The human genome comprises more than 60 Rab genes that make up one GTPase family; these genes have been studied mostly as molecular switches in intracellular membrane trafficking in various systems [1, 2]. Rab GTPases are key regulators also in animal nervous systems, exclusively in neurons across cell bodies to synapses; preferential subcellular localizations of each Rab protein are unique [3, 4]. Alterations of these Rab GTPases have been recognized to be associated with virtually all neuronal activities in health and disease. A well-known example is RAB7A mutations, which cause the neurodegenerative defects that are associated with the autosomal dominant Charcot-Marie-Tooth disease type 2B (CMT2B) peripheral neuropathy [5-8]. CMT is the most common inherited peripheral neuropathy which is characterized by progressive distal muscle weakness and sensory impairment.

Rab40 proteins, consisting of four paralog members including RAB40B in humans, are an atypical subgroup of Rab proteins with unique structural and functional features. Besides the GTPase domain of the Rab family, Rab40 proteins are distinguished by an unusual domain of suppressor of cytokine signaling (SOCS) box in the C terminal part [9-11]. Group of proteins containing the SOCS box domain have typically been studied as adaptors that recruit substrates into a specific ubiquitination process of the Cullin-RING ligase 5 (CRL5) complex [12, 13]. A variety of SOCS box-containing substrate receptors are involved in each specific target degradation in various biological processes. Four important components of the E3 ligase complex are the CUL5 scaffold protein, elongin B/C, SOCS box-containing adaptors, and ring box protein 2 (RBX2) [12-14].

Among the four paralog members of human Rab40 proteins, RAB40B and RAB40C show preferentially higher expression in the nervous system and muscle tissues. However, little is known about the cellular functions of the Rab40 proteins beyond those structural features, except from studies on cellular adhesion and migration of several carcinoma cell lines. Both human RAB40A and RAB40C were shown to form complexes with CUL5 for proteasomal degradation of p21-activated kinase 4 (PAK4) and receptor for activated C kinase 1 (RACK1), respectively [10, 15]. Similarly, RAB40B-CUL5-dependent ubiquitination was found to regulate epithelial protein lost in neoplasm (EPLIN) localization and promote cell migration and invasion by altering focal adhesion and cytoskeletal dynamics in breast cancer cells [9, 16]. Other components of the CRL5 complex have also been linked to neuronal development, such as PAK4 kinase or RBX2-CUL5 complexes in proper axonal growth or neuronal migration [17, 18]. Thus, the overlapping tissue expression of Rab40s suggests a probability that those proteins may also be involved in neurodevelopmental and/or degenerative processes in the nervous system.

In the present study, we identified one RAB40B nonsense mutation from a CMT family that follows an autosomal dominant inheritance. Because there is no known evidence about neuronal functions of Rab40 proteins, we assessed the in vivo pathogenicity of the human mutation in the nervous system of model organisms, zebrafish and Drosophila. Based on the high sequence similarity and structural conservation between human and model animal Rab40 proteins, we examined the neuropathogenic effects of the RAB40B mutation using ectopic expression models.

Participants

This study examined a Korean dominant CMT2 family with seven members including two affected individuals (family ID: FC162; Fig. 1A). Informed consent was obtained from all participants according to the protocol approved by the Institutional Review Board of Sungkyunkwan University, Samsung Medical Center (2014-08-057-002) and Kongju National University (KNU-IRB-2018-62).

Clinical assessment

Neuromuscular clinical phenotypes including motor and sensory impairments, deep tendon reflexes, and muscle atrophy were obtained in the standardized methods. Strengths of flexor and extensor muscles were assessed manually using the standard medical research council (MRC) scale [19]. To determine the severity of physical disability, we measured the functional disability scale (FDS), CMT neuropathy score version 2 (CMTNSv2), CMT examination score (CMTES), Rasch-modified CMTNSv2 (CMTNSv2-R), and Rasch-modified CMTES (CMTES-R) [20-23]. Sensory impairments were assessed in terms of the level and severity of pain, temperature, vibration, and position. Age at onset was determined by asking patients for their ages when symptoms, i.e., distal muscle weakness, foot deformity, or sensory change, first appeared.

Motor and sensory conduction velocities of median, ulnar, peroneal, and sural nerves were determined by standard methods using the surface stimulation and recording electrodes. Motor nerve conduction velocities (MNCVs) and compound muscle action potentials (CMAPs) of median nerves were determined by stimulating the elbow and wrist. In the same way, the MNCVs and CMAPs of peroneal nerves were determined by stimulating the knee and ankle. Amplitudes of CMAP were measured from baseline to negative peak values. Sensory nerve conduction velocities (SNCVs) were obtained over a finger-wrist segment from ulnar nerves and were also recorded for sural nerves. Sensory nerve action potential (SNAP) amplitudes were measured from positive peaks to negative peaks.

DNA extraction and genetic screening of the CMT family

Total DNA was extracted from blood obtained from participants using a HiGene Genomic DNA Prep Kit (Biofact, Daejeon, Korea). Exome capture and next generation sequencing were carried out using a SureSelect Human All Exon 50M kit (Agilent Technologies, Santa Clara, CA, USA) and HiSeq 2500 Genome Analyzers (Illumina, San Diego, CA, USA), respectively. Frequencies of rare alleles from the whole exome sequencing (WES) data were obtained from public human genome databases of the 1000 Genomes Project (http://www.1000genomes.org), the Genome Aggregation Database (https://gnomad.broadinstitute.org), and the Korean Reference Genome Database (http://coda.nih.go.kr/coda/KRGDB/index.jsp). Possible pathogenic variations were basically evaluated according to the American College of Medical Genetics and Genomics-Associations for Molecular Pathology (ACMG-AMP) guidelines (https://wintervar.wglab.org). Candidate pathogenic mutations found in the FC162 family were confirmed by Sanger sequencing of all participants.

Microinjection in zebrafish

Zebrafish (AB strain) were maintained with a cycle of 13-h light and 11-h dark in an automatic breeding system (Genomic-Design, Daejeon, Korea) at 28.5°C and pH 7.0~7.9, according to animal research protocols approved by the Institutional Animal Care and Use Committee of Samsung Biomedical Research Institute and Sungkyunkwan University (IACUC#20201008001 and IACUC#20200916001). The mutant hRAB40B ORF was subcloned into pCS2+ vector, and its capped mRNAs were synthesized using mMESSAGE mMACHINE SP6 kit (Ambion, Foster City, CA, USA). The in vitro synthesized hRAB40B-Y83X mRNAs (2,000 pg) were introduced into zebrafish embryos at the 1~2 cell stage using a gas-powered microinjection system (PV83 Pneumatic PicoPump, SYS-PV830; World Precision Instruments, Sarasota, FL, USA).

Zebrafish behavior analysis

Moving patterns and velocity of zebrafish larvae at 84 hours post-fertilization (hpf) were analyzed using DanioVision (Noldus, Wageningen, Netherlands). Behavioral analysis was performed using light stimulation in 1 ml zebrafish breeding media. Control zebrafish or hRAB40B-Y83X mRNA-injected zebrafish were placed in individual 24-well plates and their movements were recorded for 30 min and analyzed using EthoVision XT software (Noldus). All statistical analyses were performed using GraphPad Prism 5 (GraphPad, Boston, MA, USA). Values are presented as mean±SD or as fold change relative to the mean control. Differences between control and mRNA injected groups were evaluated using unpaired Student’s t-test.

Immunohistochemistry of zebrafish tissues

Zebrafish larvae at 84 hpf were fixed in 4% paraformaldehyde in 0.1 M PBS overnight at 4°C. The larvae were permeabilized with 0.5% TritonX-100 in 0.1 M PBS (PBST) for 15 min at room temperature and washed with 1× PBS three times and blocked in 1% DMSO and 1% BSA in PBST containing 4% normal goat serum at room temperature for 1 hr. The following primary antibodies were used for incubation: mouse anti-SV2 (1:50; Developmental Studies Hybridoma Bank, Iowa City, IA, USA) and Alexa 647-conjugated α-BTX (B35450, 1:150; Invitrogen-ThermoFisher Scientific, Carlsbad, CA, USA) at 4°C overnight in blocking solution. After washing three times with PBST, larvae were treated with Alexa Fluor 488-conjugated secondary antibodies (A11001, 1:250; Life Technologies, Carlsbad, CA, USA) for 2 hr. at room temperature. The zebrafish larvae were mounted on slides with mounting solution containing n-propyl gallate and images were captured and analyzed with a confocal microscope (LSM 700; Carl Zeiss, Jena, Germany)/Zeiss ZEN imaging software and ImageJ (NIH, Bethesda, MD, USA). NMJs were analyzed by measuring the proportion of areas with merged yellow signals within the region of interest (ROI) of the zebrafish trunk.

Generation of transgenic flies

Human RAB40B cDNA (HG14600-G) was obtained from Sino Biological (Beijing, China). The p.Y83X mutation was introduced by site-directed PCR mutagenesis. RAB40B ORFs with or without the early stop codon were subcloned into pUAST-AttB transformation vector. Constructs were injected into Drosophila embryos for generation of transgenic strains (BestGene, Chino Hills, CA, USA). Genomic landing sites P40 and VK31 were used for site-directed transgene insertions on the second and third chromosome, respectively [24].

Drosophila strains

Rab40 and Cul5 RNAi strains of the TRiP collection were obtained from Bloomington Drosophila Stock Center (BDSC; Bloomington, IN, USA). GAL4 strains were also sourced from BDSC, and used for ectopic expression described by followings: Act5C-GAL4 [25] and tub-GAL4 [26] for ubiquitous expression, elav-GAL4 [27] and nSyb-GAL4 [28] for pan-neuronal expression, OK371 [29], OK307 [30], and D42-GAL4 [31] for expression in motor neurons, ppk-GAL4 for expression in multidendritic neurons [32], and repo-GAL4 for pan-glial expression [33]. Temperature-sensitive mutation of GAL80 repressor and its application for spatiotemporal gene expression were described by McGuire et al. [34]. Each GAL4 strain with the transgene of tub-Gal80^ts was crossed with UAS-hRAB40B flies at 20°C (permissive temperature), and newborn adults carrying all three transgenes were collected for temperature shift to the inducing temperature (30°C).

Drosophila locomotion analysis

Locomotion performance was measured by a climbing assay that uses the negative geotaxis behavior of flies [35]. Flies were collected daily for each genotype, put into groups of 10 to 15 individuals, and transferred to a 30°C chamber to induce transgene expression. Each group was assessed for climbing ability 2, 8, and 15 days after collection. Flies were put into test tubes of 18 cm length. After tapping tubes on a mouse pad to force out flies at the bottom, flies were allowed to climb the wall for one minute. Total numbers of flies that climbed over the marked height of 11 cm in the test tubes in three repeated tests were counted to calculate the climbing percentage. Four or six groups per genotype were assessed to calculate means and standard deviations.

Clinical manifestation in a CMT2 family with nonsense RAB40B mutation

In one CMT family (family ID: FC162), two individuals of mother and son were affected with axonal peripheral neuropathy (II-3 and III-2 in Fig. 1A). The clinical and electrophysiological features of the patients are shown in Table 1. Onset ages were 13 and 3 years old in mother and son, respectively. Muscle weakness and atrophy started and predominated in the distal portions of legs, and were noted to a lesser extent distally in upper limbs. Vibration sense was reduced to a greater extent than pain in the III-2 patient. Foot deformity and absent deep tendon reflex were shown in both patients, but scoliosis was found only in the mother (II-3). Functional disability was similar (FDS: 3), and CMTNSv2 was in the moderate category in both patients (17 and 13, respectively). However, when considering the disease duration, the son was more severely affected than his mother. Motor and sensory conduction studies were carried out in both patients (Table 1). Electrophysiological findings confirmed that both patients were axonal type of CMT2 neuropathy.

In the mutation screen by WES for the FC162 family, we identified a nonsense mutation in the RAB40B gene (rs781464100, c.249C>A, p.Y83X, Fig. 1B) specific to the affected members. The early stop codon of the deduced sequences is expected to generate a truncated product that lacks its C-terminal SOCS box domain of the atypical Rab protein (Fig. 1C). Although no other candidate mutation was identified in the exome sequencing for the familial neuropathic phenotype, more evidences are required to evaluate the pathogenicity of the RAB40B mutation.

In sequence comparison of human RAB40B to zebrafish and Drosophila homologs, high sequence similarity was confirmed along the GTPase domain and the SOCS box sequences, as well as around the nonsense mutation site (Fig. 1C). Based on the sequence and structural homology between humans and the model animal Rab40 proteins, we examined the in vivo pathogenicity of the human mutation using zebrafish and Drosophila platforms.

Transient transformation with human RAB40B-Y83X induced motility defects in zebrafish

The RAB40B mutation in the FC162 family with the neuropathic phenotypes were manifested with the nonsense mutation, which followed an autosomal dominant inheritance (Fig. 1A). Based on the inheritance pattern and the presumptive pathogenic cause of the RAB40B mutation, we examined ectopic expression phenotypes of the human mutation in model organisms of zebrafish and Drosophila.

Capped mRNAs of human RAB40B with the early stop mutation were prepared in vitro and introduced into zebrafish embryos for transformation with the probable pathogenic sequences. Using the ectopic expression model of zebrafish, we observed that exogenous human mutant hRAB40B led to nonsignificant developmental delay (Fig. 2A). But when we investigated the effect of human mutation on motility by testing the transformant’s swimming pattern. We noticed that 84 hpf-larvae transformed by the mutant hRAB40B stalled with restricted localization, whereas control zebrafish swam throughout the tested area (Fig. 2B). The zebrafish locomotion velocity was measured, and we found that the hRAB40B-Y83X transformants moved more slowly than control and Mock-expressing fish (Fig. 2C).

Based on the larval motility defect, we examined neurodevelopmental phenotypes induced by the hRAB40B-Y83X transformation in formation of neuromuscular junctions (NMJs). In zebrafish, SV2 (synaptic vesicle protein 2) staining reveals synaptic vesicles in motor axons and terminals; thus, staining has been used to label the presynaptic terminals of NMJ. Neuromuscular synapses can be visualized when presynaptic SV2-stained vesicle clusters (green in Fig. 2D) are opposed to postsynaptic α-BTX (bungarotoxin)-stained AChR clusters (red in Fig. 2D) [36, 37]. We found that hRAB40B-Y83X-transformant zebrafish were labeled less often by both pre- and postsynaptic markers in the larval NMJs (Fig. 2D, E). In addition, we analyzed the extent of NMJ innervation in the NMJ areas of hRAB40B-Y83X larvae (Fig. 2F, G) with those of control or Mock-expressing larvae. The zebrafish expressing hRAB40B-Y83X had less innervation signals than the control and Mock-expressing fish in both presynaptic and postsynaptic regions of the NMJ (Fig. 2F, G). This neurodevelopmental phenotype in NMJ formation is consistent with motility defects induced by transformation with the human mutant sequences, suggesting a possible neuropathic causality of the RAB40B mutation.

Ectopic expression of hRAB40B-Y83X reproduced defective motor performance in Drosophila

A stable transgenic model of the human RAB40B mutation was established in Drosophila to examine neuropathic impairments by mutant sequences beyond developmental phenotypes. Expression of the hRAB40B in the transgenic embryos were confirmed by RT-PCR (Fig. 3) Transgenic strains of the human RAB40B gene or its nonsense mutant form were prepared by site-directed integration of corresponding transgenes [24]. First, ectopic expression of those human genes, as well as RNAi knockdown of Drosophila Rab40, were examined for developmental effects by ubiquitous expression driven by Act5c-GAL4.

Semi-lethality was shown by ubiquitous expressions of both the wild and mutant hRAB40B (18.1% and 27.1% failure during the larval and pupal development of Act>hRAB40B (n=199) and Act>hRAB40B-Y83X (n=236), respectively); however, majority of the ectopic expression progeny succeeded in development to adults. The adult survivors showed no detectable morphological defects, except reduced lifespans (half survival rate around 20 days after hatching). In addition, delayed larval development was shown by expression of both forms of human RAB40B, as well as by RNAi knockdown of the Drosophila homolog (from two- to four days-delay on average, depending on culture vials). As for the downregulation of Drosophila Rab40, we established a balanced stock that expressed its dsRNAs ubiquitously and constantly. Except for delayed development for few days as well, the interim stock showed no detectable phenotype for generations, suggesting that the Drosophila Rab40 by itself is near a dispensable gene for normal development.

Parallel to the motility test of zebrafish transformants, we examined the larval locomotive abilities of the progeny of ectopic expressions of hRAB40B genes or Rab40 RNAi knockdown in Drosophila. Due to the delayed larval growth depending on genotypes, the early third instar larvae for the locomotion test were collected based on body size from groups of different culture days. However, it was hard to find any distinguishable difference in larval crawling ability among different genetic conditions of hRAB40B or Drosophila Rab40, including expression of human mutation (data not shown).

Then, in adult animals, we tested the dominant pathogenicity of the human mutation to find similar degenerative hallmarks by its ectopic expression in Drosophila. To avoid any developmental effects of hRAB40B expression, we applied a temporal expression system using a temperature-sensitive mutation of GAL80 (tub-Gal80^ts) for ectopic expression of the transgenes [34]. Under the tub-Gal80^ts system, locomotion abilities of flies expressing hRAB40B at the inducing temperature after hatching were examined to recapitulate neuropathic phenotypes of the human mutation. The locomotion ability was assessed quantitatively by a climbing assay. First, the human RAB40B was expressed ubiquitously by tub-GAL4. When comparing motor abilities after expression of the wild and mutant hRAB40B, we were able to discern a difference between those ectopic expressions, with a progressive decline of motor performance when mutant hRAB40B was expressed (Fig. 4A, 4B, tub-GAL4). Further, we examined the constant expression of Rab40 dsRNAs on the motor performance by adult ages. Differently from temporal expression of the human mutation, the constant downregulation of the Drosophila homolog itself did not show a locomotive decline by ages (data not shown), suggesting that the defective motor performance by the mutant hRAB40B is a gain-of-function phenotype of the mutation rather than a loss-of-function condition of Rab40.

Next, we sought to determine whether the progressive motor defect was of neuron-origin. Several specific GAL4 drivers were introduced for ectopic expression in the nervous system, such as pan-neuronal GAL4s, elav-GAL4 and nSyb-GAL4, and neuron type-specific GAL4s like OK371 or OK307 for expression in motor neurons [29, 30], ppk-GAL4 for expression in sensory neurons [32], and D42-GAL4 for expression in synapses with the musculature [31]. However, we failed to find a GAL4 driver among those that could reproduce the locomotion defect shown by ubiquitous expression of the mutant hRAB40B (Fig. 4A, 5).

We also tested expression in glial cells, the other cell type in the nervous system. A similar behavioral defect was found when the pan-glial GAL4, repo-GAL4 [33] was applied for the mutant hRAB40B expression (Fig. 4B). Ectopic expression driven by repo-GAL4 exactly repeated the progressive decline of motor performance shown by ubiquitous expression of the human mutation. Thus, the deteriorating locomotion ability of the mutant hRAB40B was not caused by its neuronal expression, but was reproduced under control of the pan-glial GAL4, suggesting an association of the Drosophila glial system with the behavioral phenotype.

Cul5 downregulation enhanced defective motor phenotype of mutant hRAB40B expression

The unique SOCS box domain of Rab40 proteins induces physical interactions with CUL5 and other essential components of the CRL5 complex to target specific substrates of the ubiquitination system [12, 13]. The pleiotropic developmental functions of Drosophila Cul5 have been reported predominantly in the female germ line, as well as in larval neuromuscular junctions [38, 39]. To examine the prepotent molecular pathway underlying the defective motor performance caused by the C-terminally truncated human RAB40B, we tested Cul5 knockdown to modify the mutant hRAB40B-induced motor phenotypes.

Since ubiquitous downregulation of Cul5 led to a partial developmental failure (data not shown), epistatic effects of the Cul5 knockdown were also assessed in the tub-Gal80^ts system. Under control of the pan-glial repo-GAL4, downregulation of Drosophila Cul5 was found to synergistically enhance the defective motor phenotype of the mutant hRAB40B (Fig. 4C). Though a direct physical interaction should be identified between the human product and the Drosophila ubiquitination unit, this genetic interaction result suggests a possibility that the locomotion defect caused by the human mutation may be associated with the CRL5-dependent proteolytic regulation that is conserved between Drosophila and vertebrates.

In this study using ectopic expression models of zebrafish and Drosophila, we assessed in vivo effects of a RAB40B mutation observed in a family with autosomal dominant axonal neuropathies. By introduction of the human mutation, we were able to identify behavioral phenotypes similar to human neuropathic symptoms such as a motility defect of zebrafish larvae and a progressive decline of locomotion ability in Drosophila. This is the first evidence that an alteration of Rab40 proteins is associated with clinical phenotypes, which is consistent with preferential expression of the atypical Rab protein in the human nervous system.

Despite the successful establishment of animal models for the RAB40B nonsense mutation, there are presently many model limitations when analyzing human neuropathic symptoms. Distinct neurodevelopmental phenotypes were manifested in the transformed zebrafish, such as defective larval motility and reduced synaptic marks in NMJs (Fig. 2); however, in the Drosophila transgenic model, a probable defect in larval crawling ability was below the detection limit, despite a portion of the developmental failure being due to hRAB40B expression. Instead, in adult flies, we were able to reproduce an ectopic expression-induced progressive deterioration of locomotion ability (Fig. 4). Moreover, despite the phenotypic similarity of the ectopic expressional traits to the human neuropathic symptoms, we do presently not have enough evidence of the molecular pathways underlying the pathogenic phenotypes by the RAB40B mutation, except for genetic interaction with the Cul5 knockdown in the Drosophila model. For this reason, strict evaluation of our established models will require further work, such as screening of target substrates that are regulated by the probable RAB40B-involved proteolytic regulation.

In the present study, we utilized zebrafish transformants and Drosophila transgenic animals. The human genome comprises four paralogs of Rab40s, and zebrafish and Drosophila retain only two and a homolog of Rab40 genes, respectively. Because these paralogs are possibly involved in different biological processes despite those high homologies, the zebrafish system may be suitable for detailed analysis of human mutation. It will thus be promising in future to develop transgenic models of zebrafish or other mammalian organisms.

Rab40 proteins, including human RAB40B, share a unique C-terminal SOCS box domain, which is a key motif to assemble the CRL5 ubiquitin ligase complex [12, 13]. Thus, mutations across the binding motif may cause a dominant-negative result against proteolytic regulation of specific targets of the Rab40 adaptors [9, 10]. In this study, we identified a genetic interaction between ectopic expression of the SOCS box-deficient hRAB40B and Drosophila Cul5 knockdown that enhances locomotion defects (Fig. 4C). When considering neurodevelopmental and degenerative phenotypes caused by alterations of human and Drosophila CUL5s [17, 18, 38, 40], this result supports functional conservation of the specific proteolytic regulation in the nervous system from Drosophila to human, and suggests that the dominant locomotive defects induced by the SOCS box-lacking RAB40B mutation may be associated with the conserved process of CRL5 ubiquitination. Though the direct interaction of the human product with Drosophila CRL5 complex remains to be clarified, the epistatic behavioral modification may be a significant clue to trace the molecular pathogenesis of the human mutation.

In the Drosophila model, the defective motor performance induced by the hRAB40B mutation was displayed not by neuronal GAL4 drivers, but was reproduced under control of the pan-glial repo-GAL4. The glial systems of Drosophila and vertebrates share crucial cellular functions of metabolic sustenance and immune support, and in regulation of synaptic plasticity [41-43]. However, the Drosophila glia comprises only limited cell numbers in adult CNS when compared to mammalians, further, Drosophila lacks myelin sheaths and Schwan cells that are exclusively developed in the vertebrate nervous system. Due to these reasons, Drosophila models for neuropathic phenotypes have been established mostly for the axonal type CMT, whereas demyelinating disorders are a representative pathogenesis for human peripheral neuropathies. However, a mitochondrial disorder in a group of glial cells was recently reported to induce progressive locomotion defect in Drosophila [44]. Like the SOCX mutants, a group of neuropathic phenotypes may be caused from glial defects but independent on demyelination. Besides developmental roles of glial cells in organization of the nervous system, little is known about glial functions for degenerating phenotypes in the adult nervous system. Though more evidence for molecular pathogenesis of the hRAB40B mutation should be elucidated to address glial causality of the mutation, the present study may provide a novel example of neuropathic phenotypes caused by an alteration in the glial system.

Taken together, our results so far evidence a probable pathogenic gain-of-function of the RAB40B mutation lacking the C-terminal SOCS box, and show that ectopic expressional locomotive defects in model animals may be associated with an alteration of the CRL5 complex-mediated proteolytic regulation in the glial system. Nevertheless, it seems that further studies are necessary to solve the controversies between the locomotion defect in the flies of pan-glial expression of the hRAB40B mutation and the axonal defective CMT phenotype in the examined family.

This study was supported by the National Research Foundation of Korea (2021R1A4A2001389, 2021R1A6A3A123041249, 2021R1A2C3004572, and 2022R1I1A2068995), Korean Health Technology R&D Project, Ministry of Health and Welfare (HR22C1363), and Future Medicine 2030 Project of the Samsung Medical Center (SMX122005).

Clinical and electrophysiological features in CMT2 patients of FC162 family