2.2.7. At 35/37 dpi

Lungs had a normal appearance (Figure 8A). Only a few, very small foci of soft, sunken, dark red pulmonary tissue (2% of total pulmonary volume in one calf and less than 1% in two calves) associated with mild focal fibrous pleuritis were found (Figures 1 and 8B).

**Figure 8.** Pulmonary lesions at 35/37 dpi. (**A**) Gross appearance of lung after regeneration. (**B**) Very small lesion in a middle lobe (short arrow) and focal fibrous pleuritis (long arrow) between middle lobe (M) and basal lobe (B). (**C**) Normal pulmonary morphology. HE-stain. (**D**) Small lesion surrounded by normal lung tissue (stars); lesions are characterized by infiltrates of macrophages and multinucleated giant cells (arrow), hyperplastic BALT (B, examples) and tertiary lymphoid tissue (T). HE-stain. (**E**). Higher magnification of the area indicated by the arrow in Figure 8D: infiltrates of macrophages and multinucleated giant cells (arrows, examples), and hyperplastic BALT (B). HE-stain.

By histology, airways, blood vessels and alveolar compartment had completely normal morphology in most of the pulmonary tissue (Figure 8C). BALT and diffuse immune cells had regressed to normal amounts. There were no tertiary lymphoid tissue and no signs of fibrosis.

The small lesions were characterized by infiltrates of macrophages and multinucleated giant cells embedded in vascularized fibrotic tissue (Figure 8D,E). Lymphocytic infiltrates were organized as peribronchial and perivascular BALT, and tertiary lymphoid tissue in the interlobular septa. Chlamydial inclusions were not detected.

#### **3. Discussion**

A bovine model of respiratory infection with the zoonotic pathogen *C. psittaci* was previously established [19]. This model has been used to investigate the course of disease, effects on pulmonary functions as well as shedding, transmission and zoonotic potential of *C. psittaci* [20–22]. The infection model was very efficient, since all calves inoculated with the pathogen developed pneumonia, but none of the controls. The model was highly reproducible, and therefore repeatedly used to examine outcomes of different treatment regimens [23,24]. The localization of lesions corresponded to the inoculation sites and was distinct from field infections that have a cranioventral distribution [25]. Thus, the model consistently allows induction of lesions at predetermined sites of the lung.

For this qualitative histological study, groups of 3 of the 21 calves inoculated were necropsied at predetermined days post inoculation. Although the number of animals is low, the validity of observations is supported by the concordance of histological findings and consistent progression patterns. The morphological evaluation of the lung at sequential time points after inoculation revealed the potential of this model to study tissue reactions from initial stages of inflammation to extensive necrosis, up to stepwise organization and regeneration. The reactions observed are matching well with the steps reported as general reactions of pulmonary tissue to an insult [26]: primary lesions are often amplified through inflammatory host reactions. If the host is capable of limiting damage and proliferation of the infectious agent, organization begins. This may result in reparation, resolution or remodeling. Reparation implies repopulation with epithelial cells, but dysfunctional tissue structure. In a remodeled lung, pulmonary tissue is replaced with connective tissue, which is dysfunctional and may cause ectasia and emphysema of adjacent tissue. A successful return to normal structure and function is termed resolution or regeneration.

In the model presented, the initial insult was the intrabronchial inoculation with *C. psittaci,* which has a tropism for epithelial cells [27]. This resulted in a limited infection of AEC1. Infected alveolar epithelial cells have the potential to produce and release cytokines, e.g., the proinflammatory cytokine IL8, which is highly chemotactic for neutrophils, chemokines and growth factors initiating the augmentation [28,29]. Chlamydiae are released during the replication cycle from AEC1 into the alveolar lumen where they get into contact with alveolar macrophages. Bacterial LPS may induce necrosis of alveolar macrophages and release of IL-1α, which lead to a loss of vascular integrity and thus increase the influx of neutrophils from the pulmonary vasculature into the alveoli and airways [30]. Neutrophils display unique migration mechanisms in the lung resulting in particularly high numbers [31]. Inflammatory mediators enhance neutrophil activity and their deleterious effect on endothelium and epithelium [32,33]. The severe exudation and the necrosis of bronchiolar epithelium observed in the calves are most likely due to host reactions and not an effect of chlamydial replication, since no chlamydial inclusions were present in vascular endothelium or airway epithelium.

The massive release of protein-rich exudate and fibrin as seen at 3 dpi and 4 dpi is a common reaction in cattle and has been attributed to an imbalance of pro- and antifibrinolytic factors [34]. Fibrin can further increase vascular permeability, influence the expression of inflammatory mediators and alter migration and proliferation of various cell types [35]. Fibrin thrombi may occlude capillaries and arterioles and thus reduce perfusion; fibrinous exudate within alveoli prevents ventilation. This results in tissue hypoxia and, eventually, necrosis as observed as fibrinonecrotic bronchopneumonia with highest severity at 4 dpi.

While the number of neutrophils continuously declined, the number of macrophages increased during this initial phase. Airway macrophages possess high phagocytic activity to remove debris and exudate, which also included chlamydiae. Processing and presentation of chlamydiae and chlamydial antigens by macrophages may enhance the pathogen-specific immune response. Macrophages also play an important role in the downregulation of immune response and in tissue repair [30,36]. Sloughed epithelial cells, dying neutrophils and microvesicles as encountered in areas of necrosis comprise a rich depot of phosphatidyl serines that can reprogram macrophages from a proinflammatory to an antiinflammatory and prorepair state [9].

In patients with localized destructive processes such as necrotizing pneumonia, which may progress to cavitating lesions, it has been postulated that there is no return to preexisting tissue architecture [37]. In our model, tissue necrosis but no cavitation occurred. Under field conditions, fibrinonecrotic bronchopneumonia in domestic animals rarely resolves completely and sequelae, like sequestra, abscesses, gangrene, fibrosis, scars or chronic pleuritis, are frequent [38]. This may be due to the ability or failure of the host to clear the pathogen and terminate chronic inflammation. After the initial infection of AEC1, chlamydial inclusions were present in continuously decreasing numbers in neutrophils and macrophages. At 14 dpi, they had been almost cleared. From 7 dpi to 14 dpi, the infiltration of lymphocytes representing an adaptive immune response continuously increased [39]. Besides diffuse lymphoid infiltrates, organized structures developed as tertiary lymphoid tissue in the interalveolar septa and as BALT around airways. This local immune reaction was transient and had disappeared at 35/37 dpi.

At the same time, small aggregates of epithelial cells, as first signs of tissue repair, were observed at the edge of necrosis. The pulmonary epithelium does not exhibit a constant turnover, but can respond robustly after injury to replace damaged cells. In the alveolar niche, especially type 2 alveolar epithelial cells (AEC2) have an enormous reparative potential [40–42]. They clonally generate more AEC2 [43]. The cuboidal AEC2 grow, proliferate, follow the basement membrane and secrete basement membrane components, but are inefficient for gas exchange [44].

The multifocal distribution of epithelial cell regenerates at the edge necrosis may be attributable to the fact, that a few AEC2 were retained at these sites and served as starting points [43]. This process may have been supported by remnants of basal membrane, an important scaffold for AEC2 [45]. Recently it was shown that even functionally mature AEC1 could replicate and generate AEC2 [46]. Thus, the delicate epithelial cells observed at 7 dpi might also represent surviving AEC1. Cuboidal epithelial cells interpreted as AEC2 increasingly replaced areas of necrosis. The process started at the periphery and moved towards the center of necrotic tissue leaving only a few small foci of necrosis at 14 dpi. Efficient epithelialization is likely to be crucial for the prevention of pathologic lung remodeling [26].

The cuboidal epithelial cells formed organized alveolar spaces. Maintenance of alveolar units requires complex interactions between various cell types including epithelial cells, endothelial cells, mesenchymal cells, macrophages and other immune cells, and a great variety of mediators [9,47]. Vascularization in areas of re-epithelialization can originate from multipotent mesenchymal stem cells or from blood vessels in adjacent tissue by sprouting. Capillary sprouts originating from the connective tissue of interlobular septa adjacent to areas of necrosis were seen as early as 10 dpi. The sprouting process involves several specialized types of endothelial cells. Vascular endothelial growth factor or inflammatory cytokines induce the formation of tip cells that use filopodia to sense environmental cues, translate these into dynamic processes and express matrix metalloproteinases for invasion [48]. Tip cells use extracellular matrix as a scaffold, provided in this case by the epithelial cells and basement membrane. Stalk cells elongate the sprouting vessels, form the lumen and connect it to the circulation, while phalanx cells stabilize the new vessels and optimize their function [48]. The combination of vascularization and epithelialization contributed to the successful neo-alveolarization observed. The final step in this process is the maturation of cuboidal alveolar epithelial cells to AEC1 and thus functional alveoli [8,40]. This took place between 14 dpi and 37 dpi when the last calves were examined.

The fibrinous exudate and growth factors, e.g., transforming growth factor β, released from inflammatory cells, such as macrophages, and endothelial cells induced immigration of fibroblasts and production of collagen and extracellular matrix. The first fibroblasts and collagen fibers were observed at 7 dpi and distinct fibrosis around arterioles and in the interalveolar septa at 14 dpi. Since there was no permanent or progressive fibrosis in our model, it does not reflect pulmonary fibrosis in humans [37,49], but it is an example that the development of fibrosis can be prevented. The time course of development and resolution of fibrosis in the calves was comparable to that in bleomycin-induced lung fibrosis in mice, where the maximal extent of fibrosis was found at 14 days after a single instillation and subsided after treatment was withdrawn [50].

There was a good correlation between clinical data, deterioration of lung function and pulmonary lesions (Table 1). The more extensive lesions during the initiation phase were associated with systemic signs, e.g., fever, increased respiratory rate and decreased tidal volume. As soon as the volume of pulmonary lesions decreased and organization started, lung function and clinical signs returned to normal values. This occurred before regeneration was complete, because cattle ventilate only about 30% of the total lung volume when breathing spontaneously under resting conditions [51].

In conclusion, the experimental infection of calves with *C. psittaci* allows (1) to dissect the tissue processes involved in the development and resolution of inflammation, lesion development, neo-alveolarization, revascularization, waxing and waning of immune cell infiltrates, and fibrosis in the lung, (2) to correlate clinical with morphological findings and (3) to investigate the influence of treatment regimens on lung regeneration. These qualitative histological data may serve as basis for further in-depth studies using more advanced methods.

#### **4. Materials and Methods**

#### *4.1. Animals*

Forty-two conventionally raised calves (Holstein-Friesian, male) were included in this study. Animals originated from a farm without history of *Chlamydia*-associated health problems. Before the study, the herd of origin was regularly tested for the presence of Chlamydiae by the National Reference Laboratory for Psittacosis. Calves were purchased at the age of 14–28 days weighing between 42.2 and 71.2 kg. Animals were included in the study after a quarantine period of at least 20 days and confirmation of a clinically healthy status.

Throughout the entire study, animals were reared under standardized conditions (room climate: 18–20 ◦C) and in accordance with international guidelines for animal welfare. Nutrition included commercial milk replacer and coarse meal. Water and hay were supplied ad libitum. None of the given feed contained antibiotics.

This study was carried out in strict accordance with European and National Law for the Care and Use of Animals. The protocol was approved by the Committee on the Ethics of Animal Experiments and the Protection of Animals of the State of Thuringia, Germany (Permit Number: 04-002/07, 18 December 2007). All experiments were done in a containment of biosafety level 2 under supervision of the authorized institutional Agent for Animal Protection. Bronchoscopy to inoculate the pathogen was strictly performed under general anesthesia. During the entire study, every effort was made to minimize suffering.

#### *4.2. Experimental Design*

Twenty-one calves were inoculated with *C. psittaci* and twenty-one calves served as controls. The inoculum was placed at eight defined pulmonary sites by bronchoscope. Each calf underwent daily clinical examination. Groups of three calves were euthanized at 2, 3, 4, 7, 10, 14 and 35 or 37 dpi. At necropsy, distribution and extent of pulmonary lesions were determined and samples collected for histological, immunohistochemical and microbiological investigations.

#### *4.3. Inoculum and Inoculation*

*C. psittaci* strain 02DC15 was isolated at the Friedrich-Loeffler-Institut, Jena, Germany, from an aborted calf fetus in 2002. The isolate was classified as *C. psittaci* genotype A-VS1 by DNA microarray testing and *omp*A gene sequencing [52]. Chlamydiae were propagated in buffalo green monkey kidney cell culture using standard procedures [53]. Frozen stocks of strain 02DC15 were diluted to the required titer in stabilizing SPGA medium and used as inoculum in the present trial.

Calves were inoculated with 10<sup>8</sup> ifu of *C. psittaci* strain 02DC15 in 6 mL stabilizing medium SPGA (containing saccharose, phosphatile substances, glucose and bovine albumin) [54]. Control calves received the same amount of SPGA containing buffalo green monkey kidney cells.

The inoculation by bronchoscope has been described in detail [19,55]. In brief, non-fed calves were anesthetized with xylazin (0.2 mg/kg body weight, Rompun 2%, Bayer Vital GmbH, Leverkusen, Germany) and ketamine (1.7 ± 0.3 mg/kg body weight, Ursotamin, Serumwerk Bernburg AG, Bernburg, Germany). A flexible video endoscope (Veterinary Video-Endoscope PV-SG 22–140, Karl Storz GmbH and Co.KG, Germany) was inserted through the oral cavity and 0.5–1.5 mL of inoculum were placed at eight defined pulmonary sites (Figure 9).

**Figure 9.** Schematic drawing of bovine lung. Inoculation sites are indicated as yellow stars (small—0.5 mL, middle-sized—1.0 mL and large—1.5 mL), sampling sites as red bars and and regional lymph nodes in green.

#### *4.4. Clinical Scoring and Pulmonary Function Testing*

Clinical observations were recorded twice daily and included feed intake, rectal temperature, respiratory rate and the presence or absence of clinical signs of diarrhea or respiratory disease. In addition, the appearance of oral mucosa, conjunctivae, skin, hair and dyspnea were assessed daily, and the heart rate was counted. Extremities, umbilicus and mandibular lymph nodes were palpated and inducement of cough was tested. Results were summarized using a 49-point clinical score consisting of subscores for general condition (maximum 8 points), respiratory system (maximum 17 points), cardiovascular system (maximum 13 points) and other organ systems (maximum 11 points) as described [19].

Pulmonary function testing was performed using the impulse oscillometry system (IOS) as described elsewhere [22].

### *4.5. Necropsy, Gross Pathology and Tissue Samples*

Always three calves inoculated with chlamydiae and three controls were euthanized at 2, 3, 4, 7, 10, 14 and 35/37 dpi. In deep anesthesia (pentobarbital-sodium, 770 ± 123 mg/10 kg body weight, intravenously, Release, WdT eG, Garbsen, Germany), the trachea was exposed and large clamps were placed distal to the larynx to prevent contamination of the airways by blood or gastric contents. Subsequently, animals were sacrificed by exsanguination. The lung was removed and macroscopic lesions were recorded. To determine the total percentage of pulmonary lesions, the percentage of lesions was subjectively assessed for each lobe and multiplied by the relative percentage of the respective pulmonary lobe of the total lung volume. The volumes had been determined in 18 age-matched calves prior to this study and were measured as displacement of water. For this, the total lung volume was determined; then the different lobes were dissected, the main bronchi closed and their volume measured. On average, the left cranial and caudal apical lobes contributed 5%, the left caudal lobe 34%, the right cranial apical lobe 5%, the right caudal apical lobe 7%, the middle lobe 8%, the right caudal lobe 35% and the accessory lobe 1% to the total lung volume (Figure 9).

Samples were collected from each lung lobe (Figure 9) and fixed in 3.5% neutral buffered formalin. Sites with macroscopic lesions were preferentially sampled. Then a complete necropsy was performed.

#### *4.6. Histopathology*

Formalin-fixed tissues were embedded in paraffin after 24 h. Formalin-fixed paraffin-embedded (FFPE) tissue sections were stained with hematoxylin and eosin to evaluate lesions. Consecutive sections were stained with azan to demonstrate collagen fibers and by PAS-reaction for glycogen-rich material.

#### *4.7. Immunohistochemistry*

Consecutive FFPE tissue sections were used to label chlamydiae, epithelial cells and blood vessels by the indirect immunoperoxidase method. As primary antibodies, the anti-chlamydial-LPS antibody ACI-P500 (Progen, Heidelberg, Germany), anti-cytokeratin antibody MNF116 (Dako Denmark, Glostrup, Denmark) and anti-factor VIII polyclonal antiserum (Dako Denmark, Glostrup, Denmark) were used. Peroxidase-labeled sheep anti-mouse IgG (NA 931, GE Healthcare Europe GmbH, Freiburg, Germany) served as secondary antibody for the monoclonal antibodies and peroxidase-labeled goat anti-rabbit IgG for the polyclonal antiserum. Sections were digested with 0.05% proteinase K (Merck, Darmstadt, Germany) for antigen retrieval. Diaminobenzidine was used as chromogen.

#### *4.8. Exclusion of Co-Infections*

The herd of origin was known to be free of bovine herpes virus 1 and bovine virus diarrhea/mucosal disease virus. Routine microbiological screening revealed that all animals were negative for *Salmonella* infections (fecal swabs) and relevant enteric parasites (fecal smears). To verify relevant respiratory co-pathogens, the presence of *Mycoplasma*, *Pasteurella* or *Mannheimia* spp. was evaluated in nasal swabs taken immediately before challenge and before necropsy as well as in lung tissue samples obtained during necropsy. Neither *Mannheimia haemolytica* nor *Mycoplasma bovis* was detected in any sample. *Pasteurella multocida* and *Mycoplasma bovirhinis* were detected in nasal swabs, but never in any lung tissue sample. Serological findings confirmed that animals did not acquire infections with respiratory viruses relevant in bovines (i.e., bovine respiratory syncytial virus, parainfluenza 3 virus or adenovirus type 3).

#### **Supplementary Materials:** Supplementary Materials can be found at http://www.mdpi.com/1422-0067/21/8/2817/ s1.

**Author Contributions:** Conceptualization, E.M.L.-T. and P.R.; Formal analysis, J.L.; Funding acquisition, K.S. and P.R; Investigation, E.M.L.-T., J.L., C.O. and P.R.; Methodology, E.M.L.-T., J.L., C.O., K.S. and P.R.; Project administration, K.S. and P.R.; Resources, E.M.L.-T., K.S. and P.R.; Supervision, E.M.L.-T.; Visualization, E.M.L.-T. and J.L.; Writing—original draft, E.M.L.-T.; Writing—review and editing, J.L., C.O., K.S. and P.R.. All authors have read and agreed to the published version of the manuscript.

*Int. J. Mol. Sci.* **2020**, *21*, 2817

**Funding:** This study was financially supported by the Federal Ministry of Education and Research (BMBF) of Germany under Grant no. 01 KI 0720 "Zoonotic *Chlamydia*—Models of chronic and persistent infections in humans and animals".

**Acknowledgments:** The authors are very grateful to Annelie Langenberg, Sylke Stahlberg, Ines Lemser, and all colleagues of the technical staff of the animal house for their excellent assistance while performing the in vivo phase of this study. We thank Sabine Scharf, Christine Grajetzki and Simone Bettermann for excellent technical assistance in preparation of the inocula, PCR and related techniques. Furthermore, we wish to express our gratitude to Kerstin Heidrich, Sabine Lied, Monika Godat for skillful technical assistance in the ex vivo phase, and Wolfram Maginot for excellent photographic support. Help in microbiological testing given by Ulrich Methner and Silke Keiling, Martin Heller and Susann Bahrmann, Mandy Elschner, Astrid Rassbach and Katja Fischer is gratefully acknowledged.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
