|Home | About | Journals | Submit | Contact Us | Français|
The proline-rich transmembrane protein (PRRT2) gene was recently identified using exome sequencing as the cause of autosomal dominant paroxysmal kinesigenic dyskinesia (PKD) with or without infantile convulsions (IC) (PKD/IC syndrome). Episodic neurologic disorders, such as epilepsy, migraine, and paroxysmal movement disorders, often coexist and are thought to have a shared channel-related etiology. To investigate further the frequency, spectrum, and phenotype of PRRT2 mutations, we analyzed this gene in 3 large series of episodic neurologic disorders with PKD/IC, episodic ataxia (EA), and hemiplegic migraine (HM).
The PRRT2 gene was sequenced in 58 family probands/sporadic individuals with PKD/IC, 182 with EA, 128 with HM, and 475 UK and 96 Asian controls.
PRRT2 genetic mutations were identified in 28 out of 58 individuals with PKD/IC (48%), 1/182 individuals with EA, and 1/128 individuals with HM. A number of loss-of-function and coding missense mutations were identified; the most common mutation found was the p.R217Pfs*8 insertion. Males were more frequently affected than females (ratio 52:32). There was a high proportion of PRRT2 mutations found in families and sporadic cases with PKD associated with migraine or HM (10 out of 28). One family had EA with HM and another large family had typical HM alone.
This work expands the phenotype of mutations in the PRRT2 gene to include the frequent occurrence of migraine and HM with PKD/IC, and the association of mutations with EA and HM and with familial HM alone. We have also extended the PRRT2 mutation type and frequency in PKD and other episodic neurologic disorders.
Paroxysmal kinesigenic dyskinesia (PKD) is the most common type of paroxysmal dyskinesia. The disorder was first reported in 1892 by Shuzo Kure1 in a 23-year-old Japanese man who had frequent movement-induced paroxysmal attacks, typical of PKD, starting from age 10 years. In 1901, Gowers2 described a similar child, and in 1967, Kertesz3 and Weber4 described families with this condition. Idiopathic or familial attacks often occur in childhood induced by sudden movements, accelerating from walking to running, or infrequently other stimuli such as sound, stress, or hyperventilation. There is often a warning of an impending attack such as numbness or paraesthesia in the affected limb or face, which then develops into pure or mixed dystonic, choreic, ballistic, or athetotic manifestations. In most patients up to 20 attacks per day occur, lasting between 30 seconds and 2 minutes.5–7
PKD is often associated with infantile convulsions (IC) or convulsions with choreoathetosis (ICCA). Attacks respond well to anticonvulsant therapy, indicating an overlap in the pathologic mechanisms that underlie these episodic disorders.6,7 In 1997, genetic linkage analysis identified a locus for IC or ICCA on chromosome 16p12-q12.8 A number of family members later developed attacks typical of PKD. Subsequently, multiple families with either PKD or ICCA syndrome from multiple populations were linked to chromosome 16, within or near the original ICCA linkage, suggesting that these disorders are allelic with variable expression, although other studies have demonstrated heterogeneity.8–17
Over the last 15 years, a considerable amount of work has been expended on the identification of the PKD/IC gene17–20 until recently, when mutations in the PRRT2 gene were identified as an important cause using exome sequencing.21–25 This is consistent with a recent report of PKD in 2 cases associated with a microdeletion of this region.26 These were mainly heterozygous loss-of-function mutations. In the homozygous state one case has been reported to cause nonsyndromic intellectual disability.27 In addition, missense PRRT2 mutations were also reported that could also be loss-of-function or consistent with a dominant-negative effect.22,23 The PRRT2 gene encodes a proline-rich transmembrane protein that is highly expressed in the CNS. Using yeast 2-hybrid screening, the PRRT2 protein has been shown to interact with the synaptosomal-associated protein 25 (SNAP25), suggesting a role in the fusion of synaptic vesicles to the plasma membrane.28
To assess the PRRT2 mutation frequency, spectrum, and associated phenotype, we analyzed this gene in 3 series of episodic neurologic disorders identifying mutations in 28 out of 58 PKD/IC, 1 out of 182 episodic ataxia (EA), and 1 out of 128 probands or sporadic cases with hemiplegic migraine (HM). There were 27 loss-of-function and 3 missense mutations identified; they were in males more frequently than females (male to female ratio 52:32). A number of families with PRRT2 mutations and PKD had associated migraine or HM, 1 family had EA with HM, and another large family had HM alone in several individuals, extending the phenotype of this disorder to involve other types of episodic neurologic conditions such as EA, migraine, and HM.
We have ethical approval for this research and patients and unaffected family members were recruited with informed consent (NHNN studies 06/N076 and 07/Q0512/26). The diagnosis of PKD or IC/ICC, migraine/HM, and EA was made using recognized criteria5,16,29–32 by the clinicians who are authors on this publication (tables 1 and 2). The EA cases are part of the channel gene UK service and had already been screened as negative for the KCNA1 and CACNA1A gene by Sanger sequencing and MLPA. The PRRT2-positive cases were also negative for these 2 genes.
DNA was extracted by a standard phenol chloroform method from blood in affected and unaffected patients. PCR was employed to analyze the coding exons of the PRRT2 gene. The longest PRRT2 transcript was used for primers design and sequencing, Genbank NM_145239, Ensemble (PRRT2 ENSG00000167371) transcript PRRT2-001–ENST00000358758. Primers were designed to exons and flanking introns of the coding exons 2 and 3 of the PRRT2 gene. Synthesized primers were from Sigma Genesis (Sigma-Aldrich Co. LLC, USA) (Prrt2_2f CCTATCTCCTCCTCTTCCAG, Prrt2_2r CTCCAGAGGCTCTATTGCAG, Prrt2_3f CTTACCCGCCATCTATGG, Prrt2_3r AGGCTCCCTTGGTCCTTAGG). PCR analysis was performed using 10 pmol of both forward and reverse genomic primers and FastStart Taq DNA polymerase (www.roche-applied-science.com). Then each purified product was sequenced using forward or reverse primers, as well as sequencing primers to sequence the middle part of exon 2 (Prrt2_2fmid AAGAGGCCACTGCAGACCAG and Prrt2_2rmid TGGTTGAAGGGCTGGCTTG) with Applied Biosystems BigDye terminator v3.3 sequencing chemistry as per the manufacturer's instructions.33 The resulting reactions were resolved on a ABI3730XL genetic analyser (Applied Biosystems, Foster City, CA) and analyzed with SeqScape v2.5 software (Gene Codes, VA). Mutations were verified in both directions, repeat sequencing carried out for specific exons in the proband and other family members to confirm and verify segregation. Mutation position was labeled from the start ATG of the PRRT2 gene according to the standard nomenclature,34,35 Genbank accession number NM_145239, Ensemble transcript PRRT2-001–ENST00000358758. None of these mutations were present on sequencing the PRRT2 gene in 475UK and 96 Asian controls. Several reported SNPs were identified and these include P75P, P138A, P216L, and C276C.
The PRRT2 gene was sequenced in 58 PKD/IC, 182 EA, and 128 HM family probands or sporadic cases (tables 1 and 2). The majority of families and sporadic individuals were of English origin but mutations were also identified in families from Malaysia, India, Wales, Somalia, Pakistan, Kenya, Poland, Malta, Austria, Philippines, Ireland, Scotland, and Australia (table 1). Families 6, 11, and 20 were used in the initial identification of the PRRT2 gene.24 Few patients had associated IC or ICC, reflecting a bias toward adult movement disorders in our department. The p.R217Pfs*8 mutation was by far the most common change, identified in 24 cases/probands from multiple ethnic origins (table 1) as previously reported.21–24 Fifteen of these were in families where the mutation segregated with the disease and was also present in 2 clinically unaffected members, suggesting incomplete penetrance. Males were more often affected with PRRT2 mutations compared to females (ratio 52:32). Additional loss-of-function heterozygous mutations were identified in one PKD family with a heterozygous p.C332insGAC, one sporadic PKD patient had a p.L171Lfs*3 mutation, and an additional family with PKD had an exon 3 mutation, c.1011C>T, that results in abnormal splicing. Three missense mutations were identified, a P215R and a P216H mutation in different sporadic PKD cases and a G305W mutation in a child with early onset PKD very responsive to carbamazepine. Although all 3 were predicted to be damaging using SIFT and Polyphen,36 these families were not large enough for segregation and the PRRT2 mutations should be considered as variants of unknown significance on this basis until further proof is obtained. None of these mutations were present on sequencing the PRRT2 gene in 475 UK and 96 Asian control individuals.
The majority of patients with PRRT2 mutations had typical PKD, but we observed a number of cases/families with associated migraine. Two PKD cases with HM were observed in the same family (table 2), a mother and child had severe migraine headaches with PKD and ipsilateral hemimotor and sensory features associated with the p.R217Pfs*8 mutation. Two cases with EA and migraine or HM were also identified. The proband with EA and HM (tables 1 and 2, F18) had the p.R217Pfs*8 mutation with episodic balance difficulties starting at the age of 18 years with unilateral headaches and hemiplegic episodes. She had frequent attacks every day of involuntary movement and balance problems, and cerebellar ataxia on examination but normal imaging. The mother had a past history of epilepsy and severe headaches, and one of her daughters had HM.
An additional family (tables 1 and 2, F30) had HM in a number of individuals in association with infantile convulsions in the proband (tables 1 and 2; figure, B) and we also identified a child with PKD, Asperger syndrome, and attention-deficit/hyperactivity disorder (ADHD), associated with the p.R217Pfs*8 mutation.
The frequency of PRRT2 mutations in ethnically diverse PKD families and sporadic cases was 28 out of 58 (48%), similar to the previous report of 50.4%.24 The PRRT2 mutation frequency in EA (1/182) and in the HM (1/128) series was low, with less than 1% of patients screened in each group. It is still possible that PRRT2 defects exist in the other 30 PKD cases, as whole gene26 or exon deletions or as deep intronic or regulating mutations in the untranslated promoter regions of the gene. No large PKD/IC families remain in our series that are negative for PRRT2.15 The previously reported Indian PKD family, linked to a different region of chromosome 16 than the PKD/IC kindreds, was found to have the p.R217Pfs*8 mutation.14
The p.R217Pfs*8 mutation was identified in a large number of cases from many different ethnic backgrounds. We identified a number of other novel loss-of-function mutations as well as missense amino acid changing mutations in the PRRT2 gene. These additional mutations were consistent with the previous reports where other types of indel and nonsense loss-of-function changes were seen, as well as damaging missense mutations. These data suggest that a mechanism of PRRT2 haploinsufficiency as well as a possible dominant-negative mutational effect is associated with PKD.
The majority of patients with PRRT2 mutations identified here and before had typical PKD (tables 1 and 2). However, in addition to infantile convulsions, the phenotype is frequently complicated by other episodic neurologic disorders such as epilepsy, migraine, or HM. In total, 12 out of 30 probands/cases had migraine type headaches; 3 of these families had HM. It is possible that these associations are coincidence, given the high incidence of migraine in the general population, although the frequency is high in these cases and migraine associated with hemiplegia is very rare. The association is not surprising as the coexistence of movement disorders, migraine, and epilepsy is often described in the neurologic channelopathies,37 indicating an overlap in etiologic pathways as well as clinical features.38
The function of PRRT2 is poorly understood. The gene is known to interact with the SNAP25 and both are highly expressed in the basal ganglia. SNAP25 is a presynaptic protein involved in synaptic vesicle release playing an important role in calcium triggered exocytosis.28 The interaction with SNAP25 and the possible disruption of neurotransmitter release associated with PRRT2 mutations is consistent with the pathogenic pathways involved in other paroxysmal movement disorders. In paroxysmal nonkinesigenic dyskinesia (PNKD), mutations in the PNKD gene are associated with disruption of synaptic protein regulated exocytosis39 and SNAP25 has also been shown to be associated with ADHD,40 which was seen in one of our PKD cases with the p.R217Pfs*8 mutation. Although further work is required, the genetic data and extension of the clinical phenotypes associated with the PRRT2 genes indicates an overlap in the pathways with other similar disorder of channel or synapse dysfunction such as epilepsy, migraine, and PNKD.
The authors thank the patients and controls for their participation in this study.
This study was designed and funding obtained by HH, KB, NWW, MGH. KB, MS, RCD, MAK, SAS, GMW, TC, SDS, EMV, LSM, AHS, HAT, SR, JWS, AL, TW, DK, NWW, MH, HH collected samples and assessed patients clinically. AG and HH conducted experiments and performed data analysis. The manuscript was written by HH with input and revision from KB, MS, RCD, MAK, TC, SDS, LSM, JWS, DK, NWW and MH.
Supported by grants from The Medical Research Council (MRC), The Wellcome Trust, The Muscular Dystrophy Campaign, Action Medical Research, The Dystonia Medical Research Foundation (DMRF), National Organisation for Rare Disorders (NORD), The Brain Research Trust (BRT), and the Italian Ministry of Health. L.S.M. was supported by a Reta Lila Weston Fellowship. S.A.S. is supported by a Robert Bosch Fellowship. This study was also supported by the National Institute for Health Research (NIHR) UCLH/UCL Comprehensive Biomedical Research Centre.
The authors report no disclosures relevant to the manuscript. Go to Neurology.org for full disclosures.