Cannabinoid CB1 receptor in dorsal telencephalic glutamatergic neurons drives overconsumption of palatable food and obesity

The study addresses how endocannabinoid signaling via cannabinoid type 1 receptors (CB1) in dorsal telencephalic glutamatergic neurons regulates hedonic feeding and contributes to diet-induced obesity (DIO). Obesity stems from chronic overnutrition and is associated with psychiatric and cognitive comorbidities. Feeding is regulated by homeostatic circuits (hypothalamus/brainstem) and hedonic circuits (mesolimbic dopamine), which interact extensively. Endocannabinoids are endogenous orexigenic signals; elevated endocannabinoid tone is linked to obesity, and CB1 blockade reduces body weight and metabolic disturbances in rodents and humans. Sensory processing, especially olfaction, critically contributes to food reward and is modulated by metabolic state. Prior work showed CB1 in dorsal telencephalic glutamatergic neurons mediates orexigenic effects during hunger and enhances odor detection of food, yet mice lacking CB1 in these neurons did not differ in body weight on low-fat diet. The current study tests whether CB1 in this neuronal population drives overconsumption of palatable food and the development of obesity, and explores the involvement of olfactory cortex.

Background literature indicates: (i) endocannabinoids and CB1 signaling orchestrate energy balance centrally and peripherally; obesity correlates with elevated endocannabinoid tone; pharmacological or genetic CB1 blockade reduces food reinforcement, motivation, and body weight in animal models and humans. (ii) Olfactory function influences hedonic evaluation of food and is modulated by metabolic state, affecting appetite and feeding behavior. (iii) CB1 in dorsal telencephalic glutamatergic neurons mediates orexigenic effects during fasting and regulates odor detection, linking hunger to enhanced olfactory processing. These findings frame the hypothesis that CB1 in cortical glutamatergic neurons, potentially within olfactory circuits, promotes hedonic overconsumption leading to obesity.

Animals: Male C57BL/6N, Glu-CB1-WT (CB1 floxed/floxed), and Glu-CB1-KO (CB1 floxed/floxed; heterozygous Nex-Cre) mice, 2–5 months old. Housing: 23 ± 1 °C, 12-h light/dark cycle. Diets: low-fat diet (LFD; 13.9 kJ/g: 4.0% fat, 21.1% protein, 56.6% carbohydrate), chocolate diet (16.1 kJ/g: 7.3% fat, 18.6% protein, 61.2% carbohydrate), high-fat diet (HFD; 21.1 kJ/g: 35.0% fat, 21.3% protein, 29.5% carbohydrate). Ethical compliance per EU 2010/63/EU and local committees.

Food intake and body weight: Measured weekly. Pair-feeding: After 4 weeks of ad libitum HFD, WT mice were daily given the average amount consumed by KO mice (starting week 5); body weight and intake recorded weekly.

Indirect calorimetry: 10–12-week-old males on HFD for 2–3 weeks were single-housed in TSE Systems metabolic chambers. After 48 h acclimation, VO2, VCO2, locomotor activity, food and water intake were recorded every 15 min for 48 h. Ad libitum HFD and water provided. Energy expenditure (kcal/h) analyzed with ANCOVA using body weight as covariate; respiratory exchange ratio (RER) and activity assessed.

Glucose and insulin tolerance tests: For GTT, overnight fast (12–16 h) followed by i.p. glucose 2 mg/g; tail blood glucose measured at set intervals. For ITT, 3–5 h fast, i.p. insulin 0.75 IU/kg; glucose measured over time.

Operant self-administration: Conducted during dark phase in Med Associates mouse operant chambers using chocolate-flavored pellets (14.4 kJ/g: 20.6% protein, 12.7% fat, 66.7% carbohydrate) as reinforcers. After 7 days on LFD or HFD, mice trained FR1 (1 h/day, 5 days) then FR5 (1 h/day, 5 days) with cue-light. Motivation assessed under progressive ratio (PR) with escalating response requirements (sequence: 1, 5, 12, 21, 33, 51, 75, 90, 120, 155, 180, 225, 260, 300, 350, 410, 465, 540, 630, 730, 850, 1000, 1200, 1500, 1800, 2100, 2400, 2700, 3000, 3400, 3800, 4200, 4600, 5000, 5500). Session max 5 h or until 1 h without responding; breaking point (BP) defined as last ratio completed. PR tested before and after 9 days chocolate withdrawal and following HFD withdrawal (5 consecutive PR sessions).

Fluorescence immunohistochemistry and confocal imaging: Coronal 40 µm cryosections incubated with rabbit anti-CB1 (1:500) or rat anti-HA (1:200). Imaging with Leica DMRA fluorescence microscope and spinning disk confocal (Nikon Ti; Yokogawa CSU-W1; Prime BSI sCMOS). Excitation: 488/561 nm; emission filters 525/30 and 570 LP. Tile scans with 20x/0.75 NA (pixel 670 nm, binning 2); zoomed 60x/1.2 NA water (pixel 111 nm). Images processed in Fiji/ImageJ.

RNA isolation and qPCR: Total RNA from olfactory cortex and brain punches. TaqMan mouse CB1 probe (Mm00432621_s1) for endogenous expression; SYBR Green primers recognizing mouse and rat CB1 transgene (forward TTC AAA CTG GGT GGG GTT AC; reverse GCC TGG TGA CGA TCC TCT TA) to assess AAV-mediated rescue.

Synaptosomal preparation: Olfactory bulb homogenized in 320 mM glucose TVEP buffer; sequential centrifugations to isolate crude synaptosomes and post-lysis synaptosomal fraction; fractions stored/analyzed.

Western blot: Olfactory cortex lysates in RIPA buffer; SDS-PAGE and transfer to nitrocellulose; probed for CB1 (1:1000), synaptotagmin (1:500), and β-actin (1:1000); HRP-conjugated secondaries; ECL detection; densitometry with Fusion/Bio-1D; CB1 normalized to β-actin.

Olfactory habituation test: Novel odor (butter-vanillin in 50% mineral oil); mineral oil as control. Food removed 2 h prior. 10 µl applied to cotton swab. Sequence: 30-min habituation with swab; presentation of solvent; then two 3-min presentations of odor (Q1 and Q2) with 5-min inter-trial intervals. Exploration defined as nose within <1 cm or touching swab.

Buried food-seeking test: Mice habituated to chocolate pellets; food removed 2 h prior. Habituation 5 min in test cage with fresh bedding; chocolate pellet buried 2 cm beneath bedding in a corner; mouse placed in opposite corner; latency to retrieve pellet recorded.

AAV-mediated CB1 re-expression: Adult males received bilateral stereotaxic injections of AAV-stop-HA-rCB1 or AAV-stop-empty into medial (+2.6 AP, ±0.5 ML, −4.0 DV) and lateral (+2.3 AP, ±1.5 ML, −4.2 DV) anterior olfactory nucleus (AON) of Glu-CB1-KO and Glu-CB1-WT mice during week 4 of HFD. Verification of HA-CB1 expression by FIHC and qPCR in AON and OB; outcomes assessed 8 weeks post-injection.

Statistics: Data as mean ± SEM. Unpaired two-tailed Student’s t-test, one-/two-/three-way ANOVA with appropriate post hocs (Bonferroni, Sidak, Newman–Keuls), repeated measures ANOVA for longitudinal data. ANCOVA for energy expenditure (kcal/h) and food intake (g/week) with body weight as covariate. PR FR1/FR5 analyzed by three-way repeated measures ANOVA; BP by two-way ANOVA (diet × genotype). Significance p < 0.05. Software: GraphPad Prism 5.0, SPSS v15.0.

• CB1 deletion in dorsal telencephalic glutamatergic neurons protects against DIO: Glu-CB1-KO mice on HFD gained less weight and consumed less food than Glu-CB1-WT. Food intake difference remained significant after ANCOVA controlling for body weight (genotype F(1,272)=25.996, p<0.001).
• Metabolic tolerance improved in KO: HFD-fed Glu-CB1-KO mice were protected from glucose intolerance and insulin resistance relative to WT (GTT and ITT; AUC difference significant, figure-reported).
• Pair-feeding eliminated genotype weight differences: When WT mice were pair-fed to KO intake starting week 5 of HFD, body weight differences were blunted, indicating reduced intake as primary driver of leanness.
• No early differences in energy expenditure or activity: During weeks 2–3 of HFD, energy expenditure (ANCOVA with body weight as covariate), RER, and locomotor activity showed no significant genotype differences.
• No phenotype on low-fat diet: Under LFD, no differences in body weight or intake between genotypes.
• Palatable diet paradigms reveal hedonic feeding deficit in KO: • Isocaloric chocolate diet: KO showed decreased weight gain and reduced intake vs WT; intake difference significant by ANCOVA (genotype F(1,157)=8.575, p<0.01). • Free-choice (LFD + HFD ad libitum): Both preferred HFD, but WT had higher initial preference and consumed more HFD, leading to greater weight gain; intake difference significant by ANCOVA (genotype F(1,226)=5.269, p<0.05).
• Motivation after HFD withdrawal: Short-term LFD or HFD exposure did not alter PR breaking point among groups. After HFD withdrawal, WT displayed increased motivation (higher BP) for chocolate pellets across five sessions, with significant differences in the last two sessions; this sensitization was absent in KO, indicating protection from HFD withdrawal–induced reward enhancement.
• CB1 expression reduced in olfactory circuits of KO: FIHC and western blots showed strong reduction of CB1 in OB, especially granule cell layer (GCL); qPCR showed significantly decreased CB1 mRNA in AON (mean diff −0.845, 95% CI −1.22 to −0.471, p<0.001). OB total and synaptosomal CB1 protein levels were ~4-fold lower in KO vs WT (genotype F(1,12)=15.55, p<0.01).
• Odor-guided behavior: Under LFD or early HFD, no genotype differences in odor habituation. When obesity emerged, obese HFD-fed WT showed higher Q1 exploration (enhanced odor detection) than lean KO; similar patterns observed on chocolate and free-choice diets. Body weight correlated with exploration time in WT (r(33)=0.508, p=0.002). In fasting, LFD-fed WT outperformed HFD-fed WT; genotype differences under HFD were blunted by fasting.
• Buried food-seeking: HFD-fed KO exhibited longer latency to find buried chocolate pellets vs obese WT (t(28)=0.224, p=0.034), consistent with reduced odor detection and/or motivation; similar phenotype on LFD.
• AAV rescue in olfactory cortex partially restores phenotype: Re-expressing CB1 in AON glutamatergic neurons of KO mice (AAV-stop-HA-CB1) increased cumulative weight gain and food intake vs AAV-empty KO, blunting differences with WT. Two-way ANOVA showed genotype×AAV interactions for cumulative weight gain (F(1,27)=4.262, p=0.049), food intake (F(1,28)=7.684, p=0.001), and GTT AUC (F(1,22)=4.466, p=0.046). Rescue also improved Q1 odor exploration partially.

Findings demonstrate that CB1 in dorsal telencephalic glutamatergic neurons is a key driver of hedonic overconsumption of palatable diets leading to obesity. The protection from DIO in Glu-CB1-KO mice stemmed primarily from reduced food intake rather than increased energy expenditure, as confirmed by pair-feeding and metabolic phenotyping. The absence of phenotype on LFD underscores a specific role in hedonic, not homeostatic, feeding. Behavioral assays revealed that HFD withdrawal enhances motivation for palatable rewards in WT but not KO mice, implicating CB1 in cortical excitatory neurons in reward sensitization after obesogenic diet exposure. Mechanistically, substantial reductions of CB1 at presynaptic sites of olfactory circuits (AON projections to OB) in KO mice, coupled with diminished odor detection and food-seeking behaviors, link olfactory processing to hedonic feeding control. Partial restoration of obesity, intake, and odor responses by targeted CB1 re-expression in olfactory cortex supports a causal contribution of olfactory cortical CB1 to overfeeding and DIO. Comparisons with other CB1 manipulations indicate complementary mechanisms across neuronal populations: forebrain/sympathetic CB1 can increase energy storage and thermogenesis without altering intake, whereas cortical glutamatergic CB1 chiefly enhances palatable food consumption, shaping the development of obesity through hedonic pathways.

CB1 expression in dorsal telencephalic excitatory neurons critically regulates hedonic feeding and promotes overconsumption of palatable food, driving diet-induced obesity. Deleting CB1 in these neurons reduces intake and protects against obesity and metabolic impairments without altering energy expenditure, and prevents HFD withdrawal–induced increases in food motivation. Olfactory circuits are a key locus: CB1 loss reduces presynaptic CB1 in OB and impairs odor-guided behavior; targeted CB1 re-expression in olfactory cortex partially restores weight gain, intake, and odor detection. These results position cortical glutamatergic CB1, particularly within olfactory pathways, as a pivotal modulator of hedonic feeding. Future work should dissect contributions of other reward-related regions (striatum, amygdala, hypothalamus), determine causal directionality between olfactory changes and obesity, and evaluate sex-specific effects.

• Regional specificity: Although CB1 deletion targets dorsal telencephalic glutamatergic neurons, multiple cortical and subcortical projection areas are affected; contributions from regions beyond olfactory cortex (e.g., striatum, amygdala, hypothalamus) cannot be excluded.
• Causality between olfaction and obesity: While olfactory CB1 rescue altered both odor detection and obesity phenotypes, it remains unclear whether enhanced olfaction drives overconsumption or reflects consequences of weight gain.
• Sex limitation: Experiments were performed in male mice; potential sex differences in CB1’s role in hedonic feeding were not assessed.
• Behavioral ceiling/floor effects: Under ad libitum HFD access, motivation for pellets may be limited, potentially masking genotype differences prior to diet withdrawal.
• Partial rescue: AAV-mediated CB1 re-expression produced partial rather than complete restoration, which may reflect targeting efficiency, expression levels, or involvement of additional circuits.