|Home | About | Journals | Submit | Contact Us | Français|
NDRG1 is an intracellular protein that is induced under a number of stress and pathological conditions, and it is thought to be associated with cell growth and differentiation. Recently, human NDRG1 was identified as a gene responsible for hereditary motor and sensory neuropathy-Lom (classified as Charcot-Marie-Tooth disease type 4D), which is characterized by early-onset peripheral neuropathy, leading to severe disability in adulthood. In this study, we generated mice lacking Ndrg1 to analyze its function and elucidate the pathogenesis of Charcot-Marie-Tooth disease type 4D. Histological analysis showed that the sciatic nerve of Ndrg1-deficient mice degenerated with demyelination at about 5 weeks of age. However, myelination of Schwann cells in the sciatic nerve was normal for 2 weeks after birth. Ndrg1-deficient mice showed muscle weakness, especially in the hind limbs, but complicated motor skills were retained. In wild-type mice, NDRG1 was abundantly expressed in the cytoplasm of Schwann cells rather than the myelin sheath. These results indicate that NDRG1 deficiency leads to Schwann cell dysfunction, suggesting that NDRG1 is essential for maintenance of the myelin sheaths in peripheral nerves. These mice will be used for future analyses of the mechanisms of myelin maintenance.
NDRG1, an intracellular protein composed of 394 amino acids, is highly conserved among multicellular organisms, and its expression is induced by stress stimuli. Previously, we showed that NDRG1 is induced by homocysteine, 2-mercaptoethanol, and tunicamycin in cultured human endothelial cells (13). This pattern is similar to that seen for molecular chaperones in the endoplasmic reticulum. Subsequently, NDRG1 was found to be upregulated in a human lung cells following treatment with nickel compounds (31). This change in expression reflected an increase in hypoxia-inducible factor 1 caused by hypoxia or the subsequent elevation of intercellular calcium concentrations (23, 24). NDRG1 expression is also induced by p53 expression and DNA damage, and its expression is inhibited under conditions of cell growth (14). These results suggest that NDRG1 is involved in the cellular stress response mechanisms.
Conversely, Ndrg1 was also identified as a downstream target of N-myc (25). In N-myc knockout mouse embryos, NDRG1 expression is upregulated. During the early stages of differentiation of some tissues, it seems that N-myc activity leads to decreased NDRG1 expression as tissue differentiation progresses. Indeed, NDRG1 has been identified as a gene whose expression is downregulated in tumors (14, 27). Furthermore, NDRG1 expression is induced by differentiation stimuli in cancer cells (21, 29). NDRG1 was also reported to be a metastasis suppressor gene (3, 6). In this regard, the effects of NDRG1 are thought to reflect its potential role in cell differentiation.
Recently, a nonsense mutation of human NDRG1 was reported to be causative for hereditary motor and sensory neuropathy-Lom (9), which is a severe peripheral neuropathy identified in the Gypsy community of Lom, a small town in northwest Bulgaria (10, 11). The hereditary motor and sensory neuropathy-Lom is classified as Charcot-Marie-Tooth disease type 4D (4). Patients with this disease exhibit an early-onset peripheral neuropathy that progresses to severe disability in adulthood, characterized by muscle weakness, sensory loss, and neural deafness. These symptoms are caused by demyelination of peripheral nerves. These observations suggest that NDRG1 is necessary for axonal survival.
To clarify the function(s) of NDRG1, we generated Ndrg1-deficient mice by gene targeting. The Ndrg1-deficient mice exhibited progressive demyelination in peripheral nerves. Moreover, we showed that NDRG1 was significantly expressed in the cytoplasm of Schwann cells, suggesting that NDRG1 deficiency is a primary cause of Schwann cell dysfunction.
We isolated genomic clones carrying the mouse Ndrg1 gene and characterized the promoter and the first exon (25). An 8.9-kb EcoRI fragment encompassing the promoter and exon 1 was used to construct the targeting vector. The initiating Met codon for NDRG1 translation exists in exon 2. A loxP-flanked pSTneoB (26) cassette was inserted at the BamHI site 1.2 kb downstream of the transcriptional start site, and an additional loxP plus BamHI site sequence was inserted at an EcoRV site 1.2 kb upstream of the start site. This resulted in three loxP sites in the vector, so that the Cre recombinase should excise the promoter and exon 1 (Fig. (Fig.1A).1A). The sequence was inserted into the diphtheria toxin A fragment cassette vector (30), and the DNA was linearized by SalI digestion for electroporation.
R1 embryonic stem (ES) cells (18) were electroporated with the targeting vector and selected with 250 μg of G418 per ml. G418-resistant ES colonies were selected, and correctly targeted clones were identified by Southern blotting with a Gene Images Random-Prime system (Amersham Biosciences) with 5′-external, 3′-external, and inner probes (Fig. (Fig.1A).1A). These ES cells were injected into blastocysts to obtain mouse chimeras, which were crossed with wild-type C57BL/6 mice (SLC Japan) for germ line transmission of the floxed Ndrg1 allele. These mice were further crossed with EIIa-Cre deleter mice (15) to excise the promoter and exon 1 region of the Ndrg1 gene together with the neomycin resistance cassette (7). Successful excision of the sequences in the offspring was confirmed by PCR analysis of DNA isolated from punctured ear lobes with primers P1 (5′-AGCAGGCTCTTAAAGCGGCTCC-3′), P2 (5′-CCGCCTCTGTCAAATTAGTAGCTG-3′), and P3 (5′-GGGAGAGCTGAAGGCTGTTCTAGG-3′). Those heterozygous mice with the excised Ndrg1 allele were backcrossed with wild-type C57BL/6 mice. The heterozygous offspring lacking the Cre gene were used for this study. The experiments were conducted in accordance with the current guidelines for the care and use of experimental animals of the National Cardiovascular Center in Japan.
Male mice aged 2 months (wild-type, heterozygous, and homozygous mice) were sacrificed, and their kidneys and sciatic nerves were excised. For extraction of total RNA, whole kidneys or sciatic nerves were immediately homogenized in Trizol reagent (Invitrogen). Isolated total RNA was electrophoresed in a 1% agarose gel containing 2% formaldehyde (10 μg/lane) and transferred to a nylon membrane. To make a specific probe, a partial cDNA fragment (904 bp) was amplified by PCR with primers 5′-CTCAGACACCAAACTGCCAAAAC-3′ and 5′-AATGCTACAAACCCAGTCAGCAG-3′, with the full-length Ndrg1 cDNA used as a template. The fragment obtained was labeled with fluorescein-12-dUTP (PerkinElmer Life Sciences), and hybridization and detection procedures were performed as previously described (32).
For extraction of total protein, the excised organs were homogenized in lysis buffer as described before (12). The protein lysates were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (10 to 20% gradient gel) and transferred to a polyvinylidene difluoride membrane (Bio-Rad). After blocking with 3% skim milk in phosphate-buffered saline (PBS) with 0.05% Tween 20, the membrane was incubated with a 1:1,000 dilution of anti-NDRG1 antiserum (2) and then with a 1:1,000 dilution of peroxidase-conjugated goat anti-rabbit immunoglobulin G (Zymed). Detection was performed with the Western Lightning Chemiluminescence Reagent Plus (PerkinElmer Life Sciences) with the LAS-1000plus image analyzer.
Male mice aged 1, 2, and 5 weeks and 3 and 6 months (wild-type, heterozygous, and homozygous mice) were anesthetized with nembutal (Abbott Laboratories) and perfused with ice-cold PBS containing 4% paraformaldehyde, and sciatic nerve fragments were excised. For light and electron microscopy, the fragments were fixed with 2.5% glutaraldehyde for 2 h at 4°C. After being washed with PBS, the samples were cut into small pieces and fixed with 2% OsO4 for 2 h at 4°C. After dehydration in an ascending ethanol series, the samples were embedded in Quetol812 resin. For light microscopy, semithin sections (1-μm thickness) of sciatic nerves were stained with 0.1% toluidine blue. Slides were examined with an Axioplan 2 microscope (Carl Zeiss). For electron microscopy, ultrathin sections (90-nm thickness) on mesh grids were stained with uranyl acetate and lead acetate and examined with an H-300 electron microscope (Hitachi).
For immunofluorescence microscopy, paraformaldehyde-fixed sciatic nerve specimens from mice aged 3 weeks (wild-type and homozygous mice) were washed with PBS at 4°C and embedded in OCT compound (Sakura Finetek) at −80°C. Frozen sections (5-μm thickness) were washed with PBS. After being blocked with 10% normal goat serum for 15 min at room temperature, the sections were incubated with a 1:200 dilution of anti-NDRG1 antiserum and a 1:100 dilution of rat anti-myelin basic protein (MBP) monoclonal antibody (Chemicon) overnight at 4°C and then with a 1:100 dilution of Alexa Fluor 488-conjugated anti-rabbit immunoglobulin G antibody (Molecular Probes) and a 1:100 dilution of Alexa Fluor 546-conjugated anti-rat immunoglobulin G antibody (Molecular Probes) for 1 h at room temperature. Fluorescence was detected with the Axiovert 200 microscope and photographed with the AxioCam (Carl Zeiss).
Semithin sections of sciatic nerves from three homozygous and three wild-type mice aged 3 weeks and 3 months were photographed, and myelinated axons in a fixed area were counted manually. The diameter of the axons and the thickness of nerve fibers (axon plus myelin) were analyzed with ImageGauge software (Fujifilm). To assess the thickness of the myelin sheath, the g ratio (axon diameter/fiber diameter) was calculated (1, 5). Complex figures with folded myelin were excluded. The significance of differences between mean values was determined by the F test.
Total RNA was extracted from the sciatic nerves, kidneys, and brains of homozygous and wild-type mice aged 5 weeks. Reverse transcription-PCR (RT-PCR) was performed with total RNA (50 ng) as the template and a SuperScript One-Step RT-PCR kit with Platinum Taq (Invitrogen). The primer pairs were 5′-ACCCTGAGATGGTAGAGGGTCTC-3′ and 5′-CCAATTTAGAATTGCATTCCACC-3′ for Ndrg1, 5′-ATTCTTGGACATCTTTTCAGCCA-3′ and 5′-TGCAGGAAGTACTTGAAAGCCTC-3′ for Ndrg2, 5′-CATTAACATTGACCCGTGTGCTA-3′ and 5′-TTGTATTTATAGGGTCGAGGCGA-3′ for Ndrg3, 5′-AAGTACGTGATTGGCATTGGAGT-3′ and 5′-CAGGTGCATTATCTCCGACTACC-3′ for Ndrg4, and 5′-GGAGAAACCTGCCAAGTATGATG-3′ and 5′-CTAGGCCCCTCCTGTTATTATGG-3′ for Gapd.
In the wire-hanging test, we assessed the grip strength of the mice as described before (8). Each mouse was placed on wire netting (20 by 30 cm) taped around the edge. The wire netting was shaken three times and turned upside down. The amount of time that each mouse held onto the wire netting was recorded up to a maximum of 300 s. In the rotorod test, we assessed the ability of the mice to maintain balance on a rotating cylinder as described before (17). The accelerating Rota-Rod (model 7650; Ugo Basile) consists of a 3-cm-diameter cylinder with knurls. Each mouse was placed on the cylinder, which turned at a constant rotation (5 rpm) for 1 min for training, and then the rotation speed was increased over a 300-s period from 5 to 30 rpm. The amount of time that the mouse remained on the accelerating cylinder was recorded. Mice that fell in less than 15 s were given a second trial. Mice that did not fall during the 300-s trial period were given a score of 300 s. Two sets of the rotorod test were performed in the same day, and the higher score for each mouse is reported. The body weight of each mouse was also measured. Both motor activity tests were carried out once every 2 weeks from 5 to 19 weeks of age.
We made a targeting vector to eliminate Ndrg1 expression by deletion of the promoter and exon 1 region with Cre-loxP excision (Fig. (Fig.1A).1A). R1 ES cells were electroporated with the targeting vector and selected with G418. One hundred forty-five G418-resistant ES colonies were selected, and correctly targeted clones were identified by Southern blotting with a 5′-external sequence probe (Fig. 1A and B). Five independently targeted clones were isolated, and the genomic organization of their Ndrg1 locus was further confirmed by Southern blotting with 3′-external and internal probes. After Cre-loxP excision, two lines of mice heterozygous for deletion of the Ndrg1 promoter and exon 1 were obtained. Mice with the deleted Ndrg1 allele and without the Cre gene (Ndrg1+/−) were backcrossed with wild-type C57BL/6 mice (Ndrg1+/+). The Ndrg1+/− mice were crossed to generate homozygous Ndrg1-deficient mice (Ndrg1−/−).
Ndrg1−/− mice were born normally with the expected Mendelian frequency. Both male and female Ndrg1−/− mice were fertile. We confirmed the elimination of NDRG1 expression in the kidney, where NDRG1 was abundantly expressed in wild-type mice (Fig. 1C and D) (32). However, in the sciatic nerve, a faint signal of Ndrg1 mRNA was detectable in Ndrg1−/− mice by Northern blot analysis (Fig. (Fig.2A).2A). Leaky transcription of NDRG1 might be possible because the Ndrg1−/− mice still showed normal organization of the Ndrg1 gene downstream of exon 2, containing the initiating Met codon. To confirm this possibility, we performed Western blot analysis on lysates from the sciatic nerves of Ndrg1+/+, Ndrg1+/−, and Ndrg1−/− mice. As shown in Fig. Fig.2B,2B, faint bands corresponding to the normal size of NDRG1 were observed. These data suggested that a small amount of full-length NDRG1 was expressed in the sciatic nerves of Ndrg1−/− mice. In this regard, Ndrg1−/− mice are hypomorphic, at least in the sciatic nerves.
Despite the low-level expression of NDRG1, Ndrg1−/− mice began showing hind limb weakness at 3 months of age. We also detected an abnormal limb clasping phenotype in Ndrg1−/− mice upon tail suspension (Fig. (Fig.1E).1E). These observations suggested neurological abnormalities. These phenotypes were progressive, that is, 1-year-old or older Ndrg1−/− mice exhibited substantial impairment in hind limb function (i.e., dragging of their legs) and leg muscle atrophy. Two independent lines of Ndrg1−/− mice exhibited similar phenotypes. Ndrg1+/− mice were indistinguishable from Ndrg1+/+ mice in both appearance and behavior.
To address possible peripheral nerve dysfunction, we performed histological analyses. Severe degeneration of the sciatic nerves in Ndrg1−/− mice was seen at 3 months of age. We observed a large number of thinly myelinated axons (Fig. 3C and F). The myelinated axons were significantly decreased in density, and the g ratios of the neuronal fibers were significantly increased in Ndrg1−/− mice at 3 months of age (Table (Table1).1). Electron microscopy of the sciatic nerve showed onion bulb pathology with Schwann cell processes, thin myelin sheaths, endoneurial collagenization, and infiltration of macrophages in Ndrg1−/− mice (Fig. (Fig.4C).4C). A similar demyelinating phenotype is seen in human Charcot-Marie-Tooth disease type 4D patients (11).
To investigate the process of demyelination in Ndrg1−/− mice, we assessed the initial formation of the neuronal myelin sheath at a younger age. Histological analysis showed that, at 1 and 2 weeks of age, the sciatic nerves of Ndrg1−/− mice did not differ from the nerves of Ndrg1+/+ mice; both displayed normal growth of Schwann cells and normal formation of the myelin sheath (Fig. 5A to D). At 3 weeks of age, the density of myelinated axons was not affected in Ndrg1−/− mice, and no significant difference in the g ratios was observed (Table (Table1).1). However, at 5 weeks of age, the myelin sheaths of Ndrg1-deficient mice began to degenerate (Fig. (Fig.5F).5F). The observed demyelination was incomplete and sporadic but was especially pronounced in the myelin sheaths of relatively thick axons. These results indicated that Schwann cell proliferation and myelination were normal and that some defect in the maintenance of the myelin sheath occurred in Ndrg1−/− mice. Older Ndrg1−/− mice exhibited more severe disease in their sciatic nerves, demonstrating that this peripheral nerve degeneration is progressive (Fig. 5G to J).
NDRG1 is ubiquitously expressed in various tissues (13, 32). In particular, the kidney is reported to be a site of high NDRG1 expression in mice and humans (16, 32). Although the peripheral nerves of Ndrg1−/− mice demonstrated clear pathology, no apparent morphological abnormality was observed in their kidneys (data not shown).
To examine which cells, neurons or Schwann cells, are responsible for the demyelinating defects, we investigated the expression of NDRG1 in the sciatic nerve by immunohistochemical analysis. At 3 weeks of age, NDRG1 was abundantly expressed in the cytoplasm of Schwann cells rather than in myelin sheaths or axons in Ndrg1+/+ mice (Fig. 6A to C). We confirmed that MBP, a myelin marker protein, was normally expressed in the myelin sheath of the sciatic nerves of both Ndrg1+/+ and Ndrg1−/− mice (Fig. 6B and E). These results suggested that the cytoplasmic expression of NDRG1 in Schwann cells is essential for the maintenance of myelin structure. Thus, defects in Schwann cells caused by NDRG1 deficiency could be a primary cause of the neural degeneration seen in Ndrg1−/− mice.
To compare the expression patterns of all NDRG family members, we performed RT-PCR analysis on RNA samples from the sciatic nerve, brain, and kidney of Ndrg1+/+ and Ndrg1−/− mice (Fig. (Fig.7).7). In Ndrg1+/+ mice, Ndrg1 was expressed in the sciatic nerves as much as in the kidney but relatively less expressed in the brain. In contrast, Ndrg2, Ndrg3, and Ndrg4 were abundantly expressed in the brain of both Ndrg1+/+ and Ndrg1−/− mice but less in the sciatic nerves. Upregulation of Ndrg2, Ndrg3, and Ndrg4 was not observed in Ndrg1−/− mice.
To examine the muscle strength of the legs, the wire-hanging test was performed. Ndrg1−/− mice were able to hang onto the wire for a significantly shorter period than Ndrg1+/+ at all ages tested. Ndrg1−/− mice demonstrated that their muscle strength was quite decreased. Male mice of both genotypes tended to fall sooner than females due to their heavier body weight (Fig. 8A and C). To measure more complicated motor activities and motor learning in the same mice, a rotorod test was carried out. The scores of older Ndrg1−/− mice in this test were slightly lower than those of the age-matched Ndrg1+/+ controls, though the differences were minor compared to those seen in the wire-hanging test (Fig. (Fig.8B8B).
In this study, we successfully generated Ndrg1−/− mice. The Ndrg1−/− mice exhibited a progressive demyelinating disorder of the peripheral nerves. Histological and quantitative analyses revealed that Schwann cell proliferation and the initial myelination of Ndrg1−/− mice were normal after birth (Fig. (Fig.55 and Table Table1).1). However, sporadic degeneration began by 5 weeks of age (Fig. (Fig.5).5). These results strongly suggest that the ability to form myelin sheaths is retained but some defect in the maintenance of the myelin sheath is present in the Schwann cells of Ndrg1−/− mice. Therefore, NDRG1 is essential for maintenance of the myelin sheath.
It has been reported that NDRG1 expression is induced by differentiation or stress stimuli (21, 27, 29). NDRG1 has also been proposed to shuttle between the cytoplasm and the nucleus in cells (14). Furthermore, phosphorylation of NDRG1 depends on extracellular stimuli (2). These observations imply that NDRG1 may have a role in signal transduction. Recently, it was reported that rat NDRG1 is expressed in astrocytes only in the regions where neurons existed (28). This observation suggests that NDRG1 may also play a similar role in neuronal survival in the brain. We demonstrated that NDRG1 was abundantly expressed in the Schwann cell cytoplasm rather than in myelin sheaths (Fig. (Fig.6).6). This expression pattern is unique compared to that of other Charcot-Marie-Tooth disease-responsible proteins, such as peripheral myelin protein 22, myelin protein zero, connexin 32, and L-periaxin (4). These proteins are localized to the plasma membrane of Schwann cells and are thought to have a role in the formation and/or stabilization of the myelin sheaths. Cytoplasmic expression and phosphorylation of NDRG1 implies its association with intracellular signal transduction in Schwann cells. The NDRG1-mediated signals in Schwann cells related to axonal cross talk could be important for the maintenance of myelin sheaths and axonal survival.
Ndrg1−/− mice exhibited muscle weakness, whereas the complicated motor abilities were relatively retained (Fig. (Fig.8).8). These results indicate that NDRG1 deficiency causes peripheral nerve degeneration leading to muscle weakness. This suggests that peripheral nerves may be quite vulnerable to NDRG1 deficiency but that some degree of functional redundancy for NDRG1 may exist within the central nervous system. NDRG1 is one of four NDRG family members exhibiting different expression patterns (20, 22, 32). We previously demonstrated that NDRG4 is abundantly expressed in neurons in the brain but not in the peripheral nerves (32). NDRG4 expression is induced by homocysteine and reduced both the proliferation and migration rates of cultured cells (19), suggesting that NDRG4 could play a role similar to that of NDRG1 in the brain. NDRG2 and NDRG3 were expressed less in the sciatic nerve than in the brain (Fig. (Fig.7).7). Indeed, no apparent morphological abnormality of the brain was detected in Ndrg1−/− mice (data not shown). NDRG1 deficiency may be compensated for by other NDRG members in the brain.
Although the Ndrg1−/− mice exhibited reductive depletion of NDRG1, a nonsense mutation of human NDRG1 (R148X) is responsible for Charcot-Marie-Tooth disease type 4D (9). The phenotypes of patients with this disease (10, 11) and of Ndrg1−/− mice in peripheral nerves were similar. This suggests that the C-terminal region of NDRG1 may be essential for NDRG1 function.
In conclusion, we found that NDRG1 deficiency leads to a peripheral neuropathy characterized by demyelination, though the initial formation of the myelin sheaths was normal. NDRG1 is abundantly expressed in the cytoplasm of Schwann cells and plays an essential role in maintenance of myelin sheaths. Although the exact molecular functions of NDRG1 are still under investigation, the Ndrg1−/− mouse will be a good model for Charcot-Marie-Tooth disease type 4D and may be used for future analysis of human peripheral nerve neuropathy as well as provide insight into potential therapies.
We thank A. Nagy (Mt. Sinai Hospital) for providing R1 ES cells and loxP and Cre plasmids. We are grateful to H. Westphal (NIH) for the EIIa-Cre mice. We also thank Y. Imai (Riken BSI) for critical discussion.
This work was supported in part by grants-in-aid from the Ministry of Health, Labor, and Welfare of Japan and the Ministry of Education, Culture, Sports, Science, and Technology of Japan and by the Program for Promotion of Fundamental Studies in Health Science of the Organization for Pharmaceutical Safety and Research of Japan.