Phenotypic variability and genetic characteristics of PRRT2-associated paroxysmal movement disorders in 13 Chinese families

PRRT2-associated paroxysmal movement disorders (PRRT2-PxMDs) include paroxysmal kinesigenic dyskinesias (PKD), benign familial infantile epilepsy (BFIE), infantile convulsions and choreoathetosis (ICCA), episodic ataxia (EA), paroxysmal nonkinesiasgenic dyskinesias (PNKD), and, in addition, other childhood-onset movement disorders and different types of seizures may be caused by PRRT2 mutations, suggesting that the understanding of the spectrum of PRRT2-PxMDs is still evolving. We collected and analyzed retrospectively the clinical of children diagnosed with paroxysmal movement disorders by the Department of Neurology of Anhui Provincial Children's Hospital from January 2015 to June 2020. The genetic tests were performed in the probands and their family members. Thirteen children and their family members, 30 in total, were tested. Twenty six patients and 4 (13.34%) carriers from 13 families were identied, 14 (46.67%) were diagnosed with BIE, 7 (23.33%) with PKD, 2 (6.67%) with ICCA, 1 (3.33%) with epilepsy (focal), 1 (3.33%) with infantile spasm (IS), and 1 (3.33%)was diagnosed with PKD and PNKD, Eight different variants were identied in 13 families, and NM_145239.2:c.640-641insC was found in 4 families while recurrent mutation c.649dupC was not found. Three novel mutation, c.884G > C, c.865G > C, and c.-65-1G > C were identied in this study. This study conrmed that there is no clear genotype-phenotype correlation in patients with PRRT2-PxMDs, in addition, the clinical ndings show variable phenotype within families, including the families affected due to the newly identied pathogenic variants in this study.

Most PRRT2-PxMDs are caused by heterozygous pathogenic variants in the PRRT2 gene or, rarely, the 16p11.2, including PRRT2, recurrent deletion are inherited in an autosomal dominant manner, and approximately 10% of pathogenic variations are de novo and 90% are inherited. However, high phenotypic variability is common in affected family members due to reduced penetrance and variable expressivity. Besides, in some patients typically with a more severe phenotype, PRRT2-PxMD is inherited in an autosomal recessive manner.

Subjects
We collected the data of clinical feature and genetic testing of patients diagnosed with paroxysmal movement disorders in the Department of Neurology of Anhui Provincial Children's Hospital from January 2015 to June 2020.
The diagnostic criteria for BIE were (Vigevano 2005): a) onset age between 3 and 12 months; b) normal development of intelligence and movement before and after the onset of the disease; c) focal attack or focal secondary generalized attack, sometimes as cluster attacks (i.e. more than 2 episodes within 24 hours); d) normal EEG background. Epileptiform discharges were found in EEG during paroxysms, but no epileptiform discharge was found in interictal periods; e) no abnormality in cranial imaging; and f) exclusion of convulsions caused by metabolic disorders such as hypocalcemia and hypoglycemia.
The diagnostic criteria for PKD were (Bruno et a1.2004): muscular dystonia or choreographic athetosis, induced by sudden movement and lasting for several seconds to 1 minute; no disturbance of consciousness; and exclusion of other organic diseases. This study was approved by the medical ethics committee of Anhui Children's Hospital. All subjects were fully aware of the purpose and content of the study, and written informed consent was obtained from parents and/or guardians of all the subjects before enrolling.
The diagnosis of other types of PRRT2-PxMDs was depended on the results of genetic testing.

Molecular genetic analysis
This study was performed with the Anhui Provincial Children's Hospital. Peripheral blood samples were obtained from the patient upon receiving written informed consent from the parents or guardians, and whole-exome sequencing (WES), as described previously (Tong et a1. 2018), were performed by Chigene (Beijing, China) in the patients and their family members. xGen Exome Research Panel v1.0(IDT, Iowa, USA) was used to construct whole exome library and Next-generation sequencing (NGS) was performed on Illumina Novaseq 6000 platform (Illumina, Santiago, USA). Quantitative PCR or Sanger sequencing were used to con rm the candidate variation.

Variants analysis
Raw data cleaning was following the standard instruction recommended by Illumina, followed by paired-end alignment using Burrows-Wheeler Aligner (BWA) and variants calling by using software package GATK, and SNPs and short, < 50bp, indels were screened. Using the online analysis platform provided by Chigene (https://chigene.org), variation were annotated by using multiple databases including the 1000 genomes project database, dbSNP, ESP, ExAC, OMIM, Orphanet, and etc., and the signi cance of the pathogenicity of variations was predicted using tools SIFT (http://sift.jcvi.org/), PolyPhen-2 3. Results

Clinical features
Patients and their family members, 30 sample in total, from 13 unrelated families were analyzed, and 26 patients were diagnosed with PRRT2-PxMDs (Table  1), and the pedigrees were showed in Figure 1.
Patient 12, from family 6, was a boy diagnosed at age 6 months and 10 days, the only child of an unrelated couple. He was born in a natural childbirth after 36 weeks plus 5 days of labor without history of asphyxia. His father had a history of BIE and PKD, and his aunt had a history of PKD. He was found having nodding and embracing seizures in clusters, 6-8 times/cluster, lasting 1-2 minutes, and 3-5 clusters/day, at age 5 months and 26 days, most of which occurred when he was just awake or sleepy. The patient had normal head MRI, genetic metabolic blood screening and urine screening, electrocardiogram and heart color ultrasound tests. Video electroencephalogram(VEEG) showed extensive medium-high amplitude slow wave bursts for about 1 second, accompanied by an electromyography(EMG) burst during the attacks of nodding and hug movements (Figure 2A) , concurrent with. After treatment with ACTH(2.0IU/kg/d, two weeks) and sodium valproate (VPA) for (20mg/kg/d, one mouth), his seizure stopped at the age of 8 months and the EEG returned to normal ( Figure 2B). At one year of follow-up, the patient had no seizures or neuromotor developmental delay.
The summarization and distribution of 26 patients with PRRT2-PxMDs were showed in Table 2.

Genetic analyses
Eight PRRT2 variations were found in 30 of the 49 tested members, including 4 carriers without phenotype , and the recurrent variant c.640-641insC was identi ed in 10 patients from 5 families, which was previously reported correlating with PKD or BFIS (PMID:22623405;22744660) Two of 3 novel variants, c.884G>C, c.865G>C, except c.-65-1 G>C, are VUS (variation of uncertain signi cance). (Table 3) . All variations were con rmed by using quantitative PCR or Sanger sequencing.

Discussion
Precisely describing PRRT2-associated paroxysmal movement disorders (PRRT2-PxMDs) is still a challenge ( Ebrahimi-Fakhari et al. 2015). Apart from the core phenotypes, kinesigenic dyskinesia (PKD), benign familial infantile epilepsy (BFIE), paroxysmal kinesigenic dyskinesia with infantile spasm (PKD/IS), and infantile convulsions and choreoathetosis (ICCA) of PRRT2-PxMDs, PRRT2 pathogenic variants have been identi ed in patients diagnosed with several childhood-onset movement disorders and seizures, including paroxysmal nonkinesigenic dyskinesia (PNKD) ( Delcourt et al.2015 ;, paroxysmal exertion-induced dyskinesia ( Liu et al,2012), and episodic ataxia ( Delcourt et al.2015). Patients diagnosed with PKD who have heterozygous PRRT2 deletion (16p11.2 deletion) tend to have additional disorders including developmental delay (DD), intellectual disability (ID), and/or autism spectrum disorder (ASD) ( Termsarasab et al,2014;Weber et al,2013;Silveira-Moriyama et al ,2013), and individuals with biallelic PRRT2 mutations usually have a more severe phenotype including ID, and seizures. ( Delcourt et al.2015) However, there is no comprehensive evidence for genotype-phenotype correlations in core phenotypes of PRRT2-PxMDs and phenotypic variation is common within or between families affected due to the same PRRT2 mutation. In our study, BIE and PKD are usually diagnosed in the same family, in siblings (in family 1), actually, indicating reduced penetration.
The PRRT2 gene is located in 16p11.2, containing 4 exons with a total length of 3794bp and encodes 340 amino acids. Protein PRRT2 is mainly expressed in cerebral cortex, basal ganglia and cerebellum, and, it is's functioning transferred from cortex to basal ganglia and cerebellum during the development of nervous system. Age-depend decreasing levels of PRRT2 in the nervous system may explain the age-dependence of BFIE and PKD. We identi ed no c.649dupC, the recurrent mutation reported previously in most of patients ( Ono et al,2012;Méneret et al,2012;Lee et al,2012), but the truncating mutation c.640-c.641insC (c.640dupC) in 4 families (family 1, 4, 5, and 7) instead, however, the diagnosis ranges from nonsyndromic BFIE to relatively severe PKD in these families, not to mention PKD due to the novel mutation p.A289P we identi ed in family 8, for example, which was supposed to cause a milder phenotype.
Infantile spasm (IS) is usually considered a serious epileptic encephalopathy, most of the seizures are refractory, and patients with IS are at the high risk of having ID. In this study, patient 12, from family 6, was diagnosed with IS whowithout history of hypoxia, as previously reported, who had normal development until the last follow-up, while his father was diagnosed with ICCA.After taking anti-epilepsy treatment and one year of follow-up, the seizures ceased and his mental motor development was normal.

Conclusion
PRRT2-PxMDs include PKD, BFIE, ICCA, EA, PNKD, FHM, PED, benign spasmodic torticollis, and other diseases, and genotype-phenotype correlation is still unclear, which was proved in this study. We found that c.640_c.641insC occurred in most of our patients, instead of the reccurrent insertion at c.649, and, in addition, we identi ed novel variants c.884G>C, c.865G>C, and c.-65-1G>C that cause variable phenotype within the families. In short, our research expanded the spectrum of pathogenic variation of PRRT2, however, the clinical ndings strengthened the point that PRRT2-PxMDs, in the context of dominant inherited disease, have no genotype-phenotype correlation.

6.Ethics approval and consent to participate
This study was approved by the medical ethics committee of Anhui Provincial Children's Hospital. The guardians of all subjects fully understand the purpose and content of the study, and obtain written informed consent from the guardians of all subjects before participating in the study.

7.Consent for publication
The manuscript has been approved by all authors for publication.

8.Availability of data and materials
The datasets used during the current study are available from the corresponding author on reasonable request.

9.Con icts of interest
None of the authors has any con ict of interest to disclose.

10.Fund
This research did not receive any speci c grant from funding agencies in public, commercial, or not-for-pro t sectors.

11.Authors' contributions
Lulu Wu: Data collection and article writing; Xinyu Yang: Data collection and article writing; Xiaocui Wang: Guiding clinical data collection; Shuang Yu: Participating in data collection; Bin Yang: Guiding research process and modi ng articles

12.Acknowledgements
Thanks to Tianchen Xu for his guidance on English grammar. Thanks to Beijing Zhiyin Oriental Translational Medicine Research Center for providing technical guidance.