Vol. 83, Issue 5, 1538-1544, 1997

Nitric oxide decreases lung liquid production in fetal lambs

James J. Cummings

Department of Pediatrics, State University of New York at Buffalo, Buffalo, New York 14222

ABSTRACT

INTRODUCTION

METHODS

RESULTS

DISCUSSION

ACKNOWLEDGEMENTS

FOOTNOTES

REFERENCES

ABSTRACT

Cummings, James J. Nitric oxide decreases lung liquid production in fetal lambs. J. Appl. Physiol. 83(5): 1538-1544, 1997.To examine the effect of nitric oxide on fetal lung liquid production, I measured lung liquid production in fetal sheep at 130 ± 5 days gestation (range 122-137 days) before and after intrapulmonary instillation of nitric oxide. Thirty-one studies were done in which net lung luminal liquid production (Jv) was measured by plotting the change in lung luminal liquid concentration of radiolabeled albumin, an impermeant tracer that was mixed into the lung liquid at the start of each study. To see whether changes in Jv might be associated with changes in pulmonary hemodynamics, pulmonary and systemic pressures were measured and left pulmonary arterial flow was measured by an ultrasonic Doppler flow probe. Variables were measured during a 1- to 2-h control period and for 4 h after a small bolus of isotonic saline saturated with nitric oxide gas (10 or 100%) was instilled into the lung liquid. Control (saline) instillations (n = 6) caused no change in any variable over 6 h. Nitric oxide instillation significantly decreased Jv and increased pulmonary blood flow; these effects were sustained for 1-2 h. There was also a significant but transient decrease in pulmonary arterial pressure. Thus intrapulmonary nitric oxide causes a significant decrease in lung liquid and is associated with a decrease in pulmonary vascular resistance. In a separate series of experiments either amiloride or benzamil, which blocks Na⁺ transport, was mixed into the lung liquid before nitric oxide instillation; still, there was a similar reduction in lung liquid production. Thus the reduction in lung liquid secretion caused by nitric oxide does not appear to depend on apical Na⁺ efflux.

pulmonary circulation; ion transport; birth transition; fetus

INTRODUCTION

THE FETAL LUNG IS DISTENDED with fluid that is actively secreted by the pulmonary epithelium (38). Before effective air breathing can occur in the newborn, this fluid must be rapidly removed and net luminal secretion must cease (5). Failure to adequately remove this luminal fluid can result in respiratory distress that lasts from hours to days (2). There is increasing evidence that this reversal from net liquid secretion to net resorption is an active process, under hormonal control (3, 4, 7, 8, 12, 14, 31, 32, 41). In addition, the resorption of lung liquid that occurs during spontaneous labor is dependent on active Na⁺ transport (12).

At the time of birth, dramatic changes must also occur within the pulmonary circulation to enable the lung to function as the organ of gas exchange. In the fetus, pulmonary blood flow is minimal because of high vascular resistance (35, 36). The transition from fetal to newborn life is accompanied by a decrease in pulmonary vascular resistance that results in a severalfold increase in pulmonary blood flow and a decrease in pulmonary arterial pressure (17, 18, 22). One or more mediators, including prostaglandins and nitric oxide, may be involved (1, 9, 13, 20, 27, 28, 36, 37). We and others have shown that certain prostaglandins decrease net lung liquid production (15, 24), suggesting a relationship between these two critical perinatal events. However, the effect of nitric oxide, perhaps a major mediator of the transitional pulmonary circulation, on lung liquid production is unknown.

In the present study I measured lung liquid production in fetal lambs from 122 to 137 days gestation, before and after instillation of an intrapulmonary bolus of nitric oxide. I also measured pulmonary and systemic hemodynamics and did repeat studies by using the Na⁺-transport inhibitors amiloride and benzamil. Nitric oxide caused a significant decrease in both lung liquid production and pulmonary vascular resistance. These effects were not altered by Na⁺-transport inhibition.

METHODS

All operative procedures and experimental protocols were approved by the Institutional Animal Care and Use Committee at the University at Buffalo.

3

4

RESULTS

Table  1.   Data from control and NO groups by experimental period Group Experimental Period Gestation, days n Mean Carotid Arterial Pressure, mmHg Mean Pulmonary Arterial Pressure, mmHg Mean Left Atrial Pressure, mmHg Mean Heart Rate, beats/min Control (saline) BaselineNO/salineRecovery 129 ± 5  6 44 ± 4 42 ± 5 43 ± 4  43 ± 5 42 ± 5 42 ± 4  3 ± 1 2 ± 1 2 ± 1  156 ± 15 155 ± 22 154 ± 25  10% NO (low dose) BaselineNO/salineRecovery 131 ± 4  8 46 ± 3 44 ± 4* 45 ± 4  47 ± 3 43 ± 4* 45 ± 3  3 ± 1 3 ± 1 3 ± 1  169 ± 13 170 ± 11 166 ± 16  100% NO (high dose) BaselineNO/salineRecovery 128 ± 7  5 42 ± 6 38 ± 7* 43 ± 6  44 ± 4 39 ± 6* 43 ± 6  2 ± 1 1 ± 1 2 ± 1  166 ± 11 171 ± 9 167 ± 17  10% NO after amiloride BaselineAmiloride NO/salineRecovery 130 ± 5  8 43 ± 3 42 ± 4 41 ± 3 43 ± 3 45 ± 3 44 ± 4 43 ± 3 44 ± 4 2 ± 2 3 ± 3 3 ± 2 3 ± 2 160 ± 16 158 ± 15 157 ± 18 159 ± 17 Values are means ± SD; n = no. of studies. NO, nitric oxide. * Significantly different from baseline period within same group, P < 0.05.

DISCUSSION

These studies show that nitric oxide instilled into the lung liquid of fetal sheep causes significant decreases in both lung liquid production and pulmonary vascular resistance. These effects are sustained for up to 90 min and can be elicited in fetuses as young as 122 days. This confirms previous work in which Iwamoto and Morin (23) documented the effects of instilled nitric oxide on the fetal pulmonary circulation and is the first study to look at its effects on lung liquid production.

Both lung liquid resorption and pulmonary vasodilation need to occur before effective pulmonary gas exchange, and these events appeared to be triggered at the moment of birth (6, 19, 25). Nitric oxide has recently been implicated in the birth-related decrease in pulmonary vascular resistance. Abman et al. (1) found that infusion of N^G-nitro-L-arginine, a selective inhibitor of nitric oxide production from L-arginine, into late-gestation fetal lambs immediately before cesarean delivery blunted the normal rise in pulmonary blood flow and drop in pulmonary vascular resistance. More recently, Cornfield et al. (13) found that the effects of birth-related stimuli, such as expansion of the lungs with gas, increasing Pa_O2, and increasing shear forces within the pulmonary circulation, on pulmonary blood flow and pulmonary vascular resistance were markedly attenuated if fetal lambs were pretreated with N^G-nitro-L-arginine. In both studies, however, giving nitric oxide to the animals restored the normal physiological responses.

In this study, we gave nitric oxide in aqueous form into the lung liquid. That the epithelial surface may be the best route of exogenous nitric oxide administration has been suggested by Frostell et al. (21), who noted that this route would prevent immediate scavenging by oxyhemoglobin. The dose chosen was based on preliminary observations by Iwamoto et al. (23), who studied the cardiovascular effects of nitric oxide in fetal lambs. They found that a local concentration of ~10⁵ M caused consistent decreases in pulmonary vascular resistance (23). In our study of tracheal instillation of nitric oxide, the pulmonary epithelial surface was exposed to concentrations in the range of 10⁵ to 10⁴ M. This concentration had measurable effects on the pulmonary circulation.

Interestingly, the decrease in pulmonary arterial pressure after nitric oxide instillation was not as marked in our Na⁺-blocker experiments. When either amiloride or benzamil was given before nitric oxide instillation, pulmonary pressure decreased, on average, by 2 mmHg, compared with 4 mmHg when Na⁺ blockers were not given. Whether this represents a true difference between these groups is unclear; when pulmonary vascular resistances are compared, they decreased by virtually the same amount (48 vs. 50%, respectively).

A small but significant decrease in systemic pressure was also seen after nitric oxide instillation and was more marked at the higher dose. It is unlikely that this was a direct effect of nitric oxide on the systemic circulation because any nitric oxide reaching the systemic circulation (e.g., by absorption into the bronchial circulation) would be rapidly scavenged within the fetal bloodstream. Rather, it may be a secondary effect of the marked increase in pulmonary blood flow, effectively causing a "steal" phenomenon from the systemic circulation (i.e., by decreasing in right-to-left ductal shunting). Consistent with this notion is the observation that the largest drops in systemic pressure occurred in fetuses that had the largest increases in pulmonary blood flow. Still, a direct effect on the systemic circulation cannot be ruled out.

We measured pulmonary blood flow continuously and found that the effect of nitric oxide was seen within 15 s; this probably represents the time it takes for the gas to diffuse through the lung liquid and reach the epithelial surface. The method by which lung liquid production is measured does not allow for continuous measurement; nevertheless, the reduction in lung liquid production could usually be seen at the first time point (10 min). The effects on both pulmonary hemodynamics and lung liquid production began to decline by 1 h and returned to baseline by 2 h after instillation.

In this study, instillation of nitric oxide caused only a modest reduction in lung liquid production; in no case was liquid production reversed. However, the effect of nitric oxide on pulmonary hemodynamics was also modest; pulmonary blood flow increased two to three times, compared with the seven- to ninefold increase normally seen at birth. Although it is possible that higher concentrations of nitric oxide may have more pronounced effects, I did not see any significant difference in pulmonary effects between our low- and high-dose experiments. Furthermore, it is likely that higher doses of nitric oxide, even when given directly to the pulmonary epithelial surface, might have adverse systemic effects.

In the fetus at term, lung liquid production represents a balance between Cl secretion, which increases luminal liquid volume, and Na⁺ reabsorption, which reduces liquid volume (38). In the present studies, blocking Na⁺ transport with amiloride or benzamil did not prevent the reduction in lung liquid production after nitric oxide instillation, raising the possibility that nitric oxide may have reduced lung liquid production by affecting Cl transport. This possibility could not be explored in the in vivo model used and would have to be confirmed or refuted by in vitro studies of ion transport by respiratory epithelia.

Conceivably, the negative results seen in the Na⁺-transport blockade might have resulted from inactivation of those agents by nitric oxide, which is known to be a very reactive molecule. This possibility was lessened by the fact that the blocker was given well before the nitric oxide is instilled so that Na⁺-transport blockade was in place before the epithelia were exposed to nitric oxide. Nevertheless, subsequent to the above studies, I performed four experiments in three fetuses (132 ± 3 days) in which I instilled a -adrenergic agent (terbutaline) after first giving amiloride and nitric oxide. If amiloride were inactivated and no longer blocked Na⁺ transport, there would have been an adrenergic-induced lung liquid resorption, but there was no change in net liquid production after terbutaline in any of the four experiments.

My baseline values for lung liquid production are generally higher than some others have reported, but I think that the data are reliable, for several reasons. First, my estimates of total lung luminal fluid volume by using this method were well within reported values for fetal sheep at this gestation, 25-30 ml/kg (5, 11, 19, 25, 29, 32). Second, I ruled out significant transepithelial tracer leakage by sampling fetal plasma at the beginning and end of each study. Third, I ruled out tracer leakage around the Foley balloon or across the tracheal suture site by measuring tracer activity in amniotic fluid at the beginning and end of each study. Finally, although the values for lung luminal liquid production measured in this study are generally higher than some studies using a tracheal loop, they are within the range reported by others using tracer-dilution techniques (10, 30, 32). It is possible that the differences I found relate to differences between my system and the "tracheal loop" system that others (10, 12, 41), including myself (16), have used. Previous preparations used a large-bore tracheal loop that permanently diverted drainage from the distal trachea. In this tracheal loop system, the trachea is snugly sutured around the catheter at its proximal and distal ends, and fluid efflux is diverted into a loop of tubing that can be from 100 to 200 cm long. Not only does this loop system divert efflux from the normal pathway but it also increases the resistance to lung liquid efflux; it is possible that these factors may alter the subsequent measurement of lung liquid production. I think my innovation using a small-caliber Foley-type catheter represents an improvement over the previous loop method; during studies only a very small section of upper trachea is omitted, and between studies fluid efflux is not diverted into an exterior loop of tubing but is allowed to occur along the natural path.

When it was noted that some of our values for Jv seemed rather high, we started to look for loss of label into the fetal/maternal circulation as well. In 13 studies, we used gamma detection on fetal serum, maternal serum, and the maternal thyroid before and immediately after each experiment. In no case did we find a significant change in activity. This, along with our negative data from amniotic fluid, verified that our label remained within the lung during the course of the experiment.

This study shows that nitric oxide, a 3,5-cyclic guanosine monophosphate-dependent pulmonary vasodilator, reduces lung liquid production. We have previously shown that nitric oxide-3,5-cyclic guanosine monophosphate-independent pulmonary vasodilators, including prostaglandin D₂ and the leukotriene blocker FPL-55712, also reduce lung liquid production (15). Conversely, mechanical withdrawal of lung liquid in the fetal lamb increases pulmonary blood flow (34, 40). Taken together, these findings suggest that control of lung liquid production and pulmonary vascular resistance may be interrelated.

During spontaneous labor, reduction in lung liquid production is dependent on Na⁺ transport (12). However, in this study I found that the reduction in lung liquid production caused by nitric oxide instillation was unaffected by amiloride or benzamil, Na⁺-transport blockers. This finding, in addition to its modest effect on lung liquid production, suggests that nitric oxide alone is not responsible for the dramatic changes in lung liquid production that occur in the perinatal period. Just as a variety of mediators may be important in the transitional pulmonary circulation, it is likely that lung liquid production is regulated by a combination of agents that act in concert at birth to promote lung liquid resorption. For example, we and others have previously shown that certain prostaglandins, which may be important mediators of the transitional circulation (26, 27), also reduce lung liquid production (15, 24).

ACKNOWLEDGEMENTS

The author thanks H. Wang and D. Swartz for technical assistance.

FOOTNOTES

Address for reprint requests: J. J. Cummings, Dept. of Pediatrics, Div. of Neonatology, Children's Hospital of Buffalo, 219 Bryant St., Buffalo, NY 14222.

Received 14 May 1997; accepted in final form 26 June 1997.

REFERENCES