|Home | About | Journals | Submit | Contact Us | Français|
Previous studies have demonstrated that systemically administered immunotherapy can protect mice from systemic challenge with the bacterial pathogen Francisella tularensis. However, for protection from inhalational challenge with this bacterium, we wondered if mucosally administered immunotherapy might be more effective. Therefore, we administered cationic liposome–DNA complexes (CLDC), which are potent activators of innate immunity, intranasally (i.n.) and assessed the effectiveness of protection from lethal inhalational challenge with F. tularensis. We found that pretreatment by i.n. administration of CLDC 24 h prior to bacterial challenge elicited nearly complete protection of BALB/c mice from lethal challenge with F. tularensis LVS strain. We also observed that mucosal CLDC immunotherapy provided a statistically significant increase in survival time in mice challenged with the highly virulent F. tularensis Schu4 strain. Protection was associated with a significant reduction in bacterial burden in the lungs, liver, and spleen. Mucosal administration of CLDC elicited significantly increased expression of IL-12, IFN-γ, TNF-α, IFN-β and IFN-α genes in the lung as detected by real-time quantitative PCR. In vitro treatment of F. tularensis infected macrophages with CLDC-elicited cytokines also significantly suppressed intracellular replication of F. tularensis in infected macrophages. In vivo, depletion of NK cells prior to administration of CLDC completely abolished the protective effects of CLDC immunotherapy. CLDC-elicited protection was also dependent on induction of IFN-γ production in vivo. We conclude therefore that activation of local pulmonary innate immune responses is capable of eliciting significant protection from inhalational exposure to a virulent bacterial pathogen.
Francisella tularensis is a Gram-negative facultative intracellular bacterium that causes the dangerous and occasionally fatal disease tularemia. While cases of infection in humans are relatively rare, concern over the use of F. tularensis as a biological weapon has brought renewed interest in this pathogen . Inhalation of as few as 10 organisms has been shown to cause an acute and fatal disease in mice . The high infectivity and virulence of inhaled F. tularensis has resulted in the organism being classified as a category A priority pathogen. Despite the importance of the inhaled route of infection, immune responses associated with protection from pneumonic tularemia are not well understood .
Two F. tularensis subspecies cause the majority of human infections: tularensis and holarctica, also referred to as ‘type A’ and ‘type B,’ respectively. A third subspecies novicida is attenuated in humans but is studied as a model pathogen of mice. A live vaccine strain (LVS) was derived from F. tularensis ssp. holarctica during the 1940s and 1950s by sequential passage on agar plates followed by passage in mice . LVS is attenuated in humans but retains virulence for mice, although it is less virulent in mice than wild-type A and B strains. Because LVS causes a disease in mice which mimics virulent disease in humans, it has been studied extensively as a model intracellular pathogen [5,6]. However, LVS is not licensed as a vaccine for widespread use due to concerns regarding immune reactivity and a lack of information on the molecular basis for attenuation. Thus the lack of an effective licensed vaccine has resulted in a need for alternative immunotherapeutic approaches to prevent pneumonic tularemia.
Though F. tularensis is susceptible to treatment with several different antibiotics, non-specific immunotherapeutics that stimulate mucosal innate immunity for defense against F. tularensis would be desirable for several reasons. Appropriate innate immune responses have the potential to provide immediate and potent defense at mucosal sites of infection. In addition, such innate immune responses are often pathogen non-specific and capable of providing protection against a broad range of different organisms. This is particularly desirable in a biodefense situation in which the identity of an outbreak-causing pathogen may not be known. The potential for stimulation of innate immunity to provide host defense against F. tularensis was demonstrated by Elkins et al. who found that parenteral administration of immunostimulatory CpG oligonucleotides protected mice from intraperitoneal challenge with F. tularensis LVS [7,8]. However, to provide defense against respiratory infection, it may be advantageous and more effective to stimulate mucosal as opposed to systemic immune responses. For example, two research groups demonstrated that IL-12 administered intranasally to mice 24 h prior to challenge provided IFN-γ-dependent protection from lethal i.n. challenge with F. tularensis LVS or F. tularensis ssp. novicida [9,10]. However, repeated administration of high doses of recombinant cytokines is not likely to be an effective or easily administered option for immunotherapy of inhaled bacterial infections. In addition, a synthetic TLR4 agonist administered intranasally to mice provided protection from F. tularensis ssp. novicida but did not address whether this agonist could provide protection from F. tularensis type A and B strains . Thus, it appears possible to control or prevent pneumonic tularemia using mucosal immunotherapy. However, an effective immunotherapeutic must be easily and safely administered and must demonstrate efficacy against F. tularensis type A and B strains.
Therefore, we evaluated the use of an immunotherapeutic (cationic lipid–DNA complexes) with a demonstrated record of safety in multiple mammalian species including humans (J. Fair-man, personal communication) that was capable of potently and broadly activating host innate immunity with potential to be applied to mucosal immunotherapy for F. tularensis. Previous studies have shown that combining cationic liposomes with nucleic acids (e.g., with CpG ODN, plasmid DNA, or synthetic RNA) markedly increases immune stimulation when administered systemically or as a vaccine adjuvant [12–15]. Moreover, we have recently demonstrated that mucosally administered CLDC can protect mice from pathogenic Burkholderia species . Therefore, we wondered if administration of CLDC to the airways might be effective as an immunotherapeutic for pneumonic tularemia.
To address this question, we administered CLDC immunotherapy to mice and assessed the effectiveness of protection from inhalational challenge with F. tularensis using infection models with F. tularensis LVS and the highly virulent F. tularensis ssp. tularensis Schu4 strain. In particular, we investigated the influence of route of administration and timing of CLDC administration on protection from pneumonic tularemia. In addition, immunological mechanisms of protection were investigated, with an emphasis on mucosal mechanisms of protection. We report here that mucosally administered immunotherapy can be used to effectively prevent or delay infection with virulent F. tularensis. Thus, mucosal immunotherapy may provide an additional option for non-specific control or prevention of inhaled pneumonic bacterial pathogens.
Specific pathogen-free 6–8-week-old BALB/cJ, 129SvEv, and ICR mice were purchased from the Jackson Laboratory (Bar Harbor, ME), Taconic (Hudson, NY), and Harlan (Indianapolis, IN), respectively. IFN-γ−/− mice in BALB/cJ background were purchased from the Jackson Laboratory. Mice lacking the type I interferon receptor (IFN-α/βR−/−) on the 129SvEv background were kindly provided by Dr. Phillipa Marrack (National Jewish, Denver, CO) and Dr. Michel Auget (Swiss Institute for Biomedical Research, Lucerne). All studies involving mice were approved by the Institutional Animal Care and Use Committee at Colorado State University.
Cationic liposomes (100 mM DOTIM lipid + cholesterol) in 10% sucrose solution were provided by Juvaris BioTherapeutics (Pleasanton, CA). Cationic liposome–DNA complexes (CLDC) were freshly prepared for each experiment as follows. Liposomes were diluted 1:5 in sterile Tris-buffered 5% dextrose water (pH 7.4). Non-coding plasmid DNA (pMB75.6 empty vector, 3 mg/ml) was then added to a final concentration of 0.2 mg/ml causing spontaneous formation of CLDC. For intranasal administration of CLDC, mice were deeply anesthetized by intraperitoneal injection of 100 mg/kg ketamine (Fort Dodge Animal Health, Overland Park, KS) with 10 mg/kg xylazine (Ben Venue Labs, Bedford, OH) and 20 µl CLDC was rapidly placed in sequential droplets on alternating nares, allowing mouse to inhale the droplets. Intraperitoneal and subcutaneous injection of CLDC was accomplished with 200 µl injected with an 18-gauge needle. Subcutaneous injection was given in the skin dorsal to the cervical region of the mouse spine. Intravenous injection of 100 µl 0.5X (half of the lipid and DNA concentration) CLDC was given through the lateral tail vein.
Frozen stocks of F. tularensis LVS and strain Schu4 were prepared as previously described [17,18]. Bacteria were thawed just before use, diluted in PBS and kept on ice prior to challenge. For intranasal challenge, mice were first deeply anesthetized with 100 mg/kg ketamine plus 10 mg/kg xylazine administered intraperitoneally. Then 20 µl of diluted bacteria was placed in sequential droplets on alternating nares, allowing inhalation of the droplets. The 20 µl challenge inoculum contained either 105 colony forming units (CFU) F. tularensis LVS or 40 CFU Schu4. All mice were monitored twice daily for disease symptoms and euthanized at a pre-determined humane endpoint.
Lungs, spleen and liver were collected and homogenized in 5 ml sterile PBS using a stomacher (Teledyne Tekmar, Mason, OH). Bacterial counts were determined by plating 100 µl of serial 10-fold dilutions of organ homogenate on modified Mueller–Hinton agar and incubating at 37 °C with 5% CO2 for 72 h. The lower limit of detection using this method was 50 CFU per organ.
Individual mouse tissues were placed in 1 ml Trizol Reagent (Invitrogen, Carlsbad, CA) and frozen at −80°C. Tissues were thawed and homogenized using a Polytron PT10-35 rotorstator homogenizer (Kinematica, Lucerne, Switzerland). RNA was extracted following the manufacturer’s instructions for Trizol Reagent. Any remaining DNA was then eliminated by treatment with DNase I Amplification Grade (Invitrogen) using 4U in a 100 µl volume. RNA preparations were further purified using the RNeasy Mini Kit (Qiagen, Valencia, CA) RNA cleanup protocol and quantified using a SmartSpec 3000 Spectrophotomer (Bio-Rad, Hercules, CA) at OD260.
For qRT-PCR, 4 µg RNA was reverse transcribed to cDNA using the SuperScript III First-Strand Synthesis Kit for RT-PCR (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions using the provided oligo dT to prime reverse transcription. Cytokine cDNA was quantified using specific primers and FAM dye-labeled Taq-Man probes purchased from Applied Biosystems Inc., Foster City, CA. Primer and probe sequences for IFN-α4 and IFN-β were published previously . All other primers and probes were designed by Applied Biosystems Inc. Reactions were prepared with Taqman Universal PCR Master Mix (Applied Biosystems Inc.) and run in 96-well format on an iCycler (Bio-Rad, Hercules, CA). Cytokine mRNA expression was normalized to abundance of the hypoxanthine guanine phosphoribosyl transferase 1 (HPRT1) housekeeping gene to control for different levels of mRNA recovery between samples. The ΔΔCt method of relative quantification was used to calculate fold change in transcript abundance of treated samples compared to untreated samples.
NK cells were systemically depleted from mice by intraperitoneal administration of 50 µl anti-asialo-GM1 antibody (Wako, Osaka, Japan) diluted to 200 µl in sterile PBS. Anti-asialo-GM1 antibody was administered 24 h prior to CLDC treatment and then administered again 5 days later to assure continued NK cell depletion. A control group of mice received i.p. injection of 50 µg ChromPure rabbit IgG antibody (Jackson ImmunoResearch, West Grove, PA) diluted to 200 µl in PBS.
Spleen supernatants used for in vitro experiments were produced using ICR mice. Mice were given 200 µl CLDC by intravenous injection. At 3 h post-treatment, mice were euthanized and spleens were removed. Single-cell suspensions were prepared from spleens by mechanical disruption and screening through a 100 µm nylon mesh screen (BD Biosciences, San Jose, CA) followed by NH4Cl lysis. Cells were resuspended in 24-well plates at 5 × 106 cells per ml in complete MEM (Invitrogen, Carlsbad, CA) containing 10% FBS (Gemini Bio-Products, West Sacramento, CA), 2mM l-glutamine (Invitrogen), 1X non-essential amino acids (Invitrogen), 0.075% sodium bicarbonate (Fisher Scientific, Pittsburgh, PA), 100U/ml penicillin, and 100 µg/ml streptomycin (Invitrogen). After 20 h incubation at 37 °C, supernatants were removed, centrifuged to remove cells, combined and stored in frozen aliquots at −80 °C. Each aliquot was thawed for one use and then discarded. For some experiments IFN-γ and TNF-α in spleen supernatant were neutralized by the addition of anti-IFN-γ (clone R4.6A2) or anti-TNF-α (clone TN3-19.12) antibodies (eBioscience, San Diego, CA). Isotype antibodies for anti-IFN-γ (clone eBRG1) and anti-TNF-α (clone eBio299Arm) were used as controls (eBioscience). Antibodies were incubated at a concentration of 10 µg/ml in spleen supernatant at 37 °C for 30 min prior to treatment of cells.
Mouse alveolar macrophage AMJ.2 cells were pre-treated prior to infection by plating 200,000 cells per well of a 24-well plate in antibiotic-free complete media (see previous paragraph) containing dilutions of CLDC or control spleen supernatant. Cells were incubated 16 h at 37°C in 5% CO2, supernatants were removed and cells were washed with 1 ml sterile PBS. To infect AMJ.2 cells, 500 µl antibiotic-free media containing F. tularensis LVS at a multiplicity of infection (MOI) of 50 CFU per cell was added to each well. Plates were centrifuged at 2400 × g for 2 min and then incubated at 37 °C for 1.5 h. Infection medium was removed and cells were washed twice with sterile PBS. To remove remaining extracellular bacteria, 1 ml of media containing 50 µg/ml gentamicin (Sigma–Aldrich, St. Louis, MO) was added and cells were incubated for 1 h at 37 °C. The supernatant was then removed, cells were washed with sterile PBS and 1 ml media containing 10 µg/ml gentamicin was added to cells to prevent extracellular bacterial growth during incubation. Cells were incubated for 48 h at 37 °C in 5% CO2 to allow for intracellular growth of F. tularensis LVS. To quantify intracellular bacteria, cells were washed three times with sterile PBS and then lysed in 1 ml sterile de-ionized H2O. Cell lysates were serially diluted 10-fold in PBS, plated on modified Mueller-Hinton agar and incubated for 72 h at 37 °C in 5% CO2. In addition, to account for any remaining extracellular bacteria, the final PBS wash fluid was also serially diluted and plated. Intracellular F. tularensis LVS CFU/well was calculated by subtracting the final PBS wash CFU/well from the cell lysate CFU/well.
All statistical analyses were conducted using Prism v. 5.0 (Graph-Pad Software, La Jolla, CA). Kaplan–Meier survival and log rank analysis was used to compare survival times. In survival analyses involving comparisons of more than 2 groups, the Bonferroni corrected threshold was applied to determine the appropriate level of significance. For experiments involving comparison of two groups, Student’s two-tailed t test was used for statistical analysis. Comparison of multiple treatment groups was performed using ANOVA, followed by Tukey’s multiple means comparison test. Serial mouse weight measurements were analyzed by two-way ANOVA with repeated measures and Bonferroni post-tests. Group differences were considered statistically significant for p <0.05.
We used a high-dose F. tularensis LVS inhalational challenge model to determine whether CLDC administered directly to the nasal and lung mucosa could generate rapid protective immunity. For intranasal (i.n.) delivery of CLDC, BALB/c mice were anesthetized and given 20 µl of CLDC (10 µl per nostril). This volume of CLDC contained 4 µg non-coding plasmid DNA. Control mice were administered 20 µl of the 5% dextrose water diluent used to prepare CLDC. At 24 h post-treatment, mice (BALB/c, n = 5 per group) were challenged with 105 colony forming units F. tularensis LVS via the i.n. route. We and others have found that this dose consistently results in 100% lethality for BALB/c mice [18,20–22]. Infected mice were monitored twice daily for disease symptoms and euthanized at a pre-determined humane endpoint. Untreated and diluent-treated mice all developed severe infection and were euthanized at days 7–8 post-challenge (Fig. 1a). In contrast, all mucosal CLDC-treated mice were protected from disease. Organ bacterial burdens were determined at day 6 post-challenge for diluent and CLDC-treated mice. CLDC treatment resulted in a greater than one to two log decrease in bacterial burden in lungs, liver and spleen (Fig. 1b). These data indicate that mucosal CLDC treatment provides innate immune protection from pneumonic F. tularensis LVS.
To determine whether CLDC-treated mice cleared F. tularensis LVS infection, CLDC-treated mice surviving challenge were euthanized at day 40 and lungs, liver and spleen were homogenized and plated on modified Mueller–Hinton agar. In two separate experiments (n = 5 mice per group), we observed that all organs were free of bacteria, suggesting that CLDC-treated mice effectively and completely cleared the F. tularensis infection.
We also investigated whether clearance of the initial challenge dose was associated with development of long-lasting adaptive immunity to F. tularensis which might protect against future infection. Therefore, survivors from the initial CLDC-treated mice were subjected to re-challenge on day 40 after the first challenge. The mice (BALB/c, n = 5 per group) were challenged i.n. with 105 or 106 CFU F. tularensis LVS. In two separate experiments, all mice (n = 5 per group) survived re-challenge at 105 and 106 CFU with no disease symptoms during 40 days of observation (data not shown). The ability of previously CLDC-treated and infected mice to resist re-challenge with F. tularensis LVS at a 10-fold higher dose than the initial 100% lethal dose with no apparent signs of disease strongly suggests that these mice mounted a protective adaptive immune response. Thus, CLDC-treated and challenged mice were able to develop effective adaptive immune responses against subsequent re-challenge with F. tularensis.
Next, we assessed the impact of route of delivery on the effectiveness of immunotherapy with CLDC. Mice (n = 5 per group) were pre-treated 24 h before infection with CLDC administered by the i.n. route (20 µl), the i.p. route (200 µl), the s.c. route (200 µl) and the i.v. route (100 µl). These doses of CLDC were chosen because they elicit significant immune stimulation, without toxicity (Dow, S., unpublished observations). Twenty-four hours after administration of CLDC, mice were challenged i.n. with 105 CFU F. tularensis LVS. We observed that immunotherapy with CLDC administered via the i.v., i.p. and s.c. routes did not significantly increase survival time compared to untreated controls (Fig. 2a). In contrast, i.n. administration of CLDC 24 h before challenge elicited complete protection from lethal infection (Fig. 2a). Thus, mucosal administration of CLDC immunotherapy was much more effective than systemic treatment for generating protection from pneumonic F. tularensis infection.
We next investigated the effects that timing of CLDC administration had on generating protective immunity. Mice (n = 5 per group) were treated with 20 µl CLDC administered i.n. 72 h prior to challenge, 24 h prior to challenge, 10 min after challenge (0 h time point) or 24 h after challenge. Whereas pre-treatment 24 h prior to challenge resulted in 100% protection from lethal infection, CLDC immunotherapy administered 72 h before challenge resulted in only 60% survival (Fig. 2b). Administration of CLDC concurrently with infection (0 h time point) or administration 24 h after infection resulted in 40% and 60% survival, respectively (Fig. 2b). These findings suggest that there is a window of maximal activity for mucosal CLDC immunotherapy, which diminishes partially with increasing time of administration before or after challenge.
Studies were conducted next to characterize the CLDC-induced innate immune responses in the lung. Systemically administered CLDC have been shown previously to provoke a strong Th1 type immune response with production of pro-inflammatory cytokines [13,15]. Several of these cytokines including IFN-γ, IL-12 and TNF-α have been demonstrated to function in host defense against pneumonic F. tularensis [9,23]. Thus, we sought to determine whether mucosal administration of CLDC resulted in increased expression in the lung of these important protective cytokines. In addition, we examined expression of the type I interferons IFN-α and IFN-β in the lung in response to mucosal CLDC immunotherapy. Recent studies suggest that type I IFNs can play an important role in either increasing or decreasing the susceptibility of animals to bacterial infections [24–31]. Moreover, CLDC are known to be potent inducers of type I IFN production [32,33].
Cytokine gene expression in lung tissues following mucosal administration of CLDC was determined using quantitative realtime PCR (qRT-PCR). Induction of cytokine gene expression by CLDC treatment was assessed by comparing mRNA levels in lungs from CLDC-treated mice to lungs from untreated mice, resulting in a relative measure of the fold change in mRNA expression. At 24 h after CLDC treatment, lungs from treated mice (n = 4 mice per group) had significant increases in expression of IFN-γ, IL-12b(p40), TNF-α, IFN-α4 and IFN-β (Fig. 3). For example, cytokine induction in lungs of CLDC-treated mice ranged from a high of 46-fold induction for IFN-γ to 4-fold for IFN-α4 (Fig. 3). These data demonstrate that mucosal CLDC treatment resulted in the upregulation of key pro-inflammatory cytokine genes with potential antibacterial activity in the lung. The role of these key cytokines in CLDC-induced protection was examined next.
Previous studies have highlighted the critical importance of IFN-γ for defense against F. tularensis [23,34–36]. In addition, CLDC have been shown in earlier studies to potently induce production of IFN-γ by NK cells . Therefore, we hypothesized that IFN-γ production might be a key factor in the protection elicited by mucosally administered CLDC. This possibility was tested by assessing the ability of CLDC immunotherapy to protect mice unable to produce IFN-γ against F. tularensis challenge. Wild-type BALB/c mice and IFN-γ−/− mice (n = 5 per group) on the BALB/c background were administered CLDC by the i.n. route, then subjected to lethal i.n. challenge with 105 CFU F. tularensis LVS 24 h after CLDC treatment. We observed that CLDC immunotherapy was completely ineffective in protecting IFN-γ−/− mice from F. tularensis challenge, with survival times in CLDC-treated IFN-γ−/− mice similar to those of untreated wild-type mice (Fig. 4a). In addition, CLDC treatment of IFN-γ−/− mice failed to reduce their bacterial burdens below the levels present in infected but sham-treated wild-type mice (Fig. 4b). These data indicate that CLDC-elicited production of IFN-γ is required for protection against lethal F. tularensis pulmonary challenge.
We have found previously that NK cells are the primary source of IFN-γ production in CLDC-treated animals . Thus, we hypothesized that NK cells might play an important role in protection mediated by mucosally delivered CLDC. This possibility was examined by means of systemic NK depletion experiments. Mice were depleted of NK cells by i.p. administration of 50 ul of anti-asialo-GM1 antibody, a treatment which we have found results in >80% depletion of lung NK cells at 24 h (data not shown). Additionally, other investigators have demonstrated that systemic NK cell depletion after administration of the asialo-GM1 antibody lasts for at least 5 days [37–39]. Wild-type BALB/c mice (n = 5 per group) were administered anti-asialo-GM1 antibody or an equivalent amount of control rabbit IgG 24 h prior to CLDC treatment. The mice were administered CLDC i.n and then subjected to i.n. challenge 24 h later with 105 CFU F. tularensis LVS. Five days after the first NK cell depletion, the mice were treated with a second dose of NK depleting or control antibody to assure continued NK depletion. Mucosal immunotherapy with CLDC treatment generated complete protection in mice pre-treated with irrelevant rabbit IgG and in non pre-treated mice (Fig. 5a). In contrast, NK cell depletion eliminated the protective effect of CLDC treatment. For example, all NK cell-depleted and CLDC-treated mice were euthanized on approximately the same days as diluent treated mice (Fig. 5a). In addition, suppression of bacterial replication and dissemination from the lungs to spleen and liver by CLDC immunotherapy was also significantly abrogated in the NK cell depleted mice (Fig. 5b). Thus, these data indicate a key role for NK cells in mediating effective control of bacterial growth and dissemination in response to mucosally administered CLDC immunotherapy.
While the type I interferons (IFN), IFN-α and IFN-β, have been classically understood to provide broad defense against many viral infections, it has more recently become appreciated that type I interferons also may play a role in defense against certain bacterial pathogens [26,28] or have a pathogenic role in other bacterial infection [24,25,27,29]. At least one previous report indicated that mice lacking IFN-α/β receptor were more resistant to F. tularensis . However, IFN-β is also required for activation of the host-protective inflammasome in response to F. tularensis infection [40–42]. Thus we hypothesized that type I IFN production may modulate pneumonic F. tularensis infection. To address this question, we conducted challenge studies in mice which lack the common receptor for type I interferon (IFN-α/β receptor−/−). Intranasal challenge of wild-type 129 mice (n = 5 per group) demonstrated that this mouse strain is highly resistant to F. tularensis LVS. Wild-type 129 mice challenged with up to 106 CFU all survived to day 40 without developing severe disease (Fig. 6a and data not shown). However, when isogenic IFN-α/βR−/− mice were challenged with 105 CFU F. tularensis LVS, they succumbed to disease within 9 days of challenge (Fig. 6a). Thus, these results suggest that type I IFNs may play a protective role in immunity against F. tularensis infection in the context of the highly resistant 129 mouse model.
Recent studies have also demonstrated that CLDC are potent inducers of type I IFN production [32,33]. Since the preceding experiments suggested a potential role for type I IFNs in protection from F. tularensis infection, we next assessed the role of type I IFNs in CLDC-mediated protection. When IFN-α/βR−/− mice were treated i.n. with CLDC 24 h prior to challenge with 105 CFU F. tularensis LVS, we observed that they were significantly protected compared to sham-treated IFN-α/βR−/− mice (Fig. 6b). Thus, induction of type I IFN responses do not appear to be necessary for CLDC-induced protection.
Having established the importance of IFN-γ for CLDC-induced protection from F. tularensis, we next sought to further characterize the mechanism of cytokine induced protection. Previous studies have established that macrophages and dendritic cells are important primary target cells for infection with F. tularensis [18,43–45]. In addition, it has been shown that IFN-γ and TNF-α can induce macrophage intracellular killing of F. tularensis [46–50]. Therefore, we investigated the effects of CLDC-elicited cytokines on intracellular inhibition of F. tularensis LVS, using an in vitro system. Alveolar macrophages from the AMJ.2 cell line were pre-treated for 24 h with supernatants from CLDC-stimulated spleen cells, then infected with F. tularensis and the effects on numbers of viable intracellular bacteria determined 48 h later, as described in Section 2.
We found that pre-treatment of AMJ cells with supernatants from CLDC-stimulated spleen cells resulted in significant reduction in intracellular growth of F. tularensis, compared to untreated cells or cells pre-treated with supernatants from control spleens (Fig. 7). The inhibitory activity was also present even when supernatants were diluted as much as 1:100. This result indicated that soluble factors induced by CLDC promoted cellular resistance to F. tularensis LVS intracellular growth.
To identify the specific factor(s) responsible for inhibiting F. tularensis intracellular growth, we assessed the effects of neutralizing IFN-γ or TNF-α activity (or both) on the ability of supernatants to inhibit replication of F. tularensis in macrophages. Neutralization of IFN-γ activity resulted in nearly complete elimination of the F. tularensis suppressive activity in supernatants (Fig. 7), which indicated that IFN-γ was the primary CLDC-induced factor responsible for inhibiting F. tularensis LVS growth. Neutralization of TNF-α activity induced a small but not significant reduction in F. tularensis suppressive activity (Fig. 7), indicating that TNF-α was not the major active cytokine. However, neutralization of both IFN-γ and TNF-α activity completely abrogated the inhibitory effects of CLDC-stimulated spleen supernatants, suggesting a possible cooperative effect between the two cytokines. These findings suggest that one effect of mucosal CLDC immunotherapy may be to inhibit the replication of F. tularensis in an early target cell population of alveolar macrophages in an IFN-γ dependent fashion.
The preceding results indicated that mucosal immunotherapy was effective in generating rapid protection against F. tularensis LVS challenge. However, F. tularensis LVS, while fully virulent in mice, is significantly attenuated in humans [4,51]. Therefore, results of experiments conducted with F. tularensis LVS, while mechanistically informative, may not fully predict outcomes in humans exposed to more virulent strains of F. tularensis. Therefore, we next evaluated the effectiveness of CLDC immunotherapy in a challenge model employing the highly virulent Schu4 strain of F. tularensis .
BALB/c mice (n = 5 per group) were treated with CLDC or diluent administered by the i.n. route. Twenty-four hours later, the mice were challenged with an LD100 dose of 40 CFU F. tularensis Schu4, administered by the i.n. route. We observed that all diluent treated mice succumbed to disease by 4.5 days post-challenge (Fig. 8a). In contrast, mice treated with CLDC survived significantly longer (p = 0.0027) than control mice, though they ultimately did succumb to infection. We also assessed the impact of CLDC immunotherapy on body weight changes in infected mice and found that CLDC-treated mice had a significant reduction in F. tularensis-induced weight loss, compared to control mice (Fig. 8b). In addition, CLDC treatment resulted in a more than 10-fold reduction in bacterial burdens in the lung, liver and spleen at day 3 after bacterial challenge (Fig. 8c). Collectively, these data demonstrate that mucosal CLDC treatment is capable of eliciting at least partial protection even against a highly virulent strain of F. tularensis. These results suggest that additional refinements to the CLDC immunotherapy platform, or the use of combination therapy, could result in greater protective effects.
In this study we found that delivery of immunostimulatory cationic lipid-DNA complexes to the lung mucosa protected mice from pulmonary challenge with F. tularensis. In the F. tularensis LVS challenge model, mucosal CLDC treatment provided complete protection from lethal pulmonary challenge. In the case of the highly virulent F. tularensis Schu4 strain, mucosal CLDC treatment resulted in a statistically significant increase in survival time and decrease in bacterial burden. CLDC-protected mice surviving LVS challenge were immune to subsequent high dose pulmonary challenge. Thus, in the LVS challenge model innate immune protection with mucosal CLDC was not sterilizing and allowed the development of an adaptive immune response which was host-protective.
While previous studies have demonstrated that vaccination and other strategies utilizing antigen-specific immunotherapeutic approaches can provide protection from inhaled F. tularensis [4,21,52–56], fewer studies have examined the ability of pathogen non-specific innate immunity to generate protection from inhalational challenge [9–11]. Previous work has established that systemic immunotherapy can protect mice from systemic challenge with F. tularensis [7,8]. However, in the case of inhalational challenge, we found that only mucosally administered CLDC immunotherapy provided complete protection while systemic routes of immunotherapy provided incomplete protection. Targeting of CLDC to the lung mucosa likely has several advantages over systemic delivery for protection from pulmonary F. tularensis. By directing immunotherapy to the initial site of infection, the immune response has an opportunity to limit bacterial growth in the lung prior to systemic dissemination of the infection. In this study we demonstrated that CLDC promoted a pro-inflammatory environment in the lung characterized by increased gene expression of IL-12, IFN-γ, TNF-α, IFN-α and IFN-β. We also found that CLDC-elicited cytokines induced alveolar phagocytes to resist F. tularensis intracellular growth in vitro. Alveolar phagocytes are important initial target cells for F. tularensis infection in the lung [3,18,45,57], and therefore CLDC-induced activation of these cells may serve to help restrict initial infection. In contrast, our own recent studies and those of others suggest that F. tularensis pulmonary infection typically results in an immunosuppressive phenotype in the lung characterized by a lack of pro-inflammatory cytokines and an increase in the immunosuppressive cytokine TGF-β which likely favors bacterial growth and dissemination [17,18,58,59]. Thus mucosal CLDC immunotherapy promotes a pro-inflammatory environment in the lung which favors bacterial killing and likely at least partially counteracts the immunosuppressive environment induced by F. tularensis.
The timing of mucosal CLDC administration was critical in influencing the outcome of F. tularensis LVS challenge. Mucosal CLDC provided complete protection when given 24 h prior to challenge while administration at later times concurrent with and following challenge only provided partial protection. Thus pre-exposure prophylaxis with mucosal CLDC would likely be very effective. Pre-exposure prophylaxis might be practical and useful in situations in which a threat of exposure is known such as in the early stages of an epidemic when only a few individuals have become infected but many more could be at risk of infection. The partial efficacy of therapy 24 h after challenge also suggests that this therapy might potentially be beneficial for post-exposure treatment. For post-exposure therapy, mucosal CLDC might be used in combination with other therapies to provide augmented protection. The greater efficacy of pre-exposure prophylaxis may in part be explained by F. tularensis infection causing an immunosuppressive phenotype in the lung [17,18,58,59]. Pre-exposure prophylaxis would therefore provide the opportunity for mucosal CLDC to generate a pro-inflammatory anti-bacterial Th1 type response in the lung prior to F. tularensis-induced immunosuppression. In contrast, post-exposure CLDC treatment may not be able to fully counteract the already established F. tularensis-induced immunosuppression.
In this study we found that mucosal CLDC treatment induced Th1 cytokine gene expression in the lung including IL-12, IFN-γ, TNF-α, IFN-α and IFN-β. Of the tested genes, IFN-γ was the most up-regulated. The important role of IFN-γ in defense against F. tularensis is very well established [9,23,34–36,60]. As previously described [9,23], we found that IFN-γ−/− mice were highly susceptible to inhaled challenge with F. tularensis LVS resulting in a high day 6 bacterial burden (approximately 109 CFU per lung, spleen or liver). We found that treatment of IFN-γ−/− mice with mucosal CLDC did not increase survival and importantly did not alter bacterial burden in any tested organ. Thus it appeared that IFN-γ was absolutely necessary for mucosal CLDC-induced protection. We further investigated the role of cytokines in defense against F. tularensis using an in vitro model of LVS infection in alveolar macrophages. Pre-treatment of cells with CLDC spleen supernatant resulted in a two log reduction in intracellular F. tularensis growth. Antibody depletion of IFN-γ present in CLDC spleen supernatant resulted in nearly complete restoration of F. tularensis intracellular growth, suggesting that IFN-γ is the major CLDC-induced component responsible for induction of cellular resistance. Numerous studies corroborate this data by showing that IFN-γ induces an anti-F. tularensis state in phagocytes [46,48,50]. Cellular killing of F. tularensis by IFN-γ-activated phagocytes is typically dependent on nitric oxide (NO) [48,50,61], although one study found that IFN-γ-induced killing of F. tularensis by murine alveolar macrophages was NO independent .
We assessed the role of NK cells in mucosal CLDC-induced protection by systemically depleting NK cells from BALB/c mice with anti-asialo-GM1 antibody. NK cell depletion resulted in a significant and marked loss of CLDC-induced protection. Thus, we concluded that NK cells were required for mucosal CLDC-induced protection. We have previously found that NK cells are the source of the majority of systemic CLDC-induced IFN-γ in spleen and lung . Therefore, it is likely that NK cells are the major source for IFN-γ production in response to mucosal CLDC as well. However, we cannot exclude the possibility that IFN-γ might also be produced by another cell type such as CD8+ T cells or that alternative IFN-γ-independent mechanisms may play a role in CLDC-induced protection from F. tularensis growth and dissemination. Nonetheless, the effect of NK cell depletion on CLDC-treated mouse survival implicates an important role for NK cells in CLDC-induced protection.
CLDC are potent inducers of the type I interferons, IFN-α and IFN-β [32,33]. While type I interferons have classically been understood to provide host defense against viruses, more recently it has been demonstrated that type I interferons can also have important impacts on bacterial infections as well . For some bacteria, type I interferons exert host-protective effects [28,30,31] while in other bacterial systems they have been shown to enhance pathogenesis [24,25,27,29].In F. tularensis infection, IFN-β appears to play a key role in activation of the inflammasome for innate immune defense [40–42]. However, at least one report has demonstrated that BALB/c mice lacking IFN-α/β receptor were more resistant to F. tularensis LVS . In contrast, our finding that IFN-α/β receptor−/− mice on a 129SvEv background were more susceptible to inhaled F. tularensis LVS infection supports the idea that type I interferons may play a protective role in defense against pneumonic tularemia. The basis for the discrepancy in results between this study and that of Metzger et al.  is not known. It is possible that a difference in the mouse strain studied may influence the role of type I interferons in protection/pathogenesis. For example, we found that 129SvEv mice survived challenge with an inhaled dose of F. tularensis LVS which was 10-fold higher than the LD100 for BALB/c mice. The basis for this high level of natural resistance is not known and would be intriguing to investigate. Regardless of the effect of type I interferons on infection, we found that type I interferons were not necessary for mucosal CLDC-induced protection, suggesting that CLDC-induced IFN-γ responses are likely sufficient for protection elicited by mucosal CLDC treatment.
In summary, we found that mucosal CLDC immunotherapy provided protection against lethal pulmonary F. tularensis. Mucosal administration of CLDC resulted in a lung innate immune response characterized by upregulation of pro-inflammatory cytokines. We found that CLDC-induced protection was dependent on the ability of mice to produce IFN-γ. Additionally, the ability of CLDC to limit F. tularensis growth in alveolar macrophages was highly dependent on IFN-γ. NK cell depletion experiments demonstrated that NK cells also play an important role in CLDC-induced protection. To our knowledge no truly antigen non-specific innate immunity has previously demonstrated the ability to protect mice against lethal challenge with the virulent Schu4 strain of F. tularensis. Therefore, our finding that mucosal CLDC provides partial protection against pulmonary Schu4 challenge represents a step forward in the development of immunotherapeutics to combat virulent pulmonary F. tularensis. New formulations of CLDC which combine additional pattern recognition receptor agonists may provide enhanced potency. While immunotherapy alone may not be sufficient to provide complete protection from highly virulent pulmonary F. tularensis, CLDC immunotherapy might potentially be used to augment or enhance post-exposure vaccine or antibiotic therapy. We are currently studying whether CLDC-based immunotherapy might be effective in enhancing the potency of antibiotic therapy or allowing delayed initiation of antibiotic therapy.
The authors thank Shayna Warner and Amber Troy for their excellent technical contributions. Andrew Goodyear provided critical BSL3 training and discussion.. This work was supported by a National Institute of Allergy and Infectious Diseases SBIR grant (1 R43 AI060146-01A2).