HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 24, 21821-21828, June 14, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/24/21821    most recent M112109200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Ji, Z.-S. Articles by Mahley, R. W. Search for Related Content PubMed PubMed Citation Articles by Ji, Z.-S. Articles by Mahley, R. W. Social Bookmarking                      What's this?

Apolipoprotein E4 Potentiates Amyloid  Peptide-induced Lysosomal Leakage and Apoptosis in Neuronal Cells^*

Zhong-Sheng Ji, R. Dennis Miranda, Yvonne M. Newhouse, Karl H. Weisgraber^§¶, Yadong Huang^¶, and Robert W. Mahley^§¶**

From the  Gladstone Institute of Neurological Disease, ^§ Cardiovascular Research Institute, and Departments of  Medicine and ^¶ Pathology, University of California, San Francisco, California 94141-9100

Received for publication, December 18, 2001, and in revised form, March 8, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

We assessed the isoform-specific effects of apolipoprotein (apo) E on the response of Neuro-2a cells to the amyloid  peptide (A1-42). As determined by the intracellular staining pattern and the release of -hexosaminidase into the cytosol, apoE4-transfected cells treated with aggregated A1-42 showed a greater tendency toward lysosomal leakage than neo- or apoE3-transfected cells. A1-42 caused significantly greater cell death and more than 2-fold greater DNA fragmentation in apoE4-secreting than in apoE3-secreting or control cells. H₂O₂ or staurosporine enhanced cell death and apoptosis in apoE4-transfected cells but not in apoE3-transfected cells. A caspase-9 inhibitor abolished the potentiation of A1-42-induced apoptosis by apoE4. Similar results were obtained with conditioned medium from cells secreting apoE3 or apoE4. Cells preincubated for 4 h with a source of apoE3 or apoE4, followed by removal of apoE from the medium and from the cell surface, still exhibited the isoform-specific response to A1-42, indicating that the potentiation of apoptosis required intracellular apoE, presumably in the endosomes or lysosomes. Studies of phospholipid (dimyristoylphosphatidylcholine) bilayer vesicles encapsulating 5-(and-6)-carboxyfluorescein dye showed that apoE4 remodeled and disrupted the phospholipid vesicles to a greater extent than apoE3 or apoE2. In response to A1-42, vesicles containing apoE4 were disrupted to a greater extent than those containing apoE3. These findings are consistent with apoE4 forming a reactive molecular intermediate that avidly binds phospholipid and may insert into the lysosomal membrane, destabilizing it and causing lysosomal leakage and apoptosis in response to A1-42.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Human apolipoprotein (apo)¹ E exists in three major isoforms, apoE2, apoE3, and apoE4, which are encoded by three apoE alleles (2, 3, and 4) (1). The 4 allele is a susceptibility gene for Alzheimer's disease (2-5) and other neurodegenerative disorders (6-9). The apoE isoforms have differential effects on neurite outgrowth in cultured neurons and on the stability of the intracellular cytoskeleton and microtubular system (10-14). Recently, we demonstrated that a bioactive form of apoE (E4 to a greater extent than E3) can escape an intracellular membrane compartment, enter the cytosol, and interact with cytoskeletal components to form neurofibrillary tangle-like neuronal inclusions (15). Our studies in transgenic mice have indicated a role for apoE4 in neuronal degeneration and behavioral activity (16, 17). In addition, apoE is critical in the deposition of amyloid  peptide (A) in transgenic mice overproducing the amyloid precursor protein (18-20).

The neurotoxicity of A1-42 appears to be an important factor in the pathogenesis of Alzheimer's disease (for review, see Refs. 21 and 22). A1-42 is a proteolytic product of the amyloid precursor protein and is a normal component of plasma and cerebrospinal fluid (23, 24). A deposition or amyloid plaque formation is found in Alzheimer's disease cases (25, 26) and in transgenic animal models of A overproduction (27, 28). In in vitro studies, A induces neuronal death, in part through apoptotic pathways (29-33). Caspase-2 (34), caspase-3 (35), caspase-6 (36), caspase-8 (37), and caspase-12 (38) have been implicated in A-induced apoptosis. The c-Jun amino-terminal kinase, which phosphorylates c-Jun, activates transcription, and leads to cell death, is also activated in neurons exposed to A (37, 39). However, the precise molecular mechanism of A-induced apoptosis remains controversial.

Aggregated A is more toxic than soluble A in vitro (40-42), but soluble A can also cause cell death (43-45). A-induced oxidation appears to contribute to apoptotic cell death by increasing the production of oxygen radicals and H₂O₂, which oxidize other molecules (e.g. lipids and proteins), resulting in damage of intracellular membranes (46-54). A-induced intracellular oxidative stress may be caused by endogenously synthesized or exogenously added A (38).

As a means of reducing toxicity, A is removed from interstitial fluids surrounding central nervous system cells by extracellular proteolysis and receptor-mediated endocytosis and degradation (55). Although several membrane proteins bind A, two receptors are primarily involved. Soluble A is taken up by the low density lipoprotein receptor-related protein (56-59), and aggregated A is internalized by scavenger receptors (60). A uptake mediated by the low density lipoprotein receptor-related protein requires a ligand that can form a complex with A and mediate its internalization (61). One such ligand is apoE, which binds to the low density lipoprotein receptor-related protein and other members of the low density lipoprotein receptor gene family and is a key protein for transport and redistribution of lipids among cells (62-64). Another is ₂-macroglobulin (65, 66). Both are important in the internalization of A (64, 67). Although lysosomal degradation of A had been considered a mechanism for reducing A toxicity, A1-42 is resistant to degradation by lysosomes (50). In vitro, the accumulation of A1-42 in lysosomes damages lysosomal membranes, resulting in leakage and ultimately in cell death (51). This finding correlates with the observation that the endosomes and lysosomes in neurons in Alzheimer's disease brains are structurally abnormal, e.g. the number of endosomes and lysosomes is increased 2-8-fold in cells in all areas of the brain that are susceptible to the neuropathologic effects of Alzheimer's disease (52).

In this study, we examined the interactions between apoE isoforms and A and assessed the effects of apoE3 and apoE4 on lysosomal stability, cell death, and apoptosis in cultured neuronal cells. We show that apoE4, in concert with A1-42, stimulates lysosomal leakage and potentiates cell death and apoptosis.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Reagents-- Z-VAD was purchased from Enzyme System Products (Livermore, CA); 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), H₂O₂, and Lucifer Yellow were purchased from Sigma. A1-40 and A1-42 were from Bachem (Torrance, CA). 1,2-Dimyristoylphosphatidylcholine (DMPC) was from Avanti (Alabaster, AL), and 5-(and-6)-carboxyfluorescein (CF) dye was from Molecular Probes (Eugene, OR). A1-42 was iodinated with IODO-GEN (Pierce), as described (68), to a specific activity of ~1300 cpm/ng A peptide. Rabbit -very low density lipoproteins (-VLDL) were prepared as described (69). Anti-apoE serum, staurosporine, caspase-3 inhibitor I (catalog no. 235423) and caspase-9 inhibitor II (catalog no. 218776) were purchased from Calbiochem (La Jolla, CA). Recombinant human apoE2, apoE3, and apoE4 were expressed in Escherichia coli and purified as described (70). The purified apoE was incubated with 2% -mercaptoethanol at room temperature for 4 h and then dialyzed against 0.1 M phosphate-buffered saline (4 liters × 4 changes, pH 7.4) for 24 h prior to use.

Cell Culture-- Neuro-2a cells were maintained in NB medium (50% Dulbecco's modified Eagle's medium and 50% F-12 medium) containing 10% fetal bovine serum. Neuro-2a cells were transfected with apoE3 or apoE4 genomic DNA by the LipofectAMINE method as described (71). Stably transfected cells were selected in 10% NB medium containing 400 µg/ml G418. The amount of apoE secreted into the culture medium by transfected cells was measured by immunoblot (71). ApoE3- and apoE4-transfected cells secreting 40 or 80 ng of apoE/ml of medium/24 h were chosen for the studies; cells secreting 80 ng of apoE/ml of medium/24 h were used unless otherwise noted.

A1-42 or A1-40 (1 mg) was dissolved in 100 µl of dimethyl sulfoxide and diluted in water to 1 ml. A was incubated at 37 °C for 72 h to form aggregates before use (46, 72, 73).

Determination of Cell Survival-- Cell survival was estimated with an MTT colorimetric assay (74).

DNA Fragmentation Assay-- DNA fragmentation of apoptotic cells was determined with Cell Death Detection enzyme-linked immunosorbent assay kits (Roche Molecular Biochemicals).

Measurement of Lysosomal Membrane Stability-- Cells were treated with A1-42 or apoE as described, and membrane stability and leakage of lysosomes were measured in the cytosol by Lucifer Yellow release and -hexosaminidase activity (75, 76). Briefly, the cells were incubated with Lucifer Yellow (100 µg/ml) overnight, washed three times, and incubated with A1-42 or A1-40 (20 µM) at 37 °C for 24 h. After incubation, the cells were observed by fluorescence (Zeiss) or confocal microscopy (MRC-1024, Bio-Rad) to determine whether Lucifer Yellow had leaked into the cytoplasm from membrane-limited lysosomes. The percentage of cells showing a diffuse fluorescence staining pattern indicative of lysosomal leakage was determined by counting ~30 cells in each of 15 fields (magnification, ×60) under a fluorescence microscope.

In some experiments, cells were grown in 10% NB medium to ~90% confluence and washed twice with serum-free medium. The cells were incubated with or without A1-42 (20 µM) at 37 °C for 24 h, scraped from the plates, and homogenized with a Kontes homogenizer pestle B (eight strokes). The cytosolic fraction was obtained by ultracentrifugation (51), and the cytosolic -hexosaminidase activity was measured as described (75, 76).

Cell Association and Degradation of ¹²⁵I-A1-42 by Transfected Neuro-2a Cells-- Neo-, apoE3-, and apoE4-transfected Neuro-2a cells were grown to ~90% confluence, washed once with N2 medium (NB medium plus supplement), and incubated with ¹²⁵I-A1-42 (20 µM) at 37 °C for 6, 12, or 24 h. The cells were placed on ice, and the culture medium was collected for degradation studies. The cells were washed five times with 0.2% bovine serum albumin in phosphate-buffered saline and then dissolved in 0.1 N NaOH. The cell association of ¹²⁵I-A was measured with a  counter. The degradation of A1-42 was determined as described (77).

Preparation of DMPC Vesicles Containing Fluorescent Dye-- DMPC was dissolved in benzene, lyophilized, and resuspended in 20 mM Tris-HCl containing 0.15 M NaCl and 1 mM EDTA, pH 7.4. DMPC was sonicated, and CF dye was encapsulated by the DMPC as described (78). Briefly, the lyophilized DMPC and CF were mixed to final concentrations of 10 mg/ml and 100 mM, respectively. Small vesicles were prepared by sonicating the DMPC solution at 24 °C for 20 min. The DMPC and the DMPC vesicles containing CF dye were purified by gel filtration on a Superdex 200 column (Amersham Biosciences). The phospholipid concentration of DMPC was measured with an enzymatic colorimetric method (Wako, Richmond, VA).

Release of CF Dye from DMPC Vesicles-- The release of CF dye from the DMPC vesicles was measured with a Hitachi F-2000 spectrofluorometer at excitation and emission settings of 480 and 518 nm, respectively. Fluorescence was recorded after 10 µl of the protein solution (at various concentrations as indicated) was added to a jacketed cuvette containing 500 µl of DMPC solution (60 µM final concentration). All solutions and protein samples were kept at 23.9 °C, the transition temperature for DMPC. Release of fluorescent dye by apoE or A was quantitated and expressed as a percentage of the total amount of dye released by 2% Triton (20 µl), which represented the control (100%), as described (79).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

ApoE4 Potentiates A-induced Lysosomal Leakage-- To determine whether apoE has isoform-specific effects on A-induced lysosomal leakage, neo-, apoE3-, and apoE4-transfected Neuro-2a cells were incubated first with Lucifer Yellow, a marker of fluid-phase pinocytosis that accumulates in secondary lysosomes (75), and then with 20 µM A1-42 (or A1-40) for 20 h at 37 °C, and examined by confocal microscopy. As shown in Fig. 1, untreated control cells displayed a punctate pattern of fluorescence, revealing small, circumscribed vesicular structures resembling intact lysosomes. No significant lysosomal leakage was observed in untreated apoE3- or apoE4-transfected cells. After treatment with A1-42, however, it was readily apparent that more apoE4-transfected Neuro-2a cells than neo- or apoE3-transfected cells displayed a diffuse intracellular pattern of fluorescence, indicating lysosomal leakage into the cytosol (Fig. 1, arrows).

View larger version (28K): [in this window] [in a new window]   Fig. 1.   A1-42-induced release of Lucifer Yellow from lysosomes of Neuro-2a cells. Neo-, apoE3-, and apoE4-transfected cells that were not treated with A1-42 displayed a discrete punctate pattern of fluorescence (top panels). A treatment of the apoE-transfected cells caused lysosomal leakage as shown by fluorescence that diffused into the cytoplasm; apoE4-transfected cells showed the most marked lysosomal leakage (bottom panels, arrows). The cells were labeled with Lucifer Yellow (100 µg/ml) at 37 °C for 20 h and washed. The cells were then incubated with 20 µM A1-42 at 37 °C for 20 h and visualized by confocal microscopy.

To confirm this finding, we counted the number of cells in multiple fields that showed the diffuse pattern of fluorescence after treatment with A1-42. Lysosomal leakage was seen in only 15% of the neo- and apoE3-transfected cells but in more than 25% of the apoE4-transfected cells (Fig. 2A). ApoE4 also potentiated lysosomal leakage induced by A1-40 (Fig. 2A), albeit to a lesser extent than that induced by A1-42.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   A-induced lysosomal leakage in apoE3- and apoE4-transfected cells. A, quantitation of neo- and apoE-transfected cells showing a diffuse cytosolic staining pattern with Lucifer Yellow, indicating lysosomal leakage. Control and apoE-transfected Neuro-2a cells were incubated with Lucifer Yellow and treated with A1-42 or A1-40 as described in Fig. 1, and the cells were observed by fluorescence microscopy. The percentage of cells with a diffuse pattern of Lucifer Yellow staining was calculated by observing 15 different fields of cells (60× objective). A positive cell was one in which a diffuse pattern of fluorescence was observed. Values are the mean ± S.D. of two separate experiments. *, apoE4-transfected cells treated with A1-42 versus neo- and apoE3-transfected cells treated with A1-42 (p < 0.05). B, quantitation of the lysosomal enzyme -hexosaminidase activity in the cytosol, indicating lysosomal leakage. The neo-, apoE3-, and apoE4-transfected cells were grown in 100-mm dishes to ~90% confluence and incubated with 20 µM A1-42 for 24 h. After incubation the cells were washed and cytosolic fractions were isolated as described under "Materials and Methods." The enzymatic activity of -hexosaminidase was assayed in 20 µg of cytosolic protein for each sample. Values are the mean ± S.D. of two separate experiments performed in duplicate. *, apoE4-transfected cells treated with A1-42 versus neo- and apoE3-transfected cells treated with A1-42 (p < 0.001).

The effects of apoE3 and apoE4 on A1-42-induced lysosomal leakage were also assayed by measuring the lysosomal enzyme -hexosaminidase in the cytosol (Fig. 2B). A treatment increased cytosolic -hexosaminidase activity to a significantly greater extent in apoE4-transfected cells than in neo- and apoE3-transfected cells (~85% versus ~40 and ~30%, respectively; p < 0.001). The differences observed for the neo- versus the apoE3-transfected cells treated with A were not statistically significant.

ApoE4 Potentiates A1-42-induced Cell Death and Apoptosis-- Transfected Neuro-2a cells were incubated with 20 µM A1-42 for 18 h at 37 °C, and cell survival was measured with an MTT assay (Fig. 3). Cell survival (as a percentage of control) was significantly lower in apoE4-transfected cells than in neo- or apoE3-transfected cells (48% versus 62 and 70%, respectively; p < 0.05). In all three cell lines, cell death was reduced by pretreatment with Z-VAD, an inhibitor of various caspases (80) (Fig. 3), indicating that apoptosis caused a significant portion of the A-induced cell death.

View larger version (34K): [in this window] [in a new window]   Fig. 3.   ApoE4 promotes A1-42-induced cytotoxicity in Neuro-2a cells. Control and apoE-transfected cells were grown in 96-well plates, washed with fresh medium, and then incubated for 2 h with or without Z-VAD (100 µg/ml). A1-42 (20 µM) was added to the medium for an additional 18 h. Cell survival was measured with an MTT assay. Control cells were not treated with A or Z-VAD, and the percentage was set at 100%. Values are the mean ± S.D. of two separate experiments performed in quadruplicate. *, apoE4-transfected cells treated with A versus neo- or apoE3-transfected cells treated with A (p < 0.05).

To investigate more specifically the differential effects of apoE3 and apoE4 on A-induced apoptosis, we measured DNA fragmentation in neo-, apoE3-, and apoE4-transfected Neuro-2a cells 18 h after the addition of 20 µM A1-42. DNA fragmentation was increased to a much greater extent in apoE4- than in neo- and apoE3-transfected cells (~250% versus ~140 and 110%, respectively, of that in Z-VAD-treated control cells) (Fig. 4). There was only a trend toward apoE3 being protective; however, the potentiation of A1-42-induced apoptosis by apoE4 was highly significant (p < 0.001). Pretreatment with Z-VAD greatly reduced the A-induced DNA fragmentation in all three cell lines and abolished almost all of the potentiation seen in the apoE4-transfected Neuro-2a cells (Fig. 4).

View larger version (25K): [in this window] [in a new window]   Fig. 4.   ApoE4 enhances A-induced apoptotic DNA fragmentation. Neuro-2a cells were incubated first with or without Z-VAD (100 µg/ml) for 2 h and then with A1-42 for 18 h. Control cells were not treated with A1-42. Apoptotic cell death was measured with a DNA fragmentation assay. Values are the mean ± S.D. of three separate experiments. The effects of Z-VAD treatment alone were compared with results obtained in untreated control cells and showed no effect in any of the cell lines.

ApoE4 Potentiates and ApoE3 Protects against Apoptosis Induced by H₂O₂ and Staurosporine-- Next, we examined the effects of apoE3 and apoE4 secretion on the response of Neuro-2a cells to H₂O₂ (Fig. 5A). DNA fragmentation induced by H₂O₂ (100 µM) was increased in the apoE4-transfected cells and reduced in the apoE3-transfected cells compared with the neo-transfected cells. These results were confirmed in a Neuro-2a cell line that secreted a lower level of apoE into the medium (40 versus 80 ng of apoE/ml of culture medium/24 h). To ascertain whether apoE modulates the cellular response to other agents capable of inducing apoptosis, we treated the Neuro-2a cell lines with staurosporine (0.3 µM). Again, apoE4 potentiated and apoE3 protected against DNA fragmentation (Fig. 5B).

View larger version (38K): [in this window] [in a new window]   Fig. 5.   ApoE4 potentiates and apoE3 protects against H2O2- and staurosporine-induced DNA fragmentation in transfected Neuro-2a cells. Neo-, apoE3-, and apoE4-transfected cells were grown in 96-well plates (25,000 cells/well) in 10% NB medium for 24 h. A, cells were incubated with or without H2O2 (100 µM) at 37 °C for 18 h. ApoE3- and apoE4-transfected cells secreting about 40 ng (hatched bars) or 80 ng (black bars) of apoE/ml of medium/24 h were used. After incubation, DNA fragmentation of apoptotic cells was assayed. Values are the mean ± S.D. of two separate experiments (each with six separate wells for each condition). B, the cells were incubated with or without staurosporine (0.3 µM) at 37 °C for 18 h, and DNA fragmentation was assayed. The apoE-transfected cells secreted ~80 ng of apoE/ml/24 h. *, apoE3-transfected versus neo-transfected cells (p < 0.005); , apoE4-transfected versus neo- or apoE3-transfected cells (p < 0.05).

Caspase Inhibitors Inhibit the Potentiation of Apoptosis by ApoE4-- Inhibition of caspase-3 reduced the DNA fragmentation induced by the A1-42 in all three cell lines, including the potentiation of apoptosis induced by A in the apoE4-transfected cells (Fig. 6A). Inhibition of caspase-9 had little effect on A-induced DNA fragmentation in neo- and apoE3-transfected cells but abolished the potentiation of A-induced apoptosis in apoE4-transfected cells (Fig. 6B). Because caspase-9 activation requires the release of mitochondrial cytochrome c (81, 82), this suggests that the association of apoE4 and A1-42 treatment of the cells may enhance DNA fragmentation secondary to lysosomal leakage and oxidative stress related to mitochondrial dysfunction and activation of caspase-9.

View larger version (13K): [in this window] [in a new window]   Fig. 6.   Effects of caspase inhibitors on A1-42-induced DNA fragmentation in Neuro-2a cells. Neo-, apoE3-, and apoE4-transfected cells were incubated with or without caspase inhibitors (100 µM) at 37 °C for 2 h. A1-42 was then added to the medium, and the incubation was continued for 18 h. Apoptotic cells were measured with a DNA fragmentation assay. A, caspase-3 inhibitor. Values are the mean ± S.D. of two separate experiments performed in quadruplicate. B, caspase-9 inhibitor. Values are the mean ± S.D. of two experiments (each with six separate wells for each condition). The effects of the caspase inhibitors alone are compared with results obtained in untreated control cells. Results are presented as percentage change in DNA fragmentation.

Conditioned Medium from ApoE3- and ApoE4-secreting Neuro-2a Cells: ApoE4 Potentiates Apoptosis-- Next, we considered the possibility that apoE generated within the secretory pathway of the transfected cells might be responsible for the results seen after A1-42 treatment. ApoE3- and apoE4-secreting Neuro-2a cells were cultured for 24 h, and the conditioned media were transferred to neo-transfected cells; 20 µM A1-42 was added, and DNA fragmentation quantitated after 18 h. A-induced DNA fragmentation was significantly greater in cells incubated with apoE4-conditioned medium than in those incubated with neo- or apoE3-conditioned medium (314% versus 232 and 202%, respectively; p < 0.05). There was a trend toward less DNA fragmentation in cells incubated in apoE3-conditioned medium than in those incubated in neo-conditioned medium (p = 0.067).

Internalized ApoE4 Is Required to Potentiate A-induced Apoptosis-- Next, we assessed the effects of exposing the cells to apoE before treatment with A1-42. Neo-transfected cells were incubated with apoE3 or apoE4 (7.5 µg of protein/ml) plus -VLDL (5 µg of protein/ml) for 4 h at 37 °C, washed, and treated with suramin to remove cell surface-bound apoE and -VLDL. A1-42 was added in fresh medium without apoE, the cells were incubated for 18 h at 37 °C, and DNA fragmentation was determined. The DNA fragmentation increased 203, 160, and 283% in the cells preincubated with A1-42 alone, with apoE3 plus -VLDL plus A, and with apoE4 plus -VLDL plus A, respectively (Fig. 7). Thus, preexposure of the Neuro-2a cells to apoE3 slightly decreased DNA fragmentation compared with the cells treated with A1-42 alone, whereas apoE4 potentiated the A-induced apoptosis (p < 0.001).

View larger version (16K): [in this window] [in a new window]   Fig. 7.   Internalized apoE4 + -VLDL stimulates A1-42-induced DNA fragmentation. Neuro-2a cells were preincubated with apoE3 + -VLDL or apoE4 + -VLDL (7.5 µg + 5 µg/ml) at 37 °C for 4 h, washed, and treated with 10 µM suramin (to remove surface-bound lipoproteins and apoE). Cells were then incubated with 20 µM A1-42 for 18 h, and DNA fragmentation was determined. Control cells were not preincubated with apoE + -VLDL or treated with A. The -VLDL/E3- and -VLDL/E4-treated cells without A1-42 were not different from the control cells. Values are the mean ± S.D. of two separate studies performed in quadruplicate. *, p < 0.001 versus cells treated with A1-42 alone.

Cell Association and Degradation of ¹²⁵I-A1-42 by Neuro-2a Cells Incubated with -VLDL plus ApoE3 or ApoE4 Are Not Different-- Neuro-2a cells were incubated with -VLDL (5 µg of protein/ml) and apoE3 or apoE4 (7.5 µg of protein/ml) along with ¹²⁵I-A1-42, and cell association and degradation of the A were determined after 24 h at 37 °C. The amount of A bound, internalized, and degraded was not significantly different in the presence of apoE3 or apoE4 (Table I).

View this table: [in this window] [in a new window]   Table I Cell association and degradation of 125I-A1-42 in 24 h by Neuro-2a cells incubated with -VLDL, -VLDL + apoE3, or -VLDL + apoE4

Possible Mechanism for Potentiation of A-induced Lysosomal Leakage by ApoE4-- The release of fluorescent dye from phospholipid vesicles, which serve as artificial bilayer membranes, has been used to monitor the interaction and remodeling of the vesicles by apolipoproteins, including apoE (78, 83). When the phospholipid DMPC and CF are sonicated together, the CF becomes trapped inside the DMPC vesicles. At high concentrations of CF (~100 mM) in the DMPC vesicles, its fluorescence is self-quenched (55). DMPC vesicles can be disrupted by interaction with apolipoproteins, and a fluorescent signal is released into the solution.

The addition of apoE3 or apoE4 to the CF-containing DMPC vesicles resulted in rapid dose-dependent release of CF fluorescence as the vesicles were remodeled into disks (Fig. 8). In four separate studies, the fluorescence released by apoE4 was 135% higher than that released by apoE3 (p < 0.001), indicating its greater ability to disrupt the DMPC vesicles. In other studies, we compared the ability of apoE2, apoE3, or apoE4 to release fluorescent dye from the DMPC vesicles. ApoE2 was the least active, and apoE4 was the most active. In two separate assays, the mean fluorescent dye release obtained 30-60 s after addition of apoE2, apoE3, and apoE4 (0.002 µM) was 45, 55, and 68%, respectively (p < 0.01).

View larger version (18K): [in this window] [in a new window]   Fig. 8.   Release of CF dye from DMPC vesicles by apoE3 and apoE4. DMPC vesicles with CF dye trapped in the aqueous space were prepared as described under "Materials and Methods." Fluorescence released from the DMPC vesicles was measured with a spectrofluorometer after the addition of 10 µl of apoE at the indicated concentrations. After a 10-s baseline measurement, apoE (10 µl) was added to the DMPC, and the fluorescence released was recorded for 1 min.

When added to DMPC vesicles, A can release trapped intravesicular CF very effectively. Like apoE and other lipoproteins (84), A1-42 is amphipathic and can bind to lipids (85). When mixed with A, apoE4 had a significantly greater ability than apoE3 to release the fluorescent dye (data not shown).

Next, we pretreated CF-containing DMPC vesicles with very low concentrations of apoE3 or apoE4 that caused the release of similar, very small amounts of fluorescent dye (Fig. 9). When A1-42 was then added, there was a marked, prompt release of dye. Importantly, in the presence of A, the DMPC vesicles pretreated with apoE4 were more unstable and more dye was released than from the vesicles pretreated with apoE3. Results from three separate studies revealed that the release of fluorescence induced by A was 160% greater for the apoE4-pretreated DMPC vesicles than for those pretreated with apoE3 (p < 0.001). A1-42 (0.05 µM) added alone to the DMPC vesicles caused a release of CF very similar to that obtained with apoE3 pretreatment followed by addition of A (~30-35% of fluorescence released). These results suggest that apoE4 is acting in concert with A to destabilize and disrupt lipid membranes. Thus, the role of apoE4 in enhancing A-induced destabilization of lysosomal membranes may allow A and apoE4 (and lysosomal enzymes) to enter the cytosol and activate cell death and apoptotic pathways.

View larger version (20K): [in this window] [in a new window]   Fig. 9.   Release of CF dye from the DMPC vesicles by sequentially adding apoE3 or apoE4 and then A. ApoE3 or apoE4 (10 µl, 0.0001 µM) was added into the solution of DMPC vesicles containing CF dye 10 s after the start of the experiment. Twenty seconds after the addition of the apoE, A (20 µl, 0.05 µM) was added, and the released fluorescence was recorded for an additional 30 s.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The effects of apoE isoforms on A-induced neurotoxicity remain a critical area of investigation in which published data appear to be in conflict. In some studies, lipid-free apoE4 bound more avidly than apoE3 to A (86) and had greater ability to form A fibrils (87-89). This finding suggested that apoE4 accelerates amyloid deposition and neuropathology. In other studies, lipid-free or lipidated apoE3 bound A much more avidly than apoE4 (44, 58, 59, 90-92). This finding suggested that apoE3 protects against A-induced cell death and apoptosis by enhancing the clearance and degradation of A (58, 59) or by reducing the interaction of A with cell-surface membranes (44, 92). Other effects of apoE3 have been attributed to its ability to protect against oxidative stress, which contributes to A-induced cytotoxicity (93-96). Some of the reported differences in apoE isoform-specific effects may reflect differences in the source and physical state of apoE or in the origin and handling of the cells. For example, it is known that apoE can be unstable and that care must be exercised to maintain it as a monomeric, unaggregated protein, as described under "Materials and Methods." In addition, we use apoE-transfected cells as a primary system in which to explore initially an activity of apoE. The apoE assumes a native conformation and is lipidated as it is synthesized and secreted by the cells. Furthermore, the cells used in our studies synthesized low levels of apoE (40-80 ng/ml/24 h), which may approximate physiological conditions more closely. Alternatively, it is reasonable to speculate that some of the differences in published data may demonstrate that apoE acts through different pathways depending on the specific cell type (neurons, astrocytes, or microglia) and the specific physical state of the protein (conformation, degree of lipidation) under different physiological or pathological conditions. Thus, apoE4 may have a detrimental role under one set of conditions and apoE3 may exhibit a protective function under another set of circumstances.

This study shows that apoE4 potentiates A1-42-induced cell death and apoptosis in Neuro-2a cells and that apoE3 has little or no protective effect regardless of whether the cells were synthesizing and secreting low levels of apoE, were exposed to conditioned medium from apoE-transfected cells, or were incubated with apoE-enriched -VLDL. On the other hand, apoE3 protected the transfected Neuro-2a cells from H₂O₂- and staurosporine-induced cell death and apoptosis, whereas apoE4 increased cell death and apoptosis in response to these agents. Others have shown that apoE3 protects cells against H₂O₂ cytotoxicity (93), but not against staurosporine (59, 97). The reasons for the differences remain to be explained, as discussed above.

Previously, it was shown that A accumulates in lysosomes and is slowly degraded (50). We have now shown that apoE4 potentiates A-induced lysosomal leakage, as determined by Lucifer Yellow fluorescent staining of lysosomes and by measurement of the lysosomal enzyme -hexosaminidase in the cytosol of the Neuro-2a cells. ApoE3 plus A1-42 gave results similar to A1-42 treatment of the cells, at best revealing only a trend toward protection against lysosomal leakage. Our demonstration that A1-42 alone can cause lysosomal leakage is consistent with the ability of A to disrupt lysosomal membrane integrity (51).

The enhanced lysosomal leakage associated with A1-42 treatment of the apoE4-expressing cells may be linked directly with A-induced cell death and apoptosis. Reactive oxygen species and free radical formation have been proposed to explain at least a portion of the cytotoxicity of A (46-54). Oxidative stress induced by H₂O₂ treatment of the apoE4-secreting cells also potentiated DNA fragmentation. Apoptosis may be initiated by escape of A1-42, apoE4, or the lysosomal enzymes into the cytosol. Several lysosomal enzymes, such as cathepsin B, D, or L, and lysosomal extracts have been shown to cause apoptosis under various conditions and in several different cell types (98-103). Lysosomal proteases have been shown to activate caspases (100) and to directly cleave the Bcl-2 family member, Bid (99). The lysosomal leakage may then affect mitochondrial function and increase the release of cytochrome c, the activation of caspase-9, and the induction of apoptosis (81, 104, 105). Previously, it has been shown that A treatment of cells can release cytochrome c (104). Our data demonstrate that caspase-9 is involved and that the caspase-9 inhibitor totally blocks the A1-42-induced apoE4-potentiated increase in DNA fragmentation. There is considerable evidence to suggest that A can activate several apoptotic pathways (for review, see Refs. 106 and 107); however, we would suggest that in association with lysosomal leakage at least the mitochondrial cytochrome c-caspase-9 pathway is operative.

Importantly, potentiation of A1-42-induced apoptosis required that apoE4 be internalized by the cells. Incubation of Neuro-2a cells with conditioned medium from the apoE-transfected cells gave results very similar to those obtained with the transfected cells that synthesized and secreted apoE. Most revealing, a 4-h preincubation of Neuro-2a cells with -VLDL plus apoE4, followed by a complete removal of apoE from the medium and the cell surface by suramin treatment, still resulted in the apoE4 potentiation of DNA fragmentation. This finding strongly suggests that intracellular apoE within the endosomes or lysosomes acts in concert with the A to give the apoE4 isoform-specific results: lysosomal leakage and cytotoxicity.

We speculate that apoE4 can potentiate these effects because it forms reactive folded intermediates more readily than apoE3 and is less stable (108). Extensive data now exist demonstrating that protein molecules can be partially unfolded to form a distinct stable physical state (referred to as the molten globule) with unique properties (109, 110). This state can exist in living cells and has been linked to important physiological processes (110). These reactive intermediates bind avidly to phospholipids and membranes, alter membrane-associated processes, and translocate across membranes (110). We speculate that apoE4 forms a reactive intermediate that destabilizes the lysosomal membrane and, in concert with A1-42, causes membrane disruption and leakage. Consistent with this hypothesis, Weisgraber and associates (111) have demonstrated that apoE4 forms phospholipid disks about 2-fold more rapidly than apoE3 and have suggested that apoE undergoes several steps when it interacts with phospholipid by opening up the four-helix bundle of the protein molecule and forming a bioactive intermediate. Furthermore, at pH 4, which is similar to the pH of many lysosomes, apoE4 becomes more unstable and more reactive and binds more avidly to phospholipids than apoE3. In addition, the reactive intermediate of apoE4 may avidly interact with the phospholipid bilayer of the cell organelles or interact more avidly with A to form reactive co-fibers of apoE and A to disrupt the lysosomal membranes.

Like the apolipoproteins, A is amphipathic and can bind to lipids (43, 44, 112, 113). A can also interact with itself to form fibrils. In fact, apoE4 enhances the formation of A fibrils displaying uniform distribution of apoE along or within the fibrillar structure (87). ApoE3 has less ability to form complex A fibrils. In addition, A may form reactive intermediates or apoE-A intracellular complexes may form, interact with membranes, and alter the stability of the cell organelle.

Our studies of DMPC phospholipid vesicles support the hypothesis that apoE4 is more reactive than apoE3 in remodeling (disrupting) phospholipid bilayers. ApoE4 was more active than apoE3, which was more reactive than apoE2, in releasing the fluorescent dye from DMPC vesicles. The enhanced ability of apoE4 (apoE4 > apoE3 > apoE2) to destabilize phospholipid bilayers may contribute to its detrimental effects in the context of A1-42, whereas apoE3 and apoE2, with less ability to remodel the DMPC vesicles, may not. Likewise, pretreatment of the DMPC vesicles with apoE4 enhanced the ability of A1-42 to disrupt the bilayer and release the dye. The A1-42-induced, apoE4-potentiated instability of lysosomal membranes and lysosomal leakage are consistent with this observation. However, a more precise understanding of the mechanism requires knowledge of the physical state of apoE and A in lysosomes, how they may interact with each other in this organelle, and how they remodel and destabilize the membranes.

	ACKNOWLEDGEMENTS

We thank Sylvia Richmond and Catharine Evans for manuscript preparation, Gary Howard and Stephen Ordway for editorial assistance, John C. W. Carroll and Jack Hull for graphics, and Stephen Gonzales and Chris Goodfellow for photography.

	FOOTNOTES

^* This work was supported in part by Program Project Grant HL47660 from the National Institutes of Health and Grant 8RT-0130 from the University of California Tobacco-related Disease Research Program.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^** To whom correspondence should be addressed: Gladstone Inst. of Neurological Disease, P. O. Box 419100, San Francisco, CA 94141-9100. Tel.: 415-826-7500; Fax: 415-285-5632; E-mail: rmahley@gladstone.ucsf.edu.

Published, JBC Papers in Press, March 23, 2002, DOI 10.1074/jbc.M112109200

	ABBREVIATIONS

The abbreviations used are: apo, apolipoprotein; A, amyloid  peptide; CF, 5-(and-6)-carboxyfluorescein; DMPC, 1,2-dimyristoylphosphatidylcholine; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; VLDL, very low density lipoprotein.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES