The clinical spectrum of female epilepsy patients with PCDH19 mutations in a Chinese population

Mutations in PCDH19, which encodes protocadherin 19, have been identified in epilepsy, mainly in affected females. We summarized the clinical spectrum of female epilepsy patients with PCDH19 mutations in a Chinese population. We screened for PCDH19 mutations in 75 girls diagnosed with Dravet syndrome (DS) without a SCN1A mutation and 29 girls with fever‐sensitive and cluster seizures. We identified 11 novel and 7 reported mutations in 21 of 104 probands (20.2%), including 6 (6/75, 8%) DS girls and 15 (15/29, 51.7%) girls with fever‐sensitive epilepsy. The mutations were inherited in 9 probands, de novo in 11, and undetermined in the remaining patient. Shared clinical features included early onset seizures (5–18 months), seizures sensitive to fever, focal seizures or generalized tonic–clonic seizures in clusters and brief seizures. Mental retardation was present in 17 probands. Three patients had autistic features. Two of the nine probands with inherited mutations had no family history of epilepsy, one inherited the mutation from her transmitting father and the other inherited from her asymptomatic mother. Our results confirmed that the clinical spectrum of PCDH19 mutations includes female DS patients, epilepsy and mental retardation limited to females, epilepsy with normal development and asymptomatic female carriers.

The clinical manifestations of PCDH19-related epilepsy are heterogeneous, but the cardinal features include early seizure onset, generalized or focal seizures highly sensitive to fever, and brief seizures occurring in clusters repeating over the course of several days (5). The phenotype ranges from mild epilepsy to epileptic encephalopathy (8). Brief generalized tonic-clonic seizures (GTCS) and/or focal seizures in clusters are the frequent seizure types. Other seizure types such as myoclonic seizures, atypical absences and atonic seizures are rarely observed (8)(9)(10)(11). Cognitive development was reported to range from normal through mild to severe intellectual disability, with or without behavioral and psychiatric problems (8)(9)(10). The clinical manifestations of patients with PCDH19 mutations could overlap those of DS, which is mainly caused by SCN1A mutations (2,3).
In the present study, we screened the PCDH19 mutations in 75 unrelated female DS patients without a SCN1A mutation and 29 female epilepsy patients with fever-sensitive and cluster seizures in the Chinese population. We aimed to assess the rate of PCDH19 mutations in these two groups of Chinese patients and further understand the clinical features of female epilepsy patients with PCDH19 mutations.

Subjects
A total of 104 female probands, with either sporadic or familial fever-sensitive epilepsy were recruited to be screened for PCDH19 mutations, including 75 female DS patients without SCN1A mutations. All patients were identified in the Pediatric Clinic of Peking University First Hospital from February 2005 to December 2015. All 75 of the DS patients met the previously described diagnostic criteria for DS (12,13). The other 29 girls with fever-sensitive epilepsy met the following inclusion criteria (5,6,11,14): (i) seizures begin in infancy or early childhood (range: 4-60 months); (ii) seizure types mainly consist of generalized tonic, clonic, tonic-clonic, and/or focal seizures; (iii) seizures are in clusters and are highly sensitive to fever; and (iv) mental retardation present or absent (n = 19 and n = 10, respectively). Clinical information and blood samples were collected from the probands and their relatives when possible. The study protocol was approved by the Ethics Committee of Peking University First Hospital. Written informed consent was obtained from the participants or parents before enrollment.

Genetic analysis
Genomic DNA was extracted from peripheral blood lymphocytes by a simple salting-out procedure (15). The SCN1A mutation analysis in female DS patients was previously performed using Sanger screening and the multiplex ligation-dependent probe amplification (MLPA) method (SALSA MLPA P137) (13). The six coding exons of the PCDH19 gene and their flanking introns (reference sequence NM_001184880.1) were amplified and sequenced in both directions using an ABI 3730 Genetic Analyzer (Applied Biosystems, Foster City, CA). Five PCR primer pairs were used to amplify exon 1, which is longer than 2 kb. Primer sequences and PCR/sequencing conditions are available upon request. The novelty of the mutations and non-pathogenic variants found in the present study was examined using the Leiden open variation database (http://www.lovd.nl/ PCDH19), the dbSNP database and the Exome Aggregation Consortium (Exac) database. Segregation analysis was performed for all available members of a pedigree. The novel PCDH19 missense substitutions were not found in a cohort of 100 controls. Sequence alignment with PCDH19 genes in other vertebrates was performed using ClustalW (http://www.ebi.ac.uk/ clustalw/). The pathogenicity of the identified variants was predicted using Mutation Taster Server (http://www .mutation taster.org), Polyphen-2 (http://genetics.bwh .harvard.edu/pph2/) and SIFT (http://sift.jcvi.org/).
To detect PCDH19 deletions and duplications, the MLPA method was applied using the SALSA MLPA P330-A2 probe mix in those patients determined to be PCDH19 mutation-negative by Sanger sequencing.

The inheritance of PCDH19 mutations in 9 families
The mutations were inherited in 9 probands (45%, 9/20), de novo in 11 probands, and undetermined in the remaining patient (patient 10) because the patient's father's DNA was unavailable. The segregation of inherited variants was performed in families as shown in Figure 4. Two of the nine probands with inherited mutations had no family history of epilepsy, one (proband 9) inherited the mutation from her transmitting father and the Fig. 4. Pedigrees of the nine families with inherited PCDH19 mutations. Arrow, proband; a dot in a square, asymptomatic mutation-carrying male; a dot in a circle, asymptomatic mutation-carrying female; black circle, affected mutation-carrying female; gray-filled circle, affected female without PCDH19 mutation. m/+: individuals heterozygous or mosaic for the mutation; m: individuals hemizygous for the mutation; +/+ and +: individuals homozygous and hemizygous for the wild-type allele. other (proband 14) inherited the mutation from her asymptomatic mother. For the seven familial cases, two (probands 5 and 19) were three-generation pedigrees, two (probands 4 and 6) were two-generation pedigrees and three (probands 7, 12 and 13) were single-generation pedigrees.
The clinical phenotype of the females with PCDH19 mutations showed wide inter-and intra-familial variability, even between twins. Proband 5 was from a three-generation family, and her paternal grandmother never experienced febrile seizures or epilepsy but had behavioral disturbances including mild depression symptoms, agitation and aggressive features. In the three families in which the probands inherited the PCDH19 mutations from their mother, two mothers were affected and one was asymptomatic. Variability between siblings was also evident, even in twins. In our study, the mother and aunt of proband 6 were monozygotic twins and proband 12 and her sister were dizygotic twins. The mother of proband 6 had febrile seizures with normal cognitive function starting at approximately 3 years of age. Her monozygotic twin sister (no DNA sample available) also experienced fever-sensitive seizures beginning at approximately 3 years, had cognitive impairment and died of status epilepticus at 21 years. Proband 12 experienced her first brief focal febrile seizure cluster at 5 months and later developed frequent febrile and afebrile seizures before 6 years. She had intellectual disability and distinct autistic characteristics. However, her twin sister experienced her first brief focal afebrile seizure cluster at 18 months. Seizures were controlled by VPA and LEV treatment at 3 years. She had normal cognitive function except mild attention deficit.

Discussion
In this study, we searched for PCDH19 mutations in DS patients without SCN1A mutations and girls with fever-sensitive and cluster seizures in the Chinese population. PCDH19 mutations were detected in 21/104 (20.2%) females with fever-sensitive epilepsy. The percentage of PCDH19 mutations in 6 DS girls without SCN1A mutations (8%) is much lower than that (51.7%) in 15 girls whose clinical features fulfill the diagnosis of PCDH19-related epilepsy.
Up to now, more than 150 PCDH19 mutations including missense, nonsense, small deletions and insertions, splice site mutations, intragenic deletion and whole gene deletions, have been reported, with most of the mutations (>90%) located in the largest exon, exon 1, which encodes the extracellular cadherin domain of the protein (6). In this study, we identified 18 mutations clustered in exon 1, including nine missense mutations, four nonsense mutations, three small deletions, and two small insertions. Eleven were novel mutations and 7 were reported previously (1, 2, 5, 9-11, 14, 16-19). The missense mutation p.N340S was identified in three unrelated patients (two DS and one non-DS), and the frameshift mutation p.Y366LfsX10 was identified in two unrelated patients. Both of these mutations were reported previously (1, 2, 5, 9-11, 14, 16-18). The high occurrence of recurrent mutations in our cohort and previous studies suggests that PCDH19 is an important gene in epilepsy. The majority of PCDH19 mutations in our patients are unique. Hence, our reports expand the spectrum of PCDH19 mutations associated with epilepsy in females.
Previous studies have claimed that chromosomal rearrangements may occur (2,4,20); thus, screening for PCDH19 rearrangements should be performed in addition to direct sequencing of the entire coding sequence. Nonetheless, our MLPA screening involving the PCDH19 gene did not identify any deletions or duplications.
Recently, the increasing number of PCDH19-related epilepsy patients have been sporadic cases, rather than familial cases (5,11). De novo mutations accounted for more than half of the sporadic cases in our study; and the remaining mutations were inherited from asymptomatic fathers, asymptomatic mothers, or from mothers with only a history of febrile seizures (2-5, 8, 9, 11, 19, 21). In this study, 14 of the 21 probands were isolated cases. In 11 out of the 20 probands in whom inheritance could be assessed, the mutation arose de novo. Two out of the nine probands with inherited mutations had no family history of epilepsy, one was inherited from her transmitting father and the other was inherited from her asymptomatic mother. The transmitting fathers were presumably asymptomatic due to the unusual X-linked inheritance of PCDH19. However, sporadic asymptomatic mother carriers have also been reported (1,4,20,22). Completely skewed X-inactivation or a somatic mosaicism may explain the asymptomatic mothers with the mutations (22,23).
Genotype-phenotype correlations were difficult to establish in this study because of distinct phenotypic variability, even in individuals with the same mutation. In our study, phenotypic variability was even evident in members of the same families including a twin sibling. The grandmother (II-4) of proband 5 carrying the PCDH19 mutation had only mild behavioral problems without seizures. The mother of patient 14 was asymptomatic, and the mother of patient 6 had only febrile seizures in her childhood. Partial or completely skewed X-inactivation may play a great role in females with a mild phenotype or asymptomatic carrier (4). Other genes or environmental factors may also be involved in the phenotypes associated with PCDH19 mutations (9).
Our six DS patients with PCDH19 mutations all fulfilled the main criteria for DS including age of seizure onset (before one-year-old), seizures triggered by fever, multiple seizure types, and cognitive impairment. As previously reported, DS patients with PCDH19 mutations had a slightly older onset age (median age of 9.5 months, with a range from 7.5 to 12 months, vs 5.3 months with a range of 2-11 months in our series of SCN1A-positive DS patients), less frequent status epilepticus (SE), fewer myoclonic seizures, fewer photosensitivity, and a better long-term outcomes than DS patients with SCN1A mutations (2,5,13). In our six DS patients, the seizure onset was 6.5 months, only one had SE, three had myoclonic seizures, and none had photosensitivity. Two DS patients also exhibited autistic features. We followed up with four of the six DS patients with PCDH19 mutations for 1-6 years. In the first few years, seizure frequency increased along with cognitive impairment. Seizures then became less frequent, and cognitive impairment was less severe than in those DS patients with SCN1A mutations. Specchio et al. reported that antiepileptic treatment could not prevent the recurrence of seizure clusters and that the seizures improved with growth and fewer episodes of febrile illnesses (9). Therefore, PCDH19 mutation screening should be performed in DS female patients without SCN1A mutations.
The other 15 patients with PCDH19 mutations were recruited from a group of female epilepsy patients with fever-sensitive and cluster seizures, including six familial EFMR cases. The clinical features of these patients were in accord with the features of reported cases with PCDH19 mutations (2-5, 8, 9, 14, 18). Seizures were resistant to treatment during infancy and childhood, but seizure frequency and intractability tended to decrease over time. Similarly, most of the patients presented a variable degree of mental retardation, though some patients may have normal cognitive function (8,9,18). Only one of our 15 patients had SE during the follow-up period. Female sporadic epilepsy patients with clusters of attacks and fever sensitivity should also be tested for PCDH19 mutations.
In summary, the clinical spectrum of our patients with PCDH19 mutations include female DS patients, familial or isolated epilepsy and mental retardation limited to females (EFMR), epilepsy with normal development and asymptomatic female carriers. Our data confirmed that female DS patients without SCN1A mutations and females with fever-sensitive and cluster seizures should be tested for the PCDH19 mutation. In addition to the clinical features of epilepsy described for PCDH19-related patients, females with PCDH19 mutations may present behavioral problems without seizures or may even be asymptomatic.