Homer 1a gates the induction mechanism for endocannabinoid-mediated synaptic plasticity.

At hippocampal excitatory synapses, endocannabinoids (eCBs) mediate two forms of retrograde synaptic inhibition that are induced by postsynaptic depolarization or activation of metabotropic glutamate receptors (mGluRs). The homer family of molecular scaffolds provides spatial organization to regulate postsynaptic signaling cascades, including those activated by mGluRs. Expression of the homer 1a (H1a) immediate-early gene produces a short homer protein that lacks the domain required for homer oligomerization, enabling it to uncouple homer assemblies. Here, we report that H1a differentially modulates two forms of eCB-mediated synaptic plasticity, depolarization-induced suppression of excitation (DSE) and metabotropic suppression of excitation (MSE). EPSCs were recorded from cultured hippocampal neurons and DSE evoked by a 15 s depolarization to 0 mV and MSE evoked by a type I mGluR agonist. Expression of H1a enhanced DSE and inhibited MSE at the same synapse. Many physiologically important stimuli initiate H1a expression including brain-derived neurotrophic factor (BDNF). Treating hippocampal cultures with BDNF increased transcription of H1a and uncoupled homer 1c-GFP (green fluorescent protein) clusters. BDNF treatment blocked MSE and enhanced DSE. Thus, physiological changes in H1a expression gate the induction pathway for eCB-mediated synaptic plasticity by uncoupling mGluR from eCB production.

Modulation of excitatory synaptic transmission by Delta 9-tetrahydrocannabinol switches from agonist to antagonist depending on firing rate.

Delta(9)-tetrahydrocannabinol (THC), the principal psychoactive ingredient in marijuana, acts as a partial agonist on presynaptic cannabinoid type 1 (CB1) receptors to inhibit neurotransmitter release. Here, we report that THC inhibits excitatory neurotransmission between cultured rat hippocampal neurons in a manner highly sensitive to stimulus rate. THC (1 microM) inhibited excitatory postsynaptic currents (EPSCs) and whole-cell I(Ca) evoked at 0.1 Hz but at 0.5 Hz THC had little effect. The cannabinoid receptor full agonists [(R)-(+)-[2,3-dihydro-5-methyl-3[(4-morpholinyl)methyl]pyrrolo[1,2,3-de]-1,4-benzoxazinyl]-(1-naphthalenyl)methanone mesylate salt] (Win55212-2) (100 nM) and 2-arachidonylglycerol (1 microM) inhibited EPSCs independent of stimulation at 0.1 or 0.5 Hz. THC occupied CB1 receptors at 0.5 Hz, but the receptors failed to couple to presynaptic Ca(2+) channels. Consequently, 1 microM THC blocked the inhibition of EPSC amplitude by Win55212-2 when EPSCs were evoked at 0.5 Hz. A depolarizing prepulse to 0 mV reversed THC inhibition of I(Ca), but reversal of the inhibition produced by Win55212-2 required a pulse to +80 mV, suggesting that the voltage-dependent reversal of Gbetagamma inhibition of voltage-gated Ca(2+) channels accounts for the frequency-dependence of cannabinoid action. THC blocked depolarization-induced suppression of EPSCs evoked at 0.5 Hz, indicating that it inhibited retrograde endocannabinoid signaling in a frequency-dependent manner. Thus, THC displayed a state-dependent switching from agonist to antagonist that may account for its complex actions in vivo.

Excitotoxic loss of post-synaptic sites is distinct temporally and mechanistically from neuronal death.

Dendritic degeneration and loss of synaptic proteins are early events correlated with functional decline in neurodegenerative disease. The temporal and mechanistic relationship between synapse loss and cell death, however, remains unclear. We used confocal microscopy and image processing to count post-synaptic sites on rat hippocampal neurons by expressing post-synaptic density protein 95 fused to green fluorescent protein. Fluorescent puncta co-localized with neurotransmitter release sites, NMDA-induced Ca2+ increases and NMDA receptor immunoreactivity. During excitotoxic neurodegeneration, synaptic sites were lost and synaptic transmission impaired. These changes were mediated by NMDA receptors and required Ca2+-dependent activation of the proteasome pathway. Tracking synapses from the same cell following brief neurotoxic insult revealed transient loss followed by recovery. The time-course, concentration-dependence and mechanism for loss of post-synaptic sites were distinct from those leading to cell death. Cells expressing p14ARF, which inhibits ubiquitination of post-synaptic density protein 95 and prevents loss of synaptic sites, displayed an increased sensitivity to glutamate-induced cell death. Thus, excitotoxic synapse loss may be a disease-modifying process rather than an obligatory step leading to cell death. These results demonstrate the importance of assessing synaptic function independent of neuronal survival during neurodegeneration and indicate that this approach will be useful for identifying toxins that degrade synaptic connections and for screening for agents that protect synaptic function.