Parkinson’s disease (PD) is the most common neuro-
degenerative movement disorder, affecting ~1% of the
population above the age of 60. The classical form of
the disease is characterized clinically by rigidity, resting
tremor, bradykinesia and postural instability. Its patho-
logical hallmarks are the preferential loss of dopamin-
ergic neurons of the substantia nigra pars compacta and
formation of Lewy bodies — intracytoplasmic inclusion
bodies that are mainly composed of fibrillar α-synuclein1.
The clinical symptoms of PD arise by a threshold effect,
whereby denervation of the corpus striatum by dopamin-
ergic neuronal loss reduces dopamine levels to below 70%
of wild type2.
Mitochondrial dysfunction has long been implicated in
the pathogenesis of PD. Evidence first emerged following
the accidental exposure of drug abusers to 1-methyl-
4-phenyl-1,2,3,4-tetrahydropyridine (MPTP) — an
environ mental toxin that results in an acute and irrevers-
ible parkinsonian syndrome3. The active metabolite of
MPTP, the 1-methyl-4-phenylpyridinium ion (MPP+)
is an inhibitor of complex I of the mitochondrial electron
transport chain and a substrate for the dopamine transporter
(DAT). It therefore accumulates in dopaminergic neurons,
where it confers toxicity and neuronal death through
complex I inhibition4. This has many deleterious conse-
quences, including increased free radical production and
oxidative stress; and decreased ATP production, which
causes increased intracellular calcium concentration, exci-
totoxicity and nitric oxide related cellular damage. MPP+
also causes enhanced dopamine release, which leads to
further oxidative damage5. Crucially, a biochemical link
between MPTP toxicity and idiopathic PD was established
when several groups worldwide reported that complex I
was decreased in the substantia nigra, skeletal muscle
and platelets of patients with PD6–9. However, it remained
to be shown whether this systemic complex I deficiency,
as observed in patients with PD, was causally related to
dopaminergic cell loss in PD. Greenamyre’s laboratory
showed that rats administered the highly selective com-
plex I inhibitor rotenone developed a PD-like syndrome
characterized by neuronal degeneration and the forma-
tion of α-synuclein-rich inclusion bodies, however, the
degree of complex I inhibition was partial and not suf-
ficient to impair brain mitochondrial function; the effects
were therefore more likely to be conferred by increased
production of free radicals and oxidative stress10.
Mitochondrial DNA (mtDNA) encodes 13 proteins
that are all components of the electron transport chain,
and there have been several reports of mtDNA mutations
in rare maternally-inherited pedigrees of parkinsonism,
including the 12SrRNA gene in one family with par-
kinsonism, deafness and neuropathy11. More recently,
mutations in the nuclear encoded mitochondrial gene,
DNA polymerase γ (POLG), were reported in families
with parkinsonism associated with progressive external
ophthalmoplegia and multiple mitochondrial deletions
in affected members12. However, the clinicopathological
phenotype of such pedigrees is distinct from that of
idiopathic PD, and there is little convincing evidence to sup-
port a causal role of mtDNA variants in the pathogenesis
of idiopathic PD.
Despite the evidence of complex I inhibition and con-
comitant free radical production resulting in increased
oxidative stress in cellular, animal and human studies,
Department of Molecular
Neuroscience, Institute of
Neurology, Queen Square,
London WC1N 3BG, UK.
Correspondence to N.W.W.
reductase is an enzyme
complex consisting of more
than 40 polypeptides that
spans the inner mitochondrial
membrane. It oxidizes NADH,
resulting in the transfer of
electrons from NADH to
A collective term describing
the mitochondrial enzymes
(also known as complexes I–IV)
that are needed to generate
the electron and proton
‘gradient’ that is used by
complex V to generate ATP.
(DAT). A monoamine
transporter, the function of
which is the clearance of the
out of a synapse into a
presynaptic neuron or a
Expanding insights of mitochondrial
dysfunction in Parkinson’s disease
Patrick M. Abou-Sleiman, Miratul M. K. Muqit and Nicholas W. Wood
Abstract | The quest to disentangle the aetiopathogenesis of Parkinson’s disease has been
heavily influenced by the genes associated with the disease. The α-synuclein-centric theory
of protein aggregation with the adjunct of parkin-driven proteasome deregulation has, in
recent years, been complemented by the discovery and increasing knowledge of the
functions of DJ1, PINK1 and OMI/HTRA2, which are all associated with the mitochondria and
have been implicated in cellular protection against oxidative damage. We critically review
how these genes fit into and enhance our understanding of the role of mitochondrial
dysfunction in Parkinson’s disease, and consider how oxidative stress might be a potential
unifying factor in the aetiopathogenesis of the disease.
NATURE REVIEWS | NEUROSCIENCE
VOLUME 7 | MARCH 2006 | 207
Complex I deficiency
A reduction in the enzymatic
activity of complex I compared
with the remaining respiratory
chain complexes, as
determined by in vitro
it remained unclear whether this observed dysfunction
was a primary process in the pathogenesis of the disease
or a secondary process.
A major leap in our understanding of the aetio-
pathogenesis of the disease came when mutations were
identified in α-synuclein in 1997, followed by mutations
in parkin a year after that13,14. The demonstration that
α-synuclein is the main constituent of Lewy bodies in
the same year suggested a primary role for α-synuclein
aggregation, however, later studies revealed close inter-
play between α-synuclein aggregation and oxidative
stress in the pathogenesis of PD1,15. The identification of
mutations in DJ1 (Parkinson’s disease (autosomal reces-
sive, early onset) 7, a possible redox sensor) in 2003 and
phosphatase and tensin homologue (PTEN)-induced
kinase 1 (PINK1, a mitochondrial kinase) in 2004 pro-
vided strong evidence that mitochondrial dysfunction
and oxidative stress might have a primary role in the
pathogenesis of PD, although how mutations in these
genes cause neuronal degeneration is still unclear16,17.
So, although classically regarded as an archetypical
non-genetic disease due to the high proportion of
sporadic cases, hugely significant advances in our
understanding of PD have stemmed directly from the
study of these genes associated with a small proportion
of familial cases.
So far, five PD-associated genes have been con-
clusively identified (TABLE 1). A sixth gene, ubiquitin
carboxyl-terminal esterase L1 (UCHL1), has been
associated with the disease, but the genetic evidence for
its pathogenicity is weak as only a single mutation has
been identified in one family18. Recently, variation and
mutation of the HtrA serine peptidase 2 (HTRA2, also
known as OMI) gene has also been tentatively associated
In this review, we present an overview of the pub-
lished functional data on α-synuclein, parkin, DJ1,
PINK1, OMI/HTRA2 and leucine-rich repeat kinase 2
(LRRK2) in order to place them in the context of the
current theories on the aetiopathogenesis of the disease.
However, rather than attempting to review the entire
known biochemistry of PD, we focus on potential areas
of overlap between the different pathways. As a conse-
quence, we highlight the involvement of primarily oxida-
tive stress and, by association, the mitochondria in all of
the parkinsonian pathways described so far. We therefore
conclude that oxidative stress is a good candidate as the
unifying factor for several diverse but overlapping path-
ways to PD, and suggest that an improved understanding
of the mechanisms involved in the production of reac-
tive oxygen species (ROS) and the protection against
them, which is primarily mitochondrial, might result in
improved treatment of the disease.
α α-Synuclein, protein aggregation and PD
During the early 1990s, the relevance of rare Mendelian
forms of PD to idiopathic PD remained under-appreciated
and the prevailing view was that environmental factors
were the most important influence in disease aetiology.
In 1997, the discovery of an alanine-to-threonine mis-
sense mutation (A53T) in α-synuclein prompted a shift
in this belief13. This shift gained significant momentum
by the discovery that α-synuclein was a main component
of Lewy bodies in both familial and sporadic disease1,
which suggests that abnormalities of α-synuclein might
be crucial for the pathogenesis of both rare and common
forms of PD.
Although the function of α-synuclein is still unclear,
exciting recent work has revealed that α-synuclein can
rescue mice from neurodegeneration caused by deletion
of the synaptic co-chaperone, cysteine-string protein-α
(CSPα), and that it might act with CSPα to serve a pro-
tective role against injury at nerve terminals20. Given that
α-synuclein assumes a fibrillar β-pleated sheet structure
in Lewy bodies in PD and related α-synucleinopathies21,
the leading hypothesis for its pathogenicity is the forma-
tion of toxic aggregates. Under certain in vitro conditions,
wild-type or mutant α-synuclein can form protofibrillar
Table 1 | Parkinson’s disease-associated genes
Involved in synaptic
An E3 ligase
Age of onset: 30–60 years
Lewy bodies: ++
Age of onset: ~30 years
‡Lewy bodies: –
Age of onset: 30–50 years
Lewy bodies: ?
PARK2 ParkinAutosomal recessive
PARK6 Phosphatase and tensin
induced kinase 1 (PINK1)
early onset) 7 (DJ1)
kinase 2 (LRRK2)
Autosomal recessiveA mitochondrial
Autosomal recessiveInvolved in oxidative
Age of onset: 20–40 years
Lewy bodies: ?
PARK8 Autosomal dominant A protein kinaseAge of onset: 40–60 years
Lewy bodies: + variable
Age of onset: 44–70 years
Lewy bodies: ?
UnmappedHtrA serine peptidase
2 (HTRA2, also known
A serine protease
and/or involved in
*PARK1 and 4 share an entry because they have been shown to be caused by the same gene. ‡There has been one reported case of
a parkin-positive patient with Lewy bodies. ++ Fulminant Lewy body pathology. + Lewy bodies present.
208 | MARCH 2006 | VOLUME 7
Autosomal recessive PD
(ARPD). A familial form of PD
with an autosomal recessive
mode of inheritance.
RING finger proteins
Specialized zinc finger proteins
that bind two atoms of zinc.
Proteins containing RING
fingers are involved in
of many subunits that are
involved in the degradation
(oligomeric) or fibrillar conformations. However, contro-
versy remains about which species is the toxic culprit22,23.
The missense mutants A53T and A30P (alanine to
phenylalanine) in α-synuclein both promote protofibril
formation, but only the A53T mutation promotes fibril
formation, and, in fact, A30P inhibits conversion to
fibrils23. How might protofibrils be toxic? α-Synuclein
has been shown to bind synaptic vesicles, and protofibrils
can form pores that could lead to permeabilization of the
vesicle membranes, thereby releasing excess dopamine
into the cytosol24. Formation of protofibrils is enhanced
and stabilized by dopamine quinones derived from the
oxidation of dopamine, and this could account for the
selective toxicity of α-synuclein in the substantia nigra15.
Protofibril formation is also enhanced by missense and
triplication mutants of α-synuclein. Protofibrils were
recently reported in vivo, in brains from patients with
PD who had the α-synuclein triplications25.
Toxicity associated with increased α-synuclein
expression has important ramifications for the genetic
predisposition to idiopathic PD, as α-synuclein pro-
moter polymorphisms have been associated with the
disease26. There is also accumulating evidence that
phosphorylation of α-synuclein promotes the formation
of α-synuclein protofibrils and filaments in vitro, and
α-synuclein is extensively and selectively phosphorylated
at serine residue 129 in vivo in the brains of patients with
PD and related synucleinopathies27. Recently, Chen
and Feany found that phosphorylation of α-synuclein
at Ser129 significantly enhanced α-synuclein toxicity
in vivo in a Drosophila model of PD, and, interestingly,
showed that blockade of Ser129 phosphorylation was
associated with reduced toxicity and inclusion forma-
tion28. In Drosophila, the G-protein-coupled receptor
kinase 2 (GPRK2) phosphorylated α-synuclein at Ser129
(REF. 28), however, the identity of the kinase responsible
for α-synuclein phosphorylation in the human brain
The Drosophila model of α-synuclein also lends
weight to the fibril hypothesis, as expression of either
wild-type or mutant A30P or A53T α-synuclein resulted
in flies with progressive dopaminergic cell loss, motor
deficits, premature death and fibrillar α-synuclein-
positive inclusions similar to Lewy bodies29. Almost all
of the transgenic α-synuclein mouse models have failed
to recapitulate all the features of PD, most notably the
dopaminergic cell loss. Nevertheless, expression of wild-
type or A53T α-synuclein in mice generally results in
the neurological dysfunction and early death associated
with the formation of widespread α-synuclein-positive
inclusions that can be fibrillar30,31. However, transgenic
expression of the protofibrillogenic A30P α-synuclein
resulted in no phenotype in mice, in contrast to parallel
A53T mouse lines that were associated with widespread
α-synuclein inclusions and death32. The reason for dis-
cordance in phenotypes between the A30P mice and
Drosophila remains unknown.
It is still not clear whether the aggregates in PD are
toxic or protective. In α-synuclein-mutant flies, the
chaperone heat-shock protein 70 (HSP70) suppressed
the PD phenotype and reduced dopaminergic neuronal
degeneration. However, there was no difference in the
frequency or distribution of α-synuclein inclusions in
surviving flies33. Moreover, the work of Chen and Feany
also suggests that, at least in flies, the formation of
α-synuclein inclusions might be a protective response
to α-synuclein toxicity in vivo28. The ubiquitin ligase
parkin has also been shown to reduce α-synuclein tox-
icity in vitro and in vivo, but this is associated with an
enhancement of α-synuclein aggregation, including an
increase in phosphorylated inclusions in rats transduced
with lentiviruses expressing the A30P α-synuclein muta-
tion34,35. Although these latter studies do not rule out the
possibility that protofibrils are toxic in these systems and
that the protective response is mediated by their conver-
sion to non-toxic fibrillar forms, they do at least raise the
possibility that additional cellular defects might occur
secondary to altered α-synuclein expression.
α-Synuclein is localized predominantly in synaptic
terminals and in the cytosol of the cell body36. Although
there is no evidence for mitochondrial localization,
dysfunction of α-synuclein has been shown to indirectly
but significantly impact on neuronal mitochondrial
function. There is accumulating evidence for a close
relationship between α-synuclein and oxidative damage;
overexpression of mutant α-synuclein sensitizes neurons
to oxidative stress and damage by dopamine and mito-
chondrial toxins such as MPP+ and 6-hydroxydopamine,
resulting in increased protein carbonylation and lipid
peroxidation in vitro and in vivo37,38. Interestingly,
α-synuclein-knockout mice were initially found to have
marked resistance to MPTP, and a recent study also
showed resistance to other mitochondrial toxins, includ-
ing malonate and 3-nitropropionic acid39,40. The mecha-
nism of this resistance seems to be due to α-synuclein
deficiency resulting in a reduction of oxidative stress
— α-synuclein has previously been shown to modulate
the release of dopamine from synaptic vesicles into the
cytosol, where it auto-oxidizes, leading to increased pro-
duction of free radicals and oxidative stress41. Moreover,
the generation of ROS and oxidative stress is likely to
exacerbate the toxic effect of the α-synuclein muta-
tions, including aggregation in an amplification loop
(see below). Therefore, it can be envisaged that cellular
pathways in dopamine neurons can conspire to cause
catastrophic neuronal demise in the presence of defects
in either the α-synuclein or oxidative stress pathways.
Homozygous mutations in the parkin gene were discov-
ered in families with autosomal recessive PD (ARPD)14. In
contrast to α-synuclein, parkin mutations are common,
and account for almost half of all cases of ARPD, espe-
cially those with onset before 21 years of age42,43. The
465 amino acid protein contains two RING (really inter-
esting new gene) fingers separated by an in-between
RING (IBR) domain at the carboxyl (C) terminus,
which, like other RING finger proteins, functions as an
E3 ubiquitin ligase44. The amino (N) terminus bears a
ubiquitin-like domain that binds to the RPN10 subunit
of the 26S proteasome45. In the cell, E3 ubiquitin ligases
are one component of the ubiquitin–proteasome system
NATURE REVIEWS | NEUROSCIENCE
VOLUME 7 | MARCH 2006 | 209
(UPS), a main cellular pathway that promotes removal of
damaged or misfolded proteins46. E3 ligases catalyse the
addition of ubiqutin molecules to lysine residues of dam-
aged target proteins, and the presence of a polyubiquitin
chain provides a signal for its removal and degradation
by the proteolytic complex, the 26S proteasome46.
A striking variety of homozygous and compound
hetero zygous mutations have been reported, includ-
ing gene rearrangements and missense mutations42,43.
Despite the fact that there have been few postmortem
studies in parkin-related ARPD, it is becoming clearer
that the nature of the parkin mutation is crucial for inter-
preting the pathology, and that mutations that abolish
parkin activity seem to be associated with a lack of Lewy
bodies. However, in mutations that reduce but do not
abolish parkin activity, Lewy bodies can occur47.
The lack of Lewy bodies in ARPD patients led some
to postulate that parkin-related ARPD was a distinct
clinical syndrome from sporadic PD. However, recent
insights in parkin suggest that the two types of PD might
have shared aetiological pathways. Complex I is selec-
tively reduced in peripheral leukocytes of patients with
parkin-related ARPD48, and parkin is generally found
in Lewy bodies of patients with sporadic and familial
PD49,50. A Drosophila model has revealed a role for parkin
in maintaining mitochondrial function and preventing
oxidative stress — pathways heavily implicated in spo-
radic PD. Parkin null mutants had severe mitochondrial
pathology associated with reduced lifespan, apop-
tosis, flight muscle degeneration and male sterility51.
Microarray analysis in these flies revealed upregula-
tion of genes involved in oxidative stress and electron
transport, including a homologue of the mammalian
peripheral benzodiazepine receptor. A genomic screen
for modifiers of lifespan in the parkin null flies found
the strongest modifier to be loss-of-function mutations
of glutathione S-transferase (GSTS1)52. Reanalysis of
the same flies revealed progressive degeneration of a
select cluster of dopaminergic neurons and evidence of
increased oxidative damage with increased protein car-
bonyls compared with controls53. Furthermore, neuro-
degeneration was enhanced in GSTS1 null mutants,
whereas GSTS1 overexpression significantly rescued
the parkin phenotype53.
Mammalian models also support a role for parkin in
maintaining mitochondrial function. Deletion of exon 3
of parkin in mice results in nigrostriatal dysfunction and
reduced expression of several proteins involved in mito-
chondrial function and oxidative stress, including subu-
nits of complexes I and IV. The mice also had decreased
mitochondrial respiratory capacity and showed evidence
of increased oxidative damage54. Intriguingly, parkin
deficiency in these mice did not cause dopaminergic
degeneration, which is also observed in exon 2 and
other exon 3 deletion models55–57. It will therefore be
interesting to determine whether neurodegeneration in
these models requires an environmental insult such as an
oxidative stressor. A similar mechanism might operate in
parkin heterozygotes, as they seem to be at increased risk
of PD and have nigrostriatal dysfunction, as visualized
by positron emission tomography (PET)58.
The mechanism by which parkin might regulate
mitochondrial function is unclear. Parkin might be
directly involved in maintaining mitochondrial integ-
rity: it has been localized to the outer mitochondrial
membrane (OMM), where it has a crucial role in pre-
venting mitochondrial swelling and rupture secondary
to ceramide toxicity59. Furthermore, the Drosophila
homologue of the peripheral benzodiazepine receptor
(PBR) was upregulated in parkin mutant flies, and
PBR is a component of the mitochondrial permeability
transition pore (mPTP), where it is important in mPTP
opening, thereby regulating oxidative damage secondary
to mitochondrial dysfunction and OMM rupture60. It
is not known whether PBR is a substrate for parkin.
Ubiquitylation also mediates the insertion of mitochon-
drial proteins into the OMM and parkin might also have
a role in this61. However, further studies are required to
confirm mitochondrial localization of parkin. So far,
most studies have localized parkin to other subcellular
organelles, most notably the endoplasmic reticulum,
where it has been shown to have a neuroprotective role
against endoplasmic reticulum stress62.
It is equally likely that parkin could indirectly maintain
mitochondrial function. As free radical-induced oxidative
damage is a normal consequence of the electron transport
chain, parkin might have a role in removing oxidatively
damaged proteins and mutant parkin might lead to the
accumulation of such proteins, which might lead to further
oxidative stress and apoptosis. This mechanism would be
most compatible with the known function of E3 ligases in
the UPS. Moreover, the related E3 ubiquitin ligase, HOIL1,
has been shown to degrade its oxidized substrate, iron
responsive element-binding protein 2 (IRP2)63. Parkin
expression is upregulated after exposure to the complex I
inhibitor MPP+ in neuronal cells64. However, parkin func-
tion itself can be modified by oxidative stress: exposure to
nitric oxide generates free radicals that modify cysteine
residues in RING1, which results in significant alteration
of parkin’s E3 ligase activity, including inactivation65,66.
Furthermore, nitrosylated forms of parkin are detectable
in both the brains of patients with PD and the brains from
MPTP and rotenone treated animals, which suggests
that inactivation of parkin might be a crucial step in the
pathogenesis of sporadic PD65,66. Moreover, dopamine has
recently been shown to covalently modify and function-
ally inactivate parkin E3 ligase activity by increasing parkin
insolubility in vitro and in the brains of patients with PD67.
Another recent study also found that a wide array of
oxidative stressors, including rotenone, MPP+, 6-hydroxy-
dopamine, paraquat, nitric oxide and iron, as well as
dopamine, all altered parkin solubility and caused parkin
aggregation, thereby suggesting a mechanism for par-
kin dysfunction in the pathogenesis of idiopathic PD68.
Interestingly, modulation of parkin’s E3 ligase activity
has also been shown for the interactor, BCL2-associ-
ated athano gene 5 (BAG5). BAG5 is a member of the
BAG family of proteins and has been shown to interact
with parkin, as well as the chaperone HSP70. Moreover,
BAG5 seems to promote neurodegeneration by inhibiting
parkin’s E3 ligase activity and promoting sequestration of
parkin into aggregates69.
210 | MARCH 2006 | VOLUME 7
Proteins that share sequence
homology to the bacterial ThiJ
domain. Functions include
protein chaperones, catalases,
proteases and ThiJ kinases.
(Isoelectric point). The pH of a
solution at which a dissolved
charged molecule has no
electric charge and will
therefore not move in an
Ever since parkin was found to be an E3 ligase, it
has been proposed that mutations of parkin or parkin
dysfunction might lead to the toxic accumulation of
its substrate. E3 ligases confer specificity in the UPS,
usually by targeting one protein70. Surprisingly, many
disparate substrates for parkin have been discovered,
with strong replicated evidence for the septin CDCREL1
and PAELR71. Moreover, overexpression of CDCREL1
and PAELR in vivo mediates dopaminergic neuro-
degeneration72,73, and both also accumulate in the brains
of patients with parkin-related ARPD74,75. However,
neither of these substrates accumulates in Drosophila
or mammalian parkin-knockout models, and it is not
known whether any of these substrates are oxidatively
modified following oxidative stress or can be appro-
priately degraded by parkin. By contrast, the parkin
substrate, aminoacyl-tRNA synthetase cofactor p38, is
upregulated in the midbrain of parkin null mice as well
as in the brains of patients with ARPD and idiopathic
PD76. Moreover, adenovirus-mediated overexpression
of p38 in the substantia nigra of mice induced loss of
α-Synuclein might unify the oxidative stress and
substrate accumulation hypotheses for the media-
tion of parkin dysfunction. Parkin can protect against
α-synuclein-induced neurotoxicity in vitro and
in vivo73,77. However, there is no replicated data that
parkin interacts directly with α-synuclein. Parkin has
been shown to interact with a glycosylated form of
α-synuclein (known as sp22)78, however, it is not clear
whether α-synuclein toxicity is mediated by sp22. The
interaction of parkin and α-synuclein might be direct
if parkin can degrade oxidatively modified forms of
α-synuclein. Alternatively, the interaction might be
indirect and α-synuclein has been shown to inhibit
the proteasome (although it is not clear how), and it
has been observed that parkin can rescue neurons
from α-synuclein induced proteasomal dysfunction77.
What is becoming clear from numerous studies is that
loss of parkin function might impact on many cellular
pathways, rendering dopaminergic neurons sensitive
to neurotoxicity and death. Although UPS dysfunc-
tion and mitochondrial dysfunction might be the main
pathogenetic pathways, the discovery of a wide range of
substrates implicates additional pathways. Future studies
should illuminate these pathways, and it will be of sig-
nificant interest to see whether and how these additional
pathways impact and/or converge on the two currently
accepted pathways of UPS and oxidative stress.
Mutations of DJ1 were identified in two consanguineous
European families with early-onset ARPD16, and several
pathogenic mutations have since been identified, including
homozygous and heterozygous missense mutations and
exonic deletions. In terms of the overall mutation burden
in PD, DJ1 mutations seem to be quite rare, accounting
for only an estimated 1–2% of early-onset cases79.
The precise cellular distribution and subcellular
localization of the protein has been hotly contested
since it was first associated with PD. It is certain that
DJ1 is widely expressed in both the brain and peripheral
tissues. Beyond that, reports are conflicting80–83 — the
only study carried out on the endogenous protein so far
has detected DJ1 expression in both neurons and glia,
although expression in the latter was weak. Subcellular
distribution of the endogenous protein is primarily cyto-
plasmic with a smaller pool of mitochondrial-associated
Structurally, DJ1 is a member of the ThiJ/PfpI/DJ1
superfamily and is highly conserved across species. It
has limited homology to several prokaryotic proteins,
including heat shock protein chaperones and ThiJ/PfpI
proteases84. DJ1 has been ascribed various functions85,86,
but perhaps the most relevant in terms of the pathogene-
sis of PD is its potential role in oxidative stress response,
either as a redox sensor or antioxidant protein87,88. At
present, there are two lines of evidence to support this
theory, the first of which is that in mammalian cells
exposed to an oxidative stressor, such as paraquat
or H2O2, DJ1 undergoes an acidic shift in pI-value
by modifying its cysteine residues, which quench
ROS and protect cells against stress-induced death.
Its Drosophila homologues DJ1α and DJ1β are also
oxidized at their cysteine residues after exposure to the
same stresses both in vivo and in culture89,90. Further
evidence has been obtained from various model systems.
Mammalian cell cultures, and mouse and Drosophila
knockouts all indicate that the ablation of functional
DJ1 either by small interfering RNA (siRNA) or gene
deletion sensitizes cells to oxidative stress. These cells
can be rescued by overexpression of wild-type but not
mutant (L166P) DJ1 (REF. 91). Therefore, loss of DJ1
increases levels of intracellular ROS and increases sus-
ceptibility to dopaminergic neuron degeneration in vivo
following exposure to exogenous sources of oxidative
stress92. However, it is unlikely that DJ1 exerts its protec-
tive function through simple antioxidation. Its ability to
quench ROS is modest82,89, and several lines of evidence
indicate that it is more likely to be implicated in the
regulation of apoptosis. It is plausible that, through
modification of its cysteine residues on exposure to
H2O2, DJ1 is acting as a sensor of cellular ROS levels. DJ1
functions in the phosphatidylinositol 3-kinase (PI3K)
survival pathway as a negative regulator of PTEN. The
PTEN tumour suppressor regulates the PI3K pathway
by dephosphorylating phosphatidylinositol-3,4,5-
trisphosphate, which is required for the activation of
a survival kinase, protein kinase B (PKB, also known
as Akt)93. In cultured cells, tissue from patients with
cancer and Drosophila DJ1α RNAi knockdowns, DJ1
expression correlates with the phosphorylation of PKB.
In keeping with a possible modulatory effect, it has been
shown that increased PI3K/Akt signalling capacity in
Drosophila DJ1α RNAi knockdowns reduced cell death,
whereas reduced PI3K/Akt expression enhanced the
phenotype89. How DJ1 interacts with the PI3K–PTEN–
Akt pathway remains to be determined. One possibility,
given that PTEN function can be modulated by expo-
sure to H2O2, is that DJ1 exerts a redox effect on PTEN
through either its protease activity or redox-sensitive
NATURE REVIEWS | NEUROSCIENCE
VOLUME 7 | MARCH 2006 | 211
(∆Ψm). A chemiosmotic
gradient of protons across the
membrane. The energy this
creates is used for ATP
synthesis by the electron
DJ1 is further implicated in both apoptosis and cellular
response to oxidative stress through its binding part-
ners in the brain, death-associated protein 6 (DAXX)82,
54 kDa nuclear RNA-binding protein (p54NRB), pyri-
midine tract-binding protein-associated splicing factor
(PSF)94 and topoisomerase I binding protein (TOPORS,
also known as p53BP3)95. DJ1 acts as a potent inhibitor
of the DAXX–ASK1 (apoptosis signal-regulating kinase
1) pathway by sequestering DAXX to the nucleus,
away from its cytoplasmic effector, and also inhibits
PSF-induced apoptosis by cooperating with p54NRB to
activate PSF-silenced transcription82.
DJ1 mutations disrupt protein activity by either
destabilizing the protein or affecting its subcellular
localization. Wild-type DJ1 exists as a homodimer
in vitro96–99. However, the L166P mutation impairs the
ability of DJ1 to self-interact, which results in a highly
unstable protein, which is degraded by the 20S/26S
proteasome99–101. The reduced protein levels result
in loss of function, as demonstrated by the impaired
regulation of the DAXX–ASK1 pathway seen in L116P
mutants82. Incorrect subcellular localization might also
result in loss of DJ1 function — the L166P, M26I and
D149A mutations all show reduced nuclear localization
in favour of mitochondrial localization16,94. The reduced
access to nuclear proteins, such as p54NRB and PSF
might increase PSF-induced apoptosis as a result of the
increased mitochondrial localization of the mutants82. As
the mitochondrial function of DJ1 remains to be deter-
mined, it is not clear whether the mutagenecity associ-
ated with increased mitochondrial localization is due to
a mitochondrial gain of function or to a loss of access
to binding partners in different cellular compartments.
In addition to cellular response to oxidative damage
and apoptosis, DJ1 also functions as a redox sensitive
molecular chaperone that is capable of preventing the
aggregation of α-synuclein and the neurofilament subu-
nit NFL102. However, as the majority of these studies have
been carried out on cell lines overexpressing DJ1, which
are prone to artefact, these putative roles remain to be
confirmed in vivo.
PINK1 and OMI/HTRA2
Mutations in PINK1 were initially identified in three
large consanguineous families with ARPD — one
Spanish and two Italian17. The gene had previously been
cloned by two groups of researchers who independently
analysed differential expression profiles of cancer cell
lines103,104. PINK1 was shown to be transcriptionally
activated by PTEN, but its expression was not sufficient
to suppress the growth of cancer cells103. It is noteworthy
that DJ1 has recently been identified as a suppressor of
PINK1 encodes a ubiquitously expressed 581 amino
acid protein, which consists of an N-terminal mitochon-
drial targeting motif, a highly conserved serine/threo-
nine kinase domain and a C-terminal autoregulatory
Mutation reports from around the world indicate
that the frequency of PINK1 mutations lies somewhere
between those of DJ1 and parkin, with the notable
exception of the L347P PINK1 mutation, for which
a carrier frequency of 8% has been reported in the
Philippino population106–109. The reported mutations do
not show any obvious clustering within the gene, and
most are distributed throughout the kinase domain with
a subset located in the N-terminal region between the
mitochondrial targeting motif and the kinase domain
(approximately amino acid residues 30–150). At present,
it is not clear how the mutations located outside the
kinase domain affect enzyme function. One possibility
is that they could disrupt mitochondrial localization or
processing. Most nuclear encoded mitochondrial pro-
teins contain a cleavable N-terminal ‘pre-sequence’ of
between 20 and 60 amino acid residues that directs the
protein towards its target after translation on cytoplas-
mic ribosomes (for a recent review on mitochondrial
import, see REF. 110). The mechanism of PINK1 mito-
chondrial import remains unknown, and the peptidase
cleavage sites unmapped. These mutations might either
disrupt interaction with the translocase complex of the
inner membrane (TIM complex) or be cleaved by the
mitochondrial processing peptidase (MPP). Western
blot analysis of cells overexpressing PINK1 suggests that
the processed protein is ~10 kDa smaller than the pre-
protein, and the 10 kDa fragment translates to a cleav-
age product of ~100 amino acids108, potentially placing
mutations such as C92F at or around the cleavage site.
As the kinase domain is the only functional domain
in PINK1, and the site of most of the mutations, disrup-
tion of the kinase activity is the most probable disease
mechanism. Kinases are defined by 12 highly conserved
subdomains that fold to form a common catalytic
core structure that is required for phosphorylation111.
Alignments of the PINK1 protein against other known
kinases indicate that it has all the necessary subdomains
to form an active serine/threonine kinase. At present,
confirmation of the bioinformatic prediction is limited
to studies that use in vitro autophosphorylation assays;
these studies demonstrate that PINK1 is active104,105,108.
However, these assays are too crude to distinguish any
significant functional effects of the mutations on kinase
activity. Some functional data has been obtained from
indirect assays, which show that loss of PINK1 func-
tion adversely affects mitochondrial function and cell
viability under stress. Mitochondrial membrane potential
(∆ψm) (FIG. 1) and levels of cell death were measured in
a neuroblastoma cell line overexpressing G309D PINK1
after exposure to an exogenous source of cellular stress,
MG-132, a proteasome inhibitor. Cells overexpressing
G309D PINK1 had significantly reduced ∆ψm com-
pared with the wild type and increased levels of cell
death following exposure to stress, but not under basal
conditions. Moreover, cells overexpressing wild-type
PINK1 had higher ∆ψm and lower levels of cell death
than cells transfected with vector alone17. Consistent
with these results, overexpression of wild-type PINK1
was subsequently shown to reduce the release of cyto-
chrome c from mitochondria under basal conditions and
staurosporine-induced stress. The reduced cytochrome c
levels also result in a decrease in the production of the
pro-apoptotically cleaved caspase 3, caspase 7, caspase 9
212 | MARCH 2006 | VOLUME 7
and poly (ADP-ribose) polymerase (PARP). Significantly,
two PINK1 mutations, E240K and L489P, were shown to
abrogate the protective effect112. Finally, the L347P muta-
tion decreases the half-life of PINK1 from ~2 h to ~0.25 h,
which supports the idea that loss of PINK1 function
might contribute to PD108.
Studies on PINK1 mutations strongly support a
direct involvement of mitochondria in the pathogenesis
of idiopathic PD. The relationship between PINK1 and
the ∆ψm remains to be determined. It might centre
around the electron transport chain or, more likely, the
mitochondrial pro-apoptotic pathways. Regulation of
the electron transport chain might involve phosphory-
lation of at least two of its subunits, complexes I and V
(REFS 113,114), in order to maintain the electrochemical
gradient that generates the ∆ψm. Depolarization of the
∆ψm is associated with opening of the mitochondrial
permeability transition pore (mPTP), which occurs
in necrotic and apoptotic cell death (see hypothetical
model in FIG. 2). Both the voltage-dependent anion chan-
nel (VDAC) and the adenine nucleotide translocator
(ANT) undergo phosphorylation in response to stress
and might represent potential areas for PINK1 involve-
ment. Recent studies of isolated mitochondria from
selective ANT-knockout mice revealed preservation of
mPTP activation and cytochrome c release, suggesting
that ANT is a non-essential structural component of
the mPTP. However, ANT might still have a role in the
regulation of mPTP opening, and any post-translational
modifications of ANT, such as phosphorylation, might
be important for this regulation115.
One example of a procaspase that is released from the
intermembrane space by the opening of the mPTP is the
OMI/HTRA2 protein. It has previously been implicated
in neurodegeneration, and recently associated with
predisposition to PD19. OMI/HTRA2, a PDZ domain-
containing serine protease, also contains an N-terminal
mitochondrial-targeting motif and a reaper-like motif.
OMI/HTRA2 is thought to localize to the mitochon-
drial intermembrane space, where it is released into
the cytosol during apoptosis to relieve the inhibition of
caspases by binding to inhibitor of apoptosis proteins
(IAPs). OMI/HTRA2 is also able to induce cell death
through its proteolytic activity.
OMI/HTRA2 was initially shown to interact with
proteins associated with Alzheimer’s disease, such
as presenillin 1 and amyloid-β116,117. Recently, OMI/
HTRA2-knockout mice have been shown to display
parkinsonian phenotypes, including rigidity and
tremor118, which, together with its localization in the
PARK3 (Parkinson’s disease 3) locus made it a candidate
for mutation screening in patients with PD. Mutations
in OMI/HTRA2 were not found in the PARK3 fami-
lies. However, a mutation at Gly399 was found in four
Figure 1 | Schematic representation of the mitochondrial electron transport system. Mitochondria are the main
providers of cellular energy, which is generated through the flow of electrons down the electron transport system (ETS).
The ETS is located on the inner mitochondrial membrane (IMM) and consists of 4 membrane spanning enzyme complexes.
These comprise complex I (NADH–ubiquinone reductase), which oxidizes NADH, complex II (succinate–ubiquinone
oxidoreductase), which oxidizes FADH2, complex III (ubiquinol cytochrome c oxidoreductase) and complex IV
(cytochrome c oxidase). The ETS also contains two hydrophobic electron carriers, coenzyme Q10 (CoQ) and cytochrome c
(Cyt c), both of which are encoded by nuclear DNA. The ETS transfers electrons through a series of oxidation–reduction
reactions, culminating in the reduction of oxygen to produce water. The oxidation–reduction reactions are coupled to the
transfer of protons (H+) across the IMM, and this proton efflux creates a proton electrochemical gradient known as the
protomotive force140. The protomotive force consists mainly of an electrical component called the mitochondrial
membrane potential (∆Ψm) and a transmembrane pH gradient. The ∆Ψm, which is maintained at about –150 to –180 mV
negative to the cytosol, is central to mitochondrial function and provides the force that drives the influx of protons and
Ca2+ into the mitochondria as well as determining the generation of O2
through a proton channel of the F1F0-ATP synthase (also known as complex V). The re-entry of protons depolarizes the
∆Ψm and induces the phosphorylation of matrix ADP to generate ATP. Therefore, the transfer of electrons, generation of
the ∆Ψm and ATP synthesis are all closely coupled. Pi, phosphate group.
–. Protons traverse the mitochondrial membrane
NATURE REVIEWS | NEUROSCIENCE
VOLUME 7 | MARCH 2006 | 213
Figure 2 | Hypothetical schematic of the mitochondrial permeability transition
pore. The mitochondrial permeability transition pore (mPTP) is a conductance pore that
spans the inner (IMM) and outer (OMM) mitochondrial membranes. It consists of
membranous elements such as the voltage-dependent anion channel (VDAC) on the
OMM, the adenine nucleotide translocator (ANT) on the IMM, and cyclophilin D (Cyclo
D) in the matrix. Other proteins, such as the peripheral benzodiazepine receptor (PBR),
hexokinase (HK) and creatine kinase (CK), might also be associated with the mPTP. It is
not clear whether the mPTP has a role in normal mitochondrial physiology. However,
under a combination of pathophysiological conditions, such as high Ca2+ concentration,
increased oxidative stress, low ATP, and mitochondrial depolarization, the complex forms
an open pore between the inner and outer membranes, allowing free diffusion of solutes
across the membranes. The opening of the mPTP ultimately results in mitochondrial
swelling, mitochondrial Ca2+ efflux and the release of apoptogenic proteins, such as
cytochrome c and procaspases, from the intermembrane space141. ANT-knockout studies
suggest that it might not be an essential structural component but might have a role in
the regulation of mPTP opening115. ∆Ψm, mitochondrial membrane potential; BAX,
BCL2-associated X protein; ROS, reactive oxygen species.
• ROS and oxidative stress
• Increased Ca2+ concentration
• Misfolded mitochondrial proteins
Short (~40) amino acid motifs
that form beta-propeller
structures, which are thought
to serve as rigid scaffolds for
protein interactions. WD40
repeat-containing proteins can
therefore coordinate the
assembly of multi-protein
patients with sporadic PD, and a polymorphism at
Ala141 was found at higher frequencies in patients with
PD19. Both mutations are predicted to affect regulation of
the proteolytic activity of OMI/HTRA2, thereby modu-
lating cell death. At the cellular level, the mutations have
been shown to increase susceptibility to stress, as shown
by decreased ∆ψm after exposure to staurosporine. It
is tempting to speculate that PINK1 and OMI/HTRA2
share a common pathway in the mitochondrial response
to cellular stress and modulation of apoptosis, perhaps
with OMI/HTRA2 as a downstream target for PINK1,
which initiates apoptosis when levels of oxidative stress
become too high. However, whether such interaction
exists in PD remains to be determined.
Leucine-rich repeat kinase 2
Mutations of LRRK2 have recently been shown to cause
autosomal dominant PD previously linked to the PARK8
locus119,120. LRRK2 mutations are estimated to account
for 5–6% of cases with a positive family history, and
a significant minority of apparently sporadic cases (up
to 1.6%)121. LRRK2 encodes a complex multi-domain
protein that consists of N-terminal leucine-rich repeats,
a GTPase ROC/COR domain, a mitogen-activated pro-
tein kinase kinase kinase (MAPKKK) and C-terminal
WD40 repeats120. At present, little is known about LRRK2
function, however, some interesting preliminary data has
begun to emerge from in vitro overexpression systems.
LRRK2 encodes a kinase and is capable of autophosphory-
lation. It might be associated with the outer mitochon-
drial membrane (OMM)122,123 and can bind parkin124.
Significantly, three PD-associated mutations, two in
the kinase domain (G2019S and I2020T) and one in the
ROC/COR GTPase domain (R1441C) increase LRRK2
autophosphorylation, hinting at a dominant gain-of-
function mechanism122,123. Indeed, overexpression of
R1441C, Y1699C or G2019S LRRK2 was sufficient to
induce neuronal degeneration in mouse primary cortical
neurons124. Although these data still require validation
in vivo, the delineation of the LRRK2 signalling pathway
holds great promise for furthering our understanding of
the aetiology of the disease.
Pathways of cell death in PD
The discovery of PD genes has raised important ques-
tions, including what the principal pathways are and how
the genes fit into them. Important clinicopathological
and epidemiological observations had already marked
out the main pathways before the recent advances in
genetics. Pathological studies show that the loss of
dopaminergic neurons is ubiquitous in all forms of
human parkinsonism. Considerable evidence shows
that the mitochondria-dependent apoptotic pathway
is predominantly activated125, although this has been
controversial, as there have been many studies that have
also failed to confirm the presence of apoptosis in the
substantia nigra of patients with PD (for a review, see
REF. 126). The identification of Lewy bodies in the brains
of people with PD suggested that aggregation was also
important in the pathogenesis of dopaminergic cell death.
Interestingly, recent pathological insights from families
with LRRK2 mutations have revealed distinct path-
ology with neurofibrillary tangles, which suggests that
mutant LRRK2 might have a role in tau phosphory lation
As the apoptotic pathway is substantially conserved
in all cells, primary dysfunction in this cascade alone
was not thought likely to account for the selective
cell loss in patients with PD. Moreover, despite many
studies, it was still unclear whether Lewy bodies were
toxic or protective for dopamine neurons. Therefore,
additional pathways, perhaps upstream of these two
pathological endpoints, were thought more important.
Epidemiological studies unambiguously demonstrated
that ageing is the greatest risk factor in the development
of sporadic PD127. The molecular mechanisms underly-
ing ageing are complex, but elegant studies in Drosophila
and Caenorhabditis elegans, in which the expression of
superoxide dismutase (SOD) was manipulated, as well
as the positive effects of caloric restriction in mammals,
suggest that oxidative stress is a main cellular pathway
in the ageing process128,129. Postmortem studies of the
brains of patients with sporadic PD reveal increased
214 | MARCH 2006 | VOLUME 7
oxidative damage of lipids (peroxidation) and proteins
(carbonylation)127, and the finding of decreased complex
I activity in the brains of people with PD suggested that
mitochondrial dysfunction might exacerbate the oxida-
tive stress in PD. The remarkably exclusive degeneration
of dopaminergic neurons following exposure to mito-
chondrial neurotoxins (such as MPTP and rotenone)
suggests that dysfunction of the mitochondrial pathway
could confer the selective vulnerability of dopaminer-
gic neurons in PD, although the mechanism remains
obscure3,10. One possible mechanism is the pro-oxidant
environment in dopaminergic neurons due to the pres-
ence of the oxidant neurotransmitter dopamine that
is normally packaged in vesicles. Exposure of neurons
to increased cytosolic dopamine has many deleterious
effects, including increases in oxidative stress as well
as promotion and stabilization of potentially toxic
Ubiquitin–proteasome system and mitochondria.
Although α-synuclein and parkin mutations confirm
that protein misfolding and UPS dysfunction are part
of a significant upstream pathway en route to dopamin-
ergic degeneration, the discovery of PINK1, DJ1 and
OMI/HTRA2 mutations confirmed that mitochondrial
dysfunction is another main upstream pathway to par-
kinsonism. An intriguing question is whether all of the
known genes converge to form a common pathogenetic
pathway. In view of their differential subcellular localiza-
tions, direct interactions seem unlikely. However, they
might interact via overlapping pathways, and there is
significant evidence for a close relationship between the
UPS and mitochondrial function. For example, protea-
somal stress can result in increased sensitivity of neurons
to MPTP and, conversely, complex I defects can result in
decreased proteasome activity130,131.
The mechanism through which mitochondrial and
proteasomal impairment lead to dopamine cell loss
is becoming clearer, and the generation of oxidative
stress might be common to both and might be allied to
apoptotic cell death in PD (FIG. 3). Evidence of increased
oxidative damage has been demonstrated following
mitochondrial or proteasomal impairment in vivo132,133.
Furthermore, the pathways might be interdependent,
which would result in either feedback or feedforward
loop mechanisms between the UPS and mitochondria
such that dysfunction of one pathway would have
inevitable deleterious consequences for the other. As
the UPS requires ATP, mitochondrial dysfunction and
ATP depletion are likely to lead to UPS dysfunction.
Treatment of human neuroblastoma cell lines with the
complex I inhibitor rotenone was associated with ~20%
ATP depletion, an increase in ROS and oxidized pro-
teins, and a marked reduction in proteasome activity134.
Inactivation of the proteasome is therefore likely to create
a feedforward amplification loop, with further damage
from failure to clear the excess oxidized protein species
leading to further ROS generation. Detergent-insoluble
aggregates of α-synuclein accumulate in cells following
mitochondrial inhibition with rotenone or oligomycin,
and disappear after subsequent washout of inhibitors
paralleling recovery of mitochondrial metabolism32.
Perhaps the most compelling evidence for this interplay
occurring in PD is the demonstration that chronic roten-
one exposure in rats results in the formation of Lewy
body-like aggregates in addition to parkinsonism10.
Evidence is also accumulating for a converse mecha-
nism, whereby UPS dysfunction can result in secondary
mitochondrial dysfunction and damage. Treatment of
cultured primary rat cortical neurons with proteasome
inhibitors was sufficient to induce the redistribution of
cytochrome c into the cytosol, and this was associated
with depolarization of the ∆ψm, which led to activa-
tion of caspase 3 and resulted in apoptotic cell death135.
Of particular relevance to PD, when the expression of
A30P mutant α-synuclein was induced in PC12 cells
treated with a proteasome inhibitor, cells underwent
mitochondrial apoptosis and ∆ψm depolarization. Both
∆ψm depolarization and apoptosis were blocked with
cyclosporin A, which suggests that proteasome inhibi-
tion resulted in ∆ψm depolarization associated with
opening of the mPTP136 (FIG. 2).
The mechanism by which proteasome inhibition
leads to mitochondrial injury and apoptosis remains
unclear. Several pro-apoptopic proteins, such as p53 and
BCL2 (B-cell leukaemia/lymohoma 2) family members,
including BAX, BID and SMAC, are normally degraded
by the UPS137. Therefore, the UPS has a crucial role in
homeostasis by preventing the accumulation of these
potentially toxic molecules. Proteasome inhibition leads
to the accumulation of these proteins, which result in
mPTP opening, ∆ψm depolarization and apoptosis137.
In addition, p53 has been shown to transcriptionally
activate the expression of several target genes, including
BAX, NOXA (adult T cell leukaemia-derived PMA-
responsive) and PUMA (BCL2 binding component 3),
which can, in turn, cause mitochondrial dysfunction138.
Further studies are required to determine the significance
of p53 alterations in proteasome inhibition, particularly
in PD and in vivo models of the disease.
The interplay between the UPS and mitochondria
might also be mediated in part by the known PD
genes. Parkin seems to be crucial for maintenance
of mitochondrial function, as parkin-knockout mice
develop mitochondrial deficits and oxidative damage,
which does not lead to neurodegeneration. Although
mitochondrial proteins are not targeted by the UPS,
parkin might regulate proteins on the OMM that are
accessible to the UPS. Parkin in neuronal PC12 cell
lines localized to the OMM, where it protected cells
from the damaging effects of ceramides by delaying
mitochondrial swelling and cytochrome c release59.
We have also shown in neuroblastoma cells that pro-
teasome inhibition induces ∆ψm depolarization and
apoptosis, and that wild-type PINK1 protein, but not
the missense mutant G309D, protected cells from this
stress17. Similarly, RNAi knockdown of DJ1 also sensi-
tized neurons to proteasomal inhibiton91.
DJ1 has recently been identified as a regulator
of p53 transcriptional activity by binding TOPORS/
p53BP3, and might, therefore, represent a molecular
bridge between the UPS and the mitochondrial stress
NATURE REVIEWS | NEUROSCIENCE
VOLUME 7 | MARCH 2006 | 215
Proteasome dysfunction or ↓ATP
The ubiquitin pathway The mitochondrial pathway
response95. Parkin might also be an important bridging
protein between these two systems, which are important
for ubiquitinating and degrading oxidized mitochondrial
proteins, thereby ameliorating oxidative stress. Parkin is
itself sensitive to oxidative stress, and is inactivated by
nitric oxide-mediated nitrosylation, which could lead
to a simultaneous increase in UPS and mitochondrial
Figure 3 | Pathways to parkinsonism. The discovery of Mendelian inherited genes has enhanced our understanding of
the pathways that mediate neurodegeneration in Parkinson’s disease. One main pathway of cell toxicity arises through
α-synuclein, protein misfolding and aggregation. These proteins are ubiquitinated and initially degraded by the ubiquitin–
proteasome system (UPS), in which parkin has a crucial role. However, there is accumulation and failure of clearance by the
UPS over time, which leads to the formation of fibrillar aggregates and Lewy bodies. α-Synuclein protofibrils can also be
directly toxic, leading to the formation of oxidative stress that can further impair the UPS by reducing ATP levels, inhibiting
the proteasome, and by oxidatively modifying parkin. This leads to accelerated accumulation of aggregates.
Phosphorylation of α-synuclein-containing or tau-containing aggregates might have a role in their pathogenicity and
formation, but it is not known whether leucine-rich repeat kinase 2 (LRRK2) mediates this. Another main pathway is the
mitochondrial pathway. There is accumulating evidence for impaired oxidative phosphorylation and decreased complex I
activity in Parkinson’s disease, which leads to reactive oxygen species (ROS) formation and oxidative stress. In parallel,
there is loss of the mitochondrial membrane potential. This leads to opening of the mitochondrial permeability transition
pore (mPTP), release of cytochrome c from the intermembrane space to the cytosol, and activation of mitochondrial-
dependent apoptosis resulting in caspase activation and cell death. There is evidence that recessive-inherited genes, such
as phosphatase and tensin homologue (PTEN)-induced kinase 1 (PINK1), Parkinson’s disease (autosomal recessive, early
onset) 7 (DJ1) and HtrA serine peptidase 2 (HTRA2, also known as OMI), might all have neuroprotective effects against the
development of mitochondrial dysfunction, although the exact site of their action remains unknown. Parkin has also been
shown to inhibit the release of cytochrome c following ceramide-induced stress, and is itself modified by the interacting
protein BCL2-associated athanogene 5 (BAG5). Dysfunction of both pathways leads to oxidative stress, which causes
further dysfunction of these pathways by feedback and feedforward mechanisms, ultimately leading to irreversible
cellular damage and death. I–IV, mitochondial electron transport chain complexes I–IV; α-syn(PO4)n, phospho-α-synuclein;
A30P, alanine to proline substitution at α-synuclein amino acid residue 30; A53T, alanine to threonine subsitution at
α-synuclein residue 53; E1, ubiquitin activating enzyme; E2, ubiquitin conjugating enzyme; E46K, glutamic acid to lysine
substitution at α-synuclein residue 46; NO, nitric oxide; 3n/4n, 3 or 4 copies of α-synuclein; Tau(POi)n,Tau (POi)n, phospho-
Tau; UCHL1, ubiquitin carboxyl-terminal esterase L1.
216 | MARCH 2006 | VOLUME 7
Despite this, parkin’s exact role in mitochondria is
unknown and, in particular, no mitochondrial sub-
strates have been identified. Interestingly, LRRK2 has
been shown to associate with the OMM, and the same
research group have also reported that LRRK2 interacts
with parkin, although this appears to occur in the cyto-
plasm122,124. Moreover, parkin appeared to increase the
amount of ubquitinated LRRK2-containing aggregates
in cells124. It is still not clear whether parkin might inter-
act with the recently discovered PINK1, DJ1 or OMI/
HTRA2 (REFS 16,17,19). Parkin has been localized to the
OMM, and overexpressed DJ1 is translocated from the
cytosol to the OMM after oxidative stress, which might
mediate DJ1’s neuroprotective function88. Furthermore,
parkin has been found to associate with mutant DJ1,
which might enhance its stability in vitro. The detergent-
insoluble levels of DJ1 seem to be differentially regu-
lated by parkin in vivo, although parkin does not seem
to ubiquitinate DJ1 directly139. Recently, endo genous
DJ1 was detected in the matrix and intermembranous
space of mitochondria, where it is thought it might have
a role in the maintenance of mitochondrial function83.
Deleting DJ1 expression in mammalian systems in vivo
does not lead to dopaminergic degeneration, which
suggests that DJ1 is not crucial for mitochondrial integ-
rity. However, it might protect neurons from exposure
to oxidative stress, as DJ1-knockout mice were more
sensitive to MPTP neurotoxicity93. This suggests that
DJ1 acts as a free radical scavenger, which prevents the
accumulation of free radicals derived from the mito-
chondrial electron transport system. This would be
consistent with its localization in the intermembranous
space93. It will be interesting to investigate the effect of
ageing on DJ1-knockout mice.
Delineating the mitochondrial pathways involved in PD
will require a lot of further work, and identifying the sub-
strates of the known genes will provide an ideal starting
point. Although mitochondria might be key to neuronal
integrity and survival, defects in disparate areas of mito-
chondrial signalling are likely to have deleterious effects
that converge due to feedback and feedforward mecha-
nisms. It is likely that the main dysfunctional pathway
will centre around the response to oxidative stress and
its numerous associated facets. These could include
the deregulation of the electron transport chain, which
can cause a dual effect of reduced ATP production and
increased oxidative stress. The reduction of ATP might
lead to UPS dysfunction, which, in combination with
oxidative stress, is likely to lead to further mitochon-
drial dysfunction, leading to pathological activation of
pathways involved in aggregation and apoptosis (FIG. 3).
The initial defects might be mild, but could increase
through the actions of feedback mechanisms and the
effects of ageing. Furthermore, oxidative stress could
affect mitochondrial integrity, which is maintained by
the mitochondrial membrane potential and the mPTP.
This might be the second main site of regulation after
the electron transport chain, and an important molecu-
lar link between mitochondria and the UPS. The death
of dopaminergic neurons, which results in PD, is almost
certainly due to a combination of exogenous stressors
(which probably include dopamine itself) and a genetic
predisposition (which renders the cells less capable of
dealing with the stress). The common endpoint to
these cellular insults might converge on the initiation
of the mitochondrial apoptotic pathways and release
of pro-apoptotic proteins from the intermembranous
space, including cytochrome c and SMAC, that induce
caspase-dependent cell death; and apoptosis inducing
factor (AIF) and endonuclease G that translocate to the
nucleus and induce caspase-independent nuclear DNA
fragmentation126. We now have an understanding of the
basic molecular framework to the aetiopathogenesis of
the disease, further research into the mitochondrial
pathways should lead to more clarity for designing
effective drugs to treat PD.
Spillantini, M. G. et al. α-Synuclein in Lewy bodies.
Nature 388, 839–840 (1997).
The first study to show α α-synuclein in Lewy bodies,
linking synuclein biology and dysfunction to
Bernheimer, H., Birkmayer, W., Hornykiewicz, O.,
Jellinger, K. & Seitelberger, F. Brain dopamine and the
syndromes of Parkinson and Huntington. Clinical,
morphological and neurochemical correlations.
J. Neurol. Sci. 20, 415–455 (1973).
Langston, J. W., Ballard, P., Tetrud, J. W. & Irwin, I.
Chronic parkinsonism in humans due to a product of
meperidine-analog synthesis. Science 219, 979–980
The first major study to show that MPTP can
induce parkinsonism in humans.
Nicklas, W. J., Vyas, I. & Heikkila, R. E. Inhibition of
NADH-linked oxidation in brain mitochondria by
1-methyl-4-phenyl-pyridine, a metabolite of the
tetrahydropyridine. Life Sci. 36, 2503–2508 (1985).
Fiskum, G., Starkov, A., Polster, B. M. & Chinopoulos, C.
Mitochondrial mechanisms of neural cell death and
neuroprotective interventions in Parkinson’s disease.
Ann. NY Acad. Sci. 991, 111–119 (2003).
Mizuno, Y. et al. Deficiencies in complex I subunits of
the respiratory chain in Parkinson’s disease.
Biochem. Biophys. Res. Commun. 163, 1450–1455
Parker, W. D. Jr, Boyson, S. J. & Parks, J. K.
Abnormalities of the electron transport chain in
idiopathic Parkinson’s disease. Ann. Neurol. 26,
Schapira, A. H. et al. Mitochondrial complex I
deficiency in Parkinson’s disease. Lancet 1, 1269
The first study to show complex I deficiency in the
brains of patients with PD, linking the MPTP model
to sporadic PD.
Orth, M. & Schapira, A. H. Mitochondrial involvement
in Parkinson’s disease. Neurochem. Int. 40, 533–541
10. Betarbet, R. et al. Chronic systemic pesticide exposure
reproduces features of Parkinson’s disease. Nature
Neurosci. 3, 1301–1306 (2000).
Reports perhaps the best animal model of PD —
parkinsonism in rats caused by the complex I
11. Thyagarajan, D. et al. A novel mitochondrial 12SrRNA
point mutation in parkinsonism, deafness, and
neuropathy. Ann. Neurol. 48, 730–736 (2000).
12. Luoma, P. et al. Parkinsonism, premature menopause,
and mitochondrial DNA polymerase γ mutations:
clinical and molecular genetic study. Lancet 364,
13. Polymeropoulos, M. H. et al. Mutation in the
α-synuclein gene identified in families with Parkinson’s
disease. Science 276, 2045–2047 (1997).
The first report of a single gene mutation resulting
14. Kitada, T. et al. Mutations in the parkin gene cause
autosomal recessive juvenile parkinsonism. Nature
392, 605–608 (1998).
15. Conway, K. A., Rochet, J. C., Bieganski, R. M. &
Lansbury, P. T. Jr. Kinetic stabilization of the
α-synuclein protofibril by a dopamine-α-synuclein
adduct. Science 294, 1346–1349 (2001).
An important paper showing that dopamine
modulates α α-synuclein aggregation, which suggests
a mechanism for the regional selectivity of PD.
16. Bonifati, V. et al. Mutations in the DJ-1 gene
associated with autosomal recessive early-onset
parkinsonism. Science 299, 256–259 (2003).
17. Valente, E. M. et al. Hereditary early-onset Parkinson’s
disease caused by mutations in PINK1. Science 304,
The first report of a PD gene unequivocally
localized to mitochondria.
18. Leroy, E. et al. The ubiquitin pathway in Parkinson’s
disease. Nature 395, 451–452 (1998).
19. Strauss, K. M. et al. Loss of function mutations in the
gene encoding Omi/HtrA2 in Parkinson’s disease.
Hum. Mol. Genet. 14, 2099–2111 (2005).
20. Chandra, S., Gallardo, G., Fernandez-Chacon, R.,
Schluter, O. M. & Sudhof, T. C. α-Synuclein cooperates
with CSPα in preventing neurodegeneration. Cell 123,
NATURE REVIEWS | NEUROSCIENCE
VOLUME 7 | MARCH 2006 | 217
A groundbreaking study that suggests a role for
α α-synuclein in the prevention of neurodegeneration.
21. Spillantini, M. G., Crowther, R. A., Jakes, R.,
Hasegawa, M. & Goedert, M. α-Synuclein in
filamentous inclusions of Lewy bodies from Parkinson’s
disease and dementia with Lewy bodies. Proc. Natl
Acad. Sci. USA 95, 6469–6473 (1998).
22. Conway, K. A., Harper, J. D. & Lansbury, P. T.
Accelerated in vitro fibril formation by a mutant
α-synuclein linked to early-onset Parkinson disease.
Nature Med. 4, 1318–1320 (1998).
23. Caughey, B. & Lansbury, P. T. Protofibrils, pores, fibrils,
and neurodegeneration: separating the responsible
protein aggregates from the innocent bystanders.
Annu. Rev. Neurosci. 26, 267–298 (2003).
24. Lashuel, H. A., Hartley, D., Petre, B. M., Walz, T. &
Lansbury, P. T. Jr. Neurodegenerative disease: amyloid
pores from pathogenic mutations. Nature 418, 291
25. Miller, D. W. et al. α-Synuclein in blood and brain from
familial Parkinson disease with SNCA locus
triplication. Neurology 62, 1835–1838 (2004).
26. Farrer, M. et al. α-Synuclein gene haplotypes are
associated with Parkinson’s disease. Hum. Mol. Genet.
10, 1847–1851 (2001).
27. Fujiwara, H. et al. α-Synuclein is phosphorylated in
synucleinopathy lesions. Nature Cell Biol. 4, 160–164
An important study showing that phosphorylation
of synuclein at Ser129 may mediate pathogenicity
of synuclein in vivo.
28. Chen, L. & Feany, M. B. α-Synuclein phosphorylation
controls neurotoxicity and inclusion formation in a
Drosophila model of Parkinson disease. Nature
Neurosci. 8, 657–663 (2005).
29. Feany, M. B. & Bender, W. W. A Drosophila model of
Parkinson’s disease. Nature 404, 394–398 (2000).
30. Masliah, E. et al. Dopaminergic loss and inclusion
body formation in α-synuclein mice: implications for
neurodegenerative disorders. Science 287,
31. Giasson, B. I. et al. Neuronal α-synucleinopathy with
severe movement disorder in mice expressing A53T
human α-synuclein. Neuron 34, 521–533 (2002).
32. Lee, M. K. et al. Human α-synuclein-harboring familial
Parkinson’s disease-linked Ala-53 → Thr mutation
causes neurodegenerative disease with α-synuclein
aggregation in transgenic mice. Proc. Natl Acad. Sci.
USA 99, 8968–8973 (2002).
33. Auluck, P. K., Chan, H. Y., Trojanowski, J. Q., Lee, V. M.
& Bonini, N. M. Chaperone suppression of α-synuclein
toxicity in a Drosophila model for Parkinson’s disease.
Science 295, 865–868 (2002).
34. Lo Bianco, C. et al. Lentiviral vector delivery of parkin
prevents dopaminergic degeneration in an α-synuclein
rat model of Parkinson’s disease. Proc. Natl Acad. Sci.
USA 101, 17510–17515 (2004).
35. Oluwatosin-Chigbu, Y. et al. Parkin suppresses wild-
type α-synuclein-induced toxicity in SHSY-5Y cells.
Biochem. Biophys. Res. Commun. 309, 679–684
36. Goedert, M. α-Synuclein and neurodegenerative
diseases. Nature Rev. Neurosci. 2, 492–501 (2001).
37. Orth, M. & Tabrizi, S. J. Models of Parkinson’s disease.
Mov. Disord. 18, 729–737 (2003).
38. Tabrizi, S. J. et al. Expression of mutant α-synuclein
causes increased susceptibility to dopamine toxicity.
Hum. Mol. Genet. 9, 2683–2689 (2000).
39. Dauer, W. et al. Resistance of α-synuclein null mice to
the parkinsonian neurotoxin MPTP. Proc. Natl Acad.
Sci. USA 99, 14524–14529 (2002).
40. Klivenyi, P. et al. Mice lacking α-synuclein are resistant
to mitochondrial toxins. Neurobiol. Dis. 17 Nov 2005
41. Sulzer, D. et al. Neuromelanin biosynthesis is driven
by excess cytosolic catecholamines not accumulated
by synaptic vesicles. Proc. Natl Acad. Sci. USA 97,
42. Abbas, N. et al. A wide variety of mutations in the
parkin gene are responsible for autosomal recessive
parkinsonism in Europe. French Parkinson’s Disease
Genetics Study Group and the European Consortium
on Genetic Susceptibility in Parkinson’s Disease. Hum.
Mol. Genet. 8, 567–574 (1999).
43. Lucking, C. B. et al. Association between early-onset
Parkinson’s disease and mutations in the parkin gene.
French Parkinson’s Disease Genetics Study Group.
N. Engl. J. Med. 342, 1560–1567 (2000).
44. Shimura, H. et al. Familial Parkinson disease gene
product, parkin, is a ubiquitin–protein ligase. Nature
Genet. 25, 302–305 (2000).
45. Sakata, E. et al. Parkin binds the Rpn10 subunit of
26S proteasomes through its ubiquitin–like domain.
EMBO Rep. 4, 301–306 (2003).
46. Ciechanover, A. The ubiquitin–proteasome pathway:
on protein death and cell life. EMBO J. 17,
47. Farrer, M. et al. Lewy bodies and parkinsonism in
families with parkin mutations. Ann. Neurol. 50,
48. Muftuoglu, M. et al. Mitochondrial complex I and IV
activities in leukocytes from patients with parkin
mutations. Mov. Disord. 19, 544–548 (2004).
49. Shimura, H. et al. Immunohistochemical and
subcellular localization of Parkin protein: absence of
protein in autosomal recessive juvenile parkinsonism
patients. Ann. Neurol. 45, 668–672 (1999).
50. Schlossmacher, M. G. et al. Parkin localizes to the
Lewy bodies of Parkinson disease and dementia with
Lewy bodies. Am. J. Pathol. 160, 1655–1667 (2002).
51. Greene, J. C. et al. Mitochondrial pathology and
apoptotic muscle degeneration in Drosophila parkin
mutants. Proc. Natl Acad. Sci. USA 100, 4078–4083
The first study to link parkin to mitochondrial
52. Greene, J. C., Whitworth, A. J., Andrews, L. A.,
Parker, T. J. & Pallanck, L. J. Genetic and genomic
studies of Drosophila parkin mutants implicate
oxidative stress and innate immune responses in
pathogenesis. Hum. Mol. Genet. 14, 799–811
53. Whitworth, A. J. et al. Increased glutathione
S-transferase activity rescues dopaminergic neuron
loss in a Drosophila model of Parkinson’s disease.
Proc. Natl Acad. Sci. USA 102, 8024–8029 (2005).
54. Palacino, J. J. et al. Mitochondrial dysfunction and
oxidative damage in parkin-deficient mice. J. Biol.
Chem. 279, 18614–18622 (2004).
55. Goldberg, M. S. et al. Parkin-deficient mice exhibit
nigrostriatal deficits but not loss of dopaminergic
neurons. J. Biol. Chem. 278, 43628–43635 (2003).
56. Itier, J. M. et al. Parkin gene inactivation alters
behaviour and dopamine neurotransmission in the
mouse. Hum. Mol. Genet. 12, 2277–2291 (2003).
57. Perez, F. A. & Palmiter, R. D. Parkin-deficient mice are
not a robust model of parkinsonism. Proc. Natl Acad.
Sci. USA 102, 2174–2179 (2005).
58. Khan, N. L. et al. Progression of nigrostriatal
dysfunction in a parkin kindred: an [18F]dopa PET and
clinical study. Brain 125, 2248–2256 (2002).
59. Darios, F. et al. Parkin prevents mitochondrial swelling
and cytochrome c release in mitochondria-dependent
cell death. Hum. Mol. Genet. 12, 517–526 (2003).
60. Casellas, P., Galiegue, S. & Basile, A. S. Peripheral
benzodiazepine receptors and mitochondrial function.
Neurochem. Int. 40, 475–486 (2002).
61. Zhaung, Z. P. & McCauley, R. Ubiquitin is involved in
the in vitro insertion of monoamine oxidase B into
mitochondrial outer membranes. J. Biol. Chem. 264,
62. Takahashi, R., Imai, Y., Hattori, N. & Mizuno, Y. Parkin
and endoplasmic reticulum stress. Ann. NY Acad. Sci.
991, 101–106 (2003).
63. Yamanaka, K. et al. Identification of the ubiquitin-
protein ligase that recognizes oxidized IRP2. Nature
Cell Biol. 5, 336–340 (2003).
64. Hyun, D. H., Lee, M., Halliwell, B. & Jenner, P. Effect of
overexpression of wild-type or mutant parkin on the
cellular response induced by toxic insults. J. Neurosci.
Res. 82, 232–244 (2005).
65. Chung, K. K. et al. S-nitrosylation of parkin regulates
ubiquitination and compromises parkin’s protective
function. Science 304, 1328–1331 (2004).
66. Yao, D. et al. Nitrosative stress linked to sporadic
Parkinson’s disease: S-nitrosylation of parkin regulates
its E3 ubiquitin ligase activity. Proc. Natl Acad. Sci.
USA 101, 10810–10814 (2004).
67. LaVoie, M. J., Ostaszewski, B. L., Weihofen, A.,
Schlossmacher, M. G. & Selkoe, D. J. Dopamine
covalently modifies and functionally inactivates parkin.
Nature Med. 11, 1214–1221 (2005).
An important study showing that parkin activity is
inactivated by dopamine, which suggests a
mechanism for parkin dysfunction in sporadic PD.
68. Wang, C. et al. Stress-induced alterations in parkin
solubility promote parkin aggregation and
compromise parkin’s protective function. Hum. Mol.
Genet. 14, 3885–3897 (2005).
69. Kalia, S. K. et al. BAG5 inhibits parkin and enhances
dopaminergic neuron degeneration. Neuron 44,
70. Robinson, P. A. & Ardley, H. C. Ubiquitin–protein
ligases. J. Cell Sci. 117, 5191–5194 (2004).
71. von Coelln, R., Dawson, V. L. & Dawson, T. M. Parkin-
associated Parkinson’s disease. Cell Tissue Res. 318,
72. Dong, Z. et al. Dopamine-dependent
neurodegeneration in rats induced by viral vector-
mediated overexpression of the parkin target protein,
CDCrel-1. Proc. Natl Acad. Sci. USA 100,
73. Yang, Y., Nishimura, I., Imai, Y., Takahashi, R. & Lu, B.
Parkin suppresses dopaminergic neuron-selective
neurotoxicity induced by Pael-R in Drosophila. Neuron
37, 911–924 (2003).
74. Imai, Y. et al. An unfolded putative transmembrane
polypeptide, which can lead to endoplasmic reticulum
stress, is a substrate of Parkin. Cell 105, 891–902
75. Choi, P. et al. SEPT5_v2 is a parkin-binding protein.
Brain Res. Mol. Brain Res. 117, 179–189 (2003).
76. Ko, H. S. et al. Accumulation of the authentic parkin
substrate aminoacyl-tRNA synthetase cofactor,
p38/JTV-1, leads to catecholaminergic cell death.
J. Neurosci. 25, 7968–7978 (2005).
77. Petrucelli, L. et al. Parkin protects against the toxicity
associated with mutant α-synuclein: proteasome
dysfunction selectively affects catecholaminergic
neurons. Neuron 36, 1007–1019 (2002).
78. Shimura, H. et al. Ubiquitination of a new form of
α-synuclein by parkin from human brain: implications
for Parkinson’s disease. Science 293, 263–269
79. Abou-Sleiman, P. M., Healy, D. G., Quinn, N.,
Lees, A. J. & Wood, N. W. The role of pathogenic DJ-1
mutations in Parkinson’s disease. Ann. Neurol. 54,
80. Rizzu, P. et al. DJ-1 colocalizes with tau inclusions: a
link between parkinsonism and dementia. Ann.
Neurol. 55, 113–118 (2004).
81. Bandopadhyay, R. et al. The expression of DJ-1
(PARK7) in normal human CNS and idiopathic
Parkinson’s disease. Brain 127, 420–430 (2004).
82. Junn, E. et al. Interaction of DJ-1 with Daxx inhibits
apoptosis signal-regulating kinase 1 activity and cell
death. Proc. Natl Acad. Sci. USA 102, 9691–9696
83. Zhang, L. et al. Mitochondrial localization of the
Parkinson’s disease related protein DJ-1: implications
for pathogenesis. Hum. Mol. Genet. 14, 2063–2073
84. Cookson, M. R. The biochemistry of Parkinson’s
disease. Annu. Rev. Biochem. 74, 29–52 (2005).
85. Nagakubo, D. et al. DJ-1, a novel oncogene which
transforms mouse NIH3T3 cells in cooperation with
ras. Biochem. Biophys. Res. Commun. 231, 509–513
86. Hod, Y., Pentyala, S. N., Whyard, T. C. &
El-Maghrabi, M. R. Identification and
characterization of a novel protein that regulates
RNA–protein interaction. J. Cell Biochem. 72,
87. Mitsumoto, A. & Nakagawa, Y. DJ-1 is an indicator for
endogenous reactive oxygen species elicited by
endotoxin. Free Radic. Res. 35, 885–893 (2001).
88. Canet-Aviles, R. M. et al. The Parkinson’s disease
protein DJ-1 is neuroprotective due to cysteine-sulfinic
acid-driven mitochondrial localization. Proc. Natl
Acad. Sci. USA 101, 9103–9108 (2004).
89. Yang, Y. et al. Inactivation of Drosophila DJ-1 leads to
impairments of oxidative stress response and
phosphatidylinositol 3-kinase/Akt signaling. Proc. Natl
Acad. Sci. USA 102, 13670–13675 (2005).
90. Meulener, M. et al. Drosophila DJ-1 mutants are
selectively sensitive to environmental toxins
associated with Parkinson’s disease. Curr. Biol. 15,
91. Yokota, T. et al. Down regulation of DJ-1 enhances cell
death by oxidative stress, ER stress, and proteasome
inhibition. Biochem. Biophys. Res. Commun. 312,
92. Kim, R. H. et al. Hypersensitivity of DJ-1-deficient
mice to 1-methyl-4-phenyl-1,2,3,6-
tetrahydropyrindine (MPTP) and oxidative stress.
Proc. Natl Acad. Sci. USA 102, 5215–5220 (2005).
93. Kim, R. H. et al. DJ-1, a novel regulator of the
tumor suppressor PTEN. Cancer Cell 7, 263–273
94. Xu, J. et al. The Parkinson’s disease-associated DJ-1
protein is a transcriptional co-activator that protects
against neuronal apoptosis. Hum. Mol. Genet. 14,
218 | MARCH 2006 | VOLUME 7
95. Shinbo, Y., Taira, T., Niki, T., Iguchi-Ariga, S. M. &
Ariga, H. DJ-1 restores p53 transcription activity
inhibited by Topors/p53BP3. Int. J. Oncol. 26,
96. Honbou, K. et al. The crystal structure of DJ-1, a
protein related to male fertility and Parkinson’s
disease. J. Biol. Chem. 278, 31380–31384 (2003).
97. Tao, X. & Tong, L. Crystal structure of human DJ-1, a
protein associated with early onset Parkinson’s
disease. J. Biol. Chem. 278, 31372–31379 (2003).
98. Wilson, M. A., Collins, J. L., Hod, Y., Ringe, D. &
Petsko, G. A. The 1.1-Å resolution crystal structure of
DJ-1, the protein mutated in autosomal recessive
early onset Parkinson’s disease. Proc. Natl Acad. Sci.
USA 100, 9256–9261 (2003).
99. Miller, D. W. et al. L166P mutant DJ-1, causative for
recessive Parkinson’s disease, is degraded through the
ubiquitin–proteasome system. J. Biol. Chem. 278,
100. Macedo, M. G. et al. The DJ-1L166P mutant protein
associated with early onset Parkinson’s disease is
unstable and forms higher-order protein complexes.
Hum. Mol. Genet. 12, 2807–2816 (2003).
101. Moore, D. J., Zhang, L., Dawson, T. M. & Dawson, V. L.
A missense mutation (L166P) in DJ-1, linked to
familial Parkinson’s disease, confers reduced protein
stability and impairs homo-oligomerization.
J. Neurochem. 87, 1558–1567 (2003).
102. Shendelman, S., Jonason, A., Martinat, C., Leete, T. &
Abeliovich, A. DJ-1 is a redox-dependent molecular
chaperone that inhibits α-synuclein aggregate
formation. PLoS Biol. 2, e362 (2004).
103. Unoki, M. & Nakamura, Y. Growth-suppressive effects
of BPOZ and EGR2, two genes involved in the PTEN
signaling pathway. Oncogene 20, 4457–4465 (2001).
104. Nakajima, A., Kataoka, K., Hong, M., Sakaguchi, M. &
Huh, N. H. BRPK, a novel protein kinase showing
increased expression in mouse cancer cell lines with
higher metastatic potential. Cancer Lett. 201,
105. Silvestri, L. et al. Mitochondrial import and enzymatic
activity of PINK1 mutants associated to recessive
parkinsonism. Hum. Mol. Genet. 14, 3477–3492
106. Rogaeva, E. et al. Analysis of the PINK1 gene in a
large cohort of cases with Parkinson disease. Arch.
Neurol. 61, 1898–1904 (2004).
107. Bonifati, V. et al. Early-onset parkinsonism associated
with PINK1 mutations: frequency, genotypes, and
phenotypes. Neurology 65, 87–95 (2005).
108. Beilina, A. et al. Mutations in PTEN-induced putative
kinase 1 associated with recessive parkinsonism have
differential effects on protein stability. Proc. Natl
Acad. Sci. USA 102, 5703–5708 (2005).
109. Li, Y. et al. Clinicogenetic study of PINK1 mutations in
autosomal recessive early-onset parkinsonism.
Neurology 64, 1955–1957 (2005).
110. Chacinska, A., Pfanner, N. & Meisinger, C. How
mitochondria import hydrophilic and hydrophobic
proteins. Trends Cell Biol. 12, 299–303 (2002).
111. Hanks, S. K. & Hunter, T. Protein kinases 6. The
eukaryotic protein kinase superfamily: kinase
(catalytic) domain structure and classification.
FASEB J. 9, 576–596 (1995).
112. Petit, A. et al. Wild-type PINK1 prevents basal and
induced neuronal apoptosis, a protective effect
abrogated by Parkinson’s disease-related mutations.
J. Biol. Chem. 280, 34025–34032 (2005).
113. Schulenberg, B., Aggeler, R., Beechem, J. M.,
Capaldi, R. A. & Patton, W. F. Analysis of steady-state
protein phosphorylation in mitochondria using a novel
fluorescent phosphosensor dye. J. Biol. Chem. 278,
114. Chen, R., Fearnley, I. M., Peak-Chew, S. Y. &
Walker, J. E. The phosphorylation of subunits of
complex I from bovine heart mitochondria. J. Biol.
Chem. 279, 26036–26045 (2004).
115. Kokoszka, J. E. et al. The ADP/ATP translocator is not
essential for the mitochondrial permeability transition
pore. Nature 427, 461–465 (2004).
116. Gray, C. W. et al. Characterization of human HtrA2, a
novel serine protease involved in the mammalian
cellular stress response. Eur. J. Biochem. 267,
117. Park, H. J., Seong, Y. M., Choi, J. Y., Kang, S. &
Rhim, H. Alzheimer’s disease-associated amyloid β
interacts with the human serine protease HtrA2/Omi.
Neurosci. Lett. 357, 63–67 (2004).
118. Martins, L. M. et al. Neuroprotective role of the
Reaper-related serine protease HtrA2/Omi revealed
by targeted deletion in mice. Mol. Cell Biol. 24,
119. Paisan-Ruiz, C. et al. Cloning of the gene containing
mutations that cause PARK8-linked Parkinson’s
disease. Neuron 44, 595–600 (2004).
120. Zimprich, A. et al. Mutations in LRRK2 cause
autosomal-dominant parkinsonism with pleomorphic
pathology. Neuron 44, 601–607 (2004).
121. Gilks, W. P. et al. A common LRRK2 mutation in
idiopathic Parkinson’s disease. Lancet 365, 415–416
122. West, A. B. et al. Parkinson’s disease-associated
mutations in leucine-rich repeat kinase 2 augment
kinase activity. Proc. Natl Acad. Sci. USA 102,
123. Gloeckner, C. J. et al. The Parkinson disease causing
LRRK2 mutation I2020T is associated with increased
kinase activity. Hum. Mol. Genet. 15, 223–232
124. Smith, W. W. et al. Leucine-rich repeat kinase 2
(LRRK2) interacts with parkin, and mutant LRRK2
induces neuronal degeneration. Proc. Natl Acad. Sci.
USA 102, 18676–18681 (2005).
125. Tatton, W. G., Chalmers-Redman, R., Brown, D. &
Tatton, N. Apoptosis in Parkinson’s disease: signals for
neuronal degradation. Ann. Neurol. 53 (Suppl. 3),
S61–S70; discussion S70–S72 (2003).
126. Vila, M. & Przedborski, S. Targeting programmed cell
death in neurodegenerative diseases. Nature Rev.
Neurosci. 4, 365–375 (2003).
127. Bowling, A. C. & Beal, M. F. Bioenergetic and oxidative
stress in neurodegenerative diseases. Life Sci. 56,
128. Jazwinski, S. M. Genetics of longevity. Exp. Gerontol.
33, 773–783 (1998).
129. Khalyavkin, A. V. & Yashin, A. I. How the analysis of
genetic mutations can help us to solve basic problems
in gerontology? II. Life extending genetic modifications
in budding yeast S. cereviseae, fruit fly
D. melanogaster and laboratory mice M. musculus.
Adv. Gerontol. 12, 46–56 (2003).
130. Hoglinger, G. U. et al. Dysfunction of mitochondrial
complex I and the proteasome: interactions between
two biochemical deficits in a cellular model of
Parkinson’s disease. J. Neurochem. 86, 1297–1307
131. Sullivan, P. G. et al. Proteasome inhibition alters
neural mitochondrial homeostasis and mitochondria
turnover. J. Biol. Chem. 279, 20699–20707
132. Beal, M. F. Limited-time exposure to mitochondrial
toxins may lead to chronic progressive
neurodegenerative diseases. Mov. Disord. 15,
133. McNaught, K. S., Perl, D. P., Brownell, A. L. &
Olanow, C. W. Systemic exposure to proteasome
inhibitors causes a progressive model of Parkinson’s
disease. Ann. Neurol. 56, 149–162 (2004).
134. Shamoto-Nagai, M. et al. An inhibitor of mitochondrial
complex I, rotenone, inactivates proteasome by
oxidative modification and induces aggregation of
oxidized proteins in SH-SY5Y cells. J. Neurosci. Res.
74, 589–597 (2003).
135. Qiu, J. H. et al. Proteasome inhibitors induce
cytochrome c-caspase-3-like protease-mediated
apoptosis in cultured cortical neurons. J. Neurosci.
20, 259–265 (2000).
136. Tanaka, Y. et al. Inducible expression of mutant
α-synuclein decreases proteasome activity and
increases sensitivity to mitochondria-dependent
apoptosis. Hum. Mol. Genet. 10, 919–926
137. Jesenberger, V. & Jentsch, S. Deadly encounter:
ubiquitin meets apoptosis. Nature Rev. Mol. Cell Biol.
3, 112–121 (2002).
138. Vogelstein, B., Lane, D. & Levine, A. J. Surfing the p53
network. Nature 408, 307–310 (2000).
139. Moore, D. J. et al. Association of DJ-1 and parkin
mediated by pathogenic DJ-1 mutations and
oxidative stress. Hum. Mol. Genet. 14, 71–84
140. Nicholls, D. G. Mitochondrial function and dysfunction
in the cell: its relevance to aging and aging-related
disease. Int. J. Biochem. Cell Biol. 34, 1372–1381
141. Duchen, M. R. Mitochondria in health and disease:
perspectives on a new mitochondrial biology. Mol.
Aspects Med. 25, 365–451 (2004).
The authors gratefully acknowledge grants from the Medical
Research Council (P.M.A.-S. and N.W.W.) and the Parkinson’s
Disease Society (N.W.W.), and a clinical training fellowship
to M.M.K.M., also provided by the Medical Research
Competing interests statement
The authors declare no competing financial interests.
The following terms in this article are linked online to:
Entrez Gene: http://www.ncbi.nlm.nih.gov/entrez/query.
α-synuclein | BAG5 | CDCREL1 | DAXX | DJ1 | HTRA2 | LRRK2 |
p54NRB | PAELR | parkin | PINK1 | POLG | PSF | PTEN | TOPORS
The Institute of Neurology: http://www.ion.ucl.ac.uk/
Access to this interactive links box is free online.
NATURE REVIEWS | NEUROSCIENCE
VOLUME 7 | MARCH 2006 | 219