MINIREVIEW

Cellular models to investigate biochemical pathways in

Parkinsons disease
Tiziana Alberio1, Leonardo Lopiano2 and Mauro Fasano1
1 Department of Biomedical, Informatics and Communication Sciences, and Centre of Neuroscience, University of Insubria, Busto Arsizio,
Italy
2 Department of Neuroscience, University of Torino, Italy

Keywords
cellular models; dopamine; mitochondrial
impairment; mitophagy; oxidative stress;
neurodegeneration; Parkinsons disease;
unfolded protein stress; a-synuclein
Correspondence
M. Fasano, Department of Biomedical,
Informatics and Communication Sciences,
and Centre of Neuroscience, University of
Insubria, via Alberto da Giussano 12, 21052
Busto Arsizio, Italy
Fax: +39 0331 339459
Tel: +39 0331 339450
E-mail: mauro.fasano@uninsubria.it
(Received 12 October 2011, revised 19
January 2012, accepted 31 January 2012)
doi:10.1111/j.1742-4658.2012.08516.x

Cellular models are instrumental in dissecting a complex pathological process into simpler molecular events. Parkinsons disease is multifactorial and
clinically heterogeneous; the aetiology of the sporadic (and most common)
form is still unclear and only a few molecular mechanisms have been claried so far in the neurodegenerative cascade. In such a multifaceted picture,
it is particularly important to identify experimental models that simplify
the study of the different networks of proteins genes involved. Cellular
models that reproduce some of the features of the neurons that degenerate
in Parkinsons disease have contributed to many advances in our comprehension of the pathogenic ow of the disease. In particular, the pivotal biochemical pathways (i.e. apoptosis and oxidative stress, mitochondrial
impairment and dysfunctional mitophagy, unfolded protein stress and
improper removal of misfolded proteins) have been widely explored in cell
lines, challenged with toxic insults or genetically modied. The central role
of a-synuclein has generated many models aiming to elucidate its contribution to the dysregulation of various cellular processes. In conclusion, classical cellular models appear to be the correct choice for preliminary studies
on the molecular action of new drugs or potential toxins and for understanding the role of single genetic factors. Moreover, the availability of
novel cellular systems, such as cybrids or induced pluripotent stem cells,
offers the chance to exploit the advantages of an in vitro investigation,
although mirroring more closely the cell population being affected.

Introduction
Parkinsons disease (PD) is the most common neurodegenerative movement disorder, affecting approximately
six million people worldwide [1,2]. Although clinically
heterogeneous, PD is characterized by the cardinal
signs of tremor at rest, bradykinesia and rigidity. In
late stages of the disease, the symptoms are accompanied by postural instability, gait dysfunction, swallow-

ing and speech difculties, autonomic disturbances

and, in up to 80% of the patients, by cognitive impairment [1]. Parkinsonian motor symptoms originate from
the degeneration of approximately 60% of dopaminergic neurons of the substantia nigra pars compacta
(SNpc), about 450 000 in adult human SNpc, which
is paralleled by the loss of 80% of dopamine in the

Abbreviations
CMA, chaperone-mediated autophagy; DAQ, dopamine quinone; DAT, dopamine transporter; iPS, induced pluripotent stem cells;
LB, Lewy bodies; LRRK2, leucine-rich repeat kinase 2; MPP+, 1-methyl-4-phenylpyridinium; PD, Parkinsons disease; ROS, reactive oxygen
species; SNpc, substantia nigra pars compacta; UCH-L1, ubiquitin carboxy-terminal hydrolase L1; UPR, unfolded protein response;
UPS, ubiquitin-proteasome system.

1146

FEBS Journal 279 (2012) 11461155 2012 The Authors Journal compilation 2012 FEBS

T. Alberio et al.

striatum (caudate nucleus and putamen). The selective

degeneration of SNpc neurons causes a reduced stimulation of the direct pathway (mediated by D1-like
receptors) and a reduced inhibition of the indirect
pathway (mediated by D2-like receptors), both leading
to an increased GABAergic tone in the output motor
nuclei of the basal ganglia and a consequent lower
activity of the thalamus with respect to facilitating the
initiation of movement [1,2]. The excessive subthalamic
nucleus glutamatergic activity, consequent to dopamine
depletion, might add a reex excitotoxic insult to SNpc
neurons [3]. Several compensatory mechanisms in this
complex network mask the onset of clinical symptoms
until 6070% of SNpc neurons are destroyed [4]. Thus,
treatments aiming to protect SNpc neurons and stop,
revert or modify the progression of PD are doomed to
failure unless the diagnosis occurs before the onset of
motor signs [5]. Currently available therapies are
mainly focused on the compensation of motor symptoms. Consequently, non-motor complications are the
major cause of disability in advanced PD patients.
Motor symptoms are not unique of PD, although PD
is the major cause of the parkinsonian syndrome. The
occurrence of tremor may be of iatrogenic or psychogenic origin, and patients affected by essential tremor
are sometimes erroneously diagnosed as PD patients.
Similarly, the occurrence of parkinsonian signs might
indicate that the patient is affected by atypical degenerative parkinsonisms, such as multiple system atrophy
or progressive supranuclear palsy [6].
PD is a complex, multifactorial disease in which different factors concur to the pathogenetic process.
In vitro models (established cell lines, primary cell cultures or stem cells) offer the advantage of a controlled
environment but may lack the cellular microenvironment critical to disease development. By contrast, a
suitable microenvironment may be present in animal
models, although differences in brain structure and in
the methods used to induce parkinsonism constitute
important limitations [7]. Cellular models are particularly useful for exploring single pathogenetic mechanisms and the genes proteins involved. Indeed, cellular
investigations have the intrinsic advantage of being
fast and reproducible and directly target the specic
molecular pathways at the basis of the progression of
PD [5].
The human neuroblastoma SH-SY5Y cell line has
been widely used as a cellular model to reproduce
impaired dopamine homeostasis, which is a possibly
pivotal aspect in the pathogenesis of PD. Indeed,
SH-SY5Y cells possess a complete dopaminergic system. In particular, they couple the good activity of the
dopamine transporter (DAT) with the low activity of

Biochemical pathways in Parkinsons disease

the vesicular monoamine transporter type 2, so that

the cytoplasmic dopamine concentration may be raised
by administration of exogenous dopamine in the culture medium [8,9]. The effect of specic PD-related
proteins has been investigated, after exposure to dopamine or other toxins in SH-SY5Y cells, as well as in
other catecholaminergic human neuroblastoma cell
lines, such as SK-N-BE or BE2-M17 cells. Further
neuron-like cellular models currently used in PD
research include PC12, a cell line derived from a pheochromocytoma of the rat adrenal medulla, and MES,
comprising hybrid rat mesencephalic-neuroblastoma
cells. Additionally, primary neuronal cultures derived
from animal models can constitute helpful cellular systems for investigating PD pathogenesis at the cellular
level, especially when exploring the role of specic
genes with the aim of determining neuron susceptibility
to different stress conditions [10].
In addition to the possibility of being used as a new
therapeutic approach, stem cells also represent a valuable tool for drug discovery. Human induced pluripotent stem cells (iPS) offer the possibility of obtaining
in vitro models of dopamine neurons. The development
of protocols to produce standardized iPS phenotypes
represents a remarkable advancement towards obtaining patient-specic stem cell lines for studying various
disease mechanisms [11].
Appropriate models for investigating the biochemical mechanisms of PD progression are instrumental in
the search for novel pharmacological targets, with the
aim of developing disease-modifying therapies and
addressing non-motor issues of PD [1,5,6]. Cellular
models are far from being able to reproduce the complexity of PD. However, they can provide valuable
insights for validation in animal models [7] and or in
human specimens [12]. In this review, we discuss how
cellular models may shed light on biochemical pathways that are impaired during PD pathogenesis,
including the formation and role of Lewy bodies, the
susceptibility of dopaminergic neurons, and the crosstalk of dopaminergic neurons and other cell types.

Genetic aetiology of PD
PD has long been considered as a sporadic, nongenetic
disease, and familiarity was an exclusion criterion in
diagnostic guidelines. On the other hand, environmental factors have been considered to strongly impact
PD aetiology since parkinsonism induced by mitochondrial toxins has been reported [1]. Additional support
for the environmental hypothesis was provided by
evidence of an increased PD risk in postencephalitic
patients. In 1996, a clear demonstration of a

FEBS Journal 279 (2012) 11461155 2012 The Authors Journal compilation 2012 FEBS

1147

Biochemical pathways in Parkinsons disease

T. Alberio et al.

genetically-linked form of PD was reported in a family

bearing a point mutation in the a-synuclein gene. Over
the next 15 years, the list of PD-linked mutations grew
rapidly, and genetic forms of the disease account for
up to 20% of PD cases to date. Several loci have been
associated with PD. In many cases, penetrance is not
complete, thus accounting for an intrinsic difculty in
their identication [1,2]. Nevertheless, the association
with PD is generally recognized for the proteins listed
in Table 1. The accumulating evidence indicates that
the molecular pathways of neurodegeneration triggered
by each mutation may be shared by several genetic
forms of PD and may also play a role in the common
sporadic disease [2]. Information obtained in genetic
approaches are useful for obtaining new insights into
sporadic PD because genetic factors may contribute to
the common biochemical pathways involved in PD
pathogenesis. Therefore, it would be important to be
able to integrate the neurodegenerative process in a
unifying perspective.
Autosomal recessive parkinsonisms are mainly
linked to mutations in the genes encoding parkin,
PINK1 and DJ-1, and are associated with early onset.
Carriers of these mutations display the upsurge of PD
motor symptoms before 50 years of age and the available post-mortem studies often show an atypical PD
pathology in that Lewy bodies (LB) may be absent [2].
Homozygous and compound heterozygous mutations
in the parkin gene cause 50% of familial and 20% of
early-onset PD cases, with more than 100 mutations
being reported so far. Parkin participates in a multiprotein E3 ubiquitin ligase complex and is involved
both in protein degradation through the ubiquitinproteasome system (UPS) and the autophagic degrada-

tion of dysfunctional depolarized mitochondria.

PINK1 is also involved in this process and is responsible for parkin recruitment to dysfunctional mitochondria [13]. Mutations in the gene encoding PINK1 are
the second most common cause of early-onset PD,
accounting for 7% of cases, whereas mutations in the
gene encoding DJ-1 are a rare cause of recessive PD
[2]. Many functions have been attributed to DJ-1, a
peroxiredoxin-like antioxidant enzyme with a protective role, mainly associated with oxidative stress
defence [9]. An enhanced cytoprotective effect in
response to mild oxidative stress was recently observed
in a C-terminal cleaved form of DJ-1 [14]. Homozygous and compound heterozygous mutations in the
gene for ATP13A2, which encodes for a lysosomal cation-transporting ATPase, have been shown to cause a
rare, atypical form of recessive juvenile parkinsonism
with dementia (PARK9), also known as KuforRakeb
syndrome [2]. PARK9 is associated with neurodegeneration and brain iron accumulation, which indicates a
role for ATP13A2 in mitochondrial function. Thus,
the protein might participate in mitochondrial quality
control, which also involves parkin, PINK1 and DJ-1
[15,16].
Autosomal dominant PD forms are linked to mutations in three principal genes: SNCA encoding a-synuclein (PARK1 for point mutations, PARK4 for
duplication or triplication of the gene), LRRK2
(PARK8) and the recently discovered GBA. Leucinerich repeat kinase 2 (LRRK2) is a large protein kinase
with a GTPase domain, involved in intracellular signalling pathways. Mutations in LRRK2 are the most
common cause of dominant familial PD and account
for 1% of sporadic PD cases [5]; mutations outside of

Table 1. Parkinsons disease-related proteins. Gene and protein names are indicated for genetic PD forms, where the gene product has
been identified. For gene products, official Uniprot protein names are listed. Proteins are cited in the text using their aliases.
PARK locus

Gene name

Gene product

Alias

PARK1 PARK4
PARK2
PARK5
PARK6
PARK7
PARK8
PARK9
PARK11
PARK13
PARK14
PARK15
PARK17
PARK18
Gauchers locus

SNCA
PARK2
UCHL1
PINK1
PARK7
LRRK2
ATP13A2
GIGYF2
HTRA2
PLA2G6
FBXO7
VPS35
EIF4G1
GBA

Alpha-synuclein
E3 ubiquitin-protein ligase parkin
Ubiquitin carboxyl-terminal hydrolase isozyme L1 (UCH-L1)
Serine threonine-protein kinase PINK1
Protein DJ-1
Leucine-rich repeat serine threonine-protein kinase 2
Probable cation-transporting ATPase 13A2
PERQ amino acid-rich with GYF domain-containing protein 2
Serine protease HTRA2 (HtrA2)
85 kDa calcium-independent phospholipase A2 (IPLA2)
F-box only protein 7
Vacuolar protein sorting-associated protein 35 (hVPS35)
Eukaryotic translation initiation factor 4 gamma 1 (EIF-4G1)
Glucosylceramidase

a-synuclein
parkin
UCH-L1
PINK1
DJ-1
LRRK2
ATP13A2
GIGYF2
HtrA2
IPLA2
FBXO7
hVPS35
EIF-4G1
GBA

1148

FEBS Journal 279 (2012) 11461155 2012 The Authors Journal compilation 2012 FEBS

T. Alberio et al.

Biochemical pathways in Parkinsons disease

the enzymatic domains have not been shown to segregate in a Mendelian way with the disease [17]. Mutations of GBA, a lysosomal enzyme, in turn link PD to
lysosomal storage disorders, as it is the case for the
lysosomal ATPase ATP13A2. Homozygous and compound heterozygous mutations in the gene for GBA
are linked to Gauchers disease, the most common
lysosomal storage disorder, whereas heterozygous carriers have an up to 10-fold increased risk of PD. Interestingly, among Ashkenazi Jewish PD patients, 20%
have GBA mutations and 15% have LRRK2 mutations [2].

Molecular mechanisms of PD
pathogenesis
Molecular mechanisms leading to degeneration of
SNpc neurons are not fully understood; however, the
available evidence indicates mitochondrial impairment,
UPS dysfunction, altered calcium homeostasis and oxidative stress as major factors [5,18]. The temporal and
the mechanistic relationships among these factors are
not well characterized and this still hampers the clarication of the causeeffect scheme that leads to neurodegeneration in PD. Dopamine spontaneously
oxidizes, giving rise to various molecules that can act
as endogenous toxins; thus, nigral dopaminergic
neurons are particularly susceptible to oxidative stress.
Iron appears to play a key role in the increased susceptibility of the pigmented, neuromelanin-containing neurons of SNpc; the dopamine-derived neuromelanin
pigment, in turn, appears to be a critical regulator of
iron homeostasis [19].
Dopamine homeostasis might be altered by a-synuclein oligomers (prebrillar intermediates) that associate
with vesicles and induce dopamine leakage [2]. Therefore, any process controlling the concentration of
a-synuclein may potentially affect the pathogenetic
oligomerization of the protein and dopamine-induced
toxicity. This includes point mutations (PARK1);
duplication and triplication of the a-synuclein gene
(PARK4); proteasomal and autophagic lysosomal
impairment; and regulation of expression at the transcriptional level [20,21]. In this view, the dopamine
a-synuclein interplay is a main player in the pathogenetic process [8] (Fig. 1). Moreover, a-synuclein
aggregation and bril formation can be regulated by
several post-translational modications. The current
literature associates a-synucleindopamine quinone
(DAQ) adducts with protobril formation; however, it
has been recently shown that noncovalent adducts possess a predominantly random coil conformation, which
shifts the bril formation equilibrium towards the solu-

Fig. 1. Evidence for biochemical mechanisms of neurodegeneration

in PD gained from cellular models. The interplay between altered
homeostasis of dopamine (DA) and a-synuclein aggregation into
protofibrils is at the basis of the oxidative stress condition. Protofibrils interact with DA storage vesicles and promote its release into
the cytosol, where it promptly oxidizes to DAQ. a-Synuclein variants derived from gene mutations (PARK1) or covalent modifications by DAQ are more prone to aggregate. Increased amounts of
a-synuclein, which might be up-regulated at the genetic (duplication triplication of the gene, PARK4) or transcriptional level, also
favour its aggregation. Proteolytic systems such as UPS and the
autophagy system might fail in the control of protein levels, either
for genetic reasons (loss of function of lysosomal proteins, i.e.
ATP13A2 and GBA) or because they are engulfed by a-synuclein
adducts, possibly modified by the addition of DAQ. The overall
result of this process is the increase of ROS together with an insufficient removal of damaged mitochondria through UPS and autophagy. a-Synuclein aggregation also regulates mitochondrial
dynamics in a process that is rescued by the Parkin PINK1
DJ-1 complex. Because of oxidative stress conditions, damaged
mitochondria will accumulate, leading to neuronal death. The same
effect is reached when mitophagy is directly impaired without
involvement of the a-synuclein cascade, as is the case for mutations of the genes for parkin, PINK1 or DJ-1, or when mitochondrial
functionality is specifically impaired in dopaminergic neurons (e.g.
with MPP+). Arrows indicate causal relationships.

ble, unfolded form [22]. By contrast, covalent addition

of DAQ stabilizes the b-sheet conformation that is
more prone to aggregate into protobrils. In both
cases, DAQ adducts prevent the protein from assuming its physiological conformation and possibly exerting its function.
The role of a-synuclein phosphorylation in the pathogenesis of PD and other synucleinopathies, such as
multiple system atrophy, has been extensively debated.

FEBS Journal 279 (2012) 11461155 2012 The Authors Journal compilation 2012 FEBS

1149

Biochemical pathways in Parkinsons disease

T. Alberio et al.

a-Synuclein phosphorylation may occur at distinct

tyrosine and serine residues. Most of the a-synuclein in
LB is phosphorylated on S129, although in vitro studies have recently demonstrated that phosphorylation at
S87, Y125, S129, Y133 and Y136 all inhibit a-synuclein bril formation, thus suggesting that phosphorylation does not promote a-synuclein aggregation.
Indeed, S129 phosphorylation by the PD-related kinase
LRRK2 may eliminate a-synuclein-induced nigrostriatal degeneration [23].
The generation of reactive oxygen species (ROS) by
altered dopamine homeostasis leads to mitochondrial
impairment, which in turn enhances the oxidative
stress condition. Toxins such as 1-methyl-4-phenylpyridinium (MPP+) and rotenone are able to damage
dopaminergic cells because they are taken up via the
DAT and block the mitochondrial respiratory chain,
with consequent activation of apoptotic cell death.
Conversely, oxidative stress conditions, such as those
triggered by a-synuclein aggregation, are the cause of
mitochondrial depolarization (Fig. 1).
Impaired protein clearance is a major feature that
contributes to neural loss in PD [21,24]. The identication of parkin as an ubiquitin E3 ligase, together with
the reported mutations of the deubiquitinating enzyme
ubiquitin carboxy-terminal hydrolase L1 (UCH-L1) in
rare cases from a single PD family, has suggested that
a failure in the UPS might signicantly contribute to
PD pathogenesis. Indeed, impaired UPS activity has
been shown in PD patients. a-Synuclein, either aggregated into protobrils or covalently modied by dopamine oxidation products, may engulf the UPS, thus
leading to an anomalous accumulation of undisposed
substrates. Modied a-synuclein was also reported to
interfere with autophagic protein disposal mechanisms
[25]. Interaction of dopamine-a-synuclein adducts with
the autophagosome interferes with the disposal of
damaged or misfolded proteins and leads to their accumulation. Accordingly, the specic inhibition of autophagic protein clearance leads to the accumulation of
aggregated proteins such as a-synuclein protobrils
(Fig. 1). Eventually, both UPS and autophagy protein
clearance systems are intrinsically less effective in the
elderly, thus exacerbating the pathogenetic cascade
[21,24].
Mitochondrial dysfunction and autophagy are
strictly connected. Indeed, damaged mitochondria are
removed through macroautophagy (mitophagy), aiming to reduce the activation of apoptosis. If this clearance process is impaired, either at the lysosomal level
(ATP13A2, GBA) or at the mitochondrial level (parkin, PINK1, DJ-1), an enhancement of oxidative stress
is expected to occur.
1150

Overall, there does not appear to be a single pathway to pathogenesis. However, we suggest that a main
path (i.e. the principal branch of a multibranched
pathway) can be identied that starts from a-synuclein
aggregation and leads to mitochondrial impairment.
This biochemical cascade may be activated at different
stages in various forms of PD: at its beginning, in the
classical pathology of the disease or directly at the
mitochondrion, as occurs in recessive parkinsonisms
[2] (Fig. 1). Noteworthy, a-synuclein pathology has
been recently shown to spread to neighbouring cells
via several cell-to-cell transmission mechanisms, thus
accounting for a possible amplication of this pathogenetic process [26].

Cellular models of PD pathogenetic

mechanisms
A common issue in PD research is the choice of the
best model system. A number of cellular and animal
models have been developed to clarify specic aspects
of the neurodegenerative process and to better characterize the sequence of pathological mechanisms
because they allow investigators to dissect the complex
pathogenetic cascade into simpler molecular events.
Several cellular models are able to reproduce biochemical alterations as a result of mutations of PD-related
proteins, whereas others reproduce cellular effects of
PD-related toxins, such as MPP+, the natural herbicide rotenone, 6-hydroxydopamine, DAQ and dopamine itself, combined or not with genetic factors.
Studies in cellular models can be performed by means
of targeted (molecular, biochemical or pharmacological) or unbiased (proteomic or transcriptomic)
approaches [27]. Cell lines are also useful for preliminary screening of putative therapeutic compounds and,
with respect to animal models, have the advantage of
possessing a human genetic background. However,
they are simplied systems, which cannot fully reproduce a dopaminergic neuron network or recapitulate
the entire pathogenetic ow of PD. Below, we describe
the contribution of PD models based on cellular systems to the understanding of molecular neurodegeneration mechanisms in PD.
Apoptosis and oxidative stress
Dopamine neuronal death in the SNpc has been frequently linked to an increase in oxidative stress that
overwhelms cellular protective mechanisms and
induces apoptosis. a-Synuclein may variously contribute to this pathogenetic mechanism. Indeed, the
expression of either wild-type or mutant protein in

FEBS Journal 279 (2012) 11461155 2012 The Authors Journal compilation 2012 FEBS

T. Alberio et al.

various cell lines has demonstrated that a-synuclein

modulates dopamine toxicity, which is related to ROS
produced by dopamine oxidation. Nevertheless, lack of
a-synuclein is detrimental for cell viability and several
protective functions have been proposed [28]. a-Synuclein is involved in vesicle trafcking at the synapse and
exhibits a chaperone-like function in the formation of
the SNARE complex [29], inhibits apoptosis activation
through the oligomerization with cytochrome c at the
mitochondrial surface [30] and exerts a protective function by modulating S-phase checkpoint responses at
the nuclear level [28]. Many of the toxins that are commonly used to generate experimental parkinsonian syndromes increase a-synuclein expression [9]. In this
vision, the normal function of a-synuclein is partially
neuroprotective and the up-regulation seen in chronic
neurodegenerative diseases is a compensatory mechanism that neurons adopt to protect themselves from
chronic oxidative stress [8,28]. The question whether
the interplay between dopamine and a-synuclein is
benecial or not has been widely tackled using dopaminergic cellular models in which the genetic and the
toxic factors were combined. Results have often been
controversial, as shown in recent studies [8,27]. These
ndings corroborate the view that a-synuclein dosedependently modulates the susceptibility of dopaminergic cells to dopamine toxicity and dopamine-induced
apoptosis. Clearly, it is crucial to understand the physiological functions of this protein, as well as how the
levels of a-synuclein and dopamine vary in SNpc neurons, especially with age. The observed discrepancies
in the literature might be a result of the chosen cellular
model. For example, it is still a matter of debate
whether established cell lines better mirror the disease
in their proliferative (tumour-like) or differentiated
(neuron-like) state (retinoic acid in the presence or not
of 12-O-tetradecanoylphorbol-13-acetate is commonly
used to differentiate SH-SY5Y cells) [31]. The expression of dopamine receptors is higher in undifferentiated cells, thus leading to the entry of a greater
amount of possibly toxic dopamine and 6-hydroxydopamine in the cell. Consequently, undifferentiated neuroblastoma cells are more sensitive to oxidative stress
and more informative on this pathogenetic mechanism.
By contrast, differentiated cells better recapitulate
other specic aspects of the pathophysiology [10,31].
SH-SY5Y cells have been employed in studies specically targeting the role of various proteins in the
oxidative stress response. The noted features of these
cells (dopaminergic system, differentiability) have led
to the generation of several experimental toxic models
aiming to test the action of potential apoptosis inhibitors after oxidative insult. The inspection of these

Biochemical pathways in Parkinsons disease

models has allowed the identication of proteins and

functional categories relevant to neuroprotection or
neurodegeneration [27]. For example, the activation
of a protective unfolded protein response (UPR) and
endoplasmic reticulum stress pathways helps to tolerate a moderate oxidative damage, with DJ-1 playing
a signicant role in such protective action [32]. After
oxidative challenge, DJ-1 has been observed to translocate to mitochondria and subsequently to the
nucleus. Exposure to nontoxic dopamine levels specically modies the two-dimensional electrophoresis
pattern of DJ-1, thus indicating a dopamine-dependent modication of the protein towards the activation of protective mechanisms [9]. The interplay of a
high cytosolic dopamine load and increased a-synuclein expression elicits biochemical alterations that,
through a bioinformatic network enrichment procedure, have been related to the involvement of the
NF-jB pathway [8].
The fact that LRRK2 is a protein kinase suggested
that PD may result from a decit in signal transduction mechanisms. Nevertheless, cellular models overexpressing LRRK2 mutants did not permit the
unambiguous identication of the transduction pathways involved [33]. LRRK2 may specically phosphorylate a-synuclein on the S129 residue. This
post-translational modication has been investigated in
cellular models. Apoptotic SH-SY5Y cells treated with
rotenone show increased S129-phosphorylated a-synuclein levels. Conversely, over-expression of S129A
mutant a-synuclein (i.e. a mutant unphosphorylated in
this position) decreases the activation of endoplasmic
reticulum stress and apoptosis [34]. All this points to
S129 phosphorylation being a pathogenetic factor.
Similarly, co-transfection of LRRK2 and a-synuclein
in SH-SY5Y cells induces increased a-synuclein aggregation and transmission of aggregates to neighbouring
cells [35]. However, these ndings are in contrast to
several observations supporting the role of S129-phosphorylated a-synuclein as the species endowed with
neuroprotective action [23].
The emerging development of iPS from mutation
carriers is unravelling the contribution of distinct genes
to the generation of oxidative stress and the consequent apoptosis process. Dopamine neurons derived
from LRRK2 mutant iPS show increased expression of
a-synuclein and key oxidative stress-response genes
and a higher sensitivity to cell death after exposure to
stress agents [36]. Moreover, dopamine neurons have
recently been generated from a-synuclein triplication
patients, giving rise to a very efcient tool for studying
mechanistic events in degeneration and new therapeutic agents [37].

FEBS Journal 279 (2012) 11461155 2012 The Authors Journal compilation 2012 FEBS

1151

Biochemical pathways in Parkinsons disease

T. Alberio et al.

Mitochondrial impairment
Impaired mitochondrial function has been implicated
since the very beginning in neurodegeneration in PD
[1]. In particular, alterations of the electron transport
chain enzyme complex I appear to be specically relevant to this disease. Mitochondrial dynamics might be
affected in several ways, including depolarization as a
result of mitochondrial toxins, perturbation of the
ssionfusion homeostasis, alteration of damaged
mitochondrion disposal through mitophagy and
reduced functionality as a result of inherited or
somatic mutations in the mtDNA [38].
MPP+ and rotenone neurotoxins, used to produce
models of PD, both disrupt mitochondrial function,
inhibiting complex I activity. They cause the specic
degeneration of dopaminergic neurons because they
are actively taken up via the DAT. The generation of
mitochondrial impairment induces the production of
ROS and a greater oxidative stress. This creates a
link between two of the major pathogenetic mechanisms in PD [2]. Moreover, specic mitochondrial
proteins were found to be covalently modied by
dopamine and its oxidized form, DAQ. This condition may trigger mitochondrial damage and play a
role in the enhanced vulnerability of dopaminergic
neurons [39].
Recent in vitro studies have delineated the role of
mitochondrial fragmentation in neuronal damage and
connected it with PD-related proteins: a-synuclein
directly affects membrane dynamics and induces a
mitochondrial fragmentation that is rescued by the
coexpression of PINK1, parkin or DJ-1 [40]. The
view that a-synuclein is an important regulator of
mitochondrial function is also supported by evidence
showing that its overexpression affects the expression
levels of mitochondrion-associated proteins. Very
recent studies have linked parkin, PINK1 and DJ-1
with mitochondrial autophagy (mitophagy) using
cellular models. Functional PINK1 has been shown
to be a prerequisite for parkin translocation to the
mitochondrion and the complex PINK1 parkin was
revealed to be crucial for proper mitophagy [13]. Furthermore, DJ-1 participates in this process, although
it works in parallel to the PINK1 parkin pathway to
maintain mitochondrial function [41]. In general, the
accumulation of damaged mitochondria as a result of
altered mitophagy might be a factor leading to autosomal recessive PD [24]. Experiments performed in
SH-SY5Y cells revealed that PINK1 is able to exert
its protective role through the phosphorylation of
Akt via activation of the mammalian target of rapamycin complex 2 [42]. A recent study using iPS cells,
1152

generated from patients with PINK1-linked PD or

genetically-matched normal subjects, has conrmed
the reduced recruitment of parkin to mitochondria
and a probable impairment of the degradation of
depolarized mitochondria [11].
Hybrid cell lines (cybrids), obtained from the fusion
of cells that lack mtDNA with platelet mtDNA from
patients [10], constitute a valuable novel model for
unravelling the causes of mitochondrial decits in
patients and retracing the molecular cascades generated by neurotoxins. These studies have offered new
insights into oxidative stress generation and susceptibility, as well as mitochondria morphology. Generation of ROS by mitochondria might be the cause of
the accumulation of mutations in mtDNA, which
would give rise to complex I dysfunctions and PD neurodegeneration [10].
Unfolded protein stress and removal of misfolded
proteins
Autophagy plays a fundamental role in the disposal of
damaged mitochondria, and as a mechanism for
removing misfolded, unfolded or damaged proteins, in
collaboration with the other major proteolytic system
(i.e. the UPS). Similar to other neurodegenerative diseases, PD is characterized by the presence of ubiquitinylated inclusions that point to a failure in protein
disposal mechanisms, including (but not limited to)
UPS [24]. Indeed, evidence of the accumulation of
autophagic vacuoles in PD might indicate autophagy
imbalance, either as a result of excessive activation or
a reduced completion of the proteolytic pathway.
Recent ndings linking mutations in lysosomial
enzymes (i.e. ATP13A2 and GBA) to PD give support
to this view [2].
The link between protein disposal failure and PD
has been investigated in several cellular models. The
presence of misfolded proteins is recognized as a central event in the activation of a defence mechanism,
leading to apoptosis in several neurodegenerative diseases, including PD. On this basis, the activation of
this mechanism after a-synuclein overexpression was
investigated both in SH-SY5Y cells and in non-neuronal HEK-293 that lack a-synuclein expression; these
studies indicated that a-synuclein misfolding and
aggregation may rst act as a molecular sensor for the
activation of proteolytic pathways and, subsequently,
apoptotic cell death [43]. a-Synuclein is physiologically
degraded by chaperone-mediated autophagy (CMA)
and dysfunction of CMA results in a-synuclein aggregation and compensative macroautophagy activation.
Moreover, stable expression of A53T a-synuclein in

FEBS Journal 279 (2012) 11461155 2012 The Authors Journal compilation 2012 FEBS

T. Alberio et al.

PC12 cells leads to the accumulation of autophagic

vacuoles [24]. Eventually, a-synuclein that has been
covalently modied by dopamine quinone species
interferes with CMA and leads to the accumulation of
native a-synuclein [25]. Consequently, a-synuclein
accumulation leads to an engulfment of protein disposal mechanisms that exacerbate the UPR.
The UPS was considered to be the main player in
PD pathogenesis for three reasons: UCHL1 was considered as a PD-linked protein, parkin was identied
to be a ubiquitin E3 ligase and reduced proteasome
activity was reported in human subjects who were
affected by PD. In the PC12 model, the proteasomal
inhibitor PSI induced cell apoptosis and the appearance of cytoplasmic LB-like inclusions, thus recapitulating two primary features of PD. These inclusions
were characterized by the identication of 56 proteins,
including 20 previously reported protein components
of LB, 28 newly-identied proteins and eight unknown
proteins [44]. Recently, the role of UPS in PD pathogenesis has gained a new relevance as a major factor
in mitophagy. Indeed, dysfunctional mitochondria activate mitophagy through the proteasome: several mitochondrial outer-membrane proteins are ubiquitinylated
by parkin and further degraded through the UPS.
These proteins may serve either as inhibitors of mitochondrial fragmentation (e.g. mitofusins) or as negative modulators of the interaction with the
autophagosome (e.g. the voltage-dependent anion
channel, VDAC1) [13]. Because both processes are
essential for the activation of mitophagy, it is conceivable that inhibition of the UPS leads to the accumulation of a protein cargo that should be eliminated and
simultaneously impairs the elimination of dysfunctional mitochondria.

Conclusions and perspectives

Cellular models can provide a wide portfolio of opportunities for studying pathogenetic mechanisms in PD
and be particularly useful for unravelling the role of
PD-related proteins or the mechanism of action of a
specic neurotoxin. Cell culture models may be divided
into three categories: established cell lines (either differentiated or not), stem cells and patient-derived cell
models [10]. Independent of the technical differences
among them, they can be regarded as powerful tools
for exploring PD pathophysiology, provided that the
limitations of such a simplistic approach are taken into
account.
Cellular models have contributed to a comprehensive explanation of the PD pathogenetic cascade and
delineation of the common functional aspects of the

Biochemical pathways in Parkinsons disease

various PD-related proteins. The central role of a-synuclein in the PD neurodegenerative process has led to
the generation of many models aiming to elucidate its
contribution to the dysregulation of various cellular
processes. A problem in clarifying its role in PD pathogenesis is its physiological function, which still
remains unclear. The protein has been localized in
many different cellular structures and appears to have
different functions. A dysregulation of a-synuclein concentration, aggregation state and post-translational
modications may produce neural suffering both for
the loss of its protective functions and for the gain of
new toxic properties. a-Synuclein is involved in all
three main pathogenetic mechanisms underlying PD.
Indeed, it is able to modulate dopamine toxicity and
inuence the cellular oxidative load, to interact directly
with mitochondrial membrane and change mitochondrial network dynamics and to affect the UPR. DJ-1
also protects against oxidative toxicity specically
induced by dopamine. Moreover, together with parkin
and PINK1, DJ-1 participates in the removal of damaged mitochondria. Parkin acts at the crossroad of the
UPS and the mitophagy process. Its activity is necessary for triggering mitochondrial disposal by the elimination of outer membrane proteins via the UPS. Of
note, all species that interfere with the cascade of
events described above (e.g. mitochondrial toxins,
ROS, UPS inhibitors, etc.) can mimic the pathogenetic
mechanisms, thus leading to cellular phenotypes similar to those determined by genetic mutations of
PD-related proteins.
On the basis of the overall view depicted above,
classical cellular models appear to be the correct
choice for preliminary studies aiming to investigate
the molecular action of new drugs or potential toxins
and, at the same time, aiming to understand the role
of single genetic factors in the pathogenetic process at
the cellular level. Still, it is important to stress the
impossibility of reproducing the complexity of the disease in vitro. The availability of new cellular systems,
such as cybrids or iPS cells, offers the chance of
exploiting the advantages of an in vitro investigation,
although more closely mirroring the affected cell population. These systems also represent a new step
towards personalized medicine, with the aim of testing each drug in a model with the same genetic background of the patient.

Acknowledgements
We are grateful to Professor Riccardo Fesce for critically reading the manuscript. The authors declare that
they have no conicts of interest.

FEBS Journal 279 (2012) 11461155 2012 The Authors Journal compilation 2012 FEBS

1153

Biochemical pathways in Parkinsons disease

T. Alberio et al.

References
1 Obeso JA, Rodriguez-Oroz MC, Goetz CG, Marin C,
Kordower JH, Rodriguez M, Hirsch EC, Farrer M,
Schapira AH & Halliday G (2010) Missing pieces in the
Parkinsons disease puzzle. Nat Med 16, 653661.
2 Shulman JM, De Jager PL & Feany MB (2011) Parkinsons disease: genetics and pathogenesis. Annu Rev
Pathol 6, 193222.
3 Blandini F (2010) An update on the potential role of
excitotoxicity in the pathogenesis of Parkinsons disease.
Funct Neurol 25, 6571.
4 George JL, Mok S, Moses D, Wilkins S, Bush AI,
Cherny RA & Finkelstein DI (2009) Targeting the progression of Parkinsons disease. Curr Neuropharmacol 7,
936.
5 Dawson TM, Ko HS & Dawson VL (2010) Genetic animal models of Parkinsons disease. Neuron 66, 646661.
6 Morgan JC, Mehta SH & Sethi KD (2010) Biomarkers
in Parkinsons disease. Curr Neurol Neurosci Rep 10,
423430.
7 Blandini F & Armentero MT (2012) Animal models of
Parkinsons disease. FEBS J, this issue.
8 Alberio T, Bossi AM, Milli A, Parma E, Gariboldi MB,
Tosi G, Lopiano L & Fasano M (2010) Proteomic analysis of dopamine and a-synuclein interplay in a cellular
model of Parkinsons disease pathogenesis. FEBS J 277,
49094919.
9 Alberio T, Colapinto M, Natale M, Ravizza R, Gariboldi MB, Bucci EM, Lopiano L & Fasano M (2010)
Changes in the two-dimensional electrophoresis pattern
of the Parkinsons disease related protein DJ-1 in
human SH-SY5Y neuroblastoma cells after dopamine
treatment. IUBMB Life 62, 688692.
10 Schule B, Pera RA & Langston JW (2009) Can cellular
models revolutionize drug discovery in Parkinsons disease? Biochim Biophys Acta 1792, 10431051.
11 Seibler P, Graziotto J, Jeong H, Simunovic F, Klein C
& Krainc D (2011) Mitochondrial Parkin recruitment is
impaired in neurons derived from mutant PINK1
induced pluripotent stem cells. J Neurosci 31,
59705976.
12 Ravid R & Ferrer I (2012) Brain banks as key player in
biochemical and molecular studies on cerebral cortex
involvement in Parkinsons disease. FEBS J, this issue.
13 Geisler S, Holmstrom KM, Skujat D, Fiesel FC, Rothfuss OC, Kahle PJ & Springer W (2010) PINK1 Parkin-mediated mitophagy is dependent on VDAC1 and
p62 SQSTM1. Nat Cell Biol 12, 119131.
14 Chen J, Li L & Chin LS (2010) Parkinson disease protein DJ-1 converts from a zymogen to a protease by
carboxyl-terminal cleavage. Hum Mol Genet 19,
23952408.
15 Schneider SA, Paisan-Ruiz C, Quinn NP, Lees AJ,
Houlden H, Hardy J & Bhatia KP (2010) ATP13A2

1154

16

17

18

19

20
21

22

23

24

25

26

27

28

29

mutations (PARK9) cause neurodegeneration with

brain iron accumulation. Mov Disord 25, 979984.
Branco DM, Arduino DM, Esteves AR, Silva DF,
Cardoso SM & Oliveira CR (2010) Cross-talk between
mitochondria and proteasome in Parkinsons disease
pathogenesis. Front Aging Neurosci 21, 217.
Cookson MR (2010) The role of leucine-rich repeat
kinase 2 (LRRK2) in Parkinsons disease. Nat Rev
Neurosci 11, 791797.
Surmeier DJ, Guzman JN, Sanchez-Padilla J & Goldberg
JA (2010) What causes the death of dopaminergic neurons in Parkinsons disease? Prog Brain Res 183, 5977.
Double KL, Reyes S, Werry EL & Halliday GM (2010)
Selective cell death in neurodegeneration: why are some
neurons spared in vulnerable regions? Prog Neurobiol
92, 316329.
Cookson MR & Bandmann O (2010) Parkinsons disease:
insights from pathways. Hum Mol Genet 19, R21R27.
Cuervo AM, Wong ES & Martinez-Vicente M (2010)
Protein degradation, aggregation, and misfolding. Mov
Disord 25(Suppl 1), S49S54.
Bisaglia M, Tosatto L, Munari F, Tessari I, Polverino
de Laureto P, Mammi S & Bubacco L (2010) Dopamine quinones interact with alpha-synuclein to form
unstructured adducts. Biochem Biophys Res Commun
394, 424428.
Oueslati A, Fournier M & Lashuel HA (2010) Role of
post-translational modications in modulating the structure, function and toxicity of alpha-synuclein: implications for Parkinsons disease pathogenesis and
therapies. Prog Brain Res 183, 115145.
Xilouri M & Stefanis L (2011) Autophagic pathways in
Parkinson disease and related disorders. Expert Rev
Mol Med 13, e8.
Martinez-Vicente M, Talloczy Z, Kaushik S, Massey
AC, Mazzulli J, Mosharov EV, Hodara R, Fredenburg
R, Wu DC, Follenzi A et al. (2008) Dopamine-modied
alpha-synuclein blocks chaperone-mediated autophagy.
J Clin Invest 118, 777788.
Dunning CJ, Reyes JF, Steiner JA & Brundin P (2011)
Can Parkinsons disease pathology be propagated from
one neuron to another? Prog Neurobiol
doi:10.1016/j.pneurobio.2011.11.003.
Alberio T & Fasano M (2011) Proteomics in Parkinsons disease: an unbiased approach towards peripheral
biomarkers and new therapies. J Biotechnol 156,
325337.
Liu X, Lee YJ, Liou LC, Ren Q, Zhang Z, Wang S &
Witt SN (2011) Alpha-synuclein functions in the
nucleus to protect against hydroxyurea-induced replication stress in yeast. Hum Mol Genet 20, 34013414.
Burre J, Sharma M, Tsetsenis T, Buchman V, Etherton
MR & Sudhof TC (2010) Alpha-synuclein promotes
SNARE-complex assembly in vivo and in vitro. Science
329, 16631667.

FEBS Journal 279 (2012) 11461155 2012 The Authors Journal compilation 2012 FEBS

T. Alberio et al.

30 Bayir H, Kapralov AA, Jiang J, Huang Z, Tyurina YY,

Tyurin VA, Zhao Q, Belikova NA, Vlasova II, Maeda
A et al. (2009) Peroxidase mechanism of lipid-dependent cross-linking of synuclein with cytochrome C: protection against apoptosis versus delayed oxidative stress
in Parkinson disease. J Biol Chem 284, 1595115969.
31 Martins Lopes F, Schroder R, Conte da Frota ML Jr,
Zanotto-Filho A, Muller CB, Pires AS, Meurer RT,
Delevati Colpo G, Pens Gelain D, Kapczinski F et al.
(2010) Comparison between proliferative and neuronlike SH-SY5Y cells as an in vitro model for Parkinson
disease studies. Brain Res 1337, 8594.
32 Bauereis B, Haskins WE, Lebaron RG & Renthal R
(2011) Proteomic insights into the protective mechanisms of an in vitro oxidative stress model of early stage
Parkinsons disease. Neurosci Lett 488, 1116.
33 Berwick DC & Harvey K (2011) LRRK2 signaling
pathways: the key to unlocking neurodegeneration?
Trends Cell Biol 21, 257265.
34 Sugeno N, Takeda A, Hasegawa T, Kobayashi M,
Kikuchi A, Mori F, Wakabayashi K & Itoyama Y
(2008) Serine 129 phosphorylation of alpha-synuclein
induces unfolded protein response-mediated cell death.
J Biol Chem 283, 2317923188.
35 Kondo K, Obitsu S & Teshima R (2011) a-Synuclein
aggregation and transmission are enhanced by leucinerich repeat kinase 2 in human neuroblastoma SH-SY5Y
cells. Biol Pharm Bull 34, 10781083.
36 Nguyen HN, Byers B, Cord B, Shcheglovitov A, Byrne
J, Gujar P, Kee K, Schule B, Dolmetsch RE, Langston
W et al. (2011) LRRK2 mutant iPSC-derived DA neurons demonstrate increased susceptibility to oxidative
stress. Cell Stem Cell 8, 267280.
37 Devine MJ, Ryten M, Vodicka P, Thomson AJ, Burdon
T, Houlden H, Cavaleri F, Nagano M, Drummond NJ,
Taanman JW et al. (2011) Parkinsons disease induced

Biochemical pathways in Parkinsons disease

38

39

40

41

42

43

44

pluripotent stem cells with triplication of the a-synuclein locus. Nat Commun 2, 440.
Schapira AH & Gegg M (2011) Mitochondrial contribution to Parkinsons disease pathogenesis. Parkinsons
Dis 2011, 159160.
Hastings TG (2009) The role of dopamine oxidation in
mitochondrial dysfunction: implications for Parkinsons
disease. J Bioenerg Biomembr 41, 469472.
Kamp F, Exner N, Lutz AK, Wender N, Hegermann J,
Brunner B, Nuscher B, Bartels T, Giese A, Beyer K
et al. (2010) Inhibition of mitochondrial fusion by
a-synuclein is rescued by PINK1, Parkin and DJ-1.
EMBO J 9, 35713589.
Thomas KJ, McCoy MK, Blackinton J, Beilina A, van
der Brug M, Sandebring A, Miller D, Maric D,
Cedazo-Minguez A & Cookson MR (2011) DJ-1 acts in
parallel to the PINK1 parkin pathway to control mitochondrial function and autophagy. Hum Mol Genet 20,
4050.
Murata H, Sakaguchi M, Jin Y, Sakaguchi Y, Futami
J, Yamada H, Kataoka K & Huh NH (2011) A new
cytosolic pathway from a Parkinson disease-associated
kinase, BRPK PINK1: activation of AKT via
mTORC2. J Biol Chem 286, 71827189.
Bellucci A, Navarria L, Zaltieri M, Falarti E, Bodei S,
Sigala S, Battistin L, Spillantini M, Missale C & Spano
P (2011) Induction of the unfolded protein response by
a-synuclein in experimental models of Parkinsons disease. J Neurochem 116, 588605.
Li X, Zhang Y, Xie P, Piao J, Hu Y, Chang M, Liu
T & Hu L (2010) Proteomic characterization of an
isolated fraction of synthetic proteasome inhibitor
(PSI)-induced inclusions in PC12 cells might offer
clues to aggresomes as a cellular defensive response
against proteasome inhibition by PSI. BMC Neurosci
11, 95.

FEBS Journal 279 (2012) 11461155 2012 The Authors Journal compilation 2012 FEBS

1155