Electrophysiological studies of neurosecretory cells in the cat hypothalamus

Crosstalk in the magnocellular system during osmotic stimulation of one supraoptic nucleus

Neural connections linking the four magnocellular nuclei, i.e., the paired supraoptic (SON) and paraventricular (PVN) nuclei, may contribute to the simultaneous and parallel changes in firing patterns of oxytocinergic neurons during reflex milk ejection. To investigate these neural connections in the absence of suckling, intranuclear release of oxytocin (OT) was stimulated by microdialysis of hypertonic CSF containing 1 M NaCl (HS-CSF) into the right SON area and glucose metabolism of both SONs and PVNs and the neural lobe of virgin and lactating (10-12 day) rats was mapped by the autoradiographic [14C]deoxyglucose (DG) method. OT in the microdialysates and in plasma, obtained before and after 80-90 min of dialysis with CSF or HS-CSF, was quantified by RIA. In both virgin and lactating rats, microdialysis of HS-CSF unilaterally into the SON area significantly (p < 0.05) increased release of OT in the nucleus and into plasma, which was associated with enhanced (p < 0.05) metabolic activity in the ipsilateral and contralateral SON and the neural lobe but not in either PVN. Compared with virgins, lactating rats were less active, had lower (p < 0.05) glucose utilization in the hypothalamo-neurohypophysial system, and less (p < 0.05) OT in plasma during microdialysis of HS-CSF into the SON area. The osmotic stimulus did not activate neural structures (suprachiasmatic and medial amygdaloid nuclei) near the SON in either hemisphere. Thus, neural connections or, less likely, transport of OT via the subarachnoid space, may function to recruit activation of cells in the contralateral SON following hypertonic stimulation of cells in the other SON.

Gastric related properties of rat paraventricular neurons

Central control of gastric acid secretion (GAS) was formerly attributed to specific neurons in the lateral hypothalamus (LHA), but the rostral part of the paraventricular nucleus (r-PVN) of the hypothalamus was shown to be another site that modulates central control of GAS. In the present study, the characteristics of 145 spontaneously firing r-PVN neurons were investigated in 22 rats. Discharges were: high frequency regular or irregular, low frequency regular or irregular. About half of the regular discharges were phasic. Electrical stimulation in the r-PVN evoked responses in 407 LHA units. Response latencies ranged from 4.7 to 78.1 msec; indicating mono- and polysynaptic, and myelinated and nonmyelinated fiber connections. Gastric related and non-related r-PVN neurons were observed. It was also shown, for the first time, that electrophoretic application of various chemicals, especially glucose, affected chemosensitive neurons in the r-PVN. PVN neuron responses to LHA repetitive stimulation were classified as excitatory (E), excitatory-inhibitory (E-I), I and I-E. PVN neuronal discharges might be modulated by gastric type LHA neurons. Electrophoretically applied norepinephrine (NE) increased PVN neuronal activity and suppressed GAS. Results suggest that rostral PVN neurons might affect LHA control of GAS, with NE as the probable transmitter or modulator.

gamma-Aminobutyric acid (GABA) causes consistent depolarization of neurons in the guinea pig supraoptic nucleus due to an absence of GABAB recognition sites

The action of gamma-aminobutyric acid (GABA) in the supraoptic nucleus was investigated using guinea pig brain slices. GABA produced a membrane depolarization accompanied by a decrease in the input resistance. The action of GABA was concentration-dependent throughout a wide range of concentrations (10(-7)-10(-3) M). In none of the cells examined, a membrane hyperpolarization was observed. The reversal potential for the depolarization induced by GABA was about 25 mV positive to the resting membrane potential. The amplitude of the GABA-induced depolarization was increased to 1.5 X the control by reducing the external Cl- from 134.2 mM to 10.2 mM. The action of GABA was readily antagonized by relatively low concentrations of bicuculline (10(-5) M). The action of GABA in the hippocampus or in the anterior hypothalamus was markedly different from that in the supraoptic nucleus, i.e. GABA produced both depolarizing and hyperpolarizing responses in the hippocampus and consistently a hyperpolarization in the anterior hypothalamus. The depolarizing but not the hyperpolarizing response in the hippocampus was selectively blocked by picrotoxin (2 X 10(-5) M) or by bicuculline (10(-5) M). The depolarizing component was dependent on the external Cl- concentration and had a reversal potential similar to that of the depolarization induced by GABA in the supraoptic nucleus. The hyperpolarizing component was resistant to bicuculline and had a reversal potential about 30 mV negative to the resting membrane potential.(ABSTRACT TRUNCATED AT 250 WORDS)

Two classes of arginine vasopressin binding sites on rat brain membranes

Specific binding sites for arginine vasopressin (AVP) were demonstrated on rat brain membranes using [3H]AVP of high specific activity. At pH 7.4 in the presence of 5 mM MgCl2, one class of sites was measured with a KD of 0.56 nM and a Bmax of 4.3 fmol/mg protein. At pH 8.0 in the presence of MgCl2, two distinct sites were observed, having KD values of 0.42 and 13 nM and Bmax values of 5.6 and 68 fmol/mg protein, respectively, and similar results were obtained at pH 7.4 after repeatedly freezing and thawing the membranes. Binding increased with pH, apparently representing increased occupancy of the high capacity, lower affinity site. Binding to the lower affinity site was also enhanced by freezing and thawing membranes, or by adding 5 mM NiCl2 or 10 microM ZnCl2 to the incubation medium, whereas binding to the high affinity site was dependent on the addition of Mg. AVP was over 35 times more active in displacing 0.4 nM AVP than oxytocin or arginine-vasotocin, and 10,000 times more active than somatostatin. A number of other peptides had no effect on [3H]AVP binding at concentrations up to 10(-5) M. Autoradiography and regional dissection studies revealed a marked concentration of high affinity AVP-binding sites in the supraoptic and paraventricular nuclei of the hypothalamus, and Mg significantly enhanced the binding in these regions.

Electrophysiologic evidence for neural connections between the paraventricular nucleus and neurons of the supraoptic nucleus in the rat

These experiments were part of a study to examine the extrahypothalamic projections of the paraventricular nucleus in the rat. Thirty-nine cells, histologically identified as situated within the supraoptic nucleus, were recorded. Nine of those cells were classified as secreting vasopressin by their phasic patterns of electrical discharge. Six of the phasically firing cells were excited after electrical stimulation of the paraventricular nucleus and the other three were inhibited. Eight cells were antidromically identified as projecting to the paraventricular nucleus; most of these were not spontaneously active. Of the remaining 22 cells, the majority were spontaneously active, showing continuous patterns of firing, and most were inhibited after stimulation of the paraventricular nucleus. The data supply evidence for the presence of functional neural connections between the two major nuclei of the tuberohypophyseal neurosecretory system.

Synaptic inputs and action potentials of magnocellular neuropeptidergic cells: intracellular recording and staining in slices of rat hypothalamus

Excitatory postsynaptic potentials (EPSPs) and action potentials of magnocellular neuropeptidergic cells (MNCs) in the paraventricular (PVN) and supraoptic nuclei (SON) were studied with intracellular recording in coronal slices of rat hypothalamus. The fluorescent dye Lucifer Yellow (LY) was injected intracellularly and the cells were subsequently identified as magnocellular (somata greater than 15 x 15 micrometer). These cells generally had a large cytoplasm-to-nucleus ratio. In PVN it was frequently possible to trace filled dendrites to the ependyma of the third ventricle, and occasionally dendritic spines could be seen. Electrical stimuli in areas dorsolateral and ventrolateral to the fornix column evoked EPSPs in some anatomically identified MNCs of PVN, which indicates that presynaptic fibers innervating MNCs approach PVN from this region. Short-latency (less than 1 msec) spikes could be evoked in many MNCs of PVN by stimulation near SON, which is consistent with the known projection to the neurohypophysis of many MNCs. Action potentials in MNCs of PVN and SON had significantly longer durations at one-third spike height (mean +/- S.D. = 2.06 +/- 0.6 msec) than hippocampal CA1 pyramidal cells (1.17 +/- 0.29 msec). This suggests that neuroendocrine cells in mammals and some lower vertebrates and invertebrates are similar in this regard.

Dye-marked paraventricular neuroendocrine cells in vivo in cat hypothalamus

In antidromically identified neurons in the cat hypothalamus we recorded and injected fluorescent dye-markers (Lucifer Yellow, LY; Procion Yellow, PY) intracellularly. The dye-filled neurons lay in the rostral portion of the magnocellular paraventricular nucleus of the hypothalamus. We observed two morphological cell types of similar size based on the intracellular injections of LY or PY: a bipolar cell type with fusiform perikaryon and a multipolar cell type with a polygonal perikaryon. These morphological cell types correspond to previous descriptions of immunocytochemically identified vasopressinergic and oxytocinergic magnocellular neurons in mammals. This study demonstrates the feasibility of in vivo intracellular dye-marking and electrophysiological recordings from mammalian hypothalamic neurons. We have here a basis for correlating morphological characteristics with the physiological traits of single magnocellular neuroendocrine cells.

Electrophysiological study of hepatic vagal projection to the medulla

Medullary projection of the hepatic branch of vagus was examined electrophysiologically. Analysis of compound field potentials elicited by the hepatic vagal stimulation revealed that this nerve branch projects dominantly in the left side of the medulla. Afferent neurons in the solitary tract nucleus were identified by the electrical stimulation of the nerve branch. A collision test was also employed to separate the afferent from the efferent. Some afferent neurons which innervate the liver were responsive to the portal infusion of isotonic glucose solution, hypertonic saline and water. Their response patterns were almost similar to those of the hepatic afferent nerves.

The descending afferent connections of the paraventricular nucleus of the hypothalamus (PVN)

The distribution of afferents to the paraventricular nucleus of the hypothalamus (PVN) was studied following iontophoresis of horseradish peroxidase (HRP) into the nucleus. This report describes the descending connections to this structure. The PVN receives a substantial input from limbic system structures, especially from the lateral septal nucleus and the ventral portion of the subicular cortex. A small number of labeled cells were found in the amygdala, primarily the medial nucleus. Of the circumventricular organs, the subfornical organ and OVLT both project to the PVN, the former very strongly. In both the preoptic area and hypothalamus, medial cell groups have more input than do the lateral areas. Labeled cells were found in the medial and hypothalamus, medial cell groups have more input than do the lateral areas. Labeled cells were found in the medial and lateral preoptic areas, suprachiasmatic nucleus, ventromedial nucleus, arcuate nucleus, retrochiasmatic and lateral hypothalamic areas. Of particular interest is the finding that the PVN receives input from the contralateral PVN and ipsilateral SON. Of the other diencephalic structures only the zone incerta showed a consistent number of labeled cells. The results are discussed in relationship to the possible neural structures that might mediate the response of the PVN neurons to adrenalectomy.

Intracellular recordings from the paraventricular nucleus in slices of rat hypothalamus

1. The electrical activity of thirty-five neurones in the lateral area of the paraventricular nucleus (p.v.n.) was recorded intracellularly in vitro from slices of rat hypothalamus. 2. Spontaneously occurring action potentials were observed in twenty-four of the neurones. The temporal pattern of action potentials was generally slow and irregular; occasionally some cells fired bursts of action potentials. 3. Depolarizations with a fast rising phase and slow decay occurred spontaneously in most cells. These depolarizations exhibited a wide range of amplitudes in each cell (up to 33 mV), showed temporal summation, and could serve as pre-potentials for spontaneously occurring action potentials. Presumably, these depolarizations were excitatory post-synaptic potentials (e.p.s.p.s.). 4. Depolarizing current injection could evoke action potentials. Extracellular stimuli dorsolateral to the fornix column occasionally elicited action potentials which had a short and invariant latency and which could respond to stimulation rates of 100 Hz. In some cases, extracellular stimuli in the same area evoked depolarizations which had long and variable latency and were similar to those occurring spontaneously. These two types of responses probably represent antidromic and orthodromic activation respectively. 5. Intracellular injections of horseradish peroxidase suggest that these recordings were obtained primarily, but not exclusively, from magnocellular neuroendocrine cells. This is consistent with previous anatomical studies on the location of magnocellular elements in p.v.n.

Spontaneous and reflex activity of paraventricular nucleus units in cycling and lactating rats

Extracellular recording of paraventricular neurones after electrical stimulation of the neurohypophysis reveals three cell types: type I cells, invaded by an antidromic action potential; type II cells, whose evoked response is orthodromic; and type III cells, which do not respond to stimulation. During the oestrous cycle of the rat, type I cells are the only ones in the paraventricular nucleus (PVN) whose spontaneous activity (expressed in number of action potentials per second) varies. It is significantly higher during lactation (3.6 "/- 0.6) and pro-oestrus (3.4 "/- 0.5) than during oestrus (2.0 +/- 0.5), metoestrus (1.2 +/- 0.3) and dioestrus (1.5 +/- 0.4). There are significantly more phasic units during prooestrus. The mean spontaneous firing rate of type II and III cells, remain unchanged. The reflex activity of those neurones was studied. Vaginal dilatation, which evokes a release of oxytocin31, increases the activity of a small percentage of type I cells, but there are no differences in this response at different stages of the cycle and lactation. Silent and phasic type I units are never activated. Only the evoked activity of type III cells is affected by the endocrine state. These results are discussed in the light of the known or hypothetical roles of these different cells.

A study of hypothalmic neurosecretory cells of bullfrogs in vitro

Neurosecretory cells of preoptic nuclei of bullfrogs were studied in isolated hypothalamo-hypophysial preparations under constant perfusion with oxygenated Ringer solution at 15-17 degress C. Antidromic potentials were recorded following stimuli applied to the posterior lobe of the pituitary or the stalk. 2. Intracellularly and extracellularly recorded potentials resembled those obtained in vivo from neurosecretory cells of the mammalian hypothalamus. They were unique in that the antidromic potential had a long duration (10-20 msec) and a distinct notch on the rising phase (between A and B spikes). The conduction velocity of the stalk fibres in vitro at this temperature was 0-1--0-2 m/sec. 3. When two successive stimuli were given to the posterior lobe or to the stalk separated by intervals of between 30 and 65 msec, the test (second) response showed a longer delay of the B spike. This delay between the A and B components was as long as 10 msec. Further shortening of stimulus intervals produced block of B spikes in the test response. A complete separation of A and B spikes occur spontaneously in a few instances. 4. Evidence indicated that inhibitory recurrent axon collaterals play a role in the control of bullfrog neurosecretory cells. Antidromic potentials were inhibited by a 'conditioning' stimulus for as long as 300-400 msec, even when the stimulus did not evoke an antidromic potential. 5. It was found that in addition to the inhibitory interaction there is a facilitatory recurrent axon collateral system which operates within the nuclei. The evidence for this is: (1) stimulation of the posterior lobe, with single subthreshold pulses evoked an action potential if preceded by another stimulus of subthreshold or just threshold intensity. The durations of such facilitatory effects were found to be 20--400 msec; (2) a single pulse given to the posterior lobe did occasionally evoke two spikes from neurosecretory cells; the second spike which occurred 15-30 msec after the first had the characteristics of a trans-synaptically produced potential; (3) gradual changes in the intensity of stimuli applied to the neural lobe produced a sudden shift in latencies ranging between 15 and 30 msec. The potentials having long latency also showed characteristics of those transsynaptically excited. In addition, an increase in excitability of neurosecretory cells by antidromic stimulation was confirmed by using orthodromically induced action potentials in in vivo studies. Possible functional significance of inhibitory and excitatory recurrent collateral system in neurosecretory cells was discussed. 6. Two to threefold increase in NaCl concentration of a perfusate slightly increased the latency and refractory period of antidromic potential but not the shape of the potential. Norepinephrine added to a perfusate (1 mug/ml). augmented the separation of A and B spikes of the antidromic potential. Acetylcholine at a concentration of 1 mug/ml. did not have an appreciable effect on the antidromic potential.

Electrophysiological identification of cell bodies of the tubero-infundibular neurones in the rat

1. Electrical stimulation of the median eminence induced unit responses in the suprachiasmatic, the ventromedial and the arcuate nuclei of the hypothalamus in the female rat anaesthetized with urethane.2. Arcuate units with constant latency responded in one to one fashion to repetitive stimuli up to 286 Hz.3. A conditioning stimulus just above threshold increased the threshold and latency for responses produced by a test stimulus given within 6 msec after the conditioning stimulus. Absolute refractory period was 1.4-2.0msec.4. A conditioning stimulus just below threshold did not alter the threshold and latency of the responses to test stimuli at any interval.5. A stimulus given within a time period following a spontaneously occurring spike failed to induce a unit response.6. It is concluded that the arcuate nucleus of the hypothalamus includes cell bodies of origin of the tubero-infundibular axons in the rat.

Studies of antidromically identified neurosecretory cells of the hypothalamus by intracellular and extracellular recordings

1. Neurosecretory neurones in supraoptic (SON) and paraventricular (PVN) nuclei of the hypothalamus of cats, anaesthetized with chloralose, and dogs, anaesthetized with Nembutal, were studied. These neurosecretory neurones were identified by action potentials evoked antidromically following stimulation of the posterior lobe of the pituitary gland. Reactions of 158 such neurones in cats and 228 in dogs were analysed.2. The latencies of antidromic potentials evoked in neurosecretory neurones by posterior lobe stimulation were between 10 and 22 msec for SON and between 14 and 28 msec for PVN cells. Approximate speed of conduction in the axons was 0.4-0.9 m/sec. The absolute refractory period for the soma-dendritic (SD) spike was 5-10 msec. These cells followed repetitive stimulation up to a rate of 100/sec.A notch was generally present on the rising phase of antidromic potentials and when the antidromically conducted signal fell in the relative refractory period of the preceding response, a complete separation between this first small A-spike and later large B-spikes, probably soma-dendritic spike, frequently occurred. Thus, two responses, a small and a large, sometimes appeared with more than 10 msec intervening. When the second antidromic response fell in the absolute refractory period of the first, the B-spike was blocked and only the A-spike appeared.3. Intracellular recordings from neurosecretory cells, mainly from SON in the dog, showed that these neurones possess resting membrane potentials of 50-80 mV, and action potentials of the same magnitude. In spontaneously firing neurosecretory cells separate A- and B-spikes also occurred and could be recorded intracellularly.4. Neurosecretory cells were excited by current injected intracellularly through a micro-electrode. The rheobase was 1-10 nA. A low intensity of stimulation only induced a small A-spike, but as the current was increased the full sized spike was evoked. Application of suprathreshold depolarizing current produced repetitive discharges. The intervals between spikes shortened with an increase in applied current intensity.5. There were a few neurones excited by stimulation of the posterior pituitary whose potentials did not meet the adopted criteria of antidromic potentials. These units were not classified as neurosecretory cells. The characteristics of cells giving the atypical ;antidromic potentials' were: the neurones discharged repetitively to antidromic stimulation, but with fluctuating and very long latencies.6. Neurosecretory cells in both SON and PVN were orthodromically excited by single pulse stimulations of the septal area, mid-brain reticular formation (RF), central gray, anterior commissure and hippocampus. The orthodromic responses generally consisted of two to three spikes with latencies of 10-30 msec. Excitation was followed by an inhibition, of ;spontaneous' discharges as well as of subsequent antidromic excitation, lasting 100-500 msec. Intracellular recordings from neurosecretory cells showed that stimulations of the septal area and RF produced action potentials or EPSPs of short duration followed by long lasting IPSPs. Hyperpolarization was always longer than the preceding EPSP, and its duration was generally 80 msec. Large IPSPs of 20 mV could be recorded occasionally.7. Antidromic excitation of neurosecretory cells by stimulation of the posterior pituitary was followed by the inhibition of ;spontaneous' discharges of the cells. This inhibition usually lasted for 100 msec. A corresponding IPSP was recorded during this inhibitory phase. These findings indicate the existence of recurrent collaterals in neurosecretory cells.8. This conclusion that recurrent collaterals exist was also supported by other evidence, namely, that certain neurones were found in the SON and PVN which responded to a single pulse antidromic stimulation of the posterior pituitary with five to seven discharges at a rate of between 500 and 800/sec. Weaker stimuli produced fewer spikes. Such cells resembled in their behaviour ;Renshaw cells' of the spinal cord. RF stimulation had an inhibitory effect on some of these neurones and an excitatory effect on others.9. Neurosecretory cells in the SON and PVN were excited by osmotic stimulation. Other neurones in close proximity were also found to be osmosensitive but they were either interneurones or neurosecretory cells whose axons ended in areas other than the posterior pituitary since they were not antidromically excitable.

Recurrent inhibition of antidromically identified rat supraoptic neurones

1. In anaesthetized male rats, the hypothalamus and pituitary stalk were exposed by a transpharyngeal approach. The compound field potential of the supraoptic nucleus evoked by stimulation of the pituitary stalk, was recorded with glass electrodes inserted near the origin of the anterior cerebral artery.2. The mean latency of 169 antidromically evoked action potentials isolated from the field was 9.9 msec with an extreme range of 6-26 msec. Although the wave form of the antidromic action potential showed a variety of shapes and sizes and the initial wave could be of either polarity, the majority were strikingly similar in form. The initial wave was positive with an inflexion on the rising phase and was followed by a shallow rather longer lasting negative potential.3. The antidromic nature of the action potential was confirmed when the action potential evoked at constant latency after the stimulus was observed to be cancelled by another occurring spontaneously. Although the antidromic action potentials followed stimulation frequencies greater than 100 Hz, the response to high frequency stimulation was seldom tested since the amplitude of the action potential was greatly reduced at frequencies above 30 Hz if the number of shocks exceeded a critical number, as few as 3-6 at 100 Hz.4. Stimulation of the pituitary stalk at intensities below and near threshold for antidromic invasion of the cell under study was shown by means of post-stimulus time histograms to be associated with an inhibitory period lasting on average 80 msec (S.D. = 13, N = 30).5. An increase in the intensity and duration of the inhibitory period occurred as the intensity of the stimulation was increased as might be expected if the response was mediated synaptically. The inhibitory pathway is believed to involve the recurrent collateral axons already demonstrated anatomically since the stimulation intensities necessary to produce either a marked inhibitory response or antidromic invasion of the cell in question are in most instances nearly the same.

Action potentials and release of neurohypophysial hormones in vitro

1. Isolated rat neurohypophyses were studied in vitro and the hormones released on electrical stimulation of the pituitary stalk or on exposure to excess potassium were estimated by a milk-ejection assay.2. The stalk was stimulated with trains of 500 stimuli, or multiples thereof, applied at different frequencies. Below frequencies of ca. 35 c/s, hormone release was found to depend on the total number of stimuli applied as well as on the frequency of stimulation. Above ca. 35 c/s, identical numbers of stimuli were progressively less effective as the frequency of stimulation was increased, and the dependence of the hormone output on the total number of stimuli was less apparent.3. The amplitude of the compound action potential recorded from the neurohypophysis following electrical stimulation of the stalk was found to decrease as a function of the frequency of stimulation. Stimulation at 50 c/s reduced its amplitude about sevenfold within 30 sec.4. The addition of tetrodotoxin (TTX) to the incubation media abolished the compound action potential recorded from the neural lobe as well as the release of hormones evoked by electrical stimulation. Resting release, however, was unaffected by TTX.5. In TTX-treated neural lobes, excess potassium was still effective in eliciting graded secretory responses. This indicates the independence of the release process from the action potential generating mechanism and suggests that TTX-paralysed preparations represent a useful model for the study of hormone release in the absence of conducted action potentials.6. The release of hormones from the neurohypophysis and the release of neurotransmitters at chemical synapses both depend on the entry of calcium into the nerve terminals following their depolarization by invading action potentials. In both systems, experimental separation of the release mechanism can be achieved by the use of TTX. These and other parallels suggest that the release process is similar.

Supraoptic neurosecretory cells: adrenergic and cholinergic sensitivity

Adrenergic and cholinergic agonists and antagonists were applied microelectrophoretically to over 700 neurons in the cat supraoptic nucleus, 20 percent of which were antidromically identified as neurosecretory cells. Norepinephrine uniformly depressed all sensitive cells. Acetylcholine caused both muscarinic depression and nicotinic excitation which were antagonized by atropine and dihydro-beta-erythroidine, respectively. These results support the hypothesis that norepinephrine and acetylcholine are directly involved in controlling the release of antidiuretic hormone.

Supraopticneurosecretory cells: autonomic modulation

Neurosecretory cells in the supraoptic nuclei of anesthetized cats were antidromically identified by electrical stimulation of the posterior pituitary. The responses of these units to afferent volleys from vagal and carotid sinus nerves were examined with a computer of average transients. Synaptic excitation of neurosecretory cells by stimulation of these pathways demonstrates the existence of excitatory inputs from vagal and carotid sinus nerve afferents. Since these pathways are involved in the release of antidiuretic hormone, the results support the hypothesis that its release is related to an increase in discharge frequency of supraoptic neurosecretory cells.