Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Chronic hypoxia-induced pulmonary hypertension (HPH) develops as a consequence of functional and structural alterations in the pulmonary vasculature, including increased vasomotor tone, thickening of the media and adventitia of small muscularized arteries, and extension of muscle into previously nonmuscularized peripheral arteries (1). All these changes contribute to increased pulmonary vascular resistance, thereby impairing right ventricular (RV) ejection and causing RV hypertrophy. The vascular remodeling process has been shown to involve medial smooth muscle cell (SMC) hypertrophy and hyperplasia (2, 3), fibroblast proliferation (2, 4), and matrix protein synthesis (4). A unique feature of HPH is its reversibility after return to normoxia, with a rapid decrease in pulmonary artery pressure and gradual regression of RV hypertrophy and distal pulmonary vessel muscularization (7, 8).

Angiotensin II (Ang II), an important mediator of systemic vascular remodeling, may contribute to hypoxia- induced medial hypertrophy of pulmonary arteries. This hypothesis is based largely on data from chronically hypoxic rats showing that: (1) angiotensin-converting enzyme (ACE) expression is locally increased in the walls of newly muscularized small pulmonary arteries (9); and (2) treatment with ACE inhibitors reduces pulmonary arterial pressure (10), medial thickening of muscular pulmonary arteries (10, 13), and cardiac hypertrophy (10, 12, 14).

It is well documented that Ang II acts as an agonist of both positive and negative growth processes of many cell types, including vascular SMCs (17), endothelial cells (18), cardiac fibroblasts (19), and myocytes (20). Moreover, there is evidence that Ang II may influence both vasoconstrictor (reviewed in Reference 21) and vasodilator responses (22, 23). In this regard, two pharmacologically distinct subtypes of the Ang II receptor, designated AT1 and AT2, have been described based on their affinity for selective receptor antagonists and their sensitivity to reducing agents (reviewed in References 21 and 24). Whereas AT1 mediates the vasoconstrictor and growth promoting effects of Ang II (21), AT2 has been shown to mediate vasodilator effects (22, 23, 25, 26) and to exert antigrowthantiproliferative (17), antihypertrophic (20), or apoptotic (27) action, depending on the cell type. These observations suggest that the development and reversal of HPH and vascular remodeling may be associated with antithetical patterns of expression of vascular AT1 and AT2 receptors. Data on Ang II receptor expression in this setting are extremely meager (28). The purpose of this study was to directly compare the time courses of Ang II receptor accumulation and of distal smooth muscle extension. Moreover, because the hypoxia-induced rise and the reoxygenation-induced decrease in pulmonary artery pressure may depend on the modulated number and/or on the activity of each Ang II receptor subtype, we evaluated the impact of AT1 and AT2 receptors on pulmonary vascular tone during chronic exposures both to hypoxia and post-hypoxic normoxia.

To address these questions, the levels of AT1 and AT2 transcripts and receptors were quantitated in the lungs of rats subjected to chronic normobaric hypoxia for 1, 2, or 3 wk, and of rats subjected to 3 wk of hypoxia followed by a return to normoxia for 15 h or 1, 2, or 3 wk by in situ autoradiographic binding and reverse transcriptase/polymerase chain reaction (RT-PCR) assays, respectively. The same lungs were subjected to a histologic evaluation of the percentage of vessels undergoing muscularization. Finally, we took advantage of an isolated perfused whole lung system from chronically hypoxic and normoxia-recovered rats to analyze the vasotonic action of AT1 and AT2 receptors.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Experimental Model

All animals were studied in accordance with procedures established by the Animal Care and Use Committee of our institution. Male Wistar rats (Charles River, Saint Aubun les Elbeuf, France) weighing 250-300 g at the beginning of the experiments were randomly divided into three sets. Two sets were exposed to chronic hypoxia: in one, rats were subjected to hypoxia for 1, 2, or 3 wk (hypoxic groups) and studied within 1 h of removal from the hypoxic chamber; in the other, rats subjected to 3 wk of hypoxia were returned to normoxia for 15 h or 1, 2, or 3 wk (normoxia-recovery groups). Finally, control rats were continuously exposed to room air for 1 to 6 wk. Hypoxia (10% O₂) was obtained in a ventilated chamber (500-liter; Flufrance, Cachan, France) as previously described (29) and monitored using an oxygen analyzer (model no. OA150; Servomex, Crowborough, UK). The chamber was opened once a day for 1 h to clean the cages and replenish food and water stores. Control and normoxia-recovery groups were kept in the same room, with the same light-dark cycle. Rat chow and tap water were provided ad libitum. At each experimental time point, the rats were anesthetized by intraperitoneal injection of sodium pentobarbital (20 mg/kg); the thorax was opened and the heart and lungs quickly removed en bloc. The heart was dissected free, and the right ventricle was carefully separated from the left ventricle and septum. The fresh ventricular tissues were blotted dry then weighed separately to determine the index of RV hypertrophy (Fulton index) based on the RV free wall weight-to-body weight ratio (RV/BW), left ventricular free wall plus septum weight-to- body weight ratio (LV+S/BW), and RV weight-to-left ventricular plus septum weight ratio (RV/LV+S). Because the Fulton index did not vary in normoxic animals within the 6 wk of experiments, only one control group was used. One of the lungs was immediately blotted dry, frozen in liquid nitrogen, and stored at 80°C for RNA preparations; the other lung was processed for histologic analyses or in situ receptor binding assay.

Light Microscopy Study of Pulmonary Arteries

Fresh lungs were fixed in the distended state by infusion of 4% aqueous buffered formalin into the trachea at 25 cm H₂O pressure. The entire piece was placed in a bath of the same fixative for 1 wk. A midsagittal slice of the right lung, including the apical, azygous, and diaphragmatic lobes, was processed for paraffin embedding. Sections (5 µm thick) were cut for light microscopy and stained with hematoxylin-phloxin-saffron and orcein-picro-indigo-carmine. For each rat, a total of 35 to 65 intraacinar vessels accompanying either an alveolar duct or alveolus was examined. Their type was identified as muscular or nonmuscular. Muscular arteries were defined as having a complete or partial layer of SMCs bound by two orcein-stained elastic lamina, and nonmuscular arteries were defined as free of SMCs. SMCs were identified as elongated cells stained red by phloxin and containing square-ended nuclei.

In Situ Autoradiographic Quantitative Receptor Binding Assay

Fresh lungs distended by infusion of Tissue-Tek (Miles, France), which was diluted by half in phosphate-buffered saline, into the trachea were rapidly frozen in isopentane at 30°C and stored at 80°C. Transverse sections (15 µm thick) from the peripheral part of the lung were cut at 20°C in a cryostat, thaw-mounted on gelatin-coated glass slides, and dried at 4°C for 1 h in a desiccator or overnight. Ang II binding sites were labeled with (3-[¹²⁵I]iodotyrosyl-4,Sar-1, Ile-8) Ang II (Amersham, Orsay, France; specific activity, 2,000 Ci/mmol) as previously described (19) with a few modifications. Briefly, consecutive sections were preincubated for 15 min at room temperature in 10 mM sodium phosphate buffer (pH 7.4) containing 150 mM NaCl, 1 mM ethylenediaminetetraacetic acid (EDTA), 0.2% proteinase-free bovine serum albumin (Sigma, Saint Quentin Fallavier, France), and 0.3 mM bacitracin (Sigma) to remove endogenous bound ligand. The sections were transferred and preincubated for an additional 30 min in the same mixture without the bacitracin, in the presence of the AT1-specific nonpeptidic antagonist losartan (0.5 × 10⁶ M; Merck Sharp and Dohme-Chibret, Paris, France), or the AT2-specific nonpeptidic antagonist PD123319 (0.5 × 10⁶ M; Parke Davis, Suresnes, France), or unlabelled Ang II (0.5 × 10⁶ M; Sigma). Finally, the sections were incubated for 1.5 h in the presence of 32 pM ([¹²⁵I]Sar-1, Ile-8)-Ang II and losartan, PD123319, or unlabelled Ang II (in the same concentrations as above) to identify AT2, AT1, and total Ang II receptor binding sites, respectively. After incubation, the sections were washed four times for 1 min in 50 mM fresh Tris-HCl buffer (pH 7.4, 4°C), once in water at 4°C for 30 s, then dried for a minimum of 2 h. They were exposed to imaging plates with a range of known amounts of ([¹²⁵I]Sar-1, Ile-8)-Ang II from the same medium as that used for lung section incubation, and were finally analyzed and quantified using the Bio-Imaging Analysis System (Fuji MacBAS 1000; Fuji Medical Systems, Clichy, France), a computing densitometer equipped with a helium-neon laser beam that is more sensitive and linear than conventional X-ray film autoradiography. The main advantage of the MacBAS system is that uptake is not saturable, which leads to more precise quantification of the radioactive signals in the form of units per pixels. The Quant analysis method supplied with the instrument was used to measure radioactivity within each lung section by enclosing the section in a freehand curve. Pixels accumulated within the entire section were normalized for surface area and converted to fmol/ mm² by direct density comparison with the ([¹²⁵I]Sar-1, Ile-8)-Ang II calibration curve. Therefore, the data are means of receptor binding capacity per unit of surface and are expressed in fmol/mm². The images can be displayed in levels of gray or in pseudocolors. Ang II-specific total binding corresponds to the total (AT) minus the nonspecific (NS) labeling revealed with cold Ang II. AT1-specific binding was measured in the presence of PD123319, and AT2-specific binding was measured in the presence of losartan. Furthermore, we checked that AT1+AT2 binding data were close to the values for specific Ang II total binding, as indicated by the high correlation curve r² = 0.92 (n = 39).

RNA Preparation and Quantitative RT-PCR Assay

RNA was extracted according to the procedure described by Chomczynski and Sacchi (30) from peripheral lung. Peripheral lung was obtained by sharp dissection of the outer 3 to 5 mm of lung tissue free from the remaining tissue. RNA concentration was determined using standard spectrophotometric techniques, and RNA quality was assessed by visual inspection of denaturing agarose gels stained with ethidium bromide. RNAs were then stored at 20°C as a suspension in H₂O/diethylpyrocarbonate (DEPC).

A quantitative RT-PCR assay was perfected to allow evaluation of Ang II receptor subtype gene expression. The assay involved simultaneous RT-PCR coamplification of AT1 or AT2 messenger RNA (mRNA) and of mRNAs encoding the housekeeping enzyme malate dehydrogenase (MDH). Selection of MDH mRNA as the internal quantitative standard was based on a previous study in which MDH mRNA pulmonary levels remained unchanged during hypoxia (31). For each sample, increasing concentrations of total RNA (125, 250, 500, and 750 ng) previously denatured by a 2-min exposure to 95°C were reverse-transcribed using 200 U of Moloney murine leukemia virus reverse transcriptase (Life Technologies, Cergy Pontoise, France), for 30 min at 42°C. The reaction was performed in 20 µl of an H₂O/DEPC mixture containing 1 µg of the oligo(dT) primers (Boehringer, Meylan, France), 20 U of RNase inhibitor (Promega, Charbonnières, France), 1.25 mM deoxynucleotide triphosphate (dNTP)-Na salts, 5 mM MgCl₂, and 2 µl of PCR buffer (×10) (Boehringer). The reaction was stopped by heating the reverse-transcribed products (complementary DNAs [cDNAs]) for 5 min at 99°C. Total cDNA pools (20 µl) were then coamplified by simultaneously using AT1 (or AT2) and MDH sense and antisense primers (Oligo Express, Paris, France). For AT1, the antisense primer was 5'-GCACAATCGCCATAATTATCC-3' (extending from base 719 to base 739 of the coding sequence) and the sense primer was 5'-GGAAACAGCTTGGTGGTG-3' (extending from base 132 to base 149 of the coding region), yielding a DNA fragment of 607 bp. Both AT1 primers were designed from the cDNA sequences common to AT1a (32) and AT1b (33) rat receptors. For AT2, the antisense primer was 5'-CCCATAGCTATTGGTCTTCAGCAGATG-3' and the sense primer was 5'-GCATGAGTGTTGATAGGTACCAATCGG-3', respectively extending from base 709 to base 735 and from base 410 to base 436 of the rat coding region (34, 35), and yielding a DNA fragment of 325 bp. Finally, the MDH antisense and sense primers, which were derived from the nucleotidic sequence of rat pre-MDH (36), were 5'-TTTCAGCTCAGGGATGGCCTCG-3' and 5'-CAAGAAGCATGGCGTATACAACCC-3', respectively (37), and yielded a DNA fragment of 507 bp. The coamplification reaction was with 5 U of Taq DNA polymerase (Life Technologies) in 100 µl of an H₂O mixture containing 300 ng of each of four PCR primers in a final concentration of 0.2 mM dNTP-Na salts, 2 mM MgCl₂, and 10 µl of PCR buffer (×10). Thirty-one amplification cycles were performed in a Perkin Elmer apparatus (Perkin Elmer, Courtaboeuf, France) as follows: denaturation at 94°C for 1 min, annealing at 60°C for 2 min, and extension at 72°C for 2 min. A total of 45 µl of each final PCR was homogenized, removed in a reproducible manner, and run on 1% ethidium bromide-stained agarose gel at 80 V for 2.5 h. Signals of both separated PCR product bands (AT1 and MDH, or AT2 and MDH) were analyzed using densitometry with a computer-based imaging system (Gel Doc 1000; Biorad, Ivry sur Seine, France).

Pharmacologic Studies on Isolated Lungs

Rats from the control group plus the 2- and 3-wk hypoxia groups subjected or not to subsequent 3 wk of normoxia recovery were anesthetized with sodium pentobarbital (40 mg intraperitoneally). After tracheal cannulation, ventilation was started with warmed normoxic gas (95% air, 5% CO₂) at 60 breaths/min with an inspiratory pressure of 9 cm H₂O and an expiratory pressure of 2.5 cm H₂O. A midline sternotomy was performed, and 100 IU heparin was administered into the right ventricle. The heart and lungs were removed and suspended in a humidified chamber at 37°C. The lungs were perfused through a pulmonary artery cannula with a peristaltic pump, at a constant flow of 0.05 ml/g body weight/min. The recirculated perfusate was warm (38°C) physiologic saline (116 mM NaCl, 4.7 mM KCl, 19 mM NaHCO₃, 0.83 mM MgSO₄, 1.8 mM CaCl₂, 2 mM H₂O, 1.04 mM NaH₂PO₄, 5.5 mM glucose, 0.11 g/liter phenol red Na, and 4 g/100 ml Ficoll); meclofenamate (3.2 µM) was added to the perfusate at the beginning of the experiments. Effluent perfusate was drained from the left ventricular cannula into a reservoir. Mean perfusion pressure was continuously recorded from a side port of the pulmonary artery (P23 XL transducer; Gould, Ballainvilliers, France). Pulmonary venous pressure was assumed to be zero. After a constant 30-min equilibration period, each lung preparation was used for only one of the studies described below.

Vasotonic response to Ang II. Ang II was injected into the arterial line as 50-µl boluses containing increasing doses of Ang II (from 5 × 10¹⁰ to 5 × 10⁸ M) at 10-min intervals. Pressor responses were examined in the presence or absence of losartan (10⁵ M). Losartan was added to the perfusate reservoir 10 min before determination of pressor-response curves to Ang II.

Vascular reactivity to the AT2-specific peptidic agonist CGP42112A. Vasotonic effects of CGP42112A (Ciba Geigy, Rueil Malmaison, France) were investigated by directly injecting the compound into the arterial line as 50-µl boluses containing increasing doses of CGP42112A (5 × 10¹¹ to 5 × 10⁹ M) at 10-min intervals. In another series of experiments, the vasodilator response to CGP42112A was examined on lungs preconstricted by infusion of the endoperoxide analog U-46619. U-46619, first diluted in 20 ml of physiologic saline, was infused into the pulmonary arterial line at a constant rate of 50 pmol/min. Pulmonary artery pressure increased gradually in response to U-46619, without reaching a plateau. CGP42112A was then injected into the arterial line as 50-µl boluses containing increasing doses (from 5 × 10¹¹ to 5 × 10⁹ M) at 3-min intervals.

Statistical Analysis

Results for the various groups of rats are expressed as means ± standard error of the mean (SEM). For Fulton indexes, muscularization, binding, and RT-PCR assays, differences across groups were evaluated using one-way analysis of variance (ANOVA), and group-to-group comparisons were done using Scheffé's test. To compare effects on pressure changes induced by Ang II versus its vehicle in isolated rat lungs, two-way ANOVA with repeated measurements was performed: we tested for drug effect, dose effect, and interaction. Because interactions were significant, the nonparametric Kruskal-Wallis test was used to compare groups at each Ang II dose level. When a significant difference was observed, comparisons were performed using the Mann-Whitney test. For whole results, the accepted level of significance was P < 0.05, P < 0.01, or P < 0.001.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Assessment of Hypoxia-Induced RV Hypertrophy (Fulton Index)

The degree of RV hypertrophy was significantly increased by 1.4-fold (P < 0.05) and by 1.7-fold (P < 0.001) after 1 and 2 wk of hypoxia, respectively, as compared with control values, reaching a 1.9-fold increased value by 3 wk (P < 0.001) (Figure 1). The Fulton index was sharply decreased (1.4-fold, P < 0.001) after 1 wk of return to normoxia, as compared with the 3 wk of exposure, with small additional decreases during the second and third weeks.

View larger version (55K): [in this window] [in a new window]   Figure 1.   Graph showing the Fulton index variations in control, hypoxic, and normoxia-recovered rat groups within the time of experiments. Control group (C); week (W); right ventricular weight (RV); left ventricular weight plus septal weight (LV+S); arbitrary unit (AU). Data are expressed as mean ± SEM obtained from five to 11 animals. *P < 0.05; **P < 0.001 versus control group; § P < 0.001 versus 3-wk hypoxic group.

Histologic Analysis

Figure 2 shows light microscopy pictures of hematoxylin and orcein-stained lung sections obtained from control (A), 3-wk hypoxic (B), and 3-wk normoxia-recovered (C) rats. Pulmonary vessel thickening owing to muscularization was visible clearly in the 3-wk hypoxic lungs as compared with the control and 3-wk normoxia-recovered lungs. Determination of counts of muscular and nonmuscular vessels in each experimental group (n = 5 rats per group) after 2 and 3 wk of hypoxia revealed a 2.8-fold (P < 0.05) and 3-fold (P < 0.01) increase in the percentage of alveolar duct muscular vessels and a 3.7-fold (P < 0.05) and 4-fold (P < 0.01) increase in the percentage of alveolar wall muscular vessels, respectively, as compared with corresponding control values (Figures 2D and 2E). One week after return to normoxia, percentages of muscular vessels remained significantly higher (P < 0.05) in alveolar ducts and alveolar walls as compared with control values. However, the significant reversal (P < 0.01) in the percentage of muscular vessels observed in alveolar walls after 3 wk of normoxia recovery did not reach the control value.

View larger version (119K): [in this window] [in a new window]   Figure 2.   (A-C) Photomicrographs of lung sections from 3-wk control, 3-wk hypoxic, and 3-wk normoxia-recovered rats stained with orcein and phloxin. Muscular arteries have a complete layer of SMCs, seen in black. The small vessels are indicated by white arrowheads, whereas larger vessels are indicated by black arrows. b = bronchioles. For each rat, percentages of distal muscularization of normally nonmuscular arteries in alveolar ducts (D) and alveolar walls (E) were obtained by examining a total of 35 to 65 intraacinar vessels accompanying either an alveolar duct or an alveolus. Data are expressed as mean ± SEM of five separate experiments. *P < 0.05; **P < 0.01 versus control group; § P < 0.01 versus 3-wk hypoxic group.

Ligand Binding Assay

As shown in Figure 3A, [¹²⁵I]Ang II binding was present at a low level throughout the lung under basal conditions (column AT, row C), increased dramatically and uniformly after 2 wk of hypoxia (row H 2wk), decreased by hypoxia 3 wk (row H 3wk) and 15 h after return to normoxia (row N 15h), and returned to the basal level 3 wk after return to normoxia (row N 3wk). Pretreatment of lung sections with 0.5 × 10⁶ M Ang II prevented radioligand binding, revealing a very low level of nonspecific binding in each group (column NS). In each group, total [¹²⁵]Ang II binding capacity was only minimally affected by pretreatment with PD123319 to identify AT1 receptors (column AT1) but fell sharply after losartan pretreatment to identify AT2 receptors (column AT2), suggesting that Ang II binding total capacity was far more heavily dependent on AT1 receptors than on AT2 receptors (Figure 3A). Densitometric analysis of radioactive signals from each lung group revealed that the amount of AT2 receptors (Figure 3B) was approximately 10 times less than that of AT1 receptors (Figure 3C). A time-course analysis (Figure 3) indicated that both AT1-related and AT2-related signals reached maximal intensity after 2 wk of hypoxia. AT1 and AT2 receptor levels were significantly (P < 0.05) increased after 2 wk of hypoxia: the increases versus corresponding controls were 16-fold for AT1 (0.523 ± 0.289 [n = 3] versus 0.033 ± 0.012 fmol/mm² [n = 3]) and 18-fold for AT2 (0.0371 ± 0.0150 [n = 3] versus 0.002 ± 0.002 fmol/mm² [n = 3]). Both levels started to decrease during the third week of hypoxia and decreased further during the first 15 h of normoxia recovery, reaching control values after 1 wk (0.013 ± 0.008 fmol/mm² [n = 5] for AT1 and 0.003 ± 0.002 fmol/mm² [n = 5] for AT2, P < 0.05 versus 2 wk of hypoxia).

View larger version (83K): [in this window] [in a new window]   Figure 3.   In situ autoradiographic quantitative binding analysis of rat pulmonary Ang II receptor and Ang II receptor subtype (AT1 and AT2) densities during chronic hypoxia and normoxia recovery. (A) Transverse pulmonary sections (15 µm thick) from control (C), 2-wk hypoxic (H 2wk), and 3-wk hypoxic (H 3wk) rats and from 15-h (N 15h) and 3-wk normoxia-recovered (N 3wk) rats were incubated with 32 pM ([125I]Sar-1, Ile-8)-Ang II (column AT). Incubation with 0.5 µM unlabelled Ang II prevented radioligand binding (column NS). Identification of Ang II receptor subtypes AT1 and AT2 was determined in competition binding experiments based on sensitivity to subtype-selective concentrations of PD123319 or losartan. Specific [125I]Ang II binding sensitive to PD123319 or losartan was estimated as the density of AT1 receptor (column AT1) or AT2 receptor (column AT2), respectively. Specific AT2 (B) and AT1 (C) binding was calculated as the difference between total and nonspecific binding. Values are given as the mean ± SEM of three to five separate experiments. §P < 0.05 versus control and all normoxia-recovered groups; *P < 0.05 versus 2-wk normoxia-recovered group.

Ang II Receptor Subtype mRNA Accumulation

Figure 4A shows typical gels obtained from the analysis by the AT1-PCR and AT2-PCR assays of various amounts of RNA extracted either from a normoxic or a 2-wk hypoxic lung. To obtain quantitative results, we systematically analyzed RT-PCR products fitting into the linear part of each AT1, AT2, and MDH amplification curve (Figure 4B), showing that available signals were yielded from 250 and 500 ng initial RNA, whatever the experimental condition. In addition, we verified that MDH signal density for both RNA doses was unaffected by experimental conditions (Figure 4C), indicating that use of the MDH signal to normalize AT1 and AT2 mRNA signal densities was appropriate. For each independent RNA sample, definitive AT1/MDH and AT2/MDH values were the means of values systematically determined from both 250 and 500 ng RNA doses.

View larger version (40K): [in this window] [in a new window]   Figure 4.   (A) Four typical ethidium bromide-stained gels obtained from the analysis by the AT1-PCR and AT2-PCR assays of various amounts of RNA extracted either from a normoxic (left) or a 2-wk hypoxic (right) lung. (B) Representative graph showing the AT1, AT2, and MDH amplification curves obtained from the analysis of a 2-wk hypoxic lung RNA sample. (C) Graph comparing MDH mRNA signal modulation in the control, hypoxia, and normoxia recovery groups. Comparisons were performed depending on initial quantities (250 or 500 ng) of peripheral lung RNA. No significant modulation in MDH mRNA signal as compared with normoxic conditions was observed, whatever the initial quantity of RNA, under hypoxia or normoxia recovery, indicating that use of the MDH signal to normalize AT1 and AT2 mRNA signal densities was appropriate. RT-PCR determination of AT2 (D) and AT1 (E) mRNA concentrations in peripheral lung RNA. Data are expressed as mean ± SEM of four to six separate experiments and are expressed in arbitrary units (AU) based on the densitometric scores. *P < 0.05.

Densitometric signal analysis confirmed that in normoxic lungs AT1 mRNA was predominant, whereas AT2 mRNA was found only in insignificant amounts (Figures 4D and 4E), in agreement with previous results (38). A transient and significant (P < 0.05) increase in AT2 mRNA (Figure 4D) was detected after 2 wk of hypoxia (0.535 ± 0.028 [n = 4] versus 0.001 ± 0.001 AU [n = 5] in control lungs). Subsequently, AT2 mRNA concentrations decreased to values similar to control values. Conversely, AT1 mRNA concentrations remained stable during early hypoxia but showed a marked decrease (P < 0.05) after 3 wk of hypoxia (5.4-fold versus 2 wk, from 0.821 ± 0.021 [n = 4] to 0.152 ± 0.152 [n = 5]) and remained at low concentrations during normoxia recovery (Figure 4E).

Pharmacologic Analysis of Pulmonary Vascular Tone in Perfused Lungs Isolated from Control, Hypoxic, and Normoxia-Recovered Rats

To determine whether changes in respiratory oxygen level had a quantitative or qualitative influence on the pulmonary vasotonic effects of Ang II, we looked at AT1 and AT2 receptor effects on vascular tone in perfused lungs isolated from control rats, and from 2-wk and 3-wk hypoxic rats subjected or not to subsequent 3 wk of normoxia recovery. Because AT1 and AT2 receptors have been shown to mediate vasoconstrictor and vasodilatator actions of Ang II respectively, we studied, in each group of lungs, the effects of specific AT1 antagonist losartan and AT2 agonist CGP42112A on the pulmonary vasopressive response to Ang II and to endoperoxide analog U-46619, respectively.

Pulmonary vasoconstrictive response to Ang II. After the equilibration periods had elapsed, mean baseline pulmonary artery pressures were 7.6 ± 0.2 mm Hg in lungs from normoxic rats, 12.5 ± 0.7 and 13.2 ± 0.3 mm Hg in lungs from 2-wk and 3-wk hypoxic rats, respectively, and 11.2 ± 1.0 mm Hg in lungs from 3-wk hypoxic rats subjected to subsequent 3 wk of normoxia (n = 11 to 17 animals per group). These pressures were significantly higher in hypoxic (P < 0.01) and normoxia-recovered (P < 0.05) rat lungs than in control lungs. Repeated administration of incremental Ang II doses induced a concentration-dependent increase in pulmonary arterial pressure (Ppa) that was more marked in lungs from hypoxic and normoxia-recovered rats than in those from control rats, with no difference between 2-wk and 3-wk hypoxic groups (Figure 5). The pressor response to the highest Ang II dose was increased 1.7-fold and 1.5-fold in lungs from 2-wk and 3-wk hypoxic rats, respectively (P < 0.05), and 2-fold in lungs from normoxia-recovered rats (P < 0.01), as compared with lungs from normoxic rats. Interestingly, the pressor response to Ang II was significantly higher in lungs from normoxia- recovered rats (P < 0.05) than in those from 3-wk hypoxic rats. In the four groups, the vasopressor response to all Ang II dose levels was completely abolished in the presence of losartan, in keeping with the known vasoconstrictive action of AT1 receptors (21, 24). It is noteworthy that the vasopressor response to Ang II did not differ between normoxic and hypoxic animals when expressed as percent change from baseline tone.

View larger version (15K): [in this window] [in a new window]   Figure 5.   Pressure changes induced by Ang II in lungs from control, 2-wk, and 3-wk hypoxic, and 3-wk normoxia-recovered rats in the absence (solid symbols) or presence (open symbols) of 105 M of the selective AT1 receptor antagonist losartan. Values are expressed as mean ± SEM of six separate experiments.

Vascular reactivity to CGP42112A. There was no effect of CGP42112A (5 × 10¹¹ to 5 × 10⁹ M) on baseline Ppa in lungs from control normoxic rats, 2-wk and 3-wk hypoxic rats, or 3-wk hypoxic rats returned to normoxia. Addition of CGP42112A to the perfusate during vascular tone increase by U-46619 infusion also failed to alter perfusion pressure in the three groups of rats. CGP42112A failed to induce vasodilation whether or not the lungs were pretreated with meclofenamate. When Ang II was added during vascular tone increase by U-46619, no vasodilation was seen with any of the doses tested (10¹¹ to 10⁸ M), but there was an additional vasopressive response with doses higher than 10¹⁰ M. In contrast, bolus administration of sodium nitroprusside caused a dose-dependent decrease in Ppa in lungs from control normoxic rats, hypoxic rats, and hypoxic rats returned to normoxia (data not shown).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

We found that (1) AT1 and AT2 receptors accumulated transiently during chronic normobaric hypoxia in the rat lung, (2) this accumulation accompanied the initial stage of distal muscularization, and (3) neither receptor subtype accumulated during normoxia recovery, refuting the hypothesis of antithetical AT1 and AT2 expression during hypoxia and normoxia recovery. We also found that the vasoconstrictor response to Ang II, which is increased in isolated perfused lungs from hypoxic rats and from hypoxic rats returned to normoxia, was completely inhibited by the selective AT1 receptor antagonist losartan, and that the AT2 receptor agonist CGP42112A did not alter pulmonary vascular tone in any of the study conditions. These findings extend our knowledge of the hypoxia-related and normoxia recovery-related regulation of Ang II receptor expression.

Previous studies have examined renin-angiotensin system involvement during development of HPH. Increased ACE expression in the walls of small pulmonary arteries has been reported during chronic hypoxia (9), as well as a protective effect of ACE inhibitors against pulmonary hypertension and increased arterial muscularization in rats (10, 13, 15, 16). Moreover, infusion of the AT1 receptor antagonists losartan and GR 138950C reduced distal muscularization in rats subjected to chronic hypoxia for 7 or 14 d (14, 28). These findings point to an important role for AT1 receptor subtypes during HPH. Very few studies have examined the expression and role of AT2 receptors in this setting.

We found that densities of both AT1 and AT2 receptors were low in normoxic and 1-wk hypoxic lungs, significantly increased after 2 wk of hypoxia, and decreased beyond 2 wk of hypoxia. The only previous study that examined the subject of Ang II receptor expression found a small increase in AT1 receptor density in whole lung membranes from rats exposed to hypoxia for 7 d (28). Our results showing that the maximum was reached after 2 wk and increased more than 15-fold from baseline extend these previous findings. The downregulation of AT1 and AT2 expression demonstrated during the late stage of hypoxia was not reversed after pulmonary reoxygenation, and no accumulation of Ang II receptors was observed subsequently during normoxia recovery. This finding of AT1 and AT2 receptor accumulation exclusively during chronic hypoxia suggests that Ang II may participate through both Ang II receptor subtypes in the remodeling processes associated with the development of HPH, and that this peptide is probably not involved in mechanisms governing reversal during normoxia recovery.

The distal spread of smooth muscle into normally nonmuscular pulmonary arteries in alveolar ducts and walls, a feature common to almost all forms of pulmonary hypertension (39), is one of the earliest structural pulmonary vasculature changes observed during hypoxia and is closely correlated with a sustained elevation in Ppa (40). Within this context, a meaningful observation from our in situ binding studies was a relatively uniform increase in AT1 and AT2 radioactive staining throughout the lung sections at 2 to 3 wk of hypoxia, suggesting that both receptor subtypes are found throughout the pulmonary microvasculature undergoing muscularization and other marked structural modifications during HPH (39). This diffuse distribution of Ang II receptors is in agreement with previous results showing that after 2 wk of hypoxia, ACE expression was increased in alveolar duct and alveolar wall arteries (9) and that muscularization in these vessels was prevented by AT1-specific antagonists (14). To investigate the factual relation between Ang II receptor accumulation and distal muscularization, we compared the time courses of these two processes in the same lungs during hypoxia and normoxia recovery. Accumulation of both AT1 and AT2 occurred concomitantly with the rapid muscularization process characteristic of the first 2 wk of hypoxia. During the third week of hypoxia, however, distal muscularization continued to progress, whereas AT1 and AT2 binding capacities decreased slightly. These data suggest that Ang II may participate in the initiation and rapid extension of the remodeling processes responsible for production of "new" smooth muscle. This possibility is consistent with a report by Zakheim and colleagues (41) of a transient increase in circulating Ang II levels in the early stages of exposure to hypoxia.

Although muscularization of normally nonmuscular arteries in alveolar ducts and walls has been shown to depend largely on AT1-mediated signal transmission, the presence and increase of AT2 receptors in hypoxic lungs raise questions regarding the role of this subtype. Little is known about the physiologic function(s) of this receptor, which is abundantly expressed in fetal tissues, but in adults is present only in low concentrations and in a limited number of tissues (reviewed in References 21 and 42). In particular, AT2 transcripts, which are widely expressed in the developing cardiopulmonary system, are chiefly located in vascular structures (38). Nevertheless, the role of AT2 receptors in the pulmonary vasculature has never been examined. Various physiologic responses mediated by AT2 receptors have been reported, including growth inhibition of cultured coronary microvascular endothelial cells (18) and cardiac fibroblasts (19), apoptosis of pheochromocytoma PC12W and fibroblast R3T3 cell lines (27), and inhibition of collagen synthesis by cardiac fibroblasts in myocardial fibrosis (19) and in tissue repair in adults (21). AT2 receptors have been shown to antagonize the growth effects of AT1 receptors in both the adult and fetal aortic wall under normal physiologic conditions and after balloon injury (17, 43). Conversely, reports of AT2-mediated arterial hypertrophy in hypertensive rats (44, 45) strongly suggest that AT2 function depends on the cell mitotic index and/or on cell differentiation. These data are consistent with the possibility that AT2 receptors may protect from AT1-mediated muscularization by means of growth-antagonizing and/or apoptosis-enhancing effects, contributing to maintain vascular homeostasis during distal extension of smooth muscle. Support for this hypothesis has been provided by undocumented data from Morrell and coworkers (Table 3 in Reference 14) showing that infusion of the specific AT2 antagonist PD123319 resulted in a greater increase in the percentage of muscularized alveolar wall arteries in rats subjected to hypoxia.

Our main finding of increased AT1 and AT2 expression during the early phase (first 2 wk) of chronic hypoxia argues for a prominent regulatory role of hypoxia. The peak in AT1 and AT2 receptor levels seen after 2 wk of hypoxia may coincide with completion of the main processes ensuring adaptation to hypoxia. Exposure to hypoxia for 2 wk was associated with an increase in AT2 transcripts in the absence of any significant change in the level of AT1 mRNA. Concentrations of both AT1 and AT2 mRNA were decreased after 3 wk of hypoxia and remained low throughout normoxia recovery. That AT2 transcripts and AT2 receptors accumulated at the same time indicates that hypoxia-induced transcriptional mechanisms may activate the AT2 gene via a hypoxia-induced factor (HIF)-response element similar to that described for phosphoglycerate kinase, muscle-type lactate dehydrogenase, and the erythropoietin genes (reviewed in Reference 46). On the other hand, our data suggest that the accumulation of AT1 receptors after 2 wk of hypoxia may result from post-transcriptional regulatory mechanisms associated with hypoxia-induced activation of elongation factors, as previously described for elongation factor EF1 in hypoxic plant tissue (47). The decreases in AT1 and AT2 transcripts and in the corresponding receptor levels seen after 3 wk of hypoxia may be due chiefly to a decrease in the transcriptional activity of both genes reflecting complete adaptation to hypoxia. Alternatively, the rapid decrease in AT expression may result from massive degradation of transcripts. Interestingly, because the Ang II receptor densities were still high after 3 wk of hypoxia, we were able to show that the decreases started at the end of the hypoxic period and became more pronounced within the first 15 h after return to normoxia, suggesting that pulmonary reoxygenation may also alter receptor stability.

Finally, to investigate the potential role of AT1 and AT2 receptors on pulmonary vessel tone, we studied the vasomotor response to angiotensin II in isolated perfused lungs. Ang II-induced vasoconstriction was completely inhibited by losartan in both normoxic and hypoxic lungs, suggesting that it was entirely mediated by AT1 receptors. In contrast, the AT2 receptor-specific agonist CGP42112A had no effect on basal vascular tone and did not cause vasodilation in lungs preconstricted by U-46619. This lack of effect is at variance with reports of a vasodilating effect of AT2 manifesting as an abnormally high mean arterial pressure in rats coinfused with subpressor Ang II doses plus PD123319 (22, 23) and in AT2 knockout neonatal mice (25, 26). However, the possibility remains that AT2 receptors may mediate distal vasodilator effects normally masked by massive vasoconstriction owing to the 10-fold higher number of AT1 as compared to AT2 receptors in hypoxic lungs. An alternative explanation is that AT2- mediated vasodilation may be masked by exposure to hypoxia for 2 wk because it is normally mediated through an endothelium-dependent process known to be altered by hypoxia (29). As expected from receptor binding data, AT1-induced vasoconstriction was enhanced in lungs from hypoxic rats. In contrast, the persistent increase in the vasoconstrictor response to Ang II after return to normoxia is difficult to explain in regard to our results showing a decreased AT1 number by 3 wk. As previously shown for ACE, the pulmonary distribution of AT1 receptors may be differently altered during hypoxia and normoxia recovery. After return to normoxia, the remaining of AT1 receptors could be exclusively localized in the distal muscularized vessels, resulting in a persistent increase in the vasoconstrictor response to Ang II.

In conclusion, to our knowledge this is the first study designed to investigate the pulmonary expression and functional role of Ang II receptor subtypes during HPH development and subsequent reversal upon return to normoxia. Our main finding of upregulated expression of AT1 and AT2 receptors during hypoxia may provide a new basis for understanding the mechanisms by which Ang II plays a role in structural and hemodynamic changes in pulmonary vasculature during hypoxia and after return to normoxia. Among other endogenous vasoconstrictor and growth-promoting substances, endothelin (ET)-1 is also considered an important mediator of chronic HPH (48, 49). Functional alterations in endothelin receptor subtypes ETA and ETB have also been described in the lungs from chronically hypoxic rats (48), with a loss of ETB-mediated vasodilation and enhanced ETA-mediated vasoconstriction. Because Ang II has been found to increase endothelin concentration in vitro from endothelial cells (50), it is likely that both mediators act in synergy to favor pulmonary vasoconstriction and smooth muscle proliferation during development of HPH.