Animal Models of Parkinson´s Disease

Introduction

Parkinson’s disease (PD) was the first neurological disease to be modelled in animals. Early models of PD used toxins which selectively targeted dopaminergic neurons, such as reserpine, 6-hydroxydopamine, and 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine. These initial models have greatly contributed to the current understanding of the pathogenesis of PD and have proven to be valuable tools in the development of novel therapeutic approaches, but have failed to mimic important characteristics of PD. Recently, it has been found that chronic systemic exposure to the pesticide rotenone can reproduce specific features of PD in rodents. Moreover, the association of α-synuclein mutations with some cases of familial PD have motivated the development of genetic models of PD in mice and Drosophila. The present essay gives a brief survey of the clinics and pathophysiology of PD, discusses the different animal models of PD currently available, and briefly compares the suitability the rodents and primates as models for human PD.

Parkinson’s disease

Parkinson’s disease (PD), first described by James Parkinson in 1817, is the second most common neurodegenerative disease, affecting about 120.000 people in the UK, with approximately 10.000 newly diagnosed cases each year, the vast majority of them in elderly people. PD is a chronically progressive disorder which develops gradually until the clinical symptomatology becomes manifest. The cardinal symptoms in fully developed parkinsonism comprise rest tremor, rigor, and bradykinesia. Further clinical manifestations are gait abnormalities and postural instability. Moreover, cognitive and psychological impairments such as dementia and depressive symptomatology have been reported in about 30% (Aarsland et al. 1996) and 40% (Cummings 1992) of PD patients respectively. The primary causal mechanism underlying these deficits is a progressive degeneration of the nigrostriatal dopaminergic (DA) pathway (see Fig.1) However, the full clinical expression of parkinsonism does not occur until the striatal DA level is reduced by about 80% (Hornykiewicz 1993) (see Fig.2). Dopamine depletion below this level is counterbalanced by compensatory mechanisms, such as increased activity of the remaining DA neurons, or an elevated expression of postsynaptic DA receptors (Deumens et al. 2002).

illustration not visible in this excerpt

Another pathological feature of PD is the occurrence of intracellular, eosinophilic inclusions, so called Lewy-bodies, a major component of which is a protein called α-synuclein. Although PD is usually considered a disorder specific to nigral DA neurons, it has been demonstrated that neuronal degeneration in PD patients also occurs in other DA systems, such as the ventral tegmental area (Agid et al. 1990), as well as in non-dopaminergic cell populations, including noradrenergic neurons of the locus coeruleus and dorsal vagal nucleus, serotonergic neurons of the dorsal raphe, and the cholinergic pathway from the nucleus basalis Meynert (see Jellinger 1990, for a review). These non-nigral lesions are considered to account for the cognitive and psychological aspects of PD.

illustration not visible in this excerpt

Although the aetiology of PD remains unknown, there is evidence for the implication of oxidative stress and dysfunction of complex I of the mitochondrial respiratory chain in the pathogenesis of PD (Betarbet 2002). Both environmental and genetic factors have been suggested to underlie PD. The idea of an environmental contribution to the evolvement of PD derived from the discovery of the toxin 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP), which causes selective degeneration of nigral DA neurons in humans, associated with motor disturbances characteristic for PD. Other toxins that have been shown to induce nigral DA depletion include carbon monoxide (Choi and Cheon 1999), β-carolines (Collins and Neafsey 2002), and rotenone (Betarbet et al. 2000). On the other hand, rare cases of familial PD have recently been linked to point-mutations of the α-synuclein gene (Mashlia et al. 2000)^[1], indicating that there may as well be a genetic contribution to the pathogenesis of PD^[2], though genetic mutations cannot today explain sporadic and late onset cases (Tanner et al. 1999).Taken together, these findings suggest that various exogenous and endogenous^[3] compounds can induce PD when a susceptible background in the form of a genetic predisposition exists.

Animal models of Parkinson’s disease: General considerations

Why use animal models?

Animal models are important for two reasons: to study the pathogenic mechanisms underlying PD and to develop novel therapeutic approaches. For instance, the early reserpine-model of PD provided first evidence for a linkage of striatal DA depletion with parkinsonian motor symptomatology (Carlsson et al. 1957), and enabled the discovery of the therapeutic effects of the dopamine precursor levodopa (L-DOPA), which still represents the gold-standard in the treatment of PD (Carlsson et al 1957). Therefore, the development of valid animal models of PD have assisted (a) to enable a deeper understanding of the pathophysiology of PD, and (b) discovering novel therapeutic strategies for PD.

The ideal animal model of PD

In order to allow for a maximal transferability of the insights derived from experimental models to human PD, an experimental model of PD should simulate both the neuropathic and phenotypic alterations observed in human PD. According to Beal (2001), the ideal animal model for PD should exhibit the following features: (1) a normal development of DA populations at birth that show progressive degeneration in adulthood, (2) the major motor deficits of PD, including rigor, tremor at rest, and bradykinesia, (3) the development of characteristic Lewy-bodies, (4) in case of a genetic model, it should be based on a single mutation, to allow for a straightforward regulation of the genes by expression under the control of particular enhancer or suppressor genes, (5) a relatively short disease course in the range of months, allowing for a rapid screening of therapeutic agents. As yet, none of the animal models available exhibit all of these features. Rather, each animal model is valuable only to the extent to which it accurately mimics the pathogenic, histological, biochemical, and clinical features which are currently in the interest of the investigator.

The reserpine model

Reserpine causes a depletion of monoamines in the brain (Carlsson et al 1957) and induces transient hypokinesia and muscular rigidity in rodents (Colpaert 1987), thus providing a pharmacological model of PD. Moreover, it has been shown that these symptoms can be abolished by administration of L-DOPA, indicating a causal relationship between DA depletion in the brain and the motor deficits observed in the reserpine-treated animals. It is now clear that reserpine interferes with the storage of monoamines in presynaptic intracellular vesicles (Carlsson 1975), resulting in a depletion of monoamines in nerve terminals and the induction of a transient hypokinetic state. Moreover, the reserpine model has provided an indication of the mode of action of DA-releasing drugs^[4]. To date, the majority clinically used antiparkinsonian drugs, such as amantadine, trihexiphenidyll, L-DOPA and DA receptor agonists have been demonstrated to ameliorate motor deficits in this model (Menzaghi 1997). However, a major drawback of the reserpine model is that the motor deficits are temporary, and there is no damage to the nigrostriatal tract.

The 6-hydroxydopmanine model

General considerations

6-hydroxydopamine is the hydroxylated analogue of the natural DA neurotransmitter and was the first agent discovered that had specific neurotoxic effects on catecholaminergic agents. Since systematically administered 6-OHDA is unable to cross the blood-brain barrier, it must be injected directly into the nigrostriatal pathway. Although 6-OHDA-induced lesions have been applied to mice, cats, dogs, and monkeys, rats are most commonly used because of established stereotactic techniques. Sites used for injections are the substantia nigra pars compacta (SNPC), the medial forebrain bundle (MFB), or the striatum (Deumens et al. 2002). Unilateral 6-OHDA induced lesions are usually extensive, reliably resulting in the following well-described symptomatic: (1) an extensive DA depletion in the ipsilateral striatum of about 80-90%, (2) denervation supersensitivity of the postsynaptic DA receptors in the ipsilateral striatum, and (3) a characteristic turning behaviour in response to both D-amphetamine and apomorphine (Deumens et al. 2002).

Mechanism of action

6-OHDA is taken up by both DA and noradrenergic (NE) via a transport mechanism that is selective for DA or NE, respectively^[5]. Within the neuron, 6-OHDA causes degeneration of the nerve terminals, and can also affect the cell body (see Fig.3). Its neurotoxicity is assumed to be mediated by several effects. It has been shown that 6-OHDA produces oxidative stress in vivo as well as in vitro, either due the generation of hydrogen peroxide and derived hydroxyl radicals or via a auto-oxidation process that leads to the formation of several toxic species, such as quinones (see Blum et al. 2001, for a review). The subsequent oxidative stress leads to a reduction in cellular antioxidative capabilities, impairs intracellular redox potential regulation, and causes lipid peroxidation. Moreover, 6-OHDA inhibits complex I and complex IV of the mitochondrial respiratory chain, leading to decreased ATP levels and, more importantly, to the production of superoxide free radicals^[6] (Glinka and Youdim 1995).

Motor deficits

Following unilateral 6-OHDA-induced lesions, rats initially tend to turn toward the site of the lesion. Administration of drugs acting on the DA system leads to an active rotational behaviour due to a physiologic imbalance between the lesioned and the unlesioned striatum, with the animal rotating away from the side of greater activity. Accordingly, subcutaneous administration of the DA-releasing agent D-amphetamine creates a DA imbalance that favours the unlesioned striatum and thus produces ipsilateral rotations. In contrast, the postsynaptic agonist apomorphine induces rotations to the lesioned side due to stimulation of denervation-induced upregulated D2 receptors in the denervated striatum. This turning behaviour can be quantified and correlates with the degree of the lesion (Ungerstedt 1968)^[7]. This allows a convenient and reliable control of the extent of the lesion and of the effects of therapeutic treatment^[8]. 6-OHDA lesions have also been shown to affect place (cognitive) and cue (sensorimotor) orientation in the Morris water maze task, which might be seen as a correlate to the cognitive impairments in some human PD cases (Wishaw and Dunnett 1985).

The fact that the degree and severity of the lesion correlates with the rotational behaviour makes the model particularly suitable for testing the efficacy of novel antiparkinsonian compounds, and also for examinating the effects of DA cell transplantation (Betarbet et al. 2002).

Drawbacks of the 6-OHDA model

While 6-OHDA induces nigrostriatal degeneration, it does not affect non-dopaminergic cell populations which are commonly affected in PD patients, such as the locus coeruleus or the nucleus basalis Meynert, nor does it reproduce the characteristic Lewy-bodies. Moreover, 6-OHDA leads to an acute loss of DA neurons, occurring within 2-3 days after administration (Faull et al. 1969), which is in marked contrast to the progressive and gradual degeneration of DA neurons seen in idiopathic PD. Finally, the 6-OHDA model differs from idiopathic PD in that it is subject to compensatory mechanisms, which is especially true for the unilateral model where the sprouting of axons from the intact site to the contralateral may serve as a compensatory mechanism (Van Oosten and Cools, 1999).

Despite these limitations, the 6-OHDA model has provided a valuable tool to investigate the efficacy of potentially antiparkinsonian drugs (Schwartig and Huston 1996). Furthermore, it has been used to examine the effects of cell transplantation, and for testing neurotrophic factors which promote survival of the degenerating DA neurons.

[...]

^[1] However, most neurodegenerative disorders with Lewy-bodies are associated with abnormal accumulation of wild-type, not mutant, α-synuclein (Mashlia 2000).

^[2] PD has also been associated with parkin and UCH-L1 mutations, indicating that the ubiquitin and the proteasome pathways may be relevant to the evolvement of Parkinson's disease.

^[3] For instance, studies reported the presence of the neurotoxin 6-OHDA in both rat (Senoh and Witkop 1959) and human (Curtius et al. 1974) brain, which is probably due to a non-enzymatic reaction between DA, hydrogen peroxide and free iron (Jellinger et al. 1995; Linert et al. 1996).

^[4] Compounds that release DA from vesicular stores are less active in the reserpine model than in normal rats, reflecting the decreased availability of DA, whereas compounds that release DA from cytoplasmic stores maintain a strong activity in the reserpine model.

^[5] Selectivity for DA neurons is achieved by pre-treatment with desimipamine, a NE transporter blocker, which selectively inhibits the uptake of 6-OHDA into NE neurons.

^[6] It has been shown that complex I activity should be reduced to >70% to cause severe ATP depletion (Davey and Clark, 1996), and that MPTP in vivo causes only a transient 20% reduction in murine striatal and midbrain ATP levels (Chan et al. 1991), raising the possibility that formation of ROS rather than MPP+-related ATP deficit accounts for the MPTP-induced dopaminergic neuronal death.

^[7] Low doses of apomorphine rather than D-amphetamine are more sensitive to the size of the lesions, making apomorphine a better predictor of the extent of the lesion (Hudson et al. 1993).

^[8] An alternative measure has been introduced by Chang et al. (1999), who used an adjusted steps task rather than drug-induced rotational-behaviour as an indicator of nigrostriatal DA-depletion. This task is sensitive to submaximal lesions in which striatal DA levels are decreased by about 60-80%, while extensive lesions of the striatum (> 90% loss of DA fibre density) and concomitantly SN (>50% loss of DA neurons) are believed to be required to demonstrate rotational behaviour (Hudson et al. 1993).