Skip to main content
NCBI home page
Infection and Immunity logoLink to Infection and Immunity
. 2019 Jul 23;87(8):e00198-19. doi: 10.1128/IAI.00198-19

Development, Characterization, and Standardization of a Nose-Only Inhalation Exposure System for Exposure of Rabbits to Small-Particle Aerosols Containing Francisella tularensis

Katherine J O’Malley a, Jennifer D Bowling a, Eileen M Barry b, Karsten R O Hazlett c, Douglas S Reed a,✉
Editor: Craig R Royd
PMCID: PMC6652755  PMID: 31085702

Inhalation of Francisella tularensis causes pneumonic tularemia in humans, a severe disease with a 30 to 60% mortality rate. The reproducible delivery of aerosolized virulent bacteria in relevant animal models is essential for evaluating medical countermeasures. Here we developed optimized protocols for infecting New Zealand White (NZW) rabbits with aerosols containing F. tularensis.

KEYWORDS: Francisella tularensis, aerosols, animal models, inhalation, rabbits, tularemia

ABSTRACT

Inhalation of Francisella tularensis causes pneumonic tularemia in humans, a severe disease with a 30 to 60% mortality rate. The reproducible delivery of aerosolized virulent bacteria in relevant animal models is essential for evaluating medical countermeasures. Here we developed optimized protocols for infecting New Zealand White (NZW) rabbits with aerosols containing F. tularensis. We evaluated the relative humidity, aerosol exposure technique, and bacterial culture conditions to optimize the spray factor (SF), a central metric of aerosolization. This optimization reduced both inter- and intraday variability and was applicable to multiple isolates of F. tularensis. Further improvements in the accuracy and precision of the inhaled pathogen dose were achieved through enhanced correlation of the bacterial culture optical density and the number of CFU. Plethysmograph data collected during exposures found that respiratory function varied considerably between rabbits, was not a function of weight, and did not improve with acclimation to the system. Live vaccine strain (LVS)-vaccinated rabbits were challenged via aerosol with human-virulent F. tularensis SCHU S4 that had been cultivated in either Mueller-Hinton broth (MHB) or brain heart infusion (BHI) broth. LVS-vaccinated animals challenged with SCHU S4 that had been cultivated in MHB experienced short febrile periods (median, 3.2 days), limited weight loss (<5%), and longer median survival times (∼18 days) that were significantly different from those for unvaccinated controls. In contrast, LVS-vaccinated rabbits challenged with SCHU S4 that had been cultivated in BHI experienced longer febrile periods (median, 5.5 days) and greater weight loss (>10%) than the unvaccinated controls and median survival times that were not significantly different from those for the unvaccinated controls. These studies highlight the importance of careful characterization and optimization of protocols for aerosol challenge with pathogenic agents.

INTRODUCTION

Francisella tularensis is a Gram-negative coccobacillus that causes a severe zoonotic disease in humans known as tularemia (also known as rabbit fever). Transmission of F. tularensis is by direct contact with infected animals, tissues, or arthropods, by ingestion, or by inhalation; tularemia is not typically spread from person to person. Inhalation of as few as 15 organisms is sufficient to cause disease; F. tularensis is readily grown to a high concentration in culture and relatively stable in aerosol, properties considered ideal in a biological weapon. For these reasons, both the former Soviet Union and the United States (prior to 1969) included F. tularensis in their offensive biological weapons programs (1). There are no licensed vaccines for tularemia; antibiotics are typically given and show good efficacy against most isolates of F. tularensis. Considering the high morbidity/mortality and infectivity by aerosol, as well as the possibility of the existence of antibiotic-resistant strains, F. tularensis is a tier 1 select agent; tier 1 select agents are those for which there is the most concern about malicious use (2).

We have previously shown that rabbits are a relevant model of the human disease caused by inhalation of a small-particle aerosol containing strain SCHU S4, the virulent prototype type A strain of F. tularensis. Very low doses caused fever, weight loss, dehydration, anorexia, and pneumonia in rabbits (3). Clinical findings included elevated erythrocyte sedimentation rates, lymphopenia, and thrombocytopenia. Radiographs confirmed the development of bacterial pneumonia, and naive rabbits died within 4 to 7 days after exposure. The cause of death in the rabbits appeared to be severe septic shock and/or acute respiratory distress syndrome. Additionally, we have shown that attenuated strains of F. tularensis can protect rabbits against aerosol challenge with virulent SCHU S4 (4–6). The degree of protection against morbidity and mortality is a function of the attenuated strain used, the number of vaccinations, and the route of vaccination. An aerosol prime/boost vaccination regimen with one particular SCHU S4-derived recombinant strain, S4 ΔaroD, protected 66 to 85% of rabbits against aerosolized SCHU S4 at robust challenge doses (20 to 200 50% lethal doses [LD50]) (4).

The incidence of natural tularemia, particularly for inhalation exposure, is so rare and the bacterium causes such severe disease that the licensure of vaccines or therapeutics for treating tularemia will be possible only via FDA’s Animal Rule (7, 8). The Animal Rule allows for the use of multiple animal models in pivotal efficacy studies when human studies are not ethically or logistically possible. These pivotal studies must be conducted under the auspices of good laboratory practices (GLP), which require substantial documentation with an implicit goal of ensuring rigor and reproducibility. In more recent guidance, FDA has created an Animal Model Qualification Program which allows animal models to be developed and qualified for use in licensing more than one product (7). Requirements include the condition that the route of challenge must be the same as the probable route of exposure in humans and that, as much as possible, well-characterized pathogens that are minimally passaged and of human origin should be used. For tularemia, as a bioweapon concern, the route of exposure most likely to be employed in a bioweapon attack is dispersal of the bacterium in a small-particle aerosol (mass median aerodynamic diameter [MMAD], ≤5 μm) that can penetrate deep into the alveolar regions of the lung (9).

Most in vivo research on tularemia has been with mice, although rats, rabbits, and nonhuman primates (NHP) have also been studied (10). The NHP and rabbit studies have all utilized aerosol delivery for respiratory challenge, whereas most of the rodent studies have used intranasal (i.n.) or intratracheal (i.t.) inoculations for respiratory challenge. While the i.n. and i.t. methods for inoculation of the respiratory tract are relatively easy to perform and do not require the expensive equipment required for aerosol exposures, substantial differences in deposition based on the methodology and the individual operator have been noted (11). Penetration into the alveolar region, in particular, is highly variable, and it is unclear whether the liquid delivered with the pathogenic agent might alter lung physiology, further complicating matters. Studies using aerosols of F. tularensis were employed extensively in the 1950s, 1960s, and 1970s (12–23). There has been renewed interest in F. tularensis since letters containing anthrax were sent in 2001.

We previously showed that brain heart infusion (BHI)-grown F. tularensis has both a higher and a more consistent aerosol performance than F. tularensis grown in Mueller-Hinton broth (MHB) or Chamberlain’s chemically defined medium (24). Aerosol performance is often measured by use of the spray factor (SF), the ratio of the aerosol concentration (typically measured by an impinger) and the nebulizer concentration; this allows for comparisons between exposures and between different chambers, environmental settings, etc. (25). For F. tularensis, relative humidity (RH) is important, with better performance being seen at a higher RH. Similar aerosol performance was seen between the live vaccine strain (LVS) and strain SCHU S4. The improvement in aerosol performance with BHI-grown F. tularensis did not appear to impact virulence in naive animals, as assessed by exposure of naive BALB/c mice to lethal doses of LVS. We report here our efforts to develop standardized protocols for aerosol challenge of rabbits and to demonstrate the reproducibility of these protocols. We have previously shown that rabbits are a relevant outbred model of human tularemia (3) and have a pulmonary architecture similar to that of humans (26). As part of these efforts, we report here that while the choice of culture medium for the aerosolized challenge agent did not affect the survival of unvaccinated rabbits, it did affect the survival of vaccinated rabbits. The latter observations mirror those reported for the intranasal challenge of vaccinated inbred mice (27). The implications of these findings and the impact on conducting pivotal vaccine efficacy studies for tularemia and other high-priority, aerosolized agents are discussed.

RESULTS

Environmental data and aerosol particle size.

During our initial characterization of the aerosol system used for the rabbit studies, we evaluated the environmental conditions inside the rabbit nose-only (RNO) exposure chamber. Table S1 in the supplemental material shows the mean, median, standard deviation, and coefficient of variation (CV) for relative humidity, temperature, airflow rates, and pressure during tularemia aerosols in the RNO exposure chamber. Even without supplemental humidification, a high relative humidity was achieved (75%), with little variability from day to day (CV = 0.12). Even when conditions were drier at the aerosol start, relative humidity increased rapidly once an aerosol was initiated and remained elevated for the remaining aerosol exposures on that day (Fig. 1A). Aerosol particle size was measured using 1-μm fluorescein beads as a surrogate, and the results are shown in Fig. 1B. The mass median aerodynamic diameter (MMAD) of the particles (3.05 μm) was larger than what would be expected for a Collision 3-jet nebulizer, although the geometric square deviation of 1.77 suggests a relatively homogeneous particle size, as is typical for the Collison nebulizer (28, 29). The larger-than-expected particle size was likely a function of the short distance between the Collison nebulizer and the RNO exposure chamber and of the high RH. However, 80% of the aerosol particles had a MMAD of less than or equal to 5 μm, so they would still be expected to reach the deep lung of the rabbit respiratory tract. Aerosol performance was compared using the spray factor (SF), the ratio of the aerosol concentration to the nebulizer concentration. For BHI-grown SCHU S4, SF was higher in the RNO exposure chamber without supplemental humidification than in BHI-grown SCHU S4 aerosols in a rodent whole-body (RWB) exposure chamber with supplemental humidification (Fig. 1C). We also evaluated SCHU S4 grown in MHB and found a similar improvement in SF when using the rabbit nose-only tower without supplemental humidification. The SF achieved with MHB-grown SCHU S4 in the RNO exposure chamber was more than a 1-log improvement over that achieved with MHB-grown SCHU S4 aerosolized into an RWB chamber with supplemental humidification (4.12 × 10−7 and 2.67 × 10−8, respectively) (24). However, the SF for MHB-grown SCHU S4 in the RNO exposure chamber was still significantly lower than that for BHI-grown SCHU S4 in the RNO exposure chamber (P < 0.0001). A higher SF is indicative of better recovery of the agent from an aerosol and a lower nebulizer concentration required to achieve the same desired dose (9, 25).

FIG 1.

FIG 1

Open in a new tab

Aerosol performance of the RNO exposure chamber. (A) Change in relative humidity at 5-s intervals over 10 consecutive 10-min aerosol exposures in the RNO exposure chamber with F. tularensis in BHI. (B) Aerosol particle size in the RNO exposure chamber, as measured by the aerodynamic particle sizer using 1-μm fluorescein microspheres. The solid black line is the relative mass of the particles, as measured by the concentration (left y axis) compared to the particle size (x axis); the dotted line is the cumulative percentage of aerosol particles (right y axis) compared to the particle size (x axis). GSD, geometric square deviation. (C) SF for F. tularensis SCHU S4 in the RWB and RNO exposure chambers after being grown overnight in either BHI or MHB. Asterisks indicate results that were significantly different (**, P < 0.05) or highly significantly different (***, P < 0.001) by one-way ANOVA with multiple comparisons.

Reproducibility of system performance.

We next sought to evaluate the variability and reproducibility of the RNO exposure chamber for aerosols of F. tularensis. Figure 2A shows the SF for multiple aerosol exposures (runs) conducted on 14 different days using BHI-grown SCHU S4, with the SF mean across all aerosols being 9.48 × 10−7. One-way analysis of variance (ANOVA) using Brown-Forsythe’s test found that there was a significant difference between the means for the different exposure days (P = 0.0065) (Table S2). Tamhane’s T2 multiple comparisons determined that of the 91 comparisons made between days, significant differences were found in only 12 of those comparisons (Table S3). We further compared whether SF varied from earlier runs to later runs, evaluating whether placement of the samples in an ice bath over the course of several hours affected SF. The results of this comparison are shown in Fig. 2B. As the results show, BHI-grown SCHU S4 had a remarkably stable SF across runs over the course of the day, with no significant difference from the mean being found using a one-way ANOVA with Brown-Forsythe’s test (P = 0.7637) as well as Tamhane’s T2 multiple comparisons. We determined whether the SF in the RNO exposure chamber varied between strains and isolates of F. tularensis using BHI-grown SCHU S4 and LVS as well as recombinant, attenuated mutants of SCHU S4 (Fig. 3). In agreement with what we had previously observed in an RWB chamber, there was no significant difference in SF for any of the isolates by one-way ANOVA with Brown-Forsythe’s test (P = 0.4076).

FIG 2.

FIG 2

Open in a new tab

Variation between aerosol exposures of F. tularensis SCHU S4. (A) The graph shows the SF for individual F. tularensis aerosol exposures (gray circles) on 14 different days, with the mean and standard deviation (lines and error bars, respectively) being shown for each day. The dotted line shows the mean SF (9.48 × 10−7) across all SCHU S4 aerosols. (B) The graph shows the SF for aerosol runs over multiple days; the values plotted are the individual SF for that run from different days with the mean and standard deviation. The dotted line shows the mean SF (9.48 × 10−7) across all SCHU S4 aerosols.

FIG 3.

FIG 3

Open in a new tab

No significant variation in SF between different isolates of F. tularensis. SCHU S4, LVS, or recombinant derivative isolates of SCHU S4 were grown in BHI and aerosolized in the RNO exposure chamber. The results shown are the individual SF for different exposures with the mean and standard deviation for each isolate. The dotted line shows the mean SF (9.48 × 10−7) for SCHU S4 aerosols.

Improving accuracy and precision in the challenge dose.

The most predominant concerns with aerosol exposure of animals to infectious agents are (i) achieving the desired target inhaled dose and (ii) ensuring that all animals receive comparable doses. The inhaled dose (also known as the presented dose) is calculated as the product of the aerosol concentration of the agent, the duration of the exposure, and the minute volume (the mean volume of air inhaled per minute) of the animal (25). Depending on the agent, the environmental parameters and the choice of culture medium for propagating the agent can affect the aerosol concentration, which impacts the inhaled dose. These parameters can be controlled, and their effects can be calculated. In our initial tularemia vaccine study done in rabbits, the mean inhaled challenge dose (11,070 CFU) delivered to the rabbits was 11-fold higher than the desired target dose of 1,000 CFU (Fig. 4A). Analysis of the aerosol parameters revealed a 2-fold better than predicted SF (a 7.8 × 10−7 actual SF versus a 4.4 × 10−7 predicted SF) but also a nebulizer concentration that was 4-fold higher than desired (a 9.8 × 105-CFU/ml actual concentration versus a 2.4 × 105-CFU/ml desired concentration).

FIG 4.

FIG 4

Open in a new tab

Improvements in inhaled dose based on improved determination of bacterial concentration using the OD600 within the linear range. (A) Aerosol doses in rabbits for the first vaccine study (gray circles are results for individual rabbits); the actual dose is on the y axis, while the target dose is on the x axis. (B) OD (x axis) and CFU (y axis) readings of different cultures of SCHU S4 grown in BHI used to determine the nebulizer concentration for the results shown in panel A. (C) Narrowed OD-to-CFU range to fit linear regression analysis; the resulting formula is shown on the graph. The r2 value was 0.99. (D) Aerosol doses in the subsequent vaccine study (gray circles are the results for individual rabbits); the actual dose is on the y axis, while the target dose is on the x axis.

Although there was variability in the SF from what we had predicted, the data suggested that a more precise estimation of the nebulizer concentration would improve our accuracy in achieving the target dose. The bacterial concentration in broth culture is often estimated using the optical density (OD) of the culture (for F. tularensis, the OD is read at 600 nm) and comparing that with the colony growth on agar. Typically, after an 18-h culture of SCHU S4 in BHI, the OD at 600 nm (OD600) of the culture is ∼0.9. In our determination of the OD-to-CFU ratio, this was beyond that portion of the curve that was linear, which was between 0.05 and 0.3 (Fig. 4B and C). We felt that making dilutions of the broth culture to get the OD within that range would produce the desired concentration.

In a subsequent repetition of the vaccine study, rabbits were split into 2 dose groups; one group was to receive 1,000 CFU, while the second was to receive 10,000 CFU. Twofold dilutions were made from the overnight broth culture until the OD600 fell within the linear range, which was used to then back-calculate the broth culture concentration. For the 1,000-CFU-target-dose group, the actual mean challenge dose was 1,025 CFU, while for the 10,000-CFU-target-dose group, the actual mean challenge dose was 9,038 CFU (Fig. 4D). For all subsequent vaccine studies, using this refined method for determining the broth culture concentration, challenge doses averaged 2.6-fold higher than the target dose (Fig. S2 and Table S4).

While this refined methodology for determining the bacterial concentration in the overnight broth culture thereby improved the average actual inhaled challenge dose, it did not improve the precision in dosing between animals, as evidenced by the range of inhaled challenge doses (coefficient of variation; Table S4). Since the nebulizer concentration and SF were within the range of the predicted values and the aerosol duration was fixed at 10 min, the only remaining variable was the respiratory function of the rabbits. Figure 5 shows the plethysmography data collected in real time during the rabbit aerosol challenges. Predictive formulas, such as Guyton’s, Alexander’s, and Bide’s formulas, are a function of body weight (30–32). As Fig. 5A shows, while there was a trend of increased minute volume with weight, for individual rabbits, body weight would not be particularly helpful in predicting the minute volume. Minute volumes between two rabbits of the same age and weight can vary by more than 1 liter/min. When the rabbit minute volume was plotted by aerosol exposure days, there was day-to-day variation, as shown in Fig. 5B, and these differences were significant by one-way ANOVA using Brown-Forsythe’s test (P < 0.0001). In the midst of these studies, we also evaluated inhalation of small-particle aerosols containing attenuated strains as a potential means of delivery for live vaccines (4). Using the data acquired during those studies, we evaluated whether repeated aerosol exposures of rabbits in the nose-only tubes would alter respiratory function (Fig. S3). There were significant variations by one-way ANOVA in breathing rate (P = 0.0437), tidal volume (P = 0.0217), and minute volume (P = 0.0086) across the three aerosol exposures. However, there was no trend across the three exposures, either positive or negative; therefore, acclimation did not appear to alter respiratory function in this circumstance.

FIG 5.

FIG 5

Open in a new tab

Averaged real-time minute volume in rabbits during aerosol exposures to F. tularensis. (A) Averaged minute volume (y axis) over 10-min aerosol exposures for individual rabbits (gray circles) compared to the baseline weight (x axis); the black line is the minute volume calculated from body weight using Guyton’s formula. (B) Averaged minute volume for individual rabbits on the day of exposure. The dotted line is the mean minute volume across all rabbits.

The challenge broth medium impacts the survival of LVS-vaccinated rabbits.

We recently reported that LVS-vaccinated, inbred mice survived better if they were challenged with MHB-grown SCHU S4 than BHI-grown SCHU S4 (27). Using the parameters defined above, we sought to evaluate whether the same pattern of vaccine-mediated protection would be seen in aerosol-challenged outbred rabbits. Rabbits were vaccinated with LVS using a prime-boost aerosol vaccination approach, with the boost given 14 days after the prime. Thirty days after the booster vaccination, rabbits were challenged with 500 CFU of SCHU S4 grown in MHB or BHI. LVS-vaccinated rabbits in both groups developed fevers after challenge, but those challenged with BHI-grown SCHU S4 had fever responses persist far longer than those challenged with MHB-grown SCHU S4 (Fig. 6A). The persistence and severity of the fever response in the BHI-grown SCHU S4-challenged LVS-vaccinated rabbits are further demonstrated in the heat map shown in Fig. 6B. On days 8 and 9 postchallenge, the median temperature for the LVS-vaccinated, MHB-grown SCHU S4-challenged rabbits (38.6°C and 38.7°C for days 8 and 9, respectively) was significantly different from that for the LVS-vaccinated, BHI-grown SCHU S4-challenged rabbits (40.8°C and 40.9°C for days 8 and 9, respectively) (P = 0.0040 for day 8 and P = 0.0023 for day 9). Mock-vaccinated rabbits developed fever at about the same time regardless of how SCHU S4 was cultured and succumbed at roughly the same time (between days 5 and 6). LVS-vaccinated rabbits in both groups lost weight after challenge, but beginning on day 7, weight started to recover in the MHB-grown SCHU S4-challenged rabbits, while BHI-grown SCHU S4-challenged rabbits continued to lose weight through day 15 (Fig. 6C). At day 20, the last date that weight was recorded, the BHI-grown SCHU S4-challenged rabbits stopped losing weight but did not begin to regain the weight that they had lost. Three of six LVS-vaccinated rabbits survived challenge with MHB-grown SCHU S4, while only one of six LVS-vaccinated rabbits survived challenge with BHI-grown SCHU S4 (Fig. 6D). The survival for LVS-vaccinated, MHB-grown SCHU S4-challenged rabbits was statistically significantly different from that for the historical controls (P = 0.0016). Although fewer LVS-vaccinated, BHI-grown SCHU S4-challenged rabbits survived, the survival for this group was also significantly different from that for the historical controls (P = 0.0038). The differences in survival between the two LVS-vaccinated groups were not significant (P = 0.5954).

FIG 6.

FIG 6

Open in a new tab

BHI-grown F. tularensis is a more rigorous challenge than MHB-grown F. tularensis for vaccinated rabbits. Rabbits were vaccinated by aerosol prime-boost delivery of LVS 2 weeks apart and then challenged 30 days after the boost with ∼500 CFU of SCHU S4 grown in MHB (black lines, A and C) or BHI (gray lines, A and C). (A) Mean body temperature by group with error bars for the standard deviation for the first 20 days postchallenge (x axis); (B) mean body temperature by group, including mock-vaccinated groups, as a heat map for the first 20 days postchallenge; (C) mean percent weight change by group with error bars for the standard deviation for the first 20 days postchallenge (x axis); (D) time to death in each vaccine and control group compared to historical mock-vaccinated controls challenged with equivalent doses of BHI-grown SCHU S4. The time to death for the LVS-vaccinated, MHB-grown SCHU S4-challenged rabbits was significantly different from that for the historical controls by one-way ANOVA (P < 0.05). The time to death for the other groups was not significantly different from that for the historical controls.

DISCUSSION

We report here our efforts to develop and characterize a standardized, reproducible method for aerosol challenge of rabbits with virulent F. tularensis. Using our protocol of growing F. tularensis for 2 days on cysteine heart agar (CHA) followed by an overnight broth culture in BHI, we found while there is some variability in SF between aerosol exposures, there is none between runs on a given exposure day. This consistency will be quite useful in pivotal efficacy studies, where standardization and reproducibility will be critical. Refining our ability to pinpoint the concentration of bacteria in our broth culture also allowed us to improve our accuracy in achieving the desired target dose, with only a 2.6-fold difference between the actual and the target challenge doses being achieved over 13 subsequent vaccine studies. The data reported here show the value in developing a well-characterized aerosol challenge system, particularly for vegetative bacteria.

Our protocol did not eliminate all variability, as there was some day-to-day variation in SF as well as other factors. There is the inherent error in the current methods used for quantifying bacteria via either spectrophotometry or growth on agar, which can add some variability. Our refinement in determining the broth culture concentration and the subsequent nebulizer concentration improved the accuracy of the mean actual doses obtained in relation to the desired target dose but did not improve precision between animals. Differences in the minute volume between individual animals would also contribute to variability when aerosols are for a fixed time duration (in our case, 10 min). Altering the duration of the exposure based on differences in respiratory function would correct for this variability, ensuring that all animals breathe in the same volume of experimental air and improving precisions. There are currently two possibilities for adjusting the duration of an aerosol. First, respiratory function can be measured immediately prior to the aerosol, and this value can be put into the aerosol system, adjusting the length of the aerosol. However, this cannot correct for differences in respiratory function that might occur during the aerosol. The second option is to measure respiratory function in real time and pass those data to the system controlling the aerosol, adjusting the duration of the aerosol on the fly. This has already been demonstrated for NHP using the same Biaera AeroMP system that we used (33), and we are currently evaluating a new version of the FinePointe software (Data Sciences International, St. Paul, MN) that would do the same for rabbits. The only caveat to this is that with the current system, rabbits would have to be done one at a time if the aerosol duration were tied to respiratory function. This would double the amount of time required to complete a vaccine challenge compared to that required with our current challenge system.

One argument against using the rabbit for tularemia vaccine studies had been that vaccination with LVS extended the time to death but not survival (5, 6, 10, 21). Our own prior studies are included in those findings, including the observation that oral inoculation or inhalation of LVS did extend the rabbit time to death more than subcutaneous inoculation, with one oral LVS-vaccinated rabbit surviving challenge (5). These data do fit with what has been reported in humans and monkeys, where oral and aerosol LVS vaccination offered better protection against subsequent respiratory challenge (17, 20, 34, 35). The data presented here demonstrate that the apparent failure of LVS vaccination to protect in the rabbit model was at least partly due to our use of SCHU S4 grown overnight in BHI broth for aerosolization. MHB is the predominant broth medium used for the propagation of F. tularensis, although modified casein partial hydrolysate (MCPH), Trypticase soy broth (TSB), and BHI are also used. To our knowledge, TSB and MCPH have not been as carefully evaluated for the propagation of SCHU S4 as BHI or MHB has been. We had originally chosen BHI, as F. tularensis grown in BHI had a higher, more consistent SF than F. tularensis grown in MHB, and the growth of F. tularensis in BHI did not appear to alter its virulence or pathogenicity in naive mice, although it did decrease the time to death (24, 36). BHI-grown F. tularensis has been shown to more closely resemble F. tularensis from infected macrophages than MHB-grown F. tularensis, as shown by protein and gene expression profiles (36). BHI-grown F. tularensis stimulated fewer proinflammatory mediators in naive macrophages than MHB-grown F. tularensis. Both electron microscopy and flow cytometry found that MHB-grown F. tularensis lacks structural integrity, whereas BHI-grown F. tularensis does not, and this could explain the differences in the stimulation of inflammatory responses in vitro and in vivo (36, 37). We recently reported that LVS-vaccinated mice survived low-dose challenge with MHB-grown SCHU S4 better than they survived challenge with comparable doses of BHI-grown SCHU S4 (27). Here, we did not see 100% survival of LVS-vaccinated rabbits when they were challenged with ∼500 CFU (∼25 LD50) of MHB-grown SCHU S4, and even those that did survive developed a fever and lost weight postchallenge. However, the severity and duration of both the fever and weight loss seen in LVS-vaccinated rabbits challenged with MHB-grown SCHU S4 were significantly less than those seen in LVS-vaccinated rabbits challenged with BHI-grown SCHU S4. These data suggest that BHI-grown SCHU S4 is a more rigorous challenge for demonstrating protection in vaccine studies than MHB-grown SCHU S4. Considering that we have previously shown that attenuated derivatives of SCHU S4 outperform LVS and protect rabbits well against both morbidity and mortality at even higher challenge doses of SCHU S4 grown in BHI, this further illustrates the superior performance of these attenuated derivatives as potential vaccine candidates (4–6). Both the survival of LVS-vaccinated rabbits and the consistent/higher SF from the aerosol characterization studies argue that BHI-grown F. tularensis should be used in pivotal F. tularensis efficacy studies to achieve a rigorous, reproducible aerosol challenge sufficient to demonstrate the efficacy of potential vaccine candidates in animals.

MATERIALS AND METHODS

Biosafety and regulatory information.

All work with live F. tularensis was conducted at biosafety level 3 (BSL-3) or animal biosafety level 3 (ABSL-3). For respiratory protection, all personnel wore powered air-purifying respirators (3M GVP-1 powered air purifying respirator with an L-series bump cap) or used a class III biological safety cabinet. Vesphene IIse detergent (1:128 dilution; Steris Corporation, Erie, PA) was used to disinfect all liquid wastes and surfaces associated with the agent. All solid wastes, used caging, and animal wastes were steam sterilized. Animal carcasses were digested via alkaline hydrolysis (Peerless Waste Solutions, Holland, MI). The University of Pittsburgh Regional Biocontainment Laboratory (RBL) is a Registered Entity with the CDC/USDA for work with tier 1 select agents. Work with recombinant F. tularensis strains was approved by the University of Pittsburgh’s Institutional Biosafety Committee.

Rabbits.

Young adult male and female New Zealand White (NZW) rabbits (Robinson Services, Inc.) were housed in the University of Pittsburgh RBL at ABSL-3 for the duration of the studies. All studies were performed under protocols approved by the University of Pittsburgh’s Institutional Animal Care and Use Committee. The University of Pittsburgh is fully accredited by the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC). Rabbits were monitored at least once daily prior to infection and at least twice daily after infection until clinical signs resolved or the animal became moribund. Prior to vaccination, IPTT-300 temperature/identifier chips (BioMedic Data Systems, Seaford, DE) were implanted subcutaneously. Body weight was recorded once in the morning, and body temperature was recorded twice daily. The body temperature was read using a DAS-7000 reader (BioMedic Data Systems). Rabbits that were determined to be moribund (i.e., rabbits with any of the following clinical signs: weight loss of ≥20%, body temperature of <34°C, no response to prodding, respiratory distress) were first anesthetized with isoflurane (2 to 5%) and then euthanized promptly by barbiturate overdose (100 mg/kg of body weight sodium pentobarbital given intravenously or intracutaneously).

Bacteria.

F. tularensis LVS or SCHU S4 were originally obtained from Gerald Nau (38) and the Dynport Vaccine Company, respectively, and were stored as single-passage stocks; the SCHU S4 isolate used is also available from the Biodefense and Emerging Infections (BEI) Repository, catalog number NR-10492 (3, 39). F. tularensis SCHU S4 recombinant derivatives, generated in the laboratories of Eileen M. Barry and Wayne Conlan, were shipped to the University of Pittsburgh for these experiments. F. tularensis was grown first on cysteine heart agar (CHA) for 2 days prior to growth overnight (18 h) in a shaker incubator at 200 rpm, using brain heart infusion (BHI) broth supplemented with ferric pyrophosphate and l-cysteine in baffled, vented polycarbonate Erlenmeyer flasks (3). After the exposures were completed, nebulizer and all-glass impinger (AGI) contents were quantified on CHA.

Aerosol exposures.

Exposures to aerosols of F. tularensis were conducted inside a class III biological safety cabinet (Baker Co., Sanford, ME) located inside the RBL as previously described using a Biaera AeroMP exposure system (Biaera Technologies, Hagerstown, MD) (3). Figure S1 in the supplemental material shows a technical schematic of the aerosol system used for these studies. Rabbits were exposed two at a time for 10 min in a nose-only exposure chamber (CH Technologies, Westwood, NJ) using a vertical discharge 3-jet Collison nebulizer controlled by the AeroMP system (Fig. S2A and B). Aerosol samples were collected in an all-glass impinger containing BHI and antifoam A. These samples were plated on CHA to determine the bacterial aerosol concentration. The presented or inhaled dose was determined as described previously as the product of the aerosol concentration, the minute volume (see “Plethysmography” below) of the rabbit, and the duration of the aerosol (40).

Plethysmography.

Respiratory function data were collected in real time during the aerosol exposures using Buxco XA or FinePointe software (Data Sciences International, St. Paul, MN) (Fig. S2C). A pneumotach was mounted on each nose-only tube and calibrated using a 10-ml syringe per the manufacturer’s instructions before the aerosol was started. The minute volume, the mean volume of air inhaled into the lungs per minute, was used to determine the inhaled dose. Data were exported from the software into an Excel spreadsheet and averaged for the duration of the exposure.

Aerosol particle size.

Mass median aerodynamic diameter (MMAD) was determined as a measure of particle size using an aerodynamic particle sizer 3321 (TSi, Shoreview, MN) attached to the nose-only exposure chamber; 1-μm fluorescein-coated beads in BHI broth were aerosolized as a surrogate for F. tularensis. Data were transmitted to the AeroMP and recorded at 2-min 30-s intervals during the exposure.

Statistical methods.

Data were collected and organized using spreadsheets in Microsoft Excel software; graphing and statistical analyses were conducted using Prism (version 8.0) software (GraphPad, San Diego, CA).

Supplementary Material

Supplemental file 1
IAI.00198-19-s0001.pdf (638.8KB, pdf)

ACKNOWLEDGMENTS

The research described herein was sponsored by National Institute of Allergy and Infectious Diseases, National Institutes of Health, grants U01 AI077909-01, R01 A102966-01A1, AI 123129, and AI 100138.

Footnotes

Supplemental material for this article may be found at https://doi.org/10.1128/IAI.00198-19.

REFERENCES

  • 1.Dennis DT, Inglesby TV, Henderson DA, Bartlett JG, Ascher MS, Eitzen E, Fine AD, Friedlander AM, Hauer J, Layton M, Lillibridge SR, McDade JE, Osterholm MT, O'Toole T, Parker G, Perl TM, Russell PK, Tonat K. 2001. Tularemia as a biological weapon: medical and public health management. JAMA 285:2763–2773. doi: 10.1001/jama.285.21.2763. [DOI] [PubMed] [Google Scholar]
  • 2.Federal Select Agent Program. 2011. Select agents and toxins list. HHS and USDA https://www.selectagents.gov/SelectAgentsandToxinsList.html. Accessed 23 August 2018.
  • 3.Reed DS, Smith L, Dunsmore T, Trichel A, Ortiz LA, Cole KS, Barry E. 2011. Pneumonic tularemia in rabbits resembles the human disease as illustrated by radiographic and hematological changes after infection. PLoS One 6:e24654. doi: 10.1371/journal.pone.0024654. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.O'Malley KJ, Bowling JL, Stinson E, Cole KS, Mann BJ, Namjoshi P, Hazlett KRO, Barry EM, Reed DS. 2018. Aerosol prime-boost vaccination provides strong protection in outbred rabbits against virulent type A Francisella tularensis. PLoS One 13:e0205928. doi: 10.1371/journal.pone.0205928. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Stinson E, Smith LP, Cole KS, Barry EM, Reed DS. 2016. Respiratory and oral vaccination improves protection conferred by the live vaccine strain against pneumonic tularemia in the rabbit model. Pathog Dis 74:ftw079. doi: 10.1093/femspd/ftw079. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Reed DS, Smith LP, Cole KS, Santiago AE, Mann BJ, Barry EM. 2014. Live attenuated mutants of Francisella tularensis protect rabbits against aerosol challenge with a virulent type A strain. Infect Immun 82:2098–2105. doi: 10.1128/IAI.01498-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.FDA. 2014. Guidance for industry: animal models—essential elements to address efficacy under the animal rule. FDA, Silver Spring, MD: http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM399217.pdf. [Google Scholar]
  • 8.FDA. 2002. New drug and biological drug products; evidence needed to demonstrate effectiveness of new drugs when human efficacy studies are not ethical or feasible. Final rule. Fed Regist 67:37988–37998. [PubMed] [Google Scholar]
  • 9.Roy CJ, Reed DS, Hutt JA. 2010. Aerobiology and inhalation exposure to biological select agents and toxins. Vet Pathol 47:779–789. doi: 10.1177/0300985810378650. [DOI] [PubMed] [Google Scholar]
  • 10.Lyons CR, Wu TH. 2007. Animal models of Francisella tularensis infection. Ann N Y Acad Sci 1105:238–265. doi: 10.1196/annals.1409.003. [DOI] [PubMed] [Google Scholar]
  • 11.Miller MA, Stabenow JM, Parvathareddy J, Wodowski AJ, Fabrizio TP, Bina XR, Zalduondo L, Bina JE. 2012. Visualization of murine intranasal dosing efficiency using luminescent Francisella tularensis: effect of instillation volume and form of anesthesia. PLoS One 7:e31359. doi: 10.1371/journal.pone.0031359. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Hood AM. 1961. Infectivity of Pasteurella tularensis clouds. J Hyg (Lond) 59:497–504. doi: 10.1017/S002217240003919X. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Cox CS. 1976. Inactivation kinetics of some microorganisms subjected to a variety of stresses. Appl Environ Microbiol 31:836–846. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Cox CS, Goldberg LJ. 1972. Aerosol survival of Pasteurella tularensis and the influence of relative humidity. Appl Microbiol 23:1–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Cox CS. 1971. Aerosol survival of Pasteurella tularensis disseminated from the wet and dry states. Appl Microbiol 21:482–486. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Hood AM. 1977. Virulence factors of Francisella tularensis. J Hyg (Lond) 79:47–60. doi: 10.1017/S0022172400052840. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Hornick RB, Eigelsbach HT. 1966. Aerogenic immunization of man with live tularemia vaccine. Bacteriol Rev 30:532–538. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Eigelsbach HT, Tulis JJ, McGavran MH, White JD. 1962. Live tularemia vaccine I.: host-parasite relationship in monkeys vaccinated intracutaneously or aerogenically. J Bacteriol 84:1020–1027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Saslaw S, Eigelsbach HT, Prior JA, Wilson HE, Carhart S. 1961. Tularemia vaccine study. II. Respiratory challenge. Arch Intern Med 107:702–714. doi: 10.1001/archinte.1961.03620050068007. [DOI] [PubMed] [Google Scholar]
  • 20.Eigelsbach HT, Tulis JJ, Overholt EL, Griffith WR. 1961. Aerogenic immunization of the monkey and guinea pig with live tularemia vaccine. Proc Soc Exp Biol Med 108:732–734. doi: 10.3181/00379727-108-27049. [DOI] [PubMed] [Google Scholar]
  • 21.Baskerville A, Hambleton P. 1976. Pathogenesis and pathology of respiratory tularaemia in the rabbit. Br J Exp Pathol 57:339–347. [PMC free article] [PubMed] [Google Scholar]
  • 22.Day WC, Berendt RF. 1972. Experimental tularemia in Macaca mulatta: relationship of aerosol particle size to the infectivity of airborne Pasteurella tularensis. Infect Immun 5:77–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Schricker RL, Eigelsbach HT, Mitten JQ, Hall WC. 1972. Pathogenesis of tularemia in monkeys aerogenically exposed to Francisella tularensis 425. Infect Immun 5:734–744. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Faith SA, Smith LP, Swatland AS, Reed DS. 2012. Growth conditions and environmental factors impact aerosolization but not virulence of Francisella tularensis infection in mice. Front Cell Infect Microbiol 2:126. doi: 10.3389/fcimb.2012.00126. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Roy CJ, Pitt MLM. 2012. Infectious disease aerobiology: aerosol challenge methods, p 65–80. In Swearengen JR. (ed), Biodefense research methodology and animal models, 2nd ed CRC Press LLC, Boca Raton, FL. [Google Scholar]
  • 26.Kamaruzaman NA, Kardia E, Kamaldin NA, Latahir AZ, Yahaya BH. 2013. The rabbit as a model for studying lung disease and stem cell therapy. Biomed Res Int 2013:691830. doi: 10.1155/2013/691830. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Holland KM, Rosa SJ, Kristjansdottir K, Wolfgeher D, Franz BJ, Zarrella TM, Kumar S, Sunagar R, Singh A, Bakshi CS, Namjoshi P, Barry EM, Sellati TJ, Kron SJ, Gosselin EJ, Reed DS, Hazlett K. 2017. Differential growth of Francisella tularensis, which alters expression of virulence factors, dominant antigens, and surface-carbohydrate synthases, governs the apparent virulence of Ft SchuS4 to immunized animals. Front Microbiol 8:1158. doi: 10.3389/fmicb.2017.01158. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Thomas RJ, Webber D, Sellors W, Collinge A, Frost A, Stagg AJ, Bailey SC, Jayasekera PN, Taylor RR, Eley S, Titball RW. 2008. Characterization and deposition of respirable large- and small-particle bioaerosols. Appl Environ Microbiol 74:6437–6443. doi: 10.1128/AEM.01194-08. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.May KR. 1973. The Collison nebulizer: description, performance and application. J Aerosol Sci 4:235–243. doi: 10.1016/0021-8502(73)90006-2. [DOI] [Google Scholar]
  • 30.Guyton A. 1947. Measurement of the respiratory volumes of laboratory animals. Am J Physiol 150:70–77. doi: 10.1152/ajplegacy.1947.150.1.70. [DOI] [PubMed] [Google Scholar]
  • 31.Alexander DJ, Collins CJ, Coombs DW, Gilkison IS, Hardy CJ, Healey G, Karantabias G, Johnson N, Karlsson A, Kilgour JD, McDonald P. 2008. Association of Inhalation Toxicologists (AIT) working party recommendation for standard delivered dose calculation and expression in non-clinical aerosol inhalation toxicology studies with pharmaceuticals. Inhal Toxicol 20:1179–1189. doi: 10.1080/08958370802207318. [DOI] [PubMed] [Google Scholar]
  • 32.Bide RW, Armour SJ, Yee E. 2000. Allometric respiration/body mass data for animals to be used for estimates of inhalation toxicity to young adult humans. J Appl Toxicol 20:273–290. doi:. [DOI] [PubMed] [Google Scholar]
  • 33.Bohannon JK, Honko AN, Reeder RJ, Cooper K, Byrum R, Bollinger L, Kuhn JH, Wada J, Qin J, Jahrling PB, Lackemeyer MG. 2016. Comparison of respiratory inductive plethysmography versus head-out plethysmography for anesthetized nonhuman primates in an animal biosafety level 4 facility. Inhal Toxicol 28:670–676. doi: 10.1080/08958378.2016.1247199. [DOI] [PubMed] [Google Scholar]
  • 34.Tulis JJ, Eigelsbach HT, Hornick RB. 1969. Oral vaccination against tularemia in the monkeys. Proc Soc Exp Biol Med 132:893–897. doi: 10.3181/00379727-132-34331. [DOI] [PubMed] [Google Scholar]
  • 35.Hornick RB, Dawkins AT, Eigelsbach HT, Tulis JJ. 1966. Oral tularemia vaccine in man. Antimicrob Agents Chemother (Bethesda) 6:11–14. [DOI] [PubMed] [Google Scholar]
  • 36.Hazlett KR, Caldon SD, McArthur DG, Cirillo KA, Kirimanjeswara GS, Magguilli ML, Malik M, Shah A, Broderick S, Golovliov I, Metzger DW, Rajan K, Sellati TJ, Loegering DJ. 2008. Adaptation of Francisella tularensis to the mammalian environment is governed by cues which can be mimicked in vitro. Infect Immun 76:4479–4488. doi: 10.1128/IAI.00610-08. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Singh A, Rahman T, Malik M, Hickey AJ, Leifer CA, Hazlett KR, Sellati TJ. 2013. Discordant results obtained with Francisella tularensis during in vitro and in vivo immunological studies are attributable to compromised bacterial structural integrity. PLoS One 8:e58513. doi: 10.1371/journal.pone.0058513. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Horzempa J, Carlson PE Jr, O'Dee DM, Shanks RM, Nau GJ. 2008. Global transcriptional response to mammalian temperature provides new insight into Francisella tularensis pathogenesis. BMC Microbiol 8:172. doi: 10.1186/1471-2180-8-172. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Hesselbrock W, Foshay L. 1945. The morphology of Bacterium tularense. J Bacteriol 49:209–231. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Roy CJ, Pitt LM. 2005. Infectious disease aerobiology: aerosol challenge methods, p 61–76. In Swearengen JR. (ed), Biodefense: research methodology and animal models. CRC Press LLC, Boca Raton, FL. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental file 1
IAI.00198-19-s0001.pdf (638.8KB, pdf)

Articles from Infection and Immunity are provided here courtesy of American Society for Microbiology (ASM)

RESOURCES