Abstract
Parkinson’s disease (PD) is a common neurodegenerative disease with a heterogeneous etiology that involves genetic and environmental factors or exogenous. Current LRRK2 PD animal models only partly reproduce the characteristics of the disease with very subtle dopaminergic neuron degeneration. We developed a new model of PD that combines a sub-toxic MPTP insult to the G2019S-LRRK2 mutation. Our newly generated mice, overexpressing mutant G2019S-LRRK2 protein in the brain, displayed a mild, age-dependent progressive motor impairment, but no reduction of lifespan. Cortical neurons from G2019S-LRRK2 mice showed an increased vulnerability to stress insults, compared with neurons overexpressing wild-type WT-LRRK2, or non-transgenic (nTg) neurons. The exposure of LRRK2 transgenic mice to a sub-toxic dose of MPTP resulted in severe motor impairment, selective loss of dopamine neurons and increased astrocyte activation, whereas nTg mice with MPTP exposure showed no deficits. Interestingly, mice overexpressing WT-LRRK2 showed a significant impairment that was milder than for the mutant G2019S-LRRK2 mice. L-DOPA treatments could partially improve the movement impairments but did not protect the dopamine neuron loss. In contrast, treatments with an LRRK2 kinase inhibitor significantly reduced the dopaminergic neuron degeneration in this interaction model. Our studies provide a novel LRRK2 gene-MPTP interaction PD mouse model, and a useful tool for future studies of PD pathogenesis and therapeutic intervention.
Introduction
Parkinson’s disease (PD) is the second most prevalent neurodegenerative disease after Alzheimer’s disease, and affects 5–10 million people worldwide (1). PD is characterized by tremor, rigidity and bradykinesia due to a severe and selective loss of dopaminergic neurons of the subtantia nigra (2,3). The disease is progressive and fatal, and no cure is currently available. Existing treatments, such as L-DOPA, are only effective in controlling the symptoms. Most cases of PD are sporadic, but mutations in specific genes have been identified in a familial form of the disease.
Mutations in the PARK8 gene encoding for Leucine-rich repeat kinase 2 (LRRK2) are the most common cause of familial PD (4). The most prevalent mutation, G2019S, causes ~1% of idiopathic PD in North America and 5 % of familial PD (4). LRRK2 is a large multi-domain protein with GTPase and kinase activities. The G2019S mutation has been shown to increase the kinase activity of LRRK2 (5). The LRRK2 locus has also been linked to an increased risk of apparently sporadic PD in several genome-wide association studies across different populations, likely via alterations in expression level (6–9). We have previously reported a clear neurotoxicity in cell models by expressing mutant (but not wild-type) LRRK2, suggesting the importance of kinase activity for cell toxicity (10,11).
The pathogenesis of PD remains incompletely understood, but both genes and environment are believed to contribute (12,13). Among environmental factors, exposure to pesticides (14), toxins, pathogens (15,16) or head trauma (17) have been proposed to contribute to neurotoxicity, possibly via oxidative stress (18) or mitochondrial dysfunction (19,20). There is increasing evidence of the interaction between genes and environment in the physiopathology of PD. However, lacking a reliable mouse in vivo model prevents further dissecting of the mechanisms underlying the gene-environmental interaction in LRRK2-linked PD pathogenesis and therapeutic development.
So far, most animal models of PD focus on the either sporadic PD or familial PD. Several groups reported transgenic mouse models using mutant LRRK2 (such as R1441C, R1441G and G2019S) which demonstrated relatively mild, but interesting phenotypes with subtle dopaminergic neuron changes and no shortened lifespan compared with control mice (21–26). LRRK2 KO mice failed to show any neurodegeneration or deficiency of the dopaminergic system (27), and they displayed an unchanged sensitivity to MPTP (28). Recently, a catecholaminergic neuron-specific, Tet-inducible, conditional transgenic LRRK2-G2019S mouse showed a robust phenotype with a cell autonomous kinase-dependent neurodegeneration of dopaminergic neurons (29). Other existing models of PD rely on the use of toxic compounds such as 6-OHDA (30,31) and large toxic doses of MPTP (30–33) that can induce a selective degeneration of dopaminergic neurons in a short time. However, there are substantial behavioral variations in these models without production of genetic vulnerability (33), which prevents them from being useful for drug development with LRRK2-related compounds.
In this study, we investigated the interaction between a toxin, MPTP and genetic factors of PD both in vivo and in vitro. We developed a new mouse model of PD that combines treatment with a sub-toxic dose of MPTP with an overexpression of PD-linked mutant LRRK2. We further determined this model to have robust neurodegeneration and impairment of motor activity, resembling PD-like symptoms, by pathological and behavioral approaches. This study also provides a novel, in vivo, model for further studies of PD pathogenesis and therapeutics.
Results
Generation of transgenic mice overexpressing human mutant G2019S-LRRK2 protein in brains
To generate G2019S transgenic mice, we cloned a full-length human LRRK2 cDNA, with a Flag tag in its N-terminal, into a mouse prion protein promoter (MoPrP) vector. This vector was previously successfully used for transgenic models, including our HD and DRPLA transgenic models, both of which have robust, progressive, fatal neurodegenerative phenotypes (34–37). Additionally, the pattern of expression is similar to LRRK2 and includes high levels in the brainstem. The clone was engineered and confirmed to carry the G2019S mutation (Fig. 1A), which is the most common missense mutation in large family pedigrees, with nearly complete penetrance. Among independent transgenic mouse lines, we identified and selected the heterozygous 243 and 231 lines, expressing highest LRRK2 protein.
Generation of human wild-type WT-LRRK2 and G2019S-LRRK2 transgenic mice. (A) Flag-WT-LRRK2 or G2019S-LRRK2 gene was cloned into a mouse prion promotor (MoPrP) vector to generate transgenic mice expressing LRRK2 variants predominantly in neurons. (B) Lysates derived from whole brain of nTg, WT-, and G2019S-LRRK2 (Flag-tagged) heterozygous (Hetero) and homozygous (Homo) transgenic mice were analyzed by western blot using anti-LRRK2 antibodies. (C) Quantitation of LRRK2 protein levels in transgenic mice normalized to nTg mice based on three independent western blot experiments using anti-LRRK2. The LRRK2 protein level in the brain of G2019S homozygous transgenic mice is about 2–3-fold of heterozygous mice and about 5-fold of nTg mice. (D) Average weight of the mice at 4, 9 and 12 months. No difference has been observed by ANOVA.
To generate the homozygous G2019S mice, we mated heterozygous 243 lines with each other and validated the homozygous line by genotyping their offspring. By western blots of whole brain homogenates performed in parallel on extracts prepared from the heterozygous transgenic line and non-transgenic (nTg) controls, we confirmed the homozygous G2019S mice expressed LRRK2 approximately 5-fold above the level of endogenous mouse LRRK2 and 2- and 3-fold above the level of LRRK2 in heterozygous G2019S mice (Fig. 1). All the generated mice were fertile, did not present any gross abnormalities and had comparable weight to their nTg littermates throughout their lives (Fig. 1D, One-way ANOVA analysis, 4 months: F(2, 33) = 0.09813; 9 months: F(2, 33) = 0.007019; 12 months: F(2, 33) = 0.2754).
G2019S-LRRK2 transgenic mice display progressive movement impairment with aging
To assay the effect of the LRRK2 mutation on the mice, we measured the behavioral performance of the animals throughout their lives. The homozygous G2019S mice did not present any gross abnormalities or reduction in their life expectancy. The general activity of the mice, and the rearing activity measured using open field testing were reduced in the transgenic mice at 6 months of age. This deficit was increased at 12 months (Fig. 2A). In addition, the motor performance of the transgenic mice, when assayed using the accelerating rotarod protocol, showed a significant decrease at 9 and 12 months. The deficits in the G2019S-LRRK2 transgenic mice appeared to be progressive and age dependent since the animals did not show any difference at 4 months (Fig. 2B). Even at 12 month of age, the WT-LRRK2 transgenic mice had subtle decreases in performance, but there was no statistical significance compared with none transgenic mice.
G2019S-LRRK2 mice display motor impairment with aging. (A) Locomotor activity was monitored by open field testing for transgenic mice at ages of 6 and 12 months. Twenty-four mice in each group, half male and half female. G2019S mice showed significant reductions in both locomotor activity and rearing activities. P < 0.05 by ANOVA versus nTg mice. (B) Rotarod testing was done for mice at ages of 4, 9 and 12 months. n=24, half male and half female. G2019S mice showed an age-dependent impairment. T-test, ***P < 0.001 by ANOVA versus nTg mice.
Primary neurons from G2019S-LRRK2 mice cortex were more vulnerable to stress insults
To test whether neurons from G2019S-LRRK2 transgenic mice altered the cell viability to various stress insults, primary cortical neurons at 7 days in vitro were exposed to various stressors for 48 h. The survival percentage was measured by nuclear condensation assay. Stressors included an autophagy inhibitor—3MA (0.5–5 μM), an excitotoxic stressor—NMDA (1–10 μM), a depolarization agent—KCl (1–20 mM), an ER stressor—DTT (0.5–5 μM), a mitochondrial stressor—MPTP-derived, MPP+ (0.1–1 μM) and 24 h deprivation of B27, a trophic factor cocktail from the culture medium.
For each specific kind of stress, G2019S-LRRK2-derived neurons showed a greater vulnerability to stressors than the nTg neurons or the neurons derived from transgenic expressing WT-LRRK2 (Fig. 3). Neurons expressing G2019S-LRRK2 were vulnerable to all stressors. In contrast, neurons overexpressing WT-LRRK2 were only vulnerable to DTT (Fig. 3A, Two-way ANOVA: interaction F(4, 45) = 2,977; dose: F(2, 45) = 28,60; genotype: F(2, 45) = 125,3), MPP+ (Fig. 3F, Two-way ANOVA: interaction F(8, 75) = 5,377; F(4, 75) = 159,2; genotype: F(2, 75) = 38,47) and B27 deprivation (Fig. 3E, One-way ANOVA: F(2, 15) = 18,53) than nTg control neurons, but they showed no change to 3MA (Fig. 3A, Two-way ANOVA: interaction F(4, 45)= 0.7679; dose: F(2, 45) = 12,47; genotype: F(2, 45) = 35,79), NMDA (Fig. 3B, Two-way ANOVA: interaction F(4, 45) = 16,89; dose: F(2, 45) = 116,4; genotype: F(2, 45) = 60,61) and KCl insults (Fig. 3C, Two-way ANOVA: interaction F(4, 45) = 9,851; dose: F(2, 45) = 40,11; genotype: F(2, 45) = 41,28) compared with nTg neurons. These results demonstrate stress insults that cause more cell death in neurons expressing PD mutant G2019S-LRRK2 than controls, suggesting a genetic mutation could enhance the effects of environmental stressors.
Primary cortical neurons from LRRK2 transgenic mice were more vulnerable to stress insults. Primary cortical neurons isolated from nTg, LRRK2 and G2019S-LRRK2 at 7 DIV were treated with different stressors. (A) Autophagy inhibitor, 3MA, (B) excitotoxic stressor, NMDA, (C) depolarization agent, KCl and (D) ER stressor, DTT were applied at the indicated concentrations for 48 h. (E) Primary neurons were deprived of trophic factor supplement, B27 in the culture medium for 24 h. (F) Primary neurons were exposed to increasing concentration of MPP+ for 48 h. Quantification of cell death was done by nuclear condensation assay. Two-way ANOVA analysis, *P < 0.05 versus nTg, **P < 0.01 versus nTg and ***P < 0.001 versus nTg; #P < 0.05 versus G2019S, ##P < 0.01 versus G2019S and ###P < 0.001 versus G2019S.
LRRK2 transgenic mice treated with a sub-toxic dose of MPTP developed robust motor impairment and dopamine neuron degeneration
To investigate whether the expression of LRRK2 alters the ability of MPTP to induce PD-like phenotypes in vivo, a sub-toxic (low) dose of MPTP, which does not induce abnormality in nTg mice, was used. A cohort of 4-month-old male mice expressing WT-LRRK2 or G2019S-LRRK2, and nTg mice were injected with MPTP at 2.5 mg/kg sub-cutaneous, 2 injections at 24 h interval. Control mice of each genotype were injected with a similar volume of saline according to the same protocol. Five-day post-injection, the mice were tested on the accelerating rotarod over 3 days (Fig. 4). There was no significant change on rotarod performances of vehicle-treated mice among all experimental groups at the testing age. The nTg mice with a sub-toxic dose of MPTP showed no change on rotarod performance compared with the vehicle-treated control mice. In contrast, MPTP exposure of wild-type LRRK2 and G2019S-LRRK2 mice significantly decreased the rotarod performance compared with nTg mice with MPTP exposure. Mutant G2019S-LRRK2 mice displayed about 80% reduction in rotarod performance (Fig. 4B, Two-way ANOVA, interaction F(2, 42) = 3,203; treatment: F(1, 42) = 10,26; genotype: F(2, 42) = 5,812). These results demonstrated that the expression of LRRK2 proteins significantly enhanced the toxic effects of MPTP on mouse movement behavior.
Sub-toxic dose of MPTP exposure induced motor impairment in LRRK2 transgenic mice. (A) Experimental flow chart. Mice at age of 4 months (nTg, LRRK2 and G2019S-LRRK2 mice, n = 10 per group) were injected with MPTP (2.5 mg/kg, sc, 2 doses 24 h apart). Rotarod testing was performed from 5 to 7 days post-injection (3 trials per day, 3 days in a row). (B) Shown are the average latency to fall ± sem. Two-way ANOVA *P < 0.05 versus nTg, ***P < 0.001 versus nTg and ###P < 0.001 versus G2019S.
To assess whether theabove-mentioned movement impairment is due to the loss of dopaminergic neurons, substantia nigra sections from above mice at 8 days post-MPTP exposure were probed with an anti-tyrosine hydroxylase (TH, marker for dopaminergic neurons) antibody (Fig. 5A). The TH-positive neurons were quantified through substantia nigra sections. In striatal sections, dopaminergic fibers were visualized in a diffuse staining (Fig. 5B). Both WT-LRRK2 and G2019S-LRRK2 transgenic mice with MPTP exposure displayed significant loss of TH-positive neurons compared with nTg mice with MPTP exposure (Fig. 5C, Two-way ANOVA: interaction F(2, 42) = 17,27; treatment: F(1, 42) = 55,55; genotype: F(2, 42) = 18,38). Mutant G2019S mice had more severe dopaminergic neuron loss than control mice with over 60% in the substantia nigra (Fig. 5C) and loss of up to 80% of the dopaminergic fibers in the striatum (Fig. 5D, Two-way ANOVA: interaction F(2, 42) = 24,99; treatment: F(1, 42) = 97,04; genotype: F(2, 42) = 26,54). We further confirmed these findings by counting DAPI staining in TH-positive neurons in substantia nigra (Fig. 5E, two-way ANOVA: interaction F(2, 42) = 8,850, treatment: F(1, 42) = 29,19, genotype: F(2, 42) = 9,639). There was no significant TH-positive neuron loss in nTg mice with MPTP exposure. There was no change of dopaminergic neuron loss among the various genotype mice with vehicle treatment. In addition, there was no other neuron loss in substantia nigra, and there was no cell loss in other regions of the mouse brain (cortex, hippocampus, striatum and cerebellar) among all experimental groups.
Sub-toxic dose of MPTP exposure induced the loss of TH-positive immunoreactivity in LRRK2 transgenic mouse brains. (A) Representative images of TH neurons in the substantia nigra after MPTP exposure. Left and middle: TH green 20X; right: TH Red 40X. (B) Quantification of TH-positive cells in the substantia nigra. Results are normalized to control and expressed as mean ± sem. (C) and (D) Quantification of TH-positive immunoreactivity in the striatum. (E) DAPI and anti-TH staining of substantia nigra. The number of cells was quantified by counting DAPI. Results are normalized to control and expressed as mean ± sem. Two-way ANOVA analysis, *P < 0.05 versus nTg, **P < 0.01 versus nTg and ***P < 0.001 versus nTg; #P < 0.05 versus G2019S, ##P < 0.01 versus G2019S and ###P < 0.001 versus G2019S.
Astrocyte activation is another hallmark of neurodegeneration (38). Substantia nigra sections from above mice at 8 day post-MPTP exposure were subjected to immunostaining using anti-glial fibrillar protein (GFAP, a marker for astrocyte activation) antibodies. There was no change in GFAP immune reactivity in any mice of the three genotypes that received vehicle treatment (Fig. 6). There was no significant change in nTg mice with MPTP exposure compared with untreated groups. However, the exposure of mice expressing WT-LRRK2 or G2019S-LRRK2 to a sub-toxic dose of MPTP induced a significant increase in astrocyte activation compared with nTg mice (Fig. 6B, two-way ANOVA: interaction F(2, 42) = 66,44; treatment: F(1, 42) = 179,7; genotype: F(2, 42) = 69,27). PD mutant G2019S-LRRK2 induced the greatest astrogliosis, at up to 60 times the control levels.
Sub-toxic dose of MPTP exposure increased astrocytic activation in the brains of LRRK2 transgenic mice. (A) Representative images of GFAP (Red) staining in brains of transgenic mice. (B) Quantification of (A). Results are normalized to control and expressed as mean ± sem. Two-way ANOVA analysis, ***P < 0.001 versus nTg; ###P < 0.001 versus G2019S.
L-DOPA and a LRRK2 kinase inhibitor suppress the PD-like phenotypes in transgenic mice exposed to a sub-toxic dose of MPTP
Treatment with a LRRK2 kinase inhibitor, LRRK2-in-1, significantly improved motor impairment, reduced TH-positive neuron loss and reduced astrogliosis in G2019S-LRRK2 mice with MPTP exposure compared with untreated mice (Fig. 7). Although LRRK2-in-1 generally has a low blood brain barrier (BBB) penetration rate (39), MPTP has been reported to induce the leaking of the BBB that could result in LRRK2-in-1 entering brain tissue (40). Treatment with L-DOPA significantly improved motor-impairment in the rotarod performance testing G2019S-LRRK2 mice with MPTP exposure (Fig. 7B, two-way ANOVA: interaction F(6, 60) = 5,359; treatment: F(3, 60) = 11,69; genotype: F(2, 60) = 27,19). Unlike LRRK2-in-1 treatment, L-DOPA did not rescue TH-positive neuron loss (Fig. 7C, two-way ANOVA: interaction F(6, 60) = 5,149; treatment: F(3, 60) = 13,04; genotype: F(2, 60) = 30,85) and astroglia activation (Fig. 7D, two-way ANOVA: interaction F(6, 60) = 13,28; treatment: F(3, 60) = 34,31; genotype: F(2, 60) = 86,27) in G2019S-LRRK2 mice with MPTP exposure compared with untreated mice. These results demonstrated that L-DOPA and LRRK2-in-1 suppress PD-like phenotypes with distinct mechanisms.
Treatment of LRRK-IN-1 or L-DOPA attenuated MPTP and G2019S-LRRK2-induced PD-like phenotypes in mice. (A) Experimental flow chart. A cohort of mice from each group (n = 10) at age of 4 months were injected with MPTP (2.5 mg/kg, sc, 2 doses 24 h apart), and treated with a LRRK2 Kinase inhibitor (LRRK2-IN-1) or L-DOPA. (B) Rotarod testing was performed from 5 to 7 days post-injection (3 trials per day, 3 days in a row). (C) Quantification of TH-positive cells in the substantia nigra. (D) Quantification of GFAP. Two-way ANOVA analysis, *P < 0.05 versus nTg, **P < 0.01 versus nTg and ***P < 0.001 versus nTg; #P < 0.05 versus G2019S, ##P < 0.01 versus G2019S and ###P < 0.001 versus G2019S.
Discussion
With the emergence of large-scale genetic studies of the whole genome, there is increasing evidence of the interaction between genes and environmental factors or exogenous in the physiopathology of PD. In particular, the significant role of LRRK2 in both familial and sporadic form of the disease has been demonstrated (41,42). Overall, the phenotype observed in the LRRK2 transgenic mouse model is progressive and mild with little to no dopaminergic neuronal loss even in late life stages (24,43). There is, however, a strong body of evidence for LRRK2 toxicity using in vitro models. In this study, we generated the homozygous wild-type-LRRK2 and G2019S-LRRK2 transgenic mice under mouse-PrP promotor which lead to the ectopic expression of LRRK2 predominantly in neurons. Our transgenic mice overexpressing mutant G2019S-LRRK2 showed a mild motor phenotype after 9 months of age. Importantly, with two injections of a sub-toxic dose of MPTP, our G2019S-LRRK2 transgenic mice at 4 months of age displayed severe motor impairment, loss of dopaminergic neurons and astrogliosis in the brain. This LRRK2 gene-MPTP interaction model showed a short-time course and good response to L-DOPA and LRRK2 kinase inhibitor (LRRK2-in-1) treatments. Thus, this model provides a powerful tool for further drug screen and pathogenesis studies.
LRRK2 has been associated with toxicity through many processes such as mitochondrial failure (44), excitotoxicity (45), ER stress (46), autophagy (47) and more globally ROS production. When cultured in vitro, primary neurons overexpressing G2019S-LRRK2 showed an increased sensitivity to every stress tested, whereas WT-LRRK2 neurons showed a higher sensitivity only to MPP+ and serum withdrawal. These two conditions share an increased production of ROS, and impaired mitochondrial function suggesting that the mitochondria may be the most critical cellular compartment in LRRK2 toxicity genotype specific vulnerability of neurons cannot be explained solely by an increase in LRRK2 activity levels because when treated with DTT, which induces ER stress and an increase of LRRK2 levels (46); there is no extra sensitivity of the WT-LRRK2 overexpressing neurons compared with the nTg controls. Comparing the different genotypes during MPTP treatments both in vivo and in vitro may help to distinguish the molecular mechanisms that are specific to the mutation of LRRK2 from the ones common to all forms of PD.
In addition to the dopaminergic degeneration, it was important to show that the motor phenotype observed in the MPTP-treated animals could be rescued by L-DOPA. Both WT-LRRK2 and G2019S-LRRK2 mice responded to L-DOPA injections with a strong improvement in the rotarod performance. However, L-DOPA could not protect the dopaminergic neuron loss or reduce the astrogliosis. In contrast, a LRRK2 inhibitor, LRRK2-in-1 has been shown to be protective in vitro (11) as well as in some models of PD mice (48,49). Here, we showed that LRRK2-IN-1 can improve the motor phenotype of treated mice, and protect against dopaminergic neuron loss. The actual inhibition of the kinase activity remains debated, and the possibility remains that the observed protection might be due to a reduction of LRRK2 levels by proteosomal degradation (50). A recent study suggests that the G2019S mutant form of LRRK2 imparts resiliency to LRRK2 inhibitors (51). Additionally, long-term treatments with LRRK2 inhibitors have peripheral effects, most notably in the lungs and kidneys of rats (52) and non-human primates (53), which resemble what is observed in LRRK2 KO animals. Taken together, these results suggest a potential downside to kinase inhibition and may limit the use of the inhibitors in human therapeutics. Our new model that combines genetic modification and sub-toxic dose of MPTP may help to sort out the discrepancies and elucidate the role of LRRK2 kinase activity in sporadic as well as familial PD.
In conclusion, we find that the overexpression of LRRK2 in mice increases the vulnerability of dopaminergic neurons to degeneration when exposed to sub-toxic doses of MPTP. This novel model of the LRRK2 gene/MPTP interaction with a relatively short-time period provides a useful tool for better understanding the molecular and cellular mechanisms in both sporadic and familial PD, and has potential for testing novel therapeutics for PD intervention.
Material and Methods
Generation of LRRK2 transgenic mice
Human wild-type-LRRK2 and G2019S-LRRK2 gene were cloned into a mouse PrP promoter-derived plasmid and injected into a mouse gem line to generate the transgenic mouse line as previously described (34). The heterozygous G2019S-LRRK2 mouse was identified by genotype PCR (LRRK2 primer: ATGATGACTGCACTGGGTTCCCTGAAGAATG; PrP sense primer: GGGACTATGTGGACTGATGTCGG; PrP antisense primer: CCAAGCCTAGACCACGAGAATGC). The mice were maintained on a B6C3F1 background. G2019S-LRRK2 homozygous mice were generated by crossing male and female heterozygous G2019S-LRRK2 mice and identified by genotype PCR. Mice were housed in the Center for Comparative Medicine at Johns Hopkins University School of Medicine. Handling procedures were in accordance with NIH Guide for the Care and Use of Laboratory Animals and approved by the Institutional Animal Care and Use Committees of the Johns Hopkins Medicine Institute. Mice were maintained in a pathogen-free facility and exposed to a 12 h light/dark cycle with food and water provided ad libitum.
Western blot.
Mice were decapitated and the whole brain was dissected on ice. Hemi-brain was quickly immersed in a 2 ml lysis buffer (10% Chaps in pH 7.4 PBS with protease inhibitor cocktail and phosphatase inhibitor) and homogenized by Dounce Tissue Homogenizer (Eberbach). Brain lysate was rotated in 4°C for 1 h and then was centrifuged at 14 000 rpm for 15 min. Supernatant was harvested and protein concentration was determined by the BCA method (Pierce). An aliquot of 20 μg protein per sample was loaded onto gels. Mouse anti-Flag (Sigma F1804, 1:2000) and rabbit anti-LRRK2 antibody (MJFF-3, 1:2000) were used as primary antibodies.
Behavior analysis
Open field test.
Male and female transgenic mice and their respective littermate controls (6 and 12 months) were tested for general locomotor activity. Individual mice were put into the 16 × 16 in. cage of a PAS Open field monitor system (San Diego Instruments, San Diego, CA) in a bright room. The mouse’s horizontal and vertical movements were monitored and recorded for 60 min (sum data per 5 min) by a grid of 32 infrared beams at ground level and 16 elevated (3 in.) beams. These horizontal and vertical values were averaged and chartered as total activity and rearing activity, respectively.
Rotarod testing.
Transgenic mice and their respective littermate controls (4, 9 or 12 months) were tested for rotarod performance. The mice were trained before the first session of the test. The training session was a single trial in which the speed was set to 4 rpm and the mice were continuously placed back on for 5 min. For the testing sessions, mice walked on a computer-driven rotating rod (Columbus Instruments, Columbus, OH) at speeds increasing from 4 to 40 rpm (0.1 RPM increase per second) for 3 trials at 1 h intervals. The measurements were repeated for 3 consecutive days. The latency to fall from the rod was recorded, 9 attempts per mouse were kept and the arithmetic mean of these 9 values set aside for further analysis. The same rotarod protocol was used for MPTP treated mice.
Primary neuron culture and treatment
Whole brains were dissected from embryonic day 18 (E18) pregnant mice, and cortex was isolated, dissociated by trypsin. The cells were grown on coverslips pre-coated with poly-D-lysine (20 ng/ml; Life Technology, Camarillo, CA, USA) in media consisting of Neurobasal (Invitrogen, Waltham, MA, USA), B27 supplement (2% w/v), L-glutamine (500 mM) and penicillin/streptomycin (100 U/ml). At DIV 7, the cells were treated with various stressors for 24 h. Cell death was measured by nuclear condensation assay, as described previously (54). At 5 DIV, the neurons were treated with increasing doses of stressors agents: 3MA for autophagic stress, NMDA for excitotoxicity, KCl for depolarizing stress, DTT for ER stress and MPP+ for a mitochondrial stress. All the chemical were purchased from MilliporeSigma (Burlington, MA, USA). In addition, for a general stress effect, some neurons were deprived of trophic factor by replacing the medium with neurobasal without B27. After 48 h, cells were fixed for 30 min in 1× PBS solution containing 4% PFA and the nuclei were stained using Hoechst. Image acquisition was performed automatically on an Axiovert 200 (Zeiss, Thornwood, NY, USA) and the intensity of every nucleus was measured using Volocity (PerkinElmer, Boston, MA, USA). Results are presented in percentage of cell death.
Immunohistochemistry and neuron quantification
LRRK2 Transgenic and normal control mice were anesthetized, then followed intracardial perfusion with 50 ml of cold PBS and 50 ml of 4% paraformaldehyde (PFA). Brains were post-fixed in PFA overnight and then cryopreserved in 30% sucrose at 4°C for 48 h. Midbrain sections containing substantia nigra pars compacta were frozen on dry ice and sliced on a sliding microtome (Leica SM2000, Leica Biosystems Inc. Buffalo Grove, IL, USA), collected consecutively and preserved in the PBS buffer with NaN3 at 4°C. Tyrosine Hydroxylase was detected using anti-TH monoclonal antibody (Cell Signaling, Danvers, MA, USA) and GFAP using anti-GFAP polyclonal antibody (Abcam 7260, 1:500). Pictures were taken on an Axiovert 200 inverted microscope using a 63X objective. For cell counting, five consecutive sections containing the substancia nigra were selected from 5 to 10 animals per group.
MPTP exposure and drug treatment experiments
Male mice at 2 to 4 months of age were injected with MPTP at 2.5 mg/kg sub-cutaneous, in two separate injections 24 h apart in a hood dedicated to MPTP use only. Control mice were injected with a similar volume of saline according to the same protocol. This MPTP protocol with only two injections via subcutaneous route provided further rescuing compound with IP injections. Moreover, pilot study results that a normal dose of 10 mg/kg described previously by other group lead to the death of all LRRK2 transgenic mice (WT and G2019S) within a few hours. The 2.5 mg/kg dose was selected because no abnormality was observed in nTg mice, which provided a basic tolerant dose for LRRK2 gene and MPTP interaction studies. To avoid any human contact with MPTP, mice were kept in the hood and monitored while recovering for 72 h before being put back in their housing room. Five days after injections, the rotarod testing was performed over 3 days. The mice were sacrificed at 8 days post-injection for further analysis.
For the chemical rescue experiments, the same MPTP treatment protocol was used with additional injections of compounds. LRRK2-IN-1 (50 mg/kg) was injected intraperitoneally once a day, every day from the day before the first MPTP injection until the sacrifice day. L-DOPA (20 mg/kg) treatments were done once a day intraperitoneally, once a day starting from day 4 post-injection.
Statistical analysis
All the experiments were performed with the operators blinded to genotypes and treatments. Open field data were analyzed by two-tailed, unpaired Student’s t-test by comparison of nTg and transgenic mice for each condition or data interval. P < 0.05 was considered significant. Other assays were analyzed using one- or two-way ANOVA.
Conflict of Interest statement.
Wild-type-LRRK2 and G2019S-LRRK2 transgenic mice are the tangible property of Johns Hopkins University.
Funding
National Institutes of Health (R01 NS055252 to C.A.R., R01NS093383 to W.W.S.).
References
Author notes
Present address: Department of Neurology, The First Affiliated Hospital, Sun Yat-sen University, Guangzhou, China.
