H. Plun-Favreau, Department of Molecular Neuroscience, Institute of Neurology, University College London, Queen Square, London WC1N 3BG, UK Fax: +44 0207 278 5616 Tel: +44 0207 837 3611; ext. 3936 E-mail: email@example.com
Rare, inherited mutations causing familial forms of Parkinson’s disease have provided insight into the molecular mechanisms that underlie the genetic and sporadic forms of this disease. Loss of protein function resulting from autosomal-recessive mutations in PTEN-induced putative kinase 1 (PINK1), Parkin and DJ-1 has been linked to mitochondrial dysfunction, accumulation of abnormal and misfolded proteins, impaired protein clearance and oxidative stress. Accumulating evidence suggests that wild-type PINK1, Parkin and DJ-1 may be key components of neuroprotective signalling cascades that run in parallel, interact via cross talk or converge in a common pathway.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
phosphatase and tensin homologue deleted on chromosome 10
tumour necrosis factor receptor-associated protein 1
ubiquitin C-terminal hydrolase L1
ubiquitin proteasomal system
Parkinson’s disease (PD) is a common neurodegenerative disorder with no known cure, estimated to affect 4 million people worldwide. The disease is characterized by the degeneration of dopaminergic neurons in the substantia nigra pars compacta and the presence of protein inclusions called Lewy bodies. The death of dopamine neurons in the substantia nigra pars compacta alters neurotransmitter balance in the striatum resulting in the progressive loss of movement control, the principal hallmark of PD, encompassing clinical features such as resting tremor, bradykinesia, postural instability and rigidity.
The most common form of PD is sporadic; there are, however, inherited forms of PD, accounting for 5–10% of cases. Little is known about how or why neurons die in PD, but similarities between both forms of the disease have led researchers to believe that a common set of molecular mechanisms may underlie PD aetiology.
To date, six genes have been implicated in the pathogenesis of PD, α-synuclein, Parkin, PTEN-induced putative kinase 1 (PINK1), DJ-1, leucine-rich repeat kinase 2 (LRRK2) and ATP13A2. Mutations in the genes encoding α-synuclein, LRRK2 and ATP13A2 cause autosomal-dominant forms of parkinsonism. Mutations in the genes encoding Parkin, DJ-1 and PINK1 all cause autosomal-recessive parkinsonism of early onset and are the focus of this minireview.
Autosomal-recessive Parkinson’s disease genes and proteins
Mutations in PARK2 were first reported in patients with autosomal-recessive juvenile-onset PD (AR-JP)  and are now known to be the predominant cause of early-onset parkinsonism. A large number of pathogenic mutations have been identified in Parkin, present in ∼ 50% of individuals with AR-JP, and 77% of sporadic cases with disease onset before the age of 20 . Clinically, PD patients with mutations in PARK2 suffer a slow progression of the disease commonly associated with early-onset dystonia and are l-Dopa responsive . Pathological studies on AR-JP patients with Parkin mutations have revealed a lack of Lewy body inclusions  except in some later onset cases [5,6].
Parkin localizes predominantly to the cytosol and cellular vesicles [7–9]. However, part of the cellular Parkin pool associates with the outer mitochondrial membrane . Parkin is an E3 ubiquitin ligase, an essential component of the ubiquitin-proteasomal system (UPS) . Parkin also has a proteasome-independent role and a number of putative substrates for Parkin have been described, including proteins implicated in PD such as synphilin-1 and a glycosylated form of α-synuclein . It is worth noting, however, that the only Parkin substrates known to accumulate in Parkin-null mice are the aminoacyl tRNA synthase cofactor p38 and far upstream-element binding protein 1 .
Mutations in PARK6 are the second most-common cause of autosomal-recessive PD after Parkin. Initially, three pedigrees were described with mutations in the PINK1 gene: a G309D point substitution in one family and a truncation mutation (W437X) in two additional families . Subsequently, several studies have described other pathogenic mutations in the PINK1 gene . Patients with PINK1 mutations respond well to l-Dopa treatment but do not have typical AR-JP phenotype, for example, dystonia at onset . The presence of a mitochondrial targeting sequence first suggested its precise subcellular location before Gandhi et al.  provided evidence that PINK1 is located in the mitochondrial membranes in human brain tissue. Although a cytoplasmic pool of PINK1 has been described [16,17]. PINK1 is of great interest to research into mitochondrial dysfunction in PD. PINK1 contains a putative catalytic serine–threonine kinase domain and shares homology with calmodulin-dependant protein kinase 1. In addition, preliminary evidence by Valente et al.  suggested that PINK1 protected mitochondria and cells against stress.
DJ-1 (PARK 7)
Mutations in PARK7 are associated with AR-JP and are a rare cause of familial PD [18–20]. One reported DJ-1 mutation is a large deletion unlikely to produce any protein. The other, a point mutation (L166P), has been studied extensively. Later, several studies led to the identification of a number of other pathogenic mutations causing familial PD . Clinically, age of onset is usually in the third decade with a slow disease progression and a good response to l-Dopa. DJ-1 is localized to both the nucleus and cytoplasm in different cell types [22,23], although a pool of wild-type DJ-1 has been shown to localize to the mitochondria . The L166P mutant protein has been shown to be associated with loss of nuclear localization and translocation to mitochondria  although this was not confirmed in other studies . Conversely, localization of wild-type DJ-1 at the mitochondria is suggested to be a requirement for neuroprotection . DJ-1 has been ascribed various functions, notably in resistance to oxidative stress , but also transcription, cell signalling, apoptosis [27,28] and aggregation of α-synuclein . The protein may also act as a chaperone. Finally, studies suggested that DJ-1 could possess cysteine protease activity. However, the protease activity of DJ-1 is still a matter of debate [30,31]. But perhaps the most important function with regard to PD is its putative role in oxidative stress. DJ-1 is thought to protect neurons from oxidative stress [19,32,33] although exactly how it exerts its protective effects remains to be determined.
Molecular pathways of neurodegeneration in PD
The study of autosomal-recessive PD genes has provided valuable insight into the molecular mechanisms of dopaminergic degeneration. The absence of normal proteins resulting from mutations in these genes causes a range of different but overlapping pathological effects in neurons, namely mitochondrial impairment, proteasomal dysfunction, oxidative stress and protein phosphorylation . These processes are being intensively examined, partly in the hope that they will shed light on the more common sporadic form of PD.
Mitochondrial dysfunction has been implicated in the pathogenesis of a wide range of neurodegenerative diseases, particularly PD . Defects in mitochondrial complex I have been closely linked to PD. Environmental toxins causing parkinsonism such as 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine and rotenone inhibit complex I of the mitochondrial electron transport chain, leading to oxidative stress, impaired energy metabolism, proteasomal dysfunction and, eventually, death of dopaminergic neurons [35,36]. Their administration in vivo mimics the pathological effects of PD [37,38]. Interestingly, susceptibility to rotenone toxicity is increased in neurons from Parkin-null mice . PINK1 suppression using small interfering RNA decreased cell viability and significantly increased 1-methy-4-phenylpyridinium and rotenone-induced cytotoxicity . Furthermore, it has been reported very recently that germline deletion of the PINK1 gene in mice significantly impairs mitochondrial functions and provides critical protection against oxidative stress [41,42]. Neurons with reduced levels of endogenous DJ-1 were also sensitized to toxicity elicited by rotenone  and Drosophila DJ-1 mutants were selectively sensitive to environmental toxins associated with PD .
Parkin and PINK1 have been shown to be located, at least in part, to the mitochondria. In Drosophila models of PINK1, several studies [45–47] strongly suggested that PINK1 acts upstream of Parkin in a common pathway that influences mitochondrial integrity in a subset of tissues (including flight muscle and dopaminergic neurons). Recent studies suggest that the PINK1/Parkin pathway regulates mitochondrial morphology in Drosophila and mammalian models [48–50].
DJ-1 does not seem to operate in the same pathway as Parkin and PINK1. Muscle and dopaminergic phenotypes associated with Drosophila PINK1 inactivation can be suppressed by the overexpression of Parkin, but not DJ-1 . Although there is less evidence for a direct role of DJ-1 in mitochondrial function, the fact that Drosophila lacking DJ-1 exhibit increased sensitivity to environmental mitochondrial toxins [44,51] does point to a role for DJ-1 in mitochondrial function.
Drosophila studies suggest that PINK1 is required for mitochondrial function and that the PINK1/Parkin pathway regulates mitochondrial morphology [45–47]. In this connection, a coherent hypothesis is that these two proteins might act directly at the mitochondrion, through their respective phosphorylation or ubiquitination activities. Alternatively, PINK1 might need to be released into the cytosol in order to fulfil its function under conditions of stress. This is the case for mitochondrial proteins such as Smac/Diablo and Omi/HtrA2 . The mature form of these proteins can be generated by proteolysis. During apoptosis, mature Omi/HtrA2 and Smac/Diablo are released from the mitochondria into the cytosol where they exhibit a pro-apoptotic function. PINK1 is cleaved  and this cleavage seems to play a crucial role in its protective function against various stressors [53,54]. However, the protease responsible for PINK1 cleavage as well as the PINK1 cleavage site remains to be identified advances which would shed much light on PINK1s role in the cell. It is possible that PINK1 could exhibit an extra-mitochondrial role, interacting with Parkin, DJ-1 and other signalling molecules in the cytosol, which in turn regulate mitochondrial function.
Given that mitochondria have crucial roles in multiple cellular processes, including metabolism, regulation of cell cycle and apoptosis, Ca2+ homeostasis, ATP production and cellular signalling, it is likely that Parkin, PINK1, DJ-1 and interactors such as Omi/HtrA2  play a part in these processes.
Proteasomal dysfunction and proteolytic stress
The proteasome is a large multi-catalytic proteinase complex found in the nucleus and cytoplasm of eukaryotic cells [56,57]. UPS dysfunction and proteolytic stress are likely to contribute to dopaminergic neurodegeneration . Moreover, mutations in two components of the UPS; Parkin and ubiquitin C-terminal hydrolase-L1 (UCH-L1)  in familial PD strongly supports the hypothesis that proteasomal dysfunction may contribute to PD aetiology .
Notably knockdown of DJ-1  and Parkin [61,62] enhances susceptibility to proteasome inhibition in cell models. In addition, DJ-1-deficient mice treated with the mitochondrial complex I inhibitor paraquat display decreased proteasome activities and increased levels of ubiquitinated protein . Finally, the UPS has also been shown to be important for the regulation of PINK1 stability  and the degradation of DJ-1 [30,64], PINK1  and Parkin [66,67] mutant proteins.
Chaperones may be key players in PD pathogenesis. PINK1 has been shown to interact with the Hsp90 molecular chaperone and it was proposed that the inhibition of this interaction might contribute to the pathogenesis of PD . Furthermore, PINK1 has been suggested to protect against oxidative stress by phosphorylating the mitochondrial chaperone tumour necrosis factor receptor-associated protein 1 (TRAP1)  as well as playing an important role in the regulation of HtrA serine peptidase 2 (HtrA2) protease activity . Moreover, in light of evidence that PINK1 acts upstream of Parkin in the same biological pathway it is often speculated that PINK1 might phosphorylate Parkin.
Structural studies indicate that HtrA2 has similarities to its bacterial homologues DegS and DegP  which function as both molecular chaperones and proteases. DJ-1 also has been shown to have similarities to its stress adaptive homologue Hsp31  suggesting that both HtrA2 and DJ-1 may degrade unfolded proteins, performing crucial functions with regard to protein quality control in different cell compartments. Finally, several chaperones have been shown to be Parkin substrates [70,71] and Parkin folding seems to be dependent on chaperones .
It is therefore tempting to speculate that proteins such as Parkin, PINK1, DJ-1, Hsp90, TRAP1 or HtrA2 might participate in the detoxification of proteins either directly through their putative chaperone function or indirectly through their interactions with chaperone molecules.
Oxidative damage to lipids, proteins and DNA occurs in PD . This stress can directly impair protein ubiquitination and degradation systems and the toxic products of oxidative damage induce cell-death mechanisms.
Many lines of evidence suggest that DJ-1 functions as an antioxidant. Oxidative stress causes an acidic shift in the isoelectric point of DJ-1 [26,32,74] suggesting self-oxidation. Embryonic stem cells deficient in DJ-1 display increased sensitivity to oxidative stress and proteasome inhibition . Following exposure to oxidative stress, DJ-1 associates with Parkin, potentially linking these proteins into a common molecular pathway leading to nigral degeneration and PD . Parkin knockout mice have revealed an essential role for Parkin in oxidative stress  and Drosophila Parkin mutants show increased sensitivity to oxidative stress . Implication of PINK1 in oxidative stress processes has also been strongly suggested: inactivation of Drosophila PINK1 using RNAi suggested that PINK1 maintains neuronal survival by protecting neurons against oxidative stress . In mammalian cell culture, PINK1 protects against oxidative stress-induced cell death by suppressing cytochrome c release from mitochondria, with the protective action of PINK1 depending on its ability to phosphorylate the mitochondrial chaperone TRAP1 .
Protein phosphorylation and signalling pathways
PINK1 has a strongly predicted, conserved serine/threonine kinase domain  and has been shown to exhibit autophosphorylation activity [15,80,81] in vitro. In vivo, PINK1 has been shown to phosphorylate the mitochondrial chaperone TRAP1, protecting against oxidative stress-induced apoptosis  and to be important for the phosphorylation of HtrA2 upon activation of the p38 pathway, preventing against mitochondrial stress .
PINK1 was originally identified by an analysis of expression profiles from cancer cells after the introduction of exogenous phosphatase and tensin homologue deleted on chromosome 10 (PTEN), a tumour suppressor that is involved in the regulation of the phosphatidylinositol 3-kinase signalling pathway . Interestingly, Parkin, DJ-1 and HtrA2, although devoid of kinase activity, have also been shown to be regulated and/or regulators of the phosphatidylinositol 3-kinase pathway. A genetic screen of Drosophila gain-of-function mutants has shown that DJ-1 was a negative regulator of PTEN , and an impairment of phosphatidylinositol 3-kinase/Akt signalling has been observed in a DJ-1 and Parkin Drosophila model of PD . The phosphatidylinositol 3-kinase/Akt pathway has also been shown to be reduced in Parkin knockout mouse brain , suggesting a common molecular event in the pathogenesis of PD. In addition, HtrA2 might be directly regulated by Akt . Nevertheless, whether the phosphatidylinositol 3-kinase signalling pathway is important for the regulation of Parkin, PINK1, DJ-1 and HtrA2 activity remains to be determined.
Parkin can be phosphorylated by a number of kinases including casein kinase 1, protein kinase A, protein kinase C  and cyclin-dependant kinase 5 . Phosphorylation of Parkin by CDK5 may regulate its ubiquitin-ligase activity and therefore contribute to the accumulation of toxic Parkin substrates and decreased ability of dopaminergic cells to cope with toxic insults in PD . To date, no direct phosphorylation of DJ-1 or PINK1 has been reported.
A common pathway to parkinsonism?
There has been a great deal of interest from the PD scientific community in linking the familial-associated genes in a common pathogenic pathway of neurodegeneration. To date, however, a single pathway unifying these proteins has not been fully mapped out.
PINK1 and Parkin seem to function, at least in part, in the same pathway, with PINK1 acting upstream of Parkin. Moreover, a recent study has proposed a role for Cdc37/Hsp90 chaperones and Parkin on PINK1 subcellular distribution, providing further evidence for a Parkin/PINK1 common pathogenic pathway in recessive PD . The role of the PINK1–Parkin pathway in regulating mitochondrial function underscores the importance of mitochondrial impairment as a key molecular mechanism underlying PD. Overexpression experiments in SH-SY5Y human neuroblastoma cells have shown that DJ-1 specifically interacts with Parkin under stress conditions. Specifically, this association is mediated by pathogenic DJ-1 mutations and oxidative stress . These data suggest a link DJ-1 and Parkin in a common pathway in mammals. A described case of autosomal-recessive PD with digenic inheritance, suggested that DJ-1 and PINK1 might physically interact and collaborate to protect cells against stress . However, the muscle and dopaminergic phenotypes associated with Drosophila PINK1 inactivation, can be rescued by overexpression of Parkin but not DJ-1, suggest that PINK1 and DJ-1 do not function in the same pathway, at least in flies . Finally, PINK1 has been shown to interact with HtrA2 and both seem to be components of the same mitochondrial stress-sensing pathway . Several mutations implicating HtrA2 in PD have been identified . However, the evidence that mutations in HtrA2 modulate PD risk was later questioned and continues to be an area of debate. Sanchez et al. effectively demonstrated that HtrA2 is not a PD risk-gene in an extended series of North American PD cases . However, Bogaerts et al. examined the contribution of genetic variability in HtrA2 to PD risk in an extended series of Belgian PD patients and control individuals. This mutational analysis identified a new mutation (Arg404) strengthening a role for the HtrA2 mitochondrial protein in PD susceptibility .
Each molecular event occurring between genetic mutation and nigral cell degeneration is intimately linked to other components of the degenerative process. The challenge for scientists is therefore to determine whether there is a single pathway unifying these proteins or whether the situation is more complicated, for example, involving cross-talk from other pathways (Fig. 1). If the latter is the case, are there parallel pathways leading to the same or similar pathological effects or are there multiple pathways converging at a common point? Answering these questions requires a good PD model. Drosophila and more recently zebrafish  models have recapitulated many of the phenotypic and pathologic features of PD, however, these models are far-removed from human DA neurons. Both primary neurons and human neuronal cell lines better represent the cell types involved in PD, but have major limitations . Advances in the field of stem cell research might open up a new route to develop a cell model that more closely mirrors the disease situation in humans. The use of induced pluripotent stem cells as a research tool has become very promising following a number of publications showing re-programming of human fibroblasts carrying mutations to induced pluripotent stem cells [94,95] and recently their differentiation into specific neuronal subtypes .
Understanding the exact function of Parkin, PINK1, DJ-1 and HtrA2 proteins in age-matched healthy volunteer (and ideally relatives) neurons compared with the neurons of patients with AR-JP may allow us to dissect biochemical pathways that lead to these diseases and will be a major step forward in our understanding of the pathogenesis of PD and ultimately to the development of novel therapeutic approaches.
The authors wish to thank Professor Nicholas Wood for his comments.