Mitochondrial Dysfunction as a Predictor and Driver of Alzheimer’s Disease-Like Pathology in OXYS Rats

Abstract

Growing evidence suggests that mitochondrial dysfunction is an early event in sporadic Alzheimer’s disease (AD), but the impact of mitochondrial dysfunction on the transition from healthy aging to AD remains elusive. Here we estimated the influence of mitochondrial dysfunction on the initiation of AD signs in OXYS rats, which simulate key characteristics of sporadic AD. We assessed the mitochondrial ultrastructure of pyramidal neurons of the hippocampus at the age preceding the development (age 20 days), during manifestation (4–5 months), and at the well-pronounced stages (18–24 months) of the AD-like pathology in OXYS rats. Ultrastructural alterations were collated with the amounts of proteins mediating mitochondrial dynamics [mitofusins (MFN1 and MFN2) and dynamin-1-like protein (DRP1)]; with activity of respiratory chain complexes I, IV, and V in the hippocampal mitochondria; with reactive oxygen species (ROS) production; and with expression of uncoupling protein 2 (UCP2) regulating ROS production. Already at the preclinical stage, OXYS rats showed some characteristic changes in hippocampal mitochondria, which increased in size with the manifestation and progression of AD-like pathology, including decreased activity of respiratory complexes against the background of greater fusion and formation of larger mitochondria. Signs of AD developed simultaneously with increasing dysfunction of mitochondria, with a dramatic decrease in their number, and with increased fission but without upregulation of ROS production (observed only in 20-day-old OXYS rats). Summarizing the data from our present and previous studies, we conclude that mitochondrial dysfunction appears to mediate or possibly even initiate pathological molecular cascades of AD-like pathology in OXYS rats and can be considered a predictor of the early development of the late-onset form of AD in humans.

Keywords

Alzheimer’s disease mitochondrial dysfunction senescence-accelerated OXYS rats

INTRODUCTION

The age-related impairments of mitochondrial functions are called the missing link between brain aging and the development of the sporadic form of Alzheimer’s disease (AD), the most common type of age-related dementia worldwide, which has no cure [1]. There is convincing evidence that mitochondria may mediate, drive, or contribute to a variety of AD-associated pathologies. The central role of mitochondria in aging was first proposed by Denham Harman, on the basis of his original theory that aging is caused by the accumulation of damage resulting from reactive oxygen species (ROS) [2]. On the other hand, recent findings suggest that formation of ROS is neither the primary nor the initial cause of aging [3]. Moreover, transient stress on mitochondria, including mitochondrial ROS (playing a critical role in a number of intra- and extracellular processes), elicits beneficial changes that extend the lifespan [4,5, 4,5]. Clearly, mitochondrial function regulates the rate of aging, and mitochondrial dysfunction takes the center stage in the pathophysiology of age-related neurodegenerative disorders, but the underlying mechanisms remain unclear [6].

Growing evidence suggests that mitochondria at least mediate or possibly even initiate pathological molecular cascades of sporadic AD; this evidence supports the hypothesis of a primary mitochondrial cascade, which means that mitochondrial dysfunction exists independently of amyloid-β (Aβ), and potentially is upstream of Aβ deposition [7]. Numerous studies have revealed that dysfunctional mitochondria can contribute to aging independently of ROS, and not only mitochondrial metabolic dysfunction but also mitochondrial dynamics and mitochondrial calcium uptake can be involved in the neurodegeneration in AD [8 –10]. In addition, mitochondrial deficiencies directly influence cellular signaling including the cascade of apoptosis and interorganellar crosstalk, by affecting mitochondrion-associated membranes that constitute an interface between the outer mitochondrial membrane and the endoplasmic reticulum (ER) [11]. Besides, mitochondria are the central regulators in cellular protein signaling pathways that modulate mitochondrial fusion and fission, which are necessary for mitochondrial health. Thus, the relation between mitochondrial dysfunction and AD is obvious, but dissecting its details remains a major challenge for research on the disease pathogenesis.

According to the author of the “mitochondrial cascade hypothesis,” R.H. Swerdlow, the place of mitochondrial dysfunction is at the apex of the pyramid of AD-associated pathological changes [12]. Our studies on AD-like pathology in senescence-accelerated OXYS rats (which simulate the key characteristics of the sporadic form of AD [13 –15] suggest that mitochondrial dysfunction can mediate or possibly initiate pathological molecular cascades of AD, and this way, may form the base of the above-mentioned pyramid. Recently, we reported that an age-dependent increase in the levels of Aβ_1–42 and extracellular Aβ deposits in the brain of OXYS rats occur later than do mitochondrial structural abnormalities, hyperphosphorylation of the tau protein, synaptic loss, and neuronal cell death [16]. The development of early disturbances in mitochondrial function in OXYS rats is indicated by the significant increase in the prevalence of a common deletion (4834 bp) in mitochondrial DNA (mtDNA), especially at the stage of completion of brain development in the postnatal period [17]. This mutation may be the cause of accelerated brain aging in OXYS rats if this deletion results in energy deficiency. Nevertheless, a nuclear magnetic resonance study on brain energy metabolism has not detected any signs of energy deficiency in the brain of postnatal and young adult OXYS rats [18].

In the present study, using OXYS rats, we explored the ways of violation of mitochondrial function and its impact on the initiation or progression of pathological molecular cascades of AD. We assessed the mitochondrial ultrastructure of pyramidal neurons of the CA1 region of the hippocampus during several key periods: 1) in 20-day-old animals, i.e., at the age preceding the development of AD-like pathology in OXYS rats; 2) in 4- to 5-month-old animals, i.e., the age of unfolding manifestation of AD-like pathology; and 3) at the age of 18–24 months: the period of active AD-like pathology [15, 16]. Ultrastructural alterations were collated with the levels of principal proteins mediating mitochondrial dynamics [mitofusins (MFN1 and MFN2) and dynamin-1-like protein (DRP1)] and activity of respiratory chain complexes I, IV, and IV in hippocampal mitochondria. Mitochondrial function was also evaluated by measuring ROS production together with uncoupling protein 2 (UCP2) expression, which has been demonstrated to regulate ROS production [19].

MATERIALS AND METHODS

Animals

Male senescence-accelerated OXYS rats and age-matched male Wistar rats (parental control strain) were obtained from the Breeding Experimental Animal Laboratory of the Institute of Cytology and Genetics (ICG), the Siberian Branch of the Russian Academy of Sciences (Novosibirsk, Russia). All the experiments on rats were carried out according to Animal Care Regulations of the ICG, Novosibirsk. The OXYS rat strain was derived from the Wistar strain at the ICG as described earlier [20] and was registered in the Rat Genome Database (http://rgd.mcw.edu/). At the age of 4 weeks, the pups were taken away from their mothers and housed in groups of five animals per cage (57×36×20 cm) and kept under standard laboratory conditions (22°C±2°C, 60% relative humidity, and 12-h light/12-h dark cycle; lights on at 9 a.m.). The animals were provided with standard rodent feed (PK-120-1; Laboratorsnab, Ltd., Moscow, Russia) and water ad libitum.

Tissue sample preparation

After the rats were euthanized via CO₂ asphyxiation, brains were carefully excised, and the hemispheres were separated along the midline. For electron microscopic examination, the hippocampus samples (n = 3) from 20-day-old and 5- and 24-month-old Wistar and OXYS rats were fixed with 2.5% glutaraldehyde in sodium cacodylate buffer. For western blot analysis, the hippocampus of 20-day-old and 5- and 24-month-old rats (n = 9) was separated from the brain. Mitochondrial fraction was immediately isolated from hippocampal tissue and stored at –70°C until analysis. For immunohistochemical assays, the hemispheres of 20-day-old and 5- and 24-month-old rats were immediately fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS) for 48 h, followed by cryoprotection in 30% sucrose in PBS at 4°C for 48 h, and then the brains were frozen and stored at –70°C until analysis. The fixed hemispheres were sliced at 20 μm thickness (n = 4 to 5; 4–6 serial tissue sections per animal) using a Microm HM-505 N cryostat (Microm International GmbH, Germany). For the measurement of ROS production rate, we used Wistar and OXYS rats at ages 20 days, 3 months, and 24 months (n = 5 to 7). During isolation of brain mitochondria from 20-day-old rats, because the mass of the brain of a single animal was too small, we combined brain tissues from two animals and homogenized them together to obtain a single data point. For an enzyme-linked immunosorbent assay (ELISA), the hippocampus of 20-day-old and 3- and 18-month-old rats (n = 6) was separated from the brain, placed in a microcentrifuge tube, and was stored at –70°C until experiments. For assays of activities of mitochondrial complexes I, IV, and V, the mitochondrial fraction was isolated from the hippocampus of rats at ages 20 days, 4 months, and 24 months (n = 6).

Electron microscopy, calculations, and image analysis

For electron microscopic examination, small samples from the hippocampus (2×2×2 mm) containing the CA1 region were fixed with 2.5% glutaraldehyde in 0.1 M cacodylate buffer for 1 h at room temperature, washed twice in the buffer, then postfixed with 1% osmium tetroxide in the same buffer containing of few crystals of potassium ferricyanide (K₃[Fe(CN)₆]) for 1 h at room temperature and finally incubated in a 1% aqueous solution of uranyl acetate overnight. Next day, the samples were dehydrated in an ethanol series and in acetone and embedded in the Agar 100 Resin (Agar Scientific, Essex, UK). Complete polymerization of the samples was conducted by keeping them in a 60°C oven for 3 days. Ultrathin (70 nm) sections were obtained by means of a Leica Ultracut ultra-microtome (Leica Microsystems GmbH, Germany) and were stained with Reynolds lead citrate and uranyl acetate. The tissue slices were examined under a JEOL JEM-100SX transmission electron microscope (JEOL Ltd., Japan) at 60 kV at the Interinstitutional Shared Center for Microscopic Analysis of Biological Objects (Institute of Cytology and Genetics, Novosibirsk, Russia) at low magnification (3500×) so that the whole soma of a neuron could be analyzed at once. In the ImageJ software (National Institutes of Health open software; http://rsbweb.nih.gov/ij), we estimated the extent of mitochondrial alteration and quantified parameters reflecting the functional state of the mitochondria: quantity of the mitochondria per unit area of the cytoplasm, the average area of a mitochondrial cross-section, the number of mitochondria undergoing fission or fusion, the number of filamentous mitochondria, and the number of mitochondrial contacts with the ER and with the outer nuclear membrane. The results were processed by statistical analysis. At least ≥30 randomly chosen cells were imaged in each group.

Isolation of total protein from the hippocampal mitochondrial fraction

Hippocampal samples were immersed in a buffer (250 mM sucrose; 1 mM EDTA; 70 mM KCl in PBS, pH 7.4) with 1:10 protease inhibitors (Protease inhibitor cocktail P8340; Sigma-Aldrich), sheared with scissors, and homogenized with a mechanical homogenizer (# Z359971, Sigma, USA) for 3 min at 4°C on ice. The homogenates were centrifuged at 800×g for 10 min. Supernatants were decanted and centrifuged at 15,000×g for 20 min. The pellets were resuspended in 120 μl of a buffer (50 mM Tris-HCl pH 7.4; 150 mM NaCl; 2 mM EGTA; 0.2% Triton X-100; 0.3% NP-40), and incubated on ice for 30 min. The supernatants were transferred to new tubes as the mitochondrial fraction and stored at −70°C. Total protein was quantified by means of the Bio-Rad Bradford Kit (Bio-Rad Laboratories, USA).

Western blotting

For this purpose, approximately 25 μg samples of the mitochondrial fraction were resolved by SDS-PAGE (15% gel) in Tris-glycine running buffer (TGB; 25 mM Tris base, 190 mM glycine, and 0.1% SDS) and transferred to a nitrocellulose membrane (Hybond-CExtra, Amersham, USA). The membrane was blocked with a 5% solution of bovine serum albumin (Sigma–Aldrich, USA) in PBS with 0.1% Tween 20 (TBS-T) for 1 h at room temperature, and then was incubated for 3 h at room temperature with an antibody against Drp1, VDAC, mitofusin 1, or mitofusin 2 (catalog ## ab56788, ab14734, ab57602, or ab50838; Abcam, USA) diluted 1:1000 in the above-mentioned blocking solution. Secondary antibodies were the corresponding horseradish peroxidase (HRP)-conjugated antibodies (## ab6721, ab6808; Abcam, USA). After incubation with the respective secondary antibody, chemiluminescent signals were measured and scanned, and intensity of the bands was quantified in the ImageJ software (http://rsb.info.nih.gov/ij/). GAPDH and VDAC served as internal loading controls.

Immunofluorescent staining

This procedure was performed by a standard indirect method as described previously [16]. Primary antibodies and dilutions were as follows: anti-DRP1 (1:250; # ab56788, Abcam), anti-Mfn1 (1:250; # ab57602, Abcam), anti-Mfn2 (1:250; # ab50838, Abcam). After incubation with the respective secondary antibodies conjugated with DyLight-650 or Alexa Fluor 488 (## ab96886, ab150073, Abcam) diluted 1:250, the slices were coverslipped with the Fluoro-shield mounting medium containing 4^′,6-diamidino-2-phenylindole (DAPI; # ab104139, Abcam) and examined under an Axioplan 2 microscope (Zeiss).

Assays of activity of mitochondrial complexes I, IV, and V

Activities of these mitochondrial complexes were measured using enzyme assay kits according to manufacturer’s protocols (## ab109721, ab109911, 109907; Abcam, USA). Protein concentrations of mitochondrial lysates were estimated. Enzymatic activities were measured spectrophotometrically on a CLARIOStar spectrophotometer (BMG Labtech, Germany) and expressed as changes of absorbance per minute per milligram of mitochondrial protein.

Isolation of brain mitochondrial fraction

Animals were anesthetized with halothane and decapitated. All further manipulations were performed at 4°C according with procedure described by A. Panov [21] Briefly, after the brain was excised, the cerebellum and brainstem were cut out and discarded. The forebrain was immersed in ice-cold sucrose buffer (75 mM mannitol, 175 mM sucrose, 10 mM MOPS, pH 7.2, 1 mM EGTA), sheared with scissors, and homogenized in a glass Dounce homogenizer. The homogenate was centrifuged at 1000 × g for 7 min, the supernatant was then centrifuged at 10 800 × g for 12 min. The pellet was resuspended in 15% Percoll (in sucrose buffer) and layered atop 23% and 40% layers. The mitochondrial fraction between 40% and 23% layers was collected, washed in sucrose buffer twice with centrifugation at 16100 × g and 8500 × g, respectively, for 11 min. The final pellet was resuspended in 350 μl of the incubation medium (120 mM KCl, 10 mM NaCl, 2 mM MgCl₂×6H₂O, 2 mM KH₂PO₄, 20 mM MOPS pH 7.2, 1 mM EGTA, 0.7 mM CaCl₂×2H₂O) and subjected to measurement of hydrogen peroxide production. H₂O₂ was quantified with Amplex red as described in [22].

ROS measurement

ROS formation was quantified as H₂O₂ production kinetics measured on a Varian Cary Eclipse (Varian, USA) spectrofluorometer. A sample of a mitochondrial suspension equivalent to 50 μg of total protein was added to cuvettes containing HRP, Amplex Red (2 mM), and a mixture of the following substrates: glutamate (10 mM), malate (2 mM), pyruvate (2.5 mM), and succinate (5 mM). H₂O₂ generated during mitochondrial oxidative activity converted Amplex Red to resorufin with excitation/emission maxima 570/585 nm. The slope of a fluorescence change graph constructed in the Cary Eclipse software was used to calculate H₂O₂ generation speed in pmol H₂O₂/min per mg of protein.

ELISA

Tissue samples were homogenized in PBS on ice, sonicated twice (Vibra-Cell VCX 130, Sonics & Materials, USA), and centrifuged at 5000 × g for 5 min. UCP2 protein levels were assayed with the ELISA Kit for Uncoupling Protein 2, Mitochondrial (UCP2) (Cusabio, PRC). UCP2 levels were normalized to the total protein.

Statistical analysis

The data were subjected to ANOVA (Statistica 10.0 software). The Newman–Keuls test was applied to significant main effects and interactions to assess the differences between some sets of means. The data are presented as mean±SEM. The differences were considered statistically significant at p < 0.05.

RESULTS

Decreased mitochondrial content and impaired mitochondrial structure in the hippocampal neurons of OXYS rats

To evaluate mitochondrial impairment during the development of AD-like pathology in OXYS rats, we conducted comparative analysis of the content (Fig. 1A) and ultrastructural state (Fig. 1B) of mitochondria in the hippocampal CA1 region in OXYS and Wistar rats at ages 20 days and 5 and 24 months. First, we calculated a relative quantity of mitochondria, which reflects total mitochondrial content of a neuronal soma. We found that mitochondrial content in the hippocampus was lower in OXYS rats (F_1,188 = 7.4, p = 0.007). The maximal quantity of mitochondria in OXYS rats was seen at the age of 20 days, and this content decreased with age (p = 0.042; Fig. 1A). A significant decrease of this parameter in the hippocampus of OXYS rats occurred at the age of 5 months (25%; p = 0.0055) as compared to Wistar rats. This parameter reached a twofold interstrain difference at the age of 24 months (p = 0.01) during the increase of mitochondrial content in the hippocampus of Wistar rats (nonsignificant).

Fig.1

OXYS rats experience notable mitochondrial depletion starting from the age of 5 months, with this change becoming larger at the age of 24 months owing to an increase in the mitochondrial number in Wistar rats while the depletion in OXYS rats persists. Note damaged mitochondria appearing in aged animals (insets). Data are presented as mean±SEM, per 100 μm² of the cytoplasmic area. White scale bar represents 5 μm, black scale bar represents 1 μm. White arrowheads indicate lipofuscin. Note the indentation in the nuclear membrane: a black arrow. ^#Significant differences between the strains of the same age, ^*significant age-related differences within a strain. Abbreviations: d, days; mo, months; dm, damaged mitochondria; Mv, multivesicular body.

To assess structural integrity of mitochondria in OXYS rats, we chose a classification described previously [23]. We determined the shares (%) of intact mitochondria (structurally intact cristae), of moderately damaged mitochondria (electron-lucent matrix, ruptured cristae, and signs of vacuolation), and of severely damaged mitochondria (significantly rarefied matrix, near-complete loss of the cristae structure, and marked swelling). These parameters changes with age in both rat strains (F_2,187 = 23.9, p < 0.0001, F_2,187 = 21.2, p < 0.0001, and F_2,187 = 9.26_, p = 0.00015, respectively). The proportion of intact mitochondria decreased in Wistar and OXYS rats from the age of 5 months to age 24 months (74.7% ±3.2% to 59.3% ±4.1%, p = 0.0036, and 76.6% ±3.1% to 54.5% ±4.9%, p = 0.0003, respectively) while the proportion of moderately damaged mitochondria increased (18.5% ±2.1% to 29.0% ±2.8%, p = 0.0032, and 15.9% ±2.4% to 28.9% ±3.1%, p = 0.0013, respectively). Thus, we didn’t found the differences in the proportion of intact and moderately damaged mitochondria between OXYS and Wistar rats. However, in OXYS rats, these age-dependent changes occurred on the background of significantly lower amount of mitochondria than that in Wistar rats. Therefore, the percentage of intact mitochondria in OXYS rats translates into less mitochondrial biomass when compared to age-matched Wistar rats. Moreover, the proportion of severely damaged mitochondria significantly increased only in OXYS rats from the age of 5 months to 24 months (7.5% ±1.6% to 16.6% ±3.3%, p = 0.013).

The average area of a mitochondrial cross-section reflects the volume of mitochondria in the cell; this parameter was higher in OXYS rats as compared to Wistar rats at the age of 20 days (0.26±0.02 and 0.2±0.02 μm², respectively, p = 0.017). With age, the area of mitochondrial cross-sections increased in both rat strains (F_2,187 = 23.7, p < 0.0001), and at the age of 24 months, this parameter was 0.46±0.05 μm² in Wistar rats and 0.47±0.03 μm² in OXYS rats.

Because mitochondria constitute the main source of cellular ATP, their colocalization with other cellular structures may indicate increased energy demand of those structures [24, 25]. Thus, the number of mitochondria adjacent to the nucleus reflects the extent of nuclear ATP consumption. This parameter increased with age in both rat strains (F_2,187 = 3.36, p < 0.00001), but only in OXYS rats at age 5 months, was the number of mitochondria adjacent to the nucleus threefold greater when compared to the age of 20 days (p = 0.012) and was twofold greater when compared to age-matched Wistar rats (p = 0.011). This age-dependent increase persisted only in OXYS rats: at the age of 24 months, this parameter was threefold greater when compared to the age of 20 days (p = 0.012).

The amount of filamentous mitochondria forming during a fusion of several smaller organelles decreased with age in both rat strains (F_2,188 = 12.06, p = 0.00001), but only 24-month-old OXYS rats manifested a trend toward a decrease of this parameter when compared to 5-month-old animals (1.92% ±0.8% and 4.56% ±1.04%, respectively, p = 0.051).

Impaired mitochondrial fusion and fission in the hippocampal neurons of OXYS rats

Morphologically, the events of mitochondrial dynamics could be observed as two mitochondria came into direct contact (Fig. 2A). Thus, the frequency of such contacts reflects the intensity of mitochondrial dynamics. At the age of 5 months, the number of mitochondrial contacts was lower in OXYS rats than in Wistar rats (p = 0.0017), and by the age of 24 months this parameter reached a twofold interstrain difference (7.8±1.6 and 14.2±2.1 respectively, p = 0.0014; Fig. 2B).

Fig.2

The proportion of mitochondria participating in the events of mitochondrial dynamics in OXYS rats decreases at the age of 5 months and remains at the low level until the age of 24 months. Data are presented as mean±SEM of the percentage of all mitochondria of a neuronal soma. Arrowhead marks lipofuscin. The scale bar represents 500 nm. ^#Significant differences between the strains of the same age; d: days, mo: months.

Next, we examined age-related changes in the level of proteins directly mediating mitochondrial dynamics (MFN1, MFN2, and DRP1). In 20-day-old OXYS rats the level of MFN1 was higher than that in Wistar rats (p = 0.01; Fig. 3A, E); this result might indicate upregulation of mitochondrial fusion, complementary to the increase in the mitochondrial cross-section area in OXYS rats at the age of 20 days. This finding was supported by the increased MFN1/DRP1 ratio (reflecting the balance between fission and fusion) in 20-day-old OXYS rats as compared to Wistar rats (F_1,34 = 6.6, p = 0.015, Fig. 3C). With age, the level of MFN1 increased only in Wistar rats (p < 0.03), as did the MFN1/DRP1 ratio (p < 0.02). In contrast, with age, in OXYS rats, the amount of DRP1 increased (p < 0.05; Fig. 3B) while the MFN2/DRP1 ratio decreased (p < 0.05; Fig. 3D), indicating upregulation of mitochondrial fission. In 20-day-old and 4- and 24-month-old OXYS rats, the changes of MFN1, MFN2 and DRP1 amounts in the hippocampus revealed by immunohistochemical analysis (Fig. 3F, G) were consistent with the protein levels measured by western blot analysis.

Fig.3

At age 20 days, MFN1 fusion protein content is higher in OXYS rats compared to Wistar rats (A), whereas with age, OXYS rats experience upregulation of the fission-promoting DRP1 protein (B). The ratio of fusion protein MFN1 to fission protein DRP1 is elevated in OXYS rats at the preclinical stage (C); the fusion-to-fission machinery ratio in terms of MFN2 decreased in 24-month-old OXYS rats, revealing advanced signs of AD when compared to age-matched Wistar rats (D). Data are presented as mean±SEM. ^# Significant differences between the strains of the same age. The scale bar represents 20 μm. ^&Significant differences from 20-day-old animals; d: days, mo: months, W: Wistar, O: OXYS.

ER–mitochondria contacts are the main site of molecular interaction between mitochondria and the rest of the cell. At the age of 5 months, the number of ER–mitochondria contacts (Fig. 4A, B) was 1.5-fold higher in OXYS rats than in age-matched Wistar rats (p = 0.0017). By contrast, with age, this parameter in OXYS rat decreased by half (p = 0.0004), and at age 24 months, the number of ER–mitochondria contacts was lower in the hippocampus of OXYS rats than in the hippocampus of Wistar rats (p = 0.0008). At the same time in Wistar rats, this parameter increased only by the age of 24 months (p = 0.008) when it approached the value observed in 5-month-old OXYS rats (Fig. 4B).

Fig.4

The number of mitochondria forming contacts with the ER in OXYS rats increases at age 5 months but shows a sharp drop at the age of 24 months. ^#Significant differences between the strains of the same age. Data are presented as mean±SEM of the percentage of all mitochondria in a neuronal soma. Arrowheads trace ER–mitochondria contact sites. The black arrow indicates lipofuscin. The scale bar represents 500 nm. *Significant age-related differences within a strain; d: days, mo: months.

ROS production by OXYS rat mitochondria does not increase with the development of AD-like signs

Mitochondria are major sites of production and accumulation of reactive oxygen species, and mitochondrial dysfunction causes increased radical production. ROS production by brain mitochondria was assayed using several substrate mixtures [22]: pyruvate+malate, glutamate+malate, pyruvate+glutamate+malate, succinate, succinate+malate, succinate+pyruvate+glutamate, and pyruvate+glutamate+succinate+malate. Only at the age of 20 days, was ROS production by mitochondria (isolated from the brain of OXYS rats) slightly higher than that in age-matched Wistar rats (∼30%, p < 0.05) during incubation with a single substrate mixture: pyruvate+malate. Mitochondria from both 3- and 24-month-old OXYS rats showed no difference in the ROS production rate as compared to Wistar rats (Table 1.). To some extent, these results can be explained by specific details of the isolation of mitochondria by centrifugation, especially for the 24-month-old rats’ mitochondria. The bloated, disintegrating mitochondria may remain in the upper layer of the Percoll density gradient where they mix with a synaptosomal fraction, while intact mitochondria move downward where they form a distinct layer, as evidenced by electron microscopy analysis of a suspension of mitochondrial (Fig. 5A) and synaptosomal fractions (Fig. 5B).

Table 1

Generation of ROS by rat brain mitochondria oxidizing various substrates and substrate mixtures

Age	Strain	ROS generation rate in pmol H₂O₂/min per mg of protein
		PM	GM	PGM	S	SM	PGS	PGSM
20 d	Wistar	119.3±8.6	84.4±10.6	154.4±13.3	33.2±9.1	47.5±7.8	116.3±15.7	133.2±29.4
	OXYS	154.5±7.5^#	101.82±9.2	195.3±14.4	38.3±6.6	60.7±11.7	117.0±15.1	120.2±29.7
3 mo	Wistar	167.0±29.5	132.0±15.4	225.5±57.3	44.3±8.7	68.4±11.0	195.1±40.8	191.7±23.1
	OXYS	151.5±11.0	109.3±12.7	214.9±26.0	45.8±9.0	92.9±12.9	196.0±29.8	243.7±28.9
24 mo	Wistar	66.4±2.7	96.5±12.3	92.1±8.4	44.2±8.8	–	–	97.9±10.2
	OXYS	57.6±13.9	99.2±36.5	75.7±9.9	44.7±14.7	–	–	83.3±17.0

Data are presented as mean±SEM. ^#Significant differences between the strains of the same age; P, pyruvate; G, glutamate; S, succinate; M, malate; d, days; mo, months.

Fig.5

Representative micrographs of pellets from mitochondrial and synaptosomal fractions. The mitochondrial fraction (A) contains condensed well-preserved mitochondria. The synaptosomal fraction (B) with large vesicles derived from synaptic terminals (asterisks) contains occasional bloated mitochondria (arrowheads). Scale bar: 1 μm. UCP2 protein content (C) in the hippocampus of both Wistar and OXYS rats increases with age. Data are presented as mean±SEM. *Significant age-related differences within a strain; d: days, mo: months.

To further evaluate the state of mitochondrial function in OXYS rats and to explore the possible mechanisms by which they may regulate ROS production rates, we have measured hippocampal content of a member of the uncoupling protein family (UCP2), which is considered a key regulator of ROS production and is expressed in a wide range of cells and tissues [19]. According to ANOVA, UCP2 protein content increased with age in the hippocampus of both Wistar rats and OXYS rats (F_2,25 = 61, p < 0.001): at age 3 months, the level of protein was twice higher when compared to 20-day-old animals (p < 0.001 and p < 0.03, respectively; Fig. 5C). A further increase from 3 to 18 months of age was seen in the hippocampus of Wistar and OXYS rats (p < 0.001 for both rat strains). In 18-month-old OXYS rats, we observed a tendency toward diminished UCP2 protein content when compared to age-matched Wistar rats (p = 0.058).

Decreased activity of respiratory chain complexes I, IV, and V in the hippocampus of OXYS rats

Activity of the respiratory chain is the step that most directly determines cellular ATP output. We have measured age-related changes in the activities of the respiratory chain complexes I, IV, and V in the hippocampal mitochondria of Wistar and OXYS rats. This is because a decrease in these activities is an important sign of the impairment of mitochondrial function in the AD brain.

According to ANOVA, the activity of complex I was lower in OXYS rats (F_1,23 = 17.4, p < 0.0007) but did not change with age in either rat strain (Fig. 6A). At the age of 24 months, the activity of complex I was lower in OXYS rats than in age-matched Wistar rats (p < 0.03). Likewise, we found that complex IV activity was decreased in OXYS rats (F_1,23 = 36.5, p < 0.0001) and further decreased with age in both rat strains (F_2,23 = 4.8, p < 0.02; Fig. 6B). Already at the age of 20 days, OXYS rats had lower activity of complex IV (p < 0.002), and this parameter remained lower than that in Wistar rats at the ages of 4 months (p < 0.002) and 24 months (p = 0.08; Fig. 6B). According to ANOVA, complex V activity also decreased in OXYS rats (F_1,39 = 12.2, p = 0.001), with the interstrain difference being significant at age 4 months (p < 0.05) when this activity was 30% lower in OXYS rats than in Wistar rats (Fig. 6C).

Fig.6

OXYS rats are characterized by decreased activity of respiratory chain complexes in hippocampal mitochondria. Complex I activity (A) is decreased at the age of 24 months whereas complex IV activity (B) is impaired in both young and adult animals, whereas Wistar rats experience a similar decrease only at age 24 months. Likewise, complex V activity (C) shows a tendency toward a decrease, with 5-month-old OXYS rats experiencing impaired activity. Data are presented as mean±SEM. ^# Significant differences between the strains of the same age; ^&significant differences from 20-day-old animals. OD: optical density.

DISCUSSION

Normal brain aging and AD are both marked by mitochondrial impairments, albeit to varying degrees [1]. Neuropathological studies in animal models and AD patients suggest that mitochondrial dysfunction is an early event in most of late-onset AD cases [26], but its impact on the transition from “healthy” aging to AD pathology has not yet been completely elucidated. According our recent data, the progression of AD-like pathology in OXYS rats is deeply involved in mitochondrial abnormalities. The purpose of this study was to estimate the influence of mitochondrial dysfunction on the initiation or development of the AD-like pathology in OXYS rats. Our attention was focused on the period of the development of disease signs in OXYS rats when energy deficiency plays a significant role and may be the core event eventually leading to the manifestation of clinical symptoms.

We found that already at the age of 20 days, i.e., at the preclinical stage of the development of AD-like pathology, OXYS rats showed some characteristic changes in hippocampal mitochondria, which increased in size during the manifestation (the age of 5 months) and progression of these pathological changes (18 months). First, even at this early age, OXYS rats show significantly lower respiratory complex IV activity, which is incidentally the most consistent finding in aged tissues [3] and a tendency toward a decrease in the activity of complexes I and V. Furthermore, the persistent depression of respiratory chain activity in the hippocampal mitochondria of OXYS rats is observed throughout the lifespan. Simultaneously this decreased activity, OXYS rats show increased fusion, which leads to formation of larger mitochondria. Such fused highly integrated mitochondrial phenotype is thought to be geared toward upregulation of energy supply via ATP synthesis.

Therefore, we probably did not detect any signs of energy deficiency in OXYS rats’ brain up to the age of 3 months, when we estimated the brain energy metabolism by ³¹P NMR spectroscopy [17]. What’s more, we have previously detected an increased phosphate potential in 2- to 3-week-old OXYS rats compared to Wistar rats (elevated creatine phosphate to inorganic phosphate [P_Cr/P_i] and P_Cr/ATP ratios) which could be a consequence of adaptation of the brain to hypoxic conditions [18]. Later, we have shown that the prevalence of the mitochondrial common deletion (4834 bp in mtDNA; which is the most typical form of mtDNA oxidative damage associated with aging and degenerative diseases) in the hippocampus of OXYS rats is increased, especially at the stage of completion of brain development in the postnatal period [17]. This level remained elevated not only at the stages preceding the manifestation and the development of the signs of AD but also during their progression. Nonetheless, at age 24 months, there were no detectable differences in the prevalence of mtDNA 4834 bp deletion between the OXYS and Wistar rats.

By the age of 20 days, final maturation of the brain takes place, during which many brain regions, including the hippocampus, undergo development and histogenesis associated with the formation of interneuronal contacts and elimination of “transitional” cell populations by apoptosis [28]. High apoptotic activity [28] and, according to the latest data, neurogenesis [29], point to enhanced ROS production. It is noteworthy that in the present study we found that early age (20 days) is the only period when brain mitochondria from OXYS rats (and only when oxidizing pyruvate + malate) show signs of increased production of ROS. It should be noted that increased production of ROS occurs much earlier than amyloid plaque accumulation, tau hyperphosphorylation, or behavioral abnormalities in OXYS rats [16]. Additionally, cognitive impairments in OXYS rats manifest earlier than significantly increased level of brain oxidative stress markers [30]. This is consistent with recent studies suggesting that oxidative damage events in the AD brain, at least in part, can be related to inflammatory responses rather than directly from mitochondrial dysfunction [31].

For a long time overproduction of ROS has been regarded as the cornerstone of mitochondrial pathology and of neurodegenerative diseases in which mitochondrial pathology plays an important part [32]. In contrast, we did not see enhanced formation of ROS by mitochondria isolated from the brain of 5- and 24-month-old rats when the AD-like pathology developed and progressed [15, 16] and well-pronounced structural disturbances in mitochondria were revealed in the hippocampus of OXYS rats by electron microscopy. Nevertheless, the results of this study are consistent with other studies [33] indicating that OXYS liver mitochondria from 4- and 12-month-old rats produce less ROS simultaneously with a slight decrease in the transmembrane electrical potential. We hypothesized that the state of mitochondria in OXYS rats may be characterized as mild uncoupling, which can decrease ROS production by lowering the membrane potential; this change increases the rate of electron transfer and oxygen consumption while maintaining respiratory complexes in a more oxidized state [34].

Mitochondrial uncoupling proteins (UCPs) are believed to perform a major function in this process. UCPs are a group of five mitochondrial inner-membrane transporters with variable tissue expression, which seem to function as regulators of energy homeostasis and antioxidants. In particular, these proteins uncouple respiration from ATP production, allowing stored energy to be released as heat. Data from experimental models have suggested that UCPs may strongly influence the aging rate and lifespan [35 –37]. Polymorphisms in UCP2 associated with longevity in humans have been revealed [38]. Recently, it was shown that expression of both UCP2 and UCP4 is significantly lower in the AD brain as compared to non-AD brains [39]. Our present assays revealed that UCP2 levels significant increase with age in the hippocampus of both OXYS and Wistar rats. On the other hand, in OXYS rats, this increase was less significant in the period of a significant decline in the number of mitochondria in the hippocampus: 1.5-fold during the manifestation and twofold during progression of AD-like pathology. Therefore, accurate quantification of interstrain differences in UCP2 amounts on the inner mitochondrial membrane is problematic. We can only assume that simultaneously with the decreased volume of mitochondria in the hippocampus, the amount of UCP2 on the inner membrane of mitochondria in 4-month-old OXYS rats is higher than that in Wistar rats. Our results seem to be in agreement with recent observations by Lores-Arnaiz et al. [40] showing that UCP2 protein expression increases both in synaptosomes and in nonsynaptic mitochondria of 17-month-old mice as compared with young animals. UCP2 upregulation seems to be a possible mechanism by which mitochondria may become resistant to oxidative damage during aging and this trend toward an increase of inducible uncoupling with age may be a general adaptive mechanism of aging cells and supports the “uncoupling to survive” hypothesis [41].

Data on the change of mitochondrial mass during brain aging are limited but overall suggest that with advancing age, this parameter increases or tends to increase [42]. In 2-year-old Wistar rats, we observed a tendency toward an increase in the number of mitochondria in hippocampal neurons. Furthermore, in OXYS rats, their amount decreased by age 5 months, remained at this level at 24 months and was significantly lower than that in Wistar rats. Starting at the age of 4 months and until age 24 months, we observed a shift from mitochondrial fusion toward fission as evidenced by increasing DRP1 content. Decreased mitochondrial fusion and formation of a fragmented mitochondrial phenotype is a characteristic sign of many models of neurodegenerative diseases [43]. Mitochondrial fission itself may also contribute to the removal of mitochondria through subsequent mitophagy contributing to the diminishing number of mitochondria. Overall, the observed events appear to signify a mismatch between the intensities of biogenesis and mitophagy; this phenomenon as a whole is also considered a hallmark of age-associated neurodegenerative diseases [44].

Mitochondria are physically and biochemically in contact with ER via formation of mitochondria-associated membranes harboring various effector and signaling proteins positioned in lipid rafts. In addition to mitochondrial dynamics, ER contact sites play a pivotal role in mitophagy, membrane lipid transport, and calcium signaling [45]. It was demonstrated that a moderate increase in mitochondrial calcium supplied by the ER during formation of a membrane contact site also increases respiratory activity and ATP synthesis [46]. Thus, increased formation of ER contact sites on neuronal mitochondria of OXYS rats during manifestation of AD signs (age 5 months) may be a part of an ongoing compensatory reaction aimed at increasing the supply of ATP and improving the quality control of the remaining mitochondria. Similarly, OXYS neuronal mitochondria starting from a young age are more frequently positioned next to the nucleus, with the frequency increased threefold in 5- and 24-month-old rats, suggesting that starting at the age of the first signs of AD, the demand for ATP in the neuronal nucleus increases. Moreover, increased formation of ER contact sites is seen in human AD patients: both in fibroblasts from patients with familial or sporadic AD [47] and in postmortem brain tissue, as well as in a mouse model of AD [48]. Additionally, a mitofusin-2 knockdown increases ER–mitochondria contacts and decreases Aβ production [49]. Notably, in OXYS rats, formation of ER contact sites after an increase at the age of 5 months underwent a significant decrease later, at the age of active progression of AD-like pathology. This change in ER contact sites is accompanied by signs of AD, such as Aβ accumulation, tau hyperphosphorylation, synaptic impairment, and neuronal death. These data suggest that by the late age (24 months), an initial compensatory response aimed against Aβ accumulation no longer functions. Additionally, at this age, OXYS rats accumulate more severely damaged mitochondria as compared to Wistar rats; this finding indicates that OXYS mitochondria are under a significant workload, where quality control mechanisms become insufficient. It is worth noting that Wistar rats also manifest accumulation of damaged mitochondria with age as well as an increase in formation of ER–mitochondria contact sites by age 24 months. By contrast, in OXYS rats, these changes develop against the background of a persistently decreasing number of mitochondria and depressed OXPHOS activity; the latter phenomenon may be important for the development of AD-like pathology.

Summarizing the results of the present and previous studies in Fig. 7, we compared the age-dependent changes in neuronal mitochondria with the development of AD signs in OXYS rats. We can see that already at the age of 20 days in the absence of any signs of AD, OXYS rats show decreased activity of complex IV of the respiratory chain, increased ROS production, and accumulation of deletions in mitochondrial DNA. Manifestation of behavioral aberrations and deterioration of cognitive abilities in OXYS rats take place at the age of 3–4 months during the increase in mitochondrial aberrations and hyperphosphorylation of tau protein. Increased accumulation of Aβ in the brain of OXYS rats occurs at age 7–12 months, concurrently with well-pronounced neurodegenerative changes, and becomes a secondary event in relation to mitochondrial dysfunction. The key role of mitochondrial dysfunction in the pathophysiology of AD is confirmed by the ability of mitochondria-targeted antioxidant SkQ1 to alleviate the neurodegenerative alterations via improvement of structural and functional state of mitochondria in OXYS rats. Namely, SkQ1 prevents the neuronal loss and synaptic damage, enhances neurotrophic supply, decreases the amounts of toxic forms of Aβ and tau hyperphosphorylation in the hippocampus of OXYS rats, thereby resulting in improvement of cognitive function [14, 50]. Collectively, these data allow us to conclude that mitochondrial dysfunction appears at least to mediate or possibly even initiate pathological molecular cascades of AD-like pathology in OXYS rats, but we do not know the molecular genetic basis of the main culprit of mitochondrial alterations. The identification of the genes responsible for mitochondrial dysfunction will enable us to further delineate the events leading to sporadic-AD–related aberrations Nevertheless, we assume that mitochondrial dysfunction may be considered a predictor of the early development of the late-onset form of AD in humans. If so, mitochondrial dysfunction actually constitutes a mechanistic link between brain aging and AD.

Fig.7

The age-dependent changes in neuronal mitochondria and the key periods of the development of AD-like pathology in OXYS rats. At the preclinical stage (age 20 days), OXYS rats show decreased respiratory chain activity, increased ROS production and mtDNA 4834 bp deletion, and upregulation of mitochondrial fusion. The decrease in respiratory-chain activity, in the number of mitochondria, and in mitochondrial dynamics, increased mtDNA 4834 bp deletion, and increasing contacts with the ER during manifestation of the signs of AD-like pathology in OXYS rats (age 3–5 months) take place simultaneously with hyperphosphorylation of tau protein, synaptic and neuronal damage, behavioral abnormalities, and a decline of cognitive abilities. At the progressive stage (the age of 18–24 months), OXYS rats show a decreased number of mitochondria, lower respiratory chain activity, and smaller contact with ER, but increased levels of APP, accumulation of Aβ, formation of amyloid plaques, synaptic loss, neuronal cell death, behavioral abnormalities, and deterioration of cognitive abilities.

Footnotes

ACKNOWLEDGMENTS

The authors would like to thank Ekaterina A. Rudnitskaya, who helped with the immunohistochemical analysis. This work was supported by a Russian Scientific Foundation grant (16-15-10005).

Authors’ disclosures available online ().

References

Grimm

, Friedland

, Eckert

(2016) Mitochondrial dysfunction: The missing link between aging and sporadic Alzheimer’s disease. Biogerontology 17, 281–296.

Harman

(2009) Origin and evolution of the free radical theory of aging: A brief personal history, 1954-2009. Biogerontology 10, 773–781.

Payne

BAI

, Chinnery

(2015) Mitochondrial dysfunction in aging: Much progress but many unresolved questions. Biochim Biophys Acta 1847, 1347–1353.

Ristow

, Zarse

(2010) How increased oxidative stress promotes longevity and metabolic health: The concept of mitochondrial hormesis (mitohormesis). Exp Gerontol 45, 410–418.

Rose

, Santoro

, Salvioli

(2017) Mitochondria and mitochondria-induced signalling molecules as longevity determinants. Mech Ageing Dev 165, 115–128.

Theurey

, Pizzo

(2018) The aging mitochondria. Genes (Basel) 9, 22.

Gao

, Wang

, Liu

, Xie

, Su

, Wang

(2017) Abnormalities of mitochondrial dynamics in neurodegenerative diseases. Antioxidants 6, 25.

Correia-Melo

, Passos

(2015) Mitochondria: Are they causal players in cellular senescence? Biochim Biophys Acta 1847, 1373–1379.

Gonzalez-Freire

, De Cabo

, Bernier

, Sollott

, Fabbri

, Navas

, Ferrucci

(2015) Reconsidering the role of mitochondria in aging. J Gerontol A Biol Sci Med Sci 70, 1334–1342.

10.

Ziegler

, Wiley

, Velarde

(2015) Mitochondrial effectors of cellular senescence: Beyond the free radical theory of aging. Aging Cell 14, 1–7.

11.

Raffaello

, Rizzuto

(2011) Mitochondrial longevity pathways. Biochim Biophys Acta 1813, 260–268.

12.

Swerdlow

, Koppel

, Weidling

, Hayley

, Ji

, Wilkins

(2017) Mitochondria, cybrids, aging, and Alzheimer’s disease. Prog Mol Biol Transl Sci 146, 259–302.

13.

Kozhevnikova

, Korbolina

, Stefanova

, Muraleva

, Orlov

, Kolosova

(2013) Association of AMD-like retinopathy development with an Alzheimer’s disease metabolic pathway in OXYS rats. Biogerontology 14, 753–762.

14.

Stefanova

, Muraleva

, Skulachev

, Kolosova

(2014) Alzheimer’s disease-like pathology in senescence-accelerated OXYS rats can be partially retarded with mitochondria-targeted antioxidant SkQ1. J Alzheimers Dis 38, 681–694.

15.

Stefanova

, Kozhevnikova

, Vitovtov

, Maksimova

, Logvinov S

, Rudnitskaya

, Korbolina

, Muraleva

, Kolosova

(2014) Senescence-accelerated OXYS rats: A model of age-related cognitive decline with relevance to abnormalities in Alzheimer disease. Cell Cycle 13, 898–909.

16.

Stefanova

, Muraleva

, Korbolina

, Kiseleva

, Maksimova

, Kolosova

(2015) Amyloid accumulation is a late event in sporadic Alzheimer’s disease-like pathology in nontransgenic rats. Oncotarget 6, 1396–413.

17.

Loshchenova

, Sinitsyna

, Fedoseeva

, Stefanova

, Kolosova

(2015) Influence of antioxidant SkQ1 on accumulation of mitochondrial DNA deletions in the hippocampus of senescence-accelerated OXYS rats. Biochemistry 80, 596–603.

18.

Sergeeva

, Bagryanskaya

, Korbolina

, Kolosova

(2006) Development of behavioural dysfunctions in accelerated-senescence OXYS rats is associated with early postnatal alterations in brain phosphate metabolism. Exp Gerontol 41, 141–150.

19.

Mailloux

, Harper

M-E

(2011) Uncoupling proteins and the control of mitochondrial reactive oxygen species production. Free Radic Biol Med 51, 1106–1115.

20.

Stefanova

, Fursova

, Kolosova

(2010) Behavioral effects induced by mitochondria-targeted antioxidant SkQ1 in Wistar and senescence-accelerated OXYS rats. J Alzheimers Dis 21, 479–491.

21.

Panov

(2013) Practical Mitochondriology: Pitfalls and Problems in Studies of Mitochondria, CreateSpace Independent Publishing Platform.

22.

Panov

, Schonfeld

, Dikalov

, Hemendinger

, Bonkovsky

, Brooks

(2009) The neuromediator glutamate, through specific substrate interactions, enhances mitochondrial ATP production and reactive oxygen species generation in nonsynaptic brain mitochondria. J Biol Chem 284, 14448–14456.

23.

Kolosova

, Tyumentsev

, Muraleva

, Kiseleva

, Vitovtov

, Stefanova

(2017) Antioxidant SkQ1 alleviates signs of Alzheimer’s disease-like pathology in old OXYS rats by reversing mitochondrial deterioration. Curr Alzheimer Res 14, 1283–1292.

24.

Prachař

(2003) Intimate contacts of mitochondria with nuclear envelope as a potential energy gateaway for nucleo-cytoplasmic mRNA transport. Gen Physiol Biophys 22, 525–534.

25.

Dzeja

, Terzic

(2003) Phosphotransfer networks and cellular energetics. J Exp Biol 206, 2039–2047.

26.

Santos

, Correia

, Zhu

, Smith

, Moreira

, Castellani

, Nunomura

, Perry

(2013) Mitochondrial DNA oxidative damage and repair in aging and Alzheimer’s disease. Antioxid Redox Signal 18, 2444–2457.

27.

Chen

, Gong

, Stice

, Knowlton

(2009) Mitochondrial OPA1, apoptosis, and heart failure. Cardiovasc Res 84, 91–99.

28.

Kim

, Sun

(2011) Programmed cell death during postnatal development of the rodent nervous system. Dev Growth Differ 53, 225–235.

29.

Walton

, Shin

, Tajinda

, Heusner

, Kogan

, Miyake

, Chen

, Tamura

, Matsumoto

(2012) Adult neurogenesis transiently generates oxidative stress. PLoS One 7, e35264.

30.

Kolosova

, Shcheglova

, Sergeeva

, Loskutova

(2006) Long-term antioxidant supplementation attenuates oxidative stress markers and cognitive deficits in senescent-accelerated OXYS rats. Neurobiol Aging 27, 1289–1297.

31.

Desler

, Lillenes

, Tønjum

, Rasmussen

(2017) The role of mitochondrial dysfunction in the progression of Alzheimer’s disease. Curr Med Chem 24. doi: 10.2174/0929867324666170616110111

32.

Srivastava

(2017) The mitochondrial basis of aging and age-related disorders. Genes (Basel) 8, 398.

33.

Men’shchikova

, Shabalina

, Zenkov

, Kolosova

(2002) Generation of reactive oxygen species by mitochondria in senescence-accelerated OXYS rats. Bull Exp Biol Med 133, 175–177.

34.

Korshunov

, Skulachev

, Starkov

(1997) High protonic potential actuates a mechanism of production of reactive oxygen species in mitochondria. FEBS Lett 416, 15–18.

35.

Echtay

, Roussel

, St-Pierre

, Jekabsons

, Cadenas

, Stuart

, Harper

, Roebuck

, Morrison

, Pickering

, Clapham

, Brand

(2002) Superoxide activates mitochondrial uncoupling proteins. Nature 415, 96–99.

36.

Andrews

(2010) Uncoupling protein-2 and the potential link between metabolism and longevity. Curr Aging Sci 3, 102–112.

37.

Hirose

, Schilf

, Lange

, Mayer

, Reichart

, Maity

, Jöhren

, Schwaninger

, Scharffetter-Kochanek

, Sina

, Sadik

, Köhling

, Miroux

, Ibrahim

(2016) Uncoupling protein 2 protects mice from aging. Mitochondrion 30, 42–50.

38.

Rose

, Crocco

, de Rango

, Montesanto

, Passarino

(2011) Further support to the uncoupling-to-survive theory: The genetic variation of human UCP genes is associated with longevity. PLoS One 6, 1–7.

39.

Thangavel

, Kempuraj

, Zaheer

, Raikwar

, Ahmed

, Selvakumar

, Iyer

, Zaheer

(2017) Glia maturation factor and mitochondrial uncoupling proteins 2 and 4 expression in the temporal cortex of Alzheimer’s disease brain. Front Aging Neurosci 9, 1–10.

40.

Lores-Arnaiz

, Lombardi

, Karadayian

, Orgambide

, Cicerchia

, Bustamante

(2016) Brain cortex mitochondrial bioenergetics in synaptosomes and non-synaptic mitochondria during aging. Neurochem Res 41, 353–363.

41.

Brand

(2000) Uncoupling to survive? The role of mitochondrial inefficiency in ageing. Exp Gerontol 35, 811–820.

42.

Swerdlow

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

43.

Knott

, Perkins

, Schwarzenbacher

, Bossy-Wetzel

(2008) Mitochondrial fragmentation in neurodegeneration. Nat Rev Neurosci 9, 505–518.

44.

Sebastián

, Palacín

, Zorzano

(2017) Mitochondrial dynamics: Coupling mitochondrial fitness with healthy aging. Trends Mol Med 23, 201–215.

45.

Rowland

, Voeltz

(2012) Endoplasmic reticulum-mitochondria contacts: Function of the junction. Nat Rev Mol Cell Biol 13, 607–615.

46.

Bravo

, Vicencio

, Parra

, Troncoso

, Munoz

, Bui

, Quiroga

, Rodriguez

, Verdejo

, Ferreira

, Iglewski

, Chiong

, Simmen

, Zorzano

, Hill

, Rothermel

, Szabadkai

, Lavandero

(2011) Increased ER-mitochondrial coupling promotes mitochondrial respiration and bioenergetics during early phases of ER stress. J Cell Sci 124, 2511–2511.

47.

Area-Gomez

, Del Carmen Lara Castillo

, Tambini

, Guardia-Laguarta

, De Groof

AJC

, Madra

, Ikenouchi

, Umeda

, Bird

, Sturley

, Schon

(2012) Upregulated function of mitochondria-associated ER membranes in Alzheimer disease. EMBO J 31, 4106–4123.

48.

Hedskog

, Moreira Pinho

, Filadi

, Rönnbäck

, Hertwig

, Wiehager

, Larssen

, Gellhaar

, Sandebring

, Westerlund

, Graff

, Winblad

, Galter

, Behbahani

, Pizzo

, Glaser

, Ankarcrona

(2013) Modulation of the endoplasmic reticulum-mitochondria interface in Alzheimer’ s disease and related models. Proc Natl Acad Sci U S A 110, 7916–7921.

49.

Leal

, Schreiner

, Pinho

, Filadi

, Wiehager

, Karlström

, Pizzo

, Ankarcrona

(2016) Mitofusin-2 knockdown increases ER-mitochondria contact and decreases amyloid β-peptide production. J Cell Mol Med 20, 1686–1695.

50.

Stefanova

, Muraleva

, Maksimova

, Rudnitskaya

, Kiseleva

, Telegina

, Kolosova

(2016) An antioxidant specifically targeting mitochondria delays progression of Alzheimer’s disease-like pathology. Aging (Albany NY) 8, 2713–2733.