Evaluation of insulin sensitivity by hyperinsulinemic-euglycemic clamps using stable isotope-labeled glucose

Dear Editor, Insulin resistance is a critical factor in the pathogenesis of metabolic diseases such as obesity, nonalcoholic fatty liver disease (NAFLD) and type 2 diabetes (T2D). For many years, the hyperinsulinemic-euglycemic clamp has been used as a “gold standard” method to accurately measure insulin action in vivo. It is widely used in humans, dogs, rats and mice. During the clamp, glucose kinetics, including the rates of endogenous glucose production and disposal in mice, are conventionally assessed with tracers. The radioactive tracer [3-H]glucose is commonly used because it is sensitive and massless, but it is harmful to our environment, and cannot be used in humans because it is hazardous if introduced into the body. Therefore, medical research has turned to stable isotopes as alternative tracers. Although many studies have successfully established the clamp method using stable isotopes in humans, no method has been developed for laboratory mice because of the limitations of mass spectrometry, which requires the infusion of a large dose of stable isotope and a large volume of blood. In this study, we have successfully devised a sensitive method using [6,6-H]glucose as a tracer in mice. [6,6-H] glucose is a stable (non-radioactive) naturally occurring isotope with no known harmful effects and similar metabolic effects to normal glucose. It can be distinguished from natural isotopomers of glucose (i.e., with other isotopic fine structures) using a high resolution mass spectrometer. To establish this method, we tested it in a high-fat diet (HFD)-induced obese mouse model, which is well known and widely used in metabolism research. Compared to mice fed with a regular diet (RD), HFD-induced obese mice had significantly higher body weight, plasma insulin levels, glucose production measured by pyruvate tolerance test (PTT), glucose intolerance evaluated by glucose tolerance test (GTT) and insulin insensitivity assessed by insulin tolerance test (ITT) (Fig. 1a–e). All the results indicate that glucose production and insulin resistance were dramatically increased in HFD-fed mice compared to RD-fed animals. To evaluate insulin action and glucose metabolism in vivo, we performed hyperinsulinemic-euglycemic clamp studies using [6,6-H]glucose as a tracer (Fig. 1f, g). A bolus of [6,6-H]glucose (600 μg kg) was administered via catheter followed by continuous infusion of [6,6-H]glucose at the rate of 30 μg kg min for 90min to maintain steady-state conditions. During the clamp, we monitored the blood glucose levels, plasma insulin levels and glucose infusion rate (GIR) (Fig. 1h, i, Supplementary Figure S1a). At the steady state, GIR in HFD-fed mice is 12.17± 1.09 mg kg min, which is much lower than that in RD-fed mice (22.10± 1.23mg kg min), indicating that insulin sensitivity is decreased in HFD-fed mice (Fig. 1i). In our experiments, levels of glucose and [6,6-H]glucose were simultaneously measured using mass spectrometer. To distinguish [6,6-H]glucose from natural isotopomers of glucose (M+ 2), we used a high resolution mass spectrometer (Orbitrap) coupled with an ultra-high performance liquid chromatography (UPLC) system. It was recently reported that the Orbitrap machine has a mass resolution of 100,000, and can therefore resolve isotopic fine structures. We performed this assay based on a mass resolution of 140,000, which enables us to differentiate oxygen-18 (O), C and H isotopes of

RSD, relative standard deviation. c Immunoblots showing the reduced pAKT levels in liver, skeletal muscle and white adipose tissue (WAT) extracts of mice fed a HFD.

Mice
C57BL/6J mice were housed in colony cages under 12-hr light/dark cycle. Mice with diet-induced obesity were obtained by feeding a diet containing 60 kcal% fat (Research Diets, D12492) for 16 weeks. All animal experiments were carried out with the approval of the Animal Care and Use Committee at Tsinghua University.

Hyperinsulinemic-Euglycemic Clamp
Mice were anesthetized with Pentobarbital sodium salt (Sigma, P3761). Hairs at the incision site were removed and skin was disinfected by a betadine scrub.
An incision was made in the skin and the right jugular vein was identified. A catheter (PE-10) filled with heparinized saline (200 units heparin ml -1 saline) was inserted into the vein toward the side of the chest. A suture was placed at each end of the vessel. The hairs were removed from the area around the shoulder blades and the catheter was tunneled under the skin from the right jugular to the interscapular incision on the back. The catheter was plugged with a stainless steel plug and settled on the back. All skin wounds were sutured and the mouse was allowed to recover for 4-5 days. Mice that lost < 4% of their precannulation weight after recovery were used for clamp studies.
The mice were fasted overnight, then weighed to calculate the insulin dose and placed in a restrainer. The setup and time-line for the experiment are shown in Figure 1F. An equilibration syringe was connected to the catheter and an initial bolus of [6,6-2 H]glucose (600 μg kg -1 ) was administered, followed by continuous infusion of [6,6-2 H]glucose (30 μg kg -1 min -1 ). Isotopic enrichment was achieved at approximately 60 min after the onset of isotope infusion, and blood samples for isotope measurements were collected at t = -10 and 0 min. Following this basal infusion period, insulin was constantly infused at the rate of 6 mU kg -1 min -1 until termination of the study. [6,6-2 H]glucose was infused together with glucose at various rates until the blood glucose concentration reached a constant level of about 100 ± 5 mg dl -1 . Blood samples were taken at t = 110 and 120 min.

Mass spectrometry
All plasma samples (15 μl) were deproteinized by gently mixing them with 60 μl cold methanol (pre-chilled at -80 ºC) and incubating for 1-2 hrs at -80 ºC. The samples were then centrifuged with 14,000 g at 4 ºC for 10 min. The supernatant was transferred to a new tube and lyophilized to produce a pellet.
High resolution mass spectrometry (Q Exactive, Thermo) coupled with Ultimate 3000 ultra-high performance liquid chromatography (UPLC) was used for the analysis. Negative ion mode with a mass resolution of 140,000 was performed for glucose detection. Flow rate of sheath gas and aux gas was set as 35 and 10 respectively. Spray voltage was 2.8 kV. For LC separation, an amide column (2.1 mm × 100 mm, Waters) was used for analysis of glucose. Column temperature was 40 ºC. Mobile phases contained 80% acetonitrile in A and 30% acetonitrile in B with 5 mM ammonium acetate as modifier. The flow rate was 250 µL min -1 and the gradient was as follows: 0-3 min: 10% B; 3-4 min: 10-60% B; 4-6 min: 60% B; 6.1-8 min: 10% B. A total of 2 μL sample was injected for analysis. Accurate masses of 179.0561 and 181.0686 with mass tolerance of 3 ppm were used for extraction of glucose and [6,6-2 H]glucose in negative mode.

Calculation
Under steady-state conditions, the rate of glucose appearance is equal to the rate of disappearance. Our calculations are all based on Steele's equations for steady-state conditions 16,17 . Because [6,6-2 H]glucose is not massless, the glucose disposal rate (GDR) is equal to the constant isotope infusion rate (F) divided by the isotope enrichment (atom percent excess, APE) minus F 16, 17 .
During the basal infusion period, while there is no glucose infusion, GDR is equal to the rate of hepatic glucose production (HGP). During infusion of unlabeled glucose mixed with [6,6-2 H]glucose, GDR is equal to the sum of the rate of HGP and the glucose infusion rate (GIR). Therefore, the HGP is calculated by subtracting the GIR from the calculated GDR. The calculation was performed as indicated in Figure 1G.
Samples were loaded on SDS-PAGE gels and then transferred to nitrocellulose membranes. Immunoblotting was done in gelatin buffer (50 mM Tris HCl pH 7.4, 150 mM NaCl, 5 mM EDTA, 0.05% Tween-20) with the corresponding antibodies. The rabbit polyclonal antibody against anti-pAKT (13038) was purchased from Cell Signaling Technology. The rabbit polyclonal antibody against anti-AKT (SC-8312) was from Santa Cruz. The mouse monoclonal antibody against anti-Tubulin (T6199) was from Sigma-Aldrich.

Measurement of Plasma insulin and free fatty acids
Plasma insulin (Mercodia, 10-1247-01) and free fatty acid levels (Wako, 294-636) were measured according to the manufacturer's instructions.

Statistical analyses
Age-and weight-matched male mice were randomly assigned for the experiments. The number of animals used in each experiment is outlined in the corresponding figure legends. No animals were excluded from statistical analyses, and the investigators were not blinded in the studies. All studies were performed on at least three independent occasions. Results are reported as mean ± s.e.m. Comparison of different groups was carried out using two-tailed unpaired Student's t-test. Differences were considered statistically significant at