A Francisella novicida Mutant, Lacking the Soluble Lytic Transglycosylase Slt, Exhibits Defects in Both Growth and Virulence

Abstract

Francisella tularensis is the causative agent of tularemia and has gained recent interest as it poses a significant biothreat risk. F. novicida is commonly used as a laboratory surrogate for tularemia research due to genetic similarity and susceptibility of mice to infection. Currently, there is no FDA-approved tularemia vaccine, and identifying therapeutic targets remains a critical gap in strategies for combating this pathogen. Here, we investigate the soluble lytic transglycosylase or Slt in F. novicida, which belongs to a class of peptidoglycan-modifying enzymes known to be involved in cell division. We assess the role of Slt in biology and virulence of the organism as well as the vaccine potential of the slt mutant. We show that the F. novicida slt mutant has a significant growth defect in acidic pH conditions. Further microscopic analysis revealed significantly altered cell morphology compared to wild-type, including larger cell size, extensive membrane protrusions, and cell clumping and fusion, which was partially restored by growth in neutral pH or genetic complementation. Viability of the mutant was also significantly decreased during growth in acidic medium, but not at neutral pH. Furthermore, the slt mutant exhibited significant attenuation in a murine model of intranasal infection and virulence could be restored by genetic complementation. Moreover, we could protect mice using the slt mutant as a live vaccine strain against challenge with the parent strain; however, we were not able to protect against challenge with the fully virulent F. tularensis Schu S4 strain. These studies demonstrate a critical role for the Slt enzyme in maintaining proper cell division and morphology in acidic conditions, as well as replication and virulence in vivo. Our results suggest that although the current vaccination strategy with F. novicida slt mutant would not protect against Schu S4 challenges, the Slt enzyme could be an ideal target for future therapeutic development.

Keywords: Francisella; Francisella novicida; cell division; cell morphology; lytic transglycosylase; peptidoglycan (PG); tularemia; virulence.

FIGURE 1

Francisella novicida Slt homology model. (A) Amino acid sequence of the mature F. novicida Slt protein; U domain, linker (L) domain, and catalytic (C) domain are annotated in green, gray, and blue, respectively. (B) Homology model of F. novicida Slt onto the structure of Slt70 from E. coli. U, L, and C domains are color-coded according to sequence shown in (A). Model was generated using Cn3D software v4.3.1 based on alignment of F. novicida Slt with the crystal structure of Slt70 from E. coli (PDB accession 1QSA). The helix cylinders are color-coded according to the domains shown in (A). The amino acid backbone is rendered according to identity; identical residues in red and dissimilar residues in blue.

FIGURE 2

Francisella novicida slt exhibits pH-dependent growth defects. F. novicida WT, slt, and slt-C were grown in BHI broth (A) and Chamberlain’s defined medium adjusted to pH 7 (B), 6.2 (C), or 5 (D), and monitored by both optical density and plating for CFU at specific time points. OD_600nm readings are shown as points with connecting lines, while CFU/mL at 0, 8, and 16 h growth are shown as bars. Results represent average with standard error of the mean based on quadruplicate well measurements. These data are representative of at least three experiments. CFU counts were compared using one-way ANOVA applied to log-transformed data; counts for the slt strain were significantly different from WT at 0, 8, and 16 h in all conditions tested (red bars). OD values were fitted to a logistic growth equation, and fold-change was determined by two-way ANOVA; ^∗P ≤ 0.01, slt compared to WT and slt-C, ^∗∗P ≤ 0.001, slt compared to WT and slt-C.

FIGURE 3

Scanning electron microscopy analysis of Fn strains. Fn WT, slt, and slt-C strains were grown to exponential phase in CDM adjusted to pH 6.2 (top panels; A–C), or pH 7.0 (bottom panels; D–F), and analyzed by SEM. Images were acquired under 10,000× magnification; scale bar, 1 μm.

FIGURE 4

Surface structure and morphology of Fn strains. Fn slt cells grown in CDM pH 6.2 show significantly ruffled surface structure and fusion of cells (B, arrowheads), as well as tube-like surface projections (B, arrows). Equal magnification images of Fn WT (A) and Fn slt-C (C) strains grown in CDM pH 6.2 show smooth surface appearance of cells and uniform appearance. Images were acquired under 30,000× magnification; scale bar, 0.5 μm.

FIGURE 5

Live/Dead staining analysis of Fn WT and slt. Cells were grown in CDM adjusted to either pH 5.0 or 7.0 and subjected to Live/Dead staining and confocal microscopy at 0 and 6 h time points. (A) Representative confocal images of WT and slt strains grown in CDM pH 5.0. (B) The percentage of viable cells was calculated as number of green-stained cells divided by the total number of cells, and then used to determine the percent change in viability from 0 to 6 h of growth. The mean and standard error were plotted for three independent experiments, wherein at least 100 cells were counted for each sample. Significance was determined by student’s t-test; ^∗∗P < 0.01 for slt pH 5.0 compared to each other sample.

FIGURE 6

Replication of F. novicida in J774.1 murine macrophages. Macrophages were infected with WT or slt strains at an MOI of ∼100:1 and incubated in the presence of gentamicin to eliminate extracellular bacteria. At 4 and 24 h time points, cells were lysed and plated for recovered bacteria, and percent increase of bacteria from 4 to 24 h was calculated. Results from three independent experiments performed in triplicate wells are shown, and statistical significance was determined by student’s t-test; ^∗P < 0.05.

FIGURE 7

Fn slt is attenuated during intranasal infection. Groups of BALB/c mice (n = 10) were challenged with 5 doses of (A) Fn WT, (B) Fn slt, and (C) Fn slt-C strains, and monitored for 21 days. LD₅₀ values from these experiments are included in Table 2.

FIGURE 8

Dissemination of F. novicida WT and slt strains during intranasal challenge. BALB/c mice were infected intranasally with 42 CFU WT or 52 CFU slt mutant. At days 1, 2, 3, 4, 7, 14, and 21, mice were euthanized and lungs and spleens were harvested. Serial dilution and plating of homogenized lung and spleen were performed to determine recovery of bacteria. Results are shown as CFU per organ based on 5 mice per time point, except for day 4 WT; only four Fn WT-challenged mice remained and the rest had succumbed to infection. Error bars represent standard error of the mean; ^∗∗P < 0.001, average CFU/spleen days 2, 3, 4; ^∗P < 0.05 average CFU/lung days 1, 2, 4, as assessed by t-test. WT-infected mice were only carried out until day 4, based on previous experiments showing a median time until death of 5 days for WT F. novicida.

FIGURE 9

Fn slt provides protection against challenge with Fn WT. Mice surviving the initial Fn slt exposure (n = 10, and n = 6 for the 620 CFU group) were challenged after 28 days with 35 CFU of the Fn parent strain.

FIGURE 10

Robust Th17- and Th2-type cytokine response in splenocytes re-stimulated with Fn slt or Ft Schu S4. BALB/c mice were infected intranasally with mock/PBS (n = 5), 10 (n = 5), 94 (n = 5), and 935 (n = 3) CFU Fn slt mutant. At day 21 post infection, mice were euthanized and spleens were harvested. Splenocytes (10⁶ cells) were re-stimulated for 48 h in the presence of (A) irradiated Fn slt (5 μg/ml) or (B) Ft Schu S4 (5 μg/ml) and the levels of cytokines measured by the Luminex bead-based suspension assay. Graph and Table values are expressed as a fold change compared to PBS control group for each analyte. Red text indicates cytokine profiles which markedly differed between Fn slt and Ft Schu S4 re-stimulation. ^∗Th17-like cytokines; ^∗∗Th2-like cytokines. A P-value of <0.05 was determined for every analyte by t-test of log10 values against PBS only controls.

FIGURE 11

Serum IgG, IgG1, and IgG2a antibody titer levels by ELISA. Sera were collected from BALB/c mice vaccinated with mock/PBS (n = 5), 10 (n = 5), 94 (n = 5), and 935 (n = 3) CFU Fn slt mutant. The reactivity of (A) total IgG, (B) IgG1 and IgG2a in serum was tested against Fn slt or Ft Schu S4 by endpoint ELISA. Statistical significance was determined by t-test on log10 values compared against PBS only control; ns, not significant, ^∗P < 0.5, ^∗∗P < 0.01, ^∗∗∗P < 0.001.

FIGURE 12

Fn slt does not protect against re-challenge with Ft Schu S4. Groups of mice receiving 10, 94, and 935 CFU of Fn slt vaccine or PBS only (n = 10) were challenged 28 days post-vaccination with either a high dose (37 CFU) or low dose (6 CFU) of Ft Schu S4 via the intranasal route. Log rank comparison and survival regression analyses identified no significant difference between vaccinated and control groups.