Actions of acetylcholinesterase in the guinea-pig cerebellar cortex in vitro

0306-4522/92 s5.00 + 0.00 Pcrgamon Press plc 0 1992 IBRO

Newo~ience Vol. 47, No. 2, pp. 291-301, 1992 Printed in Great Britain

ACTIONS OF ACETYLCHOLINESTERASE IN THE GUINEA-PIG CEREBELLAR CORTEX IN VITRO M. APPLEYARD* and H. JAHNSEN~ Institute of Neurophysiology,

Blegdamsvej 3c, DK-2200 Copenhagen, Denmark

Abatraet-Acetylcholinesterase is released in a calcium-dependent manner when afferents of the cerebellar cortex are stimulated. Since cholinergic transmission is probably insignificant in the cerebellar cortex, the ester= itself might serve as a transmitter or modulator. Therefore, the effect of acetylcholinesterase in the cerebellum was investigated in slices of guinea-pig cerebella during intracellular recording from Purkinje cell somata or dendrites. Addition of acetylcholinesterase (20 U/ml) to the superfusion medium did not change the membrane potential or the input resistance of the Purkinje cells. Thus, ester= does not act like a classical transmitter. The threshold for Na+ spikes generated by intracellular current injection was unaffected, but the threshold for Ca2+ spikes was increased. This increase was abolished by tetrodotoxin (1 PM). Furthermore, when Ca2+ currents were blocked by substituting Mn*+ for Ca2+ (2mM) a decrease in a Na+ plateau potential was seen in the presence of esterase. The effect of acetylcholinesterase on Ca2+ spikes is therefore most likely due to a reduction of the non-inactivating Na+ current of the Purkinje cell membrane. When present this current contributes to activation of Ca2+ spikes in dendrites. Acetylcholinesterase also enhanced the response of Purkinje cells to the excitatory amino acids glutamate and aspartate thought to be transmitters in the cerebellar cortex. The responses tecame larger and faster in the presence of esterase. Responses to climbing fibre stimulation were also enhanced

by acetylcholinesterase. The late part of this synaptic response was increased. The potentiation by esterase of responses of Purkinje cells to excitatory amino acids and to climbing fibrc stimulation may be mediated through interference with transmitter uptake, because it was prevented by treatment with DL-2-amino-4phosphonobutyric acid (0.5 mM) and di-hydrokainate (0.1 mM). None of the effects of esterase was due to hydrolysis of acetylcholine because irreversible inhibition of the catalytic site of the enzyme with soman did not prevent the actions. The observations were specific for acetylcholinesterase. Butyrylcholinesterase (20-40 U/ml) showed none of the effects. It is concluded that acetylcholinesterase in the cerebellar cortex seems to mediate a novel type of modulation by two separate mechanisms. Esterase reduces the tendency towards Ca2+ spike gem&ion in Purkinje cells. Ca2+ spikes are followed bv afierhvoeroolarizations and in their absence. firina of Na+ spikes at-higher freque&ies is possible. Secbndly, t&e-is an enhancement of the action of eicitatory transmitters so that the extended operating range can be utilized.

It is generally accepted that acetylcholinesterase (EC 3.1.1.7, AChE) plays a role in cholinergic transmission by hydrolysing the transmitter acetylcholine (ACh), thus terminating its action. However, the distribution of AChE within the brain is not restricted to cholinergic systems,B*35suggesting that the enzyme may have additional functions. For instance, in the substantia nigra AChE in a soluble form (mol. wt 80,000)8 is present in non-cholinergic,2g dopaminergic’ neurons. Evidence from both in vitro and in vivo studies shows that AChE is secreted from the dendrites and terminals of these nigrostriatal neurons.1’J3,“‘*36 Intracellular recording studies of nigral neurons have demonstrated that AChE induces a long-lasting hyperpolarization with a *Present address: Department of Physiology, Royal Free Hospital School bf Medicine, I\owla% Hili Street, London NW3 2PF, U.K. tTo whom correspondence should be addressed. Abbreviations: ACh, acctylcholine; AChE, acetylcholinesterase; APB, DL-2-amino4qhosphonobutyric acid; BChE, butyrylcholineaterase; CFR; climbing fibre response; ChAT, choline acetyltransferse; DHK, dihydrokainate; TTX, tetrodotoxin.

concurrent fall in input resistance, an action that is independent of the hydrolysis of ACh.lg*” A similar situation may exist in the cerebellar cortex where levels of AChE are high, but there is very little ACh and choline acetyltransferase (ChAT), the synthesizing enzyn~?~~ (Appleyard, unpublished observations). ChAT staining’* and in vivo electrophysiological recording9 have provided no convincing evidence for cholinergic transmission in the cerebellar cortex, and in an in vitro investigation it was concluded that it is unlikely that ACh is a transmitter of the major afferent systems of the cerebellar corte~.‘~ In high doses, however, ACh can produce small depolarizations with an increase in membrane resistance in Purkinje cell~.‘~ In vivo studies using intracerebellar push-pull cannulae have demonstrated Ca*+ -dependent secretion of AChE from guinea-pig cerebellar cortex.4 One of the principal sources of AChE within the guinea-pig cerebellar cortex is the climbing fibre system27~35which originates in the inferior olive and projects onto cerebellar Purkinje cells.’ Purkinje cells respond to this input with low-frequency regular discharge composed of complex spikes3’ Systemic

291

292

M.

APPLEYAW

administration of harmaline induces an oscillatory activity in olivary cells which produces a regular discharge in Purkinje cells and a characteristic tremor.28,32 Such treatment also produces a marked increase in the secretion of AChE from the cerebellar cortex, with a similar time-course to the tremor produced.4 This suggests that AChE is secreted from the climbing fibres and, therefore, could modulate the activity of target neurons. In an attempt to investigate possible modulatory functions of secreted AChE within the cerebellar cortex, we have investigated the effects of exogenously applied AChE upon the membrane properties of guinea-pig Purkinje cells recorded intracellularly in oitro. The effects of AChE upon the response of Purkinje cells to climbing fibre stimulation, and to iontophoretic application of putative excitatory amino acid transmitters endogenous to the cerebellar cortex have also been investigated. In order to determine whether any observed effects of AChE were related to the hydrolysis of ACh, some experiments were performed with AChE that had been pre-treated with an irreversible inhibitor (soman). Butyrylcholinesterase was also examined for similar properties. Preliminary reports of some of these findings have previously been published as abstracts.2.23 EXPERIMENTAL

PROCEDURES

Cerebellar slices

Cerebellar slices were obtained from albino guinea-pigs (Statens Seruminstitut) of either sex weighing 250-450g. The method for obtaining slices has been published elsewhere.% Briefly, the animal was anaesthetized with ether and killed with a blow from a metal rod to the thorax. The brain was removed and the cerebellar vermis cut out and placed in a Vibratome. Sagittal slices with a thickness of 350-400 nrn were cut. At the beginning of a recording session one slice was transferred to an interface-type recording chamber with a volume of 1.5 ml and a temperature of 365°C which was constantly perfused with pre-heated, oxygenated Ringer at a rate of 1.5 ml/min by means of a peristattic pump. The slice was illuminated with a cold light source from above and viewed using a stereo-microscope. which allowed the Purkinje cell layer, molecular layer and granule cell layer to be easily discerned. The recording electrode was thus positioned in the Purkinje cell layer under visual guidance; cells were found blindly by moving down through the tissue with the electrode in steps of 2 pm. It was possible to identify whether the electrodes impaled cell soma or dendrites by observing the relative heights of the Na+ and Ca*+ spikes, somatic recordings having relatively large Nat spikes and small Ca2+ spikes whilst the reverse was true for dendritic recordings.Z1~‘0~3’ Recording system

Electrodes of resistance SO-130 MR were pulled from filamented glass tubes and tilled with a 3 M solution of potassium acetate. Recordings were performed with a conventional bridge-balance amplifier with a cut-off fnzcurency of 8 kHz. The-recorded signals were stored on a four-track Racal tape recorder with a band width of O-S kHz. The final display of records involved digitizing the signals on a computer at 5-100 kHz.

and

H.

JAHNSEN

Silent cells or cells showing no spontaneous activity with a hyperpolarizing constant current of less than I nA injected through the recording electrode were accepted into the data base. In cases where this bias current was needed it was kept constant throughout the experiment. The input resistance of the cell was determined using small amplitude hyperpolarizing current pulses injected through the recording electrode. The membrane potential was determined by measuring the change in potential seen when the electrode was pulled out of the cell or when the cell was accidentally lost. In most experiments the Na+ spike peak was used to check whether there was a significant drift in the reference potential of the recording system. The drift was found to be insignificant. When AChE was introduced in the chamber Na+ and CaL’ spike thresholds were measured at IO-min intervals in one of two ways: either by changing the bias current manually in a depolarizing direction or by injection of a single slow ramp depolarizing current. In some cells, where the electrode characteristics permitted it, Lissajous figures were obtained by injection of sequences of depolarizing and hyperpolarizing ramp currents and plotting the resulting voltage against the injected current. Data for plots of current vs frequency characteristics were obtained by injecting depolarizing current pulses lasting 2 s at 10-s intervals. Climbing fibre responses were elicited by stimulating the white matter with a tungsten microelectrode at the base of the folium containing the recorded cell. Glutamate or aspartate were applied onto the cell through an iontophoresis electrode positioned in the molecular layer 0.2-I 5 mm from the recording electrode, and at a similar depth in the slice. Stimulation by either climbing fibre activation or by iontophoretic application of amino acids was continued throughout the experiment at a rate of 0.1 or 0.2 Hz. At the end of each recording the extracellular potential was measured, after removing the electrode from the cell. Any stimulus artefact was also measured. If AChE had been applied during the course of the experiment the next recording was performed in a new slice, after thoroughly washing the recording chamber with standard Ringer solution. Solulions

The standard Ringer solution contained (in mM): NaCl, 130; KCl, 2; NaHC03, 20; KH,PO, , 1.25; CaCl, , 2; MgCl, , 2 and glucose, 4. The preparation of slices was performed in modified Ringer solution containing (in mM): NaCI, 120; KCl, 2; NaHCO,, 20; KH,PO,, 1.25; CaCI,, 2; MgCl,, 2; glucose, 4 and HEPES, 10 (pH 7.2). When CaCl, was substituted with MnCI,, KH,PG., was replaced with KC1 on an equimolar basis. The solutions were bubbled with 95% 0,/5% CO,, and this gas mixture also filled the space above the medium in the recording chamber. Tetrodotoxin (‘FIX, Sigma), DL-2-amino-4-phosphonobutyric acid (APB, Cambridge Research Biochemicals) and dihydrokainate (DHK, Cambridge Research Biochemical@ were added to the medium immediately before it was used for superfusion. Cholinesrerase preparations

After obtaining stable recordings from Purkinje cells, the following choline&erase preparations, made up in oxygenated standard Ringer, were infused into the recording chamber for a period of 30 min by means of a continuously perfused two-way switch: 20 U/ml AChE (Sigma, Electric Eel type V-S, approx. 1400 U/mg protein); 4OU/ml BChE (Sigma Horse Serum, 500-1000 U/mg protein); AChE inactivated by pretreatment with soman (soman-AChE was a gift from Dr Mary French) so that it could no longer hydrolyse acetylcholine.‘9

A~tyl~o~nes~r~ RESULTS

I~enti~~ation and typical features of celfs Stable intracellular recordings were obtained from a total of 112 cells situated in the Purkinje cell layer. Of these 112 cells 51 were lost accidentally at various stages during the experiments and they were therefore excluded from the final analysis. The 61 cells were found in slices from 49 animals. All cells displayed the electroresponsive properties characteristic of Purkinje cells, as previously described.soJ’ Input resistance, determined by injection of square current hyperpolarixing pulses, ranged from 13 MR to 36 MQ for somatic recordings and from 8.6MZZ to 23.6 Mlf for dendritic recordings. The majority of recordings (51 cells) were from the somatic region since the fast Na+ spikes, which originate in the soma, were of large amplitude (3 30mV) compared to the slow Ca*+ spikes which originate in the dendrites.)O Dendritic recordings (10 ceils) displayed smaller amplitude Na+ spikes and large amplitude (> 20 mV) Ca2+ spikes (Fig. 1).

A

293

in the cerebellar cortex

Several cells displaying these properties were intra~ll~~ly filled with Lucifer Yellow; subsequent histological examination showed that they exhibited typical Purkinje cell morphology. Eflects of acetylcho~i~ster~e properties of Purkinje cells

on the basic mem~r~e

Ringer containing AChE (20 U/ml) was introduced into the recording chamber during recordings from 46 of the 61 cells, after ensuring that stable responses were obtained for at least 20 min. In contrast to observations made in the substantia nigra19 there were no significant changes in the membrane potential or input resistance of the recorded cells, either during the 30-min period of AChE application, or in the subsequent washout period. The membrane potential before AChE averaged -67.4 rt 8.6 mV (S.D.) and -66.5 + 8.5 mV after AChE in 26 somatic recordings. The input resistance was 20.5 & 5.6 MQ before AChE and 21.4 & 6.4 MQ after AChE in 17 somata. In six dendrites the input resistance was 17.0f 5.1 MQ before AChE and 17.1 + 5.3 MIZ after. There was no obvious effect of AChE upon the

B AChE

control

20 min

I

20mV

2.89nA

3.15nA

O.JnA I

Fig. 1. Method of measuring the current threshold for Ca2+ spikes. A and B are from a dendrite. In the control solution the threshold was 2.89 nA and after 20 min in 20 U/ml AChE it increased to 3.15 nA. The asterisk marks the onset of Na+ spike firing. Note that the apparent decline in voltage during current injection in A is probably an artifact due to electrode rectification. C-E show that TTX abolished the effect of ACbE on the Caz+ spike threshold. C is in control medium. Na+ spike firing begins at the asterisk. At a larger current injection, Caz* spikes are produad. (D) TTX (1 @i) caused an increase in Ca*+ spike threshold. (E) After TIX treatment AC&E did not increase Ca2+ spike threshold. The dashdot line shows the kvel of current needed for Ca2+ spike generation before TIX. and the dotted line is the level after TTX and in TTX+ACbE.

M. APPLEYARD

294

and H. JAHNSEN found in the Purkinje cell membrane.~~ When this conductance is blocked, a depolarizing current injected through the recording electrode produces less depolarization of the membrane and thus the current threshold for Ca’” spikes is increased. This is true even when the electrode is placed in a dendrite. As seen in the dendritic recordings in Fig. lA, B and C, the Na’ spiking begins well before the Cal* spiking and since the non-inactivating Na+ current is activated at the same voltage levei as the Na+ spikes this current must also have contributed to the depolarization leading to Cal+ spiking. AChE may increase the current threshold for Ca2+ spikes, as TTX does, by reducing the non-inactivating Na+ conductance. This hypothesis is corroborated by the observations shown in Fig. 3. The steady-state current vs firing frequency plot obtained before and after AChE was added to the medium revealed that the firing frequency was reduced when currents nearly sufficient for Ca2+ spike generation were injected (Fig. 3A). However, when even larger currents were used, the cell could fire at frequencies not attainable in the control solution since in the presence of AChE the Nat spikes were not interrupted by Ca” spikes. In order to study the effect of AChE on Na-’ currents without interference from Ca2+ currents the experiments of Fig. 3B-D were performed. They all involved blockade of Ca*+ (and Ca2+-dependent K * ) currents by substitution of Ca?+ with Mn”‘. In the current vs voltage plot of Fig. 3B the reduction in the Nat plateau potential by AChE is seen as a difference between the control plot and the AChE plot just below the firing threshold which was about -45 mV in this cell. The highest point in each plot represents the average inter-spike voltage level during spiking at the lowest possible frequency. Figure 3C is a direct demonstration that blockade of Ca*+ currents by substitution of Ca2+ by MnZi does enhance the size of the Na’ plateau potential in Purkinje ceils because even though the Ca” component is no longer present, the concomitant

response to small amplitude depolarizing current pulses, no change in the threshold level for Na+ spike firing, the shape of the Na+ spikes, or in the afterdepolarization (not shown). We obtained Lissajous figures for three cells treated with AChE in normal medium using double ramp currents with a cycle time of 1 or 3 s and an amplitude large enough to give voltage excursions about 30 mV around the holding potential, which was - 65 to - 70 mV in these cells. After AChE there was no change in the slope of the figure in either the depolarizing or hyperpolarizing direction, indicating no change in the current vs voltage ~lationship of the cells in voltage ranges below the firing threshold (not shown). Thresholds for Na+ and Ca2+ spike firing were measured as previously described at IO-min intervals before and after application of AChE to 13 cells. Both thresholds were found to be stable in the period preceding AChE application, and for up to 2 h in control cells in which both solutions perfusing the two-way switch were standard Ringer solution. Whereas the presence of AChE had no effect upon the threshold for Na+ spike firing it produced a signi~cant increase in the current needed to generate Ca2+ spikes in all but two of the cells examined (Figs 1,2), regardless of whether the recording was at the level of the soma or the dendrites. The onset of this effect was 10-20 min after commencement of AChE perfusion and the threshold continued to increase until it levelled out approximately 20-30 min later (Fig. 2C). In a few cells the recording was continued for 2-3 h following the 30-min period of AChE application. The increased Ca*’ spike threshold persisted throughout this period. In three out of four cells where lo-* M TTX was added first, the AChE-induced increase in Ca’+ spike threshold was no Ionger observed (Figs I C-E, 2B). TTX itself, however, appeared to increase the current threshold for Ca2+ spikes. This is presumably because it blocks the non-inactivating Na+ conductance

f3 1.5

ti

t

?!

$ z = g

_L1 20

Time In AChE (min)

TTX

1.0

0.5

0.0

i

~ 0

-10

Time

I 10

in AChE (min)

: 20

o.ov10f 0

20

Tlmo

JO

’

40

SO

In AChE (mln)

Fig. 2. Increase in current threshold for Ca*+ spikes. (A) Effect of AChE. The data were pooled from 13 cells. The current threshold in the control solution was normalized. The threshold was signticantly increased after 10 min (P c 0.05, Pratt’s teat, see Ref. 34) and after 20 min (P < 0.01). (B) TlX also increased the current threshold of the Ca*+ spike and blocked the affect of AChE on the threshold. Data from four cells, current was normalized as in A. (C) Time-course of the increase in Ca*+ spike threshold induced by AChE. Data taken from three cells. All bars show S.E.M.

60

295

Acetylcholinesterase in the cerebellar cortex

0’ 0.0

1

I

I

I

1

0.5

1.0

1.5

2.0

2.5

Current

(nA)

C

2+

Mn

Mn=+ + AChE

m

Mn2+ + AChE

Fig. 3. Effect of AChE on the active membrane properties. (A) Current vs frequency plots from a Purkinje cell before (open circles) and after AChE @led circles). The frequency was measured as an average over 100 ms during the final part of a 2-s square wave current stimulus. AChE reduced tiring frequency at a stimulus intensity giving maximal firing frequency in control medium. However, due to lack of Ca2+ spikes the cell could be driven to higher firing frequencies in the presence of AChE. (B) Current vs voltage plot in Mn*+ (2 mM) and no Ca*+ obtained with current pulses lasting 200 ms from a d.c. level of -0.7 nA. The firing level was about -45 mV. In the control solution a current of -0.4 nA was sticient to activate a plateau potential with spikes (the two upper data points). In AChE the plateau and firing was reached only after current increased to -0.3 nA (see text for further explanation). (C) Lissajous plot in control medium (upper trace) and in 2mM Mn*+ (lower trace). Mn*+ isolated and enhanced the Na+ plateau potential seen as an increase in hysteresis. The membrane potential reached -9omV (control) and -84 mV @in’+ ) at the most negative levels. The current reached -0.8 nA at the trough. Cycle time was 1 s. The arrows indicate the depolarizing and hyperpolarizing phases. (D) Lissajous plots obtained as in C. In Mn2+ the Na+ plateau potential was evident (upper trace), but it disappeared a&r 20 U/ml AChE (middle trace). When the ramp current was increased about 17% the Na+ plateau reappeared (lower trace). Trough membrane potential was - 80 mV at - 1.2 nA in the two upper traces and - 85 mV at - 1.3 nA in the lower trace. Calibration for C is the same as for D.

reduction of the Ca2+-dependent K+ current enables the Na+ component to express itself to a larger extent. This is seen as an increased hysteresis in the lower Lissajous plot compared to the upper plot of Fig. 3C. When AChE was added under these circumstances the hysteresis was completely abolished (Fig. 3D, upper and middle plots). However, a plateau could still be generated (Fig. 3D, lower plot), but it required an increase in stimulating current. AChE that had been pre-treated with the irreversible inhibitor soman, in order to block its esterase activity, also produced an increase in the Ca2+ spike threshold in the two cells examined. In another experiment, however, where the AChE solution was passed through a coarse glass filter with a large surface to adsorb any proteinaceous material before being applied to the cell, there was no change in the

Ca2+ spike threshold, or in any other observed variable. Thus, the observations were probably not due to some sort of contaminant in the drug solution. Similarly, the effects on Ca2+ spike threshold were specific for AChE. Superfusion of BChE onto four cells produced no alterations in input resistance, membrane potential or thresholds for Na+ and Ca2+ spikes. Acetylcholinesterase responses

and

excitatory

amino

acid

Iontophoresis of either glutamate or aspartate onto the dendritic region of the recorded Purkinje cells resulted in a dose-dependent depolarization leading to Na+ spike thing and, with sufficiently high and/or prolonged doses, Ca2+ spike firing. A standard dose producing a small depolarization was

M. APPLEYAKD

296

chosen for each cell, and this was then applied every fifth or 10th second throughout the experiment. The response to excitatory amino acid was shown to be stable for at least 20min prior to application of AChE-containing Ringer (20 U/ml). Five to 20 minutes following application of AChE there was a marked enhancement of the response to glutamate iontophoresis in five out of the six cells examined, regardless of whether the recording was at the level of the soma or the dendrites. Typically, in the presence of AChE an earlier onset of the depolarization was observed, and the duration and amplitude of the response were also increased. The one cell which did not respond to AChE in this way was the same cell that did not exhibit an increase in Ca”” spike threshold in response to AChE. Application of AChE also introduced a marked increase in the response to aspartate, again with a delay of 5-20 min, in all five cells examined. A marked increase in both the amplitude and duration of the response was observed, regardless of whether the recording was at the level of the soma or the dendrites. The time-course of the effect of AChE is shown in Fig. 4A. The effect is first observable after

and H.

JAHNSEN

approximately 5 min and is fully developed at about 20 min. When AChE that had been me-treated with soman was used (five cells) there was still a marked enhancement of the response to excitatory amino acids (Fig. 4B). If AChE was first passed through an adsorbent of proteinaceous material, no effect on the responses of Purkinje cells to excitatory amino acids was observed. Ringer solution with 40U/ml BChE produced no alteration in the response to glutamate (two cells) or aspartate (two cells), as seen in Fig. 4C. Effect of acetylcholinesternse on the climbing fihre response

Electrical stimulation of the white matter at the base of the folium generated an all-or-none response characteristic of climbing fibre activation of Purkinje cells (Fig. 5). i4s30 The response was sometimes preceded by an antidromic action potential. The duration of the response was prolonged by depolarization. In some cells the climbing fibre response (CFR) itself was followed by a long-lasting (SO-100 ms) depolarization or plateau component which was

A Cl min

A 20 mln

6 min

EOmV

I

c

sop SOnA

arp

SOnA

Fig. 4. Effect of AChE on responses of Purkinje cells to an excitatory amino acid. (A) Time-course of the development of the enhanced response to aapartate. The effect was evident after 5 min and was fuhy developed after 20 min. Recording from soma or proximal dendtite. Holding potential -63 mV. (B) inhibition of esterase activity did not prevent the efTbct of AChE on excitatory amino acid responses. Response to iontophoresis of aspartate recocded from Punkinje cell soma before and 20min after soman-treated AChE was introduced in the recording chamber. Though the &erase activity was completely eliminated the response to aapartate became larger and faster after AChE. Holding potential -60 mV. (C) Butyrylcholinesterase had no effect on the responses of Furkinje celb to excitatory amino acids. Single responses to iontophoretic p&es of aqartate recorded from soma before and during treatment with BCBE. Holding potential - 65 mV.

Acetylcholinesterase

A

291

in the cerebellar cortex

B rverage

of ten 8weev8

!\ AChE 20UIml / I

i

20mV

t control 7s

control

?Omr

ACM 2OUlml hyperpol8rlred

I

20mV

Fig. 5. Enhancement of CFRs by AChE. (A) Superimposed single CFRs obtained in control solution and in a solution containing AChE. Note that the late plateau-like part of the CFR was enhanced. Holding potential - 72 mV. (B) Superimposed averages obtained before. and during AChE superfusion. Same cell as in A. (C-E) In this cell AChE produced an enhancement that was so large that the Furkinje cell tired Na+ spikes during the late part of the CFR (C, D). Even when the cell was hyperpolarized by injection of inward current through the recording electrode the cell generated Na+ spikes after AChE treatment (E). The holding potential was -71 mV in C and -75 mV in E.

enhanced by depolarization and abolished by hyperpolarization; a similar climbing fibreinduced plateau potential has been described in rabbit Purkinje cells6 Climbing fibre responses were shown to be stable for at least 20min prior to application of AChE. Ten to 20 minutes following start of perfusion with an AChEcontaining Ringer solution (20 U/ml) the failing phase of the CFR was prolonged and, when present, the late plateau-like component was enhanced, both in amplitude and duration (Fig. 5A, B). This was observed in eight out of the 11 cells examined. Following AChE application a train of Na+ spikes riding upon the plateau potential could sometimes be observed (Fig. 5C, D). The Na+ spikes persisted even if the cell was hyperpolarized 3-5 mV by current injection (Fig. 5E). Larger hyperpolarizations abolished the plateau and the spikes. No change in the early spike burst component of the CFR was observed following application of AChE. Effect of amino acid uptake blockers Enhancement of synaptic responses could be produced by mechanisms in any of the following four categories. First, it could be due to a presynaptic action such as facilitation of transmitter release from the terminals. Second, it could be due to a transsynaptic action on the Purkinje cells through interneurons. Third, the increase could be due to a postsynaptic action such as an increase in the number

of available receptors or ion channels for the response or, fourth, the effect could be caused by an increase in the time the transmitter was available for interaction with the postsynaptic receptors. The tirst possibility seems unlikely to be the only effect because there should not be an increase in the response to putative transmitter released from iontophoretic pipettes. A transsynaptic effect is unlikely because intemeurons in the cerebellum are inhibitory and a disinhibition of the Purkinje cell would cause changes in the membrane potential and input resistance; this is not observed. The effect of AChE on Purkinje cells, therefore, must be due to postsynaptic actions or to interference with systems involved in removal of excitatory amino acids from the synaptic cleft. The latter possibility was addressed in the following way. There is a Ca2+-dependent membranebound transport mechanism for excitatory amino acids in both neurons and glia and this system can be blocked with APB.” We were unable to hold a neuron under stable conditions for a period long enough to measure first the effect of APB on iontophoretic responses and then the effect of both APB and AChE. APB itself changed the input resistance of Purkinje cells and thus the iontophoretic response could not be compared. Therefore we instead added APB to the control medium before the control period and we then tested the effect of AChE in the presence

298

M. APPLEYARD

and H. JAHNSEN

B

A glu 20nA

control

AChE 20min

APB 0.5mM

APB 0.5 mM

E

D

C control

DHK 0.1 mM

DHK 0.1 mM

k

20mV

I

10 ms

Fig. 6. Effect of APB and DHK. All sweeps are averages of eight trials. (A) Control response of a Purkinje cell to iontophoretic pulse of glutamate in the presence of 0.5 mM APB. (B) The rwpotue to the same dose of glutamate 20 min after AC&E was added to the APB-containing medium. Note that AChE did

not enhance the response to glutamate in the presence of APB. On the contrary, the response beMme somewhat smaller. Holding potential -7OmV. (C, D) DHK prolonged the falling phase of the CFR (arrow). (E) in the presence of DHK, AChE failed to change the CFR. Holding potential -62 mV.

of APB. When APB was added to the medium at a

concentration of OSmM prior to the addition of AChE the effect of the response of the Furkinje cells to excitatory amino acids was blocked in three experiments (Fig. 6A, B). In fact, in two of these experiments there was a reduction of the response to the amino acid in the late part of the experiment. It should be pointed out that the fact that AChE was prevented by APB from enhancing depolarizations by excitatory amino acids does not prove that AChE acts through interference with uptake mechanisms, because APB is known to have postsynaptic and presynaptic effects as well. The reduction of the glutamate response seen in Fig. 6B may be a result of such an action. The result merely demonstrates that we could not rule out the possibility of uptake inhibition by AChE. DHK also blocks uptake of excitatory amino acids.25 Indeed, when 0.1 mM DHK was added to the extracellular medium, the falling phase of a CFR was prolonged (Fig. 6C, D, arrow). Under these circumstances AChE in three experiments did not potentiate the CFR (Fig. 6E). This is expected if AChE acts like DHK by reducing amino acid uptake but, again, presynaptic or postsynaptic possibilities are not ruled out by these experiments. DISCUSWON

The cells used in this study were located in the cerebellar cortex and exhibited the same eiectro-

responsive characteristics as identified Purkinje cells described by other authors.7,‘4,2’*M.3’ The main findings of our study were that AChE does not act like an ordinary transmitter in the cerebellar cortex. Neither membrane potential nor input resistance of the neurons was changed. However, the current threshold for Ca2+ spike generation was increased by AChE so that the probability that the Furkinje cells would respond with CaZ+ spike firing during a given stimulation was decreased. As a consequence, the cells could sustain a higher Na+ spike tiring frequency because Ca*+ spikes followed by after-hyperpolariaations did not interfere with firing. Furthermore, the responses of Purkinje cells to excitatory amino acids thought to be transmitters in the cerebellar cortex were increased, and when the excitatory synapses formed by a climbing fibre and a Purkinje ceIi were activated the late part of the response of the Purkinje cell became larger. Acetylcholinesterase Purkinje cells

and the membrane properties

of

The lack of change in membrane potential and input resistance of the Purkinje cells is a major difference between our results and observations from dopaminergic neurons in the substantia nigra.‘91o Of the active membrane properties the maximmn attainable Na+ spike firing frequency and the threshold for Ca2+ spikes were affected. It is not known if AChE has these effects in the substantia nigra.

299

Acetylcholinesterase in the cerebellar cortex What is the underlying mechanism by which AChE induces a selective change in firing properties without affecting the membrane potential or input resistance measured below the Na+ spike threshold? The increase in Ca”+ spike threshold was observed in both somatic and dendritic recordings indicating that it was not produced by a caption of the el~troto~~ conduction of the stimulus from the soma to the dendritic regions of the cell where the Ca’+ spikes originate.2’*3’ When the Naf plateau potential was isolated and enhanced by the use of extracellular Mn2+ it was obviously affected by AChE. We have not shown that there is no effect of AChE on the voltage-sensitive Ca2+ channels responsible for the CaZf spike. However, in view of the fact that the AChE-induced increase in Ca2+ spike threshold is TV&sensitive and therefore Na+-dependent and that AChE suppressed the Na’ plateau potential, it is reasonable to conclude that AChE changes the firing pattern of Purkinje cells from one dominate by Na+ and Caa+ spikes to a pattern with Na+ spikes only and an increased firing range by suppression of the non-inactivating Na+ current. This current normally contributes to activation of Ca2+ spikes in the dendrites. These spikes in turn slow down firing of Na+ spikes by activating a Ca2+-dependent I(+ current. Acetylcholinesterase and the responses of Purkinje cells to excitatory amino acids We have investigated the effect of AChE on responses to the two excitatory amino acids aspartate and glutamate applied at the dendritic level. Numerous studies using selective destruction of neuron populations and analysis of intracellular amino acid content have shown that aspartate is probably the transmitter released from climbing fibres and glutamate is the transmitter of the parallel fibre system (for a review see Ref. 22). Furthermore, the excitatory synaptic currents of Purkinje cells are suppressed by the non-N-methyl-D-aspartate blocker 6-cyano-2,3dihydro-7-nitroquinoxalinedione (CNQX).33 Thus, these two amino acids are the principal excitatory transmitters of the cerebellar cortex. Responses to both transmitters were enhanced by AChE. It was di~c~t to quantify the enhan~ment because of the highly non-linear properties of the Purkinje cell membrane. In cells hyperpolarized beyond - 65 mV, i.e. into a range where the non-linearities are small, the increase in amplitude of the response was in the order of 10%200% of the control value. The mechanism behind this enhancement is at present unknown. It is unlikely that it is a result of the action of AChE upon the non-inactivating Na+ current since such an action would be expected to decrease the extent of membrane depolarization produced by the excitatory amino acids. The experiments with APB and DHK indicate that the effect may be due to reduced or removal of excitatory amino acid from the vicinity of the receptors on the Purkinje cells. However, it must be pointed out that these experiments are not con-

clusive and they do not rule out other mechanisms such as direct effects on the postsynaptic membrane. Actions of DHK and in particular APB may not be restricted to blockade of amino acid uptake; these drugs may also interfere with other sites for excitatory amino acids. D-APB, for example, is a weak non-selective antagonist of excitatory amino acid receptors.‘6 Indeed we did observe some reduction of aspartate and glutamate responses upon addition of APB in some cells. A clarification of this issue must await further studies of the Purkinje cell responses including single channel recordings. Acetylcholinesterase and the climbing jibre response The late part of the CFR which was enhanced by AChE has strong components of intrinsic membrane properties of the Purkinje cells. The non-inactivating Na+ and Ca2+ currentsM*31 probably contribute to this potential. On the other hand, the increased current th~hold for Ca’+ spikes is best explained by a reduction of the non-inactivating Na+ current which should lead to a decrease in the potential. We cannot at present explain this apparent discrepancy, but it may simply be a result of different underlying mechanisms. The DHK experiment shown in Fig. 6C-E suggests that the CFR is enhanced because the natural transmitter stays longer in the vicinity of the postsynaptic receptors. However, the interpretation of this result is complicated by the fact that DHK may have other actions than uptake inhibition. The slight reduction of the CFR after AChE points in this direction. The only safe ~on~l~ion is that we cannot rule out an effect of AChE on amino acid uptake systems. The late plateau-like component of the CFR has been described by Cambell et a1.6in rabbit Purkinje cells. It has been suggested, on the basis of studies utilizing different time intervals between climbing fibre and parallel fibre stimulation,ls that this plateau potential may be responsible for the decreased responsiveness of Purkinje cells to stimulation of their parallel fibre input following climbing fibre stimulation. We observed, however, that the excitability of the Purkinje cell was increased during the p1at~u-like potential when AChE was present. Thus, it seems that AChE drives the Purkinje cell firing CFRs from a state of decreased excitability towards one of increased excitability. CONCLUSION

The results presented in this paper are a direct demonstration of a modulatory effect of AChE upon the response of a cell, in this case the Purkinje cell, to stimulation of its afferents and the transmitters endogenous to them. Our results with AChE suggest that the function of AChE in the cerebellar cortex is to increase the ability of Purkinje cells to respond to excitatory input and fire Na+ action potentials over an expanded range of

300

M. APPLEYARDand H. JAHNSEN

frequencies.

CIearly,

have important

such an action

consequences

of AChE

could

for the activity

of

target cells. It is now necessary to investigate other regions of the brain where secretion of AChE has been suggested, such as the substantia nigra and the hippocampus,j to determine whether AChE

produces

similar

effects

to those

observed

in the

cerebellum. Acknowledgements-The

project was supported by the Wellcome and Carl&erg Trusts, the Trustees of Bruin, The Danish Research Academy and the Danish Medical Research Council.

REFERENCES

1. Allen G. I. and Tsukahara N. (1974) Cerebrocerebellar communication systems. Physiol. Rev. M, 957-1006. 2. Appleyard M. E. and Jahnsen H. (1989) Non-cholinergic effects of acetylcholinesterase upon the membrane properties of mammalian cerebellar Purkinje cells. Eur. J. Neurosci., Suppl. 2, 129. 3. Appleyard M. E. and Smith A. D. (1987) Spontaneous and carbachol-evoked in uiuo secretion of acetylcholinesterase fr&n the hippocampus of the rat. Neurochim. Znt. 11, 397406. 4. AoDlevard M. E.. Vercher J. L. and Greenfield S. A. (1988) Release of acetvlcholinesterase from the guinea-Die ~ .., &ibeilum in vivd. Neuroscience 25, 133-138. 5. Butcher L. L. and Woolf N. J. (1982) Monoaminergic-cholinergic relationships in the chemical communication matrix of the substantia nigra and neostriatum. Brain Res. Bull. 9, 475-492. 6. Cambell N. C., Ekerot C. F., Hesslow G. and Oscarsson 0. (1983) Dendritic plateau potentials evoked in Purkinje cells by parallel fibre volleys in the cat. J. Physiol. 340, 209-223. 7. Chan C. Y., Hounsgaard J. and Nicholson C. (1988) Effects of electric fields on transmembrane potential and excitability of turtle cerebellar Purkinje cells in vitro. J. Physiol. 402, 751-771. 8. Chatonnet A. and Lockridge 0. (1989) Comparison of butyrylcholinesterase and acetylcholinesterase. Biochem. J. 260, I

625-634.

9. Crawford J. M., Curtis D. R., Voorhueve P. E. and Wilson V. J. (1966) Acetylcholine sensitivity of oerebellar neurones in the cat. J. Physiol. 186, 139-165. 10. Crepe1 F. and Dhanjal S. S. (1982) Cholinergic mechanisms and neurotransmission in the cerebellum of the rat. An in vitro study. Brain Res. 244, 59-68. 11. Cue110A. C., Romero E. and Smith A. D. (1981) In vitro release of acetylcholinesterase from the rat substantia nigra. J. Physiol. 312, 14P-15P.

12. Cue110 A. C. and Sofroniev M. (1985) In Neurofrunsmitters in Action (ed. Bousfield D.), pp. 309-318. Elsevier, Amsterdam. 13. De Sarno P., Giacobini E. and Downen M. (1987) Release of acetylcholinesterase from the caudate nucleus of the rat. J. Neurosci. Rex 18, 578-590. 14. Eccles J. C., Llinls R. and Sasaki K. (1966) The excitatory synaptic action of climbing fibres on the Purkinje cells of the cerebellum. J. Physiol. 182, 268-296. 15. Ekerot C. F. (1985) Climbing fibre actions of Purkinje cells--plateau potentials and long-lasting depression of parallel fibre responses. In Cerebellar Functions (eds Dichgans J., Bloedel J. R. and Precht W.), pp. 268-274. Springer, Berlin. 16. Evans R. H., Francis A. A., Jones A. W., Smith D. A. S. and Watkins J. C. (1982) The effects of a series of w-phosphonic a-carboxylic amino acids on electrically evoked and excitatory amino acid-induced responses in isolated spinal cord preparations. Br. J. Pharm. 75, 65-75. 11 Foster A. C. and Fagg G. E. (1984) Acidic amino acid binding sites in mammalian neuronal membranes: their characteristics and relationship to synaptic receptors. Bruin Res. Rev. 7, 103-164. 18. Greenfield S. A., Gnmewald R. A., Foley P. and Shaw S. G. (1983) Origin of various enzymes released from the substantia nigra and caudate nucleus: effects of 6-hydroxydopamine lesions of the nigro-striatal pathway. J. camp. Neural. 214, 87-92. 19. Greenfield S. A., Jack J. J. B., Last A. T. J. and French M. (1988) An electrophysiological action of acetylcholinesterase independent of its catalytic site. Expl Brain Res. 70, 44-444. 20. Greenfield S. A., Nedergaard S., Webb C. and French M. (1989) Pressure ejection of acetylcholinesterase within the guinea-pig substantia nigra has non-classical actions on the pars compacta cells independent of selective receptor and ion channel blockade. Neuroscience 29, 21-25. 21. Hounsgaard J. and Midtgaard J. (1988) Intrinsic determinants of firing pattern in Purkinje cells of the turtle cerebellum in vitro. J. Physiol. 402, 731-749.

22. Ito M. (1984) The Cerebellum and Neuronal Control. Raven Press, New York. 23. Jahnsen H. and Appleyard M. E. (1989) Acetylcholinesterase modulates membrane properties and synaptic transmission in mammalian cerebellar cortex. Sot. Neurosci. Absrr. Part 2 15, 973. 24. Jahnsen H. and Laursen A. M. (1981) Brain slices. In Current Methods in Ceffular Neurobiology Vol. 3 (eds Barker J. L. and McKelvy J. F.). John wile;, New York. 25. Johnston G. A. R.. Kennedv S. M. E. and Twitchin B. (1979) Action of the neurotoxin kainic acid on high affinity uptake of L-glutamic acid in rat brain slices. J. Neurochem. $2, 121-128. 26. Karczmar A. G. (1969) Is the central choline&z nervous system overexploited? Fedn Proc. 28, 147-157. 27. Kasa P. and Csillik B. (1965) Cholinergic excitation and inhibition in the cerebellar cortex. Nature %@I,695-696. 28. Lamarre Y., De Mont&y C., Dumont M. and Weiss M. (1971) Harmaline-induced rhythmic activity of cerebellar and lower brain stem neurons. Brain Res. 32, 246-250. 29. Levey A. I., Wainer B. H., Rye D. B., Mufson E. J. and Mesulam M.-M. (1984) Choline acetyltransferase-immunoreactive neurons intrinsic to rodent cortex and distinction from acetylchoiinesterase-positive neurons. Neuroscience 13, 341-353.

30. Llinls R. cerebellar 31. Llinis R. cerebellar

and Sugimori M. (1980) Electrophysiological properties of in vitro Purkinje cell somata in mammalian slices. J. Physiol. 305, 171-195. and Sugimori M. (1980) Electrophysiological properties of in vitro Purkinje cell dendrites in mammalian slices. J. Physiol. 305, 197-213.

Acetyicholinesterase

in the cerebellar cortex

301

32. Llinis R. and Volkind R. H. (1973) The olivo-cerebellar system: functional properties as revealed by harmaline-induced tremor. Expl Brain Res. 18,69-87. 33. Perkel D. J., Hestrin S., Sah P. and Nicholl R. A. (199Q) Excitatory synaptic currents in Purkinje cells. Proc. R. Sot. Land. R. 241, 116-121. 34. Rahe A. J. (1974) Tables of critical values for the Pratt Matched Paired Signed Rank Statistic. J. Am. Stat. Ass. 69, 368-373. 35. Silver A. (1974) Z%e Biology of Chol~srer~es.

North Holland, Ams~rdam. 36. Taylor S. and Greenfield S. A. (1989) Release of acetylcholinesterase from the guinea-pig substantia nigra during peripheral nerve stimulation. Brain Res. 482, 356-358. 37. Thatch W. T. (1972) Cerebellar output: properties, synthesis and use. Brain Res. 40, 89-97. (Accepted 18 September 1991)