Mitophagy is the highly specialized form of autophagy responsible for the selective degradation of damaged, dysfunctional, or superfluous mitochondria. This process represents a critical quality control mechanism that maintains cellular homeostasis by eliminating mitochondria that have accumulated damage from normal metabolic activity, reactive oxygen species, or mutation. The term "mitophagy" was first coined in 1998 when researchers observed mitochondrial degradation during nutrient deprivation, and subsequent research has revealed mitophagy as a fundamental process essential for cellular health in organisms ranging from yeast to humans [@mizushima1998].
The importance of mitophagy in the nervous system cannot be overstated. Neurons are uniquely dependent on mitochondrial quality control for several reasons: they are post-mitotic cells that cannot divide to dilute accumulated damage, they have extreme longevity with some neurons surviving for the entire human lifespan, and they have exceptionally high energy demands requiring constant mitochondrial ATP production. These characteristics mean that neurons must rely on robust quality control mechanisms—including mitophagy—to maintain function over decades of continuous activity.
Mitochondrial dysfunction is among the earliest and most consistent pathological features in neurodegenerative diseases including Alzheimer's disease (AD), Parkinson's disease (PD), amyotrophic lateral sclerosis (ALS), and Huntington's disease (HD). The discovery that mutations in genes encoding mitophagy proteins cause familial forms of PD provided definitive genetic evidence that this pathway is central to disease pathogenesis [@narendra2008].
Mitochondrial quality control begins with the detection of damage. Multiple parallel sensing mechanisms ensure robust activation of mitophagy when mitochondria are compromised:
Mitochondrial membrane potential loss: The inner mitochondrial membrane maintains a critical electrochemical gradient (Δψm) essential for ATP synthesis. When this potential drops below approximately -150 mV—typically due to mitochondrial damage from oxidative stress, calcium overload, or mitochondrial DNA mutations—the translocase of outer membrane (TOM) complex fails to import PINK1 (PTEN-induced kinase 1). Instead, PINK1 accumulates on the outer mitochondrial membrane (OMM), where it can phosphorylate its substrates and initiate the mitophagy cascade.
Reactive oxygen species (ROS): Mitochondria are the primary source of cellular ROS as a byproduct of oxidative phosphorylation. Excessive ROS damages mitochondrial proteins, lipids, and mitochondrial DNA (mtDNA). Oxidatively damaged proteins require replacement, and oxidized cardiolipin exposure on the OMM serves as an "eat-me" signal. The accumulation of mitochondrial ROS is particularly damaging in neurons due to their high metabolic rate and relatively low antioxidant capacity.
Mitochondrial DNA damage: Damage to mtDNA can result in the release of mitochondrial dsRNA and dsDNA into the cytoplasm, where they activate innate immune sensors including cGAS and MDA5. This triggers inflammatory responses and provides additional motivation for mitochondrial turnover—a mechanism potentially relevant to the neuroinflammation observed in neurodegenerative diseases.
Calcium dysregulation: Excessive cytosolic calcium accumulation in mitochondria leads to opening of the mitochondrial permeability transition pore (mPTP), causing membrane potential collapse and releasing intermembrane space proteins that signal damage.
The PINK1-Parkin pathway represents the best-characterized mechanism of mitophagy initiation, and mutations in both genes cause familial PD [@kitada1998]:
PINK1 stabilization: Under normal conditions, newly synthesized PINK1 is imported into mitochondria through the TOM and TIM (translocase of inner membrane) complexes and rapidly degraded by proteases in the inner membrane. When mitochondrial membrane potential collapses, import is blocked and PINK1 accumulates on the OMM. Here, PINK1 can phosphorylate its substrates, including Parkin and ubiquitin.
Parkin recruitment and activation: PINK1 directly phosphorylates Parkin (encoded by the PARK2 gene) at Ser65 within its ubiquitin-like domain, triggering a conformational change that activates its E3 ligase activity. PINK1 also phosphorylates ubiquitin at Ser65, creating phospho-ubiquitin chains that enhance Parkin activation in a feedforward manner.
Ubiquitin chain formation: Activated Parkin ubiquitinates multiple OMM proteins including:
Autophagy receptors bridge ubiquitinated mitochondria to the forming autophagosome. Multiple receptors participate in mitophagy:
p62/SQSTM1: The most extensively studied autophagy receptor, p62 contains a PB1 domain for oligomerization, a ZZ-type zinc finger domain, a TBK1-binding domain, an LIR motif, and a C-terminal UBA domain that binds ubiquitin chains. p62 also has the ability to sequestrate ubiquitinated cargo into aggregates, facilitating their delivery to autophagosomes.
OPTN (Optineurin): This receptor is particularly important in neurons due to its calcium-sensitive binding to ubiquitin. OPTN mutations cause familial ALS, highlighting its relevance to neurodegeneration. TBK1 phosphorylates OPTN, enhancing its affinity for ubiquitin chains and LC3.
NDP52 (CALCOCO2): Originally characterized as a binding partner for damaged DNA, NDP52 specifically recognizes ubiquitin chains generated by Parkin and recruits TANK-binding kinase 1 (TBK1) to initiate autophagy.
TAX1BP1: This receptor cooperates with p62 and NDP52 in mitophagy, with phosphorylation enhancing its function.
Once receptors are recruited, the autophagosome formation machinery is engaged:
LC3 lipidation: LC3 (microtubule-associated protein 1A/1B-light chain 3) is synthesized as pro-LC3 and processed by ATG4 proteases to LC3-I. ATG7 activates LC3-I (E1-like), ATG3 transfers it to phosphatidylethanolamine (PE), generating LC3-II which is incorporated into the expanding phagophore membrane.
ATG5-ATG12 conjugation: The ATG12-ATG5-ATG16L1 complex functions as an E3 ligase, facilitating LC3 lipidation at the site of autophagosome nucleation.
Phagophore expansion: The isolation membrane grows by the addition of lipids from multiple sources, including ER-mitochondria contact sites (MAMs), Golgi-derived vesicles, and plasma membrane-derived vesicles.
Lysosomal fusion: The mature autophagosome fuses with lysosomes through the action of SNARE proteins (STX17, SNAP-29, VAMP8), V-ATPases (proton pump), and lysosomal membrane proteins.
While the PINK1-Parkin pathway is the best characterized, multiple alternative pathways can induce mitophagy:
BNIP3 (BCL2/adenovirus E1B 19kDa interacting protein 3) and its homolog NIX (BNIP3L) are OMM proteins that can directly induce mitophagy independent of ubiquitination. Both contain LIR motifs that interact with LC3/GABARAP proteins. BNIP3 is induced by hypoxia through HIF-1α, making it particularly relevant to the hypoxic microenvironment of neurodegenerative disease brains.
FUNDC1 (FUN14 domain-containing protein 1) is an OMM protein with a LIR motif that can bind LC3 under stress conditions. FUNDC1 is regulated by phosphorylation—Src kinase phosphorylates FUNDC1 at Tyr18, inhibiting its interaction with LC3. Dephosorylation by PGAM5 activates FUNDC1-mediated mitophagy.
Ceramide, a sphingolipid that accumulates in neurodegenerative conditions, can directly activate mitophagy through multiple mechanisms. Ceramide binds to LC3 and also inhibits mTORC1, creating a pro-autophagic environment.
Mitochondrial movement along microtubules is coupled to mitophagy quality control. Damaged mitochondria are stationary, while healthy mitochondria are mobile. The Miro1 protein links mitochondria to motors, and its degradation by Parkin is essential for mitophagy initiation.
Alzheimer's disease represents the most common cause of dementia worldwide. While the amyloid cascade hypothesis has dominated AD research for decades, emerging evidence highlights mitochondrial dysfunction and impaired mitophagy as critical early events in disease pathogenesis. The relationship between mitophagy and AD is bidirectional—amyloid-beta and tau pathology both impair mitophagy, while failing mitophagy accelerates protein aggregation.
Amyloid-β (Aβ) peptides directly impact mitochondrial function and mitophagy at multiple levels. Aβ oligomers bind to mitochondrial proteins, disrupting the electron transport chain and causing mitochondrial membrane potential loss. This membrane depolarization should theoretically activate PINK1-Parkin mitophagy, but Aβ actively suppresses this pathway. Aβ interferes with PINK1 stabilization on the OMM by altering mitochondrial membrane lipid composition and reduces Parkin expression at both mRNA and protein levels through transcriptional dysregulation.
Aβ also disrupts mitochondrial dynamics by altering the expression and post-translational modification of fusion and fission proteins. Mitochondrial fragmentation is observed in AD neurons and is exacerbated by impaired mitophagy. The MFN1/2 and OPA1 proteins involved in fusion are downregulated, while fission protein DRP1 is upregulated and hyperactive.
Neurofibrillary tangles composed of hyperphosphorylated tau protein are the second major pathological hallmark of AD. Tau pathology correlates more strongly with cognitive decline than amyloid plaques, and tau directly impairs mitophagy through multiple mechanisms. Hyperphosphorylated tau accumulates in mitochondria where it binds to and inhibits key mitophagy proteins. Tau physically interacts with PINK1, reducing its ability to accumulate on damaged mitochondria. Parkin recruitment is also impaired in the presence of tau pathology, as tau interferes with ubiquitin chain formation on OMM proteins.
Furthermore, tau pathology disrupts the cytoskeletal infrastructure necessary for mitophagy. Tau hyperphosphorylation destabilizes microtubules, which are required for autophagosome transport and lysosomal fusion.
Post-mortem studies of AD brains consistently show signs of impaired mitophagy. LC3-II levels are elevated in AD brains, suggesting a block in autophagic flux rather than reduced initiation. This is consistent with the accumulation of autophagic vacuoles observed in AD neurons, which represent failed degradation rather than increased initiation.
PINK1 and Parkin protein levels are reduced in AD brain tissue, particularly in regions with high tau pathology (entorhinal cortex, hippocampus). Genetic studies have identified variants in mitophagy-related genes as risk factors for AD, including PINK1 and PARK2 polymorphisms that modify age of onset.
Parkinson's disease is the second most common neurodegenerative disorder. The discovery that mutations in PINK1 (PARK6) and Parkin (PARK2) cause familial PD provided definitive evidence that mitophagy dysfunction is central to disease pathogenesis. Unlike AD, where mitophagy impairment is likely secondary to protein aggregation, in PD, primary genetic defects in the mitophagy machinery are sufficient to cause disease.
Biallelic loss-of-function mutations in PINK1 account for approximately 1-2% of early-onset familial PD and up to 5% of early-onset sporadic cases. PINK1 deficiency prevents the initiation of mitophagy in response to mitochondrial damage, leading to progressive accumulation of dysfunctional mitochondria.
PINK1 knockout mice show subtle mitochondrial abnormalities but do not develop overt dopaminergic neuron loss, suggesting compensatory mechanisms in mice that are absent in humans. However, PINK1-deficient Drosophila develop dramatic mitochondrial pathology and dopaminergic neuron degeneration.
Parkin mutations cause autosomal recessive juvenile-onset PD in approximately 50% of cases. Over 200 pathogenic mutations have been identified throughout the gene, including point mutations, deletions, and duplications. Most pathogenic mutations disrupt the E3 ligase activity required for ubiquitin chain formation.
Parkin-deficient models show accumulation of damaged mitochondria, increased oxidative stress, and progressive dopaminergic neuron loss. Interestingly, parkin deficiency also leads to increased susceptibility to mitochondrial toxins including MPTP and rotenone.
LRRK2 (leucine-rich repeat kinase 2) mutations are the most common cause of familial PD, accounting for 5-10% of cases. LRRK2 phosphorylates multiple autophagy proteins including p62, OPTN, and ATG16L1. PD-associated LRRK2 mutations impair mitophagy through multiple mechanisms. Enhanced LRRK2 kinase activity leads to hyperphosphorylation of autophagy receptors, disrupting their function.
Heterozygous mutations in GBA (glucocerebrosidase) are the most significant genetic risk factor for PD, increasing risk 5-20 fold. GBA mutations impair lysosomal function, reducing the degradative capacity of the mitophagy pathway.
α-Synuclein aggregation into Lewy bodies is the pathological hallmark of PD. Oligomeric α-synuclein inhibits PINK1 stabilization on the OMM by interfering with the TOM complex. It also prevents Parkin recruitment and disrupts autophagosome-lysosome fusion. The relationship between α-synuclein and mitophagy is bidirectional—impaired mitophagy leads to α-synuclein accumulation, while α-synuclein aggregation further impairs mitophagy.
ALS is associated with mutations in genes encoding autophagy receptors (OPTN, TBK1, p62) and mitophagy proteins. OPTN mutations are a cause of familial ALS, directly linking mitophagy to motor neuron disease. TDP-43 proteinopathy, the hallmark inclusion in ALS, disrupts autophagic flux.
Mutant huntingtin protein impairs mitophagy by sequestering key autophagy proteins into aggregates. PINK1 and Parkin function is compromised in HD models, and enhancing mitophagy has shown promise in cellular and animal models.
FTD shares many pathological features with ALS, including TDP-43 inclusions and autophagy dysfunction. Mutations in genes regulating autophagy (GRN, CHMP2B) cause familial FTD.
Multiple therapeutic strategies are being developed to enhance mitophagy in neurodegenerative disease:
mTOR inhibitors: Rapamycin and rapalogs inhibit mTORC1, relieving its repression of autophagy initiation. These compounds enhance mitophagy in models but have significant immunosuppressive side effects.
NAD+ boosters: Nicotinamide riboside (NR) and nicotinamide mononucleotide (NMN) increase NAD+ levels, activating sirtuins and enhancing mitochondrial quality control.
USP30 inhibitors: USP30 is a deubiquitinase that removes ubiquitin chains from mitochondria, opposing Parkin activity. USP30 inhibitors are in preclinical development.
Natural compounds: Resveratrol, curcumin, and other polyphenols enhance mitophagy through AMPK activation.
USP25 inhibition: A groundbreaking 2026 study demonstrated that USP25 inhibition restores mitophagy and ameliorates Parkinson's disease [@xu2026]. USP25 is a deubiquitinase that specifically removes ubiquitin from mitochondrial proteins, opposing PINK1-Parkin signaling. Unlike USP30 (which targets TOM complex), USP25 acts on a broader set of mitochondrial substrates. Small molecule USP25 inhibitors are now in development as PD therapeutics.
Viral vector delivery of PINK1, Parkin, or autophagy receptors is being explored. AAV-mediated gene delivery has shown promise in animal models.
High-throughput screens have identified small molecules that enhance mitophagy, including:
A Phase 2 trial of urolithin A in PD (NCT04578101) has completed enrollment, with results pending. Nicotinamide riboside has been evaluated in AD with mixed results.
The development of biomarkers for mitophagy status in humans is critical for patient selection and response monitoring:
Molecular biomarkers: Circulating mitochondrial DNA levels, mitophagy-associated proteins in extracellular vesicles, and mitochondrial-derived peptides.
Imaging biomarkers: PET ligands targeting autophagy components or mitochondrial proteins are in development.
Functional biomarkers: Measurements of mitochondrial respiration and ATP production can indirectly assess mitophagy function.
Developing robust biomarkers for mitophagy in humans remains challenging. The inability to directly measure mitophagy in the human brain limits patient selection and response monitoring in clinical trials.
Neurons and astrocytes have different mitophagy requirements and vulnerabilities. Understanding cell-type-specific regulation is essential for developing targeted therapies.
Aging is the primary risk factor for neurodegenerative diseases and is associated with diminished mitophagy capacity. The relationship between aging-related mitophagy decline and disease onset is an active research area.
The following diagram shows the key molecular relationships involving Mitophagy discovered through SciDEX knowledge graph analysis: