PARKIN, PINK1 and DJ1 analysis in early onset Parkinson’s disease in Ireland

Abstract

Variants in PARKIN, PINK1, and DJ1 are associated with early onset Parkinson’ disease (EOPD, age-at-onset<45). We previously reported a single PINK1 and a single DJ1 heterozygous variant carriers. We aimed to expand upon our previous EOPD studies and investigate for any genotype-phenotype correlations in Irish PD. 314 PD patients were recruited from Dublin Neurological Institute, Ireland. Genetic analysis was performed at the Mayo Clinic, Jacksonville, USA. We screened 81 patients with young onset PD (age-at-onset<50), of which 58 had EOPD. We identified 4 patients with homozygous/compound heterozygous variants and 3 heterozygote carriers (pathogenic PINK1/DJ1 variants were not found). Expansion of one of the pedigrees showed a novel variant in exon 9, in a symptomatic patient. We identified 6.89% PARKIN variant carriers associated with EOPD. These findings suggest that PINK1 and DJ1 are rarely associated with Irish YOPD, while PARKIN variant frequency is similar to that reported worldwide.

Introduction:

Young onset Parkinson’s disease (YOPD) is defined by the age-at-onset <50 years and early onset Parkinson’s disease (EOPD) by the age-at-onset <45 years.¹ While variants in PARKIN (PARK2), PINK1 (PARK6) and DJ1 (PARK7) are associated with EOPD, PARKIN variants are more common (found in 50% of autosomal recessive (AR) monogenic PD and 15% of sporadic EOPD).² With advancing age, the prevalence of PARKIN variants decreases decreases (77% <20 years, and 2–7% at 30–45 years).³ Over 200 PARKIN variants have been described including: multiexonic deletions, point mutations, truncating mutations and copy number variations.⁴ Single heterozygous PARKIN variant carriers occur in 0–7.9% of PD and 0–3.7% of controls and their significance remains controversial.^5–7 The clinical characteristics of PARKIN-PD include: dystonia, which may predate parkinsonism, ⁸ early dyskinesia in feet and legs, abnormal gait pattern that does not vary with “on” and “off”, ⁹ symmetric onset, hyperreflexia, diurnal variation with sleep benefit ¹⁰ and slow disease course. ⁸ The response to levodopa is usually excellent and the lowest possible levodopa dose should be used for treatment. A higher frequency of compulsive shopping, binge eating, and punding/hobbyism was recently reported. ¹¹PARKIN disease is a predominantly ventral nigropathy with the involvement of locus coeruleus. In contrast to non-Parkin PD, it has a distinct pathology characterized by the absence (or rare occurrence) of the Lewy Bodies. This corresponds to a more benign phenotype with rare dementia. ¹²

Variants in PINK1 are the second most common cause of AR PD.² While the prevalence of PINK1 variants in familial AR PD was reported at 4.26%, the frequency in sporadic PD ranges from 0.53%1.2% and 0.39% in controls.¹³ There have been 40 point variants reported (mostly missense and nonsense) and, in contrast to PARKIN, structural deletions are rare.^14,15 Recent studies suggest that heterozygote variants in PINK1 may also increase the PD risk three-fold.¹⁶

Similarly to PARKIN-PD, PINK1-PD is characterized by the early age-at-onset (AAO) (however onset in the eight decade was described in one patient with compound heterozygote mutation), ¹⁷ slow progression and good levodopa response; however in contrast to PARKIN-PD dystonia early on in the disease is not typical. ¹⁸ Moreover, psychiatric comorbidity (anxiety, dysphoria, depression, cognitive decline) and gait problems are common in PINK1-PD in comparison to PARKIN-PD and may precede the development of motor symptoms. ^{19, 20}

The frequency of DJ1 variants in EOPD is estimated at 0.2%–1%. ^21,22 There have been 20 causative DJ1 variants described to date.²³ The phenotype is similar to that of PARKIN with PD onset between the ages 20 and 40, good levodopa response and slow progression. ²³ However psychiatric (cognitive impairment, anxiety and depression) ²⁴ and atypical (blepharospasm, laterocollis) features may also occur. ²⁵ The ¹⁸F-dopa uptake on the functional neuroimaging was reported as normal in the asymptomatic heterozygous DJ-1 carriers implying that loss of DJ1 protein function is needed to cause the disease. ²⁶

To date, EOPD screening in Ireland has identified only one PINK1 heterozygote (novel R147H variant in exon 2, absent in controls) and one DJ1 heterozygote (variant of unknown significance-VUS in exon 3, c.212C>T, p. P73L, rs367584305).^27,28 Our objective was to expand upon our previous studies in EOPD and investigate for any genotype-phenotype correlations in Irish PD.

Materials and Methods:

The ethical approval was granted by the the Mater Misericordiae University Hospital (MMUH), Dublin, Ireland (IRB#1/378/1300). Written consent was obtained. 314 unrelated Irish PD patients (UK Brain Bank Criteria) > age 18 were recruited from Dublin Neurological Institute (DNI) at the MMUH (www.neurologicalinstitute.ie). We studied 81 patients with YOPD. Of these 58 had EOPD with the AAO <45 years (mean 37.81 years (±6) (range16–44)). We isolated genomic DNA from whole blood using the QIAmp Blood Midi Kit^® (Qiagen, Germantown, MD, USA).

We performed genetic screening at the Mayo Clinic, Jacksonville, FL, USA. All patients were pre-screened for the most common causes of autosomal dominant disease (LRRK2, SNCA, VPS35). Three hundred and fourteen patients were tested for specific point mutations; LRRK2 p.G2019S, SNCA p. A53T, p.A30P and VPS35 p.D620N (TaqMan Allelic Discrimination Assay, on Quant Studio System Real-Time PCR System). Bi-directional Sanger sequencing of all coding exons plus 25 base pair of exon-intron boundary of PARKIN (12 exons) (NM_004562.2), PINK1 (8 exons) (NM_032409.2) and DJ1 (6 exons) (NM_001123377.1)²⁹ was performed in the 81 YOPD patients using the Applied Biosystems DNA Analyzer 3730xl. We analysed all sequence data using Seqscape v2.5 (Life Technologies, Carlsbad, CA, USA). Multiplex ligation-dependent probe amplification (MLPA) was performed using SALSA MLPA Kit P051-C1/P052-C1 (MRC-Holland, Netherlands) (PARKIN, PINK1, DJ1 and SNCA). The pathogenicity of the variants was determined based on the current literature, and in sillico prediction tools (Polyphen v2, ³⁰ SIFT, ³¹ Mutation Taster ³²).

Statistical analysis was performed using IBM SPSS Statistics, 22.0. Bivariate associations were calculated using χ2 /Fisher’s exact tests or t-tests/Mann Whitney U-tests appropriately, with significance where p<0.05. Linear and logistic regression models were used for genotype-phenotype correlation.

Results:

No pathogenic mutations in SNCA, LRRK2 or VPS35 were identified in 314 patients. We screened 81 unrelated patients with YOPD of which 58 had EOPD (Table 2) for PARKIN, PINK1 and DJ1 with a genotyping success rate of 100%. We identified 7 PARKIN variants: 3 missense variants (p.R275W, p.G430D, and p.P437L), 1 frame shift deletion causing the premature introduction of a stop codon (p.Pro113fs), and 3 exonic deletions (exon 4 & 5 and exon 3). The distribution of PARKIN variants were in Hardy-Weinberg equilibrium. There was one homozygote carrier (p.G430D), two compound heterozygote carriers (p.Leu112fsX163/p.R275W and p.G430D/Ex4&5del) and one heterozygote carrier for p.R275W/Ex3 del (the family of the p.R275W/Ex3 del carrier was not available to screen for an inheritance pattern) (Suppl.Table 2). Additionally, 3 heterozygote carriers (p.R275W(n=1) and p.P437L (n=2)) were identified. We did not identify any pathogenic PINK1 or DJ1 variants.

The homozygous G430D carrier and carriers of multiple heterozygous PARKIN variants (n=4) had an AAO< 40 years, therefore we compared these 4 carriers to the group of 54 non-carriers with EOPD (n=58) (Table 2). Three probands harbouring single heterozygote PARKIN R275W and P437L mutations were excluded from the comparison (PARKIN heterozygosity is controversial). ³³

Three (75%) PARKIN biallelic variants carriers were women in comparison to 18 (33.3%) of non-carriers (p=0.13, Fisher’s exact test). The number of carriers was limited, but statistical analysis yielded two significant results. The mean AAO in 4 PARKIN biallelic variants carriers was lower 24.2 (±8.5) (range 16–35) than in non-carriers 38.8 (±4.5) (range 28–44), p<0.000001, OR 14.3 (95%CI 9.2–19.5) and the disease duration was longer mean 30.2 years (±20.4) (range 6–50) versus 11.4 years (±8.5) (range 2–39)(p=0.04, Mann-Whitney U test). Three of 4 carriers (75%) had family history of PD versus 34/54 (63%) non-carriers. The first symptom in all 4 PARKIN biallelic variants carriers was tremor, 75% had tremor-predominant subtype, and UPDRS-III score <32, 50% had levodopa-induced “ dancing feet” dyskinesia and 25% wearing off (Table 1). None of the carriers had dystonia, freezing of gait or dementia and 1 carrier exhibited impulsive compulsive behaviours (on a dopamine agonist). There were no prodromal symptoms reported. We identified 1 heterozygous carrier of a PARKIN p.P437L mutation with AAO of 49 years (no other PARKIN carriers had AAO between 45–50 years). In the current study we found that homozygous PARKIN patients and patients with multiple heterozygous pathogenic PARKIN variants were responsible for 6.89% (4/58) of EOPD in Ireland. We performed segregation analysis in two families.

Table 1.

Genotype-phenotype assessment in PARKIN pathogenic variants carriers with PD versus non-carriers

	EOPD (AAO<45)
	PARKIN pathogenic variants carriers n=4	Non-carriers n=54	p value
Gender	n (%)	n (%)
Women	3 (75%)	18 (33.3%)
Men	1 (25%)	36 (66.7%)	0.13b
	median (range)	median (range)
Age-at-assessment	53.5 (40–71)	49.5 (32–79)	0.75c
Age-at-onset	23 (16–35)	40 (28–44)	<0.01 c
	median (range)	median (range)
Disease duration	32.5 (6–50)	9.5 (2–39)	0.04 c
Family history	n (%)	n (%)
yes	3 (75%)	34 (63%)
no	1 (25%)	20 (37%)	1b
Cognition	n (%)	n (%)
MCI-PD	2 (50%)	6 (11.1%)	0.88b
PDD	0 (0%)	9 (16.7%)	1b
Hallucinations	1 (25%)	6 (11.1%)	0.41b
Motor complications	2 (50%)	26 (48.1%)	1b
Wearing off	1 (25%)	19 (35.2%)	1b
Dyskinesia	2 (50%)	17 (31.5%)	0.59b
Dystonia	0 (0%)	5 (9.3%)	1b
Freezing of gait	0 (0%)	8 (14.8%)	1b
DBS in situ	0 (0%)	3 (5.6%)	1b
	median (range)	median (range)
UPDRS-III score	26.5 (8–44)	16 (3–51)	0.34c
UPDRS-III categories	n (%)	n (%)
<32 (mild)	3 (75%)	44 (81.5%)
33–58 (moderate)	1 (25%)	10 (18.5%)
>58 (severe)	0 (0%)	0 (0%)	1b
Falls	1 (25%)	7 (12.1%)	0.45b
Medications	n (%)	n (%)
Levodopa	4 (100%)	43 (79.6%)	1b
MAOBI	2 (50%)	35 (64.8%)	0.61b
Dopamine agonists	2 (50%)	27 (50%)	1b
Madopar
(levodopa+benserazide)	0 (0%)	3 (5.6%)	1b
Anticholinergics	0 (0%)	8 (14.8%)	1b
Amantadine	1 (25%)	4 (7.4%)	0.31b
Duodopa	1 (25%)	1 (1.9%)	0.13b
Apomorphine	0 (0%)	4 (7.4%)	1b
Subtype
Tremor-predominant	3 (75%)	27 (50%)
PIGD	1 (25%)	16 (29.6%)
Mixed	0 (0%)	11(20.4%)	0.34d
First symptom
Tremor	4 (100%)	35 (64%)	0.29b
Bradykinesia	0 (0%)	14 (25.9%)	0.56b
Stiffness	0 (0%)	1 (1.9%)	1b
Writing difficulties	0 (0%)	2 (3.7%)	1b
Pain	0 (0%)	0 (0%)	N/a
Speech problems	0 (0%)	0 (0%)	1b
Gait problems (all)	0 (0%)	2 (3.7%)	1b
Loss of arm swing	0 (0%)	0 (0%)	N/a
Shuffling	0 (0%)	1 (1.9%)	1b
Dragging a leg	0 (0%)	1 (1.9%)	1b
Balance problems	0 (0%)	0 (%)	N/a

Legend:

^a:Pearson Chi Square test

^b:Fisher Exact test

^c:Mann Whitney U test

^d:Linear by Linear Armitage exact trend test

bold: statistically significant result.

Family 1

The proband was a 41-year-old woman (III:1, Fig.1a) who developed right hand tremor at age 35. Her brother (III:2, Fig.1a) developed tremor at age 12, and dystonia at 17 (diagnosed with EOPD at age 21). The proband’s other brother (III:3, Fig.1a) had dyskinesia of frontalis (pre-levodopa exposure), corrugators supercilii, orbicularis oris muscles and foot dystonia at age 31 (diagnosed with PD two years later). The proband’s maternal grandfather (I:3 in Fig.1a) was diagnosed with PD in his sixth decade. The proband (III:1, Fig.1a), two affected brothers (III:2; III:3, Fig.1a) and an unaffected sister (III:6, Fig.1a) carried PARKIN p.Leu112fsX163/p.R275W variant. In another unaffected sister (III: 7, Fig.1a), we identified the 40bp deletion in exon 3, but not R275W variant indicating that p.Leu112fsX163 and p.R275W are not co-inherited on the same allele but are in a compound heterozygous status. Both of the probands parents were deceased at the time of the study.

Fig. 1.

a Family 1 pedigree, b Family 2 pedigree, red: affected with Parkinson–s disease, blue: PARKIN+

Family 2

The proband was a 66-year-old woman (III:6, Fig.1b) who developed a bilateral hand tremor at age sixteen. She was diagnosed with EOPD at 24-years and commenced on a low dose of levodopa. Overtime she required levodopa 200mg/day, ropinirole 8mg/day, amantadine 100mg/day and selegiline 5mg/day. Prominent lower limb peak dose Irish ”dancing‟-style dyskinesia developed (at age 61). The proband’s sister (III:8, Fig.1b) developed bilateral hand tremor at age 30 and Irish dancing dyskinesia. She developed postural instability, anxiety and depression with dementia. A paternal cousin had a daughter (IV:1, Fig.1b), who developed foot dystonia in her twenties and later on developed right hand tremor, torticollis and ”dancing‟ legs dyskinesia (on levodopa treatment). The proband (III:6, Fig. 1b) and affected sister (III:8, Fig. 1b) were positive for a compound heterozygous p.G430D/Ex4&5 del PARKIN variant. Two unaffected daughters of the affected sister (III:8, Fig.1b) carried the exon 4 & 5 deletions but not the maternal exon 12 missense variant present in their mother III:8, Fig. 1b). Analysis of IV:1, a daughter of the probands first cousin, revealed the missense variant in exon 12 (c.1289G>A, p.Gly430Asp) along with a novel nonsense variant in exon 9 (c.1045A>T, p.Lys349Stop). This nonsense variant inserts a stop codon and has not been reported in the exome variant server.²⁹ This stop codon produces a false impression of a heterozygous deletion of exon 9 on MLPA analysis.

Discussion:

We previously reported PARKIN variants frequency in Ireland occurring at 17.5%. ³⁴ We also reported one single PINK1 heterozygous variant (p.R147H) carrier (AAO of 51 years) in 290 Irish PD patients, 86 of whom had AAO<45 years, ²⁷ and 1 heterozygous DJ1 variant carrier with AAO of 50 years. ²⁸ In this study we identified 4/58 (6.89%) PARKIN variant carriers associated with EOPD. We did not identify any pathogenic variants in PINK1 or DJ1, however caution in interpretation is needed as the sample size was small.

When we combine the results we found PARKIN variant frequency of 11.22% (11/98) in Irish EOPD.³⁴ In both studies, ³⁴ we found marginally more PARKIN variant carriers with family history of PD (combined n=7/11, 63.6%). In this current study the AAO was younger [age-at-onset mean 24.2 years (SD±8.5) and range 16–35] in carriers versus non-carriers [mean 38.8 years (SD±4.5) and range 28–44] in comparison to our prior study [carriers mean 35 years (SD±7), non-carriers mean 35 years (SD±8)], ³⁴ We also found a longer disease duration in carriers than in non-carriers in both studies [current study mean 30.2 years (SD±20.4) (range 6–50) versus non-carriers mean 11.7 years (SD±8.9) (range 2–39) and in our Wiley et al., study mean 17.11 years (SD±11) in carriers and mean 12 (SD±8) years in non-carriers) ³⁴ in keeping with the earlier onset and more benign disease.

We found 3 PARKIN heterozygotes carriers and 2 with AAO <45 (3 in our prior study). The overall frequency of PARKIN heterozygote carriers in Irish EOPD was 5.1% (5/98) (AAO <45) and 6.12% (6/98) in YOPD (AAO <50). Expansion of one of the pedigrees led to the discovery of a novel nonsense variant in exon 9, c.1045A>T, Lys349Stop in a symptomatic patient, and this finding needs to be further investigated. Two patients had “Irish dancing” dyskinesia that has been reported with PARKIN phenotype. ³⁵

Conclusion:

The findings of the current study suggest that PINK1 and DJ1 are rarely associated with YOPD in Ireland, while PARKIN variant frequency is similar to that reported worldwide. These results have implications for clinicians caring for patients with EOPD in Ireland, and elsewhere, as the recognition of a genetic cause of PD affects treatment decisions (e.g., avoidance of high dose levodopa in Parkin-PD), prognosis and appropriate genetic counseling in these younger patients.

Availability of data and material:

The data will be provided upon a reasonable request.