Parkinson's--Divergent Causes, Convergent Mechanisms

See allHide authors and affiliations

Science  21 May 2004:
Vol. 304, Issue 5674, pp. 1120-1122
DOI: 10.1126/science.1098966

Parkinson's disease (PD) is a progressive neurodegenerative illness that affects about 1 million people in North America. PD is associated with a profound and selective loss of dopaminergic neurons in the nigrostriatal pathway of the brain, as well as a more widespread (but inadequately characterized) neuronal loss in other brain regions. Clinical manifestations include motor abnormalities (tremor, rigidity, slowness, balance problems), autonomic disturbances, psychiatric sequelae (usually depression), and cognitive impairment. Although environmental risk factors for PD have received considerable attention, the importance of the genetics underlying susceptibility to PD is increasingly recognized. Despite the overall rarity of the familial forms of PD (<10% of cases), the identification of single genes linked to the disease has yielded crucial insights into possible mechanisms of PD pathogenesis. Now, on page 1158 of this issue, Valente et al. (1) report that a newly identified familial form of PD is caused by a mutation in a putative mitochondrial protein kinase called PINK1 (PTEN-induced kinase 1).

Postmortem studies have consistently implicated oxidative damage (2) in PD pathogenesis, but the source of this damage has not been clear. Leading candidates for production of reactive oxygen species include dopamine metabolism and dysfunction of mitochondria. In 1982, after a group of intravenous drug users developed acute, permanent parkinsonism from injecting a contaminant (MPTP) of a synthetic opiate (3), it became clear that “environmental” chemicals might be the culprits in some cases. Epidemiological studies also suggested that environmental chemicals, such as pesticides, might be crucial factors in PD pathogenesis. The discovery that MPTP inhibits the first enzyme complex of the mitochondrial electron-transfer chain (complex I) prompted several groups to uncover complex I mitochondrial defects in the brains and platelets of people with PD (4). But could an apparent systemic defect in a mitochondrial enzyme cause the selective neurodegeneration that characterizes PD? The puzzle was resolved with the discovery that chronic systemic administration of the lipophilic complex I inhibitor, rotenone, could reproduce many of the features of PD, including the cytoplasmic proteinaceous inclusions called Lewy bodies that are characteristic of the disease (5). Together, these studies suggest that environmental chemicals, disrupted mitochondrial complex I activity, and oxidative stress may all participate in the killing of dopaminergic neurons in PD.

Genetic studies of PD have led in other directions. The first causative but rare mutation was found in the α-synuclein gene (6). Subsequently, α-synuclein, a phosphoprotein of uncertain function, was found to be a major component of Lewy bodies, even in the more common sporadic cases of PD where no mutation has been found. Moreover, overexpression of wild-type α-synuclein can also cause PD (7). In PD patients, the protein appears to be oxidatively and nitratively modified and cross-linked in insoluble aggregates. The formation of dopamine-quinone adducts may be important in this process. Another familial PD mutation affects ubiquitin carboxyl-terminal hydrolase-1 (UCHL1) (8)—a component of the cell's ubiquitin-proteasome system (UPS) that degrades damaged proteins; UCHL1 has both hydrolase and ubiquitin ligase activities. A third, and much more common causative mutation, affects another component of the UPS, a ubiquitin E3 ligase called parkin (9). Finally, pathogenic mutations have been reported in DJ-1, a protein that participates in the oxidative stress response (10). Thus, disease-causing mutations implicate aberrant protein handling and oxidative stress as key events in PD pathogenesis.

The diverse causes of PD may be related mechanistically (see the figure). As expected, normal mitochondrial activity may be affected by environmental chemicals (both natural and synthetic), and by mitochondrial DNA polymorphisms (11) and nuclear genes. Less expected are the findings that α-synuclein overexpression and inactivation of parkin can also cause mitochondrial dysfunction (12, 13). A common by-product of many types of mitochondrial impairment is increased production of reactive oxygen species (“free radicals”), and this may be the source of the oxidative damage found in PD brains. Inhibition of mitochondrial complex I leads to increased production and aggregation of α-synuclein (14). In dopaminergic neurons, aggregation may be promoted by dopamine metabolites and, perhaps, by the formation of highly reactive dopamine-quinones. The latter can form adducts with proteins, such as α-synuclein (15), cross-link them, and facilitate their aggregation. In addition to aberrant protein modification, dopamine oxidation can affect mitochondrial function (16) and increase oxidative stress. Inhibition of complex I can also impair the UPS, apparently by yielding reactive oxygen species that cause oxidative damage to proteins, perhaps including the components of proteasomes (17). The normal version of DJ-1 protects against oxidative stress, but the mutant protein does not. Thus, mitochondrial dysfunction and oxidative stress can bring into play the gene products and pathways implicated by disease-causing mutations (see the figure). Conversely, PD mutations may lead to mitochondrial impairment and oxidative stress.

Common themes in PD.

Five nuclear genes are known to carry mutations that cause PD. These genes encode α-synuclein, UCHL1, parkin, DJ-1, and PINK1. Mutations or altered expression of these proteins contributes to PD pathogenesis through common mechanisms that result in mitochondrial impairment, oxidative stress, and protein mishandling. Similar mechanisms also may be at work in toxin or pesticide-induced PD. α-Synuclein, a cytosolic and vesicle-associated protein, is oxidatively and nitratively modified and may form adducts with dopamine quinones, all of which accelerate its aggregation. UCHL1 and parkin are components of the ubiquitin-proteasome system (UPS) that degrades misfolded or damaged proteins. Inactivation of parkin results in mitochondrial impairment, and parkin may be S-nitrosylated and inactivated in PD. DJ-1 protects against oxidative stress and may become localized in mitochondria during times of oxidative stress. PINK1 is a nuclear-encoded, mitochondrial protein kinase, the substrates of which remain to be defined. Polymorphisms in mitochondrial genes dramatically alter the risk of developing PD. Environmental toxins, such as rotenone, impair mitochondrial function, cause oxidative stress, and lead to aggregation of proteins, including α-synuclein. Cytosolic dopamine (DA) may be oxidized to highly reactive dopamine quinones, and this may help to determine the selective loss of nigrostriatal dopaminergic neurons in PD. Red arrows indicate putative primary causes of PD; dashed arrows indicate effects that are probably secondary. Blue arrows indicate mechanisms of PD that are probably secondary to primary genetic or environmental causes. mtDNA, mitochondrial DNA.


So, where does PINK1 fit in? It is the first nuclear-encoded mitochondrial protein to be unambiguously implicated in PD pathogenesis. But this discovery prompts many new questions. Is the localization of PINK1 exclusively mitochondrial? And are the mitochondria where mutant PINK1 exerts its pathogenic effects? Is PINK1 really a protein kinase? What are its substrates? How does the loss of its putative serine/threonine kinase activity affect mitochondrial function? Valente and colleagues hypothesize that “PINK1 may phosphorylate mitochondrial proteins in response to cellular stress, protecting against mitochondrial dysfunction.” However, current data seem just as compatible with the idea that loss of the phosphorylation of normal mitochondrial proteins might cause mitochondrial dysfunction, perhaps through increased production of reactive oxygen species. In this regard, it is noteworthy that several subunits of complex I, including one that is essential for proper assembly of the complex, are known to be reversibly phosphorylated (18, 19). There is also a precedent for phosphorylation-dependent control of mitochondrial respiration (20) and of reactive oxygen species production by the electron-transfer chain (21).

There seem to be multiple, divergent causes of PD, yet the pathogenesis of this disease appears to be converging on common mechanisms—mitochondrial impairment, oxidative stress, and protein mishandling, all of which are tightly linked. At this point, exactly how PINK1 fits into the pathogenic cascade is unclear. Nevertheless, the discovery of PINK1 as a putative mitochondrial protein kinase that is important in PD pathogenesis opens new avenues of investigation into both basic mitochondrial biology and neurodegenerative disease.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
View Abstract

Navigate This Article