Progress in Neurobiology Vol. 38, pp. I to 17, 1992 Printed in Great Britain. All rights ~-served

0301-0082/92/$15.00 © 1991 Pergamon Press pig

INTRINSIC MECHANISMS INVOLVED IN THE ELECTROPHYSIOLOGICAL PROPERTIES OF THE VASOPRESSIN-RELEASING NEURONS OF THE HYPOTHALAMUS P. LEGENDRE*,~"and D . A. POULAIN~ tLaboratoire de Biologie Cellulaire, INSERM U 261, D~partment des Biotechnologies, Institut Pasteur, 28 rue du Dr. Roux 75724 Paris C~dex 15, France $Laboratoire de Neuroendocrinologie Morphofonctionelle, UniversitJ de Bordeaux II, 146 rue L~o-Saignat, 33076 Bordeaux Cbdex, France (Received 3 April 1991)

CONTENTS 1. 2. 3. 4.

Introduction Passive membrane properties of vasopressin neurons Action potentials Endogenous mechanisms underlying phasic activity 4.1. The plateau potential 4.1.1. General description 4.1.2. Mechanisms of plateau potential activation 4.2. Modulation of plateau potential activity 4.3. Ionic conductances involved in plateau potential activity 5. Voltage-dependent ionic currents from the soma of supraoptic neurons 5.1. Sodium and calcium currents 5.2. Potassium currents 5.2.1. Delayed rectifying outward potassium current 5.2.2. Transient outward potassium current 5.2.3. Calcium-dependent potassium current 5.2.4. Inward rectifying currents 6. Conclusion References

4 4 4 4

5 8

9 10 10 11 12 12 13 13 14 15

As illustrated in Fig. 1, phasic activity is a bursting activity, characterized by bursts of action potentials which last between 5 and l l5 sec, with a mean of 20-25 sec in most cells. The intervening silent periods are also variable, between 5 and 50sec, with a mean of 20-25sec (Fig. 1A,C). Scatter diagrams show no particular relationship between the duration of these two parameters and the correlation coefficient is non significant for most of the cells. It is interesting to note that, whatever the stimulus drive, the mean values for burst and silence durations are similar in all vasopressin cells, which may be indicative of a strong trend to maintain these two parameters around such values (for a detailed analysis of phasic patterns, see Poulain et al., 1988). The organization of spike activity into bursts, results in a bimodai distribution of firing rates which is typical of such activity among magnocellular neurons. On a histogram of firing rates (Fig. 1B), the first peak at zero spikes/see corresponds to the silent periods, whereas the second peak, around 5-15 spikes/see, corresponds to the mean firing rate within a burst. A notable feature of spike activity within a burst is that each burst starts with a high firing rate

I. INTRODUCTION Magnocellular vasopressin-secreting cells are located in discrete areas of the hypothalamus, the supraoptic and paraventricular nuclei. These neurons are particular in that they can be characterized by their ability to evolve a specific pattern of electrical activity in relation to physiological stimuli and hormone release. This pattern of electrical activity, described in detail by extracellular recordings under the name of phasic activity is characterized by a more or less regular succession of long lasting bursts of action potentials separated by intervals of electrical silence (Arnauld et aL, 1974; Poulain et al., 1977; for review see Poulain and Wakerley, 1982). Phasic activity occurs in response to hyperosmotic and cardiovascular stimuli and has a direct impact on the pattern of vasopressin release. Thus, the amount of hormone released per action potential is significantly enhanced during phasic activity in comparison to that released during a regular and continuous pattern of firing with the same frequency of discharge (Dutton and Dyball, 1979). *Author to whom correspondence should be addressed. n'N 3SlI--A

1

2

1

2

P. LEGENDREandD. A. POULAIN

A o~ 0

B

25 a

0

C

1800

seconds 5o

seconds

a

200

b cv=47~

b m 82

0

25_c

Sp/sec

25

c

0

msec

250

d cv=?~

cv=22~

O3 0

seconds

70 0

seconds

70 0

Sp/sec

25

0

seconds

2~

FIG. 1. Phasic activity in a supraoptic vasopressin neuron. (A) Sequential histogram of firing rates (spikes/see) recorded extracellularly /n vivo during 30 min. Note the variability of burst and silence durations, as well as the initial high firing rate at the onset of each burst. (B, a) The periodicity of firing computed by an expectation density function indicates a fairly stable periodicity despite the variability in burst duration. The distribution histogram of firing rates (B, b) displays a typical bimodal form, while the distribution of interspike intervals (B, c) is skewed towards short intervals, as a result of the clustering of spikes into bursts. (C) distribution histograms of burst durations (a), silence durations (b), and mean intraburst firing rates (c). m: mean, SD: standard deviation, CV: coefficient of variation. (C, d) mean profile of firing rates for a series of bursts revealing the initial peak at the beginning of bursts as well as the stability of the stationary phase. (From Poulain and Wakerley, 1982; with permission.) up to 15-30 spikes/sec, that lasts 1-2 sec; there is then a more steady discharge at a lower frequency with, however, a certain degree of clustering during the stationary phase (Fig. 1C). It is important to realize that vasopressin cells do not fire in a phasic pattern under normal conditions. Most of the time, under conditions of normal body fluid homeostasis, the quasi totality of vasopressin neurons have a slow irregular discharge of 1-2 spikes/sec, with a slow irregular pattern. Even then, the activity is not random, and there is a significant clustering of spikes with short interspike intervals, below 50 msec. When stimulated, vasopressin cells are progressively recruited into a phasic activity, which clearly depends not only on their afferent stimulation, but also on the level of activity of each particular neuron. Thus, practically all neurons whose mean firing rate reaches 2.5 spikes/sec display a phasic activity. In acute experiments with a powerful stimulus such as hemorrhage, the neurons are first driven into a continuous activity, and it takes about 10-15rain for the cells to evolve a phasic pattern. That phasic activity depends on afferent input is also indicated by electrical stimulation experiments, showing that stimulation of an inhibitory pathway can disrupt the phasic pattern (Poulain et al., 1980). while antidromic stimulation can either prematurely interrupt or trigger a burst, depending on the time of application (Leng, 1981). During the past years, different in vitro model have been developed to characterize the electrical

membrane properties of vasopressin neurons: slices (Andrew and Dudek, 1983; Erickson et aL, 1990), perfused explants (Bourque and Renaud, 1983) and primary cell cultures (Legendre et al., 1982; Theodosis et al., 1983). In this review, we will describe results obtained with intracellular or patch clamp recording techniques on such preparations, that allow us to understand, to a certain extent, the intrinsic mechanisms which underlie phasic bursting activity in vasopressin neurons.

2. PASSIVE MEMBRANE PROPERTIES OF VASOPRESSIN NEURONS Intracellular recordings obtained from magnocellular neurons in cell cultures, explants of slices from mice or rats show a resting membrane potential of - 6 0 m V , an input resistance in the range of 50-200 Mfl and a time constant of 9-10 msec (Legendre et al., 1982; Andrew and Dudek, 1984; Bourque, 1987; Armstrong and Smith, 1990a). In guinea pig slices, the input resistance of vasopressin neurons is higher (0.5 C ~ ) with a time constant of 14msec (Erickson et al., 1990). Recently, the electrotonic properties of supraoptic neurons were described using intracellular current clamp recordings (Armstrong and Smith, 1990b). Interestingly, voltage transient responses are described by two exponentials (Fig. 2A), which allows estimation of a dendritic electrotonic length of 1.02

VASOPKESSIN-~J.,EASING

NEURONSOF THE HYPOTHALAMUS

and of a dendritic to somatic conductance ratio of 4.1 (Armstrong and Smith, 1990b). The dendritic arborization of these neurons thus contributes significantly to their electrical behavior, by transfer of

dentritic depolarizations to the cell body and a spike trigger zone (Armstrong and Smith, 1990b). Studies of voltage-current relationships also reveal a complex behavior of the input resistance. Vaso-

A Um

lO

\

I

ovi

1.0

IVm-Vmmcl Vtmuc

to-

~ t

13.18 1~

OA

0.01

~

0.00

. 9.10

,

....

18.20

27.30

36.40

"-...

.-.'... :,.:.. , ..~, .,.,

45.50

54.60

63.70

,

....

72.80

,

....

81.90

91.00

B

Comrol

Ba lOOpM

Cs 2mM

Washout

_JO. lOnA 10mY 200am

-7

r

I

I--

J-- --1

i

FIG. 2. (A) Dendritic contribution to the electrotonic behavior of supraoptic neurons. Membrane potential voltage transients produced by a hyperpolarizing current step were computer averaged (n = 8). Vm iS the pre-pulse resting membrane potential; Vm~, the maximal deviation. Calibration bars indicate time scale and amplitude. The semi-logarithmic plots below have been obtained by calculating the proportion of the voltage deviation (Vm-VmJVm~) accounted for by each voltage point, to is an estimate of the membrane time constant tm. However, note that the voltage transient is not exactly fitted by one exponential. The difference between the fitted fine and the data points was used to calculate a second exponential with tm corresponding to the equalizing time constant of the supraoptic neuron. (From Armstrong and Smith, 1990b; with permission.) (B) Time-dependent inward rectification shown by current-clamp recording from a vasopressin cell in slice. The rectification is not blocked by Ba2+, but reversibly suppressed by Cs +. (From Erickson et al., 1990; with permission.)

4

P. LEGENDREand D. A. POULAIN

B

A I

_.j

/

nA

3( mV

10.

-/ y , -~o . ~ -~o~ 0.5

~ . o. .~ .

lnA

-20

-40

-60

FIG. 3. (A) Current-voltage relationship in a cultured vasopressin cell in presence of Cd2+ . Changes in membrane potential at the end of 100 msec constant current pulses are plotted against the current used. Actual recordings are shown in the inset. In cells showing plateau potential activity, the curve is non linear in the hyperpolarizing direction, as a result of an instantaneous inward rectification. (From Legendre et al., 1982; with permission.) (B) Steady state I - V relationship in a neuron recorded from a hypothalamic oxplant during a voltage clamped depolarizing after-potential (DAP). Data points wore measured at the peak of the DAPs (500 msec: filled circles) or after 10 sec following the onset of voltage clamping of the DAP (squares). At the peak of DAPs, there is a negative resistance in the I - V relationship (arrow), coinciding with the threshold for spike activation. (From Bourque, 1986; with permission.) pressin neurons display an instantaneous (primary cell cultures, Legendre etal., 1982) or a delayed (slices of guinea pig, Erickson et al., 1990) anomalous rectification (Figs 2B and 3A). This is observed for a membrane potential lower than - 8 0 / - 9 0 mV and reflects the presence of inward rectifier currents (see 5.2.4). On the other hand, a strong calcium-dependent inward rectification (Fig. 3B) can be evoked by depolarizing the cell at a holding potential close to the threshold for action potentials (Bourque, 1986). This suggests that a calcium-related current is strongly active during membrane potential depolarization and spike firing. Such a rectification is also observed in monolayer cultures (Legendre et al., 1982).

3. ACTION POTENTIALS Action potentials intracellularly recorded from vasopressin neurons in slices and perfused explants are underlain by a low threshold sodium current blocked by tetrodotoxin (TTX) and a high threshold calcium current (Bourque and Renaud, 1985). The calcium current appears to be responsible for a prominent shoulder on the repolarizing phase of the action potential (Fig. 4). Spike duration ranges between 1 and 4 mscc and increases with time during a burst (Bourque and Renaud, 1985). A similar phenomenon is also observed in neurohypophysial terminals and appears to be of significance to hormone release facilitation (Gainer et al., 1986). The mechanism of action potential broadening is not clearly understood. It may represent a frequencydependent modulation of potassium conductances and/or a facilitation of calcium currents. Action potentials recorded from the cell bodies are followed by a prominent calcium-dependent after-

potential hyperpolarization (Andrew and Dudek, 1984b; Andrew, 1987a). The after-potential hyperpolarization is also observed at the end of a suprathreshold depolarizing pulse and may be involved in the modulation of phasic activity. (Fig. 5A, see Section 4.2).

4. ENDOGENOUS MECHANISMS UNDERLYING PHASIC ACTIVITY One can postulate at least two mechanisms to explain the ability of vasopressin neurons to display phasic activity. On the one hand, the neurons could be considered as passively driven output neurons, transforming afferent synaptic drive into spike trains. On the other hand, the neurons could possess intrinsic nonlinear properties actively shaping spike trains. From intracellular recordings, the latter hypothesis appears most probable, without excluding the role of synaptic modulation. The variable duration of the bursts and the lack of a regular rhythmic pattern suggest that bursts of spikes are caused by an endogenous sustained depolarization. Such a long lasting depolarization, the plateau potential, was originally described in cultures of immunocytochemically identified vasopressin neurons (Legendre et aL, 1982; Theodosis e t a l . , 1983). It was also studied in slice (Andrew and Dudek, 1983, 1984a) and ¢xplant preparations (Bourque et aL, 1986). 4.1. THE PLATEAUPOTENTIAL 4.1.1. General description

Plateau potentials are an endogenous phenomenon. Their threshold is close to - 4 5 mV (Legendre

VASOPRESSIN-RELEASINONEURONSov THEHYPOTHALAMUS

5

A 4o mV 30s

1 ms

t ,

30 mV 2s

1~5 ms

FIG. 4. Action potential broadening recorded from two supraoptic neurones in slices. (A) spontaneous bursts recorded from a phasically active neuron. (B) Trains of action potentials evoked by a depolarizing current pulse. In A and B, the top trace shows the trains of action potentials; the lower panel has been obtained by superimposingaction potentials. Note the notch on the falling phase of action potentials, and the progressivespike broadening which takes place during the spontaneous bursts and the current induced train of spikes. (From Bourque and Renaud, 1985; with permission.) et al., 1988). They are evoked in an all or none manner, by excitatory synaptic inputs or by brief depolarizing pulses of current. Plateau potentials recorded from monolayer cultures reach a depolarizing level of - 2 0 / - 3 0 m V (Legendre et al., 1982) instead of - 4 5 / - 4 0 m V in other preparations (Andrew and Dudek, 1984; Bourque et aL, 1986). However, their duration and periodicity is similar in all preparations. The discrepancy in amplitude is probably due to incomplete cell maturation in cell cultures (Figs 5B and 6). In slices, plateau potentials can also be triggered by antidromic action potentials, or when evoked depolarization is above spike threshold (Andrew and Dudek, 1984a): this suggests that the threshold value of the plateau potential recorded from adult cells is closely similar to the value obtained from cultured neurons. Plateau potentials require 50-100msec to reach their peak value. In regards to the onset of action potentials, this time to peak is slow (Legendre et aL, 1982), indicating that the voltage-dependent ionic currents underlying plateau potential depolarization have slow kinetics. Plateau potentials have a duration varying from 20 to 90 sec, which is closely similar to that of the bursts of action potentials recorded extracellularly /n vivo

(Fig. 1). In cultured vasopressin neurons, the steadystate depolarization of plateau potentials is characterized by a very gradual repolarization (0,1 mV/sec) which leads eventually to a relatively rapid repolarization (2 mV/sec; Fig. 6). While usually not followed by hyperpolarization (but see Andrew, 1987a), plateau potentials are followed by a refractory period lasting 10o30 sec. Duration of the plateau potential is also voltage-dependent and decreases when membrane potential decreases (Legendre et aL, 1982; Bourque, 1986). This may be due to the increase in the holding current since the input resistance increases with time (Legendre et aL, 1982), and may reflect the voltage-dependence of the kinetics of the ionic conductances underlying the plateau potentials (Bourque, 1986). 4.1.2. Mechanisms of plateau potential activation Bursts of action potentials /n vivo are greatly variable in duration and are observed only in response to specific stimuli (see Poulain and Wakerley, 1982). Thus, although plateau potentials appear to be an endogenous phenomenon, their onset and duration is obviously modulated by extrinsic stimuli. Although a pacemaker potential was first suspected to be responsible for the periodic plateau potentials,

6

P. LEGENDREand D. A. POULAIN

A

0A nA

hA0"6 %:.

~__~__

.,~_~.L...,

..

,,

~. . . . . .

;a

, .......

CO~T,NU0US

B

%04 •

n;~

~;*=''-:.~''~

'

--~:-'ll

"

. . . . . .

_ . ~ J 30 mV 2$

cell 19

J

SO mV

Ss

FIG. 5. (A) Modification of the phasic pattern by brief suprathreshold depolarizing pulses of current. The pulses increase the burst interval duration. Note that each evoked depolarization with its superimposed action potentials is followed by a large hyperpolarization. (B) Bursts triggered by brief depolarizing current pulse recorded from a supraoptic neuron in slice. In a silent neuron, a 150 msec suprathreshold pulse (arrow) induced a sustained burst with an underlying plateau potential (p). (From Andrew and Dudek, 1982; with permission.)

A V I

.J I

B

I

............................................................................................................................................................................................................ IlnA 5sec

FIG. 6. Plateau potentials recorded from cultured vasopressin neurons. (A) Spontaneous activity showing plateau potentials and between them, slow potentials (arrows) in responses to synaptic inputs. (B) The application of repetitive hyperpolarizing pulses at IHz shows the decrease in input resistance during the plateau potential depolarization. (From Legendre e t a/., 1982; with permission.)

VASOPRESSIN-RELEASING NEURONS OF THE HYPOTHALAMUS

several experimental observations did not confirm this hypothesis. Plateau potential duration increases with membrane depolarization (Legendre et al., 1982; Andrew and Dudek, 1984b) whereas burst duration decreases in bursting pacemaker neurons (Frazier et aL, 1967; Gainer, 1972; Lcgendre et ai., 1985). Furthermore, vasopressin neurons do not possess a region of negative resistance in the steady state I / V relationship (Bourque, 1987), characteristic of pacemaker neurons (Wilson and Wachtel, 1974; Thompson and Smith, 1976; Adams and Lcvitan, 1982; Gorman et al., 1982). The issue is apparently not completely solved since plateau potential activity can be induced by tonic depolarization while post-synaptic potentials are blocked by TTX (Andrew, 1987a). However, this observation is not strong enough to conclude that plateau potential depolarizations arc driven by a pacemaker potential since regular membrane potential oscillations were never observed in the presence of TTX. Membrane potential fluctuations observed at - 5 0 m V were quite irregular (Andrew, 1987a). This instability may be due to fluctuation of some ionic conductances, such as a transient calcium current with a low threshold and a residual activation probability at a holding potential of - 5 5 / - 5 0 mV (Fox et al., 1987). Nevertheless, in a standard medium (without TTX), a pacemaker-like activity was recorded from vasopressin neurons in slices. These membrane potential Oscillations were related to post-synaptic potentials since high Mg 2+ conccn-

7

trations or TTX applications blocked both the oscillations and the post-synaptic potentials (Andrew, 1987b). This rhythmic post-synaptic potential activity suggests that there exist some neurons connected with vasopressin cells, which may display a bursting pacemaker-like activity. Indeed, bursting pacemaker neurons (not identified as vasopressin neurons) have been described in cultured explants of the hypothalamus (Gahwiler and Dreifuss, 1979), but no bursting pacemaker neurons have yet been recorded extracellularly in vivo within the area corresponding to that included in coronal slices (Andrew, 1987b). It has been proposed therefore that there may be non-spiking cells which would act as pacemakers within the supraoptic nuclei (Andrew, 1987b). Their endogenous voltage fluctuations would lead to graded neurotransmitter release (Simmers, 1981), and hence to isopcriodic fluctuations of post-synaptic activity. Plateau potential depolarizations cannot be evoked directly by short depolarizing events but need an apparent summation of depolarizing after-potentials (Andrew and Dudek, 1984b). This other endogenous potential (DAP, Fig. 7), with an amplitude of 10-15 mV and a duration of about 2 sec is evoked by post-synaptic potentials, action potentials, antidromic spikes and depolarizing pulses of current. DAPs are clearly observed between plateau potentials, and are reminiscent of slow potential depolarizations (Fig. 6) observed in primary cultures (Legendre et al., 1982). Although it was proposed that DAPs represent partially activated plateau potentials (Andrew

cell 10

u

H

H

m

H

TJs5

mv

FIG. 7. Interruption of plateau potentials by hyperpolarizing current pulses in a vasopressin neuron recorded from a slice. In these examples, the electrical activity was recorded at high gain, so that action potentials and the hyperpolarization resulting from current injection (0.15 hA) are clipped by the chart recorder. The dashed line represents the base-line resting membrane potential. (A) A brief (0.4 sec) hyperpolarizing pulse (H, left) only briefly interrupted the plateau potential, which resumed its level very rapidly. A larger pulse (H, right) ended the plateau. (B) Action potentials followed by a depolarizing after-potential (DAP, arrow) were observed during the silent period which followed the interruption of the plateau. (C) After an interrupted plateau, a slow depolarization leads to another burst of action potentials. (From Andrew and Dudek, 1982; with permission.)

8

P. LEGENDREand D. A. POULAIN

and Dudek, 1984b), they may also be a distinct phenomenon since depolarizations of an intermediate duration between that of a plateau and that of a DAP have never been observed (Andrew and Dudek, 1984b). In primary cultures, slow potentials, as plateau potentials, reach a high level of depolarization ( - 3 0 mY), and they are observed only during the refractory period (Legendre et al., 1988). Slow potentials are also all or none phenomena, triggered by exogenous stimuli and with a voltage-dependent duration. They were also described as "trigger potentials" in primary cultures since the onset of slow potentials is superimposed on the onset of plateau potentials in response to depolarizing pulses (Legendre et al., 1982). Finally, contrary to DAPs, summation of slow potential depolarizations cannot be clearly observed, since they are high enough to trigger a plateau potential (Legendre et al., 1988). 4.2. MODULATIONOF PLATEAUPOTENTIALACTIVITY The mechanisms underlying the time course and the periodicity of plateau potentials have been investigated systematically only in cultures of vasopressin neurons (Legendre et al., 1988). Depolarizing pulses of current have no effect on plateau potential duration (Legendre et al., 1982). The interruption of plateau potentials by large hyperpolarizing currents (Figs 7 and 8) was seen in cultures, in hypothalamic

A

J•'•

homv

J

,,J, ~

~__.

hnA

5sec

slices and explants (Logendr¢ et al., 1982; Andrew and Dudek, 1984b; Bourque et al., 1986). However, in cultures, the effect of hypcrpolarizing pulses applied during plateau potentials depends on the timing of application (Lcgendrc et al., 1988). During the onset of depolarization, or at the end of the slow repolarizing phase, they interrupt the plateau potential, evoking a rapid repolarization of the membrane potential with a sigmoidal time course (Fig. 8A). During the slow repolarizing phase of the plateau, they induce a large hyperpolarization followed by a slow depolarization returning to the normal level of the plateau potential depolarization. Each plateau potential is followed by a refractory period during which no other plateau potential can be evoked. When vasopressin neurons in cultures or slices exhibit spontaneous plateau potentials, the duration of the refractory period is quite variable and apparently independent of the duration of the plateau (Andrew and Dudek, 1984b; Legendre et al., 1988). For cultured vasopressin cells, suprathreshold stimuli applied immediately after the plateau repolarization trigger only isolated action potentials; later on, they induce slow potential depolarization (Fig. 8B). The minimal duration of the refractory period lasts from 5 to 10 sec. However, the time interval between two successive plateau potentials is increased by increasing the frequency of depolarizing pulses during the refractory period. Conversely, the refractory period is

homv

•~:!~L.,..~-~ Ilomv

hn,~

I

II

5sec

5sec

B

d I

~ _

I

I

,

i

[

I

II i i i , i i I

i t I ~ I I I ~ I J [ I ~ i I ] I I I I iiiiiiiIiii111~111111111t11111111!1111;;,1111111111111111111111111111

i

i

i

[

[

[

I

hnA

5sec

V

I

120mV

i

i

I

,

i

i

JL ~ . t j ~ . u ~ j ~ u

i

i

I

i

i

, hnA IOsec

FIG. 8. Effect of current pulses on plateau potential activity in cultured vasopressin neurons. (A) Plateau potentials evoked by depolarizing pulses of current, Left: control. Hyperpolarizing pulses interrupted the plateau potential when they were applied at the beginning (middle trace) or at the end (left trace) of the plateau. (B) the duration of the refractory period was increased by increasing the frequency of stimulation between two plateau potentials. Note that depolarizing pulses applied during the refractory period evoked only action potentials or slow potentials. (C) Conversely, the interruption of a plateau potential by application of a hyperpolarizing c ~ t pulse reduced the following refractory period. (From Legendre et aL, 1988; with permission.)

VASOPRESSIN-RELEASING NEURONS OF THE HYPOTHALAMUS

shortened when the preceding plateau is interrupted by application of a hyperpolarizing pulse (Legendre et al., 1988). These observations may provide an explanation for the modulation of phasic activity by various stimuli. Thus, bursts of phasic activity can be prematurely interrupted by a volley of antidromic spikes which is followed by a long lasting post-hyperpolarization (Dreifuss et al., 1976; Andrew and Dudek, 1984a), in a manner similar to the interruption of the plateau potential by the application of a pulse of hyperpolarizing current. Moreover, repetitive antidromic stimulation prolongs the refractory period between bursts (Leng, 1981). Thus, excitatory and inhibitory postsynaptic potentials may not only permit or prevent the onset of plateau potentials but also modulate their duration and refractory period (Poulain et al., 1980). 4.3.

IONIC CONDUCTANCES INVOLVED IN PLATEAU POTENTIAL ACTIVITY

The ionic conductances underlying plateau potential activity have been studied in primary cell cultures,

9

in slices and in pert'used explants, using current clamp experiments and intracellular recordings. For the sake of clarity, we will consider the various ionic conductances involved by considering in turn the onset of the plateau potential, the stationary phase, the fast repolarizing phase, and finally the refractory period (Legendre et al., 1982). The onset of plateau potentials is superimposed to the slow potentials which are both sodium and calcium dependent. During the stationary phase, characterized by a progressive repolarization, there is a decrease in the input resistance due to the involvement of several ionic conductances. Plateau potential depolarization is calcium dependent: it is insensitive to TTX, and can be reversibly suppressed by calcium current blockers such as Co 2+ or Cd 2+ (Fig. 9; Legendre et al., 1982; Bourque et al., 1986; Andrew, 1987a). Interestingly, in perfused explants, plateau potentials similar in amplitude and duration to those observed in cultures can be obtained after substitution of [Ca2+]0 by [Ba2+]0 (Fig. 10; Bourque et al., 1986). Ba e + can block calcium-dependent potassium currents and is known to enhance current through calcium channels (Connor, 1979; Hagiwara and

A

.2o+

V I

l

,,11

I

I

r-t

11hA

I 0.2 Kg(::m2 5sec

TTX

B

l

C p

I

I02K0~cn~

P

rCo+÷

3

V O __J

I Co*+

120mY

V

1021~m~. Suc

P

120mY

~ Co+*

10.2Kg/cm* 5~:

FiG. 9. (A) Evoked plateau potentials recorded from cultured vasoprcssin neurons before, during and after tetrodotoxin application (TTX, 1 mM). Note that plateau potential activity was not modified by TTX. (13, C) Plateau potentials are calcium-dependent. (B) A spontaneous plateau potential was shortened by Co2+ (10raM, applied during the depolarizing phase of the plateau. (C) Upper trace: two successive plateau potentials driven by synaptic input. Lower trace; in the same neuron, the plateau potential terminated by Co2+ was followed by hyperpolarization. (From Legendre et al., 1982; with permission.)

10

P. LEGENDREand D. A. POULAIN

A I G0mV lOs

B

10s FIG. 10. Spontaneous plateau potentials recorded from vasopressin neurons in slices after substitution of Ca2+ by Ba2+. (A) When Na + current was not blocked by TTX, action potentials were followed by an extended shoulder which increased in duration with depolarization of the membrane potential. Extreme prolongation resulted in a sustained plateau depolarization. Note the spontaneous hyperpolarizing events occuring during the plateau potential depolarization. (B) When TTX was added to the medium, the cell displayed recurring plateau potentials and slow potentials closely similar to those described in cultured vasopressin neurons. (From Bourque et al., 1986; with permission.)

Homori, 1982). However, in perfused explants, plateau potentials failed to appear during K + channel blockade by intracellular Cs + loading or by bath application of TEA (which inhibits the voltage and/or calcium-dependent K + channels; Thompson, 1977). This suggests that plateau potentials (at least in presence of Ba 2+) are underlain by a largely unopposed and slowly inactivating inward current, probably carried through Ca 2+ channels (or unspecific cationic channels) (Bourque et al., 1986). The duration of the slow repolarizing phase appears to be mainly under the control of K + conductances (Fig. 11; Legendre et al., 1982). At least two K + conductances appear to be involved since the duration of the plateau is prolonged by the application of TEA or Ba 2+, two inhibitors of the voltage and/or calcium-dependent K + channels. However, in contrast to Ba 2+, TEA application increased the depolarizing level of the plateau potential. Progressive activation of the K + conductances may repolarize the cell membrane to a critical level at which Ca 2+ channels close. The fast repolarizing phase which ends the plateau results from the closure of ionic conductances. These are probably calcium or unspecific cationic conductances since K + conductances can be ruled out in view of the lack of effect of TEA during the fast repolarization itself (Legendre et al., 1982). In primary cell cultures, the refractory period is controlled essentially by particular calcium-dependent K + conductances. The refractory period is considerably reduced in the presence of TEA or Ba 2+ and this clearly indicates that the refractory period is not due to Ca 2+ conductance inactivation (Fig. l l C ; Legendre et al., 1988). Since no hyperpolarization was observed after a plateau, the cal-

cium-dependent K + conductances involved are probably voltage-dependent, as for the potential-and Ca 2+-dependent K + current described in vertebrate sympathetic neurons (Adams et al., 1982). This may explain the lengthening of the refractory period when repetitive slow potentials are evoked during that period. The increased accumulation of intracellular Ca 2+ during repetitive depolarization may maintain to a high level the activation likelihood of the K + conductance. Conversely the shortening of the refractory period observed when a plateau potential is prematurely interrupted may result from the prevention of intracellular Ca 2+ accumulation.

5. VOLTAGE-DEPENDENT IONIC CURRENTS THE S O M A NEURONS

FROM

OF SUPRAOPTIC

Somatic voltage-dependent ionic currents of supraoptic neurons were characterized using intracellular voltage-clamp recording techniques in perfused explants and slices, or using the patch-clamp recording technique (whole cell configuration) in primary cultures. To our knowledge, no information exists about dendritic currents. Moreover, most observations were obtained without any discrimination between vasopressin or oxytocin neurons (but see Erickson eta[., 1990). 5.1. SODIUM AND CALCIUM CURRENTS

Voltage dependent Na + and Ca 2+ currents are only partially characterized, mainly because of space clamp problems encountered in this type of neuron (Armstrong and Smith, 1990b).

11

VASOPRF_,SS1N-RELEASINGNEURONSOF THE HYPOTHALAMUS

A V

~

....

~

V

120mV

,~ ' - ' ~ ' ~

120mV

2}

,

I

p

hnA

i

I

lO.2k~,cm2

P

I--1

llnA

,

I

,10.2 kO/cm2

p

hnA lo2k~cm

TEA

4)

3) ,

I

p

B

['-7

TEA

~

~

I~A ,F1

1¢2 k~..m2

TFA

5sec

t l I l i i l II! I L [ l ~ [

V

I

_la=~

~+

C v

120my

h~

[7

TEA

I-I

TEA

_la2~:,~.

81m¢

FIG. 11. Effect of potassium conductance blockers on evoked plateau potential activity. (A) AI: control; TEA (10 mM, applied before (A2) or during (A3) a plateau potential, increased its amplitude and duration. Such application had no effect during the fast repolarizing phase which terminated the plateau (A4). (B) Ba:+ (6 raM) increased the duration of the plateau but did not change its amplitude. Note that the duration of the refractory period was reduced after Ba2+ application. (C) Effect of TEA on the duration of the refractory period. Application of TEA immediately after the end of a plateau reduced dramatically the duration of the refractory period. (From Legendre et al., 1988; with permission.)

Cultured neonatal or adult supraoptic neurons possess voltage-activated Na + currents which are completely blocked by 1 mM TTX (Cobbett and Mason, 1987; Cobbett and Weiss, 1990). The Na + current has a threshold of - 59 mV, is fully activated at membrane potentials close to - 1 0 m V and is activated and inactivated within 5 msec of the onset of a depolarizing voltage step (Cobbett and Mason, 1987; Cobbett and Weiss, 1990). Two classes of Ca 2+ currents are also described using Ba2+ as the main charge carrier (Mason et al., 1988). These currents may be differentiated by their thresholds and their inactivation kinetics. One current has a low threshold ( - 4 5 mV), is fully activated at - 30 mV and shows inactivation with time during a sustained depolarizing voltage step. The second

current has a higher threshold ( - 2 0 mV), is maximum at - 1 0 mV and does not inactivate with time (Mason et al., 1988). These Ca 2+ currents probably produce the notch or shoulder seen on the falling phase of the action potential and may also be involved in the Ca 2+dependent DAP. Moreover, the non-activating Ca 2+ current may also be implicated in plateau potential depolarizations. 5.2. POTA,.~IUM CURRENTS

As stated above, K + currents are strongly implicated in the control of plateau potential activity. Voltage clamp studies of K + currents were performed in slices, in isolated explants and in cell

P. LEGENDREand D. A. POULAIN

12

cultures of the supraoptic area. Only a few K + currents are fully characterized using the whole cell recording patch clamp technique in primary cell cultures (Cobbett et al., 1989). Three types of currents were identified on the basis of their kinetics, voltage sensitivities, Ca 2+ dependence and pharmacology: a delayed current, a transient current and a calciumdependent current. 5.2. l. Delayed rectifying outward potassium current The delayed outward K ÷ current (Ik) is similar to the delayed rectifying K + current described in many neurons (see for review Hille, 1984). This current is blocked by 50% by TEA application (10 mM) and its onset develops with a sigmoidal time course (Fig. 12). The kinetics of this current are best described by three successive close states giving rise to the open state (Cobbett et al., 1989). Moreover, the tail current of lk is well described by one exponential and is voltagedependent. This suggests as a first approximation that there is only one open state for this class of channel.

The delayed potassium current decays during long lasting voltage steps ( > 15 see). The decay, due to inactivation of Gk is described by more than one exponential (Cobbett et al., 1989), which may reflect the activity of several classes of ion channels with different inactivation kinetics, or one class of channels with a complex inactivation process. The slowly inactivating I k may be activated when membrane potential is more than--30mV. This current is therefore largely responsible for the repolarizing of the action potential (Hille, 1984). It is probably also involved in the control of the plateau potential depolarization. However, since this current inactivates with time, it cannot control alone any long-lasting depolarization. 5.2.2. Transient outward potassium current The second type of potassium current (IA) observed in cultured supraoptic neurons resembles in some respects the transient outward potassium current described in invertebrate neurons (Connor and

A a

Control

500 pA

b

TI:A

500 pA

100 mV !

Vh ffi - 6 0 m V

!

50 m s

B 1-0

Im (nA)

0.5

-50

0

50

Vtest ( m Y )

FIG. ]2. Whole ceU recording of the delayed potassium current (Ik) from a cultured supraopfic neuron in calcium free medium. (Aa) Current responses (top traces) to positive voltage steps (lower traces) from

a holding potential (Vh) of --60 mV. Note the sigmoidal time course of the activation o f l k. (Ab) Outward current Obtained with the same protocol than in A, a but in the presence of 10 mM TEA. (B) I/V relationship of Ik before (open circle) and during (filled circle) TEA application. (From Cobbett et al., 1989; with permission.)

13

VASOPRESSIN-RELEASING NEURONS OF THE HYPOTHALAMUS

C

A Vh = -100 mV

Im (mA) 1.0 i

-60 mV,_.l -100 mV--"

, , 25 ms

~,,.e

] 100 mV

B

i

-60

D

0

5O

Vte=t (mY)

500 pA

] 100 mV Vh = -100 mV

I

10

ms

I

-60

0 Vtost (mV)

FiG. 13. (A) Separation of the outward transient current IA from the total K + current in calcium frcc medium. The current evoked by a voltage step to 0 mV from a holding potential of -60 mV was substracted from the current obtained by a voltage step to 0 mV from a holding potential of -100 inV. The difference represents I A. (B) IA evoked by the application of positive voltage steps after subtraction of Ik. (Vh=--100mV). I/V relationship of IA obtained from recording shown in B. (D) Plot of the conductance underlying I Aagainst the test potentials. (Vt~t). (From Cobbott et al., 1989; with permission.)

Stevens, 1971; Neher, 1971; Thompson 1977) and in some vertebrate neurons (Gustafsson et al., 1982; Galvan and Scdlmeier, 1984). Its threshold is - 7 0 mV and it reaches a maximum at +30 mV (Fig. 13). I^ is totally inactivated at a holding potential of - 60 mV. As previously described in other cell types (Thompson, 1977), kinetics analysis of this current reveals four closed states prior to the open state and at least one inactivated state. However, unlike in invertebrate cells, the activation time constant is voltage-dependent. Moreover, 4-aminopyridine (4-AP) blocks I^ in supraoptic cells, and this compound is more potent on these neurons than on snail neurons (Thompson, 1977) and on bullfrog sympathetic cells (Adams et al., 1982). The transient current recorded in cultured postnatal and adult supraoptic neurons differs in its calcium sensitivity. This current in adult neurons is completely blocked by Cd 2+ (Bourque, 1988) while IA in cultured cells is unaltered by reducing [Ca2+]0 or by Co 2+ application (Cobbett et al., 1989). However, when direct comparison is made, they appear to have similar kinetics, and it has been proposed that these two currents represent different developmental stages of the same channels (Cobbett et al., 1989). The transient potassium current is known to control the rate of firing through its effect on the rate of depolarization between two successive action potentials. (Connor, 1978; Salkoff and Wyman, 1980; Gustafsson et al., 1982). I^ may therefore control plateau potential initiation and firing frequency within a burst, maintaining spike discharge at the

relatively low frequency which characterizes vasopressin neurons. 5.2.3. Calcium-dependent potassium current In cultured supraoptic neurons, the Ca 2+-dependent potassium current recorded in the presence of 2 mM [Ca2 +]0 is suppressed by Co 2+ application or by lowering [Ca2+]0. The current does not inactivate with time and its voltage sensitivity is similar to that previously described in many cells (Thompson, 1977; Hille, 1984). I / V c u r v e for this current is typically bell shaped with a maximum at + 15 mV (Cobbett et al., 1989), which reflects the modulation of this current by an intracellular accumulation of free Ca 2+ (Fig. 14). In vasopressin neurons, Ca 2+ may enter the cell via voltage-gated Ca 2+ channels in two situations: first, during Ca 2+-dependent plateau potential depolarization; second during each action potential within a burst. Since the current does not inactivate with time, its activation by progressive intracellular accumulation of Ca 2+ during a plateau may help the gradual repolarization of the membrane to the critical value for Ca 2+ current deactivation (Legendre et ai., 1982) (see Section 4.3). 5.2.4. Inward rectifying currents An inward rectifying current was recently described in supraoptic neurons of the guinea pig (Fig. 2), using both current clamp and voltage clamp intracellular recordings (Erickson et al., 1990). The

P. LEGENDREand D. A. POULAIN

14

B

A

0.6

a

a Control

/m (nA)

b + Co 2*

500 pA

.

.

-50

] 100 m Y V~ = -30 mV

i

50 ms

.

.

.

0 50 Vtest (mY)

luO

0 50 Vtest (mY)

100

b

i

t

-50

FIG. 14. Ca2+ activated outward K + current (A) Effect of 2 mM Co2+ on outward K + current evoked by voltage steps in the presence of 2 m~Ca 2+]0. (Vh = --30 mV). (Ba) I / V relationship of total K + current before (open circle) and during (filled circle) Co2+ application. (Bb) 1/V relationship of the Ca' + activated K + component of the total K + current obtained after substraction of the curve shown in (Ba). (From Cobbett et al., 1989; with permission.)

current activates with slow kinetics at membrane potentials lower than - 7 5 m V . It is completely blocked by 2 m_M Cs application, is not suppressed by application of 0.5raM Ba 2+ and is enhanced by increasing extracellular K + concentration. The current resembles the Ih current (Brown and DiFrancesco, 1980; DiFrancesco and Ojeda, 1980; Yanagihara and Irisawa, 1980; Mayer and Westbrook, 1983; Crepel and Penit-Soria, 1986; Hayashi and Fishman, 1988; Williams et al., 1988b). Clearly, its reversal potential, close to - 5 0 mV, suggests that the channels are not selective for K + (Erickson et al., 1990). This current, observed in the cell body of vasopressin neurons may play an important role in plateau potential activity. It is known to depolarize the cell membrane at hyperpolarized levels and thus may prevent the interruption of a plateau by hyperpolarizing after-potentials which follow action potentials within a burst. Furthermore, its amplitude and its voltage dependence can be modulated by the intracellular accumulation of Ca 2+ (Hagiwara and Irisawa, 1989) and neurotransmitter release (William et al., 1988a; Rudy, 1988; Zidichouski et al., 1990; Bauer et al., 1990). The presence of instantaneous or delayed inward rectifications is also reported in supraoptic neurons, in cultures as well as in pcrfused explants (Legendre et al., 1982; Bourque, 1987). It is thus possible that supraoptic neurons possess several types of inward rectifying currents with different kinetics and ionic selectivity (Rudy, 1988), some of them being active at resting membrane potential (Bauer et al., 1990). For example, deactivation of the inward rectifying current may account for the apparent increase in the input resistance observed during plateau potential depolarization in slices (Andrew, 1987a). However, this apparent increase of ~ was interpreted as a voltagedependent closure of calcium channels which was evoked by the hyperpolarizing current pulses

(Andrew, 1987a). Nevertheless, the input resistance in these neurons increased slowly during sustained depolarization (see Fig. 8 in Andrew, 1987a). Further analysis, using voltage clamp techniques, are necessary to address this issue.

6. CONCLUSION All the observations described above clearly indicate that the phasic electrical activity of vasopressin neurons results from complex endogenous ionic mechanisms modulated by synaptic activity. The plateau potential depolarization appears to be the main endogenous mechanism underlying bursting activity. Its duration accounts for the duration of the burst of action potentials, but may be prematurely shortened by inhibitory synaptic input. The occur= fence of a plateau (which determines eventually the periodicity of the bursts) depends not only on the synaptic excitatory drive necessary to trigger each plateau, but also on other endogenous mechanisms, such as depolarizing after-potentials (DAP) or slow potentials. Furthermore, the periodicity of the plateau partly depends on the refractory period, which is under the control of voltage and Ca 2+dependent K + conductances. The refractory period, in turn, may be either prolonged by synaptic drive, which reactivates the voltage and Ca2+-dependent K + conductances, or shortened, if inhibitory synaptic inputs have prematurely interrupted the preceding plateau potential. Nevertheless, several aspects of the endogenous membrane properties of vasopressin neurons remain unclear. As in motoneurons (Kiehn, 1991) and cortical neurons (Schwindt et aL, 1988), plateau depolarizations in supraoptic neurons appear to be caused by the activation of a high threshold and slowly inactivating calcium current or an aspecific inward cationic

VASOPRESSIN-RELEASING NEURONSOF THEHYPOIMALAMUS current (Schwindt et al., 1988). In adult vasopressin neurons, the level of depolarization of plateau potentials ( - 4 0 m V ) makes unlikely such a mechanism since the high threshold and slowly inactivating calcium current (see Section 5.1) is deactivated at membrane potentials lower than - 2 0 / - 3 0 mV (Fox et aL, 1987). This is not the case in primary cell cultures of fetal hypothalami, where plateau potentials have a high level of depolarization ( - 2 0 mV) and repolarize when the membrane potential reaches a critical value ( - 3 0 mV). The large amplitude of depolarization may reflect a process of maturation. One hypothesis is that there is a low density of potassium channels in immature ceils (Bourque et al., 1986). This suggestion is not supported by experiments performed in slices, which showed that blockade of potassium currents alone did not evoke large plateau potentials (Bourque et al., 1986). Another hypothesis is that the large plateau recorded from the soma of cultured vasopressin neurons reflects the incomplete migration of the Ca 2÷ or aspecific cationic channels from the cell body to the dendrites. Dendrites make a significant contribution to the electrotonic behavior of supraoptic neurons (Armstrong and Smith, 1990b). It is thus tempting to speculate that in adult vasopressin neurons, the low amplitude plateau potentials result from the electrotonic propagation of a dendritic plateau potential depolarization similar in amplitude to the plateau potential described in cell culture. Such a spatial distribution of ionic channels was already suspected in supraoptic neurons (Mason and Leng, 1984). Finally, as mentioned in the introduction of this article, vasopressin cells do not fire in a phasic pattern under normal conditions. Vasopressin cells are progressively recruited into phasic activity when afferent stimulation increases tonically. It appears therefore that the ability to generate plateau potential activity is a conditional property of the membrane. Under stimulatory conditions, the neurotransmitters released by the various excitatory synaptic terminals (noradrenaline, serotonine, glutamate and various peptides; see for review Renaud and Bourque, 1991) trigger post-synaptic excitatory potentials, but in addition, probably induce a bistable behavior of the membrane by having a long term effect on membrane conductances. For instance, in vivo (Arnauld et al., 1983) and in vitro (Hailer and Wakerley, 1980), phasic activity is induced by prolonged application of glutamate, a neurotransmitter present in presynaptic buttons impinging on supraoptic neurons (Van Den Pol et al., 1990). In motoneurons (Kiehn, 1991), long plateau potentials are observed only during the application of neurotransmitters such as serotonine or glutamate receptors agonist as N M D A . In this model, serotonine appears to induce a bistable behavior by removing an opposing outward K ÷ current, thus allowing plateau potential depolarization (Kiehn, 1991). The development of new models such as adult dispersed cell preparations (Cobbett and Weiss, 1990), and the success of several groups to perform whole cell patch clamp recordings in adult slices (Edwards, 1989) will allow us to test these hypotheses and further clarify the various issues discussed in this review. Nevertheless, the studies which have been

15

briefly summarized here clearly indicate that the final common pathway that vasopressin neurons constitute is not passively driven by afferent input. On the contrary, the complex endogenous properties that the neurons possess reveal that vasopressin neurons are actively involved in shaping the final reponse into patterns of action potentials adapted to induce most efficiently hormone release upon reaching the terminals.

REFERENCES ADAMS,P. R., BROWN,D. A. and CONSTANt,A. (1982) M. Current and other potassium currents in bulfrog sympathetic neurons. J. Physiol., Lond. 330, 537-572 ADAMS,P. R., CONSTANTI,A., BROWN,D. A. and CLARK, R. B. (1982) IntraceUuiar Ca 2+ activates a fast voltage sensitive K + current in vertebrate sympathetic neurons. Nature, Lond. 296, 746--749. ADAMS,W. B. and L~VITAN,I. B. (1982) Voltage and ion dependence of the slow currents which mediate bursting in aplysia neurone R15. J. Physiol., Lond. 327, 157-171. ANDREW,R. D. (1987a) Endogenous bursting by rat supraoptic neuroendocrine cells is calcium-dependent. J. Physiol., Lond. 384, 451-465. ANDREW, R. D. (19878b) Isoperiodic bursting by magnocellular neuroendocrine cells in the rat hypothalamic slice. J. Physiol., Lond. 384, 467-477. ANDREW,R. D. and DUDEK,F. E. (1983) Burst discharge in mammalian neuroendocrine cells involves an intrinsic regenerative mechanism. Science 221, 1050-1052. ANDREW,R. D. and DUDEIC,F. E. (1984) Intrinsic inhibition in magnocellular neuroendocrine cells of rat hypothalamus. 3. Physiol., Lond. 353, 171-185. ANDREW, R. D. and DUDEK, F. E. (1984) Analysis of intracellularly recorded phasic bursting by mammalian neuroendocrine cells. 3". Physiol., Lond. 51, 552-566. ARMSTRONG,W. E. and S~na'H,B. N. (1990) Tuberal supraoptic neurons-I. Morphological and electrophysiological characteristics observed with intracellular recording and biocytin filling in vitro. Neuroscienee 38, 469-483. ARMSTRONG,W. E. and StaTS, B. N. (1990) Tuberal supraoptic neurons-II. Electrotonic properties. Neuroscience 38, 485-494. ARNAULD, E., VINCENT,J. D. and DREnn.Jss, J. J. (1974) Firing patterns of hypothalamic supraoptic neurons during water deprivation in monkeys. Science 250, 535-537. ARNAULD,E., CIRINO,M., LAYTON,B. S. and RENAUD,L. P. 0983) Contrasting actions of amino acids, acetylcholine, noradrenaline and leucine enkephalin on the excitability of supraoptic vasopressin-secreting neurons. Neuroendocrinology 36, 187-196. BAUER, C. K., MEYERHOF,W. and SCHWARZ,J. R. (1990) An inward-rectifying current in clonal rat pituitary cells and its modulation by thyrotrophin-releasing hormone. J. Physiol., Lond. 429, 169-189. BOURQUE, C. W. 0986) Calcium-dependent spike aftercurrent induces burst firing in magnoceUular neurosecretory cells. Neurosei. Lett. 70, 204-209. BOURQUE, C. W. (1987) Intrinsic features and control of phasic burst onset in magnocellular neurosecretory cells. Neurol. Neurobiol. 31, 387-396. BougQUE, C. W. (1988) Transient calcium-dependent potassium current in magnocellular neurosecretory cells of the rat supraoptic nucleus. J. Physiol., Lond. 397, 331-347. BOURQUE,C. W. and RENAUD,L. P. (1983) A perfused in vitro preparation of rat hypothalamus for electrophysiological studies on neurosecretory neurons. J. Neurosci. Meth. 7, 203-214.

16

P. LEGENDREand D. A. POULAIN

BOURQUE,C. W. and RENAUD,L. P. (1985) Activity dependence of action potential duration in rat supraoptic neurosecretory neurons recorded in vitro. J. Physiol., Lond. 363, 429-439. BOURQUE, C. W., BROWN, D. A. and RENAUD, L. P. (1986) Barium ions induce prolonged plateau depolarizations in neurosecretory neurons of the adult rat supraoptic nucleus. J. Physiol., Lond. 375, 573-586. BROWN, H. and DIFRANCESCO, D. (1980) Voltage-clamp investigations of membrane currents underlying pacemaker activity in rabbit sinoatrial node. J. Physiol., Lond. 308, 331-351. COBBETT, P. and MASON, W. T. (1987) Whole-cell voltage clamp recordings from cultured neurons of the supraoptic area of neonatal rat hypothalamus. Brain Res. 409, 175-180. COBBETT,P., LEGENDRE,P. and MASON,W. T. (1989) Characterization of three types of potassium current in cultured neurons of rat supraoptic nucleus area. J. Physiol., Lond. 410, 443-462. COBBETT, P. and WEISS, M. L. (1990) Voltage-clamp recordings from identified dissociated neuroendocrine cells of the adult rat supraoptic nucleus. J. Neuroendocr. 2, 267-269. CONNOR, J. A. (1979) Calcium current in molluscan neurons: measurement under condition which maximize it visibility. J. Physiol., Lond. 268, 41~0. CONNOR, J. A. (1978) Slow repetitive activity from fast conductance changes in neurons. Fed. Proc. 37, 2139-2145. CONNOR, J. A. and STEVENS, C. F. (1971) Voltage clamp studies of a transient outward membrane current in gastropod neural somata. J. PhysioL, Lond. 213, 21-30. CREPEL, F. and PEN1T-SORIA, J. (1986) Inward rectification and low threshold calcium conductance in rat cerebellar purkinje cells. An in vitro study. 3. Physiol., Lond. 372, 1-23. DIFRANCESCO, D. and OJEDA, C. (1980) Properties of the current If in the sinoatrial node of the rabbit compared with those of the current Ik2 in purkinje fibers. J. Physiol., Lond. 308, 353-367. DREIFUSS,J. J., TRIBOLLET,E., BAERTSCHI,A. J. and LINCOLN, D. W. (1976) Mammalian endocrine neurons: control of phasic activity by antidromic action potentials. Neurosci. Lett. 3, 281-286. DUTTON, A. and DYBALL, R. E. J. (1979) Phasic firing enhances vasopressin release from the rat neurohypophysis. J. Physiol., Lond. 290, 433-440. EDWARDS, F. A. , KONNERTH, A., SAKMANN, B. and TAKAHASm, T. (1989) A thin slice preparation for patch clamp recordings from neurones of the mammalian central nervous system. Pfliigers Arch. 414, 600~512. ERICKSON, K. R., RONNEKLIEV, O. K. and KELLY, M. J. (1990) Inward rectification (Ih) in immunocytochemically-identified vasopressin and oxytocin neurons of guinea-pig supraoptic nucleus. J. Endocrinol. 2, 261-265. FRAZIER,W. T., KANDEL, E. R., KUPFERMANN,I., WAZIRI,K. and COGGESHALL,R. E. (1967) Morphological and functional properties of identified neurons in the abdominal ganglion of Aplysia california. J. Neurophysiol. 30, 1288-1351. Fox, A. P., NOWYCKY,M. C. and TSmN, R. W. (1987) Kinetic and pharmacological properties distinguishing three types of calcium currents in chick sensory neurons. J. Physiol., Lond. 394, 149-172. GAINER, H. (1972) Electrophysiological behavior of an endogenously active neurosecretory cell. Brain Res. 39, 403-418. GAINER, W., WOLFE, S. A., JR, OB~dD, A. L. and SALZBERO, B. M. (1986) Action potential and frequency-dependent secretion in the mouse neurohypophysis. Neuroendocrinology 43, 557-563.

G.~d-IWlLER,B. H. and DgEWUSS,J. J. (1979) Phasically firing neurons in long-term cultures of the rat hypothalamic supraoptic area: pacemaker and follower cells. Brain Res. 177, 95-103. GALVAN, M. and SEOLMEIER,C. (1984) Outward currents in voltage-clamped rat sympathetic neurons. J. Physiol., Lond. 356, 115-134. GORMAN, A. L. F., HERMANN, A. and THOMAS, M. V. (1982) Ionic requirements for membrane oscillations and their dependence on the calcium concentration in a molluscan pace-maker neurone. J. Physiol., Lond. 327, 185~17. GUSTAFSSON, B., GALVAN, P., GRAFE, P. and WIGSTROM, H. (1982) A transient outward current in a mammalian central neurone blocked by 4-aminopyridine. Nature 299, 252-254. HAGIWARA, S. and HOMORI, H. (1982) Studies of calcium channels in rat clonal pituitary cells with patch electrode voltage clamp. J. PhysioL, Lond. 331, 231-252. HAGlWARA, S. and ImSAWA,H. (1989) Modulation by intracellular Ca: + of the hyperpolarization-activated inward current in rabbit single sinoatrial node cells. J. Physiol., Lond. 409, 121-141. HALLER, E. W. and WAKERLEY,J. B. (1980) Electrophysiological studies of paraventricular and supraoptic neurones recorded in vitro from slices of rat hypothalamus. J. Physiol., Lond. 302, 347-362. HAYASHI,H. and FISHMAN,H. i . (1988) Inward rectifier K + channel kinetics from analysis of the complex conductance of Aplysia neuronal membrane. Biophys. J. 53, 747-757. HILLE, B. (1984) lonic Channels of Excitable Membranes. Sinauer: Sunderland, MA, U.S.A. KIEHN, O. (1991) Plateau potentials and active integration in the "final common pathway" for motor behavior. TINS 14, 68-73. LEGENDRE, P., COOKE, I. M. and VINCENT, J. n . (1982) Regenerative response of long duration recorded intracellularly from dispersed cell cultures of fetal mouse hypothalamus. J. Neurophysiol. 48, 1121-I141. LEGENDRE, P., McKENZIE, J. S. and VINCENT, J. D. (1985) Evidence for bursting pacemaker neurons in cultured spinal cord cells. Neuroscience 16, 753-767. LEGENDRE P., POULAIN,D. A. and VINCENT, J. D. (1988) A study of ionic conductances involved in plateau potential activity in putative vasopressinergic neurons in primary cell culture. Brain Res. 457, 386-391. LENG, G. (1981) The effect of neural stalk stimulation upon firing patterns in rat supraoptic neurons. Exp. Brain Res. 41, 135-145. MASON,W. T. and LENG,G. (1984) Complex action potential waveform recorded from supraoptic and paraventricular neurones of the rat: evidence for sodium and calcium spike components at different membrane sites. Exp. Brain Res. 56, 135-143. MASON, W. T., COBBETT,P., INENAGA,K. and LEGENDRE, P. (1988) Ionic current in cultured supraoptie neurons: actions of peptides and transmitters, Brain Res. Bull. 20, 757-764. MAYER, M. L. and WESTBROOK, G. L. (1983) A voltageclamp analysis of inward (anomalous) rectification in mouse spinal sensory ganglion neurons. J. Physiol., Lond. 340, 19-45. NEHER, F. (1971) Two fast transient current components during voltage clamp of snail neurons. J. Gen. Physiol. 58, 36-53. POULAIN, D. A., BROWN, D. and WAKERLEY, J. M. (1988) Statistical analysis of patterns of electrical activity in vasopressin and oxytocin-secreting neurones. In: Pulsatility in Neuroendocrine Systems, pp. 119--154. Ed. LLmG,G. CRC Press: Florida.

VASOPRESglN-RELEAS1NGNEURONSOF THE HYPOTHALAMUS POUL~N, D. A., ELL~CDOIUW, F. and VINCENT, J. D. (1980) Septal connections with identified oxytocin and vasopressin neurons in the supraoptic nucleus of the rat. An electrophysiological investigation. Neuroscience 5, 379-387. POULA[N,D. A., W.~ERLEY,J. B. and DYBALL,R. E. J. (1977) Electrophysiological differentiation of oxytocin- and vasopressin-secreting neurons. Prec. R. Soc. (Lend.) 196, 367-384. POUL~aN, D. A. and W~,ERLEY, J. B. (1982) Electrophysielegy of hypothalamic magnocellular neurons secreting oxytocin and vasopressin. Neuroscience 7, 773408. I~NAUD, L. P. (1987) Magnocellular neuroendocrine neurons: update on intrinsic properties, synaptic imputs and neuropharmacology. TINS 10, 498-502. I~NAUD, L. P. and BOURQUE,C. (1991 ) Neurophysiology and neuropharmacology of hypothalamic magnoceUular neurons secreting vasopressin and oxytocin. Prog. Neurobiol. 36, 131-169. RUDY, B. (1988) Diversity and ubiquity of K + channels. Neuroscience 25, 729-749. SALKOn~,L. and WYMAN,R. (1980) Facilitation of membrane electrical excitability in Drosophila. PNAS 77, 6216-6220. SCI-IWINDT,P. C., SPAIN, W. J., FOEHRING,R. C., CHUBB, M. C. and CRILL, W. E. (1988) Slow conductances in neurons from cat sensorimotor cortex in vitro and their role in slow excitability changes. J. Neurophysiol. 59, 450-467. SIMMERS,J. A. (1981) Non-spiking interactions in crustacean rhythmic motor system. In: Neurones Without Impulses Eds. A. ROBERTSand B. M. H. BUSH.pp. 171-186. Society for Experimental Biology: U.K.

JPN38/1--B

17

Tm~ODOSlS,D. T., I_~G~,~DRE,P., VINCENT,J. D. and CooI~, I. M. (1983) Immunocytochemically identified vasopressin neurons in culture show slow, calcium-dependent electrical responses. Science 221, 1052-1054. THOMPSON, S. H. (1977) Three pharmacologically distinct potassium channels in molluscan neurons. J. Physiol., Lend. 265, 465-488. THOMPSON,S. H. and SMITH,S. J. (1976) Depolarizing afterpotentials and burst production in molluscan pace-maker neurons. J. Neurophysiol. 39, 153-161. VAN DEN POL, A., WUARIN,J. P. and DUDEK, F. E. (1990) Glutamate, the dominant excitatory transmitter in neuroendocrine regulation. Science 250, 1276-1278. WILLIAMS, J. T. COLMERS,W. F. and PAN, Z. Z. (1988a). Voltage- and ligand-activated inwardly rectifying currents in dorsal raphe neurons in vitro. J. Neurosci. g, 3499-3506. WILLIAMS,J. T., NORTH,A. and TOKIMASA,T. (1988b) Inward rectification of resting and opiate-activated potassium currents in rat locus coeruleus neurons. J. Neurosci. 8, 4299-4306. WILLIAMS, W. A. and WACHTEL, H. (1974) Negative resistance characteristics essential for the maintenance of slow oscillations in bursting neurons. Science 186, 932-934 YANAGmARA,K. and IPaSAWA, H. (1980) Inward current activated during hyperpolarization in the rabbit sinoatrial node cell. Pfliiger Arch. 385, 11-19. ZIDICHOUSKI,J. A., CHEN,H. and SMITH,P. A. (1990) Neuropeptide Y activates inwardly-rectifying K + channels in C-cells of amphibian sympathetic ganglia. Neurosci. Lett. 117, 123-128.