Skip to main content
Download PDF

Gene therapy in Aβ-induced cell and mouse models of Alzheimer’s disease through compensating defective mitochondrial complex I function

Journal of Translational Medicine volume 22, Article number: 760 (2024) Cite this article

Abstract

Background

Alzheimer’s disease (AD) is the most common neurogenerative disorder without effective treatments. Defects in mitochondrial complex I are thought to contribute to AD pathogenesis. The aim of this study is to explore whether a novel gene therapy transducing yeast complex I gene NDI1 can be used to treat AD with severely reduced complex I function in cell and animal models.

Methods

The differentiated human neural cells were induced by Aβ1–42 to establish the AD cell model, and adeno-associated virus serotype 9 (AAV9) was used to transduce yeast NDI1 into the cell model. Aβ1–42 was injected into the hippocampus area of the brain to establish the AD mouse model. AAV9-NDI1 was injected stereotaxically into the hippocampus area to test the therapeutic effect.

Results

The expressed yeast complex I had an ameliorating effect on the defective function of human complex I and cellular pathological characteristics in the AD cell model. Furthermore, AAV9-NDI1 gene therapy in the hippocampus had a therapeutic effect on various aspects of mitochondrial function, histopathological characteristics and neurological defects in the AD mouse model. In addition, AAV9-NDI1 injection into the hippocampus of normal mice did not cause any adverse effect.

Conclusions

Compensating mitochondrial complex I function with yeast NDI1 is effective for gene therapy in Aβ-induced AD cell and mouse models. The results of this study offer a novel strategy and approach for treating AD types characterized by complex I abnormalities.

Introduction

Alzheimer’s disease (AD) is the most common dementia type [1,2,3] without effective treatments. Its clinical manifestations mainly include progressive memory loss and cognitive function decline. Some AD patients are accompanied by mental and behavioral abnormalities [4]. Late-stage patients lose their ability to take care of themselves and impose heavy burdens on their families and society. AD is a neurodegenerative disease. Its pathological characteristics mainly include extracellular plaques formed by deposition of β-amyloid (Aβ) and intraneuronal fibrillary tangles formed by hyperphosphorylation of tau protein, accompanied by neuroinflammation, mitochondrial abnormalities, synaptic function loss, and neuronal loss [5]. Since the cause of AD is complex, there are various competing hypotheses. The Aβ cascade hypothesis is not comprehensive [6, 7], while the mitochondrial cascade hypothesis is based on early-occurring mitochondrial abnormalities [8,9,10,11]. Aβ pathology, Tau pathology, and defects in mitochondrial complexes may influence each other, forming a vicious cycle [12]. Different AD patients have different driving causes and should be treated individually [6, 9]. Functional defect of the mitochondrial respiratory chain complex I is the main cause of disease in some AD patients [13, 14]. Currently, there is no gene therapy for AD that can alternatively compensate the function of complex I.

Yeast complex I consists of a single subunit, NDI1, whereas mammalian (including humans) complex I has 45 subunits. Our previously published paper [15] showed that transduction of the yeast NDI1 gene can functionally compensate the defective complex I induced by rotenone in human neural cells. Especially for the rotenone-induced Parkinson’s disease mouse model, the expressed yeast complex I had therapeutic effects according to improved histopathology and neurobehaviors. In addition, AAV5-NDI1 gene therapy in the substantia nigra area of the brain does not cause any adverse effect in mouse [15].

In this study, Aβ1–42 was used to induce the differentiated human neural cells to establish the AD cell model. The recombinant adeno-associated virus serotype 9 (AAV9-NDI1) was used to study the effect of yeast complex I on various functions of human complex I and cytopathological characteristics in the AD cell model. Furthermore, we studied the rehabilitating effects of NDI1 gene therapy in the hippocampus of the brain on various aspects of mitochondrial functions, histopathological characteristics and neurological features using an AD mouse model that was established by injecting Aβ1–42 into the hippocampus of the brain. The results of our study show that gene therapy with yeast NDI1 is a promising approach to treat AD with complex I abnormalities.

Materials and methods

Preparation of Aβ1–42

According to the manufacturer’s instruction, human Aβ1–42 (Sigma, USA) was dissolved in DMSO to prepare a 5 mM stock solution and stored at -80℃. To generate the working solution containing the oligomeric form of Aβ1–42, the stock solution was diluted to 134 µM in 0.9% NaCl and incubated at 4 °C for 24 h.

Packaging of recombinant adeno-associated virus

HA tagged NDI1, with the HA tag inserted behind the mitochondrial targeting sequence of NDI1, was subcloned into AAV9 vector, generating the rAAV9-NDI1 plasmid. rAAV9-NDI1 and AAV9-vector viruses with the titer of ~ 2 × 1012 vg/mL were packaged by Wuhan Primus Brain Science and Technology Co., Ltd.

Cell culture, establishment of Aβ1-42-induced AD cell model, and transduction of recombinant adeno-associated virus

Human neuronal cells SH-SY5Y were purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA) and cultured in Dulbecco’s modified Eagle’s medium (DMEM) (Gibco, California, USA) supplemented with 10% fetal bovine serum (FBS) (Gibco) under 5% CO2 condition.

SH-SY5Y (2 × 105) cells were seeded in a 6-well plate, and treated with 10 µM trans-retinoic acid (trans-RA) (Sigma, St Louis, MO, USA) to induce cell differentiation by replacing with fresh medium containing 10 µM trans-RA every 2 days. After treated with trans-RA for 24 h, cells were transduced with AAV9-vector or rAAV9-NDI1 viral particles with a multiplicity of infection (MOI) of 4 × 104. On the 8th day of trans-RA treatment, cells were treated with DMSO or 10 µM Aβ1–42 for 24 h before harvested. Experiments using AD cell model were set up in triplicate, in 3 groups including DMSO + vector (normal), Aβ1–42 + vector (model), and Aβ1–42 + NDI1 (therapy) groups.

Establishment of Aβ1-42-induced AD mouse model and stereotaxic injection of recombinant adeno-associated virus

Male C57BL/6 mice (6–8 weeks old) were purchased from Shanghai Slack Experimental Animal Co., Ltd. The mouse experiments in this study had been approved by the Institutional Animal Care & Use Committee of Wenzhou Medical University (Approval No.: wydw2022-0236).

On day 0, mice were anesthetized by 1.5% isoflurane inhalation and fixed in a stereotaxic apparatus (Stoelting, Wood Dale, IL, USA). AAV9-vector or rAAV9-NDI1 viral particles (3 µL, titer ~ 2 × 1012 vg/mL) was injected into the right hippocampus (posterior to bregma − 1.7 mm, lateral to midline − 1.0 mm, subdural − 1.5 mm). Two weeks after AAV9 injection (on day 14), DMSO or Aβ1–42 (400 pmol) was injected into the right hippocampus (coordinates were the same as above). Behavioral experiments were conducted 2 weeks after injection of Aβ1–42 (on day 28). On day 35, mice were euthanized to dissect out brain tissues for subsequent analyses. Experiments using AD mouse model were set up in 3 groups including DMSO + vector (normal), Aβ1–42 + vector (model), and Aβ1–42 + NDI1 (therapy) groups. There were at least 3 mice in each group for each experiment.

Determination of ATP content

The cellular ATP content was measured by using a luciferin/luciferase chemiluminescence ATP determination kit (TermoFisher Scientific, USA) as previously described [15]. Briefly, cells (1 × 106) were lysed and clarified by centrifugation. The supernatant was mixed with the standard reaction solution, and the chemiluminescence was detected. In addition to determining the basal ATP synthesis (Base), cells were treated with oligomycin to determine the ATP synthesis after the inhibition of ATP synthase (Oligo-resistant).

Determination of ROS content

Approximately 2.5 × 105 cells were collected and incubated in 500 µL of 5 µM MitoSOX™ Red (mitochondrial superoxide indicator, TermoFisher Scientific, USA) in the dark at 37 °C for 25 min. After washing with PBS, the cell pellet was resuspended, and 5000 cells were analyzed by flow cytometry (BD Biosciences, USA) to obtain the median fluorescence intensity (MFI) of each sample.

Novel object recognition test

The novel object recognition test was performed according to the published protocol [16]. In the habituation phase, each mouse was allowed to explore freely for 5 min in a test box. In the training phase, each mouse was allowed to explore for 5 min in the test box with two identical objects (A1, A2) placed in the opposite areas, and the cumulative exploration time for each object was recorded. In the testing phase (1 h after the training phase), each mouse was placed in the test box with the old object (A2) replaced with a novel object (B2), and the cumulative exploration time for each object in 5 min was recorded. There were 6–8 mice in each group. Analysis was performed using SMART software (Pan Lab, Barcelona, Spain).

Morris water maze test

Morris water maze test was performed according to the published protocol [17]. On day 1–5, with a small platform hidden in the South/West (SW) quadrant, a spatial acquisition test (hidden platform trial) was conducted. The time for each mouse to find the hidden platform in 1 min was recorded, with each mouse staying on the hidden platform for at least 10 s. On the 6th day, a reference memory test (probe trial) was performed after the hidden platform was withdrawn. The entry position for each mouse was the North/East quadrant, opposite to the target quadrant (SW) of the original position of the hidden platform. The cumulative time for each mouse passing through the target quadrant (SW), and the number of times for each mouse passing the original position of the hidden platform in 1 min, were recorded. There were 6–8 mice in each group. A camera was used to record the swimming trajectory of the mice, and the DigBehv animal behavior software system was used for analysis.

Western blot analysis

Mice were anesthetized with 1.5% isofurane inhalation. Fresh brain tissues encompassing the hippocampus were coronally dissected using mouse brain mold, then fresh hippocampus tissues were dissected according to mouse brain map. Western blot analysis was performed as described [15]. Briefly, cells, or fresh hippocampus tissues from mice (at least 3 per group), were homogenized, lysed, and centrifuged. The supernatants (lysates) were separated on 10% SDS-PAGE and transferred to PVDF membrane. The membrane blots were blocked and incubated with primary antibodies including anti-HA (1:1000, Cell Signaling Technology), anti-p-Tau (Ser202, Thr205) (1:500, TermoFisher Scientific), anti-p-Tau (Thr181) (1:1000, TermoFisher Scientific), anti-Tau (1:500, TermoFisher Scientific), anti-LC3B (1:1000, Cell Signaling Technology), anti-P62 (1:1000, Cell Signaling Technology), anti-LAMP1 (1:100, Santa Cruz Biotechnology), anti-β-actin (1:2000, OriGene Technologies), anti-GFP (1:1000, Beyotime Institute of Biotechnology) or anti-GAPDH (1:5000, Proteintech, Wuhan, China) respectively, overnight at 4 ℃. Subsequently, the membranes were incubated with appropriate secondary antibodies, and developed by ECL reagent. The gray value for each band was analyzed using ImageJ software.

Immunofluorescent staining

Immunofluorescent staining was performed as previously described [15]. Briefly, cells cultured on coverslips were incubated with 100 nM Mito-Tracker Red CMXRos for 30 min at 37 ℃, fixed, permeabilized, and blocked. The cells were subsequently incubated with anti-HA antibody (1:100, Cell Signaling Technology) overnight at 4 ℃, with secondary antibody, and stained with DAPI. Image acquisition was performed using a laser scanning confocal microscope. Mice (at least 3 for each group) were anesthetized with 1.5% isofurane inhalation and transcardially perfused with 4% paraformaldehyde. Brain tissues encompassing the hippocampus were coronally dissected using mouse brain mold, fixed, embedded, and sectioned. Briefly, the hippocampus tissue sections were pretreated for antigenic retrieval, permeabilized, blocked, and incubated with anti-HA (1:100; Cell Signaling Technology) or anti-Tom (1:400; Proteintech, Wuhan, China) antibody, overnight at 4 ℃. The subsequent steps were the same as the immunofluorescent staining protocol for cultured cells.

Detection of complex I-dependent oxygen consumption or total oxygen consumption

The oxygen consumption of cells or hippocampus tissue mitochondria was measured using the Oxygraph-2k cellular respiration apparatus. For detecting complex I-dependent oxygen consumption, 2 × 106 cells were placed in 2 ml of respiratory buffer (20 mM Hepes, 250 mM sucrose, 2 mM KH2PO4, 10 mM MgCl2, 1 mM ADP, pH = 7.1 ), and transferred into the detection chamber. The cells were permeabilized with 2% digitonin, and mixed with complex I substrate (1 M glutamate and 1 M malic acid, 4:1) to record complex I-dependent oxygen consumption. Subsequently, rotenone at the final concentration of 100 µM was added to record the oxygen consumption after inhibiting human complex I. Lastly, flavonoid at the final concentration of 0.225 M was added to record the oxygen consumption after further inhibiting yeast complex I. Total oxygen consumption measurement was performed as previously described [15]. Briefly, cells (5 × 106) were resuspended and transferred to the detection chamber. The level of basal oxygen consumption was recorded (Base). Then, after oligomycin treatment, the oxygen consumption level (Oligo-resistant) was recorded. Lastly, after FCCP treatment, the uncoupling, maximal oxygen consumption (FCCP) was determined. Fresh hippocampal tissues from mice were homogenized and centrifuged to obtain crude mitochondria. The oxygen consumptions of mitochondria including basal oxygen consumption of complexes I and II (CI + CII), oxygen consumption after oligomycin treatment (Oligo-resistant), and oxygen consumption after FCCP treatment (FCCP) were determined. There were at least 3 mice in each group. Data analysis was performed using DatLab software.

Immunohistochemistry

The immunohistochemistry was performed as previously described [15]. The primary antibodies included anti-HA (1: 500, Cell Signaling Technology), anti-GFAP (1: 1000, Abcam) or anti-Iba-1 (1:1000, Abcam). There were 3–5 mice in each group. There were 3–5 brain sections for each group, and one field was randomly selected from each section, from which the average value of each group was calculated.

Measurement of acetylcholinesterase (AchE) activity and acetylcholine (Ach) concentration

The lysates of fresh hippocampal tissues of mice (at least three mice in each group) were obtained by homogenization and centrifugation. AchE assay kit and Ach assay kit (Nanjing Jianjian Bioengineering Institute) were used to measure AchE activity and Ach concentration according to the manufacturer’s instructions.

Detection of mitochondrial complex I enzyme activity

Fresh mouse hippocampus tissues (at least 3 mice in each group) were homogenized and centrifuged to obtain mitochondria. Mitochondrial complex I enzyme activity was measured as previously described [15]. Briefly, the isolated mitochondria were resuspended, and frozen and thawed repeatedly. The rate of NADH oxidation in a reaction mixture was detected by a spectrophotometer, reflecting complex I enzyme activity. The complex I enzyme activity was normalized by the citrate synthase activity.

Statistical analysis

P values were calculated by using the SPSS 22.0 software. The t-test was used to compare means between two groups. The one-way ANOVA was used to compare means among three groups. First, the homogeneity of variances was tested. If the variances were equal, Tukey was used to calculate the P values. Otherwise, Tamhane’s T2 was used to calculate the P values. If P < 0.05, the difference was considered significant.

Results

Yeast NDI1 was successfully transduced into differentiated SH-SY5Y cells and localized in mitochondria

The structural diagram of recombinant expression vector rAAV9-CMV-HA-NDI1 was shown in Supplementary Fig. 1A. On the 8th day of trans-RA treatment, which was the 6th day after rAAV9-HA-NDI1 transduction, HA-tagged yeast NDI1 (HA-NDI1) was easily detected by western blot in human neural cell line SH-SY5Y (Supplementary Fig. 1B), with 70–75% of the cell population being HA-NDI1-positive examined by fluorescence microscopy (Supplementary Fig. 1C). Importantly, the HA-NDI1 was co-localized with MitoTracker after examined using laser scanning confocal microscopy (Supplementary Fig. 1D). These results demonstrated that the NDI1 gene was introduced by AAV9 into differentiated SH-SY5Y cells with high efficiency, and was highly expressed and localized in mitochondria.

Fig. 1
figure 1

Transduction of NDI1 can protect cell morphology, cell viability and reduce Tau phosphorylation in AD cell model induced by Aβ1-42. (A): The cell morphological changes were examined under an inverted phase contrast microscope (scale bar: 50 μm). (B): The cell viability was analyzed by trypan blue exclusion staining. (n=3). (C): Phosphorylated Tau (p-Tau) and Tau protein levels were detected by western blot. (D): Statistical analysis of the quantified p-Tau levels. (n=3). (E): Statistical analysis of quantified Tau levels. (n=3). Data were shown as mean ± SD, and analyzed by one-way ANOVA. ns: no significant difference, *: P<0.05, **: P<0.01, ***: P<0.001

Full size image

Transduction of NDI1 protects cell morphology, cell viability and reduces Tau phosphorylation in AD cell model induced by Aβ1–42

The cell bodies in DMSO + vector (normal) group were long and spindle-shaped, exhibiting extended synapses interconnected in a network-like manner. In contrast to the normal group, the cell bodies of Aβ1–42 + vector (model) group became rounded, with broken or disappeared synapses, fewer intercellular connections, and abnormal morphology. Compared with the model group, the cell bodies in Aβ1–42 + NDI1 (therapy) group were spindle-shaped, and the synapses became longer. However, compared with the normal group, the synapse length was not fully protected, with rare long synapses and relatively loose intercellular connections, in the therapy group. The results showed that transducing NDI1 can protect in cell morphology (Fig. 1A).

The cell survival percentage in the model group was significantly lower than that in the normal group (P = 0.0002), whereas in the therapy group it was significantly higher than that in the model group (P = 0.0075) (Fig. 1B). These results indicated that transducing NDI1 can protect cell viability.

Western blot analysis (Fig. 1C) of p-Tau (Ser202, Thr205) and Tau protein levels showed that p-Tau level was significantly increased in the model group compared with that in the normal group (P = 0.0279). In contrast, p-Tau level was significantly decreased in the therapy group compared with the model group (P = 0.0247, and decreased to levels similar to the normal group (Fig. 1D). Tau protein levels were not significantly different between groups (Fig. 1E). These results demonstrated that transducing NDI1 can reduce Tau protein phosphorylation.

Transduction of NDI1 compensates for defective complex I function and mitochondrial oxidative phosphorylation in AD cell model induced by Aβ1–42

The complex I-dependent oxygen consumption of cells was detected by a cellular respiration apparatus (Fig. 2A). The oxygen consumption was significantly lower in Aβ1–42 + vector (model) group than in DMSO + vector (normal) group (P = 0.0003). In contrast, it was significantly higher in the Aβ1–42 + NDI1 (therapy) group than in the model group (P = 0.0040). In order to determine the ratio of endogenous mammalian complex I to exogenous yeast complex I, the mammalian complex I inhibitor rotenone was first added, and then the general (including yeast) complex I inhibitor flavonoids was added to the chamber, to detect the sensitivity ratio in complex I-dependent oxygen consumption. Almost all of the complex I-dependent oxygen consumptions in the normal and the model groups were sensitive to rotenone, indicating the major role of endogenous complex I. However, in the therapy group, the rotenone-sensitive complex I-dependent oxygen consumption was significantly decreased compared with that in the normal or the model group (P < 0.001), whereas the flavonoids-sensitive complex I-dependent oxygen consumption was significantly increased compared with that in the normal or the model group (P < 0.001), which accounted for the majority of complex I-dependent oxygen consumption in the therapy group (Fig. 2B). These data indicated that exogenous complex I played the major role in the therapy group by compensating for a large proportion of the impaired endogenous complex I function.

Fig. 2
figure 2

Transduction of NDI1 can compensate for complex I function and mitochondrial oxidative phosphorylation function in AD cell model induced by Aβ1-42. (A): The complex I-dependent oxygen consumption of cells was analyzed by cellular respiration apparatus. (n=3). (B): The proportion of complex I-dependent oxygen consumption, sensitive to rotenone or flavonoids. (n=3). (C): The total oxygen consumption of cells was determined by cellular respiration apparatus, including basal oxygen consumption (Base), oligomycin-resistant oxygen consumption (Oligo-resistant) and maximum oxygen consumption after FCCP treatment (FCCP). (Oligo-sensitive) was calculated as Base- (Oligo-resistant), representing mitochondria-associated oxygen consumption. (n=3). (D): Respiratory control rate (RCR) was calculated as base/oligo-resistant. (n=3). (E): Leak control rate (LCR) was calculated as oligo-resistant/FCCP. (n=3). (F): The ATP level of cells was detected by using an ATP determination kit, including basal ATP (Base), oligomycin-resistant ATP (Oligo-resistant). (Oligo-sensitive) was calculated as Base- (Oligo-resistant), representing mitochondria-associated ATP. (n=4). (G): The proportion of oligomycin-sensitive ATP was calculated as oligo-sensitive/base. (n=4). Data were shown as mean ± SD, and analyzed by one-way ANOVA. ns: no significant difference, *: P<0.05, **: P<0.01, ***: P<0.001

Full size image

The total oxygen consumptions of cells including basal oxygen consumption (Base), oligomycin-resistant oxygen consumption (Oligo-resistant) and maximum oxygen consumption after FCCP treatment (FCCP) were also examined by the cellular respiration apparatus (Fig. 2C). Compared with DMSO + vector (normal) group, the Base, Oligo-sensitive, and FCCP levels were significantly lower in the Aβ1–42 + vector (model) group (P = 0.0011, P = 0.0004, P = 0.0202). In contrast, in the Aβ1–42 + NDI1 (therapy) group, these three oxygen consumption levels were significantly higher than those in the model group (P = 0.0042, P = 0.0034, P = 0.0456) (Fig. 2C). These results showed that transducing NDI1 can increase cellular basal oxygen consumption, mitochondria-related oxygen consumption, and maximum oxygen consumption to the levels similar to the normal group.

In order to further explore the mitochondrial ATP synthase coupling efficiency and the proton leakage, the respiratory control rate (RCR) and the leakage control rate (LCR) were calculated respectively. The respiratory control rate (Fig. 2D) was significantly lower in the model group than in the normal group (P = 0.0004), whereas it was significantly increased in the therapy group compared with the model group (P = 0.0474). Conversely, the leakage control rate (Fig. 2E) was significantly higher in the model group than in the normal group (P = 0.0009), whereas it was significantly lower in the therapy group (P = 0.0137), and reduced to level similar to the normal group.

The ATP level of cells was determined by using an ATP determination kit (Fig. 2F). Basal ATP level (Fig. 2F, Base) was significantly lower in Aβ1–42 + vector (model) group than in DMSO + vector (normal) group (P = 0.0001), whereas it was significantly higher in Aβ1–42 + NDI1 (therapy) group than in the model group (P = 0.0015). Similarly, the oligomycin-sensitive ATP level (Fig. 2F, Oligo-sensitive) and the ratio of oligomycin-sensitive ATP content (Fig. 2G) were significantly lower in the model group than in the normal group (P = 0.0002, P = 0.0124), whereas it was significantly higher in the therapy group than in the model group (P = 0.0039, P = 0.0066), and increased to level similar to the normal group.

Transduction of NDI1 reduces mitochondrial ROS content and protects from the reduction in final autophagy level in AD cell model induced by Aβ1–42

Mitochondrial ROS content was evaluated according to median fluorescence intensity (MFI) detected by flow cytometry. The mitochondrial ROS content was significantly higher in Aβ1–42 + vector (model) group than in DMSO + vector (normal) group (P = 0.0002), whereas it was significantly lower in Aβ1–42 + NDI1 (therapy) group than the model group (P = 0.0008), and reduced to levels similar to normal group (Fig. 3A). The results showed that transducing NDI1 can reduce mitochondrial oxidative stress level.

Fig. 3
figure 3

Transduction of NDI1 can reduce mitochondrial ROS content and protect from the reduction in final autophagy level in AD cell model induced by Aβ1-42. (A): The mitochondrial ROS was analyzed by flow cytometry. (n=3). (B): The LC3-I, LC3-II, and P62 protein levels were detected by western blot. (C): Statistical analysis of quantified LC3-II /I levels. (n=3). (D): Statistical analysis of quantified P62 levels. (n=3). (E): LAMP1 protein level was detected by western blot. (F): Statistical analysis of quantified LAMP1 levels. (n=3). Data were shown as mean ± SD, and analyzed by one-way ANOVA. ns: no significant difference, *: P<0.05, ***: P<0.001

Full size image

Western blot analysis was used to examine protein levels of autophagosome markers (LC3-II and P62) and lysosomal marker LAMP1 (Fig. 3B, E). LC3-II/I and P62 levels were significantly higher in Aβ1–42 + vector (model) group than in DMSO + vector (normal) group (P = 0.0383, P = 0.0100), whereas they were significantly lower in Aβ1–42 + NDI1 (therapy) group than in the model group (P = 0.0306, P = 0.0400) (Fig. 3C, D), and reduced to levels similar to normal group. Conversely, LAMP1 protein level was significantly lower in the model group than in the normal group (P = 0.0139), whereas it was significantly higher in the therapy group than in the model group (P = 0.0303) (Fig. 3F), and increased to level similar to normal group. These results suggested that Aβ1–42 promoted the formation of autophagosomes but damaged lysosomes, causing defective cellular autophagy due to the failed degradation of autophagosomes through the lysosomal pathway. In the therapy group, expression of yeast NDI1 reduced the number of autophagosomes and increased the number of lysosomes, so the autophagosomes could be completely degraded via the lysosomal pathway, protecting from the reduction in final cellular autophagy level.

NDI1 was successfully expressed in the right hippocampus of AD mouse model induced by Aβ1–42 and located in the mitochondrial membrane of hippocampal neurons

In order to estimate the kinetics of AAV9-NDI1 expression in mouse brain, AAV9-GFP virus was stereotaxically injected into the right hippocampus of mice, and GFP expression was examined by western blot at different times post AAV9-GFP injection. The results showed that GFP was detectable on day 14. Its expression reached the highest level on day 21 and could maintain until day 49 (Fig. 4A).

Fig. 4
figure 4

NDI1 was successfully expressed in the right hippocampus of AD mouse model induced by Aβ1-42 and located in the mitochondrial membrane of hippocampal neurons. (A): The expression kinetics of AAV9-GFP (used to estimate AAV9-NDI1) was detected by western blot. (B): The expression of HA (NDI1) was detected by western blot. (C): The HA (NDI1) expression (brown) was detected by immunohistochemistry. (Scale bar in the upper images: 500 μm, scale bar in the lower images: 50 μm). (D): The co-localization of HA (NDI1) (green) with mitochondrial membrane protein Tom20 (red) was examined by immunofluorescent staining, and counterstained with DAPI (blue) to visualize nuclei. Scale bar: 25 μm

Full size image

On the 35th day after injection of AAV9-HA-NDI1, HA (NDI1) in the right hippocampus of mouse brain was easily detected by western blot and immunohistochemistry. The results of western blot showed that HA (NDI1) was expressed in Aβ1–42 + NDI1 group (Fig. 4B). Immunohistochemistry also showed that HA (NDI1) was expressed in the cytoplasmic area of hippocampal neurons (Fig. 4C). The colocalization of HA (NDI1) with mitochondrial membrane protein Tom20 was examined by immunofluorescence, indicating that HA (NDI1) was successfully expressed in the mitochondrial membrane of hippocampal neurons (Fig. 4D).

Transduction of NDI1 improves short-term cognitive memory and long-term spatial memory in AD mouse model induced by Aβ1–42

The short-term cognitive memory of mice was assessed by the novel object recognition test. During the training phase, there was no significant difference in each group between the exploration time percentage for object A1 and that for object A2 (Fig. 5A). During the test phase, the exploration time percentage for new object B2 was significantly higher than that for old object A1 in DMSO + vector (normal) group or Aβ1–42 + NDI1 (therapy) group (P < 0.0001, P = 0.0008), whereas there was no significant difference in Aβ1–42 + vector (model) group (Fig. 5B). Likewise, during the test phase, the discrimination index (DI) for A1 and B2 objects was significantly lower in the model group than in the normal group or the therapy group (P = 0.0019, P = 0.0059), whereas there was no significant difference in DI between the therapy group and the normal group (Fig. 5C). The results showed that NDI1 can improve the short-term cognitive memory of AD mouse to the level similar to normal mouse.

Fig. 5
figure 5

Transduction of NDI1 can improve short-term cognitive memory and long-term spatial memory in AD mouse model induced by Aβ1-42. (A): The percentage of exploration time for object A1 or A2 during the training phase in the novel object recognition test. (n=8). (B): The percentage of exploration time for object A1 or B2 during the testing phase. (n=8). (C): The discrimination index (DI) for object A1 and B2 during the testing phase. DI was calculated as (B2-A1)/ (A1+B2). (n=8). (D): The time for a mouse to find the hidden platform in the hidden platform trial of Morris water maze test. Comparison between Aβ1-42+vector group and DMSO+vector group, *P<0.05; comparison between Aβ1-42+ NDI1 group and Aβ1-42+vector group, #P<0.05. (n=6). (E): The route map of a mouse in the probe trial. The red point was the starting position, and the blue position was the end position. (F): The cumulative time for a mouse to cross the target quadrant in the probe trial. (n=6). (G): The total number of times for a mouse to cross the target quadrant in the probe trial. (n=6). Data were shown as mean ± SD, and analyzed by one-way ANOVA. ns: no significant difference, *: P<0.05, **: P<0.01, ***: P<0.001

Full size image

The long-term spatial memory of mice was evaluated by the Morris water maze test. In the spatial acquisition test (hidden platform trial), the time for a mouse to find the hidden platform on days 4 and 5 was significantly increased in Aβ1–42 + vector (model) group compared with DMSO + vector (normal) group or Aβ1–42 + NDI1 (therapy) group (P = 0.016, P = 0.0466, P = 0.0169, P = 0.0324), whereas there was no significant difference between the therapy group and the normal group (Fig. 5D). The route map of the reference memory test (probe trial) on day 6 was shown in Fig. 5E. The cumulative time (Fig. 5F) or the number of times (Fig. 5G) for a mouse crossing the target quadrant was significantly reduced in the model group than in the normal group or the therapy group (P = 0.0348, P = 0.0196, P = 0.0007, P = 0.0071), whereas there was no significant difference between the therapy group and the normal group. The results showed that NDI1 can improve the long-term spatial memory of AD mouse to the level similar to normal mouse.

Transduction of NDI1 reduces Tau protein phosphorylation and neuroinflammatory response and improves cholinergic level in the hippocampus of AD mouse model induced by Aβ1–42

The phosphorylated-Tau (p-Tau) (Thr181) and Tau protein levels in the right hippocampus of mouse were detected by western blot (Fig. 6A). The p-Tau level was significantly higher than in Aβ1–42 + vector (model) group than in DMSO + vector (normal) group (P = 0.001), and was significantly lower in Aβ1–42 + NDI1 (therapy) group than in the model group (P = 0.0013), reaching to the low level similar to the normal group (Fig. 6B). In contrast, there was no significant difference in the Tau levels among all groups (Fig. 6C). These results showed that NDI1 can reduce Tau phosphorylation in the AD mouse model.

Fig. 6
figure 6

Transduction of NDI1 can reduce Tau protein phosphorylation and neuroinflammatory response and improve cholinergic level in the hippocampus of AD mouse model induced by Aβ1-42. (A): The p-Tau and Tau protein levels were analyzed by western blot. (B): Statistical analysis of quantified p-Tau levels. (n=3). (C): Statistical analysis of quantified Tau levels. (n=3). (D): GFAP positive cells (brown in cytoplasm) and Iba-1 positive cells (brown in plasma membrane and extended synapses) were detected by immunohistochemistry. (Scale bar in the images of low magnification: 50 μm, scale bar in the images of high magnification: 150 μm). (E): Statistical analysis of the percentage of GFAP positive cells (astrocytes). (n=5). (F): Statistical analysis of the percentage of Iba-1 positive cells (microglia). (n=5). (G): AchE activity was analyzed by AchE kit. (n=3). (H): Ach level was analyzed by Ach kit. (n=3). Data were shown as mean ± SD, and analyzed by one-way ANOVA. ns: no significant difference, *p<0.05,**p<0.01, ***p<0.001

Full size image

To examine the effects on neuroinflammatory response, immunohistochemistry was used to detect GFAP positive cells (astrocytes) and Iba-1 positive cells (microglia) in the right hippocampus (Fig. 6D). The percentage of GFAP (Fig. 6E) or Iba-1 (Fig. 6F) positive cells was significantly higher in Aβ1–42 + vector (model) group than in DMSO + vector (normal) group (P = 0.0095, P = 0.009), and was significantly lower in Aβ1–42 + NDI1 (therapy) group than in the model group (P = 0.0057, P = 0.0089), reaching to the low level similar to the normal group. These results showed that NDI1 can reduce the inflammatory response in the AD mouse model.

The acetylcholinesterase (AchE) activity and acetylcholine (Ach) concentration are considered to be the neuropathogenic markers for AD [18, 19]. To examine the effects on cholinergic level, AchE activity and Ach concentration were measured in the right hippocampus. The AchE activity was significantly elevated in Aβ1–42 + vector (model) group compared to DMSO + vector (normal) group (P = 0.0042), whereas it was significantly decreased in Aβ1–42 + NDI1 (therapy) group compared to the model group (P = 0.0062), reaching to the low level similar to the normal group (Fig. 6G). Conversely, Ach concentration was significantly lower in the model group compared to the normal group (P < 0.0001), whereas it was significantly higher in the therapy group compared to the model group (P = 0.0031) (Fig. 6H). These results indicated that NDI1 can also improve the hippocampal cholinergic level in the AD mouse model.

Transduction of NDI1 improves complex I activity and mitochondrial oxidative phosphorylation function in the hippocampus of AD mouse model induced by Aβ1–42

The mitochondrial complex I enzyme activity in the right hippocampus was detected using a UV spectrophotometer. Compared with the DMSO + vector (normal) group, complex I activity was significantly lower in the Aβ1–42 + vector (model) group (P = 0.0096) (a decrease of about 67%). However, complex I activity was significantly higher in the Aβ1–42 + NDI1 (therapy) group than in the model group (P = 0.0496), reaching to the level similar to the normal group (Fig. 7A). These results showed that NDI1 can compensate for mitochondrial complex I enzyme activity in the AD mouse model.

Fig. 7
figure 7

Transduction of NDI1 can improve complex I activity and mitochondrial oxidative phosphorylation function in the hippocampus of AD mouse model induced by Aβ1-42. (A): The complex I enzyme activity was measured using a UV spectrophotometer. (n=3). (B): The total oxygen consumption of mitochondria was detected by cellular respiration apparatus, including basal oxygen consumption of complexes I and II (C I+ C II), oligomycin-resistant oxygen consumption (Oligo-resistant), and maximum oxygen consumption after FCCP treatment (FCCP). (Oligo-sensitive) was calculated as (C I+ C II)- (oligo-resistant). (n=3). (C): Respiratory control rate (RCR) was calculated as (C I + C II)/oligo-resistant. (n=3). (D): Leakage control rate (LCR) was calculated as oligo-resistant/FCCP. (n=3). Data were shown as mean ± SD, and analyzed by one-way ANOVA. ns: no significant difference, *: P<0.05, **: P<0.01, ***: P<0.001

Full size image

The total oxygen consumption of mitochondria in the right hippocampus was determined by cellular respiration apparatus, including the basal oxygen consumption of complexes I and II (C I + C II), oligomycin-resistant oxygen consumption (Oligo-resistant), and maximum oxygen consumption after FCCP treatment (FCCP). Compared with DMSO + vector (normal) group, oxygen consumption levels of C I +C II, Oligo-sensitive, and FCCP (P = 0.0342, P = 0.0091, P = 0.0003) were significantly lower in the Aβ1–42 + vector (model) group. These three oxygen consumption levels were significantly higher in the Aβ1–42 + NDI1 (therapy) group than in the model group (P = 0.0197, P = 0.026, P = 0.0105) (Fig. 7B). Mitochondrial coupling efficiency was evaluated by the respiratory control rate (RCR) and the leakage control rate (LCR). RCR, representing the coupling of substrate oxidation and ATP synthase, was significantly lower in the model group than in the normal group (P = 0.0037), whereas it was significantly higher in the therapy group than in the model group (P = 0.0419) (Fig. 7C). LCR, representing proton leakage, was significantly higher in the model group than in the normal group (P = 0.0017), whereas it was significantly lower in the therapy group than in the model group (P = 0.0063) (Fig. 7D). These results showed that NDI1 can improve the total oxygen consumption, ATP synthase-related oxygen consumption, uncoupled maximum oxygen consumption and mitochondrial coupling efficiency in AD mouse to the levels similar to normal mouse.

Injection of AAV9-NDI1 into the hippocampus of mouse does not affect the short-term cognitive memory and long-term spatial memory abilities in normal mouse

The novel object recognition test was used to assess the short-term cognitive memory of mouse. During the training phase, there was no significant difference in each group between the exploration time percentage for object A1 and that for object A2 (Supplementary Fig. 2A). During the testing phase, the exploration time percentage for new object B2 was significantly higher than that for old object A1 in vector group (P = 0.0007) or NDI1 group (P = 0.0058) (Supplementary Fig. 2B). During the testing phase, there was no significant difference in the discrimination index (DI) for A1 and B2 objects between vector group and NDI1 group (Supplementary Fig. 2C).

The Morris water maze test was used to evaluate the long-term spatial memory of mouse. In the spatial acquisition test (hidden platform trial) (day 1–5), there was no significant difference between vector group and NDI1 group in the time for a mouse to find the hidden platform between vector group and NDI1 group (Supplementary Fig. 2D). The route map of the reference memory test (probe trial) (day 6) was shown in Supplementary Fig. 2E. There was also no significant difference in the cumulative time (Supplementary Fig. 2F) or the number of times (Supplementary Fig. 2G) for a mouse crossing the target quadrant between vector group and NDI1 group. These results indicate that injection of AAV9-NDI1 into the hippocampus does not affect the short-term and long-term memories in normal mouse.

Injection of AAV9-NDI1 into the hippocampus of mouse does not cause neuronal damage or neuroinflammatory response in normal mouse

The morphology of pyramidal neurons in CA1 area of hippocampus was examined by Hematoxylin-Eosin (H&E) stain. Neurons in both vector group and NDI1 group were arranged in an orderly manner, and showed round or oval shape with large/round nuclei and dark blue nucleoli (Supplementary Fig. 3A). Immunohistochemistry of the mouse hippocampus (Supplementary Fig. 3B) revealed that there were no significant difference in percentages of GFAP positive astrocytes (Supplementary Fig. 3C) or Iba-1 positive microglia (Supplementary Fig. 3D) between vector group and NDI1 group. These results indicate that injection of AAV9-NDI1 into the hippocampus does not cause any obvious neuronal defects in normal mouse.

Discussion

In recent years, increasing evidence has shown that mitochondrial abnormalities play a crucial role in the pathogenesis of AD [8, 10, 20,21,22,23,24]. It has been found that mitochondrial abnormalities are early events in AD pathological changes based on studies on brain tissue of AD patients, transgenic mice, and transgenic cell lines, and the mitochondrial cascade hypothesis of AD pathogenesis has been proposed [8,9,10, 25, 26]. Therefore, our study focused on the gene therapy of AD targeting mitochondrial defects.

Complex I is the rate-limiting entrance for electron transfer in the respiratory chain. In addition, complex I is also involved in oxidative stress, apoptosis, and autophagy. Functional defects of mitochondrial complex I can lead to diseases of the nervous, muscular and other systems including Parkinson’s disease, AD, Leigh syndrome, and Leber optic neuropathy etc. [27]. Quantitative analysis of human brain tissue protein profiles showed that the levels of up to 12 subunits of complex I was significantly reduced in the brains of patients with late-onset AD, whereas the levels of complexes II, III, IV, and V did not show significant changes [13]. A recent study using the newly developed imaging marker [18F] BCPP-EF revealed that there were indeed defects in complex I in the brains of AD patients [14]. And this study revealed that these complex I defects were closely related to neurodegenerative changes (brain atrophy, cognitive impairment, etc.) in mild AD, suggesting that complex I deficiency could be a novel therapeutic target for early AD [14].

Aβ oligomerization and accumulation in both hereditary AD and sporadic AD coupled with Tau hyperphosphorylation is believed to cause mitochondrial complex defects [12]. In a vicious cycle, the mitochondrial complex defects can aggravate the deposition of Aβ to form plaques and promote the phosphorylation of Tau to form neurofibrillary tangles, resulting in defective oxidative phosphorylation, decreased ATP, and ultimately leading to abnormal synaptic function and neuronal cell death [12, 28]. The results of our study showed that Aβ1–42 causes Tau pathology (Fig. 1C, D E, Fig. 6A, B,C) and mitochondrial complex I pathology (Figs. 2D , E , 7A) in our AD cell and mouse models. Among them, the activity of complex I in the AD model mice decreased by about 67% compared with that in the normal mice (Fig. 7A), which forms the basis of our study by compensating the function of mitochondrial complex I to prevent the vicious cycle of complex defects with Aβ pathology and Tau pathology. However, the timing of the highest NDI1 expression was the same as the timing of Aβ1–42 treatment in the vitro experiments of this study. Our experiments in vitro were to some extent preventative. The timing of the highest NDI1 expression was 7 days after the start of Aβ1–42 treatment in the vivo experiments of this study. Our experiments in vivo were to some extent rescuing. We plan to use a transgenic AD mouse model after their AD phenotypes being well developed, to examine the effect of AAV9-NDI1 gene therapy for rescuing the AD phenotypes in our future study.

New AD treatments have been actively developed [29, 30]. In the clinical trials of drugs targeting Aβ, the vast majority of Aβ antibodies are unable to effectively change the course of the disease [6]. Tau-targeted therapy also attracts attention in recent years, but no significant therapeutic effects have been reported [6, 21, 29]. Small molecule modulators of mitochondrial function have been tested in clinical trials for AD treatment [9, 21, 29,30,31]. According to the Clinicaltrials.gov, most AD clinical trials have focused on Aβ with less trials targeting tau, APOE, and mitochondria [6, 21, 29]. Considering the multifactorial complexity and heterogeneity of AD, AD patients with different causative factors should be treated separately with individualized therapies [6, 9]. Our study adopted a novel gene therapy strategy to compensate for defective mitochondrial complex I, which may be used to treat AD patients with severely low mitochondrial complex I function selected by imaging marker [14] in the future.

Currently reported potential targets of gene therapy for AD include Aβ metabolism, lipid and protein metabolism (such as APOE), growth factors (such as NGF), synaptic function, inflammation, cell survival, and tau [32]. Among them, only AAV-NGF, Tau antisense oligonucleotide, and AAV-APOEε2 have processed to clinical trials [6, 33, 34]. According to a published report [35], the survival rate of PS/APP transgenic AD mice with p66Shc gene knockout increased at 15 months of age, and Aβ-induced cognitive impairment was completely recovered. Importantly, these beneficial effects on survival and cognitive performance were independent of Aβ levels and amyloid plaque deposition, but were associated with improved brain mitochondrial respiration, a reversal of mitochondrial complex I dysfunction, restored adenosine triphosphate production and reduced ROS levels [35]. Our study of complex I gene therapy showed that compensating the defective complex I directly can in fact improve short-term cognitive memory ability, long-term spatial memory ability, Tau histopathology, and neural tissue inflammation in our Aβ-induced AD mouse model to the levels similar to the normal mice (Figs. 5 , 6).

Up to now, one of the treatment strategies for mitochondrial complex I defects is complex I gene therapy. Guy et al. conducted clinical trial on ND4 gene therapy for Leber disease [36]. This approach can only target the defect of one subunit of complex I. Our study used yeast complex I NDI1 gene for gene therapy, which can target defects in all subunits of complex I. Studies showed [37] that yeast complex I could compensate for the function of the entire complex I consisting of 45 subunits in mammalian cells. Transgenic NADH dehydrogenase yeast NDI1 can restore oxygen regulation of breathing in mice with mitochondrial complex I-deficiency [38]. Using the NDI1 gene to target the defective complex I could resolve the problem of axon damage and neuron loss in Multiple Sclerosis disease model [39]. Studies showed that the expression of yeast NDI1 in rats does not cause immune response, possibly because the exogenous protein is located in the mitochondria and evades immune surveillance [40]. Our previously published paper [15] showed that yeast complex I can compensate for the rotenone-induced complex I functional defects in human neural cells. Especially in the rotenone-induced Parkinson’s disease mouse model, yeast NDI1 had good therapeutic effects in the aspects of histopathology and neurobehaviors, and AAV5-mediated delivery of yeast NDI1 in the substantia nigra area of the mouse brain did not cause any adverse effects.

Clinical trials have proven that the safety, efficacy, and expression persistence of AAV in gene therapy for Parkinson’s disease and AD are satisfactory [41]. Currently, in phase I and phase II clinical trials of gene therapy for AD and Parkinson’s disease, recombinant AAV2 is generally injected directly into the brain. Long-term observations of all these clinical trials have not found serious side effects or adverse reactions [41,42,43]. It was reported [44] that AAV type 9 spread in the hippocampus of mice and had a higher infection efficiency compared with other AAV serotypes. Recent studies showed that injection of AAV9-Survivin [45] or AAV9-MICU3 [46] into the hippocampus of AD model mice achieved optimal therapeutic efficacy. AAV9 has been used in clinical trials [47]. In our study, AAV9-NDI1 was stereotaxically injected into the hippocampus of the AD mouse model. In order to test the safety of AAV9-mediated yeast NDI1 gene therapy in the hippocampus of brain, AAV9-NDI1 was also injected into the hippocampus of normal mice. Our results showed that AAV9-NDI1 did not cause significant side effects (Supplementary Figs. 2 and 3).

In conclusion, our study can offer a novel strategy, and a potential effective and safe gene therapy approach for treating AD characterized by complex I abnormalities. Additional research is still necessary to determine whether this finding can be translated into clinical practice.

Data availability

The datasets used and analyzed during the current study are available from the corresponding authors upon reasonable request.

Abbreviations

AAV:

adeno-associated virus

rAAV:

recombinant adeno-associated virus

Aβ:

β-amyloid

Ach:

acetylcholine

AchE:

acetylcholinesterase

AD:

Alzheimer’s disease

FCCP:

carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone

LCR:

leakage control rate

NDI1:

NADH-quinone oxidoreductase

RCR:

respiratory control rate

ROS:

reactive oxygen species

References

  1. Long JM, Holtzman DM. Alzheimer Disease: an update on pathobiology and treatment strategies. Cell. 2019;179(2):312–39.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  2. Cui L, Hou NN, Wu HM, Zuo X, Lian YZ, Zhang CN, et al. Prevalence of Alzheimer’s Disease and Parkinson’s Disease in China: an updated Systematical Analysis. Front Aging Neurosci. 2020;12:603854.

    Article  PubMed  PubMed Central  Google Scholar 

  3. Collaborators GBDDF. Estimation of the global prevalence of dementia in 2019 and forecasted prevalence in 2050: an analysis for the global burden of Disease Study 2019. Lancet Public Health. 2022;7(2):e105–25.

    Article  Google Scholar 

  4. Saleem U, Akhtar R, Anwar F, Shah MA, Chaudary Z, Ayaz M, et al. Neuroprotective potential of Malva neglecta is mediated via down-regulation of cholinesterase and modulation of oxidative stress markers. Metab Brain Dis. 2021;36(5):889–900.

    Article  CAS  PubMed  Google Scholar 

  5. Congdon EE, Sigurdsson EM. Tau-targeting therapies for Alzheimer disease. Nat Rev Neurol. 2018;14(7):399–415.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  6. Pleen J, Townley R. Alzheimer’s disease clinical trial update 2019–2021. J Neurol. 2022;269(2):1038–51.

    Article  PubMed  Google Scholar 

  7. van Dyck CH, Swanson CJ, Aisen P, Bateman RJ, Chen C, Gee M, et al. Lecanemab in Early Alzheimer’s Disease. N Engl J Med. 2023;388(1):9–21.

    Article  PubMed  Google Scholar 

  8. Swerdlow RH. Mitochondria and mitochondrial cascades in Alzheimer’s Disease. J Alzheimers Dis. 2018;62(3):1403–16.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Lim JW, Lee J, Pae AN. Mitochondrial dysfunction and Alzheimer’s disease: prospects for therapeutic intervention. BMB Rep. 2020;53(1):47–55.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Desler C, Lillenes MS, Tonjum T, Rasmussen LJ. The role of mitochondrial dysfunction in the progression of Alzheimer’s Disease. Curr Med Chem. 2018;25(40):5578–87.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. Swerdlow RH. The Alzheimer’s disease mitochondrial Cascade hypothesis: a current overview. J Alzheimers Dis. 2023;92(3):751–68.

    Article  CAS  PubMed  Google Scholar 

  12. Eckert A, Schulz KL, Rhein V, Gotz J. Convergence of amyloid-beta and tau pathologies on mitochondria in vivo. Mol Neurobiol. 2010;41(2–3):107–14.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  13. Adav SS, Park JE, Sze SK. Quantitative profiling brain proteomes revealed mitochondrial dysfunction in Alzheimer’s disease. Mol Brain. 2019;12(1):8.

    Article  PubMed  PubMed Central  Google Scholar 

  14. Terada T, Therriault J, Kang MS, Savard M, Pascoal TA, Lussier F, et al. Mitochondrial complex I abnormalities underlie neurodegeneration and cognitive decline in Alzheimer’s disease. Eur J Neurol. 2022;29(5):1324–34.

    Article  PubMed  Google Scholar 

  15. Li H, Sun B, Huang Y, Zhang J, Xu X, Shen Y, et al. Gene therapy of yeast NDI1 on mitochondrial complex I dysfunction in rotenone-induced Parkinson’s disease models in vitro and vivo. Mol Med. 2022;28(1):29.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Lueptow LM. Novel object Recognition Test for the investigation of learning and memory in mice. J Vis Exp. 2017(126).

  17. Vorhees CV, Williams MT. Morris water maze: procedures for assessing spatial and related forms of learning and memory. Nat Protoc. 2006;1(2):848–58.

    Article  PubMed  PubMed Central  Google Scholar 

  18. Tatulian SA. Challenges and hopes for Alzheimer’s disease. Drug Discov Today. 2022;27(4):1027–43.

    Article  PubMed  Google Scholar 

  19. Hampel H, Mesulam MM, Cuello AC, Farlow MR, Giacobini E, Grossberg GT, et al. The cholinergic system in the pathophysiology and treatment of Alzheimer’s disease. Brain. 2018;141(7):1917–33.

    Article  PubMed  PubMed Central  Google Scholar 

  20. Wang W, Zhao F, Ma X, Perry G, Zhu X. Mitochondria dysfunction in the pathogenesis of Alzheimer’s disease: recent advances. Mol Neurodegener. 2020;15(1):30.

    Article  PubMed  PubMed Central  Google Scholar 

  21. Perez Ortiz JM, Swerdlow RH. Mitochondrial dysfunction in Alzheimer’s disease: role in pathogenesis and novel therapeutic opportunities. Br J Pharmacol. 2018;176(18):3489–507.

    Article  Google Scholar 

  22. Scheltens P, De Strooper B, Kivipelto M, Holstege H, Chetelat G, Teunissen CE, et al. Alzheimer’s disease. Lancet. 2021;397(10284):1577–90.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Ashleigh T, Swerdlow RH, Beal MF. The role of mitochondrial dysfunction in Alzheimer’s disease pathogenesis. Alzheimers Dement. 2023;19(1):333–42.

    Article  CAS  PubMed  Google Scholar 

  24. Reiss AB, Ahmed S, Dayaramani C, Glass AD, Gomolin IH, Pinkhasov A, et al. The role of mitochondrial dysfunction in Alzheimer’s disease: a potential pathway to treatment. Exp Gerontol. 2022;164:111828.

    Article  CAS  PubMed  Google Scholar 

  25. Han Y, Liu D, Cheng Y, Ji Q, Liu M, Zhang B, et al. Maintenance of mitochondrial homeostasis for Alzheimer’s disease: strategies and challenges. Redox Biol. 2023;63:102734.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  26. Guo W, Hassan LA, Chu YH, Yang XP, Wang SX, Zhu HX, et al. Mapping trends and hotspots of mitochondrial dysfunction in Alzheimer’s disease from 2013 to 2022: a bibliometric analysis of global research. Front Neurosci. 2023;17:1199625.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Rodenburg RJ. Mitochondrial complex I-linked disease. Biochim Biophys Acta. 2016;1857(7):938–45.

    Article  CAS  PubMed  Google Scholar 

  28. Fisar Z. Linking the Amyloid, Tau, and Mitochondrial Hypotheses of Alzheimer’s Disease and Identifying Promising Drug Targets. Biomolecules. 2022;12(11).

  29. Cummings J, Lee G, Zhong K, Fonseca J, Taghva K. Alzheimer’s disease drug development pipeline: 2021. Alzheimers Dement (N Y). 2021;7(1):e12179.

    Article  PubMed  Google Scholar 

  30. Cummings J, Lee G, Nahed P, Kambar M, Zhong K, Fonseca J, et al. Alzheimer’s disease drug development pipeline: 2022. Alzheimers Dement (N Y). 2022;8(1):e12295.

    Article  Google Scholar 

  31. Cummings J, Lee G, Ritter A, Sabbagh M, Zhong K. Alzheimer’s disease drug development pipeline: 2020. Alzheimers Dement (N Y). 2020;6(1):e12050.

    Article  PubMed  Google Scholar 

  32. Congdon EE, Ji C, Tetlow AM, Jiang Y, Sigurdsson EM. Tau-targeting therapies for Alzheimer disease: current status and future directions. Nat Rev Neurol. 2023.

  33. Loera-Valencia R, Piras A, Ismail MAM, Manchanda S, Eyjolfsdottir H, Saido TC, et al. Targeting Alzheimer’s disease with gene and cell therapies. J Intern Med. 2018;284(1):2–36.

    Article  CAS  PubMed  Google Scholar 

  34. Griciuc A, Federico AN, Natasan J, Forte AM, McGinty D, Nguyen H, et al. Gene therapy for Alzheimer’s disease targeting CD33 reduces amyloid beta accumulation and neuroinflammation. Hum Mol Genet. 2020;29(17):2920–35.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  35. Derungs R, Camici GG, Spescha RD, Welt T, Tackenberg C, Spani C, et al. Genetic ablation of the p66(shc) adaptor protein reverses cognitive deficits and improves mitochondrial function in an APP transgenic mouse model of Alzheimer’s disease. Mol Psychiatry. 2016;22(4):605–14.

    Article  PubMed  Google Scholar 

  36. Koilkonda RD, Yu H, Chou TH, Feuer WJ, Ruggeri M, Porciatti V, et al. Safety and effects of the vector for the Leber hereditary optic neuropathy gene therapy clinical trial. JAMA Ophthalmol. 2014;132(4):409–20.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Mentzer RM Jr., Wider J, Perry CN, Gottlieb RA. Reduction of infarct size by the therapeutic protein TAT-Ndi1 in vivo. J Cardiovasc Pharmacol Ther. 2014;19(3):315–20.

    Article  CAS  PubMed  Google Scholar 

  38. Jimenez-Gomez B, Ortega-Saenz P, Gao L, Gonzalez-Rodriguez P, Garcia-Flores P, Chandel N, et al. Transgenic NADH dehydrogenase restores oxygen regulation of breathing in mitochondrial complex I-deficient mice. Nat Commun. 2023;14(1):1172.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  39. Talla V, Koilkonda R, Guy J. Gene Therapy with single-subunit yeast NADH-Ubiquinone oxidoreductase (NDI1) improves the visual function in experimental autoimmune encephalomyelitis (EAE) mice model of multiple sclerosis (MS). Mol Neurobiol. 2020;57(4):1952–65.

    Article  CAS  PubMed  Google Scholar 

  40. Marella M, Seo BB, Flotte TR, Matsuno-Yagi A, Yagi T. No immune responses by the expression of the yeast Ndi1 protein in rats. PLoS ONE. 2011;6(10):e25910.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Qu Y, Liu Y, Noor AF, Tran J, Li R. Characteristics and advantages of adeno-associated virus vector-mediated gene therapy for neurodegenerative diseases. Neural Regeneration Res. 2019;14(6):931–8.

    Article  CAS  Google Scholar 

  42. Rocco MT, Akhter AS, Ehrlich DJ, Scott GC, Lungu C, Munjal V, et al. Long-term safety of MRI-guided administration of AAV2-GDNF and gadoteridol in the putamen of individuals with Parkinson’s disease. Mol Ther. 2022;30(12):3632–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  43. Rafii MS, Tuszynski MH, Thomas RG, Barba D, Brewer JB, Rissman RA, et al. Adeno-Associated viral vector (serotype 2)-Nerve growth factor for patients with Alzheimer Disease: a Randomized Clinical Trial. JAMA Neurol. 2018;75(7):834–41.

    Article  PubMed  PubMed Central  Google Scholar 

  44. Aschauer DF, Kreuz S, Rumpel S. Analysis of transduction efficiency, tropism and axonal transport of AAV serotypes 1, 2, 5, 6, 8 and 9 in the mouse brain. PLoS ONE. 2013;8(9):e76310.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  45. Lee Y, Ju YJ, Gee MS, Jeon SH, Kim N, Koo T et al. Survivin enhances hippocampal neurogenesis and cognitive function in Alzheimer’s disease mouse model. CNS Neurosci Ther. 2023.

  46. Zhou G, Ye Q, Xu Y, He B, Wu L, Zhu G, et al. Mitochondrial calcium uptake 3 mitigates cerebral amyloid angiopathy-related neuronal death and glial inflammation by reducing mitochondrial dysfunction. Int Immunopharmacol. 2023;117:109614.

    Article  CAS  PubMed  Google Scholar 

  47. Gomez Limia C, Baird M, Schwartz M, Saxena S, Meyer K, Wein N. Emerging perspectives on Gene Therapy Delivery for neurodegenerative and neuromuscular disorders. J Pers Med. 2022;12(12).

Download references

Acknowledgements

Not applicable.

Funding

This work was supported by the Wenzhou Municipal Science and Technology Bureau of China [grant number ZY2022002]; the National Natural Science Foundation of China [grant number 81971291]; and Key Discipline of Zhejiang Province in Medical Technology [First Class, Category A].

Author information

Author notes

  1. Hongzhi Li, Zhuo Chen and Yuqi Shen contributed equally to this work and share first authorship.

Authors and Affiliations

  1. Key Laboratory of Laboratory Medicine, Ministry of Education, Wenzhou Key Laboratory of Cancer Pathogenesis and Translation, School of Laboratory Medicine and Life Sciences, Wenzhou Medical University, Wenzhou, 325035, China

    Hongzhi Li, Zhuo Chen, Yuqi Shen, Ting Xiong, Andong Chen, Lixia Chen, Yifan Ye & Qingyou Jiang

  2. Brain Center, Wenzhou Central Hospital, Wenzhou, 325000, China

    Yaxi Zhang & Jun Sun

  3. Department of Internal Neurology, Beijing Friendship Hospital, Capital Medical University, 95 Yongan Road, Xicheng District, Beijing, 100050, China

    Luxi Shen

  4. School of Laboratory Medicine and Life Sciences, Wenzhou Medical University, Northern Zhongxin Road, Chashan University Town, Wenzhou, Zhejiang, 325035, China

    Hongzhi Li

Authors
  1. Hongzhi Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Zhuo Chen
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Yuqi Shen
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Ting Xiong
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Andong Chen
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Lixia Chen
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Yifan Ye
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Qingyou Jiang
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Yaxi Zhang
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Jun Sun
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. Luxi Shen
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Hongzhi Li, Luxi Shen and Jun Sun conceived and designed the research; Zhuo Chen, Yuqi Shen, Ting Xiong, Andong Chen, Lixia Chen, Yifan Ye, Qingyou Jiang and Yaxi Zhang performed the research and acquired the data, Zhuo Chen and Yuqi Shen analyzed and interpreted the data. Zhuo Chen, Yuqi Shen and Luxi Shen drafted the manuscript, Luxi Shen, Hongzhi Li and Jun Sun revised the manuscript. All authors have read and approved the final version of the manuscript.

Corresponding authors

Correspondence to Hongzhi Li, Jun Sun or Luxi Shen.

Ethics declarations

Ethics approval and consent to participate

This study was conducted according to the guidelines of Declaration of Helsinki, and approved by the Institutional Animal Care & Use Committee of Wenzhou Medical University in Wenzhou, China (wydw2022-0236). Not applicable for consent to participate.

Consent for publication

Not applicable.

Competing interests

All authors declare that they have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Li, H., Chen, Z., Shen, Y. et al. Gene therapy in Aβ-induced cell and mouse models of Alzheimer’s disease through compensating defective mitochondrial complex I function. J Transl Med 22, 760 (2024). https://doi.org/10.1186/s12967-024-05571-3

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12967-024-05571-3

Keywords

Journal of Translational Medicine

ISSN: 1479-5876

Contact us