Elodie Chaillou

Pulsatile Cerebrospinal Fluid and Plasma Ghrelin in Relation to Growth Hormone Secretion and Food Intake in the Sheep

As in other species, exogenous administration of ghrelin, an endogenous ligand for the growth hormone (GH) secretagogue receptors can stimulates feeding behaviour and GH secretion in the sheep. However, the importance of endogenous ghrelin for these two functions as well as its central or peripheral origin remained to be established. In the present study, cerebrospinal fluid (CSF) ghrelin concentrations were measured in five anoestrous ewes and found to be more than 1000‐fold lower than circulating plasma levels, in keeping with the even lower concentration in hypothalamic compared to abomasum tissue extracts. Cluster analysis indicated that CSF ghrelin levels were markedly pulsatile, with a greater number of peaks than plasma ghrelin. Pulsatility parameters were closer for GH and CSF ghrelin than between GH and plasma ghrelin. Plasma ghrelin and GH levels were significantly correlated in three out of five ewes but CSF ghrelin and GH in one ewe only. Half of the CSF ghrelin episodes were preceded by a ghrelin peak in plasma with a 22‐min delay. Cross‐correlations between plasma GH and plasma or CSF ghrelin did not reach significance but a trend towards cross‐correlation was observed from 20 to 0 min between plasma and CSF ghrelin. At 09.00 h, when food was returned to ewes, voluntary food intake did not elicit a consistent change in plasma or CSF ghrelin levels. By contrast, a peripheral ghrelin injection (1 mg, i.v.) immediately stimulated feeding behaviour and GH secretion. These effects were concomitant with a more than ten‐fold increase in plasma ghrelin levels, whereas CSF ghrelin values only doubled 40–50 min after the injection. This suggests that peripherally‐injected ghrelin crosses the blood–brain barrier, but only in low amount and with relatively slow kinetics compared to its effects on GH release and food intake. Taken together, the results obtained in the present study support the notion that, in the ovariectomised‐oestradiol implanted sheep model, peripheral ghrelin injection rapidly induces GH secretion, and feeding behaviour, probably by acting on growth hormone secretagogue receptor subtype 1 located in brain regions in which the blood–brain barrier is not complete (e.g. the arcuate nucleus).

Effect of feeding on Fos protein expression in sheep hypothalamus with special reference to the supraoptic and paraventricular nuclei: an immunohistochemical study

The hypothalamus plays an important role in the control of food intake in different species, but there is little relevant information for ruminants like sheep. In order to study the putative role of several hypothalamic nuclei in food intake in sheep, Fos expression, a marker of cellular activity, was compared by immunohistochemistry between fed and unfed ewes. The expression of Fos protein was stimulated in the supraoptic nucleus of fed ewes, whereas it was increased in the paraventricular nucleus of unfed animals. In the latter nucleus, Fos immunoreactivity was mainly localized close to the third ventricle, an area corresponding to the parvocellular system of the nucleus, but never in the magnocellular system. In the paraventricular nucleus, the number of corticotrophin releasing factor‐immunoreactive neurons and the number of Fos/corticotrophin releasing factor double‐labelled neurons were not affected by feeding or lack of feeding. The number of Fos‐immunoreactive neurons was higher in the lateral septum, the infundibular, the ventromedial and in the dorsomedial nuclei of unfed ewes than in those of fed ewes. Our results show for the first time that the dorsomedial and ventromedial nuclei are involved in the control of feeding in sheep as in rodents. The supraoptic nucleus of sheep is activated by the same conditions as in rodents but, conversely, the paraventricular nucleus is activated in unfed sheep, whereas in rodents and primates, this nucleus is activated by satiety as well as by fasting. In sheep, unlike in rodents, corticotrophin releasing factor did not appear to be involved in short‐term regulation of food intake.

Distribution of galanin immunoreactivity in the sheep diencephalon

Although the physiological role of galanin has been demonstrated in several endocrine regulations in sheep, the anatomical characteristics of this neuronal system has never been studied. The distribution of galanin-containing neurones was described by immunohistochemistry using galanin antiserum in the diencephalon of adult ewes, both ovariectomized or treated with colchicine. Galanin-immunoreactivity was found throughout the diencephalon. In the ovariectomized ewes, galanin-immunoreactive neurones were mainly observed in the medial preoptic area and the infundibular nucleus. The highest density of immunoreactive fibres was found in the external layer of the median eminence. Numerous galanin-immunoreactive fibres were also observed in the preoptic area, the mediobasal hypothalamus, the periphery of the supraoptic and the paraventricular nuclei. With colchicine treatment, the number of labelled neurones increased, and additional galanin-immunoreactive perikarya were observed in the bed nucleus of the stria terminalis, the lateral septum, the supraoptic, the paraventricular and the periventricular nuclei and the paraventricular nucleus of the thalamus. In the caudal part of the diencephalon, the density of labelled neurones was lower in both groups of animals than in other species studied. Regardless of treatment, labelling was not seen in the suprachiasmatic nucleus and only rarely in the ventromedial nucleus. These results describe, for the first time, the distribution of galanin-immunoreactive neurones in the sheep diencephalon. Compared to other species studied, distribution in the sheep diencephalon has several distinct differences. In ovariectomized animals, the medial preoptic area presents more labelled neurones in sheep than in monkeys, whereas in the supraoptic nucleus the density of labelled neurones is lower in sheep than in humans or opossums. After colchicine treatment only very few differences were observed between sheep and rats, but in contrast to other species, the suprachiasmatic nucleus of the sheep does not contain labelled neurones.

Afferent and efferent connections of the cortical and medial nuclei of the amygdala in sheep.

The cortical (CoA) and the medial (MeA) nuclei of the amygdala are involved in the processing of olfactory information relevant to social recognition in the ewe. To better understand the neural pathways responsible for these effects, the connections of both CoA and MeA with the telencephalic and diencephalic regions were studied by injecting an anterograde (Biotin-Dextran-Amine, BDA) or a retrograde (Fluorogold, FG) neuronal tracer into either the CoA or the MeA. Concerning the primary olfactory structures, the CoA receives inputs from both the main olfactory bulb and the accessory olfactory bulb (AOB), while the MeA is innervated by cells only from the AOB. Among the other olfactory structures, only the entorhinal cortex and the tenia tecta are connected with both the CoA and the MeA. With respect to the other secondary olfactory structures, the connections with the CoA and the MeA show segregating neuronal routes. The CoA is connected with the accessory olfactory nucleus, the piriform, the endopiriform and the orbitofrontal cortices while the MeA exhibited connections with the nucleus of the lateral olfactory tract, the perirhinal and the insular cortices. Concerning the diencephalic structures, only the MeA receives projections from the PVN and the MBH. On the other hand, we showed that the BNST is the major site of connection with both the CoA and the MeA. Reciprocal projections were observed between the CoA and the MeA and between both nuclei and the basal or the lateral nuclei of the amygdala with the exception of the CoA which does not send inputs to the lateral nucleus. These data are discussed in relation with olfactory learning in the context of sexual and maternal behavior in sheep.

Body fat content and feeding level interact strongly in the short- and medium-term regulation of plasma leptin during underfeeding and re-feeding in adult sheep

Circulating leptin is regulated by food intake in the long, medium and short term; however, little is known about putative remnant effects of these successive regulations at any given time. To clarify this, two experiments were conducted in adult sheep, during which body condition parameters and plasma leptin were measured. During experiment 1, twenty ewes with normal body condition were either well fed (101 % of maintenance energy requirements (MER)) or underfed (41 % MER) for 166 d, then rapidly re-fed (at a mean of 208 % MER) for 3 d. Leptinaemia decreased after 14 d of underfeeding, remained depressed until day 166 and did not increase after 3 d re-feeding, whereas it was increased (+153 %; P < 0.05) by re-feeding the previously well-fed ewes. During experiment 2, twenty-four fat or lean ewes were either well fed (114 % MER) or underfed (52 % MER) for 94 d, and gradually re-fed for 2 d and maintained at a high feeding level (235 % MER) for 9 d. Underfeeding decreased leptinaemia in fat (from 4.19 to 2.63 ng/ml) but not lean ewes, and re-feeding increased leptinaemia after 5 d in lean previously well-fed (+123 %; P < 0.05) but not underfed ewes. In fat ewes, the impact of re-feeding was rapid (+144 %; P < 0.001 at 5 d) in previously well-fed ewes, whereas it was more gradual with a maximum at 11 d (+162 %; P < 0.01) in previously underfed ewes. In conclusion, leptinaemia is modulated by short-term energy intake level in interaction with long-term regulations involving nutritional history and body fatness, suggesting that a biological threshold of adiposity (about 20 %) is necessary to allow short- and medium-term leptin regulation.

Long-Term Undernutrition Followed by Short-Term Refeeding Effects on the Corticotropin-Releasing Hormone Containing Neurones in the Paraventricular Nucleus: An Immunohistochemical Study in Sheep

The effect of nutritional level on the immunoreactivity of corticotropin‐releasing hormone (CRH) in neurones of the hypothalamic paraventricular nucleus was described in sheep, a ruminant, whose feeding strategy differs from that of monogastric species. Two groups of ewes were underfed (40%), or fed at maintenance (100%) for 167 days, after which one‐half of each group was killed or ad libitum refed (at least 150% of maintenance) for 4 days before killing. The presence of CRH in the paraventricular nucleus was examined by immunohistochemistry. The number of CRH immunoreactive neurones was increased in underfed ewes, but without modification of the plasma concentration of cortisol, indicating that the rise of CRH was not released in the portal blood nor linked to the pituitary‐adrenal axis. Refeeding did not modify significantly the number of CRH immunoreactive neurones in the nucleus although these neurones were increased, only in refed ewes that were previously underfed. These data differ from those for rats and mice where CRH expression is decreased or not modified by underfeeding which could reflect different effects of undernutrition on CRH immunoreactive neurones in monogastric compared to ruminants species.

Multiple exposures to familiar conspecific withdrawal is a novel robust stress paradigm in ewes

Paradigms used so far to study the effects of social isolation in sheep confound the effects of social isolation with those of other stressors (e.g. new environment) and showed contradictory effects after multiple social isolations. We propose here to characterize and examine the repeatable effects of social isolation induced by the familiar conspecific withdrawal (FCW). This latter test consists of socially isolating the ewe by the removal of group mates from the room test for 3 h. Behavioral and endocrine responses of adult ovariectomized-estradiol implanted ewes were compared 90 min before and 90 min after FCW, which was applied three times every fourteen days. We observed that each FCW induced significant increases in plasma cortisol level, in the number of vocalizations, foot pawing, circling attempts and a significant decrease in time spent lying down. An increase in plasma cortisol levels and decrease in duration of maintenance behaviors were significantly lower after the third FCW than after the first one. These differences could be explained by higher plasma cortisol levels and lower duration of maintenance behaviors before the third FCW than before the first FCW suggesting an anticipation of the social isolation period. These data indicate that social isolation is sufficient to induce distress with stable stressful responses after multiple exposures to familiar conspecific withdrawal.

Presence of galanin in dopaminergic neurons of the sheep infundibular nucleus: a double staining immunohistochemical study.

The distribution of tyrosine hydroxylase (TH) and of galanin immunoreactive (IR) neurons were examined in the sheep infundibular nucleus. Antisera raised against TH and galanin were used on adjacent sections and for double immunohistochemical staining of the same sections. There was considerable overlap in the distribution of TH and galanin-IR neurons in the medial part of the nucleus. Most of the galanin-IR neurons were also TH-IR, but less than 50% of the TH-IR neurons also expressed galanin immunoreactivity. Neurons immunoreactive to TH alone were observed close to the third ventricle and in the rostral part of the infundibular nucleus. In the median eminence, TH and galanin-IR fibres overlapped mainly in the lateral and dorsal parts of the external layer, but the colocalisation of both antigens could not be assessed on the available material. Thus, in sheep, the population of catecholaminergic neurons of the infundibular nucleus may be subdivided into different subpopulations according to their peptide content, but does not appear segregated as in rat and human.

Stress during pregnancy alters dendritic spine density and gene expression in the brain of new-born lambs.

Rodent studies show how prenatal stress (PS) can alter morphology in the cortico-limbic structures that support emotional and cognitive functions. PS-induced alteration is less well described in species with a gyrencephalic brain and complex earlier fetal development, and never in sheep at birth to rule out postnatal environment effects or influences of maternal behavior. This study aimed to assess the consequences of a mild chronic stress in pregnant ewes on the neurobiological development of their lambs at birth. During the last third of gestation, 7 ewes were exposed daily to various unpredictable and negative routine management-based challenges (stressed group), while 7 other ewes were housed without any additional perturbation (control group). For each group, a newborn from each litter was sacrificed at birth to collect its brain and analyze its expression levels of genes involved in neuronal dendritic morphology (Dlg4, Rac1, RhoA, Doc2b), synaptic transmission (Nr1, Grin2A, Grin2B) and glucocorticoid receptor (Nr3C1) in hippocampus (HPC), prefrontal cortex (PFC) and amygdala (AMYG). Results revealed that lambs from stressed dam (PS lambs) showed under-expression of Rac1 and Nr1 in PFC and overexpression of Dlg4 in AMYG compared to controls. To assess the morphological consequences of gene dysregulations, the dendritic morphology of pyramidal neurons was explored by Golgi-Cox staining in HPC and PFC. PS lambs had higher dendritic spine density in both structures and more stubby-type spines in the CA1 area of HPC than controls. This is the first demonstration in sheep that PS alters fetal brain, possibly reflecting functional changes in synaptic transmission to cope with adversity experienced in fetal life.

Rapid effects of melatonin on hormonal and behavioral stressful responses in ewes

Sheep are gregarious mammals with complex social interactions. As such, they are very sensitive to social isolation and constitute a relevant animal model to study specifically the biological consequences of social stress. We examined previously the behavioral and endocrine responses in ewes isolated socially in the familiar conspecific withdrawal model (FCW) and showed that stressful responses increased and maintenance behaviors decreased, confirming that social isolation is a strong stressor in sheep. Melatonin synchronizes seasonal and circadian rhythms; and several studies reported its implication in cognitive processes as emotion. Here we investigated its role in the modulation of social stressful responses. Firstly, we studied ewes in the FCW model during the day (characterized by low melatonin levels) and the night (characterized by high melatonin levels). We found lower stressful responses (significant lower levels of cortisol plasma, number of foot pawings, of circling attempts) during the night as compared to the day. To investigate whether these effects were due to melatonin or to darkness, we submitted ewes to FCW during the night with lights on, a condition that suppresses melatonin secretion. Ewes infused with melatonin under these conditions showed decreased stressful responses (significant lower levels cortisol plasma, number of vocalizations, time spent with the head out of the cage) as compared to ewes infused with saline. These findings demonstrate that melatonin diminishes the endocrine and behavioral impact of social isolation in ewes and support the idea that melatonin has a calming effect in socially stressful situations.