The finding that mutations in the Gaucher’s Disease (GD) gene GBA1 are a strong risk factor for Parkinson’s Disease (PD) has allowed for unique insights into pathophysiology centered on disruption of the autophagic-lysosomal pathway. Protein aggregations in the form of Lewy bodies and the effects of canonical PD mutations that converge on the lysosomal degradation system suggest that neurodegeneration in PD is mediated by dysregulation of protein homeostasis. The well-characterized clinical and pathological relationship between PD and the lysosomal storage disorder GD emphasizes the importance of dysregulated protein metabolism in neurodegeneration, and one intriguing piece of this relationship is a shared phenotype of autophagic-lysosomal dysfunction in both diseases. Translational application of these findings may be accelerated by the use of midbrain dopamine neuronal models derived from induced pluripotent stem cells (iPSCs) that recapitulate several pathological features of GD and PD. In this review, we discuss evidence linking autophagic dysfunction to the pathophysiology of GD and GBA1-linked parkinsonism and focus more specifically on studies performed recently in iPSC-derived neurons. The finding that mutations in the Gaucher’s Disease (GD) gene GBA1 are a strong risk factor for Parkinson’s Disease (PD) has allowed for unique insights into pathophysiology centered on disruption of the autophagic-lysosomal pathway. Protein aggregations in the form of Lewy bodies and the effects of canonical PD mutations that converge on the lysosomal degradation system suggest that neurodegeneration in PD is mediated by dysregulation of protein homeostasis. The well-characterized clinical and pathological relationship between PD and the lysosomal storage disorder GD emphasizes the importance of dysregulated protein metabolism in neurodegeneration, and one intriguing piece of this relationship is a shared phenotype of autophagic-lysosomal dysfunction in both diseases. Translational application of these findings may be accelerated by the use of midbrain dopamine neuronal models derived from induced pluripotent stem cells (iPSCs) that recapitulate several pathological features of GD and PD. In this review, we discuss evidence linking autophagic dysfunction to the pathophysiology of GD and GBA1-linked parkinsonism and focus more specifically on studies performed recently in iPSC-derived neurons.
Dr. Mochizuki:
Protein aggregation and deposition are common characteristic of many neurodegenerative disorders, including Parkinson's disease (PD). Protein misfolding commonly occurs within cells, and cells have evolved a range of mechanisms to maintain intracellular protein homeostasis. As one of the major degradation pathways, autophagy mediates the elimination of abnormal or unwanted protein from cells. A growing number of studies in disease models and patients have implicated impaired autophagy machinery in the pathogenesis of PD. We have invited three experts in this field to discuss about the intriguing link between autophagy and PD-related genes, α-synuclein, parkin and GBA1 and the potential value of targeting autophagy pathway as a therapeutic strategy for PD.
What is your view on the inter-relationship between α-Syn and the autophagy-lysosomal pathway?
Evidence from human postmortem samples and cellular/animal models of PD link α-synucleinaccumulation to impaired clearance due to alterations in the proteolytic machineries. Is autophagy lysosomal pathway (ALP) rather than the ubiquitin-proteasome system (UPS) the main route responsible for α-synuclein degradation? How much does autophagy contribute to the evolution and progression of α-synuclein pathology? On the other hand, how does α-synuclein expression influence the function of autophagy lysosomal pathway under pathological conditions?
Dr. Xilouri:
Both UPS and ALP have been proposed to clear α-synuclein; however to a different extent and in a cell-, conformation- and tissue- specific manner. α-Synuclein can undergo both ubiquitin-dependent and ubiquitin-independent degradation via the 26S/20S proteasome, whereas only a small proportion of soluble-cell-derived intermediate α-synuclein oligomers, not including monomeric α-synuclein, are targeted to the 26S proteasome for degradation. In vivo, the UPS seems to be the main degradation route for α-synuclein under normal conditions, while with increased α-synuclein burden the ALP is recruited. On the other hand, the wild-type α-synuclein and not the PD-linked A53T and A30P forms, the phosphorylated or the dopamine-modified α-synuclein, is degraded via the selective process of chaperone-mediated autophagy (CMA). CMA can degrade only monomeric or dimeric forms of the protein, whereas macroautophagy is the only process that can clear oligomeric α-synuclein through the process of aggrephagy. Not only mutations but also post-translational modifications such as phosphorylation, sumoylation and ubiquitination may also alter the partitioning of α-synuclein to proteasomal or lysosomal degradation. Besides the UPS and the ALP, other proteases such as calpains and neurosin have been implicated in the cleavage of normal or aggregated forms of intracellular α-synuclein. Such cleavage may promote the generation of truncated α-synuclein species with pathogenic significance. Moreover, secreted neurosin and metalloproteinases have been found to cleave extracellular α-synuclein at selective sites, generating fragments with increased tendency to aggregate.
As total α-synuclein protein load is a critical determinant for the development of pathology, factors that control its levels, such as ALP, may affect the evolution and progression of synucleinopathies. To this regard, genetic and human post-mortem studies suggest that ALP is impaired in synucleinopathies and seems to be involved in α-synuclein aggregation. Disturbances in ALP may also affect the cell-to-cell transmission of α-synuclein, through increased secretion of free- or exosome-associated non-degraded forms of the protein, thus facilitating the propagation of α-synuclein-linked pathology in different brain regions of the CNS.
Autophagy not only degrades various forms/conformations of α-synuclein, but can also be a target of the protein’s aberrant effects. Increased α-synuclein has been reported to impair macroautophagy both in vitro and in vivo, possibly through its interaction with Rab1a, which causes the mislocalization of Atg9, an autophagosome formation-related protein. There is also evidence suggesting that the PD-linked α-synuclein mutations could affect macroautophagy machinery at different stages (macroautophagy initiation, formation of autophagosomes, autophagosome maturation and fusion of autophagosomes with the lysosome). Furthermore, the A53T mutation has been shown to dysregulate the selective removal of mitochondria through mitophagy, resulting in massive mitochondrial removal accompanied by bioenergetics deficits and neuronal degeneration. In regards to CMA, it has been reported that the A30P and A53T PD-linked mutant α-synuclein forms bind more tightly to LAMP2A, CMA’s specific receptor, but are not taken up and degraded within lysosomes, thus becoming toxic by inhibiting the CMA-mediated degradation of other cytosolic substrate proteins. Moreover, post-translational modifications of wild-type α-synuclein, such as oxidation and nitration of the protein, alter its binding and uptake into lysosomes, while phosphorylation and dopamine-modification almost completely prevents its CMA-dependent degradation. In addition, extracellular α-synuclein can trigger autophagic impairment in recipient cells upon its uptake, as observed by an accumulation of enlarged deficient lysosomes, further contributing to the spread of the pathology. Aberrant α-synuclein can also affect lysosomal activity per se, by altering the activity of lysosomal enzymes, such as cathepsins and GCase.
Whether ALP malfunction is the primary cause of synucleinopathies or occurs secondarily, in part through impairment of these pathways by aberrant α-synuclein species, remains to be studied further.
How does autophagy-lysosomal pathway dysfunction connect the link between GBA1 and PD?
Mutations in the Gaucher’s disease gene glucocerebrosidase (GBA1) increase risks for PD. Glucocerebrosidase (GCase) deficiency has also been observed in idiopathic PD patients without GBA1 mutation. Notably, autophagic lysosomal dysfunction is a shared phenotype in both diseases. How does glucocerebrosidase deficiency influence the autophagy lysosomal pathway and eventually contribute to PD pathogenesis?
Dr. Krainc:
Mechanistically, loss of GCase activity results in accumulation of glucosylceramide (GluCer) which stabilizes toxic soluble high molecular weight oligomeric forms of α-synuclein. Importantly, a similar effect on α-synuclein is not seen with other sphingolipids, and accumulation of these toxic intermediate forms of α-synuclein appears to be reversible with depletion of GluCer, suggesting that GluCer has a specific connection to α-synuclein accumulation. While reduced GCase activity results in α-synuclein accumulation, α-synuclein in turn can impair lysosomal GCase activity in neurons. In particular, α-synuclein disrupts trafficking of GCase and other lysosomal hydrolases from the ER, resulting in accumulation of the immature ER form of GCase and a decrease in post-ER GCase. Reduced lysosomal GCase activity in turn results in accumulation of lipid substrates and lysosomal dysfunction. Impaired trafficking of GCase in response to α-synuclein accumulation is at least in part mediated by the SNARE protein ykt6, which plays a key role in ER-Golgi trafficking. Ykt6 is inactivated by α-synuclein resulting in impaired trafficking of GCase and other lysosomal enzymes, and thus decreased GCase activity in the lysosomal compartment. Additional evidence has demonstrated that GCase and α-synuclein can directly interact with each other as well.
Since α-synuclein accumulation is common to all forms of PD, the effect of α-synuclein on lysosomal GCase activity can be seen even in those PD individuals without pathogenic GBA1 mutations. Indeed, GCase activity is decreased in the brains of PD patients both with and without GBA1 mutations, including patients with sporadic PD with no known pathogenic mutation. The bidirectional relationship between GCase activity and α-synuclein can result in a critical pathogenic feedback loop resulting in persistent α-synuclein accumulation and lysosomal dysfunction, ultimately leading to progressive neuronal loss.
Dysregulated metabolism of dopamine itself can at least in part explain the vulnerability of dopaminergic neurons in the synucleinopathies, as dopamine-dependent toxic cascade has been shown to cause lysosomal dysfunction. Interestingly, oxidized dopamine has a direct impact on GCase as evidenced by the finding that incubation of recombinant GCase with dopamine reduces GCase enzymatic activity, and that dopamine quinones modify cysteine residues in the catalytic site of this enzyme.
How are the autophagy receptor and autophagy-related proteins (ATGs) involved in parkin-mediated mitophagy under normal and pathological condition?
Mutations in Parkin cause early-onset PD. Parkin binds, ubiquitinates, and targets damaged mitochondria for autophagic clearance. Could you please update us on how autophagy receptors and autophagy-related proteins (ATGs) recruit ubiquitinated mitochondria to autophagosome formation site?
Dr. Matsuda:
In mammalian cells, the causative gene products of hereditary Parkinson's disease, PINK1 (a serine kinase) and Parkin (an E3 ubiquitin ligase), work together to ubiquitylate damaged mitochondria and by doing so trigger mitophagy (Ref 1, Onishi et al., 2021). Five types of autophagy adaptor proteins (p62/SQSTM1, NBR1, NDP52/CALCOCO2, OPTN/optineurin, and TAX1BP1) have been identified in mammalian cells that couple ubiquitylated proteins with autophagic degradation. In 2015, CRISPR/Cas9-mediated experiments revealed that two of these adaptors, NDP52 and OPTN, are critical for Parkin-dependent mitochondrial autophagy (i.e., mitophagy) (Ref 2, Lazarou et al., 2015). The next piece of the puzzle is to determine which autophagy-related factors (ATGs) cooperate with NDP52 and OPTN in mitophagy.
It is well-accepted that autophagy adaptors (p62/SQSTM1, NBR1, NDP52/CALCOCO2, OPTN/optineurin, and TAX1BP1) bind to both ubiquitin and LC3, and that these interactions play a role in autophagic degradation of ubiquitylated targets. LC3 is an evolutionary conserved autophagy-related protein that localizes on the isolation membrane and autophagosomes. In this scenario, the binding of LC3 by autophagy adaptors is thought to promote the recruitment of existing isolation membranes to areas in close vicinity to mitochondria as an early step in mitophagy.
The current model, however, is limited as it does not sufficiently explain all experimental results. Although all autophagy adaptors can be transported to damaged mitochondria and bind LC3 (Ref 3, Yamano et al., 2020), only NDP52 and OPTN are required for mitophagy (Ref 2, Lazarou et al., 2015). This suggests that NDP52 and OPTN have important functions other than “binding LC3” that promote ubiquitin-dependent mitophagy in mammalian cells.
Recently, NDP52 and OPTN were reported to bind autophagy initiation factors in addition to LC3 (Ref 3, Yamano et al., 2020; Ref 4, Vargas et al., 2019). A 2019 study found that a SKITCH domain in NDP52 binds FIP200 (Ref 4, Vargas et al., 2019), a component of the ULK1 kinase complex that acts as an autophagy initiation factor. Similarly, a 2020 study showed that OPTN binds to ATG9 via a leucine zipper domain (Ref 3, Yamano et al., 2020). ATG9 is unique among ATG proteins in that it is a multiple-transmembrane protein that is essential for autophagy and is thought to be play a role the expansion of vesicles (ATG9 vesicles) that serve as seeds for the autophagosome.
In summary, two processes are important for Parkin-dependent mitophagy in mammalian cells. One is NDP52/OPTN dependent binding of LC3, which recruits existing isolation membranes to mitochondria. The other process is also NDP52/OPTN dependent but involves the binding of autophagy initiation factors such as ATG9 and FIP200. These interactions are important for promoting the accumulation of ATG9 vesicles and ULK1 complexes near mitochondria, which is required for de novo synthesis of the isolation membrane and more efficient induction of mitophagy.
The finding that mutations in the Gaucher’s Disease (GD) gene GBA1 are a strong risk factor for Parkinson’s Disease (PD) has allowed for unique insights into pathophysiology centered on disruption of the autophagic-lysosomal pathway. Protein aggregations in the form of Lewy bodies and the effects of canonical PD mutations that converge on the lysosomal degradation system suggest that neurodegeneration in PD is mediated by dysregulation of protein homeostasis. The well-characterized clinical and pathological relationship between PD and the lysosomal storage disorder GD emphasizes the importance of dysregulated protein metabolism in neurodegeneration, and one intriguing piece of this relationship is a shared phenotype of autophagic-lysosomal dysfunction in both diseases. Translational application of these findings may be accelerated by the use of midbrain dopamine neuronal models derived from induced pluripotent stem cells (iPSCs) that recapitulate several pathological features of GD and PD. In this review, we discuss evidence linking autophagic dysfunction to the pathophysiology of GD and GBA1-linked parkinsonism and focus more specifically on studies performed recently in iPSC-derived neurons.
Autophagy has been described as a double-edged sword as both deficient and excessive autophagy can be detrimental. What are your opinions on the potential of developing therapeutic agents that boost autophagy activity for PD?
Dr. Xilouri:
Enhancing α-synuclein degradation represents an attractive candidate option for therapy aiming to reduce not only the intracellular protein levels, but also the extracellular protein pool, which is increasingly being recognized as a potential key player in synucleinopathy propagation. Various approaches to target the ALP have been proven successful in both cellular and animal a-synucleinopathy models. In particular, boosting macroautophagy via mTOR-dependent (rapamycin) or mTOR-independent pharmacological and nutritional modulators (Metformin, Nilotinib, AICAR, Trehalose, Resveratrol, Pomegranate, C1) has been reported to enhance autophagosome formation, lysosome biogenesis, and lysosome function thus promoting α-synuclein clearance. In addition, molecular modulation of macroautophagy via Atg7, Beclin1 or TFEB overexpression is also reported to exert beneficial effects on α-synuclein-related toxicity. On the other hand, targeting CMA through molecular upregulation of LAMP2A expression or application of chemical modulators such as retinoic acid receptor alpha (RARα) antagonists has been also proven successful in alleviating α-synuclein-associated toxicity. In regard to CMA enhancement it seems that this approach has the potential to reduce a-synuclein levels and to mitigate the protein’s detrimental effects on lysosomal function at the same time, in effect “killing two birds with one stone”. Lastly, restoration of proper enzymatic activity of GCase has been shown to improve lysosomal function and lessen α-synuclein levels.
However, such approaches should be handled with caution, since uncontrolled manipulation of the global a-synuclein levels may lead to neurotoxicity, due to the prevailing role of the protein in synaptic neurotransmission. In regard to the degradation machineries in particular, experimental findings surmise that such strategies imply extensive knowledge about dosage and timing of application, a fact that may limit their current therapeutic applicability. Others have reported that under specific circumstances induction of macroautophagy can have detrimental effects, thus the therapeutic utility of chemical modulators of macroautophagy or even CMA should be examined with carefully given that they may be involved in diverse range of processes. Collectively, the targeted manipulation of selective ALP components, rather than broad autophagy stimulation, may be a safer strategy for the development of novel pharmacological therapies in PD.
Dr. Matsuda:
Autophagy is involved in numerous biological processes, and as pointed out, both deficient and excessive autophagy can be detrimental. Therefore, as a therapeutic strategy for Parkinson’s disease, enhancing autophagic activity might be difficult to achieve without detrimental effects. Instead, we might consider a therapeutic strategy that increases the activity of Parkin itself. In general, it is difficult to activate protein function. However, in the case of Parkin, its enzymatic activity is repressed via auto-inhibition, and thus can be activated by "inhibition of the auto-inhibitory mechanisms". Indeed, several "hyperactive mutations (V224A, W403A, and F146A)" that release the auto-inhibition and activate the E3 function of Parkin have been found and, intriguingly, the impaired mitophagic activity of pathogenic Parkin mutations was rescued by incorporation of a hyperactive mutation (Ref 5, Yi et al., 2019). By searching for small molecule compounds that promote effects on Parkin activity similar to the V224A, W403A, and F146A mutations, it thus might be possible to develop a therapeutic strategy for Parkinson's disease via Parkin activation.
Conclusion
Recent advances in the field have revealed that several PD-related genes and impaired autophagy form a bidirectional pathogenic loop, aggravating PD progression. The crucial question to be addressed is the therapeutic potentials of autophagy enhancing agents against PD. While increased autophagy activity might compensate the reduced degradation of pathogenic proteins, overactivation of autophagy might cause excessive lysosomal clearance and subsequently cell stress. Future drug development targeting autophagy should aim at determining the threshold that dictates whether autophagy trends the role of pro-survival or pro-death. Alternative option would be to target selective ALP molecules, such as parkin and GCase that require less but not completely no concern about their threshold effects. In general, studies have shown that parkin overexpression exerts a prominent protective role on neuronal survival. Also, there is an ongoing clinical trial of Ambroxol, a pharmacological chaperone that can enhance GCase activity for PD patients with or without GBA1 mutations. But as always, dosage, efficacy and safety of each potential therapeutic agent have to be carefully determined.
References:
1. Onishi M, Yamano K, Sato M, Matsuda N, Okamoto K. 2021. Molecular mechanisms and physiological functions of mitophagy. EMBO J 40: e104705.
2. Lazarou M, Sliter DA, Kane LA, Sarraf SA, Wang C, Burman JL, Sideris DP, Fogel AI, Youle RJ. 2015. The ubiquitin kinase PINK1 recruits autophagy receptors to induce mitophagy. Nature 524:309-314.
3. Yamano K, Kikuchi R, Kojima W, Hayashida R, Koyano F, Kawawaki J, Shoda T, Demizu Y, Naito M, Tanaka K, Matsuda N. 2020. Critical role of mitochondrial ubiquitination and the OPTN-ATG9A axis in mitophagy. J Cell Biol 219(9):e201912144.
4. Vargas JNS, Wang C, Bunker E, Hao L, Maric D, Schiavo G, Randow F, Youle RJ. 2019. Spatiotemporal Control of ULK1 Activation by NDP52 and TBK1 during Selective Autophagy. Mol Cell 74:347-362 e346.
5. Yi W, MacDougall EJ, Tang MY, Krahn AI, Gan-Or Z, Trempe JF, Fon EA. 2019 The landscape of Parkin variants reveals pathogenic mechanisms and therapeutic targets in Parkinson's disease. Hum Mol Genet. 2019 28(17):2811-2825.