Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Since the isolation of endothelin (ET)-1 from porcine endothelial cells by Yanagisawa and coworkers in 1988, the role of this potent vasoconstrictor peptide in the etiology of hypertension and hypertensive end-organ damage has been investigated with increasing interest. Subsequently, the molecular structures of the peptide family; the receptor subtypes; and the regulation, production, and release of ET and its various physiologic effects have been studied in detail (1). In addition to the putative role of ET in the cardiovascular system, the ET system seems to play an important role in the pathogenesis of lung disease. There is growing evidence that the ET system is involved in the pathogenesis of primary and also secondary pulmonary hypertension (2, 3). In vitro studies also supported a role for ET-1 as a growth promoting factor for bronchial smooth-muscle cells (4). There is also evidence that the lung ET system is involved in the development of pulmonary fibrosis (5) and in bronchiolitis obliterans associated with lung transplant rejection (6, 7). However, it is still unknown whether endogenous activation of the paracrine lung ET system is a cause or a consequence of lung injury. A powerful method to answer this question is an animal model with activated lung ET system.

We have recently generated human ET-1 transgenic (tg) mice (8). The transgene was expressed predominantly in the brain, lung, and kidney. Renal transgene expression was associated with a pathologic phenotype manifested by signs such as age-dependent development of interstitial fibrosis and glomerulosclerosis, leading to a progressive decrease in glomerular filtration rate in a blood pressure- independent manner (8). The aim of the present study was to analyze the expression of the transgene in the lung in more detail and to analyze whether or not the activated lung ET system in these mice causes lung pathology.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Materials

[¹²⁵I]-ET-1 (2,000 Ci/mmol) was obtained from Du Pont (Bad Homburg, Germany). Unlabeled ET-1 was from Peninsula Laboratories, Inc. (Germany). The selective ET receptor ligands (BQ 123 and BQ 3020) were from California Peptides, Inc. Unless otherwise stated, all reagents were of analytical grade and were purchased from Merck (Darmstadt, Germany), Boehringer-Mannheim (Mannheim, Germany), or Sigma (Munich, Germany).

Animals

Male heterozygous human ET-1 tg mice, line 856 (8), and their corresponding littermates were used. Animals were fed standard breeding rodent chow and water ad libitum. Genotype was confirmed by polymerase chain reaction (PCR) of genomic DNA using standard techniques (8). We analyzed the lungs of 3- and 12-mo-old ET-1 tg mice and their corresponding littermates. Mice were killed by decapitation and the lungs were carefully removed. The lungs were then immediately frozen in liquid nitrogen and stored at 80°C, or fixed in a formalin buffer for paraffin-embedding.

Reverse Transcription/PCR

Expression analysis in tg mice was performed by reverse transcription (RT)-PCR. RNA was prepared using TRIzol reagent (GIBCO BRL). RNA samples were reverse transcribed (M-MLV Reverse Transcriptase; GIBCO BRL), and the resultant complementary DNA was amplified by a PCR with specific human ET-1 primers as recently described (8): sense, 5'-TCC AGA GAG CGT TAT GTG AC-3'; antisense, 5'-TTC TGC TGA GAG CAT TG-3'. After pretreatment (94°C for 5 min), a step program (58°C for 60 s, 72°C for 75 s, and 94°C for 60 s; 28 cycles) was performed, followed by a final extension reaction (72°C for 10 min).

Tissue ET-1

Tissue preparation. The probes (200 mg) were stored in liquid nitrogen until further analysis. The frozen tissue was powdered in the presence of liquid nitrogen. The powdered samples were suspended and subsequently homogenized using a motor-driven pestle homogenizer in 2 ml buffer solution at 4°C (0.14 mol/liter NaCl, 2.6 mmol/liter KCl, 8 mmol/liter Na₂HPO₄, 1.4 mmol/liter KH₂PO₄, and 1% Triton-X 100; pH 7.4). The homogenates were centrifuged at 4°C for 60 min at 100,000 × g. The supernatants were retained for ET-1 enzyme-linked immunosorbent assay (ELISA). The recovery rate for ET-1 after preparing tissue ET-1 as described earlier was always between 97.5 and 99%.

ELISA. The commercially available enzyme immunoassays for ET-1 suitable for direct measurement of ET-1 in plasma and after tissue preparation were obtained from BIOMEDIC GmbH (Vienna, Austria) and were performed as recently described (9).

Immunohistochemistry

Slides with 3-µm-thick paraffin sections were dewaxed in xylol (twice for 5 min), and rehydrated in 100% ethanol (twice for 5 min), 75% ethanol (5 min), and tap water (5 min). After washing in Tris-buffered saline (TBS) (twice for 3 min), paraffin sections were digested with 0.0025% trypsin in 0.1% calcium chloride, pH 7.8, for 15 min at 37°C, followed by washing in TBS buffer, pH 7.5 (5 min). The sections were denaturated with 4 mol/liter HCl for 15 min. Nonspecific binding sites were blocked with incubation buffer (bovine serum albumin [BSA], Triton X100, and phosphate-buffered saline [PBS]; three times for 5 min) at room temperature. Excess buffer was removed and the sections were circled with a waterproof pen (Dako, Glostrup, Denmark) and incubated with primary monoclonal antibody (anti-ET-1 antibodies, rat antimouse mature macrophage antibodies [10], anti-CD4 antibodies, anti-CD8 antibodies, anti-CD19 antibodies, and antineutrophil antibody [Ly-6G; PharMingen/Becton-Dickinson, Hamburg, Germany]) for 2 h at 37°C, then overnight at 4°C. Negative slides were incubated with incubation buffer alone. After washing in PBS the slides were incubated with biotinylated rabbit antirat immunoglobulin G (Jackson ImmunoResearch Labs, West Baltimore, MD), followed by alkaline-phosphatase-conjugated streptavidin complex (Jackson ImmunoResearch) and Fast Red Tablets (Boehringer-Mannheim) according to the user's manual. Double-stained lung sections were analyzed by light microscopy and by a computer-aided image analysis system. We measured the relation of red-stained cells (positive cells) to total cell number of the whole lung section using a computer-aided image analysis system (8).

Binding Assays for ETA and ETB Receptors

To analyze the expression of ET receptor subtypes (ETA and ETB) in the lung, binding assays were performed in the presence or absence of the subtype-specific ET receptor ligands BQ123 (3 µmol/liter) and/or BQ3020 (3 µmol/liter) as recently described (11). The assay buffer for binding studies contained 1 mg/ml bacitracin, 100 mmol/liter Tris-HCl, 5 mmol/liter MgCl₂, and 1 g/liter BSA, pH 7.4, in a total volume of 150 µl. The [¹²⁵I]-ET-1 tracer concentration was kept constant at 40,000 counts/min/tube, whereas the concentration of unlabeled ET-1 was increased from 0 to 25 nmol/liter (competition studies with "cold saturation"). Samples from crude plasma membranes were used at a protein concentration of 0.5 mg/ml. Binding studies were performed at room temperature for 120 min. Nonspecific binding was assessed in the presence of excess ET-1 (5 µmol/liter). After adding 1 ml of cold binding buffer, free and receptor-bound radioactivity was separated by centrifugation at 30,000 × g (4°C) for 15 min, and the pellets thus obtained were washed two additional times with 1 ml of cold binding buffer. [¹²⁵I] was counted in a Packard Gamma Counter (78% counting efficiency for [¹²⁵I]). Receptor density (B_max) and binding affinity (K_D) values were calculated by linear-regression analysis of Scatchard plots.

Histologic Evaluation

For pathohistologic evaluation, all samples were embedded in paraffin, cut in 3 µm sections, and submitted to hematoxylin-eosin (HE) staining, periodic acid-Schiff staining, and Sirius Red staining (8, 9). The media/lumen ratio of the intrapulmonary arteries and the bronchi was analyzed using a video microscope connected to a personal computer. After Sirius Red staining, the severity of pulmonary fibrosis was evaluated using an image analyzing system (Quantimed 500). We measured the relation of red-stained area (connective tissue) to total area of the whole lung section. The data thus obtained were analyzed using the NIH Image 1.61 program (8, 9).

Hydroxyproline Determination

Hydroxyproline (HYP), an amino acid common to all collagens, was quantified colorimetrically in duplicates from lung tissue as recently described (12). Briefly, tissue was homogenized in 4 ml 6 N HCl and hydrolyzed at 110°C for 16 h. The hydrolysate was filtered, 50-µl aliquots were evaporated under vacuum, and residual HCl was removed after addition of methanol. The sediment was redissolved in 1.2 ml of 50% isopropanol and incubated with 0.2 ml of 0.84% chloramine-T in 42 mmol/liter sodium acetate, 2.6 mmol/liter citric acid, and 39.5% (vol/vol) isopropanol, pH 6.0, followed by incubation for 10 min at room temperature. Next, 0.248 g of p-dimethylaminobenzaldehyde, dissolved in 0.27 ml of 60% perchloric acid, and 0.73 ml isopropanol were added and incubated at 50°C for 90 min. HYP was then quantitated photometrically at 558 am from a standard curve with the amino acid alone and against a reagent blank. For calculations, the mean of two aliquots from the right lung was used.

Detection of Apoptotic Cells

A standard in situ terminal deoxyribonucleotidyl transferase (TdT)-mediated deoxyuridine triphosphate (dUTP) nick-end labeling (TUNEL) method was used to identify apoptotic cells within lung tissue sections. With some modifications TUNEL was performed according to the method of Gavrieli and colleagues (13). Tissue samples were fixed in Bouin's fixans (2.5% copper acetate and 4% picric acid in 3.5% formaldehyde in distilled water) and embedded in paraffin. For the positive control, TUNEL was performed after deoxyribonuclease treatment. On the other hand, for the negative control, TUNEL was done without addition of TdT. Briefly, 3-µm-thick paraffin sections were dewaxed, incubated with 30 µg/ml proteinase K for 30 min at room temperature, and washed. Endogenous peroxidase was then inactivated by covering the sections with 2% H₂O₂ for 5 min at room temperature, washing and treating with 0.1% Triton X100 in 0.1% sodium citrate for 3 min at room temperature, washing again, and then labeling with TdT (1:200) and Biotin-dUTP labeling mixture (1:100) in the TdT-reaction buffer for 60 min at 37°C. The reaction was terminated by immersing the slides in Tris buffer at room temperature for 15 min. After washing in PBS for 5 min at room temperature, sections were blocked with 2% BSA for 10 min at room temperature and then incubated with avidin-biotinilated horseradish peroxidase for 10 min at 37°C, washed with PBS, and developed with 3'-amino-9-ethyl-carbazole solution to produce a brown product. We counted TUNEL-positive cells and the total cell number in a given anatomic structure (arteries, bronchi, interstitial tissue) to calculate an apoptotic index (TUNEL-positive cells/total cell number).

In Vivo Bromodeoxyuridine Incorporation

We injected bromodeoxyuridine (BrdU) (200 mg/kg) into an animal to label the DNA in vivo, then after 22 h the animals were killed and paraffin-embedded tissue samples were prepared. Briefly, 3-µm-thick paraffin sections were dewaxed and enzymatic digestion was done with 0.1% trypsin and 0.1% CaCl₂ in TBS buffer (0.1 M Tris, 0.1 M NaCl, and 0.05 M MgCl₂), and incubated for 15 min at 37°C to obtain best results. After washing with Aqua dest. and TBS buffer the sections were denaturated with 4 M HCl for 15 min at room temperature. After denaturation, we incubated the specimen with incubation buffer (PBS containing 0.5% BSA and 0.1% Tween 20) for 15 min at room temperature to simultaneously neutralize the pH and block unspecific binding. Then probes were incubated with 50 µl of anti- BrdU-AP antibody solution (1:5 diluted with buffer) and incubated 35 min at 37°C in a humid chamber. We washed the slides three times in PBS for 2 min, then substrate reaction was started with a sufficient amount of the freshly prepared substrate solution and incubated at room temperature for 20 min until a clearly visible color developed depending on the amount of BrdU incorporated into the DNA. The slides were washed with PBS at room temperature for 5 min followed by Aqua dest. and embedded with glycerine/PBS. We scored double-stained lung sections by AP-positive cells with a light microscope (×400 magnification) (14).

Right Ventricular Pressure Measurements

Right ventricular pressure was measured as recently described (15). Mice were anesthetized with ketamine/xyalazine (100 and 15 mg/kg) in a single hindquarter intramuscular injection. After an adequate level of sedation was achieved, mice were placed in a supine position, spontaneously breathing room air, and after calibration of the zero point of the pressure transducer to the mid-anteroposterior (AP) diameter of the chest, a 26-gauge needle was introduced percutaneously into the thorax via a subxyphoid approach. Right and left ventricular pressures were measured using a pressure transducer (Gulton-Statham, Costa Mesa, CA) recorded on a multichannel recorder (Grass Institute Co., Quinen, MA). The heart rate under these conditions was between 300 and 600 beats/min (bpm). If the heart rate fell below 300 bpm, it was assumed that the level of anesthesia or trauma was inhibiting cardiac function and those measurements were excluded from analysis.

Blood Gas Analysis

Arterial samples (0.4 ml) were obtained by percutaneous left ventricular sampling from lightly anesthetized mice (3-min inhalation of metofane) while spontaneously breathing room air (altitude 5,280 ft). Pa_O2, PCO₂, and pH were measured using a clinical blood gas analyzer (Radiometer, Copenhagen, Denmark) as recently described (15).

Analysis of Data

The unpaired Student's t test determination of statistical differences of group means was used. In addition, analysis of variance followed by t test was used if appropriate. Results were considered significantly different at a value of P < 0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Pulmonary Expression of the Transgene and ET Receptors

The expression of ET-1 in the lung was analyzed by RT-PCR, measurement of tissue ET-1 concentrations, and immunohistochemistry. RT-PCR revealed transgene expression in the lungs of ET-1 tg mice. No human ET-1 messenger RNA signal was detectable in nontransgenic littermates. Lung tissue ET-1 concentrations were significantly increased in 3-mo-old ET-1 tg mice (87.9 ± 8.7 pg/g in ET-1 tg mice and 59.3 ± 9.9 pg/g in non-tg littermates; P < 0.05), whereas there was only a tendency toward increased tissue ET-1 concentrations in 12-mo-old ET-1 tg mice (82.6 ± 12.9 pg/g in ET-1 tg mice and 71.8 ± 8.9 pg/g in non-tg littermates; not significant). Immunohistochemistry revealed that ET-1 expression was detectable within all parts of the lungs (blood vessels, alveoli, bronchial wall, and interstitial tissue) of 12-mo-old ET-1 tg mice. However, by far the strongest expression was detected within the bronchial wall in the lungs of ET-1 tg mice (Figure 1).

View larger version (108K): [in this window] [in a new window]   View larger version (39K): [in this window] [in a new window]   Figure 1.   Pulmonary expression of the transgene was analyzed by RT-PCR and immunohistochemistry. (A) RT-PCR products from lungs of ET-1 tg mice had the expected size (321 bp); Lane 1: 3-mo-old ET-tg mouse, lane 2: 3-mo-old control mouse, lane 3: 12-mo-old ET-tg mouse, lane 4: 12-mo-old control mouse, lane 5: 3-mo-old ET-tg mouse, lane 6: 3-mo-old control mouse, lane 7: 12-mo-old ET-tg mouse. (B) A typical lung section from a ET-1 tg mouse stained with an ET-1 antibody. (C) A typical lung section from a control mouse stained with an ET-1 antibody. (D) A typical lung section from a tg mouse stained without the primary antibody (negative control) demonstrated a very low nonspecific background staining.

Scatchard analysis revealed that the B_max and K_D of both ET receptors were similar in 3- and 12-mo-old ET-1 tg mice and their corresponding controls (Table 1).

Pulmonary Fibrosis and Bronchial and Vascular Remodeling in ET-1 Transgenic Mice

The media/lumen ratio of pulmonary arteries in 3- and 12-mo-old ET-1 tg mice was similar to those seen in the corresponding littermates. The right ventricular weight was also not increased in ET-1 tg mice (Table 2).

ET-1 tg mice were, on the other hand, characterized by a progressive accumulation of extracellular matrix proteins within the lung, as seen after computer-aided image analysis of Sirius Red-stained lung sections. Fibrotic tissue was predominantly located in the perivascular and peribronchial space (Figure 2) around medium- and large-sized pulmonary arteries in ET-1 tg mice. Analysis of lung HYP content likewise demonstrated that a primary pulmonary overexpression of ET-1 causes fibrosis (Figure 2). The bronchial media/lumen ratio was similar in ET-1 tg mice and their corresponding littermates (Figure 3). The morphology of the alveoli in ET-1 tg mice was normal.

View larger version (26K): [in this window] [in a new window]   View larger version (32K): [in this window] [in a new window]   View larger version (115K): [in this window] [in a new window]   Figure 2.   Age-dependent development of pulmonary fibrosis in ET-1 tg mice. (A) Bar graphs showing the ratio of fibrotic area to total lung area in 3- and 12-mo-old ET-1 tg mice in comparison with age-matched littermates. n = 10 in each group. Values are means ± standard error of the mean (SEM). ##P < 0.01 compared to age-matched littermates. (B) Bar graphs showing right lung hydroxyproline content in 3- and 12-mo-old ET-1 tg mice in comparison to age-matched littermates. n = 7 in each group. Values are means ± SEM. ##P < 0.01 compared with age-matched littermates. (C) Typical lung section of a 12-mo-old ET-1 tg mouse after Sirius Red staining showing enhanced pulmonary fibrosis. Fibrosis was predominantly located in the perivascular space in ET-1 tg mice. (D) Typical lung section from an age-matched non-transgenic control mouse. (E) Typical lung section of a 12-mo-old ET-1 tg mouse after trichrome staining showing likewise enhanced pulmonary fibrosis. (F) Typical lung section from an age-matched non-tg control mouse (trichrome staining).

View larger version (26K): [in this window] [in a new window]   Figure 3.   Bar graph showing the bronchial smooth muscle surface/lumen ratio in 3- and 12-mo-old ET-1 tg mice in comparison with age-matched littermates. n = 10 in each group. Values are means ± SEM.

Inflammatory Cell Infiltration in the Lungs of ET-1 tg Mice

Analysis of lung sections by light microscopy after HE staining and trichrome staining revealed that the lungs of ET-1 tg mice are characterized by an infiltration of mononuclear cells. The infiltrates were mainly located around medium-sized pulmonary arteries. These infiltrates were not seen in lungs of age-matched littermates that lived under the same conditions as the ET-1 tg mice. This makes infectious diseases very unlikely as a cause of these infiltrates. Infectious diseases were also excluded by periodical regularly performed microbiologic testing of blood samples in our tg facility. To characterize these infiltrates in more detail, we analyzed lung sections by immunohistochemistry using antibodies against neutrophils, macrophages, and CD4-, CD8-, and CD19-positive cells, followed by a computer-aided image analysis. These studies revealed that the density of CD4-positive cells in the whole lung was significantly increased in 3- and 12-mo-old ET-1 tg mice as compared with age-matched littermates (Figure 4) (2.1- and 2.3-fold, respectively; P < 0.05 in each case), whereas the density of neutrophils, macrophages, and CD8- and CD19-positive cells was not affected (data not shown).

View larger version (144K): [in this window] [in a new window]   Figure 4.   Pulmonary infiltration in 3-mo-old ET-1 tg mice and control mice was analyzed after trichrome staining (A and B) and after immunohistochemistry (C, D, and E). (A) Lung section showing infiltrates with mononuclear cells in a ET-1 tg mouse. The infiltrates were mainly located around medium-sized pulmonary arteries. (B) These infiltrates were not seen in age- and sex-matched controls. (C) A typical lung section from a ET-1 tg mouse stained with a CD4 antibody is shown. The density of CD4-positive cells was markedly enhanced compared with age- and sex-matched littermates. (D) A typical lung section from a control mouse stained with a CD4 antibody. (E) Staining without the primary CD4 antibody revealed a very low background signal. (Original magnification of trichrome-stained sections: ×200; original magnification of section stained by immunohistochemistry: ×400.)

ET-1-Dependent Regulation of Growth and Programmed Cell Death in the Lungs of ET-1 tg Mice

Cell growth measured by in vivo BrdU incorporation and programmed cell death (apoptosis) measured by the TUNEL method were analyzed in different compartments of the lung (interstitial tissue, pulmonary arteries, and bronchi). The major finding was a highly increased cell proliferation rate in bronchial cells (including epithelial cells, smooth-muscle cells, and peribronchial cells such as fibroblasts) in 3- and 12-mo-old ET-1 tg mice compared with the age-matched littermates (2.4-fold in 3-mo-old ET-1 tg mice and 3.2-fold in 12-mo-old ET-1 tg mice, respectively; Table 3). Cell proliferation of interstitial cells was only slightly increased in ET-1 tg mice. Pulmonary arteries did not show an increased BrdU incorporation either in 3-mo-old or in 12-mo-old ET-1 tg mice (Table 3).

Apoptotic cells were, in general, rarely seen in the lungs of 3- and 12-mo-old ET-1 tg mice and their corresponding littermates. The number of apoptotic cells within the bronchi, pulmonary interstitial tissue, and pulmonary arteries was not different in tg and non-tg mice (Table 3).

Right Ventricular Pressure and Blood Gases

Right ventricular pressure was studied in ET-1 tg mice and their age-matched littermates exposed to inspired O₂ equivalent to that of Denver, CO (altitude 5,280 ft, 21% O₂). There were no differences in right ventricular systolic pressure between ET-1 tg mice and their corresponding littermates at any age (Figure 5). Blood gases (PO₂ and PCO₂) and arterial pH were also similar in ET-1 tg mice and their corresponding littermates (Table 4).

View larger version (33K): [in this window] [in a new window]   Figure 5.   Right ventricular systolic pressure in littermates and human ET-1 tg mice. The mice were raised at an altitude of 5,280 ft breathing room air. n = 4-6 in each group. Values are means ± SEM.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The paracrine lung ET system is thought to be involved in the pathogenesis of several lung diseases such as primary and secondary pulmonary hypertension (2, 3), pulmonary fibrosis (5), asthma and chronic obstructive lung disease (16), and bronchiolitis obliterans (6, 7). However, it is still unknown whether endogenous activation of the paracrine ET system is a cause or a consequence of lung injury, and in particular it is also unknown which parts of lung pathology (airway remodeling, fibrosis, vascular remodeling, and/or inflammatory reaction) are mainly mediated via an activated ET system. Our study demonstrated that a chronically activated lung ET system caused increased pulmonary matrix synthesis and a chronic inflammatory lung disease characterized by an increased infiltration of CD4-positive cells. Pulmonary hypertension, on the other hand, was not seen in ET-1 tg mice.

Pulmonary Fibrosis

An activated paracrine ET system has been implicated in the pathogenesis of diseases characterized by a progressively fibrotic remodeling, such as liver fibrosis/cirrhosis (17), arteriosclerosis (18), and kidney fibrosis (1, 8). There are also reports demonstrating an activated paracrine lung ET system in animal models of pulmonary fibrosis (5, 19) as well as in humans with pulmonary fibrosis (20). However, the definitive proof of the concept that long-term activation of the pulmonary ET system causes pulmonary fibrosis is missing so far because studies using ET receptor antagonists in animal models of pulmonary fibrosis had conflicting results (19, 24). Our study, on the other hand, indicates for the first time that primary chronic overexpression of ET-1 within the lung causes lung fibrosis. The localization of fibrosis primarily in bronchovascular structures differs from the pathology seen in humans with idiopathic pulmonary fibrosis, where the alveolar septae are a principal site of fibrosis. However, our tg mice were generated using a promotor that localized expression primarily to vascular structures and large airways. It remains uncertain whether overexpression of ET-1 in the alveoli would result in interstitial fibrosis. Further studies are necessary to analyze in more detail which matrix proteins are involved in ET-1-induced fibrotic remodeling of the lung in ET-1 tg mice.

Pulmonary Hypertension

This study indicates that an activated lung ET system itself without any other stimuli does not cause pulmonary hypertension. Indirect parameters such as vascular morphology of intrapulmonary arteries and the weight of the right ventricle also support the finding of normal pulmonary arterial blood pressure in ET-1 tg mice. This finding was unexpected because there are several studies in humans as well as in animal models of pulmonary hypertension, suggesting an important role of the pulmonary ET system in the pathogenesis of pulmonary hypertension (25, 33). Several possibilities could explain this seeming contradiction. First, the level of ET-1 overexpression in the lungs was modest, and may not have been sufficient to produce a pulmonary vascular phenotype. Second, counteracting mechanisms belonging to the ET system itself, such as increased ETB/ ETA receptor density, might explain this finding. We thus analyzed the expression of ET receptors within the lung and demonstrated that the density and affinity of both ET receptors were not altered in ET-1 tg mice. This suggests that altered receptor density is not the explanation, although we could not exclude differences in receptor localization or in postreceptor signal transduction pathways.

A third possibility is induction of other counter-regulatory pathways. There are indications of secondary activation of the nitric oxide (NO) system in ET-1 tg mice. This activated NO system contributes to the maintenance of normal systemic arterial blood pressure in ET-1 tg mice (28). Similar findings, a secondarily activated NO system resulting in normal systemic blood pressure, were seen in ET-2 tg rats (29), also indicating that a chronically activated ET system causes a compensatory activation of the NO system. We therefore suggest that NO also might attenuate the vasoconstrictor effects of ET-1 in the pulmonary circulation as well. An activated pulmonary ET system most probably requires additional factors/conditions for the manifestation of pulmonary hypertension. We suggest that the balance between the activation of the paracrine vascular lung ET system and the lung NO system, rather than the absolute activity of the vascular lung ET system, is important for the development of pulmonary hypertension. Involvement of the lung NO system in ET-1 tg mice will require further study.

Airway Remodeling

Measurement of the media/lumen ratio of the bronchi in ET-1 tg mice does not confirm in vitro observations that ET-1 promotes mitogenesis in airway smooth-muscle cells (4, 30, 31). Our observation, however, agrees with a recent study demonstrating that ET-1 has minor mitogenic effects on its own on vascular smooth-muscle cells in vitro but dramatically potentates the mitogenic effect of platelet-derived growth factor (PDGF)-BB (32). We therefore suggest that these discrepancies between our in vivo data and the cell culture studies mentioned earlier might be at least partially explained by the cell culture conditions (presence of growth factors such as PDGF-BB, etc.) in these studies (4, 30, 31). In addition, it is also possible that the relatively moderate overexpression of ET-1 in the lungs in our tg mice might be too low to induce bronchial smooth-muscle cell growth. In any case, our data suggest that pulmonary fibrosis and recruitment of inflammatory cells are the main in vivo long-term effects of an activated lung ET system. ET-induced growth of pulmonary smooth-muscle cells seems to be, at least in mice, of minor relevance in vivo.

Recruitment of Inflammatory Cells

Several proinflammatory mediators (interleukin [IL]-1, IL-1, and tumor necrosis factor-; see References 1 and 33) are known to stimulate the ET system. It has also been postulated that ET-1 itself causes recruitment of inflammatory cells (34). The present study revealed for the first time that an activated pulmonary ET system causes, without any further stimuli, an infiltration of mononuclear cells primarily around medium-sized pulmonary arteries. Immunohistochemistry revealed that these infiltrates consisted mainly of CD4-positive cells. The molecular mechanisms leading to a recruitment of CD4-positive lymphocytes around medium-sized pulmonary arteries are unknown so far. Interestingly, it has been shown that ET-1 induces the expression of E-selectin and intercellular adhesion molecule-1, facilitating migration of inflammatory cells (34). We suggest that the ET-1-induced recruitment of inflammatory cells might also contribute to lung pathology.

To sum up, the present study revealed that overexpression of the entire human ET-1 gene within the lungs of tg mice on its own does not cause pulmonary hypertension. In contrast, pulmonary ET-1 overexpression resulted in a development of pulmonary fibrosis and recruitment of inflammatory cells (predominantly CD4-positive cells). The in vivo ET-1 growth-promoting effects on smooth-muscle cells are, in comparison with the ET-1-induced pulmonary profibrotic effects as well as the ET-1-induced recruitment of inflammatory cells, less important. The present study suggests that ET-1 may have an etiologic role in pulmonary diseases associated with fibrosis and inflammation, such as bronchiolitis, as well as primary or secondary pulmonary fibrosis.