Introduction

Paroxysmal kinesigenic choreoathetosis (PKC; MIM 128200) is a paroxysmal movement disorder characterized by recurrent and brief attacks of unilateral or bilateral involuntary movements, including dystonic posturing, chorea, athetosis, and ballism, which are precipitated by the sudden onset of movements (Kato et al. 2006). The attacks can last as long as a few seconds to a few minutes, occur up to 100 times daily, but usually manifest in childhood or early adolescents, and commonly decrease with age. There is no loss of consciousness during these attacks. The attacks are responsive to anticonvulsants such as carbamazepine or phenytoin. Electroencephalogram (EEG) analysis demonstrates normal or nonspecific abnormalities. Neuroimaging and neuropathological studies resulted in unremarkable findings (Sadamatsu et al. 1999; Nagamitsu et al. 1999). The etiology and pathophysiology of PKC still remain unclear. Some neurologists consider PKC as a form of reflex epilepsy, whereas others believe that basal ganglia dysfunction may play a role in its cause (Kato et al. 2006). Most (40–70%) were familial cases in which PKC was transmitted in an autosomal dominant mode of inheritance with incomplete penetrance (Tomita et al. 1999; Valente et al. 2000). Males are affected more often than females, with an estimated ratio of 3–4:1 (Bhatia 1999).

We previously performed a genome-wide linkage and haplotype analysis in eight Japanese families with PKC and defined the disease locus within a 12.4-cM region between D16S3093 and D16S416 at 16p11.2-q12.1 (Tomita et al. 1999). This PKC-critical region (PKCCR) was confirmed by others (Bennett et al. 2000; Swoboda et al. 2000; Valente et al. 2000; Cuenca-Leon et al. 2002). In addition, mapped regions for other conditions probably allelic to PKC, such as infantile convulsions and paroxysmal choreoathetosis (ICCA; MIM 602066) and benign familial infantile convulsions (BFIC2; MIM 605751), shared with that for PKC (Lee et al. 1998; Hattori et al. 2000; Swoboda et al. 2000; Caraballo et al. 2001; Weber et al. 2004). Nevertheless, mutations in any genes within the PKCCR have remained uncovered.

Here, we describe the results of the mutation analyses of seven PKC families for a total of 157 genes located at or around the PKCCR, together with linkage/haplotype analysis of four newly identified families.

Materials and methods

Subjects

The subjects studied included seven families (Families A–G) with PKC. Two of them (Families F and G) corresponded respectively to Families 1 and 3 in our previous report (Tomita et al. 1999), and five other families (Families A–E, Fig. 1) were those which were newly collected. A total of 21 members, including 16 PKC patients from four of the five families (Families A–D), underwent a linkage and haplotype analysis. In addition, one of each affected individual (we further call them representative patients) chosen from all seven families was subjected to mutation analysis. Blood samples were collected from all participants after obtaining written informed consent, and the study protocol was approved by the Committee for Ethical Issues on Human Genome and Gene Analysis at Nagasaki University.

Fig. 1
figure 1

Pedigree of the five families (Families A–E) with paroxysmal kinesigenic choreoathetosis (PKC). The closed squares and circles denote individuals affected with PKC. Although not shown here, Families F and G correspond to Families 1 and 3 reported previously (Tomita et al. 1999)

Genotyping and linkage analysis

Genomic DNA was isolated from the blood lysate of the 21 participants by phenol-chloroform extraction, followed by ethanol precipitation. To try to narrow down the PKCCR, we performed genotyping and haplotype analysis in the four newly collected PKC families, using 13 microsatellite markers (Table 1) located to a 36-cM region at 16p12-q21 to which PKC has been shown to be linked, as well as using three additional markers (AC007353-M1, AC092368-M1, and AC092721-M3) that were designed by us according to the human genome sequence (http://genome.cse.ucsc.edu/). Sample DNA was polymerase chain reaction (PCR) amplified for each marker locus with fluorescence-labeled primers. PCR was performed on DNA Thermal Cycler Model 9700 (Applied Biosystems, Foster City, CA) in a 10-μl reaction mixture containing 1×PCR buffer (Takara Bio, Otsu, Japan), 200 μM each of dNTP, 0.5 μM each of primer, 10 ng DNA, and 0.25 units ExTaq DNA polymerase HS-version (Takara Bio, Otsu, Japan) under the conditions of denaturation at 94°C for 2 min, 35 cycles of 94°C for 30 s, 58°C for 30 s, and 72°C for 1 min, and final extension at 72°C for 7 min. PCR products were run on an auto-sequencer Model 3100 (Applied Biosystems, Foster City, CA). Allele sizes were analyzed by GeneScan and Genotyper software (Applied Biosystems, Foster City, CA) to determine the genotypes.

Table 1 Primer sequences of the 12 microsatellite markers used for genotyping and linkage analysis

Mutation analysis

We performed PCR-based mutation analysis of the seven representative patients with typical features of PKC. Sequences examined for mutations among the seven patients included those of 1,371 coding exons, excluding the 3′-UTR and 5′-UTR in 117 genes, located at the PKCCR between D16S3093 and D16S416 (Table 1). The analysis in five families (Families A and C–F) was expanded to an additional 192 exons of 40 other genes, whereas because of depletion of genomic DNA, an expanded analysis was not done in the remaining two families. Thus, a total of 1,563 exons in 157 genes were analyzed for mutation (primer sequences are available on request).

Real-time quantitative PCR

We performed real-time quantitative PCR in six representative patients from Families A and C–G. Six pairs of primers and TaqMan probes were designed for exons 1, 6, and 13 of SCNN1G, and for exons 1, 16, and 30 of ITGAL. PCR was carried out in a 10-μl reaction mixture containing 5 μl of 2×TaqMan Universal PCR Master mix (Applied Biosystems, Foster City, CA), 0.4 μM each of primer, a 0.2-μM probe, and 10 ng DNA under the conditions of 2 min at 50°C, 10 min at 95°C, 40 cycles of 15 s at 95°C, and 1 min at 60°C, with a 7900HT Sequence Detection System (Applied Biosystems, Foster City, CA).

Results

The haplotype analysis showed that all affected individuals in the four new families share an allele at each locus examined between D16S3131 and D16S408 (Fig. 2). One end of the shared region was defined by a recombination between D16S3131 and D16S3093 in individual III-3 in Family A, and the other end by a recombination between D16S514 and D16S408 in individual III-1 in the same family. These results defined a minimum PKCCR in the four new families within an approximately 24-cM segment between D16S3131 and D16S408. Therefore, the present linkage/haplotype analysis did not contribute to narrow down the previously defined PKCCR.

Fig. 2
figure 2

Haplotype analysis of the four new PKC families (Families A–D). The numbers in boxes represent putative disease haplotypes. The heavy short lines indicate recombination sites

Among a total of 1,563 exons of 157 genes analyzed, we detected 243 base alterations in the seven representative patients (Fig. 3), 36 of which were base substitutions in coding regions and have not been reported in the dbSNP database (http://www.ncbi.nlm.nih.gov/SNP/). Seven of the alterations in six genes were nonsynonymous substitutions resulting in amino-acid substitutions (Table 2). Five of such nonsynonymous substitutions in four genes were observed both in some patients and among 100 normal control individuals. The remaining two, i.e., 6186C>A in exon 3 of SCNN1G (the gene for sodium channel, nonvoltage-gated 1, gamma) and 45842A>G in exon 29 of ITGAL (the integrin alpha L precursor gene), observed in Families C and F, respectively, were not observed among more than 400 normal controls. The real-time quantitative PCR analysis did not detect a duplication or a deletion within the two genes. Of the 35 intronic base changes we identified in the seven patients, none were located at the acceptor or donor splice sites (Fig. 3).

Fig. 3
figure 3

Classification of 243 base alterations in 157 candidate genes. Information of the newly found single nucleotide polymorphisms (SNPs) is shown in Table 2. None of the novel intronic SNPs are located at any of the acceptor or donor splice sites

Table 2 List of genes for mutation analysis in the PKC patients, and novel SNPs identified, their positions, nucleotide changes, and amino acid changes

G-banding chromosome analysis at the 400-band level and C-banding analysis revealed that all five patients from Families B–D, F, and G had a normal-sized heterochromatin block on chromosome 16 without an inversion (data not shown).

Discussion

The PKCCR was assigned to a segment between D16S3093 and D16S416 in eight Japanese families (Tomita et al. 1999). It was also mapped between D16S3100 and D16S771 in an Afro-Caribbean family (Bennett et al. 2000), between D16S3131 and D16S3396 in 11 families of diverse ethnicity (Swoboda et al. 2000), between D16S3145 and GATA140E03 in a Spanish family (Cuenca-Leon et al. 2002), and a 24-cM segment between D16S3131 and D16S408 in the present study (Fig. 4). Thus, the shortest region of overlap (SRO) did not become narrower than a 12.4-cM segment detected by Tomita et al. (1999). Valente et al. (2000) assigned a form of PKC (a second PKC locus, EKD2), in an Indian family, to a segment between D16S416 and D16S503, the region distinct from those mapped by Tomita et al. (1999) and by Swoboda et al. (2000). Furthermore, two other clinical entities, ICCA and BFIC2, were assigned to a region encompassing the centromere of chromosome 16 (Lee et al. 1998; Hattori et al. 2000; Swoboda et al. 2000; Caraballo et al. 2001; Weber et al. 2004). Since all of these loci were confined to a relatively small region, it is likely that all of these paroxysmal movement disorders actually belong to one disorder and are allelic, as suggested previously (Tomita et al. 1999).

Fig. 4
figure 4

The PKC-critical region (PKCCR) summarized by five mapping studies (Tomita et al. 1999; Bennett et al. 2000; Swoboda et al. 2000; Cuenca-Leon et al. 2002, present study), as well as a seemingly second PKC locus (EKD2) by Valente et al. (2000). The location of markers and intermarker distances are from the Généthon map (Dib et al. 1996)

We searched for mutations in almost all protein-coding genes mapped at the PKCCR. In addition, we also analyzed four ion-channel-related genes (CACNG3, SCNN1B, SCNN1G, and CNGB1), albeit located outside the PKCCR, since many episodic neurologic disorders, such as muscle diseases, epilepsy, and movement disorders, are known as ion-channel abnormalities (Bhatia et al. 2000). However, 14 coding exons in seven genes (Table 3) were not analyzed because of difficulties in PCR-amplification.

Table 3 List of exons that have not been sequenced in the mutation analysis

In the present study on a total of 157 genes, we failed to identify any causative mutations that can explain PKC in all of the seven families examined. However, two nonsynonymous substitutions, 6186C>A in exon 3 of SCNN1G and 45842A>G in exon 29 of ITGAL, which were co-segregated with PKC in Families C and F, respectively, which were not found in normal control individuals, might be implicated in PKC. In other words, they were not able to be totally ruled out from the candidacy for PKC. It thus remains to be investigated whether another mutation in either gene is found in other PKC families.

Although the mapping of PKC was successful in at least nine studies, causative mutations have been uncovered. This may imply that PKC is caused by aberrations other than exonic mutations, such as a deletion or insertion, in the promoter regions, including the 5′-UTR or 3′-UTR. However, there is still a possibility for usual exonic mutations in a novel gene not annotated in public databases. As PKC itself is, generally, a viable disorder with which patients may show high reproductive fitness, such a mutated allele may be transmitted through many generations. A chromosomal rearrangement is another possibility. The pericentromeric region of chromosome 16 has a large heterochromatin (C-band) block that contains several duplicated regions, through which, frequent chromosomal rearrangements occur (Loftus et al. 1999). It remains also to be seen whether PKC patients within a family share such a variant.