¹Dept of Experimental and Applied Pharmacology, University of Pavia, Italy

²Pharma Research CNS, Bayer AG, Wuppertal, Germany

Correspondence to: M Racchi, Dept of Experimental and Applied Pharmacology, University of Pavia, Viale Taramelli 14, 27100 Pavia, Italy. E-mail: racchi@unipv.it

We have investigated the acute and chronic effect of metrifonate (MTF) and dichlorvos (DDVP), respectively the prodrug and active acetylcholinesterase inhibitor, on the secretory processing of the amyloid precursor protein (APP) in SH-SY5Y neuroblastoma cells. We demonstrate that the acute treatment of SH-SY5Y cells with both compounds results in an increased secretion of the soluble fragment of APP (sAPP) into the conditioned media of cells, with a pattern correlated to the level of acetycholinesterase inhibition. The regulation of APP processing in these conditions is mediated by an indirect cholinergic effect on muscarinic receptors, as demonstrated by inhibition with atropine. We have also followed APP expression and metabolism after long-term treatment with metrifonate. Treated cells showed reduced AChE activity after 24, 48 h and also following 7 days of repeated treatment, a time point at which increased AChE expression was detectable. At all time points sAPP release was unaffected suggesting that enhanced sAPP release by MTF is transitory, nevertheless the sensitivity of cholinergic receptors was unchanged, as indicated by the fact that cholinergic response can be elicited similarly in untreated and treated cells. APP gene expression was unaffected by long-term AChE inhibition suggesting that increased short-term sAPP release does not elicit compensatory effects.

Molecular Psychiatry (2001) 6, 520-528.

amyloid precursor protein; cholinergic transmission; acetylcholinesterase inhibitors; neuroblastoma cell lines; protein kinase C

Introduction

Alzheimer's disease (AD) is a progressive neurodegenerative disorder that affects a large proportion of the elderly population. Although the involvement of several neurotransmitter systems has been demonstrated in this disease, interest has been focussed on the decline in the cholinergic system. As of today the current available therapy for AD consists of the administration of acetylcholinesterase (AChE) inhibitors.

The formation and growth of brain amyloid lesions, the senile plaques, is widely believed to be a crucial event in the pathogenesis of Alzheimer's disease (AD).¹ The primary constituent of amyloid plaques is a 39-43 aminoacid hydrophobic peptide, named A, which is produced as the cleavage product of a much larger protein, the amyloid precursor protein (APP).^2,3 Cleavage of APP is complex and can be regulated by several protease activities. The pathways that generate A (reviewed in Racchi and Govoni⁴) are regulated by the sequential action of two enzymes, named and -secretase. Alpha secretase, instead, cleaves APP within the A sequence and releases a soluble N-terminal fragment, termed sAPP,^5,6 a process that constitutes the so-called non amyloidogenic pathway.

Extensive studies on APP processing established that the two proteolytic pathways may be balanced⁴ and that the stimulation of sAPP release is associated with reduced formation of amyloidogenic peptides. This hypothesis has been demonstrated, for example, for muscarinic-agonist-induced sAPP secretion. In human embryonic kidney (HEK) cell lines stably transfected with individual muscarinic receptor subtype, activation by carbachol of M1 and M3, but not M2 and M4 receptor subtypes increased sAPP release.⁷ Stimulation of sAPP secretion from M1 receptor-expressing HEK cells by carbachol was accompanied by a decrease in the release of A fragments,⁸ suggesting that these agents activate a pathway that cleaves APP within the A domain and hence might prevent amyloid formation. Furthermore in APP-transfected HEK293 and in teratocarcinoma (NT2)-derived differentiated human neuron cultures, NT2N, muscarinic receptor-mediated sAPP release diminishes A levels.^9,10 A cholinergic effect on sAPP secretion has been also demonstrated for cholinesterase inhibitors. It has been shown that superfused rat cortical slices release sAPP in response to stimulation by physostigmine, eptastigmine and dichlorvos.^11,12 Conversely, the secretion of sAPP in a number of cell lines was inhibited using the anticholinesterase drug tacrine,¹³ however, this work can lead to confounding interpretations in this context since tacrine is not only an AChE inhibitor, but also has an effect on multiple neurotransmitter systems.

Metrifonate [O,O-dimethyl-(1-hydroxy-2,2,2-tri- chloroethyl)-phosphonate] or MTF is a long-acting and well tolerated cholinesterase inhibitor that was reported to improve cognitive function and behavioural deficits in AD patients. This drug non enzymatically dehydrochlorinates in aqueous alkaline solutions and spontaneously rearranges to the active compound dichlorvos (2,2-dichlorovinyl-dimethyl-phosphate) or DDVP.¹⁴ This is the product that inhibits AChE, by a competitive interaction with the catalytic site of the enzyme leading to irreversible inhibition within minutes.

We have investigated the possibility that metrifonate and dichlorvos may modulate APP metabolism exploring the response to drugs in the SH-SY5Y neuroblastoma cells, after short- and long-term treatment. While short-term treatment with either MTF or DDVP increases the secretion of sAPP, with a pattern consistent with the AChE inhibitory effect of the compounds and with an activity due to indirect cholinergic stimulation, we observed that long-term treatment does not result in changes in APP mRNA expression and in downregulation of the response to cholinergic stimulation of endogenous muscarinic receptors, in the presence of an increased expression of AChE at the longest time points of MTF exposure.

Materials and methods

Materials

Culture media, foetal calf serum (FCS), supplements and molecular biology reagents were obtained from GIBCO Life Technologies (Paisley, UK). Electrophoresis reagents were obtained from Bio-Rad (Hercules, CA, USA). Metrifonate [O,O-dimethyl- (1-hydroxy-2,2,2-trichloroethyl)-phosphonate] and dichlorvos (2,2-dichlorovinyl-dimethyl phosphate) were a kind gift of Dr Bernard H Schmidt. All other reagents were of the highest grade available and were purchased from Sigma Chemical Co (St Louis, MO, USA) unless otherwise specified.

SH-SY5Y neuroblastoma cells

Cells were cultured in minimum essential medium (MEM) supplemented with 10% FCS, penicillin (100 U m^-1), streptomycin (100 g m^-1), Na pyruvate (1 mM) at 37°C in a 5% CO₂ incubator. Cells were seeded at a density of 1 ´ 105 cm^-2 in 60-mm dishes for the short-term treatment and in 100-mm dishes for the long-term treatment, and used the day after plating.

Experimental treatments

Short-term treatment: Confluent monolayers of cells were washed twice with PBS and treated for 2 h at 37°C in serum-free MEM in the presence of vehicle alone or in the presence of the compound of interest. Metrifonate (MTF), dichlorvos (DDVP), carbachol and atropine were dissolved directly in serum-free MEM to working concentrations (see Results) at the time of use.

Phorbol 12, 13 dibutyrate (PdBu) and GF-109203X were prepared as stock solution (5 mM and 2.5 mM respectively) in DMSO and diluted in serum-free MEM at the time of use.

Long-term treatment: For this set of experiments, the cells were treated with metrifonate 100 M or vehicle alone in MEM + 10% FCS according to a pulsed protocol as depicted in Figure 1. Briefly, cells were washed twice with PBS and treated with the compounds of interest for 2 h (pulse). Following this step cells were washed with PBS, and fresh MEM + 10% FCS without additions was replaced and the cells left to recover for 22 h. The same pattern was repeated twice to perform experiments including two pulse periods and a total of 48 h of AChE inhibition. Also we performed an experiment in which the same scheme was repeated for 7 days. At the end of each incubation cells were washed twice with PBS and treated for 2 h at 37°C in serum-free MEM (with or without addition of test compounds), that was finally collected for the analysis of sAPP release by Western blot. Finally we repeated the pulsed treatment for 7 consecutive days, determining at the end of treatment the levels of AChE activity, the levels of AChE protein and the modulation of APP metabolism as detailed below.

Preparation of the conditioned media

The conditioned media were collected after 2 h of incubation and centrifuged at low speed to remove debris and detached cells. Proteins from conditioned media were prepared by DOC (deoxycholate)/TCA (trichloroacetic acid) precipitation.¹⁵

Immunodetection of sAPP

Normalization of protein loading on each blot was obtained by loading an equal volume of samples of conditioned medium standardized to total cell lysate protein concentration. Proteins were subjected to SDS/PAGE (10% gel) and then transferred to polyvinylidene difluoride (PVDF) membranes (NEN-Dupont, Boston, MA, USA). For the detection of the secreted APP, either the monoclonal antibodies 22C11 (Boehringer-Mannheim, Mannheim, Germany) or 6E10 (Senetek, St Louis, MO, USA) were used, and the blots were incubated overnight at 4°C. Detection was carried out by incubation with alkaline phosphatase-conjugate goat anti mouse IgG (Kirkegaard and Perry Labs, Gaithersburg, MD, USA) for 1 h. The blots were then washed extensively, and sAPP was visualized using an enhanced chemiluminescence method (NEN). The same immunoreactive bands detected by the antibody 22C11 were detected in Western blots by the antibody 6E10 and were also immunoprecipitated by the antiserum ER31-16 (M Racchi, unpublished). Both 6E10 and ER31-16 recognize epitopes in the first 16 aminoacids of A, a site that constitutes the C-terminal of sAPP cleaved at the ' site', therefore the identified bands can be assumed to be the ' secretase' cleaved form of sAPP.

Preparation of RNA and cDNA

The confluent monolayers were washed twice with TBS (20 mM Tris-HCl (pH 7.5), 500 mM NaCl) and then cells were lysed directly in a culture dish by adding 1 ml of TRIZOL reagent (Gibco), a mono-phasic solution of phenol and guanidine isothiocyanate, and passing the cell lysates several times through a pipette. Chloroform (0.2 ml) was then added to the homogenized samples. The samples were centrifugated to separate the aqueous phase and an organic phase. After transfer of the aqueous phase, the RNA was recovered by precipitation with isopropyl alcohol. Two micrograms of RNA were reverse transcribed by Superscript II RNase H Reverse Transcriptase (Gibco). Reverse transcription was carried out in a final volume of 20 l containing 1 l of oligo-(dT) primer (0.5 g ml^-1), 4 l buffer (250 mM tris-HCl pH 8.3, 375 mM KCl, 15 mM MgCl₂), 2 l DTT 0.1 M and distilled water to volume. Following an initial heating at 70°C for 10 min of RNA and oligo-(dT), the other compounds were added and samples were heated at 42°C for 50 min and at 70°C for 15 min. One microlitre of samples (100 ng of cDNA) was used for the PCR.

Polymerase chain reaction for APP

PCR was carried out in a 25 l volume containing 100 ng cDNA, 2.5 l of PCR buffer (100 mM Tris-HCl pH 8.3, 500 mM KCl), 1.5 mM MgCl₂, 200 M of each dNTP, 20 pmol of the forward primer h-APP-F (5'-CACCACAGAGTCTGTGGAAGA-3'), 20 pmol of the reverse primer h-APP-R (5'-AGGTGTCTGAGAT ACTTGT-3') and 2.5 U Ampli Taq Gold (Perkin Elmer). PCR was carried out in a GeneAmp PCR System (Perkin Elmer). Following an initial denaturation of 10 min at 95°C, samples were amplified for 30 cycles, each consisting of 30 s at 95°C, 30 s at 58°C and 30 s at 72°C. After a final extension step of 5 min at 72°C, amplification products were electrophoresed on 3% agarose (Sigma) gel dissolved in TAE (40 mM tris-acetate, 2 mM EDTA) buffer. The DNA bands were visualized with ethidium bromide (0.5 g ml^-1, Sigma) staining and UV illumination. Semiquantitative estimates of band density were obtained with a CCD camera coupled to a Macintosh computer running NIH Image software.

Polymerase chain reaction for GPDH (glucose 6-phosphate dehydrogenase)

PCR reactions were performed using 10 ng of cDNA, 2.5 l of PCR buffer (100 mM Tris-HCl pH 8.3, 500 mM KCl), 3.5 mM of MgCl₂, 200 M of each dNTP, 0.5 M h-GPDH-R (5'-CCACCCATGGCAATTCCATGGCA-3'), 0.5 M h-GPDH-F (3'-TCTAGACGGCAGGTCAG GTCCACC-5') and 2.5 U Ampli Taq Gold (Perkin Elmer) in a final volume of 25 l. Samples were subjected to one cycle of 10 min at 95°C, then to 35 cycles of 30 s at 95°C, 30 s at 60°C, and 1 min at 72°C. After a final extension step of 7 min at 72°C, amplification products were electrophoresed on 1% agarose (Sigma) gel dissolved in TAE (40 mM tris-acetate, 2 mM EDTA) buffer. The DNA bands were visualized with ethidium bromide (0.5 g ml^-1, Sigma) staining and UV illumination.

Determination of acetylcholinesterase activity

Following the recovery of conditioned media the cell monolayers were washed with phosphate-buffered saline and cells were scraped and collected in phosphate buffer pH 8.0. The cell suspension was kept on ice and used for the determination of acetylcholinesterase activity according to Ellman.¹⁶ An aliquot of the cell suspension was dissolved in NaOH 1 M and following neutralization, protein concentration was determined using the method of Bradford (BioRad protein assay kit). The levels of expression of AChE were determined by Western blot as described above using an antibody for AChE from Chemicon according to the manufacturer's instructions.

Densitometric and statistical analysis

The relative density of immunoreactive bands on Western blots was calculated following acquisition of the blot image through a Nikon CCD video camera module and analysis by means of the Image 1.47 program (Wayne Rasband, NIH, Research Services Branch, NIMH, Bethesda, MD, USA). Normalization between blots was achieved by dividing each individual peak area by the average peak area of control samples within the respective blot, or within the respective group of treatments. Data were analyzed using the analysis of variance test followed, when significant, by an appropriate post hoc comparison such as the Dunnet's t-test; a P value <0.05 was considered significant. The data reported are expressed as the mean ± SD of triplicate/ quadruplicate samples in two/three independent experiments.

Results

Characterization of the effect of MTF and DDVP on the release of sAPP

The SH-SY5Y human neuroblastoma cell line expresses choline acetyltransferase activity, acetylcholinesterase activity and cholinergic receptors.¹⁷ Cells treated with carbachol release sAPP in a concentration-dependent manner and at a concentration of 1 mM reach levels approximately twice that of the basal release (Figure 2). The pattern of sAPP released by SH-SY5Y cells is that of a doublet band with apparent molecular masses of 110 and 130 kDa, consistent with other reports and consistent with the pattern of expression of the major APP isoforms in this cell line (see below).^13,18,19 Treatment of the cells for 2 h with MTF results in a similar concentration-dependent increase in sAPP release and a MTF concentration of 100 M promotes an increase of sAPP release quantitatively similar to that elicited by 1 mM carbachol (Figures 2 and 3). Because MTF in aqueous solution decomposes spontaneously originating the active AChE inhibitor dichlorvos (DDVP), we treated the SH-SY5Y neuroblastoma cells with DDVP (10 nM-1 M) for 2 h. Similarly to what was observed with MTF, DDVP promoted the non amyloidogenic metabolism of APP with a maximal effect observed for a concentration of 1 M, corresponding to sAPP secretion of 2.7-fold above basal (Figure 3).

We evaluated the AChE activity on the same set of cells used to characterise the effect of MTF on sAPP release. The data demonstrate that the increase in sAPP release was inversely related to the levels of AChE activity. In particular the concentration of 100 M MTF, that elicited the maximal effect on sAPP secretion, was associated with a 75% inhibition of AChE activity (Figure 3).

Pharmacological characterization of the cholinergic effect of MTF

As suggested by the data described above, the mechanism of action of MTF through its active metabolite DDVP is the inhibition of AChE, thus increasing the availability of ACh that, through muscarinic receptors, may modulate APP metabolism. We treated the cells with 100 M MTF in the presence or absence of 10 M atropine. The effect of MTF was inhibited by simultaneous treatment with the muscarinic receptor antagonist atropine, that was able to inhibit the sAPP release induced by MTF to basal level. These data support an indirect cholinergic receptor-mediated effect of MTF, and therefore of DDVP, on sAPP release. The effect of MTF on APP metabolism was also blocked by GF-109203X, a specific inhibitor of protein kinase C (PKC) (Figure 2) suggesting that MTF promotes the secretion of non amyloidogenic APP fragments by an indirect cholinomimetic effect coupled to protein kinase C.

Long-term effect of pulsed treatment with MTF on AChE activity, sAPP release and APP mRNA expression

To determine the effect of prolonged AChE inhibition on parameters related to APP metabolism and expression, we treated the cells with a pulse of 2 h in the presence of metrifonate 100 M or vehicle alone. The media were replaced with fresh MEM + 10% FCS and the cells allowed to recover for 22 h. The analysis of sAPP release and the measurements of AChE inhibition were carried out after the 24-h cumulative period as described in Methods. The level of residual AChE activity 24 h following treatment with MTF was 37% ± 23 (SD) of the activity of untreated cells, corresponding to an inhibition of AChE activity averaging 63%, a level comparable to that described in short-term experiments. The levels of AChE activity returned to normal if the cells were left to recover for additional 24 h (not shown). In a second set of experiments we repeated twice the treatment with MTF or vehicle for 2 h followed by replacement with fresh MEM + 10% FCS for 22 h. At the 48-h time point the residual AChE activity was 20% ± 14 (SD) of the activity of untreated cells, corresponding to an inhibition averaging 80%.

The long-term inhibition of AChE activity did not modify the constitutive metabolism of APP. Following both treatment protocols, cells were incubated in serum-free MEM and the levels of sAPP in the conditioned medium analysed as described previously. Basal release of sAPP was unaffected in cells treated either with one or two pulses of MTF at 24 and 48 h respectively (Figure 4).

Confluent monolayers of vehicle- or MTF-treated cells were lysed, RNA was purified and used for RT-PCR analysis of APP and GPDH (see Methods). As shown in Figure 4c, the specific primers used amplify three cDNA fragments of 310, 253, 86 bp, corresponding respectively to APP mRNAs containing exon 7 (encoding the KPl domain) and exon 8, isoforms containing only exon 7, and APP mRNAs where both exons are spliced out. The figure shows that the prolonged AChE inhibition resulting from pulsed treatment with MTF did not modify the expression of the three isoforms of APP. Artefacts due to unequal concentrations of RNA during RT-PCR were ruled out by the analysis of GPDH expression. Equal ratio of APP/GPDH signals was obtained from cells treated either with one or two pulses of MTF at 24 and 48 h respectively.

Effect of long-term AChE inhibition on the expression of AChE and the modulability of APP metabolism

Cells treated with either of the long-term protocols show unmodified parameters concerning expression of APP mRNA and constitutive APP metabolism. We addressed the possibility that prolonged AChE inhibition could alter the physiological and pharmacological modulability of APP metabolism. Following either a single or a double pulse of MTF and a protocol of 7 consecutive days of treatment, we challenged vehicle- and MTF-treated cells with 300 nM PdBu, a direct PKC activator, and with 1 mM carbachol. We have obtained the same results for all treatment protocols and Figure 5 shows the results of double pulse at 48 h and at 7 days of treatment. These data demonstrate that both types of treatment were able to elicit responses, in terms of sAPP release, similar to those obtained from naive cells, suggesting that the cholinergic receptor and PKC-dependent modulation of the non amyloidogenic APP secretory pathway were unaffected by the prolonged cholinesterase inhibition. In the samples exposed for 7 days the final activity of AChE was 19.8% ± 13 (corresponding to an inhibition averaging 80%) of control cells and at this time point we observed that the expression of AChE was increased in the treated cells (Figure 6).

Discussion

The current available strategies for AD therapy are substantially limited to drugs that improve central cholinergic neurotransmission.²⁰ Acetylcholinesterase inhibitors (AChEI) prevent the hydrolysis of released acetylcholine, increasing the efficiency of cholinergic transmission.²¹ This action is supposed to control the key symptoms of Alzheimer's disease, namely memory and cognitive problems. These are the main functions of an inhibitor of AChE providing potentiation of the cholinergic transmission. On the other hand the literature suggest that cholinergic activities may also be involved in the regulation of APP metabolism.

A key paper in 1992 by Nitsch and coworkers⁷ demonstrated that cholinergic agonists could also affect AD pathogenesis by modulating APP metabolism. It was demonstrated that treatment of M1 and M3 receptor-transfected cells with carbachol resulted in a stimulated release of sAPP that was dependent on ligand-receptor interaction and on protein kinase C (PKC) activation. Electrical depolarization of brain slices provided stronger support to the idea that neuronal activity could determine a modulation of sAPP release. Superfused slices from rat cortex and hippocampus released sAPP upon electrical stimulation, in a frequency-dependent, tetradotoxin-sensitive manner.²² The model of brain slices helped to demonstrate that cholinergic stimulation elicited an increase in sAPP release, with a pharmacological profile suggestive of a complex interaction between different muscarinic receptor subtypes. In fact nonspecific cholinergic activation with carbachol was effective in stimulating sAPP release only in the presence of the M2 subtype receptor antagonist gallamine or with a selective M1 receptor agonist.²² In the same system an increased sAPP release was also obtained following treatment with acetylcholinesterase inhibitors.¹¹ These observations may have important pharmacological implications in AD. In fact the AChE inhibition would have additional advantages in correcting intrinsically defective APP metabolism, possibly modifying the progression of the illness in addition to the effect on the immediate symptom of cholinergic deficiency. This hypothesis claims attention to the short- and long-term effects of AChE inhibition, that we addressed using as reference molecule the prodrug metrifonate (MTF) and its active metabolite dichlorvos (DDVP), a compound that has demonstrated clinical efficacy toward cognitive and behavioural functions in patients with Alzheimer's disease.^23,24

In our investigation we have observed that the short-term treatment with metrifonate or its active product dichlorvos can affect APP metabolism by increasing the rate of -secretase processing, resulting in the stimulation of sAPP release from SH-SY5Y neuroblastoma cells. The treatment with DDVP elicited a maximal increase of sAPP secretion at a concentration (1 M) of two orders of magnitude lower than that of MTF (100 M), in agreement with the difference in potency toward AChE inhibition, as demonstrated in other studies on MTF and DDVP pharmacology.¹⁴ The measurements of AChE activity in the presence of metrifonate revealed an inverse relationship between the levels of AChE inhibition and the levels of sAPP released. The difference in potency observed for the two drugs, and the apparent inverse correlation between AChE activity and sAPP release, suggest that the effect of metrifonate on APP processing was dependent on its conversion to dichlorvos and thus on anti-ChE activity. Similar results concerning the effect of MTF have been recently communicated by Pakaski et al,²⁵ suggesting also a short-term effect of MTF on the levels of PKC.

APP processing can be modulated by stimulation of muscarinic receptors, both via PKC-dependent and -independent mechanisms.^7,26 We provide evidence that the metrifonate-induced secretory metabolism of APP is sensitive to inhibition by atropine, the non specific muscarinic receptor antagonist. These results suggest an indirect receptor-mediated effect of metrifonate, and therefore of dichlorvos, on the proteolitic pathway regulated by -secretase. Furthermore our experiments carried out treating the cell line with MTF in the presence or absence of the specific inhibitor of PKC, indicate that this treatment promotes an increase of APP secretion by a PKC-dependent mechanism.

In many cell types the increase in sAPP secretion (mostly demonstrated by PKC activation) is paralleled by a reduction of A production, supporting the concept that the two pathways are reciprocally regulated. This is the case also for the muscarinic agonist-induced sAPP release that, in muscarinic receptor-transfected cells as well as in human NT2nN cells, is accompanied by a reduction of A release.¹⁰ The same appears to be true for tacrine that reduces the levels of A secreted by SH-N-SK neuroblastoma cells.²⁷ We were unable to measure detectable levels of A secreted by untransfected SH-SY5Y cells following our experimental scheme, and further investigation on this matter is being addressed. However we choose to perform our experiments without manipulating the overexpression of APP, mainly because our primary goal was to address questions concerning the effect of AChEI on APP expression itself. Cholinergic modulation of APP metabolism occurs in rat fetal brain neurons,²⁸ however the regulation of APP metabolism was demonstrated to be dependent on the maturation in vitro of the neurons. Basal sAPP release and APP holoprotein levels increased over maturation of the cells in vitro reaching a plateau by 7 days. By that time the modulation of sAPP release by carbachol was no longer evident. The findings introduce the concept of saturability of the secretory mechanism for sAPP, a concept that is more clearly appreciated when studying cells derived from Down's syndrome (DS) individuals.

Fibroblasts from DS patients overexpressing APP at mRNA and holoprotein levels, secrete sAPP approximately twice as much as fibroblasts from normal individuals. These cells, likely because of a saturation of the secretory mechanism, do not respond with increased sAPP release when stimulated with phorbol esters.²⁹

Almost all the experiments on in vitro pharmacology⁴ are focussed on the acute effect of several drugs on APP metabolism. The results of the second part of our work were specifically designed to address the effect of chronic treatment with a drug on APP metabolism. In particular we have investigated the effect of long-term AChE inhibition following treatment with MTF/DDVP, by exploring its effects at different times of treatment on APP mRNA expression, as well as on APP basal and pharmacologically regulated metabolism. Finally we also addressed the possibility that prolonged exposure to AChE inhibitors may affect the expression of AChE itself.

Since the action of MTF/DDVP is to irreversibly inhibit AChE,¹⁴ we set up experiments where a pulse (2 h) of metrifonate was given to the cells and then the effects on APP metabolism, AChE activity and APP expression, were explored after a determined time period (22 h). In previous experiments we have observed that AChE activity recovers within 48 h after the first addition of the drug, so we persisted with a second pulse of the drug, ie, the protocol was repeated once (24 h pulsed) or twice (48 h pulsed). The results of both pulsed experiments demonstrate that, as expected, the kinetic of inhibition of AChE by DDVP shows that the inhibition is maintained at 24 and 48 h, at the end of the experimental treatment. Our concern in these sets of experiments was to investigate whether the treatment had any effect on the levels of expression of APP mRNA and the results suggested that no significant alteration of APP mRNA levels can be observed either at 24 or 48 h. Another concern, possibly the most important of our investigation, was to determine whether or not the prolonged AChE inhibition would affect the physiological and pharmacological modulability of APP metabolism. Toward this goal we have also set up an experiment where cells were exposed to MTF in a pulsed protocol for 7 consecutive days, to increase further the time of exposure and reproduce more closely (as much as an in vitro cellular model can do) the in vivo situation. The prediction from preliminary data was that AChE expression would actually be increased and this was the case at 7 days of treatment, however AChE inhibition was still substantial. Even in these conditions, as well as 24 and 48 h, we demonstrate that the normal physiological and pharmacologically modulated metabolism of APP is not altered by long-term AChE inhibition. The acute effect demonstrated after 2 h of MTF/DDVVP treatment appears indeed to be transitory, possibly because of an homeostatic adaptation of the cells to the effects of AChE inhibition, and no further accumulation of sAPP occurs after the short-term treatment as evidenced by our pulsed experimental scheme. Another possible effect of long-term AChE inhibition could have been the desensitization of cholinergic receptors and therefore a blunting of the 'cholinergic'-induced non amyloidogenic APP metabolism. We demonstrate that the pharmacological modulation of sAPP release is not affected negatively by long-term AChE inhibition, because at 24 h (not shown), at 48 h and at 7 days, either phorbol esters and more importantly carbachol can elicit secretory responses similar to those elicited in naive cells. As seen in Figure 6, the relative ratio of response to carbachol and phorbol esters vs basal release produces similar results.

In conclusion our studies demonstrate that AChE inhibition by metrifonate can acutely promote the non amyloidogenic metabolism of APP by an indirect cholinergic effect through an intracellular mechanism coupled to protein kinase C. The results described in the second part of our work are the first to describe the effect of a long-term pharmacological treatment on APP metabolism.

The results cannot directly link the effect of AChE inhibition in patients to a positive modulation of APP metabolism, rather suggest that for AChE inhibition induced by metrifonate the acute effect is probably transitory. However the data indicate that a negative influence of prolonged AChE inhibition on the modulability of APP metabolism is unlikely, because even after 7 days of treatment its physiology and pharmacology remain unaffected.

Acknowledgements

This research was partially funded by MURST Cofin99, 'Fondo Ateneo Ricerca' and 'Progetto Ateneo' University of Pavia, Telethon E866 to SG.

1 Selkoe DJ. Amyloid beta-protein and the genetics of Alzheimer's disease. J Biol Chem 1996; 271: 18295-18298, MEDLINE

2 Tanzi RE, McClatchey AI, Lamperti ED, Villa-Komaroff L, Gusella JF, Neve RL. Protease inhibitor domain encoding by an amyloid protein precursor mRNA associated with Alzheimer disease. Nature 1988; 331: 115-126,

3 Kang J, Lemaire HG, Unterbeck A, Salbaum JM, Masters CL, Greschik KH et al. The precursor of Alzheimer's disease amyloid A4 protein resembles a cell surface receptor. Nature 1987; 325: 733-736, MEDLINE

4 Racchi M, Govoni S. Rationalizing a pharmacological intervention on the amyloid precursor protein metabolism. TIPS 1999; 20: 418-423,

5 Esch FS, Keim PS, Beattie EC, Blacher RW, Culwell AR, Oltersdorf T et al. Cleavage of amyloid -peptide during constitutive processing of its precursor. Science 1990; 248: 1122-1124, MEDLINE

6 Anderson JP, Esch FS, Keim PS, Sambamurti K, Lieberburg I, Robakis NK. Exact cleavage site of Alzheimer precursor protein in neuronal PC 12 cells. Neurosci Lett 1991; 128: 126-128,

7 Nitsch RM, Slack BE, Wurtman RJ, Growdon JH. Release of Alzheimer amyloid precursor derivatives stimulated by activation of muscarinic acetylcholine receptors. Science 1992; 258: 304-307, MEDLINE

8 Hung AY, Haass C, Nitsch RM, Qiu WQ, Citron M, Wurtman RJ et al. Activation of protein kinase C inhibits cellular production of the amyloid beta-protein. J Biol Chem 1993; 268: 22959-22962,

9 Buxbaum JD, Koo EH, Greengard P. Protein phosphorylation inhibits production of Alzheimer amyloid A4 peptide. Proc Natl Acad Sci USA 1993; 89: 3050-3059,

10 Wolf BA, Wertkin AM, Jolly YC, Yasuda RP, Wolfe BB, Konrad RJ et al. Muscarinic regulation of Alzheimer's disease amyloid precursor protein secretion and amyloid -protein production in human neuronal NT2N cells. J Biol Chem 1995; 270: 4916-4922,

11 Mori F, Lai CC, Fusi F, Giacobini E. Cholinesterase inhibitors increase secretion of APPs in rat brain cortex. Neuroreport 1995; 6: 633-636,

12 Giacobini E. Cholinesterase inhibitors do more than inhibit cholinesterase. In: Becker R, Giacobini E (eds). Alzheimer's Disease: From Molecular Biology to Therapy. Birkhauser: Boston, 1996, pp 187-204.

13 Lahiri DK, Lewis S, Farlow MR. Tacrine alters the secretion of the beta-amyloid precursor protein in cell lines. J Neurosci Res 1994; 37: 777-787,

14 Hinz VC, Grewing S, Scmitdt BH. Metrifonate induces cholinesterase inhibition exclusively via slow release of dichlorvos. Neurochem Res 1996; 21: 331-337,

15 Bergamaschi S, Binetti G, Govoni S, Wetsel WC, Battaini F, Trabucchi M et al. Defective phorbol ester-stimulated secretion of beta-amyloid precursor protein from Alzheimer's disease fibroblasts. Neurosci Lett 1995; 201: 1-5,

16 Ellman GI, Courtney KD, Andres VJ, Featherstone RM. A new, rapid colorimetric determination of acetylcholinesterase activity. Biochem Pharmacol 1961; 7: 88-95,

17 Ehrich M. Using neuroblastoma cell lines to address differential specificity to organophosphates. Clin Exp Pharmacol Physiol 1995; 22: 291-292,

18 Weidemann A, Konig G, Bunke D, Fischer P, Salbaum JM, Masters CL et al. Identification, biogenesis, and localization of precursors of Alzheimer's disease A4 amyloid protein. Cell 1989; 57: 115-126, MEDLINE

19 Ganter U, Strauss S, Jonas U, Weidemann A, Beyreuther K, Volk B et al. Alpha 2-macroglobulin synthesis in interleukin 6 stimulated human neuronal (SH-SY5Y neuroblastoma) cells. Potential significance for the processing of Alzheimer beta amyloid precursor protein. FEBS Lett 1991; 282: 127-131,

20 Knopman DS, Morris JC. An update on primary drug therapies for Alzheimer disease. Arch Neurol 1997; 54: 1406-1409,

21 Benzi G, Moretti A. Is there a rationale for the use of acetylcholinesterase inhibitors in the therapy of Alzheimer's disease? Eur J Pharmacol 1998; 346: 1-13,

22 Farber SA, Nitsch RM, Schulz JG, Wurtman RJ. Regulated secretion of -amyloid precursor protein in rat brain. J Neurosci 1995; 15: 7442-7451,

23 Morris JC, Cyrus PA, Orazem J, Mas J, Bieber F, Ruzicka BB et al. Metrifonate benefits cognitive, behavioural, and global function in patients with Alzheimer's disease. Neurology 1998; 50: 1222-1230,

24 Schmidt BH, Roland Heinig R. The pharmacological basis for metrifonate's favourable tolerability in the treatment of Alzheimer's disease. Dementia and Geriatric Cognitive Disorders 1998; 9 (Suppl 2): 15-19,

25 Pakaski M, Rakonczay Z, Fakla I, Papp H, Kasa P. In vitro effects of metrifonate on neuronal amyloid precursor protein processing and protein kinase C level. Brain Res 2000; 863: 266-270,

26 Slack BE, Breu J, Petryniak MA, Srivastava K, Wurtman RJ. Tyrosine phosphorylation-dependent stimulation of amyloid precursor protein secretion by the m3 muscarinic acetylcholine receptor. J Biol Chem 1995; 270: 8337-8344,

27 Lahiri DK, Farlow MR, Sambamurti K. The secretion of amyloid beta-peptides is inhibited in the tacrine-treated human neuroblastoma cells. Brain Res Mol Brain Res 1998; 62: 131-140,

28 Salvietti N, Cattaneo E, Govoni S, Racchi M. Changes in beta amyloid precursor protein secretion associated with the proliferative status of CNS derived progenitor cells. Neurosci Lett 1996; 212: 199-203,

29 Racchi M, Johnston JA, Flood FM, Cowbum RF, Govoni S. Amyloid precursor protein metabolism in fibroblasts from individuals with one, two or three copies of the amyloid precursor protein (APP) gene. Biochem J 1999; 338: 777-782,

Figure 1 Scheme of long-term experimental treatment with metrifonate. For the 7 days treatment the scheme described was repeated for 7 consecutive days without the collection of samples indicated at 26 h.

Figure 2 Metrifonate-induced sAPP release is dependent on the activation of muscarinic receptors. SH-SY5Y cells were treated for 2 h with 1 mM CCh in serum-free MEM in the absence (lane 2) or in the presence (lane 3) of 10 M atropine. Another set of cells was incubated for 2 h with 100 M MTF alone in serum-free MEM (lane 4) or in presence of 10 M atropine (lane 5) or 2.5 M GF-109203X (lane 6).

Figure 3 Effect of acute treatment with metrifonate (MTF) and dichlorvos (DDVP) on sAPP release and acetylcholinesterase activity. Cells were treated for 2 h with serum-free MEM alone or increasing concentrations of the compounds in serum-free MEM. Analysis of sAPP release was done by Western blot of proteins released into the conditioned media. The acetylcholinesterase activity was determined according to the Ellman method on cellular lysates. Results are expressed as a percent of basal (sAPP release) or of control (AChE activity) ± SD of three independent experiments. AChE activity has been measured on the same samples used for sAPP release following MTF treatment.

Figure 4 Effect of long-term treatment with 100 M MTF on sAPP basal release and AChE activity. (a) 24 h pulsed treatment: SH-SY5Y cells were incubated for 2 h in medium in presence or absence of 100 M metrifonate, then the media were replaced with fresh MEM + 10% FCS for 22 h. (b) 48 h pulsed treatment: the treatment for 2 h with MTF or vehicle alone followed by replacement with fresh MEM + FCS 10% for 22 h has been repeated twice. Proteins released into the conditioned media were collected, precipitated, and subjected to Western blot analysis. (c) Expression of APP isoforms is not modified by 24 h and 48 h pulsed treatment with 100 M metrifonate. The 'pulsed treatments' were carried out as described in Methods. The example of RT-PCR of mRNA for APP shows that the specific APP primers used amplify three cDNA fragments of 310, 253, 86 bp, corresponding to the alternatively spliced APP mRNA as indicated in the text. The results shown in the figure were confirmed in three independent experiments.

Figure 5 Effect of long-term AChE inhibition on the modulability of APP metabolism. Cells were treated for 2 h (pulse) with 100 M MTF (lanes 4-9) or with vehicle alone (lanes 1-3). The media were then replaced with fresh MEM + 10% FCS for 22 h. After this time we repeated another pulsed treatment (lanes 4-6) followed by replacement of fresh media for 22 h. We have also repeated the pulsed treatment for 7 consecutive days (lanes 7-9). At the end of the incubation times the cells were treated with serum-free MEM (lanes 1,4,7), 300 nM PdBu (lanes 2,5,8) or 1 mM carbachol (lanes 3,6,9) in serum-free MEM. Proteins released in the conditioned media were collected, precipitated and subjected to Western blot analysis. Protein loading was normalised separately in each treatment group (control vs MTF treatments)

Figure 6 Expression of AChE following prolonged treatment with MTF. Cells were treated for 7 consecutive days as described in Methods and elsewhere in the paper. After the last day of treatment cell lysates normalised for protein content were loaded on SDS PAGE and subjected to Western blot analysis.