Glycine effects on glutamate-receptor elicited acetylcholinesterase release from slices and synaptosomes of the spinal ventral horn

JOURNAL

OF THE

NEUROLOGICAL SCIENCES ELSEVIER

Journal of the Neurological Sciences 139 (Suppl.) (1996) 76-82

Glycine effects on glutamate-receptor elicited acetylcholinesterase release from slices and synaptosomes of the spinal ventral horn Daniel Rodriguez-Ithurralde as* J, Silvia Olivera a, Anabel1 La Paz a, Oscar Vincent ‘, Amalia Rondeau a of Molecular

’ L.aboratoty b Dkision

of Neuromyology

Neuroscience, Institute

Instituto

de Inwstigaciones Biol6gicaJ Clemente Estable (IIBCE), Auenida It&a 3318, C. Postal 11600, Montecideo, Uruguay de Incestigaciones Bioldgicas Clemente Estable (IIBCEI, Avenidu ltalia 3318, C. Postal 11600, MontecGdeo, Urugua?;

Received 23 March 1996

Abstract To study the mechanisms by which glutamate-elicited acetylcholinesterase release (GEAR) might play a part in the pathogenesis of excitotoxically triggered motor neurone disease, and to investigate the interaction of GEAR with spinal glycinergic mechanisms, we measured acetylcholinesterase (AChE) and cholinergic markers, after stimulating ventral horn slices and synaptosomes from the mouse spinal cord, with both glutamate- and glycine-receptor agonists. Glutamate (GLU), kainate and AMPA, as well as glycine (GLY) evoked dose-related, calcium-dependent liberation of soluble forms of AChE from both slices and synaptosomes. GLY-evoked AChE release showed remarkable age-related postnatal changes. In the immature slice of the ventral horn, GLY potentiated the GEAR response in the presence of strychnine, suggesting N-methyl-D-aspartate (NMDA) receptor involvement, and was also able to evoke a strychnine-sensitive AChE release in the abscence of exogenous GLU. After the 28th postnatal day, nearly all the AChE secreted was released either after the activation of non-NMDA glutamate receptors or by strychnine-sensitive GLY-evoked AChE release mechanisms. Both GEAR and GLY-evoked AChE release might impair the negative feedback loop which modulates the overactivation of motor neurones, and cause prolonged extracellular rises of soluble AChE. These effects might augment the vulnerability of motor neurones to excitotoxic stress, promote fiber outgrowth, and eventually accelerate the metabolic exhaustion of lower motor neurones. It is possible that the mechanisms described are operative at the spinal cord of ALS/MND patients. Keywords: Acetylcholinesterase; ALS; a-Amino-3.hydroxy-5.methylisoxazole-4-propionic acid (AMPA)/kainate Glutamate receptor; Glycine; MND; Motor neuron; Mouse; NMDA receptor; Spinal cord slice

1. Introduction Excessive activation of glutamate receptors (GluR), one of the major potential mechanisms implicated in the pathogenesis of sporadic Amyotrophic Lateral Sclerosis/Motor Neurone Disease (ALS/MND) (Rothstein et al., 1993; Rothstein and Kuncl, 1995; Doble, 1995; Roman0 et al., 1995; Kwak and Nakamura, 19951, consistently elicits the release of acetylcholinesterase (AChE, EC 3.1.1.7) from cerebral and spinal cord slices (Sapriza et al., 1992; Yannicelli et al., 1992; Rodriguez-Ithurralde et al., 1995). More-

* Corresponding author. Tel: +598 (2) 471.616, ext. 125; Fax: f598 (2) 475-461 or 475.548. e-mail: [email protected] ’ Director of the Laboratory of Molecular Neuroscience (PEDECIBA) at IIBCE.

receptors; AMPA;

Excitotoxicity;

over, glutamate-elicited AChE release (GEAR) shows remarkable age-dependent changes during postnatal development (Rodriguez-Ithurralde and Vincent, 1994), which parallel the developmental periods of selective vulnerability of motor neurones to excitotoxic amino acids found by McDonald et al. (1991). These observations showed both a close qualitative association and a quantitative correlation between GluR activation and AChE secretion, and for that reason we have suggested that the GEAR response might play a part in the pathogenesis of excitotoxically-triggered motor neurone disease (Rodriguez-Ithurralde et al., 1995). In fact, if GEAR operates in vivo, increased extracellular levels of soluble AChE could have profound consequences on modulatory circuits of the spinal cord, since feed-back inhibition of motor neurones is activated by cholinergic synapses of motor neurone axon collaterals on glycinergic

0022-510X/96/$15.00 Copyright 0 1996 Elsevier Science B.V. All rights reserved. PII SOO22-5 10X(96)00095-0

D. Rodriguez-Ithurralde

et al./

Journal

qf the

inhibitory interneurones. Thus, an early rise in soluble AChE could theoretically block these cholinergic synapses, impairing the modulation of GluR-mediated motor neurone overactivation (Rodriguez-Ithurralde et al., 1995). HOWever, the main sources of glycine (GLY) in the ventral horn, i.e., glycinergic intemeurones and some glia cells, could be activated by GluR agonists (Cull-Candy, 1995; Kwak and Nakamura, 1995) or by acetylcholine (ACh) before any hypothetical AChE-mediated impairement of cholinergic transmission takes place. The consequent glycine (GLY) release might also be present in vivo and be relevant to ALS/MND pathogenesis, since GLY can participate in either protective or lesional mechanisms, according to the type of GluR and/or GLY receptor (GlyR) involved. Despite their potential importance for neurodegenerative pathology, effects of GLY on AChE release, GluR subtypes involved in GEAR, and GLY-GEAR interactions, are not well understood mechanisms. The aim of this work was to investigate them in controlled conditions, by using isolated slices and synaptosomes from the ventral region of the spinal cord.

2. Methods 2. I. Animals used und tissue preparation Male albino mice of the IIBCE strain, l-60 days of postnatal age (DPN) and adults of 90- 100 DPN were used. They were killed by painless decapitation, and the cervical (or in some experiments the lumbar) enlargements of the spinal cord were quickly dissected out. Longitudinal (columnar, 900 pm) slices were obtained on a cold plate from the ventral half of the spinal cord (the ‘ventral horn’ slices) and quickly washed in cold (0-4°C) oxygenated (95% 0,/S% CO,) artificial cerebrospinal fluid containing 10 mM glucose (ACSF), as described (RodriguezIthurralde and Vincent, 1994). In other experiments, the cervical spinal cord was sliced at 900 p.m following the transversal plane and the transversal slices so obtained further sectioned on a cold plate to isolate the ventral horn plus surrounding white matter from the periependymal and dorsal regions, with similar results in incubation experiments. 2.2. Pharmacologic stimulation of slices Ventral horn slices were either superfused with or incubated into oxygenated ACSF (35°C) containing 10 mM glucose to assess, after a 25 min stabilisation period, the effects of different concentrations of glutamate receptor (GluR) agonists and/or antagonists on the release of proteins, cholinesterases, lactate dehydrogenase (LDH) and other markers. The same experimental design was used to assess the effects of GLY on AChE release and to study the interactions of GLY with the GEAR response. Drug

Neurological

Sciences

I39

(Suppl.)(1996)

76-82

11

stimulation exploited recent knowledge on both spinal distribution and pharmacological potentiation of the different GluR subtypes and subunits (Henley et al., 1992; Kalb et al., 1992; Jakowec et al., 1995a,b; Tijlle et al., 1995a,b). In timed superfusion experiments, 1 to 3 ventral slices were placed into a superfusion chamber (0.4 ml) which dropped into a fraction collector. After superfusion with basal ACSF (35”C), ACSF (35°C) containing a known concentration of a GluR agonist and/or antagonist was used (Rodriguez-Ithurralde and Vincent, 1994). In incubation experiments, after stabilization of the slices in basal oxygenated ACSF (25 min, 35°C) the slices were incubated into oxygenated (95% 0,/5% CO, > ACSF containing different concentrations of GluR agonists and/or antagonists, as in Rodriguez-Ithurralde et al. (1995). 2.3. Subcellular fractionation by differential centrifugation and synaptosomal stimulation Fractionation procedures were based on previously described protocols (Yee et al., 1989). Ventral halves of spinal cords from adult male mice were homogenized with a glass-teflon homogenizer in 10 volumes of ice-cold 0.32 M sucrose (0-4°C) and then spun for 10 min at 1000 g. All centrifugation steps were carried out at 4°C and all homogenizations were in ice-cold buffer. The supematant was saved and the pellet was rehomogenized in 10 volumes of 0.32 M sucrose and recentrifuged under the same conditions. The pellet (PI) consists of nuclei and cell debris. Supematants were pooled and centrifuged for 10 min at 20000 g. The supematant was saved and the pellet (P2) was resuspended in 1 volume of 0.32 M sucrose and layered in 4 ml aliquots onto discontinuous sucrose gradients (4 ml 1.2 M sucrose, 4 ml 0.8 M sucrose) in 12 ml Beckman SW40 centrifuge tubes. The gradients were centrifuged at 58 000 g for 110 min. The 0.8/1.2 M sucrose interface, comprising synaptosomal membranes, was resuspended to 50 vol in ice-cold ASCF and then either used directly in incubation experiments or stored in aliquots at - 80°C. For studying drug effects on AChE release, synaptosomal aliquots were incubated into oxygenated (95% 0,/5% CO,) artificial cerebra-spinal fluid (ACSF, 35°C) containing different concentrations of glutamate receptor (GluR) and/or glycine receptor (GlyR) agonists and/or antagonists. In some fractionations, the spinal cord tissue was homogenized in 0.32 M sucrose in 1mM sodium phosphate buffer containing 0.1 mM EDTA and the supernatants from the 1000 g centrifugation were spun at 11 500 g after pooling. 2.4. Sample processing, cholinesterase assay and molecular form isolation Timed ACSF fractions were either collected from the superfusion system outlet or sequentially sampled from the incubation media (Rodriguez-Ithurralde and Vincent,

78

D. Rodriguez-Ithurralde

et al./.lournal

of the Neurological Sciences 139

1994). After centrifugation, soluble enzymatic activities and proteins present in the ACSF used in the experiments were either assayed or concentrated, desalted and fractionated by gel filtration chromatography. Fractionations of AChE molecular forms were based on previously described methods. Media from incubation experiments were concentrated and desalted, and then fractionated by gel filtration on Sephacryl S-300 columns as previously described (Karlsson et al., 1985). AChE and butyrylcholinesterase (BuChE, EC 3.1.1.8) activities were assayed by a modification of the spectrophotometric method of Ellman et al. (1961) using acetylthiocholine iodide or butyrylthiocholine iodide as a substrate, respectively, using previously published protocols (Rodrfguez-Ithurralde et al., 1983). Blanks for non-specific hydrolysis and control incubations in the presence of inhibitors were performed as recommended by Ellman et al. (1961). Inhibition controls using fasciculin 2 for detecting fasciculin-insensitive cholinesterases (Rodriguez-Ithurralde et al., 1983, Karlsson et al., 1985) and 2 X 10m5 M ethopropazine for inhibiting BuChE activities (Biagioni et al., 1995) were also employed. Erythrocyte ghosts and purified AChE from electric eel (Sigma) were used as AChE activity standards. Erythrocyte ghosts were prepared as in Karlsson et al. (1985) using HIV-free blood from the Hospital Italiano, Montevideo, Uruguay. Total spinal-cord content in cholinesterases was determined in non-incubated slices and after 30 min incubations. Tissue was collected in ten volumes of ice-cold 0.1 M potassium phospate buffer (pH 7.4) and sonicated. The sonicate was centrifuged for 10 min at 500 g, and AChE and BuChE activities present in the primary soluble pool were determined. Total content was assayed upon membrane disruption with 0.5, 1 and 5 % Triton X- 100. 2.5. Lactate dehydrogenase (LDH) assay and protein determination LDH was measured in aliquots of incubation and superfusion media as in Biagioni et al. (1995). The absence of LDH activity change in the medium was taken as an indication that the AChE activity increase was not attributable to unspecific changes in cell membrane permeability. Protein was estimated by the method of Lowry et al. (195 1) with bovine serum albumin as standard. All biochemical assays were run in quintuplicate, and data shown in figures illustrated means of at least 5 experiments unless otherwise indicated. Error bars were represented when higher than 10% of the mean. 2.6. Histochemical methods For determining in vitro effects of drugs on histochemitally demonstrable cholinesterases, control and stimulated slices were incubated as described in the section Pharmacologic stimulation of slices (2.21, and sectioned in a

(Supp1.I(1996) 76-82

cryostat (5 pm) either fresh of after fixation in Holt’s buffered formaline solution (0-4°C). Sections were processed for localisation of cholinesterases, using BuChE and/or AChE inhibitors at concentrations described for the method of Ellman et al. (1961) in a previous paper (Rodtiguez-Ithurralde et al., 1983). In other sections, a final concentration of 2 X lop5 M ethopropazine was used to inhibit BuChE (Biagioni et al., 1995). 2.7. Materials Acetylcholine (ACh), acetylthiocholine, butyrylthiocholine, purified AChEs from electric eel and from bovine erythrocite, bovine plasma albumin, diisopropylfluorophosphate (DFP), 5,5’dithiobis-(-2-nitrobenzoic acid), ethopropazine, etylene diaminotetracetic acid (EDTA), glutamate, glycine, kainate, physostigmine sulfate, Tris-(hydroxymethyl)-aminomethane and Triton X- 100 were purchased from Sigma (St. Louis, MO, USA). Fasciculin 2 was purified in our laboratory from D. angusticeps crude venom (Rodriguez-Ithurralde et al., 1995). Gel filtration chromatography media were from Pharmacia (Uppsala, Sweden). All other chemicals were of the purest grade available.

3. Results 3.1. Biochemical cholinesterases

characterization

of

released

Ventral spinal cord slices showed no noticeable AChE release when superfused with basal ACSF; however, when ACSF containing GLU was superfused, a peak of AChE activity appeared 2-4 min after the start of the pharmacological stimulation (not shown). Upon incubations in the presence. of the GluR agonists GLU, AMPA and kainate, and also after parallel incubations in the presence of GLY, both ventral horn slices and synaptosome preparations showed Ca2+- dependent releases of cholinesterases, as compared with control identical preparations incubated in similar conditions but in the absence of amino acid. The AChE/BuChE ratio and molecular forms released upon exposure to either GluR agonists or GLY were undistinguisable. None of the agents caused a significant increase in the LDH activity of the ACSF. Upon incubations with both types of agents, more than 97% of the cholinesterases secreted by ventral slices was true AChE (EC 3.1.1.7), and globular form-mainly G4-always represented more than 90% of the total secreted AChE activity, irrespective of the drug which elicited the release. The molecular form pattern of AChE deducted from gel filtration chromatography was the following: G4 > Gl > G2. This pattern and the quantitative ratio amongst molecular forms were almost identical to those obtained from spinal cord slices

D. Rodriguez-Ithurralde

Concentration

et al./Journal

Glycine

(moles/l)

Fig. 1, Dose-response characteristics of acetylcholinesterase (AChE) release evoked by L-glutamate (GLU) and kainate from ventral horn slices. Slices were prepared from cervical spinal cords of male, 36-day-old (postnatal, PN) mice. After stabilization in basal (35°C) artificial cerebrospinal fluid (ACSF), slices were incubated in ACSF (35°C) containing either 0.00, 0.01, 0.05, 0.25, 0.50, 1.00 or 5.00X 10m3 mol/l GLU. Kainate curve shows results obtained in the same conditions, but using either 0.1, 1, 10, 100, 500 or 1000X lOm6 mol/l kainate instead of glutamate. AChE activity released after 30 min of stimulation was kainate; n assayed and expressed as AA,, z / m g of wet tissue: A-A, - n , glutamate. Each point represents the mean of at least five determinations. Error bars are represented when higher than 10% of the mean.

that included dorsal horns (Rodriguez-Ithurralde 1995).

79

of the Neurological Sciences 139 (Suppl.) (1996) 76-82

et al.,

3.2. Pharmacologic features of GEAR GLU (2 X 1O-6 M to 4 X lop3 M,) and kainate (1 X lo-’ M to 5 X 10m4 M) elicited dose-related, Caz f-dependent releases of soluble AChE (mainly the G4 molecular form), with maximal responses at 5 X lop4 M kainate and at 1 X 10m3 M GLU. Fig. 1 shows a typical experiment, showing remarkable decreases of the response at concentrations higher than 1 X lop3 M for kainate and 5 X 10m3 M for GLU. Age-dependent changes in GEAR were detected in the only-ventral slices, and were similar to those described for the whole spinal cord (RodriguezIthurralde and Vincent, 1994, Rodriguez-Ithurralde et al., 1995). For instance, GLY potentiation of GEAR is present in ventral slices prepared from mice aged 12-28 DPN, but absent in 28 days and older mice (Fig. 3).

concentration

(PM)

Fig. 2. Dose-response curve of glycine (GLY)-evoked AChE release from ventral horn slices and synaptosomes of adult mice (90 DPN). After stabilization in basal ACSF as in Fig. 1, different preparations of slices and synaptosomal fractions were simultaneously exposed to either 0, 25, 50, 75, 100 or 250X 10e6 mol/l GLY. The increase in AChE activity after 30 min was expressed as AA 4, 2 /mg of wet tissue as compared to the activity of control, non-stimulated preparations: slices (W - W ) and synaptosomes (A -A ). Each point represents the mean of at least five determinations.

the absence of exogenous GLU (Fig. 2). GLY-evoked AChE release evidenced remarkable age-related changes and therefore, a precise determination of mice age (between birth and 40 DPN) is necessary for its analysis. Fig. 2 represents GLY-evoked release in the abscence of exogenous GLU and strychnine from adults, the stage of minimal response to GLY. Maximal secretion was attained between 50-75 M GLY. A minor part of this GLY-evoked release was strychnine-resistant (not shown). 3.3.2. GLY potentiation of GEAR As shown for slices including dorsal horns (RodriguezIthurralde et al., 1995) in ventral slices from immature

3.3. Glycine actions on AChE release Slices from the ventral half of the spinal cord showed 2 main types of AChE release requiring the presence of GLY in the incubation medium: (1) GLY-evoked AChE release (independent of the presence of exogenous GLU), and (2) GLY potentiation of GEAR. Both types of actions varied with developmental age of mice, and reach adult stable values at 40 DPN (Figs. 2-4). 3.3.1. GLY-eooked AChE release GLY was able to evoke a dose-related, Ca*+-dependent release of AChE from ventral slices and synaptosomes in

Fig. 3. Effects of Ca’+, strychnine, and Mg*+ on GLU- and GLYevoked AChE release from ventral horn slices. Slices from male, 29 DPN-old mice. After stabilisation, slices were incubated into either ACSF containing Ca*+ at a final concentration of 1.2 mmol/l (hatched bars) or Ca2+-free ACSF (solid bars). Both ACSF contained the following final concentrations of drugs (in kmol/l): in condition Glu, glutamate 500; in condition Glu Gly, glutamate 500 +glycine 60; in condition Glu Gly Sky, glutamate 500+glycine 60+strychnine 60; in condition Glu Gly Mg, as in Glu Gly plus Mg2+ 8.0 mmol/l; and in condition Gly, glycine 60. After incubation and centrifugation, total cholinesterase activity an specific AChE activity were measured. AChE activity was expresed as percent of increase as compared to control release (from non-stimulated slices).

D. Rodriguez-Ithurralde

80

2

15

20

25

Postnatal

30 age

35

et al./Joumal

40

of the Neurological Sciences 139

45

(days)

Fig. 4. Postnatal evolution of GLY-evoked AChE release from ventral horn slices. Slices were obtained from mice at different postnatal ages and incubated for 30 min in LCRA (35°C) containing 60 pmol/l GLY. AChE activity was measured and the activity released to the LCRA was expressed as increase per cent as compared with the activity shown by slices from mice of the same age, incubated in LCRA without GLY. Each point represents the mean of at least six determinations.

mice (12 -28 DPN) 3-6 X IO-’ M GLY caused a dual (two-component) response in the presence of GLU: (1) GLY potentiated GEAR, and this action was not blocked by 3-6 X lo-’ M strychnine, an effect usually attributed to NMDA receptor involvement. (2) In addition, GLY was able to trigger, even in the presence of GLU, an apparent increase in the GEAR response which was abolished by the presence of 3-6 X 10m5 strychnine. After the 29 DPN, strychnine abolished GLY potentiation of GEAR. In order to discriminate amongst the different GLY actions on release, the interactions between GLU and GLY effects were investigated both in the presence and in the absence of both strychnine, Ca2+ and Mg*‘. Fig. 3 illustrates that analysis carried out at a fixed age (29 DPN) to avoid age-dependent changes. At 29 DPN and thereafter, apparent increases in GEAR elicited by GLY were abolished in the presence of 6 X 10m5 M strychnine (Fig. 3, GLU GLY STRY condition). Ca2+-dependency was evident in most conditions, and Mg2+ blocked more than half of the strychnine-resistant release caused by GLU plus GLY (Fig. 3). 3.3.3. Age-dependency of GLY effects on AChE release GLY-evoked AChE release showed important age-related changes during development. Postnatal evolution of GLY-evoked release measured in the absence of strychnine is shown in Fig. 4. Two main peaks occurred, one at 14 DPN and a second one at 27-29 DPN. Adult values were attained at approximately 40 DPN.

4. Discussion 4.1. Pharmacological characteristics of GEAR and GLYevoked AChE release GEAR pharmacologic features of ventral slices and synaptosomes confirmed earlier data from dorso-ventral 1994; slices (Rodriguez-Ithurralde and Vincent,

(Suppl.)(1996,6)76-82

Rodriguez-Ithurralde et al., 1995), demonstrating that GluR activation is necessary for this Ca*+ - dependent response, and suggesting the involvement of both NMDA and nonNMDA glutamate receptors in the immature animals, and the almost exclusive participation of non-NMDA receptors after 28 DPN. These observations are consistent with recent reports on the anatomical distribution and age-dependent modulation of both NMDA and non-NMDA receptor subunits in the rodent ventral horn (Kalb et al., 1992; Tiille et al., 1995b; Jakowec et al., 1995a,b). NMDA receptors are present during the first 2-3 postnatal weeks, to almost disappear later (Kalb et al., 1992), as occurs with GLY-resistant potentiation of GEAR. On the other hand, GluR,-, subunit expression is subjected to remarkable regulation during the first postnatal weeks, and AMPA receptors lacking the GluR, (GluR,) subunit, being therefore permeable to Ca2+, become relatively abundant in many motor neurones (Tiille et al., 1995a). GLY-evoked AChE release deserve special consideration, since it showed striking postnatal changes (Fig. 4), including 2 release peaks centred at 14 and 28 DPN (Fig. 4). They are in close temporal correlation with peaks of selective vulnerability to excitotoxic agents, previously described by McDonald et al. (1991), and might also be related to developmental periods of increased plasticity of spinal synapses, as proposed for transient increases in excite-vulnerability (Rodriguez-Ithurralde and Vincent, 1994). After 28 DPN, most GLY-evoked AChE release is strychnine-sensitive and can be adscribed to classical (inhibitory) GlyR, present at the motor neurone membrane. 4.2. Site and mechanism of AChE release Our present data from ventral horn slices and synaptosomes show unequivocally that dorsal horn structures are neither necessary for GEAR nor for GLY-evoked AChE release. Although the precise site of secretion is unknown, the motor neurone is the principal cell candidate for GEAR (Rodriguez-Ithurralde et al., 1995). Several in vitro and in vivo studies carried out in different species have shown the ability of motor neurones to release AChE upon varied stimuli (Appleyard, 1992; Sapriza et al., 1992; Massoulii: et al., 1993; Biagioni et al., 1995). Moreover, the body of our results-e.g., the secretion kinetics, its Ca*+-dependency, the molecular form pattern, the blocking by Mg*+, and particularly the release from synaptosomal preparations-point to an intraneuronal, possibly presynaptic origin of the AChE, and synapses from motor neurone axon collaterals on local intemeurones have been proposed as release locations (Rodriguez-Ithurralde et al., 1995). However, dendrites can not be discarded, since they are able to secrete AChE in the substantia nigra and other regions (Taylor et al., 1990; Appleyard, 1992; Biagioni et al., 1995). On the other hand, due to the fact that motor neurones are enriched in GlyR sites, and because most local intemeurones release GLY, the simplest explanation

D. Rodriguez-Ithurralde

et al. / Journal

oj’the Neurological

of data on GLY-evoked AChE release is to postulate that this type of response originates also in motor neurones. However, both glia cells and cholinoceptive glycinergic intemeurones bear GluR and cannot be discarded as release sites. These interneurones, as many cholinoceptive cells, could be able to secrete AChE, and might be activated either directly by extracellular GLU rises or as a consequence of GLU-evoked motor neurone activation (Rodriguez-Ithurralde et al., 1995). Most mechanisms discussed can theoretically be present in the same cell type, and even circunscribed to the axon terminal. Therefore, the simplest common path for explaining most forms of AChE release, including K+-evoked secretion (Biagioni et al., 1995) GEAR as well as GLY potentiation of GEAR, is to postulate that the respective stimuli are coupled to AChE secretion through a cytosolic increase in Ca2+. In fact, both NMDA and AMPA receptors lacking the GluR, (GluR,) subunit are permeable to Ca*+ and depolarizations caused by both kainate receptors and K+ can help opening voltage-dependent NMDA receptor permeabilities (Henley et al., 1992). However, strychnine-sensitive GLY-evoked AChE release can not be triggered directly by inward Ca2+ fluxes. 4.3. Significance of GEAR and GLY-euoked AChE release GEAR has been demonstrated in other CNS regions, including neocortex, hippocampus, striatum and brain stem from a wide range of species representative of fishes, amphybia, birds and mammals (Sapriza et al., 1992; Yannicelli et al., 1992; Rodn’guez-Ithurralde and Vincent, 1994) whereas AChE release evoked by other stimuli acting both in vivo and in vitro, was also shown to be widespread in the mammal CNS (Sapriza et al., 1992; Massoulie et al., 1993; Biagioni et al., 1995). It is unlikely that early AChE secretion represents a sign of cellular damage because it occurred even following physiological (movement) and distant electrical stimulation (see, for a review, Appleyard, 1992) and because the simultaneous release of other intracellular proteins, as LDH, was negligeable (Rodriguez-Ithurralde and Vincent, 1994; Biagioni et al., 1995). On the contrary, it might form part of the repertoire of chemical signalling of the motor neurone, and subserve an important role (Biagioni et al., 1995) given the phylogenetic conservation and wider anatomical distribution of the response (Sapriza et al., 1992) which could be triggered in vivo either by rises in GLU levels or by motor neurone overactivation (Rodriguez-Ithurralde et al., 1995). K+-evoked AChE release has been related to neuronal activity (Taylor et al., 1990; Appleyard, 1992) as well as to neural development and differentiation (Sapriza et al., 1992 Biagioni et al., 1995) whereas GEAR has been associated with developmental synaptic plasticity (Rodriguez-Ithurralde and Vincent, 1994) and with overactivation of GluR. In this context, it is worth noting that excitatory amino acids (Henley et al., 1992; Jakowec et al.,

Sciences

139 (Suppl.)

(1996)

76-82

81

1995a,b) and other transmitters (Lipton and Kater, 1989; Lauder, 1993) and modulators (Layer, 1991) participate in activity-dependent phenomena regulating development, synaptic plasticity and stabilization of neural circuits. It has been postulated that released ACh inhibits neurite outgrowth (Lipton and Kater, 1989) whereas AChE would promote their elongation (Layer, 1991) but decreasing their ability to establish synaptic contacts (Biagioni et al., 1995). All these observations and our data, taken together, suggest a possible link, i.e., the release of AChE, between activity-dependent stimulation of GluR and plastic modifications of neural circuits and synapses. 4.4. Significance of GEAR for ALS/ MND and neurodegeneratiue pathology There are several possible ways in which an extracellular AChE excess can contribute to aggravate an excitotoxitally triggered injury of the spinal cord: (1) Impairing feed-back inhibition of motor neurones (Rodriguez-Ithurralde et al., 1995) which normally modulates their overactivation, by blocking cholinergic transmission from motor neurone axon collaterals to cholinoceptive, glycinergic intemeurones. In this context, it is important to note that ALS patients sera have frequently shown increased AChE levels (Festoff, 1982). (2) Triggering a sustained cycle of AChE synthesis and secretion by motor neurones. If extracellular GLU levels rise in the spinal cord of ALS/MND patients, it is conceivable that not only GEAR would be triggered, but GLY-releasing intemeurones might be activated as well, either directly by GLU acting on their GluR or by motor neurone axon collaterals, causing prolonged increases both in extracellular GLY and in GLY-evoked AChE release from motor neurones. (3) Interfering with regulatory mechanisms that modulate the balance between synaptic plasticity and stability, since the disbalance AChE/ACh in favor of the enzyme can trigger the predominance of outgrowth mechanisms (Lipton and Kater, 1989; Biagioni et al., 1995). (4) Secreting inactive AChE. In laboratory rodents, AChE inactivation by tight-binding inhibitors, e.g., fasciculins, leads to the production of more or less generalized fasciculations (Rodriguez-Ithurralde et al., 1983; Karlsson et al., 1985) as seen in ALS patients. Mechanisms (2) (3) and (4) can also contribute to metabolic stress of motor neurones. (5) Activating non-classical (Appleyard, 1992) noncholinergic functions of cholinesterases (Sapriza et al., 1992; Massoulii: et al., 1993; Rodriguez-Ithurralde and Vincent, 1994; Dajas et al., 1993) which can have important roles in the pathogenesis of neurodegenerative damage, and must be taken into account when designing neuroprotective strategies (Rothstein and Kuncl, 1995) and selecting new therapeutic agents against neurodegenerative diseases (Millard and Broomfield, 1995).

82

D. Rodriguez-lthurralde

et al. /Journal

of the Neurological Sciences 139 (Suppl.) (19961 76-82

Acknowledgements The authors wish to thank Mr. Mario Lalinde for his excellent technical assistance with the illustrations. This study was partially supported by the European Commission (EC Contract CIl * -CT94-0005) to DRI, and by CONICYT of Uruguay, PEDECIBA (Area Biologia) of Uruguay, and the British Council.

References Appleyard, M. (1992) Secreted acetylcholinesterase: non-classical aspects of a classical enzyme. Trends Neurosci., 15: 485-490. Biagioni, S., Bevilacqua, P., Scarsella, Cl., Vignoli, A.L. and AugustiTocco, G. (1995) Characterization of acetylcholinesterase secretion in neuronal cultures and regulation by high K+ and soluble factors from target cells. J. Neurochem., 64: 1528-1535. Cull-Candy, S. (1995) NMDA receptors: do glia hold the key? Curr. Biol. 5: 841-843. Dajas, F., Silveira, R., Costa, G., CastelI& M., Jerusalinsky, D., Medina, J., Levesque, D. and Greenfield, S. (1993) Differential cholinergic and non-cholinergic actions of acetylcholinesterase in the substantia nigra revealed by fasciculin-induced inhibition. Brain Res., 616: 1-5. Doble, A. (1995) Excitatory amino acid receptors and neurodegeneration. Thtrapie, 50: 319-337. Ellman, G.L., Courtney, D.K., Andres, V. and Featherstone, M. (1961) A new and rapid calorimetric determination of acetylcholinesterase activity. B&hem. Pharmacol., 7: 88-95. Festoff, B. (1982) Release of acetylcholinesterase in amyotrophic lateral sclerosis. In: L.P. Rowland (Ed.), Human Motor Neuron Disease. Raven Press, New York, NY, pp. 503-516. Henley, J., Ambrosini, A., Rodriguez-Ithurralde, D., Sudan, H., Lu, Q., Usherwood, P.N.R. and Barnard, E.A. (1992) Purified unitary kainate/alpha-amino-3-hidroxy-5-methylisoxazole propionate (AMPA) and kainate/AMPA/N-methyl-paspartate (NMDA) receptors with interchangeable subunits. Proc. Natl. Acad. Sci. USA, 89: 4806-48 10. Jakowec, M.W., Yen, L and Kalb, R.G. (1995a) In situ hybridization analysis of AMPA receptor subunit gene expression in the developing rat spinal cord. Neuroscience, 67: 909-920. Jakowec, M.W., Fox, A.J., Martin, L.J. and Kalb, R.G. (1995b) Quantitative and qualitative changes in AMPA receptor expression during spinal cord development, Neuroscience, 67: 893-907. Kalb, R.G., Lidow, M.S., Halsted, M.J. and Hocktield, S. (1992) Nmethyl+aspartate receptors are expressed in the developing spinal cord ventral horn. Proc. Natl. Acad. Sci. USA, 89: 8502-8506. Karlsson, E., Mbugua, P. and Rodriguez-Ithurralde, D. (1985) Anticholinesterase toxins. Pharmac. Ther., 30: 259-276. Kwak, S. and Nakamura, R. (1995) Acute and late neurotoxicity in the rat spinal cord in viva induced by glutamate receptor agonists. J. Neurol. Sci., 129 (Suppl.): 99-103. Lauder, J.M. (19931 Neurotransmitters as growth regulatory signals: role of receptors and second messengers. Trends Neurosci. 6, 233-240.

Layer, P.G. (1991) Cholinesterase during development of avian nervous system. Cell. Mol. Neurobiol. 11: 7-33. Lipton, S.A. and Kater, S.B. (1989) Neurotransmitter regulation of neuronal outhgrowth, plasticity and survival. Trends Neurosci. 12: 265-270. Lowry, O.H., Rosebrough, N.J., Farr, A. and Randall, R. (1951) Protein measurement with the Folin phenol reagent. J. Biol. Chem., 193: 265-275. Massoulit, J., Pezzamenti, L., Bon, S., Krejci, E. and Vallete, F.M. (1993) Molecular and cellular biology of cholinesterases. Prog. Neurobiol., 41: 31-91. McDonald, J.W., Trescher, W.H. and Johnston, M.V. (1991) The pattern and degree of selective vulnerability to excitoxic brain injury is dependent upon developmental age. In: Excitatory Amino Acids (B.S. Meldrum et al., eds.) p. 609-614, Raven Press: N.York. Millard, C.B. and Broomfield, C.A. (1995) Anticholinesterases: Medical Applications of neurochemical Principles. J. Neurochem. 64: 19091918. Romano, C., Price, M.T. and Olney, W. (1995) Delayed excitotoxicity neurodegeneration induced by excitatory amino acid agonists in isolated retina. J. Neurochem. 65: 59-67. Rodriguez-Ithurralde, D. and Vincent, 0. (1994) Excitotoxicity and motoneuronal chemical markers during motor neurone (MN) programmed death. .I. Neural. Sci., 124 (Suppl.): 52-53. Rodriguez-Ithurralde, D., Silveira, R., Barbeito, B. and Dajas, F. (1983) Fasciculin, a powerful anticholinesterase polypeptide from Dendroaspis angusticeps venom. Neurochem. Int., 5: 267-274. Rodriguez-Ithurralde, D., Olivera, S., Migues, V., Vincent, 0. and Salazar, R. (1995) Glutamate-receptor elicited acetylcholinesterase release in the mouse spinal cord slice, a model of early excitotoxic injury. J. Neurol. Sci., 129 (Suppl): 104-106. Rothstein, J.D., Jin, L., Dykes-Hoberg, M. and Kuncl, R.W. (19931 Chronic inhibition of glutamate uptake produces a model of slow neurotoxicity. Proc. Natl. Acad. Sci. USA., 90: 6591-6595. Rothstein, J.D. and Kuncl, R.W. (1995) Neuroprotective strategies in a model of chronic glutamate-mediated motor neuron toxicity. J. Neurochem., 65: 643-651. Sapriza, C., Yannicelli, B., Vincent, 0. and Rodriguez-Ithurralde, D. (1992) Nuevas funciones para un fen6meno generalizado: secreci6n de acetilcolinesterasa en el sistema nervioso central de 10s vertebrados. Bol. Sot. Zool. Uruguay, 7: 90. Taylor, S.J., Jones, S.A., Haggblad, J. and Greenfield, S.A. (1990) ‘On-line’ measurement of acetylcholinesterase release from the substantia nigra of the freely-moving guinea-pig. Neuroscience, 37: 71-76. Tiille, T.R., Berthele, A., Zieglglnsberger, W., Seeburg, P.H. and Wisden, W. (1995a) Flip and flop variants of AMPA receptors in the rat lumbar spinal cord. Eur. J. Neurosci., 7: 1414-1419. Tiille, T.R., Berthele, A., Laurie, D.J., Seeburg, P.H. and Zieglgansberger, W. (1995b) Cellular and subcellular distribution of NMDARl splice variant mRNA in the rat lumbar spinal cord. Eur. J. Neurosci., 7: 1414-1419. Yannicelli, B., Sapriza, C., Vincent, 0. and Rodriguez-Ithurralde, D. (1992) Molecular forms of AChE released by central neurons (Spanish). Bol. Sot. Zool. Uruguay, 7: 96. Yee, D.K., Pastuszko, A., Nelson, D. and Wilson, D.F. (1989) Effects of o,L-2-amino-5-phosphonovalerate on metabolism of catecholamines in synaptosomes from rat brain. J. Neurochem., 52, 54-60.