Ghrelin Recruits Specific Subsets of Dopamine and GABA Neurons of Different Ventral Tegmental Area Sub-nuclei

Accepted Manuscript Research Article Ghrelin recruits specific subsets of dopamine and gaba neurons of different ventral tegmental area sub-nuclei María Paula Cornejo, Franco Barrile, Pablo Nicolás De Francesco, Enrique Leo Portiansky, Mirta Reynaldo, Mario Perello PII: DOI: Reference:

S0306-4522(18)30632-8 https://doi.org/10.1016/j.neuroscience.2018.09.027 NSC 18655

To appear in:

Neuroscience

Received Date: Revised Date: Accepted Date:

14 June 2018 10 September 2018 19 September 2018

Please cite this article as: M. Paula Cornejo, F. Barrile, P. Nicolás De Francesco, E. Leo Portiansky, M. Reynaldo, M. Perello, Ghrelin recruits specific subsets of dopamine and gaba neurons of different ventral tegmental area subnuclei, Neuroscience (2018), doi: https://doi.org/10.1016/j.neuroscience.2018.09.027

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

GHRELIN RECRUITS SPECIFIC SUBSETS OF DOPAMINE AND GABA NEURONS OF DIFFERENT VENTRAL TEGMENTAL AREA SUB-NUCLEI María Paula Cornejo1, Franco Barrile1, Pablo Nicolás De Francesco1, Enrique Leo Portiansky2, Mirta Reynaldo1, Mario Perello1 1Laboratorio

de Neurofisiología del Instituto Multidisciplinario de Biología Celular, Consejo Nacional

de Investigaciones Científicas y Técnicas de Argentina, Universidad Nacional de La Plata y Comisión de Investigaciones Científicas–Provincia de Buenos Aires, 1900 La Plata, Buenos Aires, Argentina. 2Laboratorio

de Análisis de Imágenes, Facultad de la Ciencias Veterinarias, Universidad Nacional de

La Plata y Consejo Nacional de Investigaciones Científicas y Técnicas de Argentina, 1900 La Plata, Buenos Aires, Argentina Running title: Neuroanatomical aspects of ghrelin action in the VTA Corresponding Author: Dr. Mario Perelló Laboratory of Neurophysiology, Multidisciplinary Institute of Cell Biology Calle 526 S/N entre 10 y 11-PO Box 403 La Plata, Buenos Aires, Argentina 1900 Phone +54 221 4210112 Email: [email protected] or [email protected] ABSTRACT Ghrelin is a stomach-derived hormone that regulates rewarding behaviors and reinforcement by acting on the ventral tegmental area (VTA). The VTA is a complex midbrain structure mainly comprised of dopamine (DA) and gamma-aminobutiric acid (GABA) neurons that are distributed in several VTA sub-nuclei. Here, we investigated the neuroanatomical distribution and chemical phenotype of ghrelin responsive neurons within the VTA. In wild-type mice, we found that: 1) ghrelin binding cells are present in most VTA sub-nuclei but not in its main target, the nucleus accumbens (Acb); 2) systemically-injected ghrelin increases food intake but does neither affect locomotor activity nor the levels of the marker of neuronal activation c-Fos in the VTA sub-nuclei; 3) centrally-injected ghrelin increases food intake, locomotor activity and c-Fos levels in non-DA neurons of all VTA subnuclei; 4) intra-VTA-injected ghrelin increases food intake, locomotor activity and c-Fos levels in nonDA neurons of all VTA sub-nuclei; 5) both centrally- and intra-VTA injected ghrelin increases c-Fos levels in DA neurons of the parabrachial pigmented VTA sub-nucleus. In genetically modified mice in which a subset of GABA neurons expresses the red fluorescent protein tdTomato, we found that centrally-injected ghrelin increases c-Fos levels in GABA neurons of the interfascicular VTA subnucleus. These results suggest that ghrelin can recruit specific subsets of VTA neurons in order to modulate food intake and locomotor activity. Keywords: mesolimbic pathway, locomotor activity, food intake, dopamine, GABA INTRODUCTION

Ghrelin is stomach-derived hormone that acts via its unique known receptor, the growth hormone secretagogue receptor (GHSR), which is highly expressed in the central nervous system (Yanagi et al., 2018). Ghrelin signaling in the brain regulates a variety of functions including growth hormone secretion, food intake, locomotor activity and stress responses, among others (Yanagi et al., 2018). Many evidence showed that ghrelin also regulates mesolimbic circuitries and, as a consequence, various reward-related aspects of eating (Perello et al., 2010). In this regard, the ventral tegmental area (VTA) has emerged as a key target of ghrelin signaling (Liu and Borgland, 2015; Perello and Dickson, 2015). Different studies have documented the presence of GHSR mRNA expression and ghrelin binding sites in the VTA of rats and mice (Abizaid et al., 2006; Cabral et al., 2013; Guan et al., 1997; Zigman et al., 2006; Chuang et al., 2011; Perello M et al., 2012; Mani et al. 2014). In mice, intra-VTA administration of ghrelin acutely increases appetite, alcohol intake, locomotor activity and sexual behaviors (Jerlhag et al., 2006, 2008, 2009; Egecioglu et al., 2010; Lockie et al., 2015; Prieto-Garcia et al., 2015). Ghrelin action in the VTA has been proposed to increase the consumption of palatable food independently of its action in the hypothalamic circuits (Denis et al., 2015). Acute increments of plasma ghrelin levels could act in the VTA, as indicated by studies showing that systemically-injected ghrelin induces locomotor activity, conditioned place preference and accumbal dopamine (DA)-overflow in mice (Jerlhag, 2008; Jerlhag et al., 2010; Perello et al., 2010; Disse et al., 2011; Lockie et al., 2015). In addition, intra-VTA administration of GHSR antagonists block the orexigenic effects of systemically-injected ghrelin in rats (Abizaid et al., 2006). Ghrelin signaling in the VTA seems to be physiologically relevant under some conditions. VTAlesioned rats fully increase food intake in response to centrally administered ghrelin, but they display smaller peanut butter intake and shorter time exploring for rewarding foods (Egecioglu et al., 2010). Also, mice expressing GHSR only in VTA neurons display normal body weight, food intake or locomotor activity but they display increased food intake in response to novelty stress and increased cocaine-induced locomotor activity, as compared to GHSR-knockout mice (Skov et al., 2017). The specific neuronal targets mediating these effects of ghrelin in the VTA are currently uncertain. The VTA is a complex and heterogeneous midbrain area. Based on cytoarchitectonic features, the VTA can be subdivided into five sub-nuclei named paranigral (PN), parabrachial pigmented (PBP), interfascicular (IF), rostral linear (RLi) and caudal linear (CLi) (Phillipson, 1979; Sanchez-Catalan et al., 2014). The VTA, as a whole, mainly innervates the nucleus accumbens (Acb) and olfactory tubercle, but also other brain areas including the striatum, prefrontal cortex and hippocampus (Lammel et al., 2008; Yetnikoff et al., 2014). Importantly, VTA sub-nuclei differentially project to specific Acb subdivisions: the medial portion of the VTA, which comprises the IF and medial parts of the PN and PBP, mainly innervates the medial Acb shell (MAcbSh), while the lateral portion of the VTA, which includes the lateral PN and PBP, mostly innervates the lateral Acb shell (LAcbSh) and Acb core (AcbC) (Sanchez-Catalan et al., 2014). In the Acb, different subdivisions play specific but overlapping functions: the MAcbSh is more linked to the reward processing itself while the LAcbSh and, particularly, the AcbC are more linked to the cognitive processing of motor functions related to

reward and reinforcement (Kelley, 2004; Ikemoto, 2007). The main neuronal type within the VTA is DA neurons, which represent ~60% of all VTA neurons and are particularly enriched in the PBP, the PN, the IF and the CLi (Fu et al., 2012; Sanchez-Catalan et al., 2014). The activation of VTA DA neurons results in increments of DA release in the Acb and behavioral changes, such as increments in locomotor activity (Adamantidis et al., 2011; Boekhoudt et al., 2016). The VTA also contains gamma-aminobutyric acid (GABA) neurons that represent ~35% of all VTA neurons and are interspersed with the DA neurons (Olson and Nestler, 2007; Nair-Roberts et al., 2008). VTA GABA neurons play a major role regulating VTA DA neurons and, as a consequence, can affect locomotor activity (Shank et al., 2007). Some evidences indicate that both VTA DA and GABA neurons are targeted by ghrelin (Abizaid et al., 2006; Zigman et al., 2006). Furthermore, the presence of GHSR exclusively in tyrosine hydroxylase (TH)-expressing neurons, which includes but is not restricted to VTA DA neurons, is sufficient to mediate some of ghrelin actions on food-reward related behaviors (Chuang et al., 2011). Currently, the distribution of ghrelin responsive neurons within the VTA and their chemical phenotype is uncertain. Such information would help to clarify the specific neuronal circuits of the mesolimbic pathway that are recruited by ghrelin and their behavioral implications. The goal of the current study was to gain neuroanatomical insights on the action of ghrelin on the VTA of mice. First, we mapped the distribution of ghrelin binding neurons within the VTA of wild-type mice. Since VTA sub-nuclei display different connectivity and, as a consequence, play specific roles, we hypothesized that combining a careful neuroanatomical analysis of the ghrelin responsive regions of the VTA-to-Acb circuit together with the assessment of some VTA-dependent behaviors could clarify the actions of ghrelin in the VTA. Thus, we analyzed the induction of the marker of neuronal activation c-Fos in both VTA DA and VTA non-DA neurons as well as in neurons of the Acb of wildtype mice that were injected with ghrelin either subcutaneously, centrally or directly into the VTA. In all cases, we assessed the effect of ghrelin on food intake and locomotor activity. In addition, we used transgenic mice, in which a subset of GABA neurons expressing the glutamic acid decarboxylase 2 (gad2) enzyme are fluorescently labeled, to investigate if ghrelin induces c-Fos in VTA GABA neurons. EXPERIMENTAL PROCEDURES Animals. Mice were generated in the animal facility of the IMBICE and included: 1) 113 Wildtype (WT) C57BL/6J mice and 2) 12 Gad2-tdTomato mice, in which tdTomato red fluorescent protein is expressed in GABA Gad2-expressing neurons. Gad2-tdTomato mice were generated by crossing Gad2-CreER mice and Ai14 mice. Gad2-CreER mice [Jackson Laboratory, Gad2tm1(cre/ERT2)Zjh/J; Stock# 010702] express a tamoxifen-inducible Cre recombinase under the endogenous promoter elements of the Gad2 gene (Taniguchi et al., 2011). Ai14 mice [Allen Institute, 129S6Gt(ROSA)26Sortm14(CAGtdTomato)Hze/J;

Stock#

007908]

harbor

a

targeted

mutation

of

the

Gt(ROSA)26Sor locus with a loxP-flanked STOP cassette preventing the transcription of the tdTomato, which is expressed only following Cre-mediated recombination (Madisen et al., 2010). All mice were in a pure C57BL/6J background. In order to induce Cre recombination, 2-month-old Gad2-

tdTomato mice received a daily injection of tamoxifen (70 mg/kg body weight (BW), i.p.) or vehicle (sesame oil) for 4 consecutive days. After 3 weeks, the pattern of red fluorescent cells displayed the distribution of Gad2 expression, as previously shown (Taniguchi et al., 2011). Experiments were performed with adult (8-12-week-old) male mice, which were housed in a 12-h light/dark cycle with regular chow and water available ad libitum, except when indicated. Of note, different sets of mice were used to assess food intake and locomotor activity because successive treatments and/or behavioral assessments can cause carry-over effects on the reactivity of the mesolimbic pathway and/or behavioral sensitization (Steketee and Kalivas, 2011). Also, mice used to assess locomotor activity were not used for neuroanatomical studies because the duration of this assessment (see below) is shorter than the minimum time required to detect the presence of c-Fos protein in the cell nucleus (Hoffman et al., 1994). Experiments were carried out in strict accordance with the recommendations of the Guide for the Care and Use of Laboratory Animals of the National Research Council, USA (Animals, 2011), and all efforts were made to minimize suffering. All protocols received approval from the Institutional Animal Care and Use Committee of the IMBICE (approval ID 10-0112). Drugs. Fluorescein-ghrelin[1-18] (hereafter referred to as F-ghrelin) is an 18 residues analog of the hormone with a fluorescent dye attached through a Lys19 at its C-terminus that behaves similarly to the endogenous ghrelin in terms of GHSR affinity (McGirr et al., 2011). F-ghrelin was provided by Dr. Luyt from the University of Western Ontario (Canada). F-ghrelin and ghrelin (Global Peptide, cat# PI-G-03) were dissolved in isotonic phosphate-buffered saline, pH 7.4 (PBS) and prepared fresh on experimental days. Based on our previous dose-response studies (Chuang et al., 2011; Cabral et al., 2014) a dose of 60 pmol/g BW was chosen for systemic ghrelin administration as this dose induces a reliable increment in food intake and transiently increases plasma ghrelin levels resembling concentrations observed in mice under energy deficit conditions. As we have done in the past (Cabral et al., 2016), a 300 pmol/mouse dose was used for central ghrelin administration as this is the minimum dose that induces the maximal food intake response in our experimental conditions (unpublished observations). A 30 pmol/side dose was chosen for intra-VTA ghrelin administration because this is the minimum dose that increases food intake when microinjected in the ARC (Wren et al., 2001). Ghrelin binding assay. Ghrelin binding sites were labeled using F-ghrelin as described before (Cabral et al., 2013). Briefly, anesthetized WT mice (n=7) were intra-cerebroventricularly (ICV, coordinates antero-posterior: -0.3 mm; medio-lateral: +1.00 mm; dorso-ventral: -2.30 mm) injected into the lateral ventricle with 2 µL of vehicle (artificial cerebrospinal fluid, aCSF) containing 60 pmols of F-ghrelin. After 30 minutes, mice were deeply anesthetized with chloral hydrate (500 mg/kg i.p.) and transcardially perfused. For this purpose, the thoracic field was initially exposed by making incisions through the skin along the thoracic midline and along the base of the ribcage. Then, the breastbone was rostrally cut extending to the level of the clavicles. Afterwards, the diaphragm was separated from the chest wall and the heart was exposed. Then, a 24G needle was inserted in the heart and clamped to the left ventricle. Immediately after, heparinized saline was infused and the right

atrium was cut. Mice were perfused with 20 mL of heparinized saline and then with 50 mL of 4% neutral buffered formalin. Brains of perfused mice were removed, post-fixed 2 h in fixative, immersed overnight in 20% sucrose, frozen and coronally cut at 35 µm into four equal series on a sliding cryostat as previously described (Cabral et al., 2012). One series of coronal sections was used for immunohistochemistry. Briefly, sections were pretreated with 0.5% H2O2 and then treated with blocking solution (3% normal donkey serum and 0.25% Triton X-100). Next, sections were incubated with a goat anti-fluorescein antibody (Molecular Probes, cat# A-11096, 1:1,500) 48 h at 4ºC. Then, sections were incubated with a biotinylated donkey anti-goat antibody (Vector Laboratories, cat# BA5000, 1:1,000), and with reagents of the Vectastain Elite ABC kit (Vector Laboratories, cat# PK-6200), according to manufacturer’s protocols. Finally, sections were incubated with 3-3’-diaminobencidine (DAB)/nickel solution in order to generate a black precipitate in fluorescein immunoreactive (IR) cells. Negative controls were performed using the same procedure but omitting primary or secondary antibodies. Sections were sequentially mounted on glass slides and coverslipped with mounting media. We have validated the specificity and accuracy of this experimental strategy in mice (Cabral et al., 2013). Briefly, we have shown that: a) F-ghrelin is fully bioactive in vivo; b) the pattern of Fghrelin labeling agrees with the reported central distribution of the GHSR mRNA; c) F-ghrelin labeling is reduced by an excess of ghrelin; and d) F-ghrelin fails to label brain nuclei of GHSR-deficient mice. Analysis of the GHSR mRNA in situ hybridization histochemistry (ISHH) dataset from the Allen Institute collection. The distribution of neurons expressing GHSR mRNA was assessed by the analysis

of

the

coronal

ISHH

dataset

available

at

http://mouse.brain-

map.org/experiment/show/74511739 from the Allen Institute collection (Lein et al., 2007). For the VTA, 6 full sized images from the slice subset were used. Images corresponded to 25 µm coronal brain sections that were 100 µm apart from each other. Subcutaneous administration of ghrelin. Initially, WT mice were individually housed for, at least, 3 days prior to experimental procedures. Mice were made accustomed to handling by performing daily subcutaneous injections of saline over the shoulders, into the loose skin over interscapular area. On the experimental day, mice were subcutaneously (SC)-injected with saline alone or containing ghrelin. Injections took no more than 20 seconds and caused no discomfort in mice. Here, 12 mice were used to measure food intake as described below (vehicle: n=7, ghrelin: n=7). Two hours after treatment, mice were perfused and their brains were processed as described above. Brain sections were stored in cryopreservant solution until used for immunostainings. An independent set of 12 WT mice was used exclusively to assess locomotor activity. In this case, each mouse received injections of vehicle and ghrelin in a randomized crossover design in order to reduce the influence of confounding covariates. ICV administration of ghrelin. Mice were first stereotaxically implanted with a single indwelling sterile guide cannula (RWD Life Sciences) into the lateral ventricle using the placement coordinates described above. After surgery, mice were individually housed and allowed to recover for at least 5 days. In these days, mice were made accustomed to handling by removal of the dummy cannula and

connection to an empty cannula connector. On the experimental day, mice were ICV-injected, through a 30-gauge needle, with 2 µL of aCSF alone or containing ghrelin. In particular, mice were manually restrained and dummy cannulas were removed from guide cannulas and replaced by injector cannulas. Injections were slowly performed over the course of a minute in order to avoid rapid pressure changes inside the cranial cavity. Then, the injector cannulas were left in place for 2 minutes in order to prevent back flow of the injected solution. Finally, the injector cannulas were removed and replaced by the dummy cannulas. For neuroanatomical studies, 23 WT mice were ICV-injected and remained with free access to food (vehicle: n=10, ghrelin: n=13), while other 13 WT mice were ICVinjected and had no access to food (vehicle: n=3, ghrelin, n=10). Two hours after treatment, mice were perfused and their brains were processed, as described above. Brain sections were stored in cryopreservant solution until used for immunostainings. In addition, 12 Gad2-tdTomato mice were ICV-injected with aCSF alone or containing ghrelin, as described above (vehicle: n=6, ghrelin: n=6). Two hours after treatment, gad2-tdTomato mice were perfused and their brains were also processed for immunostaining. An independent set of 10 WT mice was exclusively used to assess locomotor activity and received ICV injections of vehicle or ghrelin different days in a randomized crossover design. In all cases, the correct location of ICV injections was verified post-mortem. Intra-VTA administration of ghrelin. WT mice were first stereotaxically implanted with a double indwelling guide cannula (Plastics One) above the VTA. Placement coordinates for intra-VTA guide cannulas were as follows: 1) antero-posterior: -3.4 mm, which is centered in the antero-posterior axis of the VTA that goes from bregma -2.9 mm to bregma -3.9 mm, 2) medio-lateral: ±0.5 mm, which allows the symmetric infusion in the lateral axis of the VTA that laterally goes ±1.0 mm from the midline, and 3) dorso-ventral: -1.2 mm, which allows injections to be made 500 µm dorsal to the VTA since injectors extend 2.8 mm below the guide cannula. The location of the injectors avoids VTA damage and allows injected ghrelin to reach the entire VTA as predicted by a report using a dextran with molecular mass similar to ghrelin injected in similar conditions as used here (Carare et al., 2008). After surgery, mice were individually housed and allowed to recover for at least 5 days. These mice were also made accustomed to handling by removal of the dummy cannula as described above. On the experimental day, mice were intra-VTA injected using a double 30-gauge injector with 0.4 µL of aCSF containing FluoSpheresTM 0.25% (Red Fluorescent, Invitrogen, cat# F8793) with or without ghrelin. Intra-VTA injections were performed similarly as described for ICV injections and also took ~5 minutes per mouse. Two hours after treatment, mice were perfused and their brains were processed as described above. The accuracy of VTA injections was verified post-mortem. The inclusion criteria were: 1) visible bilateral tracts of injectors ~500 µm above the VTA, 2) presence of FluoSpheresTM in the VTA, and 3) absence of c-Fos induction in the ARC, which was considered as an indicator of absence of ghrelin diffusion into the CSF. Initially, 34 mice were implanted with intraVTA cannulas. However, 25 mice remained in this experiment because 9 mice did not fill all the inclusion criteria. In particular, 11 WT mice were included in the analysis of the food intake

experiments (vehicle: n=6, ghrelin: n=5) and 14 mice were included in the analysis of the locomotor activity experiments (vehicle: n=7, ghrelin: n=7). Food intake assessment. Food intake was assessed in ad libitum fed mice that were individually housed with limited amount of bedding in order to easily visualize chunks of pellets. On the experimental day, all food pellets were removed from the home cage hoppers and the bedding was confirmed to be free of chow remains. Then, mice were subjected to the different treatments and exposed to a single pre-weighed chow pellet (~1500 mg) in the floor of the home cages. Mice remained undisturbed for 2-hour, when chow pellets and any additional chow spillage were collected and weighed. Weighing was performed in calibrated scales that had a precision of 1 mg. Two-hour food intake was calculated subtracting the remaining weight of the pellet to the initial weight, and expressed in mg. Locomotor activity assessment. Locomotor activity was registered in recording cages (19×28×13 cm) that were exclusively dedicated to each mouse and placed in a ventilated and acoustically-isolated monitoring box (55×35×90 cm) equipped with an overhead camera and dimmable LED illumination. Locomotor activity was assessed in mice that were individually housed before the experiment and had regular chow available ad libitum, except when transferred to the recording cage. No food was provided in the recording cage because mice could interact with it and alter its overall behavior. Before assessing locomotor activity, each mouse was habituated to the recording environment. In particular, each mouse was transferred from its home cage (19×28×13 cm) to its dedicated recording cage that, in turn, was placed in the monitoring box during 1 h on 2 consecutive days. Autoclaved bedding was added to each recording home cage in the first day of habituation and maintained through the whole experiment. On the experimental day, mice were transferred to its dedicated recording cage and placed in the monitoring box; after 40 min, they were SC, ICV or intra-VTA administered with vehicle or ghrelin. Locomotor activity was registered during 35 min after treatment, with two cages recorded in parallel each time. Videos were imported into Fiji (Schindelin et al., 2012), sampled at 8 frames per second, and processed using custom-made macros to extract total distance traveled for each mouse, which was expressed in meters. Immunostainings. One series of coronal sections was used for double c-Fos and tyrosine hydroxylase (TH) immunostaining which was performed as described before (Valdivia et al., 2015). Briefly, sections were treated with H2O2 and blocking solution, as described above, and then incubated with an anti-c-Fos antibody (Santa Cruz Biotechnology, cat# sc-7202, 1:2,000) for 48 h at 4°C. Next, sections were sequentially incubated with: a biotinylated donkey anti-rabbit antibody (Vector Laboratories, cat# BA-1000, 1:1,000), reagents of the Vectastain Elite ABC kit and a DAB/nickel solution, as described above, in order to generate a black nuclear precipitate in c-Fos-IR cells. Next, sections were incubated with a rabbit anti-TH antibody (Santa Cruz, cat# sc-14007, 1:20,000) for 48 h at 4ºC. Finally, sections were sequentially incubated with the biotinylated donkey anti-rabbit antibody, reagents of the Vectastain Elite ABC kit, and a DAB solution, without nickel, in

order to generate a brown precipitate in the cytoplasm of TH-IR cells. Brain sections of Gad2tdTomato mice were used to perform fluorescent c-Fos immunostaining, as described before (Cabral et al., 2017a) . Briefly, sections were treated with blocking solution and incubated with the above described anti-c-Fos antibody (1:1,000) for 48 h at 4°C. Then, sections were incubated with an Alexa Fluor488 anti-rabbit antibody (Molecular Probes, cat# A-11008; 1:1,000) for 2 h. Brain sections were sequentially mounted on glass slides and coverslipped with mounting media. Quantitative Analysis. Bright-field images were acquired with 10X/0.30 and 60X/0.80 objectives using a Nikon Eclipse 50i and a DS-Ri1 Nikon digital camera with a 0.45X adapter. Fluorescent images were acquired with a 10X/0.45 and a 20X/0.80 objectives in a Zeiss AxioObserver D1 microscope equipped with an Apotome 2 structured illumination module and an Axiocam 506 monochrome camera. Alexa 488 was excited at 450-490 nm and detected in the 500-550 nm range, while tdTomato was excited at 533-558 nm and detected in the 570-640 nm range. All images were taken in comparable areas and under the same optical and light conditions. Blind quantitative analyses were performed independently by two observers in one out of four complete series of coronal sections of each mouse containing the hypothalamic arcuate nucleus (ARC, between bregma -1.22 and -1.94 mm), the Acb (between bregma 1.34 and 0.86 mm) and the VTA (between bregma 3.08 and -3.80 mm). For the analysis, the Acb was subdivided into MAcbSh, LAcbSh and AcbC while the VTA was subdivided into PN, PBP, IF, CLi and RLi sub-nuclei, according to previous descriptions (Kelley, 2004; Ikemoto, 2007). Neuroanatomical limits of each brain region were identified using a mouse brain atlas (Paxinos and Franklin, 2001). In brain sections immunostained against c-Fos and TH, TH-IR signal was used to better define the Acb subdivisions and VTA sub-nuclei. The total number of IR cells in each area was estimated using the Konigsmark’s formula, where the total cells are equal to the number of cells counted multiplied by the total number of sections through the nucleus and divided by the number of sections in which cells were counted (Konigsmark, 1970). All data were corrected for double counting, according to the method of Abercrombie (Abercrombie, 1946) where the ratio of the actual number of neurons or cell nuclei to the observed number is represented by T/(T+h) where T= section thickness, and h= the mean diameter of the neuron or cell nuclei along the z-axis. The mean diameter of the neurons or cell nuclei was determined using Fiji. The number of fluorescein-IR cells was estimated in each VTA sub-nucleus. The number of c-Fos-IR cells in the VTA and in the Acb was also estimated in each sub-nucleus while the number of c-Fos-IR cells in the ARC was estimated per section. For double TH and c-Fos immunostainings, all TH-IR cells positive and negative for c-Fos were counted in each VTA sub-nucleus, and results were expressed as the total number of double c-Fos-IR/TH-IR cells or as percentages representing TH-IR neurons positive for cFos compared to the total number of either TH-IR or c-Fos-IR neurons in each VTA sub-nucleus. For c-Fos immunostainings in Gad2-tdTomato mice, red fluorescent neurons were referred as tdTomato cells and tdTomato cells positive for c-Fos were referred as c-Fos-IR/tdTomato cells. Here, all tdTomato cells positive and negative for c-Fos were counted in each VTA sub-nucleus, and results

were expressed as a percentage that represents tdTomato neurons positive for c-Fos compared to the total number of tdTomato neurons in each VTA sub-nucleus. Statistical analyses. For ICV experiments, neuroanatomical quantitative data were grouped as follows: 1. Ghrelin+food, which includes data of mice ICV-injected with ghrelin and free access to food; 2. Ghrelin-food, which includes data of mice ICV-injected with ghrelin and without access to food; and 3. Vehicle, which pools data of vehicle-injected mice with and without access to food. Data normality was tested using Shapiro-Wilk test. When a normal distribution was found, data were expressed as mean±SEM and compared with Student’s paired or unpaired t tests or regular one-way ANOVA, depending on the number of groups or experimental design. When a normal distribution was not found, data were expressed as median and interquartile range and compared with the nonparametric Mann-Whitney or Kruskal-Wallis test. The tests used for each comparison were indicated in the figure legends. Effect sizes were reported for all statistically significant results. The number of mice used in all experiments was larger than the number theoretically required to adequately test the null hypotheses with a minimum effect size of η2p=0.4 (for ANOVA) or Cohen’s d=0.8 (for t-test), which are the minimum recommended values to detect large effects (Lakens, 2013). Statistical analysis was performed using GraphPad Prism 6.0 and G*Power software was used to calculate the effect sizes. Differences were considered statistically significant when p<0.05. RESULTS 1) F-ghrelin-labeled neurons are distributed in different VTA sub-nuclei. We first performed a detailed analysis of the presence of ghrelin binding neurons in the different sub-nuclei of the VTA and the Acb. The neuroanatomical analysis of F-ghrelin labeling was performed in the VTA and the Acb as well as in the ARC, where GHSR is highly expressed and represents an internal positive control of the assay. Examination of the brain slices of F-ghrelin-injected mice in the fluorescence microscope revealed the presence of clearly labeled cell bodies mainly in the ARC, as previously shown (Cabral et al., 2013). In order to improve the visualization of F-ghrelin labeling in the VTA and Acb, we performed a chromogenic immunostaining against fluorescein that provides higher sensitivity due to the enzymatic amplification steps of the reaction. Under these conditions, fluorescein-IR cell bodies were found in the ARC and the VTA while no signal was detected in the Acb (Figure 1, column 2). The number of fluorescein-IR cells in the ARC was estimated in ~1650, and they were relatively homogeneous in shape, small sized (~7 µm), dark stained and enriched in the ventromedial ARC (Figure 1 A2 and G). Fluorescein-IR cells were detected in the VTA of 5 out of 7 F-ghrelin-injected mice. As expected based on the observed fluorescent signal, signal intensity of the fluorescein-IR cells was weaker in the VTA as compared to the ARC (Figure 1 G-J). The number of fluorescein-IR cells in the VTA was estimated in ~520, and they were particularly enriched in the IF (~70 %) but also present in the PBP (~20 %) and the PN (~10 %). No fluorescein-IR cells were found in the CLi and the RLi. Interestingly, fluorescein-IR cells of each VTA sub-nucleus displayed different cytoarchitectonic features. Fluorescein-IR cells in the IF were homogeneous, round shaped, small sized (~9 µm) and densely packed (Figure 1 H). Fluorescein-IR cells in the PBP and the PN were

more heterogeneous in size (ranging between 7 and 12 µm) and spindle-shaped with no uniform orientation (Figure 1 I-J). Similarly as indicated by the binding assay, the analysis of the in situ hybridization immunohistochemistry data reported by the Allen Mouse Brain Atlas (Lein et al., 2007) revealed that GHSR-expressing neurons are enriched in the ARC, present in the RLi, the IF, the PN and the PBP and absent in the Acb (Figure 1, column 3). 2) SC-injected ghrelin does not affect locomotor activity nor the number of c-Fos-IR cells in the VTA and the Acb. To test if systemic ghrelin acts in the VTA, we investigated the effects of SC-injected ghrelin. SCinjected ghrelin induced a ~2.5-fold increase of 2-h food intake in ad libitum-fed mice (Figure 2A), and did not affect locomotor activity in mice that had no access to food (Figure 2B). As we previously shown (Cabral et al., 2014), SC-injected ghrelin significantly increased the number of c-Fos-IR cells in the ARC, as compared to the number found in the ARC of vehicle-treated mice (not shown). In order to test if ghrelin acts on the VTA, we quantified c-Fos in the VTA sub-nuclei containing fluorescein-IR cells as well as in the Acb, a key VTA target. The quantitative analysis was performed in brain sections immunostained against both c-Fos and TH since the observation of TH-IR signal in the VTA allows not only to estimate the effect of ghrelin on the DA neurons but also to better define the boundaries of VTA sub-nuclei because the cytoarchitectonic features of DA neurons in each sub-nucleus are different (Figure 2C). In particular, PN TH-IR cells are homogeneous, medium sized (~10 µm), medium-to-dark stained and semi-laterally oriented, PBP TH-IR cells are large (~13 µm), medium-to-low stained and with no uniform orientation, and IF TH-IR cells are small (~9 µm), mediumto-dark stained and densely packed along the midline (Figure 2C). In the Acb, the distribution of TH-IR fibers helps to better delineate the boundaries of its subdivisions (Figure 2D). We found that SC-injected ghrelin did not affect the number of c-Fos-IR cells neither in the VTA nor in the Acb, as compared to the numbers found in vehicle-treated mice. 3) ICV-injected ghrelin increases locomotor activity as well as the number of c-Fos-IR cells in the VTA and the Acb. To test if central ghrelin acts in the VTA, we investigated the effects of ICV-injected ghrelin. ICVinjected ghrelin induced a ~6.5-fold increase of 2-h food intake in ad libitum-fed mice (Figure 3A) and induced a ~4-fold increase of locomotor activity in mice that had no access to food (Figure 3B). Locomotor activity in mice receiving central treatment with vehicle was significantly reduced as compared to locomotor activity in mice SC-injected with vehicle likely as a consequence of the differences inherent to the procedure required for each type of injection. Since food intake itself could affect the number c-FosIR cells in some brain areas, the quantitative analysis of c-Fos-IR cells was performed in brain samples of ghrelin-injected mice with or without access to food. ICV-injected ghrelin similarly increased the number of cFos-IR cells in the ARC of mice with or without access to food, as compared to vehicle treatment (not shown). As compared to vehicle treatment (Figure 3C), ICV-injected ghrelin induced a ~3.5-, ~5- and ~9-fold increase of the number of c-Fos-IR cells in the IF (Figure 3D), the PN (Figure 3E) and the PBP (Figure 3F), respectively, of mice with or without access to food. ICV-injected ghrelin induced a ~6-fold increase of the number of double c-Fos-IR/TH-IR cells, as compared to vehicle treatment, exclusively in the PBP of mice without access to food (32±7 vs. 5±4 cells, respectively; H(3,36)=14.05, p=0.0009, η2p:0.3847); these double c-Fos-IR/TH-IR cells in the PBP represented 5.0±1.1% of the total PBP c-Fos-IR cells. As compared to vehicle treatment (Figure 3G),

ICV-injected ghrelin induced a ~4.5-fold increase of the number of c-Fos-IR cells in the AcbC (Figure 3H) of mice without access to food as well as a ~3-fold increase of the number of c-Fos-IR cells in the MAcbSh (Figure 3I) and a ~5-fold increase of the number of c-Fos-IR cells in the LAcbSh (Figure 3J) of mice with or without access to food. 4) Intra-VTA-injected ghrelin increases locomotor activity as well as the number of c-Fos-IR cells in the VTA and the Acb. To test if ghrelin can directly act in the VTA, we administered the hormone into this brain area (Figura 4A). Intra-VTA-injected ghrelin induced a ~3.5-fold increase of 2-h food intake in ad libitumfed mice (Figure 4B) and induced a ~2-fold increase of locomotor activity in mice that had no access to food (Figure 4C). Intra-VTA-administered ghrelin did not affect the number of c-Fos-IR cells in the ARC, as compared to vehicle-treated mice (11±3 vs. 14±4 c-Fos-IR cells/section, respectively). As compared to vehicle treatment (Figure 4D), intra-VTA-injected ghrelin induced a ~5.5-, ~10- and ~6-fold increase of the number of c-Fos-IR cells in the IF (Figure 4E), the PN (Figure 4F) and the PBP (Figure 4G), respectively. Here, intra-VTA-injected ghrelin increased the number of double c-Fos-IR/TH-IR cells, as compared to vehicle treatment, exclusively in the PBP sub-nucleus (74±31 vs. 16±10 cells, respectively; U(6,5)=4.000, p=0.0238, Cohen’s d=1.3245); double c-Fos-IR/TH-IR cells in the PBP of ghrelin-treated mice represented 4.3±2.1% of the total PBP TH-IR cells and 9.9±3.5% of the total PBP c-Fos-IR cells. In the Acb (Figure 4H), intra-VTAinjected ghrelin induced a ~3.5-, ~3- and ~2-fold increase of the number of c-Fos-IR cells in the AcbC (Figure 4I), the MAcbSh (Figure 4J) and the LAcbSh (Figure 4K), respectively, as compared to the numbers found in vehicle treatment. 5) ICV-injected ghrelin increases c-Fos levels in the IF GABA neurons. Since ghrelin increased c-FosIR signal mainly in non-DA neurons, we tested if it affects c-Fos levels in tdTomato neurons of the VTA of the Gad2-tdTomato mice. The number of tdTomato neurons in the PN, IF and PBP sub-nuclei of Gad2-tdTomato mice was estimated in ~200, ~500 and ~650 respectively (Figure 5A). Interestingly, cytoarchitectonic features of the tdTomato neurons in each region of the VTA were slightly different: PN tdTomato cells were mostly small sized (~9 µm) and semi-laterally oriented, PBP tdTomato cells were mostly medium sized (~11 µm) and with no uniform orientation, and IF tdTomato cells were small (~9 µm) and densely packed along the midline (Figure 5A). As compared to vehicle treatment, ICV-injected ghrelin increased the number of c-Fos-IR cells in the VTA of the Gad2-tdTomato mice similarly as seen in WT mice (not shown). ICV-injected ghrelin increased the fraction of c-Fos-IR/tdTomato cells, as compared to vehicle treatment, exclusively in the IF subnucleus (16.5±4.6 vs. 1.0±1.0 % of the total IF tdTomato cells, respectively; U(6,6)=3.000, p=0.0076, Cohen’s d=2.0868, Figure 5B). ICV-injected ghrelin did not affect the fraction of c-Fos-IR/tdTomato cells neither in the PN (11.4±5.2 vs 2.0±2.0 % of the total tdTomato cells, respectively; U(6,6)=10.50, p=0.0909) nor in the PBP (7.0±5.3 vs 0.0±0.0 % of the total tdTomato cells, respectively, U(6,6)=9.000, p=0.0909) of Gad2tdTomato mice compared to the fraction found in vehicle-treated mice. DISCUSSION The mesolimbic pathway connects the VTA and the Acb and regulates cognitive processes including incentive salience, motivation, reinforcement learning and goal-seeking behaviors, among others. The VTA contains different sub-nuclei that preferentially innervate and activate some Acb

subdivisions that, in turn, could regulate different behaviors. Ghrelin signaling is known to target the mesolimbic pathway and modulate several reward-related aspects of eating (Perello and Dickson, 2015). However, the specific components of the VTA-to-Acb circuit regulated by ghrelin are currently uncertain. Here, we analyzed the distribution of ghrelin binding neurons within the VTA and the distribution of ghrelin-induced c-Fos within the VTA and the Acb in order to gain neuroanatomical insights on the action of ghrelin on the mesolimbic circuit. Current binding assay as well as the data available at the Allen Mouse Brain Atlas for GHSR ISHH confirmed that the mouse VTA expresses GHSR while the Acb lacks GHSR-expressing neurons, similarly as already showed by most previous studies that had used either ISHH, real time PCR, binding strategies or transgenic mouse models in which GHSR-expressing neurons are fluorescent (Zigman et al., 2006; Chuang et al., 2011; Perello M et al., 2012; Cabral et al., 2013; Mani et al., 2014, 2017). Thus, previous and current data seem to suggest that a major target of ghrelin within the mesolimbic pathway is the VTA. We found that ghrelin binding neurons are enriched in the IF, and that ICV- and intra-VTA-injected ghrelin increase c-Fos levels in the IF as well as in the MAcbSh. Thus, ghrelin signaling in the VTA seems to predominantly regulate reward-related processes. We also found that ghrelin binding neurons are present in the PN and the PBP, and that ICV- and intra-VTA-injected ghrelin increase c-Fos levels in these VTA subnuclei. In addition, both ICV-injected ghrelin in the absence of food and intra-VTA-injected ghrelin increase c-Fos levels in the AcbC. Thus, ghrelin signaling in the VTA could also affect motor functions aimed to obtain food. In line with this possibility, we found that ICV- and intra-VTA injected ghrelin increase locomotor activity. Notably, it has been reported that intra-Acb-injected ghrelin increases peanut butter intake in mice (Prieto-Garcia et al., 2015). Since presynaptic mechanisms are known to mediate the effect of ghrelin on other neuronal populations, such as the hypophysiotropic corticotrophin-releasing factor producing neurons (Cabral et al., 2012, 2016), it is possible that ghrelin could also locally activate Acb neurons by acting on GHSR present in presynaptic terminals innervating them. Further studies are required to test this possibility. DA neurons are the predominant neuronal type within the VTA followed by GABA neurons. Here, we found that only a subset of PBP DA neurons increase the level of c-Fos in response to ICVor intra-VTA-injected ghrelin, while DA neurons of the IF and the PN do not increase c-Fos levels under these experimental conditions. Previous studies have shown that DA neurons of an undefined VTA sub-nucleus express the GHSR and increase the frequency of action potentials in response to ghrelin (Abizaid et al., 2006; Zigman et al., 2006). In addition, intra-VTA ghrelin administration induces Acb DA release in mice and rats (Jerlhag et al., 2006, 2008; Kawahara et al., 2009). Current observation that intra-VTA-injected ghrelin increases c-Fos levels in the Acb could be another indication that ghrelin acts on VTA DA neurons since D1 dopamine receptor activation in the Acb increases c-Fos levels in this brain region (Graybiel et al., 1990). The fraction of DA neurons of the PBP that display an increase of c-Fos levels in response to ghrelin was relatively small; thus, it is possible that ghrelin actions on VTA DA neurons mainly affects electrical activity or neurotransmitter release, independently of c-Fos gene transcription (Hoffman and Lyo, 2002). In line with this

possibility, early studies showed that activation of the mesolimbic pathway can induce robust DAdependent increments of c-Fos levels in striatal regions without significant increments of c-Fos in VTA DA neurons (Kiba and Jayaraman, 1994; Hunt and McGregor, 1998). We found that non-DA neurons of the VTA showed a robust increase of c-Fos levels in response to ICV- and intra-VTA-injected ghrelin. The use of the Gad2-tdTomato mice, in which GABA Gad2-expressing neurons are red fluorescent, allowed us to identify that ghrelin increases c-Fos levels in a subset of IF GABA neurons. Previous studies showed that VTA GABA neurons also express GHSR and that ghrelin activates VTA DA neurons, in part, via inhibition of the local GABA tone (Abizaid et al., 2006). VTA GABA neurons includes inhibitory interneurons that are regulators of intra-VTA neuronal circuits as well as GABA neurons that project to and regulate other brain nuclei targets (Yetnikoff et al., 2014; Morales and Margolis, 2017). Further studies are needed to reveal the role of ghrelin responsive VTA GABA neurons. Notably, we previously showed that other subsets of GABA neurons express GHSR and mediate specific actions of ghrelin (Cabral et al., 2016, 2017a; Cornejo et al., 2018). Thus, GABA neurons seem to represent a key neuronal population by which ghrelin regulates different physiological functions. Importantly, GABA neurons are specifically and efficiently labeled in Gad2Cre mice (Taniguchi et al., 2011). However, the number of GABA cells observed in the VTA may be underestimated in this mouse model since some VTA GABA neurons may express Gad1, the other isoform of the GABA synthetizing enzyme (Margolis et al., 2012). Thus, the use of traditional labeling studies using ISHH for different markers of GABA neurons may provide a different estimate of the actual number and distribution of ghrelin responsive GABA cells of the VTA. We found here that administration of ghrelin in the VTA increases both food intake and locomotor activity in mice, as it has been previously observed using higher doses of intra-VTA-injected ghrelin in similar experimental conditions (Jerlhag et al., 2006, 2008; Egecioglu et al., 2010; Lockie et al., 2015). Here, each VTA side was injected with 30 pmol of ghrelin, which is the minimum amount of hormone that increases food intake when microinjected in the ARC (Wren et al., 2001), and the criteria to include mice as hits within the intra-VTA experimental group was very strict. However, we cannot rule out that a fraction of injected ghrelin diffused to nearby areas, such as the substantia nigra that is located in close apposition to the VTA (Ryczko and Dubuc, 2017; Suda et al., 2018). Notably, we found that intra-VTA injected ghrelin induced a smaller increment of food intake and locomotor activity as compared to the effect of 300 pmol of ICV-injected ghrelin, which induces the maximal food intake response in our experimental conditions (unpublished observations). Although we did not test if higher doses of intra-VTA-injected ghrelin can induce higher effects, it is likely that the orexigenic effect of ICV-injected ghrelin involves not only the VTA but also others brain targets, including the ARC which is a major target of the orexigenic effects of ghrelin (Cabral et al., 2014). Similarly, the effect of ICVinjected ghrelin on locomotor activity likely involves additional targets, such as the laterodorsal tegmental nucleus or the substantia nigra, which can also mediate ghrelin´s effects on locomotor activity (Jerlhag et al., 2006, 2008; Suda et al., 2018). Interestingly, ICV-injected ghrelin fully increases locomotor activity in mice lacking AgRP neurons, which is the main ghrelin-responsive

neuronal type within the ARC, suggesting that the neuronal circuits mediating ghrelin effects on locomotor activity and food intake can be dissociated (Denis et al., 2015). The understanding of the precise neuronal circuits recruited by the action of ghrelin in the VTA in order to increase food intake and locomotor activity requires further studies. We found that SC-injected ghrelin increases food intake of mice but does neither increase cFos levels in the VTA nor affect locomotor activity. Such observation agrees with our previous studies showing that SC-administered F-ghrelin does not reach the VTA (Cabral et al., 2014). In contrast, some previous studies did detect that systemically injected ghrelin increases locomotor activity in mice (Jerlhag, 2008; Jerlhag et al., 2010; Disse et al., 2011). In addition, systemically-injected ghrelin has been shown to affect other behaviors linked to the mesolimbic pathway in mice, such as conditioned place preference for palatable diets, intake of palatable foods and DA release in the Acb (Jerlhag, 2008; Disse et al., 2010; Jerlhag et al., 2010; Perello et al., 2010; Denis et al., 2015; Lockie et al., 2015). Notably, these referred studies have tested higher doses of ghrelin that, in some cases, were administered via intraperitoneal injections, which have a faster absorption rate as compared to SC administration (Hedrich and Bullock, 2004). Since the amount of ghrelin that gains access to the brain depends on its plasma concentration (Cabral et al., 2014), it is reasonable to hypothesize that the experimental procedures that induce larger increments of plasma ghrelin more likely unmask effects of the hormone that depend on its action on deeper brain areas. It is interesting to note that ghrelin is not synthetized in the brain (Cabral et al., 2017b); thus, the physiological relevance of GHSR present in areas that are not easily accessible to circulating factors, such as the VTA, is uncertain. GHSR-expressing neurons of the VTA likely respond to ghrelin present in the cerebrospinal fluid (Cabral et al., 2013). In addition, GHSR displays a high constitutive activity, in the absence of ghrelin (Holst et al., 2003), and is able to heterodimerize with other G protein coupled receptors in order to modulate their signaling (Schellekens et al., 2015). Thus, constitutive GHSR activity and/or GHSR interactions with other receptors in the VTA may also have impact on the activity of the mesolimbic pathway under particular conditions, which remain to be determined. In summary, we provided here the first neuroanatomical analysis of the distribution of ghrelin responsive cells within the mouse VTA. We found that SC-injected ghrelin did not increase c-Fos in the VTA or the Acb nor locomotor activity in mice suggesting that transient elevations of plasma ghrelin, such as those induced in our experimental conditions, do not reach the VTA. Thus, it would be interesting to investigate the impact on the mesolimbic pathway of controlled plasma ghrelin elevations that mimic changes in hormone concentration observed in animals under ordinary energy deficit circumstances, such as a sporadic fasting. Here, we not only confirmed that ICV- and intraVTA-injected ghrelin acts in the VTA but also provide important neuroanatomical insights of this action. In particular, we found that ghrelin activates different types of VTA neurons, which are located in specific sub-nuclei. Ghrelin action in the VTA may directly recruit DA neurons or indirectly regulate them via modulation of local GABA neurons, and it can be hypothesized that segregated neuronal circuits within the VTA sub-nuclei could independently mediate the effect of ghrelin on food intake or locomotor activity. The

use of GHSR-conditional knockout mice in which GHSR is expressed exclusively in a single VTA neuronal subpopulation or a single VTA sub-nucleus using the Cre-loxP technology may help to clarify the intricacies of ghrelin action on the VTA. FIGURE LEGENDS Figure 1. Panels A-F show images of coronal brain sections of the mouse ARC (A), VTA (B to E) and Acb (F). Column 1 displays images of each brain area, with the corresponding subdivisions, for the indicated bregma level according to the Mouse Brain Atlas (Paxinos and Franklin, 2001). Column 2 shows representative photomicrographs of brain slices of mice ICV-injected with F-ghrelin and subjected to IHC against fluorescein. Arrows point to fluorescein-labeled neurons. Column 3 displays images of slices from the Allen Mouse Brain Atlas showing the ISHH against GHSR mRNA. Arrowheads point to GHSR mRNA positive cells. Scale bar: 100 μm. Panels G-J show high magnification photomicrographs of the areas marked in images of column 2 depicting fluoresceinlabeled cells of the ARC (G) and the IF (H), the PN (I) and the PBP (J). Arrows point to fluoresceinlabeled cells. Scale bar: 10 μm. Figure 2. Panel A shows a scatter dot plot of 2-h food intake of mice SC-injected with vehicle or ghrelin. Circles represent food intake of each mouse. Median and interquartile range is shown for each group. Mann-Whitney test, U(7,5)=2.000, p=0.0051, Cohen’s d=2.0806. Panel B displays a scatter dot plot of the distance traveled by mice SC-injected with vehicle or ghrelin. Circles represent locomotor activity of each mouse, and lines link both values of each mouse as this experiment was performed in a crossover fashion. Mean±SEM are shown for each group. Panel C and D display representative photomicrographs of brain coronal sections containing the VTA and Acb, respectively, subjected to immunostaining against c-Fos (black) and TH (brown). Subdivisions of the VTA and the Acb are overlayed following the delineation described in the Mouse Brain Atlas (Paxinos and Franklin, 2001). Insets in Panel C depict high magnification images of the areas marked in low magnification images corresponding to the three sub-nuclei of the VTA: IF (1), PN (2) and PBP (3). Scale bar: 100 μm (low magnification) and 10 μm (high magnification). Figure 3. Panel A shows a scatter dot plot of 2-h food intake of mice ICV-injected with vehicle or ghrelin. Circles represent food intake of each mouse, and mean±SEM are shown for each group. Student’s t-test with Welch’s correction, t(14.2)=4.707, p=0.0002, Cohen’s d=1.8413. Panel B displays a scatter dot plot of the distance traveled by mice ICV-injected with vehicle or ghrelin. Circles represent locomotor activity of each mouse, and lines link both values of each mouse as this experiment was performed in a crossover fashion. For each group, mean±SEM are shown. Student’s paired t-test, t(9)=2.773, p=0.0108, Cohen’s d=1.2770. Panels C and G show representative photomicrographs of brain coronal sections containing the VTA and the Acb, respectively, subjected to immunostaining against c-Fos (black) and TH (brown). Subdivisions of the VTA and the Acb are overlayed following the delineation described in Mouse Brain Atlas (Paxinos and Franklin, 2001). Insets in Panel C depict high magnification images of the areas marked in low magnification images corresponding to the IF (1), PN (2) and PBP (3). Arrows point to TH-IR cells while arrowheads point

to c-Fos-IR cells. Scale bars: 100 µm (low magnification) and 10 µm (high magnification). Ghrelininjected mice were allowed (ghrelin ICV + food) or not (ghrelin ICV - food) to consume chow after ICV treatment. Panels D to F display scatter dot plots of the number of c-Fos-IR cells in the IF, PN and PBP, respectively, of mice ICV-treated with vehicle or ghrelin. In each panel, circles represent the number of c-Fos of each mouse. Mean±SEM are shown for IF and PN data while median and interquartile range is shown for PBP data. For the IF, Kruskal-Wallis test was performed to analyze the data with H(3,36)=15.91, p=0.0004, η2p=0.3886; Dunn`s post-test, **, p=0.0044 and ***, p=0.0009 vs. vehicle ICV. For the PN, Kruskal-Wallis test was performed to analyze the data with H(3,36)=18.84, p<0.0001, η2p=0.4025; Dunn’s post-test, **, p=0.0024 and ***, p=0.0002 vs. vehicle ICV. For the PBP, Kruskal-Wallis test was performed to analyze the data with H(3,36)=22.82, p<0.0001, η2p=0.5134; Dunn`s post-test, ***, p=0.0002 and ****, p<0.0001 vs. vehicle ICV. Panels H to J show the number of c-Fos-IR cells in the AcbC, MAcbSh and LAcbSh, respectively, of mice ICVtreated with vehicle or ghrelin. Mean±SEM are shown for MAcbSh data while median and interquartile range is shown for AcbC and LAcbSh data. For the AcbC, Kruskal-Wallis test was performed with H(3,36)=17.40, p=0.0002, η2p=0.4324; Dunn’s post-test, ****, p<0.0001 vs. vehicle ICV. For the MAcbSh, one-way ANOVA was performed with F(2,33)=10.76, p=0.0003, η2p=0.3948; Tukey’s posttest, *, p=0.0363 and ***, p=0.0002 vs. vehicle ICV. For the LAcbSh, Kruskal-Wallis test was performed with H(3,36)=15.82, p=0.0004, η2p=0.4324; Dunn’s post-test, *, p=0.0156 and ***, p=0.0004 vs. vehicle ICV. Figure 4. Panel A displays a schematic diagram of the location of the guide cannula (black) and the injector cannula (red) in a coronal brain section containing the VTA (yellow/green). Panel B shows a scatter dot plot of 2-h food intake of mice intra-VTA-injected with vehicle or ghrelin. Circles represent food intake of each mouse. Median and interquartile range is shown for each group. Mann-Whitney test, U(6,5)=1.000, p=0.0043, Cohen’s d=2.9917. Panel C shows a scatter dot plot of the distance traveled by mice intra-VTA-injected with vehicle or ghrelin. Circles represent locomotor activity of each mouse and mean±SEM are shown for each group. Student’s t-test, t(12)=3.007, p=0.0055, Cohen’s d=1.9451. Panels D and H show representative photomicrographs of coronal brain sections containing the VTA and the Acb, respectively, subjected to immunostaining against c-Fos (black) and TH (brown). Subdivisions of the VTA and the Acb are overlayed following the delineation described in Mouse Brain Atlas (Paxinos and Franklin, 2001). Insets in Panel D depict high magnification images of the areas marked in low magnification images corresponding to the IF (1), PN (2) and PBP (3). Arrows point to TH-IR cells, arrowheads point to c-Fos-IR cells and needle points to c-Fos-IR/THIR cells. Scale bars: 100 µm (low magnification) and 10 µm (high magnification). Panels E to G display scatter dot plots of the number of c-Fos-IR cells in the IF, PN and PBP, respectively, of mice intra-VTA-treated with vehicle or ghrelin. Circles represent the number of c-Fos of each mouse and median and interquartile range is shown for each group. For the IF, Mann-Whitney test was performed with U(6,5)=0.0, p=0.0022, Cohen’s d=2.7669. For the PN, Mann-Whitney test was performed with U(6,5)=0.0, p=0.0022, Cohen’s d=1.9784. For the PBP, Mann-Whitney test was performed with

U(6,5)=1.000, p=0.0043, Cohen’s d=2.7589. Panels I to K show scatter dot plots of the number of cFos-IR cells in the AcbC, MAcbSh and LAcbSh, respectively, of mice intra-VTA-treated with vehicle or ghrelin. Circles represent the number of c-Fos of each mouse and median and interquartile range is shown for each group. For the AcbC, Mann-Whitney test was performed with U(6,5)=1.0000, p=0.0043, Cohen’s d=1.4114. For the MAcbSh, Mann-Whitney test was performed with U(6,5)=3.0000, p=0.0152, Cohen’s d=1.4721. For the LAcbSh, Student’s t-test was performed with t(9)=3.372, p=0.0309, Cohen’s d=1.4277. Figure 5. Panel A shows a representative photomicrograph of a coronal brain section containing the VTA of a Gad2-tdTomato mouse. Insets show high magnification images of the areas marked in the low magnification photomicrograph, which correspond to the IF (1), PN (2) and PBP (3). Panel B displays representative photomicrographs of coronal brain sections of Gad2-tdTomato mice ICVtreated with vehicle (upper row) or ghrelin (lower row) subjected to immunostaining against c-Fos. Left, middle and right photomicrographs show c-Fos-IR cells (green), tdTomato cells (red) and the merge of both images, respectively. Subdivisions of the VTA are overlayed following the delineation described in Mouse Brain Atlas (Paxinos and Franklin, 2001). Insets in each panel shows in high magnification the areas marked in low magnification photomicrographs. Arrowheads point at c-FosIR cells and arrows point at c-Fos-IR/tdTomato cells. Scale bars: 100 µm (low magnification) and 10 µm (high magnification). ACKNOWLEDGEMENTS This work was supported by the PICTO2013-0065 grant, which is co-financed by the National Agency of Scientific and Technological Promotion of Argentina and GlaxoSmithKline, to MP. We would like to thank to Dr. María Jose Tolosa and Dr. Guadalupe García Romero for their technical assistance. MPC was supported by CONICET. REFERENCES Abercrombie M (1946) Estimation of nuclear population from microtome sections. Anat Rec 94:239– 247. Abizaid A, Liu Z-W, Andrews ZB, Shanabrough M, Borok E, Elsworth JD, Roth RH, Sleeman MW, Picciotto MR, Tschöp MH, Gao X-B, Horvath TL (2006) Ghrelin modulates the activity and synaptic input organization of midbrain dopamine neurons while promoting appetite. J Clin Invest 116:3229–3239. Adamantidis AR, Tsai H-C, Boutrel B, Zhang F, Stuber GD, Budygin EA, Touriño C, Bonci A, Deisseroth K, de Lecea L (2011) Optogenetic interrogation of dopaminergic modulation of the multiple phases of reward-seeking behavior. J Neurosci Off J Soc Neurosci 31:10829– 10835. Animals NRC (US) C for the U of the G for the C and U of L (2011) Guide for the Care and Use of Laboratory Animals. National Academies Press (US). Available at: https://www.ncbi.nlm.nih.gov/books/NBK54050/ [Accessed May 28, 2018]. Bariselli S, Bellone C (2017) VTA DA neuron excitatory synapses in Shank3 Δex4-9 mouse line. Synap N Y N 71.

Boekhoudt L, Omrani A, Luijendijk MCM, Wolterink-Donselaar IG, Wijbrans EC, van der Plasse G, Adan RAH (2016) Chemogenetic activation of dopamine neurons in the ventral tegmental area, but not substantia nigra, induces hyperactivity in rats. Eur Neuropsychopharmacol J Eur Coll Neuropsychopharmacol 26:1784–1793. Cabral A, Cornejo MP, Fernandez G, De Francesco PN, Garcia-Romero G, Uriarte M, Zigman JM, Portiansky E, Reynaldo M, Perello M (2017a) Circulating Ghrelin Acts on GABA Neurons of the Area Postrema and Mediates Gastric Emptying in Male Mice. Endocrinology 158:1436– 1449. Cabral A, De Francesco PN, Perello M (2015) Brain Circuits Mediating the Orexigenic Action of Peripheral Ghrelin: Narrow Gates for a Vast Kingdom. Front Endocrinol 6 Available at: http://www.frontiersin.org/Neuroendocrine_Science/10.3389/fendo.2015.00044/full [Accessed June 8, 2018]. Cabral A, Fernandez G, Perello M (2013) Analysis of brain nuclei accessible to ghrelin present in the cerebrospinal fluid. Neuroscience 253:406–415. Cabral A, López Soto EJ, Epelbaum J, Perelló M (2017b) Is Ghrelin Synthesized in the Central Nervous System? Int J Mol Sci 18:638. Cabral A, Portiansky E, Sánchez-Jaramillo E, Zigman JM, Perello M (2016) Ghrelin activates hypophysiotropic corticotropin-releasing factor neurons independently of the arcuate nucleus. Psychoneuroendocrinology 67:27–39. Cabral A, Suescun O, Zigman JM, Perello M (2012) Ghrelin Indirectly Activates Hypophysiotropic CRF Neurons in Rodents. PLOS ONE 7:e31462. Cabral A, Valdivia S, Fernandez G, Reynaldo M, Perello M (2014) Divergent Neuronal Circuitries Underlying Acute Orexigenic Effects of Peripheral or Central Ghrelin: Critical Role of Brain Accessibility. J Neuroendocrinol 26:542–554. Carare RO, Bernardes-Silva M, Newman TA, Page AM, Nicoll JAR, Perry VH, Weller RO (2008) Solutes, but not cells, drain from the brain parenchyma along basement membranes of capillaries and arteries: significance for cerebral amyloid angiopathy and neuroimmunology. Neuropathol Appl Neurobiol 34:131–144. Chuang J-C, Perello M, Sakata I, Osborne-Lawrence S, Savitt JM, Lutter M, Zigman JM (2011) Ghrelin mediates stress-induced food-reward behavior in mice. J Clin Invest 121:2684– 2692. Cornejo MP, De Francesco PN, García Romero G, Portiansky EL, Zigman JM, Reynaldo M, Perello M (2018) Ghrelin receptor signaling targets segregated clusters of neurons within the nucleus of the solitary tract. Brain Struct Funct. Denis RGP et al. (2015) Palatability Can Drive Feeding Independent of AgRP Neurons. Cell Metab 22:646–657. Disse E, Bussier AL, Deblon N, Pfluger PT, Tschöp MH, Laville M, Rohner-Jeanrenaud F (2011) Systemic ghrelin and reward: Effect of cholinergic blockade. Physiol Behav 102:481–484. Disse E, Bussier A-L, Veyrat-Durebex C, Deblon N, Pfluger PT, Tschöp MH, Laville M, RohnerJeanrenaud F (2010) Peripheral ghrelin enhances sweet taste food consumption and preference, regardless of its caloric content. Physiol Behav 101:277–281.

Egecioglu E, Jerlhag E, Salomé N, Skibicka KP, Haage D, Bohlooly-Y M, Andersson D, Bjursell M, Perrissoud D, Engel JA, Dickson SL (2010) Ghrelin increases intake of rewarding food in rodents. Addict Biol 15:304–311. Fu Y, Yuan Y, Halliday G, Rusznák Z, Watson C, Paxinos G (2012) A cytoarchitectonic and chemoarchitectonic analysis of the dopamine cell groups in the substantia nigra, ventral tegmental area, and retrorubral field in the mouse. Brain Struct Funct 217:591–612. Graybiel AM, Moratalla R, Robertson HA (1990) Amphetamine and cocaine induce drug-specific activation of the c-fos gene in striosome-matrix compartments and limbic subdivisions of the striatum. Proc Natl Acad Sci 87:6912–6916. Guan XM, Yu H, Palyha OC, McKee KK, Feighner SD, Sirinathsinghji DJ, Smith RG, Van der Ploeg LH, Howard AD (1997) Distribution of mRNA encoding the growth hormone secretagogue receptor in brain and peripheral tissues. Brain Res Mol Brain Res 48:23–29. Hedrich HJ, Bullock GR eds. (2004) The laboratory mouse. Amsterdam ; Boston: Elsevier Academic Press. Hoffman GE, Le W-W, Abbud R, Lee W-S, Susan Smith M (1994) Use of Fos-related antigens (FRAs) as markers of neuronal activity: FRA changes in dopamine neurons during proestrus, pregnancy and lactation. Brain Res 654:207–215. Hoffman GE, Lyo D (2002) Anatomical markers of activity in neuroendocrine systems: are we all “fos-ed out”? J Neuroendocrinol 14:259–268. Holst B, Cygankiewicz A, Jensen TH, Ankersen M, Schwartz TW (2003) High constitutive signaling of the ghrelin receptor--identification of a potent inverse agonist. Mol Endocrinol Baltim Md 17:2201–2210. Hunt GE, McGregor IS (1998) Rewarding brain stimulation induces only sparse Fos-like immunoreactivity in dopaminergic neurons. Neuroscience 83:501–515. Ikemoto S (2007) Dopamine reward circuitry: two projection systems from the ventral midbrain to the nucleus accumbens-olfactory tubercle complex. Brain Res Rev 56:27–78. Jerlhag E (2008) Systemic administration of ghrelin induces conditioned place preference and stimulates accumbal dopamine. Addict Biol 13:358–363. Jerlhag E, Egecioglu E, Dickson SL, Douhan A, Svensson L, Engel JA (2006) Ghrelin administration into tegmental areas stimulates locomotor activity and increases extracellular concentration of dopamine in the nucleus accumbens. Addict Biol 12:6–16. Jerlhag E, Egecioglu E, Dickson SL, Engel JA (2010) Glutamatergic regulation of ghrelin-induced activation of the mesolimbic dopamine system: Mechanisms for ghrelin-induced reinforcement. Addict Biol 16:82–91. Jerlhag E, Egecioglu E, Dickson SL, Svensson L, Engel JA (2008) Alpha-conotoxin MII-sensitive nicotinic acetylcholine receptors are involved in mediating the ghrelin-induced locomotor stimulation and dopamine overflow in nucleus accumbens. Eur Neuropsychopharmacol 18:508–518. Jerlhag E, Egecioglu E, Landgren S, Salomé N, Heilig M, Moechars D, Datta R, Perrissoud D, Dickson SL, Engel JA (2009) Requirement of central ghrelin signaling for alcohol reward. Proc Natl Acad Sci U S A 106:11318–11323.

Kawahara Y, Kawahara H, Kaneko F, Yamada M, Nishi Y, Tanaka E, Nishi A (2009) Peripherally administered ghrelin induces bimodal effects on the mesolimbic dopamine system depending on food-consumptive states. Neuroscience 161:855–864. Kelley AE (2004) Ventral striatal control of appetitive motivation: role in ingestive behavior and reward-related learning. Neurosci Biobehav Rev 27:765–776. Kiba H, Jayaraman A (1994) Nicotine induced c-los expression in the striatum is mediated mostly by dopamine D 1 receptor and is dependent on NMDA stimulation. Mol Brain Res:13. Konigsmark BW (1970) Methods for the Counting of Neurons. In: Contemporary Research Methods in Neuroanatomy (Nauta WJH, Ebbesson SOE, eds), pp 315–340. Springer Berlin Heidelberg. Available at: http://link.springer.com/chapter/10.1007/978-3-642-85986-1_14 [Accessed May 28, 2015]. Lakens D (2013) Calculating and reporting effect sizes to facilitate cumulative science: a practical primer for t-tests and ANOVAs. Front Psychol 4 Available at: http://journal.frontiersin.org/article/10.3389/fpsyg.2013.00863/abstract [Accessed July 26, 2018]. Lammel S, Hetzel A, Häckel O, Jones I, Liss B, Roeper J (2008) Unique properties of mesoprefrontal neurons within a dual mesocorticolimbic dopamine system. Neuron 57:760– 773. Lein ES et al. (2007) Genome-wide atlas of gene expression in the adult mouse brain. Nature 445:168–176. Liu S, Borgland SL (2015) Regulation of the mesolimbic dopamine circuit by feeding peptides. Neuroscience 289:19–42. Lockie SH, Dinan T, Lawrence AJ, Spencer SJ, Andrews ZB (2015) Diet-induced obesity causes ghrelin resistance in reward processing tasks. Psychoneuroendocrinology 62:114–120. Madisen L, Zwingman TA, Sunkin SM, Oh SW, Zariwala HA, Gu H, Ng LL, Palmiter RD, Hawrylycz MJ, Jones AR, Lein ES, Zeng H (2010) A robust and high-throughput Cre reporting and characterization system for the whole mouse brain. Nat Neurosci 13:133–140. Mani BK, Osborne-Lawrence S, Mequinion M, Lawrence S, Gautron L, Andrews ZB, Zigman JM (2017) The role of ghrelin-responsive mediobasal hypothalamic neurons in mediating feeding responses to fasting. Mol Metab 6:882–896. Mani BK, Walker AK, Lopez Soto EJ, Raingo J, Lee CE, Perelló M, Andrews ZB, Zigman JM (2014) Neuroanatomical characterization of a growth hormone secretagogue receptor-green fluorescent protein reporter mouse. J Comp Neurol 522:3644–3666. Margolis EB, Toy B, Himmels P, Morales M, Fields HL (2012) Identification of Rat Ventral Tegmental Area GABAergic Neurons. PLOS ONE 7:e42365. McGirr R, McFarland MS, McTavish J, Luyt LG, Dhanvantari S (2011) Design and characterization of a fluorescent ghrelin analog for imaging the growth hormone secretagogue receptor 1a. Regul Pept 172:69–76. Morales M, Margolis EB (2017) Ventral tegmental area: cellular heterogeneity, connectivity and behaviour. Nat Rev Neurosci 18:73–85.

Nair-Roberts RG, Chatelain-Badie SD, Benson E, White-Cooper H, Bolam JP, Ungless MA (2008) Stereological estimates of dopaminergic, GABAergic and glutamatergic neurons in the ventral tegmental area, substantia nigra and retrorubral field in the rat. Neuroscience 152:1024–1031. Olson VG, Nestler EJ (2007) Topographical organization of GABAergic neurons within the ventral tegmental area of the rat. Synap N Y N 61:87–95. Paxinos G, Franklin KBJ (2001) The mouse brain in stereotaxic coordinates, 2nd ed. San Diego: Academic Press. Perello M, Dickson SL (2015) Ghrelin Signalling on Food Reward: A Salient Link Between the Gut and the Mesolimbic System. J Neuroendocrinol 27:424–434. Perello M, Sakata I, Birnbaum S, Chuang J-C, Osborne-Lawrence S, Rovinsky SA, Woloszyn J, Yanagisawa M, Lutter M, Zigman JM (2010) Ghrelin Increases the Rewarding Value of High-Fat Diet in an Orexin-Dependent Manner. Biol Psychiatry 67:880–886. Perello M, Scott Michael M., Sakata Ichiro, Lee Charlotte E., Chuang Jen‐Chieh, Osborne‐Lawrence Sherri, Rovinsky Sherry A., Elmquist Joel K., Zigman Jeffrey M. (2012) Functional implications of limited leptin receptor and ghrelin receptor coexpression in the brain. J Comp Neurol 520:281–294. Phillipson OT (1979) A Golgi study of the ventral tegmental area of Tsai and interfascicular nucleus in the rat. J Comp Neurol 187:99–115. Prieto-Garcia L, Egecioglu E, Studer E, Westberg L, Jerlhag E (2015) Ghrelin and GHS-R1A signaling within the ventral and laterodorsal tegmental area regulate sexual behavior in sexually naïve male mice. Psychoneuroendocrinology 62:392–402. Ryczko D, Dubuc R (2017) Dopamine and the Brainstem Locomotor Networks: From Lamprey to Human. Front Neurosci 11 Available at: https://www.frontiersin.org/articles/10.3389/fnins.2017.00295/full [Accessed June 6, 2018]. Sanchez-Catalan MJ, Kaufling J, Georges F, Veinante P, Barrot M (2014) The antero-posterior heterogeneity of the ventral tegmental area. Neuroscience 282:198–216. Schellekens H, De Francesco PN, Kandil D, Theeuwes WF, McCarthy T, van Oeffelen WEPA, Perelló M, Giblin L, Dinan TG, Cryan JF (2015) Ghrelin’s Orexigenic Effect Is Modulated via a Serotonin 2C Receptor Interaction. ACS Chem Neurosci 6:1186–1197. Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, Preibisch S, Rueden C, Saalfeld S, Schmid B, Tinevez J-Y, White DJ, Hartenstein V, Eliceiri K, Tomancak P, Cardona A (2012) Fiji: an open-source platform for biological-image analysis. Nat Methods 9:676–682. Shank EJ, Seitz PK, Bubar MJ, Stutz SJ, Cunningham KA (2007) Selective ablation of GABA neurons in the ventral tegmental area increases spontaneous locomotor activity. Behav Neurosci 121:1224–1233. Skov LJ, Jensen M, Christiansen SH, Ratner C, Woldbye DPD, Holst B (2017) Exploring the Behavioral and Metabolic Phenotype Generated by Re-Introduction of the Ghrelin Receptor in the Ventral Tegmental Area. Int J Mol Sci 18.

Steketee JD, Kalivas PW (2011) Drug Wanting: Behavioral Sensitization and Relapse to DrugSeeking Behavior. Pharmacol Rev 63:348–365. Suda Y, Kuzumaki N, Narita M, Hamada Y, Shibasaki M, Tanaka K, Tamura H, Kawamura T, Kondo T, Yamanaka A, Narita M (2018) Effect of ghrelin on the motor deficit caused by the ablation of nigrostriatal dopaminergic cells or the inhibition of striatal dopamine receptors. Biochem Biophys Res Commun 496:1102–1108. Taniguchi H, He M, Wu P, Kim S, Paik R, Sugino K, Kvitsani D, Fu Y, Lu J, Lin Y, Miyoshi G, Shima Y, Fishell G, Nelson SB, Huang ZJ (2011) A Resource of Cre Driver Lines for Genetic Targeting of GABAergic Neurons in Cerebral Cortex. Neuron 71:995–1013. Valdivia S, Cornejo MP, Reynaldo M, De Francesco PN, Perello M (2015) Escalation in high fat intake in a binge eating model differentially engages dopamine neurons of the ventral tegmental area and requires ghrelin signaling. Psychoneuroendocrinology 60:206–216. Wren AM, Small CJ, Abbott CR, Dhillo WS, Seal LJ, Cohen MA, Batterham RL, Taheri S, Stanley SA, Ghatei MA, Bloom SR (2001) Ghrelin Causes Hyperphagia and Obesity in Rats. Diabetes 50:2540–2547. Yanagi S, Sato T, Kangawa K, Nakazato M (2018) The Homeostatic Force of Ghrelin. Cell Metab 27:786–804. Yetnikoff L, Lavezzi HN, Reichard RA, Zahm DS (2014) An update on the connections of the ventral mesencephalic dopaminergic complex. Neuroscience 282:23–48. Zigman JM, Jones JE, Lee CE, Saper CB, Elmquist JK (2006) Expression of ghrelin receptor mRNA in the rat and the mouse brain. J Comp Neurol 494:528–548.

Highlights:  Ghrelin binding cells were present in most VTA sub-nuclei.  SC-injected ghrelin did not affect locomotor activity nor c-Fos in the VTA.  ICV- and intraVTA-injected ghrelin increased locomotor activity.  ICV- and intraVTA-injected ghrelin increased c-Fos in non-DA neurons and in a subset of DA neurons of the VTA.  ICV-injected ghrelin increased c-Fos in a subset of GABA neurons of the VTA.