The Lack of β-Adrenoceptors Results in Enhanced Insulin Sensitivity in Mice Exhibiting Increased Adiposity and Glucose Intolerance

We obtained 3- to 4-month old wild-type and β-less mice as previously described (10). Briefly, β₁^+/−/β₂^+/−/β₃^+/− mice were crossed to generate β₁^+/+/β₂^+/+/β₃^+/+ and β₁^−/−/β₂^−/−/β₃^−/− mice. Several couples were then established from these homozygous mice, and experiments were performed on β₁^+/+/β₂^+/+/β₃^+/+ (wild-type) and β₁^−/−/β₂^−/−/β₃^−/− (β-less) offspring. Mice were housed individually and kept on a 12-h light/dark cycle in a temperature-controlled room at 24°C. They were allowed ad libitum access to water and a standard laboratory chow, unless otherwise stated. All procedures used were approved by the Office Vétérinaire Fédéral et Cantonal (Geneva).

Mice were anesthetized with a solution of ketamine (150 mg/kg body wt) and xylazine (16 mg/kg body wt) and scanned using a densitometer (Lunar Piximus). Fat content was determined as the percentage of body mass.

Blood glucose levels were determined using a glucose meter (One Touch; LifeScan, Milpitas, CA). Plasma insulin concentrations were measured by enzyme-linked immunosorbent assay (Crystal Chem, Chicago, IL). Plasma corticosterone levels were determined by a double-antibody radioimmunoassay (Immunodiagnostic Systems, Boldon, U.K.). Triglycerides and NEFAs were measured using an enzymatic triglycerides PAP 150 kit (BioMérieux, Lyon, France) and a NEFA-C kit (Wako Chemicals, Neuss, Germany), respectively.

Measurements of glycogen in liver were performed as previously described (11).

Total RNA was isolated by the method of Chomczynski and Sacchi (12). Oligo-dT first-strand cDNA were synthesized using 5 μg total RNA and a Superscript II RNase H reverse transcription kit (Invitrogen Life Technologies, Basel, Switzerland) according to the manufacturer’s instructions. Real-time PCR was performed using an ABI rapid thermal cycler system and a SYBR Green PCR master mix according to the manufacturer’s instructions. The real-time PCR conditions were a step at 50°C for 2 min, followed by a denaturing step at 95°C for 10 min, 40 cycles of denaturation at 95°C for 15 s, and annealing/extension at 60°C for 1 min. A melting curve was generated at the end of the PCR cycles to verify that a single product was amplified. Cyclophilin A was used as a control to account for any variation caused by the efficiencies of the reverse transcription and PCR. The PEPCK oligonucleotide primers used were: upstream 5′-tggtggccgtagacctgaa-3′ and downstream 5′-tgtttgtaggagcagccatgag-3′, covering the nucleotides 14391–522 of PEPCK cDNA (GenBank accession no. NM_011044). Cyclophilin A oligonucleotide primers used were: upstream 5′-caaatgctggaccaaacacaa-3′ and downstream 5′-ccatccagccattcagtcttg-3′, covering the nucleotides 343–412 of cyclophilin A (GenBank accession no. XM 355936). The upstream and downstream oligonucleotide primers were chosen on both sides of an intron to prevent amplification of possible contaminating genomic DNA.

After an overnight fast, glucose (1 mg/g body wt) in normal saline (0.9% NaCl) was administered intraperitoneally. Blood glucose levels were measured using a glucose meter (One Touch II; LifeScan) from tail blood at 0, 15, 30, 60, 90, and 120 min after glucose injection.

After a 4-h fast, insulin (0.75 mU/g body wt; Novo Nordisk, Bagsvaerd, Denmark) in normal saline (0.9% NaCl) was administered intraperitoneally. Blood glucose levels were determined using a glucose meter (One Touch II; LifeScan) from the tail blood at 0, 20, 40, 60, and 120 min after insulin injection.

A catheter was placed into the jugular vein of the mice under anesthesia (pentobarbital, 100 mg/kg i.p.). This catheter was connected to the insulin infusion pump and the glucose and tracer d-[3-³H] glucose infusion pumps (Perkin Elmer Life and Analytical Sciences, Boston, MA). On the day of the experiment, the mice were fasted for 8 h. Mice were continuously infused with the tracer at a rate of 30 μCi · kg⁻¹ · min⁻¹ during the whole experiment starting at time = −45 min. The insulin infusion (18 mU · kg⁻¹ · min⁻¹) started at time = 0 min. Euglycemia was maintained by the use of a concomitant, variable, glucose (20% solution) infusion. Glycemia was determined in 3.5 μl of blood sampled from the tip of the tail vein every 5 min using a glucose meter (One Touch II; LifeScan). In addition, 5 μl of blood was recovered at time = −10, 0, 50, and 60 min, and specific activity of the tracer was measured in deproteinized blood samples, as previously reported (13). Whole-body glucose turnover was then calculated by dividing the tracer infusion rate by the plasma glucose specific activity. Hepatic glucose production was calculated as the difference between the glucose turnover rate (R_d) and the glucose infusion rate in steady-state conditions.

After an overnight fast, glucose (1 mg/g body wt) in normal saline (0.9% NaCl) was administered intraperitoneally. Blood glucose and plasma insulin levels were measured from the tail blood at 0, 5, 15, 30, 60, and 120 min after glucose injection.

The results are given as the means ± SE. Statistical analysis was performed using the Mann-Whitney rank sum test or the two-tailed unpaired Student’s t test, except for food restriction experiments, which were analyzed using paired Student’s t test. The same group of wild-type or β-less mice was used for the insulin tolerance test (ITT) in ad libitum and restricted states. The calculations were performed using Sigma STAT software (SPSS, Chicago, IL). A P value <0.05 was considered statistically significant.

We studied 3- to 4-month-old triple knockout male mice lacking the β₁/β₂/β₃-adrenoceptors (β-less) and corresponding control wild-type mice. At this age, there was no difference in body weight, nor in food intake between the two groups of mice (Table 1). However, dual-energy X-ray absorptiometry analysis revealed that the percentage of body fat of β-less mice was highly increased in comparison with wild-type mice. Corroborating this result, the weight of the epididymal fat pad was found to be increased in β-less mice (Table 1).

We then looked at whether glucose tolerance was altered in β-less mice. For this purpose, we performed glucose tolerance tests (GTTs) in the two groups of mice. As shown in Fig. 1, the glycemia of β-less mice was significantly higher in comparison with controls during the GTT. We reasoned that the glucose intolerance of β-less mice could be explained either by a defect in glucose-induced insulin secretion and/or by a decrease in insulin sensitivity. To test the first hypothesis, we looked at potential changes in insulin levels after an intraperitoneal bolus of glucose (1 mg/g). In the fasted state, we observed that the insulin levels of β-less mice were not statistically different from those of wild-type mice (0.40 ± 0.07 and 0.56 ± 0.04 ng/ml in wild-type and β-less mice, respectively; NS). In contrast, the increase over basal insulin levels was significantly lower in β-less than in control mice after the glucose bolus, and the difference was maintained at 120 min postinjection (Fig. 2).

Because the physiological situation corresponding to the glucose administration is the fed state and because β-less mice showed impaired insulin release in response to a glucose challenge, we examined whether the lack of insulin secretion during the fed state would lead to an increase in hepatic glucose production. For this purpose, we examined the mRNA expression of PEPCK, a key gluconeogenic enzyme whose expression has been shown to be suppressed by insulin (14) in wild-type and β-less fed mice. Using quantitative real-time PCR, we observed that PEPCK mRNA levels were 1.5-fold higher in β-less than in wild-type mice (Fig. 3A). No difference in hepatic glycogen content was found between the two groups of mice (37.8 ± 4.3 and 31.8 ± 7.6 mg/g of liver in wild-type and β-less mice, respectively; NS), suggesting that hepatic glycogenolysis was not different between β-less and wild-type mice.

Because glucocorticoids are also important regulators of hepatic glucose output via their stimulatory effects on gluconeogenesis, we measured corticosterone plasma levels in fed wild-type and β-less mice. We found that β-less mice have increased glucocorticoid levels (Fig. 3B). In contrast, free fatty acid (0.53 ± 0.02 and 0.53 ± 0.07 mEq/l in wild-type and β-less mice, respectively; NS) and triglyceride (128.3 ± 16.2 and 101.2 ± 10.9 mg/dl, respectively; NS) concentrations were not significantly different between wild-type and β-less mice.

To test the second hypothesis that may also explain the glucose intolerance, namely a decrease in insulin sensitivity in β-less mice, we performed ITTs in the two groups of mice. Surprisingly, as shown in Fig. 4, the fall of glycemia after insulin administration was of greater magnitude in β-less than in wild-type mice. Moreover, the area under the ITT curve of β-less mice was significantly smaller than that of wild-type mice (Fig. 4, inset).

Dietary restriction has been shown to increase insulin sensitivity. Because β-less mice displayed increased insulin sensitivity, we hypothesized that the beneficial effect of food restriction on whole-body insulin-mediated glucose clearance, which is thought to be mediated, at least in part, by a decrease in sympathetic tone, would be less pronounced in β-less than in wild-type mice. As shown in Fig. 5A, we observed that food restriction (20% reduction for 6 days) was able to ameliorate insulin sensitivity of wild-type animals, but it had no effect on this parameter in β-less mice (Fig. 5B).

To further substantiate the observations made with the ITTs, we performed euglycemic-hyperinsulinemic clamps. The metabolic parameters before and at the end of the clamps are shown in Table 2. During the clamps, we observed that a higher glucose infusion rate was required to maintain normal glucose levels in β-less compared with control mice (23.2 ± 1.4 and 32.2 ± 3.0 mg · min⁻¹ · kg⁻¹ in wild-type and β-less mice, respectively; P < 0.05). As shown in Fig. 6A, this difference was caused by an increase in the rate of glucose disappearance (R_d), whereas hepatic glucose production was similarly suppressed in both groups (Fig. 6B).

We have shown that 3- to 4-month-old mice lacking the β₁/β₂/β₃-adrenoceptors are glucose intolerant compared with wild-type controls, as demonstrated by the results of the GTT. We reasoned that this observation could be caused either by a defect in glucose-induced insulin secretion and/or by a decrease in insulin sensitivity.

Concerning the former hypothesis, we observed that β-less mice actually display a defect in insulin secretion. This last observation is consistent with previous results from our laboratory showing that high fat–fed mice lacking the β₃-adrenoceptor also exhibit a glucose intolerance caused by impaired insulin secretion (M.J., unpublished data). In humans, the β₃-adrenoceptor has been shown to be expressed in the pancreatic islets, and it has been proposed that its activation stimulates insulin secretion by β-cells (15). Thus, it is possible that the lack of β₃-adrenoceptors is directly responsible for the impaired insulin secretion observed in our mice. However, it should be stated that β-cell–specific β₃-adrenergic receptor expression has not yet been clearly demonstrated in mice (M.J., unpublished data) (16). Because β₁/β₂-adrenergic stimulation of β-cells is thought not to play a major role in controlling insulin secretion, we can speculate that the observed effect results from the activation of the β-adrenergic system on tissues that may indirectly regulate insulin secretion. With regard to this possibility, it has been proposed that a factor secreted by white adipose tissue, whose expression is under the regulation of the β₃-adrenoceptor, could mediate the β-adrenergic effect on insulin secretion (16). Finally, we could also suppose that there is an upregulation of the α₂-adrenoceptors in the pancreatic β-cells of our β-less mice. Because these receptors are known to inhibit insulin secretion, they could be responsible for the observed defect.

In addition, together with the defect of glucose-induced insulin secretion, we observed that, in a fed state, PEPCK mRNA levels were 1.5-fold higher in the liver of β-less mice than in controls. This increase suggests that the lack of glucose-induced insulin secretion in β-less mice during the GTT leads to a less efficient inhibition of PEPCK gene expression in the liver. Another possible explanation for the increase of PEPCK expression in β-less mice is the hypercorticosteronemia that we observed in these animals. Indeed, glucocorticoids are known to stimulate hepatic expression of the PEPCK gene both via a glucocorticoid responsive element on its promoter and through the recruitment of proliferator-activated receptor-γ coactivator (17,18).

Concerning a possible involvement of a defect in insulin sensitivity for explaining the glucose intolerance of β-less mice, we observed that β-less mice are surprisingly more sensitive to insulin, as demonstrated by the results of the euglycemic-hyperinsulinemic clamps and ITTs performed in fed wild-type and β-less mice, as well as by those obtained in food-restricted mice. Indeed, in the present study, food restriction, which is thought to decrease sympathetic activity in both humans and rodents (19–21), ameliorated insulin tolerance of wild-type mice, whereas it had no effect on insulin sensitivity in β-less animals. Taken together, these results suggested that β-adrenoceptors are essential in the regulation of insulin-mediated glucose utilization. The enhanced insulin sensitivity of β-less mice is surprising because these animals display increased fat mass, which has been shown to be one of the major factors contributing to the development of insulin resistance. However, it has also been reported that β-adrenergic stimulation inhibits insulin-stimulated glucose uptake in skeletal muscle and in white adipose tissue. Indeed, it has been shown that epinephrine decreases insulin-stimulated glucose transport in rat skeletal muscle and in humans by altering the intrinsic activity of surface GLUT4 transporters (22–25). In addition, the β-adrenergic agonist isoproterenol was shown to inhibit insulin-stimulated glucose transport in rat adipose cells (4,5). Thus, it is possible that the suppression of β-adrenergic stimulation directly contributes to the enhanced peripheral insulin sensitivity of β-less mice.

The augmentation in insulin sensitivity is of particular interest with regard to the use of β-blockers in humans. It has been shown that they are highly effective in reducing the risk of cardiovascular events and death in post–myocardial infarction patients with diabetes (26). Indeed, the insulin resistance associated with type 2 diabetes leads to high insulin levels, resulting in increased sympathetic activation and thus norepinephrine release (27–30). Furthermore, norepinephrine has been shown to cause hypertension, to increase the risk of sudden death by lowering the threshold to lethal ventricular arrhythmias, and to augment arterial wall damage predisposing to atheroma. In this context, β-blockade could inhibit the harmful actions of sustained high norepinephrine activity (31). It would therefore be interesting to further explore whether our β-less mice have enhanced cardiovascular protection. However, this potential benefit of the β-blockade could be to the detriment of obesity and glucose intolerance, as suggested by the phenotype of β-less mice.

Altogether, our results obtained with β-less mice show that an intact β-adrenergic system is required to regulate overall glucose homeostasis, and they complement previous data obtained with pharmacological agonists or antagonists. In particular, the glucose intolerance of β-less mice seems to be related to defective glucose-induced insulin secretion as well as a glucocorticoid-induced increase in gluconeogenesis. In addition, β-less mice are more sensitive to insulin in terms of the rate of glucose utilization (R_d). Future studies in insulin-sensitive tissues of β-less mice could provide important information on the cross talk between the β-adrenergic and insulin signaling pathways.

This work was supported by Swiss National Science Foundation Grant 3100AO-105889 (to F.R.-J.), the Fondation Boninchi (to P.M.), the Fondation du Centenaire de la Société Suisse d’Assurances Générales sur la Vie Humaine Pour la Santé Publique et les Recherches Médicales (to P.M.), and European Community FP6 contract LSHM-CT-2003-503041 (to F.R.-J.). This study was part of the Geneva Programme for Metabolic Disorders.

We are indebted to Pr. Jean-Paul Giacobino for helpful discussions and critical comments of this manuscript, and we thank Dr. Dominique Pierroz for giving us access to the dual-energy X-ray absorptiometry scanner. We are grateful to Dr. Kobilka for providing us with the double-knockout mice. Finally, we thank Marcella Klein for her excellent technical assistance.