Discussion
PRRT2 is an important gene recently identified in neurological paroxysmal disorders.
PRRT2 is the major causative gene of PKD, BFIE, and ICCA [
5‐
8,
12‐
19], and is also responsible for several familial or sporadic cases with paroxysmal non-kinesigenic dyskinesia (PNKD), paroxysmal exercised-induced dyskinesia (PED), sporadic BIE, and hemiplegic migraine (HM) [
20‐
24]. In this study, we observed that
PRRT2 mutations are also common in Chinese families with BFIE and ICCA.
Including our mutations, 62 different
PRRT2 mutations have been described in literature in B(F)IE (Additional file
2: Table S2). To date, 170 BFIE families and 32 sporadic BIE cases have been screened for
PRRT2 mutations.
PRRT2 mutations were identified in 74.1% (126/170) of BFIE families and 34.4% (11/32) of sporadic BIE cases (Additional file
3: Table S3). Fifteen different
PRRT2 mutations were identified in BFIE or BIE (Additional file
2: Table S2), including two insertion mutations, five deletion mutations, four missense mutations, one nonsense mutation, and three splice site mutations. References cited in Additional file
2: Table S2 and Additional file
3: Table S3 were listed in the Additional file
4: Supplemental references. The mutation rates of
PRRT2 in BFIE ranged from 54.5% to 85.7% in previous studies [
6,
12,
17].
Our study provides the first report of the deletion mutation c.649delC and the insertion mutation c.904_905insG being associated with a phenotype of BFIE only. The mutation c.649delC has been previously reported in PKD, ICCA, and HM [
13,
22,
25‐
27], but not in families with BFIE. This mutation was observed in our three BFIE families (Families 10–12). The mutation c.904_905insG was previously described in one ICCA family [
21]. We detected this mutation in one BFIE family (Family 13). The most described mutation c.649_650insC is a hotspot for
PRRT2 mutation. This was also present in our BFIE families, accounting for 69.2% (9/13) of
PRRT2 mutation-positive families. In previous studies, the percentage of c.649_650insC in BFIE families with a
PRRT2 mutation ranged from 85.7% to 92.9% [
6,
12,
24]. This mutation was observed mainly with a BFIE, PKD and ICCA phenotype [
6,
8,
12,
13,
28,
29], and but also in a few familial or sporadic cases of PNKD, PED, HM and episodic ataxia [
20‐
22]. This site appears to be particularly prone to mutation: it occurs in a tract of nine cytosines preceded by four guanines, facilitating the formation of a hairpin loop and slippage during DNA replication. Besides the deletion mutation c.649delC and the insertion mutation c.649_650insC, a nonsense mutation c.649C > T affecting the same nucleotide has been reported [
13,
15,
25,
30‐
33]. Incomplete penetrance was found in five of our BFIE families (Families 4, 5, 6, 8 and13). This phenomenon has also been described in other studies [
6,
12,
28]. It was reported that a penetrance of
PRRT2 mutations in BFIE was 82% [
12].
In our study, two probands of BFIE families experienced a single diarrhea-induced seizure. We cannot differentiate whether the late relapse of seizures in these two patients is a manifestation of convulsions with mild gastroenteritis (CwG), or the presentation of BFIE itself. CwG is a well-recognized infant seizure disorder associated with mild diarrhea. It is characterized by [
34]: (1) previously healthy infants and young children aged 6 months to 3 years having afebrile generalized convulsions associated with symptoms of gastroenteritis; (2) seizures occuring sometimes in cluster; (3) normal laboratory examination results including electrolytes, blood glucose and cerebrospinal fluid; (4) normal interictal electroencephalography; and (5) excellent seizure and developmental outcomes. The incidence of CwG is thought to be higher in Asian populations than in people from Western countries [
35‐
38]. Okumura reported that approximately 10% of children with BIE experienced CwG [
39]. To explore whether
PRRT2 mutations are associated with CwG, several groups conducted
PRRT2 mutation analysis in patients with CwG, but no
PRRT2 mutation was found [
18,
24,
40].
In our study,
PRRT2 mutations were present in 59.1% of Chinese BFIE families, presumably relating to the ethnic background. For the absence of a mutation in
PRRT2 in our remaining nine BFIE families, microdeletion of the
PRRT2 gene or other BFIE-related genes (
SCN2A,
KCNQ2, and
KCNQ3) should be to be screened [
29]. Microdeletions of
PRRT2 gene have been found in sporadic cases of PKD and ICCA [
41], and also in one PKD family [
33].
Including our ICCA families, 95 families with ICCA has been reported (Additional file
3: Table S3). 23 different
PRRT2 mutations have been reported in the ICCA phenotype (Additional file
2: Table S2): seven insertion mutations, four deletion mutations, four missense mutations, six nonsense mutations, one splice site mutation, and one microdeletion. Overall,
PRRT2 is reported to be mutated in 83.3%-100% of ICCA families [
6,
8,
18,
31], although one study reported that the much lower rate of 37.5% [
17].We reviewed the
PRRT2 mutation rate in BFIE and ICCA families described in literature and our study (Additional file
3: Table S3), and found that it was higher (91.6%, 87/95) in ICCA families than in BFIE families (74.1%, 126/170).
The mutation c.514-517delTCTG has been previously observed in one family with PKD [
5], and we found this mutation in an ICCA family. The mutation c.649delC has been reported in one sporadic ICCA case. In our study, we found three BFIE (Families 10–12) and three ICCA families (Families 26–28) with this mutation. This suggests that this mutation is the second most common mutation in Chinese families with BFIE and ICCA. We identified a novel stop codon mutation c.1023A > T (X341C) in one ICCA family. As far as we know, this mutation is the first reported mutation of
PRRT2 exon 4. It is predicted to change the stop codon into a cysteine, and would introduce a new stop codon after a 28-amino acids elongation of the C-terminal tail, which forms the extracellular domain of the PRRT2 protein. Functional studies of this mutation should be performed in the future.
We observed some uncommon phenotypes in our ICCA families. Family 23 is an ICCA family with c.649_650insC mutation. In this family, migraine co-existed with ICCA in one individual and with PKD in another. Both relatives had the 649_650inC mutation. The phenotype of migraine (with or without aura) has been reported in BFIE, PKD or ICCA families with a
PRRT2 mutation [
17,
19,
20,
22,
33,
42,
43].
PRRT2 mutation was identified in one large family with isolated HM [
20]. The reported frequency of migraine among
PRRT2 mutation carriers is significantly higher (27.1%) than in the overall population with epilepsy (8%-15%) [
19,
44]. Therefore, the association of migraine with
PRRT2 mutation may not be coincidental. In Family 27, the proband (III-4) with ICCA also had an episode of febrile seizure (FS) at 7 years old, conforming to the diagnosis of febrile seizures plus (FS+). Patients with FS or FS + have been reported in families with BFIE [
12,
23,
29,
40,
45], PKD [
25,
30,
46] or ICCA [
15,
17,
18]. Some carried
PRRT2 gene mutations. However, FS may not be associated with
PRRT2 mutations in ICCA families. For example,
PRRT2 mutation did not co-segregate with FS in a large ICCA family [
15]. In a previously described ICCA family, only two of the four FS patients had a
PRRT2 mutation [
18]. FS are more common in patients with epilepsy than in the general population [
47]. It is therefore plausible that the occurrence of febrile seizures in patients with a
PRRT2 gene mutation is not solely caused by this mutation, but by the occurrence of epilepsy itself. It could also be caused by a mutation in another gene. The proband in Family 28 had PED. In ICCA families, paroxysmal dyskinesias are mostly of the kinesigenic type, but families with PED have also been reported [
48]. A phenotype of PED with
PRRT2 mutation has been reported in two Chinese ICCA families [
21].
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
YZ designed the study and contributed to the initial draft of the manuscript. YZ, XX, SW, ZY, YW, XL and XW assessed the patients clinically, performed the phenotyping, and collected the DNA samples. XY extracted DNA from peripheral blood, analyzed the clinical and genetic data, and drafted the manuscript. All authors read and approved the final manuscript.