Animal models of Parkinsons disease - Wiley Online Library

MINIREVIEW

Animal models of Parkinson’s disease Fabio Blandini1 and Marie-Therese Armentero1,2 1 Interdepartmental Research Center for Parkinson’s Disease, IRCCS National Neurological Institute C. Mondino, Pavia, Italy 2 Laboratory of Functional Neurochemistry, IRCCS National Neurological Institute C. Mondino, Pavia, Italy

Keywords Lewy bodies; MPTP; non-mammalian models; Parkinson’s disease; pesticides; substantia nigra; toxic models; transgenic models; 6-OHDA; a-synuclein Correspondence F. Blandini, IRCCS National Neurological Institute C. Mondino, Via Mondino 2, 27100 Pavia, Italy Fax: +39 0382 380448 Tel: +39 0382 380416 E-mail: [email protected] (Received 12 October 2011, revised 19 December 2011, accepted 12 January 2012) doi:10.1111/j.1742-4658.2012.08491.x

Animal models of Parkinson’s disease (PD) have been widely used in the past four decades to investigate the pathogenesis and pathophysiology of this neurodegenerative disorder. These models have been classically based on the systemic or local (intracerebral) administration of neutoxins that are able to replicate most of the pathological and phenotypic features of PD in mammals (i.e. rodents or primates). In the last decade, the advent of the ‘genetic era’ of PD has provided a phenomenal enrichment of the research possibilities in this field, with the development of various mammalian (mice and, more recently, rats) and non-mammalian transgenic models that replicate most of the disease-causing mutations identified for monogenic forms of familial PD. Both toxic and transgenic classes of animal PD models have their own specificities and limitations, which must be carefully taken into consideration when choosing the model to be used. If a substantial and reproducible nigrostriatal lesion is required (e.g. for testing therapeutic interventions aimed at counteracting PD-related cell death), a classic toxic model such as one based on the administration of 1-methyl-4-phenyl1,2,3,6-tetrahydropyridine or 6-hydroxydopamine will adequately serve the purpose. On the other hand, if selected molecular mechanisms of PD pathogenesis must be investigated, transgenic models will offer invaluable insights. Therefore, until the ‘perfect’ model is developed, indications to use one model or another will depend on the specific objectives that are being pursued.

Introduction Although almost two centuries have passed subsequent to the original description by James Parkinson, the exact cause of Parkinson’s disease (PD) remains unknown. A restricted number of converging pathogenic mechanisms, including oxidative stress, mitochondrial defects, proteolytic stress and neuroinflammation, have been suggested for sporadic PD, the most frequent form of the disease. More recently, the existence of familial

(genetic) forms of PD, accounting for 10–15% of diagnosed PD cases, has been recognized. This not only made the general picture of the disease more complex, but also generated new perspectives for a deeper understanding of the molecular mechanisms of the disease, which may translate into new therapeutic options. Most of our current knowledge on the potential pathogenic and pathophysiological mechanisms of PD

Abbreviations 6-OHDA, 6-hydroxydopamine; BBB, blood–brain barrier; DA, dopamine; DAT, dopamine transporter; KO, knockout; LB, Lewy body; L-DOPA, 3,4-dihydroxyphenylalanine; LPS, lipopolysaccharide; LRRK2, leucine-rich repeat kinase 2; MPP+, 1-methyl-4-phenylpyridinium; MPTP, 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine; PD, Parkinson’s disease; PINK1, PTEN-induced putative kinase 1; ROS, reactive oxygen species; SNpc, substantia nigra pars compacta; TH, tyrosine hydroxylase.

1156

FEBS Journal 279 (2012) 1156–1166 ª 2012 The Authors Journal compilation ª 2012 FEBS

F. Blandini and M.-T. Armentero

derives from innumerable studies conducted, in the past four decades, on experimental models of PD. In vivo (animal) models, in particular, have not only provided (and are still generating) invaluable information, but also the possibility of testing innovative therapeutic approaches. Such models have been classically based on the use of neutoxins that are able to replicate most of the pathological and ⁄ or phenotypic features of PD in mammals (i.e. rodents or primates); in the last decade, the advent of the ‘genetic era’ of PD has provided a phenomenal enrichment of the experimental possibilities, with a number of transgenic models that have been made available to the scientific community. Both classes of animal PD models (i.e. toxic and transgenic) have their own specificities and limitations; indication to use one model or the other, therefore, depends on the specific objectives that are being pursued.

Toxic animal models of PD Toxic models represent the classic (and oldest) experimental PD models; they aim to reproduce the pathological and behavioural changes of the human disease in rodents or primates by using pharmacological agents (neurotoxins) that induce the selective degeneration of nigrostriatal neurones. These toxins can be administered either systemically or locally, depending on the type of agent used and the species involved. Below, we describe the main toxic models that are currently in use.

Animal models of Parkinson’s disease

Systemic administration 1-Methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) A classic systemic model is based on the administration of MPTP, with selective toxicity for dopaminergic neurones (Fig. 1). The selective toxicity of MPTP for the nigrostriatal tract was recognized in the mid1980s, after Langston et al. [1] described, in young drug users from Northern California, the occurrence of marked parkinsonism caused by the intravenous injection of a street preparation of an analogue of the narcotic meperidine containing MPTP. After crossing the blood–brain barrier (BBB), MPTP is transformed by monoamine oxidase B into its active metabolite, 1-methyl-4-phenylpyridinium ion (MPP+) (Fig. 1), which is then carried by the dopamine (DA) transporter (DAT) into dopaminergic neurones of the substantia nigra pars compacta (SNpc), where it blocks mitochondrial complex I activity. This discovery indicated that the mitochondria of dopaminergic neurones were a preferential target of toxicity, paving the way to myriads of studies exploring mitochondrial function in PD in the subsequent decades, and providing a formidable animal model of PD. When administered to primates through repeated bilateral intra-carotid injections, MPTP causes a 3,4-dihydroxyphenylalanine (l-DOPA) responsive parkinsonian syndrome, characterized by all of the cardinal symptoms of PD, which represents the best PD-like clinical picture obtainable in experimental animals. MPTP can also be administered to rodents. Mice are used, in this case, because

Fig. 1. Structures of dopamine and major dopaminergic toxins used to replicate features of Parkinson’s disease in animal models.


1157



rats are highly resistant to MPTP, for reasons that are not yet clearly understood. Indeed, rats are sensitive to MPP+, although only if injected directly into the SNpc. In mice the toxin, which is usually administered intraperitoneally through repeated bolus injections over a short period of time (i.e. four MPTP doses, every 2 h over 1 day), substantially causes the same selective lesions seen in MPTP-treated monkeys. A major limitation of the MPTP model, which is shared by the toxic models in general (possibly with one exception; see below), is that the SNpc lesion is not accompanied by the formation of Lewy body (LB)-like cytoplasmic inclusions. Therefore, a crucial neuropathological hallmark of PD would be missing. Various studies have tried to solve this problem by modifying the MPTP treatment regimens, with controversial results being obtained. Fornai et al. [2] reported that a 30-day administration of MPTP (30 mgÆkg)1Æday)1) via osmotic minipumps induced the formation of nigral inclusions immunoreactive for ubiquitin and a-synuclein; Shimoji et al. [3] failed to identify LB-like inclusions in mice treated either with acute (1 day), semi-chronic (5 days) or chronic (35 days) regimens of MPTP treatment, although they did not use osmotic minipumps for continuous delivery. Analogously, Alvarez-Fischer et al. [4] failed to identify LB-like inclusions or even substantial SNpc cell loss in mice chronically treated with MPTP (40 mgÆkg)1Æday)1), this time using minipumps. Moderate neuronal loss, still in the absence of LB-like pathology, occurred only when MPTP was co-administered with probenecid, an uricosuric agent that potentiates the effects of the toxin by blocking its clearance from the circulation. Indeed, chronic coadministration of MPTP with probenecid (ten doses over 5 weeks) has been proposed to overcome another limitation of the MPTP procedure. Indeed, the dopaminergic nigrostriatal deficits obtained with acute (four injections over 1 day) or sub-acute administration of MPTP (single daily dose for 5–10 days) tends to be reversible. The addition of probenecid potentiates the effects of MPTP, allowing the gradual loss of SNpc neurones associated with a substantial loss of striatal DA and DA uptake, which lasts for at least 6 months after withdrawal from treatment. Pesticides Another major merit of the MPTP model involved drawing attention to the potential role of environmental toxins in the pathogenesis of PD. Structural similarities were noted between MPTP, MPP+ and certain pesticides, in particular paraquat (in the 1960s, MPP+ 1158

was even tested as an herbicide under the name of cyperquat) (Fig. 1). Subsequent epidemiological analyses confirmed that subjects with a history of chronic exposure to pesticides have an increased risk of developing PD. This observation led to numerous groups exploring the ability of agricultural pesticides to induce neuropathological modifications that may be reminiscent of the human disease, thereby generating additional toxic models of PD. Betarbet et al. [5] described a new model based on chronic (5 weeks), intravenous administration of rotenone to rats. Rotenone (Fig. 1) is a flavonoid found in the roots and stems of several plants and used as a broad-spectrum pesticide [5]. Being highly lipophilic, rotenone easily crosses the BBB and, unlike MPP+, does not depend on DAT to enter dopaminergic neurone. Once in the cell, rotenone blocks complex I activity, causing massive formation of reactive oxygen species (ROS), and inhibits proteasome activity, thereby generating proteolytic stress. In rats, rotenone caused the selective degeneration of nigrostriatal dopaminergic neurones and, for the first time in the field of PD toxic models, LB-like cytoplasmic inclusions containing ubiquitin and a-synuclein. The development of a behavioural phenotype characterized by hypokinesia, hunched posture and severe rigidity was also observed. Similar changes were achieved using subcutaneous or intraperitoneal routes of administration. The rotenone model has generated great resonance in the field of experimental PD and provided crucial insights into the mitochondria-related mechanisms of neurodegeneration in PD. On the other hand, a number of critical issues have been highlighted. For example, there is a substantial variability in the individual response to the toxin, which translates into substantial oscillations in the percentage of animals developing a satisfactory nigrostriatal lesion; this makes the rotenone model unsuitable if a reproducible and standardized lesion is required (e.g. to test potentially neuroprotective treatments). Mortality is another issue, with rotenone being highly toxic to various organs, including the heart, liver, kidneys and gastrointestinal tract. A third criticism is related to the specificity of the rotenone-induced lesions, which have been questioned on the basis that the degenerative process may involve other neuronal populations within the basal ganglia, thereby inducing a pattern of pathological changes more similar to atypical parkinsonism, rather than PD. Based on these considerations, the rotenone model remains a milestone in the field of PD toxic models, although a higher degree of reproducibility across laboratories would favour a more widespread adoption of this procedure.



Another systemic model of PD is based on the administration of paraquat. As noted above, paraquat, which is a nonselective bipyridyl contact herbicide, shares structural similarities with MPP+, although the mechanisms of action are quite different. Being a charged molecule, paraquat does not cross the BBB; it may therefore enter the brain via a neutral amino acid transporter, such as the system L carrier (i.e. LAT-1); a sodium-dependent transport mechanism would carry paraquat inside the neurone, independently of DAT. At the cytosolic level, paraquat generates massive oxidative stress by acting as a redox cycling compound (via the formation of superoxide anion), as well as by impairing recycling of oxidized glutathione to its reduced form, which hampers the efficiency of intracellular antioxidant systems. Complex I blockade does not appear to play a significant role in paraquatinduced neurotoxicity because this toxin has low affinity to mitochondrial complex I and only at high doses. More recently, a pro-apoptotic activity involving mainly the intrinsic mitochondrial pathway has been shown for paraquat, leading to the induction of pro-apoptotic proteins Bax and Bak, followed by cytochrome C release and caspase-9 activation [6]. Paraquat is highly selective for nigrostriatal dopaminergic neurones, although cell loss is generally moderate (20–30%) and detectable only after multiple injections. Reduced motor activity and dose-dependent loss of striatal dopaminergic nerve fibres have been reported in paraquat-treated mice, without a substantial decrease of striatal dopamine; increased expression and aggregation of a-synuclein have also been detected in the SNpc of these animals [7]. Subsequent studies have shown that paraquat toxicity can be exacerbated by the simultaneous administration of fungicide manganese ethylenebis(dithiocarbamate). When administered together to mice, intraperitoneally, in a subchronic fashion, the two toxins induce synergistic effects at the nigrostriatal level, causing more profound neuronal death in the SNpc and deficits in dopaminergic neurotransmission [8]. Local administration 6-hydroxydopamine (6-OHDA) The prototypical model based on local (i.e. intracerebral) injection of a neurotoxin is the 6-OHDA model, which was also the first PD animal model ever generated [9]. 6-OHDA is a hydroxylated analogue of DA with high affinity for DAT, which transports the toxin inside dopaminergic neurones (Fig. 1). Local injection is required because 6-OHDA does not cross the BBB. After injection into the SNpc or, preferably, into the


medial forebrain bundle that conveys the efferent fibres from nigral cell bodies to the striatum, 6-OHDA causes massive anterograde degeneration of the nigrostriatal pathway. SNpc neurones begins to die within the first 12 h post-injection, whereas marked lesion of striatal dopaminergic terminals, paralleled by DA depletion, is established within 2–3 days. This procedure grants the highest level of nigral cell loss and striatal DA depletion obtainable in PD animal models (90–100%). The injection is commonly carried out unilaterally, with the contralateral hemisphere serving as control, because a high mortality rate is associated with bilateral injections. In the mid-1990s, a variant of the original procedure was proposed, in which 6-OHDA is injected into the striatum [10]. This induces prompt damage of striatal terminals, followed by delayed, progressive cell loss of SNpc neurones, which are secondarily affected through a ‘dying back’ mechanism. The degree of SNpc damage obtained with this procedure is less marked compared to the intra-medial forebrain bundle injection, remaining confined to 50–70% of the nucleus, and evolves over a period of 4–6 weeks. This alternative modality thereby provides a progressive model of nigrostriatal degeneration, which is more similar to the gradual evolution of the neurodegenerative process of human PD [11]. The mechanism of action of 6-OHDA is substantially related to its prooxidant properties. Once in the neurone, 6-OHDA accumulates in the cytosol and undergoes prompt auto-oxidation, promoting a high rate of hydrogen peroxide formation. As an additional mechanism, 6-OHDA can accumulate in the mitochondria, where it inhibits complex I activity. The lesion obtained with 6-OHDA is highly reproducible, which represents a considerable added value when new therapeutic strategies are to be investigated and clear neuroprotective effects must be demonstrated. The unilateral lesion caused by 6-OHDA induces typical stereotypies (i.e. turning behaviour) in response to drugs that are able to stimulate striatal DA receptors, either directly, such as apomorphine, or indirectly, such as amphetamine, which prompts the release of DA from striatal terminals of the intact hemisphere. In the first case, animals will rotate contralaterally to the lesioned side, whereas, in the other case, they will rotate ipsilaterally. Clear spontaneous limb asymmetries, as a result of motor impairment of limbs contralateral to the injected hemisphere, are also present and can be evaluated by using various behavioural tests. The motor impairment affecting 6-OHDA infused animals is permanent only in animals with a complete nigrostriatal lesion (> 95%); indeed, partial recovery of function,


1159



evolving in the 2–3 months after the 6-OHDA infusion, can be observed in animals bearing partial lesions of the nigrostriatal tract. Lipopolysaccharide Inflammation has recently emerged as a key player in PD pathogenesis and all neurotoxins currently used in experimental models generate a neuroinflammatory response in the nigrostriatal tract. These considerations have prompted the introduction of another toxic model, in which bacterial endotoxin lipopolysaccharide (LPS), which causes intense tissue inflammation, is directly infused into the nigrostriatal pathway of rats. Intranigral injection of LPS results in activation of microglia and degeneration of the dopaminergic system. LPS-induced neurotoxicity is mediated by the microglial activation and the resulting release of cytotoxic molecules because LPS per se is not toxic to neurones. More recently, Hunter et al. [12] have shown that the injection of LPS into the striatum of rats induces progressive degeneration of the nigrostriatal pathway, characterized by a 41% cell loss in the injected SNpc, at week 4 post-injection, and a 42% reduction in the striatal levels of DA; accumulation of a-synuclein and ubiquitin in surviving SNpc neurones was also reported, along with marked rotational behaviour, ipsilateral to the lesioned side, in response to the systemic administration of amphetamine. Transgenic animal models of PD Although toxic models of PD have provided an invaluable bulk of information on disease pathology, the lack of an age-dependent, slowly progressive lesion and the fact that LBs are typically not observed in these models represent major drawbacks. The discovery of monogenic ‘Mendelian’ forms of PD has provided considerable insights into the disease pathogenesis and the recent burst of genomewide association studies has provided evidence that familial and sporadic forms of PD may share common genetic backgrounds. These considerations have prompted the development of new animal models, in particular mice, which reproduce known PD-related mutations. Below, we describe transgenic mice models, as well as alternative models developed in non-mammalian organisms, such as Drosophila, Caernorhabditis elegans and zebrafish (Danio rerio), which recapitulate monogenic mutations observed in familial PD patients. A list of known genes and related protein products involved in familial PD is provided by Alberio et al. [13]. 1160

Mouse genetic models a-Synuclein a-Synuclein mutations cause a rare form of autosomal dominant PD and SNCA was the first gene to be linked to familial PD. Three PD-related point mutations (A30P, A53T and E46K) have been identified so far. Duplication or triplication of the gene can cause PD as well, suggesting that the level of expression of the protein is also a causal factor of disease. To date, various a-synuclein transgenic mice have been developed, although no significant nigrostriatal degeneration has been obtained. Expression of wild-type or mutated a-synuclein under the tyrosine hydroxylase (TH) promoter, which allows selective expression of the protein in catecholaminergic neurones, leads to down-regulation of TH expression without evidence of neuronal death. Similarly, the expression of truncated a-synuclein on a null background reduced DA levels, although it had no effect on nigral neurones [14]. On the other hand, the use of thy-1 promoter, which prompts expression of wild-type a-synuclein in cortical and subcortical neurones, as well as in SNpc, leads to a-synuclein inclusions in the olfactory bulb, SNpc and locus coeruleus. Time-dependent decrease of striatal DA content, reduced TH expression in the striatum and movement slowness were also detected [15], along with increased susceptibility to MPTP toxicity. Transgenic mice in which neuronal a-synuclein expression is driven by the prion promoter do not replicate the typical neurodegeneration of dopaminergic system but, instead, exhibit characteristic motor dysfunctions and are responsive to DA treatment [14]. Interestingly, some of these models also reproduced the non-motor dysfunctions observed at early stages of PD, such as gastrointestinal alterations and olfactory deficits. Leucine-rich repeat serine ⁄ threonine kinase 2 (LRRK2) Mutations in the gene for LRRK2 represent the most prevalent cause of autosomal dominant PD. To date, six disease-causing mutations have been identified (R1441G, R1441C, N1437H, Y1699C, G2019S, I2020T). Overexpression of wild-type LRRK2 in BAC transgenic mice induces increased DA release in the striatum and motor hyperactivity, whereas overexpression of the G2019S mutated protein causes age-dependent reduction of the striatal content, release and uptake of DA, thereby suggesting a role for LRRK2 in DA transmission [16]. Similarly, BAC transgenic mice overexpressing R1441G mutant LRRK2 protein showed age-dependent and progressive motor-activity



deficits, with apparent immobility by 10 months of age, reminiscent of akinesia in advanced PD patients, which could be reversed by dopaminergic agents. The pathological function of mutated LRRK2 may involve interactions with a-synuclein, possibly mediated by the altered kinase activity associated with the mutations; indeed, mutated LRRK2 prompts the aggregation of a-synuclein and phosphorylation of the protein on serine 129, which is typically found in LBs [17,18]. Overexpression of the wild-type protein or of its mutated form, however, fails to induce neuronal cell death. LRRK2 knockout (KO) mice are viable, and do not show major abnormalities or increased susceptibility to MPTP. PTEN-induced putative kinase 1 (PINK1) Mutations in the gene encoding for PINK1 cause the second most common form of autosomal-recessive PD and lead to loss of function of the protein. PINK1 is a Ser ⁄ Thr kinase with a dual subcellular localization (mitochondrial and cytosolic), probably reflecting compartment-specific functions. PINK1 KO mice do not exhibit major abnormalities; in particular, no changes in the number of DA neurones or striatal DA levels are observed, whereas mild mitochondrial and nigrostriatal neurotransmission deficits can be present. PINK1 KO mice, however, show increased susceptibility to oxidative stress and ROS production [19]. Recently, an agedependent reduction of DA levels accompanied by reduced locomotor activity was observed in G309DPINK1 transgenic mice, in which PINK1 function is completely abolished [20]. These mice showed no LB formation or nigrostriatal degeneration for up to 18 months of age; however, a progressive reduction of mitochondrial pathogenesis was detected. Parkin Parkin mutations cause the most common autosomal recessive form of PD. To date, more than 100 alleged disease-causing mutations have been reported in the gene for parkin, most of which lead to loss of function of the protein and usually cause early-onset PD. Parkin is an E3 ubiquitin ligase and participates in the ubiquitin proteasome system. Moreover, in combination with PINK1, parkin is directly involved in the mitochondria quality control. Several parkin KO mice have been developed, although none have shown substantial dopaminergic or behavioural abnormalities, except for subtle nigrostriatal and locus coeruleus alterations. These mice show a higher susceptibility both to neurotoxins and inflammatory stimuli, suggest-


ing that parkin mutations may sensitize dopaminergic neurones to cellular insults. On the other hand, transgenic mice overexpressing the Q311X parkin mutation selectively in dopaminergic neurones developed progressive motor deficits and age-dependent nigrostriatal degeneration, as well as a-synuclein pathology, thereby suggesting that mutant parkin proteins may act as dominant-negative modulators [21]. DJ-1 Point mutations (L166P, D149A) in the gene for DJ-1 cause rare autosomal-recessive PD with early onset. DJ-1 is a redox-sensitive molecular chaperones protein, localized in the cytoplasm, which associates with mitochondria and the nucleus upon oxidation. Mutations cause a loss of function of DJ-1 by inducing instability of the dimeric, functional form of the protein or lack of expression. Mutations also affect the serine protease activity of DJ-1, which represents another crucial function of this protein [13]. DJ-1 KO mice are more sensitive to toxins and oxidative stress, but, similar to parkin and PINK1 KO mice, they do not display major neuronal abnormalities. To the best of our knowledge, no transgenic mouse lines overexpressing wild-type or mutant DJ-1 have been reported to date. DJ-1 overexpression in nigral neurones of mice has been induced either directly, using adeno-associated vectors [22], or, indirectly, by administration of the histone deacetylase inhibitor phenylbutyrate-induced [23]. In both cases, DJ-1 overexpression was associated with increased protection against toxin-induced neurodegeneration. In keeping with this latter observation, both in vitro and in vivo data indicate that toxic insults lead to the enhanced expression of DJ-1, probably as a defence mechanism to counteract increased oxidative stress, suggesting that the protein may be protective but not essential. Multiple transgenic mice Multiple transgenic mouse lines have also been developed, although no consistent results have been obtained. Crossing of a-synuclein transgenic mice with parkin or DJ-1 KO mice, or the simultaneous silencing of PINK-1, DJ-1 and parkin, did not result in nigrostriatal degeneration. Overexpression or KO of LRRK2 enhanced or reduced, respectively, the protein aggregation normally observed in A53T a-synuclein transgenic mice, suggesting that LRRK2 may be involved in synuclein toxicity [24]. To date, no LRRK2 transgenic or KO line in a parkin, PINK1 or DJ-1 KO background has been reported.


1161



MitoPark mouse As noted above, the mitochondrion represents a crucial crossroad in PD pathogenesis. For this reason, a novel rodent model, the MitoPark mouse, has been developed in which mitochondrial function is selectively disrupted in dopaminergic neurones. This is obtained through the elimination of the nuclear genome-encoded mitochondrial Tfam gene, which promotes mitochondrial DNA replication and transcription. Animals bearing this selective deletion survive to adulthood and progressively show a PD-like phenotype, including degeneration of nigrostriatal pathways, reduced striatal DA and intracellular dense bodies in dopaminergic neurones reminiscent of LBs. Interestingly, l-DOPA treatment normalizes motor deficits of MitoPark mice and its efficacy weakens with age, possibly as a consequence of the progressive neurodegenerative process [25]. Electrophysiological and neurochemical alterations within the nigrostriatal circuit, including a reduced capacity for DA release in the striatum, are detected before the emergence of overt behavioural deficits. Although MitoPark mice do not reflect specific PD-related mutations, they recapitulate alterations of cellular pathways knowingly affected in PD, thereby suggesting that this genetic model may represent a valuable tool for the development and screening of new therapeutic strategies in PD.

RAT genetic models Transgenic rat models reproducing monogenic PD mutations have been developed recently, some of which are now commercially available. These rat models represent a step forward for two reasons: (a) compared to mice, the rat neuronal circuitry more closely recalls that of humans and (b) compared to mice, rats are less prone to anxiety, which represents a major advantage for behavioural evaluation. Transgenic rats expressing human A53T and A30P mutated a-synuclein under the TH promoter have been recently described. These animals showed no major motor impairments, up to 25 months of age, although significant olfactory deficits were observed, reminiscent of those that can be detected in the early phases of human PD [26]. A temporal G21019S LRRK2 transgenic rat has also been developed very recently [27] in which transgene expression was turned on in 5-monthold rats, impairing DA uptake by DAT. No signs of dopaminergic cell loss, however, were detected in these animals. On the other hand, when neurone-specific expression of G2019S mutant LRRK2 is driven in adult rats by adenoviral vectors, progressive degenera1162

tion of nigral dopaminergic neurones is found [28]. Interestingly, constitutive expression of G2019S LRRK2 has no effect, suggesting the presence of significant compensatory mechanisms.

Alternative transgenic models The lack of nigrostriatal degeneration observed in transgenic mice has recently encouraged the development of alternative, non-mammalian models, with reduced genomic complexity and a greater ease of manipulation. These models are characterized by lower costs compared to rodents or nonhuman primates, and allow the possibility of conducting high-throughput experiments. Drosophila In the past decade, the fruit fly Drosophila melanogaster has emerged as a suitable model for studying mechanisms of PD-related neurodegeneration. Clusters of dopaminergic neurones are detectable in the developing and adult fly, and metabolic pathways for DA synthesis are conserved between Drosophila and humans. The first PD-like Drosophila model was first generated by neuronal overexpression of wild-type or mutant (A53T or A50P) human a-synuclein, for which no fly homologue exists [29], using the GAL4 ⁄ upstream activation sequence system. Transgenic flies showed age-dependent and selective loss of dopaminergic neurones, and also the formation of fibrillary inclusions containing a-synuclein, as well as a progressive loss of climbing activity (index of motor deficit), counteracted by l-DOPA or DA-agonists. The importance of post-translational modification, such as phosphorylation on serine 129 and tyrosine 125, on a-synuclein oligomerization and toxicity has emerged using the Drosophila model. For example, Karpinar et al. [30] showed that soluble oligomers variants of a-synuclein correlate with enhanced dopaminergic degeneration in Drosophila, suggesting for the first time that the early, soluble forms of aggregates represent the most toxic form of a-synuclein. Mutational analyses of a-synuclein in Drosophila have permitted an extended evaluation of the protein domains involved and ⁄ or required for toxicity showing, for example, that truncated forms of a-synuclein have a central hydrophobic region between residues 71 and 82 that is essential for the formation of oligomeric and fibrillary forms of the protein and toxicity. Most of the genes implicated in familial PD, with the exception of a-synuclein, have at least one fly homologue. Mutations that induce a loss of function or inactivation of the fly homologues of PINK1,



parkin, DJ-1 or LRKK2 selectively in dopaminergic neurones have been associated with a loss of DA neurones and motor deficits. In general, loss of the Drosophila homologue gene causes a weaker phenotypic alteration compared to the corresponding form of familial PD and no evidence of LB formation, possibly because of the absence of an a-synuclein fly homologue [31]. Drosophila parkin null mutants show a reduced life span, mitochondrial defects and flight muscle degeneration that leads to severe flight and climbing disabilities, as well as reduced proteasome 26S activity. Similarly, overexpression of mutant but not wild-type human parkin in Drosophila induces dopaminergic degeneration and motor deficits, suggesting a dominant negative effect of the mutated protein in PD pathology. PINK1 mutant flies share phenotypic characteristics with parkin mutant flies, including dopaminergic neurone degeneration and flying deficits. Studies in Drosophila were shown to be important for identifying the role that parkin and PINK1 play in the regulation of mitochondrial physiology, including fission ⁄ fusion processes [31]. Overexpression of PINK1 in Drosophila has also been shown to suppress a-synuclein-dependent phenotypes, including dopaminergic cell loss and climbing ability, also pointing to a potential role of PINK1 in synaptic function. Unlike mammals, Drosophila expresses two DJ-1 homologues: DJ-1a, restricted to male germline, and DJ-1b that, as in mammals, is ubiquitously expressed. Different mutations in both genes have been induced. DJ-1b KO flies show an enhanced susceptibility to cytotoxins, such as paraquat, H2O2 and rotenone, further supporting the protective redox function of DJ-1. Similarly, DJ-1b mutations that cause a loss of protein function protein lead to the accumulation of ROS in the fly brain. Inconsistent results have been obtained in KO LRRK2 Drosophila and also in flies overexpressing wild-type or mutant forms of the human protein, with flies showing no overt dopaminergic degeneration. Interaction studies in Drosophila have also revealed that LRKK2, PINK1, parkin and DJ-1 genetically interact with each other and that LRRK2 is able to suppress PINK1 and parkin pathology in flies, most likely through the activation of the 4E(eIF4E) binding protein. C. elegans C. elegans comprises an optimal model for genetic studies because it is a multicellular organism that is sufficiently simple to be studied in great detail, with a well characterized genome. Strains can be frozen and


subsequently thawed, remaining viable, thus allowing long-term storage. In addition, C. elegans is one of the simplest organisms with a nervous system and it is relatively easy to disrupt the function of specific genes via RNA interference by injecting, soaking or feeding the worms with double-stranded DNA containing the sequence of the targeted gene. One-third (302 out of 10 000) of the cells that constitute the adult worm are neurones, of which exactly eight are dopaminergic and are involved in motor activity. Mutations within the gene for TH abolish the typical decrease in locomotion (basal slowing) observed when the worm enters a bacterial lawn, indicating that DA, as in mammals, is involved in movement control. C. elegans is transparent and dopaminergic neurones can be easily observed in live animals by driving the expression of green fluorescent protein under the DAT promoter, thereby allowing rapid assessment of neurodegenerative cell loss. Homologues of human PD-related proteins, including parkin, LRRK2, PINK1 and DJ-1, but not a-synuclein, have been identified and are conserved in C. elegans. Overexpression of wild-type and mutated (A53T, A30P) human a-synuclein induces a loss of dopaminergic neurones, an accumulation of total and phosphorylated a-synuclein in cell bodies and dendrites (reminiscent of LB-like inclusions), as well as deficits in basal slowing that can be relieved by dopaminergic treatment [30]. In addition, although C. elegans overexpressing A56P or A76P mutated human a-synuclein are unable to form a b-sheet structure and form aggregate deposits, these variants have higher toxicity in vivo. These data again suggest that soluble oligomers, rather than a-synuclein deposits, may be the toxic a-synuclein species in PD. KO parkin C. elegans shows increased susceptibility to mitochondrial inhibitors [32]. Similarly, C. elegans bearing a parkin mutation that causes the expression and aggregation of a truncated form of the protein, reminiscent of parkin mutations in PD patients, shows hyper-sensitivity to endoplasmic reticulum stress [33]. Interestingly, the expression of human a-synuclein in this parkin mutant C. elegans background caused increased developmental defects and lethality. Similar to parkin, KO DJ-1 and PINK1 C. elegans show increased susceptibility to PD-inducing toxins [32] and this phenotype can be rescued by overexpression of wild-type LRRK2, indicating the protective nature of LRRK2. Overexpression of LRKK2 G2019S mutant in C. elegans enhances dopaminergic neurodegeneration, reduces DA levels and enhances vulnerability to rotenone, whereas wild-type LRKK2 overexpression enhances resistance to the toxin.


1163



Analyses of KO and mutant C. elegans allow rapid assessment of genetic and environmental interactions and can be predictive of a cross-link between these two elements. For example, KO of LRK-1, the C. elegans homologue of human LRRK2, suppresses all phenotypic aspects of PINK1 loss-of-function mutants. Conversely, the hypersensitivity of LRK1 mutant animals to endoplasmic reticulum stressors can be reduced in a PINK1 mutant background, suggesting an antagonistic role of PINK-1 and LRK-1. C. elegans has also been used for large-scale screening of potential modifiers of a-synuclein. Such large-scale functional genomic analyses sustain the value of C. elegans for discovering novel proteins and pathways, especially in the context of a-synuclein-related pathology. Zebrafish Zebrafish is a popular aquarium fish, 3–4 cm in length, with features that make it an interesting and simple model for evaluating pathological mechanisms in PD. Embryos develop externally and are transparent, thus allowing the direct use of fluorescent reporter to visualize morphology and ⁄ or the phenotypic changes induced by genetic or pharmacological manipulations. Fish larvae can be breed in 96-well plates and used to perform high-throughput screenings of innovative therapeutic agents. Importantly, extensive information has demonstrated similarities to the mammalian central nervous system and key areas of zebrafish brain are astonishingly conserved compared to their human counterparts; for example, certain areas of the telencephalon are homologous to regions of mammalian basal ganglia involved in motor movements. TH-positive neurones located in the ventral diencephalon are homologous to mammalian midbrain SNpc and ventral tegmental area neurones and project to the proencephalon, which displays anatomical similarities to the striatum. Dopaminergic neurones (14 in total) are sensitive to PD-inducing toxins. Similar to Drosophila and C. elegans, extensive genomic data are available on zebrafish; PD-related protein homologues, including parkin, DJ-1, PINK1 and LRKK2, are detected in zebrafish. Although three synuclein genes are present in the fish, no homologue to human a-synuclein has been observed. In zebrafish, overexpression of parkin protein protects the fish from cellular stress. Parkin KO causes a moderate (20%) loss of dopaminergic neurones, reduced mitochondrial complex I activity and increased susceptibility to toxins [34]. As in mice, DJ-1 KO in zebrafish embryos does not cause detectable DA cell death, although it induces an increased suscep1164

tibility of dopaminergic neurones to toxins, thereby supporting the protective function of the protein [35]. Similarly, PINK1 knockdown in zebrafish does not induce dopaminergic cell loss but, instead, alters dopaminergic projections and induces locomotor deficits [36] and can be rescued by the expression of human LRRK2 protein. Zebrafish with PINK1 point mutations that cause a loss of kinase domain show a reduced number of dopaminergic neurones and mitochondrial function at the larval stage [37]. Interestingly, KO of LRKK2 causes embryonic lethality, whereas the deletion of the functional WD40 domain of the protein, required for dimerization of the protein, induces the loss of DA neurones and motor deficits [38]. Zebrafish can readily absorb hydrophilic compounds added to their tank water. The model therefore represents a potential tool for high-throughput testing and screening of novel molecules that may counteract pathological mechanisms in PD.

Concluding remarks The availability of an experimental model mimicking all the major pathological and phenotypic features of PD is a crucial need that remains to be addressed. Such a tool would be instrumental for a full understanding of PD pathogenesis, which would lead to the identification of disease-modifying therapies. Various toxic and transgenic models of PD are currently available, all with significant advantages and disadvantages. If we consider toxic models, substantial nigrostriatal degeneration is generally obtained, with good replication of PD motor symptoms (particularly in MPTP-treated monkeys), although no consistent LB-like formation is detected, with the possible exception of rotenone. On the other hand, transgenic models offer astonishing insights into selected molecular aspects of PD pathogenesis, particularly for the familial forms, and LB-like inclusions can be observed, at least in a-synuclein overexpressing animals. However, the absence of consistent neuronal damage in the nigrostriatal pathway remains a major limitation for these models. Thus, until a ‘perfect’ model is developed, any suitable research strategy will rely on the accurate selection of the model to be used based on the specific research needs. For example, if a neuroprotective treatment must be tested, a model granting a reproducible, toxin-induced nigrostriatal lesion will be used. By contrast, if the role of selected proteins involved in PD pathogenesis is to be investigated, then a specific transgenic model will be the right choice.




Conflict of interest

tions following intrastriatal injection of 6-hydroxydopamine in the rat: new clues from an old model. Eur J Neurosci 25, 397–405. Hunter RL, Cheng B, Choi DY, Liu M, Liu S, Cass WA & Bing G (2009) Intrastriatal lipopolysaccharide injection induces parkinsonism in C57 ⁄ B6 mice. J Neurosci Res 87, 1913–1921. Alberio T, Lopiano L & Fasano M (2012) Cellular models to investigate biochemical pathways in Parkinson’s disease. FEBS J 279, 1146–1155. Chesselet MF (2008) In vivo alpha-synuclein overexpression in rodents: a useful model of Parkinson’s disease? Exp Neurol 209, 22–27. Lam HA, Wu N, Cely I, Kelly RL, Hean S, Richter F, Magen I, Cepeda C, Ackerson LC, Walwyn W et al. (2011) Elevated tonic extracellular dopamine concentration and altered dopamine modulation of synaptic activity precede dopamine loss in the striatum of mice overexpressing human alpha-synuclein. J Neurosci Res 89, 1091–1102. Li X, Patel JC, Wang J, Avshalumov MV, Nicholson C, Buxbaum JD, Elder GA, Rice ME & Yue Z (2010) Enhanced striatal dopamine transmission and motor performance with LRRK2 overexpression in mice is eliminated by familial Parkinson’s disease mutation G2019S. J Neurosci 30, 1788–1797. Kumar A & Cookson MR (2011) Role of LRRK2 kinase dysfunction in Parkinson disease. Expert Rev Mol Med 13, e20. Qing H, Wong W, McGeer EG & McGeer PL (2009) Lrrk2 phosphorylates alpha synuclein at serine 129: Parkinson disease implications. Biochem Biophys Res Commun 387, 149–152. Gautier CA, Kitada T & Shen J (2008) Loss of PINK1 causes mitochondrial functional defects and increased sensitivity to oxidative stress. Proc Natl Acad Sci USA 105, 11364–11369. Gispert S, Ricciardi F, Kurz A, Azizov M, Hoepken HH, Becker D, Voos W, Leuner K, Muller WE, Kudin AP et al. (2009) Parkinson phenotype in aged PINK1deficient mice is accompanied by progressive mitochondrial dysfunction in absence of neurodegeneration. PLoS One 4, e5777. Lu XH, Fleming SM, Meurers B, Ackerson LC, Mortazavi F, Lo V, Hernandez D, Sulzer D, Jackson GR, Maidment NT et al. (2009) Bacterial artificial chromosome transgenic mice expressing a truncated mutant parkin exhibit age-dependent hypokinetic motor deficits, dopaminergic neuron degeneration, and accumulation of proteinase K-resistant alpha-synuclein. J Neurosci 29, 1962–1976. Paterna JC, Leng A, Weber E, Feldon J & Bueler H (2007) DJ-1 and Parkin modulate dopamine-dependent behavior and inhibit MPTP-induced nigral dopamine neuron loss in mice. Mol Ther 15, 698–704.

The authors declare that they have no conflicts of interest.

12

References 1 Langston JW, Ballard P, Tetrud JW & Irwin I (1983) Chronic Parkinsonism in humans due to a product of meperidine-analog synthesis. Science 219, 979–980. 2 Fornai F, Lenzi P, Ferrucci M, Lazzeri G, di Poggio AB, Natale G, Busceti CL, Biagioni F, Giusiani M, Ruggieri S et al. (2005) Occurrence of neuronal inclusions combined with increased nigral expression of alpha-synuclein within dopaminergic neurons following treatment with amphetamine derivatives in mice. Brain Res Bull 65, 405–413. 3 Shimoji M, Zhang L, Mandir AS, Dawson VL & Dawson TM (2005) Absence of inclusion body formation in the MPTP mouse model of Parkinson’s disease. Brain Res Mol Brain Res 134, 103–108. 4 Alvarez-Fischer D, Guerreiro S, Hunot S, Saurini F, Marien M, Sokoloff P, Hirsch EC, Hartmann A & Michel PP (2008) Modelling Parkinson-like neurodegeneration via osmotic minipump delivery of MPTP and probenecid. J Neurochem 107, 701–711. 5 Betarbet R, Sherer TB, MacKenzie G, Garcia-Osuna M, Panov AV & Greenamyre JT (2000) Chronic systemic pesticide exposure reproduces features of Parkinson’s disease. Nat Neurosci 3, 1301–1306. 6 Fei Q, McCormack AL, Di Monte DA & Ethell DW (2008) Paraquat neurotoxicity is mediated by a Bakdependent mechanism. J Biol Chem 283, 3357–3364. 7 Manning-Bog AB, McCormack AL, Li J, Uversky VN, Fink AL & Di Monte DA (2002) The herbicide paraquat causes up-regulation and aggregation of alphasynuclein in mice: paraquat and alpha-synuclein. J Biol Chem 277, 1641–1644. 8 Thiruchelvam M, Brockel BJ, Richfield EK, Baggs RB & Cory-Slechta DA (2000) Potentiated and preferential effects of combined paraquat and maneb on nigrostriatal dopamine systems: environmental risk factors for Parkinson’s disease? Brain Res 873, 225–234. 9 Ungerstedt U, Ljungberg T & Steg G (1974) Behavioral, physiological, and neurochemical changes after 6-hydroxydopamine-induced degeneration of the nigrostriatal dopamine neurons. Adv Neurol 5, 421–426. 10 Sauer H & Oertel WH (1994) Progressive degeneration of nigrostriatal dopamine neurons following intrastriatal terminal lesions with 6-hydroxydopamine: a combined retrograde tracing and immunocytochemical study in the rat. Neuroscience 59, 401–415. 11 Blandini F, Levandis G, Bazzini E, Nappi G & Armentero MT (2007) Time-course of nigrostriatal damage, basal ganglia metabolic changes and behavioural altera-

13

14

15

16

17

18

19

20

21

22


1165



23 Zhou W, Bercury K, Cummiskey J, Luong N, Lebin J & Freed CR (2011) Phenylbutyrate up-regulates the DJ-1 protein and protects neurons in cell culture and in animal models of Parkinson disease. J Biol Chem 286, 14941–14951. 24 Lin X, Parisiadou L, Gu XL, Wang L, Shim H, Sun L, Xie C, Long CX, Yang WJ, Ding J et al. (2009) Leucine-rich repeat kinase 2 regulates the progression of neuropathology induced by Parkinson’s-disease-related mutant alpha-synuclein. Neuron 64, 807–827. 25 Good CH, Hoffman AF, Hoffer BJ, Chefer VI, Shippenberg TS, Backman CM, Larsson NG, Olson L, Gellhaar S, Galter D et al. (2011) Impaired nigrostriatal function precedes behavioral deficits in a genetic mitochondrial model of Parkinson’s disease. FASEB J 25, 1333–1344. 26 Lelan F, Boyer C, Thinard R, Remy S, Usal C, Tesson L, Anegon I, Neveu I, Damier P, Naveilhan P et al. (2011) Effects of human alpha-synuclein A53T-A30P mutations on SVZ and local olfactory bulb cell proliferation in a transgenic rat model of Parkinson disease. Parkinsons Dis 2011, 987084. 27 Zhou H, Huang C, Tong J, Hong WC, Liu YJ & Xia XG (2011) Temporal expression of mutant LRRK2 in adult rats impairs dopamine reuptake. Int J Biol Sci 7, 753–761. 28 Dusonchet J, Kochubey O, Stafa K, Young SM Jr, Zufferey R, Moore DJ, Schneider BL & Aebischer P (2011) A rat model of progressive nigral neurodegeneration induced by the Parkinson’s disease-associated G2019S mutation in LRRK2. J Neurosci 31, 907–912. 29 Feany MB & Bender WW (2000) A Drosophila model of Parkinson’s disease. Nature 404, 394–398. 30 Karpinar DP, Balija MB, Kugler S, Opazo F, RezaeiGhaleh N, Wender N, Kim HY, Taschenberger G, Falkenburger BH, Heise H et al. (2009) Pre-fibrillar

1166

31

32

33

34

35

36

37

38

alpha-synuclein variants with impaired beta-structure increase neurotoxicity in Parkinson’s disease models. EMBO J 28, 3256–3268. Hirth F (2010) Drosophila melanogaster in the study of human neurodegeneration. CNS Neurol Disord Drug Targets 9, 504–523. Ved R, Saha S, Westlund B, Perier C, Burnam L, Sluder A, Hoener M, Rodrigues CM, Alfonso A, Steer C et al. (2005) Similar patterns of mitochondrial vulnerability and rescue induced by genetic modification of alpha-synuclein, parkin, and DJ-1 in Caenorhabditis elegans. J Biol Chem 280, 42655–42668. Springer W, Hoppe T, Schmidt E & Baumeister R (2005) A Caenorhabditis elegans Parkin mutant with altered solubility couples alpha-synuclein aggregation to proteotoxic stress. Hum Mol Genet 14, 3407–3423. Fett ME, Pilsl A, Paquet D, van Bebber F, Haass C, Tatzelt J, Schmid B & Winklhofer KF (2010) Parkin is protective against proteotoxic stress in a transgenic zebrafish model. PLoS One 5, e11783. Bretaud S, Allen C, Ingham PW & Bandmann O (2007) p53-dependent neuronal cell death in a DJ-1-deficient zebrafish model of Parkinson’s disease. J Neurochem 100, 1626–1635. Xi Y, Ryan J, Noble S, Yu M, Yilbas AE & Ekker M (2010) Impaired dopaminergic neuron development and locomotor function in zebrafish with loss of pink1 function. Eur J Neurosci 31, 623–633. Xi Y, Noble S & Ekker M (2011) Modeling neurodegeneration in zebrafish. Curr Neurol Neurosci Rep 11, 274–282. Sheng D, Qu D, Kwok KH, Ng SS, Lim AY, Aw SS, Lee CW, Sung WK, Tan EK, Lufkin T et al. (2010) Deletion of the WD40 domain of LRRK2 in Zebrafish causes Parkinsonism-like loss of neurons and locomotive defect. PLoS Genet 6, e1000914.
