α-Synuclein (PARK1)

PARK1 involves mutations in the α-synuclein gene.^20,21 The first families with an α-synuclein mutation originated from southern Italy and Greece, and with the description of several additional families with the same A53T mutation, it is believed that they have a common founder. Age at onset was young (mean, 46 years) with a low frequency of tremor and relatively rapid disease progression, although symptoms were levodopa responsive.²² Pathological analysis of the few brains available revealed the presence of Lewy bodies in the substantia nigra and locus ceruleus.²³ Linkage to markers on chromosome 4q21-q23 was demonstrated and the locus designated PARK1.²⁴ Further analyses identified a mutation in exon 4 of the gene encoding α-synuclein resulting in an alanine-to-threonine substitution at codon 53 (A53T), which was also found in affected members of three Greek families with earlyonset autosomal dominant PD.²⁰ A second mutation (A30P) was later found in a small German family with PD,²¹ but extensive study in large groups of PD families and sporadic cases has not identified other patients or families with this mutation.^25–28

α-Synuclein is a protein of 140 amino acids that is predominantly expressed in neurons and is one of the most common brain proteins. Its normal function remains unclear, although it plays a role in synaptic plasticity based on song-learning studies in the zebra finch²⁹ and vesicular regulation of dopamine release from knockout mice.³⁰ A key observation linking α-synuclein to PD was the demonstration that it is one of the principal components of Lewy bodies.³¹ Lewy bodies are intracytoplasmic aggregates comprising several proteins, including ubiquitin and α-synuclein, and have supported the notion that abnormal protein handling might be important in PD pathogenesis. Furthermore, mutant isoforms of α-synuclein more readily oligomerize and it has been suggested that its tendency to aggregate into misfolded structures may confer toxic properties to the protein. Indeed, overexpression of wild-type or mutant protein in transgenic mice or Drosophila reproduces many of the behavioral and pathological features of PD.^32,33

Multiplications of the wild-type α-synuclein gene have been described in PD families. A triplication of the gene was identified in a large autosomal dominant kindred with PD and tremor,³⁴ and duplication of the gene was found in one of 42 familial probands of early onset PD.³⁵ A third α-synuclein point mutation (E46K) has been reported in an autosomal dominant family with parkinsonism and Lewy body dementia.³⁶

Several models of abnormal α-synuclein expression have been developed. Knockout of the gene in mice resulted in no detectable abnormality other than an alteration of dopamine release in response to rapid stimulation, although this has no clear functional correlate.³⁰ Overexpression of wild-type human α-synuclein in mice resulted in loss of dopaminergic terminals, intranuclear and cytoplasmic ubiquitin-rich nonfibrillar α-synuclein inclusions in the substantia nigra, hippocampus, cortex, and a rotor-rod motor deficit at 1 year.³² Overexpression of human wild-type and mutant α-synuclein in flies caused a loss of dopaminergic neurons, Lewy body–like inclusions with fibrillar α-synuclein, and a motor deficit with no significant difference between wild-type and mutant α-synuclein.³³ Additional mouse models of α-synuclein expression have demonstrated inclusion formation and spinal cord pathology but no dopaminergic cell loss, and motor deficit at late stage.³⁷ Virus-mediated overexpression of α-synuclein induces nigral degeneration in rodents.³⁸

The α-synuclein mutations result in the protofibrillar form, which is considered the more toxic form of the protein. The A53T and A30P α-synuclein mutations promote protofibril formation and A30P inhibits conversion to fibrils.³⁹ Catecholamines, including dopamine and levodopa, inhibit fibril formation in vitro, and this is reversed by antioxidants; that is, catechol oxidation promotes protofibril formation.⁴⁰ This observation would support a protective role for Lewy bodies in PD. An important observation revealing a potential toxic mechanism for α-synuclein is that the mutant A30P form increases toxicity to dopamine, increasing cell death and free radical–mediated damage.⁴¹ The authors proposed that the mutation impaired vesicular uptake of dopamine, resulting in higher cytoplasmic or extravesicular synaptic concentrations of dopamine that would in turn cause free radical–mediated damage. Phosphorylation at the Ser129 residue is required to mediate the toxicity of α-synuclein and increased the formation of inclusions in SHSY-5Y cells.⁴² This phosphorylated form of α-synuclein is present in Lewy bodies.⁴³ Prevention of this phosphorylation by substitution of an alanine residue reduced inclusion formation in the SHSY-5Y model, and in the Drosophila model this same mutation at 129 that prevents phosphorylation protected against dopaminergic neuronal loss.⁴⁴

Parkin (PARK2)

PARK2 gene mutations were first identified in autosomal recessive juvenile-onset parkinsonism (ARJPD).⁴⁵ ARJPD has been most commonly seen in the Japanese population and is characterized by onset before age 40 years, symptomatic improvement following sleep, mild dystonia, and a good response to levodopa.⁴⁶ Resting tremor is seen less frequently than in idiopathic PD, and patients may have brisk tendon reflexes but no other pyramidal, cerebellar, or autonomic features. The disease is often symmetrical and dyskinesias develop early but progression is usually slow. Pathologically, there is dopaminergic cell loss in the SNc and locus ceruleus but no Lewy bodies are seen.⁴⁷ The gene responsible for ARJPD was mapped to 6q25.2-q27,⁴⁸ and in 1998 the gene was discovered and named parkin.⁴⁵ Affected patients carry deletions or point mutations in various parts of the parkin gene.^49,50 The absence of Lewy bodies in ARJPD may simply reflect the limited time over which the pathology has evolved. However, the relationship of parkin mutations to idiopathic PD has been highlighted by the identification of parkin mutations in apparently sporadic cases of PD and by the description of Lewy bodies in parkin positive patients with later-onset disease than ARJPD.^51,52 Parkin mutations are a common cause of PD under age 25 years but rare over age 40 years.^53,54 Parkin-related PD has been reported in multiple generations in families without consanguinity, suggesting a pseudo-autosomal dominant mode of inheritance for some mutations.^55,56 Fluorodopa positron emission tomography (PET) in parkin patients demonstrates reduced uptake in the striatum, although there is some discordance regarding the symmetry and pattern of this reduction^57,58 However, the rate of loss of fluorodopa PET signal was slower in the parkin patients than in sporadic PD.⁵⁹ Asymptomatic heterozygous parkin mutation carriers had intermediate levels of striatal fluorodopa PET uptake compared with normal controls and homozygous symptomatic patients.⁵⁸ This suggests an intermediate stage of nigrostriatal dysfunction that may interact with other genetic or environmental factors to induce PD. The frequency of heterozygous parkin mutation carriers is not known.

PARK2 encodes parkin, which functions as an E3 ligase, ubiquinating proteins for destruction by the proteosome.^60,61 Several substrates for parkin have been identified, including a 22-kDa glycosylated form of α-synuclein, parkin-associated endothelin receptor-like receptor (Pael-R), and CDCrel-1. Overexpression of Pael-R causes it to become ubiquinated, insoluble, and unfolded and leads to endoplasmic reticulum stress and cell death.⁶² It has been demonstrated to accumulate in its insoluble form in the brains of patients with parkin mutations, suggesting a possible toxic mechanism. CDCrel-1 is a protein involved in cytokinesis and may influence synaptic vesicle function.⁶¹ Overexpression of parkin protected against dopaminergic loss in rodents coexpressing α-synuclein, suggesting a protective role for parkin.⁶³

A parkin knockout mouse model has been described.⁶⁴ This showed an increase in striatal extracellular dopamine, a reduction in synaptic excitability, and a mild nonprogressive motor deficit at 2 to 4 months. There was no loss of dopaminergic neurons and no inclusion formation. Dopamine receptor binding affinities and parkin E3 ligase substrate levels were normal. Interestingly, these mice had decreased striatal mitochondrial respiratory chain function and reductions in specific respiratory chain and antioxidant proteins.⁶⁵ Parkin knockout flies developed muscle pathology, mitochondrial abnormalities and apoptotic cell death.⁶⁶ Overexpression of parkin in PC12 cells indicated that it is associated with the mitochondrial outer membrane.⁶⁷ Parkin-positive patients have decreased lymphocyte complex I activity.⁶⁸ The ability of parkin to ubiquinate proteins may be impaired by S-nitrosylation, which in turn may be a consequence of excitotoxicity-mediated damage.⁶⁹

UCH-L1 (PARK5)

A further mutation in the gene encoding ubiquitin carboxyhydrolase (UCH)-L1 again supported the relevance of the ubiquitin-proteosomal system (UPS) in PD pathogenesis.⁷⁰ UCH-LI is an enzyme that hydrolyzes the C-terminus of ubiquitin to generate ubiquitin monomers that can be recycled to clear other proteins. A missense mutation was identified in two siblings with typical PD in a German family demonstrating apparent autosomal dominant inheritance.⁷⁰ Age at onset was 49 years in one and 51 years in the other. The mutant form of UCH-L1 was shown to have reduced enzyme activity resulting in impaired protein clearance through the ubiquitin-proteasome pathway. However, no other mutations in this gene have been identified in other families, suggesting it is a rare cause of PD.^71,72 Given that no further cases of PD have been described with mutations in this gene, some doubt has been cast on the relevance of UCH-L1 to PD.

PINK1 (PARK6)

The PARK6 locus (chromosome 1p36⁷³) was first identified in a large consanguineous Italian family and subsequently in an additional three Italian families and others from Europe and elsewhere, including Asia.^74–78 The mean age at onset ranges from 21 to 57 years. Progression is usually slow and patients exhibit a good response to levodopa. PARK6 mutations appear to be a rare cause of PD.

The PINK1 (PTEN-induced kinase 1) gene is ubiquitously transcribed and is believed to encode a mitochondrial kinase.^74,79 It is believed that PINK1 may play a role in protecting cells against stress conditions that affect mitochondrial membrane potential, but the downstream targets through which PINK1 mediates its protection have not been identified. As 11 of the 14 reported mutations fall into the kinase domain of PINK1,^74,76,77 altered phosphorylation of target proteins probably represents a key pathogenic mechanism, leading to abnormal stress response and neurodegeneration. The reversible phosphorylation of proteins is an important method of regulating cellular activities.⁸⁰ Up to 30% of eukaryotic proteins are phosphorylated,⁸¹ and there are more than 500 human genes encoding protein kinases.⁸² The phosphorylation of mitochondrial proteins is considered pivotal to the regulation of respiratory activity in the cell and to signaling pathways leading to apoptosis, as well as for other vital mitochondrial processes. For instance the phosphorylation of α-synuclein is an important step in mediating its toxicity (see earlier), and Lewy bodies do contain the phosphorylated form of this protein.

DJ-1 (PARK7)

The PARK7 locus on chromosome 1p36, only about 25 cM from the PARK6 locus, was first identified in a small group of young-onset PD patients in a remote region of Holland.⁸³ Average age at onset is 32 years, with a currently reported range of 25 to 40 years. Onset is asymmetrical, progress is slow, and there is a good response to levodopa. Tremor is infrequent, and psychiatric disturbances have been described in some. Fluorodopa PET scans demonstrate a symmetrical reduction in uptake. No pathological studies of PARK7 patients have been undertaken at the time of writing.

PARK7 encodes DJ-1; mutations are autosomal recessive and comprise both deletions and point mutations that result in a loss or inactivation of the protein. Its function is unknown, but it is widely distributed and conserved. It can protect against toxicity mediated by free radicals and transfers to the outer mitochondrial membrane under conditions of oxidative stress.^84,85 Wild-type DJ-1 is also located in the mitochondrial matrix and intermembraneous space, and this distribution is not altered by mutations in the protein.⁸⁶

LRRK2 (PARK8)

Mutations in the LRRK2 gene are the most common cause of either familial or “sporadic” PD identified to date. The LRRK2 G2019S mutation alone has been reported in 2.8% to 6.6% of autosomal dominant PD families^87–89 and in 2% to 8% of sporadic cases.^90–92 The G2019S mutation has variable penetrance, with 17% at 50 years and 85% at 70 years, a profile that mimics idiopathic, sporadic PD. Although other LRRK2 mutations are described, the G2019S mutation remains the most common cause of either sporadic or familial PD. This mutation has not been seen in Alzheimer’s disease or in parkinsonian syndromes other than idiopathic PD.^93,94

Many of the reported cases of LRRK2 mutations have typical features of PD with asymmetrical onset of tremor, bradykinesia, and rigidity. As noted, the age at onset is variable with occasional very late onset cases⁸⁹ and a report of one carrier male reaching 89 years with only subtle neurological changes.⁹⁵ Patients have a good response to levodopa but develop motor complications including dyskinesias. Fluorodopa PET and imaging using ligands for the dopamine transporter with single-photon emission computed tomography (SPECT) demonstrate changes typical of those seen in idiopathic PD.⁹⁶ Although all LRRK2 mutant brains examined to date demonstrate loss of dopaminergic neurons in the SNc, one of the morphological hallmarks of idiopathic PD, additional pathology may also be seen. Pure nigral neuronal degeneration was found in the first family linked to this locus⁹⁷; neurofibrillary tangles, abnormal tau deposits, and widespread Lewy body synucleinopathy have been described in others, including one family with anterior horn cell loss.^98,99 Three brains of “sporadic” PD with G2019S LRRK2 mutations have had pathological examination and all have demonstrated nigral neuronal loss and Lewy body formation typical of PD.⁹⁰ All of these subjects had PD based on clinical criteria.

The LRRK2 gene encodes a 286-kDa cytoplasmic protein that is widely expressed in the brain.¹⁰⁰ LRRK2 is a member of the ROCO protein family and appears to have multiple functions, at least by virtue of its predicted structure. These include a Ras/GTPase domain involved in cytoskeletal responses to external stimuli, vesicular trafficking and the stimulation of stress-activated kinase.¹⁰¹ The leucine-rich motif may have numerous functions, including protein-protein interactions and substrate binding for ubiquitination. The LRRK2 kinase domain belongs to the MAPKKK family of kinases with catalytic activity for both serine/threonine and tyrosine residues.

The G2019S mutation changes a highly conserved glycine at the start of the kinase activation segment, and it has been postulated that this has an activating effect causing a “gain of function” compatible with its autosomal dominant inheritance pattern.⁸⁹ LRRK2 also has a WD40 domain, which again may be involved in cytoskeletal assembly and signal transduction.

PARK9, PARK10, and PARK11

The PARK9 locus on chromosome 1p36 was described in an autosomal recessive, juvenile-onset parkinsonian disorder with pyramidal features, ophthalmoplegia, and dementia.^102,103 PARK10 on chromosome 1p32 was identified in the families of Icelandic PD patients with late-onset disease.¹⁰⁴ PARK11 was obtained by association studies, and little information is available on phenotype.

Genetic Associations

Only a minority of cases of PD are part of a clear familial pedigree. Some of the single-gene mutations described above may account for a proportion of the remaining patients. However, our current understanding is that such single-gene causes of PD will remain in the minority. Thus, the large proportion of PD patients may develop their disease as a result of environmental factors, polygenic influences, or a combination of the two. There have been several genetic association studies attempting to determine significant polymorphisms that may increase or decrease the risk for PD. Further evidence for the role of genes in PD comes from genome-wide screens.^105,106 The first found 174 families with a minimum of two clinically affected individuals with PD per family and identified a marker from the intronic region of parkin in an earlyonset group and a region on 9q in a dopa-resistant group. In the total sample, areas of interest were found on chromosomes 5q, 8p, and 17q. Data from the second genome-wide linkage study used a sample of 113 affected sibling pairs with PD and identified suggestive linkage on chromosomes 1, 9, 10, and 16, with no evidence implicating the regions containing parkin, α-synuclein, or tau genes. However, additional studies have shown that α-synuclein promoter region variants can influence the risk for PD.^107,108 Those alleles that increase α-synuclein expression lead to an increased risk for developing PD, an observation in line with the multiplications of the gene causing familial PD (see earlier). Similarly, variants that influence parkin expression can also modulate the risk for PD: in this case, those alleles that lower parkin expression enhance the risk for PD.¹⁰⁹ Mutations in the gene for glucocerebrosidase have been associated with an increased risk for PD among Ashkenazi Jews.¹¹⁰ Glucocerebrosidase deficiency causes type 1 Gaucher’s disease. Thirty-one of 99 Ashkenazi PD patients had one or two mutant alleles of the glucocerebrosidase gene (GBA), and 95 of 1543 controls (Ashkenazi) were carriers. Those with GBA mutations had onset of their PD around age 60 years and clinical features typical of idiopathic disease.

Genetic Causes of Parkinsonism

Several disorders that include parkinsonism in their phenotype have been characterized at the genetic level (Table 71-2). The frontotemporal dementias are discussed in detail in Chapter 73, the dystonias in Chapter 35, Huntington’s disease in Chapter 67, Wilson’s disease in Chapter 108 and the inherited ataxias in Chapter 68, so these will be discussed only briefly here.

TABLE 71-2 Secondary Familial Parkinsonism

The frontotemporal dementias and parkinsonism linked to chromosome 17 (FTDP-17) usually have onset in the fifth decade, although the age range is wide (25 to 75 years). Symptoms are of gradual onset and include motor dysfunction in the forms of parkinsonism, behavioral, and personality disorders and cognitive decline.¹¹¹ Patients may exhibit apathy, depression, aggression, disinhibition, obsessive-compulsive disorder, executive dysfunction, and nonfluent aphasia. Patients and families tend to fall into either the predominantly parkinsonian or dementia types. Pathological examination shows severe frontotemporal atrophy and degeneration that includes the substantia nigra and the basal ganglia. There are tau accumulations in the remaining neurons and glia. FTDP-17 is autosomal dominant and is due to mutations in the tau gene. Although the genotype-phenotype relationship is relatively loose, those with the parkinsonism predominant form more commonly have exon 10′ or 5′ mutations.

Certain of the spinocerebellar ataxias (SCAs) are associated with parkinsonism and indeed may even manifest with this feature. SCA2 dopa-responsive parkinsonism is most often observed in the Chinese Asian population.^112,113 Patients may have asymmetrical disease; a resting tremor and the presence of ataxia and abnormal eye movements may make differentiation from other parkinsonian disorders difficult if genetic testing is not performed. Imaging with fluorodopa PET has produced variable results from changes typical of those seen in PD¹¹⁴ to severe involvement of the caudate.¹¹⁵ The demonstration of an abnormally expanded CAG repeat in the ataxin-2 gene confirms SCA2. SCA3 (Machado-Joseph disease) mutations in conjunction with parkinsonism have been found most often in Caribbean populations.

Fragile X mental retardation complex is a common cause of mental retardation. It is an X-linked disease caused by an abnormal CGG expansion in the FMRI gene, which results in reduced gene expression. Intermediate length repeats can be a cause of tremor-ataxia parkinsonism in men. About 60% of these patients have a postural tremor, ataxia, autonomic dysfunction, impaired cognition, and symmetrical parkinsonism.^116,117

Patients with very large expansions of CAG repeats in the huntingtin gene can present with juvenile-onset Huntington’s disease, known as the Westphal variant. The predominant features are those of bradykinesia and rigidity with little, if any, response to levodopa.

Dystonia in association with parkinsonism is seen in a number of genetic diseases. X-linked dystonia parkinsonism was first reported in men from an island in the Philippines who had early onset of action tremor, dystonia, blepharospasm, and parkinsonism in 40% (Lubag) with poor response to levodopa. This disease is referred to as DYT3.¹¹⁸ Rapid-onset dystonia parkinsonism (DYT12) is an autosomal dominant disorder associated with bulbar features including dysarthria and dysphagia, dystonia, postural instability, and bradykinesia. Symptoms progress rapidly over hours and may be precipitated by physical or emotional stress. There is usually a poor response to levodopa. Mutations in the ATP1A3 gene that encodes a subunit of the sodium-potassium channel have been described in this disorder.¹¹⁹

Mitochondria have their own DNA, and the mitochondrial genome encodes 13 proteins of the oxidative phosphorylation system in addition to 2 ribosomal and 22 transfer RNAs. The discovery of complex I deficiency in PD substantia nigra (see later) raised the possibility that the mutation of genes (nuclear or mitochondrial) encoding complex I subunits might be involved in determining the enzyme’s defective activity. As mitochondrial DNA (mtDNA) is inherited in a strictly maternal pattern, if there were full penetrance of such a mtDNA gene defect, mitochondrial inheritance should be identifiable in pedigrees with parkinsonism. Such maternal inheritance has been described in PD¹²⁰ but appears rare. However, it is known that 40% of patients with proven mitochondrial diseases and mtDNA mutations present as sporadic cases. Thus maternal inheritance is not a sine qua non of mtDNA gene defects. However, molecular genetic investigations of mtDNA have so far been unable to identify any specific mutation that clearly co-segregates with PD.

Studies using age-matched controls found no increase in the 5-kilobase mitochondrial deletion mutation in PD substantia nigra.¹²¹ Several studies have sequenced mtDNA in PD, but these have all used unselected patients in terms of their complex I activity.^122,123 Although some reports have suggested an increased frequency of certain mtDNA polymorphisms in PD, this has not been replicated in all studies.^124–128 Two studies have demonstrated a relationship between mtDNA haplotypes and the risk for developing PD. The first showed a reduced risk for PD in individuals with haplotypes J and K,¹²⁹ and the second, a 22% decrease in PD in those with the UKJT haplotype cluster.¹³⁰ In contrast, a smaller study reported an increased risk for PD with haplotypes J and T.¹³¹

Mutations in the gene for mtDNA polymerase gamma (POLG) have been demonstrated in patients with progressive external ophthalmoplegia and parkinsonism. Autosomal dominant or recessive inheritance of progressive external ophthalmoplegia with age at onset ranging from 10 to 54 years was followed some years later (range, 6 to 40 years) by the development of an asymmetrical, levodopa-responsive bradykinetic rigid syndrome together with resting tremor in some patients. Additional features included variable limb, pharyngeal or facial weakness, cataracts, ataxia, peripheral neuropathy, and premature ovarian failure.¹³² Muscle biopsy demonstrated ragged red, cytochrome oxidase–negative fibers in all patients with multiple mtDNA deletions on Southern blotting. Symmetrically reduced striatal 18-fluorodopa PET was seen in two patients. Brain histology was available on an additional two patients; both showed severe loss of substantia nigral dopaminergic neurons but without the development of Lewy bodies or other synuclein aggregates. Four families had the same A2864G mutation inherited in autosomal fashion in three and with a founder effect in the fourth. Mutations in the exonuclease or polymerase portions of the gene were identified in the autosomal recessive families. Another patient with autosomal dominant progressive external ophthalmoplegia parkinsonism and an A2492G mutation has been reported.¹³³

Oxidative Stress and Damage

There are several lines of evidence that suggest increased oxidative stress and oxidative damage to biomolecules in PD substantia nigra (Table 71-3).

TABLE 71-3 Oxidative Damage in Parkinson’s Disease

The free radical gas nitric oxide (NO^•) is present in many tissues, including the central nervous system. NO^• is generated by the conversion of L-arginine to L-citrulline by nitric oxide synthase (NOS). At least three NOS isoforms are recognized and are all expressed within the brain. NO^• acts as an atypical molecular messenger but at higher concentrations may have a toxic role and be implicated in the neurodegeneration that occurs in PD. As a free radical, NO^• could potentially contribute to dopaminergic neuronal death by mechanisms such as increased lipid peroxidation, release of iron(II), and damage to DNA. It can also inhibit a number of enzymes such as cytochrome c oxidase and superoxide dismutase and affects mitochondrial function by inhibiting complexes II, III, and IV. Animal studies have implicated NO^• in nigrostriatal neuronal loss. In addition to its possible neuroprotective effect with regard to 1-methyl-4 phenylpyridinium (MPP⁺) toxicity, 7-nitroindazole (7-NI) also protects in the methamphetamine animal model of PD.¹⁸⁶ NOS activity is at its highest in the nigrostriatal system in nonhuman primates and humans. Attempts to demonstrate altered levels of NO^• in the brains of PD patients have been inconclusive with both decreased and increased levels of cerebrospinal fluid nitrate, a marker of NOS activity, being seen.^187–189

Mitochondrial Dysfunction

The first link between mitochondria and PD was made in 1989 when a defect in the activity of respiratory chain protein complex I was identified in substantia nigra from patients with PD.^190,191 This study has been expanded over the years, and results to date show that there is a specific defect of approximately 35% complex I deficiency in PD nigra.¹⁷⁷ This defect in complex I activity in PD brain does not affect any other part of the respiratory chain. In addition, no defect in mitochondrial activity has been identifiable in any other part of PD brain, including the caudate putamen, globus pallidus, tegmentum, cortex, cerebellum, and substantia innominata.¹⁹²

Following the report of complex I deficiency in PD substantia nigra, respiratory chain abnormalities were described in skeletal muscle mitochondria from PD patients. This particular area has proved very contentious, with several groups either describing similar defects or no abnormality whatsoever (see Schapira¹⁹³ for review). Two magnetic resonance spectroscopy studies on skeletal muscle mitochondrial function in PD have shown conflicting results.^194,195 Finally, mitochondrial complex I deficiency was also identified in platelet mitochondria of PD patients.^196–198 In contrast to skeletal muscle, there is a consensus among several laboratories that complex I deficiency does exist in PD platelet mitochondria. The majority of studies, however, suggest that this deficiency, as least based on a group-to-group analysis, is modest (about 20% to 25%) (see Schapira¹⁹³ for review). The complex I deficiency in PD lacks the sensitivity to allow its use as a biomarker of PD.

The discovery of complex I deficiency in PD and the role of mitochondria in PD has been enhanced by the subsequent identification of mutations in genes encoding mitochondrial proteins, such as PINK1 and DJ1, as causes of autosomal recessive PD, and by the mitochondrial abnormalities associated with α-synuclein and parkin expression (see earlier). Furthermore, the environmental toxins causing parkinsonism identified so far are all mitochondrial inhibitors of complex I, such as MPTP, rotenone, and annonacin.