Degenerative mechanism in dopaminergic neurons

1.

INTRODUCTION
Parkinson’s disease (PD) is a widely occurring neurodegenerative disease that

was first described by James Parkinson (Parkinson, 1817). The main clinical symptoms
displayed by PD patients include bradykinesis (difficulty in initiating and carrying out
movements), tremor, muscle rigidity and jerky movements. The more common late-onset,
sporadic form of PD affects approximately 2% of the world’s population over the age of 65
years (Hughes et al., 1993). The young are not spared from this disease although the earlyonset familial forms are rare. The clinical symptoms were later found to be due to the
selective degeneration of neurons containing dopamine as their neurotransmitter within the
substantia nigra pars compacta (SNpc), a midbrain structure. These cells form part of an
interconnecting neuronal circuitry within the brain, which functions to control voluntary
motor movement. With the progressive loss of up to 80% of these dopaminergic neurons,
this circuitry is adversely affected, resulting in the manifestation of the clinical symptoms.
Given the wide spread nature of this disease, epidemiological studies revealed
that the economical burden of PD amounts to between US$7.1b to US$24.5b in USA alone
(Siderwof, 2001). Much more are lost indirectly through the decrease in productivity and
informal care by family members. The cause of PD is unknown. No therapeutic agents
present are able to cure or delay the onset of PD. The best treatment to date is the use of LDOPA, a dopamine precursor. L-DOPA is taken orally and it serves to replace the
dopamine lost in the SNpc due to the death of the dopaminergic cells (Ball, 2001).
However, this treatment could only to reduce the symptoms associated with PD, and is not
exactly a long-term cure for it. Chronic administration of L-DOPA was recently shown to
cause motor fluctuations as well as neuropsychiatric problems such as cognitive

1


impairment (Riley and Lang, 1993). As such, the search for better therapeutic treatment for
the cure of PD continues. To achieve this, it is imperative to first understand the cause of
PD and the basis of the selective loss of dopaminergic cells in the brain. Until further

studies are done to affirm the molecular pathways of neuronal cell death, it will not be
possible to design appropriate and effective therapeutic agents to rescue this disease.

1.1

A Short Review on Nigral Degeneration – The Vulnerability of
Dopaminergic Neurons
Although the etiology of the selective degeneration of dopaminergic cells

remained elusive, constant efforts are being made to unravel the phenomena underlying the
degeneration process. Studies have thus been done on dopaminergic neurons alone, most
trying to understand why; of all types of neurons, only the dopaminergic neurons are
affected in PD. Many factors hence surfaced.
The utilization of dopamine as the neurotransmitter by these cells is in fact
an intrinsic stress factor to the cells (Cohen et al., 1997). Dopamine is metabolized by the
enzyme monoamine oxidase (MAO). This is a natural metabolic process to clear the
intracellular dopamine within the cytosol. However, it was found to generate a significant
amount of hydrogen peroxide (H202). The H202 can be further broken down to generate
glutathione disulfide (GSSG). Both metabolic products are potentially damaging. GSSG
reacts spontaneously with thiol groups in proteins to form protein mixed disulfides. This
reaction compromises the functions of the affected proteins. In addition, increased iron
level has been noted in the brains of PD patients (Dexter et al., 1989). The H202 generated
could react with the iron to generate toxic hydroxyl radicals, which are responsible for
damages to membrane proteins via lipid peroxidation. Mitochondrial membranes could

2


thus be greatly affected, which in turns, affect its functions. The location of MAO on the
outer membrane of the mitochondria could thus potentially evoke changes, whether

directly or indirectly, at the distant inner membrane. These processes have the potential to
affect several enzymes located within the mitochondria. ATPase, an important enzyme
responsible for the generation of cellular energy ATP, is one of them. A further postulation
is that disturbance to the mitochondria function could lead to the impairment of proton
pumping, resulting in a decrease in the mitochondrial membrane potential. This would in
turn release cytochrome-c, leading to the onset of apoptosis (Desagher and Martinou,
2000).
Parkinsonian brains have also been shown to possess decrease level of
glutathione (GSH) (Riederer et al., 1989). GSH is an anti-oxidant, and may be involved in
the detoxification of H202 generated from dopamine turnover. A decrease in the level of
GSH would mean a higher basal level of H202 within the cells, thus leading to damages of
cellular macromolecules and their subsequent peroxidation.
The predisposition of dopaminergic neurons to oxidative damages does not
stop here. The proximity of the mitochondria to these reactive oxygen species (ROS)
increases the mutation rate of the mitochondria DNA by 10-20 times (Ozawa et al., 1997).
The lack of protective histone-like proteins (Clayton et al., 1974) and the poor DNA repair
system in the mitochondria (Shadel and Clayton, 1997) could evoke additive damaging
effect to the mitochondrial genome. The mitochondrial genome encodes many proteins of
the oxidative phosphorylation system. This system is responsible for the production of
ATP. A decline in ATP production due to damages to the mitochondria would result in
cellular degeneration, whether it is by apoptosis or necrosis.

3


All in all, the above factors have been determined and could act
synergistically to produce deleterious effect in dopaminergic cells. This increase
susceptibility to either environmental or endogenous agents could help explain the
selective degeneration of dopaminergic neurons in PD. Again, although these facts are
known, the exact degenerative mechanisms involved remained enigmatic. Thus, efforts are

still needed to unravel the mysterious mode of cell death underlying the selective nigral
degeneration.

1.2

Concluding Remarks on PD Research Review
After many years of research, the degenerative mechanisms underlying the

selective death of dopaminergic neurons remained a mystery. Up till now, there are still no
conclusive evidences that state the exact mechanism of death in these cells. However,
many models have been constantly generated to provide more clues to the understanding of
PD. With the progressive refinement of such models and the use of more powerful research
techniques, the etiology of PD should one day come to light. The focus now for scientists
working on PD should be to continue in their area of research, thus contributing to the pool
of information with their expertise. This is thus the aim of the present investigation. With
the collective efforts of scientists all around the world, the understanding of PD should be
within reach, hence giving the many sufferers of PD hope in returning to a normal lifestyle
they once had.

1.3

The Use of Neurotoxins in PD Research
In an effort to understand the etiology of PD, the ideal scenario is of course to

extract information directly from patients suffering from PD. However, due to the

4


inaccessibility of the mid-brain section and the lack of non-evasive diagnostic tools to

study the brain, it has been difficult to obtain valuable information about the real-time
events that are happening in the neurons of these PD patients. In addition, ethical concerns
about the use of PD patients as subjects for studies add on to the current problem. As such,
pharmacological agents and neurotoxins have been used to develop experimental models in
a wide variety of species in order to avoid the problems faced with live human subjects.
These neurotoxins must demonstrate their abilities to induce and mimic at least certain
clinical characteristics found in PD patients, e.g. rigidity or the specific loss of
dopaminergic neurons in the SNpc. There are currently four neurotoxins which have been
established to induce most of the characteristics of PD. The choice of the neurotoxins
employed in each study is dependent on the aspects in which the studies are approached.

1.3.1

The Use of Dopamine in PD Research
Given the review presented in section 1.1 regarding the intrinsic stress of

dopaminerigic cells, it is no surprise that dopamine, the neurotransmitter itself, is used as a
neurotoxin to study the disease. Dopamine is synthesized and contained in vesicles within
the neurons. These vesicles serve to regulate the concentration of the neurotransmitter in
the cytoplasm and the synaptic cleft (Jonsson, 1971). Dopamine is metabolized over time
and this metabolism via MAO can potentially lead to the formation of H202 and
dihydroxyphenylacetic acid (Maker et al, 1981). The presence of H2O2 leads to oxidative
stress. Oxidative stress occurs when the level of reactive oxygen species (ROS) such as
H2O2 and hydroxyl radicals is beyond the threshold of what the cells can handle. Increased
level of ROS is harmful to the cells because it can cause severe damages to DNA, proteins
and lipids in the membranes. The reaction of the H2O2 with the high levels of iron found in

5

the SNpc region produces hydroxyl radical, which can then immediately react with lipids,
DNA, and susceptible amino acids in proteins, resulting in cellular damage (Halliwell,
1992). In addition, the catechol ring of the dopamine molecule can also undergo oxidation
spontaneously in the presence of transition metals like iron or enzymatically, to form DA
quinone and more ROS (Hastings, 1995). The oxidative stress effects of dopamine can be
attenuated by antioxidants such as GSH (Gabbay et al., 1996). Thus, dopamine metabolism
can lead to the generation of high levels of ROS within the dopaminergic neurons, thereby
leading to oxidative stress and cellular damage. Therefore, the exposure of dopaminergic
neurons to its own neurotransmitter, even at physiological concentrations (0.1 – 1mM) can
induce oxidative stress and subsequent cell death (Ziv et al., 1994).
The mode of cell death induced by dopamine seems to indicate towards
apoptosis. Apoptosis is a regulated mode of cell suicide, the details of which will be
discussed in section 1.4. The exposure of both neuronal and non-neuronal cells to
dopamine induces several morphological and biochemical hallmarks of apoptosis (Ziv et
al., 1994; Stokes et al., 2000). DNA damage, a common downstream effect of oxidative
stress was observed with the increased level of p53, a molecule involved in this event
(Daily et al., 1999). Bax, a protein also involved in apoptosis, was shown to be activated
and the overexpression of its inhibitor Bcl-2, was able to block the dopamine-induced
apoptosis (Offen et al., 1997).
Although there are clues to the mechanisms in which dopamine causes cell
death, the event is not completely elucidated. Moreover, the role of dopamine as a
neurotoxin to study PD is plagued by the belief that the death mechanisms observed after
exposing cells or animals to exogenous dopamine at high concentrations is not

6


representative of the true events happening within the brain of PD patients. It has been
argued that the main factors for concern should be the intracellular sources of dopamine, as
well as its redistribution within the cells. Although this is in part true, the relevance of

dopamine and its metabolism to the pathology of PD should not be underestimated. Studies
using dopamine should continue in order to further understand its effects on the intrinsic
environment within and around these dopaminergic neurons.

1.3.2

The Use of 6-hydroxydopamine (6-OHDA) in PD Research
6-OHDA is a hydroxylated analogue of dopamine and is extensively used as a

model to study PD. In experimental models of PD, 6-OHDA is directly injected into the
striatum, the substantia nigra, or the ascending medial forebrain bundle (for rats). Although
6-OHDA cannot cross the blood-brain barrier, these direct intracerebral injections can
reproduce the phenomena of striatal neuronal degeneration, dopamine depletion and the
motor impairments that come along with it. The relevance of 6-OHDA for PD-related
study not only stems from its ability to induce parkinsonism, but also, it has been found to
occur naturally in both rat and human brains (Senoh and Witkop, 1959; Curtius et al.,
1974) and in the urine of L-DOPA treated PD patients (Andrew et al., 1993). 6-OHDA can
be produced by a non-enzymatic reaction between dopamine, H202 and free iron (Linert et
al., 1996). Thus, this compound may even potentially play a very important role either in
the onset and/or the progress of the disease. The use of this compound should yield results
relevant to the etiology of PD.
The presence of 6-OHDA in the striatal region has been suggested to cause
nigrostriatal dopaminergic lesions via the generation of ROS e.g. H202 (Heikkila and
Cohen, 1971). The oxidative stress produced from the high levels of ROS generated in vivo

7


(Kumar et al., 1995) and in vitro (Choi et al., 1999b) can be nullified by the cointroduction of antioxidants (Yamada et al., 1997; Mayo et al., 1999). 6-OHDA was also
found to induce biochemical and morphological hallmarks of apoptosis e.g. chromatin

condensation as detected by TUNEL assays, in the SNpc of 6-OHDA injected rats (Zuch et
al., 2000). Similar to dopamine-induced cell death, the level of p53 and Bax are also
increased in models using 6-OHDA (Blum et al., 1997). Therefore, the mechanisms of cell
death induced by both dopamine and 6-OHDA appear to be overlap. This is not surprising
since both compounds are very related in structure. Like many models, 6-OHDA does not
completely mimic all the clinical and pathological features of PD. However, its natural
existence and the potentially harmful effects of its presence still make it a good neurotoxin
to use for the study of nigral degeneration.

1.3.3

The Use of Rotenone in the Study of PD
Rotenone has been one of the most successful neurotoxins which is able to

induce almost all of the characteristics of PD. Rotenone is a naturally occurring highaffinity complex I inhibitor. It is an organic pesticide and is commonly used to kill
nuisance fishes in lakes and reservoirs. Owning to its extreme lipophilic nature, it can cross
biological membranes easily and its movement is rapid and independent of transporters (as
compared to MPTP/MPP+). As such, rotenone is a systemic complex I inhibitor which can
act on all parts of the brain (Talpade et al., 2000). However, its administration causes only
the selective degeneration of the striatal neurons (Betarbet et al., 2000). This is remarkable
as it indicates that nigrostriatal neurons are particularly vulnerable to complex I inhibitors.
Therefore, exposure to such compounds could potentially be a factor in PD onset.

8


As for other neurotoxins, the exact mechanism in which it causes cell death is
still unknown. However, since it is a specific complex I inhibitor, evidences had surfaced
which demonstrate its effects on inducing oxidative stress. Chronic exposure of SH-SY5Y,
a human neuroblastoma cell line, to rotenone over the course of four weeks greatly reduced

the level of GSH. In addition, there were observable increase in DNA oxidation and
protein damage (Sherer et al., 2002). These rotenone-treated cells also showed apoptotic
characteristic like the release of cytochrome-c and the activation of caspase-3. The same
observations were demonstrated by another study using the same cell line (Kitamura et al.,
2002). In this study, caspase-9, caspase-3 as well as caspase-12 were activated in the
presence of rotenone. Apart from caspase activation, other apoptotic characteristics e.g.
DNA fragmentation was also evident. Thus, the mode of action of rotenone seems to
suggest that a complex I-induced oxidative damage will subsequently lead to caspase
activation and apoptosis.
Rotenone has been an excellent model for the study of PD based on three
main observations. Firstly, rotenone is a systemic complex I inhibitor. This is in agreement
with the observed loss of complex I activity in other types of cells i.e. in the platelets of PD
patients (Haas et al., 1995). Secondly, the specific and chronic degeneration of the SNpc
induced by rotenone indeed recapitulate the nature of the progressive dopaminergic cell
loss in PD patients. Last but not least, rotenone-treated animals showed the presence of
Lewy bodies, a cytoplasmic inclusion consisting of mainly the alpha-synuclein and
ubiquitin proteins (Betarbet et al., 2000). This is one of the features that the other
neurotoxins are unable to reproduce (Betarbet et al., 2002).

9


Although rotenone has been able to reproduce most of the features of PD, its
requirement for a chronic dosage regime into animal models is very capital- and labourintensive. In addition, the major disadvantage of using rotenone is that not all animals and
cells respond to its toxicity and develop SNpc leisions. In summary, this neurotoxin can
indeed recapitulate most of the features of PD. Its niche in inducing Lewy bodies remained
unparalleled. However, the specific study of cell death would still require a neurotoxin
which can cause a uniformed response. This thus leaves a stage for the use of dopamine, 6OHDA, as well as MPTP/MPP+, in models for PD research.

1.4

The Use of 1-Methyl-4-Phenyl-1,2,3,6-Tetrahydropyride (MPTP) in the
Understanding of PD
The use of MPTP in PD research has been extensive. It was first discovered

in 1979, when a group of young Californians addicted to a synthetic heroin analog was
found to develop PD symptoms. Post-mortem studies on the brains of these victims
revealed the loss of cells in the substantia nigra similar to that found in PD (Davis et al.,
1979). This led to the discovery of the neurotoxin MPTP. MPTP was found to induce
many of the biochemical and neuropathological changes that are observed in postmortem
brains of PD patients. Similarly, these changes were also observed in the striatal neuronal
circuit of MPTP-treated animal models such as in monkeys and mice. The changes include
typical PD characteristics such as marked reduction of dopaminergic neurons and
dopamine content. The treated animals also manifest PD symptoms such as bradykinesis
and muscle rigidity (Kopin and Markey, 1988; Langston, 1996). From then on, MPTP has
been widely used to induce Parkinsonism in a variety of animal and cell models in order to
study the death mechanisms behind the nigralstriatal cell death. So far, the studies have

10


been promising in providing many important clues to the understanding of the etiology and
pathogenesis of PD. The comparison of idiopathic and MPTP induced PD symptoms are
summarized in Table 1.
Idiopathic PD
Bradykinesia
Rigidity
Postural instability
Rest tremor
Loss of dopaminergic neurons

Response to L-DOPA
Lewy bodies
(α-synuclein and ubiquitin positive)
Decrements in complex I

MPTP Intoxication
Bradykinesia
Rigidity
Postural instability
Rest tremor
Loss of dopaminergic neurons
Response to L-DOPA
Lewy bodies
(α-synuclein and ubiquitin positive)
Decrements in complex I

Table 1. Comparison of pathological changes observed in idiopathic PD and MPTPinduced toxicity. (This table is adapted from Zhang et al., 2000b).
The most widely used animals to generate models to study the neurotoxicity
of MPTP are monkeys and mice. When administered to animals, MPTP is oxidized to
MPP+ (1-methyl-4-phenylpyridinium) by the enzyme monoamine oxidase B (Heikkila et
al., 1984). MPP+ then enters the dopaminergic cells via the dopamine uptake sites on the
surface of the cells (Chiba et al., 1985). The MPP+ finally enters the mitochondria. There,
MPP+ exerts its effect by inhibiting the complex I of the oxidative phosphorylation chain
(Nicklas et al., 1985). The mitochondrial inhibition leads to a decrease in ATP production,
an important cellular form of energy. The loss of mitochondrial membrane potential then
ensues (Mizuno et al., 1988).
Similarly to the other neurotoxins discussed above, the neurotoxicity effect of
MPTP is attributed to the generation of oxidative stress. Inhibition of the complex I could
lead to the formation of superoxide anion, also a ROS (Hasegawa et al., 1990). Byproducts generated with the reaction of the superoxide anion with other ROS e.g. nitric
oxide (NO) could produce more damaging effect within the cell (Beckman, 1994). The

11


reaction of superoxide anion and NO produces peroxynitrite, a potent oxidant (Crow and
Beckman, 1995). To confirm the harmful effects of ROS generated by MPTP, factors that
could reduce the level of ROS within the cells were found to protect the cells against the
neurotoxicity of MPTP (Przedborski et al., 1992; Matthews et al., 1997; Grunewald and
Beal, 1999). In addition to the generation of superoxides and other potent ROS, MPTP was
found to potentiate the harmful effects of ROS by inducing a decrease in GSH content
(Desole et al., 1993). GSH is a thiol, an antioxidant capable of countering the harmful
effects of ROS. MPTP was also shown to increase the level of free iron in the SNpc
(Mochizuki et al., 1994). The free iron available could potentially catalyze the generation
of the toxic hydroxyl radicals from the ROS present. The hydroxyl radicals generated are
capable of causing severe damages to the neighboring membranes via lipid peroxidation.
The damage caused by ROS seems to induce the activation of common proteins such as
p53 and Bax. As like in the other three neurotoxins discussed, the levels of both proteins
were found to be elevated in response to MPTP/MPP+ treatment. Kitamura et al. (1998)
demonstrated a strong increase in p53 expression in MPP+-treated SH-SY5Y cells. In
mice, nigral Bax mRNA and protein levels were increased after MPTP exposure
(Hassouna et al., 1996). In addition, there was a rise in the activities of caspase-3, caspase8 and caspase-1 in the substantia nigra of MPTP-treated mice (Viswanath et al., 2000).
Thus, the data presented seems to suggest that apoptosis might be the degenerative
mechanisms involved in PD.
A study by Lotharius and O’Malley (2000) produced evidences for an
additional mechanism of MPTP toxicity. MPP+ was found to induce the redistribution of
dopamine within the neurons. Dopamine is usually stored inside vesicles located in the

12

cytoplasm. The vesicular displacement of dopamine into the surrounding cytosol could
result in more ROS being generated, thus leading to cellular damage and ultimately death.
MPP+ was also found to decrease the DNA content of the mitochondria to about one-third
in HeLa S3 cells (Miyako et al., 1997). The mitochondria genome is very compact and its
DNA encodes several components of the respiratory chain. A decrease in the DNA content
would greatly affect the normal function of the mitochondria. Recently, an endogenous
MPTP-like compound was found to be present in the cerebrospinal fluid of untreated PD
patients (Naoi and Maruyama, 1999). As such, exposure to endogenous or environmental
agents such as MPTP or other MPTP-like substances could well lead to pathological
effects such as those experienced in PD.
In summary, the action of MPTP on neuronal degeneration encompassed a
wide variety of mechanisms. Its effects not only mimic the selective degeneration of the
SNpc in animal models thus resulting in the behavioral and motor deficits, it also brings
together the various phenomenon observed in the different neurotoxins i.e. the effects of
dopamine and its redistribution within the cells; and the effects of a potent complex I
inhibitor and the subsequent oxidative stress generated. Thus, the use of MPTP can result
in all these effects using just one neurotoxin. On the other hand, MPTP is not at all the
omnipotent neurotoxin that could provide all clues to the understanding of PD. For one,
MPTP administration does not result in the formation of Lewy body, an intracellular
occlusion of protein aggregate found associated with many forms of PD. However, the
ultimate purpose of this study is to understand the process of cell death. MPTP hence
possess the ability to induce a condition where all the factors which may be critical in the

13


pathology of PD, can be reproduced in a single model. This significant advantage thus
makes it an excellent model to use for the specific study of neuronal cell death.

1.5

Apoptosis and Its Regulation
Even though the effects of the various neurotoxins are known, the mode of

cell death caused by it is still under much investigation. However, many studies have
favoured the postulation that the death of dopaminergic cells caused by neurotoxin
treatment is via apoptosis. Apoptosis is a term coined by Kerr et al (1972) to describe an
intrinsic cell suicide program. Apoptosis is a regulated mode of cell death that occurs
throughout development in all multicellular organisms. It is also involved in the death of
cells during diseases or when they are damaged. Cells dying via apoptosis show distinct
morphological characteristics such as chromatin condensation, nuclear fragmentation, cell
shrinkage, cytoplasmic condensation and apoptotic body formation. The apoptotic bodies
are then phagocytosed by macrophages, therefore preventing inflammation (Clarke, 1990).
On the other end of the spectrum, the process of necrosis describes an unordered mode of
cell death. Necrosis is characterized by the swelling of the cytoplasm and nucleus. The
plasma membrane finally burst and cellular contents are released into the surrounding,
resulting in the activation of the immune response. Since apoptosis has been implicated as
the mode of cell death in PD, it is thus helpful to understand the basic mechanisms
involved in the execution of this mode of cell death. This will in turn result in a deeper
knowledge of the factors affecting the apoptotic process in these diseased neuronal cells.

14


1.5.1

Caspases – The Mediators of Apoptosis
The execution of apoptosis requires the activities of a few main families of

proteins and caspases is one of them. Caspases is a distinct, highly conserved class of

intracellular cysteine protease. This family of proteins is characterized by their almost
absolute specificity for aspartic residues in their substrate (Earnshaw et al., 1999). They are
the subfamily of the interleukin-1β-converting enzyme (ICE). There are at least 14
members identified, including 11 human forms. Caspases are constitutively expressed as
inactive proenzymes (Weil et al., 1996). They are only activated upon reception of death
stimuli e.g. serum withdrawal. Activation of caspases requires the proteolytic cleavage at
the two sites within the protein (Nicholson and Thornberry, 1997). The released subunits
then assemble to form heterotetramers that are enzymatically active (Wilson et al., 1994).
Caspases are divided into two subgroups based on their functions in the
apoptotic pathways. The ‘initiator’ caspases consisting of caspase-1, -2, -4, -5, -8, -9, -10, 11, and -12 receive the initial death stimulus. The activation of this cascade of caspases
usually converges to activate a common set of the downstream ‘effector’ caspases. Effector
caspases (consisting of caspase-3, -6, -7 and -14) are directly responsible for the cleavage
of housekeeping proteins, causing death to the cells and also giving rise to the
morphological characteristics observed during apoptosis. They can also inactivate proteins
which inhibit apoptosis by cleaving them into fragments. In general, three different types
of death stimuli exist which are able to activate distinct sets of caspases within the cells.
These lead to the separate grouping of the caspases into three different apoptotic pathways.
The three pathways are namely the death receptor (Fas-ligand)-mediated, the

15


mitochondria-mediated, and the recently discovered endoplasmic reticulum (ER)-mediated
pathways. The three pathways are illustrated in Figure 1.

a) Death Receptor

b) Mitochondria

Death Ligand

c) Endoplasmic Reticulum
e.g. unfolded
protein

e.g. DNA damage

Ca2+
Plasma
Membrane

CD95

SOD

Bcl-2

Bax

Intracellular
membrane

Bak?

Cytochrome-c

?

FLIPS
Apaf-1

FADD

Caspase-12
FLIPL

Caspase-8

APOPTOSIS

Caspase-9

APOPTOSIS

APOPTOSIS

Figure 1. An illustration of the different proteins involved in the various apoptotic
pathways. Legend: SOD, silencer of death domain; FLIPS, short splice variant of Faslinked inhibitory protein; FLIPL, long splice variant; FADD, Fas-associated death domain.
(This figure is adapted from Daniel, 2000).
The death receptor-mediated pathway is activated when ligands such as the
tumor-necrosis factor (TNF) is bound to the death receptor found on the surface of the cell
(see Figure 1a). This leads to the recruitment of adaptor proteins such as FADD (Fasassociated death domain) to the receptor (Blagosklonny, 2000). The adaptor proteins in
turn recruit procaspases which are then activated to form the enzymatically active
heterotetramers. Caspase-8 and caspase-10 are the caspases recruited in this pathway, with
caspase-8 playing a more major role. Although the initial part of this pathway is mapped

16


out, the downstream proteins involved are still unclear. However, activation of this

pathway will eventually lead to cell death via apoptosis.
On the other hand, more information has been gathered regarding the
mitochondria-mediated pathway (see Figure 1b). The major mechanism by which the
mitochondria activate the apoptotic pathway is through the release of cytochrome-c from
the inter-membrane space of the mitochondria, into the cytosol (Green and Reed, 1998).
The cytosolic presence of cytochrome-c will result in the formation of the apoptosome,
consisting of Apaf-1, procaspase-9 and cytochrome-c itself. In the presence of ATP,
procaspase-9 is cleaved resulting in the activation of caspase-3 (Li et al., 1997). When this
occurs, apoptosis ensues. This particular pathway has been implicated as the apoptotic
pathway involved in the degeneration of neurons in several neurodegenerative diseases,
including PD.
The third subset of caspases are involved in the response to ER stress.
Conditions such as disturbed calcium homeostasis, malfolded proteins, and exposure to
free radicals can induce ER stress (see Figure 1c). Caspase-12, which is localized to the ER
is activated in response to stress in the ER (Nakagawa et al., 2000). As with the deathreceptor mediated pathway, the downstream members involved are not clear. However, the
end result of the activation of this pathway is the death of cells via apoptosis.

1.5.2

Mitochondria Factors – The Potential Initiators of Apoptosis
The importance of the mitochondria in apoptosis was observed in a cell-free

study where nuclear condensation and DNA fragmentation were found to be dependent on
the presence of this cellular organelle (Newmeyer et al., 1994). Within the mitochondria,
there exist certain factors which, when released into the cytosol, will initiate apoptosis.

17


This release of the mitochondria factors are stimulated when the mitochondria is damaged

beyond a certain threshold. The known factors that trigger the release include disruption of
electron transport, oxidative phosphorylation, and alteration of cellular reduction-oxidation
(redox) potential. Since the various neurotoxins discussed can inhibit complex I and lead to
the generation of ROS, the above factors stated may have been induced. The death events
of the dopaminergic neurons may thus be initiated by these mitochondria factors.
Among the mitochondria factors which are released is cytochrome-c.
Cytochrome-c, when present in the cytosol, forms an essential part of the vertebrate
‘apoptosome’ which is composed of Apaf-1, cytochrome-c itself and caspase-9. This
apoptosome complex can trigger the activation of both caspase-9 and caspase-3, leading to
apoptosis (Li et al., 1997). Another potent factor which can induce apoptosis is the
apoptosis-inducing factor (AIF). AIF is a protein which when released, may lead to the
activation of caspases (Fulda et al., 1998). Also, it can be translocated from the
mitochondria into the nucleus to induce DNA fragmentation. This action is independent of
caspases (Susin et al., 1999). The release of these two mitochondria factors is potent
enough to result in a full-scale apoptotic event, which can lead to the death of the cells
within hours.

1.5.3

The Guardians of Mitochondria – The Bcl-2 Family
AIF is a recently characterized mitochondria factor and the control of its

release has not yet been fully unraveled. However, clues of the release of cytochrome-c
and its regulation are somewhat known. The release of cytochrome-c and the subsequent
activation of the caspase-9/3 pathway is a critical step in the commitment of the cells to
apoptotic death. As such, this process is highly regulated, and the Bcl-2 family of proteins

18

is responsible for this regulation. The Bcl-2 family will determine if the apoptosome could
assemble by regulating the release of cytochrome-c from the mitochondria. Bcl-2 and BclXL have been shown to inhibit the activation of caspases by inhibiting the release of
cytochrome-c (Yang et al., 1997; Vander et al., 1997). Structural analysis of Bcl-2 and BclXL showed the similarity of these proteins to the pore-forming domains of diphtheria toxin
and colicins A and E1 (Muchmore et al., 1996). It was then postulated that these two
proteins could form channels in the mitochondrial membrane and prevent the release of
cytochrome-c from the mitochondria. The mechanisms of how the channels are formed are
not fully understood yet. Apart from these two anti-apoptotic members, other proapoptotic
family members do exist. These proteins function to promote the release of cytochrome-c,
thereby enhancing the apoptotic process. The proapoptotic members include Bax and Bad.
Bcl-XL was shown to sequester Bax, thus promoting cell survival (Oltvai et al., 1993).
However, upon reception of death stimulus, Bad accelerate apoptosis by displacing Bax
from Bcl-XL (Yang et al., 1995).
The Bcl-2 family of proteins thus possesses wide range of potential to
regulate apoptosis by influencing the release of mitochondria factors. However, even
though the molecular mechanisms underlying the pro- and anti-apoptotic functions of Bcl2 members have been under intense scrutiny, their functions are not entirely understood.
The current understanding is that a balance between the pro- and anti-apoptotic members
serves to provide a more refine mode of regulation. The regulation of apoptosis is thus a
complex network of interacting proteins which involves yet another protein family, namely
the IAP (inhibitor of apoptosis) protein family.

19


1.5.4

Keeping Caspases in Check - The IAP (Inhibitor of Apoptosis) Protein Family
The IAP family of proteins is a newly discovered family due to its homology

to the baculoviral apoptotic inhibitor (Clem and Duckett, 1997). All family members
possess a RING finger domain in their C-terminus and at least one BIR (baculovirus IAP

repeat) domain in the N-terminus (Uren et al., 1998). To date, six family members have
been identified, namely X-linked IAP (XIAP), HIAP1, HIAP2, NIAP, BIR-repeat
containing ubiquitin conjugating enzyme (BRUCE) and survivin. Studies on this family of
proteins have revealed some very powerful properties of these proteins in regulating
apoptosis. XIAP is one of such member. XIAP has been demonstrated to regulate both
initiator and effector caspases (Takahashi et al., 1998; Srinivasula et al., 2001). It is thus a
good therapeutic target to work on in relation to the various diseases that involves the
activation of apoptosis. The involvement of IAP in the death mechanism of dopaminergic
neurons in PD has not been demonstrated. However, with their unique anti-apoptotic
properties, their involvements are very highly possible.

1.5.5

Signaling Molecules – Kinases and Its Counterparts
Kinases, especially the c-jun N-terminal kinase (JNK) group, play a big role

in cell signaling in response to stress. JNK is a group of mitogen-activated protein (MAP)
kinases and is encoded by three different genes jnk1, jnk2 and jnk3 (Gupta et al., 1996).
Each of these three genes gives rise to alternative spliced products of 46kDa and 54kDa.
The jnk1 and jnk2 genes are ubiquitously expressed while jnk3 expression is found only in
the brain, heart and testis. JNK are activated by the phosphorylation of two residues found
within the protein by MEKK (MAP kinase kinase) (Kyriakis et al., 1994). The
phosphorylation of JNK seems to be somewhat related to the mitochondria. Early JNK

20


activation was found to be dependent on the mitochondrial adenine nucleotide translocator
activity (ANT) (Cassarino et al., 2000). The inhibition of ANT activity by bongkrekic acid,
a mitochondrial transition pore inhibitor, was able to attenuate JNK activity. Therefore, the

early activation of JNK is responsive to mitochondria damage, an event indicated to have
taken place in PD. The role of JNK in apoptosis is that it is able to antagonize the activity
of Bcl-2 possibly via phosphorylation (Maundrell et al., 1997; Park et al., 1997) and
activate p53 (Milne et al., 1995). These actions will lead to accelerated cell death via
apoptosis. Both of these phenomena have been observed in models used to study PD and
these events could potentially take place in the intrinsically-stressed dopaminergic neurons.
On the other hand, MEKK proteins are able to induce the phosphorylation and activation
of NF-κB, a protein capable of inducing the transcription of anti-apoptotic proteins such as
XIAP (Xiao et al., 2001). XIAP can inhibit the cleavage and activation of key caspases
such as caspase-3 (Takahashi et al., 1998). Thus, similar to the Bcl-2 family, the balance
between the activation of the various members and targets of the MAP kinase family of
proteins plays a part in regulating the commitment of the cells to apoptosis.

1.6

Is Apoptosis Really Involved in PD?

1.6.1

Evidences from Human Postmortem Tissues
The data from human postmortem tissues have led to several evidences for

the role of apoptosis in PD. Both Mochizuki et al. (1996) and Kingsbury et al. (1998)
showed the presence of DNA fragmentation in the substantia nigra of PD patients using the
TUNEL assay, a method that detects apoptosis. Kingsbury et al. (1998) also observed
concurrent characteristics of apoptosis such as chromatin condensation and irregular
nuclear morphology. Staining of the postmortem tissues for the detection of key players of

21

apoptosis (i.e. activated effector caspases e.g. caspase-3) was also a method of choice.
Hartmann et al. (2000) managed to show a significant decrease of caspase-3-positive
pigmented neurons in the SNpc of PD patients, hence indicative of their death through the
apoptotic pathway.
However, not all studies on postmortem tissues were able to show the
presence of apoptotic characteristics. Both Banati et al. (1998) and Kosel et al. (1997)
failed to show any morphological signs of apoptosis in the tissues of their subject. It has
been proposed that many parameters, such as 1) the speed of the apoptotic process (usually
occurring between 1-2 hours); 2) the time lag between the death of the patient and the
fixation of the tissues; and 3) the use of formalin or paraffin embedded tissues, could
influence the detection of apoptotic cells (Banati et al., 1998). Furthermore, it was
postulated that the degeneration of the dopamine neurons might be much earlier than the
onset of the disease. By the time the patients die from PD, most of the vulnerable cells
would have degenerated. If the mechanism of cell death were indeed apoptosis, the
phagocytosis of the apoptotic bodies would leave no clues. Most importantly, visualization
of these apoptotic characteristics does not reveal anything about the molecular events
leading to apoptosis. Considering the difficulty in extracting and interpreting data from
human tissues, animal models thus come as the next best choice for the study of PD.

1.6.2

Clues from Animal Models
Using MPTP treated mouse model, Tatton and Kish (1997) demonstrated the

presence of apoptotic nuclei in the SNpc. Further evidences of the involvement of the
apoptotic pathway were produced by Vila et al. (2001). This particular study described the
induction of nigral Bax mRNA in mice after MPTP exposure. Overexpression of Bcl-2

22

was able to attenuate the neurotoxicity of MPTP in mice (Yang et al., 1998). The
activation of Bax and Bcl-2 indicate the involvement of the mitochondria-mediated
apoptotic pathway. In line with this, Viswanath et al. (2000) was able to show a rise in
caspase-3 activities in MPTP treated mice. Caspase-8 and caspase-1 were also shown to
have increased activities. The use of other neurotoxins such as 6-OHDA in vivo also
showed the presence of apoptotic cells. 6-OHDA-treated rats was found to be TUNEL
positive upon one to fourteen days after 6-OHDA administrations (He et al., 2000). Marti
et al. (1997) was also able to demonstrate nuclear condensation within the brains of 6OHDA treated rats. All these findings seem to suggest the possible involvement of the
apoptotic pathway in the selective degeneration of the dopaminergic cells in the SNpc.

1.6.3

More Evidences from Cell Models
The amount of work done using cell models for the study of MPTP/MPP+ in

relation to PD is huge. The commonly used cell lines include the mouse neuroblastoma
N2a, the human neuroblastoma SH-SY5Y, and the dopaminergic cancer cell line of rat
origin, PC12. Evidences from the use of cell models largely point to the involvement of the
apoptotic pathway in the degenerating cells, although contradicting evidences do exist.
Gomez et al. (2001) managed to demonstrate the activation of caspase-9 and a slight
increase in cytosolic cytochrome c. In the same study, MPTP-induced cell death was
attenuated using broad caspase inhibitor. Several other evidences also seem to support the
involvement of mitochondria-mediated apoptotic pathways. Studies specifically targeted at
the mitochondria, especially in the area of the mitochondria transition pore (MTP), were
also done. The opening of MTP has been implicated with the release of cytochrome-c
during apoptosis (Cassarino et al., 1999). In line with this, blockage of the MTP opening

23

using cyclosporin A was shown to reduce MPP+-induced apoptosis (Seaton et al., 1998).
Likewise, the involvement of caspases was also demonstrated in primary dopaminergic
neuronal cultures (Du et al., 1997). Even though many more of such studies have provided
evidences to the involvement of the mitochondria-mediated apoptotic pathways in MPTP
models, contradicting results do exist. Recent work by Hurelbrink et al. (2001)
demonstrated that the use of caspase-3 inhibitors did not promote cell survival in vitro,
suggesting that apoptosis might not be the major determinant of neuronal cell death. In
addition, electron microscopic studies on an MPP+-challenged cell model revealed that the
cells might be dying via necrosis instead of apoptosis (Choi et al., 1999a).
The use of 6-OHDA also produced evidences of caspase activation. A study
by Ochu et al. (1998) demonstrated the presence of caspase-3-like activities. In addition,
the cleavage of poly-ADP-robose polymerase, one of the targets of caspase activity, was
also reported in SH-N-SH cells after exposure to 6-OHDA (Bruchelt et al., 1991).
Rotenone, as mentioned earlier, was also able to induce the activation of caspase-12,
caspase-9 and caspase-3 upon exposure of SH-SY5Y cells to this neurotoxin (Kitamura et
al., 2002).
The evidences presented so far have indicated the presence of apoptotic
characteristics and the activation of proteins involved in apoptosis, although there are
exceptions. Thus, even though the effects of the neurotoxicity of the various neurotoxins
are well studied, the molecular mechanisms involved remained ambiguous. As such, more
work is needed to gain more insights into the actual mode of cell death undertaken during
neuronal degeneration in PD.

24


1.7

Objective of Present Study
Since the mode of cell death (i.e. apoptosis or necrosis) is still not defined

with certainty, the first objective of the present study is to try to determine the mode of cell
death using a cell model challenged with MPP+. The cell line employed in this study is a
mouse dopaminergic neuronal cell line of mesencephalic origin, MN9D. Simple LDH
(lactate dehydrogenase) assay will be used to test for the presence of necrosis. Activation
of the apoptotic pathway will be assayed using antibodies specific to the caspases
involved. If apoptosis is proven to be the mode of cell death, cell biology studies such as
the use of inhibitors against the various caspases will be used to map out the main
apoptotic pathway involved. The sequential activation of the caspases will also be studied
to map out the pathway involved in the apoptotic event. On the other hand, if necrosis is
the mode of cell death, the stress status of the cell will be studied via the use of antibodies
against the various stress indicators such as ERK1/2, p38 and JNK. This will serve to give
more insights into involvement of the stress signaling pathways.
The hypotheses to be tested in this study are: (1) the death of MN9D cells
exposed to MPP+ is via apoptosis which is caused by mitochondria damage and the
resulting ATP depletion; and (2) the apoptotic death is via the mitochondrial-mediated
apoptotic pathway.

25