**1. Introduction**

Pneumonia affects large proportions of the human population worldwide [1]. Community acquired pneumonia (CAP), which is defined as infection not acquired during hospitalization or ventilation, is the most common type of pneumonia and the most frequent cause of morbidity and mortality by infection in developed countries [2]. The annual incidence in adults is 1.5–1.7 per 1000 in Europe [3]. CAP can be caused by viral and bacterial pathogens. Infections with *Streptococcus pneumoniae* and *Haemophilus influenzae* are especially frequent [2,4], but a variety of other bacteria including the intracellular pathogens, *Mycoplasma pneumoniae, Chlamydia (C.) pneumoniae, C. psittaci, Coxiella burnetii* and *Legionella pneumophila* can also be involved [3]. Studies focusing on *Chlamydia* and *Mycoplasma* in patients with CAP indicate that these infections are more frequent than commonly reported, especially in children [5–7]. Infections with *C. psittaci* are about twice as frequent as with *C. pneumoniae* [6].

Infection models involving these pathogens are relevant to elucidate host reactions to the pathogen, development and resolution of tissue lesions and to evaluate treatment options. Only animal models can truly reflect the complex cellular interactions during lung injury and repair [8,9]. The search for appropriate animal models of respiratory disease in humans has a long history and is still ongoing [10]. Criteria to optimize experimental research were redefined recently [11]. While mouse models are most frequently used for being cost and time efficient and offering many options for genetic tracing and immunological monitoring, models in domestic animal species, as presented here, possess their own specific advantages, i.e., lung physiology and structure resembling more closely that of humans and spontaneous disease and lesions being comparable to those occurring in humans [8,11,12].

Symptoms of pneumonia are the result of pulmonary alveoli filling with exudate and thus preventing gas exchange. One of the complications of bacterial CAP is necrotizing pneumonia [13,14]. In humans, necrotizing pneumonia is most commonly seen in *Staphylococcus aureus*, *Streptococcus pneumoniae* or *Klebsiella pneumoniae* infections and may occur in both children and adults [13,15]. Conservative versus surgical treatment is subject to controversial debate [13,16,17]. While long-term effects with higher rates of mortality have been reported in elderly patients [18], full resolution within a few months occurred in most children [16].

In cattle and pigs, necrotizing lesions are observed in a number of bacterial pneumonias. The progression to necrotizing lesions may be due to virulence factors of microorganisms or to host reactions, which are more exudative especially in cattle [13]. Experimental infection of calves with *C. psittaci* was reported to progress to necrotizing lesions, with clinical course, pulmonary dysfunction and systemic host reactions having been well characterized [19–22]. In contrast to the general consensus that the outcome of pulmonary necrosis consists of fibrotic scars, sequestra or abscesses, complete healing was observed. In the following, the time sequence from tissue injury to regeneration is described based on qualitative histological data with the aim of providing a fundamental characterization of a necrotizing pneumonia model.

#### **2. Results**

#### *2.1. Clinical Signs, Acute Phase Response and Pulmonary Dysfunctions*

Details of clinical signs and pulmonary dysfunctions induced by the pathogen were reported elsewhere [19–22]. In brief, intrabronchial inoculation of 10<sup>8</sup> inclusion forming units (ifu) of *C. psittaci* per calf resulted in acute respiratory illness characterized by fever, dyspnea, dry cough, hyperemic conjunctivae and enlarged mandibular lymph nodes. Respiratory signs were accompanied by signs of a systemic inflammatory response, i.e., elevated heart rates (mild tachycardia), reduced appetite and dullness.

During the period of acute illness (i.e., 2–4 days post inoculation, dpi), blood gas analysis revealed hypoxemia. Pulmonary function testing indicated both obstructive and restrictive pulmonary disorders. The resulting pattern of spontaneous breathing was characterized by a reduction of tidal volume by about 25%, a doubling of respiratory rate, and consequently by a significant increase of minute ventilation to about 150%. Although acute clinical signs decreased and general health improved rapidly from 5 dpi onwards, alterations in respiratory mechanics, acute phase reaction (decreased blood concentrations of albumin and elevated blood concentration of lipopolysaccharide binding protein) and disorders in acid-base equilibrium lasted until 10–11 dpi. By the end of this study (37 dpi), the remaining three calves appeared clinically inconspicuous.

Data of rectal temperature, respiratory rate and tidal volume are given in Table S1 in absolute numbers, while relative changes of these parameters are included in Table 1.



<sup>1</sup> Absolut change of rectal temperature between the average rectal temperature measured at two different days in the week before challenge (baseline) and the rectal temperature measured at the day of necropsy. <sup>2</sup> Respiratory rate was counted in resting animals (in stable). The relative change of respiratory rate was calculated between baseline values (= mean of individual measurements at two different days in the week before challenge) and the respiratory rate counted at the day of necropsy. <sup>3</sup> body weight; <sup>4</sup> not available (calves necropsied at 2 dpi were too sick to undergo pulmonary function testing).

#### *2.2. Pulmonary Lesions*

#### 2.2.1. At 2 dpi

In two calves 17% of pulmonary tissue and in one calf 30% of pulmonary tissue were dark red, firm with a wet cut surface and pus draining from airways indicating purulent bronchopneumonia (Figures 1 and 2A). Lesions were located around bronchioles at the inoculation sites and often had cylindrical shape (Figure 2B). They were associated with mild fibrinous pleuritis in two calves.

**Figure 1.** Percentage of pulmonary tissue with lesions at 2, 3, 4, 7, 10, 14 and 35/37 dpi. Each point in the diagram represents one or two (2×) individual calves.

By histology, lesions had a lobular distribution with numerous neutrophils and protein-rich exudate in alveoli, bronchioli and bronchi (Figure 2C,D). The walls and perivascular spaces of small arterioles at the periphery of lesions were thickened by protein-rich edema, fibrin precipitates and neutrophils (Figure 2C,E). Fibrin thrombi obstructed blood vessels. There were areas of necrosis in the vascular walls. Lymphatics in subpleural and interlobular connective tissue contained fibrin thrombi and a few neutrophils. Low numbers of chlamydial inclusions were found in type 1 alveolar epithelial cells (AEC1) and occasionally in neutrophils and macrophages in the exudate (Figure 2F).

**Figure 2.** Pulmonary lesions at 2 dpi. (**A**) Macroscopic appearance and distribution of pulmonary lesions (dark red, arrows). (**B**) Section through the left basal lobe. A circumscribed, dark red and sunken lesion (thick arrow) is centered around a bronchus (thin arrow). (**C**) A bronchiolus (B) and the surrounding alveoli are distended by neutrophils. The wall of the arteriole (A) is infiltrated with neutrophils. Inset: higher magnification of neutrophils in an alveolus. HE-stain. (**D**) Numerous neutrophils in the lamina propria (LP), epithelium (E) and lumen (L) of a bronchus. HE-stain. (**E**). Fibrin precipitates (brown) surrounding a small arteriole (A). IHC, factor VIII. (**F**) Few chlamydial inclusions in type 1 alveolar epithelial cells (arrows, examples). Inset: higher magnification of a chlamydial inclusion. IHC, chlamydia.
