INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

INCREASED PRODUCTION OF CATECHOLAMINES by the sympathetic nervous system and/or adrenal medulla is characteristic of physiological situations such as chronic hypoxemia and plays an important role in fetal adaptation to reduced nutrient and oxygen availability (12, 13, 18, 25, 29). Recent investigations show that prolonged increases in epinephrine (Epi) and norepinephrine (NE) cause marked retardation of tissue growth in well-nourished fetal sheep (2), so catecholamines could be largely responsible for the pattern of growth retardation observed during chronic hypoxemia (3, 25). Prolonged inhibition of insulin secretion (2, 17) and inhibitory effects on the action of insulin-like growth factor I (IGF-I) (17) has been observed during infusions of Epi and NE into normoxemic fetal sheep and proposed to explain these adverse effects on growth. Prolonged hyperinsulinemia has been shown to modify some of the responses of fetal sheep to short-term hypoxemia (29). Insulin and glucagon were also infused during studies of NE actions on protein metabolism in fetal sheep (22) to offset the pancreatic hormone changes usually associated with NE infusion, but the effects of this were not assessed. The consequences of preventing or correcting the profound hypoinsulinemia that occurs in fetal sheep when catecholamine concentrations remain high for prolonged periods (2), therefore, remain unknown.

To examine the hypothesis that prolonged inhibition of insulin release may be a primary reason for adverse effects of Epi and NE on metabolism and growth in the fetus, we administered insulin to chronically cannulated fetal sheep to restore blood insulin concentration to the normal range for periods of up to 7 days during prolonged infusions of either Epi or NE at the same rate as used previously (2). Effects on blood pressure, heart rate, blood gases, and metabolite and hormone concentrations in the fetus were monitored throughout these prolonged infusions and compared with the changes observed in fetal sheep infused with Epi or NE alone (2), as well as with those in control twin fetuses subject to the same maternal environment. Effects on fetal development were determined at termination of the infusions.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animal preparation. All surgical procedures and experimental protocols used in the studies were carried out in accordance with a project licence approved by the UK home office under the terms of the Animals (Scientific Procedures) Act 1986. Ten mule × Suffolk crossbred ewes, of accurately known gestational age mated with Polled Dorset rams and diagnosed as twin-pregnant by ultrasound scanning at 70-90 days gestation, were fasted for 24 h before surgery at 118-120 days gestation (full-term is 145-147 days in these breeds). Anesthesia was induced by intravenous injection of 20 ml of 5% thiopentone sodium (Intraval sodium) and maintained with halothane (1.5% in oxygen) after tracheal intubation. The surgical insertion of fetal and maternal vascular cannulas, administration of antibiotics, and care of the sheep during the postoperative period were carried out as described earlier (2). To provide an estimate of fetal size, the length of both hind feet (metatarsal) was measured against a ruler during surgery. After surgery, ewes were housed in metabolism cages and offered water and hay ad libitum. Beginning on the next day, ewes were also offered 300 g of a barley-based concentrate mixture twice daily to provide sufficient metabolizable energy and protein for maintenance and pregnancy during the final month of gestation. Four days after surgery, daily blood sampling of the ewe and fetuses before the morning feed was begun and was continued until termination of each study to monitor blood gases and pH (1 ml) and to determine plasma metabolite and hormone concentrations (4 ml) during the experiments. Samples were collected into heparinized syringes and placed on ice immediately. Plasma for determination of metabolite and hormone concentrations was separated by centrifugation at 4°C and stored at 20°C until required for analysis. Arterial pressure and heart rate were monitored continuously in both fetuses and recorded as described previously (2). Mean values over each hour, calculated after removal of out-of-range observations are used in results presented here.

Experimental procedures. To permit evaluation of the effect of insulin on responses to Epi or NE, rates of Epi or NE infusion used and the changes in them during these experiments were the same as those used in our previous studies in which Epi and NE were infused alone (2). In the first experiment, Epi was infused into seven fetuses (6 twins and 1 single) for 3 days at a rate of 1 µg/min. The rate of Epi infusion was then increased to 2 µg/min for the rest of the study. At this time, infusion of insulin (0.5 mU/min) was begun via a separate lumen of the femoral vein catheter. This rate of insulin infusion, which was in the range that restored plasma insulin concentration and growth to normal in pancreatectomized fetal sheep (8), was continued for 6 days. The insulin concentration of the infusion solution was then doubled, and infusion continued at twice the initial rate for a further 2 days to assess the concentration dependence of responses to insulin. Epi infusion was maintained for another day before termination. Control fetuses were infused with similar volumes of acidified saline diluent throughout. Previous investigations (2) provided no evidence that prolonged infusion of either Epi or NE into one fetus at the rates used here has significant effects on any of the measured parameters in the diluent-infused control fetus.

Because plasma concentrations of NE are much greater than those of Epi in fetal sheep during prolonged stressful conditions (13, 18), the second study was designed primarily to investigate how cardiovascular, metabolic, and endocrine responses to NE infusion may be modified by prevention of hypoinsulinemia. It was also designed to examine in greater detail how the catecholamine-infused fetus responds to cessation of insulin infusion, a response only studied on the last day of Epi infusion during the first experiment. A crossover design was used, so both fetuses from each twin-pregnant ewe could be infused with NE for a period of 7 days, insulin also being infused for part of this period. After collection of initial preinfusion control observations, one twin fetus was infused intravenously with NE at a rate of 2 µg/min for 4 days and then at twice this rate for a further 3 days. This fetus was also infused with insulin at 0.5 or 0.75 mU/min (0.2 ± 0.03 mU · min¹ · kg¹ estimated fetal wt) via a separate lumen of the venous catheter for the first 2 days of NE infusion. The insulin infusion was then stopped and switched to the control twin for the next 2 days while the NE-infused fetus received saline diluent. Insulin infusion to the NE-infused fetus was resumed at twice the initial rate at the same time as the NE infusion rate was doubled. Insulin infusion continued at this rate for 2 days but was stopped again for the last day of NE infusion. The other fetus, infused with diluent except on days 3 and 4 when it received insulin, served as a control. During the second week of the experiment, the same sequence of NE and insulin infusions was administered to the control fetus from the first phase of the study while the other fetus now served as control. The experiment was terminated 24 h after cessation of NE infusion to the second fetus. Six fetuses were infused with NE and insulin, and each provided control observations for its twin during the other week of study. Because all six fetuses in this study were administered NE and insulin, effects of the infusions on their growth could only be evaluated by comparisons of their growth with that of saline-infused control fetuses from other studies in our laboratory.

Infusions all were given at the same rate (1.2 ml/h) via a separate lumen of the venous catheter. Changes in infusions were always made after collection of daily blood samples and feeding of the sheep. Infusion solutions of Epi and NE were prepared and infusions carried out as described previously (2). In each experiment, the control twin was infused with diluent (0.3% ascorbic acid in 0.9% sterile saline). The choice of cannulated twin for infusion of catecholamine or diluent was random.

Insulin infusion solutions were prepared fresh daily by dilution from a stock solution (1.0 U/ml) prepared from Neutral Bovine Insulin (100 IU/ml; Hypurin, CP Pharmaceuticals, Wrexham, UK). Fetal plasma from the fetus to be infused, sterilized by passage through a 0.2-µm sterile Puradisc membrane filter (Whatman International, Maidstone, Kent, UK), was added at 0.5% to the acidified 0.9% sterile saline diluent before addition of insulin to minimize insulin adsorption to infusion catheters and syringes. Control twins were infused with this diluent at the same rate (1.2 ml/h) during periods of insulin infusion.

Experiments were terminated by rapid intravenous administration of 20 ml of a barbiturate anesthetic (Euthatal, Rhône Mérieux, Harlow, Essex, UK) to the ewe. The uterus was removed with catheters still attached, so individual fetuses could be identified. Fetuses were removed, toweled dry, and weighed before dissection to recover and weigh the principal organs. When a third uncannulated fetus was found at autopsy, this was treated as other fetuses and used to provide additional control information on organ and carcass tissue development. Before dissection, the length of both fetal hindlimbs was measured to provide an index of fetal limb growth since the time of surgery. Fetal carcasses were frozen and stored at 20°C for later dissection. To obtain a more specific evaluation of carcass tissue development, carcasses were thawed to permit removal of a selection of muscles or muscle groups and bones. After removal of the skin, the crown-rump length of each fetus was measured. Then, as described previously (2), selected individual muscles or muscle groups from the hindlimb and back on one side of the carcass were carefully removed and weighed. These muscles represent ~30% of the skeletal muscle on one side of the carcass. The pelvis, femur, tibia, and metatarsal from one hindlimb were also removed, weighed, and measured.

Analyses. Blood Pa_O2, Pa_CO2, and pH values were determined at 39°C using a Corning 168 blood gas analyzer (Corning Medical and Scientific, Halstead, Essex, UK), whereas blood hemoglobin, O₂ saturation (Sa_O2), and O₂ content were determined using an OSM2 Hemoximeter (Radiometer, Crawley, West Sussex, UK) and hematocrits determined using a microhematocrit centrifuge (Hawksley, Lancing, Sussex, UK). Plasma glucose, lactate and nonesterified fatty acid (NEFA) concentrations were measured using the same spectrophotometric methods as used in earlier studies (2). Also, as in earlier studies (2), plasma insulin was determined by radioimmunoassay using anti-porcine insulin antiserum, a monocomponent ovine insulin standard (Lilly), ¹²⁵I-labeled porcine insulin, and a dextran-coated charcoal separation. Standard curves for bovine insulin could be superimposed on those for ovine insulin. The sensitivity of the assay was 10 pmol/l. Within assays, the coefficient of variation, determined from differences between duplicate determinations, was ±18% within the range 10-60 pmol/l and 10.2% within the range 60-240 pmol/l. The measured insulin concentration of a sample, assayed in duplicate in 26 assays (mean = 68 pmol/l), varied by ±16.7% .

Calculations. To calculate the rate of growth of fetuses during the period of catecholamine infusion, we used an estimate of the initial weight of the fetus at surgery, calculated from measurements of metatarsal length on both hindlimbs made at this time by an equation similar to that of Santucci et al. (28), derived from measurements on saline-infused control fetal sheep in our flock (2). This estimate and fetal body weight at autopsy were then used to calculate the rate of fetal weight gain and estimates of muscle and bone growth rates during catecholamine infusion. These calculations, using equations reported previously (2), depended on assumptions that growth in weight of diluent-infused control fetuses was linear throughout the period of study and was similar in Epi- or NE-infused fetuses before catecholamine infusion. It was also assumed that the proportion of skeletal muscle and bone in the fetal body of control fetuses did not change during the study. The growth rate of uncannulated triplet fetuses could not be determined because their metatarsals were not measured at surgery.

Statistical analysis. Concentrations of lactate, NEFA, and insulin were converted to logarithms before calculation of means or any other statistical procedures, because absolute values are not normally distributed, and changes in concentration are multiplicative. Values reported in the text are geometric means, and logarithmic scales are used for graphical presentation of geometric means and standard errors. General linear model (GLM) repeated-measures ANOVA procedures for unequal subclass numbers and covariance, where necessary, have been used for statistical evaluation of fetal body and tissue weights at termination as well as responses to infusion of insulin with Epi or NE. For the latter, mean values calculated for each individual for defined periods during the infusion were used. The estimated weight of the fetus at surgery or the mean preinfusion control value calculated from values over the 4 days before the start of infusion for each fetus, as appropriate, was included as a covariate in analyses when there was evidence of significant covariance. Pairwise comparisons between estimated marginal means for each treatment group were used for post hoc assessment of differences between groups. Some responses to infusion of Epi or NE together with insulin were compared with responses to infusion of Epi or NE alone at the same rate during studies carried out under similar conditions in our laboratory (2). Similar GLM procedures were used for the analyses. Probabilities <0.05 were regarded as significant. SPSS for Windows (version 7.5, Advanced Statistics 7.51) was used for all calculations.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effects on fetal blood hormone and metabolite concentrations. Infusion of insulin at 0.5 mU/min to the fetus after 3 days of Epi infusion returned plasma insulin concentration from the low value established during the first 3 days to the same range as that of saline-infused control twins, even though the Epi infusion rate was doubled at the same time (Fig. 1). Doubling of the insulin infusion rate after 6 days resulted in a further increase in plasma insulin, but mean plasma insulin concentrations established were still within the normal range for normoglycemic fetal sheep at this stage of gestation. Restoration of normoinsulinemia during Epi infusion significantly decreased (P < 0.001) fetal plasma glucose to 0.9 ± 0.06 mM from the value observed during the first 3 days of Epi infusion (1.4 ± 0.03 mM; Fig. 1). Doubling of the insulin infusion rate decreased plasma glucose further to the control range, despite continued Epi infusion. Because fetal plasma NEFA and lactate responses to Epi infusion were already largely attenuated after 2 days of Epi infusion, there were no significant differences between control and Epi-infused fetuses during the period of insulin infusion (Fig. 2). However, when insulin infusion was stopped after 8 days, fetal plasma concentrations of glucose and lactate increased significantly (P < 0.002). The plasma glucose concentration 24 h later (1.3 ± 0.08 mM) was not significantly different from that during the first 3 days of Epi infusion and was comparable with that of fetal sheep infused continuously with Epi alone over the same period, whereas the lactate concentration was comparable with that observed after the first day of Epi infusion and was much higher than that of fetuses infused continuously with Epi alone for 7 days or more (Fig. 2). Plasma NEFA also increased, but this increase was more variable and was not significant.

View larger version (2625K): [in this window] [in a new window]   Fig. 1.   Plasma insulin and glucose concentrations (means ± SE) in fetal sheep infused with epinephrine (Epi) and insulin (, n = 7; A) or with norepinephrine (NE) and insulin (, n = 6; B) compared with mean values in their control twins infused with saline diluent (, n = 6). Periods of infusion are indicated by horizontal bars at top and bottom. Insulin infusion was switched from NE-infused fetus to control fetus during middle 2 days of infusion (indicated by dotted line in B). Ranges for plasma insulin and glucose concentrations (means ± SE) during prolonged infusion of Epi or NE alone (dotted lines) into similar fetal sheep and their control twins (dashed lines) reported earlier (2) are illustrated in A and B. Full details of infusion rates and calculation of mean values are given in METHODS.

View larger version (2524K): [in this window] [in a new window]   Fig. 2.   Plasma lactate and nonesterified fatty acid (NEFA) concentrations (means ± SE) of fetal sheep infused with Epi and insulin (, n = 7; A) or with NE and insulin (, n = 6; B) compared with concentrations in their saline-infused control twins (, n = 6). Ranges for plasma lactate and NEFA concentrations (means ± SE) in similar fetuses infused with Epi or NE alone (dotted lines) and their control twins (dashed lines) during earlier studies (2) are illustrated in A and B. Periods of catecholamine and insulin infusion are illustrated by horizontal bars on each figure. Full details of infusion rates and calculation of mean values are given in METHODS.

When insulin was infused with NE from the start of a prolonged infusion, plasma insulin concentration was maintained in the same range as that of control fetuses (Fig. 1). Cessation of insulin infusion on the third and fourth days resulted in a rapid decline of plasma insulin to the range observed during infusion of NE alone. Subsequent doubling of infusion rates of both NE and insulin increased plasma insulin to values similar to those observed during infusion of insulin with Epi. When insulin was infused with NE, fetal plasma concentrations of glucose (Fig. 1), lactate, or NEFA (Fig. 2) did not differ significantly from control values. However, plasma glucose, lactate, and NEFA concentrations increased significantly (P < 0.05) when insulin infusion was stopped on the third and fourth days of NE infusion. Resumption of insulin infusion at twice the initial rate, in association with a doubling of the NE infusion rate, resulted in a return of glucose, lactate, and NEFA concentrations to the control range. On the seventh day, the cessation of insulin infusion was again associated with significant increases in plasma glucose (Fig. 1), lactate, and NEFA (Fig. 2) concentrations to high values 24 h later. Cessation of catecholamine infusion at the end of the study was followed by a rapid recovery in endogenous insulin secretion. One hour after switch-off, the geometric mean plasma insulin concentration had increased from 45 to 163 pmol/l in seven fetuses (4 Epi- and 3 NE-infused fetuses) in which this was followed. After cessation of NE infusion, insulin and metabolite concentrations returned to the control range within 24 h (Figs. 1 and 2). Switching of the insulin infusion to control fetuses on the third and fourth days of NE infusion had no significant effect on their plasma glucose (Fig. 1), lactate, or NEFA (Fig. 2) concentrations, even though plasma insulin concentration increased (Fig. 1).

Effects on fetal blood gases and pH. Infusion of Epi increased fetal Pa_O2, Sa_O2, and the O₂ content of fetal arterial blood significantly and decreased Pa_CO2 during the first 3 days of infusion (Table 1). Restoration of normoinsulinemia by insulin infusion appeared to limit the magnitude of these changes (Table 1), but values did not return to the normal range even when the insulin infusion rate was doubled. The infusions had no significant effect on fetal arterial blood pH (results not shown). In the second experiment, infusion of insulin with NE for the first 2 days and then again on the fifth and sixth days, when infusion rates of both NE and insulin were doubled, also limited the magnitude of changes in fetal blood Pa_CO2 and O₂ content compared with responses to NE alone in the earlier study, but Sa_O2 increased markedly. However, cessation of insulin infusion did increase blood Pa_O2, Sa_O2, and O₂ content and reduce Pa_CO2 significantly when NE infusion continued alone at either rate, as illustrated by combined means for the two periods. There were no significant changes in any of these measures during the week when fetuses were infused with diluent (Table 1).

Effects on blood pressure and heart rate. Infusion of Epi alone for 3 days resulted in changes in mean arterial pressure (MAP) and heart rate comparable with those observed during infusion of Epi alone (2); so MAP in the infused fetus was already significantly increased compared with that of control fetuses before the start of insulin infusion (Table 2). Commencement of insulin infusion to the fetus, in association with a doubling in the Epi infusion rate, did not prevent a further increase in MAP like that observed during infusion of Epi alone, but MAP declined more rapidly from the maximum value when insulin was infused with Epi (Fig. 3). In fetuses infused with insulin and Epi, MAP was decreased significantly from the second to the sixth day of infusion in comparison to the mean value during the first 24 h of infusion or that before commencement of insulin infusion (Table 2). By contrast, addition of insulin to the Epi infusion increased fetal heart rate significantly, an increase that was not evident during infusion of Epi alone (Fig. 3). This increase was maintained until the infusion of insulin with Epi ceased (Table 2). Doubling of the rate of insulin infusion on the seventh day did not alter MAP or heart rate further, but cessation of insulin infusion was followed by a significant increase in MAP from 46 ± 2.47 to 49 ± 1.98 mmHg (P < 0.05) and a decrease in heart rate from 190 ± 4.2 to 170 ± 3.9 beats/min (P < 0.001) in the five fetuses for which measurements in all periods were available.

View larger version (36K): [in this window] [in a new window]   Fig. 3.   Changes in hourly mean arterial pressure and heart rate of fetal sheep relative to mean value during 6 h before start of insulin infusion at 0.5 mU/min and an increase (2×) in Epi infusion rate to 2 µg/min (, n = 7) compared with changes observed in fetuses infused with Epi alone (, n = 10) during an earlier study (2). Vertical bars illustrate average ± SE range for hourly mean values. Details of Epi infusion rates used are given in METHODS. , Change.

Infusion of insulin from the start of the NE infusion did not prevent a rapid increase in MAP during the first 2-4 h of infusion, but there was also a large and prolonged increase in heart rate (Fig. 4). The mean heart rate of infused fetuses during the first 48 h of insulin and NE infusion was 189 ± 6.8 beats/min, some 14 ± 4.1 beats/min greater (P < 0.02) than the mean value on the day before commencement of infusion. Comparison of increases in MAP and heart rate of fetuses infused with NE plus insulin during the first 48 h of infusion with those observed in fetuses infused with NE alone at the same rate (Fig. 5) shows that the increase in MAP was less than that in fetuses infused with NE alone, although this difference was not significant. However, the increase in mean heart rate was significantly greater than that during infusion of NE alone. Mean heart rate from 24-48 h of infusion was 14 ± 4.6 beats/min greater in fetuses infused with NE plus insulin than in fetuses infused with NE alone (P < 0.05). Switching of the insulin infusion to the control twin after 48 h resulted in a rapid and significant (P < 0.005) decrease in heart rate of the NE-infused fetuses and a small, but not significant, increase in their MAP over the following 2 days (Fig. 4). There was a small increase in mean heart rate of control fetuses at this time, but this was not consistent nor statistically significant.

View larger version (37K): [in this window] [in a new window]   Fig. 4.   Hourly mean arterial pressure and heart rate in twin fetal sheep infused with NE for 7 days and insulin from 0 to 48 h and from 96 to 144 h of NE infusion (, n = 6) compared with values in same fetuses during preceding or subsequent week when they received saline diluent with insulin added between 48 and 72 h of infusion period (, n = 6). Horizontal bars and vertical arrows indicate periods of infusion and times when NE or insulin infusion rates were doubled. Full details of NE infusion rates used are given in METHODS. Vertical bars illustrate average ± SE ranges for each group.

View larger version (30K): [in this window] [in a new window]   Fig. 5.   Hourly mean arterial pressure and heart rate changes in twin fetal sheep infused with NE + insulin for 48 h (, n = 6) relative to mean values during 3 h before infusion started compared with changes over first 48 h of infusion in similar twin fetal sheep infused with NE alone (, n = 6) during a study reported earlier (2). NE was infused at 2 µg/min during both studies and insulin at 0.5 mU/min. Full details of infusions are given in METHODS. Vertical bars illustrate average ± SE ranges for hourly means over period of observation.

Restoration of insulin infusion to NE-infused fetuses and doubling the rate of both NE and insulin infusions on day 5 was followed acutely by increases in both MAP and heart rate (Fig. 4). There was no significant change in MAP during the 48 h of infusion at these higher rates, but mean heart rate (185 ± 7.1 beats/min) increased significantly (P < 0.02) relative to that in the previous 48 h. Cessation of insulin infusion on the seventh day resulted in a significant increase (P < 0.05) in MAP to a value (48.4 ± 1.55 mmHg) comparable with that observed during the first 24 h of NE infusion and in a steady decrease in heart rate to the control range (Fig. 4). Cessation of NE infusion after 7 days was followed by rapid and significant decreases (P < 0.01) in MAP and heart rate to values initially below the normal range, recovery to values comparable with those of control fetuses taking some 24 h (Fig. 4). Responses of fetuses to the insulin and NE infusions during the first week of the study did not differ from those observed in fetuses infused with insulin and NE during the second week of the study, although there was evidence for a steady developmental increase in MAP and a decline in heart rate of control fetuses over the period of the study.

Effects on fetal growth and development. The infusion of insulin for 8 of the 12 days of Epi infusion had no significant effect on fetal weight at termination nor on the estimated rate of fetal growth during infusion relative to that of cannulated control fetuses (Table 3). Although three of the insulin- and Epi-infused fetuses were triplets, the estimated marginal mean for the rate of weight gain during Epi infusion (53 ± 16.3 g/day; n = 6) was significantly greater (P < 0.02) than that of fetuses infused with Epi alone (19 ± 12.5 g/day; n = 7) in earlier experiments (2) and did not differ significantly from the mean estimated rate of weight gain (86 ± 9.6 g/day; n = 21) for all cannulated control fetuses during the two series of experiments.

Growth of the six fetuses infused with insulin for 4 of the 7 days of NE infusion did not appear to be affected adversely, although no direct comparison with untreated control fetuses within this study was possible. The mean weight of these fetuses at termination was large (4.6 ± 0.21 kg; n = 6) and not different from that of all control fetuses (4.2 ± 0.16 kg; n = 21) from the Epi-infusion experiment and the earlier study (2), but was significantly greater (P < 0.005) than that of fetuses infused with NE alone (3.5 ± 0.21; n = 6). The estimated marginal mean rate of weight gain of these fetuses during NE infusion (Table 3) was comparable with that of control fetuses during those studies.

Infusion of insulin and Epi had no significant influence on the weight of the perirenal fat at termination nor on the weight of muscles and hindlimb bones isolated from the fetal carcass at termination (Table 4). The weight of muscle and bone isolated from fetuses infused with insulin and NE was larger than in the other groups, but tissue weights relative to fetal body weight did not differ among any of the groups (Table 4). During the period of infusion, the estimated rate of weight gain for the selected skeletal muscles from fetuses infused with insulin and Epi was 2.1 ± 0.65 g/day (n = 6) and was 2.1 ± 0.67 g/day (n = 6) for fetuses infused with insulin and NE. These values did not differ significantly from the growth rate of these muscles in cannulated control fetuses [2.6 ± 0.84 g/day (n = 4)]. Estimated growth in weight of the four hindlimb bones during infusion was also unaffected, being 1.4 ± 0.18, 1.2 ± 0.14, and 1.7 ± 0.15 g/day in cannulated controls, insulin and Epi-, or insulin and NE-infused fetuses respectively. Bone diameters were unaffected by the infusions. Proportionate development of the major visceral organs was also largely unaffected (Table 5), but insulin infusion does not appear to overcome adverse effects of Epi on the weight of either the lung or spleen of infused fetuses.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Results of the experiments reported here provide strong support for the hypothesis that inhibition of insulin release is responsible for many of the metabolic changes and adverse effects on growth that occur when plasma concentrations of Epi or NE are increased in fetal sheep for long periods. During prolonged Epi or NE infusion, the high plasma glucose concentrations of fetuses (2) are directly comparable with those of hypoinsulinemic pancreatectomized (7, 9, 10) or streptozotocin-treated fetal sheep (14, 15, 26). Insulin infusion restores plasma glucose concentration to the normal range and prevents adverse effects on growth in the pancreatectomized fetus (8). In the present study, insulin infusion at a similar rate corrected the hypoinsulinemia, restored the plasma glucose concentration to normal, and prevented the inhibition of growth by prolonged Epi or NE infusion. This similarity provides strong evidence that inhibition of insulin secretion by Epi and NE plays an essential role in the restoration of fetal homeostasis in situations in which sympathetic activity is stimulated. The importance of this inhibition of insulin release by Epi and NE to their actions is reinforced by the marked changes in plasma metabolite concentrations that occur when insulin infusion is switched off. Plasma lactate and NEFA increase markedly, and glucose returns to the high values observed when either Epi or NE was infused alone (2).

The magnitude of the changes in MAP and heart rate, which occur when insulin infusion is switched off, suggests counterregulation of the selective actions of insulin on peripheral vascular resistance and the distribution of cardiac output in the fetal sheep (21, 29) could be at least as important in bringing about growth retardation and restoration of homeostasis by the catecholamines as is counterregulation of direct actions on cellular metabolism. Nevertheless, the observation that insulin infusion failed to normalize fetal arterial blood gases completely does indicate that not all actions of the catecholamines on the circulation are completely attenuated by restoration of normoinsulinemia.

Acute increases in plasma glucose, NEFA, and lactate concentrations dominate the metabolic response of the fetal sheep to Epi or NE administration during the first 24 h. However, the sustained hypoinsulinemia and consequent failure of plasma glucose to return to the normal range may be of greater relevance to the growth retardation observed when circulating catecholamine concentrations remain elevated for prolonged periods (2). Glucose use by fetal sheep has not been measured during prolonged Epi or NE administration, but it is likely that glucose consumption would have been reduced by the hypoinsulinemia and decreased maternal-to-fetal glucose concentration gradient. Fetal plasma glucose concentration increases to a similar extent in hypoinsulinemic pancreatectomized fetal sheep (7-10), but glucose use by fetal tissues is reduced in untreated pancreatectomized (7, 10) or streptozotocin-treated fetal sheep (14, 15) and is restored to normal by fetal insulin administration (7). Glucose use in the placenta of Epi- or NE-infused fetal sheep, by contrast, may have increased, because uteroplacental glucose consumption is proportional to fetal glucose concentration and is unaffected by insulin (6, 14). Insulin alters the distribution of glucose use amongst tissues in fetal sheep by increasing cardiac output (21, 24) and selective actions on the vasculature of skeletal muscle (1, 27) as well as through tissue-specific stimulation of glucose transport. Adverse changes in the absolute rate of glucose consumption by insulin-sensitive tissues, such as skeletal muscle, are thus likely to have been greater than any change in the overall rate of fetal glucose use during Epi or NE infusion-induced hypoinsulinemia, whereas glucose consumption via insulin-insensitive pathways in other tissues will likely have increased because of the higher fetal glucose concentration.

Tissue blood flow was not measured, but measurements on fetal sheep have shown that increases in insulinemia result in a reduction in carcass vascular resistance (29) and in significant increases in combined ventricular output and blood flow to the carcass tissues (21, 29). Superimposition of hypoxemia on hyperinsulinemia does not overcome all these effects of hyperinsulinemia. Blood flow to carcass tissues remained markedly increased in hyperinsulinemic fetuses relative to that observed in hypoxemic control fetuses, despite large increases in plasma Epi and NE concentrations (29). The more rapid attenuation of the MAP responses to Epi or NE infusion by insulin administration and the increases in MAP observed after its cessation, as well as the significantly lower MAP maintained during insulin administration to Epi- or NE-infused fetuses reported here, indicates that insulin concentrations within the physiological range have an important influence on the regulation of peripheral vascular resistance and the distribution of cardiac output in the presence of high Epi or NE concentrations, as they do in normal fetuses (21). It is not known whether the relative magnitude of these tissue-specific effects of insulin on carcass tissue blood flow is maintained when Epi or NE concentrations are increased. The catecholamines themselves may also promote vasodilatation, and their vasoconstrictor effects may be antagonized by insulin (4, 30). Evidence from postnatal studies in vivo indicates that relationships between insulin and Epi in the regulation of skeletal muscle blood flow and glucose (19) or protein (11) metabolism are complex and may involve changes in the distribution of blood flowing to nutritive or nonnutritive capillary networks within the muscle (4). The significance of such observations for the distribution of blood flow and metabolism within the skeletal musculature and its relativity to that of other tissues in the fetus is unknown.

These investigations, in agreement with earlier studies (21, 22, 29), show that infusion of insulin results in increased fetal heart rates. However, despite coinfusion of Epi or NE, the increases in fetal heart rate were considerably less than those observed during prolonged hyperinsulinemia (29) or when normoinsulinemia was restored by insulin infusion during short infusions of NE into fetal sheep (22). Whatever the reason for the differences among these studies, it is evident from the prolonged infusion studies that counterregulation of insulin's effects on the heart, as well as the systemic circulation, must contribute to the redistribution of metabolic activity by Epi and NE.

Plasma insulin concentrations were only restored to the normal range by insulin infusion for part of the period of Epi or NE infusion in these investigations. Some disruption to fetal growth might, therefore, have been expected on the basis of earlier findings (2) and may well have occurred. However, there was no evidence in either study for any significant disruption to fetal growth like that observed when similar fetal sheep were infused with only Epi or NE (2). Mean body weights of fetuses in the first experiment were less at termination than those of control fetuses in the earlier investigation (2), due to the inclusion of undiagnosed triplet-bearing ewes. However, fetuses given exogenous insulin for 8 of the 12 days of Epi infusion in the studies reported here had an estimated rate of growth that did not differ significantly from that of control fetuses in either study, whereas Epi-infused fetuses in the previous study had failed to gain any weight during infusion (2). Infusion of exogenous insulin to restore fetal plasma insulin concentration for 4 of the 7 days of NE infusion also appeared to prevent significant reduction in the estimated growth rate of these fetuses relative to that of normal control fetuses in our laboratory, although no direct comparison with saline-infused control fetuses within the experiment was possible. There was no evidence that either Epi or NE had significant inhibitory effects on the weight or estimated growth rate of skeletal muscles or hindlimb bones isolated from the fetal carcass when exogenous insulin was administered to restore normal plasma insulin concentrations during their infusion. Infusion of Epi or NE alone resulted in a ~30% reduction in muscle growth during the infusion period (2). In the present study, by contrast, estimated rates of muscle growth during Epi or NE infusion were comparable with those of control fetuses in the earlier study. The mechanism by which infusion of exogenous insulin at ~0.2 mU · kg¹ · min¹ restores fetal tissue growth during Epi or NE infusion remains uncertain, but is entirely consistent with the restoration of growth in pancreatectomized fetal sheep administered exogenous insulin via implanted minipumps (8).

Whatever the mechanisms involved, the negligible effect of Epi or NE infusion on fetal growth, and particularly on growth of the skeletal muscles, observed when normoinsulinemia is restored by insulin infusion, contrasts strikingly with the marked adverse effects observed when either Epi or NE was infused alone (2). This emphasizes the importance of insulin for maintenance of peripheral protein anabolism, whether this results from direct actions on tissue protein synthesis, which remains in doubt (20), or is consequent on increased delivery of substrates and oxygen to peripheral tissues (31). When placental transfer of glucose or oxygen is unrestricted, the high glucose and oxygen concentrations maintained during prolonged Epi or NE infusions into fetal sheep (2) and their reduction towards the normal range during insulin infusion make it clear that the profound inhibition of insulin release by Epi or NE is a primary determinant of their adverse effect on carcass tissue accretion. The observations reported here imply that redistribution of metabolic activity away from carcass tissues, consequent on the inhibition of insulin release, must be a primary determinant of the decreased carcass tissue accretion observed when Epi and NE concentrations are increased, as they are during hypoxemia (12, 13, 18, 25). Increased production of IGF binding protein I (IGFBP-1) and decreased bioactivity of IGF-I have been proposed to explain adverse effects of catecholamines on growth (17), but inhibition of insulin release by catecholamines (2, 17) seems likely to be the primary regulatory mechanism by which the selective inhibition of fetal growth (2, 25) and tissue DNA synthesis (16) is brought about during hypoxemia. Insulin itself stimulates the production of IGF-I and inhibits synthesis of IGFBP-1 (4), so restoration of growth by insulin infusion in both Epi- or NE-infused fetal sheep and in pancreatectomized fetal sheep (8) will likely involve modulation by increased activity of IGF-I and decreased production of IGFBP-1.

In conclusion, these experiments show that establishment and maintenance of hypoinsulinemia by Epi and NE infusion are essential to their adverse effects on the growth of the fetal sheep. To the extent that curtailment of the metabolically expensive processes of protein turnover in the fetus may be vital for successful adaptation to reduced nutrient and oxygen availability (23), they also show that inhibition of insulin release by Epi and NE may be crucial to their success in permitting adaptation of the fetus to prolonged undernutrition or hypoxemia.

Perspectives