Search

The involvement of melatonin in mammalian brain pathophysiology has received growing interest, but information about the anatomical distribution of its two G-protein-coupled receptors, MT1 and MT2 , remains elusive. In this study, using specific antibodies, we examined the precise distribution of both melatonin receptors immunoreactivity across the adult rat brain using light, confocal, and electron microscopy. Our results demonstrate a selective MT1 and MT2 localization on neuronal cell bodies and dendrites in numerous regions of the rat telencephalon, diencephalon, and mesencephalon. Confocal and ultrastructural examination confirmed the somatodendritic nature of MT1 and MT2 receptors, both being localized on neuronal membranes. Overall, striking differences were observed in the anatomical distribution pattern of MT1 and MT2 proteins, and the labeling often appeared complementary in regions displaying both receptors. Somadendrites labeled for MT1 were observed for instance in the retrosplenial cortex, the dentate gyrus of the hippocampus, the islands of Calleja, the medial habenula, the suprachiasmatic nucleus, the superior colliculus, the substantia nigra pars compacta, the dorsal raphe nucleus, and the pars tuberalis of the pituitary gland. Somadendrites endowed with MT2 receptors were mostly observed in the CA3 field of the hippocampus, the reticular thalamic nucleus, the supraoptic nucleus, the inferior colliculus, the substantia nigra pars reticulata, and the ventrolateral periaqueductal gray. Together, these data provide the first detailed neurocytological mapping of melatonin receptors in the adult rat brain, an essential prerequisite for a better understanding of melatonin distinct receptor function and neurophysiology., (© 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd.)

Purpose: Earlier studies have shown that positron emission tomography (PET) imaging with the radioligand [(18)F]MPPF allows for measuring the binding potential of serotonin 5-hydroxytryptamine(1A) (5-HT(1A)) receptors in different regions of animal and human brain, including that of 5-HT(1A) autoreceptors in the raphe nuclei. In the present study, we sought to determine if such data could be obtained in rat, with a microPET (R4, Concorde Microsystems)., Methods: Scans from isoflurane-anaesthetised rats (n = 18, including six test-retest) were co-registered with magnetic resonance imaging data, and binding potential, blood to plasma ratio and radiotracer efflux were estimated according to a simplified reference tissue model., Results: Values of binding potential for hippocampus (1.2), entorhinal cortex (1.1), septum (1.1), medial prefrontal cortex (1.0), amygdala (0.8), raphe nuclei (0.6), paraventricular hypothalamic nucleus (0.5) and raphe obscurus (0.5) were comparable to those previously measured with PET in cats, non-human primates or humans. Test-retest variability was in the order of 10% in the larger brain regions (hippocampus, medial prefrontal and entorhinal cortex) and less than 20% in small nuclei such as the septum and the paraventricular hypothalamic, basolateral amygdaloid and raphe nuclei., Conclusions: MicroPET brain imaging of 5-HT(1A) receptors with [(18)F]MPPF thus represents a promising avenue for investigating 5-HT(1A) receptor function in rat.

The serotonin1B receptor knockout (5-HT1B KO) mouse is a valuable animal model of addiction to psychostimulants. We previously found selective increases in dopamine (DA) turnover in the nucleus accumbens of these mice, in addition to several changes in their central serotonin system. Here, we searched for further DA adaptations by measuring D1 and D2 receptor as well DA plasma membrane transporter (DAT) sites by ligand binding autoradiography, and G-protein coupling to D1 and D2 receptors by [35S]GTP gamma S autoradiography. Except for a slight increase in the lateral septum, D1 receptor binding did not differ from wild-type in twenty-one other neocortical, limbic or basal ganglia regions examined in the KO. Nor were there changes in D1 agonist-stimulated G-protein coupling in any of these regions, including the lateral septum. Increases in D2 binding sites, presumably involving GABAergic projection neurons, were measured in the nucleus accumbens, olfactory tubercle and ventral tegmental area of the 5-HT1B KO. However, no activation of the efficacy of D2 receptor coupling to G-protein could be measured in these and other brain regions. Binding to DAT was unchanged throughout brain. Because of their implication in cocaine addiction, the functionality of mu-opioid and GABAB receptors was also assessed by [35S]GTP gamma S autoradiography. 5-HT1B KO showed selective decreases in G-protein coupling to mu-opioid receptors in the paraventricular thalamic nucleus, and to GABAB receptors in the basolateral nucleus of amygdala. It is likely that these latter changes underlie some aspects of the addictive behavior of the 5-HT1B KO mouse.

Immuno-electron microscopic and beta-microprobe studies have demonstrated that the internalization of serotonin 5-HT(1A) autoreceptors, after acute treatment with the selective 5-HT(1A) receptor agonist 8-OH-DPAT or with the specific serotonin reuptake inhibitor (SSRI) fluoxetine, is associated with a marked decrease in the in vivo binding of [(18)F]MPPF in the nucleus raphe dorsalis (NRD) of rat. To determine whether this event might be amenable to brain imaging, the present [(18)F]MPPF positron emission tomographic (PET) study was carried out in anesthetized cats given or not a single dose (5 mg/kg, i.v.) or chronically treated with fluoxetine (5 mg/kg, s.c. for 21 days). Compared to control, [(18)F]MPPF binding potential was considerably (and visibly) decreased in the cat NRD after acute fluoxetine treatment, while it remained unchanged in other brain regions. Unexpectedly, after chronic fluoxetine treatment, [(18)F]MPPF binding potential was not affected in any brain region. In parallel immuno-electron microscopic experiments carried out in rat, the density of 5-HT(1A) autoreceptors on the plasma membrane of NRD dendrites was comparable to control after chronic fluoxetine treatment. If the decrease in [(18)F]MPPF binding at the onset of SSRI treatment was detectable by PET imaging, it could potentially serve as a biological index of efficacy.

Enhancing cerebral serotonin (5-hydroxytryptamine, 5-HT) neurotransmission is a common property of antidepressant treatments and the basis for their efficacy. 5-HT1A receptors located on the cell body and dendrites of 5-HT neurons (autoreceptors) play a key role in this regard. Because they normally mediate an inhibition of neuronal firing, their desensitization is a prerequisite to the delayed enhancement of 5-HT neurotransmission upon treatment with monoamine oxidase (MAOI) inhibitors or specific serotonin reuptake inhibitors (SSRI). Using beta-sensitive microprobes in vivo, we measured a significant decrease (-30%) in binding sites for the 5-HT1A PET radioligand [18F]MPPF associated with an equivalent reduction (-34%) in the cell surface density of 5-HT1A receptor immunoreactivity (internalization), in the nucleus raphe dorsalis (autoreceptors), but not hippocampus (heteroreceptors), of rats given a single dose of the specific 5-HT1A receptor agonist, 8-OH-DPAT (0.5 mg/kg, iv). This effect was completely blocked by pretreatment with the selective 5-HT1A antagonist WAY 100635. Having ruled out that this decreased density of [18F]MPPF binding in the nucleus raphe dorsalis of 8-OH-DPAT-treated rats resulted from a local blood flow effect, we obtained autoradiographic evidence indicating that the total amount of specific binding of [18F]MPPF in tissue sections was unaffected by the 8-OH-DPAT treatment in either NRD or hippocampus. It was therefore concluded that the internalization of 5-HT1A autoreceptors accounted for the decreased binding in vivo of [18F]MPPF in the nucleus raphe dorsalis of rats treated with 8-OH-DPAT. Thus, PET imaging might provide a mean to measure 5-HT1A receptor internalization in human brain and thus assess responsiveness to antidepressant treatment., (Copyright 2004 Elsevier Inc.)

The firing of central serotonin (5-hydroxytryptamine, 5-HT) neurons and their capacity to release 5-HT are subjected to a receptor-mediated auto-control via 5-HT(1A) and 5-HT(1B) receptors respectively located on the somata/dendrites (5-HT(1A) autoreceptors) and preterminal axon arborizations (5-HT(1B) autoreceptors) of these neurons. To further characterize mutual adaptations of these two receptor subtypes in the absence of one of them, activation of G-protein coupling by agonist was measured and compared to wild-type (WT) in 5-HT(1A) and 5-HT(1B) homozygous knockout (KO) mice. As expected, in WT, the non-selective 5-HT(1A/1B) receptor agonist 5-carboxyamidotryptamine (5-CT) stimulated guanosine 5'-O-(gamma-[(35)S]thio)triphosphate ([(35)S]GTP(gamma)S) incorporation in many brain regions endowed with one and/or the other receptor. In the respective KOs, no stimulation was measured in regions known to express only or mainly the deleted receptor. In the 5-HT(1A) KOs, the amplitude of G-protein activation in regions endowed with 5-HT(1B) receptors was unchanged by comparison to WT. In the 5-HT(1B) KOs, the magnitude of the 5-CT stimulation was the same as WT in all regions containing 5-HT(1A) receptors, except in the amygdala, where it was significantly lower, even if this region was one of the most strongly activated in the WT. A similar result was obtained in the amygdala of 5-HT(1B) KOs after activation by the selective 5-HT(1A) receptor agonist R-(+)8-hydroxy-2-(di-n-propylamino) tetralin (8-OH-DPAT). Under these conditions, however, there was in addition a significant lowering of the stimulated (but not basal) [(35)S]GTP(gamma)S incorporation by comparison to WT in all regions endowed with 5-HT(1A) receptors, including the dorsal raphe nucleus. Thus, eventhough agonist radioligand binding to either 5-HT(1A) or 5-HT(1B) receptors is unchanged in the reciprocal KOs, it appears that a compensatory decrease in the efficiency of G-protein coupling to 5-HT(1A) receptors has developed in the 5-HT(1B) mutant. This could represent the first indication of a cross-talk between these two 5-HT receptor subtypes, at least in brain regions where they are co localized in the same neurons.

5-HT1A knockout (KO) mice display an anxious-like phenotype, whereas 5-HT1B KOs are over-aggressive. To identify serotoninergic correlates of these altered behaviors, autoradiographic measurements of 5-HT1A and 5-HT1B serotonin (5-HT) receptors and transporter (5-HTT) were obtained using the radioligands [3H]8-OH-DPAT, [125I]cyanopindolol and [3H]citalopram, respectively. By comparison to wild-type, density of 5-HT1B receptors was unchanged throughout brain in 5-HT1A KOs, and that of 5-HT1A receptors in 5-HT1B KOs. In contrast, decreases in density of 5-HTT binding were measured in several brain regions of both genotypes. Moreover, 5-HTT binding density was significantly increased in the amygdalo-hippocampal nucleus and ventral hippocampus of the 5-HT1B KOs. Measurements of 5-HT axon length and number of axon varicosities by quantitative 5-HT immunocytochemistry revealed proportional increases in the density of 5-HT innervation in these two regions of 5-HT1B KOs, whereas none of the decreases in 5-HTT binding sites were associated with any such changes. Several conclusions could be drawn from these results: (i) 5-HT1B receptors do not adapt in 5-HT1A KOs, nor do 5-HT1A receptors in 5-HT1B KOs. (ii) 5-HTT is down-regulated in several brain regions of 5-HT1A and 5-HT1B KO mice. (iii) This down-regulation could contribute to the anxious-like phenotype of the 5-HT1A KOs, by reducing 5-HT clearance in several territories of 5-HT innervation. (iv) The 5-HT hyperinnervation in the amygdalo-hippocampal nucleus and ventral hippocampus of 5-HT1B KOs could play a role in their increased aggressiveness, and might also explain their better performance in some cognitive tests. (v) These increases in density of 5-HT innervation provide the first evidence for a negative control of 5-HT neuron growth mediated by 5-HT1B receptors.

The 5-HT1A and 5-HT1B receptors of serotonin play important roles as auto- and heteroreceptors controlling the release of serotonin itself and of other neurotransmitters/modulators in the central nervous system (CNS). To determine the precise localization of these receptors, we examined their respective cellular and subcellular distributions in the nucleus raphe dorsalis and hippocampal formation (5-HT1A) and in the globus pallidus and substantia nigra (5-HT1B), using light and electron microscopic immunocytochemistry with specific antibodies. Both immunogold and immunoperoxidase preembedding labelings were achieved. In the nucleus raphe dorsalis, 5-HT1A immunoreactivity was found exclusively on neuronal cell bodies and dendrites, and mostly along extrasynaptic portions of their plasma membrane. After immunogold labeling, the density of membrane-associated 5-HT1A receptors could be estimated to be at least 30-40 times that in the cytoplasm. In the hippocampal formation, the somata as well as dendrites of pyramidal and granule cells displayed 5-HT1A immunoreactivity, which was also prominent on the dendritic spines of pyramidal cells. In both substantia nigra and globus pallidus, 5-HT1B receptors were preferentially associated with the membrane of fine, unmyelinated, preterminal axons, and were not found on axon terminals. A selective localization to the cytoplasm of endothelial cells of microvessels was also observed. Because the 5-HT1A receptors are somatodendritic, they are ideally situated to mediate serotonin effects on neuronal firing, both as auto- and as heteroreceptors. The localization of 5-HT1B receptors to the membrane of preterminal axons suggests that they control transmitter release from nonserotonin as well as serotonin neurons by mediating serotonin effects on axonal conduction. The fact that these two receptor subtypes predominate at extrasynaptic and nonsynaptic sites provides further evidence for diffuse serotonin transmission in the CNS.

Recent ultrastructural data demonstrate the largely asynaptic character of the cholinergic innervation in many regions of adult rat brain. These data favour the hypothesis of a diffuse transmission/modulation by acetylcholine in the CNS and, by way of consequence, that of a persistent, low level of acetylcholine in the extracellular space.

Recent immunoelectron microscopic studies have revealed a low frequency of synaptic membrane differentiations on ACh (ChAT-immunostained) axon terminals (boutons or varicosities) in adult rat cerebral cortex, hippocampus and neostriatum, suggesting that, besides synaptic transmission, diffuse transmission by ACh prevails in many regions of the CNS. Cytological analysis of the immediate micro-environment of these ACh terminals, as well as currently available immunocytochemical data on the cellular and subcellular distribution of ACh receptors, is congruent with this view. At least in brain regions densely innervated by ACh neurons, a further aspect of the diffuse transmission paradigm is envisaged: the existence of an ambient level of ACh in the extracellular space, to which all tissue elements would be permanently exposed. Recent experimental data on the various molecular forms of AChE and their presumptive role at the neuromuscular junction support this hypothesis. As in the peripheral nervous system, degradation of ACh by the prevalent G4 form of AChE in the CNS would primarily serve to keep the extrasynaptic, ambient level of ACh within physiological limits, rather than totally eliminate ACh from synaptic clefts. Long-lasting and widespread electrophysiological effects imputable to ACh in the CNS might be explained in this manner. The notions of diffuse transmission and of an ambient level of ACh in the CNS could also be of clinical relevance, in accounting for the production and nature of certain cholinergic deficits and the efficacy of substitution therapies.

WIN 35428 and GBR 12935, two uptake blocker ligands of the membrane transporter for dopamine (DA), were evaluated as quantitative markers of DA innervation density in CNS tissue. From alternate rat brain slices respectively processed for either light microscope or film autoradiography, counts of DA axon terminals (varicosities) labeled by uptake/storage of [3H]DA were matched with densitometric measurements of the specific binding of [3H]WIN 35428 and [3H]GBR 12935 in the same anatomical areas. The relation between the two parameters was examined in 1) the normal cingulate cortex; 2) the neostriatum severely DA-denervated by unilateral intramesencephalic injections of 6-hydroxydopamine; and 3) the neostriatum, partly DA-reinnervated by an intrastriatal graft of fetal mesencephalic neurons after prior 6-hydroxydopamine lesion. For technical reasons, the hyperdense DA innervation of normal striatum was not amenable to such correlative testing. Data were subjected to multilevel analysis. Specific [3H]WIN binding at 37 degrees C was tightly and linearly correlated with the number of DA varicosities over the full range of DA innervation densities tested. The regression lines for intact cortex and for DA-denervated as well as DA-reinnervated neostriatum had the same slope and crossed the ordinate near zero. In contrast, [3H]GBR 12935 binding at 37 degrees C showed no correlation with the number of DA varicosities. A linear correlation could be obtained after incubation with [3H]GBR 12935 at 4 degrees C in the presence of ZnSO4, but the intercept of this regression line remained significantly above zero at origin, indicating extraneous binding to non-DA transporter sites. Providing that the hyperdense DA innervation of the normal neostriatum does not generate a particular problem in vivo as it does in vitro. WIN 35428, but not GBR 12935, might satisfy the selectivity and sensitivity requirements of a quantitative marker of DA innervation density for eventual use in positron emission tomographic studies.

Direct counting of axon terminals (varicosities) labeled by uptake/storage of a tritiated monoamine provides a means to test radioligands of the corresponding membrane transporter as quantitative markers of regional monoamine innervation density in brain tissue. In autoradiographs from alternate rat brain slices, counts of [3H]5-HT-labeled axon terminals were matched with densitometric measurements of the specific binding of tritiated cyanoimipramine (CYI), citalopram (CITAL), and 6-nitroquipazine (6-NTQ), under conditions of hypo-, normo-, or hyper-5-HT innervation of the neostriatum. A total of 267 pairs of data were subjected to a multilevel analysis (iterative generalized least square procedure). With all three ligands, there was a linear relationship between the density of 5-HT innervation and the density of specific binding and no change in the slope of the regression lines as a function of 5-HT innervation density. Thus, none of these ligands gave any sign of down- or up-regulation of the 5-HT transporter consequent to 5-HT hypo- or hyper-innervation. The regression lines for CYI and CITAL were not significantly different from one another and crossed the ordinate near zero, whereas the regression line for 6-NTQ was less steep and had a higher intercept with the ordinate. In addition, the dispersion of values around the regression line (residuals) was lower with CYI and CITAL than 6-NTQ. It was concluded that both CYI and CITAL may serve as quantitative markers of 5-HT innervation density, at least in vitro, whereas 6-NTQ demonstrates a certain lack of specificity and sensitivity. Further work will be needed to assess the potential of CYI and CITAL for positron emission tomographic studies of living brain. Such empirical testing should also be applicable for screening radioligands of the dopamine or the noradrenaline transporters.

Adult mice transgenic for the human form of neurofilament light protein display abnormal perikaryal immunoreactivity for this protein in many regions of the CNS and notably the thalamus. To determine the sequence of development of these anomalies, we have compared normal and transgenic mice of different postnatal ages (P0-P70), using immunocytochemistry with primary antibodies recognizing both murine and human sequence of neurofilament light protein (NR-4) or the human form only (DP5-1-12). In normal mouse brainstem, several nuclei displayed immunoreactive perikarya at P0. The number of these perikarya culminated at P10, followed by a general decrease, some nuclei having lost all perikaryal immunostaining in adults. In transgenic mouse brainstem, the distribution of perikaryal immunoreactivity already resembled at P0 that of P10 in normal mouse, and remained unchanged in adults. Differences between normal and transgenic mice were even more pronounced in the forebrain. Some nuclei of normal mouse basal forebrain that were weakly immunopositive at P10 or P20, but no longer in adults, were already labeled at P0 and remained so or became more intense at later stages in transgenic mice. In the thalamus of normal mouse, perikaryal labeling was faint, confined to a few nuclei, and detected only transiently at P10, whereas in transgenics, it was already observed in some nuclei at P0, increased in intensity and extended to other nuclei at P10, and persisted thereafter. Strongly immunoreactive, inflated perikarya with excentric nuclei were prominent in these thalamic nuclei at P20, and even larger in size at P70. In the cerebral cortex of normal mice, layers II-III and layer V of many cytoarchitectonic areas showed immunoreactive cell bodies at P10, a distribution which became gradually restricted to the parietal cortex in adults. In transgenic mice, immunopositive cortical cell bodies were first detected at P3, filled layers II-III of numerous cortical areas at P10, and then rapidly decreased in number to approach the adult pattern at P20. In the cortex as well as thalamus of P10 transgenic mice, differences between the patterns of cellular staining with clones NR4 and DP5-1-12 antibodies indicated that both the murine and human proteins were accumulated in these neurons. Thus, neurofilament light protein accumulation in the transgenic mouse brain generally involved neurons displaying perikaryal immunoreactivity for the protein at least at some point during normal postnatal development.(ABSTRACT TRUNCATED AT 400 WORDS)

A recently developed technique of immunoautoradiography on nitrocellulose transfers of serial frozen sections was used to determine tryptophan hydroxylase concentration in selected areas of the adult rat brain following neonatal 6-hydroxydopamine destruction of nigrostriatal dopamine neurons. Particular attention was paid to the neostriatum, known to be serotonin-hyperinnervated under these conditions, and to the nucleus raphe dorsalis, containing the cell bodies of origin for these nerve terminals. The hippocampus was also investigated as a territory of structurally intact serotonin innervation arising primarily from the nucleus raphe medianus. Tryptophan hydroxylase protein was measured at successive transverse levels across the entire caudorostral extent of all these regions. Similar measurements of tyrosine hydroxylase protein across the substantia nigra and the neostriatum verified the disappearance of the nigrostriatal dopamine neurons. The average tryptophan hydroxylase tissue concentration in the dorsal third of the serotonin-hyperinnervated neostriatum was up by 36% above control, i.e. significantly less than the number of its serotonin axon terminals or varicosities. This was therefore indicative of a lowering of the tryptophan hydroxylase protein content per serotonin ending. Interestingly, a tight correlation between the respective level-by-level concentrations of tryptophan hydroxylase and tyrosine hydroxylase protein in the control neostriatum allowed the prediction the tryptophan hydroxylase concentration after dopamine denervation with a serotonin hyperinnervation. Tryptophan hydroxylase concentration was also significantly reduced in both the nucleus raphe dorsalis and nucleus raphe medianus, notably at those raphe dorsalis levels known to give rise to the serotonin hyperinnervation of neostriatum. It is hypothesized that the lower steady-state level of tryptophan hydroxylase inside the terminals and cell bodies of hyperinnervating serotonin neurons was the result of a feedback inhibition of the synthesis of the enzyme by its end-product, presumably because of the increased amount of serotonin in these terminals.

Light microscopic immunocytochemistry with monoclonal antibodies recognizing both murine and human light neurofilament proteins (mNF-L and hNF-L) or hNF-L only was used to examine the distribution of NF-L in the CNS of adult mice, normal or transgenic for the human gene. In normal mice, major fiber bundles were immunoreactive to the first antibody, with few exceptions such as the internal capsule, anterior commissure, and corpus callosum. Strong immunoreactivity was also present in the perikarya of motoneurons in the spinal cord and brainstem, as well as in other brainstem nuclei. Faint cell body staining was visible in layers II, III, and V of the parietal cortex and layers V and VI of the retrosplenial cingulate cortex. In transgenic mice, all forebrain as well as brainstem fiber tracts were intensely immunoreactive to both antibodies. Cell body staining was more intense than in normal mouse and involved additional forebrain and brainstem regions, including extended areas of cerebral cortex. Abnormal cell body labeling was particularly striking in several thalamic nuclei, where numerous darkly stained perikarya were considerably enlarged by accumulated immunoreactive material and exhibited eccentric and fragmented nuclei. At the electron microscopic level, these perikarya were filled with disarrayed filaments displacing all other organelles against the cytoplasmic membrane. Such aberrant accumulation of NF-L was presumably the result of an overexpression in selective subpopulations of CNS neurons. It was compatible with prolonged survival of the animal and could provide a new experimental model of neurodegenerative disease.

Neonatal destruction of nigrostriatal dopamine neurons by cerebroventricular injection of 6-hydroxydopamine (6-OHDA) results in a serotonin (5-HT) hyperinnervation of the rostral neostriatum in adult rat. Quantitative ligand-binding autoradiography was used to compare the density of various 5-HT receptor subtypes in the adult brain of control and neonatally 6-OHDA-lesioned rats. 5-HT1A, 5-HT1B, 5HT1nonAB and 5-HT2 sites were labeled with [3H]8-OH-DPAT, [125I]cyanopindolol, [3H]5-HT and [125I]DOI, respectively, and measured in the rostral and caudal halves of neostriatum and selected forebrain or midbrain regions. 5-HT1A binding, measured after 6 months, was unchanged in all regions examined including the dorsal raphe nucleus. Three months after the lesion, 5-HT1B binding was increased throughout the neostriatum (30%), but also in the substantia nigra (50%) and globus pallidus (30%), suggesting an up-regulation and an increased axonal transport of these receptors in neostriatal projection neurons. 5-HT1nonAB binding was also increased throughout the neostriatum (40%) and in the substantia nigra (50%), but unchanged in the globus pallidus, as if this up-regulation preferentially involved striatonigral as opposed to striatopallidal neurons. 5-HT2 binding showed an even greater increase (60%), which was restricted to the rostral half of neostriatum and also seemed imputable to an up-regulation as heteroreceptors. Even though the exact cause(s) of these receptor increases could not be determined, their anatomical distribution suggested that they were somehow related to the initial dopamine denervation in the case of the 5-HT1B and 5-HT1nonAB receptors, and more tightly linked to the 5-HT hyperinnervation in the case of the 5-HT2 receptors. Such receptor changes could participate in adaptive mechanisms implicating other transmitters and behavioral disturbances observed in this particular experimental model. Interestingly, they could also account for an enhancement of neostriatal 5-HT function even in a condition where extracellular levels of 5-HT apparently remain normal because of increased uptake.

The high-affinity binding of [3H]paroxetine to membranes was measured in different regions of the rat and rabbit brain: cingulate, frontal, parietal, piriform, entorhinal, and visual cortical areas; dorsal and ventral hippocampus; rostral and caudal halves of neostriatum (rat) or caudate nucleus and putamen (rabbit); ventral mesencephalic tegmentum; and midbrain raphe nuclei region. The tissue concentrations of serotonin (5-HT), 5-hydroxyindole-3-acetic acid (5-HIAA) and 5-hydroxy-l-tryptophan (5-HTP) were also determined by high-performance liquid chromatography (HPLC) in the same brain samples. The regional density of [3H]paroxetine binding varied in both species; the highest values (Bmax) were found in the midbrain raphe region and ventral mesencephalic tegmentum. The cortical values ranged from moderate to low, with a significantly higher density in the cingulate cortex of the rat compared with rabbit. In the rat, there was also a higher density in the ventral than dorsal hippocampus, and the caudal than rostral neostriatum. In the rabbit, the hippocampal and neostriatal values were generally lower and more uniform. In both species, there was an excellent correlation between regional 5-HT levels and specific [3H]paroxetine binding (r = 0.87 in the rat and 0.96 in the rabbit). Considering the available quantitative data on the number of 5-HT nerve cell bodies and axon terminals in different regions of the rat brain, it appears likely that the high amount of [3H]paroxetine binding in the midbrain raphe region and ventral mesencephalic tegmentum reflects the presence of 5-HT uptake sites on 5-HT nerve cell bodies and dendrites as well as axon terminals. In other brain regions, the heterogeneous distribution of [3H]paroxetine binding parallels that of the number of 5-HT axon terminals, emphasizing the potential usefulness of this radioligand as a marker of 5-HT innervation density.

A systematic comparison of the effects of iontophoresed dopamine (DA) was carried out in the neostriatum (NS), nucleus accumbens (Acb) and anterior cingulate (ACg), prefrontal (PF) and parietal (Par) cortex of urethane-anesthetized rats, before and after treatment with the specific DA uptake blockers GBR 12909 and Bupropion. Similar experiments were also conducted after DA denervation with 6-hydroxydopamine and after DA depletion with alpha-methyl-p-tyrosine. The average rate of spontaneous neuronal firing was comparable in all regions, except in the NS after DA depletion. A majority of the units were inhibited by DA in every region and condition tested. As assessed with the IT50 index, the responsiveness to DA was not markedly different between regions, indicating that the postsynaptic sensitivity to this amine is independent of the density of DA receptors and of DA innervation. In contrast, the average duration of DA inhibitions (RT90) was considerably longer (5-fold) in the intact ACg than in the PF, Par, NS, or Acb. Moreover, treatment with both DA uptake blockers reduced the duration of DA inhibitions in ACg (4- to 9-fold); while lengthening it in PF, NS and Acb; and having no apparent effect in Par. DA depletion and DA denervation also reduced the duration of the DA inhibitions in ACg without effect in Par. Taken together, these results provide further evidence for the existence of a presynaptic, positive-feedback mechanism in ACg, triggered by DA, and favouring the further release of this transmitter upon its reuptake in DA nerve terminals.

Paraventricular and paracisternal regions of adult rat central nervous system were investigated by light- and electron-microscope radioautography after intraventricular administration of tritiated adrenaline. In tissue primarily fixed by glutaraldehyde perfusion and post-fixed by immersion in osmium tetroxide, there were no aggregates of silver grains indicative of intraneuronal accumulation of the tracer, except over perivascular nerve terminals at the base of the brain. In contrast, when both fixation and postfixation were carried out by rapid vascular perfusion, preferentially labeled nerve cell bodies and axonal varicosities (i.e. terminals) were detected in various anatomical areas known to contain dopaminergic and/or noradrenergic neurons. Serotoninergic axonal varicosities in the supraependymal plexus and subcommissural organ, as well as a small group of nerve cell bodies of undetermined chemical identity in the n. paraventricularis thalami were also found to be labeled. Addition of a ten-fold higher concentration of non-radioactive serotonin to the solution of [3H]adrenaline suppressed the reactivity in the subcommissural organ and the supraependymal plexus but had no such effect elsewhere in brain. Lesioning of the nigrostriatal dopaminergic system with 6-hydroxydopamine prior to [3H]adrenaline injection eradicated axon terminal labeling in the ipsilateral neostriatum. Electron-microscopic examination of [3H]adrenaline-labeled varicosities in the neostriatum, lateral septum, arcuate nucleus and median eminence extended earlier observations on the ultrastructure of the catecholaminergic innervation of these regions. It was concluded that both dopaminergic and noradrenergic neurons as well as certain serotonin-containing axon terminals can take up and retain [3H]adrenaline, although they probably have lesser affinity for this amine than for their own transmitter. Due to the fact that presumptive adrenergic neurons are intermingled with dopaminergic and noradrenergic elements, further work will be needed to determine to which extent they also contributed to [3H]adrenaline uptake in the present experimental conditions.

1. Most of the current knowledge concerning the fine structure of serotonin (5-HT) neurons in mammalian CNS is derived from radioautographic studies following uptake of tritiated 5-HT in vivo. 2. (3H)5-HT-labelled perikaria and proximal dendrites have thus far been visualized ultrastructurally in nuclei raphe dorsalis, paragigantocellularis lateralis and dorsomedialis hypothalami. In none of these regions were they found to possess cytological features which would constitute a hallmark of their serotoninergic nature. 3. Most central 5-HT neurons give rise to thin and unmyelinated, albeit long and ramified axons. These innervate the entire neuraxis, issuing billions of varicose terminals along their paths. Such varicosities usually consist of small axonal enlargements (0.5-1 micrometers in diameter) containing clustered vesicular organelles. These organelles include small (15-50 nm), pleomorphic agranular vesicles and larger (50-130 nm), round or flattened dense-cored vesicles. 4. A salient feature of central 5-HT varicosities is that they often lack the junctional complexes classically considered as the morphological substrate for chemical transmission in CNS. The occurrence of synaptic specializations varies from one brain region to another. When present, synaptic contacts are generally found on dendritic branches or spines of nonserotoninergic neurons, but axo-somatic or axo-axonic synapses, as well as contacts with 5-HT elements, have also been reported. The frequency with which 5-HT varicosities establish junctional relationships may be more dependent on the territory of innervation than characteristic of any subset of 5-HT fibers or nerve cell bodies. 5. It is currently assumed that both non-junctional and junctional 5-HT varicosities can release their endogenous amine upon axonal depolarization. This mode of action, which evokes that of a local neurohormone, might subtend the modulatory effects of 5-HT in the CNS.

Catecholaminergic neurons, which take up and retain exogenous norepinephrine labeled with tritium, were studied by means of high resolution radioautography, in the substantia nigra, the substantia grisea periventricularis, and the locus coeruleus of the rat. Under the conditions required for the radioautographic detection of exogenous norepinephrine-(3)H, it was established that (1) glutaraldehyde was the most suitable fixative for preserving the labeled amine in situ; (2) norepinephrine-(3)H itself, rather than metabolites, accounted for most of the reactions detected in catecholaminergic neurons. At various time intervals after an intraventricular injection of norepinephrine-(3)H, the tracer reached a concentration 15-100 times higher, and disappeared at a slower rate, in presynaptic axons (t(1/2):4 hr) than in nerve cell bodies (t(1/2):0.8-1.3 hr). After pretreatment with a monoamine oxidase inhibitor, the radioautographic reactions increased and persisted longer, especially in the preterminal axons. Within neurons, the labeled amine was ubiquitously distributed in the nerve cell body and concentrated in presynaptic axons and synaptic terminals of various morphological types. Although large granular vesicles were usually present in the labeled axonal bulbs, no structural characteristic could be specifically ascribed to catecholaminergic neurons. It is suggested that exogenous norepinephrine bound to macromolecular complexes is present in all parts of catecholaminergic neurons and mainly concentrated within presynaptic axons.