Regular Exercise Enhances Cognitive Function and Intracephalic GLUT Expression in Alzheimer’s Disease Model Mice

Abstract

Brain energy metabolic impairment is one of the main features of Alzheimer’s disease (AD) and is considered an underlying factor involved in cognitive impairment. Therefore, brain energy metabolism may represent a new therapeutic target for AD medical interventions. Among nutrients providing energy, glucose, the primary energy source, cannot cross the blood-brain barrier freely without specific glucose transporters (GLUTs), which are essential for the maintenance of cerebral energy metabolism homeostasis. Several converging lines of evidence suggest that GLUT1 deficiency in mice leads to synapse reduction and dysregulation coupled with mitochondrial morphological changes. In this study, the results revealed that regular exercise (RE) decreased the expression of amyloid-β and phosphorylated tau by western blot, and enhanced the spatial learning and exploration ability of AD model mice as assessed by Morris water maze test. Mitochondrial cristae and edges were clear and intact, ATP production in the brain raised, the number of synapses increased, and GLUT1 and GLUT3 expression levels improved in the central nervous system (CNS) in AD model mice after RE. Changes in GLUT1 and GLUT3 expression at the protein level after RE are an important part of energy metabolic adaptation in AD model mice. Learning and memory improvement are highly associated with mitochondrial integrity and sufficient synapses in the CNS. This research suggests that increased brain energy metabolism attributed to RE exhibits promising therapeutic potential for AD.

Keywords

Alzheimer’s disease energy metabolism glucose transporter mitochondria synapse

INTRODUCTION

Alzheimer’s disease (AD) is a central nervous system (CNS) degenerative disease that is characterized by progressive memory deficits and cognitive impairments [1]. According to statistics, there were greater than 30 million AD patients worldwide in 2010, and this number is expected to reach 106 million by 2050 [2]. However, no treatment method is available that can effectively prevent the occurrence and development of AD [3]. Thus, effective and safe treatments for AD are urgently needed.

Brain glucose metabolism dysregulation is one of the main features of AD [4, 5], and it is related to severity of AD pathology or symptom expression, such as cognitive impairment, accumulations of amyloid-β (Aβ) peptide and neurofibrillary tangles [6], reduced brain synapses, and mitochondrial dysfunction [7, 8]. Thus, the author posited the idea that glucose homeostasis in the CNS can be regarded as a potential target for disease-modifying treatments in AD. Glucose is the primary source of energy closely related to body metabolism; however, glucose must cross the cell membrane with the help of glucose transporters (GLUTs) to participate in energy metabolism [9, 10]. Simultaneously, monosaccharides, polyols, and other small carbon compounds also require members of the GLUT family to cross the membranes of eukaryotic cells in vivo [11]. GLUTs belong to the SCL2A family of sodium-independent facilitated hexose transporters. The SCL2A family includes 14 members (GLUT1–GLUT14), of which GLUT1 and GLUT3 belong to the solute carrier family 2 (SLC2) and are the two most important GLUT proteins in brain tissues [12]. GLUT1 (SCL2A1 gene) is expressed both in astrocytes and endothelial cells [13], whereas GLUT3 (SCL2A3 gene) is highly expressed in neurons [14]. Interestingly, several studies have demonstrated that the level of glucose metabolism is significantly reduced in brains of AD model mice [15]. These studies above show that the level of GLUTs is crucial for the maintenance of metabolic homeostasis in the brain. In other words, the level of GLUTs is closely related to the occurrence and development of AD symptoms. In addition, a series of observational studies have confirmed that the emergence of AD in elderly patients is closely related to unhealthy living habits, and lack of physical exercise is one of the main causes [16]. In the present work, mitochondrial integrity was observed by electron microscopy and ATP levels were analyzed by ATP Assay Kit to reveal the metabolic state in the CNS. Based on these findings, the current study aimed to reveal whether RE could alleviate cognitive impairment via increasing brain regional GLUT1- and GLUT3-mediated glucose metabolism.

MATERIALS AND METHODS

Materials and reagents

Polyclonal rabbit anti-GLUT3 (cat. no. AF5463), polyclonal rabbit anti-Aβ (cat. no. AF6084), polyclonal rabbit anti-P-Tau (cat. no. AF3148), polyclonal rabbit anti-PSD-95 (cat. no. AF5283), and polyclonal rabbit anti-Synapsin (SYN) (cat. no. AF6201) were obtained from Affinity (OH, USA), and polyclonal goat anti-GLUT3 (cat. no. sc31840) was obtained from Santa Cruz (CA, USA). Polyclonal rabbit anti-Brain Derived Neurotrophic Factor (BDNF) (cat. no. PAA011Mu01) was obtained from CLOUD-CLONE (Wuhan, China). Polyclonal rabbit anti-GLUT1 (cat. no. ab652), polyclonal mouse anti-GFAP (cat. no. 4648), and NeuN (cat. no. ab104224) were obtained from Abcam (Cambridge, UK). Polyclonal mouse anti-GLUT1 (cat. no. sc377228) was obtained from Santa Cruz. The secondary antibodies (goat anti-rabbit cat. no. ab150077; goat anti-mouse cat. no. ab97035) were obtained from Abcam (Cambridge, UK).

Animals and ethics

APP/PS1 double-transgenic mice can accurately simulate various pathological features of human AD patients, and various pathological features begin to appear in the brain of 5-6-month-old APP/PS1 double-transgenic mice. Six-month-old male APP/PS1 double-transgenic mice and age-matched wild-type littermates were purchased from Nanjing Biomedical Research Institute of Nanjing University (Nanjing, China, no. 201602397). Each mouse was housed in a plastic cage in a controlled environment (22–25°C; 50±10% relative humidity and automatic 12-h light/dark cycle) with unlimited access to food and water. Furthermore, the experiments were strictly performed in accordance with international ethical guidelines and the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Regular exercise

Mice were randomly divided into four groups based on random number tables (n = 25 each group): wild-type mice group (WT-NT), wild-type mice group with regular exercise (WT-T), APP/PS1 mice group (AD-NT), and APP/PS1 mice group with regular exercise (AD-T). Groups of WT-NT and AD-NT mice with no access to exercise were used as controls. Mice in the AD-T and WT-T groups received swim training for 1 h six days/week in a pool with sufficient width for a total of four weeks. The pool was filled with water (21–23°C) to a depth of 40 cm. If a mouse stopped moving, the intervention was provided to allow the mouse to continue its exercise until 1 h of swimming was complete.

Morris water maze test

After 4 weeks of regular exercise, cognitive function was tested using the Morris water maze apparatus (Zhengzhou University, Henan, China), which is a stainless-steel circular tank with a diameter of 120 cm and a height of 50 cm. The tank was filled with water (21–23°C) to a depth of 40 cm and divided into four equal quadrants. A platform (diameter, 12 cm) was placed in the second quadrant and submerged 1 cm below the water surface. For the spatial memory learning test, mice were trained for six days. Each trial was started by placing mice in one of the four quadrants. Mice were allowed to swim in the pool for a 60-s period to find the hidden platform. If a mouse did not find the platform within 60 s, it would be removed from the water and placed by researchers on the platform for 15 s. On the seventh day, the probe trail test was performed to assess spatial reference learning and memory, and the platform was removed. Each mouse was allowed to swim freely for 60 s. The frequency that each mouse crossed the position where the platform was once placed and the time that it spent in the target quadrant was recorded.

Electron microscopy

Slices that were 50–70 nm thick were obtained using an ultra-thin microtome for electron microscopy. The slices were dyed with 2% uranyl acetate for 25 min in the dark. The slices were washed 5 times quickly, and then the slices were placed on filter paper to dry for 10 min. The slices were dyed with 0.3% lead citrate for 5 min in the dark and washed 5 times quickly. Sections were observed using an HT7700 electron microscope (HITACHI, Japan). Photographs of the hippocampus and cortex were obtained at random. The average amount of synapses from three slices for each mouse was used for statistical analysis.

ATP levels assays

ATP was measured using an ATP Assay kit (S0026B, Bwyotime, China) according to the manufacturer’s instructions. It was based on the fact that luciferase catalyzes the formation of light from luciferin need the energy provided by ATP, and the emitted light intensity is linearly related to the ATP concentration. Briefly, the cortex and hippocampus of mice were homogenized in lysis buffer (7 μL/mg) and centrifuged at 12000 g for 10 min at 4°C to collect the cell supernatant. An aliquot (100 μL) of an ATP detection working solution was added to each well of a black 96-well plate. After incubating the plate for 5 min at room temperature, 30 μL samples of the collect cell supernatant were added to the wells, and the luminescence was measured immediately using Synergy2 (BioTek, USA). The relative ATP level was normalized to the total protein according to the following formula: ATP level = ATP concentration/protein concentration. The protein concentration of the sample was measured at 562 nm with Synergy2 (BioTek, USA).

Western blotting

Five mice were randomly selected from each group for the protein level experiment. Protein was extracted from the hippocampus and cortex and homogenized in radio-immunoprecipitation assay buffer (cat. no. CW2334S, CWBIO, China). The protein concentration was determined using the bicinchoninic acid assay (cat. no. CW0014S, CWBIO), and a total of 30 μg protein was separated by electrophoresis on 10% SDS-PAGE gels (cat. no. CW0022, CWBIO). Specific parameters include the use of a 3% concentrated gel (80 V, 30 min) and 10% separated gel (120 V, 60 min), and a 10 to 245 KD molecular weight standard protein marker (cat. no. PR1920, Solarbio, China) was used for electrophoresis. Proteins were then transferred onto polyvinylidene fluoride membranes (cat. no. IPVH00011, Millipore, USA). The membranes were blocked for 2 h with 5% skimmed milk at room temperature and incubated with antibodies against GLUT1 (1:200), GLUT3 (1:200), Aβ (1:500), P-Tau (1:500), PSD-95 (1:500), SYN (1:1000), BDNF (1:500), and β-actin (1:10000) at 4°C overnight followed by HRP-conjugated secondary antibody (1:500) incubation for 1 h. The protein bands were visualized with enhanced chemiluminescence (cat. no. PE0010, Solarbio) and imaged with the Bio-Image Analysis system (Zhengzhou University, Henan, China). The ratios of protein band intensities to β-actin were determined.

Immunofluorescence staining

Briefly, 25 μm coronal slices of brain tissue were obtained in a freezing microtome (cat. no. CM1520, Leica, Germany) for immunofluorescence. The slices were blocked for 2 h with PBS with 10% fetal bovine serum and 0.3% Triton X-100 at room temperature and then incubated with anti-GLUT1 (dilution of 1:300), anti-GLUT3 (dilution of 1:200), anti-GFAP (dilution of 1:200), and anti-NeuN (dilution of 1:500) primary antibodies overnight at 4°C. On the next day, the slices were incubated with secondary antibodies conjugated to Alexa Fluor 488 (dilution of 1:200, ab150077, Abcam, USA) and Cy3 (dilution of 1:200, ab97035, Abcam, USA) fluorophores for 2 h at room temperature in the dark. The tissue images were captured using a confocal fluorescence microscope (LSM710, Carl Zeiss, Germany). The average labelled positive area (positive area/ total area of field) from three slices for each mouse with three fields of view/slice was used for statistical analysis.

Statistical analysis

The results for each group are expressed as the means±SEM. All statistical analyses were conducted using SPSS 21.0 software (SPSS IBM, Armonk, NY, USA). The latency performance was analyzed using repeated measures analysis of variance (RM-ANOVA) across days. Probe trial time in the target quadrant compared with time in other quadrants and the times of passing the hidden platform position were assessed using one-way ANOVA. LSD-t tests were used to analyze western blotting, Electron microscopy and ATP levels assays results, and unpaired Student’s t-tests were employed to determine intergroup differences in immunostaining. The significance level was set at ^*p < 0.05, ^**p < 0.01, or ^***p < 0.001.

RESULTS

RE reduced Aβ and P-Tau expression in the brain of AD model mice

The expression of Aβ and P-Tau in the cortex and hippocampus of AD mice were quantified by western blot. The AD-NT group showed a significant increase in Aβ and P-Tau expression in the cortex (^***p < 0.001, Fig. 1A-C) and hippocampus (^***p < 0.001, Fig. 1D-F); however, after RE, the expression of Aβ and P-Tau of the AD-T group were decreased in the cortex (^##p < 0.01, Fig. 1A-C) and hippocampus (^#p < 0.05, Fig. 1D-F) compared with the AD-NT group. These results suggested that RE treatment could delay the pathological features of AD mice.

Fig.1

The expression Aβ and P-Tau. Representative immunoblots of Aβ and P-Tau in the cortex (A-C) and hippocampus (D-F) in each group. Protein immunoreactivity was normalized to β-actin. Individual data are presented as the mean±S.E.M. from four individual mice in each group. (^*p < 0.05, ^**p < 0.01, ^***p < 0.001 compared with the WT-NT group; ^#p < 0.05, ^##p < 0.01 compared with the AD-NT group; each group, n = 5).

RE improved learning and memory ability in AD model mice

The effects of RE treatment on spatial reference learning and memory ability were evaluated. In the space navigation trials, navigation paths of training provided evidence that the learning ability of the AD-T group was similar to that of WT-NT group and increased compared with the AD-NT group. As shown in Fig. 2A and B, the escape latency and total distance in all groups exhibited a downward trend in the T/training group compared with the NT/non-training group in both WT and AD model mice groups. This finding demonstrated that the T/training group showed better spatial reference learning performance (shorten the escape latency and total distance) than the NT/non-training group (^*p < 0.05, ^#p < 0.05 in Fig. 2A, B). During the hidden platform test, the representative navigation paths at the last day of testing showed that the frequency of passing the hidden platform position was increased in the AD-T group compared with the AD-NT group (^#p < 0.05, Fig. 2C, D). Simultaneously, compared with the AD-NT group, the AD-T group spent more time in the target quadrant than other quadrants. These results suggested that RE treatment obviously attenuated cognitive impairments in AD model mice.

Fig.2

The detection of learning and memory ability. The escape latency (A) and total distance (B) during orientation navigation in the MWM test. C) Representative navigation traces on Day 7 for the MWM test. D) The times of crossing the hidden platform position. (^*p < 0.05, ^**p < 0.01, ^***p < 0.001 compared with the WT-NT group; ^#p < 0.05, ^##p < 0.01, ^## #p < 0.001 compared with the AD-NT group; each group, n = 25).

RE increased synapse density in the brain of AD model mice

Electron microscopy technology was used to further verify the effects of regular exercise on synapse density in AD brain. The density of synapses in the cortical area (^***p < 0.001, Fig. 3A, B, Arrow: Synapse) and hippocampus (^***p < 0.001, Fig. 3C, D) in the AD-NT group were decreased obviously compared with the WT-NT group, whereas the number of synapses were increased in the cortex (^##p < 0.01, Fig. 3A, B) and hippocampus region (^#p < 0.05, Fig. 3C, D) in the AD-T group compared with the AD-NT group. The expressions of PSD-95, SYN, and BDNF in the cortex are shown in Fig. 4. The expression levels of PSD-95, SYN, and BDNF in the AD-NT group were significantly lower than those of the WT-NT group in cortex (^**p < 0.01, Fig. 4A-D) and hippocampus (^**p < 0.01, ^***p < 0.001, Fig. 4E-H); after RE, the AD-T group showed significantly higher expression levels of PSD-95, SYN, and BDNF than those of the AD-NT group in cortex (^#p < 0.05, Fig. 4A-D) and hippocampus (^#p < 0.05, ^##p < 0.01, Fig. 4E-H), but there was no statistical significance between WT-NT and WT-T group. These results suggested that RE treatment obviously enhanced synaptic plasticity and synaptic density in the cortex and hippocampus of AD model mice.

Fig.3

Synapse density. A) Synapse density in the cortex were observed by electron microscopy. Arrow: synapse. B) The average number of synapses from three slices for each mouse with three fields of view/slice was used for statistical analysis. (^*p < 0.05, ^***p < 0.001 compared with the WT-NT group; ^##p < 0.01 compared with the AD-NT group; each group, n = 5). C) Synapse density in the hippocampus were observed by electron microscopy. Arrow: synapse. D) The average number of synapses from three slices for each mouse with three fields of view/slice was used for statistical analysis. (^***p < 0.001 compared with the WT-NT group; ^#p < 0.05 compared with the AD-NT group; each group, n = 5). Scale bar, 2 μm.

Fig.4

The expression of PSD-95, SYN, and BDNF. Representative immunoblots of PSD-95, SYN, and BDNF in the cortex (A-D) and hippocampus (E-H) in each group. Protein immunoreactivity was normalized to β-actin. Individual data are presented as the mean±S.E.M. from 4 individual mice in each group. (^**p < 0.01, ^***p < 0.001 compared with the WT-NT group; ^#p < 0.05, ^##p < 0.01 compared with the AD-NT group, each group, n = 5).

RE ameliorated mitochondrial integrity and ATP levels in the brain of AD model mice

Increasing evidence has shown that mitochondrial dysfunction is one of the earliest pathological features. Electron microscopy technology was used to further verify the effects of regular exercise on mitochondrial integrity in AD brain, mitochondrial edges were clear, and mitochondrial cristae were dense and integral in the WT-NT group (Fig. 5A, D, Arrowhead: Mitochondria), while the mitochondria in AD-NT group had blurred edges and broken mitochondria cristae (Fig. 5B, E). The edges of the mitochondria became clear and the mitochondrial cristae became dense and integral in the AD-T group (Fig. 5C, F), and regular exercise improved mitochondrial dysfunction in the AD-T group. Simultaneously, the ATP levels in the AD-NT group were significantly lower than the WT-NT group in cortex (^***p < 0.001, Fig. 5H) and hippocampus (^***p < 0.001, Fig. 5G); after RE, the AD-T group showed significantly higher ATP than the AD-NT group in cortex (^##p < 0.01, Fig. 5 H) and hippocampus (^##p < 0.01, Fig. 5G). These results proved that regular exercise ameliorated energy metabolism in brains.

Fig.5

Mitochondrial integrity and ATP levels assays. A, D) Mitochondrial integrity in the cortex were observed by electron microscopy in WT-NT group. Arrowhead: mitochondria. B, E) Mitochondrial integrity in the cortex were observed by electron microscopy in AD-NT group. Arrowhead: mitochondria. C, F) Mitochondrial integrity in the cortex were observed by electron microscopy in AD-T group. Arrowhead: mitochondria. H) ATP levels detected with ATP kit in cortex (^*p < 0.05, ^**p < 0.01, ^***p < 0.001 compared with the WT-NT group; ^##p < 0.01 compared with the AD-NT group; each group, n = 5). G) ATP levels detected with ATP kit in hippocampus (^***p < 0.001 compared with the WT-NT group; ^##p < 0.01 compared with the AD-NT group; each group, n = 5). The data were presented with mean±SEM. (Scale bar, A, B, C 1 μm; D, E, F 500 nm).

RE increased GLUT1 and GLUT3 expression in the brain of AD model mice

Two regions related to learning and memory, namely, the hippocampus and cortex, were selected to assess whether RE affects GLUT1 and GLUT3 expression in these areas. The results showed that GLUT1 and GLUT3 expression was significantly reduced in the hippocampus and cortex in the AD-NT group compared with the WT-NT group (^*p < 0.05, ^***p < 0.001, Fig. 6A-D). Furthermore, increased GLUT1 and GLUT3 expression was evident in the hippocampus and cortex of the AD-T group compared with the AD-NT group (^#p < 0.05, ^##p < 0.01, Fig. 6A-D). These results revealed that RE enhanced GLUT1 and GLUT3 expression in AD model mice in the cortex and hippocampus.

Fig.6

GLUT1 and GLUT3 expression. Representative immunoblots of GLUT1 (A, B) and GLUT3 (C, D) in the cortex and hippocampus in each group. Protein immunoreactivity was normalized to β-actin. Individual data are presented as the mean±S.E.M. from 5 individual mice in each group. (^*p < 0.05, ^***p < 0.001 compared with the WT-NT group; ^#p < 0.05 compared with the AD-NT group, ^##p < 0.01 compared with the AD-NT group).

Double immunofluorescence labelling of GLUT1 and GLUT3

To further detect the influence of regular exercise on GLUT1 and GLUT3 expression in AD model mice, a double immunofluorescence labelling assay was used. The positive labelling of GLUT1 and GLUT3 is significantly reduced in AD model mice compared with WT mice (^*p < 0.05, ^**p < 0.01, ^***p < 0.001, Figs. 7 and 8, Supplementary Figures 1–4), and the results showed that the positive areas of GLUT1 and GLUT3 were increased in the cortex of the AD-T group compared with the AD-NT group (^#p < 0.05, Figs. 7 and 8). In addition, these areas were also increased in the hippocampus of the AD-T group compared to the AD-NT group (^#p < 0.05, ^##p < 0.01, Supplementary Figures 1–4). Furthermore, these results suggested that regular exercise ameliorated the expression of GLUT1 and GLUT3, which mediate energy metabolism in AD model mice.

Fig.7

GLUT1 and GFAP co-expression in the cortex. A) Representative immunofluorescence of GLUT1 and GFAP in the cortex in each group. GLUT1-positive cells are green. GFAP-positive cells are red. GLUT1 and GFAP double-positive cells are yellow. B) GLUT1-positive labelled areas were analyzed in cortex. Individual data are presented as the mean±S.E.M. from 5 individual mice in each group. (^*p < 0.05, ^**p < 0.01 compared with the WT-NT group; ^#p < 0.05 compared with the AD-NT group; each group, n = 5). Scale bar, 100 μm.

Fig.8

GLUT3 and NeuN co-expression in the cortex. A) Representative immunofluorescence of GLUT3 and NeuN in the cortex in each group. GLUT3-positive cells are green. NeuN-positive cells are red. GLUT3 and NeuN double-positive cells are yellow. B) GLUT3-positive labelled areas were analyzed in cortex. Individual data are presented as the mean±S.E.M. from 5 individual mice in each group. (^***p < 0.001 compared with the WT-NT group; ^#p < 0.05 compared with the AD-NT group; each group, n = 5). Scale bar, 100 μm.

DISCUSSION

AD is characterized by a progressive decline in cognitive and physical function and causes loss of independence [17]. The disease is costly to families, individuals, and the health care system [18, 19]. To date, medical treatment has demonstrated limited efficacy, and more attention is increasingly given to nonpharmacological interventions to improve the occurrence and development of AD [20]. In addition, unhealthy lifestyles are closely related to the incidence of AD, in which lack of exercise is one of the most important pathogens for AD patients. [21]. Moreover, physical exercise contributes to cognitive improvements in AD patients [22 –24]. To date, the therapeutic efficacy of exercise for AD and the underlying mechanisms have not been completely characterized.

Several studies have confirmed that abnormal brain glucose homeostasis is an intrinsic factor involved in the pathogenesis of AD and closely related to the occurrence of clinical symptoms [25, 26], such as, AD is characterized by accumulations of Aβ peptide known as plaques and neurofibrillary tangles. It has been proposed that exercise can reduce accumulations of Aβ peptide and neurofibrillary tangles in some studies [27]. In this article, the results show that RE reduced the expression of Aβ and P-Tau in AD model mice, both in the cortex and hippocampus. The density of synapses is closely connected with brain function [28, 29], using electron microscopy, the present study viewed that the density of synapses decreased in AD model mice, including cortex and hippocampus; however, after RE treatment the density of synapses were increased in these regions. At the same time, RE also increases the expression of synaptic-related proteins, such as post synaptic density protein 95 (PSD-95), synapsin (SYN), and brain derived neurotrophic factor (BDNF). The pre- and post-synaptic markers (SYN and PSD-95) are related to the formation of synapse [30], and the decrease of SYN and PSD-95 linked to synaptic plasticity will influence recognition memory [31, 32]. BDNF has been emerged as a key mediator of cognition as it is highly expressed in the regions of the brain, such as the hippocampus and cortex, and regulates structural and functional neuronal processes, in both the short and long term [33]. Western blot was utilized to detect the expression of SYN, PSD95, and BDNF, the results show that the expression of SYN, PSD95, and BDNF decreased in cortex and hippocampus of AD model mice; however, after RE treatment, the expression of SYN, PSD95, and BDNF were increased. The changes of SYN, PSD95, and BDNF expression were consistent with changes in synaptic density observed by electron microscopy.

A lot of literature has confirmed that the glucose utilization rate in AD brain is significantly reduced [34]. The structural function of mitochondria in the brain is changed in response to dysregulated energy metabolism in the brain [35, 36], Mitochondria are maternally inherited intracellular organelles with key roles in energy metabolism and second messenger signaling. Although genetically determined mitochondriopathies are best recognized as childhood disorders and managed by pediatric specialists, mitochondrial impairment is increasingly recognized in adult-onset neurodegenerative disorders, including AD and Parkinson’s disease [37 –39]. In this study, mitochondria in the AD-NT group exhibit blurred edges and broken mitochondria cristae compared with the WT-NT group; in the AD-T group, the mitochondria cristae became denser and more complete, and mitochondrial edges became distinct. Related literature confirms that exercise can improve glycolysis and oxidative phosphorylation in AD brain [40, 41]. This phenomenon is also closely related to the improvement of mitochondrial function, mitochondrial dysfunction evidenced by the decline in pyruvate dehydrogenase (PDH) and cytochrome c oxidase (COX) regulatory enzymes was found in AD model mice, and RE enhanced the expression of COX and PDH [42 –44]. This study further found that the ATP levels were reduced in AD model mice through ATP levels assays, whether cortex or hippocampus; however, the ATP levels were increased in these regions of AD-T group. These results further demonstrate that there is energy metabolism disorder in the brain of AD model mice, while RE can ameliorate energy metabolism disorder in the brain of AD model mice.

In the nervous system, the main source of energy is aerobic metabolism of glucose [45]. The interaction among neurons, astrocytes, and endothelial cells plays an important role in the energy supply of neuronal activity [46]. Different subtypes of glucose transporters (GLUTs) are expressed in these cells to mediate the transport of glucose. GLUT1 and GLUT3 are mainly expressed in the nervous system; GLUT1 is expressed in endothelial cells and astrocytes [47], while GLUT3 is expressed in neurons [48]. Astrocytes absorb glucose from the blood and metabolize it into lactate, which is then delivered to neurons. The relative contribution of this astrocyte-to-neuron lactate shuttle as a major source of energy to maintain neuronal physiology is a controversial issue compared with direct glucose uptake by neurons; nevertheless, the role of GLUT1 in astrocytes is underscored by the range of clinical phenotypes associated with GLUT1 deficiency [49]. In this study, the density of GLUT1 and GLUT3 were decreased in the hippocampus and cortex in AD model mice compare to that expressed in the wild-type mice. There are several reviews on the biochemical characteristics of GLUTs, the role of GLUTs in astrocytes and neuronal metabolism [50, 51], the clinical consequences of GLUT1 deficiency, and the effect of GLUT1 in other neurologic disorders [52, 53]. Thus, GLUT1 and GLUT3 are critical for cerebral glucose homeostasis. Clearly, regular physical activity provides important health benefits. Behavioral deficits and plaque deposition are reduced in the human APP transgenic mouse model in response physical exercise [54]. In this study, regular exercise improved cognitive impairments in AD-T mice, particularly learning and memory ability. Moreover, increased GLUT1 and GLUT3 expressions were observed in the cortex and CA1 and CA3 of the hippocampus in AD-T model mice. Accordingly, if physical exercise is beneficial to physical and psychological health, it could also be considered an effective therapy for delaying AD progression by improving glucose metabolism.

Based on our results, we propose that the failure of glucose utilization due to impaired glycolysis is a fundamental feature of AD. At regionally specific threshold levels of brain glucose and impaired glycolytic flux, accumulating pathology in vulnerable brain regions triggers the onset of AD symptoms. Regular exercise can delay cognitive impairment by ameliorating glucose metabolism. Our results have important translational implications and set the stage for future studies that may uncover therapeutic interventions targeting brain glucose dysregulation in AD.

Footnotes

ACKNOWLEDGMENTS

This project was supported by the National Natural Science Foundation of China (No. 81401015). The authors are very grateful to the teachers and classmates who helped and supported our experiments, especially our mentor Professor Cheng Chang. The authors are deeply affected by the rigorous atmosphere in the laboratory.

Authors’ disclosures available online ().

The supplementary material is available in the electronic version of this article: .

References

Corsi

, Di Raimo

, Di Lorenzo

, Rapp-Ricciardi

, Archer

, Ricci

, Businaro

(2016) Cognitive disability in Alzheimer’s disease and its management. Clin Ter 167, e123–e126.

Norton

, Matthews

, Barnes

, Yaffe

, Brayne

(2014) Potential for primary prevention of Alzheimer’s disease: An analysis of population-based data. Lancet Neurol 13, 788–794.

Schneider

, Mangialasche

, Andreasen

, Feldman

, Giacobini

, Jones

, Mantua

, Mecocci

, Pani

, Winblad

, Kivipelto

(2014) Clinical trials and late-stage drug development for Alzheimer’s disease: An appraisal from 1984 to 2014. J Intern Med 275, 251–283.

Hunt

, Schonknecht

, Henze

, Seidl

, Haberkorn

, Schroder

(2007) Reduced cerebral glucose metabolism in patients at risk for Alzheimer’s disease. Psychiatry Res 155, 147–154.

Herholz

(2010) Cerebral glucose metabolism in preclinical and prodromal Alzheimer’s disease. Expert Rev Neurother 10, 1667–1673.

Lane

, Hardy

, Schott

(2018) Alzheimer’s disease. Eur J Neurol 25, 59–70.

Selkoe

(2002) Alzheimer’s disease is a synaptic failure. Science 298, 789–791.

Balietti

, Tamagnini

, Fattoretti

, Burattini

, Casoli

, Platano

, Lattanzio

, Aicardi

(2012) Impairments of synaptic plasticity in aged animals and in animal models of Alzheimer’s disease. Rejuvenation Res 15, 235–238.

Chen

, Cheung

, Feng

, Tanner

, Frommer

(2015) Transport of sugars. Annu Rev Biochem 84, 865–894.

10.

Benarroch

(2014) Brain glucose transporters: Implications for neurologic disease. Neurology 82, 1374–1379.

11.

Duran-Aniotz

, Hetz

(2016) Glucose metabolism: A sweet relief of Alzheimer’s disease. Curr Biol 26, R806–809.

12.

Mueckler

, Thorens

(2013) The SLC2 (GLUT) family of membrane transporters. Mol Aspects Med 34, 121–138.

13.

Winkler

, Nishida

, Sagare

, Rege

, Bell

, Perlmutter

, Sengillo

, Hillman

, Kong

, Nelson

, Sullivan

, Zhao

, Meiselman

, Wendy

, Soto

, Abel

, Makshanoff

, Zuniga

, De Vivo

, Zlokovic

(2015) GLUT1 reductions exacerbate Alzheimer’s disease vasculo-neuronal dysfunction and degeneration. Nat Neurosci 18, 521–530.

14.

Fidler

, Campbell

, Funari

, Dunne

, Balderas Angeles

, Middleton

, Chaudhuri

, Weyrich

, Abel

(2017) Deletion of GLUT1 and GLUT3 reveals multiple roles for glucose metabolism in platelet and megakaryocyte function. Cell Rep 20, 881–894.

15.

, Varma

, Casanova

, Dammer

, Pletnikova

, Chia

, Egan

, Ferrucci

, Troncoso

, Levey

, Lah

, Seyfried

, Legido-Quigley

, O’Brien

, Thambisetty

(2018) Evidence for brain glucose dysregulation in Alzheimer’s disease. Alzheimers Dement 14, 318–329.

16.

Riedel

(2014) Preventing cognitive decline in preclinical Alzheimer’s disease. Curr Opin Pharmacol 14, 18–22.

17.

Querfurth

, LaFerla

(2010) Alzheimer’s disease. N Engl J Med 362, 329–344.

18.

Mangialasche

, Solomon

, Winblad

, Mecocci

, Kivipelto

(2010) Alzheimer’s disease: Clinical trials and drug development. Lancet Neurol 9, 702–716.

19.

Hickman

, Faustin

, Wisniewski

(2016) Alzheimer disease and its growing epidemic: Risk factors, biomarkers, and the urgent need for therapeutics. Neurol Clin 34, 941–953.

20.

Ashford

, Mahoney

, Burkett

(2015) A role for complementary and integrative medicine in Alzheimer’s disease prevention. J Alzheimers Dis 48, 13–14.

21.

Winblad

, Amouyel

, Andrieu

, Ballard

, Brayne

, Brodaty

, Cedazo-Minguez

, Dubois

, Edvardsson

, Feldman

, Fratiglioni

, Frisoni

, Gauthier

, Georges

, Graff

, Iqbal

, Jessen

, Johansson

, Jonsson

, Kivipelto

, Knapp

, Mangialasche

, Melis

, Nordberg

, Rikkert

, Qiu

, Sakmar

, Scheltens

, Schneider

, Sperling

, Tjernberg

, Waldemar

, Wimo

, Zetterberg

(2016) Defeating Alzheimer’s disease and other dementias: A priority for European science and society. Lancet Neurol 15, 455–532.

22.

Hauer

, Schwenk

, Zieschang

, Essig

, Becker

, Oster

(2012) Physical training improves motor performance in people with dementia: A randomized controlled trial. J Am Geriatr Soc 60, 8–15.

23.

Vreugdenhil

, Cannell

, Davies

, Razay

(2012) A community-based exercise programme to improve functional ability in people with Alzheimer’s disease: A randomized controlled trial. Scand J Caring Sci 26, 12–19.

24.

Liu

, Zhao

, Zhang

, Shi

(2013) Long-term treadmill exercise inhibits the progression of Alzheimer’s disease-like neuropathology in the hippocampus of APP/PS1 transgenic mice. Behav Brain Res 256, 261–272.

25.

Kerti

, Witte

, Winkler

, Grittner

, Rujescu

, Floel

(2013) Higher glucose levels associated with lower memory and reduced hippocampal microstructure. Neurology 81, 1746–1752.

26.

Gardener

, Sohrabi

, Shen

, Rainey-Smith

, Weinborn

, Bates

, Shah

, Foster

, Lenzo

, Salvado

, Laske

, Laws

, Taddei

, Verdile

, Martins

(2016) Cerebral glucose metabolism is associated with verbal but not visual memory performance in community-dwelling older adults. J Alzheimers Dis 52, 661–672.

27.

Zhang

, Guo

, Wang

, Song

, Zhang

, Du

(2018) Long-term treadmill exercise attenuates Abeta burdens and astrocyte activation in APP/PS1 mouse model of Alzheimer’s disease. Neurosci Lett 666, 70–77.

28.

Sarkar

, Jun

, Rellick

, Quintana

, Cavendish

, Simpkins

(2016) Expression of microRNA-34a in Alzheimer’s disease brain targets genes linked to synaptic plasticity, energy metabolism, and resting state network activity. Brain Res 1646, 139–151.

29.

Cheng

, Edin

, Lauritzen

, Aspmodal

, Christoffersen

, Jian

, Rasmussen

, Pettersen

, Xiaoqun

, Bergersen

(2012) Alterations of monocarboxylate transporter densities during hypoxia in brain and breast tumour cells. Cell Oncol (Dordr) 35, 217–227.

30.

Sinclair

, Tayler

, Love

(2015) Synaptic protein levels altered in vascular dementia. Neuropathol Appl Neurobiol 41, 533–543.

31.

Song

, Augustine

(2015) Synapsin isoforms and synaptic vesicle trafficking. Mol Cells 38, 936–940.

32.

Wang

, Peng

(2016) Basic roles of key molecules connected with NMDAR signaling pathway on regulating learning and memory and synaptic plasticity. Mil Med Res 3, 26.

33.

Dong

, Zhao

, Lin

, Liang

, He

, Zheng

, Sui

, Zhuang

, Yan

(2018) Exercise improves recognition memory and synaptic plasticity in the prefrontal cortex for rats modelling vascular dementia. Neurol Res 40, 68–77.

34.

Liu

, Zhuo

, Li

, Jin

, Lin

, Zhang

, Liang

, Wu

, Huang

, Wang

, Lin

, Chen

, Tao

(2017) Activation of brain glucose metabolism ameliorating cognitive impairment in APP/PS1 transgenic mice by electroacupuncture. Free Radic Biol Med 112, 174–190.

35.

Cardoso

, Correia

, Carvalho

, Moreira

(2017) Mitochondria in Alzheimer’s disease and diabetes-associated neurodegeneration: License to heal!. Handb Exp Pharmacol 240, 281–308.

36.

Guo

, Tian

, Du

(2017) Mitochondrial dysfunction and synaptic transmission failure in Alzheimer’s disease. J Alzheimers Dis 57, 1071–1086.

37.

Parker

, Parker

Jr , Overman

(1990) Intracranial diffuse axonal injury at autopsy. Ann Clin Lab Sci 20, 220–224.

38.

Swerdlow

, Parks

, Miller

, Tuttle

, Trimmer

, Sheehan

, Bennett

Jr , Davis

, Parker

Jr (1996) Origin and functional consequences of the complex I defect in Parkinson’s disease. Ann Neurol 40, 663–671.

39.

Swerdlow

, Khan

(2009) The Alzheimer’s disease mitochondrial cascade hypothesis: An update. Exp Neurol 218, 308–315.

40.

Alkadhi

(2018) Exercise as a positive modulator of brain function. Mol Neurobiol 55, 3112–3130.

41.

Kinni

, Guo

, Ding

, Konakondla

, Dornbos

3rd , Tran

, Guthikonda

, Ding

(2011) Cerebral metabolism after forced or voluntary physical exercise. Brain Res 1388, 48–55.

42.

Bernardo

, Marques-Aleixo

, Beleza

, Oliveira

, Ascensao

, Magalhaes

(2016) Physical exercise and brain mitochondrial fitness: The possible role against Alzheimer’s disease. Brain Pathol 26, 648–663.

43.

Bishop

, Granata

, Eynon

(2014) Can we optimise the exercise training prescription to maximise improvements in mitochondria function and content? Biochim Biophys Acta 1840, 1266–1275.

44.

, Kang

, Jiang

, Wang

, Zhang

, Ji

(2014) Exercise-induced neuroprotection of hippocampus in APP/PS1 transgenic mice via upregulation of mitochondrial 8-oxoguanine DNA glycosylase. Oxid Med Cell Longev 2014, 834502.

45.

Homenko

, Susin

, Kataeva

, Irishina

, Zavolokov

(2017) Characteristics of cerebral glucose metabolism in patients with cognitive impairment in Parkinson’s disease. Zh Nevrol Psikhiatr Im S S Korsakova 117, 46–51.

46.

Hubener

, Bonhoeffer

(2014) Neuronal plasticity: Beyond the critical period. Cell 159, 727–737.

47.

Shiohama

, Fujii

, Takahashi

, Nakamura

, Kohno

(2013) Reversible white matter lesions during ketogenic diet therapy in glucose transporter 1 deficiency syndrome. Pediatr Neurol 49, 493–496.

48.

Simpson

, Dwyer

, Malide

, Moley

, Travis

, Vannucci

(2008) The facilitative glucose transporter GLUT3: 20 years of distinction. Am J Physiol Endocrinol Metab 295, E242–E253.

49.

Nagele

, D’Andrea

, Lee

, Venkataraman

, Wang

(2003) Astrocytes accumulate A beta 42 and give rise to astrocytic amyloid plaques in Alzheimer disease brains. Brain Res 971, 197–209.

50.

Caspersen

, Wang

, Yao

, Sosunov

, Chen

, Lustbader

, Xu

, Stern

, McKhann

, Yan

(2005) Mitochondrial Abeta: A potential focal point for neuronal metabolic dysfunction in Alzheimer’s disease. FASEB J 19, 2040–2041.

51.

Braak

, Del Trecidi

(2015) Neuroanatomy and pathology of sporadic Alzheimer’s disease. Adv Anat Embryol Cell Biol 215, 1–162.

52.

Allen

, Messier

(2013) Plastic changes in the astrocyte GLUT1 glucose transporter and beta-tubulin microtubule protein following voluntary exercise in mice. Behav Brain Res 240, 95–102.

53.

Choeiri

, Staines

, Miki

, Seino

, Messier

(2005) Glucose transporter plasticity during memory processing. Neuroscience 130, 591–600.

54.

Koo

, Kang

, Oh

, Yang

, Cho

(2017) Treadmill exercise decreases amyloid-beta burden possibly via activation of SIRT-1 signaling in a mouse model of Alzheimer’s disease. Exp Neurol 288, 142–152.