Insulin resistance during pregnancy is counteracted by enhanced insulin secretion. This condition is aggravated by obesity, which increases the risk of gestational diabetes. Therefore, pancreatic islet functionality was investigated in control nonpregnant (C) and pregnant (CP), and cafeteria diet-fed nonpregnant (Caf), and pregnant (CafP) obese rats. Isolated islets were used for measurements of insulin secretion (RIA), NAD(P)H production (MTS), glucose oxidation (14CO2 production), intracellular Ca2+ levels (fura-2 AM), and gene expression (real-time PCR). Impaired glucose tolerance was clearly established in Caf and CafP rats at the 14th wk on a diet. Insulin secretion induced by direct depolarizing agents such as KCl and tolbutamide and increasing concentrations of glucose was significantly reduced in Caf, compared with C islets. This reduction was not observed in islets from CP and CafP rats. Accordingly, the glucose oxidation and production of reduced equivalents were increased in CafP islets. The glucose-induced Ca2+ increase was significantly lower in Caf and higher in CafP, compared with all other groups. CP and CafP islets demonstrated an increased Ca2+ oscillation frequency, compared with both C and Caf islets, and the amplitude of oscillations was augmented in CafP, compared with Caf islets. In addition, Cavα1.2 and SERCA2a mRNA levels were reduced in Caf islets. Cavα1.2, but not SERCA2a, mRNA was normalized in CafP islets. In conclusion, cafeteria diet-induced obesity impairs insulin secretion. This alteration is related to the impairment of Ca2+ handling in pancreatic islets, in especial Ca2+ influx, a defect that is reversed during pregnancy allowing normalization of insulin secretion.
- insulin resistance
- pancreatic β-cells
- metabolic syndrome
pregnancy is associated with peripheral insulin resistance, which is compensated by increased insulin secretion during normal circumstances (3, 19, 30). Pancreatic islets undergo major structural and functional changes during pregnancy to fulfill this increased demand for insulin (1, 30, 34). These alterations in rats peak at around days 14–16 of pregnancy, and the placental lactogens and/or prolactin hormones play an important role in this process (6, 8, 23, 33).
The mechanisms responsible for increasing the capacity of the β-cells to respond to a higher demand of insulin during pregnancy is very relevant in the context of Type 2 diabetes studies, since it may provide clues for potential therapeutics. The inability of maternal β-cells to respond to this increased demand for insulin can lead to the development of glucose intolerance and to gestational diabetes mellitus (10, 11).
Several models of experimental obesity have been used to date. Among them, the use of the fat-rich diet is an interesting approach, since it closely resembles the overfeeding intake of humans, affecting specific tissues involved in the regulation of energy expenditure (9, 36). An improvement for this model is the use of the cafeteria diet, which is even closer to human food intake, since it is more palatable, strongly increases adiposity, and has been proposed to be the rodent model that best fits human obesity (25, 27, 28).
In this study, we investigated glucose homeostasis and pancreatic islet functionality in nonpregnant and pregnant cafeteria diet-induced obese rats. We found that glucose tolerance is impaired in obese cafeteria-fed pregnant and nonpregnant rats. Cafeteria diet-induced obesity impairs insulin secretion induced by glucose, tolbutamide, and KCl in freshly isolated islets. However, this inhibitory effect is overcome in islets from pregnant rats, probably as a result of an increase in metabolic activity, associated with a better intracellular Ca2+ handling, in especial Ca2+ influx.
MATERIALS AND METHODS
All applicable institutional and governmental regulations concerning the ethical use of animals were followed during this research study. The experimental procedures were approved by the university's Committee for Ethics in Animal Experimentation from the State University of Campinas (protocol number 1198-1).
d-[U-14C]glucose and 125I-human recombinant insulin were purchased from G. E. Health Care (Little Chalfont, Buckinghamshire, UK). MTS/PMS preparation was from CellTiter96 aqueous assay (Promega, Madison, WI). Standard commercial kits were used for measurement of plasma total cholesterol (CHOL), triglycerides (TG) (both from Roche Diagnostics; Mannheim, Germany) free fatty acids (FFA; Wako Chemicals, Neuss, Germany), and albumin and total proteins (Laborlab; Guarulhos, SP, Brazil). Fura-2 AM was purchased from Invitrogen (Carlsbad, CA). Routine reagents were purchased from Sigma-Aldrich (St. Louis, MO).
Animals, dietary regimen and composition of diet.
Virgin female Wistar rats (70 days old, obtained from the State University of Campinas Central Breeding Centre) were randomly assigned to two diet groups: a standard rodent chow and water ad libitum or cafeteria diet for 14 wk. The cafeteria diet group received soft drinks ad libitum, instead of water, alternated daily (Coca-Cola and Guaraná Antarctica) and were fed a pellet made of 37.5% standard rodent chow, 25% peanuts, 25% chocolate, and 12.5% cookies, offered together with palatable food items comprising wafer, snacks, cakes, and biscuits, totaling 4.41 kcal/g (43.1% from carbohydrates; 12.1% from proteins, and 46.9% from fats) as opposed to the 2.63 kcal/g of the standard chow diet Nuvilab CR-1 (Nuvital, Colombo, PR, Brazil). To guarantee the success of this protocol, leftovers were collected and replaced with new items daily. During the feeding period, rats’ weight gain was measured weekly. After 12 wk of treatment, the rats were mated by housing males with females overnight during 3 days, and the presence of sperm in vaginal smears was verified each morning. On the occasion of presence of sperm (day 1 of pregnancy), the female rat was housed separately for confirmation of pregnancy and further utilized in experiments on the 15th/16th day of pregnancy, corresponding the 14th wk of treatment. During the experimental period, rats were housed at 22 ± 2°C on a 12:12-h light-dark cycle (lights on 0600–1800).
At the end of the feeding period, and on the 15th/16th day of pregnancy, the rats were killed by decapitation, and perigonadal and retroperitoneal fat pad weights were measured. Blood glucose levels were measured using a glucose analyzer (Accu-Check Advantage II, Roche, Basel, Switzerland). Plasma CHOL, TG, FFA, albumin, and total proteins were measured using standard commercial kits, according to the manufacturer's instructions. Insulin was measured by RIA using rat insulin as standard.
Intraperitoneal glucose tolerance test.
At 14–15 days after the onset of pregnancy, all groups of rats were submitted to an intraperitoneal glucose tolerance test (ipGTT). Food was withdrawn 12 h before the experiment, and then the rats were weighed, and a basal blood sample was taken from the tip of the tail (t = 0 min). Subsequently, each rat received a glucose solution load (2 g/kg ip body wt), and additional blood samples were collected at 15, 30, 60, and 120 min after injection. Glucose levels during the test were measured immediately. The area under the curve was calculated from values for each rat.
Islet isolation, insulin secretion, and insulin content.
Islets were isolated from fed rats (pregnant or not; 14 wk of treatment, 15th/16th day of pregnancy) by collagenase digestion of pancreas and then selected with a micropipette under a microscope to exclude any contaminating tissues. Groups of four islets were first incubated for 45 min at 37°C in Krebs-Ringer bicarbonate buffer (KRB) containing glucose 5.6 mmol/l and equilibrated with 95% O2-5% CO2, pH 7.4. The solution was then replaced with fresh KRB, and the islets were incubated for a further 90-min period with medium containing 2.8, 5.6, 8.3, 11.1, 16.7, or 27.7 mmol/l glucose; 2.8 mmol/l glucose plus 40 mmol/l KCl; or 2.8 mmol/l glucose plus 100 μmol/l tolbutamide. The incubation medium contained (mmol/l): 115 NaCl, 5 KCl, 24 NaHCO3, 2.6 CaCl2, 1 MgCl2, and 25 HEPES; pH 7.4, supplemented with BSA (0.3% wt/vol; Sigma). For measurement of the total insulin content, groups of 10 islets were collected and transferred to tubes of 1.5 ml. Alcohol-acid solution (1 ml; final concentration of 20% of ethanol and 0.2 mmol/l of HCl) was added to the samples followed by sonication of the pancreatic islets (3 times, 10-s pulses). The insulin in the media was measured by RIA.
Pancreatic islets were homogenized through short bursts of sonication in 500 μl of buffer composed of 50 mmol/l Tris·HCl, 10 mM EDTA, 1% SDS (pH 8.1). DNA was extracted in phenol/chloroform, precipitated in ethanol, and resuspended in low Tris-EDTA buffer. RNA was subsequently removed by digestion with 1 μg of RNase A (Sigma) for 30 min at 37°C. Thus, DNA was quantified using a commercial kit (Quant-iT PicoGreen, Invitrogen), according to the instructions in the manual.
Groups of 25 islets were incubated for 2 h at 37°C in KRB supplemented with 11.1 mmol/l glucose with trace amounts of d-[U-14C]glucose (20 μCi/ml) for 14CO2 formation. Islet glucose metabolism was stopped with HCl (1 N) with consequent cell cleavage. 14CO2 released was absorbed by NaOH (1 mol/l) for 1 h at 4°C, obtaining NaH14CO3. Scintillation fluid was added, and radioactivity was counted in a liquid scintillation counter.
Metabolic activity of islets was assessed by measuring reducing equivalents, namely NAD(P)H, through reduction of a water-soluble tetrazolium salt, MTS (3-[4,5,dimethylthiazol-2-yl]-5-[3-carboxymethoxy-phenyl]-2-[4-sulfophenyl]-2H -tetrazolium, inner salt) to its respective formazan product in a living tissue system (29, 33). For this purpose, groups of 100 freshly isolated islets were incubated for 150 min in Krebs/HEPES sterile buffer containing 11.1 mmol/l glucose, 15% MTS, and 1% phenazine methosulfate (PMS). The absorbance was acquired at 490 nm, every 10 min after the addition of the reagent solution, according to the manufacturer's instructions.
Measurement of intracellular Ca2+.
Freshly isolated islets were incubated in RPMI 1640 medium supplemented with 5% fetal calf serum, 11.1 mmol/l glucose, 100 UI of penicillin/ml, 100 μg of streptomycin/ml at 37°C in a 95% O2-5% CO2 for 4–6 h. Thereafter, islets were transferred to plates containing KRB medium (mmol/l concentrations were 115 NaCl, 5 KCl, 24 NaHCO3, 1 MgCl2, 2.6 CaCl2, 25 HEPES; pH 7.4) with 5 mmol/l glucose and 5 μmol/l fura-2 AM for 1 h. Islets were then transferred to a thermostatically regulated open chamber (37°C), placed on the stage of an inverted microscope (Nikon UK, Kingston, UK), and perifused with KRB at a flow rate of 1.5 ml/min. A ratio image was acquired approximately every 5 s with an ORCA-100 charge-coupled device camera (Hammamatsu Photonics Ibérica, Barcelona, Spain), in conjunction with a Lambda-10-CS dual-filter wheel (Sutter Instrument Company, Novato, CA), equipped with 340 and 380 nm, 10-nm bandpass filters, and a range of neutral density filters (Omega Opticals, Stanmore, UK). Data were obtained using the ImageMaster5 software (Photon Technology International, Birmingham, NJ). The amplitude and frequency of Ca2+ oscillations were determined after the first peak following the introduction of 11.1 mmol/l glucose in the medium. For the amplitude, the first increase in Ca2+ concentration was calculated subtracting basal values (2.8 mmol/l) from values obtained at the zenith (11.1 mmol/l). The amplitude of oscillations was calculated by subtracting the highest from the lowest value of individual oscillation at 11.1 mmol/l glucose. We considered as oscillation every increment in Ca2+ fluorescence ratio (F340/F380) at least two times greater than any change in F340/F380 at basal glucose concentration. The frequency was calculated by counting the number of oscillations during 11.1 mmol/l glucose, and it was expressed as oscillations per minute. In the experiments with tolbutamide, the area under the curves was calculated during the period that the drug was present in the medium, after subtracting the basal values (2.8 mmol/l glucose).
Total cellular RNA was extracted from groups of 500 islets using TRIzol reagent. Two micrograms of total RNA were reverse transcribed using a reverse transcriptase and random hexamer primers. Real-time PCR reactions were performed in a total volume of 15 μl using the Fast SYBR Green technology (Applied Biosystems, Foster City, CA). Samples were denatured at 94°C for 10 min followed by 40 PCR cycles at 95°C/60°C. PCR amplifications were performed in duplicate. The purity of the amplified PCR products was verified by melting curves. The expression of the target genes was normalized against the expression levels of the housekeeping gene GAPDH. Sequence of the primers used were (5′-3′): sarcoplasmic/endoplasmic reticulum Ca2+ ATPase 2a (SERCA2a) forward: TGGTACTGGCTGATGATAACTTCTCC, reverse: TGTTGTTGTAGATGGCACGGC; L-type-α1.2 subunit voltage-sensitive Ca2+ channel (CaVα1.2) forward: GACACAGAGAGGAAGTTCAAGGG, reverse: GCGTGGGCTCCCATAGTTG; L-type-β2 subunit voltage-sensitive Ca2+ channel (CaVβ2) forward: TGCACTGGAGTATCCAAGCG, reverse: CCACTTCGTCTCAGCCACTC.
Data were expressed as means ± SE for the number of rats and samples (n) indicated. Statistical analysis was performed by the Student's t-test or two-way ANOVA, followed by the Newman-Keuls posttest. P < 0.05 was considered statistically significant.
Cafeteria-diet-fed rats showed a greater weight gain than control chow diet-fed rats from the 3rd until the 12th wk of the feeding period, when females were mated with males (Fig. 1). At the end of the diet period (14th wk), body weight, retroperitoneal, and perigonadal fat pad weights were heavier in Caf and CafP rats, compared with the C and CP groups (Table 1). CafP and CP rats presented higher body weight than Caf and C rats, respectively, because of the contribution of the weight of the fetuses. After 12 h of fasting, plasma insulin, TG, FFA, albumin, and total plasma protein levels were similar for the groups (Table 2). Fasting blood glucose and plasma (CHOL) levels were lower in CP and CafP than the C and Caf groups. In the fed state, plasma insulin and FFA levels were higher in Caf and CafP than C and CP rats. Fed blood glucose levels were diminished in both CP and CafP rats, compared with nonpregnant groups. In CP rats, fed TG plasma levels were higher than those of the C group. CHOL, albumin, and total plasma proteins values were similar in the groups during the fed state. Insulin/glucose index was higher in Caf and CafP than C and CP groups in both fasting (12 h) and fed conditions.
As installation of insulin resistance in diet-induced obesity is time dependent and tissue specific (25), our first studies were performed in rats fed on a cafeteria diet for 8 wk. After this period, we observed an increase in weight gain, but only a marginal impairment in glucose tolerance in the cafeteria-diet-fed rats (data not shown) and, for this reason, we decided to extend the diet period for 6 additional weeks. After 14 wk, cafeteria diet-induced obesity provoked impaired glucose tolerance in pregnant and nonpregnant rats (Fig. 2A). This effect was dependent only on obesity, rather than on the pregnancy state or on the interaction between both conditions. At 30 min after glucose challenge, blood glucose concentrations were higher in Caf than C rats (P < 0.05) and in CafP than C and CP rats (P < 0.05). After 1 h, glycemia was still higher in Caf and CafP rats than C and CP rats (Fig. 2A). The area under the curve in response to glucose load was higher in Caf and CafP than in control groups [32,370 ± 1,680 Caf (n = 19), 31,280 ± 3,015 CafP (n = 6) vs. 22,800 ± 1,095 C (n = 18), 20,500 ± 885 CP (n = 6), respectively; mg·dl−1·min−1] (P < 0.01). This effect, at least in the CafP rats, was probably due to an increase in insulin resistance, since plasma insulin during the GTT (Fig. 2B) in CafP rats at 15 min (interaction among obesity and pregnancy) and 1 h (effect of obesity) was higher than that of the other groups.
Islet insulin content and insulin secretion induced by glucose, KCl, and tolbutamide.
Pregnancy enhanced total insulin content in islets from CafP rats compared with the other groups (43.7 ± 3.7 vs. C 15.6 ± 1.2; CP 19.1 ± 1.5; Caf 12.6 ± 1.1, ng/ng DNA P < 0.001; n = 14–19). This effect was dependent on both obesity and the pregnancy states and also on the interaction between both conditions. Total insulin content was similar in C, CP, and Caf islets. The insulin released by islets during static incubation in response to increasing glucose concentrations (2.8–27.7 mmol/l), normalized by the islet DNA content, is shown in Fig. 3. In all groups, insulin secretion was represented by a sigmoidal curve. At 2.8 mmol/l glucose, insulin released was higher in Caf and CafP compared with C and CP islets (P < 0.05). At 5.6 mmol/l glucose, secretion was higher in islets from pregnant (CP and CafP) than nonpregnant rats (C and Caf) (P < 0.05). The insulin secretion, in the presence of 8.3 mmol/l glucose, was lower in Caf compared with CP and CafP groups (P < 0.01). At 11.1 mmol/l glucose, the insulin secretion was significantly lower in Caf compared with C, CP, and CafP islets (P < 0.05). Finally, at 16.7 mmol/l glucose, but not at 27.7 mmol/l, insulin secretion was higher in CafP than Caf islets (P < 0.05). Pregnancy shifted the dose-response curve to the left in both chow (CP) and cafeteria diet (CafP), compared with nonpregnant rats (C and Caf groups). When the insulin secretion was expressed as a percentage of the maximal release inside each group (Fig. 3, inset), the half-maximal release (EC50) was higher in Caf compared with other groups (12.3 ± 0.6, 10.6 ± 0.2, 9.3 ± 0.1, and 9.6 ± 0.5 mmol/l, respectively for Caf, C, CP, and CafP; P < 0.05). It is of note that normalization of insulin secretion by total insulin content shifted the glucose dose-response curve to the right in the CafP group (not shown). However, it seems that this kind of normalization does not reflect the capacity of individual islets to secrete insulin, as judged by the results normalized to total DNA content. Insulin secretion stimulated by 40 mmol/l KCl or 100 μmol/l tolbutamide, in the presence of 2.8 mmol/l glucose, was strongly reduced in Caf islets compared with all of the other groups, and higher in CafP than CP islets (P < 0.05; n = 8–10) (Fig. 4). These results indicate that Caf islets also secreted less insulin than the other groups when the islet cells were depolarized by agents that bypass nutrient metabolism and that pregnancy restores insulin secretion in those experimental conditions.
Glucose oxidation and metabolic activity.
Because insulin secretion from β-cells is closely linked to metabolism of glucose, we analyzed some metabolic indicators in islets from obese and pregnant rats. At 11.1 mmol/l glucose, the d-[U-14C]glucose conversion to 14CO2 (glucose oxidation) was higher in islets from CafP rats, compared with the other groups (CafP 2.9 ± 0.1 vs. C 2.1 ± 0.2; CP 2.0 ± 0.1, and Caf 1.9 ± 0.1 pmol/ng DNA·2 h; P < 0.001; n = 12). This effect was dependent on both obesity and the pregnancy state and also on the interaction between both conditions. We also measured the generation of reducing equivalents [i.e., NAD(P)H] in the presence of 11.1 mmol/l glucose, as an indicator of the metabolic activity of the islets (33). CafP showed a higher metabolic activity compared with C and CP islets already at 10 min, and this activity was higher compared with all other groups from 30 up to 150 min (effect of both obesity and pregnancy with interaction) (Fig. 5). It is of note that CP also showed higher glucose oxidation and metabolic activity compared with C group in absolute values, a difference not seen after normalization to total DNA content (not shown).
Glucose- and tolbutamide-induced cytoplasmic Ca2+ alterations.
After exposure of pancreatic islets to 11.1 mmol/l glucose, cytosolic Ca2+ concentrations were increased in all groups (Fig. 6, A–D). The amplitude of the Ca2+ alterations, measured after 3–6 min glucose exposure, was significantly lower in islets from Caf, compared with the other three groups (Fig. 6E) (P < 0.05; with interaction among obesity and pregnancy). The frequency of oscillations was similar between islets from C and Caf and increased significantly in islets from both CP and CafP groups (Fig. 6F) (P < 0.05). Finally, the amplitude of the oscillations was higher in CafP, compared with Caf islets (P < 0.05). The cytosolic Ca2+ concentrations were also increased by tolbutamide (100 μmol/l) in all groups, and the area under the curve was lower in Caf compared with C islets (0.72 ± 0.14 vs. 1.20 ± 0.10 F340/F380·min; P < 0.05; n = 6), an effect of obesity without interaction.
Cavα1.2, Cavβ2, and SERCA2a gene expression.
A reduction of both Cavα1.2 (64% vs. C; P < 0.01) and SERCA2a (27% vs. C; P < 0.05) gene expression was observed in Caf compared with C islets (Fig. 7, A and C). Pregnancy enhanced both Cavα1.2 (56% vs. C; P < 0.05) and SERCA2a (27% vs. C; P < 0.05) islet gene expression in control rats (Fig. 7, A and C). Cavα1.2 gene expression was augmented in CafP compared with Caf islets (77% vs. Caf; P < 0.001), while no differences were found in SERCA2a between these groups. These results showed separated effects of obesity and pregnancy (without interaction) in SERCA2a gene expression, while in Cavα1.2 gene expression, there was an effect of both nutritional and physiological conditions with interaction. Cavβ2 gene expression was similar among the groups (Fig. 7B).
Obesity is reaching epidemic proportions worldwide, mainly due to changes in lifestyle and easy access to both imbalanced food and overfeeding (7). Many diseases are associated with or have increased risks for health in obese patients, including Type 2 diabetes (4, 13). Under normal situations, the organism is able to compensate the demand for insulin, increasing the hormone secretion from pancreatic β-cells (24). Defects or inability of the β-cells to respond to glucose certainly have a role in the onset of Type 2 diabetes. Therefore, the mechanisms that make β-cells capable of overcoming insulin resistance during pregnancy are of great importance within the scope of biomedical sciences, having potential implications for humans. To address this question, we used an experimental model in which obesity was achieved by feeding female rats on a high-fat/high-calorie cafeteria diet (25, 35). Because the metabolic disorders induced by the cafeteria diet closely resemble those observed in humans, this is an attractive model to study obesity and also its consequence upon pregnancy. Our results show that cafeteria-induced obesity impairs insulin secretion stimulated by glucose and also by direct depolarizing agents, which is ameliorated in obese rats during pregnancy, and point to an important role of increased metabolic activity and especially of Ca2+ influx and mobilization in this process.
The impaired glucose tolerance observed in obese rats, both pregnant and nonpregnant, is in accordance with several observations showing that obesity induces insulin resistance (4, 16, 22, 25). Interestingly, fasted and fed plasma glucose is normal in the insulin-resistant rats, especially in pregnant groups, indicating that the higher plasma insulin levels are enough to maintain normoglycemia. The resistance to endogenous insulin in rats develops between days 16 and 19 of pregnancy (18), justifying the normal insulin levels observed in the pregnant chow-diet rats during ipGTT at the 14–15th day of pregnancy.
Reductions in insulin secretion stimulated by glucose in islets from rats fed on high-carbohydrate, high-protein or high-lipid diets have been previously observed (31). Thus, it is expected that obese cafeteria rats also present lower insulin release capacity. Nevertheless, the recovery of this capacity during pregnancy is of crucial interest and the information regarding intracellular events leading to this enhanced insulin secretion is not yet known. Therefore, our results showing increased metabolic activity, improved Ca2+ handling, and recovery of insulin secretion in pancreatic β-cells of obese rats during pregnancy are, to our knowledge, the first data to contribute to the clarification of the molecular mechanism of this phenomenon. On the other hand, it is at variance with decreased insulin secretion observed in high-saturated-fat-fed pregnant obese rats (15). We understand that different findings may be the outcome of different scientific strategies (type and time of diet, and time of pregnancy), but it is clear that a cafeteria diet is much closer to human overfeeding than other experimental models.
It is well established that the redox signaling is important for normal cellular functions, and it is of the greatest importance in the case of β-cells, where the oxidative metabolism is the main signal for insulin secretion. The majority of the NADPH inside the cells comes from the pentose phosphate pathway owning to the oxidative phase of reactions. At the molecular levels, NADPH can be generated by mitochondrial transhydrogenase or in the cytosol from glucose-6-phosphate by several enzymes (37). It has been known since long ago that intracellular NADPH increases following a load of glucose on pancreatic islets, and it is assumed that the increase in the ratio of [NADPH/NADP+] is important for insulin secretion (14, 20). Conversely, the [NADH/NAD+] ratio seems to be of no importance for proper insulin secretion (14). Recently, it has been proposed that NADPH is important for Ca2+ homeostasis (37). Indeed, NADPH is essential for Ca2+-induced insulin exocytosis, an effect that seems to be mediated by the NADPH-dependent protein glutaredoxin-1 (26). Therefore, the increase in NADPH showed in this study may partly explain the recovery of the insulin secretion capacity of pregnant obese rats. We speculate that a possible increase in the nonoxidative phase of pentose phosphate pathway may also provide additional substrate (fructose-6-phosphate) to glycolysis. This hypothesis is in line with the increase in glucose oxidation in pregnant obese rats, albeit it cannot be excluded by our data that these pathways are independently regulated.
Several of the steps involved in the insulin secretion process are modulated during pregnancy (1, 5, 8, 30). Ca2+ has a multitude of effects on the cells and plays an essential role on exocytosis (12). A correlation between the oscillatory insulin secretion pattern and oscillations of free Ca2+ within β-cells is well established (32). Cafeteria diet-induced obesity provoked a defect in Ca2+ handling by the islets. This effect seems to be dependent on a reduction of Ca2+ influx, as well as internal Ca2+ mobilization. Our experiments showing that direct depolarization of the cell membrane by KCl and tolbutamide (circumventing glucose metabolism) is impaired in Caf islets associated to a reduction in the L-type Ca2+ channel Cavα1.2 subunit expression, support this hypothesis. In accordance, reduced insulin secretion is also associated with defective glucose-dependent cytosolic Ca2+ handling in animal models of glucose intolerance (21), a defect linked to alterations in the activity and expression of the SERCA, which is involved in the regulation of Ca2+ handling by β-cells (17, 21). Actually, pregnancy increases SERCA2 protein expression in pancreatic islets, as shown in this study, and the inhibition of this protein decreases the first phase of insulin secretion, suggesting that intracellular Ca2+ stores in pancreatic islets from pregnant rats play a role in the enhancement of insulin release capacity (2). It seems that normalization of the Cavα1.2 gene expression in CafP islets is sufficient to restore cytoplasmic Ca2+ levels and insulin secretion in these islets, independent of modifications on the SERCA2a expression.
Perspectives and Significance
The present observations indicate that cafeteria diet induces obesity and insulin resistance, as well as diminishes insulin secretion stimulated by glucose and other depolarizing agents. The inhibitory effect of obesity on insulin secretion seems to be due to a defect in Ca2+ mobilization by these islets, independently of alterations in islets metabolism. Pregnancy recovers the secretory capacity in islets from obese rats, which is linked to the restoration of the β-cell ability to manage Ca2+, especially by increasing the capacity of Ca2+ uptake. This phenomenon seems to be, at least in part, dependent on the augmentation of the expression of the L-Type Ca2+ channels sub-unit α1.2. In addition, during pregnancy pancreatic islet metabolic activity is enhanced in obese rats, which may increase the glucose responsiveness of pancreatic β-cells. Further studies aiming at clarifying the deleterious effects of obesity on Ca2+ movements and insulin secretion, as well as the discovery of agents that help the islet cells to recover such an ability during the pregnant state, may help in developing strategies to maintain normoglycemia in Type 2 diabetic obese patients.
No conflicts of interest are declared by the authors.
This work was supported by the Brazilian foundations Fundação de Amparo à Pesquisa do Estado de São Paulo, Conselho Nacional de Pesquisa, and Instituto Nacional de Obesidade e Diabetes. We thank Mr. L. Domingos for technical assistance and Dr. N. Conran for English editing. M. L. Bonfleur is a Ph.D. student on leave from the State University West of Paraná, Brazil.
- Copyright © 2010 the American Physiological Society