The role of endothelin-A receptors in cardiovascular responses to acute hypoxaemia in the late gestation sheep fetus

Abstract

1. In unanaesthetized chronically instumented fetal sheep (118-121 days gestation) we investigated the effect of acute isocapnic hypoxaemia (arterial P_o2, 12.5 ± 0.6 mmHg) on heart rate (FHR), mean systemic arterial blood pressure (MABP), carotid and femoral blood flows (CBF and FBF, respectively), and carotid and femoral vascular resistances (CVR and FVR, respectively) with the infusion of either the endothelin-A (ETA) receptor antagonist FR139317, or saline vehicle.

2. During normoxaemia FHR (P < 0.05) and CBF (P < 0.01) were greater, and CVR (P < 0.01) was lower with FR139317 than with vehicle infusion. CVR remained lower with FR139317 than with vehicle infusion during hypoxaemia (P < 0.01) and recovery (P < 0.05). During hypoxaemia the rapid initial bradycardia, the increase in MABP and FVR and the decrease in FBF were similar with vehicle and FR139317 infusion. In both groups plasma endothelin-1 concentration ([ET-1]) was unaltered by hypoxaemia. The increase in CBF during hypoxaemia with vehicle (P < 0.01) was absent with FR139317 infusion.

3. Thus in the late gestation ovine fetus endogenous ET-1 modulates basal FHR, CBF and CVR via ETA receptors. Modulation of CBF and CVR persists during hypoxaemia but ETA receptors do not appear to contribute to the decrease in femoral blood flow measured during acute hypoxaemia.

Under conditions of reduced oxygen supply it is vital that the fetus mounts appropriate cardiovascular responses for its survival, growth and development. These responses occur in two phases. First, there is a rapid decrease in fetal heart rate (FHR) and increase in peripheral (e.g. femoral) vascular resistance which are predominantly carotid chemoreflexes mediated via vagal and α-adrenergic efferents, respectively (Giussani, Spencer & Hanson, 1994). The second phase is slower and is characterized by maintained femoral vasoconstriction and vasodilatation of carotid, cerebral, myocardial and adrenal vascular beds. Carotid chemoreceptors do not appear to mediate the femoral or carotid vascular responses (FVR and CVR, respectively). The vasoconstriction is likely to be due to an increase in plasma catecholamines, arginine vasopressin (AVP) and angiotensin II (AII) (Giussani et al. 1994).

From adult studies a concept has emerged that no single mechanism is responsible for cardiovascular control at rest (Paller & Linas, 1984) or during stress such as haemorrhage (Scroop, Stankewytsch-Janusch & Marker, 1992). In the fetus reflex mechanisms initiated during hypoxaemia appear to operate against a background of other hormonal mechanisms, for example while plasma AII concentration increases during hypoxaemia, it is only once the carotid chemoreceptor input has been removed that a role for AII in the peripheral vascular responses to hypoxaemia becomes apparent (Green, McGarrigle, Bennet & Hanson, 1994). Some of this modulation may occur at the vascular endothelium which acts as a hypoxaemic sensor and effector system, since endothelial cells produce a number of vasoactive autocoids, e.g. endothelin-1 (ET-1), nitric oxide (NO) and prostacyclin (Vane, Anggard & Botting, 1990).

Endothelin-1 is released from the abluminal surface of cultured endothelial cells (Wagner et al. 1992), although there is also evidence that smooth muscle cells produce it (Hahn, Resnik, Scott-Burden, Powell, Dohi & Buhler, 1990). These mechanisms of release support a paracrine and/or autocrine rather than endocrine action of ET-1. Endothelin-1 is released during hypoxaemia from adult rat resistance vessels in vitro (Rakugi et al. 1990) and elevated plasma [ET-1] is observed during hypoxaemia in adult rats (Shirakami et al. 1991; Horio et al. 1991). Endothelin-1 concentration is greater in fetal than in maternal plasma (Haegerstrand, Hemsen, Gillis, Larsson & Lundburg, 1989; Nakamura et al. 1990) which suggests that the fetus is able to produce ET-1. It is therefore conceivable that the release of fetal ET-1 during hypoxaemia may be of sufficient magnitude to influence circulating [ET-1]. Thus the first aim of this study was to measure circulating [ET-1] during normoxaemia and acute hypoxaemia.

The development of selective ET receptor blockers has provided a new approach to determining a physiological role for ET-1 in the fetus (Ihara et al. 1992). Endothelin-A (ETA) receptor activation is implicated in the regulation of basal arterial pressure and renal vascular resistance in the adult anaesthetized rat (Pollock & Opgenorth, 1993). Endothelin-1 is implicated in the maintenance of high fetal pulmonary vascular tone in vivo (Ivy, Kinsella & Abman, 1994; Wong, Fineman & Heymann, 1994) but appears to have no role in determining basal FVR (Ivy et al. 1994). Endothelin-1 contributes to hypoxaemic vasoconstriction of fetal pulmonary vessels in vitro (Wang, Coe, Toyoda & Coceani, 1995) and preliminary evidence suggests that ET-1 contributes to the increase in fetal systemic vascular resistance after 3 h of hypoxaemia (Jones, 1995). In the adult rat abluminal administration of ET-1 causes a vasoconstriction of the cerebral microvasculature (Willette & Sauermelch, 1990). The second aim of this study was to address the contribution of endogenous ET-1 to the systemic cardiovascular responses to acute isocapnic hypoxaemia using FR139317, a specific ETA receptor antagonist (Sogabe et al. 1992).

METHODS

Surgical preparation

Five pregnant ewes (Suffolk × Blue-faced Leicester) were instrumented at 118-121 days gestation (term, 147 days) under general anaesthesia (1 g thiopentone i.v. for induction followed by 2 % halothane in oxygen at 2 l min⁻¹ for maintenance). An incision was made in the mid-line of the lower abdominal wall and the uterus palpated to determine the number and position of fetuses. A fetus was partially exteriorized through an incision in the uterine wall. Heparinized catheters (i.d., 1.0; o.d., 2.0 mm translucent vinyl tubing, Portex Ltd) were placed in a fetal carotid artery, a jugular vein and a brachial vein, and in the amniotic cavity. Ultrasound flow probes (Transonic System Inc., USA) were placed around the uncatheterized carotid artery, as this provides a good estimate of cerebral blood flow in the fetus (vanBel, Roman, Klautz, Teitel & Rudolph, 1994), and a femoral artery, which gives an indication of hindlimb blood flow. Stainless-steel electrodes were sewn subcutaneously onto the chest and a hindlimb to record ECG. Catheters and electrodes were exteriorized through the maternal flank and secured to the ewe's back in a plastic bag. Following surgery, antibiotics were administered i.m. to the ewe (4 ml Streptopen (procaine penicillin G and digdrostreptomycin), Pitman-Moore Ltd, UK) and amniotic cavity (600 mg Crystapen (sodium benzyl penicillin), Britannia Pharmaceuticals Ltd, UK; 80 mg Gentamicin (gentamicin sulphate, sodium metabisulphate and disodium edetate), DBL, UK). A period of at least 5 days post-operative recovery was allowed prior to experimentation during which daily antibiotic treatment was given to the ewe (300 mg Crystapen i.v.) and fetus (150 mg Crystapen i.v.) and amniotic cavity (150 mg Crystapen). Gentamicin was administered into the amniotic cavity (40 mg) and to the ewe (40 mg, i.v.) on days 1 and 2 only. Patency of catheters was maintained by a continuous infusion of heparinized saline (50 U heparin per 1 ml of 0.9 % NaCl at 0.125 ml h⁻¹) and fetal arterial blood was collected daily for blood gas analysis.

Drugs

ET-1 (50 μg; Human, porcine; Sigma) was dissolved in 10 ml degassed saline with two drops of glacial acetic acid (Fisons, UK). Ten millilitres of 2 mg ml⁻¹ sheep albumin (Fraction V powder; Sigma) was added to give a final [ET-1] of 2.5 μg ml⁻¹. This stock solution was divided into 200 μl aliquots, each containing 500 ng ET-1, and stored at -20°C (shelf-life of ca. 3 months). On the day of study the required number of aliquots were resuspended in 2 ml saline.

On the day of experimentation FR139317 (a gift from Fujisawa Pharmaceutical Co. Ltd, Japan) was dissolved in saline and equimolar NaOH (Sigma) (5 drops per 10 ml) to give a final solution of 700 μg ml⁻¹ (Nirei, Hamada, Shoubo, Sogabe, Notsu & Ono, 1993).

Experimental procedure

A preliminary study was undertaken in two fetuses in an attempt to determine the dose of FR139317 required to antagonize the cardiovascular responses to 1.5 μg ET-1 (i.v.). In one fetus (126 days gestation) three dose regimes of FR139317 (i.v.) were investigated: dose 1, 3 μg bolus followed by 0.5 μg min⁻¹ infusion; dose 2, 30 μg bolus followed by 5 μg min⁻¹ infusion; dose 3, 300 μg bolus followed by 50 μg min⁻¹ infusion. Endothelin-1 caused a 32 % decrease in femoral blood flow (FBF), a 7 % decrease in carotid blood flow (CBF) and a 5 % increase in mean arterial blood pressure (MABP). Following the infusion of the lowest dose of FR139317 the FBF response to ET-1 was attenuated (5 % decrease), while an increase in MABP (4 %) to ET-1 became apparent. However, after doses 2 and 3 of FR139317 the administration of ET-1 caused an increase in FBF (4 and 2 %, respectively) and a small increase in MABP (1 and 2 %, respectively). Following the infusion of all three doses of FR139317, ET-1 produced an increase in CBF (15, 7 and 10 %, respectively). In a second fetus (133 days gestation) a maximal dose of FR139317 (3 mg bolus followed by 500 μg min⁻¹ infusion) caused a massive vasoconstriction and resulted in the death of the fetus approximately 30 min after the onset of infusion. This pressor effect has been observed before with high FR139317 doses in the conscious adult rat (Gardiner, Kemp, March, Bennet, Davenport & Edvinsson, 1994). Thus we used the 300 μg bolus followed by 50 μg min⁻¹ infusion regime for the subsequent hypoxaemia studies in order to maximize ETA receptor antagonism.

In five fetuses two experiments were conducted on separate days. Day 1, hypoxaemia with the infusion of vehicle (126 ± 0.9 days gestation, mean ±s.e.m.; 4 ml h⁻¹; 5 drops of 150 mm NaOH per 10 ml saline), and day 2, hypoxaemia with infusion of FR139317 (126 ± 0.5 days gestation, 300 μg in 2 ml bolus followed by 50 μg min⁻¹ infusion). In two of the five fetuses the order of experiments on days 1 and 2 was reversed. Fetal oxygenation was manipulated by placing a loosely tied polyethylene bag over the ewe's head into which a controlled mixture of air, N₂ and CO₂ was passed at 44 l min⁻¹. Thus, after the onset of vehicle or FR139317 infusion (0 min) an initial 1 h normoxaemic period was established while the ewe was breathing air only, followed by the induction of 1 h of fetal isocapnic hypoxaemia (arterial O₂ pressure (P_a,O2) to ca. 12 mmHg) by reducing maternal inspired O₂ fraction (inspirate: 14-18 l min⁻¹ air; 22 l min⁻¹ N₂; 1.2 l min⁻¹ CO₂). FR139317 infusion was stopped at ca. 125 min (recovery).

Arterial blood pressure, CBF, FBF and ECG were monitored continuously throughout the procedure using MacLab Chart software (ADInstruments Pty Ltd) on a Macintosh LCIII (Apple Computers Inc.). Samples of fetal arterial blood (2-3 ml) were collected prior to the onset of drug infusion (at 0 min) and during normoxaemia (at 15 and 45 min), hypoxaemia (at 75 and 105 min) and recovery (at 135 and 165 min). Blood was transferred immediately into chilled EDTA tubes and spun at 4°C (2000 g) for 10 min. Plasma was decanted into tubes and stored at -20°C for subsequent hormonal analysis. A further 0.6 ml blood was collected at these times and at 90 min (hypoxaemia) for the analysis of pH, blood gases (BG Electrolytes system, Instrumentation Laboratory), haemoglobin (Hb) and haematocrit (Hct; Co-oximeter, Instrumentation Laboratory. Calibrated for fetal haemoglobin).

At the end of a study ewes were killed by an overdose of barbiturate (30-40 ml i.v., 200 mg ml⁻¹ pentobarbitone sodium BP, Rhône Mérieux, UK).

Endothelin-1 radioimmunoassay

Aliquots of plasma (0.6 ml) were acidified with 1.0 ml of 4 % acetic acid. The acidified samples were then passed through C18 columns and the columns rinsed with 3 ml of 25 % ethanol. The columns were then washed with 1 ml of 4 % acetic acid in 86 % ethanol (to extract endothelin) and the eluate collected in glass tubes. The eluate was evaporated at 37°C in a stream of air. The dried extracts were reconstituted in 0.22 ml of assay buffer.

Plasma [ET-1] was measured using an ET(1-21)-specific ¹²⁵I-radioimmunoassay (Nichols Institute, Diagnostics BV, Saffron Walden, UK). Samples were assayed in a single batch. Recoveries averaged 88 %. The assay sensitivity was 3.2 pmol l⁻¹. The intra- and interassay coefficients of variation reported for the assay are 4.5 % (5.9 pmol l⁻¹) and 5.8 % (4.4 pmol l⁻¹), respectively. The assay was specific for ET(1-21) and cross-reacted with ET-1 (100 %), ET-2 (67 %) and ET-3 (84 %).

Data analysis

Values are expressed as means ±s.e.m. Blood gas and pH data during normoxaemia, hypoxaemia and recovery were reduced to summary measures (Matthews, Altman, Campbell & Royston, 1990) and Student's paired t test used to compare pre- with post-infusion values during normoxaemia, between normoxaemia and hypoxaemia and between normoxaemia and recovery.

Student's t test was used to compare plasma [ET-1] in hypoxaemia (at 75 and 105 min) and recovery (at 135 and 165 min) with normoxaemia (at 45 min). The Bonferroni method of correction was used for multiple comparisons.

Vascular resistances (in mmHg min ml⁻¹) were calculated using the formula MABP/flow. Cardiovascular data in normoxaemia, hypoxaemia and recovery hours were reduced to summary measures (Matthews et al. 1990) by calculating the mean for individual fetuses over a period of time. Summary measures were then tested using Student's paired t test. Student's paired t test was also used to test rapid transient changes (60 vs. 65 min) in FHR.

RESULTS

Blood measurements

In both vehicle and FR139317 groups fetal P_a,O2 (P < 0.01) decreased during hypoxaemia. Arterial P_CO2 (P_a,CO2) was unaltered throughout the course of the procedure in both groups, although there was a small but significant reduction in pH during hypoxaemia in the FR139317 group (P < 0.05), and during hypoxaemia (P < 0.01) and recovery (P < 0.01) in the vehicle group (Table 1).

Table 1.

Fetal arterial blood measurements in normoxaemia and hypoxaemia in vehicle- and FR139317-infused groups

	Pa,O2 (mmHg)	Pa,CO2 (mmHg)	pH	[Hb] (g dl−1)	Hct (%)
Preinfusion
Vehicle	21.8 ± 1.1	49.5 ± 1.6	7.37 ± 0.01	8.3 ± 0.3 *	28.6 ± 1.2
FR139317	21.8 ± 0.9	50.2 ± 1.4	7.37 ± 0.00	7.8 ± 0.4	26.6 ± 2.6
Normoxaemia
Vehicle	21.8 ± 1.1	48.7 ± 1.2	7.36 ± 0.01	7.8 ± 0.4	26.7 ± 1.8
FR139317	20.7 ± 0.9	48.5 ± 1.1	7.36 ± 0.00	7.5 ± 0.5	26.6 ± 2.2
Hypoxaemia
Vehicle	12.5 ± 0.6 **	47.4 ± 1.4	7.32 ± 0.01 **	6.5 ± 0.3 **	28.7 ± 1.8
FR139317	12.4 ± 0.5 ††	48.0 ± 1.2	7.33 ± 0.01 †	6.3 ± 0.3 ††	28.7 ± 2.0 ††
Recovery
Vehicle	21.1 ± 1.6	48.0 ± 1.2	7.29 ± 0.02 **	7.8 ± 0.3	26.3 ± 2.1
FR139317	20.7 ± 1.3	47.8 ± 0.7	7.32 ± 0.02	7.1 ± 0.4 ‡	25.9 ± 1.7

Values are shown as means ±s.e.m.; n= 5.

^*P < 0.05

^**P < 0.01vs. normoxaemia in vehicle group

^†P< 0.05

^††P< 0.01 vs. normoxaemia in FR139317 group

^‡P< 0.05 FR139317 vs. vehicle group.

Arterial [Hb] decreased during hypoxaemia in both vehicle and FR139317 groups. Haematocrit increased during hypoxaemia with FR139317 (P < 0.01) but not with vehicle infusion (Table 1). In the FR139317 group plasma [ET-1] was lower during recovery than normoxaemia (P < 0.01) (Fig. 1). There was no change in plasma [ET-1] with the onset of FR139317 infusion. During hypoxaemia [ET-1] remained unaltered from normoxaemic levels of ca. 13 pmol l⁻¹ in both vehicle and FR139317 groups (Fig. 1).

Figure 1. The effect of FR139317 on plasma [ET-1] during normoxaemia and hypoxaemia.

Plasma [ET-1] during normoxaemia (45 min), hypoxaemia (shaded region, 75 and 105 min) and recovery (135 and 165 min) with vehicle (□) and FR139317 (▪) infusion. Columns and error bars indicate means and s.e.m., respectively. n= 5 unless indicated otherwise by number in parentheses. *P < 0.0125, significantly different from normoxaemia (45 min) by Student's unpaired t test.

Fetal heart rate and blood pressure

During normoxaemia FHR was greater in the FR139317- than in the vehicle-infused group (P < 0.05, Fig. 2 and Table 2). At the onset of hypoxaemia there was a rapid initial decrease in FHR in both vehicle (P < 0.05) and FR139317 (P < 0.01) groups. As hypoxaemia continued, FHR returned to normoxaemic levels although there was no difference in FHR between vehicle and FR139317 groups. During recovery a rebound tachycardia was apparent in both groups but this only reached significance with vehicle infusion (recovery vs. normoxaemia, P < 0.01). Moreover there was no difference in FHR between the vehicle and FR139317 groups during recovery.

Figure 2. The effect of FR139317 on fetal heart rate and mean arterial blood pressure during normoxaemia and hypoxaemia.

Fetal heart rate and MABP responses to acute hypoxaemia (shaded region) during vehicle (○, n= 5) and FR139317 (•, n= 5) infusion. Symbols and error bars indicate means and s.e.m., respectively. The dashed line indicates the onset of infusion at 0 min. The horizontal bars show the time periods over which data were reduced to a summary measure. *P < 0.05 and **P < 0.01vs. normoxaemia in vehicle group by Student's paired t test; ††P < 0.01vs. normoxaemia in FR139317 group by Student's paired t test; §P < 0.05, 60 vs. 65 min in vehicle group by Student's paired t test; P < 0.01, 60 vs. 65 min in FR139317 group by Student's paired t test; ‡P < 0.05 FR139317 vs. vehicle group by Student's paired t test.

Table 2.

Summary measures of cardiovascular parameters during normoxaemia and hypoxaemia in vehicle- and FR139317-infused groups

	FHR (beats min−1)	MABP (mmHg)	FBF (ml min−1)	FVR (mmHg min ml−1)	CBF (ml min−1)	CVR (mmHg min ml−1)
Normoxaemia
Vehicle	166 ± 6	50.2 ± 1.6	52.49 ± 5.40	1.01 ± 0.11	75.91 ± 5.99	0.68 ± 0.04
FR139317	186 ± 2 ‡	49.4 ± 0.5	55.55 ± 7.97	0.99 ± 0.16	94.53 ± 7.85 ‡‡	0.55 ± 0.04 ‡‡
Hypoxaemia
Vehicle	155 ± 7	60.6 ± 3.5 *	30.10 ± 3.99 **	2.21 ± 0.35 **	96.99 ± 8.45 **	0.65 ± 0.06
FR139317	164 ± 10	58.3 ± 2.1 ††	36.66 ± 6.29 ††	1.89 ± 0.40 †	110.60 ± 12.03 ‡	0.55 ± 0.05 ‡‡
Recovery
Vehicle	195 ± 5 **	55.6 ± 3.5	59.24 ± 4.00	0.99 ± 0.11	83.05 ± 9.07	0.72 ± 0.08
FR139317	197 ± 7	49.6 ± 3.5 ‡	60.09 ± 8.35	0.93 ± 0.17	91.29 ± 5.49	0.56 ± 0.03 ‡

Values are shown as means ±s.e.m.; n= 5.

^*P < 0.05

^**P < 0.01vs. normoxaemia in vehicle group

^†P < 0.05

^††P < 0.01vs. normoxaemia in FR139317 group

^‡P < 0.05

^‡‡P< 0.01 FR139317 vs. vehicle group.

There was no change in MABP after the onset of FR139317 infusion. MABP rose during hypoxaemia with vehicle (P < 0.05) and FR139317 (P < 0.01) infusion (Fig. 2 and Table 2). Pressure was greater with vehicle than with FR139317 infusion during the recovery period (P < 0.05).

Femoral circulation

During normoxaemia and hypoxaemia FBF tended to be greater with FR139317 than with vehicle infusion although there was no significant difference in FBF or FVR between these groups throughout the course of the procedure (Fig. 3 and Table 2). In hypoxaemia there was a rapid decrease in FBF (vehicle and FR139317, P < 0.01) and increase in FVR (vehicle, P < 0.01; FR139317, P < 0.05) which was sustained throughout hypoxaemia.

Figure 3. The effect of FR139317 on femoral and carotid blood flow and vascular resistance during normoxaemia and hypoxaemia.

Femoral and carotid blood flow and vascular resistance responses to acute hypoxaemia (shaded region) during vehicle (○, n= 5) and FR139317 (•, n= 5) infusion. Symbols and error bars indicate means and s.e.m., respectively. The dashed line indicates the onset of infusion at 0 min. The horizontal bars show the time periods over which data were reduced to a summary measure. **P < 0.01vs. normoxaemia in vehicle group by Student's paired t test; †P < 0.05 and ††P < 0.01vs. normoxaemia in FR139317 group by Student's paired t test; ‡P < 0.05 and ‡‡P < 0.01, FR139317 vs. vehicle group by Student's paired t test.

Carotid circulation

During normoxaemia CBF was higher (P < 0.01) and CVR was lower (P < 0.01) with FR139317 than with vehicle infusion. This modulation of CBF (P < 0.05) and CVR (P < 0.01) persisted throughout hypoxaemia (Fig. 3 and Table 2). Furthermore CVR remained lower with antagonist than with vehicle infusion during the recovery hour (P < 0.05). During hypoxaemia CVR did not change from normoxaemia in either group. The increase in CBF during hypoxaemia with vehicle infusion (P < 0.01) did not reach significance with FR139317 infusion.

DISCUSSION

The results of this study show that ETA receptor antagonism alters basal FHR, CBF and CVR. This modulation of carotid vascular tone persists during hypoxaemia and recovery. On the other hand ETA receptor activation does not appear to modulate or mediate basal or hypoxaemia-stimulated changes in MABP or FVR. We did not observe any change in circulating [ET-1] during hypoxaemia in these fetuses.

In humans basal plasma [ET-1] is higher in the fetal than in the maternal circulation which implies that the fetus is able to produce ET-1 (Nakamura et al. 1990). A range of basal fetal plasma [ET-1] levels has been reported: mixed human umbilical plasma, ca. 9 pmol l⁻¹ (Haegerstrand et al. 1989); cord umbilical artery, ca. 3 pmol l⁻¹ (Nakamura et al. 1990); and late gestation ovine plasma, ca. 3 pmol l⁻¹ (Jones, Abman & Wilkening, 1994). In this study we have reported basal plasma [ET-1] of between 10 and 14 pmol l⁻¹. Previous adult in vivo and in vitro studies have suggested that hypoxaemia stimulates ET-1 production (Rakugi et al. 1990; Shirakami et al. 1991) although we did not measure maternal plasma [ET-1] in the present study. We have shown that there is no change in circulating fetal [ET-1] after 1 h of hypoxaemia in agreement with preliminary studies by Jones et al. (1994). However, since cultured endothelial cells release the majority of ET-1 towards the abluminal surface, circulating [ET-1] may not accurately reflect the local concentration of the peptide within the blood vessel wall. Alternatively it may be that any increase in local ET-1 production during hypoxaemia does not pass into the systemic circulation in high enough concentrations to be detected (Wagner et al. 1992). Measurement of ET-1 peptide and receptor gene expression (Li et al. 1994) in specific vascular beds would be a more appropriate technique but this was beyond the scope of the present study.

During normoxaemia FHR increased with the infusion of FR139317 compared with the vehicle group in contrast to other fetal studies in which the ETA receptor antagonist BQ 123 did not change resting FHR (Ivy et al. 1994; Wong et al. 1994). The increased FHR observed in the present study was not accompanied by a change in MABP which implies a negative-chronotropic effect of endogenous ET-1 via ETA receptors by a direct action on the fetal heart. Indeed ETA and ETB receptor subtypes have been localized to adult atrial and ventricular myocardium, the atrioventricular conducting system and endocardial cells (Golfman, Hata, Beamish & Dhalla, 1993).

Systemic infusion of ET-1 (Chatfield, McMurtry, Hall & Abman, 1991) and big ET-1 (Jones & Abman, 1994) produces a sustained hypertension in the sheep fetus. However in the present study we observed no effect of ETA receptor antagonism on peripheral vascular resistance, demonstrated by the similarity of FBF and MABP measurements between vehicle and FR139317 groups after the onset of infusion during normoxaemia. These findings agree with those of Ivy et al. (1994) using BQ 123 in the late gestation ovine fetus. On the other hand during normoxaemia, CBF was higher and CVR was lower in the FR139317 than in the vehicle group which suggests a tonic vasoconstrictory action of ET-1 via ETA receptors in the carotid vascular bed. Indeed abluminal administration of ET-1 in the adult rat vasoconstricts the microvasculature of the cerebral cortex (Willette et al. 1990). Alternatively, ET-1 can also act via ETA receptors to inhibit induced NO synthesis in cultured vascular smooth muscle cells (VSMC) (Ikeda, Yamamoto, Maeda, Shimpo, Kanbe & Shimada, 1997) and we have recently shown that NO synthesis regulates resting carotid vascular tone (Green, Bennet & Hanson, 1996). In fetal sheep ET-1 stimulates renal vasodilatation via ETB receptor-mediated NO release (Bogaert et al. 1996). The relative contribution of ETA and ETB receptor subtypes to vascular responses to exogenous ET-1 in the conscious adult rat is dependent upon the vascular bed examined (Gardiner et al. 1994). Such vascular bed selectivity might explain the apparent differing contribution of endogenous ET-1 to basal femoral and carotid vascular tone in the present study. However, while the dose of FR139317 used in the present study antagonized the action of an exogenous bolus of ET-1, we cannot address the extent to which the action of local concentrations of endogenous ET-1 was antagonized.

A role for ET-1 has been implicated in the decrease in ovine fetal hindlimb blood flow during 3 h of hypoxaemia induced by maternal common iliac artery occlusion (Jones, 1995) and in the pulmonary hypertensive response of the lamb (Wang et al. 1995). However in the present study ETA receptor antagonism did not affect the MABP response to hypoxaemia since pressure rose to a similar level in vehicle and FR139317 groups. Furthermore while FBF during hypoxaemia tended to be greater with FR139317 than with vehicle infusion this did not reach significance. Thus we suggest that ETA receptor activation does not play a large role in the fetal peripheral vascular changes during acute hypoxaemia. However, there is evidence of a small number of ETB receptors on porcine aortic smooth muscle cells which mediate vasoconstriction, and if present these would not have been antagonized by FR139317 in the present study (Shetty, Okada, Webb, DelGrande & Lappe, 1993).

Endothelin-1 modulated carotid vascular tone, probably via ETA receptors, during normoxaemia and this action persists during hypoxaemia and recovery. There was no change in CVR during hypoxaemia in vehicle- or FR139317-infused groups, but our results show that the increase in CBF seen with vehicle infusion during hypoxaemia does not reach significance in the FR139317 group, presumably because the basal CBF was higher and there was no difference in MABP between the two groups during hypoxaemia. However, in this study we did not investigate the role of ETB receptors directly. The vasodilatory action of ET-1 via ETB receptors is mediated via NO release (Bogaert et al. 1996). Furthermore, ET-1 acts via ETA receptors to inhibit induced NO synthesis in VSMCs (Ikeda et al. 1997) and our recent observations show that an increase in NO synthesis plays a key role in the vasodilatation seen in the fetal carotid vascular bed in hypoxaemia (Green et al. 1996).

In summary we have shown that antagonism of ETA receptors alters baseline FHR, CBF and CVR. This modulation of carotid vascular tone persists throughout hypoxaemia and recovery. However, ETA receptors do not play a role in mediating the increase in CBF or the decrease in peripheral blood flow during mild hypoxaemia. These findings are supported by the lack of change in plasma [ET-1] during hypoxaemia. Future studies will need to assess the production of ET-1 and its action via receptor subtypes at the local tissue level, especially during more intense and prolonged hypoxaemia.