T B P ' D: HE Iochemistry of Arkinson S Isease

27 Apr 2005 3:58 AR AR261-BI74-02.
tex XMLPublishSM (2004/02/24) P1: JRX

10.1146/annurev.biochem.74.082803.133400
Annu. Rev. Biochem. 2005. 74:29–52

doi: 10.1146/annurev.biochem.74.082803.133400
Copyright c 2005 by Annual Reviews. All rights reserved
First published online as a Review in Advance on January 17, 2005
THE BIOCHEMISTRY OF PARKINSON’S DISEASE∗

Mark R. Cookson
Cell Biology Section, Laboratory of Neurogenetics, National Institute on Aging,
Bethesda, Maryland 20892; email: [email protected]
Annu. Rev. Biochem. 2005.74:29-52. Downloaded from www.annualreviews.org
Key Words Lewy body, mitochondria, parkinsonism, proteasome, protein

aggregation
by Pennsylvania State University on 10/02/13. For personal use only.
■ Abstract Several genes have been identified for monogenic disorders that variably
resemble Parkinson’s disease. Dominant mutations in the gene encoding α-synuclein
enhance the propensity of this protein to aggregate. As a consequence, these patients
have a widespread disease with protein inclusion bodies in several brain areas. In
contrast, mutations in several recessive genes ( parkin, DJ-1, and PINK1) produce
neuronal cell loss but generally without protein aggregation pathology. Progress has
been made in understanding some of the mechanisms of toxicity: Parkin is an E3
ubiquitin ligase and DJ-1 and PINK1 appear to protect against mitochondrial damage.
However, we have not yet fully resolved how the recessive genes relate to α-synuclein,
or whether they represent different ways to induce a similar phenotype.
CONTENTS
DEFINITIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
α-SYNUCLEIN AND AGGREGATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
PARKIN AND THE PROTEASOME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
DJ-1 AND PINK1: MITOCHONDRIAL CONNECTIONS? . . . . . . . . . . . . . . . . . . . . 42
PATHWAYS TO PARKINSONISM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
CONCLUDING REMARKS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
DEFINITIONS
For many of us, our opinions about diseases are greatly informed by the infectious
diseases, where we can rely on Koch’s postulates to tell us causation. We can
isolate and culture the microorganism responsible for infection, introduce it into
a host, and see the disease recapitulated. However, these ideas are less useful in
understanding pathogenesis in the neurodegenerative disorders in which sporadic
cases occur without readily identifiable causes. We can circumvent this to some
extent by examining monogenic forms of the disease. Here there is a much simpler
∗
The U.S. Government has the right to retain a nonexclusive, royalty-free license in and to
any copyright covering this paper.
0066-4154/05/0707-0029$20.00 29
27 Apr 2005 3:58 AR AR261-BI74-02.tex XMLPublishSM (2004/02/24) P1: JRX
30 COOKSON
cause and effect relationship: inherit a faulty copy (or copies) of the gene and
develop disease. Knowing the gene identity and the nature of mutations, one can
introduce these into animals or other models and induce a facsimile of the disease.
An underlying assumption is that the disease process might also be similar for the
sporadic forms of these disorders.
There is a special problem of definition when discussing Parkinson’s disease
(PD). Although there are multiple loci with PARK designation (Table 1), the
TABLE 1 Monogenic forms of Parkinson’s disease
Locusa Gene Protein function Phenotypeb

PARK1 SNCAc Synaptic? PD/DLBD

PARK4 (AD) Lipid binding? Onset from age 30 to 60,

rapid course.
Fulminant Lewy bodies
PARK2 (AR) Parkin E3 ligase Parkinsonism
Onset from teenage to 40s,
slow course.
No Lewy bodies, except one
case
PARK3 (AD) Unknown; PD, dementia
Chr2p13d — Onset from age 50 to 60s
Lewy bodies, tangles and
plaques.
PARK5 (AD?) UCHL1 Ubiquitin hydrolase/ligase Typical PD
Onset at about age 50
Unknown pathology
PARK6 (AR) PINK1 Protein kinase Parkinsonism
Onset from age 30 to 50s
Unknown pathology
PARK7 (AR) DJ-1 Oxidative stress response? Parkinsonism
Onset from age 20 to 40s,
slow course.
Unknown pathology
PARK8 (AD) Unknown; Parkinsonism
Chr12p1-q13 — Onset from age 40 to 60s
Variable pathologye
a
PARK loci are shown for the monogenic forms. Inheritance is shown in parentheses below each loci: AD, autosomal dominant;
and AR, autosomal recessive.
b
PD, Parkinson’s disease; DLBD, diffuse Lewy body disease.
c
SNCA is the gene name for α-synuclein.
d
The PARK3 linkage was described in Reference 148.
e
PARK8-linked families have been described by two groups as either Lewy body negative or variably Lewy body and tangle
positive. See References 149–150.
PARKINSON’S DISEASE 31
syndromes for each are distinct. The genes for six of the loci are known and
will be discussed in the text. Additionally, there are two extant loci for which the
underlying genetic lesion has not been identified. Each gene will be discussed sep-
arately, but broadly there are two clinicopathological components. Firstly, there is
clinically defined parkinsonism, a syndromic term which encompasses the cardi-
nal features of the Parkinsonian movement disorder. These are a resting tremor,
bradykinesia (slowness of movement), rigidity and postural instability, all prob-
lems in initiating or stopping movement. Their pathological correlate is the loss
of dopaminergic neurons in the substantia nigra pars compacta (SNpc). Although
there are many other neuronal groups affected in different brain regions in PD
(1), nigral degeneration and resultant parkinsonism is a consistent feature of all
the genetic conditions discussed here. Secondly, PD is marked postmortem by the

presence of Lewy bodies (LBs) and Lewy neurites in surviving neurons (2). These
are intracellular aggregations of lipids and proteins that were first identified by
eosin staining and now by immunostaining for their protein components including
ubiquitin and α-synuclein. Pathological definitions of PD require the presence of
α-synuclein-positive Lewy pathology in surviving nigral neurons, combined with
nigral cell loss and intact striatal neurons (3). α-Synuclein also marks the lesions
found in a range of related disorders, referred to as synucleinopathies (4). Impor-
tantly, in diffuse Lewy body disease (DLBD), Lewy pathology is found in many
brain areas and is associated with the expanded phenotype of these cases, such as
dementing illness and fluctuations in consciousness (5).
PD is therefore a disease with two parts, parkinsonism and Lewy pathology.
Neither of these alone precisely defines PD as both occur independently in other
neurological settings. However, understanding these two components is critical to
understanding the biochemistry of PD and how the genetic forms of parkinsonism
relate to this process. Conveniently, α-synuclein is both part of Lewy pathology
and a cause of dominantly inherited disease.
α-SYNUCLEIN AND AGGREGATION

In 1997, Polymeropolous and colleagues identified a dominant mutation in the
α-synuclein gene in a number of families of Greek or Sicilian ancestry (6). This
gene had been cloned previously as a precursor to a small peptide found in the
Alzheimer’s brain (7). Other homologues had been cloned as a gene upregulated
by song learning in the zebra finch (8) or as an abundant protein from the electric
eel neuromuscular synapse (9), where the protein was named for its localization to
synapses and nuclei. In fact, α-synuclein is part of a gene family including β- and
γ -synucleins and synoretin (10). All synucleins have a series of imperfect repeats
including the sequence motif KTKEGV and a variable C-terminal tail, which is
highly acidic in α-synuclein. Synucleins are also basally phosphorylated at serine
and tyrosine residues. Some of these structural motifs in α-synuclein are shown in
Figure 1. There is little or no detectable secondary structure in solution, and hence
α-synuclein is referred to as natively unfolded.
32 COOKSON
Figure 1 Motifs in the α-synuclein protein. The natively unfolded α-synuclein protein
is shown in a linear form. PINK1 shaded areas represent the imperfect KTKEGV repeats.
Human mutations are shown in red and map to the repeat region. At the C-terminal end
of the protein is an acidic tail, containing several sites of phosphorylation (green). The C
terminus also contains the alternatively spliced exon V and a calpain I cleavage site. The
acidic tail tends to decrease protein aggregation, whereas a hydrophobic region near the
imperfect repeats promotes aggregation.
The functions of the synucleins are not well understood. α-Synuclein binds
to lipid membranes, forming an amphipathic helix (11). Given the location of
a pool of α-synuclein at synaptic membranes, there may be a synaptic role for
the protein. In support of this idea, α-synuclein knockout mice have synaptic
deficits. An early report suggested a loss of dopamine release in the striatum (12),
whereas a more recent study showed a decrease in paired-pulse facilitation in the
hippocampus (13). This may correlate with a loss of synaptic markers in antisense-
treated hippocampal cultures (14). Taken together, these observations suggest that
α-synuclein plays a role in regulating the reserve pool of synaptic vesicles in brain.
Possibly related to lipid binding, α- and β-synucleins also inhibit phospholipase
D2 at physiological concentrations (15, 16). However, knockout animals show no
PD-like symptoms, suggesting that loss of protein function does not cause disease.
The first α-synuclein mutation that was discovered is an A53T point substitution
(6). An unusual aspect of the mutation is that the amino acid is already a threonine
in rodents and other species (17). Subsequently, three additional mutations have
been found: A30P in German kindred (18), E46K in a Spanish kindred (19), and a
triplication of the wild-type gene in a large family from Iowa (20). Pathology from
three of these kindreds is available and shows α-synuclein-positive Lewy bodies
in the brainstem as well as nigral cell loss. However, α-synuclein pathology is not
limited to the nigra in many of these cases. In fact, the clinical descriptions of
many of the patients with α-synuclein mutations reflect a disease with prominent
dementia, presumably a reflection of the widespread cortical Lewy bodies in these
cases. The Spanish E46K mutation was reported as “Lewy body dementia” (19).
In the Iowan kindred, glial cell inclusions were found (21), which would otherwise
be typical of multiple system atrophy (MSA), another synucleinopathy. Therefore,
mutations in α-synuclein produce a fulminant disease that includes parkinsonism
but is much more widespread and may resemble DLBD. The disorder is also more
progressive and tends to have an earlier onset than sporadic PD.
α-Synuclein is one of several proteins associated with neurodegenerative dis-
eases that have a high propensity to aggregate. β-pleated sheet-like bonding
stabilizes the aggregated forms. This contrasts with the unstructured protein in
solution or folding when bound to lipid, earning α-synuclein the title of “a protein
chameleon” (22). A central hydrophobic region of α-synuclein, near the repeats,
tends to self-associate, contributing to aggregation (23, 24). This region is not

shared with (for example) β-synuclein, and consequently these homologues vary
radically in their propensity to aggregate. In fact, β-synuclein can prevent α-
synuclein aggregation in vivo (25) and in vitro (26). The C-terminal acidic tail of
α-synuclein inhibits aggregation and, hence, truncated forms are more prone to
aggregate (27, 28). This ability of wild-type α-synuclein to aggregate presumably
explains its presence in the several sporadic synucleinopathies.
The end product of α-synuclein aggregation is the formation of heavily insolu-
ble polymers of protein known as fibrils. It is thought that fibrillar α-synuclein is the
building block of Lewy bodies. The most direct evidence for this is immunogold
labeling showing that α-synuclein is present at sites along fibrils isolated from
Lewy bodies (29). Lewy bodies contain many proteins other than α-synuclein,
including neurofilaments and other cytoskeletal proteins, suggesting there are co-
precipitants that might be important in aggregation. However, fibrils can be formed
in vitro from α-synuclein alone, suggesting that this protein is sufficient to form
inclusions [reviewed in (22)]. α-Synuclein is also the most sensitive marker for
Lewy bodies, implying that it is necessary for Lewy body formation (30).
Although the A53T mutation promotes the formation of such fibrillar species,
A30P does not. In fact, A30P slows the rate of fibril accumulation but strongly
promotes the formation of oligomeric species (31, 32). No studies on the E46K
mutation have been performed to date, but the pathology in these cases suggests
the mutation would have an effect on fibril formation. Because aggregation is a
concentration- and nucleation-dependent process (33), the Iowan triplication is
predicted to promote accumulation of oligomers and fibrils. We have recently
shown that oligomeric α-synuclein is deposited in detergent-insoluble fractions
from brains of patients with the triplication mutation (34). Figure 2 shows an
outline of the α-synuclein aggregation pathway, with an emphasis on the effects
of different mutations.
The fact that all mutations promote the formation of oligomeric rather than
fibrillar species has led some to suggest that oligomers, not fibrils, are toxic.
Oligomers, also referred to as protofibrils (35), can form annular structures that
34
27 Apr 2005 3:58
AR
COOKSON
AR261-BI74-02.tex
XMLPublishSM (2004/02/24)
P1: JRX
may have pore-like properties and might damage membranes (36). Recently, annu-
lar synuclein oligomers have been isolated from human brain samples (37). There
is some additional evidence to support the idea that oligomers are the toxic species
for α-synuclein. In most cell culture models, toxicity is seen without heavily aggre-
gated α-synuclein, and it has been suggested that soluble species mediate toxicity
(38). α-Synuclein aggregation and deposition into insoluble fractions occurs later
than cell death in vitro (39). Conversely, Lewy body formation involves deposits
of fibrillar α-synuclein into very insoluble fractions. Therefore, α-synuclein ag-
gregation is the key step that drives both pathology and cellular damage, but these
two outcomes can be dissociated from each other.
Common effects between mutations, such as the formation of oligomers, are
important to note, as they are more likely to be related to the pathogenic mech-
anism than differential effects. However, this is true only if both mutations are
unambiguously causal. The A30P family is relatively small, leading to concerns

about whether the mutation is truly causal, although the mutation does segregate
with disease with a reasonably high penetrance (40). A second difficulty is the
lack of autopsy studies. Positron emission tomography data show presynaptic ni-
gral cell loss, and the clinical phenotype is reported to be “to that of sporadic” PD
(40). However, as PD is defined primarily by pathology, the diagnostic accuracy
is unclear. If oligomeric species mediate toxicity to neurons and A30P promotes
only oligomer formation, and if Lewy bodies are a consequence of the formation
of fibrillar species, then one might predict that the A30P family would have much
fewer Lewy bodies than the other mutations and might be defined as parkinson-
ism not PD. However, assuming that the A30P mutation is causal, then all of the
α-synuclein mutations share the same causal agent and so have the same etiology.
It is clear from the above discussion that protein aggregation is important in
the disease, and hence it is critical to elucidate both genetic and nongenetic fac-
tors that increase this aggregation. Promoter alleles that increase expression of
←
Figure 2 The pathogenic cascade of α-synuclein aggregation. α-Synuclein exists in
solution as an unstructured monomer, shown as a linear structure, similar to Figure 1.
Inside the cell, the monomer is in equilibrium with membrane-associated forms with
higher helical content, shown schematically as an amphipathic helix. In the helix, blue
and red circles indicate charged residues, gray circles are nonpolar and hydrophobic
amino acids. The A30P mutation disfavors membrane binding. The green arrows in-
dicate the pathogenic formation of aggregated species. All mutations reported to date
increase the rate of formation of oligomers or protofibrils, which may also produce
pores. Oligomers and other intermediates are kinetically stabilized by dopamine (DA).
However, these are transient species that further aggregate to form mature fibrils, which
are stabilized by β-sheet-like interactions and are highly insoluble. The formation of
Lewy bodies is presumed to be a consequence of fibrillization. Events such as the
attachment of ubiquitin (black dot in the figure) are thought to be secondary to the
initial aggregation and deposition processes.
36 COOKSON
α-synuclein (41), therefore increasing protein concentration, are associated with

sporadic PD (42). Metals, pesticides, and oxidizing conditions all promote α-
synuclein aggregation [reviewed in (43)], suggesting a reason why these appear
to play a role in sporadic PD. α-Synuclein also undergoes a number of posttrans-
lational modifications, some of which enhance aggregation. The phosphorylated
form of α-synuclein is found in Lewy bodies (44). However, given that α-synuclein
is basally phosphorylated, whether phosphorylation precedes aggregation and/or
deposition is unclear. Tyrosine nitration (45) and methionine oxidation (46, 47) also
occur, although again whether nitration or oxidation is required for α-synuclein
toxicity has not been directly tested. Finally, cleavage of α-synuclein by calpain I
is also associated with deposition (48). Because the C-terminal region of the pro-
tein prevents aggregation, removal is predicted to promote aggregation. Calpain I

cleavage has also been associated with neuroprotection in vitro (49).
Although α-synuclein is expressed in many tissues, symptoms are restricted

to particular neuronal cells. Part of the reason is that aggregation does not occur
equally in all tissues. For example, α-synuclein is not as heavily aggregated in
blood samples as in the brain from patients with the triplication mutation (34).
Why is the brain vulnerable to these processes when other tissues also express the
same protein? There are likely to be several contributing factors. Expression of
α-synuclein is high in the brain compared to other tissue (50). Brain also has a high
concentration of macromolecules, which may promote aggregation by molecular
crowding (51, 52). The brain may have a higher oxidative stress level than other
tissues. Some neuronal groups may be highly vulnerable because catechols, in-
cluding dopamine, can stabilize oligomeric intermediates of aggregation (53).
Conversely, inhibition of dopamine synthesis ameliorates α-synuclein toxicity
(38). These observations reinforce the idea that oligomeric α-synuclein species are
toxic.
One aspect that deserves special discussion is the potential role of lipids. As
stated above, α-synuclein adopts a helical conformation when bound to lipid mem-
branes. This is a fairly stable conformation, which would inhibit conversion of
protein into fibrillar forms (54). However, lipids can also promote α-synuclein
aggregation, especially forming oligomeric species (55, 56). If oligomers were the
toxic species, then lipids would promote α-synuclein toxicity. In support of this,
genes that suppress α-synuclein toxicity in yeast are clustered in families related
to lipid metabolism and vesicle transport (57). Relating this to human pathology,
Lewy bodies contain lipids at their core (58). The A30P mutation, but not A53T,
decreases α-synuclein’s affinity for artificial lipid membranes (59) and promotes
its cytosolic accumulation in yeast (57).
This discussion suggests that α-synuclein is the causative agent in PD/DLBD.
Stretching the analogy to Koch’s postulates further, can we reintroduce α-synuclein
mutations into a susceptible organism and reproduce the disease? α-Synuclein
transgenic animals have been made in a number of species, including mice,
Drosophila melanogaster, and Caenorhabditis elegans [reviewed in (1, 60)]. Viral
delivery of α-synuclein induces nigral degeneration in rats (61) and primates (62).
These apparently solve the problem of whether α-synuclein is the causal agent, as
the disease is replicated in the model organism.
A difficulty is that the point mutations apparently behave differently in different
contexts. For example, in two models, A53T α-synuclein promotes damage to the
spinal cord and leads to complex motor phenotypes without nigral cell loss (63,
64). Despite matched levels of expression, A30P α-synuclein does not produce an
equivalent phenotype. This observation has been used to suggest either that A30P
is not pathogenic (64) or that fibrils are more toxic than protofibrils (65). However,
in other models A30P produces several effects, including intracellular inclusions
and neuronal damage (66). In Drosophila, A30P is more toxic than wild type or
A53T (67). Further complicating the picture, nigral neurons remain intact in all
of the transgenic mouse models to date, but loss of dopaminergic neurons is seen
in transient, viral-induced experiments in rats or primates. Why these models are
quite so variable is unclear.

An important point in thinking about the different mutations in α-synuclein is
that the difference between wild-type and mutant proteins is quantitative rather
than qualitative. In other words, the properties of α-synuclein that cause dam-
age to neurons are not specific to the mutants but are already a function of the
wild-type protein. This is neatly illustrated in experiments using inducible pro-
moters to drive α-synuclein expression and toxicity in human cell lines (38). At
lower expression levels, mutant α-synuclein is more toxic than wild type, but at
higher expression levels, all variants are similarly damaging. Although it is difficult
to directly compare the different transgenic models, mutant proteins are associated
with greater amounts of pathology until the wild-type protein is expressed at high
levels. The corresponding dosage curve for human studies may come from studies
of multiplication mutations. The triplication, with double the protein load, includes
Lewy pathology and cellular damage in many brain regions. At the other end of
the range, different promoter variants may more subtly affect expression (42) and
are associated with idiopathic PD.
In summary, α-synuclein is a protein with a natural tendency to aggregate into
oligomers that can then further aggregate into fibrils that are deposited as Lewy
bodies and similar pathologies. Mutations variably produce either oligomers/proto-
fibrils or fibrils. The fibrillar mutations (A53T, triplication of the wild-type alleles,
and probably E46K) produce a widespread Lewy body disease in humans that
includes parkinsonism but also has elements of DLBD. PD/DLBD is therefore a
protein aggregation disorder, which could lead to therapeutics aimed at preventing
or reversing α-synuclein aggregation. For example, peptides directed at the central
portion of α-synuclein can prevent or limit aggregation and toxicity (68), and
antibodies to aggregated α-synuclein (69) may also be beneficial if they can be
expressed inside cells. There are other possible routes to interfere in the pathogenic
process, such as any of the downstream effects caused by the presence of aggregated
proteins in a neuron. Because these might overlap with the cellular functions
impacted by the recessive mutations associated with inherited parkinsonism, these
are discussed in the next two sections.
38 COOKSON
PARKIN AND THE PROTEASOME

The first gene that causes recessive parkinsonism was named parkin by Mizuno and
colleagues (70). These patients had early-onset parkinsonism (teens to twenties)
with slow progression and additional features such as dystonia. Subsequent studies
suggest that parkin mutations are the most numerous cause of recessive, early-onset
parkinsonism [reviewed in (71)]. Identification of additional mutations shows that
the phenotype can be expanded to include cases with features more typical of
sporadic PD (72). One discrepancy between parkin disease and PD is the absence
of Lewy bodies, although there is one exception (73). This suggests Lewy body
formation is not required to evoke nigral cell death, i.e., that there are other ways
to kill neurons. So parkin disease is a phenocopy of PD, parkinsonism without

Lewy bodies. There are other forms of parkinsonism with these features, notably
exposure to the toxin 1-methyl-4-phenyl-1,2,4,6,-tetrahydropyridinium (MPTP)

(74).
Subsequent studies have shown that parkin encodes an E3 protein-ubiquitin
ligase (75, 76). E3 ligases control the key step in the cycle of ubiquitin-mediated
hydrolysis of damaged or misfolded proteins that are degraded via the proteasome.
The reaction promoted by E3 ligases is the addition of a lysine-linked chain of
four or more ubiquitin molecules to the target protein, which is recognized by
subunits in the proteasome lid. An outline of the reaction scheme catalyzed by
parkin is shown in Figure 3. Parkin’s domain structure reflects this role. At the
N terminus of parkin is a ubiquitin-like domain (Ubl) that interacts with Rpn10
→
Figure 3 A parkin-centric view of the ubiquitin-proteasome system. Ubiquitin is
activated by the enzyme E1 (yellow), represented here by a red circle around the black
dot of ubiquitin. After activation, ubiquitin is transferred to an E2 enzyme (blue), which
docks with E3s including parkin (green). Parkin contains two RING (really interesting
new gene) domains separated by an IBR (in-between ring) motif, and the E2 is probably
recruited to this region. Substrates (red diamond; see text for description of the different
candidates) then bind to the same region of parkin. For simplicity, substrates are shown
binding to RING1 and E2 to RING2, but as there are no data on how parkin is folded,
we cannot be sure of the exact spatial arrangement of these components. Activated
ubiquitin is transferred from the E2 to the target and, by analogy to other E3 ligases,
there is no transfer to parkin itself. This process is repeated until a string of four
or more ubiquitin molecules, linked by lysines to each other and to the substrate, is
formed. This is recognized by the proteasome, which degrades the protein into small
peptides and amino acids. During all or some of this process, Parkin may be tethered to
the proteasome by interactions of its N-terminal ubiquitin-like (Ubl) domain. Prior to
substrate degradation, the polyubiquitin chain is removed and recycled to monomeric
ubiquitin by a series of enzymes including the ubiquitin C-terminal hydrolases. Parkin
is also freed to participate in further reactions.
in the proteasome cap (77). It is likely that the Ubl domain tethers parkin close
to the proteasome, directing poly-ubiquitylated proteins toward their proteolytic
end. At least one recessive parkin mutation (R42P) disrupts this interaction (77).
At the C-terminal region of parkin, there are two RING (Really Interesting New
Gene) fingers. RING fingers are found in a number of E3 ligases and have varying
numbers of cysteine and histidine residues that coordinate a structurally important
zinc atom (78). Parkin has two RING fingers of the C3 HC4 type separated by an in-
between RING (IBR) domain, again typical of E3 ligases. The RING-IBR-RING
40 COOKSON
motif is important in recruiting substrates and the E2 enzyme (UbcH7 or UbcH8)

that carries activated ubiquitin. The precise mechanism by which parkin promotes
ubiquitylation of its substrates is not fully defined, in part because the structure
of parkin has not been solved. However, other E3 ligases that have similar RING
domains act as scaffolds to bring the ubiquitin bound to the E2 close to target
lysines on the substrate protein. These E3 ligases are not catalysts as the reaction
does not proceed via a thiol intermediate as on the E3 but is based on a proximity
effect. A by-product of this reaction scheme is that these E3 enzymes undergo
autoubiquitylation, a phenomenon that is easily seen with parkin.
There are several targets for parkin’s E3 ligase activity. These include the septins
CDC-rel1 and CDC-rel2 (75, 79), cyclinE (80), p38 tRNA synthase (81), Paelr1
(82), synaptotagmin XI (83), and synphilin-1 (84), as well as parkin itself (75, 76). It
is notable that some of these proteins are synaptic, as parkin is tethered to synaptic
densities by an interaction with the PDZ protein Cask (85). This implies that
parkin could have a role in synaptic function. In most cases, parkin can ubiquitylate
substrates without additional binding proteins, but parkin does require an additional
protein (hSel10) for activity against cyclin E (80). Parkin may be present as part of a
modular complex in the brain with additional proteins that act to control substrate
specificity. At least one additional protein can also bind to parkin, the U-box
protein CHIP (86). CHIP interacts with Hsp70 to ubiquitylate misfolded proteins.
Hsp70 and CHIP are important components of the decision-making machinery of
the cell to direct either protein refolding in the endoplasmic reticulum (ER) or
retrotranslocation to the cytosol and degradation.
There are suggestions that parkin inactivation may play a role in typical PD.
Two groups have recently shown that exposure to nitric oxide (NO) alters parkin’s
E3 ligase activity (87, 88). The mechanism involves NO-derived radical species
that attack cysteine residues in RING1. As these cysteines are important in coor-
dinating the structural zinc molecule, this will affect protein folding and, hence,
enzyme activity. In one study, decreased activity of parkin toward synphilin-1 was
noted (87), whereas increased autoubiquitylation was seen in the other study (88).
Whether small differences in experimental conditions are responsible for these ap-
parent discrepancies remains to be resolved. One notable difference is that shorter
exposures to nitrosylating agents produce increased activity, whereas longer expo-
sures inhibit activity. Parkin is nitrosylated both in human tissue from PD patients
and in the MPTP and rotenone animal models (87, 88). NO-derived radicals are
important mediators of MPTP toxicity (1). Another observation suggesting parkin
may play a role in sporadic PD is the reported association with promoter polymor-
phisms (89), although conflicting results have been reported (90).
By implication, loss of parkin function results in the accumulation of one or
more of its substrates, which then leads to cell death of nigral neurons. Support
for this idea comes from experiments where overexpression of the parkin sub-
strates Paelr1 (91) produces dopaminergic cell death that can be rescued by parkin
but not its E3 inactive mutants. Another substrate, CDC-rel1, also causes cell
death restricted to the nigra in vivo (92). One genetic finding that has been widely
discussed in support of the proteasome hypothesis is the mutations in UCHL1

in two siblings with PD (93). This gene encodes a neuronal-specific ubiquitin
C-terminal hydrolase, whose enzymatic roles include the recycling of ubiquitin
chains back to monomeric ubiquitin and adding ubiquitin to already monoubiqui-
tylated α-synuclein (94). UCHL1 expression is restricted to neurons, making it an
interesting candidate for neurodegeneration. However, there is controversy with
the UCHL1 mutation, discussed in an excellent recent review (95). There have
been no additional families with UCHL1 mutations, despite extensive searches,
that would strengthen the case that this gene can be pathogenic for PD. Therefore,
whether UCHL1 mutations are truly causal for PD is unclear. There is, however, a
relatively common S18Y polymorphism in UCHL1 that has been associated with
risk for sporadic PD in several, but not all, studies (95). It is possible that UCHL1
plays some role in PD, but this requires further clarification.
Parkin is also neuroprotective against stresses in which the direct relationship

to its substrates is unclear. Because of its unambiguous contribution to dominantly
inherited PD, several laboratories have examined whether there is a relationship
between α-synuclein and parkin. For example, we showed that parkin, but not its
E3 inactive mutants, protects cells against mutant α-synuclein (96), which has been
confirmed in other laboratories (49, 87). Others have shown that parkin suppresses
mutant α-synuclein toxicity in Drosophila models (91, 97). Demonstrating again
that the difference between wild-type and mutant α-synuclein is qualitative rather
than quantitative, parkin can suppress the toxicity associated with expression of
high levels of α-synuclein in vitro (98).
The simplest explanation for this observation is that α-synuclein might be a
parkin substrate. The steady-state level of α-synuclein protein is not affected by
the expression of parkin in cell lines (84) or in Drosophila (91). Although there
have been suggestions that α-synuclein protein levels might respond to proteasome
inhibition in vitro, most studies have not noted any effect. One study suggested
that α-synuclein can be degraded by the proteasome in an ubiquitin-independent
fashion (99), which would not require an E3. There is also evidence for α-synuclein
degradation by lysosomal proteases (100, 101). Set against these observations is
the experimental evidence that there is a glycosylated form of α-synuclein, sp22,
that can be purified from human brain and is a substrate for parkin (102). To
date, there are no replications of this important result, so it is not quite clear whether
sp22 is present in some of the model systems where α-synuclein is demonstrably
toxic and, hence, whether formation of sp22 is required for α-synuclein toxicity.
If the relationship between parkin and α-synuclein is not one of enzyme and
substrate, then what are the mechanisms involved in neuroprotection? There are
several possibilities, which will be discussed later in this chapter.
Creating knockout models for parkin is one way to model the disease and
understand the pathogenic process. Two groups have produced mice with tar-
geted deletion of exon 3 of parkin. Neither shows loss of nigral neurons, although
there are subtle changes in dopaminergic neurotransmission (103, 104). Recently,
mice with an exon 7 deletion have been reported to have loss of neurons in the
42 COOKSON
locus coeruleus with attendant behavioral changes (105). Although there is no loss
of catecholaminergic neurons, Drosophila parkin knockouts show mitochondrial
damage and apoptosis of flight muscles (106). Interestingly, one of the knockout
mouse models shows deficits in mitochondrial respiration (107). Perhaps surpris-
ingly, then, the major phenotype reported in parkin models is mitochondrial. Parkin
rather specifically prevents mitochondrial cytochrome c release and apoptosis
in vitro (108). This observation has gained greater weight as it has been real-
ized that other recessive genes for parkinsonism also impact on mitochondrial
function.
DJ-1 AND PINK1: MITOCHONDRIAL CONNECTIONS?

Since the discovery of parkin mutations, two additional genes have been found
associated with recessive parkinsonism, DJ-1 (109) and PINK1 (110). Patients
with mutations in either of these two genes have similar phenotypes to each other
and to parkin. Onset is generally early (from ages 30 to 50), and the course is benign
with long disease duration. Individuals with DJ-1 (111) or PINK1 mutations (112)
have loss of presynaptic dopaminergic function, although no autopsy studies are
yet available.
Because it is recessive, we might expect mutations in DJ-1 to be loss-of-function
mutations. Hence, understanding this normal function is critical. DJ-1 was cloned
as an interactor of GAPDH, playing a role in mRNA regulation by stabilizing
mRNA species after transcription (113). DJ-1 becomes more acidic under oxida-
tive conditions, one of only several proteins that do so (114). These two facts led
Bonifati and colleagues (109) to suggest that DJ-1 maintians neuronal viability
by modulating gene expression under conditions of cellular stress. DJ-1 has lim-
ited homology to several prokaryotic proteins, including cysteine proteases and
chaperones (115). Weak protease (116) and chaperone (117) activities have been
reported, although whether these are physiologically relevant has not been estab-
lished. Several groups have crystallized the DJ-1 dimer, and the structure indicates
that while there is a cysteine (C106) that might be catalytically active (117), it is
not adjacent to a histidine residue as seen in authentic proteases in the superfamily
(118).
The first description of DJ-1 mutations suggested that one point mutation,
L166P, promotes localization of the normally cytoplasmic DJ-1 protein to mi-
tochondria (109). This implies that loss of cytoplasmic function is sufficient to
induce parkinsonism. However, L166P more dramatically destabilizes the protein
(116, 119–123). If both cytoplasmic and mitochondrial pools of the protein are
depleted, then we cannot be sure that cytosolic DJ-1 is neuroprotective. In our
hands, the effect of L166P in redirecting the protein to mitochondria was minor;
both wild-type and mutant proteins could be found in either mitochondria or cy-
tosol (119). As an aside, although the proteasome may be involved in degradation
of L166P DJ-1, this does not imply that its function is related to parkin because
the proteasome is a major degradation route for many mutant proteins.
Some progress has come from recent work showing that DJ-1 oxidation is
directed toward a specific cysteine residue, C106 (124, 125). Oxidation converts
the sulfhydryl group of C106 to a sulfonic or sulfinic acid, which correlates with
the acidic pI shift of the protein. In our hands, DJ-1 is normally excluded from
mitochondria in the majority of cultured cells but moves to the outer mitochondrial
membrane under oxidative conditions, although this has not yet been confirmed
in vivo. Correspondingly, DJ-1 protects cells against mitochondrial complex I
inhibitors and other oxidative stresses (126, 127). Mutations at C106 that block the
ability of DJ-1 to respond to oxidative stress exclude the protein from mitochondria
and have a dominant negative effect on cell viability in response to mitochondrial
damage (125).
These observations support the contention that loss of function of DJ-1 sensi-
tizes neurons to oxidative stress (109). This would suggest that DJ-1 and parkin
have effects on distinct neuronal survival pathways. However, others have reported
that knockdown of DJ-1 sensitizes cells to both oxidative stress and proteasome
inhibition or ER stress (126). Therefore, there is some clarification required to
establish whether DJ-1 has a general effect on cell death or is specific to certain
types of stresses. DJ-1 does not seem to have an effect on staurosporine-induced
cell death, hence it does not suppress apoptosis per se (126).
Whether specific or general, the mechanism by which DJ-1 protects cells is
not clear. One possibility is that DJ-1 acts as an antioxidant, scavenging hydrogen
peroxide or other radical species (123, 127). Whether a protein that contains one
readily oxidized cysteine residue would have a substantial effect in a cellular con-
text where there are many other low-molecular-weight thiols is not known. Also,
at least in our hands, DJ-1 becomes oxidized under conditions where cell death is
not prominent. Perhaps oxidation of DJ-1 is a bellwether for the cell—warning of
more damaging conditions to come and protecting cells before conditions become
unmanageable.
What does DJ-1 do once oxidized? The recruitment to the outer mitochondrial
membrane (125), if reproducible, might suggest that there is a suppression of pro-
cell death pathways. Neuronal damage resulting from mitochondrial toxins such
as MPTP involves some of these pathways [reviewed in (1)]. However, no binding
partners at the outer mitochondrial membrane have been identified. Alternatively,
DJ-1 might act as a chaperone (117), although this has been challenged (116).
One could imagine that DJ-1 might play a role in suppressing neuronal damage by
promoting the refolding of damaging proteins. DJ-1 has been found in association
with intraneuronal inclusions formed by the microtubule-binding protein tau (66,
128) but only rarely labels Lewy bodies (129). As the neuropathology of DJ-1 cases
is not yet reported, whether DJ-1-mediated disease includes the protein inclusion
pathology resulting from lack of chaperone activity is not clear. DJ-1 may also be
acting as a protease, but the physiological targets for this activity have not been
identified. Returning to the hypothesis suggested by Bonifati and colleagues, DJ-
1 might alter the transcriptional profile of cells by binding to GAPDH (113) or
to components of the transcriptional machinery such as the SUMO-conjugating
44 COOKSON
enzyme PIASxα (130). Some of these binding partners also affect cell death;
GAPDH promotes apoptosis in several models. Given these many possibilities,
we can be sure only that DJ-1 is neuroprotective, arguably having something to do
with mitochondrial function.
The third of the trio of recessive genes is PINK1 (110). Protein function is rea-
sonably clear for PINK1, as it contains a serine/threonine-directed protein kinase
domain. Human PINK1 and its mouse homologue had been cloned previously and
possess authentic kinase activity against PINK1 itself (131, 132). PINK1 also has
an N-terminal mitochondrial localization signal and transfected PINK1 is found
in the mitochondria (110). A fusion protein with an N-terminal myc tag is present
as a single band and thus is likely to be the preprotein. This preprotein is present
in mitochondria, implying that PINK1 is imported and then cleaved; similar pro-
cesses are well described for other mitochondrial matrix proteins. Although there is
only one study to date, Valente and colleagues (110) reported that PINK1 protects
cells against apoptosis induced by exposure to proteasome inhibitors. Critically,
recessive mutations that are predicted to lack kinase activity were not neuropro-
tective. At the time of writing, there are some obvious gaps in our knowledge. The
first question is whether this protection is specific to proteasome-induced stress
or is more general. It is possible, for example, that PINK1 suppresses apoptotic
signaling in response to several different stressors. Secondly, it will be critical
to identify the physiological substrate for this mitochondrial kinase. Presumably,
the substrate(s) plays some role in mitochondrial responses to stress, which might
clarify whether the observed protection is specific or general.
PATHWAYS TO PARKINSONISM
The above discussion has deliberately separated the biochemistry of PD into two
components: α-synuclein aggregation and cell death of susceptible neurons. As
stated earlier, human genetics can indicate the proximal events in these diseases.
Therefore, in this bipartite view of the disease process, dominant synuclein muta-
tions account for the intracellular protein inclusion pathology, whereas recessive
mutations tell us more about the pathways that lead to cell death. The remaining
question is: Can we link these two sets of processes together experimentally or
logically?
To discuss this question, I start with the most recent discoveries and work back-
wards. Although it is too early to be certain if our initial predictions about DJ-1
and PINK1 will hold true in vivo, both of these proteins protect cells against loss
of mitochondrial function. It is unlikely that PINK1 and DJ-1 physically inter-
act because they are probably on opposite sides of the mitochondrial membrane.
The mitochondrial leader peptide of PINK1 should direct the kinase through the
mitochondrial import machinery into the mitochondria. In contrast, when DJ-1
overlaps with mitochondria, it localizes to the outer mitochondrial surface (125).
Therefore, if there is a common mitochondrial pathway for DJ-1 and PINK1, it is
at the level of the whole organelle. This is reasonable given the prominent roles
that mitochondria play in determining cellular life or death. However, if we sup-
pose that all these genes suppress cell death under many circumstances, it would
be surprising that cell death is restricted to a subset of neurons and not a more
general phenomenon. It is interesting that DJ-1 might be excluded from mitochon-
dria unless cells are stressed (125). If the cell survival pathway is more important
under specific conditions, then one might expect the pattern of cellular damage
to be restricted to cells that undergo these stresses or are especially vulnerable to
them.
If mitochondrial pathways are implicated in this scheme, we would predict that
the neurons affected in recessive parkinsonism would be susceptible to mitochon-
drial damage. Although exposure to MPTP produces a Parkinsonian syndrome,

toxicity is dependent on uptake by the dopamine transporter (74), so restriction
to dopaminergic neurons is not surprising. However, rotenone also inhibits mito-

chondrial complex without requiring uptake via the dopamine transporter, so this
is a test of whether some neurons are more sensitive than others to mitochondrial
damage. Chronic administration of rotenone selectively damages dopaminergic
neurons in the nigra (133), although some studies also report more generalized
toxicity. The mechanism appears to involve free radicals (134); hence DJ-1 should
protect neurons from rotenone toxicity. Therefore, mitochondrial complex I with
attendant oxidative damage might account for some of the neuronal damage in
Parkinsonian conditions, and we can relate this to loss of function mutations in
DJ-1 and PINK1.
Parkin does not easily resolve into this scheme. Given the ubiquitylation activity
of this enzyme, cell death is most obviously related to proteasome function. We
can ask whether proteasome inhibition would be sufficient to induce cell death and
whether such cell death would be restricted to the mosaic of cells susceptible in
PD. This experiment was performed recently, and the patterns of cell death closely
resemble those in sporadic PD (135). Proteasome inhibitors also preferentially
affect catecholaminergic neurons in some (96), but not all (136), in vitro models.
This implies that susceptible neurons in PD are linked by sensitivity to proteasome
dysfunction.
Therefore, there are at least two pathways that can lead to Parkinsonian syn-
dromes: mitochondrial and proteasomal. Logically, there are three ways to con-
sider these two pathways. Firstly, perhaps each is sufficient to induce cell death
and is independent of the other. In this scheme, the fact that some groups of neu-
rons are affected by these stresses is coincidental, and the human syndromes are
phenocopies of each other. A second possibility is that both mitochondrial and
proteasomal damage are required for neuronal damage, having initially distinct
events but converging on a later, common pathway. The third possibility is the
neatest: The three genes mark a single pathway that we can connect in an ordered
way.
However, it is hard to separate proteasomal from mitochondrial damage because
they interact with each other. Addition of proteasome inhibitors increases the
46 COOKSON
sensitivity of catecholaminergic neurons to rotenone or MPTP in vitro (136). In

this study, complex I inhibitors caused a decrease in proteasome activity. This may
be the result of ATP depletion, as the ubiquitin-proteasome system is very heavily
ATP dependent, or a consequence of oxidation, or both. Reciprocally, proteasome
inhibitors have been reported to cause mitochondrial damage (137). Therefore,
proteasomal and mitochondrial damage interact in both directions to converge on
cell death as an outcome.
It is also possible that we have misunderstood one or more of these components.
The evidence that parkin has an effect on mitochondria (138) is surprising for an
E3 ligase with no mitochondrial substrates. The effects of parkin can be specific, as
parkin is effective against apoptosis pathways that proceed through mitochondrial
signaling but not other triggers (108). A more powerful example is when pro-
teomics was used to examine the brains of parkin knockout mice. Although many
proteins were present on two-dimensional gels, mitochondrial proteins were specif-

ically represented (107). These observations led to the idea that mitochondria may
be important in parkin disease as well as DJ-1 and PINK1, but specificity is unclear.
Such considerations become much more complex when we attempt to add α-
synuclein into these schemes. The Dawson laboratory recently articulated these
difficulties by proposing that there are two logical models (139). In the “unify-
ing model,” parkin and synuclein have differential effects on a common pathway,
whereas in the “distinguishing model,” PD and recessive parkinsonism have differ-
ent pathways (139). This discussion centers on whether parkin plays an essential
role in α-synuclein disease and vice versa. Alternatively stated, the problem is:
What causes disease in these different conditions?
We can be sure that α-synuclein is causal in the PD/DLBD families, and it
is likely that protein aggregation underscores the disease process. What happens
downstream of protein aggregation to cause cell death is less clear. α-Synuclein
has detrimental effects on both proteasomal [(57, 96) and references therein] and
mitochondrial (140, 141) function. Interactions between mitochondria and pro-
teasomal function were discussed above, but aggregated α-synuclein can inhibit
the proteasome in vitro, suggesting that it might directly affect the ubiquitin pro-
teasome system (142, 143). Adding to the confusion, mitochondrial damage may
exacerbate α-synuclein aggregation, promoting the accumulation of the protein
posttranslationally (100). This leads to many schemes of the pathogenesis of PD
that evoke amplifying circles of mitochondrial damage, proteasomal dysfunction,
and protein accumulation.
If α-synuclein affects both mitochondria and the proteasome and if mitochon-
drial/proteasomal genes cause parkinsonism, does this mean that α-synuclein
mediates neuronal damage in the recessive diseases? α-Synuclein is a good can-
didate for being an endogenous stressor, as we know the wild-type protein can be
toxic when present at high levels. One might imagine that lack of protective gene
products (such as parkin, DJ-1, or PINK1) might make some neurons susceptible
to the same causal agent at normal expression levels. An argument against α-
synuclein involvement in cell death is the lack of obvious Lewy pathology in most
parkin cases, although there are reported exceptions (73). However, if Lewy body
formation is not required for toxicity, as implied by A30P, then we might not need
to see inclusion body pathology for α-synuclein to be toxic. Another argument is
that parkinsonism is a component of diseases caused by other aggregating pro-
teins, including tau mutations (144). There are several parallels between tau and
α-synuclein. Both are intracellular proteins with a tendency to aggregate, perhaps
coaggregating (145), and both proteins can cause cell death (146). Parkinsonism
can be a component of the phenotype of patients’ mutations in spinocerebellar
ataxia genes (147), which are associated with aggregations of polyglutamine pro-
teins. Therefore, α-synuclein is not the only brain protein that can aggregate and
kill nigral cells, although it is one of the few that aggregates so readily in its
wild-type form; tau is the other major one in brain.
In the absence of α-synuclein pathology in parkin, DJ-1, and PINK1 cases, we

cannot be certain about the causative agent in the same way as α-synuclein muta-
tions. However, parkin can protect cells against α-synuclein toxicity. To my mind,
this implies that the pathways triggered by the aggregating protein must converge
at some point on the positive effects of parkin and other recessive gene products.
A critical set of experiments will be to compare whether all three recessive parkin-
sonism genes are important in protecting against α-synuclein toxicity specifically
or against toxic proteins in general. More importantly, we need to better define the
relationships between the different recessive gene products and understand where
their effects are specific and where they only coincidentally affect the same cellular
processes.
Why should we care about the distinction between the concept of a single
pathway and multiple roads to the same output, even if it is a tractable problem?
The most practical benefit from really understanding the nature of cell death in PD
and related disorders is the possibility of providing new therapeutic avenues. In this
sense it is not critical whether events are early or late in the pathogenic process; each
is an avenue for intervention. Identifying the earliest and most specific events that
cause neuronal loss in these disorders might also indicate where to aim strategies
with the highest level of specificity. No matter what the underlying cause of cell
loss in parkin, DJ-1, or PINK1 disease, all three of these genes impact neuronal
ability to survive in the face of stress, and I think it can be no coincidence that
all three produce such a similar phenotype. I suspect that recessive genes tell us
why neurons are damaged in Parkinsonian conditions but that our thinking about
these is quite crude at this point. Certainly, we have yet to identify why different
neuronal groups rely on these proteins more than others, which may be key to the
problem of parkinsonism in many diseases.
CONCLUDING REMARKS
Identifying monogenic forms of PD has led us to a causative agent, α-synuclein.

It is probable that causation is shared with sporadic PD in which intracellular
aggregates of the same protein are found throughout the brain. Gene dosage is
important, and this is correlated with increased toxicity of the protein at higher
48 COOKSON
expression levels. There is currently an unclear relationship between α-synuclein

disease and recessive forms of parkinsonism. The challenges for the PD field are
to describe in detail the routes that lead to toxicity in these different situations and
answer whether, or not, these different pathogenic cascades overlap.
The Annual Review of Biochemistry is online at

http://biochem.annualreviews.org
LITERATURE CITED
1. Dauer W, Przedborski S. 2003. Neuron 17. Hamilton BA. 2004. Genomics 83:739–42
39:889–909 18. Kruger R, Kuhn W, Muller T, Woitalla

2. Braak H, Del Tredici K, Rub U, de Vos D, Graeber M, et al. 1998. Nat. Genet.
RA, Jansen Steur EN, Braak E. 2003. Neu- 18:106–8

robiol. Aging 24:197–211 19. Zarranz JJ, Alegre J, Gomez-Esteban JC,
3. Hughes AJ, Daniel SE, Ben-Shlomo Y, Lezcano E, Ros R, et al. 2004. Ann. Neu-
Lees AJ. 2002. Brain 125:861–70 rol. 55:164–73
4. Trojanowski JQ, Lee VM. 2003. Ann. NY 20. Singleton AB, Farrer M, Johnson J, Sin-
Acad. Sci. 991:107–10 gleton A, Hague S, et al. 2003. Science
5. McKeith I, Mintzer J, Aarsland D, Burn 302:841
D, Chiu H, et al. 2004. Lancet Neurol. 21. Gwinn-Hardy K, Mehta ND, Farrer M,
3:19–28 Maraganore D, Muenter M, et al. 2000.
6. Polymeropoulos MH, Lavedan C, Leroy Acta Neuropathol. 99:663–72
E, Ide SE, Dehejia A, et al. 1997. Science 22. Uversky VN. 2003. J. Biomol. Struct.
276:2045–47 Dyn. 21:211–34
7. Iwai A, Yoshimoto M, Masliah E, Saitoh 23. Giasson BI, Murray IV, Trojanowski JQ,
T. 1995. Biochemistry 34:10139–45 Lee VM. 2001. J. Biol. Chem. 276:2380–
8. George JM, Jin H, Woods WS, Clayton 86
DF. 1995. Neuron 15:361–72 24. Bodles AM, Guthrie DJ, Greer B, Irvine
9. Maroteaux L, Campanelli JT, Scheller GB. 2001. J. Neurochem. 78:384–95
RH. 1988. J. Neurosci. 8:2804–15 25. Hashimoto M, Rockenstein E, Mante M,
10. George JM. 2002. Genome Biol. 3:RE- Mallory M, Masliah E. 2001. Neuron
VIEWS3002 32:213–23
11. Clayton DF, George JM. 1998. Trends 26. Park JY, Lansbury PT Jr. 2003. Biochem-
Neurosci. 21:249–54 istry 42:3696–700
12. Abeliovich A, Schmitz Y, Farinas I, Choi- 27. Murray IV, Giasson BI, Quinn SM, Kop-
Lundberg D, Ho WH, et al. 2000. Neuron paka V, Axelsen PH, et al. 2003. Biochem-
25:239–52 istry 42:8530–40
13. Cabin DE, Shimazu K, Murphy D, Cole 28. McLean PJ, Kawamata H, Hyman BT.
NB, Gottschalk W, et al. 2002. J. Neu- 2001. Neuroscience 104:901–12
rosci. 22:8797–807 29. Crowther RA, Daniel SE, Goedert M.
14. Murphy DD, Rueter SM, Trojanowski JQ, 2000. Neurosci. Lett. 292:128–30
Lee VM. 2000. J. Neurosci. 20:3214–20 30. Sampathu DM, Giasson BI, Pawlyk AC,
15. Payton JE, Perrin RJ, Woods WS, George Trojanowski JQ, Lee VM. 2003. Am. J.
JM. 2004. J. Mol. Biol. 337:1001–9 Pathol. 163:91–100
16. Jenco JM, Rawlingson A, Daniels B, Mor- 31. Giasson BI, Uryu K, Trojanowski JQ, Lee
ris AJ. 1998. Biochemistry 37:4901–9 VM. 1999. J. Biol. Chem. 274:7619–22
32. Conway KA, Lee SJ, Rochet JC, Ding Mizuno Y, et al. 2003. J. Biol. Chem.
TT, Williamson RE, Lansbury PT Jr. 278:41890–99
2000. Proc. Natl. Acad. Sci. USA 97:571– 50. Ueda K, Fukushima H, Masliah E, Xia Y,
76 Iwai A, et al. 1993. Proc. Natl. Acad. Sci.
33. Wood SJ, Wypych J, Steavenson S, Louis USA 90:11282–86
JC, Citron M, Biere AL. 1999. J. Biol. 51. Shtilerman MD, Ding TT, Lansbury PT
Chem. 274:19509–12 Jr. 2002. Biochemistry 41:3855–60
34. Miller DW, Hague SM, Clarimon J, Bap- 52. Uversky VN, Cooper EM, Bower KS, Li
tista M, Gwinn-Hardy K, et al. 2004. Neu- J, Fink AL. 2002. FEBS Lett. 515:99–103
rology 62:1835–38 53. Conway KA, Rochet JC, Bieganski RM,
35. Lashuel HA, Petre BM, Wall J, Simon Lansbury PT Jr. 2001. Science 294:1346–
M, Nowak RJ, et al. 2002. J. Mol. Biol. 49
322:1089–102 54. Zhu M, Fink AL. 2003. J. Biol. Chem.

36. Volles MJ, Lansbury PT Jr. 2002. Bio- 278:16873–77
chemistry 41:4595–602 55. Cole NB, Murphy DD, Grider T, Rueter

37. Pountney DL, Lowe R, Quilty M, Vick- S, Brasaemle D, Nussbaum RL. 2002. J.
ers JC, Voelcker NH, Gai WP. 2004. J. Biol. Chem. 277:6344–52
Neurochem. 90:502–12 56. Sharon R, Bar-Joseph I, Frosch MP,
38. Xu J, Kao SY, Lee FJ, Song W, Jin LW, Walsh DM, Hamilton JA, Selkoe DJ.
Yankner BA. 2002. Nat. Med. 8:600–6 2003. Neuron 37:583–95
39. Gosavi N, Lee HJ, Lee JS, Patel S, Lee SJ. 57. Willingham S, Outeiro TF, DeVit MJ,
2002. J. Biol. Chem. 277:48984–92 Lindquist SL, Muchowski PJ. 2003. Sci-
40. Kruger R, Kuhn W, Leenders KL, Spren- ence 302:1769–72
gelmeyer R, Muller T, et al. 2001. Neu- 58. Gai WP, Yuan HX, Li XQ, Power JT,
rology 56:1355–62 Blumbergs PC, Jensen PH. 2000. Exp.
41. Chiba-Falek O, Touchman JW, Nussbaum Neurol. 166:324–33
RL. 2003. Hum. Genet. 113:426–31 59. Bussell R Jr, Eliezer D. 2004. Biochem-
42. Farrer M, Maraganore DM, Lockhart P, istry 43:4810–18
Singleton A, Lesnick TG, et al. 2001. 60. Lotharius J, Brundin P. 2002. Nat. Rev.
Hum. Mol. Genet. 10:1847–51 Neurosci. 3:932–42
43. Di Monte DA. 2003. Lancet Neurol. 61. Lo Bianco C, Ridet JL, Schneider BL, De-
2:531–38 glon N, Aebischer P. 2002. Proc. Natl.
44. Fujiwara H, Hasegawa M, Dohmae N, Acad. Sci. USA 99:10813–18
Kawashima A, Masliah E, et al. 2002. Nat. 62. Kirik D, Annett LE, Burger C, Muzyczka
Cell Biol. 4:160–64 N, Mandel RJ, Bjorklund A. 2003. Proc.
45. Giasson BI, Duda JE, Murray IV, Chen Q, Natl. Acad. Sci. USA 100:2884–89
Souza JM, et al. 2000. Science 290:985– 63. Giasson BI, Duda JE, Quinn SM, Zhang
89 B, Trojanowski JQ, Lee VM. 2002. Neu-
46. Hokenson MJ, Uversky VN, Goers J, ron 34:521–33
Yamin G, Munishkina LA, Fink AL. 2004. 64. Lee MK, Stirling W, Xu Y, Xu X, Qui
Biochemistry 43:4621–33 D, et al. 2002. Proc. Natl. Acad. Sci. USA
47. Uversky VN, Yamin G, Souillac PO, Go- 99:8968–73
ers J, Glaser CB, Fink AL. 2002. FEBS 65. Giasson BI, Lee VM. 2003. Cell 114:1–8
Lett. 517:239–44 66. Neumann M, Muller V, Gorner K, Kret-
48. Mishizen-Eberz AJ, Guttmann RP, Gias- zschmar HA, Haass C, Kahle PJ. 2004.
son BI, Day GA 3rd, Hodara R, et al. 2003. Acta Neuropathol. 107:489–96
J. Neurochem. 86:836–47 67. Feany MB, Bender WW. 2000. Nature
49. Kim SJ, Sung JY, Um JW, Hattori N, 404:394–98
50 COOKSON
68. El-Agnaf OM, Paleologou KE, Greer B, Gu WJ, Fon EA. 2002. J. Biol. Chem.
Abogrein AM, King JE, et al. 2004. 277:486–91
FASEB. J. 18:1315–17 86. Imai Y, Soda M, Hatakeyama S, Akagi T,
69. Emadi S, Liu R, Yuan B, Schulz P, Hashikawa T, et al. 2002. Mol. Cell 10:
McAllister C, et al. 2004. Biochemistry 55–67
43:2871–78 87. Chung KK, Thomas B, Li X, Pletnikova
70. Kitada T, Asakawa S, Hattori N, Matsum- O, Troncoso JC, et al. 2004. Science
ine H, Yamamura Y, et al. 1998. Nature 304:1328–31
392:605–8 88. Yao D, Gu Z, Nakamura T, Shi ZQ, Ma
71. West AB, Maidment NT. 2004. Hum. Y, et al. 2004. Proc. Natl. Acad. Sci. USA
Genet. 114:327–36 101:10810–14
72. Klein C, Pramstaller PP, Kis B, Page CC, 89. West AB, Maraganore D, Crook J,
Kann M, et al. 2000. Ann. Neurol. 48:65– Lesnick T, Lockhart PJ, et al. 2002. Hum.
71 Mol. Genet. 11:2787–92
73. Farrer M, Chan P, Chen R, Tan L, Lin- 90. Oliveira SA, Scott WK, Nance MA, Watts
coln S, et al. 2001. Ann. Neurol. 50:293– RL, Hubble JP, et al. 2003. Arch. Neurol.
300 60:975–80
74. Langston JW. 1996. Neurology 47:S153– 91. Yang Y, Nishimura I, Imai Y, Takahashi
60 R, Lu B. 2003. Neuron 37:911–24
75. Zhang Y, Gao J, Chung KK, Huang H, 92. Dong Z, Ferger B, Paterna JC, Vogel D,
Dawson VL, Dawson TM. 2000. Proc. Furler S, et al. 2003. Proc. Natl. Acad. Sci.
Natl. Acad. Sci. USA 97:13354–59 USA 100:12438–43
76. Shimura H, Hattori N, Kubo S, Mizuno 93. Leroy E, Boyer R, Auburger G, Leube B,
Y, Asakawa S, et al. 2000. Nat. Genet. Ulm G, et al. 1998. Nature 395:451–52
25:302–5 94. Liu Y, Fallon L, Lashuel HA, Liu Z, Lans-
77. Sakata E, Yamaguchi Y, Kurimoto E, bury PT Jr. 2002. Cell 111:209–18
Kikuchi J, Yokoyama S, et al. 2003. 95. Healy DG, Abou-Sleiman PM, Wood
EMBO Rep. 4:301–6 NW. 2004. Cell Tissue Res. 318:189–94
78. Marin I, Ferrus A. 2002. Mol. Biol. Evol. 96. Petrucelli L, O’Farrell C, Lockhart PJ,
19:2039–50 Baptista M, Kehoe K, et al. 2002. Neu-
79. Choi P, Snyder H, Petrucelli L, Theisler ron 36:1007–19
C, Chong M, et al. 2003. Mol. Brain Res. 97. Haywood AF, Staveley BE. 2004. BMC
117:179–89 Neuroscience 5:14
80. Staropoli JF, McDermott C, Martinat C, 98. Oluwatosin-Chigbu Y, Robbins A, Scott
Schulman B, Demireva E, Abeliovich A. CW, Arriza JL, Reid JD, Zysk JR.
2003. Neuron 37:735–49 2003. Biochem. Biophys. Res. Commun.
81. Corti O, Hampe C, Koutnikova H, Dar- 309:679–84
ios F, Jacquier S, et al. 2003. Hum. Mol. 99. Tofaris GK, Layfield R, Spillantini MG.
Genet. 12:1427–37 2001. FEBS Lett. 509:22–26
82. Imai Y, Soda M, Inoue H, Hattori N, 100. Lee HJ, Khoshaghideh F, Patel S, Lee SJ.
Mizuno Y, Takahashi R. 2001. Cell 2004. J. Neurosci. 24:1888–96
105:891–902 101. Paxinou E, Chen Q, Weisse M, Giasson
83. Huynh DP, Scoles DR, Nguyen D, Pulst BI, Norris EH, et al. 2001. J. Neurosci.
SM. 2003. Hum. Mol. Genet. 12:2587–97 21:8053–61
84. Chung KK, Zhang Y, Lim KL, Tanaka Y, 102. Shimura H, Schlossmacher MG, Hattori
Huang H, et al. 2001. Nat. Med. 7:1144– N, Frosch MP, Trockenbacher A, et al.
50 2001. Science 293:263–69
85. Fallon L, Moreau F, Croft BG, Labib N, 103. Goldberg MS, Fleming SM, Palacino JJ,
Cepeda C, Lam HA, et al. 2003. J. Biol. 120. Macedo MG, Anar B, Bronner IF, Can-
Chem. 278:43628–35 nella M, Squitieri F, et al. 2003. Hum. Mol.
104. Itier JM, Ibanez P, Mena MA, Abbas N, Genet. 12:2807–16
Cohen-Salmon C, et al. 2003. Hum. Mol. 121. Moore DJ, Zhang L, Dawson TM, Daw-
Genet. 12:2277–91 son VL. 2003. J. Neurochem. 87:1558–67
105. Von Coelln R, Thomas B, Savitt JM, Lim 122. Gorner K, Holtorf E, Odoy S, Nuscher B,
KL, Sasaki M, et al. 2004. Proc. Natl. Yamamoto A, et al. 2004. J. Biol. Chem.
Acad. Sci. USA 101:10744–49 279:6943–51
106. Greene JC, Whitworth AJ, Kuo I, An- 123. Takahashi-Niki K, Niki T, Taira T, Iguchi-
drews LA, Feany MB, Pallanck LJ. 2003. Ariga SM, Ariga H. 2004. Biochem. Bio-
Proc. Natl. Acad. Sci. USA 100:4078–83 phys. Res. Commun. 320:389–97
107. Palacino JJ, Sagi D, Goldberg MS, Krauss 124. Kinumi T, Kimata J, Taira T, Ariga H,
S, Motz C, et al. 2004. J. Biol. Chem. Niki E. 2004. Biochem. Biophys. Res.
279:18614–22 Commun. 317:722–28
108. Darios F, Corti O, Lucking CB, Hampe C, 125. Canet-Aviles RM, Wilson MA, Miller
Muriel MP, et al. 2003. Hum. Mol. Genet. DW, Ahmad R, McLendon C, et al. 2004.
12:517–26 Proc. Natl. Acad. Sci. USA 101:9103–
109. Bonifati V, Rizzu P, van Baren MJ, 8
Schaap O, Breedveld GJ, et al. 2003. Sci- 126. Yokota T, Sugawara K, Ito K, Takahashi
ence 299:256–59 R, Ariga H, Mizusawa H. 2003. Biochem.
110. Valente EM, Abou-Sleiman PM, Caputo Biophys. Res. Commun. 312:1342–48
V, Muqit MM, Harvey K, et al. 2004. Sci- 127. Taira T, Saito Y, Niki T, Iguchi-Ariga SM,
ence 304:1158–60 Takahashi K, Ariga H. 2004. EMBO Rep.
111. Dekker M, Bonifati V, van Swieten J, 5:213–18
Leenders N, Galjaard RJ, et al. 2003. Mov. 128. Rizzu P, Hinkle DA, Zhukareva V, Boni-
Disord. 18:751–57 fati V, Severijnen LA, et al. 2004. Ann.
112. Khan NL, Valente EM, Bentivoglio AR, Neurol. 55:113–18
Wood NW, Albanese A, et al. 2002. Ann. 129. Bandopadhyay R, Kingsbury AE, Cook-
Neurol. 52:849–53 son MR, Reid AR, Evans IM, et al. 2004.
113. Hod Y, Pentyala SN, Whyard TC, El- Brain 127:420–30
Maghrabi MR. 1999. J. Cell. Biochem. 130. Niki T, Takahashi-Niki K, Taira T, Iguchi-
72:435–44 Ariga SM, Ariga H. 2003. Mol. Cancer
114. Mitsumoto A, Nakagawa Y, Takeuchi A, Res. 1:247–61
Okawa K, Iwamatsu A, Takanezawa Y. 131. Unoki M, Nakamura Y. 2001. Oncogene
2001. Free Radic. Res. 35:301–10 20:4457–65
115. Bandyopadhyay S, Cookson MR. 2004. 132. Nakajima A, Kataoka K, Hong M, Sak-
BMC Evol. Biol. 4:6 aguchi M, Huh NH. 2003. Cancer Lett.
116. Olzmann JA, Brown K, Wilkinson KD, 201:195–201
Rees HD, Huai Q, et al. 2004. J. Biol. 133. Betarbet R, Sherer TB, MacKenzie G,
Chem. 279:8506–15 Garcia-Osuna M, Panov AV, Greenamyre
117. Lee SJ, Kim SJ, Kim IK, Ko J, Jeong CS, JT. 2000. Nat. Neurosci. 3:1301–6
et al. 2003. J. Biol. Chem. 278:44552–59 134. Sherer TB, Betarbet R, Testa CM, Seo BB,
118. Wilson MA, Collins JL, Hod Y, Ringe D, Richardson JR, et al. 2003. J. Neurosci.
Petsko GA. 2003. Proc. Natl. Acad. Sci. 23:10756–64
USA 100:9256–61 135. McNaught KSP, Perl DP, Brownell A-L,
119. Miller DW, Ahmad R, Hague S, Baptista Olanow CW. 2004. Ann. Neurol. 56:149–
MJ, Canet-Aviles R, et al. 2003. J. Biol. 62
Chem. 278:36588–95 136. Hoglinger GU, Carrard G, Michel PP,
52 COOKSON
Medja F, Lombes A, et al. 2003. J. Neu- 144. Hutton M. 2001. Neurology 56:S21–25
rochem. 86:1297–307 145. Lee VM, Giasson BI, Trojanowski JQ.
137. Sullivan PG, Dragicevic NB, Deng JH, 2004. Trends Neurosci. 27:129–34
Bai Y, Dimayuga E, et al. 2004. J. Biol. 146. Hardy J. 2003. J. Mol. Neurosci. 20:203–6
Chem. 279:20699–707 147. Gwinn-Hardy K. 2004. Arch. Neurol. 61:
138. Shen J, Cookson MR. 2004. Neuron 25–26
43:301–4 148. Gasser T, Muller-Myhsok B, Wszolek
139. Von Coelln R, Dawson VL, Dawson TM. ZK, Oehlmann R, Calne DB, et al. 1998.
2004. Cell Tissue Res. 318:175–84 Nat. Genet. 18:262–65
140. Hsu LJ, Sagara Y, Arroyo A, Rocken- 149. Funayama M, Hasegawa K, Kowa H,
stein E, Sisk A, et al. 2000. Am. J. Pathol. Saito M, Tsuji S, Obata F. 2002. Ann. Neu-
157:401–10 rol. 51:296–301
141. Tanaka Y, Engelender S, Igarashi S, Rao 150. Wszolek ZK, Pfeiffer RF, Tsuboi Y, Uitti
RK, Wanner T, et al. 2001. Hum. Mol. RJ, McComb RD, et al. 2004. Neurology
Genet. 10:919–26 62:1619–22

142. Snyder H, Mensah K, Theisler C, Lee J, 151. Paisan-Ruiz C, Jain S, Evans EW, Gilks
Matouschek A, Wolozin B. 2003. J. Biol. WP, Simon J, et al. 2004. Neuron 44:595–
Chem. 278:11753–59 600
143. Lindersson E, Beedholm R, Hojrup P, 152. Zimprich A, Biskup S, Leitner P, Lichtner
Moos T, Gai W, et al. 2004. J. Biol. Chem. P, Farrer M, et al. 2004. Neuron 44:601–
279:12924–34 7
NOTE ADDED IN PROOF

Since writing this chapter, the gene for the PARK8 locus has been cloned (151,
152). LRRK2 is a large complex kinase and is currently not characterized. The
protein has been termed Dardarin by Paisan-Ruiz et al. (151).
P1: JRX
April 29, 2005 5:26 Annual Reviews AR261-FM
Annual Review of Biochemistry

Volume 74, 2005
CONTENTS
FROM PROTEIN SYNTHESIS TO GENETIC INSERTION,
Paul Zamecnik 1
THE BIOCHEMISTRY OF PARKINSON’S DISEASE,

Mark R. Cookson 29
APPLICATIONS OF DNA MICROARRAYS IN BIOLOGY,

Roland B. Stoughton 53
ZONA PELLUCIDA DOMAIN PROTEINS, Luca Jovine, Costel C. Darie,
Eveline S. Litscher, and Paul M. Wassarman 83
PROLINE HYDROXYLATION AND GENE EXPRESSION,
William G. Kaelin Jr. 115
STRUCTURAL INSIGHTS INTO TRANSLATIONAL FIDELITY,
James M. Ogle and V. Ramakrishnan 129
ORIGINS OF THE GENETIC CODE: THE ESCAPED TRIPLET THEORY,
Michael Yarus, J. Gregory Caporaso, and Rob Knight 179
AN ABUNDANCE OF RNA REGULATORS, Gisela Storz, Shoshy Altuvia,
and Karen M. Wassarman 199
MEMBRANE-ASSOCIATED GUANYLATE KINASES REGULATE ADHESION
AND PLASTICITY AT CELL JUNCTIONS, Lars Funke, Srikanth Dakoji,
and David S. Bredt 219
STRUCTURE, FUNCTION, AND FORMATION OF BIOLOGICAL
IRON-SULFUR CLUSTERS, Deborah C. Johnson, Dennis R. Dean,
Archer D. Smith, and Michael K. Johnson 247
CELLULAR DNA REPLICASES: COMPONENTS AND DYNAMICS AT THE
REPLICATION FORK, Aaron Johnson and Mike O’Donnell 283
EUKARYOTIC TRANSLESION SYNTHESIS DNA POLYMERASES:
SPECIFICITY OF STRUCTURE AND FUNCTION, Satya Prakash,
Robert E. Johnson, and Louise Prakash 317
NOD-LRR PROTEINS: ROLE IN HOST-MICROBIAL INTERACTIONS AND
INFLAMMATORY DISEASE, Naohiro Inohara, Mathias Chamaillard,
Christine McDonald, and Gabriel Nuñez 355
v
P1: JRX
April 29, 2005 5:26 Annual Reviews AR261-FM
vi CONTENTS
REGULATION OF PROTEIN FUNCTION BY GLYCOSAMINOGLYCANS—AS

EXEMPLIFIED BY CHEMOKINES, T.M. Handel, Z. Johnson, S.E. Crown,
E.K. Lau, M. Sweeney, and A.E. Proudfoot 385
STRUCTURE AND FUNCTION OF FATTY ACID AMIDE HYDROLASE,
Michele K. McKinney and Benjamin F. Cravatt 411
NONTEMPLATE-DEPENDENT POLYMERIZATION PROCESSES:
POLYHYDROXYALKANOATE SYNTHASES AS A PARADIGM,
JoAnne Stubbe, Jiamin Tian, Aimin He, Anthony J. Sinskey,
Adam G. Lawrence, and Pinghua Liu 433
EUKARYOTIC CYTOSINE METHYLTRANSFERASES, Mary Grace Goll
and Timothy H. Bestor 481

MONITORING ENERGY BALANCE: METABOLITES OF FATTY ACID
SYNTHESIS AS HYPOTHALAMIC SENSORS, Paul Dowell, Zhiyuan Hu,

and M. Daniel Lane 515
STRUCTURE AND PHYSIOLOGIC FUNCTION OF THE LOW-DENSITY
LIPOPROTEIN RECEPTOR, Hyesung Jeon and Stephen C. Blacklow 535
COPPER-ZINC SUPEROXIDE DISMUTASE AND AMYOTROPHIC LATERAL
SCLEROSIS, Joan Selverstone Valentine, Peter A. Doucette,
and Soshanna Zittin Potter 563
THE STRUCTURE AND FUNCTION OF SMC AND KLEISIN COMPLEXES,
Kim Nasmyth and Christian H. Haering 595
ANTIBIOTICS TARGETING RIBOSOMES: RESISTANCE, SELECTIVITY,
SYNERGISM, AND CELLULAR REGULATION, Ada Yonath 649
DNA MISMATCH REPAIR, Thomas A. Kunkel and Dorothy A. Erie 681
GENE THERAPY: TWENTY-FIRST CENTURY MEDICINE, Inder M. Verma
and Matthew D. Weitzman 711
THE MAMMALIAN UNFOLDED PROTEIN RESPONSE, Martin Schröder
and Randal J. Kaufman 739
THE STRUCTURAL BIOLOGY OF TYPE II FATTY ACID
BIOSYNTHESIS, Stephen W. White, Jie Zheng, Yong-Mei Zhang,
and Charles O. Rock 791
STRUCTURAL STUDIES BY ELECTRON TOMOGRAPHY: FROM CELLS
TO MOLECULES, Vladan Lučić, Friedrich Förster,
and Wolfgang Baumeister 833
PROTEIN FAMILIES AND THEIR EVOLUTION—A STRUCTURAL
PERSPECTIVE, Christine A. Orengo and Janet M. Thornton 867

T B P ' D: HE Iochemistry of Arkinson S Isease

Uploaded by

T B P ' D: HE Iochemistry of Arkinson S Isease

Uploaded by

27 Apr 2005 3:58 AR AR261-BI74-02.

tex XMLPublishSM (2004/02/24) P1: JRX

Annu. Rev. Biochem. 2005. 74:29–52

THE BIOCHEMISTRY OF PARKINSON’S DISEASE∗

Key Words Lewy body, mitochondria, parkinsonism, proteasome, protein

TABLE 1 Monogenic forms of Parkinson’s disease

Locusa Gene Protein function Phenotypeb

PARK1 SNCAc Synaptic? PD/DLBD

PARK4 (AD) Lipid binding? Onset from age 30 to 60,

the genetic conditions discussed here. Secondly, PD is marked postmortem by the

α-SYNUCLEIN AND AGGREGATION

tends to self-associate, contributing to aggregation (23, 24). This region is not

unambiguously causal. The A30P family is relatively small, leading to concerns

α-synuclein (41), therefore increasing protein concentration, are associated with

tein prevents aggregation, removal is predicted to promote aggregation. Calpain I

Although α-synuclein is expressed in many tissues, symptoms are restricted

quite so variable is unclear.

PARKIN AND THE PROTEASOME

to kill neurons. So parkin disease is a phenocopy of PD, parkinsonism without

exposure to the toxin 1-methyl-4-phenyl-1,2,4,6,-tetrahydropyridinium (MPTP)

motif is important in recruiting substrates and the E2 enzyme (UbcH7 or UbcH8)

discussed in support of the proteasome hypothesis is the mutations in UCHL1

Parkin is also neuroprotective against stresses in which the direct relationship

DJ-1 AND PINK1: MITOCHONDRIAL CONNECTIONS?

drial damage. Although exposure to MPTP produces a Parkinsonian syndrome,

to dopaminergic neurons is not surprising. However, rotenone also inhibits mito-

sensitivity of catecholaminergic neurons to rotenone or MPTP in vitro (136). In

proteins were present on two-dimensional gels, mitochondrial proteins were specif-

In the absence of α-synuclein pathology in parkin, DJ-1, and PINK1 cases, we

Identifying monogenic forms of PD has led us to a causative agent, α-synuclein.

expression levels. There is currently an unclear relationship between α-synuclein

The Annual Review of Biochemistry is online at

39:889–909 18. Kruger R, Kuhn W, Muller T, Woitalla

RA, Jansen Steur EN, Braak E. 2003. Neu- 18:106–8

322:1089–102 54. Zhu M, Fink AL. 2003. J. Biol. Chem.

chemistry 41:4595–602 55. Cole NB, Murphy DD, Grider T, Rueter

Genet. 10:919–26 62:1619–22

NOTE ADDED IN PROOF

Annual Review of Biochemistry

THE BIOCHEMISTRY OF PARKINSON’S DISEASE,

APPLICATIONS OF DNA MICROARRAYS IN BIOLOGY,

REGULATION OF PROTEIN FUNCTION BY GLYCOSAMINOGLYCANS—AS

and Timothy H. Bestor 481

SYNTHESIS AS HYPOTHALAMIC SENSORS, Paul Dowell, Zhiyuan Hu,

You might also like