SEARCH

SEARCH BY CITATION

Keywords:

  • dopaminergic neuron;
  • endoplasmic reticulam stress;
  • mitochondrial complex I;
  • Pael-R;
  • parkin

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

Parkin, a ubiquitin ligase, is responsible for autosomal recessive juvenile parkinsonism (AR-JP). We identified parkin-associated endothelin receptor-like receptor (Pael-R) as a substrate of parkin, whose accumulation is thought to induce unfolded protein response (UPR) -mediated cell death, leading to dopaminergic neurodegeneration. To create an animal model of AR-JP, we generated parkin knockout/Pael-R transgenic (parkin-ko/Pael-R-tg) mice. parkin-ko/Pael-R-tg mice exhibited early and progressive loss of dopaminergic as well as noradrenergic neurons without formation of inclusion bodies, recapitulating the pathological features of AR-JP. Evidence of activation of UPR and up-regulation of dopamine and its metabolites were observed throughout the lifetime. Moreover, complex I activity of mitochondria isolated from parkin-ko/Pael-R-tg mice was significantly reduced later in life. These findings suggest that persistent induction of unfolded protein stress underlies chronic progressive catecholaminergic neuronal death, and that dysfunction of mitochondrial complex I and oxidative stress might be involved in the progression of Parkinson’s disease. parkin-ko/Pael-R-tg mice represents an AR-JP mouse model displaying chronic and selective loss of catecholaminergic neurons.

Abbreviations used
AR-JP

autosomal recessive juvenile parkinsonism

DOPAC

3, 4-dihydroxyphenylacetic acid

ER

endoplasmic reticulum

HVA

homovanillic acid

LC

locus coeruleus

PD

Parkinson’s disease

PDGF

platelet-derived growth factor

TH

tyrosine hydroxylase

UPR

unfolded protein response

Parkinson’s disease (PD) is an age-related movement disorder including tremor, rigidity, bradykinesia and postural instability. A variety of evidence has strongly demonstrated that these motor symptoms are related to the selective degeneration of dopaminergic neurons in the substantia nigra pars compacta (SNpc) (Olanow and Tatton 1999). However, the underlying molecules and cellular pathways that mediate neuronal death remain poorly defined. Although the majority of PD cases appear to present without a family history of disease, several genes such as α-synuclein, parkin, DJ-1, PINK-1, and LRRK2 have been identified as the genes responsible for familial PD. Among genes identified to date, mutations of parkin have been shown to be the cause of autosomal recessive juvenile parkinsonism (AR-JP) (Kitada et al. 1998; Lucking et al. 2000). Parkin functions as a protein-ubiquitin ligase (E3) and loss of E3 activity causes AR-JP, suggesting that substrate(s) of parkin, which cannot be properly degraded and accumulates in AR-JP patients, may cause dysfunction and eventually the death of susceptible neurons (Imai et al. 2000; Shimura et al. 2000; von Coelln et al. 2004a). In addition, increasing evidence suggests that partial loss of parkin E3 function resulting from S-nitrosylation or covalent modification by dopamine may also play a role in late-onset and sporadic PD (Chung et al. 2004; Yao et al. 2004; LaVoie et al. 2005). Parkin-associated endothelin receptor-like receptor (Pael-R/GPR37) was identified as an intracellular substrate of parkin, which has a propensity to accumulate in an unfolded form in the endoplasmic reticulum (ER) (Imai et al. 2001). When over-expressed in cultured cells, Pael-R tends to become misfolded and insoluble, inducing ER stress, and ultimately leading to cell death. Parkin ubiquitinates and promotes the degradation of misfolded species of Pael-R, resulting in the suppression of ER stress-induced cell death. The finding that panneuronal expression of Pael-R in Drosophila causes progressive selective loss of dopaminergic neurons further strongly supports a pathogenetic role for Pael-R in AR-JP (Yang et al. 2003). However, none of parkin null mouse models demonstrates either alteration in gross brain morphology or dopaminergic neuronal loss except for a recent report by Rodriguez-Navarro et al. (Goldberg et al. 2003; Itier et al. 2003; Von Coelln et al. 2004b; Perez and Palmiter 2005; Rodriguez-Navarro et al. 2007). A number of α-synuclein transgenic (tg), DJ-1- or PINK1- null mice models have also been created (Fernagut and Chesselet 2004; Goldberg et al. 2005; Kim et al. 2005). Although each of these models reproduces some of the pathological features of PD, obvious degeneration of dopaminergic neurons is not observed. The apparent preservation of dopaminergic neurons in these genetically modified animals suggests that obvious dopaminergic neuronal death may occur over a more protracted time scale than the average life spans of experimental animals, or that additional pathogenic event(s) may be required to induce such cell death.

Given that RNA interference (RNAi)-mediated down-regulation of endogenous parkin enhances the neurodegeneration of Pael-R tg Drosophila (Yang et al. 2003), deletion of the parkin gene in Pael-R tg mice may enhance the accumulation of Pael-R, resulting in neuronal degeneration. To test this hypothesis, we generated parkin-deficient/Pael-R-over-expressing double-mutant mice by crossbreeding parkin knockout (ko) mice with Pael-R tg mice. Here we show that parkin-ko/Pael-R-tg double-mutant mice exhibit early and progressive loss of dopaminergic neurons without formation of inclusion bodies, recapitulating the pathological characteristics of AR-JP. We provide compelling in vivo evidence of a mechanism linking progressive neuronal degeneration with persistent chronic ER stress. We also report that parkin-ko/Pael-R-tg double-mutant mice exhibit down-regulation of Ndufs4 and Ndufa10, two phosphorylated subunits of mitochondrial complex I, resulting in decrease in activity of mitochondrial complex I, suggesting that impairment of complex I activity might be the common pathway involved in various forms of PD including AR-JP. Moreover, the dopamine up-regulation caused by Pael-R over-expression seems to enhance oxidative stress, contributing to selective dopaminergic neuronal death.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

Generation of double-mutant mice over-expressing human Pael-R on a parkin null background

Transgenic mice that express human Pael-R under the control of the murine prion (PrP) or platelet-derived growth factor (PDGF) promoter (Imai et al. 2007) as well as the exon 3-deleted parkin null mice (Kitao et al. 2007) were produced as described. Double-mutant mice were generated by crossbreeding these two existing mouse lines. In the first step, parkin ko mice were bred to heterozygous Pael-R transgenic mice. Double-heterozygous mice generated in the first round of breeding were once more crossed with parkin ko mice to generate heterozygous Pael-R transgenic mice on a parkin null background (parkin-ko/non-tg or parkin-ko/Pael-R-het-tg). The parkin-ko/PrP- or PDGF-Pael-R-hetero-tg were crossbred with each other to generate parkin null mice without Pael-R transgene (parkin-ko/Pael-R-non-tg), heterozygous or homozygous Pael-R transgenic mice lacking parkin (parkin-ko/PrP- or PDGF-Pael-R-het-tg and parkin-ko/PrP- or PDGF-Pael-R-homo-tg, respectively). In two successive breeding steps, cohorts of littermates with or without endogenous parkin expression and with or without transgenic Pael-R expression were generated in ratios consistent with Mendelian principles. Age-matched littermate mice were used in all experiments. All procedures involving animals conformed to the guidelines of the Institutional Animal Care Committee of RIKEN BSI and Kyoto University Graduate School of Medicine.

Immunohistochemistry

Mice were injected with pentobarbital (100 mg/kg; Sigma, St Louis, MO, USA) and perfused transcardially with ice-cold phosphate-buffered saline (PBS, pH 7.4), followed by 4% paraformaldehyde in PBS. Serial coronal sections at 16-μm thickness were collected on slides. Using standard avidin–biotin peroxidase method (Elite standard kit SK6100; Vector Laboratories, Burlingame, CA, USA), deparaffinized sections were stained with primary antibodies against tyrosine hydroxylase (TH, Chemicon, Temecula, CA, USA), α-synuclein (BD Transduction Laboratories, Lexington, KY, USA), BiP (Stressgen, Collegeville, PA, USA), CHOP (Santa Cruz Biotechnology, Santa Cruz, CA, USA) and ubiquitin (Dako, Carpinteria, CA, USA). For co-localization of TH with BiP or CHOP, deparaffinized sections were doubly stained with TH and BiP or CHOP, followed by reaction with Alexa 488- and 546-conjugated secondary antibodies (Molecular Probes, Eugene, OR, USA), and then examined with a LSM 510 confocal laser-scanning microscope (Carl Zeiss, Inc., Minneapolis, MN, USA).

Stereological analysis

Total numbers of TH-positive or Nissl-positive neurons in SNpc and locus coeruleus (LC) were determined using an unbiased optical fractionator method (Stereoinvestigator, MicroBrightField) as previously described (West 1993; Goldberg et al. 2003; Von Coelln et al. 2004b).

Behavioral tests

The mouse cohort for behavioral tests comprised 20 parkin-ko/Pael-R-non-tg mice, parkin-ko/PrP-Pael-R-het-tg mice and parkin-ko/PrP-Pael-R-homo-tg mice each. All tests were carried out by investigators blinded to the genotype of the animals being tested.

Western blot

Immunolabeling was performed using primary antibodies against α-synuclein (BD Transduction Laboratories), Pael-R (Imai et al. 2001), PKR-like ER-resident kinase (PERK, Santa Cruz), XBP1 (Santa Cruz), BiP (BD Transduction Laboratories), caspase-12 (Oncogene, Cambridge, MA, USA) JNK1/2 (Santa Cruz), phospho-JNK1/2 (Santa Cruz), CHOP (BioLegend, San Diego, CA, USA), TH (Chemicon), dopamine transporter (Chemicon), and vesicular monoamine transporter 2 (VMAT2, Chemicon), as well as horseradish peroxidase-conjugated secondary antibodies and ECL solutions (Amersham Pharmacia Biotech, Piscataway, NJ, USA). For densitometric analysis, images were scanned and densitometry was performed using the NIH IMAGE 1.4 software (Scion Corporation, Frederick, MD, USA).

RNA extraction and real-time PCR

Total RNA was isolated from freshly dissected midbrains using TRIzol Reagent (Invitrogen, Carlsbad, CA, USA) and first strand cDNA was synthesized from 2 μg of total RNA using SuperScriptTM II Rnase H- Reverse Transcriptase (Invitrogen) according to the manufacturer’s instructions. Real-time PCR analysis was performed in triplicate on the ABI prism 7000 sequence detection system (Applied Biosystems, Foster City, CA, USA) using the SYBR Green PCR Master mix (Applied Biosystems). Results of real-time PCR were normalized against those for actin and plotted as ratio versus parkin-ko/Pael-R-non-tg.

Mitochondrial preparation and complex I, II–III, and IV activity assays

The mitochondrial complexes I, II–III, and IV activity assays were performed in triplicate on isolated mitochondria preparations as previously described (Hsu et al. 2005). The mitochondrial complexes and complex I subunits were detected using Total OXPHOS Complexes Detection Kit (MitoScience, Eugene, OR, USA), anti-8 kDa subunit monoclonal antibody (MitoScience), anti-18 kDa subunit monoclonal antibody (MitoScience) and anti-42 kDa subunit polyclonal antibody (Biocompare). Anti-COXIV monoclonal antibody (Molecular Probes) and anti-porin monoclonal antibody (Calbiochem, San Diego, CA, USA) were used as loading controls.

Measurement of catecholamines (HPLC)

To determine the concentration of catecholamines in striatal tissues by HPLC with electrochemical detection, male mice (n = 12 each) were decapitated, striata were dissected. The tissue was weighed and sonicated in 0.5 ml of ice-cold 0.1 M perchloric acid, to which 3, 4-dihydroxybenzylamine (DHBA) (Sigma) was added as the internal standard. DA and metabolites were detected with series coulometric detector (ESA, Inc., Chelmsford, MA, USA). Data were collected and processed on a CHROMELEONTM Chromatography Data Systems 6.40 (Dionex, Sunnyvale, CA, USA).

Protein carbonyl assay

Brain homogenate was assayed for protein carbonyls according to the manufacturer’s instructions (OxyBlotTM Protein Oxidation Detection Kit, Chemicon).

Statistics

All values are presented as the mean ± SEM. Results were tested for significance using one-way Anova, followed by the Bonferoni post hoc test (spss 15.0 software). A significance level of < 0.05 was used.

Results

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

Generation of parkin-ko/Pael-R-tg double-mutant mice

Transgenic mice over-expressing human Pael-R driven by murine PrP or PDGF promoter, designated PrP-Pael-R tg or PDGF-Pael-R tg mice, respectively, were generated as previously shown (Imai et al. 2007). PrP-Pael-R tg mice exhibited mild but significant loss of dopaminergic neurons (Fig. 1a). Neurodegeneration was not observed in PDGF-Pael-R tg mice even at the age of 2 years (Fig. 1b), probably due to relatively low expression level of Pael-R (Imai et al. 2007). Two lines of PrP- and PDGF-Pael-R tg mice were bred with parkin null mice (Kitao et al. 2007) to generate parkin null without Pael-R transgene (parkin-ko/Pael-R-non-tg), parkin null with Pael-R heterozygous tg (parkin-ko/PrP- or PDGF-Pael-R-het-tg) and Pael-R homozygous tg (parkin-ko/PrP- or PDGF-Pael-R-homo-tg) mice. We confirmed that 26-month-old parkin-ko mice did not display dopaminergic cell loss as compared with wild-type mice (Fig. 1c). Pael-R tended to be more insoluble in the midbrain region (Fig. S1), and throughout life the steady-state levels of both endogenous Pael-R and over-expressed human Pael-R remained unchanged (data not shown). Although parkin deletion had no effect on endogenous Pael-R monomer levels, both Triton X-100 soluble and insoluble fractions of over-expressed Pael-R monomer were significantly increased in parkin-ko/PrP-Pael-R-het-tg mice compared with PrP-Pael-R-het-tg mice with wild-type (wt) parkin (Fig. 1d and e). A higher molecular-weight band corresponding to the size of Pael-R dimer was constantly observed in all the samples (Fig. 1d). The high-molecular-weight bands were also observed in Pael-R null mice, indicating that they at least partially represent non-specific signals (Fig. 1f). However, given that non-specific signals are assumed to be almost identical between Pael-R null and wild-type mice, the decreased signal intensity in higher molecular-weight bands in Pael-R null mice compared with wt mice indicates that the higher molecular-weight bands in Pael-R expressing mice comprise both Pael-R dimer and the non-specific signals, and the former species are increased in the Triton X-100 insoluble fraction of parkin null mouse brains (Fig. 1d–f). Taken together, these data in Fig. 1(e) suggest that although less prominent than Pael-R transgene, insoluble species of endogenous Pael-R aggregates are likely to be increased in parkin null mice from 6 months of age.

image

Figure 1.  Age-dependent loss of TH-positive neurons in Pael-R tg mice and absence of parkin increases steady-state levels of Pael-R. (a) Number of TH-positive neurons in Pael-R tg mice driven by PrP promoter. Values are the mean ± SEM (n = 10). *, < 0.05, versus wild-type. (b) Number of TH-positive neurons in Pael-R tg mice driven by PDGF promoter (n = 6). (c) Number of TH-positive neurons in parkin ko mice (n = 9). (d) Representative western blot analysis of Triton X-100 soluble and insoluble lysates from midbrain of 6- and 18-month-old mice was performed using antibody recognizing both mouse and human Pael-R. The molecular weight was noted on the left. An antibody against actin was used as a loading control. (e) The expression levels of Pael-R (monomer, the band at approximately 38 kDa; total, the bands including Pael-R monomer, non-specific/Pael-R dimer, as well as smeared bands between Pael-R monomer and dimer) were quantified using optical density and normalized to that of actin. The protein levels are relative to those of parkin wt/Pael-R non-tg defined as 1. The data are presented as the mean ± SEM (n = 6). *, < 0.05; **, < 0.01. (f) Western blot analysis was performed on brain lysates from wt or Pael-R ko mice. The molecular weight was noted on the left.

Download figure to PowerPoint

Age-related neurodegeneration of dopaminergic neurons in parkin-ko/Pael-R-tg double-mutant mice

Progressive catecholaminergic neuronal loss was observed in both PrP-Pael-R and PGDF-Pael-R tg mice crossed with parkin null mice.

Parkin-ko/PrP-Pael-R-tg mice exhibited decreased number of TH-positive neurons in the substantia nigra pars compacta (SNpc) and locus ceruleus (LC) regions compared with parkin-ko/Pael-R-non-tg mice (Fig. 2a–d). Unbiased stereological analyses revealed no difference in the number of TH-positive neurons in the SNpc of 3-month-old mice as assessed by either TH or Nissl staining. However, the number of TH-positive neurons exhibited age-dependent reduction beginning as early as 6 months in the parkin-ko/PrP-Pael-R-homo-tg double-mutant mice (Fig. 2e). The finding of similar loss of Nissl-positive neurons confirmed that the decrease of TH-positive neurons did not result from reduction of TH immunoreactivity, but from the loss of neurons per se (Fig. 2f). Similar loss of TH-positive neurons in the ventral tegmental area (VTA) and LC regions was also observed (Fig. S2a and b). Obvious loss of TH-positive neurons was observed at 18 months and 12 months in the PrP-Pael-R tg mice with and without parkin, respectively, indicating that loss of parkin predated the onset of neurodegeneration. The enhancement of Pael-R toxicity by parkin deficiency was more prominent in Pael-R tg mice driven by PDGF. Loss of TH-positive neurons became evident at 18 months of age in parkin-ko/PDGF-Pael-R-het-tg mice (Fig. 2g and h); whereas no frank neurodegeneration was observed in PDGF-Pael-R-het-tg mice expressing endogenous parkin even at 2 years of age (Fig. 1b). The number of hippocampal neurons in the dentate gyrus region, which also expressed high levels of Pael-R transgene (Fig. S1), was unaltered in parkin-ko/PrP-Pael-R-tg double-mutant mice (Fig. S3), suggesting that the cell loss occurred in catecholaminergic neuron-specific manner. Since parkin-ko/PrP-Pael-R-tg demonstrated more robust phenotype, we mainly analyzed parkin-ko/PrP-Pael-R-tg mice in the following analyses.

image

Figure 2.  Degeneration of TH-positive neurons in parkin-ko/PrP-Pael-R-tg double-mutant mice. (a and b) Representative photomicrographs of TH-immunoreactivity in SNpc of 24 months parkin-ko/Pael-R-non-tg and parkin-ko/PrP-Pael-R-homo-tg mice, respectively. (c and d) Representative photomicrographs of TH-immunoreactivity in LC of 24 months parkin-ko/Pael-R-non-tg and parkin-ko/PrP-Pael-R-homo-tg mice, respectively. (e and f), Number of TH-positive (e) or Nissl-positive neurons (f) in SNpc of parkin-ko/PrP-Pael-R-tg double-mutant mice. (g and h), Number of TH-positive (g) or Nissl-positive neurons (h) in SNpc of parkin-ko/PDGF-Pael-R-tg double-mutant mice. Data are expressed as mean ± SEM (n = 10). *, < 0.05, **, < 0.01 versus parkin-ko/Pael-R-non-tg.

Download figure to PowerPoint

Tyrosine hydroxylase optical density in the striatum was also reduced in parkin-ko/PrP-Pael-R-homo-tg mice (Fig. 3a–c). This reduction in TH protein levels was confirmed by western blot at 18- and 24-month-old double-mutant mice. Moreover, there were significant decreases in dopamine transporter and VMAT2 protein levels at these ages, consistent with the loss of dopaminergic nerve terminals (Fig. 3d and e).

image

Figure 3.  Reduced TH-immunoreactivity in the striatum of parkin-ko/PrP-Pael-R-tg double-mutant mice. (a–c) Representative images of TH-immunoreactivity in the striatum of 24 months parkin-ko/Pael-R-non-tg, parkin-ko/PrP-Pael-R-het-tg and parkin-ko/PrP-Pael-R-homo-tg of mice, respectively. (d) Representative western blot images of TH, DAT and VMAT2 in young (6 M) and aged (24 M) mouse striatum. Actin was used as a loading control. (e) The expression levels of TH, DAT and VMAT2 (normalized to actin) were quantified using optical density. Data are expressed as mean ± SEM (n = 6). *, < 0.05, **, < 0.01 versus parkin-ko/Pael-R-non-tg.

Download figure to PowerPoint

No obvious behavioral defects in parkin-ko/PrP-Pael-R-tg double-mutant mice

Although observation of spontaneous, voluntary movements over 30 min in the open field test revealed that parkin-ko/PrP-Pael-R-tg mice tended to be more active at younger ages and less active at old ages compared with parkin-ko/Pael-R-non-tg mice, the difference failed to reach statistical significance (data not shown). In addition, these mice were similarly able to maintain their balance on the rotarod before falling off when young (data not shown). Although a tendency towards poor performance was observed in parkin-ko/PrP-Pael-R-tg mice at later stages, no significant difference was observed compared with parkin-ko/Pael-R-non-tg mice (data not shown).

Evidence for ER stress and activation of unfolded protein response in parkin-ko/PrP-Pael-R-tg double-mutant mice

We hypothesized that perturbation of ER homeostasis and triggering activation of unfolded protein response (UPR) underlies the progressive loss of dopaminergic neurons in parkin-ko/PrP-Pael-R-tg double-mutant mice. In the midbrain of parkin-ko/PrP-Pael-R-tg double-mutant mice, significant increase in BiP mRNA levels was clearly detected at the age of 6 months and maintained higher levels of transcription throughout life (Fig. 4a). On the other hand, the spliced form of XBP1 mRNA produced by 26-nucleotide splicing from primary XBP1 mRNA via the ribonuclease activity of IRE1 was increased only in early stages (Fig. 4a). Correspondingly, levels of BiP protein persistently increased in parkin-ko/PrP-Pael-R-tg double-mutant mice, whereas levels of XBP-1 increased only in the early stages (Fig. 4b and c). Collectively, these findings suggest that the IRE1 and ATF6 pathway of the UPR is activated in response to Pael-R accumulation in vivo.

image

Figure 4.  Chronic and persistent activation of UPR in parkin-ko/Pael-R-tg double-mutant mice. (a) Real-time PCR was performed on mRNA extracted from the midbrain of parkin-ko/Pael-R-tg double-mutant mice. Data are expressed as mean ± SEM (n = 6). *, < 0.05, **, < 0.01 versus parkin-ko/Pael-R-non-tg. (b) Representative western blot images of BiP, phosphorylated form of PERK (p-PERK) and XBP-1 in the middle brain region of young (6 M) and aged (24 M) mice. Actin was used as a loading control. (c) The expression levels of BiP, p-PERK and XBP-1 protein (normalized to actin) were quantified using optical density. Data are expressed as mean ± SEM (n = 6). *, < 0.05, **, < 0.01 versus parkin-ko/Pael-R-non-tg.

Download figure to PowerPoint

On the other hand, PERK phosphorylation was unaltered at all time points examined (Fig. 4b and c), suggesting that global translational suppression induced by PERK phosphorylation may be transient in parkin-ko/Pael-R-tg double-mutant mice.

Evidence for ER-stress-mediated cell death: activation of JNK2, caspase-12 and CHOP/GADD153 in parkin-ko/Pael-R-tg double-mutant mice

Levels of CHOP were already substantially higher in parkin-ko/PrP-Pael-R-tg double-mutant mice at 6 months of age, followed by a nearly identical pattern of expression throughout life (Fig. 5a). The mRNA levels of caspase-12 and JNK2 were also increased at 12 months (Fig. 5a). The increase in levels of CHOP, JNK2 and caspase-12 mRNAs were accompanied by up-regulation of CHOP protein, phosphorylated JNK2 and cleaved form of caspase-12 (Fig. 5b and c).

image

Figure 5.  Up-regulation of CHOP, JNK2 and caspase-12 in parkin-ko/PrP-Pael-R-het-tg and parkin-ko/PrP-Pael-R-homo-tg mice. (a) Real-time PCR was performed on the same samples as in Fig. 4a. Values were standardized to the level of actin mRNA and data are expressed as mean ± SEM (n = 6). *, < 0.05, **, < 0.01 versus parkin-ko/Pael-R-non-tg. (b) Representative western blot images of caspase-12 (Casp-12), CHOP, and phosphorylated form of JNKs (p-JNK1/2) in the midbrain region of young (6 M) and aged (24 M) mice. Actin was used as a loading control. The pan-JNK1/2 antibody was also used to assess the total amount of JNK proteins. The asterisk indicates a non-specific signal. (c) The expression levels of CHOP, p-JNK2 and active forms of caspase-12 (p42 and p20) (normalized to actin) were quantified using optical density. Data are expressed as mean ± SEM (n = 6). *, < 0.05, **, < 0.01 versus parkin-ko/Pael-R-non-tg.

Download figure to PowerPoint

Specific activation of UPR in dopaminergic neurons in parkin-ko/Pael-R-tg double-mutant mice

To confirm the specific occurrence of ER stress in dopaminergic neurons of the SNpc, double staining of TH and BiP or CHOP was performed. Up-regulation of BiP was restricted to TH-positive neurons and almost all TH-positive neurons exhibited intense expression of BiP in aged parkin-ko/PrP-Pael-R-tg double-mutant mice (Fig. S4a). Intense expression of CHOP was also observed in dopaminergic neurons in parkin-ko/PrP-Pael-R-tg double-mutant mice (Fig. S4b).

Evidence for impairment of complex I in parkin-ko/Pael-R-tg double-mutant mice

An approximately 30% reduction in complex I activity was observed in mitochondria isolated from parkin-ko/PrP-Pael-R-tg double-mutant mouse midbrain as well as the whole brain obtained from 18- and 24-month-old mice (Fig. 6a and data not shown), whereas there was no significant differences in complex II-III or IV activities (data not shown).

image

Figure 6.  Impaired mitochondrial complex I in parkin-ko/PrP-Pael-R-het-tg and parkin-ko/PrP-Pael-R-homo-tg mice. (a) Complex I activity analysis was performed on mitochondria isolated from parkin-ko/PrP-Pael-R-tg double-mutant mouse midbrains at later stages. Data are expressed as mean ± SEM (n = 6). *, < 0.05 versus parkin-ko/Pael-R-non-tg. (b) Real-time PCR was performed on mRNA extracted from 18–24 months of parkin-ko/PrP-Pael-R-tg double-mutant mouse midbrains. The values were standardized to the level of actin mRNA and expressed as mean ± SEM (n = 6). *, < 0.05 versus parkin-ko/Pael-R-non-tg. (c) Representative western blot images of subunits of complex I (Ndufa10, Ndufs4 and CI-8 kDa) and levels of complexes. Porin and COXIV were used as loading controls. (d) The expression levels of Ndufa10, Ndufs4, CI-8 kDa and complex I–V (normalized to COXIV) were quantified using optical density. Data are expressed as mean ± SEM (n = 6). *, < 0.05 versus parkin-ko/Pael-R-non-tg.

Download figure to PowerPoint

To investigate how Pael-R over-expression reduces complex I activity, we used a microarray approach to search for molecular markers and found that two transcripts, encoding Ndufs4 and Ndufa10, were significantly down-regulated. Consistent with the GeneChip data, semi-quantitative real time PCR confirmed decreased expression of Ndufs4 and Ndufa10 in parkin-ko/PrP-Pael-R-tg double-mutant mice when values were normalized to those for the house-keeping genes actin or 18S RNA (Fig. 6b). The reduction in expression of Ndufs4 (18 kDa) and Ndufa10 (42 kDa) proteins was confirmed by immunoblot analysis (Fig. 6c and d). We further used a mixture of monoclonal antibodies directed against various proteins in complexes of the electron transport chain to investigate the assembly of complex I in parkin-ko/PrP-Pael-R-tg mice. Quantitative band densitometry of western blot images revealed minor but significant reduction of protein from complex I, but not those from complexes II-V (Fig. 6c and d).

Abnormality of DA and its metabolites in parkin-ko/Pael-R-tg double-mutant mice

Striatal levels of DA were significantly increased in younger parkin-ko/PrP-Pael-R-het-tg and parkin-ko/PrP-Pael-R-homo-tg mice compared with age-matched parkin-ko/Pael-R-non-tg mice. Over time, however, the levels of DA in parkin-ko/PrP-Pael-R-tg double-mutant mice gradually decreased, and eventually significantly decreased at 24 months. Accordingly, the levels of 3, 4-dihydroxyphenylacetic acid (DOPAC) and homovanillic acid (HVA), the major metabolites of DA, were significantly increased in parkin-ko/PrP-Pael-R-homo-tg and parkin-ko/PrP-Pael-R-het-tg at younger stages (Fig. 7).

image

Figure 7.  HPLC analysis of DA and its metabolites in the striatum of parkin-ko/PrP-Pael-R-tg double-mutant mice. Data are expressed as mean ± SEM (n = 12). *, < 0.05, **, < 0.01 versus parkin-ko/Pael-R-non-tg.

Download figure to PowerPoint

Evidence for oxidative damage in parkin-ko/Pael-R-tg double-mutant mice

It is widely believed that dopamine can induce neurotoxic effects via the formation of highly reactive oxygen species, quinones and semiquinones generated by dopamine auto-oxidation or via its enzymatic metabolism by MAO, leading to oxidative stress. The finding of increased levels of dopamine and its metabolites in parkin-ko/PrP-Pael-R-tg double-mutant mice therefore prompted us to investigate whether increased levels of oxidized proteins could be detected in these mice. Both parkin-ko/Pael-R-non-tg and parkin-ko/PrP-Pael-R-tg double-mutant mice represented age-dependent increases in levels of protein carbonyls, a general marker of oxidative damage. Levels of protein carbonyls were significantly higher in parkin-ko/PrP-Pael-R-tg double-mutant mice compared with parkin-ko/Pael-R-non-tg controls in the midbrain region (Fig. 8a), but not in the cortex region (Fig. 8b), suggesting that dopamine and/or its metabolites play important roles in the production of protein carbonyls.

image

Figure 8.  Increased oxidative damage in parkin-ko/PrP-Pael-R-tg double-mutant mice. Carbonyl proteins were evaluated using lysates isolated from midbrain (a) and cortex (b). Incubation of the same blots with anti-actin antibody confirmed equivalent loading of proteins in each lane.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

Recently, we have published a mouse model of PD by infecting Pael-R encoding adenovirus in the substantia nigra and shown that Pael-R over-expression in vivo leads to ER stress-induced death of dopaminergic neurons over a couple of weeks (Kitao et al. 2007). To confirm the results in a genetic mouse model, we have generated parkin-ko/Pael-R-tg double-mutant mice, in which ER stress is evoked and selective and progressive catecholaminergic neuronal death without inclusion body formation occurs over 2 years, which recapitulates the main features of AR-JP. We obtained multiple lines of evidence indicating that chronic and persistent ER stress causes progressive loss of dopaminergic neurons over a long period of time in parkin-ko/Pael-R-tg double-mutant mice, providing a new genetic animal model to further explore the pathogenesis of PD.

In this study, parkin deletion promoted the accumulation of both soluble and insoluble Pael-R derived from the transgene, further supporting the idea that Pael-R is the substrate of parkin. Regarding the dopaminergic cell loss, it starts in PrP-Pael-R-het-tg mice at the age of 18 months, whereas it does in Pael-R-het -tg/parkin-ko double mutant mice at the age of 12 months (Figs 1a and 2e). Moreover, there were no difference in the dopaminergic cell number between 26-month-old PDGF-Pael-R-het-tg and wt mice, whereas 24-month-old PDGF-Pael-R-het-tg mice crossed with parkin-ko mice displayed reduced dopaminergic cell number compared with parkin-ko mice of the same age (Figs 1b, 2g and h). These data indicate that parkin deletion promotes neuronal loss by Pael-R accumulation.

Very recently, it was reported that around 35% TH-positive cell loss in the substantia nigra accompanied by motor behavioral abnormalities in foot print analysis occur in 24-month-old parkin-ko mice, raising the possibility that neurodegeneration phenotype of parkin-ko was simply added to that of Pael-R-tg mice(Rodriguez-Navarro et al. 2007). However, this is unlikely in our system, since 26-month-old parkin-ko mice displayed no dopaminergic neurodegeneration (Fig. 1c). The reasons for the discrepancies are unclear at this moment. Although we did not perform foot print analyses, they might have detected abnormalities in parkin-ko/Pael-R-homo-tg double mutant mice which display 40% TH-positive cell loss at the age of 24 months.

Speculation regarding the involvement of ER stress in neuronal death has grown recently, due in part to reports of activation of the UPR in in vitro and in vivo models of neurodegenerative diseases (Nakagawa et al. 2000; Imai et al. 2001; Southwood et al. 2002; Takahashi and Imai 2003; Rao and Bredesen 2004; Tessitore et al. 2004). BiP has been shown to accompany ER stress and to be anti-apoptotic, while the failure of cells to counteract ER stress initiates activation of multiple pathways that lead to apoptosis (Breckenridge et al. 2003). CHOP is a member of CCAAT/enhancer-binding protein family that is induced by ER stress and participates in ER stress-mediated apoptosis (Oyadomari and Mori 2004). Excessive ER stress can also activate caspase-12, which resides on the outside of ER membrane (Nakagawa et al. 2000). Moreover, it has been shown that the ER transmembrane kinase/nuclease IRE1 can activate the c-Jun N-terminal kinase (JNK) by recruiting TRAF2 in response to ER stress (Urano et al. 2000). The early and consistent up-regulation of BiP, accompanied by activation of caspase-12, CHOP and JNK2 in parkin-ko/PrP-Pael-R-tg double-mutant mice might represent cellular efforts to relieve ER stress. Over time, however, the cellular mechanisms fail to correct the continuous protein-folding defects, eventually leading to activation of multiple ER stress-mediated apoptotic processes. The finding of lack of alteration of phosphorylation of PERK is interesting, and suggests that persistent UPR induced in parkin-ko/PrP-Pael-R-tg double-mutant mice is not identical to conventional acute UPR.

Chronic and mild ER stress is known to induce UPR which allows for adaptation, instead of apoptosis, although UPR is designed to facilitate both adaptation to stress and apoptosis. Kaufmann and his colleagues have recently reported that survival is favored during chronic stress as a result of the intrinsic instabilities of mRNAs and proteins that promote apoptosis (Rutkowski et al. 2006; Rutkowski and Kaufman 2007). Consistent with their data, the scale of BiP up-regulation of double mutant mice at 24 months of age was greater than that of CHOP (Figs 4b and 5b).

It has been reported by several groups that complex I is decreased in the substantia nigra, skeletal muscle and platelets of patients with PD. Moreover, complex I inhibitors such as 1-methyl-4-phenyl-1,2,2,6-tetrahydropyridine and rotenone have been shown to cause dopaminergic cell death (Mizuno et al. 1998; Betarbet et al. 2000; Schapira 2001; Dauer and Przedborski 2003). More recently, phosphatase and tensin homologue-induced kinase 1 (PINK1), a mitochondrial protein, and DJ-1, a protein involved in oxidative stress partly located at mitochondria, turned out to be the genes responsible for familial PD termed PARK6 and PARK7, respectively (Dawson and Dawson 2003; Miller et al. 2003; Shen and Cookson 2004; Valente et al. 2004). These findings strongly support the idea that mitochondrial dysfunction, especially complex I deficiency plays a crucial role in the pathogenesis of PD. In this study, we found that complex I activity was decreased in Pael-R over-expressing mice, suggesting an important link between ER stress and mitochondrial dysfunction, both of which are thought to be involved in the pathogenetic mechanisms underlying PD.

In parkin-ko/Pael-R-tg double-mutant mice, the decrease in complex I activity was ascribable to transcriptional down-regulation of Ndufs4 and Ndufa10 subunits. Although we examined whether tunicamycin, thapsigargin or Pael-R over-expression-induced UPR is responsible for the down-regulation of Ndufs4 and Ndufa10 in cultured cells, only negative results were obtained (data not shown). The mechanism underlying down-regulation of Ndufs4 and Ndufa10 subunits of complex I in Pael-R tg mice is thus still unknown.

It is worth noting that down-regulated subunits Ndufs4 and Ndufa10 in PrP-Pael-R-tg mice are two exclusively phosphorylated proteins among complex I subunits (Schulenberg et al. 2003, 2004; Smeitink et al. 2004). Phosphorylation of mitochondrial proteins is pivotal to the regulation of respiratory activity in cells, and to signaling pathways leading to apoptosis, as well as for other vital mitochondrial processes. Ndufs4 has been suggested to be involved in assembly of functional complex I (Scacco et al. 2003). In addition, cyclic AMP-dependent intracellular signal transduction via phosphorylation of Ndufs4 has been reported to regulate the activity of complex I (Papa et al. 2001; Smeitink et al. 2001). The functional consequences of Ndufa10 phosphorylation await further investigation but might affect the binding affinity to NADH and in turn regulate the amount of fully active complex I (Schulenberg et al. 2003). Ndufa10 has appeared late in mitochondrial evolution and has been referred to as a “mammalian-specific” subunit of complex I (Cardol et al. 2004), consistent with a more regulatory role for this phosphorylated protein. It should also be noted that the phosphorylation sequence of Ndufa10 is likely to be a casein kinase I-like consensus motif, giving rise to the possibility that PINK1 is responsible for its phosphorylation (Schilling et al. 2005). Moreover, recent studies indicated that PINK1 and Parkin function, at least in part, in the same pathway, with PINK1 functioning upstream of Parkin, based on the observations that PINK1 and Parkin deficient Drosophila exhibit the identical phenotype with male sterility, apoptotic muscle degeneration and defects in mitochondrial morphology (Clark et al. 2006; Park et al. 2006; Yang et al. 2006). In this regard, whether Ndufs4 and Ndufa10 are substrates of PINK1 is an important issue to be clarified.

One feature common to dopaminergic neurons is the constitutive synthesis of dopamine within their cytoplasm. This is potentially important, given that metabolism of dopamine gives rise to various molecules that can act as endogenous toxins. If not properly handled, cytoplasmic dopamine might provoke neuronal damage through the generation of reactive oxygen species and, therefore, through mechanisms of oxidative stress (Shen and Cookson 2004). It has been shown that dopamine facilitates the transition of non-toxic α-synuclein protofibrils to toxic fibrils present in Lewy bodies (Lee et al. 2001; Sulzer 2001). It has also been reported that covalent modification of Parkin by dopamine lead to substantial inhibition of its E3 activity (LaVoie et al. 2005). Moreover, reduced level of VMAT2, leading to increase of cytoplasmic dopamine, is shown to result in progressive nigrostriatal neurodegeneration in mice (Caudle et al. 2007). This suggests the possibility that inappropriate metabolism of dopamine or its signaling or both might contribute to the selective degeneration of dopaminergic neurons. Notably, panneuronal expression of Pael-R in Drosophila causes age-dependent selective degeneration of dopaminergic neurons, and knockdown of parkin exacerbates this phenotype (Yang et al. 2003). Moreover, we have recently found that Alpha-methyl-p-tyrosine (AMPT), a TH inhibitor, ameliorates dopaminergic cell death induced by infection of adenovirus encoding Pael-R, implicating the pathological role of dopamine and its metabolites (Kitao et al. 2007). parkin-ko/PrP-Pael-R-tg double-mutant mice demonstrate higher levels of DA, DOPAC and HVA early in the disease process and maintain higher levels of DOPAC and HVA throughout the lifetime. Correspondingly, these mice show higher levels of protein carbonyls, a well-known marker of oxidative damage specifically in the midbrain (Fig. 9). It is recently reported that oxidative stress compensatory mechanisms are impaired in 24-month-old parkin ko mice, suggesting that they may contribute to the increase of oxidative stress at the end-stage of double mutant mice(Rodriguez-Navarro et al. 2007).

image

Figure 9.  The mechanisms underlying neuronal death in parkin-ko/PrP-Pael-R-tg double-mutant mice.

Download figure to PowerPoint

Based on these observations, both chronic ER stress and excessive dopamine-mediated oxidative stress are likely to contribute to dopaminergic neuron-specific degeneration (Fig. 9). It is an intriguing question whether Pael-R is involved in the pathogenesis of sporadic PD, since Pael-R is localized to Lewy bodies(Murakami et al. 2004). The examination of Pael-R level accumulated in post-mortem brain of patients with sporadic PD will provide important clues to this question in the future. Taken together, parkin-ko/Pael-R-tg double-mutant mice provide an excellent opportunity to dissect the molecular mechanisms underlying AR-JP as well as other degenerative diseases caused by chronic ER stress.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

This work was supported by the Grant-in aid from the Ministry of Health and Labour, Grant-in-Aid for Scientific Research on Priority Areas (Research on Pathomechanisms of Brain Disorders) from the MEXT of Japan to R.T. (18023020), Grant-in-Aid for Scientific Research to R.T. (18390255) from JSPS, Research Grant to R.T. from Takeda Science Foundation, Research Grant from RIKEN BSI to R.T., Grant-in-Aid for Young Scientists from the MEXT of Japan to H.-Q.W. (18700351).

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

Fig. S1Expression of Pael-R in distinct regions of brain. A, Immunoblot analysis of Triton X-100 soluble and insoluble fractions from cerebral cortex, midbrain and striatum of 6 months mice was performed using antibody recognizing both mouse and human Pael-R. Antibodies to transferrin receptor (TfR) and calnexin were used as loading controls for insoluble and soluble fractions, respectively.

Fig. S2Loss of TH-positive neurons in parkin-ko/PrP-Pael-R-tg double-mutant mice. A, Number of TH-positive neurons in the VTA of parkin-ko/PrP-Pael-R-tg double-mutant mice. B, Number of TH-positive neurons in the LC of parkin-ko/PrP-Pael-R-tg double-mutant mice. Data are expressed as mean ± SEM (n = 10). Statistical analysis was performed by ANOVA, followed by the Bonferoni post hoc test. *, < 0.05, **, < 0.01 vs. parkin-ko/Pael-R-non-tg.

Fig. S3 No reduction of Nissl-positive neurons in 24 months parkin-ko/PrP-Pael-R-tg double-mutant mice in the dentate gyrus region of hippocampus. Data are expressed as mean ± SEM (n = 10). Statistical analysis was performed by Anova, followed by the Bonferoni post hoc test.

Fig. S4 Specific activation of UPR in parkin-ko/PrP-Pael-R-tg mice. A, Representative images of BiP in TH-positive neurons in the SNpc. B, Representative images of CHOP in TH-positive neurons in the SNpc.

Please note: Blackwell Publishing are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

FilenameFormatSizeDescription
JNC_5607_sm_figures.doc2682KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.