Chronic Central Melanocortin-4 Receptor Antagonism and Central Neuropeptide-Y Infusion in Rats Produce Increased Adiposity by Divergent Pathways

All procedures were approved by the Animal Experimentation Ethics Committee of the Garvan Institute/St Vincent’s Hospital and are in keeping with the National Health and Medical Research Council of Australia’s guidelines on animal experimentation. Male Wistar rats were purchased from Animal Resources Center (Perth, Australia) at weights of 250–280 g. They were housed in groups in plastic cages on pellet paper bedding and under conditions of controlled temperature (23°C) and illumination (0600–1800). They were allowed ad libitum access to standard laboratory diet (Norco Stockfeeds, South Lismore, Australia) and water, unless otherwise stated.

Animals were anesthetized with an intraperitoneal injection of ketamine and xylazine at 60 and 10 mg/kg, respectively (Mavlab, Brisbane, Australia, and Troy Laboratories, Sydney, Australia, respectively), for the placement of a cannula in the right lateral cerebral ventricle (11), which was used for all ICV administrations. Another cannula was placed in the right jugular vein for blood sampling. Rats received an intramuscular injection of penicillin at 0.3 mg/kg (Troy Laboratories) as well as a subcutaneous analgesic, buprenorphine (Temgesic 0.025 mg/kg) (Reckitt and Coleman, Hull, U.K.), and were left to recover to presurgery weights (7–10 days) in individual cages, with daily handling to minimize stress.

A subset of rats were used in acute (4-h) feeding studies after bolus ICV injection of porcine NPY (1.2 nmol) (Auspep, Melbourne, Australia), synthetic MC4 receptor antagonist HS014 (1.5 nmol) (Auspep), or vehicle (5 μl of 0.9% NaCl) for control rats. Injections were made over a period of 60 s, animals were then returned to their cages, and total ad libitum food intake was measured after 4 h.

Rats were anesthetized with halothane (Veterinary Companies of Australia, Sydney, Australia) for implantation of subcutaneous osmotic minipumps (model 2001; Alza, Palo Alto, CA) (11), which were connected via polyethylene tubing to the ICV cannula for central infusion of porcine NPY (3.5 nmol/day), HS014 (4.8 nmol/day), or vehicle (24 μl/day of 0.9% NaCl) for control rats. Food intake of the NPY- and HS014-infused rats was restricted to that of the vehicle-infused animals (30–32 g/day) to ensure that any differences observed were not caused by increased food intake. Rats were fasted for 2–3 h (from ∼0930 to 1230). Basal blood samples (0.3 ml) were collected daily from the jugular cannula between 1130 and 1230 each day using sodium citrate (6.1 mg/ml in 0.9% NaCl) as an anticoagulant. Then, 3–4 days after minipump implantation, an intravenous glucose tolerance test was performed where blood samples were taken 1 and 5 min after intravenous glucose injection (300 mg/kg). These two time points were chosen because we have previously shown that insulinemia attains a peak at 1 min after such glucose injection and returns toward baseline values within 5 min (31). Incremental areas under the resultant insulin curves were calculated (by subtracting baseline values) between 0 and 5 min after glucose injection. All plasma samples were frozen in liquid N₂ and stored at −20°C until analysis. Plasma concentrations of triglyceride (TG) (Sigma Diagnostics, St. Louis, MO), nonesterified fatty acid (Wako, Osaka, Japan), and glucose (glucose oxidase method; Trace Scientific, Melbourne, Australia) were measured using commercial kits. Plasma corticosterone and leptin, basal, and glucose-stimulated insulin levels were measured by radioimmunoassay kits (ICN Biomedicals, Costa Mesa, CA, and Linco Research, St. Louis, MO, respectively).

After 5–6 days of minipump infusion, rats were anesthetized with halothane, the interscapular BAT and liver were then removed and weighed, and portions were freeze-clamped and stored at −80°C until analysis. Epididymal and retroperitoneal fat pads were then excised and weighed.

Total RNA was extracted from BAT samples using TRI-Reagent (Sigma) according to the manufacturer’s instructions. The mouse uncoupling protein-1 (UCP-1) probe was generated by polymerase chain reaction from genomic DNA. The 182-bp long fragment corresponding to exon 3 was labeled with [α-³²P]dCTP, using a megaprime DNA labeling kit (Amersham, Buckinghamshire, England), and Northern blot analysis was carried out under standard conditions. Blots were washed under stringent conditions before autoradiographic exposure to Kodak BioMax film at −80°C for 2–3 days. As a control for loading, the same blot was hybridized with glyceraldehyde 3-phosphate dehydrogenase (GAPDH) probe, and all ratios of UCP-1–to–GAPDH optical density were expressed as a percentage of control values. All films were analyzed with a computer-assisted image analysis system connected to a GS-690 Imaging Densitometer (Bio-Rad, Hercules, CA).

Liver samples (50 mg) were homogenized in a glass hand-held homogenizer in a 4-ml chloroform:methanol (2:1) solution, and lipids were left to extract into the organic phase for 12–16 h at room temperature. Two milliliters of 0.6% NaCl solution was added, and then the tubes were vortexed and centrifuged for 10 min at 4°C (825g). The lower organic phase containing lipids was transferred to glass scintillation vials and dried under nitrogen gas. The dried lipids were then resuspended in 500 μl ethanol and stored at −20°C until a TG assay was performed (Peridochrome TG kit; Roche Diagnostics).

Statistics were performed using Statview version 4.5 (Abacus Concepts). Repeated measures or factorial analysis of variance tests were used for the statistical analysis of results, with Fisher’s post hoc tests where appropriate. Data are expressed as means ± SD, and the results were considered statistically significant at P < 0.05.

To standardize the biological potency of NPY and HS014, we chose doses of these compounds that produce similar and maximal degrees of hyperphagia in ad libitum–fed rats. After acute ICV injection of a maximum dose of NPY (1.2 nmol) (32), rats consumed sevenfold as much diet as vehicle-injected rats in the ensuing 4 h (Fig. 1). A similar 4-h food intake was obtained after ICV injection of 1.5 nmol of HS014 (Fig. 1), and higher doses did not produce greater hyperphagia (data not shown). For chronic infusion studies, three times this maximally effective acute dose was infused over each 24-h period as previously described for NPY (11). Because all animals were fed the same amount as vehicle-infused controls during chronic infusions, latency to feed was used as an index of biological potency of NPY and HS014. In both NPY- and HS014-infued rats, latency to feed was <30 s on each of days 2–6 of infusion compared with >60 s in vehicle-infused controls (data not shown). If the latency to feed was >30 s, animals were excluded from the NPY- or HS014-infused groups (<10% of rats).

There was no significant difference among NPY-, HS014-, and saline-infused rats with respect to body weight before infusion (NPY 354 ± 8, HS014 372 ± 7, and control 367 ± 6 g, n = 9–16, NS) or with respect to body weight gain during the 6 days of infusion (Fig. 2A). This is consistent with data showing that whereas chronic ICV NPY or MC4 receptor antagonist infusion significantly increases body weight in ad libitum–fed rats (8,24,25,33), no such increase is observed when hyperphagia is prevented by pair-feeding with vehicle-infused controls (11,33,34). However, 6 days of central infusion of NPY or HS014 with pair-feeding to control rats significantly increased fat accumulation in the epididymal and retroperitoneal white adipose tissue (WAT) depots compared with saline-infused controls (Fig. 2B). Furthermore, the retroperitoneal fat depot weight was significantly greater in HS014-infused than in NPY-infused rats (Fig. 2B). The 6 days of ICV infusion of both NPY and HS014 resulted in significant increases in plasma leptin levels compared with controls, but the effect was significantly greater in NPY-infused rats than in rats infused with HS014 (Fig. 2C).

The plasma concentrations of corticosterone during the 6 days of ICV infusion are shown in Fig. 3A. Compared with the vehicle-infused control rats, pair-fed NPY-infused rats had an increase in plasma corticosterone levels during the first 2 days of infusion. In contrast, long-term ICV infusion of the MC4 receptor antagonist HS014 with pair-feeding to control rats had no significant effect on corticosteronemia at any time point. NPY-infused rats exhibited significantly greater plasma insulin levels compared with control and HS014-infused rats over the 6 days of central infusion (Fig. 3B). Rats chronically ICV-infused with HS014 exhibited no significant difference from control values in plasma insulin levels over the 6 days of infusion (Fig. 3B). Levels of glycemia were not significantly changed from control values in ICV NPY- or HS014-infused rats (Fig. 3C).

Immediately before glucose injection, basal insulinemia was greater than control values in NPY- but not HS014-infused rats (Fig. 4A). Intravenous glucose-induced plasma insulin levels were more elevated than control values for both NPY- and HS014-infused rats at 1 and 5 min postglucose injection (Fig. 4A), and the difference was significant at 5 min. The incremental areas under the insulin curves shown in Fig. 4 are NPY 3,960 ± 440, P = 0.08 vs. control; HS014 5,130 ± 330, P < 0.001 vs. control; and control 2,750 ± 470 pmol/l × 5 min, with no significant difference between NPY- and HS014-infused rats. Plasma glucose levels of NPY- and HS014-infused rats were no different from control values either before or at any time after glucose injection (Fig. 4B).

The plasma concentrations of free fatty acids (FFAs) and TGs are shown in Fig. 5. Plasma FFA levels in NPY-infused rats were significantly reduced compared with saline and HS014-infused rats for the first 3 days of ICV infusion (Fig. 5A). Plasma FFA levels in HS014-infused rats were not significantly different from control levels for the entire 6 days of infusion (Fig. 5A). Plasma TG levels of NPY-infused rats were significantly elevated compared with control and HS014-infused rats on days 1–5 of central infusion (Fig. 5B). Plasma TG levels of HS014-infused rats were not significantly different relative to controls during the infusion period (Fig. 5B).

Hepatic TG content was significantly increased over control values in HS014- but not NPY-infused rats (Fig. 5C). There was no significant difference among the three groups with respect to liver weight after 5–6 days of infusion (NPY 13.2 ± 0.3, HS014 12.6 ± 0.5, and control 12.0 ± 0.3 g, n = 4–8, NS).

BAT weight was significantly increased over control values after 5–6 days of ICV NPY or HS014 infusion (NPY 0.36 ± 0.05, HS014 0.33 ± 0.02, and control 0.16 ± 0.02 g, n = 5–6, both P < 0.01 vs. control). Both NPY- and HS014-infused rats exhibited a significantly lower level of UCP-1 mRNA in the BAT compared with that of saline-infused controls, with no significant difference between the NPY- and HS014-infused groups (Fig. 6).

This work investigated whether aspects of the obesity syndromes induced by chronically increased NPY-ergic activity or central MC4 receptor antagonism are mediated by independent or common pathways. To this end, we compared the time course of hormono-metabolic consequences of chronic ICV infusion of NPY and HS014 (an analogue of the endogenous MC3/MC4 receptor antagonist AgRP) in rats, where hyperphagia had been prevented by pair-feeding.

It was demonstrated that 6 days of central infusion of either NPY or HS014 in rats resulted in significant increases in WAT mass, even though NPY- and HS014-infused rats were prevented from overeating by pair-feeding with vehicle-infused control rats. As with another very recent study using 7-day ICV AgRP injections (33), we demonstrated that MC4 receptor antagonist-infused rats show an increase in WAT mass in the absence of increased food intake. Our study also extends this observation by making a direct comparison of the effects of central HS014 infusion with those of NPY and by time course comparison of endocrine and metabolic effects. Because ICV HS014 induced at least as great an increase in fat mass as ICV NPY and yet had divergent hormonal and metabolic effects, we conclude that MC4 receptor antagonism does not induce obesity solely by regulation of the endogenous NPY-ergic system.

Because plasma concentrations of leptin are positively correlated with percent body fat (35), increased plasma leptin levels in NPY- and HS014-infused rats probably reflected the increased adiposity. However, plasma leptin levels of rats infused with the MC4 receptor antagonist HS014 were significantly lower than those of NPY-infused rats, even though adiposity in the HS014 group was at least as great. A similar outcome was recently reported in comparison of effects of ICV SHU9119 and NPY in ad libitum–fed rats (25). This highlights the fact that leptinemia is regulated by many factors besides fat mass (35). Corticosterone and insulin, plasma levels of which were shown to be increased in NPY- but not HS014-infused rats, are both known to positively regulate plasma leptin levels (35). Additionally, other researchers have shown that plasma testosterone levels are significantly reduced by NPY but not by SHU9119 infusion (12,25), and because testosterone is an inhibitor of leptin secretion (35), this difference may also contribute to the lower plasma leptin levels of the HS014 group.

It has recently been reported that chronic central infusion of the MC3/MC4 receptor antagonist SHU9119 increased plasma corticosterone levels in ad libitum–fed rats (25). This effect was probably mediated by SHU9119-induced hyperphagia because feeding is known to increase corticosteronemia (36), and because in the present study MC4 receptor antagonist infusion with concomitant pair-feeding caused no such hypercorticosteronemia. Our finding is in keeping with reports that animal models with chronic MC4 receptor antagonism, such as MC4 receptor knockout mice and agouti obese yellow mice, do not have elevated basal plasma corticosterone levels (17).

High plasma levels of insulin are known to contribute to obesity because attenuation of the hyperinsulinemia of genetically obese rodents (37,38) reduced the obese phenotype in the absence of significant effects on feeding. Furthermore, administration of exogenous insulin to animals and humans results in increased fat accumulation and other metabolic defects characteristic of obesity (39,40,41). NPY-infused rats exhibited two- to threefold increases over controls in basal plasma insulin levels over the infusion period. This hyperinsulinemia, in addition to reduced thermogenesis (9,42,43,44,45), is the likely primary mechanism by which NPY-infused rats become obese in the absence of hyperphagia. In contrast to the NPY-infused group, rats that received HS014 infusions did not exhibit any significant change in insulinemia relative to control rats, indicating that hyperinsulinemia may not be the only contributor to the observed increase in adiposity. Glucose-stimulated plasma insulin concentrations were elevated in both NPY- and HS014-infused rats in the absence of any difference from controls in glycemia, an indication of insulin resistance. This increased insulin response to glucose in both obesity models may be mediated by insulin resistance, which is commonly observed in situations of increased fat mass, as in ICV NPY-infused rats (11). Therefore, the late-onset hyperinsulinemia observed in MC4 receptor knockout and agouti obese yellow mice is unlikely to result from a primary defect in insulin secretion, and hyperinsulinemia probably develops secondary to insulin resistance.

In addition to the hormonal differences discussed above, differences in metabolic profiles were also observed between ICV NPY- and MC4 receptor antagonist-induced increases in adiposity, suggesting independent etiologies. Whereas central NPY infusion significantly reduced plasma FFA and increased plasma TG levels (possibly mediated by hyperinsulinemia and/or reduced sympathetic nervous activity), the plasma concentrations of FFAs and TGs in HS014-infused rats were not significantly different from controls, reflective of the unchanged plasma insulin profile. Unlike NPY-infused rats, the total liver TG content observed in rats infused with HS014 was increased relative to controls, despite no change in tissue weight.

It is well established that central NPY administration decreases sympathetically mediated thermogenic activity in BAT (9,42,43,44,45), most likely by decreasing the expression of UCP-1 (9). This effect is likely to contribute to NPY-induced obesity because BAT ablation per se also results in an obesity syndrome (46). This study has shown that UCP-1 mRNA levels are also decreased in BAT after 6 days of central MC4 receptor antagonism with HS014, extending a recent report that 7-day ICV AgRP injection with pair-feeding decreased the levels of UCP-1 protein in BAT (33). Furthermore, the extent of BAT UCP-1 mRNA downregulation in HS014-infused rats was at least as great as that induced by 6-day ICV NPY infusion. Recently, it was reported that ICV NPY infusion or antagonism of central MC3 and MC4 receptors with AgRP could reduce the plasma levels of thyroid-stimulating hormone and T4 (33,47,48). Reduced thyroid function could reduce whole-body energy expenditure and contribute to increased energy retention as fat.

The heterogeneity of hormonal and metabolic consequences of central NPY infusion or MC4 receptor antagonism shown in this study are supported by a recent comparison of brain sites of c-Fos activation in response to NPY or the MC3/MC4 receptor antagonist AgRP-(83–132) (49). c-Fos immunoreactivity was significantly increased over control values in the hypothalamic PVN of both NPY- and AgRP-injected rats, measured at 2 h after ICV injection. In contrast, only NPY significantly increased c-Fos activation in the NTS at this time point, although AgRP did produce a lesser degree of c-Fos activation in this site at 24 h after injection. Only AgRP resulted in significantly increased c-Fos immunoreactivity in the accumbens shell and the lateral septum (49). The PVN is implicated as a site regulating BAT thermogenesis (9) and activity of the thyrotropic axis (48), whereas the NTS is implicated in mediation of NPY-induced insulin secretion (50). Collectively, these data fit the hypothesis that both NPY and MC receptor antagonism reduce thyroid function (47,48) and BAT UCP-1 expression by effects on the PVN, yet only NPY has significant effects on basal hyperinsulinemia, mediated by neuronal activation in the NTS.

In conclusion, the differences in hormonal and metabolic perturbations that develop during the early stages of ICV NPY and HS014 infusions, using doses that produce similar increases in adiposity, suggest that NPY and MC4 receptor antagonism induce increases in fat mass via at least partially divergent pathways. In the NPY obesity model, hypercorticosteronemia and hyperinsulinemia are early onset factors. In contrast, the increase in adiposity induced by chronic central MC4 receptor antagonism does not appear to involve hypercorticosteronemia or basal hyperinsulinemia. However, it is likely that the etiology of both obesity models involves increased glucose-induced insulin secretion, possibly secondary to increased fat mass and insulin resistance, and downregulation of BAT UCP-1 mRNA expression, which may be mediated by changes in thyroid function as recently reported (33,47,48).

The National Health and Medical Research Council of Australia has supported this work via a Center Grant to the Garvan Institute and a postdoctoral fellowship awarded to A. Sainsbury (Peter Doherty Fellowship 987122).

We thank Dr. Herbert Herzog, Marjorie Liu, Michelle Couzens, and Elizabeth Smith (Garvan Institute, Sydney, Australia) for provision of the UCP-1 and GAPDH probes and help with Northern analysis. We thank Dr. Julie Ferguson and Dr. Jenny Kingham for their invaluable veterinary advice and the staff of the Garvan Institute Biological Testing Facility.