Developmental Brain Research, 43 (1988) 273-285 Elsevier 273 BRD 50813 Electrophysiology and ultrastructure of mouse hypothalamic neurons in culture: a correlative analysis during development P. Legendre 1, A. Tixier-Vidal 2, J.L. B r i g a n t 1 a n d J . D . V i n c e n t 1 li. N. S. E. R.M. U. 176, Bordeaux (France) and 2Laboratoire de Neuroendocrinologie cellulaire et mol~culaire, U.A. 1115 CNRS, Colldge de France, Paris (France) (Accepted 7 June 1988) Key words: Hypothalamus; Dissociated cell culture; Development; Electrical activity; Extracellular patch recording; Electron microscopy The development of the electrical activity of hypothalamic neurons in dissociated cell cultures obtained from 14 day old mice foetuses was studied using patch extracellular and intracellular recording techniques. Electrophysiological data were compared with morphological observations obtained by electron microscopy. During patch recording, excitability of the cells was tested by the application of a 40 mM KC1 solution. Tetrodotoxin (TTX, 10-6 M in the delivery pipette) and Co z+ (10-2 M in the delivery pipette) were applied to the recorded cell by pressure in order to study the involvement of sodium and calcium currents in the electrical activity during the in vitro development. From the first day of incubation, TTX and Co 2+ were able to block reversibly the spontaneous electrical activity. However, only TTX application inhibited action potentials which suggests that calcium currents could be poorly involved in the action potential generation at the beginning of neuronal differentiation. Three different phases were found in the electrophysiological development of hypothalamic neurons in culture. The first phase (between the 1st and the 5th day of incubation) was characterized by an increase in the ratio of the spontaneously active cells (15% at day 1 and 90% at day 5). This increase paralleled the increase of the ratio of excitable cells. During this period no post-synaptic activity was detected. Morphologically, at 36 h, no synaptic contact was observed and growth cones were found to be very primitive. The second phase, between the 6th and the 9th day of culture, was characterized by a decrease in the ratio of spontaneously active cells and by the appearance, in a few cases, of a postsynaptic potential activity. During this phase the majority of the silent cells were excitable. At this stage neurons formed well differentiated neurites and growth cones. Synaptogenesis had already started and several stages of synapse formation could be seen. The third phase of the development, from 10 days of incubation, was characterized by an increase in post synaptic potential activity. During this period, numerous mature synapses could be observed although most of the synaptic contacts were located on neurites. In addition, some synapses were apposed onto degenerated structures. In conclusion, hypothalamic neurons in culture appear to differentiate in 3 steps: a primitive stage during which spontaneous electrical activity and excitability develop without any synaptic contact; a 2nd stage during which synaptic contacts develop, followed by a third stage of synapse maturation where mature synapses are formed whereas transient synapses degenerate. INTRODUCTION lar to that o b s e r v e d in vivo 22. In particular, some large n e u r o n s c o n t a i n i n g a m a t e r i a l i m m u n o r e a c t i v e The ability of foetal central m a m m a l i a n n e u r o n s to differentiate in p r i m a r y culture a n d to reach a ' m a ture' state of synaptic o r g a n i z a t i o n 2,2t in c o n j u n c t i o n with the direct access to cells for m o r p h o l o g i c a l control has p r o v i d e d a r e m a r k a b l e m o d e l for study of the d e v e l o p m e n t of n e u r o n a l electrophysiological properties. This is the case for h y p o t h a l a m i c n e u r o n s from foetal mice. Such cells synthetize in vitro n e u r o t r a n s mitters a n d n e u r o h o r m o n e s 16,29 a n d display after 4 weeks of i n c u b a t i o n a n d electrical activity very simi- to vasopressin a n t i b o d i e s exhibit electrical activity characterized by r e c u r r e n t p l a t e a u d e p o l a r i z a t i o n s which are related to the phasic electrical activity recorded in vivo 14,25.29. H o w e v e r , while n e u r o n - l i k e cells could be identified in the culture from the third day of i n c u b a t i o n , their electrical activity h a d n o t b e e n s t u d i e d before the age of 4 weeks, b e c a u s e of the difficulty of recording intraceilularly from a cell smaller t h a n 5 ktm in diameter. T h e i n t r o d u c t i o n of p a t c h - c l a m p Correspondence: P. Legendre, I.N.S.E.R.M.U. 176, 1 rue Camille Saint SaChs, 33077 Bordeaux Cedex, France. 0165-3806/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division) 274 techniques has allowed the recording of ionic current from small cells 2°. These techniques also allowed the recording of small external currents in the close vicinity of nerve bundles 2°, or of small cell membranes t3 with a definition closely similar to that obtained with intracellular recordings 17. Using these techniques, we have studied the spontaneous and evoked electrical activity of dissociated hypothalamic neurons developing in culture during the first 3 weeks of incubation. The observations were correlated with morphological data concerning their ultrastructural development and synaptogenesis. (35 °C) by a controlled air flux. Stable intracellular recordings were obtained from cells of 10-2(/~tm in diameter (4 weeks of incubation) with 3 M KC1 or 4 M C H 3 COOK-filled micropipettes having resistances of 50-60 MQ and 80-100 MQ respectively. Recordings were performed using a conventional bridge circuit ( D A G A N 8500) allowing current stimulation and voltage recording with the same microelectrode. The membrane potential behavior and the current of stimulation were recorded on tape (Racat TDI) and photographed from a digital oscilloscope (Tectronix 5D 10). Patch recordings MATERIALS AND METHODS Cell cultures The procedures used for preparing dissociated cell cultures of mouse hypothalamus are adapted from those of Benda et al. 2. Briefly, foetuses were removed from IOPS/OF1 mice on the 14th day of gestation. Foetal hypothalami were dissociated mechanically and the cell suspension was dispersed into 35 mm sterile Petri dishes (Lux) at approximately 106 cells per dish. The nutriment medium consisted of H A M F 12 (Gibco) supplemented with 20% fetal calf serum (IBF), 0.6% glucose and adjusted at pH 7.3, 330 mOsm. Cultures were incubated at 37 °C in an atmosphere of 95% air and 5% CO 2. The medium was not removed during the first 3 days and then was changed twice a week. The patch-clamp amplifier (LIST EPC7), especially designed to detect small ionic current through membrane patches was used as a sensitive currentvoltage converter to record external currents during spontaneous or evoked electrical activity. Electrodes (2-5 Mr2 resistance) were filled with the external recording medium. The leakage through the tip of the electrode was diminished after suction and the seal input resistance increased to a final value ranging from 20 to 60 Mr2 t3 independently of the age ol the culture. Signals were photographed from a digital oscilloscope (Tectronix 5110 5DI0) and were stored on FM tape (Racal TDI). They were filtered at 2 kHz (basel filter 8 poles) before recording on a polygraph (Gould 240(I). The noise level generally did not exceed t - 3 pA. Drug application Ultrastructural techniques Fixation, embedding and sectioning of selected area were performed as previously described 2'27. Small areas, in horizontal sections, were selected according to the same phase contrast criteria as for electrophysiological recordings. Intracellular recordings Intracellular recordings were made under direct visualization using an inverted phase microscope (200x magnification). The preparation was continuously perfused using gravity field with a balanced salt solution consisting of Hank's medium (Gibco), supplemented with 1% glucose, 3 mM CaCI 2 (final concentration) buffered at pH 7.1 with 10 mM HEPES (Sigma) and maintained at constant temperature Tetrodotoxin (TTX, 10-~ M, Sigma), CoCI: (5 mM) and KCI (30 mM) were diluted in the recording medium and were locally applied by pressure through multi-barrelled micropipettes using micropumps (Medical Instrument BH2). The ejection pressure was recorded on the tape recorder. In some experiments, medium alone was applied from one barrel to rule out pressure ejection artifacts. RESULTS Morphology and ultrastructure Phase contrast observation. Using phase contrast microscopy, 24 h after plating, the cultures appeared as numerous clusters of 10-30 cells of small size (<5 ~m; Fig. 1A). The celt clusters were directly attached 275 Fig. 1. Phase contrast photomicrographs of hypothalamic cell cultures at various times of incubation (A) 24 h; (B) 6 days; (C) 21 days. Bar = 20tim. 276 Fig. 2. Electron microscope pictures of sections performed in small aggregates attached to the plastic surface at 36 h. a: presumptive neuron as identified by the presence of rosettes of ribosomes, the development of the Golgi zone (G) which displays a few dense core vesicles (---~). b: at the periphery of the aggregate, one can see transversal and longitudinal profiles of neurites, some of them containing dense core vesicles (--,). *. A presumptive primitive growth cone. Bar = 5 0 t i m 277 Fig. 3. Electron microscope pictures of horizontal sections performed in a small area of a culture taken at the 6th day in vitro. Different steps of synapse formation are illustrated, a: a growth cone (GC) initiates a synaptic contact with a neurite (N) as revealed by the presence of a few synaptic-like vesicles apposed to a presynaptic m e m b r a n e thickening (---~) which is associated to a coated pit. Note the presence in the growth cone of s m o o t h reticulum and dense core vesicles, b: a further step of synapse differentiation is illustrated. O n e notices the n u m b e r of synaptic vesicles, the differentiation of the presynaptic and postsynaptic membranes. However one can see also some structures characteristic of growth cones such as smooth reticulum and dispersed clear vesicles, c: a synapse displays a dense accumulation of synaptic vesicles reminiscent of a pre-degenerative stage. Bar = 50tim. 278 to the plastic dish. After 48 h of incubation small cells with ovoid perikaryon migrated from the clusters on a limited basal layer of non-refringent flat cells. These small cells were either bipolar or multipolar and some thick processes arising from the clusters began to be observed. During the first week of incubation, numerous cells migrated from the clusters. After a week of incubatiom neurons still had a grey cell body of various forms from which arose several short processes. At this stage of culture, the flat cells had proliferated and a continuous layer was observed (Fig. 1B). From the first to the second week of incubation, the number of intermingled fibers increased and a complex network was seen. The cell body of the neurons increased in size and became refringent (Fig. 1C). After 3 weeks of culture the perikarya of the neurons reached their maximal size and remained constant after this age. At this stage, the neurons consisted either of small bipolar cells (5-10 #m diameter), or larger and more complex multipolar cells (15-25 #m diameter). Some of these large cells corn tained a vasopressin-like material as previously described by immunohistochemical methods zS, Ultrastructural correlates of electrical development Culture dishes were examined at 3 selected steps of neuronal development in culture: 36 h, 6 days and 13 days. The choice of these steps was based on the electrophysiological data. After in situ embedding of the monolayers, small areas containing neuron-like cells were selected according to the same phase contrast features as for electrophysiological recordings. Thirty-six hours. Aggregates examined in 36 h culture consisted in small neuro-epithelial cells with a high nuclear/cytoplasm ratio. Some of them displayed a more developed and a clearer cytoplasm than the others; these cells possessed typical rosettes of ribosomes and a conspicuous Golgi zone with some small dense core vesicles (Fig. 2a). In the core and at the surface of the aggregate, transversal and irregularly shaped profiles of cytoplasmic extension formed a loose network. Very few elongated cytoplasmic extensions were observed. Some of these contained longitudinal profiles of microtubules and dense core vesicles (Fig. 2b). Growth cones were very primitive, containing only large irregular vacuoles and were devoid of smooth reticulum. No syn- apses were found at this stage. Six days. At this age of culture (Fig. 3), the neuronal perikarya exhibited an increase in the cytoplasmic volume and an enlargement of the Golgi zone. They formed well differentiated neurites with bundles of microtubules and a network of smooth endoplasmic reticulum which was particularly apparent in varicosities and growth cones (Fig. 3a). In addition, varicosities and growth cones displayed numerous electron lucent as well as dense core vesicles. Synaptogenesis had already started and several stages of synapse formation could be easily seen (Fig. 3a-c). These contacts were identified by the differentiation of the presynaptic and postsynaptic densities and by the disymmetric opposition of electron lucent vesicles to the presynaptic membrane: The diameter and number of these vesicles varied greatly from one contact to another indicating several steps of synapse differentiation. In a few cases, a compact organization of synaptic vesicles was even observed (Fig. 3c). The postsynaptic-like component was either a narrow neurite or a varicosity. Thirteen days. The main difference between 6-dayold and 13-day-old culture resided in the develop, ment of mature synapses as shown by the number and the organization of synaptic vesicles and their relationship with the presynaptic membrane (Fig. 4a,b). Terminal boutons also contained large dense core vesicles which were generally located rather far from the presynaptic membrane, Synaptic boutons were generally apposed onto neurites of various diameter (axon-like or dendrite-like) or onto varicosities. Synapses on dendritic spines or on neuronal soma were not found. In addition, synapses apposed onto a degenerated structure were often found at this stage, suggesting the occurrence of neuronal death. They were characterized by the accumulation of densely packed synaptic vesicules (Fig. 4c,d). General electrical activity" Spontaneous electrical activity was recorded extracellutarly as from the first day of culture. Two different types of spontaneous electrical activity were detected extracellulafly. The first type can be described as a complex wave form with two successive peaks. The second peak was either positive or negative, depending on the recorded cell (Fig. 5A). The maximal amplitude of 279 Fig. 4. Electron microscope pictures of horizontal sections performed in small areas of cultures taken at the 13th day in vitro, a and b illustrate two fully differentiated synapses as revealed by the great number of synaptic vesicles, the presence of large dense core vesicles outside the synaptic vesicle clusters and the differentiation of the presynaptic and postsynaptic membranes (compare to Fig. 3). c and d show two examples of degenerating synapses. One notice on the presynaptic side the presence of irregular small vesicles (----~). In contrast the post-synaptic side is greatly dilated and rather empty of intracellular membranes ('~r). Bar = 50/~m. 280 ! A l t ! 7 .,.........j i ,t,.,......_. 150pA i ,¢, -Jr, :1 t l / J 25mV ~'~'~'~_] 25 mV 2ms 5ms Fig. 5. Oscilloscope traces of simultaneous extracellular (upper trace) and intracellular (down trace) recordings of action potential activity from the same cell, in 28-day-old cell culture. A: examples of two types of action potential extracellularly recorded. Left: biphasic waveform with positive and negative peaks. Right: positive waveform, composed of two successive positive peaks. Note the lack of correlation existing between the shapes of the intracellularly and extracellularly recorded action potentials. B: examples of repetitive action potential firing. these signals ranged between 40 and 300 p A depending on the age of the culture. Maximal duration did not exceed 3 ms and was independent on the state of. neurone maturation. Since the shape of extracellular signals did not change with the age of the culture, simultaneous intracellular and extracellular recordings were performed on large cells from 25- to 30-day-old cultures (n = 21) in order to correlate the extracellularly recorded events with m e m b r a n e potential changes. We thus observed that this type of signal recorded extracellularly with patch electrode corresponded to action potentials recorded intracellularly on the same cell (Fig. 5A,B). However, no correlation was found between the shape of the extracellular signals and the shape of action potentials intracellularly recorded, even during repetitive firing (Fig. 5B). The second type of electrical activity detected with patch electrode appeared as a simple deviation in either a positive or a negative direction (Fig. 6). These events were of low amplitude (5-10 pA) but of greater duration (4-15 ms) than the action potentials recorded with patch electrode. With double recording techniques (n -- 21), in 25- to 30-day-old culture, such electrical activity was found to represent postsynaptic potential. Excitatory and inhibitoty postsynaptic potentials observed with intracellular recording corresponded respectively to extracellularly recorded positive and negative electrical events (Fig. 6A,B), Inversion in polarity between intracellularly and extracellularly recorded synaptic events were never observed. Effect of sodium and calcium current blockers on developing electrical activity Since it was not possible electrically to stimulate cells by patch electrode because of the low input resistance of the seal, a solution of 30 m M K + was applied to the surface of the cell to investigate its excitability. Such applications evoked a burst of action potentials on spontaneously active cells (Fig. 7A) and on some silent cells (Fig. 7B). Using micropressure application, the effects of T I ' X (10 -6 M) and CoCI 2 (5 raM) on spontaneous or evoked electrical activity of neurons were studied in A B it d i , i;(r i li o'~i ' -*'--,~--.~- __Jlomv ' 20ms !i } 20ms Fig. 6. Oscilloscope traces of excitatory (A) and inhibitory ~(B) postsynaptic potential activities extracellularly (upper trace) and intracellularly (down trace) recorded (28-day-old culture). 281 A , 1,, 'LJ 13 KCI m _-.. Kci r-I .- _- l ll,t 1 . .- ~ I ,50pA 5s¢c • 1 i 2 5 i IOOmsec A 5sec / [! i p L ...... .....l IOOmsec Fig. 7. Burst of action potentials evoked by pressure application of a 30 mM KCI solution onto a spontaneously active neuron (5-dayold-culture) (A) and onto a silent neuron (9-day-old culture) (B). Note that during evoked spike discharge, the amplitude of the action potential decreases with time. 1-, 4-, 6-, 10- and 21-day-old culture. At these ages, brief application of TTX on the surface of a spontaneously active cell immediately suppressed electrical activity for 20-60 s (Fig. 8A). At the beginning of the recovery period, spontaneous action potentials reappeared with two distinct modifications (Fig. 8B): the amplitude of the extracellular signal was reduced and the frequency of firing was greatly affected. Complete restauration of the initial amplitude and frequency occurred only after a few minutes. During TTX effect, the application of 30 mM K + failed to evoke action potentials (Fig. 8A) which suggests that a TTX-sensitive sodium current is indispensable for action potential generation as early as the first day of incubation. Cobalt application (5 mM) suppressed spontaneous firing for a short period of time (10-30 s) as from the first day of culture (Fig. 8B). When postsynaptic potentials were recorded, they were also suppressed for a similar period of time, but contrary to the TTX effect, the action potentials reappeared at the beginning of the recovery period with the same shape and the same amplitude as those observed before cobalt application. The application of 30 mM K + was able to evoke a burst of action potentials during the cobalt-induced suppression of spontaneous electrical activity (Fig. 8B). These important differences between the effect of TTX and cobalt on the electrical activity of the developing hypothalamic cell in culture suggest that the involvement of calcium currents in action potential generation is considerably reduced in comparison to that of sodium currents. Evolution of the electrical activity with the age of the culture The development of the electrical activity was followed during the first 3 weeks of culture. The percentage of excitable cells, spontaneously active cells and cells presenting synaptic activity were estimated on four different culture dishes per age studied. During each experiment, 20 cells or more were recorded. The results are summarized in Fig. 9. At 24 h of incubation, 15% of the cells tested were found to be spontaneously active (Fig. 9A). Action potentials were of low amplitude (40.2 pA + 4.8; n = 28) and occurred with a low frequency. No regular rhythmic activity was observed. During the first 5 days of culture, the percentage of spontaneously active cells increased exponentially to reach 95% of the cells tested at the end of this stage. The firing frequency increased but no post-synaptic potentials were observed (Fig. 9A). Throughout this period no significant difference was found between the percentage of spontaneously active cells and of excitable cells (excitability was detected by the application of the K + solution). This suggests that silent cells were not excitable at the beginning of the differentiation in culture or that ionic currents were too low to be detected (Fig. 9B). In order to determine if the spontaneous electrical activity observed in absence of synaptic inputs was related to electrotonic coupling, 31 pairs of ceils were recorded with two patch electrodes at day 1 (n = 12), day 2 (n = 8), day 4 (n = 6) and day 7 (n = 5). Generally, in cultures aged from 1 to 4 days, one of the two 282 recorded cells was silent (63%) and, when the two cells were spontaneously active no evidence of synchronization was found. Between the 5th and the 9th day of incubation, the percentage of spontaneously active cells decreased dramatically to reach at the end of this stage 15~i of the cells tested. The firing frequency also decreased but a majority of neurons remained excitable ( 8 0 ~ at day 9). The spontaneous activity was completely restored A TTX KCI n r"q after the 10th day of incubation and remained constant until the 3rd week of culture Postsynaptic potential activit~ began to be observed after a week. It increased slowly during the second and the third week of incubation and was observed in 50% of the cells tested at the 21st day (Fig. 9C). During development, the amplitude of postsynaptic potentials remained constant (40 pA _+_5:,,7 = 60). The majority of synaptic events were excitatory. Negative signals were only significantly observed af- Ji i , ,IzspA IOsec j__,25pA 'I 25msec B CoCI 2 I I 25rnsec KCl I r-1 ,, . . . . . . : 12spa ~L .... IOsec 25msec IOOmsec 25msec Fig. 8. A: effect of T T X app[ication (10 -~ M) on spontaneous electrical activity (4-day-old culture). Under these conditions, KCt appli: cation failed to evoke a burst of spikes, and action potentials were of lower amplitude at the beginning of the recovery period, b: effect of CoCI2 application (5 raM) on spontaneous electrical activity (6-day-old culture): Note that KCI application was still able :to evoke a burst of action potentials during CoC12application. ] , .~43 ter the 13th day and were recorded in 50% of cells presenting a synaptic activity after three weeks of incubation. I00 Spon!oneous octive cells 50 DISCUSSION Since the patch electrode does not damage the cell membrane because of the low seal resistance (the cell can be recorded twice or more with different electrodes), there is good reason to suppose that electrical activity recorded in hypothalamic neurons during development in culture reflect the on-going electrical activity. This electrophysiological approach was first developed on spinal cord cell culture L3and the parallel electrophysioiogical and neurochemical studies ascertained the role of electrical activity in neuronal development 3,11. In this study we have thus described 3 different stages during development of electrophysiological activity in the culture: a first stage (1-5 days) during which only spontaneous action potential activity develops; a second stage, between the 5th and the 9th day characterized by a decrease in the spontaneous action potential activity, while cells remain excitable. and a third stage during which postsynaptic electrical activity develops. As already observed for mouse foetal hypothalamic cells taken on the 16th day of gestation and grown in serum-supplemented medium 2'2~''2s or in serumfree medium 27, our cell cultures derived from 14-dayold embryos displayed fully differentiated synapses after 10 days of incubation. However, formation of synaptic contacts started around the 6th day of incubation, which may explain that no postsynaptic potentials could be recorded before a week of culture. On the one hand, these morphological data rule out the hypothesis of a synaptically driven spontaneous electrical activity at the beginning of cell differentiation in culture. On the other hand, no evidence of electrotonic coupling has been observed. Since spontaneous, but not evoked action potentials were blocked by cobalt, we may therefore speculate that such electrical activity at the beginning of the culture is linked to a calcium-dependent release of excitatory compound from growth cones. Growth cones of embryonic chick ciliary ganglion neurons ~2 and Xenopus embryonic neurons 3~ spontaneously release excitatory neurotransmitter (acetyl- ,. . . . I . . . . . 3 . . . . . 5 . . . . . . 7 . . . . . . . .// II 9 . . . . 20 13 t00 I 50 o~ . . . . . . . . . . . . . . . . . . . . . I 3 5 7 9 [I It; . 20 13 I00 PSPs o} L~ 50 i l ~ . i,/T l/// I 3 5 7 9 II I3 20 Age (days) Fig. 9. Evolution of evoked or spontaneous electrical activity with the age of the culture. A: evolution of the proportion of excitable cells. B: evolution of the proportion of cells displaying spontaneous action potential activity. C: evolution of the proportion of cells displaying postsynaptic electrical activity. Each point represents the mean of a experiments. Evolution of the electrical activity with culture age was analysed for 20 cells per experiment. Note that postsynaptic electrical activity became apparent during the transitory decrease in the proportion of cells displaying spontaneous action potential activity. choline) in the absence of a target. This 'hormonal" release was found to be concomitant with neurite extension 7"9'24, which corresponds to the structural feature of the first development stage in our culture. Similarly, it was recently shown that mouse hypothalamic neurons grown in serum-free medium spontaneously release [3H]dopamine but do not respond to K+-induced depolarization before 5 days in culture2< A comparable leakage of excitatory substances may be sufficient to bring the cell membrane potential close to action potential threshold. In primitive neurons, the input resistance is highl'< and such conditions are optimal for attainment of a high depolarization with low change in ionic conductances. At this age of the culture, calcium plays no role in action potential generation and cobalt action might 284 be indirect, through the postulated calcium-dependent release of excitatory substances. However, since with our recording techniques, the slow membrane potential fluctuations cannot be recorded, we cannot exclude an artefactual effect of cobalt application such as a non-detectable membrane potential hyperpolarization or a screening effect of divalent cations on the surface of the membrane ~°'15. Spontaneous firing was observed in only 15% of the cells tested at day 1. This ratio then increased during the first 5 days of incubation. Since no difference was found between the number of spontaneously active cells and that of excitable cells, it is possible that this small ratio of excitable cells, observed on the first day, reflects a mixture of recorded glial precursor cells and neuronal precursors, because of impossibility to distinguish them by phase contrast microscopy. However, it is likely that the ratio observed after 4 days of incubation (35%) correspond to recorded neurons which then become identifiable by their neurites (see Fig. 1). In this way, it is possible that this phenomenon reflects a maturation of the neuronal membrane and especially the development of sodium channels 24,3°. In support of this hypothesis, TTX is able to block action potentials reversibly from the first day of culture, and at this age, all recorded cells possess sodium channels, albeit at a very low density (personal communication). The second stage of electrophysiological development is characterized by a transitory decrease of the spontaneous electrical activity and the occurrence of postsynaptic potentials. This period corresponds to the time at which synaptic contacts were first observed using electron microscopy and it may represent a stage of synaptic differentiation. This transitory disappearance of the spontaneous electrical activity may he initiated by early synaptic contact. At early stages of end plate reinnervation, neuromuscular junctions in culture do not seem to display excitatory postsynaptic potentials ~. Xenopus embryonic neurons in culture lose their capacity for release of neurotransmitter through the whole cell membrane area after the first contact with muscle cells 6. Thus, a similar mechanism could be involved in our model and the decrease in the spontaneous electrical activity would reflect a decrease in the spontaneous release of excitatory substances due to the maturation of synaptic contact. This was indeed observed for the release of [3H]dopamine in similar cultures after 5 days of incubation 23. In the same way, receptor accumulation at the postsynaptic element could take a few days, as in nerve muscle cultures, during which time synapse efficacy would be low 1. This hypothesis would explain why synaptic activity develops progressively, between the 1st and the 3rd week of incubation. After 2 weeks in culture, mature synaptic contacts increase in number. However at this age axo-somatic contacts are rare in comparison to axodendritic synapses, which may also explain why only 50% of spontaneously active cells recorded from the soma display postsynaptic potential activity, because the space constant of the cell membrane. At the same time, degeneration of postsynaptic elements was frequently observed under electron microscopy, which suggests either a neuronal death or another mechanism of regulation of synapse number in culture. This might explain the transient decrease of the ratio of excitable cells between the 5th and the 9th day of incubation. The complex morphological and electrophysiological developmental processes which take place in hypothalamic cell cultures clearly follow the synaptic differentiation already described in other neuronal systems 4'5"18and give evidence that hypothalamic cell cultures represent an interesting model for analysis of the development of the electrophysiological properties of neuroendocrine cells. We have described the development of the spontaneous electrical activity in conditions where intraceilular medium was not modified; we need now to study the development of ionic currents, using whole cell recording techniques, to clarify the relationship between the maturation of the cell membrane and the synaptic differentiation, involved in the development of the general electrical activity of hypothalamic neurons in culture. ACKNOWLEDGEMENTS We thank B. Dupouy, R, Picart and C. Pennarun for their skilfull technical assistance, R. Miguelez for art work and D. Hackenauer for typing the manuscript. This work was supported by I.N.S.E.R.M. and C.N.R.S. (Greco 85 and U.A. 1115). 2~5 REFERENCES 1 Anderson. M.J. and Cohen, M.W., Nerve-induced and spontancous redistribution of acetylcholine receptors on cultured muscle cells, J. Physiol. (Lond.), 268 (19771 757-773. 2 Benda, P., De Vilry, F., Picart, R. and Tixier-Vidal, A,. Dissociated cell cultures from foetal mouse hypothalamus. Patterns of organization and ultrastructural features, Lrp. Brain Res.. 23 11975) 29-47. 3 Bergey, G.K., Fitzgerald, S.C., Sihrier, B.K. and Nelson, P.G., Neuronal maturation in mammalian cell culture is dependent on spontaneous electrical activity. Brain Res.. 2(17 119811 49-58. 4 Changeux, J.P.. Courrege, P. and Danchin, A., A theory of the epigenesis of neuronal networks by selective stabilization of synapses, Proc. Natl. Acad. Sci. U.S.A., 70 (1973) 2974. 5 Changeux, J.P. and Danehin, A., Selective stabilisation of developing synapses as a mechanism for the specification of neuronal networks, Nature (Lond. 1,264 (1976) 705-712. 6 Chow, I. and Poo, M.M., Release of acety[choline from embryonic neurons upon contact with muscle cell, J. Neurosci., 5 11985) 11/76-t1/82. 7 Cohen, S.A., Early nerve-muscle synapses in vitro release transmitter over post-synaptic membrane having low acetylcholine sensitivity, Proc. Natl. Acad. Sci. U.S.A., 77 (1980) 644-648. 8 Dennis, M.J. and Miledi, R., Non-transmitting neuromuscular junctions during an early stage of end-plate reinnervation, J. Physiol. (Lond.L 239 (19741 553-570. 9 Fenwick, E.M.. Marty, A. and Neher, E., A patch-clamp study of bovine chromaffin cells of their sensitivity to aeetylcholine, J. Physiol, (Lond.), 331 ( 19821 577-597. 10 Frankenhauser, B. and Hodgkin, A.L.. The action of calcium on the electrical properties of squid axon, ,1. Physiol. (Lond.), 137 (1957) 218-244. 11 Henderson, L.P., Smith, M.A. and Spitzer, N.C., The absence of calcium blocks impulse-evoked release of acetylcholine but not de novo formation of functional neuromuscular synaptic contacts in culture. J. Neurosci., 4 119841 3141)-3150. 12 Hume, R I . , Role, L.W. and Fischbach, G.D., Acetylcholine release from growth cones detected with patches of acetylcholine receptor-rich membranes, Nature (Lond.), 305 (1983) 632-634. 13 Jackson, M.B., Lecar, H., Brenneman, D.E., Fitzgerald, S. and Nelson, P.G.. Electrical development in spinal cord cell culture, J. Neurosci., 2 (1982) 1051-1061. 14 Legendrc, P., Cooke, I.M. and Vincent, J.D., Regenerative responses of long duration recorded intracellularly from dispersed cell cultures of feotal mouse hypothalamus, J. Neurophysiol., 48 (1982) 1121-1141. 15 Legendre, P,, McKenzie, J.S., Dupouy, B. and Vincent, J.D., Evidence for bursting pacemaker neurones in cul- tured spinal cord cells, Neuroscience, 16 (t985) 753-767. 16 Loudes, C., Faivre-Bauman, A. and Tixier-Vidal. A , Techniques for culture of hypothalamic neurons, Methods Enzymol.. 1113119831 313-314, 17 Mallart, A., Electric current flow reside permcurial sheaths of mouse motor nerves, .I, Phvsiof, ~lond, l, 368 119851 565-575~ 18 Mark, R.F., Synaptic repression at n.~-uromuscular junelions, PhysioL Rev.. 611 (198/i) 355-394= 19 McCormick. D.A. and Prince, D.A,. P~Jst-natal dcvclopmenl of electrophysiological properties of rat cerebral corlical pyramidal neurones, J Physiol. f lond. j, 393 (19871 743-762. 21) Ncher, E. and Sakmann, B., Single channel currents r c corded from membrane of denervated frog inuscle fibers, Nature (Lond.), 2611(1976) 799-8112. 21 Nelson. P,G, and Lieberman, M,, Excitable (:ells in Fis~uc (ulture, Plenum, New York, 1981. 22 Poulain, D.A., Wakerley, J.B. and Dyball, R.E.J., F-lectrophysiological differentiation of oxytocin- and vasopressin-secreting neurones, Proc. R. So~. L o n d , 196 119771 3(~7- 384. 23 Puymirat, J,, Barret, A., Faivre-Bauman, A. and TixierVidal, A., Biochemical characterizati~m of the uptake and release of 3H dopamine by dopamincrgic hypothalamic neurons: a developmental study using serum-free medium cultures, Dev. Biol., 119 11987) 75-84. 24 Sun, Y.A. and Poo, M.M., Non-quantal release of ace@choline at a developing neuromuscular synapse in culture, J. Neurosci.. 5 (1985) 634-642. 25 Theodosis. I).T., Legendre, P., Vincent, J D . and Cooke, I.M., lmmunocytochemically identified vasopressin neurons in culture shown slow, calcium-dependent electrical responses, Science, 221 (1983) 11152-11154. 26 Tixier-Vidal, A, and de Vitry, F,, Hypothalamic neurons in cell culture, Int. Rev. Cytol., 58 119791 291-331. 27 Tixier-Vidal, A., Picart, R., koudes, C. and Faivre-Bauman, A., Effects of polyunsaturated fatty acids and hormones on synaptogenesis in serum-free medium cultures of mouse fetal hypothalamic cells, Neuroscience, 17 11986) 115-132, 28 Tixier-Vidal. A.. Picart, R., Lourdes, (2., Puymirat. J. and Faivre-Bauman, A., Etude morphologique et fonctionnelle de la synaptog6n6se dans des cultures de cellules hypothalamiques foetales en milieu chimiquement d6fini, Ann. d'Endocrinol., 45 ( 19841 35N. 29 Vincent, J . D , Legendre, P., Poulam, D.A., Arnauld, E. and Theodosis, D.T., Electrophysiology of vasopressin-secreting cells, Front. Horm. Res.. 13 (1985) 52-68. 3tl Yashida, S., Action potentials dependent on monovalent cations in developing mouse embryos, Dev, Biol,: /111 ( 19851 200-206. 31 Young, S.H. and Mu-Ming, P., Spontaneous release of transmitter from growth cones of embryonic neurones, Nature (Lond, L 305 11983) 634-637.