Selective expression of Parkinson's disease-related Leucine-rich repeat kinase 2 G2019S missense mutation in midbrain dopaminergic neurons impairs dopamine release and dopaminergic gene expression

Abstract

Preferential dysfunction/degeneration of midbrain substantia nigra pars compacta (SNpc) dopaminergic (DA) neurons contributes to the main movement symptoms manifested in Parkinson's disease (PD). Although the Leucine-rich repeat kinase 2 (LRRK2) G2019S missense mutation (LRRK2 G2019S) is the most common causative genetic factor linked to PD, the effects of LRRK2 G2019S on the function and survival of SNpc DA neurons are poorly understood. Using a binary gene expression system, we generated transgenic mice expressing either wild-type human LRRK2 (WT mice) or the LRRK2 G2019S mutation (G2019S mice) selectively in the midbrain DA neurons. Here we show that overexpression of LRRK2 G2019S did not induce overt motor abnormalities or substantial SNpc DA neuron loss. However, the LRRK2 G2019S mutation impaired dopamine homeostasis and release in aged mice. This reduction in dopamine content/release coincided with the degeneration of DA axon terminals and decreased expression of DA neuron-enriched genes tyrosine hydroxylase (TH), vesicular monoamine transporter 2, dopamine transporter and aldehyde dehydrogenase 1. These factors are responsible for dopamine synthesis, transport and degradation, and their expression is regulated by transcription factor paired-like homeodomain 3 (PITX3). Levels of Pitx3 mRNA and protein were similarly decreased in the SNpc DA neurons of aged G2019S mice. Together, these findings suggest that PITX3-dependent transcription regulation could be one of the many potential mechanisms by which LRRK2 G2019S acts in SNpc DA neurons, resulting in downregulation of its downstream target genes critical for dopamine homeostasis and release.

Introduction

Parkinson's disease (PD) is the most common degenerative movement disorder and presents with four cardinal features: resting tremor, rigidity, slowness and paucity of voluntary movement and postural instability (1,2). These motor symptoms are attributed to a substantial loss of substantia nigra pars compacta (SNpc) dopaminergic (DA) neurons and associated deficiency in dopamine transmission (2,3). Although dopamine replacement therapy via the supplement of dopamine precursor, l-3,4-dihydroxyphenylalanine (l-DOPA), is the standard for PD therapy (4), l-DOPA treatment does not prevent the degeneration of SNpc DA neurons and becomes less effective over time (5). Therefore, there is a need to better understand the underlying mechanisms involved in SNpc DA neuronal death in PD in order to improve current therapies and to prevent the worsening of its associated motor symptoms.

Clinically, there are several genetic mutations that have been implicated in PD etiology (6). Multiple missense mutations in Leucine-rich repeat kinase 2 (LRRK2) are the most common genetic abnormality, particularly the LRRK2 G2019S mutation, which has been linked to a late-onset autosomal dominant familial form of PD (7). LRRK2 encodes a large cytosolic protein that contains both GTPase and protein kinase domains, and the LRRK2 G2019S mutation resides in the kinase domain and may potentiate the kinase activity (8). LRRK2 regulates several cellular activities such as protein synthesis (9–11), endoplasmic reticulum (ER)–Golgi export (12), cytoskeleton dynamics (13), endocytosis (14) and autophagy (15).

In neurons, there is increasing evidence suggesting that LRRK2 regulates synaptic transmission through both pre- and post-synaptic mechanisms (16–18). Additionally, mouse models that harbor LRRK2 G2019S or R1441G/C mutations display abnormalities in dopamine transmission (19–21). Although LRRK2 is expressed abundantly in striatal neurons that receive dopamine input from SNpc DA neurons (22,23) and modulates both synaptogenesis and dopamine receptor activation (17), genetic deletion of Lrrk2 has no deleterious effects on rodent SNpc DA neurons (24–26). Furthermore, several lines of LRRK2 transgenic and knockin mice have been generated; however, none develops substantial loss of SNpc DA neurons (27,28). Such studies raise the question of whether PD-related LRRK2 mutations, namely, LRRK2 G2019S, affect the cell autonomous function and/or survival of SNpc DA neurons.

Given the strong clinical association of LRRK2 G2019S and PD, we hypothesized that the minimal impact of G2019S mutation on the survival of SNpc DA neurons in vivo may be due to a lack of robust transgene overexpression. Therefore, in this study, we overexpressed wild-type human LRRK2 (WT) as well as human LRRK2 G2019S (G2019S) selectively in mouse midbrain DA neurons using a tetracycline-dependent binary gene expression system (25). In the LRRK2 G2019S model, we detected a more than 6-fold increase in the LRRK2 protein expression in the SNpc DA neurons compared with the non-transgenic (nTg) littermate controls. However, after 18 months, we did not observe overt motor/behavioral abnormalities or substantial SNpc DA cell death.

Interestingly, the G2019 mice were characterized by a significant reduction in dopamine content and release. Therefore, to understand the molecular mechanisms underlying this phenotype, we assessed the expression of proteins responsible for dopamine synthesis, transport and degradation. We found that dopamine regulatory proteins tyrosine hydroxylase (TH), vesicular monoamine transporter 2 (VMAT2), dopamine transporter (DAT) and aldehyde dehydrogenase 1 (ALDH1A1) were significantly downregulated in G2019S mice. Similarly, the LRRK2 G2019S transgene also reduced the expression of transcription factor paired-like homeodomain 3 (PITX3), a key regulator of dopamine gene expression in SNpc DA neurons (29). In light of these observations, our findings suggest that the LRRK2 G2019S mutation may suppress DA transmission in part through downregulation of key DA genes.

Results

Murine models of selective human LRRK2 WT and G2019S overexpression in midbrain DA neurons

We have previously inserted tetracycline transactivator (tTA) cDNA into the 3′-untranslated region of Pitx3, a gene exclusively expressed by midbrain DA neurons in the brain (30), to generate Pitx3-IRES2-tTA knockin mice. In these mice, the transcription of tTA is selectively expressed by midbrain DA neurons via the Pitx3 promoter (31). In this so-called ‘tet-off’ system, tTA can turn on the expression of any transgene under the control of tetracycline operator (tetO) (31).

We have also generated transgenic mice with either LRRK2 WT or LRRK2 G2019S under the control of tetO: tetO-LRRK2 WT and tetO-LRRK2 G2019S, respectively (25). Thus, to selectively express human LRRK2 in mouse midbrain DA neurons, we crossed Pitx3-IRES2-tTA heterozygous knockin mice with either tetO-LRRK2 WT or tetO-LRRK2 G2019S mice to generate Pitx3-IRES2-tTA/tetO-LRRK2 WT (WT mice) and Pitx3-IRES2-tTA/tetO-LRRK2 G2019S (G2019S mice) bigenic mice (Fig. 1A).

Figure 1.

Transgenic expression of human LRRK2 selectively in mouse midbrain DA neurons. (A) Schematic diagram depicts the generation of LRRK2 bigenic mice using a tetracycline-regulated gene expression system. Sample images show staining of HA-tagged LRRK2 (green) and TH (red) in sagittal section of 1-month-old tetO-G2019S single transgenic and G2019S bigenic mice. Scale bars: 400 µm (top two panels) and 1 mm (bottom panel). (B) Subcellular distribution of LRRK2 in SNpc DA neurons. Immunofluorescent images show the distribution of transgenic human LRRK2 at soma, dendrites (den.), axons and axon terminal striatum of SNpc DA neurons in 1-month-old G2019S mice. Human LRRK2 (green) was revealed by HA antibody staining; DA neurons were visualized by TH immunostaining (red). Scale bar: 20 µm (top panel) and 10 µm (lower panels). (C) Western blots show the levels of transgenic and endogenous LRRK2 expression in the striatum (ST) and midbrain (MB) homogenates of 1-month-old nTg, WT and G2019S mice. The cerebral cortex (CX) of Lrrk2 homozygous knockout (KO) mice was used as a negative control for the specificity of LRRK2 antibody. TH, dopamine- and cAMP-regulated neuronal phosphoprotein (DARPP32) and β-actin were used as the loading controls. Bar graph estimates the level of LRRK2 overexpression (normalized against β-actin expression) in the striatum and midbrain of nTg and bigenic mice. (D) Representative images show mouse and human LRRK2 expressions in SNpc DA neurons of 1-month-old G2019S and nTg mice. Scale bar: 20 µm. Scatter plot estimates the LRRK2 protein expression levels. Data were presented as mean ± SEM. (E) Western blots show the expression of transgenic and endogenous LRRK2 proteins in the midbrain of post-natal day 1 (P1), 1-month (1M) and 3-month (3M)-old nTg and G2019S bigenic mice treated with (+) or without (−) doxycycline (DOX). TH and β-actin were used as the loading controls.

In order to facilitate the identification of human LRRK2 proteins, their respective C-terminals were tagged with hemagglutinin (HA) epitopes (25). Staining for HA revealed that expression of the human LRRK2 G2019S protein was restricted to the TH-positive midbrain DA neurons of 1-month-old G2019S mice (Fig. 1A and Supplementary Material, Fig. S1). HA staining was mainly accumulated in the soma and proximal dendrites, but sparsely distributed along axon bundles (Fig. 1B). No obvious staining was detected in the DA axon terminals at the striatum (Fig. 1B). Western blot analyses further confirmed that HA-tagged LRRK2 proteins were predominantly accumulated in the midbrain tissues (Fig. 1C). The same regional expression pattern and subcellular localization of LRRK2 WT proteins were observed in WT transgenic mice, and no expression of HA-tagged LRRK2 was detected in both tetO-WT and tetO-G2019S mice (data not shown).

Using an antibody recognizing both human and mouse LRRK2 proteins, we found an approximate 4-fold increase in LRRK2 protein expression in the midbrain homogenate of 1-month-old LRRK2 WT and G2019S transgenic mice, compared with age-matched controls (Fig. 1C). We further quantified the fluorescent intensity of LRRK2-positive signals in SNpc DA neurons and found a more than 6-fold increase in LRRK2 protein expression in G2019S mice when compared with controls (Fig. 1D). Although the initial expression of Pitx3 occurs in early embryonic stages (29), no apparent accumulation of HA-tagged LRRK2 proteins was detected in post-natal day 1 (P1) G2019S mice (Fig. 1E). As expected in a tet-off system, administration of doxycycline (DOX), a member of tetracycline class antibiotics, suppressed the expression of transgenic LRRK2 proteins (Fig. 1E).

G2019S mice have normal motor function and do not present with midbrain DA neurodegeneration

G2019S mice developed normally and survived for the expected life span (data not shown) and had normal body weight compared with other genotypes (Fig. 2A). Additionally, G2019S mice showed comparable horizontal, vertical and fine movements to control mice in the Open-field test (Fig. 2B–D). G2019S mice also displayed normal balance and motor coordination in rotarod test (Fig. 2E) and exhibited regular gait patterns (Fig. 2F and G). In contrast, WT transgenic mice displayed moderate increase of horizontal and vertical movements starting at 2 months of age (Fig. 2H and I).

Figure 2.

G2019S mice displayed no apparent motor abnormalities. (A–G) At 2, 6, 12 and 18 months of age, a cohort of male nTg, Pitx3-IRES2-tTA (tTA), tetO-G2019S and G2019S mice (n ≥ 8 animals in each genotype) were repeatedly weighted (A); tested for horizontal movement (B), vertical movement (C) and fine movement (D) in Open-field; examined by rotarod test (E) and measured for stride time (F) and length (G). Data were presented as mean ± SEM. Two-way ANOVA was used for statistical analysis. (H and I) The horizontal (H) and vertical movements (I) of male WT (n = 6) and control nTg (n = 10) mice were measured using the Open-field test at 2, 6, 12 and 18 months of age. Data were presented as mean ± SEM. One-way (F and G) and two-way ANOVAs were used for statistical analysis.

We next stained DA neurons for TH in series of midbrain coronal sections obtained from 20-month-old nTg, WT and G2019S transgenic mice (Fig. 3A). Using an unbiased stereological approach (31), we counted the numbers of TH-positive DA neurons in the SNpc and ventral tegmental area (VTA). Neither WT nor G2019S transgenic mice presented with significant loss of DA neurons in either the SNpc or VTA, compared with controls (Fig. 3B–E). Furthermore, there was no apparent astrocytosis or microgliosis in the midbrain of 20-month-old G2019S mice (Fig. 3F–I). Taken together, these data demonstrate that overexpression of the PD-related LRRK2 G2019S mutation in midbrain DA neurons did not induce substantial motor abnormalities or neuron loss in aged mice.

Figure 3.

G2019S mice showed no substantial midbrain DA neuron loss. (A) TH staining (brown) of midbrain coronal sections of 20-month-old nTg, WT and G2019S mice. Scale bar: 200 µm. (B and C) Numbers of TH-positive neurons remain in the SNpc (B) and VTA (C) of nTg and G2019S mice at 20 months of age (n = 3–9 animals per genotype per region). Data were presented as mean ± SEM. Unpaired t-test was used for statistical analysis. (D and E) Numbers of TH-positive neurons remain in the SNpc (D) and VTA (E) of nTg and WT mice at 20 months of age (n = 3 animals per genotype per region). Data were presented as mean ± SEM. Unpaired t-test was used for statistical analysis. (F) Sample images show GFAP and Iba1 staining in the midbrain of 20-month-old nTg and G2019S mice. Scale bar: 200 µm (left two columns) and 20 µm (right two columns). (G and H) Scatter plots depict the density of GFAP-positive astrocytes (G) and Iba1-positive microglia cells (H) in the SNpc of nTg and G2019S mice (n = 4 animals per genotype and four sections per animals). Data were presented as mean ± SEM. Unpaired t-test was used for statistical analysis. (I) Scatter plot shows cell body size of microglia in (H). Data were presented as mean ± SEM. Unpaired t-test was used for statistical analysis.

G2019S mice display age-dependent loss of DA axonal terminals

We next examined the integrity of midbrain DA axon bundles and terminals in nTg, WT and G2019S mice. The majority of midbrain DA axon fibers form transverse bundles across substantia nigra pars reticulata (SNpr) and globus pallidus and terminate at the striatum (32). The 18-month-old WT and G2019S had no apparent alterations of TH-staining pattern in DA axon bundles (Supplementary Material, Fig. S2). However, the density of TH-positive DA axon terminals at the striatum was decreased in G2019S mice compared with age-matched nTg and WT mice (Fig. 4A and B), although there was no reduction in DA axon terminals observed in 1-month-old G2019S mice (Fig. 4B). TH staining also revealed abnormally enlarged varicosities in the striatum of 18-month-old G2019S mice (Fig. 4C). These TH-positive varicosities were also co-stained with dopamine transporter VMAT2 (Fig. 4C), suggesting that overexpression of LRRK2 G2019S in midbrain DA neurons may lead to axon terminal degeneration in aged, not young, mice.

Figure 4.

G2019S mice exhibited degeneration of DA axon terminals. (A) Sample images show TH (green) staining in the striatum of 18-month-old nTg, WT and G2019S mice. Scale bar: 6 µm. (B) The bar graph depicts the fraction of striatum occupied by TH-positive DA axon terminals of 1- and 18-month-old nTg, WT and G2019S mice. N ≥ 3 animals per genotype per age group and n ≥ 5 sections per animal. Data were presented as mean ± SEM. One-way ANOVA plus Tukey's post hoc test. *P < 0.05 and **P < 0.01. (C) Sample images show VMAT2 (green) and TH (red) co-staining in the striatum of 18-month-old nTg and G2019S mice. Topro3 was used to stain the nucleus (blue). Arrowhead marks a large varicosity. Scale bar: 10 µm.

G2019S mice exhibit age-dependent decreases in dopamine release

In light of these findings, we assessed the dopamine content in the dorsal striatum of 1- and 12-month-old nTg, WT and G2019S mice. Although 12-month-old G2019S presented with a modest decrease in dopamine compared with the other cohorts, WT mice displayed elevated dopamine levels compared with nTg controls at both 1 and 12 months of age (Fig. 5A). In addition, levels of the main dopamine metabolite 3,4-dihydroxyphenylacetic acid (DOPAC) were modestly decreased in 1-month-old G2019S mice compared with age-matched nTg controls (Fig. 5B).

Figure 5.

G2019S mice showed decrease of dopamine content and release. (A) HPLC measures the steady levels of dopamine in the dorsal striatum of 1- and 12-month-old nTg, WT and G2019S mice. The DA content was high in young and old WT transgenic mice compared with controls. In aged mice, the G2019S mutant showed lower levels of DA in the dorsal striatum (nTg versus WT: t₍₁₂₎ = 2.317, *P < 0.05; nTg versus G2019S: t₍₁₁₎ = 11.07, *P < 0.05; WT versus G2019S: t₍₉₎ = 4.002, **P < 0.01). Data were presented as mean ± SEM. One-way ANOVA plus Tukey's post hoc test. (B) HPLC measures the steady levels of DOPAC in the dorsal striatum of 1- and 12-month-old nTg, WT and G2019S mice. The 1-month-old G2019S mice showed reduced DOPAC levels in the striatum compared with nTg controls. Data were presented as mean ± SEM. One-way ANOVA plus Tukey's post hoc test. *P < 0.05. (C) Voltammetry monitors the release of dopamine in the dorsal striatum of 1- and 12-month-old nTg, WT and G2019S mice. Young WT mice increased significantly the amount of DA release in response to higher stimulus intensities [ANOVA interaction: F_(4,30) = 4.35; Bonferroni's post hoc test P < 0.05* and P < 0.01**], but young G2019S mice were comparable to control mice. In aged mice, the DA release in G2019S mutants was lower compared with WT and control groups [ANOVA genotype factor: F_(1,33) = 15.13; Bonferroni's post hoc test P < 0.05*]. (D) Uptake indexes, expressed as τ values, were comparable between groups in both the age considered. One-way ANOVA was used for statistical analysis.

Next, dopamine release kinetics was studied. Fast-scan cyclic voltammetry was used to evaluate transient increases in extracellular dopamine concentrations, evoked by single electrical pulse stimulation in slices containing the dorsolateral striatum (31). We found a significant decrease in dopamine release in 12-month-old G2019S transgenic mice, compared with age-matched nTg and WT mice (Fig. 5C). However, there was no change in dopamine release in 1-month-old G2019S mice (Fig. 5C). Contrasting the phenotype of G2019S mice, WT transgenic mice showed increased dopamine release at 1 month of age, but no change at 12 months, compared with nTg mice (Fig. 5C), although the dynamics of dopamine uptake was comparable between different genotypes within each age group (Fig. 5D). As controls, we found no significant alteration of dopamine release in 12-month-old Pitx3-tTA and tetO-G2019S single transgenic mice (Supplementary Material, Fig. S3). Together, these data suggest that LRRK2 G2019S mutation specifically impairs both dopamine homeostasis and release in aged mice.

G2019S mice show age-dependent alterations of DA protein and gene expression

Both DA neuron soma and axon terminals contain high levels of TH, VMAT2, DAT and ALDH1A1 proteins critical for dopamine synthesis, transport and degradation, respectively (33,34). Therefore, to better understand how LRRK2 G2019S affects dopamine homeostasis and release, we examined the expression levels of these factors in the striatum of 1-, 6- and 18-month-old G2019S mice, as well as age-matched nTg and WT mice.

The expression of TH, DAT and ALDH1A1 proteins was significantly increased in 1-month-old G2019S mice, compared with nTg controls (Fig. 6A and B). However, the levels of TH, VMAT2, DAT and ALDH1A1 proteins were substantially decreased in 18-month-old G2019S mice, compared with nTg controls and WT mice (Fig. 6A and C). Therefore, the expression of TH, VMAT2, DAT and ALDH1A1 proteins appears to show biphasic responses to introduction of the LRRK2 G2019S transgene.

Figure 6.

Alternations of DA gene expression in G2019S mice. (A) Western blots show TH, VMAT2, DAT and ALDH1A1 expression in striatum homogenates from 1-, 6- and 18-month-old nTg, WT and G2019S (GS) mice. β-actin was used as the loading control. (B and C) Scatter plots quantify the relative expression levels of TH, VMAT2, DAT and ALDH1A1 in striatum homogenates of nTg, WT and G2019S (GS) mice at 1 (B) and 18 months of age (C). N ≥ 3 animals per genotype per age point. Data were presented as mean ± SEM. One-way ANOVA plus Tukey's post hoc test. *P < 0.05, **P < 0.01 and ***P < 0.001. (D) Scatter plots show expression of Th, Vmat2, Dat and Aldh1a1 mRNAs in the LCM-isolated SNpc DA neurons from 12-month-old control, WT and G2019S (GS) mice. N = 3 mice per genotype and n > 200 SNpc DA neurons per animal. β-actin was used to normalize the gene expression. Data were presented as mean ± SEM. One-way ANOVA plus Tukey's post hoc test. **P < 0.01 and ***P < 0.001.

To understand how LRRK2 G2019S regulates DA protein expression, we isolated individual SNpc DA neurons from 12-month-old nTg, WT and G2019S mice by laser-capture microdissection (LCM) and quantified the expression of Th, Vmat2, Aldh1a1 and Dat mRNAs. We found that the expression of Vmat2 and Aldh1a1 mRNAs was substantially decreased in the SNpc DA neurons of G2019S mice when compared with nTg controls (Fig. 6D). The expression of Th and Dat was also lower in G2019S mice, but this was not statistically significant (Fig. 6D). These data suggest that the LRRK2 G2019S mutant suppresses the transcription of key DA genes, leading to downregulation of DA protein expression during aging.

G2019S mice exhibit age-dependent reduction of PITX3 expression

As Th, Vmat2, Dat, Aldh1a1 and other genes essential for the function and maintenance of midbrain DA neurons are under the control of transcription factors PITX3 and nuclear receptor-related 1 protein (NURR1) (35), we next assessed the expression of PITX3, and later NURR1, in the SNpc DA neurons of 1- and 12-month-old nTg, Lrrk2 homozygous knockout (KO), WT and G2019S mice. PITX3 staining localized primarily in the nucleus of DA neurons, and there was no PITX3 staining in the DA neurons of Pitx3-deficient mice, validating the specificity of our antibody (Fig. 7A and Supplementary Material, Fig. S4). Subsequent image analyses revealed a significant reduction of PITX3 expression in the SNpc DA neurons of 12-month-old G2019S and KO mice, compared with age-matched nTg, Pitx3-tTA, tetO-WT, tetO-G2019S and WT mice (Fig. 7B). No apparent alterations of PITX3 protein expression were observed in the DA neurons of 12-month-old Pitx3-tTA, tetO-WT and tetO-GS single transgenic mice (Supplementary Material, Fig. S5). This decrease is age-dependent, as there was no significant alteration in PITX3 expression in 1-month-old G2019S mice (data not shown). We further examined the expression of Pitx3 mRNA in LCM-isolated SNpc DA neurons and found a similar reduction of Pitx3 mRNA in the DA neurons of 12-month-old G2019S mice (Fig. 7C). Additionally, we found that the expression of PITX3 was also downregulated in DA neurons overexpressing the PD-related α-synuclein A53T mutation (Fig. 7D). Given that the Pitx3 deficiency leads to SNpc DA neuron loss (30), these results suggest that PITX3 is an important pathogenic target for PD-related mutant LRRK2 and α-synuclein.

Figure 7.

Overexpression of LRRK2 G2019S mutation suppressed PITX3 and Nurr1 expression. (A) Sample images show co-staining of PITX3 (green), LRRK2-HA (red) and TH (blue) in SNpc DA neurons of 12-month-old nTg, KO, WT and G2019S mice. Scale bar: 10 µm. (B) Box and whisker plots depict the average intensity of PITX3 staining in the nucleus of TH-positive mDA neurons of 12-month-old nTg, KO, WT and G2019S mice. Data were presented as mean ± SEM. One-way ANOVA plus Tukey's post hoc test. ***P < 0.001. (C) Scatter plot shows the level of Pitx3 mRNA expression in the LCM-isolated SNpc DA neurons from 12-month-old control, WT and G2019S (GS) transgenic mice. N = 9 mice per genotype and n > 200 cells per animal. β-actin was used to normalize the gene expression. Data were presented as mean ± SEM. One-way ANOVA plus Tukey's post hoc test. *P < 0.05. (D) Box and whisker plots depict the average intensity of PITX3 staining in the nucleus of TH-positive mDA neurons of 1-month-old Pitx3-IRES2-tTA/tetO-SNCA A53T bigenic mice. N = 92 nuclei in the nTg group (n = 3 mice) and n = 205 nuclei in the A53T bigenic group (n = 5 mice). Data were presented as mean ± SEM. Unpaired t-test. **P < 0.01. (E) Box and whisker plots depict the average intensity of NURR1 staining in the nucleus of TH-positive mDA neurons of 1- and 12-month-old nTg (n = 212 and 391 nuclei in the 1- and 12-month age group), WT (n = 230 and 390) and G2019S (n = 206 and 287) mice. Data were presented as mean ± SEM. One-way ANOVA plus Tukey's post hoc test. **P < 0.01 and ***P < 0.001. (F) Scatter plot shows the level of Nurr1 mRNA expression in the LCM-isolated SNpc DA neurons from 12-month-old control, WT and G2019S (GS) transgenic mice. N = 9 mice per genotype and n > 200 cells per animal. β-actin was used to normalize the gene expression. Data were presented as mean ± SEM. One-way ANOVA plus Tukey's post hoc test. **P < 0.01.

As the expression of NURR1 is also downregulated in PD and α-synuclein transgenic mice (36), we also examined NURR1 expression in G2019S mice. Like PITX3, NURR1 staining was restricted to the nucleus of DA neurons (Supplementary Material, Fig. S6), and NURR1 protein expression was decreased in 1-month-old G2019S mice when compared with nTg controls. No significant change was found in 12-month-old G2019S mice (Fig. 7E). Contrastingly, NURR1 protein levels were upregulated in both 1- and 12-month-old WT mice (Fig. 7E). Furthermore, Nurr1 mRNA expression showed a decrease in 12-month-old G2019S mice, but no alteration was seen in WT mice (Fig. 7F). These data suggest that LRRK2 G2019S may suppress the expression of NURR1, whereas WT LRRK2 has the opposite effect.

Discussion

Since the initial identification of PD-linked LRRK2 mutations, numerous strains of LRRK2 mutant mice have been generated to study their effects on SNpc DA neurodegeneration (27,28). However, although one model of LRRK2 G2019S transgenic mice developed significant (∼17%) neuron loss (37), in the majority of these model systems, the extent of DA neuron loss is quite modest (21,37–39), and only one line of G2019S BAC mice shows mild reduced extracellular dopamine content (38). Although it is possible that the LRRK2 G2019S mutation is simply less pathogenic than other PD-associated mutations such as the α-synuclein A53T mutation [that leads to >40% SNpc DA neuron loss in 12-month-old mice (31)], there is strong clinical data linking the LRRK2 G2019S mutation to late-onset PD (40,41). However, LRRK2 G2019S mutations are not necessarily sufficient to cause PD as families harboring the G2019S mutations develop PD with only 67% penetrance (42).

In light of these observations, we hypothesized that the minimal impact of the G2019S mutation on the survival of DA neurons in vivo may be attributed to a lack of robust transgene expression. Therefore, we employed a tet-off binary gene expression system under the control of the midbrain DA neuron promoter Pitx3 and managed to achieve more than 6-fold overexpression of LRRK2 G2019S proteins selectively in the SNpc DA neurons. However, even with robust LRRK2 G2019S expression, these mice also failed to develop any substantial degeneration of SNpc DA neurons.

Alternatively, LRRK2 G2019S mutation may exert its pathogenicity not only in the DA neurons but also in other neurons and non-neuronal cells. The LRRK2 G2019S mutation-induced DA neuron loss may result from both cell autonomous and non-autonomous mechanisms. Compared with midbrain DA neurons, LRRK2 is more abundant in the striatal projection neurons (SPNs) and regulates the post-synaptic responses to dopamine transmission (17,18,43,44). In the striatum, LRRK2 is particularly enriched in the SPNs located in the striosome/patch compartment (45,46). Neuron tracing studies indicate that striosomal SPNs can form monosynaptic innervation onto the SNpc DA neurons, suggesting that LRRK2 may regulate DA neuron activity and dopamine release in a non-DA neuron-autonomous manner (47,48). The expression of LRRK2 is also substantially elevated in microglia upon inflammatory stimulation (49,50). Microglia are the immune cells in the CNS, and excessive activation of microglia is implicated in the pathogenesis of PD and other neurological disorders (51). The influence of mutant LRRK2 in SPNs and microglia on PD-related SNpc DA neuron dysfunction/loss, however, remains to be further investigated.

Yet, despite the lack of neurodegeneration, this new line of G2019S mice allowed us to systematically evaluate the impact of the gross LRRK2 G2019S overexpression on both the viability and the functionality of SNpc DA neurons. G2019S mice showed a modest decrease of dopamine release and the associated downregulation of TH, VMAT2, DAT and ALDH1A1 proteins critical for the synthesis, transport and degradation of dopamine in DA neurons (29). These observations are consistent with other animal models of LRRK2 mutations, many of which show similar alterations in dopamine transmission (20,21,38,52). Most notably, behavioral and physiological hyperactivity of the DA system has been reported in LRRK2 BAC transgenic mice overexpressing human WT LRRK2, but not G2019S mutation (21). We also observed differential effects of WT and G2019S LRRK2 on dopamine release in 1-month-old bigenic mice and found that overexpression of LRRK2 G2019S in DA neurons led to substantial decrease of the dopamine content and release at DA axon terminals in 12-month-old mice.

Early studies indicate that LRRK2 is involved in presynaptic transmission of glutamatergic synapses (16,53). In DA neurons, however, LRRK2 was primarily located in the soma and proximal dendrites and axons, suggesting that LRRK2 may not directly modulate dopamine release at the axon terminals. Previous studies have also shown that LRRK2 regulates cytoskeleton dynamics and ER–Golgi transport, which may also affect the trafficking of synaptic vesicles to the axon terminals (12,13). However, we did not observe gross morphological alterations in ER, Golgi or mitochondria in LRRK2 G2019S overexpressing SNpc DA neurons (Supplementary Material, Fig. S7).

Although there were no apparent morphological abnormalities in axon fibers (Supplementary Material, Fig. S2), LRRK2 G2019S overexpression was associated with reduced striatal DA axon terminal density and abnormally large varicosities. These structural abnormalities, in combination with the aforementioned repression of DA regulatory proteins, further substantiate the decrease in dopamine transmission observed in G2019S mice; however, the mechanism underlying these events is unclear. As TH, VMAT2, DAT and ALDH1A1 are under the control of NURR1 and PITX3 responsive, we hypothesized that LRRK2 G2019S mutation may lead to disruptions in NURR1 and PITX3 signaling.

NURR1 and PITX3 are transcription factors critical for both differentiation and maintenance of SNpc DA neurons (29,36). It has previously been shown that genetic ablation of Nurr1 or Pitx3 diminishes the expression of TH, VMAT2 and other DA genes and induces the progressive loss of SNpc DA neurons (30,54,55). Clinically, NURR1 proteins are substantially downregulated in the SNpc DA neurons of patients with PD and α-synuclein transgenic mice (36). In our study, we found that both NURR1 and PITX3 expression is reduced in the background of LRRK2 G2019S. Combined, these observations offer at least one possible explanation for the disrupted dopamine transmission observed in G2019S mice, and the relationship between LRRK2 G2019S and NURR1/PITX3 warrants further study. The levels of PITX3 proteins were increased in the TH-positive neurons in LRRK2 WT transgenic mice, but decreased in the G2019S mice, suggesting that the PD-related LRRK2 G2019S mutation likely compromises the normal function of LRRK2 in promoting the expression of PITX3. In support of this notion, the intensity of PITX3 staining was also reduced in the TH-positive neurons of Lrrk2 KO mice. We speculate that LRRK2 GTPase and kinase inhibitors may also suppress the expression of PITX3. Together, these data indicate that the G2019S mutation may not simply enhance the kinase activity of LRRK2; it perhaps also impairs the normal function of LRRK2 through a potentially dominant negative mechanism.

In summary, we presented a new line of LRRK2 G2019S transgenic mice grossly overexpressing the most common PD-associated mutation, selectively in midbrain DA neurons. Although G2019S mutant mice are not robust models of PD-related DA neuron loss, these mutant mice may serve as a useful model system to investigate the dopamine transmission deficiency occurring at the pre-clinical stages of the disease. Additionally, our initial findings suggest that PITX3, a master control gene in SNpc DA neurons, is among the downstream targets of mutant LRRK2, suggesting deficiency in PITX3 as a common pathogenic pathway in PD.

Materials and Methods

Animals

Pitx3^+/IRES2-tTA knockin mice (31) and tetO-HA tag LRRK2 WT and tetO-LRRK2 G2019S transgenic mice were created, as described previously (25). The Pitx3^+/IRES2-tTA mice were crossbred with tetO-LRRK2 WT or G2019S transgenic mice to obtain Pitx3^+/IRES2-tTA/tetO-LRRK2 WT (Abbr.: WT) or Pitx3^+/IRES2-tTA/tetO-LRRK2 G2019S (Abbr.: G2019S) bigenic mice. Pitx3^ak mice were purchased from The Jackson Laboratory (Bar Harbor, ME, USA) (strain B6xC57BLKS-ak/J). All mice were housed in a 12 h light/dark cycle and fed regular diet ad libitum. All mouse work follows the guidelines approved by the Institutional Animal Care and Use Committees of the National Institute of Child Health and Human Development, NIH.

Genotyping

Genomic DNA was prepared from tail biopsy using Direct PCR Lysis Reagent (Viagen Biotech, Inc., Los Angeles, CA, USA) and subjected to polymerase chain reaction (PCR) amplification using specific sets of PCR primers for each genotype, including Pitx3-IRES2-tTA transgenic mice (Pitx3-F: GACTGGCTTGCCCTCGTCCCA and Pitx3-R: GTGCACCGAGGCCCCAGATCA) and LRRK2 transgenic mice (LRRK2-F: TATTGGGCTACAACCGGAAA and LRRK2-Prp-R: CTCCATCAAAGGGACCTGAA).

Immunohistochemistry and light microscopy

As described previously (34), mice were sacrificed and then perfused via cardiac infusion with 4% paraformaldehyde in cold phosphate-buffered saline. To obtain frozen sections, brain tissues were removed and submerged in 30% sucrose for 48 h and sectioned at 30 μm thickness using a cryostat (Leica CM1950, Buffalo Grove, IL, USA). Antibodies specific to TH (1:1000, Pel-Freez; 1:200, Sigma-Aldrich; 1:50, Santa Cruz), HA (16B12, 1:500, Covance Inc.; 1:500, Cell Signaling), GFAP (1:500, Sigma-Aldrich), Iba1 (1:500, Wako), PITX3 (1:500, Life Technologies), VMAT2 (1:1000) (56), GM130 (1:500, BD), GLG1 (1:500, Sigma-Aldrich), COX IV (1:500, Life Technologies), Calreticulin (1:500, Cell Signaling) and NURR1 (1:200, Santa Cruz) were used, as suggested by manufacturers. Alexa Fluor 488, Alexa Fluor 546 or Alexa Fluor 647-conjugated secondary antibody (1:500, Life Technologies) was used to visualize the staining. Nucleus was stained by To-pro3 (1:1000, Life Technologies). Fluorescence images were captured using a laser scanning confocal microscope (LSM 510; Zeiss, Thornwood, NJ, USA). The paired images in all the figures were collected at the same gain and offset settings. Post-collection processing was applied uniformly to all paired images. The images were either presented as a single optic layer after acquisition in z-series stack scans from individual fields or displayed as maximum intensity projections to represent confocal stacks.

Image analysis

For the quantitative assessment of various marker protein accumulations and distributions, images were taken using identical settings and exported to Image J (NIH, Bethesda, MD, USA) for imaging analyses. Images were converted to an eight-bit color scale (fluorescence intensity from 0 to 255) using Image J. Areas of interest were first selected by polygon or freehand selection tools and then subjected to measurement by mean optical intensities. The mean intensity for the background area was subtracted from the selected area to determine the net mean intensity.

Terminal analysis

Fluorescence images of DA terminal (striatum sections) stained with a TH antibody (1:1000, Pel-Freez, Rogers, AR, USA) were captured using a laser scanning confocal microscope (63×, LSM 510; Zeiss). The paired images were collected at the same gain and offset settings. Area fraction of each image was analyzed by ImageJ.

Stereology

According to the mouse brain in stereotaxic coordinates (3rd edition, Keith B.J. Franklin and George Paxinos), a series of coronal sections across the midbrain (40 μm per section, every fourth section from Bregma −2.54 to −4.24 mm, 10–12 sections per case) were processed for TH (1:1000, Pel-Freez) staining overnight and subsequently with ABC reagents (Vector Laboratories) for an additional hour. Visualization was performed using DAB kit (SK-4100, Vector Laboratories) for 5 min at room temperature. The number of TH-positive neurons was assessed using the optical fractionator function of Stereo Investigator 10 (MicroBrightField Inc., Williston, VT, USA). Three or more mice were used per genotype at each time point. Counters were blinded to the genotypes of the samples. The sampling scheme was designed to have coefficient of error less than 10% in order to obtain reliable results.

Tissue fractionation and western blot

Striatum and midbrain tissues were homogenized with 10 volumes of RIPA buffer plus protease and phosphatase inhibitor cocktail and centrifuged at 10 000g for 10 min. Protein concentrations in supernatant were measured by BCA (Thermo Fisher Scientific). Proteins were size-fractioned by 4–12% NuPAGE Bis–Tris-polyacrylamide gel electrophoresis (Invitrogen) using MES running buffer (Invitrogen). After transfer to nitrocellulose membranes, the membranes were immunoblotted with appropriate dilutions of the primary antibody: HA (16B12, Covance Inc.), LRRK2 (MJFF3), DARPP32 (Cell Signaling Technology), TH (Pel-Freez), ALDH1A1 (Sigma-Aldrich), DAT (Millipore), VMAT2 (1:1000) (56), β-actin (Cell Signaling) and β-tubulin (Cell Signaling) at 4°C. Signals were visualized by enhanced chemiluminescence development and quantified with ImageJ.

Behavior tests

Open-field test

As described previously (34), the ambulatory and rearing activities of mice were measured by the Flex-Field activity system (San Diego Instruments, San Diego, CA, USA). Flex-Field software was used to trace and quantify mouse movement in the unit as the number of beam breaks per 30 min.

Rotarod test

As described previously (34), mice were placed onto a rotating rod with auto-acceleration from 0 to 40 rpm for 5 min (San Diego Instruments). The length of time the mouse stayed on the rotating rod was recorded, across 10 trials.

Gait analysis

As described previously (57), the TreadScan Gait Analysis System (Clever Sys, Reston, VA, USA) was used to assess the nature of gait behaviors of mice. Each mouse was trained for 15 s at 8 cm/s. After 1 min of rest, the mouse would be recorded for 20 s at 100 frames/s at 8 cm/s. Stride length and gait angle were analyzed.

Fast-scan cyclic voltammetry

Following isoflurane anesthesia, brains were removed and 400 μm thick coronal sections through the striatum were prepared in carbogen-bubbled, cold high-sucrose solution (in mm: sucrose 194, NaCl 30, KCl 4.5, MgCl₂ 1, NaHCO₃ 26, NaH₂PO₄ 1.2, glucose 10). Slices were then transferred to a chamber filled with artificial oxygenated cerebrospinal fluid [aCSF, in mm: 126 NaCl, 2.5 KCl, 1.2 NaH₂PO₄, 2.4 CaCl₂, 1.2 MgCl₂, 25 NaHCO₃, 11 glucose, 20 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid, 0.4 l-ascorbic acid] at 32°C and allowed to recover for 1 h. Slices were then maintained at room temperature until used for recording. Cylindrical carbon-fiber microelectrodes (50–100 µm of exposed fiber) were prepared with T650 fibers (6 µm diameter; Goodfellow, Coraopolis, PA, USA) and inserted into a glass pipette. The carbon-fiber electrode was held at −0.4 V, and the potential was increased to 1.2 V and back at 400 V/s every 100 ms using a triangle waveform. Dopamine release was evoked by rectangular, electrical pulse stimulation (200–600 µA, 0.6 ms per phase, biphasic unless otherwise noted) applied every 5 min. Data collection and analysis were performed using the Demon Voltammetry and Analysis software suite (58). Ten cyclic voltammograms of charging currents were recorded as background before stimulation, and the average of these responses was subtracted from data collected during and after stimulation. Maximum amplitudes of extracellular dopamine transients were obtained from input/output function (I/O) curves. I/O curves were constructed by plotting stimulus current against concentration of DA response amplitude over a range of stimulus intensities. Dopamine uptake was analyzed by the decay time constants (τ) of the evoked signals. Following experiments, electrodes were calibrated using solutions of 1 and 10 µm dopamine in aCSF. Two-way analysis of variance (ANOVA) with Bonferroni's post hoc comparison in the case of significant interaction or Student's t-test was used for statistical comparison between groups.

High-performance liquid chromatography (HPLC)

After obtaining the wet weight of the tissue, the samples were homogenized in 500 µl of 0.1 N perchloric acid containing 100 µM EDTA. The tissue was sonicated; after centrifugation, the supernatant was frozen and stored at −80°C until assayed for dopamine and DOPAC by liquid chromatography with electrochemical detection (59). Briefly, mobile-phase solution containing octanesulfonic acid as an ion-pairing agent was pumped isocratically through a reversed-phase liquid chromatographic column. DA and DOPAC were quantified by the current produced after exposure of the eluate to a flow-through electrode set to oxidizing and then reducing potentials in series, with recordings from the last electrode reflecting reversibly oxidized species.

LCM and quantitative reverse transcriptase–PCR (qRT–PCR) assay

Brains of Pitx3/H2BjGFP control double-transgenic, Pitx3/H2BjGFP/LRRK2WT and Pitx3/H2BjGFP/G2019S triple-transgenic mice were quickly dissected out, and the frozen brains were sectioned at 20 µm thickness by a cryostat onto a PAN membrane frame slide (Applied Biosystems, Foster City, CA, USA) and stored at −80°C until LCM. The GFP-positive cells in SNpc were selected and captured by an ArturusXT microdissection system with fluorescent illumination (Applied Biosystems) onto LCM Macro Caps (Applied Biosystems) separately at the following working parameters: spot size, 7–25 µm; power, 50–70 mW and duration, 20–40 µs. The total RNA was extracted with the PicoPure Isolation kit (Applied Biosystems) after the protocol provided by the manufacturer. The cDNA was synthesized from 50 ng of RNA by the First Strand kit (QIAGEN, Valencia, CA, USA) after genomic DNA elimination. The SYBR Green real-time PCR detection method was used to quantitate the Aldh1a1, Th, Vmat2 and Dat expression levels in the control and LRRK2 transgenic SNpc DA neurons and normalized by β-action (ActB) expression. Aldh1a1, Th, Vmat2, Dat and ActB primers used were from QIAGEN and tested by the manufacturer.

Statistical analysis

Statistical analysis was performed using Graphpad Prism 5 (Graphpad Software, Inc., La Jolla, CA, USA). Data were presented as mean ± SEM. Statistical significances were determined by comparing means of different groups using Student's t-test, one-way and two-way ANOVAs and post hoc analysis.

Supplementary Material

Supplementary Material is available at HMG online.

Funding

This work was supported in part by the intramural research programs of the National Institutes of Health (NIH)—National Institute on Aging (grants AG000928 and AG000944 to H.C.) and NIH—National Institute on Alcohol Abuse and Alcoholism (D.M.L.).

Acknowledgements

We thank the Combined Technical Research Core of National Institute of Dental and Craniofacial Research (NIDCR) for the use of the Laser Capture Microdissection equipment, Dr David Goldstein and Ms Patricia Sullivan of NINDS for early HPLC analysis, Dr Daniel R. Principe for editing and proofreading the manuscript and Cai laboratory members for their constructive suggestions.

Conflict of Interest statement. The authors have declared that no conflict of interest exists.

Author notes

†

These authors contributed equally to this work.

‡

Present address: Synapse and Neural Circuit Research Unit, National Institute of Neurological Diseases and Stroke, National Institutes of Health, Bethesda, MD 20892, USA.

¶

Present address: Department of Anatomy, Zhongshan School of Medicine, Sun Yat-Sen University, Guangzhou, China.