Article Text

Original article
A type I interferon signature identifies bilateral striatal necrosis due to mutations in ADAR1
  1. John H Livingston1,
  2. Jean-Pierre Lin2,
  3. Russell C Dale3,
  4. Deepak Gill3,
  5. Paul Brogan4,5,
  6. Arnold Munnich6,
  7. Manju A Kurian7,8,
  8. Victoria Gonzalez-Martinez9,
  9. Christian G E L De Goede10,
  10. Alastair Falconer11,
  11. Gabriella Forte12,
  12. Emma M Jenkinson12,
  13. Paul R Kasher12,
  14. Marcin Szynkiewicz12,
  15. Gillian I Rice12,
  16. Yanick J Crow12,13
  1. 1Department of Paediatric Neurology, Leeds General Infirmary, Leeds, UK
  2. 2General Neurology & Complex Motor Disorders Service, Evelina Children's Neurosciences Centre, Evelina London Children's Hospital, Guy's and St Thomas' NHS Foundation Trust, London, UK
  3. 3TY Nelson Department of Neurology, and Neuroimmunology group, The Children's Hospital at Westmead, University of Sydney, Sydney, New South Wales, Australia
  4. 4Department of Paediatric Rheumatology, University College London, Institute of Child Health, London, UK
  5. 5Great Ormond Street Hospital NHS Foundation Trust, London, UK
  6. 6Departement de Génétique, INSERM U781, Université Paris Descartes-Sorbonne Paris Cité, Institut Imagine, Hôpital Necker-Enfants Malades (AP-HP), Paris, France
  7. 7Department of Neurology, Great Ormond Street Hospital NHS Foundation Trust, London, UK
  8. 8Neurosciences Unit, Institute of Child Health, University College London, London, UK
  9. 9Department of Neurosurgery, Hôpital Gui de Chauliac, University Hospital of Montpellier, Montpellier, France
  10. 10Department of Paediatric Neurology, Royal Preston Hospital, Preston, UK
  11. 11Department of Paediatrics, Scarborough General Hospital, Scarborough, UK
  12. 12Department of Genetic Medicine, Faculty of Medical and Human Sciences, Institute of Human Development, University of Manchester, Manchester, UK
  13. 13Manchester Academic Health Science Centre, Central Manchester University Hospitals NHS Foundation Trust, St. Mary's Hospital, Manchester, UK
  1. Correspondence to Professor Y J Crow, Department of Genetic Medicine, University of Manchester, St Mary's Hospital, 6th Floor, Oxford Road, Manchester M13 9WL, UK; yanickcrow{at}mac.com

Abstract

Background We recently observed mutations in ADAR1 to cause a phenotype of bilateral striatal necrosis (BSN) in a child with the type I interferonopathy Aicardi-Goutières syndrome (AGS). We therefore decided to screen patients with apparently non-syndromic BSN for ADAR1 mutations, and for an upregulation of interferon-stimulated genes (ISGs).

Methods We performed Sanger sequencing of ADAR1 in a series of patients with BSN presenting to us during our routine clinical practice. We then undertook detailed clinical and neuroradiological phenotyping in nine mutation-positive children. We also measured the expression of ISGs in peripheral blood from these patients, and in children with BSN who did not have ADAR1 mutations.

Results Nine ADAR1 mutation-positive patients from seven families demonstrated an acute (five cases) or subacute (four cases) onset of refractory, four-limb dystonia starting between 8 months and 5 years of age. Eight patients were developmentally normal at initial presentation. In seven cases, the disease was inherited as an autosomal recessive trait, while two related patients were found to have a heterozygous (dominant) ADAR1 mutation. All seven mutation-positive patients assayed showed an upregulation of ISGs (median: 12.50, IQR: 6.43–36.36) compared to controls (median: 0.93, IQR: 0.57–1.30), a so-called interferon signature, present many years after disease onset. No interferon signature was present in four children with BSN negative for mutations in ADAR1 (median: 0.63, IQR: 0.47–1.10).

Conclusions ADAR1-related disease should be considered in the differential diagnosis of apparently non-syndromic BSN with severe dystonia of varying evolution. The finding of an interferon signature provides a useful screening test for the presence of ADAR1 mutations in this context, and may suggest novel treatment approaches.

  • Clinical genetics
  • Immunology (including allergy)
  • Molecular genetics
  • Neurology
View Full Text

Statistics from Altmetric.com

Introduction

We have previously suggested the grouping of Mendelian disorders in which an upregulation of type I interferons is apparently central to their pathogenesis.1 In this context, we recently identified mutations in ADAR1 to cause a radiological phenotype of bilateral striatal necrosis (BSN) in a child with the type I interferonopathy Aicardi-Goutières syndrome (AGS).2 We therefore decided to screen patients with apparently non-syndromic BSN for ADAR1 mutations.

BSN refers to the acute or subacute onset of a dystonic/rigid movement disorder associated with radiological evidence of symmetrical abnormalities in the corpus striatum and sometimes globus pallidus.3 These are apparent on CT as low density, and on MRI as high signal on T2 weighted and FLAIR sequences, and low signal on T1 weighted images. Swelling of the involved structures is often apparent in the acute stages, later followed by shrinkage and persisting signal change.

Onset can occur at any time from the neonatal period through to adolescence. The radiological phenotype is typically associated with developmental regression, dystonia, choreoathetosis, spastic quadriparesis and, in some cases, death.

Isolated BSN is most frequently attributed to the sequelae of infection, for example, due to mycoplasma, measles or streptococcus. The pathogenic basis of this association is not well understood, although some of these cases may be autoimmune mediated.4 Familial infantile striatal degeneration can be inherited as an autosomal recessive or mitochondrial trait, with recognised genetic causes due to mutations in the mitochondrial genes ATP6,5 ND1,6 ND67 and NDUFV1,8 as well as NUP62 (encoding a nuclear pore complex protein),9 SLC19A310 and SLC25A1911 (encoding thiamine transporter proteins), and the acyl dehydrogenase encoding gene GCDH.12

Here, we describe eight patients with previously unexplained BSN due to mutations in ADAR1. We also report a ninth mutation-positive patient, a relative to one of our index cases, with a severe dystonia phenotype for whom neuroimaging was not available. ADAR1-related BSN/dystonia is associated with the presence of an interferon signature, which likely represents a reliable screening test for the condition, and highlights possible approaches to the treatment of this devastating disease.

Methods

A total of nine ADAR1 mutation-positive patients were identified through our own clinical practice. A further four mutation-negative patients with previously unexplained dystonia were also ascertained. Twelve of these 13 patients had clinical and radiological features of BSN. Patient 8 had a severe dystonia phenotype, but no extant cranial MR and/or CT data. Clinical and radiological data were collated, and the cranial imaging reviewed together by two co-authors (JHL and YJC).

For Sanger sequencing, primers were designed to amplify the coding exons of ADAR1 (see online supplementary table S1). Purified PCR amplification products were sequenced using BigDye terminator chemistry and an ABI PRISM 3730 xL Genetic Analyzer (96-capillary system). Mutation description was based on the reference cDNA sequence NM_001111.4, with nucleotide numbering beginning from the first A in the initiating ATG codon.

The expression of six genes known to be interferon-stimulated was assessed in whole blood. Total RNA was extracted from whole blood using a PAXgene (PreAnalytix) RNA isolation kit. RNA concentration was assessed using a spectrophotometer (FLUOstar Omega, Labtech). Quantitative reverse transcription PCR analysis was performed using the TaqMan Universal PCR Master Mix (Applied Biosystems), and cDNA derived from 40 ng total RNA. The relative abundance of target transcripts, measured using TaqMan probes for IFI27 (Hs01086370_m1), IFI44L (Hs00199115_m1), IFIT1 (Hs00356631_g1), ISG15 (Hs00192713_m1), RSAD2 (Hs01057264_m1), and SIGLEC1 (Hs00988063_m1), was normalised to the expression level of HPRT1 (Hs03929096_g1) and 18 s (Hs999999001_s1), and assessed with the Applied Biosystems StepOne Software V2.1 and Applied Biosystems Data Assist Software V.3.01.

Patient data were expressed relative to the average of 29 normal controls. The median fold change of the six ISGs, when compared to the median of the 29 healthy controls, was used to create a score for each patient. The mean interferon score of the controls plus two SDs above the mean (+2 SD) was calculated. Scores above this value (>2·466) were designated as positive.

Clinical information and samples were obtained with informed consent.

Results

Clinico-radiological details of nine patients with mutations in ADAR1 are described below (see online supplementary table S2).

Five children presented, at 9 months (patient 1), 9 months (patient 2), 10 months (patient 3), 11 months (patient 4), 16 months (patient 5) and 18 months (patient 6) of age, respectively, with an acute onset of severe generalised dystonia in the context of a febrile illness. Patient 1, and patients 3–6 were developmentally completely normal prior to the onset of their disease.

In patient 1, neurological disease followed 2 weeks after demonstrating typical skin lesions of chicken pox. Specifically, this child developed a tremor and stiffening of the left arm, within 7 days the right arm was also involved, and within a further 2 weeks she demonstrated a severe generalised rigid/dystonic movement disorder, and had lost all her previous motor skills. On a background of mild to moderate global developmental delay, after a 1-week episode of possible bronchiolitis at age 9 months, patient 2 experienced acute onset and rapid progression to severe four-limb dystonia, necessitating emergency management, over a period of 7 days. Patients 3 and 4, full brothers to one another, demonstrated a remarkably stereotyped presentation. Both experienced a febrile illness (following an ear infection in patient 3, and a chest infection in patient 4), after which there was a rapid loss over a 2–4-week period, of previously acquired developmental skills (at the time of onset of their disease both children were cruising furniture and starting to say single words). Patient 3 is now severely affected with profound four-limb dystonia. His brother is more mildly affected, being left with some useful residual hand function, an ability to communicate through sign language and gesture, and having developed full continence. Following a diarrhoeal illness, over a 2-week period, age 16 months, patient 5 stopped walking, lost the use of one, then the other arm, and then developed a generalised increase in tone and rigidity. At the age of 18 months, patient 6 developed progressive motor deterioration 2 weeks following a febrile illness associated with a rash and red eyes. Over a 4-week period she developed a generalised four-limb dystonia. Her condition stabilised after 6 weeks and, at the time of writing, 4 months following diagnosis, there has been some clinical improvement.

After previously normal development, two further children (patients 7 and 9) developed a severe, progressive dystonia over a period of a few months, while patient 8 showed a gradual deterioration over a several-year interval.

Patient 7 presented at the age of 5 years with a limp following a fall. Over the next 4 months she developed tremor of all four limbs and rigidity of her legs. Patient 8, a paternal half-sister to patient 7, experienced a slowly progressive motor difficulty beginning at age 1 year, worsening to the point of admission for investigations at 18 months of age. Deterioration was gradual but relentless thereafter, so that she was toe walking at age 3 years, required support to walk by 7 years of age, and was confined to a powered wheel chair by age 12 years. Patient 9 became unwell at 14 months of age when he lost his balance and coordination and kept falling to the left side. Over a week, he became unable to pick himself up, and within 3 months he was unable to sit. Over a period of 1 week, he then made what was thought to be a full recovery, following which he was up and walking. However, 3 days later he experienced a rapidly progressive deterioration over a period of a few weeks, resulting in a severe four-limb dystonia.

Three of the patients have been treated with the insertion of bilateral globus pallidus internus deep brain stimulators, and one had an intrathecal baclofen pump implant fitted for medically refractory dystonia. Patient 1 developed vasculitic skin lesions on the feet after 2 years of age. None of the other children demonstrated any chilblain-like lesions. In retrospect, patients 3, 7 and 8 were noted to have cutaneous pigmentation with freckling of the face and hands.

Available cranial MRI from seven ADAR1 mutation-positive patients showed the features of BSN with symmetrical signal change in the caudate and putamen, often associated with swelling and later shrinkage (figure 1 and see online supplementary figure S1). In patient 2 at the time of presentation, imaging showed moderate cerebral atrophy and high signal in the cerebral white matter. Patient 6 also showed patchy high signal on T2, and FLAIR sequences in the deep cerebral white matter. In all other patients there were no additional MR abnormalities. In patient 1, CT scan at the time of presentation aged 9 months demonstrated symmetrical calcification in the putamen and caudate (figure 2). Cranial CT imaging of patient 3 at the age of 1 year was normal, while at age 13 years a subtle blush of high signal in the basal ganglia was observed, which was thought to represent calcification (see online supplementary figure S1). Central pontine calcification was observed on CT in patient 5 at age 5 years. This had not been present previously. Patient 7 also showed bilateral basal ganglia calcification at presentation aged 5 years. Where available, follow-up imaging demonstrated shrinkage of the putamen with persisting signal change, and the development of mild to moderate generalised cerebral atrophy in 4 of 5 patients. Patient 2 showed mild diffusely abnormal high signal in the cerebral white matter, while patients 1 and 9 showed patchy non-specific high signal in the deep cerebral white matter.

Figure 1

Representative MR images from ADAR1 mutation-positive patients. T2 axial MR images at time of presentation (top row) and on follow-up at variable times after presentation (bottom row). Top row: (A) Patient 1 aged 9 months, symmetrical high signal and swelling of putamen and caudate. (B) Patient 2 aged 9 months. Global cerebral atrophy with symmetrical high signal in putamen and caudate. Slight abnormal high signal in cerebral white matter. (C) Patient 5 aged 16 months. Symmetrical high signal in putamen and caudate. (D) Patient 7 aged 5 years. Slight cerebral atrophy and mild swelling of putamen and caudate. (E) Patient 9 aged 14 months. Symmetrical high signal and swelling in putamen and caudate. (F) Patient 6 at 19 months showing bilateral putaminal high signal and some patchy signal change within the deep frontal white matter. Bottom row: (A) Patient 1 at 3 years of age showing persistence of high signal and slight shrinkage of putamen. (B) Patient 2 at 2 years 8 months showing progression of atrophy and shrinkage of putamen and caudate. (C) Patient 5 at 5 years showing mild cerebral atrophy and persisting high signal and shrinkage of putamen. (D) Patient 7 at 8 years showing mild cerebral atrophy and subtle signal change in globus pallidus. (E) Patient 9 at 5 years showing high signal and shrinkage in putamen.

Figure 2

Representative CT images from three ADAR1 mutation-positive patients demonstrating intracranial calcification. (A) Patient 1 aged 9 months. Bilateral calcification in putamen and head of caudate. (B) Patient 5 aged 5 years. Calcification in central pons. (C) Patient 7 aged 5 years. Bilateral calcification in globus pallidus and putamen.

Extensive investigations were undertaken, including muscle biopsy and respiratory chain enzyme analysis (five patients), cerebrospinal fluid (CSF) analysis (seven patients), and liver biopsy (one patient). All results were essentially normal, except that patient 2 was noted to have a slightly raised level of CSF neopterin indicative of an inflammatory process. Of note, CSF neopterin analysis in patients 3 and 4 was normal during the first few months of presentation.

Seven individuals, patients 1–6 and patient 9, were shown to carry biallelic ADAR1 variants (table 1). Patients 7 and 8, paternal half-sisters to one another through an apparently asymptomatic father, were found to have a c.3019G>A transition (p.G1007R) in the heterozygous state. These variants were considered to be pathogenic on the basis of species conservation (see online supplementary figure S2) and the output of pathogenicity prediction packages (see online supplementary table S3). Where samples were available, all parents tested were heterozygous for one putative mutation. Of further note, both the p.P193A and the p.G1007R substitutions have been reported in patients with a clinical diagnosis of AGS. The p.P193A mutation is recorded in 41 subjects (32 of 8568 European–Americans, and 9 of 4397 African–Americans) annotated on the Exome Variant Server database, while none of the other ADAR1 mutations present in the patients reported here were recorded in more than 13 000 control alleles.

Table 1

Ancestry, family members tested, consanguinity status and sequence alterations in ADAR1 mutation-positive patients with bilateral striatal necrosis/dystonia

Four patients with dystonia, all of whom had typical radiological features of striatal necrosis, had no ADAR1 mutations (see online supplementary table S5 and supplementary figure S3).

All seven ADAR1 mutation-positive patients assayed, including patients 7 and 8 harbouring a heterozygous p.G1007R mutant allele, demonstrated an upregulation of ISGs (median: 12.50, IQR: 6.43–36.36), when assayed between 1 and 24 years after the onset of neurological features, compared with controls (median: 0.93, IQR: 0.57–1.30). No interferon signature was present in four children with BSN who were negative for mutations in ADAR1 (median: 0.63, IQR: 0.47–1.10) (figure 3 and see online supplementary table S4).

Figure 3

(A) Quantitative reverse transcription PCR (qPCR) of a panel of six interferon stimulatory genes (ISGs) in whole blood measured in ADAR1 mutation-positive (1–3, and 5–8) and mutation-negative (10–13) patients, with ages in years at time of testing stated in brackets. Bars show RQ values (error bars are relative quantification (RQ) min and RQ max) for patients as compared with healthy control. RQ is equal to 2−ΔΔCt, with -ΔΔCt±SDs, that is, the normalised fold change relative to a calibrator. (B) Interferon scores calculated from the median fold change in RQ values for the panel of six ISGs.

Discussion

BSN refers to the acute or subacute onset of a dystonic movement disorder associated with radiological abnormalities comprising symmetrical signal change and often swelling in the corpus striatum (caudate and putamen), and less commonly globus pallidus, on MRI. The prognosis for BSN is variable, with some patients making a complete recovery, whereas others develop a severe dystonia or a more akinetic-rigid phenotype.

CSF pleocytosis is common in BSN, and in some patients evidence of infection can be found, for example, with mycloplasma or streptococcus. Other cases may be caused by acute decompensation of a metabolic state such as related to mitochondrial disease, or an organic aciduria. However, in most patients, the aetiology of BSN remains unknown. The prevalence of sporadic BSN has been estimated at 1–9/1 000 000, while the familial form of the disease has been estimated at less than 1/1 000 000.13 It should be noted, that ‘sporadic’ instances of a disease may have a genetic basis, so that the use of this term is not necessarily helpful.

We have identified mutations in the gene encoding the editing protein adenosine deaminase acting on RNA (ADAR1) in eight children with radiologically defined BSN, and in a ninth patient, a paternal half-sister to patient 7, with severe dystonia whose imaging status is unknown. The characteristics of the phenotype are: acute or subacute onset of striatal necrosis, which may occur in the context of an identified infectious trigger; either previously normal development or prior developmental delay; and the evolution of severe and progressive dystonia associated with extreme loss of function and relative cognitive preservation. Clinically, these patients were not obviously different from many other children presenting with BSN.

All eight ADAR1 mutation-positive cases with cranial scan data demonstrated features of BSN. As a possible clue to the diagnosis, three children had calcification in the basal ganglia, and a further child showed a single focus of calcification in the pons. The fact that calcification was not present at the time of the initial presentation in either patient 3 or patient 5 indicates that calcification is not mandatory for the diagnosis (we note that not all patients in our series underwent cranial CT scanning). Mild white matter changes were seen in four patients, and most patients developed mild progressive cerebral atrophy.

The prevalence of ADAR1-related disease in patients with BSN is unknown. Our study was subject to ascertainment bias in that three children were selected for testing because of the presence of intracranial calcification (albeit minimal and pontine in patient 5—see above), while a fourth child, patient 2, was noted to have a slightly raised level of CSF neopterin indicative of an inflammatory process.14 However, our identification of nine cases, seven of whom were previously considered to have ‘idiopathic’ BSN/dystonia, suggests that ADAR1 mutations may represent an important cause of this severe phenotype.

ADAR1 was recently identified as a sixth gene, mutations in any of which may cause AGS.2 Patient 1 was initially ascertained as a possible ‘atypical case’ of AGS on the basis of the presence of basal ganglia calcification, and the development of vasculitic, chilblain-like, lesions on the feet, which became manifest more than 1 year after the onset of her neurological disease. None of the other patients described here have the ‘typical’ clinical15 or radiological16 phenotype of AGS and, until this time, they were all considered as cases of non-syndromic BSN/severe dystonia. Presentation with BSN has not been described in AGS patients with mutations in the other AGS-related genes (TREX1, RNASEH2A/B/C, SAMHD1),17 possibly indicating a particular susceptibility to striatal necrosis in ADAR1 mutation-positive patients.

More than 130 different ADAR1 mutations have been documented in patients with dyschromatosis symmetrica hereditaria 1 (DSH), an autosomal-dominant disorder characterised by the childhood onset of hypopigmented and hyperpigmented macules on the face and dorsal aspects of the extremities.18 Uniquely, the same heterozygous p.G1007R mutation seen in patients 7 and 8 was previously described in two patients with DSH, demonstrating neurodegeneration with dystonia and intracranial calcification. Specifically, Tojo et al19 described a female patient presenting at age 17 years with gait disturbance and dystonic posturing of the legs, who experienced intellectual deterioration from age 21 years, and became wheelchair bound by 22 years of age. Meanwhile, Kondo et al20 reported a male child presenting with loss of intellectual skills and axial torsion dystonia beginning at age 3 years. Of note, the mother of this child carried the same ADAR1 mutation, but did not demonstrate a neurological phenotype (at least into young adulthood), thus highlighting the existence of remarkable age-related penetrance/non-penetrance. Similarly, our patients 7 and 8 are half-sisters related through a father who, when last seen in his fifth decade, was asymptomatic. As shown previously in the context of AGS, the p.G1007R substitution likely confers a dominant-negative effect specific to this particular mutation.2

The skin lesions typical of DSH were not recognised initially in any of our patients. In retrospect, three children are considered to demonstrate cutaneous hyperpigmentation in the form of freckling of the hands and face. However, we find it hard to differentiate this sign from ‘normal’ freckling, so that we are not certain of its clinical utility. DSH has only very rarely been reported outside Japan and China. Moreover, even within known families segregating the DSH phenotype, a marked variability in expression is well recognised.21

Adenosine deaminases acting on RNA (ADARs) catalyse the deamination of adenine to inosine in dsRNA, and are thought to be involved in modifying the content and structure of cellular RNAs.22 Loss of editing, or non-editing, ADAR1 activity may lead to an increase in immunoreactive dsRNA, which might, in turn, induce type 1 interferon production.23 ,24 It is speculated that ADAR1 has a role in the metabolism of endogenous retroelements, as has been demonstrated for TREX1, and suggested for SAMHD1 and RNase H2.1 Why a loss of ADAR1 function should predispose to the development of BSN is unclear. ADAR1 is expressed throughout the brain including the basal ganglia,25 and it has been previously shown that a loss of ADAR1 makes cells more susceptible to apoptosis following stress, including infection.26 ,27 Further work is necessary to explore these possibilities in the context of the BSN phenotype described here.

The demonstration of an interferon signature in patients with ADAR1-related BSN is important for two reasons. First, it may be possible to use this blood marker as a screening test for BSN due to ADAR1 mutations. We emphasise that the interferon signature seen in six of the seven cases assayed here was recorded many years (24 years in the case of patient 8) after disease onset, that is, it was not related to an infectious agent (although we do not rule out the possibility that infection might be responsible for triggering an initial decompensation). Furthermore, the absence of a similar signature in four other children with BSN and no mutations in ADAR1 suggests that this feature may be specific to ADAR1-related disease.

Second, this observation points to a novel approach to treatment which, perhaps if given early enough, might prevent the very severe dystonia that is a characteristic of all the patients whom we describe. Specifically, it has recently been demonstrated that treatment of TREX1-null mice with antiretroviral therapy can prevent the development of the associated lethal inflammatory disorder.28 A pilot study of antiretroviral treatment in AGS patients is currently under ethical review. We speculate that such treatments might also be relevant in ADAR1-associated BSN, and that an interferon signature may provide a biomarker by which to monitor treatment efficacy.

In conclusion, our findings indicate that ADAR1 mutations should be sought in the context of BSN and severe dystonia, irrespective of whether an associated infection has been identified. The observation of an interferon signature likely provides a useful clinical marker of ADAR1-related disease, and may suggest novel therapeutic approaches as are being considered in other type I interferonopathies.

Acknowledgments

We sincerely thank the patients and their families included in this research. We also thank all other clinicians who have contributed samples and clinical data to our studies, particularly Cynthia Sharpe, Starship Children's Hospital, Auckland, New Zealand. YJC acknowledges the Manchester Biomedical Research Centre and the Greater Manchester Comprehensive Local Research Network. We also thank the Guy's and St Thomas’ Charity Complex Motor Disorders Service.

References

View Abstract

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:

Footnotes

  • Contributors JHL and YJC collated and reviewed all clinical and radiological data. GIR performed quantitative PCR analysis. GMAF, EMJ, PRK and MS performed all other experimental work. J-PL, RCD, DG, PB, AM, VG-M, MK, CGELDeG and AF provided clinical samples and critically reviewed clinical and immunological patient data. JHL, J-PL and YJC conceived the study, planned, designed and interpreted experiments, and wrote the initial draft with the assistance of GIR and PRK. All authors critically reviewed the manuscript and agreed to its publication.

  • Funding European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement 241779. European Research Council (GA 309449: Fellowship to YJC). Great Ormond Street Hospital Children's Charity (2011-NAT-02). Guy's and St Thomas’ Charity New Services and Innovation Grant (Project Number G060708). The Dystonia Society UK.

  • Competing interests None.

  • Ethics approval The study was approved by the Leeds (East) Research Ethics Committee (reference number 10/H1307/132).

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.