416
Brain Research, 585 (1992) 416-420 •"> 1992 Elsevier Science Publishers B.V. All rights reserved 0006-8993/92/$05.00
BRES ~i250
Differential actions of acetylcholinesterase on the soma and dendrites of dopaminergic substantia nigra neurons in vitro Mihfily Haj6s * and Susan Greenfield Unil'ersity Department of Pharmacology, Oxford (UK) (Accepted 31 March 1992)
Key words: Acetylcholinesterase; Substantia nigra; Intracellular recording; Guinea pig, Dendrite
In the substantia nigra, acetylcholinesterase (ACHE) has non-cholinergic action on dopaminergic neurons. The subset of neurons particularly sensitive to AChE are characterized by functionally active apical dendrites extending into the pars reticulata and generating a powerful calcium conductance. This study thus attempted to establish directly the importance of these dendrites regarding the action of ACHE. Segregation of the pars compacta from the pars reticulata did not affect the AChE-induced hyperpolarization on this sub-set of dopaminergic neurons, However, the ionic basis of the hyperpolarization was related to the integrity of the neurons: AChE caused an opening of potassium channels in intact cells. On the other hand when the pars reticulata containing apical dendrites was removed, an action of AChE involving the closure of calcium/sodium channels was revealed. The results demonstrate that the net effect of AChE need not be related to any particular segment of the dopaminergic neurons, whereas the nature of the mechanism underlying that effect depends on the presence, or otherwise, of the apical dendrites.
A soluble form of acctylcholinesterase (ACHE) is released from the dendrites of dopaminergic neurons within the substantia nigra (SN) :,'~, The released protein subsequently modifies the neuronal activity of the same neurons in a non-esteratic capacity: electrophysiological studies in vitro showed that AChE can hyperpolarize SN dopaminergic neurons and inhibit their spontaneous firing, again independent of esteratic activity ~':~. More recently, a specific sub-set of SN dopaminergic neurons were identified as particularly sensitive to ACHE: these neurons are mostly located within the rostral part of the nucleus at the level of the mammillary body :;), Although complete differential morphological and electrophysiological profiles of subsets of SN dopaminergic neurons have not been estab. lished, some clear differences between these cells have been recognised I,"),~t'. The characteristics of rostrally located dopaminergic neurons appear different to those recorded more caudaP,~: they are electrophysiologically conspicuous by their wider range of spontaneous firing activity, a noa-rhythmic more phasic pattern of
firing, shorter width of action potential and the pros¢nc¢ of a low threshold calcium conductance 12,~4,2", Morphologically, these neurons possess long apical dendrites, running into the pars reticulata '~,~s,~, which arc regarded as the primary location in generation of a low tt.'cshold calcium conductance ",Is. The aim of the present study was to clarify the relative importance of the apical dendrites of SN dopaminergic neurons as regards the effects of ACHE. Thus, intracellular recordings were performed from SN dopaminergic neurons following segregation of the pars compacta and pars reticulata ('SNr-cut') compared to intact control neurons. Mid-brain slices (400 p.m) were obtained from albino guinea pigs of either sex (300-500 g). Recordings were performed from slices cut at the level of the mamillary body, this level is shown to contain dopamine-immunoreactive cell bodies of the A9 group, in the guinea pig I~, Individual slices were placed under a dissection microscope and the pars reticulata and crus cerebri on one side was cut away. The substantiae
* Permanent address: Department of Physiology, A. Szent-Gy6rgyi Medical University, Szeged, Hungary. Correspemh,nce: M. Haj6s, University Department of Pharmacology, Mansfield Rd., Oxford, OX! 3QT, UK.
417 nigrae where the pars compacta and pars reticulata were segregated are denoted as 'SNr-cut'. Slices were maintained for at least 2 h in the following artificial cerebrospinal fluid (acsf) solution (in raM): NaCI 120, KC! 5, NaHCO 3 20, CaC! 2 2, MgSO 4 2, glucose 10, HEPES acid 6.7, HEPES salt 3.3. At the time of recording, slices were transferred to an organ bath and perfused with the following perfusate: (in mM) NaCl 123, KCI 2, NaHCO 3 26, KH2PO 4 1.25, MgSO 4 1.3, CaC! 2 2.4, glucose 10, at a rate of approximately 2.0 ml/min, saturated constantly with carbogen (95% 0 2 and 5% CO 2) and maintained at 33 + 1.0°C. Recordings were made using glass microelectrodes filled with 3 M potassium acetate (series resistances measured in acsf: 50-120 Mn). The electrodes were mounted in micromanipulators and placed with the aid of a dissecting microscope in the pars compacta region of the substantia nigra. Intracellular potential was fed into a Neurodata (IR-283) preamplifier with an active bridge circuit to allow simultaneous measurements of the membrane potential and intracellular injection of constant current through the recording elecUode. Only those neurones that had an apparent resting membrane potential greater or equal to - 4 5 mV and displayed an overshoot action potential were retained for experimentation. Passive and active membrane properties of all cells in the study were recorded as described previously ~L~2'2°. AChE (Sigma, Electric Eel Type VS) was applied in a final concentration of 125 U / m l by adding to the perfusate reservoir and was recycled from the effluent tube back into a closed circuit 2°. The effects of AChE were assessed in terms of modifications to resting potential, firing rate and input resistance. Input resistance was calculated by measuring the voltage drop to square pulses of hyperpolarizing current (0.1-0.5 hA, 200 ms duration). Data were stored on a video recorder and transferred offline using a D / A interface (CED 1401), personal computer (Research Machines, Nimbus VX/2) and employing Sigavg software (CED). All numerical data is expressed as the mean :t: standard error of the mean. For statistical analysis unpaired Student's t-test was used. Following experiments where the electrode had contained biocytin (Sigma), impaled neurons were filled with the dye using depolarizing pulse commands (0.5 nA, 500 ms, 1 Hz) for 5-10 rain. Slices were subjected to a procedure for visualization of the biocytin filled cell and the contralateral side for immunohistochemical staining for tyrosine hydroxylase, as described previously 12,14,20. Tyrosine hydroxylase immunoreactive cells were detectable in the slices cut at the level of the mamillary body of the SN. These cells displayed variety in the
¸
Fig. 1. Coronalsection(100/~m) of guinea-pigmesencephalonat the level of the mammillary bodies, showing tyrosine hydro~iase immunoreactivity in the pars compacta region in intact SN (A) and in SN followingremoval of the pars reticulata (B). C: cell filled with biocytinfrom SNr-cut slice. Scalebar: 100 ~m. shape of the somata and were comparable with the previously decribed forms 3'4'14'19. Moreover, a high number of immunoreactive dendrites of the neurons were observed in the SN pars reticulata, which is especially characteristic of rostrally located dopaminergic neurons 3'14. The morphology of cells filled with biocytin was indistinguishable from that of the neurons positively identified as dopaminergic pars compacta neurons. In SNr-cut slices tyrosine hydroxylase immunoreactive neurons, or neurons filled with biocytin were located at the edge of the sectioning with clear absence of the SN pars reticulata (Fig. 1). A total of 29 cells, 18 control and 11 SNr-cut neurons were recorded and analysed from the rostrai part of the SN. There was no significant difference between the two groups of neurons in their membrane ~otential, spontaneous firing rate or input resistance. In control cells membrane potential was - 51.85 + 2.17 mV (mean and S.E.M.), firing rate 6.58 4- 2.33 Hz and input resistance 120.7 :l: 9.9 M•, whereas the values in the SNr-cut group were -50.88 + 2.88 mV, 2.56 :l: 0.98 Hz and 165 :l: 25.8 M n , respectively. In half of the control neurons, termination of a 200 ms hyperpolariz-
418 ing current pulse resulted in generation of action po. tentials. Similarly, if the membrane potential was held using negative DC current at a hyperpolarized level, additional hyperpolarizing or depolarizing pulses generated a burst of action potentials, described as a 'Low Threshold Spike' (LTSgCa2+) II. By contrast, none of the SNr-cut neurons generated LTS activity, although these cells displayed identical electrophysiologicai characteristics, were recorded from the same level as the intact cells and indeed were frequently in the same slice of midbrain. Application of AChE resulted in cessation of the spontaneous firing and hyperpolarization of the majority of the. recorded cells. Of the control neurons 14 cells responded by diminution of the firing rate and/or hyperpolarization whilst 4 cells were insensitive to AChE (Fig. 2). Of the 11 SNr-cut neurons 8 cells were responsive. The firing rate was significantly reduced both in control (0.33 + 1.9 Hz; P < 0.02) and lesioned cells (0.89 :t: 0.56 Hz; P < 0.05). There was no significant difference either in the AChE.induced hyperpolarization between the two groups of cells: mean values were 5.83 + 1.04 mV and 7.38 + 1.77 mV in control and SNr-cut neurons, respectively, in control cells hyperpolarization of the membrane was associated with a decrease in the input resistance: AChE application reduced the input resistance by 27.2 :l: 7.3 M/~. Analysis of current-voltage relations of neurons indicated a reversal at -88.5:1:9.9 mV, which approximates most closely the equilibrium potential for potassium ions, ( - 100 mV) given K~ - 3.2 mM at 33°C and assuming K , - 140 raM. in SNr-cut neurons, however, AChE induced an increase in input resistance of 85.0 ± 33,2 M/~. Analysis of the current-voltage relation revealed a reversal potential of -28,2 ± 5.8 mV which indicates
significant changes in membrane conductance for other ions beside potassium (Fig. 3). The difference between the reversal potentials for the effect of AChE with and without SNr intact was highly significant (P < 0.001). Slices recorded in this study contained a high proportion of tyrosine-hydroxylase positive cells: furthermore the morphology of these dopaminergic neurons was comparable to those injected with biocytin. It seems most probable therefore that the cells used in this study were indeed dopaminergic. In addition, the morphology of the perikarya and the proximal dendrites of lesioned cells stained for tyrosine-hydroxylase or marked with biocytin strongly resembled the control cells, but clearly indicated the cut of apical dendrites. Electrophysiological characteristics of the cells from SNr-cut slices did not differ significal~tly from control cells, regarding firing rate, action potential width or input resistance, in agreement with previous observations s'14. It seems highly unlikely that the SNr-cut and control cells were not from the same neuronal population. The only difference seen following the sectioning procedure was that none of SNr-cut cells displayed a low-threshold calcium conductance: this finding is in accordance with previous studies localising this conductance in the apical dendrites s.ls. AChE displayed an inhibitory effect on control SN dopaminergic neurons predominantly via an opening of potassium channels, as shown previously 6.7.2°, However, in this study AChE was also found here to be as effective on the SNr.cut cells as on control neurons. This finding suggests that the apical dendrites of SN neurons are not the exclusive target of the protein. However, the mode of the action of AChE appears different at the soma, compared to that at the apical SNr dendrites, AChE appears to hyperpolarize the
AChE ..o°v.,
....
B
AChE
-45 m V ~ . F . . ~ ~ ,
'0mvL. -
.....
5 mln Fig. 2. Recording of membrane potential of a control neuron (A) and a SNr-cut neuron (B) during application of ACHE. These two cells were recorded from each respective side of the same midbrain slice, AChE induced hyperpolarization and reduced firing rate of the recorded neurons in both cases; after termination of AChE application a partial recovery can be observed within 10 min.
419
Control
A
AChE
L
-50 mV . . , i ~
.
.
.
.
.
.
.
.
~
-
. . . .
-_
~
-_':
i20 mV I. . . . . . . . .
..
:_l
j0,2 nA
100 msec
Control
B
AChE
lJ (_
_
_C-'.Z,'
_.__a0.20 nA 100 mseo D
C
nA
nA -0.0
-0.4
-0.|
0:0
-0.6 .4O
-0.4
-0.2
0.0
0.2
.7 • MChl
° ~
[.4o
'4 0
'410
.100
mV .120
-120
Fig. 3. Effects of AChE on membrane properties of control (A) and SNr-cut (B) nigral neurons. These neurons are the same as shown in Fig. 2. A: Traces show characteristic responses of the intact neuron under control conditions, prior to AChE application (left). Note characteristic low-threshold spikes in response to injection of hyperpolarizing current. Following application of AChE (right) there is a hyperpolarization and a fall in input resistance. B: Traces show characteristic responses of the SNr-cut neurons under control conditions. Application of AChE induced hyperpolarization with a marked increase in input resistance. C and D: plots of voltage-current relationships prior to (open circles) and during AChE application (filled circles) in control and SNr-cut neurons, respectively (same cells as in A and B). Each data point is an average of measurements made at 175 ms from 5 successive pulses; lines fitted to data are simple regression lines. The voltage-current relationships intersect at - 102 mV in control cell (C) and at - 22 mV in SNr-cut cell (D).
420 somatic membrane through a net closure of ion channels. However, since the mean reversal potential was still negative, this effect cannot be attributed either solely to Ca 2+ (R.P.: + 132 mV) or Na + (R.P.: +41 mV). The most parsimonious explanation would be that, the predominant closure of Na + or Ca 2+ channels was slightly offset by the modest opening of some somatic CI- and/or K + channels. These findings suggest that AChE can have dual actions on pars compacta neurons, depending on its site of action on the cell. The action of AChE within the pars compacta (at the level of cell body and proximal dendrites) itself seems to involve no single ion channel and indeed would be masked in a physiological situation by the principal effects seen when the dendrites are intact. The results demonstrate that the dominant channel via which AChE acts is located on the apical dendrites. However the net action is deten~ined not by an exclusivity of AChE for any single ion channel since the protein produces effects just as marked at the level of the soma. it is more probable that the distal dendrites are more readily affected since this part of the cell is the most likely origin of released AChE t3. Two conclusions arise from this study, first there are multiple ionic bases for the action of ACHE, and second the apical dendrites could have a critically active role in neuronal functions, distinct from physico-chemical events at the soma. This work was supported by BristoI.Mayers.Squibb Co (USA), We would like to thank Mr. C. Wehb for excellent technical assistance and Mrs. P. Cordery for histological services. I Chtodo, L.A, Antelman, S.M., Caggiula, A.R. and Lineberry, C.G., Sensory stimulation alters the discharge rate of dopamine (DA) neurons: evidence of two functional types of DA cells in the substantia nigra, Brain Res., 189 (1980) 544-549. 2 Chubb, I.W., Goodman, S. and Smith, A.D., is acetylcholinesterase secreted from central neurons into the cerebrospinal fluid? Neuroscience, I (1976) 57-62. 3 Gerfen, C.R., Baimbridge, K.G. and Thibault, J., The neostriatal mosaic: !11. Biochemical and developmental dissociation of patch-matrix mesostriatal systems, .I. Neurosci., 7 (1987) 39353944. 4 Grace, A.A. and Onn, S-P, Morphology and electrophysiologicai properties of immunocytochemically identified rat dopamine neurons recorded in vitro, J. Neurosci., 9 (1989) 3463-3481.
5 Greenfield, S.A., A non-cholinergic action of acetylcholinesterase (ACHE) in the brain: from neuronal secretion to the generation of movement, Cell. Mol. Neurobiol., 11 (1991) 55-77. 6 Greenfield, S.A., Jack, JJ.B., Last, A.TJ. and French, M., An electrophysiological action of acetylcholinesterase independent of its catalytic site, Exp. Brain Res., 70 (1988) 441-444. 7 Greenfield, S.A., Nedergaard, S., Webb, C. and French, M., Pressure ejection of acetylcholinesterase within the subtantia nigra has non-classica| actions on the pars compacta cells independent of selective receptor and ion channel blockade, Neuroscience, 29 (1989) 21-25. 8 Harris, N.C., Ramsay, S., Kelion, A. and Greenfield, S.A., Electrophysiological evidence for a dendritic localization of calcium conductance in guinea pig substantia nigra neurons in vitro, Exp. Brain Res., 74 (1989) 411-416. 9 Kita, T., Kita, H. and Kitai, S.T., Electrical membrane properties of rat substantia nigra compacta neurons in an in vitro slice preparation, Brain Res., 372 (1986) 21-30. 10 Kubota, Y., Kang, Y. and Kitai, S.T., Two types of substantia nigra compacta neurons based on electrophysiological membrane characteristics, Soc. Neurosci. Abstr., 15 (1989) 900P. 11 Llinas, R.R., Greenfield, S.A. and Jahnsen, H., Electrophysiolo~ of pars compacta cells in the in vitro substantia nigra: a possible mechanism for dendritic release, Brain Res., 294 (1984) 127-132. 12 Murphy, K.P.SJ. and Greenfield, S.A., ATP-sensitive potassium channels counteract anoxia in neurons of the substantia nigra, Exp. Brain Res., 84 (1991) 355-358. 13 Nedergaard, S., Bolam, J.P. and Greenfield, S.A., Facilitation of a dendritic calcium conductance by 5-HT in the substantia nigra, Nature, 333 (1988) 174-177. 14 Nedergaard, S. and Greenfield, S.A., Subpopulations of pars compacta neurons in the substantia nigra: the significance of qualitatively and quantitatively distinct conductances, Neuroscience 48 91992) 423-437. 15 Preston, RJ., McRea, R.A., Chang, H.T. and Kitai, S.T., Anatomy and physiology of substantia nigra and retrorubral neurons studied by extra and intracellular recording and by horseradish peroxidase labelling. Na~roscience, 6 (1981) 331-344. 16 Shepard, P.D. and German, D.C., Electrophysiological and pharmacological evidence Ibr the existence of distinct sublmpulations of nigrostrlatal dopamincrgtc neuron in the rat, Ne~rosclt,nce, 27 (1988) 537-546. 17 Smits, R.J.P.M., Steinbusch, H.W.M. and Mulder, A.H., Distribution of dopamine-tmmunoreactive cell bodies in the guinea pig brain. J. Chem. Neuroanat., 3 (1990) 101-123. 18 Walsh, J.P., Cepeda, C, Buchwald, N,A. and l~vine, M.S,, Neurophysiological maturation of cat substantia nigra neurons: evi. dence from in vitro studies, Synapse, 7 (1991) 291-300. 19 Wassef, M., Berod, A. and Sotelo, C., Dopaminergic dendrites in the pars reticulata of the rat substantia nigra and their striatal input. Combined immuno~tochemical localization of tyrosine hydro~lase and anterograde degeneration, Neuroscience, 6 (1981) 2125-2139. 20 Webb, C. and Greenfield, S.A., Non-cholinergic effects of acetylcholinesterase in the substantia nigra: a possible role for an ATP-sensitive potassium channel, Exp. Brain Res. 89 (1992) 49-58.
Brain Research, 585 (1992) 416-420 •"> 1992 Elsevier Science Publishers B.V. All rights reserved 0006-8993/92/$05.00
BRES ~i250
Differential actions of acetylcholinesterase on the soma and dendrites of dopaminergic substantia nigra neurons in vitro Mihfily Haj6s * and Susan Greenfield Unil'ersity Department of Pharmacology, Oxford (UK) (Accepted 31 March 1992)
Key words: Acetylcholinesterase; Substantia nigra; Intracellular recording; Guinea pig, Dendrite
In the substantia nigra, acetylcholinesterase (ACHE) has non-cholinergic action on dopaminergic neurons. The subset of neurons particularly sensitive to AChE are characterized by functionally active apical dendrites extending into the pars reticulata and generating a powerful calcium conductance. This study thus attempted to establish directly the importance of these dendrites regarding the action of ACHE. Segregation of the pars compacta from the pars reticulata did not affect the AChE-induced hyperpolarization on this sub-set of dopaminergic neurons, However, the ionic basis of the hyperpolarization was related to the integrity of the neurons: AChE caused an opening of potassium channels in intact cells. On the other hand when the pars reticulata containing apical dendrites was removed, an action of AChE involving the closure of calcium/sodium channels was revealed. The results demonstrate that the net effect of AChE need not be related to any particular segment of the dopaminergic neurons, whereas the nature of the mechanism underlying that effect depends on the presence, or otherwise, of the apical dendrites.
A soluble form of acctylcholinesterase (ACHE) is released from the dendrites of dopaminergic neurons within the substantia nigra (SN) :,'~, The released protein subsequently modifies the neuronal activity of the same neurons in a non-esteratic capacity: electrophysiological studies in vitro showed that AChE can hyperpolarize SN dopaminergic neurons and inhibit their spontaneous firing, again independent of esteratic activity ~':~. More recently, a specific sub-set of SN dopaminergic neurons were identified as particularly sensitive to ACHE: these neurons are mostly located within the rostral part of the nucleus at the level of the mammillary body :;), Although complete differential morphological and electrophysiological profiles of subsets of SN dopaminergic neurons have not been estab. lished, some clear differences between these cells have been recognised I,"),~t'. The characteristics of rostrally located dopaminergic neurons appear different to those recorded more caudaP,~: they are electrophysiologically conspicuous by their wider range of spontaneous firing activity, a noa-rhythmic more phasic pattern of
firing, shorter width of action potential and the pros¢nc¢ of a low threshold calcium conductance 12,~4,2", Morphologically, these neurons possess long apical dendrites, running into the pars reticulata '~,~s,~, which arc regarded as the primary location in generation of a low tt.'cshold calcium conductance ",Is. The aim of the present study was to clarify the relative importance of the apical dendrites of SN dopaminergic neurons as regards the effects of ACHE. Thus, intracellular recordings were performed from SN dopaminergic neurons following segregation of the pars compacta and pars reticulata ('SNr-cut') compared to intact control neurons. Mid-brain slices (400 p.m) were obtained from albino guinea pigs of either sex (300-500 g). Recordings were performed from slices cut at the level of the mamillary body, this level is shown to contain dopamine-immunoreactive cell bodies of the A9 group, in the guinea pig I~, Individual slices were placed under a dissection microscope and the pars reticulata and crus cerebri on one side was cut away. The substantiae
* Permanent address: Department of Physiology, A. Szent-Gy6rgyi Medical University, Szeged, Hungary. Correspemh,nce: M. Haj6s, University Department of Pharmacology, Mansfield Rd., Oxford, OX! 3QT, UK.
417 nigrae where the pars compacta and pars reticulata were segregated are denoted as 'SNr-cut'. Slices were maintained for at least 2 h in the following artificial cerebrospinal fluid (acsf) solution (in raM): NaCI 120, KC! 5, NaHCO 3 20, CaC! 2 2, MgSO 4 2, glucose 10, HEPES acid 6.7, HEPES salt 3.3. At the time of recording, slices were transferred to an organ bath and perfused with the following perfusate: (in mM) NaCl 123, KCI 2, NaHCO 3 26, KH2PO 4 1.25, MgSO 4 1.3, CaC! 2 2.4, glucose 10, at a rate of approximately 2.0 ml/min, saturated constantly with carbogen (95% 0 2 and 5% CO 2) and maintained at 33 + 1.0°C. Recordings were made using glass microelectrodes filled with 3 M potassium acetate (series resistances measured in acsf: 50-120 Mn). The electrodes were mounted in micromanipulators and placed with the aid of a dissecting microscope in the pars compacta region of the substantia nigra. Intracellular potential was fed into a Neurodata (IR-283) preamplifier with an active bridge circuit to allow simultaneous measurements of the membrane potential and intracellular injection of constant current through the recording elecUode. Only those neurones that had an apparent resting membrane potential greater or equal to - 4 5 mV and displayed an overshoot action potential were retained for experimentation. Passive and active membrane properties of all cells in the study were recorded as described previously ~L~2'2°. AChE (Sigma, Electric Eel Type VS) was applied in a final concentration of 125 U / m l by adding to the perfusate reservoir and was recycled from the effluent tube back into a closed circuit 2°. The effects of AChE were assessed in terms of modifications to resting potential, firing rate and input resistance. Input resistance was calculated by measuring the voltage drop to square pulses of hyperpolarizing current (0.1-0.5 hA, 200 ms duration). Data were stored on a video recorder and transferred offline using a D / A interface (CED 1401), personal computer (Research Machines, Nimbus VX/2) and employing Sigavg software (CED). All numerical data is expressed as the mean :t: standard error of the mean. For statistical analysis unpaired Student's t-test was used. Following experiments where the electrode had contained biocytin (Sigma), impaled neurons were filled with the dye using depolarizing pulse commands (0.5 nA, 500 ms, 1 Hz) for 5-10 rain. Slices were subjected to a procedure for visualization of the biocytin filled cell and the contralateral side for immunohistochemical staining for tyrosine hydroxylase, as described previously 12,14,20. Tyrosine hydroxylase immunoreactive cells were detectable in the slices cut at the level of the mamillary body of the SN. These cells displayed variety in the
¸
Fig. 1. Coronalsection(100/~m) of guinea-pigmesencephalonat the level of the mammillary bodies, showing tyrosine hydro~iase immunoreactivity in the pars compacta region in intact SN (A) and in SN followingremoval of the pars reticulata (B). C: cell filled with biocytinfrom SNr-cut slice. Scalebar: 100 ~m. shape of the somata and were comparable with the previously decribed forms 3'4'14'19. Moreover, a high number of immunoreactive dendrites of the neurons were observed in the SN pars reticulata, which is especially characteristic of rostrally located dopaminergic neurons 3'14. The morphology of cells filled with biocytin was indistinguishable from that of the neurons positively identified as dopaminergic pars compacta neurons. In SNr-cut slices tyrosine hydroxylase immunoreactive neurons, or neurons filled with biocytin were located at the edge of the sectioning with clear absence of the SN pars reticulata (Fig. 1). A total of 29 cells, 18 control and 11 SNr-cut neurons were recorded and analysed from the rostrai part of the SN. There was no significant difference between the two groups of neurons in their membrane ~otential, spontaneous firing rate or input resistance. In control cells membrane potential was - 51.85 + 2.17 mV (mean and S.E.M.), firing rate 6.58 4- 2.33 Hz and input resistance 120.7 :l: 9.9 M•, whereas the values in the SNr-cut group were -50.88 + 2.88 mV, 2.56 :l: 0.98 Hz and 165 :l: 25.8 M n , respectively. In half of the control neurons, termination of a 200 ms hyperpolariz-
418 ing current pulse resulted in generation of action po. tentials. Similarly, if the membrane potential was held using negative DC current at a hyperpolarized level, additional hyperpolarizing or depolarizing pulses generated a burst of action potentials, described as a 'Low Threshold Spike' (LTSgCa2+) II. By contrast, none of the SNr-cut neurons generated LTS activity, although these cells displayed identical electrophysiologicai characteristics, were recorded from the same level as the intact cells and indeed were frequently in the same slice of midbrain. Application of AChE resulted in cessation of the spontaneous firing and hyperpolarization of the majority of the. recorded cells. Of the control neurons 14 cells responded by diminution of the firing rate and/or hyperpolarization whilst 4 cells were insensitive to AChE (Fig. 2). Of the 11 SNr-cut neurons 8 cells were responsive. The firing rate was significantly reduced both in control (0.33 + 1.9 Hz; P < 0.02) and lesioned cells (0.89 :t: 0.56 Hz; P < 0.05). There was no significant difference either in the AChE.induced hyperpolarization between the two groups of cells: mean values were 5.83 + 1.04 mV and 7.38 + 1.77 mV in control and SNr-cut neurons, respectively, in control cells hyperpolarization of the membrane was associated with a decrease in the input resistance: AChE application reduced the input resistance by 27.2 :l: 7.3 M/~. Analysis of current-voltage relations of neurons indicated a reversal at -88.5:1:9.9 mV, which approximates most closely the equilibrium potential for potassium ions, ( - 100 mV) given K~ - 3.2 mM at 33°C and assuming K , - 140 raM. in SNr-cut neurons, however, AChE induced an increase in input resistance of 85.0 ± 33,2 M/~. Analysis of the current-voltage relation revealed a reversal potential of -28,2 ± 5.8 mV which indicates
significant changes in membrane conductance for other ions beside potassium (Fig. 3). The difference between the reversal potentials for the effect of AChE with and without SNr intact was highly significant (P < 0.001). Slices recorded in this study contained a high proportion of tyrosine-hydroxylase positive cells: furthermore the morphology of these dopaminergic neurons was comparable to those injected with biocytin. It seems most probable therefore that the cells used in this study were indeed dopaminergic. In addition, the morphology of the perikarya and the proximal dendrites of lesioned cells stained for tyrosine-hydroxylase or marked with biocytin strongly resembled the control cells, but clearly indicated the cut of apical dendrites. Electrophysiological characteristics of the cells from SNr-cut slices did not differ significal~tly from control cells, regarding firing rate, action potential width or input resistance, in agreement with previous observations s'14. It seems highly unlikely that the SNr-cut and control cells were not from the same neuronal population. The only difference seen following the sectioning procedure was that none of SNr-cut cells displayed a low-threshold calcium conductance: this finding is in accordance with previous studies localising this conductance in the apical dendrites s.ls. AChE displayed an inhibitory effect on control SN dopaminergic neurons predominantly via an opening of potassium channels, as shown previously 6.7.2°, However, in this study AChE was also found here to be as effective on the SNr.cut cells as on control neurons. This finding suggests that the apical dendrites of SN neurons are not the exclusive target of the protein. However, the mode of the action of AChE appears different at the soma, compared to that at the apical SNr dendrites, AChE appears to hyperpolarize the
AChE ..o°v.,
....
B
AChE
-45 m V ~ . F . . ~ ~ ,
'0mvL. -
.....
5 mln Fig. 2. Recording of membrane potential of a control neuron (A) and a SNr-cut neuron (B) during application of ACHE. These two cells were recorded from each respective side of the same midbrain slice, AChE induced hyperpolarization and reduced firing rate of the recorded neurons in both cases; after termination of AChE application a partial recovery can be observed within 10 min.
419
Control
A
AChE
L
-50 mV . . , i ~
.
.
.
.
.
.
.
.
~
-
. . . .
-_
~
-_':
i20 mV I. . . . . . . . .
..
:_l
j0,2 nA
100 msec
Control
B
AChE
lJ (_
_
_C-'.Z,'
_.__a0.20 nA 100 mseo D
C
nA
nA -0.0
-0.4
-0.|
0:0
-0.6 .4O
-0.4
-0.2
0.0
0.2
.7 • MChl
° ~
[.4o
'4 0
'410
.100
mV .120
-120
Fig. 3. Effects of AChE on membrane properties of control (A) and SNr-cut (B) nigral neurons. These neurons are the same as shown in Fig. 2. A: Traces show characteristic responses of the intact neuron under control conditions, prior to AChE application (left). Note characteristic low-threshold spikes in response to injection of hyperpolarizing current. Following application of AChE (right) there is a hyperpolarization and a fall in input resistance. B: Traces show characteristic responses of the SNr-cut neurons under control conditions. Application of AChE induced hyperpolarization with a marked increase in input resistance. C and D: plots of voltage-current relationships prior to (open circles) and during AChE application (filled circles) in control and SNr-cut neurons, respectively (same cells as in A and B). Each data point is an average of measurements made at 175 ms from 5 successive pulses; lines fitted to data are simple regression lines. The voltage-current relationships intersect at - 102 mV in control cell (C) and at - 22 mV in SNr-cut cell (D).
420 somatic membrane through a net closure of ion channels. However, since the mean reversal potential was still negative, this effect cannot be attributed either solely to Ca 2+ (R.P.: + 132 mV) or Na + (R.P.: +41 mV). The most parsimonious explanation would be that, the predominant closure of Na + or Ca 2+ channels was slightly offset by the modest opening of some somatic CI- and/or K + channels. These findings suggest that AChE can have dual actions on pars compacta neurons, depending on its site of action on the cell. The action of AChE within the pars compacta (at the level of cell body and proximal dendrites) itself seems to involve no single ion channel and indeed would be masked in a physiological situation by the principal effects seen when the dendrites are intact. The results demonstrate that the dominant channel via which AChE acts is located on the apical dendrites. However the net action is deten~ined not by an exclusivity of AChE for any single ion channel since the protein produces effects just as marked at the level of the soma. it is more probable that the distal dendrites are more readily affected since this part of the cell is the most likely origin of released AChE t3. Two conclusions arise from this study, first there are multiple ionic bases for the action of ACHE, and second the apical dendrites could have a critically active role in neuronal functions, distinct from physico-chemical events at the soma. This work was supported by BristoI.Mayers.Squibb Co (USA), We would like to thank Mr. C. Wehb for excellent technical assistance and Mrs. P. Cordery for histological services. I Chtodo, L.A, Antelman, S.M., Caggiula, A.R. and Lineberry, C.G., Sensory stimulation alters the discharge rate of dopamine (DA) neurons: evidence of two functional types of DA cells in the substantia nigra, Brain Res., 189 (1980) 544-549. 2 Chubb, I.W., Goodman, S. and Smith, A.D., is acetylcholinesterase secreted from central neurons into the cerebrospinal fluid? Neuroscience, I (1976) 57-62. 3 Gerfen, C.R., Baimbridge, K.G. and Thibault, J., The neostriatal mosaic: !11. Biochemical and developmental dissociation of patch-matrix mesostriatal systems, .I. Neurosci., 7 (1987) 39353944. 4 Grace, A.A. and Onn, S-P, Morphology and electrophysiologicai properties of immunocytochemically identified rat dopamine neurons recorded in vitro, J. Neurosci., 9 (1989) 3463-3481.
5 Greenfield, S.A., A non-cholinergic action of acetylcholinesterase (ACHE) in the brain: from neuronal secretion to the generation of movement, Cell. Mol. Neurobiol., 11 (1991) 55-77. 6 Greenfield, S.A., Jack, JJ.B., Last, A.TJ. and French, M., An electrophysiological action of acetylcholinesterase independent of its catalytic site, Exp. Brain Res., 70 (1988) 441-444. 7 Greenfield, S.A., Nedergaard, S., Webb, C. and French, M., Pressure ejection of acetylcholinesterase within the subtantia nigra has non-classica| actions on the pars compacta cells independent of selective receptor and ion channel blockade, Neuroscience, 29 (1989) 21-25. 8 Harris, N.C., Ramsay, S., Kelion, A. and Greenfield, S.A., Electrophysiological evidence for a dendritic localization of calcium conductance in guinea pig substantia nigra neurons in vitro, Exp. Brain Res., 74 (1989) 411-416. 9 Kita, T., Kita, H. and Kitai, S.T., Electrical membrane properties of rat substantia nigra compacta neurons in an in vitro slice preparation, Brain Res., 372 (1986) 21-30. 10 Kubota, Y., Kang, Y. and Kitai, S.T., Two types of substantia nigra compacta neurons based on electrophysiological membrane characteristics, Soc. Neurosci. Abstr., 15 (1989) 900P. 11 Llinas, R.R., Greenfield, S.A. and Jahnsen, H., Electrophysiolo~ of pars compacta cells in the in vitro substantia nigra: a possible mechanism for dendritic release, Brain Res., 294 (1984) 127-132. 12 Murphy, K.P.SJ. and Greenfield, S.A., ATP-sensitive potassium channels counteract anoxia in neurons of the substantia nigra, Exp. Brain Res., 84 (1991) 355-358. 13 Nedergaard, S., Bolam, J.P. and Greenfield, S.A., Facilitation of a dendritic calcium conductance by 5-HT in the substantia nigra, Nature, 333 (1988) 174-177. 14 Nedergaard, S. and Greenfield, S.A., Subpopulations of pars compacta neurons in the substantia nigra: the significance of qualitatively and quantitatively distinct conductances, Neuroscience 48 91992) 423-437. 15 Preston, RJ., McRea, R.A., Chang, H.T. and Kitai, S.T., Anatomy and physiology of substantia nigra and retrorubral neurons studied by extra and intracellular recording and by horseradish peroxidase labelling. Na~roscience, 6 (1981) 331-344. 16 Shepard, P.D. and German, D.C., Electrophysiological and pharmacological evidence Ibr the existence of distinct sublmpulations of nigrostrlatal dopamincrgtc neuron in the rat, Ne~rosclt,nce, 27 (1988) 537-546. 17 Smits, R.J.P.M., Steinbusch, H.W.M. and Mulder, A.H., Distribution of dopamine-tmmunoreactive cell bodies in the guinea pig brain. J. Chem. Neuroanat., 3 (1990) 101-123. 18 Walsh, J.P., Cepeda, C, Buchwald, N,A. and l~vine, M.S,, Neurophysiological maturation of cat substantia nigra neurons: evi. dence from in vitro studies, Synapse, 7 (1991) 291-300. 19 Wassef, M., Berod, A. and Sotelo, C., Dopaminergic dendrites in the pars reticulata of the rat substantia nigra and their striatal input. Combined immuno~tochemical localization of tyrosine hydro~lase and anterograde degeneration, Neuroscience, 6 (1981) 2125-2139. 20 Webb, C. and Greenfield, S.A., Non-cholinergic effects of acetylcholinesterase in the substantia nigra: a possible role for an ATP-sensitive potassium channel, Exp. Brain Res. 89 (1992) 49-58.
