Thus, the bovine lung does present an advantageous model to assess pathophysiological consequences of both airway obstructions and pulmonary restrictions. (ii) This large animal model offers the great potential to perform non-invasively and almost painless long-term studies allowing a simultaneous within-subject approach of functional changes of both the organ and the systemic level.
In this particular model signs of acute respiratory illness were maximal 2–4 dpi and did not last longer than one week after challenge [
38]. Although the severity of illness was mainly driven by respiratory signs, additional systemic reaction, similar to atypical pneumonias in human medicine [
39] were observed. The most striking symptoms included dry cough, tachypnea, fever, reduced appetite, and tachycardia [
38].
Pulmonary dysfunctions
Pulmonary function techniques from human medicine were applied to spontaneously breathing animals with body weights comparable to adult humans. Thus basic parameters of pulmonary functions (for example airflows and lung volumes) were more comparable and transferable to human patients compared to results obtained from murine models.
Furthermore, a functional differentiation between airway resistance and tissue mechanics of the lung was possible in the present study, while assessment of compliant properties of the lung in mice would require medical or surgical treatment [
41,
42]. During the acute period of illness (3–4 dpi) due to
C. psittaci infection, the pattern of breathing was characterised by a significant decrease in tidal volume and a significant increase in both respiratory rate and airflows (clinically seen as short and rapid breathing cycles, i.e. dyspnoea). Per breath, dead space volume was significantly elevated while alveolar volume was reduced by about 10% indicating alveolar hypoventilation that was confirmed by a decreasing end-tidal concentration of CO
2. However, global hypoventilation was not confirmed. In contrast, partial pressure of CO
2 in peripheral blood decreased, too, indicating hypocapnia due to global hyperventilation. The latter was caused by an increase in minute ventilation by 50% due to the strong increase in respiratory rate. The elevated minute ventilation was most likely the attempt to compensate for hypoxaemia induced by
C. psittaci infection as shown previously [
40].
Alterations in respiratory mechanics after inoculation of
C. psittaci included both obstructive and restrictive components and lasted longer than the clinically visible changes in the pattern of respiration. Restriction was assessed by decreasing respiratory reactance (Xrs) which indicates limitations in elasticity or compliance of the lung-thorax system [
23]. This loss in elasticity was predominantly a result of inflammatory reactions, such as cell infiltration, accumulations of fibrin and protein-rich fluid or signs of regeneration described for this model in detail elsewhere [
7,
43]. In the present study, the statistically significant decrease of Xrs at all frequencies (3–15 Hz) in
C. psittaci challenged calves continued until 11 dpi. Thus, the duration of reduced compliant properties of lung tissue exceeded the presence of acute clinical signs [
38] by about one week.
In the acute phase of respiratory illness, the loss of pulmonary compliance was most likely accompanied by stiffness of the peripheral respiratory system due to small airways narrowing or constriction [
44]. Indeed, peripheral airflow was limited in calves inoculated with
C. psittaci compared to control calves during the acute phase of disease which was indicated by an increase of Rrs ≤ 5 Hz and Rdist (significant at 3 dpi and at 3–4 dpi, respectively). The negative frequency dependence of Rrs, i.e. an increase only at low frequencies (Rrs < 5 Hz), is a valid diagnostic tool to identify peripheral airways obstruction in both humans [
45] and calves [
21]. In addition to obstruction in distal airways, calves experimentally challenged with
C. psittaci also suffered from obstruction in central or upper airways as indicated by an increase Rprox. These finding are in good agreement with reports in literature associating chlamydial infections in calves with both upper respiratory tract disease [
3] and obstruction of peripheral airways [
15]. Moreover, it was also shown in experimentally
C. suis challenged pigs that peripheral airways obstruction during the acute phase (3 dpi) were followed by upper airways obstruction (at 7 dpi). To our knowledge lung function data of humans suffering from acute chlamydial pneumonia are not available, but taking these findings together obstruction of the upper and lower respiratory tract might probably also be involved in pathogenesis of acute chlamydial pneumonia in humans.
In parallel to the presence of airways obstruction, FRC increased significantly at 3–4 dpi. Baseline data of about 40 mL/kg b.w. measured in this study in calves are in good agreement with data reported for the healthy bovine lung in adult cows (38.6 ± 3.1 mL/kg; [
46]). After
C. psittaci-infection, FRC increased significantly by 17% to 45.7 mL/kg b.w. (3 dpi) which is moderate compared to FRC data reported in cows with severe bronchiolitis and an expanded lung field (56.5 ± 7.7 mL/kg; [
47]). In calves, due to the lack of collateral airways, the presence of fibrin, inflammatory cells, detritus and protein rich fluid in the airways and/or alveoli during the acute phase of this model [
43] resulted in narrowed peripheral airways which can easily result in the development of trapped air. In the present model the increase in FRC was transient, thus hyperinflation or over-distension of alveoli is indicated rather than the presence of emphysema [
48]. As over-distension might reduce the recoil of elastic fibers it is likely that hyperinflation also contributed to reduced lung compliance described above. An increase of FRC was also reported for
C. suis infected swine [
49]. Radiographically-impressive distension of the lung with air is found in cases of
C. trachomatis pneumonia in children, which despite the mild respiratory symptoms in infancy is associated with obstructive limitations up to 7–8 years after hospitalisation (i.e. increased FRC, forced and peak expiratory flow rates) [
50,
51]. A long-term impairment of lung function and structure after chlamydial infection was also shown for naturally
Chlamydia-infected calves [
15] and experimentally challenged mice [
52]. In human medicine, asthma is a common chronic inflammatory disease of the airways, and the involvement of
C. pneumoniae in asthma pathogenesis is still largely discussed [
53,
54].
None of the lung functions assessed in control animals was significantly influenced by intrabronchial inoculation of BGM cell suspension. Pulmonary function data in control calves revealed physiological changes over time due to lung growth and development (gain in body weight during the study was 0.6 kg per day in average). In control calves Xrs increased significantly over time, displaying increasing compliant properties of lung and thorax. These findings are in line with fundamental understanding from the very beginning of veterinary pulmonology showing that lungs are easier to stretch with enhanced body or lung size [
55,
56]. It has been shown for growing calves that Xrs increased with increasing body weight [
57]. During the period of pulmonary maturation (until a body weight of about 300 kg [
58,
59]) bronchiolar diameters were also shown to increase [
60], resulting in decreased airway resistance.
Acid–base imbalances
Compared to other studies [
28,
61‐
63] the control values of pH, pCO
2, HCO
3–, base excess and A
tot (Alb) or A
tot (Prt) are in the ranges reported whereas AG and SIDm
3,
4 were lower and SIG
(Alb) or SIG
(Prt) were higher in absolute values than those described in literature. In calves experimentally infected with
C. psittaci, most of the effects assessed in venous blood were slight or moderate in amplitude and were mostly related to either the acute phase (2–4 dpi) or the resolution phase (7–10 dpi) after inoculation of the pathogen. Nevertheless the investigated parameters accurately assessed the influence of
C. psittaci on the acid–base balance of the host organism.
Partial pressure of CO
2 provides information regarding ventilation or respiratory component of acid–base balance in the Henderson-Hasselbalch equation as well as in the strong ion approach. Despite no access to arterial blood in this study, venous blood was informative enough identifying venous hypocapnia (pCO
2(v)↓) 2–3 dpi as a result of hyperventilation (the latter was proved by pulmonary function testing). In general, hyperventilation can be caused primary by stimulation of pulmonary nociceptive receptors related to pulmonary disease and impairment of gas exchange (hypercapnia, hypoxaemia) or secondary for recovery from metabolic acidosis [
64]. Natural compensatory mechanisms probably never overcompensate, and as a general rule, the pH will vary in a direction similar to the primary component disorder [
65]. Therefore, it is more plausible that hyperventilation occurred to compensate for hypoxaemia, a known consequence of experimentally induced pulmonary disease in this model as reported previously by our group [
40]. As a result, blood pH increased slightly 2 dpi. Decreases in both cHCO
3- and cHCO
3-(st) at 3 dpi, together with decreased cBase and cBase (Ecf), can traditionally be interpreted as compensatory mechanisms to return to normal pH. In the period 7–10 dpi, cHCO
3-, cBase and cBase (Ecf) increased but pH was not influenced. In conformity with the more modern approach, cHCO
3-, cBase and cBase (Ecf) are described as dependent (strong ion) variables that cannot be regulated independently of pCO
2, while SID and A
tot are independent variables [
66]. Only the independent variables influence the system and they are not influenced by the system. A
tot and SID reflect the metabolic system. A
tot (Alb) decreased 2–10 dpi and produced an alkalotic effect caused by hypo-albuminaemia. Albumin is a negative acute-phase protein, i.e. a marker of inflammation. This finding supplements our previously reported results identifying LBP (lipopolysaccharide binding protein) as a suitable marker of the acute phase in bovines [
38,
40]. In addition, albumin is the most important buffer in plasma [
29]. A
tot (Prt) was less affected because of hyper(gamma)globulinaemia, a spontaneous immune response. Our findings are in good agreement with data reported recently by Poudel et al. (2012) demonstrating that both the lowered plasma albumin and the increased globulin concentrations were associated with the intensity of
C. pecorum infection in calves, and were attributed to ongoing systemic inflammation and its detrimental effects on liver function caused by chlamydiae [
67].
SIDm
3, SIDm
4 and SIDm
5 decreased 3 dpi (acidotic effect) caused by hyponatraemia (without change in haematocrit; data not shown) which dominated the concurrent hypochloraemia (alkalotic effect). SIDm
3, SIDm
4 and SIDm
5 increased slightly 10–14 dpi due to continuing hypochloraemia and normalised sodium concentrations which led to a mild alkalotic effect as seen at the same time in a slightly increased pH. Effects of increased potassium and lactate were compensated and were not seen in SIDm
3. The adaptive retention of acid during sustained hypocapnia is normally accompanied by a loss of sodium into the urine [
68]. To maintain the electroneutrality in blood in the presence of hyponatraemia, the HCO3
– concentration must decrease concurrently. The measured values of cNa
+ and base excess agree with studies by Funk (2007) [
34], which showed that a decrease of sodium by 10 mmol/L explains a decrease of Base excess by −3 mmol/L.
This experimental study demonstrated again that often multiple acid–base disturbances exist concurrently and that mixed acid–base disturbances traditionally cannot be detected when the blood pH is unchanged. Strong ion theory provides evidence about the presence of acid–base imbalances, but only the selective view on the single parameters, which are required to calculate the strong ion variables, help to understand the complex response of the host organism and interactions between numerous variables.