Dominique A. Poulain

Activity-Dependent Structural and Functional Plasticity of Astrocyte-Neuron Interactions

Observations from different brain areas have established that the adult nervous system can undergo significant experience-related structural changes throughout life. Less familiar is the notion that morphological plasticity affects not only neurons but glial cells as well. Yet there is abundant evidence showing that astrocytes, the most numerous cells in the mammalian brain, are highly mobile. Under physiological conditions as different as reproduction, sensory stimulation, and learning, they display a remarkable structural plasticity, particularly conspicuous at the level of their lamellate distal processes that normally ensheath all portions of neurons. Distal astrocytic processes can undergo morphological changes in a matter of minutes, a remodeling that modifies the geometry and diffusion properties of the extracellular space and relationships with adjacent neuronal elements, especially synapses. Astrocytes respond to neuronal activity via ion channels, neurotransmitter receptors, and transporters on their processes; they transmit information via release of neuroactive substances. Where astrocytic processes are mobile then, astrocytic-neuronal interactions become highly dynamic, a plasticity that has important functional consequences since it modifies extracellular ionic homeostasis, neurotransmission, gliotransmission, and ultimately neuronal function at the cellular and system levels. Although a complete picture of intervening cellular mechanisms is lacking, some have been identified, notably certain permissive molecular factors common to systems capable of remodeling (cell surface and extracellular matrix adhesion molecules, cytoskeletal proteins) and molecules that appear specific to each system (neuropeptides, neurotransmitters, steroids, growth factors) that trigger or reverse the morphological changes.

Oxytocin neuron activation and fos expression: A quantitative immunocytochemical analysis of the effect of lactation, parturition, osmotic and cardiovascular stimulation

As c-fos expression is generally thought to be linked to neuronal activation, we compared Fos immunoreactivity in identified oxytocinergic and vasopressinergic neurons of female rats under various conditions known to elicit particular patterns of electrophysiological and secretory activity in these neurons. In suckled lactating animals, Fos immunoreactivity was visible only in rare oxytocinergic and vasopressinergic neurons of the paraventricular and supraoptic nuclei, even after interruption of suckling for 18-72 h. On the other hand, many Fos-positive cells were visible in the nuclei of parturient rats; they involved about 25% of supraoptic oxytocinergic elements. Even more Fos-positive elements were visible in the nuclei of lactating rats that had also undergone 24 h water deprivation or haemorrhage. This involved about 75% vasopressinergic neurons and 25% oxytocinergic neurons of the supraoptic nucleus. Fos immunoreactivity was particularly conspicuous in oxytocin neurons of the anterior commissural nucleus after haemorrhage. After water deprivation or haemorrhage, Fos-positive oxytocinergic neurons in the supraoptic nucleus were significantly more numerous in virgin rats than in lactating rats. Our observations show that suckling, although a most potent stimulus for oxytocin neuron activation and oxytocin release, is inefficient in inducing Fos synthesis in magnocellular neurons, even after a period of interruption. On the other hand, parturition, water deprivation and haemorrhage were more potent stimuli for both neurosecretory systems. However, under each type of stimulation, only part of the neuronal populations within each nucleus were Fos-positive, suggesting that different stimulus-specific pathways are involved in these regulations. In so far as electrical activity is one possible mechanism for c-fos expression, comparison of the patterns of c-fos activation with the known electrophysiological behaviour of hypothalamic magnocellular neurons suggests that Fos synthesis in these neurons is linked to the number of action potentials generated over a period of time, more than to the pattern of electrical activity, whatever the physiological impact of this pattern. Furthermore, within a group of neurons, the heterogeneity of the response in terms of Fos synthesis may be correlated to the variability of the electrophysiological response within this group.

Septal release of vasopressin in response to osmotic, hypovolemic and electrical stimulation in rats

The central release of vasopressin was studied in anesthetized rats using push-pull perfusions and radioimmunoassay of the hormone. A basal release was observed in the lateral septum and in the lateral ventricle, whereas no vasopressin was detected in the perfusates from the caudate nucleus. Under osmotic stimulation, vasopressin release increased up to 12 and 60 times basal levels following i.p. injections of 5 ml and 10 ml/kg b.wt. of 2 M NaCl, respectively. This increase was blocked by using a calcium-free perfusion medium containing 0.1 mM EGTA. In the lateral ventricle, osmotic stimulation (5 ml/kg of 2 M NaCl i.p.) had the same effect as in the septum. In the caudate nucleus, no release was observed. Hemorrhage also increased the septal release of vasopressin in 5 out of 6 animals tested. Electrical stimulation of the pituitary stalk and of the supraoptic nucleus was used to evoke the release of vasopressin into the bloodstream. Septal release slightly decreased during pituitary stalk stimulation, whereas it did increase during stimulation of the supraoptic region. Our results show that systemic stimuli for vasopressin release evoke both a peripheral and a septal release of the hormone. The dissociation of the effects of electrical stimulation of the pituitary stalk and of the supraoptic nucleus suggests, however, that the vasopressinergic neurones responsible for septal release are distinct from those which project to the neurohypophysis.

The Glutamatergic Innervation of Oxytocin‐ and Vasopressin‐secreting Neurons in the Rat Supraoptic Nucleus and its Contribution to Lactation‐induced Synaptic Plasticity

The present ultrastructural study analysed the distribution of glutamatergic synapses on oxytocin‐ and vasopressin‐secreting neurons in the rat supraoptic nucleus (SON) after post‐embedding immunogold labelling for glutamate and GABA, oxytocin or vasopressin. About 20% of SON axo—somatic synapses were enriched in glutamate immunoreactivity, visible over synaptic‐like vesicles, mitochondria and synaptic densities. Double labelling for glutamate and GABA showed that putative glutamatergic terminals were distinct from GABAergic terminals. In ultrathin sections stained for glutamate and either oxytocin or vasopressin, the proportion of glutamatergic synapses was similar on oxytocinergic and vasopressinergic somata in virgin rats under basal conditions of peptide release as well as in lactating rats, in which oxytocin secretion is enhanced. Cross‐sectional soma areas were significantly increased in lactating rats: oxytocinergic profiles were, on average, ˜40% larger than in virgin rats. However, the incidence of axo—somatic glutamatergic synapses (assessed as mean number of synapses per 100 μm of plasmalemma or proportion of somatic surface apposed to synaptic active zones) did not diminish, indicating that there was a compensatory increase of synapses during lactation. Also, we found an increase in the number of glutamatergic terminals making synaptic contact simultaneously onto two or more oxytocinergic elements in the same plane of section. Our observations therefore indicate that SON oxytocinergic and vasopressinergic neurons are innervated to a similar extent by a relatively large proportion of glutamatergic synapses. They reveal, moreover, that glutamatergic afferents participate in the lactation‐induced synaptic plasticity of the oxytocinergic system.

Oxytocin-secreting neurones: a physiological model for structural plasticity in the adult mammalian brain

Abstract It is now evident that the adult brain is capable of structural plasticity in response not only to lesions, but also to physiological stimuli. The hypothalamic oxytocinergic system offers a striking example of an adult mammalian neuronal system that can undergo such plasticity, since its basic architecture and synaptic circuitry is reversibly modified under particular physiological conditions. During parturition, lactation, and even prolonged dehydration, glial coverage of oxytocinergic neurones markedly diminishes, and their surfaces are left in extensive juxtaposition; concurrently, synaptic remodelling associates two or more of the neurones by creating common presynaptic terminals. The numerous and extensive neuronal juxtapositions and the shared synaptic input may have important functional consequences in facilitating the synchronized electrical activity of the neurones under certain physiological conditions, such as lactation. In the neurohypophysis, where neurosecretory terminals release hormone directly into the circulation, stimulation induces glial retraction from the perivascular space, thus enlarging the neurohaemal contact area. Although the mechanisms by which these anatomical changes occur are still unknown, oxytocin itself appears to be of primary importance in the reorganization of its own system, at least within the hypothalamus.

Influence of Ovarian Steroids on the Ultrastructural Plasticity of the Adult Rat Supraoptic Nucleus Induced by Central Administration of Oxytocin

In earlier studies, we showed that continuous intracerebroventricular infusion of oxytocin, for several days, into the third ventricle of normally hydrated, non‐lactating adult female rats significantly reduced glial coverage of magnocellular oxytocinergic neurons in the hypothalamus. It also induced synaptic remodelling whereby many oxytocinergic neurons became synaptically contacted by the same presynaptic terminals (shared synapses). Such changes were closely similar to those observed in the oxytocinergic system when it is physiologically activated, as during parturition and lactation.

Oxytocinergic innervation of the rat spinal cord. An electron microscopic study

Pre- and postembedding immunocytochemical procedures were used, together with antisera raised against oxytocin or its neurophysin, to characterize oxytocinergic pathways in the rat spinal cord, at the electron microscopic level. Pre-embedding immunoperoxidase staining performed on vibratome sections revealed oxytocin- and neurophysin-positive axonal profiles and terminals scattered predominantly in laminae I and II of the dorsal horn and in the central gray (lamina X). They were also visible, but to a lesser extent, in the intermediolateral columns, at thoracic and lumbar levels. Postembedding immunogold staining performed directly on ultrathin sections of the same areas, fixed in osmium and embedded in resin, permitted to show clearly that the oxytocinergic axons made symmetrical and asymmetrical synaptic contacts onto dendritic profiles. It also allowed subcellular localization of the neuropeptide immunoreactivities which were restricted to relatively large, electron-dense vesicles in the immunopositive terminals. Oxytocinergic terminals were never seen to participate in glomerular configurations in the superficial layers of the dorsal horn nor were immunoreactive cell bodies visible in any spinal area. Our results provide direct morphological evidence that oxytocinergic pathways make synapses in several regions of the spinal cord, thus supporting the contention that oxytocin may exert neurotransmitter/neuromodulator actions in this area of the CNS.

Neuronal-glial and synaptic plasticity in the adult rat paraventricular nucleus.

Using quantitative ultrastructural analysis on cells identified by immunogold postembedding immunocytochemistry, we show that magnocellular oxytocinergic neurons in the adult rat paraventricular nucleus (PVN) undergo significant neuronal-glial and synaptic changes upon stimulation. Thus, during lactation, the surface membranes of most PVN oxytocinergic somata and dendrites were directly juxtaposed; many were also contacted synaptically by the same axonal terminal (shared synapses). Non-oxytocinergic profiles showed few plasmalemma juxtapositions and shared synapses. These ultrastructural changes are similar to those that modify oxytocin neurons in the supraoptic nucleus under the same conditions, and indicate that the whole oxytocinergic system in the hypothalamus is capable of neuronal-glial and synaptic plasticity when stimulated to release its neurohormone.

Synaptic and neuronal-glial plasticity in the adult oxytocinergic system in response to physiological stimuli.

Magnocellular oxytocinergic neurons in the hypothalamus offer a striking example of a mammalian neuronal system whose basic architecture and synaptic circuitry can be reversibly modified in adulthood. During parturition, lactation and prolonged osmotic stimulation, glial coverage of oxytocinergic neurons markedly diminishes and their surfaces are left in extensive juxtaposition; concurrently, there is formation of new synapses, which are predominantly GABAergic and which couple two or more oxytocinergic neurons simultaneously. These structural changes do not permanently modify the anatomy of the system since upon cessation of stimulation, neuronal juxtapositions and shared synapses disappear, to reappear upon new stimulation. At present, we can only speculate about the cellular mechanisms and factors responsible for these reversible neuroanatomical changes. However, oxytocin itself appears to be of primary importance since it can induce similar anatomical changes when chronically infused into the third ventricle.

Effects of estrogen on the electrical activity of identified and unidentified hypothalamic units.

Experiments performed on unanesthetized ovariectomized female rabbits demonstrated the effects of estradiol benzoate (EB; 20 microng i.v.) on the electrical activity of hypothalamic units which send their axons to the median eminence. Of a total of 1,840 cells recorded in hypothalamic and preoptic areas, 46 (2.5%) were antidromically activated by stimulating the median eminence. Under the present experimental conditions, EB induced a progressive diminution in the mean firing rate of these cells observed throughout the recording period (30-120 min). In addition to cells projecting to the median eminence, neurons which could not be antidromically invaded using our techniques were observed to be sensitive to estrogen. Estrogen administration produced a long-lasting inhibition of antidromically activated cells and a depression of much shorter duration (15-20 min) of unidentified nonstimulated units. These data suggest the existence of two types of hypothalamic neurons sensitive to estrogen.