One of the promises of genetic association studies is better animal models for research and for development of therapeutics. Such promises are often slowly fulfilled. Twelve genetic loci that can cause Parkinson disease (PD) or parkinsonism have been identified in the last decade, but a model that fully recapitulates the pathology and behavior of the human disease has yet to emerge. Now, the latest transgenic mouse comes close. In the June 7 Nature Neuroscience, researchers led by Chenjian Li at Cornell University, New York, debut a model based on the LRRK2 gene, which was identified in 2004 (see ARF related news story). The mouse exhibits many of the defining characteristics of PD, including pathology in the axons that extend from the substantia nigra (the site of dopaminergic neuron loss in PD) to the striatum, and age-dependent movement problems that respond to L-DOPA. The latter “suggests a possible overlap with mechanisms seen in human PD patients,” wrote Mark Cookson, National Institute of Health, Bethesda, Maryland, in a comment to Alzforum (see below). “These mice provide a valid model of Parkinson disease and are a resource for the investigation of pathogenesis and therapeutics,” write the authors.

First author Yanping Li and colleagues used bacterial artificial chromosomes (BACs) containing human mutated LRRK2 (R1441G mutation that causes PD) to generate transgenic mice. The mice developed normally despite robust expression of human LRRK2 in several areas of the brain, such as cortex, cerebellum, striatum, and the ventral midbrain. By 10-12 months motor impairment was apparent, however, which cleared up with treatment with L-DOPA or the dopamine agonist apomorphine. Li and colleagues used a dopamine re-uptake blocker and microdialysis to confirm that the animals released significantly less dopamine compared to wild-type animals. Histological analysis showed that axons of dopaminergic neurons looked abnormal, containing spheroids and dystrophic neurites. The researchers also found that those axons stained positive for hyperphosphorylated tau, which has also been found in patients with the LRRK2 mutation (for more on tau in PD, see ARF spectrum series story). One thing the mice do not seem to recapitulate is the neuronal loss seen in PD. For more on this model, read Q&A on PD Online Research with senior author Chenjian Li.—Tom Fagan


  1. A progressive, and eventually dramatic, depletion of dopamine-containing neurons that project from the substantia nigra pars compacta to the striatum is one hallmark of Parkinson disease. Nigral cell loss is thought to be the pathological event leading to many of the movement problems in PD and is the basis of mainstay therapy using L-DOPA for many patients. Several toxins, including 6-OHDA, MPTP and rotenone, have been used for several years to induce dramatic damage to the substantia nigra as a model of parkinsonism (1). Although there are some problems with these models, they have been very important in understanding mechanisms related to denervation in the basal ganglia. In contrast, genetic-based models have produced generally mild and sometimes inconsistent phenotypes (2). Furthermore, none that I know of produce a dramatic lesion of the nigra or a particularly strong behavioral output. Delivery of α-synuclein, a gene thought to be causal for PD, directly to the substantia nigra using viral vectors produces the most striking cell loss phenotype (3) but even this is incomplete. One can argue how critical it is to have this specific phenotype, especially as we know that PD involves many different brain regions, but most would acknowledge that an L-DOPA responsive phenotype based on nigral cell loss would at least be a preferred effect of genetic manipulation.

    Into this history comes the genetically modified mouse from the Cornell group (4) showing a dramatic phenotype of very slow spontaneous movement. The LRRK2 R1441G mutation used here was first found in a series of families from the Basque region of Northern Spain (5) and is definitively pathogenic. There are several important aspects to this paper by Li et al; first, it supports the contention that mutant LRRK2 can be toxic in vivo as previously suggested in cell models; second, a strong and age-related behavioral effect is seen in the presence of moderate cell loss and axonal damage; third, the behavioral output is rescued by L-DOPA, suggesting a possible overlap with mechanisms seen in human PD patients. Although not always compared side-by-side, mice overexpressing wild-type LRRK2 do not show any of these effects (4).

    Given the strong effects, it is strongly hoped that other researchers are able to access these or similar mouse models and this is apparently in place (6). Comparison with other pathogenic mutations is a critical next step, as is the important requirement to show that phenotypes are stable across multiple generations of this line, which would make them very useful. One surprise that needs to be resolved is where the behavioral phenotype is localized anatomically. Returning to the point about toxin models, in 6-OHDA lesions a very substantive loss of nigral projections can be achieved (up to 90 percent in some paradigms) without necessarily inducing the same level of akinetic behavior. This might indicate that other brain circuits are involved. One might propose that the striatal target cells are also damaged in this model; this seems highly unlikely given that L-DOPA is effective, and this drug is thought to require an intact striatum to work. Excitingly, the Li et al. mice now give us a platform where such experiments can feasibly be performed.

    See also: 

    PD Online Research


    . Parkinson's disease: genetic versus toxin-induced rodent models. FEBS J. 2008 Apr;275(7):1384-91. PubMed.

    . Strengths and limitations of genetic mouse models of Parkinson's disease. Parkinsonism Relat Disord. 2008;14 Suppl 2:S84-7. PubMed.

    . Viral vectors, animal models and new therapies for Parkinson's disease. Parkinsonism Relat Disord. 2008;14 Suppl 2:S169-71. Epub 2008 Jun 27 PubMed.

    . Mutant LRRK2(R1441G) BAC transgenic mice recapitulate cardinal features of Parkinson's disease. Nat Neurosci. 2009 Jul;12(7):826-8. PubMed.

    . Cloning of the gene containing mutations that cause PARK8-linked Parkinson's disease. Neuron. 2004 Nov 18;44(4) PubMed.

  2. LRRK2 R1441C Mice Recapitulate PD Motor Deficits
    The recent article by C.J. Li’s group reports on the first transgenic LRRK2 mouse model. The LRRK2 mouse uses a BAC system to express WT or R1441C LRRK2, which is a mutation in the GTPase domain of LRRK2. The benefit of the BAC system is that it permits use of the LRRK2 promoter, which allows for an expression pattern that is elevated but exhibits a distribution that recapitulates the pattern of endogenous LRRK2. The mouse is notable in several respects. The most important observation is that the mouse exhibits age-dependent motor deficits that are responsive to L-DOPA, which is a classic phenotype observed in patients with Parkinson disease and shows that the motor deficits derive from dysfunction of dopaminergic neurons. The nature of the dysfunction, though, is not entirely clear. The group reports a modest decrease in dopamine release from the neurons. This phenotype is reminiscent of phenotypes observed for parkin and PINK1 mice, and might be a preliminary indication that deficits in dopamine release are a common feature of genes associated with PD. A second encouraging observation is that mice show signs that are commonly associated with neuronal degeneration. The researchers observe tyrosine hydroxylase positive spheroids, and increased tau phosphorylation, shown with the AT8 antibody.

    These degenerative phenotypes are interesting for what they do and do not show. On the positive side, the morphologic changes are consistent with a hypothesis that there is degeneration of dopaminergic neurons. Spheroids are a common, non-specific sign of neurodegeneration. Tau phosphorylation is also an early sign of neuronal degeneration, or at least neuronal stress. From my perspective, the evidence of dopaminergic degeneration is quite believable because it is consistent with what we observe in our C. elegans model, which has been in review for over a year and hopefully will be out soon. The presence of phospho-tau reactivity is also interesting because neurofibrillary tangles are observed in some cases of LRRK2-associated parkinsonism, and because polymorphisms in tau are strongly associated with Parkinson disease, being the second strongest result observed in multiple GWAS studies after those in α-synuclein. On the negative side, the manuscript is striking for the absence of any mention of loss of dopamine neurons or the presence of α-synuclein inclusions or Lewy body pathology. This might mean that expressing mutant LRRK2 alone is insufficient to recapitulate Parkinson disease and one needs to also have the human proteins present, such as α-synuclein and tau. Alternatively, this might reflect the vagaries of transgenic models, and some other models might show α-synuclein pathology (although I am not aware of any reports of α-synuclein pathology in the other mouse models that are in development). Nevertheless, the presence of a L-DOPA responsive transgenic mouse model derived from a genetic mutation associated with Parkinson disease represents an important advance for the field.

    View all comments by Benjamin Wolozin

Make a Comment

To make a comment you must login or register.


News Citations

  1. PARK8 is Cloned: Introducing…"Dardarin"
  2. Et tu, Brute? Parkinson’s GWAS Fingers Tau Next to α-Synuclein

External Citations

  1. PD Online Research

Further Reading

Primary Papers

  1. . Mutant LRRK2(R1441G) BAC transgenic mice recapitulate cardinal features of Parkinson's disease. Nat Neurosci. 2009 Jul;12(7):826-8. PubMed.