For procurement of tissue for histological examination, LepRb-EYFP adult male mice (20–30 g, 4–8 weeks old, n = 4) were housed with ad libitum access to both food and water in a light (12 hours on, 12 hours off) and temperature (21.5–22.5°C)-controlled environment. For the experiments investigating the LepRb-expressing cells responsive to peripherally administered ghrelin, ad libitumfed mice were injected with ghrelin (2 μg/g body weight, s.c.) or saline and perfused 2 hours later, as follows. Mice were deeply anesthetized with an intraperitoneal injection of chloral hydrate (350 mg/kg) and perfused transcardially with diethylpyrocarbonate (DEPC)-treated phosphate-buffered saline (PBS) followed by 10% neutral buffered formalin. Brains were removed, postfixed in the same fixative for 4–6 hours at 4°C, immersed in 20% sucrose in DEPC-treated PBS, pH 7.0 at 4°C, and cut coronally at 25 μm into five equal series on a sliding microtome. Tissue was stored at 20°C in antifreeze solution (Elias et al., 1998) until processed.

We generated the GHSR cRNA probes as previously explained in detail (Zigman et al., 2006), although here a 916-bp fragment of cDNA amplified with GHSR-specific primers (mGHSR1047, 5′-GTGGTGTTTGCTTTCATCCTC-3′ and mGHSR1962, 5′-CATGCTCAAATTAAATGCATCC-3′) was used as template. The amplified PCR products were gel-purified and then subcloned into PCR4-TOPO vector (Invitrogen, Carlsbad, CA) according to the manufacturer's protocol. The sequences and directionalities of the inserts were confirmed by DNA sequencing at the core DNA sequencing facility at UTSW Medical Center. To generate antisense ³⁵S-labeled cRNA to use as probes, the plasmid was linearized by restriction digestion and then subjected to in vitro transcription with either T3 or T7 RNA polymerases according to the manufacturer's protocol (Ambion, Austin, TX). Control sense riboprobes were similarly generated.

Free-floating sections of mouse brains were processed sequentially by ISHH and then IHC using a protocol reported previously by our laboratories (Elias et al., 1998). Series of three different mouse brains were processed for GHSR-LepRb coexpression. Briefly, brain sections were first rinsed in DEPC-treated PBS, pH 7.0, and were pretreated with 1% sodium borohydride (Sigma, St. Louis, MO) in DEPC-treated PBS for 15 minutes at room temperature. After thorough washing in DEPC-treated PBS, the sections were rinsed in 0.1 M triethanolamine (TEA, pH 8.0), incubated in 0.25% acetic anhydride in 0.1 M TEA for 10 minutes, then washed again in 2× saline-sodium citrate (SSC). The ³⁵S-labeled GHSR mouse cRNA riboprobe was diluted to 10⁶ cpm/ml in a hybridization solution containing 50% formamide, 10 mM Tris-HCl, pH 8.0, 5.0 mg tRNA (Invitrogen), 10 mM dithiothreitol (DTT), 10% dextran sulfate, 0.3 M NaCl, 1 mM EDTA, pH 8.0, and 1× Denhardt's solution. Next the sections were incubated at 57°C for 12–16 hours in the hybridization solution. Subsequently, sections were rinsed in 4× SSC and incubated in 0.002% RNase solution A (Roche Molecular Biochemicals, Indianapolis, IN) with 0.5 M NaCl, 10 mM Tris-HCl, pH 8.0, and 1 mM EDTA for 30 minutes, followed by a 30-minute incubation in the same buffer minus the RNase. Sections were rinsed with 2× SSC and then with 50% formamide in 0.2× SSC at 50°C. The sections were then submitted to stringency washes as follows: 2× SSC at 50°C for 1 hour; 0.2× SSC at 55°C for 1 hour; 0.2× SSC at 60°C for 1 hour. IHC was begun on the same ISHH-processed sections after first washing them in PBS, pH 7.4. Sections were pretreated with 0.3% hydrogen peroxide in PBS, pH 7.4, for 30 minutes at room temperature and then were incubated in 3% normal donkey serum (Jackson ImmunoResearch Laboratories, West Grove, PA) with 0.25% Triton X-100 in PBS (PBT) for 1 hour. Next, the slides were incubated overnight at room temperature in polyclonal anti-GFP antiserum made in rabbit (Molecular Probes/Invitrogen, Eugene, OR; cat. no. A-6455, lot 71B1, 1:20,000 in PBT-azide). As described previously, this antiserum crossreacts with EYFP (Scott et al., 2009). Furthermore, according to our prior report (Scott et al., 2009), the GFP antiserum does not show reactivity in the absence of EYFP transgene expression. After washing in PBS, sections were incubated in biotinylated donkey antirabbit IgG (Jackson ImmunoResearch Laboratories; 1:1,000) for 1 hour at room temperature, followed by incubation for 1 hour in a solution of avidin-biotin complex (Vectastain Elite ABC Kit, Vector Laboratories, Burlingame, CA; 1:500) diluted in PBS. The sections were next washed in PBS and incubated in a solution of 0.04% diaminobenzidine tetrahydrochloride (DAB; Sigma) and 0.01% hydrogen peroxide in PBS. Brain sections were mounted onto SuperFrost Plus slides (Fisher Scientific, Pittsburgh, PA), and slides were placed in x-ray film cassettes with BMR-2 film (Kodak, Rochester, NY) for 2 days. Slides were then dipped in NTB2 photographic emulsion (Kodak), dried, and stored in desiccant-containing, foil-wrapped slide boxes at 4°C for 4 weeks. Control experiments to confirm the specificity of this protocol involved hybridization with a sense GHSR riboprobe, which showed no evidence of nonspecific labeling. Also, we performed control experiments to test the specificity of the IHC reactions; they included IHC in wildtype brain samples and LepRb-YFP brain samples with omission of primary antiserum. In neither of these controls was staining observed.

ISHH patterns were visualized first on autoradiographic film and then by observing slides dipped in photographic emulsion for direct, cellular visualization. Brain sections were viewed with both a Zeiss Axioskop and a Zeiss Stemi 2000-C dissecting microscope using both brightfield and darkfield optics. Photomicrographs were produced with a Zeiss digital camera attached to the microscopes and a Dell desktop computer. Criteria used to determine whether a GFP-IR cell coexpressed GHSR mRNA included both 1) brightfield visualization of silver granules overlying the DAB-stained cell at 5× the background density of silver granule deposition, and 2) conformation of the overlying silver granules to the shape of the DAB-stained cell. Cell counts were performed on every fifth section of each mouse brain then multiplied by 5 to obtain an estimate of the absolute cell number for each CNS location, as done in the past (Scott et al., 2009). Estimates of cell counts were performed using a 10× objective. The data were corrected for double counting, according to the method of Abercrombie (1946), whereby the ratio of the actual number of neurons to the observed number is represented by T/T+h where T = section thickness, and h = the mean diameter of the neuron along the axis perpendicular to the plane of section. It is important to note that the double-label studies are inherently qualitative. Thus, our results provide data for relative comparisons of cell numbers between brain nuclei and are not accurate counts of absolute cell numbers. Data are presented as average ± standard error of the mean (SEM) (of three mice). An image editing software program, Adobe Photo-Shop 7.0 (San Jose, CA), was used to combine the photo-micrographs into plates, adjust sharpness, contrast and brightness, and remove any obvious dust artifacts from the darkfield images.

GHSR-null and wildtype littermate study animals were derived from crosses of animals heterozygotic for the recombinant, GHSR-null allele. Animals were weaned at 3 weeks of age. One set of GHSR-null and wildtype littermates was provided with ad libitum regular chow and water. Food intake in response to leptin was evaluated when these mice were 10–12 weeks of age. An independent set of GHSR-null and wildtype mice was provided with ad libitum water and high-fat diet (HFD, 88137 Western diet; Harlan Teklad, Houston, TX, which provides 5.3 Kcal/g and 42% Kcal from fat) beginning at age 4 weeks. Responses to leptin were evaluated in these mice after 16 weeks exposure to HFD. All study animals were initially group housed and then were moved into individual housing (one mouse per cage) 3 days before the below leptin administration experiments to acclimate them to being alone and to allow food intake measurements. Mouse leptin was obtained from Dr. E. Parlow (National Institute of Diabetes and Digestive and Kidney Diseases and the National Hormone and Pituitary Program, Torrance, CA). We assessed the response to leptin administration using three different experimental conditions in three different cohorts of mice: 1) acute leptin administration to ad libitum-fed mice. For this study, mice fed with either regular chow or HFD were used. Each mouse was used for both leptin and vehicle treatments on 2 independent days. On the first experimental day, food was removed at 4 pm and animals (n = 6–7 per group per treatment) were weighed and injected with either leptin (at 2 μg/g BW, intraperitoneal [i.p.]) or PBS (total volume = 150 μl). Immediately before the dark cycle started, at 6 pm, a weighed amount of food was reintroduced in the cages. Sixteen hours later the animals and the remaining food were weighed. Five days later the same protocol was repeated in a cross-over fashion in which mice previously injected with leptin were injected with PBS and vice versa. We have successfully studied the actions of leptin on food intake and body weight with this protocol previously (Enriori et al., 2007). 2) Chronic leptin treatment in ad libitum-fed mice. For this study, mice fed with either regular chow or HFD were used. On the experimental day, animals (n = 6–8 per group) were injected twice daily at 8:00 am and 4:30 pm with leptin (1 μg/g BW, i.p.) or PBS (total volume = 150 μl) according to the following scheme: PBS injections for 3 days, followed by leptin injections for 3 days, followed by PBS injections for 2 days. Animals and food were weighed daily during the injection period at 10:00 am. This protocol has previously been established by others (Bjornholm et al., 2007). 3) Acute leptin treatment in fasted-refed mice. This study was performed with mice maintained on regular chow. On the experimental day, animals (n = 6–8 per group) were weighed and regular chow was removed. Twenty-four hours later the animals were weighed and injected with either leptin (at 2 μg/g BW, i.p.) or PBS (total volume = 150 μl). One hour later mice were provided with a weighed amount of chow and food intake was measured after 24 hours. Body weights were also measured at 24 hours. Plasma acyl-ghrelin levels of wildtype and GHSR-null mice in ad libitum-fed and 24-hour fasting conditions were measured in EDTA- and protease inhibitor-treated, acid-stabilized plasma samples using an EIA kit according to the manufacturer's instructions (#10006307, Cayman Chemical, Ann Arbor, MI), as previously described (Sakata et al., 2009).

The level of colocalization of GFP-IR and GHSR mRNA expression was particularly high in the Arc nucleus, a key player in body weight and food intake regulation. It has been shown that leptin and ghrelin have opposite roles in the regulation of the food intake. Also, it has been hypothesized that ghrelin might affect leptin sensitivity (and vice versa). Thus, we decided to evaluate the body weight and food intake responses of GHSR-null mice to peripherally administered leptin. The experiment was performed with ad libitum-fed GHSR-null and wildtype littermates, which had similar body weights (24.9 ± 0.6 and 23.9 ± 0.5 g, respectively; P = NS). A single injection of leptin (2 μg/g BW, i.p.) significantly reduced body weight and overnight food intake in wildtype mice, as expected from previous studies using this same protocol (Enriori et al., 2007) (Fig. 2A,B). Leptin had the same effect in GHSR-null mice and no significant differences were found between genotypes (Fig. 2A,B). Next, we tested the body weight and food intake response of ad libitum-fed GHSR-null and wildtype littermates to a more chronic treatment with leptin (1 μg/g BW, i.p.; twice daily injections for 3 successive days). This latter treatment schedule significantly reduced body weight and food intake in wildtype mice, as previously shown using this same protocol (Bjornholm et al., 2007), and it had the same effect in GHSR-null mice (Fig. 2C,D). No significant differences between genotypes were found.

We previously reported that GHSR-null mice on a mixed genetic background fed HFD gain less body weight than do wildtype littermates. Chronic maintenance of wildtype C57Bl6/J mice on an HFD results in diet-induced obesity and leptin resistance (Collins et al., 2004). Thus, we decided to test if GHSR-null mice (which in the current study are on a pure C57Bl6/J genetic background) and wildtype littermates fed with HFD have a differential sensitivity to the anorectic actions of leptin. These studies were performed in mice exposed to HFD for 16 weeks. As reported previously, GHSR-null mice fed on HFD were leaner than wildtype mice (35.5 ± 1.2 and 38.3 ± 1.2 g, respectively; P = 0.05), although the magnitude of the body weight difference was no longer as great as that observed previously using mice on a mixed genetic background (Zigman et al., 2005). A single injection of leptin (2 μg/g BW, i.p.) failed to reduce either body weight or overnight food intake in either genotype (Fig. 3A,B, respectively). Next, we exposed the HFD-treated GHSR-null and wildtype littermates to the previously described chronic leptin protocol. Three days of twice daily leptin administration (1 μg/g BW, i.p.) slightly reduced body weight in both groups of mice (Fig. 3C). Food intake was not affected by chronic leptin treatment (Fig. 3D). No significant differences between genotypes were found.

In this report we identified a population of medial Arc neurons that highly express both leptin receptors and ghrelin receptors and are well positioned to integrate these physiologically antagonistic signals. The exact chemical identity or identities, however, of these Arc ghrelin- and leptin-coresponsive neurons remains to be determined. Multiple lines of evidence suggest that these cells likely are NPY (neuropeptide Y)- and AgRP (agouti-related gene product)-producing neurons. NPY/AgRP neurons are known to predominate in the more medial aspects of the mouse Arc, whereas POMC/CART neurons localize more to the lateral aspects (Elias et al., 1998, 1999). Within the various subpopulations of Arc neurons, GHSR is primarily expressed in NPY/AgRP neurons (Willesen et al., 1999). It has been shown that ghrelin directly depolarizes and stimulates transcription of c-Fos, along with NPY and AgRP in these neurons (Dickson and Luckman, 1997; Kamegai et al., 2000; Seoane et al., 2003; Cowley et al., 2003). Consistent with ghrelin action in the Arc, blockade of both NPY's and AgRP's actions abolishes ghrelin-induced feeding (Nakazato et al., 2001). Furthermore, ablation of the Arc significantly blunts the orexigenic activity of ghrelin (Tamura et al., 2002). With respect to leptin sensing, most NPY/AgRP neurons also express LepRb (Cheung et al., 1997). It has been shown that leptin directly hyperpolarizes NPY/AgRP neurons and inhibits NPY transcription, AgRP transcription, and release of NPY and AgRP from these neurons (Spanswick et al., 1997; Elias et al., 1999; Enriori et al., 2007). Thus, the current finding of a high degree of ghrelin receptor and leptin receptor coexpression within medial Arc neurons is not unexpected. In contrast, the finding of only a few other extra-Arc brain sites containing only a small number of ghrelin receptor- and leptin receptor-coex-pressing cells is surprising.

The lack of colocalization of ghrelin and leptin receptors within the vast majority of VTA neurons suggests that ghrelin- and leptin-mediated actions on food intake via direct actions in the VTA likely involve different pathways. Infusion of leptin directly into the VTA of rats results in a significant reduction of food intake and body weight acutely (Morton et al., 2009). Correspondingly, leptin has been shown to reduce action potential frequency in in vivo recordings of putative dopaminergic neurons, implying that a direct action of leptin inhibiting dopamine neuron activity underlies its effect on feeding (Hommel et al., 2006). Furthermore, knockdown of VTA leptin receptor expression has been shown to increase feeding along with the rewarding value of sucrose in a sucrose preference test (Hommel et al., 2006). Ghrelin, meanwhile, has also been shown to directly affect dopamine neuronal function, increasing dopamine release in the nucleus accumbens and enhancing action potential generation in VTA dopamine neurons (Abizaid et al., 2006; Skibicka et al., 2011; Jerlhag et al., 2006, 2007). Injection of ghrelin into the VTA increases food intake acutely, intake of a rewarding diet over regular chow, and operant lever pressing for a sucrose reward (Naleid et al., 2005; Abizaid et al., 2006; Egecioglu et al., 2010; Skibicka et al., 2011), while antagonism of GHSR activity in the VTA blunts the orexigenic capacity of peripherally administered ghrelin and also decreases operant lever pressing for a sucrose reward normally induced by an overnight fast (Abizaid et al., 2006; Skibicka et al., 2011). Also, HFD food reward, as measured in calorically restricted mice by conditioned place preference (CPP), has been shown to be dependent on ghrelin receptor signaling, as GHSR-null mice and GHSR antagonist-administered wildtype mice demonstrate a lack of conditioning that is likely VTA-dependent (Perello et al., 2010). Interestingly, leptin and ghrelin appear to regulate different aspects of rewarding behaviors when evaluated by CPP. In particular, we have shown that the physiological increases in ghrelin associated with caloric restriction enhance mainly the associative learning of the reward value of HFD (or rather, the acquisition of CPP for HFD), but are not required for its retrieval (or rather, expression of the learned association) (Perello et al., 2010). In contrast, leptin specifically blocks the expression of food CPP but not the acquisition of food CPP (Figlewicz et al., 2004). In light of these studies, we propose that the physiologically antagonistic actions of leptin and ghrelin in the VTA on food intake and food reward involve the modulation of distinct populations of neurons. Segregation of receptor expression in the VTA, as demonstrated in this report, implies that the circuits and dopamine neurons that effect the actions of leptin and ghrelin are fundamentally different. The investigation of the projection patterns of both ghrelin and leptin receptor expressing cells in the VTA will be crucial to understanding how leptin and ghrelin signals converge on circuits regulating feeding and food reward. For example, prior work has demonstrated leptin receptor expressing VTA neurons project to the amygdala (Leshan et al., 2010) while the projection pattern of GHSR VTA neurons remains to be explored thoroughly.

Similar to the expression of leptin receptors and ghrelin receptors in the VTA, the current study demonstrates a segregation of receptor expression in the dorsal hindbrain. While leptin receptor expression was observed in the medial part of the NTS, ghrelin receptor expression was visualized dorsolateral to this area. It is thus likely that the populations of neurons that express leptin receptors and ghrelin receptors differ both functionally and neurochemically. For example, previous reports have shown that a significant number of leptin-responsive neurons in the hindbrain also coexpress the neuropeptide glucagon-like peptide (GLP-1) (Huo et al., 2008). These leptin-responsive, GLP-1-expressing neurons were absent from the dorsolateral NTS (Huo et al., 2008), which is the population of neurons that express GHSR. Thus, similar to that observed in other areas of the CNS, leptin and ghrelin appear to activate distinct circuits within the hindbrain.