INTRODUCTION

Top Abstract Introduction References

CLINICAL PULMONARY EDEMA in the presence of heart failure primarily occurs because of elevated left atrial pressure (1, 34, 38). However, the relationship of elevated left atrial pressure to the resolution of alveolar edema is not well understood. Since work from several investigators has demonstrated that alveolar fluid transport depends on active Na⁺ transport across the alveolar epithelium (2, 9, 12, 18, 20, 23, 24, 26, 40), it is possible that active fluid transport protects against alveolar flooding in the presence of elevated lung microvascular pressures. With the exception of one study in newborn lambs (29), the effect of left atrial hypertension on lung liquid clearance has not been studied.

Therefore, the first objective was to determine the effect of moderate left atrial hypertension on alveolar fluid clearance in anesthetized, ventilated sheep. Because the initial experiments demonstrated that alveolar fluid clearance persisted at a normal rate despite moderate left atrial hypertension, the second objective was to investigate the potential role of endogenous epinephrine release in maintaining alveolar fluid clearance at normal levels during moderate left atrial hypertension. Two lines of evidence supported a role for endogenous epinephrine in sustaining normal alveolar fluid clearance in the setting of moderately elevated left atrial pressure. First, plasma epinephrine levels were increased during left atrial hypertension. Second, bilateral adrenalectomies reduced alveolar fluid clearance in the presence of left atrial hypertension by 30%. The third objective was to evaluate the potential effect of aerosolized ₂-agonist therapy as a method for stimulating alveolar fluid clearance to a maximal level in the presence of left atrial hypertension. However, aerosolized salmeterol (a lipophilic -adrenergic agonist) did not increase alveolar fluid clearance in the presence of left atrial hypertension, whereas aerosolized salmeterol did increase alveolar fluid clearance in sheep with normal left atrial pressure. Therefore, the fourth objective was to test the hypothesis that release of atrial natriuretic factor (ANF) in the presence of left atrial hypertension may blunt the normal response of the alveolar epithelium to ₂-adrenergic agonist stimulation. This hypothesis was suggested by prior in vivo (25) and in vitro (37) studies in which ANF reduced alveolar epithelial Na⁺ transport. In the sheep experiments in these studies, there was a modest twofold increase in plasma ANF after left atrial hypertension. In addition, in vitro studies with rat alveolar epithelial type II cells indicated that ANF inhibited Na⁺ pump activity in the presence of ₂-adrenergic stimulation.

The in vivo experiments were done in anesthetized, ventilated sheep so that systemic and pulmonary hemodynamics could be measured at the same time that alveolar and lung liquid clearance was measured. Measurement of lung lymph flow and the lymph-to-plasma protein concentration ratio made it possible to determine the effect of increased left atrial pressures on lung vascular filtration as well as to estimate absorption of protein-free alveolar fluid into the lung lymph, as we have done before (3, 20).

	METHODS

Sheep Preparation and General Experimental Protocol

1

1

The sheep were surgically prepared for collection of lung lymph as described previously (3, 20). The surgical preparation of the sheep usually required 2 h. A Foley catheter (20-Fr., Bard Urological Division, Covington, GA) was inserted into the left atrium to approximate the pathological condition of left atrial hypertension. The Foley catheter was inserted and secured in the left atrium, as we have done before (15).

In all experiments, after the surgical preparation, there was a 2-h baseline of stable heart rate, systemic blood pressure, pulmonary vascular pressures, and arterial blood gases. One hour into the baseline, the left atrial pressure was raised by inflation of the left atrial Foley catheter to reach a left atrial pressure of either 18 or 24 cmH₂O. A catheter was placed in the left atrium and recorded the left atrial pressure continuously. At the end of the baseline period, the test solution was instilled into both lungs.

Fiber-optic bronchoscopy was used to instill the test solution (3 ml/kg) directly into both the right lower and left lower lobes of the sheep lung. In some sheep, either aerosolized salmeterol (5 mg) or saline was delivered over a 1-h period after fluid instillation. Aerosolization was done with a whisper jet nebulizer system (Marquest, model 123014) in the inspiratory limb of the ventilatory circuit with a driving flow rate of 10 l/min.

Lymph samples were taken at 15-min intervals throughout the study, and blood was sampled hourly.

At the end of the 4-h experimental period, the sheep was exsanguinated. A median sternotomy was done to excise the lungs. Alveolar fluid samples from the distal air spaces of each lung were obtained by gently guiding a catheter to a wedged position and aspirating fluid, as we have done before in the sheep studies (3, 20). The gravimetric lung water in the lungs was then determined.

In selected experiments, serum samples for measurements of epinephrine levels were obtained once the sheep had stabilized in the baseline period and at 2 and 4 h after fluid instillation. Salmeterol levels were measured in the plasma, alveolar fluid, and lung lymph in selected sheep experiments. In two sheep, ANF was measured in the plasma at baseline and 2 and 4 h after elevation of left atrial pressure to 24 cmH₂O.

Preparation of the Instillate

Specific Experimental Protocols

Group I: Left atrial hypertension (24 cmH₂O; n = 5) vs. normal left atrial pressure (n = 4). Left atrial pressure was increased to 24 cmH₂O. After the baseline period, 3 ml/kg of the 5% albumin solution were instilled with a bronchoscope into each lower lung lobe. After instillation, 5 ml of 0.9% saline were nebulized over 60 min (nebulization was done as a control for the experiment in groups III and IV). Then, the sheep were processed as described in Sheep Preparation and General Experimental Protocol. The same protocol was followed in the four control sheep with no increase of left atrial pressure.

Group II: Left atrial hypertension (24 cmH₂O) in the presence of bilateral adrenalectomy (n = 4). After the surgical preparation, the bilateral adrenalectomies were done through posterior incisions. Left atrial pressure was increased to 24 cmH₂O. After the baseline period, 3 ml/kg of 5% albumin solution were instilled into both lower lung lobes as previously described. Then, 5 ml of 0.9% saline were nebulized over 60 min. Neosynephrine (phenylephrine HCl injection, Elkins-Sinn, Cherry Hill, NJ), an -adrenergic agonist, was used at a dose of 100 µg/min to support systemic blood pressure after the bilateral adrenalectomies. To provide controls for the bilateral adrenalectomies and neosynephrine infusion, three sheep with normal left atrial pressure underwent bilateral adrenalectomies with neosynephrine infusion. Alveolar fluid clearance in these three sheep was 30 ± 6%, similar to control sheep.

Group III: Left atrial hypertension (18 and 24 cmH₂O) with aerosolized or instilled salmeterol (n = 12). Initially, left atrial pressure was increased to 24 cmH₂O. Then, 3 ml/kg of the 5% albumin solution were instilled into each lower lung lobe. Salmeterol (5 mg in 5 ml 0.9% saline) was nebulized over 60 min. Initially, studies were done only with a left atrial pressure of 24 cmH₂O (n = 4). Because there was no effect of aerosolized salmeterol, left atrial pressure was reduced to a lower level, i.e., 18 cmH₂O (n = 5). Because again there was no effect, in additional sheep with left atrial pressure of 24 cmH₂O, a solution of 10⁶ M salmeterol was dissolved in the instillate (n = 3). All sheep were processed as described in Sheep Preparation and General Experimental Protocol.

Group IV: Normal left atrial pressures with aerosolized salmeterol (n = 4). After the baseline period, 3 ml/kg of the 5% albumin solution were instilled into each lower lobe as previously described. Salmeterol, 5 mg in 5 ml of 0.9% saline, was nebulized over 60 min. Then, the sheep were processed as described in Sheep Preparation and General Experimental Protocol.

Measurements

Radioactivity studies to measure flux of intravascular protein into the air spaces of the lung. Although the elevated hydrostatic pressure increases transvascular fluid filtration without altering barrier permeability to protein (as indicated in the lymph to plasma protein concentration ratios in Fig. 1), we wanted to verify that the integrity of the alveolar epithelium was maintained in the presence of moderately elevated left atrial pressure. Therefore, in three control sheep and in three sheep with left atrial hypertension (24 cmH₂O), we injected 10 µCi iv of ¹³¹I-labeled human serum albumin (¹³¹I-albumin; Frosst Laboratories, Montreal, Canada) during the baseline period. The quantity [counts · min¹ (cpm) · g¹] of ¹³¹I-albumin in the alveolar fluid was measured after 4 h. There were no differences in the counts of ¹³¹I-albumin in the alveolar fluids of sheep with left atrial hypertension compared with the counts in the alveolar fluids from sheep with normal left atrial pressure. Also, the total counts, expressed as a ratio of alveolar to plasma ¹³¹I-albumin at 4 h (39), indicated that there was <1 ml of plasma in the air spaces of the lung.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   A: alveolar liquid clearance measured by increase in alveolar protein concentration (final-to-instilled protein concentration ratio) in aspirated fluid from distal air spaces in sheep with and without left atrial hypertension. Final-to-instilled protein concentration was similar in sheep with normal left atrial pressures, compared with sheep with moderate left atrial hypertension. B: lung liquid clearance was similar in sheep with normal left atrial pressure vs. left atrial hypertension. C: lung lymph flow in sheep without () and with () left atrial hypertension. Shaded area indicates time in the experiment when Foley catheter was inflated to increase left atrial pressure to 24 cmH2O. Left atrial pressure remained elevated for the remainder of experiment. Arrow indicates start of instillation of the 5% albumin solution into distal air spaces. Lung lymph flow increased significantly more in sheep with left atrial hypertension than in control sheep (P < 0.05). D: lung lymph-to-plasma protein ratio in sheep without () and with () left atrial hypertension. Lung lymph-to-plasma protein ratio decreased significantly more after left atrial hypertension than in control sheep (P < 0.05). Data are means ± SD.

Trichloroacetic acid precipitation was carried out on the instillates and on selected samples from each experiment; it was established that the tracer ¹³¹I was always >98% bound to the protein.

Measurement of alveolar liquid clearance. As in previous studies (3, 20), alveolar liquid clearance was estimated by measuring the increase in the final alveolar protein concentration, compared with the initial alveolar (instilled) protein concentration (raw data shown in Figs. 1A, 2A, 3A, and 4A). Alveolar liquid clearance (ALC) was calculated as

(1)

where V_i is the volume of the instilled fluid (ml), Fw_i is the water fraction of the instilled fluid, and V_f is the volume of the final alveolar fluid (ml). V_f was estimated as V_f = (V_i × TP_i)/Tp_f, where TP refers to the total protein concentration of the initial (Tp_i) and of the final alveolar fluid (Tp_f).

Lung liquid clearance (excess lung water measurement). To determine the extravascular water in the lungs of the sheep, standard methods were used, as in our prior studies (3, 20). The excess water (E) in the lungs was calculated by using the following equation

(2)

where lung liquid clearance was determined by first calculating the excess lung water in the instilled lungs; W_e is the water weight of the instilled lungs; D_e is the dry weight of the instilled lungs, and P is the weight of the protein instilled with the instillate. W_c and D_c are the reference data from control lungs, as in our prior studies (3, 19), or from studies with left atrial hypertension with a left atrial pressure of 24 cmH₂O (10, 16). Thus lung liquid clearance was the difference between the instilled volume and the excess lung water remaining in the lungs 4 h after the 5% albumin instillation, divided by the instilled volume.

Determination of plasma concentration of epinephrine. Plasma epinephrine was measured by HPLC by a laboratory technician blinded to the conditions of the experiments. Plasma was taken from the animal at the baseline and at 2 and 4 h. One milliliter of blood was collected in a heparinized tube, as we have done before (28). Blood samples were immediately centrifuged at 3,000 g for 5 min at +4°C; 0.5 ml of plasma was transferred to an Eppendorf tube and quickly frozen to 70°C in acetone and dry ice. Samples were stored at 70°C until analyzed. Plasma samples were spiked with an internal standard and absorbed on activated alumina at alkaline pH. Epinephrine was eluted by 0.1 M perchloric acid, analyzed by reverse-phase HPLC using a C₈ column, and measured by the amperometric method with the use of an electrochemical detector. Correlation coefficient and detection limit of this method were 0.96 and 10 pg/ml, respectively. Salmeterol measurements were done courtesy of Glaxo Ware, Herfordshire, UK. Measurements were done by HPLC with fluorescence detection after sample preparation by solid-phase extraction, as previously described (7).

Determination of plasma concentrations of ANF. A radioimmunoassay (13) was used to measure concentrations of ANF in duplicate in plasma samples after prior extraction on Sep-Pak cartridges.

In Vitro Studies With Isolated Rat Alveolar Epithelial Type II Cells

Na⁺-K⁺-ATPase activity. Activity of the Na⁺-K⁺-ATPase was quantified by a radiometric monitoring of ouabain-sensitive ATP hydrolysis at maximal velocity (Vmax), as optimized previously (35). In brief, a crude cell homogenate was obtained by sonicating the cells. ATP hydrolysis was analyzed by monitoring P_i release with the use of [-³²P]ATP (ICN Biochemicals, Montreal, Quebec, Canada) as a tracer. Na⁺-K⁺-ATPase activity was calculated as the difference between the slopes of the regression lines of P_i release obtained in the presence and absence of 2 mM ouabain (Sigma Chemical). The data were standardized to cellular protein content determined by the method of Bradford (35).

12

7

12

6

Statistics

	RESULTS

Alveolar and Lung Liquid Clearance With Left Atrial Hypertension

Alveolar and Lung Liquid Clearance With Left Atrial Hypertension After Bilateral Adrenalectomy

View larger version (21K): [in this window] [in a new window]   Fig. 2.   A: alveolar liquid clearance was measured by the increase in alveolar protein concentration (final-to-instilled protein concentration ratio) in aspirated fluid from distal air spaces in sheep with left atrial hypertension with and without adrenal glands. Alveolar liquid clearance was reduced in sheep that had bilateral adrenalectomy done before instillation. (*P < 0.05). B: lung liquid clearance was also significantly reduced in sheep with bilateral adrenalectomies (*P < 0.05). C: lung lymph flow in sheep with bilateral adrenalectomy after left atrial hypertension was induced () and in sheep with left atrial hypertension with their adrenal glands intact (). Shaded area indicates time when Foley balloon catheter was inflated to increase left atrial pressure to 24 cmH2O. Left atrial pressure remained elevated for the rest of experiment. Arrow indicates instillation of 5% albumin solution into distal air spaces. Lung lymph flow was not different among sheep with adrenalectomies and sheep with intact adrenal glands. D: lung lymph-to-plasma protein ratio in sheep with bilateral adrenalectomy after left atrial hypertension () and in sheep with left atrial hypertension with intact adrenal glands (). Lung lymph-to-plasma protein ratio decreased in both groups compared with baseline, although there was a greater decline in lymph-to-plasma protein ratio in sheep with the intact adrenal glands (P < 0.05). Data are means ± SD.

Alveolar and Lung Liquid Clearance With and Without Left Atrial Hypertension in the Presence of Aerosolized Salmeterol

View larger version (22K): [in this window] [in a new window]   Fig. 3.   A: alveolar liquid clearance was measured by increase in alveolar protein concentration (final-to-instilled protein concentration ratio) in aspirated fluid from distal air spaces in sheep with left atrial hypertension treated with salmeterol. Alveolar liquid clearance was normal in presence of moderate left atrial hypertension and was not increased by aerosolized or instilled salmeterol in sheep with left atrial hypertension. B: there was no significant difference in lung liquid clearance. C: lung lymph flow in sheep with aerosolized salmeterol and left atrial hypertension () and with aerosolized saline and left atrial hypertension (). Shaded area indicates time in the experiment when left atrial pressure was elevated to either 18 or 24 cmH2O; pressure was maintained at that level for the duration of experiment. Arrow indicates instillation of 5% albumin solution into distal air spaces. D: lung lymph-to-plasma protein ratios in sheep with aerosolized salmeterol and left atrial hypertension () and with aerosolized saline and left atrial hypertension (). Data are means ± SD.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   A: alveolar liquid clearance was measured by increase in alveolar protein concentration (final-to-instilled protein concentration ratio) in aspirated fluid from distal air spaces in sheep with normal left atrial pressures with and without salmeterol. Alveolar liquid clearance was significantly increased in animals treated with aerosolized salmeterol (*P < 0.05). B: lung liquid clearance was also significantly increased by aerosolized salmeterol in these sheep (*P < 0.05). C: lung lymph flow in sheep with aerosolized salmeterol () and in control sheep with aerosolized NaCl (). Arrow indicates instillation of 5% albumin solution into distal air spaces. Lung lymph flow tended to increase more in salmeterol-treated sheep than in control sheep, although difference did not reach statistical significance. D: lung lymph-to-plasma protein ratios in sheep with aerosolized salmeterol () and in control sheep with aerosolized NaCl (). Lymph-to-plasma protein ratio was significantly less in salmeterol-treated sheep from 90 to 150 min in the study (P < 0.05). Lung lymph-to-plasma protein ratio decreased significantly more with aerosolized salmeterol than in control sheep. Data are means ± SD.

Hemodynamic Measurements With and Without Left Atrial Hypertension

Measurements of Lung Lymph Flow With and Without Left Atrial Hypertension

Salmeterol Levels

6

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Levels of salmeterol as measured in alveolar fluid, lung lymph, and plasma in sheep with normal and elevated left atrial pressure that were treated with aerosolized salmeterol. Box plots represent median and 25th to 75th percentiles; error bars span 10th to 90th percentiles.

ANF Studies

The binding assay that was carried out on membrane preparations of the alveolar type II cells suggests that C-ANF does not compete with ¹²⁵I-ANF, since only very high concentrations partially displaced ¹²⁵I-ANF from membrane receptors, thus suggesting that clearance receptors are absent on alveolar type II cells (Fig. 6). To discriminate between both subtypes of guanylyl cyclase receptors that could be present, we evaluated the displacement of ¹²⁵ANF by CNP, considered a natural ligand for guanylyl cyclase receptors of subtype B. The results indicated that only a very high concentration (10⁷ M) of CNP could inhibit the binding (Fig. 6). Overall, these results suggest that the guanylyl cyclase receptor of subtype A is the main ANF receptor on these cells. To determine whether this receptor was functional, the release of cGMP was measured after ANF stimulation. As shown in Fig. 7, there was a gradual increase in cGMP release with ANF stimulation.

View larger version (13K): [in this window] [in a new window]   Fig. 6.   Representative competitive binding curves of atrial natriuretic factor (ANF) on rat alveolar epithelial type II cells. Binding of 125I-labeled rat ANF [ANF-(99128)] to membranes of rat alveolar type II cells was progressively inhibited by increasing concentration of unlabeled ANF (), C-type natriuretic peptide (22-amino acid peptide; ), and C-ANF (). Binding is represented as %B/B0, where B and B0 represent binding with or without displacing peptides, respectively. Each point represents mean ± SE, duplicate determination.

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Measurement of cGMP release from alveolar epithelial type II cells stimulated with different concentrations of ANF for 90 min. There was a significant dose-dependent increase in cGMP release with ANF stimulation. Results are means ± SE of 5 separate experiments.

The impact of ANF on basal and stimulated Na⁺-K⁺-ATPase activity was then determined. Although ANF (10⁷ M) did not inhibit basal Na⁺K⁺-ATPase activity, it did inhibit the increase in activity induced by terbutaline (10² M) (Fig. 8). This effect could be reproduced by incubating the cells with dibutyryl-cGMP (10³ M) (data not shown).

View larger version (16K): [in this window] [in a new window]   Fig. 8.   Effect of -adrenergic agonist stimulation [terbutaline (terb)] on Na+-K+-ATPase activity of alveolar type II cells pretreated with ANF (107 M). Pretreatment with ANF inhibited stimulation of Na+-K+-ATPase by terbutaline. Results are means ± SE of 7 different experiments. * P < 0.05 vs. ANF+terbutaline, control, or ANF.

	DISCUSSION

Although several mechanisms that regulate fluid movement across the lung endothelial barrier in the presence of elevated lung vascular pressure have been identified (4, 34, 38), there is very little information regarding the regulation of lung fluid balance across the alveolar epithelial barrier in the presence of elevated lung vascular pressure. The experiments in this study were designed to determine the effect of acute elevations of left atrial pressure on net alveolar fluid clearance in sheep. Left atrial pressure was increased to 24 cmH₂O so that the effect of moderate left atrial pressure could be studied under steady-state conditions over 4 h in anesthetized, ventilated sheep. Lung lymph flow was measured to provide a quantitative index of lung vascular filtration as well as to provide an index of alveolar fluid reabsorption on the basis of a decline in the lymph-to-plasma protein ratio, as we have reported in prior studies (3, 20, 32). The first objective was to determine the effect of moderate left atrial hypertension on alveolar fluid clearance. Interestingly, the results indicated that the net alveolar fluid clearance was maintained at a normal level in the presence of moderate elevations of left atrial pressure. This was a remarkable finding, since lung lymph flow was increased nearly threefold over baseline levels (Fig. 1), indicating a marked increase in the transvascular fluid filtration in the lung. In addition, since lung liquid clearance remained at a normal level, the removal of the excess fluid transported to the interstitial space from the alveolar space was also maintained.

In the only previous work by Raj and Bland (29) on elevated pulmonary vascular pressure and lung liquid clearance, newborn lambs were used to assess the impact of a hydrostatic stress. These investigators found that elevated left atrial pressure decreased the clearance of instilled isotonic saline (6 ml/kg) in the first 2 h after instillation, although at 6 h there was no difference. In contrast, our experiments utilized a 5% albumin in Ringer lactate solution as the instillate, so that we could measure both alveolar and lung liquid clearance. Their 6-h results in the newborn lamb are similar to our 4-h data in the adult sheep in terms of no change in the lung liquid clearance in the presence of left atrial hypertension.

To determine the mechanism responsible for maintenance of normal alveolar fluid clearance in the presence of moderate left atrial hypertension, we tested the hypothesis that endogenous release of epinephrine might be responsible in part for sustaining alveolar fluid clearance. There were two lines of evidence that supported a role for endogenous epinephrine in sustaining alveolar fluid clearance in the setting of moderate left atrial hypertension. First, plasma epinephrine levels were elevated during left atrial hypertension to approximately two- to threefold normal levels. Second, bilateral adrenalectomies reduced alveolar fluid clearance in the presence of left atrial hypertension by ~30% (Fig. 2). Thus endogenous release of epinephrine appears to account for ~30% of the alveolar fluid clearance that occurred in the presence of moderate left atrial hypertension. It is remarkable that alveolar fluid clearance persisted at 70% of normal rates in the sheep with elevated left atrial pressures (Fig. 1).

Interestingly, aerosolized salmeterol, a potent lipophilic ₂-adrenergic agonist, did not increase alveolar fluid clearance in the sheep with left atrial hypertension, even when the pressure was increased to only 18 cmH₂O (Fig. 3). In contrast, aerosolized salmeterol markedly increased alveolar and lung liquid clearance when left atrial pressure was normal (Fig. 4). It is possible that the failure to respond to exogenous salmeterol in the presence of left atrial hypertension reflected maximal stimulation from the modest increase in endogenous epinephrine levels. This interpretation is, however, unlikely, since recent work indicates that progressively higher levels of epinephrine in dogs produce a dose-dependent increase in alveolar liquid clearance (17). Second, the response to endogenous catecholamines was submaximal, so, theoretically, additional ₂-adrenergic stimulation should have increased alveolar fluid clearance. An alternative explanation is that the effect of salmeterol was inhibited by a circulating endogenous factor released by left atrial hypertension.

It is well known that left atrial hypertension leads to ANF release (22), and it has been shown recently that ANF can inhibit both unstimulated and stimulated Na⁺ transport in the isolated perfused rat lung (25) as well as across cultured alveolar epithelial type II cells (37). To test the hypothesis that the failure of salmeterol to stimulate alveolar liquid clearance in these sheep could be related to ANF secretion, we first measured the level of circulating ANF in the sheep with left atrial hypertension. Plasma ANF increased by ~100% in two sheep after left atrial hypertension. Although the increase in plasma ANF was less than the increase seen in animals developing heart failure secondary to a cardiomyopathy (14), the magnitude of cardiac dysfunction was also less in our sheep. In view of the elevated plasma ANF levels, we thought it was reasonable to examine the effect of ANF on -adrenergic stimulation of Na⁺ transport in a well-established in vitro model of cultured rat alveolar type II cells. To evaluate this issue, we first determined whether there were functional ANF receptors on rat alveolar type II cells. In the intact lung, the most predominant receptor for ANF is the clearance receptor (27); however, in alveolar type II cells, the guanylyl cyclase-transducing receptors are exclusively present (Fig. 6). In fact, like Tharaux et al. (37), we found that the guanylyl cyclase receptor of subtype A is the predominant receptor on alveolar type II cells. These receptors are functional, since ANF stimulation caused a significant increase in cGMP release (Fig. 7). We then determined whether ANF could inhibit ₂-adrenergic stimulation of Na⁺ transport in alveolar type II cells. To address this question, we measured the activity of Na⁺-K⁺-ATPase, since it is an essential component of transcellular Na⁺ transport and it can be activated by -adrenergic agonists (35). Interestingly, as in the work of Tharaux et al. (37), ANF by itself did not inhibit Na⁺-K⁺-ATPase. However, ANF inhibited the stimulating effect of terbutaline on Na⁺-K⁺-ATPase (Fig. 8). Thus this in vitro data support the possibility that the lack of effect of ₂-agonist stimulation with salmeterol in sheep (Fig. 4) could be secondary to the presence of elevated levels of ANF in the lung. These results would then also support the hypothesis raised by two previous investigators suggesting that ANF may inhibit Na⁺ absorption in the lung (25, 37).

However, this inhibitory action of ANF in our studies or in the experiments by Tharaux et al. (37) was achieved with concentrations of ANF at 10⁷ M. This concentration exceeded the levels of ANF measured in the plasma in our sheep studies. Therefore, although the in vitro results are consistent with a possible role of ANF, they are not conclusive. The need to use larger quantities of ANF in vitro could be explained by a rapid degradation of the peptide when it is applied to the cell. This would mean that a greater concentration of ANF is necessary to obtain significant receptor stimulation in vitro. This is possible, since ANF is metabolized by the neutral endopeptidase 24-11 (16), an enzyme that has been shown to be present in the membrane of alveolar type II cells (11). However, another possibility is that the effect of ANF in sheep in vivo is determined by tissue concentrations of ANF, and, of course, our plasma assays do not accurately assess tissue or receptor concentrations. The plasma data simply demonstrate that there is more ANF released with left atrial hypertension. Interestingly, it has been shown in hamsters with a cardiomyopathy that concentrations of ANF in lung tissue increase with the evolution of heart failure (14). So, local tissue concentration of ANF may be higher and, therefore, could be responsible for the inability of salmeterol to stimulate clearance. In summary, the in vitro data suggest that an elevation of ANF may prevent the stimulatory effect of -adrenergic agonist therapy on alveolar liquid clearance during moderate left atrial hypertension.

A basic issue in our experiments was to explain how alveolar fluid could be removed from the air spaces and the interstitium of the lung in the presence of markedly elevated lung vascular filtration. First, there are several mechanisms that protect against alveolar flooding in the presence of moderate left atrial hypertension. There is a pressure gradient in the lung interstitium from the alveolar to the extra-alveolar interstitium, even in the presence of edema (5). Therefore, a significant fraction of the microvascular filtrate moves to the extra-alveolar compartment, where increases in lung lymph flow can facilitate the removal of some of the excess interstitial fluid (15, 34). Also, some of the interstitial edema fluid flows across the low-resistance visceral pleural mesothelium into the pleural space (6). Furthermore, it is well known that the alveolar epithelial barrier is very tight, resisting the movement of both low- and high-molecular-weight molecules (19, 20, 33, 36). Finally, based on the results of this study, the active ion-transport properties of the alveolar epithelial barrier provide an additional mechanism for removing alveolar fluid. Evidently, net alveolar epithelial fluid clearance can be maintained, as illustrated in these experiments, in the presence of a marked increase in lung vascular filtration with moderate elevations of left atrial pressure. Obviously, when left atrial pressure is increased to markedly higher levels (1, 38), progressive interstitial edema will result in a rise in interstitial pressure to a critical level that breaks the epithelial barrier at the distal airway and/or alveolar level and will result in bulk flow of interstitial edema fluid into the distal air spaces (8, 34). However, under conditions of moderate left atrial hypertension, overt alveolar flooding does not seem to occur (8). The data in this study support the critical role of alveolar epithelial fluid transport in preventing accumulation of edema fluid in the air spaces of the lung in the presence of moderate elevation of left atrial pressure. It is possible that some alveolar flooding occurs at moderate levels of left atrial hypertension, but it is likely, based on the data in this study, that mild alveolar flooding can be efficiently and rapidly removed across the alveolar barrier.

As already discussed, an important objective of this study was to determine whether aerosolized salmeterol, a long-acting lipophilic -adrenergic agonist, could increase alveolar fluid clearance in the presence of elevated left atrial pressure. There was no effect of aerosolized salmeterol in the presence of either 18 or 24 cmH₂O left atrial hypertension. However, the data indicate a substantial increase in alveolar fluid clearance in the presence of aerosolized salmeterol in sheep with normal left atrial pressure (Fig. 4). Delivery of salmeterol to the distal air spaces of the lung was reasonably efficient, with measured levels of salmeterol in the alveolar fluid compartment at a mean level of ~700 ng/ml. Interestingly, this level was equivalent to a concentration of 10⁶ M, a level that is similar to the concentration needed for maximal alveolar clearance in the ex vivo human lung with salmeterol (31). The studies with aerosolized salmeterol demonstrated the expected decline in the lymph-to-plasma protein ratio (Fig. 4) that is associated with the transport of more protein-free alveolar fluid from the air spaces into the lung interstitium, reflected by a decrease in the lymph-to-plasma protein ratio, as we reported before (3). The results of these studies in sheep predict that aerosolized ₂-adrenergic agonist therapy is likely to be ineffective until left atrial pressure is lowered to a normal range. Thus it appears that exogenous ₂-adrenergic therapy may be beneficial, but only when left atrial pressure is normalized.

There are some limitations to these studies. First, although the studies were done in anesthetized sheep, endogenous release of epinephrine was not prevented. Second, positive-pressure ventilation maintained lung volume at a higher functional residual capacity than would probably occur in the setting of spontaneous ventilation. At lower lung volumes, interstitial edema might be cleared less efficiently, a factor that could lower the threshold for alveolar flooding and/or slow the rate of alveolar fluid clearance. Although steady-state conditions were achieved over 4 h, longer-term studies of fluid clearance (12-24 h) were not done. Thus the rate of clearance that we measured over 4 h might not be sustained for longer time periods. Finally, halothane (2%) itself may cause a modest reduction in the rate of alveolar liquid clearance, at least in rats (30), although the levels of halothane used in this study were low (1%). Also, the basal rates of alveolar fluid clearance in these anesthetized sheep were similar to our published data for alveolar liquid clearance over 4 h in unanesthetized, spontaneously ventilating sheep (19).

In summary, alveolar epithelial fluid transport is maintained at normal levels in anesthetized, ventilated sheep for 4 h in the presence of moderate left atrial hypertension (24 cmH₂O). The normal alveolar fluid clearance was sustained, in part, by endogenous release of epinephrine in the presence of left atrial hypertension. Even without epinephrine stimulation, however, alveolar fluid clearance continued at 70% of normal levels in the presence of moderate left atrial hypertension. Aerosolized ₂-adrenergic therapy was only effective in accelerating the rate of alveolar fluid clearance in the presence of normal left atrial pressure. The inability of exogenous ₂-adrenergic therapy to stimulate alveolar fluid clearance in the presence of left atrial hypertension may have been related to the release of ANF, a factor that has been shown to downregulate vectorial Na⁺ transport across the alveolar epithelium both in vivo and in vitro. On balance, these in vivo studies in sheep support the conclusion that alveolar epithelial fluid transport is an important mechanism that prevents or minimizes the accumulation of alveolar edema fluid in the presence of moderate left atrial hypertension.