IA, however, appears to be different from the outward currents found mv. The time course of recovery from inactivation was complex with full

Transcription

1 Journal of Physiology (1991), 432, pp With 5 figures Printed in Great Britain A FAST, TRANSIENT K+ CURRENT IN NEUROHYPOPHYSIAL NERVE TERMINALS OF THE RAT BY PETER J. THORN*, XIAOMING WANG AND JOSE R. LEMOS From the Worcester Foundation for Experimental Biology, Neurobiology Division, 222 Maple Avenue, Shrewsbury, MA 01545, USA (Received 23 July 1990) SUMMARY 1. Nerve terminals of the rat posterior pituitary were acutely dissociated and identified using a combination of morphological and immunohistochemical techniques. Macroscopic terminal membrane currents and voltages were studied using the whole-cell patch clamp technique. 2. In physiological solutions, depolarizing voltage clamp steps, from a holding potential (-80 mv) similar to the normal terminal resting potential, elicited a fast, inward followed by a fast, transient, outward current. 3. The threshold of activation for the outward current was -60 mv. The outward current quickly reached a peak and then decayed more slowly. The decay was fitted by two exponentials with time constants of and ms. These decay constants did not show a dependence on voltage. The time to peak of the outward current decreased and the amplitude increased with increasingly depolarized potential steps. 4. The outward current was blocked by the substitution of K+ with Cs+ and its reversal potential was consistent with a potassium current. 5. The transient outward current showed steady-state inactivation at more depolarized (than -80 mv) holding potentials with 50 % inactivation occurring at mv. The time course of recovery from inactivation was complex with full recovery taking > 16 s Aminopyridine (4-AP) blocked the transient outward current in a dosedependent manner (- IC50 = 3 mm), while charybdotoxin (4 jug/ml) and tetraethylammonium (100 mm) had no effect on the current amplitude. 7. Lowering external [Ca2+] had no effect on the fast, transient outward current nor did the calcium channel blocker Cd2+ (2 mm). 8 The neurohypophysial outward current reported here corresponds most closely to IA and not to the delayed rectifier or Ca2+-activated K+ currents. Neurohypophysial IA, however, appears to be different from the outward currents found in the cell bodies in the hypothalamus which project their axons to the posterior pituitary. * Present address: University of Liverpool, Dept of Physiology, MRC Secretory Control Group, Brawnlow Hill, PO Box 147, Liverpool L69 3BX. MS 8171

2 314 P. J. THORN, X. WANG AND J. B. LEMOS 9. Under current clamp, evoked action potential duration increased (122 %) upon application of 5 mm-4-ap, indicating that IA is involved in neurohypophysial spike repolarization. 10. The existence of this current could help explain why maximal peptide release only occurs in response to bursts of electrical activity invading the nerve terminals. INTRODUCTION Neurotransmitter release occurs as a result of axonal spikes invading nerve terminals. Even at the neuromuscular junction, however, there is no simple linear relationship between the amount of acetylcholine released and the number of action potentials stimulating the motor ending (Katz, 1966). Other systems, in particular peptide-releasing neurones, maximally release transmitter only in response to specific patterns of activity (Cazalis, Dayanithi & Nordmann, 1985). In vivo, oxytocin and vasopressin are each released most effectively by distinctive, complex, bursting patterns of action potentials which are generated in somata located in the paraventricular and supraoptic nuclei of the hypothalamus and transmitted to their terminals in the neurohypophysis (Poulain & Wakerley, 1982; Renaud, Bourque, Day, Ferguson & Randle, 1985). Optical recordings of frog neurohypophysis (Salzberg, Obaid, Senseman & Gainer, 1983) and extracellular electrode recordings in rat neurohypophysis (Nordmann & Stuenkel, 1986) have demonstrated spike propagation in the nerve terminals. Bourque (1990), using intracellular recordings in rat neurohypophysis, demonstrated modulation of the incoming action potential at the nerve terminals. At stimulus frequencies of 10 Hz an action potential broadening, a decrease in latency of terminal spikes, a decrease in rate of rise of action potentials, a decrease in amplitude of spikes and finally a failure to generate spikes was observed (Bourque, 1990). All these changes represent a modulation of the incoming spike signal and, thus, the study of the ionic mechanisms underlying these changes is important in our understanding of peptide release. On the basis of somata recordings, it has been proposed that the complex interaction of a number f inward and outward currents underlie vertebrate bursting patterns of action ntial activity (Llinas, 1988). These outward currents appear to be carried primarily by potassium. Several K+ channels with different kinetic and pharmacological properties have been described in a variety of vertebrate neurones (Llina's, 1988), and the three most prevalent types were originally termed the 'delayed rectifier' (Hodgkin & Huxley, 1952), 'Ca2l-activated K+' (Meech & Standen, 1975), and 'A' (Connor & Stevens, 1971) currents. Fast transient outward potassium currents (IA) originally found in invertebrates (Hagiwara, Kusano & Saito, 1961; Connor & Stevens, 1971; Neher, 1971) have also been described in vertebrate neurones (Adams, Brown & Constanti, 1982; Gustafsson, Galvan, Grafe & Wigstrom, 1982). Currents classified as transient all show a decay in response to prolonged depolarization. However differences such as sensitivity to extracellular calcium concentrations (Mayer & Sugiyama, 1988), block by 4-aminopyridine (4-AP; Belluzzi, Sacchi & Wanke, 1985) and voltage inactivation (Zbicz & Weight, 1985), indicate a family of potassium currents. At this time, the identity of the channels which underlie patterned synaptic activity in vertebrates is not known. An important step is to characterize which ionic

3 NEUROHYPOPHYSIAL IA 315 currents are found in nerve terminals in order to understand how specific patterns of electrical activity govern transmitter release. Although electrophysiological studies have been possible on the squid giant synapse (Llinas, Steinberg & Walton, 1981; Augustine, Charlton & Smith, 1987) and crab sinus gland terminals (Lemos, Nordmann, Cooke & Stuenkel, 1986), it has been more difficult to characterize the ionic channels of vertebrate nerve terminals due to their small size and inaccessibility. In this study we took advantage of an acutely dissociated preparation of rat neurohypophysial nerve terminals (Nordmann, Cazalis, Dayanithi, Castanas, Giraud, Legros & Louis, 1986 a; Nordmann, Dayanithi & Cazalis, 1986b; Nordmann, Dayanithi & Lemos, 1987; Brethes, Dayanithi, Letellier & Nordmann, 1987) to whole-cell patch clamp the terminals and study the properties of their outward currents. Two abstracts have appeared (Thorn & Lemos, 1988; Lemos & Thorn, 1989). METHODS Preparation of isolated neurohypophysial nerve terminals Male CD rats ( g) (Charles River; Boston, MA, USA) were killed with C02 followed by decapitation. As previously described (Nordmann, Desmazes & Georgescault, 1982; Nordmann et al a, b, 1987; Cazalis, Dayanithi & Nordmann, 1987), the posterior pituitary was dissected free of pars intermedia and the anterior lobe and then homogenized in a solution containing (in mm): sucrose, 300; HEPES, 10; EGTA, 0-2; ph, 7-4. The homogenate was placed in a 35 mm dish that had been coated with poly-l-lysine (Sigma; 1 mg/to ml distilled water) overnight. The terminals stuck to the bottom of the dish within 5 min and the bathing solution was then changed to normal Locke solution (in mm): NaCl, 145; KCl, 5; MgCl2, 2; CaCl2, 2-2; glucose, 10; HEPES, 10; ph, 7 0. Using immunohistochemistry, the preparation is judged to be nearly exclusively nerve terminals (Nordmann et al. 1987; Wang, Treistman & Lemos, 1990). Any blood cells were easily recognized while pituicytes and pars intermedia cells contained visible nuclei. Electrophysiological recording Standard patch clamp recording techniques (Hamill, Marty, Neher, Sakmann & Sigworth, 1981) were used at room temperature ( ). Patch electrodes (Jencon) were double pulled (Kopf 700C) to resistances around 5 MCI, and filled with 'intracellular-like' solution: (in mm) KCl, 150; MgCl2, 2; EGTA, 10; HEPES, 10; MgATP, 2; cyclic AMP, 0-25; ph, 7-4. Whole-cell configurations were produced by gentle mouth suction from cell-attached patches of resistances greater than 10 Gfl. The currents were recorded using a List EPC-7 patch clamp amplifier and filtered at 3 khz corner frequency, -3 db, with an 8-pole Bessel filter (Frequency Devices 902LPF). Currents were digitized at between 1-25 Hz (e.g. Fig. 1A) to 20 khz (Fig. 4C). Interpulse intervals varied between 8 and 30 s. A custom program written using Clab (Indec Systems) was used to control voltage protocols and analyse the data. Low (3 pm) Ca2+ Locke solution was the same as normal above, except that it contained: (in mm) EGTA, 2 and CaCl2, 1-8. The pipette was usually filled with 'intracellular-like' solution, but in some experiments KCl was substituted with CsCl to block any potassium currents. Tetraethylammonium (TEA), tetrodotoxin (TTX), MgATP, cyclic AMP and 4-AP were also obtained from Sigma. Charybdotoxin was a kind gift of Dr Chris Miller. RESULTS Whole-cell voltage clamp recordings were made on nerve terminals acutely isolated from the rat posterior pituitary. Terminals were identified under phase contrast and interference optics by appearance, being phase bright, spherical and lacking nuclei. Immunohistochemistry, using labelled antibodies against neurophysin (Nordmann et al. 1987), showed fluorescence in morphologically identical structures. Immunoblotting techniques also verified that we recorded from

4 316 P. J. THORN, X. WANG AND J. R. LEMOS A X pa 80ms as II v M. B Time to peak (ms) 80 C Step potential (mv) Fig. 1. Kinetics of neurohypophysial terminal macroscopic outward currents. A, an example of whole-cell voltage clamp recordings from a nerve terminal. From a holding potential of -80 mv, the imposed voltage step (-60 to + 30 mv) protocol (lower) elicited a family of currents (upper). All currents, unless otherwise stated, are leak subtracted by appropriate scaling of the currents obtained during hyperpolarizing steps. An inward current is quickly followed by an outward current that increases in amplitude

5 NEUROHYPOPHYSIAL IA vasopressin- or oxytocin-containing terminals (Wang et al. 1990). A total of fortyeight terminals, ranging in diameter from 5-12 /sm, were successfully whole-cell clamped. The mean input resistance of the isolated nerve terminals was GfQ (mean+ standard error of the mean, S.E.M.; n = 5). Kinetics of outward current Under voltage clamp an inward current followed by a fast, transient, outward current were activated by depolarizing steps from a holding potential of -80 mv (Fig. IA). The threshold for activation of the outward currents was -60 mv (Figs IA and 2B). The outward current rose to a peak amplitude whose magnitude increased with more depolarizing potentials. The outward current time to peak decreased when stepping to more depolarized membrane potentials (Fig. 1B). After reaching a peak the current decayed to a steady non-zero amplitude indicating a residual, non-inactivating component of the outward current. The decay of the outward current (Fig. 1C) was fitted by the sum of two exponentials of the form: y = A +B1 e-t/tl +B2 e-t/t2, where A is a constant, B1 and B2 are amplitudes, t is time and T1 and T2 are time constants. The time constants of decay were not strongly voltage dependent (Fig. 1C), although a slight decrease in both T1 and T2 was sometimes observed at more depolarized membrane potentials. The mean time constants of decay measured at a holding potential of -80 mv with a step to a membrane potential of + 30 mv, were T, = and T2 = ms (mean±+s.e.m., n = 6; correlation coefficient, r = 0 995). Inactivation of the outward current The steady-state inactivation of the transient outward current was studied using a range of holding potentials (-100 to -20 mv), by applying a voltage step to a constant membrane potential of + 20 mv, in twelve terminals (Fig. 2A). The holding potential was applied for either 180 ms or 8 s before making the step; the results using either protocol were the same, indicating that the inactivation process took less than 180 ms.. At more depolarized holding potentials the transient outward current decreased in amplitude (Fig. 2A). The current at the end of the voltage step was measured and used as an index of the steady-state current. The steady-state plot (Fig. 2B) showed a dependence (k = 15X8 mv) on holding potential with 50% inactivation occurring at mv. No other outward current component was during more depolarized step membrane potentials. The outward currents decay to a steady, non-zero level at the end of the pulse. B, graph of time to peak plotted as a function of the step membrane potential. The relationship shows a decline with depolarizing voltage. C, the decay of the outward current was approximated by the sum of two exponentials (see Results). This figure shows the current decay at two different step membrane potentials, 0 mv (lower) and + 20 mv (upper). The continuous lines on the graph represent the fitted decays with time constants of 29-9 and 87-1 ms at 0 mv and 31-5 and 88-2 ms at + 20 mv, with correlation coefficients of and , respectively. 317

6 318 P. J. THORN, X. WANG AND J. R. LEMOS A 160 pa ms B k I, --,, i 800 E T t j ;!T, ' \i 0 *l O4!0T '5 600 c12 4_Qc 400 "X 20 U Voltage (mv) Fig. 2. Voltage dependence of terminal outward currents. A, steady-state inactivation was measured using a range of holding potentials and stepping to a constant membrane potential. In this example the range (lower) of holding potentials was between -80 and -25 mv with depolarizing steps to + 30 mv. Increasingly depolarizing holding potentials inactivate the current (upper) and reduces its peak amplitude. B, graph of the peak outward current plotted against holding potential. The peak (minus steady-state) currents were expressed as a proportion of the maximum current obtained at the most hyperpolarized holding potential (-100 mv) used in this study. The steady-state current was the amplitude of the current at the end of the voltage pulse. The points (0) are means+s.e.m. (n = 8). The line ( ) is a fit to a Boltzmann distribution of the form 1/Imax = [1 + exp (V- Vo/kJ-' Where I is the peak current measured, Im.x is the maximum peak current, V0 is the half-maximal activation voltage, V is the holding potential, and k is the slope factor. Effective valency of the gating particle was determined to be 2-4. Also included is a current-voltage (I-V) relationship of the peak outward current plotted against step membrane potential. The points (0) are the mean+s.e.m. (n = 17). The holding potential was -80 mv. The non-linear relationship shows the threshold for activation at -60 mv.

7 NEUROHYPOPHYSIAL IA 319 A 200 pa! B ) _ Time (s) Fig. 3. Recovery from inactivation. A, the time course of recovery from inactivation was measured using a two-voltage pulse protocol. Each step had a duration of 200 ms. The initial currents (first panel) obtained were superimposed for each two-pulse presentation. The time interval between the first voltage step and the second was increased from s, to 0-338, 0-464, 0-713, 1P2, 2-2, 4-2, 8-2 and 16-2 s, respectively. For illustrative purposes, only the current segments that contain the steps are shown and the interpulse intervals and the capacitative transients are omitted. B, graph of the amplitude of the second peak (minus steady-state) current expressed as a proportion of the amplitude of the first peak (minus steady-state) current. The points are the mean + S.E.M. (n = 3). The line is the sum of two negative exponentials of the form: y = c+aj(i -e(-tltl)) + a2(1 e(-t/t2)), where t is time; constants c = 0-25, a, = 0-31, a2 = 1-53 and time constants T1 = 2 s and T2 = 33.3 s. observed at depolarized holding potentials (e.g. -25 mv) where the fast, transient current was inactivated, even during pulses of long (800 ms) duration (Fig. 2A). Time course of recovery from inactivation The time course of recovery from inactivation was measured (n = 11) using a twopulse voltage protocol. From a holding potential of -80 mv, a step to a membrane potential of + 20 mv activated the outward current and this was maintained long

8 320 P. J. THORN, X. WANG AND J. R. LEMOS A 20 ms j- I B a. : a, -o E Au a. C; cu FD C Membrane potential (mv) Fig. 4. Reversal potential of the transient outward current was measured from the tail current amplitude (peak minus the steady-state current at the end of the second pulse) elicited by the second pulse in a two-pulse protocol. A, the current responses (upper) were elicited by voltage steps (lower) from a holding potential of -80 mv to a potential of + 30 mv and then stepping back to a range of potentials between -110 and -60 mv. The tail current amplitude at the second voltage step was zero at the reversal potential of the current. These currents were not leak subtracted. B, the relationship between tail current amplitude and step membrane potential measured from the experiment above

9 NEUROHYPOPHYSIAL IA enough to completely inactivate the transient component. The terminal was then returned to the holding potential for intervals of between 0-21 and 16-2 s before again stepping to a membrane potential of + 20 mv (Fig. 3A). Between each two-pulse trial there was a gap of 30 s in order to ensure complete recovery of the currents. The peak outward current at the second step was expressed as a proportion of the peak current elicited by the first step and plotted as a function of the interpulse interval (Fig. 3B). Increased interpulse interval durations elicited an outward current of progressively larger peak amplitude indicating a recovery from inactivation of the current. The time course of recovery was approximated by the sum of two negative exponentials with time constants of 2-0 and 33-3 s although full recovery was often seen by 16 s (Fig. 3B). A 214 ms return to the holding potential was able to remove a significant proportion (30%) of the inactivation. Permeability of the outward current The reversal potential of the outward current was determined using a two-voltage step protocol. From a holding potential of -80 mv the terminal was stepped to a membrane potential of + 30 mv for 10 ms to activate the outward current and then stepped back to a range of membrane potentials between -110 and -65 mv (Fig. 4A). The amplitude of the tail currents were measured 400,us after the onset of the second step. This allowed complete decay of the capacitance artifact and gives an approximate measure of the instantaneous current which was then plotted as a function of the membrane potential. There was no attempt to fit the tail currents. Their reversal potential was obtained by linear regression analysis (Fig. 4B). The mean reversal potential was mv (n = 4) which is close to the calculated Nernst equilibrium potential of -86 mv (150:5 mm-k+). Substitution of K+ with Cs+ in the pipette rapidly blocked all outward currents, after breaking through to form the whole-cell configuration, leaving a transient inward current (Fig. 4C). This current decayed to 0 within 8 ms and, thus, does not significantly contaminate measurements of the outward current. The block by Cs+ (Fig. 4 C) and the agreement between the outward current reversal potential and the terminal potassium equilibrium potential indicate that the fast, transient, outward current is potassium selective. Pharmacology of the transient outward current The transient outward current (Fig. 5A) was blocked (Fig. 5B) (n = 12) in a dosedependent manner by bath-applied 4-AP (Fig. 5C). Low concentrations (< 1 mm) of 4-AP had no effect on the outward current. A 50 % block (giving the apparent IC50) of the current was observed at a concentration of 3 mm. The total outward current, 321 was approximately linear and was fitted using linear regression to determine a reversal potential of -82 mv in this experiment. C, caesium (CsCl substituted for KCI in the pipette) blocks all the nerve terminal outward currents and isolates the inward currents (upper) elicited in response to depolarizing voltage clamp steps (lower). After breaking through into the whole-cell configuration outward currents were observed to be rapidly blocked as Cs+ entered the terminal. The holding potential was -80 mv and steps were made to between -40 and + 30 mv in 10 mv intervals. It PHY 432

10 322 P. J. THORN, X. WANG AND J. R. LEMOS A 200 pa 102 ms C 120 r 100 x 80 al) 60 B -- " M- m, o L ) log [4-API (mm),0 D E ]-20 m\/v 2 ms Fig. 5. Effect of 4-AP on the transient outward current and terminal action potentials. A, control, before addition of 4-AP, shows current (upper) responses to command step depolarizations (-60 to + 30 mv) from a holding potential of -80 mv. B, after addition of 5 mm-4-ap to the bathing solution (normal Locke solution containing 1 /tm-ttx to block the fast, inward Nal current (see Lemos & Nowycky, 1989)) the same voltage steps (lower) elicit almost no currents (upper) indicating an almost total block of the outward current (compare with A). Currents were not leak subtracted. Scale bars are for both A and B. C, dose-response curve for 4-AP effects on peak outward currents. The amplitude of the current elicited by a step to 0 mv from a holding potential of -90 mv in normal Locke solution with no added 4-AP was set to 100%. Each point includes S.E.M. (n = 4-12). D, under current clamp, repeated depolarizing steps of 75 pa lasting 10 ms in a dissociated neurohypophysial terminal elicited action potentials lasting 3-34 ms (± 0 14 ms), measured at -40 mv. Note that the terminal resting potential is -90 mv, and that the dashed line indicates 0 mv. E, after addition of 5 mm-4-ap, which inhibited IA in this terminal, the same depolarizing steps evoked action potentials with a mean duration of 7-43 ms (± 0 97 ms). including the non-inactivating component, was blocked by higher concentrations (5 mm) of 4-AP (compare Fig. 5A and B). In contrast, charybdotoxin at a concentration (4,ug/ml) shown to block Ca2+-activated K+ currents (Miller, Moczydlowski, Latorre & Phillips, 1985) did not affect any of the currents when bath applied (n = 3). TEA, at up to 100 mm, was also ineffective (n = 3). Under current clamp, action potentials could be elicited by passing depolarizing current (- 75 pa) steps of 10 ms duration into the isolated terminals (Fig. 5D). Upon

11 NEUROHYPOPHYSIAL IA addition of maximally effective concentrations (5 mm) of 4-AP, which blocked IA in these terminals (see Fig. 5C), the spike duration increased by 122% from 3-34 to 7-43 ms (Fig. 5E). Dependence of outward current on calcium In view of the calcium dependence of the fast, transient, outward current of supraoptic nuclei (SON) cell bodies (Bourque, 1988), it was of interest to study the effects of external calcium on the transient outward current of the terminals. Lowering (n = 5) the calcium concentration (3/MM) or bath applying (n = 7) the Ca2+-channel blocker Cd2+ (at 2 mm) had no effect on the amplitude or the activation/inactivation kinetics of the nerve terminal fast, transient, outward current. However, any Ca2+ effects could have been 'buffered out' (see below) because of the use of 10 mm-egta in the recording pipette. 323 DISCUSSION The results presented in this paper demonstrate the presence of a fast, transient potassium current in nerve terminals of the rat posterior pituitary. The block by internal Cs' (Fig. 4C) and the close agreement between the current's reversal (Fig. 4B) and the terminal's potassium equilibrium potentials are both evidence that the fast transient outward current is carried by K+. The failure of TEA and charybdotoxin to affect the current coupled to the block by 4-AP of all the outward current (Fig. 5) helps identify this K+ current as IA (Rogawski, 1985; Rudy, 1988). Comparison with other outward currents The neurohypophysial transient outward current is similar in its kinetics to other neuronal IAS (Rogawski, 1985). The current time to peak of 10 ms at maximal activation voltages is similar to that reported in hippocampal CA3 neurones (Gustafsson et al. 1982) although much slower than the ms reported in rat sympathetic neurones (Beluzzi et al. 1985). The current decay recorded in the terminals was described by the sum of two exponentials with time constants of 20 and 140 ms; the initial time constant is very similar to the 30 ms time constant reported by Bourque (1988) recorded in the SON cell bodies and in rat nodose ganglion (Cooper & Schrier, 1985). There are two major differences between the neurohypophysial terminal IA and currents recorded in other cells. First, the terminal IA half-maximal inactivation was mv, more depolarized than reported in other neurones which ranged from -110 mv in bull-frog sympathetic neurones (Adams et al. 1982) to -60 mv in hippocampal CA3 neurones (Zbicz & Weight, 1985). Additionally the inactivation curve for the terminal IA was much less steep (with k = 15-8) than that reported in other neurones. Secondly, the recovery from inactivation in terminals followed a complex time course. About 30 % of the current was observed to quickly recover, but subsequent recovery was best described by the sum of two exponentials with time constants of 2 and 33 s. Previous studies indicated recovery time constants of 42 ms (Belluzzi et al. 1985) and 20 ms (Galvan & Sedlmeir, 1984) in rat sympathetic 11-2

12 324 P. J. THORN, X. WANG AND J. R. LEMOS neurones and 150 ms in bull-frog sympathetic neurones (Adams et al. 1982). Most interestingly, in the SON cell bodies Bourque (1988) reported a recovery from inactivation described by the sum of two exponentials the longer of which was only 320 ms. The terminal IA was not found to be dependent on the extracellular calcium concentration, and not affected by extracellular Cd2. In contrast, in the SON cell bodies the IA was substantially reduced in a zero calcium media (Bourque, 1988). Mayer & Sugiyama (1988) demonstrated a shift in the activation/inactivation curves, which was dependent on extracellular divalent cations. This was not described by Bourque (1988). The apparent calcium independence of the terminal IA1 however, could be due to the use of the whole-cell configuration which would effectively buffer the intracellular calcium concentration through the use of EGTA in the pipette. Role of IA in nerve terminal function The whole-cell recordings from terminals in this study indicate that the neurohypophysial IA carries a substantial proportion of the outward current. Block with 4-AP abolished all outward currents and no evidence for a delayed rectifier potassium current was seen. Furthermore, the nerve terminals have a low resting membrane potential of -90 (see Fig. 5D; Lemos & Nowycky, 1989) to -60 mv (Bourque, 1988). The inactivation curve of the terminal IA indicates that a substantial proportion of the current (greater than 50 %) would be available at these resting membrane potentials. This and the spike broadening caused by 4-AP (Fig. 5E) suggest that IA is the major current responsible for terminal repolarization after a spike. Intracellular recordings in the terminals by Bourque (1990) demonstrate spike broadening and failure of propagation of spikes at high frequencies of neural stalk stimulation. The time course of recovery from inactivation in the terminal IA indicates that inactivation would build up with repeated stimulation, lead to a reduction in the available channels and a subsequent broadening of the spikes. After prolonged stimulation at high frequencies repolarization by IA would fail because of inactivation and lead to a failure of spike propagation. The time intervals of several seconds required between bursts of spikes for maximal peptide release would enable the full removal of inactivation. The IA voltage kinetics would appear to fit well with the described modulation of spikes at the nerve terminals by Bourque (1990) and could explain the effectiveness of such bursts of incoming activity in eliciting maximal release of peptide transmitters from the neurohypophysial terminals (Cazalis et al. 1985; Bondy, Gainer & Russell, 1987). Such a mechanism acting presynaptically has been suggested, furthermore, to directly modulate release of transmitter. Shimahara (1983) found in Aplysia that presynaptic hyperpolarizations decreased the amplitude of evoked synaptic potentials. This effect was attributed to activation of IA at the axon terminal. Interestingly, the squid giant synapse showed the opposite effect, presumably because that terminal lacks this current (Llinas, Steinberg & Walton, 1981). In the neurohypophysis the existence of IA assures that peptides will only be released in response to high-frequency patterns of action potentials invading the terminal and this could help explain the facilitation seen in vivo by bursting activity (Shaw, Bicknell & Dyball, 1984). Therefore the nerve terminal is not just a passive element

13 NEUROHYPOPHYSIAL IA at the end of the axon which releases chemical messengers in a 'one-to-one' fashion in response to stimulation initiated in the neuronal cell body, but is rather another active integration site helping to control synaptic transmission. We wish to thank Dr S. Treistman for helpful discussions and critical comments on the manuscript, and Dr J. Nordmann for helpful discussions. We also thank S. Gatti-Favereau for excellent technical assistance. This work was supported by grants from NIH and NSF to J. R. L. 325 REFERENCES ADAMS, P. R., BROWN, D. A. & CONSTANTI, A. (1982). M-currents and other potassium currents in bullfrog sympathetic neurones. Journal of Physiology 330, AUGUSTINE, G. J., CHARLTON, M. P. & SMITH, S. J. (1987). Calcium action in synaptic transmitter release. Annual Review of Neuroscience 10, BELLUZZI, O., SACCHI, 0. & WANKE, E. (1985). A fast transient outward current in the rat sympathetic neurone studied under voltage clamp conditions. Journal ofphysiology 358, BONDY, C. A., GAINER, H. & RUSSELL, J. T. (1987). Effects of stimulus frequency and potassium channel blockage on the secretion of vasopressin and oxytocin from the neurohypophysis. Neuroendocrinology 46, BOURQUE, C. W. (1988). Transient calcium-dependent potassium current in magnocellular neurosecretory cells of the rat supraoptic nucleus. Journal of Physiology 397, BOURQUE, C. W. (1990). Intraterminal recordings from the rat neurohypophysis in vitro. Journal of Physiology 421, BRETHES, D., DAYANITHI, G., LETELLIER, L. & NORDMANN, J. J. (1987). Depolarization-induced Ca2+ increase in isolated neurosecretory nerve terminals measured with Fura-2. Proceedings of the National Academy of Sciences of the USA 84, CAZALIS, M., DAYANITHI, G. & NORDMANN, J. J. (1985). The role of patterned burst and interburst interval on the excitation-coupling mechanism in the isolated rat neural lobe. Journal of Physiology 369, CAZALIS, M., DAYANITHI, G. & NORDMANN, J. J. (1987). Hormone release from isolated nerve endings of the rat neurohypophysis. Journal of Physiology 390, CONNOR, J. A. & STEVENS, C. F. (1971). Inward and delayed outward membrane currents in isolated neural somata under voltage clamp. Journal of Physiology 213, COOPER, E. & SCHRIER, A. (1985). Single-channel analysis of fast transient potassium currents from rat nodose neurones. Journal of Physiology 369, GALVAN, M. & SEDLMEIR, C. (1984). Outward currents in voltage clamped rat sympathetic neurones. Journal of Physiology 356, GUSTAFSSON, B., GALVAN, M., GRAFE, P. & WIGSTROM, H. (1982). A transient outward current in a mammalian central neurone blocked by 4-AP. Nature 299, HAGIWARA, S., KuSANO, K. & SAITO, N. (1961). Membrane changes of Onchidium nerve cell in potassium-rich media. Journal of Physiology 155, HAMILL, 0. P., MARTY, A., NEHER, E., SAKMANN, B. & SIGWORTH, F. (1981). Improved patchclamp techniques for high resolution current recordings from cells and cell-free membrane patches. Pflugers Archiv 391, HODGKIN, A. L. & HUXLEY, A. F. (1952). A quantitative description of membrane current and its application to conductance and excitation in nerve. Journal of Physiology 117, KATZ, B. (1966). Nerve, Muscle, and Synapse, pp McGraw-Hill Book Company, New York. LEMOS, J. R. & NORDMANN, J. J. (1986). Ionic channels and hormone release from peptidergic nerve terminals. Journal of Experimental Biology 124, LEMOS, J. R., NORDMANN, J. J., COOKE, I. M. & STUENKEL, E. L. (1986). Single channels and ionic currents in peptidergic nerve terminals. Nature 319, LEMOS, J. R. & NOWYCKY, M. C. (1989). Two types of calcium channels coexist in peptide-releasing vertebrate nerve terminals. Neuron 2, LEMOS, J. R. & THORN, P. J. (1989). Fast, transient K' current in neurohypophysial nerve terminals. Society for Neuroscience Abstracts 15, 76. LLINA'S, R. R. (1988). The intrinsic electrophysiological properties of mammalian neurons: insights into central nervous system function. Science 242,

14 326 P. J. THORN, X. WANG AND J. R. LEMOS LLINAs, R. R., STEINBERG, I. Z. & WALTON, K. (1981). Presynaptic calcium currents in squid giant synapse. Biophysical Journal 33, MAYER, M. L. & SUGIYAMA, K. (1988). A modulatory action of divalent cations on transient outward current in cultured rat sensory neurones. Journal of Physiology 396, MEECH, R. W. & STANDEN, N. B. (1975). Potassium activation in Helix aspersa neurones under voltage clamp: a component mediated by calcium influx. Journal of Physiology 249, MILLER, C., MOCZYDLOWSKI, E., LATORRE, R. & PHILLIPS, M. (1985). Charybdotoxin, a protein inhibitor of single Ca-activated K channels from mammalian smooth muscle. Nature 313, NEHER, E. (1971). Two fast transient current components during voltage clamp on snail neurons. Journal of General Physiology 58, NORDMANN, J. J., CAZALIS, M., DAYANITHI, G., CASTANAS, E., GIRAUD, P., LEGROS, J. J. & LouIs, F. (1986 a). Are opioid peptides co-localized with vasopressin or oxytocin in the neural lobe of the rat? Cell Tissue Research 246, NORDMANN, J. J., DAYANITHI, G. & CAZALIS, M. (1986b). Do opioid peptides modulate, at the level of the nerve endings, the release of neurohypophysial hormones? Experimental Brain Research 61, NORDMANN, J. J., DAYANITHI, G. & LEMOS, J. R. (1987). Isolated neurosecretory nerve endings as a tool for studying the mechanism of stimulus-secretion coupling. Bioscience Reports 7, NORDMANN, J. J., DESMAZES, J. P. & GEORGESCAULT, D. (1982). The relationship between the membrane potential of neurosecretory nerve endings, as measured by voltage-sensitive dye, and the release of neurohypophysial hormones. Neuroscience 7, NORDMANN, J. J. & STUENKEL, E. L. (1986). Electrical properties of axons and neurohypophyseal nerve terminals and their relationship to secretion in the rat. Journal ofphysiology 380, POULAIN, D. A. & WAKERLEY, J. B. (1982). Electrophysiology of hypothalamic magnocellular neurones secreting oxytocin and vasopressin. Neuroscience 7, RENAUD, L. P., BOURQUE, C. W., DAY, T. A., FERGUSON, A. V. & RANDLE, J. C. R. (1985). Electrophysiology of mammalian hypothalamic supraoptic and paraventricular neurosecretory cells. In The Electrophysiology of the Secretory Cell, ed. POISNER, A. M. & TRIFARO, J. M., pp Elsevier, New York. ROGAWSKI, M. A. (1985). The A-current: how ubiquitous a feature of excitable cells is it? Trends in Neurosciences 8, RUDY, B. (1988). Diversity and ubiquity of K channels. Neuroscience 25, SALZBERG, B. M. & OBAID, A. L. (1988). Optical studies of the secretory event at vertebrate nerve terminals. Journal of Experimental Biology 139, SALZBERG, B. M., OBAID, A. L., SENSEMAN, D. M. & GAINER, H. (1983). Optical recording of action potentials from vertebrate nerve terminals using potentiometric probes provides evidence for sodium and calcium components, Nature 306, SHAW, F. D., BICKNELL, R. J. & DYBALL, R. E. J. (1984). Facilitation of vasopressin release from the neurohypophysis by application of electrical stimuli in bursts. Neuroendocrinology 39, SHIMAHARA, T. (1983). Presynaptic modulation of transmitter release by the early outward potassium current in Aplysia. Brain Research 263, THORN, P. J. & LEMOS, J. R. (1988). Ca-activated K-channel in rat neurohypophysial nerve terminals. Society for Neuroscience Abstracts 14, 641. WANG, X., TREISTMAN, S. N. & LEMOS, J. R. (1990). Direct detection of vasopressin from individual nerve terminals of the rat neurohypophysis after whole-cell patch-clamp recording. Society for Neuroscience Abstracts 16, ZBIcz, K. L. & WEIGHT, F. F. (1985). Transient voltage and calcium-dependent outward currents in hippocampal CA3 pyramidal neurones. Journal of Neurophysiology 53,