Introduction

Hereditary motor and sensory neuropathy with proximal dominancy (HMSN-P) was first reported by Takashima et al. (1997) as an autosomal dominant form of hereditary motor and sensory neuropathy with dominant proximal involvement (MIM 604484). It was characterized by autosomal dominant inheritance, slowly progressive proximal neurogenic atrophy and weakness, and sensory disturbance such as paresthesia and vibration loss, leading to bedridden incapacity. Neuropathological examination revealed a decrease in the number of anterior horn cells and marked loss of myelinated fibers in the posterior funiculus. The disease locus underlying HMSN-P was mapped to 3p14.1-q13 with LOD scores 4.04 and 3.10 at D3S1284 and D3S1591, respectively (Takashima et al. 1997). The subsequent fine mapping narrowed the candidate region to a 3.1 cM interval between D3S1591 and D3S1281, which certainly lies on 3q13.1, and the existence of a single founder was also suggested by information about shared marker alleles among patients and the geographical specificity of the disease (Takashima et al. 1999). The same group has recently refined the candidate region for HMSN-P in the coordinate system of the human genomic DNA sequence (H. Takashima et al. unpublished data).

Since HMSN-P was found in Okinawa, Japan, this disease has been regarded as having clinical features resembling those of Kennedy syndrome (MIM 313200), with the addition of an autosomal mode of inheritance; it is sometimes referred to as a part of the spectrum of Charcot–Marie–Tooth disease (CMT2P) (Gemignani and Marbini 2001). Thus far, however, HMSN-P has not been reported in other areas in the world. Recently, we newly found two large families with multiple affected members developing symptoms similar to those of HMSN-P in a small mountain village in the Kansai area of the mainland of Japan. Detailed electrophysiological and electromyography (EMG) examinations suggest that the disease is neurogenic as seen in HMSN-P found in Okinawa (Materials and methods). Moreover, neuropathological examination revealed extensive neuronal loss in the anterior horn of the spinal cord. Thus, we refer to the disease as the “Kansai-type” HMSN-P and the original HMSN-P in Okinawa as the “Okinawa-type.”

Although we did not find any clinical heterogeneities and inconsistency of the mode of inheritance between the Kansai-type and Okinawa-type, there was no record of immigration from Okinawa to the western village (or vice versa) or affinal connection between them. This poses a question about the genetic heterogeneity, namely, a question whether the disease locus of the Kansai-type overlaps with that of the Okinawa-type. Therefore, to identify the disease locus for the Kansai-type, we decided to begin by examining linkage of the Kansai-type HMSN-P with eight microsatellite markers covering the candidate locus (3q13.1) for Okinawa-type. We report the results of linkage analysis and of the identification of a disease haplotype shared among affected members in the pedigrees studied.

Materials and methods

Families and clinical features

Two large families with affected individuals were located in the Kansai area, ∼1,200 km from Okinawa (Fig. 1). These families consist of 5–6 generations including at least 40 affected members (Fig. 2). The inheritance pattern of the disease in the pedigrees exhibits an autosomal dominant mode of inheritance (Fig. 2). Moreover, this disease showed no obvious clinical heterogeneity with the HMSN-P in Okinawa (Table 1), which shows malignant fasciculation associated with a disease process affecting peripheral motor nerves. Electrophysiological evaluation, including a nerve conduction study and EMG in the upper and lower extremities, revealed that compound muscle action potential (CMAP) amplitudes decreased while conduction velocities remained relatively intact and that sensory nerve action potential (SNAP) amplitudes also decreased or were absent due to involvement of the dorsal root ganglia. Examination via electromyography of affected muscles revealed prominent spontaneous activity, in the form of fibrillation potentials and positive sharp waves. Decreased recruitment of motor unit action potentials (MUAPs) with high amplitude and long-duration motor units was observed, evidence of prior axonal loss.

Fig. 1
figure 1

Geographic distributions of the two types of HMSN-P in Japan. The two parts are 1,200 km from each other, separated by the sea. Thus far, in the small mountain village in the western part of Japan, no other family with the disease has been found

Full size image
Fig. 2
figure 2

Two large pedigrees with “Kansai-type” HMSN-P

Full size image
Table 1 Clinical information on patients with Kansai-type and Okinawa-type HMSN-P
Full size table

Subjects, DNA extraction, and microsatellite typing

A total of 16 affected and 31 unaffected members from the two families with HMSN-P in a mountain village participated in this study. In order to compare the disease haplotypes of the Kansai-type with Okinawa-type HMSN-P cases, we also genotyped an affected Okinawa kindred sib-pair. Informed consent was obtained from all patients and healthy volunteers, and the study was approved by all of the participating institutes. Genomic DNA was extracted from lymphocytes of peripheral blood using a QIAamp DNA Blood Maxi Kit (Qiagen, Valencia, CA, USA) as per the manufacturer’s instructions. Hereafter, we will refer to each individual member using the corresponding identifier of Pedigree:Generation:Sample in Fig. 2. Of eight microsatellite markers, seven were selected around 3q13.1 from the GDBS database (http://jbirc.jbic.or.jp/gdbs/top.jsp) (Tamiya et al. 2005), and one was novel. We genotyped the eight markers for linkage analyses on the ABI3700 and all the information regarding these microsatellite makers is listed in Table 2. In order to refine a position of obligate recombination observed in Pedigree 1, we also examined four additional microsatellite markers between D3S0490i and D3S1496i. We also carried out confirmatory examination for D3S1291 (5′-TTT AAC AAT CCC AAT CAG G-3′ for forward primer and 5′-GGC AAA ATG TCT GCT GTA-3′ for reverse primer) in a few members in the Kansai-type families. The genotype data at D3S1291 in Okinawa-type patients was compiled from the previous works (Takashima et al. 1997, 1999). The PCR condition and fragment analysis were according to a standard procedure described previously (Tamiya et al. 2005).

Table 2 Information on the microsatellite marker on chromosome 3p12.2–3q13.11 for linkage analysis
Full size table

Linkage and haplotype analysis

We used eight microsatellite markers located on the chromosomal region 3p12-q13 (Fig. 3, Table 2). Since genetic map positions for all markers except D3S3652 had not been described, we extrapolated sex-averaged map positions for each such marker linearly using the nearest flanking microsatellite markers from the deCODE map (Kong et al. 2002). Mendelian segregation of marker genotypes was confirmed using the “Pedigree Check” program in Pedtool (Fishelson and Geiger 2002). Parametric two-point linkage analysis was performed using SUPERLINK-ONLINE (http://bioinfo.cs.technion.ac.il/superlink-online/) (Fishelson and Geiger 2002; Silberstein et al. 2006). Disease inheritance was assumed to be autosomal dominant with complete penetrance and without phenocopy. We assumed the disease allele frequency to be 0.0001 following the previous study (Takashima et al. 1997). For marker alleles, we assumed equal allele frequencies at each locus.

Fig. 3
figure 3

The intervals co-segregated with the Kansai-type and Okinawa-type of HMSN-P. The genetic distances were linearly interpolated using the nearest flanking markers in the deCODE map. Physical distances were calculated from the coordinate system of the human genomic DNA sequence in NCBI build 35. Black boxes in the map bars represent microsatellite markers used in the linkage analysis. Gray boxes represent markers for additional typing. Open boxes represent markers used in previous work (Takashima et al. 1997, 1999). The closed circle indicates the centromere

Full size image

Results

Two-point LOD scores are shown in Table 3. In Pedigree 1, six of the eight-microsatellite markers showed significant LOD scores (LOD > 3.3) exhibiting evidence of linkage of the 3p12-q13 region to the Kansai-type HMSN-P and the Okinawa-type. The maximum LOD scores for the two pedigrees were 6.72 in Pedigree 1 and 1.72 in Pedigree 2, obtained at the same marker D3S1592i with zero recombination fraction, θ = 0. Combining two pedigrees, we see that five markers from D3S1496i to D3S3652 showed the clear evidence of linkage to the Kansai-type HMSN-P (Table 3).

Table 3 Two-point LOD scores between phenotype and markers on chromosome 3p12.2–3q13.11
Full size table

To infer a disease haplotype (i.e., a haplotype cosegregating with disease) based upon the evidence of linkage, we assume that if there is more than one possible diplotype configuration, a configuration with a minimum number of recombination events is more likely. We abbreviate this assumption to the MREC (minimum recombination events configuration). This assumption may be appropriate for our data with short marker distances and is analogous to the minimum recombinant haplotype configurations in the pedigree considered in Qian and Beckmann (2002). Inspecting nuclear families with affected offspring 1:V:9, 1:V:22, and 1:V:35 for which both parents were genotyped, we find that an eight-marker haplotype inherited from their affected parents is 121222[1]1, where [1] at D3S1083i stands for ambiguity in 1:V:22 under the MREC assumption (Fig. 2). Note here that an obligate recombination between D3S0490i and D3S1496i is observed for an inheritance from an affected parent 1:IV:40 to either of dizygotic twins 1:V:34 or 1:V:35, which resulted in the LOD score of minus infinity at D3S0490i (Table 3). This means that the linkage of D3S0490i to the disease is rejected, so we consider only the remaining seven markers: a candidate disease haplotype is 21222[1]1 from D3S1496i to D3S0532i. Under the MREC assumption, all the affected members in Pedigree 1 may share the same haplotype 21222[1]1 without introducing any additional recombinations in the pedigree (Fig. 2). In Pedigree 2, only one affected member 2:V:4 cannot have this haplotype for the terminal 2 markers (D3S1083i and D3S0532i), resulting in the exclusion of linkage for these markers (Table 3). For the remaining affected members 2:IV:8 and 2:V:2, the seven-marker disease haplotype is 2[1]22211, where [1] at D3S1496i indicates ambiguity in 2:IV:8. Thus, under the MREC assumption, we identify a unique five-marker (from D3S1496i to D3S3652) haplotype 2[1]222 cosegregated with disease and shared by all the affected members of both pedigrees.

We summarize the result of additional genotyping with genotype results for affected siblings of Okinawa-type in Table 4. It is evident that a recombination occurred between D3S0490i and D3S1048i in the inheritance from 1:IV:40 to 1:V:34. The location of this event depends on the paternal allele of 1:V:35 at D3S1070i.

Table 4 Inference of disease haplotype and additional genotyping result
Full size table

Discussion

In this study, we demonstrated that the disease locus for the Kansai-type HMSN-P is strongly linked to 3q13, overlapping with a region previously assigned as a candidate locus for Okinawa-type HMSN-P (Takashima et al. 1999). This coincidence of the disease locus might reflect the clinical overlap of the two diseases. The coincidence of the disease locus and the clinical homogeneity suggest that the causative gene for both types is identical and is located within the 7.3 Mb interval between D3S1488i and D3S1083i (Fig. 3), which includes at least 21 known genes from FLJ35799 to ZPLD1. Our result from the linkage analysis warrants further study, such as an extensive mutation search in expressed genes including these known genes within this interval that co-segregate with the disease.

The existence of a unique disease haplotype shared among all the affected members in the two western families is suggestive of a founder mutation in the western population, as with the case of Okinawa-type (Takashima et al. 1997). It seems, however, that disease haplotypes are not shared between the two types (Table 4). Of four microsatellite markers used in previous works (Takashima et al. 1997, 1999), the precise position of A281WA5 in the coordinate system of the human genomic sequence was never determined, i.e., was unmapped. On the other hand, D3S3654 and D3S1563 did not locate within the interval linked to the Kansai-type HMSN-P (Fig. 3). The disease alleles at D3S1291 in the linked interval seem to be different between Kansai-type (186 in size) and Okinawa-type (188 from the previous works by Takashima et al.), which is consistent with our finding. Such a circumstance is not unusual when the same locus is linked to the disease for different pedigrees. Therefore, the allelic heterogeneity is suggested in the two types of HMSN-P that share the same disease locus, as is often the case for rare Mendelian diseases. We also note that, although HMSN-P has been reported only in Japan, there is a possibility that HMSN-P has already been found in other areas in the world but has been erroneously classified into other disease categories, such as hereditary motor neuropathy (HMN), spinal muscular atrophy (SMA), limb-girdle muscular dystrophy (LGMD), or amyotrophic lateral sclerosis (ALS). Although HMN, SMA, LGMD, and ALS are also progressive muscular atrophic diseases, these diseases do not involve the sensory system primarily. On the other hand, HMSN-P involves peripheral sensory nerves severely when the disease progresses. Diagnosis of HMSN-P is based on existing sensory nerve disturbance both subjectively, and objectively by EMG examination, in addition to proximal dominant muscular atrophy. Until a specific genetic test is available, this hypothesis can be verified by a confirmatory screen using electrophysiological examination in patients with these diseases. The confirmatory screening of the allelic heterogeneity of HMSN-P will be important to facilitate identification of the causative gene and investigation of the gene function in the same scenario as CMT2F and distal HMNs (Evgrafov et al. 2004).