Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Several lines of evidence implicate endothelin-1 (ET-1), acting via the ET_A receptor (ET_AR) subtype, in the pathogenesis of pulmonary hypertension (Chen and Oparil, 2000). ET-1 constricts pulmonary resistance arteries and its action accords, for potency and pattern, with a mediator role in hypoxic vasoconstriction (Wang et al., 1995). Synthesis of the peptide and expression of ET_AR also increase in response to hypertensive stimuli (Ivy et al., 1998; Black et al., 2000; Chen and Oparil, 2000). ET-1 antagonists, on the other hand, may reverse pulmonary hypertension of various etiologies in both newborn and adult animals (Di Carlo et al., 1995; Miyahara et al., 1999; Jasmin et al., 2001). In the neonate in particular, ET-1 antagonists are effective against the hypertension resulting from hypoxia (lamb and piglet) (Wang et al., 1995; Perrault et al., 2001), experimental diaphragmatic hernia (lamb) (Thébaud et al., 2000), and increased blood flow (lamb) (Petrossian et al., 1999) and against the rebound hypertension consequent to abrupt discontinuation of NO inhalation (lamb) (McMullan et al., 2001).

A host of ET-1 antagonists have been developed lately differing in specificity for the ET-1 receptor subtypes, physicochemical properties, oral bioavailability, and rate of disposal by the body (Warner et al., 1994; Cheng et al., 1997). Among them, certain butenolide and benzothiazine analogs are of particular interest for high specificity against ET_AR and effectiveness by both oral and parenteral routes (Doherty et al., 1995; Patt et al., 1997; Repine et al., 1998).

The present study was undertaken in the fetal and newborn lamb with the 2-fold objective of verifying and comparing the efficacy against hypoxic pulmonary vasoconstriction of two representative ET-1 antagonists, belonging to the butenolide (PD 156707) and benzothiazine (PD 180988) families. These antagonists share selectivity for ET_AR but differ in half-life within the host (Doherty et al., 1995; Patt et al., 1997; Repine et al., 1998). One of them, PD 156707, has already shown pulmonary antihypertensive properties in the adult rat (Haleen et al., 1998). Our ultimate aim, however, was to provide the experimental framework for a possible use of these, or allied compounds sharing the same versatility for pharmaceutical formulation, in the management of pulmonary hypertension in infants. An ET_AR rather than a dual ET_AR/ET_BR antagonist was chosen for investigation since animal data, obtained by us (Wang et al., 1995) and others (Ivy et al., 2000, 2001), assign to the ET_BR population a protective function against pulmonary hypertension. The choice of the sheep was motivated by the fact that in this species, as in humans, ET-1 exerts its pulmonary constrictor effect primarily, if not exclusively, via ET_AR (Fukuroda et al., 1994; Russell and Davenport, 1995; Wang et al., 1995). In addition, the same species lends itself well to an experimental approach combining suitable in vivo and in vitro preparations (Wang et al., 1995). A preliminary account of the results with PD 180988 has been reported (Coe et al., 2000).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

In vivo work was carried out in newborn lambs of Suffolk/Rambouillet/Dorset crossbreed. Their average age at surgery was 4 days (range, 1-13 days), whereas the actual study was performed at 10-43 days of age (average, 19 days). Fetal lambs at term (gestation age, 136-139 days, term 145 days) of Southdown or Southdown/Dorset stock were used for in vitro work. Surgical procedures and experimental protocols were approved by the Animal Care Committee of our institutions.

Solutions and Drugs. Krebs medium for the isolated pulmonary resistance arteries had the following composition: 118 mM NaCl, 4.7 mM KCl, 1 mM KH₂PO₄, 0.9 mM MgSO₄, 2.5 mM CaCl₂, 11.1 mM glucose, and 25 mM NaHCO₃. The solution was bubbled with gas mixtures containing either no O₂ or 12.5% O₂ plus 5% CO₂ in N₂, and the resulting partial pressure of O₂ (pO₂) was, respectively, 7 ± 2 and 56 ± 1 mm Hg (pH 7.4). As shown previously, the O₂-containing mixture duplicated the neonatal condition, whereas the zero-O₂ mixture ensured a full-fledged hypoxic contraction (Wang et al., 1995). pO₂ was measured with a gas analyzer (model 1610; Instrumentation Laboratory, Lexington, MA). Sterile Hanks' balanced salt solution (without CaCl₂, MgCl₂, MgSO₄, and phenol red) was obtained from ICN Biomedicals Inc. (Costa Mesa, CA) and was supplemented with antibiotic-antimycotic solution (100 units/ml penicillin G, 100 µg/ml streptomycin, and 0.25 µg/ml fungizone) (PSF from ICN Biomedicals Inc.) and a serine protease inhibitor (100 µM Pefabloc; Roche Molecular Biochemicals, Indianapolis, IN).

Intact Animal. Newborn lambs (n = 30), with an average weight of 4.1 kg (range, 2.7-7.1), were chronically instrumented according to a published protocol (Wang et al., 1995). In brief, the animals were anesthetized and, through a left thoracotomy, had a flow probe (C and C Instruments, Culver City, CA or Transonic Systems Inc., Ithaca, NY) placed around the main pulmonary artery (PA). Catheters were positioned in both the left atrium and the right lower lobe pulmonary vein (PV) and, concomitantly, the constricted ductus arteriosus was ligated. Four to 21 days after instrumentation (average, 8 days), at the age of 7 to 28 days (average, 12 days) and under anesthesia, indwelling vascular sheaths were placed in both carotid arteries and external jugular veins. Afterward, the lambs were allowed to recover for 1 to 29 days (average, 10 days) before being tested with either ET_AR antagonist. On the study day, using the vascular sheaths as a guide, the animals were fitted with additional catheters in the PA, the aortic root (Ao), and the right atrium (RA) (two or three distinct lines). Separate RA lines served for pressure (P) monitoring, for the delivery of the ET_AR antagonist, and when required, for administration of ONO-11113. A high-fidelity catheter (Millar Instruments, Houston, TX) was also positioned in the left ventricle for measuring dP/dt. Pressure lines were connected to strain gauge transducers (model 23ID; Gould Instrument Systems Inc., Cleveland, OH), the flow probe to its own flowmeter (Gould model SP2202 or Transonic model T201D), and the Millar catheter to a Millar control unit (model TC-510). All hemodynamic signals including heart rate were displayed on a Gould recorder (model ES2000) and were also stored digitally on a microcomputer (CVSOFT program; Odessa Computer Co., Calgary, AB, Canada) for offline analysis. Cardiac output, being derived from PA flow, was normalized for body weight (cardiac index). Pulmonary vascular resistance (PAR) was calculated by dividing the difference between mean PAP and mean left atrium pressure by cardiac index (expressed as mm Hg per milliter per minute per kilogram, i.e., units per kilogram). Systemic vascular resistance (AoR) was calculated by dividing the difference between mean AoP and mean RA pressure by cardiac index (expressed as units per kilogram). Arterial blood gases were measured periodically with a Nova Stat Profile 5 blood gas analyzer (Nova Biomedical Corp., Waltham, MA), and vascular lines were flushed intermittently with heparinized saline (4 units/ml). Blood samples were withdrawn from the PV for the assay of ET_AR antagonists and ET-1. Studies were performed in the conscious lamb breathing spontaneously either room air (normoxia) or a mixture of approximately 12% O₂ and 5% CO₂ in N₂ to elicit a steady-state hypoxic response (normocapnic hypoxia) over a 60-min period.

Assay of ET_AR Antagonists and ET-1. When studying the normoxic animal, PV blood samples were collected before ET_AR antagonist administration and at the end of the treatment (protocol 1). In hypoxia experiments (protocol 2), sampling was carried out first in normoxia and then during sustained hypoxia at the beginning and the end of the 60-min period (hypoxia control) or just before ET_AR antagonist administration ("zero time") and at the end of the 60-min treatment (hypoxia treatment). With rebound evaluation experiments (protocol 3), blood samples were obtained in normoxia, during sustained hypoxia before and at the end of the 30-min treatment with the ET_AR antagonist, and again after a further 30-min hypoxia in the absence of treatment. In those instances in which pulmonary vascular tone was raised with ONO-11113 (protocol 4), blood was withdrawn first during the normoxic control, then after 30 min of ONO-11113 infusion, and last after a further 30-min period in which ONO-11113 infusion and ET_AR antagonist treatment were combined. In all cases, arterial blood (2 ml) was collected in heparinized tubes and centrifuged immediately in the cold to separate the plasma fraction. Plasma was then stored at 80°C for ET_AR antagonist assay. ET-1 was measured in the same samples in most experiments.

Isolated Resistance Arteries. Rings of pulmonary resistance arteries (sixth generation) were obtained from fetal lambs at term (n = 11) and were prepared for mechanical recording as previously reported (Wang and Coceani, 1992). In brief, the vessel was suspended between two 25-µm tungsten wires inside a jacketed organ bath (capacity, 6.5 ml), and the solution was gassed with the O₂ mixture (i.e., 12.5%) mimicking the neonatal condition. One of the wires was connected to an isometric force transducer (model DSC-6BE4-110; Kistler Morse, Redmond, WA). Once mounted, the vessel was stretched to about 30% of the expected transmural pressure in vivo and was equilibrated for at least 60 min. Afterward, this load was removed, the dimensions at rest were measured (internal diameter, 172 ± 4 µm; length, 735 ± 38 µm; wall thickness, 33 ± 1 µm), and the vessel was stretched again in stepwise fashion until a functionally relevant tension, as predicted by the La Place equation, was achieved. The preparation was equilibrated for 30 min more, or until a stable baseline had been reached, before starting the actual experiment.

Endothelin Receptor Radioligand Binding Assay. The main pulmonary artery was collected from term fetal lambs (n = 25) and was freed of loose connective tissue and fat. It was rinsed with sterile Hanks' balanced salt solution and was then transferred to a sterile plastic tube where it was frozen in liquid nitrogen for storage (at 80°C). When required, individual specimens were pulverized in a freezer/mill (model 6700; Spex Industries, Edison, NJ), and the powdered tissue was resuspended in 10 ml of buffer containing 20 mM Tris, 2 mM EDTA, and 0.1% BSA (pH 7.4). The suspension was centrifuged at 500g (5 min, 4°C), and the supernatant was filtered through cheesecloth. The filtrate was centrifuged again at 30,000g (20 min, 4°C), and the pellet was resuspended in buffer containing protease inhibitors (200 µM Pefabloc, 10 µM phosphoramidon, 10 µM leupeptin, and 1 µM pepstatin). The resulting tissue fraction (about 3 mg of protein/ml) was divided in 1-ml aliquots and frozen at 80°C for further work up.

Analysis of Data. Data are expressed as the mean ± S.E.M., where n is the number of experiments. Statistical comparison of two means was done using Student's t test for paired or unpaired observations. Multiple comparisons were made with an analysis of variance followed by Duncan's multiple range test. Binding data were analyzed using Kaleidagraph (Synergy Software, Reading, PA), and IC₅₀ values were calculated using a one-site fit. Differences are considered significant when p < 0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Intact Animal. Chronically instrumented lambs were conscious at the time of the study and showed no sign of discomfort. In the absence of any treatment, blood gas (pO₂ 73 ± 3 mm Hg; pCO₂ 34 ± 3 mm Hg; pH 7.50 ± 0.01) and hemodynamic (Table 1) variables were within normal limits and remained stable throughout the period of observation. When breathing the low-O₂ gas mixture, the animals exhibited the expected fall in arterial pO₂ but no alterations in the acid/base status (pO₂ 44 ± 1 mm Hg; p < 0.01 versus control; pCO₂ 34 ± 2 mm Hg; pH 7.49 ± 0.01). Coincidentally, PAP and PAR rose and remained elevated for as long as hypoxia was maintained (Fig. 1A). Little or no change was noted instead in the other hemodynamic variables, specifically those pertaining to systemic vascular function (Fig. 1A; Table 1). Elevations in PAP and PAR, comparable in magnitude to those resulting from exposure to alveolar hypoxia, were also observed in the normoxic animal during intra-RA infusion of ONO-11113 (Fig. 2). Again, this constrictor response, as for that elicited by hypoxia, appeared to be confined to the pulmonary vascular district and was sustained.

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Hemodynamic response to alveolar hypoxia in the control and PD 156707-treated conscious lamb. Hypoxia was started at approximately 10 min (dashed vertical line) and produced a peak effect by time 0, which, in the absence of treatment, was maintained for as long as the animal was made hypoxic (i.e., up to 60 min). A, control; B, treatment. In both panels, values at 15 min refer to the normoxia control, whereas values at time 0 (hypoxia control) are obtained immediately before administering PD 156707.

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Effect of PD 156707 on pulmonary hypertension secondary to ONO-11113 treatment.

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Persistence of ETAR antagonist effect after discontinuation of treatment in the course of hypoxia. Hypoxia was started at approximately 10 min (dashed vertical line) and produced a peak effect by time 0, which, in the absence of treatment, would be maintained for as long as the animal was made hypoxic (see Fig. 1). A, PD 156707; B, PD 180988. In both panels, values at 15 min refer to the normoxia control, whereas values at time 0 (hypoxia control) are obtained immediately before administering the ETAR antagonist.

View larger version (23K): [in this window] [in a new window]   Fig. 4.   Effect of PD 180988 on the pulmonary constrictor response to hypoxia. Hypoxia was started at approximately 10 min (dashed vertical line) and produced a peak effect by time 0 which, in the absence of treatment, would be maintained for as long as the animal was made hypoxic (see Fig. 1). A, PD180988, low dose; B, PD 180988, high dose (for details, see text). In both panels, values at 15 min refer to the normoxia control, whereas values at time 0 (hypoxia control) are obtained immediately before administering PD 180988.

Blood ET-1. ET-1 levels were consistent among animals and increased when lowering blood pO₂ from normoxic to hypoxic values (Table 2). The latter effect developed rapidly, as one may infer from the presence of an upward trend as soon as pulmonary hypertension had reached the steady state (3.07 ± 0.5 pg/ml at zero time, n = 4). Conversely, no such increase was seen with the pulmonary hypertension secondary to ONO-11113 administration, and ET-1 values in this case (2.53 ± 0.33 pg/ml, n = 3) equaled those of the normotensive animal (Table 2). ET_AR antagonists had no effect by themselves on ET-1, and regardless of the test condition, blood concentrations for the treatment groups overlapped with those of controls (Table 2).

Blood PD 156707 and PD 180988. ET_AR antagonists were measurable in blood from all treated animals. Their levels, under any condition, were about 10-fold higher with PD 180988 than PD 156707 and were also more variable (Table 3). In addition, the values for PD 180988 showed no clear difference depending on the dosage, low versus high, used in the experiment. With either compound, however, blood levels attained after the 60-min treatment period did not vary significantly between normoxia and hypoxia (Table 3). Furthermore PD 156707 and PD 180988 behaved similarly in being more abundant at the 30- than the 60-min interval through the treatment in hypoxia (i.e., protocol 3, see Materials and Methods) (PD 156707, 158 ± 28 ng/ml, n = 3; PD 180988, 3097 ± 1715 ng/ml, n = 3). The latter finding implies that the initial bolus of the drug contributes in large measure to the observed values. Consistent with this conclusion is also the fact that concentrations being detected in blood 30 min after the cessation of a 30-min treatment (PD 156707, 35 ± 14 ng/ml, n = 4; PD 180988, 539 ± 396 ng/ml, n = 3) do not depart significantly from those found at the end of a 60-min period of uninterrupted treatment (see Table 3).

Isolated Resistance Arteries. Pulmonary resistance arteries contracted when the pO₂ of the medium was lowered from neonatal to hypoxic values (Fig. 5). As expected from vessels with an intact endothelium (see Wang et al., 1995), this contraction started after some delay (average, 14 min, range, 4-21) and progressed slowly to a steady plateau (peak in 22 min, range, 18-26). In contrast, the contraction of the normoxic vessel to ONO-11113 was immediate in onset and development (peak in 15 min from application, range 13-19), and it also attained higher magnitude (Fig. 5). When tested on the sustained response to hypoxia, PD 156707 reversed gradually the contractile tone (average, 42 min for maximal effect) in all, but one, experiment (Fig. 5). This exceptional preparation, however, was characterized by an unusually strong hypoxic contraction (1.18 mN/mm before and during treatment) and was not included in the computation. No such effect of PD 156707 was seen when the tone of the vessel was raised with ONO-11113 (Fig. 5). In fact, there was a modest, albeit significant, further elevation in tone due to the antagonist (Fig. 5).

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Contractile response of isolated pulmonary resistance arteries to hypoxia and 100 nM ONO-11113 before and during treatment with 0.1 µM PD 156707. In both cases, PD 156707 was tested on the sustained contraction. Note that the hypoxia group does not include one experiment in which the antagonist was without effect (see text). However, even after including this exceptional result in the computation the difference between control and treatment values remains significant (p < 0.01). , p < 0.05; , p < 0.01 versus control.

Endothelin Receptor Radioligand Binding Study. The relative specific binding of 30 pM of [¹²⁵I]ET-1 (ET_AR + ET_BR binding) and [¹²⁵I]ET-3 (ET_BR binding) to isolated fetal lamb PA tissue (radiolabeled ET-3/radiolabeled ET-1 binding) demonstrated a small proportion of ET_B (6.5%) versus ET_A (93.5%) receptor sites. Accordingly, in competition binding experiments, ET-1 inhibited [¹²⁵I]ET-1 binding to isolated fetal lamb PA tissue with an IC₅₀ of 0.24 nM, and this value was comparable to that of 0.3 nM for PD 156707 and 0.89 nM for PD 180988.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

This study validates the concept that ET-1, acting via the ET_AR subtype, is a critical mediator for the contraction elicited by hypoxia in the pulmonary vasculature of the perinatal animal. By extension, it provides the experimental foundation for the use of ET_AR antagonists as pulmonary antihypertensive agents in infants. Several findings support our conclusion: 1) blood levels of ET-1 are increased during the hypoxia- but not the ONO-11113-triggered pulmonary hypertension, 2) ET_AR antagonists of diverse chemical structure reverse completely, or nearly completely, the hypoxic pulmonary response both in vitro and in vivo, and 3) the same compounds have, in contrast, no significant relaxant effect on the normoxic pulmonary vasculature whether in its resting state or during the contraction by ONO-11113. Furthermore, effective doses of the ET_AR antagonists accorded with individual half-life values in the body, with the shorter acting among the compounds tested (PD 156707) requiring a higher dosage, and attendant blood levels were also within the expected range for an antihypertensive action (about 90 and 300 ng/ml for PD 156707 and PD 180988, respectively; S. J. Haleen and K. M. Welch, unpublished data). Normalization of pulmonary hemodynamics obtained with the ET_AR antagonists during hypoxia implies the absence in the vasculature of an ET_BR subtype mediating contraction. Our earlier work on isolated pulmonary resistance arteries, i.e., the putative prime target for the hypoxic stimulus, failed to show any ET_BR-based response (Wang et al., 1995). This agrees with the results of the radioligand binding assay being reported here. An ET_B set of receptors, on the other hand, was found in the small veins, but its activation resulted in relaxation (Wang et al., 1995). Hence, we may conclude that there is a dichotomy in the pulmonary vascular ETR system of the lamb whereby the two receptor subtypes subserve generally different functions (i.e., ET_AR, contraction; ET_BR, relaxation). A similar arrangement has been reported in humans (Fukuroda et al., 1994; Russell and Davenport, 1995), and this makes the lamb a better experimental model, compared with other species such as the rabbit (Fukuroda et al., 1994), for testing ET_AR antagonists with potential therapeutic use as pulmonary antihypertensive agents.

Translated to the clinical situation, our findings support the feasibility of introducing an ET_AR antagonist for the management of infants with pulmonary hypertension. The two compounds, chosen for our study, provide a good lead insofar as they pave the way to different, oral versus parenteral, pharmaceutical formulations and to short- versus long-acting therapeutic interventions. Several advantages, both conceptual and practical, may derive from adopting a mechanism-based treatment over the current, palliative use of vasodilators such as nitric oxide and prostacyclin (Barst et al., 1996, 1999; Clark et al., 2000). There is lesser likelihood, once the dose of the ET_AR antagonist has been matched to the patient, of systemic vasodilatation complicating the treatment course. This is particularly true when considering parenteral prostacyclin as the alternative. Within the lungs, this new therapy would exert its effect on the constricted vasculature with the attendant lesser risk, on one hand, of blood being shunted locally and, on the other hand, of hypertension reoccurring as soon as administration of the drug is discontinued. In fact, rebound hypertension is a potentially troublesome complication of NO inhalation (Cueto et al., 1997), which we expect not to take place, at least in the same abrupt fashion, with any orally or parenterally administered ET_AR antagonist. The latter assumption is based not only on the observation made here that the antihypertensive effect of the antagonists persists beyond the treatment period but also on the notion that a mechanism-based intervention may reverse any hypertension-induced down-regulation of NO- and prostaglandin I₂-linked pulmonary vasodilator systems (Shaul et al., 1997; Tuder et al., 1999). Hence, even in the case of a transient withdrawal from drug action, the pulmonary vasculature may be able to counter more effectively any drive toward excess contraction. Last, ET_AR antagonists would prove particularly useful with any pathological condition (e.g., congenital diaphragmatic hernia) where the pulmonary hypertension fails to respond to NO (Neonatal Inhaled Nitric Oxide Study Group, 1997).

Against all of these positive features, there are possible drawbacks. As mentioned above, ET_AR antagonists may not remain selective in their action on the hypertensive pulmonary vascular district, and a significant systemic hypotension has complicated studies in the animal (Petrossian et al., 1999; McMullan et al., 2001) as well as the clinical trial of a dual ET_AR/ET_BR antagonist (i.e., bosentan) in the adult (Williamson et al., 2000). However, judging from our present work and the outcome of a recent clinical study (Channick et al., 2001), any such complication should be avoidable by adjusting the dosage. Another difficulty may derive from the involvement of alternative mechanisms, such as the interference with calcium-sensitive potassium channels, in the pathogenesis of pulmonary hypertension (Cornfield et al., 2000). Their contribution, should it become significant, could lessen the impact of any ET_AR antagonist-based therapy. ET-1, on the other hand, may modulate central neural pathways involved in respiratory control (Kuwaki et al., 1996) and the response to stress (Kurihara et al., 2000), and this raises the question of an adverse effect of the ET_AR antagonist on brain. However, apart from the sedation, we could not detect any central action at doses effective on the pulmonary circulation. Significant in this context is also the fact that these compounds cross sparingly the blood-brain barrier, as one can infer from whole body autoradiography and the direct assay of cerebrospinal fluid (S. J. Haleen and K. M. Welch, unpublished data). A final, possible complication, also originating from experimental data (Coceani et al., 1999; Takizawa et al., 2000) is that the ET_AR antagonist could reopen the ductus arteriosus in the young infant. Such an event, however, seems remote since closure of the vessel becomes irreversible beyond the immediate neonatal period.

In conclusion, then, it is safe to assume that the anticipated benefits in using the ET_AR antagonists for neonatal pulmonary hypertension outweigh any potential drawback. The new classes of antagonists, studied by us, lend themselves well to this particular application, because they combine the required efficacy with versatility in both the route of administration and the time course of action. Conceivably, our data have implications not only for the management of the sick infant, but also for patients suffering from pulmonary hypertension in adult age.