Editorial Commentary:Ready to Burst?
J. Russell,G. Leng
DOI: https://doi.org/10.1046/j.1365-2826.2001.00874.x
2002-12-01
Journal of Neuroendocrinology
Abstract:During parturition in the rat there is a sustained increase in oxytocin secretion, superimposed on which are pulses of secretion, each associated with the birth of a pup. The sustained increase is the result of an increase in the background electrical activity of magnocellular oxytocin neurones, and the pulses are the result of intermittent high-frequency bursts of activity. The bursts occur approximately synchronously in all the magnocellular oxytocin neurones, in both supraoptic nuclei and both paraventricular nuclei (1). Whereas the increase in ‘background’ firing is typical of the responses of oxytocin neurones to many other stimuli, bursting is elicited in oxytocin neurones only during parturition or in response to suckling. During parturition, bursts appear to accompany the passage of a foetus through the birth canal, but bursts can also be elicited by suckling in latepregnant or parturient rats (2). Thus the ability of oxytocin neurones to fire in bursts is already present before births start. Because bursting is only exhibited in late pregnancy and parturition, it is often suspected that pregnancy induces specific adaptations that favour the expression of bursting behaviour. This is a superficially attractive hypothesis but may be quite wrong: oxytocin cells even during lactation do not exhibit bursts in response to any stimulus other than suckling; in lactating rats oxytocin neurones respond to most stimuli in much the same way as they do in virgin rats. Nevertheless, there are a number of well documented changes that seem to reflect significant changes in the intrinsic properties of oxytocin neurones, or significant changes in the neuronal networks that directly regulate them. What these changes mean is less clear, though the interpretation often offered is that these changes contribute to the ability to generate bursts. For instance, there is an increase in the incidence of GABA synapses on oxytocin cells by the end of pregnancy, but there seems to be a reduction in the efficacy of GABA as the tonic GABA current density in each oxytocin cell is not altered (3). The potentiation by progesterone metabolites at these synapses in pregnancy seems to guard against premature activation of oxytocin neurones rather than contributing to the burst-firing mechanism, because progesterone secretion collapses prior to parturition. There are marked changes in the glial architecture of the supraoptic nucleus at parturition that may favour dendro–dendritic interactions between oxytocin neurones mediated by oxytocin release, and which may favour glutamatergic excitation by reducing extracellular glutamate uptake (4); but then again, oxytocin released from dendrites is believed to be a key element in bursting, and this suppresses glutamate release (5). Not everything that happens to oxytocin neurones in pregnancy will be part of the explanation for bursting, and perhaps none of the observed changes have much to do with it. For all that we really know, the only reason why we don’t see bursting in oxytocin neurones in a virgin rat or a male rat is because they’re not giving birth or being suckled. However, the explanation of the ability of oxytocin neurones to fire in bursts has proved to be a challenging problem, precisely because this phenomenon is only observed in such exceptional circumstances. In oxytocin neurones generally, every spike is followed by a hyperpolarization that results from activation of Ca-dependent K channels, and this hyperpolarization limits the firing rate of oxytocin cells. There are, broadly, two components to the postspike hyperpolarization, a large fast transient hyperpolarization (the hyperpolarizing afterpotential, HAP), and a small long-lasting hyperpolarization (the afterhyperpolarization, or AHP). Under almost all known physiological circumstances, there will be an interval of at least 30 ms between successive spikes, leading to a maximum firing rate of less than 40 Hz, and in practice the maximum sustainable discharge rate of oxytocin cells is closer to 15 Hz. During a milk-ejection burst, interspike intervals as short as 6 ms are observed, and intervals of less than 10 ms are common. Between bursts however, such short intervals are not seen. A paper in this issue of The Journal (6) adds to the catalogue of changes in oxytocin neurones that may reflect adaptations to facilitate bursting, or which may reflect adaptations to facilitate interactions between oxytocin neurones. Compared with virgin rats, oxytocin neurones from late pregnant rats or from lactating rats have broader action potentials reflecting enhanced voltage-gated Ca influx, have larger AHPs, but with faster decay times, and more of them display depolarizing after-potentials (DAPs). A cell exhibiting a HAP, DAP and AHP will show a sequence of hyperpolarization, depolarization and afterhyperpolarization following an individual spike. DAPs are rarely seen in oxytocin neurones normally, but are commonly seen in neighbouring vasopressin neurones. The consequences of an enhanced AHP with a more rapid decay are not easy to predict; because the AHP is a small, long-lasting hyperpolarization, for it to exert a significant effect on firing rate requires summation following repetitive spike activity. The AHP seems to serve as a damping, smoothing influence on firing rate, and the observed changes in AHP decay rate may suggest that greater irregularity in firing rate will result. However the enhanced amplitude of the AHP would certainly not favour the ability to express bursts. On the other hand, the greater incidence of a DAP will tend to make these neurones more excitable, and may enhance the effect of any positive-feedback effects of dendro–dendritic interaction between oxytocin neurones. The DAP may be part of a local positive-feedback loop in which the broader action potentials in oxytocin neurones in pregnancy lead to more Journal of Neuroendocrinology, 2002, Vol. 14, 927–928