Neuronal projections to the lateral retrochiasmatic area of sheep with special reference to catecholaminergic afferents: immunohistochemical and retrograde tract-tracing studies

Heterogeneous expression of dopaminergic markers and Vglut2 in mouse mesodiencephalic dopaminergic nuclei A8-A13

Co-transmission of glutamate by brain dopaminergic (DA) neurons was recently proposed as a potential factor influencing cell survival in models of Parkinson's disease. Intriguingly, brain DA nuclei are differentially affected in Parkinson's disease. Whether this is associated with different patterns of co-expression of the glutamatergic phenotype along the rostrocaudal brain axis is unknown in mammals. We hypothesized that, as in zebrafish, the glutamatergic phenotype is present preferentially in the caudal mesodiencephalic DA nuclei. Here, we used in mice a cell fate mapping strategy based on reporter protein expression (ZsGreen) consecutive to previous expression of the vesicular glutamate transporter 2 (Vglut2) gene, coupled with immunofluorescence experiments against tyrosine hydroxylase (TH) or dopamine transporter (DAT). We found three expression patterns in DA cells, organized along the rostrocaudal brain axis. The first pattern (TH-positive, DAT-positive, ZsGreen-positive) was found in A8-A10. The second pattern (TH-positive, DAT-negative, ZsGreen-positive) was found in A11. The third pattern (TH-positive, DAT-negative, ZsGreen-negative) was found in A12-A13. These patterns should help to refine the establishment of the homology of DA nuclei between vertebrate species. Our results also uncover that Vglut2 is expressed at some point during cell lifetime in DA nuclei known to degenerate in Parkinson's disease and largely absent from those that are preserved, suggesting that co-expression of the glutamatergic phenotype in DA cells influences their survival in Parkinson's disease.

First evidence of neuronal connections between specific parts of the periaqueductal gray (PAG) and the rest of the brain in sheep: placing the sheep PAG in the circuit of emotion

The periaqueductal gray (PAG) is a mesencephalic brain structure organised in subdivisions with specific anatomical connections with the rest of the brain. These connections support the different PAG functions and especially its role in emotion. Mainly described in territorial and predatory mammals, examination of the PAG connections suggests an opposite role of the ventral and the dorsal/lateral PAG in passive and active coping style, respectively. In mammals, the organisation of PAG connections may reflect the coping style of each species. Based on this hypothesis, we investigated the anatomical connections of the PAG in sheep, a gregarious and prey species. Since emotional responses expressed by sheep are typical of active coping style, we focused our interest on the dorsal and lateral parts of the PAG. After injection of fluorogold and fluororuby, the most numerous connections occurred with the anterior cingulate gyrus, the anterior hypothalamic region, the ventromedial hypothalamic nucleus and the PAG itself. Our observations show that the sheep PAG belongs to the neuronal circuit of emotion and has specific parts as in other mammals. However, unlike other mammals, we observed very few connections between PAG and either the thalamic or the amygdalar nuclei. Interestingly, when comparing across species, the PAG connections of sheep were noticeably more like those previously described in other social species, rabbits and squirrel monkeys, than those in territorial species, rats or cats.

Neurokinin B acts via the neurokinin-3 receptor in the retrochiasmatic area to stimulate luteinizing hormone secretion in sheep

Recent data have demonstrated that mutations in the receptor for neurokinin B (NKB), the NK-3 receptor (NK3R), produce hypogonadotropic hypogonadism in humans. These data, together with reports that NKB expression increases after ovariectomy and in postmenopausal women, have led to the hypothesis that this tachykinin is an important stimulator of GnRH secretion. However, the NK3R agonist, senktide, inhibited LH secretion in rats and mice. In this study, we report that senktide stimulates LH secretion in ewes. A dramatic increase in LH concentrations to levels close to those observed during the preovulatory LH surge was observed after injection of 1 nmol senktide into the third ventricle during the follicular, but not in the luteal, phase. Similar increases in LH secretion occurred after insertion of microimplants containing this agonist into the retrochiasmatic area (RCh) in anestrous or follicular phase ewes. A low-dose microinjection (3 pmol) of senktide into the RCh produced a smaller but significant increase in LH concentrations in anestrous ewes. Moreover, NK3R immunoreactivity was clearly evident in the RCh, although it was not found in A15 dopaminergic cell bodies in this region. These data provide evidence that NKB stimulates LH (and presumably GnRH) secretion in ewes and point to the RCh as one important site of action. Based on these data, and the effects of NK3R mutations in humans, we hypothesize that NKB plays an important stimulatory role in the control of GnRH and LH secretion in nonrodent species.

Neural systems mediating seasonal breeding in the ewe

Seasonal reproduction in ewes is caused by a dramatic increase in response to oestradiol (E(2)) negative feedback during the nonbreeding (anoestrous) season. Considerable evidence supports the hypothesis that A15 dopaminergic neurones in the retrochiasmatic area (RCh) play a key role in these seasonal changes. These A15 neurones are stimulated by E(2) and inhibit gonadotrophin-releasing hormone (GnRH) secretion in anoestrus, but not the breeding season. Because A15 neurones do not contain oestrogen receptors-alpha (ER alpha), it is likely that E(2)-responsive afferents stimulate their activity when circulating E(2) levels increase during anoestrus. Retrograde tract tracing studies identified a limited set of ER alpha-containing afferents primarily found in four areas [ventromedial preoptic area, RCh, ventromedial and arcuate (ARC) nuclei]. Pharmacological and anatomical data are consistent with GABA- and glutamate-containing afferents controlling A15 activity in anoestrus, with E(2) inhibiting GABA and stimulating glutamate release at this time of year. Tract tracing demonstrated that A15 efferents project posteriorly to the median eminence and the ARC, suggesting possible direct actions on GnRH terminals or indirect actions via kisspeptin neurones in the ARC to inhibit GnRH in anoestrus. Identification of this neural circuitry sets the stage for the development of specific hypotheses for morphological or transmitter/receptor expression changes that would account for seasonal breeding in ewes.

Androgen receptor expressing neurons that project to the paraventricular nucleus of the hypothalamus in the male rat

Androgen receptors are distributed throughout the central nervous system and are contained by a variety of nuclei that are known to project to or regulate the paraventricular nucleus (PVN) of the hypothalamus, the final common pathway by which the brain regulates the hypothalamic-pituitary-adrenal (HPA) response to homeostatic threat. Here we characterized androgen receptor staining within cells identified as projecting to the PVN in male rats bearing iontophoretic or crystalline injections of the retrograde tracer FluoroGold aimed at the caudal two-thirds of the nucleus, where corticotropin-releasing hormone-expressing neurons are amassed. Androgen receptor (AR) and FluoroGold (FG) double labeling was revealed throughout the limbic forebrain, including scattered numbers of cells within the anterior and posterior subdivisions of the bed nuclei of the stria terminalis; the medial zone of the hypothalamus, including large numbers of AR-FG-positive cells within the anteroventral periventricular and medial preoptic cell groups. Strong and consistent colabeling was also revealed throughout the hindbrain, predominantly within the periaqueductal gray and the lateral parabrachial nucleus, and within various medullary cell groups identified as catecholaminergic, predominantly C1 and A1 neurons of the ventral medulla. These connectional data predict that androgens can act on a large assortment of multimodal inputs to the PVN, including those involved with the processing of various types of sensory and limbic information, and provide an anatomical framework for understanding how gonadal status could contribute to individual differences in HPA function.

A map of the major nuclei of the fetal sheep brainstem

The fetal sheep has been used to investigate a wide range of developmental and pathological processes such as the effect of severe hypoxia, asphyxia, or intrauterine infection on the brain but, until now, there has been no complete description of the normal anatomical organisation of neuronal groups to facilitate interpretation of these studies. In this paper, we describe the major nuclei of the fetal sheep brainstem based on a study of 5 fetal sheep at 140 days of gestation (G140: term is G147). Nuclei were identified with the aid of brain atlases available for other species, and from the previously published, partial descriptions available for the sheep. Fifty-five distinct nuclei were identified after Nissl (thionin) staining, and their caudal and rostral margins were defined. This paper provides an easy reference to the position of the major nuclei within the fetal sheep brainstem, and can be used as a guide for future studies examining the organisation of neuronal populations under normal and pathological conditions in this animal model.

Noradrenergic regulation of hypothalamic cells that produce growth hormone-releasing hormone and somatostatin and the effect of altered adiposity in sheep

The growth hormone (GH) axis is sensitive to alteration in body weight and there is evidence that central noradrenergic systems regulate neurones that produce growth hormone-releasing hormone (GHRH) and somatostatin (SRIF). This study reports semiquantitative estimates of the noradrenergic input to neuroendocrine GHRH and SRIF neurones in the sheep of different body weights. We also studied the effects of altered body weight on expression of dopamine beta-hydroxylase (DBH), the enzyme that produces noradrenalin from dopamine. Ovariectomised ewes were made Lean (39.6 +/- 2.6 kg; Mean +/- SEM) by dietary restriction, whereas Normally Fed animals (61.2 +/- 0.8 kg) were maintained on a regular diet. Brains were perfused for immunohistochemistry and in situ hybridisation. The Mean +/- SEM number of GHRH-immunoreactive (-IR) cells was lower in Normally Fed (65 +/- 7) than in Lean (115 +/- 14) animals, whereas the number of SRIF-IR cells was similar in the two groups (Normally Fed, 196 +/- 17; Lean 230 +/- 21). Confocal microscopic analysis revealed that the percentage of GHRH-IR cells (Normally Fed 36 +/- 1.5% versus Lean 32 +/- 4.6%) and percentage of SRIF-IR cells (Normally Fed 30 +/- 40.4% versus Lean 32 +/- 2.3%) contacted by noradrenergic fibres did not change with body weight. FluoroGold retrograde tracer injections confirmed that noradrenergic projections to the arcuate nucleus are from ventrolateral medulla and noradrenergic projections to periventricular nucleus arise from the ventrolateral medulla, nucleus of solitary tract, locus coeruleus (LC) and the parabrachial nucleus (PBN). DBH expressing cells were identified using immunohistochemistry and in situ hybridisation and the level of expression (silver grains/cell) quantified by image analysis. The number of DBH cells was similar in Normally Fed and Lean animals, but the level of expression/cell was lower (P < 0.02) in the PBN and LC of Lean animals. These results provide an anatomical basis for the noradrenergic regulation of GHRH and SRIF cells and GH secretion. Altered activity or noradrenergic neurones in the PBN and LC that occur with reduced body weight may be relevant to the control of GH axis.

Progesterone-receptive beta-endorphin and dynorphin B neurons in the arcuate nucleus project to regions of high gonadotropin-releasing hormone neuron density in the ovine preoptic area

Progesterone inhibits gonadotropin-releasing hormone (GnRH) secretion through interneuronal systems located in the mediobasal hypothalamus in ewes. Endogenous opioid peptides are implicated in this inhibition of GnRH secretion. The distributions of endogenous opioid peptides are known to overlap with progesterone receptors (PR) in the arcuate nucleus. We investigated whether PR is expressed by beta-endorphin and dynorphin B neurons in the arcuate nucleus and if a subset of double-labeled cells projects to the preoptic area where most GnRH neurons are detected. Injection of a retrograde tracer, Fluorogold, into the rostral preoptic area was performed in ovariectomized ewes pretreated with estrogen and progesterone. Brain sections were processed using double immunocytochemistry. Only brains of ewes with an injection site encompassing at least 80 GnRH neurons were processed for PR and then either beta-endorphin or dynorphin B immunocytochemistry. Antigen retrieval is essential for PR detection but causes Fluorogold to fade. Thus, quantitative analysis was performed on photographs taken before and after antigen retrieval. We found that 25-30% of PR-containing neurons, 20% of beta-endorphin cells and 22% of dynorphin B neurons in the arcuate nucleus project toward the preoptic area. From the PR/beta-endorphin double-labeled cells that represent 25 and 36% of PR and beta-endorphin cells, respectively, 35% were labeled with Fluorogold. From the PR/dynorphin B double-labeled cells that account for 39 and 62% of PR and dynorphin B neurons, respectively, 26% contained Fluorogold. These data strongly support the hypothesis that progesterone acts in the arcuate nucleus through beta-endorphin and dynorphin B neurons to affect preoptic area GnRH neurons.

Progesterone-receptive dopaminergic and neuropeptide Y neurons project from the arcuate nucleus to gonadotropin-releasing hormone-rich regions of the ovine preoptic area

Progesterone inhibits gonadotropin-releasing hormone (GnRH) secretion in sheep through an interneuronal system located in the mediobasal hypothalamus. This study focused on known inhibitors of GnRH secretion in sheep, dopamine and neuropeptide Y (NPY). As the distributions of tyrosine hydroxylase (TH)- and NPY-immunoreactive neurons overlap with progesterone receptors (PR) in the arcuate nucleus, we hypothesized that, if these neurons mediate, at least partially, the inhibitory feedback signal of progesterone, then they should co-express PRs. Fluorogold (FG), a retrograde tracer, was injected into the preoptic area of ovariectomized ewes pretreated with estrogen and progesterone. When the FG injection site encompassed at least 80 GnRH neurons, sections from the arcuate nucleus were processed using dual immunocytochemistry for PR and either TH or NPY. We found that 30% of PR-immunoreactive, 12% of TH-containing and 21% of NPY-synthesizing neurons project toward this GnRH-rich region. Of the PR/TH dual-labeled cells, which represent 21% of PR and 31% of TH cells, respectively, 22% displayed FG labeling. Of the PR/NPY neurons, which account for 19% of PR and 67% of NPY neurons, respectively, 26% were FG fluorescent. This study suggests that subsets of arcuate nucleus dopaminergic and NPY neurons may transduce, at least in part, the progesterone-mediated inhibition of GnRH secretion.

Differential expression of cocaine- and amphetamine-regulated transcript and agouti related-protein in chronically food-restricted sheep

Recently, much attention has focused on the role of the melanocortin system in the regulation of energy homeostasis, especially the satiety effects of the pro-opiomelanocortin (POMC)-derived peptide alpha-melanocyte stimulating hormone (alpha-MSH). We have found that POMC mRNA levels are similar in fat and thin sheep and the current study sought to further characterize the effects of nutritional status on the melanocortin system. To this end, we studied the expression of agouti-related peptide (AGRP) (an endogenous antagonist of alpha-MSH) and cocaine- and amphetamine-regulated transcript (CART), which is co-localized within POMC cells of the arcuate nucleus (ARC) in rodents. Twelve ovariectomized ewes were randomly divided into two groups and fed a maintenance (n=6) or restricted diet (n=6). At the time of experimentation, the animals had significantly (P<0.0001) different bodyweights (53.4+/-2.2 kg, ad libitum vs. 30.4+/-1.2 kg, food-restricted), which was largely due to altered body fat deposits. In situ hybridization was used to study the expression of POMC, AGRP and CART. The expression of POMC in the ARC was similar in ad libitum and food-restricted animals but the expression of AGRP was profoundly increased in the food-restricted group. The expression of CART was abundant throughout the hypothalamus but was not found in the ARC. In food-restricted animals, the expression of CART was lower in the retrochiasmatic nucleus (P<0.01), paraventricular nucleus (P<0.001), the dorsomedial nucleus and the lateral hypothalamic area (P<0.05), but was higher (P<0.01) in the posterior hypothalamic area. Thus, long-term changes in nutritional status have profound effects on the expression of AGRP and CART in the hypothalamus.