- Email: [email protected]
Characteristics of CA1 activation through the hippocampal trisynaptic pathway in the unanaesthetized rat
Brain Research, 413 (1987) 75- 86
75
Elsevier BRE 12612
Characteristics of activation through the hippocampal trisynaptic pathway in the unanaesthetized rat O. Herreras, J.M. Solis, R. Martin del Rio and J. Lerma* Departamento de lnvestigaci6n, Centro Ram6n y Cajal, Madrid (Spain) (Accepted 28 October 1986)
Key words." Neuronal transmission; Trisynaptic circuit; Hippocampus; 0 Rhythm; Evoked potential
The hippocampal CA I field is activated by the entorhinal cortex mainly through the hippocampal excitatory trisynaptic circuit. Field responses of the CA~ region were evoked by ipsilateral CA 3 or perforant path volleys (mono- or trisynaptic activation, respectively) in paralyzed, locally anaesthetized rats and studied as a function of the stimulus patterns presented. The relationship of these responses with the concomitant EEG was also explored. Results showed that mono- and especially trisynaptically evoked responses were progressively enhanced by increasing the stimulus frequency from 0.1 to 1.0 Hz. At specific intensities the trisynaptically evoked population spike (PS) was present only with a rather fixed frequency of stimulation (--~0.5 Hz). PS was produced in 100~4 of the responses using 0.7 Hz, indicating the existence of a threshold-like level for this stimulus parameter. The frequency of presented paired pulses differentially affected pair-pulse facilitation of mono- and trisynaptically evoked excitatory postsynaptic potentials (EPSP): higher frequency decreased the former and increased the latter. All evoked responses studied (i.e. EPSP and PS) showed steep increments and decrements in amplitude, clearly developing several clusters. Moreover, the amplitude distribution of trisynaptic PS often varied spontaneously from maximal to negligible values, showing an all-or-none distribution. Clustering was interpreted as evidence of the existence in the hippocampus of functional neuronal aggregates. All-or-none distribution of trisynaptic PS was found to be associated with the EEG pattern. PS amplitude being maximal during irregular EEG activity and minimal during 0 rhythm. Present results suggest that (1) the entorhinal cortex may exert modulatory actions on CA 1 by a mechanism widely based on the frequency of the input; (2) information transfer from the entorhinal cortex to other brain areas throughout the hippocampus is biased by hippocampal EEG: and (3) electrotonic coupling may be functionally predominant in the hippocampus.
0 r h y t h m in this structure, abolishes C A I p o p u l a t i o n
INTRODUCTION Single as well as r e p e t i t i v e stimulation of hippo-
spikes 2~ and diminishes excitability t h r o u g h o u t a penicillin-induced epileptic focus 25.
c a m p a l p a t h w a y s m o d i f i e s the r e s p o n s e of h i p p o c a m -
Previous studies h a v e d e s c r i b e d the a n a t o m i c a l
pal n e u r o n s to s u b s e q u e n t testing with e i t h e r short-
and functional l a m e l l a r o r g a n i z a t i o n of intrinsic con-
or l o n g - t e r m t i m e courses. S e v e r a l types of short-
nections in the h i p p o c a m p u s 5"3°'37, including an exci-
t e r m m o n o s y n a p t i c alterations of n e u r o n a l transmis-
tatory polysynaptic circuit in which cortical i n f o r m a -
sion h a v e b e e n d e s c r i b e d in the rat h i p p o c a m p u s , the most widely studied b e i n g pair pulse facilitation 3'12" 28,40.47, f r e q u e n c y facilitation 2"3"712"18"36"47 and fre-
tion, e n t e r i n g via the p e r f o r a n t path, m a y be elabo r a t e d by the h i p p o c a m p u s . T h e s e l a m e l l a e b e c o m e functional w h e n e v e r a n u m b e r of granule cells of
q u e n c y defacilitation or h a b i t u a t i o n - l i k e changes 1'2 19.45,46. In addition, changes in h i p p o c a m p a l transmis-
d e n t a t e gyrus are activated b e c a u s e their axons, the
sion h a v e r e c e n t l y b e e n r e p o r t e d in relation to the hippocampal
turn may trigger the discharge of C A I n e u r o n s . Since the p e r f o r a n t p a t h w a y r e p r e s e n t s the m a j o r hippo-
E E G 11'22'26"~9. Sensory stimulation, which g e n e r a t e s
campal input w h e r e a s C A 1 m a y be c o n s i d e r e d the
animal's
behavioral
state
and
the
mossy fibers, excite C A 3 p y r a m i d a l cells, which in
* Present address: Instiluto de Neurobiologia S. Ram6n y Cajal, C.S.1.C. Vel~izquez 144, 28006-Madrid, Spain.
Correspondence: O. Herreras. Depto. Investigaci6n, Ctro. Ram6n y Cajal. Ctra. Colmenar Km 9, 28034-Madrid. Spain. 0006-8993/87/$03.50 © 1987 Elsevier Science Publishers B.V. (Biomedical Division)
76 hippocampal output, it should be useful to determine to what extent the patterns of impulses incoming to the hippocampus affect C A 1 sensitivity to these same volleys, considering that activation of CA 1 pyramidal cells from the entorhinal cortex occurs mostly through the hippocampal trisynaptic circuit 6,17. The aims of this study were thus to characterize the polysynaptic activation of CA 1 by the perforant pathway, and to determine this activation's degree of dependence on the ongoing EEG. Trisynaptic and monosynaptic CA I activation induced by ipsilateral stimuli were also compared. MATERIALS AND METHODS Thirty-four Sprague-Dawley rats weighing 250-300 g were used. Under continuous ether anaesthesia, the animals were placed on a stereotaxic frame and pressure points and surgical wounds infiltrated with the potent and long-acting local anaesthetic bupivacaine. The ear bars were removed after a metal plate, fastened to the stereotaxic instrument, was fixed to the skull with dental cement. Trephine holes were drilled in the skull at preselected stereotaxic coordinates. Stimulating electrodes were of the bipolar concentric type, formed by a 26G cannula (varnish-coated except at the cut surface) and a 100 j~m stainless steel inner wire (teflon-coated except 500 ktm at the tip), protruding 1 mm from the cannula. An electrolytically sharpened stainless steel, varnish-coated microelectrode (3-5 #m at the tip) was used to record CA~ evoked potentials. After electrode implantation, animals were paralyzed with i.p, D-tubocurarine (5 mg/kg b. wt.) and ether anaesthesia was discontinued. Mechanical ventilation was applied by means of a special nasal adaptor attached to the stereotaxic frame. The room was maintained quiet and in partial darkness to reduce uncontrolled sensory stimuli (see ref. 20). Heart rate was continuously monitored as an indicator of the preparation's stability and body temperature was kept constant with a heater plate. Doses of local anaesthetic were periodically infiltrated into surgical wounds in order to avoid possible distress. Control animals, maintained under these conditions for up to 6 h, showed no signs of discomfort since they sporadically displayed slow wave sleep episodes. The recording electrode was connected to a field
effect transistor, the output of which was amplified and band-pass filtered at 1-30 Hz to record the EEG and at 1-3000 Hz for evoked potentials. EEG was continuously monitored on paper by an electroencephalograph and evoked potentials were simult~neously displayed on the storage screen of an oscilloscope, Both EEG and evoked responses were taped for analysis. Stimulating electrodes were positioned on CA 3 and/ or on the perforant pathway to induce the mono- or trisynaptic activation of the (?Am field, respectively. Since electrodes were usually implanted on both areas in each animal, it was demonstrated that the CA:~ electrode did not alter the trisynaptic ('A~ activation. Fhe recording electrode was first placed in the pyramidal cell layer to study somatic potentials (i.e. population spike, PS) and subsequently lowered to the CA t stratum radiatum in order to record dendritic potentials (i.e. field excitatory postsynaptic potentials, EPSP). The arrangement of stimulating and recording electrodes is shown in Fig. 1A. Typical responses recorded at both somatic and dendritic levels, evoked by stimulating either the CA 3 or perforant pathway, are presented in Fig. lB. These field potentials have been shown to correspond to the extracellular current summation of synchronous unitary events evoked by an afferent volley 4'2v. Monosynaptic somatic responses (Fig. 1, R~-S 0 consisted of a sharp 4 - 6 ms latency negative wave (PS) preceded and followed by positive waves. Dendritic responses (Fig. 1, R2-S ~) appeared a~ a wider negative wave lasting 4-5 ms (EPSP). Trisynaptically evoked responses (Fig. 1, $2) showed longer latency (13-17 ms) and were preceded by a short-latency (3-4 ms) triphasic wave corresponding to the granule cell PS monosynaptically triggered by the perforant path input and recorded by volume conduction ¢''~7. The trisynaptic nature of these perforant path-evoked CAj responses has been reported previously 5'¢''17. Population spike amplitude was measured as the voltage from the shoulder of the first positive wave to the peak of maximal negativity. In dendritic records, the voltage from the baseline to the inaximal negativity was taken as the EPSP amplitude. In some experiments, EPSP was also evaluated as the maximum rate of rise (slope) of the negative-going wave. Since similar results were obtained by both methods, the former was chosen in order to facilitate and acceler-
77
A
R1
•
•
MONO
0 - - - - - - 0 TRI
~ . ~
e...~-.--~---~e
e
•
A
Sl
E
~-
t:E
a. W
/
2
*
.9 .~---o . . . . -0 /
I
0.1
Sl
05
S2
STIMULUS
04
0~5
INTIENSITY ( m A )
B
R1
B
0.1mA R2 J
0.4mA.~~@-"'~'~~
e ~
0.8mA~~-~~~ Fig. 1. A: schematic diagram of a hippocampal lamella showing the arrangement of recording and stimulating electrodes. Evoked field potentials were recorded from CA 1 stratum pyramidale (R 0 or CA I stratum radiatum (R2). Electrical stimuli were applied on the CA3 field (S0, to evoke the monosynaptic activation of CA l, or on the perforant pathway ($2), to induce CA] response trisynaptically. B: typical averaged CA l potentials (n = 20) evoked by stimuli applied to either CA 3 (S 0 or perforant pathway ($2) and recorded at either pyramidal (Rp population spike) or dendritic (R2, field EPSP) layers. Stimuli are indicated by black dots. Note in records showing trisynaptic activation ($2), the presence of an additional short-latency triphasic wave, corresponding to the monosynaptic-evoked potential propagated from the dentate gyms. Calibrations: 5 ms and 4 inV.
ate EPSP measurements. At the end of each experiment, the exact electrode locations were histologically verified using the Prussian blue method (see ref. 48). Experiment I was designed to determine the effect of the presentation pattern of afferent stimuli on both mono- and trisynaptic CA] responses. Afferent stimuli consisted of 0.1 ms square pulses, with frequency (0.1-1.0 Hz) and intensity (0.05-1.0 mA) steeply varied. In all stimulation series, first 10-15 records
6 ~ e
MONO
0-----o
TRI
i ~ .-,--'~'e ( 0 4 Hz )
A E
//
I-
I
111=0~1 (0 4 HI)
3E
1 y/ oz
04 STIMULUS
//i 11 ~11/ .'~" Ilit
06
do
INTENSITY (mA)
Fig. 2. Stimulus-response curves of mono- and trisynaptic activation of CA 1. Variations of field-EPSP (A) and population spike (B) amplitudes were measured as a function of stimulus intensity. Stimuli were delivered at 0.1 Hz in A and as indicated in B, bottom. Examples of averaged trisynaptic responses (n = 20) depicted at the top of B correspond to those used for the 0.5 Hz curve. Calibration: 5 ms, 2 inV.
78 were excluded to allow response stabilization (see Fig. 3A). T a p e d records were acquired in a programable digital oscilloscope (Tektronix 7D20) at a sampling rate of 2 0 - 5 0 kHz, and averages (n = 20) were obtained and p h o t o g r a p h e d . Tests were always performed during continuous irregular E E G activity ( I A ) , i.e. in absence of 0 rhythm, and were discontinued if a change in the E E G p a t t e r n was o b s e r v e d (see below). E x p e r i m e n t 11 was designed to study mono- and trisynaptic CA] activation in relation to different E E G patterns. Long-lasting records (up to 40 min) of simultaneously t a p e d E E G and e v o k e d potentials were analyzed. In this e x p e r i m e n t , stimulation p a r a meters for studying the PS were adjusted at an intensity of 2 x threshold and a frequency of 0 . 4 - 0 . 7 Hz, whereas the same frequencies with subthreshold in-
A
tensities were used for studying thc EPSP. Spontaneous time course of PS amplitude was represented~ and taped E E G was later divided into segments with stable and similar PS amplitudes. In o r d e r to characterize the E E G segments, autocorrelation functions ( A C F ) and p o w e r spectra (PwS) were obtained by a PDP-11 c o m p u t e r and interpreted as described previously 16'25. In addition, amplitude histograms of evoked responses were constructed for these longlasting records. RESULTS
Experiment 1. Dependence o/'the (:A t response on the stimulation parameters To analyze the C A I response e v o k e d either monoor trisynaptically, 30-35 single pulses were applied at
B
C
"••
I
0.4 Hz 0.8 Hz
I ,
,
E
D 100,
~__ IIO,
.
)i /
~
i"
-
~
.,..-~
~_
!
q[
.o
O/O~
/
//
~ |O
o---o
o'z o~4 o:6 STIMULUS FI~OUENG'Y (HI)
'm~
o~
~ X'
,'o
:
"
O~2
t/
, 0.4
STIMULUS
,(
-
.o,o
-
....
/ / /
¢
L -" O~S
FREQUENCY
:
, 011
TEl
-
io
(Hi)
Fig. 3. Effects of the variation in the stimulus frequency on mono- and trisynaptic CA l responses. A: oscilloscope traces of the first 8 trisynaptic responses, recorded at the CA] pyramidal layer, after starting the perforant pathway stimulation at 0~6 Hz: Note the gradual increase of population spike. B, C: averages (n = 20) of dendritic field EPSP (B) and somatic population spike(C) both trisynaptically evoked at different frequencies but with a constant stimulation intensity. Note that the short-latency triphasic wave showed an opposite behavior as did the trisynaptic population spike with the frequency. D: percentage of increase in field EPSP amplitude as a function of the stimulus frequency, taking as 100% that amplitude obtained at 1 Hz. E: variation of the amplitude (as percentage of the maximal population spike) and percentage of occurrence of the population spike with the frequency of stimulation. Calibrations: 5 ms, 2 mV.
79 each intensity and/or frequency of stimulation. Only the last 20 responses were used for m e a s u r e m e n t s because first responses showed considerable variability, gradually increasing in amplitude to a steady response, as seen clearly in trisynaptic records (Fig. 3A). The effect of stronger stimulus strength on monoand trisynaptically evoked potentials is p r e s e n t e d in Fig. 2. Monosynaptic responses (both EPSP and PS) saturated at lower intensities than trisynaptic responses. The threshold for monosynaptic PS was 2 - 5 times lower than that of trisynaptically e v o k e d PS, the latter being highly d e p e n d e n t on the stimulus frequency (Fig. 2B, bottom). E x a m p l e s of intensity dependence of trisynaptic PS e v o k e d at 3 different strengths of perforant path stimulus are shown at the top of Fig. 2B, where a concomitant e n h a n c e m e n t in the triphasic short-latency wave ( p r o p a g a t e d monosynaptic granule cell PS; see refs. 6, 17) is also evident. Stimulus frequency greatly affected trisynaptically evoked responses, whereas it only slightly influenced monosynaptic ones. In o r d e r to determine the effect of stimulus frequency on the EPSP, a preselected stimulus intensity was used, which always evoked 50% of the maximal amplitude EPSP (as d e t e r m i n e d at a frequency of 0.1 Hz, Fig. 2A). As seen in Fig. 3D, EPSP amplitude increased as a function of stimulus frequency. The effect of stimulus frequency on PS amplitude as well as on the percentage of spike occurrence is shown in Fig. 3E. The intensity of stimuli selected was the same as in the EPSP study, but 0.4 Hz were used, which in most cases e v o k e d measurable PS. 100% PS was obtained at 0.7 Hz. The modulation of the trisynaptic response by the stimulus frequency is more evident considering that the monosynaptic granule cell response (the short-latency triphasic wave in Fig. 3B, C; see refs. 1, 19, 45) decreased simultaneously. A clear pair pulse facilitation was observed at an interpulse interval of 60 ms with a wide range of stimulus intensities, the effect of frequency of pair presentation on both mono- and trisynaptic pair pulse facilitation was tested. F o r this study we used a selected stimulation intensity in such a way that at 0.4 Hz the test EPSP reached 50% of its maximal amplitude. Sixty ms of interpulse interval was chosen in order to diminish the effect of recurrent inhibition on the den-
tate gyrus (see ref. 28) during perforant path activation. Pair pulse facilitation of trisynaptic EPSP increased as a function of pair frequency, while the degree of facilitation of the monosynaptic C A I EPSP decreased (Fig. 4). A comparative study of pair pulse facilitation of mono- and trisynaptic PS could not be
A
o.,xz,
,x/'"
0.7Hz:'~~'4 B
0.1Hz 0.4Hz 0.7Hz
I 2SO -
:
C
= MONO
i
-"
U
IsO
~OO
o~,
d. STIMULUS
d.
o',
~'o
FRr~QUI[NCY ( H z )
Fig. 4. Effects of the variation in the stimulus frequencv on the pair pulse facilitation of mono- and trisynaptic field EF'SP. A, B: average (n = 20) of tri- (A) and monosynaptic (B) CA] responses obtained with paired pulses applied with an interpulse interval of 60 ms, at indicated frequencies. The used stimulus intensity selected was that which evoked a 50% maximal amplitude EPSP test determined at 0.4 Hz. C: frequency variation of the pair pulse facilitation of EPSP. Facilitation was estimated as the percentage of the test with respect to the conditioning response. Note the opposite frequency effect on facilitation of mono- and trisynaptic field EPSP. Calibrations: 10 ms, 2 inV.
80
1
°d
\
Twin pulse paradigm is usually employed to study recurrent inhibition by varying the interpulse interval. In the present study, this goal was not attempted because, during trisynaptic CA/ activation, pair pulse facilitation of the PS was strong enough to completely mask recurrent inhibition at any interval, and this inhibition was only detectable in the first few responses.
Experiment IL Relationship of CAt response with the ongoing EEG
8-
Fig. 5. Combination of consecutive single and twin pulses on the perforant pathway to illustrate the powerful facilitatory action of paired stimuli on the trisynaptic population spike. Once the response to single subthreshold stimuli was stabilized (1), 4 twin-pulses were applied (2-5), after which single stimulation was continued (6-9). Note that trisynaptic population spike appeared from the first twin pulse in the test response and even in the conditioning (4-5), outlasting the paired pulse stimulation (6). The arrow denotes the probable CA3 spike (see text), on which paired stimuli had a facilitatory effect. Stimulus frequency was 0.2 Hz. Calibration: 10 ms, 4 mV.
performed due to the special features that the trisynaptic PS showed for this plasticity. Twin pulses applied on the perforant pathway at both low intensity and low frequency (subthreshold for PS in control response) were enough to evoke a trisynaptic PS in the test response. In most cases (12 out of t5), after several twin pulses, there was also a fully developed PS in the control response which outlasted pair pulse stimulation. This phenomenon is shown in Fig. 5, in which a combination of consecutive single and twin pulses is presented. In this animal, the recording electrode was situated in a CAt zone near the CA 3 pyramidal layer. An additional 7-8 ms latency spike (arrow in the first record of Fig. 5), most probably generated in the CA 3 region, was concomitantly observed. This C A 3 spike (presumably disynaptically evoked) exhibited behavior similar to that of the trisynaptic CA1 population spike.
Long-lasting simultaneous records of hippocampal E E G and CA 1 field potentials, evoked at fixed parameters of stimulation (see Materials and Methods), were performed. During these long-lasting records, trisynaptic PS amplitude spontaneously varied usually from zero (no reading) to its maximal value, and there was a clear tendency toward clustering. Fig. 6 shows superpositions in the oscilloscope storage screen of trisynaptic potentials (top) and their corresponding PS amplitude histograms (bottom). In A, several clusters of the PS were evident in both oscilloscope records and histogram. In C, a clearly bimodal distribution of trisynaptic PS amplitude can be observed, and a mixed pattern of distribution is presented in B. These 3 patterns of amplitude distribution were often displayed by the same animal, depending on the E E G pattern present in the hippocampus (see below), and whenever clustering was evident (as shown in Fig. 6A), the minimum voltage interval between two clusters of amplitude was 400-600 pV. Although amplitude variation was less in monosynaptic PS than in trisynaptic PS, a slight tendency to clustering was also observed (not shown). This PS behavior was also observed in the field EPSP evoked either mono- (Fig. 7A) or trisynaptically (Fig. 7B). Previous studies in our laboratory have demonstrated a close relation between the hippocampal 0 rhythm, evoked by sensory stimulation, and the amplitude of CA 1 field responses z°. We therefore studied the possibility that the spontaneous amplitude variations observed in evoked potentials were related to spontaneous changes in the hippocampal E E G pattern. Fig. 8A shows the time course of trisynaptic PS amplitude with spontaneous distribution of PS amplitude similar to that in Fig. 6C. E E G was divided into segments, and labeled 1 for those during
81
A
B
."_ J
1751P
, , I
N=326
20 1
C
I•
N-307
~, ~ Amplitude
N=457
40
,,. ;
_J
_]
l
i~
.. . 2 4 e Amplitude
*mV'
o.,v
,
s s Amplitude
r.,v
Fig. 6. Spontaneous amplitude variations of trisynaptic CA t population spike• Superpositions of evoked potentials in the storage screen of an oscilloscope (upper row) and their corresponding amplitude histograms (lower row) are shown. For this and the following figure, the oscilloscope beam was greatly attenuated in intensity for better visualization of the slower and more repetitive waves of the evoked potentials. A, B and C are representative examples in 3 different animals showing tendency to amplitude clustering (A), all-ornone distribution (C) and a mixed pattern (B) of the CA 1populations spike. Arrowheads indicate the most frequent amplitude figures. Thin arrows indicate the maximal amplitude values. Number of superimposed records and samples represented in histograms were the same (N). Calibrations: 2 ms, 2 mV.
which PS amplitude was maximal, and lI for those during which no PS appeared in the evoked potential• Typical autocorrelation functions (ACF) and power spectra (PwS) obtained from each E E G segment are presented in Fig. 8 together with an example of raw
=
/
~30~1
Tl IL~T
N=269
=10
301 10
2
3 4 mV Amplitude
l l
|IT
N=343 II I ?
1 Amplitude 2 3
4 mV
Fig. 7. Spontaneous amplitude variations of mono- (A) and trisynaptic (B) field EPSP. Upper row shows the superposition of 269 (A) and 60 (B) records in the storage screen of an oscilloscope. The corresponding amplitude histograms are presented in the lower row. Arrowheads indicate the most frequent amplitude figures. 'N' denotes samples taken for histograms. In A, broken-line band indicates the loose traces of EPSPs. Black dots mark the instant of stimulation. Calibrations: 2 ms, 2 mV.
E E G and the corresponding evoked potentials. Both the rhythmic A C F and the prominent 4.7-7 Hz spectral bars in PwS of Fig. 8 demonstrated that during spontaneous hippocampal 0 rhythm, there was a drastic reduction in trisynaptic CA~ response, whereas amplitude was maximal when the hippocampus displayed irregular E E G activity (i.e. flat A C F and PwS). Similar relationships between E E G and monosynaptically evoked PS were observed, although these amplitude variations were less pronounced (not shown). DISCUSSION
The main results of this study demonstrate that CA~ population output, driven trisynaptically by stimulating the perforant pathway: (a) is highly dependent on stimulation frequency; (b) is clearly biased by the ongoing E E G , being maximal during IA and minimal during 0 rhythm; and (c) tends to show amplitude clustering• Monosynaptic CAI responses, provided by stimulating the CA 3 field, present similar but less marked characteristics.
82
10- A >8E 6
1
A
"O
II
• •
, ,'.
,
.
.-
..
,
•
.
"o
, ,
..
",*
• o,
•
,*
•
,
•
, •
.
..
•
..
.
•
...
.
,
.
."
.
*, °°
°
°
°"
°
,
4• °
°°
*
°
¢Z 2
E <0
°,#
- -
0
B
%
•
°
•
°
*
°o
I
°
.
°
2
3
EEG
°
...
'4
ACF
. :
°°
5 PwS
°°
6 min EP
/
t
-ls
i'
0
1
ls
125Hz
Fig. 8. Relationship between the trisynaptic population spike amplitude and the ongoing hippocampal EEG activity. A: spontaneous variation of population spike amplitude through time in a representative animal which had shown the same amplitude distribution presented in Fig. 6C. Note periods during which evoked potentials showed population spikes of maximal amplitude~l) and others during which these are absent (II). Examples of raw ongoing hippocampal EEG, their corresponding autocorrelation functions (ACF) and power spectra (PwS), and concomitant superimposed evoked potentials (EP) are presented in the upper (period I)and lower (period II) rows of B. During irregular EEG activity (fiat ACF and PwS), population spike was maximal while during O rhythm (rhythmic ACF and 4.2-7 Hz peak in PwS) no population spike appeared. Large spectral bars present in the PwS at 1.6-3.2 Hz were due to the slow EEG wave which followed stimulus artifact. Calibrations: 1 s and 300~V for EEG; 1.5 mW for ACF; 0.3 mW for PwS; 5 ms and 2 mV for EP. Some of the described features of monosynaptic responses are congruent with results previously rep o r t e d from in vitro or anaesthetized p r e p a r a t i o n s 2,3,7'12"28,36,45 and although m o d u l a t i o n s of trisynaptically e v o k e d responses may be a consequence of the former, to our k n o w l e d g e these findings have not previously been r e p o r t e d . T h e r e are, however, several differential characteristics which should be stated. Trisynaptic PS p r e s e n t e d a clear thresholdlike level for frequency not o b s e r v e d in the monosynaptic PS. Trisynaptic PS could usually not be e v o k e d with stimuli frequencies below 0.3 Hz, even with
quite high intensities but were fully d e v e l o p e d (Fig. 3) within a narrow range of stimulus frequency ( 0 . 3 - 0 . 7 Hz). The powerful facilitatory effect of gradually increased frequency on the C A t population spike occurred in spite of the concomitant decrease in granule cell e v o k e d discharge (Fig, 3; see also refs. 1, 2, 19.45.46). A l t h o u g h stimulus intensity obviously has a powerful effect on C A I responses (Fig. 2}, the entorhinal cortex may exert a strong m o d u l a t o r y action on C A l output by means of a mechanism based on the frequency pattern• These findings therefore clarify the functional rote of the hippocam-
83 pal excitatory trisynaptic circuit. While a massive synchronous discharge of granule cells evoked by perforant path volleys would not be enough to induce the trisynaptic CA t PS, an attenuated but more frequent evoked discharge of granule cells would be able to trigger the C A t . As shown (Fig. 3), frequency facilitation was much more effective on the CA 1 population spike evoked through the tri- than the monosynaptic pathway. Considering that CA3 responses (disynaptically evoked) appeared enhanced (Fig. 5; see also ref. 2) whereas dentate gyrus responses decreased with stimulus frequency, a frequency facilitatory mechanism similar to CA 1 may well occur in CA 3. Thus, the CA 3 field may participate in frequency facilitation of trisynaptically evoked CA l responses by acting as an amplifier. In addition, direct afferent fibers from the entorhinal cortex to CA 3 and CA t, which are part of the perforant pathway t°'21"3°'37, probably contribute to the frequency facilitation of trisynaptic CA t responses. It should be emphasized that although the perforant path is incapable of triggering CAt cells monosynaptically 6, it may depolarize the cell membrane, potentiating the delayed volleys (e.g. those entering to the CAt trisynaptically) to bring cells to discharge 6. In many studies frequencies of 0.5-1 Hz have been considered standard stimulus frequencies for monitoring control responses. Alger and Teyler 2, however, showed in in vitro experiments that background rates of stimulation (frequencies as low as 0.2 Hz) modified the responses. This finding together with our in vivo results indicate that frequencies above 0.2 Hz should not be used to test hippocampal control responses. Our result concerning the relationship between the E E G and the amplitude of evoked potentials (Fig. 8) is in agreement with other f i n d i n g s 2°'26'29. I n the presence of 0 rhythm, induced by sensory stimulation, a decrease in the CA1 evoked potentials occurs along with possible diminution of general excitability 2°,25. Studies of unitary hippocampal activity support the evidence of 0-related decrease of CA 1 evoked responses. For example, a firing reduction of complex spike cells, which presumably correspond to pyramidal neurons iS, has been shown during 0-associated behaviors while these cells increased their discharge during non-0 behaviors 13,35,3s,at. Preliminary results
showed that electrophysiologically identified CA 1 pyramidal cells greatly decrease their responses to Schaffer volleys during sensory-induced 0 rhythm (unpublished observations), indicating that the decrease in population spike amplitude during 0 rhythm is mainly the consequence of a reduction in the number of cells which fire after the stimulus. On the other hand, an increase in CA 1 PS evoked by commissural stimulation has been observed during some 0 behaviors L1. The contrasting result in this study could be the outcome of a different pathway being stimulated and the displayed hippocampal 0 pattern is known to have different behavioral correlates (see ref. 20, for detailed discussion). Variations of excitability in a specific neuronal population may actually be due to the influence of several converging pathways. It is difficult, at present, to know to what extent each afferent pathway contributes to the final output of a neuronal population. Our results suggest that perforant path input to the hippocampus may play an important role in the hippocampal output. Although frequency is not the only input parameter, it is undoubtly one of the most important. Since frequencies used to activate the perforant pathway in our study may occur spontaneously, modulation of CA 1 output could be exerted from the entorhinal cortex through the perforant pathway by slightly varying the rate of its ongoing discharge. Indeed, entorhinal pyramidal cells fire up to 20 spikes/s (see e.g. ref. 14). However, mechanisms underlying changes in population responses during different behavioral states or hippocampal E E G patterns remain to be clarified. Taking into account that evoked responses also change with the hippocampal EEG pattern 11,2°,26.49(Fig. 8), modulation of the population excitability by the input frequency and/or EEG pattern probably influence hippocampal modulation of behavioral processes, influencing information transfer from the entorhinal cortex to other brain areas. During spontaneous 0 rhythm, hippocampal output is closed, signifying a restricted transfer from the entorhinal cortex to other structures, whereas cortical information transfer would be fully accomplished when the hippocampus displays IA. The tendency toward clustering of CA 1 evoked potentials should have important functional implications. Since evoked potentials are the result of synchronous extracellular unitary currents, discrete var-
84 iations in the recorded voltage are probably due to the activation of groups of cells operating as functional unities. The mechanism by which a group of cells discharges as a functional unit is unclear. Several characteristics of the hippocampus, however, such as electrotonic coupling T M , ephaptic phenomena 23'39'43'44 o r local circuit interactions, i.e. recurrent excitation 24'3t, may account for the synchronization of a neuronal group, Recurrent excitation is probably of minor importance in evoking clustering because it requires a relatively long time to recruit neurons 24'3~. Although electrical field effects could contribute to the neuronal synchronization 23'39'43'44, their participation in the present phenomenon is as yet uncertain. Measurable transmembrane depolarizations (induced by field effect) occur whenever population spikes are elicited 39"43, however, field EPSPs, ew~ked by subthreshold stimuli, also showed clustering (Fig. 7). In addition, clustering of field EPSPs, recorded at somatic levels, occur without any population spike in the evoked potential (Fig. 8B). Nevertheless, the net result of electric field effects would be to reinforce and amplify any mechanism that triggers and synchronizes action potentials in a neuronal population, as Taylor and Dudek 43 have recently stressed. Electrotonic coupling of neurons T M provides the hippocampus with an important requisite to display such a phenomenon because inputs to the coupled cells, even though smoothly graded, induce them to fire synchronously (see ref. 9, for review) inasmuch as all of the coupled cells would receive these inputs. On the other hand, such a neuronal group becomes less excitable when an element of the group receives a hypothetical inhibitory action. Consequently, an input may or may not activate an aggregate of coupled cells, causing a discrete step in the recorded voltage. It is important to note that voltage intervals between
REFERENCES 1 Abraham, W.C. and Bliss, T.V.P., An analysis of the increase in granule cell excitability accompanying habituation in the dentate gyrus of the anesthetized rat, Brain Research, 331 (1985) 303-313. 2 Alger, B.E. and Teyler, T.J., Long-term and short-term plasticity in the CA 1, CA3 and dentate regions of the rat hippocampal slice, Brain Research, 110 (1976) 463-480. 3 Andersen, P., Interhippocampal impulses. II. Apical dendritic activation of CA t neurons, Acta Physiol. Scand,, 48 (1960) 463-480.
clusters were quite similar in most animals, indicating that a limited number of neuronal units are involved in each voltage step. The amount of neurons forming such functional aggregates is at present difficult to ascertain since the extracellular current generated by every neuron induces a different voltage at the recording electrode, depending on the relative distance between them. Data from experiments correlating the number of firing units and the PS size indicate that 4 neurons firing synchronously contribute about 500 ktV to the PS amplitude a. This finding together with dye-coupling results in the hippocampus s'32"3a may indicate that each step in clustering is generated by the synchronous activity of 3-5 neurons since intervals between PS clusters were 400600/~V (see Fig. 6). Clustering was more evident in trisynaptic than in monosynaptic potentials (Fig. 7). Electrotonic coupling has been observed in the 3 subfields forming the trisynaptic pathway (i.e. dentate gyrus, CA 3 and CAI) 8'33'34'42. Therefore, connections among the electrotonically coupled cell populations may explain this result. Finally, if the above interpretation is correct, it may be hypothesized that electrotonic coupling is functionally predominant in the hippocampus. Although most neurons are reportedly dye-coupled in pairs 8'32'34, it should be remembered, as discussed by Andrew et al. s, that dye-coupled cells may actually underestimate the number of electrically coupled neurons.
ACKNOWLEDGEMENTS We wish to thank to Ms. C. Stoddard Delgado for reviewing the English text. O.H. is recipient of a fellowship from the Spanish F.1.S.S.S, (85/632).
4 Andersen, P., Unit analysis of hippocampal population spikes, Exp. Brain Res., 13 (1971) 208-221. 5 Andersen, P., Bliss, T.V.P. and Skrede, K.K., Lamellar organization of hippocampal excitatory pathways, Exp. Brain Res., 13 (1971) 222-238. 6 Andersen, P., Holmqvist, B. and Voorhoeve, P.E.. Excitatory synapses on hippocampal apical dendrites activated by entorhinal stimulation, Acta Physiol. Scand.. 66 (1966) 461-472. 7 Andersen, P. and Lomo, T., Control of hippocampal output by afferent volley frequency, Progr, Brain Res.. 27 (1967) 400-412.
85 8 Andrew, R.D., Taylor, C.P., Snow, R.W. and Dudek, F.E., Coupling in rat hippocampal slices: dye transfer between CA I pyramidal cells, Brain Res. Bull., 8 (1982) 211-222. 9 Bennett, M.V.L., Electrical transmission: a functional analysis and comparison with chemical transmission. In E.R. Kandel (Ed.), Handbook of Physiology, Vol. 1, Section 1, Williams and Wilkins, Baltimore, MD, 1977, pp. 357-416. 10 Blackstad, T.W., On the termination of some afferents to the hippocampus and fascia dentata, Acta Anat. (Basel), 35 (1958) 202-214. 11 Buzs'aki, G., Grasty'an, E., Czopf, J., Kell6nyi, L. and Prohaska, O., Changes in neuronal transmission in the rat hippocampus during behavior. Brain Research, 225 (1981) 235-247. 12 Creager, R.. Dunwiddie, T. and Lynch, G., Paired-pulse and frequency facilitation in the CA 1 region of the in vitro rat hippocampus, J. Physiol. (London), 299 (1981)) 409-424. 13 Delacour, J., Associative and non-associative changes in unit activity of the rat hippocampus, Brain Res. Bull., 8 (1982) 367-373. 14 Finch, D.M. and Babb, T.L., Inhibition in subicular and entorhinal principal neurons in response to electrical stimulation of the fornix and hippocampus, Brain Research, 196 (1980) 89-98. 15 Fox, S.E. and Ranck, Jr. J.B., Electrophysiological characteristics of hippocampal complex-spike and theta cells, Exp. Brain Res., 41 ( 1981 ) 399-410. 16 Fuentes, J., Bufio, Jr. W. and Garcia-Austt, E., Simulation of postsynaptic activities in hippocampal cells during theta rhythm. Brain Res. Bull., 7 (1981) 157-162. 17 Fujita, Y., Synaptic activation of dentate granule cells and its effect upon pyramidal cells in rabbit. In Physiologie de l'ttippocampe, Coll. Int. C.N.R.S.. (Paris), 1962, pp. 47-69. 18 Gloor, P., Vera, C.L. and Sperti, L., Electrophysiological studies of hippocampal neurons. III. Responses of hippocampal neurons to repetitive perforant path volleys, Electroenceph. Clin. Neurophysiol., 17 (1964) 353-370. 19 Harris, E.W., Lasher, S.S. and Steward, O., Analysis of the habituation changes in transmission in the temporodentate pathway of the rat, Brain Research, 162 (1979) 21-32. 20 Herreras, O., Soils, J.M. and Lerma, J., Abolition of CA I population spike by sensory stimulation, Exp. Brain Res., 61 (1986) 654-657. 21 Hjort-Simonsen, A. and Jeune, B., Origin and termination of the hippocampal perforant path in the rat studied by silver impregnation, J. Comp. Neurol., 144 (1972) 215-232. 22 Jeantet, Y. and Jaffard, R., Relations quantitatives entre la densit~ spectrale de l'activit6 spontan~e et I'excitabilit~ des synapses commissurales du champ CA I de l'hippocampe dorsal chez la Souris 6veill~Se, C.R. Acad. Sci. Paris, t. 293 (1981) 545-548. 23 Jefferys, J.G.R., Influence of electric fields on the excitability of granule cells in guinea-pig hippocampal slice, J. Phystol. (London), 319 (1981) 143-152. 24 Lebovitz. R.M., Dichter, M. and Spencer. W.A., Recurrent excitation in the CA 3 region of cat hippocampus, Int. J. Neurosci., 2 (1971) 99-108. 25 Lerma. J., Herreras. O., Mufioz, D. and Soils, J.M., Interactions between hippocampal penicillin spikes and theta rhythm, Electroenceph. Clin. Neurophysiol., 57 (1984)
532-540. 26 Leung, L.S., Behavior-dependent evoked potentials in the rat hippocampal CA~ region of the rat. 1. Correlation with behavior and EEG, Brain Research, 198 (1980) 95-117. 27 Lomo, T., Patterns of activation in a monosynaptic cortical pathway: the perforant path input to the dentate area of the hippocampal formation, Exp. Brain Res., 12 (1971) 18-45. 28 Lomo, T., Potentiation of monosynaptic EPSPs in the perforant path-dentate granule cell synapse, Exp. Brain Res., 12 (1971) 46-63. 29 Lopes da Silva, F.H., Wadman, W.J., Arnolds, D.E.A.T., Veeken, C. and Holsheimer, J., Hippocampus: behavior and modulation of local circuits. In F. Reinoso-Suarez and C. Ajmone-Marsan (Eds.), Cortical Integration, Raven New York, 1984, pp. 147-170. 30 Lorente de N6, R., Studies on the structure of the cerebral cortex. II. Continuation of the study of the Ammonic system, J. Psychol. Neurol., 46 (1934) 113- 177. 31 MacVicar, B.A. and Dudek, F.E., Local synaptic circuits in rat hippocampus: interactions between pyramidal cells, Brain Research, 184 (1980) 220-223. 32 MacVicar, B.A. and Dudek, F.E., Dye-coupling between CA 3 pyramidal cells in slices of rat hippocampus, Brain Research, 196 (19811) 494-499. 33 MacVicar, B.A. and Dudek, F.E., Electrotonic coupling between pyramidal cells: a direct demonstration in rat hippocampal slices, Science, 213 (1981) 782 785. 34 MacVicar, B.A. and Dudek, F.E., Elcctrotonic coupling between granule cells of rat dentate gyrus: physiological and anatomical evidence, J. Neurophysiol.. 47 (1982) 579-592. 35 Mink, W.D., Best, P.J. and Olds, J., Neurons in paradoxical sleep and motivated behavior, Science, 158 (1967) 1335-1337. 36 Racine, R.J. and Milgram, N.W., Short-term potentiation phenomena in the rat limbic forebrain, Brain Research, 260 (1983) 201-216. 37 Ram6n y Cajal, S., Histologie du Systi'me Nerveuse de l'Homme et des VertebrOs, Vol. 1I, Maloine, Paris, 1911. 38 Ranck, Jr. J.B., Studies on single neurons in dorsal hippocampal formation and septum of unrestrained rats. I. Behavioral correlates and firing repertoires, Exp. Neurol., 41 (1973) 461-531. 39 Richardson, T.L., Turner, R.W. and Miller, J.J., Extracellular fields influence transmembrane potentials and synchronization of hippocampal neuronal activity, Brain Research, 294 (1984) 255-262. 40 Steward, O., White. W.F. and Cotman, C.W., Potentiation of the excitatory synaptic action of commissural, associational and entorhinal afferents to dentate granule cells, Brain Research, 134 (1977) 551-560. 41 Suzuki, S.S. and Smith, G.K., Single-cell activity and synchronous bursting in the rat hippocampus during waking behavior and sleep, Exp. Neurol., 89 (1985) 71-89. 42 Taylor, C.P. and Dudek, F.E., A physiological test for electrotonic coupling between CA t pyramidal cells in rat hippocampal slices, Brain Research, 235 (1982) 351-357. 43 Taylor, C.P. and Dudek, F.E., Excitation of hippocampal pyramidal cells by an electrical field effect. J. Neurophysiol., 52 (1984) 126-142. 44 Taylor, C.P., Krnjevic, K. and Ropert, N., Facilitation of hippocampal CA 3 pyramidal cell firing by electrical fields generated antidromically, Neuroscience, 11 (1984) 101-109.
86 45 Teyler, T.J. and Alger, B.E., Monosynaptic habituation in the vertebrate forebrain: the dentate gyrus examined in vitro, Brain Research, t 15 (1976) 413-425. 46 Turner, R.W, and Miller, J.J., Effects of extracellular calcium on low frequency induced potentiation and habituation in the in vitro hippocampal slice preparation, ('an. J. Physiol., 60 (1982) 266-275. 47 Turner, R.W., Richardson, T.L. and Miller, J.J., Ephaptic interactions contribute to paired pulse and frequency po-
tentiation of hippocampal field potentials, Exp. Brain Res., 54 (1984) 467-570. 48 Winson, J., Influence of raphe nuclei on neuronal transmission from perforant pathway through dentate gyrus, J? Neurophysiol., 44 (1980) 937-950. 49 Winson, J. and Abzug, C., Neuronal transmission through hippocampal pathways dependent on behavior. J. Neurophysiol.. 41 (1978) 716-732.
75
Elsevier BRE 12612
Characteristics of activation through the hippocampal trisynaptic pathway in the unanaesthetized rat O. Herreras, J.M. Solis, R. Martin del Rio and J. Lerma* Departamento de lnvestigaci6n, Centro Ram6n y Cajal, Madrid (Spain) (Accepted 28 October 1986)
Key words." Neuronal transmission; Trisynaptic circuit; Hippocampus; 0 Rhythm; Evoked potential
The hippocampal CA I field is activated by the entorhinal cortex mainly through the hippocampal excitatory trisynaptic circuit. Field responses of the CA~ region were evoked by ipsilateral CA 3 or perforant path volleys (mono- or trisynaptic activation, respectively) in paralyzed, locally anaesthetized rats and studied as a function of the stimulus patterns presented. The relationship of these responses with the concomitant EEG was also explored. Results showed that mono- and especially trisynaptically evoked responses were progressively enhanced by increasing the stimulus frequency from 0.1 to 1.0 Hz. At specific intensities the trisynaptically evoked population spike (PS) was present only with a rather fixed frequency of stimulation (--~0.5 Hz). PS was produced in 100~4 of the responses using 0.7 Hz, indicating the existence of a threshold-like level for this stimulus parameter. The frequency of presented paired pulses differentially affected pair-pulse facilitation of mono- and trisynaptically evoked excitatory postsynaptic potentials (EPSP): higher frequency decreased the former and increased the latter. All evoked responses studied (i.e. EPSP and PS) showed steep increments and decrements in amplitude, clearly developing several clusters. Moreover, the amplitude distribution of trisynaptic PS often varied spontaneously from maximal to negligible values, showing an all-or-none distribution. Clustering was interpreted as evidence of the existence in the hippocampus of functional neuronal aggregates. All-or-none distribution of trisynaptic PS was found to be associated with the EEG pattern. PS amplitude being maximal during irregular EEG activity and minimal during 0 rhythm. Present results suggest that (1) the entorhinal cortex may exert modulatory actions on CA 1 by a mechanism widely based on the frequency of the input; (2) information transfer from the entorhinal cortex to other brain areas throughout the hippocampus is biased by hippocampal EEG: and (3) electrotonic coupling may be functionally predominant in the hippocampus.
0 r h y t h m in this structure, abolishes C A I p o p u l a t i o n
INTRODUCTION Single as well as r e p e t i t i v e stimulation of hippo-
spikes 2~ and diminishes excitability t h r o u g h o u t a penicillin-induced epileptic focus 25.
c a m p a l p a t h w a y s m o d i f i e s the r e s p o n s e of h i p p o c a m -
Previous studies h a v e d e s c r i b e d the a n a t o m i c a l
pal n e u r o n s to s u b s e q u e n t testing with e i t h e r short-
and functional l a m e l l a r o r g a n i z a t i o n of intrinsic con-
or l o n g - t e r m t i m e courses. S e v e r a l types of short-
nections in the h i p p o c a m p u s 5"3°'37, including an exci-
t e r m m o n o s y n a p t i c alterations of n e u r o n a l transmis-
tatory polysynaptic circuit in which cortical i n f o r m a -
sion h a v e b e e n d e s c r i b e d in the rat h i p p o c a m p u s , the most widely studied b e i n g pair pulse facilitation 3'12" 28,40.47, f r e q u e n c y facilitation 2"3"712"18"36"47 and fre-
tion, e n t e r i n g via the p e r f o r a n t path, m a y be elabo r a t e d by the h i p p o c a m p u s . T h e s e l a m e l l a e b e c o m e functional w h e n e v e r a n u m b e r of granule cells of
q u e n c y defacilitation or h a b i t u a t i o n - l i k e changes 1'2 19.45,46. In addition, changes in h i p p o c a m p a l transmis-
d e n t a t e gyrus are activated b e c a u s e their axons, the
sion h a v e r e c e n t l y b e e n r e p o r t e d in relation to the hippocampal
turn may trigger the discharge of C A I n e u r o n s . Since the p e r f o r a n t p a t h w a y r e p r e s e n t s the m a j o r hippo-
E E G 11'22'26"~9. Sensory stimulation, which g e n e r a t e s
campal input w h e r e a s C A 1 m a y be c o n s i d e r e d the
animal's
behavioral
state
and
the
mossy fibers, excite C A 3 p y r a m i d a l cells, which in
* Present address: Instiluto de Neurobiologia S. Ram6n y Cajal, C.S.1.C. Vel~izquez 144, 28006-Madrid, Spain.
Correspondence: O. Herreras. Depto. Investigaci6n, Ctro. Ram6n y Cajal. Ctra. Colmenar Km 9, 28034-Madrid. Spain. 0006-8993/87/$03.50 © 1987 Elsevier Science Publishers B.V. (Biomedical Division)
76 hippocampal output, it should be useful to determine to what extent the patterns of impulses incoming to the hippocampus affect C A 1 sensitivity to these same volleys, considering that activation of CA 1 pyramidal cells from the entorhinal cortex occurs mostly through the hippocampal trisynaptic circuit 6,17. The aims of this study were thus to characterize the polysynaptic activation of CA 1 by the perforant pathway, and to determine this activation's degree of dependence on the ongoing EEG. Trisynaptic and monosynaptic CA I activation induced by ipsilateral stimuli were also compared. MATERIALS AND METHODS Thirty-four Sprague-Dawley rats weighing 250-300 g were used. Under continuous ether anaesthesia, the animals were placed on a stereotaxic frame and pressure points and surgical wounds infiltrated with the potent and long-acting local anaesthetic bupivacaine. The ear bars were removed after a metal plate, fastened to the stereotaxic instrument, was fixed to the skull with dental cement. Trephine holes were drilled in the skull at preselected stereotaxic coordinates. Stimulating electrodes were of the bipolar concentric type, formed by a 26G cannula (varnish-coated except at the cut surface) and a 100 j~m stainless steel inner wire (teflon-coated except 500 ktm at the tip), protruding 1 mm from the cannula. An electrolytically sharpened stainless steel, varnish-coated microelectrode (3-5 #m at the tip) was used to record CA~ evoked potentials. After electrode implantation, animals were paralyzed with i.p, D-tubocurarine (5 mg/kg b. wt.) and ether anaesthesia was discontinued. Mechanical ventilation was applied by means of a special nasal adaptor attached to the stereotaxic frame. The room was maintained quiet and in partial darkness to reduce uncontrolled sensory stimuli (see ref. 20). Heart rate was continuously monitored as an indicator of the preparation's stability and body temperature was kept constant with a heater plate. Doses of local anaesthetic were periodically infiltrated into surgical wounds in order to avoid possible distress. Control animals, maintained under these conditions for up to 6 h, showed no signs of discomfort since they sporadically displayed slow wave sleep episodes. The recording electrode was connected to a field
effect transistor, the output of which was amplified and band-pass filtered at 1-30 Hz to record the EEG and at 1-3000 Hz for evoked potentials. EEG was continuously monitored on paper by an electroencephalograph and evoked potentials were simult~neously displayed on the storage screen of an oscilloscope, Both EEG and evoked responses were taped for analysis. Stimulating electrodes were positioned on CA 3 and/ or on the perforant pathway to induce the mono- or trisynaptic activation of the (?Am field, respectively. Since electrodes were usually implanted on both areas in each animal, it was demonstrated that the CA:~ electrode did not alter the trisynaptic ('A~ activation. Fhe recording electrode was first placed in the pyramidal cell layer to study somatic potentials (i.e. population spike, PS) and subsequently lowered to the CA t stratum radiatum in order to record dendritic potentials (i.e. field excitatory postsynaptic potentials, EPSP). The arrangement of stimulating and recording electrodes is shown in Fig. 1A. Typical responses recorded at both somatic and dendritic levels, evoked by stimulating either the CA 3 or perforant pathway, are presented in Fig. lB. These field potentials have been shown to correspond to the extracellular current summation of synchronous unitary events evoked by an afferent volley 4'2v. Monosynaptic somatic responses (Fig. 1, R~-S 0 consisted of a sharp 4 - 6 ms latency negative wave (PS) preceded and followed by positive waves. Dendritic responses (Fig. 1, R2-S ~) appeared a~ a wider negative wave lasting 4-5 ms (EPSP). Trisynaptically evoked responses (Fig. 1, $2) showed longer latency (13-17 ms) and were preceded by a short-latency (3-4 ms) triphasic wave corresponding to the granule cell PS monosynaptically triggered by the perforant path input and recorded by volume conduction ¢''~7. The trisynaptic nature of these perforant path-evoked CAj responses has been reported previously 5'¢''17. Population spike amplitude was measured as the voltage from the shoulder of the first positive wave to the peak of maximal negativity. In dendritic records, the voltage from the baseline to the inaximal negativity was taken as the EPSP amplitude. In some experiments, EPSP was also evaluated as the maximum rate of rise (slope) of the negative-going wave. Since similar results were obtained by both methods, the former was chosen in order to facilitate and acceler-
77
A
R1
•
•
MONO
0 - - - - - - 0 TRI
~ . ~
e...~-.--~---~e
e
•
A
Sl
E
~-
t:E
a. W
/
2
*
.9 .~---o . . . . -0 /
I
0.1
Sl
05
S2
STIMULUS
04
0~5
INTIENSITY ( m A )
B
R1
B
0.1mA R2 J
0.4mA.~~@-"'~'~~
e ~
0.8mA~~-~~~ Fig. 1. A: schematic diagram of a hippocampal lamella showing the arrangement of recording and stimulating electrodes. Evoked field potentials were recorded from CA 1 stratum pyramidale (R 0 or CA I stratum radiatum (R2). Electrical stimuli were applied on the CA3 field (S0, to evoke the monosynaptic activation of CA l, or on the perforant pathway ($2), to induce CA] response trisynaptically. B: typical averaged CA l potentials (n = 20) evoked by stimuli applied to either CA 3 (S 0 or perforant pathway ($2) and recorded at either pyramidal (Rp population spike) or dendritic (R2, field EPSP) layers. Stimuli are indicated by black dots. Note in records showing trisynaptic activation ($2), the presence of an additional short-latency triphasic wave, corresponding to the monosynaptic-evoked potential propagated from the dentate gyms. Calibrations: 5 ms and 4 inV.
ate EPSP measurements. At the end of each experiment, the exact electrode locations were histologically verified using the Prussian blue method (see ref. 48). Experiment I was designed to determine the effect of the presentation pattern of afferent stimuli on both mono- and trisynaptic CA] responses. Afferent stimuli consisted of 0.1 ms square pulses, with frequency (0.1-1.0 Hz) and intensity (0.05-1.0 mA) steeply varied. In all stimulation series, first 10-15 records
6 ~ e
MONO
0-----o
TRI
i ~ .-,--'~'e ( 0 4 Hz )
A E
//
I-
I
111=0~1 (0 4 HI)
3E
1 y/ oz
04 STIMULUS
//i 11 ~11/ .'~" Ilit
06
do
INTENSITY (mA)
Fig. 2. Stimulus-response curves of mono- and trisynaptic activation of CA 1. Variations of field-EPSP (A) and population spike (B) amplitudes were measured as a function of stimulus intensity. Stimuli were delivered at 0.1 Hz in A and as indicated in B, bottom. Examples of averaged trisynaptic responses (n = 20) depicted at the top of B correspond to those used for the 0.5 Hz curve. Calibration: 5 ms, 2 inV.
78 were excluded to allow response stabilization (see Fig. 3A). T a p e d records were acquired in a programable digital oscilloscope (Tektronix 7D20) at a sampling rate of 2 0 - 5 0 kHz, and averages (n = 20) were obtained and p h o t o g r a p h e d . Tests were always performed during continuous irregular E E G activity ( I A ) , i.e. in absence of 0 rhythm, and were discontinued if a change in the E E G p a t t e r n was o b s e r v e d (see below). E x p e r i m e n t 11 was designed to study mono- and trisynaptic CA] activation in relation to different E E G patterns. Long-lasting records (up to 40 min) of simultaneously t a p e d E E G and e v o k e d potentials were analyzed. In this e x p e r i m e n t , stimulation p a r a meters for studying the PS were adjusted at an intensity of 2 x threshold and a frequency of 0 . 4 - 0 . 7 Hz, whereas the same frequencies with subthreshold in-
A
tensities were used for studying thc EPSP. Spontaneous time course of PS amplitude was represented~ and taped E E G was later divided into segments with stable and similar PS amplitudes. In o r d e r to characterize the E E G segments, autocorrelation functions ( A C F ) and p o w e r spectra (PwS) were obtained by a PDP-11 c o m p u t e r and interpreted as described previously 16'25. In addition, amplitude histograms of evoked responses were constructed for these longlasting records. RESULTS
Experiment 1. Dependence o/'the (:A t response on the stimulation parameters To analyze the C A I response e v o k e d either monoor trisynaptically, 30-35 single pulses were applied at
B
C
"••
I
0.4 Hz 0.8 Hz
I ,
,
E
D 100,
~__ IIO,
.
)i /
~
i"
-
~
.,..-~
~_
!
q[
.o
O/O~
/
//
~ |O
o---o
o'z o~4 o:6 STIMULUS FI~OUENG'Y (HI)
'm~
o~
~ X'
,'o
:
"
O~2
t/
, 0.4
STIMULUS
,(
-
.o,o
-
....
/ / /
¢
L -" O~S
FREQUENCY
:
, 011
TEl
-
io
(Hi)
Fig. 3. Effects of the variation in the stimulus frequency on mono- and trisynaptic CA l responses. A: oscilloscope traces of the first 8 trisynaptic responses, recorded at the CA] pyramidal layer, after starting the perforant pathway stimulation at 0~6 Hz: Note the gradual increase of population spike. B, C: averages (n = 20) of dendritic field EPSP (B) and somatic population spike(C) both trisynaptically evoked at different frequencies but with a constant stimulation intensity. Note that the short-latency triphasic wave showed an opposite behavior as did the trisynaptic population spike with the frequency. D: percentage of increase in field EPSP amplitude as a function of the stimulus frequency, taking as 100% that amplitude obtained at 1 Hz. E: variation of the amplitude (as percentage of the maximal population spike) and percentage of occurrence of the population spike with the frequency of stimulation. Calibrations: 5 ms, 2 mV.
79 each intensity and/or frequency of stimulation. Only the last 20 responses were used for m e a s u r e m e n t s because first responses showed considerable variability, gradually increasing in amplitude to a steady response, as seen clearly in trisynaptic records (Fig. 3A). The effect of stronger stimulus strength on monoand trisynaptically evoked potentials is p r e s e n t e d in Fig. 2. Monosynaptic responses (both EPSP and PS) saturated at lower intensities than trisynaptic responses. The threshold for monosynaptic PS was 2 - 5 times lower than that of trisynaptically e v o k e d PS, the latter being highly d e p e n d e n t on the stimulus frequency (Fig. 2B, bottom). E x a m p l e s of intensity dependence of trisynaptic PS e v o k e d at 3 different strengths of perforant path stimulus are shown at the top of Fig. 2B, where a concomitant e n h a n c e m e n t in the triphasic short-latency wave ( p r o p a g a t e d monosynaptic granule cell PS; see refs. 6, 17) is also evident. Stimulus frequency greatly affected trisynaptically evoked responses, whereas it only slightly influenced monosynaptic ones. In o r d e r to determine the effect of stimulus frequency on the EPSP, a preselected stimulus intensity was used, which always evoked 50% of the maximal amplitude EPSP (as d e t e r m i n e d at a frequency of 0.1 Hz, Fig. 2A). As seen in Fig. 3D, EPSP amplitude increased as a function of stimulus frequency. The effect of stimulus frequency on PS amplitude as well as on the percentage of spike occurrence is shown in Fig. 3E. The intensity of stimuli selected was the same as in the EPSP study, but 0.4 Hz were used, which in most cases e v o k e d measurable PS. 100% PS was obtained at 0.7 Hz. The modulation of the trisynaptic response by the stimulus frequency is more evident considering that the monosynaptic granule cell response (the short-latency triphasic wave in Fig. 3B, C; see refs. 1, 19, 45) decreased simultaneously. A clear pair pulse facilitation was observed at an interpulse interval of 60 ms with a wide range of stimulus intensities, the effect of frequency of pair presentation on both mono- and trisynaptic pair pulse facilitation was tested. F o r this study we used a selected stimulation intensity in such a way that at 0.4 Hz the test EPSP reached 50% of its maximal amplitude. Sixty ms of interpulse interval was chosen in order to diminish the effect of recurrent inhibition on the den-
tate gyrus (see ref. 28) during perforant path activation. Pair pulse facilitation of trisynaptic EPSP increased as a function of pair frequency, while the degree of facilitation of the monosynaptic C A I EPSP decreased (Fig. 4). A comparative study of pair pulse facilitation of mono- and trisynaptic PS could not be
A
o.,xz,
,x/'"
0.7Hz:'~~'4 B
0.1Hz 0.4Hz 0.7Hz
I 2SO -
:
C
= MONO
i
-"
U
IsO
~OO
o~,
d. STIMULUS
d.
o',
~'o
FRr~QUI[NCY ( H z )
Fig. 4. Effects of the variation in the stimulus frequencv on the pair pulse facilitation of mono- and trisynaptic field EF'SP. A, B: average (n = 20) of tri- (A) and monosynaptic (B) CA] responses obtained with paired pulses applied with an interpulse interval of 60 ms, at indicated frequencies. The used stimulus intensity selected was that which evoked a 50% maximal amplitude EPSP test determined at 0.4 Hz. C: frequency variation of the pair pulse facilitation of EPSP. Facilitation was estimated as the percentage of the test with respect to the conditioning response. Note the opposite frequency effect on facilitation of mono- and trisynaptic field EPSP. Calibrations: 10 ms, 2 inV.
80
1
°d
\
Twin pulse paradigm is usually employed to study recurrent inhibition by varying the interpulse interval. In the present study, this goal was not attempted because, during trisynaptic CA/ activation, pair pulse facilitation of the PS was strong enough to completely mask recurrent inhibition at any interval, and this inhibition was only detectable in the first few responses.
Experiment IL Relationship of CAt response with the ongoing EEG
8-
Fig. 5. Combination of consecutive single and twin pulses on the perforant pathway to illustrate the powerful facilitatory action of paired stimuli on the trisynaptic population spike. Once the response to single subthreshold stimuli was stabilized (1), 4 twin-pulses were applied (2-5), after which single stimulation was continued (6-9). Note that trisynaptic population spike appeared from the first twin pulse in the test response and even in the conditioning (4-5), outlasting the paired pulse stimulation (6). The arrow denotes the probable CA3 spike (see text), on which paired stimuli had a facilitatory effect. Stimulus frequency was 0.2 Hz. Calibration: 10 ms, 4 mV.
performed due to the special features that the trisynaptic PS showed for this plasticity. Twin pulses applied on the perforant pathway at both low intensity and low frequency (subthreshold for PS in control response) were enough to evoke a trisynaptic PS in the test response. In most cases (12 out of t5), after several twin pulses, there was also a fully developed PS in the control response which outlasted pair pulse stimulation. This phenomenon is shown in Fig. 5, in which a combination of consecutive single and twin pulses is presented. In this animal, the recording electrode was situated in a CAt zone near the CA 3 pyramidal layer. An additional 7-8 ms latency spike (arrow in the first record of Fig. 5), most probably generated in the CA 3 region, was concomitantly observed. This C A 3 spike (presumably disynaptically evoked) exhibited behavior similar to that of the trisynaptic CA1 population spike.
Long-lasting simultaneous records of hippocampal E E G and CA 1 field potentials, evoked at fixed parameters of stimulation (see Materials and Methods), were performed. During these long-lasting records, trisynaptic PS amplitude spontaneously varied usually from zero (no reading) to its maximal value, and there was a clear tendency toward clustering. Fig. 6 shows superpositions in the oscilloscope storage screen of trisynaptic potentials (top) and their corresponding PS amplitude histograms (bottom). In A, several clusters of the PS were evident in both oscilloscope records and histogram. In C, a clearly bimodal distribution of trisynaptic PS amplitude can be observed, and a mixed pattern of distribution is presented in B. These 3 patterns of amplitude distribution were often displayed by the same animal, depending on the E E G pattern present in the hippocampus (see below), and whenever clustering was evident (as shown in Fig. 6A), the minimum voltage interval between two clusters of amplitude was 400-600 pV. Although amplitude variation was less in monosynaptic PS than in trisynaptic PS, a slight tendency to clustering was also observed (not shown). This PS behavior was also observed in the field EPSP evoked either mono- (Fig. 7A) or trisynaptically (Fig. 7B). Previous studies in our laboratory have demonstrated a close relation between the hippocampal 0 rhythm, evoked by sensory stimulation, and the amplitude of CA 1 field responses z°. We therefore studied the possibility that the spontaneous amplitude variations observed in evoked potentials were related to spontaneous changes in the hippocampal E E G pattern. Fig. 8A shows the time course of trisynaptic PS amplitude with spontaneous distribution of PS amplitude similar to that in Fig. 6C. E E G was divided into segments, and labeled 1 for those during
81
A
B
."_ J
1751P
, , I
N=326
20 1
C
I•
N-307
~, ~ Amplitude
N=457
40
,,. ;
_J
_]
l
i~
.. . 2 4 e Amplitude
*mV'
o.,v
,
s s Amplitude
r.,v
Fig. 6. Spontaneous amplitude variations of trisynaptic CA t population spike• Superpositions of evoked potentials in the storage screen of an oscilloscope (upper row) and their corresponding amplitude histograms (lower row) are shown. For this and the following figure, the oscilloscope beam was greatly attenuated in intensity for better visualization of the slower and more repetitive waves of the evoked potentials. A, B and C are representative examples in 3 different animals showing tendency to amplitude clustering (A), all-ornone distribution (C) and a mixed pattern (B) of the CA 1populations spike. Arrowheads indicate the most frequent amplitude figures. Thin arrows indicate the maximal amplitude values. Number of superimposed records and samples represented in histograms were the same (N). Calibrations: 2 ms, 2 mV.
which PS amplitude was maximal, and lI for those during which no PS appeared in the evoked potential• Typical autocorrelation functions (ACF) and power spectra (PwS) obtained from each E E G segment are presented in Fig. 8 together with an example of raw
=
/
~30~1
Tl IL~T
N=269
=10
301 10
2
3 4 mV Amplitude
l l
|IT
N=343 II I ?
1 Amplitude 2 3
4 mV
Fig. 7. Spontaneous amplitude variations of mono- (A) and trisynaptic (B) field EPSP. Upper row shows the superposition of 269 (A) and 60 (B) records in the storage screen of an oscilloscope. The corresponding amplitude histograms are presented in the lower row. Arrowheads indicate the most frequent amplitude figures. 'N' denotes samples taken for histograms. In A, broken-line band indicates the loose traces of EPSPs. Black dots mark the instant of stimulation. Calibrations: 2 ms, 2 mV.
E E G and the corresponding evoked potentials. Both the rhythmic A C F and the prominent 4.7-7 Hz spectral bars in PwS of Fig. 8 demonstrated that during spontaneous hippocampal 0 rhythm, there was a drastic reduction in trisynaptic CA~ response, whereas amplitude was maximal when the hippocampus displayed irregular E E G activity (i.e. flat A C F and PwS). Similar relationships between E E G and monosynaptically evoked PS were observed, although these amplitude variations were less pronounced (not shown). DISCUSSION
The main results of this study demonstrate that CA~ population output, driven trisynaptically by stimulating the perforant pathway: (a) is highly dependent on stimulation frequency; (b) is clearly biased by the ongoing E E G , being maximal during IA and minimal during 0 rhythm; and (c) tends to show amplitude clustering• Monosynaptic CAI responses, provided by stimulating the CA 3 field, present similar but less marked characteristics.
82
10- A >8E 6
1
A
"O
II
• •
, ,'.
,
.
.-
..
,
•
.
"o
, ,
..
",*
• o,
•
,*
•
,
•
, •
.
..
•
..
.
•
...
.
,
.
."
.
*, °°
°
°
°"
°
,
4• °
°°
*
°
¢Z 2
E <0
°,#
- -
0
B
%
•
°
•
°
*
°o
I
°
.
°
2
3
EEG
°
...
'4
ACF
. :
°°
5 PwS
°°
6 min EP
/
t
-ls
i'
0
1
ls
125Hz
Fig. 8. Relationship between the trisynaptic population spike amplitude and the ongoing hippocampal EEG activity. A: spontaneous variation of population spike amplitude through time in a representative animal which had shown the same amplitude distribution presented in Fig. 6C. Note periods during which evoked potentials showed population spikes of maximal amplitude~l) and others during which these are absent (II). Examples of raw ongoing hippocampal EEG, their corresponding autocorrelation functions (ACF) and power spectra (PwS), and concomitant superimposed evoked potentials (EP) are presented in the upper (period I)and lower (period II) rows of B. During irregular EEG activity (fiat ACF and PwS), population spike was maximal while during O rhythm (rhythmic ACF and 4.2-7 Hz peak in PwS) no population spike appeared. Large spectral bars present in the PwS at 1.6-3.2 Hz were due to the slow EEG wave which followed stimulus artifact. Calibrations: 1 s and 300~V for EEG; 1.5 mW for ACF; 0.3 mW for PwS; 5 ms and 2 mV for EP. Some of the described features of monosynaptic responses are congruent with results previously rep o r t e d from in vitro or anaesthetized p r e p a r a t i o n s 2,3,7'12"28,36,45 and although m o d u l a t i o n s of trisynaptically e v o k e d responses may be a consequence of the former, to our k n o w l e d g e these findings have not previously been r e p o r t e d . T h e r e are, however, several differential characteristics which should be stated. Trisynaptic PS p r e s e n t e d a clear thresholdlike level for frequency not o b s e r v e d in the monosynaptic PS. Trisynaptic PS could usually not be e v o k e d with stimuli frequencies below 0.3 Hz, even with
quite high intensities but were fully d e v e l o p e d (Fig. 3) within a narrow range of stimulus frequency ( 0 . 3 - 0 . 7 Hz). The powerful facilitatory effect of gradually increased frequency on the C A t population spike occurred in spite of the concomitant decrease in granule cell e v o k e d discharge (Fig, 3; see also refs. 1, 2, 19.45.46). A l t h o u g h stimulus intensity obviously has a powerful effect on C A I responses (Fig. 2}, the entorhinal cortex may exert a strong m o d u l a t o r y action on C A l output by means of a mechanism based on the frequency pattern• These findings therefore clarify the functional rote of the hippocam-
83 pal excitatory trisynaptic circuit. While a massive synchronous discharge of granule cells evoked by perforant path volleys would not be enough to induce the trisynaptic CA t PS, an attenuated but more frequent evoked discharge of granule cells would be able to trigger the C A t . As shown (Fig. 3), frequency facilitation was much more effective on the CA 1 population spike evoked through the tri- than the monosynaptic pathway. Considering that CA3 responses (disynaptically evoked) appeared enhanced (Fig. 5; see also ref. 2) whereas dentate gyrus responses decreased with stimulus frequency, a frequency facilitatory mechanism similar to CA 1 may well occur in CA 3. Thus, the CA 3 field may participate in frequency facilitation of trisynaptically evoked CA l responses by acting as an amplifier. In addition, direct afferent fibers from the entorhinal cortex to CA 3 and CA t, which are part of the perforant pathway t°'21"3°'37, probably contribute to the frequency facilitation of trisynaptic CA t responses. It should be emphasized that although the perforant path is incapable of triggering CAt cells monosynaptically 6, it may depolarize the cell membrane, potentiating the delayed volleys (e.g. those entering to the CAt trisynaptically) to bring cells to discharge 6. In many studies frequencies of 0.5-1 Hz have been considered standard stimulus frequencies for monitoring control responses. Alger and Teyler 2, however, showed in in vitro experiments that background rates of stimulation (frequencies as low as 0.2 Hz) modified the responses. This finding together with our in vivo results indicate that frequencies above 0.2 Hz should not be used to test hippocampal control responses. Our result concerning the relationship between the E E G and the amplitude of evoked potentials (Fig. 8) is in agreement with other f i n d i n g s 2°'26'29. I n the presence of 0 rhythm, induced by sensory stimulation, a decrease in the CA1 evoked potentials occurs along with possible diminution of general excitability 2°,25. Studies of unitary hippocampal activity support the evidence of 0-related decrease of CA 1 evoked responses. For example, a firing reduction of complex spike cells, which presumably correspond to pyramidal neurons iS, has been shown during 0-associated behaviors while these cells increased their discharge during non-0 behaviors 13,35,3s,at. Preliminary results
showed that electrophysiologically identified CA 1 pyramidal cells greatly decrease their responses to Schaffer volleys during sensory-induced 0 rhythm (unpublished observations), indicating that the decrease in population spike amplitude during 0 rhythm is mainly the consequence of a reduction in the number of cells which fire after the stimulus. On the other hand, an increase in CA 1 PS evoked by commissural stimulation has been observed during some 0 behaviors L1. The contrasting result in this study could be the outcome of a different pathway being stimulated and the displayed hippocampal 0 pattern is known to have different behavioral correlates (see ref. 20, for detailed discussion). Variations of excitability in a specific neuronal population may actually be due to the influence of several converging pathways. It is difficult, at present, to know to what extent each afferent pathway contributes to the final output of a neuronal population. Our results suggest that perforant path input to the hippocampus may play an important role in the hippocampal output. Although frequency is not the only input parameter, it is undoubtly one of the most important. Since frequencies used to activate the perforant pathway in our study may occur spontaneously, modulation of CA 1 output could be exerted from the entorhinal cortex through the perforant pathway by slightly varying the rate of its ongoing discharge. Indeed, entorhinal pyramidal cells fire up to 20 spikes/s (see e.g. ref. 14). However, mechanisms underlying changes in population responses during different behavioral states or hippocampal E E G patterns remain to be clarified. Taking into account that evoked responses also change with the hippocampal EEG pattern 11,2°,26.49(Fig. 8), modulation of the population excitability by the input frequency and/or EEG pattern probably influence hippocampal modulation of behavioral processes, influencing information transfer from the entorhinal cortex to other brain areas. During spontaneous 0 rhythm, hippocampal output is closed, signifying a restricted transfer from the entorhinal cortex to other structures, whereas cortical information transfer would be fully accomplished when the hippocampus displays IA. The tendency toward clustering of CA 1 evoked potentials should have important functional implications. Since evoked potentials are the result of synchronous extracellular unitary currents, discrete var-
84 iations in the recorded voltage are probably due to the activation of groups of cells operating as functional unities. The mechanism by which a group of cells discharges as a functional unit is unclear. Several characteristics of the hippocampus, however, such as electrotonic coupling T M , ephaptic phenomena 23'39'43'44 o r local circuit interactions, i.e. recurrent excitation 24'3t, may account for the synchronization of a neuronal group, Recurrent excitation is probably of minor importance in evoking clustering because it requires a relatively long time to recruit neurons 24'3~. Although electrical field effects could contribute to the neuronal synchronization 23'39'43'44, their participation in the present phenomenon is as yet uncertain. Measurable transmembrane depolarizations (induced by field effect) occur whenever population spikes are elicited 39"43, however, field EPSPs, ew~ked by subthreshold stimuli, also showed clustering (Fig. 7). In addition, clustering of field EPSPs, recorded at somatic levels, occur without any population spike in the evoked potential (Fig. 8B). Nevertheless, the net result of electric field effects would be to reinforce and amplify any mechanism that triggers and synchronizes action potentials in a neuronal population, as Taylor and Dudek 43 have recently stressed. Electrotonic coupling of neurons T M provides the hippocampus with an important requisite to display such a phenomenon because inputs to the coupled cells, even though smoothly graded, induce them to fire synchronously (see ref. 9, for review) inasmuch as all of the coupled cells would receive these inputs. On the other hand, such a neuronal group becomes less excitable when an element of the group receives a hypothetical inhibitory action. Consequently, an input may or may not activate an aggregate of coupled cells, causing a discrete step in the recorded voltage. It is important to note that voltage intervals between
REFERENCES 1 Abraham, W.C. and Bliss, T.V.P., An analysis of the increase in granule cell excitability accompanying habituation in the dentate gyrus of the anesthetized rat, Brain Research, 331 (1985) 303-313. 2 Alger, B.E. and Teyler, T.J., Long-term and short-term plasticity in the CA 1, CA3 and dentate regions of the rat hippocampal slice, Brain Research, 110 (1976) 463-480. 3 Andersen, P., Interhippocampal impulses. II. Apical dendritic activation of CA t neurons, Acta Physiol. Scand,, 48 (1960) 463-480.
clusters were quite similar in most animals, indicating that a limited number of neuronal units are involved in each voltage step. The amount of neurons forming such functional aggregates is at present difficult to ascertain since the extracellular current generated by every neuron induces a different voltage at the recording electrode, depending on the relative distance between them. Data from experiments correlating the number of firing units and the PS size indicate that 4 neurons firing synchronously contribute about 500 ktV to the PS amplitude a. This finding together with dye-coupling results in the hippocampus s'32"3a may indicate that each step in clustering is generated by the synchronous activity of 3-5 neurons since intervals between PS clusters were 400600/~V (see Fig. 6). Clustering was more evident in trisynaptic than in monosynaptic potentials (Fig. 7). Electrotonic coupling has been observed in the 3 subfields forming the trisynaptic pathway (i.e. dentate gyrus, CA 3 and CAI) 8'33'34'42. Therefore, connections among the electrotonically coupled cell populations may explain this result. Finally, if the above interpretation is correct, it may be hypothesized that electrotonic coupling is functionally predominant in the hippocampus. Although most neurons are reportedly dye-coupled in pairs 8'32'34, it should be remembered, as discussed by Andrew et al. s, that dye-coupled cells may actually underestimate the number of electrically coupled neurons.
ACKNOWLEDGEMENTS We wish to thank to Ms. C. Stoddard Delgado for reviewing the English text. O.H. is recipient of a fellowship from the Spanish F.1.S.S.S, (85/632).
4 Andersen, P., Unit analysis of hippocampal population spikes, Exp. Brain Res., 13 (1971) 208-221. 5 Andersen, P., Bliss, T.V.P. and Skrede, K.K., Lamellar organization of hippocampal excitatory pathways, Exp. Brain Res., 13 (1971) 222-238. 6 Andersen, P., Holmqvist, B. and Voorhoeve, P.E.. Excitatory synapses on hippocampal apical dendrites activated by entorhinal stimulation, Acta Physiol. Scand.. 66 (1966) 461-472. 7 Andersen, P. and Lomo, T., Control of hippocampal output by afferent volley frequency, Progr, Brain Res.. 27 (1967) 400-412.
85 8 Andrew, R.D., Taylor, C.P., Snow, R.W. and Dudek, F.E., Coupling in rat hippocampal slices: dye transfer between CA I pyramidal cells, Brain Res. Bull., 8 (1982) 211-222. 9 Bennett, M.V.L., Electrical transmission: a functional analysis and comparison with chemical transmission. In E.R. Kandel (Ed.), Handbook of Physiology, Vol. 1, Section 1, Williams and Wilkins, Baltimore, MD, 1977, pp. 357-416. 10 Blackstad, T.W., On the termination of some afferents to the hippocampus and fascia dentata, Acta Anat. (Basel), 35 (1958) 202-214. 11 Buzs'aki, G., Grasty'an, E., Czopf, J., Kell6nyi, L. and Prohaska, O., Changes in neuronal transmission in the rat hippocampus during behavior. Brain Research, 225 (1981) 235-247. 12 Creager, R.. Dunwiddie, T. and Lynch, G., Paired-pulse and frequency facilitation in the CA 1 region of the in vitro rat hippocampus, J. Physiol. (London), 299 (1981)) 409-424. 13 Delacour, J., Associative and non-associative changes in unit activity of the rat hippocampus, Brain Res. Bull., 8 (1982) 367-373. 14 Finch, D.M. and Babb, T.L., Inhibition in subicular and entorhinal principal neurons in response to electrical stimulation of the fornix and hippocampus, Brain Research, 196 (1980) 89-98. 15 Fox, S.E. and Ranck, Jr. J.B., Electrophysiological characteristics of hippocampal complex-spike and theta cells, Exp. Brain Res., 41 ( 1981 ) 399-410. 16 Fuentes, J., Bufio, Jr. W. and Garcia-Austt, E., Simulation of postsynaptic activities in hippocampal cells during theta rhythm. Brain Res. Bull., 7 (1981) 157-162. 17 Fujita, Y., Synaptic activation of dentate granule cells and its effect upon pyramidal cells in rabbit. In Physiologie de l'ttippocampe, Coll. Int. C.N.R.S.. (Paris), 1962, pp. 47-69. 18 Gloor, P., Vera, C.L. and Sperti, L., Electrophysiological studies of hippocampal neurons. III. Responses of hippocampal neurons to repetitive perforant path volleys, Electroenceph. Clin. Neurophysiol., 17 (1964) 353-370. 19 Harris, E.W., Lasher, S.S. and Steward, O., Analysis of the habituation changes in transmission in the temporodentate pathway of the rat, Brain Research, 162 (1979) 21-32. 20 Herreras, O., Soils, J.M. and Lerma, J., Abolition of CA I population spike by sensory stimulation, Exp. Brain Res., 61 (1986) 654-657. 21 Hjort-Simonsen, A. and Jeune, B., Origin and termination of the hippocampal perforant path in the rat studied by silver impregnation, J. Comp. Neurol., 144 (1972) 215-232. 22 Jeantet, Y. and Jaffard, R., Relations quantitatives entre la densit~ spectrale de l'activit6 spontan~e et I'excitabilit~ des synapses commissurales du champ CA I de l'hippocampe dorsal chez la Souris 6veill~Se, C.R. Acad. Sci. Paris, t. 293 (1981) 545-548. 23 Jefferys, J.G.R., Influence of electric fields on the excitability of granule cells in guinea-pig hippocampal slice, J. Phystol. (London), 319 (1981) 143-152. 24 Lebovitz. R.M., Dichter, M. and Spencer. W.A., Recurrent excitation in the CA 3 region of cat hippocampus, Int. J. Neurosci., 2 (1971) 99-108. 25 Lerma. J., Herreras. O., Mufioz, D. and Soils, J.M., Interactions between hippocampal penicillin spikes and theta rhythm, Electroenceph. Clin. Neurophysiol., 57 (1984)
532-540. 26 Leung, L.S., Behavior-dependent evoked potentials in the rat hippocampal CA~ region of the rat. 1. Correlation with behavior and EEG, Brain Research, 198 (1980) 95-117. 27 Lomo, T., Patterns of activation in a monosynaptic cortical pathway: the perforant path input to the dentate area of the hippocampal formation, Exp. Brain Res., 12 (1971) 18-45. 28 Lomo, T., Potentiation of monosynaptic EPSPs in the perforant path-dentate granule cell synapse, Exp. Brain Res., 12 (1971) 46-63. 29 Lopes da Silva, F.H., Wadman, W.J., Arnolds, D.E.A.T., Veeken, C. and Holsheimer, J., Hippocampus: behavior and modulation of local circuits. In F. Reinoso-Suarez and C. Ajmone-Marsan (Eds.), Cortical Integration, Raven New York, 1984, pp. 147-170. 30 Lorente de N6, R., Studies on the structure of the cerebral cortex. II. Continuation of the study of the Ammonic system, J. Psychol. Neurol., 46 (1934) 113- 177. 31 MacVicar, B.A. and Dudek, F.E., Local synaptic circuits in rat hippocampus: interactions between pyramidal cells, Brain Research, 184 (1980) 220-223. 32 MacVicar, B.A. and Dudek, F.E., Dye-coupling between CA 3 pyramidal cells in slices of rat hippocampus, Brain Research, 196 (19811) 494-499. 33 MacVicar, B.A. and Dudek, F.E., Electrotonic coupling between pyramidal cells: a direct demonstration in rat hippocampal slices, Science, 213 (1981) 782 785. 34 MacVicar, B.A. and Dudek, F.E., Elcctrotonic coupling between granule cells of rat dentate gyrus: physiological and anatomical evidence, J. Neurophysiol.. 47 (1982) 579-592. 35 Mink, W.D., Best, P.J. and Olds, J., Neurons in paradoxical sleep and motivated behavior, Science, 158 (1967) 1335-1337. 36 Racine, R.J. and Milgram, N.W., Short-term potentiation phenomena in the rat limbic forebrain, Brain Research, 260 (1983) 201-216. 37 Ram6n y Cajal, S., Histologie du Systi'me Nerveuse de l'Homme et des VertebrOs, Vol. 1I, Maloine, Paris, 1911. 38 Ranck, Jr. J.B., Studies on single neurons in dorsal hippocampal formation and septum of unrestrained rats. I. Behavioral correlates and firing repertoires, Exp. Neurol., 41 (1973) 461-531. 39 Richardson, T.L., Turner, R.W. and Miller, J.J., Extracellular fields influence transmembrane potentials and synchronization of hippocampal neuronal activity, Brain Research, 294 (1984) 255-262. 40 Steward, O., White. W.F. and Cotman, C.W., Potentiation of the excitatory synaptic action of commissural, associational and entorhinal afferents to dentate granule cells, Brain Research, 134 (1977) 551-560. 41 Suzuki, S.S. and Smith, G.K., Single-cell activity and synchronous bursting in the rat hippocampus during waking behavior and sleep, Exp. Neurol., 89 (1985) 71-89. 42 Taylor, C.P. and Dudek, F.E., A physiological test for electrotonic coupling between CA t pyramidal cells in rat hippocampal slices, Brain Research, 235 (1982) 351-357. 43 Taylor, C.P. and Dudek, F.E., Excitation of hippocampal pyramidal cells by an electrical field effect. J. Neurophysiol., 52 (1984) 126-142. 44 Taylor, C.P., Krnjevic, K. and Ropert, N., Facilitation of hippocampal CA 3 pyramidal cell firing by electrical fields generated antidromically, Neuroscience, 11 (1984) 101-109.
86 45 Teyler, T.J. and Alger, B.E., Monosynaptic habituation in the vertebrate forebrain: the dentate gyrus examined in vitro, Brain Research, t 15 (1976) 413-425. 46 Turner, R.W, and Miller, J.J., Effects of extracellular calcium on low frequency induced potentiation and habituation in the in vitro hippocampal slice preparation, ('an. J. Physiol., 60 (1982) 266-275. 47 Turner, R.W., Richardson, T.L. and Miller, J.J., Ephaptic interactions contribute to paired pulse and frequency po-
tentiation of hippocampal field potentials, Exp. Brain Res., 54 (1984) 467-570. 48 Winson, J., Influence of raphe nuclei on neuronal transmission from perforant pathway through dentate gyrus, J? Neurophysiol., 44 (1980) 937-950. 49 Winson, J. and Abzug, C., Neuronal transmission through hippocampal pathways dependent on behavior. J. Neurophysiol.. 41 (1978) 716-732.