INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

THE FETAL LUNG is a liquid-filled organ with a high resistance to pulmonary blood flow. Significant adaptive changes occur in the fetal lung at birth to prepare it for the function of gas exchange during postnatal life. These changes include distension of the lung with air and a rapid increase in pulmonary blood flow to provide an optimum ventilation/perfusion match. A rapid and sustained decrease in pulmonary vascular resistance (PVR) and pulmonary artery pressure during the first breath facilitate this adaptation. The pulmonary vasodilation is initiated by a number of birth-related stimuli, such as an increase in oxygen tension (3), distension of the lung (17, 30), establishment of an air-liquid interface (6), and hemodynamic shear stress (1). Oxygen appears to be the primary stimulus for this transition from a low to a high O₂ environment (3, 20).

The effect of oxygen on fetal pulmonary circulation appears to be mediated primarily by the release of nitric oxide (NO) (2, 29) and vasodilator prostaglandins (17) from the vascular endothelium. A number of factors including steroids (7) and estrogens (16, 22) enhance the response of the pulmonary circulation to physiological stimuli in fetal lambs. However, these factors appear to prepare the pulmonary vascular bed for vasodilation but are not required for the rapid decrease in PVR that occurs at birth in response to oxygen (21). On the basis of the rapid increase in oxygen tension in the fetal lung during the transition process, we proposed the hypothesis that oxygen exerts its effects on pulmonary circulation by increasing the rate of oxidative phosphorylation and release of ATP. Because birth-related pulmonary vasodilation is initiated by several factors, including O₂, lung distension, and shear stress, we investigated the role of oxidative phosphorylation and ATP release in mediating vasodilation caused by exposure to these stimuli independent of each other. The objectives of our study were to investigate the effects of inhibition of oxidative phosphorylation on plasma ATP levels and pulmonary hemodynamics during exposure of fetal lambs to 1) oxygen alone, 2) lung distension without a change in arterial PO₂ (Pa_O2), 3) lung distension accompanied by an increase in Pa_O2, and 4) shear stress caused by a transient increase in pulmonary blood flow without lung distension or oxygenation. In addition, we assessed the potential nonspecific effects of inhibitors of oxidative phosphorylation on the pulmonary vascular response to infusion of ATP and of the endothelium-dependent and -independent vasodilators acetylcholine and S-nitroso-N-acetylpenicillamine (SNAP), respectively.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Study design. Fifty-nine fetal lambs were studied at 132 ± 2 days of gestation, term gestation being 140 days. Thirty lambs were divided into three groups. Ten lambs were controls to investigate the effects of lung distension and oxygenation on pulmonary vascular pressures, pulmonary blood flow, and plasma ATP levels. Ten lambs each were pretreated with 2,4-dinitrophenol (DNP), an uncoupler of oxidative phosphorylation, or antimycin-A, an inhibitor of cytochrome electron transport. Each lamb had sequential ventilation with 10% O₂-90% N₂, 21% O₂-79% N₂, and 100% O₂. Fifteen additional lambs were divided into three groups. Five lambs were studied to assess the effect of administration of 100% O₂ to the ewe on fetal pulmonary hemodynamics and plasma ATP levels. Five animals each had pretreatment with DNP or antimycin-A before studies with O₂ exposure. In addition, two groups of five animals each had the response to shear stress, acetylcholine, and SNAP determined before and after pretreatment with DNP or antimycin-A. These studies were done to assess the potential nonspecific effects of DNP and antimycin-A on endothelium-dependent and -independent vasodilation. Four fetal lambs had determination of the pulmonary vascular response to infusion of ATP solution into the pulmonary artery before and after pretreatment with DNP or antimycin-A. This study was approved by the Animal Care and Use Committee of the Medical College of Wisconsin (Milwaukee, WI) and the Animal Investigation Committee of Wayne State University (Detroit, MI).

Surgical preparation. The ewe underwent a sterile surgical procedure under general anesthesia with isoflurane and oxygen administered via an endotracheal tube. The fetus was exposed by a subumbilical midline incision and an incision on the uterus close to the fetal head. Catheters made from Tygon tubing (Baxter Diagnostics; McGraw Park, IL) were inserted into the ascending aorta and right atrium via the carotid artery and jugular vein. In the 30 animals that underwent ventilation studies, the fetal trachea was exposed through the same incision, and a 4.0-mm endotracheal tube was inserted into the trachea and secured by a purse string suture. The proximal end of the trachea was also cannulated and attached to the endotracheal tube with a long polyvinyl chloride tube exteriorized to the mother's back as a loop. This allowed uninterrupted drainage of fetal lung fluid into the upper airway and amniotic cavity during recovery from surgery. A left lateral thoracotomy was then done on the fetus to insert catheters into the left pulmonary artery and left atrium and an ultrasonic flow transducer (size 6 S, Transonic Systems; Ithaca, NY) around the left pulmonary artery. An inflatable vascular occluder was placed around the ductus arteriosus (Invivo Metric Systems; Heldsberg, CA) in the 10 animals that had the evaluation of response to shear stress. A large-bore catheter was placed in the amniotic cavity to replace the amniotic fluid with lactated Ringer solution and to measure the amniotic cavity pressure, which served as a zero reference for fetal intravascular pressures. The catheters and cable for the flow transducer and vascular occluder were exteriorized to the flank of the ewe. The fetus was returned to the uterine cavity, and the uterine and abdominal incisions were closed. Antibiotics (2,000,000 units of procaine penicillin and 250 mg of gentamicin im) were administered to the ewe on the day of surgery and on each postoperative day.

Study protocol. Each animal was allowed to recover for 3 days before experiments were done. For the study, the ewe was placed in a cage in the study area and was allowed free access to alfalfa pellets and water. The fetal intravascular catheters and amniotic fluid catheter were connected to strain-gauge manometers (P23 XL, Spectramed Electronics, Critical Care Division; Oxnard, CA), and the flow transducer was connected to an ultrasonic transit time blood flowmeter (Transonic Systems). Zero flow output from the flowmeter was used as the baseline, and the mean flow output above baseline was used as the mean pulmonary blood flow. Aortic, pulmonary arterial, left atrial and right atrial pressures, and left pulmonary blood flow were recorded on a Grass model 7D polygraph (Grass Instruments; Quincy, MA). The PVR for the left lung was calculated as the difference between the mean left pulmonary arterial and left atrial pressures divided by left pulmonary blood flow, measured as milliliters per minute.

Drug preparation. DNP and antimycin-A were obtained in powder form (Sigma; St. Louis, MO), and solutions for infusion were made on each study day by dissolving the total dose of drug (1 mg DNP and 0.1 mg antimycin-A) initially in 1 ml ethanol and dilution with 19 ml of 0.9% saline to a final concentration of 50 µg/ml for DNP and 5 µg/ml for antimycin-A. For control experiments, a solution of 1 ml ethanol in 20 ml of 0.9% saline was used for infusion at the same time point. Acetylcholine and SNAP (Sigma) were dissolved in 0.9% saline to final concentrations of 10 and 1 µg/ml, respectively. The disodium salt of ATP (Sigma) was dissolved in sterile water for a final concentration of 1 mg/ml (15).

Studies with oxygen exposure of the ewe. For each experiment, the ewe was brought to the study area, and the fetal lamb was given either 1) 1 mg DNP, 2) 0.1 mg antimycin-A, or 3) vehicle used to dissolve DNP and antimycin-A. Baseline hemodynamic variables were recorded, and blood samples for ATP levels were drawn from the pulmonary artery and left atrium 10 min after the completion of drug infusion. Hemodynamic variables had returned to baseline by this time. The experiment with oxygen exposure alone was done with 100% O₂ administered to the ewe via a plastic bag placed loosely over its head. Oxygen was given at a flow rate of 15-20 l/min, and blood gas tensions and blood pH were determined in arterial blood samples from the fetus at baseline and 10, 20, and 30 min during the exposure to oxygen. Our previous study (11) has demonstrated that exposure of the ewe to O₂ in this manner results in a 7 ± 2 Torr increase in fetal Pa_O2 and a 20% increase in arterial O₂ saturation. Blood samples were drawn from the pulmonary artery at baseline and at 10, 20, and 30 min to determine the plasma ATP levels. Hemodynamic variables were measured at the same time intervals.

Protocol for ventilation studies. The ewe was brought to the study area, and the fetal lamb was given infusions of DNP, antimycin-A, or vehicle for these agents as described above. Baseline hemodynamic variables were recorded, and blood samples for ATP levels were drawn from the pulmonary artery and left atrium 10 min after the completion of drug infusions. The tubing connected to the endotracheal tube was then detached from the tube in the proximal end of the trachea. The fetal lung fluid was allowed to drain by gravity. Artificial surfactant (Exosurf Neonatal, Burroughs Wellcome; Research Triangle Park, NC) was given in a dose of 100 mg/kg and flushed in with 10 ml of air. Ten minutes after administration of the surfactant, ventilation of the lungs was initiated with 10% O₂ using a time-cycled, pressure-limited ventilator (Sechrist infant ventilator model IV-100B, Sechrist Industries; Anaheim, CA) at a frequency of 40-60 breaths/min, inspiratory time of 0.5 s, peak inspiratory pressure of 30-40 cmH₂O, and a positive end-expiratory pressure of 4 cmH₂O. The ventilator settings were adjusted to keep the Pa_O2 at 30-40 Torr and the pH at 7.30-7.40. The ventilator pressure, rate, and inspiratory time were not changed after the initial adjustments were made. After 10 min of ventilation at 10% O₂, the fraction of inspired oxygen (FI_O2) was increased to 21% and 100% O₂ at 10-min intervals. Hemodynamic variables, arterial blood gas tensions, and ATP levels were measured 10 min after the beginning of ventilation at each FI_O2. These FI_O2 were chosen from preliminary studies showing no change in Pa_O2 at 10% O₂, a two- to threefold increase in Pa_O2 at 21% O₂, which mimics the change seen at birth, and a fivefold increase in Pa_O2 at 100% O₂, which mimics the increase seen at 2-4 h of age in the newborn.

Evaluation of response to hemodynamic shear stress. Baseline variables were recorded over a 20-min period during the infusion of DNP, antimycin-A, or vehicle. A rapid inflation of the vascular occluder around the ductus arteriosus was then achieved with 1 ml normal saline injected into the lumen of the occluder cable, and the inflation was maintained for 30 min. Previous studies (1) have demonstrated that the rapid increases in pulmonary artery pressure and pulmonary blood flow and the decrease in PVR during this mechanical constriction of the ductus arteriosus are maintained for at least 30 min. The hemodynamic variables were continuously recorded for 30 min. Blood samples for determination of ATP levels were drawn 30 min after the inflation of the occluder in control experiments to determine whether shear stress is associated with an increased release of ATP. We found that the increase in ATP levels and pulmonary blood flow in the fetus during exposure of the ewe to oxygen peak at 30 min (11). The occluder was then deflated, and the flow and pressure were allowed to return to baseline before further experiments were done.

Effects of DNP and antimycin-A on response to endothelium-dependent and -independent vasodilators. These studies were done on the animals assigned to shear response studies on a different day. Animals assigned to DNP and antimycin-A groups were kept in the same treatment group. Each animal was given an infusion of DNP, antimycin-A, or vehicle into the left pulmonary artery, and baseline variables were recorded. An infusion of acetylcholine was then given into the left pulmonary artery at doses of 1, 2, and 5 µg/min for 2 min at each dose, and the hemodynamic variables were recorded at the end of the 2-min infusion. Doses were given in a random order, and the animal was allowed to recover for 5-10 min between each dose. The hemodynamic variables returned to baseline between each dose of acetylcholine. After a 30-min recovery period, each animal was given an infusion of SNAP in a dose of 0.1, 0.2, and 0.5 µg/min for 2 min into the left pulmonary artery (13). Hemodynamic variables were recorded at the end of a 2-min period, and the animal was allowed to recover for 5-10 min between each dose of SNAP.

Effect of DNP and antimycin-A on the response to ATP infusion. Fetal lambs were pretreated with DNP, antimycin-A, or vehicle on separate days, followed by infusion of ATP at 0.1-10 µM concentrations into the left pulmonary artery. These concentrations of ATP reproduce the increase in plasma ATP levels noted during the transition of the pulmonary circulation at birth (14) and produce minimal and maximal changes in PVR in fetal lambs (15). The order of treatments on different days was changed for each animal in the study. Hemodynamic variables were recorded at the end of 10 min of infusion of each dose of ATP. Lambs were allowed to recover for 10 min between each dose of ATP as described previously (13, 15).

Assay for ATP. Blood samples for ATP levels were collected rapidly into heparinized syringes and were placed immediately into EDTA tubes kept at 4°C. The samples were immediately centrifuged at 800 g and 4°C, and plasma was separated and kept frozen at 20°C. Assay for ATP was done using the firefly luciferin-luciferase method as we previously reported (11, 14). ATP is consumed and light is emitted when firefly luciferase catalyzes the oxidation of D-luciferin. The amount of light emitted during the reaction is proportional to the availability of ATP. Samples were diluted 1,000-fold and were run in triplicate against a 10-point standard curve. Internal standardization for each sample was also done in triplicate using a 1 µM solution of ATP (10 µl/100 µl sample). Luminescence was measured using a Berthold Luminometer (Lumat LB 9501, EG&G Berthold Analytical Instruments; Nashua, NH). Plasma ATP levels were expressed as micromoles per liter. The intra-assay coefficient of variation was 1.3-2.6%, and the interassay coefficient of variation was 7%. The smallest detectable concentration of ATP using this method was 1 pM.

Statistical analysis. Data are shown as means ± SD. Comparison of baseline data with those obtained during ventilation at each FI_O2 was done by single-factor ANOVA for repeated measures. Comparison of control data with those obtained after each drug infusion was done by two-way ANOVA (32); the two factors affecting the outcome were FI_O2 and the presence or absence of inhibitors of oxidative phosphorylation. When significant differences (P < 0.05) were found, a Duncan's multiple-range test was done to determine which means were different.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The fetal lambs were of normal weight (2.9 ± 0.5 kg) and had normal blood pH (7.33 ± 0.08) and PO₂ (19 ± 4 mmHg or 2.53 ± 0.53 kPa) at the time of study. Effects of exposure to oxygen without lung distension were as follows: administration of 100% O₂ to the ewe resulted in a 5 ± 2-mmHg increase in fetal Pa_O2 (Table 1) and a 19 ± 4% increase in arterial O₂ saturation. The plasma ATP levels in the pulmonary artery increased significantly in control animals during exposure of the ewe to 100% O₂ (Fig. 1). Fetal lambs treated with DNP and antimycin-A had no increase in plasma ATP levels in the pulmonary artery during oxygen exposure. In independent experiments, DNP and antimycin-A did not affect the measurement of known amounts of ATP added to fetal lamb plasma. There was a fourfold increase in pulmonary blood flow and a sixfold decrease in PVR in the control experiments during O₂ exposure (Fig. 2). Pretreatment of animals with DNP and antimycin-A attenuated the changes in pulmonary blood flow and PVR in response to O₂ exposure (Fig. 2).

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Plasma ATP levels in the pulmonary artery (PA) measured during exposure of the ewe to oxygen (A) and during ventilation of fetal lungs with different fractions of inspired oxygen (FIO2; B). Data are means ± SD for 5 animals in each group for oxygen exposure and 10 in each group for ventilation studies. ATP levels increased during oxygen exposure and during ventilation with 21% and 100% O2 in the control group only. 2,4-Dinitrophenol (DNP) and antimycin-A (AMA) significantly inhibited the increase in ATP levels during oxygen exposure and ventilation with 21% and 100% O2. *P < 0.05 vs. baseline (BL); #P < 0.05 vs. the control group.

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Changes in left PA blood flow (A) and pulmonary vascular resistance (PVR; B) in the left lung of the fetus during exposure of the ewe to 100% O2. Data are means ± SD for 5 animals in each group. Left PA blood flow increased and PVR decreased during O2 exposure in the control group, whereas DNP and AMA inhibited these changes during O2 exposure. *P < 0.05 vs. baseline; #P < 0.05 vs. the control group.

Effects of lung distension alone and with oxygenation. Ventilation of fetal lambs with 10% O₂ did not change the Pa_O2 significantly, whereas the Pa_O2 increased by twofold during ventilation with 21% O₂ and sixfold during ventilation with 100% O₂ (Table 1). There was a significant decrease in Pa_O2 and increase in pH during ventilation at all three FI_O2 (Table 1). The plasma ATP levels in the pulmonary artery increased significantly in control animals during ventilation of fetal lungs with 21% and 100% O₂ (Fig. 1) but not during ventilation with 10% O₂. Fetal lambs treated with DNP and antimycin-A had no increase in plasma ATP levels in the pulmonary artery during ventilation with 21% and 100% O₂ (Fig. 1).

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Changes in left PA blood flow (A) and PVR (B) of the left lung during ventilation of fetal lungs with 10%, 21%, and 100% O2. Data are means ± SD for 10 animals in each group. Left PA blood flow increased and PVR decreased during ventilation of the fetal lungs at all three FIO2. DNP and AMA attenuated the increase in PA blood flow and the decrease in PVR during ventilation with 21% and 100% O2 but not with 10% O2. *P < 0.05 vs. baseline; #P < 0.05 vs. the control group.

View larger version (27K): [in this window] [in a new window]   Fig. 4.   Changes in PA pressure (A) and systemic vascular pressure (B) during ventilation of fetal lungs with 10%, 21%, and 100% O2. Data are means ± SD for 10 animals in each group. The PA pressure decreased in the control group during ventilation with 21% and 100% O2. DNP and AMA inhibited the decrease in PA pressure during ventilation. Systemic pressure did not change during ventilation of the lungs. *P < 0.05 vs. baseline; #P < 0.05 vs. the control group.

Response to shear stress. Rapid constriction of the ductus arteriosus resulted in significant increases in pulmonary artery pressure and pulmonary blood flow and a decrease in PVR (Fig. 5). Constriction of the ductus was not associated with significant changes in systemic arterial pressure or Pa_O2 (data not shown). The plasma ATP levels did not change significantly during the increase in pulmonary blood flow caused by ductal constriction (2.4 ± 0.8 µM at baseline and 3.1 ± 0.8 µM at 30 min after ductal constriction). Pretreatment with DNP did not attenuate the response to shear stress, whereas antimycin-A caused partial attenuation of the increase in pulmonary blood flow and the decrease in PVR caused by shear stress (Fig. 5).

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Response of left PA pressure (A) and left PA flow (B) to partial compression of the ductus arteriosus in the fetus. Data are means ± SD for 5 animals in each group. The left PA pressure and blood flow increased significantly with ductal compression in all three groups. The increase in PA blood flow 30 min after ductal compression was attenuated in the AMA-treated group but not in DNP-treated group. #P < 0.05 vs. the control group.

Response to acetylcholine and SNAP. Acetylcholine caused a significant increase in pulmonary blood flow and a decrease in PVR at the 2 and 5 µg/min infusion rates (Fig. 6). The vasodilator response to acetylcholine was not altered by pretreatment with DNP or antimycin-A. SNAP also caused a significant increase in pulmonary blood flow and a decrease in PVR at the 0.2 and 0.5 µg/min infusion rates. DNP did not alter the response to SNAP, whereas antimycin-A attenuated the response to the 0.5 µg/min dose of SNAP (Fig. 6).

View larger version (21K): [in this window] [in a new window]   Fig. 6.   Response of left PA blood flow to infusions of acetylcholine (A) and the nitric oxide donor S-nitroso-N-acetyl-penicillamine (SNAP). Data are means ± SD for 5 animals in each group. Left PA blood flow increased during infusion of acetylcholine and SNAP in all three groups. AMA but not DNP attenuated the increase in PA blood flow at the highest dose of SNAP. *P < 0.05 vs. baseline; #P < 0.05 vs. the control group.

Response to ATP infusion. Infusion of ATP caused a significant increase in pulmonary blood flow and significant decreases in pulmonary artery pressure and PVR at the doses of 0.5-10 µM (Fig. 7). Pretreatment of fetal lambs with DNP and antimycin-A did not alter this response to ATP infusion (Fig. 7).

View larger version (19K): [in this window] [in a new window]   Fig. 7.   Effects of DNP and AMA on the pulmonary vascular response to ATP infusion. Data are means ± SD for 4 animals. Left PA blood flow increased (B) and PA pressure decreased (A) during ATP infusion. The response to ATP was not altered by DNP or AMA. *Significant change from baseline.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Our study demonstrated that oxidative phosphorylation and ATP release contribute to oxygen-induced pulmonary vasodilation in fetal lambs. The plasma ATP levels increased significantly when the fetal Pa_O2 was increased by oxygenation alone or accompanied by lung distension. The plasma ATP levels did not change during pulmonary vasodilation induced by lung distension alone in the absence of changes in Pa_O2. Inhibition of oxidative phosphorylation and ATP release during oxygen exposure attenuated the pulmonary vasodilation in response to oxygen.

The mechanism by which fetal pulmonary vessels dilate during the first breath is not clearly known. Several birth-related stimuli play an integral part in the pulmonary vasodilation occurring at birth. These include lung distension (17), removal of fetal lung liquid (31), establishment of an air-liquid interface (6), hemodynamic shear stress (1), and exposure to higher oxygen tension (3, 20). Oxygen appears to be the most important of these physiological stimuli in causing perinatal pulmonary vasodilation (20, 30). An increase in fetal Pa_O2 to postnatal levels, even in the absence of lung distension, causes pulmonary vasodilation of the magnitude seen at birth (20). A number of previous studies have indicated that release of NO and vasodilator prostaglandins in response to these birth-related stimuli play a critical role in this adaptation (2, 17, 29). However, the factors that mediate the effects of oxygen and lung distention on endothelial release of NO and prostaglandins are not known. Because the fetus is making a transition from a low to high oxygen environment at birth, we proposed the hypothesis that oxygen causes pulmonary vasodilation by increasing the rate of oxidative phosphorylation and synthesis and release of ATP. Our previous studies (11, 14) have shown that exposure of the fetus to oxygen, alone or in combination with lung distension, increases the plasma ATP levels in the fetal pulmonary artery. Infusion of exogenous ATP into the fetal pulmonary artery increases the pulmonary blood flow to postnatal levels (15). We have also demonstrated that inhibition of the purine receptors that mediate the effects of extracellular ATP and adenosine causes inhibition of birth-related pulmonary vasodilation (14). The current study provides evidence that inhibition of oxidative phosphorylation and ATP synthesis also attenuates the oxygen-induced pulmonary vasodilation in the fetus.

The plasma ATP levels demonstrated a time-dependent increase during exposure of the fetus to oxygen in the absence of ventilation. The increase in ATP levels was maximal at 30 min and coincided with the maximum increase in the pulmonary blood flow. Although the fetal Pa_O2 increased on average by only 5 Torr during exposure of the ewe to oxygen, the fetal arterial saturation increased by nearly 20%, indicating a significant increase in the availability of oxygen for ATP synthesis. However, further increases in Pa_O2 during ventilation of fetal lungs with 21% and 100% O₂ did not induce a larger increase in plasma ATP levels. These data suggest that the ATP release plateaus during oxygen exposure, with potential involvement of other mechanisms in this pulmonary vasodilation. The attenuation of pulmonary vasodilation by DNP and antimycin-A was greater during exposure of unventilated lungs to oxygen compared with the inhibition observed during ventilation with 21% and 100% O₂. These data also support a redundancy in the mechanisms that contribute to vasodilation during the complex interaction of birth-related stimuli with the fetal pulmonary circulation.

Our data are consistent with two previous studies that evaluated the influence of oxygenation and ventilation on plasma adenosine levels in fetal lambs (19, 26). These studies demonstrated that basal plasma adenosine levels in the fetus are higher than in adults (19) and that they decrease significantly during oxygenation of the fetus (26). Because high adenosine levels represent a low oxidative phosphorylation state, a fall in adenosine levels during oxygenation suggests an increase in oxidative phosphorylation, as demonstrated in our study.

The source of the increased ATP levels in plasma may be fetal red blood cells (RBC) or vascular endothelial cells. Our previous in vitro study (14) has indicated that fetal RBC are capable of increasing ATP production with an increased availability of oxygen. Fetal RBC are known to contain enzymes of oxidative phosphorylation, unlike adult RBC (5). Our previous study (14) indicated that both whole blood and plasma ATP levels increase during oxygen exposure in the fetus, indicating that fetal RBC may be a source of increased plasma ATP, because RBC are the most abundant cell type in whole blood. ATP release from erythrocytes appears to play a significant role in the regulation of vascular tone in the systemic and pulmonary vascular beds of adult animals (9, 27, 28). Mature RBC release ATP in response to a decrease in oxygen tension in the systemic circulation, facilitating vasodilation and increased supply of oxygen to tissue (9). Our in vitro studies with fetal RBC have shown that an increase in oxygen tension is associated with an increased release of ATP from fetal RBC. The difference in the effect of oxygen on ATP release between fetal and mature RBC may be due to a loss of enzymes of oxidative phosphorylation in mature RBC (5). Thus fetal RBC appear to be uniquely suited to facilitate the birth-related transition in the pulmonary circulation. ATP derived from deformation of RBC during their passage through the pulmonary microcirculation plays a significant role in the regulation of NO release and pulmonary vasodilation in adult rabbit lungs (27, 28). Ralevic et al. (24) have previously demonstrated that vascular endothelial cells also release ATP when they are exposed to shear stress. Flow-induced release of ATP occurs in the rat pulmonary circulation perfused with blood-free Krebs solution, suggesting that endothelial cells are capable of ATP release in response to shear stress (8). Inhibition of receptors for ATP attenuated the pulmonary vasodilation that occurred in response to stepwise increases in flow (8). Thus ATP release may be an important signal for the initiation of pulmonary vasodilation in response to diverse physiological signals.

ATP causes pulmonary vasodilation in the fetus both by activation of G protein-coupled P2Y receptors on the endothelial cells and by degradation of ATP to its vasoactive metabolites, ADP, AMP, and adenosine. The metabolites of ATP also cause significant pulmonary vasodilation in the fetal lamb (12). Our previous studies have suggested that the receptors activated by these nucleotides and the nucleoside adensoine are of P2Y2 and A_2A subtypes (10, 12). The NO synthase antagonist N-nitro-L-arginine attenuates the vasodilation caused by ATP and its metabolites in the fetal pulmonary circulation (13). ATP stimulates the release of NO and vasodilator prostaglandins from cultured human vascular endothelial cells (4, 23). Therefore, ATP may mediate, in part, the endothelium-dependent vasodilation caused by oxygen in the fetal pulmonary circulation.

A limitation of our study is that the inhibitors of oxidative phosphorylation administered to animals with an intact circulation in our study will be distributed to other tissues besides RBC and vascular endothelial cells. These drugs may, therefore, alter the reactivity of the pulmonary circulation independent of their effects on the release of ATP into plasma. We attempted to address this potential limitation by using doses of DNP and antimycin-A that were lower than those previously used in other studies (18, 25). We found that the doses of DNP used in our study inhibited the ATP release during oxygen exposure without altering the vasodilator response to other physiological stimuli such as lung distension and shear stress. We also did not find a significant alteration in the response to NO-dependent and -independent vasodilators by DNP at the doses used in our study. We found that antimycin-A caused some attenuation of the response to shear stress and SNAP. Because NO plays a significant role in shear-induced vasodilation and because SNAP is a NO donor, it is possible that antimycin-A has nonspecific effects on NO-mediated vasodilation in fetal lambs. The strength of the intact animal model used in our study is that it allowed us to investigate the role of oxidative phosphorylation in the context of complex interactions that occur in the lung between physical and biochemical events during the first breath. The birth-related transition in the pulmonary circulation is a net result of these unique interactions and may not be duplicated in the in vitro systems (30).

In conclusion, our study provides evidence for a role of increased oxidative phosphorylation and release of ATP in mediating the pulmonary vasodilation that occurs in the fetus in response to oxygen exposure. The signaling mechanisms involved in activation of purine receptors by ATP and the release of vasoactive mediators from the endothelium remain to be determined.