- Email: [email protected]
Detection of an atropine-resistant component of the hippocampal theta rhythm in urethane-anesthetized rats
Brain Research, 500 (1989) 55-60 Elsevier
55
BRES 14871
Detection of an atropine-resistant component of the hippocampal theta rhythm in urethane-anesthetized rats Mark Stewart and Steven E. Fox Department of Physiology, SUNY ttealth Science Center, Brooklyn, NY 11203 (U.S.A.
(Accepted 14 March 1989) Key words: Medial septum; Rhythmic cell; Acetylcholine: 7-Aminobutyric acid
An important pharmacological feature of the hippocampal 0 rhythm in urethane-anesthetized animals is tts apparent sensitivity to antimuscarinic drugs. This sensitivity may be partly due to a masking of the 0 frequency by increases in both higher and lower frequency EEG components that are unrelated to any residual 0 rhythm. The discovery of atropine-resistant, rhythmic medial septal neurons has provided a physiological trigger for averaging EEG and unit activity after large atropine doses. Such averaging has permitted the detection of an atropine-resistant component of the hippocampa[ 0 rhythm in urethane-anesthetized rats. The postatropine 0 activity recorded from both CA1 (superficial to the pyramidal cell layer) and dentate (near the hippocampal fissure) in 15 rats was typically reduced in amplitude, but the recordings from the two locations maintained their phase relations to the septal units and to each other. The presence of this residual 0 component after doses as large as 100 mg/kg indicates that it cannot be mediated by muscarinic cholinergic receptors. The coupling of the signal to the atropine-resistant septal cells strengthens our previous suggestion that these septo-hippocampal neurons are not cholinergic, and are therefore probably GABAergic.
INTRODUCTION The E E G of the h i p p o c a m p a l formation is characterized by a very regular 4 - 1 2 Hz sinusoidal pattern that reflects the synchronous activity of a great many synapses. This activity, referred to as the 0 rhythm or rhythmical slow activity, can be recorded with depth electrodes in awake and urethane-anesthetized animals. The 0 rhythm r e c o r d e d during urethane anesthesia has been characterized by its sensitivity to large doses of antimuscarinic drugs v'm. A n a t o m i c a l l y , both cholinergic and G A B A e r g i c s e p t o - h i p p o c a m p a l neurons have been identified 1' 2,s,~.ls Of the electrophysiologically identifiable cells of the medial septal nuclei, only atropineresistant and atropine-sensitive rhythmically bursting cells have been found to be driven antidromically from the h i p p o c a m p u s 1+'. We have p r o p o s e d that there is a c o r r e s p o n d e n c e between cholinergic and atropine-sensitive cells on the one hand and G A B A -
ergic and atropine-resistant cells on the o t h e r ~<'. This conclusion is the simplest explanation for a great variety of data, Some of the m o r e i m p o r t a n t points for making this argument are r e i t e r a t e d here. The direct application of a t r o p i n e only to the septum can apparently eliminate the h i p p o c a m p a l 0 rhythm t6. This indicates that a change in the firing pattern of atropine-sensitive septal cells, from rhythmic to non-rhythmic, can eliminate the 0 rhythm, in spite of unblocked cholinergic receptors within the hippocampus. The continued rhythmic activity of atropine-resistant septal cells alone, must not be sufficient to organize h i p p o c a m p a l cell activity which would result in a rhythmic E E G , nor do the atropine-resistant cells a p p e a r to be able to directly generate much of a rhythmic E E G signal. This is consistent with recent anatomical evidence that the G A B A e r g i c s e p t o - h i p p o c a m p a l p r o j e c t i o n is almost exclusively onto interneurons ('. It occurred to us that some c o m p o n e n t of the 0
(Jorrespondence: M. Stewart,Department of Physiology,Box 31, SUNY Health Science Center, 450 Clarkson Avenue, Brooklyn, NY 11203, U.S.A.
1i006-8993/89/$03.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)
36 rhythm might remain after atropine, masked by increases in Large-amplitude irregular activity in the EEG. We attempted to detect any contribution of the atropine-resistant septal neurons to the EEG by using the continued rhythmic activity of these cells as a trigger for averaging the post-atropine hippocampal EEG. MATERIALS AND METHODS Male Sprague-Dawley albino rats (310-450 g) were anesthetized with urethane (1.0 g/kg i.v. or 1.5 g/kg i.p.). Each of the 15 rats was held in a Kopf stereotaxic frame and maintained at 37 °C with an isothermal heating pad (Braintree Scientific, Braintree, MA) and/or shielded heating lamp. Electrode implants were made using stereotaxic coordinates 13. Extracellular recordings were taken from rhythmically bursting medial septal neurons with microelectrodes (1-2 Mr'2 stainless steel, Microprobe, Ciarksburg, MD) that were lowered through the septal nuclei (AP level 9.2-9.8 mm anterior to interaural line) at an angle of 20° off the vertical in the coronal plane. Hippocampal E E G was recorded monopolarly from each of a pair of stainless steel wires (125 ~m diameter) that were varnish insulated and cut square. The tips were vertically separated by about 1 mm and referred to a screw over cerebellum. The E E G electrode pair was lowered vertically (AP 4.7, ML 3.0) until the 0 rhythm recorded on the deeper electrode was phase-reversed with respect to that on the superficial electrode, and maximal in amplitude. The deeper electrode tip was then near the hippocampal fissure ( referred to as 'dentate' EEG) and the superficial electrode was dorsal to the pyramidal cell layer of CA1 (referred to as 'CAI' EEG) 7. On occasion, the deeper electrode was positioned closer to the granule cell layer. This deeper location was apparent in the EEG as a decrease in 0 rhythm amplitude from positions nearer the fissure and an increase in amplitude of the higher frequencies (30-50 Hz). Electrode locations for all units (one per rat) were marked at the ends of the experiments by passing current through the recording electrodes and visualizing the individual lesions with the Prussian blue reaction in Nissl-stained 40-/~m frozen sections. All units were confirmed to be within the medial septal
nucleus-nucleus of the diagonal band (vertical limb). Atropine sulfate (Sigma) was prepared as a 25 or 50 mg/ml solution in water for i.v. (25-100 mg/kg) or i.p.(50 mg/kg) injection. Postatropine recordings of septal unit activity and concurrent EEG were started at least 5 min after the full dose of atropine had been given and lasted for 10 rain to 1 h. Unit activity and E E G recordings were amplified, filtered (single units: bandpass = 500 Hz to 10 kHz; -24 dB/octave high pass, -6 dB/octave low pass. EEG: bandpass = 0.1 Hz to 10 kHz; -6 dB/octave rolloffs), and stored on tape for off-line analysis. A Hewlett Packard 9836U computer system was used for the construction of magnitude spectra as well as spike- and sine-wave triggered histograms. For magnitude spectra computations, taped data were lowpass filtered (100 Hz) and digitized at 200 Hz. Spectral averages were computed using the fastFourier transform (FFT) on 61 Hann-windowed segments of 512 data points, overlapped by 50%. A detailed description of the construction of phase histograms has been published 5. For spike-triggered histograms, a 45-ms duration rolling average of the discriminated and digitized septal unit spike train was computed. Theta scores 1~', estimates of rhythmic quality based on autocorrelations of cosine-windowed 1.024-s epochs of EEG made every 64 ms, were used to identify periods of 0 frequency activity. Peaks of identified '0 cycles' in the converted spike train or in a simultaneously digitized, independent sine wave were detected, and acceptable 0 cycles were those whose periods fell within pre-set time windows (typically 176-272 ms). Cycles were averaged by normalizing slow-wave amplitude data and spike data collected during each 0 cycle, redistributing them into 32 histogram bins. Final scaling yielded the slow-wave amplitude at each 11.25 ° segment of the average 0 cycle. RESULTS Fifteen rats received massive doses of systemic atropine. Doses ranged from 25-100 mg/kg i.v. (one rat received 50 mg/kg i.p.). The hippocampal 0 rhythm appeared to be completely eliminated when judged by inspection of the raw E E G (Fig. 1) or by frequency analysis (Fig. 2). Close inspection of the
57
pre- and post-atropine magnitude spectra indicates that the 0 activity might in fact be masked by large increases in both higher and lower frequencies after atropine. U n f o r t u n a t e l y , frequency analyses or autocorrelation methods cannot easily resolve this question. We used the spike trains of atropine-resistant
10
S
i',
c
CA 1 EEG
O -O
05
~
X
?re Atropine -- Post Atropine
N O
E
septal cells as a trigger to average the hippocampal 10
E E G . This method permitted the detection of a residual c o m p o n e n t of the 0 rhythm after doses of atropine as large as 100 mg/kg i.v. (Fig. 3, top and middle). We refer to this residual signal as 0 'activity' or as a ' c o m p o n e n t ' of the 0 rhythm for several
(b
1.0
7
30
40
I,
;"
DENTATE EEG
i
c
20
O 05 ~O ® N
.,_.
...... Post-Atropine
O
E
Pre-Atropine
0
J _ [ill ..l.J i,[, li.,l[ _ lliU . I]iJ.;, ,IILL_..IL ]1-- IN IF ~ll ll'[~ " ]l~ F'] IF1 i' ,
~
',-,,,,
~4~ ~r
10
20
30
40
Frequency(Hz) Fig. 2. Magnitude spectral plots for the CA1 (top) and dentate (bottom) EEGs, before and after atropine injection. Eighty seconds of EEG for each of the two conditions illustrated in Fig.1 were used for these magnitude spectra. The loss of 0 rhythm is seen in these overlaid plots to potentially result, at least in part, from a ' masking' of the 0 frequency by increases in lower and higher frequencies. For clear illustration of the low frequency components, only the first 102 of the 256 real frequency points are plotted. Normalized magnitudes: CA1 EEG - 1.0 = 0.46 mV/bin; Dentate EEG - 1.0 = (I.41 mV/bin. Bin width = 0.39 Hz.
~,t
Post-Atropine
reasons: (a) the 0 frequency is not obvious in the raw E E G or magnitude spectra; (b) as we will indicate, I
,I
'! ~ r]
'
I
Fig. 1. Septal unit and hippocampal EEG activity before and after 100 mg/kg i.v. atropine sulfate. The filtered septal unit recording is the top trace of each panel. The middle trace is EEG recorded superficial to the pyramidal cell layer of CA1 ('CAI' EEG) and the bottom trace is EEG recorded deep to the hippocampal fissure ('dentate' EEG). Normally dentate EEG recordings were taken near the amplitude maximum which is close to the hippocampal fissure. This recording was taken deeper than usual as evidenced by the relatively smaller amplitude compared with the CA1 recording and the large amount of high frequency activity. The top panel shows 0 rhythm in the hippocampal EEG along with rhythmic bursting in the scptal unit. The bottom panel shows the apparent absence of hippocampal 0 rhythm and continued rhythmicity in the septal unit, after atropine injection. Calibrations: l mV for the unit, 0.2 mV for both EEG channels, and 0.5 s. Positivity is up.
the relative amplitudes of CA1 and dentate recordings are not preserved; and (c) the rhythmic contribution of atropine-sensitive septal cells has been eliminated. The average n u m b e r of 0 cycles collected per rat was 351 cycles pre-atropine (S.D. = 275, range = 39 - 1166, median = 333) and 339 cycles post-atropine (S.D. = 277, range -- 127 1160, median --- 258). In order to address the possibility of recovering a spurious sinusoid by averaging with a periodic trigger, we triggered the averages of the E E G and the septal unit with a sine wave whose period was within + 10 ms of the average bursting period of the units , and matched the total n u m b e r of averaged cycles ( Fig. 3, bottom). The larger amplitude of the spike-triggered average compared to the sine wave-triggered aver-
58
CA1 EEG
":'~ -01
"
+0.1
DT EEG (mY)
i
"'_'"
"
.~
-0.1
(SP'UNylITbin)o~ r J ....... ~ 1 ~ ~ ' • 0
90
180
Lt~-: 270
360
+0.05]
CA1
I
~
VLo.osJ
-
+0.05
. - .
EEG (mY) -0.05 0.63' MS-DB UNIT (sp/cy/bin) 0 C
0
90
180
270
360
0
90
180
270
360
+0.05 CA1 EEG (mVl
-0.05 +0.05 DT EEG (mY)
-0,05 0,17~
MS-DB
UNIT (sp/cy/bin) 0
PHASE{deg) Fig. 3. Detection of postatropine urethane 0 activity by averaging the hippocampal EEG, triggered on the rhythmic bursting of the atropine-resistant septal unit. Dotted lines in the histograms are 1/4 cycles of the averaged EEG or spike histogram redrawn on either end for clarity. Panel A shows the averaged CA1 and dentate EEGs triggered on the septal unit, before atropine injection. Peak-to-peak amplitudes of each channel are approximately 200/~ V. Number of cycles in the average is 238. Mean bursting frequency for the unit was 4.4 Hz. Panel B shows the averaged EEG, triggered on the unit, after atropine. Note the 0 cycle amplitudes are diminished, but their phase relations to the unit and to each other are maintained. Number of cycles in the average is 475. Mean bursting frequency for the unit was 4.8 Hz. Panel C shows the averages of EEG activity triggered on a simultaneously sampled 4.6 Hz sine wave. Number of cycles in the average is 475. Positivity is up for the EEG averages.
age and the fact that the phases of the two E E G records were essentially identical to the p r e - a t r o p i n e condition only in the spike-triggered average, indicate that the recovered 0 rhythm is not an artifact of the averaging method. It is difficult to provide much quantitative d a t a on the residual 0 c o m p o n e n t , since there are only two recording locations and they are similar, but not identical across rats. We will however, present some crude quantitative data and some interesting observations. The fraction of the p e a k - t o - p e a k 0 cycle amplitude remaining after a t r o p i n e was 0.44 for CA1 recordings (n = 15, S.D. = 0.34, range = 0.09- 1.25, median = 0.27) and 0.29 for d e n t a t e recordings (n = 15, S.D. = 0.13, range = 0.12 - 0.50, median = 0.28). The phase of the positive p e a k relative to the spike trigger shifted an average of 6.8 ° (n = 15, S.D. = 47.3) in CA1 and 28.5 ° (n = 15, S.D. = 41.1) in the dentate. The average shifts were in the same direction for the two locations: increased lag of the 0 activity relative to the unit bursts. The p e a k - t o - p e a k amplitude of the averaged post-atropine 0 cycle was larger than 2 S . E . M . of the averaged 0 cycle in 10 of 15 rats for C A I 0 rhythm and 13 of 15 rats for d e n t a t e 0 rhythm. W h e n only these rats were used, means and S.D. for the fraction of the p e a k - t o - p e a k a m p l i t u d e preserved after atropine were essentially identical to those values obtained when all animals were included. The amplitude of the residual 0 c o m p o n e n t is clearly reduced c o m p a r e d to the p r e - a t r o p i n e condition. We e m p h a size however, that these values are only estimates, and are likely to be overestimates of the residual 0 component. There was no relation b e t w e e n the a m p l i t u d e of the residual 0 activity and the dose of atropine. It should also be n o t e d that the dentate 0 activity (typically r e c o r d e d at the level of the h i p p o c a m p a l fissure) was a t t e n u a t e d m o r e than the 0 activity r e c o r d e d in CA1 (dorsal to the p y r a m i d a l cell layer), but when the d e n t a t e E E G e l e c t r o d e was closer to the granule cell layer, a larger fraction of the p r e - a t r o p i n e amplitude a p p e a r e d to be preserved. DISCUSSION
In u r e t h a n e - a n e s t h e t i z e d rats, the h i p p o c a m p a l 0 rhythm has been characterized by its sensitivity to
59 large doses of atropine or scopolamine. The loss of the 0 rhythm is apparent in the raw E E G and in all of the methods commonly used to demonstrate its presence, such as power spectral analyses and autocorrelation methods 9"1°. One group of rhythmic septal cells has been shown to project to the hippocampus and to remain rhythmic after such large doses of atropine 15"16. Using the spike trains from these atropine-resistant septai neurons, we have demonstrated the existence of an atropineresistant component of urethane 0 rhythm in rats. The presence of this component even after doses of atropine as high as 100 mg/kg (i.v.) indicates that this residual 0 activity cannot be due to the action of muscarinic cholinergic synapses. The tight coupling of the residual 0 activity to the rhythmic firing of the atropine-resistant septal cells indicates that these cells are involved in its production, either directly or indirectly. Further, the presence of an atropineresistant component of the urethane 0 rhythm strengthens the previously reported suggestion that the atropine-resistant septal cells are not cholinergic 1~. The mechanism by which the atropine-resistant septal cells ultimately generate the atropine-resistant 0 activity is unclear. Synaptic contacts of the atropine-resistant cells directly onto pyramidal and granule cells is one possible mechanism for generating a 0 E E G component. Another possibility is predicted by our previously reported model 16 where we suggested that the atropine-resistant septal cells are GABAergic and terminate exclusively on interneurons in the hippocampus (see Introduction). The atropine-resistant rhythmic septal neurons may indirectly produce a rhythmic E E G component by modulating the residual firing of hippocampal 0 cells (presumed interneurons). The postatropine 0 cell firing rate may be sufficiently low to prevent such rhythmic modulation from being detected with spike train analysis. Averages of 0 cell firing triggered on atropine-resistant septal cell bursts will be necessary to identify such continued modulation. Vanderwolf and Baker ~7 have suggested that a REFERENCES 1 Amaral, D.G. and Kurz, J., An analysis of the origins of the cholinergic and noncholinergie septal projections to the hippocampal formation of the rat, J.Comp.Neurol.. 240
serotonergic projection from the median raph6 is responsible for the atropine-resistant 0 rhythm of freely-moving rats. Our finding that atropine-resistant medial septal cells are likely responsible for the atropine-resistant 0 activity of urethane-anesthetized rats seems to justify consideration of the possibility that the atropine-resistant septal cells ~pace' the atropine-resistant 0 rhythm of walking rats as well. In fact, we suggest that the "pacemaker mechanism' may be indentical for walking and urethane-anesthetized rats. By "pacemaker mechanism' we mean the mechanism by which the rhythmic cells of the septum (regardless of their frequency) entrain the neurons of the hippocampal formation. We suggest that this mechanism involves the rhythmic modulation of 0 cell firing by both atropine-sensitive and atropine-resistant septal neurons. In this scenario, the data on the effects of serotonergic contributions to the hippocampal E E G might be explained by an effect of serotonergic agents on the atropine-resistant cell population in the medial septal nuclei. In support of this strictly modulatory action for serotonin is the sensitivity of some septal neurons to serotonergic agents ~4 and the lack of rhythmic activity in median raphd neurons 4. It is not our intention to assert that walking 0 rhythm is identical to urethane 0 rhythm. The apparent effects of urethane on glutamate synapses ~?, differences in the firing phases of hippocampal neuronsS~ and the depth profiles of the EEG 3'7'b~, make this impossible. Rather, we suggest that the mechanism of 'pacing' may be identical in awake and anesthetized animals and attributable ultimately to the two populations of rhythmic septal neurons.
ACKNOWLEDGEMENTS Supported by National Institutes of Health Grants NS 17095 to S. E. F. and NS07117 (Neurophysioiogy Training Grant). We thank Dr. James B. Ranck, Jr. for comments on an earlier version of the manuscript. (1985) 37-59. 2 Baisden, R.H., Woodruff, M.L. and Hoover, D.B., Cholinergic and non-cholinergic septohippocampal projections: a double-label horseradish peroxidase-acetylcholinesterase study in the rabbit, Brain Research, 290 (1984)
6(J 140-151. 3 Bland, B.H. and Whishaw, I.Q., Generators and topography of hippocampal theta (RSA) in the anesthetized and freely moving rat, Brain Research, 118 (1976) 259-280. 4 Crunelli, V. and Segal, M., An electrophysiological study of neurons in the rat median raphO and their projections to septum and hippocampus, Neuroscience. 15 (1985) 47-6(I. 5 Fox, S.E.. Wolfson, S, and Ranck Jr., J.B., Hippocampal theta rhythm and the firing of neurons in walking and urethane anesthetized rats, Exp. Brain Res., 62 (1986) 495-508. 6 Freund, T.E and Antal, M., GABA containing neurons in the septum control inhibitory interneurons in the hippocampus, Nature (Lond.), 336 (1988) 170-173. 7 Green, K.E and Rawlins, J.N.P., Hippocampal theta in rats under urethane: generators and phase relations, EEG Clin. Neurophysiol., 47 (1979) 420-429. 8 K6hler, C., Chan-Palay, V. and Wu, J.-Y., Septal neurons containing glutamic acid decarboxylase immunoreactivity project to the hippocampal region in the rat brain, Anat.Embryol., 169 (1984) 41-44. 9 Kramis, R., Vanderwolf, C.H. and Bland, B.H., Two types of hippocampal rhythmical slow activity in both the rabbit and the rat: relations to behavior and effects of atropine, diethyl ether, urethane, and pentobarbital, Exp. Neurol.. 49 (1975) 58-85. 10 Leung, L.-W.S., Spectral analysis of hippocampal EEG in the freely moving rat: effects of centrally active drugs and relations to evoked potentials, EEG Clin. Neurophysiol., 60 (1985) 65-77.
11 Mesulam, M.-M., Mufson, E.J., Wainer, B.tt. ~md Lcvcy. A.I., Central cholinergic pathways in the rat: an overview based on an alternative nomenclature (Chl-Ch(~), ,Vc~r(~science, 10 (1983) 1185-12(11. 12 Moroni, E, Corradetti, R., Casamenti, F., Moncti, G. and Pepeu, G., The release of endogenous GABA and glutamate from the cerebral cortex in the rat, Naunyn-Schrniedeberg's Arch. Pharmacol., 316 (I 981 ) 235-239 13 Paxinos, G. and Watson, C., The Rat Brain in Stereotaxic Coordinates, Academic, Sydney, 1982. 14 Segal, M., Properties of rat medial septal neurons recorded in vitro, J. Physiol. (Lond.), 379 (1986) 309-330. 15 Stewart, M. and Fox, S.E., Two populations of rhythmically bursting neurons in the septal nuclei are revealed by atropine, Soc. Neurosci. Abstr.. 12 (1986) 1527. 16 Stewart, M and Fox, S.E., Two populations of rhythmically bursting neurons in the rat medial septum are revealed by atropine, J. Neurophysiol., 6l (1989) 982-993. 17 Vanderwolf, C.H. and Baker, G.B., Evidence that serotonin mediates non-cholinergic neocortical low voltage fast activity, non-cholinergic hippocampal rhythmical slow activity and contributes to intelligent behavior, Brain Research, 374 (1986) 342-356. 18 Wainer, B.H., Levey, A.I., Rye, D.B., Mesulam, M.-M. and Mufson, E.J., Cholinergic and non-cholinergic septohippocampal pathways, Neurosci. Lett., 54 (1985) 45-52. 19 Winson, J.. Patterns of hippocampal theta rhythm in the freely moving rat, EEG Clin. Neurophysiol., 36 (1974) 291-301.
55
BRES 14871
Detection of an atropine-resistant component of the hippocampal theta rhythm in urethane-anesthetized rats Mark Stewart and Steven E. Fox Department of Physiology, SUNY ttealth Science Center, Brooklyn, NY 11203 (U.S.A.
(Accepted 14 March 1989) Key words: Medial septum; Rhythmic cell; Acetylcholine: 7-Aminobutyric acid
An important pharmacological feature of the hippocampal 0 rhythm in urethane-anesthetized animals is tts apparent sensitivity to antimuscarinic drugs. This sensitivity may be partly due to a masking of the 0 frequency by increases in both higher and lower frequency EEG components that are unrelated to any residual 0 rhythm. The discovery of atropine-resistant, rhythmic medial septal neurons has provided a physiological trigger for averaging EEG and unit activity after large atropine doses. Such averaging has permitted the detection of an atropine-resistant component of the hippocampa[ 0 rhythm in urethane-anesthetized rats. The postatropine 0 activity recorded from both CA1 (superficial to the pyramidal cell layer) and dentate (near the hippocampal fissure) in 15 rats was typically reduced in amplitude, but the recordings from the two locations maintained their phase relations to the septal units and to each other. The presence of this residual 0 component after doses as large as 100 mg/kg indicates that it cannot be mediated by muscarinic cholinergic receptors. The coupling of the signal to the atropine-resistant septal cells strengthens our previous suggestion that these septo-hippocampal neurons are not cholinergic, and are therefore probably GABAergic.
INTRODUCTION The E E G of the h i p p o c a m p a l formation is characterized by a very regular 4 - 1 2 Hz sinusoidal pattern that reflects the synchronous activity of a great many synapses. This activity, referred to as the 0 rhythm or rhythmical slow activity, can be recorded with depth electrodes in awake and urethane-anesthetized animals. The 0 rhythm r e c o r d e d during urethane anesthesia has been characterized by its sensitivity to large doses of antimuscarinic drugs v'm. A n a t o m i c a l l y , both cholinergic and G A B A e r g i c s e p t o - h i p p o c a m p a l neurons have been identified 1' 2,s,~.ls Of the electrophysiologically identifiable cells of the medial septal nuclei, only atropineresistant and atropine-sensitive rhythmically bursting cells have been found to be driven antidromically from the h i p p o c a m p u s 1+'. We have p r o p o s e d that there is a c o r r e s p o n d e n c e between cholinergic and atropine-sensitive cells on the one hand and G A B A -
ergic and atropine-resistant cells on the o t h e r ~<'. This conclusion is the simplest explanation for a great variety of data, Some of the m o r e i m p o r t a n t points for making this argument are r e i t e r a t e d here. The direct application of a t r o p i n e only to the septum can apparently eliminate the h i p p o c a m p a l 0 rhythm t6. This indicates that a change in the firing pattern of atropine-sensitive septal cells, from rhythmic to non-rhythmic, can eliminate the 0 rhythm, in spite of unblocked cholinergic receptors within the hippocampus. The continued rhythmic activity of atropine-resistant septal cells alone, must not be sufficient to organize h i p p o c a m p a l cell activity which would result in a rhythmic E E G , nor do the atropine-resistant cells a p p e a r to be able to directly generate much of a rhythmic E E G signal. This is consistent with recent anatomical evidence that the G A B A e r g i c s e p t o - h i p p o c a m p a l p r o j e c t i o n is almost exclusively onto interneurons ('. It occurred to us that some c o m p o n e n t of the 0
(Jorrespondence: M. Stewart,Department of Physiology,Box 31, SUNY Health Science Center, 450 Clarkson Avenue, Brooklyn, NY 11203, U.S.A.
1i006-8993/89/$03.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)
36 rhythm might remain after atropine, masked by increases in Large-amplitude irregular activity in the EEG. We attempted to detect any contribution of the atropine-resistant septal neurons to the EEG by using the continued rhythmic activity of these cells as a trigger for averaging the post-atropine hippocampal EEG. MATERIALS AND METHODS Male Sprague-Dawley albino rats (310-450 g) were anesthetized with urethane (1.0 g/kg i.v. or 1.5 g/kg i.p.). Each of the 15 rats was held in a Kopf stereotaxic frame and maintained at 37 °C with an isothermal heating pad (Braintree Scientific, Braintree, MA) and/or shielded heating lamp. Electrode implants were made using stereotaxic coordinates 13. Extracellular recordings were taken from rhythmically bursting medial septal neurons with microelectrodes (1-2 Mr'2 stainless steel, Microprobe, Ciarksburg, MD) that were lowered through the septal nuclei (AP level 9.2-9.8 mm anterior to interaural line) at an angle of 20° off the vertical in the coronal plane. Hippocampal E E G was recorded monopolarly from each of a pair of stainless steel wires (125 ~m diameter) that were varnish insulated and cut square. The tips were vertically separated by about 1 mm and referred to a screw over cerebellum. The E E G electrode pair was lowered vertically (AP 4.7, ML 3.0) until the 0 rhythm recorded on the deeper electrode was phase-reversed with respect to that on the superficial electrode, and maximal in amplitude. The deeper electrode tip was then near the hippocampal fissure ( referred to as 'dentate' EEG) and the superficial electrode was dorsal to the pyramidal cell layer of CA1 (referred to as 'CAI' EEG) 7. On occasion, the deeper electrode was positioned closer to the granule cell layer. This deeper location was apparent in the EEG as a decrease in 0 rhythm amplitude from positions nearer the fissure and an increase in amplitude of the higher frequencies (30-50 Hz). Electrode locations for all units (one per rat) were marked at the ends of the experiments by passing current through the recording electrodes and visualizing the individual lesions with the Prussian blue reaction in Nissl-stained 40-/~m frozen sections. All units were confirmed to be within the medial septal
nucleus-nucleus of the diagonal band (vertical limb). Atropine sulfate (Sigma) was prepared as a 25 or 50 mg/ml solution in water for i.v. (25-100 mg/kg) or i.p.(50 mg/kg) injection. Postatropine recordings of septal unit activity and concurrent EEG were started at least 5 min after the full dose of atropine had been given and lasted for 10 rain to 1 h. Unit activity and E E G recordings were amplified, filtered (single units: bandpass = 500 Hz to 10 kHz; -24 dB/octave high pass, -6 dB/octave low pass. EEG: bandpass = 0.1 Hz to 10 kHz; -6 dB/octave rolloffs), and stored on tape for off-line analysis. A Hewlett Packard 9836U computer system was used for the construction of magnitude spectra as well as spike- and sine-wave triggered histograms. For magnitude spectra computations, taped data were lowpass filtered (100 Hz) and digitized at 200 Hz. Spectral averages were computed using the fastFourier transform (FFT) on 61 Hann-windowed segments of 512 data points, overlapped by 50%. A detailed description of the construction of phase histograms has been published 5. For spike-triggered histograms, a 45-ms duration rolling average of the discriminated and digitized septal unit spike train was computed. Theta scores 1~', estimates of rhythmic quality based on autocorrelations of cosine-windowed 1.024-s epochs of EEG made every 64 ms, were used to identify periods of 0 frequency activity. Peaks of identified '0 cycles' in the converted spike train or in a simultaneously digitized, independent sine wave were detected, and acceptable 0 cycles were those whose periods fell within pre-set time windows (typically 176-272 ms). Cycles were averaged by normalizing slow-wave amplitude data and spike data collected during each 0 cycle, redistributing them into 32 histogram bins. Final scaling yielded the slow-wave amplitude at each 11.25 ° segment of the average 0 cycle. RESULTS Fifteen rats received massive doses of systemic atropine. Doses ranged from 25-100 mg/kg i.v. (one rat received 50 mg/kg i.p.). The hippocampal 0 rhythm appeared to be completely eliminated when judged by inspection of the raw E E G (Fig. 1) or by frequency analysis (Fig. 2). Close inspection of the
57
pre- and post-atropine magnitude spectra indicates that the 0 activity might in fact be masked by large increases in both higher and lower frequencies after atropine. U n f o r t u n a t e l y , frequency analyses or autocorrelation methods cannot easily resolve this question. We used the spike trains of atropine-resistant
10
S
i',
c
CA 1 EEG
O -O
05
~
X
?re Atropine -- Post Atropine
N O
E
septal cells as a trigger to average the hippocampal 10
E E G . This method permitted the detection of a residual c o m p o n e n t of the 0 rhythm after doses of atropine as large as 100 mg/kg i.v. (Fig. 3, top and middle). We refer to this residual signal as 0 'activity' or as a ' c o m p o n e n t ' of the 0 rhythm for several
(b
1.0
7
30
40
I,
;"
DENTATE EEG
i
c
20
O 05 ~O ® N
.,_.
...... Post-Atropine
O
E
Pre-Atropine
0
J _ [ill ..l.J i,[, li.,l[ _ lliU . I]iJ.;, ,IILL_..IL ]1-- IN IF ~ll ll'[~ " ]l~ F'] IF1 i' ,
~
',-,,,,
~4~ ~r
10
20
30
40
Frequency(Hz) Fig. 2. Magnitude spectral plots for the CA1 (top) and dentate (bottom) EEGs, before and after atropine injection. Eighty seconds of EEG for each of the two conditions illustrated in Fig.1 were used for these magnitude spectra. The loss of 0 rhythm is seen in these overlaid plots to potentially result, at least in part, from a ' masking' of the 0 frequency by increases in lower and higher frequencies. For clear illustration of the low frequency components, only the first 102 of the 256 real frequency points are plotted. Normalized magnitudes: CA1 EEG - 1.0 = 0.46 mV/bin; Dentate EEG - 1.0 = (I.41 mV/bin. Bin width = 0.39 Hz.
~,t
Post-Atropine
reasons: (a) the 0 frequency is not obvious in the raw E E G or magnitude spectra; (b) as we will indicate, I
,I
'! ~ r]
'
I
Fig. 1. Septal unit and hippocampal EEG activity before and after 100 mg/kg i.v. atropine sulfate. The filtered septal unit recording is the top trace of each panel. The middle trace is EEG recorded superficial to the pyramidal cell layer of CA1 ('CAI' EEG) and the bottom trace is EEG recorded deep to the hippocampal fissure ('dentate' EEG). Normally dentate EEG recordings were taken near the amplitude maximum which is close to the hippocampal fissure. This recording was taken deeper than usual as evidenced by the relatively smaller amplitude compared with the CA1 recording and the large amount of high frequency activity. The top panel shows 0 rhythm in the hippocampal EEG along with rhythmic bursting in the scptal unit. The bottom panel shows the apparent absence of hippocampal 0 rhythm and continued rhythmicity in the septal unit, after atropine injection. Calibrations: l mV for the unit, 0.2 mV for both EEG channels, and 0.5 s. Positivity is up.
the relative amplitudes of CA1 and dentate recordings are not preserved; and (c) the rhythmic contribution of atropine-sensitive septal cells has been eliminated. The average n u m b e r of 0 cycles collected per rat was 351 cycles pre-atropine (S.D. = 275, range = 39 - 1166, median = 333) and 339 cycles post-atropine (S.D. = 277, range -- 127 1160, median --- 258). In order to address the possibility of recovering a spurious sinusoid by averaging with a periodic trigger, we triggered the averages of the E E G and the septal unit with a sine wave whose period was within + 10 ms of the average bursting period of the units , and matched the total n u m b e r of averaged cycles ( Fig. 3, bottom). The larger amplitude of the spike-triggered average compared to the sine wave-triggered aver-
58
CA1 EEG
":'~ -01
"
+0.1
DT EEG (mY)
i
"'_'"
"
.~
-0.1
(SP'UNylITbin)o~ r J ....... ~ 1 ~ ~ ' • 0
90
180
Lt~-: 270
360
+0.05]
CA1
I
~
VLo.osJ
-
+0.05
. - .
EEG (mY) -0.05 0.63' MS-DB UNIT (sp/cy/bin) 0 C
0
90
180
270
360
0
90
180
270
360
+0.05 CA1 EEG (mVl
-0.05 +0.05 DT EEG (mY)
-0,05 0,17~
MS-DB
UNIT (sp/cy/bin) 0
PHASE{deg) Fig. 3. Detection of postatropine urethane 0 activity by averaging the hippocampal EEG, triggered on the rhythmic bursting of the atropine-resistant septal unit. Dotted lines in the histograms are 1/4 cycles of the averaged EEG or spike histogram redrawn on either end for clarity. Panel A shows the averaged CA1 and dentate EEGs triggered on the septal unit, before atropine injection. Peak-to-peak amplitudes of each channel are approximately 200/~ V. Number of cycles in the average is 238. Mean bursting frequency for the unit was 4.4 Hz. Panel B shows the averaged EEG, triggered on the unit, after atropine. Note the 0 cycle amplitudes are diminished, but their phase relations to the unit and to each other are maintained. Number of cycles in the average is 475. Mean bursting frequency for the unit was 4.8 Hz. Panel C shows the averages of EEG activity triggered on a simultaneously sampled 4.6 Hz sine wave. Number of cycles in the average is 475. Positivity is up for the EEG averages.
age and the fact that the phases of the two E E G records were essentially identical to the p r e - a t r o p i n e condition only in the spike-triggered average, indicate that the recovered 0 rhythm is not an artifact of the averaging method. It is difficult to provide much quantitative d a t a on the residual 0 c o m p o n e n t , since there are only two recording locations and they are similar, but not identical across rats. We will however, present some crude quantitative data and some interesting observations. The fraction of the p e a k - t o - p e a k 0 cycle amplitude remaining after a t r o p i n e was 0.44 for CA1 recordings (n = 15, S.D. = 0.34, range = 0.09- 1.25, median = 0.27) and 0.29 for d e n t a t e recordings (n = 15, S.D. = 0.13, range = 0.12 - 0.50, median = 0.28). The phase of the positive p e a k relative to the spike trigger shifted an average of 6.8 ° (n = 15, S.D. = 47.3) in CA1 and 28.5 ° (n = 15, S.D. = 41.1) in the dentate. The average shifts were in the same direction for the two locations: increased lag of the 0 activity relative to the unit bursts. The p e a k - t o - p e a k amplitude of the averaged post-atropine 0 cycle was larger than 2 S . E . M . of the averaged 0 cycle in 10 of 15 rats for C A I 0 rhythm and 13 of 15 rats for d e n t a t e 0 rhythm. W h e n only these rats were used, means and S.D. for the fraction of the p e a k - t o - p e a k a m p l i t u d e preserved after atropine were essentially identical to those values obtained when all animals were included. The amplitude of the residual 0 c o m p o n e n t is clearly reduced c o m p a r e d to the p r e - a t r o p i n e condition. We e m p h a size however, that these values are only estimates, and are likely to be overestimates of the residual 0 component. There was no relation b e t w e e n the a m p l i t u d e of the residual 0 activity and the dose of atropine. It should also be n o t e d that the dentate 0 activity (typically r e c o r d e d at the level of the h i p p o c a m p a l fissure) was a t t e n u a t e d m o r e than the 0 activity r e c o r d e d in CA1 (dorsal to the p y r a m i d a l cell layer), but when the d e n t a t e E E G e l e c t r o d e was closer to the granule cell layer, a larger fraction of the p r e - a t r o p i n e amplitude a p p e a r e d to be preserved. DISCUSSION
In u r e t h a n e - a n e s t h e t i z e d rats, the h i p p o c a m p a l 0 rhythm has been characterized by its sensitivity to
59 large doses of atropine or scopolamine. The loss of the 0 rhythm is apparent in the raw E E G and in all of the methods commonly used to demonstrate its presence, such as power spectral analyses and autocorrelation methods 9"1°. One group of rhythmic septal cells has been shown to project to the hippocampus and to remain rhythmic after such large doses of atropine 15"16. Using the spike trains from these atropine-resistant septai neurons, we have demonstrated the existence of an atropineresistant component of urethane 0 rhythm in rats. The presence of this component even after doses of atropine as high as 100 mg/kg (i.v.) indicates that this residual 0 activity cannot be due to the action of muscarinic cholinergic synapses. The tight coupling of the residual 0 activity to the rhythmic firing of the atropine-resistant septal cells indicates that these cells are involved in its production, either directly or indirectly. Further, the presence of an atropineresistant component of the urethane 0 rhythm strengthens the previously reported suggestion that the atropine-resistant septal cells are not cholinergic 1~. The mechanism by which the atropine-resistant septal cells ultimately generate the atropine-resistant 0 activity is unclear. Synaptic contacts of the atropine-resistant cells directly onto pyramidal and granule cells is one possible mechanism for generating a 0 E E G component. Another possibility is predicted by our previously reported model 16 where we suggested that the atropine-resistant septal cells are GABAergic and terminate exclusively on interneurons in the hippocampus (see Introduction). The atropine-resistant rhythmic septal neurons may indirectly produce a rhythmic E E G component by modulating the residual firing of hippocampal 0 cells (presumed interneurons). The postatropine 0 cell firing rate may be sufficiently low to prevent such rhythmic modulation from being detected with spike train analysis. Averages of 0 cell firing triggered on atropine-resistant septal cell bursts will be necessary to identify such continued modulation. Vanderwolf and Baker ~7 have suggested that a REFERENCES 1 Amaral, D.G. and Kurz, J., An analysis of the origins of the cholinergic and noncholinergie septal projections to the hippocampal formation of the rat, J.Comp.Neurol.. 240
serotonergic projection from the median raph6 is responsible for the atropine-resistant 0 rhythm of freely-moving rats. Our finding that atropine-resistant medial septal cells are likely responsible for the atropine-resistant 0 activity of urethane-anesthetized rats seems to justify consideration of the possibility that the atropine-resistant septal cells ~pace' the atropine-resistant 0 rhythm of walking rats as well. In fact, we suggest that the "pacemaker mechanism' may be indentical for walking and urethane-anesthetized rats. By "pacemaker mechanism' we mean the mechanism by which the rhythmic cells of the septum (regardless of their frequency) entrain the neurons of the hippocampal formation. We suggest that this mechanism involves the rhythmic modulation of 0 cell firing by both atropine-sensitive and atropine-resistant septal neurons. In this scenario, the data on the effects of serotonergic contributions to the hippocampal E E G might be explained by an effect of serotonergic agents on the atropine-resistant cell population in the medial septal nuclei. In support of this strictly modulatory action for serotonin is the sensitivity of some septal neurons to serotonergic agents ~4 and the lack of rhythmic activity in median raphd neurons 4. It is not our intention to assert that walking 0 rhythm is identical to urethane 0 rhythm. The apparent effects of urethane on glutamate synapses ~?, differences in the firing phases of hippocampal neuronsS~ and the depth profiles of the EEG 3'7'b~, make this impossible. Rather, we suggest that the mechanism of 'pacing' may be identical in awake and anesthetized animals and attributable ultimately to the two populations of rhythmic septal neurons.
ACKNOWLEDGEMENTS Supported by National Institutes of Health Grants NS 17095 to S. E. F. and NS07117 (Neurophysioiogy Training Grant). We thank Dr. James B. Ranck, Jr. for comments on an earlier version of the manuscript. (1985) 37-59. 2 Baisden, R.H., Woodruff, M.L. and Hoover, D.B., Cholinergic and non-cholinergic septohippocampal projections: a double-label horseradish peroxidase-acetylcholinesterase study in the rabbit, Brain Research, 290 (1984)
6(J 140-151. 3 Bland, B.H. and Whishaw, I.Q., Generators and topography of hippocampal theta (RSA) in the anesthetized and freely moving rat, Brain Research, 118 (1976) 259-280. 4 Crunelli, V. and Segal, M., An electrophysiological study of neurons in the rat median raphO and their projections to septum and hippocampus, Neuroscience. 15 (1985) 47-6(I. 5 Fox, S.E.. Wolfson, S, and Ranck Jr., J.B., Hippocampal theta rhythm and the firing of neurons in walking and urethane anesthetized rats, Exp. Brain Res., 62 (1986) 495-508. 6 Freund, T.E and Antal, M., GABA containing neurons in the septum control inhibitory interneurons in the hippocampus, Nature (Lond.), 336 (1988) 170-173. 7 Green, K.E and Rawlins, J.N.P., Hippocampal theta in rats under urethane: generators and phase relations, EEG Clin. Neurophysiol., 47 (1979) 420-429. 8 K6hler, C., Chan-Palay, V. and Wu, J.-Y., Septal neurons containing glutamic acid decarboxylase immunoreactivity project to the hippocampal region in the rat brain, Anat.Embryol., 169 (1984) 41-44. 9 Kramis, R., Vanderwolf, C.H. and Bland, B.H., Two types of hippocampal rhythmical slow activity in both the rabbit and the rat: relations to behavior and effects of atropine, diethyl ether, urethane, and pentobarbital, Exp. Neurol.. 49 (1975) 58-85. 10 Leung, L.-W.S., Spectral analysis of hippocampal EEG in the freely moving rat: effects of centrally active drugs and relations to evoked potentials, EEG Clin. Neurophysiol., 60 (1985) 65-77.
11 Mesulam, M.-M., Mufson, E.J., Wainer, B.tt. ~md Lcvcy. A.I., Central cholinergic pathways in the rat: an overview based on an alternative nomenclature (Chl-Ch(~), ,Vc~r(~science, 10 (1983) 1185-12(11. 12 Moroni, E, Corradetti, R., Casamenti, F., Moncti, G. and Pepeu, G., The release of endogenous GABA and glutamate from the cerebral cortex in the rat, Naunyn-Schrniedeberg's Arch. Pharmacol., 316 (I 981 ) 235-239 13 Paxinos, G. and Watson, C., The Rat Brain in Stereotaxic Coordinates, Academic, Sydney, 1982. 14 Segal, M., Properties of rat medial septal neurons recorded in vitro, J. Physiol. (Lond.), 379 (1986) 309-330. 15 Stewart, M. and Fox, S.E., Two populations of rhythmically bursting neurons in the septal nuclei are revealed by atropine, Soc. Neurosci. Abstr.. 12 (1986) 1527. 16 Stewart, M and Fox, S.E., Two populations of rhythmically bursting neurons in the rat medial septum are revealed by atropine, J. Neurophysiol., 6l (1989) 982-993. 17 Vanderwolf, C.H. and Baker, G.B., Evidence that serotonin mediates non-cholinergic neocortical low voltage fast activity, non-cholinergic hippocampal rhythmical slow activity and contributes to intelligent behavior, Brain Research, 374 (1986) 342-356. 18 Wainer, B.H., Levey, A.I., Rye, D.B., Mesulam, M.-M. and Mufson, E.J., Cholinergic and non-cholinergic septohippocampal pathways, Neurosci. Lett., 54 (1985) 45-52. 19 Winson, J.. Patterns of hippocampal theta rhythm in the freely moving rat, EEG Clin. Neurophysiol., 36 (1974) 291-301.