Brain Research, 169 (1979)247-260 (L) Elsevier/North-HollandBiomedical Press
247
ELECTROPHYSIOLOGY OF DISSOCIATED HIPPOCAMPAL CULTURES FROM FETAL MICE
JOHN H. PEACOCK Department Of Neurology, Stat~['ord University School o[" Medicine, StatT[brd, Calif 94305 (U.S.A.)
(Accepted October, 5th 1978)
SUMMARY Action potentials, postsynaptic potentials (PSPs) and burst potentials have been recorded intracellularly from over 200 neurons in hippocampal cell cultures prepared from fetal mice of 13-18 days gestational age. Repetitively firing action potentials are elicited by intracellular electrical stimulation and often are preceded by stereotyped prepotentials which probably are generated on processes remote from the cell body. In some cells, action potentials are succeeded by long duration depolarizing afterpotentials (0.3-2 sec) with additional superimposed action potentials. Postburst afterhyperpolarization can last up to 2.5 sec. Action potentials are short (0.6-1.2 msec) with peak rates &rise from 64 to 267 V/sec (mean 139 ± 13 V/sec, 24 cells) and corresponding rates of fall from 21 to 133 V/sec (mean 70:5 7 V/sec, 24 cells). Following single action potentials, the afterhyperpolarization is usually less than 10 msec. Inhibitory PSPs occur frequently (up to 70 ~o incidence), have reversal potentials o f - - 3 0 to --40 mV, and can be evoked in synaptically coupled cell pairs. Excitatory PSPs can initiate prepotentials and action potentials, suggesting dendritic and somatic loci respectively. Neural networks exhibit a broad range of electrophysiologic phenomenology including reciprocal innervation, multiple innervation and synchronous bursting among a widespread population of neurons.
INTRODUCTION The hippocampus with its morphologically distinct cell types has become an increasingly attractive structure for investigation of the electrophysiology of central neurons and the abnormal functioning of these neurons during epileptogenesis3, 6. Results from studies in situ3, 7, in the slice preparation 9 and in organotypic cultures 13 are mutually complementary and encourage development of a dissociated hippo-
248 campal system in which a correlation of electrophysiology with morphology could be favorably attempted. Culture of hippocampus offers the possibility f~gr selective enrichment of pyramidal cells which constitute the major neuronal class in the prenatal hippocampus of rats" and mice 1 because relatively few granule cells have proliferated before birth. Previously it has been demonstrated that neurons in dissociated hippocampal cultures prepared from prenatal rats contain a major population of pyramidal cells 2 and this condition should apply similarly to dissociated cultures prepared from prenatal mice. In a companion study of such mouse cultures:,, hippocampal neurons displayed a marked degree of dendritic differentiation at both an optical and electron microscopic level. Some electrophysiologic properties of hippocampal cell cultures are similarly well-developed and are reported here. METHODS
Cell culture Hippocampi from fetus of 13-18 days gestational age were removed, mechanically dissociated by aspiration, and dispersed as single cells and small aggregates of cells into collagen-coated plastic dishes (Falcon 30011. Full details of the culture methodology have been separately enumerated 5. Electrophysiology The recording conditions for the culture chamber, the optical configuration for combined phase-contrast and ultraviolet illumination, and the preparation of Lucifer Yellow CH dye-containing microelectrodes have been previously deseribedZ. For routine electrophysiologic recording, microeleetrodes were filled immediately before use with 4 M potassium acetate (pH 7.0) containing 10 % KC1 (3 M). Microelectrode resistances ranged from 125 to 150 MfL In most cases a bridge circuit was used to pass currents across the membrane and simultaneously record voltage responses. On occasion separate current and voltage microelectrodes were placed in the same cell. Statistical analysis Data are reported in terms of the mean :~ standard error of the mean. Linear or exponential functions were fitted to the data by the method of least squares. RESULTS
Passive membrane properties Resting membrane potentials recorded from 228 hippocampal neurons ranged from --10 to --75 mV with a mean o f - - 3 8 . 2 ~z 0.79 mV. These data were collected from potassium acetate filled microetectrodes, and were selected on the basis of a stable membrane potential for several to many minutes even when the value =was low: Clearly in the lower range, the neurons were most likely injured; but often ~arge spontaneously occurring inhibitory PSPs could be recorded from them and thus these cells were included in the data base.
249 Hippocampal neurons in culture are small and fragile (15-20/~m). Successful penetration is critically dependent upon several factors. First, with the microelectrodes used in this study there appears to be a threshold for reliable recording which develops sometime between the second and third week in culture. Cells that are less mature than that age are often fatally damaged by microelectrode entry and fire only a brief barrage of action potentials before dying. It is possible that improved microelectrodes will allow recording from y o u n g cultures. Secondly, the total divalent cation strength appears to be important for recording stability. The total divalent cation concentration in normal medium in 2.8 m M which empirically is too low for stable recordings in these cells and routinely was increased to 4.5 mM by addition of Ca (final concentration 3.5 m M). Increases in divalent cation strength above 5 m M appeared to add little additional stability to the recordings. The third factor is microelectrode resistance. These cells are injured if the microelectrode resistance, measured in the culture medium, is less than 100 ME2 and the optimal range is from 125 to 150 M f L Some microelectrode tips were bevelled, but this maneuver did not provide a consistent technical benefit. The input resistance of these cells was estimated at 30-40 ME~ by single microelectrode recording using the bridge circuit, in several cells an attempt was made to use separate intracellular and voltage microelectrodes for a more precise evaluation o f the input resistance and other passive membrane properties. Stable recordings with two microelectrodes were achieved in two cells in which the input resistance was linear over a 15-25 mV range (r 2 = 0.99) studied with hyperpolarizing current pulses; input resistances were 34 and 44 Mr2 respectively (Table I). In one o f the cells, voltage transients larger than - - 2 5 mV were followed by a decrease in amplitude after about 15 msec. A similar sag in input resistance was occasionally seen in other cells studied with single microelectrodes and has been reported for dorsal root ganglion cells 8. The membrane time constants for each cell (Table I) were calculated by plotting Vo--V the exponential decay of - - with respect to time where Vo is the steady state voltage Vo and V the instantaneous voltage. Values of 1.5 and 1.6 msec closely fitted an TABLE
I
Membrane electrical properties o f cultured hippocarnpal neurons measured with two intracellular microelectrodes #l each cell Cell 1
Cell 2
Input resistance (Mt)) 34 Membrane time constant 1.5 Soma surface area (>~ 10 ~ cm2)* 12.57 Membrane resistance (~ cm2)** 1500 Input conductance (nS) 29.4 Action potential rate of rise (V/sec) 91 Summed process conductance (nS) 21.0 Soma conductance (nS) 8.4
44 1.6 12.57 1600 22.7 133 14.9 7.8
* Cells were considered to have the surface area of a sphere 20/~m in diameter. Obtained with one microelectrode prior to penetration with the second one.
**
250 e x p o n e n t i a l relationship I r ~ = 0.99) a n d c o r r e s p o n d respectively to the m e m b r a n e time constants of these cells. A s h o r t e r initial e x p o n e n t i a l decay was o b t a i n e d by a peeling p r o c e d u r e a n d was 1.0 ~ra 0.97) a n d 0.6 msec (r 2 0.981 for each cell respectively. The finding o f a two time c o n s t a n t charging curve suggests a dendritic structure which c o n t r i b u t e s to the passive electrical p r o p e r t i e s o f these cetts ~. The ratio o f process to s o m a c o n d u c t a n c e (Table 1) is a b o u t 2: this r a t i o also predicts that processes play a m a j o r role in the electrical structure of these cells. The s u m m e d process c o n d u c t a n c e value is o b t a i n e d by s u b t r a c t i n g the s o m a c o n d u c t a n c e (surface a r e a divided by m e m b r a n e resistance) f r o m the i n p u t c o n d u c t a n c e I reciprocal o f i n p u t resistance).
Active electrical membrane properties The usual m o r p h o l o g y o f h i p p o c a m p a l n e u r o n s in m a t u r e cultures is that o f a r o u n d cell b o d y (Fig. t A ) w i t h o u t m u c h evidence o f process d e v e l o p m e n t . F r o m these cells, however, a c o n s i d e r a b l e variety o f electrophysiologic p h e n o m e n a can be observed. N o t only are s p o n t a n e o u s l y occurring synaptic or action potentials (Fig. I E) r e c o r d e d f r o m m o s t cells, but also repetitively firing a c t i o n potentials are c o m m o n l y
F
i
0.2nA
,
Fig. 1. Combined fluorescent dye-staining with Lucifer Yellow CH and intracellular electrical recording from an isolated neuron shown in phase-contrast in A before penetration and in B in phase-fluorescence. C and D show the fluorescent cell (printed as positive image) plus and minus the cell body. Eis a chart writer record of spontaneous repetitive potentials and F is an oscilloscope record of the-same activity displayed at a faster time base than in E. The culture was 19 days old and was prepared from fetuses of 14 and 17 days gestational age.
251 elicited in response to a current step of about 50 msec duration. Some action potentials are initiated by stereotyped prepotentials (Fig. 1F, arrows) which suggests that their generation sites are remote from the cell body and that these cells have a more extensive process formation than can be visualized with phase-contrast optics. The existence of well developed processes was demonstrated in some cells by fluorescent dye-staining (Fig. IB and C). The neuron in Fig. 1C actually has a single major process which leaves the underside of the soma and then trifurcates (Fig. I D after removal of the cell body). It is possible that the hump on the falling phase of the action potentials as well as the low amplitude depolarizing potentials following the second and third action potentials have been generated on processes. This would give a total of three separate potentials, each of which may have originated from one of the three major (perhaps dendritic) branches or from the smooth (perhaps axonal) process exiting to the left lower side of the cell body (Fig. 1C). Although the prepotentials in these neurons have an electrical configuration which appears to represent an active electrogenic response, it is possible that they are recurrent synaptic potentials. But, if so, prepotentials should not be present after blockade of synaptic transmission. Thus one culture was examined in 10 m M Mg and 1.8 m M Ca. No synaptic potentials were recorded from any of 11 cells studied. However, the finding that prepotentials still occurred in 3 out of 10 repetitively firing cells suggests that a mechanism of recurrent synaptic excitation is not the usual basis for generation of prepotentials in these cells. Prepotentials (Fig. 2B and C) can also be activated by a preceding synaptic potential (Fig. 2A). Here the time to peak for the PSP is 3 msec compared to the time to peak for the prepotentials of 1 msec. Prepotentials in turn trigger action potentials in this cell (Fig. 2D). The peak rate of rise of the action potential with respect to time was measured in 24 cells by electronic differentiation of the action potential at membrane potentials of about - - 8 0 mV. Although values ranged broadly from 64 to 267 V/sec (mean 139 ~ 13
A
C
B
D ,
Fig. 2. Intracellular recording of postsynaptic potentials (A) followed by prepotentials in B and C, and triggering an action potential after the prepotential in D. In each set of tracings the coupling artifact is shown at the beginning of the trace. This cell was in a 21-day-old culture prepared from a fetus of 14 days gestational age.
252 V/sec), t h e r e clearly is a w e l l - d e v e l o p e d e l e c t r o g e n i c m e c h a n i s m
in these cells.
S i m i l a r l y these a c t i o n p o t e n t i a l s fell r a p i d l y f r o m p e a k values. R a t e s o f fall f o r the 24 ceils r a n g e d f r o m 2t to 133 V / s e c ( m e a n 70 -~ 7 V secl. T h e r a t i o o f t h e rise t o fall rates was 2.2 -~- 0 . 1 4 T h e d u r a t i o n s o f a c t i o n p o t e n t i a l s w h i c h w e r e elicited by a s h o r t p r e c e d i n g s t i m u l a t i n g pulse w e r e m e a s u r e d in 7 cells f r o m t h e o n s e t o f t h e active r e s p o n s e to the
B
+6mV
~
~
.L
_
_
, 1
-58mY
C
L-
~ •
,
.
2 O m V
-
%-
- - -
Fig. 3. Spontaneously occurring inhibitory PSPs in B, upper trace, recorded from the cell shown in the phase-contrast photomcirograph in A, arrow. Membrane holding current was adjusted by a second electrode placed in the same cell to the membrane voltages indicated at the left hand of the 3 traces shown in B. This cell was in a 35-day-old culture from a fetus of 17 days gestational age. In C, upper trace, and D, upper 2 traces, are shown transynapticaUy evoked PSPs from another cell. The presynaptic action potential is in the bottom trace of D. The responding PSPs are shown at two levels of membrane potential indicated at the left hand site of the tracing. In C an action potential occurred when the postsynaptic cell was hyperpolarized to about - - 7 0 InV. The bottom trace in C shows the stimulating current used to elicit the presynaptic action potential in both C andD. This cell was from a 26-day-old culture prepared from a fetus of 17 days gestational age.
253 termination of the falling phase and were 0.6 to 1.3 msec with a mean of 0.9 ~ 0.09 msec. Following the action potential, the duration of the afterhyperpolarization in these 7 cells ranged from 2 to 10 msec. Long duration afterhyperpolarizations to single action potentials were not observed under the present recording conditions.
Sj,naptic potentials There is abundant evidence for spontaneously occurring inhibitory and excitatory PSPs in these cultures, and occasionally linked excitatory-inhibitory PSPs occur also. The incidence of spontaneous synaptic activity was about 70 ~ in cultures from several fetal ages grown during the same period of time and well matched in terms of culture environment and testing conditions. In this series, inhibitory PSPs predominated and were recorded from 50/72 consecutively penetrated cells. Inhibitory PSPs are usually of greatest amplitude ( --10 to - - I 5 mY) immediately after penetration and then often decline in amplitude over several seconds to a steady state amplitude at resting potentials o f - - 3 0 to --40 inV. In some cells a spontaneous reversal in sign occurs within this membrane potential range and is followed by positive-going potentials at higher resting membrane potentials. These inhibitory PSPs can be reversed by application of constant hyperpolarizing currents, but the accuracy of the reversal potentials is open to question when microelectrodes with high resistances are used in a bridge circuit. This problem can be surmounted with two intracellular microelectrodes (Fig. 3A) by using one microelectrode for stimulation and one for recording purposes. In this case the reversal potential was about --34 mV (Fig. 3B, center trace). Such spontaneously occurring potentials from other cells have been reversibly blocked by 20 mM Mg applied directly from large tip micropipettes (about 20 ,urn tip diameter). Inhibitory PSPs can also be transsynaptically evoked by intracellular stimulation of a presynaptic cell. In Fig. 3D, a presynaptic action potential, lower record, is followed after a brief latency (0.4 msec) in a second cell by a PSP whose sign is determined at the membrane potentials given in the upper and middle traces. An action potential (Fig. 3C) is evoked with further hyperpolarization and is followed by a depolarizing afterpotential of much greater duration than the PSPs in Fig. 3D. The amplitude of PSPs range from I to 20 mV in different cells at membrane potentials o f - - 6 0 to --80 inV. Such positive-going PSPs may include both excitatory as well as reversed inhibitory PSPs. No systematic attempt to classify PSPs on the basis of reversal potential was made in the present series. Potentials smaller than 1 mV would not have been reliably detected because of the recording noise introduced by technical requirements of high resistance microelectrodes. It would appear that the PSPs observed have been generated in response to electrogenic activity in the presynaptic cell because no PSPs have been detected in 1/~g/ml tetrodotoxin or 8.5-10 mM Mg. Synaptic delays in 19 synaptically coupled cell pairs ranged from 0.4 msec (Fig. 3D) for a probable monosynaptic junction to 17 msec which most likely represents the summation of multiple synaptic delays. The variation in synaptic delay has been illustrated in data recorded from one cell, Fig. 4A1--4, in which 4 separate presynaptic
254
1
2
3
4
! . . . .
B
Fig. 4. A~-~ shows the postsynaptic response recorded in the same neuron, upper traces, to 4 separate presynaptic inputs, lower traces. B1 shows the same cell pair as in A1 but at different oscilloscope gains in order to display the excitatory PSPs in the upper traces which followed the action potentials in the lower traces. B2 and B3 from the same cell pair show transynaptically evoked burst firing, upper traces, and reciprocal activation of the presynaptic cell, lower traces. B ~ again shows the same cell pair as B1 but in this case the other cell is electrically stimulated and is followed by a synaptic potential recorded from the original presynaptic member of this cell pair. This cell was in a culture 21 days old prepared from a fetus of 14 days gestational age. The calibration in A1 applies also to A2,3, ~. The amplitude calibration in B2 holds also for B~ but the time base calibration in B2 has a value of 100 msec for B:~.
inputs generated similarly a p p e a r i n g responses but after progressively longer latencies f r o m the p r e s y n a p t i c action potential. The response c o m p l e x c o n s i s t e d at a m i n i m u m o f an excitatory P S P followed by two a c t i o n potentials a n d their prepotentials (see Fig. 2 w h i c h is t a k e n f r o m the cell p a i r in Fig. 4A4). S y n a p t i c delays in Fig, 4A1 4 are 2 msec, 7 msec, 10 msec a n d 17 msec respectively.
Burst potentials Burst potentials were r e c o r d e d in a t o t a l o f 61 cells, some o f which were f r o m cultures g r o w n and tested in D M E M a n d o t h e r s in M E M . Bursting occurs s p o n t a n e ously in some cells, a n d in others can be elicited b y direct or p r e s y n a p t i c i n t r a c e l l u l a r stimulation. In some cultures, m a n y n e u r o n s are linked t o g e t h e r in a p a t t e r n o f s y n c h r o n o u s l y occurring b u r s t i n g activity. A n e x a m p l e o f this p h e n o m e n o n is s h o w n in Fig. 5 in which 25 cells are s y n c h r o n o u s l y c o u p l e d within a m i c r o s c o p i c field o f 800 # m d i a m e t e r a n d over a distance o f 400 # m between cells. The e x p e r i m e n t a l sequence was as follows. F r o m 6 cells, s y n c h r o n o u s activity relative to cell 1 was r e c o r d e d a n d the last m e m b e r o f this g r o u p was called cell 2. The m i c r o e l e c t r o d e in cell 1 was m o v e d to cell 3 where bursting c o n t i n u e d to be s y n c h r o n o u s to cell 2. Then while c o n t i n u o u s l y r e c o r d i n g from cell 3 as a reference point, 22 other cells were p e n e t r a t e d in repetitive sequence using the s a m e microelectrode. Twenty o f these n e u r o n s were s y n c h r o n o u s l y c o u p l e d to cell 3, a n d 4 examples are shown in Fig. 5E a n d F. Bursting occurred n e a r l y s i m u l t a n e o u s l y (within 1 r e s e t ) for all cell pairs except 2 a n d 3 (Fig. 5C) where cell 2 h a d a n earlier onset t h a n cell 3 T h e response to i n t r a c e l l u l a r s t i m u l a t i o n in cell 2 (Fig. 5B) consisted o f an action p o t e n t i a l f o l l o w e d by a local response but n o t a fully d e v e l o p e d s e c o n d action
CELL 2
B
5 msec
lOmV~
0.SnA Jr
C
D
I
~mV 40~
CELL 3
,o j
££JJJ,
0.5 nA
E
F
10
SECONDS
Fig. 5. Synchronous bursting recorded from multiple cells within a single microscopic field of 800 t~m diameter. A is a phase-contrast photomicrograph of the central portion of the field with identified cells described in the text. B shows an action potential, upper trace, elicited in cell 2 by intracellular current stimulation, lower trace. The middle trace shows the first derivative of the voltage with respect to time (electronically differentiated). Calibration here is as noted in D, middle trace. C shows oscilloscopic recordings of spontaneously occurring bursting activity in cells 2 and 3. D shows repetitive action potentials, upper trace, elicited by transmembrane current stimulation, bottom trace, and with electronic differentiation, middle trace, as in B. E displays chartwriter record of bursting potentials recorded from cell 3 on 4 different occasions with respect to synchronous bursting activity recorded in F from 4 separate cells. The number of the cell is indicated to the left of the tracing. Resting membrane potentials for cells 3 and 14 were --65 mV and --60 mV respectively. The other cells shown in F had lower membrane potentials than cell 14 as has been indicated in the figure by positioning relative to cell 14. This was a 55-day-old culture from a fetus of 13 days gestational age.
256 potential. This voltage configuration was observed in 6 other cells in the overall series of recordings. In contrast to cell 2. the action potential (Fig. 5D) in cell 3 continued to fire repetitively when periodically checked over the 2.3 h period of recording. The bursting observed in cell 3 was not affected by stimulation of any of the cells penetrated with the roving microelectrode, indicating that none of these cells was the pacemaker responsible for the bursting activity. An example of bursting elicited by direct irttracellular stimulation is presented in Fig. 6A. After application of the stimulus during which repetitive firing occurs, there is a prolonged depolarization lasting up to 500 msec. Superimposed upon thi~ do~ depolarizing waveform are brief depolarizing voltage transients which may +epre.sent PSPs or local active responses, and action potentials. In addition there are depolarizing afterpotentials which succeed some action potentials in the burst (Fig. 6A, B, C arrowheads) regardless of whether the burst in this cell is triggered by direct ~Fig. 6A) or transynaptic (Fig. 6B,C) stimulation. Some attenuated depolarizing afterpotentials can be observed following other action potentials in the bursts. Transynaptically evoked bursts occurred in 4 19 synaptically coupled paws+ In one cell pair separated by a distance of about 100 #m (Fig. 7At) the excitatory PSP occurred after a delay of about 13 msec and was followed in 26 out of 57 trials by an action potential and a long rectangular lover 150 msecl depolarizing afterpotential terminated by a rather abrupt onset of the falling phase (Fig. 7A~). When the cell was hyperpolarized the amplitude of the rectangular response was larger and. in addition.
250
k
rnsec
I
i
t
C _Ppsy
PRE
-,
]~e~ I
Fig. 6. Chartwriter records of two burst firing patterns in the same cell. A shows burst firing elicited by intraceUular stimulation indicated by a bar below voltage recording. In B, POST, the burst is synaptically evoked by a spike on B, PRE, to direct stimulation again indicated by bar below+recording. The portion of the burst denoted by arrowheads in B is expanded 5-fold in C to clarify t h e d e p o l a d z i n g afterpotentials following spikes. The calibration+value of 250 msec in A applies to time base brackets in B and C. The duration of the event marker signal in B and C is 25 msec. Alltraces are displayed at the amplification shown in A. Culture age was 2t days and fetal gestational age was 14 days.
257
,
~lOmV
I 150msec
B
C
Fig. 7. Transynapticallyevoked burst firing in the cell labeled as POST by electrical stimulation of the cell labelled as PRE in A1 which is a phase-contrast photomicrograph taken after the microelectrodes were placed intracellularly. All electrical records are postsynaptic responses which begin after the coupling artifacts shown at the left hand margin of the trace. A3 was obtained at a resting potential of--50 mV whereas all other traces were recorded at a membrane potential of about --70 inV. The calibrating bar value of 30 msec should be used for As and C1,2,3. The calibrating bar value of 150 msec should be used for A~ and B~,,:,and B3. These data were recorded on FM tape at a recording speed of 3 in./sec which produces mild distortion of the waveform. This culture was 43 days old and was prepared from a fetus of 17 days gestational age. action potentials were superimposed to give a burst configuration to the response. The excitatory PSP sometimes triggered an action potential (Fig. 7B2 and 3, a n d C2 and 3), sometimes did n o t (Fig. 7B~ and C1), and sometimes appeared d u r i n g or after onset of the first action potential (Fig. 7B3, C3). Nevertheless the burst still occurred which suggested that the first action potential was n o t necessary for burst generation in this pair. However, the fact that n o bursting occurred in 22 trials in which the presynaptic action potential failed to evoke an excitatory PSP indicates that synaptic excitation, at least, is required for bursting in this cell. The excitatory PSP recorded from the soma may serve only as a m o n i t o r of synaptic i n p u t which m a y be distributed as well to dendritic sites for generation of burst firing. The m e a n burst d u r a t i o n was 331.4 ± 8.82 msec; the m e a n n u m b e r of action potentials per burst was 14.1 ~ 0.48.
258 Reciprocal interaction of varying complexity was observed in 3 out of 19 cell pairs. One such pair is shown in Fig. 4, B1 and B4, in which each member reciprocally evokes excitatory PSPs in the other; note persistence of the stereotyped doublet action potentials in Fig. 4B4 with intracellular stimulation as had also occurred with transynaptic stimulation in Fig. 4Al-.,~. On occasion, after the excita~/ory P s p (Fig. 4Bt upper trace), a series of potentials was generated (Fig. 4B2 ana a, upper traces) which in turn elicited potentials on the presynaptic cell (lower traces) and in Other cells could lead to reciprocal burst firing. DISCUSSION
The primary result from this work is that a wide diversity of electrical activity develops or has been maintained in parallel with the morphologic differentiation previously described for these dissociated cultures from fetal mouse hippocampus~. It is probable that many, if not most, of the intracellutar recordings in this study have been from neurons of pyramidal cell origin. Whether pyramidal cells which were postmitotic at the time of culture have been able to survive for two months or whether the majority of the ceils present at that time completed their final cell divisions shortly after being placed in culture has not yet been determined. Nevertheless. some features of the electrical recordings are strikingly similar to some of the characteristics of intact hippocampal pyramidal cells a,6.9. Spontaneously occurring action potentials which occurred singly or in short bursts without a depolarizing shift of the membrane potential were recorded from many cells. In these cells repetitive action potentials could be elicited by long pulses of depolarizing currents. Action potentials had fast rates of rise. were about I msec m duration, and had correspondingly fast rates of fall. In some cells the acticua potential was preceded by a brief depolarizing potential. These prepotentials were highly stereotyped, occurred with reliability in a gwen cell. did not have a large increase in amplitude on hyperpolarization, and were found under conditions of a high Mg to Ca ratio which otherwise abolished ongoing synaptic activity. On this basis it is concluded that these prepotentials are active electrical responses which are generated on portions of the membrane located on cell processes at a distance from the site of the microelectrode in the cell body. Additionally, fluorescent dye filling and electron mlcl'OScopy 5 strongly suggest the development of dendrites in these cells. On the basis of this combined electrical and morphologic evidence, these prepotentials very likely are generated on dendrites and possibly are equivalent to the fast prepotentials described by Spencer and KandeP °. but an axonal origin of the prepotentials should also be considered. In cultured hippocampal neurons, the initial dendritic branching points are often near the soma. as in Fig. ID. Thus functional parallels between neurons in culture and the intact mature hippocampus should be drawn with a caution similar to that previously discussed 7 in a comparison of the fast prepotentiats of immature and mature hippocampal pyramidal cells. Spontaneously occurring synaptic potentials can be recorded in up to 70 ~ of the neurons. This synaptic activity includes the frequent occurrence of inhibitory PSPs
259 with reversal potentials between --30 and --40 inV. Inhibitory synaptic activity is observed in medium containing glycine, a condition whereby glycine responses and presumably glycine-mediated inhibitory PSPs are not found in dissociated spinal cord cultures ~. Therefore glycine would seem an unlikely candidate for the inhibitory transmitter in hippocampal cultures and y-aminobutyric acid, which is not a component of the medium and which is a putative inhibitory transmitter in hippocampal pyramidal cells ~2, is a better candidate. Excitatory PSPs can trigger action potentials recorded from the cell body either with or without preceding dendritic action potentials suggesting that excitatory synaptic input occurs at both cell somata and cell dendrites. Ultrastructural evidence for these synaptic contacts has been discussed previouslyz. More complex synaptic potentials have been observed in which an excitatory PSP is abruptly terminated by onset of an inhibitory PSP. Such closely linked synaptic potentials suggest that synaptic contacts mediating those potentials are spaced near each other, perhaps on dendritic structures as illustrated in Fig. 10C of the preceding report 5. Transynaptic stimulation can elicit these complex synaptic potentials as well as elicit monophasic excitatory PSPs triggering action potentials and inhibitory PSPs with demonstrated reversal potentials. Given this abundance of synaptic activity it is not surprising that reciprocal innervation can be demonstrated in nerve cell pairs. A significant feature of the spontaneous electrical activity in hippocampal cultures is the widespread occurrence of bursting potentials with an accompanying depolarizing shift of the membrane potential. These bursts can be elicited by direct intracellular stimulation as well as by stimulation of a presynaptic cell. Bursting appears to be mediated by depolarizing afterpotentials which trigger additional action potentials in order to maintain the burst for periods lasting up to 800 msec. In those cases of transynaptic activation of burst firing, reciprocal bursting on the presynaptic cell has also been observed. Some features of burst firing in dissociated hippocampal cultures resemble that reported for cultured hippocampal explants from fetal mouse by Zipser and Crain la. In that study, burst firing occurred spontaneously, in response to stimulation of other neurons, and in the presence of low concentrations of strychnine and bicuculline, The use of Lucifer Yellow CH 11 as an agent for fluorescent dye staining, along with concomitant recording of electrical activity during the time that cell morphology is being disclosed, offers a powerful technique for identification of cell types based on criteria of structure and function. In addition, studies on innervation in these cultures will be greatly enhanced by the capability to locate synaptically coupled ceils which are contacted by the dye-filled neuron. The work of this study taken together with that of the preceding report 5 presents a survey of certain well-developed morphologic and electrophysiologic characteristics of dissociated hippocampal cultures from fetal mice. Among possible avenues for further work, this culture system offers opportunities for detailed study of dendritic structure and function, examination of mechanisms of burst firing in hippocampa[ neurons, and studies of synaptic interaction in a population probably dominated by pyramidal cells.
260 ACKNOWLEDGEMENTS This w o r k was s u p p o r t e d by N I H G r a n t N S 1215l. I t h a n k Dr. D a v i d Prince for his helpful review o f the m a n u s c r i p t and a m grateful to Dr. W a l t e r S te w a r d for his gift o f Lucifer Yellow C H , Reed Pike helped with p h o t o g r a p h y and Cheryl J o o typed the m an u scr i p t .
REFERENCES 1 Angevine, J. B.. Development of the hippocampal region. In R. L. Isaacson and K, H, Pribrana (Eds.), The Hippocampus, Vol. l, Plenum Press, New York, 1975. pp. 61-90. 2 Banker, G. A. and Cowan, W. M., Rat hippocampal neurons in dispersed cell cuIture, Brain Research. 126 (1977) 397-425. 3 Kandel, E. R. and Spencer, W. A., Electrophysiologic properties of an archicortical neuron. A n n N. Y. Acad. Sci.. 94 (1961) 570-603. 4 Nelson, P. G., Ransom, B. R.. Henkart. M. and Bullock, P. N.. Mouse spinal cord in celt culture. IV. Modulation of inhibitory synaptic function, J, Neurophysiol., 40 (1977) 1178-.1187~ 5 Peacock, J. H., Rush. D. F. and Mathers, L. H., Morphology of dissociated hippocampal cultures from fetal mice. Brain Research. 169 (1979) 247-260. 6 Prince, D. A., Neurophysiology of epilepsy, Ann. Rev. Neurosci., 1 (1978) 395-415. 7 Purpura. D. P.. Prelevic. S. and Santini, M.. Postsynaptic potentials and spike variations in the feline hippocampus during postnatal ontogenesis, Exp. Neurol.. 22 (1968) 408 422. 8 Ransom, B. R., Neale, E., Henkart, M., Bullock, P, N. and Nelson, P, G., Mouse spinal cord in cell culture. I. Morphology and intrinsic neuronal electrophysiologic properties. J. Neurophysiol., 40 (1977) t 132-1150. 9 Schwartzkroin, P. A., Characteristics of CAl neurons recorded intracellular[y in the hippocampaJ in vitro slice preparation, Brain Research, 85 (1975J 423-436. 10 Spencer, W. A. and Kandel, E. R., Electrophysiology ofhippocampal neurons. IV. kas~ prepotenrials, J. NeurophysioL, 24 (196lj 272-285. 11 Stewart, W. W.. Functional corrections between cells as revealed by dye-coupling with a highly fluorescent naphthalamide tracer. Cell, 14 (1978) 741-759. 12 Straughan, D. W., Neurotransmitter and the hippocampus. In R L. Isaacson and K. H, Pribram (Eds.), The Hippocampus, Vol. t. Plenum Press, New York. 1975. pp. 239-268. 13 Zipser, B., Crain, S. M. and Bornstein, M. B., Directly evoked "paroxysmal" depolarizations of mouse hippocampal neurons in synaptically organized explants in long-term culture. Brain Research, 60 ( 19731 489-495.
247
ELECTROPHYSIOLOGY OF DISSOCIATED HIPPOCAMPAL CULTURES FROM FETAL MICE
JOHN H. PEACOCK Department Of Neurology, Stat~['ord University School o[" Medicine, StatT[brd, Calif 94305 (U.S.A.)
(Accepted October, 5th 1978)
SUMMARY Action potentials, postsynaptic potentials (PSPs) and burst potentials have been recorded intracellularly from over 200 neurons in hippocampal cell cultures prepared from fetal mice of 13-18 days gestational age. Repetitively firing action potentials are elicited by intracellular electrical stimulation and often are preceded by stereotyped prepotentials which probably are generated on processes remote from the cell body. In some cells, action potentials are succeeded by long duration depolarizing afterpotentials (0.3-2 sec) with additional superimposed action potentials. Postburst afterhyperpolarization can last up to 2.5 sec. Action potentials are short (0.6-1.2 msec) with peak rates &rise from 64 to 267 V/sec (mean 139 ± 13 V/sec, 24 cells) and corresponding rates of fall from 21 to 133 V/sec (mean 70:5 7 V/sec, 24 cells). Following single action potentials, the afterhyperpolarization is usually less than 10 msec. Inhibitory PSPs occur frequently (up to 70 ~o incidence), have reversal potentials o f - - 3 0 to --40 mV, and can be evoked in synaptically coupled cell pairs. Excitatory PSPs can initiate prepotentials and action potentials, suggesting dendritic and somatic loci respectively. Neural networks exhibit a broad range of electrophysiologic phenomenology including reciprocal innervation, multiple innervation and synchronous bursting among a widespread population of neurons.
INTRODUCTION The hippocampus with its morphologically distinct cell types has become an increasingly attractive structure for investigation of the electrophysiology of central neurons and the abnormal functioning of these neurons during epileptogenesis3, 6. Results from studies in situ3, 7, in the slice preparation 9 and in organotypic cultures 13 are mutually complementary and encourage development of a dissociated hippo-
248 campal system in which a correlation of electrophysiology with morphology could be favorably attempted. Culture of hippocampus offers the possibility f~gr selective enrichment of pyramidal cells which constitute the major neuronal class in the prenatal hippocampus of rats" and mice 1 because relatively few granule cells have proliferated before birth. Previously it has been demonstrated that neurons in dissociated hippocampal cultures prepared from prenatal rats contain a major population of pyramidal cells 2 and this condition should apply similarly to dissociated cultures prepared from prenatal mice. In a companion study of such mouse cultures:,, hippocampal neurons displayed a marked degree of dendritic differentiation at both an optical and electron microscopic level. Some electrophysiologic properties of hippocampal cell cultures are similarly well-developed and are reported here. METHODS
Cell culture Hippocampi from fetus of 13-18 days gestational age were removed, mechanically dissociated by aspiration, and dispersed as single cells and small aggregates of cells into collagen-coated plastic dishes (Falcon 30011. Full details of the culture methodology have been separately enumerated 5. Electrophysiology The recording conditions for the culture chamber, the optical configuration for combined phase-contrast and ultraviolet illumination, and the preparation of Lucifer Yellow CH dye-containing microelectrodes have been previously deseribedZ. For routine electrophysiologic recording, microeleetrodes were filled immediately before use with 4 M potassium acetate (pH 7.0) containing 10 % KC1 (3 M). Microelectrode resistances ranged from 125 to 150 MfL In most cases a bridge circuit was used to pass currents across the membrane and simultaneously record voltage responses. On occasion separate current and voltage microelectrodes were placed in the same cell. Statistical analysis Data are reported in terms of the mean :~ standard error of the mean. Linear or exponential functions were fitted to the data by the method of least squares. RESULTS
Passive membrane properties Resting membrane potentials recorded from 228 hippocampal neurons ranged from --10 to --75 mV with a mean o f - - 3 8 . 2 ~z 0.79 mV. These data were collected from potassium acetate filled microetectrodes, and were selected on the basis of a stable membrane potential for several to many minutes even when the value =was low: Clearly in the lower range, the neurons were most likely injured; but often ~arge spontaneously occurring inhibitory PSPs could be recorded from them and thus these cells were included in the data base.
249 Hippocampal neurons in culture are small and fragile (15-20/~m). Successful penetration is critically dependent upon several factors. First, with the microelectrodes used in this study there appears to be a threshold for reliable recording which develops sometime between the second and third week in culture. Cells that are less mature than that age are often fatally damaged by microelectrode entry and fire only a brief barrage of action potentials before dying. It is possible that improved microelectrodes will allow recording from y o u n g cultures. Secondly, the total divalent cation strength appears to be important for recording stability. The total divalent cation concentration in normal medium in 2.8 m M which empirically is too low for stable recordings in these cells and routinely was increased to 4.5 mM by addition of Ca (final concentration 3.5 m M). Increases in divalent cation strength above 5 m M appeared to add little additional stability to the recordings. The third factor is microelectrode resistance. These cells are injured if the microelectrode resistance, measured in the culture medium, is less than 100 ME2 and the optimal range is from 125 to 150 M f L Some microelectrode tips were bevelled, but this maneuver did not provide a consistent technical benefit. The input resistance of these cells was estimated at 30-40 ME~ by single microelectrode recording using the bridge circuit, in several cells an attempt was made to use separate intracellular and voltage microelectrodes for a more precise evaluation o f the input resistance and other passive membrane properties. Stable recordings with two microelectrodes were achieved in two cells in which the input resistance was linear over a 15-25 mV range (r 2 = 0.99) studied with hyperpolarizing current pulses; input resistances were 34 and 44 Mr2 respectively (Table I). In one o f the cells, voltage transients larger than - - 2 5 mV were followed by a decrease in amplitude after about 15 msec. A similar sag in input resistance was occasionally seen in other cells studied with single microelectrodes and has been reported for dorsal root ganglion cells 8. The membrane time constants for each cell (Table I) were calculated by plotting Vo--V the exponential decay of - - with respect to time where Vo is the steady state voltage Vo and V the instantaneous voltage. Values of 1.5 and 1.6 msec closely fitted an TABLE
I
Membrane electrical properties o f cultured hippocarnpal neurons measured with two intracellular microelectrodes #l each cell Cell 1
Cell 2
Input resistance (Mt)) 34 Membrane time constant 1.5 Soma surface area (>~ 10 ~ cm2)* 12.57 Membrane resistance (~ cm2)** 1500 Input conductance (nS) 29.4 Action potential rate of rise (V/sec) 91 Summed process conductance (nS) 21.0 Soma conductance (nS) 8.4
44 1.6 12.57 1600 22.7 133 14.9 7.8
* Cells were considered to have the surface area of a sphere 20/~m in diameter. Obtained with one microelectrode prior to penetration with the second one.
**
250 e x p o n e n t i a l relationship I r ~ = 0.99) a n d c o r r e s p o n d respectively to the m e m b r a n e time constants of these cells. A s h o r t e r initial e x p o n e n t i a l decay was o b t a i n e d by a peeling p r o c e d u r e a n d was 1.0 ~ra 0.97) a n d 0.6 msec (r 2 0.981 for each cell respectively. The finding o f a two time c o n s t a n t charging curve suggests a dendritic structure which c o n t r i b u t e s to the passive electrical p r o p e r t i e s o f these cetts ~. The ratio o f process to s o m a c o n d u c t a n c e (Table 1) is a b o u t 2: this r a t i o also predicts that processes play a m a j o r role in the electrical structure of these cells. The s u m m e d process c o n d u c t a n c e value is o b t a i n e d by s u b t r a c t i n g the s o m a c o n d u c t a n c e (surface a r e a divided by m e m b r a n e resistance) f r o m the i n p u t c o n d u c t a n c e I reciprocal o f i n p u t resistance).
Active electrical membrane properties The usual m o r p h o l o g y o f h i p p o c a m p a l n e u r o n s in m a t u r e cultures is that o f a r o u n d cell b o d y (Fig. t A ) w i t h o u t m u c h evidence o f process d e v e l o p m e n t . F r o m these cells, however, a c o n s i d e r a b l e variety o f electrophysiologic p h e n o m e n a can be observed. N o t only are s p o n t a n e o u s l y occurring synaptic or action potentials (Fig. I E) r e c o r d e d f r o m m o s t cells, but also repetitively firing a c t i o n potentials are c o m m o n l y
F
i
0.2nA
,
Fig. 1. Combined fluorescent dye-staining with Lucifer Yellow CH and intracellular electrical recording from an isolated neuron shown in phase-contrast in A before penetration and in B in phase-fluorescence. C and D show the fluorescent cell (printed as positive image) plus and minus the cell body. Eis a chart writer record of spontaneous repetitive potentials and F is an oscilloscope record of the-same activity displayed at a faster time base than in E. The culture was 19 days old and was prepared from fetuses of 14 and 17 days gestational age.
251 elicited in response to a current step of about 50 msec duration. Some action potentials are initiated by stereotyped prepotentials (Fig. 1F, arrows) which suggests that their generation sites are remote from the cell body and that these cells have a more extensive process formation than can be visualized with phase-contrast optics. The existence of well developed processes was demonstrated in some cells by fluorescent dye-staining (Fig. IB and C). The neuron in Fig. 1C actually has a single major process which leaves the underside of the soma and then trifurcates (Fig. I D after removal of the cell body). It is possible that the hump on the falling phase of the action potentials as well as the low amplitude depolarizing potentials following the second and third action potentials have been generated on processes. This would give a total of three separate potentials, each of which may have originated from one of the three major (perhaps dendritic) branches or from the smooth (perhaps axonal) process exiting to the left lower side of the cell body (Fig. 1C). Although the prepotentials in these neurons have an electrical configuration which appears to represent an active electrogenic response, it is possible that they are recurrent synaptic potentials. But, if so, prepotentials should not be present after blockade of synaptic transmission. Thus one culture was examined in 10 m M Mg and 1.8 m M Ca. No synaptic potentials were recorded from any of 11 cells studied. However, the finding that prepotentials still occurred in 3 out of 10 repetitively firing cells suggests that a mechanism of recurrent synaptic excitation is not the usual basis for generation of prepotentials in these cells. Prepotentials (Fig. 2B and C) can also be activated by a preceding synaptic potential (Fig. 2A). Here the time to peak for the PSP is 3 msec compared to the time to peak for the prepotentials of 1 msec. Prepotentials in turn trigger action potentials in this cell (Fig. 2D). The peak rate of rise of the action potential with respect to time was measured in 24 cells by electronic differentiation of the action potential at membrane potentials of about - - 8 0 mV. Although values ranged broadly from 64 to 267 V/sec (mean 139 ~ 13
A
C
B
D ,
Fig. 2. Intracellular recording of postsynaptic potentials (A) followed by prepotentials in B and C, and triggering an action potential after the prepotential in D. In each set of tracings the coupling artifact is shown at the beginning of the trace. This cell was in a 21-day-old culture prepared from a fetus of 14 days gestational age.
252 V/sec), t h e r e clearly is a w e l l - d e v e l o p e d e l e c t r o g e n i c m e c h a n i s m
in these cells.
S i m i l a r l y these a c t i o n p o t e n t i a l s fell r a p i d l y f r o m p e a k values. R a t e s o f fall f o r the 24 ceils r a n g e d f r o m 2t to 133 V / s e c ( m e a n 70 -~ 7 V secl. T h e r a t i o o f t h e rise t o fall rates was 2.2 -~- 0 . 1 4 T h e d u r a t i o n s o f a c t i o n p o t e n t i a l s w h i c h w e r e elicited by a s h o r t p r e c e d i n g s t i m u l a t i n g pulse w e r e m e a s u r e d in 7 cells f r o m t h e o n s e t o f t h e active r e s p o n s e to the
B
+6mV
~
~
.L
_
_
, 1
-58mY
C
L-
~ •
,
.
2 O m V
-
%-
- - -
Fig. 3. Spontaneously occurring inhibitory PSPs in B, upper trace, recorded from the cell shown in the phase-contrast photomcirograph in A, arrow. Membrane holding current was adjusted by a second electrode placed in the same cell to the membrane voltages indicated at the left hand of the 3 traces shown in B. This cell was in a 35-day-old culture from a fetus of 17 days gestational age. In C, upper trace, and D, upper 2 traces, are shown transynapticaUy evoked PSPs from another cell. The presynaptic action potential is in the bottom trace of D. The responding PSPs are shown at two levels of membrane potential indicated at the left hand site of the tracing. In C an action potential occurred when the postsynaptic cell was hyperpolarized to about - - 7 0 InV. The bottom trace in C shows the stimulating current used to elicit the presynaptic action potential in both C andD. This cell was from a 26-day-old culture prepared from a fetus of 17 days gestational age.
253 termination of the falling phase and were 0.6 to 1.3 msec with a mean of 0.9 ~ 0.09 msec. Following the action potential, the duration of the afterhyperpolarization in these 7 cells ranged from 2 to 10 msec. Long duration afterhyperpolarizations to single action potentials were not observed under the present recording conditions.
Sj,naptic potentials There is abundant evidence for spontaneously occurring inhibitory and excitatory PSPs in these cultures, and occasionally linked excitatory-inhibitory PSPs occur also. The incidence of spontaneous synaptic activity was about 70 ~ in cultures from several fetal ages grown during the same period of time and well matched in terms of culture environment and testing conditions. In this series, inhibitory PSPs predominated and were recorded from 50/72 consecutively penetrated cells. Inhibitory PSPs are usually of greatest amplitude ( --10 to - - I 5 mY) immediately after penetration and then often decline in amplitude over several seconds to a steady state amplitude at resting potentials o f - - 3 0 to --40 inV. In some cells a spontaneous reversal in sign occurs within this membrane potential range and is followed by positive-going potentials at higher resting membrane potentials. These inhibitory PSPs can be reversed by application of constant hyperpolarizing currents, but the accuracy of the reversal potentials is open to question when microelectrodes with high resistances are used in a bridge circuit. This problem can be surmounted with two intracellular microelectrodes (Fig. 3A) by using one microelectrode for stimulation and one for recording purposes. In this case the reversal potential was about --34 mV (Fig. 3B, center trace). Such spontaneously occurring potentials from other cells have been reversibly blocked by 20 mM Mg applied directly from large tip micropipettes (about 20 ,urn tip diameter). Inhibitory PSPs can also be transsynaptically evoked by intracellular stimulation of a presynaptic cell. In Fig. 3D, a presynaptic action potential, lower record, is followed after a brief latency (0.4 msec) in a second cell by a PSP whose sign is determined at the membrane potentials given in the upper and middle traces. An action potential (Fig. 3C) is evoked with further hyperpolarization and is followed by a depolarizing afterpotential of much greater duration than the PSPs in Fig. 3D. The amplitude of PSPs range from I to 20 mV in different cells at membrane potentials o f - - 6 0 to --80 inV. Such positive-going PSPs may include both excitatory as well as reversed inhibitory PSPs. No systematic attempt to classify PSPs on the basis of reversal potential was made in the present series. Potentials smaller than 1 mV would not have been reliably detected because of the recording noise introduced by technical requirements of high resistance microelectrodes. It would appear that the PSPs observed have been generated in response to electrogenic activity in the presynaptic cell because no PSPs have been detected in 1/~g/ml tetrodotoxin or 8.5-10 mM Mg. Synaptic delays in 19 synaptically coupled cell pairs ranged from 0.4 msec (Fig. 3D) for a probable monosynaptic junction to 17 msec which most likely represents the summation of multiple synaptic delays. The variation in synaptic delay has been illustrated in data recorded from one cell, Fig. 4A1--4, in which 4 separate presynaptic
254
1
2
3
4
! . . . .
B
Fig. 4. A~-~ shows the postsynaptic response recorded in the same neuron, upper traces, to 4 separate presynaptic inputs, lower traces. B1 shows the same cell pair as in A1 but at different oscilloscope gains in order to display the excitatory PSPs in the upper traces which followed the action potentials in the lower traces. B2 and B3 from the same cell pair show transynaptically evoked burst firing, upper traces, and reciprocal activation of the presynaptic cell, lower traces. B ~ again shows the same cell pair as B1 but in this case the other cell is electrically stimulated and is followed by a synaptic potential recorded from the original presynaptic member of this cell pair. This cell was in a culture 21 days old prepared from a fetus of 14 days gestational age. The calibration in A1 applies also to A2,3, ~. The amplitude calibration in B2 holds also for B~ but the time base calibration in B2 has a value of 100 msec for B:~.
inputs generated similarly a p p e a r i n g responses but after progressively longer latencies f r o m the p r e s y n a p t i c action potential. The response c o m p l e x c o n s i s t e d at a m i n i m u m o f an excitatory P S P followed by two a c t i o n potentials a n d their prepotentials (see Fig. 2 w h i c h is t a k e n f r o m the cell p a i r in Fig. 4A4). S y n a p t i c delays in Fig, 4A1 4 are 2 msec, 7 msec, 10 msec a n d 17 msec respectively.
Burst potentials Burst potentials were r e c o r d e d in a t o t a l o f 61 cells, some o f which were f r o m cultures g r o w n and tested in D M E M a n d o t h e r s in M E M . Bursting occurs s p o n t a n e ously in some cells, a n d in others can be elicited b y direct or p r e s y n a p t i c i n t r a c e l l u l a r stimulation. In some cultures, m a n y n e u r o n s are linked t o g e t h e r in a p a t t e r n o f s y n c h r o n o u s l y occurring b u r s t i n g activity. A n e x a m p l e o f this p h e n o m e n o n is s h o w n in Fig. 5 in which 25 cells are s y n c h r o n o u s l y c o u p l e d within a m i c r o s c o p i c field o f 800 # m d i a m e t e r a n d over a distance o f 400 # m between cells. The e x p e r i m e n t a l sequence was as follows. F r o m 6 cells, s y n c h r o n o u s activity relative to cell 1 was r e c o r d e d a n d the last m e m b e r o f this g r o u p was called cell 2. The m i c r o e l e c t r o d e in cell 1 was m o v e d to cell 3 where bursting c o n t i n u e d to be s y n c h r o n o u s to cell 2. Then while c o n t i n u o u s l y r e c o r d i n g from cell 3 as a reference point, 22 other cells were p e n e t r a t e d in repetitive sequence using the s a m e microelectrode. Twenty o f these n e u r o n s were s y n c h r o n o u s l y c o u p l e d to cell 3, a n d 4 examples are shown in Fig. 5E a n d F. Bursting occurred n e a r l y s i m u l t a n e o u s l y (within 1 r e s e t ) for all cell pairs except 2 a n d 3 (Fig. 5C) where cell 2 h a d a n earlier onset t h a n cell 3 T h e response to i n t r a c e l l u l a r s t i m u l a t i o n in cell 2 (Fig. 5B) consisted o f an action p o t e n t i a l f o l l o w e d by a local response but n o t a fully d e v e l o p e d s e c o n d action
CELL 2
B
5 msec
lOmV~
0.SnA Jr
C
D
I
~mV 40~
CELL 3
,o j
££JJJ,
0.5 nA
E
F
10
SECONDS
Fig. 5. Synchronous bursting recorded from multiple cells within a single microscopic field of 800 t~m diameter. A is a phase-contrast photomicrograph of the central portion of the field with identified cells described in the text. B shows an action potential, upper trace, elicited in cell 2 by intracellular current stimulation, lower trace. The middle trace shows the first derivative of the voltage with respect to time (electronically differentiated). Calibration here is as noted in D, middle trace. C shows oscilloscopic recordings of spontaneously occurring bursting activity in cells 2 and 3. D shows repetitive action potentials, upper trace, elicited by transmembrane current stimulation, bottom trace, and with electronic differentiation, middle trace, as in B. E displays chartwriter record of bursting potentials recorded from cell 3 on 4 different occasions with respect to synchronous bursting activity recorded in F from 4 separate cells. The number of the cell is indicated to the left of the tracing. Resting membrane potentials for cells 3 and 14 were --65 mV and --60 mV respectively. The other cells shown in F had lower membrane potentials than cell 14 as has been indicated in the figure by positioning relative to cell 14. This was a 55-day-old culture from a fetus of 13 days gestational age.
256 potential. This voltage configuration was observed in 6 other cells in the overall series of recordings. In contrast to cell 2. the action potential (Fig. 5D) in cell 3 continued to fire repetitively when periodically checked over the 2.3 h period of recording. The bursting observed in cell 3 was not affected by stimulation of any of the cells penetrated with the roving microelectrode, indicating that none of these cells was the pacemaker responsible for the bursting activity. An example of bursting elicited by direct irttracellular stimulation is presented in Fig. 6A. After application of the stimulus during which repetitive firing occurs, there is a prolonged depolarization lasting up to 500 msec. Superimposed upon thi~ do~ depolarizing waveform are brief depolarizing voltage transients which may +epre.sent PSPs or local active responses, and action potentials. In addition there are depolarizing afterpotentials which succeed some action potentials in the burst (Fig. 6A, B, C arrowheads) regardless of whether the burst in this cell is triggered by direct ~Fig. 6A) or transynaptic (Fig. 6B,C) stimulation. Some attenuated depolarizing afterpotentials can be observed following other action potentials in the bursts. Transynaptically evoked bursts occurred in 4 19 synaptically coupled paws+ In one cell pair separated by a distance of about 100 #m (Fig. 7At) the excitatory PSP occurred after a delay of about 13 msec and was followed in 26 out of 57 trials by an action potential and a long rectangular lover 150 msecl depolarizing afterpotential terminated by a rather abrupt onset of the falling phase (Fig. 7A~). When the cell was hyperpolarized the amplitude of the rectangular response was larger and. in addition.
250
k
rnsec
I
i
t
C _Ppsy
PRE
-,
]~e~ I
Fig. 6. Chartwriter records of two burst firing patterns in the same cell. A shows burst firing elicited by intraceUular stimulation indicated by a bar below voltage recording. In B, POST, the burst is synaptically evoked by a spike on B, PRE, to direct stimulation again indicated by bar below+recording. The portion of the burst denoted by arrowheads in B is expanded 5-fold in C to clarify t h e d e p o l a d z i n g afterpotentials following spikes. The calibration+value of 250 msec in A applies to time base brackets in B and C. The duration of the event marker signal in B and C is 25 msec. Alltraces are displayed at the amplification shown in A. Culture age was 2t days and fetal gestational age was 14 days.
257
,
~lOmV
I 150msec
B
C
Fig. 7. Transynapticallyevoked burst firing in the cell labeled as POST by electrical stimulation of the cell labelled as PRE in A1 which is a phase-contrast photomicrograph taken after the microelectrodes were placed intracellularly. All electrical records are postsynaptic responses which begin after the coupling artifacts shown at the left hand margin of the trace. A3 was obtained at a resting potential of--50 mV whereas all other traces were recorded at a membrane potential of about --70 inV. The calibrating bar value of 30 msec should be used for As and C1,2,3. The calibrating bar value of 150 msec should be used for A~ and B~,,:,and B3. These data were recorded on FM tape at a recording speed of 3 in./sec which produces mild distortion of the waveform. This culture was 43 days old and was prepared from a fetus of 17 days gestational age. action potentials were superimposed to give a burst configuration to the response. The excitatory PSP sometimes triggered an action potential (Fig. 7B2 and 3, a n d C2 and 3), sometimes did n o t (Fig. 7B~ and C1), and sometimes appeared d u r i n g or after onset of the first action potential (Fig. 7B3, C3). Nevertheless the burst still occurred which suggested that the first action potential was n o t necessary for burst generation in this pair. However, the fact that n o bursting occurred in 22 trials in which the presynaptic action potential failed to evoke an excitatory PSP indicates that synaptic excitation, at least, is required for bursting in this cell. The excitatory PSP recorded from the soma may serve only as a m o n i t o r of synaptic i n p u t which m a y be distributed as well to dendritic sites for generation of burst firing. The m e a n burst d u r a t i o n was 331.4 ± 8.82 msec; the m e a n n u m b e r of action potentials per burst was 14.1 ~ 0.48.
258 Reciprocal interaction of varying complexity was observed in 3 out of 19 cell pairs. One such pair is shown in Fig. 4, B1 and B4, in which each member reciprocally evokes excitatory PSPs in the other; note persistence of the stereotyped doublet action potentials in Fig. 4B4 with intracellular stimulation as had also occurred with transynaptic stimulation in Fig. 4Al-.,~. On occasion, after the excita~/ory P s p (Fig. 4Bt upper trace), a series of potentials was generated (Fig. 4B2 ana a, upper traces) which in turn elicited potentials on the presynaptic cell (lower traces) and in Other cells could lead to reciprocal burst firing. DISCUSSION
The primary result from this work is that a wide diversity of electrical activity develops or has been maintained in parallel with the morphologic differentiation previously described for these dissociated cultures from fetal mouse hippocampus~. It is probable that many, if not most, of the intracellutar recordings in this study have been from neurons of pyramidal cell origin. Whether pyramidal cells which were postmitotic at the time of culture have been able to survive for two months or whether the majority of the ceils present at that time completed their final cell divisions shortly after being placed in culture has not yet been determined. Nevertheless. some features of the electrical recordings are strikingly similar to some of the characteristics of intact hippocampal pyramidal cells a,6.9. Spontaneously occurring action potentials which occurred singly or in short bursts without a depolarizing shift of the membrane potential were recorded from many cells. In these cells repetitive action potentials could be elicited by long pulses of depolarizing currents. Action potentials had fast rates of rise. were about I msec m duration, and had correspondingly fast rates of fall. In some cells the acticua potential was preceded by a brief depolarizing potential. These prepotentials were highly stereotyped, occurred with reliability in a gwen cell. did not have a large increase in amplitude on hyperpolarization, and were found under conditions of a high Mg to Ca ratio which otherwise abolished ongoing synaptic activity. On this basis it is concluded that these prepotentials are active electrical responses which are generated on portions of the membrane located on cell processes at a distance from the site of the microelectrode in the cell body. Additionally, fluorescent dye filling and electron mlcl'OScopy 5 strongly suggest the development of dendrites in these cells. On the basis of this combined electrical and morphologic evidence, these prepotentials very likely are generated on dendrites and possibly are equivalent to the fast prepotentials described by Spencer and KandeP °. but an axonal origin of the prepotentials should also be considered. In cultured hippocampal neurons, the initial dendritic branching points are often near the soma. as in Fig. ID. Thus functional parallels between neurons in culture and the intact mature hippocampus should be drawn with a caution similar to that previously discussed 7 in a comparison of the fast prepotentiats of immature and mature hippocampal pyramidal cells. Spontaneously occurring synaptic potentials can be recorded in up to 70 ~ of the neurons. This synaptic activity includes the frequent occurrence of inhibitory PSPs
259 with reversal potentials between --30 and --40 inV. Inhibitory synaptic activity is observed in medium containing glycine, a condition whereby glycine responses and presumably glycine-mediated inhibitory PSPs are not found in dissociated spinal cord cultures ~. Therefore glycine would seem an unlikely candidate for the inhibitory transmitter in hippocampal cultures and y-aminobutyric acid, which is not a component of the medium and which is a putative inhibitory transmitter in hippocampal pyramidal cells ~2, is a better candidate. Excitatory PSPs can trigger action potentials recorded from the cell body either with or without preceding dendritic action potentials suggesting that excitatory synaptic input occurs at both cell somata and cell dendrites. Ultrastructural evidence for these synaptic contacts has been discussed previouslyz. More complex synaptic potentials have been observed in which an excitatory PSP is abruptly terminated by onset of an inhibitory PSP. Such closely linked synaptic potentials suggest that synaptic contacts mediating those potentials are spaced near each other, perhaps on dendritic structures as illustrated in Fig. 10C of the preceding report 5. Transynaptic stimulation can elicit these complex synaptic potentials as well as elicit monophasic excitatory PSPs triggering action potentials and inhibitory PSPs with demonstrated reversal potentials. Given this abundance of synaptic activity it is not surprising that reciprocal innervation can be demonstrated in nerve cell pairs. A significant feature of the spontaneous electrical activity in hippocampal cultures is the widespread occurrence of bursting potentials with an accompanying depolarizing shift of the membrane potential. These bursts can be elicited by direct intracellular stimulation as well as by stimulation of a presynaptic cell. Bursting appears to be mediated by depolarizing afterpotentials which trigger additional action potentials in order to maintain the burst for periods lasting up to 800 msec. In those cases of transynaptic activation of burst firing, reciprocal bursting on the presynaptic cell has also been observed. Some features of burst firing in dissociated hippocampal cultures resemble that reported for cultured hippocampal explants from fetal mouse by Zipser and Crain la. In that study, burst firing occurred spontaneously, in response to stimulation of other neurons, and in the presence of low concentrations of strychnine and bicuculline, The use of Lucifer Yellow CH 11 as an agent for fluorescent dye staining, along with concomitant recording of electrical activity during the time that cell morphology is being disclosed, offers a powerful technique for identification of cell types based on criteria of structure and function. In addition, studies on innervation in these cultures will be greatly enhanced by the capability to locate synaptically coupled ceils which are contacted by the dye-filled neuron. The work of this study taken together with that of the preceding report 5 presents a survey of certain well-developed morphologic and electrophysiologic characteristics of dissociated hippocampal cultures from fetal mice. Among possible avenues for further work, this culture system offers opportunities for detailed study of dendritic structure and function, examination of mechanisms of burst firing in hippocampa[ neurons, and studies of synaptic interaction in a population probably dominated by pyramidal cells.
260 ACKNOWLEDGEMENTS This w o r k was s u p p o r t e d by N I H G r a n t N S 1215l. I t h a n k Dr. D a v i d Prince for his helpful review o f the m a n u s c r i p t and a m grateful to Dr. W a l t e r S te w a r d for his gift o f Lucifer Yellow C H , Reed Pike helped with p h o t o g r a p h y and Cheryl J o o typed the m an u scr i p t .
REFERENCES 1 Angevine, J. B.. Development of the hippocampal region. In R. L. Isaacson and K, H, Pribrana (Eds.), The Hippocampus, Vol. l, Plenum Press, New York, 1975. pp. 61-90. 2 Banker, G. A. and Cowan, W. M., Rat hippocampal neurons in dispersed cell cuIture, Brain Research. 126 (1977) 397-425. 3 Kandel, E. R. and Spencer, W. A., Electrophysiologic properties of an archicortical neuron. A n n N. Y. Acad. Sci.. 94 (1961) 570-603. 4 Nelson, P. G., Ransom, B. R.. Henkart. M. and Bullock, P. N.. Mouse spinal cord in celt culture. IV. Modulation of inhibitory synaptic function, J, Neurophysiol., 40 (1977) 1178-.1187~ 5 Peacock, J. H., Rush. D. F. and Mathers, L. H., Morphology of dissociated hippocampal cultures from fetal mice. Brain Research. 169 (1979) 247-260. 6 Prince, D. A., Neurophysiology of epilepsy, Ann. Rev. Neurosci., 1 (1978) 395-415. 7 Purpura. D. P.. Prelevic. S. and Santini, M.. Postsynaptic potentials and spike variations in the feline hippocampus during postnatal ontogenesis, Exp. Neurol.. 22 (1968) 408 422. 8 Ransom, B. R., Neale, E., Henkart, M., Bullock, P, N. and Nelson, P, G., Mouse spinal cord in cell culture. I. Morphology and intrinsic neuronal electrophysiologic properties. J. Neurophysiol., 40 (1977) t 132-1150. 9 Schwartzkroin, P. A., Characteristics of CAl neurons recorded intracellular[y in the hippocampaJ in vitro slice preparation, Brain Research, 85 (1975J 423-436. 10 Spencer, W. A. and Kandel, E. R., Electrophysiology ofhippocampal neurons. IV. kas~ prepotenrials, J. NeurophysioL, 24 (196lj 272-285. 11 Stewart, W. W.. Functional corrections between cells as revealed by dye-coupling with a highly fluorescent naphthalamide tracer. Cell, 14 (1978) 741-759. 12 Straughan, D. W., Neurotransmitter and the hippocampus. In R L. Isaacson and K. H, Pribram (Eds.), The Hippocampus, Vol. t. Plenum Press, New York. 1975. pp. 239-268. 13 Zipser, B., Crain, S. M. and Bornstein, M. B., Directly evoked "paroxysmal" depolarizations of mouse hippocampal neurons in synaptically organized explants in long-term culture. Brain Research, 60 ( 19731 489-495.
