Genetic spectrum of Charcot–Marie–Tooth disease associated with myelin protein zero gene variants in Japan

Abstract We aimed to reveal the genetic features associated with MPZ variants in Japan. From April 2007 to August 2017, 64 patients with 23 reported MPZ variants and 21 patients with 17 novel MPZ variants were investigated retrospectively. Variation in MPZ variants and the pathogenicity of novel variants was examined according to the American College of Medical Genetics standards and guidelines. Age of onset, cranial nerve involvement, serum creatine kinase (CK), and cerebrospinal fluid (CSF) protein were also analyzed. We identified 64 CMT patients with reported MPZ variants. The common variants observed in Japan were different from those observed in other countries. We identified 11 novel pathogenic variants from 13 patients. Six novel MPZ variants in eight patients were classified as likely benign or uncertain significance. Cranial nerve involvement was confirmed in 20 patients. Of 30 patients in whom serum CK levels were evaluated, eight had elevated levels. Most of the patients had age of onset >20 years. In another subset of 30 patients, 18 had elevated CSF protein levels; four of these patients had spinal diseases and two had enlarged nerve root or cauda equina. Our results suggest genetic diversity across patients with MPZ variants.


| INTRODUCTION
Myelin protein zero (MPZ) protein is a major structural component of myelin and encoded by MPZ gene, which is expressed by Schwann cells. 1 MPZ protein is classified as a member of immunoglobulin superfamily and an essential membrane protein comprising 248 amino acids. 2 The final structure of MPZ protein consists of three domains: extracellular domain comprising 124 amino acids, transmembrane domain comprising 26 amino acids, and intracellular domain comprising 69 amino acids located at the C-terminus. 3,4 Charcot-Marie-Tooth disease (CMT) is the most common inherited peripheral neuropathy. CMT is commonly divided into two groups: demyelinating type with slower median nerve conduction velocity (<38 m/s) and axonal type with maintained median nerve conduction velocity (>38 m/s). 5 MPZ variants contribute to the cause of demyelinating neuropathy CMT1B (OMIM 118200) or axonal neuropathy CMT2I/J (OMIM 607677/607736) and also the more severe, juvenile-onset Dejerine-Sottas syndrome (OMIM 145900) and hypomyelinating neuropathy, congenital, 2 (OMIM 618184). 1,6 Moreover, MPZ variants are associated with dominant intermediate Charcot-Marie-Tooth disease D (CMTDID) (OMIM 607791). 7 The phenotype of CMT caused by MPZ variants varies from severe pediatric onset to mild adult onset. 1 To date, about 250 variants of this gene have been described as the cause of inherited peripheral neuropathy (https://portal.biobaseinternational.com/hgmd/pro/). There are limited studies that analyzed large number of patients with MPZ variants. 6,8 Our laboratory analyzed the genetic spectrum of Japanese patients with CMT. 9 In this study, we investigated 85 patients to clarify the genetic spectrum of inherited peripheral neuropathy associated with MPZ variants in Japan. In addition, we also investigated the age of onset, cranial nerve involvement, serum creatine kinase (CK), and cerebrospinal fluid (CSF) protein in 77 patients with reported and novel pathogenic variants.

| Subjects
We examined 1657 Japanese patients who were considered to have inherited peripheral neuropathy from April 2007 to August 2017. All patients and family members provided written informed consent to participate in the study. Before starting this study, patients suspected to have demyelinating CMT with median motor nerve conduction velocity (median MCV) below 38 m/s were checked for duplication or deletion of PMP22 using fluorescence in situ hybridization or multiplex ligation-dependent probe amplification, and patients with duplication or deletion of PMP22 were excluded. Clinical information and blood/DNA samples were collected by neurologists or pediatricians and referred to our genetic laboratory at Kagoshima University Hospital. Using the Gentra Puregene Blood kit (QIAGEN), genomic DNA derived from patients and their families was extracted from peripheral blood cells according to the manufacturer's instructions.

| Microarray sequencing and whole-exome sequencing
From April 2007 to April 2012, variant screening was conducted in 417 patients using customized MyGeneChip, CustomSeq, Resequencing Array (Affymetrix, Inc.), targeting 28 disease-causing or related genes of CMT. We have described the procedure of sequencing and data analysis previously. 10 However, this methodology could not identify some variants due to the false negative hybridization and a low-detection efficiency of the DNA microarray in our laboratory. 11 Thus, we combined whole-exome sequencing to overcome these issues. Whole-exome sequencing was performed by HiSeq2000 (Illumina Inc., San Diego). Using the Burrows-Wheeler Aligner, we aligned the sequences to human genome reference (NCBI37/hg19) and used SAM tools (http://www.htslib.org) for calling the variants.
The called variants annotation was performed using CLC Genomic Workbench software program (Qiagen, Hilden, Germany) and an inhouse script. Whole-exome sequencing was performed as indicated in the previous study. 12

| Targeted resequencing
In May 2012, we introduced the Illumina MiSeq platform (Illumina Inc.), targeting all coding exons and exon-intron junctions of 60 disease-causing or candidate genes of inherited peripheral neuropathies.
We have described this system previously. 13  To analyze the copy number variations of MPZ, we screened the 803 patients using CovCopCan software. 14

| Data analysis and variant interpretation
All MPZ variants were checked against the Human Gene Mutation Database (https://portal.biobaseinternational.com/hgmd/pro/gene).
We then used Sanger sequencing to validate the suspected variants, and segregation analysis was conducted where possible. Variants were classified according to the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/ AMP) guidelines published in 2015. 15 The types and frequency of reported MPZ variants in our study were compared with previous studies. We also referred to the reports  Figure S1 shows the schematic diagram of this study.

| Analysis of variants
In 1657 Japanese patients with suspected inherited peripheral neu- Asp75Gly, and p.Ile112Val were classified as axonal CMT (Table 3).
We       variants had similar phenotype as the reported variants, which are occurred near or at the same codon. Patients with p.Glu37Lys, p.
Arg36Gly, or p.Arg36Trp had adult onset and they were classified as axonal CMT. Patients with p.Asp75Gly or p.Asp75Val also had the similar phenotype. Although, patients with p.Asn122Asp or p.
Asn122Ser were associated with adult onset, the electrophysiological classification varied in each patient. Meanwhile, patients with p.
A patient had compound heterozygous variant (p.Asp75Val/p. The patient's daughter did not have weakness, sensory impairment and decreased tendon reflexes, and denied to undergo electrophysiological examination. Presently, the age of the daughter is 50 years old and she does not have symptoms associated with neuropathy. The nephew's age of onset (56 years) was older than that of the patient (56 years old vs 30 years old; Figure S4).

| Clinical and laboratory findings
We assessed 77 patients with inherited peripheral neuropathy comprising 64 with reported MPZ variants and 13 with novel pathogenic MPZ variants (Table S2). The onset age of these patients indicated a bimodal distribution (Figure 2(A)). Prominent clustering in the first decade and slight clustering between the third and fifth decade were evident, in line with large genetic profiles of Japanese CMT patients. 9 Cranial nerve involvement was confirmed in 20 patients (Table S3). Dysarthria was detected in seven patients, while dysphagia was confirmed in four patients. Hearing loss was also detected in four patients ( Table 4). The most common MPZ variant in patients Asp61Asn, p.Asp75Val, p.Phe19Ser/p.Asp75Val, p.Gly103Glu, p.
Ile112Val, p.Ile114Thr, p.Thr124Met, p.Asp128Asn, p.Lys130Arg, and p.Leu170Arg showed symptoms associated with cranial nerve dysfunction. The clinical information of patient with p.Thr124Met has been described elsewhere. 18 We analyzed serum CK levels in 30 patients. Of them, eight (26.7%) showed elevated CK levels, with the levels being <1000 U/L in most cases. Most of the patients with elevated CK levels had neuropathic symptoms in their middle age (Figure 2(B)). The proportion of patients with adult onset was greater in the elevated CK group than in the normal CK group (p = 0.039) (Figure 2(C)). However, patients with elevated CK were not statistically associated with axonal CMT (p = 0.57) (Table S4).
We analyzed CSF protein levels in  (Table S5).

| DISCUSSION
We investigated 85 patients with inherited peripheral neuropathy associated with MPZ variants in Japan. In this study, we focused on the distribution of MPZ variants in the world to compare Japanese patients with known MPZ variants included in our case series. Interestingly, there were differences in the types of MPZ variants between Japan and other countries. In the present study, we confirmed 13 variants, which have been reported only in Japan. However, one of the 13 variants (p.Leu48Val) was reported from Russia. 19 Therefore, patients with 12 variants were considered to be concentrated in Japan. Three of the 12 variants (p.Asp75Val, p.Gly93Glu, and p. Leu170Arg) were also detected in our study. Patients with p.Asp75Val were frequently observed and described in a study of axonal CMT in Japan. 20,21 p.Gly93Glu was detected in a Japanese CMT1B family with low-nerve conduction velocities. 22  There are several points to consider in this study. First, we analyzed the patients with novel variants and assessed their pathogenicity in accordance with the ACMG guidelines. Although the exact pathogenicity of novel variants should be assessed by functional studies, we were unable to perform functional studies for novel MPZ variants. Also, we were unable to perform the clinical assessment for severity such as CMT neuropathy score. Further, we were able to analyze serum CK, CSF protein levels and MRI findings only in limited patients. Due to the design of this study, these data were insufficient in this study. These points will be addressed in future studies. and Japan society for the promotion of science (26461275, 18H02742).

CONFLICT OF INTEREST
The authors declare no financial or other conflicts of interest.

ETHICAL STATEMENT
The study protocol was reviewed and approved by the Institutional Review Board of Kagoshima University. All patients and family members provided written informed consent to participate in the study.

PEER REVIEW
The peer review history for this article is available at https://publons. com/publon/10.1111/cge.13881.

DATA AVAILABILITY STATEMENT
All data generated or analysed during this study are included in this published article and its supplementary information files.