Skip to main content
PMC home page
Proceedings of the National Academy of Sciences of the United States of America logoLink to Proceedings of the National Academy of Sciences of the United States of America
. 2009 Sep 29;106(41):17528–17533. doi: 10.1073/pnas.0905523106

SH3TC2/KIAA1985 protein is required for proper myelination and the integrity of the node of Ranvier in the peripheral nervous system

Estelle Arnaud a, Jennifer Zenker a,b, Anne-Sophie de Preux Charles a, Claudia Stendel c, Andreas Roos d, Jean-Jacques Médard a, Nicolas Tricaud c, Henning Kleine e, Bernhard Luscher e, Joachim Weis f, Ueli Suter c, Jan Senderek c,d,f, Roman Chrast a,1
PMCID: PMC2765159  PMID: 19805030

Abstract

Charcot–Marie–Tooth disease type 4C (CMT4C) is an early-onset, autosomal recessive form of demyelinating neuropathy. The clinical manifestations include progressive scoliosis, delayed age of walking, muscular atrophy, distal weakness, and reduced nerve conduction velocity. The gene mutated in CMT4C disease, SH3TC2/KIAA1985, was recently identified; however, the function of the protein it encodes remains unknown. We have generated knockout mice where the first exon of the Sh3tc2 gene is replaced with an enhanced GFP cassette. The Sh3tc2ΔEx1/ΔEx1 knockout animals develop progressive peripheral neuropathy manifested by decreased motor and sensory nerve conduction velocity and hypomyelination. We show that Sh3tc2 is specifically expressed in Schwann cells and localizes to the plasma membrane and to the perinuclear endocytic recycling compartment, concordant with its possible function in myelination and/or in regions of axoglial interactions. Concomitantly, transcriptional profiling performed on the endoneurial compartment of peripheral nerves isolated from control and Sh3tc2ΔEx1/ΔEx1 animals uncovered changes in transcripts encoding genes involved in myelination and cell adhesion. Finally, detailed analyses of the structures composed of compact and noncompact myelin in the peripheral nerve of Sh3tc2ΔEx1/ΔEx1 animals revealed abnormal organization of the node of Ranvier, a phenotype that we confirmed in CMT4C patient nerve biopsies. The generated Sh3tc2 knockout mice thus present a reliable model of CMT4C neuropathy that was instrumental in establishing a role for Sh3tc2 in myelination and in the integrity of the node of Ranvier, a morphological phenotype that can be used as an additional CMT4C diagnostic marker.

Keywords: Charcot–Marie–Tooth disease, peripheral neuropathy, Schwann cell

With an estimated prevalence of 1 in 2,500 persons, Charcot–Marie–Tooth (CMT) neuropathies [also called hereditary motor and sensory neuropathies (HMSNs)] are among the most common inherited neurological disorders (1). Over the last 25 years, progress in genetics has led to the identification of more than 30 genes responsible for different CMT forms (www.molgen.ua.ac.be/CMTMutations). Identification of these genes, which encode proteins involved in various glial and axonal functions, contributed tremendously both to basic understanding of peripheral nerve function and to genetic counseling in affected families.

CMT4C is a demyelinating, autosomal recessive CMT neuropathy. The clinical manifestations observed in CMT4C patients include delayed age of walking, muscular atrophy, areflexia, sensory impairment, foot deformities (pes cavus), and reduced motor and sensory nerve conduction velocities. The histopathological examination of patient sural nerve biopsies revealed a demyelinating type of neuropathy. Electron microscopic analysis also demonstrated the presence of nonmyelinating Schwann cell complexes with abnormal cell processes and demyelinated and remyelinated axons surrounded by onion bulbs consisting of empty basal lamina sheaths (2–5). Disease severity and the age of onset vary in and between families, but usually the sensorimotor neuropathy can be detected in the first decade of life (4–6). In addition to the peripheral neuropathy, almost all patients affected by CMT4C develop early-onset severe scoliosis.

CMT4C has originally been mapped in two large Algerian families to a 13-cM linkage interval on chromosome 5q23-q33 (7). By homozygosity mapping and allele-sharing analysis, this interval was narrowed down to a 1.7-megabase region, leading to the identification of mutations in a thus far uncharacterized transcript called KIAA1985, or SH3TC2 (4). SH3TC2 encodes a protein of 1,288 aa containing two Src homology 3 (SH3) and 10 tetratricopeptide repeat (TPR) domains sharing no overall significant similarity to any other human protein with known function. The presence of SH3 and TPR domains suggests that SH3TC2 could act as a scaffold protein. SH3TC2 is well conserved among vertebrate species, whereas no nonvertebrate orthologs were identified (4).

The role of SH3TC2 in the peripheral nervous system remains largely unknown. Thus, we decided to analyze its function in vivo by creating a Sh3tc2ΔEx1/ΔEx1 mouse, a model of CMT4C disease. By detailed characterization of Sh3tc2ΔEx1/ΔEx1 peripheral nerve development and function, we found that these mice mimic neuropathy phenotypes present in CMT4C patients. Moreover, the characterization of the Sh3tc2ΔEx1/ΔEx1 mice led to the finding of lengthened nodes of Ranvier in these mice. This nodal phenotype was confirmed in biopsies from CMT4C patients, providing an additional clinical marker that may improve the diagnostic approach to this disease.

Results

Mice with a Disrupted Sh3tc2 Gene Develop a Peripheral Neuropathy.

Based on described mutations in exon 1 of SH3TC2 in neuropathic patients affected by CMT4C leading to a premature stop codon (4), we decided to develop a mouse knockout model of CMT4C by replacing exon 1 of the Sh3tc2 gene with an enhanced GFP (eGFP)-Neo cassette (Fig. 1A). Mice homozygous for the targeted allele (Sh3tc2neoΔEx1/neoΔEx1) were crossed with the deleter strain nestin-Cre to eliminate the neo gene, leading to the generation of a Sh3tc2ΔEx1 allele. The mating of Sh3tc2ΔEx1/+ mice led to a normal Mendelian ratio of the offspring genotypes. The generated Sh3tc2 mutant mice (Sh3tc2ΔEx1/ΔEx1) lacked exon 1 and expressed eGFP instead of Sh3tc2 mRNA (Fig. S1 A–C). We took advantage of the eGFP cassette replacing Sh3tc2 expression in heterozygote and homozygote animals (Sh3tc2ΔEx1/+ and Sh3tc2ΔEx1/ΔEx1) to determine which cell types express Sh3tc2. In sciatic nerve cross-sections of Sh3tc2ΔEx1/+ or Sh3tc2ΔEx1/ΔEx1 mice, we found a typical croissant-shape staining of the Schwann cell cytoplasm (Fig. 1B). In all other tested tissues (liver, brain, optic nerve, spinal cord, and dorsal root ganglia), eGFP was undetectable which, together with the endoneurium-restricted expression of endogenous Sh3tc2 as detected by qPCR (Fig. S1D), revealed the specificity of the Sh3tc2 promoter.

Fig. 1.

Fig. 1.

Open in a new tab

Sh3tc2ΔEx1/ΔEx1 mice develop a peripheral neuropathy. (A) Schematic diagram showing the targeting vector containing a 5′ homology arm up to the ATG start site, the eGFP and LoxP-neo-LoxP cassettes necessary for monitoring of the eGFP expression in vivo and Geneticin (G418) selection in ES cells, respectively, and a shorter 3′ homology arm. After homologous recombination in embryonic stem cells, exon 1 (Ex1) of Sh3tc2 gene was eliminated. The neo cassette was removed by crossing the resulting mice with a nestin-Cre deleter strain. Primers used for the ES cell screening (5F and 5R, and 3F and 3R), the mouse genotyping (F, Rwt, and Rmt), and the validation of the neo gene deletion (nF and nR) are indicated by arrows. (B) Amplification of GFP signal using anti-GFP antibody on sciatic nerve cross-cryosections. Typical croissant-shaped staining assessed the Schwann cell specificity of the Sh3tc2 promoter in the homozygous knockout nerve (Upper). No signal was detected in the Sh3tc2+/+ nerves (Lower). (C and D) Nerve conduction velocity of Sh3tc2ΔEx1/ΔEx1 and control mice (n = 4) during their first year of life. At 4 weeks, motor [MNCV, (C)] and sensory [SNCV, (D)] nerve conduction velocities were already significantly slowed down in Sh3tc2ΔEx1/ΔEx1 mice.

Mutant mice developed normally, were fertile, and lived as long as their wild-type or heterozygous littermates. In contrast to CMT4C-affected patients, who often develop a scoliosis, no skeletal abnormalities were detected by X-ray analysis of mutant mice (Fig. S2). Although appearing normal overall, Sh3tc2ΔEx1/ΔEx1 mice could be distinguished from control mice by an abnormal clenching of toes and clasping of hind limbs upon tail suspension (Fig. S3), suggesting peripheral nervous system (PNS) abnormalities. To address this issue further, we measured both motor nerve conduction velocity (MNCV) and sensory nerve conduction velocity (SNCV) in these animals. At 4 weeks after birth, Sh3tc2 mutant mice already displayed electrophysiological characteristics of neuropathy, as indicated by a significantly lower nerve conduction velocity compared with control littermates (10-m/s decrease in mutant MNCV and 4-m/s decrease in SNCV). The neuropathy slowly progressed, reaching a reduction of 15 m/s for MNCV and 13 m/s for SNCV at 1 year of age in Sh3tc2 mutant compared with control animals (Fig. 1 C and D).

Peripheral Nerves of Sh3tc2ΔEx1/ΔEx1 Mice Are Hypomyelinated.

The reduction of nerve conduction velocity suggested abnormalities in peripheral nerve fiber myelination in Sh3tc2ΔEx1/ΔEx1 mice. Light-microscopic examination of semithin sections of sciatic nerves isolated from animals 10 (P10), 23 (P23), 56 (P56), and 365 (P365) days old revealed obvious hypomyelination at P56 and at P365 (Fig. S4A). Ultrastructural analysis of sciatic nerve by electron microscopy at P56 demonstrated that in particular, bigger axons were substantially hypomyelinated, and this phenotype was even more visible at P365. However, even after 1 year of duration of the disease, the axons themselves appeared to be healthy (Fig. 2A). Rare onion bulbs and elongated, abnormally branched processes of nonmyelinating Schwann cells as described in CMT4C patients could be found in some sections of Sh3tc2ΔEx1/ΔEx1 sciatic nerve at P365. To evaluate the onset of the hypomyelination phenotype more quantitatively, we performed a morphometric analysis at all four developmental stages (P10–P365). This analysis revealed a slightly higher g ratio in Sh3tc2ΔEx1/ΔEx1 sciatic nerves already at P10 and P23 that became more obvious at P56 and P365 (Fig. S4B). Moreover, representation of myelin thickness as a function of axonal diameter during development (P10–P365) showed that the Sh3tc2ΔEx1/ΔEx1 axons were rarely covered by a myelin sheath thicker than 1 μm, whereas we found that the sheath reached almost 2 μm in the largest control nerve fibers at P56 and P365 (Fig. 2B). The observed hypomyelination did not lead to significant axonal loss in mutant animals, as determined by counting the total number of axons per nerve at P56 (Sh3tc2ΔEx1/ΔEx1, 3,772 ± 209, n = 3; controls, 4,089 ± 160, n = 3). However, a slight shift of the axonal size distribution was observed toward the population of small axons (<3 μm) in Sh3tc2ΔEx1/ΔEx1 nerve starting from P23 (Fig. S4B).

Fig. 2.

Fig. 2.

Open in a new tab

Hypomyelination of sciatic nerve of Sh3tc2ΔEx1/ΔEx1 mice. (A) Electron micrographs of sciatic nerves isolated from wild-type and Sh3tc2ΔEx1/ΔEx1 mice at P56 and at 1 year of age (P365). Although reduced myelin sheath thickness is clearly visible at both stages, the axons themselves remain preserved, even in 1-year-old Sh3tc2ΔEx1/ΔEx1 animals. (B) The scatter plots display myelin thicknesses of individual axons as a function of their respective diameters determined at P10 (n = 4), P23 (n = 4), P56 (n = 2), and P365 (n = 3). Each point corresponds to one fiber (gray points, Sh3tc2ΔEx1/ΔEx1; black points, Sh3tc2+/+). Even at 1 year of age, the myelin in Sh3tc2ΔEx1/ΔEx1 animals did not reach the mature thickness.

Optic nerve axons of Sh3tc2ΔEx1/ΔEx1 mice were myelinated normally, suggesting that the observed myelination defects are PNS-specific (Fig. S4C).

Sh3tc2 Localizes to the Plasma Membrane and the Endocytic Recycling Compartment.

To identify its potential function, we have determined the subcellular localization of Sh3tc2 in cultured Schwann cells (MSC80 cell line). After transfection, a C-terminal FLAG-tagged form of Sh3tc2 (Sh3tc2-FLAG) was localized at the plasma membrane in a dotted pattern and in the pericentrosomal region, as assessed by γ-tubulin costaining (Fig. 3A Left). To further characterize the pericentriolar localization of Sh3tc2, we used Rab11 as a marker of the endocytic recycling compartment (8). We found that Rab11-eGFP partially colocalized in the endocytic recycling compartment with the FLAG-tagged Sh3tc2 (Fig. 3A Right). The identity of this cellular compartment was further confirmed by a transferrin uptake assay (Fig. 3B). Furthermore, analysis of the SH3TC2 amino acid sequence revealed a putative myristoylation site at the N terminus (4). We hypothesized that myristoylation could be responsible for the observed plasma membrane localization of Sh3tc2. An in vitro myristoylation assay revealed that SH3TC2 could indeed be myristoylated, whereas replacing the putative myristoylation site (glycine at position 2) with the nonmyristoylable amino acid alanine (G2A) led to a loss of myristoylation capacity (Fig. 3C). In addition, transient transfection of COS-7 cells with the mutant or wild-type SH3TC2 protein showed that the localization at the plasma membrane was lost when SH3TC2 could not be myristoylated, whereas the localization in the endocytic recycling compartment was partially retained (Fig. 3D).

Fig. 3.

Fig. 3.

Open in a new tab

Subcellular localization of Sh3tc2 at the plasma membrane and endocytic recycling compartment. (A) The mouse Schwann cell line (MSC80) cells were transfected with Sh3tc2-FLAG (Left) and with Sh3tc2-FLAG in combination with Rab11-eGFP (Right). Transfected cells were immunostained with anti-FLAG antibody and γ-tubulin and costained with DAPI to visualize the nuclei. Sh3tc2 localized at the membrane surface and in a cytoplasmic compartment near the centrosome (Left, arrow). This compartment is the endocytic recycling compartment, as shown by the colocalization of Sh3tc2-FLAG with Rab11-eGFP (Right). (B) Transferrin (red) was internalized by COS-7 cells that were transfected with Rab11-HA (blue) and SH3TC2-Myc (green). The overlay shows colocalization of the three molecules in the endocytic recycling compartment. (C) Autoradiography after in vitro myristoylation assay in COS-7 cells using [3H]myristic acid (3H-Myr) showed that wild-type SH3TC2 (wt) can be myristoylated at the N terminus, whereas myristoylation is lost if the N-terminal motif is mutated (G2A). (Lower) A Western blot documenting equal expression of mutant and wild-type SH3TC2-HA. (D) When transfected into COS-7 cells, mutation of the SH3TC2 myristoylation site (G2A-HA) led to diffuse cytoplasmic staining and loss of the plasma membrane localization, whereas the localization at the recycling endosome was partially preserved.

Sh3tc2 Deletion Affects the Myelination-Related Gene Expression Program.

To identify genes and pathways involved in the pathological mechanisms leading to CMT4C disease, we performed a microarray analysis of mRNA populations isolated from P28 sciatic nerve endoneuria of three Sh3tc2ΔEx1/ΔEx1 and three control mice. Analysis of signals obtained from Sh3tc2 and eGFP probes confirmed the inactivation of the Sh3tc2 gene in Sh3tc2ΔEx1/ΔEx1 mice (Fig. S5A). Focused analysis of the expression of myelination-related genes revealed slight down-regulation of some of the genes encoding the main structural myelin proteins, including myelin basic protein (Mbp) and myelin protein zero (Mpz), whereas the increased expression of transcription factors Scip/Oct-6 and Krox24 suggested the presence of incompletely differentiated Schwann cells in P28 Sh3tc2ΔEx1/ΔEx1 sciatic nerve (Fig. 4A). These data were confirmed by qPCR (Fig. S5B). Moreover, genes involved in cholesterol biosynthesis, which plays a crucial role in normal myelin membrane synthesis (9), were all down-regulated in absence of Sh3tc2 (Fig. 4B). We also analyzed the global transcriptomic changes in Sh3tc2 mutants. Despite the observed hypomyelination phenotype in Sh3tc2ΔEx1/ΔEx1 mice at P23, as indicated by altered g ratio (Fig. S4B), the transcriptome was not substantially affected by Sh3tc2 depletion at P28, which is a relatively early stage of the disease. The expression of 179 probes differed significantly between Sh3tc2ΔEx1/ΔEx1 and controls (fold change >2; P < 0.05). Interestingly, annotation and classification of the corresponding genes based on biological functions revealed enrichment in genes involved in cell signaling and adhesion processes (Fig. 4C and Table S1).

Fig. 4.

Fig. 4.

Open in a new tab

Transcriptional changes in Sh3tc2ΔEx1/ΔEx1 sciatic nerve endoneurium at P28. (A) Variation in the level of transcription of genes involved in myelination. Microarray data are represented as a fold change (positive values represent genes overexpressed and negative values genes underexpressed in Sh3tc2ΔEx1/ΔEx1 mice). Scip, Krox24, and Krox20 encode transcription factors involved in initiation and progression of myelination. Mobp, Mbp, Mpz, Mag, Prx, Pmp22, and Cx32 genes encode myelination-related proteins. (B) Synchronous down-regulation of expression of transcripts encoding enzymes involved in cholesterol synthesis. (C) Representation of the distribution of different functional categories obtained after annotation of genes varying more than 2-fold (P < 0.05) between Sh3tc2ΔEx1/ΔEx1 and control samples. The number of genes in each category is depicted as a fraction of the total number of regulated genes and is also indicated in each fraction of the pie. Some genes were represented by multiple array probes listed in Table S1. Genes with unknown function and biological categories represented by fewer than three genes are not represented.

Sh3tc2 Is Required for the Integrity of Nodes of Ranvier.

Because adhesion plays a crucial role in myelin organization and in regions of close interactions between the Schwann cell and the underlying axon, we examined these structures more closely. By electron microscopy analysis, we could not detect alterations in either compact myelin or Schmidt–Lanterman incisures (which connect Schwann cells' perinuclear cytoplasm to the adaxonal cytoplasm) in Sh3tc2ΔEx1/ΔEx1 compared with wild-type sciatic nerve (Fig. S6). However, we observed structural and molecular changes at the nodes of Ranvier. Sciatic nerve teased fibers were labeled with Nile red, a fluorescent lipophilic stain that intensively labels lipid-rich structures, such as myelin. Whereas internodal length was not significantly affected (Sh3tc2ΔEx1/ΔEx1, 605 ± 141 μm; controls, 676 ± 160 μm; fiber diameter, 3–6 μm), this analysis revealed substantially wider nodal spaces between two consecutive Schwann cells in mutant fibers (Fig. 5A and Table S2). We therefore extended our investigation by using markers specific for nodal (pan-Nav, phospho-ERM), paranodal (Caspr), and juxtaparanodal (Kv1.2) regions. The Caspr staining of teased fibers confirmed the presence of enlarged nodes (Fig. 5B and Table S2), whereas the Na+ channels remained tightly clustered at the nodes (Nav staining). The staining of Schwann cell microvilli with phospho-ERM showed reduced intensity in mutant nerves. In addition, potassium channel distribution was perturbed at the juxtaparanodes of Sh3tc2ΔEx1/ΔEx1 fibers (Kv1.2 staining). Previous data demonstrated that potassium channels are more concentrated near the paranodes and diffuse slightly away toward the internodes (10). We confirmed this staining in the wild-type situation, but in mutants, the staining appeared less intense and more diffuse than in controls. To determine whether the abnormalities observed by immunofluorescence correspond to alterations in the ultrastructure of the axoglial contact at the nodes of Ranvier, we performed electron microscopy of sciatic nerve longitudinal sections (1-year-old Sh3tc2ΔEx1/ΔEx1 mice, n = 5, and control mice, n = 3). All nodes of Ranvier (n = 50) analyzed in Sh3tc2ΔEx1/ΔEx1 sciatic nerve presented the enlarged phenotype. Both classes I and II of nodes (as previously defined in Phillips et al., ref. 11) were affected in Sh3tc2ΔEx1/ΔEx1 sciatic nerve, leading to a substantially larger nodal gap in mutant animals (2.69 ± 1.70 μm) compared with controls (0.77 ± 0.27 μm; Fig. 5C). Such a nodal phenotype has not been described in CMT4C patients. Thus, we evaluated the structure of the nodes of Ranvier by using longitudinal sections of nerve biopsies from five CMT4C patients with SH3TC2 mutations (Table S3). Electron microscopy revealed changes at the nodes of Ranvier in three of the five CMT4C patients that are remarkably similar to those observed in Sh3tc2ΔEx1/ΔEx1 mice (Fig. 5D).

Fig. 5.

Fig. 5.

Open in a new tab

Nodes of Ranvier are substantially widened in Sh3tc2ΔEx1/ΔEx1 mice and CMT4C patients' sciatic nerves. (A) Nile red (Nr) staining of myelin of single teased fibers revealed larger nodes of Ranvier (arrowheads) between consecutive myelin sheaths in 5-month-old mutant (Sh3tc2ΔEx1/ΔEx1) compared with control (Sh3tc2+/+) nerve fibers. DAPI staining (blue) enables the visualization of Schwann cell nuclei. (B) Immunostaining of the nodes on teased fibers isolated from 1-year-old animals. Caspr staining confirmed widening of the nodes. Na+ channels were still well clustered at the nodes of Sh3tc2ΔEx1/ΔEx1 fibers, as assessed by Nav staining. Phospho-ERM staining was less intense in mutant nodes, and Kv1.2 distribution at the juxtaparanode was substantially more diffuse in Sh3tc2ΔEx1/ΔEx1 fibers. (C) Longitudinal sections of sciatic nerve isolated from 1-year-old mice analyzed by EM showed wider nodes in Sh3tc2ΔEx1/ΔEx1 axons compared with control samples. Inset shows that the attachment of paranodal loops to the axon, including the presence of transverse bands (arrowheads), is preserved in mutant nerve. Basal lamina covering the node was not affected in Sh3tc2ΔEx1/ΔEx1 animals. (Magnification: Inset, 64,000×.) (D) Nodes of Ranvier observed in human sural nerve biopsy specimens from two CMT4C patients show remarkably similar changes compared with the Sh3tc2ΔEx1/ΔEx1 mouse.

Discussion

During the last two decades, genetic studies have continuously documented new CMT4C cases (4, 5, 7, 12–15). Although the recent identification of SH3TC2 as the gene mutated in CMT4C (4) immediately led to substantially improved clinical diagnostic and genetic counseling in the affected families, the functional role of the SH3TC2 protein in peripheral nerves remained unknown. Here, we describe the generation and analysis of a mouse model of CMT4C disease that led to the discovery of a role for Sh3tc2 in myelination and in the structure of the node of Ranvier.

Our mutant mice show that Sh3tc2 is required for normal myelination in mouse peripheral nerve. In the absence of this gene, mice develop a peripheral neuropathy. Nerve conduction velocity measurements, morphometric analyses, and transcriptional analysis indicated the presence of the neuropathy already at 4 weeks of age, suggesting an early onset of the disease in Sh3tc2ΔEx1/ΔEx1 mice. The neuropathy was then slowly progressive; adult mutant mice presented MNCV and SNCV values ≈15 m/s below controls, which is a decrease comparable to the values recorded in CMT4C patients (4). Whereas CMT4C disease is commonly classified as “demyelinating” disease, we did not observe any signs of demyelination (e.g., myelin debris, Schwann cell onion bulbs, or macrophage infiltration) in the early stages of the disease in Sh3tc2ΔEx1/ΔEx1 mice. Also, the DAPI staining of longitudinal sections or cross-sections of the Sh3tc2ΔEx1/ΔEx1 nerve did not show increased numbers of nuclei, and no cell cycle activators (e.g., cyclin-D1) were overexpressed in our microarray data. Moreover, internodal length was not reduced. Our data therefore suggest that conduction defects in Sh3tc2ΔEx1/ΔEx1 mice are primarily a consequence of hypomyelination and/or of disrupted structure of nodes of Ranvier (see discussion below). The demyelinating features, such as onion bulbs present in the patient biopsies (4, 5), which were absent in younger animals, were infrequently observed in Sh3tc2ΔEx1/ΔEx1 mice at 1 year of age. An explanation for this difference might be the progressive aspect of the disease; the demyelination phenotype present in symptomatic patient biopsies could be the consequence of an early hypomyelination. In addition, a substantial heterogeneity in the age of onset and severity of the disease was observed among patients with CMT4C (4, 5, 12–15).

Sh3tc2 is specifically expressed in Schwann cells in the mouse. Our in vitro experiments demonstrated its punctate localization at the plasma membrane and in the endocytic recycling compartment. Myelin biogenesis and maintenance is a complex process involving coordinated exocytosis, endocytosis, and cytoskeletal dynamics. Perturbations in the myelin protein trafficking and/or turnover are associated with major myelin pathologies (16, 17). Current hypotheses suggest that recycling endosomes play a central role in protein sorting and trafficking, both during plasma membrane recycling and as an intermediate step during cargo transport from the trans-Golgi network to the plasma membrane (18). Sh3tc2 could be transported from endosomal storage sites to the plasma membrane when needed for myelin formation, as described recently for proteolipid protein (PLP) in oligodendrocytes (19). Alternatively, Sh3tc2 could be involved in the regulation of cargo transport through the recycling endosome. Such a transport route has been described recently for the myelin protein Mog in oligodendrocytes (20). Interestingly, transcripts encoding Gcc2 and Vps39, proteins that are involved in Rab GTPase-dependent vesicle trafficking (21, 22), were significantly down-regulated in Sh3tc2ΔEx1/ΔEx1 nerve (Table S1).

Analysis of the Sh3tc2ΔEx1/ΔEx1 nerve fibers led to the discovery of an interesting phenotype at the nodes of Ranvier. Using EM and immunohistological localization of markers of different parts of the nodal region, we observed that the nodes were substantially wider, whereas the clustering of sodium channels and the paranodal structures as stained by Caspr and analyzed by EM at the level of transverse bands were preserved in Sh3tc2-deficient mice. In addition, we detected less phospho-ERM in the microvilli and substantially more diffuse Kv1.2 staining at the juxtaparanodes in Sh3tc2ΔEx1/ΔEx1 mice. Such widening of the nodes associated with decreased phospho-ERM has been described in Laminin 2-deficient mice, where it is associated with impaired clustering of the sodium channels (23). Our data do not allow us to distinguish whether Sh3tc2 is directly involved in the formation and/or maintenance of the node of Ranvier or whether the observed nodal widening is a secondary effect of hypomyelination. However, previous data show that disruption of the nodal gene Nfasc in the CNS results in a reduction of myelination, suggesting that the nodal phenotype could be the primary defect also in CMT4C (24). Moreover, our transcriptomic analysis revealed a 4-fold increase of expression of the nodal matrix gene Hapln2 already in 1-month-old Sh3tc2ΔEx1/ΔEx1 mice (Table S1), suggestive of the presence of early modifications in the nodal environment (25). We noticed in some of the 1-year-old Sh3tc2ΔEx1/ΔEx1 animals a sporadic presence of nodes with a particularly wide gap (>4 μm), absence of microvilli, and a disrupted axoglial junction. These later changes suggest that, as described previously in the diphtheria toxin model (26), nodal lengthening may evolve toward segmental demyelination of the mutant nodes. We found similar or even more severe defects at the node of Ranvier in nerve biopsies from CMT4C patients. In these cases, several nodes were widened, and considerable stretches of axonal surface were covered just by Schwann cell basal lamina but not by cytoplasmic processes of Schwann cells, suggesting a more advanced stage of the disease in these patients compared with analyzed Sh3tc2ΔEx1/ΔEx1 animals. To our knowledge, such alteration of the structure of the node of Ranvier in CMT disease patients has so far been documented only in dominant X-linked CMT disease caused by mutations in the gap junction protein-β1, 32 kDa (GJB1), also called connexin 32 (Cx32) (27). Data from different CMT animal models (CMT1A, refs. 28 and 29; CMT4B1, ref. 30; and, recently, CMT4E, ref. 31) show that the eventual defects observed at the nodes vary between different mutants. Systematic analysis of the nodes of Ranvier would therefore be of great value to improve the characterization of different CMT forms that often remain difficult to discriminate.

In summary, our characterization of the Sh3tc2ΔEx1/ΔEx1 mice demonstrated that these animals represent a valid model of CMT4C neuropathy and allowed us to unveil the Sh3tc2 function in the myelination process and in the node of Ranvier integrity. The Sh3tc2ΔEx1/ΔEx1 mice thus represent an in vivo model for further examination of CMT4C disease mechanisms and for evaluation of any potential therapeutic approach to treat this disease. Moreover, this study led to the discovery of a previously undescribed pathological feature of CMT4C disease that may lead to improvement of its diagnosis.

Experimental Procedures

Generation of Sh3tc2-Null Mice.

We screened the 129S6/SvEvTac mouse genomic BAC library (RPCI-22; BACPAC Resources Center, Children's Hospital Oakland Research Institute, Oakland, CA) and isolated clones containing the genomic region around exon 1 of the mouse Sh3tc2 gene. We then used a recombineering-based approach (32) to obtain the targeting vector (Fig. 1) containing a 5′ homology arm up to the ATG start site (4,655 bp) and the eGFP and LoxP-Neo-LoxP cassettes, allowing us to monitor Sh3tc2 expression and Geneticin (G418) selection in ES cells, respectively, as well as a shorter 3′ homology arm (1,714 bp). The linearized targeting construct was electroporated into 129Sv-derived ES cells by using standard procedures. A total of 384 G418-resistant ES clones were screened by PCR using primers 5F and 3R, located outside the 5′ and 3′ homology arms, respectively, and 5R and 3F, located in the eGFP-Neo cassette (Fig. 1A). Positive clones were confirmed by Southern blotting using ES cell genomic DNA digested with NheI and hybridized with 5′ external probe SBP (Fig. 1A and Fig. S1A). Two correctly targeted clones were used for the injection into C57BL/6J blastocysts that led to the subsequent development of mice with the targeted allele (Sh3tc2neoΔEx1). To eliminate influence of the neo cassette on the expression of nearby genes, we crossed the generated mice with the Nes-Cre deleter (33). This led to the generation of Sh3tc2ΔEx1/ΔEx1 mice, which are also referred to as “Sh3tc2 mutants.” Genotyping PCR conditions are described in SI Materials and Methods.

Morphological, Electrophysiological, and Biochemical Analyses.

Morphometric analysis, Nile red staining, electron microscopy, X-ray analysis, nerve conduction velocity measurements, immunohistochemistry, RT-PCR, qPCR, Sh3tc2 subcellular localization, myristoylation assay, and transferrin uptake assay were performed according to standard methods, details of which and of the ensuing quantifications are described in SI Materials and Methods.

Microarray Analysis.

Sciatic nerve endoneurium of three wild-type and three knockout males were dissected at P28 after 24 h of fasting. A total of 300 ng of total endoneurium RNA was used to synthesize cRNA by using the Illumina TotalPrep RNA amplification kit (Ambion). The Mouse WG-6 v1.1 Expression Beadchips (Illumina) were hybridized with biotin-streptavidin Cy3-labeled cRNA and scanned by the Illumina BeadArray Reader. Microarray data analysis is described in detail in SI Materials and Methods.

Supplementary Material

Supporting Information
supp_106_41_17528__index.html (1.1KB, html)

Acknowledgments.

We thank the transgenic animal facility (Faculty of Biology and Medicine and the University Hospitals, Lausanne, Switzerland) for help in generating the Sh3tc2ΔEx1/ΔEx1 mice; Drs. P. Descombes and C. Delucinge Vivier (Genomics Platform, NCCR Frontiers in Genetics, Geneva, Switzerland) for their help with the microarray experiment; Y. Krempp for his help with the development of the g-ratio calculator software; Dr. S. Anrejevic for help with X-ray analysis; Dr. H. Kleine for help with the myristoylation assay; Dr. M. T. Damiani (Universidad Nacional de Cuyo, Mendoza, Argentina) for the Rab11-eGFP construct; Dr. A. Trumpp (Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland) for providing us with Nes-Cre deleter strain; F. Schupfer, D. Marek, and I. Labgaa for technical assistance; Dr. G. Lemke for support in the initial phase of the project; and Dr. J. Beckmann for his advice. This work was supported by Swiss National Science Foundation (SNSF) Grant PP00A-106714 (to R.C.), by grants from the SNSF, the Deutsche Forschungsgemeinschaft (DFG), and the National Center for Excellence in Research–Neural Plasticity and Repair (to U.S.), the DFG (to J.W.) the START program of Aachen University of Technology (to J.S.), and the Interdisciplinary Centre for Clinical Research BIOMAT within the Faculty of Medicine, Aachen University of Technology (to J.W. and J.S.). J.S. is a Heisenberg fellow and C.S. a postdoctoral fellow of the DFG. A.R. received a Ph.D. scholarship from Aachen University of Technology.

Footnotes

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

This article contains supporting information online at www.pnas.org/cgi/content/full/0905523106/DCSupplemental.

References

  • 1.Skre H. Genetic and clinical aspects of Charcot-Marie-Tooth's disease. Clin Genet. 1974;6:98–118. doi: 10.1111/j.1399-0004.1974.tb00638.x. [DOI] [PubMed] [Google Scholar]
  • 2.Gabreels-Festen AA, Gabreels FJ, Jennekens FG, Joosten EM, Janssen-van Kempen TW. Autosomal recessive form of hereditary motor and sensory neuropathy type I. Neurology. 1992;42:1755–1761. doi: 10.1212/wnl.42.9.1755. [DOI] [PubMed] [Google Scholar]
  • 3.Kessali M, et al. A clinical, electrophysiologic, neuropathologic, and genetic study of two large Algerian families with an autosomal recessive demyelinating form of Charcot-Marie-Tooth disease. Neurology. 1997;48:867–873. doi: 10.1212/wnl.48.4.867. [DOI] [PubMed] [Google Scholar]
  • 4.Senderek J, et al. Mutations in a gene encoding a novel SH3/TPR domain protein cause autosomal recessive Charcot-Marie-Tooth type 4C neuropathy. Am J Hum Genet. 2003;73:1106–1119. doi: 10.1086/379525. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Houlden H, et al. The phenotype of Charcot-Marie-Tooth disease type 4C due to SH3TC2 mutations and possible predisposition to an inflammatory neuropathy. Neuromuscul Disord. 2009;19:264–269. doi: 10.1016/j.nmd.2009.01.006. [DOI] [PubMed] [Google Scholar]
  • 6.Guilbot A, et al. The autosomal recessive form of CMT disease linked to 5q31–q33. Ann N Y Acad Sci. 1999;883:453–456. [PubMed] [Google Scholar]
  • 7.LeGuern E, et al. Homozygosity mapping of an autosomal recessive form of demyelinating Charcot-Marie-Tooth disease to chromosome 5q23–q33. Hum Mol Genet. 1996;5:1685–1688. doi: 10.1093/hmg/5.10.1685. [DOI] [PubMed] [Google Scholar]
  • 8.Ullrich O, Reinsch S, Urbe S, Zerial M, Parton RG. Rab11 regulates recycling through the pericentriolar recycling endosome. J Cell Biol. 1996;135:913–924. doi: 10.1083/jcb.135.4.913. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Verheijen MH, Chrast R, Burrola P, Lemke G. Local regulation of fat metabolism in peripheral nerves. Genes Dev. 2003;17:2450–2464. doi: 10.1101/gad.1116203. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Poliak S, Peles E. The local differentiation of myelinated axons at nodes of Ranvier. Nat Rev Neurosci. 2003;4:968–980. doi: 10.1038/nrn1253. [DOI] [PubMed] [Google Scholar]
  • 11.Phillips DD, Hibbs RG, Ellison JP, Shapiro H. An electron microscopic study of central and peripheral nodes of Ranvier. J Anat. 1972;111:229–238. [PMC free article] [PubMed] [Google Scholar]
  • 12.Gabreels-Festen A, et al. Study on the gene and phenotypic characterisation of autosomal recessive demyelinating motor and sensory neuropathy (Charcot-Marie-Tooth disease) with a gene locus on chromosome 5q23–q33. J Neurol Neurosurg Psychiatry. 1999;66:569–574. doi: 10.1136/jnnp.66.5.569. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Azzedine H, et al. Spine deformities in Charcot-Marie-Tooth 4C caused by SH3TC2 gene mutations. Neurology. 2006;67:602–606. doi: 10.1212/01.wnl.0000230225.19797.93. [DOI] [PubMed] [Google Scholar]
  • 14.Colomer J, et al. Clinical spectrum of CMT4C disease in patients homozygous for the p. Arg1109X mutation in SH3TC2. Neuromuscul Disord. 2006;16:449–453. doi: 10.1016/j.nmd.2006.05.005. [DOI] [PubMed] [Google Scholar]
  • 15.Gosselin I, et al. Founder SH3TC2 mutations are responsible for a CMT4C French-Canadians cluster. Neuromuscul Disord. 2008;18:483–492. doi: 10.1016/j.nmd.2008.04.001. [DOI] [PubMed] [Google Scholar]
  • 16.Berger P, Niemann A, Suter U. Schwann cells and the pathogenesis of inherited motor and sensory neuropathies (Charcot-Marie-Tooth disease) Glia. 2006;54:243–257. doi: 10.1002/glia.20386. [DOI] [PubMed] [Google Scholar]
  • 17.Scherer SS, Wrabetz L. Molecular mechanisms of inherited demyelinating neuropathies. Glia. 2008;56:1578–1589. doi: 10.1002/glia.20751. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Ang AL, et al. Recycling endosomes can serve as intermediates during transport from the Golgi to the plasma membrane of MDCK cells. J Cell Biol. 2004;167:531–543. doi: 10.1083/jcb.200408165. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Trajkovic K, et al. Neuron to glia signaling triggers myelin membrane exocytosis from endosomal storage sites. J Cell Biol. 2006;172:937–948. doi: 10.1083/jcb.200509022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Winterstein C, Trotter J, Kramer-Albers EM. Distinct endocytic recycling of myelin proteins promotes oligodendroglial membrane remodeling. J Cell Sci. 2008;121:834–842. doi: 10.1242/jcs.022731. [DOI] [PubMed] [Google Scholar]
  • 21.Burguete AS, Fenn TD, Brunger AT, Pfeffer SR. Rab and Arl GTPase family members cooperate in the localization of the golgin GCC185. Cell. 2008;132:286–298. doi: 10.1016/j.cell.2007.11.048. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Brett CL, et al. Efficient termination of vacuolar Rab GTPase signaling requires coordinated action by a GAP and a protein kinase. J Cell Biol. 2008;182:1141–1151. doi: 10.1083/jcb.200801001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Occhi S, et al. Both laminin and Schwann cell dystroglycan are necessary for proper clustering of sodium channels at nodes of Ranvier. J Neurosci. 2005;25:9418–9427. doi: 10.1523/JNEUROSCI.2068-05.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Zonta B, et al. Glial and neuronal isoforms of neurofascin have distinct roles in the assembly of nodes of Ranvier in the central nervous system. J Cell Biol. 2008;181:1169–1177. doi: 10.1083/jcb.200712154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Oohashi T, et al. Bral1, a brain-specific link protein, colocalizing with the versican V2 isoform at the nodes of Ranvier in developing and adult mouse central nervous systems. Mol Cell Neurosci. 2002;19:43–57. doi: 10.1006/mcne.2001.1061. [DOI] [PubMed] [Google Scholar]
  • 26.Allt G. In: The Peripheral Nerve. Landon DN, editor. London: Chapman and Hall; 1976. pp. 683–695. [Google Scholar]
  • 27.Hahn AF, Ainsworth PJ, Bolton CF, Bilbao JM, Vallat JM. Pathological findings in the x-linked form of Charcot-Marie-Tooth disease: A morphometric and ultrastructural analysis. Acta Neuropathol. 2001;101:129–139. doi: 10.1007/s004010000275. [DOI] [PubMed] [Google Scholar]
  • 28.Neuberg DH, Sancho S, Suter U. Altered molecular architecture of peripheral nerves in mice lacking the peripheral myelin protein 22 or connexin32. J Neurosci Res. 1999;58:612–623. doi: 10.1002/(sici)1097-4547(19991201)58:5<612::aid-jnr2>3.0.co;2-x. [DOI] [PubMed] [Google Scholar]
  • 29.Devaux JJ, Scherer SS. Altered ion channels in an animal model of Charcot-Marie-Tooth disease type IA. J Neurosci. 2005;25:1470–1480. doi: 10.1523/JNEUROSCI.3328-04.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Bolino A, et al. Disruption of Mtmr2 produces CMT4B1-like neuropathy with myelin outfolding and impaired spermatogenesis. J Cell Biol. 2004;167:711–721. doi: 10.1083/jcb.200407010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Baloh RH, et al. Congenital hypomyelinating neuropathy with lethal conduction failure in mice carrying the Egr2 I268N mutation. J Neurosci. 2009;29:2312–2321. doi: 10.1523/JNEUROSCI.2168-08.2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Liu P, Jenkins NA, Copeland NG. A highly efficient recombineering-based method for generating conditional knockout mutations. Genome Res. 2003;13:476–484. doi: 10.1101/gr.749203. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Dubois NC, Hofmann D, Kaloulis K, Bishop JM, Trumpp A. Nestin-Cre transgenic mouse line Nes-Cre1 mediates highly efficient Cre/loxP mediated recombination in the nervous system, kidney, and somite-derived tissues. Genesis. 2006;44:355–360. doi: 10.1002/dvg.20226. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting Information
supp_106_41_17528__index.html (1.1KB, html)
0905523106_ST1_PDF.pdf (72.2KB, pdf)
0905523106_ST2_PDF.pdf (40.5KB, pdf)
0905523106_ST3_PDF.pdf (26.5KB, pdf)
0905523106_ST4_PDF.pdf (10.3KB, pdf)
0905523106_0905523106SI.pdf (3.8MB, pdf)

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences

RESOURCES