Research Articles, Cellular/Molecular

Galanin-Expressing GABA Neurons in the Lateral Hypothalamus Modulate Food Reward and Noncompulsive Locomotion

Emily Qualls-Creekmore, Sangho Yu, Marie Francois, John Hoang, Clara Huesing, Annadora Bruce-Keller, David Burk, Hans-Rudolf Berthoud, Christopher D. Morrison and Heike Münzberg
Journal of Neuroscience 21 June 2017, 37 (25) 6053-6065; https://doi.org/10.1523/JNEUROSCI.0155-17.2017

Abstract

The lateral hypothalamus (LHA) integrates reward and appetitive behavior and is composed of many overlapping neuronal populations. Recent studies associated LHA GABAergic neurons (LHAGABA), which densely innervate the ventral tegmental area (VTA), with modulation of food reward and consumption; yet, LHAGABA projections to the VTA exclusively modulated food consumption, not reward. We identified a subpopulation of LHAGABA neurons that coexpress the neuropeptide galanin (LHAGal). These LHAGal neurons also modulate food reward, but lack direct VTA innervation. We hypothesized that LHAGal neurons may represent a subpopulation of LHAGABA neurons that mediates food reward independent of direct VTA innervation. We used chemogenetic activation of LHAGal or LHAGABA neurons in mice to compare their role in feeding behavior. We further analyzed locomotor behavior to understand how differential VTA connectivity and transmitter release in these LHA neurons influences this behavior. LHAGal or LHAGABA neuronal activation both increased operant food-seeking behavior, but only activation of LHAGABA neurons increased overall chow consumption. Additionally, LHAGal or LHAGABA neuronal activation similarly induced locomotor activity, but with striking differences in modality. Activation of LHAGABA neurons induced compulsive-like locomotor behavior; while LHAGal neurons induced locomotor activity without compulsivity. Thus, LHAGal neurons define a subpopulation of LHAGABA neurons without direct VTA innervation that mediate noncompulsive food-seeking behavior. We speculate that the striking difference in compulsive-like locomotor behavior is also based on differential VTA innervation. The downstream neural network responsible for this behavior and a potential role for galanin as neuromodulator remains to be identified.

SIGNIFICANCE STATEMENT The lateral hypothalamus (LHA) regulates motivated feeding behavior via GABAergic LHA neurons. The molecular identity of LHAGABA neurons is heterogeneous and largely undefined. Here we introduce LHAGal neurons as a subset of LHAGABA neurons that lack direct innervation of the ventral tegmental area (VTA). LHAGal neurons are sufficient to drive motivated feeding and locomotor activity similar to LHAGABA neurons, but without inducing compulsive-like behaviors, which we propose to require direct VTA innervation. Our study integrates galanin-expressing LHA neurons into our current understanding of the neuronal circuits and molecular mechanisms of the LHA that contribute to motivated feeding behaviors.

Introduction

The lateral hypothalamic area (LHA) is implicated in feeding, food reward, and other motivated behaviors. Early research established that the LHA is necessary to maintain homeostatic feeding behaviors (Anand and Brobeck, 1951). Furthermore, rodents show impressive motivation to self-administer electric stimulations to the LHA (Olds and Milner, 1954) and the LHA is well established for its interaction with the mesolimbic dopamine reward circuitry in the brain (Peyron et al., 1998; Korotkova et al., 2003; Leinninger et al., 2009; Kempadoo et al., 2013). The LHA is a large region with a heterogeneous molecular makeup and a diverse pattern of inputs and outputs (Swanson et al., 2005; Hahn and Swanson, 2010; Bonnavion et al., 2016). Much research has focused on defining the molecular signature of LHA neurons and their neural projections that regulate feeding and food reward (Berthoud and Münzberg, 2011).

Recent studies have focused on GABAergic LHA (LHAGABA) neurons and associated neural circuits for their involvement in feeding and especially food reward (Jennings et al., 2015; Nieh et al., 2015). Optogenetic and chemogenetic stimulation and inhibition of LHAGABA neurons had opposing effects on operant food seeking and food consumption. Interestingly, in vivo visualization of LHAGABA neuronal activity highlighted that food-seeking versus food-consumption behaviors are represented by activation of anatomically distinct GABA populations (Jennings et al., 2015). Similarly, activation of all ventral tegmental area (VTA)-projecting LHA neurons enhanced sucrose seeking, while selective activation of VTA-projecting GABAergic LHA neurons promotes food consumption but not sucrose seeking (Nieh et al., 2015).

We recently described a subpopulation of LHA neurons that express the neuropeptide galanin (LHAGal) together with leptin receptors and neurotensin (Berthoud and Münzberg, 2011; Laque et al., 2013). These LHAGal neurons modulate food reward without promoting overall food consumption (Laque et al., 2015). Also, LHAGal neurons do not project to the VTA, but instead innervate the locus ceruleus (LC) and local LHA neurons (Laque et al., 2015). Galanin often acts as an inhibitory neuropeptide, and many galanin neurons coexpress the inhibitory neurotransmitter GABA (Melander et al., 1986; Sherin et al., 1998). Thus, we hypothesized that LHAGal neurons may represent a distinct subset of LHAGABA neurons that selectively modulates food reward, but not overall food consumption.

To test this hypothesis, we compared operant food-seeking behavior and overall chow consumption following chemogenetic activation of LHAGal neurons or LHAGABA neurons. Here we show that LHAGal neurons are sufficient to promote operant food seeking, while overall chow consumption remained unchanged. In addition, we discovered that activation of LHAGal or LHAGABA neurons robustly enhanced locomotor activity, but that GABA-induced hyperactivity was associated with a striking, compulsive-like behavior not observed with LHAGal neuron-induced hyperactivity.

Thus, our data suggest that LHAGal neurons represent a subpopulation of LHAGABA neurons that promotes food reward independent of direct VTA innervation, while VTA innervating LHAGABA neurons may be required to induce overall chow consumption. Furthermore, we show that targeting exclusively galanin-expressing LHA neurons attenuates compulsive-like locomotor behavior, while maintaining food-seeking behavior. The clear behavioral distinction of molecularly defined subsets of LHAGABA neurons could have important implications for treatment strategies of feeding and psychiatric disorders.

Materials and Methods

Animals

All experimental procedures were approved by the Institutional Animal Care and Use Committee at the Pennington Biomedical Research Center. Mice were bred and housed at 22°C on a 12 h light/dark cycle. Food and water were available ad libitum unless otherwise specified. GalCre mice were generated in house as previously described (Laque et al., 2015) and VgatCre mice were generously provided by Dr. Bradford B. Lowell, Beth Israel Deaconess Medical Center and Harvard Medical School (Vong et al., 2011). Experimental animals were bred and used from homozygous breeding colonies. GalTgGFP/VgatTom mice were generated by crossbreeding GalTgGFP mice [Stock Tg(Gal-EGFP)109Gsat, #0163420UCD, Mutant Mouse Regional Resource Center; Laque et al., 2013] with VgatCre mice and Rosa-tomatof/f mice [B6.Cg-Gt(ROSA)26Sor<tm14(CAG-tdTomato)Hze>/J, Jackson Laboratory, stock #007914] and were used for histology and cell-count experiments as hemizygous GalTgGFP, double heterozygous VgatTom mice.

Genotyping

Genotype for experimental animals was determined by PCR using the following primers: GalCre: cre, forward, 5′-CCT CTC CAC CCA AGC GGC CGG AGA ACC-3′; cre, reverse, 5′-CCG GCT CCG TTC TTT GGT GGC CCC TTC GCG-3′; wild type, forward, 5′-TCC TGA GAC CAT GTC CAC TG-3′; wild type, reverse, 5′-CTG CCA CTC CTG TGA TCT GA-3′; VgatCre: wild type, forward, 5′-CTT CGT CAT CGG CGG CAT TCT G-3′; wild type, reverse, 5′-CAG GGC GAT GTG GAA TAG AAA-3′; cre, reverse, 5′-ATC GAC CGG TAA TGC AGG CAA-3′; GalTgGFP: forward, 5′-TAT ATC ATG GCC GAC AAG CA-3′; reverse, 5′-GAA CTC CAG CAG GAC CAT GT-3′; Rosa-tomatof/f: wild type, forward, 5′-AAG GGA GCT GCA GTG GAG TA-3′; wild type, reverse, 5′-CCG AAA ATC TGT GGG AAG TC-3′: loxP, forward, 5′-GGC ATT AAA GCA GCG TAT CC-3′; loxP, reverse, 5′-CTG TTC CTG TAC GGC ATG G-3′.

Stereotaxic surgery

Viral injections were done as described previously (Rezai-Zadeh et al., 2014); briefly, mice were anesthetized using inhaled isofluorane (2% isofluorane/1 L O2/min) and mounted into a stereotaxic frame (Stereotaxic Alignment System, Kopf Instruments) to enhance anatomically reproducible injections into the LHA (±0.9 mm lateral, −1.35 mm caudal, and −5.4 mm ventral to bregma based on the Paxinos mouse-brain atlas; Paxinos and Franklin, 2004). Either AAV5-hSyn-DIO-hM3Dq-mCherry [DREADD (designer receptor exclusively activated by designer drugs)-Gq, titer 6 × 1012 cfu/ml] or AAV5-hSyn-DIO-mCherry (control, titer 6 × 1012 cfu/ml) were injected bilaterally (200 nl per site at a rate of 20 nl/30 s) into the LHA. Guide cannula and injector remained in place for 5 min following injection to allow for diffusion into the parenchyma and to prevent backflow up the cannula tract. Skull access holes were filled using bone wax (Lukens, #901) and the incision was closed with stainless steel wound clips (CellPoint Scientific, #203-1000) and postoperative care was given (bupivicaine/lidocaine: 2.5–12.5 mg/kg once at incision; carprofen: 5 mg/kg once daily for 72 h; once warm saline, i.p.). Mice recovered for 2–3 weeks to allow sufficient viral expression before further experimental testing.

Two cohorts of virus-expressing mice were generated. Cohort 1 had exclusively male mice (GalCre: n = 4 control, n = 9 DREADD-Gq; VgatCre: n = 4 control, n = 8 DREADD-Gq virus), Cohort 2 had mixed male and female mice (GalCre: control virus, n = 3 males, 3 females; DREADD-Gq virus, n = 4 males, 5 females).

Behavioral experiments

Clozapine-N-oxide dose selection.

Initial repeatable measurements, such as those related to locomotor activity, energy expenditure, and food intake, were performed with multiple doses of the designer drug clozapine-N-oxide (CNO; ligand for the DREADD-Gq receptor) to identify the most efficient CNO dose. Nonrepeatable behavioral experiments (operant conditioning, ensure intake, marble burying) were performed with a single CNO dose (1 mg/kg), which has been also used by several other studies (Jennings et al., 2015; Nakajima et al., 2016). For LHAVgat operant testing experiments, 1 mg/kg CNO induced excessive compulsive behaviors (gnawing) in all animals and interfered with the mouse's ability to perform the operant task. Thus, all LHAVgat mice were tested at multiple CNO doses (0.1–1 mg/kg) to identify individual CNO doses where mice were able to properly perform the operant task. For the marble burying, which can be only performed once, we chose the lowest CNO dose (0.1 mg/kg) tested, to minimize excessive compulsive behavior.

Metabolic parameters.

Animals from cohort 1 were adapted to metabolic chambers (TSE Systems) for 1 week before data were collected for locomotor activity (beam breaks) and oxygen consumption (VO2 ml/kg/h) in 30 min intervals. For stimulation experiments, animals were intraperitoneally injected with PBS or CNO (0.1 mg/kg or 1 mg/kg) at ∼1000 h.

Locomotor activity was also measured with subcutaneously implanted telemetric radio transmitters (G2 E-Mitter, Starr Life Sciences), detected with individual receivers (Series 4000, Starr Life Sciences), and recorded with VitalView software (Starr Life Sciences). Telemetric radio transmitters were surgically implanted on the dorsal surface of each mouse in the interscapular area. Mice recovered for ≥1 week before measurement of locomotor activity via telemetry. Measurements occurred in the home cage and, for neuronal stimulation experiments, mice were intraperitoneally injected with PBS or CNO (0.1 mg/kg or 1 mg/kg, i.p.) at ∼1000 Zeitgeber time and locomotor activity was recorded continuously (every minute) for 24 h.

Motivated food intake.

DREADD-Gq animals from cohort 1 (n = 8 per genotype) were food-restricted during the training phase of the experiment. This greatly facilitates operant conditioning learning by encouraging exploratory behavior and increasing motivation to work for food. Food restriction started 2 weeks before the operant training session with 2.5–3.5 g/d standard laboratory chow (Laboratory Rodent Diet 5001, LabDiet; distributed during the light cycle at 1200 Zeitgeber time) until animals reached 90% of their baseline body weight. Operant conditioning chambers (Med Associates) were ∼15.9 × 14.0 × 12.7 cm, with a house light, wire bar floor, two lighted nose-poke holes with infrared beams, and a pellet dispenser that released reward pellets into a hopper located between the nose-poke holes. Mice were acclimated to sucrose reward pellets in their home cage 2 d before operant training to prevent neophobia. Training occurred in the following four steps with illuminated nose-poke holes marking the “correct” (rewarded) hole: Step 1—autoshaping: one reward per nose poke, both holes are marked correct; Step 2—fixed ratio 1 (FR1): one reward per one correct nose poke, only one hole is marked correct; Step 3—fixed ratio 5 (FR5): one reward per five consecutively correct nose pokes, only one hole is marked correct; Step 4—progressive ratio (PR): one reward per increasing consecutive correct nose pokes on an exponential scale (i.e., 1, 2, 4, 6, 9, 12, 15, 20, 25, 32, 40, 50, 62, 77, 95, 118 …; Richardson and Roberts, 1996). Autoshaping was performed for 1 d in a 30 min session to acclimate the mice to the chambers and reinforce nose-poke behavior; FR1, FR5, and PR training were each performed daily for 5–7 d until mice achieved stable performance (≤10% variation in number of correct nose pokes and a 3:1 ratio of correct/incorrect responses over a 3 d period). Fixed ratio training consisted of 30 min sessions and PR sessions were 60 min or ended when the mouse did not complete any nose poke for a 5 min period (break point). Once training was completed, mice were returned to ad libitum feeding until returning to baseline bodyweight. PR testing occurred during the light cycle at 1200 Zeitgeber time. Stimulation of virally infected neurons was done with intraperitoneal injections of PBS or CNO at doses from 0.05 to 1 mg/kg body weight 1 h before PR testing. Data used for analysis included the total number of correct nose pokes (collectively leading to reward), total number of incorrect nose pokes (not leading to reward), and total number of rewards achieved during the PR testing session.

Palatable food intake.

Mice from cohort 1 had access to a liquid palatable diet (Ensure, 97 kcal/ml) for 1 h daily over 5 d. On day 4 and 5, mice received counterbalanced intraperitoneal injections of either PBS or CNO (0.1 or 1 mg/kg body weight) 1 h before Ensure access; intake (in grams) was measured following each training or testing session.

Chow intake.

Intake (in grams) of standard laboratory chow (Purina 5001) was measured in metabolic cages (TSE Systems) and analyzed as cumulative food intake for 3 and 24 h following intraperitoneal injection of either PBS or CNO (dose range of 0.1–1 mg/kg body weight).

Marble-burying test.

Standard mouse cages were filled with fresh bedding to a depth of 5 cm and 16 marbles (∼1 cm diameter) were evenly spaced across the bedding. VgatCre mice from cohort 1 (n = 4 control, n = 8 DREADD-Gq) and GalCre mice from cohort 2 (n = 6 control, n = 9 DREADD-Gq) were placed in the cage for 30 min and allowed to ambulate freely. Injection of CNO (1 mg/kg in GalCre mice, 0.1 mg/kg in VgatCre mice) was administered 1 h before beginning the test. Please note, the marble-burying test in part reflects the animal's response to a novel environment and is therefore only testable once per animal. Thus, in this test we compared DREADD animals to a control virus group also treated with CNO. A repeated PBS control injection would have confounded the novelty of the marble-burying test. A photograph was taken of the cage before and after the mouse was in the cage and data were analyzed to determine the percentage of marbles buried while the mouse occupied the cage. ImageJ (Abramoff et al., 2004) was used to quantify the visible marble area in each photograph before and after the test. Data were expressed as percentage area buried after testing time to reflect changes in compulsive-like behaviors (Angoa-Perez et al., 2013).

Perfusion and immunohistochemistry

Two hours before perfusion, mice were injected with CNO (0.1 mg/kg or 1 mg/kg, i.p.) for later histochemical confirmation of CNO-induced neural activation as indicated by induction of the early-response gene cFos (Rezai-Zadeh et al., 2014; Yu et al., 2016). Mice were deeply anesthetized and transcardially perfused with ice-cold saline followed by 10% formalin. Brains were harvested, postfixed in 10% formalin for 24 h, and then transferred into 30% sucrose for cryopreservation until sectioning into four representative series (30 μm per section) throughout the brain using a sliding microtome. One (of four) series of brain sections was processed using free-floating immunohistochemistry. Sections were pretreated with 1% H2O2 in methanol, 0.3% glycine, and 0.03% SDS, before blocking in 3% normal donkey serum, followed by incubation in primary antibody, goat anti-c-Fos (1:200; Santa Cruz Biotechnology). Staining was developed with the Vectastain Elite ABC kit (Vector Laboratories) and 3,3′-diaminobenzidine tetrahydrochloride kit (Metal-Enhanced DAB Substrate Kit, Life Technologies) according to the manufacturers' specification. After that, mCherry was further stained with rabbit anti-dsRed (1:1000; Clontech) and detected with fluorescent-labeled secondary antibody (1:200; Alexa Fluor 594 donkey anti-rabbit, Thermo Fisher Scientific). Brains from GalTgGFP/VgatTom mice (n = 4) were stained for GFP using chicken anti-GFP (1:1000; Abcam) and detected with fluorescent-labeled secondary antibody (1:200; Alexa Fluor 488 donkey anti-chicken, Jackson ImmunoResearch Laboratories) and tdTomato was stained for with rabbit anti-dsRed (1:1000; Clontech), which was detected with fluorescent-labeled secondary antibody (1:200; Alexa Fluor 594 donkey anti-rabbit, Thermo Fisher Scientific).

Cell counts

For quantification of colocalization of galanin and Vgat (marker for GABA), we counted the number of galanin neurons colocalized with Vgat. Within the LHA, galanin is concentrated around the fornix [perifornical area (PFA)], and also in the anterior lateral LHA (indicated as LHA). For cell counting, we sampled both of these regions, which are located in two anatomical sections rostrocaudally. Images of LHA and PFA regions were collected using a Leica SP5 confocal microscope with 488 and 594 nm lasers in sequential scanning mode. HyD detector gain was fixed for all samples, z-section thickness was fixed at 2.6 μm per step, and the zoom was fixed at 1.4×. The microscope had a 20×, 0.70 numerical aperture, dry objective. Brightness and contrast were modified with identical settings (Adobe Photoshop, Adobe Systems) between brains to better visualize GFP and tdTomato staining. We first counted the total galanin-GFP-expressing neurons, and performed a second count of galanin-GFP neurons also expressing tdTomato.

Data analysis

Statistical differences were analyzed using SigmaPlot (Systat Software). Analysis included paired and independent t tests, two-way ANOVA, and two-way repeated-measures ANOVA as further indicated in figure legends. For datasets that did not follow a normal distribution, nonparametric analyses were applied using the Mann–Whitney rank sum test. A p value <0.05 was considered statistically significant and denoted as follows: *p < 0.05, **p < 0.01, ***p < 0.001.

Results

Galanin neurons in the LHA are partially GABAergic

The neuropeptide galanin acts mostly inhibitory and galanin-expressing neurons are often GABAergic. To directly address the molecular identity of LHAGal neurons, we generated double reporter mice (GalTgGFP/VgatTom mice, n = 4; 1 male, 3 females) with transgenic GFP expression in galanin neurons (Laque et al., 2013) and conditional tdTomato expression in vesicular GABA transporter (Vgat)-positive neurons (Vong et al., 2011). We found that approximately half of LHA (45 ± 9%) and ∼30 ± 6% of PFA galanin neurons were GABAergic neurons and that GABAergic galanin neurons represent a small fraction of overall LHAGABA neurons (Fig. 1 A–F). Importantly, LHAGABA neurons heavily innervate the VTA (Nieh et al., 2015), while LHAGal neurons do not directly project to the VTA (Laque et al., 2015). Instead, LHAGal neurons heavily innervate LHA orexin neurons and the noradrendergic LC (Goforth et al., 2014; Laque et al., 2015). These results indicate that LHAGal neurons represent a GABAergic subpopulation with a unique molecular identity and previous work also indicates a distinct projection profile that can be distinguished from the larger heterogeneous population of GABAergic neurons in the LHA.

Figure 1.
Figure 1.

Galanin neurons in the LHA coexpress GABA. A, B, Overview of galanin-expressing (green) and Vgat-expressing (red) neurons in the hypothalamus of GalTgGFP/VgatTom mice (scale bar, 200 μm). C, D, Confocal image of the LHA depicting the PFA (C) and the LHA dorsal of the PFA (D) in GalTgGFP/VgatTom mice. Sampled regions are indicated by white boxes in A and B. Galanin-expressing neurons are shown in green (middle) and Vgat-expressing neurons are shown in red (right). The merged image (left) shows colocalization of galanin-expressing and Vgat-expressing neurons (yellow, open arrow), even though several non-Vgat-expressing galanin neurons (green, striped arrow) and non-galanin-expressing Vgat neurons (red, filled arrow) were also present. E, F, Estimates of cell counts for total galaninGFP neurons and double-labeled galaninGFP/tdTomato neurons in the LHA and PFA of GalTgGFP/VgatTom mice expressed as number of cells (E) or as percentage of total galanin neurons (F; n = 4). Mean ± SEM. 3V, Third ventricle; DMH, dorsomedial hypothalamus; ARC, arcuate nucleus; VMH, ventromedial hypothalamus; fx, fornix; PFA, perifornical area; LHA, lateral hypothalamic area; Vgat, vesicular GABA transporter; TdTom, TdTomato reporter gene; Gal, galanin; tgGFP, transgenic green fluorescence protein.

Chemogenetic activation of LHAGal neurons enhances food-seeking behavior

Recent data have highlighted the entire population of LHAGABA neurons as an important regulator of feeding behavior, emphasizing the LHAGABA projections to the VTA (LHAGABA > VTA projection) as particularly important in motivated feeding behaviors. Thus, here we tested whether activation of non-VTA projecting LHAGal neurons would be sufficient to modulate sucrose-seeking behavior.

We first targeted DREADD-Gq or control virus into the LHA of GalCre mice (LHAGal DREADD-Gq), which allowed us to selectively activate LHAGal neurons with systemic CNO injections. Viral expression was visualized via the fluorescent marker mCherry to verify correct targeting throughout the LHA area (Fig. 2A,B). We verified neuronal stimulation capability of DREADD-Gq histologically with CNO-induced cFos expression in mCherry-positive neurons, which was prominent in the PFA (Fig. 2C, left) and extended through the LHA with some cFos-positive neurons in the zona incerta, while mice injected with control virus (LHAGal Control) showed little cFos in mCherry+ neurons (Fig. 2C, right).

Figure 2.
Figure 2.

Activation of LHAGal neurons increases motivated feeding behavior. A, Illustration of viral injection sites to the LHA in GalaninCre mice with AAV5-hSyn-DIO-hM3Dq-mCherry (DREADD-Gq) or AAV5-hSyn-DIO-mCherry (control). B, Representative overview of hM3Dq-mCherry expression in the LHA of GalaninCre mice. C, Representative images of cFos (green) induction in mCherry (red) neurons following CNO (1 mg/kg) in LHAGal DREADD-Gq (left) and LHAGal Control (right) mice. D, Experimental procedure scheme for operant conditioning using fixed ratio (FR) and PR training. E–G, In the operant conditioning PR test, LHAGal DREADD-Gq mice injected with CNO (1 mg/kg, i.p.) significantly increases the total number of sucrose pellet rewards earned (E) and total number of correct nose pokes (F), while the total number of incorrect nose pokes was unchanged (G) compared with PBS-injected LHAGal DREADD-Gq mice. H, Total 1 h intake of palatable fat solution (Ensure) was unchanged in CNO (1 mg/kg)-injected LHAGal DREADD-Gq or LHAGal Control mice. I, Acute (3 h) chow intake was unchanged in CNO (0.3–1 mg/kg, i.p.)-injected LHAGal DREADD-Gq or LHAGal Control mice. Also 24 h chow intake (inset; 1 mg/kg CNO) was unchanged. J, K, In the food zone experiment (J), we confirmed that CNO (1 mg/kg) had no effect on the time spent in the food zone during 3 h after injections (K). E–G, *p < 0.05, n = 8, mean ± SEM, paired t test. H–K, n = 4 control, n = 9 DREADD-Gq, two-way ANOVA or two-way repeated-measures ANOVA for dose–response experiments. 3V, Third ventricle; ME, median eminence; DMH, dorsomedial hypothalamus; ARC, arcuate nucleus; VMH, ventromedial hypothalamus; fx, fornix; LHA, lateral hypothalamic area; FR, fixed ratio; PR, progressive ratio; DREADD, Designer Receptors Exclusively Activated by Designer Drugs; W, water; F, food; Gal, galanin; PBS, phosphate-buffered saline; CNO, clozapine-N-oxide.

To quantify sucrose-seeking behavior, we trained the mice for operant responding with palatable sucrose pellets (Fig. 2D) and to perform increasing numbers of nose pokes to receive a sucrose pellet (PR testing). After stable PR performance, we activated LHAGal neurons with systemic CNO injections (1 mg/kg) and compared their motivated behavior with PBS-injected mice. Activation of LHAGal neurons significantly increased the total number of sucrose rewards earned (Fig. 2E; paired t test, n = 8, t(7) = −2.78, p < 0.05) and the total number of correct nose pokes was likewise increased following CNO injections (Fig. 2F; paired t test, n = 8, t(7) = −2.58, p < 0.05). However incorrect nose pokes (no sucrose reward; Fig. 2G) were not changed, indicating a true increase of motivated behavior (sucrose seeking), rather than increased impulsive responding.

The operant driven PR task focuses on active food seeking and less on food consumption. To contrast food seeking with food consumption, we exposed animals to liquid Ensure (a palatable complete nutrition beverage) for 1 h over 5 d. CNO or PBS injections were administered during the last 2 d in DREADD-Gq-expressing mice and mice that received a control virus injection. However, palatable food consumption remained unchanged by activation of LHAGal neurons (Fig. 2H). This suggests that activation of LHAGal neurons specifically enhances the motivational aspect of food seeking rather than consumption of palatable food.

We further wanted to distinguish feeding of palatable food, which is mainly driven by a hedonic mechanism, from feeding of regular chow, which is mainly driven by homeostatic mechanisms. This was done in metabolic chambers to allow continuous (24 h) recording of food intake following CNO (0.3–1 mg/kg) and PBS injections in control and LHAGal DREADD-Gq mice. CNO-induced activation of LHAGal neurons did not change 3 or 24 h chow intake (Fig. 2I). We also measured the amount of time that the mice spent in proximity to the feeding container (Fig. 2J, Food Zone) over the 3 h following CNO or PBS injections, but the CNO did not affect the time spent in the food zone of the cage (Fig. 2K).

Chemogenetic activation of LHAGal neurons increases locomotor activity and energy expenditure

LHAGABA neurons have also been implicated in the regulation of arousal and wakefulness, and we therefore implanted mice with telemetric radio transmitters to assess the effect of galanin neuron stimulation on locomotor activity. Animals were implanted with telemetric radio transmitters for minute-by-minute measurement of locomotor activity. LHAGal DREADD-Gq and control mice were injected with PBS or CNO (1 mg/kg, i.p.). Indeed, CNO robustly induced locomotor activity that peaked at 1 h postinjection and returned to the baseline activity levels ∼3 h postinjection (Fig. 3A). Cumulative locomotor activity (over the first 3 h postinjection) showed a significant 2.5-fold increase in locomotor activity in CNO-injected LHAGal DREADD-Gq mice, but not in control mice (Fig. 3B; LHAGal Control, n = 4; LHAGal DREADD-Gq, n = 9; ANOVA for the interaction between treatment and group: F(1,22) = 5.071, p < 0.05).

Figure 3.
Figure 3.

Activation of LHAGal neurons increases locomotor activity. A, Locomotor activity before and after injection of PBS or CNO (1 mg/kg) in LHAGal DREADD-Gq mice (n = 9) as measured by telemetry in 1 min increments. B, Cumulative locomotor activity (3 h postinjection, measured by telemetry) was significantly increased following CNO (1 mg/kg) injection in LHAGal DREADD-Gq mice (n = 9), but not in LHAGal Control mice (n = 4; 2-way ANOVA, pairwise comparisons by Holm–Sidak method). C, Locomotor activity as measured by beam breaks was increased in LHAGal DREADD-Gq mice following CNO (1 mg/kg) injection compared with LHAGal Controls. D, Different CNO doses (0.3–1 mg/kg, i.p.) were tested for CNO-induced locomotor activity measured by beam breaks and no reduction in behavioral effects was observed with the tested doses. E, F, Oxygen consumption measurement in metabolic chambers showed only mild effects of CNO (1 mg/kg)-induced locomotor activity on energy expenditure (E), which was significant only at the highest CNO dose tested (1 mg/kg; F). C–F, n = 4 LHAGal Control; n = 9 LHAGal DREADD-Gq, two-way repeated-measures ANOVA, pairwise comparisons by Holm–Sidak method. All data are represented as mean ± SEM, *p < 0.05, **p < 0.01, and ***p < 0.001. LHA, lateral hypothalamic area; DREADD, Designer Receptors Exclusively Activated by Designer Drugs; Gal, galanin; PBS, phosphate-buffered saline; CNO, clozapine-N-oxide.

We also measured locomotor activity by beam breaks in the TSE System, which measures 3D movements including up-and-down rearing, in contrast to the telemetric measurements that exclusively captures 2D movements in the cage. We confirmed an increase in locomotor activity following CNO injection, which returned to PBS baseline levels at 3 h postinjection (Fig. 3C; n = 4 LHAGal Control; n = 9 LHAGal DREADD-Gq; repeated-measures ANOVA for the interaction between group and interval, F(7,77) = 6.494; p < 0.001). Comparing cumulative locomotor activity we found a doubling of locomotor activity in LHAGal DREADD-Gq mice at all tested CNO doses compared with control mice (Fig. 3D; repeated-measures ANOVA for the interaction between group and level of drug, F(3,33) = 4.584; p < 0.01).

We further analyzed VO2 consumption in these mice as locomotor activity drives VO2 consumption as a measure for energy expenditure. CNO consistently increased VO2 mildly compared with PBS-treated mice (Fig. 3E,F), but only reached significance at a CNO dose of 1 mg/kg (Fig. 3F; repeated-measures ANOVA for the interaction between group and level of drug, F(3,33) = 4.413, p < 0.01). This further suggests that increased locomotor activity may also induce locomotion with less effect on oxygen consumption (e.g., head movements).

Thus, activation of LHAGal neurons showed a clear increase in locomotor activity that is in line with increased wakefulness during the light cycle, where nocturnal mice are typically asleep. Indeed, while CNO-injected LHAGal DREADD-Gq animals were awake and engaged in various behaviors, control or PBS-injected animals were often asleep. Notably, activation of LHAGal neurons did not induce compulsive-like behavior, in contrast to the compulsive eating or foraging locomotor behavior induced by activation of LHAGABA neurons as reported by others (Jennings et al., 2015; Venner et al., 2016).

Chemogenetic activation of LHAGABA neurons increases food-seeking and food-consumption behavior

To allow a more precise comparison of LHAGABA neurons to LHAGal neurons, we performed chemogenetic activation of LHAGABA neurons using the exact experimental setup as used for chemogenetic activation of LHAGal neurons.

LHAGABA neurons were targeted by viral injections into VgatCre mice (Vong et al., 2011) using the same stereotaxic coordinates as described for GalCre mice, with correct viral targeting to the LHA validated histochemically (Fig. 4A–C). Functional validation was confirmed by comparing CNO-induced cFos in mCherry+ neurons in LHAVgat DREADD-Gq mice with LHAVgat Control animals. Only LHAVgat DREADD-Gq mice showed robust CNO-induced cFos in mCherry+ neurons, while CNO-injected control mice showed only few cFos+ neurons (Fig. 4B,C).

Figure 4.
Figure 4.

Activation of LHAGABA neurons increases motivated feeding and chow intake. A, Illustration of viral injection sites to the LHA in VgatCre mice with AAV5-hSyn-DIO-hM3Dq-mCherry (DREADD-Gq) or AAV5-hSyn-DIO-mCherry (control). B, Representative overview of hM3Dq-mCherry expression in the LHA of VgatCre mice. C, Representative images of cFos (green) induction in mCherry (red) neurons following CNO (1 mg/kg) in DREADD-Gq (left) and control (right) injected VgatCre mice. D–F, In the operant conditioning PR test, LHAVgat DREADD-Gq mice injected with CNO (0.05–0.2 mg/kg, i.p.) significantly increases the total number of sucrose pellet rewards earned (D) and total number of correct nose pokes (E; CNO dose–response experiment was performed and data are presented as maximal effective CNO dose per individual; green line, 0.05 mg/kg; blue, 0.1 mg/kg; red, 0.2 mg/kg), while the total number of incorrect nose pokes was unchanged (F) compared with PBS-injected LHAVgat DREADD-Gq mice. G, Total 1 h intake of palatable fat solution (Ensure) was unchanged in CNO (0.1 mg/kg)-injected LHAVgat DREADD-Gq or LHAVgat Control mice. H, Acute (3 h) chow intake was increased in CNO (0.1 and 0.3 mg/kg, i.p.)-injected LHAVgat DREADD-Gq compared with LHAVgat Control mice and PBS-injected LHAVgat DREADD-Gq mice. However, 24 h chow intake (inset; 0.1 mg/kg CNO) was unchanged. I, In the food zone experiment, we confirmed that CNO (0.1 mg/kg) had no effect on the time spent in the food zone during 3 h postinjection (K). D–F, *p < 0.05, n = 8, mean ± SEM, paired t test. G–I, n = 4 LHAVgat Control; n = 8 LHAVgat DREADD-Gq; two-way ANOVA or Mann–Whitney rank sum test for dose–response experiments. 3V, Third ventricle; ME, median eminence; DMH, dorsomedial hypothalamus; ARC, arcuate nucleus; VMH, ventromedial hypothalamus; fx, fornix; LHA, lateral hypothalamic area; DREADD, Designer Receptors Exclusively Activated by Designer Drugs; Gal, galanin; PBS, phosphate-buffered saline; CNO, clozapine-N-oxide.

Food-seeking behavior in LHAVgat DREADD-Gq mice was examined using the same operant conditioning schedule as for GalCre mice (Fig. 2D). Initially, when testing operant responding with a CNO dose of 1 mg/kg, all mice exhibited an erratic, compulsive-like locomotor behavior (see below) and failed to perform any significant number of nose pokes (data not shown). In an effort to moderate this compulsive-like locomotor behavior, we tested all animals with decreasing CNO doses and measured PR responding for all animals. In seven of eight mice, we found a CNO dose that increased operant responding. However, the CNO dose sufficient to enhance operant responding and low enough to moderate compulsive locomotor activity was highly variable, possibly due to minute differences in viral expression levels and anatomical locations. To visualize CNO-induced operant behavior, we plotted the maximal CNO-induced operant response for each animal in Figure 4D,E. Note that the individual data in Figure 4D,E were induced by varying CNO doses. For clarity, we have color-coded the individual response data for the used CNO dose in the figure legend. With this analysis we confirm earlier reports (Jennings et al., 2015) and show that activation of LHAGABA neurons increased total number of sucrose rewards earned (Fig. 4E; paired t test, n = 8, t(7) = −2.73, p < 0.05) and correct nose pokes (Fig. 4E; paired t test, n = 8, t(7) = −2.75, p < 0.05), but did not alter the number of incorrect nose pokes (not leading to sucrose pellet; Fig. 4F). Neither CNO dose induced incorrect nose pokes, indicating that the erratic behavior did not induce impulsive responding but rather prevented the animals from engaging in the learned task at all.

Activation of LHAGABA neurons did not change hedonic food consumption in a paradigm with 1 h access to liquid Ensure (Fig. 4G), but chow consumption over 3 h but not 24 h was increased by CNO (Fig. 4H; n = 4 LHAVgat Control; n = 8 LHAVgat DREADD-Gq, for 0.1 mg/kg CNO: U(3,7) = 3.0, T = 13.0, p < 0.05; for 0.3 mg/kg CNO: U(3,7) = 4.0, T = 14.0, p < 0.05; 3 h chow intake at 1 mg failed to reach significance: U(3,7) = 6.0, T = 16.0, p = 0.10). Note, that these data were generated using decreasing doses of CNO (0.1–1 mg/kg, i.p.) as used for LHAGal neuronal activation, again to moderate compulsive locomotor behavior. There was high variability among individual mice in this experiment, with several of the mice demonstrating more robust increases in food intake during the 3 h period following CNO injection. Despite the increase in food intake, there was no increase in time spent in the food zone during the 3 h following CNO injection for the LHAVgat DREADD-Gq mice (Fig. 4I).

Chemogenetic activation of LHAGABA neurons increases locomotor activity more robustly than activation of LHAGal neurons

We further measured locomotor activity following CNO injection in LHAVgat DREADD-Gq and LHAVgat Control mice using telemetry in their home cages and beam breaks in metabolic chambers. We found both measures showed a strong increase in locomotor activity, which continued to be elevated for ∼6 h following injection. Using telemetric monitoring of locomotor activity, LHAVgat DREADD-Gq and LHAVgat Control mice were injected with PBS or CNO (0.1 mg/kg, i.p.). We observed a CNO-induced increase in locomotor activity that peaked at 1 h postinjection and this increase continued until 5 h postinjection before returning to the baseline activity levels (Fig. 5A). Cumulative locomotor activity (over the first 3 h postinjection) was measured during the 3 h following CNO or PBS injection showing a significant 2.5-fold increase in locomotor activity in LHAVgat DREADD-Gq mice, but not in LHAVgat Control mice (Fig. 5B; LHAVgat Control, n = 4; LHAVgat DREADD-Gq, n = 8; ANOVA for the interaction between treatment and group: F(1,20) = 18.663, p < 0.001). These data were similar to data for locomotor activation of LHAGal neurons. In contrast, locomotor activity measured by beam breaks (adding up-and-down movements like rearing) in the TSE System reflected a stronger locomotor behavior after LHAGABA neuron activation compared with activation of LHAGal neurons. Locomotor activity increased quickly following CNO injection and returned to PBS baseline levels at 6 h postinjection (Fig. 5C; LHAVgat Control, n = 4; LHAVgat DREADD-Gq, n = 8; repeated-measures ANOVA for the interaction between group and interval, F(8,80) = 2.771; p < 0.01). Comparing cumulative increases (sum of beam breaks during 3 h following injection) in locomotor activity, LHAGABA neuronal activation caused a fourfold increase in locomotor activity at multiple CNO dosages (Fig. 5D; LHAVgat Control, n = 4; LHAVgat DREADD-Gq, n = 8; repeated-measures ANOVA for the interaction between group and level of drug, F(3,47) = 8.389; p < 0.001), in contrast to the twofold induction after LHAGal neuron activation (Fig. 3D).

Figure 5.
Figure 5.

Activation of LHAGABA neurons robustly increases locomotor activity. A, Locomotor activity before and after injection of PBS or CNO (1 mg/kg) in LHAVgat DREADD-Gq mice (n = 8) as measured by telemetry in 1 min increments. B, Cumulative locomotor activity (3 h postinjection, measured by telemetry) was significantly increased following CNO (0.1 mg/kg) injection in LHAVgat DREADD-Gq mice (n = 8), but not in LHAVgat Control mice (n = 4; 2-way ANOVA, pairwise comparisons by Holm–Sidak method). C, Locomotor activity as measured by beam breaks was increased in LHAVgat DREADD-Gq mice following CNO (1 mg/kg) injection compared with LHAVgat Controls. D, Different CNO doses (0.1–1 mg/kg, i.p.) were tested for CNO-induced locomotor activity measured by beam breaks and no reduction in behavioral effects was observed with the tested doses. E, F, Oxygen consumption measurement in metabolic chambers showed significant effects of CNO (0.1 mg/kg)-induced locomotor activity on energy expenditure (E), which was significant at all CNO doses tested (0.1–1 mg/kg; F). C–F, n = 4 LHAVgat Control; n = 8 LHAVgat DREADD-Gq; two-way repeated-measures ANOVA, pairwise comparisons by Holm–Sidak method. All data are represented as mean ± SEM, *p < 0.05, **p < 0.01, and ***p < 0.001. LHA, lateral hypothalamic area; DREADD, Designer Receptors Exclusively Activated by Designer Drugs; Gal, galanin; PBS, phosphate-buffered saline; CNO, clozapine-N-oxide.

Energy expenditure (VO2) also increased with activation of LHAGABA neurons, which extended ∼4 h following CNO injection (Fig. 5E; LHAVgat Control, n = 4; LHAVgat DREADD-Gq, n = 8; repeated-measures ANOVA for the interaction between group and interval, F(8,80) = 4.428; p < 0.001). Mean VO2 was likewise increased following CNO injection at each dosage of drug tested (0.1–1 mg/kg; Fig. 5F; LHAVgat Control, n = 4; LHAVgat DREADD-Gq, n = 8; repeated-measures ANOVA for the interaction between group and level of drug, F(3,30) = 12.026, p < 0.001). LHAGABA neuronal activation induced a stronger and more prolonged effect on locomotor activity and VO2 than did activation of LHAGal neurons, suggesting that galanin neurons in the LHA represent a molecularly distinct subset of the overall GABA population in the LHA. It may also suggest a role for a different transmitter system that coexpresses with GABA, but is distinct from galanin neurons.

LHAGABA neurons, but not LHAGal neurons, induce compulsive-like behavior

LHAVgat DREADD-Gq mice showed CNO-induced hyperactive compulsive-like behavior (i.e., digging in bedding, gnawing on inedible objects in the cage), and was clearly distinct from CNO-induced locomotor activity in LHAGal DREADD-Gq mice. To further examine the nature of the hyperactive behavior following LHAGABA neuronal activation, we used the marble-burying test. The marble-burying test is a measure of stereotypical behavior and is accepted as an indicator of psychiatric conditions, such as anxiety and obsessive–compulsive disorder. Rodent behavior in the marble-burying test can be manipulated with common pharmacological modulators of these conditions, such as anxiolytics and 5-HT compounds (Deacon, 2006; Angoa-Perez et al., 2013). LHAGal DREADD-Gq and LHAVgat DREADD-Gq mice demonstrated a striking difference in CNO-induced marble-burying behavior. Compared with control mice, activation of LHAGal neurons (CNO dose: 1 mg/kg) resulted in significantly fewer buried marbles (Fig. 6A,C; n = 6 LHAGal Control; n = 9 LHAGal DREADD-Gq; T(13) = 4.126, p < 0.001), while activation of LHAGABA neurons (CNO dose 0.1 mg/kg) resulted in significantly more buried marbles (Fig. 6B,C; n = 4 LHAVgat Control; n = 8 LHAVgat DREADD-Gq; T(10) = −2.273, p < 0.05). These results provide strong support for a physiological distinction between LHAGal and the larger LHAGABA neuron population.

Figure 6.
Figure 6.

Activation of LHAGal neurons reduces compulsive burying. In contrast, LHAGABA activation increases burying. A, Marble-burying behavior following CNO injection (1 mg/kg) in LHAGal DREADD-Gq mice (n = 9) is decreased compared with LHAGal Control mice (n = 6). B, Marble-burying behavior following CNO injection (0.1 mg/kg) LHAVgat DREADD-Gq mice (n = 8) is increased compared with LHAVgat Control mice (n = 4). All data are represented as mean ± SEM, paired t test, *p < 0.05, ***p < 0.001. C, Representative images of marble placement at test start (far left), and at the end of 30 min testing period for LHAGal control (left), LHAGal DREADD-Gq (middle), LHAVgat control (right), and LHAVgat DREADD-Gq (far right). LHA, lateral hypothalamic area; DREADD, Designer Receptors Exclusively Activated by Designer Drugs; Gal, galanin.

Discussion

The LHA modulates food reward and consumption behaviors; however, the exact role of anatomically and molecularly distinct regions of LHA neurons and their contribution to these behaviors still remain unclear. Our data confirm that LHA neurons and specifically LHAGABA neurons modulate food reward and consumption behaviors. In addition, we distinguish the larger population of LHAGABA neurons from LHAGal neurons, which represents a subpopulation of LHAGABA neurons that does not innervate the VTA directly. Overall our data demonstrate that activation of LHAGal neurons is sufficient to enhance the motivation to work for food and thus does not require direct VTA innervation. In contrast, activation of LHAGal neurons did not modulate overall food consumption and thus may require direct VTA interaction. This further suggests that VTA-projecting LHAGABA neurons may mediate the homeostatic food consumption independently of motivational food seeking. While our study did not directly address the possible role of the neuropeptide galanin per se, our study is the first to use genetically targeted approaches to investigate the contribution of LHAGal circuitry to modulate behavioral output.

Modulation of food seeking versus food consumption

Recent studies demonstrated that LHAGABA neurons are potent modulators of feeding (Jennings et al., 2015; Nieh et al., 2015; Barbano et al., 2016) where activation of overall LHAGABA neurons induced food reward and consumption (Jennings et al., 2015), while select activation of LHAGABA > VTA circuits exclusively induced food consumption, but not food reward (Nieh et al., 2015). These data suggested that non-VTA-projecting LHAGABA neurons are responsible for mediating food reward. We show that activation of LHAGal neurons is sufficient to selectively enhance food reward, but not food consumption. Collectively our data provide evidence that non-VTA-projecting LHAGal neurons represent LHAGABA neurons that mediate food-reward behavior and can be contrasted to VTA-projecting LHAGABA neurons that promote food consumption.

It should be noted that LHAGal neurons strongly innervate LHA orexin neurons as well as noradrenergic LC neurons (Laque et al., 2015), which modulate both goal-directed and motivated behavior (Harris et al., 2005; Hofmeister and Sterpenich, 2015). Orexin-mediated food reward involves at least in part projections to the VTA (Peyron et al., 1998; Korotkova et al., 2003; Zheng et al., 2007; Kempadoo et al., 2013). Nevertheless, the dense orexin input to the LC that regulates arousal (Bourgin et al., 2000) could certainly contribute to the overall effect on motivated behavior and food reward and this possibility will need to be addressed in future studies.

We also tested palatable food intake after stimulation of either LHAGal or LHAGABA neurons, but we were unable to confirm increased consumption of palatable, fat-rich Ensure solution with LHAGABA neurons (Jennings et al., 2015). However, due to the large consumption of Ensure (2–3 ml) per test session, we cannot rule out the possibility that this is due to a ceiling effect that prevents detection of increased consumption. Together our data suggest that motivated behavior and overall chow intake is regulated by distinct LHA neuronal populations and that motivated feeding does not require direct interaction with VTA neurons.

LHAGal and LHAGABA neuronal activation both increase locomotor activity, with differing behavioral characteristics

Chemogenetic activation of both LHAGal and LHAGABA neurons robustly increased locomotor activity, which was most pronounced in the LHAGABA DREADD-Gq mice. Recent studies using chemogenetic and optogenetic methods to investigate LHAGABA neurons did not evaluate locomotor behavior (Jennings et al., 2015; Nieh et al., 2015). Locomotor activity may depend on a variety of behavioral changes. While physical activity and exercise is tightly bound to changes in energy expenditure, other locomotor behaviors may rather indicate arousal, grooming, and anxiety, as well as compulsive behaviors. Yet, routine locomotor assessment (beam breaks, telemetry) is not sufficient to understand the cause for changes in locomotor activity.

Indeed, activation of LHAGal or LHAGABA neurons similarly induced locomotor activity, even though we observed visually distinct behavior. Activation of LHAGal neurons increased a range of typical mouse behavior, such as grooming, digging in the bedding, and walking, with frequent transitions from one behavior to the other. In contrast, activation of GABA neurons resulted in repetitive behavior dependent on the environment (e.g., availability of marbles), but was characterized by compulsive maintenance of one behavior over a long time (e.g., digging, gnawing).

We speculate that the difference in locomotor behaviors is also based on the presence of LHA > VTA projections that seem to promote compulsive-like locomotor behavior, while the absence of LHA > VTA projections diminish compulsive locomotor behavior. This hypothesis is further supported by numerous publications linking increased dopamine release in the nucleus accumbens to obsessive–compulsive and repetitive behavior as well as overall locomotor activity (Molloy and Waddington, 1987; White et al., 1988; Colle and Wise, 1991; Berridge et al., 2005; Aliane et al., 2009; Kim et al., 2015). Indeed, chemogenetic activation of GABAergic LHA neurotensin neurons that predominantly act via VTA projections induces dopamine-dependent locomotor activity (Patterson et al., 2015) and also increases food-reward behavior (Leinninger et al., 2011; Kempadoo et al., 2013; Opland et al., 2013), even though compulsivity was not tested in these studies.

Increased locomotor activity may also reflect arousal, as recently suggested by data showing that LHAGABA neurons innervate the ventral medial preoptic area (VLPO; Venner et al., 2016). Future studies will have to determine whether LHAGal > orexin versus LHAGABA > VLPO circuits contribute to different aspects of arousal and locomotor activity and whether this may also differentially affect repetitive compulsive behavior. Several studies mentioned induction of compulsive-like behaviors following chemogenetic or optogenetic activation of LHAGABA or LHAGABA > VTA projections (Nieh et al., 2015; Navarro et al., 2016; Venner et al., 2016), so that an LHA > VTA-mediated compulsive behavior is strongly suggested.

Anatomical considerations and other neurotransmitters

Our data collected from the GalTgGFP/VgatTom mice show that only half of LHAGal neurons coexpress GABA, so that other neurotransmitters and/or neuropeptides may contribute to the observed effects. LHAGal and LHAGABA neurons do not express hypocretin/orexin (Hcrt/Ox) or melanin-concentrating hormone, but do innervate Hcrt/Ox neurons (Jennings et al., 2015; Laque et al., 2015). A portion of LHAGal neurons coexpress neurotensin and cocaine-and-amphetamine-regulated transcript, as well as the long-form leptin receptor (LepRb; Laque et al., 2013), which are all mediators of anorexigenic action.

Our previous work suggested that LepRb-expressing LHAGal neurons mediate food reward via inhibitory inputs to Hcrt/Ox (Laque et al., 2015), either via inhibitory GABA or galanin signaling (Goforth et al., 2014). However, activation of LHAGal neurons induced food reward so that indirect actions of GABAergic LHAGal neurons are likely. Indeed, activation of LHAGal or LHAVgat both seemed to induce cFos in Hcrt/Ox neurons (data not shown), suggesting indirect actions involving disinhibition via interneurons or feedforward mechanisms via effects on the first-order projection sites in other brain regions (e.g., LC or VTA). We further observed strong cFos induction in non-Hcrt/Ox neurons, with some of these likely reflecting primary activated LHAGal or LHAVgat neurons. However, given the extent of cFos-positive non-Hcrt/Ox neurons, we cannot rule out the possibility that other, unidentified LHA populations are equally involved in mediating reward behavior.

Thus, the role of Hcrt/Ox neurons to mediate LHAGal signals requires further investigation. LHAGal neurons, similar to Hcrt/Ox neurons, strongly innervate the LC (Carter et al., 2012; Laque et al., 2015). The Hcrt/Ox > LC axis mediates arousal (Carter et al., 2012) and the LC is generally associated with the modulation of goal-directed behavior (Usher et al., 1999; Bouret and Richmond, 2015). Thus, further studies will have to clarify whether the LC is a relevant mediator of the observed effects in food-reward and locomotor activity.

Recent studies have shown that within the LHAGABA population, discrete groups of GABA neurons are differentially activated during select feeding behaviors (Jennings et al., 2015), suggesting the existence of functionally distinct neuronal populations. Indeed, discrete populations of LHAGABA > VTA terminals respond differentially to varying degrees of optogenetic parameters, such as laser frequency and duration (Barbano et al., 2016). While LHA > VTA projections generally enhanced food reward and consumption, LHAGABA > VTA projection selectively modulated food consumption, indicating that non-VTA-projecting LHAGABA neurons are required for LHAGABA-induced food reward.

Our data indicate that LHAGal neurons represent these non-VTA-projecting LHAGABA neurons that selectively modulate food reward. Furthermore, activation of LHAGABA neurons, but not LHAGal neurons, induced compulsive locomotor activity. We further propose that compulsive locomotor behavior requires LHAGABA innervation of the VTA, which is supported by Nieh and colleagues reporting that activation of LHAGABA > VTA terminals increased maladaptive gnawing behavior (Nieh et al., 2015).

In summary, we show that LHAGal neurons represent a population of LHAGABA neurons that do not directly innervate the VTA and are sufficient to modulate food reward. We propose that the effect of LHAGal stimulation on motivated behavior relies on direct or indirect projections to Hcrt/Ox and/or the LC (Fig. 7). Further, LHAGal neurons induce motivated feeding and locomotor activity without changes in daily caloric intake or compulsive/repetitive behaviors, which could have important implications for treatment strategies for feeding disorders and compulsive behaviors.

Figure 7.
Figure 7.

Proposed circuit for LHAGal and LHAGABA behavioral phenotypes. We propose that LHAGABA/galanin-positive neurons modulate motivated feeding via non-VTA projections such as the LC or Hcrt/Ox population within the LHA. Nongalaninergic LHAGABA neurons modulate consummatory behavior along with compulsive behaviors via VTA projections. LHA, lateral hypothalamic area; ARC, arcuate nucleus; VMH, ventromedial hypothalamus; fx, fornix; opt, optic tract; mt, mammillothalamic tract; VTA, ventral tegmental area; Hcrt/Ox, Hypocretin/orexin; LC, locus coeruleus; Gal, galanin.

Footnotes

  • This work was supported by American Heart Association Grant AHA053298N and National Institutes of Health (NIH) Grants P/F DK020572-30, R01-DK092587 (H.M.), P20-GM103528 (H.M. and S.Y.), and T32-DK064584 (EQC). This work used the facilities of the Cell Biology and Bioimaging Core, supported in part by the Centers of Biomedical Research Excellence (NIH Grant P20-GM103528) and the Clinical Nutritional Research Unit (NIH Grant 1P30-DK072476) center grants from the NIH. Partial support was provided through the Animal Phenotyping Core supported through National Institute of Diabetes and Digestive and Kidney Diseases Nutritional Obesity Research Centers Grant #2P30 DK072476 (“Nutritional Programming: Environmental and Molecular Interactions”) at the Pennington Biomedical Research Center. H.-R.B. was partially supported by NIH Grant R01-DK047348 and C.D.M. was partially supported by R01-DK105032.

  • The authors declare no competing financial interests.

  • Correspondence should be addressed to either Heike Münzberg, PhD, or Emily Qualls-Creekmore, PhD, Pennington Biomedical Research Center, Louisiana State University System, 6400 Perkins Road, Baton Rouge, LA 70808. Heike.Munzberg{at}pbrc.edu or Emily.Qualls-Creekmore{at}pbrc.edu

References

  1. ↵
    2. Abramoff MD,
    3. Magalhaes PJ,
    4. Ram SJ
    (2004) Image Processing with ImageJ. Biophotonics Int 11:36–42.
    OpenUrl
  2. ↵
    2. Aliane V,
    3. Pérez S,
    4. Nieoullon A,
    5. Deniau JM,
    6. Kemel ML
    (2009) Cocaine-induced stereotypy is linked to an imbalance between the medial prefrontal and sensorimotor circuits of the basal ganglia. Eur J Neurosci 30:1269–1279. doi:10.1111/j.1460-9568.2009.06907.x pmid:19769590
    OpenUrlCrossRefPubMed
  3. ↵
    2. Anand BK,
    3. Brobeck JR
    (1951) Hypothalamic control of food intake in rats and cats. Yale J Biol Med 24:123–140. pmid:14901884
    OpenUrlPubMed
  4. ↵
    2. Angoa-Perez M,
    3. Kane MJ,
    4. Briggs DI,
    5. Francescutti DM,
    6. Kuhn DM
    (2013) Marble burying and nestlet shredding as tests of repetitive, compulsive-like behaviors in mice. J Vis Exp 50978. doi:10.3791/50978 pmid:24429507
    OpenUrlCrossRefPubMed
  5. ↵
    2. Barbano MF,
    3. Wang HL,
    4. Morales M,
    5. Wise RA
    (2016) Feeding and reward are differentially induced by activating GABAergic lateral hypothalamic projections to VTA. J Neurosci 36:2975–2985. doi:10.1523/JNEUROSCI.3799-15.2016 pmid:26961951
    OpenUrlAbstract/FREE Full Text
  6. ↵
    2. Berridge KC,
    3. Aldridge JW,
    4. Houchard KR,
    5. Zhuang X
    (2005) Sequential super-stereotypy of an instinctive fixed action pattern in hyper-dopaminergic mutant mice: a model of obsessive compulsive disorder and Tourette's. BMC Biol 3:4. doi:10.1186/1741-7007-3-4 pmid:15710042
    OpenUrlCrossRefPubMed
  7. ↵
    2. Berthoud HR,
    3. Münzberg H
    (2011) The lateral hypothalamus as integrator of metabolic and environmental needs: from electrical self-stimulation to opto-genetics. Physiol Behav 104:29–39. doi:10.1016/j.physbeh.2011.04.051 pmid:21549732
    OpenUrlCrossRefPubMed
  8. ↵
    2. Bonnavion P,
    3. Mickelsen LE,
    4. Fujita A,
    5. de Lecea L,
    6. Jackson AC
    (2016) Hubs and spokes of the lateral hypothalamus: cell types, circuits and behaviour. J Physiol 594:6443–6462. doi:10.1113/JP271946 pmid:27302606
    OpenUrlCrossRefPubMed
  9. ↵
    2. Bouret S,
    3. Richmond BJ
    (2015) Sensitivity of locus ceruleus neurons to reward value for goal-directed actions. J Neurosci 35:4005–4014. doi:10.1523/JNEUROSCI.4553-14.2015 pmid:25740528
    OpenUrlAbstract/FREE Full Text
  10. ↵
    2. Bourgin P,
    3. Huitrón-Résendiz S,
    4. Spier AD,
    5. Fabre V,
    6. Morte B,
    7. Criado JR,
    8. Sutcliffe JG,
    9. Henriksen SJ,
    10. de Lecea L
    (2000) Hypocretin-1 modulates rapid eye movement sleep through activation of locus coeruleus neurons. J Neurosci 20:7760–7765. pmid:11027239
    OpenUrlAbstract/FREE Full Text
  11. ↵
    2. Carter ME,
    3. Brill J,
    4. Bonnavion P,
    5. Huguenard JR,
    6. Huerta R,
    7. de Lecea L
    (2012) Mechanism for hypocretin-mediated sleep-to-wake transitions. Proc Natl Acad Sci U S A 109:E2635–E2644. doi:10.1073/pnas.1202526109 pmid:22955882
    OpenUrlAbstract/FREE Full Text
  12. ↵
    2. Colle LM,
    3. Wise RA
    (1991) Circling induced by intra-accumbens amphetamine injections. Psychopharmacology (Berl) 105:157–161. doi:10.1007/BF02244302 pmid:1796121
    OpenUrlCrossRefPubMed
  13. ↵
    2. Deacon RM
    (2006) Digging and marble burying in mice: simple methods for in vivo identification of biological impacts. Nat Protoc 1:122–124. doi:10.1038/nprot.2006.20 pmid:17406223
    OpenUrlCrossRefPubMed
  14. ↵
    2. Goforth PB,
    3. Leinninger GM,
    4. Patterson CM,
    5. Satin LS,
    6. Myers MG Jr.
    (2014) Leptin acts via lateral hypothalamic area neurotensin neurons to inhibit orexin neurons by multiple GABA-independent mechanisms. J Neurosci 34:11405–11415. doi:10.1523/JNEUROSCI.5167-13.2014 pmid:25143620
    OpenUrlAbstract/FREE Full Text
  15. ↵
    2. Hahn JD,
    3. Swanson LW
    (2010) Distinct patterns of neuronal inputs and outputs of the juxtaparaventricular and suprafornical regions of the lateral hypothalamic area in the male rat. Brain Res Rev 64:14–103. doi:10.1016/j.brainresrev.2010.02.002 pmid:20170674
    OpenUrlCrossRefPubMed
  16. ↵
    2. Harris GC,
    3. Wimmer M,
    4. Aston-Jones G
    (2005) A role for lateral hypothalamic orexin neurons in reward seeking. Nature 437:556–559. doi:10.1038/nature04071 pmid:16100511
    OpenUrlCrossRefPubMed
  17. ↵
    2. Hofmeister J,
    3. Sterpenich V
    (2015) A role for the locus ceruleus in reward processing: encoding behavioral energy required for goal-directed actions. J Neurosci 35:10387–10389. doi:10.1523/JNEUROSCI.1734-15.2015 pmid:26203134
    OpenUrlFREE Full Text
  18. ↵
    2. Jennings JH,
    3. Ung RL,
    4. Resendez SL,
    5. Stamatakis AM,
    6. Taylor JG,
    7. Huang J,
    8. Veleta K,
    9. Kantak PA,
    10. Aita M,
    11. Shilling-Scrivo K,
    12. Ramakrishnan C,
    13. Deisseroth K,
    14. Otte S,
    15. Stuber GD
    (2015) Visualizing hypothalamic network dynamics for appetitive and consummatory behaviors. Cell 160:516–527. doi:10.1016/j.cell.2014.12.026 pmid:25635459
    OpenUrlCrossRefPubMed
  19. ↵
    2. Kempadoo KA,
    3. Tourino C,
    4. Cho SL,
    5. Magnani F,
    6. Leinninger GM,
    7. Stuber GD,
    8. Zhang F,
    9. Myers MG,
    10. Deisseroth K,
    11. de Lecea L,
    12. Bonci A
    (2013) Hypothalamic neurotensin projections promote reward by enhancing glutamate transmission in the VTA. J Neurosci 33:7618–7626. doi:10.1523/JNEUROSCI.2588-12.2013 pmid:23637156
    OpenUrlAbstract/FREE Full Text
  20. ↵
    2. Kim IH,
    3. Rossi MA,
    4. Aryal DK,
    5. Racz B,
    6. Kim N,
    7. Uezu A,
    8. Wang F,
    9. Wetsel WC,
    10. Weinberg RJ,
    11. Yin H,
    12. Soderling SH
    (2015) Spine pruning drives antipsychotic-sensitive locomotion via circuit control of striatal dopamine. Nat Neurosci 18:883–891. doi:10.1038/nn.4015 pmid:25938885
    OpenUrlCrossRefPubMed
  21. ↵
    2. Korotkova TM,
    3. Sergeeva OA,
    4. Eriksson KS,
    5. Haas HL,
    6. Brown RE
    (2003) Excitation of ventral tegmental area dopaminergic and nondopaminergic neurons by orexins/hypocretins. J Neurosci 23:7–11. pmid:12514194
    OpenUrlAbstract/FREE Full Text
  22. ↵
    2. Laque A,
    3. Zhang Y,
    4. Gettys S,
    5. Nguyen TA,
    6. Bui K,
    7. Morrison CD,
    8. Münzberg H
    (2013) Leptin receptor neurons in the mouse hypothalamus are co-localized with the neuropeptide galanin and mediate anorexigenic leptin action. Am J Physiol Endocrinol Metab 304:E999–E1011. doi:10.1152/ajpendo.00643.2012 pmid:23482448
    OpenUrlAbstract/FREE Full Text
  23. ↵
    2. Laque A,
    3. Yu S,
    4. Qualls-Creekmore E,
    5. Gettys S,
    6. Schwartzenburg C,
    7. Bui K,
    8. Rhodes C,
    9. Berthoud HR,
    10. Morrison CD,
    11. Richards BK,
    12. Münzberg H
    (2015) Leptin modulates nutrient reward via inhibitory galanin action on orexin neurons. Mol Metab 4:706–717. doi:10.1016/j.molmet.2015.07.002 pmid:26500842
    OpenUrlCrossRefPubMed
  24. ↵
    2. Leinninger GM,
    3. Jo YH,
    4. Leshan RL,
    5. Louis GW,
    6. Yang H,
    7. Barrera JG,
    8. Wilson H,
    9. Opland DM,
    10. Faouzi MA,
    11. Gong Y,
    12. Jones JC,
    13. Rhodes CJ,
    14. Chua S Jr.,
    15. Diano S,
    16. Horvath TL,
    17. Seeley RJ,
    18. Becker JB,
    19. Münzberg H,
    20. Myers MG Jr.
    (2009) Leptin acts via leptin receptor-expressing lateral hypothalamic neurons to modulate the mesolimbic dopamine system and suppress feeding. Cell Metab 10:89–98. doi:10.1016/j.cmet.2009.06.011 pmid:19656487
    OpenUrlCrossRefPubMed
  25. ↵
    2. Leinninger GM,
    3. Opland DM,
    4. Jo YH,
    5. Faouzi M,
    6. Christensen L,
    7. Cappellucci LA,
    8. Rhodes CJ,
    9. Gnegy ME,
    10. Becker JB,
    11. Pothos EN,
    12. Seasholtz AF,
    13. Thompson RC,
    14. Myers MG Jr.
    (2011) Leptin action via neurotensin neurons controls orexin, the mesolimbic dopamine system and energy balance. Cell Metab 14:313–323. doi:10.1016/j.cmet.2011.06.016 pmid:21907138
    OpenUrlCrossRefPubMed
  26. ↵
    2. Melander T,
    3. Hökfelt T,
    4. Rökaeus A,
    5. Cuello AC,
    6. Oertel WH,
    7. Verhofstad A,
    8. Goldstein M
    (1986) Coexistence of galanin-like immunoreactivity with catecholamines, 5-hydroxytryptamine, GABA and neuropeptides in the rat CNS. J Neurosci 6:3640–3654. pmid:2432203
    OpenUrlAbstract
  27. ↵
    2. Molloy AG,
    3. Waddington JL
    (1987) Pharmacological characterization in the rat of grooming and other behavioural responses to the D1 dopamine receptor agonist R-SK&F 38393. J Psychopharmacol 1:177–183. doi:10.1177/026988118700100304 pmid:22158979
    OpenUrlCrossRefPubMed
  28. ↵
    2. Nakajima K,
    3. Cui Z,
    4. Li C,
    5. Meister J,
    6. Cui Y,
    7. Fu O,
    8. Smith AS,
    9. Jain S,
    10. Lowell BB,
    11. Krashes MJ,
    12. Wess J
    (2016) Gs-coupled GPCR signalling in AgRP neurons triggers sustained increase in food intake. Nat Commun 7:10268. doi:10.1038/ncomms10268 pmid:26743492
    OpenUrlCrossRefPubMed
  29. ↵
    2. Navarro M,
    3. Olney JJ,
    4. Burnham NW,
    5. Mazzone CM,
    6. Lowery-Gionta EG,
    7. Pleil KE,
    8. Kash TL,
    9. Thiele TE
    (2016) Lateral hypothalamus GABAergic neurons modulate consummatory behaviors regardless of the caloric content or biological relevance of the consumed stimuli. Neuropsychopharmacology 41:1505–1512. doi:10.1038/npp.2015.304 pmid:26442599
    OpenUrlCrossRefPubMed
  30. ↵
    2. Nieh EH,
    3. Matthews GA,
    4. Allsop SA,
    5. Presbrey KN,
    6. Leppla CA,
    7. Wichmann R,
    8. Neve R,
    9. Wildes CP,
    10. Tye KM
    (2015) Decoding neural circuits that control compulsive sucrose seeking. Cell 160:528–541. doi:10.1016/j.cell.2015.01.003 pmid:25635460
    OpenUrlCrossRefPubMed
  31. ↵
    2. Olds J,
    3. Milner P
    (1954) Positive reinforcement produced by electrical stimulation of septal area and other regions of rat brain. J Comp Physiol Psychol 47:419–427. doi:10.1037/h0058775 pmid:13233369
    OpenUrlCrossRefPubMed
  32. ↵
    2. Opland D,
    3. Sutton A,
    4. Woodworth H,
    5. Brown J,
    6. Bugescu R,
    7. Garcia A,
    8. Christensen L,
    9. Rhodes C,
    10. Myers M Jr.,
    11. Leinninger G
    (2013) Loss of neurotensin receptor-1 disrupts the control of the mesolimbic dopamine system by leptin and promotes hedonic feeding and obesity. Mol Metab 2:423–434. doi:10.1016/j.molmet.2013.07.008 pmid:24327958
    OpenUrlCrossRefPubMed
  33. ↵
    2. Patterson CM,
    3. Wong JM,
    4. Leinninger GM,
    5. Allison MB,
    6. Mabrouk OS,
    7. Kasper CL,
    8. Gonzalez IE,
    9. Mackenzie A,
    10. Jones JC,
    11. Kennedy RT,
    12. Myers MG Jr.
    (2015) Ventral tegmental area neurotensin signaling links the lateral hypothalamus to locomotor activity and striatal dopamine efflux in male mice. Endocrinology 156:1692–1700. doi:10.1210/en.2014-1986 pmid:25734363
    OpenUrlCrossRefPubMed
  34. ↵
    2. Paxinos G,
    3. Franklin KBJ
    (2004) The mouse brain in stereotaxic coordinates. New York: Elsevier Science.
  35. ↵
    2. Peyron C,
    3. Tighe DK,
    4. van den Pol AN,
    5. de Lecea L,
    6. Heller HC,
    7. Sutcliffe JG,
    8. Kilduff TS
    (1998) Neurons containing hypocretin (orexin) project to multiple neuronal systems. J Neurosci 18:9996–10015. pmid:9822755
    OpenUrlAbstract/FREE Full Text
  36. ↵
    2. Rezai-Zadeh K,
    3. Yu S,
    4. Jiang Y,
    5. Laque A,
    6. Schwartzenburg C,
    7. Morrison CD,
    8. Derbenev AV,
    9. Zsombok A,
    10. Münzberg H
    (2014) Leptin receptor neurons in the dorsomedial hypothalamus are key regulators of energy expenditure and body weight, but not food intake. Mol Metab 3:681–693. doi:10.1016/j.molmet.2014.07.008 pmid:25352997
    OpenUrlCrossRefPubMed
  37. ↵
    2. Richardson NR,
    3. Roberts DC
    (1996) Progressive ratio schedules in drug self-administration studies in rats: a method to evaluate reinforcing efficacy. J Neurosci Methods 66:1–11. doi:10.1016/0165-0270(95)00153-0 pmid:8794935
    OpenUrlCrossRefPubMed
  38. ↵
    2. Sherin JE,
    3. Elmquist JK,
    4. Torrealba F,
    5. Saper CB
    (1998) Innervation of histaminergic tuberomammillary neurons by GABAergic and galaninergic neurons in the ventrolateral preoptic nucleus of the rat. J Neurosci 18:4705–4721. pmid:9614245
    OpenUrlAbstract/FREE Full Text
  39. ↵
    2. Swanson LW,
    3. Sanchez-Watts G,
    4. Watts AG
    (2005) Comparison of melanin-concentrating hormone and hypocretin/orexin mRNA expression patterns in a new parceling scheme of the lateral hypothalamic zone. Neurosci Lett 387:80–84. doi:10.1016/j.neulet.2005.06.066 pmid:16084021
    OpenUrlCrossRefPubMed
  40. ↵
    2. Usher M,
    3. Cohen JD,
    4. Servan-Schreiber D,
    5. Rajkowski J,
    6. Aston-Jones G
    (1999) The role of locus coeruleus in the regulation of cognitive performance. Science 283:549–554. doi:10.1126/science.283.5401.549 pmid:9915705
    OpenUrlAbstract/FREE Full Text
  41. ↵
    2. Venner A,
    3. Anaclet C,
    4. Broadhurst RY,
    5. Saper CB,
    6. Fuller PM
    (2016) A novel population of wake-promoting GABAergic neurons in the ventral lateral hypothalamus. Curr Biol 26:2137–2143. doi:10.1016/j.cub.2016.05.078 pmid:27426511
    OpenUrlCrossRefPubMed
  42. ↵
    2. Vong L,
    3. Ye C,
    4. Yang Z,
    5. Choi B,
    6. Chua S Jr.,
    7. Lowell BB
    (2011) Leptin action on GABAergic neurons prevents obesity and reduces inhibitory tone to POMC neurons. Neuron 71:142–154. doi:10.1016/j.neuron.2011.05.028 pmid:21745644
    OpenUrlCrossRefPubMed
  43. ↵
    2. White FJ,
    3. Bednarz LM,
    4. Wachtel SR,
    5. Hjorth S,
    6. Brooderson RJ
    (1988) Is stimulation of both D1 and D2 receptors necessary for the expression of dopamine-mediated behaviors? Pharmacol Biochem Behav 30:189–193. doi:10.1016/0091-3057(88)90442-X pmid:2902644
    OpenUrlCrossRefPubMed
  44. ↵
    2. Yu S,
    3. Qualls-Creekmore E,
    4. Rezai-Zadeh K,
    5. Jiang Y,
    6. Berthoud HR,
    7. Morrison CD,
    8. Derbenev AV,
    9. Zsombok A,
    10. Münzberg H
    (2016) Glutamatergic preoptic area neurons that express leptin receptors drive temperature-dependent body weight homeostasis. J Neurosci 36:5034–5046. doi:10.1523/JNEUROSCI.0213-16.2016 pmid:27147656
    OpenUrlAbstract/FREE Full Text
  45. ↵
    2. Zheng H,
    3. Patterson LM,
    4. Berthoud HR
    (2007) Orexin signaling in the ventral tegmental area is required for high-fat appetite induced by opioid stimulation of the nucleus accumbens. J Neurosci 27:11075–11082. doi:10.1523/JNEUROSCI.3542-07.2007 pmid:17928449
    OpenUrlAbstract/FREE Full Text
Back to top

In this issue

Email
Print
View Full Page PDF
Citation Tools
Respond to this article
Galanin-Expressing GABA Neurons in the Lateral Hypothalamus Modulate Food Reward and Noncompulsive Locomotion
Emily Qualls-Creekmore, Sangho Yu, Marie Francois, John Hoang, Clara Huesing, Annadora Bruce-Keller, David Burk, Hans-Rudolf Berthoud, Christopher D. Morrison, Heike Münzberg
Journal of Neuroscience 21 June 2017, 37 (25) 6053-6065; DOI: 10.1523/JNEUROSCI.0155-17.2017
Twitter logo Facebook logo Mendeley logo

Jump to section

Keywords

Responses to this article

Respond to this article

Jump to comment:

No eLetters have been published for this article.