Modulation of the endoplasmic reticulum-mitochondria interface in Alzheimer's disease and related models.

It is well-established that subcompartments of endoplasmic reticulum (ER) are in physical contact with the mitochondria. These lipid raft-like regions of ER are referred to as mitochondria-associated ER membranes (MAMs), and they play an important role in, for example, lipid synthesis, calcium homeostasis, and apoptotic signaling. Perturbation of MAM function has previously been suggested in Alzheimer's disease (AD) as shown in fibroblasts from AD patients and a neuroblastoma cell line containing familial presenilin-2 AD mutation. The effect of AD pathogenesis on the ER-mitochondria interplay in the brain has so far remained unknown. Here, we studied ER-mitochondria contacts in human AD brain and related AD mouse and neuronal cell models. We found uniform distribution of MAM in neurons. Phosphofurin acidic cluster sorting protein-2 and σ1 receptor, two MAM-associated proteins, were shown to be essential for neuronal survival, because siRNA knockdown resulted in degeneration. Up-regulated MAM-associated proteins were found in the AD brain and amyloid precursor protein (APP)Swe/Lon mouse model, in which up-regulation was observed before the appearance of plaques. By studying an ER-mitochondria bridging complex, inositol-1,4,5-triphosphate receptor-voltage-dependent anion channel, we revealed that nanomolar concentrations of amyloid ß-peptide increased inositol-1,4,5-triphosphate receptor and voltage-dependent anion channel protein expression and elevated the number of ER-mitochondria contact points and mitochondrial calcium concentrations. Our data suggest an important role of ER-mitochondria contacts and cross-talk in AD pathology.

Effects of a full-body electrostimulation garment application in a cohort of subjects with cerebral palsy, multiple sclerosis, and stroke on upper motor neuron syndrome symptoms.

OBJECTIVES: Dysfunction of the central nervous system may inflict spastic movement disorder (SMD). Electrical stimuli were identified as promising therapeutic option. Electrical stimulation provided by a 58-electrode full body garment was investigated based on data from regular trial fittings. METHODS: Data from 72 testees were investigated. Age averages 36.6 (19.8) ys with 44 females. The cohort spans infantile cerebral paresis (CP) (n=29), multiple sclerosis (MS) (n=23) and stroke (n=20). Data were stratified by etiology and an entry BBS Score<45. RESULTS: Effect sizes (Cohen`s d) related BBS, TUG, FGA, 10mWT, WMFT, EQ5D5L and Pain. Significance levels are indicated by *: p<0.05, **: p<0.01, ***: p<0.001, (t): p<0.1: CP: 1.64***, 0.29*, 1.59***, 0.76(t), 1.00***, 0.5*, 1.28***; MS: 1.83***, 0.83***, 1.28**, 1.07***, 0.93*, 1,11**, 0.78*; Stroke: 1.28**, 0.78**, 0.89, 0.92**, 0.71, 1.26*, 0.78*. CONCLUSIONS: Multi-site transcutaneous electrical stimulation may increase ambulation related skills in subjects with SMD stemming from CP, MS and stroke. The results indicate effects on static and dynamic balance, fall risk, mobility, upper extremity improvement and an overall increase in health utility and a reduction in spasticity related pain. Effects are immediate as well as sustained. These results may inspire individual trial fittings and inform further controlled trials.

A pan-European study of the C9orf72 repeat associated with FTLD: geographic prevalence, genomic instability, and intermediate repeats.

We assessed the geographical distribution of C9orf72 G(4) C(2) expansions in a pan-European frontotemporal lobar degeneration (FTLD) cohort (n = 1,205), ascertained by the European Early-Onset Dementia (EOD) consortium. Next, we performed a meta-analysis of our data and that of other European studies, together 2,668 patients from 15 Western European countries. The frequency of the C9orf72 expansions in Western Europe was 9.98% in overall FTLD, with 18.52% in familial, and 6.26% in sporadic FTLD patients. Outliers were Finland and Sweden with overall frequencies of respectively 29.33% and 20.73%, but also Spain with 25.49%. In contrast, prevalence in Germany was limited to 4.82%. In addition, we studied the role of intermediate repeats (7-24 repeat units), which are strongly correlated with the risk haplotype, on disease and C9orf72 expression. In vitro reporter gene expression studies demonstrated significantly decreased transcriptional activity of C9orf72 with increasing number of normal repeat units, indicating that intermediate repeats might act as predisposing alleles and in favor of the loss-of-function disease mechanism. Further, we observed a significantly increased frequency of short indels in the GC-rich low complexity sequence adjacent to the G(4) C(2) repeat in C9orf72 expansion carriers (P < 0.001) with the most common indel creating one long contiguous imperfect G(4) C(2) repeat, which is likely more prone to replication slippage and pathological expansion.

Altered enzymatic activity and allele frequency of OMI/HTRA2 in Alzheimer's disease.

The serine-protease OMI/HTRA2, required for several cellular processes, including mitochondrial function, autophagy, chaperone activity, and apoptosis, has been implicated in the pathogenesis of both Alzheimer's disease (AD) and Parkinson's disease (PD). Western blot quantification of OMI/HTRA2 in frontal cortex of patients with AD (n=10) and control subjects (n=10) in two separate materials indicated reduced processed (active, 35 kDa) OMI/HTRA2 levels, whereas unprocessed (50 kDa) enzyme levels were not significantly different between the groups. Interestingly, the specific protease activity of OMI/HTRA2 was found to be significantly increased in patients with AD (n=10) compared to matched control subjects (n=10) in frontal cortex in two separate materials. Comparison of OMI/HTRA2 mRNA levels in frontal cortex and hippocampus, two brain areas particularly affected by AD, indicated similar levels in patients with AD (n=10) and matched control subjects (n=10). In addition, we analyzed the occurrence of the OMI/HTRA2 variants A141S and G399S in Swedish case-control materials for AD and PD and found a weak association of A141S with AD, but not with PD. In conclusion, our genetic, histological, and biochemical findings give further support to an involvement of OMI/HTRA2 in the pathology of AD; however, further studies are needed to clarify the role of this gene in neurodegeneration.

Glucocerebrosidase variant T369M is not a risk factor for Parkinson's disease in Sweden.

INTRODUCTION: Genetic variants in the Beta-glucocerebrosidase gene (GBA1) is a known risk factor for Parkinson's disease. The GBA1 mutations L444P, N370S and many other have been shown to associate with the disease in populations with diverse background. Some GBA1 polymorphisms have a less pronounced effect, and their pathogenicity has been debated. We have previously found associations with L444P, N370S and E326K and Parkinson's disease in Sweden. METHOD: In this study we used pyrosequencing to genotype the T369M variant in a large Swedish cohort consisting of 1,131 patients with idiopathic Parkinson's disease, and 1,594 control subjects to evaluate the possibility of this variant conferring an increased risk for Parkinson's disease. RESULTS: The minor allele frequency was 2.15% in patients and 1.76% in controls. Statistical analysis showed that there was no significant difference in allele frequency between patients and control subjects, p-value 0.37, Odds Ratio 1.23 with a 95% confidence interval of 0.82-1.83. CONCLUSION: Our results suggest that T369M is not a risk factor for Parkinson's disease in the Swedish population.

Cerebellar alpha-synuclein levels are decreased in Parkinson's disease and do not correlate with SNCA polymorphisms associated with disease in a Swedish material.

Alterations of brain and plasma alpha-synuclein levels and SNCA gene variability have been implicated in the pathogenesis of Parkinson's disease (PD). We therefore measured alpha-synuclein protein levels in postmortem PD and control cerebellum tissue using Western blot and investigated whether the levels correlated to SNCA genotype. We found markedly decreased alpha-synuclein levels in PD patients (n=16) compared to gender- and age-matched controls (n=14; P=0.004) normalized to alpha-tubulin. We also performed an association study of the noncoding polymorphisms rs2737029 (A/G) and rs356204 (A/G) (intron 4), and of rs356219 (T/C) (3'-region) of SNCA in a Swedish PD case-control material. Using a two-sided chi(2) test, we found significant association of rs2737029 (P=0.003; chi(2)=9.07) and rs356204 (P=0.048; chi(2)=3.91) with disease, strengthening the involvement of SNCA polymorphisms in sporadic PD. Stratification of the human postmortem brain material by genotype of the three investigated polymorphisms, did not indicate any influence of genotype on alpha-synuclein protein levels when comparing PD with controls. Taken together, our findings demonstrate that the investigated Parkinson patients have markedly reduced levels of alpha-synuclein in cerebellum, and that this reduction is general, rather then correlated to the investigated polymorphisms, although two of the polymorphisms also associated with disease in a Swedish material.

Lrrk2 and alpha-synuclein are co-regulated in rodent striatum.

LRRK2, alpha-synuclein, UCH-L1 and DJ-1 are implicated in the etiology of Parkinson's disease. We show for the first time that increase in striatal alpha-synuclein levels induce increased Lrrk2 mRNA levels while Dj-1 and Uch-L1 are unchanged. We also demonstrate that a mouse strain lacking the dopamine signaling molecule DARPP-32 has significantly reduced levels of both Lrrk2 and alpha-synuclein, while mice carrying a disabling mutation of the DARPP-32 phosphorylation site T34A or lack alpha-synuclein do not show any changes. To test if striatal dopamine depletion influences Lrrk2 or alpha-synuclein expression, we used the neurotoxin 6-hydroxydopamine in rats and MitoPark mice in which there is progressive degeneration of dopamine neurons. Because striatal Lrrk2 and alpha-synuclein levels were not changed by dopamine depletion, we conclude that Lrrk2 and alpha-synuclein mRNA levels are possibly co-regulated, but they are not influenced by striatal dopamine levels.

Parkinson's disease: a genetic perspective.

Parkinson's disease (PD) is a common neurodegenerative disorder in the aging population, affecting more than 1% over the age of 65 years. Certain rare forms of the disease are monogenic, representing 5-10% of PD patients, but there is increasing evidence that multiple genetic risk factors are important also for common forms of PD. To date, 13 genetic loci, PARK1-13, have been suggested for rare forms of PD such as autosomal dominant and autosomal recessive PD. At six of these loci, genes have been identified and reported by several groups to carry mutations that are linked to affected family members. Genes in which mutations have been linked to familial PD have also been shown to be candidate genes for idiopathic forms of PD, as those same genes may also carry other mutations that merely increase the risk. Four of the PARK genes, SNCA at PARK1, UCH-L1 at PARK5, PINK1 at PARK6 and LRRK2 at PARK8, have been implicated in sporadic PD. There are indeed multiple genetic risk factors that combine in different ways to increase or decrease risk, and several of these need to be identified in order to begin unwinding the causative pathways leading to the different forms of PD. In this review, we present the molecular genetics of PD that are understood today, to help explain the pathways leading to neurodegeneration.

Expression of multi-drug resistance 1 mRNA in human and rodent tissues: reduced levels in Parkinson patients.

The membrane transporter multi-drug resistance 1 (MDR1, P-gp) regulates the bioavailability of endogenous and exogenous compounds and has been implicated in disorders such as Parkinson's disease, cancer, epilepsy, human immunodeficiency virus disease, and inflammatory bowel disease. To promote further understanding of the role of MDR1 in disease, we have characterized cellular MDR1 mRNA expression in post-mortem human and fresh-frozen Sprague-Dawley rat tissues by using radioactive oligonucleotide probe in situ hybridization. We report MDR1 mRNA in human and rat endothelial cells of small vessels in the brain and pia mater. Mdr1 mRNA is also expressed in the blood vessel walls of rat sensory dorsal root and sympathetic ganglia. In peripheral tissues, we have observed MDR1 mRNA in human and rat liver and renal tubules and in human adrenal cortex and the epithelial lining of rat intestine. In female and male reproductive tissues of rat, strong gene activity has been found in steroid-hormone-synthesizing cells. Quantification of MDR1 mRNA in human striatum has revealed reduced levels in Parkinson patients compared with control individuals. The high expression of MDR1 mRNA in blood vessels of the nervous system, in tissues involved in absorption and excretion, and in tissues forming barriers to the environment support the physiological role of MDR1 as a regulator of intracellular levels of endogenous and exogenous compounds.

High and complementary expression patterns of alcohol and aldehyde dehydrogenases in the gastrointestinal tract: implications for Parkinson's disease.

Parkinson's disease (PD) is a heterogeneous movement disorder characterized by progressive degeneration of dopamine neurons in substantia nigra. We have previously presented genetic evidence for the possible involvement of alcohol and aldehyde dehydrogenases (ADH; ALDH) by identifying genetic variants in ADH1C and ADH4 that associate with PD. The absence of the corresponding mRNA species in the brain led us to the hypothesis that one cause of PD could be defects in the defense systems against toxic aldehydes in the gastrointestinal tract. We investigated cellular expression of Adh1, Adh3, Adh4 and Aldh1 mRNA along the rodent GI tract. Using oligonucleotide in situ hybridization probes, we were able to resolve the specific distribution patterns of closely related members of the ADH family. In both mice and rats, Adh4 is transcribed in the epithelium of tongue, esophagus and stomach, whereas Adh1 was active from stomach to rectum in mice, and in duodenum, colon and rectum in rats. Adh1 and Adh4 mRNAs were present in the mouse gastric mucosa in nonoverlapping patterns, with Adh1 in the gastric glands and Adh4 in the gastric pits. Aldh1 was found in epithelial cells from tongue to jejunum in rats and from esophagus to colon in mice. Adh3 hybridization revealed low mRNA levels in all tissues investigated. The distribution and known physiological functions of the investigated ADHs and Aldh1 are compatible with a role in a defense system, protecting against alcohols, aldehydes and formaldehydes as well as being involved in retinoid metabolism.

Association study of two genetic variants in mitochondrial transcription factor A (TFAM) in Alzheimer's and Parkinson's disease.

Mitochondrial (mt) dysfunction has been implicated in Alzheimer's (AD) and Parkinson's disease (PD). Mitochondrial transcription factor A (TFAM) is needed for mtDNA maintenance, regulating mtDNA copy number and is absolutely required for transcriptional initiation at mtDNA promoters. Two genetic variants in TFAM have been reported to be associated with AD in a Caucasian case-control material collected from Germany, Switzerland and Italy. One of these variants was reported to show a tendency for association with AD in a pooled Scottish and Swedish case-control material and the other variant was reported to be associated with AD in a recent meta-analysis. We investigated these two genetic variants, rs1937 and rs2306604, in an AD and a PD case-control material, both from Sweden and found significant genotypic as well as allelic association to marker rs2306604 in the AD case-control material (P=0.05 and P=0.03, respectively), where the A-allele appears to increase risk for developing AD. No association was observed for marker rs1937. We did not find any association in the PD case-control material for either of the two markers. The distribution of the two-locus haplotype frequencies (based on rs1937 and rs2306604) did not differ significantly between affected individuals and controls in the two sample sets. However, the global P-value for haplotypic association testing indicated borderline association in the AD sample set. Our data suggests that the rs2306604 A-allele could be a moderate risk factor for AD, which is supported by the recent meta-analysis.

DJ-1 and UCH-L1 gene activity patterns in the brains of controls, Parkinson and schizophrenia patients and in rodents.

DJ-1 (PARK7) has been implicated in early onset and familial cases of Parkinson's disease (PD). We therefore mapped cellular activity patterns of the DJ-1 gene in human and rodent brain tissue with radioactive in-situ hybridization. In all three mammals mRNA expression was restricted mainly to neurons in all regions analyzed. White matter, such as crus cerebri and capsula interna appeared negative, suggesting that glial cells express DJ-1 at levels below the detection limit of our method. We compared DJ-1 mRNA expression to the neuronal marker UCH-L1, which has also been implicated in PD, and found lower levels for DJ-1 but very similar patterns of expression. Measurement of the signal intensity revealed that human frontal cortex of control cases expressed DJ-1 mRNA more abundantly than other regions such as substantia nigra in the midbrain. Comparing DJ-1 expression in dopamine neurons on hemi-sections from controls and patients we could not detect any difference between 14 controls, 8 idiopathic Parkinson and 5 schizophrenia cases. Of note, DJ-1 is expressed in several other tissues such as the liver, gastrointestinal tract, adrenal and pituitary gland and during embryonic development, while UCH-L1 has a strictly neuronal expression also outside the CNS. We conclude that DJ-1 and UCH-L1, like other genes linked to PD, are not expressed specifically in DA neurons, but instead generally in neurons. The abundant expression of DJ-1 in certain peripheral tissues and of UCH-L1 in peripheral neurons may also be of relevance for the spectrum of symptoms in different forms of PD.

S18Y in ubiquitin carboxy-terminal hydrolase L1 (UCH-L1) associated with decreased risk of Parkinson's disease in Sweden.

Ubiquitin carboxy-terminal hydrolase L1 (UCH-L1) is a neuron-specific enzyme that removes ubiquitin from the C-terminal end of substrates and a component of the ubiquitin-proteasome system. A protective effect of a UCH-L1 variant, S18Y, was suggested since the common variant was found to be inversely associated with sporadic Parkinson's disease (PD). We investigated the association of S18Y in our Swedish PD material. The tyrosine variant was significantly inversely associated with PD (P=0.049) and with a low age of onset (50 years) (P=0.017) in the case-control material, supporting the hypothesis of a protective function.

Genetic Variations and mRNA Expression of NRF2 in Parkinson's Disease.

Nuclear factor erythroid 2-like 2 (NRF2) encodes a transcription factor regulating mechanisms of cellular protection and is activated by oxidative stress. NRF2 has therefore been hypothesized to confer protection against Parkinson's disease and so far an NRF2 haplotype has been reported to decrease the risk of developing disease and delay disease onset. Also NRF2 adopts a nuclear localization in Parkinson's disease, which is indicative of increased NRF2 activity. We have investigated the association between NRF2 and Parkinson's disease in a Swedish case-control material and whether NRF2 expression levels correlate with NRF2 genetic variants, disease, or disease onset. Using pyrosequencing, we genotyped one intronic and three promoter variants in 504 patients and 509 control subjects from Stockholm. Further, we quantified NRF2 mRNA expression in EBV transfected human lymphocytes from patients and controls using quantitative real-time reverse transcription PCR. We found that one of the promoter variants, rs35652124, was associated with age of disease onset (Χ2 = 14.19, p value = 0.0067). NRF2 mRNA expression levels however did not correlate with the rs35652124 genotype, Parkinson's disease, or age of onset in our material. More detailed studies on NRF2 are needed in order to elucidate how this gene affects pathophysiology of Parkinson's disease.

Strong association between glucocerebrosidase mutations and Parkinson's disease in Sweden.

Several genetic studies have demonstrated an association between mutations in glucocerebrosidase (GBA), originally implicated in Gaucher's disease, and an increased risk of Parkinson's disease (PD). We have investigated the possible involvement of genetic GBA variations in PD in the Swedish population. Three GBA variants, E326K, N370S, and L444P were screened in the largest Swedish Parkinson cohort reported to date; 1625 cases and 2025 control individuals. We found a significant association with high effect size of the rare variant L444P with PD (odds ratio 8.17; 95% confidence interval: 2.51-26.23; p-value = 0.0020) and a significant association of the common variant E326K (odds ratio 1.60; 95% confidence interval: 1.16-2.22; p-value = 0.026). The rare variant N370S showed a trend for association. Most L444P carriers (68%) were found to reside in northern Sweden, which is consistent with a higher prevalence of Gaucher's disease in this part of the country. Our findings support the role of GBA mutations as risk factors for PD and point to lysosomal dysfunction as a mechanism contributing to PD etiology.

Novel progranulin mutations with reduced serum-progranulin levels in frontotemporal lobar degeneration.

Frontotemporal lobar degeneration (FTLD) is a progressive neurodegenerative disease with an age at onset generally below 65 years. Mutations in progranulin (GRN) have been reported to be able to cause FTLD through haploinsufficiency. We have sequenced GRN in 121 patients with FTLD and detected six different mutations in eight patients: p.Gly35Glufs*19, p.Asn118Phefs*4, p.Val200Glyfs*18, p.Tyr294*, p.Cys404* and p.Cys416Leufs*30. Serum was available for five of the mutations, where the serum-GRN levels were found to be >50% reduced compared with FTLD patients without GRN mutations. Moreover, the p.Cys416Leufs*30 mutation segregated in an affected family with different dementia diagnoses. The mutation frequency of GRN mutation was 6.6% in our FTLD cohort.

Genetic Screening of the Mitochondrial Rho GTPases MIRO1 and MIRO2 in Parkinson's Disease.

MIRO1 and MIRO2 (mitochondrial Ras homolog gene family, member T1 and T2) also referred to as RHOT1 and RHOT2, belong to the mitochondrial Rho GTPase family and are involved in axonal transport of mitochondria in neurons. Because mitochondrial dysfunction is strongly implicated in Parkinson's disease (PD), MIRO1 and MIRO2 can be considered as new candidate genes for PD. We analyzed two non-synonymous polymorphisms and one synonymous polymorphism in MIRO1 and two non-synonymous polymorphisms in MIRO2, in a Swedish Parkinson case-control material consisting of 241 patients and 307 neurologically healthy controls. None of the analyzed polymorphisms in MIRO1 and MIRO2 were significantly associated with PD. Although we did not find a significant association with PD in our Swedish case-control material, we cannot exclude these Rho GTPases as candidate genes for PD or other neurodegenerative disorders.

Association of a protective paraoxonase 1 (PON1) polymorphism in Parkinson's disease.

Pesticide exposure has been suggested to increase the risk to develop Parkinson's disease (PD). The arylesterase paraoxonase 1 (PON1) is mainly expressed in the liver and hydrolyzes organophosphates such as pesticides. The polymorphism Leu54Met (rs854560) in PON1, impairing enzyme activity and leading to decreased PON1 expression levels, has been reported to be associated with Parkinson's disease (PD). PON1 is part of a cluster on chromosome 7q21.3 together with PON2 and PON3. We investigated the occurrence of four additional polymorphisms in PON1 and two in PON2 in a Swedish PD case-control material. We found a significant association (p=0.007) with a PON1 promoter polymorphism, rs854571. The minor allele was more common among controls than PD cases which suggest a protective effect. This is strengthened by the fact that rs854571 is in strong linkage disequilibrium with another PON1 promoter polymorphism, rs854572, reported to increase PON1 gene expression. Our findings support the hypothesis that PON1 is involved in the etiology of PD and that higher PON1 levels are reducing the risk for PD.

Adh1 and Adh1/4 knockout mice as possible rodent models for presymptomatic Parkinson's disease.

Alcohol dehydrogenases (ADH) catalyze the reversible metabolism of many types of alcohols and aldehydes to prevent the possible toxic accumulation of these compounds. ADHs are of interest in Parkinson's disease (PD) since these compounds can be harmful to dopamine (DA) neurons. Genetic variants in ADH1C and ADH4 have been found to associate with PD and lack of Adh4 gene activity in a mouse model has recently been reported to induce changes in the DA system. Adh1 knockout (Adh1-/-) and Adh1/4 double knockout (Adh1/4-/-) mice were investigated for possible changes in DA system related activity, biochemical parameters and olfactory function compared to wild-type (WT) mice. Locomotor activity was tested at ∼7 (adult) and >15 months of age to mimic the late onset of PD. Adh1-/- and Adh1/4-/- mice displayed a significantly higher spontaneous locomotor activity than WT littermates. Both apomorphine and d-amphetamine increased total distance activity in Adh1-/- mice at both age intervals and in Adh1/4-/- mice at 7 months of age compared to WT mice. No significant changes were found regarding olfactory function, however biochemical data showed decreased 3,4-dihydroxyphenylacetic acid (DOPAC)/DA ratios in the olfactory bulb and decreased homovanillic acid (HVA)/DA ratios in the olfactory bulb, frontal cortex and striatum of Adh1/4-/- mice compared to WT mice. Our results suggest that lack of Adh1 alone or Adh1 and Adh4 together lead to changes in DA system related behavior, and that these knockout mice might be possible rodent models to study presymptomatic PD.

Genetic studies of the protein kinase AKT1 in Parkinson's disease.

The protein kinase AKT1 belongs to the Akt family and is a potent mediator of cell growth and survival and fully activated when phosphorylated. The AKT family has been found to be phosphorylated to a lesser extent in the dopaminergic cells of Parkinson's disease patients compared to control individuals, which might influence cell survival. Several publications support the implication of AKT1 in disorders of the dopaminergic system including bipolar disease and schizophrenia. In 2008 an association study performed in a Greek Parkinson's disease case-control material reported the identification of a protective AKT1 haplotype. Based on their work we have performed a replication study in a Swedish Parkinson's disease cohort. We genotyped the four single nucleotide polymorphims (SNPs): rs2494743, rs2498788, rs2494746 and rs1130214 in a case-control material consisting of 243 Parkinson patients and 315 controls. We did not find any associations with Parkinson's disease for either the individual SNPs or any of the haplotypes. In contrast to previously published results, our data do not support the hypothesis of genetic variants in AKT1 confering protection against Parkinson's disease.