95
Brain Research, 587 (1992) 95-101 © 19q2 Elsevier Science Publishers B.V. All rights reserved 0006-8993/92/$05.(}0 BRES 17943
Acetylcholine modulates averaged sensory evoked responses and perforant path evoked field potentials in the rat dentate gyrus T o m C. F o s t e r a and Sam A. D e a d w y l e r
b
" Department of Psychology, Unh'ersityof Virginia, Charlottest'ille, VA 22904 (USA) and b Department of Physiology and Phannacolo~,, Bowman Gray School of Medicine, Winston-Salem, NC 27103 (USA) (Accepted 10 March 1992)
Key words: Acetylcholine: Dentate gyrus: Averaged evoked potential: Field potential
The effect of localized application of acetylcholine (ACh) on well characterized components of sensory evoked and electrically induced potentials in the dentate gyrus was investigated in rats while performing a tone discrimination task. Local pressure application of ACh to the granule cell layer of the dentate gyrus through the recording pipette increased the amplitude of perforant path evoked population spikes without changing the amplitude of the field EPSP. When the pipette was relocated to the outer molecular layer of the dentate gyrus (OM), ACh application decreased the amplitude of the perforant path field EPSP, Two major components of the averaged auditory evoked potential (AEP) recorded during criterion performance of the discrimination task were significantlychanged by dendritic application of ACh. The N z component of the OM AEP which has been shown to reflect perforant path synaptic activity decreased in amplitude while the Nz component which represents activity from septal connections, was significantly increased. These effects were not due to the pressure ejection procedure nor drug related changes in behavioral performance of the task, The results suggest that ACh may act to differentially modulate the synaptic excitability of dentate granule cells, allowing them to acquire responses to sensory stimulation during the establishment and maintenance of discrimination learning.
INTRODUCTION
Previous studies indicate that the perforant path and septo-hippocampai projection to the dentate gyrus of the rat hippocampus mediate two characteristic negative components, N t and N z respectively, of the auditory evoked potential (AEP) recorded from this region during performance of a tone discrimination task s" m The early N t component reflects activity of perforant path synapses on dentate granule cells and exhibits a high degree of amplitude fluctuation related to a 'sequential dependency' during performance of twotone discrimination task m The amplitude of the late N z component is maximum in animals trained to criterion on a tone discrimination task, and is modulated by medial septai connections during acquisition and extinction of the behavior. Furthermore, the perforant path and septal inputs to the granule cells have been shown to differentially influence the discharge pattern
of dentate granule cells during performance of the two-tone discrimination task ~'~. The perforant path is the major excitatory afferent to the granule cells of the dentate gyrus :'~'~ and utilizes glutamate and possibly aspartate as the transmitter ':. The other major afferent input originates from cells in the medial septum, and diagonal band. It consists largely of cholinergic z2 and GABAergic fibers t,~. Heteros~,naptie interactions between septal and perforant path inputs have been demonstrated in the dentate gyrus L4.t~.zs. Prior activation of septo-hippocampal fibers increases the amplitude of perforant path evoked population spikes recorded in the granule cell layer t'~t. Disinhibitory influences on hippocampal neurons have been hypothesized for the medial septal projections a.4.zs, however, it has not been determined in vivo whether such disinhibitory actions facilitate or depress sensory driven activity in the dentate gyrus. The present experiment was designed to examine
Correspondence: S.A. Deadwyler, Department of Physiology and Pharmacology, Bowman Gray School of Medicine, Wake Forest University Medical Center, 300 S. Hawthorne St., Winston.Salem, NC 27103, USA, Fax: (i)(919) 748-7738.
g6 the influence of directly applied acetylcholine (ACh) on tone evoked potentials and granule cell population responses (field EPSPs and population spikes) recorded in the hippocampus of awake freely moving rats, The results provide evidence that cholinergic transmission in the dentate cyrus: (1) modifies granule cell discharge characteristics to perforant path synaptic inputs, (2) changes stimulation evoked synaptic responses in the dentate cyrus, and (3) alters well characterized sensory evoked potentials in the molecular layer of the dentate cyrus during performance of a tone discrimination task.
MATERIALS AND METHODS Specific procedures a~Ldcoordinates for in)plantation of slimulating and recording electrodes h~tvc been previously published t3 Briefly, h adult male Spraguc-D:,wley rats were anesthetized with sodium pentobarbital for impl;mlatkm of a removable microdrive assembly positioned on the skull over the hippocampus 7, The microdrive ;Ls~,'mbly could be mounted on the animal each day prior to the recording session. The assembly was n)odified to hold glass recording pipette microelcctrodcs which could be recorded from :is well ;is deliver cholinergic drugs to the recording site via micro-pressure ejection (Fizz. I). Bipolar stimulating cleelrod~Js were permanently impl:mted in Ihe ,regular bundle of one hemisphere for localization of rite cleclrodc tip during insertion, After recovery from surgery. :mim:ds were water deprived to ~Sq~ .d lib body well, hi end trained Io a criteria of less than 3 rcspon,~es per trial on a single tone diserimin:ition task, A 5 s lone (3 kilt)wits presellted on avertlge once every riO s (range 3l)~t)O s). Nose.pokes Ihrough a photocell beam during lone presentation fl,,livei~d a water reward and terminated the lone, Recording sessions conSislcd of Illll= 150 Irials per day. Fig, I ,,,howL,,the ret:ordin~ and drug delivery ± ~ystcut as well as the hchavioral condititming pttrtldigm, Glass pipeltes (lip diumeter I~.~ # m l filled with AC'h dissolved itt ~alhle (~(10 raM, pll %,% or saline triune were loaded into tile mlerodrive assembly immediately prior to the recordiltg session, Th~ eh:clrode lip Was localized by physiological criteria to eilher the middle of the mt)lecular layer or the grunule cell layer utilizing depth and field potential profiles evoked via electrical activation of the pcrforant path ~'~.A complelely sealed silaslic tube (PE IOI connected the pipette to the pressure delivery system which was connected through a modified commutator. A chloride coaled silver wire carried cleetrophysiological signals fr-m the saline in the pipette to the electrical cable through a small pinhole in ,ac tubing which was sealed around the wire to prevent leakage, Local application of drugs via pressure/ejection (3l psi, 20-35 ms) was regulated by a Picospritzer (General Valve Co.) and delivery times were controlled by a computer which synchronized delivery to behavioral and elcctrophysiological events (Fig. I ). Signals from the pipette were amplified, low pass filtered (l).5 l l z - I . t l kllz), displayed on an oscilloscope and stored on tape or computer disk for off-line analysis. Prior |o each behavioral recording session the tip of Ihe pipette was positioned in the dentate cyrus and i,put-output curves of field EPSPs a,d population spike amplitudes to pcrforant path stimulalion (I00 hiS, (l,2 llz) were conslrnclcd heft)re and after delivery of ACh. Mea~urenle,t of field potentials consisted of an average of 5 stimulus elicited potentials preceding (baseline) and 5 potentials at a given stimulus intensity hallowing ACh or saline application through the recording pipette. Granule cell popnlation spike amplitudes were measured from spike onset to peak and field EPSPs were measured from baseline to the amplitude at 2.0 ms after onset. Seven different stimulus intensities which ranged from threshold to maximum population spike and field EPSP amplitudes were used. Effects on field EPSP amplitudes after ACh or saline application are reported as
Pi.n.o
~
~
-.,~(~-Er
SCOPE
i ..qF ,u0
Tone Response.,,
ns.e,~
I I I
~el"Z~ .f~l~l~
I
Slim ~ r ]
/-II,...
I
...~l~
I p,,o,o,I ~,,l, (-50o ms~
iV i/~''Fiei d Polentlal
I I' • 1 '.0
f I
v
"2
,
0,=
o.,
. c.6sEc
Fig. I. Illustration of ACh and saline application and electrophysiological recording techniques. Top: ACh delivery and electmphysiological recording device. Pipettes filled with saline and/or ACh in s,line were lowered into the hippocampus via the indicated microdrive mounted on the skull Electmphysiological signals were led via a silver wire. through the PE tubing to a connector and cable assembly to II1~ amplifier lind subsequently displayed and registered ill Ihe computer. Delivery of drug out of the pipette was produced via air pressure ,ppllcd In Ihc PU tubing, which was controlled by Ih~ computer. A commutator and swivel through which the air line and cleetrophysitflotfletd signals pass allowed the atdmal freedom of movement during pcrfiwmanec of the behavioral task. Bottom: diagram of behavioral paradigm und ACh and saline application procedure. Tone onset (Tone) signaled the start of rite trial. A nose-poke through the light beam of a photocell (Response) during tone presentation resulted in the delivery of a drop of water (Reward) and terminated the tone. The perforant path was stimulated (Slim) 50(| ms prior to tone onset. The perforant path field potential is shown on an expanded time scale ( -51)(| ms) and in relation to the N I and N, components of the a M AEP flower trace). ACh or ~line was delivered through the recording pipette 15 s prior to tone onset on randomly selected trials during the session.
percent change from baseline to better illustrate specific actions of ACh on different amplitude potentials, During behavioral recording sessions, pipettes were localized to file outer molecular layer of the dentate cyrus ( a M ) utilizing field potential profiles, ACh and saline delivery trials were presented randomly on 30~ of the trials throughout the 100 trial session. ACh or saline delivery was initiated 15 s prior to tone onset. Field EPSPs and population spikes were elicited by electrical stimulation of perforant path 500 ms prior to tone-onset (Fig. I. bottom). Stimulus intensity was adjusted to elicit field EPSPs of approximately 1.0 mV amplitude. The amplitudes of N l and N2 components of sensory e~t)ked a M AEPs were stored on computer, sorted and averaged according to ACh or saline vs non-delivery trials within the same session,
97 differences in behavioral performance levels for any of the 6 animals tested as a function of local delivery of ACh or saline. The results indicate that changes in behavioral performance were not responsible for modification of electrophysiological measures observed during ACh delivery trials described below.
RESULTS
Lack of effect of ACh or saline application on behavior Behavioral performance on the tone discrimination task was maintained at criterion performance levels which was an average of < 3 responses per tone presentation trial throughout the recording sessions, it was possible that ACh might have a behavioral as well as a physiological effect. Hence saline vs. ACh trials were sorted and compared in terms of the following behavioral measures, latency to respond during the tone stimulus, intertrial responding after ejection of ACh or saline, and mean number of responses/trial for ACh vs saline trials. In addition, ACh or saline delivery trials were compared to non-delivery trials to determine possible non-specific effects on discrimination behavior. Examination of the effects of ACh, saline or non-delivery trials revealed no significant
Effects of ACh application on dentate gyrus field EPSPs and population spikes Fig. 2 illustrates the effect of application of ACh to the granule cell layer of the dentate gyrus on the perforant path evoked population spike. Electrical stimulation of the perforant path resulted in the characteristic negative-going population spike superimposed on the rising slope of field EPSP (Fig. 2). An increase in population spike amplitude and associated late positivity was observed at all stimulus intensities after local delivery of ACh through the recording
U
'" 1701
T
°"°I
Y O
25'
5 10 15 20 25 30 O SECONDS
b
12.
12.
8"
B.
4 mV
0
8
o
1'6
~
;6
8
~'o
TIMI~ (ms)
e
f 12
Z~ 20' < U .J
15,
o
-5
1 10 20 ,40 60 80 100 PERCENT OF" M A X I M U M CONTROL AMPLITUDE
0,~,,.,,.
1~
1'6
o
Electrical Stimulus
Fig. 2. Time course of enhanced amplitude of dentate gyrus population spike after application of ACh averaged over ;ill animals. The effects of ACh on perforant path input-output curves was determined by averaging 5 potentials before drug application as a control for each of 6 different stimulus intensities, a: mean percent increase in population spike amplitude across animals and stimulus intensity relative to control during 35 s after ACh ejection, h: mean proportional increase in perforant path population spike 15 s after drug application for different intensities of perforant path stimulation ranging from spike threshold to 100% of maximum baseline amplitude. Error bars in a and h represent ± S.E.M. c-f: time course of facilitation for the perforant path population spike after delivery of ACh from the recording pipette Iocaled in the granule cell layer, c: before, d-f: S, 10, 15 and 25 s after application of ACh. The increase in population spike amplitude was evident 5 s after ACh application reaching a maximum at 15 s. ACh application to the granule cell layer did not significantly alter the field EPSP. The arrow in c indicates the point at which EPSP amplitude was assessed.
98
lation spike was specific to ACh and not due to the pressure-ejection procedure. A one-way ANOVA of the percent change in population spike amplitude averaged across stimulus intensities on sessions in which either ACh or saline was delivered showed a significant (Fl,17 = 7.61, P < 0.05) increase following ACh application (mean change ACh 137 + 11.3% S.E.M.; mean change saline 99 + 4.9%). There was no consistent effect of ACh or saline on the field EPSPs recorded in the granule cell layer (Fig. 2c-e). Following tests in the granule cell layer, the recording pipette was raised 150 p m to the OM and the effects on stimulus elicited and sensory evoked potentials evaluated on a trial-by-trial basis during performance of the tone discrimination task. Utilizing the time-course of the effect on the population spike (Fig.
pipette. The population spike enhancement by ACh was detectable within the first 5 s (125 :t: 6% S.E.M.) after delivery and the maximum effect (156 + 9%) occurred within 10-15 s (Fig 2a). The total duration of the increase was 35 s, at which point population spike amplitude returned to baseline (Fig. 2c-e). Fig. 2b shows the interaction of population spike amplitude with magnitude of ACh produced increase. The largest proportional change ((postACh-preACh)/maximum baseline amplitude) in population spike amplitude was in the mid-range of the input-output curve as shown in Fig. 2b. The increase in the late positive response was assumed to reflect a corresponding increase in inhibition as a result of the increased population of neurons activated by the stimulus following ACh application. An increase in individual granule cell spontaneous discharge was observed for a brief time (5-20 s) following ACh application in a few instances where isolated cells could be identified and activated by the stimulus at intensities below population spike threshold. No consistent effect on field EPSP or population spike was observed after saline delivery from the pipette, indicating that the enhancement of the popu-
2a) to determine the time of A C h delivery, a pressure ejection pulse to the recording pipette occurred 15 s p r i o r to tone onset on randomly selected trials d u r i n g
the behavioral session. In 10 different sessions involving 3 different animals dendritic field EPSPs from electrical stimulation of the perforant path were elicited 500 ms prior to tone onset (Fig. 1). in 7 of 7 instances
b
N 1
I
S01,ne
t
f~//~"'~ ~//~a',
';, .,
>~
1001
- 100
L
......
- 200 ACh
~aUne v$ Cont r o l
v$ Cont rol
N2 :~,'L),~[\;
~",, ,'1 1 T o n e on
~/
" IV"
1 " 200
' mse¢
>
=
[
~-,oo
.t AC~ V$ ContrOl
Saline V$ Control
Fig. 3. a: mean tone evoked OM AEPs (solid line) and $.E.M, (dashed line) average across animals and sessions after delivery of saline (top) or ACh (bottom), 15 s beE)re tone onset. No significant difference in amplitude was observed for N t and N 2 across non-delivery or saline delivery trials, b: group mean difference in the amplitude of N t (top) and N~ (bottom). Delivery of ACh significantly decreased the amplitude of N I and increased N 2 amplitude, Bars represent S,E,M.s.
99 small but consistent (t 6 = 31.75, P < 0.0001) reductions in the amplitude of the dendritic field EPSPs were observed following application of ACh to the OM (150 p~m above 9-cell layer). The field EPSP amplitude was measured at the same point as in Fig. 2. Saline delivery produced no significant change in the perforant path field EPSP in the same number of animals. Effect of ACh application on the tone eL,eked OM A E P The conditioned tone stimuli reliably elicited the well characterized waveform of the tone evoked OM AEP previously reported during criterion performance of this task ~'='. Fig. 3a shows the distinct N R and N 2 negative components of this evoked potential on trials in which saline or ACh was delivered from the pipette. A marked reduction in N~ amplitude occurred following ACh delivery 15 s prior to tone onset. A one-way ANOVA examining changes in the amplitude of Nj indicated a significant decrease during ACh delivery trials compared to non-delivery trials (F~,~ = 9.13, P < 0.01). The magnitude of the ACh induced decrease in NI amplitude varied for different recording sessions and different animals (range - 6 to -319/zV; mean difference - 126.22 4- 38.98 #V S.E.M.). Alternatively no significant differences in N= amplitude were found comparing saline delivery with non-delivery trials. The differences which did occur following saline delivery were in the opposite direction to those related to ACh (mean difference 50.5 + 47.38 ,¢V; Fig. 3b). A marginally significant difference in N~ amplitude between ACh and saline delivery trials was obtained when evaluated 'across test sessions (Fl,17---8.44, P < (I.(12). Finally, the amplitude of the N a pot0ntial on non-delivery trials was not significantly different for either ACh or saline test sessions. Thus the effects of ACh were limited to delivery trials only and were dissipated before the next (non.delivery) trial occurred. The influence of ACh on the longer latency N 2 component of the OM AEP was the reverse of the effect on N=. N 2 amplitude increased after ACh application relative to non-delivery trials in 8 out of 9 tests (mean increase 127.33 4- 59.26/.~V S.E.M.). The difference was not significant by ANOVA due to the large range of amplitude changes ( - 384-114 p,V), however, when the differences in N 2 amplitude were normali~,.ed relative to percent change between ACh delivery vs non-delivery trials a sighificant increase in N 2 amplitude was observed (t 8 = 2.485, P < 0.037). In comparison to saline delivery trials mean N 2 amplitude was increased by 144/~V on ACh delive.,'y trials; however, this difference is somewhat artificial since saline and ACh trials were administered during different recording sessions. N, amplitude was not significantly changed
on saline delivery vs non-delivery trials (Fig. 3b), nor did it change significantly across sessions on non-delivery trials when either ACh or saline were in the recording pipette (mean difference in N 2 amplitude = 2.98 p.V for ACh non-delivery vs. saline non-delivery trials). Thus ACh significantly and differentially altered the amplitudes of both N! and N 2, the two well characterized components of the OM AEP. The data indicate that the amplitudes of Nj and N 2 were not affected by leakage of ACh from the pipette in that the changes were specific to delivery vs. non-delivery trials. In addition, the lack of behavioral effects across different animals and sessions in which a relatively high concentration of ACh (500 raM) was retained in the recording pipette supports the likelihood that the effects of ACh application were not secondarily associated with alterations in motivational or attentional factors. Although attempts were made to block the effects of ACh on tone evoked potentials with systemic applications of the muscarinic cholinergic antagonists, scopolamine and.atropine, the experiments were confounded by the fact that animals stopped responding after injections of both these drugs at doses which influenced evoked potentials.
DISCUSSION Local application of ACh to physiologically identified regions of the dentate gyrus through the recording pipette significantly altered electrically elicited, perforant path mediated, field potentials as well as identified components of the sensory evoked OM AEP in awake freely moving rats. In no instances were these changes associated with a change in behavioral perfor. mance of the tone discrimination task. The differential influence of ACh delivery on the N t and N 2 components and the lack of effect on saline delivery trials indicate that the effects were not due to the method of application through the recording pipette. These results support and extend previous findings that the Nm and N 2 components result from different neural substrates acting on the dentate granule cells 8-~.1.~ Moreover, the differential change in amplitude of the two components (decreased Nj and increased N 2) after dendritic application of ACh was similar to changes observed in these components during acquisition and retraining of the task and opposite those observed during extinction of the tone discrimination task s-~0 Thus, the activation of an endogenous cholinergic projection to the dentate gyrus could be responsible for
!110 the change in amplitude of these potentials which occurs during establishment of the behavioral significance of the tone stimuli through associative conditiontire. The ACh induced reduction in the amplitude of the Nt component of the OM AEP and the perforant path elicited field EPSP provides further evidence that the Nt component represents activation of pefforant path-granule cell synapses. Depth profile analysis and lesion data in previous studies demonstrated that the site of origin of Nt is in the outer two-thirds of the dentate gyrus molecular layer ~.13.These findings were further substantiated by a recent report which showed that ACh reduces the field EPSP from medial but not lateral perforant path activation in the hippocampal slice preparation ~. The increase in population spike amplitude recorded in the granule cell layer in the current study is similar to other findings showing cholinergic enhancement of the population spike 21,26.3o in the CAI region, but does not agree with earlier reports of a lack of cholinergic influence via microiontophoretic application of ACh on field potentials which did not exhibit population spikes (see ref. 21, Fig. 9) in the dentate gyrus 20,21 It is therefore quite possible that the influence of ACh is dependent on the presence of a small population spike in order for facilitation of the population spike to be observed (i,e, ,rimulation must exceed threshold), The A C h induced facilitation of the population spike and depression of the field EPSP in the CAI region has been attributed to (1) a reduction in the release of inhibitory and excitatory neurotransmitters respectively ,~,4,tv,:,~ and (2) a slow depolarization via inhibi. tion of K* channels through muscarinic receptor mediated inactivation of the M-current zr,,:4,:7, The widespread terminvtion of septal cho!inergie fibers in the dentate gyrus and ~he strategic localization of these terminals on both granule cells and inhibitory interneu. rons 14 supports the concept of widespread cholinergic modulation of granule cell activity 29. Frotscher and Leranth 14 identified choline acetyltransferase positive synapses on the dendrites and cell bodies of granule cells indicating both direct and indirect autions of ACh on granule cells via local circuit processes. The observed increase in the perforant path evoked population spike following ACh delivery in this experiment agrees with prior in vivo studies in which activation of the septal-cholinergic pathway increased the responsiveness of granule cells to perforant path synaptic inputs in anesthetized animals z.n.25. Previous reports from this laborator~j have shown that tone evoked granule cell discharge is significantly increased during presentation of reinforced vs non-reinforced
tones 12 in animals performing a two-tone discrimination task. Lesions of the entorhinal cortex and perforant path eliminate all tone evoked granule cell activity, while lesions of the medio-lateral septum result in non-differentiated discharges to either reinforced or non-reinforced tones ~3. The current results provide further insight into the role of ACh in controlling these learning related changes in granule cell activity, and suggest that cholinergic projections to the dentate gyrus during acquisition of a tone discrimination task are partially responsible for altering the reactivity to sensory stimuli associated during such procedures, it is not insignificant that ACh application was sufficient to increase the septal influence reflected by N 2 as well as simultaneously reduce the influence of sensory stimuli via the perforant path synapses on granule cell dendrites as indicated by the reduction in Nz. Whether or not such reciprocal changes in tone evoked activity result from functional interaction within the local circuitry of the dentate gyrus, or the operation of feedback connections between the dentate gyrus, hippocampus, medial septum and entorhinal cortex, is a question for further investigation. Acknowh,dgements. Figure I was drawn by M, West. This re.';earch was supported by NIDA Grunts DA03502, DA04441 and DA00119 and a gift from RJ Reynolds'tobacco Co, to S,A.D.
REFERBNCES I Alvarez.Leefmun,~, FJ, and Oardm:r.Medwin, A,R,, Influence of the septum on the hlppo~ampal dentate ar~a width are unaccom. panled by field put~ntials,/, PhysioL, 249 (t975) 1~. 16. 2 Andersen, P,, Holmqpist, B. and, Voorhoeve, P,B, Entorhinal activation of dentate granule cells, Acta Phystol Sound, 66:44846l), 1966. 3 Ben-Ari, Y,, Krnjevi~, K,, Reinhardt, W, and Ropert, N,, Intracellular observations on the disinhibitory action of acetylcholine in the hippocampus, Neuroscicnce, 6 (1981) 24?$-2484. 4 Bilkey, D,K, and Goddard, G.V., Medial septal facilitation of hippocampal granule cell actb;b i~, mediated by inhibition of inhibitory interneurons, groin Rcs., 361 (1985) 99-106. $ Blackstad, T,W,, Ultrastructural studies on the hippocampal region, Ping. Brain Rex., 3 (1963) 122-148, 6 Consolo, S,, Wang, G.X., Fusi, R., Vinci, R,, Forloni, G. and Ladinsky, H., In vitro and in vivo evidence for the existence of presynaptic muscarinic eholinergie receptors in the rat hippocampus, Brain Res., 309 (1984) 147-151, 7 Deadwyler, S.A., Biela, J., Rose, G., West, M. and Lynch, G.A., Microdrive for use with glass or metal microelectrodes in recording from freely-moving rats, El~tmencephalol~r, Clin, Neumphysiot,, 4'7 (1979) 752-754, 8 Deadwyler, S,A, West, M, and Lynch, O,, Synapfically identified slow potentials during behavior, Brain Res., 161 (1979) 211-225, 9 Dcadwyler, S,A,, West, M,O, and Robinson, J.H., Entorhinal and
septal inputs differentiallycontrol sensory.evokedresponses in the rat dentate 8Yrus,Science, 211 (1981) !181-1183. i0 Deadwyler, S,A,, West, M.O,, Christian, E,P,. Hampson, R.E, and Foster, T,C,, Sequence-related changes in the sensory-evoked potentials in the dentate gyrus: a mechanism for item-specific short-term information storage in the hippocampus, Behav. Neural Biol,, 44 (1985) 201-212.
101 11 Fantie, B.D and Goddard, G.V., Septal modulation of the population spike in the fascia dentata produced by perforant path stimulation in the rat, Brain Res., 252 (1982) 227-237. 12 Fonnum, F., Glutamate: a neurotransmitter in mammalian brain, J. Neurochem., 42 (1984) i - I I . 13 Foster, T.C, Hampson, R.E., West, M.O. and Deadwyler, S.A., Control of sensory activation of granule cells in the fascia dentata by extrinsic affercnts: septal and entorhinal inputs, J. Neurosci., 8 (1988) 3869-3878. 14 Frolscher, M. and Leranth, C., The cholinergic innervation of ihe rat fascia dentata: identification of target structures on granule cells by combining choline acetyitransferase immunocytochemistry and golgi impregnation, J. Comp. Neurol., 243 (1986) 58-70. 15 Gottlieb, D.I. and Cowan, W.M., On the distribution of axonal terminals containing spheroidal and flattened synaptic vesicles in the hippocampus and dentate g.vrus of the rat and cat, Z. Zellforsch. Mikrosk. Anat., 129 (1972) 413-419. 16 Halliwell, J.V, and Adams, P.R., Voltage-clamp analysis of muscarinic excitation in hippocampal neurons, Brain Res., 250 (1982) 71-92, 17 Hounsgaard, J., Prcsynaptic inhibitory action of acetylcholine in area CAI of the hippocampus, Exp. Neuroi., 62 (19'78) 787-797. 18 Kahle, .I.S. and Cotman, C.W., Carbachol depresse~ synaptic responses in the medial but not the lateral perforant path, Brain Res., 482 (1989) 159-163. 19 Kohler, C., Chan-Palay, V, and Wu, J,Y., Septal neurons containing glutamic acid decarboxyla~ immunoreactivity project to the hippocampal region in the rat brain, Anat. Embryol., 169 (1984) 41-44, 20 Krnjevi~, K., Reiffenstein, R.J. and Ropert, N., Disinhibitory action of acet),lcholine in the rat's hippocampus: Extracellular observations, Nearoscience, 6 ( 1981) 2465-2474.
21 Krnjevi~, K. and Ropert, N., Electrophysiological and pharmacological characteristics of facilitation of hippocampal pepulation spikes by stimulation of the medial septum, Neuroscience, 7 (1982) 2165-2183. 22 Lewis, P.R. and Shute, C.C.D., The cholinergic limbic system: projections to hippocampal, medial cortex, nuclei of the ascending cholinergic reticular system and the subfornical organ and supra-optic crest, Brain, 90 (1967) 521-540. 23 Lomo, T., Patterns of activation in a monosynaptic cortical pathway: the perforant path input to the dentate area o[ the hippocampal formation, Exp. Brain Res., 12 (1971) 18-45. 24 Madison, D.V., Lancaster, B. and Nicoll, R.A., Voltage clamp analysis of cholinergic action in the hippocampus, J. Neurosci., 7 (1987) 733-741. 25 Robinson, G.B. and Racine, RJ., Interactions between septal and entorhinal inputs to the rat dentate gyrus: facilitation effects, Brain Res., 379 (1986) 63-67. 26 Ropert, N. and Krnjevi~, K., Pharmacological characteristics of facilitation of hippocampal population spikes by cholinomimetics, Neuroscience, 7 (1982) 1963-1977. 27 Segal, Mo, Synaptic activation of a cholinergic receptor in rat hippocampus, Brain Res., 452 (1988) 79-86. 28 Stewart, M. and Fox, S.E. Do septal neurones pace the hippocampal theta rhythm? Trends Neurosci., 13 (1990) 163-168. 29 Valentino, R.J. and Dingledine, R., Presynaptic inhibitory effect of acetylcholine in the hippocampus, J. Neurosci., 1 (1981)784= 792. 30 Xie, X,, Inhibition of long-term potentiation of hippocampal CA1 EPSP by a negative cooperativity, Brain Res., 445 (1988) 367-369.
Brain Research, 587 (1992) 95-101 © 19q2 Elsevier Science Publishers B.V. All rights reserved 0006-8993/92/$05.(}0 BRES 17943
Acetylcholine modulates averaged sensory evoked responses and perforant path evoked field potentials in the rat dentate gyrus T o m C. F o s t e r a and Sam A. D e a d w y l e r
b
" Department of Psychology, Unh'ersityof Virginia, Charlottest'ille, VA 22904 (USA) and b Department of Physiology and Phannacolo~,, Bowman Gray School of Medicine, Winston-Salem, NC 27103 (USA) (Accepted 10 March 1992)
Key words: Acetylcholine: Dentate gyrus: Averaged evoked potential: Field potential
The effect of localized application of acetylcholine (ACh) on well characterized components of sensory evoked and electrically induced potentials in the dentate gyrus was investigated in rats while performing a tone discrimination task. Local pressure application of ACh to the granule cell layer of the dentate gyrus through the recording pipette increased the amplitude of perforant path evoked population spikes without changing the amplitude of the field EPSP. When the pipette was relocated to the outer molecular layer of the dentate gyrus (OM), ACh application decreased the amplitude of the perforant path field EPSP, Two major components of the averaged auditory evoked potential (AEP) recorded during criterion performance of the discrimination task were significantlychanged by dendritic application of ACh. The N z component of the OM AEP which has been shown to reflect perforant path synaptic activity decreased in amplitude while the Nz component which represents activity from septal connections, was significantly increased. These effects were not due to the pressure ejection procedure nor drug related changes in behavioral performance of the task, The results suggest that ACh may act to differentially modulate the synaptic excitability of dentate granule cells, allowing them to acquire responses to sensory stimulation during the establishment and maintenance of discrimination learning.
INTRODUCTION
Previous studies indicate that the perforant path and septo-hippocampai projection to the dentate gyrus of the rat hippocampus mediate two characteristic negative components, N t and N z respectively, of the auditory evoked potential (AEP) recorded from this region during performance of a tone discrimination task s" m The early N t component reflects activity of perforant path synapses on dentate granule cells and exhibits a high degree of amplitude fluctuation related to a 'sequential dependency' during performance of twotone discrimination task m The amplitude of the late N z component is maximum in animals trained to criterion on a tone discrimination task, and is modulated by medial septai connections during acquisition and extinction of the behavior. Furthermore, the perforant path and septal inputs to the granule cells have been shown to differentially influence the discharge pattern
of dentate granule cells during performance of the two-tone discrimination task ~'~. The perforant path is the major excitatory afferent to the granule cells of the dentate gyrus :'~'~ and utilizes glutamate and possibly aspartate as the transmitter ':. The other major afferent input originates from cells in the medial septum, and diagonal band. It consists largely of cholinergic z2 and GABAergic fibers t,~. Heteros~,naptie interactions between septal and perforant path inputs have been demonstrated in the dentate gyrus L4.t~.zs. Prior activation of septo-hippocampal fibers increases the amplitude of perforant path evoked population spikes recorded in the granule cell layer t'~t. Disinhibitory influences on hippocampal neurons have been hypothesized for the medial septal projections a.4.zs, however, it has not been determined in vivo whether such disinhibitory actions facilitate or depress sensory driven activity in the dentate gyrus. The present experiment was designed to examine
Correspondence: S.A. Deadwyler, Department of Physiology and Pharmacology, Bowman Gray School of Medicine, Wake Forest University Medical Center, 300 S. Hawthorne St., Winston.Salem, NC 27103, USA, Fax: (i)(919) 748-7738.
g6 the influence of directly applied acetylcholine (ACh) on tone evoked potentials and granule cell population responses (field EPSPs and population spikes) recorded in the hippocampus of awake freely moving rats, The results provide evidence that cholinergic transmission in the dentate cyrus: (1) modifies granule cell discharge characteristics to perforant path synaptic inputs, (2) changes stimulation evoked synaptic responses in the dentate cyrus, and (3) alters well characterized sensory evoked potentials in the molecular layer of the dentate cyrus during performance of a tone discrimination task.
MATERIALS AND METHODS Specific procedures a~Ldcoordinates for in)plantation of slimulating and recording electrodes h~tvc been previously published t3 Briefly, h adult male Spraguc-D:,wley rats were anesthetized with sodium pentobarbital for impl;mlatkm of a removable microdrive assembly positioned on the skull over the hippocampus 7, The microdrive ;Ls~,'mbly could be mounted on the animal each day prior to the recording session. The assembly was n)odified to hold glass recording pipette microelcctrodcs which could be recorded from :is well ;is deliver cholinergic drugs to the recording site via micro-pressure ejection (Fizz. I). Bipolar stimulating cleelrod~Js were permanently impl:mted in Ihe ,regular bundle of one hemisphere for localization of rite cleclrodc tip during insertion, After recovery from surgery. :mim:ds were water deprived to ~Sq~ .d lib body well, hi end trained Io a criteria of less than 3 rcspon,~es per trial on a single tone diserimin:ition task, A 5 s lone (3 kilt)wits presellted on avertlge once every riO s (range 3l)~t)O s). Nose.pokes Ihrough a photocell beam during lone presentation fl,,livei~d a water reward and terminated the lone, Recording sessions conSislcd of Illll= 150 Irials per day. Fig, I ,,,howL,,the ret:ordin~ and drug delivery ± ~ystcut as well as the hchavioral condititming pttrtldigm, Glass pipeltes (lip diumeter I~.~ # m l filled with AC'h dissolved itt ~alhle (~(10 raM, pll %,% or saline triune were loaded into tile mlerodrive assembly immediately prior to the recordiltg session, Th~ eh:clrode lip Was localized by physiological criteria to eilher the middle of the mt)lecular layer or the grunule cell layer utilizing depth and field potential profiles evoked via electrical activation of the pcrforant path ~'~.A complelely sealed silaslic tube (PE IOI connected the pipette to the pressure delivery system which was connected through a modified commutator. A chloride coaled silver wire carried cleetrophysiological signals fr-m the saline in the pipette to the electrical cable through a small pinhole in ,ac tubing which was sealed around the wire to prevent leakage, Local application of drugs via pressure/ejection (3l psi, 20-35 ms) was regulated by a Picospritzer (General Valve Co.) and delivery times were controlled by a computer which synchronized delivery to behavioral and elcctrophysiological events (Fig. I ). Signals from the pipette were amplified, low pass filtered (l).5 l l z - I . t l kllz), displayed on an oscilloscope and stored on tape or computer disk for off-line analysis. Prior |o each behavioral recording session the tip of Ihe pipette was positioned in the dentate cyrus and i,put-output curves of field EPSPs a,d population spike amplitudes to pcrforant path stimulalion (I00 hiS, (l,2 llz) were conslrnclcd heft)re and after delivery of ACh. Mea~urenle,t of field potentials consisted of an average of 5 stimulus elicited potentials preceding (baseline) and 5 potentials at a given stimulus intensity hallowing ACh or saline application through the recording pipette. Granule cell popnlation spike amplitudes were measured from spike onset to peak and field EPSPs were measured from baseline to the amplitude at 2.0 ms after onset. Seven different stimulus intensities which ranged from threshold to maximum population spike and field EPSP amplitudes were used. Effects on field EPSP amplitudes after ACh or saline application are reported as
Pi.n.o
~
~
-.,~(~-Er
SCOPE
i ..qF ,u0
Tone Response.,,
ns.e,~
I I I
~el"Z~ .f~l~l~
I
Slim ~ r ]
/-II,...
I
...~l~
I p,,o,o,I ~,,l, (-50o ms~
iV i/~''Fiei d Polentlal
I I' • 1 '.0
f I
v
"2
,
0,=
o.,
. c.6sEc
Fig. I. Illustration of ACh and saline application and electrophysiological recording techniques. Top: ACh delivery and electmphysiological recording device. Pipettes filled with saline and/or ACh in s,line were lowered into the hippocampus via the indicated microdrive mounted on the skull Electmphysiological signals were led via a silver wire. through the PE tubing to a connector and cable assembly to II1~ amplifier lind subsequently displayed and registered ill Ihe computer. Delivery of drug out of the pipette was produced via air pressure ,ppllcd In Ihc PU tubing, which was controlled by Ih~ computer. A commutator and swivel through which the air line and cleetrophysitflotfletd signals pass allowed the atdmal freedom of movement during pcrfiwmanec of the behavioral task. Bottom: diagram of behavioral paradigm und ACh and saline application procedure. Tone onset (Tone) signaled the start of rite trial. A nose-poke through the light beam of a photocell (Response) during tone presentation resulted in the delivery of a drop of water (Reward) and terminated the tone. The perforant path was stimulated (Slim) 50(| ms prior to tone onset. The perforant path field potential is shown on an expanded time scale ( -51)(| ms) and in relation to the N I and N, components of the a M AEP flower trace). ACh or ~line was delivered through the recording pipette 15 s prior to tone onset on randomly selected trials during the session.
percent change from baseline to better illustrate specific actions of ACh on different amplitude potentials, During behavioral recording sessions, pipettes were localized to file outer molecular layer of the dentate cyrus ( a M ) utilizing field potential profiles, ACh and saline delivery trials were presented randomly on 30~ of the trials throughout the 100 trial session. ACh or saline delivery was initiated 15 s prior to tone onset. Field EPSPs and population spikes were elicited by electrical stimulation of perforant path 500 ms prior to tone-onset (Fig. I. bottom). Stimulus intensity was adjusted to elicit field EPSPs of approximately 1.0 mV amplitude. The amplitudes of N l and N2 components of sensory e~t)ked a M AEPs were stored on computer, sorted and averaged according to ACh or saline vs non-delivery trials within the same session,
97 differences in behavioral performance levels for any of the 6 animals tested as a function of local delivery of ACh or saline. The results indicate that changes in behavioral performance were not responsible for modification of electrophysiological measures observed during ACh delivery trials described below.
RESULTS
Lack of effect of ACh or saline application on behavior Behavioral performance on the tone discrimination task was maintained at criterion performance levels which was an average of < 3 responses per tone presentation trial throughout the recording sessions, it was possible that ACh might have a behavioral as well as a physiological effect. Hence saline vs. ACh trials were sorted and compared in terms of the following behavioral measures, latency to respond during the tone stimulus, intertrial responding after ejection of ACh or saline, and mean number of responses/trial for ACh vs saline trials. In addition, ACh or saline delivery trials were compared to non-delivery trials to determine possible non-specific effects on discrimination behavior. Examination of the effects of ACh, saline or non-delivery trials revealed no significant
Effects of ACh application on dentate gyrus field EPSPs and population spikes Fig. 2 illustrates the effect of application of ACh to the granule cell layer of the dentate gyrus on the perforant path evoked population spike. Electrical stimulation of the perforant path resulted in the characteristic negative-going population spike superimposed on the rising slope of field EPSP (Fig. 2). An increase in population spike amplitude and associated late positivity was observed at all stimulus intensities after local delivery of ACh through the recording
U
'" 1701
T
°"°I
Y O
25'
5 10 15 20 25 30 O SECONDS
b
12.
12.
8"
B.
4 mV
0
8
o
1'6
~
;6
8
~'o
TIMI~ (ms)
e
f 12
Z~ 20' < U .J
15,
o
-5
1 10 20 ,40 60 80 100 PERCENT OF" M A X I M U M CONTROL AMPLITUDE
0,~,,.,,.
1~
1'6
o
Electrical Stimulus
Fig. 2. Time course of enhanced amplitude of dentate gyrus population spike after application of ACh averaged over ;ill animals. The effects of ACh on perforant path input-output curves was determined by averaging 5 potentials before drug application as a control for each of 6 different stimulus intensities, a: mean percent increase in population spike amplitude across animals and stimulus intensity relative to control during 35 s after ACh ejection, h: mean proportional increase in perforant path population spike 15 s after drug application for different intensities of perforant path stimulation ranging from spike threshold to 100% of maximum baseline amplitude. Error bars in a and h represent ± S.E.M. c-f: time course of facilitation for the perforant path population spike after delivery of ACh from the recording pipette Iocaled in the granule cell layer, c: before, d-f: S, 10, 15 and 25 s after application of ACh. The increase in population spike amplitude was evident 5 s after ACh application reaching a maximum at 15 s. ACh application to the granule cell layer did not significantly alter the field EPSP. The arrow in c indicates the point at which EPSP amplitude was assessed.
98
lation spike was specific to ACh and not due to the pressure-ejection procedure. A one-way ANOVA of the percent change in population spike amplitude averaged across stimulus intensities on sessions in which either ACh or saline was delivered showed a significant (Fl,17 = 7.61, P < 0.05) increase following ACh application (mean change ACh 137 + 11.3% S.E.M.; mean change saline 99 + 4.9%). There was no consistent effect of ACh or saline on the field EPSPs recorded in the granule cell layer (Fig. 2c-e). Following tests in the granule cell layer, the recording pipette was raised 150 p m to the OM and the effects on stimulus elicited and sensory evoked potentials evaluated on a trial-by-trial basis during performance of the tone discrimination task. Utilizing the time-course of the effect on the population spike (Fig.
pipette. The population spike enhancement by ACh was detectable within the first 5 s (125 :t: 6% S.E.M.) after delivery and the maximum effect (156 + 9%) occurred within 10-15 s (Fig 2a). The total duration of the increase was 35 s, at which point population spike amplitude returned to baseline (Fig. 2c-e). Fig. 2b shows the interaction of population spike amplitude with magnitude of ACh produced increase. The largest proportional change ((postACh-preACh)/maximum baseline amplitude) in population spike amplitude was in the mid-range of the input-output curve as shown in Fig. 2b. The increase in the late positive response was assumed to reflect a corresponding increase in inhibition as a result of the increased population of neurons activated by the stimulus following ACh application. An increase in individual granule cell spontaneous discharge was observed for a brief time (5-20 s) following ACh application in a few instances where isolated cells could be identified and activated by the stimulus at intensities below population spike threshold. No consistent effect on field EPSP or population spike was observed after saline delivery from the pipette, indicating that the enhancement of the popu-
2a) to determine the time of A C h delivery, a pressure ejection pulse to the recording pipette occurred 15 s p r i o r to tone onset on randomly selected trials d u r i n g
the behavioral session. In 10 different sessions involving 3 different animals dendritic field EPSPs from electrical stimulation of the perforant path were elicited 500 ms prior to tone onset (Fig. 1). in 7 of 7 instances
b
N 1
I
S01,ne
t
f~//~"'~ ~//~a',
';, .,
>~
1001
- 100
L
......
- 200 ACh
~aUne v$ Cont r o l
v$ Cont rol
N2 :~,'L),~[\;
~",, ,'1 1 T o n e on
~/
" IV"
1 " 200
' mse¢
>
=
[
~-,oo
.t AC~ V$ ContrOl
Saline V$ Control
Fig. 3. a: mean tone evoked OM AEPs (solid line) and $.E.M, (dashed line) average across animals and sessions after delivery of saline (top) or ACh (bottom), 15 s beE)re tone onset. No significant difference in amplitude was observed for N t and N 2 across non-delivery or saline delivery trials, b: group mean difference in the amplitude of N t (top) and N~ (bottom). Delivery of ACh significantly decreased the amplitude of N I and increased N 2 amplitude, Bars represent S,E,M.s.
99 small but consistent (t 6 = 31.75, P < 0.0001) reductions in the amplitude of the dendritic field EPSPs were observed following application of ACh to the OM (150 p~m above 9-cell layer). The field EPSP amplitude was measured at the same point as in Fig. 2. Saline delivery produced no significant change in the perforant path field EPSP in the same number of animals. Effect of ACh application on the tone eL,eked OM A E P The conditioned tone stimuli reliably elicited the well characterized waveform of the tone evoked OM AEP previously reported during criterion performance of this task ~'='. Fig. 3a shows the distinct N R and N 2 negative components of this evoked potential on trials in which saline or ACh was delivered from the pipette. A marked reduction in N~ amplitude occurred following ACh delivery 15 s prior to tone onset. A one-way ANOVA examining changes in the amplitude of Nj indicated a significant decrease during ACh delivery trials compared to non-delivery trials (F~,~ = 9.13, P < 0.01). The magnitude of the ACh induced decrease in NI amplitude varied for different recording sessions and different animals (range - 6 to -319/zV; mean difference - 126.22 4- 38.98 #V S.E.M.). Alternatively no significant differences in N= amplitude were found comparing saline delivery with non-delivery trials. The differences which did occur following saline delivery were in the opposite direction to those related to ACh (mean difference 50.5 + 47.38 ,¢V; Fig. 3b). A marginally significant difference in N~ amplitude between ACh and saline delivery trials was obtained when evaluated 'across test sessions (Fl,17---8.44, P < (I.(12). Finally, the amplitude of the N a pot0ntial on non-delivery trials was not significantly different for either ACh or saline test sessions. Thus the effects of ACh were limited to delivery trials only and were dissipated before the next (non.delivery) trial occurred. The influence of ACh on the longer latency N 2 component of the OM AEP was the reverse of the effect on N=. N 2 amplitude increased after ACh application relative to non-delivery trials in 8 out of 9 tests (mean increase 127.33 4- 59.26/.~V S.E.M.). The difference was not significant by ANOVA due to the large range of amplitude changes ( - 384-114 p,V), however, when the differences in N 2 amplitude were normali~,.ed relative to percent change between ACh delivery vs non-delivery trials a sighificant increase in N 2 amplitude was observed (t 8 = 2.485, P < 0.037). In comparison to saline delivery trials mean N 2 amplitude was increased by 144/~V on ACh delive.,'y trials; however, this difference is somewhat artificial since saline and ACh trials were administered during different recording sessions. N, amplitude was not significantly changed
on saline delivery vs non-delivery trials (Fig. 3b), nor did it change significantly across sessions on non-delivery trials when either ACh or saline were in the recording pipette (mean difference in N 2 amplitude = 2.98 p.V for ACh non-delivery vs. saline non-delivery trials). Thus ACh significantly and differentially altered the amplitudes of both N! and N 2, the two well characterized components of the OM AEP. The data indicate that the amplitudes of Nj and N 2 were not affected by leakage of ACh from the pipette in that the changes were specific to delivery vs. non-delivery trials. In addition, the lack of behavioral effects across different animals and sessions in which a relatively high concentration of ACh (500 raM) was retained in the recording pipette supports the likelihood that the effects of ACh application were not secondarily associated with alterations in motivational or attentional factors. Although attempts were made to block the effects of ACh on tone evoked potentials with systemic applications of the muscarinic cholinergic antagonists, scopolamine and.atropine, the experiments were confounded by the fact that animals stopped responding after injections of both these drugs at doses which influenced evoked potentials.
DISCUSSION Local application of ACh to physiologically identified regions of the dentate gyrus through the recording pipette significantly altered electrically elicited, perforant path mediated, field potentials as well as identified components of the sensory evoked OM AEP in awake freely moving rats. In no instances were these changes associated with a change in behavioral perfor. mance of the tone discrimination task. The differential influence of ACh delivery on the N t and N 2 components and the lack of effect on saline delivery trials indicate that the effects were not due to the method of application through the recording pipette. These results support and extend previous findings that the Nm and N 2 components result from different neural substrates acting on the dentate granule cells 8-~.1.~ Moreover, the differential change in amplitude of the two components (decreased Nj and increased N 2) after dendritic application of ACh was similar to changes observed in these components during acquisition and retraining of the task and opposite those observed during extinction of the tone discrimination task s-~0 Thus, the activation of an endogenous cholinergic projection to the dentate gyrus could be responsible for
!110 the change in amplitude of these potentials which occurs during establishment of the behavioral significance of the tone stimuli through associative conditiontire. The ACh induced reduction in the amplitude of the Nt component of the OM AEP and the perforant path elicited field EPSP provides further evidence that the Nt component represents activation of pefforant path-granule cell synapses. Depth profile analysis and lesion data in previous studies demonstrated that the site of origin of Nt is in the outer two-thirds of the dentate gyrus molecular layer ~.13.These findings were further substantiated by a recent report which showed that ACh reduces the field EPSP from medial but not lateral perforant path activation in the hippocampal slice preparation ~. The increase in population spike amplitude recorded in the granule cell layer in the current study is similar to other findings showing cholinergic enhancement of the population spike 21,26.3o in the CAI region, but does not agree with earlier reports of a lack of cholinergic influence via microiontophoretic application of ACh on field potentials which did not exhibit population spikes (see ref. 21, Fig. 9) in the dentate gyrus 20,21 It is therefore quite possible that the influence of ACh is dependent on the presence of a small population spike in order for facilitation of the population spike to be observed (i,e, ,rimulation must exceed threshold), The A C h induced facilitation of the population spike and depression of the field EPSP in the CAI region has been attributed to (1) a reduction in the release of inhibitory and excitatory neurotransmitters respectively ,~,4,tv,:,~ and (2) a slow depolarization via inhibi. tion of K* channels through muscarinic receptor mediated inactivation of the M-current zr,,:4,:7, The widespread terminvtion of septal cho!inergie fibers in the dentate gyrus and ~he strategic localization of these terminals on both granule cells and inhibitory interneu. rons 14 supports the concept of widespread cholinergic modulation of granule cell activity 29. Frotscher and Leranth 14 identified choline acetyltransferase positive synapses on the dendrites and cell bodies of granule cells indicating both direct and indirect autions of ACh on granule cells via local circuit processes. The observed increase in the perforant path evoked population spike following ACh delivery in this experiment agrees with prior in vivo studies in which activation of the septal-cholinergic pathway increased the responsiveness of granule cells to perforant path synaptic inputs in anesthetized animals z.n.25. Previous reports from this laborator~j have shown that tone evoked granule cell discharge is significantly increased during presentation of reinforced vs non-reinforced
tones 12 in animals performing a two-tone discrimination task. Lesions of the entorhinal cortex and perforant path eliminate all tone evoked granule cell activity, while lesions of the medio-lateral septum result in non-differentiated discharges to either reinforced or non-reinforced tones ~3. The current results provide further insight into the role of ACh in controlling these learning related changes in granule cell activity, and suggest that cholinergic projections to the dentate gyrus during acquisition of a tone discrimination task are partially responsible for altering the reactivity to sensory stimuli associated during such procedures, it is not insignificant that ACh application was sufficient to increase the septal influence reflected by N 2 as well as simultaneously reduce the influence of sensory stimuli via the perforant path synapses on granule cell dendrites as indicated by the reduction in Nz. Whether or not such reciprocal changes in tone evoked activity result from functional interaction within the local circuitry of the dentate gyrus, or the operation of feedback connections between the dentate gyrus, hippocampus, medial septum and entorhinal cortex, is a question for further investigation. Acknowh,dgements. Figure I was drawn by M, West. This re.';earch was supported by NIDA Grunts DA03502, DA04441 and DA00119 and a gift from RJ Reynolds'tobacco Co, to S,A.D.
REFERBNCES I Alvarez.Leefmun,~, FJ, and Oardm:r.Medwin, A,R,, Influence of the septum on the hlppo~ampal dentate ar~a width are unaccom. panled by field put~ntials,/, PhysioL, 249 (t975) 1~. 16. 2 Andersen, P,, Holmqpist, B. and, Voorhoeve, P,B, Entorhinal activation of dentate granule cells, Acta Phystol Sound, 66:44846l), 1966. 3 Ben-Ari, Y,, Krnjevi~, K,, Reinhardt, W, and Ropert, N,, Intracellular observations on the disinhibitory action of acetylcholine in the hippocampus, Neuroscicnce, 6 (1981) 24?$-2484. 4 Bilkey, D,K, and Goddard, G.V., Medial septal facilitation of hippocampal granule cell actb;b i~, mediated by inhibition of inhibitory interneurons, groin Rcs., 361 (1985) 99-106. $ Blackstad, T,W,, Ultrastructural studies on the hippocampal region, Ping. Brain Rex., 3 (1963) 122-148, 6 Consolo, S,, Wang, G.X., Fusi, R., Vinci, R,, Forloni, G. and Ladinsky, H., In vitro and in vivo evidence for the existence of presynaptic muscarinic eholinergie receptors in the rat hippocampus, Brain Res., 309 (1984) 147-151, 7 Deadwyler, S.A., Biela, J., Rose, G., West, M. and Lynch, G.A., Microdrive for use with glass or metal microelectrodes in recording from freely-moving rats, El~tmencephalol~r, Clin, Neumphysiot,, 4'7 (1979) 752-754, 8 Deadwyler, S,A, West, M, and Lynch, O,, Synapfically identified slow potentials during behavior, Brain Res., 161 (1979) 211-225, 9 Dcadwyler, S,A,, West, M,O, and Robinson, J.H., Entorhinal and
septal inputs differentiallycontrol sensory.evokedresponses in the rat dentate 8Yrus,Science, 211 (1981) !181-1183. i0 Deadwyler, S,A,, West, M.O,, Christian, E,P,. Hampson, R.E, and Foster, T,C,, Sequence-related changes in the sensory-evoked potentials in the dentate gyrus: a mechanism for item-specific short-term information storage in the hippocampus, Behav. Neural Biol,, 44 (1985) 201-212.
101 11 Fantie, B.D and Goddard, G.V., Septal modulation of the population spike in the fascia dentata produced by perforant path stimulation in the rat, Brain Res., 252 (1982) 227-237. 12 Fonnum, F., Glutamate: a neurotransmitter in mammalian brain, J. Neurochem., 42 (1984) i - I I . 13 Foster, T.C, Hampson, R.E., West, M.O. and Deadwyler, S.A., Control of sensory activation of granule cells in the fascia dentata by extrinsic affercnts: septal and entorhinal inputs, J. Neurosci., 8 (1988) 3869-3878. 14 Frolscher, M. and Leranth, C., The cholinergic innervation of ihe rat fascia dentata: identification of target structures on granule cells by combining choline acetyitransferase immunocytochemistry and golgi impregnation, J. Comp. Neurol., 243 (1986) 58-70. 15 Gottlieb, D.I. and Cowan, W.M., On the distribution of axonal terminals containing spheroidal and flattened synaptic vesicles in the hippocampus and dentate g.vrus of the rat and cat, Z. Zellforsch. Mikrosk. Anat., 129 (1972) 413-419. 16 Halliwell, J.V, and Adams, P.R., Voltage-clamp analysis of muscarinic excitation in hippocampal neurons, Brain Res., 250 (1982) 71-92, 17 Hounsgaard, J., Prcsynaptic inhibitory action of acetylcholine in area CAI of the hippocampus, Exp. Neuroi., 62 (19'78) 787-797. 18 Kahle, .I.S. and Cotman, C.W., Carbachol depresse~ synaptic responses in the medial but not the lateral perforant path, Brain Res., 482 (1989) 159-163. 19 Kohler, C., Chan-Palay, V, and Wu, J,Y., Septal neurons containing glutamic acid decarboxyla~ immunoreactivity project to the hippocampal region in the rat brain, Anat. Embryol., 169 (1984) 41-44, 20 Krnjevi~, K., Reiffenstein, R.J. and Ropert, N., Disinhibitory action of acet),lcholine in the rat's hippocampus: Extracellular observations, Nearoscience, 6 ( 1981) 2465-2474.
21 Krnjevi~, K. and Ropert, N., Electrophysiological and pharmacological characteristics of facilitation of hippocampal pepulation spikes by stimulation of the medial septum, Neuroscience, 7 (1982) 2165-2183. 22 Lewis, P.R. and Shute, C.C.D., The cholinergic limbic system: projections to hippocampal, medial cortex, nuclei of the ascending cholinergic reticular system and the subfornical organ and supra-optic crest, Brain, 90 (1967) 521-540. 23 Lomo, T., Patterns of activation in a monosynaptic cortical pathway: the perforant path input to the dentate area o[ the hippocampal formation, Exp. Brain Res., 12 (1971) 18-45. 24 Madison, D.V., Lancaster, B. and Nicoll, R.A., Voltage clamp analysis of cholinergic action in the hippocampus, J. Neurosci., 7 (1987) 733-741. 25 Robinson, G.B. and Racine, RJ., Interactions between septal and entorhinal inputs to the rat dentate gyrus: facilitation effects, Brain Res., 379 (1986) 63-67. 26 Ropert, N. and Krnjevi~, K., Pharmacological characteristics of facilitation of hippocampal population spikes by cholinomimetics, Neuroscience, 7 (1982) 1963-1977. 27 Segal, Mo, Synaptic activation of a cholinergic receptor in rat hippocampus, Brain Res., 452 (1988) 79-86. 28 Stewart, M. and Fox, S.E. Do septal neurones pace the hippocampal theta rhythm? Trends Neurosci., 13 (1990) 163-168. 29 Valentino, R.J. and Dingledine, R., Presynaptic inhibitory effect of acetylcholine in the hippocampus, J. Neurosci., 1 (1981)784= 792. 30 Xie, X,, Inhibition of long-term potentiation of hippocampal CA1 EPSP by a negative cooperativity, Brain Res., 445 (1988) 367-369.
