Vasopressin-induced calcium signaling in cultured cortical neurons

Earlier autoradiographic studies from our laboratory detected vasopressin recognition sites in the mammalian cerebral cortex [R.E. Brinton, K.W. Gee, J.K. Wamsley, T.P. Davis, H.I. Yamamura, Regional distribution of putative vasopressin receptors in rat brain and pituitary by quantitative autoradiography, Proc. Natl. Acad. Sci. U.S.A., 81 (1984) 7248–7252; C. Chen, R.D. Brinton, T.J. Shors, R.F. Thompson, Vasopressin induction of long-lasting potentiation of synaptic transmission in the dentate gyrus, Hippocampus, 3 (1993) 193–204]. More recently, we have detected mRNA for the V1a vasopressin receptors (V1aRs) in cultured cortical neurons [R.S. Yamazaki, Q. Chen, S.S. Schreiber, R.D. Brinton, V1a Vasopressin receptor mRNA expression in cultured neurons, astroglia, and oligodendroglia of rat cerebral cortex, Mol. Brain Res., 45 (1996) 138–140]. To determine whether these recognition sites are functional receptors, we have pursued the signal transduction mechanism associated with the V1a vasopressin receptor in enriched cultures of cortical neurons. Results of these studies demonstrate that exposure of cortical neurons to the selective V1 vasopressin receptor agonist, [Phe²,Orn⁸]-vasotocin, (V₁ agonist) induced a significant accumulation of

Full-size image

inositol-1-phosphate (

Full-size image

IP₁). V₁ agonist-induced accumulation of

Full-size image

IP₁ was concentration dependent and exhibited a linear dose response curve. Time course analysis of V₁ agonist-induced accumulation of

Full-size image

IP₁ revealed a significant increase by 20 min which then decreased gradually over the remaining 60 min observation period. V₁ agonist-induced accumulation of

Full-size image

IP₁ was blocked by a selective V1a vasopressin receptor antagonist, (Phenylac¹,

Full-size image

-Tyr(Me)², Arg^6,8, Lys-NH₂⁹)-vasopressin. Results of calcium fluorometry studies indicated that V₁ agonist exposure induced a marked and sustained rise in intracellular calcium which was abolished in the absence of extracellular calcium. The loss of the rise in intracellular calcium was not due to a failure to induce PIP₂ hydrolysis since activation of the phosphatidylinositol pathway occurred in the absence of extracellular calcium. V₁ agonist activation of calcium influx was then investigated. V₁ agonist-induced

Full-size image

uptake was concentration dependent with a biphasic time course. Preincubation with the L-type calcium channel blocker, nifedipine, blocked V₁ agonist-induced calcium influx suggesting V₁ agonist-induced L-type calcium channel activation in cortical neurons. Furthermore, V₁ agonist-induced calcium influx was blocked by both bisindolyleimide I (PKC inhibitor) and U-73122 (PLC inhibitor) suggesting a modulation of V₁ agonist-induced L-type calcium channel activation by downstream components of the phosphatidylinositol signaling pathway such as protein kinase C. These results indicate that in cultured cortical neurons, V1a vasopressin receptor activation leads to induction of the phosphatidylinositol signaling pathway, influx of extracellular calcium via L-type calcium channel activation, and a rise in intracellular calcium which is dependent on V1a receptor activated influx of extracellular calcium. These data are the first to demonstrate an effector mechanism for the V₁ vasopressin receptor in the cerebral cortex and provide a potential biochemical mechanism that may underlie vasopressin enhancement of memory function.     

Vasopressin-induced cytoplasmic and nuclear calcium signaling in cultured cortical astrocytes

We sought to determine whether vasopressin V(1a) receptor (V(1a)R) mRNA detected in cortical astrocytes [Brain Res. Mol. Brain Res. 45 (1997) 138] was translated into functional receptors by investigating the effector calcium signaling cascade associated with the vasopressin V(1a) receptor subtype. Analysis of intracellular calcium dynamics using the calcium-sensitive dye fura-2 AM indicated that exposure of cortical astrocytes to V(1) vasopressin receptor agonist, [Phe(2),Orn(8)]-oxytocin, induced a marked dose-dependent increase in intracellular calcium which was abolished by depletion of extracellular calcium. V(1) agonist treatment induced a rapid increase in calcium signal in both the cytoplasm and nucleus, which was followed by an accumulation of the calcium signal in the nucleus, suggesting translocation of cytoplasmic calcium into the nucleus. The nuclear calcium signal was sustained for several seconds followed by translocation back to the cytoplasm. Following the nuclear-to-cytoplasmic calcium translocation, total free intracellular calcium concentration decreased. The dynamic calcium cytoplasmic and nuclear localization was confirmed by laser scanning confocal microscopy coupled with the calcium-sensitive dye fluo-3 AM. To determine the source of calcium, V(1) agonist-induced (45)Ca(2+) uptake and [(3)H]IP(1) accumulation were investigated. V(1) agonist induced significant and rapid uptake of (45)Ca(2+) and a significant dose-dependent increase in [(3)H]IP(1) accumulation in cortical astrocytes. To our knowledge, this is the first documentation of a vasopressin receptor-induced calcium signaling cascade in cortical astrocytes and the first documentation of vasopressin receptor induction of nuclear calcium signaling.     

Vasopressin-induced motor effects: Localization of a sensitive site in the amygdala

Arginine vasopressin (AVP) induces motor effects when administered into the cerebral ventricles, the ventral septal area (VSA), or the vestibular cerebellum of the rat brain. Because AVP-like immunoreactivity and AVP-binding sites exist in the central medial amygdala (cmeA), and because the amygdala can be kindled to produce motor effects, we hypothesized that the amygdala might play a role in AVP-induced motor effects. This hypothesis was tested by observing motor behavior in response to injection of AVP into the central medial region of the amygdala. Our results demonstrate that an initial injection of AVP into the cmeA caused minor effects, including immobility, prostration and ataxia, whereas a similar injection, given 24 h later, caused severe motor effects including barrel rotations and myoclonic/myotonic-like convulsive behavior. A potential receptor basis for the AVP-induced motor and sensitization effects in the cmeA was investigated using AVP analogues. A V₁ antagonist, d(CH₂)₅Tyr(Me)AVP), blocked both the motor and sensitization effects produced by cmeA AVP injection. A V₂ receptor agonist, DDAVP, did not affect motor activity upon cmeA injection, but did, however, sensitize animals to subsequent cmeA AVP injection. These results suggest that the cmeA is a sensitive site for AVP-induced motor effects and that these motor effects are sensitized by prior exposure to AVP. While the motor effects observed after cmeA AVP injection are mediated via AVP receptors that resemble the V₁ type, the sensitization effect may be mediated via multiple receptor system. This study syggests that AVP may play a neurotransmitter or a neuromodulator role in motor behavior, and that the cmeA may function as one CNS site for the mediation of such behavior.     

The cannabinoid agonist Win55,212-2 inhibits calcium channels by receptor-mediated and direct pathways in cultured rat hippocampal neurons

The effects of the cannabinoid receptor agonist Win55,212 on Ca²⁺ channels were studied in rat hippocampal neurons grown in primary culture. Win55,212-2 inhibited whole-cell Ba²⁺ currents through Ca²⁺ channels by both CB1 receptor-mediated and direct mechanisms. The concentration dependent inhibition of the current showed two clear phases, a high-affinity receptor-mediated phase (IC₅₀=14±2 nM) that was stereoselective and sensitive to a CB1 receptor antagonist, 300 nM SR141716, and a non-saturating phase that was neither stereoselective nor inhibited by SR141716. These concentration-dependent effects were paralleled by Win55212-induced inhibition of glutamatergic synaptic transmission. Win55,212-2 (100 nM) inhibited both ω-agatoxin IVA- and ω-conotoxin GVIA-sensitive currents. Thus, activation of cannabinoid receptors inhibits N- and P/Q-type Ca²⁺ channels. Activation of cannabinoid receptors inhibited only a fraction of the whole-cell Ca²⁺ channel current (17±2%) even though more than half of the whole-cell Ba²⁺ current was carried by N- and P/Q-type Ca²⁺ channels. Concentrations of agonist greater than 1 μM inhibited Ca²⁺ channels directly.     

Intact spatial memory in mice with seizure-induced partial loss of hippocampal pyramidal neurons

We generated defined neuronal loss in hippocampus of genetically identical mice by pilocarpine injections and studied the impact of these seizures on the performance of mice in spatial learning and memory. The numbers of TUNEL-positive degenerating cells paralleled the severity of the seizures. When compared to the numbers found for not-seizured control mice, mild, moderate, and severe seizures produced significant increases in TUNEL-positive neurons in CA1 and CA3 regions by 19, 25, and 63%, respectively. Water maze learning was abolished after the severe seizures. However, spatial learning was normal after mild or moderate seizures. Therefore, there was no linear correlation between the impairment of learning and memory performance with the number of degenerating neurons in hippocampus. Our data suggest that normal spatial learning and memory can be achieved without the full number of hippocampal pyramidal neurons in partially lesioned hippocampus.     

Distinct signaling pathways involved in multiple effects of basic fibroblast growth factor on cultured rat hippocampal neurons

We investigated possible involvement of voltage-dependent Ca(2+) channels (VDCCs) and several intracellular signaling mechanisms in multiple actions of basic fibroblast growth factor (bFGF), such as survival promotion, induction of calbindin D(28k) expression as well as acceleration of neuritic branch formation of cultured rat hippocampal neurons. Immunocytochemical staining with anti-gamma-aminobutyric acid (GABA) antibody showed that the promotion of neuron survival by bFGF in high cell-density cultures were exerted exclusively on GABA-negative neurons. Nicardipine (5 microM) attenuated the effect of bFGF on neuronal survival and formation of neurite branches, suggesting that the activity of L-type VDCCs is required for these effects. In contrast, stimulation of calbindin expression by bFGF was not attenuated by nicardipine. A phospholipase C inhibitor U73122 (1 microM) prevented the effect of bFGF on neurite branch formation, but not on neuronal survival or calbindin expression. On the other hand, chronic application of phorbol-12-myristate-13-acetate (1 microM) inhibited the effect of bFGF on neuronal survival, without inhibiting the other bFGF actions. Forskolin (100 microM) attenuated the effect of bFGF on neuronal survival and neurite branch formation, indicating that cyclic AMP plays negative regulatory roles in these actions of bFGF. Taken together, these results suggest that multiple biological actions of bFGF on hippocampal neurons are exerted through, and modulated by, distinct signaling pathways.     

Vasopressin V1 receptor in rat hippocampus is regulated by adrenocortical functions

Two subtypes of arginine vasopressin (AVP) receptors (V1 and V2) have been distinguished. In this study, we examined the characteristics of AVP binding in rat hippocampus and the effects of bilateral adrenalectomy and adrenal steroids on its [³H]AVP binding. [³H]AVP binding to rat liver and the hippocampal membranes was strongly inhibited by the V1 antagonist, OPC-21268. ADX resulted in a significant decrease in theB_max of AVP binding in the hippocampus. Chronic treatment with aldosterone and corticosterone restored the ADX-induced reduction, but treatment with dexamethasone did not. These results suggest that the AVP V1 receptor in the hippocampus is regulated by adrenocortical neuroregulatory function.     

α2-Adrenergic receptor activation inhibits calcitonin gene-related peptide expression in cultured dorsal root ganglia neurons

Calcitonin gene-related peptide (CGRP), a potent vasodilator, is produced in dorsal root ganglia (DRG) neurons which extend nerves peripherally to blood vessels and centrally to the spinal cord. We previously reported that neuronal CGRP expression is significantly reduced in the spontaneously hypertensive rat (SHR) which could contribute to the elevated BP. Other studies suggest that the enhanced activity of the sympathetic nervous system in the SHR may mediate, at least in part, this reduction in neuronal CGRP expression via activation of α₂-adrenoreceptors (α₂-AR) on DRG neurons. To test this hypothesis in vitro we employed primary cultures of adult rat DRG neurons. Neuronal cultures were initially exposed (24 h) to either the α₂-AR agonist UK 14,304 (10⁻⁶ M) or vehicle; however, no changes in CGRP mRNA content or immunoreactive CGRP (iCGRP) release were observed. Using the rationale that in vivo DRG neurons receive a continuous supply of target tissue derived nerve growth factor (NGF), which stimulates CGRP synthesis, the cultured neurons were treated (24 h) with either vehicle, NGF (25 ng/ml) alone, or NGF plus UK. NGF treatment increased CGRP mRNA accumulation 5.5±0.9-fold (p<0.001) and iCGRP release 2.9±0.4-fold (p<0.001) over control levels. The stimulatory effects of NGF were markedly attenuated, but not abolished, by UK (NGF+UK vs. control, CGRP mRNA, 2.9±0.4-fold, p<0.05; iCGRP, 1.7±0.2-fold, p<0.05). These values were also significant (p<0.05) when compared to NGF treatment alone. Experiments performed using the α₂-antagonist yohimbine confirmed that the effects of UK were mediated by the α₂-AR. These results, therefore, demonstrate that α₂-AR activation attenuates the stimulatory effects of NGF on CGRP expression in DRG neurons.     

Reduced mastication stimulates impairment of spatial memory and degeneration of hippocampal neurons in aged SAMP8 mice

The involvement of reduced mastication in senile dementia was evaluated by examining the effect of cutting off the upper molars (molarless) on spatial memory and numbers of hippocampal neurons in aged SAMP8 mice. Molarless mice showed a decrease in both learning ability in a water maze and neuron density in the hippocampal CA1 region compared with control mice. These changes increased the longer the molarless condition persisted. The data suggest a possible link between reduced mastication and hippocampal neuron loss that may be one risk factor for senile impairment of spatial memory.     

Acute blockade of hippocampal glucocorticoid receptors facilitates spatial learning in rats

Corticosteroids can facilitate or impair learning and memory processes. We found that the glucocorticoid receptor antagonist RU38486 injected locally into the dorsal hippocampus dose-dependently improved the performance of male Wistar rats in the water maze 24 h after treatment. This observation suggests a discrete specificity of hippocampal glucocorticoid receptors in facilitation of memory.     

Central effects of vasopressin and 1-desamino-8-d-arginine vasopressin (DDAVP) on interleukin-1 fever in the rat

The intracerebroventricular administration of arginine vasopressin suppressed significantly the fever evoked by interleukin-1. This antipyretic action of arginine vasopressin was blocked completely by the antivasopressor analog d(CH2)5Tyr(Me)arginine vasopressin, an antagonist of the V1 subtype of peripheral vasopressin receptor. However, in contrast to AVP, the V2 receptor agonist, 1-desamino-8-D-arginine vasopressin, did not alter the normal time course or magnitude of interleukin-1 fever. These data suggest that arginine vasopressin induced antipyresis is mediated via central receptors which may resemble the V1 subtype of peripheral vasopressin receptor. The V2 subtype of vasopressin receptor is unlikely to be involved since an agonist of this receptor did not exhibit any antipyretic activity against interleukin-1 fever.     

Modulation of calcium current, intracellular calcium levels and cell survival by glucose deprivation and growth factors in hippocampal neurons

Basic fibroblast growth factor (bFGF) and nerve growth factor (NGF) can protect CNS neurons against ischemic/excitotoxic insults, but the mechanism of action is unknown. Imaging of the calcium indicator dye fura-3 and whole-cell patch clamp recordings of calcium currents were used to examine the mechanisms whereby hypoglycemia damages and growth factors protect cultured rat hippocampal neurons. When cultures were deprived of glucose, massive neuronal death occured 16–24 h following the onset of hypoglycemia. Early hypoglycemia-induced changes included calcium current inhibition and a reduction in intracellular free calcium levels ([Ca²⁺]_i) without morphological signs of neuronal damage. Later changes included a large elevation of [Ca²⁺]_i which was causally involved in neuronal damage. NGF and bFGF prevented or reduced both early and later responses to hypoglycemia. The growth factors increased calcium (barium) current and [Ca²⁺]_i to normal limits during the early stages of hypoglycemia and prevented the later elevation in [Ca²⁺]_i and neuronal damage. Nifedipine, but not omega-conotoxin, blocked calcium currents. The increased calcium current caused by the growth factors was apparently not sufficient to protect neurons against hypoglycemic damage since K⁺ depolarization during the early stages of hypoglycemia did not prevent and, in fact exacerbated, the subsequent neuronal damage. In addition, exposure of neurons to K⁺, NGF or bFGF only during the first 1 h of hypoglycemia did not protect against hypoglycemic damage. Taken together, the data suggest that neurons initially respond to hypoglycemia with a reduction in calcium currents which may provide a means to maintain [Ca²⁺]_i within a concentration range conducive to cell survival. Prolonged energy deprivation eventually results in a failure of calcium extrusion systems, glutamate receptor activation and a loss of neuronal calcium homeostasis. Taken together, the data indicate that the mechanism of growth factor protection against energydeprivation involves of the late prevention rise in [Ca²⁺]_i.     

Vasopressin induces human mesangial cell growth via induction of vascular endothelial growth factor secretion

Vasoactive hormones, growth factors, and cytokines are important in promoting mesangial cell growth, a characteristic feature of many glomerular diseases. Vascular endothelial growth factor (VEGF) is an endothelial mitogen and promoter of vascular permeability that is constitutively expressed in human glomeruli, but its role in the kidney is still unclear. In the present study, we investigated the ability of vasopressin (AVP) to stimulate VEGF secretion by and correlation with AVP-induced cell growth in human mesangial cells. AVP caused time- and concentration-dependent increases in VEGF secretion from human mesangial cells, which was in turn potently inhibited by a V_1A receptor-selective antagonist, confirming that this secretion is a V_1A receptor-mediated event. VEGF also induced mesangial cell growth which was completely inhibited on administration of an anti-VEGF neutralizing antibody. Further, AVP-induced mesangial cell growth was completely abolished by the V_1A receptor-selective antagonist and partially inhibited by an anti-VEGF neutralizing antibody. These results suggest that AVP stimulates VEGF secretion by human mesangial cells via V_1A receptors. This secreted VEGF may function as an autocrine hormone to regulate mesangial cell growth, a mechanism by which AVP might contribute to progressive glomerular diseases such as diabetic nephropathy.     

GABA(B) receptor activation inhibits N- and P/Q-type calcium channels in cultured lamprey sensory neurons

In lamprey, sensory transmission from mechanosensory receptors (dorsal cells) to central neurons is presynaptically inhibited by GABA(B) receptor activation. The mechanisms underlying this effect were investigated using isolated dorsal cells, where voltage-dependent calcium currents were recorded in the whole-cell configuration. Activation of GABA(B) receptors by baclofen decreased the peak amplitude of high voltage-activated (HVA) calcium currents and slowed the activation phase. The role of G-proteins in mediating the effects of baclofen was examined. Intracellular dialysis of GTPgammaS occluded the effects of baclofen. Intracellular dialysis of GDPbetaS and preincubation in pertussis toxin both attenuated the effect of baclofen. Specific calcium channel blockers were used to study the types of HVA calcium channels involved in the GABA(B)-mediated modulation. The baclofen-induced inhibition was not affected by the L-type calcium channel antagonist nimodipine, but was partially blocked by the N-type blocker omega-conotoxin GVIA, and completely occluded by omega-conotoxin MVIIC, a blocker of both N- and P/Q-type channels. The pharmacology of dorsal cell GABA(B) receptors was studied using two agonists, baclofen and CGP 27492, and four antagonists, CGP 35348, CGP 55845, phaclofen and saclofen. The inhibition induced by either of the two agonists was blocked by CGP 55845, phaclofen and saclofen. The antagonist CGP 35348 completely blocked the inhibition of HVA calcium current induced by the agonist CGP 27492, but had no effect on baclofen-induced GABA(B) receptor activation. This study thus demonstrates that GABA(B) receptor activation in lamprey mechanosensory neurons inhibits N- and P/Q-type calcium channels in a voltage- and G-protein-dependent manner.     

Desensitization of somatostatin-induced inhibition of low extracellular magnesium concentration-induced calcium spikes in cultured rat hippocampal neurons

Neuronal excitability is inhibited by somatostatin, which might play important roles in seizure and neuroprotection. The possibility of whether the effect of somatostatin on neurotransmission is susceptible to desensitization was investigated. We tested the effects of prolonged exposure to somatostatin on 0.1 mM extracellular Mg(2+) concentration ([Mg(2+)](o))-induced intracellular free Ca(2+) concentration ([Ca(2+)](i)) spikes in cultured rat hippocampal neurons using fura-2-based microfluorimetry. Reducing [Mg(2+)](o) to 0.1 mM elicited repetitive [Ca(2+)](i) spikes. These [Ca(2+)](i) spikes were inhibited by exposure to somatostatin-14. The inhibitory effects of somatostatin were blocked by pretreatment with pertussis toxin (PTX, 100 ng/ml) for 18-24 h. Prolonged exposure to somatostatin induced a desensitization of the somatostatin-induced inhibition of [Ca(2+)](i) spikes in a concentration-dependent manner. The somatostatin-induced desensitization was retarded by the nonspecific protein kinase C (PKC) inhibitor staurosporin (100 nM) or chronic treatment with phorbol dibutyrate (1 microM) for 24 h, but not by the protein kinase A inhibitor KT5720. The desensitization was significantly retarded by the novel PKCepsilon translocation inhibitor peptide (1 microM). In addition, suramin (3 microM), an inhibitor of G-protein-coupled receptor kinase 2 (GRK2), caused a reduction in the desensitization. After tetrodotoxin (TTX, 1 microM) completely blocked the low [Mg(2+)](o)-induced [Ca(2+)](i) spikes, glutamate-induced [Ca(2+)](i) transients were slightly inhibited by somatostatin and the inhibition was desensitized by prolonged exposure to somatostatin. These results indicate that the prolonged activation of somatostatin receptors induces the desensitization of somatostatin-induced inhibition on low [Mg(2+)](o)-induced [Ca(2+)](i) spikes through the activation of GRK2 and partly a novel PKCepsilon in cultured rat hippocampal neurons.     

Hyperexcitability in CA1 of the rat hippocampal slice following hypoxia or adenosine

Participation of adenosine receptors in the depression of synaptic transmission during hypoxia, and the production of multiple populations spikes in the pyramidal neurons following hypoxia, has been investigated in the CA₁ area of the rat hippocampal slice. A method is presented for analysing such hyperexcitability, using input/output curves of the second population spike. This method provides evidence that rebound hyperexcitability following hypoxia or prolonged adenosine-mediated inhibition results from an increase in excitability of the CA₁ pyramidal neurons rather than from an increase in excitatory neurotransmitter release. Hypoxia-induced depression of the synaptic components of evoked field potentials was blocked in a concentration dependent manner by the selective A₁ receptor antagonist 8-cyclopenthyltheophylline (8-CPT), demonstrating extracellular accumulation of adenosine during hypoxia. Upon reoxygenation of slices following 30 min hypoxia, multiple population spikes were evoked by a single orthodromic stimulus in slices that exhibited only a single population spike prior to hypoxia. Such post-hypoxic hyperexcitability was not prevented by superfusion of slices with 8-CPT during hypoxia. Depression of synaptic transmission by 30 min superfusion of slices with 50 μM adenosine was also followed, upon washout, by the appearance of multiple population spikes. However, such hyperexcitability could not be produced by superfusion with adenosine analogues selective for A₁ receptors, cyclopentyladenosine, selective for A_2a receptors, 2-p-(2-carboxyethyl)phenetheylamino-5′-ethylcarboxamidoadenosine (CGS 21680), or active at A_2a and A_2b receptors,N⁶-[2-(3,5-dimethyoxyphenyl)-2-(2-methyl-phenyl)ethyl]adenosine, suggesting that adenosine receptors other than the A₁, A_2a or A_2b subtypes are involved in its generation.     

A postsynaptic G-protein in hippocampal long-term potentiation

The effects of guanosine triphosphate (GTP)-binding protein (G-protein) blockade on hippocampal LTP at stratum radiatum-CA₁ synapses was studied. Bath application of 20 mM lithium chloride (LiCl) inhibited long-term potentiation (LTP) of extracellularly-recorded excitatory postsynaptic potentials (EPSPs). Inclusion of 100 mM LiCl in intracellular recording electrodes was shown to block postsynaptic G-proteins by bath-application of baclofen, an agonist at the G-protein linked γ-aminobutyric acid (GABA_B) receptor. Under normal conditions, GABA_B receptor activation causes a hyperpolarization postsynaptically, and a decrease in neurotransmitter release presynaptically. With LiCl in the recording electrodes, the postsynaptically-mediated hyperpolarization was blocked, while the presynaptically-mediated depression of EPSPs was unaffected. With postsynaptic G-proteins blocked in this manner, LTP at these synapses was inhibited. These studies provide evidence for the involvement of a postsynaptic G-protein in LTP of stratum radiatum-CA₁ synapses.     

Autoradiographic detection of vasopressin binding sites, but not of oxytocin binding sites, in the sheep olfactory bulb

Oxytocin facilitates maternal behaviour in sheep. In the present study, we searched for the presence of oxytocin and vasopressin binding sites in the sheep olfactory bulb, a brain area which is thought to be involved in specific bond formation between the ewe and its lamb. Using in vitro autoradiography, we observed binding of tritiated vasopressin to the glomerular layer of the olfactory bulb. Competition studies performed with structural analogues and the use of a 125I-labelled linear vasopressin antagonist suggested that sites which bind vasopressin are V1 type receptors. In contrast, specific binding sites for oxytocin in the olfactory bulb could be detected neither in control females, nor in ovariectomized females treated with estradiol nor in postparturient ewes, although such sites were present in the uterus.