INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The cilium-associated respiratory (CAR) bacillus is an unclassified, extracellular, gram-negative, gliding bacterium that was first characterized in an aging rat colony with chronic respiratory disease (35). Morphologically similar bacteria have since been described colonizing the respiratory epithelium of other rodent species (4, 11, 20), rabbits (16), and domesticated livestock (6, 12, 26). CAR bacillus was named based on its characteristic pattern of colonization parallel to and between the cilia of the upper respiratory tract epithelium (8). Chronic respiratory disease due to CAR bacillus exhibits characteristic histologic lesions in rodents. In the early phases of disease there are mild peribronchiolar lymphoid infiltrates that progress to severe bronchopneumonia with bronchiectasis in the chronic stages (8). These lesions are virtually identical to those seen in Mycoplasma pulmonis-infected rodents (29). In addition to the histologic evidence of an immune response, rodents develop a systemic antibody response following CAR bacillus infection (8, 11); however, this apparent immune response is ineffective at clearing the infection.

Extracellular bacteria, such as CAR bacillus, are typically eliminated by the host's innate and acquired immune responses (24). In the early phases of disease, macrophage-mediated phagocytosis and T-cell-independent B-cell antibody production mediate elimination or neutralization of extracellular bacteria. Acquired responses promote T-cell-dependent B-cell antibody production and enhanced macrophage activation via cytokines produced by T cells (1). Cytokine responses from cytotoxic and helper T cells have been described as being type 1 or type 2 (22, 23). Type 1 responses are characterized by the production of gamma interferon (IFN-) and interleukin-12 (IL-12). Type 2 responses are characterized by production of IL-4, IL-5, IL-6, and IL-10.

Cytokine responses have been shown to either confer resistance or contribute to disease severity, depending on the bacterium causing the pneumonia (30, 31). For example, type 1 responses with production of IFN- have been associated with resistance to disease in murine models of respiratory infection with Bordetella pertussis (21) and Shigella flexneri (32). Similarly, tumor necrosis factor alpha (TNF-), a proinflammatory cytokine, enhances the clearance of Pseudomonas aeruginosa (33) and Klebsiella pneumoniae (18) in murine pneumonia models. Cytokine responses that contribute to disease have been demonstrated in studies of murine mycoplasmosis. Susceptible mice infected with Mycoplasma pulmonis developed elevated TNF- and IFN- levels with increased disease severity and Mycoplasma colonization (5, 27). Systemic antibody responses have also been associated with disease severity in murine mycoplasmosis. Susceptible C3H/HeN mice have increased serum immunoglobulin G1 (IgG1) and IgG2a antibody responses, whereas resistant C57BL/6 mice demonstrate a minimal antibody response (3).

The presence of severe disease in CAR bacillus-infected mice with concurrent production of antibody suggests that the immune response is ineffective. To begin characterization of the host immune response to CAR bacillus infection, two mouse strains (BALB/c and C57BL/6) were experimentally infected with CAR bacillus by intratracheal inoculation. Severity of disease, systemic humoral immune response, and pulmonary cytokine production were measured. BALB/c mice developed severe disease with elevations in CAR bacillus-specific antibody and persistent elevations in local TNF-, IFN-, and IL-4. C57BL/6 mice developed minimal disease, had a limited antibody response, and showed no detectable cytokine elevations. These results suggest that neither systemic humoral immunity nor cytokine responses are critical in resistance to CAR bacillus-induced disease; however, elevations of cytokines may play a role in disease pathogenesis and development of disease.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

CAR bacillus culture. A CAR bacillus isolate (provided by Tom Spencer, National Institutes of Health) originally obtained from a mouse was maintained in cell culture on murine 3T3 fibroblasts in Dulbecco modified Eagle medium supplemented with 10% fetal bovine serum (8, 28). Prior to inoculation, flasks containing CAR bacillus were scraped and cellular debris was removed by centrifugation at 900 × g for 10 min. The bacteria were pelleted by centrifugation at 20,000 × g for 10 min and resuspended in 1 ml phosphate-buffered saline (PBS; 15 mM NaH₂PO₄, 108 mM Na₂HPO₄, 1.4 M NaCl; pH 7.4). The bacterial concentration was determined microscopically with a hemocytometer, and bacterial suspensions were diluted with PBS to achieve a final inoculum concentration of 10⁵ CAR bacillus organisms per 30 µl of PBS.

Mycoplasma screen. To ensure that the CAR bacillus inoculum was free of mycoplasma, PCR assays to detect Mycoplasma DNA were performed on the cell cultures prior to inoculation. Mammalian cell pellets and bacterial cell pellets from CAR bacillus cultures were resuspended in 200 µl of PBS, and DNA was isolated using a Qiamp Tissue Kit (Qiagen, Santa Clarita, Calif.) according to the manufacturer's instructions. Mycoplasma PCR assays using primers known to amplify all species of Mycoplasma were performed on isolated DNA as previously described (34). All inoculated mice were screened for antibodies to M. pulmonis by enzyme-linked immunosorbent assay (ELISA; Research Animal Diagnostic and Investigative Laboratory, University of Missouri, Columbia, Mo.). Selected mice inoculated with CAR bacillus were further assessed for exposure to M. pulmonis by culture of lung swabs on selective modified Dutch agar medium at 37°C for 14 days. PCR analysis of paraffin-embedded lung sections was also performed to detect M. pulmonis contamination.

Experimental model. Six-week-old female BALB/c and C57BL/6 mice, free of known pathogens, including the respiratory pathogens CAR bacillus, M. pulmonis, Sendai virus, and pneumonia virus of mice, were obtained from the Frederick Cancer Research and Development Facility (Frederick, Md.). Mice were group housed in microisolator cages in accordance with the Guide for the Care and Use of Laboratory Animals (25). Three separate experiments were performed evaluating four to eight infected mice and three to four controls every 7 days for 49 days. Days 14 and 42 postinoculation (p.i.) were eliminated from the third experiment. A total of 94 BALB/c and 114 C57BL/6 mice were administered 30 µl of bacterial inoculum by intratracheal injection. For intratracheal inoculations, mice were anesthetized with isoflurane, and a skin incision was made on the ventral surface of the neck to expose the trachea. After intratracheal instillation of inocula, the skin incision was closed with surgical adhesive (Nexaband; Veterinary Products Laboratories, Phoenix, Ariz.). Control mice from each strain were inoculated with 30 µl of PBS. A dose of 10⁵ CAR bacillus organisms was used for all studies. This was the minimum dose found to cause consistent disease in pilot studies using BALB/c mice (data not shown).

Colonization determination. Since CAR bacillus does not grow on cell-free media (8, 28), PCR amplification was used to determine if mice were colonized with CAR bacillus. Fifty-micron sections of paraffin-embedded lung were assayed by PCR for the presence of CAR bacillus DNA. Paraffin was extracted from embedded sections of lung with xylene, and DNA was isolated according to the procedures outlined in the Qiamp Tissue Kit (Qiagen). CAR bacillus PCR was performed on the isolated DNA using previously published procedures (7). If an inoculated mouse was PCR negative, then Southern blot analysis of the PCR products was done to confirm the colonization status.

Histology. Samples of lung were embedded in paraffin, sectioned at 5 µm, and stained with hematoxylin and eosin and with a silver staining method. Lesions were scored on a scale of 1 to 7 based on criteria described in Table 1. Silver-stained sections of lung were evaluated to subjectively assess the degree of colonization in infected mice.

CAR bacillus-specific antibody isotyping. Serum IgM, IgG1, IgG2a, IgG2b, IgG3, and IgA levels were estimated in mice that had detectable antibodies to CAR bacillus as determined by ELISA. Falcon Probind plates (Becton Dickinson, Lincoln Park, N.J.) were coated overnight at 4°C with either 1.0 µg of purified CAR bacillus whole-cell antigen per ml for quantitation of antigen-specific isotypes or 1.5 µg of antibody to the heavy and light chains of mouse immunoglobulin (Jackson ImmunoResearch Laboratories, Inc., West Grove, Pa.) per ml for the isotype standards. Plates were blocked with 0.5% dry milk in PBS for 30 min at room temperature. Serum samples were diluted 1:50 for the determination of IgM and IgG isotypes and 1:10 for IgA. Standards were prepared from murine myeloma cells secreting IgM, IgG1, IgG2a, IgG2b, IgG3, or IgA (Zymed Laboratories, Inc., South San Francisco, Calif.) at concentrations of 0, 4, 8, 16, 32, and 64 ng/ml. All samples were run in duplicate, and standards were run in triplicate. Portions (100 µl) of diluted sample or standard were added to the appropriate wells and incubated at 37°C for 2 h. The plates were washed three times with 0.05% Tween 20 in PBS. The appropriate horseradish peroxidase-labeled goat anti-mouse secondary antibody was applied (Southern Biotechnology Associates, Inc., Birmingham, Ala.) and incubated for 2 h at 37°C. Plates were washed five times and incubated at room temperature for 30 min with 100 µl of azino-diethyl-benzthiazoline sulfonate substrate (Kirkegaard & Perry Laboratories, Gaithersburg, Md.). Optical density values were determined on a Bio-Kinetics reader (Bio-Tek Instruments, Inc., Winooski, Vt.) at 405 nm.

Cytokine evaluation. Local pulmonary TNF-, IFN-, and IL-4 protein levels were measured in lung homogenates by ELISA using commercially available kits and following the manufacturer's instructions (Genzyme Diagnostics, Cambridge, Mass.).

Statistical analysis. Statistical analysis was performed using SigmaStat Statistical Software package (SPSS Marketing, San Rafael, Calif.). Differences in median histologic lesion scores of BALB/c and C57BL/6 mice were statistically analyzed by Mann-Whitney rank sum analysis and were considered significant at a P of <0.05. The correlation between antibody isotype and histologic lesion scores was assessed using the Spearman rank correlation test and was considered significant at a P of <0.01. Significant differences in serum antibody isotypes were determined using a Student's t test when the data were normally distributed or a Mann-Whitney rank sum test when the data were not normally distributed. P values of <0.05 were considered significant. A Student's t test was used to determine differences in pulmonary cytokines between infected and control mice and was considered significant at a P of <0.05.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Screening for mycoplasma. Mycoplasma contamination is common in cell cultures and is a significant concern in CAR bacillus infection studies because of the similarities between CAR bacillus- and M. pulmonis-induced disease (29). Therefore, it was important to ensure that the inoculum for these experiments was free of Mycoplasma contamination. Prior to inoculation, CAR bacillus cultures and 3T3 fibroblast cell lines were screened by PCR for Mycoplasma contamination. PCR amplification of cell cultures was uniformly negative. Furthermore, no serum antibodies to M. pulmonis were detected in any inoculated mice. Lung cultures from selected inoculated mice with gross lung lesions were negative for M. pulmonis. Selected mice with histologic lesions were uniformly negative for Mycoplasma organisms as determined by PCR analyses of paraffin-embedded lung sections. Based on these results, we concluded that the inoculum and the mice were free of Mycoplasma contamination and that CAR bacillus was the sole pathogen responsible for disease.

Inoculation and colonization. Mice were anesthetized and inoculated by intratracheal instillation with either 30 µl of PBS or 30 µl of PBS containing 10⁵ CAR bacillus organisms. At weekly intervals, control mice and inoculated mice from each strain were euthanized. Since histology is an insensitive indicator of CAR bacillus colonization, two other criteria were used to assess the colonization status of inoculated mice: (i) examination for CAR bacillus colonization by PCR analyses of paraffin-embedded tissues and (ii) Southern blot analysis of the PCR products. Mice were considered colonized if they were positive by PCR or by Southern blot analysis (Table 2). Since we were interested in the antibody and cytokine profiles associated with CAR bacillus-induced disease, antibody isotyping and cytokine determinations were only performed on colonized mice.

Histologic lesions associated with CAR bacillus disease develop more rapidly and are more severe in BALB/c mice than in C57BL/6 mice. Samples of the left lung lobe at the level of the bronchus were examined for histologic lesions, and lesions were scored on a scale of 1 to 7 (Table 1). Sham-inoculated BALB/c and C57BL/6 mice had a median histologic score of 1 at all time points of the study (range, 1 to 2). The median histologic lesion scores of infected mice are given in Table 3, and the distribution of lesion scores after day 21 p.i. is given in Fig. 1. Lesions in BALB/c mice became evident by 21 days p.i. with a median lesion score of 3 and became more severe during the course of the study. In contrast, lesions in C57BL/6 mice were not evident until 35 to 42 days p.i. and were mild compared to the BALB/c mice with a median histologic score of 2 at day 49 p.i. To determine if C57BL/6 mice developed lesions comparable to BALB/c mice at extended times p.i., lesions were assessed in infected C57BL/6 mice at 70 days p.i. Disease in infected C57BL/6 at 70 days p.i. had not progressed, and the median lesion score remained at 2 (range, 1 to 3) (data not shown).

View larger version (19K): [in this window] [in a new window]   FIG. 1.   Distribution of lung lesion scores of CAR bacillus-infected BALB/c mice () and C57BL/6 mice (). Numbers above the symbols in parentheses indicate the number of mice with that lesion score; symbols with no number represent one animal.

View larger version (106K): [in this window] [in a new window]   FIG. 2.   Silver-stained sections of lung demonstrating the difference in the degree of colonization between susceptible BALB/c (A) and resistant C57BL/6 (B) mice chronically infected with CAR bacillus. Arrows indicate argyrophilic colonies of CAR bacillus in the cilia of the bronchial epithelium.

Antibody isotypes associated with CAR bacillus-infected BALB/c and C57BL/6 mice. CAR bacillus-specific antibody isotypes in serum were estimated by a semiquantitative ELISA in colonized mice that seroconverted for each of the three separate experiments. Because the trends were similar in each of the three experiments, the ELISAs were repeated, testing all mice at the same time. These data are presented in Tables 3 and 4.

Histologic lesions correlate with antibody responses. The most severe lesions and most prominent antibody responses were seen in BALB/c mice. A positive correlation (r = 0.75, P < 0.01) between total serum IgG antibody response and the severity of the histologic lesions was demonstrated in BALB/c mice (Table 3). C57BL/6 mice that were colonized with CAR bacillus developed less-severe histologic lesions with mild to moderate serum antibody responses compared to BALB/c mice. While the median histologic lesion scores in C57BL/6 mice remained between 1 and 2, a positive correlation (r = 0.75, P < 0.01) between the total serum IgG antibody response and the histologic lesions also existed. For example, at day 42 p.i., two C57BL/6 mice developed advanced histologic lesions and had higher levels of total serum IgG (data not shown). The correlation between disease severity and antibody response suggests that the humoral immune response is ineffective at eliminating infection.

Elevated cytokines are associated with CAR bacillus infection. Lung homogenates from six infected BALB/c and six infected C57BL/6 mice were assayed at each time point for TNF-, IFN-, and IL-4 by ELISA, and the levels were compared to those from sham-inoculated mice. BALB/c mice colonized with CAR bacillus had eightfold elevations in the TNF- level and twofold elevations in both the IL-4 and IFN- levels (Fig. 3). TNF- and IL-4 levels were elevated over control values by day 21 p.i. and remained elevated throughout the course of the 49-day study. IFN- levels began to increase at day 28 p.i. and persisted through day 49. There was no increase in cytokine production in the lungs of C57BL/6 mice, with the exception of two mice that showed a twofold elevation in TNF- and IFN- at day 42 (data not shown). These were the same mice that had more severe histologic lesions and elevated serum antibodies. Similar trends were demonstrated in supplemental ELISA testing of lung homogenate cytokines in another group of six BALB/c and C57BL/6 mice at each time point. These data suggest that resistance to CAR bacillus-induced disease is not mediated by either a type 1 (IFN-) or a type 2 (IL-4) cytokine response and that cytokine perturbations may potentiate disease.

View larger version (15K): [in this window] [in a new window]   FIG. 3.   Cytokine levels in lung homogenates of CAR bacillus-infected BALB/c mice. Infected BALB/c mice had elevations in TNF-, IFN-, and IL-4 compared to sham-inoculated mice. These elevations persisted throughout the 49-day study. Elevations in TNF-, IFN-, or IL-4 were not detected in the lung homogenates of CAR bacillus-infected C57BL/6 mice, except for two mice at day 42 p.i. (data not shown). Significant differences in cytokine levels (*, P < 0.05; **, P < 0.10) between control mice and infected mice, as calculated using a Student's t test, are indicated.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The primary objective of these experiments was to characterize the host immune response to CAR bacillus-induced disease. BALB/c and C57BL/6 mice were used because they tend to develop type 2 humoral immunity and type 1 cell-mediated immunity, respectively (9). Pulmonary disease in BALB/c mice developed earlier and was more severe compared to C57BL/6 mice. Infected BALB/c mice produced an increase in all IgG subclass levels, with IgG1 predominating, while C57BL/6 mice had minimally detectable antibody. BALB/c mice also had detectable elevations in pulmonary TNF-, IFN-, and IL-4 that persisted through day 49. These findings were not evident in C57BL/6 mice, except for TNF- and IFN- elevations at day 42 in two mice with more severe histologic lesions.

The observation that all serum antibody isotypes were produced in infected BALB/c mice suggests that a humoral immune response occurred during CAR bacillus disease. However, this humoral immune response was apparently ineffective at controlling disease. Though the reason for this is unknown, it is conceivable that bacterial virulence factors or other cellular responses may be hindering the effectiveness of the antibody response. Mycoplasma pulmonis-infected mice also show a similar trend in antibody responses, with susceptibility to disease being more pronounced with a prominent antibody response (3). Susceptible C3H/HeN mice infected with M. pulmonis have increased antibody responses, with all isotypes represented compared to the lower antibody responses seen in resistant C57BL/6 mice. It also appears that humoral immunity to CAR bacillus does not play a role in conferring resistance to disease since resistant C57BL/6 mice had lower antibody responses compared to susceptible BALB/c mice.

Elevations of IFN- and IL-4 in BALB/c mice suggest T-cell activation of both type 1 and type 2 subsets (22). The elevations in both IgG1 and IgG2a levels suggest that these cytokine perturbations were functionally significant, since IL-4 and IFN- act as isotype switch factors for the B-cell production of IgG1 and IgG2a, respectively. The elevations of IFN- and IL-4 with persistent CAR bacillus colonization suggest a minimal role of T cells in the clearance of bacteria. Furthermore, as described above, the serum antibody response appears to be ineffective at clearing the bacteria. Because acquired immunity appears to be ineffective in the control of CAR bacillus-induced disease, it is likely that innate immunity plays a large role in the control of CAR bacillus infection.

Innate immune responses occur within the first 96 h after infection and are the first line of defense against extracellular bacterial pathogens. Within the respiratory tract, the alveolar macrophage is the primary effector cell of innate immunity, and TNF- production is an indicator of macrophage activation (19). The importance of innate immunity in the control of extracellular bacterial infections of the respiratory tract has been demonstrated for several bacteria (18), including M. pulmonis, which has a pathogenesis similar to that of CAR bacillus. In the acute phases of murine mycoplasmosis, resistant C57BL/6 mice have a significant amount of TNF- detectable within the first 24 h of infection compared to susceptible mice (5). This early TNF- response appears to be important in controlling M. pulmonis infection in mice, since resistant C57BL/6 mice treated with intratracheal instillation of clodronated liposomes to deplete macrophage function have increased disease severity that is equivalent to that seen in susceptible mouse strains (13). Enhanced disease with diminished innate immunity has also been demonstrated in murine pneumonia models of Klebsiella pneumoniae (15) and Pseudomonas aeruginosa (2).

The suggestion that early immune responses to CAR bacillus lead to disease resistance is further supported by the fact that the more resistant C57BL/6 mice have a genetic predisposition for a robust cell-mediated immunity and macrophage function (9, 17, 36). In the CAR bacillus disease model, cytokine responses need to be evaluated at earlier times to clarify the role of innate immune responses in the development of disease resistance or susceptibility.

Whether TNF- has a protective or deleterious role in the face of disease is dependent upon the organism present and the timing of the response (30). The aforementioned examples demonstrate the important role of alveolar macrophages and TNF- in protection against disease; however, excessive production may have deleterious effects on host cells. For example, TNF- appears to promote disease progression in susceptible mice chronically infected with M. pulmonis (27) and in pulmonary models of Shigella flexneri (32). Since chronic respiratory disease caused by CAR bacillus results in a persistent increase in TNF- levels, it appears that TNF- is contributing to the pathogenesis of disease. The persistent elevation of TNF- in BALB/c mice is likely the result of the constant presence of bacterial antigens, such as lipopolysaccharide, causing overproduction of TNF-. Future experiments with mice with TNF- depleted are required to clarify its role in the pathogenesis of CAR bacillus-induced disease.

In conclusion, we have identified an apparent strain-related resistance to CAR bacillus disease. The resistant C57BL/6 mice develop mild histologic lesions without strong antibody or local cytokine responses. Susceptible BALB/c mice develop a more pronounced serum antibody response and elevations in TNF-, the type 1 cytokine IFN-, and the type 2 cytokine IL-4. These apparent immune responses are ineffective at clearing CAR bacillus, causing BALB/c mice to develop chronic, progressive, severe histologic disease with persistent colonization. These data suggest that acquired antibody responses and local T-cell responses are ineffective at eliminating CAR bacillus-induced disease. Furthermore, because there were marked elevations in TNF- and a strong humoral immune response associated with severe disease, it is conceivable that the host immune response actually contributes to disease pathogenesis and progression.