INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The use of inhaled nitric oxide (NO) as a selective pulmonary vasodilator in patients with pulmonary hypertension has been studied extensively, and is now undergoing clinical application (1). The effect of NO is limited to the pulmonary vasculature because it is rapidly inactivated by combination with hemoglobin (Hb), forming NO-Hb, which is then oxidized to methemoglobin in the presence of oxygen. Inhaled NO dilates only blood vessels that perfuse well-ventilated lung units, thus decreasing pulmonary artery pressure and intrapulmonary shunt. Inhalation of NO has proved to decrease mean pulmonary arterial pressure in a variety of pathologic conditions, such as persistent pulmonary hypertension of the newborn, congenital heart disease, primary or secondary pulmonary hypertension, and acute respiratory distress syndrome. Whereas the likelihood of toxicity at concentrations of NO in the therapeutic range appears minimal, numerous clinical studies report the occurrence of rebound increase in pulmonary hypertension above pre-NO controls with the withdrawal of inhaled NO therapy (2- 4). A recent study reports an almost constant rise in pulmonary arterial pressure during withdrawal of inhaled NO in infants with pulmonary arterial hypertension due to congenital heart disease (3). Two possible mechanisms by which this acute pulmonary vasoconstriction occurs have been suggested: inhaled NO might decrease the release of endogenous NO (5, 6), and/or decrease the sensitivity of the smooth muscle cells to endogenous NO (7).

The present study was designed to test these hypotheses. By using the isolated rat lung model, we have investigated the effect of a 48 h NO inhalation period at a dose of 30 ppm in intact rats: (1) on the pulmonary reactivity to vasoconstricting stimuli; (2) on the modulation by the endothelial endogenous NO release of the pulmonary vasoconstrictor and vasodilator responses; (3) on the sensitivity of the pulmonary artery smooth muscle to NO.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Male adult sprague dawley rats (200-350 g; Iffa Credo, Paris, France) were randomly assigned to one of the three experimental groups: control group, NO exposed group, and NO recovery group. All animals received humane care in compliance with the principles of Laboratory Animal Care formulated by the National Society for Medical Research and the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health (NIH Publication No. 80-23, revised 1978). All animals were kept 2 d in a 120 L inhalation chamber. The rats continuously breathed fresh air with (NO group; n = 39) or without (control group; n = 43) addition of NO at the concentration of 30 ppm. In the recovery group (n = 15) animals were studied 6 h after cessation of NO inhalation.

NO Exposure

The gases were blended by separate flowmeters for air and NO before they were introduced in the chamber. NO gas (CFPO; Paris, France) was stored in nitrogen (N₂) at the concentration of 900 ppm. The gases exiting the exposure chamber were discharged outside by using a plastic hose. FIO₂ and FICO₂ were measured with a fuel cell analyzer and an infrared analyzer, respectively (CPX/D; Medical Graphics, St. Paul, MN). NO and NOx concentrations were measured by an electrochemical method (Polytrons NO and NO₂; Dräger, Lübeck, Germany). The detection range is 0 -50 ppm for NO and NO₂ in 0.1 ppm increments. The electrochemical method is sensitive to about 0.1- 1 ppm (8, 9). Soda lime was added into the chamber to reduce CO₂ and NO_x concentrations. The high fresh gas flow allowed atmospheric exchange every 30 min. High chamber gas turnover rates and adding soda lime to the chamber reduced CO₂ and NO_x concentrations to < 1% and 2 ppm, respectively. NO concentration was 30 ppm and remained stable during the two-day inhalation period. FIO₂ were similar in the NO-exposed and control groups (20 ± 0.5% versus 20.5 ± 0.5%, respectively). Temperature in the chamber remained at 22- 24° C. During the exposure period rat chow and tap water were provided ad libitum and animal body weight remained stable. All the rats were studied within 30 min of removal of the chamber.

Isolated Perfused Rat Lungs

The rats were anesthetized with pentobarbital sodium (50 mg/kg intraperitoneally), intubated through a tracheostomy, and ventilated with room air. After sternotomy and intracardiac injection of 100 UI heparin, cannulas were inserted into the pulmonary artery and the left atrium. Heart and lungs were then dissected and suspended in a humidified chamber thermostated at 37° C. The lungs were perfused through the pulmonary artery cannula using a peristaltic pump (Ismatec; Bioblock, Paris, France) at a constant flow of 0.4 ml/g body weight. The perfusate was a physiological salt solution (PSS) of the following composition (in mM): 126 NaCl, 5.4 KCl, 0.83 MgSO₄, 19.0 NaHCO₃, 1.8 CaCl₂, 5.5 glucose, bovine albumin (4 g/100 ml). Meclofenamate 10 mg/ml was added to the perfusate at the start of the experiment to inhibit prostaglandin synthesis. The lungs were ventilated using a Harvard rodent ventilator (tidal volume 4 ml, respiratory rate 60 breaths/min, positive end-expiratory pressure 2.5 cm H₂O) with warmed and humidified gas (20% O₂-5% CO₂-75% N₂). The initial 20 ml of perfusate were discarded before recirculation was begun with the remaining 25 ml of solution. Perfusate temperature was maintained at 38° C. Pulmonary arterial pressure (Ppa) and pulmonary venous pressure (Ppv) were continuously monitored with P23 ID transducers (Statham, Paris, France). A cannulating probe (model 1517-025; Statham) connected to an electromagnetic flowmeter (model SP 2202; Statham) was placed in series with the perfusing circuit for continuous pulmonary perfusate flow (Q) monitoring. Flow and pressure signals were recorded on a multichannel chart recorder (Allco EN 48; Paris, France).

Fifteen minutes of equilibration were allowed before starting any experiment. Baseline Ppa was required to be  10 mm Hg or the lung was discarded.

Experimental Protocol

Vasoconstrictor response to angiotensine II (ANG II). Vasoconstrictor response to ANG II was examined in lungs from rats of control (n = 6), NO (n = 6) and recovery groups (n = 4). ANG II (1 mg in 100 ml) was delivered as an arterial bolus injection. Three boluses of ANG II were injected at 8 min intervals. The pressor response to the third ANG II challenge was recorded. After the third challenge of ANG II, 10 mMol of L-NAME were added into the reservoir. The pulmonary vasoconstrictor response to ANG II and L-NAME was reassessed 15 min later. Vasoreactivity to ANG II was also examined in five rats exposed to NO and five control rats who were treated during 2 d with the endothelin receptors inhibitor Bosentan (100 mg · kg¹ · d¹) administered once a day by gastric gavage.

Vasoconstrictor responses to U-46619. Dose-response curves in response to the thromboxane analogue U-46619 were studied in rats of control group (n = 6) and NO group (n = 6). Three doses of U-46619 were administered into the reservoir at 3 min intervals (0.08 · 10⁶, 0.16 · 10⁶, 0.24 · 10⁶ M).

Vasodilator response to arginine vasopressine (AVP). U-46619 was used to increase basal pulmonary arterial pressure. U-46619 was given in small increments until perfusion pressure increased by ~ 10 mm Hg. After the lung was stabilized, AVP was added to the perfusate reservoir as a 100 ml bolus of increasing doses, separated by 2 min intervals (final concentrations: 100 to 30,000 pg · ml¹). Vasodilation to AVP was investigated in rats of control (n = 6), NO (n = 6), and recovery (n = 6) groups. Moreover, four animals in the control group were tested after initial addition of 50 mM of L-NAME into the perfusate to ascertain that the vasodilation caused by AVP was endothelium dependent.

Vasodilator response to sodium nitroprusside. Vascular tone was elevated by ~ 10 mm Hg using U-46619. Sodium nitroprusside was administered as 100 ml bolus of increasing doses (final concentrations: 1.35 · 10⁶ to 1.35 · 10³ M) into the reservoir at 2 min intervals. Vasodilation to sodium nitroprusside was investigated in six animals in the control and NO groups, respectively.

Methemoglobin Determination

Venous blood from 21 rats (seven in the control group, seven in the NO group, and seven in the recovery group) was sampled in the inferior vena cava for optical determination of methemoglobin level (10).

Drugs

All drugs were obtained from Sigma Chemical (St. Louis, MO), except sodium nitroprusside (Hoffman-La Roche, Bâle, Switzerland), thiopental sodium (Abbott, Paris, France) and heparin (Choay, Paris, France). Sodium nitroprusside was dissolved in isotonic glucose. The other drugs were diluted in 0.9% saline. U-46619 was stored at 80° C and AVP at 20° C until use. Bosentan was a gift of Dr. R. Jones (Hoffman-La Roche). Bosentan was used as a suspension in 5% arabic gum and prepared fresh daily.

Statistical Analysis

All results are expressed as mean ± SE. Vasodilator responses were calculated as the percent reversal of U-46619 induced contraction. The dose-response relationships were analyzed by two-factor repeated-measures ANOVA using a modified t test as the post hoc analysis for individual comparisons. The non parametric ANOVA Kruskal Wallis test followed by a Mann and Whitney test for intergroup comparison was used to compare baseline perfusion pressure, pressor response to ANG II and methemoglobin levels. p Values less than 0.05 were considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Baseline Hemodynamics

Mean baseline pulmonary arterial pressure was not different between the NO group (5.76 ± 0.16 mm Hg), the control group (6.18 ± 0.18 mm Hg), and the recovery group (6.87 ± 0.35 mm Hg). Administration of L-NAME did not significantly modify baseline pulmonary arterial pressures in the three groups.

Response to ANG II. Isolated lungs of the NO group had a twofold increase in the vasoconstrictor response to ANG II as compared with lungs of control or recovery groups (p < 0.05) (Figure 1). The potentiation of ANG II pressor response by L-NAME (pressor response after addition of L-NAME divided by the pressor response before addition of L-NAME × 100) was lower in the NO exposed group as compared with the control and recovery groups (166 ± 12% for NO group, 231 ± 11% for control group and 237 ± 10% for the recovery group; p < 0.05). Administration of Bosentan reduced the pressor response to ANG II in the same proportion in the control and NO groups (control: 9.8 ± 0.8 mm Hg; control+Bosentan: 5.4 ± 0.7 mm Hg; NO: 20.0 ± 1.4 mm Hg; NO+Bosentan: 13.4 ± 1.5 mm Hg).

View larger version (21K): [in this window] [in a new window]   Figure 1.   Vasopressor response to angiotensin II (ANG II) in isolated rat lungs in the control, NO exposed and NO recovery groups. The pressor response (mean ± SE) is expressed as the absolute increase in pulmonary arterial pressure. *p < 0.05 versus control group; p < 0.05 versus NO recovery group.

Response to U-46619. Lungs from the NO group had a higher pressor response to U-46619 as compared with controls (Figure 2).

View larger version (12K): [in this window] [in a new window]   Figure 2.   Vasopressor response to U-46619 in isolated rat lungs in the control, and NO exposed groups. The pressor response (mean ± SE) is expressed as the absolute increase in pulmonary arterial pressure. *p < 0.05 versus control.

Response to AVP. The AVP induced vasodilation was lower in the NO group as compared with the control and recovery groups (p < 0.05) (Figure 3). Dose response curves to AVP were similar in the recovery and control groups. Pretreatment with L-NAME inhibited vasodilation to AVP in control animals which confirms that the AVP induced vasodilation is dependent on the endothelium (Figure 3).

View larger version (13K): [in this window] [in a new window]   Figure 3.   Endothelium dependent relaxation in isolated rat lungs as assessed by the dose response curves to arginine vasopressin (AVP) after precontraction with U-46619, in control, control with L-NAME, NO exposed, and NO recovery groups. Left panel: control and control with L-NAME; right panel: control, NO exposed, and NO recovery groups. Responses (mean ± SE) are expressed as percentage of increase in pulmonary perfusion pressure caused by U-46619. *p < 0.05 versus NO exposed group; p < 0.05 versus control with L-NAME.

Response to sodium nitroprusside. The ability of the non endothelium dependent vasodilator sodium nitroprusside to completely reverse the U-46619 induced vasoconstriction was similar in the NO and control groups (Figure 4).

View larger version (13K): [in this window] [in a new window]   Figure 4.   Endothelium independent relaxation in isolated rat lungs as assessed by the dose response curves to sodium nitroprusside after precontraction with U-46619 in control and NO exposed animals. Responses (mean ± SE) are expressed as percentage of increase in pulmonary perfusion pressure caused by U-46619.

Methemoglobin results. Methemoglobin was significantly higher in the NO group as compared with the control and recovery groups (0.29 ± 0.03 g/100 ml, 0.11 ± 0.04 g/100 ml and 0.06 ± 0.02 g/100 ml, respectively, p < 0.05).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study evaluated the consequence of a 48 h NO inhalation period (30 ppm) on pulmonary vasoreactivity and endothelial function. The results show that NO inhalation: (1) increases the reactivity of the pulmonary vasculature to vasoconstricting agents, (2) attenuates the potentiation of the vasoconstrictive reactivity by the eNOS inhibitor L-NAME, (3) depresses the endothelium-dependent vasodilation, whereas the endothelium-independent vasodilation remains unchanged. All these alterations were reversible 6 h after cessation of NO inhalation. These findings indicate that the release of endogenous NO was decreased by NO inhalation.

Inhaled NO was given at a low concentration (30 ppm) as recently recommended for therapeutic use (11). For most indications, concentrations of less than 10 ppm are sufficient to ameliorate pulmonary hypertension. However, concentrations of inhaled NO as high as 80 ppm may be needed for the management of neonates with persistent pulmonary hypertension (12). In order to minimize NO₂ accumulation in the housing chamber, the total flow gas was maintained at high levels and soda lime was used. This enabled us to maintain NO₂ under 2 ppm. Previous studies have indicated that continuous exposure to similar concentrations of NO and NO₂ does not induce lung endothelial damage (11, 13). Inhaled NO is rapidly inactivated by combination with hemoglobin, forming NO-Hb, which is then oxidized to methemoglobin in the presence of oxygen. As expected methemoglobin increased in the NO- exposed group but remained in the nontoxic range. Moreover, our isolated rat lungs model was perfused with a blood free perfusate to avoid potential side effects caused by methemoglobin.

The pulmonary vasoconstrictor responses to ANG II and U-46619 were markedly increased 30 min after discontinuation of inhaled NO. Several findings indicate that the increase in pulmonary vasoreactivity was related to an inhibition of the endogenous NO release: (1) The endothelium-dependent vasodilation to AVP in the NO-exposed rat lung was reduced. Numerous studies have shown that AVP elicits pulmonary vasodilation through release of NO in the isolated rat lungs (14, 15). Indeed, we found that vasodilation to AVP was blocked by the inhibitor of eNOS L-NAME in our control rat lungs. (2) The decrease in vasodilation to AVP was not due to a decrease in the smooth muscle response to NO and/or cGMP as the vasodilator response to the NO donor nitroprusside was unaffected after NO inhalation. This is in agreement with the observation of Buga showing that exposure of isolated pulmonary arteries to NO does not modify the sensitivity of the vascular smooth muscle cells to the direct relaxant effect of NO (5). (3) The potentiation of the contractile response to ANG II by the eNOS inhibitor L-NAME was much lower in the NO-exposed group indicating that the release of NO was decreased by NO inhalation. Previous studies have indicated that the potentiation of the contractile response induced by eNOS inhibitors is increased when lung endothelium release of NO is increased, and decreased or absent when the endothelium is dysfunctional or destroyed (16, 17). Our findings only indirectly demonstrate a reduction of NO release, and it will be critical to measure the level and the function of the eNOS protein to prove directly our hypothesis that eNOS enzyme activity decreases during NO inhalation.

Six hours after cessation of the NO inhalation, the pulmonary vasoconstrictor reactivity to ANG II, the endothelium-dependent vasodilatation, and the potentiation of the vasoconstrictor reactivity with the eNOS inhibitor L-NAME returned to control values. This indicates that the NO-induced inhibition of release of endogenous NO was reversible and that a 2-d NO inhalation period does not irreversibly damage pulmonary artery smooth muscle and endothelial cells.

In contrast with our findings of increased vasoconstrictor reactivity, the pulmonary basal tone was unchanged in the NO-exposed group. Indeed, despite pulmonary blood flow being kept at the same level throughout the entire study period, the baseline values of pulmonary artery pressures did not differ between the groups. This result is consistent with the observations that in rats, unlike in humans, the background release of NO is undetectable in lungs and does not seem to participate in maintaining the low pulmonary basal tone (17- 19). Indeed, inhibition of eNOS function has no effect on pulmonary basal tone in rats, but increases the responses to vasoconstrictive stimulations, most likely by decreasing the NO release stimulated by the vasoconstriction itself (17, 20). This agrees with our results showing that inhibition of eNOS with L-NAME failed to increase the basal pulmonary pressure but potentiated the vasoconstrictor response to ANG II.

A reduction in NO supply could allow other substances produced by the pulmonary endothelium to increase the vasoconstrictor reactivity. However, pretreatment of the isolated lungs with meclofenamate excludes the idea that prostanoid release played a role in the increased vasoconstrictor reactivity of the NO-exposed group. Inhibition of NO release may lead to an increased synthesis and/or release of endothelin. Indeed, experiments performed in cultured aortic endothelial cells, isolated peripheral arteries, and intact animals have shown that inhibition of NO synthesis increases the release of endothelin in the systemic circulation, and potentiates the vasoconstrictor response to endothelin (21, 22). In this study, however, administration of the endothelin receptor antagonist Bosentan similarly attenuated the pressor response to ANG II in control and NO-exposed rats, suggesting that endothelin does not account for the increase in pulmonary vasoconstrictive reactivity in the NO-exposed rats.

The results of the present study are consistent with the findings of Buga, who recently evaluated the effects of NO on relaxation of isolated pulmonary artery rings: a short (15 min) exposure of bovine intrapulmonary arterial rings to NO or NO-donor agents markedly inhibited endothelium-dependent relaxation without diminishing the sensitivity of the vascular smooth muscle to the direct relaxant effect of NO (5). By giving inhaled NO to intact animals at a concentration used in clinical practices (30 ppm) during a prolonged period (2 d), our study extends the physiological and clinical significance of the in vitro study of Buga. Roos (23) recently reported that chronic NO inhalation (20 ppm during 3 wk) decreases endothelium-dependent but also endothelium-independent vasodilation in lungs of hypoxic rats. The most likely explanation for the observed impaired NO release after NO inhalation is an inhibition of the eNOS. Indeed, recent studies using partially purified eNOS consistently demonstrated an inhibition of constitutive eNOS activity by a negative feedback mechanism (24, 25). A likely mechanism of this inhibition is NO binding to the heme iron molecule of eNOS. NO has been shown to interact with a number of heme proteins including hemoglobin forming NO-hemoglobin which is oxidized to methemoglobin (26, 27). Since methemoglobin increased in our NO-exposed rats, it suggests that NO inhalation may inhibit eNOS in a similar manner. Contrasting with our results, a recent study using a similar model indicates that 2 d of NO (100 ppm) inhalation does not alter the activity of the inducible lung NOS (28). This suggests that NO could exert a negative modulatory feedback on constitutive eNOS but not on inducible NOS function.

The present study has some limitations, and the extrapolation to human clinic should be cautious in relation to the species, the NO fraction used, and the absence of blood during the isolated lungs perfused period. However, preliminary study from our group indicates that similar results were obtained when the isolated lungs were perfused with blood, and when NO was given at 10 ppm (29).

Our findings of a down regulation of lung endogenous NO production after continued NO inhalation might account for important clinical observations during treatment with inhaled NO. Inhaled NO is an effective pulmonary vasodilator and is now undergoing clinical application in patients with primary or secondary pulmonary hypertension. Severe rebound pulmonary vasoconstriction at withdrawal of inhaled NO has been reported in some patients with pulmonary hypertension (2- 4). A down regulation of endogenous NO production or decreased smooth muscle NO sensitivity have been proposed as mechanisms of this rebound pulmonary hypertension (7). Our study does not indicate that NO inhalation reduces the sensitivity of the smooth muscle to NO; rather it demonstrates that the release of endogenous NO was decreased after NO inhalation. The lung eNOS has been implicated in man not only in the protection against pulmonary vasoconstriction as demonstrated in rats (18), but also in the maintenance of the pulmonary vascular tone (30). Moreover, a recent study demonstrates that patients with pulmonary hypertension, primary or secondary, have reduced eNOS expression in the pulmonary arteries (31). Therefore, our findings suggest that patients with pulmonary hypertension should be considered at risk of developing acute pulmonary vasoconstriction during withdrawal of inhaled NO.

In conclusion, a 48 h NO inhalation period at 30 ppm under normoxic conditions reversibly increases the pulmonary vasoconstrictor reactivity and decreases endogenous NO release. Inhibition of eNOS could be a possible mechanism.