α-synuclein
α-synuclein is encoded by
SNCA and its fibrillar form is the major component of Lowy bodies [
38]. Existence of hydrophobic non-amyloid β component domain endows α-synuclein with the propensity to aggregate [
39]. Aggregated α-synuclein can trigger neuronal death in PD [
40]. Thus, intensive researches focus on how to prevent or eliminate the aggregation of α-synuclein. Study show that ubiquitin proteasome system (UPS) is responsible for degrading monoubiquitinated α-synuclein, while macroautophagy is for removing deubiquitinated α-synuclein [
41,
42]. Further study demonstrates, in normal condition, α-synuclein is mainly degraded by UPS [
43]. Macroautophagy pathway can be activated in response to increased level of wild type or A53T mutant α-synuclein [
44,
45]. However, some studies show that increased α-synuclein can impair autophagosome synthesis by inhibiting Rab1a [
46]. Also, pre-formed α-synuclein aggregates compromise autophagosome clearance and are resistant to degradation by autophagy [
47]. The difference may be derived from the variability of models or different levels of α-synuclein. Further studies are warranted to address such discrepancy.
Abnormal α-synuclein level can also disrupt mitophagy. In PD postmortem brain tissues, α-synuclein accumulation increases oxidative stress and disturbs mitochondrial function [
48]. Moreover, both in vivo and in vitro, expression of α-isoforms of α-synuclein in neurons causes the fragmentation of mitochondria, which will eventually leads to the decline in respiration and neuronal death [
49]. Over-expression of α-synuclein can occur in mitochondria and disrupt mitochondrial membrane potential by opening the mitochondrial permeability transition pore (mPTP), thereby initiating mitophagy [
45]. α-synuclein can activate mPTP via interacting with either adenylate translocator (ANT) or voltage dependent anion channel (VDAC) [
50,
51]. Further study found that A53T mutant could impair mitochondrial function by residing in mitochondria membrane as monomers and oligomers [
52].
Increasing evidence confirms α-synuclein as a target of chaperone mediated autophagy (CMA). Hsc70 can recognize soluble α-synuclein and the affinity between Hsc70 and α-synuclein fibrils is 5-fold tighter compared with soluble α-synuclein [
53]. One recent study showed that Hsc70 and Ssa1p work like a tweezer to bind two domains within α-synuclein [
54]. Also, in vitro, one study shows the uptake of extracellular α-synuclein by neurons and their retrograde axonal transportation to neuronal soma. However, Hsc70 chaperones α-synuclein in the extracellular space and alleviates α-synuclein oligomer formation [
55]. Activating CMA activity via up-regulating LAMP2A decreases α-synuclein turnover and protects against α-synuclein over-expression induced neurotoxicity [
9]. Our study showed MEF2D, a transcription activator identified as neuronal survival factor, is the substrate of CMA. Wild type or mutant α-synuclein accumulation compromises normal turnover of MEF2D by CMA and leads to decrease in the MEF2D DNA binding ability and neuronal stress, which underlies the neuronal loss of PD [
56]. Thus, enhancing CMA pathway could be a promising therapeutic strategy for PD treatment.
PINK1/Parkin
PINK1 is a serine/threonine kinase protein and mutations in PINK1 cause a rare form of autosomal recessive PD [
57]. Parkin containing ubiquitin E3 ligase can ubiquitinate multiple substrates for degradation. Mutations of Parkin lead to accumulation of its substrates and are related with early onset juvenile autosomal recessive PD [
58].
Increasing evidence suggests the PINK1/Parkin pathway is essential for mitochondrial quality control. One recent study shows that Parkin ubiquitinates dynamin-related protein 1 (Drp1) to promote its degradation. Disruption of this interaction by Parkin mutation leads to the accumulation of Drp1 and mitochondrial fragmentation [
59]. In
Drosophila, both Parkin and PINK null mutants show a significant overall slowing of motichondrial protein turnover and mitophagy. Failure to remove the damaged mitochondrial proteins plays an important role in PD pathogenesis [
60]. PINK1 mutation affects mitochondrial complex I activity and the maintainance of the electron transport chain, which disturbs the mitochondrial membrane potential [
61]. Upregulation of Parkin protects cells against multiple stresses, including endoplasmic reticulum stress, mitochondrial stress, proteotoxicity and excitotoxicity [
62,
63]. By contrast, loss of Parkin results in mitochondrial fragmentation, decreased cellular Ca
2+ handling capability and increased cellular vulnerability to stress [
64]. All the above findings demonstrate that deficiency in PINK1/Parkin pathway leads to mitochondrial dysregulation.
Recent researches on PINK1/Parkin pathway have revealed molecular details for mitochondria protection. Once the mitochondria are impaired and the membrane potential gets depolarized, PINK1 will accumulate on the outer membrane of mitochondria (OMM). A study showed dimeric PINK1 on OMM can recruit Parkin and thereby phosphorylates Parkin at Ser65 [
65]. Research using mass spectrometry identified VDACs as the docking site for Parkin recruitment to the OMM [
66]. After Parkin translocation to mitochondria, many OMM proteins are ubiquitinated by Parkin and in turn recruited other proteins to initiate mitophagy. Then, these Parkin labeled mitochondria are brought to lysosomes for degradation. This PINK1/Parkin signaling pathway can be positively modulated by AF-6, which is lacked in caudate/putamen and SN of sporadic PD patients [
67]. Moreover, up-regulation of translocation of the OMM (TOMM) can rescue mitophagy impaired by Parkin mutations, hinting that TOMM acts as an important regulator in PINK1/Parkin mediated mitophagy [
68]. Despite extensive research, how autophagy related proteins are recruited during mitophagy process is still unclear and to what extent mitophagy dysregulation contributes to PD pathogenesis remains to be investigated.
LRRK2
Leucine-rich repeat kinase 2 (LRRK2) is one of the generic contributors to PD. As estimated, variable LRRK2 mutants contribute to over 10 % of familial and about 3 % of sporadic PD cases [
69]. So far, over 50 LRRK2 mutations have been identified in PD patients. Among these, the G2019S mutation is the most common cause of autosomal dominant familial PD cases [
70]. Also, G2019S mutation can be found in about 2 % sporadic PD. Thus, exploring LRRK2 pathogenicity is essential to understand the molecular mechanisms of PD.
Many reports show a relationship between LRRK2 and macroautophagy. In kidney of mouse, loss of LRRK2 leads to age-dependent bi-phasic alteration of macroautophagy activity [
71]. In human neuroglioma cells, inhibition of LRRK2 kinase activity can stimulate macroautophagy in the absence of any alteration in mTOR pathway, suggesting that LRRK2 regulates autophagic activity independent of mTOR signalings [
72]. Consistently, LRRK2 activates a persistent increase in autophagosome formation through a calcium associated pathway. Simultaneously, LRKR2 upregulation increases p62 and decreases the number of acidic lysosomes [
73]. Moreover, G2019S mutation increases autophagic vacuoles and shortens neurite length [
74]. The effects of G2019S on neurite length can be abolished by down-regulation of LC3 or Atg7 and enhanced by autophagy inducer rapamycin, hinting that autophagy plays an important role in regulation of neurite length [
75]. In fibroblasts, G2019S LRRK2 mutant exacerbates MPTP-induced cell death dependent of autophagic activity [
76]. Furthermore, a study showed that LRRK2 increases the number of autophagosomes by activating CaMKK/AMPK pathway [
73]. In addition, G2019S mutation augments autophagic flux by MEK/ERK signaling [
77].
LRRK2 and the PD-associated mutations can be degraded by CMA. In normal condition, both UPS and CMA are responsible for the degradation of wild-type LRRK2. However, G2019S LRRK2 disrupts degradation by these pathways [
78,
79]. In addition, both wild-type LRRK2 up-regulation and its mutations inhibit formation of the CMA translocation complex, thereby suppressing CMA activity [
80,
81]. Subsequently, the normal turnover of CMA substrates is disrupted. Although LRRK2 mutants damage CMA process, the detailed relationship with neuronal loss, especially in animal models, still needs to be further explored.
In physiological conditions, about 10 % of LRRK2 can present in the OMM, giving rise to the hypothesis that LRRK2 mutations might have influence over mitochondrial function. Study showed LRRK2 G2019S could lead to mitochondrial fragments by enhancing mitochondrial fission via phosphorylating decaprenyl diphosphate synthase subunit 2 Dlp1(Dlp1) [
82]. Also, LRRK2 mutants activate dendritic mitochondrial clearance by autophagy in neurons [
83].
DJ-1
DJ-1 belongs to the peptidase C56 family and acts as a redox-responsive cytoprotective protein. PD-related DJ-1 mutants are rare and contribute to 1–2 % of autosomal recessive PD [
84].
DJ-1 can serve as a regulator of autophagy. The effect of DJ-1 up-regulation on macroautophagy depends on cell type. In dopaminergic neurons, DJ-1 overexpression induces ERK-dependent mitophagy and protects against neurotoxin induced apoptosis [
11]. Also, in mouse embryonic fibroblasts, loss of DJ-1 suppresses basal autophagy and disrupts mitochondrial dynamics [
85]. However, in some cancer cells, DJ-1 deficiency activates autophagy via JNK signaling [
86]. Further studies are needed to address the relationship between DJ-1/autophagy and DA neuronal loss in the context of PD.
As an anti-oxidative protein, DJ-1 also plays an important role in regulating mitochondrial function. In human neuroblastoma M17 cells, DJ-1 wild-type overexpression induces elongated mitochondria while DJ-1 mutants overexpression causes mitochondrial fragmentation. Interestingly, DLP1 knockdown in these mutant DJ-1 cells rescues mitochondrial morphology and function [
87]. Also, in DJ-1 knockout cells, autophagy degradation was impaired and defective mitochondria accumulated [
88]. Loss of DJ-1 leads to mitochondrial phenotypes including reduced membrane potential, increased fragmentation and accumulation of autophagic markers. Supplementing DJ-1-deficient cells with glutathione reverses both mitochondrial and autophagic changes suggesting that DJ-1 may act to maintain mitochondrial function during oxidative stress [
89].
In our recent study [
90], we found that DJ-1, harboring CMA specific motif, is a direct substrate of CMA and mutation inside the motif can disturb the degradation of DJ-1 through CMA. In addition, we showed UPS is not the primary mechanism responsible for DJ-1 degradation. More interestingly, CMA preferentially clears the oxidatively injured DJ-1 and the extensive accumulation of the oxidized DJ-1 monomer following a reduction of lysosomal and CMA activity affects the formation or alters the balance of DJ-1 dimer. Furthermore, CMA-DJ-1 pathway plays a critical role in maintaining mitochondrial morphology and function under stress and protects against PD related neurotoxins induced cytoxicity. Our findings suggest that CMA/DJ-1 pathway is vital for mitochondrial homeostasis and dysregulation of this pathway may explain the neuronal loss during PD pathogenesis.