Dopamine production in neurotensin receptor 1 neurons is required for diet-induced obesity and increased day eating on a high-fat diet

Abstract

Objective: This study aimed to determine a dopaminergic circuit required for diet-induced obesity in mice.

Methods: We created conditional deletion mutants for tyrosine hydroxylase (TH) using neurotensin receptor 1 (Ntsr1) Cre and other Cre drivers and measured feeding and body weight on standard and high-fat diets. We then used an adeno-associated virus to selectively restore TH to the ventral tegmental area (VTA) Ntsr1 neurons in conditional knockout (cKO) mice.

Results: Mice with cKO of Th using Vglut2-Cre, Cck-Cre, Calb1-Cre, and Bdnf-Cre were susceptible to obesity on a high-fat diet; however, Ntsr1-Cre Th cKO mice resisted weight gain on a high-fat diet and did not experience an increase in day eating unlike their wild-type littermate controls. Restoration of TH to the VTA Ntsr1 neurons of the Ntsr1-Cre Th cKO mice using an adeno-associated virus resulted in an increase in weight gain and day eating on a high-fat diet.

Conclusions: Ntsr1-Cre Th cKO mice failed to increase day eating on a high-fat diet, offering a possible explanation for their resistance to diet-induced obesity. These results implicate VTA Ntsr1 dopamine neurons as promoting out-of-phase feeding behavior on a high-fat diet that could be an important contributor to diet-induced obesity in humans.

Figure 1.

Ntsr1-Cre deletion of TH and VTA restorations in male mice. (A) Body weight measurements of controls (n=10) and Ntsr1-Cre (n=8) (mixed-effects analysis genotype comparison p-value = 0.0005 [F{1, 8}=31.29]; Bonferroni multiple comparisons, weeks 0-3 p>0.9999, week 4 p=0.0018, weeks 5-8 p<0.0001) (B) Day- versus night-eating in Ntsr1-Cre cKO mice on SCD and HFD showed significant fixed effects (e.g., night vs. day p<0.0001, SCD vs HFD p =0.0026 whereas control versus cKO p=0.237). n=8-10 per group; 3-way ANOVA with Bonferroni’s multiple comparisons test **p<0.01). (C) High activity behaviors in the home-cage (jumping, rearing, hanging, and walking) on SCD and HFD (two-way ANOVA p=0.0159, with Bonferroni correction, control SCD versus control HFD p=0.354; Ntsr1 cKO SCD vs HFD p=0.849). (D) TH and DAT staining in the VTA of Ntsr1-Cre cKO and control VTA and cell counts (****p<0.0001, unpaired t test). (E) RFP and GFP staining in the VTA of an Ntsr1-Cre cKO mouse injected with Cre-dependent AAV-TH-RFP or AAV-GFP (p<0.05, unpaired t test). (F) Body weight measurements of Ntsr1 cKO mice injected with either AAV-GFP (n=8) or AAV-TH-RFP (n=5) on SCD and HFD. The effect of genotype was strong (p=0.0066 (F(1, 7)=14.56); there was a difference between percent of baseline body weight for weeks 5-8 (week 5 p=0.0275 week 6 p=0.0099, week 7 p=0.0035, week 8 p=0.0013 Bonferroni multiple comparisons). (G) Day- and night-eating in AAV-GFP and AAV-TH-RFP injected Ntsr1-Cre cKO mice (3-way ANOVA with Bonferroni’s multiple comparisons test, **p<0.01).