A Mouse Model of Acute Q Fever Following Infection Via A Non-Invasive Intratracheal Inoculation Method

Background: Q fever is a worldwide zoonosis caused by Coxiella burnetii and mainly transmitted by aerosols. This study aims at establishing a systematic and efficient mouse model of acute Q fever via intratracheal (IT) inoculation of aerosolized C. burnetii. Methods: BALB/c mice were infected with C. burnetii via IT route using a non-invasive aerosol pulmonary delivery device to directly place the living C. burnetii organisms into their tracheas. The bacterial loads, pathological lesions, and serological responses were analyzed in mice, and compared with those of mice infected via intraperitoneal (IP) route. Results: As early as at day three post-infection (pi) with a low dose of C. burnetii (1×10⁴ per mouse), a large amount of C. burnetii organisms were determined in blood, lungs, hearts, livers, and spleens of the mice. The inflammatory infiltration was observed in hearts and lungs of mice. Compared with mice infected via IP route, the mice infected via IT route exhibited a higher level of bacterial loads and more severe pathological lesions in hearts and lungs at day 3 and day 7 pi. Conclusions: These data indicated that IT route is more efficient than IP route to cause acute C. burnetii infection in mice. Overall, we successfully established a mouse model of C. burnetii infection via IT route, which is useful for investigations of pathogenesis and immunity of acute C. burnetii infection as well as evaluation of therapeutic drugs and preventive vaccines of Q fever.

of the mice [21,22], but it does not resemble the route of natural infection of C. burnetii in human. IN inoculation of C. burnetii closely resembles the route of natural infection in human because the organisms may rapidly enter the lungs of the mice and cause pneumonia. However, by IN inoculation, only the minority of C. burnetii inoculum are applied the nares and respired through the nasal passages into the lungs of the mice and the majority remains within the nasal passages [22,23].
Unlike IN inoculation, IT inoculation provides more closely resemble the route of natural C.
burnetii infection in human, which directly place a known quantity of organisms into the tracheas by passing the nasal passages of the mice [24]. In the present study, a mouse model of acute C. burnetii infection was developed with the immunocompetent mice (BALB/c) via IT inoculation of C. burnetii using a non-invasive aerosol pulmonary delivery device. The bacterial loads, pathological lesions, and serological responses were analyzed in mice infected with different doses of living C. burnetii organisms.
Methods health and well-being and approved by the Institute of Animal Care and Use Committee (IACUC) of Academy of Military Medical Science (AMMS). Mice were randomly divided into 3 groups (9 mice per group) and they were infected with 1×10 4 (low dose), 1×10 6 (medium dose), and 1×10 8 (high dose) C. burnetii organisms suspended in 50 µl PBS via IT route, respectively. The suspension of C. burnetii was inoculated directly into the tracheas of BALB/c mice using the MicroSprayer aerosolizer (Huironghe Company, Beijing, China) as previously described [26]. Briefly, mice were anaesthetized by intraperitoneal injection of pentobarbital sodium (100 mg/kg body weight), and then each mouse was placed on the slanted board (60° angle from the horizontal direction) and supported by a nylon band under its upper incisors, and footpad reflex was checked to confirm it was asleep. After a clear view of the trachea under laryngoscope (Huironghe Company, Beijing, China), the mouth cavity was lightened with shades of orange-red and the trachea appeared as a white light spot, and then the MicroSprayer aerosolizer was inserted 25 mm from the larynx (near the tracheal bifurcation) and 50 μl of suspension was sprayed. In addition, 9 mice were intraperitoneally infected with 1×10 8 C. burnetii organisms as control.
At day 3, 7, and 14 post-infection (pi), 3 mice in each group were euthanized through cervical dislocation, respectively, and their blood samples were collected to separate sera, their bodies and spleens were weighted, and their hearts, livers, spleens, and lungs were aseptically collected individually. Each organ was divided into two parts, one for extraction of total genomic DNA used in the real-time quantitative polymerase chain reaction (qPCR) and the other for histopathological analysis. Animal carcasses were safely disposed by the IACUC.
qPCR specific for C. burnetii Ten mg of each organ sample and 100 μl of blood per mouse were used to extract total genomic DNA with a DNeasy Blood &Tissue kit (Qiagen, GmbH, Germany) as described by the manufacturer. Each DNA sample was eluted from the DNA extraction column with 200 μl elution buffer. Two μl per DNA sample was used for each reaction in qPCR with primers dotAF (5'-CCATGGCCCCAATTCTCTT -3') and dotAR (5'-GCGCAATACGCTCAATCACA -3') and a probe (5'6-FAM CCGGAGATACCGGCGGTGGG 3'TAMRA-N) targeting the dotA gene of C. burnetii.

Histopathological analysis
Each organ sample was fixed in 10% buffered formalin for a minimum of 48 h, and then the sample was sectioned and embedded in paraffin. The paraffin-embedded tissue was sliced into a small thickness allowed to adhere to slides, and the slides were stained with hematoxylin-eosin (HE) to observe pathological changes under light microscopy.

Detection of IgGs to C. burnetii by ELISA
By enzyme-linked immunosorbent assay (ELISA), IgGs to phase I or phase II antigen of C.

Statistical analysis
The results of qPCR, the ratios of spleen to body weight and the thickness of diaphragmatic walls of the mice were expressed as the mean ± SD of 3 mice per group.
The differences in C. burnetii loads, in OD 450 of antibody response in sera, in the thickness of diaphragmatic walls in the lungs and in the ratios of spleen to body weight were assayed using the two-way analysis of variance (ANOVA) test followed by least significant difference (LSD) test and Student-Newman-Keuls (SNK) test in Statistical Package in the Social Sciences ver. 19 software (SPSS, Chicago, IL). P<0.05 was considered significantly different.

Clinical symptom of mice
To compare the clinical symptom of C. burnetii infection by different dose and different route, we infected mice with 1×10 4 , 1×10 6 , or 1×10 8 of C. burnetii via IT route and 1×10 8 of C. burnetii via IP route and monitored the development of symptoms and survival of the mice. Approximate three days after infection with C. burnetii, all mice showed the signs of illness, including ruffled coats and decreased activity, but none of the mice died. The mice infected via IT route exhibited more severe clinical signs of Q fever than the mice infected via IP route.

The C. burnetii distribution in mice
To compare the bacterial dissemination in mice infected with different dose (1×10 4 , 1×10 6 , 1×10 8 ) of C. burnetii via IT route, 3 mice per group were killed at day 3, 7, or 14 pi, and the bacterial loads in their blood, hearts, lungs, livers and spleens were detected by qPCR, and their bodies and spleens were weighted. As shown in Fig. 1, a large amount of C. burnetii were detected in these organs of all the mice infected via IT route at day 3 pi. The bacterial loads in 4 organs at day 7 pi were higher than those at day 3 or 14 pi, respectively. There were not significantly increased or decreased in bacterial loads in hearts, lungs, livers, or spleens of the mice infected via IT route with each dose of C. burnetii between three time points pi. A large amount of C. burnetii were also detected in blood, but the results calculated were not normally distributed (data not shown).
Next, we compared the bacterial dissemination in mice infected with high dose (1×10 8 /per mouse) of C. burnetii via IT route and via IP route. Interestingly, the bacterial loads in hearts, but not in other organs, of mice infected via IT route were significantly higher than those of the mice infected via IP route at day 3 and 7 pi, while the bacterial loads in lungs of mice infected via IT route were modestly higher than those of mice infected via IP route at each time point pi, but not significantly different (Fig. 2a, b, c, d). In addition, the ratio of spleen to body weight of mice infected via IT route at day 3 and 14 pi were significantly higher than those of mice infected via IP route (Fig. 2e).

Histopathological findings
To evaluate the histopathological lesions caused by C. burnetii infection via IT route, we examined the stained tissue slices from heart, lung, spleen, and liver of mice using a light microscopy. The infiltration of inflammatory cells was found in epicardium ( The infiltration of inflammatory cells, mainly neutrophils and macrophages, and the increased thickness of diaphragmatic walls were observed in lungs of mice infected with C. burnetii via IT route at day 3, 7, and 14 pi, while these histopathological lesions in lung of mice infected via IP route cannot be observed until day 7 pi (Fig. 4a). Inflammation cells and epithelial cells appeared in tracheas of mice at day 14 pi (Fig. 4b). There was no significant difference in numbers of inflammatory cells and thickness of diaphragmatic walls in lungs between mice infected with different dose of C. burnetii via IT route (data not shown). However, the thickness of diaphragmatic walls in lungs of mice infected with high dose of C. burnetii by IT route was significantly higher than that of mice infected with high dose of C. burnetii by IP route (Fig. 2f).
The pathological lesions were also observed in livers and spleens of mice infected with C. burnetii by IT route at day 7 pi. The infiltration of inflammation cells, mainly mononuclear cells, was focused on the portal area in livers (Fig. 5). Red pulp widened and white pulp atrophied, and splenic sinusoid extended with hyperemia, and the boundary became blurred in spleens (Fig. 6). The pathological lesions in livers and spleens were much severer at day 14 pi, specifically with high dose of C. burnetii. The pathological lesions in livers and spleens of mice infected with C. burnetii via IP route were observed at day 7 pi but they were lighter than those in mice infected via IT route.

Serologic response to infection of C. burnetii
To better understand the humoral immune response of mice against C. burnetii infection via IT route, we determined the antibody levels of IgGs to phase I antigen and to phase II antigen by ELSIA. As shown in Fig. 7, the antibody levels to phase I antigen (Fig. 7a) and to phase II antigen (Fig. 7b) in sera from mice infected with C. burnetii via both IT and IP route increased gradually throughout the infection. The antibody levels to phase II antigen were higher than those to phase I antigen, indicating an acute stage of C. burnetii infection. At the same time point, the antibody levels of mice infected via IT route were not significantly different from those of mice infected via IP route.

Discussion
The animal models of acute C. burnetii infection, simulating the natural route of infection and clinical presentations associated with human acute Q fever, had been developed with guinea pigs, mice, or nonhuman primates, and mice are the most attractive for being easy to perform for various research purposes in laboratories [17,28,29]. Previous studies showed that the aerosol inhalation procedure of infection in mice is of relevance to the majority of the acute Q fever cases in human, and the clinical signs, bacterial burdens, and pathological alterations in certain mouse models may be used to monitor progression of the disease. The whole-body aerosol inhalation procedure was used to prepare the mouse models of the acute C. burnetii infection [12,17,30], but it may cause systemic dissemination of the infection in mice by other routes besides the respiratory tract and also its inoculum dose of C. burnetii is not standardized. In the present study, the mouse model of the acute C. burnetii infection was developed via IT inoculation to directly place the living C. burnetii into the tracheas of the immunocompetent BALB/C mice using a noninvasive aerosol pulmonary delivery device.
In this study, high levels of bacterial loads were determined in all four organs at any time point pi, and the highest levels of bacterial loads in these organs were determined at day 7 pi, suggesting C. burnetii was rapidly disseminated from the lungs to the deep organs of the mice and demonstrating C. burnetii grew stably and proliferated rapidly. Similar tendency of bacterial loads were determined in spleens of BALB/c mice infected by C.
burnetii aerosolization [12]. Compared to other organs of mice infected with C. burnetii via IT route, the bacterial load in spleens was the highest, indicating a larger amount of organisms moved into the spleens through the circulatory system and then were engulfed by phagocytic cells and proliferated in the host cells in spleens [31]. In addition, the bacterial loads of mice infected via IT route at day 14 pi was lower than that at day 7 pi, suggesting that the immune system of the mice was activated to resist the infection and the mice were recovered in the late phase of the infection.
The pneumonia is a common complication during the acute C. burnetii infection in human because the lung is the earliest target organ of C. burnetii in natural infection [32]. In the mice infected with C. burnetii via IT route, the lungs were the foci of infection and the pathological lesions in lungs were more pronounced than those in livers and spleens.
Isolated hepatitis is a frequent presentation of the acute C. burnetii infection in human.
Besides granulomatous hepatitis and typical "doughnut" granulomas, atypical pathological changes were reported, such as epithelioid granuloma extensive extravasated fibrin and acute cholangitis without granuloma [33]. The infiltration of inflammation cells in livers of mice infected with C. burnetii via IT route, suggesting that the severe pathological lesions in livers of mice.
The cases of pericarditis, acute myocarditis, and acute endocarditis were reported in acute C. burnetii infection in human [34]. A large amount of C. burnetii organisms were determined in hearts and many inflammatory cells in epicardium and the infiltration of inflammatory cells in myocardium were observed of mice infected with C. burnetii via IT route at day 3 pi, suggesting early epicarditis in mice and the pathological lesions in hearts of mice were compatible with the heart diseases appearing in acute C. burnetii infection in human. It is speculated that the C. burnetii organisms arrived at the bronchi directly and rapidly reached the hearts through the pulmonary aorta without extensive circulatory system in mice infected with C. burnetii via IT route. The pathological lesions were consistent with the levels of bacterial loads in the organs of the mice, proving that the infection of C. burnetii spread hematogenously from the lungs to other organs such as the livers and spleens.
The influence of the inoculum dose in pathological lesions during C. burnetii infection was previously studied in guinea pigs [21]. A dose of 1×10 8 C. burnetii organisms was required to cause pathological lesions at the early phase of infection in the organs of mice infected by C. burnetii aerosolization [5,9], suggesting that a large amount of C. burnetii organism must be present in aerosols for the development of Q fever. In the present study, mice were infected with a low dose of C. burnetii organisms (1×10 4 per mouse) via IT route, and the pathological lesions were observed in hearts and lungs as early as at day 3 pi and in all four organs at day 7 pi, suggesting the IT route was more efficient to cause acute C.
burnetii infection in mice compared to aerosolization.
The specific IgGs to phase I and II antigen of C. burnetii were detected in sera from mice at day 7 and 14 pi even with the low dose of C. burnetii infection via IT route. The levels of the specific IgGs of mice infected with C. burnetii via IT route were not significantly different from those of mice infected via IP route, suggesting that the infection with C.
burnetii via IT route elicited specific immune responses against C. burnetii like that via IP route.

Conclusions
In conclusion, the results in the present study confirm that the BALB/c mouse model of acute C. burnetii infection was successfully established by IT inoculation with C. burnetii using a non-invasive aerosol pulmonary delivery device. C. burnetii was accurately delivered into the tracheas of the mice, which modeled the pulmonary route of acute C.    Histopathological analysis in spleens of mice infected with C. burnetii by IT route or IP route. BALB/c mice received an inoculum of 1×108 C. burnetii Xinqiao strain via IT route or IP route. Red pulp (shown with black arrows) widened and white pulp (shown with red arrows) atrophied, and splenic sinusoid extended with hyperemia, and the boundary became blurred in the spleens of mice infected with C. burnetii via IT or IP route at day 7 pi. Original magnification, 100 ×. Scale bars, Figure 7 Antibody response in sera of mice infected with C. burnetii by IT route or IP route. BALB/c Mice received an inoculum of 1×108 C. burnetii Xinqiao strain via IT route and IP route, respectively. Normal mice without infection were regarded as a negative control. Mice were killed at day 3, 7, and 14 pi, respectively. C. burnetiispecific IgG responses in sera were measured. OD450, optical density at 450 nm. a: the antibody levels to phase I antigen; b: the antibody levels to phase II antigen.

Supplementary Files
This is a list of supplementary files associated with the primary manuscript. Click to download.