Rapid binge-like eating and body weight gain driven by zona incerta GABA neuron activation

Abstract

The neuronal substrate for binge eating, which can at times lead to obesity, is not clear. We find that optogenetic stimulation of mouse zona incerta (ZI) γ-aminobutyric acid (GABA) neurons or their axonal projections to paraventricular thalamus (PVT) excitatory neurons immediately (in 2 to 3 seconds) evoked binge-like eating. Minimal intermittent stimulation led to body weight gain; ZI GABA neuron ablation reduced weight. ZI stimulation generated 35% of normal 24-hour food intake in just 10 minutes. The ZI cells were excited by food deprivation and the gut hunger signal ghrelin. In contrast, stimulation of excitatory axons from the parasubthalamic nucleus to PVT or direct stimulation of PVT glutamate neurons reduced food intake. These data suggest an unexpected robust orexigenic potential for the ZI GABA neurons.

Fig. 1.. Optogenetic activation of ZI GABA neurons rapidly evokes binge-like eating.

(A) Red fluorescent image shows restricted expression of ChIEF-tdTomato in the ZI after AAV-ChIEF-tdTomato was bilaterally injected into the ZI of VGAT-Cre mice. Scale bar, 500 μm. (B) Optogenetic activation with varying frequency of a ZI GABA neuron in a brain slice. (C) Schematic illustration showing the location of optical fiber tips implanted above the ZI on both sides of the brain. (D) High-fat food intake during 10 min and four times 10 min from control mice with tdTomato and mice with ChIEF-tdTomato, both with ZI expression. For the 10-min trial, continuous light stimulation (10 ms, 20 Hz) was supplied to the ZI. For the 4 ×10-min trial, 10-min light stimulation (10 ms, 20 Hz) was followed by 30 min without stimulation, repeated four times. (E) High-fat food intake over 10 min and four times 10-min trial as a percentage of unstimulated 24-hour intake (100%). (F) Action potentials evoked by 100-pA current injection in ZI GABA neurons in brain slices of mice fed or fasted for 24 hours. (G) Firing rate at different levels of current injection from ZI GABA neurons in brain slices of mice fed or fasted for 24 hours. (H) Excitatory postsynaptic currents (EPSCs) in ZI GABA neurons of mice fed or fasted for 24 hours. (I) EPSC frequency from ZI GABA neurons of mice fed (n = 12 cells from each of four mice) or mice fasted for 24 hours (n = 13 cells from four mice). (J) EPSC amplitude from ZI GABA neurons in mice fed (n = 12 cells from four mice) or mice fasted for 24 hours (n = 13 cells from four mice). (K) Ghrelin (100 nM) excites a ZI GABA neuron. (L) Ghrelin depolarizes ZI GABA neurons. (M) Ghrelin increases the firing rate of ZI GABA neurons. Statistical analysis for comparison between two groups: Two-way analysis of variance (ANOVA) with Bonferroni post hoc comparison for (D) and (E); unpaired t test for (G), (I), (L) and (M). *P < 0.05; **P < 0.01; ***P < 0.001.

Fig. 2.. Optogenetic activation of ZI GABA axon terminals in PVT rapidly evokes eating with preference for high-fat food.

(A) Anterograde mapping of ZI GABA neuron projections to PVT. (Top left) Schematic shows bilateral injection of AAV-ChIEF-tdTomato into ZI and placement of fiber optic tip above PVT. (Bottom left) Red fluorescent image shows strong projection to PVT from ZI-VGAT-ChIEF-tdTomato neurons. D3V, dorsal third ventricle. Scale bar, 100 μm. (Right) ZI axons in PVT. Scale bar, 5 μm. (B) Retrograde mapping of presynaptic neurons to PVT glutamate neurons. (Top left) Schematic shows the strategy for tracing presynaptic ZI projections to PVT glutamate neurons. TVA, the avian tumor virus receptor A; RVdg, glycoprotein–deleted rabies virus (RV). (Top right) Selective expression of RV-GFP (green) and TVA-mCherry (red) in PVT. Scale bar, 300 μm. (Bottom) PVT neurons detected with TVA-mCherry (left), RV-GFP (middle) and merged image (right) shows originating cells (yellow, expressing both GFP and mCherry). Scale bar, 20 μm. (C) RV-labeled presynaptic neurons in ZI. LH, lateral hypothalamus. Scale bars: top, 300 μm; bottom, 20 μm. (D) Optogenetically evoked inhibitory postsynaptic currents (IPSCs) of PVT vGlut2 neurons at 1, 5, 10, and 20 Hz (membrane potential clamped at −40 mV). In bicuculline (Bic, 30 μM), 1-Hz pulses evoked no obvious current. (E) Photostimulation (10 ms, 20 Hz) of ZI-VGAT-ChIEF neuron terminals in PVT increases food intake during 10-min trial. (F) Photostimulation of PVT has no effect on food intake of VGAT-Cre mice after control AAV-tdTomato injection into ZI. (G) Cumulative time during eating by VGAT-ChIEF mice during 10-min photostimulation (10 ms, 20 Hz). (H) Food intake induced by photostimulation is greater in first 10-min trial, and reduced in second and third trial, with a 5-min interval between photostimulations. (I) VGAT^ZI-PVT photostimulation increased preference for high-fat food. (J) Food intake for 24 hours with photostimulation of ZI VGAT neurons or VGAT^ZI-PVT terminals. For 4 ×10-min trial, 10-min light stimulation (10 ms, 20 Hz) followed by 30-min no stimulation, repeated four times. (K) Bic attenuates optogenetic stimulation of food intake in PVT. Statistical analysis for comparison between two groups: two-way ANOVA with Bonferroni post hoc comparison for (E) and (J); unpaired t test for (G); one-way ANOVA repeated measure with Bonferroni post hoc comparison for (H); two-way ANOVA repeated measure with Bonferroni post hoc comparison for (I) and (K). n.s., not significant; *P < 0.05; **P < 0.01; ***P < 0.001.

Fig. 3.. Binge-like eating evoked by optogenetic activation of ZI GABA axon terminals produces rapid increase in body weight and shows positive hedonic association.

(A) Photostimulation of ChIEF-expressing ZI axons reduces activity of PVT glutamate neuron in brain slice during repeated 10-s photostimulations (10 ms, 20 Hz) with 30-s rest interval. (B) Latency for representative mouse to rapidly initiate feeding in response to photostimulation over 30 consecutive trials. Photostimulation protocol same as in (A). (C) Latency of feeding initiation using 10-s photostimulation (10-ms pulses) at 10, 20, and 40 Hz. (D) Photostimulation increases time in lit chamber during light-dark conflict test. (E) Light-dark conflict test shows photostimulation increases high-fat food intake in brightly lit chamber. (F) Real-time place-preference data show tracks of control (left) and ZI VGAT-ChIEF mouse (right) in photostimulation-paired (photostim.) and nonpaired chambers. (G) Time (%) that control and ChIEF-tdTomato mice stay in photostimulation-paired chamber during a 20-min trial. (H) Daily food intake of control tdTomato and ChIEF-tdTomato mice with photostimulation of VGAT^ZI-PVT terminals for 5 min (20 Hz) every 3 hours repeated over 14 days (shaded box) then continued without photostimulation (unshaded). F_1,230 = 343.9, P < 0.0001, two-way ANOVA. (I) Body weight of tdTomato control and ChIEF-tdTomato mice with a 5-min (20-Hz) photostimulation of VGAT^ZI-PVT terminals every 3 hours repeated over 14 days (shaded box), then continued without photostimulation (unshaded). Data were from the same mice tested in (D). F_1,230 = 73.45, P < 0.0001, two-way ANOVA. (J) Weekly food intake from control and ZI VGAT neuron ablation. Ablation versus control: F_1,99 = 55.84, P < 0.0001. (K) Body weight gain of mice from control group and ZI VGAT neuron ablation. Ablation versus control: F_1,99 = 60.12, P < 0.0001. Statistical analysis for comparison between two groups: Paired t test for (D) and (E); unpaired t test for (G); two-way ANOVA with Bonferroni post hoc comparison for (J) and (K). *P < 0.05; **P < 0.01; ***P < 0.001.

Fig. 4.. Direct and indirect activation of glutamatergic PVT neurons postsynaptic to ZI axons inhibits food intake.

(A) Fluorescent image shows restricted expression of ChIEF-tdTomato in PVT after AAV Cre-dependent ChIEF-tdTomato injected into PVT of vGlut2-Cre mice. Scale bar, 100 μm. (B) Implanted optical fiber tip above PVT. (C) Photostimulation (10-ms pulses) at 5, 10, and 20 Hz excites vGlut2-ChIEF neurons in PVT. (D) Photostimulation of PVT vGlut2-ChIEF neurons decreases intake of normal, sweet, and high-fat (HF) foods over 1 hour. (E) Latency in eating cessation in control and PVT vGlut2-ChIEF mice. (F) (Top) Four consecutive 10-min trial periods with first and third paired with stimulation (10 ms, 20 Hz). Food intake of control and vGlut2-ChIEF mice with partial food restriction during each period. (G) Total food intake is reduced by photostimulation in PVT of vGlut2-ChIEF mice with food restriction over 40 min compared with controls. (H) Food intake ratio of photostimulated to unstimulated periods in food-restricted control and vGlut2-ChIEF mice. (I) (Left) Schematic shows strategy for tracing presynaptic PSTh projections to PVT glutamate neurons. (Middle and right) Green images show fluorescent RV-labeled presynaptic PSTh neurons. cp, Cerebral peduncle. (J) Schematic shows bilateral AAV-ChIEF-tdTomato injections in parasubthalamic nucleus (PSTh) and placement of fiber optic tip above PVT. (Left bottom) Restricted expression of ChIEF-tdTomato in PSTh after AAV-ChIEF-tdTomato injection in vGlut2-Cre mice. Scale bar, 500 μm. (Right bottom) PSTh vGlut2 neurons project to PVT. Scale bar, 100 μm. (Top) Higher magnification shows PSTh vGlut2 neuron terminals in PVT. Scale bar, 5 μm. (K) EPSCs in PVT glutamate neurons evoked by photostimulation (10-ms pulses) at 1, 5, 10, and 20 Hz (membrane potential clamped at −70 mV). In 2-amino-5-phosphonopentanoic acid (AP5) (50 μM) and 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) (10 μM), photostimulation of 1 Hz evoked no obvious current. (L) Photostimulation (10 ms, 20 Hz) of excitatory vGlut2 neuron terminals in PVT projecting from the PSTh decreased food intake. Statistical analysis for comparison between two groups: paired t test for (D) and (L); one-way ANOVA with Bonferroni post hoc comparison for (E) and (F); unpaired t test for (G) and (H). *P < 0.05; **P < 0.01; ***P < 0.001.