Exploring epileptic phenotypes in PRRT2-related disorders: a report of two cases and literature appraisal.

Background: The proline-rich transmembrane protein 2 (PRRT2) is a synaptic protein involved in neurotransmitter vesicle release. PRRT2 protein is highly expressed in the cerebellum, cerebral cortex, basal ganglia, and hippocampus. Variants in PRRT2 have been identified as a cause of several neurological disorders, including epilepsy, movement disorders, and headache. Methods: We report two families carrying two distinct PRRT2 mutations showing childhood onset of movement disorders, headache, and epilepsy. Demographics, clinical, EEG, neuroimaging, and genetic sequencing study data were collected. A systematic review of the literature was also performed to dissect the most frequently reported PRRT2-associated epileptic phenotypes. Results: two variants in PRRT2 gene (NM_145239.3:c718C>T, p.Arg240Ter; c.649dupC, p.Arg217Profs*8) were identified. The two variants altered the same extracellular domain of PRRT2 . The de novo PRRT2 mutation (c718C>T, p.Arg240Ter) was related to multi-drug-resistant epilepsy. According to the literature, homozygous, biallelic variants and 16p11.2 deletions lead to PRRT2 haploinsufficiency and a more severe phenotype. Conclusions: PRRT2 mutations can be associated with several epileptic phenotypes ranging from benign ASM-responsive form to more severe epileptic encephalopathies. Identifying PRRT2 variants in epilepsy patients may help achieve more personalized treatment approaches. However, phenotype-genotype correlations remain a challenge.


Introduction
The proline-rich transmembrane protein 2 (PRRT2) is a synaptic protein highly expressed in the central nervous system.It comprises an N-terminal proline-rich extracellular domain and two C-terminal transmembrane domains, which are highly conserved and play a pivotal role in the proteins' function (Figure 1) [1,2].At the cellular level, PRRT2 is mainly detected in presynaptic terminals of glutamatergic neurons in the cerebral cortex, basal ganglia, and cerebellum.The protein is involved in Ca2+-mediated neurotransmitter release through the interaction with 25 kDa Synaptosomal-Associated Protein (SNAP- 25) and with other synaptic proteins such as the Vesicle Associated Membrane Protein 2 (VAMP2) and the synaptotagmins Syt1 and 2. PRRT2 negatively regulates sodium Nav1.2 and Nav1.6 voltage-gated ion channels, facilitating excitatory transmission and depression of the inhibitory one.In addition, PRRT2 modulates synaptogenesis and neuronal migration during brain development [3].
PRRT2 gene is located on chromosome 16p11.2and consists of four exons.Several mutations have been identified -mainly nonsense, missense, or deletion -all leading to gene haploinsufficiency.The frameshift mutation c.649dupC (p.Arg217Profs*8) is the most common pathogenetic variant (about 80% of cases) and causes a premature stop codon.About 5% of patients carry a PRRT2 de novo variant, whereas more than 80% present a familial heterozygous autosomal dominant mutation [1,4].
PRRT2 variants have been associated with a heterogeneous clinical spectrum, including paroxysmal kinesigenic dyskinesia (PKD), PKD with infantile convulsions (PKD/IC), benign familial infantile epilepsy (BFIE), and hemiplegic migraine (HM) [2].The clinical features may differ according to the age of symptoms' onset, probably due to the different expression of PRRT2 in cortical and subcortical regions during brain development.Thus, while epilepsy is usually diagnosed during the infantile period, episodic ataxia, movement disorders, and hemiplegic migraine are generally observed in adolescence or adulthood [1].
Benign familial infantile epilepsy (BFIE) with neonatal-infantile onset is the most frequent epilepsy in individuals with PRRT2 mutations, accounting for approximately 90% of cases [2].BFIE is a self-limited epilepsy characterized by non-febrile seizures that generally begins before twelve months of age and remits within two years of age.Neurological examination, electroencephalogram (EEG) recordings, and brain imaging are usually unremarkable.Patients with BFIE often present c.649dupC frameshift mutation, which is inherited in most cases.Seizures are mainly motor focal with possible focal-to-bilateral tonic-clonic evolution.Rarely, other severe epilepsy syndromes (i.e., severe myoclonic epilepsy, absence epilepsy, West Syndrome, Continuous Spikes and Waves During Sleep) as well as non-convulsive status epilepticus have been described in patients carrying PRRT2 mutations [5].However, no specific genotype-phenotype correlation has been yet established.
In this report, we describe the epileptic phenotypes of two patients with PRRT2 variants (NM_145239.3:c718C>T,p.Arg240Ter; c.649dupC, p.Arg217Profs*8), affecting two different residues within the extracellular domain of the protein.We also summarized previous literature, exploring the epileptic phenotypes related to PRRT2 disorders with their clinical, diagnostic, therapeutic, and prognostic implications.

Searching strategy and review organization
We performed a systematic review of the literature using the following search strategy: ('epileptic phenotypes'/exp OR 'epilepsy') AND ('PRRT2-related disorders′).The following electronic databases and data sources were systematically searched: MEDLINE (accessed through PubMed), Scopus and Google Scholar.We evaluated studies which reported a confirmed PRRT2 pathogenic variant significantly associated with patient clinical features that included epileptic phenotypes.We considered only papers published in English.

Results of this systematic review have been reported following the guidelines of the Preferred Reporting
Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.The quality of the included studies was assessed using the Newcastle-Ottawa Quality Assessment Scale (NOS).According to this scale, each study has been evaluated on the basis of eight items, described as follows: 1) representativeness of the exposed cohort; 2) selection of the not exposed cohort; 3) ascertainment of exposure; 4) demonstration that outcome of interest was not present at the start of the study; 5) comparability of the cohorts included; 6) assessment of outcome; 7) adequate length of the follow-up; 8) adequacy of follow up of cohorts.This score ranges from 0 to 9, and a quality score equal to or higher than three was considered acceptable.

Patients recruitment
We recruited two individuals carrying PRRT2 mutations and epilepsy.We compared their epileptic phenotype and the genetic data with the PRRT2-mutated epileptic individuals reported in the literature.
Patients' data were collected from the referring epilepsy center by reviewing clinical files, electroencephalography (EEG), and magnetic resonance imaging (MRI) features.The families involved in the study provided informed consent.

Genetic studies
Whole-exome sequencing (WES) was performed on the probands, their parents, and the available healthy siblings.Genomic DNA was isolated from 1 ml of peripheral blood and enriched with SureSelect Clinical research exome 54 Mb (Agilent Technologies).Whole exome sequencing (WES) runs were performed in all affected families on Illumina sequencers using a standard Illumina pipeline.Paired-end reads were mapped to the reference human genome sequence (GRch37/hg19) after removal of duplicates.
Single-nucleotide polymorphisms (SNPs) and short deletion or insertion (indels) variants were called using the specific variant calling plugin and dbSNP147.The variants were filtered for in-house variant controls and GnomAD databases.To investigate the presence of homozygosity regions in consanguineous families recruited in this study, homozygosity mapping was performed analysing the proband variant call format (VCF) WES data on Homozygosity Mapper.Validation and segregation studies of the candidate variants that emerged by WES were performed by traditional Sanger sequencing.

Literature Search
The literature search reported above yielded 260 articles.Fifty abstracts were excluded because not focused on PRRT2-related disorders or epileptic phenotypes or did not report individual patients' data.
Of the 122 records screened, the full texts of 72 articles were reviewed for eligibility (Figure 2).Forty-seven articles initially considered for possible inclusion were eventually excluded and twenty-four were finally included in our review (Figure 2) which included 12 case reports, 5 cohort studies, 4 retrospective observational studies, and 3 case series (Table 1).According to NOS evaluation, 9 articles were scored 9 and 15 were scored 5 (Supp.Tab.1).

Case reports Individual 1
Individual 1 is a 16-year-old boy born at term after a spontaneous delivery with an unremarkable perinatal period.Neurological symptoms started at 6 when the patient complained of the onset of paroxysmal movements involving the upper limbs after prolonged physical activity.During the following 6 months, two episodes of tonic-clonic seizures upon awakening and several focal unaware seizures with behavioral arrest were also reported.An EEG showed high-amplitude spike and spike-and-waves complexes in the left frontotemporal regions.Hence, a diagnosis of focal unaware seizure associated with PKD was made, and therapy with lamotrigine 100 mg bid was administered with good clinical response.At the last evaluation, at 16, the proband did not refer to seizure recurrence or paroxysmal movement disorders.
The patient's family history showed several relatives complying with neurological disorders, including focal seizures, PKD, migraine, and Lewy Body dementia (LBD).In particular, the patient's younger brother -an 8-year-old boy -showed the same paroxysmal upper limb movements after intense physical activity and received a diagnosis of focal unaware seizures with behavioral arrest at the age of 6.The patient's mother complained of paroxysmal movement disorders, which were otherwise diagnosed as PKD, and had a history of hemiplegic migraine.The maternal uncles and grandmother also suffered from migraine without aura.Finally, the maternal grandfather, as well as his twin, suffered from LBD (Figure 3, Individual 1).Hence, WES revealed a heterozygous autosomal dominant mutation in PRRT2 (NM_145239.3:c.649dupC, p.Arg217Profs*8), with phenotyping variability and incomplete penetrance, inherited from the mother.Blood samples were collected from patients and parents, and then trio-WES was performed.The variant was confirmed as inherited from the mother in heterozygosity with traditional Sanger sequencing.

Individual 2
Individual 2 is a 15-year-old girl, born at term from spontaneous delivery, with a past medical history of Benign familial infantile epilepsy (BFIE).From the age of 5 months to 2 years, the patient presented several episodes characterized by focal to bilateral tonic-clonic seizures treated with lamotrigine.
During childhood, the patient started reporting a daily recurrence of focal unaware seizures characterized by behavioral arrest.A concomitant EEG showed high-amplitude spike and spike-and-waves complexes in the right temporal regions.In the following years, several treatments were tried, including lamotrigine, carbamazepine, valproate, clobazam, and phenobarbital, only resulting in mildly reduced seizure frequency.From the age of 14, several nocturnal focal-to-bilateral tonic-clonic seizures (1-2 episodes/month) and paroxysmal movements involving the upper limbs after prolonged physical activity were also reported.Of note, paroxysmal movements were sometimes reported after prolonged focal-tobilateral tonic-clonic seizures.A diagnosis of PKD plus focal epilepsy was made.The patient was treated with a combined ASM therapy with brivaracetam 100 mg bid, eslicarbazepine 800 mg mid, and topiramate 200 mg bid, resulting in a >75% reduction of seizure frequency.At the last clinical evaluation, performed at the age of 15, the patients reported frequent (2-3 episodes/week) migraine attacks, occasionally associated with aphasia and left limb weakness and poorly responsive to symptomatic therapy (e.g., triptans and nonsteroidal anti-inflammatory drugs).The MRI-angiography of the brain was unremarkable.Hence, a diagnosis of hemiplegic migraine was made, and a prophylaxis treatment with SSRI was started with a good clinical response.
By reviewing the patient's family history, only sporadic cases of migraine without aura from the mother's side were observed, without any further cases of epilepsy (Figure 3, Individual 2).WES revealed a rare de novo mutation (NM_145239.3:c718C>T,p.Arg240Ter) in the PRRT2 gene.
Patients suffering from BFIE generally presented unremarkable interictal EEG findings, even though focal epileptiform activity were described in several cases.On the other hand, ictal EEG frequently documented focal parietal-occipital epileptiform activity, eventually becoming bilateral.Brain MRI scan was usually normal in most cases.From the therapeutical point of view, carbamazepine, valproate, and phenobarbital were the most frequently employed ASM, with remission in about 97% of cases.

Discussion
In recent years, PRRT2 variants have been identified in an evolving clinical spectrum, partially defined by PRRT2 molecular studies.Several epileptic phenotypes related to PRRT2 variants can be identified according to the high incomplete penetrance and variable expressivity of PRRT2 mutations [1].Remarkably, our literature appraisal highlights that c.649dupC represents a mutational hotspot responsible for about 57% of all phenotypes identified [28].This frameshift mutation was more frequently related to self-limited epilepsy phenotypes (i.e., BFIE) and rarely detected in patients with more severe phenotypes, such as developmental and epileptic encephalopathy with continuous spikes and waves during sleep [25].
Conversely, the gene haploinsufficiency with complete PRRT2 loss of function is associated with a more severe epileptic phenotype, such as refractory epilepsy, suggesting a gene dosage effect [2].Notably, 16p11.2deletions and biallelic variants have been identified in patients with convulsive status epilepticus, absence epilepsy, and severe epileptic phenotypes with daily frequency seizures.Recent evidence [29] has revealed that copy number variations (CNVs) in chromosome 16p11.2are associated with several neurological and psychiatric manifestations.Specifically, 16p11.2deletions are linked with seizures and severe epileptic phenotypes, while 16p11.2duplications have been observed in patients with psychogenic non-epileptic seizures (PNES) and, to a lesser extent, microcephaly.Therefore, reciprocal CNVs in 16p11.2appear to produce opposing phenotypes, with deletion carriers tending to have epileptic manifestations, and duplication carriers tending to have non-epileptic (PNES) manifestations.The broad phenotypes observed with 16p11.2CNVs, particularly deletions, suggest limited evidence of a phenotype-genotype correlation and the potential involvement of additional environmental/epigenetic factors and specific genegene interactions.
In our report, individual 1 carried a PRRT2 missense variant inherited from the mother, associated with a mild self-limited epileptic phenotype.The proband achieved seizure freedom and complete control of paroxysmal movement disorder after lamotrigine administration.Contrarywise, Individual 2 exhibited a severe epileptic phenotype characterized by an epileptic encephalopathy with refractoriness to ASMs related to a de novo PRRT2 variant.Even though functional studies were not performed for patient 2, her variant may play a dominant-negative effect, presumably interacting with the protein encoded by the unaffected allele, leading to a more severe phenotype, as has recently been supposed for some less frequent mutations [1].However, this hypothesis should be investigated in animal models and/or organoids.
Therapeutic response in patients with PRRT2-related epileptic phenotypes can be variable.
According to the literature, sodium channel blockers (i.e., oxcarbazepine, lamotrigine, and phenytoin) and valproate are generally effective in seizures and PRRT2 neurological manifestations (e.g., movement disorders and migraine) control.Since ASM administration leads to a global suppression of neuronal excitability, these drugs usually exhibit variable effectiveness in PRRT2-related disorders [2].However, evidence on specific molecular mechanisms is emerging.Specifically, the effectiveness of sodium channel blockers on migraine and movement disorders is related to the lack of negative modulation of Na+ channels by PRRT2 protein, also associated with the paroxysmal nature of these disorders [1].Consistently, levetiracetam, phenobarbital, sulthiame, ethosuximide, acetazolamide, topiramate, and lacosamide seemare less effective.
Interestingly, in our cohort, individual 1 showed a full therapeutic response to lamotrigine, whereas individual 2 showed a worse ASM response according to the more severe phenotype.
Our literature overview showed that PRRT2 pathogenic variants are mainly related to self-limited epilepsies characterized by unremarkable EEG and brain MRI findings, rare neuropsychiatric comorbidities, and great ASM responses (i.e., seizure freedom achievement with single ASM, mainly carbamazepine).However, further epileptic phenotypes have been reported, ranging from FS, FS+, GEFS+, to more severe spectrums.The causal relationship between PRRT2 mutation and FS is, debated, given the epidemiological overlap between FS and PRRT2-related epilepsy onset in patients ages from 6 months to 6 years [2].PRRT2 mutation can represent an incidental finding not related to the specific phenotype.
Interestingly, other epileptic phenotypes have been described in a minority of patients with PRRT2 mutations, such as severe myoclonic or non-convulsive seizures, absence seizures, DS, West Syndrome, and CSWS.In these cases, a protein dosage effect is the only recognized mechanism that leads to these neurological manifestations.According to the Human Genome Variation Society (HGVS), 102 PRRT2 pathogenic variants have been reported, including missense (14), nonsense (25), frameshift (58), splice junction loss (4), and stoploss (1).Specifically, nonsense and/or frameshift variants cause a truncated protein without the C-terminal, decreasing the protein level.Interestingly, functional studies documented that pathogenic missense variants were located both in the transmembrane (TM) domains and the loop domain of the C-terminal, impairing the plasma membrane localization and/or the protein level.In contrast, likely benign and benign missense variants affect the N-terminal of the protein without affecting localization and/or the protein level.Presumably, the C-terminal is crucial for the protein function and modulates the correct PRRT2 localization.Pathogenic missense variants lead to a mis-localization with a protein loss of function and a decreased protein level, while nonsense and frameshift variants reduce the protein level, with PRRT2 haploinsufficiency.Moreover, biallelic mutations disrupt the PRRT2 protein more severely than heterozygous mutations.Of note, an individual described in the literature with heterozygous compound variants presented a life-threatening phenotype characterized by refractory epilepsy, status epilepticus, severe prolonged ataxia, and paroxysmal dyskinesia [27].Furthermore, PRRT2 variants exhibit high variable expressivity, incomplete penetrance, and pleiotropy, suggesting that the genotype-phenotype correlation is not strict and should consider possible environmental factors involved.
Moreover, mice studies suggested that PRRT2 modulates ion channels and synaptic function with distinct mechanisms during brain development [30].Therefore, some disorders are age-dependent and appear in distinct temporal windows, with a different protein expression between developing age and adulthood.
To date, there is no clear explanation of how individuals carrying PRRT2 heterozygous variants can manifest benign (e.g., BFIE) as well as more severe (e.g., epileptic encephalopathies) epileptic phenotypes.Some mutations could lead to a negative protein-to-protein interaction between the mutated and the unaffected allele, leading to protein dysfunction.However, further studies are needed to confirm this interaction.
Research has shown that PRRT2 pathogenic variants are linked to movement disorders such as paroxysmal dyskinesias.This suggests that the protein plays a role in a wide range of neurobiological processes, with an expression pattern in specific brain regions.The paroxysmal PRRT2 movement disorders may be related to simultaneous, induced, Ca2+-mediated neurotransmitter release after a kinaesthetic or non-kinaesthetic trigger.This occurs after a short-term potentiation in hyperexcitable cerebellar networks of affected patients.Many other genes linked to channelopathies and synaptic vesicle trafficking defects have also been associated with both epilepsy and movement disorders, indicating different neurological phenotypes related to the exact genetic cause.Further studies on PRRT2-related disorders could clarify the full molecular and neurobiological roles of PRRT2, leading to personalized strategies for disease prevention and management.

Conclusions
PRRT2 mutation can be associated with a variety of epileptic phenotypes, which can range from benign ASM-responsive forms to more severe epileptic encephalopathies.Identifying PRRT2 variants in patients who have epilepsy may help achieve a more personalized treatment approach.However, phenotypegenotype correlations remain a challenge, and functional studies will be needed to fully identify the role of PRRT2 variants in defying the protein dysfunction, above all in those cases associated with severe ASM unresponsiveness.Novel therapeutic approaches are needed to increase seizure control and neurological outcomes in PRRT2-mutated individuals.In addition, undiagnosed PRRT2-related neurological disorders, such as refractory seizures, could impair brain development, especially if they occur in a critical temporal window.

Statement of Ethics
Written informed consent was obtained from the patients for the publication of this case series and any accompanying images.The paper is exempt from ethical committee approval because it is not necessary for the publication of the case report.We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

Conflict of interest statement
The authors declare no conflict of interest.

Author contributions
GS and FD contributed to the conception and design of the study.GS, FD, and SLS wrote the manuscript.All authors contributed to manuscript revisions, and read and approved the submitted version.

Data availability Statement
The data are available from the corresponding author upon reasonable request.

Figure 1 .
Figure 1.PRRT2 gene.PRRT2 figure with the proline-rich domain and other biologically active domains; reported current variants of affected individuals.

Figure 2 .
Figure 2. Preferred Reporting Items for Systematic Reviews and Meta-Analyses diagram (PRISMA)From the 260 articles found in the literature, 122 records were screened from which 24 articles were finally included in the literature review.

Figure 3 .
Figure 3. Familial pedigree and PRRT2 variants.Familial pedigree of PRRT2 affected individuals.In the Individual 1 the variant p.Arg217Profs*8 has been inherited with per maternal via, with phenotyping variability and incomplete penetrance.In the individual 2, p.Arg240Ter represented a de novo rare mutation.LBD: Lewy body dementia; PKD: paroxysmal kinesigenic dyskinesias.