INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

FINDINGS FROM A NUMBER OF laboratories have suggested that nitric oxide (NO), a labile compound produced from the amino acid L-arginine via the action of the ubiquitous enzyme NO synthase (21), acts as a modulator of both circulatory dynamics and metabolic processes in the mature fetus. Thus inhibition of NO synthesis in late-gestation fetal lambs increases arterial blood pressure (6, 7, 9, 22, 29, 32) and reduces the combined left (LV) and right (RV) ventricular output (9, 29) without changing the degree of dominance of the RV output (29). Inhibition of NO synthesis also decreases fetal body (29) and placental blood flow (6, 7, 29). However, fetal body blood flow is reduced to a greater extent than placental blood flow, thereby redistributing systemic blood flow away from the fetus and toward the placenta (29). Moreover, the reduction in fetal body blood flow occurring with inhibition of NO synthesis is accompanied by an increase in fetal body O₂ extraction that initially elevates fetal whole body O₂ consumption and then maintains it at baseline levels in the face of falls in placental perfusion (29).

Inhibition of NO synthesis also increases arterial blood pressure and reduces the combined ventricular output in the immature, midgestation fetal lamb (10). However, several lines of evidence suggest that the physiological role of NO at midgestation may differ from late gestation in a number of other aspects. Specifically, in vitro data indicate that the potency of NO as a vasodilator in the umbilical-placental circulation is more pronounced at midgestation than at late gestation (18, 28). Because the placenta is mainly perfused by the right ventricle (24, 26, 30) and is central to fetomaternal gas exchange and fetal O₂ delivery (25), it is therefore possible that, in the immature fetus, NO plays a pivotal role in maintaining not only the predominance of RV output within the "in parallel" fetal circulation (26) and the near equal distribution of the combined ventricular output between the fetus and placenta (16, 17) but also the level of fetal oxygenation. Accordingly, the aim of the present study was to test these hypotheses in chronically instrumented midgestation fetal lambs in which hemodynamic, blood flow, and blood gas measurements were performed after incremental inhibition of constitutive NO synthesis with the stereospecific and potent NO synthase inhibitor N-nitro-L-arginine (L-NNA; see Ref. 12).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Experiments were approved by the Monash University Animal Experimentation Committee and were conducted in accord with guidelines established by the National Health and Medical Research Council of Australia.

Animal preparation. Six fetuses with known breeding dates were chronically instrumented under aseptic conditions at 89-96 days' gestation (term = 147 days). Border-Leicester cross ewes were anesthetized with intravenous propofol (5 mg/kg), intubated, and then mechanically ventilated with 1-3% halothane and a 2:1 nitrous oxide-oxygen mixture. The pregnant horn of the uterus was exposed through a midline laparotomy. After delivery of the fetal hindlimbs through a hysterotomy, polyvinyl catheters were inserted in the femoral artery bilaterally in the groin and in a femoral vein and advanced to the descending aorta and inferior vena cava, respectively. The fetus was returned to the uterus, and the fetal head, left forelimb, and upper thorax were delivered through a second hysterotomy. A fetal thoracotomy was performed in the third left interspace, and the fourth rib was removed to increase exposure of the heart and great vessels. After incision of the pericardium, a 6- or 8-mm ultrasonic flow probe (Transonic Systems, Ithaca, NY) was placed around the pulmonary trunk. A 22-guage Teflon cannula was inserted in the pulmonary trunk through an adventitial purse-string suture distal to the flow probe and was connected to a polyvinyl catheter. After insertion of a polyvinyl catheter in the left atrial cavity through a purse-string suture in the appendage, the pericardium was loosely closed, and overlying muscle layers were repaired. The left axillary artery and vein were then exposed, and catheters were passed via these vessels into the ascending aorta or brachiocephalic trunk and the superior vena cava, respectively. A polyvinyl catheter was inserted in a cotyledonary vein, and the tip was advanced to a major umbilical vein. All catheters were exteriorized, and, after suturing of a wide-bore catheter to the skin of the anterior chest wall for measurement of amniotic fluid pressure, the fetal skin and maternal uterine incisions were closed.

Experimental protocol. Ewes were placed in a mobile laboratory cart 3-4 days after surgery and were allowed free access to feed and water. Fetal ascending aortic and pulmonary trunk blood pressures, heart rate, and pulmonary trunk flow were recorded, and 0.4-ml blood samples were collected anerobically from the ascending aorta, pulmonary trunk, abdominal aorta, and the umbilical vein for Hb and blood gas analysis. Fetal ventricular outputs and blood flows to the fetal body and placenta were then measured with radioactive microspheres using the reference sample method (15). After baseline measurements, NO synthesis was inhibited with L-NNA (Sigma Chemical), which was dissolved in normal saline to a concentration of 5 mg/ml and infused continuously through the femoral venous catheter at a rate of 0.68 ml/min to achieve cumulative doses of 10 mg/kg (L-NNA₁₀) and 25 mg/kg (L-NNA₂₅) of estimated fetal body weight. At each L-NNA dose, hemodynamics were allowed to stabilize over a 5-min period, and blood pressure, heart rate, blood flow, and blood gas measurements were then repeated.

Physiological measurements. Ascending aortic and pulmonary arterial blood pressures were referenced to amniotic fluid pressure. Both pressures were monitored with silicon-chip pressure transducers (model CDX-111; Cobe Laboratories, Lakewood, CO), which were calibrated against a manometer before each experiment. Pulmonary trunk blood flow (i.e., RV output) was measured continuously with an ultrasonic flowmeter (model T208; Transonic Systems). The outputs from the pressure transducers and flowmeter were amplified using a programmable signal conditioner (Cyberamp model 380; Axon Instruments, Foster City, CA), and signals were displayed continuously on a paper recorder (model 800Z; Neomedix Systems, Sydney, New South Wales, Australia). At baseline and the two L-NNA doses, a 30-s segment of hemodynamic data was also digitized at a sampling rate of 200 Hz and stored on computer hard disk for subsequent off-line analysis using customized interactive software.

Radioactive microsphere technique. Radioactive microspheres, 15 µm in diameter and labeled with one of five gamma emitting isotopes (¹⁴¹Ce, ¹¹³Sn, ⁸⁵Sr, ⁹⁵Nb, or ⁴⁶Sc; New England Nuclear, Boston, MA), were ultrasonicated for 10-15 min before injection and injected over 30-45 s. At baseline and at L-NNA₂₅, two different microsphere labels were injected simultaneously, one in the left atrium to measure LV output and the other in the superior vena cava to measure RV output (31), whereas reference samples were drawn simultaneously from the ascending aorta, pulmonary trunk, and descending aorta for determination of fetal body and umbilical-placental flows. At L-NNA₁₀, LV output was obtained with a single microsphere label injected in the left atrium, whereas reference samples were withdrawn from the ascending and descending aorta to obtain fetal body and placental flows. Approximately 0.5-1 × 10⁶ microspheres were injected per radiolabel, and reference samples were withdrawn at a rate of 2.05 ml/min with a mechanical pump (model 901A; Harvard Apparatus, South Natick, MA). Reference sample collection was commenced 5-10 s before injection of microspheres and was continued for an additional 75 s after the end of injection. Blood withdrawn in the reference samples was replaced simultaneously with maternal blood mixed with a plasma substitute (Hemaccel; Behring, Marburg, Germany).

Calculation of fetal ventricular outputs and fetal body and placental flows. Radioactive microsphere blood flow measurements were calculated using the general relation Q_Tissue = (Q_Reference × R_Tissue)/R_Reference, where Q is flow (ml/min) and R is radioactivity (counts/min). LV and RV outputs and systemic flows were obtained using an adaptation of this general relation for the fetal circulation (31). Thus fetal LV output (Q_LV) was equal to (Q_Reference × R_LA)/R_LAAA, where R_LA is the radioactivity of the label injected into the left atrial cavity and R_LAAA is the radioactivity of the same label collected in the ascending aortic reference sample. Fetal RV output (Q_RV) was equivalent to (Q_Reference × R_RV)/R_VPT, where R_RV is the radioactivity of venous label passing into the right ventricle, calculated as the injected radioactivity of this label minus that portion crossing the foramen ovale to appear in the LV output, and R_VPT is the radioactivity of the venous label in the pulmonary reference sample (31). RV output at L-NNA₁₀ was obtained by interpolation from the measured pulmonary trunk flow, using baseline and L-NNA₂₅ ultrasonic flow probe measurements of pulmonary trunk flow and radioactive microsphere determinations of RV output (9, 10, 29).

Blood gas calculations. The O₂ content of arterial or venous blood (ml O₂/dl blood) was calculated as (1.36 × HbS × Hb/100) + (0.003 × PO₂) where HbS is Hb O₂ saturation (%), Hb is Hb level (g/dl), and PO₂ is O₂ tension (mmHg).

Statistics. Changes in hemodynamics, blood flows, and blood gas variables were analyzed with repeated-measures one-way ANOVA using the Statistical Package for Social Sciences version 9.0.1 (SPSS). The sums of squares were partitioned into individual degrees of freedom, and the significance of changes was evaluated using the Bonferroni procedure, as appropriate, for multiple tests (34). Results are reported as means ± SE, and P < 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

On the day of study, gestational age was 97 ± 1 days, fetal body weight 1.04 ± 0.09 kg, and placental weight 0.75 ± 0.06 kg.

Hemodynamics and blood gases. At baseline, mean ascending aortic blood pressure was 1.4 ± 0.3 mmHg lower than pulmonary arterial blood pressure (P < 0.005). Ascending aortic blood pressure increased by 13.5 ± 1.5 mmHg and pulmonary arterial blood pressure by 14.3 ± 1.6 mmHg at L-NNA₁₀ (both P < 0.005), but neither was altered further at L-NNA₂₅. With the exception of an increase in Hb concentration (P < 0.005), blood gas variables at L-NNA₁₀ were unchanged from baseline. However, L-NNA₂₅ was accompanied by a further rise in Hb concentration (P < 0.025), as well as falls in pH (P < 0.025) and Hb O₂ saturation (P < 0.005). Neither heart rate, PO₂, PCO₂, nor O₂ content changed significantly with L-NNA (Table 1).

View this table: [in this window] [in a new window]   Table 1.   Hemodynamics and ascending aortic blood gases before and after N-nitro-L-arginine

Ventricular outputs. The baseline combined ventricular output (539 ± 24 ml · min¹ · kg¹) fell by 164 ± 15 ml · min¹ · kg¹ at L-NNA₁₀ (P < 0.005) and by a further 93 ± 31 ml · min¹ · kg¹ at L-NNA₂₅ (P < 0.05, Fig. 1A). Comparing ventricles, baseline LV output (222 ± 20 ml · min¹ · kg¹) decreased by 74 ± 15 ml · min¹ · kg¹ at L-NNA₁₀ (P < 0.005) and was then unchanged at L-NNA₂₅, whereas resting RV output (317 ± 31 ml · min¹ · kg¹) fell by 90 ± 18 ml · min¹ · kg¹ at L-NNA₁₀ (P < 0.02) and by a further 80 ± 22 ml · min¹ · kg¹ at L-NNA₂₅ (P < 0.05). Thus, whereas RV output exceeded LV output at baseline (P = 0.03) and L-NNA₁₀ (P < 0.02), these outputs were not different at L-NNA₂₅ (P > 0.7; Fig. 1B).

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Changes in the combined ventricular output (A) and the left (open bars) and right (filled bars) ventricular outputs (B) at baseline and after administration of 10 mg/kg (L-NNA10) and 25 mg/kg (L-NNA25) N-nitro-L-arginine. *P < 0.05 and **P < 0.025, left vs. right ventricular output.

Fetal and placental blood flows. Baseline fetal body blood flow (275 ± 24 ml · min¹ · kg¹) decreased by 95 ± 25 ml · min¹ · kg¹ at L-NNA₁₀ (P < 0.01) and was unaltered at L-NNA₂₅ (Fig. 2A). By contrast, resting placental blood flow (264 ± 28 ml · min¹ · kg¹) fell by 70 ± 22 ml · min¹ · kg¹ at L-NNA₁₀ (P < 0.01) and by a further 71 ± 21 ml · min¹ · kg¹ at L-NNA₂₅ (P < 0.01; Fig. 2B). Consequently, although the fetal body-to-placental blood flow ratio was similar at baseline (1.1 ± 0.2) and L-NNA₁₀ (1.0 ± 0.2), it rose to 1.5 ± 0.3 at L-NNA₂₅ (P < 0.05; Fig. 2C).

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Fetal body (A) and placental (B) blood flow and the fetal body-to-placental blood flow ratio (C) at baseline and after infusion of L-NNA10 and L-NNA25.

Fetal O₂ content, extraction, delivery, and consumption. Fetal average arterial O₂ content, venous O₂ content, O₂ arteriovenous content difference, and O₂ extraction ratio did not change significantly with inhibition of NO synthesis (Table 2). However, in association with the blood flow changes, average fetal body O₂ delivery declined progressively from 15.4 ± 1.5 ml · min¹ · kg¹ at baseline to 10.0 ± 2.0 ml · min¹ · kg¹ at L-NNA₂₅ (P < 0.01; Fig. 3A), whereas resting whole body O₂ consumption (7.2 ± 0.6 ml · min¹ · kg¹) decreased by 1.4 ± 0.3 ml · min¹ · kg¹ at L-NNA₁₀ (P < 0.05) and by a further 1.5 ± 0.6 ml · min¹ · kg¹ at L-NNA₂₅ (P < 0.02; Fig. 3B).

View this table: [in this window] [in a new window]   Table 2.   Fetal body blood gas variables before and after N-nitro-L-arginine

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Fetal whole body O2 delivery (A) and O2 consumption (B) at baseline and after administration of L-NNA10 and L-NNA25.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The main findings of this study were that incremental inhibition of NO synthesis in midgestation fetal lambs initially lowered combined ventricular output without affecting the degree of dominance of the RV output and reduced systemic perfusion while maintaining the fetoplacental blood flow ratio near unity. With a greater degree of inhibition of NO synthesis, however, a further reduction in placental perfusion was not only accompanied by a preferential decrease in RV output that equalized the univentricular outputs but also a redistribution of systemic flow away from the placenta and toward the fetal body. Moreover, because fetal arterial O₂ content and systemic O₂ extraction were unaltered after inhibition of NO synthesis, these flow changes were accompanied by reductions in both fetal O₂ delivery and O₂ consumption. These findings suggest that, in the midgestation fetal lamb, NO plays a major role in supporting a dominant RV output, a maintenance of whole body O₂ utilization, and a near-equal distribution of systemic blood flow between the fetus and placenta.

Resting values for arterial blood pressure (2, 16, 17, 20), combined ventricular output (16, 17), the ratio of fetal body and placental flows (16), Hb concentration (16, 17, 20), aortic O₂ saturation (17, 20), and whole body O₂ consumption (17) in the present study were similar to those previously reported in chronically instrumented, midgestation fetal lambs. The baseline heart rate (188 beats/min) was, however, slightly lower than the reported range of 196-224 beats/min (2, 16, 17, 20), possibly related to a variation among different sheep breeds.

As noted in a prior report (10), aortic blood pressure increased after NO synthase inhibition in midgestation fetuses. Moreover, as in late-gestation fetuses (22, 32), aortic and pulmonary arterial blood pressures rose by a similar amount at midgestation so that the pattern of a higher pulmonary arterial blood pressure evident at baseline was also present after inhibition of NO synthesis. However, as pointed out previously (10), the lack of any significant change in heart rate with inhibition of NO synthase in midgestation fetal lambs contrasted with the reduction in heart rate observed in late-gestation fetuses (6, 7, 29), presumably reflecting an incomplete development of baroreceptor responses due to an immaturity of parasympathetic chronotropic mechanisms (33).

Similar to late-gestation lambs (7, 29), Hb concentration rose at midgestation after inhibition of NO synthase. Because no significant release of erythrocytes occurs from storage sites such as the spleen in fetal sheep (3), the most plausible explanation for this rise in Hb concentration was a reduction in plasma volume associated with a fluid shift from intravascular to extravascular compartments. However, reductions in plasma volume appeared to be more pronounced at midgestation because Hb concentration increased by 2.4 g/dl from a baseline value of 6.6 g/dl, whereas the corresponding rise was only 1.6 g/dl from a baseline value of 9.6 g/dl in late-gestation fetuses (29). Indeed, substitution of these Hb values into the formula devised by Brace (3) indicated that the reduction in blood volume during inhibition of NO synthesis at midgestation (36%) was more than double that in late-gestation fetuses (15%). The basis of this difference is not entirely clear at present but could relate to factors such as a higher permeability of fetal capillaries at midgestation, leading to a greater degree of fluid extravasation with rises in hydrostatic pressure (4) or a greater role for NO in the maintenance of normal vascular permeability (19) within the immature fetal circulation.

Combined ventricular output fell stepwise with incremental inhibition of NO synthase in midgestation fetuses, with the magnitude of the reductions being similar to that occurring at late gestation (29). As at late gestation, it is likely that increased arterial blood pressure (14, 23), hemoconcentration (11), and reduced circulating blood volume (13) contributed to this decline in the combined ventricular output. However, within the similarity in the response of the combined ventricular output, an important gestation-dependent difference was apparent in the effect of inhibition of NO synthase on the univentricular outputs. Thus, in accord with data from a range of laboratories (1, 26, 30), RV output exceeded LV output at baseline in late-gestation fetuses; furthermore, the relative contribution of these outputs to the combined ventricular output (40 and 60%, respectively) was unaffected by inhibition of NO synthesis (29). At midgestation, RV output also constituted 60% of the combined ventricular output at baseline and L-NNA₁₀. However, because RV output fell further at L-NNA₂₅ while LV output was similar to that at L-NNA₁₀, the univentricular outputs were not different at the higher dose of NO synthase inhibitor, suggesting that NO played an important role in supporting a dominance of the RV output in midgestation fetal lambs.

Two main factors could have underpinned the preferential reduction in RV output observed at L-NNA₂₅ in the present study. The first relates to the finding that rises in arterial pressure in utero produce greater reductions in RV output than in LV output (23). However, this is unlikely to be the explanation for our observation because, while RV output exceeded LV output at L-NNA₁₀ but not L-NNA₂₅, ascending aortic and pulmonary arterial pressures were unchanged between these L-NNA doses (Table 1). The second factor, which stems from the notion that the dominance of the RV output in the fetal circulation is due to the right ventricle being the major source of blood to the low-resistance placental circulation (24), is that the preferential reduction in RV output observed at L-NNA₂₅ was primarily related to the accompanying marked fall in placental perfusion. In support of this explanation, the reduction in placental perfusion between L-NNA₁₀ and L-NNA₂₅ (71 ml · min¹ · kg¹) accounted for 90% of the associated decrease in RV output (80 ml · min¹ · kg¹; Figs. 1 and 2).

The pattern of change in fetal body blood flow occurring with inhibition of NO synthesis at midgestation, namely a reduction in blood flow at L-NNA₁₀ of 30% and no further change at L-NNA₂₅, was similar to that observed at late gestation (29). By contrast, three main differences in the placental blood flow pattern were apparent on comparison of present findings with late-gestation fetal lamb data from our laboratory (29). First, resting placental flow at midgestation was 50% higher than the value of 175 ml · min¹ · kg¹ measured at late gestation. Second, in midgestation fetuses, placental flow fell by 23% at L-NNA₁₀ and by a further 30% from baseline at L-NNA₂₅ (Fig. 2), whereas, in late-gestation fetuses, placental flow was unchanged between baseline and L-NNA₁₀ and only fell by 21% at L-NNA₂₅. Third, at L-NNA₂₅, placental blood flow at midgestation (123 ml · min¹ · kg¹) was not too dissimilar from the value of 140 ml · min¹ · kg¹ measured at late gestation. Taken together, these findings imply that an NO-related mechanism not only underpinned a higher resting level of placental perfusion at midgestation compared with late gestation but also contributed to 50% of resting placental perfusion at midgestation compared with 20% at late gestation. These conclusions are in accord with in vitro human data that indicate that NO is a potent vasodilator within the umbilical-placental vascular bed (18) and that this vasodilator action diminishes with advancing gestation (18, 28).

An important consequence of the pronounced placental flow reduction occurring with inhibition of NO synthase at midgestation was a redistribution of systemic blood flow between the fetal and placental compartments. Thus, although blood flow was distributed near equally between these compartments at baseline and L-NNA₁₀, it favored the fetal body at L-NNA₂₅. This redistribution was the opposite of that observed in late-gestation fetuses, where the resting fetal-to-placental blood flow ratio fell from 1.9 to 1.4 after inhibition of NO synthesis (29). Interestingly, these divergent fetoplacental blood flow responses bear a striking resemblance to the differing systemic flow redistribution patterns observed at mid (16)- and late gestation (8) during acute fetal hypoxemia, which further supports the prior suggestion that a reduction in NO production may contribute to the physiological manifestations of acute in utero hypoxemia (29).

As at late gestation (29), arterial O₂ content was unchanged during the reduction in fetal body blood flow occurring with inhibition of NO synthase at midgestation; hence, fetal O₂ delivery was correspondingly decreased. However, in contrast to the reduction in mixed venous O₂ content and a rise in systemic O₂ extraction evident in late-gestation fetuses (29), no significant changes in mixed venous O₂ content, systemic O₂ extraction, or the O₂ extraction coefficient occurred after inhibition of NO synthesis at midgestation (Table 2). Consequently, although fetal whole body O₂ consumption in late-gestation fetuses was increased initially by 15% and then remained at near-baseline levels after a reduction in placental perfusion (29), inhibition of NO synthesis at midgestation was associated with a progressive fall in fetal whole body O₂ consumption to a level that was only 60% of baseline at the highest dose of NO synthase inhibitor (Fig. 3).

A possible explanation for the marked difference in the response of whole body O₂ consumption to inhibition of NO synthesis between mid- and late-gestation fetal lambs relates to the modulatory effects of NO on oxidative metabolism. Thus NO is known to be a potent inhibitor of enzymes within the mitochondrial respiratory chain, particularly cytochrome c oxidase (5). This action of NO appears to impose a substantial brake on oxidative metabolism in the adult because, although inhibition of NO synthesis reduces cardiac output, it results in an even greater increase in O₂ extraction, thereby raising whole body O₂ consumption (27). A similar modulatory action is apparent in late-gestation fetal lambs, with inhibition of NO synthesis resulting in an increased O₂ extraction that maintains whole body O₂ consumption at or above the baseline level despite falls in fetal perfusion (29). However, the absence of an increase in O₂ extraction in the face of reduced tissue perfusion in the present study suggests that this modulatory action was not present in midgestation fetuses, so that fetal whole body O₂ consumption exhibited a flow dependence after inhibition of NO synthesis.

Perspectives