In vitro mitochondrial failure and oxidative stress mimic biochemical features of Alzheimer disease

Page 1

Neurochemistry International 63 (2013) 112–120

Contents lists available at SciVerse ScienceDirect

Neurochemistry International

journal homepage: www.elsevier .com/locate /nc i

In vitro mitochondrial failure and oxidative stress mimic biochemicalfeatures of Alzheimer disease

0197-0186/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.neuint.2013.05.005

⇑ Corresponding author.E-mail address: [email protected] (R. Selvatici).

Rita Selvatici a,⇑, Luca Marani b, Silvia Marino b, Anna Siniscalchi b

a Department of Medical Sciences, Sections of Medical Genetics, University of Ferrara, Via Fossato di Mortara 74, 44121 Ferrara, Italyb Department of Pharmacology, University of Ferrara, Via Fossato di Mortara 74, 44121 Ferrara, Italy

a r t i c l e i n f o a b s t r a c t

Article history:Received 21 February 2013Received in revised form 13 May 2013Accepted 16 May 2013Available online 27 May 2013

Keywords:Sodium azideHydrogen peroxideTauGSK3p35/25BACE1

Primary cortical neurons exposed to the mitochondrial toxin NaN3 (0.1–3 mM) were submitted to oxida-tive stress with H2O2 (30–150 lM), to mimic conditions observed in neurodegenerative disorders. Theeffects of such treatment on a series of parameters useful in characterizing neuronal damage were inves-tigated: (i) the basal release of glutamate, evaluated as 3H-D-Aspartate efflux, was sharply, concentration-dependently, increased; (ii) the phosphorylation status of intracellular markers known to be involved inthe neurodegenerative processes, in particular in Alzheimer disease: tau and GSK3b were increased, aswell as the protein level of b-secretase (BACE1) and p35/25 evaluated by Western blotting, while (iii)the cell metabolic activity, measured with the MTT method, was reduced, in a concentration- andtime-dependent manner. The latter effect, as well as tau hyperphosphorylation, was prevented both bya mixture of antioxidant drugs (100 lM ascorbic acid, 10 lM trolox, 100 lM glutathione) and by theanti-Alzheimer drug, memantine, 20 lM. Since it is well known that hippocampal cholinergic neuronsare particularly affected in Alzheimer disease, the effects of NaN3 and H2O2 were also studied in electri-cally stimulated rat hippocampal slices, evaluating the 3H-Choline efflux, as an index of acetylcholinerelease. The neurotoxic treatment depressed the neurosecretory function and the mixture of antioxidantdrugs, as well as memantine, were able to restore it. The neuronal damage induced by the in vitro pro-tocol adopted in the present work displays peculiarities of neurodegenerative disorders, e.g. Alzheimerdisease, underlining the role of mitochondrial failure and oxidative stress, which appear to occurupstream the neurodegenerative process; such protocol could be utilized to test the efficacy of neuropro-tective treatments.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Mitochondria play a central role in producing ATP as the sourceof cellular energy; their functions become less efficient duringbrain aging, including decreased activities of the electron transportchain (ETC.) enzymes, which in turn leads to enhanced radical oxy-gen species production (Lin and Beal, 2006; Petrozzi et al., 2007).

The role of mitochondrial dysfunction and oxidative stress inmany age-related neurodegenerative diseases is widely recognized(Lin and Beal, 2006). The most consistent failure in mitochondrialETC. enzymes reported in Alzheimer disease (AD) has beencytochrome c oxidase (COX, complex IV) (Castellani et al., 2002),whose activity has been found deficient in different brain regions,in particular in the cerebral cortex and hippocampus (Kish et al.,1999; Cottrell et al., 2002). However, it has not been fully clarifiedwhether mitochondrial impairment and oxidative stress are in-volved in the onset and progression of the disorder or are secondary

to other phenomena leading to neurodegeneration (Petrozzi et al.,2007; Pickrell et al., 2009). A synergistic link between energy failureand peptide b-amyloid (Ab) accumulation has been shown in sev-eral studies: COX inhibition leads to tau hyperphosphorylationand to Ab deposition; in turn, the activity of COX is inhibited bythe increased levels of Ab, thus establishing a vicious circle leadingto neuronal death (Müller et al., 2010; Swerdlow, 2012).

Sodium azide (NaN3) is a well-known COX inhibitor (Duranteauet al., 1998), which has been utilized over the last 20 years to in-duce metabolic compromise resulting in increase in amyloid pro-duction and changes in tau phosphorylation (Gabuzda et al.,1994; Blass et al., 1990), as well as memory deficit and neurode-generation in rats (Callaway et al., 2002; Berndt et al., 2001). Inprevious work, carried out in cerebral cortex primary neuronsand slices, we used NaN3 both alone and in combination with theglycolysis blocker, 2-deoxy-glucose, to mimic brain hypoxia andischemia. Besides its ability to induce mitochondrial failure, evalu-ated as transmembrane mitochondrial potential reduction (mea-sured by JC-1 fluorescence, Selvatici et al., 2009) its effects onvarious neurochemical parameters were studied in our laboratory:

Page 2

R. Selvatici et al. / Neurochemistry International 63 (2013) 112–120 113

intracellular calcium concentration (Marino et al., 2007), release ofacetylcholine, glutamate and nitric oxide (Cavallini et al., 2005), ki-nase expression and activity (Selvatici et al., 2006; Siniscalchi et al.,2006), type of neuronal death (Selvatici et al., 2009). In this workwe associated NaN3 with oxygen peroxide (H2O2), with the aimat mimicking conditions – mitochondrial dysfunction and oxida-tive stress – occurring in neurodegenerative disorders, in particularAD (Gao et al., 2007). A series of parameters useful in characteriz-ing neuronal damage were examined in rat cerebral cortexprimary neurons: (i) the basal release of glutamate, evaluated as3H-D-Aspartate efflux (Bianchi et al., 2007); (ii) the phosphoryla-tion status of intracellular markers known to be involved in theneurodegenerative processes, such as tau and GSK3b, and the pro-tein level of b-secretase, also known as beta-site amyloid precursorprotein cleaving enzyme 1 (BACE1), and the p25:p35ratio (Sadleirand Vassar, 2012) evaluated by Western blotting; (iii) the cellmetabolic activity, measured with the MTT method. The neuropro-tective effects of anti-oxidant treatments as well as of memantine,an anti-Alzheimer drug, were also challenged. Finally, since it iswell known that cholinergic neurons are particularly affected inAD (Schliebs and Arendt, 2011), the efflux of 3H-Choline, taken asan index of neurosecretory function (Cavallini et al., 2003), wasevaluated in rat hippocampal slices submitted to the combinedtreatment with NaN3 and H2O2.

2. Experimental procedures

2.1. Cortical neurons cultures

Sprague–Dawley rats were used within 24 h after birth. All theprocedures were carried out in accordance with European Commu-nity and national laws and policies, and were approved by the Eth-ics Committee of the University of Ferrara.

Cell cultures were prepared as previously described (Bianchi etal., 2007). Plastic Nunc dishes (Ø 35 mm) coated with poly-L-lysine20 lg/ml were used to plate the cells (1 � 106 per dish). Cells werecultured in a modified Neurobasal medium supplemented withGibco B-27 serum-free supplement 2%, gentamycin sulfate 50 lg/ml and Glutamax 500 lM. Cytosine arabinoside 5 lM was added48 h after plating to prevent glial cell proliferation. Correction ofthe volume (2 ml/dish) was made by adding 0.3–0.4 ml of freshmedium on the fourth day of incubation. The cultures were keptat 37 �C in a humidified atmosphere with 5% CO2.

2.1.1. [3H]-D-Aspartate effluxThe experiments were carried out on 7–9 DIV (days in vitro)

neuronal cultures. Two hours before use, the neurobasal mediumwas replaced with Krebs–Ringer buffer (KRB) composed of:118.5 mM NaCl; 4.8 mM KCl; 2.5 mM CaCl2; 1.2 mM KH2PO4;1.2 mM MgSO4; 11 mM glucose; 25 mM NaHCO3, gassed with 5%CO2 and 95% O2, pH 7.4. Eight dishes were simultaneously em-ployed for each experiment. After 30 min of preloading at 37 �Cwith [3H]-D-Aspartate (0.1 lM, specific activity 15.6 Ci/mmol,DuPontNEN), the dishes were inserted in a superfusion system at29 �C, containing eight small chambers (0.5 ml) as previously de-scribed (Bianchi et al., 2007). The perfusion rate of the KRB solutionwas set at 0.5 ml/min; samples were collected every 5 min for110 min. A 10-min treatment with NaN3 and/or H2O2 was appliedat the 55th min of superfusion. The radioactivity of the collectedsamples and the fraction retained in the cells (1 N NaOH extracts)was determined by liquid scintillation spectrometry. The tritiumoutflow was expressed either as pmoles of [3H]-D-Aspartate/dishor as percentage of the total tritium content at the moment ofthe collection (fractional release).

2.1.2. Western blot analysisNeuronal cells cultures were incubated with 0.1 mM NaN3 for

12 h, then 30 lM H2O2 was added to the medium in the last20 min of incubation. At the end of the treatment, the cells (sixdishes for each treatment) were lysed using ice-cold RIPA buffercontaining: 50 mM Tris, pH 8, 150 mM NaCl, 1 mM EDTA supple-mented with 1% NP-40, 0.1% SDS, 0.5% Deoxycholic acid, 2 mMphenyl-methyl-sulphonylfluoride (PMSF) and inhibitor cocktailEDTA free (Roche) for 30 min on ice and centrifuged at 12,000 g.The supernatant was collected and the protein concentration wasdetermined using the BCA protein assay kit (Thermo scientificPierce, Rockford, Illinois). Samples containing equal amounts ofproteins (25 lg) were diluted with loading buffer (187.5 mMTris–HCl, pH 6.8, 15% 2-ME, 0.1% sodium dodecylsulfate, 30% glyc-erol, 0.003% bromophenol blue), subjected to gel electrophoresison a 10% gel and then transferred to PVDF membranes (Bio-RadLaboratories, Milan, Italy) by electroblotting. Membranes were im-mersed overnight in a Tris buffered saline solution (TBS: 20 mMTris and 137 mM NaCl) pH 7.6 containing 5% Blotting grade blotter(Bio-Rad Laboratories, Milan, Italy) at 4 �C, washed three timeswith TBS plus 0.1% Tween 20 (TBS–T) and incubated overnight at4 �C with rabbit polyclonal antibodies against Tau, pTau (Ser262)and pTau (Ser404) (Santa Cruz, Heidelberg, Germany), GSK3b,pGSK3b (Ser9) and p35/25 (Cell Signaling Technology, Milan, Italy)and the monoclonal anti-b actin (Sigma) in TBS–T buffer. After 3washes with TBS–T buffer, horseradish peroxidase-labeledanti-rabbit IgG for polyclonal antibodies or anti-mouse IgG formonoclonal antibody was added at room temperature for 1 h.ECL Western blotting detection reagents (Amersham-PharmaciaBiotech, Milan, Italy) were used to visualize specific hybridizationsignals. These were expressed as optical density (OD) units/mm2.

2.1.3. MTT assayYellow MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazo-

lium bromide) is enzymatically converted to the blue formazanproduct only by metabolically active cells, and the absorbance isproportional to the number of viable cells (Van Meerloo et al.,2011; Stockert et al., 2012). Neuronal cultures (7–9 DIV) were trea-ted with NaN3 and/or H2O2 according to different protocols de-scribed in the Section 3.

The neuroprotective treatments were applied for 12 h, eitheralone or simultaneously with the neurotoxic insult, and main-tained in the medium until the end of the experiment. Drug con-centrations were chosen according to literature data (Vergun etal., 2001; Zou et al., 2002; Gao et al., 2007; Selvatici et al., 2009;Brittain et al., 2012). At the end of the treatment the mediumwas removed and replaced with 1 ml medium containing MTT(0.5 mg/ml) and the blue color was allowed to develop for 4 h, then1 ml isopropylalcohol:HCl 9:1 was added to solubilize the bluecrystals. Samples were read at a test wavelength of 570 nm; theabsorbance data are expressed as percentages of control groups.

2.2. Hippocampal slices

The experiments were carried out on male Sprague–Dawley rats(200–300 g), in accordance with protocols approved by the EthicsCommittee of the University of Ferrara. After decapitation underlight anesthesia, hippocampal slices (400 lm thick) were preparedas previously described (Cavallini et al., 2003).

2.2.1. 3H-Choline effluxAfter 30 min of recovery, the slices were incubated with 0.1 lM

3H-Choline (specific activity 86 Ci/mmol, DuPontNEN) at 37 �C for30 min, then they were washed in cold KRB (mM composition:NaCl 118.5, KCl 4.7, CaCl2 1.25, MgSO4 1.2, KH2PO4 1.2, NaHCO3

25, glucose 10, pH 7.4), and set up in superfusion chambers (kept

Page 3

114 R. Selvatici et al. / Neurochemistry International 63 (2013) 112–120

at 37 �C and equipped with stimulating platinum electrodes). Theslices were superfused (0.25 ml/min) with 10 lM hemicholin-ium-3 containing KRB bubbled with 95% O2, 5% CO2 and submittedto continuous electrical stimulation (3 Hz, 1 ms, 20 mA/cm2, unlessotherwise specified), applied throughout the experiment, startingat the 40th min of superfusion. Superfusate samples were collectedevery 5 min; at the end of the experiments, the radioactivity of the5-min samples and of the slices (solubilized with 1 ml 1 N NaOH)was determined by liquid scintillation counting; the fractional re-lease, i.e. the amount of released tritium as a percentage of the tri-tium content at the onset of the respective collection period, wascalculated (Cavallini et al., 2003). The treatments with NaN3 and/or H2O2 and with neuroprotective drugs were carried out as de-tailed in the Results section.

2.3. Chemicals

Neurobasal™ medium, L-alanyl-L-glutamine (Glutamax™) andGibco™ B-27 supplement were from InvitroGen (San GiulianoMilanese, Italy); L-ascorbic acid, cytosine arabinoside, gentamycinsulfate, L-glutathione (GSH), hemicholinium-3, memantinehydrochloride, sodium azide, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), tetrodotoxin, (±)-6-Hydro-xy-2,5,7,8-tetramethylchromane-2-carboxylic acid (Trolox™, awater-soluble derivative of vitamin E) were from Sigma–AldrichChemical Company (St. Louis, MO, USA); all other chemicals werefrom standard commercial sources and were of the highest purityavailable.

2.4. Statistical analysis

Data are given as means ± SEM. The significance of differencesbetween treated and control samples has been assessed with Stu-dent’s t test for unpaired data, and with the analysis of variance(ANOVA) followed by Dunnett’s multiple comparison test.

3. Results

3.1. Cerebral cortex neuronal cultures

3.1.1. Effects of NaN3 and of H2O2 on [3H]-D-Aspartate effluxThe increase in glutamate extracellular concentration is consid-

ered an index of neuronal damage (Choi, 1988; Lau and Tymianski,2010; Wang and Qin, 2010).

In control neuronal cultures the basal release of glutamate,evaluated as [3H]-D-Aspartate efflux, was 0.052 ± 0.002 pmol/min/dish (0.51% of the tritium content) at the 50th min of superfusionand slowly decayed to 79 ± 6% at the 90th min of superfusion. Thetreatment with H2O2 (150–600 lM for 10 min at the 55th min)induced a concentration-dependent increase in basal [3H]D-Aspar-

Table 1[3H]-D-Aspartate efflux from rat cerebral cortex primary neurons in culture. Effects of H2O

Treatment (at the 55th min) None

65th

None (control) 89 ± 3 (9)H2O2 150 lM � 10 min 133 ± 12* (6)H2O2 300 lM � 10 min 144 ± 14** (6)H2O2 600 lM � 10 min 184 ± 37** (3)

Efflux data are expressed as % of the efflux at the 50th min.* P < 0.05.** P < 0.01, vs. control.� P < 0.05.�� P < 0.01, vs. NaN3 3 mM alone.§ P < 0.01 vs. the corresponding group without NaN3.

tate efflux (Table 1). The addition of 3 mM NaN3, which per se onlymoderately increased the basal [3H]-D-Aspartate efflux, potenti-ated the effect induced by H2O2 (Table 1). The effect of the associ-ation 3 mM NaN3 plus H2O2 150 lM also depended on the time ofapplication: [3H]-D-Aspartate efflux at the 65th min was 145 ± 3%of the control with a 5 min application and 274 ± 18% with a15 min application (each n = 4, P < 0.01). The observed increasesin [3H]-D-Aspartate efflux were not prevented by adding 1 lMtetrodotoxin to the superfusion medium (not shown).

3.1.2. Effects of NaN3 and of H2O2 on cell metabolic activityThe above experiments of [3H]D-Aspartate efflux indicate that

the association NaN3 + H2O2 induced acute neuronal damage.Accordingly, in cultures exposed to NaN3 3 mM + H2O2 150 lMfor 30 min (n = 5), a reduction in cell metabolic activity to49 ± 3% of the control was detected with the MTT method (VanMeerloo et al., 2011).

In order to better mimic the chronic mitochondrial failure ob-served in neurodegenerative diseases, other experimental proto-cols were tested, exposing the neurons to prolonged treatmentwith NaN3. Such experiments required lower drug concentrations;in fact, in preliminary experiments, 3 mM NaN3 was able to reducecell metabolic activity to as much as 33 ± 4% when applied alonefor 4 h and to 21 ± 5% when H2O2 150 lM was added in the last20 min (each n = 4). Fig. 1 shows the results of the experiments car-ried out to choose optimal experimental conditions, in which lower(0.01–1 mM) concentrations of NaN3 were applied for 12 h and30 lM H2O2 was added in the last 20 min. The association0.1 mM NaN3 plus 30 lM H2O2 reduced cell metabolic activity to43.8 ± 4% (n = 10) and was considered suitable to perform furtherexperiments.

3.1.3. Effects of NaN3 plus H2O2 on intracellular markers ofneurodegeneration

Primary rat cortical neurons were exposed to 0.1 mM NaN3 for12 h and to 30 lM H2O2 in the last 20 min of incubation. The phos-phorylation levels of the intracellular markers Tau262, Tau404 andGSK3b were evaluated by Western blotting; anti-b actin monoclo-nal antibody was used as loading control. As shown in Fig. 2, trea-ted samples showed greater phosphorylation levels of pTau262,pTau404 and pGSK3b in comparison to the untreated controlsamples.

Since p25 has been found to accumulate in brains of Alzhei-mer’s disease patients (Patrick et al., 1999) and the increase inBACE1 level/activity contributes to the pathogenesis of sporadicAD (Sadleir and Vassar, 2012), the conversion of p35 to p25 andthe protein level of BACE1 were also tested. As expected, theappearance of p25 and the increase of BACE1 were observed inthe NaN3/H2O2 treated neurons (Fig. 3).

2 alone and in the presence of NaN3.

NaN3 3 mM � 10 min

90th 65th 90th

79 ± 6 112 ± 10** (9) 78 ± 398 ± 3 213 ± 10�,§ (46) 108 ± 3

104 ± 9 239 ± 20��,§ (13) 117 ± 7152 ± 35** 283 ± 79�� (8) 155 ± 15�

Page 4

Fig. 1. Rat cerebral cortex primary neurons in culture: neuronal metabolic activity,evaluated with the MTT test and expressed as % of control cells. Neurons wereincubated with 0.1 mM NaN3 for 12 h and 30 lM H2O2 was added in the last 20 minof incubation. The figures over the bars indicate the number of experiments.⁄P < 0.05, ⁄⁄P < 0.01, vs. the respective controls, one way-ANOVA, followed byDunnett’s multiple comparison test.

R. Selvatici et al. / Neurochemistry International 63 (2013) 112–120 115

3.1.4. Effects of neuroprotective treatmentsThe above experiments showed that the adopted protocol

(0.1 mM NaN3 for 12 h plus 30 lM H2O2 in the last 20 min) wasable to trigger mechanisms known to be involved in chronic neu-rodegenerative diseases. The following experiments were aimedat verifying the effectiveness of neuroprotective treatments. Toprevent oxidative stress, antioxidant drugs (100 lM ascorbic acid,10 lM trolox, 100 lM GSH) were applied together to the neuronalcultures for 12 h (simultaneously with NaN3/H2O2). This treatmentis from here forward referred to as ‘‘antioxidant mix’’. Moreover, awell-known anti-Alzheimer drug, memantine, was tested at 20 lMconcentration, following the same experimental protocol. Neitherthe antioxidant mix, nor memantine were able to affect either cellmetabolic activity, or tau phosphorylation status in control cul-tures (Fig. 4 and Table 2). Conversely, in neuronal cultures exposedto the neurotoxic insult (0.1 mM NaN3 for 12 h plus 30 lM H2O2 inthe last 20 min) both treatments were able to significantly improvecell metabolic activity and to prevent tau hyperphosphorylation(Fig. 4 and Table 2).

3.2. Hippocampal slices

A failure in cholinergic function in the hippocampus is recog-nized as an important feature of age-related neurodegenerativediseases, particularly AD (Craig et al., 2011). In the following exper-iments, carried out in adult rat hippocampal slices, 3H-Choline ef-flux was studied as an index of acetylcholine release, and theeffects of the application of NaN3 + H2O2 were tested. Preliminaryexperiments, performed in order to choose the most suitable pro-tocol, showed that adult hippocampal slices do not survive as longas neuronal cultures, so that it was not possible to study long-termeffects of the treatments.

The basal tritium efflux from 3H-Choline-labeled hippocampalslices at the 35th min of superfusion was 4.6 ± 0.43 fmol/min(n = 6). To mimick physiological conditions of neuronal activity

(Badini et al., 1997; Cavallini et al., 2005), a continuous electricalstimulation was applied from the 40th min of superfusion on-wards. The electrical stimulation evoked a frequency-dependentincrease in 3H-Choline efflux: in control slices the stimulated(45th min)/basal (35th min) 3H-Choline efflux ratios were:1.11 ± 0.049 at 1 Hz, 1.72 ± 0.16 at 3 Hz, 2.08 ± 0.10 at 10 Hz (eachn = 6). In the presence of 1 lM tetrodotoxin the 45th/35th ratio in3 Hz stimulated slices was reduced to 0.85 ± 0.11 (n = 3). The 3 Hzfrequency was chosen for the following experiments. The concen-trations of NaN3 and H2O2 were chosen according to the results ofthe experiments of glutamate efflux (see Section 2.1.1) and on thebasis of our previous publications showing, respectively, thedependence of neuronal damage on concentration and time ofapplication of the toxic insult (Selvatici et al., 2009) and the effectsof NaN3 on acetylcholine release (Cavallini et al., 2005). As shownin Fig 5, the evoked 3H-Choline efflux from hippocampal slices wassignificantly reduced by 3 mM NaN3 plus 150 lM H2O2, applied to-gether for 15 min at the beginning of the superfusion.

The same protocol was used to test potential neuroprotectivetreatments. Both the antioxidant mix (100 lM ascorbic acid,10 lM trolox, 100 lM GSH), and the anti-Alzheimer drug, meman-tine (20 lM) were able to fully prevent the failure in synapticactivity induced by NaN3 plus H2O2. Neither treatment did, perse, modify the electrically (3 Hz) evoked 3H-Choline efflux (Fig. 5).

4. Discussion

The aim of the present work was to induce bioenergetic condi-tions similar to those present during the early stages of neurode-generative diseases, namely Alzheimer disease (AD), and topropose an in vitro model suitable to study the biochemicalchanges occurring at the cellular level and to assay substances withpotential therapeutic effect.

Mitochondrial dysfunction and oxidative damage occur early inAD, and seem to precede the onset of Ab aggregates, tau patholo-gies, synaptic dysfunction, and inflammation of brain (Praticò etal., 2001; Lin and Beal, 2006; Reddy, 2008). Thus, the cytochromec oxidase (COX, complex IV) inhibitor, NaN3, and H2O2 were usedto induce mitochondrial failure and oxidative stress, respectively.The most relevant findings were: in primary neuronal cells of ratcerebral cortex (i) the basal glutamate efflux was increased; (ii)the phosphorylation status of tau and of GSK3b were increased,as well as the p35/25 and BACE1 protein level; (iii) cell metabolicactivity was reduced in a concentration- and time of exposition-dependent manner; (iv) pretreatment either with a mixture ofantioxidant drugs (ascorbic acid, trolox, GSH) or with memantine,an anti-Alzheimer drug, prevented both metabolic activity reduc-tion and tau hyperphosphorylation. Finally, in electrically stimu-lated hippocampal slices, the combined treatment with NaN3 andH2O2 induced a reduction in cholinergic neuronal function, pre-vented both by the mixture of antioxidant drugs and by meman-tine. The present findings contribute to ongoing debate onwhether the mitochondrial failure occurs upstream or downstreamto tau hyperphosphorylation and Aß deposition (Müller et al.,2010; Swerdlow, 2012; Lee et al., 2012) supporting the view thatit occurs upstream.

4.1. Cortical primary neurons

The in vitro model of neuronal cultures, suitable to study bothshort- and long-term effects of treatments (Giordano and Costa,2011), was used to examine different markers of neuronal failuretaking place in different stages of the disease. Although tumor-derived cell lines are widely used in neurochemical studies, theyoften display morphological, developmental, and signaling

Page 5

Fig. 2. Phosphorylation status of Tau protein and GSK3b kinase (pSer9). (A) Representative Western blot of rat cerebral cortex primary neurons: (1) untreated (as control) and(2) treated with NaN3 (0.1 mM) for 12 h and H2O2 (30 uM) for the last 20 min of incubation. Experiments were performed in triplicate. Proteins levels were evaluated usingbactin immunodetection to normalizing bands optical densities. (B) Quantitative analysis of phosphorylation levels of pTau262, pTau404 and pGSK3b as percent of controlsample. Data are from densitometry measurements made from at least three separate experiments and plotted with respect to controls. Data are mean values ± SEM. ⁄P < 0.05vs. untreated control.

116 R. Selvatici et al. / Neurochemistry International 63 (2013) 112–120

characteristics substantially different from those of neurons, sothat their response may differ as well (LePage et al., 2005). There-fore, primary neuronal cultures have been chosen for the presentstudy, since they offer the advantage of reproducing the character-istics of brain tissue more closely.

It is well known that glutamate plays a relevant role in neuro-degenerative diseases, triggering a neurotoxicity cascade leadingto neuronal death (Lau and Tymianski, 2010; Wang and Qin,2010). In previous works, we showed the ability of NaN3 to in-crease glutamate efflux, both from cortical neuronal cultures andfrom cortical slices (Marino et al., 2007; Cavallini et al., 2005).The effect cannot be attributed to an increase in neuronal activity,as shown by the ineffectiveness of tetrodotoxin in preventing it,but it can be explained by the inversion of the glutamate trans-porter under low-energy conditions (Phillis and O’Regan, 2003).Here we demonstrate that H2O2 displays a similar effect and thata low concentration of NaN3, hardly able to modify glutamate ef-flux, potentiated the concentration-dependent increase inducedby H2O2. This finding can be interpreted in agreement with Gaoet al. (2007): the mitochondrial dysfunction induced by NaN3

treatment renders the neurons more prone to the oxidative stress

induced by H2O2. The effect of the combined treatment on gluta-mate efflux is rapid in onset and short in duration and may repre-sent the first step of neurodegeneration; in fact, a more prolongedexposition to the combined treatment leads to higher glutamateextracellular concentrations and, consequently, to neuronal death.In addition, it is noteworthy that NaN3 has previously been re-ported to release NO (Cavallini et al., 2005), so inducing nitrosativestress, which contributes to the pathogenesis of neurodegenerativediseases (Nakamura et al., 2012). The latter observation furthersupports the validity of the model used in this work.

The reduction in cell metabolic activity obtained with the com-bined treatment with NaN3 and H2O2 depended both on drug con-centrations and on incubation times. The protocol adopted in thesuccessive experiments was opportunely chosen to mimic theage-related mitochondrial failure, followed by oxidative stress, asseen in AD (Lin and Beal, 2006; Lee et al., 2012): a prolonged treat-ment with a very low concentration of NaN3 preceded a brief oxi-dative stress induced by H2O2 at the end of the incubation.

Hyperphosphorylation of tau, triggering neurofibrillar tanglesformation is a central biochemical event in AD pathogenesis(Chung, 2009). Tau is a phosphoprotein which potentially has 80

Page 6

Fig. 3. Rat cerebral cortex primary neurons in culture: effects of protective treatments on neuronal metabolic activity, evaluated with the MTT test and expressed as % ofcontrol cells. Neurons were incubated with 0.1 mM NaN3 for 12 h and 30 lM H2O2 was added in the last 20 min of incubation. Either ‘‘antioxidant mix’’ (100 lM ascorbic acid,10 lM trolox, 100 lM GSH), or 20 lM memantine were applied to the neuronal cultures throughout 12 h-incubation. The figures over the bars indicate the number ofexperiments. ⁄⁄P < 0.01, vs. NaN3 + H2O2, one way-ANOVA, followed by Dunnett’s multiple comparison test.

R. Selvatici et al. / Neurochemistry International 63 (2013) 112–120 117

serine/threonine and five tyrosine phosphorylation sites (Wang etal., 2013). A number of sites, including Ser262 (Biernat et al., 1993)and Ser396/404 (Bramblett et al., 1993), has been shown to interferewith binding to microtubules (Evans et al., 2000) and has beenassociated with neurons in the ‘‘pre-tangle’’ stadium in the ADbrains (Lauckner et al., 2003; Augustinack et al., 2002; Vingtdeuxet al., 2011). GSK-3b and cyclin-dependent protein kinase-5(cdk5)/p25, are the kinases primarily responsible in abnormalhyperphosphorylation of tau (Iqbal and Grundke-Iqbal, 2005),although other kinases such as PKC, PKA and ERK2 are also in-volved. Over-expression of p25, deriving from calpain cleavage ofp35, has been suggested to be involved not only in the hyper-phosphorylation of tau protein (Noble et al., 2003) but also in theformation of Ab in AD (Cruz et al., 2006), possibly coordinatingthe action of BACE1 (Wen et al., 2008). BACE1 levels are increasedin AD, potentially accelerating the initiating event for Ab produc-tion (Sadleir and Vassar, 2012).

Here we demonstrate that the combined treatment with NaN3

and H2O2 induces hyperphosphorylation of tau protein both inSer262 and Ser404 and increases the levels of both p25 and BACE1,suggesting that the combined treatment mimics the early events

of AD. Surprisingly, we detected a significant increase in GSK3bphosphorylation at Ser9, which generally is assumed to be associ-ated with the inactivation of this kinase (Jope and Johnson,2004). Such an increase could be explained as a compensatory phe-nomenon linked to the excessive release of glutamate observed un-der our experimental conditions and to the mechanisms of hypoxiaand cell death (Peineau et al., 2008; Iida et al., 2012). Indeed, cellsare able to counteract to harmful conditions, triggering compensa-tory strategies (Hindle, 2010).

The latter findings show that the combined treatment with NaN3

and H2O2 triggers a series of events which are characteristic ofage-related neurodegenerative diseases, in particular AD. The sameprotocol was challenged to display the efficacy of potential neuro-protective treatments, using the MTT test, which measures the abil-ity of cells to carry out an energy-dependent enzymatic reactionand is considered a marker of metabolic activity of viable cells(van Meerloo et al., 2011). With the chosen concentrations andtimes of incubation, about half of the neurons maintained theirmetabolic activity, so that possible favorable drug effects could bedetected. As expected, the mixture of antioxidants was able tosignificantly prevent the loss of metabolic activity, mainly

Page 7

Fig. 4. Rat cerebral cortex primary neurons in culture: effects of protectivetreatments on neuronal viability, expressed as % of control cells. Neurons wereincubated with 0.1 mM NaN3 for 12 h and 30 lM H2O2 was added in the last 20 minof incubation. Either ‘‘antioxidant mix’’ (100 lM ascorbic acid, 10 lM trolox,100 lM GSH), or 20 lM memantine were applied to the neuronal culturesthroughout 12 h-incubation. The figures over the bars indicate the number ofexperiments. ⁄⁄P < 0.01, vs. NaN3 + H2O2, one way-ANOVA, followed by Dunnett’smultiple comparison test.

Table 2Western blotting analysis of p-Tau (Ser262): effects of NaN3, H2O2 and of neuropro-tective treatments.

Treatment None NaN3

3 mM � 12 h

None (control) 4.23 ± 0.39 8.37 ± 0.89**

H2O2 30 lM � 20 min 3.96 ± 0.40 11.6 ± 1.08�§

Antioxidant mixa � 12 h 3.78 ± 0.41 –Antioxidant mixa � 12 h + H2O2

30 lM � 20 min– 2.77 ± 0.32#

Memantine20 lM � 12 h 4.8 ± 0.50 –Memantine 20 lM � 12 h + H2O2

30 lM � 20 min– 4.5 ± 0.48#

Neuronal cells cultures were incubated with 0.1 mM NaN3 for 12 h, then 30 lMH2O2 was added to the medium in the last 20 min of incubation. Data are expressedas optical density (OD) units/mm2, means ± SEM.

a mixture of antioxidant drugs (100 lM ascorbic acid, 10 lM trolox, 100 lMGSH).** P < 0.01, vs. control.� P < 0.05, vs. NaN3 3 mM alone.§ P < 0.001 vs. control and vs. H2O2 alone.

# P < 0.001 vs. NaN3 + H2O2, Bonferroni’s multiple comparison test.

Fig. 5. 3H-Choline efflux from superfused rat hippocampal slices. Electrical (3 Hz)stimulation was continuously applied from the 40th min onward. 3 mM NaN3 plus30 lM H2O2 were added in the fluid for 15 min at the onset superfusion; either‘‘antioxidant mix’’ (100 lM ascorbic acid, 10 lM trolox, 100 lM GSH), or 20 lMmemantine were applied throughout the superfusion. Histograms represent thefractional tritium release detected in the samples collected from the 40th to the70th min. The figures over the bars indicate the number of experiments. ⁄⁄P < 0.01,vs. control; �P < 0.05, ��P < 0.01, vs. NaN3 + H2O2, one way-ANOVA, followed byDunnett’s multiple comparison test.

118 R. Selvatici et al. / Neurochemistry International 63 (2013) 112–120

counteracting the oxidative stress induced by H2O2; a mere direct,non-biological interaction with H2O2 is unlikely, since neuroprotec-tion by antioxidant drugs has been shown even against the neuro-toxic effect induced by the treatment with NaN3 alone (Selvaticiet al., 2009). The efficacy of memantine could be attributed to theability of this anti-Alzheimer drug to inhibit excessive NMDA recep-tor activity and excessive production of NO (Gu et al., 2010). Inter-estingly, both neuroprotective treatments resulted effective as wellin preventing the increase induced by the combined treatment with

NaN3 and H2O2 on tau phosphorylation, which can be considered anindex of tangle formation in AD (Chung, 2009).

4.2. Hippocampal slices

The synaptic damage which initially characterizes the neurode-generative process in AD (Crews and Masliah, 2010) was studied inhippocampal slices. The main drawback of this model consists inthe short survival time of the isolated tissue which does not allowto perform long-term experiments, and to use very low concentra-tions of NaN3 and H2O2. However, we can point out several favor-able aspects: (i) the tissue is taken from adult animals; (ii) thebrain area chosen contains cholinergic terminals involved in cogni-tive functions which are gradually lost in AD; (iii) the slices aresubmitted to continuous electrical stimulation, to mimic neuronalactivity, at a physiological frequency (Sotty et al., 2003); (iv) bothmitochondrial failure and oxidative stress have been reported inhippocampus of AD patients (Cottrell et al., 2002; Lovell and Mar-kesbery, 2007). Thus, the reduction in 3H-Choline efflux observedin the hippocampal slices treated with the combination of NaN3

and H2O2 may represent the well-known cholinergic failure ob-served in AD. Interestingly, the efficacy of the neuroprotectivetreatments – memantine and the antioxidant mix seen in the neu-ronal cultures – was confirmed by these experiments, further add-ing to the value of the proposed model.

5. Conclusions

The treatment protocol described in this work induces condi-tions of neuronal suffering that may be considered similar to thosefound in AD, as demonstrated by typical markers, such as tauhyperphosphorylation and the increased level of p25 and BACE1.

Page 8

R. Selvatici et al. / Neurochemistry International 63 (2013) 112–120 119

The latter represents one of the last steps in the chain of eventsthat lead to the deposition of amyloid plaques (Hooper et al.,2008). Thus, mitochondrial failure and oxidative stress seem toprecede amyloid deposition and, consequently, the neurodegener-ation. However, it should be kept in mind that, in turn, amyloiddeposition increases tau phosphorylation and secretase expres-sion/activity, so triggering a vicious circle further contributing toneuronal damage (Müller et al., 2010; Lee et al., 2012).

The efficacy of neuroprotective treatments, displayed for thefirst time even in the more complex model of stimulated hippocam-pal slices, supports the usefulness of the proposed protocol as amodel to study biochemical changes induced by AD at the cellularlevel, and to test compound endowed with therapeutic potential.

Acknowledgements

This work was supported by Grants from University of Ferrara;Associazione Emma e Ernesto Rulfo per la Genetica Medica, Parma,Italy; Fondazione Cassa di Risparmio di Ferrara, Italy. The Authorswish to thank Dr. Joseph Mayo for the English revision of the text.

References

Augustinack, J.C., Schneider, A., Mandelkow, E.M., Hyman, B.T., 2002. Specific tauphosphorylation sites correlate with severity of neuronal cytopathology inAlzheimer’s disease. Acta Neuropathol. 103, 26–35.

Badini, I., Beani, L., Bianchi, C., Marzola, G., Siniscalchi, A., 1997. Post-ischemicrecovery of acetylcholine release in vitro: influence of different excitatoryamino acid receptor subtype antagonists. Neurochem. Int. 31, 817–824.

Berndt, J.D., Callaway, N.L., Gonzalez-Lima, F., 2001. Effects of chronic sodium azideon brain and muscle cytochrome oxidase activity: a potential model toinvestigate environmental contributions to neurodegenerative diseases. J.Toxicol. Environ. Health 63, 67–77.

Bianchi, C., Marani, L., Marino, S., Barbieri, M., Nazzaro, C., Beani, L., Siniscalchi, A.,2007. Serotonin modulation of cell excitability and of [3H]GABA and [3H]-D-aspartate efflux in primary cultures of rat cortical neurons. Neuropharmacology52, 995–1002.

Biernat, J., Gustke, N., Drewes, G., Mandelkow, E.M., Mandelkow, E., 1993.Phosphorylation of Ser262 strongly reduces binding of tau to microtubules:distinction between PHF-like immunoreactivity and microtubule binding.Neuron 11, 153–163.

Blass, J.P., Baker, A.C., Ko, L., Black, R.S., 1990. Induction of Alzheimer antigens by anuncoupler of oxidative phosphorylation. Arch. Neurol. 47 (8), 864–869.

Bramblett, G.T., Goedert, M., Jakes, R., Merrick, S.E., Trojanowski, J.Q., Lee, V.M.,1993. Abnormal tau phosphorylation at Ser396 in Alzheimer’s diseaserecapitulates development and contributes to reduced microtubule binding.Neuron 10 (6), 1089–1099.

Brittain, M.K., Brustovetsky, T., Sheets, P.L., Brittain, J.M., Khanna, R., Cummins, T.R.,Brustovetsky, N., 2012. Delayed calcium dysregulation in neurons requires boththe NMDA receptor and the reverse Na+/Ca2+ exchanger. Neurobiol. Dis. 46 (1),109–117.

Callaway, N.L., Riha, P.D., Wrubel, K.M., McCollum, D., Gonzalez-Lima, F., 2002.Methylene blue restores spatial memory retention impaired by an inhibitor ofcytochrome oxidase in rats. Neurosci. Lett. 332 (2), 83–86.

Castellani, R., Hirai, K., Aliev, G., Drew, K.L., Nunomura, A., Takeda, A., Cash, A.D.,Obrenovich, M.E., Perry, G., Smith, M.A., 2002. Role of mitochondrialdysfunction in Alzheimer’s disease. J. Neurosci. Res. 70 (3), 357–360.

Cavallini, S., Marino, S., Beani, L., Bianchi, C., Siniscalchi, A., 2003. Nociceptininhibition of acetylcholine efflux from different brain areas. Neuroreport 14(17), 2167–2170.

Cavallini, S., Marti, M., Marino, S., Selvatici, R., Beani, L., Bianchi, C., Siniscalchi, A.,2005. Effects of chemical ischemia in cerebral cortex slices. Focus on nitricoxide. Neurochem. Int. 47 (7), 482–490.

Choi, D.W., 1988. Glutamate neurotoxicity and diseases of the nervous system.Neuron 1, 623–634.

Chung, S.H., 2009. Aberrant phosphorylation in the pathogenesis of Alzheimer’sdisease. BMB Rep. 2 (8), 467–474.

Cottrell, D.A., Borthwick, G.M., Johnson, M.A., Ince, P.G., Turnbull, D.M., 2002. Therole of cytochrome c oxidase deficient hippocampal neurones in Alzheimer’sdisease. Neuropathol. Appl. Neurobiol. 28, 390–396.

Craig, L.A., Hong, N.S., McDonald, R.J., 2011. Revisiting the cholinergic hypothesis inthe development of Alzheimer’s disease. Neurosci. Biobehav. Rev. 35, 1397–1409.

Crews, L., Masliah, E., 2010. Molecular mechanisms of neurodegeneration inAlzheimer’s disease. Hum. Mol. Genet. 19 (R1), R12–R20.

Cruz, J.C., Kim, D., Moy, L.Y., Dobbin, M.M., Sun, X., Bronson, R.T., Tsai, L.H., 2006.P25/cyclin-dependent kinase 5 induces production and intraneuronalaccumulation of amyloid beta in vivo. J. Neurosci. 26 (41), 10536–10541.

Duranteau, J., Chandel, N.S., Kulisz, A., Shao, Z., Schumacker, P.T., 1998. Intracellularsignaling by reactive oxygen species during hypoxia in cardiomyocytes. J. Biol.Chem. 273, 11619–11624.

Evans, D.B., Rank, K.B., Bhattacharya, K., Thomsen, D.R., Gurney, M.E., Sharma, S.K.,2000. Tau phosphorylation at serine 396 and serine 404 by human recombinanttau protein kinase II inhibits tau’s ability to promote microtubule assembly. J.Biol. Chem. 275 (32), 24977–24983.

Gabuzda, D., Busciglio, J., Chen, L.B., Matsudaira, P., Yankner, B.A., 1994. Inhibition ofenergy metabolism alters the processing of amyloid precursor protein andinduces a potentially amyloidogenic derivative. J. Biol. Chem. 269 (18), 13623–13628.

Gao, J., Zhu, Z.R., Ding, H.Q., Qian, Z., Zhu, L., Ke, Y., 2007. Vulnerability of neuronswith mitochondrial dysfunction to oxidative stress is associated with down-regulation of thioredoxin. Neurochem. Int. 50, 379–385.

Giordano, G., Costa, L.G., 2011. Primary neurons in culture and neuronal cell linesfor in vitro neurotoxicological studies. Methods Mol. Biol. 758, 13–27.

Gu, Z., Nakamura, T., Lipton, S.A., 2010. Redox reactions induced by nitrosativestress mediate protein misfolding and mitochondrial dysfunction inneurodegenerative diseases. Mol. Neurobiol. 41 (2–3), 55–72.

Hindle, J.V., 2010. Ageing, neurodegeneration and Parkinson’s disease. Age Ageing39 (2), 156–161.

Hooper, C., Killick, R., Lovestone, S., 2008. The GSK3 hypothesis of Alzheimer’sdisease. J. Neurochem. 104 (6), 1433–1439.

Iida, Y., Aoki, K., Asakura, T., Ueda, K., Yanaihara, N., Takakura, S., Yamada, K.,Okamoto, A., Tanaka, T., Ohkawa, K., 2012. Hypoxia promotes glycogensynthesis and accumulation in human ovarian clear cell carcinoma. Int. J.Oncol. 40 (6), 2122–2130.

Iqbal, K., Grundke-Iqbal, I., 2005. Metabolic/signal transduction hypothesis ofAlzheimer’s disease and other tauopathies. Acta Neuropathol. 109 (1), 25–31.

Jope, R.S., Johnson, G.V., 2004. The glamour and gloom of glycogen synthase kinase-3. Trends Biochem. Sci. 29 (2), 95–102.

Kish, S.J., Mastrogiacomo, F., Guttman, M., Furukawa, Y., Taanman, J.W., Dozic, S.,Pandolfo, M., Lamarche, J., DiStefano, L., Chang, L.J., 1999. Decreased brainprotein levels of cytochrome oxidase subunits in Alzheimer’s disease and inhereditary spinocerebellar ataxia disorders: a nonspecific change? J.Neurochem. 72 (2), 700–707.

Lau, A., Tymianski, M., 2010. Glutamate receptors, neurotoxicity andneurodegeneration. Pflugers Arch. – Eur. J. Physiol. 460, 525–542.

Lauckner, J., Frey, P., Geula, C., 2003. Comparative distribution of tauphosphorylated at Ser262 in pre-tangles and tangles. Neurobiol. Aging 24 (6),767–776.

Lee, S.H., Kim, K.R., Ryu, S.Y., Son, S., Hong, H.S., Mook-Jung, I., Lee, S.H., Ho, W.K.,2012. Impaired short-term plasticity in mossy fiber Synapses caused bymitochondrial dysfunction of dentate granule cells is the earliest synapticdeficit in a mouse model of Alzheimer’s disease. J. Neurosci. 32 (17), 5953–5963.

LePage, K.T., Dickey, R.W., Gerwick, W.H., Jester, E.L., Murray, T.F., 2005. On the useof neuro-2a neuroblastoma cells versus intact neurons in primary culture forneurotoxicity studies. Crit. Rev. Neurobiol. 17 (1), 27–50.

Lin, M.T., Beal, M.F., 2006. Mitochondrial dysfunction and oxidative stress inneurodegenerative diseases. Nature 443 (7113), 787–795.

Lovell, M.A., Markesbery, W.R., 2007. Oxidative DNA damage in mild cognitiveimpairment and late-stage Alzheimer’s disease. Nucleic Acids Res. 35 (22),7497–7504.

Marino, S., Marani, L., Nazzaro, C., Beani, L., Siniscalchi, A., 2007. Mechanisms ofsodium azide-induced changes in intracellular calcium concentration in ratprimary cortical neurons. Neurotoxicology 8 (3), 622–629.

Müller, W.E., Eckert, A., Kurz, C., Eckert, G.P., Leuner, K., 2010. Mitochondrialdysfunction: common final pathway in brain aging and Alzheimer’s disease –therapeutic aspects. Mol. Neurobiol. 41, 159–171.

Nakamura, T., Cho, D.H., Lipton, S.A., 2012. Redox regulation of protein misfolding,mitochondrial dysfunction, synaptic damage, and cell death inneurodegenerative diseases. Exp. Neurol. 23, 812–821.

Noble, W., Olm, V., Takata, K., Casey, E., Mary, O., Meyerson, J., Gaynor, K.,LaFrancois, J., Wang, L., Kondo, T., Davies, P., Burns, M., Veeranna, Nixon, R.,Dickson, D., Matsuoka, Y., Ahlijanian, M., Lau, L.F., Duff, K., 2003. Cdk5 is a keyfactor in tau aggregation and tangle formation in vivo. Neuron 38 (4),555–565.

Patrick, G.N., Zukerberg, L., Nikolic, M., de la Monte, S., Dikkes, P., Tsai, L.H., 1999.Conversion of p35 to p25 deregulates Cdk5 activity and promotesneurodegeneration. Nature 402 (6762), 615–622.

Peineau, S., Bradley, C., Taghibiglou, C., Doherty, A., Bortolotto, Z.A., Wang, Y.T.,Collingridge, G.L., 2008. The role of GSK-3 in synaptic plasticity. Br. J. Pharmacol.153 (Suppl. 1), S428–S437.

Petrozzi, L., Ricci, G., Giglioli, N.J., Siciliano, G., Mancuso, M., 2007. Mitochondria andNeurodegeneration. Biosci. Rep. 27, 87–104.

Phillis, J.W., O’Regan, M.H., 2003. Characterization of modes of release of aminoacids in the ischemic/reperfused rat cerebral cortex. Neurochem. Int. 43 (4–5),461–467.

Pickrell, A.M., Fukui, H., Moraes, C.T., 2009. The role of cytochrome c oxidasedeficiency in ROS and amyloid plaque formation. J. Bioenerg. Biomembr. 41,453–456.

Praticò, D., Uryu, K., Leight, S., Trojanoswki, J.Q., Lee, V.M., 2001. Increased lipidperoxidation precedes amyloid plaque formation in an animal model ofAlzheimer amyloidosis. J. Neurosci. 21 (12), 4183–4187.

Page 9

120 R. Selvatici et al. / Neurochemistry International 63 (2013) 112–120

Reddy, P.H., 2008. Mitochondrial medicine for aging and neurodegenerativediseases. Neuromol. Med. 10, 291–315.

Sadleir, K.R., Vassar, R., 2012. Cdk5 protein inhibition and Ab42 increase BACE1protein level in primary neurons by a post-transcriptional mechanism:implications of CDK5 as a therapeutic target for Alzheimer disease. J. Biol.Chem. 287 (10), 7224–7235.

Schliebs, R., Arendt, T., 2011. The cholinergic system in aging and neuronaldegeneration. Behav. Brain Res. 221 (2), 555–563.

Selvatici, R., Falzarano, S., Franceschetti, L., Cavallini, S., Marino, S., Siniscalchi, A.,2006. Differential activation of protein kinase C isoforms following chemicalischemia in rat cerebral cortex slices. Neurochem. Int. 49 (8), 729–736.

Selvatici, R., Previati, M., Marino, S., Marani, L., Falzarano, S., Lanzoni, I., Siniscalchi,A., 2009. Sodium azide induced neuronal damage in vitro: evidence for non-apoptotic cell death. Neurochem. Res. 34 (5), 909–916.

Siniscalchi, A., Cavallini, S., Marino, S., Falzarano, S., Franceschetti, L., Selvatici, R.,2006. Effects of chemical ischemia on cerebral cortex slices: focus on mitogen-activated protein kinase cascade. Ann. N. Y. Acad. Sci. 1090, 445–454.

Sotty, F., Danik, M., Manseau, F., Laplante, F., Quirion, R., Williams, S., 2003.Glutamatergic, cholinergic and GABAergic neurons contribute to theseptohippocampal pathways and exhibit distinct electro-physiologicalproperties: novel implications for hippocampal rhythmicity. J. Physiol. 551,927–943.

Stockert, J.C., Blázquez-Castro, A., Cañete, M., Horobin, R.W., Villanueva, A., 2012.MTT assay for cell viability: intracellular localization of the formazan product isin lipid droplets. Acta Histochem. 114 (8), 785–796.

Swerdlow, R.H., 2012. Mitochondria and cell bioenergetics: increasingly recognizedcomponents and a possible etiologic cause of Alzheimer’s disease. Antioxid.Redox Signal. 16 (12), 1434–1455.

van Meerloo, J., Kaspers, G.J., Cloos, J., 2011. Cell sensitivity assays: the MTT assay.Methods Mol. Biol. 731, 237–245.

Vergun, O., Sobolevsky, A.I., Yelshansky, M.V., Keelan, J., Khodorov, B.I., Duchen,M.R., 2001. Exploration of the role of reactive oxygen species in glutamateneurotoxicity in rat hippocampal neurones in culture. J. Physiol. 531 (Pt 1),147–163.

Vingtdeux, V., Davies, P., Dickson, D.W., Marambaud, P., 2011. AMPK is abnormallyactivated in tangle- and pre-tangle-bearing neurons in Alzheimer’s disease andother tauopathies. Acta Neuropathol. 121 (3), 337–349.

Wang, Y., Qin, Z.H., 2010. Molecular and cellular mechanisms of excitotoxicneuronal death. Apoptosis 15, 1382–1402.

Wang, J.Z., Xia, Y.Y., Grundke-Iqbal, I., Iqbal, K., 2013. Abnormalhyperphosphorylation of Tau: sites, regulation, and molecular mechanism ofneurofibrillary degeneration. J. Alzheimers Dis. 33 (Suppl. 1), S123–S139.

Wen, Y., Yu, W.H., Maloney, B., Bailey, J., Ma, J., Marié, I., Maurin, T., Wang, L.,Figueroa, H., Herman, M., Krishnamurthy, P., Liu, L., Planel, E., Lau, L.F., Lahiri,D.K., Duff, K., 2008. Transcriptional regulation of beta-secretase by p25/cdk5leads to enhanced amyloidogenic processing. Neuron 57 (5), 680–690.

Zou, K., Gong, J.S., Yanagisawa, K., Michikawa, M., 2002. A novel function ofmonomeric amyloid beta-protein serving as an antioxidant molecule againstmetal-induced oxidative damage. J. Neurosci. 22 (12), 4833–4841.