INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE PATTERN OF PERINATAL BRAIN injury resulting from hypoxic-ischemic events varies during development (24). Maintaining viable levels of oxygen delivery during intrauterine periods of hypoxia and during parturition depends on adequate cerebral vasodilation. Cerebral blood flow (CBF) regulation during early development has been studied most extensively in fetal sheep, which have a gestational length of 145 days. At 93 days conceptional age (0.6 gestation), CBF and cerebral O₂ consumption are about one-third of the values in fetal sheep near term (9). Furthermore, the increase in CBF during hypoxia at 0.6 gestation relative to the low baseline CBF is disproportionately less than that occurring near term, thereby resulting in reduced cerebral O₂ delivery (10). This attenuated cerebrovascular response appears to be specific for hypoxia because the percent increase in CBF during hypercapnia is similar in fetuses at 0.6 and 0.9 gestation (15) and because pial arteriolar vasodilation to adenosine analogs (22) and acetylcholine (38) is similar at these two ages.

Because nitric oxide (NO) is an important vasodilator in the cerebral circulation, we considered the possibility that developmental changes in NO synthase (NOS) may contribute to developmental changes in hypoxic cerebral vasodilation. In fetal sheep, endothelial NOS is already well localized in blood vessels by 0.4 gestation (28). However, major changes in cell localization of neuronal NOS from the neuropil to distinct cell bodies occurs progressively between 0.4 and 0.9 days of gestation (28). Moreover, cortical NOS catalytic activity increases threefold between 0.6 and 0.9 gestation (29). Thus changes in CBF and O₂ consumption parallel changes in NOS activity during the last trimester in fetal sheep.

Studies that have addressed the role of NO in hypoxic vasodilation in fetal sheep brain have yielded different conclusions, possibly because of differences in experimental design. When NOS inhibitors were administered to fetal sheep at 0.8 (34) or 0.9 (26) gestation during hypoxia, CBF decreased considerably, leading the authors to conclude that NO contributes to cerebral vasodilation during hypoxia. However, NOS inhibitors also decrease normoxic baseline CBF in fetal sheep (29). When a NOS inhibitor was administered to near-term fetal sheep before hypoxia and then CBF during hypoxia was compared with the lowered normoxic CBF, the expected increase in CBF was observed during hypoxia (37). Thus NO may exert a tonic vasodilatory influence with both normoxic and hypoxic fetal blood gases but not necessarily contribute to hypoxic vasodilation. Hence the precise role of NO in the fetal CBF response to hypoxia remains unclear. Moreover, the effect of NOS inhibition on the CBF response to hypoxia in near-term fetuses has not been contrasted to that at 0.6 gestation when NOS catalytic activity and hypoxic vascular reactivity are not fully developed. In the present study, we tested the hypothesis that administration of the NOS inhibitor N-nitro-L-arginine methyl ester (L-NAME) before hypoxia would attenuate the percent increase in CBF during hypoxia in fetal sheep at 0.9 gestation (133 days of gestation) but not at 0.6 gestation (93 days of gestation).

The radiolabeled microsphere technique used to measure CBF also permitted blood flow to be determined to other organs during hypoxia. Blood flow to the gastrointestinal tract, kidney, and carcass remains unchanged or increases slightly during moderate fetal hypoxia but decreases markedly when arterial O₂ content is <1.5 mmol/l (30). In peripheral organs of adult animals, NO is thought to limit sympathetic vasoconstriction, possibly through modulatory effects of NO in the nervous system as well as at the blood vessels (17, 23, 40). If this modulatory effect of NO is present in the fetus, we would expect L-NAME administration to enhance vasoconstriction in selected peripheral organs during hypoxia, especially at a level of arterial O₂ content just above the 1.5 mmol/l threshold for observing severe reductions in peripheral blood flow. Thus a second hypothesis of this study was that peripheral vasoconstriction would be greater during hypoxia after L-NAME treatment in fetal sheep at 0.9 gestation. Peripheral organ blood flow was not measured in fetuses at 0.6 gestation because of the technical difficulty of catheterizing hindlimb arteries for lower body microsphere reference samples and because of limitations on the amount of blood replacement needed for microsphere reference samples at this age.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

All procedures on time-dated pregnant sheep were approved by the institutional animal care and use committee. Fetal sheep were studied at either 93 ± 1 (SD) or 133 ± 1.5 conceptional days. Two days before the experiment, the fetal vessels were chronically catheterized as previously described (10, 12-14). The ewe was anesthetized by face mask with halothane, and the trachea was orally intubated. The lungs of the ewe were mechanically ventilated with a mixture of 1.5% halothane, 30% nitrous oxide, and balance oxygen. Under sterile conditions, the uterus was exposed through a midline abdominal incision. Through an incision in the uterus, the head of the fetus was exteriorized. A small hole was made in the skull for insertion of a catheter into the sagittal sinus. Another catheter was sutured to the ear of the fetus to measure pressure in the amniotic fluid. The head was inserted back in the uterus, and each forelimb was then exposed for catheterization of the axillary arteries. Lastly, a hindlimb was exposed and a catheter was inserted into the inferior vena cava via a pedal vein. In 133 day fetuses, a catheter was also inserted into a pedal artery for postductal arterial sampling. All incisions in the fetus and the ewe were closed with sutures. Catheters were exteriorized through the ewe's flank. Procaine penicillin (1,200,000 U) was injected intramuscularly to the ewe before surgery, and ampicillin (500 mg) was infused into the amniotic fluid after surgery.

Experiments were performed with the ewe standing in a cart. Fetal arterial blood pressure was measured in reference to amniotic fluid pressure. Blood samples from axillary arterial and sagittal sinus catheters were analyzed for pH, partial pressure of CO₂ (PCO₂), and partial pressure of O₂ (PO₂) with a Radiometer ABL 30 electrode system. Hemoglobin concentration and O₂ saturation were measured with a Radiometer Hemoximeter OSM3. Oxygen content was calculated by assuming an O₂ binding capacity of 1.34 ml O₂/g hemoglobin.

Blood flow was measured by inferior vena caval injection of 1.0-1.5 million microspheres (15 µm diameter) labeled with either ¹⁵³Gd, ^114mIn, ¹¹³Sn, ¹⁰³Ru, ⁹⁵Nb, or ⁴⁶Sc (New England Nuclear Life Sciences Products). An arterial blood reference sample was obtained by continuous withdrawal of blood at a rate of 1.17 ml/min in 93 day fetuses and 2.55 ml/min in 133 day fetuses starting 0.5 min before the microsphere injection and ending 1 min after flushing the injection catheter. To provide blood replacement, blood was freshly drawn from the jugular vein of the ewe and then exchange transfused with fetal blood 1 h before the start of the experiment. Blood withdrawn from the fetus was retransfused intravenously during the experiment. This procedure minimizes changes in oxyhemoglobin affinity due to maternal blood transfusion during the experiment. The arterial reference sample obtained from the axillary catheter with the tip in the brachiocephalic artery was used for calculating blood flow to brain and heart. In 133 day fetuses, a reference sample was also obtained from the pedal artery catheter with the tip in the abdominal aorta for calculating blood flow to placenta and lower body fetal organs. A lower reference sample was not obtained in 93 day fetuses because the pedal artery could not be reliably catheterized and the additional blood volume replacement could significantly change baseline oxyhemoglobin affinity. Tissue and blood samples were counted for radioactivity with correction for spectral overlap of isotopes, and blood flow was calculated as previously described (13). Vascular resistance was calculated as the ratio of mean arterial pressure to regional blood flow. Cerebral O₂ consumption was calculated as the product of forebrain blood flow and O₂ content difference between axillary artery and sagittal sinus blood samples.

After obtaining baseline measurements, L-NAME was injected intravenously (20 mg/kg) in 93 day fetuses (n = 6) and 133 day fetuses (n = 8). Because this dose did not statistically decrease CBF in 133 day fetuses, an additional group of 133 day fetuses (n = 6) was studied with a higher dose of L-NAME (60 mg/kg), which decreases NOS catalytic activity by 89% in fetal brain (29). Baseline measurements during normoxia were repeated ~30-60 min after the L-NAME injection. Control groups of 93 day fetuses (n = 9) and 133 day fetuses (n = 7) did not receive L-NAME. To produce hypoxia, inspired O₂ concentration was decreased through a high airflow tube connected to a plastic bag secured over the head of the ewe (10). CO₂ was added to the gas mixture to help maintain fetal isocapnia during hypoxia. The level of fetal hypoxia was targeted at an ~50% reduction in arterial O₂ content (CaO₂) in 93 day fetuses and at a 30 and a 50% reduction in CaO₂ in 133 day fetuses. Measurements were made 10 min after the inspired O₂ concentration was sequentially decreased. Normoxic measurements before and after L-NAME were made with room air passing through the plastic bag over the head of the ewe.

Data were analyzed by two-way ANOVA where L-NAME treatment was a between-subject factor and hypoxia was a within-subject factor. If there was a significant group treatment effect or an interaction of hypoxia with group treatment, then a one-way ANOVA was performed among groups during normoxia and during hypoxia. Mean values were compared with the control group by the Newman-Keuls multiple-range test. If the two-way ANOVA indicated a significant effect of hypoxia, then one-way ANOVA with repeated measures was performed on each group. Differences between the normoxic and hypoxic values were determined by the Newman-Keuls multiple-range test. Vascular resistance values were logarithmically transformed to improve normality. To test the effect of L-NAME infusion during normoxia, pre- and postinfusion data were compared by paired t-test. The level of significance was set at P < 0.05 in all tests. Data are presented as means ± SE.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cerebral hemodynamics at 0.6 gestation. Administration of 20 mg/kg of L-NAME to fetuses at 93 days of gestation produced an increase in mean arterial pressure and decreases in arterial pH and O₂ saturation (Table 1). Arterial hemoglobin concentration increased, and there was no change in CaO₂. Forebrain CBF decreased by 20 ± 7%, and CVR increased by 67 ± 16%.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Mean arterial pressure vs. arterial O2 content (CaO2) during normoxia and hypoxia in control and N-nitro-L-arginine methyl ester (L-NAME)-treated groups of fetal sheep at 0.6 and 0.9 gestation. Values in the L-NAME groups were significantly greater than in the age-matched control groups at each level of CaO2.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Forebrain cerebral blood flow vs. CaO2 during normoxia and hypoxia in control and L-NAME (20 mg/kg)-treated fetal sheep at 0.6 gestation.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Percent increase in regional cerebral blood flow during hypoxia in control and L-NAME-treated fetal sheep at 0.6 gestation. There were no significant differences between groups.

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Forebrain cerebrovascular resistance vs. CaO2 during normoxia and hypoxia in control and L-NAME (20 mg/kg)-treated fetal sheep at 0.6 gestation. *P < 0.05 from control group. Inset bar graph shows no difference between groups in resistance during hypoxia expressed as a ratio of normoxic values.

Cerebral hemodynamics at 0.9 gestation. Administration of 20 mg/kg of L-NAME to 133 day fetuses decreased CBF in six of eight fetuses, but the change was not significant. Hence, we also studied a higher dose of 60 mg/kg, which did significantly reduce CBF by 23 ± 9% and increased CVR by 93 ± 31%. Both doses increased mean arterial pressure to an equivalent extent, and both doses increased CVR (Table 3). Although L-NAME increased hemoglobin concentration and acidosis was present, there were no changes in Pa_CO2, Pa_O2, O₂ saturation, or CaO₂ after L-NAME administration.

View larger version (21K): [in this window] [in a new window]   Fig. 5.   Forebrain cerebral blood flow vs. CaO2 during normoxia and 2 levels of hypoxia in control, 20 mg/kg L-NAME, and 60 mg/kg L-NAME groups at 0.9 gestation. *P < 0.05 from control group at corresponding level of hypoxia.

View larger version (34K): [in this window] [in a new window]   Fig. 6.   Percent increase in regional cerebral blood flow during the most severe level of hypoxia in the control and 60 mg/kg L-NAME groups at 0.9 gestation. Data from 20 mg/kg group were omitted for clarity. There were no differences among the 3 groups. There were also no differences at the more moderate level of hypoxia.

View larger version (30K): [in this window] [in a new window]   Fig. 7.   Forebrain cerebrovascular resistance vs. CaO2 during normoxia and 2 levels of hypoxia in control, 20 mg/kg L-NAME, and 60 mg/kg L-NAME groups at 0.9 gestation. *P < 0.05 from control group at the corresponding level of oxygenation. Inset bar graph shows resistance during moderate hypoxia (Hypoxia 1) and severe hypoxia (Hypoxia 2) expressed as a ratio of the normoxic value. The ratios were similar in the 3 groups.

Peripheral hemodynamics at 0.9 gestation. After administration of 20 mg/kg L-NAME, blood flow to placenta did not change (Table 5), although calculated placental vascular significantly increased by 39% due to the increase in mean arterial pressure (Fig. 8). During hypoxia, there were no significant changes in placental blood flow in either group. However, placental vascular resistance in the L-NAME group increased an additional 15% during hypoxia and was greater than that in the control group.

View larger version (42K): [in this window] [in a new window]   Fig. 8.   Peripheral vascular resistance in various organs at 0.9 gestation in the control group (open bars) and 20 mg/kg L-NAME group (solid bars) at the first baseline (Base1), second baseline (Base2; measured after saline infusion in the control group and after L-NAME infusion in the L-NAME group), moderate hypoxia (Hyp1), and severe hypoxia (Hyp2). *P < 0.05 from Base2 within same group. +P < 0.05 from control group at same level of oxygenation.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The present study demonstrates that administration of the NOS inhibitor L-NAME reduces baseline CBF compared with the predrug baseline value but does not reduce the percent increase in CBF during hypoxia in fetal sheep at either 0.6 or 0.9 gestation. These results were not confounded by decreases in cerebral O₂ consumption, Pa_CO2, or mean arterial pressure during hypoxia. Because the percent increases in CBF and the percent decreases in CVR were not different between control and L-NAME groups, we conclude that NOS inhibition does not impair cerebrovascular responsivity to hypoxia at either gestational age.

Others studying CBF during isocapnic hypoxia at 0.9 gestation (26, 37) and during hypercapnic hypoxia at 0.8 gestation (34) suggested that NO does contribute to vasodilation during hypoxia. However, this conclusion was based on a comparison of CBF during hypoxia in control fetuses with CBF during hypoxia after NOS inhibition. In two of these studies (26, 34), the NOS inhibitor was administered after the onset of hypoxia, thereby making it difficult to ascertain the increase in flow from a normoxic baseline after NOS inhibition. Moreover, in adult sheep, NOS inhibition during hypoxia can reduce CBF sufficiently to reduce cerebral O₂ uptake (19). Thus it is possible that administration of an NOS inhibitor during hypoxia will have a different metabolic effect than pretreatment. Our results with L-NAME pretreatment in 133 day fetuses are consistent with the previous fetal studies in that CBF after L-NAME was less than CBF without L-NAME during hypoxia. However, our interpretation of these data is that although NO exerts a tonic vasodilatory influence during both normoxia and hypoxia, NO is not a major mediator of vasodilatory reactivity to hypoxia, because the percent changes in CBF and CVR are unaltered by L-NAME.

At 0.9 gestation, the increase in CBF during hypoxia is sufficient to compensate for the decrease in CaO₂ and maintain bulk O₂ transport (33). However, at 0.6 gestation the increase in CBF is inadequate to maintain O₂ transport (10). We postulated that maturation of neuronal NOS was responsible for maturation of hypoxic vascular reactivity because 1) NOS catalytic activity increases threefold in cerebral cortex between 0.6 and 0.9 gestation (29) and 2) the transition from diffuse expression of neuronal NOS in the cortical neuropil to selective expression in distinct cell bodies occurs gradually between 0.4 and 0.9 gestation (28). Because L-NAME at a dose as high as 60 mg/kg failed to attenuate the percent increase in CBF and decrease in CVR during hypoxia at 0.9 gestation, our data do not support the hypothesis that maturation of NOS is responsible for maturation of hypoxic vascular reactivity. Because we found that this dose of L-NAME inhibited cortical NOS catalytic activity by 89% in fetal sheep (29), L-NAME should have been delivered to the fetal brain in an amount sufficient to affect vascular reactivity.

Although studies in postnatal rats and pigs suggest a role of NO in hypoxic vasodilation (3, 18, 39), others have failed to observe an effect of NOS inhibitors in piglets (25). In adult rat and dog, the percent increase in CBF during hypoxia was not reduced by NOS inhibitors (21, 27, 31). Thus our results in fetal sheep agree with those in adult rats and dogs. However, we cannot exclude regional differences in a modulatory role of NO, because Hudetz et al. (16) described reduced capillary red blood cell velocity during hypoxia in subsurface capillaries of adult rat cortex by 7-nitroindazole, a neuronal NOS inhibitor.

Administration of L-NAME at 0.6 gestation decreased basal CBF and increased mean arterial pressure. Thus NO appears to exert a tonic vasodilatory influence in both cerebral and peripheral vessels at this early age. Endothelial NOS is well expressed at this age in cerebral microvessels (28) and may contribute to tonic NO production together with neuronal NOS.

The increase in mean arterial pressure after L-NAME administration at 0.9 gestation was associated with an increase in vascular resistance to all major organs except kidney. Thus most of the major fetal organs as well as placenta contribute to the increase in arterial pressure after NOS inhibition. The lack of a significant increase in renal vascular resistance may be related to inadequate statistical power with a sample size of eight or to the relatively low renal blood flow in the fetus. The increase in vascular resistance in skeletal muscle and skin is consistent with an increase in femoral arterial vascular resistance after L-NAME administration reported by Green et al. (11). Part of the increase in peripheral vascular resistance seen in some organs after L-NAME inhibition may be the result of autoregulation to hypertension and not necessarily to inhibition of tonic NO production in a particular bed. However, in the case of stomach and small intestine, L-NAME caused a decrease in blood flow, thereby implying a contribution of tonic NO production. These results agree with those of Fan et al. (7, 8) who reported a decrease in blood flow to stomach and small intestine in fetal sheep at 0.6 and 0.9 gestation.

In the case of the placenta, NO is believed to exert a tonic vasodilatory effect (6, 36). The 39% increase in placental vascular resistance that we observed after L-NAME administration is comparable to the 51% increase reported by Chang et al. (5) in fetal sheep and is consistent with a considerable tonic influence of NO in the umbilical circulation. During severe hypoxia, we observed an additional 15% increase in placental vascular resistance in the L-NAME group, whereas there was no significant change in the control group. These results suggest that NO may act to attenuate the vasoconstrictor effects of other substances released during hypoxia (e.g., angiotensin).

In addition to placenta, L-NAME administration increased vascular resistance in skeletal muscle, skin, stomach, and intestines. Vascular resistance in these beds increased further during severe hypoxia in the L-NAME group but was not significantly changed in the control group. The lack of a substantial change in blood flow or vascular resistance in these beds in the control group when arterial O₂ content was reduced to ~2 mmol/l is consistent with a threshold of arterial O₂ content <1.5 mmol/l required for marked peripheral vasoconstriction (30). Thus, just above this threshold, NOS inhibition potentiated hypoxic vasoconstriction in many of the fetal peripheral organs. This potentiation may be attributed to 1) loss of endothelial-derived NO counteracting sympathetically and hormonally mediated vasoconstriction and 2) loss of neuronally derived NO attenuating sympathetic nerve outflow at the spinal or brain stem level (17, 23, 40). Thus NO plays a modulatory role in the peripheral vascular response to hypoxia in fetal sheep at 0.9 gestation.

L-NAME reduces the increase in blood flow in adrenal medulla during splanchnic nerve stimulation (4). Although we did not separate blood flow to adrenal medulla and cortex, the attenuated blood flow response of the entire adrenal gland to hypoxia after L-NAME administration may be related to loss of NO-dependent dilation in medulla during sympathoadrenal activation.

In the coronary vascular bed, L-NAME administration did not significantly reduce blood flow to the left ventricular wall, although coronary vascular resistance was increased. An increase in myocardial oxygen consumption secondary to an increase in afterload may have mitigated the reduction in coronary blood flow after L-NAME administration. During hypoxia, the increase in left ventricular blood flow was attenuated by L-NAME despite a greater mean arterial blood pressure. This attenuated coronary flow response is similar to that reported by Reller et al. (32). Thus NO makes a significant contribution to coronary vasodilation during hypoxia in fetal sheep at 0.9 gestation.

Perspectives

In brain, NOS inhibition reduced CBF during normoxia at both 0.6 and 0.9 gestation but did not reduce the percent increase in CBF during hypoxia. Our interpretation of these results is that NO tonically inhibits cerebrovascular tone during normoxia and hypoxia but does not mediate hypoxic cerebral vasodilation at either age. Because the percent increase in CBF during hypoxia is greater at 0.9 gestation than 0.6 gestation and because NOS catalytic activity in cortical homogenates increases threefold between these ages, we anticipated that NO might be responsible for the more robust hypoxic vasodilation at 0.9 gestation. However, our results failed to support developmental differences in the NOS system as the mechanism for developmental differences in hypoxic vasodilation. Thus maturation of other pathways presumably contributes to the development of the cerebrovascular response to hypoxia in the last trimester of fetal sheep. These vasodilatory pathways potentially could act to limit brain injury during intrauterine oxygen deprivation or difficult labor and delivery. However, baseline levels of CBF (2) and cerebral oxygen consumption (1) in human newborns are much lower than adult values and more closely match those of fetal sheep at 0.6 gestation than at 0.9 gestation. Thus, by inference, the CBF response to hypoxia may not be fully developed at birth in humans, and episodes of intrauterine hypoxia could lead to enhanced tissue hypoxia in the near-term human fetal brain.