A Mitochondrial Role of SV2a Protein in Aging and Alzheimer’s Disease: Studies with Levetiracetam

Abstract

Aberrant neuronal network activity associated with neuronal hyperexcitability seems to be an important cause of cognitive decline in aging and Alzheimer’s disease (AD). Out of many antiepileptics, only levetiracetam improved cognitive dysfunction in AD patients and AD animal models by reducing hyperexcitability. As impaired inhibitory interneuronal function, rather than overactive neurons, seems to be the underlying cause, improving impaired neuronal function rather than quieting overactive neurons might be relevant in explaining the lack of activity of the other antiepileptics. Interestingly, improvement of cognitive deficits by levetiracetam caused by small levels of soluble Aβ was accompanied by improvement of synaptic function and plasticity. As the negative effects of Aβ on synaptic plasticity strongly correlate with mitochondrial dysfunction, wehypothesized that the effect of levetiracetam on synaptic activity might be raised by an improved mitochondrial function. Accordingly, we investigated possible effects of levetiracetam on neuronal deficits associated with mitochondrial dysfunction linked to aging and AD. Levetiracetam improved several aspects of mitochondrial dysfunction including alterations of fission and fusion balance in a cell model for aging and early late-onset AD. We demonstrate for the first time, using immunohistochemistry and proteomics, that the synaptic vesicle protein 2A (SV2a), the molecular target of levetiracetam, is expressed in mitochondria. In addition, levetiracetam shows significant effect on the opening of the mitochondrial permeability transition pore. Importantly, the effects of levetiracetam were significantly abolished when SV2a was knockdown using siRNA. In conclusion, interfering with the SV2a protein at the mitochondrial level and thereby improving mitochondrial function might represent an additional therapeutic effect of levetiracetam to improve symptoms of late-onset AD.

Keywords

Aging Alzheimer’s disease amyloid-β levetiracetam mitochondrial dynamics mitochondrial function mitochondrial permeability transition synaptic vesicle protein 2A (SV2a)

INTRODUCTION

Mitochondrial dysfunction is one of the major underlying mechanism of brain aging, mild cognitive impairment (MCI), and late-onset Alzheimer’s disease (LOAD) [1 –6]. The velocity of the decline of mitochondrial function depends on individual genetic predisposition like APOE4 and environmental factors [7, 8] until mitochondrial energy generation falls below a critical threshold [1, 2]. Exceeding this threshold may lead to conditions where mitochondrial dysfunction gets further exaggerated by the combined effects of aging, free radical (ROS formation, and mildly elevated amyloid-β (Aβ) levels due to the stimulation of γ-secretase activity [9, 10]. Thus, mitochondrial dysfunction has been recognized as a major player within the interface between aging and AD [10 –12]. Improving mitochondrial dysfunction therefore has become an important strategy for the development of drugs to treat the earlier stages of cognitive decline [10 , 14]. Contradictory to this scenario of mitochondrial dysfunction are observations about elevated hippocampal and cortical activity in MCI patients and in a number of conditions that confer risk for AD (e.g., carriers of an APOE4 allele or pre-symptomatic carriers of familial AD mutations) [15]. This hyperexcitability mainly in hippocampal structures is associated with aberrant neuronal network activity and seems to be an important cause of enhanced seizure susceptibility and impaired cognition in elderly, cognitively impaired people, AD patients, but also in aged animals as well as transgenic AD-mouse models [16 –19]. The striking functional fluctuations of cognition observed in AD patients support the idea of network dysfunction and variations in activity as a contributing factor to the disease’s pathology since it is unlikely that loss of neurons accounts for reversible fluctuations [16].

As hyperexcitability might respond to drugs used to reduce neuronal overactivity, many antiepileptic drugs have been tested in transgenic AD mouse models overexpressing human Aβ and exhibiting elevated network activity (monitored via video-EEG recordings). Interestingly, only levetiracetam has been found to actively reduce spike frequency significantly, while many other antiepileptic drugs were not active (e.g., gabapentin, valproic acid) [20]. In line with these results, levetiracetam was able to reduce hippocampal overactivity and simultaneously improve memory performance in amnestic MCI patients [15]. Similarly, out of three antiepileptic drugs tested in AD patients with epileptic seizures, only levetiracetam showed positive effects on cognition, while all were effective in reducing seizures [21]. Accordingly, it has been suggested that cognitive dysfunction associated with neuronal hyperactivity might be related to dysfunctional inhibitory interneurons rather than to a general neuronal overactivity [22, 23]. This was confirmed by observations indicating hyperexcitability as a consequence of dysfunctional GABAergic interneurons [24], suggesting that increasing neuronal activity rather than dampening would be of therapeutic benefit.

As mentioned above, neuronal dysfunction is an important aspect of the whole spectrum of cognitive decline in the elderly and seems to be associated with mitochondrial dysfunction leading to reduced energy supply and enhanced ROS production. Thus, improving mitochondrial activity might be a major mechanism of the specific effects of levetiracetam on cognitive deficits. As a matter of fact, a few findings are in line with this assumption [25, 26]. Accordingly, we speculated that the beneficial effects of levetiracetam for this condition might be related to effects at the mitochondrial level. This assumption was further supported by our previous findings that the structurally very closely related metabolic enhancer, piracetam, shows pronounced neurotrophic and neuroprotective effects (enhanced neuritogenesis, reduced apoptosis [27 –29]) by ameliorating impaired mitochondrial function (enhanced mitochondrial membrane potential, elevated ATP production, improved mitochondrial dynamics [29, 30]) and by findings about improved mitochondrial functions by levetiracetam in mice following experimental seizures [31].

MATERIALS AND METHODS

Materials

Dulbecco’s Modified Eagle Medium, OptiMEM^® Reduced Serum Medium, hygromycin, penicillin, streptomycin, MEM Vitamin solution, MEM Non-Essential Amino Acids, sodium pyruvate, Mito Tracker CMXRos, Silencer^® Select Pre-designed siRNA (siRNA ID # s19182) and Lipofectamine^® RNAiMAX were purchased from Invitrogen, Karlsruhe, Germany. All chemicals were obtained from Sigma Aldrich, Hamburg, Germany unless otherwise stated. For ATP determination, the ViaLighttrademark Plus Cell Proliferation and Cytotoxicity BioAssay Kit from Lonza, Basel, Switzerland, was used. Antibodies were purchased from Millipore, Billerica, USA [anti-GAP43 (MAP347), anti-Glyceraldehyde-3-Phosphate Dehydrogenase (MAB374), and all secondaryantibodies].

Cell culture

PC12 cells were cultured in DMEM supplemented with 10% heat-inactivated fetal calf serum and 5% heat-inactivated horse serum, 60 units/ml penicillin, and 60 μg/ml streptomycin at 37°C in a humidified incubator containing 5% CO₂.

SH-SY5Y cells were stably transfected with DNA constructs containing the entire coding region of human AβPP (AβPP695) or the corresponding vector alone (pCEP4 vector) and were kindly donated by A. Eckert (Basel, Switzerland) [32]. Cells were cultured in Dulbecco’s Modified Eagle Medium supplemented with 10% heat-inactivated fetal calf serum, 0.3 mg/ml hygromycin, 60 units/ml penicillin, 60 μg/ml streptomycin, MEM Vitamin solution, MEM Non-Essential Amino Acids and 1 mM sodium pyruvate at 37°C in a humidified incubator containing 5% CO₂.

Animals

18-month-old female Naval Medical Research Institute (NMRI) mice used in this study were purchased from Charles River (Borchen, Germany). The latter were obtained at an age of 12 months and maintained at the Biocenter’s animal care facility until use. For western blot analysis, isolated hippocampus and frontal cortex were used. To obtain isolated mitochondria for proteomic analysis, whole brain homogenates were used. All animals were housed in plastic cages with water and food ad libitum and were maintained on a 12-h light/dark cycle. Animals were handled according to the German guidelines for animal care.

Neurite length

For determining neurite length, all cells were seeded on polylysine coated glass cover slips, treatment was started one day after seeding. PC12 cells were grown in 15% serum containing medium overnight. The next day, medium was changed to a medium containing 2% serum and nerve growth factor (NGF, 50 ng/ml) to induce differentiation. SH-SY5Y cells were seeded and incubated with normal 10% serum containing cell culture medium for 24 h. The next day media were changed to reduced cell culture medium containing 2% serum. All cells were fixed with phosphate buffered formalin solution (4%) for 20 min and stained with hematoxylin and eosin solutions. 30 cells from each stain were arbitrarily investigated and neurite length was detected by using Nikon NIS Elements AR 2.1 software.

Western blot

Cells were washed with phosphate buffered saline (PBS) and lysed in lysis buffer containing 1 mM ethylenediaminetetraacetic acid, 5 mM sodium fluoride, 0.5% Triton^TM X-100, 6 M urea, 0.5% sodium deoxycholate, 0.5% sodium dodecyl sulfate, 2.5 mM sodium pyrophosphate, 1 mM sodium orthovanadate, 3 μg/ml aprotinin, 0.01 mg/ml leupeptin, 0.01 mg/ml pepstatin, 0.1 mM phenylmethanesulfonyl fluoride (PMSF) in PBS. After protein determination (bicinchoninic acid method) 10 μg of protein per lane were loaded on a 4–12% Bis-Tris gel and separated by electrophoresis. Samples were transferred onto a polyvinylidene fluoride membrane and incubated for 1 h with blocking solution, washed three times with Tris buffered saline with Tween^® solution (TBST) and incubated with primary antibodies overnight at 4°C. After washing five times with TBST, membranes were treated with horseradish peroxidase conjugated secondary antibodies, washed five times with TBST and analyzed using a ChemiDoc XRS system (Bio Rad, Munich, Germany).

ATP levels

ATP levels were determined using a bioluminescence assay based on a luciferase reaction which catalyses the formation of light from ATP and luciferin. 2 × 10⁴ cells were seeded in a white walled 96 well plate and treated according to the manufacturer’s instructions. The emitted light is linear to the ATP concentration and was measured with a VICTORtrademark X2 Multilabel Plate Reader (Perkin Elmer).

Confocal laserscan microscopy

SH-SY5Y cells were seeded on polylysine coated glass cover slips for 4 days. Mitochondria were visualized by labeling with MitoTracker^® Deep Red FM (100 nM) for 2 h. Cells were fixed with phosphate buffered formalin solution (4% , 37°C, pH 7.4) for 20 min and washed three times with PBS. The samples were analyzed using a Leica TCS SP5 confocal laserscan microscope with a 63 x oil immersion objective and Image J 1.47t (National Institute of Health, USA). Mitochondrial shape was quantified by separating four groups: punctuated (0–2 μm), truncated (2–4 μm), tubular (4–10 μm), and elongated (>10 μm length); n = 100 mitochondria.

Proteomic analysis of isolated mitochondria

Protein content in mitochondrial preparations was estimated via the BCA method. Preparations from three mice were pooled at equal concentrations to give a final protein amount of 120 μg per sample. Proteins were precipitated with methanol/chloroform according to the method described by Wessel and Flügge et al. [33]. Afterwards, proteins were solubilized in 25 mM triethylammonium bicarbonate (TEAB) buffer. Disulfide bonds were reduced with dithiothreitol and free cysteines subsequently alkylated with iodoacetamide. Digestion using trypsin (enzyme to substrate ratio = 1:50) was performed at 37°C for 16 h. In order to enhance digestion efficiency of membrane-associated proteins, a slightly modified membrane digestion protocol was applied to remaining undigested fractions [34]. Briefly, digestion solutions were centrifuged at 15,000 × g and the supernatant was pipetted off and kept for further analysis. The resulting pellet was redissolved in 60% methanol and proteins from mitochondrial membranes were extracted and solubilized via alternating vortexing and ultrasonication steps. After further reduction and alkylation, proteins were digested using an additional aliquot of trypsin at 37°C for 16 h. The resulting peptide solution was combined with the supernatant from the first digestion. Peptides were labeled with TMTsixplex reagent for further quantification experiments. Subsequently, the solution was acidified and purified using Pierce C18 Spin Columns (Thermo Scientific).

Peptide separation was performed using 1 μg of material on an Ultimate 3000 nLC (Thermo Scientific) equipped with an Acclaim PepMap100 μCartridge precolumn (C18, 300 μm × 0.5 cm, 5 μm, 100 Å) and an Acclaim PepMap analytical column (C18, 75 μm × 50 cm, 2 μm, 100 Å) at a flow rate of 250 nl/min. A gradient with increasing concentration of acetonitrile over 145 min was applied. Mass spectrometric analysis was performed using a Q Exactive HF^TM mass spectrometer (Thermo Scientific) in data-dependent acquisition mode with a maximum of 15 precursors per full scan chosen for fragmentation. Mascot generic format (MGF) files were created from acquired MS/MS spectra by the mass spectrometer’s software and searched against all murine proteins listed in the Swissprot database (downloaded in July 2014) using an in-house Mascot server V.2.04 (MatrixScience Ltd.) [35] and PEAKS proteomics software V.7 (Bioinformatics Solutions Inc.). Data from two replicate runs were merged and compared to improve confidence in protein identification. Peptide false discovery rate as estimated by searching against a decoy database was adjusted to 1.0% . Only proteins identifications with a minimum of two unique peptides were considered significant. Protein identifications were extracted in table format and are available in Supplementary Table 1.

SV2a-kockdown

For siRNA experiments, 4 × 10⁵ cells per well were seeded in a six well plate and incubated overnight. The next day transfection started using a Silencer^® Select Pre-designed siRNA and Lipofectamine^® RNAiMAX according to the manufacturer’s instructions. In detail 250 μl Optimem/well were mixed with 7.5 μl Lipofectamin/well. In parallel siRNA and scrambled siRNA were diluted with Optimem leading to a final siRNA con-centration of 10 nM in each well. Diluted siRNA was mixed with diluted Lipofectamin and incubated for 10 min at room temperature. Afterwards the siRNA-lipid complex was added to the cells (500 μl per well) and incubated for further 24 h.

Mitochondrial swelling

Mitochondrial swelling in brain mitochondria was evaluated by measurement of spectrophotometric alterations in light scattering according to Hansson et al. with slight modifications [36]. Isolation of mitochondria was achieved using a Percoll gradient according to Sims & Anderson and Hansson et al. with slight modifications [36 –38]. Preparation was carried out in ice-cold solutions. In brief, NMRI mice were decapitated, brains were quickly dissected on ice, and after removing the cerebellum, brains were washed with isolation buffer (320 mM sucrose, 2 mM EGTA, 10 mM Trizma base, pH 7.4). Brains were homogenized in isolation buffer containing 12% Percoll using a Tissue Grinder Dounce (Wheaton, Millville, USA) by ten loose and ten tight strokes. Afterwards the homogenates were slowly layered directly onto previously prepared discontinuous Percoll gradients (26% Percoll layered above 40% Percoll) and centrifuged in a Beckmann J2-HS, rotor J20.1 (30,700 × g, 7 min, 4°C). The mitochondrial fraction was removed (band 3, for details see [38]), diluted with isolation buffer and centrifuged (16,700 × g, 12 min, 4°C). The resulting pellet was washed twice with isolation buffer (7,300 × g, 6 min, 4°C) and total protein content was estimated (Bradford method). Mitochondria were diluted with isolation buffer to a total protein content of 2.75 mg/ml. Mitochondrial permeability transition was monitored by measuring the decrease in 90° light scattering at 520 nm (emission and excitation) using a Aminco Bowman Series 2 Spectrometer (SLM-Aminco, Rochester, USA) over 750 s. Mitochondria (27.5 μg) were incubated in a stirred glass cuvette containing 1.1 ml measuring buffer (250 mM sucrose, 10 mM Trizma base, 20 mM MOPS, 2 mM KH2PO4, 1 mM MgCl₂, 1 μM EGTA, pH 7.2) with glutamate (5 mM) and malate (5 mM) for 3 min. Afterwards oligomycin (1 μg/ml) was added and the measurement was started. After 60 s, either the inhibitor of mPTP formation cyclosporine A (1 μM) or ethanol (absolute) as solvent control was added. After a further 60 s, ADP (20 μM) was added. Swelling was induced by the addition of calcium (1 μmol/mg protein) or atractyloside (400 μM) after 300 s. Alamethicin (16 μg/ml) was added after 500 s to induce maximal mitochondrial swelling. The absorbance before calcium injection was set as 0% , the absorbance after alamethicin injection as 100% . The quality of mitochondrial purification was verified by measuring the increase in mitochondrial respiration after adding cytochrome c (10 μM) and in parallel determining the respiratory control ratio (all mitochondrial preparations with an RCR less than 4 were discarded).

Statistical analysis

Data are given as mean ± S.E.M. For statistical comparison, Student’s t-test and two-way ANOVA followed by Bonferroni post-test for multiple comparisons were used. p values <0.05 were considered statistically significant. “n” stands for the number of independent experiments carried out.

RESULTS

Levetiracetam improves synaptic plasticity and interacts with mitochondria

For assessing possible effects of levetiracetam on mechanisms of synaptic plasticity we initially used differentiating PC12 cells as an established cell model to track changes in synaptic plasticity [28]. Since therapeutic plasma levels of levetiracetam in animals and men range between 23 μM and 176 μM [20 , 40] we chose 20 μM and 200 μM levetiracetam for our in vitro cell culture experiments to observe concentration dependent as well as maximal effects. In line with already described neurotrophic properties of levetiracetam, treating differentiating PC12 cells with levetiracetam elevates neurite outgrowth and significantly increases the expression of GAP43, a synaptic marker protein (Fig. 1). Under oxidative stress conditions (treatment with 50 μM sodium nitroprusside (SNP)) accompanied by mitochondrial dysfunction [27], neurite outgrowth was reduced. Levetiracetam ameliorated this impairment significantly and also elevated GAP43 levels drug (Fig. 1). To strengthen these observations, the effects of levetiracetam onneuritic outgrowth were also investigated in the human neuronal cell line SH-SY5Y cells. Again, levetiracetam also elevated neurite length in SY5Y-Ctl cells (Fig. 2A, B). Importantly, levetiracetam not only improved synaptic plasticity but also elevated ATP levels, as a typical indicator for improved mitochondrial function, suggesting additional effects of levetiracetam not depending on synaptic plasticity (Fig. 2C).

Mitochondrial dynamics, meaning the ability of mitochondria to undergo changes in size and form, are gaining more and more attention as an important factor regulating mitochondrial function [41] and mechanism of mitochondrial quality control [42]. Even if reports are sometimes controversial, in most cases mitochondrial fragmentation is accompanied by reduced mitochondrial function and vice versa [43]. Accordingly, shorter mitochondria are energetically unfavorable. We have previously used confocal microscopy of fixed mitochondria as a very reliable method to analyze mitochondrial dynamics and therefore used it to further characterize the effects of levetiracetam on mitochondrial function. Incubating cells with levetiracetam shifts mitochondrial shape toward elongated forms (>10 μm) and reduces the number of punctuated (<2 μm) mitochondria (Fig. 2A,D), which is in parallel with the positive effect of levetiracetam on other aspects of mitochondrialfunction.

Beneficial effects of levetiracetam in a cell model of brain aging

Due to the fact that brain aging is mainly associated with complex I impairment, treating cells with specific inhibitors of this respiratory chain complex (e.g., rotenone) can be used to mirror aging in vitro. This treatment leads to impaired neurite outgrowth and mitochondrial fragmentation [6]. In line with improved cognitive function in aged rats, levetiracetam also reversed rotenone-induced changes in SY5Y-Ctl cells: neurites from cells treated with the drug and rotenone are even longer than neurites from untreated cells (Fig. 3A, B). Additionally mitochondrial shape change induced by complex I inhibition is clearly shifted back toward control mitochondria (Fig. 3A, C). Taken together, levetiracetam seems to reverse age-associated changes in vitro, displayed by enhanced neurite outgrowth and mitochondrialelongation.

Levetiracetam’s activity in a cellular model of LOAD

Lately we were able to demonstrate that SH-SY5Y cells carrying an additional copy of the human AβPP695 gene (SY5Y-AβPP_wt) are an appropriate cell model for the initial phase of LOAD. These cells have only slightly elevated Aβ_1 - 40 and Aβ_1 - 42 levels, exhibit impaired mitochondrial respiration, decreased ATP levels, and in parallel show distinct changes in mitochondrial dynamics compared to the corresponding vector control cells (SY5Y-Ctl) [6]. In line with our other results, levetiracetam also enhances neurite outgrowth in SY5Y-AβPP_wt cells (Fig. 4A, B), elevates ATP levels to some extent, and promotes fusion even already at the lower concentrations (20 μM) (Fig. 4A, C, D). Similar observations about the beneficial effects of levetiracetam were obtained using a combination of Aβ load and additional complex I dysfunction (Supplementary Fig. 1) as a model for the later phase of LOAD [6].

SV2a, also a mitochondrial protein

Taken together, even if all these findings are suggestive for an interaction between levetiracetam and mitochondria, the mechanism of action remains obscure, especially since the synaptic vesicle protein 2a (SV2a) is generally accepted as the specific target of the drug [44]. Keeping in mind that most of our findings argue for a fairly direct interaction of levetiracetam with mitochondria, we asked if SV2a might also be expressed in mitochondria. Therefore we isolated mitochondria from NMRI mice using a Percoll density gradient centrifugation protocol developed by Sims and Anderson [38].

Remarkably, western blot analysis reveals that isolated brain mitochondria show similar SV2a protein levels as samples from homogenized mouse hippocampus and frontal cortex (Fig. 5A). To demonstrate that our mitochondrial preparation does not contain other cellular fractions we used antibodies against the respiratory chain complexes I and III to verify that mitochondrial proteins are enriched in mitochondrial fraction and a tubulin antibody to demonstrate the separation of cytosolic proteins (Fig. 5A). These results are in line with works from Noyer et al. who studied the subcellular distribution of specific [³H] levetiracetam binding by sucrose gradient centrifugation. They found that the majority of the drug shows up in the synaptic plasma membrane fraction but that the drug also appears in the mitochondrial membrane fraction (approximately 15–20%) [45]. To further strengthen these results, mitochondrial preparations were investigated using proteomic analysis, leading to the identification of over 11,000 unique peptide sequences belonging to more than 1,400 proteins (Supplementary Table 1). SV2A protein (Hit No. 391 for Mascot database search, Score = 340) was identified with eight unique peptides (Fig. 5B), even though identification of such a low-abundant protein is always difficult due to suppression by very abundant proteins such as sodium/potassium-transporting ATPases or glycerol-3-phosphate dehydrogenase. Gene ontology analysis using information available at http://www.uniprot.org revealed that isolation of mitochondria had been successful, as most identified proteins were marked as mitochondrial proteins. By proteomic analysis of mouse cortex and hippocampus samples where mitochondria have not been isolated, SV2a protein was also only identified with nine unique peptides, suggesting that its occurrence in the mitochondrial fraction is not caused by contamination or insufficient fractionation (data not shown).

SV2a knockdown impairs mitochondrial dynamics and abolishes the effects of levetiracetam

To prove that SV2a is not only present at the mitochondrial level but also has a functional impact, we downregulated SV2a using siRNA (SY5Y-KD cells). The results clearly show a substantial reduction of SV2a protein levels in SY5Y-Ctl as well as in SY5Y-AβPP_wt cells (Fig. 6A–C). Levetiracetam (24 h incubation) had no effect on protein levels for all conditions. At the same time SV2a knockdown alone had a significant effect on mitochondrial dynamics by shifting the equilibrium toward fission (more punctuated and less tubular mitochondria) (Fig. 6D, F). In SY5Y-KD cells, the effect of levetiracetam on mitochondrial dynamics is no longer present (Fig. 6E, F), in line with the assumption of SV2a also being the target at the mitochondrial level. However, we cannot completely role out the possibility that the SV2a-knockdown perturbs mitochondrial dynamics to a point where a rescue is no longer possible. Nevertheless, consequently the question arises as to what SV2a does in mitochondrial membranes and how does binding of levetiracetam influence its function there.

Levetiracetam reduces mitochondrial swelling

Very little is known about the modulation of mitochondrial dynamics by drugs [46] but in some cases inhibiting mitochondrial permeability transition pore (mPTP) opening might be relevant [47 –49]. In order to examine whether this mechanism of action is also a conceivable explanation for our observations obtained with levetiracetam, we measured mitochondrial swelling of isolated brain mitochondria by tracking spectrophotometric alterations in light scattering. This method is very useful to analyze mitochondrial permeability transition induced by typical inductors like calcium or atractyloside. Cyclosporin A (CsA) is a specific inhibitor of mPTP opening and serves as a positive control. In these experiments, isolated brain mitochondria were incubated ex vivo with 20 μM levetiracetam, swelling was induced by calcium chloride (1 μmol/mg protein) or atractyloside (400 μM), and alamethicin was used to detect maximal swelling. Confirming a direct interaction between levetiracetam and mitochondria, maybe especially between levetiracetam and the mPTP, incubated mitochondria show a significantly reduced mitochondrial swelling compared to control mitochondria (Fig. 7). This inhibitory effect is stronger when mPTP opening is induced by calcium chloride and in neither case is as strong as after incubation with CsA (Fig. 7).

DISCUSSION

Aberrant neuronal network activity associated with neuronal hyperexcitability seems to be an important cause of cognitive decline in aging and AD [16]. Out of many antiepileptic drugs, only levetiracetam was able to improve cognitive dysfunctions in patients and animal models by reducing hyperexcitability [15 , 50]. As impaired inhibitory interneuronal function seems to be the underlying cause for aberrant network activity rather than overactive neurons [15, 50], improving impaired neuronal function might be relevant. This would explain the lack of activity of most antiepileptic drugs investigated [20]. As thenegative effects of Aβ on synaptic plasticity are strongly correlated with mitochondrial dysfunction leading to impaired energy supply and enhanced free radical production [11 , 51], we hypothesized that the effect of levetiracetam on synaptic activity might also be caused by an improvement of mitochondrial function. Accordingly, we investigated possible effects of levetiracetam on synaptic deficits and plasticity associated with mitochondrial deficits typical for aging and AD. In agreement with former assumptions, our data indicate that levetiracetam improves neuritogenesis and enhances the synthesis of the synaptic marker protein GAP-43 in PC12 cells under baseline conditions but also after mitochondrial impairment induced by SNP.

We have previously shown that SH-SY5Y cells can be used to investigate synaptic dysfunction associated with impaired neuritogenesis over a broad range of the aging spectrum, from the normal situation with very low Aβ levels, to cells with impaired complex I function as a model for the aging process. Furthermore SH-SY5Y cells bearing an additional copy of the human amyloid-β precursor protein (AβPP) gene and therefore producing slightly more Aβ can be used as a model for people at risk for LOAD. Finally initiating complex I dysfunction (as a model of artificial aging) in these AβPP transgenic cells allows us to mirror the very early phase of the initiation of AD in vitro [6]. In line with data from PC12 cells where levetiracetam improved synaptic plasticity following oxidative stress, levetiracetam also enhanced neuritogenesis in SH-SY5Y cells already under baseline conditions or following complex I inhibition as well as under conditions mirroring LOAD situation.

We and others have previously shown that mitochondrial dysfunction and/or impaired synaptic plasticity and neuritogenesis was paralleled by profound changes of mitochondrial dynamics, shifting the equilibrium between mitochondrial fission and fusion to the fission site [30 , 53]. This was confirmed for the SH-SY5Y cells also used in this study for all different conditions outlined above by exhibiting a large percentage of small punctuated mitochondria after complex I inhibition or AβPP expression compared to control cells [6]. Levetiracetam, in parallel to its effects on improving synaptic plasticity, also improved mitochondrial dynamics by shifting the fission and fusion balance back to the fusion site with normalized numbers of long (fused) mitochondria. These findings together with the elevation of ATP levels not only support the beneficial effect of this antiepileptic drug on mitochondrial function but also suggest a direct (maybe additional) effect at the mitochondrial level.

The pharmacology of levetiracetam has been linked to a unique and brain specific binding site [45], which later has been identified as the synaptic vesicle protein SV2a [44]. However, the downstream consequences are not yet fully understood, but it appears that levetiracetam modulates exocytosis of transmitter-containing vesicles, probably by interfering with the fusion of synaptic vesicles with the neuronal membrane [54]. Other data from Budzinski et al. suggest that the SV2a is required for structural adaptive changes of synaptic vesicles upon loading with neurotransmitters [55]. While all these findings support a significant role of SV2a in vesicle function, only few previous findings may also suggest a role of SV2a in synaptic plasticity. Using a microRNA technique Cohen et al. finally identified SV2a as involved in several mechanisms of synaptic plasticity (spine density, PDS-95 clustering, GluR2 expression) [56]. This is in line with the findings of Sanchez et al. that the SV2a ligand levetiracetam reverses synaptic deficits associated with cognitive dysfunction in a mouse model of AD [20].

While these previous findings agree with our observations about a pronounced effect of levetiracetam on neuritogenesis as an important aspect of synaptic plasticity, our findings of the effects of levetiracetam on mitochondrial function and fusion are exciting and new because the SV2a protein was only described to be expressed in brain synaptic vesicles [44]. However, already the first publication about a brain specific high-affinity binding site of levetiracetam reported a low specific levetiracetam binding in the mitochondrial fraction [45]. These findings were not only confirmed but were substantially expanded by our experiments showing high levels of SV2a immunoreactivity at the mitochondrial level. Proteomic analysis of mitochondrial fraction was used to confirm these findings. A direct mitochondrial function of SV2a is also supported by enhanced fission in SV2a-KD cells and by the loss of the effects of levetiracetam on mitochondrial dynamics in these cells. Thus, our findings demonstrate for the first time a direct mitochondrial role of this protein. Even if it is possible that the mitochondrial effects are a downstream consequence of levetiracetam simply affecting general cell health and/or function via other yet unknown mechanisms, further findings support the idea of a relatively direct effect on mitochondria. In a large genome-wide research study about possible gene candidates associated with APOE4 and the risk for AD, SV2a was one of the few significant hits [57]. As APOE4 among other effects also regulates synaptic plasticity and mitochondrial function [8 , 59] and leads to impairment of GABAergic interneurons associated with cognitive deficits [60], interfering with both mechanisms might be a common final pathway of both proteins. In addition, levetiracetam significantly reduces mPTP opening of isolated mitochondria. These findings also argue for a mitochondrial function of SV2a.

Findings that SV2a seems to induce structural changes of vesicular membranes [55] and keeping in mind that mitochondrial and vesicular membranes are in some way related to each other (lipid bilayers and electrochemical gradient) make the hypothesis very attractive that SV2a might be a fission or fusion factor. Binding of levetiracetam to this protein thereby might shift the sensitive equilibrium of mitochondrial dynamics back to the probably energetically favorable, longer forms of mitochondria. Such a shift back from fragmented mitochondria like in AD toward elongated forms might explain elevated ATP levels as well as a reduced permeability transition [48]. Another possible explanation for the observed alterations of mPTP opening is an alteration of the assembly of the different pore components located at the inner and outer mitochondrial membrane [61]. Then, effects on mitochondrial dynamics would be secondary and due to a decreased mPTP opening.

In conclusion our findings suggest for the first time not only the presence but also a specific function of SV2a at the mitochondrial level and that beneficial effects of levetiracetam on cognition might also be in part the result of the interaction of the drug with mitochondrial SV2a protein.

DISCLOSURE STATEMENT

Authors’ disclosures available online (http://j-alz.com/manuscript-disclosures/15-0687r1).

Footnotes

Appendix

References

Swerdlow

(2012) Mitochondria and cell bioenergetics: Increasingly recognized components and a possible etiologic cause of Alzheimer’s disease. Antioxid Redox Signal 16, 1434–1455.

Swerdlow

(2011) Brain aging, Alzheimer’s disease, and mitochondria. Biochim Biophys Acta 1812, 1630–1639.

Yankner

, Lu

, Loerch

(2008) The aging brain. Annu Rev Pathol 3, 41–66.

Swerdlow

, Burns

, Khan

(2014) The Alzheimer’s disease mitochondrial cascade hypothesis: Progress and perspectives. Biochim Biophys Acta 1842, 1219–1231.

Wirz

KTS

, Keitel

, Swaab

, Verhaagen

, Bossers

(2014) Early molecular changes in Alzheimer disease: Can we catch the disease in its presymptomatic phase? J Alzheimers Dis 38, 719–740.

Stockburger

, Gold

VAM

, Pallas

, Kolesova

, Miano

, Leuner

, Müller

(2014) A cell model for the initial phase of sporadic Alzheimer’s disease. J Alzheimers Dis 42, 395–411.

Huang

(2010) Abeta-independent roles of apolipoprotein E4 in the pathogenesis of Alzheimer’s disease. Trends Mol Med 16, 287–294.

Valla

, Yaari

, Wolf

, Kusne

, Beach

, Roher

, Corneveaux

, Huentelman

, Caselli

, Reiman

(2010) Reduced posterior cingulate mitochondrial activity in expired young adult carriers of the APOE ɛ4 allele, the major late-onset Alzheimer’s susceptibility gene. J Alzheimers Dis 22, 307–313.

Leuner

, Schütt

, Kurz

, Eckert

, Schiller

, Occhipinti

, Mai

, Jendrach

, Eckert

, Kruse

, Palmiter

, Brandt

, Dröse

, Wittig

, Willem

, Haass

, Reichert

, Müller

(2012) Mitochondrion-derived reactive oxygen species lead to enhanced amyloid beta formation. Antioxid Redox Signal 16, 1421–1433.

10.

Müller

, Eckert

, Kurz

, Eckert

, Leuner

(2010) Mitochondrial dysfunction: Common final pathway in brain aging and Alzheimer’s disease–therapeutic aspects. Mol Neurobiol 41, 159–171.

11.

Reddy

, Beal

(2008) Amyloid beta, mitochondrial dysfunction and synaptic damage: Implications for cognitive decline in aging and Alzheimer’s disease. Trends Mol Med 14, 45–53.

12.

Reddy

, Manczak

, Mao

, Calkins

, Reddy

, Shirendeb

(2010) Amyloid-beta and mitochondria in aging and Alzheimer’s disease: Implications for synaptic damage and cognitive decline. J Alzheimers Dis 20(Suppl 2), S499–S512.

13.

Leuner

, Hauptmann

, Abdel-Kader

, Scherping

, Keil

, Strosznajder

, Eckert

, Muller

(2007) Mitochondrial dysfunction: The first domino in brain aging and Alzheimer’s disease? Antioxid. Redox Signal 9, 1659–1675.

14.

Eckert

, Renner

, Eckert

, Eckmann

, Hagl

, Abdel-Kader

, Kurz

, Leuner

, Muller

(2012) Mitochondrial dysfunction–a pharmacological target in Alzheimer’s disease. Mol Neurobiol 46, 136–150.

15.

Bakker

, Krauss

, Albert

, Speck

, Jones

, Stark

, Yassa

, Bassett

, Shelton

, Gallagher

(2012) Reduction of hippocampal hyperactivity improves cognition in amnestic mild cognitive impairment. Neuron 74, 467–474.

16.

Palop

, Chin

, Mucke

(2006) A network dysfunction perspective on neurodegenerative diseases. Nature 443, 768–773.

17.

Palop

, Chin

, Roberson

, Wang

, Thwin

, Bien-Ly

, Yoo

, Ho

, Yu

G-Q

, Kreitzer

, Finkbeiner

, Noebels

, Mucke

(2007) Aberrant excitatory neuronal activity and compensatory remodeling of inhibitory hippocampal circuits in mouse models of Alzheimer’s disease. Neuron 55, 697–711.

18.

Busche

, Chen

, Henning

, Reichwald

, Staufenbiel

, Sakmann

, Konnerth

(2012) Critical role of soluble amyloid-β for early hippocampal hyperactivity in a mouse model of Alzheimer’s disease. Proc Natl Acad Sci U S A 109, 8740–8745.

19.

Baxter

(2012) Quieting the overactive hippocampus restores memory in aging. Trends Cogn Sci 16, 360–361.

20.

Sanchez

, Zhu

, Verret

, Vossel

, Orr

, Cirrito

, Devidze

, Ho

, Yu

G-Q

, Palop

, Mucke

(2012) Levetiracetam suppresses neuronal network dysfunction and reverses synaptic and cognitive deficits in an Alzheimer’s disease model. Proc Natl Acad Sci U S A 109, E2895–E2903.

21.

Cumbo

, Ligori

(2010) Levetiracetam, lamotrigine, and phenobarbital in patients with epileptic seizures and Alzheimer’s disease. Epilepsy Behav 17, 461–466.

22.

Busche

, Eichhoff

, Adelsberger

, Abramowski

, Wiederhold

K-H

, Haass

, Staufenbiel

, Konnerth

, Garaschuk

(2008) Clusters of hyperactive neurons near amyloid plaques in a mouse model of Alzheimer’s disease. Science 321, 1686–1689.

23.

Palop

, Mucke

(2010) Amyloid-beta-induced neuronal dysfunction in Alzheimer’s disease: From synapses toward neural networks. Nat Neurosci 13, 812–818.

24.

Verret

, Mann

, Hang

, Barth

AMI

, Cobos

, Ho

, Devidze

, Masliah

, Kreitzer

, Mody

, Mucke

, Palop

(2012) Inhibitory interneuron deficit links altered network activity and cognitive dysfunction in Alzheimer model. Cell 149, 708–721.

25.

Zilberter

, Ivanov

, Ziyatdinova

, Mukhtarov

, Malkov

, Alpár

, Tortoriello

, Botting

, Fülöp

, Osypov

, Pitkänen

, Tanila

, Harkany

, Zilberter

(2013) Dietary energy substrates reverse early neuronal hyperactivity in a mouse model of Alzheimer’s disease. J Neurochem 125, 157–171.

26.

Janc

, Müller

(2014) The free radical scavenger Trolox dampens neuronal hyperexcitability, reinstates synaptic plasticity, and improves hypoxia tolerance in a mouse model of Rett syndrome. Front Cell Neurosci 8, 56.

27.

Keil

, Scherping

, Hauptmann

, Schuessel

, Eckert

, Müller

(2006) Piracetam improves mitochondrialdysfunction following oxidative stress. Br J Pharmacol 147, 199–208.

28.

Kurz

, Ungerer

, Lipka

, Kirr

, Schutt

, Eckert

, Leuner

, Muller

(2010) The metabolic enhancer piracetam ameliorates the impairment of mitochondrial function and neurite outgrowth induced by beta-amyloid peptide. Br J Pharmacol 160, 246–257.

29.

Leuner

, Kurz

, Guidetti

, Orgogozo

J-M

, Müller

(2010) Improved mitochondrial function in brain aging and Alzheimer disease - the new mechanism of action of the old metabolic enhancer piracetam.pii. Front Neurosci 4, 44.

30.

Stockburger

, Kurz

, Koch

, Eckert

, Leuner

, Müller

(2013) Improvement of mitochondrial function and dynamics by the metabolic enhancer piracetam. Biochem Soc Trans 41, 1331–1334.

31.

Gibbs

, Walker

, Cock

(2006) Levetiracetam: Antiepileptic properties and protective effects on mitochondrial dysfunction in experimental status epilepticus. Epilepsia 47, 469–478.

32.

Rhein

, Baysang

, Rao

, Meier

, Bonert

, Müller-Spahn

, Eckert

(2009) Amyloid-beta leads to impaired cellular respiration, energy production and mitochondrial electron chain complex activities in human neuroblastoma cells. Cell Mol Neurobiol 29, 1063–1071.

33.

Wessel

, Flügge

(1984) A method for the quantitative recovery of protein in dilute solution in the presence of detergents and lipids. Anal Biochem 138, 141–143.

34.

Blonder

, Veenstra

(2007) Computational prediction of proteotypic peptides. Expert Rev Proteomics 4, 351–354.

35.

Perkins

, Pappin

, Creasy

, Cottrell

(1999) Probability-based protein identification by searching sequence databases using mass spectrometry data. Electrophoresis 20, 3551–3567.

36.

Hansson

, Månsson

, Mattiasson

, Ohlsson

, Karlsson

, Keep

, Elmér

(2004) Brain-derived respiring mitochondria exhibit homogeneous, complete and cyclosporin-sensitive permeability transition. J Neurochem 89, 715–729.

37.

Sims

(1990) Rapid isolation of metabolically active mitochondria from rat brain and subregions using Percoll density gradient centrifugation. J Neurochem 55, 698–707.

38.

Sims

, Anderson

(2008) Isolation of mitochondria from rat brain using Percoll density gradient centrifugation. Nat Protoc 3, 1228–1239.

39.

Rossetti

, Bromfield

(2006) Determinants of success in the use of oral levetiracetam in status epilepticus. Epilepsy Behav 8, 651–654.

40.

Mink

, Muroi

, Seule

, Bjeljac

, Keller

(2011) Levetiracetam compared to valproic acid: Plasma concentration levels, adverse effects and interactions in aneurysmal subarachnoid hemorrhage. Clin Neurol Neurosurg 113, 644–648.

41.

Chen

, Chan

(2006) Critical dependence of neurons on mitochondrial dynamics. Curr Opin Cell Biol 18, 453–459.

42.

Patel

, Shirihai

, Huang

(2013) Optimal dynamics for quality control in spatially distributed mitochondrial networks. PLoS Comput Biol 9, e100310.

43.

Galloway

, Lee

, Yoon

(2012) Mitochondrial morphology-emerging role in bioenergetics. Free Radic Biol Med 53, 2218–2228.

44.

Lynch

, Lambeng

, Nocka

, Kensel-Hammes

, Bajjalieh

, Matagne

, Fuks

(2004) The synaptic vesicle protein SV2A is the binding site for the antiepileptic drug levetiracetam. Proc Natl Acad Sci U S A 101, 9861–9866.

45.

Noyer

, Gillard

, Matagne

, Hénichart

, Wülfert

(1995) The novel antiepileptic drug levetiracetam (ucb L059) appears to act via a specific binding site in CNS membranes. Eur J Pharmacol 286, 137–146.

46.

Andreux

, Houtkooper

, Auwerx

(2013) Pharmacological approaches to restore mitochondrial function. Nat Rev Drug Discov 12, 465–483.

47.

Naga

, Geddes

(2011) Dimebon inhibits calcium-induced swelling of rat brain mitochondria but does not alter calcium retention or cytochrome C release. Neuromolecular Med 13, 31–36.

48.

Picard

, Shirihai

, Gentil

, Burelle

(2013) Mitochondrial morphology transitions and functions: Implications for retrograde signaling? Am J Physiol Regul Integr Comp Physiol 304, R393–R406.

49.

Eckmann

, Eckert

, Leuner

, Muller

, Eckert

(2013) Mitochondria: Mitochondrial membranes in brain ageing and neurodegeneration. Int J Biochem Cell Biol 45, 76–80.

50.

Koh

, Haberman

, Foti

, McCown

, Gallagher

(2010) Treatment strategies targeting excess hippocampal activity benefit aged rats with cognitive impairment. Neuropsychopharmacology 35, 1016–1025.

51.

, Guo

, Yan

(2012) Synaptic mitochondrial pathology in Alzheimer’s disease. Antioxid Redox Signal 16, 1467–1475.

52.

Reddy

(2009) Amyloid beta, mitochondrial structural and functional dynamics in Alzheimer’s disease. Exp Neurol 218, 286–292.

53.

Eckert

, Eckmann

, Renner

, Eckert

, Leuner

, Muller

(2012) Dimebon ameliorates amyloid-beta induced impairments of mitochondrial form and function. J Alzheimers Dis 31, 21–32.

54.

Kaminski

, Gillard

, Klitgaard

(2012) Targeting SV2A for Discovery of Antiepileptic Drugs. In: Jasper’s Basic Mechanisms of the Epilepsies [Internet], 4th edition, Noebels

, Avoli

, Rogawski

, Olsen

, Delgado-Escueta

, eds. National Center for Biotechnology Information (US), Bethesda, MD.

55.

Budzinski

, Allen

, Fujimoto

, Kensel-Hammes

, Belnap

, Bajjalieh

, Chiu

(2009) Large structural change in isolated synaptic vesicles upon loading with neurotransmitter. Biophys J 97, 2577–2584.

56.

Cohen

, Lee

, Chen

, Li

, Fields

(2011) MicroRNA regulation of homeostatic synaptic plasticity. Proc Natl Acad Sci U S A 108, 11650–11655.

57.

Rhinn

, Fujita

, Qiang

, Cheng

, Lee

, Abeliovich

(2013) Integrative genomics identifies APOE ɛ4 effectors in Alzheimer’s disease. Nature 500, 45–50.

58.

Kim

, Yoon

, Basak

, Kim

(2014) Apolipoprotein E in synaptic plasticity and Alzheimer’s disease: Potential cellular and molecular mechanisms. Mol Cells 37, 767–776.

59.

Chen

H-K

, Ji

Z-S

, Dodson

, Miranda

, Rosenblum

, Reynolds

, Freedman

, Weisgraber

, Huang

, Mahley

(2011) Apolipoprotein E4 domain interaction mediates detrimental effects on mitochondria and is a potential therapeutic target for Alzheimer disease. J Biol Chem 286, 5215–5221.

60.

Leung

, Andrews-Zwilling

, Yoon

, Jain

, Ring

, Dai

, Wang

, Tong

, Walker

, Huang

(2012) Apolipoprotein E4 causes age- and sex-dependent impairments of hilar GABAergic interneurons and learning and memory deficits in mice. PLoS One 7, e53569.

61.

Ricchelli

, Gobbo

, Moreno

, Salet

(1999) Changes of the fluidity of mitochondrial membranes induced by the permeability transition. Biochemistry 38, 9295–9300.