Clinical Spectrum of Homozygous and Heterozygous PINK1 Mutations in a Large German Family With Parkinson Disease:  Role of a Single Hit? FREE

During the past few years, a significant genetic component has been demonstrated in Parkinson disease (PD), especially in cases with early onset (age at onset, <50 years). To date, 5 genes have clearly been associated with PD, including the gene for PTEN-induced putative kinase 1 (PINK1), causing a recessively inherited form.¹

The first reports described homozygous or compound heterozygous missense and nonsense PINK1 mutations in families with early-onset PD previously linked to the PARK6 gene from Europe^1- 2 and Japan,^3- 4 demonstrating a worldwide distribution of PINK1-associated early-onset PD. Although the mode of inheritance was autosomal recessive, ie, 2 mutations have to come together, a role for single heterozygous PINK1 mutations in the etiology of PD had been suggested even before the identification of the gene in a PARK6-linked family.⁵ Meanwhile, several patients with single heterozygous mutations have been described,^6- 7 and more patients carry a single mutation than predicted for a recessively acting gene.⁷

The PINK1-associated phenotype was initially considered to resemble idiopathic PD with a slightly earlier age at onset.^6,8 However, atypical cases have also been reported with dystonia at onset, sleep benefit, and even dementia.^7,9- 10

To address the role of single heterozygous PINK1 mutations, we investigated 19 first- and second-degree relatives of an identified homozygous mutation carrier.

After obtaining informed consent, a homozygous PINK1 mutation carrier and 19 of his family members (family W; Figure) underwent a detailed neurological examination by one of us (C.K.) and an additional member of the movement disorder team at the University of Lübeck (J.H. or A.H.). The latter 2 examiners were blinded as to mutational status and pedigree structure. All family members were videotaped using the Unified Parkinson's Disease Rating Scale (UPDRS) part III protocol, and all 20 videotapes were evaluated in a blinded random fashion by an independent movement disorder specialist (A.M.). The rater was instructed to look for and, if present, describe signs of PD. Because rigidity cannot accurately be judged on the basis of videotape review, a consensus diagnosis was established when information on the presence and degree of rigidity was shared between physical examiners and the videotape rater.

The diagnosis of definite PD was based on the United Kingdom Brain Bank diagnostic criteria, with the exception that positive family history was not regarded as an exclusion criterion.¹¹ Signs of probable PD were defined as the combination of bradykinesia with 1 additional cardinal PD sign and an unknown response to levodopa. For this, bradykinesia or the additional cardinal sign had to be at least mild (corresponding to a score of 2 on UPDRS part III).¹² Signs of possible PD included isolated mild bradykinesia or a combination of slight bradykinesia (score of 1 on UPDRS part III)¹² and slight rigidity. All individuals with parkinsonian signs on neurological examination were considered affected, and family members without any signs were considered unaffected. Subjects who were conscious of their signs of PD were classified as symptomatic, and individuals unaware of their signs were classified as asymptomatic. Parkinson disease was rated using the UPDRS part III and Hoehn and Yahr Scale; major cognitive dysfunction was excluded by applying the Mini-Mental State Examination (cutoff, 27 points).

Operational psychiatric lifetime diagnoses as provided by Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition¹³ were established by 2 experienced psychiatrists (S.S. and R.L.) using the German version of the Structured Clinical Interview for Axis I Disorders and for Personality Disorders.¹⁴

The study protocol was approved by the local ethics committee of the University of Lübeck.

The index patient of family W was screened for PINK1 mutations as part of a larger study as described previously.¹⁵ Family members underwent direct sequence analysis to test for the familial mutation. Mutations in the parkin (PARK2) gene had previously been excluded in individuals II:3 and II:9.

The index patient of family W carried a homozygous nonsense mutation in exon 7 of the PINK1 gene (c.1366C>T; p.Q456X). Test results in 4 of his 5 siblings and 15 of their children disclosed an additional 14 mutation carriers in family W. These included 3 definitely affected homozygous individuals, 2 heterozygous individuals with signs of probable PD, 4 heterozygous individuals with signs of possible PD, and 5 unaffected heterozygous family members (Figure and Table 1). Clinical details of all mutation carriers are summarized in Table 2 and Table 3. Age at onset ranged from 39 to 61 years in the 4 homozygous carriers. All of them had a benign clinical course and responded extremely well to small doses of levodopa without developing motor fluctuations after 6 to 24 years, and mild dyskinesias developed only in the 2 cases with the longest treatment duration (Table 3). Additional features in mutation carriers included psychiatric disturbances (n = 10), myalgia (n = 2), tics (n = 1), and restless legs syndrome (n = 1) (Table 3). None of the heterozygous mutation carriers complained of their signs of slight and/or mild PD, including bradykinesia and rigidity. Therefore, these signs had not attracted medical attention before our detailed examination. However, when specifically asked, several of them admitted to nonspecific soft symptoms, such as smaller or slower handwriting or neck and shoulder stiffness. Parkinsonian signs were more marked on the right side in all clinically affected (symptomatic and asymptomatic) mutation carriers. Individual III:2 was the only family member who had taken neuroleptic medication (short-term treatment with olanzapine), and he also received lithium therapy. Results of testing for the mutation in the remaining 5 relatives were negative (Figure). One brother of the index patient who did not have the familial PINK1 mutation (II:9) had a mild symmetric hypokinetic syndrome with bilateral bradykinesia and fatiguing, mild rigidity bilaterally in his arms and legs, and postural instability on backward-pull test results. He had a history of severe head trauma, followed by several weeks of unconsciousness but no other neurological deficits. Apart from 1 heterozygous case with signs of possible PD, there was no disagreement regarding the clinical status between the 2 on-site examiners and the blinded rater of the videotapes (Table 1).

Magnetic resonance images of the brain in all 4 definitely affected patients were normal. In all clinically affected PINK1 mutation carriers, ¹⁸F-labeled fluorodopa (¹⁸F-dopa) positron emission tomography (PET) disclosed a remarkably reduced presynaptic radiotracer binding in the putamen, but also a slightly decreased uptake in the caudate nucleus. Findings of striatal dopamine D₂ receptor PET with ¹¹C-labeled raclopride were normal.

Examination of large families with PD, eg, the Contursi family with an α-synuclein mutation¹⁶ or family LA with parkin mutations,^17- 18 has yielded considerable insight in the role of the respective genetic cause in clinical disease expression. In this study, we report the clinical and genetic findings in 20 family members of a pedigree with homozygous and heterozygous PINK1 mutations.

Overall, our family confirms that the PINK1-associated phenotype is variable, ranging from typical idiopathic PD-like cases to those with additional features including psychiatric abnormalities (Table 2 and Table 3), as reported previously.^3,7,10

The clinical phenotype in family W is compatible with benign idiopathic PD with a mean age at onset of 50 years, slow disease progression, excellent response to treatment, and minimal motor complications. Clinical signs were present only or were clearly more pronounced on the right side in all affected family members, suggesting that laterality of PD signs might also be genetically determined. Overall, signs of PD in heterozygous cases were milder than in homozygous cases, but their current mean age (42 years) was almost a decade younger than the average age at onset of their homozygous parents. The UPDRS part III scores ranged from 27 to 35 points (off state) in the definitely affected homozygous patients; they were 7 and 8 points in the 2 heterozygous individuals with signs of probable PD, and they ranged from 3 to 7 points in the 4 family members with signs of possible PD. Individual III:2 (with the highest score of 7 among the heterozygous carriers with signs of possible PD) scored 2 points for mild, bilateral postural hand tremor that is likely an adverse effect of his lithium therapy. Although UPDRS part III scores overall were low in the heterozygous carriers, they reflect unequivocal clinical signs that are not obtained in healthy individuals, such as unilaterally reduced or absent arm swing with flexion of the arm at the wrist and elbow and unilaterally delayed shoulder shrug. Further stressing the notion that the observed signs are indeed pathologic and not due to mere physiological clumsiness in the nondominant hand, all signs were exclusively or predominantly present on the dominant right-hand side in all affected individuals. It has recently been shown that the side of asymmetry of parkinsonian features is associated with hand dominance, at least in left-handed individuals.¹⁹

The mutation found in family W (c.1366C>T) has previously been identified in 4 Italian patients in the homozygous and heterozygous state^7,20 but not in any of 460 control chromosomes,⁷ supporting its pathogenic function.

It is conceivable that this nonsense mutation leads to a protein truncation, in which 80% of the protein remains intact, and therefore could retain considerable function. In keeping with this hypothesis, the disease onset in the homozygous mutation carriers in family W was rather late, and 2 of the 4 symptomatic patients had a disease onset even later than 50 years (late-onset PD). In contrast, the other described cases with the same mutation had a rather severe phenotype with an early onset.^7,20 This might be explained by the complete absence of the mutated transcript leading to a lack of mutated PINK1 protein, at least in the families described by Bonifati and colleagues.⁷ In our family W, such a phenomenon could not be demonstrated, in keeping with the milder phenotype.²¹ Psychiatric disturbances were reported in 3 of the 4 patients with this mutation^7,20 and were also present in 10 of the 15 mutation carriers described in the present study. It may be speculated that psychiatric complications are linked to this specific PINK1 mutation.

Surprisingly, 1 of the 5 mutation-negative family members also presented with mild, probable PD. However, he differed from the mutation carriers in several ways. His signs were symmetric and nonprogressive, and his gait was normal with bilateral physiological arm swing. One may speculate that his hypokinetic syndrome may represent a phenocopy, possibly related to his severe head injury.

Despite extensive mutational analysis, in a considerable number of index patients from other studies, only a single heterozygous PINK1 mutation was reported.^{6- 7,15,22- 23} However, it has frequently been argued that a second mutation may have been overlooked in such patients, which would be extremely unlikely in the offspring of the 4 homozygous mutation carriers in our family W. In this family, we identified 6 heterozygous mutation carriers with signs of possible or probable PD, who were all children of homozygous cases and therefore younger than the affected homozygous patients.

It has previously been argued that heterozygous PINK1 mutations may not play a role in the etiology of PD because they have also been detected in healthy controls. However, the frequency of heterozygous nonsynonymous alterations in the coding region in controls (7/470 [1.5%]) is considerably lower than that in patients with PD (15/523 [2.9%]; Table 4).^6- 7,23

The clinical findings in our family W suggest that heterozygous mutations may lead to a milder phenotype with a slower progression. Indeed, such (mild) signs and symptoms in an aging individual may never come to medical attention. Furthermore, it is conceivable that at least some of the reported mutation-positive controls had not reached the age of symptom onset at the time of inclusion in the study. Supporting this idea, extrapyramidal signs that did not fulfill the diagnostic criteria of PD were found in 2 heterozygous parkin mutation carriers in their ninth decade of life who belong to a large Brazilian pedigree of affected members with homozygous parkin mutations.²⁴ These signs were only observed at a follow-up visit 10 years after the initial examination, when they had still been free of signs.

Mild signs may even have gone undiagnosed in individuals used as controls, because controls for genetic studies often do not undergo the same rigorous specialist examination for movement disorders as the patients do, and no clinical follow-up on PINK1 mutation–bearing controls has been reported as yet. The overall mild clinical signs of PD were not immediately obvious in any of our heterozygous carriers in family W, nor in 2 of the 4 homozygous mutation carriers in their on state without a specific motor examination or without at least paying special attention to their presence. The description by Lees²⁵ of the mild but unequivocal early signs of PD in Ray Kennedy is an instructive example of the slow evolution of parkinsonian features, still compatible with this man's distinguished career as a soccer player and developing over a 14-year period before the diagnosis of PD was eventually made.

In addition to this genetic and clinical evidence of a possible pathologic role for heterozygous PINK1 mutations, shortly after the identification of the first PARK6-linked family, haploinsufficiency or a dominant negative effect of the protein was suggested and based on the detection of dopamine hypometabolism in asymptomatic, heterozygous carriers of the disease-associated haplotype.⁵ Moreover, the same report described a significantly greater ¹⁸F-dopa uptake loss in the head of the caudate and in the putamen in clinically affected PARK6 subjects compared with patients with idiopathic PD. This apparently more uniform striatal dopaminergic dysfunction was also confirmed in 4 homozygous and clinically moderately affected PARK6 subjects of family W who underwent ¹⁸F-dopa PET scan (individuals II:1, II:3, II:5, and II:7). Unfortunately, PET examinations of the heterozygous mutation carriers to verify the impact of single PINK1 mutations on dopamine metabolism could not be performed.

Taking these findings together, we favor the hypothesis that heterozygous PINK1 mutations are associated with markedly reduced penetrance and show a highly variable expressivity. The latter may range from slight slowness of fine finger movements to full-blown PD, depending on the presence or absence of certain genetic, epigenetic, and/or environmental factors promoting or compensating for a latent motor deficit. In this context, we consider heterozygous PINK1 mutations a susceptibility factor for PD, which would only be associated with a higher probability, rather than a certainty, of developing signs of PD in later life.

In conclusion, this study shows that even a heterozygous PINK1 mutation may be associated with signs of PD. If the pathogenic role of a single mutation can be further established, this will have major implications on genetic counseling of the carriers of these much more frequent heterozygous PINK1 mutations. Our family W with a large number of homozygous and heterozygous carriers will provide an excellent opportunity to further investigate this important issue.

Correspondence: Christine Klein, MD, Department of Neurology, University of Lübeck, Ratzeburger Allee 160, 23538 Lübeck, Germany (christine.klein@neuro.uni-luebeck.de).

Accepted for Publication: November 1, 2005.

Author Contributions: Drs Hedrich and Hagenah contributed equally to this study. Study concept and design: Hedrich, Hagenah, Münchau, and Klein. Acquisition of data: Hedrich, Hagenah, Djarmati, Hiller, Lohnau, Lasek, Grünewald, Hilker, Steinlechner, Boston, Kock, Schneider-Gold, Kress, Siebner, Binkofski, Lencer, Münchau, and Klein. Analysis and interpretation of data: Hedrich, Hagenah, Djarmati, Hilker, Siebner, Binkofski, Lencer, Münchau, and Klein. Drafting of the manuscript: Hedrich, Hagenah, Lencer, Münchau, and Klein. Critical revision of the manuscript for important intellectual content: Hedrich, Djarmati, Hilker, Boston, Kock, Schneider-Gold, Siebner, Binkofski, Lencer, Münchau, and Klein. Obtained funding: Binkofski, Münchau, and Klein. Administrative, technical, and material support: Hiller, Lohnau, Schneider-Gold, Kress, Binkofski, and Klein. Study supervision: Klein.

Funding/Support: This study was supported by grants from the Deutsche Forschungsgemeinschaft, the Parkinson's Disease Foundation/National Parkinson Foundation, the Bundesministerium für Bildung und Forschung (grant 01GI0201), and the University of Lübeck (all to Dr Klein); and the VolkswagenStiftung (Drs Binkofski, Münchau, and Klein).