Insulin Tolerance Test under Anaesthesia to Measure Tissue-specific Insulin-stimulated Glucose Disposal.

Insulin resistance is a pathophysiological state defined by impaired responses to insulin and is a risk factor for several metabolic diseases, most notably type 2 diabetes. Insulin resistance occurs in insulin target tissues including liver, adipose and skeletal muscle. Methods such as insulin tolerance tests and hyperinsulinaemic-euglycaemic clamps permit assessment of insulin responses in specific tissues and allow the study of the progression and causes of insulin resistance. Here we detail a protocol for assessing insulin action in adipose and muscle tissues in anesthetized mice administered with insulin intravenously.

[Abstract] Insulin resistance is a pathophysiological state defined by impaired responses to insulin and is a risk factor for several metabolic diseases, most notably type 2 diabetes. Insulin resistance occurs in insulin target tissues including liver, adipose and skeletal muscle. Methods such as insulin tolerance tests and hyperinsulinaemic-euglycaemic clamps permit assessment of insulin responses in specific tissues and allow the study of the progression and causes of insulin resistance. Here we detail a protocol for assessing insulin action in adipose and muscle tissues in anesthetized mice administered with insulin intravenously.
Keywords: Insulin, Adipose tissue, Muscle, Glucose transport, Glucose uptake, Insulin tolerance test [Background] Surrogate measures of whole-body insulin sensitivity such as homeostatic model assessment for insulin resistance (HOMA-IR) often do not reflect insulin-stimulated glucose uptake in peripheral tissues of the mouse (Lee et al., 2008;Mather, 2009). Ex vivo adipose tissue explants or isolated skeletal muscles preparations (Burchfield et al., 2018;Fazakerley et al., 2018b) can allow direct measurement of insulin responses in specific tissues if they are amenable to these procedures, but do not preserve the organism environment where finely articulated delivery of glucose and insulin via the vasculature, uptake into cells and intracellular metabolism may all contribute to the overall rate of tissue glucose uptake (Wasserman et al., 2011). Commonly employed methods to assess whole-body insulin action in mice include intraperitoneal (IP) or intravenous insulin tolerance tests (ITTs) and, the gold standard, hyperinsulinaemic-euglycaemic clamp (Ayala et al., 2010;Brandon et al., 2016). These techniques can assess whole-body insulin action by measuring either changes in blood glucose (ITT) or changes in the amount of glucose required to maintain glycemia (glucose infusion rate; hyperinsulinaemic-euglycaemic clamp) and can be adapted to measure tissue-specific insulin responses. For example, glucose tracers such as radiolabeled 2-deoxyglucose (2-DOG), which is often used as a surrogate for glucose in assessing glucose uptake as it is not metabolized through glycolysis and is 'trapped' in the cells following uptake, can be introduced to permit assessment of glucose uptake 2 www.bio-protocol.org/e3146 into specific tissues during these tests. Note that 2-DOG is only trapped in tissues that do not possess significant glucose-6-phosphatase activity, and so 2-DOG tracer is therefore not useful for assessing glucose disposal in the liver.
Here, we describe a protocol for a terminal intravenous ITT performed under pentobarbitone-induced anesthesia to assess insulin-stimulated glucose uptake into tissues of interest. The protocol can be performed with minimal delays between mice, making it amenable to assessing insulin action in large cohorts of animals. Intravenous administration of a bolus of insulin/tracer rapidly delivers insulin to tissues, minimizing the time between the start of the assay and initiation of insulin responses within tissues. Also, rapid equilibration of the 2-DOG tracer with the total blood glucose pool ensures that the tracer is immediately available for uptake into tissues. Alternative methods of insulin/tracer delivery (i.e., intraperitoneal injection, oral gavage) may result in delays or inconsistent insulin/tracer uptake into the central circulation. For example, in the case of IP injection, there may be a considerable delay before insulin reaches the circulation. These time lags may differ between mice or between injections, increasing experimental variability. Rapid and consistent tissue uptake is also particularly important when performing time-series experiments in live animals. Indeed, hepatic portal-vein administration of insulin has enabled the assessment of the time-resolved effects of insulin on the liver phosphoproteome (Humphrey et al., 2015). For oral gavage, 2-DOG may exhibit markedly different kinetics of appearance in the circulation compared to glucose since the sodium-dependent glucose transporters, which play a key role in oral glucose absorption, exhibit a strong preference for glucose (Bissonnette et al., 1996), and this may limit tracer availability to tissues during the assay.
In this protocol, mice are anesthetized with pentobarbitone and saline or insulin and radiolabeled 2-    11. Measure blood glucose (from tail) and collect 5 μl blood (and add to ZnSO4 as in Step A6) per mouse to determine blood radioactivity after 2, 5, 10, 15, 20, and 30 min. 8 www.bio-protocol.org/e3146  2. To measure tracer uptake into a tissue of interest, powder tissue in liquid nitrogen using a mortar and pestle and then weigh an aliquot of powder for analysis. Smaller tissues such as the soleus muscle can be homogenized according to Step C3 and do not require powdering.
3. Homogenize ~40 mg tissue (as little as 10 mg soleus and EDL muscle can be used) in 1 ml ddH2O in a 1.5 ml tube by sonication (90% power, 3 x 10 s, allowing sample to cool between pulses) and centrifuge at 13,000 x g for 15 min. Collect supernatant (~800 μl) and transfer to a new 2.0 ml tube. Bring up volume to 2 ml with ddH2O.
4. Prepare the phospho-2-DOG affinity elution columns by adding 1 ml AG1-X8 resin diluted in ddH2O (70% volume resin) to a 0.8 x 4 cm chromatography column using a wide-bore pipette tip (e.g., P1000 tip cut 5 mm from the tip using a scalpel blade to increase the aperture).
5. Place 5 ml tubes below the columns and add 1 ml of the supernatant to the columns followed by three 1 ml washes with ddH2O.
6. Place new scintillation vials below the columns and add 1 ml of the elution buffer (see Recipes below) to columns followed by another 1 ml of elution buffer. 7. Add 3 ml scintillation fluid (PerkinElmer) to the scintillation vials, vortex thoroughly, and measure 3 H DPM in each sample using a liquid scintillation counter to quantify [ 3 H]2-DOG-6-P.

Data analysis
The aim of data analysis is to normalize for the amount of tracer delivered to each mouse and available for uptake into tissues, and the amount of tissue analyzed. Methods for normalizing such data have been discussed extensively elsewhere (Sokoloff et al., 1977;Goodner et al., 1980;Hom et al., 1984;Cooney et al., 1985;Kraegen et al., 1985).
We describe two methods below: 1) the data are normalized to tracer availability in the blood to calculate the amount of tracer taken into the tissue of interest as a proportion of the amount of tracer available to the tissue, and 2) approximate an index of glucose uptake into tissues by using DPM and blood glucose to calculate a specific activity (DPM/mol glucose). In each case, data can be expressed per unit weight and/or protein/DNA content of the analyzed tissue. This calculation assumes that the 2-DOG tracer measured by a tail bleed is indicative of tracer available in the interstitial space for uptake into the tissue.
First, extrapolate DPM per 5 μl to 1 ml to yield DPM/ml. Since the tracer will disappear from the blood by an exponential decay, calculate the AUC by fitting the DPM/ml at measured time points to a single exponential function, and integrate this function over the experimental period. 10 www.bio-protocol.org/e3146 This estimates the change in blood DPM/ml throughout the experiment (DPM/ml·min). This AUC value provides a normalization factor that takes into account differences in tracer availability between mice.
Tissue DPMs can be normalized using this AUC value to calculate the proportion of available tracer taken up into the tissue. This can be further normalized to the weight of tissue analyzed (g). The final units are: 2-DOG clearance (ml/min/g). Data normalized using this method are presented in Figures 3A and 3B. These data show insulin-stimulated 2-DOG clearance into muscle and adipose tissues, but not into brain ( Figure 3A) and that feeding mice a diet high in fat and sucrose leads to impaired insulin-stimulated 2-DOG clearance into epididymal adipose tissue and quadriceps muscle ( Figure 3B).
2. To obtain a tissue-specific index of glucose uptake: The blood glucose concentration during this assay is non-steady-state. This calculation aims to take into account differences in blood glucose over the course of the experiment and approximate glucose uptake into tissue. This calculation assumes that there is no discrimination between 2-DOG and glucose at the glucose transporter and therefore the rate of [ 3 H]2-DOG accumulation in tissue is equivalent to the rate of glucose uptake into tissue (Ferre et al., 1985).
Since the kinetics of 2-DOG and glucose uptake may differ, we advise referring to 2-DOG accumulation as a "glucose uptake index".
First calculate the AUC for blood DPM (as above by exponential curve fitting; DPM/ml·min) and blood glucose during the ITT (by the trapezoidal method; μmol/ml·min). The average specific activity of 2-DOG in the blood can be calculated by dividing the blood DPM AUC by blood glucose AUC (DPM/μmol). Tissue DPM can then be converted to an index for the rate of glucose uptake by dividing by the average specific activity, and further normalized to the weight of tissue analyzed (g) and expressed per min or h. The final units are: Glucose uptake index (μmol/g/h). Data normalized using this method are presented in Figure 3B. These data show that feeding mice a high fat high sucrose diet for 14 d lowers the insulin-stimulated glucose uptake index in epididymal adipose tissue and quadriceps muscle ( Figure 3C). Krycer and Dr. Lake-Ee Quek for helpful discussion.