32
Brain Research, 527 (1990) 32-40 Elsevier
BRES 15818
Effects of tetrahydro-9-aminoacridine on cortical and hippocampal neurons in the rat: an in vivo and in vitro study P. Dutar, M.H. Bassant and Y. Lamour I.N.S.E.R.M., UniM 161, Paris (France)
(Accepted 6 March 1990) Key words: Tetrahydroaminoacridine; Physostigmine; Acetylcholine; Cerebral cortex; Hippocampus; Rat
The effects of tetrahydro-9-aminoacridine (THA), an anticholinesterase drug, have been studied in the rat both in vivo (cerebral cortex) and in vitro (CA1 field of the hippocampus) and compared with those of physostigmine. In the cerebral cortex THA potentiated the excitatory effect of acetylcholine in most neurons, including cortical neurons recorded from chronic unanesthetized animals. In vitro, THA (but not physostigmine) had a depolarizing, atropine- and tetrodotoxin-insensitive effect. This effect is associated with an increase in membrane resistance which suggests a direct effect of THA on hippocampal neurons. In addition THA blocked the slow inhibitory postsynaptic potential. At the same concentration THA potentiated the slow cholinergic excitatory postsynaptic potential produced by electrical stimulation of the cholinergic afferents. Its potency was, however, about 10 times lower than that of physostigmine. These results show that THA: (1) is an anticholinesterase much less potent than physostigmine; but (2) has also direct effects on central neurons, not observed with physostigmine and unrelated to its anticholinesterase activity.
INTRODUCTION A deficit of the central cholinergic system has been found in dementia of the Alzheimer type (SDAT) and might also occur with aging 2. One of the therapeutic approaches for improving cognitive functions in patients with S D A T is to act pharmacologically at the level of the central cholinergic synapse. O n e possibility is to administer anticholinesterase agents which increase the efficacy of cholinergic transmission by protecting A C h from being degraded. The cholinesterase inhibitor physostigmine has been extensively used but only modest improvements have been observed 15'24. Recently, it has been claimed that tetrahydro-9-aminoacridine ( T H A ) , a potent centrally acting anticholinesterase 13'a4 produces improvement of cognitive functions in Alzheimer's patients 26. Since other cholinesterase inhibitors (such as physostigmine) had only limited effects, it is possible that T H A may act in a different way to produce its clinical effects. It has been recently suggested that T H A may act in a way similar to 4-aminopyridine (4-AP). This monoamino derivative of pyridine has only weak anticholinesterase properties but blocks potassium conductances in excitable tissues and thereby increases the release of the neurotransmitter, particularly at the neuromuscular junction 23,27 and in central structures 7. Interestingly, 4-AP has also met with some success in the treatment of
S D A T 28. Some authors have suggested that, like 4-AP, T H A could block potassium channels in the central nervous system 22'25. In the present study, we describe the effects of T H A on two cerebral structures known to receive a dense cholinergic innervation, the cerebral cortex and the hippocampus. We compare these effects with those of the well known cholinesterase inhibitor physostigmine (esefine). We used two different approaches: first, we recorded extracellulady froni cortical neurons in vivo in awake restrained or anesthetized rats; and second, we performed intracellular recordings from hippocampal pyramidal neurons in the in vitro slice preparation.
MATERIALS AND METHODS In vivo recordings Most of the methods have been previously published 4'16. In brief, male Sprague-Dawley rats (380-420 g) were anesthetized with pentobarbital (45 mg/kg i.p.) and a chronic restraint device was implanted. The skull was trephined over the first somatosensory cortex (SI). However, a thin layer of bone was left in place to protect the dura. After a period of recovery the animal was habituated to the recording apparatus for 8-10 days. During repetitive trials of increasing duration, the head was painlessly secured to the stereotaxic frame. The rat was comfortably supported by a hammock. The state of consciousness was monitored by observation of the ECoG. Slow-wave sleep and paradoxical sleep were routinely observed, as well as period of quiet wakefulness, confirming that the restraint was well tolerated.
Correspondence: P. Dutar, I.N.S.E.R.M., Unit6 161, 2 rue d'Al6sia, 75014 Paris, France.
0006-8993/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
33 For the recording session, the remaining layer of bone was gently removed and the dura was opened under a stereomicroscope. There was no need to use a local anesthetic for this procedure. Vertical penetrations were made in the part of the SI area containing neuronal aggregates in layer IV, in the fore- or hind-limb representation. Conventional amplification methods were used to record extracellular activity using micropipettes (filled with 1 M NaCI and 2% Pontamine sky blue) attached to a multibarreled micropipette. Solutions for testing by iontophoresis included: sodium chloride (for current compensation, 165 raM), sodium L-glutamate (0.2 M; pH 8), acetylcholine chloride (0.2 M, pH 4), carbamyl choline chloride (earbachol 0.2 M, pH 4), 7-aminohutyric acid (0.1 M, pH 5) physostigmine liemisulfate (0.2 M, pH 5), all purchased from Sigma, and tetrahydro-9-aminoacridine (0.2 M, pH 5), purchased from Aldrich (U.S.A.). Spontaneously, as well as non-spontaneously active neurons (identified by the application of glutamate) were studied. Data were stored on a rectilinear strip chart recorder and on a computer for subsequent analysis using interspike interval histograms (IIH). Standard ACh and THA application were 60-80 nA and 80-100 nA for 24 s, respectively. At the end of each track, a dye deposit was
ACh 60
THA 100
made. The recording session lasted from 3 to 8 h. At the end of the recording session, a reward was given to the rat which was returned to his cage. The rat quickly engaged in his usual activities (grooming, feeding) without signs of discomfort. The experiment lasted for 4 consecutive recording sessions (4 days). Then the rat was deeply anesthetized with pentobarbital and perfused through the heart with saline followed by 10% formaldehyde in saline. Examination of the digestive tracts of the animals never revealed any ulcers. Frozen coronal brain sections were stained with safranin. Electrode penetrations were reconstructed on camera lucida drawings of these sections. The position of each neuron recorded was known from the reading of the micrometer gauge of the mierodrive (+1/~m). It was thus possible to know the laminar position of each neuron. The methods used in the anesthetized rat were similar except that the animal received urethane (1.5 g/kg i.p.) and no restraining device was implanted.
In vitro recordings All in vitro experiments were performed on hippocampal slices. Male Sprague-Dawley rats (150o250 g) were anesthetized with
ACh 60
ACh 60
A
P~V~60 ACh 100 m
ACh 100 m
THAIO0 ACh 100 m
ACh 100 m
B
ACh 100
Phy8o 60 m
ACh 100 m
ACh 100 m
ACh 100 m
C E20 ACh 100 m
D
THA 100 m
CARB
,o
ACh 100 m
ACh 100
THA
CARB
,oo _,o
,,, Imin
ACh
_,oo
THA
_,oo
ACh
_1oo
Fig. 1. Effects of THA and physostitnnine on acetyicholine-inducedneuronal responses in the somatic sensory cortex, recorded from the chronic unanesthetized rat. (Iontophoretic applications, current in nA, vertical calibration bar: 20 impulses/s.) A: layer V neuron. Effect of THA applied before ACh. Eight minutes after application, the effect of THA is not fully dissipated. B,C: layer V and VI neurons, respectively. Notice that in both cases the effect of physostigmine is more impressive than that of THA in spite of the lower amounts of current applied (C: continuous recording from one neuron). D: layer V neuron. The excitation induced by carbachol is not potentiated by THA while on the same neuron THA potentiated the ACh response.
34 halothane and decapitated. The hippocampus was quickly removed and placed in a cold medium. Slices (400/~m thick) were cut and placed in a holding chamber for at least 1 h. A single slice was then transferred to the recording chamber and held between two nylon nets, submerged beneath a continuously superfusing medium that had been warmed and pregassed with 95% O z and 5% CO 2. The composition of the medium was (mM): NaCI, 119; KC1, 2.5; MgSO4, 1.3; CaCI2, 2.5; NaHCO 3, 26.2; NaH2PO 4, 1.0; glucose, 11. The temperature of the medium was maintained between 29 and 31 °C. Drugs were dissolved directly in the medium just before use and were applied by addition to the superfusing medium. Drugs included: carbachol, physostigmine hemisulfate, atropine sulfate and tetrodotoxin, all purchased from Sigma; THA was obtained from IPSEN (France) or Aldrich (U.S.A.). Conventional intracellular recordings from CA1 pyramidal neurons were obtained using glass micropipettes filled with 2 M potassium acetate and having resistances of 60-140 Mr2. A bridge circuit (Axoclamp 2A) was used to apply current pulses for the measurement of membrane resistance and for the generation of afterhyperpolarizations (AHPs) which follow a train of spikes. Intracellular voltage and current recordings were stored on a rectilinear strip chart recorder (Gould). Fast events such as fast excitatory postsynaptic potentials (EPSPs) or inhibitory postsynaptic potentials (IPSPs) and the responses to depolarizing pulses recorded to evaluate the accommodation of the spike discharge (see Fig. 3) were stored on a digital oscilloscope (Nicolet) and plotted on a digital plotter (Hewlett-Packard). Slow cholinergic EPSPs (sEPSPs) were elicited by electrical stimulation of the cholinergic fibers located in the stratum oriens (repetitive stimulation 20-30 Hz for 0.5 s, 15-50 V). Stimulation pulses were applied between the two poles of the bipolar electrode, one pole being inserted into the slice and the other pole being just above the slice.
ACh 20
ACh 20
RESULTS
In vivo recordings U n a n e s t h e t i z e d rat. T h e effect of T H A has been studied on 46 neurons. T H A (100 n A , 24 s) had no effect on most (87%) cortical n e u r o n s when a p p l i e d alone. A weak excitation was o b s e r v e d in 4 neurons and an inhibition in two. The effect of T H A on the n e u r o n a l excitation induced by A C h has been studied in 37 neurons. T H A was applied 30 s before A C h ( T H A 80-100 n A , 24 s; A C h 60-80 n A , 24 s). Most neurons studied (n = 31) were located in infragranular layers. A p o t e n t i a t i o n of the excitation induced by A C h was o b s e r v e d in 28 neurons (76%). In most cases the p o t e n t i a t i o n of the duration of the excitation was m o r e striking than the p o t e n t i a t i o n of the a m p l i t u d e of the response (although both could h a p p e n together, see Fig. 1). The effect of the previous T H A application was p r e s e n t for a long time (5-11 min), but was, nevertheless, reversible (Fig. 1). The p a t t e r n of the n e u r o n a l discharge during the A C h - i n d u c e d excitation was not m o d i f i e d by the previous application of T H A . T h e excitation induced by carbachol (a p o t e n t cholinergic agonist which is not b r o k e n down by acetylcholinesterase) was not p o t e n t i a t e d by T H A (n = 7, Fig. 1D). Anesthetized animals. Twenty-three neurons were re-
THA ACh 50 20
ACh 20
ACh 20
A 5 rain
ACh 80
Physo 80
ACh 80
ACh 80
B1 1 min
ACh 80
ACh
Physo 60
ACh 80
20 min
Fig. 2. Same conventions as in Fig. 1. Urethane-anesthetized rat. Layer V neurons. Notice that in A and B, an inhibitory component of the neuronal response to ACh is potentiated by THA (A) or by pbysostigmine (B). B1 and B2 are continuous recordings from the same neuron.
35
A1 CONT.
2
THA 20pM
3
lmin
el
._I
4 C A R B 5pM
WASH
ls
2
l._ l lnA-
3
._J
_]
I
L_
100ms
C
THA ,501JM
~ s
3.,I Fig. 3. Effects of low concentration of THA on membrane properties of a CA1 pyramidal neuron. AI: the afterhyperpolafization (AHP) is elicited by a brief (60 ms) depolarizing pulse. A2: constant current hyperpolarizing pulses are applied at 0.5 Hz to measure the membrane resistance; THA (20/~M) is applied in the superfusion bath during the period indicated by the horizontal bar. At this concentration, THA does not induce any change in the membrane polarization but induces a small decrease in the AHP amplitude probably resulting from its anticholinesterase effect. A 3" the size of the AHP recovers after washing out THA from the bath. A4: carbachol (5/~M) blocked the AHP. BI: accommodation of the spike discharge induced by application of a long duration depolarizing pulse. THA does not modify accommodation (B2), in contrast with carbachoi (B3). Recordings in A and B are from the same neuron, membrane potential = -58 mV. C: top, effect of THA on the synaptic events induced by stimulation of Schaffer collateraYcommissuralfibers. Left, control trace; middle, THA (50/~M) selectively blocks the slow IPSP but does not affect the EPSP and the fast IPSP; fight, the slow IPSP recovers after washing out THA. Bottom, THA (50/~M) superfused for 30 rain does not affect the size or the shape of the spike.
corded from rats under urethane anesthesia, most of them from layer V. A potentiation of the response to ACh was observed in 17 neurons, and a suppression of the response to A C h in 3 cases. No effect was observed in 3 additional cases. These results differ from those obtained in unanesthetized animals on the following points. (1) A suppression of the response to ACh had never been observed in the absence of anesthesia. (2) Among the 17 responses to ACh potentiated by T H A , a potentiation of the neuronal excitation was observed in 11 cases. However, in the other 6 cases, a potentiation of an inhibitory component of the response was observed (Fig. 2). The inhibitory component could be present before T H A applications (biphasic responses
to ACh, Fig. 2) or only after T H A applications. A similar phenomenon could be observed with physostigmine (Fig.
2). A comparison of the potency of T H A and physostigmine was made in 20 neurons. In some but not all cases physostigmine was more potent than T H A (Fig. 1).
In vitro recordings Stable intracellular recordings were obtained from 43 hippocampal neurons located in the pyramidal layer of the CA1 region. The following neuronal properties were studied: (1) the membrane potential; (2) the A H P that follows a train of action potentials and is due to an increase in a calcium-dependent potassium conductance; (3) the accommodation of the spike discharge; (4) the
36
Ba._.c Fig. 4. THA blocks the baclofen-induced hyperpolarization. Top: left, EPSP and biphasic IPSP are triggered by stimulation of afferent fibers (arrow). Right, baclofen (B, 40 /aM) is applied in the superfusion medium for the time indicated by the horizontal symbol. It induced an hyperpolarization and a decrease in membrane resistance. Bottom: left, 20 min after the beginning of the superfusion with THA (50/~M), the slow IPSP is blocked. Right, in addition, the postsynaptic hyperpolarizing response to baclofen is also blocked.
CONTROL
THA 50pM
A
CONTROL
~
l iota v
500ms
I rain
A C h from these terminals on the p y r a m i d a l neurons 6. T H A , at 5 - 2 0 / ~ M a p p l i e d in the superfusion m e d i u m (n = 19) did not d e p o l a r i z e the n e u r o n s and did not
WASH
THA 7 0 p M
CONTROL
Ba._.c
.~
(20min) E P S P and IPSP triggered by electrical stimulation of Schaffer collateral/commissural fibers (the IPSPs are subdivided into a fast IPSP m e d i a t e d by G A B A acting on G A B A A r e c e p t o r s and a slow IPSP m e d i a t e d by G A B A acting on G A B A a receptors (see ref. 18); and (5) the sEPSP, a depolarizing event elicited by electrical stimulation of cholinergic afferents and due to the release of
.~
ATROPINE 2pM
PHYSO 5~JM
-
30sec
B
CONTROL
THA 51JM
PHYSO 5uM
15' r .......
"=-~
4/-
~"
r .....
~-
"//" ....ilumv 30s
Fig. 5. Effects of THA on the slow cholinerglc EPSP. A: top left, the slow cholinerglc excitatory postsynaptic potential (sEPSP) is eficited by stimulating the cholinerglc fibers in stratum oriens. Triangles indicate the stimulating pulses (20 V, 25 Hz, 0.5 s). Stimulation artefact is followed by a fast hyperpolarizing phase and then by the sEPSP (relatively small in the absence of cholinesterase inhibitor). THA (70/~M), is then applied in the supeffusion medium and 15 min later the sEPSP is markedly increased. After washing out THA, physostigmine (5/~M) dramatically increases the sEPSP; atropine (2/~M) antagonizes this depolarizing potential. Notice that physostigmine is much more effective than THA in increasing the size of the sEPSP and that its effect is more prolonged. Resting membrane potential, -54 inV. B: on this other neuron, the effects of THA and physostigmine on the sEPSP are compared at the same concentration (5/~M). Left: control trace (15 V; 25 Hz, 0.5 s). Middle trace: THA, applied in the superfusing medium does not induce any significant enhancement in the sEPSP size. Right: THA is then washed for 15 min and physostigmine is applied at the same concentration. Physostigmine strongly potentiates the sEPSP which reached the level for action potential generation. Resting membrane potential: -65 inV.
37
A,
THA 0.SmM
CONT
.•OmV lmln
2 TTX
3 ATROP
B,
CONT
3
~m,~v~ •
-eo
-eo
i (~AI
* (.AI
-[
";
~
ATROP
v.i.v~
i (ha)
r
-izo
-140
C
~
.-vm
PHYSO 0.3raM
Fig. 6. Effect of high concentration of THA on a CA1 pyramidal neuron. THA (0.5 mM) induces a slow depolarization of long duration (A 0 associated with a small increase in membrane input resistance. B 1, plot of current-voltage relationship (from records illustrated in inset) showing the voltage responses of the neuron to hyperpolarizing current pulses of variable intensity before (0) and during (O) THA appfication; the increase in the slope of the I N curve indicates an increase in the membrane resistance. The long-duration depolarizing effect of THA persists in the presence of TTX (1/~M) (A2) and is still associated with a small increase in the membrane resistance (B2) indicating that this effect does not result from a presynaptic effect of the THA releasing an excitatory neurotransmitter. The depolarizing effect (A3) and the increase in the input resistance (B3) are not antagonized by atropine (2/~M) applied by superfusion in the bath for 10 rain before THA application) indicating that THA is not acting through muscarinic receptors. The slope values for all I/V curves were calculated by linear regression analysis (R values ranging from 0.97 to 0.99). Calibrations in A: horizontal, 1 rain; vertical, 10 mV. All recordings from the same neuron; resting membrane potential = -65 mV. C: physostigmine applied at high concentration (0.3 mM) in the presence of TTX (1/zM) fails to induce any depolarization and change in membrane resistance. Resting membrane potential = -66 inV.
produce any potentiation of the slow EPSP. In 4 neurons a m o n g 14 tested, this concentration of T H A induced a small decrease in the size of the A H P (Fig. 3). Interestingly, the depression of the A H P did not depend on the concentration of T H A applied. This effect might depend on the level of activity of the cholinergic afferents impinging on the pyramidal neurons. However, at this concentration, T H A did not affect the accommodation of the spike discharge (Fig. 3). In contrast, only 1 - 5 / ~ M carbachol were necessary to block the A H P and to inhibit the accommodation of the spike discharge in the same neurons (n = 5) (Fig. 3). Moreover, carbachol never induced inhibitory responses as in the cortex. Increasing the concentration of T H A (50-70 /~m) produced more obvious effects: T H A induced a small atropine-resistant depolarization with an increase in m e m b r a n e resistance, and a potentiation of the atropine-
sensitive cholinergic sEPSP (Fig. 5). In addition, at this concentration, T H A selectively blocked the slow IPSP in 7 neurons out of 9 (Figs. 3 and 4). A m o n g these 7 neurons, 6 had their fast IPSP unaffected. In two neurons, the G A B A B agonist baclofen was added to the bath (Fig. 4). T H A blocked the baclofen-induced hyperpolarization in these two neurons as well as the slPSP. The effects on the EPSP were not consistent. They were generally widened but this could be due to the loss of the slow inhibitory potential. The size and the s h a p e of action potential were not altered (n = 8, Fig. 3C). Higher concentrations (0.1-1 m M ) of T H A were applied on 16 neurons: we observed a depolarizing effect of T H A in 12 neurons and no effect in the remaining 4. This depolarization was slow in onset and prolonged (Fig. 6), was not antagonized by atropine (n = 4; Fig. 6), and persisted in the presence of tetrodotoxin (n = 2, Fig. 6)
38 suggesting a direct effect, independent of a presynaptic release of ACh. The AHP was studied in 11 neurons: it was slightly depressed in 4, unaffected in 3 and its duration was prolonged in the remaining 4. This latter effect was observed only at high concentrations of THA (>0.5 mM). The small depression of the AHP observed in 4 neurons was not larger than the one observed when using low concentrations of THA. We compared the effects of THA and physostigmine on the sEPSP first at low concentrations (n = 3) and then at higher concentrations (0.3 mM, n -- 5). As illustrated in Figs. 5 and 6, physostigmine was much more potent in enhancing the cholinergic response. The effect induced by physostigmine was obtained at a concentration 10-fold smaller than that of THA and was more prolonged (Fig. 5). At higher concentrations, a depolarizing effect similar to the one induced by THA was never observed with physostigmine (Fig. 6). DISCUSSION The recent publication of a dramatic improvement in cognitive functions in Alzheimer patients receiving THA triggered a surge of interest for this substance. A significant number of papers describing the pharmacol, ogy of THA has been recently published. The picture emerging from these data is still quite confusing. THA has been shown to have weak agonist properties at the PCP binding site I and antagonistic properties at the muscarinic binding site 21. It has also been shown to inhibit high affinity choline uptake and ACh release 5, evoked GABA release 8 and monoamine uptake 1°. Our results can be summarized as follows. (1) THA applied by microiontophoresis on cortical neurons potentiates the excitation induced by ACh. The duration of the excitation is more dramatically potentiated than the amplitude (i.e. the intensity) of the excitation. This effect of THA is more clear-cut in the absence of anesthesia. In the presence of anesthesia (urethane), THA is also able to potentiate inhibitions induced by ACh. (2) The effects observed in vitro were clearly dose dependent. A potentiation of the sEPSP was observed at a concentration of THA higher than 50 /~M. THA behaved neither as an agonist, nor an antagonist of ACh. It has also a direct, atropine resistant, postsynaptic effect on CA1 pyramidal neurons (increase in resistance, depolarization) and blocks the slPSP. The comparison with physostigmine shows that: (1) THA is about 10 times less potent in potentiating the slow EPSP; and (2) physostigmine does not have the direct postsynaptic effects observed with THA. The present results provide evidence that THA be-
haves as a cholinesterase inhibitor: it potentiates the effect of synaptically released ACh (in vitro experiment) as well as the effect of exogenous ACh (in vivo experiment). The in vivo experiment also shows that this effect is specific for ACh: the effect of carbachol (which is not destroyed by ACHE), is not potentiated by THA. The fact that in vivo THA potentiates more the duration than the amplitude of the ACh-induced excitation suggests that THA would allow the exogenous ACh to occupy the receptors for a longer period of time, but would not induce a stronger effect on firing rate. The potentiation by THA of ACh-induced inhibitions observed under urethane anesthesia is more difficult to explain, since such a phenomenon was not observed in unanesthetized animals. It is possible that in the presence of anesthesia the activity of the ACh-sensitive inhibitory cortical interneurons is higher. Then the effect of ACh, when applied by iontophoresis, would preferentially activate these interneurons, resulting in the inhibition of the pyramidal neurons (which are the most likely to be recorded under our experimental conditions). The fact that inhibitory cortical interneurons are excited by ACh has been clearly demonstrated in vitro 17. The preferential excitation of these interneurons would explain the 'mixed' effect of iontophoretic ACh on the principal neurons. Since this effect of ACh is likely to be potentiated by THA, it would explain why the inhibitory responses are the most potentiated. THA had little effect on the neurons in the absence of ACh. A weak excitation was observed in only 8% of them. This weak excitation is not similar to the depolarization observed in vitro because the time course of these effects is quite different. Therefore it could be due to the potentiation by THA of the excitatory effect of the endogenous ACh released from the cholinergic terminals innervating the cerebral cortex. The in vitro experiment shows that at a low concentration, THA is not acting as a cholinergic agonist or antagonist on the muscarinic receptors of hippocampal neurons. At higher concentration, THA has a direct, T r x and atropine resistant, depolarizing effect on CA1 hippocampal neurons. This depolarizing effect is accompanied by an increased membrane resistance. Such an effect is consistent with a blockade of an outward potassium current, as also shown by other authors t2'22'25. The other effects (potentiation of the slow cholinergic EPSP, depression of the AHP) are consistent with a potentiation of the effect of synaptically released ACh. The present results are in full agreement with those of Stevens and Cotman 25 showing that THA causes depolarization and increases input resistance but does not affect spike accommodation. We did not see, however, any significant increase in action potential duration.
39 Again in agreement with these authors, we confirm that this effect of T H A is atropine and TI'X resistant, strongly suggesting a direct, non-cholinergic excitatory effect on CA1 hippocampal pyramidal neurons. We observed that T H A is able to selectively block the slow IPSP. This IPSP is mediated by G A B A acting on G A B A B receptor subtype, and involves activation of a potassium conductance (see references in ref. 18). Our results suggest that T H A blocks the potassium channels opened by G A B A B receptor activation. This is in agreement with the results of Halliwell and Grove 12 who show that T H A blocks a neurotransmitter-generated inhibitory potassium conductance (triggered by 5-HT, adenosine and baclofen). It has been demonstrated that the adenosine, the 5-HT and the G A B A B receptors, indeed, share the same potassium conductance (see references in ref. 18). In contrast, the calcium-dependent potassium conductance underlying the slow AHP was not blocked by THA, suggesting a selective action of T H A on the potassium channels. Higher concentrations of T H A also depressed the fast IPSP generated by activation of G A B A A receptors (data not shown). This could be due to an additional inhibitory effect of T H A on the G A B A release. Indeed, a dose-dependent inhibition of the potassium-evoked release of GABA has been demonstrated in the cerebral cortex8. We could not find any evidence for the temporary blockade of spike generation described by Rogawski 22 in hippocampal cultures. We also could not find any evidence for an antagonistic effect of T H A on muscarinic REFERENCES 1 Albin, R.L., Young, A.B. and Penney, J.B., Tetrahydro9-aminoacridine (THA) interacts with the phencylidine (PCP) receptor site, Neurosci. Lett., 88 (1988) 303-307. 2 Bartus, R.T. and Dean, R.L., Tetrahydroaminoacridine, 3, 4-diaminopyridine and physostigmine: direct comparison of effects on memory in aged primates, Neurobiol. Aging, 9 (1988) 351-356. 3 Bartus, R.T., Dean, R.L., Beer, B. and Lippa, A.S., The cholinergic hypothesis of geriatric memory dysfunction, Science, 217 (1982) 408-417. 4 Bassant, M.H., Baleyte, J.M. and I.amour, Y., Electrophysiological and microiontophoretic studies of single cortical somatosensory neurons in the unanesthetized rat, Neuroscience, in press. 5 Buyukuysal, R.L. and Wurtman, R.J., Tetrahydroaminoacridine but not 4-aminopyridine inhibits high-affinitycholine uptake in striatal and hippocampal synaptosomes, Brain Research, 482 (1989) 371-375. 6 Cole, A.E. and Nicoll, R.A., Characterization of a slow cholinergic post-synaptic potential recorded in vitro from rat hippocampal pyramidal cells, J. Physiol., 352 (1984) 173-188. 7 Damsma, G., Biessels, P.T.M., Westerink, B.H.C., De 'Cries, J.B. and Horn, A.S., Differential effects of 4-aminopyridineand 2,4-diaminopyridine on the in vivo release of acetylcholine and dopamine in freely moving rats measured by intrastriatal dialysis, Eur. J. Pharmacol., 145 (1988) 15-20. 8 De Belleroche, J. and Gardiner, I.M., Inhibitory effect of
receptors, nor any inhibitory effect on acetylcholine release. Therefore, the functional significance of the observations reported by several other authors is unclearS,9,11,19,21. However, an indirect effect of T H A on ACh release (through an increase in the inhibitory ACh-mediated feedback on presynaptic receptors) cannot be excluded. Since the potentiation of the slow EPSP and the depolarizing effect were observed in the same concentration range, both might contribute to the therapeutic effect of THA. However the actual concentration of T H A in the brain during clinical trials in human subjects is supposed to be below the micromolar range 2°. Therefore the relevance of the various experimental findings to the clinical effects is difficult to assess. It is likely that the anticholinesterase effect might be operating at low concentrations, but the role of the atropine-resistant depolarization is less clear. It is interesting to note that physostigmin¢ seems to produce more reliable effects than T H A in age-related memory impairment in monkeys 3. Our results also show that physostigmine has no depolarizing effect on CA1 pyramidal neurons comparable to that of THA. They also show that the potentiation of the slow cholinergic EPSP by physostigmine is much more efficient and occurs at much lower concentrations than for THA. These results show that THA: (1) has more complex effects than physostigmine on central neurons; and (2) is less powerful as an anticholinesterase drug than physostigmine. 1,2,3,4-tetrahydro-9-aminocridineon the depolarization-induced release of GABA from cerebral cortex, Br. J. Pharmacol., 94 (1988) 1017-1019. 9 Drukarch, B., Kits, K.S., Van Der Meer, E.G., Lodder, J.C. and Stoof, J.C., 9-Amino-l,2,3,4-tetrahydroacridine (THA), an alleged drug for the treatment of Alzheimer's disease, inhibits acetylcholinesteraseactivity and slow outward K current, Eur. J. Pharmacol., 141 (1987) 153-157. 10 Drukarch, B., Leysen, J.E. and Stoof, J.C., Further analysis of the neuropharmacological profile of 9-amino-l,2,3,4-tetrahydroacridine (THA), an alleged drug for the treatment of Alzheimer's disease, Life Sci., 42 (1988) 1011-1017. 11 Freeman, S.E., Lau, W.M. and Szilagyi, M., Blockade of a cardiac K+ channel by tacrine: interactions with muscarinic and adenosine receptors, Eur. J. Pharmacol., 154 (1988) 59-65. 12 Halliwell, J.V. and Grove, E.A., 9-Amino-l,2,3,4-tetrahydroacridine (THA) blocks agonist-inducedpotassium conductance in rat hippocampal neurones, Eur. J. Pharmacol., 163 (1989) 369-372. 13 Heilbronn, E., Inhibition of chofinesterase by tetrahydroaminacrine, Acta Chem. Scand., 15 (1961) 1386-1390. 14 Ho, A.K.S. and Freeman, S.E., Antichofinesterase activity of tetrahydroaminacdne and succinylcholine hydrolysis, Nature, 205 (1965) 1118-1119. 15 Hollander, E., Mohs, R.C. and Davis, K.I., Cholinergic approaches to the treatment of Alzheimer's disease, Br. Med. Bull., 42 (1986) 97-100. 16 Lamour, Y., Dutar, P. and Jobert, A., Excitatory effect of acetylcholine on different types of neurons in the first somato-
40 sensory neocortex of the rat: laminar distribution and pharmacological characteristics, Neuroscience, 7 (1982) 1483-1494. 17 McCormick, D.A. and Prince, D.A., Two types of muscarinic response to acetylcholine in mammalian cortical neurons, Proc. Natl. Acad. Sci. U.S.A., 82 (1985) 6344-6348. 18 Nicoll, R.A. and Dutar, P., Physiological roles of GABA A and GABA B receptors in synaptic transmission in the hippocampus. In E.A. Barnard and E. Costa (Eds.), Allosteric Modulation of Amino Acid Receptors: Therapeutic Implications, Raven, New York, 1989, pp. 195-204. 19 Nilsson, L., Adem, A., Hardy, J., Winblad, B. and Nordberg, A., Do tetrahydroaminoacridine (THA) and physostigmine restore acetylcholine release in Alzheimer brains via nicotinic receptors?, J. Neural Transm., 70 (1987) 357-368. 20 Park, T.H., Tachiki, K.H., Summers, W.K., Kling, K., Fitten, J., Perryman, K., Spidell, K. and Kling, A.S., Isolation and the fluorometric high performance liquid chromatographic determination of tacrine, Anal. Biochem., 159 (1986) 358-362. 21 Pearce, B.D. and Potter, L.T., Effects of tetrahydroaminoacridine on M 1 and M2 muscarine receptors, Neurosci. Lett., 88 (1988) 281-285.
22 Rogawski, M.A., Tetrahydroaminoacridine blocks voltage-dependent ion channels in hippocampal neurons, Eur. J. Pharmacol., 142 (1987) 169-172. 23 Soni, N. and Kam, P., 4-Aminopyridine - - a review, Anesth. Intens. Care, 10 (1982) 120-126. 24 Stern, Y., Sano, M. and Mayeux, R., Effect of oral physostigmine in Alzheimer's disease, Ann. Neurol., 22 (1987) 306-310. 25 Stevens, D.R. and Cotman, C.W., Excitatory actions of tetrahydro-9-aminoacridine (THA) on hippocampal pyramidal neurons, Neurosci. Lett., 79 (1987) 301-305. 26 Summers, W.K., Majovski, L.V., Marsh, G.M., Tachiki, K. and Kling, A., Oral tetrahydroaminocridine in long term treatment of senile dementia, Alzheimer type, New Engl. J. Med., 315 (1986) 1241-1245. 27 Thesleff, S., Aminopyridine and synaptic transmission, Neuroscience, 5 (1980) 1413-1419. 28 Wesseling, H., Agoston, S., Van Dam, G.B.P., Pasma, J., De Witt, D.J. and Havinga, J., Effects of 4-aminopyridine in elderly patients with Aizheimer's disease, New Engl. J. Med., 310 (1984) 988-989.
Brain Research, 527 (1990) 32-40 Elsevier
BRES 15818
Effects of tetrahydro-9-aminoacridine on cortical and hippocampal neurons in the rat: an in vivo and in vitro study P. Dutar, M.H. Bassant and Y. Lamour I.N.S.E.R.M., UniM 161, Paris (France)
(Accepted 6 March 1990) Key words: Tetrahydroaminoacridine; Physostigmine; Acetylcholine; Cerebral cortex; Hippocampus; Rat
The effects of tetrahydro-9-aminoacridine (THA), an anticholinesterase drug, have been studied in the rat both in vivo (cerebral cortex) and in vitro (CA1 field of the hippocampus) and compared with those of physostigmine. In the cerebral cortex THA potentiated the excitatory effect of acetylcholine in most neurons, including cortical neurons recorded from chronic unanesthetized animals. In vitro, THA (but not physostigmine) had a depolarizing, atropine- and tetrodotoxin-insensitive effect. This effect is associated with an increase in membrane resistance which suggests a direct effect of THA on hippocampal neurons. In addition THA blocked the slow inhibitory postsynaptic potential. At the same concentration THA potentiated the slow cholinergic excitatory postsynaptic potential produced by electrical stimulation of the cholinergic afferents. Its potency was, however, about 10 times lower than that of physostigmine. These results show that THA: (1) is an anticholinesterase much less potent than physostigmine; but (2) has also direct effects on central neurons, not observed with physostigmine and unrelated to its anticholinesterase activity.
INTRODUCTION A deficit of the central cholinergic system has been found in dementia of the Alzheimer type (SDAT) and might also occur with aging 2. One of the therapeutic approaches for improving cognitive functions in patients with S D A T is to act pharmacologically at the level of the central cholinergic synapse. O n e possibility is to administer anticholinesterase agents which increase the efficacy of cholinergic transmission by protecting A C h from being degraded. The cholinesterase inhibitor physostigmine has been extensively used but only modest improvements have been observed 15'24. Recently, it has been claimed that tetrahydro-9-aminoacridine ( T H A ) , a potent centrally acting anticholinesterase 13'a4 produces improvement of cognitive functions in Alzheimer's patients 26. Since other cholinesterase inhibitors (such as physostigmine) had only limited effects, it is possible that T H A may act in a different way to produce its clinical effects. It has been recently suggested that T H A may act in a way similar to 4-aminopyridine (4-AP). This monoamino derivative of pyridine has only weak anticholinesterase properties but blocks potassium conductances in excitable tissues and thereby increases the release of the neurotransmitter, particularly at the neuromuscular junction 23,27 and in central structures 7. Interestingly, 4-AP has also met with some success in the treatment of
S D A T 28. Some authors have suggested that, like 4-AP, T H A could block potassium channels in the central nervous system 22'25. In the present study, we describe the effects of T H A on two cerebral structures known to receive a dense cholinergic innervation, the cerebral cortex and the hippocampus. We compare these effects with those of the well known cholinesterase inhibitor physostigmine (esefine). We used two different approaches: first, we recorded extracellulady froni cortical neurons in vivo in awake restrained or anesthetized rats; and second, we performed intracellular recordings from hippocampal pyramidal neurons in the in vitro slice preparation.
MATERIALS AND METHODS In vivo recordings Most of the methods have been previously published 4'16. In brief, male Sprague-Dawley rats (380-420 g) were anesthetized with pentobarbital (45 mg/kg i.p.) and a chronic restraint device was implanted. The skull was trephined over the first somatosensory cortex (SI). However, a thin layer of bone was left in place to protect the dura. After a period of recovery the animal was habituated to the recording apparatus for 8-10 days. During repetitive trials of increasing duration, the head was painlessly secured to the stereotaxic frame. The rat was comfortably supported by a hammock. The state of consciousness was monitored by observation of the ECoG. Slow-wave sleep and paradoxical sleep were routinely observed, as well as period of quiet wakefulness, confirming that the restraint was well tolerated.
Correspondence: P. Dutar, I.N.S.E.R.M., Unit6 161, 2 rue d'Al6sia, 75014 Paris, France.
0006-8993/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
33 For the recording session, the remaining layer of bone was gently removed and the dura was opened under a stereomicroscope. There was no need to use a local anesthetic for this procedure. Vertical penetrations were made in the part of the SI area containing neuronal aggregates in layer IV, in the fore- or hind-limb representation. Conventional amplification methods were used to record extracellular activity using micropipettes (filled with 1 M NaCI and 2% Pontamine sky blue) attached to a multibarreled micropipette. Solutions for testing by iontophoresis included: sodium chloride (for current compensation, 165 raM), sodium L-glutamate (0.2 M; pH 8), acetylcholine chloride (0.2 M, pH 4), carbamyl choline chloride (earbachol 0.2 M, pH 4), 7-aminohutyric acid (0.1 M, pH 5) physostigmine liemisulfate (0.2 M, pH 5), all purchased from Sigma, and tetrahydro-9-aminoacridine (0.2 M, pH 5), purchased from Aldrich (U.S.A.). Spontaneously, as well as non-spontaneously active neurons (identified by the application of glutamate) were studied. Data were stored on a rectilinear strip chart recorder and on a computer for subsequent analysis using interspike interval histograms (IIH). Standard ACh and THA application were 60-80 nA and 80-100 nA for 24 s, respectively. At the end of each track, a dye deposit was
ACh 60
THA 100
made. The recording session lasted from 3 to 8 h. At the end of the recording session, a reward was given to the rat which was returned to his cage. The rat quickly engaged in his usual activities (grooming, feeding) without signs of discomfort. The experiment lasted for 4 consecutive recording sessions (4 days). Then the rat was deeply anesthetized with pentobarbital and perfused through the heart with saline followed by 10% formaldehyde in saline. Examination of the digestive tracts of the animals never revealed any ulcers. Frozen coronal brain sections were stained with safranin. Electrode penetrations were reconstructed on camera lucida drawings of these sections. The position of each neuron recorded was known from the reading of the micrometer gauge of the mierodrive (+1/~m). It was thus possible to know the laminar position of each neuron. The methods used in the anesthetized rat were similar except that the animal received urethane (1.5 g/kg i.p.) and no restraining device was implanted.
In vitro recordings All in vitro experiments were performed on hippocampal slices. Male Sprague-Dawley rats (150o250 g) were anesthetized with
ACh 60
ACh 60
A
P~V~60 ACh 100 m
ACh 100 m
THAIO0 ACh 100 m
ACh 100 m
B
ACh 100
Phy8o 60 m
ACh 100 m
ACh 100 m
ACh 100 m
C E20 ACh 100 m
D
THA 100 m
CARB
,o
ACh 100 m
ACh 100
THA
CARB
,oo _,o
,,, Imin
ACh
_,oo
THA
_,oo
ACh
_1oo
Fig. 1. Effects of THA and physostitnnine on acetyicholine-inducedneuronal responses in the somatic sensory cortex, recorded from the chronic unanesthetized rat. (Iontophoretic applications, current in nA, vertical calibration bar: 20 impulses/s.) A: layer V neuron. Effect of THA applied before ACh. Eight minutes after application, the effect of THA is not fully dissipated. B,C: layer V and VI neurons, respectively. Notice that in both cases the effect of physostigmine is more impressive than that of THA in spite of the lower amounts of current applied (C: continuous recording from one neuron). D: layer V neuron. The excitation induced by carbachol is not potentiated by THA while on the same neuron THA potentiated the ACh response.
34 halothane and decapitated. The hippocampus was quickly removed and placed in a cold medium. Slices (400/~m thick) were cut and placed in a holding chamber for at least 1 h. A single slice was then transferred to the recording chamber and held between two nylon nets, submerged beneath a continuously superfusing medium that had been warmed and pregassed with 95% O z and 5% CO 2. The composition of the medium was (mM): NaCI, 119; KC1, 2.5; MgSO4, 1.3; CaCI2, 2.5; NaHCO 3, 26.2; NaH2PO 4, 1.0; glucose, 11. The temperature of the medium was maintained between 29 and 31 °C. Drugs were dissolved directly in the medium just before use and were applied by addition to the superfusing medium. Drugs included: carbachol, physostigmine hemisulfate, atropine sulfate and tetrodotoxin, all purchased from Sigma; THA was obtained from IPSEN (France) or Aldrich (U.S.A.). Conventional intracellular recordings from CA1 pyramidal neurons were obtained using glass micropipettes filled with 2 M potassium acetate and having resistances of 60-140 Mr2. A bridge circuit (Axoclamp 2A) was used to apply current pulses for the measurement of membrane resistance and for the generation of afterhyperpolarizations (AHPs) which follow a train of spikes. Intracellular voltage and current recordings were stored on a rectilinear strip chart recorder (Gould). Fast events such as fast excitatory postsynaptic potentials (EPSPs) or inhibitory postsynaptic potentials (IPSPs) and the responses to depolarizing pulses recorded to evaluate the accommodation of the spike discharge (see Fig. 3) were stored on a digital oscilloscope (Nicolet) and plotted on a digital plotter (Hewlett-Packard). Slow cholinergic EPSPs (sEPSPs) were elicited by electrical stimulation of the cholinergic fibers located in the stratum oriens (repetitive stimulation 20-30 Hz for 0.5 s, 15-50 V). Stimulation pulses were applied between the two poles of the bipolar electrode, one pole being inserted into the slice and the other pole being just above the slice.
ACh 20
ACh 20
RESULTS
In vivo recordings U n a n e s t h e t i z e d rat. T h e effect of T H A has been studied on 46 neurons. T H A (100 n A , 24 s) had no effect on most (87%) cortical n e u r o n s when a p p l i e d alone. A weak excitation was o b s e r v e d in 4 neurons and an inhibition in two. The effect of T H A on the n e u r o n a l excitation induced by A C h has been studied in 37 neurons. T H A was applied 30 s before A C h ( T H A 80-100 n A , 24 s; A C h 60-80 n A , 24 s). Most neurons studied (n = 31) were located in infragranular layers. A p o t e n t i a t i o n of the excitation induced by A C h was o b s e r v e d in 28 neurons (76%). In most cases the p o t e n t i a t i o n of the duration of the excitation was m o r e striking than the p o t e n t i a t i o n of the a m p l i t u d e of the response (although both could h a p p e n together, see Fig. 1). The effect of the previous T H A application was p r e s e n t for a long time (5-11 min), but was, nevertheless, reversible (Fig. 1). The p a t t e r n of the n e u r o n a l discharge during the A C h - i n d u c e d excitation was not m o d i f i e d by the previous application of T H A . T h e excitation induced by carbachol (a p o t e n t cholinergic agonist which is not b r o k e n down by acetylcholinesterase) was not p o t e n t i a t e d by T H A (n = 7, Fig. 1D). Anesthetized animals. Twenty-three neurons were re-
THA ACh 50 20
ACh 20
ACh 20
A 5 rain
ACh 80
Physo 80
ACh 80
ACh 80
B1 1 min
ACh 80
ACh
Physo 60
ACh 80
20 min
Fig. 2. Same conventions as in Fig. 1. Urethane-anesthetized rat. Layer V neurons. Notice that in A and B, an inhibitory component of the neuronal response to ACh is potentiated by THA (A) or by pbysostigmine (B). B1 and B2 are continuous recordings from the same neuron.
35
A1 CONT.
2
THA 20pM
3
lmin
el
._I
4 C A R B 5pM
WASH
ls
2
l._ l lnA-
3
._J
_]
I
L_
100ms
C
THA ,501JM
~ s
3.,I Fig. 3. Effects of low concentration of THA on membrane properties of a CA1 pyramidal neuron. AI: the afterhyperpolafization (AHP) is elicited by a brief (60 ms) depolarizing pulse. A2: constant current hyperpolarizing pulses are applied at 0.5 Hz to measure the membrane resistance; THA (20/~M) is applied in the superfusion bath during the period indicated by the horizontal bar. At this concentration, THA does not induce any change in the membrane polarization but induces a small decrease in the AHP amplitude probably resulting from its anticholinesterase effect. A 3" the size of the AHP recovers after washing out THA from the bath. A4: carbachol (5/~M) blocked the AHP. BI: accommodation of the spike discharge induced by application of a long duration depolarizing pulse. THA does not modify accommodation (B2), in contrast with carbachoi (B3). Recordings in A and B are from the same neuron, membrane potential = -58 mV. C: top, effect of THA on the synaptic events induced by stimulation of Schaffer collateraYcommissuralfibers. Left, control trace; middle, THA (50/~M) selectively blocks the slow IPSP but does not affect the EPSP and the fast IPSP; fight, the slow IPSP recovers after washing out THA. Bottom, THA (50/~M) superfused for 30 rain does not affect the size or the shape of the spike.
corded from rats under urethane anesthesia, most of them from layer V. A potentiation of the response to ACh was observed in 17 neurons, and a suppression of the response to A C h in 3 cases. No effect was observed in 3 additional cases. These results differ from those obtained in unanesthetized animals on the following points. (1) A suppression of the response to ACh had never been observed in the absence of anesthesia. (2) Among the 17 responses to ACh potentiated by T H A , a potentiation of the neuronal excitation was observed in 11 cases. However, in the other 6 cases, a potentiation of an inhibitory component of the response was observed (Fig. 2). The inhibitory component could be present before T H A applications (biphasic responses
to ACh, Fig. 2) or only after T H A applications. A similar phenomenon could be observed with physostigmine (Fig.
2). A comparison of the potency of T H A and physostigmine was made in 20 neurons. In some but not all cases physostigmine was more potent than T H A (Fig. 1).
In vitro recordings Stable intracellular recordings were obtained from 43 hippocampal neurons located in the pyramidal layer of the CA1 region. The following neuronal properties were studied: (1) the membrane potential; (2) the A H P that follows a train of action potentials and is due to an increase in a calcium-dependent potassium conductance; (3) the accommodation of the spike discharge; (4) the
36
Ba._.c Fig. 4. THA blocks the baclofen-induced hyperpolarization. Top: left, EPSP and biphasic IPSP are triggered by stimulation of afferent fibers (arrow). Right, baclofen (B, 40 /aM) is applied in the superfusion medium for the time indicated by the horizontal symbol. It induced an hyperpolarization and a decrease in membrane resistance. Bottom: left, 20 min after the beginning of the superfusion with THA (50/~M), the slow IPSP is blocked. Right, in addition, the postsynaptic hyperpolarizing response to baclofen is also blocked.
CONTROL
THA 50pM
A
CONTROL
~
l iota v
500ms
I rain
A C h from these terminals on the p y r a m i d a l neurons 6. T H A , at 5 - 2 0 / ~ M a p p l i e d in the superfusion m e d i u m (n = 19) did not d e p o l a r i z e the n e u r o n s and did not
WASH
THA 7 0 p M
CONTROL
Ba._.c
.~
(20min) E P S P and IPSP triggered by electrical stimulation of Schaffer collateral/commissural fibers (the IPSPs are subdivided into a fast IPSP m e d i a t e d by G A B A acting on G A B A A r e c e p t o r s and a slow IPSP m e d i a t e d by G A B A acting on G A B A a receptors (see ref. 18); and (5) the sEPSP, a depolarizing event elicited by electrical stimulation of cholinergic afferents and due to the release of
.~
ATROPINE 2pM
PHYSO 5~JM
-
30sec
B
CONTROL
THA 51JM
PHYSO 5uM
15' r .......
"=-~
4/-
~"
r .....
~-
"//" ....ilumv 30s
Fig. 5. Effects of THA on the slow cholinerglc EPSP. A: top left, the slow cholinerglc excitatory postsynaptic potential (sEPSP) is eficited by stimulating the cholinerglc fibers in stratum oriens. Triangles indicate the stimulating pulses (20 V, 25 Hz, 0.5 s). Stimulation artefact is followed by a fast hyperpolarizing phase and then by the sEPSP (relatively small in the absence of cholinesterase inhibitor). THA (70/~M), is then applied in the supeffusion medium and 15 min later the sEPSP is markedly increased. After washing out THA, physostigmine (5/~M) dramatically increases the sEPSP; atropine (2/~M) antagonizes this depolarizing potential. Notice that physostigmine is much more effective than THA in increasing the size of the sEPSP and that its effect is more prolonged. Resting membrane potential, -54 inV. B: on this other neuron, the effects of THA and physostigmine on the sEPSP are compared at the same concentration (5/~M). Left: control trace (15 V; 25 Hz, 0.5 s). Middle trace: THA, applied in the superfusing medium does not induce any significant enhancement in the sEPSP size. Right: THA is then washed for 15 min and physostigmine is applied at the same concentration. Physostigmine strongly potentiates the sEPSP which reached the level for action potential generation. Resting membrane potential: -65 inV.
37
A,
THA 0.SmM
CONT
.•OmV lmln
2 TTX
3 ATROP
B,
CONT
3
~m,~v~ •
-eo
-eo
i (~AI
* (.AI
-[
";
~
ATROP
v.i.v~
i (ha)
r
-izo
-140
C
~
.-vm
PHYSO 0.3raM
Fig. 6. Effect of high concentration of THA on a CA1 pyramidal neuron. THA (0.5 mM) induces a slow depolarization of long duration (A 0 associated with a small increase in membrane input resistance. B 1, plot of current-voltage relationship (from records illustrated in inset) showing the voltage responses of the neuron to hyperpolarizing current pulses of variable intensity before (0) and during (O) THA appfication; the increase in the slope of the I N curve indicates an increase in the membrane resistance. The long-duration depolarizing effect of THA persists in the presence of TTX (1/~M) (A2) and is still associated with a small increase in the membrane resistance (B2) indicating that this effect does not result from a presynaptic effect of the THA releasing an excitatory neurotransmitter. The depolarizing effect (A3) and the increase in the input resistance (B3) are not antagonized by atropine (2/~M) applied by superfusion in the bath for 10 rain before THA application) indicating that THA is not acting through muscarinic receptors. The slope values for all I/V curves were calculated by linear regression analysis (R values ranging from 0.97 to 0.99). Calibrations in A: horizontal, 1 rain; vertical, 10 mV. All recordings from the same neuron; resting membrane potential = -65 mV. C: physostigmine applied at high concentration (0.3 mM) in the presence of TTX (1/zM) fails to induce any depolarization and change in membrane resistance. Resting membrane potential = -66 inV.
produce any potentiation of the slow EPSP. In 4 neurons a m o n g 14 tested, this concentration of T H A induced a small decrease in the size of the A H P (Fig. 3). Interestingly, the depression of the A H P did not depend on the concentration of T H A applied. This effect might depend on the level of activity of the cholinergic afferents impinging on the pyramidal neurons. However, at this concentration, T H A did not affect the accommodation of the spike discharge (Fig. 3). In contrast, only 1 - 5 / ~ M carbachol were necessary to block the A H P and to inhibit the accommodation of the spike discharge in the same neurons (n = 5) (Fig. 3). Moreover, carbachol never induced inhibitory responses as in the cortex. Increasing the concentration of T H A (50-70 /~m) produced more obvious effects: T H A induced a small atropine-resistant depolarization with an increase in m e m b r a n e resistance, and a potentiation of the atropine-
sensitive cholinergic sEPSP (Fig. 5). In addition, at this concentration, T H A selectively blocked the slow IPSP in 7 neurons out of 9 (Figs. 3 and 4). A m o n g these 7 neurons, 6 had their fast IPSP unaffected. In two neurons, the G A B A B agonist baclofen was added to the bath (Fig. 4). T H A blocked the baclofen-induced hyperpolarization in these two neurons as well as the slPSP. The effects on the EPSP were not consistent. They were generally widened but this could be due to the loss of the slow inhibitory potential. The size and the s h a p e of action potential were not altered (n = 8, Fig. 3C). Higher concentrations (0.1-1 m M ) of T H A were applied on 16 neurons: we observed a depolarizing effect of T H A in 12 neurons and no effect in the remaining 4. This depolarization was slow in onset and prolonged (Fig. 6), was not antagonized by atropine (n = 4; Fig. 6), and persisted in the presence of tetrodotoxin (n = 2, Fig. 6)
38 suggesting a direct effect, independent of a presynaptic release of ACh. The AHP was studied in 11 neurons: it was slightly depressed in 4, unaffected in 3 and its duration was prolonged in the remaining 4. This latter effect was observed only at high concentrations of THA (>0.5 mM). The small depression of the AHP observed in 4 neurons was not larger than the one observed when using low concentrations of THA. We compared the effects of THA and physostigmine on the sEPSP first at low concentrations (n = 3) and then at higher concentrations (0.3 mM, n -- 5). As illustrated in Figs. 5 and 6, physostigmine was much more potent in enhancing the cholinergic response. The effect induced by physostigmine was obtained at a concentration 10-fold smaller than that of THA and was more prolonged (Fig. 5). At higher concentrations, a depolarizing effect similar to the one induced by THA was never observed with physostigmine (Fig. 6). DISCUSSION The recent publication of a dramatic improvement in cognitive functions in Alzheimer patients receiving THA triggered a surge of interest for this substance. A significant number of papers describing the pharmacol, ogy of THA has been recently published. The picture emerging from these data is still quite confusing. THA has been shown to have weak agonist properties at the PCP binding site I and antagonistic properties at the muscarinic binding site 21. It has also been shown to inhibit high affinity choline uptake and ACh release 5, evoked GABA release 8 and monoamine uptake 1°. Our results can be summarized as follows. (1) THA applied by microiontophoresis on cortical neurons potentiates the excitation induced by ACh. The duration of the excitation is more dramatically potentiated than the amplitude (i.e. the intensity) of the excitation. This effect of THA is more clear-cut in the absence of anesthesia. In the presence of anesthesia (urethane), THA is also able to potentiate inhibitions induced by ACh. (2) The effects observed in vitro were clearly dose dependent. A potentiation of the sEPSP was observed at a concentration of THA higher than 50 /~M. THA behaved neither as an agonist, nor an antagonist of ACh. It has also a direct, atropine resistant, postsynaptic effect on CA1 pyramidal neurons (increase in resistance, depolarization) and blocks the slPSP. The comparison with physostigmine shows that: (1) THA is about 10 times less potent in potentiating the slow EPSP; and (2) physostigmine does not have the direct postsynaptic effects observed with THA. The present results provide evidence that THA be-
haves as a cholinesterase inhibitor: it potentiates the effect of synaptically released ACh (in vitro experiment) as well as the effect of exogenous ACh (in vivo experiment). The in vivo experiment also shows that this effect is specific for ACh: the effect of carbachol (which is not destroyed by ACHE), is not potentiated by THA. The fact that in vivo THA potentiates more the duration than the amplitude of the ACh-induced excitation suggests that THA would allow the exogenous ACh to occupy the receptors for a longer period of time, but would not induce a stronger effect on firing rate. The potentiation by THA of ACh-induced inhibitions observed under urethane anesthesia is more difficult to explain, since such a phenomenon was not observed in unanesthetized animals. It is possible that in the presence of anesthesia the activity of the ACh-sensitive inhibitory cortical interneurons is higher. Then the effect of ACh, when applied by iontophoresis, would preferentially activate these interneurons, resulting in the inhibition of the pyramidal neurons (which are the most likely to be recorded under our experimental conditions). The fact that inhibitory cortical interneurons are excited by ACh has been clearly demonstrated in vitro 17. The preferential excitation of these interneurons would explain the 'mixed' effect of iontophoretic ACh on the principal neurons. Since this effect of ACh is likely to be potentiated by THA, it would explain why the inhibitory responses are the most potentiated. THA had little effect on the neurons in the absence of ACh. A weak excitation was observed in only 8% of them. This weak excitation is not similar to the depolarization observed in vitro because the time course of these effects is quite different. Therefore it could be due to the potentiation by THA of the excitatory effect of the endogenous ACh released from the cholinergic terminals innervating the cerebral cortex. The in vitro experiment shows that at a low concentration, THA is not acting as a cholinergic agonist or antagonist on the muscarinic receptors of hippocampal neurons. At higher concentration, THA has a direct, T r x and atropine resistant, depolarizing effect on CA1 hippocampal neurons. This depolarizing effect is accompanied by an increased membrane resistance. Such an effect is consistent with a blockade of an outward potassium current, as also shown by other authors t2'22'25. The other effects (potentiation of the slow cholinergic EPSP, depression of the AHP) are consistent with a potentiation of the effect of synaptically released ACh. The present results are in full agreement with those of Stevens and Cotman 25 showing that THA causes depolarization and increases input resistance but does not affect spike accommodation. We did not see, however, any significant increase in action potential duration.
39 Again in agreement with these authors, we confirm that this effect of T H A is atropine and TI'X resistant, strongly suggesting a direct, non-cholinergic excitatory effect on CA1 hippocampal pyramidal neurons. We observed that T H A is able to selectively block the slow IPSP. This IPSP is mediated by G A B A acting on G A B A B receptor subtype, and involves activation of a potassium conductance (see references in ref. 18). Our results suggest that T H A blocks the potassium channels opened by G A B A B receptor activation. This is in agreement with the results of Halliwell and Grove 12 who show that T H A blocks a neurotransmitter-generated inhibitory potassium conductance (triggered by 5-HT, adenosine and baclofen). It has been demonstrated that the adenosine, the 5-HT and the G A B A B receptors, indeed, share the same potassium conductance (see references in ref. 18). In contrast, the calcium-dependent potassium conductance underlying the slow AHP was not blocked by THA, suggesting a selective action of T H A on the potassium channels. Higher concentrations of T H A also depressed the fast IPSP generated by activation of G A B A A receptors (data not shown). This could be due to an additional inhibitory effect of T H A on the G A B A release. Indeed, a dose-dependent inhibition of the potassium-evoked release of GABA has been demonstrated in the cerebral cortex8. We could not find any evidence for the temporary blockade of spike generation described by Rogawski 22 in hippocampal cultures. We also could not find any evidence for an antagonistic effect of T H A on muscarinic REFERENCES 1 Albin, R.L., Young, A.B. and Penney, J.B., Tetrahydro9-aminoacridine (THA) interacts with the phencylidine (PCP) receptor site, Neurosci. Lett., 88 (1988) 303-307. 2 Bartus, R.T. and Dean, R.L., Tetrahydroaminoacridine, 3, 4-diaminopyridine and physostigmine: direct comparison of effects on memory in aged primates, Neurobiol. Aging, 9 (1988) 351-356. 3 Bartus, R.T., Dean, R.L., Beer, B. and Lippa, A.S., The cholinergic hypothesis of geriatric memory dysfunction, Science, 217 (1982) 408-417. 4 Bassant, M.H., Baleyte, J.M. and I.amour, Y., Electrophysiological and microiontophoretic studies of single cortical somatosensory neurons in the unanesthetized rat, Neuroscience, in press. 5 Buyukuysal, R.L. and Wurtman, R.J., Tetrahydroaminoacridine but not 4-aminopyridine inhibits high-affinitycholine uptake in striatal and hippocampal synaptosomes, Brain Research, 482 (1989) 371-375. 6 Cole, A.E. and Nicoll, R.A., Characterization of a slow cholinergic post-synaptic potential recorded in vitro from rat hippocampal pyramidal cells, J. Physiol., 352 (1984) 173-188. 7 Damsma, G., Biessels, P.T.M., Westerink, B.H.C., De 'Cries, J.B. and Horn, A.S., Differential effects of 4-aminopyridineand 2,4-diaminopyridine on the in vivo release of acetylcholine and dopamine in freely moving rats measured by intrastriatal dialysis, Eur. J. Pharmacol., 145 (1988) 15-20. 8 De Belleroche, J. and Gardiner, I.M., Inhibitory effect of
receptors, nor any inhibitory effect on acetylcholine release. Therefore, the functional significance of the observations reported by several other authors is unclearS,9,11,19,21. However, an indirect effect of T H A on ACh release (through an increase in the inhibitory ACh-mediated feedback on presynaptic receptors) cannot be excluded. Since the potentiation of the slow EPSP and the depolarizing effect were observed in the same concentration range, both might contribute to the therapeutic effect of THA. However the actual concentration of T H A in the brain during clinical trials in human subjects is supposed to be below the micromolar range 2°. Therefore the relevance of the various experimental findings to the clinical effects is difficult to assess. It is likely that the anticholinesterase effect might be operating at low concentrations, but the role of the atropine-resistant depolarization is less clear. It is interesting to note that physostigmin¢ seems to produce more reliable effects than T H A in age-related memory impairment in monkeys 3. Our results also show that physostigmine has no depolarizing effect on CA1 pyramidal neurons comparable to that of THA. They also show that the potentiation of the slow cholinergic EPSP by physostigmine is much more efficient and occurs at much lower concentrations than for THA. These results show that THA: (1) has more complex effects than physostigmine on central neurons; and (2) is less powerful as an anticholinesterase drug than physostigmine. 1,2,3,4-tetrahydro-9-aminocridineon the depolarization-induced release of GABA from cerebral cortex, Br. J. Pharmacol., 94 (1988) 1017-1019. 9 Drukarch, B., Kits, K.S., Van Der Meer, E.G., Lodder, J.C. and Stoof, J.C., 9-Amino-l,2,3,4-tetrahydroacridine (THA), an alleged drug for the treatment of Alzheimer's disease, inhibits acetylcholinesteraseactivity and slow outward K current, Eur. J. Pharmacol., 141 (1987) 153-157. 10 Drukarch, B., Leysen, J.E. and Stoof, J.C., Further analysis of the neuropharmacological profile of 9-amino-l,2,3,4-tetrahydroacridine (THA), an alleged drug for the treatment of Alzheimer's disease, Life Sci., 42 (1988) 1011-1017. 11 Freeman, S.E., Lau, W.M. and Szilagyi, M., Blockade of a cardiac K+ channel by tacrine: interactions with muscarinic and adenosine receptors, Eur. J. Pharmacol., 154 (1988) 59-65. 12 Halliwell, J.V. and Grove, E.A., 9-Amino-l,2,3,4-tetrahydroacridine (THA) blocks agonist-inducedpotassium conductance in rat hippocampal neurones, Eur. J. Pharmacol., 163 (1989) 369-372. 13 Heilbronn, E., Inhibition of chofinesterase by tetrahydroaminacrine, Acta Chem. Scand., 15 (1961) 1386-1390. 14 Ho, A.K.S. and Freeman, S.E., Antichofinesterase activity of tetrahydroaminacdne and succinylcholine hydrolysis, Nature, 205 (1965) 1118-1119. 15 Hollander, E., Mohs, R.C. and Davis, K.I., Cholinergic approaches to the treatment of Alzheimer's disease, Br. Med. Bull., 42 (1986) 97-100. 16 Lamour, Y., Dutar, P. and Jobert, A., Excitatory effect of acetylcholine on different types of neurons in the first somato-
40 sensory neocortex of the rat: laminar distribution and pharmacological characteristics, Neuroscience, 7 (1982) 1483-1494. 17 McCormick, D.A. and Prince, D.A., Two types of muscarinic response to acetylcholine in mammalian cortical neurons, Proc. Natl. Acad. Sci. U.S.A., 82 (1985) 6344-6348. 18 Nicoll, R.A. and Dutar, P., Physiological roles of GABA A and GABA B receptors in synaptic transmission in the hippocampus. In E.A. Barnard and E. Costa (Eds.), Allosteric Modulation of Amino Acid Receptors: Therapeutic Implications, Raven, New York, 1989, pp. 195-204. 19 Nilsson, L., Adem, A., Hardy, J., Winblad, B. and Nordberg, A., Do tetrahydroaminoacridine (THA) and physostigmine restore acetylcholine release in Alzheimer brains via nicotinic receptors?, J. Neural Transm., 70 (1987) 357-368. 20 Park, T.H., Tachiki, K.H., Summers, W.K., Kling, K., Fitten, J., Perryman, K., Spidell, K. and Kling, A.S., Isolation and the fluorometric high performance liquid chromatographic determination of tacrine, Anal. Biochem., 159 (1986) 358-362. 21 Pearce, B.D. and Potter, L.T., Effects of tetrahydroaminoacridine on M 1 and M2 muscarine receptors, Neurosci. Lett., 88 (1988) 281-285.
22 Rogawski, M.A., Tetrahydroaminoacridine blocks voltage-dependent ion channels in hippocampal neurons, Eur. J. Pharmacol., 142 (1987) 169-172. 23 Soni, N. and Kam, P., 4-Aminopyridine - - a review, Anesth. Intens. Care, 10 (1982) 120-126. 24 Stern, Y., Sano, M. and Mayeux, R., Effect of oral physostigmine in Alzheimer's disease, Ann. Neurol., 22 (1987) 306-310. 25 Stevens, D.R. and Cotman, C.W., Excitatory actions of tetrahydro-9-aminoacridine (THA) on hippocampal pyramidal neurons, Neurosci. Lett., 79 (1987) 301-305. 26 Summers, W.K., Majovski, L.V., Marsh, G.M., Tachiki, K. and Kling, A., Oral tetrahydroaminocridine in long term treatment of senile dementia, Alzheimer type, New Engl. J. Med., 315 (1986) 1241-1245. 27 Thesleff, S., Aminopyridine and synaptic transmission, Neuroscience, 5 (1980) 1413-1419. 28 Wesseling, H., Agoston, S., Van Dam, G.B.P., Pasma, J., De Witt, D.J. and Havinga, J., Effects of 4-aminopyridine in elderly patients with Aizheimer's disease, New Engl. J. Med., 310 (1984) 988-989.
