Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Shifted pallidal co-release of GABA and glutamate in habenula drives cocaine withdrawal and relapse

An Author Correction to this article was published on 05 March 2020


Cocaine withdrawal produces aversive states and vulnerability to relapse, hallmarks of addiction. The lateral habenula (LHb) encodes negative stimuli and contributes to aversive withdrawal symptoms. However, it remains unclear which inputs to LHb promote this and what the consequences are for relapse susceptibility. We report, using rabies-based retrolabeling and optogenetic mapping, that the entopeduncular nucleus (EPN, the mouse equivalent of the globus pallidus interna) projects to an LHb neuronal subset innervating aversion-encoding midbrain GABA neurons. EPN-to-LHb excitatory signaling is limited by GABAergic cotransmission. This inhibitory component decreases during cocaine withdrawal as a result of reduced presynaptic vesicular GABA transporter (VGAT). This shifts the EPN-to-LHb GABA/glutamate balance, disinhibiting EPN-driven LHb activity. Selective virally mediated VGAT overexpression at EPN-to-LHb terminals during withdrawal normalizes GABAergic neurotransmission. This intervention rescues cocaine-evoked aversive states and prevents stress-induced reinstatement, used to model relapse. This identifies diminished inhibitory transmission at EPN-to-LHb GABA/glutamate synapses as a mechanism contributing to the relapsing feature of addictive behavior.

This is a preview of subscription content, access via your institution

Relevant articles

Open Access articles citing this article.

Access options

Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Figure 1: EPN-receptive LHb neurons preferentially target midbrain GABA neurons.
Figure 2: GABAergic plasticity at EPN-to-LHb synapses during cocaine withdrawal.
Figure 3: Impaired vesicular filling at EPN-to-LHb GABAergic synapses during withdrawal.
Figure 4: Overexpression of VGAT at EPN-to-LHb synapses rescues cocaine-withdrawal-induced reduction of GABA transmission.
Figure 5: EPN-to-LHb GABA plasticity mediates aversive cocaine withdrawal symptoms and stress-induced reinstatement.


  1. Lammel, S. et al. Input-specific control of reward and aversion in the ventral tegmental area. Nature 491, 212–217 (2012).

    Article  CAS  Google Scholar 

  2. Stamatakis, A.M. & Stuber, G.D. Activation of lateral habenula inputs to the ventral midbrain promotes behavioral avoidance. Nat. Neurosci. 15, 1105–1107 (2012).

    Article  CAS  Google Scholar 

  3. Ide, J.S. & Li, C.S. Error-related functional connectivity of the habenula in humans. Front. Hum. Neurosci. 5, 25 (2011).

    Article  Google Scholar 

  4. Brown, R.M., Short, J.L. & Lawrence, A.J. Identification of brain nuclei implicated in cocaine-primed reinstatement of conditioned place preference: a behaviour dissociable from sensitization. PLoS One 5, e15889 (2010).

    Article  CAS  Google Scholar 

  5. Gill, M.J., Ghee, S.M., Harper, S.M. & See, R.E. Inactivation of the lateral habenula reduces anxiogenic behavior and cocaine seeking under conditions of heightened stress. Pharmacol. Biochem. Behav. 111, 24–29 (2013).

    Article  CAS  Google Scholar 

  6. Jhou, T.C. et al. Cocaine drives aversive conditioning via delayed activation of dopamine-responsive habenular and midbrain pathways. J. Neurosci. 33, 7501–7512 (2013).

    Article  CAS  Google Scholar 

  7. Meye, F.J. et al. Cocaine-evoked negative symptoms require AMPA receptor trafficking in the lateral habenula. Nat. Neurosci. 18, 376–378 (2015).

    Article  CAS  Google Scholar 

  8. Shabel, S.J., Proulx, C.D., Trias, A., Murphy, R.T. & Malinow, R. Input to the lateral habenula from the basal ganglia is excitatory, aversive, and suppressed by serotonin. Neuron 74, 475–481 (2012).

    Article  CAS  Google Scholar 

  9. Shabel, S.J., Proulx, C.D., Piriz, J. & Malinow, R. Mood regulation. GABA/glutamate co-release controls habenula output and is modified by antidepressant treatment. Science 345, 1494–1498 (2014).

    Article  CAS  Google Scholar 

  10. Hong, S. & Hikosaka, O. The globus pallidus sends reward-related signals to the lateral habenula. Neuron 60, 720–729 (2008).

    Article  CAS  Google Scholar 

  11. Ishida, Y., Denovan-Wright, E., Hebb, M.O. & Robertson, H.A. Amphetamine-induced Fos expression is evident in gamma-aminobutyric acid neurons in the globus pallidus and entopeduncular nucleus in rats treated with intrastriatal c-fos antisense oligodeoxynucleotides. Exp. Neurol. 175, 275–281 (2002).

    Article  CAS  Google Scholar 

  12. Porrino, L.J. Functional consequences of acute cocaine treatment depend on route of administration. Psychopharmacology (Berl.) 112, 343–351 (1993).

    Article  CAS  Google Scholar 

  13. Pignatelli, M. & Bonci, A. Role of dopamine neurons in reward and aversion: a synaptic plasticity perspective. Neuron 86, 1145–1157 (2015).

    Article  CAS  Google Scholar 

  14. Micheva, K.D. & Smith, S.J. Array tomography: a new tool for imaging the molecular architecture and ultrastructure of neural circuits. Neuron 55, 25–36 (2007).

    Article  CAS  Google Scholar 

  15. Soiza-Reilly, M., Anderson, W.B., Vaughan, C.W. & Commons, K.G. Presynaptic gating of excitation in the dorsal raphe nucleus by GABA. Proc. Natl. Acad. Sci. USA 110, 15800–15805 (2013).

    Article  CAS  Google Scholar 

  16. Edwards, R.H. The neurotransmitter cycle and quantal size. Neuron 55, 835–858 (2007).

    Article  CAS  Google Scholar 

  17. Erickson, J.D., De Gois, S., Varoqui, H., Schafer, M.K. & Weihe, E. Activity-dependent regulation of vesicular glutamate and GABA transporters: a means to scale quantal size. Neurochem. Int. 48, 643–649 (2006).

    Article  CAS  Google Scholar 

  18. Wang, L., Tu, P., Bonet, L., Aubrey, K.R. & Supplisson, S. Cytosolic transmitter concentration regulates vesicle cycling at hippocampal GABAergic terminals. Neuron 80, 143–158 (2013).

    Article  CAS  Google Scholar 

  19. Tritsch, N.X., Ding, J.B. & Sabatini, B.L. Dopaminergic neurons inhibit striatal output through non-canonical release of GABA. Nature 490, 262–266 (2012).

    Article  CAS  Google Scholar 

  20. Chillon, M. & Kremer, E.J. Trafficking and propagation of canine adenovirus vectors lacking a known integrin-interacting motif. Hum. Gene Ther. 12, 1815–1823 (2001).

    Article  CAS  Google Scholar 

  21. Barr, A.M., Markou, A. & Phillips, A.G.A. A 'crash' course on psychostimulant withdrawal as a model of depression. Trends Pharmacol. Sci. 23, 475–482 (2002).

    Article  CAS  Google Scholar 

  22. Koob, G.F. & Volkow, N.D. Neurocircuitry of addiction. Neuropsychopharmacology 35, 217–238 (2010).

    Article  Google Scholar 

  23. Mantsch, J.R., Baker, D.A., Funk, D., Le, A.D. & Shaham, Y. Stress-induced reinstatement of drug seeking: 20 years of progress. Neuropsychopharmacology 41, 335–356 (2016).

    Article  CAS  Google Scholar 

  24. Hatzigiakoumis, D.S., Martinotti, G., Giannantonio, M.D. & Janiri, L. Anhedonia and substance dependence: clinical correlates and treatment options. Front. Psychiatry 2, 10 (2011).

    Article  Google Scholar 

  25. Li, B. et al. Synaptic potentiation onto habenula neurons in the learned helplessness model of depression. Nature 470, 535–539 (2011).

    Article  CAS  Google Scholar 

  26. Ji, H. & Shepard, P.D. Lateral habenula stimulation inhibits rat midbrain dopamine neurons through a GABAA receptor-mediated mechanism. J. Neurosci. 27, 6923–6930 (2007).

    Article  CAS  Google Scholar 

  27. Dunlop, B.W. & Nemeroff, C.B. The role of dopamine in the pathophysiology of depression. Arch. Gen. Psychiatry 64, 327–337 (2007).

    Article  CAS  Google Scholar 

  28. Tye, K.M. et al. Dopamine neurons modulate neural encoding and expression of depression-related behaviour. Nature 493, 537–541 (2013).

    Article  CAS  Google Scholar 

  29. Radke, A.K. & Gewirtz, J.C. Increased dopamine receptor activity in the nucleus accumbens shell ameliorates anxiety during drug withdrawal. Neuropsychopharmacology 37, 2405–2415 (2012).

    Article  CAS  Google Scholar 

  30. Twining, R.C. et al. Aversive stimuli drive drug seeking in a state of low dopamine tone. Biol. Psychiatry 77, 895–902 (2015).

    Article  CAS  Google Scholar 

  31. Rah, J.C. et al. Thalamocortical input onto layer 5 pyramidal neurons measured using quantitative large-scale array tomography. Front. Neural Circuits 7, 177 (2013).

    Article  Google Scholar 

  32. Micheva, K.D., Busse, B., Weiler, N.C., O'Rourke, N. & Smith, S.J. Single-synapse analysis of a diverse synapse population: proteomic imaging methods and markers. Neuron 68, 639–653 (2010).

    Article  CAS  Google Scholar 

  33. Soiza-Reilly, M. & Commons, K.G. Quantitative analysis of glutamatergic innervation of the mouse dorsal raphe nucleus using array tomography. J. Comp. Neurol. 519, 3802–3814 (2011).

    Article  CAS  Google Scholar 

  34. Seo, S. & Leitch, B. Synaptic changes in GABAA receptor expression in the thalamus of the stargazer mouse model of absence epilepsy. Neuroscience 306, 28–38 (2015).

    Article  CAS  Google Scholar 

  35. Woo, J. et al. The adhesion protein IgSF9b is coupled to neuroligin 2 via S-SCAM to promote inhibitory synapse development. J. Cell Biol. 201, 929–944 (2013).

    Article  CAS  Google Scholar 

Download references


We thank E. Kremer and P. Jegouzo for technical assistance. We thank A. Adamantidis, G. Ramakers, C. Bellone, J.C. Poncer and the Mameli laboratory for feedback on the manuscript and discussions. This work was supported by INSERM Atip-Avenir, the City of Paris, the European Research Council (Starting Grant SalienSy 335333) to M.M., the Fyssen Foundation to F.J.M. and the SFB1089 to M.K.S. The Mameli laboratory is part of the LabEx BioPsy.

Author information

Authors and Affiliations



F.J.M. performed and analyzed all the in vitro mapping and electrophysiological recordings with the help of T.S. and M.M. M.S.-R. performed and analyzed the array tomography experiments with the help of F.J.M. F.J.M., M.A.D. and M.M. performed and analyzed the behavioral experiments. M.K.S. generated and provided viral tools for rabies-based retrograde labeling. M.M. and F.J.M. designed the study and wrote the manuscript with the help of all authors. M.M. is a member of the Fens-Kavli Network of Excellence.

Corresponding author

Correspondence to Manuel Mameli.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Integrated supplementary information

Supplementary Figure 1 Topographical distribution of LHb→GABA and LHb→DA neurons.

(a) Schematic illustrating the principle of retrograde labeling of specific LHb neurons projecting to putative RMTg GABAergic neurons in the midbrain with RABVΔG (EnvA)-myrTomato (rabies) in a VGAT-Cre mouse line. (b) Injection sites for EPN and midbrain. (c) Mapping of rabies-retrogradely-labeled LHb neurons from two DAT-Cre mice injected with rabies in the VTA. Three coronal sections show the distribution through the anterior-posterior axis of these LHb→DA neurons. The magenta dotted line demarcates the lateral and medial parts of LHb. (d) Same as c but for LHb→GABA neurons in two VGAT-Cre mice injected with rabies in the putative RMTg. (e) Overlayed schematic representation of the distribution of LHb→GABA and LHb→DA neurons from two example brains.

Supplementary Figure 2 Cocaine withdrawal alters GABA but not glutamate signaling in EPN-to-LHb synapses.

(a) Opto-stimulation of EPN terminals onto LHb neurons reveals postsynaptic currents that are inward and/or outward depending on the holding potential. At -20 mV a biphasic inward (AMPAR-mediated, I-AMPAR) and outward (GABAaR-mediated, I-GABAaR) current is typically observed. (b) Evaluation of the extent of outward GABAaR-mediated current with respect to the total absolute response elicited at -20 mV in cocaine and saline-treated animals (Saline, 33.13± 6.95%, ncells = 28, nanimals = 7; Cocaine withdrawal, 3.30 ± 2.13%, ncells = 12, nanimals = 4; Cocaine no withdrawal, 25.86 ± 9.43%, ncells = 12, nanimals = 2; F(2,49) = 3.762, p=0.03). (c) AMPAR/NMDAR ratios for the EPN-to-LHb synapse in cocaine-withdrawal and in saline controls (Saline, 1.06 ± 0.30, ncells = 6, nanimals = 4; Cocaine, 0.94 ± 0.22, ncells = 5, nanimals = 5; F(1,9) = 0.94, p=0.767). (d) Current-voltage relationship for AMPAR and NMDAR currents in saline and cocaine conditions (I-AMPAR at +40 mV, ncells = 11-18, nanimals = 6-7; F(1,27) = 0.278, p=0.603; I-NMDAR at -20 mV, ncells = 7-10, nanimals = 4-6; F(1,15) = 0.578, p=0.459). (e) GABAR/NMDAR ratios obtained by electrical stimulation in the LHb in cocaine withdrawal and in saline control mice (Saline, 8.72 ± 1.90, ncells = 7, nanimals = 3; Cocaine, 7.91 ± 2.65, ncells = 7, nanimals = 4; F(1,12) = 0.70, p=0.796). (f) Samples of the effect of I-GABAaR on I-AMPAR decay kinetics (exponential fit indicated by red dotted line) and charge transfer (blue area) at -50 mV in saline and cocaine conditions. Bar graph quantifying the extent of I-GABA masking of I-AMPAR decay kinetics (Saline, 51.93 ± 9.47%, ncells = 7, nanimals = 5; Cocaine, 20.38 ± 5.30%, ncells = 7, nanimals = 5; F(1,14) = 5.96, p=0.031). *p<0.05 refers to the specific statistical test employed in panels indicated in parentheses. One-way ANOVA (b, c, d, e, f).

Supplementary Figure 3 Effects of cocaine withdrawal on EPN-to-LHb synaptic markers and function.

(a) Immunolabeling for VGAT, gephyrin and PSD-95 throughout adjacent ultrathin (100 nm) serial sections (dorsal-ventral plane). Asterisks indicate examples of puncta appositions between VGAT-positive axon terminals and postsynaptic sites containing gephyrin. Arrows show a nearby postsynaptic site for glutamatergic synapses containing PSD-95.

(b) Averaged puncta/µm3 obtained after array tomography for markers at EPN-to-LHb synapses in saline and cocaine-treated animals (Number of puncta per µm3; for total alpha1, F(1,6) = 0.473, p=0.517; total Vglut2, F(1,6) = 0.281, p=0.615; DsRed+/alpha1+, F(1,6) = 0.253, p=0.163; DsRed+/Vglut2+, F(1,6) = 2.72, p=0.15). (c) Sample, average time course and bar graph of EPN-to-LHb glutamatergic responses at baseline and after the 3rd train (Saline, 71.05 ± 12.02%, ncells = 6, nanimals =3; Cocaine, 71.05 ± 13.20%, ncells = 5, nanimals =3; F(1,9) = 0.128, p=0.728).

(d) Paired-pulse experiments for both EPN-driven IPSCs and EPSCs in saline and cocaine treated mice (Saline, ncells = 5-7, nanimals = 2-3; Cocaine, ncells = 5-7, nanimals = 3; F(1,7) = 1.12, p=0.325. Shown traces represent overlay of 9 sweeps with different interpulse intervals. (e) Paired-pulse ratio for the depletion experiments during 0.1 Hz stimulation obtained at baseline and after the 3rd train of 500 pulses at 4 Hz (For IPSCs: Saline, ncells = 4, nanimals = 2; Cocaine, ncells = 4, nanimals = 2. F(1,6) = 1.79, p=0.23; For EPSCs: Saline, ncells = 6, nanimals = 3; Cocaine, ncells = 5, nanimals = 3. F(1,9) = 0.476, p=0.51). (f) Representative image indicating an LHb neuron filled with Alexa-488 and the uncaging location. Rubi-GABA was uncaged (405 nm laser) at the soma (1), proximal (2) and distal (3) dendrites of LHb neurons, F(2,38) = 1.068, p=0.354. Representative traces, bar graph and scatter plot representing the effect of cocaine withdrawal on IPSC amplitudes. One-way ANOVA (b, c, e), Two-Way Repeated Measures ANOVA I.P. Treatment Main Effect and interaction (d, f).

Supplementary Figure 4 VGAT overexpression rescues GABAergic control over EPN-driven LHb spike fidelity.

(a-d) Spike fidelity of LHb neurons as a consequence of EPN stimulation with a 10 pulse train delivered at 5 Hz in all experimental groups prior (solid line) and after picrotoxin (dotted line). Quantification is represented in Fig. 3d.

Supplementary Figure 5 Anxiety during cocaine withdrawal, histology of VGAT overexpression, and cocaine-driven locomotor activity.

(a) Representative tracking plots (red line) of individual mice in the closed (grey) and open (white) arm of the EPM. Analysis of anxiety behavior in the elevated plus maze during cocaine withdrawal (% of time in open arm: Saline: 9.52 ± 1.74%, Cocaine: 4.23 ± 1.22%; F(1,18) = 6.16, p=0.023; For Entries in open arm: Saline: 20.8 ± 4.92, Cocaine: 14.3 ± 2.73; F(1,18) = 1.34, p=0.263; For Head dips: Saline: 36.1 ± 2.09, Cocaine: 29.1 ± 2.24; F(1,18) = 5.214, p=0.035; nanimals = 10 for both groups). (b) Histological verification for behavioral experiments. Examples of successful (required criterion for experimental inclusion) Cav2-Cre-driven VGAT overexpression at EPN-to-LHb terminals as indicated by mCherry fluorescence is represented in 4 sample slices of animals injected with either saline or cocaine. (c) Cocaine-driven locomotion (measured over 15 min) during CPPacq (nanimals = 8 for all groups, F(4, 112) = 6.459, p=0.001). *p<0.05; **p<0.01; ***p<0.001 refers to the specific statistical test employed in panels indicated in parentheses. One-way ANOVA (a). Two-Way Repeated Measures ANOVA, Experimental Phase * I.P. Treatment Interaction (c).

Supplementary Figure 6 VGAT deletion at LHb synapses.

(a) Timeline of experiment. (b) Histological verification of injection site. Examples of Cav2-Cre with Green fluorobeads (as a marker for injection site) infused in the lateral LHb. (c) GABAaR/AMPAR ratios for the EPN-to-LHb synapse obtained by optical stimulation in PBS-control and Cav2-Cre injected VGATfl/fl mice (PBS-control, 0.83 ± 0.20, ncells = 6, nanimals = 2; Cav2-Cre, 0.05 ± 0.02, ncells = 8, nanimals = 4; F(1,12) = 20.9, p=0.0006). (d) GABAaR/AMPAR ratios obtained by electrical stimulation and recordings in the medial LHb in PBS-control and Cav2-Cre injected VGATfl/fl mice (PBS-control, 0.52 ± 0.06, ncells = 7, nanimals = 2; Cav2-Cre, 0.59 ± 0.07, ncells = 8, nanimals = 4; F(1,13) = 0.528, p=0.48). (e) Total immobility during swim stress (nanimals = 10/group; F(1,18) = 8.365, p=0.0097). (f) Total distance during open field test (nanimals = 10/group; F(1,18) = 1.143, p=0.299). **p<0.01, ***p<0.001 refers to the specific statistical test employed in panels indicated in parentheses. One-Way ANOVA (c-f).

Supplementary information

Source data

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Meye, F., Soiza-Reilly, M., Smit, T. et al. Shifted pallidal co-release of GABA and glutamate in habenula drives cocaine withdrawal and relapse. Nat Neurosci 19, 1019–1024 (2016).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:

This article is cited by


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing