMP1-positive Francisella Containing Vacuoles (FCVs). This work provides the first evidence that IglC and PdpA are important pathogenic factors for invasion and intracellular growth of Francisella in epithelial cells, and further highlights the discrete mechanisms involved in Francisella infections between phagocytic and non-phagocytic cells.

Figure 1. Deletion of genes encoding IglC and PdpA perturb F. novicida invasion.

Murine BNL CL.2 hepatocytes were infected with wild-type F. novicida, deletion mutants (ΔiglC and ΔpdpA) as well as complement strains (ΔiglC::iglC and ΔpdpA::pdpA) for 3 h. Subsequently, samples were washed and treated with gentamicin for 1 h. At 4 h PI, lysates were plated onto agar-containing media and bacterial colonies were enumerated the following day. Error bars, S.E.M. (n = 4).

Figure 2. F. tularensis LVS ΔiglC and ΔpdpA mutants lung epithelial cell infections.

Bacteria were centrifuged onto human A549 cells and allowed to invade for 3 h. To determine the amount of invaded bacteria, gentamicin protection assay (invasion assay) was performed at 4 h PI and bacterial titre was measured after a 3-day incubation. Error bars, S.E.M. (n = 3).

Figure 3. F. novicida lacking either IglC or PdpA reduces liver epithelial cell colonization.

Samples were fixed at 24 h PI, differentially stained for intracellular and extracellular bacteria, and then visualized by fluorescence microscopy. The proportion of infected cells was tallied from over 1,000 cells. Cells containing one or more intracellular bacteria are considered ‘infected’. Error bars, S.E.M. (n = 3).

Figure 4. Intracellular growth kinetics of F. novicida mutants during hepatocyte infections.

BNL CL.2 cells were infected with wild-type F. novicida, deletion mutants (ΔiglC and ΔpdpA) and their respective complements. Bacteria were allowed to invade for 3 h after which extracellular bacteria were rapidly washed with PBS and killed with 100 µg mL⁻¹ of gentamicin for 1 h. Low concentrations of gentamicin (10 µg mL⁻¹) remained in the media (to inhibit extracellular bacteria) until experimental endpoint. Intracellular bacteria were then released by lysing host cells, diluted with TSBC, and plated for bacterial enumeration. Error bars, S.E.M. (n = 3).

Figure 5. Intracellular bacterial replication is severely compromised when genes encoding iglC and pdpA are deleted.

24 h and 48 h gentamicin protection assays were performed on liver BNL CL.2 cells infected with wild-type F. novicida, deletion mutants (ΔiglC and ΔpdpA), and their respective complement strains (ΔiglC::iglC and ΔpdpA::pdpA). Samples were then treated with gentamicin starting from 22 h post-inoculation until the experimental endpoint. After host cells were lysed, the released bacteria were diluted and plated for CFU enumeration. Error bars, S.D. (n = 4).

Figure 6. IglC and PdpA are essential for robust F. novicida growth within hepatocytes.

Phase and fluorescence microscopic images were taken of BNL CL.2 cells infected with wild-type F. novicida, deletion mutants (ΔiglC and ΔpdpA), and complement strains (ΔiglC::iglC and ΔpdpA::pdpA) for 48 h. At 22 h post-inoculation, the samples were washed with PBS and replaced with media containing gentamicin to prohibit further bacterial invasion. F. novicida (green) and DNA (blue, DAPI) were stained in the fixed samples. Each image represents a ‘maximum intensity’ Z-projection comprising a stack through the cell body. Images taken by fluorescence and phase microscopy were merged together to illustrate the cell borders. Scale bar = 10 µm.

Figure 7. During the late intracellular phase, IglC and PdpA are necessary for efficient proliferation in lung epithelial cells.

Human A549 cells were infected by wild-type F. tularensis LVS, ΔiglC, ΔpdpA, ΔiglC::iglC and ΔpdpA::pdpA. Intracellular bacteria were enumerated at 24 and 48 h time-points using gentamicin protection assay. At 24 h PI, the sample was switched to a low gentamicin concentration (10 µg mL⁻¹) in order to inhibit growth of extracellular microbes. Intracellular bacteria were titred after they were released from host cells and serial diluted onto agar-containing media. Error bars, S.E.M (n = 3).

Figure 8. Proportion of F. novicida associated with LAMP1 during murine hepatocyte infections.

(A) wild-type F. novicida, deletion mutants (ΔpdpA, ΔiglC) and complements strains (ΔpdpA::pdpA, ΔiglC::iglC) invaded BNL CL.2 cells for 3 h, after which extracellular bacteria were washed off and then killed with gentamicin (100 µg mL⁻¹). Subsequently, samples were exposed to low gentamicin concentration until the experimental endpoint was reached. Image stacks were assembled and used to determine the frequency of LAMP1-associated bacteria. For time-points 4, 8 and 12 h, between 30 and 50 intracellular bacteria were counted. For the 24 h time-point, more than 50 intracellular bacteria were counted. Error bars, S.E.M. (n = 3).