Background
Alzheimer's disease (AD) is one of the most prevalent neurodegenerative diseases and the No. 1 cause of dementia in the world [
1,
2]. With the aging of global population, the prevalence and incidence of AD keep increasing [
1‐
4]. By 2050, the prevalence of dementia will triple worldwide, putting a tremendous emotional and financial load on society [
1]. AD is mainly manifested as progressive and selective loss of neurons, resulting in memory impairment and executive dysfunction, accompanied by neuropsychiatric symptoms. The main pathological features of AD include extracellular β-amyloid (Aβ) plaques and intracellular neurofibrillary tangles (NFTs) composed of phosphorylated Tau protein (p-Tau) [
5,
6]. The complex Aβ–Tau interaction has synergistic effect on the pathogenesis of AD [
7]. However, clinical trials that aim to target Aβ and NFTs fail to obtain satisfied results, suggesting the necessity to investigate AD pathogenesis in a novel perspective [
8].
Microglia are innate immune cells in the central nervous system (CNS) and play an important role in the phagocytosis and clearance of pathogenic molecules [
9‐
11] and neuroinflammatory responses [
12]. Mounting evidence has shown that microglia play a critical role in Aβ and p-Tau-mediated neuronal dysfunction in the pathogenesis of AD [
13,
14]. Aβ has been found to interact with the receptor for advanced glycation end products (RAGE) in microglia, which may result in excessive activation of microglia and the production of pro-inflammatory molecules [
11,
13,
15].
Proinflammatory microglia substantially alter brain energy metabolism and ROS/NO production, leading to impairment of neural network function, neurodegeneration, and blood–brain barrier dysfunction that contribute to the pathogenesis of AD [
12,
16,
17]. Hence, microglia perform a variety of specialized functions in AD, including phagocytosis and clearance of Aβ and p-Tau proteins, removal of dying neurons, and release of neurotrophic factors that support neuronal cells [
5,
18]. These immune functions demand high energy, which are regulated by mitochondria.
Mitochondria are maternally inherited organelles that play critical roles in oxidative stress, energy metabolism, calcium homeostasis, and cell survival [
19]. AD has been proposed as a metabolic disease [
18‐
21]. Importantly, a strong correlation between microglial activation and metabolic dysfunction in AD has been demonstrated in both basic research and clinical studies [
5,
18,
19]. In AD, a series of mitochondrial abnormalities have been identified, including structure alteration, age-dependent accumulation of mitochondrial DNA (mtDNA) changes, altered mitochondrial membrane potential, excessive mitochondrial ROS production, reduced mitochondrial adenosine triphosphate (ATP), disrupted electron transport chain (ETC), and increased mitochondrial fragmentation, leading to defective mitophagy in microglia and other brain cells [
13,
22‐
29]. In addition, chronic exposure to Aβ and p-Tau induces dysregulated expression of late-onset AD-associated genes [
30], mitochondrial toxicity, and metabolic dysfunction in microglia [
31]. Moreover, therapies targeting basic mitochondrial processes, such as quality control (QC), show great therapeutic potential [
32]. This review aims to summarize the investigations on the association of mitochondrial dysfunction with the inflammatory responses and the phagocytic capacity of microglia, and its involvement in the onset and progression of AD. Future directions that deepen our understanding of mitochondrial dysfunction in AD and facilitate the development of novel therapeutic strategies were also provided.
Mitochondria: a novel target for AD therapeutics
Given the important roles of mitochondria mtDNA, energy metabolism, and QC in the regulation of microglial function, mitochondrial dysfunction has emerged as a promising therapeutic target of AD. To date, therapeutic strategies/drugs that alleviate mitochondrial dysfunction mainly focus on modulating the Warburg effect, restoring mitochondrial fission/fusion balance, and promoting mitophagy in brain cells, especially microglia.
The Warburg effect of immune cells is a key pathological change of AD, and several compounds that modulate aerobic glycolysis have been applied on AD cell and animal models. 2-deoxy-D-glucose (2-DG), a glucose analog with the 2-hydroxyl group replaced by hydrogen, is a well-known inhibitor of glycolysis since it binds to, but cannot be phosphorylated by, hexokinase [
125]. Seven-week-dietary intervention of 2-DG has been reported to promote ketogenesis and maintain alternative mitochondrial bioenergetic pathway in both microglia and neurons, and thus, enhance phagocytic capacity of microglia to decrease Aβ burden and oxidative stress in AD mouse models [
125]. Besides 2-DG, dimethyl fumarate (DMF), a derivative of the TCA cycle intermediate fumarate, suppresses aerobic glycolysis via inactivating the catalytic cysteine of the glycolytic enzyme glyceraldehyde 3-phosphate dehydrogenase (GAPDH) in immune cells [
126]. DMF treatments have been found to ameliorate cognitive deficits, mitigate tauo-/amyloidopathy, and inhibit microglial oxidative/inflammatory responses presumably through modulating the activities of AMPK/SIRT-1, AKT/CREB/BDNF, AKT/GSK-3β, adiponectin/Adipo1R, and NF-κB/IL-1β/ROS trajectories in AD mouse models [
127]. In addition, treatment by mTOR signaling inhibitors rapamycin and metformin also inhibits Aβ-induced microglial metabolic reprogramming and inflammatory responses, and thus, mitigates AD phenotypes [
31]. Inspiringly, clinical trials to determine the therapeutic effects of aforementioned compounds on AD were initiated or complected (e.g., ClinicalTrials.gov Identifier: NCT04629495, NCT04200911). Results from these clinical studies will determine the feasibility of utilizing microglial metabolic reprogramming as a target for AD treatment.
Imbalanced mitochondrial fission/fusion is also a potential target of AD therapeutics. P110, a selective inhibitor of mitochondrial fission and fragmentation, significantly improved mitochondrial health and decreased Aβ levels in the brains, and ameliorated cognitive impairment of AD mouse models [
119]. Moreover, microglial activation induced by neurotoxic proteins were suppressed by P110 in a mechanism dependent on the inhibition of Drp1-Fis1 interaction. Besides the inhibition of excessive mitochondrial fission, the enhancement of mitochondrial fusion by cannabidiol has also demonstrated encouraging outcomes with regard to neuroinflammation suppression [
128]. Cannabidiol enhances the level of mitofusin 2 (Mfn2), a mitochondrial fusion protein, and improves mitochondrial function in microglia, therefore mitigating neuroinflammation-induced cognitive impairment [
128].
Another therapeutic strategy is to restore impaired mitophagy in AD. With the help of an in vivo drug screening platform using C. elegans, two potent mitophagy agonists have been identified, namely urolithin A (UA), a small natural compound prominent in pomegranate, and actinonin (AC), an antibiotic that induces mitophagy through mitochondrial ribosomal and RNA decay pathways [
117]. In different in vitro AD models, UA and CA, along with two other mitophagy enhancers, tomatidine and nicotinamide riboside, improved HT22 cell OXPHOS and protected cells from A- and p-Tau-induced mitochondrial and synaptic toxicities [
115,
116]. Moreover, UA and AC treatment restored impaired mitophagy, facilitated microglial phagocytic efficiency of Aβ, mitigated neuroinflammation, inhibited Tau hyperphosphorylation, and enhanced memory function in mouse models of AD [
117]. Similarly, the Fang group developed a screening workflow combining advanced artificial intelligence (AI) and classical wet laboratory approaches to identify multiple novel mitophagy inducers, including kaempferol and rhapontigenin, as potential therapeutics for AD [
129]. Both kaempferol and rhapontigenin forestall memory loss and ameliorate Aβ and Tau pathologies in 3 × TgAD mice via increasing microglial phagocytosis and inducing mitophagy [
129]. Besides, NAD
+-boosting compounds, such as nicotinamide riboside (NR) and Olaparib (AZD), significantly reduced Aβ proteotoxicity via inducing mitophagy in Aβ-expressing neuroblastoma cells and AD C. elegans and mouse models [
62]. The treatment of melatonin also reversed pathologic phagocytosis of microglia and mitochondrial energy metabolism highly likely through restoration of mitophagy by improving mitophagosome–lysosome fusion via Mcoln1, therefore attenuating Aβ pathology and improving cognition [
130]. Notably, there are other mitophagy enhancers, such as Quercetin (Qu), a natural flavonoid, that has displayed promising anti-inflammatory effects on depression and neurodegeneration, implying Qu as a potential therapeutic for AD [
131]. These findings support that restoration of mitophagy plays a neuroprotective role in AD presumably through microglia-mediated neuroinflammation and Aβ plaque elimination, making mitophagy a promising therapeutic target for AD treatment [
13,
24].
Besides, there are many other proteins, peptides, and peptide mimetics that were identified as mitochondrial targeting systems [
132]. The therapeutic outcomes of these molecules in AD are with great scientific research and clinical application value. Thus, comprehensive investigations are urgently needed to further evaluate the therapeutic effect of these mitochondrial function-regulating drugs in different AD models and AD patients if applicable.
Conclusions and future directions
In summary, multiple hypotheses have been proposed for the pathogenesis of AD, one of the most complicated and progressive neurodegenerative disease, and among them, mitochondrial dysfunction has emerged as a hotspot. Growing evidence have demonstrated the tight association of mitochondrial dysfunction and microglia-driven neuroinflammation. Aβ deposition and other pathological changes damage mtDNA, disturb mitochondrial membrane permeability, alter mitochondrial metabolism and QC, leading to microglial activation and neuroinflammation. Moreover, when severely damaged mitochondria are not properly removed by mitophagy, activated microglia release harmful mitochondrial contents such as ROS and reactive nitrogen into the extracellular environment, which damages surrounding neurons and astrocytes to amplify the inflammatory responses. Therefore, microglial mitochondrial dysfunction-driven neuroinflammation causes neuronal loss and neural circuit disorder, resulting in AD ultimately. Inspiringly, strategies that aim to correct mitochondrial dysfunction in activated microglia have obtained convincing outcomes in vitro and in vivo, indicating mitochondrial dysfunction as a promising target for AD treatment.
Although tremendous progress has been made in microglial mitochondria and AD research, there are many open questions remaining to be answered in the future. First, whether there are abnormal movement of mitochondria in microglia in AD? The Reddy group demonstrated loss of mitochondrial axonal transport as an important cause for synaptic degeneration and cognitive decline in AD [
24,
25]. It is interesting to investigate the intracellular localization changes of mitochondria and their biological outcomes when microglia are exposed to Aβ and p-Tau. Second, whether current mitophagy enhancers and other relevant compounds have similar therapeutic effects on AD patients? Due to the paucity of AD animal models that are sufficiently similar to humans, the outcomes of current strategies that aim to restore mitochondrial function in AD on human remain unknown. This point could be addressed by multi-center clinical cohort studies or partially answered utilizing non-human primate modes of AD [
133]. Third, is there any approach to deliver drug candidates for restoring mitochondrial function to microglia specifically? One option is to package these candidates into artificial or modified natural nanoparticles for targeted delivery, although the targeting efficiency requires to be further improved [
133,
134]. Fourth, is there mitochondrial exchange between microglia and other types of brain cells particular neurons in AD? Recent studies reported transfer of mitochondria from astrocytes to neurons post stroke that participates in neuroprotection and neurorecovery [
135]. It is interesting to examine the existence of mitochondria transportation pathway between microglia and neurons, and the biological and pathological functions of this pathway in AD. Hence, more in-depth investigations on microglial mitochondria will provide a novel perspective for the pathogenesis of AD, filling a significant knowledge gap and promoting the development of effective therapeutic interventions and accurate early diagnosis of AD.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.