Main

Oxygen consumption is considerably lower in the fetus than in the infant because little metabolic activity is related to breathing, and none to thermogenesis. The main substrates of oxidative metabolism in the fetus are glucose, lactate, and amino acids. FFA and keto acids are not important sources of energy for the fetus, whereas they are postnatally(1). In the fetus, oxygen and metabolic substrates are supplied exclusively by placental uptake.

Although the effects of reducing oxygen supply to the fetus have been studied extensively, little information is available regarding the ability of the fetus to increase oxygen supply in the face of increased oxygen requirements. Lorijn and Longo(2) showed that norepinephrine infusion increases cardiac output and oxygen consumption in the fetal lamb. To determine whether fetal lambs are able to increase oxygen delivery, we elected to examine the effects of infusion of isoproterenol on β-adrenergic stimulation in fetal lambs. The choice of isoproterenol was based on studies of brown fat, which showed that β-adrenergic stimulation is the predominant stimulus to increasing metabolism. We examined its effects on fetal oxygen consumption, umbilical blood flow, oxygen delivery, and placental glucose and lactate uptake. The concentration of circulating catecholamines increases significantly during fetal stress and during labor(3,4). Because this catecholamine release may further aggravate fetal hypoxia or asphyxia by increasing fetal oxygen consumption, this subject is of clinical as well as physiologic interest.

METHODS

Animal preparation. Nine fetal sheep with gestational ages of 125-138 d (term is approximately 145 d) were studied. After fasting for 24 h, the ewes were given 0.3 mg of buprenorphine and 750-1000 mg of ketamine intramuscularly. A polyvinyl catheter (ID 0.05 in) was inserted into the pedal branch of a maternal femoral vein and advanced to the inferior vena cava, and a continuous infusion of 0.9% saline (500-1000 mL) was administered throughout the surgical procedure. Ketamine HCl (200 mg) was injected i.v. every 10 min to maintain sedation. Using local anesthesia with 2% lidocaine HCl, we opened the ewe's abdomen in the midline and exposed the fetal hindlimbs through a uterine incision. Anesthesia for the placement of all fetal catheters was provided locally with 0.5% lidocaine HCl. Polyvinyl catheters (ID 0.03 in) were inserted into the pedal artery and pedal vein and advanced into the femoral artery and inferior vena cava, respectively. A 3.5-F multiple-side-hole catheter was inserted via a cotyledonary vein and advanced into one of the major umbilical veins.

The fetal hindlimbs were then extracted, and the common umbilical artery was isolated through a left retroperitoneal approach as previously described(5). The diameter of the common umbilical artery was measured, and a precalibrated ultrasonic flow transducer of appropriate size (5-6 mm) was placed around it. The fetal and uterine incisions were closed, and amniotic fluid lost during surgery was replaced with warm saline. An amniotic catheter was inserted to lie adjacent to the fetal trunk. All catheters were filled with heparin (1000 IU/mL) and plugged. Together with the transducer cable of the flow probe, they were then directed subcutaneously to the ewe's flank, exteriorized, and stored in a pouch. On the day of surgery and each day afterward, the ewe received 2 million IU of penicillin G and 100 mg of gentamicin, one half i.v. and one half into the amniotic cavity.

Animal husbandry and the study design followed the guidelines of the National Institutes of Health and were approved by the Committee on Animal Research of the University of California, San Francisco.

Study design. Studies were performed while the ewe stood in a stall, with free access to water and food 2-5 d after surgery. For 20 min before the isoproterenol infusion, we recorded baseline values of mean umbilical blood flow, as well as fetal heart rate and pressures in the fetal descending aorta and peripheral umbilical vein. We measured arterial blood gases and pH, oxygen saturation, and glucose, lactate, and Hb concentrations in umbilical venous and aortic blood samples.

After the control period, we started a continuous infusion of isoproterenol into the inferior vena cava. To assess a possible dose dependency of isoproterenol-induced changes, we progressively increased the infusion rate after 10 min. At the end of each 10-min period, we repeated all measurements.

In preliminary studies, we observed that in some fetal lambs heart rate increased progressively as infusion rate was raised to approximately 5 µg/min/kg estimated fetal weight. We therefore infused isoproterenol in concentrations up to 5 µg/kg/min of estimated fetal weight. Each study lasted 40-50 min and consisted of the control period and of the two consecutive 10-min periods of infusion.

At the end of the experiment, we killed the ewe and the fetus by a lethal infusion of pentobarbital. The fetus was weighed so that the actual isoproterenol dosage in relation to body weight could be calculated.

Methods of measurement. Umbilical blood flow was measured with Transonics ultrasonic transducers and blood-flow meter (Transonics Systems, Ithaca, NY). Pressures in the fetal descending aorta, umbilical vein, and amniotic cavity were measured with Statham P23Db pressure transducers (Statham Instruments, Oxnard, CA). Amniotic pressure was used as the zero pressure reference. Fetal heart rate was measured with a cardiotachometer triggered by the arterial pulse pressure. Mean blood flow, pressures, and fetal heart rate were recorded continuously on a Gould direct-writing oscillographic recorder (Gould Inc., Cleveland, OH).

Blood gases and pH were measured by a Corning 175 Blood Gas Analyzer (Medfield, MA). Fetal blood gases were run at 39°C. Hb concentration and oxygen saturation were measured photometrically (Hemoximeter OSM 2, Radiometer, Copenhagen, Denmark). Blood lactate and glucose concentrations were measured by YSI model 23A glucose analyzer and YSI model 23L lactate analyzer, respectively (Yellow Springs Instruments, Yellow Springs, OH).

Oxygen content was determined according to this formula: oxygen content (mL/dL) = Hb (g/dL) × oxygen saturation × 1.36/100. Oxygen delivery (mL/min/kg) was calculated as the product of the umbilical venous oxygen content (mL/dL) and the umbilical blood flow (dL/min/kg fetal weight). Oxygen consumption (mL/min/kg) was calculated as the product of umbilical blood flow (dL/min/kg) and AVD (mL/dL) between umbilical venous and aortic oxygen content. Oxygen extraction ratio was calculated as oxygen consumption divided by oxygen delivery. Umbilical-placental vascular resistance (mm Hg/mL/min/kg) was calculated as the mean pressure difference between the descending aorta and umbilical vein (mm Hg) divided by umbilical-placental blood flow (mL/min/kg), assuming that proximal umbilical arterial blood pressure was equal to fetal aortic blood pressure.

Lactate placental uptake (µmol/min/kg) was calculated as the product of umbilical blood flow (L/min/kg) and AVD between umbilical venous and aortic lactate concentrations (µmol/L). Similarly, glucose placental uptake (µmol/min/kg) was calculated as the product of umbilical blood flow (L/min/kg) and AVD between umbilical venous and aortic glucose concentrations (µmol/L).

Analysis of data. We analyzed the data during the control period and in two ranges of isoproterenol doses: 0-1 and 1-5 µg/kg/min. Data are summarized as mean ± SD. Comparison between the three experimental periods was made by one-way analysis of variance for repeated measurements. Multiple comparison testing was performed only if the analysis of variance first rejected a multisample hypothesis of equal means with p < 0.05. Differences between groups were then analyzed with the Student-Newman-Keuls multiple comparison F test.

We used a simple regression test to examine the relationship between umbilical blood flow and umbilical venous oxygen content. We considered differences significant at p < 0.05.

RESULTS

Hemodynamic and oxygen metabolic changes induced by isoproterenol are summarized in Table 1. Isoproterenol increased heart rate and umbilical-placental blood flow and decreased umbilical-placental vascular resistance. The maximal effect was already achieved at the lowest dose administered. Higher doses of isoproterenol did not have any further significant influence on these variables. Mean arterial pressure decreased slightly at the lower dose but returned to baseline with higher doses. Oxygen consumption increased with the lowest infusion dose by 1.44 mL/kg/min, but it was not significantly higher at higher infusion rates. Oxygen delivery did not change significantly despite the increased umbilical blood flow because of a parallel fall in umbilical venous oxygen content. Umbilical venous oxygen saturation fell progressively with increasing isoproterenol dosage. Umbilical venous content correlated negatively with umbilical blood flow (r = 0.579, slope = -0.023 ± 0.0037, p < 0.0001 in 75 measurements) (Fig. 1). The oxygen extraction ratio increased progressively from 0.41 ± 0.17% during the control period to 0.48 ± 0.10%. The increased oxygen extraction occurred despite the lower umbilical venous oxygen content, and the femoral arterial oxygen content fell. With progression of the study and increasing infusion rate of isoproterenol, the fetuses became hypoxemic and acidotic.

Table 1 Effect of isoproterenol on fetal hemodynamics and oxygenation
Full size table
Figure 1
figure 1

Umbilical venous oxygen content vs umbilical blood flow. There is an inverse relationship between umbilical-placental blood flow and umbilical venous oxygen content.

Full size image

Both umbilical venous and aortic blood glucose concentrations increased progressively during isoproterenol infusion (Table 2). The blood glucose AVD across the placenta fell at higher isoproterenol infusion rates, and glucose uptake fell, but neither of these changes was statistically significant. Umbilical venous and aortic blood concentrations increased nearly 4-fold during the isoproterenol infusion (Table 2). Lactate delivery to the fetus through the umbilical vein increased markedly. However, lactate AVD across the placenta did not change, so the increased delivery was related to the high femoral arterial lactate concentration.

Table 2 Effect of isoproterenol on fetal glucose and lactate metabolism
Full size table

DISCUSSION

Numerous studies have shown that fetal oxygen consumption falls when oxygen delivery to the fetus is reduced by hypoxemia or a reduction in umbilical-placental blood flow(6,7). Little information is available, however, on the ability of the fetus to increase its oxygen consumption or on the potential for increasing oxygen delivery in the face of increased demand. Lorijn and Longo(2) reported that oxygen consumption in fetal lambs was raised by i.v. i.v. infusion of norepinephrine into the fetus. Based on the studies of isolated fetal adipocytes by Fain et al.(8), it appears that β-adrenoreceptor stimulation induces an increase in their metabolism and increases oxygen consumption.

Autonomic receptor function precedes demonstration of adrenergic innervation. Although sympathetic innervation exists in the fetal sheep, it remains immature(9). Therefore, circulating catecholamines presumably play a predominant role over neuronal release in the late-gestation fetus. β1-Adrenergic receptors have been demonstrated during midgestation in several species using the ligand [3H]dihydroalprenolol. To stimulate these β-adrenergic receptors in the fetus, we infused isoproterenol. Oxygen consumption increased by almost 25% with an infusion rate of ≤1 µg/kg/min and did not change significantly with increasing infusion rates. This increase in fetal oxygen consumption is probably related to stimulation of brown fat metabolism as occurs in nonshivering thermogenesis after birth. This lipolytic effect of adrenergic stimulation on fat metabolism in the fetal lamb was substantiated by several authors who showed an increase in FFA induced by graded infusions of norepinephrine and epinephrine(10,11). In addition, stimulation of fetal breathing movements produced by the isoproterenol infusion may have participated in the increase in oxygen consumption(12).

Postnatally, increased oxygen consumption is associated with attempts to increase oxygen delivery by respiratory stimulation and a rise in cardiac output. β-Adrenoreceptor stimulation does increase ventricular output in the fetal lamb(13), predominantly by increasing heart rate. In our study, isoproterenol infusion caused a 15-23% rise in umbilical-placental blood flow; however, this was not associated with an increase in oxygen delivery to the fetus. Oxygen delivery, the product of umbilical venous oxygen content and umbilical blood flow, did not change significantly throughout the study because umbilical venous oxygen content fell.

The inverse relationship between umbilical-placental blood flow and umbilical venous oxygen content may be explained by limitation of oxygen uptake because transit time in the umbilical-placental circulation is decreased. Figure 1 shows that the higher the umbilical blood flow and shorter the transit time, the lower the umbilical venous oxygenation.

This relationship is also noted when umbilical blood flow is changed by other means. During acute fetal blood volume reduction by hemorrhage, umbilical venous oxygen content and saturation increase(14). When umbilical blood flow was reduced progressively by compression of the umbilical cord, umbilical venous PO2 and oxygen content showed a small, but not significant, increase. Reducing umbilical flow by hemorrhage or cord compression increased the umbilical venous-to-descending aortic oxygen content difference. When umbilical blood flow was reduced to 50% of control by cord compression, AVD increased from 3.1 mL/dL during the control period to 4.75 mL/dL(7).

In contrast, when uterine blood flow is reduced acutely in sheep(15), oxygenation of umbilical venous blood is reduced significantly. Umbilical venous-to-arterial oxygen content difference falls markedly; this is related to the fact that although actual umbilical blood flow does not change significantly, the ratio of umbilical blood flow to uterine blood flow increases greatly, and this results in a fall in umbilical venous oxygen content with a reduction of oxygen delivery to the fetus. During isoproterenol infusion, the umbilical venous to arterial oxygen content difference also fell significantly by 13% from 4.42 ± 0.98 mL/dL during the control period to 3.86 ± 0.79 mL/dL at the higher infusion rates of isoproterenol.

Another possible explanation for the decrease in umbilical venous oxygen content is the opening of umbilical arteriovenous shunts, which allows for bypassing the site of gas exchange in the placenta. This is unlikely, because there is no evidence to support the existence of such shunts.

Whatever the mechanisms for this decrease in the maintenance of umbilical venous oxygen content with increase in umbilical blood flow, the result is to limit the fetal lamb's ability to increase its oxygen delivery by increasing cardiac output and/or umbilical-placental blood flow. This limitation may not be present in all species. In the sheep, the placenta is syndesmochorial, so that several layers separate umbilical and uterine blood; therefore, diffusion of oxygen could be relatively slow. In other species, with placentae that are hemochorial, as in the human, or hemoendothelial, as in the rabbit, the separation between the two circulations is considerably thinner, and there may be less limitation in the ability for umbilical venous oxygen content to be maintained with increasing flows.

The increase in umbilical blood flow during isoproterenol infusion is associated with a decrease in umbilical vascular resistance by 28%, with no change in mean arterial pressure. The presence of β-adrenoreceptors in the umbilical circulation has been reported by Parer(16), who suggested that increased plasma catecholamine concentrations were responsible for the decrease in umbilical vascular resistance during fetal hypoxemia. The decrease in umbilical resistance is due to a direct local vasorelaxant effect of isoproterenol on the umbilical-placental circulation, but the exact site of this effect has not been determined. Moreover, further vasodilation of the umbilical circulation may have resulted from nitric oxide release in response to the increased shear rate induced by the increased flow itself.

Fetal blood glucose concentrations increased by 70% of control during isoproterenol infusion. Under normal conditions, fetal glucose requirements are supplied largely by transfer across the placenta from the maternal circulation. The increased glucose concentrations are not due to increased diffusion across the placenta, because there was no significant increase in umbilical venous-to-arterial glucose difference, or to the increased umbilical blood flow. The increase in blood glucose concentrations therefore must be due to glucose delivered by fetal mechanisms, most likely by release from the liver. Under normal circumstances, the fetal liver synthesis glycogen at a slow rate; it has been estimated that in fetal lambs after 125 d gestation, glycogen synthesis is only approximately 5 mg/g liver per day(17). Little glucose is produced by the liver normally, but during fetal hypoxemia, significant amounts of glucose are released by the liver(18). This production of glucose is almost exclusively from glycogenolysis, because the fetal liver has little capability for gluconeogenesis(19).

The increased glucose concentrations associated with hypoxemia almost certainly are due to the increased blood catecholamine concentrations and sympathetic nerve stimulation resulting from hypoxemia. We have shown that infusion of epinephrine or norepinephrine into the umbilical vein of fetal lambs results in a marked increase of glucose release by the liver (Rudolph C, Roman C, Rudolph AM, unpublished observations). The current studies showing increased fetal production of glucose during the isoproterenol infusion indicate that hepatic glycogenolysis is controlled, at least in part, by β-adrenoreceptor stimulation.

Lactate is an important substrate for fetal metabolism(20), and isoproterenol resulted in a large increase in blood lactate concentrations. This isoproterenol-induced increase in lactate, which has already been reported in the fetal sheep(10), is not the result of enhanced placental production of lactate, because placental AVD did not change significantly during isoproterenol infusion. Therefore, it must result from increased production of lactate, decreased utilization of lactate, or both. Infusion of glucose into lamb fetuses has been reported to reduce fetal blood pH values(21). It is possible that the marked increase in fetal blood glucose concentrations enhances use of glucose for metabolism. With the limitation of oxygen supply, anaerobic glycolysis is increased, resulting in lactate accumulation with the ensuing fall in blood pH values. Furthermore, β-adrenoreceptor stimulation has a direct effect on lactate metabolism. In the adult rat, isoproterenol activates glucose conversion to lactate and suppresses markedly the conversion to the metabolites other than lactates, and it stimulates hepatic gluconeogenesis from lactate(22). In the fetus, the limited capability for gluconeogenesis may have further accentuated lactate accumulation in the blood.

With lower infusion rates of isoproterenol, or in the first period of infusion, oxygen consumption increases and lactate concentrations do not rise significantly. However, with higher doses, or after more prolonged infusion, oxygen consumption is not significantly different from control levels, but lactate concentrations increase.

In conclusion, we have shown that isoproterenol infusion into fetal lambs increases fetal oxygen consumption. It also raises umbilical blood flow but does not increase oxygen delivery to the fetus because umbilical venous oxygen content falls. Fetal blood glucose and lactate concentrations increase during isoproterenol administration. Glucose is probably released from the liver, and it is proposed that lactate concentrations rise because glycolysis is increased, but oxygen supply, and hence oxidative metabolic pathways, are limited.