Disruption of melanocortin (MC) signaling, such as by ectopic Agouti overexpression, leads to an obesity syndrome with hyperphagia, obesity, and accelerated body weight gain during high-fat diet. To investigate where in the brain disruption of MC signaling results in obesity, long-term Agouti expression was induced after local injections of recombinant adeno-associated viral particles in selected brain nuclei of adult rats. Agouti expression in the paraventricular nucleus, a hypothalamic region with a high density of MC receptors, induced acute onset hyperphagia and rapid weight gain that persisted for at least 6 weeks. In contrast, obesity and hyperphagia developed with a 3 week delay when Agouti was expressed in the dorsal medial hypothalamus. Agouti expression in the lateral hypothalamus (LH) did not affect food intake and body weight during regular diet, despite the presence of MC receptors in this region. However, during exposure to a high-fat diet, animals with Agouti expression in the LH exhibited a marked increase in body weight. Here we show that the LH is important for the protection against diet-induced obesity by controlling caloric intake during consumption of a high-fat diet. Together, this study provides evidence that different aspects of the Agouti-induced obesity syndrome, such as hyperphagia and diet responsiveness, are mediated by distinct brain regions and opens challenging opportunities for further understanding of pathophysiological processes in the development of the obesity syndrome.
- lateral hypothalamus
- dorsal medial hypothalamus
- paraventricular nucleus
- high-fat diet
Genetic studies have shown that the yellow coat color and the obesity syndrome in rodents with ectopic Agouti overexpression are regulated independently by Agouti antagonism of the melanocyte-stimulating hormone receptor-1 (MC1-R) and of the MC4-R, respectively (Duhl et al., 1994; Lu et al., 1994; Fan et al., 1997; Huszar et al., 1997). Although Agouti antagonism of the MC1-R in the skin causes a yellow fur by switching eumelanin into phaeomelanin pigment synthesis (Ollmann et al., 1998), it is not fully understood where and how Agouti expression generates the obesity syndrome. Because pharmacological blockade or genetic disruption of the MC4-R lead to hyperphagia and obesity (Huszar et al., 1997; Kim et al., 2000; Adage et al., 2001) and the MC4-R is expressed in distinct brain areas (Gantz et al., 1993; Mountjoy et al., 1994; Kishi et al., 2003), it is hypothesized that central Agouti expression accounts for hyperphagia and subsequent body weight gain within the yellow obese syndrome.
Although the melanocortin signaling pathway is clearly involved in body weight control by affecting food intake (Fan et al., 1997; Huszar et al., 1997) and diet responsiveness (Butler et al., 2001), it is poorly understood whether one or different brain sites mediate those MC effects. Other investigators have used intranuclear injection of melanocortin ligands to study these questions (Kim et al., 2000; Wirth et al., 2001); however, this approach is limited by the relatively short time of action of these compounds in tissues, which does not allow measurement of obesity development over weeks or months. To study the relationship between MC pathway involvement in the obesity syndrome and the functional anatomy of the MC system, we locally interfered with melanocortin receptor signaling using vector directed gene expression technology.
The paraventricular nucleus of the hypothalamus (PVN) contains a high density of MC4-Rs that have been proposed to mediate MC-induced changes in food intake. For example, single injections of MC4-R ligands in the PVN alter food intake and body weight in rodents (Kim et al., 2000; Wirth et al., 2001). Recent studies suggest that the dorsal medial hypothalamus (DMH) is also involved in the regulation of hyperphagia (Chen et al., 2004). In addition, based on the outcome of classical lesion, electric stimulation, and electrophysiological studies, the lateral hypothalamus (LH) has long been implicated in the regulation of energy metabolism (Anand and Brobeck, 1951; Delgado and Anand, 1953; Katafuchi et al., 1985). This idea is corroborated by the recent identification of LH neurons that express neuropeptides involved in eating behavior, such as hypocretins/orexins and the melanin-concentrating hormone (MCH) (Zamir et al., 1986; de Lecea et al., 1998; Sakurai et al., 1998). PVN, DMH, and LH neurons express moderate to high levels of MC4-R (Gantz et al., 1993; Mountjoy et al., 1994; Kishi et al., 2003). As a first step toward understanding where in the brain disruption of MC signaling results in obesity, rats with PVN, DMH, or LH injections of either recombinant adeno-associated virus (rAAV)-Agouti or rAAV-enhanced green fluorescent protein (EGFP) were monitored when exposed to a regular or a high-fat diet.
Materials and Methods
Recombinant adeno-associated viral vector production. Construction of pTR-cytomegalovirus (CMV)-EGFP has been described previously (Eaton et al., 2002). Agouti was PCR amplified using as template a plasmid containing the mouse Agouti cDNA (paE65; a kind gift from Roger D. Cone, Oregon Health and Science University, Portland, OR) using oligos 5′aagcttgagatctgccgcaccatggatgtcacccg and 5′gaagaagctagctcagcagttggggttg. A (BglII and NheI digested) fragment containing the Agouti cDNA was cloned into BamHI and SpeI digested pTR-CMV-EGFP (this removes the EGFP cDNA; correct sequence of Agouti was confirmed by sequence analysis), which generated pTR-CMV-Agouti. Next, rAAV was generated via a two-component, adenovirus-free packaging system using the helper plasmid pDG [kindly provided by J. Kleinschmidt, Deutsches Krebsforschungszentrum, Heidelberg, Germany (Grimm et al., 1998)].
Recombinant AAV particles were produced by the method described previously by Hermens et al. (1999). Briefly, the vector plasmid pTR-CMV-Agouti or pTR-CMV-EGFP and the helper plasmid pDG were cotransfected into human embryonic kidney (HEK) 293T cells using calcium phosphate precipitation (molar plasmid ratio, 1:1). The medium was replaced after 6 hr by fresh DMEM containing 10% FCS, and the cells were incubated for 48 hr at 37°C and 5% CO2. Next, the cells were dislodged, harvested, and freeze-thawed three times to release the AAV-particles from the cells. Cell debris was removed using low-speed centrifugation. The supernatant was loaded on a Matrex Cellufine sulfate bead column (Amicon, Danvers, MA). After several washings with PBS, the virus was eluted from the column with PBS containing 1 m NaCl. Next, viral particles were banded on an iodixanol (Nycomed Pharm, Oslo, Norway) density gradient using ultracentrifugation, and fractions of ∼300 μl were collected from the bottom of the gradient. To reduce viscosity of the iodixanol, rAAV-fractions were diluted 10 times with PBS and reconcentrated on a Centricon-100 concentrator (Amicon). The rAAV stocks contained 2 × 1012 particles/ml for rAAV-Agouti-woodchuck posttranslational regulatory element (WPRE) and 2 × 1011 particles/ml for rAAV-EGFP-WPRE.
In vitro infection, Western blot analysis, and MC receptor activation assay. HEK 293 cells were infected (multiplicity of infection, 10) with rAAV-Agouti or rAAV-EGFP. Four days after infection, supernatant was collected. Aliquots (20 μl) of the supernatant were diluted 1:1 with tricine sample buffer and separated on a 12% Tris/tricine gel. After electrophoresis, proteins were transferred to a nitrocellulose membrane and blocked by incubation for 1 hr at room temperature with 10% nonfat milk in Tris-buffered saline/0.1% Tween 20 (TBST). The membrane was incubated overnight with rabbit anti-Agouti antibody as the primary antibody (1:5000 dilution), followed by goat anti-rabbit peroxidase (1: 10,000 dilution) (Jackson ImmunoResearch West Grove, PA) for 60 min after washing with TBST. After washing with TBST, HRP-labeled antibodies were detected by chemiluminescence. Agouti- or EGFP-expressing HEK 293 cells were mixed at a 1:1 ratio with HEK 293 cells cotransfected with 100 ng of an MC4 receptor expression vector (Oosterom et al., 1999) and 7 μg of Cre-lacZ (Chen et al., 1995). Two days after transfection, melanocortin-induced (0.01-100 nm MT-II)β-galactosidase activity was quantified using a colorimetric assay (Chen et al., 1995). EC50 values of MC-induced β-galactosidase activity in the presence of Agouti- or EGFP-producing cells were determined by fitting the data to a sigmoidal curve with variable slope using GraphPad Prism 2.01 for Windows 95/NT (GraphPad Software, San Diego, CA).
Animals. Male outbred rats [strain Wistar (U:WU)] (body weights between 240 and 260 gm) were maintained in a 12 hr light/dark cycle (ambient temperature was 21.0 ± 1.0°C). Food (complete laboratory chow; Hope Farms, Woerden, The Netherlands) and water were available ad libitum. After 10 d of baseline recordings (body weight and food intake), rats were anesthetized with Hypnorm (0.1 ml/100 gm body weight, i.p.; Janssen Pharmaceutica, Beersse, Belgium) and Dormicum (0.05 ml/100 gm body weight, i.m., Hoffman-LaRoche, Mijdrecht, The Netherlands). Stereotaxic bilateral injections of rAAV-Agouti and rAAV-EGFP (as a control) were performed in the PVN [-1.8 mm anteroposterior (AP), 0.3 mm mediolateral (ML), and -8.6 mm dorsoventral (DV), using bregma as a reference for AP and ML coordinates and the skull as a reference for DV coordinates], DMH (-2.5 mm AP, 0.3 mm ML, and 8.4 mm DV), LH (-2.56 mm AP, 1.55 mm ML, and 9.32 mm DV), and 0.76 mm anterior of the PVN region (-1.04 mm AP, 0.3 and -0.3 mm ML, and -8.6 mm DV). During 5 min, 1 μl of virus (2 × 108 particles) was injected per site. After each injection, the needle remained stationary for an additional 5 min and was then removed.
After surgery, food intake and body weight were monitored at least every third day for no less than 42 d. Animals were then provided with a high-energy diet (HED) with a high-fat content for 10 consecutive days (energy content of diets, 3731.9 kcal/kg in regular chow and 4655 kcal/kg in the HED; percentage energy derived from carbohydrate/protein/fat in diets, 63/23/14 in regular chow and 19/19/62 in the HED). After these 10 d, all animals were anesthetized with sodium pentobarbital and perfused with saline, followed by perfusion with 4% paraformaldehyde (PFA). Brains were postfixed overnight with 4% PFA and cryoprotected in a 25% sucrose solution (overnight). Brains were than frozen in cold isopentane (-30°C for 20 sec). Cryostat sections (20 μm) were used for Agouti immunohistochemistry and for in situ hybridization.
Agouti antibody staining. After pretreatment with 4% fetal calf serum and 0.3% H2O2 in PBS, sections were incubated with a rabbit anti-Agouti antibody (1:5000, overnight at 4°C). After a 1 hr incubation with a secondary goat anti-rabbit-biotinylated antibody (1:100) (Jackson ImmunoResearch), sections were incubated with ABC (1:500) (Vector Laboratories, Burlingame, CA) for 1 hr at room temperature. Sections were then treated with diaminobenzidine (1:100) (Sigma, St. Louis, MO) in PBS with 30% H2O2 in PBS for 10 min. All immunohistochemistry steps described above were followed by a three times rinse with PBS of at least 5 min per rinse. Slides were dehydrated in serial ethanol solutions, cleared with xylene, and coverslipped. For histological confirmation of the injection site, Agouti immunoreactivity in the injection sites was confirmed in every other section throughout the hypothalamus for all animals.
In situ hybridization. Pretreated 20 μm cryostat sections from rat hypothalamus were hybridized with 35S-labeled antisense mRNA probes for Agouti-related protein (AgRP), NPY, MCH, and orexins according to van der Kraan et al. (1998). mRNA expression in the arcuate nucleus (for NPY and AgRP) and lateral hypothalamus (for MCH and hypocretins/orexins) was quantified using MCID-M5 (Imaging Research, St. Catharines, Ontario, Canada). mRNA levels are expressed in counts per minute as calculated from a standard curve of diluted probe mix on the same film on which the slides were measured. From each animal, two measurements per probe were taken in the region of interest (arcuate nucleus and lateral hypothalamus) and subsequently averaged to calculate the mean counts per minute per probe for that region.
Statistics. All data are expressed as mean ± SEM. Differences in food intake, body weight, and mRNA expression levels were assessed using one-way and repeated-measure ANOVA, unless indicated differently in the text. In the presence of a significant main effect, the analysis was followed by Tukey's (SPSS for Windows, version 9.0; SPSS, Chicago, IL) contrasts (α = 0.05).
Ethical commission. Animal research and care was approved by the Ethical Commission on Laboratory Animal Experiments of the School of Medicine, University of Utrecht (Utrecht, The Netherlands).
An rAAV-Agouti vector under the control of a CMV promotor was generated and tested for efficacy to release the Agouti protein after infection. In contrast to the media of HEK 293 cells infected with the rAAV-EGFP control particles, substantial Agouti protein was demonstrated in the media of cells infected with the rAAV-Agouti particles by Western blot analysis (Fig. 1A). Thus, Agouti was produced and released from cells infected with rAAV-Agouti viral particles. Injection of 1 μl of the rAAV-Agouti particles (2 × 108 particles) in selected brain regions resulted in local Agouti expression (Fig. 1B,C). It was confirmed that Agouti protein was expressed for at least 60 d with an expression onset as early as 3 d after injection. No Agouti staining in target sites of these neurons, such as the brainstem, was observed. To demonstrate efficacy of Agouti to antagonize MC-induced receptor activation, HEK 293 cells expressing either EGFP or Agouti were mixed with cells transfected with human MC-4 receptor and with Cre-lacZ (as a reporter for receptor stimulation), mimicking rAAV-Agouti infection in vivo with cells releasing Agouti in the surrounding of cells that express MC-4 receptors. The EC50 value of MC-induced MC-4 receptor stimulated Cre-lacZ in HEK 293 cells was right-shifted from 0.25 to 1.1 nm. This indicates that the released Agouti protein functionally antagonizes MC-4 receptor signaling.
Expression of the MC4-R antagonist Agouti in the PVN of adult rats increased body weight and food intake within 7 d after the injection of rAAV-Agouti compared with injection of rAAV-EGFP (Fig. 2A,B). These elevated levels of food intake and body weight gain remained for over 6 weeks. In this episode, PVN-injected rats ate on average 4 gm/d more than controls, resulting in a 50% increase in body weight gain (p = 0.0001). Histological confirmation of Agouti expression indicated that injections leading to Agouti expression just outside of the PVN region (missed; n = 5) were insufficient to induce the rapid onset hyperphagia, in contrast to animals with confirmed PVN Agouti expression (n = 7) [average food intake for the last 3d of the 6 weeks of ad libitum access to regular chow was 24.8 ± 0.6 gm (PVN-EGFP control), 24.6 ± 0.8 gm (missed PVN-Agouti), and 29.0 ± 0.7 gm (PVN-Agouti)]. Despite dense levels of MC4-R expression, we show that Agouti expression in the LH, in contrast to that in the PVN, did not affect body weight and food intake when animals were maintained on regular chow (Fig. 2C,D).
Unlike the rapid onset of obesity in PVN animals, onset was delayed when Agouti was expressed in the DMH. In the first 3 weeks after the injection, DMH-injected animals had similar levels of food intake and body weight when compared with controls. Thereafter, however, DMH-injected animals increased food intake and ate on average 2.5 gm/d more than controls. This resulted in a 30% increase in body weight gain (p = 0.0001) (Fig. 2E,F). Because the DMH is located only 0.76 mm posterior from the PVN region involved in acute onset obesity, we investigated whether the delayed onset in DMH animals was a consequence of Agouti leakage to the PVN region. To exclude this possibility, we showed that rAAV-Agouti did not increase food intake and body weight when injected 0.76 mm anterior, instead of posterior, to the PVN region. To also exclude the possibility that the delayed onset of increased food intake and body weight resulted from a later start time of expression of AAV constructs in the DMH when compared with the PVN, we confirmed that Agouti was expressed in the DMH as early as 3 d after the injection.
High-fat food is known to contribute to the development of obesity. Because Agouti expression in the LH did not contribute to food intake and body weight during regular chow, all animals were maintained on a high-energy diet with a high-fat content for 10 consecutive days to test diet responsiveness. This high-fat diet resulted in increased body weight gain in LH-Agouti animals when compared with LH-EGFP controls (Fig. 3A). Usually, this diet results in a suppression of daily food intake because of the higher caloric density of food (Adolph, 1947). However, animals with Agouti expression in the LH had a far less pronounced suppression of food intake when they were switched to a high-fat diet when compared with LH-EGFP control, DMH, and PVN-Agouti injected animals. Indeed, actual comparison of caloric intake during regular and high-fat diet in LH-Agouti and LH-EGFP control animals revealed that LH-Agouti animals had an increased caloric intake during the high-fat diet (Fig. 3B).
Local long-term Agouti expression in the PVN, DMH, and LH differentially induced hyperphagia and accelerated body weight gain without apparent changes in central neuropeptide systems that are known to stimulate food intake. We found that both AgRP and NPY mRNA levels in arcuate nucleus neurons, as well as MCH and orexin mRNA expression in the lateral hypothalamus, were not affected after long-term Agouti expression in the PVN, DMH, or LH (Fig. 4).
Our results show that Agouti expression in the PVN induced a rapid onset of hyperphagia and body weight gain. This was not unexpected, because a single local injection of a MC agonist stimulates food intake (Kim et al., 2000; Wirth et al., 2001). Surprisingly, Agouti expression in the DMH caused delayed onset of increased food intake and body weight, whereas LH-Agouti expression did not affect body weight on a regular diet but resulted in diet-induced obesity. Therefore, these data show that MC signaling in different brain regions contributes to different characteristics of the obesity syndrome and that the development of these characteristics, such as hyperphagia and diet-induced obesity, are under distinct neuroanatomical control in the brain. These findings open opportunities for additional research aimed at how these different brain regions control different aspects of energy homeostasis. Furthermore, the approach taken is readily available to investigate whether disruption of MC signaling in other brain regions implicated in regulation of food intake, such as the dorsal motor nucleus of the vagus nerve, also induce characteristics of the obesity syndrome.
Here we show that Agouti expression in the PVN and DMH both stimulates food intake and body weight; however, the timing of hyperphagia onset is significantly delayed in the DMH when compared with the PVN. Because these regions are closely located to each other, the DMH effects of Agouti may be explained by leakage of Agouti from the DMH to the PVN, resulting in a delayed onset of the PVN-induced hyperphagia. However, an acute or a delayed onset of hyperphagia and body weight increase were not found in animals with Agouti expression just outside the PVN (missed PVN injections) or 0.76 mm anterior to PVN (the DMH is located 0.76 mm posterior to PVN). Therefore, these data make it unlikely that the delayed onset of hyperphagia after DMH-Agouti expression was caused by leakage of Agouti from the DMH to the PVN and indicate that disruption of MC signaling in the DMH does not result in immediate increased food intake and body weight, such as observed after AAV-Agouti injection in the PVN. Additional studies can now focus on the mechanisms underlying this late onset obesity and the role of the DMH therein.
The present data show that inhibition of MC signaling in the LH selectively accelerates the development of obesity on a high-fat diet by affecting caloric intake. LH-Agouti injected animals had similar caloric intake and body weight gain on regular chow. However, Agouti expression in the LH resulted in an increased caloric intake and subsequent body weight gain on a high-fat diet when compared with LH-EGFP controls. Therefore, our data provide evidence that the LH is an important brain region for the protection against diet-induced obesity rather than for the induction of hyperphagia (as shown in the PVN). The increased caloric intake after inhibition of MC signaling in the LH may result from impaired satiety signaling during intake of high-fat food. In addition, recent studies suggest that the LH region integrates information about energy status and reward (Fulton et al., 2000). Additional behavioral and anatomical studies are needed to demonstrate how energy intake, energy expenditure, and possible reward mechanisms are integrated in the LH.
To further characterize the different mechanisms underlying the brain region-specific forms of Agouti-induced obesity, mRNA gene expression levels of known orexigenic neuropeptides (NPY, AgRP, MCH, and orexins) were measured in several brain regions. Although local and long-term Agouti expression in the PVN, DMH, or LH had marked and different effects on body weight and food intake, no obvious changes in mRNA expression levels of these known orexigenic neuropeptide systems were observed. One interpretation of these results is that brain region-specific disruption of MC-R signaling by Agouti leading to hyperphagia and accelerated weight gain on a high-fat diet is independent of regulation of the hypothalamic orexigenic neuropeptides NPY, AgRP, MCH, and orexins. This idea is consistent with previous data showing that blockade of MC-R signaling, although causing hyperphagia and obesity, did not lead to alterations in expression levels of a number of hypothalamic neuropeptides including, for example, NPY expression in the arcuate nucleus (Kesterson et al., 1997; Adage et al., 2001). We cannot rule out, however, that induction of Agouti expression did transiently alter expression of, at least, some of these neuropeptide systems, allowing these animals to become obese. Therefore, a more extensive analysis of time-dependent changes in these neuropeptide systems and other metabolic parameters may be required to further study physiological changes related to the observed brain region-specific forms of obesity induced by Agouti.
Together, these findings suggest that a single neuropeptide system, such as the MC system, regulates different adaptive behavioral and physiological strategies to alter an organism's energy balance toward a certain direction (e.g., energy conservation) in distinct brain nuclei. The brain homolog for Agouti, AgRP, is expressed in the arcuate nucleus of the hypothalamus and innervates a wide variety of brain regions, such as the PVN, DMH, and LH. These arcuate nucleus neurons respond to sudden changes in peripheral molecules, such as leptin, glucose, and insulin, that provide information about the energy status of an organism. Starvation, for example, leads to a drop in leptin levels (Ahima et al., 1996) and subsequent increase in AgRP gene expression (Hahn et al., 1998). This will result in a simultaneous activation of distinct brain regions that, as indicated by the present study, regulate brain region-specific aspects of physiological processes, each contributing separately to energy conservation.
This research was supported by Dutch Foundation for Scientific Research Project ZonMW 903-039-193.
Correspondence should be addressed to either Roger A. H. Adan or Martien J. H. Kas, Universiteitsweg 100, 3584 CG Utrecht, The Netherlands. E-mail:or .
Copyright © 2004 Society for Neuroscience 0270-6474/04/2410176-06$15.00/0