Olivier Bosler

Glucocorticoids up‐regulate the expression of glial fibrillary acidic protein in the rat suprachiasmatic nucleus

Immunoreactivity against glial fibrillary acidic protein (GFAP) was used as a dynamic index in adrenalectomized rats subjected or not to corticosterone replacement to investigate whether glucocorticoids may interact with astrocytes in the suprachiasmatic nucleus (SCN), the master component of the central circadian clock. GFAP staining in the SCN was significantly higher in rats having received implants that restored physiological plasma levels of corticosterone within diurnal or nocturnal limits than in non‐normalized rats. The effects of corticosterone were similar in the parvocellular portion of the paraventricular nucleus but were opposite in the hippocampus, another major site of negative feed‐back regulation of the hypothalamic‐pituitary‐adrenal axis, where a decreased GFAP staining was observed in discrete regions of the dentate gyrus. This indicates that glucocorticoids may positively or negatively regulate GFAP, depending on the target brain structure. In the SCN, that contains only few if any glucocorticoid receptors, indirect mechanisms that may involve serotoninergic neurons are probably responsible for the effects of corticosterone level. It is proposed that the corticosterone‐induced increase in GFAP staining in that nucleus accounts for dynamic changes in neurone‐astrocyte interactions that might occur in relation with natural fluctuations of glucocorticoids over the 24 h period.GLIA 29:212–221, 2000.

Circadian Binding Activity of AP‐1, a Regulator of the Arylalkylamine N‐Acetyltransferase Gene in the Rat Pineal Gland, Depends on Circadian Fra‐2, c‐Jun, and Jun‐D Expression and Is Regulated by the Clock's Zeitgebers

Abstract: The daily rhythm in circulating melatonin is driven by a circadian rhythm in the expression of the arylalkylamine N‐acetyltransferase gene in the rat pineal gland. Turning off expression of this gene at the end of night is believed to involve inhibitory transcription factors, among which Fos‐related antigen 2 (Fra‐2) appears as a good candidate. Circadian rhythms in the expression of three proteins of activating protein‐1 (AP‐1) complexes, namely, Fra‐2, c‐Jun, and Jun‐D, are shown here to account for circadian variations in AP‐1 binding activity. Quantitative variations in the Fra‐2 component over the circadian cycle were associated with qualitative variations in protein isoforms. Destruction of the suprachiasmatic nucleus resulted in decreased nocturnal AP‐1 activity, showing that AP‐1 circadian rhythm is driven by this nucleus. Exposure to light during subjective night and administration of a serotonin 5‐HT1A/5‐HT7 receptor agonist during subjective day, respectively, induced a 50% decrease and a 50% increase in both AP‐1 and Fra‐2 expression. These effects were impaired by suprachiasmatic nucleus lesions. These data show that pineal AP‐1 binding activity, which results from Fra‐2 expression, can be modulated by light and serotonin through the suprachiasmatic nucleus according to a “phase dependence” that is characteristic of the rhythm of clock sensitivity to both zeitgebers.

Convergent serotonin and GABA innervation of VIP neurons in the suprachiasmatic nucleus demonstrated by triple labeling in the rat

By means of a combination of serotonin (5-HT) uptake radioautography and dual immunocytochemistry, vasoactive intestinal polypeptide (VIP) neurons in the suprachiasmatic nucleus are demonstrated to simultaneously receive both 5-HT and GABA afferents at their somatic and dendritic levels. These data constitute a further step towards the improved characterization of the morphological substrate of the integrative function of these neurons, which are known to play an important role in the delivery of light-mediated rhythmic signals to other parts of the brain.

Catecholaminergic innervation of the suprachiasmatic nucleus in the adult rat: ultrastructural relationships with neurons containing vasoactive intestinal peptide or vasopressin

Catecholaminergic fibers in the suprachiasmatic nucleus of adult rats were investigated by use of light- and electron-microscopic immunocytochemistry. The suprachiasmatic nucleus receives a modest density of tyrosine hydroxylase-containing axons, homogeneously distributed in the nucleus and forming varicosities throughout its entire rostro-caudal extension. Immunolabeling with antibodies against dopamine showed that this catecholamine input comprises a dopaminergic component. Many tyrosine hydroxylase-positive cells were localized at the immediate periphery of the suprachiasmatic nucleus. With electron-microscopic examination, dendrites of these neurons were found within the limits of the nucleus as well as at a border zone between the suprachiasmatic nucleus proper and the optic tract where they received unlabeled synapses, providing a morphological support for a possible role of dopaminergic neurons in the integration and/or transfer of light-related signals. More than 91% of catecholaminergic axonal varicosities were found to establish morphologically defined synapses with dendrites. To investigate whether these synapses might be shared with neurons of one or both of the two main peptidergic populations of the nucleus, namely vasoactive intestinal peptide- and vasopressin-containing neurons, we carried out doublelabeling experments combining immunoperoxidase and immunogold-silver labeling. Results showed only a few cases of direct association of the catecholaminergic terminals with these peptidergic categories. In both types of dually stained sections, catecholaminergic synapses were preferentially made with unlabeled dendrites. The homogeneous distribution of tyrosine hydroxylase-immunoreactive fibers in the suprachiasmatic nucleus could therefore reflect a lack of significant catecholaminergic innervation of both vasoactive intestinal peptide- and vasopressin-synthesizing neurons.

Long-term variations of AP-1 composition after CRH stimulation: consequence on POMC gene regulation

It has been shown previously that the CRH-induced POMC gene transcription in the corticotroph cell line AtT-20 involves an increase in AP-1 DNA binding activity that remained elevated for at least 24 h, while induction of c-fos was transient. We showed here that there were dramatic changes in protein components of AP-1 including an initial recruitment of the transcriptional activators c-Fos and Jun-B then of Fra-2 and Jun-D. Changes in AP-1 composition were concomitant with a decrease in POMC mRNA. Moreover, the presence of Fra-2/Jun-D dimers suppressed the CRH-induction of c-fos mRNA expression as well as c-Fos/Jun-B recruitment in AP-1 complexes, suggesting the existence of autoregulatory loops of AP-1 composition that involve complex interactions between the different members of the Jun and Fos families. It is concluded that CRH stimulation of corticotroph cells involves successive recruitment of activators and repressors, possibly contributing to prevent over expression of POMC.

Synaptic connectivity of serotonin graft efferents in the suprachiasmatic and supraoptic nuclei of the hypothalamus.

We have previously reported that a cell suspension from the rostral part of the embryonic raphe grafted to the basal hypothalamus of 5,7-dihydroxytryptamine-denervated rats produced incomplete serotonin (5-HT) re-innervation of the suprachiasmatic nucleus (SCN) as opposed to hyper-innervation of the supraoptic nucleus (SON). We took advantage of this experimental model to investigate whether the graft-derived, 5-HT fibres retained normal ultrastructural features, and, particularly, a normal density of synaptic junctions, irrespective of the extent of target re-innervation. The intrinsic features of immunostained, graft-derived 5-HT axonal varicosities in both the SCN (ventral portion) and the SON were essentially similar to those exhibited by the respective endogenous innervation. Analysis of well-preserved varicosities in uninterrupted series of thin sections allowed us to evaluate directly the proportions of junctional to non-junctional 5-HT varicosities in both regions. Synaptic incidences were also remarkably conserved after grafting (45.5% in the SCN versus 38.5% in the SON; 48% and 38% in normal rats, respectively). Synapses were primarily reestablished on dendritic shafts, which also were identified as the major post-synaptic targets of the normal 5-HT innervations. We noted, however, a tendency toward increased numbers of symmetrical versus asymmetrical synapses in both the SCN and SON of grafted rats. Thus, irrespective of whether hypo-or hyper-innervation patterns developed post-grafting, the transplanted 5-HT neurons essentially retained normal ultrastructural features in their target territories, with a normal incidence of synaptic junctions. The data provide further support to the hypothesis that the innervation territory is the major determinant of the frequency with which ingrowing 5-HT fibres make synaptic junctions.

Post-lesion up-regulation of 5-HT1B binding sites in the suprachiasmatic nucleus may be reversed after spontaneous or graft-induced serotonin reinnervation.

We have previously reported that selective axotomy of serotoninergic neurons produced by an intraventricular injection of 5, 7-dihydroxytryptamine is followed by an increase in 5-HT1B binding sites in the suprachiasmatic nucleus of the hypothalamus. This post-lesion up-regulation is shown here to be spontaneously reversed after long-term survival in spite of an incomplete reinnervation of the nucleus. Recovery may be accelerated by fetal raphe transplants that produce more rapid reinnervation.

Intrinsic organization and monoaminergic innervation of the suprachiasmatic nucleus transplanted to adult rats. A light- and electron-microscopic study.

SummaryLight- and electron-microscopic immunocytochemistry was used to investigate grafts of foetal hypothalamic tissue implanted close to the site of the suprachiasmatic nucleus in adult rats with bilateral surgical ablation of this nucleus. The transplants contained vasoactive intestinal peptide and vasopressin cell clusters, which have previously been shown to characterize functional suprachiasmatic nucleus grafts. Vasoactive intestinal peptide and vasopressin neurons presented synaptic features that have not been described in the native suprachiasmatic nucleus. More specifically, their terminals within the graft were involved in ‘double’ synapses with separate unlabelled dendrites. Moreover, in dually stained sections, an unexpected synaptic investment of vasoactive intestinal peptide neurons by vasopressin endings was detected, which revealed reversed vasoactive intestinal peptide/vasopressin interactions compared to those described in the native nucleus. These observations could reflect some immature features of the grafted neurons. Ultrastructural relationships of monoaminergic fibres arising from host and/or intragraft neurons were also examined. Within the engrafted suprachiasmatic nucleus, tyrosine hydroxylase-labelled fibres, which probably belonged to cografted dopaminergic neurons, showed normal patterns of distribution and synaptic connections, with no preferential relationships with vasoactive intestinal peptide or vasopressin neurons. Serotoninergic axons arborized within transplants but, in agreement with previous data showing an inhibitory influence of the suprachiasmatic nucleus on ingrowing serotoninergic fibres, they had no predilection for the area corresponding to that nucleus. In spite of their relative scarcity, serotoninergic fibres within the engrafted suprachiasmatic nucleus showed an almost normal synaptic incidence, but synapses were not predominantly shared with the vasoactive intestinal peptide neurons, known to be their major targets in the native nucleus. This may contribute not only to the failure of functional grafts to synchronize with environmental conditions, but also to the inability of transplants to restore hormonal rhythms such as estrous cyclicity.

Is light-regulated AP-1 binding in the rat suprachiasmatic nucleus gated by the circadian clock?

In mammals, photic entrainment of circadian rhythms likely involves light- and clock-dependent expression of immediate early genes, including fos-like and jun-like genes, in the rat suprachiasmatic nucleus. Using an electrophoretic mobility shift assay, we evaluated whether the photic regulation of DNA-binding activity and composition of activating protein-1 (AP-1) complexes in the suprachiasmatic nucleus is also dependent on circadian phase. Phase-dependent light inducibility in the expression of fra-2 and c-fos genes and in immunoreactive Fra-2 and c-Fos protein expression was also evaluated, by in situ hybridization and immunocytochemistry. Lights effects on AP-1 DNA-binding differed both qualitatively and quantitatively according to the circadian phase at which light was applied. This phase dependence accounted for by both compartmentalization of proteins involved in constitutive AP-1 complexes within the nucleus or cytoplasm and control of the extent to which the expression of specific complexes was induced. It was then shown that the mechanisms by which the circadian clock gates the photic induction of AP-1 components differed according to the nature of the protein.