Apolipoprotein E4 Elicits Lysosomal Cathepsin D Release,Decreased Thioredoxin-1 Levels,and Apoptosis

Mice brain tissue

Brain tissue material was obtained from female ApoE TR mice, as previously described [16]. The mice purchased from Taconic Farms (USA) were on a C57BL/6J background and expressed human ApoE3 or ApoE4 under the control of the endogenous murine ApoE regulatory sequences. Mice were housed in groups of 6 under controlled conditions (12 h light and dark lighting cycles, 65% humidity and a temperature of 22±2°C) and fed a standard diet. The animals had free access to food and water. All animal work was performed at the Department of Pharmacy and Biotechnologies, University of Bologna, Italy. All procedures were approved by the local ethical committee (Veterinary Service of the University of Bologna) and in accordance with the European Community Council Directive of November 24, 1986 (86/609/EEC). Mice were sacrificed by cervical dislocation and brains were dissected. The hippocampi were removed, snap-frozen in liquid nitrogen and stored at –80°C before further analysis. For protein analysis, the tissue was homogenized in ice-cold lysis buffer containing: 50 mM Tris-HCl pH 7.5/150 mM NaCl/1% Triton X-100/2 mM EDTA/2 mM EGTA/protease inhibitor cocktail (1:100) and phosphatase inhibitor cocktail (1:100) (Sigma-Aldrich, Saint Louis, MO, USA). Homogenates were kept on ice for 30–60 min followed by centrifugation, 13,600×g for 10 min at 4°C. Protein concentration was determined using BCA protein assay kit (Thermo Scientific, Rockford, IL, USA).

Human primary cultures

Embryonic human cortical tissue (6–11 weeks post-conception) was dissected after elective routine first trimester abortions from healthy pregnant women without any known AD pathology. Written informed consent was obtained from each pregnant woman. The collection was approved by the Regional Ethical Review Board (Stockholm), in compliance with Swedish law. Neuronal cultures were grown in Neurobasal medium with supplement of L-glutamine (0.5 mM) and B27 (2%). For ICC analysis, cells were seeded separately on cover slips (20,000 cells/cm²), pre-coated with poly-D-lysine (Sigma-Aldrich) in phosphate buffered saline (PBS). For western blot analysis, approximately 100,000 cells were seeded in 6-well plates. Cells were cultured for 10 days before treatment.

Cell line culture, treatment, and lysis

Neuroblastoma cells (SH-SY5Y) were purchased from the American Type Culture Collection (ATCC) (Borås, Sweden) and cultured in MEM supplemented with Glutamax and 10% fetal bovine serum (Life Technologies, Paisley, UK). Cells were incubated at 37°C in the presence of 5% CO₂. When treated, the cells (SH-SY5Y or primary neurons) were washed once in PBS (Life Technologies, Paisley, UK) buffer before addition of 100 nM rApoE (ReliaTech, Wolfenbüttel, Germany) diluted in medium (MEM or Neurobasal) without serum. When using Pepstatin A (Sigma-Aldrich), the final concentration was 10 μg/ml and was added to the cells together with rApoE. Untreated cells were used as control. For collection, the cells were first washed in cold PBS, scraped down with a cell scraper, and centrifuged for 5 min at 5000×g. The pellet was then lysed in homogenization buffer (see above). Lysate was kept on ice for 20 min before centrifugation, 14,000×g for 15 min at 4°C, and the supernatants were stored for further analysis. Protein concentration was determined using BCA protein assay kit.

Preparation of cell media for immunoblotting

SH-SY5Y cells were grown in 6-well plates and treated with rApoE as described above. The medium was collected and centrifuged at 5000×g for 5 min to pellet floating cells and debris. Next, 300 μl of the cell medium were concentrated in a speed-vac centrifuge down to a volume of approximately 35 μl. The samples were then mixed with loading buffer and analyzed with immunoblotting as described below. Due to the low levels of Trx1 in the cell medium, the sensitivity of the CCD camera was increased to detect signal.

Immunoblotting

For mouse brain homogenates and cell lysates, equal protein amount for each sample (30 μg) was mixed with loading buffer containing: 250 mM Tris-HCl pH 6.8/10 % SDS/25 % glycerol/0.04% bromophenol blue (Sigma-Aldrich). Samples were then separated with SDS-PAGE and transferred to a polyvinylidene fluoride (PVDF) membrane (Millipore, Billerica, MA, USA). Next, the membrane was blocked in 5% skim milk in TBS-Tween (TBS-T) (1 h) before incubation with primary antibody at 4°C overnight. The antibodies used were: Trx1-06 (1:1000, IMCO, Solna Sweden), Trx1-03 (1:1000, IMCO), Actin (1:1000, Sigma-Aldrich), GAPDH (1D4) (1:1000, Enzo Life Sciences, Farmingdale, NY, USA), Daxx (1:1000, MBL Int Corp, Woburn, MA, USA), Cathepsin D (1:1000, Millipore, Temecula, CA, USA), Lamin A (1:1000, Cell Signaling, Danvers, MA, USA), and Caspase 3 (1:1000, Cell Signaling, Danvers, MA, USA). The day after, membranes were washed 3×10 min in TBS-T followed by 1-h incubation with horseradish peroxidase conjugated secondary antibodies (GE Healthcare, Little Chalfont, UK) diluted 1:1000 in TBS-T. Next, membranes were washed 3×10 min in TBS-T and incubated for 3 min with ECL western blotting substrate (Thermo Scientific). The chemiluminescence signal was detected and quantified using a CCD camera (LAS-3000, Fujifilm).

Real-time PCR (rtPCR)

To extract mRNA, cells were collected using a cell scraper and centrifuged at 5,000×g for 5 min. The pellet was then lysed and RNA extracted using RNeasy mini kit (QIAGEN, Hilden, Germany). For tissue, the mRNA was extracted using RNeasy Lipid Tissue Mini Kit (QIAGEN). Both extractions were performed according to the manufacturer’s protocol. Approximately 20 mg of each sample were used. To obtain the cDNA, High Capacity cDNA Reverse Transcription kit (Life Technologies, Foster City, CA, USA) was used. Real-time PCR analysis of target genes was performed as previously described [17] using TaqMan^® Gene Expression Assays (Life Technologies, Warrington, UK). The assays for analyzing Trx1 were: Hs01555212_g1 (Trx1 human) and Mm00726847_s1 (Trx1 Mouse). GAPDH was used as the endogenous control for mice samples and RPLP0 was used as control for cell samples. The results were analyzed using the comparative 2^–ΔΔC method [18].

Immunocytochemistry (ICC)

Cells were cultured and treated in 24-well plates on glass coverslips. For subcellular localization of Daxx, cells were washed in cold PBS and then fixed in ice cold methanol (Sigma-Aldrich) for 15 min. Blocking was done in 10% BSA in TBS-Tween (TBS-T) for 30 min before incubation with primary antibody (Daxx, MBL Int Corp, 1:100, 3% BSA in TBS-T) for 1 h at room temperature. For analysis of Lamp-2 and Cathepsin D, cells were washed in cold PBS before fixation and permeabilization in ice cold methanol for 5 min. Cells were then blocked in 1% BSA in TBS-T for 30 min and incubated overnight with primary antibody (Lamp-2, 1:100 Southern Biotech, Birmingham, AL USA; Cathepsin D, 1:200, Millipore, Temecula, CA, USA) in blocking buffer. As a negative control, the primary antibody was omitted. After primary antibody incubation, all staining procedures followed the same protocol. Cells were washed three times for 10 min in PBS and incubated with secondary antibodies (Alexa Fluor (488 and/or 546) IgG 1:1000, Life Technologies, Eugene, OR, USA) and DAPI (1:1000, Sigma-Aldrich) for 30 min at room temperature. After secondary antibody application, cells were washed again 3 times in PBS before the mounting medium application (ProLong Gold Antifade Reagent, Eugene, OR, USA). Imaging was performed with confocal microscopy using Zeiss (LSM 510 META) confocal laser scanning system.

Nuclear fractionation

The nuclear and cytoplasmic extraction reagents kit NE-PER (Pierce, Rockford, IL, USA) was used to isolate the nuclear and cytosolic fractions of cells. The procedure was performed according to the manufacturer’s protocol. Protease inhibitor cocktail (Sigma-Aldrich) was added freshly at a dilution of 1:500 to all samples. To check the purity of the fractions, immunoblotting with antibody against the nuclear marker Lamin-A was performed.

Cell viability and apoptosis measurements

To analyze the level of apoptosis we measured mono- and oligonucleosomes using an ELISA^PLUS kit (Roche Diagnostics Scandinavia AB, Bromma, Sweden) according to manufacturer’s protocol. To determine cell viability, we performed an MTT assay by measuring the formation of formazan from added tetrazolium salt. Cells were cultured in 24-well plates and treated with 100 nM rApoE peptide at the indicated time points. Cells were then washed once in PBS before phenol-red free medium containing 0.3 mg/ml 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) (Sigma-Aldrich) and 1% Glutamax (Life Technologies) were added. After incubation at 37°C for 1.5 h, the MTT solution was removed and 400 μl DMSO (Sigma-Aldrich) added to dissolve the formazan crystals. After incubation with gentle shaking at room temperature, absorbance was measured in triplicate (100 μl each) in a 96 well plate at 540 nm using a plate reader spectrophotometer (TECAN Safire II). Results were expressed as the percentage of the values obtained for the corresponding untreated controls.

Redox state assessment

To study the redox state of Trx1, cells were treated with ApoE as mentioned above and analysed as previously described [13] using AMS gel shift assay. As a positive control, cells were treated with Aβ₄₂ (10 μM) for 24 h.

Catalase activity analysis

First, cells were cultured in 6 cm² polystyrene plates and treated with 100 nM ApoE peptide for the indicated time points. Next, cells were washed in ice-cold PBS and collected as described above. The supernatant was discarded and the pellets were re-suspended in 50 mM K₃PO₄ pH 7.0, 1 mM EDTA (Sigma-Aldrich) buffer followed by sonication. The cell lysates were then centrifuged at 14,000×g for 15 min at 4°C, and supernatants were used for analysis. The assay measures the conversion of methanol and hydrogen peroxide into formaldehyde in order to determine the activity of catalase. A chromagen (Purpald) (Sigma-Aldrich) that reacts with aldehydes was used to visualize the formation of formaldehyde. Absorbance was measured at 550 nm in a plate reader spectrophotometer (TECAN Safire II). Analysis was performed in 96-well plates using formaldehyde as standard (0–75 mM).

Determination of glutathione levels

To measure the levels of glutathione, a GSH/GSSG Radio Detection Assay kit (Abcam, Cambridge Science Park, UK) was used according to the manufacturer’s protocol. The kit uses a non-fluorescent dye that becomes strongly fluorescent when it binds to reduced glutathione (GSH). A GSSG probe was used in order to determine the GSSG levels. Fluorescence was measured in multi-well plate reader (TECAN Safire II) by exciting samples at 490 nm and detecting at an emission wavelength of 520 nm.

DNA constructs and transfections

SH-SY5Y were cultured in 24-well plates and transfected with plasmids expressing Trx1, Grx-1, or Parkin. DJ-1 overexpressing neuroblastoma cells (BE(2)-M17) were kindly provided by Dr. Cookson at the laboratory of Neurogenetics at the National Institute on Aging. Lipofectaminetrademark 2000 (Invitrogen, Carlsbad, CA, USA) was used for transfection according to manufacturer’s protocol (previously described [13, 20]).

Acridin orange (AO) staining

Cells were cultured in a μ-Slide 8 well for live imaging (LRI) and treated with ApoE for 1 h. After 30 min of incubation with ApoE, AO Staining solution (P.A.R.I.S Anticorps, Compiègne, France) was added directly to the cells to a final concentration of 1:200. Imaging was performed on the live cells with red (640 nm) and green (525 nm) channels simultaneously, using a Zeiss (LSM 510 META) confocal laser scanning system.

Subcellular fractionation

Cells were cultured in 10 cm plates, washed and lysis buffer containing 250 mM sucrose/20 mM HEPES pH 7.4/10 mM KCl/1.5 mM MgCl₂/1 mM EDTA/1 mM EGTA/1 mM DTT/Protease inhibitor cocktail 1:500 (Sigma-Aldrich) was added. Cells were collected using a cell scraper and then passed through a needle (27G) ten times. The lysates were then left on ice for 20 min and centrifuged at 600×g for 10 min at 4°C to remove nuclei and unbroken cells. The supernatant was then centrifuged again at 100,000×g for 3 h at 4°C. The pellet containing the microsomal fraction was solubilized in a denaturing buffer containing 50 mM Tris-HCl pH 8.3/5 mM EDTA/0.05 % SDS/6 M Urea and the supernatant was precipitated using trichloroacetic acid (TCA) (Sigma-Aldrich). TCA was added to a final concentration of 20 % and incubated on ice for 45 min. Next, the samples were centrifuged at 14,000 rpm for 5 min and washed in cold acetone. Centrifugation and washing was repeated once. Finally, the acetone was removed and the pellet was re-suspended in denaturing buffer.

Statistical analysis

Mann-Whitney U-test or one-way ANOVA followed by Fischer’s PLSD post hoc test were used for analysis of differences. A value of p < 0.05 was considered statistically significant.

ApoE4 compromises Trx1 levels both in vivo and in vitro

Western blot analysis of hippocampal tissue from ApoE TR mice revealed that Trx1 levels were reduced in ApoE4 mice compared to the ApoE3 group (Fig. 1A). Next we examined the effect of exogenous ApoE on Trx1 in vitro by treating neuroblastoma cells (SH-SY5Y) with a physiological concentration of human recombinant ApoE (rApoE) (100 nM). Our results show that, compared to rApoE3-treated and controls, rApoE4 caused a significant decrease of Trx1 levels after 5 h treatment in SH-SY5Y cells. However, after 1 h and 24 h treatment no significant differences were observed between groups (Supplementary Figure 1A, and 1B, respectively). The rApoE4-induced decrease of Trx1 was confirmed in 5 h treated human primary cortical neurons (Fig. 1C). In addition, the amount of Trx1 in the cell medium was decreased after 5 h treatment of SH-SY5Y cells with rApoE4, compared to control and rApoE3-treated cells (Fig. 1D).

Analysis using RT-PCR revealed a significant increase of the levels of Trx1 mRNA in the hippocampus of ApoE4 TR mice compared to the ApoE3 group (Fig. 1E). This was also observed in vitro after rApoE treatment in SH-SY5Y cells (100  nM; 24 h) (Fig. 1F)

ApoE4 treatment induces Daxx translocation and initiates apoptosis

Trx1 is involved in the regulation of apoptosis via direct inhibition of Ask-1 [15]. Activation of Ask-1 leads to translocation of the downstream protein Daxx from the nucleus to the cytoplasm via activation of c-Jun N-terminal kinase-1 (JNK-1) [21]. We previously showed that Aβ₄₂ causes oxidation of Trx1 and subsequent translocation of Daxx [13, 22]. To investigate whether the ApoE4-induced down-regulation of Trx1 levels had a similar consequence, we analyzed the effects of ApoE isoform treatments (5 and 24 h) on the nuclear export of Daxx in SH-SY5Y cells by subcellular fractionation of cell lysates in cytosolic and nuclear fractions and by immunocytochemistry (ICC). After 5 h (Fig. 2A) the levels of Daxx were found to be decreased in the nuclear fraction of rApoE4 treated cells, suggesting nuclear export of the protein. Purity of the fractions was assessed by re-blotting with the nuclear specific protein Lamin A antibody. Increased levels of Daxx staining were evident in the cytosol after 24 h rApoE4 treatment using ICC analysis (Fig. 2B). In human primary cortical neurons, we could not detect stronger staining of the Daxx protein in the cytosol after rApoE4 treatment. However, we did observe a weaker staining of Daxx in the nucleus (Fig. 2C), which suggests a Daxx cytosolic translocation in agreement with the results seen using SH-SY5Y cells. Translocated Daxx is degraded by the proteasome [23]. Thus, the discrepancy between models could be due to differences in the rate of Daxx degradation after translocation to the cytosol. Cytoplasmic Daxx activates Ask-1 in a positive feedback loop to initiate apoptosis [24]. To test whether the observed translocation of Daxx in rApoE4 treated cells also leads to increased apoptosis, we measured the formation of apoptotic mono- and oligonucleosomes by ELISA in SH-SY5Y cells exposed to rApoE (100 nM) for 24 h. As seen in Fig. 2D, apoptotic markers were significantly increased in cells treated with rApoE4 compared to rApoE3, both after 5 and 24 h treatments. In addition, we analyzed the levels of cleaved Caspase 3 to further assess activation of apoptosis (Fig. 2E), and found a significant increase in cells treated for 24 h with 100  nM rApoE4.

Overexpression of Trx1, Grx1, and DJ-1 protects cells against ApoE4 toxicity

We also analyzed the effects of ApoE4 on cell viability using a 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) assay. As seen in Fig. 3, rApoE4 treatment decreased cell viability to approximately 60 % in SH-SY5Y cells (Fig. 3A) and to 77 % in BE(2)-M17 cells (Fig. 3B), compared to controls; rApoE3 treatment in contrast had no significant effect. The difference in MTT reduction between the rApoE3- and rApoE4-treated cells was statistically significant. rApoE4 effects on cell viability were counteracted by overexpressing Trx1, and unexpectedly, the treatment caused an increase in cell viability compared to non-treated cells. The overexpression of Grx1 or DJ-1, two other enzymes with redox function that can inhibit Ask-1 mediated apoptosis [25, 26], also protected against rApoE4 toxicity. Conversely, when treating cells that overexpress Parkin, a protein with no known inhibitory effect on Ask-1 activation, we did not observe any protection.

Minor effects on antioxidant systems by ApoE4 treatment

We previously reported that Aβ₄₂ treatment of neuroblastoma cells leads to a fast and transient oxidation of Trx1 [13]. Since the presence of ApoE4 has been related to increased oxidative stress, we explored the effects of ApoE4 (100 nM) on Trx1 redox state using a 4-acetamido-40-maleimidyl-stilbene-2,20-disulfonate (AMS) shift assay as previously described [13]. As seen in Fig. 4A, 1 h and 5 h treatments with rApoE4 induced minor increases of the most oxidized Trx1 species compared to rApoE3 treated and untreated cells. In contrast, Aβ₄₂ (10 μM) exposure resulted in a massive oxidation of Trx1, as previously reported. After 24 h treatment, no observable differences were detected. Next, we examined the effect of ApoE isoforms on two other antioxidant actors: catalase and glutathione. As seen in Fig. 4B, rApoE4 induced a small but not statistically significant decrease in the levels of reduced glutathione (GSH) after 24 h treatment. No change was observed in GSSG levels (Fig. 4C) or the GSH/GSSG ratio (Fig. 4D) at this time point. In addition, no changes in catalase activity were detected after exposure to rApoE isoforms (Fig. 4E).

ApoE4 causes lysosomal leakage and translocation of Cathepsin D

Thioredoxin-1 is known to be degraded by the aspartic protease Cathepsin D [27]. Cathepsin D normally resides within lysosomes and late endosomes but can translocate to the cytoplasm under stress conditions, where it initiates apoptosis [28]. On the other hand, ApoE can be endocytosed into neurons via a Rab5 mediated pathway and further trafficked to late endosomes and lysosomes [29]. Previous studies have shown that presence of ApoE4 increases lysosomal leakage and apoptosis induced by Aβ peptides [10]. With these facts in mind, we wished to test the effects of rApoE4 on lysosomal integrity. SH-SY5Y cells were treated with rApoE (100 nM) for 5 h and then analyzed with ICC using antibodies for Cathepsin D and the lysosome-associated membrane protein-2 (Lamp-2). As seen in Fig. 5A, Lamp-2 immunoreactivity showed a vesicular pattern in control (untreated) cells, consistent with lysosomal morphology and localization. The staining of Cathepsin D showed partial co-localization with Lamp-2. This pattern was not changed by treatment with rApoE3. However, treatment with rApoE4 resulted in decreased staining of both Lamp-2 and Cathepsin D with more diffuse patterns. In addition, rApoE4 exposure resulted in reduced co-localization between Cathepsin D and Lamp-2 compared to rApoE3-treated and control cells (Fig. 5A). We further tested if the integrity of lysosomes in these cells is affected by the rApoE isoforms by incubating cells with acridin orange (AO) solution for 30 min after rApoEs treatment (100 nM). AO added to cells can accumulate in the lysosomes— where it becomes protonated and emits red light after excitation with blue light. AO remaining outside of the lysosomes, on the other hand, emits green light instead and this difference can be used as an indicator of lysosomal integrity [30]. As seen in Fig. 5B, treatment with rApoE isoforms resulted in a changed pattern of red/green staining, showing less retention in lysosomal structures (red) and more binding to double-stranded DNA (green).

Analysis of microsomal and cytosolic fractions revealed that Cathepsin D levels were increased in the cytosolic fraction of rApoE4-treated cells compared to rApoE3- and non-treated cells (Fig. 5C). In rApoE3 treated cells, the levels of Cathepsin D were instead increased in the microsomal fraction compared to rApoE4 treated and control cells. Immunoblotting with an antibody against GAPDH was used as cytosolic marker to check the purity of the fractions (Supplementary Figure 1B). The appearance of bands migrating differently between the two fractions is likely due to altered glycosylation patterns, since the targeting of Cathepsin D to lysosomes is mediated by glycosylation at certain residues (reviewed in [31]).

To assess whether the rApoE4 induced translocation of Cathepsin D was linked to the decrease in Trx1 levels, we co-treated SH-SY5Y cells for 5 h with 100  nM rApoE and 10 μg/ml Pepstatin A, a specific inhibitor of Cathepsin D. Through this experiment, we found that Pepstatin A inhibited the reduction in Trx1 levels caused by rApoE4 alone (Fig. 5D). Figure 5E shows ICC analysis of human cortical primary neurons treated with rApoE for 1 h. Cells were stained with antibodies against Cathepsin D and Lamp-2. As seen in Fig. 5E, and in agreement with the results found with SH-SY5Y cells (Fig. 5A), rApoE4-treated neurons showed reduced co-localization between Lamp-2 and Cathepsin D suggesting lower levels of Capthesin D in the lysosomes.