Search

1. GABAergic innervation of oxytocin neurones is particularly abundant during lactation, but little is known about its functional role. In this study, the role of GABAA receptors in the suckling-induced bursting activity of oxytocin neurones was investigated in lactating rats. GABAA agonists or antagonists were applied by pressure injection into the immediate neighbourhood of recorded neurones while simultaneous recordings were made from oxytocin neurones in the contralateral supraoptic nucleus. 2. GABA and the GABA agonist isoguvacine decreased the basal electrical activity while application of GABAA antagonists (picrotoxin and gabazine) increased the basal electrical activity. However, in marked and unexpected contrast, application of GABA and isoguvacine facilitated or triggered milk-ejection reflex bursting activity whereas GABAA antagonists interrupted this reflex activity. 3. Systemic injection of hypertonic saline is known to increase the firing rate of neurones in the supraoptic nucleus and temporarily to interrupt suckling-induced bursting activity. Application of GABA into one supraoptic nucleus counteracted this inhibitory effect on milk ejection. 4. These observations can be explained if the role of the important GABAergic innervation of oxytocin neurones during lactation is to favour the expression of the stereotyped suckling-induced bursting activity. It might do this by attenuating inputs unrelated to suckling which are incompatible with bursts.

Central oxytocin administration has a profound facilitatory effect on the patterning of the milk-ejection reflex in the lactating rat. Lesion and microinjection studies indicate that this action is, in part, mediated via a population of limbic neurones in the bed nuclei of the stria terminalis and ventrolateral septum, which have been shown to possess oxytocin receptors and to be activated by selective oxytocin-receptor agonists in vitro. In vivo electro-physiological recordings reveal that some of these neurones display cyclical activity which is highly correlated to each milk ejection, and are rapidly activated following i.c.v. administration of oxytocin. coincident with the facilitation of milk ejection activity. A hypothetical model is proposed in which this population of limbic neurones serves to gate the activity of a pacemaker which, in turn, coordinates the bursting of hypothalamic magnocellular neurones. The oxytocin innervation of these neurones and their expression of oxytocin receptors increases in the post-partum period, and the resultant enhanced sensitivity leads to a greater facilitatory response during lactation. Inhibitory opioid and noradrenergic inputs which converge on these oxytocin-sensitive neurones may function to switch off the facilitatory circuit during periods of stress. Thus, this population of limbic neurones participates in the regulation of neuroendocrine activity during lactation by providing an appropriate degree of feedback to alter the patterning of the milk-ejection reflex.

Many neurotransmitters exert an autocrine control at axon terminals via autoreceptors that modulate activity-dependent exocytosis; less classical, and more complex, is the autocontrol exerted at soma and dendrites. One interesting example is provided by vasopressin neurones of the hypothalamo–neurohypophysial system, especially as the physiological relevance of the autocontrol has been widely studied. During sustained hormonal demand, vasopressin neurones adopt a phasic pattern of activity, with alternating active and silent periods (bursts), that each last 10–40 s. This phasic pattern is particularly efficient for the release of vasopressin, but the bursts are not co-ordinated between vasopressin neurones, so vasopressin is released continuously into the circulation, ensuring its effectiveness at peripheral targets. The phasic patterning of vasopressin neurones depends upon intrinsic membrane properties. Synaptic inputs to vasopressin neurones appear to arrive randomly, and bursts are initiated from summation of excitatory postsynaptic potentials (EPSPs). Each spike is followed by a prominent, long lasting (1–3 s), non-synaptic depolarizing after-potential (DAP) (Andrew & Dudek, 1983); consecutive spikes lead to summation of DAPs, and a resulting, long-lasting plateau potential sustains burst firing. As a burst progresses, DAPs, and plateau potentials undergo pronounced activity-dependent inactivation, and it is believed that this inactivation ultimately leads to termination of the burst. Thus, bursts are both sustained and terminated by activity-dependent modulation of intrinsic membrane properties. The intrinsic properties that underlie phasic patterning are regulated by autocrine–paracrine control. From the soma and dendrites of vasopressin neurones, neurosecretory granules are secreted by exocytosis in response to specific physiological stimuli. In these granules, vasopressin is co-stored with several molecules including ATP and dynorphin, and interestingly, the granules also contain vasopressin receptors and κ-opioid receptors. This co-localization facilitates binding of the peptides to their receptors at exocytosis. Dendritic release is partly activity dependent and partly self-sustaining, since vasopressin itself can elicit dendritic release without increasing electrical activity. In vivo, vasopressin modulates the phasic pattern of vasopressin neurones depending on their initial electrical activity: fast-firing neurones are slowed, and slow-firing neurones are excited. Thereby, vasopressin fosters the population of vasopressin neurones to express the phasic pattern of activity that is most efficient for vasopressin release from the axon terminals (Gouzenes et al. 1998). At the soma and dendrites of vasopressin neurones, vasopressin, via actions at V1a and V1b receptors that involve different intracellular second messenger pathways, induces both mobilization of intracellular Ca2+ and Ca2+ entry. The role of the V1b receptor is unclear, but there is now consensus that the inhibitory effect of vasopressin on phasic patterning involves V1a receptor activation. The effect of vasopressin on autoreceptors is complemented by a presynaptic action on afferent terminals, involving inhibition of glutamate release (Kombian et al. 2000). The autocrine–paracrine control also involves co-stored and co-released molecules, including in particular dynorphin. Both vasopressin and dynorphin (Brown et al. 1998) restrain the activity of phasic neurones. As reported in the previous issue of The Journal of Physiology, Brown & Bourque (2004) made intracellular recordings of vasopressin neurones from hypothalamic explants to explore the possibility that endogenous dynorphin restrains vasopressin neurones by an autocrine action on intrinsic membrane properties. DAPs and plateau potentials were evoked by triggering brief trains of spikes with depolarizing current pulses. DAPs that are evoked soon after a preceding evoked DAP are attenuated, and this effect is similar to that observed when DAPs are evoked soon after a spontaneous burst. The degree of attenuation depends on the number of spikes elicited before the evoked DAP, demonstrating that DAPs undergo pronounced activity-dependent inactivation that will reduce the probability of generation of spontaneous spikes as the burst progresses. Brown & Bourque (2004) then used pharmacological tools to demonstrate that dendritically released dynorphin generates this activity-dependent inhibition. First, inducing neurosecretory vesicle depletion by application of the black widow spider venom, α-latrotoxin, reduced activity-dependent DAP inhibition. Second, the action of α-latrotoxin was prevented by κ-opioid receptor antagonists but not by vasopressin receptor antagonists. Third, activity-dependent κ-opioid inhibition of DAPs does not result from actions on evoked after-hyperpolarizations since these are not affected by κ-opioid receptor antagonism. This suggests that the Ca2+-dependent K+ conductances that underlie after-hyperpolarizations are not involved in activity-dependent κ-opioid inhibition of DAPs. The study brings clear evidence of modulation of intrinsic membrane properties of vasopressin neurones by endogenous feedback, in this case by dendritic dynorphin release. It also clarifies the different roles of different, co-released, peptides: vasopressin decreases EPSCs via V1a receptors, while dynorphin inhibits DAPs via κ-opioid receptors. The development of the restraint exerted by each peptide also differs: restraint by vasopressin is sustained throughout each burst, while the effects of dynorphin emerge progressively as the burst develops (Brown et al. 2004). A final question pertains to the mechanisms underlying this temporal dissociation in the actions of vasopressin and dynorphin. This may be related to the respective concentration of the two peptides in neurosecretory vesicles (vasopressin is more abundant than dynorphin), to differential accumulation rates in the extracellular space, or to differential degradation by extracellular peptidases. However, it more probably results from the different ways that the two neuropeptides affect rhythmogenesis. Since EPSCs are randomly patterned, reduction of EPSC amplitude by vasopressin should result in a general reduction in excitability. Conversely, modulation of activity-dependent DAPs would be expected to also be activity dependent. These studies lay the foundations for further studies on action of co-stored and co-released molecules implicated in the control of neuronal excitability and provide useful trails for investigating their roles in the patterning of electrical activity of neurones in the CNS.

n/a