The Journal of Nutrition Vol. 128 No. 10 October 1998, pp. 1643-1649

Protein Deficiency and Nutritional Recovery Modulate Insulin Secretion and the Early Steps of Insulin Action in Rats^1,3

Márcia Q. Latorraca^{*, 2}, Marise A. B. Reis^*, Everardo M. Carneiro, Maria A. R. Mello, Licio A. Velloso^**, Mario J. A. Saad^**, and A. Carlos Boschero^{*, 4}

^* Departamento de Fisiologia e Biofisica, Instituto de Biologia, Universidade Estadual de Campinas, UNICAMP, Campinas, São Paulo 13083-970, Brasil,  Departamento de Educação Física, Instituto de Biociências, Universidade Estadual Paulista, UNESP, Rio Claro, São Paulo 13506-900, Brasil and ^** Departamento de Clínica Médica, Faculdade de Ciências Médicas, Universidade Estadual de Campinas, UNICAMP, Campinas, São Paulo, 13081-970, Brasil

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Maternal malnutrition was shown to affect early growth and leads to permanent alterations in insulin secretion and sensitivity of offspring. In addition, epidemiological studies showed an association between low birth weight and glucose intolerance in adult life. To understand these interactions better, we investigated the insulin secretion by isolated islets and the early events related to insulin action in the hind-limb muscle of adult rats fed a diet of 17% protein (control) or 6% protein [low (LP) protein] during fetal life, suckling and after weaning, and in rats receiving 6% protein during fetal life and suckling followed by a 17% protein diet after weaning (recovered). The basal and maximal insulin secretion by islets from rats fed LP diet and the basal release by islets from recovered rats were significantly lower than that of control rats. The dose-response curves to glucose of islets from LP and recovered groups were shifted to the right compared to control islets, with the half-maximal response (EC₅₀) occurring at 16.9 ± 1.3, 12.4 ± 0.5 and 8.4 ± 0.1 mmol/L, respectively. The levels of insulin receptor, as well as insulin receptor substrate-1 and phosphorylation and the association between insulin receptor substrate-1 and phosphatidylinositol 3-kinase were greater in rats fed a LP diet than in control rats. In recovered rats, these variables were not significantly different from those of the other two groups. These results suggest that glucose homeostasis is maintained in LP and recovered rats by an increased sensitivity to insulin as a result of alterations in the early steps of the insulin signal transduction pathway.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Epidemiological studies have suggested that reduced growth in fetal and neonatal life is related to the development of glucose intolerance and diabetes in adult life (Hales et al. 1991, Poulsen et al. 1997). This association is, at least in part, a consequence of fetal and early postnatal malnutrition, possibly due to maternal deprivation. Whether the association between malnutrition during early growth and subsequent diabetes is mediated by alterations in insulin action, defects in insulin secretion or a combination of both factors is unclear. Hales and Barker (1992) have proposed that malnutrition in fetal and early infant life can result in impaired development of the pancreatic  cells, leading to insulin deficiency in adult life. Recent evidence suggests that the malnourished fetus undergoes metabolic adaptations, from which it benefits in the short term by increasing fuel availability, and that these adaptations persist throughout life, leading to insulin resistance (Phillips, 1996).

Malnutrition in both humans and experimental animals is associated with reduced insulin secretion and disrupted glucose homeostasis (James and Coore 1970, Okitolonda et al. 1987). In rats fed a low-protein (LP) diet for a limited period (4 wk) immediately after weaning, the glucose-stimulated secretion of insulin is impaired (Carneiro et al. 1995). In addition to a reduced insulin secretion, an increased sensitivity to insulin represented by greater phosphorylation of the insulin receptor (IR),⁵ insulin receptor substrate-1 (IRS-1) and the association of IRS-1 with phosphatidylinositol 3-kinase (PI 3-kinase) have been observed in this same experimental model of malnutrition (Reis et al. 1997).

Fetal malnutrition in rat pups results in a reduced number of  cells, a diminution in the proliferation of islet cells, reduced islet size and a marked decrease in islet vascularization (Snoeck et al. 1990). Following nutritional rehabilitation, the glucose-tolerance test is still impaired in adult life (Dahri et al. 1991). However, the long-term consequences of intrauterine and postnatal protein malnutrition on the early steps of the insulin signaling pathway are unknown.

In this study, we evaluated the secretion and action of insulin in adult rats exposed to protein deficiency during gestation and lactation and examined the influence of nutritional recovery on this response. We investigated the secretion of insulin by isolated pancreatic islets in response to glucose and the ability of insulin to phosphorylate the IR and IRS-1 and to promote the association of the latter with PI 3-kinase in the muscle of protein-deprived and recovered rats.

	MATERIALS AND METHODS

Abstract Introduction Methods Results Discussion References

Materials.  P Collagenase was from Boehringer Mannheim (Indianapolis, IN). Antiserum against insulin was kindly provided by Dr. Leclercq-Meyer (Faculty of Medicine, Brussels Free University, Brussels, Belgium). Standard insulin was from Novo-Nordisk (Copenhagen, Denmark). Reagents for sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting were from Bio-Rad (Richmond, CA). Tris, phenylmethylsulfonylfluoride (PMSF), aprotinin, dithiothreitol, Triton X-100, Tween 20, glycerol, bovine serum albumin (fraction V) and other chemicals were from Sigma Chemical Co. (St. Louis, MO). Antisera against IR, PI 3-kinase (p85) and IRS-1 were from Santa Cruz Biotech (Santa Cruz, CA). Monoclonal antiphosphotyrosine antibodies were from UBI (Lake Placid, NY). [¹²⁵I]Protein A was from Amersham (Amersham, United Kingdom) and protein A Sepharose 6MB was from Pharmacia (Uppsala, Sweden). Nitrocellulose membranes (BA85, 0.2 mm) were from Schleicher & Schuell (Keene, NH). Sodium amobarbital (Amytal) and human recombinant insulin (Humulin R) were from Eli Lilly Co. (Indianapolis, IN).

Buffer A consisted of 100 mmol/L of Tris, 10 g/L of SDS, 50 mmol/L of (N-[2-Hydroxyethyl]piperazine-N'-[2-ethanesulfonic acid]) (pH 7.4), 100 mmol/L of sodium pyrophosphate, 100 mmol/L of sodium fluoride, 10 mmol/L of ethylenediaminetetraacetic acid and 10 mmol/L of sodium vanadate. Buffer B was similar to buffer A except that 10 g/L of Triton X-100 replaced 10 g/L of SDS, and 2 mmol/L of PMSF and 0.1 g of aprotinin/L were added. Buffer C contained 100 mmol/L of Tris, 10 mmol/L of sodium vanadate, 10 mmol/L of EDTA and 10 g/L of Triton X-100.

Animals.  All of the animal experiments were approved by the State University of Campinas Ethics Committee (São-Paulo, Brazil). Virgin female Wistar rats (85-90 d old) were obtained from the University's own breeding colony. Mating was performed by housing males with females overnight, and pregnancy was confirmed by the examination of vaginal smears for the presence of sperm. Pregnant females were separated at random and maintained from the first day of pregnancy until the end of lactation on isocaloric diets containing 6% protein (LP diet) or 17% protein (control diet) as described previously (Reis et al. 1997). Three groups of adult male rats (90 d old) were used in this study: i) a control group (C) consisting of rats born to and suckled by mothers fed a control diet, and subsequently fed a control diet after weaning, ii) an LP group consisting of the offspring of mothers fed an LP diet during both pregnancy and lactation and subsequently fed the same diet after weaning and iii) a recovered group (R) consisting of the offspring of mothers fed an LP diet during pregnancy and lactation, but fed a control diet after weaning. All of the offspring were weaned at the 4th wk after birth. Throughout the experimental period, the rats were given free access to food and water. They were kept under standard lighting conditions (12-h light/dark cycle) at a temperature of 24°C. The rats were weighed at birth, and at 30 and 90 d thereafter. The food intake was recorded three times per week. At the end of the experimental period, one group of rats was killed by decapitation. Blood samples were collected, allowed to clot and the sera stored at 20°C for the subsequent measurement of insulin by radioimmunoassay (Scott et al. 1981). The following determinations were performed immediately after decapitation: serum glucose (Trinder 1969), serum albumin (Doumos et al, 1971), serum-free fatty acids (Regouw et al. 1971) and liver glycogen (Hassid and Abrahams 1957, Sjörgreen et al. 1938).

Glucose-tolerance test.  An oral glucose-tolerance test (OGTT) was performed when the rats of the three groups were 90 d old. After 15 h of food deprivation, glucose (200 g/L) was administered orogastrically through a gastric catheter at a dose of 2 g/kg of body weight. Blood samples were obtained from the cut tip of the tail 0, 30, 60 and 120 min later for the determination of serum glucose and insulin concentrations (Scott et al. 1981, Trinder 1969). The glucose and insulin responses during the glucose-tolerance test were calculated by estimation of the total area under the glucose (G) and insulin (I) curves, respectively, using the trapezoidal method (Matthews et al. 1990).

Insulin-tolerance test.  An intravenous insulin-tolerance test (ivITT) was performed with 90 d-old rats after 15 h of food deprivation. The ITT consisted of a bolus injection of regular insulin (0.18 µmol/100 g of body weight) into the dorsal penile vein. Blood samples were obtained from the cut tip of the tail 0, 5, 10, 15 and 30 min later for the measurement of glucose levels. From 5 to 30 min after the intravenous injection of insulin, the serum glucose concentration declined linearly, and the rate constant for serum glucose disappearance (K_itt) was calculated according to Lundbaek (1962).

Islet isolation and insulin secretion.  The pancreas was removed from 90-d-old rats and digested with collagenase as described previously (Boschero et al. 1995). To measure insulin secretion, groups of five islets were first incubated for 30 min at 37°C in Krebs-bicarbonate solution containing 5.6 mmol/L of glucose equilibrated with a mixture of 95% of O₂-5% CO₂, pH 7.4. The solution was then replaced with fresh buffer, and the islets were further incubated for 60 min with the following glucose concentrations (mmol/L): 2.8, 5.6, 8.3, 11.1, 16.7 and 27.7. The incubation medium contained 115 mmol/L of NaCl, 5 mmol/L of KCl, 24 mmol/L of NaHCO₃, 1 mmol/L of CaCl₂, 1 mmol/L of MgCl₂ and albumin (1 g/L; bovine serum albumin, fraction V). The insulin release was measured as previously described (Scott et al. 1981) using rat insulin as the standard. The glucose concentration producing a response that was 50% of the maximum (EC₅₀) was calculated as the mean negative logarithm (pD2).

Tissue extraction, immunoblotting and immunoprecipitation.  The rats were anesthetized with sodium amobarbital (15 mg/kg of body weight) and used as soon as anesthesia was assured by loss of the pedal and corneal reflexes. The abdominal cavity was opened, the cava vein exposed and 0.5 mL of normal saline (9 g/L of NaCl) with or without 10 µmol of insulin/L was injected. The insulin dose was chosen based on previous work in which the amount of insulin required to achieve a significant signal was determined (Saad et al. 1995). The bolus injection of insulin leads to a transient rise in the peripheral concentration of this hormone to 5-10 times the postprandial levels at 90 s postinjection. The very high levels of insulin attained suggest that different levels of circulating insulin cannot explain some of the variations observed. After the insulin injection (~90 s), the hind-limb muscle (Musculus gastrocnemius) was excised and immediately homogenized in freshly prepared boiling buffer A for immunoblotting, or freshly prepared ice-cold buffer B for immunoprecipitations. Insoluble material was removed by centrifugation for 45 min at 50,000 × g at 4°C. The protein concentration in the supernatants was determined by the Bradford method (Bradford 1976).

For immunoprecipitation, samples containing 3 mg of total protein were incubated with 15 µL of anti-IR or anti-IRS-1 antiserum at 4°C overnight, followed by the addition of Protein A Sepharose 6MB and mixing for 1 h. The pellets were repeatedly washed in buffer C (five times), resuspended in 50 µL of Laemmli sample buffer and boiled for 5 min prior to loading onto the gel. For immunoblotting, samples of 150 µg of total protein were suspended in 50 µL of Laemmli sample buffer and boiled for 5 min before loading onto 6% SDS-PAGE in a miniature slab gel apparatus. Electrotransfer, blotting and signal detection were performed as previously described (Saad et al. 1995).

Statistical analysis.  The results are presented as the means ± SEM for the number of rats (n) indicated. When working with islets, n refers to the number of experiments performed (120 islets per group per experimental condition). Each experiment was performed with islets from two rats per group. When comparing the C and LP groups, nonpaired t tests were used to analyze the total food intake and total body weight gain during pregnancy, as well as the body weight of newborn rats. When comparing the C, LP and R groups, Bartlett's test for the homogeneity of variances was initially used to check the fit of the data to the assumptions for parametric analysis of variance. Body weight of the weaned and adult rats, serum insulin in the fed state and liver glycogen data were log-transformed to correct for variance heterogeneity or nonnormality (Sokal and Rohlf 1995). These data were analyzed by one-way analysis of variance, followed by the Tukey (for equal n) or Tukey-Kramer (for unequal n) test for individual differences between groups. P values <0.05 were considered to indicate significant difference.

	RESULTS

Abstract Introduction Methods Results Discussion References

Characteristics of the animals.  During pregnancy, the C and LP rats had a similar total food intake and total body weight gain. However, the body weight of newborn C rats was greater than that of LP rats (data not shown). At the beginning of the recovery phase, R and LP rats had similar body weights and in both cases these were significantly lower than that of C rats. Although R rats had a greater final body weight than LP rats at the end of the experimental period, their weights were still significantly lower than that of C rats (Table 1). The total food intake during the experimental period was significantly different among the C, R and LP rats. When expressed per gram of body weight, the food intake was significantly greater in LP rats compared to R rats, while the latter had a significantly higher food intake than C rats (17.0 ± 1.0 g/(100 g · d), n = 21, 10.4 ± 0.5 g/(100 g · d), n = 14 and 6.7 ± 0.2 g/(100 g · d), n = 17, respectively, P < 0.05). However, the feed efficiency, calculated as the ratio of weight gain (g) per gram of food intake over 60 d, was lower in LP rats compared to R rats, but the latter did not differ from C rats (0.10 ± 0.01, 0.21 ± 0.02, and 0.25 ± 0.01, respectively, P < 0.001). Considering that protein content is a major determinant of the feed efficiency, the protein intake appeared to be a limiting factor during postweaning growth. In addition to impaired growth, the LP rats manifested other features typical of protein malnutrition, including hypoalbuminemia, high-serum free fatty acid levels and a greater liver glycogen concentration than the C and R groups (Table 1). In the fed state, the serum glucose concentrations did not differ among the three groups of rats. In contrast, the serum insulin concentration was lower in LP rats than in R rats, and the latter was significantly lower than in C rats (Table 1).

OGTT and ITT.  The basal serum glucose and insulin concentrations were significantly lower in LP rats than in C rats (P < 0.01). In R rats, the basal insulin concentration did not differ from that of C rats and the serum glucose level was not different from that of C or LP rats (Table 2). Mean total areas under the G in response to an oral glucose load were not significantly different among groups. However, although the mean total areas under the I did not differ between the LP and R rats, these values were significantly lower than those of C rats (P < 0.05). This finding indicated that glucose-induced insulin secretion in vivo was altered in the LP and R rats (Table 2). The glucose disappearance rate during an i.v. insulin tolerance test (K_itt) in R rats was not significantly different from that in C rats. In contrast, in LP rats the K_itt was significantly greater (P < 0.05) than in the other two groups, thus indicating an increased sensitivity to insulin (Table 2).

View this table: [in this window] [in a new window]   Table 2. Basal serum glucose and insulin concentrations, total areas under the glucose (G) and insulin (I) curves obtained from the oral glucose-tolerance test, and glucose disappearance rates (Kitt) in control (C), low-protein (LP) and recovered (R) rats1

Glucose-induced insulin secretion.  Basal insulin secretion in the presence of 2.8 mmol/L of glucose was significantly lower in R and LP rats than in C rats (0.02 ± 0.003, 0.03 ± 0.005 and 0.1 ± 0.01 nmol/(islet · h), respectively, P < 0.001). Maximal release was obtained at 27.7 mmol/L of glucose in the three groups (0.61 ± 0.09, 0.44 ± 0.04 and 0.17 ± 0.04 nmol/(islet · h) in the C, R and LP groups, respectively). Except at the maximal dose tested (27.7 mmol/L of glucose), the secretory rates were significantly different in the C and R rats (P < 0.05). In LP rats, however, the insulin release was significantly lower than that of C rats under all experimental conditions (P < 0.001). In islets from all three groups of rats, the glucose-induced insulin secretion followed an S-shaped pattern, with the dose-response curve to increasing concentrations of glucose being shifted to the right in islets of LP and R rats (Fig. 1). The half-maximal release concentration of glucose for C, R and LP islets was 8.4 ± 0.1, 12.4 ± 0.5 and 16.9 ± 1.3 mmol/L of glucose, respectively (P < 0.001).

View larger version (18K): [in this window] [in a new window]   Fig 1. Comparative glucose dose-response curves for islets from control (C), recovered (R) and low-protein (LP) rats. Islets were incubated for 60 min with different concentrations of glucose. The values are the mean ± SEM of four independent experiments expressed as a percentage of the maximal insulin secretion within the same experiment. The half-maximal response was obtained with 8.4 ± 0.1 mmol/L, 12.4 ± 0.5 mmol/Land 16.9 ± 1.3 mmol/L of glucose, respectively, for the C, R and LP rats. The data were analyzed by one-way analysis of variance followed by the Tukey test. Significant differences (P < 0.05) were observed among the three groups. IR, insulin receptor.

Early steps of insulin signaling in rat muscle.  Figure 2A shows the levels of IR protein in rat muscle. In LP rats (n = 6), the IR protein levels detectable by immunoblotting were 42.5 ± 13.5% (P < 0.05) greater than in C rats, whereas in R rats, the levels were not significantly different from those of the other two groups. In contrast, immunoprecipitation with anti-IR antibody and immunoblotting with anti-phosphotyrosine antibody showed that there was no difference in the insulin-stimulated phosphorylation of the IR among the three groups (Fig. 3A). The nutritional status of the rats had no effect on the muscle IRS-1 protein levels as determined by immunoblotting (Fig. 2B). Immunoprecipitates of IRS-1 showed a 70 ± 8.8% greater insulin-stimulated phosphorylation in LP rats (n = 5) compared to C rats (n = 7) (P < 0.05) (Fig. 3B). Although R rats (n = 10) tended to have greater (35 ± 18.3%) insulin-stimulated IRS-1 phosphorylation than C rats, this level was not significantly different from the latter group or from LP rats. To examine the association of the 85 kDa subunit of PI 3-kinase with IRS-1, the same blot was incubated with antibodies to this subunit. As expected, in the three groups of rats, an 85 kDa band was present in the IRS-1 immunoprecipitates after exposure to insulin. Greater insulin-stimulated IRS-1-p85/PI 3-kinase association was detected in the muscle of LP than of C rats (Fig. 4, P < 0.05). In R rats, the association between IRS-1 and PI 3-kinase was not significantly different from that in either the C or LP rats (Fig. 4).

View larger version (33K): [in this window] [in a new window]   Fig 2. Fluorographs obtained following sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) of total extracts of hind-limb muscle from control (C) (Fig 2A, n = 6; Fig. 2B, n = 5), low-protein (LP) (Fig. 2A, n = 6; Fig. 2B, n = 5) and recovered (R) (Fig. 2A, n = 5; Fig. 2B, n = 6) rats. The rats were injected with saline (not shown) or insulin and 90 s later hind-limb skeletal muscle was excised and homogenized with extraction buffer A at 100°C as described in the Materials and Methods section. After centrifugation, aliquots of supernatants containing equal amounts of protein were resolved by SDS-PAGE on 6% polyacrylamide gels, transferred to nitrocellulose, and analyzed using antiinsulin receptor (A) or anti-insulin receptor substrate-1 antibodies (B), in conjunction with [125I] protein A, and then subjected to autoradiography. The means ± SEM (n = 4-6) of arbitrary scanning units are depicted at the bottom of the figure. Different letters indicate significant differences (P < 0.05) based on one-way analysis of variance followed by the Tukey-Kramer test.

View larger version (30K): [in this window] [in a new window]   Fig 3. Fluorographs obtained following sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) of immunoprecipitates from hind-limb muscle from control (C) (Fig. 3A, n = 5; Fig. 3B, n = 7), low-protein (LP) (Fig. 3A, n = 4; Fig. 3B, n = 5) and recovered (R) (Fig. 3A, n = 4; Fig. 3B, n = 10) rats. The rats were injected with saline (not shown) or insulin, and 90 s later hind-limb skeletal muscle was excised and homogenized in ice-cold extraction buffer B as described in the Materials and Methods section. After centrifugation, aliquots of the supernatants containing equal amounts of protein were immunoprecipitated using anti-insulin receptor (A) or anti-insulin-receptor substrate-1 (B) antibodies and then resolved by SDS-PAGE on 6% polyacrylamide gels. The nitrocellulose transfers were blotted using antiphosphotyrosine antibody in conjunction with [125I] protein A, and then subjected to autoradiography. The means ± SEM (n = 4-10) of arbitrary scanning units are depicted at the bottom of the figure. Different letters indicate significant differences (P < 0.05) based on one-way analysis of variance followed by the Tukey-Kramer test.

View larger version (36K): [in this window] [in a new window]   Fig 4. Fluorograph obtained following sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) of immunoprecipitates of hind-limb muscle from control (C) (n = 5), low-protein (LP) (n = 6) and recovered (R) (n = 10) rats. The rats were injected with saline (not shown) or insulin, and 90 s later hind-limb skeletal muscle was excised and homogenized in ice-cold extraction buffer B as described in the Materials and Methods section. After centrifugation, aliquots of the supernatant containing equal amounts of protein were immunoprecipitated using anti-insulin-receptor substrate (IRS)-1 antibodies and then resolved by SDS-PAGE on 6% polyacrylamide gels. The nitrocellulose transfers were blotted using anti-p85 (phosphatidylinositol 3-kinase) antibody in conjunction with [125I] protein A, and then subjected to autoradiography. The means ± SEM (n = 5-10) of arbitrary scanning units are depicted at the bottom of the figure. Different letters indicate significant differences (P < 0.05) based on one-way analysis of variance followed by the Tukey-Kramer test.

	DISCUSSION

Abstract Introduction Methods Results Discussion References

Rats fed a LP diet exhibited features typical of protein malnutrition such as low body weight, hypoalbuminemia, high liver glycogen and high serum free fatty acid levels. In recovered rats, these metabolic alterations were reversed, but insulin secretion in response to glucose was only partially restored. This finding indicates that nutritional recovery can only attenuate the damage to pancreatic  cells produced by protein deprivation early in life.

Protein deprivation during pregnancy was reported to produce pups with decreased neonatal  cell proliferation, and a reduced islet size vascularization (Snoeck et al. 1990). The maintenance of protein deprivation until adult age can decrease the insulin content of isolated pancreatic islets; no obvious alteration remains at adult age when the protein deficiency is restricted to fetal life (Rasschaert et al. 1995). Thus, the reduced islet insulin content may explain the impaired glucose-stimulated insulin release noted above, at least in the LP group. However, the shift to the right in the glucose dose-response curves observed in the LP and recovered groups suggests the involvement of one or more intrinsic defects in the mechanism of insulin secretion. Malnourished rats show reduced activities of pancreatic glucokinase (Hales et al. 1996) and  cell mitochondrial glycerophosphate dehydrogenase (Rasschaert et al. 1995), which could lead to impaired glucose metabolism. The islets of malnourished rats also show alterations in their Ca²⁺ handling (Carneiro et al. 1995). The possible detrimental effect of elevated free fatty acid levels (Zhou and Grill 1994) may contribute to the defective glucose-induced insulin release seen in the LP group.

Protein-deprived rats exhibited lower basal glucose levels but similar total areas under the glucose curves to those observed in control rats. In addition, they showed high levels of hepatic glycogen, attributed to the inefficiency of glucagon (Okitolonda et al. 1987). This impaired glucagon action may reflect the low activity of some enzymes such as glucose-6-phosphatase and liver and muscle alanine amino transferase reported in a similar model of malnutrition (Heard et al. 1977). Apart from the enzymatic changes that suggest reduced endogenous glucose production (Desai et al. 1995), the maintenance of glucose tolerance in the face of drastically decreased serum insulin levels has been ascribed to a high hepatic and peripheral sensitivity to insulin (Crace et al. 1990, Escriva et al. 1991, Okitolonda et al. 1987, Reis et al. 1997). This was confirmed in our study by the increased serum glucose disappearance rates (K_itt) following the i.v. injection of insulin in malnourished rats.

Altered enzymatic activities such as decreased glucokinase and increased phosphoenolpyruvate carboxykinase suggest that glucose production rather than utilization is predominant in protein malnutrition. Nutritional recovery does not fully restore these changes in liver enzyme activity produced by protein malnutrition in early life (Desai et al. 1995). Nevertheless, under food deprivation and feeding conditions as well as during the oral glucose-tolerance test, recovered rats maintained a low or near-normal glycemia, despite low-serum insulin levels. The maintenance of normoglycemia may result either from mechanisms that are independent of any dynamic insulin response (increased glucose effectiveness) or from mechanisms dependent on insulinemia (increased insulin sensitivity). The latter possibility was not confirmed by the K_itt value of our study.

Previous reports (Crace et al. 1990, Okitolonda et al. 1987) have suggested that nutritional recovery only partially reverses the high tissue sensitivity to insulin. We therefore investigated the changes in insulin sensitivity by evaluating the early molecular steps of insulin signaling in skeletal muscle since this tissue is one of the main targets of insulin action and comprises about 40% of the total body weight (Crace et al. 1990). The increased insulin sensitivity of liver and skeletal muscle in protein-deprived rats has been demonstrated (Crace et al. 1990). This in vivo insulin sensitivity in peripheral tissue and liver was confirmed using the euglycemic-hyperinsulinemic clamp technique (Escriva et al. 1991) and, more recently, by the insulin signaling system (Reis et al. 1997). In contrast to a previous study that showed no significant effect of protein deprivation on the levels of IR in the muscle of malnourished rats (Reis et al. 1997), we found that protein deprivation up to the 12^th wk of age significantly increased the IR levels in skeletal muscle. No change was observed in the IRS-1 protein levels nor in IR autophosphorylation in these animals. However, the phosphorylation of IRS-1 that reflects the kinase activity of the insulin receptor toward its endogenous substrate was increased and was accompanied by an increase in the association of IRS-1 with the lipid-metabolizing enzyme PI 3-kinase. Nutritional recovery partially restored the changes in the early steps of insulin action, as demonstrated by an intermediary increase in the IR level as well as in the phosphorylation of IRS-1 and its association with PI 3-kinase.

The increases in these three early steps of insulin action following hormone binding may be instrumental in the increased insulin sensitivity observed in both groups. Insulin increases glucose uptake into cells, partially through the translocation of GLUT4 from intracellular compartments to the plasma membrane in muscle and adipose tissues (Stephens and Pilch 1995). Several studies showed that PI 3-kinase is necessary for insulin-stimulated GLUT4 translocation (Okada et al. 1994, Sanches-Margalet et al. 1994). Hence, the pathway involving IRS-1/PI 3-kinase and the increased receptor level may be linked to insulin sensitivity in the muscle of malnourished and recovered rats. This conclusion agrees with a previous study (Ozanne et al. 1996) that showed that in adult rats subjected to a nutritional schedule identical to that used here (protein restriction during pregnancy and lactation, and nutritional recovery after weaning) there was an increase of insulin receptors in skeletal muscle. The mechanisms by which protein restriction induces these alterations are unknown, although it is probable that the reduced insulin secretion seen in these animals may play a role in this phenomenon. Hypoinsulinemia itself can induce alterations in the early steps of insulin action, as demonstrated by prolonged food deprivation and by streptozotocin-induced diabetes in rats (Saad et al. 1992).

Our results indicating an enhanced insulin sensitivity in LP rats disagree with findings associated with protein deficiency in childhood. Kwashiorkor is characterized by an impaired glucose tolerance and insulin secretory response (Becker et al. 1975, Milner 1972). Additional confounding factors not accounted for here may be responsible for the glucose intolerance seen in malnourished humans (Okitolonda et al. 1987). The persistently high tissue sensitivity to insulin seen in recovered animals may have resulted from an incomplete restoration of body weight. A lower body weight associated with decreased serum insulin levels apparently has the opposite, compensatory effect.

In conclusion, our results are consistent with the hypothesis that early protein deprivation produces pancreatic -cell dysfunction that is not completely restored by nutritional recovery. These data also provide direct evidence that in LP and recovered animals glucose homeostasis is maintained, at least in part, at the expense of increased insulin sensitivity. The initial steps in the insulin signal-transduction pathway contribute to the increase in insulin-stimulated glucose uptake in the skeletal muscle of protein-deprived and recovered rats. Further studies are necessary to elucidate the mechanisms involved in the secretory defect and the means by which malnutrition alters these three early steps of insulin action.

	FOOTNOTES

Manuscript received 2 March 1998. Initial reviews completed 15 April 1998. Revision accepted 10 July 1998.

	ACKNOWLEDGMENTS

The authors thank Léscio D. Teixeira and Clarice Yoshico Sibuya for the technical assistance and S. Hyslop for revising the grammar.

	LITERATURE CITED

Abstract Introduction Methods Results Discussion References

• Becker D. J., Pimstone B. L., Hansen J.D.L. The relationship between insulin secretion, glucose tolerance, growth hormone and serum proteins in protein-calorie malnutrition. Pediatr. Res. 1975; 9:35-39
• Boschero A. C., Szpak-Glasman M., Carneiro E. M., Bordin S., Paul I., Rojas E., Atwater I. Oxotremorine-m potentiation of glucose-induced insulin release from rat islets involves M3 muscarinic receptors. Am. J. Physiol. 1995; 268:E336-E342[Abstract/Free Full Text]
• Bradford M. M. A rapid and sensitive method for the quantitation of microgram quantities of protein utilising the principle of protein dye binding. Anal. Biochem. 1976; 72:248-254[Medline]
• Carneiro E. M., Mello M.A.R., Gobatto C. A., Boschero A. C. Low protein diet impairs glucose-induced insulin secretion from and ⁴⁵Ca uptake by pancreatic rat islets. J. Nutr. Biochem. 1995; 6:314-318
• Crace C. J., Swenne I., Kohn P. G., Strain A. J., Milner R.D.G. Protein-energy malnutrition induces changes in insulin sensitivity. Diabete Metab. 1990; 16:484-491[Medline]
• Dahri S., Snoeck A., Reusens-Billen B., Remacle C., Hoet J. J. Islet function in offspring of mothers on low-protein diet during gestation. Diabetes 1991; 40:115-120
• Desai M., Crowther N. J., Ozanne S. E., Lucas A., Hales C. N. Adult glucose and lipid metabolism may be programmed during fetal life. Biochem. Soc. Trans. 1995; 23:331-335[Medline]
• Doumos B. T., Watson W. A., Biggs H. G. Albumin standers and measurements of serum albumin with bromocresol green. Clin. Chim. Acta 1971; 31:87-96[Medline]
• Escriva F., Kergoat M., Bailbé D., Pascual-Leone A. M., Portha B. Increased insulin action in the rat after protein malnutrition early in life. Diabetologia 1991; 34:559-564[Medline]
• Hales C. N., Barker D.J.P. Type 2 (non-insulin-dependent) diabetes mellitus: the thrifty phenotype hypothesis. Diabetologia 1992; 35:595-601[Medline]
• Hales C. N., Barker D.J.P., Clark P.M.S., Cox L. J., Fall C., Osmond C., Winter P. D. Fetal and infant growth and impaired glucose tolerance at age 64. Br. Med. J. 1991; 303:1019-1022
• Hales C. N., Desai M., Ozanne S., Crowther N. J. Fishing in the stream of diabetes: from measuring insulin to the control of fetal organogenesis. Biochem. Soc. Trans. 1991; 24:341-350
• Hassid W. Z., Abrahams S. Chemical procedures for analysis of polysaccharides. Meth. Enzymol. 1957; 3:34-36
• Heard C.R.C., Frangi S. M., Wright P. M., McCartney P. R. Biochemical characteristics of different forms of protein energy malnutrition: an experimental model using rats. Br. J. Nutr. 1977; 37:1-21[Medline]
• James W.P.T., Coore H. G. Persistent impairment of insulin secretion and glucose tolerance after malnutrition. Am. J. Clin. Nutr. 1970; 23:386-389[Abstract]
• Lundbaek K. Intravenous glucose tolerance as a tool in definition and diagnosis of diabetes mellitus. Br. Med. J. 1962; 3:1057-1513
• Matthews J.N.S., Altman D. G., Campbell M. J., Royston P. Analysis of serial measurements in medical research. Br. Med. J. 1990; 27:230-235
• Milner R.D.G. Insulin secretion in human protein-calorie deficiency. Proc. Nutr. Soc. 1972; 31:219-223[Medline]
• Okada T., Kawano Y., Sakahihara T., Hazeki O., Ui M. Essential role of phosphatidylinositol 3-kinase in insulin induced glucose transport and antilipolysis in rat adipocytes. Studies with a selective inhibitor wortmannin. J. Biol. Chem. 1994; 269:3568-3573[Abstract/Free Full Text]
• Okitolonda W., Brichard S. M., Henquin J. C. Repercussions of chronic protein-calorie malnutrition on glucose homeostasis in the rat. Diabetologia 1987; 30:946-951[Medline]
• Ozanne S. E., Wang C. L., Coleman N., Smith G. D. Altered muscle sensitivity in the male offspring of protein-malnourished rats. Am. J. Physiol. 1996; 271:E1128-E1134[Abstract/Free Full Text]
• Phillips D.I.W. Insulin resistance as a programmed response to fetal undernutrition. Diabetologia 1996; 39:1119-1122[Medline]
• Poulsen P., Vaag A. A., Kyvik K. O., Møller Jensen D., Beck-Nielsen H. Low birth weight is associated with NIDDM in discordant monozygotic and dizygotic twin pairs. Diabetologia 1997; 40:439-446[Medline]
• Rasschaert J., Reusens B., Dahri S., Sener A., Remacle C., Hoet J. J., Malaisse W. Impaired activity of rat pancreatic islets mitochondrial glycerophosphate dehydrogenase in protein malnutrition. Endocrinology 1995; 136:2631-2634[Abstract]
• Regouw B.J.M., Cornelissen P.J.H.C., Helder A. P., Spijkers J.B.V., Weeber Y. M. Specific determination of free fatty acid in plasma. Clin. Chim. Acta 1971; 31:187-195[Medline]
• Reis M.A.B., Carneiro E. M., Mello M.A.R., Boschero A. C., Saad M.J.A., Velloso L. A. Glucose-induced insulin secretion is impaired and insulin-induced phosphorylation of the insulin receptor and insulin receptor substrate-1 are increased in protein-deficient rats. J. Nutr. 1997; 127:403-410[Abstract/Free Full Text]
• Saad M.J.A., Araki E., Miralpeix M., Rothenberg P. L., White M. F., Kahn C. R. Regulation of insulin receptor substrate-1 in liver and muscle of animal models of insulin resistance. J. Clin. Invest. 1992; 90:1839-1849
• Saad M.J.A., Hartmann L.G.C., Carvalho D. S., Galoro C.A.O., Brenelli S. L., Carvalho C.R.O. Effect of glucagon on insulin receptor substrate-1 (IRS-1) phosphorylation and association with phosphatidylinositol 3-kinase (PI 3-kinase). FEBS Lett. 1995; 370:131-134[Medline]
• Sanches-Margalet V., Goldfine I. D., Vlahos C. J., Sung C. K. Role of phosphatidylinositol 3-kinase in insulin receptor signaling: studies with inhibitor LY 294002. Biochem. Biophys. Res. Commun. 1994; 204:446-452[Medline]
• Scott A. M., Atwater I., Rojas E. A method for the simultaneous measurement of insulin release and B-cell membrane potential in single mouse islets of Langerhans. Diabetologia 1981; 21:407-475
• Sjörgreen B., Nordenskjold T., Holmgren H., Wollerstron J. . Bertrag zur kemtmis des leberrhythmic. Pflügers Arch. Gesante Menschem Tierr. 1938; 240:247
• Snoeck A., Remacle C., Reusens B., Hoet J. J. Effect of a low protein diet during pregnancy on the fetal rat endocrine pancreas. Biol. Neonate 1990; 57:107-118[Medline]
• Sokal, R. R. & Rohlf, F. J. (1995) Assumptions of analysis of variance. In: Biometry: The Principles and Practice of Statistics in Biological Research (Sokal, R. R. & Rohlf, F. J., eds.), pp. 392-450. W. H. Freeman and Company, New York.
• Stephens J. M., Pilch P. F. The metabolic regulation and vesicular transport of GUT4, the major insulin-responsive glucose transporter. Endocrine Rev. 1995; 16:529-546[Abstract/Free Full Text]
• Trinder P. Determination of blood glucose using an oxidase-peroxidase system with a non-carcinogenic chromogen. J. Clin. Path. 1969; 22:158-161[Abstract/Free Full Text]
• Zhou Y. P., Grill V. E. Long-term exposure of rat pancreatic islets to fatty acids inhibits glucose-induced insulin secretion and biosynthesis through a glucose fatty acid cycle. J. Clin. Invest. 1994; 93:870-876

0022-3166/98 $3.00 ©1998 American Society for Nutritional Sciences