Neuroscience Vol. 45, No. 3, pp. 653-662, Prmted in Great Britain

1991

0306.4522/9 I $3.00 + 0.00 Pergamon Press plc fi‘ 1991 IBRO

CHOLINERGIC MODULATION OF EXCITABILITY IN THE RAT OLFACTORY BULB: EFFECT OF LOCAL APPLICATION OF CHOLINERGIC AGENTS ON EVOKED FIELD POTENTIALS A. ELAAGOUBY,* N. RAVEL and R. GERVAIS Laboratoire

de Physiologie

Neurosensorielle, Universite Claude Bernard Villeurbanne cedex, France

Lyon I, CNRS UA 180, F-69622

Abstract-The effect of exogenously applied cholinergic agents upon mitral-granule cell complex activity of the olfactory bulb was studied in anesthetized rats. Output neurons were activated by electrical paired-pulse stimulation (4&80ms time interval) applied either to the olfactory nerve (orthodromic stimulation) or to the lateral olfactory tract (antidromic stimulation). Evoked field potentials were recorded in the granule cell layer. Cholinergic agents were introduced close to the mitral cell body layer

through a push-pull cannula. With both orthodromic and antidromic stimulations, acetylcholine in the presence of eserine (an acetylcholinesterase blocker), did not alter the conditioning volley, while it induced a significant increase in the amplitude of the test volley. This effect could be replicated using the cholinergic agonist carbachol. This attenuation of the paired-pulse inhibition is due to a reduction of the dendrodendritic inhibitory action of granule cells upon relay cells. Muscarinic and nicotinic transmission were studied using antidromic and orthodromic stimulations, respectively. The selective effect of acetylcholine on the test volley was totally abolished by the blockade of the muscarinic transmission (by atropine). The blockade of the GABAergic transmission (by picrotoxin), could also prevent the acetylcholine-induced effect. The results lead us to propose that in deep bulbar layers, acetylcholine may activate muscarinic receptors situated on second-order GABAergic interneurons. These interneurons could in turn inhibit granule cells (first-order interneurons). The nicotinic antagonist d-tubocurarine selectively enhanced the duration of the late component and did not appear to modify early components when stimulation was applied to the olfactory nerve. This effect related to both the conditioning and the test volleys and the enhancement in the duration of depolarization of granule cell dendrites suggests that normal activation of nicotinic receptors contributes to a faster repolarization of granule cells. Since nicotinic receptors belong to the outer glomerular layer, this result points to the existence of interneurons belonging to the periglomerular region where they receive nicotinic input and project to deep layers where they modulate granule cell activity. Taken together, our results suggest the presence of a phasic muscarinic and a tonic nicotinic modulation of bulbar interneuronal activity. Since both could finally reduce the inhibitory action of granule cells, the

action of cholinergic afferents would facilitate transmission of bulbar output neurons to central structures.

Acetylcholine (ACh) was one of the first identified neurotransmitters in the CNS. A large body of neurophysiological and behavioral data indicate that the telencephalic cholinergic system is important for modulating expression of neuronal plasticity’,47 and is involved in cognitive processes in humans as well as in animals (see review in Ref. 2). However, the elucidation of the role of ACh in the CNS remains a challenge, although its involvement in cortical information processing has been studied in various sensory areas including visual,40 somatosensory*’ and auditory” primary cortices. The most common finding is that ACh improves the signal to noise ratio and *To whom correspondence should be addressed. Abbreviations: ACh, acetylcholine; EFP, evoked field potential; EPL, external plexiform layer; HDB, horizontal limb of the diagonal band of Broca; IPL, internal plexiform layer; LOT, lateral olfactory tract; LOT-EFP, antidromically evoked field potential; OB, olfactory bulb; ON, olfactory nerve; ON-EFP, orthodromically

evoked field potential. 653

modifies response selectivity through muscarinic rethe physiological ceptors. I9 However, understanding role of ACh requires the characterization of its effect upon central muscarinic and nicotinic receptors with regard to both short- and long-term postsynaptic actions. In this context we investigated the effect of ACh in the rat main olfactory bulb (OB), a structure that offers considerable advantages in studying the function of ACh from the cellular to the behavioral level. As a paleocortex structure, the olfactory bulb is more simply organized than neocortical structures (Fig. 1). Output neurons (mitral and tufted cells), which receive their input directly from the olfactory neuroreceptors, make reciprocal dendrodendritic synapses with first-order interneurons (periglomerular and granule cells). This reciprocal interaction occurs at the level of the unique primary dendrite with the periglomerular cells, and at the level of the wide-spreading, numerous secondary dendrites with granule cells.38 The former contact is established in

A. ELAACOUBY et al.

654

LOT Fig. 1. Photomicrograph of the rat OB with a gold-silver impregnation (right) and schematic representation of the main cellular elements and their connectivity (left). OB layers from top to bottom: GL, glomerular; EPL, external plexiform; MiL, mitral; IPL, internal plexiform; GCL, granular. Olfactory tracts: ON, olfactory nerve; LOT, lateral olfactory tract. Cellular elements: PG, periglomerular cell; Mi, mitral cell; Gr, granule cell; CF, centrifugal fibers,

the outer glomerular layer while the latter occurs in the external plexiform layer. Granule cells and the majority of periglomemlar cells use GABA as a neurotransmitter. In this configuration, output neurons excite the first-order interneurons which in turn inhibit output neurons. The gross organization of pre- and postsynaptic choline@ elements of the OB is fairly well known. The OB does not contain choline&c cell bodies. Choline& elements are exclusively fibers originating from the most rostra1 part of the magnocellular forebrain nuclei and the medial part of the horizontal limb of the diagonal band of Broca (HDB).37,49This region corresponds to the Ch3 cholinergic subset as defined by Mesulam et al. in the rat.*’ In the OB, ACh fibers are concentrated more densely in the glomerular, internal (IPL) and external plexiform (EPL) layers.‘2,25.2*.5’It is likely that they reach bulbar intemeurons, since electron microscopic studies have shown that the centrifugal fibers synapse onto interneurons.”

Autoradiographic and binding studies have revealed two important characteristics of bulbar cholinergic receptors: both nicotinic and muscarinic receptors are present at fairly high densities, and their distribution within the OB is particularly well laminated with a low degree of overlap. In fact, muscarinic receptors are present in the IPL and EPL’4s4’” whereas nicotinic receptor binding is restricted to the glomerular layer.‘3*49 However, the cellular types comprising the receptors for ACh have not yet been identified. Given these neuroanatomical data, we may suppose that ACh modulates the activity of the output neurons by acting on periglomerular and granular cells. At the glomerular layer, ACh could activate nicotinic receptors thus resulting in a modulation of synaptic transmission from the olfactory nerve (ON) to the primary dendrite of the output neurons, whereas in deeper plexiform layers it could mainly solicit muscarinic receptors, and thus modify the inhibitory action of granule cells upon the secondary dendrites of these neurons.

Excitability

in the rat olfactory bulb

However, there have only been two recent studies which have investigated the effect of ACh on bulbar activity. Nickel1 and Shipley26 examined the effect of electrical stimulation of HDB on bulbar-evoked field potential and unit activity while Ravel et a1.” studied the effect of iontophoretic application of ACh on basal activity of identified units. The present study is an attempt to characterize the effect of ACh on bulbar responsiveness and to determine the respective participation of muscarinic and nicotinic mechanisms. In anesthetized rats, OB excitability was appraised using electrical stimulation applied to the ON and to the lateral olfactory tract (LOT), thus activating output neurons either orthodromically or antidromically. Orthodromic stimulation was used to characterize nicotinic modulation in the glomerular layer, while antidromic stimulation was used to study muscarinic modulation in the plexiform layers. Such electrical stimulation elicits well-characterized, reproducible and very stable evoked field potentials (EFPs) in the using

OB.

Cholinergic

a push-pull

agents

were

applied

locally

technique.

EXPERIMENTAL

PROCEDURES

Experiments were carried out on male Wistar rats (IffaCredo; 20&300 g). Animals were anesthetized by an i.p. injection of equithesin (pentobarbital 2.43 g, chloral

655

hydrate 1 I .62 g, ethanol 24.7 ml, propanedio199 ml, MgSO, 5.3 g and distilled H,O for 250 ml). At the beginning of the experiment the injected dose was 0.3 ml per 100 g and during the experiment 0.1 ml was injected each hour. The anesthetized animal was placed in a stereotaxic apparatus, a window was made over the OB and the dura was removed. For stimulation, the skull was perforated either over the ON just caudal to the cribriform plate or over the LOT. The coordinates used were those of Paxinos and Watson.” As illustrated schematically in Fig. 2, the stimulating electrode (a concentric bipolar enamel-coated stainless steel electrode, 0.4 mm diameter, Rhodes Instruments) was positioned according to the following stereotaxic coordinates: 4 mm posterior to the nasofrontal suture for the ON; AP 14.6, Lat 3.4 and DV 6.1 for the LOT. Its final position was adjusted to the site where stimulation produced the highest amplitude EFP. Monopolar recording electrodes were made of enamelcoated silver wires (80 pm diameter, maximum impedance 15 kR). Two recording electrodes were glued to a concentric push-pull cannula and electrode tips protruded from the inner cannula tip by 100pm and 800pm. This recording-perfusing device was first positioned over the OB so that the longer electrode touched the surface of the OB. Then the device was lowered into the OB until the tip of the push cannula was close to the mitral cell body layer as indicated by the inversion point of the EFP collected by the proximal electrode and the maximal amplitude of multiunit activity. In this configuration, the tip of the distal electrode was most often positioned in the granule cell layer as revealed by a characteristic EFP and a low amplitude of multiunit activity. The EEG activity was amplified, filtered (l-3000 Hz bypass band), audio-monitored and visualized on an oscilloscope. Electrical shocks, delivered at a frequency of 0.1 Hz, elicited EFPs which were screened on a second digital oscilloscope. A shock consisted of a O.l-ms

Conditioning

Test

push

stim

pull

II

ON (orlhodromic)

(antidromic)

Fig. 2. Left part. Schematic representation of the OB cellular organization and experimental set-up. Stimulation (stim) electrodes were positioned in the olfactory nerve (ON) and the lateral olfactory tract (LOT). Recording electrodes (ret), glued to a concentric push-pull cannula, protrude 100 pm and 800 pm from inner cannula. Open triangles, output neuron cell bodies; closed circles, cell bodies of periglomerular and granule cell interneurons. Right part. Typical LOT-EFP elicited by paired-pulse stimulation (40 ms inter-pulse time interval), and collected in the granule cell layer (GCL). In all figures each trace represents the mean value of 10 successive sweeps. The stimulation evokes a response on the relay cells (component I) which excite the granule cells through dendrodendritic synapses (excitatory postsynaptic potentials give rise to component II). The test volley evokes a relatively weak response.

A. ELAAGOLJBY et al.

656

LOT-EFP

in the GCL

Signal analysis was performed by software we developed ourselves. Pharmacological effects were studied on EFPs collected from the distal electrode, the tip of which was located in the granule cell layer. Components of EFPs were identified and their respective amplitudes determined and stored. Statistical analysis of data was performed using a paired t-test since each signal was recorded before and after infusion of each drug for a given recording site. RESULTS

Fig. 3. Effect of ACh on bulbar EFPs. Example of EFPs induced by paired-pulse stimulation, in (a) LOT-EFP (40 ms inter-pulse time interval), and in (b) ON-EFP stimulation (8Oms inter-pulse time interval). In this and the following figures the dotted line represents the control EFP, while the solid line represents the EFP obtained after the introduction of the drug. Here, ACh (1 mM) was introduced for 15 min. Note the increase in the amplitude of the test volley. GCL, granular cell layer.

square pulse. The threshold intensity was always found to be around 50 PA. In all experiments the stimulus intensity was set to four times above the threshold. Recordings were obtained by single volley and paired-volley stimulations with time intervals varying from 10 to 80ms. EFPs were digitized by an A/D converter (Bakker Electronics BE489 interfaced with an IBM-AT computer. Each sampling consisted of averaging 10 successive sweeps and the mean value was stored on disk for subsequent analysis. Local drug application was performed using a push-pull technique.j The Krebs’ medium used as a perfusate contained I18 mM NaCl, 4.7mM KCl, 1 mM Na,HPO,, 1.3 mM CaCl,, 1.2 mM MgCl,, 1.2 mM NaHCO,, and 11.Og glucose (PH 7.4), and was oxygenated for several minutes. The solution was pushed in via the inner cannula (0.4 mm diameter) and pulled out through the outer cannula (0.8 mm) by a peristaltic pump. The perfusion was maintained throughout the experiment at a flow rate of 20jtl/min. Neuroactive substances were diluted in the Krebs’ solution a few minutes prior to their use. These included acetylcholine chloride, eserine (physostigmine), d-tubocurarine chloride, nicotine, atropine, carbachol and picrotoxin (purchased from Sigma). Following a period of 45 min of equilibration, control EFPs were &corded, after which the drugs were introduced for 5- to 30-min periods depending on their chemical nature and concentration. The test recordings were made and the drug was then withdrawn while reintroducing the Krebs’ solution. In most cases the drug-induced effect disappeared within 30 to 60 min. This allowed us to study at the same recording site the effect of ACh and after recovery, the effect of ACh in conjunction with another drug (e.g. ACh followed by ACh + atropine; ACh followed by ACh + picrotoxin).

Effect of acetylcholine agonists In the first set of experiments, only ACh (1 mM) was applied to the Krebs’ solution for several tens of minutes. No significant effect was observed either on single volley or on paired pulse EFPs. Hence, eserine (a cholinesterase blocker) was added (1 mM) to the perfusing solution in order to reduce the activity of endogenous cholinesterases. Eserine produced no detectable effect when tested alone. However, when it was applied together with ACh, reproducible effects were obtained. In fact, following the introduction of ACh, with both LOT-EFP and ON-EFP, the conditioning volley of the paired pulse stimulation remained in most cases unchanged, while the amplitude of the test volley - for double shock intervals from 10 to 80ms - increased markedly (Fig. 3). This increase developed progressively after the introduction of the ACh-eserine solution, reaching its maximum within 15 min with 10 mM and within 20 min with 1 mM of ACh. The average value of data from 13 animals obtained on LOT-EFP showed that ACh did not significantly modify the amplitude of component II of the conditioning volley whereas it increased that of the test volley by 22% (P < 0.001) (Table 1). In most cases this increase persisted as long as the drug was applied. In the eight experiments where the reversibility was tested, EFPs returned to the reference value within 15-30 min after withdrawal of the ACh-eserine solution. The ACh-induced effect could be successfully reproduced using the cholinergic agonist carbachol. Even with a much lower concentration (0.05 mM) than that of ACh, and in the absence of eserine, carbachol induced a 21% (n = 4, P < 0.001) increase in the amplitude of the test volley of LOT-EFP (Table 1). After withdrawal of carbachol, recovery was observed within 1 h. Conversely, application of nicotine failed to produce any significant effect even at a relatively high concentration (10 mM) (Table 1). EfSect of acetylcholine antagonists The blockade of the muscarinic receptors by atropine (1 mM) did not significantly modify LOT-EFP (Table 1). However, when atropine was applied together with ACh-eserine solution, or when it was applied before introduction of ACh-eserine solution EFPs remained unchanged, even though the ACh-eserine solution applied in the same recording site 1 h earlier elicited its typical effect (n = 4). Moreover, when atropine was introduced 15 to 20 min

Excitability Table

1. Drug-induced

n ACh + eserine Carbachol Nicotine

variations

in the rat olfactory as percentage

[ClmM Conditioning

13 4 5

1-l 0.05 10

-2+ 11 +9+2 -2+1

Atropine ACh + atropine* d-Tubocurarinet

4 4 6

1 l-l 1

+17+ 11 -1_t12 +47 + 19

Picrotoxin ACh + eserinel

4 4

0.1 l-l

- 12 _+17 +3+3

657

bulb

evoked field potentials amplitude P
._,,

DISCUSSION

L4.0

Methodological considerations The effect on electrically evoked activity in the OB of ACh and related agents introduced through a push-pull cannula was investigated. Mechanical damage was unlikely to be responsible for the observed drug effects for several reasons. First, the characteristics of the EFP remained unchanged during several hours of perfusion with the standard Krebs’ solution. Second, drug effects were consistently seen a few minutes after their application, first at the proximal and later at the distal recording site. Third, drug-induced effects were fully reversible and reproducible. Another point which must be taken into account is the use of anesthetic agents. Indeed, the rats were anesthetized using a solution containing nembutal which is known to enhance paired pulse inhibition in the OB.4’ However, this effect has been reported to occur with doses around 50 mg/kg while here we injected only 9 mg/kg of nembutal. Drug concentrations ranged from 0.1 mM to 10 mM. These doses may seem relatively high, possibly causing non-specific effects. However, several data argue against a non-specific action. Preliminary experiments using tritiated compounds showed that the concentration of drug in the pulled solution reached as high as 90% of that in the pushed solution, thus indicating only a tenth of the applied substance to have reached the tissue. Moreover, as recorded EFPs were collected 800 pm away from the tip of the push cannula, the amount of the drug diffusing through the tissue to reach would have been substantially diluted. Finally, concentrations used in the present study were in the same range as those used in several other experiments carried out using the push-pull technique.3,8 Interpretation of the recorded activity Over the last 30 years, extensive studies have been devoted to the understanding of the generation of EFPs in the vertebrate OB. These include experiments in which the effect of ON and LOT stimulation was examined in different cell layers by recording field potential, extra- and intracellular unit Moreover, two studies have conactivity. 5,23,38.43,45 tributed to an understanding of the phenomenon by means of a theoretical reconstruction of components of LOT-EFP.3’,39 All the authors agree that the EFP is generated by cellular elements oriented perpendicularly to the surface of the OB, that is to say by relay neuron cell bodies and their primary dendrites and by granule cells. Briefly, both LOT- and ON-EFP early components are generated by the propagation of current in the relay cells. The relay cells excite the

10

rn”

m*

Fig. 5. Effect of picrotoxin on LOT-EFP elicited by single pulse stimulation, Picrotoxin (1 mbl) was infusecl for 12min. Note that components I and II were not significantly affected,while a large long-lasting negative deflection invariably appeared at around 30 ms.

granule cell dendrites through reciprocal dendrodendritic synapses. The synchronously generated excitatory postsynaptic potentials form the basis of component II. In fact, components I and II are highly correlated.“6 Therefore, in the present study, only component II was used as an indicator of drug effects. The very late component (component III) has been the least well studied. However, it seems to be generated through depolarization of granule cell bodies either by propagation of the depolarization generated in the EPL or by a feedback excitatory action of centrifugal fibers, 24 During this period granule cells exert a long-lasting GABAergic inhibition (lOO-400 ms) upon the relay neuronsu through dendrodendritic reciprocal synapses with the EPL. Muscarinic and nicotinic modulation of olfactory bulb responsiveness In the present study, local application of ACh agonists (ACh and carbachol) close to the mitral cell body layer induced a 22% increase in the amplitude of the test volley, elicited at 40 ms, without modifying the conditioning volley. This effect, consistently observed in 13 animals, vanished after the withdrawal of exogenously applied agonists. Therefore, the most likely explanation is that the addition of ACh reduced feedback inhibition probably by accelerating the repolarization of granule cells. Since cholinergic agonists were effective only following a first activation of the relay-granule cell complex, this effect was qualified as a phasic one. The effect of ACh was both prevented and reversed by atropine, showing that the attenuation of the paired pulse inhibition involves muscarinic receptors. This finding is in agreement with the data showing that the highest density of muscarinic receptors in the brain is found in the EPL.“s” It is worth mentioning that ACh produced a similar effect upon both LOT-EPP and ON-EFP. This is of importance since it has been shown that stimulation of the ON was more effective in activating tufted cells, while the LOT stimulation preferentially drove mitral cell~.~~Hence, both categories of output neurons could be influenced similarly by muscarinic transmission.

Excitability

in the rat olfactory

The nicotinic antagonist d-tubocurarine, applied around the mitral cell layer, induced an increase in the duration of the late component II of both the conditioning and the test volleys. Since only the component II (generated by excitatory postsynaptic potentials in the granule cell dendrites) was modified, non-specific effects of the drug on general membrane properties are unlikely. Otherwise, all components of the EFP would have been modified. The abrupt appearance of the effect following 20-25 min of perfusion in the core of the structure could be explained by the time required for diffusion of the drug from deep cell layers to the outer glomerular layer where the nicotinic receptors are located. Since the antagonist was effective by itself, this suggests the existence of a sustained endogenous release of ACh in the glomerular layer. This interpretation is in accordance with anatomical and electrophysiological data.33.51 A high level of endogenous ACh input is consistent with the fact that the experimentally applied nicotine produced no detectable effect, at least for the low concentration reaching the glomerular layer in this study. The effect produced by d-tubocurarine on ON-EFP did not appear to affect synaptic transmission from the ON to output neurons, as the early component was not modified. Knowing that extra- as well as intraglomerular spaces are innervated by ACh fibers,” the absence of an effect on the early component may well have been due to the inability of d-tubocurarine to penetrate the glial sheet surrounding the glomerulus. Thus, the observed effect could have resulted from a specific blockade of extraglomerular nicotinic receptors situated on neuronal elements which could influence granule cell dendrite excitability. To our knowledge, this observation represents the first electrophysiological hint of an action of some periglomerular region elements upon granule cells. Such cholinoceptive cells should have cell bodies or extensive arborization in the periglomerular region, where the nicotinic receptors are situated, and projections synapsing on granule cell dendrites in the EPL. However, among the three types of bulbar interneurons of the periglomerular region described in Golgi studies,35 superficial short axon cells appear to be the best candidates for mediating the observed effect. This interpretation is in accordance with the conclusion of Nickel1 and Shipley” in their attempt to identify juxtaglomerular, cholinesterase-positive cells. Nevertheless, difficulty in determining the shape of the neurons they observed precluded the clear-cut identification of cellular type. Does acetylcholine act directly or indirectly on granule cellsr Our results suggest that in deep cell layers, ACh and carbachol depress the activity of the granule cells following their excitation by relay neurons. At least two mechanisms might be involved. ACh might act directly on granule cells or indirectly by exciting some GABAergic second-order interneurons which would

bulb

659

inhibit granule cells. With respect to these two types of interneurons, ACh could either act on the soma or close to the synaptic contact with target cells (pre-synaptic modulation of transmitter released by interneurons). Lacking an electron microscopic description of the cholinergic innervation of the OB, we must rely on other types of data. Histochemical studies,*’ combined with electrophysiological data,16 strongly suggest that ACh fibers act mainly on second-order interneurons, which in turn control the activity of granular and periglomerular interneurons. In deep cell layers, these second-order cholinoceptive interneurons are referred to as “inframitral interneurons”. Since the results suggested ACh depresses granule cell activity, we tested the hypothesis of whether the “inframitral interneurons” were responsible for such inhibition. These interneurons, like the majority in the OB, may use GABA as a transmitter. Thus, introduction of ACh was preceded by an infusion of picrotoxin. Although picrotoxin did not affect the EFP, the addition of ACh was without effect. Therefore, we propose that the action of second-order cholinoceptive interneurons on granule cells is inhibitory. Although Nickel1 and Shipley have concluded that the second-order interneuron is possibly excitatory2S,26 at least three explanations for this discrepancy can be given. First, whereas here ACh was infused locally, Nickel1 and Shipley used electrical activation of the HDB. Among the neurons projecting to the OB from the region, only 15% are cholinergic, and some also contain gaianin which is known to act in opposition to ACh.” The rest of the population use GABA,So other peptides3’ or still unidentified neurotransmitters. Thus, the specificity of the induced effects cannot be compared. Second, whereas electrical stimulation of fibers would cause neurotransmitter release at the synaptic level, exogenously applied ACh solicits extra- as well as intrasynaptic receptors. This could result in opposite effects as observed for dopaminergic transmission in the striatum.48 Third, Nickel1 and Shipley found that the efficacy of inducing an effect depends on the frequency of stimulation. This parameter was not taken into account in the present study though it is important for neuronal functioning.’ Further experiments are required to clarify this discrepancy. GABAergic modulation of granule cell acticity When applied either at a relatively high concentration (1 mM) or for more than 10 min, picrotoxin modified both the spontaneous and evoked activity of granule cells. In the absence of any stimulation, several large negative deflections occurring about once every second, reflecting spontaneous synchronous depolarization of granule cell bodies, were seen.” Therefore, granule cells must receive a tonic GABAergic input preventing them from excessive and unusually long-lasting depolarization.4 Whether the GABA involved in this modulation is released

660

A. ELAAGOLJBY et al.

from intemeurons or from centrifugal fibers is unknown. Following LOT stimulation, the presence of picrotoxin (0.1 mM) induced no variation in components I and II of the EFP. However, a large negative deflection was Seen to develop in component III, starting at around 30 ms after the offset of stimulation (Fig. 5). This extra component, which reflects depolarization of granule cell bodies, is most likely due to the action of centrifugal fibers. In absence of the tonic GABAergic control, excitatory centrifugal inputs will not be counter-balanced.

ON

GL

BPL

Hi IPL

ACh

Functional implications

The present paper is the first investigation of the dual muscarinionicotinic modulation of OB excitability and Fig. 6 summarizes the main findings. The observed effects of muscarinic and nicotinic activation converge on granule cells where both contribute to a reduction in granule cell activity. Consequently, the action of centrifugal ACh fibers reaching the OB would lead to a reduction in inhibitory action exerted upon relay neurons. This is in accordance with a recent study showing that a brief multi-pulse stimulation applied to HDB inhibits granule cells and activates mitral cell unit activity.15 Nevertheless, the chemical nature of fibers mediating this effect has not been characterized. As a consequence of the hyperpolarization of granule cells, ACh modulation could affect at least two types of bulbar electrical activity. First, it could reduce the rate of spiking granule cell~~~*~~ thus resulting in the output neurons being more responsive in the presence of sensory stimulation when ACh is applied. This would increase the signal to noise ratio of OB output in a similar way to other sensory cortices.2’s40Second, it might affect the bulbar oscillatory activity in 7-8 Hz spectral frequency range which is generated by the mitral-granule cell complex.6 Such rhythmical activity is driven in the hippocampus through the septo-hippocampal cholinergic pathways.27 As the cholinergic cell bodies projecting to the OB and to the hippocampus, though distinct, are in relatively close proximity in the basal forebrain choline& complex, both ACh subsystems might contribute to the expression of the theta activity in their respective cortical targets. This is supported by the observation that sniffing and hippocampal theta rhythms became time-locked, particularly during sampling of olfactory information being learned by a rat solving a task. I* However, it can be argued that in the OB, the theta rhythm is mainly driven by

CC

Fig. 6. Schematic representation of neuronal mechanisms that could support cholinergic modulation of OB activity. The underlined keywords summarize the findings from the present study. These are: (i) the excitatory nicotinic action in the periglomerular region, likely on second-order interneurons (unidentified cell, UC); (ii) projections of UC to EPL inhibit granule cell dendrites (Gr); (iii) the nicotinic modulation is expressed tonically; (iv) the excitatory muscarinic action in the IPL and in the granule cell layer (GCL), likely on second-order GABAergic interneurons, inframitral cells (IMC); (v) the muscarinic modulation is expressed phasically. Both nicotinic and muscarinic modulations could contribute to a reduction in the inhibitory action exerted by Gr on output neurons (disinhibition).

sniffing, which in the rat is also a 7-8 Hz phenomenon. Nevertheless, it has been shown that the temporal distribution of OB unit activity within the respiratory cycle depends on both peripheral and the centrifugal influences. 32 Focusing attention on this slow activity in the olfactory pathways is of interest since it could be linked to processes associated with olfactory learning. In fact, this rhythm is likely of physiological relevance, as stimulation at the theta frequency is optimal for induction of long-term potentiation in the hippocampus.16~2g Thus, it may be supposed that choline& as well as noradrenergic action71” is of prime importance for the support of forms of OB plasticity following olfactory learning such as those observed in rat pups4* as well as in adult rabbits.‘O Acknowledgements-We wish to thank Prof. A. Holley and Dr G. Chouvet for the helpful comments and discussion, and Dr U. Kindermann and Dr R. Hudson for the careful review of the manuscript. The photomicrograph and the schema in Fig. 1 were kindly provided by Prof. F. Jourdan and Prof. A. Holley, respectively.

REFERENCES 1. Bear M. F. and Singer W. (1986) Modulation of visual cortical plasticity by acetylcholine and noradrenaline. Nature 3u), 172-175. 2. Collerton D. (1986) Cholinergic function and intellectual decline in Alxheimer’s disease. Neuroscience 19, I-28. 3. Errington M. L., Lynch M. A. and Bliss T. V. P. (1987) Long-term potentiation in the dentate gyrus: induction and increased glutamate release are bIocked by D( - )aminophosphonovaJerate. Neuroscience Zo, 279284. 4. Felix D. and MeLennan H. (1971) The effect of bicuculline on the inhibition of mitral cells of the olfactory bulb. Brain Res. 25, 661464.

Excitability

in the rat olfactory

bulb

661

5. Freeman W. J. (1972) Depth recording of averaged evoked potential of olfactory bulb. J. Neurophysiol. 35, 780-796. 6. Freeman W. J. (1974) Average transmission distance from mitral-tufted to granule cells in the olfactory bulb. EEG C/in. Neurophysiol. 36, 6094 18. 7. Gervais R., Holley A. and Keverne B. (1988) The importance of central noradrenergic influences on the olfactory bulb in the processing of learned olfactory cues. Chem. Senses 13, 3-12. 8. Goldfinger M. D., Simpson C. W. and Resch G. D. (1984) Recovery by push-pull perfusion of neurochemicals released within the cuneate nucleus of the cat by somatosensory stimulation. Pharmac. Biochem. Behav. 21, 117-123. 9. Gonon F. (1988) Non-linear relationship between impulse flow and dopamine released by rat midbrain dopaminergic neurons as studied by in vivo electrochemistry. Neuroscience 24, 19-28. 10. Gray C., Freeman W. J. and Skinner J. E. (1986) Chemical dependencies of learning in the rabbit olfactory bulb: acquisition of the transient spatial-pattern change depends on norepinephrine. Behav. Neurosci. 100, 585-596. 1 I. Gray C. M., Freeman W. J. and Skinner J. E. (1987) Induction and maintenance of epileptiform activity in the olfactory bulb depends on centrifugal input. Expl Brain Res. 68, 210-212. 12. Halasz H. and Shepherd G. M. (1983) Neurochemistry of the vertebrate olfactory bulb. Neuroscience 10, 579619. 13. Hunt S. and Schmidt J. (1978) Some observations on the binding patterns of a-bungarotoxin in the central nervous system of the rat. Brain Res. 157, 213-232. 14. Kobayashi R. M., Palkovits M., Hruska R. E., Rothschild R. and Yamamura H. I. (1978) Regional distribution of muscarinic cholinergic receptors in rat brain. Bruin Res. 154, 13-23. 15. Kunze W. A. A., Shafton A. D., Kemm R. E. and McKenzie J. S. (1991) Effect of stimulating the nucleus of the horizontal limb of the diagonal band on single unit activity in the olfactory bulb. Neuroscience 40, 21-27. 16. Larson J., Wong D. and Lynch G. (1986) Patterned stimulation of the theta frequency is optimal for induction of hippocampal long-term potentiation. Brain Res. 368, 347-350. 17. Levy F., Gervais R., Kinderman U., Orgeur P. and Piketty V. (1990) Importance of p-noradrenergic receptors in the olfactory bulb of sheep for recognition of lambs. Behnv. Neurosci. 104, 464469. 18. Macrides F., Eichenbaum H. B. and Forbes W. B. (1982) Temporal relationship between sniffing and the limbic 0 rhythm during odor discrimination reversal learning. J. Neurosci. 2, 1705-l 717. 19. McKenna T. M., Ashe J. H., Hui G. K. and Weinberger N. M. (1988) Muscarinic agonists modulate spontaneous and evoked unit discharge in auditory cortex of cat. Synapse 2, 5468. 20. Mesulam M. M., Mufson E. J., Wainer B. H. and Levey A. I. (1983) Central cholinergic pathways in the rat: an overview based on an alternative nomenclature (ChlCh6). Neuroscience 10, 1185~1201. 21. Metherate R., Trembley N. and Dykes R. W. (1988) The effects of acetylcholine on response properties of cat somatosensory cortical neurons. J. Neurophysiol. 59, 1231.-1252. 22. Mori K., Nowycky M. and Shepherd G. (1981) Analysis of long duration inhibitory potential in mitral cells in the isolated turtle olfactory bulb. J. Physiol., Lond. 314, 31 I-320. 23. Mori K. (1987) Membrane and synaptic properties of identified neurons in the olfactory bulb. Prog. Neurohiol. 29, 275-320. 24. Nakashima M., Mori K. and Takagi S. F. (1978) Centrifugal influence on olfactory bulb activity in the rabbit. Brain Res. 154, 301-316. 25. Nickel1 W. T. and Shipley M. T. S. (1988) Two anatomically specific classes of candidate cholinoceptive neurons in the olfactory bulb. J. Neurosci. 8, 44824491. 26. Nickel1 W. T. and Shipley M. T. (1988) Neurophysiology of the magnocellular forebrain inputs to the olfactory bulb in the rat: frequency potentiation of output neurons. J. Neurosci. 8, 44924502. 27. Nicoll R. A. (1985) The septo-hippocampal projection: a model cholinergic pathway. Trends Neurosci. 12, 533-536. 28. Ojima H., Yamasaki T., Kojima H. and Akashi A. (1988) Cholinergic innervation of the main and the accessory olfactory bulbs of the rat as revealed by a monoclonal antibody against choline acetyltransferase. Anat. Embryol. 178, 481488. 29. Pavlides C., Greenstein Y. J., Grudman M. and Winson J. (1988) Long-term potentiation in the dentate gyrus is induced preferentially on the positive phase of O-rhythm. Brain Res. 439, 383-387. 30. Paxinos G. and Watson C. (1982) The Rat Brain in Sfereotaxic Coordinates. Harcourt Brace Jovanovich, London. 31. Rail W. and Shepherd G. M. (1968) Theoretical reconstruction of field potentials and dendrodendritic synaptic interactions in the olfactory bulb. J. Neurophysiol. 31, 884-915. 32. Ravel N., Caille D. and Pager J. (1987) A centrifugal respiratory modulation of olfactory bulb unit activity: a study on acute rat preparation. Expl Brain Res. 65, 623428. 33. Ravel N., Akaoka H., Gervais R. and Chouvet G. (1990) The effect of acetylcholine on rat olfactory bulb unit activity. Brain Res. Bull. 24, l-5. 34. Rotter A., Birdsall N. J. M., Burgen A. S. V., Field P. M., Hulme E. C. and Raisman G. (1979) Muscarinic receptors in the central nervous system of the rat. I-Technique for autoradiographic localization of the binding of [‘Hlpropylbenzilcholine mustard and its distribution in the forebrain. Brain Res. Rev. 1, 141-165. 35. Schneider S. P. and Macrides F. (1978) Laminar distributions of interneurons in the main olfactory bulb of the adult hamster. Brain Res. Bull. 3, 73-82. 36. Schneider S. P. and Scott J. W. (1983) Orthodromic response properties of rat olfactory bulb mitral and tufted cells correlate with their projection patterns. J. Neurophysiol. SO, 358-378. 37. Senut M. C., Menetrey D. and Lamour Y. (1989) Cholinergic and peptide projections from the medial septum and the nucleus of the diagonal band of Broca to dorsal hippocampus, cingulate cortex and olfactory bulb: a combined wheatgerm agglutinin-apohorseradish peroxidase-gold immundhistochemical study. Neuroscience 30, 385403. 38. Shepherd G. M. (1972) Synaptic organization of the mammalian olfactorv bulb. Phvsiol. Rev. 52. 864912. 39. Shepherd G. M. and Brayton R. KT (1979) Computer simulation of a dendrodendritic synaptic circuit for self- and lateral inhibition in the olfactory bulb. Bruin Res. 175, 3777382. 40. Silito A. M. and Kemp J. A. (1983) Cholinergic modulation of the functional organization of the cat visual cortex. Brain Res. 289, 1433155. 41. Stewart W. B. and Scott J. W. (1976) Anesthetic-dependent field potential interactions in the olfactory bulb. Bruin Res. 103, 487499.

A.

662

ELAAGOUBY et al

42. Sullivan R. M., Wilson D. A. and Leon M. (1989) Norepinephrine and learning-induced plasticity in infant rat olfactory system. J. Neurosci. 9, 39984006. 43. Waldow U., Nowycky M. C. and Shepherd G. M. (1981) Evoked potential and single unit responses to olfactory nerve volleys in the isolated turtle olfactory bulb. Brain Res. 211, 267-283. 44. Wamsley J. K., Zarbin M. A. and Kuhar M. J. (1984) Distribution of muscarinic chohnergic high and low affinity agonist binding sites. A light microscopic autoradiographic study. Brain Res. Bull. 12, 233-243. 45. Wellis D. P. and Scott J. W. (1990) Intracellular responses of identified rat olfactory bulb interneurons to electrical and odor stimulation. J. Newophysiol. 64, 932-947. 46. Westecker M. E. (1970) Excitatory and inhibitory interactions in the olfactory bulb involving dendrodendritic synapses between mitral cells and granular cehs. Pftigers Arch. 317, 173-186. 47. Williams S. and Johnston D. (1988) Muscarinic depression of long-term potentiation in CA3 hippocampal neurons. Science 242, 84-87. 48. Williams G. and Millar J. (1990) Differential actions of endogenous and iontophoretic dopamine in the rat striatum. Eur. J. Neurosci.

2, 658-661.

49. Woolf N. J., Eckenstein F. and Butcher L. L. (1984) Cholinergic systems in the rat brain: I-Projections to the limbic telenc-ephalon. Brain Res. 13, 751-784. 50. Zaborszky L., Carlsen J., Braahear H. R. and Heimer L. (1986) Cholinergic and GABAergjc atTerents to the olfactory bulb in the rat with special emphasis to the projection neurons in the nucleus of the horizontal limb of the diagonal band. J. camp. Neural. 243, 817-906. 51. Zheng L. M., Ravel N. and Jourdan F. (1987) Topography of centrifugal acetylchohnesterase-positive fibers in the olfactory bulb of the rat: evidence for original projections in atypical glomeruli. Neuroscience 24, 567-578. (Accepted

5 June 1991)