There is an Erratum (March 2002) associated with this Article.
Small molecule insulin mimetics reduce food intake and body weight and prevent development of obesity
Ellen L. Air1, 5, Mathias Z. Strowski2, 5, Stephen C. Benoit3, Stacey L. Conarello2, Gino M. Salituro4, Xiao-Ming Guan2, Kun Liu3, Stephen C. Woods3
& Bei B. Zhang2
1 Departments of Biomedical Sciences and Cell Biology, Neurobiology and Anatomy, University of Cincinnati College of Medicine, Cincinnati, Ohio, USA
2 Department of Molecular Endocrinology and Metabolic Disorders, Merck Research Laboratories, Rahway, New Jersey, USA
3 Department of Psychiatry, University of Cincinnati College of Medicine, Cincinnati, Ohio, USA
4 Department of Medicinal Chemistry, Merck Research Laboratories, Rahway, New Jersey, USA
5 E.L.A. and M.Z.S. contributed equally to this paper.
Obesity and insulin resistance are major risk factors for a number of metabolic disorders, such as type 2 diabetes mellitus1,
2. Insulin has been suggested to function as one of the adiposity signals to the brain for modulation of energy balance. Administration of insulin into the brain reduces food intake and body weight3,
4,
5, and mice with a genetic deletion of neuronal insulin receptors are hyperphagic and obese6. However, insulin is also an anabolic factor; when administered systemically, pharmacological levels of insulin are associated with body weight gain in patients7. In this study, we investigated the efficacy and feasibility of small molecule insulin mimetic compounds8,
9 to regulate key parameters of energy homeostasis. Central intracerebroventricular (i.c.v.) administration of an insulin mimetic resulted in a dose-dependent reduction of food intake and body weight in rats, and altered the expression of hypothalamic genes known to regulate food intake and body weight. Oral administration of a mimetic in a mouse model of high-fat diet-induced obesity reduced body weight gain, adiposity and insulin resistance. Thus, insulin mimetics have a unique advantage over insulin in the control of body weight and hold potential as a novel anti-obesity treatment.
We used two nonpeptidyl compounds (Cpd 1 and Cpd 2) that are capable of mediating insulin-like and sensitizing effects in cells and exerting anti-diabetic effects in rodent models of diabetes8,
9. These compounds function by activating insulin receptor tyrosine kinase and downstream signaling pathways in cells. In previous studies, Cpd 1 and Cpd 2 elicited glucose lowering effects two to six hours following oral gavage, and sustained glucose lowering was observed with chronic dosing in rodent models of diabetes. Pilot studies indicated that Cpd 1 (but not Cpd 2) could be solubilized in aqueous solution suitable for central intracerebroventricular (i.c.v.) injection. Rats fed ad libitum were adapted to a regimen on which they received 30 min access to a 15% (w/v) sucrose solution. Administration of Cpd 1 one hour prior to the sucrose meal did not affect sucrose consumption between 60 and 90 minutes following the injections. In contrast, sucrose consumption was significantly decreased in a dose-dependent manner on the day following the injection (Fig. 1a). Chow intake (Fig. 1b) and body weight (Fig. 1c) over the 24 hours following the i.c.v administration of Cpd 1 were also reduced dose-dependently. These effects were all diminished by the second 24-hour period. Hence, a single i.c.v. injection of Cpd 1 reduced food intake over 24 hours. This time course is similar to that observed with i.c.v. injection of insulin10.
Figure 1. Centrally administered Cpd 1 reduces food intake and body weight.
a, Delayed effect of Cpd 1 on 30 min sucrose consumption 25 h post-injection. Sucrose intake was reduced on the day following i.c.v. Cpd 1. b, 24-hour chow intake. c, 24-hour body weight change. Animals were distributed into weight-matched groups prior to injection of vehicle or Cpd 1. The body weight changes correspond to the same 24-hour period in (b). d, POMC and e, NPY gene expression following repeated i.c.v. vehicle or Cpd 1 injections during a 74-hour fast. Cpd 1 increased POMC and decreased NPY expression relative to vehicle. f, Conditioned taste aversion. Preference for saccharin is depicted as the ratio of saccharin to total fluid consumed. LiCl but not Cpd 1 caused a significant CTA. g, Pica. LiCl, but not Cpd 1, elicited kaolin intake measured for 24 h. h, Na appetite. LiCl but not Cpd 1 suppressed need-induced sodium appetite. Data expressed as percent of vehicle consumption. *, P < 0.05; **, P< 0.01; ***, P < 0.001 versus vehicle group, n = 10−13).
Cpd 1 was administered i.c.v. to fasted animals to determine whether it altered the expression of proopiomelanocortin (POMC) or neuropeptide Y (NPY), two hypothalamic peptides known to regulate food intake and body weight. POMC expression, normally suppressed by food deprivation, increased 183% in fasted, Cpd 1-treated animals compared to those treated with vehicle (Fig. 1d). Neuropeptide Y (NPY) expression, normally elevated in fasted animals, was suppressed by Cpd 1 to 38% of fasted, vehicle-treated animals (Fig. 1e). This effect is larger than that seen following insulin administration10. These data indicate that Cpd 1 reduces food intake and body weight, at least in part, by upregulating the catabolic melanocortin system and downregulating the anabolic NPY system in the hypothalamus.
Several control experiments were completed to determine if doses of Cpd 1 that reduce food intake caused malaise. Conditioned taste aversion (CTA) is commonly used to assess whether a particular substance or treatment renders animals ill11. The animals consumed 75% of their total fluid intake as saccharin in the two-bottle test, indicating that administration of Cpd 1 at a dose that reduced food intake did not result in a CTA (Fig. 1f). In contrast, administration of lithium chloride (LiCl) produced a robust CTA. When rats are ill, they eat dirt or clay, such as kaolin12,
13. This phenomenon, termed geophagia, is a form of pica (the ingestion of nonnutritive substances) and provides an independent assessment of illness. Administration of Cpd 1 (2,363 ng) did not increase kaolin intake relative to the vehicle over 24 hours. LiCl, conversely, significantly increased kaolin consumption (Fig. 1g). Another symptom of malaise in rats is an unwillingness to consume concentrated sodium solutions in the face of an acute sodium deficit14. Consistent with the other indices of illness, LiCl significantly reduced consumption of hypertonic saline whereas i.c.v. Cpd 1 did not (Fig. 1h). Hence, a dose of Cpd 1 that reliably reduces food intake and body weight does not make rats ill by any of three separate indices. These findings provide experimental evidence that reduction in body weight following central administration of Cpd 1 was not mediated via nonspecific toxic effects.
To determine whether these compounds have anti-obesity effects when administered chronically, we utilized an analog of the molecule (Cpd 2) that can be effectively added to food. Cpd 2 is analogous to Cpd 1 in chemical structure and biological function9,
15. Male ICR mice (wild type) were fed standard chow or a high-fat diet (HFD, 36% (w/w), 58.4% calories as fat) with or without Cpd 2 (7−10 mg/kg/d) ad libitum for seven weeks. Mice given the HFD had an accelerated rate of body weight gain (Fig. 2a). The BW of the HFD mice increased 114 5% over the seven weeks compared to a 79 2% increase for the chow group (P < 0.01 and n = 20). Mice fed the HFD combined with Cpd 2 had a significantly reduced rate of BW gain (80 4% and P < 0.01 versus HFD group, n = 20) that was statistically indistinguishable from that in the chow fed mice. There was a trend towards decreased food intake in mice ingesting HFD and Cpd 2 when the data were expressed as grams of food consumed per mouse per day (Fig. 2b). However, this slight suppression in food intake did not reach statistical significance and was unlikely to account for the dramatic reduction in dietary-induced obesity. A full range of studies assessing energy expenditure will be necessary to fully understand the effect of insulin mimetic compounds on energy homeostasis.
Figure 2. Cpd 2 attenuates body weight gain and adiposity in ICR mice fed the HFD.
a, Body weight change. , HFD; , Chow; , HFD/Cpd 2. Each symbol represents mean s.e.m. of n = 20 animals. b, Food intake. Each symbol represents averaged daily food intake per mouse measured every week over a four-week period (triangles and circles represent two different cages). c, Adipose tissue content. Each bar represents mean s.e.m. of n = 6−8 animals. d, Epididymal adipose tissue weight. Each bar represents mean s.e.m. of n = 15 animals. e, Interscapular brown adipose tissue weight. Each bar represents mean s.e.m. of n = 15 animals. f, Morphological characterization of intrascapular brown adipose tissue (IBAT) and epididymal white adipose tissue (EWAT). **, P < 0.01; ***, P < 0.001 versus the HFD mice.
At the termination of the chronic experiment, body composition was determined. ICR mice consuming the HFD had increased total adipose mass relative to the chow group as detected by a dual energy X-ray absorptiometry scan (Fig. 2c). The adipose mass of mice consuming the HFD plus Cpd 2 was intermediate between that of the other two groups and significantly reduced relative to that of the HFD group. Lean body mass was comparable for all three groups (data not shown). The changes in total fat mass occurred in both epididymal white and interscapular brown adipose tissue (Fig. 2d−f). The weight of both fat depots increased by approximately two-fold in the HFD fed mice compared to the chow-fed mice. Treatment with Cpd 2 prevented the increase in the weights of both depots. Histological analysis indicated a significant hypertrophy of fat cells in both adipose depots following HFD feeding. The changes in morphology were most striking in intrascapular brown adipose tissue, where the HFD mice had significantly enlarged brown adipocytes resembling metabolically less-active white adipocytes. Such effects were attenuated in mice treated with HFD in combination with Cpd 2 (Fig. 2f). Because brown adipose tissue plays an important role in thermogenesis and energy expenditure16,
17, the abnormal morphology of brown adipocytes in HFD fed mice could contribute to the development of obesity. The normal-appearing brown adipocytes in HFD/Cpd 2-treated mice could allow more active fuel metabolism, thereby curtailing the development of obesity in these animals. Thus, coadministration of Cpd 2 prevented HFD-induced increase in total adipose tissue content and morphological alterations of brown adipose tissue.
Hyperinsulinemia, insulin resistance and hyperleptinemia are frequently associated with human obesity18,
19, and these features were also manifested in the murine model of diet-induced obesity in the current study. ICR mice fed the HFD had a four-fold increase in plasma insulin, a 22% increase in fasting blood glucose and a three-fold increase in postprandial plasma leptin levels (Table 1). Elevations in these parameters were not apparent in mice fed the HFD in the presence of Cpd 2. HFD mice were also highly glucose intolerant, and the addition of Cpd 2 significantly improved glucose tolerance (data not shown), suggesting that the compound was efficacious in ameliorating insulin resistance associated with obesity. Thus, Cpd 2 prevented development of HFD-induced insulin resistance, hyperinsulinemia, hyperglycemia and hyperleptinemia. Because insulin and leptin levels correlate positively with the degree of adiposity, the normalization of hyperinsulinemia and hyperleptinemia in HFD/Cpd 2 mice probably results from the prevention of diet-induced accumulation of adipose tissue. The prevention of development of obesity should also contribute to the improved insulin sensitivity observed in the treated mice.
Table 1. Cpd 2 reduces indices of obesity and improves insulin sensitivity
In order to evaluate whether chronic treatment of mice with HFD/Cpd 2 could cause nonspecific toxic effects that could complicate the interpretation of the results, plasma markers of hepatic, renal and pancreatic function, and serum electrolytes were also measured. Animals fed the HFD with or without Cpd 2 had similar levels of plasma aminotransferases (AST, ALT), lactate dehydrogenase, creatinine, amylase, creatine kinase (CK) and electrolytes as chow-fed mice (data not shown).
The mechanisms underlying the resistance to obesity conferred by Cpd 2 are presently unknown. Stimulation of the insulin receptor results in multiple actions throughout the body. In the periphery, a well-characterized action of insulin is to promote body weight gain by stimulating energy storage. As such, one of the common side effects of pharmacological administration of insulin to diabetic patients is weight gain.
Therefore, oral administration of an insulin mimetic/sensitizing compound may be anticipated to also lead to weight gain. Contrary to this, however, it had the opposite effect. Although seemingly paradoxical, the effects of the mimetic are consistent with the results of studies characterizing the phenotype of mice lacking the protein tyrosine phosphatase PTP-1B gene20,
21. PTP-1B has been implicated as a major negative regulatory component of insulin action. PTP-1B-deficient mice have increased insulin sensitivity in liver and muscle, a characteristic that would be expected to result in increased weight gain. However, the PTP-1B-deficient mice exhibit increased energy expenditure, decreased adiposity and resistance to diet-induced obesity, a similar phenotype to what occurs when the mimetic is available in food. The common feature is activation of the insulin-signaling pathway in both the brain and the periphery, as shown by genetic manipulation in the PTP-1B studies and by pharmacological activation of insulin receptor.
Insulin functions as an adiposity signal to the brain, and local administration of insulin into the brain reduces food intake and body weight3,
10. This central effect of insulin can be recapitulated by the i.c.v. injection of the insulin mimetic compound. In fact, either central or oral administration of insulin mimetic compounds is effective in creating a net catabolic effect and preventing the development of diet-induced obesity. Because the entry of insulin into the brain is receptor-mediated and saturable22, only a very small proportion of circulating insulin can gain access to brain insulin receptors. Hence, peripherally administered insulin mimetics would have a greater potential than peripherally administered insulin to activate the brain insulin receptor cascade. Direct administration of the mimetic into the brain results in reduced food intake and body weight, confirming that the mimetic has a central site of action. We hypothesize that the combined enhancement of both central and peripheral insulin action by insulin mimetics contributes to the prevention of high-fat-induced insulin resistance and obesity.
Obesity is highly prevalent in developed countries and represents a serious public health issue, but effective treatment for obesity is still lacking. We have shown in this study that activation of central and peripheral insulin signaling with small molecule insulin mimetic agents leads to beneficial effects on the control of body weight, food intake, adiposity and insulin sensitivity. Our data demonstrate the unique advantage of small molecule insulin mimetics over insulin in controlling body weight and provide proof-of-principle for a novel approach for the treatment of obesity and related metabolic disorders.
Methods Experimental procedures. All animal procedures were in accordance to the guidelines of Institutional Animal Care and Use Committee of the University of Cincinnati and Merck Research Laboratories. Compound 1 (Cpd 1) is 2,5-Dihydroxy-3-[7-(3-methyl-but-2-enyl)-1H-indol-3-yl] -6-[2-(1,1-dimethyl-allyl)-1H-indol-3-yl]-[1,4]benzoquinone; compound 2 (Cpd 2) is 2,5-Dihydroxy-3-(1-methyl-1H-indol-3-yl)-6-phenyl-[1,4]benzoquinone.
Central cannulation. Male Long-Evans rats (250−300 g) (Harlan, Indian-apolis, Indiana) were housed individually in tub cages at constant temperature (25 °C) on a light:dark 12:12-h schedule. Stereotaxic implantation of an intraventricular cannula aimed at the third cerebral ventricle (2.2 mm caudal to bregma and 7.5 mm ventral to the sagittal sinus) was done using a previously described method13. Placement was confirmed by i.c.v. injection of angiotensin II (10 ng). Patency was defined as the consumption of 5 ml water within 1 h of injection.
Feeding protocol. Rats were adapted to a schedule on which a 15% sucrose solution in water was available for 30 min at the same time each day. Experiments began once all the rats maintained a stable daily sucrose intake (2 weeks). All i.c.v. injections were given 1 h prior to the sucrose meal. Food (pelleted chow) was removed at the time of injection and returned at the end of the sucrose meal.
Real-time PCR expression analysis. i.c.v. injections (1 l of either vehicle or 750 ng Cpd 1; n = 9 per group) were administered every 12 h for 72 h to food-deprived rats (seven total injections). The rats were killed and brains were collected 2 h following the final injection (hour 74). Total hypothalamic RNA was isolated using Tri-Reagent (MRC, Cincinnati, Ohio), and DNA contamination was removed using DNAfree (Ambion, Austin, Texas). POMC and NPY expression were analyzed in separate reactions using the Taqman (Applied Biosystems, Foster City, California) real-time PCR chemistry and iCycler (Bio-Rad, Hercules, California) detection system (POMC forward 5' -CGCCCGTGTTTCCA-3', reverse 5' -TGACCCATGACGTACTTCC-3', probe 6-FAM-ACGGAGATGAACAGCCCTTGACT-TAMRA; NPY forward 5' -ATCTCTTAATGAGAGAAAGCACA -3', reverse 5' -AGACTGGTTTCACAGGATGA-3' and probe 6-FAM-CCCAGAACAAGGCTTGAAGACC-TAMRA). GAPDH served as the reference gene in each multiplexed reaction (forward 5' -TGCACCACCAACTGCTTAG-3', reverse 5' -GGATGCAGGGATGATGTTC-3' and probe VIC-CAGAAGACTGTGGATGGCCCCTC-TAMRA). The cycle number at which the fluorescence exceeded the threshold of detection (CT) for GAPDH was subtracted from that of the target (POMC or NPY) for each well (CT). The percent change in expression, relative to the vehicle treated group, was defined as (2-CT 100), where CT equals the group CT minus the CT of the vehicle treated group.
Conditioned taste aversion. Rats habituated to restricted water access (1 h/d) received an i.c.v. injection of Cpd 1 or its vehicle following access to a 0.5% (v/v) sodium saccharin solution, which replaced water. Intraperitoneal (i.p.) injection of LiCl or saline (2% BW of a 0.15 M solution) served as positive and negative controls, respectively. 3 d later, animals were offered both water and the saccharin solution in separate bottles during their 1-h drinking period.
Pica. Following i.c.v. injection of Cpd 1 or vehicle, a food hopper containing kaolin pellets (made from Fisher Scientific kaolin powder and water) was added to each cage. Pelleted chow was available in a separate food hopper. i.p. LiCl or saline controls were as above.
Sodium appetite. Rats were habituated to the presence of both 0.5 M NaCl solution and water in separate bottles for 1 wk. The animals were then rendered sodium deficient by two injections of 10 mg/ml furosemide (0.05% BW) given an hour apart and then maintained on sodium-free chow and water for 24 h. On day 2, injections were given as described above (i.c.v. Cpd 1 or its vehicle, or i.p. LiCl or NaCl). The 0.5 M NaCl bottles were returned to the cages 15 min following the injections, and intake was assessed after 120 min.
Dietary-induced obesity model. Male ICR (Institute of Cancer Research) mice were obtained from Taconic Farms (Germantown, NY) and housed 10 per cage in a room maintained at constant temperature (25 °C) on a light:dark 12:12-h schedule. Four-wk-old male ICR mice were fed regular chow or a high-fat diet (HFD, 36% (w/w), 58.4% calories as fat) with or without Cpd 2 (7−10 mg/kg/d) ad libitum for seven weeks. Individual body weights were measured weekly. Averaged daily food intake was measured once a week over a four-week period for two cages per treatment group (10 mice per cage). At the termination of the study, one group of mice was anesthetized, and body composition was assessed by a dual-energy X-ray absorptiometry scan. Another group of mice was dissected, and tissue weights were determined. Adipose tissue was fixed in 10% (v/v) buffered formaldehyde and embedded in paraffin. Sections (8 m) were cut and stained with standard hematoxylin/eosin procedures. Images were captured by using an Optronics DTI750 camera (Goleta, California) at 10 magnification.
Blood and plasma parameters. Blood glucose was measured using a One Touch Glucometer (Lifescan, Milpitas, California). Plasma insulin and leptin concentrations were measured with ELISA kits (Alpco, Windham, New Hampshire).
Calculations. Data are expressed as means s.e.m.. Statistical analysis was conducted using Student's t-test or ANOVA. Statistical significance was defined as P < 0.05.
Kopelman, P.G. Obesity as a medical problem. Nature404, 635–643 (2000). | Article | PubMed | ISI | ChemPort |
Woods, S.C., Lotter, E.C., McKay, L.D. & Porte, D. Jr. Chronic intracerebroventricular infusion of insulin reduces food intake and body weight of baboons. Nature282, 503–505 (1979). | PubMed | ISI | ChemPort |
Schwartz, M.W., Figlewicz, D.P., Baskin, D.G., Woods, S.C. & Porte, D. Jr. Insulin in the brain: a hormonal regulator of energy balance. Endocr. Rev.13, 387–414 (1992). | Article | PubMed | ISI | ChemPort |
Woods, S.C., Schwartz, M.W., Baskin, D.G. & Seeley, R.J. Food intake and the regulation of body weight. Ann. Rev. Psychol.51, 255–277 (2000). | Article | ISI | ChemPort |
Bruning, J.C. et al. Role of brain insulin receptor in control of body weight and reproduction. Science289, 2122–2125 (2000). | Article | PubMed | ISI | ChemPort |
DeFronzo, R. Pharmacological therapy for type 2 diabetes. Ann. Intern. Med .17, 281–303 (1999).
Zhang, B. et al. Discovery of a small molecule insulin mimetic with antidiabetic activity in mice. Science284, 974–977 (1999). | Article | PubMed | ISI | ChemPort |
Qureshi, S.A. et al. Activation of insulin signal transduction pathway and anti-diabetic activity of small molecule insulin receptor activators. J. Biol. Chem.275, 36590–36595 (2000). | Article | PubMed | ISI | ChemPort |
Sipols, A.J., Baskin, D.G. & Schwartz, M.W. Effect of intracerebroventricular insulin infusion on diabetic hyperphagia and hypothalamic neuropeptide gene expression. Diabetes44, 147–151 (1995). | PubMed | ISI | ChemPort |
Bernstein, I.L., Goehler, L.E. & Bouton, M.E. Relative potency of foods and drinks as targets in aversion conditioning. Behav. Neural. Biol.37, 134–148 (1983). | PubMed | ISI | ChemPort |
Mitchell, D. et al. Poison induced pica in rats. Physiol. Behav.17, 691–697 (1976). | Article | PubMed | ISI | ChemPort |
Stricker, E.M. & Verbalis, J.G. Central inhibitory control of sodium appetite in rats: correlation with pituitary oxytocin secretion. Behav. Neurosci.101, 560–567 (1987). | Article | PubMed | ISI | ChemPort |
Liu, K. et al. Discovery of a potent, highly selective, and orally efficacious small-molecule activator of the insulin receptor. J. Med. Chem.43, 3487–3494 (2000). | Article | PubMed | ISI | ChemPort |
Lowell, B.B. et al. Development of obesity in transgenic mice after genetic ablation of brown adipose tissue. Nature366, 740–742 (1993). | Article | PubMed | ISI | ChemPort |
Spiegelman, B.M. & Flier, J.S. Adipogenesis and obesity: rounding out the big picture. Cell87, 377–389 (1996). | Article | PubMed | ISI | ChemPort |
Considine, R.V. et al. Serum immunoreactive-leptin concentrations in normal-weight and obese humans. N. Engl. J. Med.334, 292–295 (1996). | Article | PubMed | ISI | ChemPort |
Elchebly, M. et al. Increased insulin sensitivity and obesity resistance in mice lacking the protein tyrosine phosphatase-1B gene. Science283, 1544–1548 (1999). | Article | PubMed | ISI | ChemPort |
Klaman, L.D. et al. Increased energy expenditure, decreased adiposity, and tissue-specific insulin sensitivity in protein-tyrosine phosphatase 1B-deficient mice. Mol. Cell. Biol.20, 5479–5489 (2000). | Article | PubMed | ISI | ChemPort |
Baura, G.D. et al. Saturable transport of insulin from plasma into the central nervous system of dogs in vivo. A mechanism for regulated insulin delivery to the brain. J. Clin. Invest.92, 1824–1830 (1991).
Acknowledgments We thank D. Szalkowski and Z. Li for help with necropsy and J. Ronan for help with preparation of histology slides. This work was supported in part by grants from the DFG (M.Z.S.), the American Diabetes Association (Physician Scientist Training Award to E.L.A.) and NIH (S.C.W.).
5% over the seven weeks compared to a 79 
, HFD; 
, Chow; 
, HFD/Cpd 2. Each symbol represents mean 
2 weeks). All i.c.v. injections were given 1 h prior to the sucrose meal. Food (pelleted chow) was removed at the time of injection and returned at the end of the sucrose meal.
l of either vehicle or 750 ng Cpd 1; n = 9 per group) were administered every 12 h for 72 h to food-deprived rats (seven total injections). The rats were killed and brains were collected 2 h following the final injection (hour 74). Total hypothalamic RNA was isolated using Tri-Reagent (MRC, Cincinnati, Ohio), and DNA contamination was removed using DNAfree (Ambion, Austin, Texas). POMC and NPY expression were analyzed in separate reactions using the Taqman (Applied Biosystems, Foster City, California) real-time PCR chemistry and iCycler (Bio-Rad, Hercules, California) detection system (POMC forward 5' -CGCCCGTGTTTCCA-3', reverse 5' -TGACCCATGACGTACTTCC-3', probe 6-FAM-ACGGAGATGAACAGCCCTTGACT-TAMRA; NPY forward 5' -ATCTCTTAATGAGAGAAAGCACA -3', reverse 5' -AGACTGGTTTCACAGGATGA-3' and probe 6-FAM-CCCAGAACAAGGCTTGAAGACC-TAMRA). GAPDH served as the reference gene in each multiplexed reaction (forward 5' -TGCACCACCAACTGCTTAG-3', reverse 5' -GGATGCAGGGATGATGTTC-3' and probe VIC-CAGAAGACTGTGGATGGCCCCTC-TAMRA). The cycle number at which the fluorescence exceeded the threshold of detection (CT) for GAPDH was subtracted from that of the target (POMC or NPY) for each well (
CT). The percent change in expression, relative to the vehicle treated group, was defined as (2-
100), where 
