DNAJC6 Mutations Disrupt Dopamine Homeostasis in Juvenile Parkinsonism‐Dystonia

Abstract

Classical Parkinson's disease (PD) is an age‐related neurodegenerative disorder, mainly affecting adults aged >50 years. Patients typically present with resting tremor, bradykinesia, rigidity, and postural instability. To date, a number of early‐onset genetic forms of PD (Parkin, PINK1, and DJ‐1)^{1, 2, 3} and complex parkinsonism syndromes (ATP13A2, PLA2G6, FBXO7, SLC6A3, SLC39A14, and PANK2) have been described.^{4, 5, 6, 7, 8, 9} Importantly, the study of such monogenic forms of disease have provided significant insight into the pathogenic mechanisms underlying sporadic PD.¹⁰ More recently, two genes, namely SYNJ1 ¹¹ and DNAJC6,^{12, 13, 14} encoding proteins involved in postendocytotic recycling of synaptic vesicles, have been identified in early‐onset parkinsonism.

In this study, we report on 6 children from three families, presenting with juvenile parkinsonism‐dystonia associated with novel, biallelic DNAJC6 mutations. We delineate their clinical phenotype, neuroimaging features (including ¹²³I‐FP‐CIT single‐photon emission computed tomography [SPECT; DaTScan]) and pattern of cerebrospinal fluid (CSF) neurotransmitter metabolites. Furthermore, we utilized patient fibroblasts and CSF to investigate secondary effects on auxilin, cyclin G–associated kinase (GAK), and dopaminergic proteins.

Materials and Methods

Results

Patient Cohort

A total of 232 children were referred with undiagnosed movement disorders for genetic research (Fig. ​(Fig.1A).1A). Of these, 25 children (10.7%) had juvenile parkinsonism, 16 females and 9 males with a current median age of 14 years (range, 4–28). Fourteen of 25 had additional neurological features, including dystonia (14 patients), developmental delay/learning difficulties (14 patients), and seizures (4 patients; Fig. ​Fig.11B).

Open in a separate window

Figure 1

Juvenile parkinsonism cohort: clinical features and molecular genetic investigation. (A) Flowchart demonstrating the pathway of molecular genetic investigations in a subcohort of 25 children with juvenile parkinsonism. (B) Clinical characteristics of 20 children from 17 families. Early infancy <3 months; infancy 3 to 12 months; toddler 12 to 24 months; childhood 2 to 13 years; adolescence 13 to 18 years. *Consanguineous family. M, male; F, female; N, normal; NP, not performed; NR, not reported. [Color figure can be viewed at wileyonlinelibrary.com]

Molecular Genetic Investigations

Families A and B (Fig. ​(Fig.1A)1A) were prioritized for autozygosity mapping studies. SNP genotyping revealed a 4.33‐Mb region of common homozygosity in both families on chromosome 1, between rs640407 (64,267,606 base pairs [bp]) and rs2566784 (68,602,735 bp; Fig. ​Fig.2A,B).2A,B). This region showed a common haplotype in all affected children whereas unaffected siblings had a different haplotype. It was therefore considered to be the likely disease locus (Fig. ​(Fig.22B).

Open in a separate window

Figure 2

Molecular genetic investigations and DaTSCAN imaging. (A) Family A and B SNP array results showing homozygous regions detected. For each chromosome, the start and end point is specified using the Reference SNP Cluster ID (rs number) and physical position. (B) Homozygous SNPs are represented in light blue (AA) and dark blue (BB), heterozygous SNPs in red (AB), and “no calls” in white. (C) Sanger Sequencing confirms a homozygous DNAJC6 mutation, c.766C > T (p.R256*), in all affected children of Family A (A‐III:1, A‐III:4, and A‐III:5) and Family B (B‐IV:2, B‐IV:4). Parents are heterozygous carriers. (D) I‐123‐DaTSCAN™ with SPECT imaging in a control subject, Patient A‐III:1 (19 years 3 months), Patient A‐III:4 (11 years 4 months), and Patient B‐IV:4 (17 years). In Patients A‐III:1 and B‐IV:4, DaTSCAN findings indicate virtually complete absence of tracer uptake in the basal ganglia, with very high background activity, suggesting loss of presynaptic dopaminergic terminals, whereas Patient A‐III:4 showed significantly reduced, albeit still visible, uptake in the head of caudate (left better than right, white arrows). [Color figure can be viewed at wileyonlinelibrary.com]

WES performed in Patients A‐III:1 and B‐IV:2 revealed 23,365 and 23,549 variants, respectively. Given familial consanguinity and the autozygosity mapping results, targeted analysis for recessive pathogenic variants within the putative disease locus on chromosome 1 was undertaken. A single homozygous nonsense variant c.766C>T (p.R256*) in DNAJC6 (Chr1, 65,248,219–65,415,869) was identified both in A‐III:1 and B‐IV:2 on WES, located within the common region of homozygosity. No other pathogenic changes in previously reported genes causing juvenile parkinsonism‐dystonia phenotypes were identified from the WES data. Direct Sanger sequencing confirmed the homozygous c.766C>T mutation in all 5 affected patients, and familial segregation studies revealed that all parents were obligate carriers in both kindreds, with unaffected siblings either wild type or heterozygous for the identified variant (Fig. ​(Fig.2C).2C). WES/whole‐genome sequencing data for the remainder of the parkinsonism‐dystonia cohort (n = 20) was interrogated for DNAJC6 mutations. This led to the identification of a homozygous nonsense variant (c.2416C>T) in a sixth unrelated patient (Patient C), which was confirmed on Sanger sequencing.

Delineation of the Clinical Phenotype of DNAJC6 Patients

Family A (3 Affected Patients)

Patients A‐III:1, A‐III:4, and A‐III:5 are 3 affected children born to first‐cousin parents, currently 20, 12, and 10 years old (Table ​(Table1).1). Two other brothers (A‐III:2 and A‐III:3), aged 17 and 15 years, have mild learning difficulties without evidence of a movement disorder. The paternal grandfather was diagnosed with PD in his fifties.

All 3 children were born at term after an uneventful antenatal period. Microcephaly was evident at birth (head circumference: <0.4th centile), but nonprogressive over time. All siblings had early neurodevelopmental delay and moderate learning difficulties.

A‐III:1 is the eldest daughter, aged 20 years. She presented at 10 years, with a 6‐week history of feeding difficulties, vomiting, and weight loss. Over time, she developed fever, unsteady gait, facial asymmetry, left‐sided tremor, and generalized seizures and was diagnosed with an encephalitis of uncertain etiology. She recovered from this acute illness, but subsequently had progressive bradykinesia, with tremor and rigidity, and loss of independent ambulation at 13 years, associated with cognitive decline. She is now wheelchair dependent, with generalised cogwheel‐rigidity, severe bradykinesia, multiple limb contractures and emotional lability (Video 1). She also has severe gut dysmotility, with recurrent vomiting, and required a gastrostomy for deteriorating bulbar dysfunction. CSF neurotransmitter analysis (age 11 years), while on levodopa therapy, revealed an isolated low 5‐hydroxyindoleacetic acid (5‐HIAA; Fig. ​Fig.3A).3A). At 12 years 11 months, when off l‐dopa, CSF HVA, and HVA:5‐HIAA ratio were low. Brain MRI showed evidence of right‐sided atrophy of the perisylvian region and right cerebellum by 19 years of age (Supporting Information Fig. S1). At 19 years, ¹²³I‐FP‐CIT SPECT (DaTScan) showed absent uptake in the basal ganglia when compared to normal subjects (Fig. ​(Fig.2D).2D). At this stage, while on l‐dopa treatment, her CSF HVA levels normalized (Fig. ​(Fig.3A).3A). Her condition is refractory to medical treatment, with no clinical response to trihexyphenidyl, benzhexol, procyclidine, clobazam, rotigotine, and apomorphine. l‐dopa has proven difficult to titrate because of marked drug sensitivity. She experiences an improvement in motor function and speech 30 minutes postdose, after which she returns to the off state. l‐dopa dosages >150 mg/d have resulted in drug‐related dyskinesias.

Open in a separate window

Figure 3

CSF neurotransmitter analysis and patient fibroblast and CSF immunoblotting. (A) CSF neurotransmitter analysis. Age‐related reference ranges indicated in brackets after each value. Red: abnormal result. Gray: borderline result. Symbol (“#”) indicates reference range: ¹Keith Hyland, Robert A.H. Surtees, et al. Pediatr Res 1993;34:10–14; ²Keith Hyland, Future Neurol 2006;1:593–603; ³Surtees R, Hyland K. Biochem Med Metab Biol 1990;44:192–199. (B) Scatterplot of CSF HVA and 5‐HIAA levels (nmol/L) measured by high performance liquid chromatography (patient = red shapes, control = black triangles). Medication at time of CSF sampling: A‐III‐1: co‐careldopa, melatonin, glycopyronium; A‐III‐4: none; B‐IV‐4: l‐dopa, pyridoxine; Control 1: none; Control 2: none. Immunoblot of auxilin and GAK in patient fibroblasts (C) and CSF (D) compared to controls. (E) Immunoblot of patient CSF for TH, DAT, VMAT, and D2R protein levels measured compared to controls. Graphs show mean protein percent optical density (OD) normalized to loading control in patients (red) and controls (black). LP, lumbar puncture; y, years; m, months; 5‐MTHF, 5‐methyltetrahydrofolate; NP, not performed. [Color figure can be viewed at wileyonlinelibrary.com]

Her two brothers (A‐III:4 and A‐III:5) presented with fine motor difficulties at 8 years of age. They subsequently developed positional tremor, upper limb dystonic posturing, hypophonia, hypomimia, bradykinesia, cogwheel rigidity, and postural instability over 12 months (Videos 2 and 3). Like their sister, both have gastrointestinal complications with sialorrhea, recurrent vomiting, and feeding difficulties, necessitating gastrostomy insertion. A‐III:4 is currently 12 years old and suffers from anxiety and perseveration. His CSF‐HVA levels are at the lower limit of normal, with a low HVA:5‐HIAA ratio <1.0 (Fig. ​(Fig.3A).3A). MRI brain scan was normal. ¹²³I‐FP‐CIT SPECT (DaTScan) at 11 years showed profound reduction in tracer uptake in the basal ganglia (Fig. ​(Fig.2D).2D). Both boys responded to treatment with transdermal rotigotine and oral trihexyphenidyl, but with increasing doses, both experienced dyskinesias, necessitating dose reduction.

Family B (2 Affected Patients)

Patients B‐IV:2 and B‐IV:4 are 2 affected girls, born to first‐cousin parents, and currently 28 and 19 years old. Both were born uneventfully following a normal pregnancy, presenting with early feeding difficulties, hypotonia, and delayed milestones by 6 months old. Both made slow developmental progress, achieving independent ambulation and spoken language by 3 years of age.

B‐IV:2 presented at 9 years with generalized seizures that stabilized with lamotrigine therapy. From 13 years of age, motor and cognitive deterioration ensued, with onset of parkinsonism and loss of speech and ambulation. She experienced anxiety and recurrence of seizures. She has severe antecollis, hypomimia, tremor, generalised cogwheel rigidity, bradykinesia, and positive glabellar tap (Video 4). MRI was normal until 18 years, after which there was radiological evidence of mild generalized atrophy. Several medications were tried without clinical benefit, including l‐dopa (maximum, 10 mg/kg/d), selegiline, rotigotine, and trihexyphenidyl. The younger sibling, B‐IV:4, presented at 7 years with gait deterioration, bradykinesia, and cogwheel rigidity. She lost independent ambulation and speech by 10 years (Video 5). She has developed dystonic posturing, bulbar dysfunction (necessitating gastrostomy), and a disrupted sleep pattern. The MRI brain scan was initially normal, but by 16 years showed subtle global cerebral atrophy (particularly in the posterior regions) as well as cerebellar atrophy (Supporting Information Fig. S2). ¹²³I‐FP‐CIT SPECT (DaTScan) at 17 years showed profound reduction in tracer uptake in the basal ganglia (Fig. ​(Fig.2D).2D). CSF HVA and HVA:5‐HIAA ratio were reduced at ages 4 and 14 years. CSF‐HIAA levels were reduced at age 4 years (Fig. ​(Fig.3A).3A). She showed an initial response to l‐dopa, but developed emotional lability at 5.5 mg/kg/d, leading to drug withdrawal. There was no clinical improvement observed with trihexyphenidyl or chloral hydrate. She had a modest response to pramipexole, with improved facial expression, reduced tremor, and increase in voluntary movements.

Family C

This 18‐year‐old girl is the third child of distantly related Latin American parents, with 2 healthy siblings. She initially presented with neonatal feeding difficulties and hypotonia. In infancy, she showed delay in attaining milestones and developed seizures characterized by staring episodes with loss of tone. She walked independently from 2 years, but by 10 years of age her gait deteriorated, leading to frequent falls, postural instability, and losing the ability to run. Over the next 4 years, she continued to deteriorate with worsening antecollis and bradykinesia (Video 6). She developed severe bulbar dysfunction with sialorrhea, dysarthria, and, dysphagia, leading to considerable weight loss. At 12 years, she developed generalized tonic‐clonic seizures and atypical absences, responsive to lamotrigine and zonisamide therapy. MRI demonstrated subtle generalized cerebral atrophy and CSF HVA was low (Fig. ​(Fig.3A).3A). Her movement disorder responded to l‐dopa, with improved tremor, gait, and a reduction in drooling. A maximum of 200 mg/d was tolerated, but further increases led to intolerable drug‐induced dyskinesias. After 4 months of treatment, she developed aggressive behavior and received treatment with quetiapine. By 16 years, she became increasingly sensitive to l‐dopa, with peak‐dose agitation, restlessness, and dyskinesia. Lowering the dose improved side effects, and continued to provide motor benefit, although the effects wore off 2 to 3 hours after administration. On‐off phenomena were commonly reported, and in the off state, she was often akinetic and rigid. Introduction of trihexyphenidyl improved rigidity, but not immobility.

Discussion

Juvenile parkinsonism attributed to DNAJC6 mutations has only recently been reported. Here, we report on a further 6 patients from three families, with two previously unreported homozygous nonsense mutations in DNAJC6. Moreover, our findings on ¹²³I‐FP‐CIT SPECT (DaTScan) imaging, CSF analysis, and immunoblotting suggest downstream dyshomeostasis of auxilin, GAK, and dopaminergic proteins in DNAJC6‐related disease.

Our data confirms that all reported cases of juvenile‐onset DNAJC6‐parkinsonism have core clinical characteristics (Table ​(Table1),1), including (1) clinical presentation of progressive parkinsonism toward of the first decade (median, 10 years; range, 7–13), (2) significant neurological regression thereafter, and (3) loss of ambulation in mid‐adolescence.^{12, 13, 14, 21} In contrast to adult‐onset PD, childhood parkinsonian disorders rarely present with a “pure” parkinsonian phenotype, as illustrated by the classical primary pediatric monoamine neurotransmitter disorders.²⁰ Similarly, in early‐onset DNAJC6‐related disease, parkinsonism is commonly present in tandem with a multitude of other clinical features, including dystonia, moderate learning difficulties, epilepsy, and neuropsychiatric features^{12, 13, 14, 21} (Table ​(Table1).1). Furthermore, many of our patients had evidence of bulbar dysfunction, gut dysmotility, and sleep disturbance. The majority of our patients showed limited response to l‐dopa and other standard therapies for parkinsonism‐dystonia. They experienced severe, often intolerable, side effects with dopaminergic agents, including on‐off phenomenon and severe dyskinesia, particularly at higher drug dosages.

¹²³I‐FP‐CIT SPECT (DaTScan) was performed in 3 patients, demonstrating reduced tracer uptake in the basal ganglia, suggestive of impaired presynaptic dopamine uptake and striatonigral neurodegeneration. Postmortem studies have confirmed striatal dopamine deficiency in patients with parkinsonism.^{22, 23} Together, these observations suggest a neurodegenerative process in DNAJC6 patients. MRI brain imaging further corroborates this hypothesis; the mild generalized cerebral and/or cerebellar atrophy in 4 patients suggests that DNAJC6‐related disorders may also be associated with neuronal loss in other regions of the central nervous system.

All 6 cases fit the juvenile phenotype associated with this gene, though more recently, DNAJC6 mutations have been reported in early adult‐onset PD.¹⁴ Although there are a number of overlapping features (progressive parkinsonism, psychiatric features), affected patients presented later (range, 21–42 years) and seizures and cognitive decline are not reported.

Homozygous and compound heterozygous mutations in DNAJC6 are predicted to result in loss of protein function. To date, splice‐site variants,¹² large multiexonic deletions,²⁴ truncating mutations,¹³ and missense mutations¹⁴ have been reported. All 6 patients in our cohort had nonsense mutations, predicted to cause nonsense‐mediated decay or premature protein truncation. Five of the 6 reported patients are from two consanguineous families originating from the same region in Pakistan, and all have the same nonsense mutation. SNP array confirmed a common haplotype at this disease locus for all affected children, suggesting a possible founder effect.

DNAJC6 encodes for auxilin, a neuronally expressed J‐chaperone protein involved in the uncoating of clathrin‐coated vesicles^{25, 26} (Fig. ​(Fig.4).4). Auxilin modifies the three‐dimensional conformation of heavy‐chain clathrin triskelions, leading to clathrin coat distortion, instability, and subsequent disassembly.^{25, 26} Neurotransmission involves rapid continuous recycling of synaptic vesicles through CME. Deficiency in auxilin ultimately results in impairment of synaptic vesicle recycling and impaired neurotransmission. Similarly, aberrant synaptic vesicular trafficking is also evident in other forms of early‐onset parkinsonism, including LRRK2,^{27, 28} VMAT2,²⁹ and SNCA‐related disease.³⁰ Clathrin‐mediated endocytosis is crucial for the regulation of developmental signaling pathways through internalization of receptors or ligands and is required for axon and dendrite outgrowth.³¹ Presence of developmental delay well before onset of parkinsonism in patients with DNAJC6 mutations further corroborates the notion that auxilin is likely to have a central role in neurodevelopment, given its role in CME. In Drosophila, auxilin is crucial for Notch signaling, a developmental pathway that regulates neural stem‐cell proliferation, survival, renewal, and differentiation, as well as neuronal specification of dopaminergic neurons.^{32, 33, 34}

Open in a separate window

Figure 4

Schematic representation of DNAJC6‐encoded auxilin protein function. The role of auxilin in synaptic vesicle recycling and endocytic pathway. (A) A nascent clathrin‐coated pit is formed at the presynaptic membrane, followed by membrane invagination; (B) the pit develops into a clathrin‐coated vesicle, where the clathrin lattice consists of (C) clathrin triskelions formed by three crossed ankle regions. (D) Auxilin (pale green square) binds to an exposed domain in the heavy chain of the clathrin. (E) Auxilin facilitates a conformational change in clathrin that allows binding of Hsc70 and by ATPase‐mediated activity, and the clathrin lattice is disrupted and distorted, leading to (F) clathrin disassembly allowing subsequent delivery of cargo neurotransmitters to the membrane or other vesicle in the endocytic pathway. Hsc70, heat shock cognate protein 70. [Color figure can be viewed at wileyonlinelibrary.com]

To investigate the downstream effects of DNAJC6 mutations, we studied auxilin and GAK protein levels in 2 patients using patient fibroblasts and CSF. Auxilin is a neuron‐specific protein, enriched in presynaptic terminals, whereas GAK is an ubiquituously expressed protein.^{35, 36} Auxilin and GAK are highly homologous proteins that both have the ability to bind clathrin and clathrin adaptor protein 2 in order to initiate clathrin uncoating of endocytosed vesicles.³⁷ In the auxilin knockout mouse model, it is reported that upregulation of GAK can partially compensate for the loss of auxilin and decrease mortality.³⁵ We therefore wished to determine whether a similar compensatory mechanism was evident in our patients. In our study, we observed that patient fibroblast auxilin protein levels were significantly reduced when compared to controls, as previously reported.¹⁴ We found that patient fibroblast GAK levels were slightly, but not significantly, increased, whereas patient CSF GAK protein levels were significantly increased. Our findings support upregulation of brain GAK levels in DNAJC6 patients, partially compensating for auxilin reduction, as evident in the auxilin knockout mouse model.³⁵

Diagnostic CSF neurotransmitter analysis revealed that levels of the stable dopamine metabolite (HVA) were either below the age‐related reference ranges or close to the lower limit of normal in our patients, indicating impaired dopamine turnover. Indeed, CSF‐HVA levels and HVA:5‐HIAA ratios were comparable to those observed in TH deficiency, an inherited dopamine synthesis defect associated with central dopamine deficiency.²⁰ In order to determine how DNAJC6 mutations may impact the dopaminergic system, we used patient CSF to analyze proteins involved in dopamine signaling and homeostasis. We observed that patient CSF had significantly reduced levels of VMAT, DAT, TH, and D2R when compared to controls. ¹²³I‐FP‐CIT SPECT (DaTScan) imaging additionally provides in vivo evidence of impaired DAT function in DNAJC6 patients. VMAT and DAT are both synaptic transporters recycled through clathrin‐mediated endocytosis.^{38, 39} The reduction in HVA associated with low VMAT/DAT protein levels may imply that the observed dopamine deficiency is associated with impaired clathrin‐mediated neurotransmitter recycling. D2R is also postulated to be internalized through clathrin‐mediated endocytosis.^{38, 39} Neurons internalize receptors to adjust excitability and degrade, resensitize, and recycle desensitized receptors.^{38, 39} DNAJC6 mutations thus may affect D2R protein levels and normal postsynaptic function. It is likely that presynaptic D2R autoreceptor function will also be affected, leading to aberrant TH regulation.⁴⁰

Overall, our findings suggest that the mechanisms governing DNAJC6‐associated parkinsonism are likely to be multifactorial. Another plausible explanation for the reduction in synaptic protein levels may be as a result of neurodegeneration secondary to defective chaperone function. Auxilin and other J‐chaperone proteins play a crucial role in regulating the folding and conformational change of proteins to maintain integrity in the neuron.^{41, 42} Indeed, in the auxilin knockout mouse model, there is sequestration of clathrin cages in the cerebellum.³⁵ With impaired auxilin function, a cumulative effect of sequestered misfolded proteins and accumulation of clathrin coat components in assembled coats and cages may lead to apoptotic cascades and neurodegeneration. There is growing interest in the role of such chaperone proteins in human disease and mutations in eight distinct J proteins (DNAJB2, DNAJB6, DNAJC5, DNAJC6, DNAJC12, DNAJC13, DNAJC19, and DNAJC29) have been described.^{43, 44} Future research into such “chaperonopathies” may provide further insights into neurodegenerative disorders.

In conclusion, we report on a cohort of patients with previously unreported DNAJC6 mutations associated with early neurodevelopmental delay, juvenile parkinsonism, and neurological regression in the second decade of life. We further demonstrate disturbance of dopamine homeostasis in patient‐derived CSF and report on a possible GAK‐mediated compensatory mechanism for auxilin deficiency. Mutations in DNAJC6 are rare, but a likely under‐recognized cause of parkinsonism‐dystonia in infants and children. Elucidating the genetic diagnosis has important implications for families given that early diagnosis negates the need for extensive invasive investigations, facilitates treatment strategies, and aids genetic counseling for future pregnancies. The early clinical features and CSF neurotransmitter signature observed in our patients can mimic primary neurotransmitter disorders, and DNAJC6 mutations should thus be considered as a differential diagnosis. We observed reduced auxilin and increased GAK protein levels, suggesting a possible compensatory role for GAK in this condition. Study of CSF synaptic proteins suggest downstream effects on dopamine synthesis, recycling, homeostasis, and signaling that may result from a combination of primary auxilin deficiency and neurodegeneration. Abnormal synaptic vesicle dynamics are increasingly recognized as a disease mechanism in neurodegenerative parkinsonian disorders, and future research into elucidating the pathogenesis of such conditions will no doubt assist the development of novel targeted treatments.

Author Roles

(1) Research Project: A. Conception and Design; B. Acquisition of Data; C. Analysis and Interpretation of Data; (2) Manuscript: A. Writing of the First Draft, B. Review and Critique; (3) Other: A. Statistical Analysis (for Figure ​Figure3);3); B. Drafting of Figures; C. CSF Neurotransmitter Analysis and Interpretation.

J.N.: 1A, 1B, 1C, 2A, 3A

E.C.‐S.: 1A, 1B, 1C, 2A, 3B

L.A.: 1B, 1C, 2A, 3B

P.T.: 1B, 1C

K.M.: 1B, 1C

S.S.: 1B, 1C

K.M.G.: 1B

S.J.R.H.: 3C

S.P.: 3C

L.B.: 1B, 1C

B.C.: 1B, 1C

J.C.: 1B, 1C

K.R.: 1B, 1C

H.C.: 1B, 1C

S.J.: 1B, 1C

R.J.: 1B, 1C

D.H.: 1B, 1C

À.G.‐C.: 1B, 1C

D.G.: 1B, 1C

F.L.R.: 1B, 1C

B.P.‐D.: 1B, 1C

C.D.G.: 1B, 1C

T.S.P.: 1B, 1C

E.M.: 1A, 1B, 1C, 2A

M.A.K.: 1A, 1B, 1C, 2A

Financial Disclosures

Nothing to report.

Supporting information

Acknowledgments

Notes

Dr. Ng, Dr. Cortès‐Saladelafont, and Dr. Abela contributed equally to the manuscript.

Relevant conflicts of interest/financial disclosures: Nothing to report.

Full financial disclosures and author roles may be found in the online version of this article.

Funding agencies: This work was funded by the UK Medical Research Council Doctoral Fellowship (to J.N.), a Wellcome Intermediate Fellowship and NIHR Professorship (to M.A.K.), Sir Jules Thorn Charitable Trust (to M.A.K.), Great Ormond Street Hospital Children's Charity (to J.N., E.M., and M.A.K.), Rosetrees Trust (to J.N., M.A.K.), Agustí Pedró i Pons Foundation, Universitat de Barcelona and Río Hortega 2015‐2017 Institute of Health Carlos III (to E.C.S.), Dystonia Medical Research Foundation (to P.T.), PS09/01132 ISCIII‐FEDER, Institute of Health Carlos III (to A.G.C.), European Union Marie Curie training network (to S.J.R.H.), NIHR Biomedical Research Centres, Swiss National Science Foundation Advanced Postdoc. Mobility Fellowship (to L.A.), Cambridge Biomedical Research Centre and Wellcome Trust for UK10K (to F.L.R., D.G.). This study made use of data generated by the UK10K Project award WT091310. A full list of consortium members can be found at the UK10K Project website (http://www.uk10k.org). This research was supported by the National Institute for Health Research Biomedical Research Centre at Great Ormond Street Hospital for Children NHS Foundation Trust, University College London.

References