NtsR1^flox mice were generated by the Michigan State University Transgenic and Genome Editing Facility (TEGF). A CRISPR/Cas9 strategy was used to insert loxP sites flanking exon 1 of the genomic Ntsr1 locus. Correct insertion of both loxP sites in the founder mouse was validated by DNA sequencing. The founder was crossed with C57Bl/6J mice (Jackson Stock #033365) for 3 generations to stabilize the insertion. Subsequent NtsR1^flox/+ progeny were intercrossed to generate NtsR1^flox/flox mice for studies. All mice were bred and housed under a 12 h light/12 h dark cycle and were cared for by Campus Animal Resources (CAR). We studied an approximately equal number of male and female mice on the C57Bl/6J background. Cohorts were staggered to control for any potential seasonal effects. Since we did not observe any metabolic or behavioral differences between males and females, both sexes were pooled. All animal protocols were approved by the Institutional Animal Care and Use Committee (IACUC) at Michigan State University and the Oklahoma Medical Research Foundation, in accordance with the Association for Assessment and Accreditation of Laboratory Animal Care and National Institutes of Health.

Developmental Model: NtsR1^flox mice were intercrossed with DAT^Cre mice (Jackson Stock #006660) to generate DATR1^Null mice (DAT^Cre/+;NtsR1^flox/flox), which undergo early developmental Cre-mediated deletion of NtsR1 from all DAT-expressing neurons. We also studied (DAT^+/+;NtsR1^flox/flox) littermates as controls, referred to as NtsR1^flox/flox, that retain NtsR1 expression in all cells. For brain slice electrophysiology experiments, mice were group housed in ventilated standard cages with ad libitum access to food and water. Rooms were maintained on a normal 12-h light/dark cycle (lights on at 6:00 AM) with the temperature held at 26°C.

Adult Model: Heterozygous NtsR1^flox/+ mice were mated to produce homozygous NtsR1^flox/flox mice and NtsR1^+/+ (referred to as wild type, Wt) littermates. NtsR1^flox/flox and Wt mice were then injected at ∼8 wks of age in the VTA (bilaterally) with either AAV-GFP or AAV-Cre-GFP to generate NtsR1^flox/flox (VTAR1^Null), Wt^Cre, NtsR1^{flox/flox;GFP} (VTAR1^GFP), and Wt^GFP mice. VTAR1^Null mice thus have an adult-onset deletion of NtsR1 specifically from VTA neurons.

Male and female 12–16 wk old NtsR1^flox/flox and Wt littermates were deeply anesthetized with sodium pentobarbital. Tissue from the VTA was microdissected, snap frozen on dry ice and stored at −80°C. RNA was extracted using Trizol (Invitrogen) and 200 ng samples were converted to cDNA using the Superscript First Strand Synthesis System for RT-PCR (Invitrogen). Sample cDNAs were analyzed in triplicate via quantitative RT-PCR for gene expression using Applied Biosytems TaqMan probes for Ntsr1 (Mm00444459_m1), tyrosine hydroxylase (TH; Mm00447557_m1) and GAPDH (Mm99999915_g1) using a BioRad CFX Connect Real-Time System. With GAPDH expression as an internal control, relative mRNA expression values were calculated by the 2^–ΔΔCt method.

Recordings were done on dopaminergic neurons from 14 to 20-month-old male and female NtsR1^flox/flox (as controls) and DATR1^Null mice. A total of 9 NtsR1^flox/flox cells (4 from males and 5 from females) and 6 DATR1^Null cells (4 from females and 2 from males) were included in the data analysis. On the day of the experiment, mice were anesthetized with 2% 2,2,2-tribromoethanol and transcardially perfused with ice-cold carboxygenated (95% O2 and 5% CO2) cutting solution for 45 s. The brains were then quickly extracted and placed in the same cutting solution containing the following (in mM): 2 KCl, 7 MgCl2, 0.5 CaCl2, 1.2 NaH2PO4, 26 NaHCO3, 11 D-glucose, and 250 sucrose. Horizontal midbrain slices (200 μm) containing the lateral VTA were obtained using a vibrating microtome (Leica). Slices were incubated for 30 min at 33–34°C with carboxygenated artificial cerebrospinal fluid (ACSF) that also contained the NMDA receptor antagonist MK-801 (10 μM). The ACSF solution contained (in mM): 126 NaCl, 2.5 KCl, 1.2 MgCl2, 2.4 CaCl2, 1.2 NaH2PO4, 21.4 NaHCO3, and 11.1 D-glucose. Slices were then left to stabilize at room temperature for at least 30 min.

Slices were placed in a recording chamber attached to an upright microscope (Nikon Instruments) and maintained at 33–34°C with ACSF perfused at a rate of 2 ml/min. VTA DA-ergic neurons were visually identified based on their location in relation to the midline, third cranial nerve, and the medial terminal nucleus of the accessory optic tract. Neurons were further identified physiologically by the presence of spontaneous pacemaker firing (1–5 Hz) with wide extracellular waveforms (>1.1 ms) and the presence of an outward potassium current (I_A), and a hyperpolarization-activated current (IH) of >100 pA. Recording pipettes (2.4–2.7 MΩ resistance) were constructed from thin wall capillaries (World Precision Instruments) with a PC-100 puller (Narishige International). Whole-cell recordings were obtained using an internal solution containing the following (in mM): 100 K-gluconate, 20 NaCl, 1.5 MgCl2, 10 HEPES-K, 2 ATP, 0.4 GTP, and 0.025 EGTA, pH 7.35–7.40, 269 –272 mOsm. Drugs: The voltage was held at −55 mV in the presence of the fast transmitter receptor antagonists picrotoxin (100 μM, GABAA; TCI) and DNQX (10 μM, AMPA; Sigma Aldrich). A minimum of 5 min of stable baseline was obtained before bath perfusion of NT_8–13 (300 nM; Sigma Aldrich) for 5 min, followed by at least 10 min of washout.

Developmental Model: At 4 weeks of age, study mice were individually housed with ad libitum access to water and chow (Harlan Teklad #7913) or 45% high-fat diet (HFD, Research Diets D12451). Food weight and body weight were measured once per week for DATR1^Null, NtsR1^flox/flox, DAT^Cre, and Wt littermates for 12 wks, until mice were 16 wk of age.

Adult Model: NtsR1^flox/flox mice and Wt littermates that underwent surgery at ∼8 weeks (described below) were housed individually after surgery and kept on ad libitum chow (Harlan Teklad #7913) for 8 weeks and then were switched to ad libitum 45% high-fat diet (HFD, Research Diets D12451) for the duration of experiments, unless otherwise specified. Diet weight and body weight were measured once per week starting 2 weeks after surgery and finishing 8 weeks after HFD was given.

NtsR1^flox/flox mice and Wt littermates (8–12 weeks) were anesthetized (isoflurane/oxygen mixture, 2–4%) and given analgesic (Meloxicam, 5 mg/kg) prior to bilateral stereotaxic injection of AAV1.hSyn.H1.eGFP-Cre.WPRE.SV40 (AAV-Cre-GFP, U Penn Vector Core) or AAV-GFP into the VTA (100 nL per side, A/P: –3.2, M/L: ± 0.48, D/V: –4.65) as per the mouse brain atlas of Franklin and Paxinos (Franklin and Paxinos, 2008). Brains were examined via post-hoc immunostaining for GFP and TH (described below) to verify targeting and GFP expression in VTA TH + neurons. Data from injected mice were excluded if GFP-expressing soma were not confined to the VTA.

Mice were analyzed in PhenoMaster metabolic cages (TSE Systems) that continuously monitored food and water intake, locomotor activity, wheel running, and metabolic parameters (VO2, respiratory exchange ratio [RER], and energy expenditure). Mice were acclimated in cages for 48 h prior to testing day. Body weight was measured before mice were placed in TSE cages. Ambient temperature was maintained at 20–23°C and airflow rate was adjusted to maintain an oxygen differential around 0.3% at resting conditions. Developmental model mice were tested after 16 wks on chow diet. Adult model mice were analyzed at the end of 8 wks on chow and then again at the end of 8 wks on HFD.

Chow or HFD was removed from home cages at ∼5PM and mice were given a clean cage bottom. Mice had ad libitum access to water during food-deprivation. The following morning at ∼9AM, fasted mice were given back chow or HFD pellets. Food intake and body weight were measured 1.5, 12, and 24 h after food was restored. Developmental model mice were tested after 16 wks on chow diet. Adult model mice were analyzed at the end of 8 wks on chow diet and then again at the end of 8 wks on HFD.

Mice were trained to nose-poke (active hole randomized) for unflavored 20 mg sucrose pellets (TestDiet 1811555) in operant-responding chambers (Med Associates) as previously described (Sharma et al., 2012; Woodworth et al., 2017a). Training: Mice were food restricted to 85% of their body weight during FR1 training sessions, which occurred over 10–16 consecutive days. Each FR1 training session was terminated after 1 h or when the mouse had earned 50 reinforcers. Once mice achieved 75% response accuracy with ≥20 reinforcers earned on 3 consecutive days of FR1 training, they were switched to ad libitum chow and trained on an FR5 schedule for 3 consecutive days. Mice that failed to reach FR1 criteria after 16 days were removed from the study. Progressive Ratio (PR) Testing: After FR5 testing mice were subject to a progressive ratio (PR) schedule where PR = [5e(R^*0.2)]−5 with R = number of food reinforcers earned + 1. The PR breakpoint was recorded as the highest response requirement completed for each 1 h test session. Mice were tested until they achieved stable PR, defined as <10% variation in rewards earned over 3 consecutive sessions. To determine if hunger alters operant responding, mice were fasted overnight, and then tested on the PR schedule.

Mice were placed in a quiet room under red lights for an hour before placement in open field boxes. For open field, locomotor activity was measured and analyzed for 30 min using a digital CCD camera and video-tracking software (Clever Sys) (Eagle et al., 2015). For amphetamine (AMPH) trials, developmental model mice were injected with PBS ∼8 min after being placed in the box and then again with AMPH at the ∼40-min mark. Developmental model mice were tracked for a total of 70 min. Adult HFD model mice were treated with PBS immediately before being placed in the boxes, and then with AMPH at the ∼30-min mark. Adult HFD model mice were tracked for a total of 90 min.

Mice were placed in a quiet room in a clean home cage and any nestlet material or enrichment was removed overnight. On testing day (light cycle), food and water were removed, and a pre-weighed, new cotton nestlet was added to the cage. After 30 min, intact remnants of the nestlet were removed from the cage and weighed. If the nestlet was found to be wet, it was air dried overnight and weighed the following day.

Mice were placed in a quiet room in home cage and any nestlet material or enrichment was removed overnight. On testing day (light cycle), mice were placed in the middle of a clean cage with 10 evenly dispersed marbles on top of the bedding. After 30 min mice were returned to their home cage. Photos taken before and after the test were used to determine the number and percentage of marbles buried. Photos were assigned new, non-genotype descriptive ID’s to facilitate blinded data entry. A marble was counted as buried if 2/3 of the marble was covered.

Mice were intracardially perfused with 10% formalin (n = 3 per genotype/viral injection), then brains were extracted, postfixed in 10% formalin overnight at 4°C and dehydrated with 30% sucrose in PBS for 2–3 days. Brains were then coronally sectioned (30 μm) for RNAScope, as per (Woodworth et al., 2018). Three free floating sections of the VTA per mouse were selected for application of RNAScope single-plex assay (catalog #322360, Advanced Cell Diagnostics) per the manufacturer’s protocol. Sections were washed in 1X PBS and incubated in Pretreatment 1 (H₂O₂) at room temperature (RT) until bubbling stopped (45–60 min) followed by 0.5X PBS wash and mounting on positively charged slides. Following washing in dH₂O and drying at 60°C oven, sections were incubated in 1X Pretreatment 2 (Target Retrieval Agent) for 5–10 min at 99–104°C and then washed with dH₂O, dried at RT, dipped in 100% EtOH and air dried. Next, they were incubated in Pretreatment III solution (Protease Plus) for 15 min at 40°C, followed by dH₂O wash. Sections were then incubated in NtsR1 (cat #: 422411, Advanced Cell Diagnostics) probes for 2 h in a humidified oven at 40°C. After amplification steps (Amp1-6), hybridization was visualized by application of Fast-Red-A and Red-B (60:1) for 15–20 min. Finally, after washing Fast-Red solution, slides were dehydrated by dipping into xylene and cover-slipped with antifade mounting agent.

Mice received a lethal i.p. dose of pentobarbital (Fatal Plus, Vortech) followed by transcardial perfusion with 0.2M PBS (pH 7.4) and then 10% neutral-buffered formalin (Fisher Scientific). Brains were removed, postfixed, and coronally sectioned into four series of 30 μm sections using a freezing microtome (Leica). Free-floating sections were exposed to primary antibodies including chicken-GFP (1:2000, Abcam, ab13970) and mouse anti-TH (1:1000, Millipore, AB_2201528) followed by species-specific secondary antibodies conjugated to Alexa-568 (1:200, LifeTech, AB_2534017) or Alexa-488 (1:200, Jackson ImmunoResearch, 703-545-155). Sections mounted onto slides were analyzed via an Olympus BX53 fluorescence microscope with FITC and Texas Red filters. Images were collected using Cell Sens software and a Qi-Click 12 Bit cooled camera and analyzed using Photoshop (Adobe). Masks applied in Photoshop to enhance brightness and/or contrast were applied uniformly to all samples.

Unpaired Student’s t-tests and ordinary 2-way ANOVA with Sidak post-test were calculated using Prism 7 (GraphPad). A p-value of <0.05 was considered statistically significant. *p < 0.05, ^**p < 0.01, ^***p < 0.001, ^****p < 0.0001.

We first assessed whether our strategy can developmentally and conditionally delete NtsR1 from DA neurons by crossing NtsR1^flox/flox and DAT^Cre mice. Progeny with both DAT^Cre and NtsR1^flox/flox are referred to as DATR1^Null mice and should exhibit early developmental Cre-mediated deletion of NtsR1 from all DAT-expressing neurons. In the absence of DAT^Cre, NtsR1^flox/flox mice exhibited robust Ntsr1 mRNA expression in the VTA (Figure 1A), consistent with the reported expression of NtsR1 in this brain region (Palacios et al., 1988; Lepee-Lorgeoux et al., 1999; Woodworth et al., 2017a,2018; Perez-Bonilla et al., 2021). In contrast, NtsR1 expression was absent from the VTA of DATR1^Null mice, confirming successful DAT^Cre-mediated deletion of Ntsr1, and that this strategy can be used to developmentally delete NtsR1 from Cre-expressing neurons (Figure 1B). Next, we investigated whether introducing loxP sites flanking NtsR1 exon 1 impacts DA neurons by measuring gene expression in VTA from Wt controls and NtsR1^flox/flox mice. In the absence of Cre expression, NtsR1^flox/flox mice should retain intact NtsR1 expression comparable to the Wt mice. The lack of significant differences in fold expression of Th or Ntsr1 (Figures 1C,D) suggests that loxP insertion itself does not disrupt baseline gene expression of VTA DA neurons. We then examined whether Cre-mediated deletion of NtsR1 from NtsR1^flox/flox mice functionally impacted the response of VTA NtsR1 neurons to Nts using VTA-containing slices from NtsR1^flox/flox (with intact NtsR1) and DATR1^Null (developmental deleted) mice. Bath application of Nts produced a significant inward shift in holding current in VTA DA neurons from NtsR1^flox/flox mice with intact NtsR1, but these Nts-mediated responses were absent in VTA DA neurons from DATR1^Null mice (Figures 1E–G). These data support the functional loss of NtsR1 from DATR1^Null mice and suggest that these mice can be used to assess the impact of deleting NtsR1 from DA neurons.

We then examined how developmental deletion of NtsR1 from DAT-expressing neurons (e.g., DA neurons) impacts body weight. To do this, we measured the weekly body weight and food intake of chow fed NtsR1^flox/flox and DATR1^Null mice with developmental deletion of NtsR1 from DA neurons while in their individual home cages (Figures 2A,B). Deleting NtsR1 from all DA expressing neurons did not alter body weight or chow intake over 16 weeks (Figures 2A,B). We also analyzed NtsR1^flox/flox and DATR1^Null mice via metabolic chambers (Figures 2C–H), but observed no differences between groups in chow intake, water intake, ambulatory locomotor activity, wheel rotations, energy expenditure or RER over 24 h of measurement (Figures 2C–H). Neither were any group differences noted across the light/dark cycle (data not shown). These data suggest that developmental deletion of NtsR1 from DA neurons is not sufficient to overtly disrupt metabolic phenotype.

Next, we examined whether developmental deletion of NtsR1 from DA neurons altered the ability of DATR1^Null mice to coordinate ingestive behavior. NtsR1^flox/flox control mice and DATR1^Null mice exhibited similar levels of fasting-induced refeeding and body weight gain, suggesting normal responses to hunger (Figures 3A,B). Given the important role of the DA system in motivated responding for palatable food rewards, we then examined whether developmental deletion impacted operant responding for sucrose, which is a DA-dependent behavior (Figure 3C). Further, we tested mice that were sated (baseline) and after overnight fasting to assess whether hunger altered motivated responding for sucrose. NtsR1^flox/flox control mice and DATR1^Null mice had comparable numbers of magazine entries, percentages of correct nose pokes, earned sucrose reinforcers, PR values and ate a similar percentage of sucrose pellets across baseline and fasted conditions (Figures 3D–H). Overall, these results suggest that developmental deletion of NtsR1 from DA neurons doesn’t impair coordinated feeding responses to hunger nor goal-directed behavior needed to obtain food.

We used open field chambers to determine whether developmental deletion of NtsR1 from DA neurons impacts DA-dependent locomotor activity and anxiety behaviors (Figures 4A–F). When placed in the novel open field environment, DATR1^Null mice traveled more distance compared to NtsR1^flox/flox littermates (Figure 4B), indicative of increased exploratory behavior. Since DATR1^Null mice and NtsR1^flox/flox controls spent comparable percentage of time in the periphery, it suggests that the developmental deletion of NtsR1 does not invoke anxiety-like movement (Figure 4C). Notably, constitutive NtsR1 knockout mice also exhibit elevated basal and stimulant-induced locomotor activity that is thought to be due to elevated DA signaling (Liang et al., 2010). We therefore examined whether developmental deletion of NtsR1 from DA neurons similarly promotes hyperactivity by measuring amphetamine (AMPH)-induced locomotor activity in DATR1^Null and NtsR1^flox/flox littermates. Importantly, this paradigm is well established to assess the integrity of the mesolimbic DA system. We found that DATR1^Null mice and NtsR1^flox/flox controls exhibited similar locomotor responses to PBS (control) treatment. AMPH injection (4mg/kg i.p.) significantly increased locomotor activity in both groups, as expected, but DATR1^Null mice maintained significantly higher AMPH-induced locomotor activity compared to the control mice (Figure 4D). These data indicate that loss of NtsR1 from DAT expressing neurons may enhance DA signaling, similar to compensatory increases in extracellular DA that occur with partial loss of DA neurons (Robinson et al., 1994) or constitutive NtsR1 deletion. Although hyperactivity and excessive DA release have been associated with anxiety, we found no differences in anxiety-like behaviors between DATR1^Null and NtsR1^flox/flox mice as assessed via percentage of time spent in the periphery of open field chambers (Figure 4C), nor via nestlet shredding or elevated plus maze (EPM) tests (Figures 4E,F).

NtsR1^flox/flox and DATR1^Null mice were given ad libitum high fat diet (HFD) in their home cages, which is well established to promote obesity. As expected, both groups of mice gained weight over the course of 16 weeks on HFD. However, the weight gain of DATR1^Null mice was significantly attenuated compared to NtsR1^flox/flox controls from weeks 12–16 (Figure 5A). This protection from developing diet induced obesity was not attributable to reduced caloric intake, since the amount of HFD consumed was similar between both groups (Figure 5B). Neither was the protection from weight gain due to enhanced energy expenditure, as both groups exhibited similar RER, VO₂ and CO₂ when analyzed in metabolic cages (>Supplementary Figure 1). DATR1^Null and NtsR1^flox/flox control mice also had similar levels of total open field locomotor activity (data not shown) although they spent less time in periphery. Additionally, DATR1^Null and NtsR1^flox/flox control mice exhibited similar AMPH-induced locomotor activity (Figures 5C,D). These data suggest that developmental deletion of NtsR1 from DA neurons provides some protection from diet-induced weight gain, but it is not mediated via behavioral alterations in feeding, movement or energy expenditure.

Developmental gene deletion often leads to compensatory actions that mask the importance of the gene product, which could have occurred in our developmental deletion of NtsR1 from DA neurons. Furthermore, DA neurons are distributed in several midbrain regions, hence, the method we used to delete NtsR1 from all DA neurons could not confirm the role of NtsR1 specifically via the DA neurons of the VTA. We reasoned that selectively inducing Cre in the VTA of adult NtsR1^flox/flox mice would permit adult-onset, VTA-specific deletion of NtsR1 that could be used to reveal the importance of NtsR1 via this brain region. To test the utility of this method, we administered bilateral VTA injections of AAV-GFP or AAV-Cre-GFP to Wt and NtsR1^flox/flox mice and ∼6 months later analyzed Ntsr1 mRNA expression by RNAScope (Figures 6A–F). Wt mice injected with AAV-GFP (Wt^GFP mice) or AAV-Cre-GFP (Wt^Cre–GFP mice) had similar levels of Ntsr1 expression in the VTA, confirming these methods did not delete Ntsr1 (Figures 6A–C). NtsR1^flox/flox mice injected with AAV-GFP also had intact Ntsr1 (VTAR1^GFP mice, Figure 6E). However, Ntsr1 expression was visibly absent in NtsR1^flox/flox mice that received AAV-Cre-GFP injections into the VTA, confirming sustained Ntsr1 deletion (Figure 6F). By contrast, both groups had comparable distributions of TH-expressing neurons (DA neurons) in the VTA (Figure 6G), suggesting intact DA systems. Taken together, these data confirm that this viral-mediated approach can site-specifically delete NtsR1 in adult mice but does not cause overt disruptions in the distribution of DA neurons.

To determine the impact of deleting NtsR1 from established VTA DA neurons we measured the body weight and food intake of VTAR1^GFP mice (with intact NtsR1 in VTA DA neurons) and VTAR1^Null mice (lacking NtsR1 in VTA DA neurons) on a normal chow diet (Figures 7A,B). Adult-onset deletion of NtsR1 from VTA neurons resulted in overall lower body weight (Figure 7A; y axis asterisks indicate overall virus effect) but did not reduce overall cumulative chow intake (Figure 7B). By contrast, no differences in body weight or feeding were observed in Wt mice injected with AAV-Cre-GFP or AAV-GFP (>Supplementary Figures 2A,B). These data suggest that NtsR1 contributes to modulating body weight, but it is not necessarily required to regulate homeostatic food intake. Given that food restriction enhances DA release in response to food rewards, and since NtsR1 expressing neurons are a subpopulation of all VTA DA neurons, we reasoned that deleting NtsR1 from adult VTA DA neurons might attenuate fasting-induced refeeding. Overnight fasted VTAR1^GFP and VTAR1^Null mice regained similar amounts of weight after re-feeding, but VTAR1^Null mice consumed significantly less chow overall, particularly 24 h after fasting compared to VTAR1^GFP controls (Figures 7C,D). These effects were not observed in Wt mice injected with AAV-Cre-GFP or AAV-GFP (>Supplementary Figures 3C,D), confirming that Cre-mediated recombination is necessary for the alterations in body weight and feeding behavior.

We next used open field chambers to assess whether adult-onset deletion of NtsR1 from VTA DA neurons altered DA-dependent locomotor activity and anxiety behaviors (Figures 8A–F). Wt mice injected with AAV-Cre-GFP or AAV-GFP exhibited comparable amounts of open field exploration, AMPH-induced locomotor activity, nestlet shredding, marble burying and EPM performance (>Supplementary Figures 4A–F). The VTAR1^Null and VTAR1^GFP control mice performed comparably to each other in these tests (Figures 8A–F). These data indicate that adult-onset deletion of NtsR1 from VTA DA neurons does not impair locomotor or anxiety responding in normal weight, chow-fed mice.

Lastly, we examined whether deletion of NtsR1 from adult VTA DA neurons impacts the susceptibility to develop diet-induced obesity. After completing assessment on chow diet, VTAR1^GFP and VTAR1^Null mice were switched to ad libitum HFD in their individual home cages over 8 wks (Figures 9A,B). Both groups gained weight on HFD, but the VTAR1^Null mice lacking NtsR1 in VTA neurons maintained modestly reduced body weight compared to controls (Figure 9A; y axis asterisks indicate overall virus effect). However, VTA specific deletion of NtsR1 did not reduce overall cumulative HFD intake (Figure 9B). These data suggest that NtsR1 in the established VTA contributes to body weight regulation, but independent of food intake, since loss of VTA NtsR1 expression had no impact on ad libitum HFD feeding. Lastly, we assessed whether obese mice with adult-deletion of NtsR1 from VTA DA neurons were able to appropriately coordinate energy need and feeding via testing their fasting-induced refeeding response. Overnight fasting caused VTAR1^GFP and VTAR1^Null mice to lose weight, and then to consume HFD when it was returned the following morning and regain weight. However, VTAR1^Null mice had modest reductions in post-fasting weight regain and HFD consumed compared to VTAR1^GFP control mice with intact NtsR1 HFD (Figures 9C,D). These effects were not observed in Wt mice on HFD injected with AAV-Cre-GFP or AAV-GFP (>Supplementary Figures 2A–D), indicating that viral injection itself does not impact energy balance. Overall, these data support that Cre-lox mediated adult-onset deletion of NtsR1 from VTA DA neurons modestly protects mice from diet-induced obesity and hunger-mediated feeding.