Amyloid β proteins inhibit Cl-ATPase activity in cultured rat hippocampal neurons

Authors


Address correspondence and reprint requests to Chiyoko Inagaki, Department of Pharmacology, Kansai Medical University, Fumizono-cho 10–15, Moriguchi-City, Osaka 570–8506, Japan. E-mail: inagaki@takii.kmu.ac.jp

Abstract

Cl-ATPase in the CNS is a candidate for an outwardly directed neuronal Cl transporter requiring phosphatidylinositol-4-phosphate (PI4P) for its optimal activity. To test its pathophysiological changes in a phosphatidylinositol (PI) metabolism disorder, the effects of neurotoxic factors in Alzheimer's disease (AD), amyloid β proteins (Aβs), on the Cl-ATPase activity were examined using primary cultured rat hippocampal neurons. Amyloid β proteins (1–40, 1–42 and 25–35) concentration-dependently (1–100 nm) and time-dependently (from 1 h to 6 day) decreased Cl-ATPase activity and elevated intracellular Cl concentrations ([Cl]i), Aβ25–35 being the most potent. Addition of inositol or 8-Br-cyclic GMP completely reversed these Aβ-induced changes. The recoveries in enzyme activity were attenuated by an inhibitor of PI 4-kinase, 10 µm wortmannin or 20 µm quercetin, but not by a PI 3-kinase inhibitor, 50 nm wortmannin or 10 µm LY294002. The PI, PIP and PIP2 levels of the plasma membrane-rich fraction were lower in the Aβ-treated cells as compared with each control. In the Aβ-exposed culture, but not in control, stimulation by 10 µm glutamate for 10 min significantly increased fragmentation of DNA and decreased cell viability. Addition of inositol or 8-Br-cyclic GMP prevented the effect of Aβ-treatment on the neurotoxicity of glutamate. Thus, Aβs reduce neuronal Cl-ATPase activity, resulting in an increase in [Cl]i probably by lowering PI4P levels, and this may reflect a pre-apoptotic condition in early pathophysiological profiles of AD.

Abbreviations used
Aβs

amyloid β proteins

AD

Alzheimer's disease

Ara-A

adenine-9β-arabinoside

DMEM

Dulbecco's modified Eagle's medium

EA

ethacrynic acid

LDH

lactate dehydrogenase

MAPK

mitogen activated protein kinase

MES

2-(N-morpholino)-ethanesulfonic acid

MEM

modified Eagle's medium

MQAE

N-(6-methoxyquinolyl)-acetoethyl ester

PBS

phosphate-buffered saline

PI

phosphatidylinositol

PIP

phosphatidylinositol monophosphate

PI4P

phosphatidylinositol-4-monophosphate

PKA

AMP-dependent protein kinase

PKG

GMP-dependent protein kinase

PMSF

phenylmethylsulfonyl fluoride

TCA

trichloroacetic acid

TLC

thin-layer chromatography.

Neuronal Cl-ATPase is a candidate for outwardly directed active Cl transport systems (Inagaki et al. 1985; Shiroya et al. 1989), which allow hyperpolarizing responses to transmitter-operated Cl channel opening (Bonnet et al. 1996) by keeping intracellular Cl concentrations lower than expected from passive distribution (Inoue et al. 1991; Hara et al. 1992). As it requires phosphatidylinositol-4-phosphate (PI4P) for its optimal activity in reconstituted proteoliposomes (Zeng et al. 1994), regulation by phosphatidylinositol (PI) metabolism of in situ Cl-ATPase activity has been suggested. Lithium, an inhibitor of inositol monophosphatase, was found to decrease Cl-ATPase activity in cultured rat hippocampal neurons under the conditions of stimulated PI turnover (Yagyu et al. 1999). Furthermore, Cl-ATPase activity was found to be reduced in the brains of Alzheimer's disease (AD) patients (Hattori et al. 1998), in which PI 4-kinase activity is reportedly as low as 50% of that in age-matched non-AD brains (Wallace 1994), and the levels of PI and PIP (tendency) are decreased (Prasad et al. 1998).

Alzheimer's disease is a neurodegenerative disorder characterized by the formation of senile plaques, neurofibrillary tangles and selective neuronal loss (Selkoe 1991). The main components of the plaques, amyloid β proteins (Aβs) consisting of 39–43 amino acids (Glenner and Wong 1984), are neurotoxic (Yankner et al. 1990) and induce neuronal degeneration and apoptosis (Forloni et al. 1993; LaFerla et al. 1995). In this study, we examined whether Aβs affect Cl-ATPase activity in cultured rat hippocampal neurons, paying special attention to PI metabolism.

Materials and methods

Materials

The sources of the materials used in this study were as follows: Dulbecco's modified Eagle's medium and modified Eagle's medium from Nissui, Tokyo, Japan; adenine-9β-arabinoside (Ara-A), Aβ1–40, 1–42 and 25–35, phenylmethylsulfonyl fluoride (PMSF), ethacrynic acid (EA), ATP, ouabain, LY294002, H-7 and quercetin from Sigma, St. Louis, MO, USA; aprotinin from Bayer, Leverkusen, Germany; KT5823, KT5720 and PD98059 from Calbiochem-Novabiochem, La Jolla, CA, USA; KN62 from Seikagaku Kogyo, Tokyo, Japan; wortmannin, inositol (myo-inositol), 8-Br-cyclic GMP and poly-l-lysine hydrobromide from Nacalai Tesque, Kyoto, Japan; fetal calf serum from Hyclone, Logan, Utah, USA; horse serum from Osakariken, Osaka, Japan; l-glutamine from ICN Biomedical, Aurora, Ohio, USA; 10-mm tissue culture dish from Iwaki, Chiba, Japan.

Cell culture and drug application

Hippocampal tissues of 19-day-old Wistar rat embryos were triturated in Ca2+ and Mg2+ free Hank's solution as described previously (Inoue et al. 1991). The cells were suspended in Dulbecco's modified Eagle's medium (DMEM) containing 13.3 µm inositol supplemented with 10 mm HEPES, 4 mm l-glutamine, 100 IU/mL penicillin G sulfate, 10% fetal calf serum and 10% horse serum. The cells were seeded in poly-l-lysine-coated plastic dishes at a density of 2.55 × 105 cells/cm2. After incubation for 2 days, the cells were exposed to 5 µm Ara-A in modified Eagle's medium (MEM) containing 47.9 µm inositol supplemented with 10 mm HEPES, 2 mm l-glutamine, 100 IU/mL penicillin G sulfate and 5% horse serum for 4 days to prevent the proliferation of non-neuronal cells. The Aβs and other reagents were usually applied for 2 days from the 8th day of culture. For monitoring time-dependent effects of Aβ, the peptide was applied at the indicated times before the 10th day of culture. For monitoring DNA fragmentation and cell viability, cells were exposed to glutamate (10 µm, 10 min) on the 10th day of culture with or without Aβ-treatment and assayed after another 2-day-culture in the usual media with or without Aβ.

Preparation of membrane fractions

The membrane fractions were prepared on the 10th day of culture. All procedures were performed on ice or at 4°C. Cultured cells were washed with ice-cold buffer solution containing 0.25 m sucrose, 1 mm EDTA-Tris (pH 7.4), 12.5 mm Tris-2-(N-morpholino)-ethanesulfonic acid (MES) (pH 7.4), 1 mm PMSF and 50 units/mL aprotinin. The harvested cells were homogenized in the same buffer and then centrifuged (10 000 g, 15 min; 100 000 g, 20 min). The pellets were suspended in 2 mm EDTA-Tris (pH 7.4), stirred for 15 min, and centrifuged (100 000 g, 20 min). The resulting pellets were resuspended in 2 mm EDTA-Tris (pH 7.4) and used as plasma membrane-rich fractions. The protein concentration was determined by the method of Lowry et al. (Lowry et al. 1951).

Enzyme assays

ATPase activities were determined by spectrophotometric measurement of inorganic phosphate liberated as described previously (Shiroya et al. 1989). The assay was carried out for 15 min at 37°C in 200 µL reaction buffer containing 100 mm Tris-MES (pH 7.4), 1 mm EDTA-Tris, 100 mm NaCl, 10 mm KCl, 6 mm magnesium acetate, 6 mm ATP-Tris (pH 7.4), 2 mm NaN3 and 14–20 µg membrane protein with or without 1 mm ouabain and/or 0.3 mm EA. The reaction was terminated by the addition of 10% TCA. The Na+/K+-ATPase activity was calculated by subtracting the ATPase activity in the presence of 1 mm ouabain from the total ATPase activity. The activity in the presence of 1 mm ouabain was regarded as Mg2+-ATPase activity. The Mg2+-ATPase activity in the presence of 0.3 mm EA was designated as the anion-insensitive Mg2+-ATPase activity. The difference between the activities in the presence and absence of 0.3 mm EA was designated as the Cl-ATPase activity.

Measurement of intracellular Cl concentration

Intracellular Cl concentration ([Cl]i) was measured as described previously (Hara et al. 1992). Dissociated hippocampal cells were cultured on poly-l-lysine-coated coverslips in plastic dishes, and treated following the same protocol as for membrane Cl-ATPase activity. The cells were washed with modified Krebs–HEPES buffer solution (pH 7.3) containing 128 mm NaCl, 2.5 mm KCl, 2.7 mm CaCl2, 1 mm MgSO4, 20 mm HEPES and 16 mm glucose, exposed to 5 mmN-(6-methoxyquinolyl)-acetoethyl ester (MQAE), a Cl-sensitive fluorescent dye, in the same buffer for 1 h at 37°C, and then washed with a dye-free buffer solution. Fluorescence intensity of a single pyramidal cell-like neuron was measured in a cell-chamber at room temperature (22°C) using an inverted fluorescence microscope system (Interdec, Osaka, Japan). Excitation and emission wavelengths were collected at 360 and 510 nm, respectively. The fluorescence intensity of MQAE in each neuron was calibrated for Cl concentration as described previously (Hara et al. 1992; Irie et al. 1998), and [Cl]i was estimated to be a value corresponding to the initial fluorescence intensity.

Analysis of [3H]inositol-labeled phosphoinositides

Cells were cultured for 2 days from the 8th day of culture in a low-inositol medium (MEM/inositol-free DMEM = 1:3) containing 5% horse serum and 2 µCi/mL [3H]inositol with or without test reagents. The plasma membrane-rich fractions prepared as described above were extracted with 4 volumes of chloroform/methanol/1 m HCl (20 : 40 : 1 v/v) followed by the addition of 1 volume of chloroform and 1 volume of 0.2 m KCl in 5 mm EDTA solution. After the mixture was vortexed and centrifuged for 1 min at 5500 g, the organic phase was collected. The remaining phase was re-extracted with 1 volume of chloroform. The combined organic phase was dried under a stream of nitrogen. Lipids were resolved by thin-layer chromatography (TLC) on Silica gel G (Analtech, Inc., Newark, DE, USA) pre-treated with 1% potassium oxalate and 2 mm EDTA in 50% methanol, using an acid solvent system composed of chloroform/methanol/4 N NH4OH/H2O (45 : 35 : 4 : 6). The TLC plates were then sprayed with ENHANCETM (NEN Life Science Products, Inc., Boston, MA, USA) to allow the visualization of phospholipids. The labeled lipids were detected by autoradiography at − 80°C using Kodak X-Omat AR film followed by densitometry using a color scanner and a public domain image processing and analysis program (NIH Image; National Institute of Mental Health, Bethesda, MD, USA).

DNA fragmentation assay

DNA fragmentation assay was performed as described before (Hockenbery et al. 1990; Behl et al. 1994). Briefly, cells were harvested and washed with ice-cold PBS by centrifugation (10 000 g, 12 min). The pellets were resuspended in 500 µL cell lysis buffer containing 0.5% Triton X-100, 5 mm Tris (pH 7.4) and 20 mm EDTA. The lysates were centrifuged (12 000 g, 15 min) and the supernatants were extracted once with equal volumes of phenol/chloroform and chloroform, and precipitated with 66% ethanol in the presence of 2 mm ammonium acetate. The DNA was dissolved in 10 mm Tris-HCl and 1 mm EDTA (pH 8.0) and then 1 µg DNA was digested with RNase A (20 µg/mL) for 1 h at 37°C and loaded on a 1.5% agarose gel. The DNA precipitate was visualized under UV light after staining with ethidium bromide.

Cell viability analysis

Two biochemical assays for cell viability were used, i.e. the 2-(2-methoxy-4-nitrophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium, monosodium salt (WST-8) reduction for mitochondrial activity (Ishiyama M. et al. 1993) using Cell Counting Kit-8 (Dojindo, Tokyo, Japan), and the release of lactate dehydrogenase (LDH) to detect damaged plasma membranes (Decker and Lohmann-Matthes 1988) using an LDH-Cytotoxic Test (Wako, Osaka, Japan).

Statistical analysis

Data (mean ± SE) were analyzed by the Student's unpaired t-test. The differences between mean values with p < 0.05 were considered significant.

Results

Effects of Aβs on Cl-ATPase activity in cultured brain neurons

When cultures of rat brain hippocampal neurons were incubated with Aβs (1–100 nm, 1 h-6 d), Cl-ATPase activity in plasma membrane fractions decreased dose- and time-dependently to the level of 60–70% of control without significant effects on Na+/K+- or anion-insensitive Mg2+-ATPase activities (Figs 1a and b). The inhibition reached a plateau after 2 days of Aβ-exposure (Fig. 1b) with relative potencies of Aβ25–35 > Aβ1–42 > Aβ1–40 (I50: < 1 nm, 3.5 nm and 38 nm, respectively) (Fig. 1a). Therefore, the effects of a cytotoxic fragment of Aβ, Aβ25–35 (10 nm, 2 day), were analyzed in the following experiments. Treatment with Aβs did not change the morphology of cultured cells (data not shown) or total amounts of membrane protein (µg protein/2 dishes, control 351 ± 12, 10 nm Aβ25–35 358 ± 24, n = 7). Aβs did not directly inhibit Cl-ATPase activity in vitro over the same concentration range as in the culture (µmol Pi/mg protein/h, control 3.5 ± 0.4, 100 nm Aβ25–35 3.7 ± 1.0, n = 3–6). Thus, Aβ inhibited neuronal Cl-ATPase activity probably via cellular mechanisms.

Figure 1.

 Effects of amyloid β proteins (Aβs) on ATPase activities in cultured rat hippocampal neurons. (a) Concentration dependency. Cells were exposed to Aβ1–40, 1–42 and 25–35 (1–100 nm) for 2 days. n = 3–30, (b) Time course. Cells were exposed to 10 nm Aβ1–42 for the indicated times, n = 5–17. *p < 0.01 vs. corresponding control without Aβs.

Mechanisms of Aβ-induced inhibition of Cl-ATPase activity

To test the involvement of PI metabolism in the Aβ-induced inhibition of Cl-ATPase activity, we examined effects of inositol or a permeable cyclic GMP analog, 8-Br-cyclic GMP, on this inhibition, the former being a major supplier of the cellular inositol pool for PI synthesis (Nahorski et al. 1991), and the latter being a stimulant of phosphorylation of PI via cyclic GMP-dependent protein kinase (PKG) (Vrolix et al. 1988; Imai et al. 1990). The Aβ-induced inhibition of Cl-ATPase activity was obviously reduced by the addition of 0.5 mm inositol or 10 µm 8-Br-cyclic GMP to culture media without significant changes in basal activities (Table 1). The inhibition by Aβ as well as control activity of Cl-ATPase was not affected by a different permeable cyclic nucleotide (10 µm 8-Br-cyclic AMP) or inhibitors of PKG (1 µm KT5823), cyclic AMP-dependent protein kinase (PKA) (1 µm KT5720), calmodulin-dependent protein kinase II (CaMKII) (10 µm KN62), calcium-dependent protein kinase (protein kinase C) (10 µm H7) and PI 3-kinase (10 µm LY294002 and 50 nm wortmannin). However, 10 µm wortmannin, which also inhibits PI 4-kinase (Nakanishi et al. 1995), reduced both basal activity (to 77% of no treatment) and Aβ-induced changes (to 40% of no treatment). This inhibition of basal Cl-ATPase activity was confirmed by the use of another PI 4-kinase inhibitor, quercetin, but not by an inhibitor (PD98059) of mitogen activated protein kinase (MAPK), which may also be inhibited by wortmannin (Sue-A-Quan et al. 1997) (µmol Pi/mg protein/h, control 2.9 ± 0.1, 20 µm quercetin 1.7 ± 0.3*, 10 µm PD98059 2.6 ± 0.3, n = 3–17, *p < 0.05 vs. control), suggesting that in situ Cl-ATPase activity is regulated by PI 4-kinase activity. As inositol and 8-Br-cyclic GMP did not change the resting Cl-ATPase activity (Table 1), the resting level of PI4P may be saturated for the optimal Cl-ATPase activity.

Table 1.  Effects of Aβ25–35 on Cl-ATPase activity in cultured rat hippocampal neurons in the presence or absence of inositol, cyclic nucleotides and protein or lipid kinase inhibitors
  Cl-ATPase activity
(µmol Pi/mg protein/h)
TreatmentµMBasal+ 10 nm Aβ25–35
  1. Primary cultured rat hippocampal neurons were exposed to 10 nm Aβ25−35 for 2 days in the presence or absence of the indicated agents. Membrane fractions were prepared as described in Materials and methods and assayed for Cl ATPase activity. ap < 0.01 vs. No treatment (–), *p < 0.05, **p < 0.01 vs. each control (Basal). n = 3–19.

3.0 ± 0.12.0 ± 0.1**
inositol5002.7 ± 0.22.8 ± 0.4
8-Br-cGMP103.5 ± 0.23.4 ± 0.2
8-Br-cAMP102.9 ± 0.12.2 ± 0.1*
KT582313.2 ± 0.32.5 ± 0.1*
KT572013.3 ± 0.11.7 ± 0.4*
KN62103.1 ± 0.21.7 ± 0.2**
H7103.4 ± 0.31.7 ± 0.1**
LY294002103.2 ± 0.12.0 ± 0.4*
wortmannin0.053.1 ± 0.32.2 ± 0.1*
102.3 ± 0.2a1.9 ± 0.1

The recovery by inositol of Cl-ATPase activity was blocked by 10 µm wortmannin and 20 µm quercetin, but not by 10 µm LY294002 or 50 nm wortmannin (Fig. 2). Similarly, the 8-Br-cyclic GMP-induced restoration of Cl-ATPase activity was blocked by the former PI-4-kinase inhibitors, but not by the latter PI 3-kinase inhibitors, and was considerably reduced by a PKG inhibitor, 1 µm KT5823 (Fig. 2). Thus, the effects of inositol and 8-Br-cyclic GMP via PKG stimulation appeared to be mediated by enhancement of PI-4-phosphorylation. These findings all support the idea that Aβ inhibits Cl-ATPase activity by reducing PI4P formation.

Figure 2.

 Recovery of Aβ25–35-induced inhibition of Cl-ATPase activity by inositol and 8-Br-cyclic GMP in the presence of several kinase inhibitors. Cultured rat hippocampal neurons were exposed to 10 nm Aβ25–35 for 2 days from the 8th day of culture in the presence or absence of indicated reagents; 0.5 mm inositol, 10 µm 8-Br-cyclic GMP, 10 µm LY294002, 0.05 or 10 µm wortmannin, 1 µm KT5823 and 20 µm quercetin. *p < 0.05, **p < 0.01.

Effects of Aβs on neuronal [Cl]i

Assuming that Cl-ATPase outwardly transports Cl, Aβ-induced inhibition of Cl-ATPase may be associated with an increase in neuronal [Cl]i. As shown in Fig. 3, Aβs increased the [Cl]i of pyramidal cell-like neurons with relative potencies of Aβ25–35 > Aβ1–42 > Aβ1–40 as observed in the inhibition of Cl-ATPase activity. Aβ25–35 showed the most potent increase in [Cl]i up to 26.2 ± 2.5 mm (mean ± SE, n = 15) as compared with control levels (10.6 ± 1.1 mm, n = 8). The addition of inositol or 8-Br-cyclic GMP suppressed the Aβ-induced accumulation of Cl in parallel with the recovery of Aβ-induced inhibition of Cl-ATPase activity. Thus, Aβ increased [Cl]i probably via the same mechanisms as those speculated in the inhibition of Cl-ATPase activity.

Figure 3.

 Effects of Aβs on intracellular chloride concentration ([Cl]i). Cultured rat hippocampal neurons were exposed to 10 nm Aβs1–40, 1–42 or 25–35 for 2 days from the 8th day of culture in the presence or absence of 0.5 mm inositol or 10 µm 8-Br-cyclic GMP. [Cl]i of pyramidal cell-like neurons was assayed fluorometrically using a Cl sensitive fluorescent dye (MQAE). **p < 0.01, n = 4–12.

Effects of Aβ on phosphoinositide levels

Phosphoinositide levels in the plasma membrane-rich fractions were analyzed by the [3H]inositol labeling method. In the control culture, the ratios of [3H]inositol incoporated into PI, PIP and PIP2 fractions were 69.4 ± 0.9%, 11.6 ± 1.1% and 19.0 ± 0.5% (n = 8–9), respectively, as observed in PC12 cells (Horwitz and Perlman 1987). When the cells were treated with 10 nm Aβ25–35 for 2 days, PI, PIP and PIP2 levels were significantly lowered compared with each control (Table 2). The addition of 10 µm 8-Br-cyclic GMP supressed such effects of Aβ25–35 without any effects of the nucleotide alone. Since whole cell levels of PI and/or PIP2 did not show significant changes (for example, %PI + PIP spot density/mg cell protein, control 100 ± 1.4, 10 nm Aβ25–35 98.4 ± 1.2, 10 nm Aβ25–35 + 10 µm 8-Br-cyclic GMP 99.5 ± 0.8, n = 6–7), Aβ may mainly reduce the plasma membrane phosphoinositide levels. Thus the changes in phosphorylated PI levels in the membrane fractions appeared to parallel with those in Cl-ATPase activity in the Aβ25–35- and/or 8-Br-cyclic GMP-treated cells.

Table 2.  Effects of Aβ25–35 on [3H]inositol-labeled phosphatidylinositol (PI) and phosphatidylinositol phosphates (PIP and PIP2) in the plasma membrane-rich fractions of cultured rat hippocampal neurons
 PIPIPPIP2
  1. Primary cultured rat hippocampal neurons were exposed to 10 nm Aβ25−35 and/or 10 µm 8-Br cyclic GMP for 2 days. Plasma membrane-rich fractions were prepared as described in the ‘Materials and methods’ section and assayed for PI, PIP and PIP2. The mean values of spot density/mg membrane protein of PI, PIP and PIP2 in control cultures were taken as 100%. *p < 0.05, **p < 0.01 vs. each control. n = 8–9.

 % of Control
Control100.0 ± 1.2100.0 ± 2.3100.0 ± 2.0
Aβ25–3593.1 ± 2.3*88.6 ± 3.9**82.9 ± 3.6**
Aβ25–35+ 8-Br-cGMP102.5 ± 2.4100.4 ± 3.695.1 ± 5.9
8-Br-cGMP104.7 ± 2.7104.8 ± 5.297.2 ± 4.9

DNA fragmentation and cell survival

To test whether Aβ-induced accumulation of Cl is related to apoptotic or cytotoxic stimuli, DNA fragmentation and cell survival were monitored 2 days after glutamate (10 µm, 10 min) exposure with or without Aβ-treatment (Aβ25–35, 10 nm). As shown in Fig. 4, Aβ or glutamate alone did not affect the basal conditions (lanes 2 and 3). However, when glutamate was applied to the cells after treatment with Aβ, it significantly accelerated the fragmentation of DNA (lane 4). Inositol and 8-Br-cyclic GMP co-applied with Aβ prevented the glutamate-induced DNA fragmentation despite the presence of Aβ.

Figure 4.

 Effects of Aβ25–35 and glutamate on DNA fragmentation. Cultured rat hippocampal neurons were exposed to 10 nm Aβ25–35 for 2 days from the 8th day of culture in the presence or absence of 0.5 mm inositol or 10 µm 8-Br-cyclic GMP, stimulated with 10 µm glutamate for 10 min, incubated for further 2 days in media without additional glutamate, and then DNA was extracted. Lane 1, no treatment; lane 2, Aβ25–35 treatment alone; lane 3, stimulation by glutamate (10 µm, 10 min) alone; lane 4, Aβ25–35 treatment + glutamate stimulation; lane 5, treatment with (Aβ25–35 + inositol) + glutamate stimulation; lane 6, treatment with (Aβ25–35 + 8-Br-cyclic GMP) + glutamate stimulation.

Cell viability was estimated by two quantitative methods, WST-8 cell reactivity and LDH release assays. As shown in Fig. 5(a), Aβ (Aβ25–35, 10 nm), but not glutamate (10 µm) alone, slightly reduced the cell reactivity. However, when applied to the Aβ-treated cells, glutamate significantly decreased the cell reactivity. Similarly, Aβ or glutamate alone did not affect the basal LDH release, whereas glutamate applied after the Aβ-treatment markedly increased the LDH release (Fig. 5b). Co-administration of inositol and 8-Br-cyclic GMP with Aβ again prevented the glutamate-induced changes in Aβ-treated cells.

Figure 5.

 Viability of Aβ25–35- and/or glutamate-treated cells. Cultured rat hippocampal neurons were treated as described in Fig. 4. (a) WST-8 reducing activity. (b) LDH release. **p < 0.01, n = 7–12.

Discussion

In this study, Aβs were demonstrated to inhibit neuronal Cl-ATPase activity and to increase the neuronal [Cl]i. The effective concentrations were rather low (1–10 nm) compared with those used in many reports concerning neurotoxic effects of Aβs (> 10 µm). As the concentration of Aβs in human cerebrospinal fluid is reportedly 2.5 ng/mL (625 pm, Seubert et al. 1992) to 5 nm (Nakamura et al. 1994), the concentration in early onset AD groups being higher than that in older normal groups, effects of nm concentrations of Aβs are of physiological and/or pathophysiological importance.

Mechanisms of Aβ-induced inhibition of Cl-ATPase were analyzed using several inhibitors of protein and lipid kinases as well as their effects on the basal Cl-ATPase activity (Table 1). Although Aβs are known to stimulate several protein and lipid kinases including PI-related enzymes such as PI 3-kinase (Luo et al. 1997) and protein kinase C (Nakai et al. 1999), inhibitors of these enzymes, 10 µm H-7, 10 µm LY294002 and 50 nm wortmannin, did not affect the inhibition of Cl-ATPase activity, suggesting that the involvement of these kinases in the inhibition is unlikely. As the Aβ-induced inhibition of Cl-ATPase activity as well as its basal activity were reduced by PI 4-kinase inhibition with 10 µm wortmannin and 20 µm guercetin, and reversed by stimulation of PI or PI4P synthesis with co-application of 0.5 mm inositol or 10 µm 8-Br-cyclic GMP (Table 1, text, Fig. 2), Aβ appeared to reduce PI4P production resulting in a decrease in Cl-ATPase activity. Supporting this idea, the effects of both inositol and 8-Br-cyclic GMP disappeared with further inhibition of PI4P synthesis by 10 µm wortmannin or 20 µm quercetin (Fig. 2).

Nanomolar concentration of Aβ was first demonstrated to reduce phosphoinositide levels in the plasma membrane-rich fractions of cultured neurons without significant effects of their whole cell levels. Phosphatidylinositol and its phosphorylated metabolites are produced in various cell compartments including plasma membranes, Golgi apparatus, endoplasmic membranes and nuclear membranes (Carpenter and Cantley 1990; Geharmann and Heilmeyer 1998), and phosphatidylinositol transfer proteins, PITPs, carry PI to other membrane compartments and enzymes such as PI 4-kinases and phospholipase C gamma (Kearns et al. 1998). Because Aβ1–40, but not Aβ25–35, reportedly inhibits in vitro PI 4-kinase activity in crude rat brain membrane preparations with a Ki of 27 µm (Wallace 1994), Aβ at nm concentrations may selectively affect the phosphoinositide metabolism in plasma membranes.

The Aβ-induced increase in [Cl]i was probably mediated by the inhibition of Cl-ATPase, as the changes in the ATPase activity and [Cl]i were correlated (Fig. 3). This increase in neuronal [Cl]i may attenuate hyperpolarizing responses to transmitter-operated Cl channel opening (Kaila 1994) and may modulate G-protein-coupled signaling (Higashijima et al. 1987; Lenz et al. 1997). A recent electrophysiological study demonstrated that Aβs reduce GABA-induced hyperpolarization in Aplysia neurons (Sawada and Ichinose 1996). Such attenuation of the main mechanism of inhibitory regulation in the CNS may elicit enhancement of neuronal excitability.

Excitotoxicity of glutamate has been suggested to be involved in the neurodegeneration in AD (Koh et al. 1990; Mattson et al. 1992). A relatively low concentration (10 µm) of glutamate accelerated DNA fragmentation and decreased the survival rates of neurons in Aβ-treated cells without significant changes in non-treated cells (Figs 4 and 5). As this effect of Aβ-treatment also paralleled the changes in [Cl]i, an elevated [Cl]i may be responsible for the enhanced excitotoxicity.

Neuronal Cl-ATPase activity was thus found to be reduced by a neurotoxic factor in AD, Aβ, probably via inhibition of PI metabolism, especially PIP production, in plasma membranes. The increases in neuronal [Cl]i and glutamate neurotoxicity may further suggest the importance of this ATPase in the regulation of neuronal excitability, or protection against neurodegeneration caused by excitotoxicity.

Acknowledgements

This work was supported by grants from the Japanese Ministry of Education, Science, Sports and Culture; Japanese Private School Promotion Foundation; and the Salt Science Research Foundation, Japan.

Ancillary