Lateral hypothalamic glutamatergic inputs to VTA glutamatergic neurons mediate prioritization of innate defensive behavior over feeding

Abstract

The lateral hypothalamus (LH) is involved in feeding behavior and defense responses by interacting with different brain structures, including the Ventral Tegmental Area (VTA). Emerging evidence indicates that LH-glutamatergic neurons infrequently synapse on VTA-dopamine neurons but preferentially establish multiple synapses on VTA-glutamatergic neurons. Here, we demonstrated that LH-glutamatergic inputs to VTA promoted active avoidance, long-term aversion, and escape attempts. By testing feeding in the presence of a predator, we observed that ongoing feeding was decreased, and that this predator-induced decrease in feeding was abolished by photoinhibition of the LH-glutamatergic inputs to VTA. By VTA specific neuronal ablation, we established that predator-induced decreases in feeding were mediated by VTA-glutamatergic neurons but not by dopamine or GABA neurons. Thus, we provided evidence for an unanticipated neuronal circuitry between LH-glutamatergic inputs to VTA-glutamatergic neurons that plays a role in prioritizing escape, and in the switch from feeding to escape in mice.

Fig. 1. Release of glutamate from LH-VGluT2 fibers within the VTA induces aversion.

A LH injection of AAV1-DIO-eYFP or AAV1-DIO-ChR2-eYFP and VTA optic fiber. B Timeline for place conditioning experiments. C Track plot for an eYFP and a ChR2-eYFP mouse during the first day of photostimulation conditioning (D1). D ChR2-eYFP mice (n = 11) spent significantly less time in the laser-paired chamber than eYFP control mice (n = 9) during and after photostimulation sessions (group × chamber × experimental phase: F_(12,216) = 21.46, P < 0.00001, ANOVA with Newman-Keuls post hoc test). * P < 0.05, ** P < 0.01, *** P < 0.001, against pretest; + P < 0.05, ++ P < 0.01, +++ P < 0.001, against ChR2-eYFP. Light-blue rectangles indicate photostimulation. Data are presented as mean values ± SEM. Source data are provided as a Source Data file.

Fig. 2. Release of glutamate from LH-VGluT2 fibers within the VTA induces aversion and escape attempts mediated by the VTA activation of glutamate receptors.

A LH injection of AAV1-DIO-ChR2-eYFP, intra-VTA microinjection, and VTA optic fiber. B Timeline for microinjections experiment. C Intra-VTA administration of NMDA- (AP5) and AMPA- (CNQX) receptor antagonists decreased avoidance responses induced by VTA photostimulation of LH-VGluT2 fibers in ChR2-eYFP mice (n = 7; chamber x experimental phase: F_(4,24) = 5.09, P = 0.004, ANOVA with Newman-Keuls post hoc test). * P < 0.05, ** P < 0.01, against pretest; + P < 0.05, ++ P < 0.01, against aCSF. D Intra-LH administration of lidocaine did not affect avoidance responses induced by VTA photostimulation of LH-VGluT2 fibers in ChR2-eYFP mice (n = 8; chamber x experimental phase: F_(4,28) = 43.47, P < 0.00001, ANOVA with Newman-Keuls post hoc test). ** P < 0.01, *** P < 0.001, against pretest. E VTA photostimulation of LH-VGluT2 fibers induced escape attempts (measured as jumps) in ChR2-eYFP mice (n = 9; eYFP mice: n = 9; eYFP: X²₉ = 0.00, P = 1; ChR2-eYFP: X²₉ = 18.00, P = 0.001, Friedman ANOVA). *** P < 0.001, against laser off, +++ P < 0.001, against eYFP mice. F Intra-VTA administration of AP5 + CNQX decreased escape attempts (jumps) induced by VTA photostimulation of LH-VGluT2 fibers in ChR2-eYFP mice (n = 5; treatment x experimental phase: F_(2,8) = 160.57, P < 0.00001, ANOVA with Newman-Keuls post hoc test). *** P < 0.001, against laser off; +++ P < 0.001, against AP5 + CNQX. G Intra-VTA administration of lidocaine did not affect escape attempts (jumps) induced by VTA photostimulation of LH-VGluT2 fibers in ChR2-eYFP mice (n = 7; treatment x experimental phase: F_(2,14) = 0.86, P = 0.45, n.s, ANOVA). Light-blue rectangles indicate photostimulation. Data are presented as mean values ± SEM. Source data are provided as a Source Data file.

Fig. 3. Release of glutamate from LH-VGluT2 fibers within the VTA disrupts feeding behavior.

A LH injection of AAV1-DIO-eYFP, AAV1-DIO-ChR2-eYFP or AAV1-DIO-Halo-eYFP and VTA optic fiber. B Food sated ChR2-eYFP mice (n = 10) ate significantly less food than eYFP control mice (n = 9) during VTA photostimulation of LH-VGluT2 fibers regardless of stimulation frequency (group: F_(1,17) = 11.06, P = 0.004, ANOVA with Newman-Keuls post hoc test). C Timeline for feeding test in food restricted mice. D ChR2-eYFP mice (n = 9) eating standard chow or palatable pellets (ChR2-eYFP (P), n = 11) showed a higher latency to eating initiation during VTA laser stimulation than Halo-eYFP (n = 9) or eYFP control mice (n = 6) (group × trial: F_(3,31) = 19.53, P < 0.00001, ANOVA with Newman-Keuls post hoc test). * P < 0.05, *** P < 0.001, against eYFP mice, +++ P < 0.001, against laser off. E ChR2-eYFP mice (n = 9) presented with standard chow or palatable pellets (ChR2-eYFP (P), n = 11) ate less during VTA laser stimulation than Halo-eYFP (n = 9) or eYFP control mice (n = 6; group x trial: F_(3,31) = 16.21, P < 0.00001, ANOVA with Newman-Keuls post hoc test). *** P < 0.001, against eYFP mice; +++ P < 0.001, against laser off. F Timeline to test disruption of ongoing feeding behavior in food restricted mice. G VTA photostimulation of LH-VGluT2 fibers for 30 s disrupted ongoing eating and significantly increased the time to resume eating in ChR2-eYFP mice (n = 6) but not in eYFP control mice (n = 6; group x experimental phase: F_(2,20) = 26.44, P < 0.00001, ANOVA with Newman-Keuls post hoc test). *** P < 0.001, against eYFP, +++ P < 0.001, against laser on. Light-blue rectangles indicate photostimulation, orange rectangles indicate either photostimulation or photoinhibition. Data are presented as mean values ± SEM. Source data are provided as a Source Data file.

Fig. 4. LH-VGluT2 neurons innervating the VTA signal the presence of a predator resulting in feeding disruption.

A VTA injection of retrograde vector HSV-LS1L-GCaMP6s and LH photometry fiber. B Mice were tested in the presence of an anesthetized rat. C Whole session recording of LH-VGluT2 neurons projecting to the VTA showing approaches to the anesthetized rat (top); heatmap of Ca²⁺ activity over successive rat approach trials (middle); cell population responses to rat approach onset showing increases in Ca²⁺ activity in LH-VGluT2 neurons projecting to the VTA (bottom). D Population Ca²⁺ activity (+SEM) in LH-VGluT2 neurons projecting to VTA during rat approach onset (n = 10). E AUC for Ca²⁺ activity in LH-VGluT2 neurons projecting to VTA before (−2 to 0 s) and after (0 to 2 s) onset of rat approach (n = 10; t₍₁₈₎ = −2.38, ∗P = 0.03, two-tailed t test). Food restricted Halo-eYFP and eYFP mice exposure to both anesthetized rat and food (F) and test timeline (G). H Food intake was significantly reduced in Halo-eYFP (n = 6) and eYFP control mice (n = 6) when presented with an anesthetized rat in 15-min experimental sessions. Photoinhibition restored feeding to baseline levels in Halo-eYFP mice (group x experimental phase: F_(2,20) = 3.52, P = 0.049, ANOVA with Newman-Keuls post hoc test). ** P < 0.01, against baseline, + P < 0.05, against eYFP. Green rectangle indicates photoinhibition. Data are presented as mean values ± SEM. Source data are provided as a Source Data file.

Fig. 5. VTA-VGluT2 neurons signal the presence of a predator resulting in feeding disruption.

A VTA injection of AAV-Flex-GCaMP6s and photometry fiber. B Mice were tested in the presence of an anesthetized rat. C Whole session recording of VTA-VGluT2 neurons showing approaches to the anesthetized rat (top); heatmap of Ca²⁺ activity over successive rat approach trials (middle); cell population responses to rat approach onset showing increases in Ca²⁺ activity in VTA-VGluT2 neurons (bottom). D Population Ca²⁺ activity (+SEM) in VTA-VGluT2 neurons during rat approach onset (n = 8). E AUC for Ca²⁺ activity in VTA VGluT2 neurons before (−2 to 0 s) and after (0 to 2 s) onset of the rat approach (n = 8; t₍₁₄₎ = −2.40, ∗P = 0.03, two-tailed t test). F VTA injection of AAV1-Flex-taCasp3 or AAV1-DIO-mCherry. Low (G) and high (H–H”) magnification of VTA from a control mouse (injected with AAV1-DIO-mCherry) showing neurons expressing VGluT2 mRNA (red) intermixed with TH-immunoreactive neurons (TH-IRs; green). Low (I) and high (J–J”) magnification of VTA from a mouse injected with AAV1-Flex-taCasp3 showing TH-IRs and lack of VGluT2 mRNA. K VTA-VGluT2 neurons are present in control mice (1061.89 ± 70.99; 3 mice) but are infrequent in caspase mice (228.33 ± 69.52; 3 mice; 7 sections per mouse; t₍₄₎ = 8.39, two-tailed t test). ** P = 0.001, against control mice. Food restricted caspase and control mice exposure to both anesthetized rat and food (L) and test timeline (M). N Food intake was significantly reduced in control mice (n = 12) but not in caspase mice (n = 12) when presented with an anesthetized rat (t₍₂₂₎ = −3.08, two-tailed t test). ** P = 0.005, against control mice. Data are presented as mean values ± SEM. Source data are provided as a Source Data file.

Fig. 6. LH-VGluT2 neurons establish synapses on VTA-VGluT2 neurons.

A LH injection of AAV2-DIO-ChR2-mCherry and VTA injection of AAV2-DIO-eYFP. B A VTA-VGluT2 soma (green outline with GFP detected by gold particles, arrowhead) making asymmetric synapses (green arrows) with two axon terminals (AT1–2, red outlines) from LH-VGluT2 neurons co-expressing mCherry (scattered dark material) and VGluT2 (gold particles). C Corresponding diagram from the image shown in (B). D AT1 at higher magnification. E AT2 at higher magnification.

Fig. 7. LH-VGluT2 neurons innervating VTA-VGluT2 neurons mediate increased escape responses and predator-induced disruption of feeding.

A LH injection of AAV2-DIO-ChR2-eYFP, VTA injection of AAV1-Flex-taCasp3 or AAV1-mCherry, and VTA optic fiber. B Time spent in the periphery and center zones of an open field arena were similar between control mCherry (n = 8) and VTA-VGluT2 ablated (n = 9) mice before, during and after VTA photostimulation of LH-VGluT2 fibers (group x zone x experimental phase: F_(2,30) = 0.07, P = 0.93, n.s., ANOVA with Newman-Keuls post hoc test). *** P < 0.001, against center for each experimental group. C mCherry mice (n = 8), but not VTA-VGluT2 ablated mice (n = 9), significantly increased the distance traveled in an open field arena during VTA photostimulation of LH-VGluT2 fibers (group x experimental phase: F_(2,30) = 16.47, P < 0.00001, ANOVA with Newman-Keuls post hoc test). +++ P < 0.001, against first period of laser off. D mCherry mice (n = 8), but not VTA-VGluT2 ablated mice (n = 9), significantly increased their open field speed during VTA photostimulation of LH-VGluT2 fibers (group x experimental phase: F_(2,30) = 10.63, P = 0.0003, ANOVA with Newman-Keuls post hoc test). + P < 0.05, +++ P < 0.001, against first period of laser off. E Ablation of VTA-VGluT2 neurons reversed the decrease in food intake induced by the presentation of an anaesthetized rat alone or in combination with VTA photostimulation of LH-VGluT2 fibers and partially reversed the decrease in food intake induced by VTA photostimulation of LH-VGluT2 fibers alone (mCherry, n = 11; Caspase, n = 10; group × experimental phase: F_(3,57) = 3.66, P = 0.02, ANOVA with Newman-Keuls post hoc test). * P < 0.05, *** P < 0.001, against mCherry mice; + P < 0.05, +++ P < 0.001, against baseline for each experimental group. Light-blue rectangles indicate photostimulation. Data are presented as mean values ± SEM. Source data are provided as a Source Data file.