Francisella tularensis subsp. holarctica LVS was obtained from the Centers for Disease Control (Atlanta, GA, USA). All bacterial studies were carried out with approval from the University of Arizona Biosafety Committee. Bacteria were grown at 37°C on chocolate agar supplemented with 1% IsoVitalex (Becton Dickinson). Inocula were prepared from lawn grown Francisella. Bacteria were resuspended in sterile phosphate-buffered saline (PBS) at an optical density at 600 nm (OD₆₀₀) of 1, equivalent of 1 × 10¹⁰ CFU/mL. Innocula were diluted in sterile PBS to obtain the desired bacterial dose. Challenge doses were quantified by serial dilution and plating on chocolate agar.

Primary BMDMs were cultured from femurs as previously described (9). Following differentiation, non-adherent cells were removed by multiple washes with PBS and BMDMs were removed from plates by scraping. Cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (Atlas), l-glutamine (HyClone), sodium pyruvate (HyClone), and penicillin–streptomycin (Life Technologies). Medium was replaced with antibiotic-free medium 24 h prior to inoculation with Francisella.

A total of 2 × 10⁵ BMDM/well were seeded into 96-well flat bottom plates (~80% confluent) for Francisella intracellular growth assays and incubated 2 h to allow adherence to the plate. Cultures were inoculated with LVS at 1 or 100 bacteria per cell. Infection was facilitated by centrifugation at 300 × g for 5 min. Cells were incubated for 1 h with bacteria, and the medium was then removed. Fresh medium containing 50 μg/mL gentamicin (Sigma) was added to kill extracellular bacteria. One hour after gentamicin addition, medium was removed, and cells washed twice before the addition of fresh antibiotic-free medium. To determine intracellular growth, medium was removed at indicated time points post-infection, and 200 μL of PBS was added to the cultures. Cells were removed from the plate by vigorous pipetting. Cells were lysed by vortexing at maximal speed for 1 min. Serial 1:10 dilutions of the lysate were made and plated onto chocolate agar. The resulting colonies were counted 48 h later. For determination of reactive oxygen species (ROS), BMDM were harvested and processed with CellRox Green (ThermoFisher) per the manufacturers instructions.

C57Bl/6J (B6) and B6.129-Tlr2^tm1Kir/J (TLR2^−/−) mice were obtained from The Jackson Laboratories (Bar Harbor, ME) and bred in-house. B6 Slc11a1^G169 (Slc11a1⁺) and B6 Slc11a1⁺ Tlr2/4/9^−/− (Slc11a1⁺ Tlr2/4/9^−/−) were obtained from Dr. Gregory M. Barton (University of California, Berkley) and bred in-house. The B6-Slc11a1⁺ mice are an incipient congenic, having been backcrossed to B6 five generations (10). This results in mice that are approximately 97% B6 background genes (excepting those linked to Slc11a1). The triple knockout has previously described and backcrossed to B6 (10). F1 mice (Slc11a1^± Tlr2^−/−Tlr4^± Tlr9^±) were produced by crossing Slc11a1⁺ Tlr2/4/9^−/− with Tlr2^−/−. Loss of TLR2 expression was confirmed by flow cytometry. All animal protocols were approved by The University of Arizona IACUC.

For intranasal infection, mice were anesthetized with 575 mg/kg of body weight of tribromomethanol (Avertin; Sigma) administered intraperitoneally. Mice were then intranasally (i.n.) inoculated with various doses of LVS suspended in 25 μL PBS. For footpad infection, mice were restrained and inoculated with LVS suspended in 50 μL PBS in the right footpad. Mice were weighed daily following all inoculations. Mice were sacrificed if they lost more than 25% of their starting weight, as indicated in our University of Arizona IACUC protocol.

Spleens, livers, and lungs were homogenized in sterile PBS by using a Biojector (Bioject) as previously described (11). Tenfold serial dilutions were plated onto chocolate agar. The resulting colonies were counted 72 h later. The limit of detection (LOD) was 50 CFU per organ.

Lungs were perfused with PBS to remove blood and then finely minced. Minced lung was placed into 10 mL of digestion buffer containing 0.5 mg/mL collagenase I (Worthington Biochemical), 0.02 mg/mL DNase (Sigma), and 125 U/mL elastase (Worthington Biochemical) in RPMI 1640 (HyClone). Lungs were digested for 30 min at 37°C and then vigorously pipetted prior to filtering through a 100-μm filter. Red blood cells were lysed by using ammonium chloride–potassium carbonate lysis buffer. Viable cells were determined by trypan blue exclusion using a hemacytometer.

Serum and lung cytokines (IL-6, IL-10, MCP-1, IFN-γ, TNF-α, and IL-12p70) were determined using a cytometric bead array (CBA) Mouse Inflammation Kit (Becton Dickinson) according to the manufacturers instructions. For lungs, single cell suspensions were incubated in DMEM with antibiotics for 24 h and the supernatant was harvested and used for the CBA. Serum was analyzed undiluted.

The following directly conjugated antibodies were utilized for flow cytometry analysis: CD3 (clone 145-2C11; eBioscience), CD4 (clone GK1.5; Biolegend), CD8a (clone 53-6.7; BD Biosciences), CD11b (clone M1/70; Biolegend), CD11c (clone N418; Biolegend), CD19 (clone 6D5; Biolegend), F4/80 (clone BM8; Biolegend), and GR-1 (Ly-6G) (clone RB6-8C5; eBioscience). All antibodies were titrated on normal B6 splenocytes prior to use. Lung cells were prepared as above. After preparation of single cell suspensions, cells were blocked for 30 min at 4°C with 24G2 serum to block Fc receptors. Cells were then stained with indicated antibodies for 30 min at 4°C in the dark. Cells were then washed three times with PBS + 2%BSA and analyzed on a Becton Dickinson LSRII flow cytometer. FlowJo (Treestar) was used for all flow cytometry analysis.

Since Slc11a1 has been previously implicated in control of intracellular bacterial infections and has only been superficially tested in Francisella infection, we tested the ability of LVS to replicate in BMDMs. We cultured BMDM from both B6 and Slc11a1⁺ mice and confirmed their macrophage phenotype (CD11b⁺ F4/80⁺) by flow cytometry. The BMDMs were then infected with LVS at either a multiplicity of infection (MOI) of 1 (Figure 1A) or 100 (Figure 1B). Two MOIs were used to ensure adequate infection. At 1, 4, and 16 hours post-infection (hpi), macrophages were lysed and viable bacterial enumerated. One hour after infection there were slightly fewer LVS in Slc11a1⁺ than B6 BMDM at both 100 and 1 MOI. By 4 h, there was a decrease in the number of LVS recovered from the Slc11a1⁺ BMDM at an MOI of 100, at the MOI of 1 there was no detectable LVS at 4 h in Slc11a1⁺ BMDM. The B6 BMDM had little change in burden at MOI of 100 and limited growth at the MOI of 1 at 4 hpi. By 16 h, LVS was still undetectable in Slc11a1⁺ BMDM at an MOI of 1, at the MOI of 100 there was a 100-fold decrease in bacteria recovered compared to the B6 BMDM and 10-fold compared to the Slc11a1⁺ 1 h after infection (Figure 1). Taken together these data indicate that LVS is slightly impaired in its ability to invade Slc11a1⁺ BMDM, but more importantly is unable to proliferate after invasion.

Our in vitro experiments seemed to conflict with the earlier report that Slc11a1 increased susceptibility to LVS infection in vivo (5). Those experiments used sublethal footpad infection. The report of Kovarova used different mouse strains and we needed to test their route of infection with our strains. We therefore examined the effect of Slc11a1 expression on footpad infection. Slc11a1⁺ and B6 mice were inoculated with a sublethal dose of 1,100 CFU of LVS. On days 3 and 7 post-infection, these mice were sacrificed and total bacterial burden was determined for the lung, liver, and spleen. In contrast to the previous report (5) where there was increased bacterial load in Slc11a1⁺ mice, we observed no significant difference in the burdens between Slc11a1⁺ and B6 mice in any organ tested at either time point (Figure 2).

Since respiratory infection with Francisella alters the immune response compared to cutaneous infection, we sought to determine the role of Slc11a1 in intranasal Francisella infection (7). Slc11a1⁺ and B6 mice were infected with ~1,300 CFU LVS by the intranasal route. Mice were monitored for weight loss and survival and a subset were sacrificed at day 3 post-infection to determine organ burdens. Surprisingly, Slc11a1⁺ mice did not lose weight compared to their starting weight (Figure 3A). By contrast, B6 mice lost weight starting at day 3. At day 7, 2/5 surviving B6 mice had lost greater than 25% of their starting weight and were euthanized as required by our IACUC protocol. The mice who survived started regaining weight around day 10 (Figure 3A). In addition to maintaining weight, we were not able to recover any viable LVS from Slc11a1⁺ mice (LOD, 50 CFU) in any of the organs tested at day 3 post-infection. We recovered high levels of Francisella (10⁵–10⁶ CFU) from every organ tested in B6 mice (Figure 3B). These data demonstrate that Slc11a1 plays an important role in the resistance of mice to intranasal LVS infection. This is relevant to our previous results that demonstrated important differences in immune responses between intradermal and intranasal infection (7).

TLR2 has been reported to play a critical role in Ft resistance. TLR2-deficient mice are more sensitive to Ft infection that B6 mice (8, 12). Since SLC11A1⁺ B6 mice were resistant to intranasal LVS infection, we wanted to determine if Slc11a1 expression could compensate for the loss of TLR2. We reasoned that Slc11a1⁺ Tlr2/4/9^−/− mice would be at least as sensitive as TLR2 KO mice, as TLR2 has been strongly linked to resistance to Ft infection. We challenged Slc11a1⁺ Tlr2/4/9^−/− mice along with Slc11a1⁺ and B6 mice with ~1,300 CFU of LVS i.n. as above. Similar to earlier experiments (Figure 3), most Slc11a1⁺ mice yielded no detectable LVS in the organs (LOD, 20 CFU) while we recovered mice high levels of bacteria on day 3 that increased into day 5 from B6 mice (Figures 4A,B). Surprisingly, Slc11a1⁺ Tlr2/4/9^−/−, like Slc11a1⁺ mice, yielded 1,000-fold lower organ burdens than B6 in all organs tested, in spite of lacking TLR2/4/9 (Figures 4A,B). TLR4 has little role in Francisella pathogenesis as its unique LPS structure fails to bind TLR4. While EF-Tu and DNAK have been identified as ligands for TLR4, TLR4-deficient mice show no increase in sensitivity Francisella infection compared to B6 mice (13–16). TLR2 has been previously demonstrated to be important in survival of LVS infection in B6 genetic background mice. These experiments demonstrate that the presence of functional Slc11a1 can compensate for the TLR2 defect in resistance to Ft infection, although they are higher than Tlr2⁺ Slc11a1⁺ mice.

To examine the impact of functional Slc11a1 expression on the innate immune response to LVS infection, we measured cytokine production in both the serum and lung homogenates of Ft infected mice. B6, Slc11a1⁺, and Slc11a1⁺ Tlr2/4/9^−/− were i.n. infected with 1,300 CFU of LVS. Additional B6 and Slc11a1⁺ mice were infected with 1,000 CFU of LVS by the footpad route and serum cytokines measured. As a negative control, naïve mice received PBS by either the intranasal or footpad route. On day 3 and 5 post-infection, mice were sacrificed and cytokine production determined. Following the footpad infection, there were no significant differences between the B6 and Slc11a1⁺ mice (Figure 5). Both mouse strains produced high levels of IFN-γ, MCP-1, and IL-6. This is not surprising since they had similar LVS organ burdens in previous experiments (Figure 2). By contrast, following intranasal infection, the Slc11a1⁺ and Slc11a1⁺ Tlr2/4/9^−/− mice accumulate very little cytokine in either in serum (Figure 6) or lung (Figure 7) after intranasal infection. B6 mice had significantly increased levels of IFN-γ and elevated levels of MCP-1, and IL-6 in the serum compared to Slc11a1⁺ and Slc11a1⁺ Tlr2/4/9^−/− mice (Figure 6). Lung cultures from B6 mice produced significantly more IFN-γ, MCP-1, and IL-6 compared to Slc11a1⁺ and Slc11a1⁺ Tlr2/4/9^−/− cultures (Figure 7). The lack of cytokine response seen in Slc11a1⁺ mice demonstrates it is not an increased innate cytokine response that protects Slc11a1⁺ mice and mediates bacterial clearance.

Because we found so little cytokine in Slc11a1⁺ mice we hypothesized the cellular milieu in the lung would differ after infection. To determine the total cellularity of lungs B6, Slc11a1⁺, and Slc11a1⁺ Tlr2/4/9^−/− were i.n. infected with 1,300 CFU of LVS, naïve mice received PBS as a control. At day 3 and 5 post-infection, mice were sacrificed and lungs were harvested, the right lobe was used for CFU determination (Figure 4) and the left lobe was used for lung cellularity counts and flow cytometry (Table 1). The naïve mice had no significant changes in total lung cellularity among strains. Consistent with previously published data, B6 mice demonstrated a significant increase in total lung cellularity at both days 3 and 5 (17). Strikingly in Slc11a1⁺ mice, we detected no increased cellularity compared to naïve mice at either 3 or 5 days post-infection. The lack of cellular infiltrate in the Slc11a1⁺ is striking, but consistent with the low levels of bacteria present. We detected only a moderate increase of cellularity in the Slc11a1⁺ Tlr2/4/9^−/− mice at day 3 although still significantly lower than the B6 mice. By day 5, the cellularity of all Slc11a1⁺ Tlr2/4/9^−/− mice had decreased close to naïve levels although B6 levels remained high (Table 1).

To determine which cells are responsible for this increased cellularity, samples were stained for flow cytometric analysis. The gating strategy for this analysis is in Figure S1 in Supplementary Material. Both the total number and the percentage of macrophages, dendritic cells, neutrophils, T, and B lymphocytes were determined. In B6 mice, two cell types were significantly increased, in both number and percentage. Neutrophils (CD3⁻, CD19⁻, CD11b⁺, Gr-1⁺) and monocyte/macrophages (CD3⁻, CD19⁻, CD11b⁺, Gr-1⁻, F4/80⁻) (Table 1) were increased. There was no significant difference in any other population in either Slc11a1⁺ or Slc11a1⁺ Tlr2/4/9^−/− mice. This is consistent with Figure 1 showing the inability of Francisella to replicate in the myeloid cells of Slc11a1⁺ mice. Thus, the resistance to infection is not due to increased recruitment of innate cells, but rather a cell intrinsic property of the resident infected cells of Slc11a1⁺ mice.

One mechanism used by Francisella for controlling the innate immune response is the reduction of host produced ROS (18, 19). To examine the ability of Slc11a1⁺ cells to produce ROS in response to LVS infection, BMDMs were cultured from B6 and Slc11a1⁺ mice and infected. We measured ROS production every 10 min. Control cells were treated with 70% tert-butyl hydroperoxide as a positive control or left untreated. Control cultures were harvested at 50 min after the infection. Both untreated B6 and Slc11a1⁺ BMDM produced similar low levels of ROS. In response to the positive control, tert-butyl hydroperoxide, the Slc11a1⁺ BMDM made significantly more ROS than the B6 cells. As seen previously, the B6 cells made little ROS in response to LVS infection. Strikingly, the Slc11a1⁺ BMDM made significantly more ROS in response to LVS starting at 20 min after infection and increasing throughout the assay (Figure 8). Thus the expression of functional Slc11a1 overrides the inhibition of ROS production by Francisella presumably mediating better early pathogen control both in vitro and in vivo.

Because of the low bacterial proliferation in Slc11a1⁺ mice, we hypothesized that Slc11a1⁺ mice would be resistant to high-dose LVS challenge. In order to determine the extent of protection conferred by a functional Slc11a1—B6, Slc11a1⁺, Slc11a1⁺ Tlr2/4/9^−/−, and Slc11a1^± Tlr2^−/−Tlr4^± Tlr9^± mice were challenged i.n. with increasing doses of LVS and monitored for survival. B6 mice only received 8,700 CFU of LVS (~8 LD₅₀). All B6 mice succumbed to infection by day 8 as expected. Slc11a1⁺ mice received either 8,700 CFU (~8 LD₅₀) or 87,000 (~80 LD₅₀) of LVS. All mice (5/5) survived the 8,700 CFU challenge and 80% (4/5) survived the 87,000 CFU challenge. Slc11a1⁺ Tlr2/4/9^−/−, and Slc11a1^± Tlr2^−/−Tlr4^± Tlr9^± had similar survival with all mice surviving the 8,700 CFU challenge and 60% of Slc11a1⁺ Tlr2/4/9^−/−, and 100% of Slc11a1^± Tlr2^−/−Tlr4^± Tlr9^± survived the 87,000 CFU challenge (Figure 9). This resistance to even very high-dose challenges of LVS, even in the absence of TLR2, highlights the strong effect functional Slc11a1 has the outcome of LVS infection.