PPP is a genetic disorders characterized by episodes of muscle weakness caused by mutations in genes encoding channel proteins of skeletal muscle [7]. The molecular diagnosis of PPP was previously based on hot spots, exon-by-exon screening of the reported genes or targeted panel sequencing [13]. The latest and large cohort studies on PPP are mainly concentrated in patients from the United States and Europe, and have identified a heterozygous pathogenic mutation in 60–70% of patients meeting clinical criteria [11, 12]. We performed genetic analysis for 37 Chinese patients using a gene panel and WES, which elucidated genetic causes in 59.5% (22/37) of patients in our cohort. This was consistent with another study in the Chinese population, identifying candidate gene variants in 65.0% (39/60) of patients [13].

In the United States and Europe, the most commonly affected gene in patients with HypoPP was CACNA1S, followed by SCN4A [2]. In up to 3% of patients, some mutations have been found to be associated with thyrotoxic HypoPP, most commonly affecting KCNJ18 and KCNE3 [19, 20]. It has been reported that very few patients have been found to carry RYR1 gene variants [21]. In addition to covering the CACNA1S and SCN4A genes, our panel included five additional genes associated with periodic paralysis, such as RYR1, KCNE3, KCNJ18, KCNJ2 and KCNJ5. Among 21 patients with HypoPP, 15 were classified as having HypoPP-2 with SCN4A variants, and 6 as having HypoPP-1 with CACNA1S variants. Based on these results, we speculated that the SCN4A gene may be the most common causal gene in patients with HypoPP in China, different from Europe and the United States. This result was consistent with another study in the Chinese population, indicating that SCN4A accounts for the majority of HypoPP, while CACNA1S mutations are relatively rare [13]. Brugnoni et al. found that among 51 patients with HypoPP, 4 Asian patients were found to carry pathogenic variants of the SCN4A gene [9]. This may imply that SCN4A is the most commonly affected gene in Chinese or Asian patients with HypoPP.

In our cohort, only one of the 22 probands was female, and the penetrance was 95.5% (42/44) in males, compared to 73.4% (14/19) in females. These results are consistent with previous literature reports that the penetrance of periodic paralysis was lower in women than in men, which may be associated with sex hormones [4].

The α subunits of the skeletal muscle Cav 1.1 channel and Nav1.4 channels are composed of four homologous transmembrane domains (DI-DIV). Within every domain, six transmembrane helical segments are present (S1-S6) consisting of two distinct modules: the voltage-sensor domain (S1-S4) and the pore domain (S5-S6) (Fig. 4). The S4 segment of each homologous domain contains repeating motifs of one positively charged amino acid followed by two hydrophobic residues (R/K-X-X; X-hydrophobic residues). The outward movement and rotation of the S4 helices in response to depolarization inducse a conformational change within the channel, which opens the main ion-conducting pore [22, 23]. Most of the CACNA1S and SCN4A mutations identified so far affect residues located in segment S4 of all four domains of the protein, which represents the voltage-sensing apparatus of the channel and leads to the replacement of a highly conserved positively charged arginine with a weakly charged or neutral amino acid residue [22, 24]. It has been reported that the common mutational hotspots in the CACNA1S gene include p.R174W (DI-S4), p.R528G/C/H (DII-S4), p.V876E (DIII-S3), p.R897S (DIII-S4), p.R900G/S (DIII-S4), p.R1086H (cytoplasmic), p.R1242G/S (DIV-S4), and p.R1239H/G (DIV-S4) [8, 25,26,27,28,29,30]. Similarly, the hotspots in the SCN4A gene include p.R222W (DI-S4), p.R669H (DII-S4), p.R672S/G/C/H(DII-S4), p.N820Y (cytoplasmic), p.R1129Q (DIII-S4), p.R1132Q (DIII-S4), p.R1135H/C (DIII-S4), and p.P1158S (Ex) [12, 31,32,33,34,35,36]. More than half of the variants (13/22) are located in the S4 segment of four domains and result in the replacement of an arginine in our study cohort (Supplement Table 1; Fig. 4). Our results indicate that the variants p.R675Q (DII-S4) and c.2020-5G > A may also be relatively common mutational hotpots in Chinese HypoPP-2 patients. The splicing variant c.2020-5G > A contributed to a non-frame-shift change (p.Leu673_Leu674insGln) in segment S4 of DII. The insertion of the hydrophilic amino acid Gln may disrupt the repeating motif of the S4 segment, stabilizing the hydrogen bonding interactions fo the Gln with Ser670 and Val677 (Fig. 5), eventually leading to a conformational change of the S4 helices, affecting the voltage-sensing apparatus. This variant was classified as “likely pathogenic” according to guidelines determined by the ACMG (PS4 + PM2 + PP3).

CACNA1S and SCN4A mutations included in our study are pictured in the same structural model channel. Both Cav1.1 and Nav1.4 channels are composed of four homologous transmembrane domains (DI–DIV); each containing six transmembrane domains (S1-S6). Plus (+) symbolizes the positive charges in segment 4 of the calcium/sodium channels. Black square: mutations in the Nav1.4 channel; Blue hexagon: mutations in the Cav1.1 channel. The figure is adapted from Brugnoni et al. (2022)

The 3D molecular structure of SCN4A. (A) Homology model of SCN4A mutant protein generated using the SwissModel online server. Magnified views of the wild-type (B) and mutant Leu673_Leu674insGln (C) are shown respectively. The H-bonds are shown as yellow dashed lines, and Gln is indicated by a red arrow

Initially, we started to analyze PPP patients using panel sequencing to identify pathological mutations in known genes. With the development of high-throughput sequencing technologies, we more recently began using WES to analyze the underlying genetic variations. Indeed, WES could be useful for the detection of mutations in regions or genes not included in our panel, as well as macro deletions or duplications. However, WES was not able to identify any mutation in HypoPP patients already screened for mutations using the panel. Moreover, the overall detection rate was comparable between the panel (54.5%) and WES (61.5%). To our best knowledge, this is the first study that analyzes the diagnostic utility of WES in patients affected by PPP, and the results do not support its superiority in the molecular diagnostic process of these diseases compared to the panel.

Mutations of the SCN4A are associated with various neuromuscular disorders, including HyperPP, HypoPP2, paramyotonia congenita and congenital myasthenic syndrome et al. [37,38,39,40,41]. SCN4A mutations leading to periodic paralysis or nondystrophic myotonia have been found located throughout every domain and segment of this channel [42]. It has been reported that similar or even the same mutations in the SCN4A gene can cause distinct clinical disorders [43]. For example, the variant R675Q can cause HyperPP, HypoPP-2 as well as NormoPP [42, 44,45,46]. In our study, the variant R675Q was the most common mutation site and was associated with HypoPP-2 in four families. Phenotypic differences between individuals harboring the same SCN4A mutation indicate that the genetic background and other factors perhaps contributed to the clinical expression of some variants.

HyperPP is characterized by attacks of flaccid limb weakness, hyperkalemia (serum potassium concentration > 5 mmol/L) and/or provoking/worsening of an attack by oral potassium intake, and normal serum potassium between attacks. In approximately half of the individuals, the onset is in the first decade of life, and the prevalence of HyperPP is approximately 0.17/100,000–0.06/100,000 [47]. There was only one HyperPP case (case 5) in the cohort, and included multiple patients in the family. The onset of the proband was 1 year old, and other patients were between 3 and 9 years old in the same family. The age at the appearance of the most severe symptoms was between 10 and 25 years. The relatively earlier onset time in patients with HyperPP than HypoPP in our cohort is consistent with previous reports that symptom onset is typically in the second decade for HypoPP while HyperPP tends to present earlier in childhood [11]. Mutations of SCN4A that are associated with HyperPP affect the gating behavior of this channel and produce gain-of-function defects characterized by impaired inactivation and/or enhancement of activation [48, 49]. We identified a heterozygous c.2111 C > T (p.Thr704Met) variant of SCN4A in case 5, one of the most common HyperPP mutations causing disrupted slow inactivation [50].

Patients with NormoPP often develop symptoms around the age of 10, generally in the form of paroxysmal muscle weakness with normal serum K+ concentrations. However, sometimes it was difficult to determine whether a normokalemic variant was a distinct entity or whether it was simply a form of HyperPP. One may see misleading measurement results of the serum potassium levels, when it is made too far apart from the progressive onset of the attack [21]. In a series of patients carrying the common T704M SCN4A variant, known to cause HyperPP, only 50% were shown to have hyperkalemia during attacks [49], and the diagnostic category of NormoPP was thought to be doubtful [51]. In addition, NormoPP is the rarest subtype of PPP and a further limitation of our study is the relatively small sample size. Therefore, it is understandable that no cases of NormoPP were found in our study.

In this cohort, 2 variants of SCN4A and 2 variants of CACNA1S were classified to uncertain significance according to ACMG. The Novel variant p.Ile1395Val was located in the DIV-S2 and the adjacent variant p.Ile1393Thr has been reported to be associated with paramyotonia congenita [52]. Another novel variant of SCN4A, p.Leu472Ile was located in the cytoplasmic Domain I. The remaining 2 variants in CACNA1S, p.V470M and p.Arg1302Gln, are distributed in DII-S2 and the S5-S6 loop of extracellular Domain 4, which are pore-forming domains. Although they are rare variants and might be pathogenic, a detailed co-segregation analysis and functional studies are still required to validate these results.