Exp Brain Res (1993) 93:1-5

Experimental Brain Research �9 Springer-Verlag 1993

Disinhibition of hippocampal pyramidal cells during the transition into theta rhythm M. Stewart

Department of Pharmacology, State University of New York Health Science Center, Brooklyn, NY 11203, USA

Received: 14 July 1992 Accepted: 8 September 1992

Abstract. The activity of hippocampal complex-spike cells (presmned pyramidal cells) and theta cells (presumed interneurons) was examined during transi- tions from non-theta electroencephalogram (EEG) states to theta EEG states in freely moving and sleeping rats. Theta cell firing rates were significantly depressed in a 1-s period centered on the EEG transition relative to the surrounding 1-s periods (normalized rates+SEM): 1.05 4- 0.02 for the "non-theta" period, 0.59 4- 0.03 for the "transition" period, and 1.364-0.04 for the "theta" period (n = 26 cells). Conversely, complex-spike cell firing was significantly increased during the transition period: 0.51 4- 0.11 for the "non-theta" period, 2.244- 0.19 for the "transition" period, and 0.244-0.04 for the "theta" period (n = 27 cells). This diametrically altered activity indicates that theta cells must be actively inhibited during the transition. The increased activity in complex-spike cells during the transition may be simply a release from inhibitory control by interneurons. The pattern of theta cell inhibition together with increased complex-spike cell activity appears to be a general property of transitions into the theta EEG state, irrespective of behavior. It is suggested that increased activity in septal afferents (GA- BAergic cell activity greater than cholinergic cell activity) initially inhibits hippocampal interneurons. The inhibi- tion is not sustained because of an activity-dependent decrease in the potency of the septointerneuronal inhibi- tion, leaving the rhythmic excitatory (cholinergic) sep- tointerneuronal inputs, together with principal cell in- puts, to increase interneuron firing rates.

Key words: Complex-spike cell - Theta cell Rhythmic slow activity - Rat

Introduction

Two broad classes of hippocampal neurons have been defined on the basis of their spontaneous and evoked

firing patterns. Complex-spike cells fire single-action potentials or bursts of about two to ten action potentials of decreasing amplitude and increasing duration (Ranck 1973). Interspike intervals within the burst can be less than 5 ms. In Ammon's horn, these cells are considered to be pyramidal cells because of their location (Fox and Ranck 1975), their antidromic responses to fimbria/ fornix stimulation (Fox and Ranck 1981), and their firing properties (e.g., Nfifiez et al. 1987). Theta cells appear to be interneurons, located primarily in stratum oriens and stratum pyramidale of Ammon's horn and in the dentate hilus. This conclusion is based upon their location (Fox and Ranck 1975), their orthodromic responses to tim- bria/fornix or perforant path stimulation (Fox and Ranck 1981; Buzsaki and Eidelberg 1983), and the cor- respondence of their firing properties (action potential shape, discharge patterns) with identified interneurons or nonpyramidal cells in vitro (Schwartzkroin and Mathers 1978; Lacaille et al. 1987).

Complex-spike cells and theta cells both fire in phase with the hippocampal theta rhythm (reviewed in Fox et al. 1986). It was noted in their original description (Ranck 1973) that theta cell firing rates were higher, by about a factor of 2, during theta rhythm than in non-theta electroencephalogram (EEG) states. Correspondingly, complex-spike cells were seen to fire at lower rates during hippocampal theta rhythm. This inverse relation of firing rates is consistent with theta cells inhibiting complex- spike cells. Complex-spike cells are least excitable on the phase of a theta cycle that theta cells fire (Rudell et al. 1980; Rudell and Fox 1984). It must be emphasized, however, that complex-spike cells can fire at very high rates when a rat is in a particular place in its environment (O'Keefe and Dostrovsky 1971; Muller et al. 1987). In fact, within a "place field", the firing rate of a complex- spike cell is higher (and its firing rate outside of the place field is lower) when theta rhythm is also present (Kubie et al. 1985). Only if the location-specific firing properties of complex-spike cells are excluded is the inverse relation of firing rates for complex-spike cells and theta cells apparent.

Studies of hippocampal theta rhythm generation and theta-related unit activity have relied on the sinusoidal pattern of the theta rhythm as an averaging tool. Small variations in cycle durations can be eliminated by using phase as the independent variable instead of time. This averaging ability has permitted: (1) depth profiles (EEG recordings at multiple depths) through the hippocampal and entorhinal cortices to be constructed using a single moveable microelectrode (Buzsaki et al. 1986; Fox and Stewart 1986; Brankack and Fox 1987); (2) determina- tion of phase relations for fast-firing and, more impor- tantly, slow-firing neurons (e.g., Fox et al. 1986); and (3) detection of small theta components masked by other EEG frequencies by using rhythmic unit activity as the averaging trigger (Stewart and Fox 1989b). Such studies have been used to define the principal elements involved in theta rhythm generation and how the synchronization of neurons throughout the hippocampal formation is maintained (Stewart and Fox 1990; Fox et al. 1992). These studies do not, however, address the issue of how the synchronization is initially achieved. The purpose of this study was to examine transitions from non-theta to theta EEG states as an approach to the mechanism of interneuron and projection cell synchronization.

Materials and methods

The dataset for the present study was obtained from recordings originally taken by Ranck. The methods for single-unit recordings in freely behaving rats, including behavioral training, electrode construction, surgery, and histology, have been previously pub- lished in detail (Ranck 1973; Fox and Ranck 1975).

The records were searched for "clean" EEG transitions. A clean transition was defined as 3 s of continuous non-theta EEG followed by 3 s of continuous theta EEG. Transitions involving briefer ep- isodes of non-theta or theta EEG (especially common during awake behaviors) were not included, in an attempt to minimize factors that might complicate observed changes in firing pattern or rate. At least one transition to theta EEG was recorded for each cell (range 1-6 transitions per cell). The method used to quantitate firing rate changes through the transition from non-theta to theta EEG states was deliberately simple. Automated detection of EEG transitions was not possible because previously described methods (e.g., Stew- art and Fox 1989a) based on autocorrelating brief epochs of EEG cannot precisely locate the first theta cycle. The first cycle of theta

rhythm was located by eye and its apparent beginning was used to define the midpoint of a "transition" period that was 1 s long. Since analysis was limited to transitions consisting of at least 3 s of non-theta EEG followed by at least 3 s of theta EEG, identification of the first theta cycle could be in error only by as much as one full theta cycle (approximately 125 ms), more precise than autocorrela- tion methods. Bin widths of 1 s minimize the error associated with identifying the first cycle of theta rhythm, without being so long that brief changes in firing rate would be undetectable. Action potentials (> 3 : 1 signal-to-noise ratio within a narrow amplitude range) were counted for the 1-s "transition" bin and compared with the firing rates for that cell from adjacent 1-s bins: a "non-theta" bin preceed- ing the transition bin, and a "theta" bin following it. Whenever possible, stable rates for non-theta and theta EEG states were obtained from one or more 30- to 60-s epochs of EEG.

Results

A total of 53 cells (26 theta cells; 27 complex-spike cells) were recorded from 23 rats. In four cases, a complex- spike cell and a theta cell could be discriminated at a single recording location (all in CA1).

Theta cells

Forty-seven transitions to the theta EEG state were iden- tified in the recordings taken from 26 theta cells. Loca- tions for the cells were as follows: 15 from CA1, 3 from CA3, and 8 from the dentate hilar region. Twenty-six transitions were recorded in awake rats (immobility to walking) and 21 from sleeping rats (slow-wave sleep to paradoxical sleep). The data were combined into a single group since transitions were indistinguishable from one another (except for small differences in the frequency of the theta EEG).

Transitions from theta to non-theta EEG states were not examined "quantitatively" because fewer of these were collected, but, more importantly, because no consis- tent trend in unit activity could be detected within a limited time around the EEG change. It was difficult even to locate a single point in the EEG record that might represent the transition between the states.

Figure 1 shows the firing of a theta cell and a complex- spike cell recorded simultaneously with a single elec-

L I I J l l l _ l l JL l I .L . l l lL - -L- I . -F--FIlF FN-III--F]- I l

O O �9 �9

lll F llI H ' ' " " ' ,,i, ,,,L . , , ,,1,, ,,,, .,, ,,,,,,,

A A A A

Fig. 1. Firing of a hippocampal complex-spike cell and a theta cell recorded simultaneously from the same electrode during a tran- sition into theta rhythm. The cells were located in area CA1. The rat stood motionless during the "non-theta" period and walked during the "theta" period. The observed transition from non-theta to theta electroencephalogram (EEG) states is marked below the

EEG trace with the large arrow. Small arrows mark the boundaries of the three 1-s bins referred to in the text. The large action poten- tials of the complex-spike cell are marked by four dots above each of 4 complex spikes fired by this cell during the "transition" period. Negativity up. Calibrations 500 pV; 0.5 s

2.0 Theta cells

O~ t -

.>_

rr"

1.5

1.o

0.5

o.o

Non-theta Transition Theta

3.0 ~ Complex-spike cells

2.5

2.0

c -

"= 1.5

.~_ ~ 1.0

0 .5

0 .0

Non-theta Transition Theta

Fig. 2. Firing rate changes in the transition from non-theta to theta electroencephalogram (EEG) states for hippocampal theta cells (top) and complex-spike cells (bottom). Relative rates are com- puted for each cell by dividing the cell's rate for each 1-s bin by the average rate for the three bins. The "transition" bin is centered on the observed EEG transition (see Results). The "non-theta" bin represents 1 s immediately prior to the transition bin (i.e., from 1.5 to 0.5 s prior to the EEG transition point. Similarly, the "theta" bin represents the 1-s period immediately after the transition bin. Bars indicate means+ SEM of relative rates for the group of cells. Top, theta cell firing rates (n = 26 cells) are significantly depressed during the transition period compared with rates from the adjacent periods. Relative rates: non-theta, 1.054- 0.02; transition, 0.59+ 0.03; theta, 1.36 • 0.04. ** Transition rate differs from both non-theta and theta rates (t = 9.736, 11.113 ; two-tailed; P < 0.001); * theta rate differs from non-theta rate (t=4.581; two-tailed; P< 0.001). Bottom, complex-spike cell firing rates (n= 20 cells) are significantly higher in the transition period. Seven cells whose firing rates were 0 spikes per second for all three bins were not included. Relative rates: non-theta, 0.51 • 0.11, transition, 2.24 • 0.19; theta, 0.24 • 0.04. * Transition rate differs from non-theta and theta rates (t = 6.157, 8.869; two-tailed; P < 0.001). The relative firing rates for non-theta and theta bins did not differ significantly

trode. The observed E E G transition point is marked (large arrow) together with the boundaries for the three 1-s bins (small arrows). Note the increased rate of firing (four complex spikes) in the complex-spike cell and the associated decrease rate of firing in the theta cell in the "transit ion" period. The E E G immediately preceding theta rhythm was typically low voltage, high frequency activity. Sharp waves (Buzsaki 1986) were not seen in any of the transitions.

Twenty-five of 26 theta cells had significantly de- pressed firing rates during the 1-s transition period as compared to both the non-theta and theta periods. This is illustrated in Figs. 2 (top) and 3 (open circles). In Fig. 2, the measured firing rates for each bin were normalized by dividing the rate for one bin by the average rate for all three bins. As a group, theta cell rates fell to about half the non-theta rate, then increased to be about 1.5 times greater during theta rhythm than in non-theta EEG. The normalized rates (_+ SEM) were 1.05 _ 0.02 for the non-theta period, 0.59_+0.03 for the transition period, and 1.36 -t- 0.04 for the theta period. Actual rate values ( • were 35.66_+2.67 spikes per second (spk/s) for the non-theta period, 20.50-t-1.61 spk/s for the transition period, and 42.66_+ 2.48 spk/s for the theta period. As indicated in Fig. 2, the rate decrease f rom non-theta to transition and the rate increase f rom tran- sition to theta were significant ( t=9.736, 11.113; two- tailed; P<0.001) . Also apparent in the figure is the sig- nificant rate increase f rom non-theta to theta ( t=4.581 ; two-tailed; P < 0.001). The rate increase f rom non-theta to theta E E G states was somewhat greater (1.5 times as opposed to 1.3 times) when the stable non-theta and theta rates were compared: 27.27+3.29 spk/s vs 41.66 +_ 3.30 spk/s.

The unnormalized rate differences for recorded theta cells (open circles) are shown in Fig. 3. All theta cells except one are represented in quadrant II (rate decrease followed by rate increase). Whereas this one cell showed a rate decrease followed by another rate decrease through the E E G transition, its stable non-theta and theta rates are in the typical direction: 16.17 spk/s and 19.18 spk/s, respectively.

The higher firing rates of theta cells than complex- spike cells made transitions to the theta state as clear in the spike train as they were in the EEG. Rhythmic bursting in theta cells never occurred before there were theta cycles in the EEG, and any theta cell firing prior to the first theta-related burst discharge was not clearly phase-locked. Since, at most, two to three theta cycles without associated bursts by a theta cell could be recorded at each transition, the absence of a phase relation for these cycles cannot be proved.

Complex-spike cells

A total of 46 transitions were recorded f rom 27 complex- spike cells (19 CA1, 6 CA3, 2 dentate). Nineteen transi- tions were recorded while the rats were awake (immobility to walking) and 26 while rats slept (slow-wave sleep to paradoxical sleep).

Q..

c-

._o t -

t - .:E_- x (D

ID') r - IO

5O O

O 0 O 0 O0 o

,0 (~ 0

Theta cells "00~. ( ~

i l 0

-50 -25

-25

-50

~, ' 25 50 []

Complex-spike cells

Rate change entering transition (spk/s)

Fig. 3. Actual firing rate changes through the transition from non- theta to theta electroencephalogram (EEG) states. Abscissa is the rate difference between "non-theta" and "transition" periods (a rate decrease is a negative value). Ordinate is the rate difference between transition and "theta" periods. All 26 theta cells (open circles) and 27 complex-spike cells (open squares) are plotted, including the 7 complex-spike cells with no spikes in all three bins. Note that all but one theta cell show the pattern of rate decrease followed by increase (i.e., all but one point exist in quadrant II). This theta cell did, however, show a rate increase when its stable rates were compared. Note too that most theta cell points lie to the right of the dotted line, indicating that the rate increase coming out of the transition was greater than the rate decrease going into the transition. The very low firing rates for complex-spike cells in the non-theta and theta bins makes their rate differences more symmetrical, hence their lining up along the dotted line. spk/s, spikes per second

The opposite pattern of firing rate changes through the transition was obtained for complex-spike cells: 17 of 27 cells showed a firing rate increase between non- theta and transition periods that was followed by a firing rate decrease from transition to theta periods. Eight cells showed no change across the three periods, with seven of these cells firing no action potentials at all during the EEG transition. Two cells exhibited firing rate decreases from non-theta to transition and identical low rates (< 1 spk/s) in the transition and theta periods. Relative rates for all cells except the seven cells with 0 spk/s rates in all three bins are summarized in Fig. 2 (bottom). The average relative rates (+ SEM) were: 0.51 4- 0.11 for the non-theta period, 2.24 4- 0.19 for the transition period, and 0.244-0.04 for the theta period. Rates in the tran- sition period were significantly greater than in either surrounding period (t=6.157, 8.869; two-tailed; P<0.001). Although the means differed by 53%, the non-theta and theta rates did not differ significantly. Actual rates (4-SEM) for the three periods were 1.034-0.33 spk/s, 3.434-0.60 spk/s, and 0.55=t=0.14 spk/s, respectively. Stable non-theta and theta rates were ob- tained for 25 of 27 complex-spike cells and were as follows (:t:SEM): 2.56 4- 0.47 spk/s during non-theta EEG and 1.39 + 0.38 spk/s during theta EEG.

To parallel the summary of normalized rates shown in Fig. 2, which excluded the 7 cells with no firing in any of the three bins, unnormalized rate differences for all 27 complex-spike cells are shown in Fig. 3 (open

squares). All but two cells are located in quadrant IV (rate increase followed by rate decrease) or at the origin (no rate changes). When their stable non-theta and theta rates were examined, the two outlier cells showed a rate decrease: 2.86 to 0.71 spk/s and 0.93 to 0.78 spk/s.

The dotted line in Fig. 3 represents equal and opposite rate changes through the EEG transition. The complex- spike cells fall nearly along this line in quadrant IV. This is largely due to their very low firing rates in the non- theta and theta bins. Given that most complex-spike cells fire less during theta EEG (ignoring location-specific firing), one would expect more points between the ordi- nate and the dotted line. With their higher rates of firing, hippocampal theta cells show a rate increase from non- theta to theta (more cells between the ordinate and the dotted line in quadrant II of Fig. 3).

Discussion

The acitivity of hippocampal theta cells (presumed inter- neurons) and complex-spike cells (presumed principal cells) was examined during EEG transitions from non- theta EEG states to theta EEG states in freely behaving rats. Theta cell firing rates were significantly depressed in a 1-s period centered on the EEG transition relative to surrounding 1-s periods. Conversely, complex-spike cell firing rates were significantly increased during the tran- sition period. This diametrically altered activity indicates that theta cells must be actively inhibited during the transition. The period of inhibition is brief (on the order of 1 s) and is followed by a period of increased theta cell activity that lasts for the duration of the theta EEG state. Increases in complex-spike cell acitivity seen during the transition to theta rhythm may be simply a release from inhibitory control by interneurons.

The pattern of theta cell inhibition together with in- creased complex-spike cell activity appears to be a general property of transitions to theta rhythm. No dif- ferences were seen when transitions were examined in rats starting to walk or when they went into paradoxical sleep (always tonic rapid eye movement, REM, at the onset). Two theta cells were found to exhibit the same firing rate depression during transitions to theta rhythm in urethane-anesthetized rats (M. Stewart, unpublished observations).

The firing of hippocampal interneurons is postulated to be controlled by rhythmic excitatory (cholinergic) and inhibitory (GABAergic) projections from the medial sep- tal nucleus and nucleus of the diagonal band (reviewed in Stewart and Fox 1990). Firing rate changes observed in hippocampal neurons can be accounted for on the basis of reported increases in activity between non-theta and theta EEG states in rhythmically bursting medial septal neurons (Stewart and Fox 1989c). While both atropine-resistant (presumed GABAergic) and atropine- sensitive (presumed cholinergic) septal cells increase their firing rates from non-theta to theta, average rates for the atropine-resistant (presumed GABAergic) cells are higher. This firing rate difference may even be accen- tuated in the EEG transition if cholinergic septal cells are relatively more inhibited by the faster firing GABAergic septal cells.

The result o f this difference in firing rate between cholinergic and GABAerg i c project ion cells, in terms o f h ippocampa l in terneuron firing, is tha t inhibit ion by G A B A e r g i c septointerneuronal afferents exceeds excita- t ion by cholinergic septointerneuronal afferents in the transi t ion period. It is suggested that the relatively greater inhibit ion is no t sustained because o f an activity- dependent decrease in its potency, ana logous to activity- dependent changes in G A B A e r g i c t ransmission reported elsewhere (e.g., McCar ren and Alger 1985; Deisz and Prince 1989; T h o m p s o n and G/ihwiler 1989). Weaken ing of the inhibi tory c o m p o n e n t o f the sep toh ippocampal project ion leaves the rhy thmic exci tatory (cholinergic) septointerneuronal inputs, together with principal cell inputs, to increase in te rneuron firing rates. This in- creased in terneuron (theta cell) activity dur ing theta E E G has long been though t to account for the very low firing rates o f complex-spike cells (principal cells; e.g., Buzsaki et al. 1983; Fox et al. 1986). A n explanat ion o f decreased in terneuron firing rates based on a slower act ion o f acetylcholine relative to G A B A is unlikely, given the demons t ra t ion o f fast muscar inic excitation o f h ippocampa l alveus/oriens and basket cell in terneurons (Reece and Schwar tzkro in 1991) and the in terneurons ' ability to follow 4 to 10-Hz rhythmic septal acitivity (e.g., Stewart and F o x 1990).

One possible funct ional consequence o f the finding that h ippocampa l in terneurons are actively inhibited (al- lowing principal cells to fire at high rates) dur ing tran- sitions into theta rhy thm is tha t animals should be espe- cially susceptible to seizure onset at these times. The strong associat ion o f behavior with E E G state would permit at least crude behavioral cont ro l o f this trigger mechanism.

The mos t significant impediments to the s tudy o f the transi t ional E E G is the difficulty in obta in ing clean spon- taneous E E G transi t ions and the inability to apply m a n y averaging techniques, such as the cons t ruc t ion o f depth profiles using a single moveable electrode, in the analysis. Ins tan taneous depth profiles, derived f rom arrays o f re- cording electrodes, are perhaps the best way to inves- tigate the transi t ional theta E E G in detail.

Acknowledgements. This work was supported by National Institutes of Health grants NS09175 (to M. Stewart) and NS04352 (to J.B. Ranck, Jr.) and by National Science Foundation grant GB26184 (to J.B. Ranck, Jr.). The author thanks J.B. Ranck, Jr. and S.E. Fox for comments on earlier versions of the manuscript.

References

Brankack J, Fox SE (1987) Current sources for the alternating and sustained potentials of the hippocampal theta rhythms of the rat. Soc Neurosci Abstr 13:1331

Buzsaki G (1986) Hippocampal sharp waves: their origin and sig- nificance. Brain Res 398:242-252

Buzsaki G, Eidelberg E (1983) Phase relations of hippocampal projection cells and interneurons to theta activity in the anes- thetized rat. Brain Res 266:334-339

Buzsaki G, Leung LS, Vanderwolf CH (1983) Cellular bases of hippocampal EEG in the behaving rat. Brain Res Rev 6:139-171

Buzsaki G, Czopf J, Kondakor I, Kellenyi L (1986) Laminar distri- bution of hippocampal rhythmic slow activity (RSA) in the

behaving rat: current-source density analysis, effects of urethane and atropine. Brain Res 365:125-137

Deisz RA, Prince DA (1989) Frequency-dependent depression of inhibition in guinea-pig neocortex in vitro by GABAB receptor feedback on GABA release. J Physiol (Lond) 412:513-541

Fox SE, Ranck JB Jr (1975) Localization and anatomical identifica- tion of theta and complex-spike cells in dorsal hippocampal formation of rats. Exp Neurol 49: 299-313

Fox SE, Ranck JB Jr (1981) Electrophysiological characteristics of hippocampal complex-spike cells and theta cells. Exp Brain Res 41:399410

Fox SE, Stewart M (1986) Analysis of hippocampal theta rhythm shows that correct location of current sources and sinks generat- ing EEG patterns can require DC recording. Soc Neurosci Abstr 12:1527

Fox SE, Wolfson S, Ranck JB Jr (1986) Hippocampal theta rhythm and the firing of neurons in walking and urethane anesthetized rats. Exp Brain Res 62:495-508

Fox SE, Brankack J, Stewart M (1992) Origins of the hippocampal theta rhythm: entorhinal and septal contributions. Soc Neurosci Abstr 18:319

Kubie JL, Muller RU, Fox SE (1985) Firing fields of hippocampal place cells: interim report. In: Buzsaki G, Vanderwolf CH (eds) Electrical activity of the archicortex. Akademiai Kiado, Bu- dapest, pp 221-231

Lacaille JC, Mueller AL, Kunkel DD, Schwartzkroin PA (1987) Local circuit interactions between oriens/alveus interneurons and CA1 pyramidal cells in hippocampal slices: electrophysiol- ogy and morphology. J Neurosci 7:1979-1993

McCarren M, Alger BER (1985) Use-dependent depression of IPSPs ion rat hippocampal pyramidal cells in vitro. J Neuro- physiol 53 : 557-571

Muller RU, Kubie JL, Ranck JB Jr (1987) Spatial firing patterns of hippocampal complex-spike cells in a fixed environment. J Neurosci 7:1935-1950

Ntifiez A, Garcia-Austt E, Bufio W Jr (1987) Intracellular O-rhythm generation in identified hippocampal pyramids. Brain Res 416: 289-300

O'Keefe J, Dostrovsky J (1971) The hippocampus as a spatial map: preliminary evidence from unit activity in the freely moving rat. Brain Res 34:171-175

Ranck JB Jr (1973) Studies on single neurons in dorsal hippocampal formation and septum in unrestrained rats. I. Behavioral cor- relates and firing repertoires. Exp Neurol 41:461-531

Reece LJ, Schwartzkroin PA (1991) Effects of cholinergic agonists on two non-pyramidal cell types in rat hippocampal slices. Brain Res 566:115-126

Rudell AP, Fox SE (1984) Hippocampal excitability related to the phase of the theta rhythm in urethanized rats. Brain Res 294:350 353

Rudell AP, Fox SE, Ranck JB Jr (1980) Hippocampal excitability phase-locked to the theta rhythm in walking rats. Exp Neurol 68 : 87-96

Schwartzkroin PA, Mathers LH (1978) Physiological and mor- phological identification of a nonpyramidal cell type. Brain Res 157:1-10

Stewart M, Fox SE (1989a) Two populations of rhythmically bursting neurons in the rat medial septum are revealed by atropine. J Neurophysiol 61:982-993

Stewart M, Fox SE (1989b) Detection of an atropine-resistant component of the hippocampal theta rhythm in urethane anes- thetized rats. Brain Res 500:55-60

Stewart M, Fox SE (1989c) Firing relations of medial septal neu- rons to the hippocampal theta rhythm in urethane anesthetized rats. Exp Brain Res 77:507-516

Stewart M, Fox S E (1990) Do septal neurons pace the hippocampal theta rhythm? Trends Neurosci 13:163 168

Thompson SM, Gfihwiler BH (1989) Activity-dependent disinhibi- tion. I. Repetitive stimulation reduces IPSP driving force and conductance in the hippocampus in vitro. J Neurophysiol 61:501-511