Expression of Endothelin-1 in the Lungs of Patients with Pulmonary Hypertension

Study Design

Table 1. Table 1. Characteristics of the Patients with Plexogenic Pulmonary Arteriopathy.

Table 2. Table 2. Characteristics of the Patients with Secondary Pulmonary Hypertension.

The study protocol was approved by the institutional research and ethics committees of Montreal General Hospital and Royal Victoria Hospital, Montreal. Lung samples were obtained after surgical resection or at autopsy ( ≤ 10 hours post mortem). Specimens were obtained from 28 patients with a clinical diagnosis of pulmonary hypertension¹⁷. The patients were divided into two groups on the basis of clinical and histopathological characteristics. Group 1 consisted of 11 patients with plexogenic pulmonary arteriopathy (grades 4 through 6),¹⁸ of whom 7 had a clinical diagnosis of primary pulmonary hypertension according to the criteria of the National Heart, Lung, and Blood Institute¹⁹ (Table 1). Group 2 consisted of 17 patients with secondary causes of pulmonary hypertension and no more than grade 3 pulmonary arteriopathy¹⁸ (Table 2). The patients with secondary pulmonary hypertension were older than those with plexogenic pulmonary arteriopathy and had significantly lower forced expiratory volumes, diffusing capacities for carbon monoxide, and hemoglobin levels. However, there were no significant differences in pulmonary arterial pressure and vascular resistance between the two groups. Control specimens were obtained from 5 patients (3 men and 2 women, 19 to 67 years old [mean, 50]) with pulmonary diseases other than idiopathic pulmonary fibrosis or hypertension (e.g., interstitial pneumonitis) and from the unused normal lungs of 10 organ donors (6 female and 4 male donors, 17 to 46 years old [mean, 30]). Tissues were fixed in 10 percent buffered formalin or Bouin's solution and embedded in paraffin for routine pathological examination and immunocytochemical analysis. The method of fixation did not affect the intensity of immunostaining in five specimens that were prepared by both methods. For the localization of endothelin-1 messenger RNA (mRNA), tissues were either snap-frozen in liquid nitrogen or fixed in 4 percent paraformaldehyde in phosphate-buffered saline (0.1 mol of sodium phosphate and 0.15 mol of sodium chloride per liter, pH 7.2) and washed in phosphate-buffered saline containing 15 percent sucrose and 0.01 percent sodium azide at 4 °C.

Immunocytochemical Analysis

Two polyclonal antiserums against human endothelin-1 were used: one against the C-terminal of endothelin-1 and the other against the C-terminal fragment of big endothelin-1 (big endothelin-1_22-38). The two endothelin antiserums were raised as previously described²⁰. A commercial antiserum against human endothelin-1 (Peninsula Laboratories, Belmont, Calif.) was also used. In addition, antiserum to von Willebrand factor (factor VIII-related antigen) (Dako, Santa Barbara, Calif.) was used as an endothelial-cell marker. The avidin-biotin-peroxidase complex method²¹ was used as previously described¹⁴. Negative controls were prepared with the specific antiserum absorbed with the cross-reactive endothelins or with nonimmune serum instead of primary antiserum, or by omitting steps in the avidin-biotin-peroxidase procedure.

Three sections were stained with each antiserum and graded semiquantitatively, as previously described²². The light-microscopical sections were examined for the localization of endothelin-1-like immunoreactivity, particularly in vascular endothelium and airway epithelium. Staining intensity was graded semiquantitatively from 0 to 4, with 0 representing the absence of any staining and 4 the maximal intensity, corresponding to the intensity of staining with factor VIII. The grading was performed by two of the investigators, without prior knowledge of diagnostic category or clinical information. Additional sections from each specimen were stained with hematoxylin and eosin for pathological grading according to the system of Heath and Edwards¹⁸.

In Situ Hybridization

Ten-micrometer cryostat sections from paraformaldehyde-fixed tissues were rehydrated in phosphate-buffered saline, rendered permeable with proteinase K, and immersed in 4 percent paraformaldehyde for five minutes to stop the reaction. After three washes in phosphate-buffered saline, sections were immersed in a solution of triethanolamine (0.1 mol per liter) and acetic anhydride (0.25 mol per liter) for 10 minutes, dehydrated in ethanol, and air-dried. Sections were hybridized with a preproendothelin-1 complementary RNA (cRNA) probe labeled with sulfur-35 (1 × 10⁶ cpm per section) for 16 hours at 42 °C¹⁴. Unbound cRNA probe was removed by incubation in RNase solution (20 μg per milliliter) for 30 minutes at 42 °C in 2 × saline sodium citrate (SSC; 1 × SSC is 0.15 mol of sodium chloride and 0.015 mol of sodium citrate per liter, pH 7). This was followed by further washes in graded concentrations of SSC (from 2 × SSC to 0.1 × SSC) at temperatures ranging from 20 °C to 55 °C. Autoradiograms were exposed in light-tight boxes for 5 to 10 days at 4 °C, developed in D-19 developer (Kodak, Rochester, N.Y.), fixed, counterstained with hematoxylin, dehydrated, and mounted. Lung sections hybridized with a sense probe or sections treated with RNase solution before hybridization were used as negative controls for determining the specificity of the hybridization signals.

Simultaneous Localization of Endothelin-1 mRNA and Peptide

To examine whether endothelin-1-like immunoreactivity was localized to the cells that expressed the preproendothelin-1 mRNA, cryostat sections were rehydrated in phosphate-buffered saline, rendered permeable in proteinase K for seven minutes, immersed in 4 percent paraformaldehyde, and washed in phosphate-buffered saline. Subsequently, sections were hybridized with the radiolabeled preproendothelin-1 cRNA probe and washed as described above. After the last wash in 0.1 × SSC, sections were rinsed in phosphate-buffered saline and immunostained with the endothelin-1 antiserum as described above, then processed for autoradiography. The above method was a modification of that described by Hoefler et al²³.

Northern Blot Analysis

For Northern blot analysis, total RNA was extracted from lung tissue from the patients with plexogenic pulmonary arteriopathy (group 1) and the patients with secondary pulmonary hypertension (group 2), and from the unused normal lung tissue of the organ donors. The RNA extraction and hybridization method has been described elsewhere²⁴.

Statistical Analysis

The results are presented as means ±SE. The significance of differences between groups was assessed with a two-tailed Wilcoxon signed-rank test, and Bonferroni's correction was applied for multiple comparisons as appropriate²⁵. Pulmonary vascular resistance was calculated in Wood units (the mean pulmonary arterial pressure minus the pulmonary-capillary wedge pressure in millimeters of mercury divided by the cardiac output in liters per minute), and total pulmonary resistance was calculated as the mean pulmonary arterial pressure divided by the cardiac output. The correlation between total pulmonary resistance and the grade of endothelin-1-like immunoreactivity (immunostaining) was assessed with ordinary least-squares linear regression techniques.

Light-microscopical sections of tissue from the patients in group 1 revealed arteries with medial and intimal thickening (including onion-skinning), peripheral proliferation of muscle cells, and typical plexiform lesions with, in some instances, vasculitis¹⁸. Lung tissue from the patients in group 2, who had secondary pulmonary hypertension, showed changes typical of their respective underlying diseases and demonstrated variable degrees of medial thickening and intimal proliferation (i.e., grades 1 through 3).

Figure 1. Figure 1. Endothelin-1-Like Immunoreactivity in Normal Lung Tissue and Lung Tissue from Patients with Plexogenic Pulmonary Arteriopathy (Group 1) and Secondary Pulmonary Hypertension (Group 2).

Panel A shows immunoreactivity (brown staining with antibody to the C-terminal of endothelin-1) in the vascular endothelium of normal lung (arrow) (x600), and Panels B through D immunoreactivity in pulmonary arteries from patients in group 1 (Panel B, x600; Panel C and Panel D, x300). In addition, staining was observed in association with plexiform lesions (Panel E, x600), neuroendocrine cells and the vascular endothelium of bronchial vessels (Panel F, x600), smooth-muscle cells of atherosclerotic pulmonary artery (Panel G, x600), and type II alveolar pneumocytes in lung sections from patients in group 2 (Panel H, x600).

Immunocytochemical analysis revealed very little endothelin-1-like immunoreactivity in the control subjects (Figure 1A). In contrast, substantial staining was observed in sections from the patients with pulmonary hypertension (Figure 1B, Figure 1C, Figure 1D, Figure 1E, Figure 1F, Figure 1G, and Figure 1H). The greatest degree of endothelin-1-like immunoreactivity was seen in the endothelium of muscular pulmonary arteries, particularly over vessels exhibiting severe morphologic changes (Fig. Figure 1B, Figure 1C, and Figure 1D). In addition, vascular immunostaining was seen in association with plexiform lesions (Figure 1E). The vascular endothelium of bronchial vessels also stained intensely (Figure 1F). Less commonly, endothelin-1-like immunoreactivity was observed in endothelial cells of capillaries and pulmonary veins and in neuroendocrine cells (Figure 1F), and in inflammatory cells, pulmonary epithelium, and vascular smooth-muscle cells of atherosclerotic pulmonary arteries (Figure 1G). In contrast, there was little or no staining in the vasculature of renal and myocardial tissues from the patients with pulmonary hypertension (data not shown).

Figure 2. Figure 2. Mean (±SE) Level of Endothelin-1-Like Immunoreactivity in Vascular Endothelium of Lung Tissue from Two Groups of Controls (Panel A) and Two Groups of Patients with Pulmonary Hypertension (Panel B).

The patients with pulmonary hypertension were further subdivided according to morphologic¹⁸ and clinical¹⁹ criteria into those with plexogenic pulmonary arteriopathy (group 1) and those with secondary pulmonary hypertension (group 2). Endothelin-1-like immunoreactivity was assessed in the endothelium of elastic and muscular pulmonary arteries, capillaries, pulmonary veins, and bronchial vessels. The level of endothelin-1-like immunoreactivity was significantly greater in the patients with pulmonary hypertension than in the controls in all vessels. The asterisk indicates P = 0.003 for the comparison between groups.

In Figure 2, the degree of endothelin-1-like immunoreactivity in endothelial cells from each vascular region is compared in the patient groups. The grade of immunostaining was low in the controls, with the greatest immunoreactivity observed in elastic and muscular pulmonary arteries (0.4 ±0.2 and 0.2 ±0.1, respectively). This was in sharp contrast to the endothelin-1 immunostaining observed in elastic and muscular pulmonary arteries of the patients in both groups with pulmonary hypertension (2.73 ±0.22 and 2.74 ±0.15, respectively; P = 0.003); less pronounced increases in immunoreactivity were seen in the other vascular regions.

The distribution of vascular endothelin-1-like immunoreactivity was similar in the patients in groups 1 and 2 (Figure 2B). However, immunostaining was significantly greater in the muscular pulmonary arteries of the patients with plexogenic pulmonary arteriopathy (group 1) (mean grade, 3.3 ±0.1 vs. 2.4 ±0.2; P = 0.016). Furthermore, linear regression analysis indicated a significant correlation between endothelin-1-like immunoreactivity in muscular pulmonary arteries and total pulmonary resistance (a functional measurement of the severity of the disease) only in the patients with plexogenic pulmonary arteriopathy (P = 0.008). All four patients with interstitial fibrosis and pulmonary hypertension had substantial immunostaining for endothelin-1 in alveolar epithelial cells (mean grade for type II pneumocytes, 2.5 ±0.5) (Figure 1H). This immunostaining was not seen in the controls, and only rare pneumocytes stained in the patients with pulmonary hypertension without pulmonary fibrosis.

Figure 3. Figure 3. Expression of Endothelin-1 mRNA in the Vascular Endothelium of Pulmonary Arteries from Patients with Plexogenic Pulmonary Arteriopathy (Group 1) and Secondary Pulmonary Hypertension (Group 2).

Panel A (dark-field microscopy) and Panel C show expression in the vascular endothelium of pulmonary arteries from patients in group 2, and Panel B expression in a small occluded pulmonary artery from a patient in group 1 (arrow indicates the site of endothelin-1 expression). Panel D shows a section adjacent to that shown in Panel C, hybridized with the endothelin-1 sense probe (negative control). Panel E and Panel F show the expression of endothelin-1-like immunoreactivity and mRNA in the vascular endothelium of muscular pulmonary arteries from patients in group 1. (Panels A through F, x600.) Panel G shows a negative control section adjacent to that shown in Figure 1C, immunostained with normal goat serum instead of the endothelin-1 antiserum. Panel H shows the low expression of endothelin-1 mRNA in normal lung (phase-contrast microscopy). (Panel G and Panel H, x300.).

The majority of lung sections from the patients with pulmonary hypertension stained best with antiserum to endothelin-1 or stained equally well with endothelin-1 and big endothelin-1 antiserums. Seven patients, however, had stronger endothelin staining for the precursor than for the mature peptide (three in group 1 and four in group 2). Experiments with the negative controls showed no staining with the respective antiserums (Figure 3G).

In situ hybridization revealed the expression of endothelin-1 mRNA in the patients with pulmonary hypertension, primarily in the vascular endothelium, whereas there were only scattered hybridization signals on autoradiographs from the control subjects (Figure 3H). Clusters of silver grains were found overlying vascular endothelial cells of pulmonary arteries with medial thickening and intimal fibrosis in lung sections from the patients with pulmonary hypertension (Figure 3A, Figure 3B, and Figure 3C). The overall distribution of endothelin-1 mRNA relative to that of endothelin-1 immunostaining was demonstrated on sections that were both hybridized with the endothelin-1 cRNA probe and immunostained with endothelin-1 antiserums (Figure 3E and Figure 3F). Silver grains indicating the expression of endothelin-1 mRNA and the brown color of endothelin-1 immunostaining were seen mostly in the same cells, but on occasion endothelin-1 mRNA and the mature peptide were also expressed independently. There were no specific signals detectable on the negative-control sections hybridized with the sense probe or treated with RNase before hybridization with the endothelin-1 cRNA probe (Figure 3D). In addition, Northern blot analysis showed higher levels of preproendothelin-1 mRNA in lung tissue from the patients with plexogenic pulmonary arteriopathy and the patients with secondary pulmonary hypertension than in normal lung tissue.

The present study provides direct evidence of increased local production of endothelin-1 in pulmonary hypertension and may explain earlier observations of increased circulating endothelin-1 levels in patients with this disorder¹⁶. Expression of endothelin-1 was found in the vessels that were most affected by the morphologic abnormalities of pulmonary hypertension. In addition, expression of endothelin-1 in alveolar epithelial cells was seen nearly exclusively in patients with pulmonary fibrosis, raising the possibility that increased production of endothelin-1 may be involved in the pathogenesis of a broad range of pulmonary diseases associated with cellular proliferation. Although the expression of endothelin-1 in the lungs of patients with pulmonary hypertension does not prove a cause-and-effect relation, such a relation is biologically plausible.

Although endothelin-1 immunostaining can only be graded semiquantitatively, the differences between the control group and the patients with pulmonary hypertension were striking. As in a previous report,¹⁴ there was little expression of endothelin-1 in the normal adult lung or in the pulmonary vasculature of patients with pulmonary disease but without pulmonary hypertension. This was in sharp contrast to the prominent endothelin-1-like immunoreactivity in specimens from the patients with pulmonary hypertension. Furthermore, the expression of preproendothelin-1 mRNA in the lungs of the patients with pulmonary hypertension, as demonstrated by in situ hybridization, matched closely the distribution of endothelin-1-like immunoreactivity, confirming that there is local production of endothelin-1 by the pulmonary vascular endothelium in this disorder.

The most intense endothelin-1 immunostaining was found in the vascular endothelium of the patients with plexogenic pulmonary arteriopathy (grades 4 through 6),¹⁸ many of whom had clinical findings consistent with a diagnosis of primary pulmonary hypertension according to the criteria of the National Heart, Lung, and Blood Institute¹⁹. This is in agreement with our earlier findings, based on the measurement of endothelin-1 plasma levels, which demonstrated the release of endothelin-1 across the pulmonary bed in patients with primary pulmonary hypertension¹⁶. Thus, the increased pulmonary vascular production and release of endothelin-1 in these patients point to a role for this vasoactive peptide in the functional and morphologic vascular abnormalities characteristic of pulmonary arteriopathy. Although there were no significant differences between groups 1 and 2 in pulmonary arterial pressure or pulmonary vascular resistance, only in group 1 was there a significant correlation between the increases in pulmonary resistance and the degree of endothelin-1-like immunoreactivity in pulmonary vessels. This finding suggests that the patients with primary pulmonary hypertension had not only a greater degree of endothelin-1 expression, but also a more specific association between increased local endothelin-1 production and the severity of the disease.

The distribution of endothelin-1-like immunoreactivity in the lung may provide further clues concerning its functional importance in patients with pulmonary hypertension. Regardless of patient group, the greatest degree of immunostaining occurred in the endothelium of elastic and muscular pulmonary arteries, which also demonstrated severe medial thickening and intimal proliferation, sometimes severe enough to produce luminal occlusion. To a lesser extent, staining was seen in pulmonary capillaries and veins and bronchial arteries, but no staining was found in systemic vessels in the myocardial and renal tissues of eight patients with pulmonary hypertension. Endothelin-1-like immunoreactivity was also seen in association with plexogenic lesions in the patients in group 1. This specific association of endothelin-1 expression with areas of abnormal pulmonary vascular architecture again supports the view that endothelin-1 contributes to the initiation or progression of these lesions. In a recent study, Stelzner et al.²⁶ reported an increase in endothelin-1 production in the lung in rats with idiopathic pulmonary hypertension. Endothelin-1 is not only the most potent vasoconstrictor yet identified,⁷ but it has also been shown to stimulate the proliferation of smooth-muscle cells,^8-10 and in vivo it may act in concert with other growth factors. Thus, from the standpoint of biologic activity and tissue distribution, endothelin-1 is a likely mediator of the functional and morphologic vascular abnormalities of pulmonary hypertension.

In addition to the vascular endothelium, limited expression of endothelin-1 was also found in neuroendocrine, inflammatory, and smooth-muscle cells. This is consistent with a previous report demonstrating the localization of endothelin-1-like immunoreactivity in pulmonary endocrine cells of developing and adult lungs¹⁴. Macrophages, mast cells, and polymorphonuclear leukocytes have also been shown to produce endothelin-1,^27-29 the biosynthesis of which may be mediated by polymorphonuclear leukocytes²⁹. Furthermore, the production of endothelin-1 in vascular smooth-muscle cells has been demonstrated in vitro^30,31 and in vivo, particularly in atherosclerotic arteries³². However, the appearance of endothelin-1-like immunoreactivity in the alveolar epithelial cells of patients with pulmonary fibrosis was an unanticipated and intriguing finding. This was an extremely rare occurrence in patients with pulmonary hypertension without pulmonary fibrosis and was not seen in the control group. A role for endothelin-1 has been suggested in progressive systemic sclerosis,^33,34 a disorder characterized by the abnormal production of connective tissue and fibrosis³⁵. Indeed, it has been demonstrated that endothelin-1 can induce the proliferation of fibroblasts and increase the production of fibrous tissue in vitro³³. The demonstration that alveolar epithelial cells express endothelin-1 in areas close to foci of fibrous replacement of alveolar structures suggests a role for this peptide in the pathogenesis of pulmonary fibrosis.

Our finding of striking increases in endothelin-1 mRNA and in the production of the mature peptide by pulmonary vascular endothelium provides compelling new evidence in support of the view that endothelin-1 contributes to the vascular abnormalities of pulmonary hypertension. The expression of this mitogenic peptide by alveolar epithelial cells in patients with pulmonary fibrosis raises the possibility that endothelin-1 has a broader role in pulmonary diseases associated with cellular proliferation or fibrosis. Although the precise definition of the pathophysiologic importance of endothelin-1 in these conditions awaits the availability of specific endothelin-1 antagonists or inhibitors for clinical use, the present study provides important new information about alterations in local pulmonary endothelin-1 synthesis and tissue distribution in human lung disease, which is an essential prerequisite for unraveling its biologic importance.