Francisella philomiragia Infection and Lethality in Mammalian Tissue Culture Cell Models, Galleria mellonella, and BALB/c Mice

Francisella (F.) philomiragia is a Gram-negative bacterium with a preference for brackish environments that has been implicated in causing bacterial infections in near-drowning victims. The purpose of this study was to characterize the ability of F. philomiragia to infect cultured mammalian cells, a commonly used invertebrate model, and, finally, to characterize the ability of F. philomiragia to infect BALB/c mice via the pulmonary (intranasal) route of infection. This study shows that F. philomiragia infects J774A.1 murine macrophage cells, HepG2 cells and A549 human Type II alveolar epithelial cells. However, replication rates vary depending on strain at 24 h. F. philomiragia infection after 24 h was found to be cytotoxic in human U937 macrophage-like cells and J774A.1 cells. This is in contrast to the findings that F. philomiragia was non-cytotoxic to human hepatocellular carcinoma cells, HepG2 cells and A549 cells. Differential cytotoxicity is a point for further study. Here, it was demonstrated that F. philomiragia grown in host-adapted conditions (BHI, pH 6.8) is sensitive to levofloxacin but shows increased resistance to the human cathelicidin LL-37 and murine cathelicidin mCRAMP when compared to related the Francisella species, F. tularensis subsp. novicida and F. tularensis subsp. LVS. Previous findings that LL-37 is strongly upregulated in A549 cells following F. tularensis subsp. novicida infection suggest that the level of antimicrobial peptide expression is not sufficient in cells to eradicate the intracellular bacteria. Finally, this study demonstrates that F. philomiragia is lethal in two in vivo models; Galleria mellonella via hemocoel injection, with a LD50 of 1.8 × 103, and BALB/c mice by intranasal infection, with a LD50 of 3.45 × 103. In conclusion, F. philomiragia may be a useful model organism to study the genus Francisella, particularly for those researchers with interest in studying microbial ecology or environmental strains of Francisella. Additionally, the Biosafety level 2 status of F. philomiragia makes it an attractive model for virulence and pathogenesis studies.


INTRODUCTION
Members of the Francisella genus are small, non-motile, Gramnegative coccobacilli of the gamma-proteobacteria class (Sjostedt, 2007). Francisella philomiragia was first identified in an ailing muskrat located in Utah approximately 50 years ago, following the discovery of its related species, F. tularensis . Mistakenly characterized as Yersinia philomiragia due to its 24% genomic homology with Y. pestis and similarities in morphology, it took 30 years to re-categorize the species as a member of the Francisella genus .
Francisella philomiragia has an affinity for aquatic environments which may increase its host species potential (Anda et al., 2001;Tarnvik et al., 2004). The natural range of F. philomiragia reflects its preference for aquatic environments as it is found near bodies of water, particularly brackish or salt water in the mainland United States Whipp et al., 2003;Berrada and Telford, 2010;Siddaramappa et al., 2012;Whitehouse et al., 2012). F. philomiragia may exist naturally by forming biofilms on exposed surfaces of the environment and infecting the aquatic amoeba, Acanthamoeba castellanii (Verhoeven et al., 2010).
Virulence factors in F. philomiragia have not been well studied in this species, but likely include proteins encoded by the Francisella Pathogenicity Island (FPI) and phospholipase C (Zeytun et al., 2012), similar to other members of the Francisella genus (Nano and Schmerk, 2007;Dai et al., 2010).
A related species, F. noatunensis (formerly named F. philomiragia noatunensis), is pathogenic to many fish and mollusk species, which inflicts negative economic and health effects on fisheries (Kay et al., 2006;Ostland et al., 2006;Mauel et al., 2007;Mikalsen and Colquhoun, 2009).  previously asserted that F. philomiragia subsp. noatunensis is a fish pathogen that is not lethal to mice and does not pose a threat to human health . However, in the time since that publication, F. noatunensis has been elevated to species level, which leaves the ability of F. philomiragia to infect mice in question and untested (Mikalsen and Colquhoun, 2009;Cowley and Elkins, 2011).
Near-drowning victims are susceptible to numerous bacterial infections due to the direct inoculation of the bacteria into the lungs (Ender and Dolan, 1997;Relich et al., 2015). F. philomiragia infections have been reported in otherwise healthy individuals via direct lung exposure resulting from near-drowning experiences in brackish or salty water or immunocompromised individuals with contact to contaminated water or fish Ender and Dolan, 1997;Cora et al., 2010;Kreitmann et al., 2015). Despite differences in genomic sequences (88% homologous to F. tularensis subsp. LVS and 84% to F. tularensis subsp. novicida and SchuS4; Zeytun et al., 2012;Davenport et al., 2014;Johnson et al., 2015), slightly different plasmids (Le Pihive et al., 2009), and reports that F. philomiragia does not cause disease in mice , some of these near drowning victims infected by F. philomiragia develop a severe pneumonic infection. This prompted further investigation on the similarity of F. philomiragia to F. tularensis subsp. novicida and subsp. LVS, strains related to virulent F. tularensis subsp.
SchuS4, and whether it may be an opportunistic pathogens in humans. This comparison was achieved through the use of in vitro experiments using cell models involved in tularemia infections (macrophages, lungs, and liver) and in vivo animal infection models (G. mellonella and BALB/c mice).
Infection Protocol for A549, J774A.1, HepG2, and U937 Cells Cells were infected at a multiplicity of infection (MOI) of approximately 500, as previously described (Han et al., 2008;Hegedus et al., 2008;Ahmad et al., 2010), with a 2-h preinfection and 1-h gentamicin pulse. Briefly, cells were seeded (10 5 /well) in a 48-well plate and allowed to attach overnight. After verifying successful cell attachment, culture media was gently removed and rinsed twice with culture media. Francisella strains were grown to mid-logarithmic phase, collected by centrifugation (10 min at 4000 × g, 4 • C), washed three times with 1x phosphate buffered saline (PBS), and diluted in serum-free DMEM to a verified bacterial concentration (CFU/mL). Dilutions of bacteria were used to infect each cell line at MOI = 500. Sets of three wells were prepared for each condition (n = 3). Characteristically, Francisella infects host cells inefficiently, despite its infectivity via multiple routes in animals and humans. Therefore, the standard MOI of 500 CFU was used to infect cells with the Francisella strains in order to achieve infection of most of the cells (Lai et al., 2001;Lai and Sjostedt, 2003). Cells were then incubated with bacteria at 37 • C, 5% CO 2 . After a 2-h incubation, well media was gently aspirated, washed twice with PBS, and treated with 50 µg/mL gentamicin in serum-free DMEM for 1 h to kill extracellular bacteria. Following the gentamicin pulse, cell media was gently aspirated, replaced with DMEM supplemented with 10% FBS and 5 µg/mL gentamicin, and allowed to incubate for 24 h at 37 • C, 5% CO 2 . (Lai et al., 2001;Han et al., 2008;Ahmad et al., 2010) Cells were lysed and plated on Chocolate agar for CFU determination.

Cytotoxicity Assay of Mammalian Tissue Cultured Cells Infected with Francisella
PrestoBlue Cell Viability Reagent (A-13261, Life Technologies, Carlsbad, CA, USA) was used according to the manufacturer's protocol. This reagent functions by using the reducing environment of the cell's cytosol to determine cell viability. The reagent contains a cell-permeable compound, which is blue in color. When added to viable cells, it encounters the reducing environment and modifies the reagent to become a red fluorescent, which can be detected by fluorescence or absorbance measurements. Briefly, reagent was added to infected cells 24 h post gentamicin-pulse at a 1:10 ratio. The reagent was incubated with cells at 37 • C for 2 h. Fluorescence was measured at excitation and emission spectra of 560 and 590 nm, respectively. Three wells were used per condition (n = 3). Data was averaged and a no-cell well was subtracted as background. Data was then plotted using GraphPad Prism 5 (GraphPad Software Inc., San Diego, CA, USA) to reflect the cytotoxic effect of Francisella species on eukaryotic cell lines.

EC 50 Antimicrobial Assays
Peptides used in this study were custom synthesized by ChinaPeptides Company (Shanghai, China) and had purities of ≥95% based on chromatographic analysis of the purified peptides. Antimicrobial activity (EC 50 ) assays of the antibiotic control levofloxacin, human cathelicidin LL-37, and murine cathelicidin mCRAMP were performed against F. philomiragia, F. tularensis subsp. novicida, and F. tularensis subsp. LVS as previously described (Amer et al., 2010). Briefly, 1 × 10 5 CFU/well of Francisella species were grown in BHI (pH 6.8), added to a sterile 96-well plate and incubated with serial dilutions of peptide or antibiotic in 10 mM phosphate buffer for 3 h at 37 • C. Dilutions were plated in triplicate on tryptic soy agar with 1% cysteine for 24 h; colonies were counted to determine survival (n = 3). This experiment was performed three independent times. Bacterial survival was calculated by a ratio of the number of colonies on each experimental plate to the average number of colonies on the control plates lacking peptide or antibiotic application. The EC 50 was determined using GraphPad Prism 5 (GraphPad Software Inc., San Diego, CA, USA) to plot the percent survival versus log of peptide or antibiotic concentration (log µg/mL) and fitting data to a standard sigmoidal doseresponse curve as previously described (Blower et al., 2015).

Murine Infection
BALB/c mice (Harlan, Frederick, MD, USA, five per group) were infected intranasally with 20 µL of the following concentrations of bacteria: 1 × 10 6 , 5 × 10 5 , 1 × 10 5 , 5 × 10 4 , 1 × 10 4 , 5 × 10 3 , or 1 × 10 3 CFU/20 µL. Mice were examined twice a day for signs of illness or death. Bacterial concentrations were verified via retrospective plating and counting of CFUs. Animal experiments were approved by and conducted in compliance with regulations of the Institutional Animal Care and Use Committee (Protocol # 0236) of George Mason University. All experiments were carried out in accordance with the National Research Council's Guide for the Care and Use of Laboratory Animals (2011)

Statistical Analysis
Antimicrobial EC 50 assays were performed in triplicate with n = 3 for each experiment, and representative experiments are shown. Standard deviations of the mean of each set are represented on each graph as error bars. Additionally the confidence interval (95%) is provided for EC 50 determinations to demonstrate statistical overlap of data. Student's t-test was performed and p values of p < 0.05 was considered statistically different.
The survival curves were performed with an n = 16 for G. mellonella and an n = 5 for BALB/c mice and were analyzed using the Mantel-Cox test, which is used to test the null hypothesis that survival curves are not different between groups. This test does not assume a normal distribution, allows for censored data, and is based off of the chi-squared test, which allows for a minimum of five samples.

RESULTS
During pulmonary tularemia infections, bacteria colonize the alveolar macrophages, the lungs, and the liver (Hall et al., 2007;Faron et al., 2015). F. philomiragia was evaluated to see if it infected cell lines representative of these systems in vitro: murine macrophage cells, J774A.1, human Type II alveolar epithelial cells, A549, and human hepatocyte-like cells, HepG2. These cell lines have been previously shown by us and others to be susceptible to infection by F. tularensis subsp. novicida and LVS (Qin and Mann, 2006;Han et al., 2008;Amer et al., 2010;Bradburne et al., 2013).
Francisella tularensis readily infects macrophages and proliferates within these cells (Anthony et al., 1991;Golovliov et al., 1997;Lai et al., 2001;Bolger et al., 2005). It is characteristic for Francisella replication to occur with little cytotoxicity until the cell becomes overburdened (at about 48 h post infection) and will experience cell death. F. philomiragia was found to be able to infect and proliferate in murine macrophages at 24 h to higher levels than what was seen for F. tularensis subsp. novicida and subsp. LVS (Figure 1A, p < 0.05). These results with the in vitro macrophage model suggest that F. philomiragia may be capable of infecting mammalian macrophages in vivo. Furthermore, these results suggest that F. philomiragia may be able to infect the alveolar macrophages in the lungs of near-drowning victims, which results in the clinical disease resembling tularemia that can afflict these patients.
Human alveolar epithelial cells are known to be infected by Francisella both in vitro and in vivo (Hall et al., 2007;Faron et al., 2015). Here, experiments utilizing A549 cells showed that F. philomiragia infects this cell type to a lesser extent than F. tularensis subsp. novicida but more than subsp. LVS ( Figure 1B, p < 0.05). The infection of this cell type suggests another potential mechanism by which the near-drowning infections in human could occur by this organism due to the direct inoculation of the lung.
Francisella infection of and proliferation in hepatocytes has been observed in human tularemia patients and animal models (Conlan and North, 1992;Lamps et al., 2004;Rasmussen et al., 2006;Ray et al., 2010). The fully virulent F. tularensis subsp. SchuS4 replicates well in cultured HepG2 cells (Qin and Mann, 2006). In these experiments, F. tularensis subsp. LVS infected human hepatocyte-like cells well, and replicated faster than F. tularensis subsp. novicida and F. philomiragia. However, by 24 h post infection, there were no differences between the bacterial burdens of the three Francisella species in HepG2 cells ( Figure 1C, p > 0.05). This suggests that the F. philomiragia infections could potentially lead to liver damage, consistent with a tularemia infection.
Francisella tularensis is said to be a "stealth" pathogen, promoting its intracellular survival by not causing high cytotoxicity, among other mechanisms (Sjostedt, 2006;Jones et al., 2014). It was previously shown that infections of A549 cells by F. tularensis subsp. LVS (at 500 MOI) for 24 h did not cause significant cytotoxicity, although CFU increased significantly (Han et al., 2008;Bradburne et al., 2013). This high MOI of 500 is standard for Francisella infection protocols, as it is not taken up into non-phagocytic cells readily (Lai et al., 2001;Lai and Sjostedt, 2003;Telepnev et al., 2003). These studies were expanded to all three strains of Francisella investigated here (F. tularensis subsp. LVS, F. tularensis subsp. novicida, and F. philomiragia) and J774A.1, A549, and HepG2 cells.
This study confirmed that F. tularensis subsp. LVS is not significantly cytotoxic toward A549 cells and, furthermore, it was found that F. tularensis subsp. novicida and F. philomiragia also displayed little cytotoxicity in this cell line at 24 h (Figure 2). HepG2s were also minimally affected by cytotoxic effects of F. tularensis subsp. LVS and displayed only 10 and 8% cytotoxicity from F. tularensis subsp. novicida and F. philomiragia infections at 24 h, respectively. The murine macrophage cell line, J774A.1, demonstrated greater susceptibility to the cytotoxic effects of Francisella, with all strains demonstrating about 33% cytotoxicity at 24 h (p < 0.05) consistent with previous reports (Lai et al., 2001;Lai and Sjostedt, 2003). However, additional cytotoxicity studies showed that F. philomiragia is highly cytotoxic to the human macrophage-like cell line, U937, (33%, p < 0.05) while F. tularensis subsp. LVS and subsp. novicida showed only 7 and 5% cytotoxicity, respectively. The differences between the FIGURE 2 | Cytotoxicity of Francisella strains to cultured mammalian cells. Cytotoxicity of F. philomiragia to cultured mammalian cells, compared to F. tularensis subsp. LVS and F. tularensis subsp. novicida.
human and murine macrophage cell lines are not yet understood in regard to Francisella infections. Previously, differences in Francisella intracellular replication have been noted between rat and murine macrophages (Anthony et al., 1991), however, other causes for the cytotoxicity differences other than species of origin are possible. These findings are consistent with the intracellular replication lifestyle of other Francisella species (Sjostedt, 2006).
Here, F. philomiragia was found to infect and replicate in the same cell types and have the same general level of cytotoxicity in those cell types as F. tularensis subsp. LVS and subsp. novicida, with the exception of the U937 cells (p < 0.05). The susceptibility of F. philomiragia to two antimicrobial peptides, LL-37, a human cathelicidin, and mCRAMP, a murine cathelicidin, known to be expressed by host cells and have killing activity against F. tularensis subsp. novicida and subsp. LVS was examined (Amer et al., 2010). This is important because host defense against Francisella infection relies not only on antibody production, but also on the response of the innate immune system (Allen, 1962;Metzger et al., 2007;Kirimanjeswara et al., 2008).
Francisella species, including F. philomiragia, are known to be highly susceptible to levofloxacin under MIC conditions (Nelson et al., 2010;Georgi et al., 2012), thus it was used as a control for the EC 50 antimicrobial assays. The EC 50 for levofloxacin of F. philomiragia is 0.0146 µg/mL (14.6 ng/mL; Figure 3A) while the F. tularensis subsp. LVS EC 50 is 0.00827 µg/mL (8.27 ng/mL) and the F. tularensis subsp. novicida EC 50 is 0.00843 µg/mL (8.43 ng/mL; Table 1). These values are statistically the same within the 95% confidence intervals (p > 0.05) and are consistent with the MICs previously reported (Georgi et al., 2012).
The sensitivity of F. philomiragia to cationic antimicrobial peptides has not been well studied. This organism is highly resistant to colistin and polymyxin B, which are cationic cyclic peptide antibiotics (Petersen et al., 2009;Stephens et al., 2016). It was previously demonstrated that expression of the human cathelicidin LL-37 in A549 cells is strongly induced by F. tularensis subsp. novicida infection (Amer et al., 2010). This is of interest as Francisella bacteria replicate directly in the cytosol of the infected cells (Wehrly et al., 2009), and thus the bacteria may be able to be killed by expression of these innate immunity peptides by the afflicted cell.
This increased resistance of F. philomiragia to cationic antimicrobial peptides could be due to differences in the LPS (Siddaramappa et al., 2012) or other surface properties of F. philomiragia compared to F. tularensis subsp. novicida or LVS perhaps due to differential expression of high-molecular weight carbohydrates in the "host-adapted" phenotype (Zarrella et al., 2011).
Galleria mellonella has been demonstrated to be a useful model for Francisella infection (Aperis et al., 2007;Ahmad et al., 2010;McKenney et al., 2012;Sprynski et al., 2014), thus a survival curve evaluating the mortality of G. mellonella during F. philomiragia infection was examined. As shown in Figure 4A, F. philomiragia injection is lethal to G. mellonella in a manner similar to F. tularensis subsp. LVS (Aperis et al., 2007;Ahmad et al., 2010). The LD 50 of F. philomiragia in G. mellonella is ∼1.8 × 10 3 CFU/mL or ∼18 CFU with an inoculation volume of 10 µL. For comparison, the LD 50 of F. tularensis subsp. novicida in G. mellonella is ∼1.2 × 10 2 CFU/mL or ∼1 CFU due to an inoculation volume of 10 µL (McKenney et al., 2012). The mean and median times to death for F. philomiragia are 2.79 and 3 days, respectively. Since F. philomiragia is able to infect G. mellonella similarly to other laboratory strains of Francisella, infection of BALB/c mice by the intranasal route of infection was tested for this organism.
BALB/c mice are a common experimental model for Francisella infections and they are susceptible to Francisella infection by the pulmonary route (intranasal or aerosol), among other routes; no studies have been reported for F. philomiragia infections of insect models, mice, rats, or marmosets Conlan et al., 2003;Cowley and Elkins, 2011). To conform to these standards and expand the in vivo results obtained with G. mellonella, a survival curve evaluating the lethality of F. philomiragia in BALB/c mice when delivered via intranasal administration (mimicking near-drowning experiences) was examined. As shown in Figure 4B, intranasal F. philomiragia is lethal to BALB/c mice with an approximate LD 50 of 3.45 × 10 3 CFU. This is very comparable to the intranasal F. tularensis subsp. LVS LD 50 (1 × 10 3 CFU) in the same species of mice but is higher than the 100 CFU intranasal LD 50 of F. tularensis subsp. novicida (Aperis et al., 2007).

DISCUSSION
Multiple reports of severe pneumonic infections of humans following near-drowning experiences Wenger et al., 1989) suggested that direct or large inoculation of F. philomiragia into the lung by this method is sufficient to allow for infection of normal, healthy human lungs, potentially via infection of the alveolar macrophages and/or lung epithelial cells. However, this organism is not generally regarded as a human pathogen and its ability to infect mammalian cells is generally uncharacterized. In addition, the highly related organism, F. noatunensis, was found to be unable to infect laboratory mice . Thus, F. philomiragia was compared to F. tularensis subsp. novicida and subsp. LVS regarding its ability to infect human and murine cells was further studied.
Francisella philomiragia was shown to be capable of infecting a murine macrophage cell line, J774A.1, which are commonly used for Francisella studies, with statistically higher levels (p < 0.05) than more commonly studied strains of Francisella (Hegedus et al., 2008;Pechous et al., 2008;Ahmad et al., 2010). Similarly, F. philomiragia was also found to infect a human Type II alveolar epithelial cell line, A549, at statistically higher levels than F. tularensis subsp. LVS. This is the first demonstration of F. philomiragia infecting Type II alveolar epithelial cells and is a significant contribution to the understanding of the potential interactions of F. philomiragia within the human lung. These findings suggest a potential mechanism by which near-drowning in brackish water known to contain F. philomiragia (Ottem et al., 2007) could potentially lead to infection through interaction of the bacteria with Type II alveolar epithelial cells of the lung and/or alveolar macrophages (Gentry et al., 2007;Hall et al., 2007;Craven et al., 2008;Faron et al., 2015). Furthermore, these results suggest that aerosol exposure to F. philomiragia could potentially lead to pulmonary infections in humans if inhaled via an aerosol. Given the wide distribution of F. philomiragia, in particular its known presence in various bodies of water within the United States, this potential route of infection should be further investigated.
In addition, it was demonstrated that F. philomiragia infects HepG2 cells, a human hepatocyte-like cell line. This finding suggests that F. philomiragia may be able to replicate in the liver in infected near-drowning victims. The liver is one of the main organs infected by F. tularensis strains and liver failure following overwhelming organ infection is thought to be the primary cause of death in mice suffering from tularemia (Conlan and North, 1992). Patients suffering from F. philomiragia pneumonia should be closely observed for sequelae similar to those found in tularemia infections caused by F. tularensis species.
Cytotoxicity data after 24 h of infection show that F. philomiragia is similar to F. tularensis subsp. novicida and subsp. LVS in most of the studied cell lines. Little cytotoxicity was seen in A549 cells (∼0%, similar to other species) and HepG2 cells (8%, more than subsp. LVS but similar to subsp. novicida), and moderate cytotoxicity in J774A.1 cells (32%, similar to other species). However, the U937 human macrophage-like cell line only showed high cytotoxicity (33%) from F. philomiragia and not the other Francisella strains studied. This observation will be the subject of future investigation to understand the difference in U937 susceptibility.
Susceptibility testing using the antimicrobial peptides LL-37, a human cathelicidin, and mCRAMP, a murine cathelicidin, showed that these peptides were highly active in vitro against F. philomiragia. Despite being active in killing the bacteria in vitro, this antimicrobial peptide host defense mechanism is clearly insufficient to control F. philomiragia infections in infected cells or in vivo.
Francisella philomiragia was found to be lethal for both in vivo models tested: G. mellonella and BALB/c mice. G. mellonella has been demonstrated to be a useful in vivo model for Francisella infection (Aperis et al., 2007;Ahmad et al., 2010;McKenney et al., 2012;Sprynski et al., 2014), thus the survival of G. mellonella during F. philomiragia infection was examined. In G. mellonella, F. philomiragia was shown to be fatal in concentrations similar to F. tularensis subsp. LVS, with an LD 50 of 18 CFU. This similarity supports the ability of G. mellonella to be used as an effective model for Francisella infection but also suggests that F. philomiragia is capable of infecting a range of hosts similar to other Francisella strains.
Francisella philomiragia is not generally regarded as a pathogen of humans or animals but is considered an environmental species of the genus (Anda et al., 2001;Tarnvik et al., 2004;Verhoeven et al., 2010). In some cases, F. philomiragia infections in near-drowning victims individuals are observed Ender and Dolan, 1997). An intranasal infection of mice by F. philomiragia was used to mimic lung exposure seen in drowning victims and test the susceptibility BALB/c mice to this organism. F. philomiragia was shown to be fatal in BALB/c mice by intranasal-delivered inoculum concentrations similar to F. tularensis subsp. LVS, with an LD 50 of 3.45 × 10 3 CFU; however, this is significantly higher than the 10 CFU LD 50 seen with F. tularensis subsp. novicida. Thus, contrary to the result for F. noatunensis ), F. philomiragia is able to infect laboratory mice. These results call for further studies to determine the full host range of F. philomiragia.

CONCLUSION
These studies show that F. philomiragia results in similar in vitro and in vivo infections to the F. tularensis subspecies novicida and LVS for the evaluated strains. It was demonstrated for the first time that there is potential for significant and robust F. philomiragia infection in macrophages, lung, and liver cells. F. philomiragia infection of human alveolar epithelial cells and macrophages suggests a mechanism for infection in the lungs of near-drowning patients. The high level of F. philomiragia intracellular replication in all three cell types suggests that F. philomiragia follows an infection course similar to tularemia caused by F. tularensis subspecies. It was previously demonstrated that infections of G. mellonella and pulmonary infections of BALB/c mice were fatal with similar LD 50 s to F. tularensis subsp. LVS. The results of these in vitro and in vivo experiments confirm earlier suggestions that F. philomiragia may be an emerging opportunistic human pathogen (Mailman and Schmidt, 2005;Sjodin et al., 2012) and that cellular and animal models of Francisella infection could also be used to study F. philomiragia. It would be of interest to evaluate all the available F. philomiragia strains for their ability to infect the various tissue culture and murine models.
It was found that F. philomiragia is comparable to the other Biosafety level 2 strains of Francisella in many respects but unusual in its effect on human U937 cells. This finding will open some interesting new avenues of research regarding pathogenesis and virulence of F. philomiragia. In addition, this work also positions F. philomiragia as another important organism in the field of Francisella research, especially for researchers interested in questions of microbial ecology or environmental persistence of members of the genus Francisella.

AUTHOR CONTRIBUTIONS
All authors listed have made substantial, direct and intellectual contribution to the work, and approved it for publication. MLV conceived the study; MLV and CNP wrote the manuscript; CNP, SLP, RJB, SA, and MM contributed experimental data and contributed to the manuscript.