Child abuse associates with an imbalance of oligodendrocyte-lineage cells in ventromedial prefrontal white matter.

Child abuse (CA) is a major risk factor for depression, and strongly associates with suicidal behavior during adulthood. Neuroimaging studies have reported widespread changes in white matter integrity and brain connectivity in subjects with a history of CA. Although such observations could reflect changes in myelin and oligodendrocyte function, their cellular underpinnings have never been addressed. Using postmortem brain samples from depressed suicides with or without history of CA and matched controls (18 per group), we aimed to characterize the effects of CA on oligodendrocyte-lineage (OL) cells in the ventromedial prefrontal white matter. Using immunoblotting, double-labeling immunofluorescence and stereological estimates of stage-specific markers, we found that CA is associated with increased numbers of mature myelinating oligodendrocytes, accompanied by decreased numbers of more immature OL cells. This was paralleled by an increased expression of transcription factor MASH1, which is involved in the terminal differentiation of the OL, suggesting that CA may trigger an increased maturation, or bias the populations of OL cells toward a more mature phenotype. Some of these effects, which were absent in the brain of depressed suicides with no history of CA, were also found to recover with age, suggesting that changes in the balance of the OL may reflect a transient adaptive mechanism triggered by early-life adversity. In conclusion, our results indicate that CA in depressed suicides is associated with an imbalance of the OL in the ventromedial prefrontal white matter, an effect that could lead to myelin remodeling and long-term connectivity changes within the limbic network.

Disrupted hippocampal neuregulin-1/ErbB3 signaling and dentate gyrus granule cell alterations in suicide.

This corrects the article DOI: 10.1038/tp.2017.132.

Neuropathology of suicide: recent findings and future directions.

Suicide is a major public health concern and a leading cause of death in most societies. Suicidal behaviour is complex and heterogeneous, likely resulting from several causes. It associates with multiple factors, including psychopathology, personality traits, early-life adversity and stressful life events, among others. Over the past decades, studies in fields ranging from neuroanatomy, genetics and molecular psychiatry have led to a model whereby behavioural dysregulation, including suicidal behaviour (SB), develops as a function of biological adaptations in key brain systems. More recently, the unravelling of the unique epigenetic processes that occur in the brain has opened promising avenues in suicide research. The present review explores the various facets of the current knowledge on suicidality and discusses how the rapidly evolving field of neurobehavioural epigenetics may fuel our ability to understand, and potentially prevent, SB.

Disrupted hippocampal neuregulin-1/ErbB3 signaling and dentate gyrus granule cell alterations in suicide.

Neuregulin-1 (NRG1) and ErbB receptors have been associated with psychopathology, and NRG1-ErbB3 signaling has been shown to increase hippocampal neurogenesis and induce antidepressant-like effects. In this study, we aimed to determine whether deficits in NRG1 or ErbBs might be present in the hippocampus of suicide completers. In well-characterized postmortem hippocampal samples from suicides and matched sudden-death controls, we assessed gene expression and methylation using qRT-PCR and EpiTYPER, respectively. Moreover, in hippocampal tissues stained with cresyl violet, stereology was used to quantify numbers of granule cells and of glia. Granule cell body size was examined with a nucleator probe, and granule cell layer volume with a Cavalieri probe. Unmedicated suicides showed sharply decreased hippocampal ErbB3 expression and decreased numbers of ErbB3-expressing granule cell neurons in the anterior dentate gyrus; a phenomenon seemingly reversed by antidepressant treatment. Furthermore, we found ErbB3 expression to be significantly decreased in the dentate gyrus of adult mice exposed to chronic social defeat stress. Taken together, these results reveal novel suicidal endophenotypes in the hippocampus, as well as a putative etiological mechanism underlying suicidality, and suggest that antidepressant or NRG1 treatment may reverse a potential deficit in anterior dentate gyrus granule cell neurons in individuals at risk of dying by suicide.

Neuropathology of mood disorders: do we see the stigmata of inflammation?

A proportion of cases with mood disorders have elevated inflammatory markers in the blood that conceivably may result from stress, infection and/or autoimmunity. However, it is not yet clear whether depression is a neuroinflammatory disease. Multiple histopathological and molecular abnormalities have been found postmortem but the etiology of these abnormalities is unknown. Here, we take an immunological perspective of this literature. Increases in activated microglia or perivascular macrophages in suicide victims have been reported in the parenchyma. In contrast, astrocytic markers generally are downregulated in mood disorders. Impairment of astrocytic function likely compromises the reuptake of glutamate potentially leading to excitotoxicity. Inflammatory cytokines and microglia/macrophage-derived quinolinic acid (QA) downregulate the excitatory amino acid transporters responsible for this reuptake, while QA has the additional effect of inhibiting astroglial glutamine synthetase, which converts glutamate to glutamine. Given that oligodendroglia are particularly vulnerable to inflammation, it is noteworthy that reductions in numbers or density of oligodendrocyte cells are one of the most prominent findings in depression. Structural and/or functional changes to GABAergic interneurons also are salient in postmortem brain samples, and may conceivably be related to early inflammatory insults. Although the postmortem data are consistent with a neuroimmune etiology in a subgroup of depressed individuals, we do not argue that all depression-associated abnormalities are reflective of a neuroinflammatory process or even that all immunological activity in the brain is deleterious. Rather, we highlight the pervasive role of immune signaling pathways in brain function and provide an alternative perspective on the current postmortem literature.

Postmortem evidence of cerebral inflammation in schizophrenia: a systematic review.

Schizophrenia is a psychiatric disorder which has a lifetime prevalence of ~1%. Multiple candidate mechanisms have been proposed in the pathogenesis of schizophrenia. One such mechanism is the involvement of neuroinflammation. Clinical studies, including neuroimaging, peripheral biomarkers and randomized control trials, have suggested the presence of neuroinflammation in schizophrenia. Many studies have also measured markers of neuroinflammation in postmortem brain samples from schizophrenia patients. The objective of this study was to conduct a systematic search of the literature on neuroinflammation in postmortem brains of schizophrenia patients indexed in MEDLINE, Embase and PsycINFO. Databases were searched up until 20th March 2016 for articles published on postmortem brains in schizophrenia evaluating microglia, astrocytes, glia, cytokines, the arachidonic cascade, substance P and other markers of neuroinflammation. Two independent reviewers extracted the data. Out of 5385 articles yielded by the search, 119 articles were identified that measured neuroinflammatory markers in schizophrenic postmortem brains. Glial fibrillary acidic protein expression was elevated, lower or unchanged in 6, 6 and 21 studies, respectively, and similar results were obtained for glial cell densities. On the other hand, microglial markers were increased, lower or unchanged in schizophrenia in 11, 3 and 8 studies, respectively. Results were variable across all other markers, but SERPINA3 and IFITM were consistently increased in 4 and 5 studies, respectively. Despite the variability, some studies evaluating neuroinflammation in postmortem brains in schizophrenia suggest an increase in microglial activity and other markers such as SERPINA3 and IFITM. Variability across studies is partially explained by multiple factors including brain region evaluated, source of the brain, diagnosis, age at time of death, age of onset and the presence of suicide victims in the cohort.

Glial fibrillary acidic protein is differentially expressed across cortical and subcortical regions in healthy brains and downregulated in the thalamus and caudate nucleus of depressed suicides.

There is mounting evidence to suggest aberrant astrocytic function in depression and suicide. Independent studies have reported astrocytic abnormalities in certain brain regions, but it remains unclear whether this is a brain-wide phenomenon. The present study examined this question by measuring glial fibrillary acidic protein (GFAP) expression in postmortem brain samples from suicide completers and matched non-psychiatric controls. Suicide completers were selected based on their recent characterization as low GFAP expressors in the prefrontal cortex, (Brodmann areas 8/9 and 10). Real-time PCR and immunoblotting were used to measure GFAP gene expression and protein levels in BA4 (primary motor cortex), BA17 (primary visual cortex), cerebellar cortex, mediodorsal thalamus and caudate nucleus. We found downregulation of GFAP mRNA and protein in the mediodorsal thalamus and caudate nucleus of depressed suicides compared with controls, whereas GFAP expression in other brain regions was similar between groups. Furthermore, a regional comparison including all samples revealed that GFAP expression in both subcortical regions was, on average, between 11- and 15-fold greater than in cerebellum and neocortex. Examining astrocyte morphology by immunohistochemistry showed that astrocytes in both thalamus and caudate displayed larger cell bodies and extended more ramified processes across larger domains than the previously described cortical astrocytes. This study reveals that astrocytic abnormalities are not brain wide and suggests that they are restricted to cortical and subcortical networks known to be affected in mood disorders. Additionally, our results show a greater diversity in human astrocytic phenotypes than previously thought.

MicroRNA regulation of central glial cell line-derived neurotrophic factor (GDNF) signalling in depression.

Although multiple studies have reported that peripheral glial cell line-derived neurotrophic factor (GDNF) is reduced in depression, cerebral GDNF signalling has yet to be examined in this condition. Here, we report an isoform-specific decrease in GDNF family receptor alpha 1 (GFRA1) mRNA expression, resulting in lowered GFRα1a protein levels in basolateral amygdala (BLA) samples from depressed subjects. Downregulation of GFRα1a was associated with increased expression of microRNAs, including miR-511, predicted to bind to long 3' untranslated region (3'-UTR)-containing transcripts (GFRA1-L) coding for GFRα1a. Transfection of human neural progenitor cells (NPCs) with a miR-511 mimic was sufficient to repress GFRA1-L/GFRα1a without altering GFRα1b, and resulted in pathway-specific changes in immediate early gene activity. Unexpectedly, GFRα1a knockdown did not reduce NPC responses to GDNF. Rather, it greatly enhanced mitogen-activated protein kinase signalling. This effect appeared to be mediated by GDNF/soluble GFRα1/neural cell adhesion molecule binding, and substituting the soluble GFRα1a/GFRα1b content of miR-511-transfected NPCs with that of controls rescued signalling. In light of previous reports suggesting that GFRα1b can inhibit GFRα1a-induced neuroplasticity, we also assessed the association between GFRα1 and doublecortin (DCX; a hyperplastic marker) in human BLA. Although controls displayed coordinated expression of GFRα1a and b isoforms and these correlated positively with DCX, the only significant association observed among depressed subjects was a strongly negative correlation between GFRα1b and DCX. Taken together, these results suggest that microRNA-mediated reductions of GFRα1a in depression change the quality, rather than the quantity, of GDNF signalling. They also suggest that central GDNF signalling may represent a novel target for antidepressant treatment.

Astrocytic abnormalities and global DNA methylation patterns in depression and suicide.

Astrocytes are glial cells specific to the central nervous system and involved in numerous brain functions, including regulation of synaptic transmission and of immune reactions. There is mounting evidence suggesting astrocytic dysfunction in psychopathologies such as major depression, however, little is known about the underlying etiological mechanisms. Here we report a two-stage study investigating genome-wide DNA methylation associated with astrocytic markers in depressive psychopathology. We first characterized prefrontal cortex samples from 121 individuals (76 who died during a depressive episode and 45 healthy controls) for the astrocytic markers GFAP, ALDH1L1, SOX9, GLUL, SCL1A3, GJA1 and GJB6. A subset of 22 cases with consistently downregulated astrocytic markers was then compared with 17 matched controls using methylation binding domain-2 (MBD2) sequencing followed by validation with high-resolution melting and bisulfite Sanger sequencing. With these data, we generated a genome-wide methylation map unique to altered astrocyte-associated depressive psychopathology. The map revealed differentially methylated regions (DMRs) between cases and controls, the majority of which displayed reduced methylation levels in cases. Among intragenic DMRs, those found in GRIK2 (glutamate receptor, ionotropic kainate 2) and BEGAIN (brain-enriched guanylate kinase-associated protein) were most significant and also showed significant correlations with gene expression. Cell-sorted fractions were investigated and demonstrated an important non-neuronal contribution of methylation status in BEGAIN. Functional cell assays revealed promoter and enhancer-like properties in this region that were markedly decreased by methylation. Furthermore, a large number of our DMRs overlapped known Encyclopedia of DNA elements (ENCODE)-identified regulatory elements. Taken together, our data indicate significant differences in the methylation patterns specific to astrocytic dysfunction associated with depressive psychopathology, providing a potential framework for better understanding this disease phenotype.

Impaired structural hippocampal plasticity is associated with emotional and memory deficits in the olfactory bulbectomized rat.

Disturbances in olfactory circuitry have been associated with depression in humans. The olfactory bulbectomized (OBX lesion) has been largely used as a model of depression-like behavior in the rat. However, quantitative neuronal rearrangements in key brain regions in this animal model have not been evaluated yet. Accordingly, we investigated changes in hippocampal plasticity as well as behavioral deficits in this animal model. OBX-induced behavioral deficits were studied in a battery of tests, namely the open field test (OFT), forced swim test (FST), and spatial memory disturbances in the Morris water maze (MWM). To characterize the neuronal remodeling, neuroanatomical rearrangements were investigated in the CA1 hippocampus and piriform cortex (PirC), brain regions receiving inputs from the olfactory bulbs and associated with emotional or olfactory processes. Additionally, cell proliferation and survival of newborn cells in the adult dentate gyrus (DG) of the hippocampus were also determined. OBX induced hyperlocomotion and enhanced rearing and grooming in the OFT, increased immobility in the FST as well as required a longer time to find the hidden platform in the MWM. OBX also induced dendritic atrophy in the hippocampus and PirC. In addition, cell proliferation was decreased while the survival remained unchanged in the DG of these animals. These various features are also observed in depressed subjects, adding further support to the validity and usefulness of this model to evaluate potential novel antidepressants.

Analysis of SNAP25 mRNA expression and promoter DNA methylation in brain areas of Alzheimer's Disease patients.

Alzheimer's Disease (AD) is the most common cause of dementia in elderly people. The presynaptic terminal is an important site of pathological changes in AD, leading to synaptic loss in specific brain regions, such as in the cortex and hippocampus. In this study, we investigated synaptosomal-associated protein, 25-kDa (SNAP25) mRNA levels and promoter DNA methylation in post mortem brain tissues (entorhinal and auditory cortices and hippocampus) from healthy elderly and AD subjects as well as in peripheral blood leukocytes of young, healthy elderly and AD patients. mRNA quantification was performed by quantitative Reverse Transcription Polymerase Chain Reaction (qRT-PCR) using the ΔΔC(T) method and promoter DNA methylation was quantified by mass spectrometry using the Sequenom EpiTYPER platform. We observed a significant decrease in SNAP25 expression in AD across all the three brain regions in relation to the healthy elderly subjects, suggesting impairment in synaptic function. The changes in the auditory cortex reflected those observed in the hippocampus and entorhinal cortex, the primary areas affected in AD. However, no AD-associated differences in SNAP25 promoter DNA methylation were observed suggesting that other mechanisms may be involved in mediating the observed gene expression changes.

Developmental profile of neuregulin receptor ErbB4 in postnatal rat cerebral cortex and hippocampus.

We investigated the cellular and subcellular distributions of neuregulin tyrosine kinase receptor ErbB4 in the postnatal rat frontal cortex and hippocampus by light-, confocal- and electron-microscopic immunocytochemistry. At birth, ErbB4-immunoreactivity (ErbB4-IR) was prominent in the apical cytoplasm and dendrites of cortical plate neurons and hippocampal pyramidal cells. Throughout postnatal development and in adulthood, ErbB4-IR in both regions remained confined to the somatodendritic compartment of neurons, which increased in number to reach the adult pattern by the end of the first postnatal month (P30). At all ages examined, double-labeling experiments revealed that ErbB4-IR always co-localized with the neuronal marker neuronal nuclei (NeuN) and never with glial markers Nestin or glial fibrillary acidic protein (GFAP). Immunoperoxidase labeling at the ultrastructural level confirmed the exclusive localization of ErbB4-IR in somatodendrites, and notably in dendritic spines. Immunogold labeling showed preponderant ErbB4-IR in the cytoplasm, where it was associated with microtubules. Furthermore, ErbB4-IR was abundant in the nucleus of adult cortical and hippocampal neurons, suggesting a role for ErbB4 nuclear signaling in the brain beyond embryonic development. Taken together, these results show that ErbB4 is expressed by neuronal somatodendrites in cerebral cortex and hippocampus from birth to adulthood, and support a role for neuregulins in dendritic growth and plasticity.

Evaluating the suitability of nicotinic acetylcholine receptor antibodies for standard immunodetection procedures.

Nicotinic acetylcholine receptors play important roles in numerous cognitive processes as well as in several debilitating central nervous system (CNS) disorders. In order to fully elucidate the diverse roles of nicotinic acetylcholine receptors in CNS function and dysfunction, a detailed knowledge of their cellular and subcellular localizations is essential. To date, methods to precisely localize nicotinic acetylcholine receptors in the CNS have predominantly relied on the use of anti-receptor subunit antibodies. Although data obtained by immunohistology and immunoblotting are generally in accordance with ligand binding studies, some discrepancies remain, in particular with electrophysiological findings. In this context, nicotinic acetylcholine receptor subunit-deficient mice should be ideal tools for testing the specificity of subunit-directed antibodies. Here, we used standard protocols for immunohistochemistry and western blotting to examine the antibodies raised against the alpha3-, alpha4-, alpha7-, beta2-, and beta4-nicotinic acetylcholine receptor subunits on brain tissues of the respective knock-out mice. Unexpectedly, for each of the antibodies tested, immunoreactivity was the same in wild-type and knock-out mice. These data imply that, under commonly used conditions, these antibodies are not suited for immunolocalization. Thus, particular caution should be exerted with regards to the experimental approach used to visualize nicotinic acetylcholine receptors in the brain.

Presence of Ras guanyl nucleotide-releasing protein in striosomes of the mature and developing rat.

Ras signal transduction pathways have been implicated as key regulators in neuroplasticity and synaptic transmission in the brain. These pathways can be modulated by Ras guanyl nucleotide exchange factors, (GEF) which activate Ras proteins by catalysing the exchange of GDP for GTP. Ras guanyl nucleotide-releasing protein (RasGRP), a recently discovered Ras GEF, that links diacylglycerol and probably calcium to Ras signaling pathways, is expressed in brain as well as in T-cells. Here, we have used a highly selective monoclonal antibody against RasGRP to localize this protein within the striatum and related forebrain structures of developing and adult rats. RasGRP immunolabeling was found to be widespread in the mature and developing rat forebrain. Most notably, it presented a prominent patchy distribution throughout the striatum at birth and at all postnatal ages examined. These patches were found to correspond with the striosomal compartment of the striatum, as identified by micro-opioid receptor labeling in the adult. RasGRP-immunoreactivity was also observed in the matrix-like compartment surrounding these patches/striosomes but appeared later in development and was always weaker than in the patches. In both striatal compartments, RasGRP was exclusively expressed by medium-sized spiny neurons and showed no preference for neurons that project either directly or indirectly to the substantia nigra. At the ultrastructural level, immunogold labeling of RasGRP was confined to the cell bodies and dendritic shafts of these output neurons. We conclude that the prominent expression of RasGRP in striosomes may be of significance for diacylglycerol signaling in the striatum, and could be of importance for the processing of limbic-related activity within the basal ganglia.

Cellular and subcellular localization of Ras guanyl nucleotide-releasing protein in the rat hippocampus.

Ras guanyl nucleotide-releasing protein (RasGRP) is a recently discovered Ras guanyl nucleotide exchange factor that is expressed in selected regions of the rodent CNS, with high levels of expression in the hippocampus. Biochemical studies suggest that RasGRP can activate the Ras signal pathway in response to changes in diacylglycerol and possibly calcium. To investigate potential sites for RasGRP signaling, we have determined the cellular and subcellular localization of RasGRP protein in adult rat hippocampus, and have also examined the appearance of RasGRP mRNA and protein during hippocampal development. RasGRP immunoreactivity is predominately localized to those neurons participating in the direct cortico-hippocampo-cortical loop. In both hippocampal and entorhinal neurons, RasGRP protein appeared to be localized to both dendrites and somata, but not to axons. Electron microscopy of hippocampal pyramidal cells confirmed RasGRP immunoreactivity in neuronal cell bodies and dendrites, where it appeared to be associated with microtubules. The localization of RasGRP to dendrites suggests a role for this pathway in the regulation of dendritic function. Examination of developing hippocampal structures indicated that RasGRP mRNA and protein appear synchronously during the first 2 weeks of postnatal development as these neurons become fully mature. This result indicates that the RasGRP signal transduction pathway is not required during early hippocampal development, but is a feature of mature neurons during the later stages of development.

The cholinergic innervation develops early and rapidly in the rat cerebral cortex: a quantitative immunocytochemical study.

A recently developed method for determining the length of cholinergic axons and number of cholinergic axon varicosities (terminals) in brain sections immunostained for choline acetyltransferase was used to estimate the areal and laminar densities of the cholinergic innervation in rat frontal (motor), parietal (somatosensory) and occipital (visual) cortex at different postnatal ages. This cortical innervation showed an early beginning, a few immunostained fibers being already present in the cortical subplate at birth. In the first two postnatal weeks, it developed rapidly along three parameters: a progressive increase in the number of varicosities per unit length of axon, and a lengthening and branching of the axons. Between postnatal days 4 and 16, the number of varicosities increased steadily from two to four per 10 microm of cholinergic axon. The mean densities of cholinergic axons increased from 1.4 to 9.6, 1.7 to 9.3 and 0.7 to 7.2 m/mm(3), and the corresponding densities of varicosities from 0.4 to 3.9, 0.4 to 3.5, and 0.2 to 2.6x10(6)/mm(3) in the frontal, parietal and occipital areas, respectively. The rate of growth was maximal during these first two weeks, after which the laminar pattern characteristic of each area appeared to be established. Adult values were almost reached by postnatal day 16 in the parietal cortex, but maturation proceeded further in the frontal and particularly in the occipital cortex. These quantitative data on the ingrowth and maturation of the cholinergic innervation in postnatal rat cerebral cortex substantiate a role for acetylcholine in the development of this brain region and emphasize the striking growth capacity of individual cholinergic neurons.

Cholinergic innervation in adult rat cerebral cortex: a quantitative immunocytochemical description.

A method for determining the length of acetylcholine (ACh) axons and number of ACh axon varicosities (terminals) in brain sections immunostained for choline acetyltransferase (ChAT) was used to estimate the areal and laminar densities of this innervation in the frontal (motor), parietal (somatosensory), and occipital (visual) cortex of adult rat. The number of ACh varicosities per length of axon (4 per 10 microm) appeared constant in the different layers and areas. The mean density of ACh axons was the highest in the frontal cortex (13.0 m/mm(3) vs. 9.9 and 11.0 m/mm(3) in the somatosensory and visual cortex, respectively), as was the mean density of ACh varicosities (5.4 x 10(6)/mm(3) vs. 3.8 and 4.6 x 10(6)/mm(3)). In all three areas, layer I displayed the highest laminar densities of ACh axons and varicosities (e.g., 13.5 m/mm(3) and 5.4 x 10(6)/mm(3) in frontal cortex). The lowest were those of layer IV in the parietal cortex (7.3 m/mm(3) and 2.9 x 10(6)/mm(3)). The lengths of ACh axons under a 1 mm(2) surface of cortex were 26.7, 19.7, and 15.3 m in the frontal, parietal, and occipital areas, respectively, for corresponding numbers of 11.1, 7.7, and 6.4 x 10(6) ACh varicosities. In the parietal cortex, this meant a total of 1.2 x 10(6) synaptic ACh varicosities under a 1 mm(2) surface, 48% of which in layer V alone, according to previous electron microscopic estimates of synaptic incidence. In keeping with the notion that the synaptic component of ACh transmission in cerebral cortex is preponderant in layer V, these quantitative data suggest a role for this innervation in the processing of cortical output as well as input. Extrapolation of particular features of this system in terms of total axon length and number of varicosities in whole cortex, length of axons and number of varicosities per cortically projecting neuron, and concentration of ACh per axon varicosity, should also help in arriving at a better definition of its roles and functional properties in cerebral cortex.

Quantified distribution of serotonin transporter and receptors during the postnatal development of the rat barrel field cortex.

Serotonin membrane transporter and 5-HT1B and 5-HT2A receptors were visualized and measured by autoradiography in the rat barrel field cortex at postnatal days 4, 8, 12, 16 and in adult (> P60). [3H]citalopram binding, reflecting the presence of 5-HT transporter on thalamocortical fibers, produced a clearcut barrel pattern from P4 to P16 (peak at P8), and decreased to a dispersed, low density in the adult. The patterning and temporal profile of 5-HT1B receptor binding ([125I]cyanopindolol) followed a parallel course. The 5-HT2A receptor binding ([125I]DOI) also conformed transiently to a barrel pattern; it increased in density from P8 to P16 and returned to a level as low as at P4 in the adult. These data suggest that 5-HT exerts a dual role in the developing somatosensory cortex: a local regulation of the peripherally-induced activity of thalamocortical axons via 5-HT1B receptors, and a trophic-like influence mediated by 5-HT2A receptors and possibly involving BDNF.

Melatonin in a primitive metazoan: seasonal changes of levels and immunohistochemical visualization in neurons.

Monthly day/night melatonin activity profiles were determined by radioimmunoassay over a 13-month period in the colonial anthozoan Renilla köllikeri, and no daily rhythmic oscillation was found. Averaging those monthly values yielded a seasonal quantitative rhythm in both colonial and non-colonial tissues of this cnidarian, with spring and summer levels found to be four to five times higher than autumn and winter ones. The annual rise, which occurred in two successive Aprils, coincided with the first stages of sexual maturation in R. köllikeri. Immunohistochemistry with a melatonin antibody raised in sheep revealed an exclusively neuronal distribution of melatonin-immunoreactivity (MEL-IR) in the endodermal septal filaments wrapped around gametophores, in endodermal walls of the rachis, and in the ectoderm of polyps. The MEL-IR ectodermal neurons shared many morphological features with serotonin-immunoreactive (5-HT-IR) neurons previously described in this animal but showed either weak or absent 5-HT-IR in double-labelling experiments. In contrast, MEL-IR and 5-HT-IR were strongly colocalized in endodermal neurons. These results indicate that melatonin is not a daily photoperiodic messenger but may instead act as a seasonal marker for reproduction in this cnidarian. We also provide the first evidence of a neuronal localisation of melatonin in an invertebrate, which suggests that melatonin may act as a neurotransmitter or neurohormone in the least evolved animals endowed with a nervous system.