Structural alterations and markers of endothelial activation in pulmonary and bronchial arteries in fatal asthma

Asthma is a chronic inflammatory disease associated with structural changes in the lungs, mainly at the airway level [1]. Vascular changes in asthma have been mostly described at the submucosal capillary level of the bronchial microcirculation and include angiogenesis, vascular dilation and hyperpermeability. These features are thought to contribute to the maintenance of inflammation and airway thickening [2].

It has to be emphasized that the arterial supply of the lungs occurs via two different systems. The bronchial arteries (BAs) arise from the aorta or intercostal arteries and supply blood to the airways, tracheobronchial lymph nodes, and nerves. The pulmonary artery (PA) runs within the airway connective tissue sheath, branching with the airways. It supplies the distal lung and is considered the functional artery of the lung [3]. Little is known about human pathology of the pulmonary and bronchial arteries in asthma.

In animals, house dust mite (HDM)-induced allergic inflammation caused pulmonary vascular remodelling that did not revert entirely after allergen withdrawal [4]. A study using the Severe Asthma Research Program Cohort detected pruning of the pulmonary vasculature in asthma by computed tomography that was associated with lower lung function, higher eosinophilia and an increased rate of exacerbation [5]. However, few studies have addressed the structure of the pulmonary and bronchial lung arteries in asthma. Green et al. described thickening of the intimal layer associated with narrowing of the lumen of BAs, suggesting that it could contribute to bronchial hyperresponsiveness [6]. On the other hand, Saetta et al. did not detect thickening of the peripheral PA in six patients who died of acute onset asthma [7], despite the presence of eosinophilic/mast cell inflammation [8]. No studies have addressed the composition of the extracellular matrix (ECM) that forms the artery walls in human asthma.

Additionally, there are differences in leukocyte trafficking between the bronchial (airways) and pulmonary circulation (distal lung), possibly due to the type and size of vessels, function and expression of adhesion molecules and differences in stimuli that induce inflammatory responses at these sites [9]. A large epidemiological study has shown that asthmatic patients have an increased risk of developing (local or systemic) pulmonary thromboembolism, with higher risk in asthma patients with more exacerbations and hospitalizations [10]. Signs of endothelial activation in PAs [11] and an increase in coagulation factors in severe asthma have been previously described in asthma and could account for this epidemiological finding [12]. There is no information related to the tissue expression of adhesion molecules comparing bronchial and pulmonary circulation in asthma.

Cigarette smoking is associated with several adverse effects in asthma [13]. Cigarette smoking is also known to cause alterations in pulmonary vessel structure and endothelial cell dysfunction [14], but the extent to which vascular the alterations present in smoking asthma patients contribute to asthma severity is unknown. Endothelin-1 (ET-1), a potent vasoconstrictor, is involved in a pathway of tobacco-induced vascular dysfunction in smokers [15], and it is considered a mediator of asthma [16]. Angiotensin II type 2 receptor (AT2) has been described as protective against vasoconstriction, proliferation and fibrosis in models of acute lung injury and bronchodysplasia [17, 18], but there is no information about the expression of AT2 in the lungs of asthma patients.

To address these questions, we analysed structural ECM components (collagen I, collagen III, versican, tenascin, and fibronectin), adhesion molecules [vascular cell adhesion molecule 1 (VCAM-1) and Intercellular Adhesion Molecule 1 (ICAM 1)] and markers of vascular tone/dysfunction (ET-1 and AT2) in pulmonary and bronchial arteries of smoking and non-smoking patients who died due to an acute asthma attack. We hypothesized that bronchial and pulmonary arteries have differential expression of these markers and that the differences would be more pronounced in smoking asthma patients.

The subject’s characteristics are shown in Table 2. The median (range) age of the asthma patients and controls was 44 (17–66) and 52 (44–63) years, respectively (p < 0.001). Only 25% of the asthma patients were regularly followed by a doctor; 33% had been hospitalized due to an asthma exacerbation in the previous year, and 12% had had a previous intensive care unit admission due to asthma. Smoking asthma patients had a median of 20 (13–60) pack-years of smoking. There were no significant differences in the demographic characteristics of non-smoking fatal asthma and smoking fatal asthma subjects (data not shown).

The morphological characteristics of PAs and BAs are shown in Tables 3 and 4. The diameter of the analysed arteries was not different between asthma patients and controls. There were no differences in arterial wall and lumen areas when asthma patients (smoking and non-smoking subjects) and controls were compared for both PA and BA.

There were no differences between asthma patients and controls for collagen I, collagen III, tenascin, fibronectin, versican and elastic fibre content in both PAs and BAs (Table 5). Asthma patients had increased VCAM-1 content in the BA wall (p = 0.026) but not in PAs (Fig. 1). Smoking asthma patients presented with decreased content of collagen III in both BA (p = 0.046) (Fig. 2) and PA (p = 0.010) (Fig. 3) walls compared to non-smoking asthma patients. There were no differences in the expression of ICAM-1, AT2 (Fig. 4) and ET-1 in both types of arteries for (smoking and non-smoking) asthma patients and controls (Tables 5 and 6).

There were no significant correlations of VCAM-1 and collagen III expression with age, pack-years, disease onset and duration of asthma in asthma patients or controls (data not shown).

In this study, we analysed the structure, ECM composition and markers of endothelial activation in pulmonary and bronchial arteries of patients who died due to a fatal asthma attack. There were no significant differences in structural alterations or ECM composition between asthma patients and controls in both pulmonary and bronchial arteries. Smoking asthma patients had a decreased content of collagen III in both types of artery walls. VCAM-1 expression was increased only in the BA walls, suggesting that differences exist in endothelial activation between pulmonary and bronchial arteries in asthma. To our knowledge, this is the first study addressing the structure and markers of endothelial activation in pulmonary and bronchial arteries in asthma.

There has been a recent renewed interest in better understanding vessel pathology in asthma. Ash et al. described vascular pruning in severe asthma that was related to several parameters of disease severity [5], and previous animal studies suggested the presence of remodelling in the PAs in models of allergic sensitization [24]. We could not find differences in structure or ECM composition in pulmonary or bronchial arteries in asthma. Our data confirm previous data by Saetta et al. in distal pulmonary arteries of 6 cases of fatal asthma [7]. On the other hand, Green et al. found an increased thickness of the intimal layer in bronchial artery branches in asthma that was related to the duration of disease in asthma [6]. The reasons for the discrepant data are not clear and are possibly related to patient characteristics in both cohorts, such as the use of corticosteroids. Our data suggest that the mechanisms leading to vessel pruning are probably functional rather than anatomical but do not exclude the influence of perivascular inflammation on the control of vascular tone [8].

Smoking is associated with endothelial dysfunction and pulmonary vessel remodelling [14]. In this study, we found that patients with asthma who smoked had less collagen III in both artery types. In patients with chronic obstructive pulmonary disease, markers of collagen III degradation are related to overall mortality and clinical outcomes [25]. It would also be interesting to investigate whether lung collagen degradation in smoker contributes to increased severity of asthma.

Previous evidence indicated that asthma causes inflammation of pulmonary arteries, as measured by PA rings of asthma patients who underwent surgery [11]. We have previously described an increase in eosinophils, mast cells and neutrophils in the adventitial region of distal pulmonary arteries [8]. However, in this study, VCAM-1 expression was increased only in the BA walls of asthma patients but not in the PAs, and there were no significant differences in endothelin-1 and ICAM-1 expression. VCAM-1 is an adhesion molecule that is important for regulating the arrest and recruitment of circulating eosinophils during allergic inflammation [26, 27]. The mechanisms regulating leukocyte kinetics between pulmonary and bronchial circulation are likely to be different. These include differences in the diameter of the vessels, differences in the cell types and adhesion molecules that regulate the response, and differences in the stimuli that induce the inflammatory response at these sites [9]. Our data support the idea that leukocyte efflux in the microinflammatory milieu of airways occurs differently than in the distal lung in asthma, with the involvement of different adhesion molecules. Leukocytes come in closer contact with the endothelium for prolonged periods under normal conditions in smaller pulmonary vessels and capillaries, so the cascade occurring in systemic vessels may not be relevant for pulmonary capillaries [28].

In patients with acute severe asthma, the renin-angiotensin system is activated [29]. Angiotensin II has been implicated in the muscularization of pulmonary vessels during hypoxia [30, 31]. The expression of the AT2 receptor had not been studied in human lungs with asthma thus far. More specifically, in contrast to the AT1 receptor, the AT2 receptor is believed to have antifibrotic and antiproliferative effects [32]. The beneficial role of AT2 receptor could be associated with the absence of structural changes in arteries in asthma. We detected expression of the AT2 receptor in artery walls, but no differential expression was observed in asthma patients compared to that in controls.

This study has several limitations. We did not have full clinical and functional characterization of these patients to perform significant correlations, since most of these patients had no regular treatment or were inadequately treated [33]. Due to the retrospective nature of this study, we were not able to perform stereological sampling of lungs, which would have enabled us to perform vessel volume analysis. Nevertheless, autopsy material is thus far the most appropriate way to understand larger vessel pathology, rather than bronchial capillary pathology, in asthma.

No structural changes or differences in ECM compositions were observed in pulmonary or bronchial arteries of patients who died of asthma in comparison to those in controls. Taken together, our data suggest that vascular pathology in asthma is influenced by functional/systemic factors rather than local factors. Loss of collagen III in both types of arteries was observed in smoking asthma patients, which was compatible with the degradative environment induced by cigarette smoking. Our data also reinforce the idea that mechanisms of leukocyte efflux via adhesion molecules differ between bronchial and pulmonary circulation, which might be relevant to understanding and treating the distal lung in asthma.