Dendritically released transmitters cooperate via autocrine and retrograde actions to inhibit afferent excitation in rat brain

Abstract

Oxytocin is released from supraoptic magnocellular neurones and is thought to act at presynaptic receptors to inhibit transmitter release. We now show that this effect is mediated by endocannabinoids, but that oxytocin nonetheless plays an important role in endocannabinoid signalling. WIN55,212-2, a cannabinoid receptor agonist, mimicked the action of oxytocin and occluded oxytocin-induced presynaptic inhibition. The cannabinoid action is at the presynaptic terminal as shown by alteration in paired pulse ratio, a reduction in miniature EPSC frequency and immunohistochemical localization of CB1 receptors on presynaptic terminals. AM251, a CB1 receptor antagonist, blocked both the WIN55,212-2 and the oxytocin-induced presynaptic inhibition of EPSCs. Depolarization of postsynaptic magnocellular neurones (which contain fatty acid amide hydrolase, a cannabinoid catabolic enzyme) caused a transient inhibition of EPSCs that could be blocked by both the AM251 and Manning compound, an oxytocin/vasopressin receptor antagonist. This indicates that somatodendritic peptide release and action on previously identified autoreceptors facilitates the release of endocannabinoids that act as mediators of presynaptic inhibition.

Figure 1. Cannabinoid receptor agonist inhibits EPSCs through CB₁ receptors

A, evoked EPSCs before (1), during (2) and after (3) the application of WIN 55,212-2 (1 μm). The time–effect plot of evoked EPSCs indicates the times (shown by the numbers) when the traces were taken. WIN55,212-2 was bath applied during the time indicated by the horizontal bar. B, a representative time–effect plot of evoked EPSC amplitude. Pretreatment with AM251 (1 μm) blocked the inhibitory effect of WIN55,212-2. C, summary graph showing the change in EPSC amplitude by WIN55,212-2 alone or AM251 and WIN55,212-2. * P < 0.05 versus control.

Figure 2. WIN55,212-2 increases paired pulse ratio (PPR)

A: top, representative paired pulse traces during control (left) and in the presence of WIN55,212-2 (right); bottom, the traces are scaled to the size of the first EPSC. B, WIN55,212-2-induced PPR change plotted against the change in the first EPSC amplitude. Each filled circle represents a cell.

Figure 3. Miniature EPSCs are inhibited by WIN55,212-2

Aa, sample traces showing basal mEPSCs (control) and in the presence of WIN55,212-2 as indicated. Ab and c, cumulative plots of interevent intervals (Ab) or amplitudes (Ac) of mEPSCs recorded from the same cell as shown in Aa. For A and B: black line, control; grey line, WIN55,212-2. Ba, sample traces showing mEPSCs induced by nifedipine (10 μm) and in the presence of both nifedipine and WIN55,212-2. Bb and c, cumulative plots of interevent interval (Bb) or amplitude (Bc) of mEPSCs generated from the same cell as shown in Ba. C, summary of the effect of WIN55,212-2 on the frequency of basal and nifedipine-induced mEPSCs. D, summary graph depicting the effect of WIN55,212-2 on the amplitude of basal and nifedipine-induced mEPSCs. * P < 0.05 versus control.

Figure 4. Oxytocin inhibits EPSCs by inducing endocannabinoid release

A, oxytocin (OXT; 1 μm) inhibits EPSC amplitude in a typical cell. B, AM251 (1 μm) blocks oxytocin effect. C, WIN55,212-2 (1 μm)-induced EPSC inhibition occludes oxytocin effect. D, summary of the effect of oxytocin on EPSC amplitude. The effect of oxytocin is normalized to the value prior to oxytocin application. * P < 0.05 versus control. E, Manning compound does not block WIN 55,212-2 effect. The inhibitory effect of WIN55,212-2 is reversed by AM251. F, summary graph of the effect of WIN55,212-2 on EPSC amplitude. * P < 0.05 versus control. The presence of Manning compound (MC) has no significant effect (n.s.).

Figure 5. Oxytocin has an endocannabinoid-independent excitatory effect on miniature EPSCs

A, example of a cell that responded to OXT with an increase in mEPSC frequency. B, sample traces showing the effect of OXT in the presence of AM251. C, effect of OXT on mEPSC frequency in the absence or presence of AM251. Each circle represents a cell. Mean and error bars are shown beside each group. * P < 0.05 versus control. D, effect of OXT on mEPSC amplitude in the absence or presence of AM251. Symbols as in C.

Figure 6. Postsynaptic depolarization induces inhibition of evoked EPSCs

Aa, depolarization of the postsynaptic magnocellular neurone (0 mV, 1 s) induces a transient depression of evoked EPSCs. In Aa, Ba and Ca, arrows indicate when the cells were depolarized. In Aa, Ba and Ca, all values were normalized to control in each cell. Control value was calculated as an average of 3–5 min prior to depolarization. * P < 0.05 versus control. Ba, the inhibitory effect of postsynaptic depolarization was blocked by Manning compound (10 μm), a OXT/V_1a receptor antagonist. Ca, the inhibitory effect of postsynaptic depolarization was blocked by AM251 (1 μm). Ab, Bb, Cb, representative current traces recorded from three different cells.

Figure 7. Postsynaptic depolarization induces inhibition of miniature EPSCs

A, sample traces recorded before (Control), within 1 min after depolarization (Post depolarization) and 4 min later (Recovery). B, cumulative plot of mEPSC interevent interval. For B and C: black line, control; grey line, post depolarization. The data were derived from the same cell as A. C, cumulative plot of mEPSC amplitude. D, summary graph showing the change in frequency and amplitude of mEPSCs after postsynaptic depolarization.

Figure 8. CB₁ receptor is in presynaptic processes

A1, confocal fluorescence micrographs of FAAH immunoreactivity (FAAH, 12 optical sections). A2, CB₁ receptor immunoreactivity (CB₁R, 20 optical sections). Scale bars for A1 and A2, 50 μm. B, electron micrographs displaying a bundle of axons (a) surrounded by glial processes (asterisks) and dendrites (d). One of the axons is labelled for CB₁ receptors (arrowhead; dark deposit). Scale bars in B–F, 300 nm. C and D, asymmetric synapse with CB₁ receptor-labelled synaptic terminal (st) contacting a postsynaptic cell. Arrows, postsynaptic density; ds, dendritic spine. E and F, symmetric synapse between a labelled synaptic terminal and a dendrite.

Figure 9. Postulated mechanism for cooperative action of dendritically released OXT and endocannabinoids in the supraoptic nucleus

OXT initiates the temporal and spatial spread of endocannabinoid signalling. (1) Initial depolarization of an OXT neurone induces OXT release. (2) OXT action on autoreceptors leads to synthesis and release of sufficient endogenous endocannabinoids which may be dependent or independent of Ca²⁺ release from internal stores. (3) Endocannabinoids diffuse to CB₁ receptors on presynaptic terminals and inhibit glutamate release. (4) Ca²⁺ release from internal stores also causes further release of OXT, generating a feed-forward stimulation of OXT–endocannabinoid signalling. (5) OXT may also diffuse to adjacent neurones and induce endocannabinoid release. (6) OXT has an excitatory effect on spontaneous glutamate release, which may be due to a direct action on the presynaptic terminal.