|Home | About | Journals | Submit | Contact Us | Français|
The central melanocortin system regulates body energy homeostasis including the melanocortin-4 receptor (MC4R). The lateral hypothalamic area (LHA) receives dense melanocortinergic inputs from the arcuate nucleus of hypothalamus and regulates multiple processes including food intake, reward behaviors and autonomic function. Using a mouse line in which green fluorescent protein (GFP) is expressed under control of MC4R gene promoter, we systemically investigated MC4R signaling in the LHA by combining double immunohistochemistry, electrophysiology and retrograde tracing techniques. We found that LHA MC4R-GFP neurons co-express neurotensin as well as the leptin receptor but not with other peptide neurotransmitters found in the LHA including orexin, melanin concentrating hormone and nesfatin-1. Furthermore, electrophysiological recording demonstrated that leptin, but not the MC4R agonist melanotan II, hyperpolarizes the majority of LHA MC4R-GFP neurons in an ATP-sensitive potassium channel-dependent manner. Retrograde tracing revealed that LHA MC4R-GFP neurons do not project to the ventral tegmental area, dorsal raphe nucleus, nucleus accumbens and spinal cord, and only limited number of neurons project to the nucleus of solitary tract and parabrachial nucleus. Our findings provide new insight into MC4R signaling in the LHA and its potential implication in homeostatic regulation of body energy balance.
Due to its increasing prevalence and serious co-morbidities including diabetes, cancer and cardiovascular diseases, obesity has become a major health concern facing to the developed countries (WHO, 2004). Our understanding of the mechanisms underlying body weight regulation has dramatically improved since the discovery that several peripheral hormones, including insulin, leptin and ghrelin, regulate energy balance through signaling in the brain (Saper et al., 2002; Schwartz et al., 2000; Zhang et al., 1994). It is now clear that the central melanocortin system is critical for the integration of multiple peripheral signals to maintain body weight homeostasis (Bjorbaek and Hollenberg, 2002; Cone, 2005; 2006; Elmquist et al., 1999; Schwartz et al., 2000; Shimizu et al., 2007).
The central melanocortin system consists of two G protein coupled receptors: the melanocortin-3 receptor (MC3R) and melanocortin-4 receptor (MC4R) (Cone, 2006). An endogenous ligand for both receptors is α-melanocyte-stimulating hormone (α-MSH), which is derived from the precursor proopiomelanocortin (POMC). The central melanocortin system is unique in that it also features an endogenous antagonist agouti-related protein (AgRP). AgRP and α-MSH are synthesized in distinct population of neurons within the arcuate nucleus of the hypothalamus (ARH) and project diffusely throughout the brain (Bagnol et al., 1999; Cone, 2005; Schwartz et al., 2000). Mutations in the MC4R gene are found in up to 5% of human patients with morbid obesity have mutations in the MC4R gene (Santini et al., 2009). Numerous studies in rodents have also clearly demonstrated that activation of MC4Rs suppresses feeding and increases energy expenditure, whereas pharmacologic or genetic inhibition of MC4Rs has the opposite effects (Balthasar et al., 2005; Benoit et al., 2000; Chen et al., 2000; Fan et al., 1997; Marsh et al., 1999). Thus, characterizing neural circuits affected by MC4R signaling will facilitate our knowledge on how the brain integrates peripheral signals to maintain body energy balance.
Consistent with the broad projections of ARH POMC/AgRP neurons in the central nervous system (CNS), MC4Rs are widely expressed in many regions of the rodent brain (Kishi et al., 2003; Liu et al., 2003; Mountjoy et al., 1994). Of these regions, the paraventricular nucleus (PVH) and lateral hypothalamic area (LHA) are of particular interest since they receive dense inputs from leptin responsive POMC/ AGRP neurons (Elias et al., 1999) and express MC4Rs (Kishi et al., 2003; Liu et al., 2003). While it is clear that MC4Rs are expressed in these regions, their specific functions are not completely understood. One study utilizing a Cre/loxP genetic technique demonstrated that MC4R signaling in SIM1-expressing neurons of the PVH and/or amygdala is essential for homeostatic regulation of energy balance primarily by affecting food intake but not energy expenditure (Balthasar et al., 2005). Several neuroanatomical studies have shown that PVH MC4R-positive neurons colocalize with several neuropeptides, including corticotropin-releasing hormone, oxytocin and thyrotropin-releasing hormone, and project to multiple regions including the nucleus of the solitary tract (NTS), the spinal cord and the median eminence of hypothalamus (Ghamari-Langroudi et al., 2011; Harris et al., 2001; Liu et al., 2003; Lu et al., 2003).
In contrast to our understanding of the PVH MC4R pathway, relatively little is known about the physiological importance of MC4Rs expressed in the LHA. The LHA has long been appreciated for its crucial role in the regulation of feeding, motivational behaviors and autonomic function (Bernardis and Bellinger, 1993b). A recent study demonstrated that pharmacological blockade of LHA MC4R signaling does not significantly alter food intake when animals fed regular chow (RC) (Kim et al., 2000). However a separate study found that over-expression of the melanocortin receptor antagonist Agouti in the LHA induces significant hyperphagia and body weight gain when mice were placed on a palatable high fat diet (HFD) (Kas et al., 2004). This suggests that the MC4R signaling in LHA regulates hedonic drive of feeding and might be responsible for exacerbated diet-induced obesity seen in MC4R-null mice (Butler et al., 2001; Srisai et al., 2011).
In the current study we systemically examined the neuroanatomical and electrophysiological characteristics of MC4R neurons in the LHA and provide new insights into the understanding of its potential physiological function.
Mice in which GFP expression is under control of MC4R gene promoter (MC4R-GFP) were obtained for neuroanatomical studies (Liu et al., 2003). As reported previously, nearly all of the MC4R-GFP positive neurons in this line co-express MC4R mRNA including neurons shown in a representative photograph of the perifornical area (Figure 5 in the original paper). Hence, we believe that the MC4R-GFP line faithfully identifies MC4R-expressing neurons in the LHA. Mice were housed in the University of Texas Southwestern Medical Center (UTSW) vivarium in a temperature-controlled environment (lights on: 06:00–18:00) with ad lib access to water and standard chow (SC) (4% fat diet #7001, Harlan-Teklad, Madison, WI). All animal procedures were performed in accordance with UTSW Institutional Animal Care and Use Committee guidelines.
The antibodies used in the present study are all commercially available and have been tested by different laboratories. Their key features are summarized in Table 1.
Surgical procedure was followed as previously described (Chamberlin et al., 1998; Gautron et al., 2010a). Briefly, MC4R-GFP mice were anesthetized with ketamine HCl/xylazine HCl (80:12 mg/kg, i.p.) and restrained in a Kopf (Tujunga, CA) stereotaxic apparatus. A small hole was drilled into the skull under aseptic conditions. A glass micropipette connected to an air pressure injector system or iontophoresis machine was positioned via the stereotaxic manipulator. Approximately 50–100 nL of retrograde tracer, fluorogold (Fluorochrome, CO, USA) or choleratoxin subunit B (CTb) (List Biological Laboratories), was slowly administered into the ventral tegmental area (VTA) (−3.2 mm from bregma, 1.1 mm lateral with 10 degree injection arm, −4.3 mm from the surface of the cortex), the dorsal raphe (DR) (−4.6 mm from bregma, 0.1 mm lateral, −2.7 mm from the skull), the nucleus accumbens-shell (NAc-shell) (1.5 mm from bregma, 1.5 mm lateral with 10 degree injection arm, −4.2 mm from the surface of the cortex), the nucleus of solitary tract (NTS) (+0.16 from the obex, ±0.23 mm lateral, −0.36mm from the surface of the brainstem and the parabrachial nucleus (PBN) (−5.2 mm from bregma, 1.3 mm lateral, −2.8 mm from the cerebellar cortex). For spinal cord injection, the T6–T7 facet joint was removed by surgical instrumentation and retrograde tracer FG was unilaterally injected by air pressure injection system. After injection, the glass micropipette was removed and the incision was closed with surgical staples. Mice were allowed to recover for 5–7 days post surgery and then perfused with 4% paraformaldehyde (PFA) and brain sections were processed for double fluorescent immunohistochemistry for GFP and tracers.
To enhance labeling of cell bodies with immunohistochemistry, MC4R-GFP mice were treated with intracerebroventricular (ICV) colchicine (10 ug) 48 hours prior to perfusion. For leptin-induced phospho-STAT3 and melanotan II (MT-II)-induced c-Fos experiments, MC4R-GFP mice were treated with intraperitoneal (IP) leptin (5 mg/kg, National Hormone and Peptide Program, Harbor-UCLA Medical Center, Torrance, CA), IP MT-II (3 mg/kg, Bachem, Torrance, CA IP), ICV leptin (1.5 ug), or ICV MTII (1 ug) as noted, 2 hours prior to perfusion. Mice were transcardially perfused with 4% PFA and cryopreserved in 20% sucrose. Brains were then sectioned into 30 micron coronal sections (collecting 1:5 sections) and stored in cryoprotectant at −20 °C until use. Brain sections were blocked and permeabilized with 3% normal donkey serum (NDS, Jackson Immuno Research, West Grove, PA), 0.3% Triton X-100 in PBS for 30 minutes at room temperature (RT), rinsed with PBS and incubated with primary antibodies (diluted in 3% NDS, 0.3% Tween-20 in PBS) at room temperature for 2–3 hours or at 4 °C for overnight. For leptin-induced pSTAT3, brain sections were treated with 1% NaOH, 1% H2O2 in distilled water for 20 minutes at room temperature, rinsed with PBS 3 times for 10 minutes each, incubated with 0.3% Glycine in PBS for 10 minutes, rinsed with PBS 3 times for 10 minutes each, incubated with 0.03% SDS in PBS for 10 minutes, blocked in 3% NDS, 0.3% Triton X-100 in PBS for 30 minutes at room temperature and then incubated with primary antibody for 24 hours at room temperature plus additional 48 hours at 4 °C. Images were first taken by fluorescent microscopy (Nikon Eclipse, 80i) and co-localization was determined by confocal microscope scanning (LSM510-META, Zeiss, Thornwood NY). Cells counts for pSTAT3 and GFP positive neurons were made in mid-dorsal lateral hypothalamus from three consecutive sections with the highest abundance of GFP expression.
To examine the role of leptin receptor in MC4R-positive LHA neurons, we examined the effect of leptin (100 nM) on the membrane potential of GFP-positive LHA neurons from MC4R-GFP transgenic mice. As described previously (Hill et al., 2008a; Williams et al., 2010), 4–10 weeks old mice were deeply anesthetized with chloral hydrate and transcardially perfused with a modified ice-cold artificial cerebrospinal fluid (ACSF, described below), in which an equiosmolar amount of sucrose was substituted for NaCl. The mice were then decapitated, and the entire brain was removed and immediately submerged in ice-cold, carbogen-saturated (95% O2 and 5% CO2) ACSF (126 mM NaCl, 2.5 mM KCl, 1.2 mM MgCl2, 2.4 mM CaCl2, 1.25 mM NaH2PO4, 26 mM NaHCO3, and 10 mM glucose). Coronal sections (200µM) were cut with a Leica VT1000S Vibratome and then incubated in oxygenated ACSF at 32°C for at least 1 h before recording. Slices were transferred to the recording chamber and allowed to equilibrate for 10–20 min prior to recording. The slices were bathed in oxygenated ACSF (32°C–34°C) at a flow rate of approximately 2 ml/min. Epifluorescence was briefly used to target fluorescent cells, at which time the light source was switched to infrared differential interference contrast imaging to obtain the whole-cell recording using Zeiss Axioskop FS2 Plus (Carl Zeiss; Jena, Germany) equipped with a fixed stage and a Hamamatsu C2741-60 charged-coupled device camera (Hamamatsu photonics; Hamamatsu city, Japan). Membrane potentials were recorded in the current clamp mode using an Axopatch 700B amplifier (Molecular Devices; Sunnyvale, CA, USA), low-pass filtered at 1 kHz, digitized at 10 kHz and analyzed offline on a PC with pCLAMP programs (Molecular Devices), Origin (Microcal; Piscataway, NJ, USA) or IgorPro (Wavemetrics; Lake Oswego, OR, USA). Recording electrodes had resistances of 2.5–5 MΩ when filled with the K-gluconate pipette solution which contains 120 mM K-gluconate, 10 mM KCl, 10 mM HEPES, 5 mM EGTA, 1 mM CaCl2, 1 mM MgCl2, and 5 mM Mg-ATP adjusted to pH 7.3.
Eighteen 8–9 week old C57BL/6 male mice (Jackson Laboratory, Bar Harbor, ME) were used for the study. One cohort of mice received IP injection of saline, MTII (3 mg/kg) or leptin (5 mg/kg), respectively (n=3/group), at 1800 (ZT18) just before lights off and fresh brains were collected by decapitation at 2000 (ZT20). Likewise, a second cohort of mice received IP injection of saline, MTII (3 mg/kg) or leptin (5 mg/kg), respectively (n=3/group), at 0600 AM (ZT6) just before lights on and fresh brains were collected by decapitation at 0800 (ZT8). Brains were post-fixed in 4% PFA at 4 °C for 48 hours, cryopreserved in 20% sucrose solution at 4 °C for 24 hours, and then cut into 5 series of 30 um sections. Brian sections were then stained for c-Fos and orexin by indirect biotin/avidin-HRP techniques as previously reported (Elias et al., 1999).
Adobe Photoshop CS2 was used to combine drawings and digital images into plates. The contrast and brightness of images were adjusted when necessary. In addition, red-green fluorescence images were converted to magenta-green for color-blind readers.
GraphPad Prism 5 software (GraphPad Software Inc., San Diego, CA) was used to perform statistical analyses. P< 0.05 was considered to be statistically significant.
A previous study using in situ hybridization demonstrated low to moderate expression of MC4Rs in the LHA of the rat (Kishi et al., 2003). In order to facilitate neuroanatomical studies of MC4R, a mouse line expressing tau-conjugated GFP under the control of the MC4R promoter (MC4R-GFP) was previously generated and validated (Liu et al., 2003). This line has also been used for other neuroanatomical studies (Gautron et al., 2010b). Utilizing this mouse line, we confirmed the expression of MC4R-GFP-positive neurons mainly in the dorsal LHA (Fig. 1A, F) as well as in the perifornical area (PFA) (Fig. 1H). The LHA is comprised of a heterogeneous cell population expressing multiple neurotransmitters. In order to determine the neurochemical phenotype of LHA MC4R-positive neurons, we carried out a series of co-localization studies to determine which neurotransmitters are expressed in LHA MC4R-GFP neurons. Dual-label fluorescent IHC demonstrated that LHA MC4R-GFP positive neurons are completely distinct from orexin, melanin concentrating hormone (MCH) (Fig. 1A, B), and nesfatin-1 expressing neurons (Fig 1C, D). In contrast, MC4R-GFP was co-expressed with the peptide neurotransmitter neurotensin (NT) in both dorsal LHA (Fig. 1E, F) and PFA (Fig 1E, H). Schematic drawing of distribution of NT, MC4R-GFP and double-positive cells in mid-dorsal LHA and PFA is shown in figure 1I. Detailed counting of NT, GFP and double-positive cells revealed that ~75% of MC4R-GFP neurons co-express with NT in the LHA and PFA (Table 2a). Importantly, after observing the colocalization of MC4R-GFP and NT within the LHA, we surveyed the entire brain and did not observe additional significant co-localization in any additional sites (data not shown).
Leinninger and colleagues recently identified a population of LHA neurons expressing the active form of the leptin receptor (LepRb), which is also distinct from orexin and MCH but co-localized with GAD67, a marker of GABAergic neurons (Leinninger et al., 2009). We have previously shown that perifornical MC4R-GFP neurons co-express GAD67 (Liu et al., 2003). These similarities in neurochemical phenotype between LHA LepRb-positive neurons and LHA MC4R-GFP neurons led us to hypothesize that LHA MC4R/NT/GAD67-positive neurons might co-express LepRb as well. To test this model, we conducted dual-label IHC for leptin-induced pSTAT3 and GFP in our MC4R-GFP mouse line. As predicted, there was a clear co-localization between MC4R-GFP neurons and pSTAT in the middorsal portions of LHA after leptin treatment (Fig. 2A, B). A schematic drawing of the distribution of pSTAT3, MC4R-GFP and double-positive cells in mid-dorsal LHA and PFA is shown in Figure 2C. Detailed counting of pSTAT3, GFP and double-positive cells revealed that ~80% of MC4R-GFP neurons co-express with pSTAT3 (Table 2b). In addition to the LHA, colocalization was also observed in the posterior hypothalamus and lateral periaqeductal grey (data not shown). Additionally, to understand the pharmacological effects of both MC4R and LepRb signaling within the LHA, we performed immunohistochemistry for c-Fos, a commonly used marker of neuronal activation (Kaczmarek and Nikolajew, 1990). MC4R-GFP mice were treated with either MC4R agonist (MTII) or leptin. Administration of either MTII or leptin (both ICV and IP) failed to induce c-Fos activation in LHA MC4R-GFP neurons (data not shown).
To further characterize the role of MC4R and leptin signaling within the LHA, we tested the electrophysiological responses of individual LHA MC4R-GFP neurons to leptin and MTII. We assessed membrane potential in neurons in the current clamp mode. The average resting membrane potential was −45.7 ± 1.7 mV in 16 cells. In 10 cells, leptin (100 nM) hyperpolarized the membrane potential by −8.6 ± 1.4 mV (Fig 3A–B). In the remaining 6 cells, leptin did not significantly alter the membrane potential (−0.2 ± 0.5 mV) (Fig 3C). The leptin-induced hyperpolarization lasted for approximately 10 min and was completely reversed by subsequent application of tolbutamide (200 uM, a specific KATP channel blocker) (Fig 3A–B). These results suggest that leptin directly inhibits a subpopulation of LHA MC4R-GFP neurons by activating KATP channels. Conversely, MTII (100 nM) had no effect on membrane potential in all LHA MC4R-GFP neurons recorded (Fig. 3D). These results suggest that leptin inhibits the firing of MC4R-GFP neurons in the LHA.
Next we sought to determine the innervation pattern of LHA MC4R-GFP neurons. Recent work by Leinninger and colleagues (Leinninger et al., 2009) has shown that LHA LepRb neurons mainly project to the ventral tegmental area (VTA) and dorsal raphe (DR). Since one third of leptin-induced pSTAT3 neurons in the LHA were colocalized with MC4R-GFP, we hypothesized that LHA MC4R-GFP neurons might project to VTA and/or DR as well. To test this possibility, we injected retrograde tracer into these two brain regions of MC4R-GFP mice and brain sections were processed for double IHC for the GFP and tracer focusing on the LHA. Four successfully targeted cases were obtained for the VTA and representative photography of injection site is shown in Fig 4A. While many of LHA neurons were labeled with tracer as predicted, none of the neurons were GFP-positive (Fig 4B), suggesting that LHA MC4R-GFP neurons do not project to the VTA. For the DR, we obtained three successful injection cases and representative photography of injection site is shown in Fig 4C. While some of the LHA neurons were clearly labeled with retrograde tracer after successful injection into the DR, but none of the neurons were colocalized with GFP (Fig 4D), again suggesting that LHA MC4R-GFP neurons do not project to the DR as well. Additionally, since it has been reported that hedonic overconsumption of palatable high fat diet was observed in mice with overexpression of MC4R antagonist Agouti solely in the LHA (Kas et al., 2004), we hypothesized that LHA MC4R-GFP neurons might project to nucleus accumbens (NAc) which has long been believed as a brain reward center that might regulate hedonic drive of feeding. To test this possibility, we injected the retrograde tracer fluorogold (FG) into the NAc-shell, which receives innervations from LHA neurons (Kelley et al., 2005). Four successful cases were obtained and, as shown in Fig 4E, injection sites included the NAc-shell with extension into the adjacent core region. For all four successfully injected cases, only a few LHA neurons were labeled with tracer and none of them were colocalized with GFP signal (Fig 4F), which indicates that LHA MC4R-GFP neurons do not project to the NAc.
To further explore potential targets of LHA MC4R-GFP neurons, we injected tracer into the nucleus tractus solitarius (NTS), parabrachial nucleus (PBN), and spinal cord, which all have been previously shown to be innervated by LHA neurons (Haring and Davis, 1983; Kelly and Watts, 1998; Moga et al., 1990; Tokita et al., 2009; Zheng et al., 2005). We obtained three successful injection cases for the NTS and representative injection is shown in Fig 5A. For all three successfully NTS-injected cases, retrograde tracer CTb-labeled neurons were clearly observed in the LHA (Fig 5B), but only occasional neurons (< 8%) were colocalized with GFP signal (Fig 5C), indicating that the majority of LHA MC4R-GFP neurons do not project to the NTS. Since it has previously shown that the PBN is densely innervated by LHA NT-positive neurons (Moga et al., 1990), the PBN was also considered as a possible target for LHA MC4R-GFP neurons. In all six of the successfully injected cases, LHA neurons were intensively labeled by tracer (Fig. 5G, 5H, 5I). However, we only observed a few neurons that were double positive for tracer and GFP signal (< 12%), indicating that the PBN is not a major target of LHA MC4R-GFP neurons. For the spinal cord, retrograde tracer FG was injected into thoracic spinal cord (T6–T7). In the two successfully injected cases (Fig. 5D, 5E, 5F), considerable numbers of neurons in the LHA were labeled with FG as predicted, but none of them were GFP-positive, indicating that LHA MC4R-GFP neurons do not project to the spinal cord. Since it has been reported that PVH MC4R neurons project to both NTS and spinal cord (Ghamari-Langroudi et al., 2011), we further analyzed PVH MC4R-GFP neurons in our successfully injected cases as a positive control experiment. Indeed, we confirmed clear projections of PVH MC4R-GFP neurons to both NTS and spinal cord in all of cases studied (data not shown).
In addition to the VTA and DR, orexin neurons (also known as hypocretin) in the LHA have been reported to be a direct target of LHA LepRb neurons through a local circuit (Leinninger et al., 2009; Louis et al., 2010a). Furthermore, NT-positive LHA neurons have also been shown to project to the VTA as well as locally to orexin neurons within the LHA (Leinninger et al., 2011). Because we had already ruled out the VTA as a potential target for LHA MC4R-GFP neurons, we tested the possibility that LHA MC4R-GFP neurons project to orexin neurons as well. Unfortunately because both MC4R-GFP and orexin neurons are interspersed within the LHA, we could not test this hypothesis using conventional retrograde tracing technique. Therefore, we utilized a complementary pharmacological approach to determine if the MC4R agonist MTII or leptin could affect orexin neuron activity using c-Fos IHC as a surrogate marker of activation. Orexin neurons display a diurnal activity pattern (Estabrooke et al., 2001), therefore we preformed this experiment at two different time points, ZT8 and ZT20. As expected, in the saline-injected groups, the percentage of c-Fos expression in orexin neurons was significantly higher during the early dark phase (ZT20) compared to the early light phase (ZT8) (Fig 6A–F, table 3). Both MTII and leptin increased c-Fos expression in orexin neurons at ZT8 (Fig 6G–L, table 3), while no differences in c-Fos levels were detected between groups at ZT20 (Fig 6G–L, table 3).
In the present study, we investigated anatomic and physiological properties of neurons in the LHA expressing MC4Rs. We found that LHA MC4R-GFP neurons co-express the anorexigenic peptide NT and the long form of LepRb, but not orexin, MCH or Nesfatin-1. While this population of neurons had no electrophysiological response to the MC4R agonist MTII, we found that most of the LHA MC4R-GFP neurons are inhibited by leptin. We were unable to identify the direct targets of LHA MC4R-positive neurons as they do not project to any of the brain regions we tested including NAc, VTA, DR, NTS, PBN and spinal cord. However, we provide pharmacological evidence supporting a possible local circuit with orexin neurons in the LHA.
Early lesion and electronic stimulation studies identified the LHA as an important brain region that regulates multiple processes including feeding, reward behaviors and autonomic function (Bernardis and Bellinger, 1993a). This was further supported at molecular level by the discoveries of LHA specific neuropeptides orexin and MCH which innervate throughout brain and are known to regulate various physiological functions including feeding, sleep/wake cycle, reward behaviors and body energy homeostasis (Berthoud and Munzberg, 2011). A recent study, however, reported that a population of LHA neurons expressing LepRb was completely distinct from orexin and MCH but completely overlapped with GAD67. These neurons also modulates the midbrain dopamine system to suppress feeding (Leinninger et al., 2009). In the present study, we demonstrate LHA MC4R-positive neurons are a part of this neuronal population and contain anorexigenic peptide NT. Understanding the relevant physiological function of NT release from LHA MC4R/LepR double-positive neurons is of great interest for the future studies.
The interaction of the leptin and melanocortin systems has been extensively investigated and debated. Leptin is a key regulator of POMC expression, the precursor of the melanocortin agonist α-MSH, (Schwartz et al., 1997) and administration of a melanocortin antagonist blocks the anorexigenic effect of leptin administration (Seeley et al., 1997). Additionally, central inhibition of the melanocortin system blunts leptin-induced activation of the sympathetic nervous system and dissipation of excess energy by brown adipose tissue (Dunbar and Lu, 1999; Satoh et al., 1998). In contrast, others have demonstrated that inhibition of both leptin and melanocortin signaling results in additive effect on weight gain and that leptin resistance in melanocortin signaling is a compensatory response to chronic hyperleptinemia (Boston et al., 1997; Trevaskis and Butler, 2005). In addition, deletion of leptin receptors from POMC neurons produces obesity, but is much milder than leptin deficiency suggesting that additional sites are important for leptin action (Balthasar et al., 2004; Hill et al., 2010). Finally, leptin deficient, but not MC4R-null, mice increase thermogenesis and locomotor activity after exposure to HFD (Butler et al., 2001). Our data now identify an overlap of leptin and melanocortin signaling in a small sub-set of neurons in the LHA. This overlap in MC4R and LepRb signaling suggests that this population of neurons may integrate numerous peripheral signals to regulate energy balance.
The degree of cross-talk between MC4R and LepR signaling pathways in this small subset of neurons remains to be elucidated. MC4R signaling has previously been demonstrated in vitro to enhance leptin-induced phosphorylation of Stat3 via a MAPK dependent pathway (Zhang et al., 2009b). Interestingly, in this system leptin signaling had no effect on MC4R induced levels of cAMP suggesting that while MC4R modulates LepRb, the converse is not true. Alternatively, we show here that leptin-induced hyperpolarization of LHA neurons is dependent on KATP channels. A similar mechanism of action occurs in POMC neurons, where leptin and insulin modulate activity of POMC neurons via PI3 kinase-dependent action on KATP channels (Belgardt et al., 2009; Hill et al., 2008b). Consistent with this model, MC4R signaling has been shown to activate PI3K in transfected CHO-K1 cells (Vongs et al., 2004). Finally, previous work in pancreatic β-cells has demonstrated a role for the cAMP sensor EPAC2 in modulation of KATP channel activity (Zhang et al., 2009a), suggesting that a similar mechanism might occur in LHA. However recent work has demonstrated that the melanocortin 3/4 receptor agonist MT II was not able to induce EPAC-mediated leptin resistance in hypothalamic POMC neurons (Fukuda et al., 2011).
One area of great interest in the future will be to determine the target of LHA MC4R/NT/LepRb neurons. Previous studies have identified the VTA and DR as regions innervated by LHA LepRb neurons (Leinninger et al., 2009). However, we did not observe any projections of LHA MC4R-GFP neurons to these areas. For other sites we tested, relatively few LHA MC4R-GFP neurons projected to the NTS (<8%) and PBN (<12%) and none projected to the spinal cord. One potential target of MC4R LHA neurons are orexin (also known as hypocretin) expressing neurons in the LHA. These neurons have also been identified as a target of LHA LepRb neurons via a local circuit (Louis et al., 2010b). Consistent with this possibility, we demonstrate here that leptin hyperpolarizes the majority of LHA MC4R-GFP neurons (Fig. 3). Hyperpolarization of the GABAergic neurons would be predicted to disinhibit down stream orexin neurons leading to activation. Indeed, pharmacological administration of either leptin or MT II increases c-Fos expression in orexin neurons at ZT8 (Table 3), although we can not definitely link this effect to LepRb or MC4Rs expressed in the LHA. Clearly additional neuroanatomical and functional studies will need to be done to delineate the targets of MC4R-LHA neurons. Dissecting this kind of local neuronal circuit will require the development of genetic and trans-synaptic tracing tools, such as a MC4R-Cre mouse line, to directly test the possibility that LHA MC4R-GFP neurons synapse onto local orexin neurons.
Electrophysiological properties of LHA LepR neurons in response to leptin seem to be heterogeneous. Leinninger et al (2009) have shown that leptin depolarizes and hyperpolarizes about 34% and 22% of LHA LepR neurons, respectively, and nearly half of LHA LepR neurons did not respond to leptin. Our electrophysiological recording of LHA MC4R-GFP neurons revealed that leptin hyperpolarizes the majority of neurons we recorded from (10/16) and had no effect on the remaining neurons. Since our double IHC revealed about ~75% of colocalization of LHA MC4R-GFP neurons with LepRb (as determined by pSTAT3), this indicates that leptin hyperpolarizes the majority of LHA MC4R/LepR neurons. It is very interesting that we do not observe a direct depolarization of LHA neurons by a MC4R agonist. This observation suggests the alternative possibility that MC4R signaling may act pre-synaptically at the terminals of LHA MC4R neurons to modulate neurotransmitter release. This has been previously demonstrated for melanocortin receptor expressing neurons in the paraventricular nucleus of the hypothalamus (Cowley et al., 1999) and the nucleus tractus solitarius (Wan et al., 2008). Unfortunately, because orexin neurons within the LHA are a likely target for MC4R-LHA neurons, we cannot distinguish this possibility using pharmacologic techniques. Future studies using genetic methods to specifically manipulate expression of MC4Rs in LHA neurons will need to resolve this question.
In conclusion, we describe a novel population of neurons within the LHA that express both the MC4R and the LepRb. Understanding the function of these neurons may shed important insights into energy homeostasis and the development of obesity.
Acknowledgements: This work was also supported by ML (MH084058-01A1, Disease Oriented Clinical Scholar Award and NARSAD Young Investigator Award), KWW (K01DK087780) and JKE (R01DK53301 and RL1DK081185).
Other Acknowledgements: We thank Jeffrey Friedman for use of MC4R-GFP mice and Carol F. Elias and Yuichi Sato for technical assistance.
Conflict of Interest: None