This Article Abstract Full Text (PDF) All Versions of this Article: 146/12/5151    most recent Author Manuscript (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Watson, E. Articles by Salton, S. R. J. Search for Related Content PubMed PubMed Citation Articles by Watson, E. Articles by Salton, S. R. J.

VGF Ablation Blocks the Development of Hyperinsulinemia and Hyperglycemia in Several Mouse Models of Obesity

Fishberg Department of Neuroscience (E.W., S.H., T.M.M., J.W., C.M., C.V.M., S.R.J.S.) and Brookdale Department of Geriatrics and Adult Development (T.M.M., C.V.M., S.R.J.S.), Mount Sinai School of Medicine, New York, New York 10029; and Departments of Medicine and Cell Biology (P.E.S.), Albert Einstein College of Medicine, Bronx, New York 10461

Address all correspondence and requests for reprints to: Stephen R. J. Salton, M.D., Ph.D., Fishberg Department of Neuroscience, Box 1065, Mount Sinai School of Medicine, One Gustave L. Levy Place, New York, New York 10029. E-mail: stephen.salton{at}mssm.edu.

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Targeted deletion of the gene encoding the neuronal and endocrine secreted peptide precursor called VGF (nonacronymic) produces a lean, hypermetabolic, hyperactive mouse. Because VGF mutant mice are resistant to specific forms of diet-, lesion-, and genetically induced obesity, we investigated the role that this polypeptide plays in glucose homeostasis. We report that VGF mutant mice have increased insulin sensitivity by hyperinsulinemic euglycemic clamp analysis, and by insulin and glucose tolerance testing. Blunted counterregulatory responses in VGF-deficient mice were likely influenced by their significantly lower liver glycogen levels. VGF deficiency lowered circulating glucose and insulin levels in several murine models of obesity that are also susceptible to adult onset diabetes mellitus, including A^y/a agouti, ob/ob, and MC4R^–/MC4R^– mice. Interestingly, ablation of Vgf in ob/ob mice decreased circulating glucose and insulin levels but did not affect adiposity, whereas MC4R^–/MC4R^– mice that are additionally deficient in VGF have improved insulin responsiveness at 7–8 wk of age, when lean MC4R^–/MC4R^– mice already have impaired insulin tolerance but are not yet obese. VGF mutant mice also resisted developing obesity and hyperglycemia in response to a high-fat/high-carbohydrate diet, and after gold thioglucose treatment, which is toxic to hypothalamic glucose-sensitive neurons. Lastly, circulating adiponectin, an adipose-synthesized protein the levels of which are correlated with improved insulin sensitivity, increased in VGF mutant compared with wild-type mice. Modulation of VGF levels and/or VGF signaling may consequently represent an alternative means to regulate circulating glucose levels and insulin sensitivity.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
MICE WITH DEFECTIVE signaling in the melanocortin pathway, including those that overexpress the melanocortin receptor antagonists agouti or the agouti-related protein (AGRP or ART) (1, 2), and those that fail to express the hypothalamic melanocortin-4 receptor (MC4R) (3) or its agonist -MSH [proopiomelanocortin (POMC) knockout] (4), all develop a maturity onset obesity syndrome that is associated with hyperphagia, hyperinsulinemia, and hyperglycemia. The central melanocortin system not only regulates energy balance, but also has been shown to regulate glucose homeostasis through effects on insulin release and sensitivity (5). Late onset hyperglycemia observed in male A^y/a and MC4R knockout mice shares similarities to the development of type 2 diabetes. Mice that fail to synthesize leptin (ob/ob) or those that express abnormal leptin receptors (db/db) develop early onset obesity, reduced metabolic rate, increased food intake, decreased fertility, and diabetes (6). Thus hyperglycemia and hyperinsulinemia are common features of several obesity syndromes that result from defects in leptin or melanocortin signaling pathways. Recent work has further demonstrated that overexpression of the POMC gene in the brains of transgenic mice can normalize hyperglycemia, glucose intolerance, and insulin resistance in leptin-deficient mice (7).

We have reported suppression of the obese phenotype in A^y/a agouti but not ob/ob mice through the inbreeding of mice with a genetic mutation in Vgf (8). VGF is a secreted polypeptide that is expressed throughout the brain, selectively in neurons, as well as in several endocrine and neuroendocrine tissues including the pancreas and pituitary (reviewed in Refs.9 and 10). The lean, hypermetabolic, hyperactive phenotype of VGF mutant mice suggests that VGF may regulate energy balance by damping peripheral metabolic activity. In this study, we further characterize VGF function by examining the effect that Vgf ablation has on glucose homeostasis in VGF mutant mice, and in double-mutant mice generated by crossing VGF-deficient mice with three well-characterized genetically obese mouse models, ob/ob, A^y/a (agouti), and MC4R-deficient mice. Our studies suggest that targeted deletion of VGF reduces the hyperinsulinemia and hyperglycemia that result from high-carbohydrate/high-fat diets, gold thioglucose (GTG) lesions, and genetic defects in the hypothalamic melanocortin pathway, and does so in part through mechanisms that appear to be independent of an effect to reduce obesity.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Mouse strains, housing, and diets
Targeted deletion of the mouse Vgf gene has been previously described (11). Chimeric VGF knockout males were directly crossed to 129/SvJ or repetitively backcrossed to C57BL/6 strains for 10 generations; homozygous VGF-deficient offspring of F₁₀ and F₁ heterozygotes on either background were phenotypically indistinguishable. For the experiments with A^y/a and ob/ob, mixed background C57BL/6 x 129/SvJ VGF mutant mice from F₂ and F₃ generations were used for breeding. To generate double-mutant mice, fertile heterozygous Vgf⁺/Vgf^– female mice or bilaterally ovariectomized C57BL/6 x 129/SvJ females (Jackson Laboratory, Bar Harbor, ME) that had been grafted with Vgf^–/Vgf^– ovaries were used. These were mated with A^y/a (agouti) males, or with fertile ob^–/ob^– males (both obtained from Jackson Laboratory) that had been rescued by a course of ip leptin. Recombinant murine leptin (20 µg/g body weight; Amgen Inc., Thousand Oaks, CA) was delivered to ob^–/ob^– males twice daily for 7 d, then approximately every other day for 30–60 d. MC4R knockout mice (MC4R^–/MC4R^–; provided by Millenium Pharmaceuticals, Cambridge, MA) and Vgf^–/Vgf^– mice, both backcrossed for at least 10 generations to a C57BL/6 background, were used to generate homozygous double-mutant MC4R^–/MC4R^–,Vgf^–/Vgf^– mice.

Mice were housed in a 12-h light, 12-h dark cycle, with chow and water available ad libitum unless otherwise specified. For high-calorie diet studies, mice were fed a high-fat (33.5–35.5%) and high-simple-carbohydrate (34–35.5%) diet that also contained 20% protein and 0.1% fiber (Bio-Serv no. F2685; 5.4 kcal/g). The carbohydrate content includes approximately 51.3% disaccharide (sucrose) and 28.3% polysaccharide, and calories provided by protein, fat, and carbohydrate are 15.8, 58.2, and 26.0%, respectively. Mice fed standard chow received a 4.5% fat, 55% carbohydrate, 20% protein, 4.7% fiber diet (Purina PicoLab Rodent Diet 20–5053; 4 kcal/gm; Purina, St. Louis, MO). The carbohydrate content includes 5.7% sucrose and 58.6% polysaccharide, and calories provided by protein, fat, and carbohydrate are 23.6, 11.9, and 64.5%, respectively. All animal studies were conducted in accordance with the Guide for the Care and Use of Experimental Animals, using protocols approved by the Institutional Animal Care and Use Committee at Mount Sinai School of Medicine.

Chemical lesioning
Mice, 3–4 months of age, were administered a single ip injection of GTG (0.8 mg/g body weight) essentially as previously described (12), after which mice were weighed at regular intervals and food intake was measured. Two weeks after GTG (or saline) injection, mice were killed and tissues were removed for analysis. Separate groups of wild-type and VGF mutant mice were killed 3 d after GTG injection, and brains were removed and examined histologically by Nissl staining (11) to confirm that hypothalamic lesions developed in mice of each genotype. Monosodium glutamate (MSG) injections were carried out by minor modification of previous methods (13). Mice received the following daily sc injections starting at postnatal d 2 until postnatal d 12: 2.5, 2.8, 3.2, 3.4, 3.6, 3.8, 4.0, 4.2, 4.4, 4.6, and 4.8 mg MSG/g body weight (in PBS). Mice were weighed weekly and killed at 9 months of age, hypothalamic RNAs was isolated, and organ and tissue weights were determined.

Northern analysis
Northern analysis was carried out as described (14), using probes to neuropeptide Y (NPY) (15), POMC (16), and AGRP (17). Relative mRNA levels were determined by densitometric analysis of film autoradiograms.

Tissue staining and immunohistochemistry
Mice were killed and samples of liver and pancreas (tail and body region) were removed, fixed in formalin, and embedded in paraffin. Pancreatic and liver sections (5 µm) were stained with hematoxylin and eosin (H&E). Liver sections were also stained for glycogen using the periodic acid Schiff (PAS) method and counterstained with hematoxylin according to the manufacturer’s instructions (PAS Kit; Sigma Inc., St. Louis, MO). For immunohistochemistry, 5-µm sections of formalin-fixed, paraffin-embedded pancreas were dewaxed in xylene and hydrated through a descending series of alcohol. Endogenous peroxidase activity was blocked by treating sections with 3% H₂O₂ in methanol for 20 min. Sections were incubated overnight at 4 C with rabbit polyclonal antihuman glucagon (Dako, Carpenteria, CA; 1:150) and guinea pig polyclonal antiswine insulin (Dako; 1:100). Staining was performed using an avidin:biotinylated enzyme complex (Vector ABC kit, Vector Laboratories, Burlingame, CA) with diaminobenzidine as a substrate, and then counterstained with Mayer’s hematoxylin.

Blood and serum analysis
Mice were anesthetized with avertin and blood samples collected by cardiac puncture or tail vein. Blood glucose levels were determined using a One Touch Profile meter (LifeScan Inc., Mountain View, CA) or Glucometer Elite (Bayer Inc., Pittsburgh, PA). Serum insulin, leptin, T₄, and glucagon levels were determined by RIA (ICN Biomedicals, Inc., Costa Mesa, CA) or by ELISA (Crystal Chem Inc., Downers Grove, IL; and Linco Research, Inc., St. Charles, MO). Serum nonesterified free fatty acids (FFA) were determined by a colorimetric method according to the manufacturer’s instructions (Wako Chemicals USA, Inc., Richmond, VA). Sera were assayed for adiponectin levels as previously described (18).

Insulin and glucose tolerance tests
Wild-type, VGF mutant, MC4R-deficient, and double-mutant MC4R^–/MC4R^–,Vgf^–/Vgf^– mice were fasted either for 20 h or for 4 h, before ip injection with human insulin (2 U/kg body weight). For the determination of insulin sensitivity and counterregulatory responses, blood glucose levels were measured before and after the fast, then at hourly intervals, using a glucometer (Glucometer Elite, Bayer Inc. or One Touch Profile meter, Lifescan). Glucose tolerance was examined by fasting 4-month-old VGF-deficient and wild-type mice for 20 h, then allowing the mice to refeed orally with sweet milk [Carnation Sweetened Condensed Milk (4.2 g protein, 27.2 g carbohydrate, 4.6 g fat per 50 g; 165 kcal/50 g) diluted 1:1 with water]. Mice were allowed to feed for 15 min from liquid diet feeding tubes (Bio-serve, Frenchtown, NJ), after which food was withdrawn. Because refeeding responses (grams of food consumed over time) after a 20-h fast in VGF mutant and wild-type mice were identical using the same liquid delivery feeding system (11), ad libitum consumption of sweet milk was not quantified. Glucose levels were monitored, as described above, in blood samples taken 15, 45, 90, 120, 180, and 240 min after refeeding.

Hyperinsulinemic euglycemic clamp studies
After anesthesia with avertin (20 µl/g body weight), an incision was made in the right lateral region of the neck and a SILASTIC catheter (0.012 in x 0.025 in; Dow Corning Corp., Midland, MI) was inserted into the jugular vein and sutured in place, and the distal end of the vein was tied off. The catheter was externalized via an incision made in the dorsal midline region of the neck. Catheters were flushed with heparin to prevent coagulation, and mice were allowed to recover for 2 d. Food was withdrawn 4 h before the start of glucose clamp experiments.

Hyperinsulinemic euglycemic clamp analysis was carried out by minor modification of previously described methods (19). Mice were clamped at 80 mg glucose per deciliter of plasma by infusing 18 mU/kg·min of human insulin (Novolin) at a constant rate of 0.3 ml/h, and simultaneously infusing glucose (10% dextrose) at variable rates through the inserted catheter. The rate of glucose infusion was adjusted in response to blood glucose levels, determined every 10 min for approximately 90 min by glucometer (Glucometer Elite; Bayer Inc.); average rates of glucose infusion (milligrams per kilogram per minute) were calculated.

Insulin-mediated tissue-specific 2-deoxy-[1-³H]-D-glucose uptake in vivo
2-Deoxy-1-[³H]-D-glucose (2DG) (Sigma Inc.), a nonmetabolizable glucose analog, was used as a tracer to determine tissue glucose uptake. 2DG (0.5 µCi/g body weight) was mixed with 20% dextrose (2 g/kg body weight) to provide a fixed specific activity, and was injected ip with human insulin (Novolin, 1 U/kg). Blood was collected from the tail vein and glucose levels were determined every 15 min for 1 h, at which time the animal was killed. Various tissues were removed, flash frozen on dry ice, and stored at –80 C until use. At each collection time point, 10 µl blood was mixed with 190 µl hemolysis reagent (50 mg/liter of digitonin and 100 mg/liter of maleimide), and the hemolized blood was then mixed with scintillation fluid (Opti-Fluor LSC; Packard, Meriden, CT) for the measurement of ³H radioactivity as described (20) using a Beckman model LS6500 scintillation counter. The glucose-specific activity was determined by dividing the ³H radioactivity by the glucose concentration at each particular time point (yielding disintegrations per minute per micromole). The mean glucose-specific activity was determined by dividing the phosphorylated 2-deoxyglucose ³H radioactivity by the integrated area under the plasma time-activity curve of 2DG-phosphate (2DG-P) (21).

In most tissues, the first step in glycolysis is phosphorylation, and once phosphorylated, 2DG-P is generally trapped, allowing determination of glucose uptake based on quantification of [³H]2DG-P tissue levels (22). Tissue samples weighing from 0.02–0.50 g were homogenized in 2 ml of water, and the homogenate was transferred to an equal amount of ice cold 7% perchloric acid. After 30 min, the samples were centrifuged at 2000 x g, and the supernatants were then neutralized with 2.2 mol KHO₃. One half of the supernatant was used to count total ³H radioactivity, and the other half was loaded onto an anion exchange column (AG 1-X8, 200–400 mesh, acetate form; Sigma Inc.) and washed with water to remove unphosphorylated glucose, and then 2DG-P was eluted with 0.4 M ammonium acetate/0.5 M formic acid. The tissue glucose uptake was determined from the ratio of tissue [³H]2DG-P to the mean glucose-specific activity.

Unlike other tissues, the liver contains glucose phosphatases, so 2DG incorporation into glycogen, rather than 2DG-P levels, is used as a measure of glucose uptake. Isolation of [³H]glycogen was carried out as previously described (20). Briefly, liver (0.05 g) was homogenized, neutralized, and 500 µl of the supernatant was treated with KOH containing 2.5% Na₂SO₄. Glycogen was precipitated by addition of 1 ml of ethanol to each sample, the pellet was resuspended in 1 ml of distilled water, and ³H radioactivity was counted.

Statistical analysis
Data are presented as mean ± SEM. Data were analyzed by two-way ANOVA, and when indicated by the appropriate P value (P < 0.05), groups were compared using a Tukey-Kramer post hoc test. Where indicated, paired t tests were used in the measurement of counterregulatory responses and to detect changes in insulin secretion by comparing blood glucose and/or serum insulin levels before and after receiving insulin and/or glucose injections. P < 0.05 was considered significant; P and n values are noted in the figures and/or accompanying legends.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
VGF mutant mice are resistant to developing hyperinsulinemia on a high-calorie diet
VGF-deficient mice were initially tested to determine whether a high-fat, high-carbohydrate diet would induce weight gain. Wild-type controls and VGF mutants were ad libitum fed for 5 wk, and food intake and weights were quantified. The body weights and adiposity of these VGF mutant mice, fed regular or high-fat diets, were previously reported and did not differ significantly (8). The wild-type mice fed high-calorie diets were all found to have substantially increased body weights and adiposity in comparison to mice fed regular diets (8). In addition, mean plasma glucose levels were higher in both mutant and wild-type mice fed the high-calorie diet, whereas circulating insulin and leptin levels were elevated in wild-type mice but unchanged in VGF mutant mice fed high-calorie in comparison to regular diets (Fig. 1, A–C).

View larger version (20K): [in this window] [in a new window]   FIG. 1. VGF mutant mice resist diet-induced hyperinsulinemia and obesity. Wild-type and VGF mutant mice were placed either on regular laboratory diets (Chow) or high-fat diets (HF) for 5 wk. Mice were then weighed and anesthetized, and blood and hypothalamic tissue were removed for analysis. Histograms identified by different letters are significantly different from one another (P < 0.05; ANOVA with Tukey-Kramer post hoc test).

Glucose use and insulin sensitivity may be higher in VGF mutant mice
Because VGF mutant mice have slightly but significantly lower plasma glucose levels than control mice (Ref.11 and Fig. 1A), we investigated whether the responses of VGF-deficient mice to fasting and insulin administration or to fasting and refeeding were different from control mice. To measure glucose tolerance, wild-type and VGF mutant mice were fasted for 20 h, then were refed sweet milk orally, and plasma glucose levels were monitored as shown in Fig. 2. Fasting caused a similar 3.5- to 3.7-g weight loss in both knockout and wild-type mice, and refeeding responses after fasting were similar as previously reported (11). Plasma glucose levels increased more slowly in VGF-deficient mice, plateaued at a similar level, then declined more rapidly to levels that were significantly lower than control mice (Fig. 2A). Insulin sensitivity and counterregulatory responses in the fasted mouse were also measured. Mice were fasted for 20 h then injected with insulin (2 U/kg body weight ip). Plasma glucose levels were measured before and after the fast, then at hourly intervals after insulin injection (Fig. 2B). Both wild-type and VGF mutant mice responded to insulin injection with a rapid decrease in plasma glucose. However, VGF-deficient mice appeared to be more sensitive to an identical insulin dose per kilogram of body weight, which led to a greater and more prolonged decrease in their plasma glucose levels. In wild-type mice, plasma glucose levels rose to 70 ± 11 mg/dl 4 h after insulin injection and remained at this level, whereas in VGF-deficient mice, plasma glucose levels remained between 25–31 mg/dl over the duration of the experiment.

View larger version (20K): [in this window] [in a new window]   FIG. 2. Glucose use may be higher and counterregulatory responses may be reduced in VGF mutant mice. A, Response of plasma glucose levels, in VGF-deficient (; n = 5) and wild-type mice (; n = 6), to fasting for 20 h and refeeding with sweet milk is shown. B, Counterregulatory response of plasma glucose levels, in VGF-deficient (; n = 4) and wild-type mice (; n = 5), to fasting and insulin injection is shown. Plasma glucose determinations from wild-type mice that are identified by the asterisks are significantly different from the corresponding levels in VGF mutant mice (P < 0.05; ANOVA with Tukey-Kramer post hoc test).

View larger version (107K): [in this window] [in a new window]   FIG. 3. Immunohistochemical staining of pancreatic islets with anti-insulin and antiglucagon antisera is similar in wild-type and VGF-deficient mice. Pancreatic islets from VGF mutant (A and C) and wild-type (B and D) were stained with antiinsulin antiserum (A and B) and antiglucagon antiserum (C and D; arrows point to increased glucagon immunoreactivity toward the islet periphery). Antisera recognize both mature and prohormone forms of either insulin or glucagon. Scale bars (A–D) are 100 µm in length. H&E staining of typical islets at the same magnification from mutant (E) and wild-type mice (F) demonstrates the smaller size of islets in VGF-deficient mice.

View larger version (115K): [in this window] [in a new window]   FIG. 4. Liver glycogen stores are reduced in Vgf–/Vgf mice. H&E stained liver sections from ad libitum-fed Vgf–/Vgf– (A) and age-matched Vgf+/Vgf+ (B) mice. Note the granular cytoplasm with areas of clearing in wild-type hepatocytes (B), and the absence of this in hepatocytes from VGF-deficient mice (A). C and D, Livers were stained for glycogen using the PAS method. Note that glycogen deposits (stained pink) are present in wild-type liver (D) and rarely found in livers from mutant mice (C). Scale bars (A–D) are 100 µm in length.

View larger version (16K): [in this window] [in a new window]   FIG. 5. Intraperitoneal glucose injection stimulates insulin responses in Vgf–/Vgf– and Vgf+/Vgf+ mice, and counterregulatory responses of plasma glucose levels to injected insulin are reduced but present in Vgf–/Vgf– mice. A, Vgf–/Vgf– (n = 3) and Vgf+/Vgf+ (n = 3) mice were fasted for 20 h, and serum insulin levels were measured before (white bars) and 30–45 min after (black bars) receiving an ip glucose injection (2 mg/g of body weight) (mean ± SEM, P < 0.05, paired t test). B, Ad libitum-fed Vgf–/Vgf– () and Vgf+/Vgf+ () mice were injected with insulin and fasted for 4 h. Plasma glucose levels were determined at time 0 immediately before insulin injection, and after subsequent 1-h intervals. Bars that are identified by asterisks are significantly different from the fasted glucose levels in mutant (n = 3) and wild-type (n = 3) mice (mean ± SEM, P < 0.05, paired t test).

Hyperinsulinemic euglycemic clamp analysis suggests that VGF mutant mice have increased insulin sensitivity and/or glucose use in comparison to wild-type mice
Because oral glucose tolerance tests suggested that VGF mutant mice had increased glucose tolerance and insulin sensitivity, we further characterized glucose homeostasis in these mice using hyperinsulinemic euglycemic clamp analysis (19). VGF mutant mice displayed a robust increase in glucose clearance, demonstrated by the significantly higher rate of glucose infusion that was required in VGF knockout in comparison to control mice to maintain blood glucose levels of 80 mg/dl (Fig. 6).

View larger version (16K): [in this window] [in a new window]   FIG. 6. Hyperinsulinemic euglycemic glucose clamp analysis of VGF mice is consistent with increased insulin sensitivity. The rate of glucose infusion, adjusted for body weight (milligram of glucose per kilogram of body weight per minute), during hyperinsulinemic euglycemic clamping was increased in Vgf–/Vgf– mice (n = 4) in comparison to Vgf+/Vgf+ mice (n = 9) (P 0.0001, ANOVA, mean ± SEM).

View larger version (17K): [in this window] [in a new window]   FIG. 7. Uptake of 2DG into VGF mutant mouse adipose tissues is increased compared with wild-type mice. VGF-deficient mice had increased glucose uptake into WAT (P < 0.05) and BAT (P < 0.05), decreased glucose uptake into spleen (P < 0.05), and unchanged glucose uptake into skeletal muscle (calf) in comparison to wild-type mice (A). B, Glucose incorporation into glycogen was not significantly reduced in the livers of VGF mutant mice in comparison to wild-type mice. Comparisons by ANOVA [mean ± SEM; Vgf–/Vgf– (n = 3), Vgf+/Vgf+ (n = 6)].

View larger version (13K): [in this window] [in a new window]   FIG. 8. VGF mutant mice resist hyperglycemia caused by GTG-induced, but not MSG-induced, chemical lesions. Wild-type and VGF-deficient mice were lesioned with GTG and plasma glucose levels were determined at 16 wk of age (A). B, Wild-type, heterozygous and homozygous VGF-deficient mice were lesioned with MSG and plasma glucose levels were measured at 36 wk of age. Histograms identified by different letters are significantly different from one another (P < 0.05; ANOVA with Tukey-Kramer post hoc comparison; n = 4–6 mice per treatment group).

VGF is required for the development of hyperglycemia and hyperinsulinemia in A^y/a (agouti) mice
Ectopic overexpression of the agouti polypeptide, a melanocortin receptor antagonist, results in a late onset obesity syndrome in agouti lethal yellow (A^y/a) mice. The phenotype caused by this mutation is recapitulated in other strains of mice with abnormal melanocortin signaling, including those that overexpress the endogenous MC4R antagonist called AGRP or ART, mice deficient in the MC4R receptor, and strains with a targeted deletion of the gene encoding the POMC peptide precursor (see Ref.28 and references therein). To determine whether VGF is required for the development of this form of maturity onset obesity, we generated agouti mice, agouti mice that were either heterozygous or homozygous for the targeted VGF mutation, and VGF mutant mice. Agouti mice that were also homozygous for the VGF mutation were indistinguishable from VGF mutant mice in body weight, body composition, and crown-to-rump length (8). In addition, analysis of glucose and insulin levels in the same mice revealed that plasma glucose levels were significantly decreased in VGF-deficient mice with or without the agouti mutation in comparison to wild-type or agouti mutant mice (Fig. 9A). Mean serum insulin levels in double-mutant mice were significantly lower than those measured in A^y/a agouti mice and significantly higher than VGF mutant mice, but were not significantly different from control levels (Fig. 9B).

View larger version (16K): [in this window] [in a new window]   FIG. 9. VGF is required for the development of hyperglycemia and hyperinsulinemia in Ay/a and hyperglycemia in ob/ob mice. Plasma glucose (A) and serum insulin (B) levels were measured in VGF mutant (Vgf–/Vgf–), wild-type, agouti (Ay/a), and double-mutant (Vgf–/Vgf–,Ay/a) mice. C, Plasma glucose levels were measured in VGF mutant (Vgf–/Vgf–), ob/ob, double-mutant (Vgf–/Vgf–,ob/ob), and wild-type mice. Histograms identified by different letters are significantly different from one another (P < 0.05; ANOVA with Tukey-Kramer post hoc test; n = number of independent animals sampled).

Ablation of VGF prevents the development of obesity and increases insulin sensitivity in MC4R-deficient mice
To extend and corroborate the findings obtained with A^y/a,Vgf^–/Vgf^– double-knockout mice, VGF was ablated from MC4R^–/MC4R^– mice. As noted with A^y/a agouti mice, depletion of VGF from MC4R^–/MC4R^– mice resulted in lean animals with reduced body weights and almost undetectable mesenteric (data not shown) and epididymal fat (Table 1). Blood glucose and serum insulin levels in the MC4R^–/MC4R^–,Vgf^–/Vgf^– mice were reduced in comparison to age-matched MC4R knockout and wild-type mice (Fig. 10, A and B). This occurred even at 7–8 wk of age, when the weights and epididymal fat pads of MC4R mutant mice and wild-type mice were indistinguishable (see Table 1 and Ref.5), and previous studies have reported that impaired insulin tolerance in MC4R knockout mice can already be detected (5). Even though body weight was reduced in double-mutant mice, food intake was comparable in wild-type, VGF knockout and double-knockout mice (data not shown). Histological analysis of liver from wild-type (MC4R⁺/MC4R⁺,Vgf⁺/Vgf⁺), MC4R^– knockout (MC4R^–/MC4R^–,Vgf⁺/Vgf⁺), and double-knockout (MC4R^–/MC4R^–,Vgf^–/Vgf^–) mice indicated that glycogen stores were substantially reduced in the livers of double-mutant mice (Fig. 10, C–E).

View larger version (82K): [in this window] [in a new window]   FIG. 10. VGF is required for the development of hyperglycemia and decreased insulin sensitivity in MC4R–/MC4R– mice. Blood glucose levels were compared in double-mutant (MC4R–/MC4R–,Vgf–/Vgf–), Vgf–/Vgf–, MC4R–/MC4R, and wild-type mice (A) at 7.5–10 wk of age. B, Serum insulin levels were quantified in double-mutant (MC4R–/MC4R–,Vgf–/Vgf–), Vgf–/Vgf–, and wild-type mice at 7.5–10 wk of age. A and B, Histograms identified by different letters are significantly different from one another (P < 0.05; ANOVA with Tukey-Kramer post hoc test; n = number of independent animals sampled). C and D, H&E staining of liver from double-mutant (MC4R–/MC4R–,Vgf–/Vgf–) (C), wild-type (D), and MC4R–/MC4R– (E) mice is shown. Note that areas of clearing, corresponding to glycogen deposits, are increased in MC4R–/MC4R– (E) and wild-type (D) compared with double-mutant (MC4R–/MC4R–,Vgf–/Vgf–) (C) mice.

View larger version (24K): [in this window] [in a new window]   FIG. 11. Counterregulatory responses of blood glucose levels to insulin injection are blunted in fasted, double-mutant MC4R–/MC4R, Vgf–/Vgf– mice in comparison to age-matched MC4R–/MC4R–/Vgf+/Vgf+ and wild-type mice. Sixteen-week-old mice were fasted for 20 h then injected with insulin (2 U/kg) ip, and blood glucose levels were determined hourly thereafter. Glucose levels that are marked with an asterisk from MC4R–/MC4R–/Vgf+/Vgf+ (n = 2) and wild-type mice (n = 3) are significantly different from the corresponding levels in the double-knockout mice (n = 2) (P < 0.05; mean ± SEM, paired t test).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
In this study, we investigated the effect that targeted deletion of the neuronal, endocrine, and neuroendocrine secreted polypeptide VGF has on glucose homeostasis. We previously noted that VGF mutant mice generally maintain lower-than-normal circulating glucose levels and low normal insulin levels (11). Decreased insulin levels in these mice did not appear to result from impaired processing of proinsulin and proglucagon and/or defective secretion of these hormones, as these would likely have had profound effects on glucose homeostasis (24, 35), which were not observed. In addition, histological analysis showed a normal distribution of glucagon-positive -cells and insulin-positive ß-cells in pancreatic islets (Fig. 3). Furthermore, the ability to increase circulating insulin levels after a glucose challenge and elevated circulating glucagon levels together suggest that VGF mutant mice do not have a defect in the processing and/or secretion of these hormones. Direct measurement of plasma glucose levels in fasted VGF mutant mice after either oral glucose or ip insulin administration, and higher rates of glucose infusion during hyperinsulinemic euglycemic clamping, indicated rather that these mice have increased insulin sensitivity and/or an increased rate of peripheral glucose use in comparison to wild-type mice. One might anticipate that increased glucose use could result from an increased basal metabolic rate and/or increased insulin sensitivity. Of note similar, sizeable increases in glucose infusion rates and peripheral tissue glucose uptake have been reported in lean, hypermetabolic mice with heterozygous disruption of the Gnas paternal allele (G protein -subunit G_s gene) (36). In addition, lean melanin concentrating hormone mutant (37) and neuromedin U transgenic mice (38) have improved glucose tolerances.

Consistent with increased insulin sensitivity, VGF mutant mice had increased glucose uptake into adipose tissue after insulin administration in comparison to wild-type mice. Adipose tissues rely predominately on insulin for glucose uptake via insulin-induced activation of glucose transporters (39), whereas other peripheral tissues such as skeletal muscle do not rely solely on insulin action for glucose uptake (40). BAT is the most highly metabolic tissue in rodents; the higher rate of glucose uptake into BAT could suggest increased thermogenesis and/or fuel use in this tissue. The unchanged rate of glucose uptake into skeletal muscle suggests that skeletal muscle might be using FA and sparing glucose for tissues such as brain that cannot use FA as an energy source. Lastly, hepatocytes are freely permeable to glucose and hence do not need insulin-induced activation of glucose transporters for glucose entry into the liver, although insulin signaling in the hypothalamus is required for the rapid regulation of hepatic glucose production (41). Because wild-type and VGF-deficient mice each incorporate similar amounts of radioactive glucose into hepatic glycogen (Fig. 7), VGF mutant mice can still synthesize glycogen, but likely have increased peripheral demand for glucose that leads to a decrease in hepatic glycogen storage (Fig. 4).

Alterations in both hypothalamic gene expression and plasma glucose levels were noted in VGF mutant mice fed the high-carbohydrate, high-fat diet. Because this high-calorie diet is complex and we did not experiment further to analyze the individual components, we do not know whether the increase in circulating plasma glucose levels in VGF mutant mice resulted from the increased dietary fat and/or simple carbohydrate (sucrose levels are higher, whereas the total carbohydrate content is substantially lower than regular chow; see Materials and Methods). Although blood glucose levels increased in wild-type mice on the high-calorie diet, hypothalamic POMC mRNA levels remained unchanged, consistent with previous studies that correlated increased susceptibility of C57BL/6 mice to diet-induced obesity with their failure to regulate hypothalamic POMC and NPY gene expression (23). Significant alterations in hypothalamic AGRP, NPY, and POMC mRNA levels in high-fat-fed VGF mutant mice correlated well with their resistance to diet-induced obesity (and was similar to diet-induced changes in hypothalamic gene expression in A/J mice, a strain that does not become obese on a high-fat diet) (23). Previous studies have suggested the possibility that elevated leptin levels in high-fat-fed mice may be responsible for decreases in hypothalamic AGRP and NPY (42). Because we observed altered hypothalamic gene expression in high-fat-fed VGF knockout mice in the absence of significant changes in adipose leptin mRNA or serum leptin levels, molecules in addition to leptin are likely to be involved in communicating the state of peripheral energy balance to the hypothalamus. Alternative possible effectors of hypothalamic gene expression include glucose and thyroid hormone, and of these, we noted no significant change in serum T₄ but a significant increase in plasma glucose in high-fat-fed compared with regular chow-fed VGF knockouts in the absence of changes in serum insulin. That glucose regulates hypothalamic POMC, AGRP, and NPY mRNA levels in VGF mutant mice fed a high-fat, high-carbohydrate diet is consistent with previous studies that demonstrated regulation of these mRNAs independent of leptin and insulin under certain circumstances (43), and with the demonstration that induction of NPY and AGRP mRNAs in the arcuate nucleus by glucoprivation is mediated by hindbrain, catecholaminergic, glucose-sensing neurons that project to the hypothalamus (44). Our studies provide additional support for the hypothesis that circulating glucose is an important messenger coupling appetite and metabolic state.

What causes increased insulin sensitivity in VGF-deficient mice? Reduction in fat mass is certainly a major contributor to increased insulin sensitivity, but the extent to which increased physical activity, increased fat oxidation, and increased energy expenditure each contribute to improved glucose homeostasis or to reduced adiposity in VGF mutant mice is currently unknown. Exercise coupled with reduced obesity improves insulin sensitivity and glucose disposal (45), changes circulating adipokine levels (46), and is also associated with increased rates of fat oxidation (47), so it remains possible that a significant component of the improved glucose homeostasis in these mice is a function of their increased physical activity.

Do circulating factors increase insulin sensitivity in VGF mutant mice? On the one hand, these mice have increased FFA levels, which have been linked to insulin resistance (48), whereas, on the other hand, VGF-deficient mice have increased levels of serum adiponectin, which is known to enhance FA oxidation and glucose uptake (31). The precise mechanism by which FFA regulate insulin resistance is incompletely understood; recent studies suggest that insulin resistance might result from accumulated intramuscular FFA metabolites in skeletal muscle and/or liver, rather than from a direct effect of elevated circulating FFA levels on insulin signaling (49, 50, 51). Interestingly, activation of the melanocortin pathway by central nervous system delivery of recombinant adeno-associated virus encoding POMC to Zucker rats with defective leptin receptors results in a 33% increase in circulating FFA levels, reduced adiposity, and improved insulin sensitivity (52), also suggesting that increased serum FFAs do not precisely correlate with insulin resistance. The increase in circulating FFA during the fed state in VGF-deficient mice is most likely due to a continuous breakdown of fat in adipose tissue, which would be consistent with their reduced fat mass (11); decreased FFA levels in Vgf^–/Vgf^– mice during fasting indicates that FFA are being used for fuel. It has also been reported that an increase in FFA levels is required for increased energy expenditure and lipid oxidation during ß-adrenergic-stimulated thermogenesis (53), a potential mechanism suggested to underlie the hypermetabolic state in VGF knockout mice (8).

Elevated serum adiponectin levels in VGF mutant mice may be responsible for barely detectable circulating FFA levels during fasting. Adiponectin activates AMPK in liver (54). AMPK inhibits acetyl-CoA carboxylase and activates malonyl-CoA dehydrogenase, and hence carnitine palmitoyltransferase 1 is disinhibited, allowing increased FA transport into mitochondria for FA oxidation (30). As noted above, VGF is required for the development of hyperinsulinemia and hyperglycemia in several transgenic models of obesity. Interestingly, ablation of VGF significantly elevated adiponectin levels in MC4R^–/MC4R^– mice at 8 wk of age, before the development of an obese phenotype in MC4R mutant mice. It is possible then that the rise in circulating adiponectin levels may function to inhibit the later development of obesity and insulin resistance in combined VGF- and MC4R-deficient mice. The increase in circulating ketone bodies and the reduced glycogen stores in the livers of VGF-deficient mice suggest that FA oxidation and gluconeogenesis are both increased to accommodate the demand for energy.

We noted that neonatal MSG treatment, which damages the hypothalamus and hypothalamic projections to the autonomic nervous system (12, 27, 55), blocked the development of the lean phenotype in Vgf^–/Vgf^– mice. In the light of these data, and similar findings after more selective global sympathectomy with guanethidine (Watson, E., and S. Salton, unpublished data), it is probable that an intact sympathetic nervous system is necessary for the development of the lean, hypermetabolic phenotype in VGF mutant mice. Increased expression of the adipocyte-derived circulating protein adiponectin has previously been associated with chronic sympathetic activation (56); therefore, increased circulating adiponectin levels in VGF-deficient mice could be the result of increased sympathetic tone in combination with reduced fat mass.

Targeted deletion of VGF prevented the development of hyperinsulinemia and hyperglycemia induced by diet and-specific genetic mutations by decreasing fat accumulation in older mice, and through mechanisms that may also be independent of altered adiposity. Ablation of Vgf in A^y/a (agouti) mice blocked the development of obesity, hyperglycemia, and hyperinsulinemia. Although food intake was similar (8), plasma glucose levels were lower in double-mutant mice compared with A^y/a littermates, and the hyperinsulinemia that is associated with the agouti mutation was not found in double-mutant A^y/a,Vgf^–/Vgf^– mice. In addition, double-mutant MC4R^–/MC4R^–,Vgf^–/Vgf^– mice displayed plasma glucose levels and insulin responses that were more similar to normal and VGF-deficient mice than to MC4R^–/MC4R^– mice. In older double-mutant mice (3–5 months of age), this is very likely due to the failure of these mice to develop obesity, much as had been seen for double-mutant A^y/a,Vgf^–/Vgf^– mice. However similar plasma glucose levels and insulin responses were obtained with MC4R^–/MC4R^–,Vgf^–/Vgf^– and Vgf^–/Vgf^– mice that were examined at 7–8 wk of age, before the development of increased body mass and increased epididymal fat pad weights by age-matched MC4R^–/MC4R^– mice when compared with control mice (Table 1). Previous studies have demonstrated that young, lean MC4R^–/MC4R^– mice at this age have already developed significantly increased plasma insulin levels and decreased peripheral insulin tolerance, although their body weights are indistinguishable from wild-type mice (5). Although wild-type and MC4R^–/MC4R^– mice are equivalently lean at 7–8 wk of age, double-mutant MC4R^–/MC4R^–,Vgf^–/Vgf^– have substantially smaller fat depots than either (Table 1). This reduction in fat mass below wild-type levels might alter the profile of adipokines (e.g. leptin, resistin, or adiponectin) secreted by adipose tissue into the circulation, significantly impacting insulin sensitivity. Alternatively, the effect of Vgf gene targeting in 7–8 wk old MC4R^–/MC4R^– mice to improve glucose homeostasis might be independent, at least in part, of a simple reduction in adiposity and/or body mass, and perhaps is the result of an impact on the central melanocortin system that directly regulates serum insulin levels (5, 7).

Consistent with an effect of VGF ablation on glucose homeostasis that is not solely dependant on a reduction in peripheral fat stores, ob/ob,Vgf^–/Vgf^– mice consume less food and maintain lower circulating insulin and glucose levels than do ob/ob mice (this study), although they have very similar adiposity and body composition (8). In addition, targeted deletion of VGF improved glucose homeostasis in ob/ob,Vgf^–/Vgf^– mice in the absence of leptin; whether this occurs through an enhancement of insulin receptor sensitivity, much as occurs after leptin administration for 2 d to ob/ob mice (57), remains to be investigated.

It is currently unknown whether the lower plasma glucose levels in double-mutant (ob/ob,Vgf^–/Vgf^–), (A^y/a,Vgf^–/Vgf^–), and (MC4R^–/MC4R^–,Vgf^–/Vgf^–) mice result from increased peripheral demand or increased sensitivity to circulating insulin, although the latter seems more likely. At least for double-mutant ob/ob,Vgf^–/Vgf^– mice that maintain low core body temperatures and have comparable adipose stores to ob/ob mice (8), lower circulating insulin and glucose levels are probably not a result of increased peripheral glucose use. The melanocortin pathway has been identified as an essential pathway for both glucose and energy homeostasis, and functions by reducing insulin release from the pancreas and by increasing thermogenesis in BAT (5, 52). Studies to determine whether secreted VGF and/or VGF-derived peptides modify melanocortin pathway function, and where in hypothalamic-autonomic outflow circuits these peptides act, will be required to better understand the mechanism of action of this gene product.

	Footnotes

 
This work was supported in part by grants from the National Institutes of Health (to C.V.M., P.E.S., S.H., and S.R.J.S.), American Heart Association (to S.R.J.S.), and Culpeper Foundation (to S.R.J.S.), by a National Insitutes of Health Endocrine Training Grant 5T32DK07645 Predoctoral Fellowship (to E.W.), and by a Career Scientist Award from the Irma T. Hirschl and Monique Weill-Caulier Trusts (to S.R.J.S.).

Present address for S.H.: Wyeth Research, Cardiovascular and Metabolic Diseases, 200 Cambridgepark Row, Cambridge, Massachusetts 02140.

Present address for T.M.M.: Department of Physiology, University of Manitoba, 730 William Avenue, Winnipeg, Manitoba R3E 3J7, Canada.

Present address for J.W.: Regeneron Pharmaceuticals Inc., 777 Old Saw Mill River Road, Tarrytown, New York 10591.

First Published Online September 1, 2005

Abbreviations: AGRP, Agouti-related protein; AMPK, AMP-activated kinase; BAT, brown adipose tissue; 2DG, 2-deoxy-1-[³H]-D-glucose; 2DG-P, 2DG-phosphate; FA, fatty acid; FFA, free fatty acid; GTG, gold thioglucose; H&E, hematoxylin and eosin; MC4R, melanocortin-4 receptor; MSG, monosodium glutamate; NPY, neuropeptide Y; PAS, periodic acid Schiff; POMC, proopiomelanocortin; WAT, white adipose tissue.

Received May 16, 2005.

Accepted for publication August 22, 2005.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Ollmann MM, Wilson BD, Yang YK, Kerns JA, Chen Y, Gantz I, Barsh GS 1997 Antagonism of central melanocortin receptors in vitro and in vivo by agouti-related protein. Science 278:135–138[Abstract/Free Full Text]
2. Shutter JR, Graham M, Kinsey AC, Scully S, Luthy R, Stark KL 1997 Hypothalamic expression of ART, a novel gene related to agouti, is up- regulated in obese and diabetic mutant mice. Genes Dev 11:593–602[Abstract/Free Full Text]
3. Huszar D, Lynch CA, Fairchild-Huntress V, Dunmore JH, Fang Q, Berkemeier LR, Gu W, Kesterson RA, Boston BA, Cone RD, Smith FJ, Campfield LA, Burn P, Lee F 1997 Targeted disruption of the melanocortin-4 receptor results in obesity in mice. Cell 88:131–141[CrossRef][Medline]
4. Yaswen L, Diehl N, Brennan MB, Hochgeschwender U 1999 Obesity in the mouse model of pro-opiomelanocortin deficiency responds to peripheral melanocortin. Nat Med 5:1066–1070[CrossRef][Medline]
5. Fan W, Dinulescu DM, Butler AA, Zhou J, Marks DL, Cone RD 2000 The central melanocortin system can directly regulate serum insulin levels. Endocrinology 141:3072–3079[Abstract/Free Full Text]
6. Friedman JM, Halaas JL 1998 Leptin and the regulation of body weight in mammals. Nature 395:763–770[CrossRef][Medline]
7. Mizuno TM, Kelley KA, Pasinetti GM, Roberts JL, Mobbs CV 2003 Transgenic neuronal expression of proopiomelanocortin attenuates hyperphagic response to fasting and reverses metabolic impairments in leptin-deficient obese mice. Diabetes 52:2675–2683[Abstract/Free Full Text]
8. Hahm S, Fekete C, Mizuno TM, Windsor J, Yan H, Boozer CN, Lee C, Elmquist JK, Lechan RM, Mobbs CV, Salton SR 2002 VGF is required for obesity induced by diet, gold thioglucose treatment and agouti, and is differentially regulated in POMC- and NPY-containing arcuate neurons in response to fasting. J Neurosci 22:6929–6938[Abstract/Free Full Text]
9. Salton SR, Ferri GL, Hahm S, Snyder SE, Wilson AJ, Possenti R, Levi A 2000 VGF: a novel role for this neuronal and neuroendocrine polypeptide in the regulation of energy balance. Front Neuroendocrinol 21:199–219[CrossRef][Medline]
10. Levi A, Ferri GL, Watson E, Possenti R, Salton SR 2004 Processing, distribution and function of VGF, a neuronal and endocrine peptide precursor. Cell Mol Neurobiol 24:517–533[CrossRef][Medline]
11. Hahm S, Mizuno TM, Wu TJ, Wisor JP, Priest CA, Kozak CA, Boozer CN, Peng B, McEvoy RC, Good P, Kelley KA, Takahashi JS, Pintar JE, Roberts JL, Mobbs CV, Salton SR 1999 Targeted deletion of the Vgf gene indicates that the encoded secretory peptide precursor plays a novel role in the regulation of energy balance. Neuron 23:537–548[CrossRef][Medline]
12. Bergen HT, Mizuno TM, Taylor J, Mobbs CV 1998 Hyperphagia and weight gain after gold-thioglucose: relation to hypothalamic neuropeptide Y and proopiomelanocortin. Endocrinology 139:4483–4488[Abstract/Free Full Text]
13. Pizzi WJ, Barnhart JE 1976 Effects of monosodium glutamate on somatic development, obesity and activity in the mouse. Pharmacol Biochem Behav 5:551–557[CrossRef][Medline]
14. Mizuno TM, Bergen H, Funabashi T, Kleopoulos SP, Zhong YG, Bauman WA, Mobbs CV 1996 Obese gene expression: reduction by fasting and stimulation by insulin and glucose in lean mice, and persistent elevation in acquired (diet-induced) and genetic (yellow agouti) obesity. Proc Natl Acad Sci USA 93:3434–3438[Abstract/Free Full Text]
15. Mizuno T, Bergen H, Kleopoulos S, Bauman WA, Mobbs CV 1996 Effects of nutritional status and aging on leptin gene expression in mice: importance of glucose. Horm Metab Res 28:679–684[Medline]
16. Mizuno TM, Kleopoulos SP, Bergen HT, Roberts JL, Priest CA, Mobbs CV 1998 Hypothalamic pro-opiomelanocortin mRNA is reduced by fasting and [corrected] in ob/ob and db/db mice, but is stimulated by leptin. Diabetes 47:294–297[Abstract]
17. Mizuno TM, Mobbs CV 1999 Hypothalamic agouti-related protein messenger ribonucleic acid is inhibited by leptin and stimulated by fasting. Endocrinology 140:814–817[Abstract/Free Full Text]
18. Combs TP, Berg AH, Rajala MW, Klebanov S, Iyengar P, Jimenez-Chillaron JC, Patti ME, Klein SL, Weinstein RS, Scherer PE 2003 Sexual differentiation, pregnancy, calorie restriction, and aging affect the adipocyte-specific secretory protein adiponectin. Diabetes 52:268–276[Abstract/Free Full Text]
19. Rossetti L, Barzilai N, Chen W, Harris T, Yang D, Rogler CE 1996 Hepatic overexpression of insulin-like growth factor-II in adulthood increases basal and insulin-stimulated glucose disposal in conscious mice. J Biol Chem 271:203–208[Abstract/Free Full Text]
20. Wang S, Subramaniam A, Cawthorne MA, Clapham JC 2003 Increased fatty acid oxidation in transgenic mice overexpressing UCP3 in skeletal muscle. Diabetes Obes Metab 5:295–301[CrossRef][Medline]
21. Jenkins AB, Furler SM, Kraegen EW 1986 2-Deoxy-D-glucose metabolism in individual tissues of the rat in vivo. Int J Biochem 18:311–318[CrossRef][Medline]
22. Mauvais-Jarvis F, Virkamaki A, Michael MD, Winnay JN,