Growth conditions and environmental factors impact aerosolization but not virulence of Francisella tularensis infection in mice

Abstract

In refining methodology to develop a mouse model for inhalation of Francisella tularensis, it was noted that both relative humidity and growth media impacted the aerosol concentration of the live vaccine strain (LVS) of F. tularensis. A relative humidity of less than 55% had a negative impact on the spray factor, the ratio between the concentration of LVS in the aerosol and the nebulizer. The spray factor was significantly higher for LVS grown in brain heart infusion (BHI) broth than LVS grown in Mueller-Hinton broth (MHb) or Chamberlain's chemically defined medium (CCDM). The variability between aerosol exposures was also considerably less with BHI. LVS grown in BHI survived desiccation far longer than MHb-grown or CCDM-grown LVS (~70% at 20 min for BHI compared to <50% for MHb and CCDM). Removal of the capsule by hypertonic treatment impacted the spray factor for CCDM-grown LVS or MHb-grown LVS but not BHI-grown LVS, suggesting the choice of culture media altered the adherence of the capsule to the cell membrane. The choice of growth media did not impact the LD(50) of LVS but the LD(99) of BHI-grown LVS was 1 log lower than that for MHb-grown LVS or CCDM-grown LVS. Splenomegaly was prominent in mice that succumbed to MHb- and BHI-grown LVS but not CCDM-grown LVS. Environmental factors and growth conditions should be evaluated when developing new animal models for aerosol infection, particularly for vegetative bacterial pathogens.

Keywords: Francisella tularensis; aerosol exposure; mice; respiratory infection; tularemia.

Figure 1

Spray factor of LVS with different exposure chambers and nebulizers. F. tularensis strain LVS was grown in MHb overnight and then aerosolized into a whole-body or nose-only exposure chamber (A) or with a Collison nebulizer using a 1- or 3-jet nozzle into a nose-only exposure chamber (B). Aerosol samples were collected in an AGI and plated to determine aerosol concentration of LVS. Plots show the spray factor (ratio of aerosol concentration to nebulizer concentration) for individual aerosol exposures (circles and squares) and the median (line) value.

Figure 2

Effect of relative humidity on aerosolization of LVS. Relative humidity during exposures is monitored and logged by the AeroMP system. In (A), the graph shows the change in relative humidity over time using different nebulizer/exposure chamber combinations without using the humidification loop to artificially increase humidity. In (B), the graph shows spray factors for individual exposures (squares and triangles) with the median and standard deviations at two different relative humidity levels. In (C) the graph shows the spray factors for a number of runs at a range of relative humidity values. Data in (C) were analyzed in Prism using four-parameter logistic regression.

Figure 3

Growth of LVS in different liquid broth media. LVS was cultured for 2 days on CHAH prior to being cultured overnight in (A) MHb, (B) CCDM, and (C) BHI liquid culture media. Graphs show the change in optical density at 600 nm at different times after the culture was started (circles). Boltzmann sigmoidal semi-logistic regression was used to fit the line shown on each graph.

Figure 4

Impact of growth media on aerosolization of F. tularensis. Graphs show spray factor for F. tularensis after growth in different liquid culture media. (A) Shows the spray factor for individual aerosol exposures of LVS grown in MHb (circles), CCDM (squares), and BHI (triangles) along with the mean and standard deviation for each growth condition. Lines with (*) are statistically significant based on two-tailed Student's t-test. (B) Spray factors for individual aerosol exposures of LVS or SCHU S4 grown in CCDM or BHI along with mean and standard deviation for each growth condition.

Figure 5

Growth media impact on LVS desiccation survival time. Graph in (A) shows impact of liquid broth culture on survival time of LVS from desiccation on Whatman filter paper. Data shown is mean values for two experiments with the standard deviation. Graph in (B) shows spray factor for individual aerosol exposures and means of LVS cultures that were washed and incubated for 3 days in fresh media, 0.8 or 10% saline.

Figure 6

Survival of mice after aerosol exposure to LVS grown in different culture media. Mice were exposed to aerosols containing LVS grown in (A) MHb, (B) CCDM, and (C) BHI at a range of doses sufficient for determining the median lethal dose (n = 10 mice per dose). Graphs are Kaplan–Meier plots with doses shown in lower left.

Figure 7

Weight loss in mice after aerosol exposure to LVS grown in different culture media. Mice were weighed daily beginning the day of exposure and continuing through day 10 after infection. Mice were exposed to aerosols containing LVS grown in (A,D) MHb, (B) CCDM, and (C) BHI at a range of doses (n = 10 per dose). Graphs (A–C) show average daily weight lost from baseline (day 0) for each group; error bars show standard deviation; graph (D) shows individual weights for mice in second highest dose group (3982 cfu) exposed to LVS grown in MHb.

Figure 8

Clinical scoring of disease in mice after aerosol exposure to LVS grown in different culture media. Mice were scored daily beginning the day of exposure and continuing through day 10 after infection for changes in behavior and appearance. Mice were exposed to aerosols containing LVS grown in (A) MHb, (B) CCDM, and (C) BHI at a range of doses (n = 10 per dose). Graphs show average daily score for each group; error bars show standard deviation.

Figure 9

Bacterial load in tissues of mice after aerosol exposure to LVS grown in different culture media. Mice became moribund after LVS infection were euthanized and necropsied. The liver, lung, and spleen were removed, weighed, and then analyzed to determine the bacterial titer in each tissue. Graphs in (A,B) show weight for lungs (A) and spleens (B) at time of necropsy from individual mice infected with LVS grown in MHb, CCDM, or BHI as well as the mean for each group and standard deviation. Weights of tissues from naive, uninfected mice are also shown. Graph in (C) shows a box and whiskers plot for the bacterial titer from the lung, liver, and spleen of mice that succumbed to infection with LVS grown in MHb (white boxes), CCDM (light gray boxes), or BHI (dark gray boxes).