Genetic factors and comorbid pathologies interact to drive regional mitophagy alterations in Lewy body dementia

Lewy body dementia (LBD) is a complex neurodegenerative disorder and the second most common cause of dementia in the elderly after Alzheimer's disease (AD). LBD lies on the spectrum between Parkinson's disease (PD) and AD, yet maintains distinct clinical and pathological features. Patients with LBD typically present with progressive dementia and often experience a combination of cognitive decline, spontaneous parkinsonism, visual hallucinations, and REM sleep behavior disorder [1]. Neuropathologically, LBD is characterized by the prevalent accumulation of misfolded α-synuclein in the form of Lewy bodies (LBs) and Lewy neurites [2], which extend beyond the brainstem into the limbic system and various cortical regions [3, 4]. In contrast to AD which is defined by a strong deposition of tau neurofibrillary tangles (NFTs) and amyloid β-containing senile plaques (SPs), LBD cases typically present with a mild to moderate burden of AD-type pathology, particularly for tangle-related changes [3, 5, 6]. This convergence of multiple proteinopathies contributes to the clinical and pathological heterogeneity of LBD and complicates the diagnosis, treatment, and understanding of disease mechanisms. Recent genetic studies have identified several risk loci associated with LBD, including variants in APOE, GBA, and SNCA [7 –11]. These findings suggest that LBD shares overlapping genetic risk factors with both PD and AD, while also presenting a unique genetic architecture that may influence disease onset, progression, and therapeutic response.

Mitochondrial dysfunction is increasingly recognized as a shared and early feature across aging and neurodegeneration, where it may contribute to the misfolding and aggregation of pathological proteins such as α-synuclein, tau, and amyloid β [12 –14]. To preserve mitochondrial integrity and cellular homeostasis, cells employ a selective form of autophagy, named mitophagy, to target damaged mitochondria for lysosomal degradation [15]. The best-characterized mitophagy pathway is directed by the ubiquitin kinase–ligase pair PINK1–PRKN [16, 17]. In this pathway, PINK1 accumulates on the outer mitochondrial membrane when its protein import is blocked and subsequently recruits PRKN, which catalyzes the ubiquitination of mitochondrial surface proteins. A critical step in this cascade is the phosphorylation of ubiquitin at serine 65 (pS65-Ub) by PINK1, which serves as a molecular tag on damaged mitochondria to facilitate the recruitment of autophagic receptors and promote their subsequent lysosomal clearance [18]. Under physiological conditions, PINK1–PRKN signaling ensures the efficient turnover of pS65-Ub tagged dysfunctional mitochondria, thereby maintaining relatively low baseline levels of pS65-Ub. Although typically transient, pS65-Ub levels markedly increase with normal aging in mice and in human brain [18 –20], which may result from increased mitochondrial damage and activation of PINK1–PRKN and/or from impaired degradation and inefficient flux through lysosomes. Changes in mitophagy flux during aging have recently been shown to differ across brain regions and even among specific cell types in mice [21], suggesting the presence of complex and diverse regulatory mechanisms.

Independently of the normal aging contribution, pS65-Ub levels are even further elevated in a variety of neurodegenerative diseases, including AD, PD, and frontotemporal dementia with parkinsonism [20, 22 –26]. Across this spectrum of diseases the accumulation of pS65-Ub significantly associates with α-synuclein and tau pathology [20, 23]. Given the frequent co-occurrence of LB, SP and NFT pathologies in LBD, we here aimed to assess their independent association with the mitophagy marker pS65-Ub and analyze how the core α-synuclein pathology interacts with comorbid AD-type pathology (tau or amyloid β) to impact mitophagy alteration using a cohort of 371 autopsy-defined LBD cases. We recently employed pS65-Ub levels as a quantitative disease endophenotype and through a genome-wide association study identified two genetic modifiers with opposing effects [27]. As a major disease risk factor, APOE rs429358 (APOE4) is strongly associated with higher pS65-Ub levels and greater pathological burden, while ZMIZ1 rs6480922 is associated with decreased pS65-Ub, greater brain weight, and reduced hippocampal neuropathologies, suggesting a potential resilience role. Here, we further assessed interactions between the two genetic mitophagy modifiers and the comorbid pathologies in relation to pS65-Ub levels in our LBD series.

We focused our analysis on the hippocampus and the amygdala due to their critical roles in memory and emotional regulation as well as their central involvement in the clinical manifestation of LBD including cognitive fluctuations, memory impairment, and neuropsychiatric symptoms [1, 28 –31]. More importantly, these two regions are among the early and most consistently affected regions in LBD, harboring all three types of neuropathologies including LBs SPs, and NFTs [4, 32]. In a pathologically defined large autopsy LBD cohort, our results indicate complex regional interactions between genetics, biology, and pathology. While each of the three main pathologies significantly and independently contributes to pS65-Ub alterations regardless of brain region, our findings further highlight a synergistic interaction between LB and NFT pathology specifically within the amygdala that may be partially mediated by ZMIZ1 rs6480922. In the future, it will be critical to define the biological and/or pathological mechanisms underlying the abnormal buildup of pS65-Ub signals in disease brains and to identify therapeutic targets that can restore mitophagy flux and attenuate disease progression.

In this study, using an autopsy-defined LBD cohort and pS65-Ub as an aging and disease marker, we revealed brain region-specific mitophagy alteration and distinct neuropathological profiles in the hippocampus and amygdala of LBD. We demonstrated strong, independent, yet differential associations between pS65-Ub positive cell density and α-synuclein, amyloid, and tau pathology in these regions. Multivariable linear regression models indicated that APOE4 significantly amplifies the pathological impact of tau pathology-mediated mitophagy dysfunction specifically in the hippocampus, suggesting a new molecular mechanism by which APOE4 contributes to the pathogenesis in LBD and AD. Furthermore, using two- and three-way interaction analysis we identified a region-specific synergistic interaction between α-synuclein and tau pathology, but not with amyloid pathology, in driving pS65-Ub levels in the amygdala, and nominated ZMIZ1 rs6480922 as a potential genetic driver in this pathological interplay. Our study defines regional- and pathological-specific synergism among genetics, biology, and pathology in LBD brains. While contributing to the growing recognition of mixed pathologies in neurodegenerative diseases, these findings underscore mitophagy as a potential biological convergent underlying disease pathogenesis.

pS65-Ub is produced in response to mitochondrial damage through the joint activity of PINK1 and PRKN. While the pathway remains active, pS65-Ub levels stay very low and transient under physiological conditions [34], reflecting the effective "piecemeal" elimination of mitochondrial subdomains that undergo normal wear and tear. The same mechanisms apply, just to a greater extent, upon stress-induced mitochondrial damage such as acute depolarization in cell culture or cumulative cellular stress during aging. This leads to an increased generation of pS65-Ub that may then remain elevated with sustained mitochondrial damage and/or compromised autophagic-lysosomal flux of labeled mitochondria.

This accumulation of pS65-Ub could be further exacerbated in neurodegenerative conditions such as AD, PD, and LBD with the presence of additional pathological, non-degradable substrates. Despite LBs being reportedly absent in some PINK1 or PRKN mutant PD cases, the high abundance of pS65-Ub in non-mutation carriers suggests that the PINK1–PRKN pathway remains active in LBD and even has a complex, disease-context-dependent interaction with LB pathology. While the precise mechanisms underlying the formation and function of these structures remain unclear, recent studies have found enriched mitochondrial and autophagic membrane remnants in LBs in addition to α-synuclein [35, 36], pointing to a potential link with failed mitophagy that warrants further investigation. It is also worth mentioning that although pS65-Ub typically appears as smaller granular structures, it tends to accumulate within larger granulovacuolar degeneration bodies when pre-tangles pathology is present in the same cell [27]. These structures are thought to be late-stage autophagic remnants of incomplete degradation though their functional significance is not yet fully understood [37 –39]. As such the mechanisms impairing mitophagy may differ between cells positive for α-synuclein pathology, tau pathology, or both.

We found in LBD that pS65-Ub levels were further elevated in the hippocampus compared to the amygdala. This selective vulnerability of the hippocampus may reflect its heightened bioenergetic demands to sustain synaptic remodeling and adult neurogenesis [40 –42], processes that rely heavily on mitochondrial integrity and are particularly susceptible to age- and disease-related dysfunction [41, 43, 44]. Alternatively, this elevation may result from a greater tau burden in the hippocampus given the stronger association between pS65-Ub levels and NFT density than with LB or SP density. Emerging evidence consistently supports a close interplay between tau pathology and mitophagy dysfunction in neurodegenerative diseases [23, 45 –49]. Besides the common link to granulovacuolar degeneration bodies, disease-associated tau has been shown to disrupt mitophagy by interfering with PRKN functions in cell and animal models [23, 45, 47, 48]. PRKN overexpression was able to enhance mitophagy flux and reduce tau pathological changes [48].

The APOE4 allele is a major common genetic risk factor for LBD [7, 8, 11] and has recently been identified as a strong mitophagy modifier that is associated with increased levels of pS65-Ub [27]. The APOE4 genotype is also well known to result in enhanced tau hyperphosphorylation, aggregation, spreading, and cytotoxicity [50]. Our interaction analysis revealed that the APOE4 genotype specifically exacerbated the impact of tau pathology-mediated pS65-Ub accumulation in the hippocampus of LBD. However, as a primary lipid transporter in the brain, APOE broadly regulates lipid homeostasis in an isoform-dependent manner across various cell types [51]. Disease-associated APOE4 promotes lipid accumulation, glia reactivity, oxidative stress, and metabolic shift [51 –55], and has been shown to enhance pS65-Ub levels even in model systems without neuropathology [27]. These combined biological and pathological effects of APOE4 likely converge to exacerbate mitochondrial and lysosomal dysfunction in the pathological environment and might thus promote additional pS65-Ub accumulation in the hippocampus.

We observed a significant synergy of α-synuclein and tau with regard to pS65-Ub accumulation specifically in the amygdala. Besides a strong presence of LB pathology in this region, it may partly be due to the more prominent co-occurrence of both neuropathologies in the same neurons in the amygdala [56, 57]. Pathologic α-synuclein and tau are also known to induce mitochondrial dysfunction and oxidative stress, with tau aggregates amplifying the neurotoxicity on mitochondria by interacting with α-synuclein [58, 59]. This may occur through a convergent action of these two proteins [60] on altering Drp1 localization and disrupting axonal trafficking, both of which are essential for maintaining mitochondrial homeostasis [61, 62]. Additionally, α-synuclein and tau have been reported to interact with autophagic-related proteins such as LC3 and p62/SQSTM1 [63, 64]. Both increased mitochondrial damage and impaired autophagic function may thus underlie the observed synergistic effect of LB and NFT pathology on increasing pS65-Ub in the amygdala.

Our three-way interaction analysis suggests that the ZMIZ1 rs6480922 variant may be a key driver for the observed synergy. This variant was previously found to be significantly associated with reduced pS65-Ub levels in our genome-wide association study [27]. Interestingly, carriers of the minor allele of rs6480922 in the current study again showed lower pS65-Ub levels, but only in subgroups with high NFT or high LB alone, not in the subgroup with high levels of both pathologies (Fig. 4b). ZMIZ1 encodes a transcriptional coactivator that enhances transcription mediated by the androgen receptor, p53, SMAD3, and NOTCH signaling [65]. Pathogenic mutations in ZMIZ1 have been linked to syndromic neural disorders with intellectual disability and neurodevelopmental delay [65]. Of note, knockout of ZMIZ1 in zebrafish resulted in altered expression of autophagy genes and accumulation of mitochondrial DNA [66]. ZMIZ1 was also recently shown to inhibit autophagy in cultured neurons via sumoylation of NLRP3 [67]. However, the biological significance of ZMIZ1 in brain mitophagy, its molecular contribution to the synergy between LB and NFT pathology, and its potential context-dependent protective role against mitophagy dysfunction remain largely unexplored and warrant future investigation.

Although we reveal a synergistic interplay between genetic factors and comorbid pathologies that drives regional vulnerability to mitophagy alterations in the autopsy-defined LBD brain, our study also has certain limitations. Further analysis using complementary, independent methods beyond the primary immunohistochemical readout would strengthen and expand the current findings. By characterizing pS65-Ub labeled mitochondrial damage in postmortem brain, we provide a snapshot of the cumulative burden on PINK1–PRKN-directed mitochondrial quality control at the end-stage of disease. Spatial multiomics analysis at single-cell resolution will help better define pS65-Ub substrates and resolve the complexity of these interactions at the subcellular level. To dissect the temporal and mechanistic contributions to the synergy between genetic variants and neuropathologies, dynamic, real-time analyses in model systems are needed for precise monitoring of mitochondrial damage, pS65-Ub production, trafficking of labeled cargo, and autophagic degradation over the course of disease progression. Our current analysis was also limited to mature inclusions (LBs, SPs, and NFTs), while emerging evidence suggests that oligomeric species or fibrils may exert greater cytotoxicity than mature inclusions [68 –71]. Future studies on these pathological species will provide more insight into the pathological impact on mitophagy and define potential pathological, non-degradable pS65-Ub substrates. Other comorbid pathologies such as TDP-43 and cerebral amyloid angiopathy, both implicated in different aspects of mitochondrial function [72 –75], may also be considered in future studies. While pS65-Ub is an intriguing aging and disease marker, much remains to be resolved to elucidate the nature and origin of the pS65-Ub signal and to refine our interpretation of its accumulation. Nevertheless, given the clinical, genetic, and pathological overlap among PD, LBD, and AD, targeting mitochondrial health and mitophagy may offer a promising common therapeutic strategy. A better understanding of the molecular architecture of pathological aggregates and their impact on mitochondrial homeostasis could refine our approaches to develop disease-modifying therapies.

Below is the link to the electronic supplementary material. Supplementary file1 (PDF 679 KB)