Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2020 Jan 2;130(1):126-142.
doi: 10.1172/JCI130340.

Extrahypothalamic GABAergic nociceptin-expressing neurons regulate AgRP neuron activity to control feeding behavior

Mark A Smith  1 , Agharul I Choudhury  2 , Justyna A Glegola  2 , Paulius Viskaitis  2 , Elaine E Irvine  2 , Pedro Caldas Custodio de Campos Silva  2 , Sanjay Khadayate  2 , Hanns Ulrich Zeilhofer  3   4 , Dominic J Withers  1   2
Affiliations

Extrahypothalamic GABAergic nociceptin-expressing neurons regulate AgRP neuron activity to control feeding behavior

Mark A Smith et al. J Clin Invest. .

Abstract

Arcuate nucleus agouti-related peptide (AgRP) neurons play a central role in feeding and are under complex regulation by both homeostatic hormonal and nutrient signals and hypothalamic neuronal pathways. Feeding may also be influenced by environmental cues, sensory inputs, and other behaviors, implying the involvement of higher brain regions. However, whether such pathways modulate feeding through direct synaptic control of AgRP neuron activity is unknown. Here, we show that nociceptin-expressing neurons in the anterior bed nuclei of the stria terminalis (aBNST) make direct GABAergic inputs onto AgRP neurons. We found that activation of these neurons inhibited AgRP neurons and feeding. The activity of these neurons increased upon food availability, and their ablation resulted in obesity. Furthermore, these neurons received afferent inputs from a range of upstream brain regions as well as hypothalamic nuclei. Therefore, aBNST GABAergic nociceptin neurons may act as a gateway to feeding behavior by connecting AgRP neurons to both homeostatic and nonhomeostatic neuronal inputs.

Keywords: Melanocortin; Metabolism; Neuroendocrine regulation; Neuroscience; Obesity.

PubMed Disclaimer

Conflict of interest statement

Conflict of interest: The authors have declared that no conflict of interest exists.

Figures

Figure 1
Figure 1. Photostimulation of aBNST GABAergic axons in the arcuate nucleus suppresses feeding.
(A) Diagram illustrating injections of the aBNST (red) with ChR2-mCherry. (B) Mosaic images (n = 35) of aBNST (enlarged inset on the right) and arcuate nucleus (C) from Vgat-Cre mice with ChR2-mCherry AAV injected into the aBNST. Scale bars: 200 μm (low-magnification images) and 20 μm (high-magnification images). (D) Diagram of Vgat-Cre mice crossed with Pomc-GFP or Npy-hrGFP mice with ChR2-mCherry in the aBNST. (E) Photostimulated synaptic currents from an arcuate NPY neuron in the absence (black) and presence (red) of 20 μM bicuculline (n = 10). (F) Synaptic currents in arcuate POMC (n = 9 neurons from 3 mice) and NPY (n = 10 neurons from 3 mice) neurons. Data represent the mean ± SEM. *P < 0.05 [unpaired t test, t (17) = 2.15, P = 0.046]. (G) Voltage traces (expanded below) from a NPY neuron during photostimulation of aBNST axons (n = 10). Vm, membrane potential. (H) Action potential frequency in NPY neurons (n = 10 neurons from 3 mice) before (baseline) and during photostimulation of aBNST axons. Data represent the mean ± SEM. **P < 0.01, by paired t test, t (9) = 3.33, P = 0.009. (I) Diagram illustrating injection of the aBNST with ChR2-mCherry and optical fiber implantation into the arcuate nucleus. (J) Cumulative food intake in fasted Vgat-Cre mice during (shaded area) and after photostimulation of ChR2-mCherry–expressing axons from the aBNST. Food intake was measured in mice without (red) and with (blue) photostimulation. Data represent the mean ± SEM. n = 25 mice. Two-way repeated-measures ANOVA [interaction: f (6,288) = 9.58, P < 0.0001; stimulation: f (1,48) = 3.68, P = 0.06]. *P < 0.05 and **P < 0.01, by Sidak’s post hoc test. (K) Food intake at 4 hours from nonstimulated and stimulated mice. Data represent the mean ± SEM. n = 25 mice. *P < 0.05, by paired t test, t (24) = 2.46, P = 0.021. 3V, third ventricle; a.c., anterior commissure.
Figure 2
Figure 2. Nociceptin is expressed in a subpopulation of Vgat neurons.
(A) Voltage traces (expanded below) from a NPY neuron during 0.5 μM nociceptin application, where indicated (n = 15). (B and C) Nociceptin-induced change in membrane potential (ΔVm) (B) and input resistance (C) in NPY neurons. Data represent the mean ± SEM. n = 15 neurons from 8 mice. *P < 0.05, Vm: 1-sampled t test, t (14) = 2.88, P = 0.012; **P < 0.01, input resistance: 1-sample t test, t (14) = 4.05, P = 0.0012. (D) Image of a coronal section (enlarged inset on the right) containing EGFP-expressing neurons driven by the Pnoc promoter (n = 10). Scale bars: 200 μm. (E) Diagram of Vgat-Cre mice crossed with Pnoc-EGFP mice and injected with ChR2-mCherry AAV. Low-magnification mosaic (top middle) and expanded (bottom middle) images showing Pnoc-EGFP (green) expression in a subpopulation of Vgat-Cre neurons expressing ChR2-mCherry (red). Arrowheads indicate colocalization of mCherry and Pnoc-EGFp. Scale bars: 200 μm (low-magnification images) and 20 μm (high-magnification images). Graph shows the percentage of ChR2-mCherry–expressing Vgat neurons that coexpressed Pnoc-EGFP. The mean from 2 mice is shown (338 of 1103 Vgat neurons coexpressed Pnoc-EGFP).
Figure 3
Figure 3. Nociceptin neurons project to the arcuate nucleus.
(A) Diagrams of the DOGCre technology used to express Cre recombinase in Pnoc-EGFP neurons (top) and injection of DOGCre (C-CreintG and N-CretrcintG) and ChR2-mCherry AAVs into the aBNST (bottom). (B) Mosaic images (enlarged images shown below) showing the expression of Pnoc-EGFP (green) and ChR2-mCherry (red) in the aBNST (n = 47). (C) Diagrams showing the injection of DOGCre and ChR2-mCherry AAVs into the aBNST and the corresponding arcuate nucleus section. (D) Coronal sections (enlarged insets shown below) showing ChR2-mCherry expression in axons in the arcuate nucleus (red) and EGFP driven by the Pnoc promoter (green). n = 10. Scale bars: 200 μm (low-magnification images) and 20 μm (high-magnification images).
Figure 4
Figure 4. Photostimulation of aBNST nociceptin axons in the arcuate nucleus suppresses feeding.
(A) Diagram of Pnoc-EGFP mice crossed with Npy-hrGFP mice and injected with DOGCre and ChR2-mCherry AAVs into the aBNST. (B) Ensembled synaptic currents from an arcuate NPY neuron (n = 27) in the absence (black) and presence (red) of 20 μM bicuculline. (C) Photostimulated synaptic currents in arcuate NPY neurons from male (n = 15 neurons from 5 mice) and female (n = 12 neurons from 3 mice) mice. Data represent the mean ± SEM. An unpaired t test was performed for comparison between sexes [t (25) = 0.35, P = 0.73]. (D) Voltage traces (expanded below) from an arcuate NPY neuron during photostimulation (where indicated) of ChR2-mCherry–expressing aBNST axons from Pnoc-EGFP mice (n = 28). (E) Arcuate NPY neuronal action potential frequency (n = 28 neurons from 9 mice) before (baseline) and during photostimulation of aBNST axons. Data represent the mean ± SEM. **P < 0.01, by paired t test, t (27) = 3.33, P = 0.003. (F) Diagram illustrating a sagittal section containing the aBNST injected with DOGCre and ChR2-mCherry AAVs and the arcuate nucleus implanted with optical fibers. (G) Cumulative food intake following an overnight fast of Pnoc-EGFP mice injected with DOGCre and ChR2-mCherry AAVs into the aBNST during (shaded area; 3-second, 10-Hz bursts every 4 seconds) and after photostimulation of aBNST Pnoc fibers in the arcuate nucleus. Food intake was measured without (Nonstimulated, red) and with photoexcitation (Stimulated, blue). Data represent the mean ± SEM. n = 12 mice. Two-way repeated-measures ANOVA [interaction: f (6,132) = 2.15, P = 0.052; stimulation: f (1,22) = 4.76, P = 0.040]. *P < 0.05, by Sidak’s post hoc test. (H) Food intake following the 4-hour photostimulation period for nonstimulated (red) and stimulated (blue) mice. Data represent the mean ± SEM. n = 12 mice. ***P < 0.001, by paired t test, t (11) = 4.52, P = 0.0009.
Figure 5
Figure 5. Optogenetic stimulation of aBNST nociceptin fibers in the arcuate nucleus does not induce anxiety-like behavior.
(A) Diagram illustrating the injection of Pnoc-EGFP mice and WT littermate controls with DOGCre, with or without ChR2-mCherry AAVs, into the aBNST and implantation of optical fibers into the arcuate nucleus. (B) Schematic representation of mice tethered to a 470-nm laser and photostimulated (3-second, 10-Hz bursts every 4 seconds) for 30 minutes in a novel open-field arena. (C–E) Percentage of time spent in the center (C) [unpaired t test, t (24) = 0.63, P = 0.53], total distance traveled (D) [unpaired t test, t (24) = 1.20, P = 0.24], and mouse velocity (E) [2-way repeated-measures ANOVA, interaction: f (173,4152) = 1.08, P = 0.241; ChR2 expression: f (1,24) = 1.44, P = 0.241] in the open-field arena for mice expressing (blue, n = 15 mice) or not expressing (red, n = 11 mice) ChR2-mCherry. Data represent the mean ± SEM. (F) Schematic representation of mice tethered to a 470-nm laser and photostimulated (3-second, 10-Hz bursts every 4 seconds) for 10 minutes on a novel elevated zero maze. (G) Percentage of time spent in the anxiogenic open zones [unpaired t test, t (17) = 0.99, P = 0.33] and (H) number of entries into the anxiogenic open zones [unpaired t test, t (17) = 1.74, P = 0.10] for mice expressing (blue, n = 11 mice) or not expressing (red, n = 8 mice) ChR2-mCherry. Data represent the mean ± SEM.
Figure 6
Figure 6. Activity of aBNST nociceptin neurons increases during the initiation of feeding.
(A) Diagram illustrating injection of DOGCre and jRGECO1a AAVs into the aBNST of Pnoc-EGFP mice for ex vivo imaging and recording. (B) Simultaneous measurements of electrical excitability (Vm, top) and fluorescence intensity (ΔF/F, bottom) in aBNST jRGECO1a–expressing neurons (n = 3). Action potential firing was evoked by depolarizing current injections (I-Inj.) and correlated with increased fluorescence. (C and D) Image of aBNST neuron (C) and corresponding changes in spontaneous jRGECO1a fluorescence (ΔF/F) (D) at 590 nm but not 488 nm (n = 5). (E) Diagram illustrating injection of DOGCre and jRGECO1a AAVs into the aBNST of Pnoc-EGFP mice, with implantation of optical fibers into the aBNST. (F) jRGECO1a expression in the aBNST and optical fiber location. (G) Illustration of the open-field arena containing 2 pots (with a novel object or food). (H) Fluorescence intensity (ΔF/F) in mice approaching a novel object (blue) or initiating feeding (red). Data represent the mean ± SEM. n = 7 mice. Two-way repeated-measures ANOVA [interaction: f (239,2868) = 1.44, P < 0.0001; time: f (239,2868) = 1.35, P < 0.001; feeding versus object: f (1,12) = 2.70, P = 0.14]. (I) Diagram of intercrossed Pnoc-EGFP and Agrp-Cre mice injected with DOG-flpo and flp-dependent ChR2-mCherry AAVs into the aBNST and Cre-dependent jRGECO1a AAV into the arcuate nucleus. Optical fibers were placed in the aBNST for ChR2 stimulation and in the arcuate nucleus for jRGECO1a activity recording. (J) Fluorescence intensity (ΔF/F) in AgRP neurons before and after aBNST photostimulation, where indicated. Data indicate the mean ± SEM. n = 5 mice. (K) AgRP activity (ΔF/F) for mice shown in J before (Ctrl.) and after stimulation (Stim.). Data represent the mean ± SEM. *P < 0.05, by paired t test [t (4) = 2.97, P = 0.041]. Scale bars: 20 μm (C) and 200 μm (F).
Figure 7
Figure 7. Caspase-mediated ablation of aBNST nociceptin neurons.
(A and B) Low-magnification mosaic images (left) and enlarged insets (right) for mice not expressing (A, n = 4) or expressing (B, n = 7) caspase 3 in aBNST Pnoc neurons. Images show EGFP expression driven by the Pnoc promoter (top panels), Cre-dependent expression of ChR2-mCherry (middle panels), and colocalization (bottom panels). Scale bars: 200 μm (low-magnification images) and 20 μm (high-magnification images).
Figure 8
Figure 8. Loss of aBNST Pnoc neurons increases adiposity and body weight.
(A) Number of mCherry-expressing aBNST somas in control (red, n = 4) and caspase 3–treated (blue, n = 7) mice. Data represent the mean ± SEM. ****P < 0.0001, by unpaired t test [t (9) = 11.88, P = 4.19 × 10–7]. (B) Cumulative change in body weight (percentage of pre-surgery weight) in control (red, n = 10) and caspase 3–treated (blue, n = 11) mice. Data represent the mean ± SEM. Two-way repeated-measures ANOVA [interaction: f (8,152) = 3.37, P = 0.0014; control versus caspase treatment: f (1,19) = 5.02, P = 0.037]. *P < 0.05 and **P < 0.001, by Sidak’s post hoc test. (C) Change in body weight 6 weeks after surgery in control (red, n = 10) and caspase 3–treated (blue, n = 11) mice. Data represent the mean ± SEM. **P < 0.01, by unpaired t test [t (19) = 3.07, P = 0.0063]. (D) Fat mass [unpaired t test, t (19) = 2.43, P = 0.025] (*P < 0.05) and (E) lean mass [unpaired t test, t (19) = 0.48, P = 0.63] six weeks after surgery in control mice (red, n = 10) and caspase 3–treated mice (blue, n = 11). Data represent the mean ± SEM. (F) Ad libitum food intake in control (red, n = 7) and caspase 3–treated (blue, n = 8) mice. Data represent the mean ± SEM. *P < 0.05, by unpaired t test [t (13) = 2.53, P = 0.025].
Figure 9
Figure 9. A population of aBNST Pnoc neurons receives inputs from the hypothalamus.
(A) Diagram of injection of C-CreintG [serotyped with AAV2(retro)] into the arcuate nucleus and N-CretrcintG (serotyped with AAV1) into the aBNST of Pnoc-EGFP mice to drive Cre recombinase expression. A Cre-dependent avian retroviral receptor (2A-TVA) tagged with GFP and N2c-glycoprotein [N2c(G)] tagged with GFP were simultaneously injected into the aBNST. After 4 weeks to allow Cre-dependent expression, a glycoprotein-deficient rabies virus (CVS-N2cΔG) expressing mCherry was injected into the aBNST. (B) Representative images (n = 3) of a coronal section containing EGFP-expressing neurons driven by the Pnoc promoter and/or the tagged avian receptor and glycoprotein (left). Expression of mCherry driven by the rabies virus (middle) colocalized (right) with a small number of GFP-positive neurons (as shown by the arrows). Scale bars: 200 μm (low-magnification images) and 20 μm (high-magnification images corresponding to the boxed regions in the upper panels). (C–E) Low-magnification mosaic images (top) and enlarged insets (bottom) showing the expression of mCherry (n = 3 mice) driven by the rabies virus in presynaptic neurons in the aBNST (C), pBNST (D), and LH (E). Scale bars: 200 μm (low-magnification images) and 100 μm (high-magnification images). D3V, dorsal third ventricle; opt., optic tract.
See this image and copyright information in PMC

References

    1. Andermann ML, Lowell BB. Toward a wiring diagram understanding of appetite control. Neuron. 2017;95(4):757–778. doi: 10.1016/j.neuron.2017.06.014. - DOI - PMC - PubMed
    1. Denis RG, et al. Palatability can drive feeding independent of AgRP neurons. Cell Metab. 2015;22(4):646–657. doi: 10.1016/j.cmet.2015.07.011. - DOI - PMC - PubMed
    1. Sternson SM, Nicholas Betley J, Cao ZF. Neural circuits and motivational processes for hunger. Curr Opin Neurobiol. 2013;23(3):353–360. doi: 10.1016/j.conb.2013.04.006. - DOI - PMC - PubMed
    1. Aponte Y, Atasoy D, Sternson SM. AGRP neurons are sufficient to orchestrate feeding behavior rapidly and without training. Nat Neurosci. 2011;14(3):351–355. doi: 10.1038/nn.2739. - DOI - PMC - PubMed
    1. Krashes MJ, et al. Rapid, reversible activation of AgRP neurons drives feeding behavior in mice. J Clin Invest. 2011;121(4):1424–1428. doi: 10.1172/JCI46229. - DOI - PMC - PubMed

Publication types