Advances in animal models of Parkinson ’ s disease

Parkinson ’ s disease is a complex neurodegenerative disease characterized by progressive movement impairments. Predominant symptoms encompass resting tremor, bradykinesia, limb rigidity, and postural instability. In addition, it also includes a series of non-motor symptoms such as sleep disorders, hyposmia, gastrointestinal dysfunction, autonomic dysfunction and cognitive impairment. Pathologically, the disease manifests through dopaminergic neuronal loss and the presence of Lewy bodies. At present, no significant breakthrough has been achieved in clinical Parkinson ’ s disease treatment. Exploring treatment modalities necessitate the establishment of scientifically sound animal models. In recent years, researchers have focused on replicating the symptoms of human Parkinson ’ s disease, resulting in the establishment of various experimental animal models primarily through drugs and transgenic methods to mimic relevant pathologies and identify more effective treatments. This review examines traditional neurotoxin and transgenic animal models as well as α -synuclein pre-formed fibrils models, non-human primate models and non-mammalian specie models. Additionally, it introduces emerging models, including models based on optogenetics, induced pluripotent stem cells, and gene editing, aiming to provide a reference for the utilization of experimental animal models and clinical research for researchers in this field.


Introduction
As the second most common degenerative neurological disease following Alzheimer's disease, Parkinson's disease (PD) affects more than 6 million individuals worldwide (Bloem et al., 2021).The disease typically occurs in the elderly population, with an overall incidence of approximately 0.3 % of the population, rising to 1-3 % in individuals over 65 years of age.Evidently, the incidence of PD is expected to witness a meteoric rise, with Chinese patients accounting for 1.7 % of the total number (Chen et al., 2018).James Parkinson was the first to describe the symptomatic manifestations of PD in the 19th century (Kalia and Lang, 2015).Individuals with PD typically exhibit an insidious progression of motor coordination abnormalities, primarily characterized by resting tremor, bradykinesia, limb rigidity, postural instability, and some non-motor symptoms, such as sleep disorders, hyposmia, gastrointestinal (GI) dysfunction, autonomic dysfunction and cognitive impairment (Armstrong and Okun, 2020).It is generally believed that the onset of non-motor symptoms precedes the onset of motor symptoms, strengthening the study of non-motor symptoms may identify the prodromal symptoms of PD and help us to make an early diagnosis of PD. (Schapira et al., 2017;Tolosa et al., 2021).Pathological features predominantly comprise neuron-enveloped inclusions, primarily manifesting as Lewy bodies (LB) and Lewy neurites, along with the progressive degeneration of dopaminergic (DA) neurons in brain regions such as the substantia nigra pars compacta (SNpc) and corpora striata (CS).α-synuclein (α-syn), a critical precursor of LB, undergoes phosphorylation and fibrogenic aggregation in the cytoplasm (More et al., 2016;Creed and Goldberg, 2018).Age significantly influences PD prevalence, exhibiting an apparent positive correlation with diagnosis.Men are at higher risk of developing the disease compared to women, with a prevalence in men approximately 1.5 times higher than in women.Transgenic (TG) technology and other contributory factors are likewise intimately linked to disease risk, with more than 90 pertinent causal loci identified (Blauwendraat et al., 2020).Substantial evidence implicates environmental factors such as pesticides, insecticides, water pollutants, habitual smoking, inappropriate physical exercises, and traumatic head injuries (Neurol, 2019) in PD across various populations (Poewe et al., 2017).
In most cases, PD presents in sporadic form and is attributed to multiple factors, such as age, environment, and genetics (Cacabelos, 2017).Despite significant strides in understanding the intricate pathogenesis and epidemiology of PD, its precise etiology remains to be systematically deciphered, and definitive cures or preventive treatments have not yet been developed.In recent decades, substantial efforts have been continually devoted to replicating the symptoms associated with human PD in animal models.Neurotoxin models have garnered widespread use in PD research owing to their established experimental techniques, cost-effectiveness, rapid iteration, and excellent experimental reproducibility.Contemporary advancements in TG technology have further refined genetic models, furnishing a novel paradigm in the exploration of PD.In addition, α-syn pre-formed fibrils (PFF) models, non-human primate models, non-mammalian specie models, and some emerging animal models have been developed to deepen our understanding of PD from different perspectives.Comprehending the merits and constraints of diverse animal models can advance modeling methodologies and facilitate a lucid and pragmatic selection of appropriate models for experimental purposes.This review aimed to elucidate the most commonly utilized neurotoxin, TG models, and some animal models with good prospects for development, summarizing their conception methodologies, working principles, advantages, and limitations.By doing so, it aims to provide a reference for PD researchers and contribute to the advancement of PD clinical medicine.

Neurotoxin-induced PD animal models
Empirical evidence indicates that humans exhibit a gamut of neurodegenerative clinical symptoms, ultimately culminating in PD, following the intravenous injection of pharmaceuticals contaminated with 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) (Langston et al., 1983).MPTP acts as a neurotoxin, inducing a PD-like phenotype through selective damage to DA neurons (Zeng et al., 2018).Consequently, the PD model induced by neurotoxins has garnered attention from researchers.Additionally, agrochemical compounds, such as pesticides, are prominent agents in the occupational environment, notably those associated with PD.High-frequency exposure to these pesticides increases the risk of PD in agricultural workers (Brown et al., 2006).Since the last century, researchers have found that neurotoxins, such as 6-hydroxydopamine (6-OHDA), rotenone, and paraquat (PQ), can be used to establish PD animal models.The neurotoxin-induced PD animal model, characterized by its frequent utilization, low experimental cost, and mature operational methodology, has been extensively employed in medical research.

MPTP-induced PD animal model
MPTP, an organic compound possessing analgesic properties, is nontoxic, fat-soluble, and adept at traversing the blood-brain barrier (BBB) unhindered.Upon entry into the brain, MPTP selectively targets astrocytes.The monoamine oxidase B enzyme, located in the outer mitochondrial membrane, catalyzes the conversion of MPTP into an intermediate metabolite, namely the 1-methyl-4-phenyl-2,3-dihydropyridinium ion.Subsequently, this compound undergoes spontaneous oxidation, giving rise to the 1-methyl-4-phenylpyridinium ion (MPP+) (Choi et al., 2015).MPP+ serves as a neurotoxin with a chemical structure akin to that of dopamine, enabling its recognition and binding by the dopamine transporter for subsequent transportation to DA neurons.Once internalized by neurons, MPP+ inhibits the activity of mitochondrial complex enzyme I.This inhibition not only results in a significant depletion of adenosine triphosphate levels but also triggers the excessive production of reactive oxygen species (ROS).The resultant increase in oxidative stress directly leads to the degeneration and loss of DA neurons, ultimately culminating in PD (Chiba et al., 1984) (Fig. 1).Given MPTP's specific inhibitory effect on mitochondrial function, this model holds particular significance in investigating mitochondrial dysfunction in PD (Bian et al., 2012).
The MPTP model currently stands as the most utilized model for PD research.It effectively replicates the behavioral symptoms characteristic of PD while accurately simulating the progressive degeneration of DA neurons in the SNpc and CS.Leveraging MPTP's inherent ability to efficiently traverse the BBB, this model predominantly employs diverse administration methods such as intraperitoneal, subcutaneous, and intramuscular injections, as well as intravenous infusions.By modulating injection parameters, such as intensity and duration, researchers can generate different mouse models of PD (Petroske et al., 2001).A single low dose injection of MPTP can establish a pre-PD model.Within 24 hours, multiple injection of a certain dose MPTP can establish acute PD models.Subacute and chronic PD models can be established by continuous or non-continuous injection of MPTP once daily for several days to several weeks (Nataraj et al., 2016).Nonetheless, acute models exhibit a shorter disease course, often demonstrating rapid onset and a high mortality rate.Experimental animals may be resistant to MPTP and spontaneously recover to normality within a short timeframe, thus failing to replicate the long-term progressive nature of human PD (Xie et al., 2020).However, compared to their acute counterparts, subacute models offer an extended period of neurological damage and a longer latently pathogenic stage.Studies have suggested that chronic animal models, induced through low-dose MPTP administration, can demonstrate a gradual decline of DA neurons (Masilamoni and Smith, 2018;He et al., 2019).These models consistently display reduced levels of DA neurons and exhibit behaviors characterized by markedly decreased activity, slow movement, and occasional intracellular aggregates of α-syn (Kowall et al., 2000).In terms of non-motor symptoms, some mouse and rat models of MPTP showed olfactory impairment by intranasal administration (Lindgren and Dunnett, 2012;McDowell and Chesselet, 2012), and in Non-human primate (NHP) model, they showed increased apathetic and depressive behaviors (Czernecki et al., 2002).Recent studies have elucidated that utilizing a low dose of MPTP over an extended duration can yield a model that more closely resembles human PD symptoms (Merghani et al., 2021).
Rodents are commonly utilized in research due to their costeffectiveness and well-established methods of operation and management.The immune system in rats, however, is fairly resistant to MPTP toxicity, resulting in a less pronounced manifestation of the PD phenotype with a rapid onset (Soderstrom et al., 2006).Consequently, mice have emerged as the most widely exploited experimental animal in PD research.Notably, among mouse strains, the C57BL/6 strain stands out as the most sensitive and has been extensively utilized in elucidating pathogenic mechanisms, as well as in developing diagnostic and therapeutic strategies for PD (Rodríguez-Cruz et al., 2020;Zhu and Gong, 2020).

6-OHDA-induced PD animal model
6-OHDA is a neurotoxin with a chemical structure similar to DA that can selectively destroy SNpc DA neurons.Upon entering the brain, 6-OHDA binds to the DA transporter, facilitating its transportation into the mitochondria of DA neurons.Consequently, it reduces the synthesis of mitochondrial complex enzyme I. Within the cell, 6-OHDA increases the production of ROS and inhibits the production of antioxidant enzymes through autooxidation and metabolic processes (Deumens et al., 2002).The lack of reduction by antioxidant enzymes in DA neurons allows for the peroxidation of lipids, proteins, and DNA, among others, by ROS, instigating oxidative stress and mitochondrial degeneration (Puspita et al., 2017).In addition, 6-OHDA reduces mitochondrial membrane potential by inhibiting the activity of mitochondrial complex enzyme IV.This compromises the functionality of the mitochondrial respiratory chain and promotes mitochondrial dysfunction.Upon depletion of intracellular adenosine triphosphate, 6-OHDA eventually leads to the degeneration and demise of DA neurons (Prajapati et al., 2018) (Fig. 1).
6-OHDA selectively damages DA neurons in the SNpc.Due to its inability to effectively cross the BBB, 6-OHDA is administered via stereotactic injection into the brain.PD models featuring unilateral or bilateral lesions are typically established through single or double-point injections, and the injection sites mostly encompass the SNpc, CS, and medial forebrain bundle (MFB).Rodent models exhibit both motor and non-motor deficits, depending on the administration method (Pellegrini et al., 2020).Unilateral injection stands as the most utilized method to establish the 6-OHDA model, with symptoms manifesting as movement disorders contralateral to the injection site, facilitating the observation of clinical intervention effects with minimal unrelated side effects and a high survival rate (Kuruvilla et al., 2013).Bilateral injection leads to eating and drinking disorders, resulting in high mortality rates; thus, unilateral injection prevails in this context.Forced swimming tests in 6-OHDA-induced rodent models revealed increased depressive behavior (Taylor et al., 2009), symptoms of GI dysfunction with delayed gastric emptying were also found in the 6-OHDA rat model (Karasawa et al., 2014), and also showed impairment in olfactory function (Tadaiesky et al., 2008).Studies have shown that injection of 6-OHDA into the SNpc or MFB of mice leads to a 90 % loss of DA neurons after 12 hours, resulting in fatalities due to eating and drinking disorders without appropriate care measures (Blesa et al., 2011).Direct injections into the CS lead to persistent loss of DA neurons in the SNpc.Compared with site-directed injection in the SNpc and MFB, at the CS site, the pathological process is slower and milder, aligning more closely with the prolonged and slow pathogenesis of human PD (Stott and Barker, 2014).
The advantage of the 6-OHDA model lies in its ability to induce neurodegeneration in one hemisphere of the brain in experimental animals.This induces behavioral symptoms of PD on the contralateral side of the body, facilitating a more accurate evaluation of movement disorders in the model and the effect of drug interventions.However, this model has its limitations.The primary motor impairment observed in animal models is limited to lateral rotation, which cannot fully capture the relatively complex and variable behavioral manifestations characteristic of human PD.In addition, this model predominantly exhibits acute onset and short duration, contrasting with the long-term and progressive nature of human PD.Additionally, the formation of LB, a key pathological marker of PD has not been observed in this model.
In the 6-OHDA PD model, rats are typically selected as the experimental animals.Mice, cats, dogs, and monkeys also exhibit obvious reactions to 6-OHDA.The selection of experimental animals should be based on the specific research objectives and actual conditions.

Rotenone-induced PD animal model
Rotenone is a lipid-soluble organic insecticide prevalent in the root bark of plants.Due to its ability to effectively traverse the BBB, rotenone is primarily administered via intravenous, intraperitoneal, and subcutaneous injections, or oral routes, to establish chronic PD models.Upon entry into the brain, rotenone inhibits the activity of mitochondrial complex enzyme I, reduces glutathione levels, induces the generation of a large amount of ROS, disrupts DA metabolism, triggers oxidative stress in the SNpc, and impairs mitochondrial function.Consequently, DA neurons and cells undergo degeneration and death (Fig. 1), eventually leading to the manifestation of PD symptoms (Bisbal and Sanchez, 2019).Because the extent of rotenone absorption varies between animals and different parts of the animal, it is necessary to use specific doses for long-term administration in animal models, a stable chronic S.He et al. model can be established usually after 30 consecutive days of administration.Studies have shown that the chronic rotenone-induced rat model exhibits motor symptoms similar to those of human PD, such as slow gait, stiff movements, and limb tremors (Airavaara et al., 2020).Rats injected subcutaneously with rotenone for more than 30 days showed reduced motor activity and lethargy (Yi et al., 2007), and the rotenone-induced rat model exhibited depressive behavior during forced swimming and sucrose preference tasks (Santiago et al., 2010), this model is also capable of exhibiting GI dysfunction (Drolet et al., 2009).In addition, degeneration of SNpc DA neurons has been observed, along with the presence of inclusions featuring LB aggregation (Tasselli et al., 2013) or α-syn in the cytoplasm of surviving neurons (Inden et al., 2011).
Due to its high toxicity and unstable chemical properties, rotenone poses challenges for sustained utilization in animals, with prolonged and frequent administration increasing mortality rates among experimental animals.Therefore, while this model may not fully replicate the longterm progressive characteristics of human PD, the chronic model presents symptoms similar to the early stages of PD, thus facilitating investigation into peripheral nervous system pathology in the early stages of PD.The rotenone-induced PD model is typically established in rats but can also be established in organisms such as mice, fruit flies, and zebrafish (Wang et al., 2017).

PQ-induced PD animal model
PQ is a lipid-soluble organic herbicide with a chemical structure similar to MPP+ and exhibits drug toxicity comparable to MPTP.However, PQ cannot effectively cross the BBB and is typically used to establish PD animal models through intracerebral, intraperitoneal, and subcutaneous injections, or through oral and nasal administration routes.PQ reaches the SNpc and selectively destroys DA neurons (Fig. 1), thereby manifesting PD symptoms (Zhang et al., 2016).The mechanism underlying PQ neurotoxicity remains unclear.It is extremely toxic to the human body and can substantially damage organs.PQ is generally believed to induce mitochondrial dysfunction (Wiley et al., 2016), and disrupt redox reactions of glutathione and related proteins, thereby diminishing cellular antioxidant capacity (Niso-Santano et al., 2010).Studies have shown that prolonged PQ ingestion leads to gradual depletion of DA neurons, mirroring the PD characteristic of reduced activity.This modeling approach effectively simulates the degeneration of DA neurons caused by prolonged human exposure to environmental hazards, and facilitates the study of preclinical chronic PD symptoms (Quintero-Espinosa et al., 2017).Studies using rodent models of PQ have shown that PQ models can produce non-motor symptoms of depression and anxiety (Campos et al., 2013;Tinakoua et al., 2015).Nonetheless, the experimental duration of this method is substantially extended, and the timeliness of PQ toxicity is poor.Unlike most of the acute PD models, which entail large per-administration doses, a short medication cycle, and rapid symptom development, which do not effectively reflect the nature of human PD caused by environmental factors, PQ modeling offers the advantage of promoting α-syn production and inducing LB formation in SNpc DA neurons, which is crucial in investigations (Naudet et al., 2017).Typically, mice and rats are employed for establishing animal models of PD using PQ.

TG animal model
PD is caused by multiple factors, manifesting sporadically in patients in most cases.While the genetic contribution to PD is generally considered low, familial PD only represents 10 % of all cases (Balestrino and Schapira, 2020).Nonetheless, the role of genetic factors in familial PD pathogenesis is unequivocal.Harnessing advancements in biomedical engineering, the utilization of TG technology has emerged as a crucial avenue for studying elusive diseases.Studies have shown that mutations in genes associated with mitochondrial dysfunction are pivotal contributors to PD (Tang et al., 2015a(Tang et al., , 2015b)).In the realm of PD research, scientists have made significant progress in identifying multiple disease-causing genes.By manipulating these genes via TG technology, either through overexpression or knockdown, researchers have developed targeted PD models.This approach provides invaluable technical support for unraveling the pathological alterations observed in human PD and delving deeper into the molecular intricacies underlying its pathogenesis.The selection of animal species for TG modeling is diverse, with mice, rats, Drosophila, and zebrafish serving as primary experimental organisms.

Dominant gene model
PD predominantly manifests in elderly individuals, exhibiting an onset that extends over several decades.However, laboratory animal models possess considerably shorter lifespans.Consequently, researchers employ TG technology to overexpress specific target genes in these models, thereby enhancing the pathological manifestations and expediting the development of PD-like symptoms.Numerous autosomal dominant genes associated with PD, notably synuclein Alpha gene (SNCA) and leucine-rich repeat kinase 2(LRRK2), have been extensively studied.The TG model established through their overexpression stands out as the most used (Fig. 2).

SNCA
Studies have shown that point mutations in SNCA can induce abnormal folding and accumulation of α-syn, leading to damage in DA neurons and the onset of PD in carriers of the mutation (Paumier et al., 2015).Notably, the SNCA model currently stands as the most widely employed TG PD model, distinguished by its capability to exhibit α-syn aggregation.This feature is commonly employed in investigating therapeutic strategies and molecular mechanisms associated with α-syn (Suzuki et al., 2022).α-syn is a soluble small-molecule protein often expressed around the nucleus and in the central nervous system before synapses.It constitutes a principal component of LB (Dawson et al., 2010).Deviations in α-syn levels are closely linked to familial PD.SNCA encompasses three mutation sites, namely A53T, A30P, and FA6K.Any alteration at these sites results in the presentation of a β-fold conformation in α-syn.This aberrant β-structure leads to α-syn aggregation, which cannot be efficiently cleared, thereby leading to mitochondrial dysfunction and neuronal cell demise (Leitner et al., 2019).In the realm of PD animal models established via TG technology, two notable categories exist: those involving overexpression driven by distinct promoters and those mediated by various viral vectors.Currently, TG mouse models featuring overexpression of A53T and A30P are widely utilized, often to simulate early PD symptoms in humans.The A53T TG mouse model can well reflect the pathological manifestations of α-syn，exhibits pronounced neurodegeneration and motor deterioration, along with non-motor symptoms, such as sleep and olfactory dysfunction (Taguchi et al., 2020).However, the A30P TG mouse model is capable of manifesting non-motor symptom disorders characteristic of early PD in humans, such as impairment in visual acuity, olfactory dysfunction, and mood abnormalities (Veys et al., 2021).
These pathological features closely parallel those observed in human PD (Nagoshi, 2018).In addition, PD model has been established by overexpressing the human α-syn gene in wild-type mice.Mice in this model showed olfactory deficits at 3 months of age (Chesselet et al., 2008), cognitive decline at 6 months of age (Masliah et al., 2011), and GI dysfunction at 11 months of age (Wang et al., 2008).This model exhibits decreased tyrosine hydroxylase (TH) activity in the SNpc and CS, as well as motor dysfunction (Prasad et al., 2011).Although the aforementioned models fail to fully simulate human PD-related symptoms, wild-type mouse model of PD serves as a good reference for investigating the prodromal phase of PD in clinical research.

LRRK2
LRRK2, a gene situated on autosomal chromosome 12, encodes the LRRK2 protein, primarily a cytoplasmic kinase located on the outer mitochondrial membrane.Studies have demonstrated that LRRK2related pathogenesis involves multiple signaling pathways and cellular processes, such as autophagy and mitochondrial dynamics (Price et al., 2018).The overexpression of LRRK2 in immune cells is closely related to neuroinflammation in PD (Alessi and Sammler, 2018;Wang et al., 2021).
In addition, LRRK2 affects the transmission of α-syn between neurons by regulating its uptake balance.Remarkably, this gene exhibits the highest mutation rate in hereditary PD cases, and the G2019S loci within LRRK2 are extremely commonly mutated regions, as evidenced by previous research (Tsika and Moore, 2013).Compared with other inherited PD, PD resulting from LRRK2 mutations manifests later onset, and the early symptoms mainly manifest as synaptic damage, which may cause some recovery via complex human compensatory mechanisms, leading to the failure to detect early PD (Merino-Galán et al., 2022).
Recent investigations have unearthed intriguing correlations regarding LRRK2 mutations.A study revealed a positive correlation between the likelihood of developing PD and age among individuals carrying mutations in the G2019S locus of LRRK2.Indicating that the incidence of PD increases with advancing age (Healy et al., 2008).Similarly, when examining 18-month-old mice with mutations in the R1441C locus of LRRK2, researchers observed the accumulation of α-syn oligomers in both the CS and cerebral cortex.This observation may be closely linked to issues such as lysosomal dysfunction, compromised molecular chaperone-mediated autophagy, and hindered mitochondrial clearance (Ho et al., 2020;Liu et al., 2021).Both G2019S and R1441C TG rats exhibit degenerative motor and cognitive deficits.The R1441C model mice developed gait abnormalities after 15 months and were found to have olfactory impairment during subsequent testing.(Dranka et al., 2014), while neither model demonstrates SNpc DA neuronal loss or neuropathology (Cresto et al., 2020).It is noteworthy that while this model demonstrates some motor dysfunction, the presence of DA neuronal loss and LB formation remains a subject of debate.
The symptoms of PD associated with LRRK2 mutations are primarily attributed to mitochondrial dysfunction.Although the mechanism underlying the LRRK2 mutation has received extensive attention and relevant conclusions have been proposed, a comprehensive explanation for its pathological role remains elusive (Vázquez-Vélez and Zoghbi, 2021).Further exploration of LRRK2 is therefore of great significance.It is worth mentioning that LRRK2 kinase inhibitors (Polissidis et al., 2020), currently undergoing clinical trials (Zhao et al., 2021), and strategies aimed at reducing LRRK2 expression may provide novel avenues for the treatment of PD.

Recessive genetic animal models
Autosomal recessive genes implicated in PD have been manipulated through knockout, knockdown, or silencing techniques in experimental animal models, leading to the loss of function of the target gene and the manifestation of the desired phenotype for experimental purposes, thereby establishing PD animal models.Currently, E3 ubiquitin-protein ligase parkin (PRKN), phosphatase and tensin homolog-induced kinase 1 (PINK1), and protein deglycase (DJ-1) stand out among the autosomal recessive genes extensively employed in PD research.These genes play pivotal roles in constructing models that effectively capture the complexity of PD, providing methods and references for the study of this disease (Fig. 2).

PRKN
The PRKN gene encodes the Parkin protein, and mutations in this gene have been demonstrated to initiate early-onset PD.Both familial and sporadic cases of PD have been associated with this gene (Lesage et al., 2015).Parkin is an E3 ubiquitin protein ligase, which is encoded by 465 amino acids.It can recognize and label unwanted proteins and maintain the normal function of DA neurons.(Chia et al., 2020;Tokarew et al., 2021).
The PRKN gene, located on chromosome 6, is susceptible to genetic errors, resulting in mutations, that increase intracellular ROS levels, such mutations are associated with an increased risk of various cancers among carriers (Kung-Chun Chiu et al., 2019;Kitada et al., 2023).Parkin inactivation also leads to the progressive loss of DA neurons as evidenced in PRKN knockout mouse models and PD brain samples (Brahmachari et al., 2019).Some studies have found spatial memory impairment in this model mice at 5-6 months of age, but no other significant non-motor symptoms have been found (Rial et al., 2014).In addition, aberrant expression of PRKN disrupts the homeostasis of the ubiquitin-proteasome system, increases cytotoxicity, and impairs the normal functioning of DA neurons (Jeong et al., 2022).Studies have established a positive correlation between abnormal PRKN expression and the pathogenesis of PD (Chen et al., 2020).
In studies concerning PD treatment, PRKN gene activation has been demonstrated to decelerate DA neuron degeneration in 6-OHDA mice (Bian et al., 2012).These promising findings suggest a novel therapeutic avenue for the treatment of PD.

PINK1
PINK1 is a PD-related gene located in the mitochondria and encodes the PINK1 mitochondrial protein characterized by kinase activity.PINK1 is expressed in all cells of the body and is thought to protect cells from stress-induced mitochondrial dysfunction.Both PINK1 and PRKN S.He et al. are autosomal recessive genes associated with familial PD.Mutations in PINK1 lead to detrimental effects on various facets of mitochondrial biology, resulting in mitochondrial dysfunction.Consequently, impaired mitochondria fail to undergo efficient removal processes, thereby fostering the degeneration of DA neurons and ultimately culminating in PD (Ando et al., 2017).
Exemplifying its significance, studies on PINK1 knockout mice illuminate the intricacies of its impact.These mice exhibit olfactory dysfunction, gait deficits, vocal and swallowing disorders, that similar to the prodromal symptoms of PD in human patients (Glasl et al., 2012;Kelm-Nelson et al., 2018).Furthermore, PINK1 knockout mice exhibit age-dependent deficits in DA levels in the CS.Although motor symptoms display reduced activity, no conspicuous abnormalities are observed in the levels of DA neurons in the CS (Moisoi et al., 2014).These observed phenomena parallel the symptoms observed in human patients with preclinical PD.Consequently, PINK1 knockout mice emerge as an exemplary model for investigating the preclinical manifestations of PD, providing a valuable avenue for in-depth exploration and understanding of the disease (Jiang and Dickson, 2018).

DJ-1
A total of 11 mutation types have been identified within the DJ-1 gene, primarily classified into two categories: single-point mutations and large-segment deletions.These mutations exhibit an autosomal recessive inheritance pattern, with a discernible correlation between alterations in this gene and the onset of early-stage PD (Chen et al., 2005).
DJ-1 is considered to serve as a neuroprotective factor with robust antioxidant properties.Mutations in DJ-1 lead to oxidative stress, initiating a cascade of events.This involves the degradation of protein kinases, leading to systemic imbalances, and the translocation of intracytoplasmic proteins to the mitochondria, culminating in mitochondrial dysfunction.Consequently, the activity of mitochondrial complex enzyme I is compromised, fostering the aggregation of α-syn and the formation of LB.This pathological process accelerates the demise of DA neurons, ultimately triggering PD (Lenaz et al., 2006).Intriguingly, zebrafish models with DJ-1 knockouts do not exhibit significant DA neuronal damage.However, they do display aberrant motor symptoms and heightened sensitivity to prooxidants, underscoring the multifaceted role of DJ-1 in these organisms (Hughes et al., 2020).Divergent phenotypic outcomes are evident in DJ-1 knockout mice and rats.DJ-1deficient rats manifest a pronounced loss of DA neurons, accompanied by motor deficits, in a mouse model of 13 -14 months of age cognitive defects of motor symptoms were found.(Pham et al., 2010;Dave et al., 2014).However, DJ-1 knockout mice demonstrate mild motor deficits without exhibiting DA neuronal loss (Chen et al., 2005).
DJ-1 plays a crucial role in mitigating oxidative stress, preserving mitochondrial functionality, and averting damage (Lev et al., 2008).These findings collectively shed light on the intricate and context-dependent roles of DJ-1 in the realm of neurodegenerative diseases, particularly PD.For a relevant comparison between the neurotoxin model and the TG model, refer to Table 1.

MitoPark gene model
It is well known that mitochondrial dysfunction is one of the important triggers of PD, and the MitoPark mouse model, a transgenic mitochondrial damage model established by specific inactivation of mitochondrial transcription factor A in DA neurons.The MitoPark mouse model exhibits a slow progressive loss of DA neurons, exhibiting not only motor but also non-motor symptoms.It is generally believed that non-motor symptoms appear earlier than motor symptoms and can be used as a reference for early diagnosis of PD.At present, the most used Neurotoxin-induced PD animal models show rapid neurodegeneration and severe deterioration of the disease, which is not conducive to the study of early PD symptoms.Therefore, the MitoPark mouse model is beneficial to the study of early PD symptoms (Beckstead and Howell, 2021).
Study found that, due to mitochondrial dysfunction, MitoPark mice were found to show progressive degeneration of DA neurons at the SNpc and CS sites starting at approximately 12 weeks and continuing until weeks of age, and found in the nervous tissue similar LB of inclusion body protein.(Ekstrand et al., 2007), and have study also found that DA neurons in MitoPark mice degenerated at 16 weeks of age, and the degeneration was severe at about 27 weeks of age (Lynch et al., 2018).These studies all showed similar timelines for PD development in the MitoPark mouse model and in humans.In experiments with male and female MitoPark mice, motor symptoms began to develop between and 14 weeks of age and worsened between 16 and 24 weeks of age.Not surprisingly, non-motor symptoms occurred earlier, as spatial cognitive deficits were found in 8 weeks of age male mice by the water maze test.At 10-12 weeks of age MitoPark mice develop olfactory dysfunction.After 16 weeks showed anxiety and depression phenotype.(Langley et al., 2021).A study found that in all the light or the dark cycle, MitoPark mice exhibited circadian rhythm disorders (Fifel and Cooper, 2014).In addition, MitoPark mice showed a positive response to levodopa treatment and a time-dependent decrease in its efficacy (Li et al., 2013).This also is helpful for our further study of PD medicinal drugs.
Although mitochondrial transcription factor A is not directly involved in PD pathogenesis, it is difficult for us to observe complete PD pathology in the MitoPark mouse model.However, this model has a similar onset time line to human PD, which can help us to analyze early PD symptoms and test PD treatment methods (Beckstead and Howell, 2021).Therefore, the MitoPark mouse model has great potential for development and deserves our further exploration.

Neurotoxin-induced animal models incorporating transgenes
We have come to understand that single neurotoxin, or TG, models possess inherent limitations and are unable to fully replicate the symptoms of PD observed in humans.In the pursuit of more faithfully mimicking human PD and meeting the demands of research, neurotoxins have been observed to significantly affect TG animals in comparison to their normal counterparts.Consequently, the effect of animal model preparation is more pronounced in TG animals.For instance, when DJ-1 TG mice were subjected to modeling through a combination of DJ-1 overexpression and administration of the neurotoxin MPTP, the resulting mice exhibit more pronounced degeneration of DA neurons and increased cell death compared to modeling using a single method alone (Heinemann et al., 2016).
Moreover, prolonged and chronic administration of rotenone injections in α-syn TG mice leads to the degeneration of DA neurons in the SNpc and CS, alongside the aggregation of α-syn and the emergence of dyskinesia associated with PD (Ng et al., 2009).These findings underscore the significance of employing integrated approaches and neurotoxin-induced models to more accurately recapitulate the complexity of PD in humans, thus enabling further exploration and comprehension of this debilitating neurodegenerative disorder.

α-syn PFF-induced PD animal model
LB generation in neurons represents a major pathological hallmark of PD, arising from the aggregation of α-syn.This toxic and insoluble fibril constitutes a principal component of LB.Some studies have shown that α-syn adopting this fibrillary conformation acts as a "seed," capable of diffusing and transmitting between cells.It continuously binds to endogenous α-syn monomers and accelerating the aggregation of α-syn, leading to the formation of LB at multiple sites, eventually leading to downstream neuroinflammation and nerve degeneration, and exacerbating PD (Duffy et al., 2018;Thomsen et al., 2021).Therefore, researchers have initiated investigations into the method of injecting α-syn PFF to establish PD animal models.
S. He et al.

C. elegans
The whole body is transparent, easy to culture, the nervous system is simple and complete, the experimental cost is low, and the experimental period is short.

S. He et al.
According to the TG animal model described previously, both forms, mediated by distinct promoters and viral vectors, have the capability to induce overexpression of endogenous α-syn.Although the TG model has improved our understanding of α-syn toxicity, the α-syn overexpression model tends to present an abnormal increase in overall protein levels, surpassing those observed in idiopathic PD.This pathological mechanism, operating beyond physiological thresholds, may induce the manifestation of additional redundant symptoms distinct from idiopathic PD.The animal model established via injection of α-syn PFF effectively simulates preclinical idiopathic PD.In addition, it can aid in the exploration of PD treatment through the study of the toxicity traits of endogenous α-syn.
The α-syn PFF animal model is commonly employed in rats and mice.
Similar to the 6-OHDA animal model, it entails unilateral injection of α-syn CS in the brain, leading to varying degrees of α-syn aggregation in the CS and the regions connecting CS to the brain, subsequently resulting in degeneration of DA neurons in the SNpc and CS (Tayler and Stowers, 2021).Unlike the 6-OHDA animal model, unilateral injection of α-syn PFF leads to gradual loss of function in ipsilateral but not contralateral CS DA neurons.Studies have shown a high sensitivity of SNpc DA neurons to α-syn PFF (Stoyka et al., 2020).The molecular size of α-syn PFF is a crucial determinant for modeling PD with α-syn PFF (Froula et al., 2019).Following experimentation, α-syn PFF with a molecular size ranging from 29 to 49 nm exhibited optimal modeling efficacy (Abdelmotilib et al., 2017).In addition, it is recommended to employ an injection concentration of 1 µg/mL of α-syn PFF (Marden et al., 2007).Upon injecting 12-month-old mice with 1 µg/mL of α-syn PFF, a significant increase in ROS in the brain was observed 24 hours post-injection.A behavioral evaluation performed two months later revealed typical Parkinsonian symptoms in the mice.Similarly, rats treated with this concentration for 24 hours exhibited slow degeneration of DA neurons due to ROS production (Verma et al., 2021;Ghosh et al., 2022;Murphy et al., 2022).To expedite the animal model preparation process, a research team employed a combination of α-syn PFF and a low dose of MPTP, administering injections into the CS once daily for five consecutive days.After 6 weeks, a significant loss of DA neurons was observed in the model brains (Merghani et al., 2021).This model serves as a reference for expediting the progression of synuclein-related diseases within the framework of normal α-syn levels, which may be applicable in the study of other synuclein-related diseases (Duffy et al., 2018).The α-syn PFF model may be more closely associated with idiopathic PD in terms of the toxicity and neuroinflammation resulting from α-syn aggregation.Additionally, further exploration and utilization of this model are warranted

Non-human primate animal models
Due to the advanced nature of experimental technology, establishing NHP animal models require substantial investments in both financial and labor resources.Additionally, its usage is limited by ethical considerations, and only a few studies have utilized NHP animal models.Nevertheless, compared with other model organisms, NHPs can simulate human PD symptoms more accurately and comprehensively, thereby providing superior avenues for investigating clinical treatment modalities (Qu et al., 2023).Compared with rodents and other organisms, NHPs exhibit a closer resemblance in brain structure to humans (Amiez et al., 2019;Garin et al., 2022;Zheng et al., 2022).In addition, the type and distribution of brain cells in NHPs are more complex than those in other model organisms, and the normal physiological activities and functional patterns of the brain are more similar to those of humans (Hickman et al., 2018;Geirsdottir et al., 2019;Lee and Heiman, 2022).As a close evolutionary relative of humans, NHPs remain highly similar to humans in terms of genetic, physiological, and behavioral diversity.Therefore, when examining PD-related cognition and behavior, the data obtained from NHP models provide more accurate insights into human PD (Frye et al., 2022).
For an extended period, PD was considered exclusive to humans, however, some studies have revealed analogous symptoms of early PD in humans in old monkeys (Hurley et al., 2011).In addition, significant synucleinopathy has been observed in certain aging NHPs, with naturally occurring PD cases also documented among monkeys (Li et al., 2021).These findings underscore the significant potential of NHPs in PD research.Commonly employed NHP models include macaques, and baboons, with PD models established through either neurotoxin induction or TG techniques.Given the prolonged experimental period and substantial labor and financial investments required for establishing transgenic NHP PD models, chemically induced NHP models are more frequently utilized (Pan et al., 2024).MPTP-induced NHP models have been widely employed in preclinical studies, with older monkeys exhibiting a greater susceptibility to infection compared to their younger counterparts (Huang et al., 2018).Some studies (Shi et al., 2019) have note the instability of older rhesus monkeys following intramuscular MPTP injection.However, subsequent adjustments, have led to young rhesus monkeys exhibiting motor symptoms, such as slow motion, resting tremor, and the non-motor symptoms, such as GI dysfunction, drowsiness, when intramuscular MPTP injection is combined with intravenous MPTP injection for 18 weeks (Shi et al., 2020).
However, neurotoxin-induced models manifest a relatively unstable phenotype, challenging the replication of the slow progression of PD.To address this issue, some studies have employed stereotaxic injection of MPP+ into different regions of the unilateral SNpc guided by magnetic resonance imaging (Lei et al., 2015).Subsequent experiments, have identified α-syn in the MPTP-induced monkey model of PD (Huang et al., 2018), which may advance the development of PD animal models.PINK1 expression has been found to be higher in NHPs compared to mouse models, suggesting that PINK1 may play a more vital role in NHPs than in mouse model (Yang et al., 2019a).Subsequent experiments, have demonstrated that knockdown of the PINK1 gene fragment via TG technology results in severe neuronal loss in the cerebral cortex of monkeys (Yang et al., 2019b(Yang et al., , 2020)).

Caenorhabditis elegans (C. elegans) models
With its relatively simple evolutionary lineage, C. elegans exhibits characteristics such as a transparent body, facile culturing, and a simple yet complete nervous system.C. elegans has emerged as an ideal model for studying the mechanism of neurodegenerative diseases, owing to the high degree of genetic homology with humans (60-80 % shared genes) and its cost-effectiveness and short experimental turnover, it usually takes a few days to show a stable phenotype.(Williams et al., 2017;Cooper and Van Raamsdonk, 2018;Chandler et al., 2021).C. elegans harbors various pathogenic genes homologous to PD, which can be used to establish different transgenic PD models, however it lacks expression of the SNCA.Therefore, a PD model that is not affected by endogenous α-syn can be established by overexpressing the SNCA via TG techniques, which can facilitate the elucidation of the interactions between α-syn and DA neurons (Mor et al., 2017).C. elegans overexpressing wild-type with SNCA showed motor defects as well as non-motor symptoms sensitive to stress (Cooper and Van Raamsdonk, 2018).Some studies have found that the C. elegans overexpressed wild-type with LRRK2 has better resistance to oxidative stress and a longer life span compared with the control group (Saha et al., 2009).In addition, PD in C. elegans can be also induced using various neurotoxins.C. elegans's transparency, allows for the visualization of components of interest using fluorescent proteins, facilitating comprehension of their secretion and action pathways.This model aids in the investigation of various diseases (Corsi et al., 2015).Although the C. elegans model significantly aids in the study of PD at the genetic level, its neural pathways differ substantially from those of humans due to the organism's limited cell count.Therefore, emphasis should be placed on the advantages of employing this model to compensate for its limitations (Cooper and Van Raamsdonk, 2018).

Drosophila models
Drosophila possesses a small number of genomes, yet its nervous system shares many structural and functional features with mammals.Approximately 70 % of human disease-related genes have homologs in Drosophila (Pandey and Nichols, 2011).Similar to C. elegans, Drosophila lacks expression of the SNCA, resulting in the absence of endogenous α-syn inclusion bodies.Consequently, this model relies on the importation of the SNCA through TG methods.Through TG techniques, the target gene can be imported, silenced, or overexpressed to establish a Drosophila model that meets the experimental requirements.By expressing the human SNCA, adult Drosophila exhibit DA neuron degeneration, motor deficits, and reduced lifespan (Dabool et al., 2019).The Drosophila model effectively replicates typical features of PD, including motor deficits, progressive loss of DA neurons and oxidative stress (Hewitt and Whitworth, 2017).In addition, the Drosophila model has a short experimental lifespan, which can effectively simulate the progression of age-related diseases such as PD.Drosophila models are suitable for large-scale genetic and drug screening, thereby enhancing our understanding of the genetic and molecular basis of diseases and expediting clinical drug development (Fernández-Hernández et al., 2016).Nevertheless, flies lack an immune system, which may lead to other pathophysiological reactions unrelated to PD.Therefore, to attain a more comprehensive understanding of PD progression, complementary studies utilizing the Drosophila model as well as other models are warranted (Suzuki et al., 2022).

Induced pluripotent stem cell (iPSC)-based models
Advancements in iPSC research, have made it feasible to establish a PD model using somatic cells from human patients.Animal models primarily simulate late-onset PD in humans; however, they fail to fully replicate the entire pathological process.Moreover, inherent physiological and anatomical differences exist between animals and humans.Consequently, the models derived from humans hold significant value for studying the disease (Stoddard-Bennett and Pera, 2020;Outeiro et al., 2021).iPSCs exhibit identical genetic profiles to those of the patients, enabling the creation of a model that authentically mirrors disease progression in patients, thereby advancing PD research (Liu et al., 2020;Sivandzade and Cucullo, 2021).
At present, two primary methods are employed for constructing such models.Firstly, iPSCs can be subjected to a series of regulatory and differentiation processes to simulate the growth environment of neurons in vivo.This process, ultimately directs iPSCs towards differentiation into DA neurons, establishing a 2D model (Kriks et al., 2011).After iPSC differentiated into neural stem cells, neural differentiation was further promoted and nerve clusters or rosettes were able to form within 11 days (Chambers et al., 2009).Although this 2D DA neuron cellular model cannot replicate the complex neural pathways of the intact brain, its operation is relatively rapid and cost-effective compared to animal models, thereby facilitating advancements in clinical drug research.
Alternatively, iPSCs can be directed to differentiate into progenitor cells specific to midbrain organoids (MOs), which are subsequently subjected to 3D culture phylogeny to develop mature MOs structures (Kelava and Lancaster, 2016).At present, many 3D culture methods are developed from 2D culture, and the transformation from 2D model to 3D model usually takes ten days to several months (Kim et al., 2024).MOs can maximally recapitulate the original organ in terms of cellular diversity, structural organization, and functionality, thereby enhancing the fidelity between experimental observations and human physiological conditions (Chlebanowska et al., 2020;Kim et al., 2024).This model effectively, elucidates the mechanism underlying the demise of DA neurons, and offer advantages in evaluating clinical drugs.MOs accurately simulate the primary symptoms and fundamental characteristics of human PD (Chia et al., 2020), including those associated with SNCA (Patikas et al., 2023), LRRK2 (Kim et al., 2019), PRKN (Dong et al., 2022), and PINK1 (Jarazo et al., 2022), as demonstrated in numerous contemporary investigations.

Optogenetic models
The discovery of optogenetic-related proteins has spurred interest in utilizing them for dynamically regulating pathological proteins implicated in neurodegenerative disease regulation.Recent studies have found that optogenetic regulation of the MOS model can reduce the spatial separation between α-syn monomers, thereby accelerating the aggregation of α-syn and eventually leading to PD.This method also exhibits promising efficacy in an in vitro iPSC cell model (Kim et al., 2023).This method addresses the prolonged duration inherent in iPSC-based models and introduces a novel approach to the development of PD models.

Gene editing-based models
The rapid advancement in molecular research has significantly enhanced the efficiency of DNA and RNA sequence analysis, thereby accelerating progress in gene editing technology.In brief, gene editing is primarily achieved through the targeted action of specific endonucleases on the target DNA to induce double-strand breaks, followed by the replacement or modification of the target gene with a donor sequence (Zhu et al., 2021).Gene editing can be employed to generate cell and animal models, aiding in the validation of genes associated with PD and the advancement of treatment methods (Mansour and El-Khatib, 2023).Gene editing technology allows us to further understand the relationship between phenotype and heredity and holds promise for gene therapy applications.Moreover, it furnishes novel approaches for generating pathological models.

Discussion
PD is a complex, chronic neurological disorder characterized by its progressive nature.Understanding the mechanisms underlying PD pathogenesis and developing effective treatment and prevention strategies necessitate the establishment of an animal model that closely mimics the symptoms observed in human PD.Existing neurotoxininduced animal models of PD possess limitations.They yield restricted biochemical responses and lack the distinctive pathological changes typical of PD.Moreover, these models employ highly toxic agents that inflict permanent damage on the animals' nervous systems, impeding the evaluation of the therapeutic effects of PD drugs.To date, many clinical drug studies have failed due to their reliance on exogenous neurotoxin models.These models induce rapid and extensive destruction of DA neurons contrasting with the slow progression of idiopathic PD in humans.Therefore, some scholars propose an alternative approach, employing endogenous neurotoxins to establish models that manifest slow-onset symptoms similar to idiopathic PD in human, thereby providing a congruent framework for preclinical drug investigations (Segura-Aguilar and Mannervik, 2023).Among them, aminochrome emerges as the most promising endogenous neurotoxins.Aminochrome, synthesized in DA neurons and exhibiting a limited propensity for spreading to neighboring neurons, induces neuronal dysfunction reminiscent of idiopathic PD in humans (Huenchuguala and Segura-Aguilar, 2024), necessitating a prolonged experimental period.The continued refinement of this model is anticipated.
In recent years, TG PD animal models have emerged as a promising alternative.Advancements in TG technology have refined these models, but they come with their own set of challenges.They are costlier, require more time for generation, and involve complex procedures that may result in unintended side effects during the modeling process.TG PD animal models offer greater flexibility in terms of species selection, primarily employing rodents such as mice and rats.Other organisms, such as dogs, cats, nematodes, fruit flies, and zebrafish, have also served as model organisms.However, it is worth noting that ethical considerations and cost factors are integral to selecting an appropriate model throughout the experiment.Genetic manipulation of familial PD related genes will facilitate a deeper understanding of the genetic underpinnings of PD.However, the incidence of familial PD represents about 10 % of PD cases, indicating its minor prevalence within the PD spectrum.This understanding will guide the trajectory of future research endeavors.
Excluding these factors, NHPs hold significant promise for PD research owing to their close genetic and physiological resemblance to humans.NHPs exhibit the highest degree of similarity to humans among animal models and can accurately replicate the clinical symptoms observed in human PD.Through the study of NHP models, researchers can attain realistic and reliable data regarding both the underlying pathological mechanisms and the efficacy of therapeutic drugs.This shift towards utilizing NHP models signifies a progression towards more sophisticated and comprehensive animal models for PD research.In addition, recently developed models such as MitoPark mouse model, α-syn PFF models, iPSC-based models, optogenetic models, gene editing models, and other advanced models, also exhibit substantial potential to advance PD research.

Conclusions and Future Directions
Currently, the lack of standardized protocols for PD modeling poses challenges in establishing an optimal animal model for PD research.An ideal PD animal model should feature normal DA neurons at the outset, followed by selective degeneration or absence of these neurons after modeling.Moreover, the model should manifest α-syn aggregation leading to LB formation alongside comprehensive PD motor deficits.Additionally, it is preferable for the disease progression in the model to occur within a reasonable timeframe.
With the ongoing advancements in biomedical technology, the development of an optimal animal model for PD has become a feasible goal.Scientists are diligently working towards refining and improving existing models to better replicate the characteristics and progression of PD in humans.It is anticipated that in the near future, these efforts will yield results, enabling researchers to address the challenges posed by PD and ultimately contribute to the overall welfare of humanity.

Fig. 1 .
Fig. 1.The mainstream method of neurotoxin-induced animal models of PD.After crossing the BBB, MPTP acts on astrocytes and reacts with MAO-B to generate MPTP ions.MPTP ions are further oxidized to MPP+, which can be recognized by DAT and enter DA neurons, producing a large number of ROS and eventually causing damage to DA neurons.After targeted injection into the brain, 6-OHDA enters DA neurons via DAT, and increases the production of ROS and inhibits the production of antioxidant enzymes through autooxidation and metabolic processes, 6-OHDA eventually leads to the degeneration and demise of DA neurons.Upon entry into the brain, rotenone inhibits the activity of mitochondrial complex enzyme I, induces the generation of a large amount of ROS, disrupts DA metabolism, and impairs mitochondrial function.Ultimately, DA neurons and cells undergo degeneration and death.PQ cannot effectively cross the BBB and is typically used to establish PD animal models through multiple injection, PQ reaches the SNpc and selectively destroys DA neurons, thereby manifesting PD symptoms.[Illustration created with Fig draw].

Fig. 2 .
Fig. 2. The mainstream method of TG animal models of PD.SNCA encompasses three mutation sites, namely A53T, A30P, and FA6K.Any alteration at these sites results in the presentation of a β-fold conformation in α-syn.This aberrant β-structure leads to α-syn aggregation, which cannot be efficiently cleared, thereby leading to mitochondrial dysfunction and neuronal cell demise, thereby manifesting PD symptoms.LRRK2, PRKN, PINK1 and DJ-1 gene mutations can cause varying degrees of mitochondrial dysfunction, and eventually lead to PD symptoms.The changes of DJ-1 gene can also cause the aggregation of α-syn.Whether the changes of the other three genes also cause the aggregation of α-syn needs to be further verified.[Illustration created with Fig draw].

Table 1
Comparative analysis of different animal models of PD.
(continued on next page) S.He et al.