mGluR-LTD at Excitatory and Inhibitory Synapses in the Lateral Habenula Tunes Neuronal Output

Abstract

Excitatory and inhibitory transmission onto lateral habenula (LHb) neurons is instrumental for the expression of positive and negative motivational states. However, insights into the molecular mechanisms modulating synaptic transmission and the repercussions for neuronal activity within the LHb remain elusive. Here, we report that, in mice, activation of group I metabotropic glutamate receptors triggers long-term depression at excitatory (eLTD) and inhibitory (iLTD) synapses in the LHb. mGluR-eLTD and iLTD rely on mGluR1 and PKC signaling. However, mGluR-dependent adaptations of excitatory and inhibitory synaptic transmission differ in their expression mechanisms. mGluR-eLTD occurs via an endocannabinoid receptor-dependent decrease in glutamate release. Conversely, mGluR-iLTD occurs postsynaptically through PKC-dependent reduction of β2-containing GABAA-R function. Finally, mGluR-dependent plasticity of excitation or inhibition decides the direction of neuronal firing, providing a synaptic mechanism to bidirectionally control LHb output. We propose mGluR-LTD as a cellular substrate that underlies LHb-dependent encoding of opposing motivational states.

Graphical abstract

Figure 1

mGluR-eLTD and -iLTD in the LHb (A) Schematic depicting the LHb microdissection (hipp, hippocampus; thal, thalamus) and mGluR1 and mGluR5 expression in the LHb. MW, molecular weight; bp: base pairs. (B) Sample trace, bar graph, and scatterplot representing DHPG-evoked current (−28.5 ± 4.6 pA, 12 of 22 responding neurons). (C) Sample traces representing EPSCs at baseline (a) and 20 min following DHPG (b). The timeline represents the DHPG effect (50 μM) on EPSCs. The bar graph and scatterplot show normalized averaged EPSCs ∼40 min after DHPG (66.3 ± 5%, t₁₉ = 6.306, ^∗∗∗p < 0.0001). (D) The same as (C) but for IPSCs (69.9 ± 7.3%, t₁₇ = 4.235, ^∗∗∗p < 0.0001). (E) LFS-driven (1 Hz, 15 min) eLTD. The bar graph and scatterplot show normalized averaged EPSCs ∼40 min after the protocol (65.8 ± 6.4%, t₉ = 5.3, ^∗∗∗p < 0.0001). (F) HFS-driven (100 Hz, 1 s, at 0 mV) iLTD (top). The bar graph and scatterplot show normalized averaged IPSCs ∼40 min after the protocol (74.2 ± 3.9%, t₆ = 7.086, ^∗∗∗p < 0.0001). When not indicated, the timescale represents 5 ms. Error bars represent SEM. n indicates number of recorded neurons.

Figure 2

mGluR1 and PKC-Dependent Induction for eLTD and iLTD (A) DHPG effect on EPSCs in the presence of the mGluR1 antagonist LY367385 (91.1 ± 5.6%, t₉ = 2.063, p > 0.05). (B) The same as (A) but for IPSCs (88.7 ± 6.7%, t₈ = 1.680, p > 0.05). (C) DHPG effect on EPSCs in the presence of the mGluR5 antagonist MTEP (74.5 ± 10.8%, t₈ = 2.377, ^∗p < 0.05). (D) The same as (C) but for IPSCs (71 ± 8.7%, t₆ = 3.425, ^∗p < 0.05). (E) DHPG effect on EPSCs during intracellular dialysis of PKC[19-36] (99.2 ± 14.8%, t₉ = 0.066, p > 0.5). (F) The same as (E) but for IPSCs (98.1 ± 9.1%, t₆ = 0.088, p > 0.05). (G) Effect of PMA on EPSCs (b, baseline versus PMA, 68.2 ± 2.4%, t₆ = 13.39, ^∗∗∗p < 0.0001) and subsequent occlusion of DHPG eLTD (c, PMA versus post-DHPG, 64.2 ± 5%, t₆ = 1.260, p > 0.05). (H) The same as (G) but for IPSCs (b, baseline versus PMA, 62.7 ± 8.7%, t₄ = 4.285, ^∗p < 0.05; c, PMA versus post-DHPG, 61.7 ± 14.4%, t₄ = 0.155, p > 0.05). Error bars represent SEM. n indicates number of recorded neurons.

Figure 3

mGluR-eLTD Expression via CB1-R Activation (A) Sample traces of AMPA-EPSCs at –60, 0, and +40 mV at baseline and after DHPG and average rectification index (baseline 3.6 ± 0.7 versus post-DHPG 3.5 ± 0.6, t₈ = 0.192, p > 0.05). (B) PPR of EPSCs in artificial cerebrospinal fluid (ASCF; baseline 0.62 ± 0.05 versus post-DHPG 0.9 ± 0.05, t₁₉ = 4.963, ^∗∗∗p < 0.0001); in the presence of LY367385 (baseline 0.53 ± 0.08 versus post-DHPG 0.64 ± 0.11, t₇ = 1.860, p > 0.05); of MTEP (baseline 0.59 ± 0.08 versus post-DHPG 0.79 ± 0.08, t₈ = 3.432, ^∗∗p < 0.01); of PKC[19-36] in the recording pipette (baseline 0.7 ± 0.1 versus post-DHPG 0.76 ± 0.1, t₉ = 1.214, p > 0.05); after PMA and PMA + DHPG (baseline 0.57 ± 0.08 versus PMA 0.74 ± 0.07, t₆ = 2.799,^∗p < 0.05; PMA baseline versus PMA post-DHPG, t₆ = 0.829, p > 0.05). Shown are neurons represented in Figures 1 and 2. One-way ANOVA among all baseline PPR conditions: F_{(9, 83)} = 0.485, p > 0.05. (C) Top: sample traces for mEPSCs. Cumulative probability plots show amplitudes and inter-event intervals for mEPSCs at baseline (black) and after DHPG (red). (mEPSC amplitude: baseline 30 ± 3.8 pA versus post-DHPG 32.1 ± 4.1 pA, KS test, p > 0.05; mEPSC frequency: baseline 3.8 ± 1.6 Hz versus post-DHPG 2.5 ± 1.1 Hz, KS test, ^∗p < 0.05). (D) Effect of WIN-55,212-2 on EPSCs (72.7 ± 3.8%, t₅ = 7.246, ^∗∗∗p < 0.001) and subsequent occlusion after DHPG application (68.5 ± 3.9%, t₅ = 5.559, p > 0.05). (E) PPR of EPSCs after WIN application and subsequent DHPG application (baseline 0.45 ± 0.02, post-WIN 0.67 ± 0.05, post-DHPG 0.71 ± 0.06; baseline versus post-WIN, t₅ = 3.411, ^∗p < 0.05; post-WIN versus post-DHPG, t₅ = 1.004, p > 0.05). (F) The same as (D) but in the presence of NESS-0327 (90.79 ± 9.02%, t₅ = 1.001, p > 0.05). (G) The same as (E) but in the presence of NESS-0327 (baseline 0.54 ± 0.09 versus post-WIN 0.57 ± 0.06, t₅ = 0.672 p > 0.05). (H) Effect of DHPG on EPSCs in the presence of NESS-0327 (95.9 ± 4.3%, t₁₁ = 0.766, p > 0.05). (I) PPR after DHPG in the presence of NESS-0327 (baseline 0.58 ± 0.03 versus post-DHPG 0.63 ± 0.06, t₁₁ = 1.404, p > 0.05). Error bars represent SEM. n indicates number of recorded neurons.

Figure 4

PKC Action on the GABA_A-Rs-β2 Subunit Underlies mGluR-iLTD (A) PPR of IPSCs after DHPG (baseline 0.64 ± 0.07 versus post-DHPG 0.73 ± 0.06, t₁₇ = 1.739, p > 0.05). (B) Top: sample traces of mIPSCs. Cumulative probability plots show inter-event intervals and amplitudes for IPSCs at baseline (black) and after DHPG (blue) (mIPSC amplitude: baseline 41.3 ± 5.9 pA versus post-DHPG 39.2 ± 4.83 pA, KS test, ^∗∗p < 0.01; mIPSC frequency: baseline 3.6 ± 1.09 Hz versus post-DHPG 3.45 ± 1.13 Hz, KS test, p > 0.05). (C) Effect of DHPG on IPSCs in the presence of NESS-0327 (79.1 ± 7.06%, t₁₀ = 2.621, ^∗p < 0.05). (D) DHPG effect on IPSCs in the presence of intracellular dynamin inhibitor (75.2 ± 10.2%, t₈ = 2.378, ^∗p < 0.05). (E) Example of peak-scaled NSFA of mIPSCs at baseline (black) and after DHPG (blue). Insets, overlay and average of analyzed traces. (F) Pooled data for N and γ after NSFA (N: baseline 37 ± 6.3 versus post-DHPG 39.3 ± 5.6; t₁₃ = 0.8, p > 0.05; γ: baseline 31.4 ± 1 versus post-DHPG 27.2 ± 1.4; t₁₃ = 2.4, ^∗p < 0.05). (G) The same as (D) but in the presence of intracellular GABA_A-β2 peptide (97.7 ± 10.4%, t₁₃ = 0.225, p > 0.05). (H) The same as (G) but for EPSCs (78.8 ± 6.3%, t₆ = 3.488, ^∗p < 0.05). (I) The same as (G) but in the presence of intracellular GABA_A-γ2 peptide (67.7 ± 5.5%, t₈ = 6.141, ^∗∗∗p < 0.001). (J) The same as (I) but for EPSCs (77.7 ± 7.4%, t₇ = 3.014, ^∗p < 0.05). Error bars represent SEM. n indicates number of recorded neurons.

Figure 5

mGluR-Dependent Bidirectional Control of LHb Neuronal Output (A) DHPG effect on synaptically evoked AP numbers in ACSF (97.7 ± 19.6; t₁₃ = 0.13, p > 0.05). 4 of 14 recorded neurons increased in firing (blue), and 5 of 14 decreased in firing (red) following DHPG. Sample traces indicate the bidirectional nature (blue increased firing, red decreased firing) of mGluR activation. Shown are superimposed EPSPs at baseline (black) and after DHPG (blue, increased firing; red, decreased firing). (B) Correlation between normalized mGluR-driven firing and normalized PSP area (Pearson correlation, r = 0.75, ^∗∗p < 0.01). (C) The same as (A) but in the presence of NESS-0327 and GABA_A-β2 peptide in the internal solution (109.2 ± 8.2, t₇ = 1.12, p > 0.05). Black and gray traces represent before and after DHPG. (D) The same as (B) but in the presence of NESS-0327 in the ACSF and GABA_A-β2 peptide in the internal solution (Pearson correlation, r = −0.12, p > 0.05). Fisher r-to-z transformation for (B) versus (D) correlations yielded a Z score of 2.03. ^∗p < 0.05. (E) Schematic indicating the induction and expression mechanisms for mGluR-eLTD and -iLTD and their relative contribution to LHb neuronal output. Error bars represent SEM. n indicates number of recorded neurons.