Recent advances in the genetics and pathogenesis of Parkinson disease

Genes responsible for dopa-responsive parkinsonism.

To date, specific mutations have been identified in three separate genes,3-5⇓⇓ and four additional loci have been linked to inherited PD (table).8-11⇓⇓⇓ Based on the biochemical and molecular properties of these gene products, a unifying pathogenetic mechanism has emerged despite the genetic and clinical heterogeneity. These investigations point to the critical role of protein folding and degradation through the ubiquitin proteasome pathway as a central common mechanism leading to cell death (figure 1). Accumulation of oxidatively damaged proteins and impairment of proteasome activity in the nigra of patients with sporadic PD12 substantiate the conclusions drawn from the genetic data that altered proteolysis is a key pathogenetic process in this disorder. This notion also accounts for the age-dependent incidence of PD.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Proposed pathogenetic cascades leading to neuronal death in PD. The accumulation of insoluble and/or toxic intermediates appears to be a central factor in these processes regardless of the specific gene mutation or potential environmental factor.

α-Synuclein.

α-Synuclein is a predominantly neuronal protein of 140 amino acids. Two point mutations in its cognate gene have been identified in few families with dominantly inherited PD. The first results in an alanine to threonine substitution in position 53 (A53T) found in 13 families with Italian/Greek descent consistent with a founder effect,3,13,14⇓⇓ and the second results in an alanine to proline substitution at position 30 (A30P) in a single German family.15 Phenotypically, these patients have typical dopa-responsive parkinsonism except for a relatively early age at onset (mean 46 years), and they have the pathologic hallmark features of PD, namely Lewy bodies. Although α-synuclein mutations represent a rare cause of PD, this protein is the major fibrillar component of Lewy bodies and Lewy neurites in dominantly inherited, as well as sporadic, PD.16

Although our knowledge about the normal function of α-synuclein is limited to its modulation of synaptic plasticity based on song learning studies in the zebra finch17 and regulation of vesicular dopamine release from knock-out mice,18 the structural properties of this protein have shed considerable light on its pathogenic involvement in PD. In solution, α-synuclein is an unfolded protein with no fixed conformation.19 In the presence of lipid-containing vesicles, it assumes an alpha helical structure, and in high concentrations it turns into a beta sheet typical of amyloid fibrils. This process of fibrillization is dependent on time and concentration, accelerated by pathogenic mutations,20 and nucleation dependent.21 Additional in vitro investigations have revealed the tendency of α-synuclein to form oligomers of different shapes that appear to precede the formation of fibrils. The propensity of both Parkinson-causing mutant isoforms of α-synuclein to oligomerize more than the wild-type protein suggests that perhaps these oligomers constitute toxic intermediates. Fibril formation, on the other hand, is accelerated only by the A53T mutant, and not the A30P mutant.22 α-Synuclein fibril accumulation in Lewy bodies might even represent a protective process whereby excess or unfolded α-synuclein is cleared from the neuronal cytoplasm. This notion is analogous to the acceleration of neurodegeneration in models of triplet repeat expansion disorders, such as Huntington and spinocerebellar ataxia type 1, by measures that suppress the formation of inclusions.23,24⇓ Lewy bodies have often been reported to be seen in relatively healthy-appearing nigral neurons, supporting the view that they could perhaps be formed by surviving neurons, whereas cells that cannot effectively fibrillize α-synuclein into Lewy bodies succumb to the toxic effects of other intermediate forms of this protein.

The central role of α-synuclein in the pathogenesis of PD is clearly based on the fact that: 1) mutations in its cognate gene result in dominantly inherited PD; 2) it accumulates abundantly in Lewy bodies; and 3) overexpression of wild-type or PD-causing mutants in transgenic mice or flies recapitulates many of the behavioral, pathologic, and biochemical features of human PD.25,26⇓ Therefore, consistent with the fact that most patients with PD do not carry mutations in the α-synuclein gene, these mutations are not essential for the pathogenicity of this protein. In vitro biochemical analyses suggest that the random coil structure of α-synuclein and its tendency to aggregate into misfolded structures under certain circumstances might be sufficient to confer toxic properties to this protein.

Unfolded or misfolded proteins represent a major stress to cells in general, and inefficient clearance of such proteins is a hallmark of aged cells. Such proteins with abnormal conformation, as well as otherwise damaged or oxidized proteins that are not degraded, tend to aggregate as inclusions. In addition to attenuating translation and recruiting molecular chaperones to assist in the proper refolding of proteins, cells use an efficient degradation process to eliminate proteins with unwanted conformations primarily through the ubiquitin/proteasome pathway.27,28⇓ The latter involves the tagging of target proteins with the small peptide ubiquitin through the step-wise action of three sets of enzymes, namely ubiquitin-activating enzyme E1, ubiquitin-conjugating enzyme E2, and ubiquitin-ligating enzyme E3. The latter acts as a substrate recognition molecule that helps the transfer of ubiquitin by E2. In keeping with the need and ability of cells to clear unfolded proteins, α-synuclein is ubiquitinated and degraded through the proteasome, which is a multi-subunit protease. Interestingly, a PD-causing mutation in α-synuclein results in a slightly slower degradation rate, with an approximately 50% prolongation of its half-life.29 Because concentration and time are important factors in the tendency of α-synuclein to fibrillize, accumulation of this protein in neurons over decades could promote its aggregation. Thus, mutations in α-synuclein stress neurons simply through changing its kinetics in the cell and predisposing it to aggregate, thereby conferring a toxic gain of function consistent with dominant inheritance. Conversely, α-synuclein interacts with certain proteasome subunits,30 and mutant α-synuclein causes impaired proteasomal function in dopaminergic cells.31 In addition, oxidative stress and iron, which are present within the microenvironment of nigral dopaminergic cells, promote the aggregation of α-synuclein.32 Furthermore, PD-causing mutations in α-synuclein render cells more susceptible to oxidative damage,33 partly by increasing intracellular levels of reactive oxygen species (Junn et al., submitted).34 It is conceivable that the latter mode of apoptotic cell death is mediated through misfolding of α-synuclein. In this respect, it is noteworthy that protein aggregation itself impairs the ubiquitin-proteasome system leading to cell death.35 Therefore, α-synuclein aggregation caused by a variety of factors such as oxidative stress, which could be induced by potential environmental insults, can lead to dysregulation of cellular functions.

Like many proteins in a cell, α-synuclein interacts with other molecules that influence its structure and function. These include Aβ amyloid which can seed the aggregation of α-synuclein,36 and synphilin-1 which co-aggregates with α-synuclein to form cytoplasmic inclusions.37 α-Synuclein also interacts with the membrane dopamine transporter resulting in clustering of the latter onto the plasma membrane, and more efficient dopamine uptake leading to increased susceptibility to oxidative stress.38 Recently, α-synuclein has been found to interact with fatty acids, suggesting its role in transporting these molecules between the aqueous and membraneous compartments of the neuronal cytoplasm.39

The accumulation of α-synuclein or its fragments in pathologic inclusions in the brain is not limited to PD. Immunohistochemical studies have demonstrated this protein in various inclusions of several neurodegenerative disorders, including those in diffuse Lewy body disease, the senile plaques of AD,40 and Down syndrome,41 as well as the glial cytoplasmic inclusions of multisystem atrophy.42,43⇓ Presumably, the relatively high concentration of α-synuclein in the brain and its tendency to interact with many other protein partners render it an accomplice in many neurodegenerative processes that involve protein aggregation, regardless of their primary origin. Additionally, such inclusions typically contain ubiquitin and/or component(s) of the proteasome complex, suggesting the general role of this protein clearing mechanism in the accumulation of pathogenic proteins in a host of such disorders. Further, the common structural features between α-synuclein and β-amyloid combined with their tendency to interact with and seed the aggregation of each other36 highlight the pathogenetic similarities between PD and AD.

Parkin.

Parkin is a recently identified protein of 465 amino acids. It was originally discovered through positional cloning of the genetic defect in Japanese families with autosomal recessive juvenile parkinsonism.4 Since then, mutations in the parkin gene have been identified in families with recessively inherited PD throughout the world, and are thought to represent the underlying defect in as many as 50% of such cases in certain populations.44 Various mutations are associated with PD, including deletions and point mutations. Clinically, the disease usually begins when the patient is in his or her 20s, is prominently associated with dystonia and diurnal fluctuations, and progresses slowly but has early and severe levodopa-induced dyskinesias, but no dementia. Pathologically, there is severe neuronal loss in the substantia nigra pars compacta and locus ceruleus, but no Lewy bodies or other inclusions have been identified to date.

The domain structure of Parkin includes a ubiquitin homologous domain in its N-terminus, and two RING finger domains in its C-terminus.45 Similar to many other proteins with a RING finger domain,46 Parkin has an E3 ubiquitin ligase function,45,47⇓ linking Parkin-associated PD to the ubiquitin-proteasome system as well. Because mutations in parkin are associated with recessively inherited PD, they result in loss or diminished E3 ligase function in the nigra and striatum of individuals with these mutations,45 leading to the abnormal accumulation of its substrate proteins. The search for these substrates has provided important information elucidating the biochemical consequences of parkin mutations. To date, three such substrates have been identified. Perhaps the most intuitively expected substrate is α-synuclein itself. However, the isoform of α-synuclein that interacts with Parkin and is ubiquitinated by it is the 22-kD glycosylated version, rather than the classically recognized 16-kD form.48 Identification of factors that dictate α-synuclein glycosylation in dopaminergic neurons, as well as the relative abundance of the two forms of this protein, would likely be crucial steps in our search for new therapeutic targets for Parkin-associated PD in particular, and perhaps for other PD types in general. The second substrate that interacts with and is ubiquitinated by Parkin is a putative G protein–coupled transmembrane protein named Parkin-associated endothelin receptor-like receptor (Pael-R).49 When Pael-R is overexpressed, it becomes unfolded, insoluble, and ubiquitinated. Unfolded Pael-R leads to endoplasmic reticulum stress and, consequently, to cell death. Wild-type Parkin promotes the degradation of Pael-R and, therefore, prevents insoluble Pael-R mediated cell death. Consistent with these observations, Pael-R accumulates in an insoluble form in the brains of individuals with parkin mutations. The third Parkin substrate reported to date is CDCrel-1,47 which belongs to a family of proteins originally designated as septins to reflect their role in cytokinesis and the separation of mother and daughter cells. CDCrel-1, a synaptic protein predominantly expressed in the CNS, interacts with syntaxin which is a SNARE (soluble N-ethylmaleimide-sensitive fusion attachment protein receptor) protein essential for membrane fusion. This interaction inhibits the process of synaptic vesicle docking at the plasma membrane and exocytosis.50 Parkin-induced ubiquitination and degradation of CDCrel-1 presumably leads to more transmitter release, whereas pathogenic mutations in parkin could result in inhibition of transmitter release. Whether these observations based on transfected cell systems can be substantiated in the brains of patients with PD remains to be tested. Additional studies are, therefore, needed to confirm and elucidate the role of these synaptic proteins in the pathogenesis of this disease.

While these Parkin substrates are interesting, it is not clear at present how they relate to the larger PD population with no parkin mutations. Nevertheless, intact functional Parkin appears important for the processing of these proteins and is needed for the formation of Lewy bodies. Lack of these inclusions in the brains of patients with parkin mutations who develop the disease at a much younger age than individuals with other forms of PD could, once again, support a protective role of Lewy bodies. It is conceivable that perhaps as yet unidentified factors could impair Parkin function in other forms of parkinsonism.

UCH-L1.

UCH-L1 is an enzyme that belongs to the ubiquitin C terminal hydrolase family that hydrolyzes small C terminal adducts of ubiquitin to generate ubiquitin monomers, which can then be recycled and used to clear other proteins. A missense mutation, isoleucine to methionine at residue 93 (I93M), has been found in two siblings with typical PD in a small German pedigree in which the disease is transmitted as an autosomal dominant disorder with incomplete penetrance.5 No pathologic information is available yet to determine the presence of Lewy bodies associated with this mutation. UCH-L1 is an abundant brain protein accounting for 1 to 2% of total soluble brain proteins present in all neurons.51 The mutant form of UCH-L1 has diminished enzymatic activity resulting in impaired protein clearance through the ubiquitin-proteasome pathway.5 The I93M mutation in this gene has yet to be discovered in other families with PD, suggesting that it is a very rare contributor to PD, or even raising the speculation that it could represent a harmless polymorphism occurring in a single family by chance.52 Confirmation of mutations in UCH-L1 in other families would certainly strengthen the role of this gene in PD particularly since the original report was not based on a genome wide screen.

The function and biochemical properties of the three gene products described above point to aberrations in the main protein-clearing machinery of cells as the central process leading to abnormal accumulation of these proteins or their substrates, as well as to the consequences of their interaction with their protein partners in the parkinsonian nigra (see figure 1). It is not surprising, then, that Lewy bodies consist largely of such proteins, including α-synuclein, synphilin-1, Parkin, UCH-L1, ubiquitin, and 26S proteasome subunits, among other molecules.

Selective vulnerability of dopaminergic neurons in PD.

The involvement of select groups of neuronal populations in the pathologic process of neurodegenerative disorders in general and PD in particular is an important issue for several reasons. In addition to the fact that this differential susceptibility dictates the clinical phenotype of these disorders, it raises intriguing questions about the molecular determinants of this pathologic phenomenon, and provides clues about therapeutic approaches. Available evidence suggests that the relative sensitivity of different neuronal populations to various disease states is likely multi-factorial, involving the expression levels of other proteins or signaling molecules. In the case of PD, several such factors could collectively contribute to the selective vulnerability of nigral neurons. These include the tendency of α-synculein and Parkin to be expressed in melanin containing dopaminergic neurons of the nigra,53 the presence of dopamine transporter that pumps MPP⁺ and perhaps other neurotoxins into these cells,54 lack of calbindin D_28K in the nigrosomes where neuronal loss is greatest,55 and the presence of neuromelanin as a source of iron.56 Whether any of these factors alone is sufficient to uniquely confer the cell specificity of the pathology in PD remains to be demonstrated.

The synthesis of dopamine itself by nigral neurons, which bear the brunt of the pathology, appears a critical variable underlying the death of these neurons. Dopamine is a pro-apoptotic neurotransmitter that activates a well-defined cell death cascade. This is because dopamine metabolism produces reactive oxygen species. The latter is accomplished through both an enzymatic reaction generating hydrogen peroxide (H₂O₂), which in turn is converted to the highly toxic hydroxyl radical by the Fenton reaction,57 and through a chemical reaction generating superoxide anion, which in the presence of nitric oxide forms peroxynitrite.58 When dopamine is synthesized by a cell or crosses its plasma membrane through the dopamine transporter, it generates reactive oxygen species which result in the activation of p38 mitogen-activated protein (MAP) kinase, as well as c-Jun N-terminal kinase (JNK).59,60⇓ These, in turn, cause the release of cytochrome c from mitochondria, leading to the cleavage of caspase 9 and eventually the activation of the final executioner, caspase 3 (figure 2).59

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. Dopamine-induced apoptotic cascade. Dopamine either synthesized by a cell or transported across the plasma membrane generates reactive oxygen species (ROS).

The ability of dopamine to produce oxidant damage is consistent with postmortem changes in several indices of oxidative stress in the parkinsonian nigra. These include elevations in iron, ferritin, and nitric oxide, as well as markers of general oxidant damage to proteins, lipids, and DNA. Specifically, α-synuclein itself is nitrated within Lewy bodies.61 Conversely, indices of protective markers are decreased in the parkinsonian nigra, including reduced glutathione, mitochondrial complex I, calbindin D_28K, and transferrin. Although there is no conclusive evidence that these markers are the primary initiators of neuronal death in PD, they can clearly contribute to its progression by perpetuating the effects of a primary insult. Interestingly, this state of oxidative stress increases the tendency of α-synuclein to aggregate32 presumably into toxic oligomers. Conversely, the overexpression of α-synuclein itself in cellular models leads to the generation of excessive amounts of reactive oxygen species (Junn et al. submitted),34 thereby creating a vicious cycle catapulting dopaminergic neurons to their demise. Therefore, potential environmental factors that lead to increased oxidant stress could induce the aggregation of α-synuclein in a genetically susceptible individual who expresses a certain level of this protein in his or her dopaminergic neurons, leading to cell death. Such a scenario provides one possible link between the role of environmental factors and genetic predisposition.

Specific mutations in α-synuclein, Parkin, and perhaps UCH-L1 are linked with inherited forms of PD. Reports of increased association of certain polymorphisms in these genes with sporadic PD remain to be substantiated.62 Yet, these gene products, and particularly α-synuclein, appear to play important pathogenetic roles in nonfamilial PD. Current knowledge about these proteins indicates that 1) mutations in α-synuclein are not necessary for its pathogenicity; 2) high levels of α-synuclein appear deleterious to dopaminergic neurons due to its tendency to aggregate and cause oxidant stress; and 3) normally functioning Parkin and UCH-L1 are important to maintain the homeostatic function of cells by clearing proteins at appropriate rates (see figure 1). These conclusions provide general targets for future curative therapies to prevent or slow down cell death in PD, even in sporadic cases.