Impact of negative energy balance on the reward brain system via leptin signaling

Table of Contents

Abstract...2

Introduction...3

Anorexia Nervosa...3

Leptin...3

Leptin Receptor expressing neurons in the LH...3

Activity Based Anorexia Model...4

Pharmacological and chemogenetic treatments in ABA – Leptin and CNO...4

Aims and Hypotheses...5

Experimental Procedures...6

Subjects...6

Behavioural setup: home cage; running-wheel activity, food restriction, and temperature assessments. .6 Experiment 1: Chronic Leptin from day 1...7

Experiment 2: Chronic Leptin from day 5...7

Experiment 3: DREADD Surgical procedures and Drug Description...8

Tissue preparation and immunofluorescence analysis...8

Data analysis and statistics...9

Results...9

Experiment 1: Chronic Leptin, Effects of daily leptin treatment on energy balance in ABA-exposed animals...9

Experiment 1: Chronic Leptin, Effects of daily Leptin Treatment on RWA and Thermogenesis...9

Experiment 2: Chronic Leptin from day 5, Effects of Leptin Treatment on Food Intake, Body Weight11 Experiment 2: Chronic Leptin from day 5, Effects of Leptin Treatment on RWA and Thermogenesis.11 Experiment 3: DREADD/ Chronic CNO...12

Experiment 3: DREADD, Effects of CNO Treatment on Food Intake, Body Weight...13

Experiment 3: DREADD, Effects of CNO Treatment on RWA and Thermogenesis...13

Discussion...15

Experiment 1: Chronic Leptin, Effects of daily leptin treatment on energy balance in ABA-exposed animals...15

Experiment 2: Chronic Leptin, Effects of daily leptin treatment from day 5 on energy balance in ABA- exposed animals...16

Experiment 1: Chronic CNO, Effects of LHLepR neurons depolarization on energy balance...16

Conclusions...18

References...19

Impact of negative energy balance on the reward brain system via leptin signalling

Mattia Modafferi (m.modafferi@students.uu.nl)

Dept. of Translational Neuroscience, Brain division, University Medical Center Utrecht

Abstract

Patients with Anorexia show a chronic Negative Energy Balance (i.e. reduced energy intake relative to energy expenditure). The hormone leptin regulates energy balance and related

neuroendocrine functions. When the administration of this hormone is combined with an animal model of anorexia (Activity-Based Anorexia, ABA), it allows for the investigation of the effects of leptin on AN-like symptoms. This study aims at assessing whether and how pharmacologic treatment with leptin can suppress or prevent the development of anorectic behaviours in ABA.

Additionally, we investigated whether chemogenetic manipulations of LepR in the lateral

hypothalamus (LHLepR) could be effective to modulate energy balance. To this end, in our first experiments, we will use C57BL/6J mice to determine how systemic leptin treatment impacts negative energy balance. In our last experiment, we will use the transgenic LepR-cre mouse for the same purpose, by chemogenetic manipulating LHLepR neurons.

We found that following systemic and chronic leptin treatment, both aspects of energy balance (i.e. energy expenditure, and energy intake) decrease. Furthermore, chronic and systemic leptin treatment once AN-like symptoms had already developed, can prevent a drop in food intake as well as an increase in energy expenditure. Finally, Chemogenetic manipulation of hM3DGi in LHLepR neurons does not lead to reduced FAA or impact on BW/FI.

Leptin treatment in ABA negatively influenced energy balance by decreasing RWA, but at the same time negatively influenced energy balance by decreasing food intake. Furthermore, once AN-like symptoms had already developed, leptin treatment did not negatively affect food intake while suppressing the emergence of hyperactivity during the initial phase of the leptin treatment.

Finally, Chemogenetic stimulation of hM3DGi in LHLepR neurons does not lead to a reduction in hyperactive behaviours, while contemporarily preventing the emergence of significant

differences in body weight or food intake in CNO vs vehicle mice.

Keywords: Activity Based Anorexia, Leptin, DREADD, LHLepR neurons

Introduction

Anorexia Nervosa

Anorexia nervosa (AN) is a deadly metabo-psychiatric disorder associated with significant medical morbidity and mortality, marked distress and impairment, and considerable treatment costs¹. Individuals with AN employ a range of weight loss behaviours including restricted eating, purging, and pathological exercise that is described as being very difficult to control². This state of chronic Negative Energy Balance (NEB) (i.e. reduced energy intake relative to energy expenditure) can lead to starvation³.Previous findings validate the critical role of excessive exercise in AN, strengthening the notion that the energy expenditure deriving from hyperactivity may indeed be a core clinical indicator for AN severity^3,4. Energy expenditure can be divided into individual subcomponents such as basal metabolism, thermogenesis to maintain body

temperature, and physical activity, each with its neural controls and effector pathways⁵. Little is known about the origin of excessive exercise in patients. As such, excessive exercise has been reported as being agreeable and rewarding for the patients² and could derive from a need to stay warm when consistently at a low weight⁶ or as a strategy to efficiently forage as described during periods of famine⁷.

Leptin

Research aiming to define the neurobiological bases of AN has relied largely on brain imaging of ill or recovered AN patients. These studies have highlighted alterations in the brain feeding and reward systems^8,9. Interestingly, previous studies have shown that the hormone leptin can modulate the feeding and reward brain circuits^10–23 and that substantiated hypoleptinaemia is a cardinal feature of acute AN²⁴. Leptin is an adipose tissue hormone that functions as an afferent signal in a negative feedback loop that maintains homeostatic control of adipose tissue mass and reduces food intake¹⁴. In addition to promoting satiation, leptin suppresses the incentive value of food²⁵, and decreased leptin signalling elicits weight gain by affecting both homeostatic and reward circuitry regulating food intake¹⁶. Leptin exerts homeostatic regulation also via modulation of thermogenic balance. The thermogenic effects of leptin have been described before (van Dijk 2001). Leptin-deficient ob/ob mice show hypothermia and have difficulties regulating body temperature in a cold environment. Leptin treatment of ob/ob mice increases body temperature (Harris et al 1998).

Moreover, among human patients with AN who report excessive exercise, a negative correlation has been shown between increased activity and hypoleptinaemia^26,27. Results that are also

consistent with the suppressive effect of leptin on Starvation Induced Hyperactivity (SIH) in food-restricted rats¹². SIH is a phenomenon in laboratory animals whereby rats supplied with food restricted to 60% of ad libitum food intake manage to survive, but die within a short period when exposed to a running wheel¹².

Overall, these data suggest that the modulating effects of leptin on brain areas influence the widely distributed neural network involved in the regulation of energy balance and reward processing. However, it is still unclear how low leptin levels causally influence behaviour in AN patients and what are the associated brain dysfunctions.

Leptin Receptor expressing neurons in the LH

Leptin treatment in patients with congenital leptin deficiency results in decreased fMRI-detected activation of hypothalamic nuclei²⁸. Several authors have recognized the Lateral Hypothalamus (LH) as a key brain area for modulating food intake and regulating energy balance^29,30. Indeed, LH is recognized as a brain area whose manipulation³¹ and lesion³² elicits alterations in eating behaviours¹⁰, physical activity, and thermogenesis³³. The LH contains a heterogeneous assembly of LH Lepin Receptor (LepR)-expressing (LHLepR) neurons³⁴, of which a predominant proportion are GABAergic¹¹, and preferentially innervate GABA interneurons in the VTA³⁵. LHLepR GABA neurons are recognized to control the activity of local orexin-producing neurons within the LH and regulate energy homeostasis by coordinating various physiological responses³³ such as the control of feeding, activity, and energy expenditure³⁶.

In a rodent study, De Vrind and colleagues showed that activation of LHLepR neurons decreases food consumption, and increases locomotion and thermogenesis, leading to body weight loss³⁴. Therefore, LHLepR GABA neurons are important regulators of energy homeostasis and more interestingly, they exert control over both the activity of the LH (local connectivity) and the reward system (by targeting the VTA). As LHLepR neurons are leptin sensitive and functionally involved in feeding, physical activity, thermoregulation and reward processing, they represent a good target to understand better AN-like symptoms as modelled in the mouse ABA model.

Activity Based Anorexia Model

The ABA model is a widely recognized rodent model which mimics many of the important traits of human AN³⁷. When rodents are fed a one-timed meal a day and allowed to run on an activity wheel, they reduce food intake concomitantly with a paradoxical increase in wheel activity leading to negative energy balance and starvation to death if not stopped^38–40. Importantly, this protocol also induces hypoleptinaemia³⁴. Caloric restriction is the primary driver of vulnerability to AN-like symptoms, and wheel access accelerates the expression of the vulnerable phenotype⁴¹. In ABA, serum leptin levels decrease in response to semi-starvation and exercise in itself

potentially may contribute to Hypoleptinemia¹². Food-restricted rodents display increased running activity in the hours preceding food presentation, known as food anticipatory activity (FAA)⁴², and its intensity is proportional to the reduction in food intake⁴³. A hypothesis for the emergence of hyperactivity in the context of AN (and ABA) derives from a need to stay warm when at low weight⁶. This would indicate that physical activity could be thermoregulatory in nature^44,45. Previous studies aiming at highlighting the thermogenic effects of leptin have shown how leptin-deficient ob/ob mice show hypothermia and have difficulties regulating body

temperature⁴⁶, and that leptin treatment of ob/ob mice could restore it⁴⁷. In the context of ABA, leptin has been shown to prevent starvation-induced hypothermia, at least during the initial phase (Day 1 to 3)⁴⁸, and Running wheel activity (RWA) has been proposed as a better predictor of body temperature increase, than viceversa⁴⁸. Hence, this protocol is interesting to study AN-like symptoms as it reproduces them and allows for manipulation of brain circuits to study the consequences of previously described endophenotypes of AN.

Pharmacological and chemogenetic treatments in ABA – Leptin and CNO Leptin treatment on RWA in ABA rodents was shown to reduce excessive exercise acutely^24,48 and during the daily treatment when hyperactivity had already developed¹². However, it can also reduce energy intake^12,48, generating a negative energy balance because the decrease in energy

expenditure is accompanied by a decrease in energy intake and an increase in thermogenesis⁴⁸. Therefore, whereas reduction of hyperactivity by leptin treatment would be beneficial to ABA mice, a decline in food intake and increased thermogenesis would be disadvantageous. In an ABA context, a number of questions regarding the neuronal mechanisms underlying the effects of leptin treatment on energy balance in general and on activity in particular, remain to be addressed.

Intra-LH leptin treatment has shown that leptin action via LHLepR neurons decreases both feeding and body weight²². Moreover, in the LH, ex vivo bath application of leptin in mouse

hypothalamic slices elicit a decrease in neuronal activity and firing rate, respectively^36,49.

Recently, rodent studies have worked on unravelling the brain networks involved in leptin effects on energy balance via chemogenetic manipulation of neuronal activity. In vivo, this can be done through the expression of Designer Receptors Exclusively Activated by Designer Drugs

(DREADDs), in combination with Cre-mediated homologous recombination and the activating ligand Clozapine-N-Oxide (CNO)⁵⁰. DREADDs are coupled through the G protein such as Gq or Gi that activate different intracellular pathways to stimulate or inhibit neuronal activity,

respectively⁵¹.

Prior research suggests that chemogenetic activation of selective midbrain DA neuronal

subpopulations (i.e. VTA-NAc dopaminergic projections) in mildly food-restricted rats, disrupts feeding patterns by either promoting or reducing food intake⁵². When time-limited (90 min) access to food is provided in the ABA paradigm, chemogenetic activation of the same DA subpopulations substantially increases food intake and FAA to prevent ABA-associated weight loss⁵³.

If previous research has focused on VTA-NAc dopaminergic projections, a more recent study has shown that activation of LHLepR neuronal activity (the opposite of leptin effect on these neurons) in mice exposed to a restricted amount of food, directly regulates parameters of energy balance by decreasing feeding and increasing locomotion while at the same time increasing body temperature³⁴. Therefore, activation of LHLepR neurons co-expressing hM3D(Gq) via CNO, into the LH of LepR-Cre food-restricted and ad libitum–fed mice, could result in the opposite effect of leptin.

Hence, chemogenetic manipulation aiming at neuronal targets expressing LHLepR neurons could be a potential target for reversing the detrimental effect on energy balance that was observed in previous leptin studies. Notably, to the best of our knowledge, DREADD manipulation of LHLepR neurons has never been performed in an ABA paradigm (in the context of severe negative energy balance).

Aims and Hypotheses

In light of previous findings on Leptin and the role of LHLepR neurons in the context of negative energy balance, the present study is performed to investigate the effects of pharmacologic and chemogenetic treatment on RWA, food intake, and thermogenesis in mice exposed to the ABA model. More precisely, the first aim of this proposal is to unravel whether we can prevent or rescue AN-like symptoms by manipulating leptin levels at different time points during ABA.

Moreover, to study a selective leptin-sensitive neuronal network, we investigate whether LepR- expressing neurons in the LH can be targeted to modulate energy balance via acting of RWA, feeding or thermoregulation. Considering the findings described above, we hypothesized that, in mice exposed to ABA, a decrease in LHLepR neuronal activity could improve or prevent ABA-

related behavioural alterations, especially hyperactivity. To test this hypothesis, we will express inhibitory hM3Di(Gi) DREADD receptors in LHLepR neurons of Lep-Cre mice exposed to the ABA protocol. Mice will be chronically treated with CNO and, as leptin hyperpolarizes LHLepR

neurons, we expect activation of hM3DGi to mimic leptin’s effect and reduce hyperactivity and body weight loss.

Altogether, this series of experiments would allow us to determine whether leptin treatment can be used to prevent or rescue AN-like symptoms and whether LHLepR-expressing neurons play a selective role in the development of ABA-related alterations.

Experimental Procedures

Subjects

Females, in-house bred C57BL/6J mice were used for the leptin experiments. Females, LepR-cre mice (B6.129(Cg)-Leprtm2(cre)Rck/J, Strain #008320. Jackson Laboratory) were used for the DREADD experiment. Even though no marked gender-specific difference in LepR expression has been observed⁵⁴, we used female rodents as they are relatively more sensitive to leptin treatment than males⁵⁵. All experiments were performed following Dutch laws (Herziene Wet op Dierproeven, Art 10.a2, 2014) and European regulations (Guideline 2010/63/EU) and were approved by the Animal Ethics Committee of Utrecht University.

Behavioural setup: home cage; running-wheel activity, food restriction, and temperature assessments

All mice were collectively housed in the ABA room for one week in collective cages to get used to the adapted light/dark cycle (lights on at 12 pm). Subsequently, mice were housed for 10 days individually, in plastic cages, with running wheels (diameter 14cm; width 9cm) and maintained in a temperature (23°C ± 2°C)- and humidity (60%-70%)-controlled room with ad libitum access to standard chow (CRM (E), SDS, 3.1 kcal/g) and water.

This habituation period provides a baseline for RWA, food intake, body weight and body temperature for each animal. After the habituation period, mice were maintained in the same running wheel cages for an additional 10 days. A cotton pad was provided as cage enrichment.

During this “baseline conditions” period, mice had unrestricted access to food, water and the running wheel. Baseline RWA, Food intake (gr), and %Body weight were used to randomly determine mice allocation to treatment vs control groups. Baseline body weight was used to determine Human End Point (HEP), namely, when animals lost 20% of their body weight on 2 consecutive days, they were removed from the experiment. RWA was registered by a small magnet and a counter activated when the magnet passed a detector during wheel revolution.

Individual wheel running revolutions were continuously registered using Cage Registration Software (Department of Biomedical Engineering, UMC Utrecht, the Netherlands). Body temperature was measured using a thermal camera (Flir T420; FLIR Systems, Inc.) and the accompanying software program ResearchIR (version 4.40.6.24, 64 bit; FLIR Systems). To measure temperature, mice were placed in a smaller cage while thermal pictures were manually taken by the experimenter. As a proxy for body temperature, we used eye temperature. Eye temperature is not only the highest temperature for body area (when measured externally), but it is also more stable across different voxels compared with other regions used to detect

temperature (i.e. brown adipose tissue)³⁴. After the baseline period, mice were maintained in the same running wheel cages and were weighed, and food restricted as follows.

During the ABA restriction period, mice were placed on a restricted feeding schedule (3h of daily access to food at dark onset). Food was given during the first 3h of the dark phase (12 pm) for all the treatment experiments. Body weight, food intake and body temperature were measured every day of the restricted feeding schedule period 3 times a day: before injections (i.e. Leptin or CNO); before food access; after food access. Running wheel revolutions were registered

continuously. FAA was defined as the RWA a few hours before food access⁴². The RWA in the last 6h and 3h of the light phase was used for FAA evaluation across all food experiments.

After the ABA restriction phase (i.e. after a max of 10 days or after reaching HEP), mice are returned to ad libitum food and social housing for a recovery period. After 1-2 weeks they are terminated and brain tissue dissected for further Immunohistological checks.

Experiment 1: Chronic Leptin from day 1

In the first series of leptin experiments, after 3h at the beginning of the dark phase at day 0, food was removed, and 2 months old female mice were placed on the scheduled feeding of 3h/day for the rest of the ABA experiment. A total of 20 mice which had previously established a stable activity level, were subjected to a 10-day chronic injection of either leptin (3mg/Kg, n=10;

Treatment) or saline (0.9% NaCl, n=10; Control) via subcutaneous (s.c.) injections (Fig.1.a.).

Leptin (Recombinant Mouse Leptin Protein,498-OB, R&D systems) was dissolved in sterile isotonic saline.

Fig.1.a. Scheme of experimental set-up. Grey areas indicate the night phase. The green bars indicate time with food access (3hs). The three coloured diamonds represent respectively the time points for leptin injections, body weight (Bw) and temperature measurements, and food measurements. On day 0, food was removed and both groups were placed on schedule feeding. Leptin was s.c. chronically injected every day at 9 am. Food was delivered every day at 12 pm and removed at 3 pm. Body weight and temperature measurements were taken at 9 am (previous to leptin), 12 pm, and 3 pm. Food consumption was measured at 3 pm.

Experiment 2: Chronic Leptin from day 5

In the second series of leptin experiments, the batch of mice used in experiment 1 was reused after recovery following termination of experiment 1. A total of 20, 3 months old female mice, were subjected to the same feeding procedures as the first series of leptin experiments. From day 5 of ABA, mice were subjected to chronic s.c. injection of either leptin (3mg/kg, n=10;

Treatment) or saline (0.9% NaCl, n=10; Control) (Fig.1.b). Measurements procedures were also the same.

Fig.1.b. Leptin was s.c. chronically injected from day 5 at 9 am

Experiment 3: DREADD Surgical procedures and Drug Description

Mice were anaesthetized by Ketamin-Dexdomitor (0.1ml/10g intraperitoneally or i.p.),

Carprofen (5mg/kg s.c.), 10% Lidocaine (0.5 ml s.c. locally). Lidocaine was injected on the skull to provide local analgesia (Lidocaine100mg/ml, Astra Zeneca BV, the Netherlands). An eye ointment was applied (CAF; CEVA Sante Animale BW) and mice were subsequently placed in a stereotactic apparatus (David Kopf Instruments, CA, USA). After exposing the skull to a minor incision, small holes (< 1 mm diameter) were drilled bilaterally for virus injection.

Microinjections of pAAV5-hSyn-DIO-hM3D(Gi)-mCherry (0.3 μL/side, 10⁹ genomic

copies/injection; Addgene 44362) or pAAV5-hSyn-DIO-mCherry (control virus, 0.3 μL/side, 10⁹ genomic copies/injection; Addgene 50459) were performed bilaterally in the LH of LepR-cre mice. The viruses were injected at the coordinates −1.3 mm anterior/posterior, + 1.9 mm

medial/lateral, −5.4 mm dorsal/ventral, 10° angle, at a rate of 0.1 μL/min per side, followed by a 10-minute waiting period before retracting the needles. Following surgery, mice were given atipamezole (2.5 mg/kg, intraperitoneal, SedaStop; AST Farma BV), and saline for rehydration.

During the following 3 days, mice were daily body weighed to control animal welfare and carprofen was continuously provided (5 mg/kg, 0.1ml/100ml of drinking water) diluted in water for 4-5 days. To allow for sufficient DREADD expression, there was a minimum of 3 weeks between surgery and behavioural testing. For chemogenetic inhibition experiments, mice were injected i.p. with clozapine N-oxide (CNO, Hellobio HB6149 , 2mg/kg, 0.1ml/10gr/bw, ip) at 9 am, every day, for the whole duration of the DREADD experiment. CNO was dissolved in sterile saline (0.9% NaCl).

Tissue preparation and immunofluorescence analysis

Mice from experiment 3 were deeply anaesthetized with isoflurane and transcardially perfused briefly with 1×PBS followed by 4% PFA at 10ml/min (50ml/mouse). Whole brains were

removed and post-fixed in 4% PFA/PBS overnight at 4˚C and subsequently transferred to 1×PBS for storage at 4˚C until further processing. Coronal brain sections (40 μm thick) were collected in 1×PBS using a vibratome (Leica VT1200). Freely floating slices were washed with 1×PBS (4 × 10 min) stained with fluorescent 4',6-diamidino-2-phenylindole (DAPI) (5 minutes) to visualise nuclear DNA and finally washed 2×10 minutes in 1×PBS. Slices were then mounted onto glass slides and covered using FluorSave (EMD Millipore Corp.). Confirmation of successful injection placement was determined post-mortem by virtue of the localization of the fluorescent reporters associated with the viruses, which revealed the extent of DREADD labelling in the LH.

Epifluorescent pictures of mounted brain slices (from ×2.5to ×40 magnification) were taken with a Zeiss microscope and Axiovision software to analyse expression patterns. The expression of DREADD (hM3DGi-mCherry) was checked using ImageJ.

Data analysis and statistics

Statistical parameters are presented as mean and ± standard error of the mean (SEM). Data were analysed using Microsoft Excel and GraphPad Prism (version 9, GraphPad Software Inc., San Diego, CA). Two-tailed paired or unpaired Student’s t-tests and Two-way repeated-measures analysis of variance were used where applicable, and a significance criterion of p 0.05, was adopted in all analyses. To investigate the establishment of anorexia-like symptoms over days, daily running, food intake and variation from initial body weight were analysed with a Repeated measure Two-way ANOVA. ANOVA with significant interactions were followed by Šídák's multiple comparisons post hoc tests with a significance level of 0.05 adjusted for the number of statistical tests. For survival analysis, a Log-rank Mantel-Cox test was used.

Results

For measurements of all experiments, baseline levels were not significantly different between vehicle-treated and leptin or CNO-treated groups.

Experiment 1: Chronic Leptin, Effects of daily leptin treatment on energy balance in ABA-exposed animals

The following results are obtained from mice exposed to ABA with leptin injection ( 3mg/kg s.c.) 3h before food exposure from ABA Day 1 to 10.Comparison of Exclusion (or Survival) rate of Vehicle vs Leptin treated mice revealed a significant difference in exclusion rate, p <

.05. Vehicle treated mice had a higher “survival” rate and therefore more Leptin treated animals had to be excluded from the experiment due to a 20% loss in body weight from baseline body weight (Fig.2.a.D).

To test aspects of energy intake (i.e. food intake), we measured daily food consumption during ABA exposure. A Two-Way ANOVA test revealed a significant difference in food intake (gr) between vehicle vs Leptin-treated animals, Group effect; p < .05, F(1, 18) = 5,62. We observed less food intake across the 10 days in Leptin-treated mice (Fig.2.a.D). Two-Way ANOVA tests were used to evaluate group differences in %body weight from the initial baseline at 9 am 12 pm and 3 pm. We found significant time and group %body weight differences at 9am (Injection time), 12pm (Food exposure start), and 3pm (End of the feeding period), p < .05 (Time effect, p

< .05; Group effect, p < .05; Interaction, p < .05). Post hoc comparisons of %body weight from

the initial baseline revealed significant group differences at 9am, from ABA Day 3 to ABA day 9; at 12pm, at day 9; at 3pm during the days 2,3,5, and from day 8 to day 10 (Fig.2.a.C), p < .05.

For all comparison, %body weight from initial baseline of Leptin treated mice did not significantly differ from vehicle treated mice.

Experiment 1: Chronic Leptin, Effects of daily Leptin Treatment on RWA and Thermogenesis

To dissect aspects of energy expenditure (i.e. RWA and thermogenesis), we assessed locomotor activity and body temperature. Two-Way ANOVA tests revealed significant time differences in RWA day/day and during the dark phase (Time effect, p < .05). We found a significant Time x Group Interaction during the light phase, p < 05 (Fig.2.a. E). Although non-significant, results show a positive trend of Leptin-treated mice running less than vehicle between days 5 and 8.

RWA during the dark phase of Leptin vs Vehicle treated mice did not reach any significant difference. Two-Way ANOVA tests were used to assess the difference in thermogenesis at 9 am,12 am, and 3 pm. No significant differences emerged (Fig.2.a. F). Additionally, feeding increased thermogenesis in both groups but no significant difference emerged between groups for all time of measurements.

Figure.2.a. Parameters during the Activity Based Anorexia (ABA) model in Vehicle vs chronically treated mice treated with s.c.

leptin. Leptin injection occurred at 9 am, from day 1 to day 10. (A) RWA in mice after vehicle or leptin (3 mg/day) treatment.

FAA indicates the emergence of Food Anticipatory activity. Green crosses indicate suppression of FAA in Leptin treated mice (B) Percentage of survival during ABA for Vehicle vs Leptin treated mice (C) Percentage body weight (calculated as a percentage from initial body weight) after vehicle or leptin injection at 12 pm (following FAA and before food delivery) (D) Food intake (gr) (E) Distribution of RWA during the light phase (F) Body temperature per day, at different time points (9 am,12 pm,3 pm).

Experiment 2: Chronic Leptin from day 5, Effects of Leptin Treatment on Food Intake, Body Weight

A comparison of the Survival curve of Vehicle vs Leptin-treated mice revealed a non-significant difference in survival rate (Fig.2.b. B). We assessed food intake and body weight. A Two-Way ANOVA test revealed a non-significant difference in food intake (gr) between vehicle vs chronically Leptin-treated animals both from ABA Day 0 to ABA Day 10, as well as from day ABA Day 5 to ABA Day 10 (Fig.2.b. D). Two-Way ANOVA tests were used to evaluate group differences in %body weight from the initial baseline at 9 am, 12 pm, and 3 pm. We found significant differences at 9am (i.e. Injection time) (Treatment effect, p < .05). Post hoc

comparisons revealed significant group differences at 9am at day 12, (Interaction p < .05 ; p < . 05). We also found significant differences at 3pm (i.e. End of feeding period) (Treatment effect, p < .05; Time effect, p < .05 Interaction p < .05). Post hoc comparisons revealed significant group differences at 3pm from ABA Day 9 to ABA Day 13 (Fig.2.b. C). For all comparisons,

%body weight of chronically Leptin-treated mice did not significantly differ when compared to vehicles, except on day 8, whereby leptin mice had overall a lower %body weight, p < .05.

Experiment 2: Chronic Leptin from day 5, Effects of Leptin Treatment on RWA and Thermogenesis

As in experiment 1 we measured RWA and body temperature as a readout of energy expenditure Two-Way ANOVA tests revealed significant time differences in cumulative RWA day/day (Time effect, p < .05; Interaction p < .05). Post hoc tests revealed a significant difference in RWA at ABA days 5 and 6 (p < .05). We found a significant Time effect during the light phase, p < 05 (Fig.2.b.E). During the light phase, leptin-treated mice ran less than vehicle after day 5. A Šídák's multiple comparisons test was used to assess group differences in running patterns from day 5 to day 8. An almost significant difference was found, p = .052 (Fig.2.b.F). Cumulative RWA during the dark phase from day 5 to day 8 did not reach any significance. Two-Way ANOVA tests were used to assess the difference in thermogenesis at 9 am,12 am,3 pm. No differences in body temperature were observed between the experimental groups (Fig.2.b. G).

Fig.2.b. Parameters during the Activity Based Anorexia (ABA) model in Vehicle vs chronically s.c. leptin-treated mice. Red bars indicate the period of chronic leptin injection. Leptin injection occurred at 9 am, from day 5 to day 13 (A) RWA in mice after vehicle or leptin (3 mg/day) treatment. FAA indicates the emergence of Food Anticipatory activity. Green crosses indicate suppression of FAA in chronically Leptin treated mice (B) Probability of survival during ABA for Vehicle vs chronically Leptin treated mice (C) Percentage body weight (calculated as a percentage from initial body weight) after vehicle or leptin injection at 12 pm (following FAA and before food delivery) (D) Food intake (gr) (E) Distribution of RWA during the light phase (F) Difference in distribution of RWA during the light phase, following leptin injections, from day 5 to day 8 (F) Body temperature per day, at different time points (9 am,12 pm,3 pm), Vehicle vs chronic Leptin.

Experiment 3: DREADD/ Chronic CNO

We injected LepR-cre mice with pAAV5-hSyn-DIO-hM3DGi-mCherry and AAV5-hSyn-DIO- mCherry, targeted at the LH region. All LHLepR mice (n = 20) had prominent viral expression in the LH (Fig.2.c.). Because of viral spread, expression of hM3DGi-mCherry extended in some animals to a region near the LH, the zona incerta (ZI) in hM3DGi mice (n = 4) and mCherry control mice (n = 6).

Fig.2.c. Representative images showing (A) Coronal and sagittal representation of injection sites. Mice were infused with pAAV5-hSyn-DIO-hM3DGi-mCherry or AAV5-hSyn-DIO-mCherry, respectively. Both viral injections are bilateral, and injection sites for both viral injections are in the LH (B) Summary of pAAV5-hSyn-DIO-hM3DGi-mCherry or AAV5-hSyn- DIO-mCherry injection sites for the 20 cases utilized for the study (C) Bilateral hM3D: mCherry or (C) mCherry control in LHLepR neurons. White text shoes magnification and viral

Experiment 3: DREADD, Effects of CNO Treatment on Food Intake, Body Weight

A comparison of the Survival curve of hM3DGi vs Control mice revealed a non-significant difference in survival rate (Fig.2.b. B). To test aspects of energy intake, we assessed food intake and body weight. A Two-Way ANOVA test revealed a non-significant difference in food intake (gr) between hM3DGi vs Control animals (Fig.2.d.F). Two-Way ANOVA tests were used to evaluate group differences of %body weight from the initial baseline at 12 pm and 3 pm from ABA day 0 to ABA day 10. We found significant differences at 12pm (Time effect, p < .05;

Interaction p < .05). Post hoc comparisons revealed almost significant group differences at 12pm at day 10, p = 09 (Fig.2.d.C). We also found significant differences in %body weight at 3pm (Time effect, p < .05 Interaction p < .05). Post hoc comparisons revealed significant group differences at 3pm from ABA Day 8 to ABA Day 10 (Fig.2.d.D).

Experiment 3: DREADD, Effects of CNO Treatment on RWA and Thermogenesis

As in the leptin experiments, to test aspects of energy expenditure, we assessed locomotor activity and body temperature. Two-Way ANOVA tests revealed that Chemogenetically activating LH GABA neurons expressing LepR caused significant time differences in RWA day/day (Time effect, p < .05; Interaction p < .05). Post hoc tests revealed nonsignificant differences in RWA of Gi vs Control (Fig.2.d.A). We found a significant Time effect and a significant interaction during the light phase, 3hs before dark onset, p < 05. A Šídák's multiple comparisons tests revealed significant differences at ABA days 5 and 6, p < .05 (Fig.2.d.H).

Control-treated mice ran less than Gi (Fig.2.d.F). Two-Way ANOVA tests were used to assess

the difference in thermogenesis at 9 am,12 am,3 pm. No differences in body temperature were observed between the experimental groups (Fig.2.b.E).

Fig.2.d. Parameters during the Activity Based Anorexia (ABA) model in chronically Gi vs Control i.p. CNO injected mice.

Injections occurred at 9 am, from day 0 to day 10 (A) RWA in mice after CNO (0.3mg/kg) treatment. Green cross indicates suppression of FAA in Gi animals (B) Probability of survival during ABA for Gi vs Control mice (C) Percentage body weight (calculated as a percentage from initial body weight) after CNO injection and before food access at 12 pm (following FAA and before food delivery) (D) Percentage body weight after CNO injection and following food access at 15 pm (E) Body temperature per day, at different time points (9 am,12 pm,3 pm), Vehicle vs chronic Leptin (F) Food intake (gr) (G) Distribution of RWA during the light phase, 6hs before dark onset (H) Difference in distribution of RWA during the light phase, 3hs before dark onset.

Discussion

The present study was performed to investigate the effects of pharmacologic (i.e. Leptin) and Chemogenetic (i.e. CNO) treatment on RWA, food intake, and thermogenesis, in mice exposed to the ABA model. We investigated whether leptin treatment can be used to prevent or rescue AN-like symptoms and whether LHLepR-expressing neurons play a selective role in the

development of ABA-related alterations in the context of ABA.

Experiment 1: Chronic Leptin, Effects of daily leptin treatment on energy balance in ABA-exposed animals

The first aim of this proposal was to investigate whether we could prevent AN-like symptoms by chronically and systemically manipulating leptin levels at different time points during ABA exposure. In our first series of experiments, we found that, following systemic and chronic leptin treatment, both aspects of energy balance (i.e. energy expenditure, and energy intake) decrease.

More specifically, we observed an “early negative effect” of leptin on energy intake (i.e.

decreased food intake during the initial phase of ABA), and a “late positive effect” on energy expenditure (i.e. reduction of FAA during 2nd half of the protocol). On the one hand, leptin treatment favoured a reduction in energy expenditure thanks to relative hypoactivity during the second stage of ABA. By contrast, a decrease in energy intake during the first stage and relative hypothermia were greater than the positive reduction in energy expenditure. Thus, although leptin treatment yielded a reduction in excessive locomotor activity during the food anticipatory period, it also caused a reduction in food intake and body weight. These results consolidate the strong effects of leptin on energy intake, which prevail over the homeostatic drive to eat.

The decrease in energy expenditure observed in our first series of ABA experiments partially ties in with previous studies wherein leptin was shown to suppress hyperactivity^12,48,56 while also decreasing food intake⁴⁸. In fact, we could only observe a reduction of hyperactive behaviours mostly during the last days of ABA exposure, and therefore being unable to prevent the emergence of AN-typical hyperactive behaviours.

Additionally, although previous research demonstrated that leptin treatment could be effective in preventing starvation-induced hypothermia^46,48, we were unable to validate such findings.

Previous studies interpret the reduction in body weight not solely as a consequence of reduced FI, but also as a consequence of increased energy output due to a necessity to generate heat⁴⁶ that can lead to hyperactivity. Wheel running was proposed to be a better predictor of a rise in body temperature, than low body temperature is of a rise in wheel running⁵⁷ (i.e. a homeostatic need to increase in body temperature to prevent hypothermia). We were unable to validate such findings.

In fact, what we observed, was a simultaneous increase of FAA and a decrease in body temperature in leptin treated mice (Fig.2.a. E, and Fig.2.a. F). If the hypothesis that RWA predicts hyperthermia is true, we otherwise propose that the hypothesis that a decrease in body temperature trigger RWA deserves further investigation. In fact, when giving ABA rodents voluntary access to a warm plate, hypothermia, FAA, and body weight loss can be prevented⁵⁷. Furthermore, the reduction of body weight observed in our leptin-treated mice could be driven both by a reduction in FI (i.e. the anorexigenic effect of leptin), a greater energy output deriving from hyperactivity, as well as by a greater energy output deriving by the necessity to generate heat to counterbalance hypothermia. Previous research has shown how acute and central leptin administration in the lateral ventricle or locally in the VTA can suppress RWA without

impacting feeding and body weight⁵⁶. We were unable to validate such a positive effect of leptin on the severe energy balance emerging from ABA. Thus, the unspecificity of our systemic leptin

application might have been the cause of our inability to suppress RWA without negatively impacting aspects of energy intake feeding and body weight.

Experiment 2: Chronic Leptin, Effects of daily leptin treatment from day 5 on energy balance in ABA-exposed animals

To the best of our knowledge, no previous studies investigated the effect of leptin treatment for rescuing AN-like symptoms in an ABA paradigm. To this end, the second aim of this proposal was to investigate whether we could rescue AN-like symptoms by chronically and systemically manipulating leptin levels once the symptoms had developed in the ABA model. We observed how chronic leptin treatment was effective in rescuing the drop in energy intake (i.e. food intake) that we observed in our first series of experiments. Hence, leptin treatment did not negatively affect food intake when compared to control. Additionally, following leptin treatment, we observed an early positive reduction of excessive locomotor activity for leptin-treated mice during the light phase, when compared to control. Hence, once AN-like symptoms had

developed, we were successful in preventing a negative early effect of leptin on Energy Intake, as well as observing a positive early effect of leptin on Energy expenditure. The rescuing in energy expenditure observed in the second phase of ABA is in line with previous studies wherein leptin was shown to suppress hyperactivity once this had already developed¹². However, our results go beyond previous reports showing the effect on leptin once AN symptoms had

developed not only on hyperlocomotion but also on food intake, body weight and thermogenesis.

For all factors encompassing energy balance, although non-significant, we could observe positive trends for the prevention of an early drop in energy intake, a decrease in hyperactive behaviours, as well as the same pattern of thermoregulation in leptin vs vehicle treated mice.

These results must be observed in light of some limitations. One of which is that the same cohort of mice was used following the first series of experiments. Although randomization was

executed, the increase in age could not be controlled. Hence, we could speculate that the smaller AN-like vulnerability that we observed in our second series of experiments could be age-related and not necessarily tied to leptin effect on homeostatic regulation. In fact, as hypothesized in previous research, vulnerability during adolescence (i.e. our first experiment) may be due to age- related differences in how caloric restriction and exercise induce dopamine adaptations⁵⁸.

Additionally, the development of DA-prefrontal cortical projections, which continues through young adulthood, may affect inhibitory control⁵⁸. Additionally, we could also assume that re- exposure to the same model might determine generalization in animal conditional learning when exposed to the same conditions.

Experiment 1: Chronic CNO, Effects of LHLepR neurons depolarization on energy balance

Systemic leptin acts on many different behaviours (because of its multiple targets). Thus, it is important to dissect and possibly locate the networks responsible for the selective effect on energy expenditure and energy intake. In the LH, if some LHLepR neurons are depolarized by leptin, others are hyperpolarized²². As different subpopulations of LepR neurons are present in the LH, it could be that bulk activation deriving from our systemic methodology may elicit disparate behavioural responses when compared to previous studies showing AN-like symptoms following selective leptin treatment. In light of these assumptions, the third aim of this proposal was to selectively chemogenetically manipulate neurocircuitries through which leptin regulates

the control of energy balance. More specifically, we hypothesized that Chemogenetic stimulation of hM3DGi in LHLepR neurons (i.e. Inhibitory) would reduce abnormal energy expenditure to prevent body weight loss and hypothermia.We showed that Chemogenetic stimulation of hM3DGi in LHLepR neurons does not lead to a reduction in FAA, while contemporarily preventing the emergence of significant differences in body weight or food intake in CNO vs vehicle mice. However, it must be noted a FAA tendency towards a significant increase during half of the protocol, whereby hM3DGi mice displayed higher RWA than control. Furthermore, we did not observe significant differences in body temperature between groups despite the effect of CNO treatment on RWA. The increase in body temperature following CNO treatment

observed in previous research³⁴ could be most likely explained by an increase in hyperactivity, rather than by a positive effect of CNO on rescuing the emergence of a negative energy balance.

Although non-significant, these results go beyond previous research investigating the effects of selective modulation of LH_GABA and LHLepR neurons on feeding, locomotion, and

thermogenesis. De Vrind et al.³⁴ found that selective stimulation of hM3DGq (i.e. excitatory) in LHLepR neurons increases body temperature and locomotion, while decreasing body weight, and food intake. The difference observed in our results could be attributed to the functional

characteristics of LHLepR neurons. In fact, when comparing our results to those of older studies, it must be pointed out that LepR neurons in the LH form a distinct functional component within the LH GABAergic circuitry for reward-related behaviour, that is based on their electrical response to leptin and/or projection targets²².

Our CNO study presents some limitations. We aimed at selectively targeting LHLepR neurons.

However, each subpopulation has been shown to presumably modulate a distinct aspect of overall leptin action on energy homeostasis⁶¹. We have assumed that our targeted subpopulation of LHLepR is GABAergic, as they represent the vast majority of LHLepR neurons in the LH¹⁰. However, we could not verify where these LHLepR would selectively project to (i.e. to the VTA).

Further research could selectively inhibit neuronal activity (i.e. hM3DGi) of LHLepR neurons to shed light on how LHLepR neurons may interact with downstream circuits (i.e. VTALepR) to process for behaviours related to energy balance to allow for better identification of therapeutic targets for metabopsychiatric disorders such as AN.

All of our experiments overall must be interpreted in light of general limitations deriving from the model itself, and the anorexigenic nature of leptin.

The ABA model is generally recognised as the closest analogous representation of the negative energy balance typical of the AN phenotype in a laboratory setting. More importantly, it is a fundamental tool for investigating the underlying functional neuronal correlates of spontaneous caloric restriction. However, there are several limitations when considering the translational applicability of this model. For instance, differently from AN patients, rodents would not stop eating if not food restricted. Additionally, rodents would refeed instantly when ad libitum food availability is re-established (i.e. during the recovery phase). Additionally, the nomenclature of the model (i.e. Activity Based Anorexia), although reflective of AN-like symptoms, cannot reflect the whole range of the Anorexic endophenotype. A further limitation is that of all the ABA studies (ours included), almost all present slim differences in the induction protocol. Thus, it is inevitable that the diversities of results observed and obtained can also be attributed to the heterogeneity of experimental procedures. Although these limitations, the ABA model remains a fundamental model of investigation for AN, especially when considering the possible application of not only chemogenetic but also optogenetic manipulations.

Other limitations could be addressed to the anorexigenic nature of leptin. As for the possibility to treat AN-like symptoms via Leptin administration, given its anorexigenic effects on feeding, it would be an arguable approach. Our results, in line with previous research, show that leptin treatment could be effective in reducing hyperactivity in ABA rodents. However, this does not necessarily translate into a clinical context. When trying to correlate serum leptin levels with activity levels in AN patients, mixed results emerge ^12,26,59. Therefore, although serum leptin levels in AN can be correlated with a patient’s physical activity, still no therapeutic evidence in a clinical setting has been provided about the efficiency of such treatment for the suppression of hyperactive behaviours. However, more recent clinical studies showed how cognitive,

behavioural, and emotional aspects of patients with Anorexia with low circulating leptin levels can be ameliorated following metreleptin (recombinant human leptin) treatment⁶⁰.

Hence, considering the positive effects that can be observed in a laboratory setting (i.e. ABA), leptin can indeed have strong potential as a treatment in human patients only if patients also integrate a rehabilitation approach for the willingness to eat. If that would not be the case, the strong anorexigenic effect of leptin would potentially lead to an earlier death.

Conclusions

In conclusion, the major scope of this study was to provide further knowledge on the role of leptin-sensitive neurons on hyperactive behaviours in the ABA model. We did so by

experimenting with the effects of pharmacologic and Chemogenetic treatment on RWA, food intake, and thermogenesis, in mice exposed to the ABA model. We observed that leptin

treatment might be a useful approach for the rescue of AN-like symptoms, but not to prevent the emergence of the same. In fact, in the first series of our experiments with chronic leptin

treatment, we were not effective in preventing the development of AN-like symptoms in ABA mice. In the second series of our experiments with chronic leptin treatment, we were effective in partially rescuing AN-like symptoms once they had emerged thanks to a reduction in FAA.

Finally, from our third experiment, we can conclude that inhibition of LepR-expressing neurons in the LH does not lead to reduced FAA or impact on BW/FI. So it is unlikely that this neuronal population is strongly responsible for these symptoms. We have extended previous research on the functional mechanisms via which leptin regulates the control of energy balance. Although no significant differences emerged, we can conclude that selective stimulation of LHLepR hM3DGi neurons can be considered a potential target for further investigations of the underlying

mechanisms involved in the regulation of AN-like symptoms in an ABA model.

References

1. Gorrell, S., Collins, A. G. E., Daniel, L. G. & Yang, T. T. Dopaminergic Activity and Exercise Behavior in Anorexia Nervosa. OBM Neurobiology 4, 1–19 (2020).

2. Scheurink, A. J. W., Boersma, G. J., Nergårdh, R. & Södersten, P. Neurobiology of hyperactivity and reward: Agreeable restlessness in Anorexia Nervosa. Physiology & Behavior 100, 490–495 (2010).

3. Hebebrand, J. et al. Clinical Trials Required to Assess Potential Benefits and Side Effects of Treatment of Patients With Anorexia Nervosa With Recombinant Human Leptin. Frontiers in Psychology 10, (2019).

4. Sternheim, L., Danner, U., Adan, R. & van Elburg, A. Drive for activity in patients with anorexia nervosa. International Journal of Eating Disorders 48, 42–45 (2015).

5. Berthoud, H.-R., Münzberg, H. & Morrison, C. D. Blaming the Brain for Obesity: Integration of Hedonic and Homeostatic Mechanisms. Gastroenterology 152, 1728–1738 (2017).

6. Carrera, O. et al. Hyperactivity in Anorexia Nervosa: Warming up not just burning-off calories.

PLoS ONE 7, (2012).

7. Illius, A. W., Tolkamp, B. J. & Yearsley, J. The evolution of the control of food intake.

Proceedings of the Nutrition Society 61, 465–472 (2002).

8. Kaye, W. H., Wierenga, C. E., Bailer, U. F., Simmons, A. N. & Bischoff-Grethe, A. Nothing tastes as good as skinny feels: the neurobiology of anorexia nervosa. Trends in Neurosciences 36, 110–

120 (2013).

9. Frank, G. K. et al. Increased Dopamine D2/D3 Receptor Binding After Recovery from Anorexia Nervosa Measured by Positron Emission Tomography and [11C]Raclopride. Biological Psychiatry 58, 908–912 (2005).

10. Omrani, A. et al. Identification of Novel Neurocircuitry Through Which Leptin Targets Multiple Inputs to the Dopamine System to Reduce Food Reward Seeking. Biological Psychiatry

S0006322321001219 (2021) doi:10.1016/j.biopsych.2021.02.017.

11. Vong, L. et al. Leptin action on GABAergic neurons prevents obesity and reduces inhibitory tone to POMC neurons. Neuron 71, 142–154 (2011).

12. Exner, C. et al. Leptin suppresses semi-starvation induced hyperactivity in rats: implications for anorexia nervosa. Mol Psychiatry 5, 476–481 (2000).

13. Cohen, P. et al. Selective deletion of leptin receptor in neurons leads to obesity. J Clin Invest 108, 1113–1121 (2001).

14. Domingos, A. I. et al. Leptin regulates the reward value of nutrient. Nat Neurosci 14, 1562–1568 (2011).

15. Elias, C. F. et al. Leptin Differentially Regulates NPY and POMC Neurons Projecting to the Lateral Hypothalamic Area. Neuron 23, 775–786 (1999).

16. Farooqi, I. S. et al. Leptin Regulates Striatal Regions and Human Eating Behavior. Science 317, 1355–1355 (2007).

17. Friedman, J. M. Leptin and the endocrine control of energy balance. Nat Metab 1, 754–764 (2019).

18. Fulton, S., Woodside, B. & Shizgal, P. Modulation of Brain Reward Circuitry by Leptin. Science 287, 125–128 (2000).

19. Hillebrand, J. J. G., Kas, M. J. H., van Elburg, A. A., Hoek, H. W. & Adan, R. A. H. Leptin’s effect on hyperactivity: Potential downstream effector mechanisms. Physiology & Behavior 94, 689–695 (2008).

20. Hommel, J. D. et al. Leptin Receptor Signaling in Midbrain Dopamine Neurons Regulates Feeding. Neuron 51, 801–810 (2006).

21. Krügel, U., Schraft, T., Kittner, H., Kiess, W. & Illes, P. Basal and feeding-evoked dopamine release in the rat nucleus accumbens is depressed by leptin. European Journal of Pharmacology 482, 185–187 (2003).

22. Leinninger, G. M. et al. Leptin Acts via Leptin Receptor-Expressing Lateral Hypothalamic Neurons to Modulate the Mesolimbic Dopamine System and Suppress Feeding. Cell Metabolism 10, 89–

98 (2009).

23. van der Plasse, G. et al. Modulation of cue-induced firing of ventral tegmental area dopamine neurons by leptin and ghrelin. Int J Obes 39, 1742–1749 (2015).

24. Hebebrand, J., Muller, T. D., Holtkamp, K. & Herpertz-Dahlmann, B. The role of leptin in anorexia nervosa: clinical implications. Mol Psychiatry 12, 23–35 (2007).

25. Schlögl, H. et al. Leptin Substitution in Patients With Lipodystrophy: Neural Correlates for Long-term Success in the Normalization of Eating Behavior. Diabetes 65, 2179–2186 (2016).

26. Holtkamp, K. et al. Physical Activity and Restlessness Correlate with Leptin Levels in Patients with Adolescent Anorexia Nervosa. Biological Psychiatry 60, 311–313 (2006).

27. Holtkamp, K. et al. Elevated Physical Activity and Low Leptin Levels Co-occur in Patients with Anorexia Nervosa. The Journal of Clinical Endocrinology & Metabolism 88, 5169–5174 (2003).

28. Baicy, K. et al. Leptin replacement alters brain response to food cues in genetically leptin- deficient adults. Proceedings of the National Academy of Sciences 104, 18276–18279 (2007).

29. Brobeck, J. R. Mechanism of the development of obesity in animals with hypothalamic lesions.

Physiol Rev 26, 541–559 (1946).

30. Hoebel, B. G. & Teitelbaum, P. Hypothalamic Control of Feeding and Self-Stimulation. Science 135, 375–377 (1962).

31. Hernandez, L. & Hoebel, B. G. Feeding and hypothalamic stimulation increase dopamine turnover in the accumbens. Physiology & Behavior 44, 599–606 (1988).

32. Elmquist, J. K., Elias, C. F. & Saper, C. B. From Lesions to Leptin: Hypothalamic Control of Food Intake and Body Weight. Neuron 22, 221–232 (1999).

33. Inutsuka, A. et al. Concurrent and robust regulation of feeding behaviors and metabolism by orexin neurons. Neuropharmacology 85, 451–460 (2014).

34. de Vrind, V. A. J., Rozeboom, A., Wolterink-Donselaar, I. G., Luijendijk-Berg, M. C. M. &

Adan, R. A. H. Effects of GABA and Leptin Receptor-Expressing Neurons in the Lateral Hypothalamus on Feeding, Locomotion, and Thermogenesis. Obesity 27, 1123–1132 (2019).

35. Johnson, S. W. & North, R. A. Two types of neurone in the rat ventral tegmental area and their synaptic inputs. The Journal of Physiology 450, 455–468 (1992).

36. Goforth, P. B., Leinninger, G. M., Patterson, C. M., Satin, L. S. & Jr, M. G. M. Leptin Acts via Lateral Hypothalamic Area Neurotensin Neurons to Inhibit Orexin Neurons by Multiple GABA- Independent Mechanisms. 11.

37. Dixon, D. P., Ackert, A. M. & Eckel, L. A. Development of, and recovery from, activity-based anorexia in female rats. Physiology & Behavior 80, 273–279 (2003).

38. Epling, W. F., Pierce, W. D. & Stefan, L. A theory of activity-based anorexia. International Journal of Eating Disorders 3, 27–46 (1983).

39. Routtenberg, A. & Kuznesof, A. W. Self-starvation of rats living in activity wheels on a restricted feeding schedule. Journal of Comparative and Physiological Psychology 64, 414–421 (1967).

40. Hall, J. F. & Hanford, P. V. Activity as a function of a restricted feeding schedule. Journal of Comparative and Physiological Psychology 47, 362–363 (1954).

41. Beeler, J. A. & Burghardt, N. S. The Rise and Fall of Dopamine: A Two-Stage Model of the Development and Entrenchment of Anorexia Nervosa. Front. Psychiatry 12, 799548 (2022).

42. Mistlberger, R. E. Circadian food-anticipatory activity: Formal models and physiological mechanisms. Neuroscience & Biobehavioral Reviews 18, 171–195 (1994).

43. Kas, M. J. H. et al. Mu-opioid receptor knockout mice show diminished food-anticipatory activity. European Journal of Neuroscience 20, 1624–1632 (2004).

44. Gutierrez, E. Temperature but not leptin prevents semi-starvation induced hyperactivity in rats:

implications for anorexia nervosa treatment. Scientific Reports 10, (2020).

45. Gutiérrez, E., Vázquez, R. & Boakes, R. A. Activity-based anorexia: ambient temperature has been a neglected factor. Psychon Bull Rev 9, 239–249 (2002).

46. Van Dijk, G. The Role of Leptin in the Regulation of Energy Balance and Adiposity. Journal of Neuroendocrinology 13, 913–921 (2001).

47. Harris, R. B. et al. A leptin dose-response study in obese (ob/ob) and lean (+/?) mice.

Endocrinology 139, 8–19 (1998).

48. Hillebrand, J. J. G., Koeners, M. P., de Rijke, C. E., Kas, M. J. H. & Adan, R. A. H. Leptin Treatment in Activity-Based Anorexia. Biological Psychiatry 58, 165–171 (2005).

49. Schiffino, F. L. et al. Activation of a lateral hypothalamic-ventral tegmental circuit gates motivation. PLoS ONE 14, e0219522 (2019).

50. Gore, B. B., Soden, M. E. & Zweifel, L. S. Manipulating Gene Expression in Projection Specific ‐ Neuronal Populations Using Combinatorial Viral Approaches. Current Protocols in Neuroscience 65, (2013).

51. Navarro, M. et al. Lateral Hypothalamus GABAergic Neurons Modulate Consummatory Behaviors Regardless of the Caloric Content or Biological Relevance of the Consumed Stimuli.

Neuropsychopharmacol 41, 1505–1512 (2016).

52. Boekhoudt, L. et al. Does activation of midbrain dopamine neurons promote or reduce feeding?

Int J Obes 41, 1131–1140 (2017).

53. Foldi, C. J., Milton, L. K. & Oldfield, B. J. The Role of Mesolimbic Reward Neurocircuitry in Prevention and Rescue of the Activity-Based Anorexia (ABA) Phenotype in Rats.

Neuropsychopharmacol 42, 2292–2300 (2017).

54. Bennett, P. A. et al. Differential Expression and Regulation of Leptin Receptor Isoforms in the Rat Brain: Effects of Fasting and Oestrogen. NEN 67, 29–36 (1998).

55. Clegg, D. J., Riedy, C. A., Smith, K. A. B., Benoit, S. C. & Woods, S. C. Differential Sensitivity to Central Leptin and Insulin in Male and Female Rats. Diabetes 52, 682–687 (2003).

56. Verhagen, L. A. W., Luijendijk, M. C. M. & Adan, R. A. H. Leptin reduces hyperactivity in an animal model for anorexia nervosa via the ventral tegmental area. European Neuropsychopharmacology 21, 274–281 (2011).

57. Hillebrand, J., Derijke, C., Brakkee, J., Kas, M. & Adan, R. Voluntary access to a warm plate reduces hyperactivity in activity-based anorexia. Physiology & Behavior 85, 151–157 (2005).

58. Beeler, J. A. et al. Vulnerable and Resilient Phenotypes in a Mouse Model of Anorexia Nervosa.

Biological Psychiatry S0006322320317443 (2020) doi:10.1016/j.biopsych.2020.06.030.

59. Hofmann, T. et al. Plasma kisspeptin and ghrelin levels are independently correlated with physical activity in patients with anorexia nervosa. Appetite 108, 141–150 (2017).

60. Milos, G. et al. Short-term metreleptin treatment of patients with anorexia nervosa: rapid on-set of beneficial cognitive, emotional, and behavioral effects. Transl Psychiatry 10, 303 (2020).

61. Myers, M. G., Münzberg, H., Leinninger, G. M. & Leshan, R. L. The Geometry of Leptin Action in the Brain: More Complicated Than a Simple ARC. Cell Metabolism 9, 117–123 (2009).