Persistence of impaired insulin secretion in infant rhesus monkeys that had been hyperinsulinemic in utero.

Chronic in utero hyperinsulinemia in the fetal rhesus monkey produces a number of changes in the fetus that are similar to those found in the human infant of the diabetic mother, including macrosomia, selective organomegaly, and altered insulin secretion during the neonatal period. The chronically hyperinsulinemic fetal rhesus model has been used to test the hypothesis that the effects of chronic hyperinsulinemia persist beyond the neonatal period into later life and may, in part, be responsible for the increased prevalence of impaired glucose tolerance or diabetes found in the human infant of the diabetic mother. We report that infant rhesus monkeys that had plasma insulin concentrations of approximately 10 times basal levels (2176 +/- 808 pmol compared to 172 +/- 101 pmol) exhibited reduced insulin secretion during the first 5 months of life. The integrated incremental change in plasma insulin and immunoreactive C-peptide (IRCP) concentration was significantly reduced by approximately 50% in response to i.v. glucose, arginine, and tolbutamide when given at 3, 4, and 5 months of age. The response to glucagon at 2 months of age was equivocal with a significantly reduced insulin response but without the corresponding IRCP reduction. There was no difference between groups in insulin sensitivity as measured at 6 months of age by an i.v. insulin tolerance test. The glucagon and glucose tolerance tests were repeated annually in both groups until the animals were 3 yr of age with no differences in insulin or IRCP secretion being observed. We conclude that chronic in utero euglycemic hyperinsulinemia results in impaired insulin secretion that persists beyond the neonatal period.

Impaired insulin secretion after intravenous glucose in neonatal rhesus monkeys that had been chronically hyperinsulinemic in utero.

Chronic hyperinsulinemia in the fetal rhesus monkey results in fetal macrosomia without change in fetal plasma glucose concentration. After 18 days of hyperinsulinemia, fetuses were delivered by cesarean section, at which time experimental animals had significantly (P less than 0.05) elevated umbilical artery plasma insulin concentrations of 2039 +/- 854 pM compared with 129 +/- 72 pM. Plasma immunoreactive C peptide (IRCP) was significantly reduced to 39 +/- 17 pM compared with 286 +/- 134 pM. Eight hours after the insulin-delivering pumps were removed, plasma glucose, insulin, and IRCP were the same in both the experimental and control groups. At this time, 0.5 g glucose/kg was given intravenously and insulin and IRCP secretion was measured over a 1-hr period. The secretion, as assessed by integrating the incremental response of both insulin and IRCP, was significantly (P less than 0.05) lower by 80% in the experimental animals compared with the controls. Our data show that experimentally produced in utero euglycemic hyperinsulinemia in the fetal rhesus monkey produces a defect in the glucose-mediated insulin secretory mechanism that is detectable in the neonatal period even when hyperinsulinemia is no longer present. This study provides more support for the concept that fuel/hormone-mediated fetal teratogenesis may explain some of the fetopathy of the infant of the diabetic mother.

Sympathoadrenal responses during hypoglycemia, hyperinsulinemia, and hypoxemia in the ovine fetus.

Interrelations of sympathoadrenal function and changes in glucose and insulin homeostasis were studied in chronically cannulated late gestation fetal sheep. Catecholamine secretory rates (based on direct adrenal sampling) and plasma concentrations were determined in the fetus during 2 h of insulin-induced hypoglycemia, during a period of hypoxemia, and during hyperinsulinemia per se (i.e., without hypoglycemia). Fetal insulin infusion (5-10 mU.kg-1.min-1) resulted in hypoglycemia and a significant rise in secretion of epinephrine but not of norepinephrine. By contrast, fetal hypoxemia caused a prompt and significant increase in adrenal secretion of both norepinephrine and epinephrine. Changes in peripheral plasma catecholamine levels were usually, but not always, qualitatively similar to those in adrenal secretion; the latter was a far more sensitive indicator of adrenal function. Hyperinsulinemia per se caused no change in adrenal secretory rates or plasma concentrations of catecholamines. Nevertheless, insulin infusion caused a fetal tachycardia even in the absence of hypoglycemia and hypoxemia, suggesting either a direct effect on the heart or stimulation of sympathetic nerves.

Impaired insulin secretion in the neonatal rhesus monkey after chronic hyperinsulinemia in utero.

The secretion of insulin by the pancreas of the newborn rhesus monkey that had been made experimentally hyperinsulinemic in utero was studied in 18 animals. Chronic in utero hyperinsulinemia was produced by the continuous subcutaneous delivery of 4.75 units of insulin per day for 18 +/- 1 days. After delivery, the insulin-containing pump was removed to allow neonatal insulin levels to drop to normal levels. By 6.5 +/- 1.0 hr after pump removal, plasma glucose, insulin, and C-peptide immunoreactivity (CPIR) were comparable in the control and experimental animals. At that point 300 micrograms of glucagon/kg body weight was given iv to stimulate insulin secretion. After 30 min a significant elevation (expressed as the percentage of basal levels) in plasma glucose by 250%, insulin by 200%, and CPIR by 200% was observed in the control animals. In contrast, no changes in plasma insulin or CPIR concentrations occurred, with an attenuated glucose response that was only one-fifth of the control response, in the experimental animals. These results along with the observed lowered concentrations of CPIR in the plasma and insulin in the pancreas at birth can be interpreted as evidence that insulin is an inhibitor of its synthesis and secretion in utero and that this abnormal intrauterine environment causes changes that persist into extrauterine life.

Effects of insulin infusion on plasma catecholamine concentration in fetal sheep.

To evaluate the response of the sympathoadrenal system in fetal sheep receiving exogenous insulin infusion, we measured plasma catecholamine levels in 14 chronically catheterized fetal sheep before and during an infusion of insulin for 2 days. Catecholamine values were measured in fetal arterial plasma by an electrochemical detection method. Fetal plasma norepinephrine and epinephrine concentrations increased significantly during insulin infusion. Significant inverse correlations were observed between the log norepinephrine concentration and fetal arterial oxygen content and glucose values. A significant direct correlation between the log norepinephrine concentration and fetal arterial carbon dioxide concentration was also observed. The log epinephrine concentration correlated inversely with plasma glucose concentration. Increases in fetal heart rate during both the noninfused and insulin-infused states correlated significantly with increases in norepinephrine concentration. We conclude that the sympathoadrenal system is activated during fetal insulin infusion, potentially supporting some of the fetal cardiovascular responses to insulin infusion.

Metabolism and placental transfer of 125I-proinsulin and 125I-tyrosylated C-peptide in the pregnant rhesus monkey.

125I-Proinsulin or 125I-tyrosylated-C-peptide (125I-tyr-CP) was administered to pregnant Rhesus monkeys by bolus followed by constant infusion to examine placental transfer of these peptides. At the end of each infusion, fetuses were exsanguinated in situ via the umbilical vein. The bolus-constant infusion technique produced a steady state in maternal plasma of immunoprecipitable label, measured using excess insulin or C-peptide antiserum. In animals infused with 125I-proinsulin, analysis of umbilical venous plasma revealed no apparent transfer to the fetus of immunoprecipitable label. In animals infused with 125I-tyr-CP, 3-13% of the umbilical venous plasma radioactivity was immunoprecipitable, representing 1.4-5.8% of the immunoprecipitable radioactivity in maternal plasma at delivery. Gel filtration chromatography of umbilical venous plasma revealed that the immunoprecipitated moiety was a fragment of 125I-tyr-CP. Analysis of maternal plasma showed that the predominant peak of radioactivity represented intact C-peptide. A peak corresponding to the fetal immunoprecipitable peak was also present. Analysis of simultaneous maternal arterial and uterine vein plasma samples showed that degradation of 125I-tyr-CP occurred across the uterus. Studies in one nonpregnant and three postpartum animals indicated that pregnancy increased the rate of metabolism of 125I-tyr-CP. When 125I-tyr-CP was incubated with trophoblastic cells in culture, degradation to a species corresponding on gel filtration to the immunoprecipitable fetal metabolite was found. We conclude that proinsulin, like insulin, does not traverse the placenta. Immunoreactive fragments of C-peptide do cross, however, and pregnancy alters the metabolism of 125I-tyr-CP, probably owing to placental degradation.

Preserved fetal plasma amino acid concentrations in the presence of maternal hypoaminoacidemia.

The effects on the conceptus of persistently decreased maternal plasma amino acid concentrations were studied in pregnant rats by the infusion of glucagon (0.21 mg/day) to the mother from day 14 to 20 of gestation with a subcutaneous, osmotically driven minipump. Controls received diluent. The experimental animals either had normal caloric intake and weight gain, or diminished caloric intake with no weight gain. Both experimental groups exhibited a decrease in plasma total amino acid concentration of approximately 50%. Maternal plasma glucose and insulin concentrations were unaffected except for slight decreases in the low weight gain group. At cesarean section on day 20, fetal weight was unaffected in the normal weight gain group, while the low weight gain animals exhibited intrauterine growth retardation. Fetal plasma glucose and insulin concentrations were unaffected. Despite the marked decrease in maternal plasma total amino acid concentration, fetal plasma total amino acid concentration was unaffected. Individual plasma amino acid concentrations in the normal weight gain mothers and fetuses revealed a spectrum of changes. Some maternal amino acids were decreased by more than 60% (alpha-aminobutyric acid, asparagine, threonine, glutamine, alanine) while others were unaffected (tyrosine, tryptophan, phenylalanine, histidine). In general, amino acids that were decreased in the mother exhibited no change or a lesser decrease in fetal plasma concentration, while those that were unaffected in the mothers showed increased fetal concentrations. Fetuses from the low weight gain mothers had plasma amino acid profiles that were similar to those of the normal weight gain mothers.(ABSTRACT TRUNCATED AT 250 WORDS)

Temporal response of immunoreactive erythropoietin to acute hypoxemia in fetal sheep.

Acute hypoxemia was produced in chronically catheterized sheep fetuses to determine the response time necessary to increase plasma immunoreactive erythropoietin (Ep) concentration. Sodium nitrite (0.2 mM) was infused via a fetal vein to induce fetal hypoxemia. The resultant fetal methemoglobinemia was associated with a predictable, incremental decrease in arterial oxygen content. Twelve nitrite infusions were performed in eight fetal sheep preparations (gestational ages 115-146 days). Mean methemoglobin level increased to 33% of total Hb after 1-2 h of NaNO2 infusion. These results were compared to those obtained in nine control studies in eight fetuses in which no change was observed for plasma Ep, arterial oxygen content, PaO2, pHa, or whole blood lactate. In the nitrite infused group, however, a significant and progressive increase in mean plasma Ep level over baseline levels was observed during the 4th and 5th h of hypoxemia (p less than 0.01). This change in Ep was significantly greater compared to the control group. These results, however, were confounded by the concomitant development of a lactic acidemia secondary to the fetal hypoxemia. To examine the theoretic possibility that lactic acidemia may primarily affect fetal Ep levels, an additional group of five fetuses was infused with L-lactic acid for the same time period. Although the decrements in pHa and whole blood lactate levels achieved in these fetuses were in excess of those observed during the nitrite infusions, this possibility was ruled out since no change in fetal plasma Ep levels occurred. We conclude that during the 4th h of acute fetal hypoxemia a predictable, progressive increase in plasma Ep level is observed.(ABSTRACT TRUNCATED AT 250 WORDS)

Effect of insulin on the size of skeletal muscle fibers of fetal rhesus monkey.

Increased body weight of insulin-treated fetuses is largely attributed to increased muscle mass. Skeletal muscle from fetal rhesus monkeys was analyzed for differences in the sizes of muscle fibers from monkeys treated in utero with high or low doses of insulin and compared to control (non-insulin-treated) animals. The results indicate no significant difference among the various groups studied and suggest that hyperplasia, not hypertrophy, of skeletal muscle fibers is responsible for the observed increase in muscle mass.

Effects of hyperinsulinemia in the primate fetus.

Nonhuman primate models of gestational diabetes have produced fetopathies most similar to those of the human infant of the mother with gestational diabetes (IGDM). Fetal hyperglycemia, hyperinsulinemia, macrosomia, selective organomegaly, intrauterine death, and placental hyperplasia are hallmarks of the fetopathy of the IGDM. The chronic infusion of insulin into the fetus of a normal pregnant rhesus monkey results in fetal hyperinsulinemia with normal to low plasma metabolic substrate concentrations. Under these conditions, fetal hyperinsulinemia is sufficient to cause fetal growth and hormone changes observed in the human IGDM. Our studies provide evidence that the soft tissue hyperplasia in the fetal macrosomia syndromes in humans and nonhuman primates in which fetal hyperinsulinemia is observed is the direct result of that chronic in utero hyperinsulinemia.

Induction in utero of hepatic glucose-6-phosphatase by fetal hypoinsulinemia.

The effect of fetal hypoglycemia and hypoinsulinemia on fetal rat hepatic glucose-6-phosphatase activity was studied. Fetal hypoglycemia and hypoinsulinemia were produced by inducing maternal hyperinsulinemia and hypoglycemia secondary to the exogenous administration of insulin via implantation of osmotically driven minipumps on day 15 of gestation into 15 experimental animals. 13 animals served as sham-operated controls. Cesarean sections were performed on day 20 or 21 of gestation under pentobarbital anesthesia. Liver glucose-6-phosphatase activity was increased in the hypoinsulinemic fetuses. In contrast, the hyperinsulinemic mothers had suppressed hepatic glucose-6-phosphatase activity. Hypoinsulinemia would appear to be the primary stimulus for enhanced fetal glucose-6-phosphatase in this model.

Chronic hyperinsulinemia in the fetal rhesus monkey: effects of physiologic hyperinsulinemia on fetal substrates, hormones, and hepatic enzymes.

Chronic fetal hyperinsulinemia, similar to that found in human infants of diabetic mothers, was produced in fetal rhesus monkeys during the latter third of gestation. Fetal plasma glucose and amino acid concentrations were found to be inversely logarithmically correlated with plasma insulin concentration. Fetal plasma glucagon concentrations were suppressed by hyperinsulinemia. Fetal plasma erythropoietin concentrations were increased by hyperinsulinemia in a dose/response manner. The activity of the hepatic gluconeogenic enzymes glucose-6-phosphatase and total phosphoenolpyruvate carboxykinase were reduced by hyperinsulinemia. Fatty acid synthase complex activity was, in contrast, increased by hyperinsulinemia while citrate cleavage enzyme and glucose-6-phosphate dehydrogenase were only increased when supraphysiologic hyperinsulinemia was produced. This model provides an opportunity to study the metabolic effects of hyperinsulinemia separate from those of hyperglycemia on the primate fetus, making it a useful model for the study of fetal pathologic conditions in diabetic pregnancies.

Chronic hyperinsulinemia in the fetal rhesus monkey. Effects of physiologic hyperinsulinemia on fetal growth and composition.

One of the hallmarks of the hyperglycemic-hyperinsulinemic infant of the diabetic mother (IDM) is macrosomia and selective organomegaly. Primary hyperinsulinemia, with insulin levels similar to those observed in human IDMs at delivery, was produced in the fetal rhesus monkey during the last third of gestation. The effects of this physiologically relevant hyperinsulinemia, in the absence of hyperglycemia, on fetal growth were studied. Fetal macrosomia, with a 23% increase in total body weight, was observed in physiologically hyperinsulinemic fetuses. A similar 27% increase in weight was produced by fetal insulin levels that were 10 times higher. A logarithmic correlation was observed between fetal birth weight ratio and fetal plasma insulin concentration. In contrast to this increase in weight, skeletal growth, as measured by crown-heel length and head circumference, was not affected by hyperinsulinemia. Only cardiomegaly was found in the low-dose hyperinsulinemic fetuses, whereas cardiomegaly, hepatomegaly, and splenomegaly were produced by hyperinsulinemia in which insulin levels were in the highest range. Compositional analysis of heart and skeletal muscle indicated no differences in the protein, RNA and DNA concentration, or in the protein-to-DNA ratio in hyperinsulinemic fetuses. We interpret these data as indicating that fetal insulin plays the predominant role in controlling the normal, as well as the augmented, fetal weight characteristic of the human infant of the diabetic mother.

Somatomedins and insulin in diabetic pregnancies: effects on fetal macrosomia in the human and rhesus monkey.

The concentrations of the somatomedins (SMs) insulin-like growth factors I and II (SM-C/IGF-I and IGF-II) were measured by RIA in six normal and seven insulin-dependent diabetic pregnant women and their infants at delivery. SM-C/IGF-I and IGF-II levels in the two groups of women were similar. Maternal IGF-II concentrations correlated with maternal hemoglobin AIc levels (r = 0.68) and infant birth weight ratios (actual birth weight/expected 50th percentile sex-corrected birth weight for gestation age; r = 0.54). SMC/IGF-I and IGF-II levels in umbilical plasma in infants of diabetic mothers did not differ from those in control infants, but were lower than the corresponding maternal values. In contrast, umbilical plasma levels of C-peptide immunoreactivity were significantly elevated in the infants of diabetic mothers (2.25 +/- 1.85 (+/-SD) vs. 0.34 +/- 0.15 pmol/ml; P less than 0.01). The infant birth weight ratio was logarithmically correlated with the umbilical plasma C-peptide immunoreactivity (r = 0.78). SM levels were also measured by radioreceptor assay in five normal and five hyperinsulinemic rhesus monkey fetuses. When chronic hyperinsulinemia was produced by continuous SC infusion of insulin in the fetal rhesus monkey, the fetal birth weight ratio was also found to be logarithmically correlated with the fetal plasma insulin concentration (r = 0.81). The fetal SM peptide content was elevated only in the fetuses with plasma insulin levels greater than 3000 microU/ml. The fetal weight gains in response to hyperinsulinemia in the human and rhesus are similar. Since fetal SM levels in the humans and monkeys were not significantly different in the two groups, our data suggest that insulin plays the predominant role in stimulating human and subhuman primate excess fetal weight gain of the infant of the diabetic mother during the latter part of gestation.

Glucose kinetics in glucose-infused small for gestational age infants.

To evaluate the maturation of glucose homeostasis in the small for gestational age (SGA) neonate, glucose kinetics were measured with 78% enriched D-[U-13C]glucose by the prime plus constant infusion technique in nine SGA infants and compared with the rate obtained in seven term appropriate for gestational age infants and 13 preterm appropriate for gestational age infants. All of the infants had received glucose intravenously from birth and continued to receive the glucose infusion throughout the study. Fasting plasma glucose and plasma insulin concentrations and plasma [13/12C]ratios were measured during the steady state turnover period. From this data, the glucose production rate was derived. During the turnover period, the SGA and both groups of appropriate for gestational age infants had similar average plasma glucose, plasma insulin, plasma glucagon concentrations, and similar persistent rates of glucose production during glucose infusion. We conclude that under stimulation of glucose infusion, the SGA infant and his AGA counterpart have similar hormonal regulatory responses as well as functional integrity in handling glucose during the second day after birth.

Impermeability of the rat placenta to insulin during organogenesis.

The cause of the embryopathy associated with diabetes mellitus is uncertain. To examine whether exogenously administered insulin may be teratogenic, tracer amounts of radiolabelled insulin were infused for two hours during organogenesis (day 12 1/2 of gestation) into three groups of pregnant rats: control (n = 8), diabetic (n = 5), and hyperinsulinemic (n = 4). For maternal plasma, no differences were found among the three study groups in the percentage of the protein-precipitable (insulin-containing) radioactivity. Tissue radioactivities were expressed relative to the two-hour maternal plasma sample. Maternal kidney samples had the highest total and protein precipitable counts followed in descending order by the maternal plasma, maternal liver, placenta, and embryo. No differences in radioactivities were noted among the three study groups for specific tissues studied. Protein-precipitable radioactivities in the embryo were more than 100-fold less than the maternal plasma values. In 11 of the 17 litters, the acid-insoluble fractions of the embryos were not distinguishable from background counts, and none of the remaining six were greater than twice background. These studies demonstrate that during the period of organogenesis, the rat embryo is protected from maternal insulin by the placenta, and hence, maternal insulin is an unlikely teratogen.

Post-hepatic insulin secretion in the fetal lamb.

Post-hepatic insulin secretion was measured in six chronically catheterized fetal lambs (fetal weight 2.8 +/- 0.3 kg, mean +/- S.E.M.) and the results were compared with those obtained in nine prematurely delivered newborn lambs (birth weight 3.1 +/- 0.3 kg and postnatal age 1.3 +/- 0.2 days). The fetal and neonatal lambs received either a 0.45% saline or a glucose infusion respectively, which resulted in a 2-fold increase in the plasma glucose concentration. [131I]insulin was infused for 110 min to determine the rate of insulin secretion during a steady state of plasma glucose concentration. Post-hepatic insulin secretion and the metabolic clearance rate were calculated. With the 2-fold rise in plasma glucose concentration, the post-hepatic insulin secretion rate increased significantly in the newborn lamb and in three out of four fetuses. The plasma insulin concentration increased significantly in the fetus (11 +/- 4.0 to 35 +/- 8 microU/ml, P less than 0.05) during glucose stimulation as a result of decreased metabolic clearance rate of insulin (10.6 +/- 1.9 to 6.3 +/- 1.8 ml . kg-1 . min-1) and an increase in post-hepatic insulin secretion rate. In spite of an increase in post-hepatic insulin secretion rate, the increase in plasma insulin concentration in the newborn lamb was not significant because of large variation in the values obtained. The data suggest that pancreatic beta-cells in the newborn and in the fetal lamb are equally responsive to a 2-fold increase in plasma glucose concentration.