Metabotropic suppression of excitation in murine autaptic hippocampal neurons

Abstract

Depolarization-induced suppression of excitation (DSE) and inhibition (DSI) are forms of short-term neuronal plasticity involving postsynaptic release of an endocannabinoid and the activation of presynaptic cannabinoid CB1 receptors. We have recently reported that CB1-dependent DSE can be elicited in autaptic cultures of excitatory hippocampal neurons of the mouse. We now report that the same preparation exhibits a parallel G(q)-coupled receptor-dependent production of endocannabinoids causing retrograde inhibition, also via CB1 receptors, which we will refer to as metabotropic suppression of excitation (MSE). We tested a spectrum of G(q)-coupled receptor agonists and found that both muscarinic and metabotropic glutamate receptors (group I) mediate retrograde inhibition via CB1 receptors in autaptic hippocampal neurons. Thus these neurons possess not only the pre- and postsynaptic machinery necessary for DSE but also that for MSE. This permitted a closer examination of MSE and its interaction with other aspects of the endocannabinoid retrograde signalling machinery: MSE mimics and occludes DSE and is itself occluded by the endocannabinoid 2-arachidonoyl glycerol (2-AG), consistent with 2-AG as a likely mediator of MSE. In contrast to DSE, MSE undergoes heterologous desensitization over the time course of minutes. In keeping with data reported for metabotropic suppression of inhibition (MSI) and DSI in the hippocampus, subthreshold MSE and DSE act synergistically. We additionally found that Delta9-tetrahydrocannabinol, which has been shown to attenuate DSE, antagonizes MSE. Finally, we have distinguished a neuronal subpopulation that exhibits DSE and a differential complement of MSE-mediating Gq-coupled receptors, making possible contrasting studies of MSE. Autaptic endocannabinoid signalling is rich, robust and complex in a deceptively simple package, including a previously unreported postsynaptic mechanism of adaptation in addition to known presynaptic CB1 desensitization. These adaptive sites offer novel targets for modulation of endogenous cannabinoid signalling.

Figure 1

Muscarinic receptors mediate MSE in autaptic neurons A, left panel shows time course of MSE (integral, in pC) recorded before and after treatment with muscarinic agonist oxo-M (10 μm). Inset shows relevant EPSC traces from time points 1–3 indicated on left panel. Scale bars, 2 nA and 10 ms. Right panel shows recovery time course for oxo-M treatment (t_1/2 = 38 s, n = 6) B, dose–response curve for oxo-M in wild-type neurons (▾), CB1–/– neurons (▪) and in presence of CB1 antagonist SR141716 (100 nm; ♦). C, top panel shows comparison of paired-pulse ratios (interval, 60 ms) before and after MSE stimulus. Bottom panel shows representative traces (scale bars, 1 nA and 10 ms). Traces are normalized to the 1st EPSC of the control stimulus to facilitate comparison. D, comparison of oxo-M responses (5–10 μm) and saturating DSE responses in the same cell (n = 14). E, left panel shows comparison of magnitude of response to serial oxo-M treatments in the same cell (n = 4). Right panel shows sample timecourse with brief, sequential oxo-M treatments. F, graph showing inhibition of MSE (oxo-M) after pretreatment with Ca²⁺ ATPase inhibitor thapsigargin (1 μm, 20+ min; n = 14). G, comparison of MSE under control conditions and after treatment with PLC inhibitor U73122 (3 μm, 2.5–5 min).

Figure 2

MSE desensitizes A, sample timecourse of prolonged oxo-M treatment showing gradual desensitization. Centre panel (top) shows summary of oxo-M treatment in same-cell experiments under control and desensitized conditions. Centre panel (bottom) shows DSE in the same cells before and after MSE desensitization. B, sample timecourse of desensitizing oxo-M responses with repeated/prolonged exposure, with DSE responses (inset) from same cell before and after MSE desensitization. Scale bars, 1 nA and 10 s **P < 0.005, Student's paired t test.

Figure 3

Muscarinic response in a neuronal subpopulation Type O A, timecourse of EPSC responses to oxo-M and 2-AG in Type O CB1–/– neurons. B, dose–response curve for oxo-M in wild-type Type O neurons (▾), CB1-/– Type O neurons (▪) and in presence of CB1 antagonist SR141716 (100 nm; ♦). C, top panel shows comparison of paired-pulse ratios (interval, 60 ms) before and after MSE stimulus. Bottom panel shows representative traces (scale bars, 1 nA and 10 ms). Traces are normalized to the 1st EPSC of the control stimulus to facilitate comparison. D, same-cell experiments wherein cells were treated with oxo-M (2 μm) alone and then in the presence of muscarinic M2 receptor antagonist gallamine (5 μm). *P < 0.05, Student's paired t test.

Figure 4

Metabotropic glutamate receptors (group I) mediate MSE in autaptic neurons A, timecourse of EPSCs (integral, in pC) recorded before and after treatment with metabotropic glutamatergic agonist DHPG (100 μm). B, dose–response curve for DHPG in wild-type neurons (▾), CB1–/– neurons (▪) and in the presence of the CB1 antagonist SR141716 (100 nm; ♦). C, top panel shows comparison of paired-pulse ratios (interval, 60 ms) before and after DHPG treatment. Bottom panel shows representative traces (scale bars, 1 nA and 10 ms). Traces are normalized to the 1st EPSC of the control stimulus to facilitate comparison. D, time course of EPSCs in Type O neuron recorded before and after treatment with metabotropic glutamatergic agonist DHPG (100 μm). E, dose–response curve for DHPG in wild-type Type O neurons (▾), CB1–/– neurons (▪) and in the presence of the CB1 antagonist SR141716 (100 nm; ♦). F, top panel shows comparison of paired-pulse ratios (interval, 60 ms) before and after DHPG treatment in Type O neurons. Bottom panel shows representative traces (scale bars, 1 nA and 10 ms). Traces are normalized to the 1st EPSC of the control stimulus to facilitate comparison. *P < 0.05, Student's paired t test.

Figure 5

MSE mimics DSE A, summary of 2-AG/MSE occlusion experiments showing oxo-M/MSE alone, 2-AG alone (5 μm) and oxo-M/MSE in the presence of 2-AG (5 μm, n = 5). B, representative same-cell experiment showing DSE time courses under control conditions (•) and in presence of oxo-M (5 μm, ▴). C, left panel shows a summary of oxo-M/DSE occlusion experiments (same cell) with inhibition by DSE alone and partial occlusion of DSE by oxo-M-induced MSE (n = 6). Right panel shows comparable partial occlusion of DSE by DHPG-mediated MSE in Type O neurons (n = 4). **P < 0.005, paired t test.

Figure 6

Metabotropic DSE enhancement A, Histogram showing duration of depolarization (in seconds) required to achieve a threshold value of at least 20% DSE inhibition (n = 34). B, left panel shows summary of threshold DSE responses, responses to 50 nm oxo-M and synergistic DSE responses in the presence of 50 nm oxo-M in the same neurons (n = 10). Right panel shows average time courses of threshold DSE (from left panel) under control and with 50 nm oxo-M. C, summary of threshold DSE inhibition, responses to 2 μm DHPG and threshold DSE in the presence of DHPG (same cell experiments, n = 14). D, same experiment as C, in Type O neurons (n = 4). *P < 0.05; **P < 0.001, paired t test.

Figure 7

A time-dependent DSE potentiation A, histogram showing degree of DSE at indicated times (0–1 min, n = 9; 1–3 min, n = 7; 3–6 min, n = 16, 6+ min, n = 11). B, averaged DSE time courses from A (at indicated times). One-way ANOVA with Tukey's post hoc test, significant versus 3–6 min (P < 0.001) and 6+ min (P < 0.01).

Figure 8

MSE–MSE and MSE–THC interactions A, summary of same-cell experiments at saturating MSE agonist concentrations showing responses to oxo-M (5–10 μm), DHPG (50–100 μm) and the two agonists in combination. B, summary of similar experiments at threshold concentrations (oxo-M, 50 nm; DHPG, 5 μm). C, Inhibition by DHPG (50–100 μm) in oxo-M (2–10 μm)-pretreated regular and Type O neurons showing heterologous desensitization in the former but not the latter. D, left panel shows summary of MSE responses alone, and in the presence of 50 nm, 200 nm, 500 nm and 5 μmΔ⁹-THC. Right panel shows sample time course of oxo-M-induced MSE alone and in the presence of increasing concentrations of Δ⁹-THC. P < 0.1 versus 200 nm, *P < 0.05 versus 500 nm THC, **P < 0.001 versus 5 μmΔ⁹-THC, one-way ANOVA with Dunnett's post hoc test.