Parkinson Disease 2, Autosomal Recessive Juvenile

A number sign (#) is used with this entry because of evidence that autosomal recessive juvenile Parkinson disease-2 (PARK2, or PDJ) is caused by homozygous or compound heterozygous mutation in the parkin gene (PARK2; 602544) on chromosome 6q26.

For a phenotypic description and a discussion of genetic heterogeneity of Parkinson disease, see PD (168600).

Clinical Features

An autosomal recessive form of familial juvenile parkinsonism, defined as onset before age 40 years, was described in a Japanese family by Takahashi et al. (1994). Juvenile-onset Parkinson disease is symptomatically different in several aspects from classic late-onset Parkinson disease (PD; 168600), although classic symptoms of PD, such as bradykinesia, rigidity, and tremor, are present. Takahashi et al. (1994) commented that the familial occurrence in Japanese PDJ patients with cases was approximately 40 to 50% and that the inheritance pattern appeared to be mostly autosomal recessive. They reported a family in which 4 of 5 sibs were affected and the parents were first cousins. A full pathologic examination of 1 of the sibs, a 67-year-old woman, was presented. The substantia nigra (SN) showed obvious neuronal loss and gliosis in the medial and ventrolateral regions. In the remainder of that region and in the locus ceruleus, the population of neurons was reduced and there was low melanin content in most of the neurons but no detectable gliosis or extraneuronal free melanin pigment suggestive of a neurodegenerative process. There were no Lewy bodies. The entire pathologic picture was different from that of Lewy body Parkinson disease (168601). This patient had been well until about the age of 10 years when gait disturbance appeared. By age 14, she was unable to walk long distances. By her forties, she was unable to walk without assistance. There was no evidence of dementia. She had been slow-moving and had shown frozen gait and tremor, more evident on motion, in the head and upper and lower limbs. She showed improvement of the movement disorder after waking up in the morning. When she was young, the improvement lasted until evening, but as she aged it became progressively shorter, eventually lasting only about 10 minutes. In the other sibs, gait disturbance began at the age of 8 or 9 years. One sister had died at age 42 years and a brother at the age of 27 years, both in a bedridden state.

Ishikawa and Tsuji (1996) described the clinical features of 17 patients from 12 Japanese families with familial juvenile parkinsonism. In 11 of these families affected members were products of consanguineous matings. All 12 families resided within the same geographic area, raising the possibility of founder effect. The mean age of onset was 27 years, with a range from 9 to 43. The most prominent symptoms were retropulsion, dystonia of the feet, and hyperreflexia with classic parkinsonism. Symptoms of tremor, rigidity, and bradykinesia were mild. Patients responded to levodopa but dopa-induced dyskinesias and wearing-off phenomena occurred frequently.

Bonifati et al. (1994) reported a man who presented with Parkinson disease at age 28. He was born of a consanguineous mating between a man who developed Parkinson disease at age 74 and his first cousin, who apparently was not affected with parkinsonian symptoms; however, the maternal grandfather developed Parkinson disease at age 65. Bonifati et al. (1994) speculated that the proband may have been homozygous for a defective Parkinson disease gene, which in heterozygous form gave rise to late-onset Parkinson disease in the affected father and maternal grandfather.

Mitsui et al. (1994) described a sister and a brother, the only offspring of a first-cousin mating, with atypical juvenile parkinsonism. They presented at age 38 and 40, respectively, with bradykinesia, cogwheel rigidity, and mild pyramidal and cerebellar signs. They were unresponsive to levodopa but responded very well to trihexyphenidyl, an anticholinergic drug. Mitsui et al. (1994) proposed that this was a new hereditary variant of early-onset Parkinson disease, distinct from the levodopa-sensitive form of juvenile Parkinson disease. They commented that responsiveness to anticholinergics but not levodopa may also be seen in Joseph disease (SCA3; 109150). MRI studies demonstrated atrophy of the cerebellar hemisphere and vermis, as well as high intensity areas in both pyramidal tracts. This juxtaposition of extrapyramidal and cerebellar signs usually results in classification of a disease as a multisystem atrophy or an olivopontocerebellar degeneration and is characteristic of many adult-onset spinocerebellar atrophies, most of which are transmitted as autosomal dominant traits.

Yamamura et al. (1973) reported familial cases of juvenile parkinsonism with marked diurnal fluctuation in symptoms. The disorder was found to show nigral cell loss and occurred in a setting of inbreeding; it undoubtedly represented a subtype of autosomal recessive juvenile parkinsonism (Nygaard, 1993). Matsumine et al. (1998) noted that early-onset parkinsonism with diurnal fluctuation (EPDF) is also a dopa-responsive form of parkinsonism and is associated with selective degeneration in the zona compacta of the substantia nigra without Lewy body formation. A distinguishing feature of this phenotype is a benefit from sleep with lessening of parkinsonian symptoms after awakening.

Portman et al. (2001) performed PET scans on 2 brothers with early-onset parkinsonism caused by mutations in the parkin gene and found marked reduction of fluorodopa (FDOPA) uptake in the caudate and putamen. Portman et al. (2001) noted that this was a different nigrostriatal dopaminergic pattern than that found in sporadic PD, thus suggesting a different pathophysiology for the early-onset disease.

Ohsawa et al. (2005) found that 8 of 9 patients with PARK2 mutations had significantly reduced sural sensory nerve action potential (SNAP) amplitude compared to 8 patients with idiopathic Parkinson disease. However, 6 PARK2 carriers had absence of decreased vibration sense in the foot, and only 2 had subjective sensory symptoms. Two patients with a presumptive diagnosis of idiopathic PD who showed a reduced SNAP amplitude were subsequently diagnosed with PARK2 as a result of DNA analysis. Ohsawa et al. (2005) suggested that reduced SNAP amplitude in patients with PD under 60 years of age may be a diagnostic indicator of PARK2 mutations, and concluded that sensory axonal neuropathy may be a common feature of the disorder.

Pathologic Findings

Mori et al. (1998) reported neuropathologic findings in a patient with PARK2 confirmed by genetic analysis. The patient had disease onset at age 24 years and died from unrelated complications at age 62. Grossly, the substantia nigra showed marked depigmentation. Melanin-containing neurons in the pars compacta were moderately decreased, and neuronal loss and gliosis were especially marked in the ventrolateral and medial regions. Most of the remaining nigral neurons contained melanin pigments. The pars reticulata was spared. Lewy bodies were not identified. Neurofibrillary tangles and argyrophilic astrocytes were identified in the SN, the locus ceruleus, posterior hypothalamus, the hippocampus, and various cortical areas. However, the pattern was not consistent with Alzheimer disease. Tyrosine hydroxylase (191290) activity was markedly reduced in the caudate and putamen, and modestly reduced in the SN.

Hayashi et al. (2000) reported neuropathologic findings in a Japanese patient with a mutation in the PARK2 gene (602544.0002) who had previously been reported by Ishikawa and Miyatake (1995). Loss of pigmented neurons and gliosis were most pronounced in the medial and ventrolateral regions of the substantia nigra pars compacta and in the locus ceruleus. Remaining neurons had low amounts of melanin. There was mild neuronal loss and gliosis in the substantia nigra pars reticulata. No Lewy bodies were identified. Some neurofibrillary tangles and senile plaques were observed in the cerebral cortex, although there was no clinical evidence of dementia.

Van de Warrenburg et al. (2001) reported a Dutch family with PARK2 confirmed by genetic analysis. Neuropathologic examination of the proband showed depigmentation of the substantia nigra, with severe loss of pigmented neurons in the pars compacta, deposition of extraneuronal melanin, and mild gliosis. No Lewy bodies or neurofibrillary tangles were seen. However, there was a diffuse spread of tau (MAPT; 157140)-positive thorn-shaped astrocytes in the caudate, putamen, and subthalamic nucleus, and a few tau-positive astrocytes in the SN.

Farrer et al. (2001) reported a patient who was compound heterozygous for 2 mutations in the PARK2 gene. At autopsy, Lewy body pathology typical of idiopathic Parkinson disease was found, which was noted to be unusual for this form of parkinsonism.

Sasaki et al. (2004) reported neuropathologic examination of a patient with Parkinson disease due to homozygous exon 3 deletion in the PARK2 gene (602544.0005). The patient had disease onset at age 33 years and died of respiratory failure at age 70. The substantia nigra showed marked depigmentation, and melanin-containing neurons of the pars compacta were moderately to severely depleted, particularly in the ventrolateral group and medial part. Some dopaminergic neurons remained, but most were atrophic, and free melanin was observed. The pars reticulata of the SN was spared. In the locus ceruleus, neurons were mildly decreased, and free melanin was seen. Lewy bodies were not observed in the SN or locus ceruleus. There were alpha-synuclein-positive and ubiquitin-positive, round or donut-shaped inclusions in the neuropils of the pedunculopontine nucleus, but no such inclusions were seen in the SN, locus ceruleus, or subthalamic nucleus. The inclusions were somewhat basophilic, distinguishing them from Lewy bodies, which show an eosinophilic tint. No immunoreactivity to phosphorylated tau was seen in any region of the brain. Sasaki et al. (2004) suggested that a functioning parkin protein may be required for Lewy body formation.

Other Features

Olfactory impairment is 1 of the earliest manifestations of Parkinson disease and reflects Lewy body infiltration in the olfactory bulb and tract before the disease affects the substantia nigra. Alcalay et al. (2011) tested olfactory function in 44 probands with early-onset PD, including 9 with 2 parkin mutations, 10 with 1 parkin mutation, and 25 with PD without parkin mutations, as well as 80 of their family members without PD, including 18 with 1 parkin mutation, 2 with 2 parkin mutations, and 60 without parkin mutations. Among those with PD, patients with 2 parkin mutations had significantly better olfaction than patients with 1 parkin mutation or PD patients with no parkin mutation. Among those without PD, olfaction was similar in carriers of 1 parkin mutation and those without mutations, but better than PD patients with 1 parkin mutation. Alcalay et al. (2011) suggested that PD patients with parkin mutations have better olfaction compared to other PD patients, and also suggested that heterozygosity for 1 parkin mutation does not confer an increased risk of neuropathology.

Inheritance

Recessive homozygous mutations in the parkin gene result in early-onset Parkinson disease. Heterozygous mutations in the parkin gene have been identified in some patients with later onset disease, raising the possibility that heterozygous mutations may confer increased susceptibility to the disease. Wang et al. (2008) identified mutations in the parkin gene in 25 (10.1%) of 247 probands with disease onset before age 50. Eighteen patients had heterozygous mutations. Using a kin-cohort study design, which does not require information about mutation status, they estimated the cumulative incidence of PD to age 65 in relatives of mutation carriers to be 7.0%, compared to 1.7% in noncarrier relatives and 1.1% in relatives of controls. Wang et al. (2008) noted the limitations to their study, including the inability to assess risk of PD among heterozygotes separately and the low age of some relatives.

Mapping

By linkage analysis using a diallelic polymorphism of the manganese superoxide dismutase gene (SOD2; 147460), Matsumine et al. (1997) found perfect segregation of the disease with the SOD2 locus. By extending the linkage analysis to 13 families with autosomal recessive juvenile parkinsonism, they discovered strong evidence for the localization of the gene at 6q25.2-q27, including the SOD2 locus, with the maximum cumulative pairwise lod scores of 7.26 and 7.71 at D6S305 (theta = 0.03) and D6S253 (theta = 0.02), respectively. Nucleotide sequence analysis of the coding regions of the SOD2 gene did not show causative mutations, suggesting that another, as yet unidentified gene in this region is responsible for autosomal recessive juvenile parkinsonism.

In a group of 15 families from 4 distinct ethnic backgrounds, Jones et al. (1998) found that the locus for autosomal recessive juvenile parkinsonism in all mapped to 6q25.2-q27. A full genomic screen excluded other candidate regions. They constructed a detailed genetic map of the linked region and mapped the position of the SOD2 gene. Recombination events restricted the chromosome 6 Parkinson disease locus to a 6.9-cM region and excluded SOD2. Tassin et al. (1998) likewise found linkage of the gene for PDJ to 6q25.2-q27 in 1 Algerian and 10 European multiplex families. They found the clinical spectrum of the disease in these families, with age at onset up to 58 years and the presence of painful dystonia in some patients, to be broader than that reported previously. In all patients examined, 2 of the 3 cardinal signs of PD (akinesia, rigidity, and tremor) were found. Marked improvement with levodopa treatment occurred in all except 2 untreated secondary cases found in family studies. Age at onset was less than 40 years for at least 1 affected sib.

Saito et al. (1998) narrowed the assignment of the PARK2 locus to a 4-cM region encompassing D6S1579 and D6S1599.

Matsumine et al. (1998) demonstrated that the disorder reported by Yamamura et al. (1973) (EPDF) showed linkage to the same region on chromosome 6 as did autosomal recessive juvenile Parkinson disease. In studies of 17 families with EPDF, they found a peak lod score of 14.2 at 1.0 cM telomeric to D6S305 and placed the disease locus in the 17-cM interval between D6S437 and D6S253, which is exactly the same position to which PDJ had been mapped.

Molecular Genetics

In several patients with PDJ, Kitada et al. (1998) identified deletions in the PARK2 gene (see, e.g., 602544.0001).

Hoenicka et al. (2002) found 5 different mutations in the PARK2 gene in 5 of 13 Spanish families with recessive inheritance. Two of the mutations were novel (602544.0013 and 602544.0012). Hoenicka et al. (2002) found 2 simple heterozygous PARK2 mutation carriers who developed clinical symptoms, either in late adulthood or after brief exposure to parkinsonizing agents. The authors suggested that heterozygosity may be a risk factor for PD. In 14 other Spanish kindreds with familial PD, 8 autosomal recessive, 4 autosomal dominant and 2 of uncertain inheritance, Hoenicka et al. (2002) found no mutations in the alpha-synuclein (SNCA; 163890) or UCHL1 (191342) genes related to PARK1 (168601) and PARK5 (191342), respectively.

Lucking et al. (2001) described an Italian family in which parkinsonism was associated with mutations in the PARK2 gene in a pseudodominant pattern of inheritance. The father (with disease onset at age 57 years) was homozygous for a triplication of exon 2, a previously undescribed mutation. The unaffected mother was heterozygous for deletions of exons 3 and 4, and the son (with disease onset at age 31 years) was a compound heterozygote carrying both mutations.

In 10 affected members of a consanguineous Brazilian family with early-onset parkinsonism, Chien et al. (2006) identified a homozygous splice site mutation in the PARK2 gene (602544.0020). The family was from an isolated region in northeastern Brazil, and their ancestors had originated from Portugal. One individual who was heterozygous for the mutation developed neuroleptic-induced parkinsonism, suggesting that haploinsufficiency was a predisposing factor.

Kay et al. (2007) found that heterozygous parkin mutations were as common in 301 controls as in 302 PD patients, and they replicated the finding in an independent set of 1,260 PD patients and 1,657 controls. Thirty-four variants, including 21 novel variants, were identified. Kay et al. (2007) concluded that heterozygous mutations in the parkin gene are not likely to contribute to the development of Parkinson disease. Quantitative gene dosage was not examined.

Alcalay et al. (2010) identified mutations in the PARK2 gene in 64 (6.7%) of 953 patients with early-onset PD before age 51, including 77 and 139 individuals of Hispanic and Jewish ancestry, respectively. Thirty-three patients had a heterozygous PARK2 mutation, and 27 had homozygous or compound heterozygous mutations. Four of 64 patients had a PARK2 mutation and another mutation in the LRRK2 (609007) or GBA (606463) genes. Hispanic individuals were more likely to be PARK2 mutation carriers than non-Hispanic individuals (15.6% vs 5.9%; p = 0.003). Marder et al. (2010) examined the same dataset reported by Alcalay et al. (2010), with specific focus on PARK2 mutation carriers. Of the 64 patients with parkin mutations, 37 were heterozygous (including 4 with mutations in other genes), 6 homozygous, and 21 compound heterozygous. Compound heterozygous and homozygous carriers had a significantly younger age at onset compared to heterozygous carriers. Significant associations were found between parkin mutation carrier status and Hispanic origin (odds ratio (OR) of 2.7), younger age at onset (less than 40 years) (OR of 5.0), and family history of PD in a first-degree relative (OR of 2.8). Deletions in exons 3 and 4 and 255delA (602544.0014) were common among Hispanics, specifically Puerto Ricans. In addition, those with heterozygous parkin mutations had younger age at onset compared to PD patients without parkin mutations, suggesting that heterozygosity is a susceptibility factor for disease development.

Kay et al. (2010) identified heterozygous copy number variations (CNV) in the parkin gene in 0.95% of 1,686 controls and 0.86% of 2,091 PD patients, suggesting that heterozygous PARK2 CNV mutations are not associated with PD risk.

Associations Pending Confirmation

For discussion of a possible association between autosomal recessive juvenile-onset Parkinson disease and variation in the PODXL gene, see 602632.0002.

Genotype/Phenotype Correlations

Foroud et al. (2003) identified 25 different parkin mutations in 103 affected individuals from 47 families with PD, including 41 individuals with mutations in both alleles and 62 individuals with a single mutation in only 1 allele. Individuals with 2 parkin mutations had an earlier age at disease onset and longer disease duration than those with 1 mutation. Thirty-five subjects (35%) with a parkin mutation had an age at onset of 60 years or above, with 30 of these 35 having only 1 mutant allele. The authors concluded that mutations in the parkin gene occur among individuals with PD with an older age at onset (greater than 60 years) who have a positive family history of the disease.

In 16 of 307 (5%) families with PD, Oliveira et al. (2003) identified mutations in the parkin gene, which included 18% of all early-onset and 2% of all late-onset families. Three families were homozygous, 3 families were compound heterozygous, and in 10 families, all the patients had heterozygous mutations. The results showed that mutations in exon 7 were observed primarily in heterozygous PD patients with a later age at onset. Oliveira et al. (2003) concluded that mutations in the parkin gene contribute to the common form of PD, and that heterozygous mutations act as susceptibility alleles for the late-onset form of PD.

Lohmann et al. (2003) compared 146 PD patients with parkin mutations to 250 PD patients without parkin mutations; they found that patients with the parkin mutations had a significantly earlier and more symmetric onset, a slower progression of disease, and a tendency toward greater response to L-DOPA despite lower doses. However, both groups had a similar wide range for age at onset (7 to 70 years and 12 to 76 years, respectively). In families with autosomal recessive parkinsonism, more than 80% of patients with an age at onset of 20 years or younger had parkin mutations, compared to 28% of those between the ages of 46 and 55 years. Carriers of at least 1 parkin missense mutation had a more severe phenotype than those with 2 truncating mutations, suggesting that missense mutations result in more than a loss of function. Patients with a single heterozygous parkin mutation had significantly later and more asymmetric onset and more frequent L-DOPA-induced difficulties than those with 2 parkin mutations.

Poorkaj et al. (2004) undertook a study to determine whether patients with early-onset PD should be screened for parkin mutations as part of their clinical workup. Patients with a diagnosis of PD and onset at or before 40 years of age were selected for genotyping by sequence and dosage analysis for all 12 exons. Mutations were found in 7 of 39 patients. Two of these were compound heterozygous; 5 were heterozygous. Early-onset PD accounted for 10% of PD patients, and 18% of the early-onset patients had parkin mutations. Assuming a strictly recessive inheritance, only 5% of early-onset cases had a pathogenic parkin genotype. The remaining 13% were heterozygous, and whether heterozygous parkin mutations were the cause of early-onset PD in these patients was unclear.

Using PET scan, Khan et al. (2005) found that 13 asymptomatic heterozygous carriers of a PARK2 mutation had significantly decreased fluorodopa uptake in the caudate, putamen, and ventral and dorsal midbrain compared to controls. Four of the heterozygous carriers had subtle extrapyramidal signs. Khan et al. (2005) concluded that parkin heterozygosity is a risk factor for nigrostriatal dysfunction and suggested that parkin heterozygosity may contribute to late-onset PD.

Klein et al. (2000) reported a large kindred from a remote village in the Western Alps of South Tyrol in northern Italy affected with adult-onset Parkinson disease inherited in an autosomal dominant pattern. The clinical features were indistinguishable from idiopathic Parkinson disease, and none of the patients demonstrated typical features of PARK2, such as diurnal fluctuation, sleep benefit, foot dystonia, hyperreflexia, or early susceptibility to levodopa-induced dyskinesias. Haplotype analysis implicated the parkin locus on chromosome 6q. Molecular analysis showed that 4 affected male sibs were compound heterozygous for 2 deletions in the PARK2 gene: a large deletion, which was later determined by Hedrich et al. (2001) to be a deletion of only exon 7 (602544.0010), and a 1-bp deletion in exon 9 (602544.0019). Two affected females were heterozygous for the 1-bp deletion. Klein et al. (2000) concluded that the phenotypic spectrum associated with mutations in the PARK2 gene are broad and that PARK2 may play a role in the etiology of late-onset typical PD. Pramstaller et al. (2005) provided detailed clinical and molecular follow-up of the family reported by Klein et al. (2000). Ancestors could be traced back to the year 1657; relevant clinical data were obtained from 196 individuals spanning 7 generations. The mean age at onset was 52.8 years, but ranged from 20 to 76 years. Five of 25 definitely affected individuals were found to be compound heterozygous for the 2 previously identified PARK2 deletions; 8 patients had only 1 of these deletions; the mutational status of 5 deceased patients was unknown; and 7 patients had no PARK2 mutations. Patients who were compound heterozygous had earlier onset than those with heterozygous mutations. Pramstaller et al. (2005) concluded that heterozygous mutations in the PARK2 gene contribute to idiopathic PD. Postmortem analysis of 1 of the patients reported by Pramstaller et al. (2005) with both PARK2 mutations showed SNCA-positive Lewy bodies.

Pathogenesis

Kitada et al. (1998) suggested that parkin may function similarly to ubiquitin family members, and its defect in PDJ may interfere with the ubiquitin-mediated proteolytic pathway leading to the death of nigral neurons.

Shimura et al. (2001) hypothesized that alpha-synuclein and parkin interact functionally, namely, that parkin ubiquitinates alpha-synuclein normally and that this process is altered in autosomal recessive Parkinson disease. Shimura et al. (2001) identified a protein complex in normal human brain that includes parkin as the E3 ubiquitin ligase, UBCH7 (603721) as its associated E2 ubiquitin-conjugating enzyme, and a novel 22-kD glycosylated form of alpha-synuclein (alpha-Sp22) as its substrate. In contrast to normal parkin, mutant parkin associated with autosomal recessive Parkinson disease failed to bind alpha-Sp22. In an in vitro ubiquitination assay, alpha-Sp22 was modified by normal, but not mutant, parkin into polyubiquitinated, high molecular weight species. Accordingly, alpha-Sp22 accumulated in a nonubiquitinated form in parkin-deficient Parkinson disease brains. Shimura et al. (2001) concluded that alpha-Sp22 is a substrate for parkin's ubiquitin ligase activity in normal human brain and that loss of parkin function causes pathologic accumulation of alpha-Sp22. These findings demonstrated a critical biochemical reaction between the 2 Parkinson disease-linked gene products and suggested that this reaction underlies the accumulation of ubiquitinated alpha-synuclein in conventional Parkinson disease.

Mortiboys et al. (2008) found that fibroblasts derived from PD patients with biallelic mutations in the PARK2 gene had significantly decreased mitochondrial complex I activity and ATP production compared to controls. Patient fibroblasts also showed altered morphology, including a greater degree of mitochondrial branching, as well as increased susceptibility to mitochondrial toxins. Complete knockdown of parkin using siRNA in control fibroblasts confirmed that the effects were due to parkin deficiency. In contrast, 50% knockdown of parkin, mimicking haploinsufficiency in humans, did not result in impaired mitochondrial function or morphology. Treatment with experimental neuroprotective glutathione replacement compounds resulted in restoration of the mitochondrial membrane potential.