All procedures were approved by the University Ethics Committee under protocol IBCCF 172. Hippocampal slices were produced as described previously, with slight modifications (Stoppini et al., 1991). Briefly, male Lister Hooded rats at postnatal day 6–7 were decapitated and hippocampi were dissected under sterile conditions in ice cold Hank's Balanced Salt Solution (HBSS, Gibco). Hippocampi were sliced transversally at 400 μm with a McIlwain Tissue Chopper (Mickle Laboratories). Slices were then transferred to 30-mm-diameter membrane inserts (Millicell, Millipore) and cultured for 14 days in 6-well culture trays with 1 mL of medium per well. Culture medium included 50% minimum essential medium (MEM), 25% HBSS, 25% heat inactivated horse serum, 1% penicillin, 1% streptomycin in addition to 25 mM HEPES, 36 mM D-glucose, 4 mM Na₂HCO₃, pH 7.3. Cultures were kept in a humidified incubator at 37°C and 5% CO₂. Medium was changed every 3 days.

Vector preparations were produced by the plasmid co-transfection method as shown previously (Petrs-Silva et al., 2011). Briefly, the crude iodixanol fraction with rAAV vectors was further purified and concentrated by column chromatography on a 5-ml HiTrap Q Sepharose column using an AKTA FPLC system (Amersham Biosciences, Piscataway, NJ). The vector was eluted from the column using 215 mM NaCl, pH 8.0, and the vector containing fractions were collected, pooled, concentrated, and buffer exchanged into Alcon BSS with 0.014% Tween 20, using a Biomax 100 K concentrator (Millipore, Billerica, MA). The titer of DNase-resistant vector genomes was measured by real-time PCR relative to a standard. Finally, the purity of the vector was validated by silver-stained sodium dodecyl sulfate–polyacrylamide gel electrophoresis, assayed for sterility and lack of endotoxin, and then aliquoted and stored at −80°C. Each vector contained the genome encoding green fluorescent protein (GFP) or human CHIP under the control of a ubiquitous chicken beta-actin (CBA) promoter.

Recombinant adeno-associated viral vector serotype 8 (rAAV8) was used to promote CHIP overexpression in hippocampal tissue and rAAV8GFP was used as a control. Hippocampal slices were infected with 10E9 vector genomes. A total of 1 μl virus solution at 1.36₁₂ VG/mL was applied directly to the top of each hippocampal slice, at 30 min of culture (day 0). All analysis were done 14 days after infection. Tunicamycin (TN, Sigma) an inhibitor of N-glycosylation widely used as an ER stress inducer, was diluted in DMSO and added to culture medium. Slices were incubated with the inhibitor at day 13, in various concentrations for 24 h, completing 14 days in vitro. Control slices were incubated with DMSO (0.016%) only.

Propidium iodide (PI) at concentration of 5 μg/mL was added to the organotypic hippocampal slices cultures (OHSCs) after the experimental procedures. Cell death by necrosis was identified by the uptake of PI (Macklis and Madison, 1990; Raval et al., 2003; Kosuge et al., 2008). Images were obtained with an inverted epi-illumination fluorescence microscope (Zeiss, MRMm Rev3) with a cold-CCD camera system (Axiocam) at 4x magnification. The same exposure time was used for all independent experiments. PI fluorescence was quantitatively analyzed using ImageJ Software and was expressed as a percentage of the maximum fluorescence (Ff), obtained after tissue fixation with 4% paraformaldehyde (4% PF).

Cell death (%) = (F − F0)/(Ff − F0) × 100: where F is the PI fluorescence of slices measured at 24 h of drug exposure; F0 is the background fluorescence prior to treatment.

The ApopTag® In Situ Apoptosis Detection Kit was used for TUNEL assay. Slices not treated with PI were fixed with paraformaldehyde (PF) 4% for 2 h, and then washed with PBS. The slices were then incubated in 1% Triton X-100 in PBS for 45 min. After 3 PBS washes of 5 min, they were incubated with Equilibration Buffer for 10 min at room temperature, and then with 30% of TdT Enzyme and 70% of Reaction Buffer for 2 h at 37°C. Slices were then incubated with Stop/Wash Buffer for 10 min at room temperature and washed with PBS. After incubation with Anti-Digoxigenin-Fluorescein (47%), Blocking Solution (53%) plus TO-PRO3 (1:1000) for 1 h, slices were again washed with PBS and coverslips were mounted with N-Propylgallate.

Hippocampal slices not treated with PI were fixed with PF 4% for 2 h and washed in PBS. After that, they were removed from the membrane inserts and placed in 24 well plates where they were permeabilized with 1% Triton X-100 in PBS for 2 h. Free floating slices were then incubated with 1% BSA in PBS for 2 h and incubated with primary antibodies in 1% BSA overnight at 37°C. Primary antibodies used include anti-rabbit CHOP/GADD153 (1:100; Santa Cruz) anti-mouse TUJ-1 (1:100; Sigma), anti-rabbit CHIP (1:100; Santa Cruz), anti-rabbit cleaved caspase-3 (1:100; Cell Signaling), anti-rabbit p53 (1:100, Santa Cruz). Following washes with PBS, tissues were incubated for 1 h at room temperature with Alexa Fluor 488-conjugated goat anti-rabbit, Alexa Fluor 555-conjugated goat anti-mouse antibodies (Invitrogen) diluted in PBS (1:200) plus TO-PRO3 (1:1000, Sigma) for nuclear staining. Tissues were then washed in PBS and mounted with N-propylgallate.

Slices were examined in a confocal microscope (Zeiss, LSM 510). Cells with condensed chromatin were counted in CA1 at 40x of magnification, in three distinct fields for each slice. Values represent mean percentages for slices under various treatments.

Hippocampal slices were rinsed with PBS and then homogenized on ice in RIPA lysis buffer containing 1% TritonX-100, 1% DOC, 1% NP-40, NaCl 150 mM, TrisHCl 10 mM, EDTA 5 mM, SDS 0.1%, PMSF (10 mg/mL), pepstatin (1 mg/mL), aprotinin (2 mg/mL), leupeptin (2 mg/mL), NaF (22 mg/mL) and sodium ortovanadate (92 mg/mL). Lysates were centrifuged at 12,000 g for 15 min at 4°C. Protein concentration in the supernatant was determined with by Lowry protein assay. In a 10% SDS–polyacrylamide gel, 30 μg of protein was applied per lane for electrophoresis. After that, gel was transferred to nitrocellulose membranes (Bio-Rad) and processed for western blotting. First, membrane was blocked with 5% milk in T-TBS buffer (0.1% Tween in 20 mM Tris-HCl/137 mM NaCl; pH 7.3), then overnight with primary antibodies: anti-goat BIP/GRP78 (1:500, Santa Cruz); anti-rabbit phosphorylated eIF2-α (1:1000, Bioscience anti-rabbit eIF2-α (1:1000, Santa Cruz); anti-rabbit CHIP (1:1000, Santa Cruz) or anti-rabbit ERK-2 (1:2000, Santa Cruz). Washed membranes were incubated with an HRP-conjugated secondary anti-antibody for 1 h and revealed with the ECL Western Blotting Analysis reagent (Amersham Biosciences). Optical density on the blots was measured with ImageJ Software.

Values are expressed as the mean ± S.E.M. Statistical significance was assessed with One-Way ANOVA followed by Bonferroni's multiple comparison post-test. Each experiment represents a pool of hippocampal slices obtained from four rats of the same litter. Four to six slices were used for every condition of treatment and/or infection. We analyze three independent experiments obtained from three different litters for statistics.

To confirm that treatment with tunicamycin induces endoplasmic reticulum stress and activates the UPR in hippocampal cultures, protein extracts from slices maintained in the presence of tunicamycin at either 10 or 80 μg/mL were processed for western blot. We examined the content of the ER resident chaperone BiP/GRP78, the increased expression of which is a hallmark of UPR (Jäger et al., 2012). Figures 1A,B shows that both 10 or 80 mg/mL of tunicamycin efficiently increased the levels of BiP/GRP78. We also examined CHOP/GADD153, the pro-apoptotic component of the UPR. Tunicamycin also increased the levels of CHOP/GADD153 (Figures 1D,E) when compared to control (Figure 1C). Besides the increased immunolabeling, CHOP/GADD153 co-localized with the nuclear tracer TOPRO3, consistent with the function of CHOP/GADD153 as a transcription factor activated during UPR. These results confirmed that Tunicamycin induces endoplasmic reticulum stress and activates the UPR in hippocampal slices.

To analyze cell death in hippocampus after UPR activation, hippocampus slices were maintained in vitro for 13 days and then incubated with tunicamycin at 10, 40, 80 or 120 μg/mL. After 24 h in the presence of tunicamycin we quantified necrotic cell death by the uptake of PI in non-fixed hippocampal slices. As shown by Kosuge et al. (2008), Tunicamycin treatment at 40, 80 or 120 μg/mL were able to induce cell death in hippocampal slices after 24 h of treatment, especially at the DG and CA1 region (Figures 2A,B). Tunicamycin at 80 μg/mL promoted severe ER stress and was chosen for subsequent molecular approaches (Figure 2B). We also estimated cell death through the identification of nuclei with chromatin condensation, usually equated with apoptosis. For this analysis, CA1 was chosen as a region of interest, although similar phenotypic patterns were observed in other hippocampal subregions (data not shown). Figures 2C–F shows photomicrographs of CA1 of hippocampal slices maintained either with Tunicamycin Figures 2D–F or control medium (Figure 2C) for 24 h (from day 13 to 14). Cells with condensed chromatin were counted in CA1, revealing a dose-dependent effect of Tunicamycin. A similar increase in the proportion of cell death detected either by the quantification of PI uptake or chromatin condensation was observed following higher doses of Tunicamycin (compare Figures 2G,H).

To test whether overexpression of CHIP alters cell death induced by ER stress in hippocampus slices we used adeno-associated virus serotype 8 (rAAV8) as a carrier. First we assess the transduction efficiency 14 days after infection, we used GFP as a control transgene (rAAV8-GFP) and examined the distribution of its fluorescence. Figures 3A,B shows GFP fluorescence in all hippocampus regions indicating a very efficient transduction of the tissue in our experimental condition. rAAV8-CHIP induced an increase in the content of CHIP after 14 days of infection (Figures 3C,D), which was confirmed by Western blots (Figure 3F). There was also a trend of increased CHIP content after treatment with Tunicamycin at 80 μg/mL (Figure 3E).

After 13 days post infection with rAAV8-CHIP or rAAV8-GFP, hippocampal slices were treated with Tunicamycin 80 μg/mL for 24 h. PI uptake was similar in both rAAV8-GFP infected (Figure 4C) and uninfected hipppocampal slices (Figure 4B). In contrast, rAAV8-CHIP reduced PI uptake to levels comparable with untreated slices (Figures 4A,D). These data indicate that overexpression of CHIP blocked necrosis induced by ER stress (Figure 4E).

Importantly, untreated tissue infected with rAAV8-CHIP or rAAV8-GFP did not show changes in the uptake of PI when compared to control slices, indicating that infection with the viral vector by itself does not significantly modulate cell death in this model (Figure S1).

Apoptotic cell death was also estimated after overexpression of CHIP. Slices treated with tunicamycin at 80 μg/mL had an increased percentage of cells with condensed chromatin in CA1 compared to control slices, while overexpression of CHIP reduced the proportion of apoptotic nuclei (Figure 5). rAAV8-GFP had no effect. rAAV8 infection by itself also did not induce chromatin condensation (Figure S2). We further examined immunofluorescence for cleaved caspase-3 and TUNEL as markers of apoptosis. Tunicamycin increased the number of positive cells detected by either method, whereas rAAV8-CHIP abrogated this effect (Figures 5F–M). The data altogether indicate that overexpression of CHIP in hippocampal tissue attenuates both necrosis and apoptosis induced by ER stress.

The elongation factor eIF2-α is a substrate of PERK, one of the ER stress sensors. We found that at 24 h of treatment, Tunicamycin induced a significant increase in eIF2-α phosphorylation, consistent with activation of the UPR. In turn, overexpression of CHIP, but not of GFP, abrogated eIF2-α phosphorylation as shown in Figure 6.

CHOP/GADD153 is a classical hallmark of apoptotic signalization triggered after UPR. To test whether CHIP overexpression alters the upregulation of CHOP/GADD153 induced by Tunicamycin, we immunostained slices infected with either rAAV8-GFP or CHIP, and estimated the amount of CHOP/GADD153 positive cells. In Figure 7, we show that overexpression of CHIP, but not GFP, prevented Tunicamycin-induced upregulation of CHOP/GADD153. Immunolabeling for TUJ-1 in CA1 region suggests neurodegeneration following Tunicamycin, as shown by the morphology of pyramidal cells stained with TUJ-1. Overexpression of CHIP, on the other hand, prevented such alterations. Taken together, these results suggest that the neuroprotection by CHIP is accompanied by a reduction in the activity of the pro-apoptotic UPR pathway.

Previous work has shown that p53 is also a downstream player associated with UPR activation (Li et al., 2006). We examined the content of p53 by immunofluorescence, and found that Tunicamycin increased the number of p53 positive cells in hippocampal tissue (Figure 8), whereas rAAV8-CHIP prevented the upregulation of p53.

Western blot for the chaperone BIP/GRP78 showed that Tunicamycin increased the content of BIP/GRP78 (Figure 9), an effect that was not changed by either rAAV8-GFP or rAAV8-CHIP (Figure 9). These results suggest that overexpression of CHIP did not prevent ER stress, nor the activation of the adaptative UPR in hippocampal slices following Tunicamycin treatment.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.