Main

Insulin resistance is usually accompanied by hyperinsulinemia or normal blood insulin with clinical manifestations of hyperglycemia. This resistance is common in insulin-dependent diabetic patients(1), noninsulin-dependent diabetic patients, and patients with extensive burns, acute infection(2), or surgical trauma(3). Markkola et al.(4) have suggested that hyperglycemia per se or glucose toxicity might significantly contribute to insulin resistance in type I diabetes. Hyperglycemia is also commonly observed in premature infants and may contribute to hemorrhage and increased mortality(5, 6). Results obtained from experiments using both animals(7) and humans(8) have suggested that this glucose intolerance in premature infants is secondary to insulin resistance(7). Little is known about regulation of blood glucose level in hypoxic newborn infants. To understand whether hypoxia may induce insulin resistance, the present study examined blood glucose, insulin, glucagon, and hydrocortisone as potential hormone causes of insulin antagonism during hypoxia in newborn calves. This study also investigated insulin receptor binding and tyrosine protein kinase activity for postreceptor functions in hypoxic liver as potential factors mediating insulin resistance.

METHODS

Reagents. Monoidinated 125I-insulin (specific activity 174-206 nCi/ng) was obtained from the Diabetic Research Institute of Huaxi Medical University, Chengdu, China. Unlabeled insulin was purchased from Novo Company, Denmark. [γ-32P]ATP (10 mCi/0.025 mL) was obtained from Yahui Co., Beijing, China. Hormone assay kits were purchased from DPC Co., Tianjin, China. WGA, N-acetylglucosamine, and Glu-4:Tyr-1 were purchased from Sigma Co., St. Louis, MO. Other reagents were purchased from local chemical companies. All of these reagents were analytical grade.

Animals. Fourteen newborn calves, fasted after birth, were divided randomly into two groups. The hypoxic group (n = 7) was tested at 6-10.5 h postdelivery (277-284 d of gestation, weight 35.4-45.1 kg) and the control group (n = 7) at 7-11 h postdelivery (276-284 d of gestation, weight 35.5-45.0 kg). The first group was exposed to a gas mixture containing oxygen 4.8-5.9% in nitrogen, and the second group was exposed to room air. Gas exposure was conducted for 2 h using a Y-type tube connected to a head hood. Subsequently, liver biopsies (3 g) were taken under anesthesia using sodium pentobarbital at a dose of 25-30 mg/kg. The samples were minced and quickly stored in liquid nitrogen until studied. All experiments were carried out at room temperature except those specifically indicated.

Glucose and hormone assay. Cervical venous blood samples were obtained for blood gas analysis, and measurements of whole blood glucose, serum lactate, insulin, glucagon, and hydrocortisone were made. Blood glucose was measured by a glucose oxidase method using Refloux II M of Boehringer Mannheim (Indianapolis, IN). Serum lactate was determined by a lactate oxidase assay using a YSI model 23-L lactate analyzer. Plasma insulin, glucagon, and hydrocortisone concentrations were determined using RIA kits according to the manufacturer's instructions.

Partially purified receptors preparations. The method of Freidenberg et al.(9) was used with minor modifications for the preparation of liver tissue membranes. Briefly, livers were homogenized in a buffer solution containing 25 mM HEPES, 0.25 M sucrose, 5 mM EDTA-Na2, 10 mM phenylmethylsulfonyl fluoride (pH 7.6), followed by centrifugation at 12 000 × g for 10 min at 4 °C. The supernatant was centrifuged at 50 000 × g for 45 min at 4°C. The pellet was solubilized in a solution containing 2 mg/mL bacitracin, 2 mM phenylmethylsulfonyl fluoride, and 2% Triton X-100 for 30 min and subsequently centrifuged at 120 000 × g for 30 min at 4°C. The resulting pellet was solubilized in the same solution and partly purified by the method of affinity chromatography using 3.0 mL of WGA-Sepharose. After extensive washing, the glycosylated proteins were eluted from the WGA-Sepharose column by 0.3 M N-acetylglucosamine. Each fraction, 1 mL, was collected and stored at -35 °C until analysis.

Insulin binding assay. Insulin receptor binding was assessed using the method of Sinha and Jenquin(10) with slight modification. Partially purified receptor (50 μL) was incubated with125 I-insulin (50 μL, 260 nCi/ng) in the presence or absence of unlabeled insulin (0.1-1000 nM) in buffer (25 mM HEPES, 0.05% Triton X-100, 0.1% BSA) for 18 h at 4 °C at a final volume of 250 μL. Free125 I-insulin was separated by 0.06% bovine γ-globulin and 10% polyethylene glycol. The pellet was counted in a γ-scintillation counter. Four binding parameters were estimated. These parameters included the affinity constant K1 and binding capacity R1 for the higher affinity binding site (S1), and affinity constant K2 and binding capacity R2 for the lower affinity binding site (S2). These parameters were determined using Scatchard analysis of a radioligand binding assay with a computer program which was developed by the Department of Nuclear Medicine, Shanghai Second Medical University. The number of binding sites(S1, S2) per mg of protein were calculated from the following formula: No./mg of protein =[R1(R2)(nM) × 0.25 mL × 6.02× 1011]/[weight of protein (mg)].

IR-TK activity assay. Tyrosine-specific protein kinase of the insulin receptors was determined by the method of Sinha and Jenquin(10). The partially purified receptor protein (80 μg) was incubated with insulin of increasing concentrations (0-1000 nM) for 17 h at 4 °C, then with 0.25 mg of Glu-4:Tyr-1 and 2 μCi of[γ-32P]ATP for 20 min at room temperature in two groups.

Statistical analysis. All data were expressed as mean ± SD. Independent t tests were performed to compare the difference between hypoxic and control groups. Paired t tests were used for comparing changes at different exposure times of pre- and post-experiment within one group.

RESULTS

There were no significant differences between the two groups for gestation age, birth weight, or age after birth. Blood gas analyses and blood lactate values in the two groups of animals are shown in Table 1. It is obvious that lactate accumulated, and negative values of base excess developed in the hypoxic group. There were no significant differences for blood glucose levels between pre- and post-experiment of control animals, but glucose levels were significantly higher at 2 h posthypoxia compared with prehypoxia values in the hypoxic group. As show in Figure 1, blood glucose levels were significantly higher after 2 h in the hypoxic group compared with the control group. Plasma insulin (t = 2.27,p < 0.05) and hydrocortisone levels (t = 5.74,p < 0.01) were significantly increased, but glucagon concentrations were significantly lower (t = 2.28, p < 0.05) after hypoxia, compared with those in the control animals.

Table 1 Lactate and acid-base data in control and hypoxic calves
Full size table
Figure 1
figure 1

Blood glucose and plasma insulin, glucagon, and hydrocortisone concentrations in control and hypoxic calves. Blood glucose significantly increased from 5.47 ± 1.61 to 7.97 ± 1.30 mmol/L(t = 6.21, p < 0.01), plasma insulin from 10.01± 1.75 to 12.99 ± 1.85 ng/mL (t = 5.72, p< 0.01) and hydrocortisone from 9.53 ± 3.10 to 26.67 ± 8.28μg/dL (t = 5.76, p < 0.01), but glucagon decreased from 73.50 ± 10.32 to 60.34 ± 9.59 pg/mL (t = 5.28,p < 0.01) in hypoxic group after 2-h hypoxia. Comparison of data obtained from pre-experiments between hypoxic and control groups showed no significant differences, but data from post-experiments between two groups showed significant differences.

Full size image

Nonspecific 125I-insulin binding was determined in the presence of 10-6 M insulin. There was no difference in nonspecific binding in the two groups (2.47 ± 1.33% in controls, 3.13 ± 0.78% in the hypoxic group). Specific binding was calculated as (observed counts - nonspecific binding)/total counts. Competitive inhibition by insulin (0.1-1000 nM) progressively decreased specific 125I-insulin binding, and the values of this binding were all lower in the hypoxic group than in the control animals. Basal 125I-insulin binding (in the absence of competition) was also lower in the hypoxic group than in the control group (12.71 ± 1.25% versus 15.14 ± 1.27%), as shown in Figure 2.

Figure 2
figure 2

Percent insulin binding vs unlabeled insulin concentration in control and hypoxic calves. Aliquots of 40 μL (40 μg) of purified protein were incubated with 125I-insulin and increasing concentrations (0, 0.1 to 1000 nM) of unlabeled insulin. Specific125 I-insulin binding was decreased significantly in the hypoxic group. p < 0.05.

Full size image

K1, K2, R1,R2, and the numbers of binding sites (S1,S2) per mg of partially purified receptors were calculated using the dual sites multipoint radioligand binding assay program. These results are summarized in Table 2. The values for all six parameters tended to be lower in the hypoxic animals. The number of low affinity receptor sites decreased significantly at 2 h after hypoxia compared with the controls.

Table 2 Characteristics of insulin receptor binding in control and hypoxic calves*
Full size table

To evaluate postreceptor binding, IR-TK activity was assessed in the partially purified hepatic insulin receptors. Dose response data for insulin-stimulated IR-TK activity in the two groups of animals are presented in Figure 3. IR-TK activity in the hypoxic animals was decreased compared with values in control animals. With the stimulation at various increased concentrations of insulin, IR-TK activity increased in both groups. In the presence of stimulation of insulin at maximal concentration(1000 nM), the maximal IR-TK activity was significantly lower in the hypoxic group than in controls (30.83 ± 5.80 versus 22.18 ± 2.79 fmol/mg of protein).

Figure 3
figure 3

IR-TK activity relative to increasing unlabeled insulin concentrations in control and hypoxic calves. Tyrosine kinase activity of partially purified insulin receptors (fmol/mg Glu-4:Tyr-1/mg protein/min) significantly decreased in the hypoxic group compared with control animals(p < 0.01 for each point). See text for details.

Full size image

DISCUSSION

Carbohydrate metabolism is often deranged during severe stress such as hypoxia(11, 12), and this derangement is due, at least in part, to stimulated endocrine responses. This is commonly accepted and extensively described in adult humans and in animals, but there are limited data in the neonatal population. It is known, however, that hypoxia in the newborn is, arguably, the most common cause of severe, acquired brain injury(13). The incidence of hyperglycemia is most frequent in the first 24 h after birth and was related to low Apgar scores, respiratory distress syndrome(14), and hypoxia(15). Both hyperglycemia and hypoxia may have a potential impact on long-term neuralgic sequelae, morbidity, and mortality. Moreover, Vardi et al.(16) reported that, among 12 stressed children, four developed type I diabetes within 1 y.

The mechanism of hypoxia-induced hyperglycemia is still unclear. The major goal of this study was to describe changes in blood glucose metabolism, changes in insulin, glucagon, and hydrocortisone concentrations, and insulin receptor functions including IR-TK activity after hypoxia. The usual hypoxic model utilizes oxygen concentrations between 5 and 8% for 0.5-2 h(17). We used the calf as our animal model because a substantial amount of blood supply was needed in this and other related studies being conducted in our laboratory. Calves were exposed to 4.8-5.9% oxygen for 2 h to create hypoxia and acidosis, with serum lactate approximating 5.8-13.0 mmol/L. In this environment, blood glucose increased from 3.1-7.2 to 4.8-8.2 mmol/L at 1 h and 5.9-9.3 mmol/L at 2 h. Blood glucose levels in two of seven and in five of seven calves were more than 7 mmol/L at 1 and 2 h, respectively, in the hypoxic animals. Therefore, hypoxia, indeed, induced hyperglycemia in the newborn.

Hyperglycemia or insulin resistance may result from defects at a prereceptor, receptor, or postreceptor level. Insulin counterregulatory hormones such as hydrocortisone, glucagon, growth hormone, and epinephrine may affect blood glucose. Hydrocortisone as a major factor in stress reactions, and glucagon and insulin were measured in this study to assess their possible contribution to insulin resistance at a prereceptor level. Our results show that hypoxia increased plasma hydrocortisone and insulin levels. Hydrocortisone is an important mediator of the stress response and of hyperglycemia. High hydrocortisone levels have been reported in premature infants with hyaline membrane disease(18). Lilien et al.(19) has reported that hydrocortisone levels in stressed newborns are lower in hyperglycemic than in normoglycemic infants. As demonstrated in the present study, hydrocortisone values were significantly elevated in hypoxic newborn calves at 1 and 2 h after hypoxia relative to prehypoxic levels.

The observed elevation of insulin levels may have been secondary to hyperglycemia or insulin resistance. Long et al.(20) noticed that insulin concentrations are either normal or increased, and that increases in plasma glucose are common in stressed patients in association with decreased sensitivity to insulin. Insulin is a major anabolic and anticatabolic hormone. Despite the rise in plasma insulin concentrations, the effects of insulin may be antagonized by glucagon and hydrocortisone or by reduced insulin action at the receptor binding site. Surprisingly, plasma glucagon values during 2 h of hypoxia were decreased, although it has been reported that pancreatic alpha cells in newborn infants at 2 h after birth are able to release glucagon(21) and that plasma glucagon levels rise after delivery and remain elevated throughout the first days of life(22). It may be noted that the simultaneous infusion of glucose and insulin can inhibit glucagon secretion(23, 24). The alpha cell is capable of responding to both amino acids and glucose(22). Glucose infusion in full-term and preterm newborn infants results in a prompt and sustained suppression of glucagon secretion(25). This observation may explain why the elevated glucose and insulin resulted in a fall of plasma glucagon levels in the present study.

Insulin receptor binding was measured in this study to assess its possible contribution to insulin resistance. A decrease in lower affinity receptor sites in liver tissue was observed after 2 h of hypoxia. It has been reported that decreased insulin binding was not related to insulin resistance in normal newborn dogs(26), and insulin receptors in fetal tissue were not down-regulated by hyperinsulinemia(27). One possible explanation for our result is that insulin reduced the insulin receptor response as a result of endocytosis or internalization(28). Another explanation is the possible effects of hormones such as hydrocortisone or glucagon.

The mechanism of insulin signal transformation after insulin binding remains unclear. Potential mechanisms have been proposed, including phosphorylation cascade, second messenger at the cell surface, and insulin-receptor interaction with other organelles(29). It has been reported that insulin binding stimulates insulin receptor autophosphorylation and subsequent insulin receptor activation(30). It is generally believed that the insulin receptor tyrosine kinase activity is essential for insulin action(30, 31). There is, however, little or no information about newborn infants or animals(10), especially after hypoxia. Johnston et al.(26) has reported that insulin binding (numbers and affinity) is not the limiting factor in neonatal insulin action and that an insulin receptor postbinding defect exists in insulin resistant newborns. IR-TK activity in normal newborn rat liver was reported to be similar in neonatal and adult rats(10). The present study showed that IR-TK activity was reduced in hypoxia, and that blood glucose levels increased in spite of an increase of insulin secretion. The possibility of decreased of IR-TK activity as a major contributing factor to insulin resistance in newborn calves cannot be excluded, because the decrease of insulin binding occurs only in lower affinity receptor sites, and it has been reported that the “spare” insulin receptor can be activated for the insulin reaction(32). The possible explanation for the decreased IR-TK activity is likely the decrease in insulin binding by the insulin receptor, and inhibition of IR-TK activity by certain molecular species changed or produced during hypoxia. It has been reported that the production of tumor necrosis factor-α increased in human mononuclear cells during hypoxia(33). This factor, in turn, can convert insulin receptor substrate-1 into an inhibitor of insulin receptor tyrosine kinase activity in culture cells(34). The mechanism of the postreceptor effect, however, needs further investigation.

The present study demonstrates that insulin resistance is associated with hyperglycemia and changes of endocrine metabolism in hypoxic newborn calves, and suggests that blood glucose levels should be closely monitored when stressed newborn infants receive a high dose glucose infusion and glucocorticoid infusion. Based on the mechanisms of decreased IR-TK activity, certain strategies may be developed for the prevention and management of hyperglycemia and insulin resistance in stress.